Professional Documents
Culture Documents
Scott's Standard Methods of Chemical Analysis, Fifth Edition.
Scott's Standard Methods of Chemical Analysis, Fifth Edition.
OF
CHEMICAL ANALYSIS
A Manual of Analytical Methods and General Reference for the Analytical
Chemist and for the Advanced Student
BY
FIFTH EDITION
EDITED BY
LONDON
1'HE TECHNICAL PRESS LTD.
LATE OF AVE MARIA LANE, LUDGATE HILL
GI.OUCESTER ROAD, KINGSTON HILL, SURREY
, fj9':z,h
CQPYRIGHT, 1917, 1922, 1925,
By D. VAN NOSTRAND COMP-'\NY
•
N. Howell Furman, Ph.D. Professor' of Chemistry, Princeton University,
Princeton, N. J. Editor in Chief. Contribution-Chapter on Potentio-
metric Methods; revision of material contributed by the late Professor
W. W. Scott.
SPECIAL SUBJECTS
Decomposition of the material, 1334. Reagents used-I, Water; II, Acids; III,
Fusions, 1335-1337. Properties and special uses of reagents, 1337-1338.
DETERMINATION OF MOISTURE, WATER OF CRYSTALLIZATION
. AND WATER OF CONSTITUTION
Method of loss in weight, 1338"':1339. Adsorption method, 1339-1340. Distillation
method: water in petroleum, 1340-1341; Brown-Duvel method for grains, 1841-1342;
toluene-moisture method, 1342-1343; xylene method, 1343; Kingman mllthod, 1343-
1346. Determination of moisture in gases, 1347.
ALLOYS-I. ANALYSIS OF NON-FERROUS ALLOYS
Classification of alloys, 1348. Preliminary examination; qualitative scheme, 1348-
1349. Methods of procedure, 1349-1350. Difficulty of complete separation of elements,
vii
viii CONTENTS
1350. Limit of accuracy, 1350. Sampling, 13.50-1351. Analysis of brasses and
bronzes-analysis of manganese bronze, 135l. Determination of copper, 1352; l~ad, 1354;
tin, 1356; iron, 1357; manganese, 1358; aluminum, 1359. Analysis of bronze bearing
metals, 1362. Determination of copper, 1362; tin, 1362; lead, 1364-1365; iron, 1365;
nickel, 1365-1366; zinc, 1366; phosphorus, 1366-1369; antimony, 1369-1370; sulfur,
1370-1371; arsenic, 1371; cadmium, 1371-1372. Analysis of copper, nickel, zinc alloys,
1373. Determination of tin, 1373; copper, 1373; lead, ~373; iron, 1373-1374; mangaIlese,
1374; nickel, 1374-1375; zinc, 1375. Analysis of bismuth bronze, alloy of copper, tin, bis-
muth, nickel and zinc, i375. Determination of tin, 1375; lead, 1375; copper, 1376; bis-
muth, 1376; nickel, 1377; zinc, 1377; iron, 1377; aluminum, 1378; manganese, 1378.
Copper-nickel alloys, 1378. Monel metal-determination of copper, 1378; nickel, 1378;
iron, 1379; sulfur, 1379; silicon, 1379; manganese, 1380. Tin and lead base alloys,
1381. Determination of lead, 1381; copper, 1382; antimony, 1382; tin,13S3; arsenic, 13S4:
Rapid methods for control work, 1384; arsenic, antimony, tin, 1385. The determina-
tion of lead, bismuth, iron, aluminum and nickel in tin base bearing metal, i387-1390.
Determination of antimony, zinc, copper in solder, 1390-1391. Rapid method for
control work, 1391-1393. Determination of tin in battery plate met~l, 1393. Zinc
base die casting alloys--determination of copper and aluminum, 1394; magnesium, 1395;
cadmium and iron, 1396; lead and tin, 1396. Fusible metals, 1399. AnalysiS of
'Vood's alloy--determination of lead, 1399; tin, 1399; bismuth, 1;~99; cadmium and
copper, 1400; antimony, 1401; iron, 1401; zinc, 1401; arsenic, 1402. Silver alloys--
analysis of silver solders-dctel'ruination of silver, 1402; copper, 1402; cadmium, 1403;
zinc, 1403; tin, 1404; lead, 1404; iron, 1404; nickel, 1405. Heat resisting alloys, 1405.
Analysis of nichrome. Determination of nickel, 1405; chromium, 1406; iron, 1407;
manganese, 1407; carbon, 1408; silicon, 1408; sulfur, 1409. Corrosion resisting alloys,
1409. Analysis of stellite. Determination of cobalt, 1409; iron, 1411; chromium, 1411;
molybdenum, 1411; 'manganese, 1412; silicon; 1412.
A. S. T. M. method for the ' chemical analysis of pig lead, 1413; determinatioll of
silver; bismuth; arsenic; remaining metals; determination of antimony and tin; de-
termination of copper and cadmium; iron; zinc; nickel and cobalt, 1413-1420.
CEMENT
Specifications, 1597-1599. A. S. T. M. methods for chemical analysis-Sa.mpling,
1599; determination of loss on ignition, 1600; insoluble residue, 1601; sulfuric anhydride,
1601; silica, 1601; ferric oxide and alumina, 1602; calcium oxid~, 1603; magnesia, 1604;
blank determination, 1604. Tentative methods: rapid determination of magnesi~
acid-alkali method, 1605-1607; titration of magnesium o:xyquinolate, 1607-1609.
Phosphorus, 1609; manganese, 1611; J. Lawrence Smith's Ipethod for alkalies, 1612~1614.
Rapid methods for chemical analysis of cement, 1614; silica, 1614; ferric oxide and
alUmina, 1614; calcium oxide, 1614; magnesia., 1615; other components, 1615. Rapid
method for the determination of carbonates in raw materials, 1616. Analysis of lime-
stone, lime, raw mix, etc., 1617. Free lime in ccme.!_lt!_ 1618. A table of analyses, 1619.
tung oil (Chinese wood oil)-gelation tests; Browne heat test; Worstall heat test;
analysis of various oils, 1847-1850. Examination of turpentine-A. S. T. M. methods-
color; specific gravity; refractive index; distillation; polymerization, 1850-1851. Analysis
of varnish-flash point; acid number; ash; solvent; fixed oils and resins; separati<1n of
polymerized oils and resins, 1851-1853. Other materials, 1854. The analysis of paint
pigments-classification of pigments. Analysis of white pigments-basic sulfate of
lead-volumetric determination of lead; volumetric determination of zinc; total lead
sulfate; super sublimed white lead, 1854-1858. Basic carbonate of lead-total lead
(gravimetric); total lead (volumetric); carbon dioxide; acetic acid; metallic lead. Zinc
lead and leaded zinc-moisture; lead and zinc; total soluble sulfates (absence of BaSO.;
pressure of BaS04); soluble zinc sulfate; sulfur dioxide; calculations, 1859-1862. Zinc
oxide-moisture; zinc oxide; total soluble sulfur; sulfur dioxide (Rigg Method); zinc
sulfate, 1862. Lithopone-moisture; barium sulfate; zinc sulfide; zinc oxide, 1862-
1863. Titanium pigments-barium sulfate; calcium sulfate; titanium dioxide, 1863-
1864. Antimony oxide, 1864. Zirconium pigments, 1864-1865. Silica or silex-
China clay-asbestine-moisture 1866. Whiting (Paris white)-moisture; loss on
ignition; calcium; magnesium; carbon dioxide; sulfates, 1866. Barytes and blanc
fixe-moisture; loss on ignition; barium sulfate; soluble sulfates; earbon dioxide, 18f>7.
Analysis of a composite white paint-insoluble residue; total lead; alumina and iron
oxide; zinc; calcium and magnesium; sulfate; sulfide; carbon dioxide; calculations,
1867-1869. Red and brown pigments. Red lead :lnd orange mineral-moisture;
, organic color; total lead and insoluble residue; determination of lead peroxide and
true red lead, 1869-1871. Vermillion, 1871. Iron oxides, 1872. Blue pigments,
1872. Ultramarine blue-moisture; silica; aluminum oxide; sodium oxide; total
sulfur; sulfur as sulfate, 1872-1873. Prussian blue-moisture; nitrogen; iron and
aluminum oxides; slllfuric acid; commercial analysis, 1873. Sublimed blue lead-
total lead; total sulfur; lea~ sulfate; lead sulfite; lead sulfide; lead carbonate; lead
oxide; zinc oxide; carbon and volatile matter, 1874. Yellow and orange pigments,
1874. Chrome yellows-American vermillion-basic lead chromate-moisture; in-'
soluble residue; lead; chromium; zinc, calcium and magnesium; sulfuric acid, 1874-1875.
Green pigments. Chrome green-moisture; insoluble residue; lead; iron, alumina and
chromium; calcium and magnesium; sulfuric acid; nitrogen, 1875-1876. Black pig-
ments-moisture; oil; carbon; ash; analysis of ash, 1876-1877.
SLAG ANALYSIS
Introduction, composition of slags, 2020. Decomposition of the sample and analysis
-general reverberatory slag; chilled blast furnace slag-determination of silica; lime;
iron, 2021-22. Determination of other constituents in slags-alumina; manganese;
zinc; 'magnesia; copper-?olorimetric; electrolytic and iodide methods, 2022-2025.'
WATER ANALYSIS
General considerations, 2046. Sanitary analysis-o~ganic nitrogen; chlorine; oxygen
consumed. Physical tests-turbidity; color; odor-cold, hot; table; taste, 2047-2049.
Chemical tests. Free ammonia; albuminoid ammonia; organic nitrogen; nitrogen as
nitrite; nitrogen as nitrate; oxygen consumed; chlorine as chlorides; total solid residue,
2049-2055, Interpretation of results, 2055-2057: Mineral analysis. Outline of
procedure; silica, iron; colorimetric; aluminum, calcium, magnesium; manganese,
persulfate method; bismuthate method; sulfates-benzidine method; sodium, potassium
and lithium; alkalinity, hydroxyl ion concentration; carbon dioxide-evolution method;
acidity; hydrogen-ion concentration; free carbonic acid; chlorine; nitrates; ammonia arid
its compounds; total mineral residue; hydrogen sulfide; ~il; dissolved oxygen by Winkler's
method, 2057-2080. Methods for the determination of small amounts of lead, zinc,
xvi CONTENTS
copper and tin, 20RG-2081. Hardness-total hardness;, preparation of solutions; opera-
tion; table; magnesium chloride; calcium sulfate; lime and soda value; methods of
reporting and interpretation, 2081-2088. Field assay of w!1rer, 2088. Boiler room con-
trol methods-introduction; solutions; determination of pliosphate; Buromin method;
sulfate; turbidimetric method; volumetric method, 2088-2093.
Rapid, methods for water analysis-special reagents; methods of procedure-
chloride; carbonate, bicarbonate and hydroxyl; sulfate; vlkaline earths; alkali de-
termination, 2093-2097. l
Irrigating waters, 2098.
MICROSCOPICAL EXAMINATION OF WATER ~UPPLY
Importance of the determination; table of odors of organisms, 2099-2101. Sampling
for analysis, 2102. Sedgwick-Rafter method of examination-apparatus required;
procedure; tabular outline identification of forms; li'lt of organisms; i4entification of
forms; control of microscopic organisms, 2102-2112.
Tabl~ of chemicals required for treatment of different genert, 2113. References,
2114-2115.
Bacteriological examination of water-sampling; apparatus needed; adjustment of
reaction; sterilization; nutrient broth; sugar broth; nutrient gelatin; nutrient agar;
eosin methylene blue agar; endo medium; brilliant green, lactose, peptone bile, 211&-2123:
Examination of water for total bacteria; examination forB. coli; subclassification of B.
coli. Coli-aerogenes group-reaction classification; presumptive test; confirmed test;
completed test; selection of coli-aerogenes tests; examination for B. typhi, etc., 2124-2133.
MISCELLANEOUS ANALYSES
Acetone-reagents; sa~ple; procedure; specific gravity; acidity; water; color; residue;
miscibility with water; KMn04 test; aldehydes. Analysis of acetone by Messinger's
method, 2134-2137. Alcohol-detection and estimation-iodofor.m test; Berthelot's
test; ethyl acetate test; Determination-distillation method; evaporation method;
analysis of grain alcohol (ethyl alcohol, spirits of wine, cologne spirits)-per cent ethyl
alcohol; non-volatile; acidity expressed as acetic, acid; aldehydes; qualitative; quantita-
tive-reagents; procedure; furfural; esters expressed, as ethyl acetate; fusel oil; nitrates;
sulfur compounds, 2137-2141. Specific gravity tables, 2142-2146. Methyl alcohol-
detection by method of Riche and Bardy; by oxidation method; in presence of ethyl
alcohol; analysis of mixtures of methyl acetate; methanol and water, 2147-2148. Form-
aldehyde-Leach's test; fIehner's test; morphine sulfate teat; determination, 2149.
Glycerol. Acetin method-reagents; method; calculation. Bichromate process-
reagents; method, 2150-2154. Light oil in coke oven gas-apparatus; procedure;
distillation; calculation; volume of gas required, 2154-2156. Starch and starch products-
sampling; preliminary examination; moisture; ash;,acidity; alkalinity; test for alkaline
starch or alkali; viscosity; converted starches-sampling, moisture ash; total acidity;
viscosity; added materials; sizing strength, 215&-2158. Tannic acid-reagents; total
astringency-Lowenthal-Proctor method, 2158-2159.
Bleaching powder-sampling; available chloride; chlorates; total chlorides; bases
and silica; free lime; quality;. wood and coarse impurities; proximate analysis for sand
and grit, 2159-2161. . Analysis of clay-chemical analysis; alkaJies; titanium; state of
silica. 2161-2164. Analysis ,of glass-density; sample !}nd determination o~ silica; lead
oxide; ferric oxide and alumina; lime; magnesia; alkalies; boric 'oxide; manganese dioxide;
sulfur; barium oxide, separation from lime; factors, 2164-2169. Glass-making ma~ls
CONTENTS xvii
-potash; feldspar, kaolin, China clay; barium carbonl}te; lim'c, 2169-2174. Printing
inks-separation of oil and pigment; analysis of the oil; analysis of the pigmcnt: (a)
black inks; (b) blue inks; (c) red inks; (d) green inks, 2174-2180. Dctermination of
peroxides-hydrogen peroxide-Kingzett's iodide ,method; arsenious acid method of
Jamieson,2180-2181. Sodium sulfide-complete analysis-water insoluble; total iodine
titration; sulfides; thiosulfate and sulfite; alkalinity, polysulfide sulfur and sodium
sulfate; sodium sulfhydrate liquor-calculations; sodiuin chloride; notes, 2181-2187.
CONDUCTOMETRIC ME-TEODS
The principles of conductometric titrations, 2323. Mobility of ions, table, 2324.
CONTENTS :xix
Change in conductance during titration-various cases, 2324-2325. Performance of
the titrations-apparatus, 232q-2328. Application of the method to acid-base titrations
--strong acid with it strong base; intermediate weak acids and bases; very weak acids
and bases, 2328-2331. Mixtures of acids, 2332. Replacement titrations, 2332. Pre-
cipitation and complex-formation, 2333. Applications, 2335.
GAS ANALYSIS
Sampling-tubes; pumps; containers for samples, 233~2339. Measurement of large
gas volumes':_meters; Pitot tube; rotameter; Thomas electric gas meter; orifice meter;
anemometer; flow-meter; Venturi meter, 2339-2342. Measurement of gas in small
quantities-Hempel gas burette; separatory funnel and graduate, 2342-2343. Standards
of measurement; specific gravity, 2343-2344. Absorption apparatus, tubes and pipettes,
2344. Examination of the gases-detection and determination of the various gases;
tables of groups, 2345-2348. Analysis of gases-analysis by Orsat apparatus-manipu-
lation; carbon dioxide, oxygen, carbon monoxide, hydrocarbons; accuracy, 2349-2351.
Eliott apparatus-description; manipulation; carbon dioxide, oxygen, carbon monoxide;
accuracy; time required, 2351:-2353. Hempel's apparatus-description; manipulation;
determination of oxygen in air-by phosphorus; by potassium pyrogallate; by explosion
with ·hydrogen; analysis of manufactured or city gas-carbon dioxide, illuminants,
oxygen, carbon monoxide, methane and hydrogen (a) by Hinman's method, and (b~ by
Hempel's method; calculation (a) Hinman's method; (b) Hempel's method; nitrogen;
acc~acy and time required; not~s, 2353-2359. The properties of the. various gases and
the methods for their determination-group I-ammonia, nitrous anhydride, nitrogen
dioxide, 2360-2361. Group II-chlorine, hydrochloric acid, cyanogen, hydrocyanic
acid, hydrogen sulfide, sulfur dioxide, silicon fluoride, carbon dioxide, 2362-2366.
Group III-phosphine, arsine, stibine, 2366. Group IV--oxygen, ozo~e, 2367-2370.
Group V-carbon monoxide, 2370-2373. Group VI-nitric oxide, 2373. Group VII-
hydrogen, methane, ethylene, acetylene, nitrous oxide, nitrogen, carbon oxysulfide, and
the noble gases, 2373-2380. Applications of gas analysis and interpretation of results-
'chimney and flue gases-carbonic acid indicators; temperature; composition of coal;
tables and chart-combustion efficiency; loss due to carbon monoxide; smoke, 2380-2388.
Producer and fuel gases. Blast furnace gas. Tables; determination of dust, 2388-2389.
City gas. Candle power; calorific power-Sargent calorimeter; Junkers calorimeter;
llIlanipulation; calculations; Thomas recording meter. Calculation of B.T.U. Sulfur;
hydrogen sulfide; ammonia; naphthalene; carbon dioxide; specific gravity; table of
composition of commercial gases, 2389-2399. Sulfuric acid gases-burner gases; sulfur
dioxide; nitrogen oxides; oxygen, 2399. Mine gases-methane; carbon monoxide,
2400-2402. Electrolytic galles, 2402. Acetylene-determination of oxygen, sulfur-
containing gases, phosphine, and acetylene, 2402. Atmospheric air-moisture; carbon
dioxide; carbon monoxide, 2403-2407. Natural gas, 2407, 2409-2410. Determination
of moisture in gases, 2408. Determination of gasoline vapor in gaseous mixtures, 2408.
Calculations,2411-2414. Determination of suspended matter in gas, 2414-2417. De-
termination of nitrogen by the nitrometer, 2418-2419. Compressed gases and their
handling,2420. Reagents,2421-2424. Tables, 2424-2433.
CHEMICAL MICROSCOPY
Advantages of microscopical metho.ds, 2434. Studies of single substances-identifi-
cation of substances; optical properties, 2434-2437. Studies of mixtures of substances--;
xx CONTENTS
various advantages of trw microscopic method, 2438. H{tndling small amounts of
material-technique; filtration; decantation; distillations, 2439. Microscopical qualita-
tive analysis, 2440. Methods of applying reagents, 2441.1 Useful group reagents-
potassium mercuric thiocyanate; cesium chloride; cesium, \lulfate; potassium nitrite;
silver nitrate; barium chloride, 2442-2443. Identity tests, for anions and cations-
cations, 2443-2444. Scheme for qualitative analysis of 'common non-ferrous alloys,
2445. Anions, 2445. Analysis of materials of unknown compositiop'-crystallographic
examination, 2446. Reasons for failure to obtain a positive test, ,2447. Microscopic
quantitative analysis; criteria; preparation methods; analysis by estimation, 2447-2449.
Analysis by counting particles---counting one constituent onlYi counting all constituents,
2449-2451. Analyses based on measurements of area-areal and linear analysis, 2452.
Determination of physical properties, 2453. Measurements; particle-size determina-
tions, 2454-2456. Apparatus for microscopical analysis, 2456. Bibliography, 2457-
2459.
QUANTITATIVE MICROCHEMICAL ANALYSIS
Introductory,2460. The microlabcratory, 2461. The microchemical balance, 2461.
Weighing, notes, 2463. Sampling, 2464. Drying and weighing samples-hygroscopic
substances; liquids; gases, 2464-2469. Treatment of rubber tubing; micro Neubauer
crucible, 2469. Filtering tube, 2470. Removal of traces of precipitate. Immersion
filter,2470-2471. Burettes. Standard solutions; titration. Cqlorimetry. Heating of
combustion tubes; electric furnaces. Equipn;.ent,2472-2475. 'Determination of carbon
and hydrogen in organic material-apparatHs; absorption tubes; procedure; notes;
2475-2482. Determination of carbonate carbon-apparatus; procedure; notes, 2482-
2484. Nitrogen by the micro Dumas method-apparatus: carbon dioxide generator;
combustion tube; nitrometer. Procedure; calculation; according to Pregl; according to
Nieder! and Trautz; notes, 2485-2491. Nitrogen by micro Kjehldahl method-appara-
tus; procedure; notes, 2492-2494. Determination of halogens in organic compounds-I,
according to Pregl-apparatus; procedum; notes; volumetric determination of iodine-
reagents; procedure; notes. Volu~etl'ic determination off chlorine and bromine-solu-
tions; procedure; notes; titration using adsorption iridicatdl'S; notes, 2494-2499. Micro
determination of phosphorus in organic substances-;-preparations of solutions; methods
of oxidizing organic material; treatment of solution with molybdate; volumetric method;
procedure wl).en arsenic is present, 2499-2501. Colorimetric determination of phos-.
phorus and arsenic-reage,r;tts; procedure; notes, 2501-2503. ,rhe microanalytical de-
termination of arsenic-oxidation of the substance; gravimetric detsf!.Ilination; volu-
metric determination-procedure; arsenic in cadavers. Gutzeit method, 2503-2507.
The micro determinatio~ of organic sulfur-apparatus; procedure; determination by
fusion; notes; volumetric determination of sulfur-Tel' Meulen hydrogenation method-
solutions; apparatus; procedure; notes, 2507-2510. Microelectroanalysis-cell of Clark
and Hermance; apparatus; procedure; determination of wckel; zinc; tin; traces of metals.
Traces of zinc in aluminum solutions; lead in the. presence of zinc; traces of copper in
nickel soluti?ns. Cadmium; mercury, 2510-2516. The Pregl cell-apparatus; procedure;
determination of copper in alloys;, deteI;mination of mercury; notes; gold; silver, 2516-
2518. Copper in organic or biological material-ashing; treatment of ash. The Biazzo
method for cOPPCl:-standards; notes, 2519. The carE,alnate colorimetric method, 2520.
Gmvimetric determination of nickel and copper-Brimck's method; procedure; copper,
2g:,:,. Vollmetric determination of iron-reagen~; procedure; notes, 2522. Determina-
CONTENTS xxi
tion of acetyl-method of Elek and Harte; method of Robinson and Bailey, 2523-2526.
The determination of methoxy-ethoxy-modified Zeisel procedure-apparatus; reagents;
procedure; calculations; notes, 2527-2530. The determination of molecular weight-for
solids by rise in boiling point method, 2530. Table of constants useful in ebullioscopy,
2532. Method of Rast, 2533. Molecular weights of liquids"':""micrq Victor Meyer
method-apparatus; determination; calculation; notes, 2534-2536. Micro gas analysis-
apparatus; collection and storage of samples; preparation of samples; preparation of
beads; ope~ation of apparatus; specific determinations: oxygen; water vapor; carbon
monoxide; ethylene; acetylene; hydrogen chloride; ammonia; nitrogen and inert gases;
acetaldehyde and similar compounds; the determination of combustible gases; nitrous
oxide; methane; hydrogen-oxygen, 2537-2545. References-applications; special
methods of analysis; physical methods; determination of organic radicals; fats; specific
compounds or elements, 2545-2547. c
METALLOGRAPHY
Introductory, the microscope, pho(ographic materials; preparation of specimens for
microscopical examination, 2548-2552. Etching sample for examination; reagents used
in etching, 2552. Equilibrium diagrams; photomicrograph of pure iron; eutectic
definition, 2553-255G. Copper-zinc alloYf\; brass, 2557. Iron-carbon alloys; charac-
teristics of carbon steel, 2558-2563. Influence of mechanical work; slip bands; Neumann
lines; critical or exaggerated grain growth, 2563-2566. Heat treatment of steel, 2566.
Case hardening, 2567. Influence of some elements on iron and steel, 2568;-2569.
SAMPLING UNIT 2
The sampling unit may be defined as that portion of the material which is
chosen in such a manner that there is a high probability that it will contain
the different sized particles of the material in the proportion in which they
occur in the entire bulk of the material. Providing there are no great un-
controllable irregularities in the mateJOial, the sampling unit, if taken under
carefully prescribed conditions for each different class, should be properly
representative of the material. However in order to obviate the necessity of
specifying in great detail, and to· guard against unforeseen sectional variations
in the material, it is more expedient to select a number of sampling units,
these depending upon .the size of the shipment to be sampled, and combine
these sampling units into one large sample called the" Gross Sample."
The character of the material, size of the various particles, uniformity of
composition and the prominence of anyone or more constituents are factors
governing the determination of the amount to be taken as the sampling unit.
Without considering a mathematical discussion of the modulus of precision
of a condition which would be .impossible of duplication and in view of the
1 Chapter by J. B. Barnitt .
• 2See: Chemical Analysis of Aluminum, Aluminum Research Labs., 1935, pp. 1-3.
Sampling and Evaluating Secondary Metals, T. A. Wright, Am. lnst. Min. & Met.
Engineers, Tech. Pub. No. 81 (1928). Also A. S. T. M. 1935 Proceedings, on Testing
in the Precious Alloy Field.
1301
1302 STANDARD METHODS OF SAMPLING
experimental data compiled on the sampling of coal, lit is safe to adopt, as the'
general sampling unit, an amount of material equal to approximately 500 times
the weight of the largest particle, If 'this is impractical, due to the presence
of very large particles, and the material cannot qe reduced by passing it
through a crusher, it will be necessary to reduce, bY- hand, the large particles"
at the various sampling points 8,0 that the correct proportion of large particles
can be included in the sampling unit. In some cases, however, a different
bulk of sampling unit will be found necessary. '
(a) If one container amounts to less than one sampling unit, take every
10th load and combine these for the gross sample.
• (b) If one container amounts to approximately 1-3 sampling units, take
. every 20th load and combine these for the gross sample.
(c) If one container amounts to more than three sampling units, take every
50th load and combine these for the gross sample. Note: When very large
shipments are being sampled by the above procedures very large gross samples
will result which would be difficult or inconvenient to reduce to the, laboratory
sample. In this case the gross sample may be reduced periodically as follows.
At regular intervals, representing a definite number of tons, the gross sample is
quartered, one quarter reserved and three quarters returned to the main
bulk. The total combined quarters thus reserved are mixed by shoveling
into a cone-shaped pile and then further reduced by the Standard Procedure
to the laboratory sample.
(d) When the material is composed of particles one inch or less in size,
take a specified increment from every nth container.
(e) When the material is being dumped into a crusher from which it flows
or is flowing, from a drying kiln or other process kiln, the material may be
sampled by taking specified increments from the outflowing stream at stated
regular intervals.·
( iii i ! )
'\! : : : i?
Templet
SAMPLE PUNCH
,
PUNCI-I METI-IOD
(" . , )
mr~
(Top View
Ib
I
J.{'Dia.
1
FIG. 151. Apparatus for Sampling Solids.
Seet. 1. Thief for sampling solids >£" diameter or smaller. OpeningH to corre-
spond. Both tubes hollow and to be made from is" sheet brass (hard). Inside tube
to fit snugly into outside tube. Point to be steel or iron.
Sect. la. Saw Method of Sampling Pigs. The bars are sampled in sets of five,
according to Template as shown. Saw is sharpened on Emery Wheel to size and shape
here shown. I
Sect. lb. Punch Method of Sampling Pigs. 6 The' bars selected as samples are
placed in a line, with every other bar, bottom side up. The sampling is done according
to template in sets of five bars each, as indicated above. The punch must be driven
almost through the bar. If a large sample is desired, the bars are turned over and
sampled on the other diagonal. I
Sect. Ie. Sample Punch.
6 Lear! id best sampled when molten (" gum-drop" method) rather than in' pig form.
STANDARD METHODS OF SAMPLING 1305
(a) In the Form of Pigs, Billets, Ingots, Sheets, Slabs, etc.-One pig, etc.,
.should be taken to represent each ton of metal in the lot and each of these
should be sampled by one of the following methods after thoroughly cleaning
the surface to be sampled. • ,/ .
1. By sawing completely through the specimen as illustrated by Sect. la,
Fig. 151. The sawdust from all the specimens' is then thoroughly mixed and
quartered down on a clean surface and the required amount of sample drawn.
2. By punching or drilling completely through if the equipment permits,
or halfway through from two opposite sides as illustrated by Sect. le, Ib,
Fig. 151; In this case the holes shall be spaced along a diagonal line from one
corner of the specimen to the other. Sampling in this wanner may be so
arranged that one or more holes are made in each! of several specimens or a
group in such positions that they represent consec6.tive positions on the diag-
onal of a single specimen. (See illustration in Sect. Ie.) These punchings or
drillings are carefully melted in a clean graphite crucible, and either granulated
by carefully pouring into distilled water and thoroughly drying, or by casting
into thin slabs which may be sawed completely through in several places and
the sawdust treated as described above. Drillings may'be- chipped and mixed
if convenient. .
(b) In the Form of Sheets.-In sampling brass discs, ana sheets, when
they can be crated so that the edges are flush, recourse can be made to a port-
able milling machine operated by a 7i H.P. 'electric motor run transversely
across the edges, so adjusted that 'only a very thin milling is taken from each
disc or sheet. These are coned and quartered for the sample. By this method
the serviceability of the disc or sheet is not destroyed, and the completeness
7Readily oxidizable material must be stored in similar fashion (iron pyrites, etc.).
8See E. Keller, Bur. Mines Bulletin, 122 (1916), 105 pp. For steels, ores, etc., see
Methods of the Chemists of the U. S. Steel Corporation for Sampling and Analysis
of Iron Ores, etc. Pub!. by Carnegie Steel Co., ·PittSburgh, Pa. See A. S. T; M.,
E 30-36, pp. 2-6, Methods of Chemical Analysis of Metals (1936).
STANDARD METHODS OF SAMPLING 1307
of the sampling depends only on the accuracy of adjusting the edges in the
crate.
PROCEDURE FOR REDUCING THE GROSS SAMPLE
By combining the several sampling units, the gross sample is o'btained
which should now be so large that large single chance particles of material
could be entirely foreign matter or entirely pure substance and affect only
slightly if at all the final sampTe. Increasing the size of the gross sample
increases its accuracy, but cost and convenience of collection and reduction
of sample, limit the degree to which this can be carried to advantage. The
reduction of the gross sample to the laboratory sample is an operation which
must be performed with accuracy and precision .. Automatic machinery and
labor and time saving devices may be used for this operation adhering closely
to the following general scheme with the exception that some materials of
uniform comparison need not be finally reduced as small as 6 mm.-4 mesh-
while it may be necessary to finally reduce other material to .14 mm.-IOO
mesh-or finer.
In'sampling large shipments, the gross sample may become excessively
large and unwieldy. This may be obviated by periodically reducing the gross
sample by the Standard Long Pile and Alternate Shovel Procedure, and sub-
sequently compositing these reduced portions for the final gross sample.
f
--- H
-=-
~Ii-
.. -
J
Crushers.-A jaw and toggle crusher is shown in Fig. 153. This is a sat-
isfactory crusher for coarse, medium and fine crushing and permits a con-
siderable range of adaptability to suit varying conditions and requirements.
Simple adjustments allow the capacity to be varied from about 200 pounds
per hour for the production of fine material to 700 pounds per hour for coarser
material whim considering only one of the several sizes of this type of crusher.
The parts are readily accessible for cleaning and repairing. By means 0,£ this
type of crusher it is possible to reduce the material to about 4 mesh.
10 By courtesy of Sturtevant Mill Co.
STANDARD METHODS OF SAMPLING 1311
\
1312 STANDARD METHODS OF SAMPLING
,/
STANDARD METHODS OF SAMPLING 1313
\
1314 STANDARD METHODS OF SAMPLING
MIXING AND DIVIDING SAMPLES
'1
Large samples (500 lbs. or more) are llsually mixed and divided by the
" Long Pile and Alternate Shovel '1 procedure \}r (500--;100 lbs.) by th~ "Cone ;
and Quartering" procedure. , , '
Small samples (less than 100 lbs.) are usually mixed by rolling on a clean
canvas, whereby the particles are given a rolling motion in alt.ernately opposite
directions, the dividing being done by quartering 'or, more accurately and
rapidly, by passing it through some form of a(fi~e. A convenient and service-
able type of riffle is illustrated by Fig. 161. Another and more sirriple type is
shown by Fig. 162. '
"
t§.
SC; - -----------.f5u
iCSlot
g
-- . - ------:t--" a "
RIFFLE SAMPLER
FIG. 161. FIG. 162.
FIG. 161. Riffle Sampler. "a," cover to fit tightly over one of the boxes with
slot in rounded tops-also fitted with two lugs ~hich fit in ends of riffle and help to
rock evenly. , f
FIG. 162. The cut shows the apron of the cylinder of this machine open for re-
ceiving the sample. After the sample is poured in, the apron's position is shifted, a
being moved to a' . The cylinder is then revolved counterclockwise. The closed sides
of the riffles plow through and thoroughly mix the sample, and no sample can be dis-
charged through the riffle while the cylind~r is revolving in this direction if the level of
the sample is below the axis of the cylinder. Aftel'tbe sample has been mixed the
cylinder is rotated one turn clockwise; the sample in'the cylinder is then cut by the
planes of the riffle and half of it is discharged into the receiving tray. The sample
remaining is again mixed by revolving the cylinder counterclockwise. By alternately
changing the direction, the sample is alternately mixed and halved until twice the'size
of sample required remains in the cylinder~ The tray is then emptied o{ the discarded
material, and the sample remaining, mixed and halved, and the sample caught in the
tray is bottled' and becomes the official sampl~, while the sample remaining in the cyl-
inder is bottle~ and held as the reserve sample. The interior of the machine is easy of
access and should be brushed clean after each sampling operation.
/"
Number of Mesh per
Opening Linear Diameter of Wire
Designation of Customary
Sieve, l\1etric Usage
System
Mm. Inch Centimeter Inch Mm. Inch
While the sampling process does. not require extremely accurate sieves
it is advisable to use those which conform approximately to a standard speci-
fication. Consequently whenever reference is made to a sieve it will be under-
stood to comply with the following specifications, which have been adopted
by a conference of representatives of various scientific and technical societies,
government bureaus, and private fir-ms, held at the U. S. Bureau of Standards.
1316 STANDARD METHODS OF SAMPLING
.
This sieve scale is essentially metric. The siE)Ve~having an opening of 1 mm:
is the basic one and the sieves above and belo:w this in the series are related
to it by using in general the square root of
2 (1.4142), 0):' the fourth r()()t of 2 (1.1892),
as the ratio of the width of one opelling to
the next sllla,lle~ opening. The first ratio i~
used for openings between 1 mm. and 128
mm. while the fourth root of 2 is used as the
ratio for opehings below l.mm. to give more
Outlet for sieves in that part of the scale.
Burned
Gases
CONTAINERS FOR SHIPMENT
DRYING OVEN
TO LABORATORY' .
FIG. 163. Dryer for Coarse Sam- Samples in which the moisture content
'pies. The outlet for air at the top is important should always be collected and
may be connected with a 'chimney
or any other device which will fur- . shipped in moisture-tight containers. A
nish a suitable draft. The sample galvanized iron or tin can with a screw tor
is spread on tared pans, weighed, wnich is seared with' a, rubber gasket and
and dried at 100 to 150 C. above adhesive tape is best adapted to this pur-
room temperature, and weighed pose. Glass fruit-jars sealed with rubber
again. The drying should be con-
tinued until the loss in weight is not gaskets may be used, but require very care-
more than 0.1 per cent per hour. 'ful packing to avoid breakage in transit.
J-Ieating units
FIG. 164. FIG. 165.
FIGS. 164 and 165. Drying oven, triple walled, with a layer of air entirely sur·
rounding the inner chamber, the temperature of! this layer being practically the samE
as that of the air in the chamber. This approaches an adiabatic construction whict
has been approved by the Bureau of Standards as the best type of design for calorimeter!
and other constant temperature devices. By this construction together with a heav)
outer insulation, heat exchanges between the inner chamber and the room outside are
practically entirely prevented, and as a result of uniformity of temperature is secured
in the heated space. The drying chamber sh()uld be entirely shielded from direct
,~'1diation from the heating units, and into eVCfOY'JJart of which heat should be carried
by convection current~.
STANDARD METHODS OF SAMPLING 1317
FIG. 167. 15 Vacuum oven. Supplied with a vacuum chamber which may be easily
removed, permitting the oven proper to be used as a regular drying oven when the
vacuum feature is not desired. Moisture test of flour can be· made in five hours at
75 0 C., while in an ordinary oven at 100° C. from ten to 'twelve hours are required.
The same conditions apply for organic tests in general, i.e., for fertilizers where the tests
must be made at the. lowest possible temperature to avoid driving off nitrogen com-
pounds; for powder where there is great danger of igniting the sample; for milk where
there is great danger of charring; foi' operations involving essential oils, etc..
'Vhen tests are to be conducted in an atmosphere of hydrogen, nitrogen, carbon
dioxide or other gas, it is only necessary to pass a current of the desired gas through
the vacuum chamber.
14 By courtesy of Arthur Thomas Co. 15 By courtesy of Central Scientific Co.
1318 STANDARD METHODS O;E SAMPLING
.j'Thick
Wrought Iron
-:r-¢:;~:::=:;::;t==;~=
I
Dlsc .
I
I
---;r
I
I
I I /V
I ----:-4 61ass Tubing
I I
I I
I I
I
~
I
I
,!,,'V~~~=~~_Y.
------5"------
J<'IG. 168. Sampling Liquids Appar4tus.
1320 STANDARD METHODS OF SAMPLING
(b) If 'the character and conditiQn of th-e liquid is such as too permit the
formation of zones of different composition, obtain Isamples by means of a
suitable apparatus as shown by Fig. 168,' whereby PJoper proportions, of the
liquid are taken at every point from the top to the \:lOttom. Comb!ne these
samples for the gross sample and, after thoroughly mixing, 'bottle tl;le required
samples (Fig. 168). By slowly lowering the samplirig bottle into the liquid
with the outlet (A) open, a portion of the liquid from every point passes through
(B) into the bottle, whereas if outlet ,(A) is kept closed untiIthe desired depth
is reached, the entire sample may be taken at any desired point.
Lead, wood, or any other material which is resistant to the liquid may b~
substituted for the iron parts of this device, according to the character of the
liquid being sampled. Under some conditions it will be necessary to substitute
a ground glass stopper for the rubber one. Openings of different sizes can be
attached at (A) to govern the rate at which the sample will flow into the bottIe.
(c) Take a sample from the upper, mid-
dle, and lower levels of t~e liquid by means
"
MGlas of a sampling device similar to that shown
Rod
by Fig. 168, by keeping the outlet tube A
Made of }(Pyrex Tubing closed until the container has been lowered
Length 30" to the desired depth.
48"
8/J."
2. FROM DRUMS, CARBOYS, BARRELS, ETC. "
p
~
of time. This gross sample may be mixed and a portion taken for the final
sample or by connecting two or more of the above devices in series; a small
representative sample may be collected automatically, e.g., if one million gal-
lons of liquid per day are passing through a pipe or conduit line at an irregular
rate, it could be sampled by conq.ucting the stream to a series of four sampling
weirs, each of which separates 5% or 1/20 of the flow. A final six gallon
sample would then represent the entire million gallons. The size of each weir
should be such as to give a suitable overflow and the apron over which the
flow passes for division should be fixed in a horizontal position.
(b) Bl means of the arrangement illustrated by Fig. 170, an approximate
sample ratio can be continuously diverted, from a non-uniform flow. This
device consists ,of a pipe A, inserted into the flowing stream at an angle of
70-80 0 provided with an opening B facing the flow and extending to the top
of pipe P. This opening B is constructed as shown in the cut (Fig. 170), the
width above the center of pipe P being equal to the diameter of pipe A, then
tapering from the center to the bottom of the pipe P. As the flow passes pipe
A, an approximately constant sample ratio is diverted through pipe A and is
s:1mpled by a small pipe C, the opening of which faces the flow at a point
halfway between the center and bottom of pipe A. Pipe C is fitted with a
stopcock whereby the amount of the sample flow through C may be regulated.
Pipe A reenters pipe P at some point which will give sufficient positive flow-
head through A. While this arrangement w~ll not necessarily produce an
. accurate sample ratio, an increased flow through P will increase the flow-head
through A and consequently increase the amount of sample withdrawn at C
and an approximately relative sample ratio will be maintained with one regula-
tion of stopcock C:
(C) With large pipes and large flows, the above arrangement may be modi-
fied as shown by Fig. 172. Three pipes, 1, 2, and 3, bent at 90~ angles with
openings facing the flow at different levels are in~erted in pipe P and connected
1322 STA]'iJ"DARD METHODS OF; SAMPLING
I
FIG, 172, Apparatus for Sampli~Liquids in Motion,
• ...1
STANDARD METHODS OF SAMPLING 1323
(a) When the sediment remains in. reasonably uniform suspension d.uring
the period required Tor taking the sample, the liquid is thoroughly mixed and
the sample oQtained by one of the foregoing procedures.
(b) When the sediment settles rapidly or is impossible of uniform distri-
bution it is necessary to take special precautions to secure the components of
the sample in the proper proportion. This can be accomplished with a fair
degree of accuracy by mixing the liquid thoroughly and then rapidly taking
a complete column of the liquid by means of the special thief or device shown
by Fig. 168 or 169. By repeating this operation a number of times, a fairly'
representative sample will be obtained. Another procedure of possible appli-
cation under certain conditions would be to take several portions at as many
uniformly distributed points from the top to the bottom by means of the thief
(Fig. 168 or 169). These portions are then combined and mixed for the gross
slmwle.
(c) To accurately sample a liquid containing insoluble matter, it is necessary
to filter the entire liquid, weigh the insoluble material and filtrate separately
and then take separate samples of the sediment and liquid, recording the
ratio of sediment to liquid for use in the final evaluation of the original mate-
rial. This, however, is impracticable except in such instances wherein the size
of sample allows of economical handling in the manner described.
The most practical and accurate procedure in this case is to effect a separa-
tion of the layers and combine portions of each layer in the proper ratios. If
this is impossible, the material may be transferred to a container of. uniform
cross section, the depth of each layer measured, and then the proper portion
taken from each layer, for a composite sample, by means of a sampling thief
similar to those shown in Figs. 168 and 169.
8. VOLATILE LIQUTDS
Ca) Volatile liquids are best sampled from a, continuous flow while the
liquid is being discharged or by means of a siphon from containers. If the
liquid is uniform in comppsition or can be made so by agitation in some manner,
a portion may be run to the bottom of the sampling bottle until the bottle
overflows through another tube from the top into another bottle or to the
132.4 STANDARl? METHODS OF $AMPLING
waste pipe, to such an extent as to leave it liquid if the first bottle containing
the maximum amount of the volatile constituent (see Fig. 173). 'Fhe sample
bottle should then be quickly stoppered and sealed for transportation to the
laboratory. I
REFERENCES
"Sampling Creosote Oil." Amer. Soc. /Testing Materials (1918), p. 721.
"Liquid Sampling Specifications." Am. Gas Light J., 103,391.
"Standard Methods for the Sampling and Analysis of Commercial Fats and Oils."
J. Ind. Eng. Chern., 11, 116 (1919).
, -- Portland
Threads iTmmnrrni',-L Cement
10 Attach
loSlack-"""
61a55 Tube
fa Sample
ConTaIner..'
(a) For a." grab," .or instantaneous, sample the ordinary t,vo-bottle aspira-
tor may be used (see Fig. 176) using care not to draw the sample so fast as to
produce a reduced pressure at gage B (Fig. 175).
(a) For continuous sampling the apparatus illustIT'.ted by Pig. 17i will
be found convenient.
This apparatus consists of two large galvanized metal tanks, one in the
form of an aspirating-bottle in which the s§_n'i.ple is taken, and the other,
slightly larger, which acts as a reservoir. A sample of gas taken under con-
st/tnt pr\_'ssure and at a uniform rate over any specified period of time, which
STANDARD METHODS OF SAMPLING 1327
I
may be varied at will, is obtained by filling the tanks with water, oil or other
liquids ac(wrding to the character of the gas, until the level reaches the top
of the aspirating-tank. A syphon acts
Pinch Cock
as the outlet and a uniform rate, if the
pressure of th~ gas in constant, is ob-
tained by attaching it to a wooden
float. The speed of outflow may be
regulated by the length of the syphon
and the opening or size of the outlet-
nozzle.
White rubber
tubill9
""
",.
",.
"
~::;_:t~.::~:
"
FIG. 176. Apparatus for Sampling Gas. FI<}. 177. Apparatus for Sampling Gas.
" As the water-level rises, the reservoir J is stl:1adily lowered, drawing the
gas uniformly into the sampling tube, and the mercury, thus displaced, escapes'
through the overflow-tube I to the bbttle G. By means of an aspirator a
rapid current of gas is drawn along the tube F-which is connected to the flue
or gas-main by mean~ of the sampling tube (Pig. P9), a suitable filt~l b(jing
interposed, if necessary, to retain any solid' particles. The .conne'ction A,
between the sampling
tube and the tube F, p
is preferably made of
capillary tubing 3 to
4 inches long, to pre-
vent the diffusion of u
the gas backwards from
N to F ~ If desired, a
small non-return mer-
cury valve may be in~
serted at this point.
H ,
Q
G
" To start the apparatus, the tube F is connected with the gas supply and
a rapid current of gas is drawn through F by means of the aspirator.
"The rubber tube connecting the jet V 'with the water-su~ply is then
closed by the clip S, and the water is allowed to run into T and to overflow
freely through W the height of which has previously been adjusted to the de-
sired level. Water is now run from the.'tank Q until the top of the overflow-
tube I is at the lev:el of the shoulder of the gas tube N, and J and N are now
filled with mercury by means of a small funnel in f: ~Th~ capillary A is finally
filled with mercury by raising the reservoir J, the lower stopcock of the sam-
pling-tube is closed, and J is t.hen lowered t.o its former position and filled to
the overflow-tube with mercury; the operation of sampling may now be com-
menced by opening the lower stopcock of the gas tube N and starting the flow
of water from the jet. The aspirator and/the water-supply to T remain in
operation during the whole period of the test and the apparatus requires no
further attention.
" The period of sampling may be. varied in a number of different ways:
(1) "Jets may be constructed to deliver the requisite quantity of water in
specified times; these are easily drawn from glass t~bing in a blowpipe-flame,
the final adjustment being effected by al~irng the head of water in T.
(2) "Tanks of various diameters may be employed.
(3) "The height of the sampling tube N may be varied." ,
STANDARD METHODS OF SAMPLING 1329
REFERENC;ES
Sinnatt, F. S., "A Convenient Apparatus for Obtaining an Ayerage Sample of Gas and
for Regulating the Flow of a Gas into an Evacuated VesseL" Analyst, 37, 250.
Gray, Thomas, "An Automatic Gas Sampling Apparatus." J. Soc. Chem. Ind., 32,
1092.
'White, "Gas and Fuel Analysis." International Chemical Series.
With the size of opening known, the methOd of re<:ording results is simpli-
fied. It is recommended that in reporting sieve analyses in graphic form
the sizes of sieve opening he arranged as ordinates with the largest size at the
top and that the percentages be plotted as abscissas .
• 1. Scope.-These specifications cover sieves for use in the usual cases of
precision testing for the fineness of materials (classification of materials accord-
ing to size). The sieves have been designated according to the size of openings
in terms of microns, the corresponding United States Standard Sieve Series
designation in arbitrary numhers being given for the convenience of those
familiar with the latter scheme of designation. ..
PHYSICAL REQUIREMENTS
2. Cloth and Frame.-\Vire cloth for standard sieves shall be \"oven (not
twilled, except that the cloth of 44, 53 snd 62-micron sieves may be tw;]Jed
until further notice) from .brass, bronze, or other suitable wire and mounted
on the frames without distortion. To prevent the material being sieved from
catching in the joint between the cloth and the frame, the joint shall be smoothly
filled with solder, or so made that the material will not catch. The sieve
frames should be circular, about 20 em. (8 in.) in diameter and about 5 em.
(2 in.) or 2.5 em. (1 iII.) between the top of the frame and the cloth.
X"OTE.-The fjjze specified ahove is for the regular standard test~ng sieve, but for
special purprn:;e.'> ot.})cr t-;izf'S may be used, ItS for example, the 3-in. sieve which is designed
especially for use on laboratory baln,nccs IlS in t.J~e ~esting of pigments.
WORKlIIA.NSHIP
4. The Sieves sha.n be free from imperfections in the sieve cloth or its
mounting; as for example, punctured, loose or wavy cloth, imperfections in
soldering, etc. ,
MARKlNG
5. The siev,", shall bear a label upon which the manuI8Cturer'. nominal
opening in metric units (preferably in microns) with the equivalent in inches,
together with the size of wire should be given.
NarE.-It h;'re('ommcnded that, until the new mi('run designation shall have been
thoroughly est.abli~bed, in addition the corresponding 'United States Standard Sieve
Scrjc~ number be given,
STANDARD METHODS OF SAMPLING 1331
TABLE I
REQUJREMENTR FOR SIEVE OPENINGS AND 'VIRE DIAMETERS WITH PERMIsSIBLE
VARIATIONS
- -
Tolerance on
Correl'lJond - Sieve O~ning Wire D~meter Tol~ance Wire Diam- Tolerance
ing . S. m eter, per cent in
Designation Standard Average Maximum
Sieve Series Opening, Openin~.
Number per cent per cent
mm. in. mm. in. Under Over
1
4760 micron
•
4 4.76
• •
0.187 1.27
5 6
0.050
7
• ±3
8
15
9
30
10
10
4000 " [) 4.00 0.157 1.12 0.044 ±3 15 30 10
3360 " 6 ::L36 0.132 1.02 0.040 ±3 15 30 10
2830 " 7 2.83 0.111 0.92 0.036 ±3 IS 30 10
,\METHODS OF TESTING
6. (a) To determine .whether the cloth conforms to the requirements of
these specifications, the wire diameter shall be ,measured in each direction by
means of a micrometer-microscope or projection apparatus or other suitable
apparatus of requisite precision, care being taken in all cases to avoid diffrac-
tion effects in measuring to the edge of a wire. Micrometer calipers and steel
scales should not be used for this purpose. The mesh shall be determined by
1332 STANDARD METHODS OF'SAl}1PLING
counting the number of openings for a ·distance of at least 6 in. in both dire~
tions of the sieve, using a scale, either. glass or st~el, and a microscope. The
use of scales with special rulings for obtaining mesh by counting "inter-
ference bands" is recommended in cases where lhrge numbers of any of the
sieves with the smaller openings are to be measure1d. - From the wire diameter
and the mesh the average opening'shall be computed.
NOTE.-No general rule can be given for the number of observations to be taken,
except that they be sufficient to cause no change in the result if say 3 or 4 additional
measurements are made.
(b) Measurements of large openings shall be made by means of the mi-
crometer-microscope or the projection lantern or other suitable apparatus in
the same way as wire diameters are measured.
NOTE.-In order to insure the securing of standard -sieves·it is possible to obtain
from the manufacturer or dealer sieves which have been certified by the National
Bureau of Standards.
APPENDIX
LIST OF POINTS TO BE CONSIDERED IN DRAWING UP SPECIFICATIONS FOR A
SIEV:ING TEST
I. ApPARATUS.
1. Unit of sampling.
2. Method of collecting representative gross sample.
3. Size of gross sample. "
4. Cautions to be taken in handling sarri'ple to prevent contamination
or change.
5. Reduction of gross sample.
III. PREPARATION OF THE SAMPLE FOR TESTING.
1. Drying.
2. Mixing.
3. Special treatment, if any is necessary.
4. Selection of test portion.
5. Size of test portion.
IV. PROCEDURE.
A. Dry Scree,nin(l:
1. Order in which the sieves are to be used.
2. Detailed description of method of shaking.
3. Detailed description of pffiGedure other than method of
shaking. .
4. End point.
STANDARD METHODS OF SAMPLING 1333
/
B. Wet Screening:
, 1. Liquid to be used.
2. Description of any preliminary separation by classification or
other means.
3. Order in which screens are to be used.
4. Detailed description of procedure including amount and
method of adding liquid, method of stirring or 'other
agitation.
5. End point.
1334
DECOMPOSITION OF THE SAMPLE 1335
Reagent to Materials
Effect Solution
I. Water will dissolve the alkali salts, ammonium, lithium, potassium and
sodium acetates, bromides, chlorates, chlorides, iodides, nitrates,
nitrites, sulfates. Also arsenates, arsenites, borates, carbonates,
oxalates, phosphates, and tartrates. The halides of silver, mer-
cury, and lead are insoluble in water. The arsenates, arsenites,
borates, carbonates, oxalates, phosphates and tartrates of the
heavy metals require acid treatment to effect solution. Sulfates
of Ba, Sr, Ca, Hg", Pb are insoluble in water.
II. Acids
1. Hydrochloric acid decomposes carbonate and oxide ores of B, Ca, Fe,
Mn, Mg, P, Sn (reducing agt.), Ti, U, Zn (SnCh assists Fe sol.).
2. Hydrochloric acid together with or'followed by an oxidizing reagent,
i.e., HNO a, KCIO a, etc. Alloys of AI, Sb, Bi, CUi Cr, Co and
Ni steels, ferro silicon, Ti and V steels, sulfide ores of Cu, Ce,
Pb, Mo, U, Zn. -
3. Nitric acid, dilute or concentrated. Alloys of Bi, Cd, Co, Cu, Pb,
Mn, silico manganese, ferro manganese, P and Ti steels, ores
of Cd, Cu, Mo, Co and Ni.
4. Aqua Regia.-The metals and alloys of Au, Ir, Hg, V, Pt. Ores of
Cd, Hg, Rh, W.
5. Sulfuric acid.-Ores of AI, Be, Mn, Pb, Th, Ti, U.
6. Sulfuric acid with HF, silicates, Cu ores and alloys.
7. Hydrofluoric acid, usually in conjunction with H 2S04, HCI or HNOa
useful for dissolving ores containing silicates where Si02 is not
to be determined. An HF-HNOa (5: 5: 200 H 20) is an excel-
l!Jnt solvent for antimony-tin alloys. .
8. KBr, Br, HNO a, HCl combinations to dissolve pyrites of Pb, Ni, Fe,
As, etc. Reactions 1-8 (excepting 7) may be made in glass;
7 in platinum.
9. Constant-boiling hydriodic acid, sp.gr. 1.70, attacks many of the
slightly soluble substances: HgS with brisk evolution of H 2S;
Sn02 with evolution of SnI4; the hot acid transposes the silver
halides, the majority of the "insoluble" sulfates (Pb; Ca; Sr;
Ba; Cr); dissolves CaF 2 ; CrCl a (anhydrous), etc. 2
10. Perchloric acid is a powerful oxidizing and also a dehydrating agent
when hot and concentrated. The solubility of the majority of
the perchlorates in water makes the acid a very desirable solvent.
The perc hI orates in general are soluble in organic solvents such
as alcohols, ketones, aldehydes, ethers, and esters. Anhydrous
perchloric acid is a dangerous substance because it interacts with
oxidizable matter to cause violent explosions. With proper pre-
cautions, especially the avoidance of evaporation 'of alcoholic
, solutions of perchloric acid, the use of the aqueous solutions is
without hazards. 3
2 Caley, J. Am. Ghem. Soc. 54,3240,4112 (1932); 55, 3947 (1933) Caley and Burford,
Ind. Eng. Chern., Anal. Ed. 8, 63 (1936).
3 Perchloric acid, The G. F. Smith Co. Columbus, O.
1336 DECOMPOSITION OF THE SAMPLE
III. Fusions j
1. The insoluble residue remaining from an aCid treatment of a substance
may be brought into solution, generally, by fusiotJ. with .an appropriate flux.
The fusions are made in crucibles (20 m!. capaGity) of silica, iron, nickel,
silver, " palau" (gold-palladium 80 : 20) or platinum. Potassium qisulphate
fusions may be made in silica or glass as well as iii platinum; sodium carbonate
fusions may be made 'in iron, nickel, "palau"," and platinum; hydroxide or
peroxide fusions may be made in iron, nickel or silver .crucibles; sodium car-
bonate potassium nitrate fusions may be made in platinum, but hydrochloric
acid should not be used in dissolving the fusion as the chlorine liberated will
attack the platinum.
2. In making fusions the sample is well mixed with six to twenty times its
weight of the fusion mixture and placed over a thin layer of the flux in the
crucible and then covered by a lid of the same material as that of the crucible.
Heat is applied cautiously until the mixture melts, the heat being raised during
the action. When the melt has reached a quiet fusion and the ,resulting
molten mass has become clear and homogeneous, the action is complete'.
(a) A stiff platinum rod is placed in the molten mass and held in place
until this has cooled, sufficiently to solidify around the wire. By reheating
the crucible the mass of fusion may be lifted out, cooled and together with
the crucible placed in the beaker in which the subsequent treatment is made.
(b) The molten mass may be poured into a crucible lid of sufficient capacity •
and when cooled placed in the beaker together with the crucible.
(c) The molten mass is cooled by rotating the crucible or flask in such a
way that it solidifies in a layer on the sides.
3. Water is now added to the material and the mixture heated until the
mass disintegrates. The beaker is covered by a clock glass and the ~dditional
solvent now added. This may be hydrochloric or sulfuric acids, ammonium
carbonate or any other appropriate solvent.
The following fluxes are commonly used- I
L Sodium carbonate.-Decomposes silica~'es of 'AI, Ca, Ce, Cr, etc.; halides of
Ag; sulfates of Ba, Pb.
2. Sodium and potassium carbonates.-Silicates of AI, F, Fe, Ni, Se, Te, etc.
The mixture fuses at a lower temperature than carbonate alone.
3. Sodium carbonate with oxidizing agent, i.e., KQIOa, KNO a, Na 202, ZnO,
MgO.-Ores of Sb, As, Cr, Fe, Mo, V, Zr, sulfides, pyrites, etc.
4. Sodium peroxide.-Ores of Sb, As, Cr, Mo, Ni, V, U, Sn. Certain alloys
and steel for chromium, etc. •
5. Potassium acid sulfate (with H 2S04).-Ores of AI, Sb, Cu, Cr, Co, Fe,
Mn, Ni, Rh, Ta, Ti, W. Steel with Cr, Mo, V and W, Th, phosphates,'
slags, oxides of metals of basic characrer. Silica rem~ins insoluble. In
a bisulfate fusion in a 250-ml. flask, 25-50 m!. of water and 10 m!. Hel
are added and the mixture boiled until the cake dissolves. If any dark
residue remains, tlie solution is decanted off and the residue again
fused with additional flux. As stated in pag}\ 1334, certain refractory
oxides require the addition of a redugng agent, such as carbon from a
filter paper, to assist the acid sulfate flux. It is better to make two
or more fusions on resulting residues remaining from an acid extraction
of the melt, rather than to increase the amount of flux in'the first charge.
DECOMPOSITION OF THE SAMPLE 1337
, 6. NaOH, KOH, basic flux (with or without KF).-Oxidized Sb ore, ores of
Cr, Sn, Zn, Zr. Sulfides and sulfates in sulfur determinations.,
7. Other reagents.-NH4Cl and CaCOa-alkali silicates, KCN--!tin oxide ore,
CaO, NazOz. Organic compounds for halide determination.
Reactions 1, 2, 3 may be carried out in iron, nickel or platinum crucibles,
5 in silica (porcelain) or platinum, 4, 6, and I7 in iron, nickel and silver.
Reagents.4--Ammonium acetate, NH4 CzH a0 2, is a very effective solvent of
lead salts, especially the sulfate. It is prepared by adding strong acetic acid
to strong ammonium hydroxide until the solution is just acid.
Ammonium hydroxide, NH40H, is a powerful solvent of chloride and bro-
mide of silver.
Ammonium sulfide, (NH4 hS, readily dissolves the sulfides of arsenic, anti-
mony, and tin. This reagent'decomposes slowly, losing NHa and forming
,(NH4 hS2. It should be kept in a dark cool place. To prepare the solution
pass a rapid current of pure hydrogen sulfide through a solution of ammonia
in a reagent bottle .
. Hydrochloric acid, HCI, is a powerful solvent and when mixed with HNO~
in the proportion (2-HCI : 1-HNOa) it will dissolve gold and platinum due
to the chlorine which is thus liberated in the nascent state.
Hydrofluoric aciq" HF, is used to decompose silicates, the 'silica being
volatilized.
Nitric acid, HN0 3 , is a powerful solvent and oxidizing agent. The fuming
variety is a more active oxidizer and desulfurizer.
Sulfuric acid, H 2S0 4, is a powerful solvent and is generally used in a dilute
form made by mixing concentrated acid (1.84 sp.gr.) with water in the pro-
portion I-H 2S04: 5H 20. •
Argol, KHC 4H 40 6, potassium acid tartrate commercial grade, is a powerful
reducing agent and also a basic flux.
Borax, cryst., Na 2B 40 7 ·10H 20, is an acid flux sometimes used as a cover
in place of N aCl. To dehydrate, fuse it in an iron or chalk lined clay crucible,
pour the fused mass on a clean surface and pulverize when cold. Melting
point, borax glass, 742 0 C.
Charcoal is a reducing agent and desulfurizer.
Iron, metallic, a basic flux, and desulfurizer. Nails or wire about Ys inch
dia. generally used. Melting point above 1500 0 C.
Lead, metallic, a basic flux. It is also used as a collector of the precious
metals in the assay of gold and silver ores. Granulated lead is used in scorifi-
cation assays and sheet lead in bullion assays. Its silver content should be
determined. Melting point, 327 0 C. ,
Lead flux.-Sixteen parts sodium bicarbonate, 16 parts potassium carbo-
nate, 8 parts flour and 4 parts borax glass. This mixture serves as a flux,
reducing agent, and desulfurizing agent.
Litharge, PbO, is a basic flux and an oxidizing and desulfurizing agent.
Its silver content should be determined. Melting point, 884 0 C.
Silica, Si02, is an acid flux. Powdered lime-glass can be used as a substi-
tute. Melting point, Si0 2, 1775 0 C.
4 W. J. Crook, Lefax, Philadelphia.
1338 DETERMINATION OF WATER
, The boat and combustion tube may be of any convenient material for the
purpose at hand-hard glass, quartz, porcelain or p)atinum.
Penfield Method.-A variant of the evolution and weighing method that
is frequently used in mineral analysis was devised by Penneld. 5 The 'rock or
mineral powder is placed in a bulb of hard glass; the tube connected to ~he
terminal bulb is widened out to form bulbs (one or tWb) at a distance of several
centimeters from the terminal bulb. The open end of the tube is connectEld ..
to a capillary tube by rubber during the heating of the powder. The water
that is driven off is caused to condense by cooling the bulbs; t.tte capillary
prevents loss of water. Finally the bulb with the mineral residue is removed
by sealing off the tube in a flame, and the part containing the water is weighed,
and the water is then removed from it and the weight of the glass is determined.
materials and tars. An electric heater is preferred for glass, although a gas
burner may be used. A ring gas burner is used for the metal still. Dean and
Stark prefer an electric'heater. 6 • 7
The apparatus, made according to A. S. T. M, specifications, may be pur-
chased from chemical apparatus supply houses. The condenser is a glass reflux
type of prescribed length so arranged as to discharge the condensate in a
Dean-Stark trap. The trap is 10 illl. capacity, tapered from the 2. ml. mark
to 0 ml. Above 2 ml. the inside diameter is 18 to 19 mm. The part below
the 2 ml. mark is calibrated in 0.1 ml. divisions.
The solvent used when testing petroleum products or bituminous materials
derived from petroleum shall be gasoline, free from water, and show the follow-
ing distillation range:
5% shall distill at a temperature not below 90° C. nor above 100° C.
90% shall distill below 210° C.
The solvent used when testing bituminous materials,derived from coal tar,
w.ater-gas tar, etc., shall be a coal-tar naphtha or light oil and shall show the
following distilling range:
98% shall distill between 120° C. and 250° C.
Procedure: For a regular Dean and Stark trap the sample size should be
such that not over 10 ml. of water are collected. For samples high in moisture
a longer trap may be secured or the sample size reduced. The volume of the
sample is accurately measured, transferred into the distilling vessel, and mixed
with 100 ml. of solvent by swirling, taking care to avoid loss of material. The
solvent can best be added to the still by rinsing the graduate with several,
portions. The connections between the distilling vessel, condenser and trap
are made with tight-fitting corks. The heat is applied and regUlated so that
the condensed distillate falls from the end of the condenser at the rate of 2
to 5 drops per second. The distillation should be continued at the specified
rate until no water is visible on any part of the apparatus except in the bottom
of the trap. A persistent ring of condensed water in the condenser tube shall
be removed by increasing the distillation rate for a few minutes.
The volume of the water in the trap at room temperature is meaf?ured and
calculated to the percenta'ge of the original volume of the sample.
2. Brown-Duvel Method for Grains.-The apparatus was described in
1907 by Brown and Duvel. 8 The apparatus was patented 9 by them in order
to allow free use of it by the public. The various components of the appa-
ratus, accessories, details of operation and effect of variation in certain details
are covered in a bulletin written by D. A. Coleman and E. G. Boerner. lO Since
this bulletin covers all the specifications in detail for the flasks, thermometers,
and all other parts of the apparatus and accessories as well as details of the
methods for various grains, it should be available before the apparatus is placed
in use.
The apparatus includes a heating chamber, arranged so several samples
may be handled at the same time, and a tank containing cold water through
which condensing tubes pass. The various parts are fastened on a suitable
frame. Gas burners are the usual source of heat.
1342 . DETERMINATION OF WATER
The distilling medium is a good grade of mIneral engine oil, free lfrom water,
0.900 specific gravity, viscosity 342 to 513 [seconds Saybolt at 20° C., and
open cup flash point, 3500 to 400 0 F. The fir~-point should not be over 450 0 F.
Grains are placed in the distilling flask whol~ in a weighed amount specified.
for each grain. With the standard thermometer in a definite position in a
specified quantity of oil, heating is continue4 until the thermometer in 'the
oil reaches the temperature designated for the kind of grain under test. Then
the flame is extinguished. The water is measured when the thermometer has
dropped to 1600 C. in a graduate under the small layer of oil which distills
over with the water.
The endpoint, i.e., flame extinguishing temperature for the g:r.:ain, is stand-
ardized or set against drying in a water-jacketed oven under standard con-
ditions.u
3. Toluene-Moisture Method (Bidwell-Sterling).1. 2__The receiving and
measuring tube used is a modification of the Dean and Stark tube. In order
to increase the accur~cy of the measurement of the water layer, the calibrated
lower end consists of a narrow tubular portion similar to a burette calibrated
in 0.1 ml. divisions. The volumes Illay be read. to 0.01 ml. especially if a
hand-lens is used. Toluene (B.P. 111 0 C.) is used as distilling solvent rather
than xylene (B.P. approximately 1400 C.) since the lower distilling temperature
of the solvent, apparently eliminates decomposition of many of the organic
materials under investigation.
The method was tried by Bidwell and Sterling on a large variety of organic
materials, such as raisins, flours, butter, dried milk, leather,1~ molasses, grass.
The method is satisfactory for moisture in essential oils,13 soaps (especially
liquid and jelly soaps), glycerine,14 soybeans, and other grains, fats, mineral
oils and all sorts of mixtures and emulsions.
The receiving tubes and also complete apparatus including condensers with
standard ground glass connections may be purchased from apparatus supply
houses. A 250-ml. Erlenmeyer or Soxhlet flask is satisfactory for the dis-
. tilling flask although th~ size of the flask ma,y be increased with large Ramnle
sizes. As source of ,heat, electric heaters so 'djusted that the heating element
is enclosed and non-glowing, and Argand gas burners are satisfactory.
The toluene used should have been previously heated under a reflux con-
denser connected to the flask by a Bidwell-Sterling tube until surplus mois'ture
has ceased to distill off. The receiving tube should be carefully calibrated
and before using should be absolutely clean. It isadvis_able to clean the tubes
and condenser with chromic-sulfuric acid cleaning solution followed by water
and alcohol rinses and oven drying.
Procedure.-Weigh sufficient sample to -give 2 to 5 gms. of water into the
distilling flask. Add sufficient toluene to cover the sample completely,' usually
about 75 ml. Connect the receiving tub~ and reflux condensers. Fill the
tube through the top of the condenser. Protect the side arm from the heat
of the burner by a piece of asbestos board vertically sUl?pended.
With water flowing ,through the condenser, heat gently to avoid foaming,
continue boiling slowly at about 2 drops per second until nearly all the water
has distilled over. Then increase the rat(} _to abo.U't 4 drops per second. A
boiling piece is often an advantage. Fo.r' frOme samples Bidwell and Sterling
recommend sufficient dry san.d to cover the bottom of the distilling flask.
DETERMINATION OF WATER 1343
When the water has apparently all been distilled over to the receiving tube,
wash down the condenser by pouring toluene in at the top. . Continue dis-
tillation a short time to determine if further water distills over. If it does,
repeat the ~ashing of the condenser. Any drops of water adhering to the
sides of the condenser tubing usually can be loosened by a copper wire through
the tpp of the condenser.
Allow the receiving tube to come to room temperature or bring to a definite
temperature by immersing in a bath of this temperature. Read the volume
of the water. The volume of water multiplied by the specific gravity at the
temperature at which the volume is read will give the weight of water in gms.
When used on glycerine or a soap high in glycerine, the distillate may be
corrected for its glycerine content by separating the aqueous layer and deter-
mining the glycerine content by the bichromate method. (See Methods for
Analysis of Soap.) Alcohol and acetone cause very unsatisfactory and un-
reliable results. .
In order to prevent soaps from foaming, a carefully measured amount of
concentrated hydrochloric acid in sufficient quantity to break the soap into
fatty acids is very desirable. The water content of the acid must be known
and proper correction made in the amount of 'water recovered. Soaps con-
taining much carbonate should be handled by the xylene-sodium acetate method.
Glass wool floated in the liquid will often prevent foaming without the addi-
tion of acid.
Soybeans and grains are placed in the flask whole. Distillation is continued
until water ceases to come off. The toluene in the distilling flask is allowed to
cool somewhat. The beans or grain are then strained out on a coffee strainer.
The solvent is returned to the distilling flask while the beans are given a rinse
with a small amount of petroleum ether and allowed to drain until practically
free from solvent. The sample is then quickly ground in a coffee mill and
returned to the distilling flask and distillation continued. until water ceases to
distill over.
4. Xylene Method.-Marcusson in 1904 2 used toluene as distilling agent
in a distillation method for lubricating greases. In 1905 1. 1.\ he describes a
similar method for water in oils, fats, and soaps in which xylene was used as
the distillation medium. In a book 18 by Marcusson 0927) on methods for
analyses for fats 'and oils, xylene is used as distillation solvent and the pro-
cedure is essentially that described in the previous references. If soaps are
present so that foaming will occur and carbonates are absent, dry oleic acid
or potassium bisulfate is added.
Rogers 12 and Bidwell and Sterling 1 pointed out that many organic mate-
rials which showed definite evidence of decomposition when xylene was used,
could be safely handled when toluene was used. In the soap methods,19 a
xylene method with anhydrous sodium acetate tel prevent foamingis described.
The general procedure with xylene is identical with the Bidwell-Sterling
procedure using toluene. The approximately 30° C. higher boiling of xylene
is often an advantage on materials that do not decompose with the liberation
of water at the distilling temperature and that hold the water rather tightly.
5. Kingman Method;-The method was described by W. A. Kingman. ls
The method had been used by him for several years. It differs from the other
distillation methods described in this section by the use of a ,non-inflammable
1344 DETERMINATION OF WATER
Flask
80° C., rIsmg gradually to 95-100° C., when practically all the moisture is
driven off, then going up rapidly to about 144-145° C. During the rapid rise
the heating should be slowed down; otherwise a too rapid evolution of vapor
ensues which heats the condenser and causes the water already distilled to boil
away from the receiver tube. Allow the apparatus to cool and then wash
down the condenser and connecting tube into the graduated tube of the appa-
ratus with a little coal-tar benzene. When the moisture has settled suffi-
ciently, read the column of water, making due allowance for meniscus, and
calculate the percentage.
The method is not as satisfactory for soaps as the toluene or xylene methods.
REFERENCES
1. G. L. Bidwell and W. F. Sterling, J. Ind. Eng. Chem., 17, 147 (1925).
1a. J. Marcusson, Mitt. klg. Materialprlifungsamt, 23, 58 (1905).
2. J. Marcusson, Mitt. klg. Materialprlifungsamt, 22,48 (1904); Z. angew. Chem.,
. .18, 754 (19Q5).
3. A. S. T. M. Standards, 1933, Part II, Non-metallic Materials, page 891, A. S. T. M.
D95-30.
4. A. S. T. M. Standards, 1933, Part II, Non-metallic Materials, page 827, A. S .. T. M.
'D128-27.
5. A. S. T. M. Standards, 1933, Part II, Non-metallic Materials, page 814, A. S. T. M.
DI58-28. .
6. E. W. Dean and D. D. Stark, J. Ind. Eng. Chem., 12, 486 (1920).
7. E. W. Dean, J. Ind. Eng. Chern., 10,823 (1918).
8. A Quick Method for the Determination of Moisture in Grain, Bclletin 99, Bureau
of Plant Industry, U. S. :qept. of Agriculture (1907). Reissued as Bureau of
Plant Industry, Circular 72, under title" A Moisture Tester for Grain and Other
Substances and How to Use It."
9. U. S. Patent No. 848616, March 26, 1907.
10. The Brown-Duvel Moisture Tester and How to Operate It, D. A. Coleman and
E. G. Goerner, Bulletin 1375, U. S. Dept. of Agriculture, 1926.
11. Air-oven and Water-oven Methods Specified in the U. S. Standards for Deter-
mining the Moisture Content of Grain, U. S. Dept. of Agriculture, Bureau of
Agricultural Economics, Service and' Regulatory Announcements, No. 147, April,
1935.
12. Rogers, Bureau of Chemistry, Bulletin 137, 172 (1910).
13. R. M. Reed, Oil and Soap, 9, 67, 1932.
14. L. F. Hoyt & P. C. _Clark, Soap, 7, 25, 1931.
15. W. A. Kingman, J. Ind. Eng. Chern., 18,93 (1926); Determination of Water Con-
tent of Liquid Glue (1926).
16. Standard Methods for Sampling and Analysis of Commercial Fats and Oils, J. Ind.
Eng. Chern., 18, 1348 (1926).
17. Standard Methods for Sampling and Analysis of Commercial Fats and Oils, A. C. S.,
and A. O. C. S. Lefax edition.
18. J. Marcusson, Die Untersuchung del' Fette und Ole, Dritte Auflage, Verlag von
Wilhelm Knapp, Halle (Saale) (1927). •
19. See chapter on Methods for Anal:ysis of Soap and Soap Products.
100Xa
.
c = per cent of water in the original sample, 100 -c = b.
.
Then --b- =per cent calculated from wet baSIS to dry.
1348
ANALYSIS OF -'NON-FERROUS ALLOYS 1349
(a) If the metal dissolves, antimony and tin are absent; omit their tests.
(b) If the metal forms a turbid .~olution, antimony and tin may be present. To a
portion, add several volumes of water and boil. (a) If a clear solution is formed treat
the entire solution in the same way, and omit tests for Sn and Sb.
(c) If the residue is insoluble, Sb and Sn are indicated, evaporate the entire solution
to dryness, gently ignite to expel water, add HNO" heat, dilute and filter. If the
filtrate is turbid, repeat the filtration until clear.
Practically all of the following methods given for the analysis of a brass
or a bronze call for a preliminary separation of ltin as meta-stannic acid' by
means of solution and digestion of the alloy in nitric acid. The meta-stannic
acid .so obtained is subject to contamination b~ such alloy constituents as
copper and iron and will quantitatively contain such elements as phosph.6rus,
arsenic and vanadium provided the ratio of tin to these elements in' the allpy
is high enough. In addition to the above drawbacks, tin is not always quan-
titatively precipitated as meta-stannic acid, particularly when the alloy con-
tains more than 0.25% of iron. In ordinary analyses of materjal· containing
but little iron, correction or provision is made for antimony and phosphorus
when they are present, and the contaminants and the slight solubility losses
are ignored. This procedure results in compensating errors in the determina-
tions of copper and tin; for the impurities make up for dissolved tin while the
dissolved tin plates out to a considerable extent with the copper. In umpire
analyses, or with alloys high in iron or antimony, the above weaknesses of the
method must be taken into account and provision must be made for the com-
plete recovery of all tin, as well as the complete recovery of all contaminating
elements carried down with the tin. 4
Solutions Required. Lead Acid.-Mix 300 m!. of H 2S0 4 (sp.gr. 1.84), and
1800·ml. of distilled water. Dissolve 1 g. of lead acetate, c.p.,\in 300 ml. of
distilled water and add this to the hot solution, stirring meanwhile. Let
stand at lea~t 24 ~ours and siphon through a thick asbestos filter.
Dilute Alcohol for Washing.-Mix equal parts of denatured alcohol and
distilled water.
Method.-Proceed exactly as in " Determination of Lead by the Electro-
lytic Method" up to the point where the precipitated H 2SnOa is filtered and
wash~d. .
To the combined filtrate and washings add 120 m!. of lead acid, and evap-
orate until copious fumes of H 2S0 4 are evolved. Cool, add 105 ml. of distilled.
water to dissolve the salts and to make the acid concentration the same as
in lead acid, heat· to boiling, aud allow to cool and settle for 5 hours, or over
night if convenient. Filter on a weighed porcelain Gooch crucible and wash
with lead acid. Wash out the lead acid with dilute alcohol, set the gooch
crucible inside a porcelain crucible, dry and ignite for 5 minutes at the full
heat of a Tirrill burner, cool and weigh as PbS0 4 , which contains 68.33% lead.
Accuracy.-Duplicate determinations should check within 0.01% of lead.
Results obtained by this method are likely to be about 0.01 % too low.
NOTEs.-l. It is impossible toremove the tin quantitatively as H 2SnO, in the case of
alloys containing over 0.25% of iron, but by the preliminary baking and long standing,
most of the H 2SnO, is precipitated.
2. See Note 4 under" Determination of Lead by the Electrolytic Method."
3. For the aetermination of lead as a sulfate in alloys containing over 0.2% see,
"Analysis of Bronze Bearing Metal."
DETERMINATION OF TIN
Solutions Required. Dilute Hydrochloric Acid.-Mix 500 m!. of HCI
Csp.gr. 1.18), and 500 m!. of distilled water. .
Dilute Sulfuric Acid.-Mix 300 ml. of H 2S0 4 Csp.gr. 1.84), and\600 mL of
distilled water.
Dilute Ammonia.-Mix 100 m!. of NH 40H Csp.gr. 0.90), and 900 m!. of
distilled water.
; Ammonium Acetate Solution for Washing.-Dissolve 10 g. of ammonium
acetate in 300 ml. of distilled water. Make slightly acid with acetic .acid and
saturate with H 2S gas.
Method.-In a 150-ml. beaker dissolve 2 g.. of bronze in a mixture of 10 m!.
of HCI Csp.gr. l.18), and 5 ml. of HNO g Csp,gr. L42). Dilute to about 75 m!.
with distilled water. Add NH 40H Csp.gr. ,0.90)/until the basic salts of copper
have dissolved and the solution has becorp.e a deep blue. Boil and allow to
settle, and filter on a closely woven filter paper. Wash with dilute ammonia
and with hot water. Dissolve the precipititte on the filter with hot dilute HC!.
Dilute the solution to about 100 m!. Add NH 4 0H Csp.gr. 0.90) until a per-
manent precipitate ~orms. Heat the solution to boilillg, allow to settle, filter.
and wash as before. Dissolve the precipitate on the filter with boiling hot
dilute H 2S0 4, washing the paper very thoroughly with this a.cid. Add NH 40H
Csp.gr. 0.90) cautiously until the precipitate which forms at first dissolves
rather slowly. Allow the solution to stand for some hours and if any lead
sulfate forms filter it off. Dilute the solution to about 200 m!. and saturate
it with H 2S gas. Filter the precipitated tin sulfide off on double papers and
wash with ammonium acetate solution, retainIng the filtrate and washings for
the iron determination. Dry the precipitate and place it with the papers in a
porcelain crucible. which projects part way through a hole in a piece of asbestos
board. 'Heat slowly until any free sulfur has been driven off, but without
allowing the sulfur to burn: When the !luJfur has been expelled, place the
crucible <)n a triangle and ignite slowly' at first and finally at the full heat of the
ANALYSIS OF NON-FERROUS ALLOYS 1357
burner. If the precipitato weighs more than 20 mg. heat to constant weight,
using a blast lamp Weigh as Sn02, which contains 78.77% of tin.
Accuracy.-Duplicate determinations should check within 0.06% of tin.
NOTEs.-l. If, during ignition, the sulfur is allowed to burn, some tin sulfate may
be formed, causing high results. On the other hand, low results may be caused by
too high heat, which causes Epattering and volatilizes some tin sulfide.
2. Tin and iron are best determined on the same sample, retaining the filtrate from
the tin sulfide precipitate to use for the iron determination.
3. For the determination of tin in alloys containing more tin than is present in this
alloy see" Analysis of Bronze Bearing Metal." .
4. In alloys containing an appreciable amount of antimony and in the presence of a
large excess of tin, the antimony as determined by the method given under "Analysis
of B,-onze Bearing Metal" should be calculated to Sb 20. and a corresponding amount
deducted from the weight of the ignited Sn02 before calculating to tin. Similarly
arsenic found in the alloy should be calculated to As 20 S and a corresponding correction
made.
DETERMINATION OF IRON
'Solutions Required.-See " Determination of Tin," and the following:
Dilute Sulfuric Acid for Reductor.-Mix 50 m!. of H 2S0 4 (sp.gr. 1.84),
and 1000 m!. of distilled water. This solution is used boiling hot.
Potassium Permanganate.-Dissolve 0.2 g. of KMn04 in water, filter
through asbestos, and dilute to 1000 m!. with distilled water. Standardize
against 0.020 g. portions of pure sodium oxalate. Each milliliter is equivalent
to approximately 0.00033 g. of iron. . ,
Method.-Proceed exactly as in the" Determination of Tin" through the
filtration and washing of the precipitated tin sulfide. Combine the filtrate
and washings from the tin sulfide precipitate and boil until all H 2S is expelled.
Add HNO a (sp.gr. 1 42) until the iron is oxidized, which is shown by the solu-
tion becoming clear and of a yellowish color. Add about 5 g. of NH 4CI, then
NH.OH (sp.gr. 0.90) until a permanent precipitate forms. Boil, allow to
settle, filter on a loosely woven filter paper and wash with dilute ammonia
and with hot water. Dissolve the precipitate on the paper with hot dilute
Hel and add NH 40H (sp.gr. 0.90) to the solution again until a permanent
precipitate forms. Boil, allow to settle, filter on a loosely woven filter paper,
and wash with dilute ammonia and with hot water until all NH 4 CI is removed.
Dissolve the ,precipitate and wash the filter paper with 100 m!. of dilute H 2S0 4
for reductor. Pass the solution through a Jones reductor, wash first with
150 m!. of. the H 2S0 4 for reductor and then with 100 m!. of distilled water.
Titrate with the stand!!-rd KMn04 solution.
Accuracy.-Duplicate determinations should check within 0.03% of iron.
NOTEs.-l. A blank determination should be made on corresponding amounts of acid
and water passed through the reductor and the results should be corrected accordingly.
About 0.5 ml. of the permanganate will be required to give a permanent coloration to
the solution.
2. A small quantity of liquid should always be left in the reductor funnel, and air
should never be allowed to enter the body of the reductor.
3. A description and further details of use of reductor see "Estimation by Reduction
with Jones Reductor and Oxidation by Standard Permanganate Solution" and i, Notes on
Reductor Technique" in chapter under Molybdenum. .
4. For the determination of iron in alI'oys containing more tin than is present in thil;l
alloy see" Analysis of Bronl!!e Bearing Metal."
1358 ANALYSIS OF NON-FERROUS ALLOYS
boil the sulfuric acid solution vigorously for a few minutes.) Cool and dilute
with 100 ml. of water. Filter off the PbS0 4 and aJ)y tin that may be precipI-
tated at this point, washing with hot water. Dilute the filtrate to at least
200 ml.
Pass H 2S into the filtrate for 30 minutes.. This will precipitate any copper,
lead, tin, arsenic or antimony which may be presept. Allow the sulficles to
settle and filter, washing with acidified H 2S water. Boil the filtrate to one-
half volume to expel H 2S, add 3 ml. HNOa (sp.gr. 1.42) and boil for a few
minutes to oxidize the iron.
Precipitate the iron and aluminum according to the following (Blum,
J. Am. Chem. Soc., 38, 1282, 191.6). Add 10 ml. of HCI (sp.gr. 1.18) and
just neutralize with NH 4 0H, using methyl red as indicator. Boil for tw.o
minutes and filter, washing with hot 2% NH 4 Cl. Dissolve the precipitate
from the paper with boiling hot HCI (1 : 1) and wash alternately with hot
water and HCI (1 : 1) into the beaker in which the precipitation was made.
Reprecipitate the iron and aluminum following the same procedure as before.
Dissolve the precipitate again with hot HCI (1 : 1), dilute to 100 ml. with
water, make just ammoniacal, then only add acid enough to hold iro~ in
solution. Add sufficient saturated solution of ammonium bisulfite to reduce
the iron (5 to 20 drops). Add a few drops of methyl orange, quickly bring ,to
neutrality with NH 40H and then add 6 to 7 drops of HCl (1 : 1) in excess.
Finally add from 1 to 3 ml. of phenylhydrazine and stir until the precipitate
becomes flaky. The supernatant liqu6r should be plainly acid to litmus and
the precipitate may be colored owing to organic matter. Filter and wash with
phenylhydrazine wash solution. Ignite in a weighed platinum crucible, finish-
ing with the blast for five minutes. Weigh as AI 20 a. The weighing should
be made rapidly and with the crucible covered as Al 20 a is somewhat hygro-
• f
SCOpIC.
NOTES.-1. For small amounts of aluminum a 5 g. sample should be used. In this
case dissolve in 60 ml. of "as~ay solution" of the followil).g proportions: 7 vols. HN0 3
(sp.gr. 1.42). 10 vols. H,S04 (sp.gr. 1.84), 25 vols. H,O. .
2. When phosphorus is present, the final precipitatd will be AI,03 plus P,Os. In
this case it is necessary to fuse the residue with Na;CO" dissolve with hot water, acidify
with HNO, and determine phosphorus by the alkalimetric method. The phosphorus is
calculated to P,Os and subtracted.
3. For accurate work the AI 20, residue is treated with HF and H 2S0 4 for the removal
of possible Si0 2• . ,
4. Ammonium bisulfite solution may be prepared by saturating cold NH 40H (1 : 1)
witjl SO, until yellow. .- .
5. This method is especially satisfactory for small amounts of aluminum in brass
or bronze since no reagents are used which may contaminate the precipitate with alu-
minum, iron or !'ilica. A precipitate with phenylhydrazine gives positive indication of
the presence of aluminum. '
Dilute Hydrochloric Acid (1 : 3).-Mix 200 mi. of HCI (sp.gr. 1.18) with
600 mi. of water.
Acidified H 2S Water.-Dilute 10 mi. of HCI (sp.gr. 1.18) with 1000 mi. of
water and saturate the solution with H 2S.
2% NH4 CI Solution.-Mix 30 mi. of HCI (sp.gr. 1.18) with 200 mi. of
water, add methyl red, neutralize with ammonia until the solution changes to
a distinct yellow and then dilute to 1000 mi. with water.
Method.-Dissolve 2 g. of the sample in 20 mi. of HCI (sp.gr. 1.18) and
5 mi. oLHN03 (sp.gr. 1.42). Boil the solution to expel chlorine and dilute
with 50 mi. of water. Nearly neutralize the cold solution with 2.5% NaOH
solution (or stronger if much free acid is present) and pour it slowly with
constant shaking into a 500-mi. volumetric flask containing 100 mi. of sodium
sulfide solution. Dilute to the mark with 2.5% NaOH solution and mix thor-
oughly. Filter on a large, dry, No. 42 Whatman filter paper or its equivalent
and keep the paper well filled with the solution lest iron be oxidized and
dissolved. Reject the first 20-25 mi. and catch exactly 250 mi. which will
represent 1 g. of sample. In routine analysis of material containing not
more than 1 % of aluminum the. aliquot portion may be gathered in a 250-ml.
graduate.
Transfer the aliquot portion to a 600-m!. beaker, neutralize the solution
with dilute HCI (1 : 3) and finally add 25 mi. excess. Digest at 40-60° C.
for one hour and filter and wash the paper and precipitate with acidified H 2S
water.
Boil the filtrate and washings to expel H 2S, add a few drops of methyl
red indicator and then dilute ammonia (1 : 2) until the solution is just dis-
tinctly yellow. Boil for two minutes and filter at once through a small filter.
Wash the beaker, paper and precipitate two or three times with hot 2% NH 4 CI
solution and discard the filtrate. Dissolve the precipitate in 20 mi. of hot
dilute HCI (1 : 3), wash the filter thoroughly with small portions of water and
reserve it for the second filtration. Dilute the filtrate to 50 mI., add methyl
red and precipitate with dilute ammonia as before. Filter, wash with hot 2%
NH 4CI solution and ignite and weigh.
The ignited residue is prone to carry Si0 2 and mU$t be purified before
weighing as follows. Add one or two drops of water, one drop of diluted
H 2S0 4 and 1 to 5 ml. HF, Evaporate to dryness, 'increase the heat slowly
and finally heat with a blast lamp or its equivalent. Weigh as Al 20 3. The
weight of A1 20 3 corrected for the blank and multiplied by 52.91 gives the per
cent of aluminum.
NOTEs.-The first Al(OHl, precipitate will carry down some NaCl and much of any
silica that wa~ dissolved in the NaOH solution: hence it is not safe to omit the second
precipitation and HF treatment. There is no attack on the glassware during the short
contact with the weak cool alkaline solution. It is, of cour"e, preferable that the NaOH be
free from aluminum. The most satisfactory test for it lies in running an analysis with
non-ferrous alloys containing no AI. Direct test, by acidification followed by precipita-
tion. with ammonia is sufficient if no precipitate is obtained, but is of doubtful value in
case one appears, for it may contain other elements, such as iron, which do not affect the
results.
If phosphorus is known to be present the method should be followed through the first
addition of ammonia. If aluminum is indicated, re-acidify the solution with HCl, add
macerated paper, two drops of methyl orange and 10 ml. of a solution of diammonium
1362 ANALYSIS OF NON-FERROUS ALLOYS
. .
phosphate (10%). Render the solution just ammonIacal, then just restore the pink color
with dilute Hel (1 : 3), heat to boiling and add 30 mi. of 25% ammonium acetate. Boil
for five minutes, filter on an 11 cm. No. 42 Whatm!j.n paper lind wash with hot 5%
NH 4NO. solution until 5 ml. of the washings no longer give a test for chlorides. Ignite
in platinum or porcelain, heat at about 1100 C. for JO minutes and weigh as AlP0 4•
0
This method is not strictly accurate on account of t,he uncertain composition of the
phosphate but is sufficiently accurate for all but the most painstaking analyses. .,
DETERMINATION OF TIN
(FOR SAMPLES FREE FROM PHOSPHORUS)
Method.-See the" Determination of Copper by the Electrolytic Method."
Use the papers containing the meta-stannic acid which was filtered off in
the "Determination of Copper by the Electrolytic Method." Ignite the
papers with the precipitate while moist in a porcelain or platinum crucible,
slowly at first, and finally at the full heat of the burtler. If the precipitate
8 A. S. T. M. Methods of Chemical Analysis '()f'Metals, p. 157, December, 19;36.
ANALYSIS OF NON-FERROUS ALLOYS 1363
weighs more than 20 mg., heat to constant weig!lt, using a bl~st lamp. Weigh
as Sn02, and calculate to Sn.
NOTES.-l. This method is not accurate for. alloys which contain phosphorus, or
which contain over 0.25% of iron. See" Determination of Tin, Analysis of Manganese
Bronze."
2. The ignited tin oxide contains small amounts of iron and copper as well as any
phosphorus, antimony or arsenic which may be present in the alloy. Qualitative tests
should be made for phosphorus, antimony and arsenic. In the absence of antimony
and arsenic the following" Qualitative Test for Phosphorus" is here made use of to detect
- phosphorus and to estimate its amount, if any is found.
phorus. The remainder is considered as pure Sn02. The factor for converting
phosphorus to % X P 20 5 is 152.6%, or approximately 172 times the phosphorus
content of the sample.
If the alloy contains over 0.20% of phosphorus, a special method for the
determination of tin should be used.
In alloys which contain over 0.25% ,of iron, use the method for the" Deter-
mination of Tin, Analysis df Manganese Bronze."
until the meta-stannic acid has dissolved. Filter and wash thoroughly with
dilute yellow ammonium sulfide. Acidify the cRmbined filtrate and washings
cautiously with 50% acetic acid. Warm, and allow the precipitated tin sulfide
and sulfur to settle. Filter on double papers I and wash with ammonium
acetate solution. . Dry the precipitate, a~d proceed exactly as in " Determi-
nation of Tin, Analysis of Manganese Bronze." I
NOTEs.-l. The ignited tin oxide will contain any arsenic or antimony that was"
present in the sample. For its correction see Note 4, "Determination of Tin," Analysis
of Manganese Bronze.
2. In the case of alloys containing smaller amounts of tin and over 0.25% of iron
proceed as in "Determination of Tin, Analysis of Manganese· Bronze."
DETERMINATION OF IRON
'Solutions Required. Yellow Ammonium Sulfide and Dilute Yellow Am-
monium Sulfide for Washing.-See " Determination of Tin."
Other Solutions.-See " Determination of Iron, Analysis of Manganese
Bronze." •
Method.-Proceed exactly as in " Determination of Copper" up to the-
point where the tin has been separated as H 2Sn03. '
Retain both the filtrate and the precipitate. Transfer the papers with the
precipitate to a ISO-ml. beaker, and cover them with 40 to 50 ml. of yellow
ammonium sulfide. Warm about 15 minutes, or until the meta-stannic acid
is dissolved. Filter and wash thoroughly with hot, dilute yellow ammonium
sulfi-de.
Dissolve the small black residue on the filter with hot dilute HCI, receiving
the solution in a small beaker. Add 5 ml. of- HN0 3 (sp.gr. 1.42) and boil to
oxidize sulfur and iron. Add this small solution to the original filtrate from
the meta-stannic acid. Add about 5 g. of NH 4 Cl, then NH 4 0H (sp.gr. 0.90)
and proceed as in the" Determination of Iron, Analysis of Manganese Bronze."
0.90), and add 20 ml. in excess. The final volume should be about 150 ml:
Electrolyze with a current of 0.3 ampere per solution. Test the solution for'
completion of deposition with fresh H 2S water as is done in the (( Determina-
tion of Copper by the Electrolytic Method." %move the cathode, wash with
water and dry with alcohol as in the" Determination of Copper by the Elec-
trolytic Method." Weigh the deposit as metallic, nickel.
DETERMINATION OF ZINC
After the copper has been. completely dep0sited by electrolysis, quickly
remove'the cathode and wash it thoroughly with distilled water, catching
the washings in the beaker containing the electrolyte. Heat the solution and
saturate with H 2S gas. Filter off any precipitate, and boil the solution to
expel H 2S. Make the solution barely alkaline with NH 40H, and add 25 ml.
of 85% formic acid. Dilute the solution with distilled water to 300 mI., heat'
to boiling, and saturate with H 2S gas. Filter, wash with hot water and dis-
solve the precipitate with hot HCI (1 : 1) and transfer the solution to a weighed
platinum or porcelain dish or crucible. 'Add a few drops of H 2S0 4 (sp.gr. 1.84)
and evaporate the solution until copious fumes escape. If the solution is not
clear and colorless, cool, add a few milliliters of HNO a (sp.gr. 1.42) and again
evaporate the solution until fumes of H 2S0 4 come off freely. Repeat the
treatment with HNO a, if necessaryvuntil the organic matter is destroyed, and
the solution is colorless. Remove the excess of H 2S0 4 by heating the dish
cautiously, and finally heat to dull redness. Weigh as ZnS04 and calculate
to Zn.
with HzS gas. Filter off and reject the. preci~itate: B~il the filtrate to expel
H.S and add HN0 3 (sp.gr. 1.42) untll the 1ron 1S oX1dized. Add NH 40H
(sp.~r. 0.90) until the ~olut~on is alkalin~. Boil ~nd filter on a loosely woven
filter paper. Wash w1th d~lute amrr.lOma and w1th hot water. Dissolve the
precipitate on the filter w1th hot d1lute H 2S0 4, rec(\iving the solution in a
350-m!. Erlenmeyer flask. Add NH 40H Csp.gr. 0.90) until the iron is entirely
precipitated, and then add HNO a Csp.gr. 1.42) cautiously until the solution
just becomes clear. Bring the solution to a temperature of about 80° C., and
add 40 ml. of ammonium molybdate at room temperature .. Allow to stand
for one minute, shake or agit~te for ~ minu.tes, filter ~I). a 9 cm. paper, and wash
very thoroughly (about 25 t1~es) w1th aCld a~momum sulfate. Dissolve the
precipitate on the paper usmg 50 ml. of d1lute ammonia. Add sufficient
H2S04 to adjust the acidity to 2.5% by volume and immediately pass the
solution through a Jones reductor, which has the reductor tube prolonged
and reaching nearl;y to the bottom of the flask, dipping into 50 ml. of fe_rric-
alum solution. Wash through the reductor with 150 ml. of the H 2S0 4 for
reductor and follow with an additional 100 ml. of liistilled water. Titrate
with 'standard KMn04' '
By this method the molybdenum which in passing through the reductor
R&S M.f'.D reduced f'.ntirt'Jy to the form corresponding to the oxjde .A102 0J' js
partially oxidized by the ferric alum, an equivalent amount of iron being
reduced to the ferrous condition. The resultant solution is not rapidly oxidized
through contact with air during the titration, and the, calculations can there-
fore be made on the M0 20 a - 2MoO a basis.
Accuracy.-Duplicate determinations should check within 0.01 % of .ph6s-
phorus.
NOTEs.-The ammonium-molybdate solution should be kept in a cool place and
should always be filtered before using.
A blank determination should be made on corresponciing amounts of acid and
water, passing through the reductor into the usual amount of ferric-alum solution in the
flask.
A small quantity of liquid should always be left in th() reductor funnel, and air
should never be allowed to enter the reductor.
For a description of the Jones reductor consult, "Determination of Iron, Analysis of
Manganese Bronze."
Transfer the papers containing the tin al1d antimony to a 350-ml. Erlen-
meyer flask and add 25 ml. of HNO a '(sp.gr. L~2), 5 g. of ammonium persui-
fate, and 15 ml. of H 2S0 4 (sp.gr. 1.84). Boil to strong fumes of SOa. If
the solution is brown, cool and add 5 ml. of HNO a (sp.gr. 1.42) and a little
more (1 g. or less) persuifate, if necestlary, arid boil again to strong fumes.
of SOa. When the ,solution is colorless, cool, ~dd 20 ml. of distilled water
20 ml. of HCI (sp.gr. 1.18) and 1 g. of sodiuin sulfite. Boil gently for 10
minutes to expel S02 completely. Dilute with 200 ml. of distilled water and
cool to 10 to 12° C. in ice or under running water.
Titrate with standard permanganate solution until the appearance ot a
decided pink color. '
NOTES.-l. In alloys containing a large excess of tin, the antimony is precipitated
quantitatively with the meta-stannic acid. .
2. Ammonium persulfate and, nitric acid are used to destroy the filter paper, and
subsequent boiling with sulfuric acid effects the solution of the tin.and antimony. A
mixture of fuming nitric and sulfuric acids might be used to accomplish.,the same results,
but the persulfate mixture is much more efficient. .
3. If the solution becomes cool during the first filtration, some of the meta-stannic
acid may dissolve.
4. If arsenic is present, correction should be made for it.
DETERMINATION OF SULFUR
Method.-Weigh out four 5 g. samples of bronze into· 250-ml. beakers.
Dissolve each in 25 m!. of HNO a (sp.gr. 1.42) and boil until the disappearance
of red fumes. Add 125 ml. of boiling water and allow to settle for one hour,
or longer, keeping the temperature just below the boiling point. Filter on
double 12.5 cm. closely woven papers, keeping the solution hot. Wash with
boiling water. Discard the meta-stannic acid precipitates, combine the four
filtrates in a tall 700-ml. beaker and add 0.5 g. of sodium carbonate, and
evaporate the solution until syrupy. Cool somewhat, then dilute to the lip
with warm water. There should be present a ll}'yer abo}lt 6 mm. in thickness,
composed of basic salts of copper. Electrolyze, using large stationary elec-
trodes, at a current of 4 amperes, or using the Frary solenoid at a current of
about 10 amperes. The lead must be removed at the anode as the copper is
taken out at the cathode. When copper and lead are out, evaporate the
electrolyte to small volume, cover the beaker, add 75 ml. of HCI (sp.gr. 1.18)
and boil down to small volume. Add 75 mL mo1'e·lICI and evaporate .to
dryness to dehydrate any silica present. Moisten with about 0.5 ml. of HCI
and take up with 20 ml. of hot water. Filter into a sman beaker, Heat the
filtrate to boiling and add 10 m!. of 5% BaCh solution, drop by drop, with
stirring. Allow to settle at least five hours, filter on a small ashless paper,
and wash with hot water. Ignite cautiously in a small porcelain or platinum
crucibl~, and then heat to a bright redness for 20 minutes. Cool and weigh
the BaS04 and calculate to sulfur. ' .
. NOTES.--;-l. Throughout the entire analysis, the utmost precautions must be taken tU
prevent contamination by sulfur in any form, H 2S0 4 fumes, HzS and S02 gases especially.
All heating with gas burners must be avoided. Run a blank'with each analysis, adding
the reagents and subjecting it to the same ope_!}l.tlons as ·the beakers containing the
dl'illings. Subtract the amount of BaS04 founa in the blank from that in the deter-
ANALYSIS OF NON-FERROUS ALLOYS 1371
minations. If this blank amounts to 0.002 g., it is a sign of poor work or impure re-
agents and the analysis should be repeated. 'Vash out all beakers, funnels, etc., with
distilled water before using.
2. It is necessary to keep the solution containing the meta-stannic acid hot during
filtration. If the solution becomes cool, some of the meta-stannic acid may dissolve.
DETERMINATION OF ARSENIC
Solutions Required. Ammonium Carbonate.-Dissolve 150 g. of ammo-
nium carbonate in one liter of distilled water. \
Method.-Dissolve the sample (5.00 to 10.00 g.) in a 100-ml. beaker in
10 ml. of HNO,,(sp.gr. 1.42) and 30 ml. of HCl (sp.gr. 1.18). When solution
is complete, cool, add 35 rol. of H 2S0 4 (sp.gr. 1.84), evaporate to the appear-
ance of copious fumes of H 2S0 4, and continue the fuming for from ten to fifteen
minutes. Cool, take up in -HCI (sp.gr. LIS), transfer to an SOO-ml. flask
ha.ving a two-hole stopper with' separatory funnel and outlet t'ube connecting
with a vertical condenser outside. (The part of the outlet tube within the
distilling flask should have a steam outlet hole about 1 cm. from the-open end.)
Rinse the beaker with HCl (sp.gr. 1.18) and dilute the solution with the acid
to make a total volume of 350 to 450 m!. Add 10 g. of cuprous chloride
(Cu 2CI 2), connect the flask with the condenser, and gradually raise the contents
of the flask' to boiling. Catch the distill~te in a tall 400-ml. beaker containing
100 ml. of cold water. The end of the condenser should extend several milli-
meters into the water in the beaker and the beaker should be surrounded with
ice water in order to avoid any possible volatilization of arsenious chloride.
After about, 100 m!. of distillate has passed over, remove the flame from the
distilling flask, pour 3 to 5 ml. of hypophosphorus acid through the separatory
funnel into the.distilling flask, and redistill as before, catching the distillate
in the same beaker. The volume of the second. distillate should be 50 to 100
ml. Pass a rapid current of H 2S through the distillate for one hour, allow to
settle, and filter on a weighed Gooch crucible: 'Vash twice with HCl (1 : 1),
/,w.,.'l' t~;""'i'i.\5W N'.._:t.._"1;0 ..%\1& ;<rate.", t}l'.",,%' t.._:.._W6l5' w':t}i' 8k-ml.._ml, <'NN~ }03.',,_" j}.._ID.,OS w':t.._"1;0 Cs..!.
Dry at 110 0 C. for 10 minutes, and weigh after cooling. Dissqlve the As 2S a
\"ith (NH4)z'CO a solution, wash thoroughly with water; dry for 30 minutes
at 110 0 C. and reweigh after cooling. The difference in ,,,eights ,multiplied
by 0,609 gives the weight of arsenic.
NOTE.~Blank determinations should be made on the reagents carried along as in
the method and the proper cOlTections made.
and dilute to 200 ml. with water, boil for a few minutes and allow to cool.
After two or three hours filter off any insoluble matter, collecting the filtrate in a
tall 600-ml. beaker. Discard the residue tin the paper and dilute the filtrate to
400 ml. Heat to boiling and pass a rapid ftream of H 2S through this hot solu-
tion for thirty minutes. (It is well to tho.roughly mix the copper sulfide with
the solution by means of rubber-tippeq ro(i so as to keep the precipitate from
being forced out of the solution by the gas.)
At the end of the half hour filter off the precipitated sulfides on an 18Yz
cm. paper, receiving the filtrate in a tall fiOO-ml. beaker." Wash the sulfides
once with hot water. If proper precipitation has been made, this filtration
will be rapid and easy to make and the filtrate will have no blue color. Return
the precipitated sulfides with the filter paper to the beaker i~ which precipita-'
tion was made and add 400 ml. of 10% H 2S0 4 • Boil vigorously for a half hour
and filter hot on a 19Yz cm. paper. Wash the residue thoroughly with boiling
water. Discard the residue on the paper. All of the cadmium is now present
in these two solutions.
Cool and pass a rapid stream of H 2S through the solution for ten minutes.
Add ammonia slowly to each solution until zinc sulfide begins to precipitate.
When this is certain, continue the passage of the gas for five minutes more and
then filter off the precipitates fjeparately on 12Yz cm. paper. Discard the
filtrates. Combine the two papers containing mixed sulfides and boil with
100 ml. of 10% H 2S0 4 for ten minutes, stirring frequently; Filter and wash.
Cool and dilute to 300 ml. Pass in H 2S and add ammonia drop by drop until
cadmium is entirely precipitated. Pass gas for about twenty minutes. Filter
the precipitate on an 11 cm. close texture. paper. Return the paper to the
beaker and boil with 50 ml. of 10% H 2S0 4 • Filter and wash. Dilute to 300 ml.
and pass in H 2S for twenty minutes. A drop or two of ammonia wilL be suffi-
cient to start the precipitation of cadmium. •,
Filter the cadmium sulfides on a weighed Gooch crucible. Wash twice
with cold water. Dry at 110° C. for two hours and reweigh.
Weight CdSXO.778= )Veight cadmium.
The final cadmium sulfide precipitate will have a characteristic brigh't, yellow color
and will be entirely free of other sulfides if the foregoing directions be carefully followed.
As a special precaution when arsenic or tin are suspected in the brass the final sul-
fide may be washed out of the crucible after weighing and treated with yellow ammonium
sulfide, (the addition of NH 4Cl will tend. to prevent cadmium sulfide being carried
into the ammoIJ.ium sulfide solution as a colloid,) - Filter and wash the residue and
dissolve the cadmium sulfide in 10% H 2S0 4, reprecipitate and weigh again.
ANALYSIS OF NON-FERROUS ALLOYS , 1373
DETERMINATION OF IRON
Solutions Required. Stannous Chloride.-Dissolve 450 g. of the salt in
380 ml. of concentrated HCI (sp,gr. 1.18) to which some water has been added
and dilute to 2 liters,
Mercuric Chloride.-Make a saturated solution of the salt in hot water,
allow to cool to room temperature and filter.
" Manganese Sulfate.-Dissolve 160 g. of the salt in water and dilute to
1750 m!. To this add 330 ml. of phosphoric acid (sp.gr. 1.70) and 320 m!. of
sulfuric acid (sp.gr. 1.84). This solution is to obviate the deterious action of
hydrochloric acid upon the standard potassium permanganate. The phos-
phoric acid allows the formation of iron phosphate, which, being nearly colo'rless,
renders the end reaction more distinct.
Potassium Permanganate.-Make a solution containing 0.3163 g. of salt
per liter. This is approximately N/I00.
Standardization of above Permanganate Solution.-Take measured quan-
tities of a carefully standardized ferric chloride solution, containing 5 g. of
metallic iron and 100 ml. concentrated hydrochloric acid to a liter in No. 1
beakers. About 5 m!. of this solution should prove adequate. Add from a
burette stannous chloride solution until the iron is almost reduced, told by the
color change. Then heat to boiling and reduce iron completely with 811CI 2,
using a slight excess of the reagent. Add 5 m!. of HgCh solution to oxidize
excess of 8nC1 2 • Pour solution into a 600-ml. beaker containing 6-8 ml. of
the Mn80 4 solution in 400 m!. of water. Titrate with KMn04 immediately
until a faint pink is obtained. Calculate potassium pefmaI\~anate in terms of
metallic iron.
Treatment of Sample.-Weigh out a 5 g. sample of drillings from which
all steel chips have been removed by a magnet and transfer to a 250-ml. beaker.
Dissolve in as little nitric acid (sp.gr. 1.42) as possIble (25 ml. should be suffi-
cient), heat until all nitric oxide fumes have been dliven off and then dilute
to 100 ml. with distilled water. Add 3 to 4 g. of ammonium chloride crystals,
heat to boiling, and add a slight excess of ammonia to precipitate all the iron
as hydroxide.
A small amount of zinc.is usually precipitated at this point as hydroxide
and also, if manganese is present in the alloy, it will come down as manganese
hydroxide.
After the precipitate has settled down, filter through filter paper and wash
with ammonium hydroxide 1 : 4 until all soluble salts are removed (or until'
1374 ANALYSIS OF NON-FERROUS ALLOYS
the blue copper color has been washed from the filter paper). Treat the pre-'
cipitate on the filter paper with a little hot HGI (1 : 1), catching the resulting
solution in original beaker. Wash the filter p~per alternately with hot water
and HCI (1 : 1) until the iron has been cOIfpletely removed. Dilute the
hydrochloric acid solution in the beaker to about 75 m!. and add 3-4 g. of am-
monium chloride. Heat to boiling and repre~ipitate the iron with a slight
excess of ammonium hydroxide. Filter and w'ash as before. (A double pre-
cipitation is usually sufficient to remove zinc, etc.) Dissolve the precipitate
in a small amount of hydrochloric acid (1 : 1). Add stannous chloride to
reduce the iron-2 or 3 drops are usually sufficient.
Heat almost to boiling and add 1 or 2 drops more of stannous chloride to
be sure an excess is present. Run in an excess of mercuric chloride solution
(about 5 m!.) to oxidize the stannous chloride. •.
Pour into a 600-ml. beaker containing 6-8 ml. of the manganese sulfate
solution in 400 m!. of water. .
Titrate with standard potassium permanganate solution until a slight pink
opalescence is obtained. - ,
DETERMINATION OF MANGANESE
Proceed exactly as in "Deterniination of Manganese by the Persulfate
Method" or "Determination of, Manganese by the Bismuthate Method,"
under" Analysis of Manganese Bronze."
DETERMINATION OF NICKEL
Solutions Required. Standard Solution of Potassium Cyanide.-Dissolve
26 g. of pure potassium cyanide in 1000 m!. of distilled water and add 1.25 g. of
silver nitrate. This solution is standardized against a solution of known metal
content. 7 '
Potassium Iodide.-Dissolve 100 g.lof c.p. potassium iodide crystals in
1000 m!. of distilled water. ,
Sodium Pyrophosphate.-Make a saturated solution of c.p. sodium pyro-
phosphate crystals in distilled water. I
Method.-See the" Determination of Copper, -J,ead and Tin."
Transfer the electrolyte and washings from the electrolytic determination
of copper to a 400-m!. beaker. Add 3 to 4 g. of ammonium chloride and then
50 to 75 m!. of a saturated solution of sodium pyrophosphate, depending'
upon the amount of zinc present. Ne.utralize with ammonium hydroxide
(sp.gr. 0.90) and then add enough to make the solution only slightly but yet
distinctly alkaline. A large excess of ammop.ia will hinder 9r entirely prevent
the reaction. Cool, and do not let the temperature get much above 20° C.
or the result will be irregular. Add 2 m!. of a 10% Ilolution of potassium
iodide and titrate with a standard solution of potassium cyanide containing
silver nitrate until the white cloud caused by the formation of silver iodide
has entirely disappeared and one drop of th:e: cyanide causes a clear solution.
This is the end point which is very sharp ind distinct.
ANALYSIS OF NON-FERROUS ALLOYS 1375
XOT'£S.-Tbe method of estimating nickel by means of adding potassium cyanide to
an ammoniacal solution of nickel containing silver iodide in suspension !'O that the solution
rcmaint.'<i turbid until the nickel W88 aU converted into the double cya.nide of nickel
8.n~ potassium,'after which 8. single drop of the .cyanide in eXC(!f!8 .cleared up ~he solution,
was originally propose:d by T. Moore and modified by \V. B. Price for use 1D the deter-
mination of nickel in German rover, etc. ..".~
The following elements interfere with the metboo: ~fanianese, zinc. iron. aluminum
and especially copper, which, owing to its forming cYIlnides., ·would render the method
valueless. Cobalt, if present, will be cstimn.ted fL'I;nickel: The iron, aluminum nnd
tine may be kept in solution so that they do.lnot interfere by too ure of organic adds
6uch 8.8 tartaric or citric or with sodium pyroph06pha.te.
DETERMINATION OF ZINC
Zinc is uruaUy taken by difference.
In case an aetua. determination is desired proceed exactly as in " Deter-
mination of Zinc,-Analysia of Bronzo Bearing IVletals.JI
, DETEiufINATroX OF LEAD
To the
•
filtrate from the H,SnO, separation, add 8 ml. of H,SO. (sp.gr.
1.84) and evaporate until copious fumes of H_ 2 SO, are evolved. Cool, a.dd
30 mi. of distilled water and. boil until all bismuth sulfate is dissolved. Cool,
filter on a weighed Gooch crucible and wash '"ith cold dilute H,sO. (10%).
Ytlter as soon &8 possible, otherv.-ise some basic bismuth sulfate will separate out.
Set the Gooch crucible inside a porcehdn crucible, dry, ignite for 5 minutes at
tbe fuJI heat of a Tirrill burner, cool and weigh as PbSO. which contains 68.33%
of lead.
NOTE.-A small amount of lead still reme.iDB in the solution to be recovered later.
(See Note: "Determination of Bismuth.")
1376 ANALYSIS OF NON-FERROUS ALLOYS
DETERMINATION OF COPPER
To the filtrate from the PbS0 4 separation add 0.11 g. of iron wire dissolved in
10 ml. of HNO a (sp.gr. 1.42) and thel/ add NH49H (sp.gr. 0.9) in sufficient
excess to hold the copper in solution. NoW add 10 ml. of a saturated solution
of ammonium carbonate, bring to boiling, filter I and wash with hot water
containing a small amount of (NH 4 )2CO a• Reselive the filtrate. Wash the
precipitate back into the original beaker and dissolve any remaining pre-
cipitate on the paper with 20 ml. of hot RNO. (1 : 2) catching the resulting
solution in the original beaker. Wash several times with hot water. Warm I
until the precipitate is completely dissolved and make a second separation
of iron and bismuth with ammonium hydroxide and ammonium carbonate.
Again dissolve the precipitate with nitric acid and make a third separation.
Hold the residue.
Combine the three filtrates and washings from the separation of iron and
bismuth, which should not exceed 400 ml., make just acid with H 2S0 4 (1 : 1),
add 10 ml. in excess, and evaporate down to a volume of 100 ml. Cool, add
1.5 ml. of HNO. (sp.gr. 1.42) and electrolyze for copper. (See" Determination
of Copper by the Electrolytic Method, Analysis of Manganese Bronze.")
DETERMINATION OF BISMUTH
Treat the precipitate containing tile bismuth and ii:on on the paper with a
little hot HCl (1 : 1) catching the resulting solution in a clean beaker. Wash
the paper alternately with hot water and hot dilute HCl (1 : 1) until the bis-
muth has been completely removed. Neutralize the free acid with NH 4 0H
(sp.gr. 0.9), make slightly acid with HCl and pass H 2S. Filter and wash with
H 2S water. (Discard the filtrate unless it is desired to run zinc or nickel on
this portion.)
Wash the precipitate of bismuth sulfide back into the beaker and pour
through the filter paper 30 ml. of hot HN0 3 (1 : 25 catching the solution in the
same beaker. Wash a few times with hot water! Heat the solution until the
precipitate is completely dissolved. Filter' through the same filter paper and
wash with hot dilute HNO. (1 : 50). If a black insoluble residue still remains,
transfer the residue to a small porcelain crucible, dry, ignite over a low flame,
cool, dissolve in 1 ml. of HNO. (sp.gr. 1.42) and combine with the main filtrate.
Boil the combined solutions down to a volume of 60 ml. and-allow to stand and
settle for one hour keeping the temperature just below the boiling point. If a
precipitate of H 2SnO. forms, filter, wash, dry, ignite and weIgh as Sn02' Add
the weight found to that already obtained (see" Determination of Tin.").
Evaporate the filtrate to a volume of about 50 mI., make just neutral with
NH 4 0H (1 : 3) and then add 5 ml. of HCl (1 : ,9) in excess. If a precipitate of
BiOCI forms, cautiously heat until it goes into solution. Dilute to a volume of
400 ml. with boiling hot water and allow to stand in a warm place several hours.
Filter oil a weighed porcelain Gooch crucible and wash with hot water. Finally
wash once with 95% alcohol but transfer the filtrate to' a' clean beaker before
doing so. Dry at 110° C. for 30 minutes, cooLa.nd weigh as BiOCI. Multiply
by Q.8025 for bismuth. . _. -
ANALYSIS OF NON-FERROUS ALLOYS 1377
NOTE.-The filtrate from the BiOCI separation contains a small amount of lead that
did not precipitate out as PbSO. (see "Determination of Lead"). This may be re-
covered in the folloWing maimer: make the solution just alkaline with ammonium
hydroxide, pass R 2S, filter and wash. Dissolve the precipitate from the filter paper
with 15 m!: of hot RNO, (1 : 2) and wash with hot water. Add 12 m!. of RNO, (sp.gr.
1.42), dilute to a volume of 150 m!. with distilled water and electrolyze for lead (see
"Determination of Lead by the Electrolytic Method, Analysis of Manganese Bronze").
Add the weight of lead found to that already obtained.
DETERMINATION OF NICKEL
Method.-Remove the tin and lead as in "Determination of Tin" and
" Determination of Lead." Remove the copper by electrolysis (see" Deter-
mination of Copper by the Electrolytic Method, Analysis of Manganese
Bronz~ "). Continue as in " Determination of Nickel, Analysis of German
Silver."
N OTE.-1. If the amount of lead is small, the copper and lead may be removed at the
same time by electrolysis (see" Determination of Copper and Lead Simultaneously by
the Electrolytic Method, Analysis of Manganese Bronze"). It is not necessary to
evaporate to dryness.
2. The presence of a moderate amount of bismuth will not prevent the electrolytic
precipitation of copper. The deposit, however, will be black and possibly spongy, and
may slough off in washing, in which case, filter and wash and combine washings and main
filtrate. If the spongy material has dropped off into the electrolyte it will dissolve.
In such a case filter both electrolyte and washings through the same paper, wash, com-
bine filtrate and washings, pass R 2S, filter and wash, boil to expel H.S, oxidize l "he iron
with RN0 3 (sp.gr. 1.42) and proceed ~ directed above.
DETERMINATION OF ZINC
Method.-Remove the tin, lead and copper as directed under" Determina-
tion of Nickel." Continue as in " Determination of Zinc, Analysis of Bronze
Bearing Metal."
DETERMINATION OF IRON
Method.-Proceed exactly as in "Determination of Iron, Analysis of
Bronze Bearing Metal," up to the point where the H 2SnOa has been separated
and any iron present in the H 2SnOa has been recovered and combined with the
main filtrate. Adq. 5 g. of ammonium chloride crystals, and a slight excess of
NH 40H (sp.gr. 0.9). Boil, allow to settle, filter on a loosely woven filter paper
and wash with dilute ammonium hydroxide and with hot water. Treat the
precipitate on the paper with a little hot HCl (1 : 1) catching the lesulting
solution in original beaker. Wash the filter paper alternately with hot water
and HCl (1 : 1) until the iron has been completely removed. Neutralize the
solution with NH.OH (sp.gr. 0.9), make acid with 5 m!. of Hel (sp.gr. 1.18)
in excess, dilute to a volume of 100 ml., warm and pass H 2S. Filter and wash
with hot water. Boil the filtrate to expel H 2S, add a few m!. of HNOa (sp.gr.
1.42) to oxidize the iron, dilute to a volume of 75 mI., add 3 to 4 g. of ammonium
chloride crystals, and a slight excess of NH.OH (sp.gr. 0.9) and continue as in
" Determination of Iron, Analysis of German Silver."
1378 ANALYSIS OF NON-FERROUS ALLOYS
DETERMINATION OF ALUMINUM
Proceed exactly as in" Determination of AluJinum, Analysis of Manganese
Bronze."
\
DETERMINATION. OF MAfOANESE
Proceed exactly as in" Determination of Manganese, Analysis of Manganese
Bronze." .
DETERMINA,rION OF IRON tl
Solutions Required. Cuprous C/lloride Solution.-Place into a 2-liter
heavy glass bottle, 400' g. of NaCI, 100 m!. of HCI (sp.gr. (1.18) and 20 g. of
Cu 2 Cb. Add 2 liters of warm water; shake well; place into the solution a few
copper rods and let stand over night in ~ cool place. After shaking, the solution
is ready for use. One milliliter of Cu2Cl2 should reduce 0.003 to' 0.005 g. of iron.
Standard Ferric, Chloride Solution.--Dissolve 5 grams of standard iron wire
in 100 m!. of hydrochloric acid (1 : 1), then 5 grams of potassium chlorate are
added to oxidize the iron. Boil until all the free chlorine is driven off, then cool
and make the solution up to two literG·
Method.-Dissolve one gram of drjIlings in 25 ml. of nitric acid (1 :-'1), in a
300-m!. Griffin lip beaker, add 50 m!. Of cold water and make strongly alkaline
with ammonia. Boil a few minutes, filter and wash precipitate with hot water
two or three times. The precipitate is redissolved in hot hydrochloric acid
(1-3) to which a few drops of sulfurous acid are added. After the precipitate is
dissolved, boil the solution a few minutes, add a few drops of nitric acid;Iet solu-
tion cool for 5 or 10 minutes, then make alkaline with ammonia, filter and wash
with hot water. Repeat redissolving ttrid ammonia precipitation. After final
reprecipitation, dissolve the iron hydrOxide in 30 ml. of hot hydrochloric acid
(1-3) into a 24-oz. flask, cool and add ;300 m!. of cold distilled water and titrate
\vith a standard cuprous chloride solution, using 2 drops of a saturated solution
of potassium sulfocyanate as inside indicator. The cuprous chloride solution is
standardized each time before using, \#"ith standard ferric chloride.
NOTE.-If the ferric hydroxide is dissolved in dilute- sulfuric acid and the iron re-
duced with zinc, the iron titrated with potassium permanganate is just as accurate as
above method, providing 5 grams of the drillings are taken.
DETERMINATIO:!i OF MANGANBSE·
DETERMINATION OF CARBON
cool in ice water for 10 minutes. Add 50 m!. more 'alcohol from a pipette, allow
to stand in ice water for 20 minutes and filter through 9 cm. paper into an
800-m!. beaker. Wash by decantation three timEfs with alcohol-hydrochloric
acid mixture and finally wash the paper twice with the same mixture. Wash
the PbCl 2 from the paper back into the beaker; w~sh paper several times with
hot water, catching washings in beaker with the bulk of the chloride. Finally
wash paper with hot acid ammonium acetate soluti'on. Heat until all PbCl 2 is
dissolved. Add 15 m!. of saturated solution of K 2 Cr 2 07; heat until precipitate
is of good orange color; filter on weighed Gooch crucible; wash with water,
alcohol and ether; dry at 100° C. and weigh. Calculate percentage of lead
by the empirical factor 63.75.,
NOTEs.-1. With proper manipulation, the PbCI. should contain consistently all but
0.003 g. of lead. When, however, the lead content is low as in the high tin Babbitts, it
is not necessary to use 70 to 100 ml. of mixed acid or to add alcohol after 50lution of the
alloy. In such cases, weigh th_e alloy directly into a 600-ml. beaker, add about 15 ml.
of HCI (1 : 1) warm, and then add HNO s (sp.gr. 1.42), a little at a time, until the alloy
is dissolved and the tin completely oxidized.
2. During the heating of the acid ammonium acetate to dissolve the PbCl., the
solution should remain perfectly clear, any turbidity indicating the .presence of tin or
antimony, as even 1 mg. of tin or antimony will cause a slight but distinct turbidity .•
DETERMINATION OF TIN
NOTE.-l. If the electrolysis is carried out over night, it will usually be found by
morning that the solution has become alkaline and it milY be taken for granted that the
tin has all been precipitated on the cathode. The best results are obtained by regulating
the current, or the time, so as to render the solution alkaline only a very short time
before the cathode is to be removed. The anode shouldl touch the bottom of the beaker
to insure proper agitation of the electrolyte. Ifhe catJiode should be 0.25 inch above.
2. Appreciable amounts of nickel, zinc and iron shbUld not be present.
I
B. DETERMINATION OF ARSENIC
to go over into the crucible. Now add 10 ml. more of H 2S0 4 (sp.gr. 1.84) to
the PbS0 4 in the original flask, and boil for several minutes. Cool, add slowly
30 ml. of water, and again heat to boiling for a few minutes; allow the solution
to cool to about 60° C. and completely transfer the PbS0 4 to the Gooch crucible.
Wash with lead acid, retaining the filtrate and these washings for the copper
determina,tions. Remove the beaker containing these solutions and wash out
the lead acid with dilute alcohol; set the Gooch crucible inside a porcelain
crucible; dry and ignite for five minutes at the full heat of a Tirrill burner; cool
and weigh as PbS0 4, which contains 68.33% lead.
specified for the analysis of the alloy._ . Each milliliter is equivalent to ap-
proximately 0.00500 g. of antimony or 0:00465 g.lof iron.
Ferrous Ammonium Sulfate.-Dissolve 12.4 g. of ferrous ammonium sul-
fate crystals in 950 mL of distilled water. and add 50 mL of H 2 S0 4 (sp.gr. 1.84).
Stare h.-To 1000 mL of boiling water add ~ cold suspension of 6 g. of
starch in 100 mL of distilled water; cool, add a ~ew drops of chloroform and
mix thoroughly. .
(a) Determination of Arsenic
Method.-In a 300-mL Florence flask, dissolve 0.5 g. of the alloy in exactly
8 mr. of H 2S0 4 (sp.gr. 1.84). The metal must be finely divided, preferably in
the form of thin foil-like shavings, although sawings or very fine drillings may
be used. Heat the solution to boiling; cool, add about 5 mL of water and a
bulk of about 0.5 mL of clean granulated pumice stone and boil the covered
solution very gently for about five minutes, or until the strong odor of SO! can
no longer be detected. Finally, being careful to have 5 mL of water present,
cool the solution to about 18° C. and cautiously add 20 mL of HCI (sp.gr. 1.18).
Insert in the flask a rubber stopper fitted with a thermometer and delivery tube
and connect the delivery tube with the" is "condenser. Heat the solution to
gentle boiling for from 10 to 15 minutes, keeping the vapor temp_erature at
107° C. for at least 5 minutes. Wash out the condenser into a 300-mL beaker
and add an excess of about 2 g. of NaHCO a• Bring the volume of the solution
to about 200 mr.; warm to about 27° ~C., add 5 mL of starch solution and titrate
with standard iodine solution to the appearance of a deep blue color.
N OTE.-A blank determination should be made on corresponding amounts of reagents
treated as above, and the results should be cO,rrected accordingly. .
If the sample does not contain 14% (or 0.07 g.) of antimony, add enough
dissolved SbCl a to the solution to bring the contents of antimony up to about
0.07 g. for the purpose of insuring perfect 'reduction of the tin and uniform
conditions for the titration. Then add exactly 6 mL of H 2S0 4 (sp.gr. 1.84)
and 60 ml. of HCl (sp.gr. 1.18) and add about!6 in. of clean, soft No. 14 pure
iron wire cut into 2 in. lengths and cleaned with dilute HCI just before using.
Boil gently for 30 minutes, add about 6 in. more wire and boil 30 minutes longer.
Remove from heat, close flask loosely with a rubber stopper and allow about
two minutes for all air to be expelled by hydrogen and acid vapors. Close
flask tightly and quickly place it in cold wa1&rj"'cooling the solution to about
20° C. Transfer the solution quickly to a' 500-roL beaker (leaving the iron
ANALYSIS OF NON-FERROUS ALLOYS 1387
wires in the flask) and rinse flask and contents with 150 mI. of cold, recently-
boiled distilled water. Add rinsings to main solution in the beaker, bring
final volume to about 300 to 350 mI., add 5 ml. of starch solution and titrate
quickly with standard iodine solution ~o the appearance of a strong blue color.
NOTES.-A blank determination should be made on corresponding amounts of
reagents treated as above, and the results should be corrected accordingly.
If the copper content of the alloy is as high as about 3% or over, it prevents a good
end-point in the tin titration when the tin percentage is low. Under such conditions,
just before titrating, add about 1 g. of KI crystals, stir until nearly dissolved and titrate
immediately. If allowed to stand too long at this point, trouble may be encountered
due to the precipitation of some of the other constituents.
Other forms of pure iron may be employed, but the wire as above specified has been
found convenient.
For accurate work it is desirable to reduce and titrate the solution of tin in an
atmosphere of CO 2•
Throw the filter paper and 'precipitate back into the original beaker, add
5 ml. of HNO. (sp.gr. 1.42), heat until the paper,is pulped and the sulfide dis-
solved, add 40 ml. of lead acid and evaporate to fumes of H 2S0 4 • Just before
the solution begins to fume it will become dark colored due to the action of the
H 2S0 4 on the filter paper. At this point add a; \:lrop or two of HNO. (sp.gr.
1.42) to the hot solution and continue to fume. Repeat the addition of HNO.
until the solution shows no brown coloration even upon strong fuming. Cool,
add a few mI. of water and take to fumes a second time. Again add a few IllI. of
water and fume. Finally cool, add 35 mI. of water, bring to boiling, cool, and
allow to stand several hours. Filter and wash with lead acid.
Dissolve the PbS0 4 with 30 mI. of hot ammonium acetate solution, wash
with hot water, dilute to a volume of 150 mI., bring to boiling, add 10 ml. of a
. saturated solution of K 2 Cr 207 and allow to stand in a warm place several
hours. Filter on a weighed porcelain gooch crucible and wash several times
with hot water and once with alcohol. Dry at 110° C. for one hour. Cool
and weigh as PbCr04. Multiply by 0.6375 for lead. \
Accuracy.-Duplicate determinations should check within 0.008% of lead.
N oTE.-Ir: p~eparing the Gooch crucible it is advisable to fit a small filter pape~ in
the bottom of the Gooch crucible before adding the asbestos and treat exactly as in the
determination.
DETERMINATION OF BISMUTH
NOTES.-1. When a relatively large amount of a lead base alloy or a tin base alloy
containing considerable lead is dissolved in a HCl solution, lead chloride will separate out
before the alloy is dissolved, if insufficient solution is present, and retard further solvent
action. For the solution of alloys containing lead and sj.lbsequent procedure see
"Determination of Anti'mony, Zinc and Copper in Solder."
2. If a precipitate of H 2SnO. forms when the-sutfides are dissolved in HNO., anti-
mony may be present in solution, and precipitate out with the BiOCI. In that case,
ANALYSIS OF NON-FERROUS ALLOYS 1389
filter the BiOCI on a paper, dissolve in 10 ml. of hot HCI (1 : 1), wash, dilute to 100 mI.,
pass H 2S, filter, wash, treat the sulfides, with 10 ml. of the KSH stock solution mixed
with an equal quantity of water, filter and wash. Proceed as described above.
3. See Note under "Determination of Lead."
DETERMINATION OF IRON
receptacle containing the mercury as ()a~hode, and e~ectrolyze over night using a
current of 0.5 ampere. Transfer the electrolytelto a 400-ml. beaker, wash the
mercury with water, combine electrolyte and washings, add 5 g. of NH 4 CI,
make just neutral with NH 40H using methyl rep' as an indicator, boil for two
minutes, filter and wash with hot water. Ignite in a weighed porcelain crucible
at first over a low, then over a strong Bunsen flatp_e. Cool and weigh as Al 20 a•
Multiply by 0.529 for aluminum. .
Accuracy.-Duplicate determinations ,should check within 0.003% of
aluminum.
NOTEs.-l. See Note 1 under "Determination of Bismuth."
2. Phosphorus, if present, will interfere. For its correction see "Determination of
Aluininum, Analysis of Manganese Bronze," Method (a). \
DETERMINATION OF NICKEL
A. General Method
Solutions Required. Mixed Acid.-Mix 400 ml. of HCI (sp.gr. 1.18) and
500 ml. of water and then add 100 ml. of'HNO a (1.42).
Potassium Hydroxide Solution.-Dissolve 100 g. of KOH in 500 ml. of
water. i
Copper Solution.-Dissolve 2.0 g. of CuS04'5H 20 in 500 ml. of water.
KSH Stock Solution.-See "Determination of Lead, etc., in Tin Base
Bearing Metal."
I
Methods of the National Lead Company Research Laboratories.
16
See "Specifications for Solder Metal," 1936-'BOok of A. S. T. M. Standards, part I,
17
p.695.
ANALYSIS OF NON-FERROUS ALLOYS 1391
DETERMINATION OF COPPER
B. DETERMINATION OF MAGNESIUM
Method.-Weigh 10 g. of the sawing'l.into a 400-ml. beaker, add 100 ml. of
water and 20 mL of H 2S0 4 (sp.gr. 1.84). (Add the acid in small doses.) When
decomposi~ion is complete, filter and wash with hot water. Discard the
residue. Evaporate the filtrate to a volume of about 75 ml. and proceed as in
" A " for the electrolytic separation of aluminum and zinc.
After separating most if not all of the zinc, transfer the electrolyte and
washings to a 600-ml. beaker, add about 5 g. of tartaric acid (in solution), 5 g.
of NH 4Cl and 10 ml. of a saturated solution of microcosmic salt. Make the
solution alkaline with NH 40H (sp.gr. 0.9) adding sufficient NH 4 0H to have
present 15 ml. for every 100 ml. of solution. Stir vigorously and allow to stand
over night. Filter and wash with cold dilute NH 4 0H (1 : 9), ignite and weigh
as Mg 2Pi>7. Multiply by 0.2184 for Mg. Run in duplicate.
1396 ANALYSIS OF NON-FERROUS ALLOYS
NOTE.-Shouid the solution after standing over night appear clear or should trans-
parent,crystals form, the magnesium has hot' been coIJtpletely precipitated. A vigorous
stirring at this point should completely precipitate the magnesium. Allow to stand
several hours or preferably over night and proceed as .outlined above.
\
C. DETERMINATION OF CADMIUM AND IRON
Weigh 30 g. of the1sample into a 600-ml. beaker, add 100 ml. of water'and
34 ml. of H 2S0 4 (sp.gr. 1.84). (Add the acid in small doses.) When decom-
position is complete, add 10 g. of Na 2S04, dilute to a volume of 350 ml. with
water, warm and pass H 2S for 30 minutes. Allow to stand in a warm place for
2 hours and again pass H 28. Filter and wash. ,Reserve the filtrate (Filtrate
No.1).
Throw the paper containing the precipitate back into the original beaker,
add 10 ml. of HNO a (sp.gr. 1'.42) and 10 ml. of H 2S0 4 (sp.gr. 1.84) and eva-
porate to fumes of SOa. Clear up any discoloration due to carbonizing of the
paper by adding a few drops of HNO a (sp.gr. 1.42) to the hot solution. Cool,
add 5 ml. of water, bring to boiling, cool, allow to stand an hour or so, filter and
wash with cold H 2S0 4 (1 : 9). Discard the residue.
To the filtrate add 5 m!. of HNO a (sp.gr. 1.42), dilute to a volume of 150 ml
with water and electrolyze for the removal of copper as described in " A."
Remove and wash the electrodes, combine electrolyte and washings in a 600-ml.
beaker and evaporate to fumes of :;;Oa. Cool, add 7 g. of Na 2S04, dilute to a
volume of 300 ml. with water, warm and pass H 2S as before. Allow to stand
as before. Again pass H 2S for 30 minutes, filter and wash with H 2 S water.
Reserve the filtrate (Filtrate No.2).
Dissolve the precipitate by pouring 10 ml. of hot HNO a (1 : 1) through the
paper, wash with hot water, add 5 ml. of H 2S0 4 (sp.gr. 1.84) and ev~porate to
strong fumes of SOa. Cool, add a few ml. of water and again evaporate to
fumes. Repeat the treatment with water and subsequent fuming to remove the
last traces of nitric acid. Finally cool, take up with 50 ml. of water, bring to
boiling, cool, neutralize with NH 40H (1 : 2) and then make acid with H 2S0 4
(1 : 9) having 3 ml. in. excess. Dilute to ~a volume of 100 ml. with water and
electrolyze over night using a current of ,0.5 affi_i)eres. Remove the cathode,
rinse in water and then in 95% alcohol, dry at 110° C. for 10 minutes, cool and
weigh as Cd. Run in duplicate. ,
Combine the reserved filtrates No.1 and No. 2,jJoil to expel H 2S, oxidize
the iron present by adding 10 m!. of H 2 0 2 to the boiling §olution, continue
boiling to expel H 20 2, cool and add a slight excess of a saturated solution of
KMn04 to insure removal of the H 2 0 2 • Pass through a Jones reductor and
titrate with N/20 KMn04 solution for irqn, Run a blank through the reductor
and correct for same. Run in duplicate.
NOTE.-Weigh out small pieces of metal. Sawings, drillings or' filings are likely to
be contaminated with iron from the tool used, to take the sample .
. .
D. DETERMINATION OF LEAD AND TIN
/
Solutions Required. Lead Acid.-See _:"j)etermination of Lead as Sul-
fate, Analysis of Manganese Bronze." . -
ANALYSIS OF NON-FERROUS ALLOYS 1397
. Should the precipitate still hold copper, make a third separation. When
finally the precipitate of Fe(OH)3 is freed of copper, transfer it, together with
the two residues held in reserve, to a 300-ml. cOI!e flask, add 10 ml. of HNOa
(sp.gr. 1.42) and 20 ml. of H 2S0 4 (sp.gr. 1.84) and evaporate to fumes. De-
stroy any organic discolQration with HNO a (sp.gt. 1.42). When the solution
shows no discoloration even upon strong fuming copl, add a few ml. of water and
fUlhe again.~ Repeat with the addition of water and subsequent fuming two or
three times to rid the solution of the last traces of nitric acid. Finally cool,
add 100 ml. of water and 20 ml. of HCl (sp.gr. 1.18), reduce with iron, and
proceed as described in "Determination of Tin in Battery Plate Metal."
Titrate with approximately N /20 KMn04 solution. Carry along a blank ,and
correct for same. Run in duplicate.
NOTE.-In general, the Pb0 2 results come high. Report the lead figure ob.tained
by the gravimetric method.
E. DETERMINATION OF MANGANESE
Solutions Required.-See " Determination of Manganese by the Persulfate
Method, Analysis of Manganese Bronze,;"
Method.-Weigh 10 g. of the drillings into a 400-ml. beaker, add 80 ml. of
the" solution for dissolving" and warm until completely in solution. Transfer
to a 125-ml. graduated flask, cool to'room temperature, make up to the mark
with water, mix thoroughly and measure off 50 ml. of the solution into.a 300-ml.
Erlenmeyer flask. Boil until the oxides of. nitrogen are expelled, dilute to
100 ml. with boiling hot water, add 20 ml. of AgNO a solution and 30 ml. of
(NH 4)2S208 solution and let stand over a hole on the steam bath until a full
permanganate color has developed. Cool to below 25° C. and immediately
titrate with standard arsenite solution to the disappearance of the pink color.
:Run in duplicate. 0
NOTES.-l. One g. of the metal requires 2.75 ml.jof the "solution for dissolving"
for solution. .
2. Should the aliquoted portion contain morlJ than 0.0015 g. of Mn, a precipitate of
hydrated Mn02 will form. In that case, meas1lre off a smaller portion, add sufficient
"acid mixture" to bring the acid mixture content up to 24 ml. and proceed as described
above. . I •
. 3. If the solution is allowed to stand too long a time, the color will fade. Warming
for 15 minutes should convert all the Mn to permanganic aCid. ~ ~
4. Owing to the rigid specifications called for in this class of alloys and the tediousness
of chemical analysis for impurities, especially tin, spectrographic methods are preferable.
See "Tentative Method of Test for Quantitative Spectrochemical Analysis of Zinc
Alloy Die Castings for Minor Constituents and Impurities," 1935 Book of A. S. T. M.
Tentative Standards, p. 1488. " '
ANALYSIS OF NON-FERROUS ALLOYS 1399
FUSIBLE METALS
Fusible metals are mixtures of lead and tin to which has been added bis-
muth, cadmium or both to lower the melting point. They may contain in
addition small amounts of impurities, such as copper, arsenic, antimony,iron
and zinc.
ANALYSIS OF WOOD'S ALLOY 23
(ALLOY OF LEAD, TIN, BISMUTH AND CADMIUM) ....
DETERMINATION OF LEAD
DETERMINATION OF TIN
DETERMINATION OF BISMUTH
wash with hot dilute HN0 3 (1 : 49). Discard the residue. Combine the fil-
trate with the original filtrate from the H 2Sn03 beparation:
Boil the combined filtrates down to a volume of 100 mI., make just
neutral with NH 40H (1 : 3) as shown by a piec~ of litmus paper and then add
5 mI. of HCI (1 : 9) in excess. Dilute to a volume of 400 mI. with boiling hot
. water and allow to stand in a warm place for seyeral hours. Filter and wash
with hot water. Reserve the filtrate. .
Dissolve the precipitate from the filter with hot HCI (1 : 1) and ~ash
alternately, first with hot water and then with hot HCI (1 : 1). Evaporate
the solution down to a volume of about 1 ·mI., dilute. to a volume of 400 mI.
with boiling hot water and allow to stand as before. Filter on a weighed por-
celain Gooch crucible and wash with hot water. Transfer the filtrate to a clean
beaker and reserve. Finally wash the precipitate on the Gooch once or twice
'with 95% alcohol, dry at 110° C. for 30 minutes, cool and weigh as BiOCI.
Multiply by 0.8024 for bismuth.
NOTEs.-l. During the solution of the alloy, lead nitrate may separate out and
precipitated H.Sn03 may coat undissolved alloy and prevent further solution. In such
a case, dilute with sufficient hot water to dissolve lead nitrate and detach the H.Sn03
coating from undissolved alloy with a glass rod. .
2. The fusion of the H.SnO. is best conducted by placing the small porcelain crucible
inside a larger one. Cover both crucibles. Set the covers on the crucibles in an inverted
position, first breaking off the small ring handles of the covers before placing them on
the crucibles. /
3. Start the fusion at a low heat and gradually bring the temperature up. If
started at too high a temperature the melt will froth over the top.
4. If the melt is leached with an excessive amount of water, the sulfides, especially
iron sulfide, will not separate but will remain in solution in a colloidal state. In such a
case, twaporate to a volume of about 50 ml., add about 20 ml. of KSH stock solution,
stir, and allow to stand in a warm (not hot) place until the supernatant solution is of a
clear yellow color. The addition of a few grams of NH.Cl accompanied with vigorous
stirring will accelerate the precipitation of the colloidal sulfides. (Do not boil.)
current of 1.0 ampere, using a platinum spiral for ariode and a platinum
cylinder for cathode. Remove the cathode, rinse in water and in 95% alcohol,
dry at 110° C. for 10 minutes, cool and weigh as cadmium plus copper. '
Dissolve the deposit from the cylinder with 5 ml. of HNO a (sp.gr. 1.42)
and wash the electrode free from cadmium with water. Combine the elec-
trolyte and washings and evaporate to a small volume. Wash into a small
porcelain crucible and evaporate to dryness. Take up with 1 m!. of HNO a
(1 : 1) and proceed as in " Determination of Copper in Solder."
Having determined the amount of copper present deduct from the weight
of copper plus cadmium. The difference is cadmium.
DETERMINATION OF ANTIMONY
NOTE.-Care must be exercised in rendering the alkaline sulfide solution acid with
Hel. If the acid is not added in small doses, especially when the solution is hot, the
r~pid evolution of H 2S will make the solution froth over the top of the beaker.
DETERMINATIN OF IRON
DETERMINATION OF ZINC
DETERMINATION OF ARSENIC
SILVER ALLOYS
ANALYSIS OF SILVER SOLDERS 2,4
DETERMINATION OF COPPER
I .
Method.-Transfer the filtrate from the AgCI separatIon to a 400-m!'
beaker, add 5 ml. of H 2S0 4 (sp.gr. 1.84), evaporate to fumes of H 2S0 4 , cool,
24 The methods given under" Analysis of Silver Solders" are 'A, S. T. M. methods,
excepting those for cadmium and zinc. See A. S. T. M, Methods of Chemical Analysis
of Metals, p, 193, December, 1936, .. -- ' .
25 See Specifications of Silver Solders"· 1936 Book'6fA. s. T. M. Standards, Part I,
p.69I.
ANALYSIS OF NO~-FERROUS ALLOYS 1403
dilute with 50 m!. of water, bring to boiling, transfer to a 200-m!. lipless beaker;
add 3 m!. of HNO a (sp.gr. 1.42), cool, dilute to 150 m!. with water and proceed
as in " Determination of Copper, Analysis of Manganese" Bronze."
DETERMINATION OF CADMIUM
DETERMINATION OF ZINC 26
DETERMINATION OF LEAD
DETERMINATION OF IRON
a distinct color persists, bring to boiling, pass through the reductor and proceed
as in " Determination of Iron, Analysis of Manganese Bronze."
,
DETERMINATION OF NICKEL
a 600-m!. beaker, add 25 ml. of a 20% solution of tartaric acid (enough to hold
all the iron and chromium in solution), make alkltline withNH 4 0H and then
make acid with 15 ml. of HCI (sp.gr. 1.18) in excess. Add 20 to 30 ml. of the
sodium dimethylglyoxime solution, neutralize with INH 40H and then make ivst
acid with the dilute acetic acid. Stir the solution vigorously and allow to stand
at a temperature of from 75 to 90° C., with occasio,ll'al stirring, for 30 minutes.
Filter on a weighed Gooch crucible, wash with hot water, dry at 110° C. to
constant weight, cool and weigh as nickel dimethylglyoxime. Multiply by
0.2032 for nickel. ,
If greater accuracy is desired, use a larger aliquot portion'with proportionally
larger amounts of reagents and filter the nickel dimethylglyoxime precipitate on
a 15 cm. paper and wash 18 or 20 times with hot water. Dissolve the precipi-
tate off the paper with boiling dilute HNO a (1 : 3) and wash the filter thor-
oughly with hot water. Add 10 ml. of dilute H 2S0 4 (1 : 1) and boil the liquid
gently until strong fumes of SOa are evolved. Cool, add 10 ml. of HNO a
(sp.gr. 1.42) and repeat the evaporation. Rinse the cover and 'sides of beaker
with a fine jet of water and fume the solution again to insure the expulsion of
every trace of HNO a. Cool, add approximately 50 m!. o'f cold water and boil
the contents of the beaker for several minutes, which should result in a per-
fectly clear solution. .,
Make the solution alkaline with 25 m!. of NH 40H (sp.gr. 0.9) in excess,
insert electrodes, and electrolyze with a current of from 1 to 2 amperes until
the solution has become 'colorless. Continue the electrolysis for at least 15
minutes longer, and then test the solution by adding 1 or 2 drops of it to a
solution of potassium thiocarbonate. A pink color indicates the presence of
nickeI. Remove the cathode quickly, rinse it in water and then in absolute
alcohol, dry for a few minutes at a temperature of 80 to 100° C., cool and
weigh as metallic nickel.
NOTE.-If 'an alcoholic solution of dimethylglyoxime was used to precipitate the
nickel, the contents of the beaker should be digested at a temperature just short of the
boiling point until the odor of alcohol can no longer be ,detected before filtering.
DETERMINATION OF 'CHROMIUM
drops of KMn04 solution (2.5%). Boil for 5 minutes or longer, add 20 ml. of
NaCI solution (10%) and boil for 10 minutes after the permanganic acid or
any Mn02 formed has dissolved completely, in order to make sure that all
chlorine has been expelled. Cool to room temperature and dilute with cold
water to 400 ml. Treat with 2 ml. of H 3 P0 4 (sp.gr. 1.72), add from a burette
ferrous ammonium sulfate solution until the solution turns green and then
add 5 ml. in excess. Titrate this excess back with 0.1 N' potassium per-
manganate.
In a clean flask, place 50 ml. of the ferrous ammonium sulfate, add 10 ml.
of H 2S0 4 (1 : 1) and titrate with the standard KMn04 solution. From this
titration calculate the number of ml. of KMn04 solution necessary to titrate ~
the amount of ferrous ammonium sulfate solution add~d to the sample. The
difference between the number of ml. of KMn04 solution calculated and the
number used for the titration of the sample represents the chromium present
and the percentage is found by multiplying this volume by 100 times the
chromium titre of the KMn04 solution.
NOTE.-The strength of the ferrous ammonium sulfate may change from day to day.
Its strength in terms of KMn04 solution should be ascertained each time the determina-
tion is made.
DETERMINATION OF IRON
DETERMINATION OF MANGANESE
Solutions Required.-See " Determination of Chromium."
Method.-Decompose 1. g. of the sample with 20 ml. of perchloric acid
(60%) exactly as outlin;ed in the determination of chromium. Add 200 ml. of
29 Blair, A. A., "The Chemical Analysis of Iron," 8th Ed., p. 248.
1408 ANALYSIS OF NON-FERROVS ALLOYS
warm water, an excess of about 10 m!. of NH 40H (sp.gr. 0.9) and 5 g. of am-
monium persulfate. Boil for 5 minutes, filter on twp superimposed 11 cm. filter
papers, supported on a Buchner funnel, gentle suction being employed. Wash
the precipitate thoroughly with hot water. Dissolve the precipitate of Fe(OH)a
and Mn02 in. 75 m!. of hot HNO a (sp.gr. 1.135) ~nd a .few drops of H 2SO a.
Wash the filter thoroughly with hot water. Repeat the precipItation with
NH 40H and ammonium persulfate, and the filtration. Dissolve the pre-
cipitate in 50 m!. of hot HNO a (sp.gr. 1.135) and a few drops of H 2SO a, and
wash the filter with HNO a of the same strength. Transfer to a 300-m!. Erlen-
meyer fl!j.sk and boil for several minutes to expel oxides of nitrogen. Trea~
cautiously with 0.5 g. of sodium bismuthate and heat to boiling for 1 or 2
minutes. Clear with a sufficient excess of H 2SO a and boil for 2 or 3 minutes
longer. Cool to 15° C., add 1 to 2 g. of sodium bismuthate, shake vigorously
for 1 minute, dilute with an equal volume of cold water and filter at once on an
ignited asbestos or alundum filter. Wash with ce>ld HNO a (3%). Add from a
burette 20 to 25 m!. (depending on the amount of permanganic acid) of ferrous
ammonium sulfate ,solution and titrate the excess with standard KMn04
solution.
To ascertain the amount of manganese present proceed exactly as in " De-
termination of Chromium."
DETERMINATION OF CARBON
Transfer the paper and precipitate to a small platinum crucible, dry, and
0
ignite, at first at a dull red heat and finally at 1050 to 1100 C. Cool and weigh.
Add 2 drops of H 2S0 4 (1 : 1) and several ml. of HF and evaporate until all the
H 2S04 has been expelled. Ignite the crucible again at 1050 to 1100 0 C., cool
and weigh.
The difference between the first and second weights multiplied by 100 and
divided by the weight of sample taken gives the percentage of silicon in the
sample.
DETERMINATION OF SULFUR
DETERMINATION OF COBALT
DETERMINATION OF IRON
Dissolve the precipitate obtained from the separation of the cobalt with
hot HCI (1 : 1), add NH 4 0H (sp.gr. 0.9) until a permanent precipitate forms,
boil, allow to settle, filter and wash with dilute ammonia and hot water until all
NH 4 Cl is remov~d. Proceed as in " Determination of Iron, Analysis of Man-
ganese Bronze."
NOTE.-RUIi a blank on chemicals used.
DETERMINATION OF CHROMIUM
DETERMINATION OF MOLYBDENUM
necessary steps to remove carbonaceous matter and the last traces of nitrio
acid. (See" Determination of Lead, Bismuth, e1)c. in Tin Base Bearing
MetaL") Cool, add 75 ml. of water and boil to a clear solution. Partially
reduce the molybdenum with 5 g. of granulated \zinc' and proceed as in
" Determination of Molybdenum in Wulfenite or Molybdenite" in chapter
under Molybdenum. II I
DETERMINATION OF MANGANESE 33
DETERMINATION OF SILICON
Method.-Proceed as in " Determination of Sili~on, Analysis of Nichrome."
33 See "Ford-Williams Method for Manganese" in chapter under Manganese; also,
Proc. A. S. T, M., Part I, p. 598, 1929.
ANALYSIS OF NON-FERROUS ALLOYS 1413
A. DETERMINATION OF SILVER
Weigh 100 g. of the sample into a 3-in. scorifier and heat in a muffle furnace
until the assay" covers." Pour into an iron mold and allow to cool. Free the
resulting lead button from PbO, scorify again and pour as before. The button
last obtained should not weigh over 20 g. and can be cupelled directly. Weigh
the silver button obtained upon an assay balance.
NOTE.-If the amount of silver is large, the button should be examined for gold, in the
usual manner.
B. DETERMINATION OF BISMUTH
Solutions Required. Sodium Carbonate.-Dissolve 100 g. Na 2 C0 3 in a
liter of distilled water.
Ammonium Carbonate.-Make a half saturated solution.
Alkaline Sulfide Wash Solution.-Dissolve 200 g. of KOH in a liter of
distilled water and mix one part of this solution with 4 parts of H 2S water.
Method. (a) Determinationfor Ordinary Amounts of Bismuth.-Dissolve
20 g. of the sample in a 400-ml. beaker with 100 ml. of HNO a (1 : 4), with the
aid of heat. When solution is complete, add dilute ammonia (1 : 2) with
constant stirring, drop by drop from a burette, until a faint opalescence appears.
If an actual precipitate is formed, redissolve by the addition of a small amount
of HNO a (1 : 4) and repeat the addition of ammonia. Now add 5 ml. of lICI
(1 : 9), fill the beaker with hot water, bring to boiling, and allow to stand on a
steam bath for two hours. The assay, while standing, must not reach the
boiling temperature. Filter through a 7-cm. paper. Transfer the precipitate
completely to the paper by means' of a "policeman" and wash twice with
hot water. Carefully examine the filtrate, washings, and any decanted liquid
and reject if clear. Dissolve the precipitate by dropping around the edges,
from a 5-ml. pipette, 5 ml. of boiling HCI (1 : 9), receiving the sulution in the
original beaker. Wash the paper thoroughly with hot water, fill the beaker
with water, bring to boiling, and allow to stand as before. Filter the bismuth
oxychloride upon a weighed Gooch crucible, wash thoroughly with water, once
with alcohol, once wIth ether, and dry upon the hot plate. Cool and weigh.
NOTE.-If time permits it is convenient to allow the assay to stand over night. In
that case, the precipitate of bismuth oxychloride generally settles so completely that the
clear supernatant solution can be decanted.
34 This method is reproduced through the courtesy of the American Society for
Testing Materials, 260 S. Broad St. Philadelphia, Pa.
35 These methods are issued under the fixed designation B 35; the final number indi-
cates the year of original adoption as standard or, in the case of revision, the year of
last revision. Issued as Tentative, 1919; Adopted, 1920; Revised, 1924; Reverted
Tentative, 1936. See A. S. T. M. Methods of Chemical Analysis of Metals, p. 174,
December, 1936.
1414 ANALYSIS OF NON-FERROUS ALLOYS
\
(b) Determination for Amounts of _B_ismuth Smaller than Can be Deter-
mined by (a).-Dissolve 100 g. of the sample in 500 ml. of dilute HNO g (1 : 4).
When solution is complete, allow to cool and add N a 2 CO g solution little by
little until a slight permanent precipitate has for,med. Then add 50 ml. of
the Na 2 CO g solution, bring to boiling, allow to stand warm until the super-
natant liquor is clear again, filter and reject the iJiltrate. _Dissolve the pre-
cipitate without washing by slo,vly pouring hot HNO g (1 : 4) around the edges
of the filter paper, using no more acid than necessary. Wash the paper once
with hot 'water and determine bismuth in the filtrate as described in (a).
NOTE:-When the sample contains a small amount of bismuth, it is often difficult to
ascertain when the correct amount of ammonia has been added to the nitrate solution.
In this case, place a small piece of litmus paper in the solution, and add the ammonia very
gradually until the litmus p,aper indicates a neutral reaction.
(c) Determination of Bismuth in Samples Containing Appreciable Amounts
of Tin and Antimony.-Use a piece of litmus paper as in Note under (b), and
after the first precipitation of bismuth oxychloride has been filtered off and
washed (see (a» proceed as follows: Dissolve the bismuth by dropping around
the edges of the paper 10 ml. of boiling HCI (1 : 2), receiving the solution in
the original beaker. Discard the paper after washing. Dilute the solution'
to about 200 ml. with fresh H 2 S water and then pass H 2S gas through the hot
solution for 15 minutes. Filter and wash with hot water. Remove any tin
or antimony present by washing threfYtimes with alkaline sulfide wash solution.
Wash the precipitate again with hot water, place it, together with the filter
paper, in a 100-ml. beaker, add 20 ml. of HNO a (1 : 4), boil until sulfides are
completely dissolved and the paper well pulped. Filter the solution, receiving
the filtrate in the original beaker, and wash well. Determine bismuth in the
original beaker, and wash well. Determine bismuth in the filtrate as in (a).
NOTEs.-When the sample cc;mtains comparatively large amounts of tin or antimony,1
the residue left after the solution of the lead in the nitric acid obscures the opalescence
found upon the addition of ammonia. _ I
If the original sample contains more than 0.25 per gent of bismuth, it is preferable
to use only a IO--g. charge. :
C. DETERMINATION' OF ARSENIC
Method.__f_Dissolve 111.11 g. of the sample in 550 m!. of HNO g (1 : 4).
When solution is complete wash into a graduated liteJ': flask, add 75 ml: of
H 2S0 4 (1 : 1), cool, and make up to the mark with water. -Transfer to a large
beaker, rinsing out the flask with 25 ml. of water. Mix thoroughly, allow to
settle and filter off 900 mI., equivalent to a 100-g. charge. Evaporate in a large
porcelain dish until only enough H 2S0 4 is left to moisten the residue. When
cool, wash into a small distilling flask with 60 ml. of HCI (sp.gr. 1.18) and 20
ml. of water, cleaning the dish carefully. Add.IO'g. of ferrous sulfate and distill,
boiling to as small a volume as possible. When cool add 50 ml. more of HCI
(sp.gr. 1.18) and redistill. Pass H 2S gas through the cold distillate for 45
minutes. Filter, and weigh the As 2S g on a Gooch crucible, washing with cold
water, alcohql and CS 2 • After drying and weighing, redis~olve with (NH 4)2COg
solution and reweigh the Gooch crucible, calculatiD-g-the
. _,. .loss in weight to arsenic .
NOTE.-The 25 ml. of water added is equivalent to the volume of precipitated lead
sulfate.
ANALYSIS OF NON-FERROUS ALLOYS 1415
and wash with H 2 S water slightly acidified. (N ote 5.) Separate tin, antimony
and arsenic in the precipitate with KOH solution as usual, obtaining an alkaline
sulfide solution No.3. (Note 6.)
To sum up, three alkaline sulfide solutions hare been obtained, containing
tin, antimony and arsenic, a precipitate of metallic sulfides containing copper,
lead, etc., and a solution containing iron, zinc, nickel, etc. .
NOTES.-l. A residue indicates the presence of antimony, tin, possibly arsenic, or
sulfur as lead sulfate.
2. It has been found that even this treatment occasionally fails to dissolve stannic
acid completely. .
3. If it is preferred, allow the solution to stand over night to insure the complete
solution of all soluble salts.
4. Any residue of lead sulfate may contain some tin, antimony, or possibly arsenic.
5. The filtrate will contain any iron, zinc, nickel, cobalt and manganese; while in
the precipitate will . be found any copper, cadmium, lead, silver, bismuth, tin, antimony
and arsenic.
6. When separating the sulfides of arsenic, antimony and tin from sulfides of copper,
lead, etc" it is necessary to wash all the sulfides back into the beaker in which they were
precipitated. These sulfides sometimes cling so tenaciously to the paper that in dis-
lodging them more water than the 25 m!. specified is required. In this case allow the
sulfides to settle and then decant the clear supernatant liquor through the filter until
the volume is reduced to 25 mI. Before rejecting the decanted fluid always test with
H 2Swater. '
In washing sulfide precipitates with water, much trouble is experienced from the
tendency of the precipitate to pass through the filter in .the colloidal form. This is
particularly true in washing sulfides tl{at have been ,digested with KOH. Time and
trouble will be saved by washing all the sulfides precipitated from mineral acid solutions
with H 2S water containing a little of the acid in which they were precipitated, The same
is true of sulfides precipitated in or filtered from an alkaline solution. ,
When working with alkaline solutions in which tin is to be determined, avoid the
use of Jena or other glass that contains zinc. The zinc content of the glass may influence
the result. '
and boil the solution until about 30 m!. are left. Expel the arsenic as chloride.
Now oxidize the solution with a pinch of KClO s and boil until no more chlorine
remains. Cool and add 5 m!. of potassium iodide solution. Titrate the liber-
ated iodine with 0.1 N sodium thiosulfate solution, using carbon disulfide as an
indicator.
DETERMINATION OF IRON
Method.-Evaporate the filtrate containing iron, zinc, etc., to 100 m!. and
oxidize with a few drops of HNO a• Separate the iron with ammonia as usual,
making two separations, and'receive the :fiitrate in a 500-ml. Erlenmeyer flask.
Redissolve the iron hydroxide with hot HCl (1 : 1) or dilute H 2S0 4 and deter-
mine the iron volumetrically by any of ,the standard methods.
/
DETERMINATION OF ZINC
Solutions Required. Nitric Acid (1 : 4).-Mix 200 ml. of HNO a (sp;gr.
1.42) with SOO m!. of distilled water.
Sulfuric Acid (1 : l).-Carefully pour,. with stirring, 500 m!. of' H 2S0 4
(sp.gr. 1.84) into 500 m!. of distilled water.
Acidulated Hydrogen Sulfide Water.-Add 20 m!. of HCl (sp.gr. LIS) to
1000 m!. of distilled water and saturate with hydr'ogen sulfide.
Ammonium Thiocyanate Solution (2%}.-Dfssolve 20 g. of NH 4 C1\fS in
1000 ml. of distilled water. ;
Hydrochloric Acid (1 : 3).-Mix 100 m!. of HCI (sp.gr. US) and 300 m!.
of distilled water.
Standard Zinc Solution (0.1 mg. of zinc per ml.):;:-Dissolve exactly 0.1 g.
of U. S. Bureau of Standards pure zinc in 5 ml. of HCI {sp;gr._1.1S) and dilute
to exactly 1000 m!. with distilled water.
Potassium Ferrocyanide Solution.-Dissolve 34.S g. of -K 4 Fe(CN)6' 3H 20
in 1000 ml. of distilled water.
Method.-Dissolve 222.23 g. of the sample in 1100 m!. of HNO a (1 : 4),
using a 1300-m!. beaker. When the lead is disilblved, transfer the solution to a
2000-m!. graduated flask and add slowly 150 ml. of H 2S0 4 (1 : 1). Cool, fill
the flask to the mark and then pour the solution into a clean 3000-ml. flask
prbvided with a rubber stopp~r. Rinse tHe measuring flask with exactly 50 ml.
of water, which is equivalent to the volume of lead s~lfate which is present.
Mix the solution thoroughly by shaking, allow__the-precipitate to settle and filter
through a dry filter until 1800 mI. of filtrate has been obtained.
ANALYSIS OF NON-FERROUS ALLOYS 1419
INTRODUCTION
The former division of the subje.ct of steel analysis into the. fields of plain
carbon steels and alloy steels is no longer followed"in the better treatises in this
field. Due to the cooperation of the National Bureau of Standards and the ~
various industrial laboratories and societies for the standardization of methods
of testing, the referee methods in this field are now well standardized, and for
this reason the subject of steels and ferroalloys is now covered by reproducing
the methods of the American Society for Testing Materials. There are in-
cluded in this chapter some selected methods for various elements, under the
caption" Other Methods" (p. 1470). Those who desire a still fuller selection
of methods should consult one of the more extended treatises in this field. 2
The present selection should serve for all practical purposes.
PREFACE
The following methods have been compiled as standard procedures for use
in referee analyses. These methods, however, are not intended to preclude
the use of other methods (or apparatus) that give results within the permissible
tolerances. In any case, the analyst should check the method and technique
1 Methods from the literature; arranged by N. H. Furman.
2 Lundell, Hoffman and Bright, Chemical Analysis of Iron and Steel, J. Wiley &
Sons, Inc., 1932. New York.- C. M. Johnson, Chemical Analysis of Special Steels,
John Wiley & Sons, Inc., New York, 1930: Methods of the Chemists of the U. S. Steel
Corporation, published by the Carnegie Steel Co., Pittsburgh, Pa. W. W. Scott,
Technical Methods of Metallurgical Analysis, D. Van Nostrand Co., New York, 1923.
F. Ibbotson, The Chemical Analysis of Steel-Works Materials, Longmans Green & Co.,
New York, 1920.
3 Under the standardization procedure of the Society, these methods are under the
jurisdiction of the A. S. T. M. Committee E-3 on Chemical Analysis of Metals. Repro-
duced with the permission of the American Society for Testing Materials, 260 S. Broad St.
Philadelphia, Pa.
4 This is a Tentative Standard and under the Regulations of the Society is subject to
annual revision. Suggestions for revision should be addressed to the Headquarters of
the Society, 260 S. Broad St., Philadelphia, Pa. Issued, 1936. Accepted for publication
as tentative by Committee E-lO on Standards, August 26, 1936.
1421
1422 CHEMICAL ANALYSIS OF lRON AND STEEL
/'
CHEMICAL ANALYSIS OF IRON AND STEEL 1423
and shaH be free from scale, suriace II\etal, grease, dirt, or other foreign sub-
stances. If the samples are taken by drilling, a drill not less than ~ nor more
than %: in. in diameter shall be used. Samples shall be uniform, thoroughly
mixed, and free from dust. Chips too coarse to pass a No. 20 sieve aTe not
recommended, nor shall long 'curly urillings which will not pack closely for the
carbon determination be used.
In referring samples to other analysts for check analyses, ipieces of the
original full-size section, when possible, should be submitted rather than
cuttings, unless the latter are specifically requested. ,
Location of Samples: Large Sections.-For large sections, iJlcluding blooms,
billets, slabs, rounds, squares, shapes, etc., samples shall be taken at any point
midway between the outside and the center of the piece by drilling paranel to
the axis. In cases where this is not practicable, the piece shall be drilled on the
side, see Fig. 183 (b) and (c), but the drilling. shall not be collected until they
represent the portion midway between the outside and the center. The
" tension test specimen may be used for sampling if it conforms to the abo've
conditions. ~ ..
Bored Forgings.-For bored forgings, samples shall be taken midway
between the inner and outer surface of the wall.
Thin Af aterial.-For thin material or material of small cross-section, such
as plates, shapes, bars, etc., if the method described above for large sections
is not applicable, the sample shall be taken by machining off the entire cross-
section, or, if this is not possible, by drilling entirely through the material at a
point midway between the outside and the center, See Fig. 183 (a).
Sheets Rolled Longitudinally.-For sheets rolled from slabs or bars longi-
tudinally, the specimen for sampling shall be cut 2 in. in width and' across the
full width of the sheet as rolled. The specimen shall be cleaned by pickling
or grinding and then folded once or more by bringing the ends together and
closing the bend. The sample for analysis shall be taken in the middle of this
length by milling the inside sheared edges or drilling entirely through from the
nat surface. Sampling by milling is preferable. For sheets of a light gage
more than one specimen may be taken and stacked together before folding.
Sheets Rolled Transversely.-For sheets rolled from slabs or bars transversely,
the specimen shall be cut from the side of the sheet, half way between the middle
and end as rolled, 2 in. in width and 18 in. in length. If the sheet is No .. 20
gage (24 oz. per sq. ft. or 0.037 in. in thickness) or lighter, the specim~n shall
be cut from the full length .of the sheet as rolle'd. The specimen selected shall
be cleaned by pickling or grinding and then folded. once or more by bringing
the ends together and closing the bend. The s~mple for analysis shall be taken
in the middle of this length by milling the inside sheared edges or drilling en-
tirely through from the fiat surface. Sampling by milling is preferable.
Sheets Not of the Full Size, Rolled.-Sheets cut from larger sheets alld not
of the full size rolled shall be sampled by milling or drilling the sheet in a
sufficient number of places so that the flam pIe is representative of the entire
sheet. The sampling may be facilitated by folding the sheet both ways. .
(b) Pig !ron.-In the absence of a special agreement between the manu-
facturer and the purchaser, or whenever it is necessary to sample pig iron in the
solid state, the sampling shall be carried out in the following manner: Not less
than 3 pigs shall be taken to represent any lot or shipment and for lots of more
than 30 tons one· pig shall be taken for each 10 tons of iron. The pigs shall be
selected by some means such as the knotted rope system, that eliminates the
element of personal choice; and 3 to 7 pigs taken in tile order selected, shall
constitute a unit sample. For boat o'r barge shipments arrangements between
the seller and the purchaser may be made, if desired, in regard to taking fewer
pigs per ton. If for any reason it is desirable to fracture the pigs, they may be
broken in two, and one portion of each of the original pigs reserved for the
sample for analysis. Any loose sand or other deleterious matter on the reserved
portions shall be removed, conveniently by brushing. .
The sample for analysis shall be collected after the proper discard, by drilling
each pig or portion of pig with a properly sharpened and hardened, fiat beaded,
%- or ;!i-in. highspeed tool-steel drill in a direction at right angles to the long
axis of the pig and at such a feed as to form a minimum amount of fine drillings.
The first drillings shall be discarded and only the drillings collected after the
drill is cutting into clean metal shall be reserved for the sample. The drilling
shall extend to within U in. of the opposite surface of the pig; and suitable
precautions shall be taken to collect all the ,drillings, fine as well as coarse
particles, and to avoid contamination of the sample in any way. Any pigs
or portions of pigs found too hard to drill readily may be annealed by heating
to a dark red color and cooling in air.
The sample for analysis shall be cOIlj.posed of equal portions by weight of
the drillings from the pigs forming the unit saml).le, and the weight of the
combined sample shall not be less than 75 g. The drillings.shall be thoroughly
mixed, as.. by gently grinding in a suitable mortar until all pass the No. 20
(840-micron) sieve. 5 Unless accurate determinations of combined and graphitic
carbon are required, the sample shall be carefully divided into two or more
portions by pouring in a cone-shaped pile on glazed paper and halviIlg or
quartering in the usual manner. These portions, of which one shall be reserved
for the manufacturer and one for the pllrchaser, shall be placed in a suitable
container, fully labeled for identification, and dispatched to the laboratories
without unnecessary delay and with full instructions as' to the determinations
required.
6 See Standard Specifications for Sieves for-Testing Purposes (E 11-26), 1936 Book
of A. S.;. M. Standards, Part II, p. 1413. Reproduced onpp. 1329-1333 of this Volume.
CHEMICAL ANALYSIS OF IRON AND STEEL 1425
When combined. and graphitic carbon must be determined with the highest
degree of accuracy, equal portions by weigJ:tt of the drillings from the pigs or
parts of pigs representing a unit sample shall be mixed to form a sample of at
least 100 g. which shall then be separated into coarse and fines by sieving as
described under (c), the sampling of gray-iron castings. •
(c) Gray-Iron Castings.-In accordance with the Standard Specifications
for Gray-Iron Castings (A. S. T. l'4. Designation: A 48) of the American
Society for Testing Materials,6 three test bars, 1.20 in. in diameter shall be cast
in sand from each heat after one quarter and three quarters of the heat has been
poured. .
One bar from each set having been broken, one end of each next to the
fracture shall be thoroughly cleaned !1nd the outer skin removed for a sufficient
distance from the fracture and down to clean metal. Chips then shall be
taken by means of a lathe or milling machine ll-cross the whole face of the bar
until not less than 100 g. has been collected. The same amount shall be
taken from each bar. The bar shall be so clamped as to' permit the attach-
ment. or use of any suitable device for collecting every part of the sample and
the machine shall be run slowly enough to reduce to a minimum the danger of
loss of fine particles (avoid drafts). Horizontal drilling may also be employed.
Preparation of Samples from Castings for Analysis.-The entire gross sample
• shall be weighed and then sifted on a No. 80 (177-micron), and if need be a No.
120 (125-micron), tight-fitting sieve (with cover). 7 Both sieves and cover m,ust
be tight-fitting to avoid loss of fine graphite by dusting. As an alternate pro-
cedure (a) the drillings may be sifted through two sieves of such sizes that not
under 10% nor more than 20% of the entire sample remains on the larger
sieve, and not under 10% and not over 20% shall pass through the smaller
sieve or (b) sieves may be used of such sizes that concordant results may be
obtained.
The two (or three) portions so obtained shall be separately weighed. Each
one shall then be thoroughly mixed without any loss of material and divided
by weight into two (or three) exactly equal portions, each of which shall be
placed in a clean, glass-stoppered bottle or other suitable receptacle and
appropriately labeled. Of the three sets of the subdivided sample, one shall
be placed in a clean, glass-stoppered bottle or other suitable receptacle and
appropriately labeled. Of the three sets of the subdivided sample, one shall
be retained by the manufacturer, one sent to the purchaser and one held in
reserve.
Before weighing the sample for analysis, the contents of each bottle shall
be thoroughly mixed. Carbon is determined on each of the sieved portions,
and the carbon present in the original drillings is calculated from the ratio
of the weights of the separate sieved fractions to the original iron.
As a general rule, the very fine material is richest in carbon, the coarse
material next, while the intermediate sizes are the poorest. For routine
analyses, the aim shoulct be to obtain the iron in as small particles as possible,
as for example by milling. In such case, the samples for graphite and carbon
may be taken- by forming a corie, flattening it out, and taking small dabs with
61936 Book of A. S. T. M. Standards, Part I, p: 457.
7 See Standard Specifications for Sieves for Testing Purposes (E 11-26),'1936 Book of
A. 'S. T. M. Standards, Part II, p. 1413.
1426 CHEMICAL ANALYSIS OF IRON AND STEEL .
a spatula from at least ten, representative portions for each weighed portion.
Less care is needed in taking portions for the p.etermination of the other con-
stituents; grab samples from the container after gentle tumbling are satisfactory.
(d) Malleable Iron.-Proceed as in the sarilpling of pig iron. '
(e) Wrought Iron.-As it is desirable that ~he sample represent the entire
cross-section of the specimen, sampling by millin& is preferable. Specimens
shall be freed frdm any mill scale by pickling, gtinding, or other- suitable means.
It is necessary to secure all of the millings, for the various sized particles will
differ in composition. For routine work, samples for carbon, manganese,
phosphorus, sulfur, and silicon may be taken from the thoroughly mixed.
millings. For the determination of slag and oxides, or for umpire determina-
tions of the other constituents, the entirll sample shall be weighed and then
sifted without dusting on a No. 35 (500-micron) sieve. The two fractions
shall be weighed and in taking samples for analysis, portions of these two well-
mixed fractions must be taken in the same proportion which they bear to the
gross sample.
Nickel boats shall not be made of sheet nickel containing more than 0.10%
of carbon. Before use, new boats and covers shall always be preheated in
oxygen until no more carbon dioxide is given o.ff or until a constant blank is
obtained. In order to p!event injury to thEl tube from spattering, a platinum
or nickel coveJ," open at both ends and allowing free access of oxygen is desirable.
Material for Lining Boats.-" RR Alundum" (AI203) alkali-free, specially
prepared for carbon determinations, and 50-mesh or finer, is entirely satis-
factory for the lining of boats. Ignited low-silica chrome ore, or " Zirkite "
(natural oxide of zirconium) properly sized and freed from materials causing
a blank may be employed. Ferric oxide has been recommended as permitting
combustion at low (800 0 to 900 0 C.) temperatures but is not recommended for
high temperatures. Quartz sand, owing to its liability to fuse or to slag with
oxides of iron, causing bubbles. of gas to be enclosed, is objectionable.
Oxygen.- The purity of the oxygen shall be not less than 99.5%. Organic
matter of any kind is an undesirable impurity. It is usually absent, and it
suffices to pass the oxygen through an absorbent such as soda-asbestos followed
byaI.lhydrone. If carbonaceous matter is suspected, the oxygen shall be passed
through a tube that is loosely packed with asbestos or copper oxide and heated
to about 450 0 C. before it is passed through the soda-asbestos.
Combustion of the Sample. Size of Particles of Sample.-The finer the
chips (excluding dust, which causes low values on a hot boat), the better, except
with alloys that burn too vigorously. Drillings or millings sized between 14 to
60 mesh are satisfactory.
Distribution of the Sample in the- Boai.-In general, the sample shall be
p3:cked in a small groove or furrow that has been made in the beddin'g material
in the boat. If the material burns too rapidly, satisfactory regulation of the
speed of combustion' may sometimes be obtained by spreading the sample
somewhat loosely over the bedding material or by covering with a thin layer
of the bedding material.
Addition of Accelera{ors to the Sample.-Shot tin, Ys-in. pellets, is an
'excellent accelerator, one pellet being placed on each end of the layer of drillings.
Millings of open-hearth iron, or other iron with 0.01 to 0.03% carbon are
satisfactory aids in the combustion of large drillings and refractory alloys,
Red lead, copper oxide, lead and powdered copper are also used. Red lead,
unless previously ignited at 500 0 C. and kept in ground-glass stoppered bottles,
will cause erratic blanks.
Preheating of Sample.-If the sample is allowed to come to the temperature
of the furnace before the oxygen is admitted, it usually bursts into a bright
flame and burns completely. A period of 72 to 1,72 min. of preheating suffices.
Rate of Adding Oxygen.-The rate at which oxygen is admitted is also a •
factor in the velocity of combustion. Assuming the combustion apparatus has
been heated to the temperature range above that recommended,;t is possible,
if the material is closely packed and if oxygen is admitted at too rapid a rate,
that the combustion may be so violent as to cause excessive spattering of fused
oxides and such fluidity of the molten slag that the boat or other container may
be injured or destroyed. Sufficient oxygen, however, shall be run in to insure
a current of gas through the absorber at all stages of the combustion. When tin is
employed as an accelerator, combustion is very rapid and it is therefore neces-
sary to increase the flow of oxygen during combustion. When the sample has
1428 CHEMICAL ANALYSIS OF IRON AND STEEL
burned (1 to 2 min.) the flow of oxygen may be reduced to a rate of 150 mi.
per min. . I ."
Purification of the Carbon Dioxide.-The purifiers that foll~w the combus-
tion tube must remove finely divid~d oxides of ir9n, oxides of sulfur or selenium,
dry the gases before they enter the weighed absorber, and protec~ the absorber
from outside effects. Iron oxide is screened out ~uring the passage of the gases
through the liquids or columns of solids that precede the weighed absorber.
The small amount of 80 2 that is given off from low-sulfur steels can be satis-
A B C D E F G H ·1
(a) Apparatus used at the National Bureau of Standards.
A. Oxygen tank with needle or reducing valve.
< B. Preheating furnace, 450 0 C, with platinized asbestos.
C. Ascarite tower. /
D. Mercury manometer.
E. Combustion tube and electric furnace, 1000---11500 C.
F. Special absorption tube which contains asbestos in the left arm, and in the right
arm H 2S0 4 saturated with H 2Cr04.
G. Tower with 20-30 mesh Anhydrone. .
H. Weighed tower with Ascarite and Anhydrone.
I. H 2S0 4 guard tube and gage.
02;nno r· ~n __
~E:::~
rrn
I 2 3 4 . 5 I 6 1 8 9 10
(b) Apparatus used in a representative steel·works laboratory.
1. Mercury valve.
2. Bottle containing concentrated H 2S0 4.
3. Tower, with gooseneck top, containing CaC!..
4. Tower with stick NaOH.
5. Electric furnace with combustion tube and adapters.
6. Bottle with aqueous solution of KMn04 (5%) inserted only for steels with more
than 0.05% sulfur.
7. Tower with granulated zinc.
S. Bottle with concentrated H 2S0 4.
9. Tower with P 206.
10. Weighed tower with Ascarite. I
factorily removed by H~S04 that has been saturated with chromic acid. High-
sulfur steels need other absorbents such as an aqueous solution of chromic acid
(50%), a solution of KMn04 (5%), or heated platinized silica gel that will
convert the dioxide to trioxide. The sulfur trioxide that is so formed is not
removed by anyone absorbent, but is condensed and absorbed during its
passage through the liquids or columns of solids in the train.
Absorbing Bulbs.-No special types of absorbing bulbs are recommended,
although the Fleming, Miller, Turner, and the Midvale (sometimes called
Stetzer and Norton) bulbs have proved satisfactory. When filled, the tubes
shall not weigh over 200 g. and they shall always be weighed filled with oxygen
and against a like counterpoise. Open bulbs, such as the Midvale, lose oxygen
by diffusion. They shall be filled with oxygen before weighing when not in
continuous use, and the same time interval shall be held between weighings.
Absorbents.-The most desirable absorbent for carbon dioxide is 20 to 30
mesh soda asbestos (Ascarite) followed by Anhydrone, Mg(CI0 4 )2. The latter
is needed to absorb the water which is formed during the reaction and is not
held. by the unused Ascarite.
General Arrangement of Apparatus.-Figures 184 (a) and (b) show two
typical arrangements of the apparatus. The apparatus and arrangement may
be modified, provided satisfactory results for the carbon determination will be
obtained. .For example, with oxygen free from carbonaceous matter the pre-
heater may be omitted; other forms of absorption towers may also be substi-
tuted for the special absorption tube.
Procedure. Carbon Steel (Sulfur Content less than O.l%).-After having
properly set up and tested the apparatus, spread 1.36 to 2.73-g. of the sample
on the bed material in the boat so that the particles are in intimate contact.
Cover the sample with a suitable cover and introduce the boat into the hot.
combustion tube. Close the tube and allow the sample to heat for 1 to 2 min.,
depending upon the size of the particles. Then admit the oxygen, at a rate
of 300 to 400 ml. per min. while combustion is going on. When combustion
• is complete, slow the current so that the gases leave the absorbents at the rate
of about 150 mL per min. for 6 to 8 min. in order to sweep out the carbon
dioxide. Withdraw the absorption tube filled with oxygen, place it in the
balance case for 10 min.,8 open momentarily, and weigh against a similar tube
used as a counterpoise. The increase in weight represents carbon dioxide.
Multiply this by 27.27 and divide by the weight of the sample to obtain the
percentage of carbon in the steel. Remove the boat from the tube and examine
the melt for evidences of incomplete combustion. If the drillings are not
thoroughly fused in a solid pig, the determination shall be rejected. The
proper correction for a blank run on every material used in the determination,
except the sample, shall be determined with special care in umpire analyses,
or in analyses of very low-carbon material.
High-Sulfur Steels.-Proceed as with carbon steels of low-sulfur content
but insert a special 80 2 oxidant in the train such as (1) a tube of platinized
silica gel heated to 440 0 C. and followed by a tube containing ironized asbestos
and anhydrone, (2) a tube containing about 25 m!. of an aqueous solution of
8 The tube will warm up when CO 2 is absorbed. It is not necessary to wait until it
reaches room temperature if it is in continuous use, provided the same time interval is
maintained, and approximately the same amount of CO 2 is absorbed.
1430 CHEMICAL ANALYSIS OF IRON
i
AND' STEEL
asbestos (finer than lO-mesh) that has been digested with HCI and then
thoroughly washed with water.
Potassium Hydroxide Solution (sp.gr. 1.10, approximately).-Dissolve 120
g. of stick KOH in 1000 ml. of water.
Diluted Hydrochloric Acid (1 : ZO).-Mix 50 m!. of HCI and 1000 m!. of
water.
Procedure.-Dissolve 1 to 3 g. of the sample (see Sampling) in 50 m!. of
HNOa (sp.gr. 1.20) and heat gently on the steam bath, with occasional stirring,
until action ceases. If silicic acid separates add 5 m!. of HF to facilitate filtra-
tion. Collect the residue, conveniently by suction on ignited asbestos contained,
in a fused-silica Gooch crucible of such diameter as will fit in the combustion
tube. Wash the residue thoroughly with hot water, then with a hot solution
of KOH (sp.gr. 1.10), hot water, then with diluted, HCI (1 : 20), and finally
with hot water. Dry at a temperature not exceeding 1500 C., and determine
the graphite by direct combustion at approximately 900 0 C. in the apparatus
used for the determination of total carbon, taking care to close the combustion
tube'immedjately after inserting the sample.
less 'of the KMn04 solution.) Heat to 55° to '600 C. and complete the titration
by adding permanganate until a faint pink coldr persists for 30 sec. Add th6'
last 0.5 to 1 ml. dropwise, with particular care. to allow each drop to become
decolorized before the next is introduced. Determine the excess of per-
manganate required to impart a pink color to1the solution by matching the
color in another beaker containing the sam~ a~ount of the specially treated
diluted H 2S0 4 (5 : 95) at 55 to 60° C.
Standard Sodium Arsenite Solution (0.035 N, approximately).-Dissolve
2 g. of sodium arsenite in water, filter if the solution is not clear, and dilute to
1 liter. The solution is preferably standardized on a standard steel of ap-
proximately the same manganese content as the sample. A clear greenish
yellow color which does not change when another drop of arsenite is added is
a satisfactory end point. As an alternate standardization procedure, 20-ml.
portions of the standard solution of KMn04 (0.03 N).may be titrated in a solu-
tion, containing 1 g. of iron as ferric nitrate, under the same condition of acidity
and dilution as in the above method. .The theoretical titer of the arseilite
solution cannot be used because manganese is not quantitatively r~duced to
the bivalent stage.
Procedure. Carbon Steel.- Disso~ve 1 g. of the sample in 50 ml. of diluted
HNO a (sp.gr. 1.135) and boil to expel oxides of nitrogen. Remove from the
heat, cool somewhat, add about 0.5 g. of sodium bismuthate, and boil for 2 to
3 min. (With high-carbon steels, about 1 g. of bismuthate, or an amount
sufficient to effect either a pink color or a precipitate of Mn02, should be added.)
Clear the solution of permanganic acid or manganese dioxide present by adding a
few drops of a saturated solution of S02 (freshly prepared) or other suitable
reducing agent free from manganese and chlorides, and boil to expel oxides
of nitrogen and sulfur. Cool to 15° C. or lower, add 0.5 g. of sodium bis-
muthate (or an amount equal to 26 times the weight _of manganese present)
and agitate for 1 min. Add 50 ml. of HNO a (3% and free from nitrous acid)
and filter through asbestos (or 'through glass frit or alundum crucibles, suffi-
ciently tight to prevent passage of bism~thati particles). Wash with HN0 3
(3%) until the washings run through colorless. The filtrate must be clear and
absolutely free from particles of bismuthate. , Titrate by either method (a)
or (b) as follows:
(a) Ferrous Suljate-Permanganate Titration.-Treat the filtrate with 2 ml. -
of phosphoric acid (85%) and then from a burette.J!.dd enough of the FeS04
solution to completely discharge (1 to 2 ml. excess) the color of the permanganic
acid. Titrate with the permanganate solution to the appearance of a faint
pink color. In exactly the same manner carry through a blank determination
using the same amounts of acid and bismuthate as was done with the sample.
Finally, add the exact volume of ferrous ammonium sulfate solution which
was employed and titrate with the standard ,KMn04 solution. The difference
between the volumes of KMn04 solution, required in the two titrations repre-
sents the manganese in the sample.
(b) Arsenite Titration.-After filtering off the bismuthate, immediately
titrate with the arsenite solution to a clear greenish/yellow color which does
not change upon addition of another drop of._.:,._arsenite. .
Alloy Steels.-Steels containing nicker;' molybdenum, and less than 1%
of chromium shall be treated like carbon steels, although it is preferable in the
CHEMICAL ANALYSIS OF. IRON AND STEEL 1433
co
Wash the paper a few more times withhot water, and then complete the deter-
mination as with carbon steels, including the preliminary oxidation and boiling
with bismuthate. ,
OPen-Hearth Iron and Wrought Iron.-Proc'eed as with carbon steels using
2 to 3 g. of the sample. It is advisable, befor~ filtering the analysis sample,
to treat the asbestos pad with a weak solution ofi KMn04 and wash it free from
the latter with diluted HN0 3 (3%). Very low' manganese irons shall not be
filtered through pads used previously for higher manganese alloys. Colori-
metric determinations by the periodate method on 0.5- to I-g. samples of open-
hearth iron are applicable for routine work.
Jl].anganic acid with arsenite is, like the visual, empirical. It is, therefore,
necessary to standardize the arsenite solution against a like weight of a standard
steel which has approximately the same manganese content as the sample
being analyzed and has been treated exactly as the latter.
While the potentiometric titration offers no particular advantage over the
visual method for steels containing less than 1% of manganese, it is claimed
that more reproducible end points can be obtained potentiometrically with steels
of high manganese content or in turbid solutions when graphite has not been
separated. The apparatus consists essentially of three parts, a motor stirrer,
electrodes, and a potentiometric unit. In the Larrabee apparatus the elec-
trodes consist of gold and platinum rods with small vanes at the bottom. In'
the Kelley apparatus platinum and calomel electrodes are used; it is desirable
to insert a salt bridge (N a 2S04) solution between the calomel electrode and
the titrating solution in this apparatus.
Alloy Steels: Nickel, Chromium-Nickel, Chromium-Vanadium, and
Chromium-Vanadium-Molybdenum Steels.-Proceed as with carbon steels.
C}1romium Steels (over 4% chromium).-Separate the greater part of the
chromium by the bicarbonate separation,ll using a 6 ml. excess for high chro-
mium (18%) steels. To the filtrate add 30 ml. of the arsenite acid mixture,
evaporate to 125 ml. and complete the 'determination as directed for plain
carbon steels. With high carbon (0.8%) high chromium steel, test the bicar-
bonate precipitate for manganese.n
Chromium-Tungsten Steel.-Treat 0.5-g. of the sample with 50 ml. of
diluted H 2S0 4 (1 : 9) and 3 m!. of H 3 P0 4 • Heat until all action ceases. Add
40 mt of water and 5 m!. of HN0 3 , boil until the carbides have dissolved, and
proceed as with carbon steels. With chromium-tungsten steels the arsenite
solution is preferably standardized on a steel of similar type of known manganese
content. Tungsten steels with high carbon (0.8%) and high chromium (4.0%)
~ontents may not be completely decomposed by the H 2S0 4-H aP0 4 attack. In
this case treat the sample with 10 ml. of a mixture of equal parts of HCI and
HNO a and 4 to 5 drops of HF. When action ceases add 5 m!. of HCI0 4
(60%) and evaporate just to dryness. Cool, add 5 mI. of H 2S0 4 and fume
until HCI0 4 is expelled. Cool, add 50 mI. of water, 5 m!. of H 3 P0 4 (sp.gr;
1.73), dilute to 100 mI. and heat until salts dissolve. If Mn02 appears, dissolve
it with a few drops of H 2S0 3 • Complete the determination as with carbon
steels. ,
Cast lron.-Dissolve as directed for carbon steels, dilute, filter through
a rapid paper, wash with hot water, dilute to 125 mI., and complete as with
carbon steels.
Open-Hearth lron.-Proceed ,as with carbon steels, but use a 2- to 3-g.
sample.
Wrought lron.-Treat 1 g. of the sample as directed for carbon steels.
Diluted Nitric Acid (2 : 100).-Mix 20 ml. of' HNO a and 1000 ml. of water.
Potassium Permanganate (2.5%).-Dissolve 25 g. of KMn04 in 1000 ml.
of water.
Ammonium Molybdate.-Place in an 800-m!. besker 65 g. of c.p. ammonium
molybdate crystals, (NH4)6Mo7024·4H20, 225 g. of\NH 4NO a, 15 ml. of NH 40H
(sp.gr. 0.90), and 600 ml. of water. Stir and heat igently. When the crystals
have dissolved, filter (without washing) and dilute to 1000 ml. with water.
Magnesia Mixture.-Dissolve 65 g. of MgS04 and 75 g: of (NH 4)2S04 in
500 ml. of water. Add NH 40H in slight excess and let stand overnight; filter
if a precipitate appears. Make just acid by adding a very sligl}t excess of
H 2S0 4 , dilute to 1 liter, and keep in a glass-stoppered bottle. -
Diluteil Ammonium Hydroxide (1 : 20).-Mix 100 ml. of NH 40H (sp.gr.
0.90) and 2000 m!. of water.
Potassium Nitrate (1%).-Dissolve 10 g. of KNO a in 1000 m!. of water.
Procedure. Carbon Steels.-Transfer 4 g. of the sample 12 to a,300-ml.
Erlenmeyer flask, dissolve the sample in 125 ml. of diluted HNOa (sp.gr. 1.135),
oxidize the organic matter by adding 10 ml. of the solution of KMn04 (2.5%)
and boil for 3 to 5 min. If no precipitat,e forms, add more KMn04 and boil
again. Dissolve the precipitate by the dropwise addition of a saturated solu-
tion of Na 2SOa or H 2SO a, and boil for a few minutes to expel oxides of nitrogen.
Cool to about 75° C. and add 85 mr. of molybdate reagent. Stopper the flask,
shake for 10 min., and allow to stand for 2 hr. or more. Filter through a tight
filter paper. Wash the flask, precipitate, and paper with the cold diluted
HNb a (2 : 100) and then five times with KNO a solution.
Set the £iltrate and washings aside after thorough mixing, and note whether
further separation of phosphomolybdate occ~rs. Dissolve the precipitate on
the filter and in the flask in 20 ml. of diluted NH 40H (1 : 1) to which has been
added 2 g. of citric acid, catch the solution in a 250-ml. beaker, wash the filter
several times with diluted NH 40H (1 : 20), then with hot water, and finally
several times with diluted HCI (1 : 20). If the ammoniacal solution 9f the
phosphomolybdate is not clear at this point, heat it to boiling, filter through
the same paper, wash the paper with hot watef, ignite in a small platinum
crucible, and fuse any residue with as little Na 2 CO a as possible (not over 0.5 g.);
extract the cooled melt with hot water, cooi, filter, and add the water solution
to the ammoniacal solution.
Render the ammoniacal solution acid \vith HCI, add 20 ml. of the acid
magnesia: mixture, cdol in ice water, and then add N'H 40B;_slowly until the
solution is faintly ammoniacal. Stir vigorously for 5 min. or until a crystalline
precipitate appears,1a and finally add 5 to 10 ml. of NH 40H. The volume of
the solution at this point should not exceE;ld 100 ml. Allow the ,solution to
stand in a cool place for 4 to 6 hr., or, preferably, overnight. Filter and wash
the precipitate moderately with diluted NH 4qH (1 : 20). D_issolve the pre-
cipitate on the filter in 20 ml. of diluted HCI (1 : 1), catching the solution in
the original beaker. Wash the filter thoroughly with warm diluted HCI
12 If the percentage of phosphorus is less than 0,05%, it is advantageous to treat two
or more 4-g. samples as described, and combine the yellow precipitates either by filtering
through the same paper or by' combining the precipitates after filtering on separate papers.
13 If'the amount of phosphorus is very small,·jp"may take 30 min. to 1 hr. before the
precipitate begins to appear.
CHEMICAL ANALYSIS OF IRON AND STEEL )437
(1 : 1). Add 0.5 to 1 g. of NH4Br14 and gently boil the HCI solution to It
volume of 5 to 10 ml. (but not to dryness) to eliminate arsenic. Dilute to a
volume of 50 to 75 mI., add 0.1 to 0.2 g. of citric acid, 2 to 3 mi. of the acid
Illagnesia mixture, and cool in ice water. Make ammoniacal, stir as above,
and allow to stand in a cool place fpr 4 to 24 hr.
Filter, wash with diluted NH 40H (1 : 20), and ignite the precipitate care-
fully and at as low a temperature as possible until the carbon has been destroyed
and the residue is white. Finally, ignite to constant weight at 1000 to 1100° C.
Dissolve the ignited precipitate in 5 mi. of diluted HNO a (1 : 1) and 20 ml. of
water. If no residue remains, the ignited precipitate may be regarded as
Mg2P207. If a residu~ remains, filter, wash with hot water, ignite, and weigh.
Treat with a few drops of HF, evaporate to dryness, ignite, and reweigh.
Subtract any loss of weight from the original weight of the pyrophosphate, and
calculate the percentage of phosphorus on the basis of this purified magnesium
pyrophosphate.
Alloy.Steels: Nickel, Chromium-Nickel, Stainless and Similar Alloy Steels
Contai[ling Neither Tungsten nor Vanadium.-Proceed as with carbon steels.
High-chromium steels (stainless), high-chromium high-nickel (18% chromium,
8% nickel; 20% chromium, 20% nickel), and other steels in this group that do
not dissolve in diluted HNO a (sp.gr. 1.135), shall be treated as follows: Treat
3 g. of the sample in a 500-ml. Erlenmeyer flask with 75 ml. of a mixture Qf
equal parts of HCI and HNO a, and heat gently. When decomposition is
complete, add 25 ml. of HCI0 4 (60%) and evaporate until fumes of HCI0 4 are
evolved. Continue the heating for 5 min. to oxidize chromium 'and to dehy-
drate silica. Cool somewhat, add 40 ml. of water, filter, and wash the flask,
paper, and silica with 55 ml. of diluted HNO a (sp.gr. 1.2). Add KMn04 to
the filtrate, l;lOil, add H 2SO a to destroy oxides of manganese and to reduce all
of the chromium, and proceed as usual. As an alternative method, the sample
m~y be decomposed by heating with diluted HNO a (1 : 1) and adding small
portions of either HF or HCI. •
Austenitic Manganese Steels (over 10% manganese).-Transfer 4 g. of the
sample to a 300-ml. Erlenmeyer flask, add 70 ml. of diluted HNO a (sp.gr. 1.135)
and digest until action ceases. Add 35 ml. of HCI0 4 (60%) and evaporate
just to fumes. Add HF dropwise until all of the silica is in solution and then
an excess of 5 drops. Fume so that the HCI0 4 refluxes on the sides of the flask
for 25 to 30 min. Cool, add 50 mL of water and 10 mi. of HNO a• Add a few
drops of KMn04 (2%) and boil until oxides of manganese are precipitated.
Dissolve the precipitated manganese with H 2SO a and boil a few minutes to
expel oxides of nitrogen. Cool to 70° C. and precipitate with molybdate, etc.,
as with carbon steels.
Chromium- Vanadium Steels or Other Stee~s Containing Vanadium but No
Tungsten.-Proceed as with carbon steels until the solution is ready for the
addition of the molybdate reagent. At this point 'cool to 10° C., add 5 ml. of
a solution of ferrous sulfate (100 g. of .FeSO.j· 7H 20, in 1000 m!. of diluted
H 2S0 4 (5 : 95» and 2 to 3 drops of H 2SO a• Mix, add 85 ml. of molybdate,
shake for 10 min., and allow to stand for 4 hrs. or, preferably, overnight. Filter,
and finish the determination as directed for carbon steel~.
14 With a precipitation temperature of 25° C., arsenic, when present in small amounts.
is not precipitated, and therefore the HBr treatment can be omitted.
1438 CHEMICAL ANALYSIS OF IRON AND STEEL
Stopper and let stand for 24 hrs. Decant the liquid, and standardize it against
a steel of known phosphorus content, as determined by the molybdate-magnesia
method. Appropriate National Bureau of Standards' standard steels are
recommended for this purpose. The solution may also be standardized by use
of the National Bureau of Standards' standard sample 84 of acid potassium
phthalate and the ratio 23 NaOH to 1 phosphorus. Protect the alkali solution
from carbon dioxide by means of a soda-lime or soda-asbestos tube.
Standard Nitric Acid.-Mix 7 m!. of clear HNO a with 1 liter of water and
standardize against the standard alkaF solution, using phenolphthalein as the
indicator. If desired, the acid solution may be rendered equivalent to the
NaOH by dilution with water.
Phenolphthalein Indicator.-Dissolve 0.29 g. of reagent in 100 m!. of ethyl
alcohol (50%). .
Procedure. Carbon Steels.-Transfer 2 to 3 g. of the sample to a 300-m!.
Erlenmeyer flask, dissolve 17 the sample in 65 m!. of diluted HNO a (sp.gr. 1.135)
and oxidize organic matter by adding 10 m!. of the KMn04 (2.5%) and boiling
for ~ to 3 min. If no precipitate forms, add more 'KMn04 and boil again.
Dissolve the precipitate by the dropwise addition of a saturated solution of
Na2SOa or H 2SO a and boil for a few minutes to expel oxides of nitrogen. Adjust
the volume to 60 m!. and the temperature to 45° C., and add 50 m!. of the
molybdate reagent. Stopper the flask, shake for 10 min., and allow the pre-
cipitate to settle. Filter through a 9-cm. close-texture paper. Wash the
flask, precipitate, and paper twice with 5-m!. portions of diluted HNO a (2 : 100)
and then five times with 5-m!. portions of a solution of KNO a (1 %). Finally,
wash the paper about ten times (until free from acid), directing the jet.of KNO a·
solution around the edge of the 'paper and then spirally down. Return the
paper and precipitate to the flask, add 25 m!. of water and a 2- to 5-ml: excess
of standard NaOH, both free from carbon dioxide, and shake or stir until the
'precipitate is dissolved. Dilute to about 150 ml. with water free from carbon
dioxide, add 3 drops of the phenolphthalein indicator, and discharge the pink
color with standard HN Oa.
Alloy Steels: Nickel, Chromium-Nickel, Stainless, and Similar Alloy Steels
Containing Neither Tungsten nor Vanadium.-Proceed as with carbon steels.
With high-chromium nickel, or other steels insoluble in diluted HNO a, treat
2 g. of the sample with 50 m!. of a mixture of equal parts of HNO a and HCI,
add 4 to 5 drops of HF, heat till action ceases and then add 20 m!. of HCI0 4.
Evaporate to fumes, fume 5 to 10 min., filter, etc. (as in the molybdate-
magnesia method)/8 and finish by the alkalimetric method. With steels of
appreciable titanium content (over 0.25% titanium) it is best to remove ti-
tanium before precipitating the phosphomolybdate.
Austenitic Manganese Steel (greater than 10% manganese).-Transfer
2 g. of the sample to a 400-m!. beaker, add 35 m!. pf diluted HNO a (sp.gr. 1.135)
and digest until action ceases. Add 20 m!. of HCI0 4 (60%) and evaporate
just to fumes. Add HF dropwise °until all of the" silica is dissolved and then
add an excess of 5 drops. Fume so that HCI0 4 refluxes on the sides of the
beaker for 20 to 25 min. Cool, add 50 m!. of water and 10 m!. of HNO a. Add
17 To prevent the possibility of silica precipitating in steel of high silicon content,
3 to 4 drops of HF may be added while the sample is dissolving.
18 See p. 1437.
1440 CHEMICAL ANALYSIS OF IRON AND STEEL
a few drops of KMn04 (2%) and boil until oxides of;manganese are precipitated.
Dissolve the precipitated manganese with H 2SO a and boil for a few minutes to
expel oxides of nitrogen. Transfer the solution to a 300-ml. Erlenmeyer flask
and adjust the volume to 60 ml. and the temperature to 45° C.19 Add 50 ml.
of ammoniacal molybdate solution, shake 10 min} and let stand for 20 min.
Filter, wash, and titrate as directed for carbon steels.
Chromium- Vanqdium Steel, or Other Steels Containing Vanadium but No
Tungsten.-Proceed as with carbon steels until the solution is ready for' the
addition of molybdate reagent. At this point cool to 10° C., add 5 m!. of a
solution of ferrous sulfate (100 g. FeS04' 7H 20 in 1000 ml. of diluted H 2S0 4
(5 : 95)) and 2 to 3 drops of H 2SO a. Mix, and then add 50 ml. of molybdate
reagent. Shake for 10 min., allow to settle 1 hr., filter, and finish as directed
for carbon steel.
High-SPeed Steels, or Other Steels Containing Tungsten and Vanadium.-'
For the decomposition of the sample (2 g.) and the preparation of the solution
for the precipitation of the ammonium phosphomolybdate proceed as in the
gravimetric method for· the determination of phosphorus in these steels. 20
Precipitate the phosphorus in a cool reduced solution and finish the determina-
tion as described for chromium-vanadium steels by the alkalimetric method
(see preceding paragraph). .
Cast fron.-Dissolve 0.5 to 2 g. of the sample in 65 ml. of diluted HNO a
, (sp.gr. 1.135). Filter through a paper of loose texture and catch the filtrate in
a 300-ml. Erlenmeyer flask. Wash tne paper a few times with diluted HNO a
(2 : 100) and then with about 50 ml. of hot water.21 Add 10 ml. of KMn04
. to the filtrate and boil for 3 to 5 min. Dissolve the precipitated oxides by the
dropwise' addition of a saturated solution of Na 2SOa or H 2SO a and boil for a
few minutes to expel oxides of nitrogen. Cool to 45° C., add 50 m!. of molyb-
date reagent and proceed as with carbon steels.
OPen-Hearth fron.-Proceed as with carbon steels. The result shall be
corrected by that obtained in a blank run.
Wrought fron.-Proceed as with carbon steelsf using 1 g. of the sample.
In wrought iron phosphorus will be present as phosphide and as phosphate
(in slag inclusions). If a differentiation b~tween the two forms is desired,
determine total phosphorus in one sample; 'determine the phosphate that is
left after phosphine has been driven off (by' attack with a non-oxidizing acid
as HCn from another sample, and the phosphorus occurring as phosphide is
then obtained by difference. The accuracy of the results obtained by this
phosphate-phosphide method is somewhat doubtful, but the 'method is ap-'
parently the only one available at present.
tion, and remove the tumbler. Add 2 m!. .Qf starch solution, then 40 m!. of
diluted HCI (1 : 1) and titrate immediately with th¢ iodate solution to a
permanent blue color.
• NOTEs.-Direct sunlight must be avoided if cadmium sol~tions are used, otherwise
low results are obtained. .
This method succeeds best when the evolution of gas is rapid and the conditions are
kept constant.
Trouble may occur if the absorbing solution becomes too hot. This condition can
be avoided by keeping the absorbing solution in a water bath. •
Some analysts prefer to add the diluted HCI at a temperature of 80° to 90° C. and to
heat the flask on an electric heater instead of over gas. The use of concentrated HCI
necessitates a condensing arrangement and does not offer any general advantages over
the diluted acid for carbon steel.
With most carbon steels, the results for sulfur which are obtained by the
evolution method and the use of the theoretical sulfur titer check the gravi-
metric results within ±0.002%. Some steels dissolve too slowly and some
(for example, certain steels containing high sulfur or high carbon) do not yield
all of their sulfur. In such cases the samples must be annealed as follows:
Place 5 g. of the sample in a 20-ml. porcelain crucible. Cover with a ;!.i-in.
layer of graphite, Acheson No. 38, cover the crucible and heat at 6850 C. for
20 min. Transfer the cooled annealed steel and graphite to the evolution flask
and proceed as in the regular method.
Alloy Steels.-Many alloy steels yield all their sulfur when analyzed by
the same procedure as is used for carbon steels, but some yield only a p!l.rt.
Hydrochloric acid of 2 : 1 strength, or the concentrated acid (sp.gr. 1.19) with
some alloy steels, will give slightly more evolved sulfur than does the 1 : 1 acid.
For example, some molybdenum steels yield most of their sulfur with con-
centrated HCI. For high silicon steels, concentrated HCI plus a small amount
of HF (0.5 m!.) has been recommended. The evolution method, however, can
be used only if experiment has shown that all sulfur is evolved as H 2S in the
particular type of steel under analysis. The evolution method is not satis~
factory for the determination of sulfur in selenium steels.
Cast Iron.-Most cast irons do not give up all their sulfur as hydrogen sul-
fide. The amount that is give:6. off can often be increased by annealing the
sample, although with many irons annealing yields no increase in the amount
of evolved sulfur. Annealing is done as directed in the second preceding
paragraph. ,
With high-silicon cast irons, use hot (70 0 C.) diluted HCr.....(1 : 1), heat •
rapidly to boiling, and then simmer. If the solution froths badly, add about
0.5 m!. of HF.
Concentrated HCI has been recommended for certain alloy cast irons, in
which case suitable condensing arrangements must be provided. Even then·
care must be taken to avoid excessive neutralization 6f the alkaline absorbent.
Open-Hearth Iron and Wrought Iron.-Proceed as with carbon steels.
treat with 100 m!. of diluted H 2S0 4 (1 : 4), and cover.29 Warm gently. When
reaction ceases, remove and rinse the cover, and evaporate the solution, taking
care to avoid spattering, until copious fumes of H 2S0 4 are given off. Replace
the cover and continue the fuming for 2 to 3 min., but no longer since insoluble
sulfates may be formed. Cool somewhat, and add 100 ml. of warm water
(40 0 to 50 0 C.) at one time. Stir until salts are in solution, heating gently if
necessary but never boiling. Filter immediately using a rapid ashless filter
paper for high-silicon steels, and a tighter paper for low-silicon steels. Transfer,
all the residue to the paper, scrubbing the casserole thoroughly with a rubber-
tipped rod. Wash the paper and residue alternately with hot diluted HCI
(5 : 95) and hot water until salts of iron have been removed. This usually
requires six washings with 5-ml. portions of each solution. Transfer the paper
and residue to a platinum crucible, heat carefully until the carbon is gone,
and then cover and ignite for 10 to 15 min. at 10500 to 11000 C. Cool in a
d€siccator and weigh. Add enough diluted ,H 2S0 4 (1 : 1) to moisten the Si0 2
and then 3 to 5 ml. of HF. Evaporate to dryness, carefully heat until H 2S0 4
is gone, and then ignite at 1000 0 C. Cool in a desiccator and weigh. If the
factor w~ight was taken the loss in weight multiplied by 10 represents the
percentage of silicon.
NOTE.-In the 'above and succeeding methods some silica passes into the filtrate.
The amount left in solution depends chiefly on the silicon content of the alloy, that is,
about 0.005% silicon for a 0.1 % silicon steel, approximately 0.02% for a 0.4% silicon
steel and about 0.05% or more for a 4.5% silicon steel. Hence in umpire work it is
necessary to evaporate the filtrate and washings to fumes of H 2S0 4, dilute, filter, and
ignite the washed paper and contents with the first portion.
Alloy Steels.-Proceed as with carbon steels if the sample dissolves in
diluted H 2S0 4, If it does not, use the acid mixture (given for carbon steels)
or the perchloric acid method. so The latter is especially suited for high-
chromium alloys.
Tungsten Steels.-Transfer 2.336 g. of the sample to a beaker or casserole
provided with a cover glass. Add a mixture of 20 ml. of HCl and 20 ml. of
HN0 3, and warm until all the steel has dissolved. Add 60 ml. of warm diluted
H 2S0 4 (1 : 1), and evaporate until copious fumes of H 2S0 4 appear. Cool
somewhat, cautiously add 5 ml. of HCl, and swirl gently until well mixed.
Add 125 ml. of warm water (40 0 to 500 C.), and proceed as in the sulfuric acid
method. Ignite the tungstic acid' and silica together at a temperature of about
10000 C. and weigh. Treat with HF and H 2S0 4 , ignite at 800 0 C., and again
weigh.
Cast Iron.-Proceed a_s with carbon steel, using 2.336-g. of the sample.
With gray or mottled pig iron it is desirable to use a finely divided sample and
to boil vigorously with slightly more acid at the start, as the particles tend to
become coated with gelatinous silica.
29 More rapid and equally satisfactory determinations can be made by dissolving
the sample in 40 ml. of diluted HCI (1 : 1), evaporating to low volume, then adding the
H 2S0 4 and evaporating to fumes, etc. Another alternate procedure consists in dissolving
the sample in 80 m!. of an acid mixture (600 mI. of water, 200 mI. of H 2S0 4 (sp.gr. 1.84),
100 mI. of HNO s (sp.gr. 1.42) and 100 m!. of HCI (sp.gr. 1.19», heating until solution is
complete, and then evaporating to copious fumes of H 2S04, etc. In this method 5 mI.
of Hel is added to the partially cooled residue before dilution with warm water. •
so See p. 1446.
1446 CHEMICAL ANALYSIS OF IROllT AND STEEL
saturate with H 2S as the solution is allowed to cool during the course of 25 min.
Allow the precipitate to settle, filter on paper or paper pulp, and wash a few
times with diluted H 2S0 4 (1 : 99) that· has been saturated with H 2S.
,Transfer the residue and p~per to a 25- or 35-ml. tall-form porcelain crucible,
ignite at a temperature not exceeding 550° C., and fuse with 2 to 4 g. of alkaline
pyrosulfate. Dissolve the cooled melt in 25 ml. of diluted HCl (1 : 9), dilute
the solution tq 100 ml.,·2 neutr!J.lize with a solution of NaOH (5%), and add
approximately 0.3 ml. in excess. Boil for 3 min., digest for about 30 min.,
filter, and wash five or six times with a cool solution of NaOH (0.5%).
Dissolve the precipitate in 15 to 25 ml. of hot diluted HNO. (1 : 3), wash the
paper with hot water, add 5 ml. of H 2S0 4 and evaporate until fumes of H 2S0 4 -
appear. Cool-the solution, dilute to 40 mI., add an excess of several milliliters
of NH 40H, and heat to boiling. Allow the precipitate to settle, filter into a
250-mI. beaker, and wash with hot water. If more than 4 mg. of iron, tin,
chromium, or other elements precipitable by NH 40H, appear to be present,
dissolve the precipitate in 15 to 25 ml. of hot diluted HNO s (1 : 3), and repeat
the precipitation with NH 4 0H. Filter and combine the filtrates. The further
treatment depends on whether the copper is to be determined by electrolysis,
or by precipitating as the sulfide and igniting to the oxide. The former method
is to be preferred in analyses of the highest accuracy, and if over 0.25% of
copper is present.
(b) Determination of Copper by Electrolysis: Apparatus.-Ditect current
must be provided. The most satisfactory electrodes are platinum cylinders
made from 50- to 6O-mesh, sand-blasted gauze, stiffened by doubling the gauze
for about 3 mm. at the top and, bottom and supported by a platinum stem
which is flattened and welded the entire width of the. gauze.
Procedure.-Transfer the ammoniacal solution to a 250-ml. beaker, neu-
tralize with diluted H 2S0 4 (1 : 1), and add an excess of 5 ml. of diluted H 2S0 4
-(1 : 1) and also 4 mL of diluted HNO s (1 : 1). Dilute the solution to 200 ml.
Adjust the 'anode and weighed cathode, cover the beaker with split watch
glasses, and electrolyze with a current of 0.5 amp. until the solution becomes
colorless (about ,2 hrs.). Rinse the cover glasses and the exposed stems of the
electrodes and sides of the beaker. Continue the electrolysis for 30 min. and
test for complete deposition (0.3 to 0.5 ml. of the electrolyte should not give a
brown color with 0.3 to 0.5 fil. of freshly prepared H 2S-water). When deposi-
tion is complete, lower the beaker while washing the cathode with a stream of
water. When free from' electrolyte, detach the cathode, dip it in alcohol,
shake off the excess, and dry for 1 min. at 100° C. Cool in a desiccator and
weigh. The increase in 'weight multiplied by 20 represents the percentage of
copper in the material.3s -
(c) Determination of Copper by Weighing as Oxide. (Copper less than·
O.25%).-Neutralize the ammoniacal solution obtained in (a) with diluted
H 2S0 4 (1 : 1), add 4 ml. of the diluted acid in excess, and adjust the volume to
about 100 ml. Heat to boiling and saturate with H 2S as the solution cools.
32 If an appreciable amount of silica has separated, it may be removed by filtration
before proceeding further.
33 Deposits of copper may be removed by immersing the cathode in diluted RNO,
(1 : 1), rinsing with water, then boiling with fresh RNO s for 5 to 10 min., rinsing with
water and then strongly igniting for 10 to 25 min. over one or two large Meker burners.
1448 CHEMICAL ANALYSIS OF IRON AND STEEL
Allow to settle, filter, transfer all the precipitate to the paper, and wash thor-
oughly with diluted H 2S0 4 (1 : 99) that has been siJ,turated with H 2S. Transfer
the paper and precipitate to a small porcelain or quartz crucible that has been
weighed with cover., He~t ~nder good oxidiz.in~ conditions until carbon has
been destroyed, and then'lgmte to constant weIght at 900 0 to 1000 0 C. Cover,
cool over a good desiccant, and weigh as CuO. If a 5-g. sample' was used, the
weight of CuO multiplied by 16.represents the percentage of' cop'per in the
material.
Carbon Steels.-Use a 10-g. sample and proceed as with copper steels.
Alloy Steels.-Most alloy steels may be analyzed as described for carbon
steels.
Tungsten Stee~s.-Treat 5-g. of the sample with 100 ml. of diluted HCI
(1 : 1). Carefully add 20 ml. of diluted HNO s (1 : 1), and boil gently until the
tungstic acid becomes bright yellow. Dilute the solution to 150 m!. with
boiling water, digest for a few minutes, filter and wash with diluted HCI (1 : 9).
Any copper retained in the tungstic acid is recovered as follows: Treat the
residue with NH 4 0H, add 5 g. of tartaric acid and sufficient H 2S0 4 to have an
excess of 5 m!. per 100 ml. of solution. Treat with H 2S, filter and combine
with the main sulfide precipitate.
Add 15 ml:of H 2S0 4 to the filtrate, evaporate just to fumes of H 2S0 4 , cool,
and add 100 ml. of water. If a small.residue of tungstic acid separates, filter
and wash with diluted H 2S0 4 (5 : 95). Add 5 g. of tartaric acid to the clear
filtrate, neutralize with NH 4 0H, ad.just the acidity to 5% by volume with
H 2S0 4, heat to boiling, and pass a rapid stream of H 2 S into the solution for 10
to 15 min. Let the precipitate settle, filter, wash, and complete the deter-
mination as directed for ({arbon steels.
Cast Iron.-Dissolve 5 g. of the sample in 100 ml. of diluted H 2 S0 4 (1 : 5).
When solution is complete evaporate to fumes of H 2S0 4 , cool somewhat, dilute
to 100 ml. with warm water, heat until salts have dissolved, and filter. Wash,
the residue with hot water, dilute the filtrate to 500 ml. with hot water, and
proceed as with carbon steel. In very accurate work the residue shall be
examined for copper.
Wrought Iron and OPen~Hearth Iron.-Proceed as with copper steels,
using 10-g. of the sample.
(50%) and continue the boiling for 5 to 10 min. or until the precipitate settles
rapidly. Filter immediately, transfer ali the precipitate to the paper, and wash
the paper and. precipitate well with diluted H 2S0 4 (1 : 99) that has been
saturated with H 2S. Place the paper and precipitate in a porcelain or silica
crucible, dry and ignite at a low temperature (520 0 to 5500 C. i conveniently in a
muffle furnace) until all carbon is destroyed. Cool, transfer the contents of the
crucible to a 250-ml. beaker. Add 5 to 6 ml. of diluted HN0 3 (sp.gr. 1.20) 34
to the crucible, warm gently and pour upon the residue in the beaker. Rinse
the crucible with a little water and warm the beaker and contents until the
copper oxide has dissolved. Carefully evaporate the solution to a volume
of 2 to 3 ml. in order to expel most of the acid. 34 Cool, "add 90 ml. of water,35
and add either 10 ml. of a solution of ammonium bifiuoride (20%), or 1 g. of
sodium fluoride (to prevent interference by ferric iron). Then add NH 4 0H
until the solution just reacts alkaline to litmus. Cool the solution to room
temperature. Acidify with acetic acid (80%) and add 1 ml. of the acid in
excess. Add 3 to 4 g. of KI which has been dissolved in a little water, stir well,
and immediately titrate with the standard thiosulfate. When the brown tints
have nearly disappeared, add 5 ml. of starch solution and continue the titration
until one drop changes the color from, blue to yellowish white, and remains
permanent for 15 to 20 sec.
Carbon Steel.-Use 10 g. of the sample and proceed as with copper steel.
Alloy Steels.-Most alloy steels may be analyzed as described for copper
or carbon steels.
Tungsten Steels.-Treat 5 g. of the sample with 75 ml. of diluted HCl
(1 : 1). Carefully add 15 ml. of diluted HN0 3 (1 : 1) and boil gently until
the tungstic acid becomes bright yellow. Dilute the solution to 100 ml. with
boiling water, digest for several minutes, filter, and wash with diluted HCI
(1 : 9). For recovery of copper held by tungstic acid, see the determination of
" copper in tungsten steels by the electrolytic method. 36
Add 10 ml. of H 2S0 4 to the filtrate and evaporate to fumes of sulfuric acid.
Cool, dilute to 200 mI., heat to boiling, add 60 ml. of the thiosulfate solution,
boil for 5 min., and complete the determination as described for copper steels.
Cast Iron.-Dissolve 5 g. of the sample in 100 ml. of diluted H 2S04 (1 : 4).
When solution is complete, evaporate to copious fumes of sulfuric acid. Cool
somewhat, dilute to 100 ml. with warm water, heat until salts have dissolved,
filter, and wash the residue with hot water. Dilute the filtrate to 250 mI., heat
to boiling, add 20 mI. of thiosulfate solution, continue the boiling for 5 min. and
complete the determination as described for copper steels.
Wrought Iron and OPen-Hearth Iron.-Use 10 g. of the sample and proceed
as described for copper steels.
34 Alternately, the residue on the paper may be dissolved in hot diluted HN0 3 (1 : 2)
and the solution caught in the original beaker, and 5 mi. of diluted sulfuric acid (1 : 1)
added. Then evaporate the solution to copious fumes, cool, dilute to 50 mi., treat with
HF, NHDH, etc., and complete the titration with thiosulfate as directed.
35 Since molybdenum interferes partly in this method, it is necessary, if the steel
contains more than 0.25% of molybdenum, to separate the latter from copper by treat-
ment at this point with NaOH as directed in the electrolytic or gravimetric method.
36 See p. 1448.
1450 CHEMICAL ANALYSIS OF IRON AND STEEL.
"
drain, and wash thoroughly with 50 ml. of hot water containing 1 g. of tartaric
acid. Nearly neutralize the absolutely clear solution with NH 4 0H and pre-
cipitate with dimethylglyoxime and NH 40H as before. A glyoxime precipitate
contaminated by cobalt is a darker red than a pure nickel dimethylglyoxime.
(b) Electrolytic Method.-Occasionally in umpire analyses, and with high-
nickel steels the determination is best completed by the electrolytic method
as follows: Filter the nickel dimethylglyoxime precipitate on a 12 or I5-cm ..
paper and wash thoroughly 18 to 20 times with hot water. Dissolve the
precipitate in hot diluted HNO a (1 : 3) and wash the filter thoroughly with hot
water. Add 20 ml. of diluted H 2S0 4 (1 : 1) and evaporate to strong fumes of
sulfuric acid. Cool somewhat, add 10 ml. of HNO., and repeat the evaporation,
to fumes. Rinse the cover and sides of the beaker with water and fume the
solution again to insure the expUlsion of every trace of the HNO a• Cool, add
, 50 ml. of cold water, and heat until salts are dissolved.
Neutralize with NH 4 0H and add an excess of 25 ml. of NH 40H. Dilute
to 175 ml. and electrolyze with a current of from 1 to 2 amp. 'using a weighed
platSnum gauze cathode and a spiral platinum anode. Continue the electrolysis
until the solution has become colorless (5 to 8 hI's.). The solution may be
tested for complete electrolysis by adding one or two drops of it to 1 ml. of
a solution of potassium thiocarbonate. 4o A pink or red color indicates the
presence of nickel. Without interrupting the current, wash the cathode with
cold water as it is withdrawn from the solution. Dip the cathode into a beaker
of water and then into alcohol, dry for a few minutes at 100° C., cool, and weigh.
The increase in weight of the cathode multiplied by 100 and divided by
the weight of the sample taken gives the percentage of nickel in the sample.
In very accurate work, dissolve the deposit in warm HNO a, wash the cathode
with water, then alcohol, dry for a few minutes at 80° C., and reweigh.
High-Chromium High-Nickel Steels (20% chromium, 20% nickel; 18%
chromium, 8% nickel, etc.).-Transfer 0.35 to 0.5 g. of the sample to a 400-ml.
beaker, and treat with 20 ml. of diluted HCl (1 : 1) and 20 ml. of diluted HNO a
(1 : 1). Heat until solution ensues, add 15 ml. of diluted H 2S0 4 (1 : 1), and
evaporate just to light fumes (for low-carbon, low-silicon alloys this latter step
may be omitted). Cool somewhat, and add 100 ml. of water. Warm until salts
dissolve, filter, and proceed as directed for nickel steels, sufficient dimethyl-
glyoxime being added to precipitate all of the nickel (20 to 40 ml. of a 1 %
solution).
Carbon Steels, and Other Steels with less than 0.05% Nickel.-Transfer
5 g. of the sample to a 400-ml. beaker, add 40 ml. of diluted HCl (1 : 1). Heat
until solution ensues and then carefully add 15 ml. of diluted HNO a (1 : 1).
Evaporate to a volume of about 15 ml. and add 50 ml. of diluted HCl (sp.gr.
1.12). ,Transfer. to a 200-ml. separatory funnel, rinsing the beaker with
several 15-ml. portions of the diluted HCl (sp.gr; 1.12). Cool to 10° C., add
120 ml. o(lether, and carefully shake for 1 to 2 min. in a stream of cold water.
Let settle for several minutes and then draw off the lower clear solution into
the original beaker. Gently heat the solution in the beaker to expel the ether
(avoid free flames), add 0.3 g. of KCIO a, boil until the chlorate is decomposed,
40 Prepared by saturating one-half of a solution df KOH (5%) with H.S, adding the
other half, and heating moderately with one twenty-fifth of its volume of es 2 •• The dark
red liquid is decanted from the undissolved eS 2 and kept in a well-closed flask.
1452 CHEMICAL ANALYSIS OF IRo.N AND STEEL
dilute to 1()0 mI., and add 3 g. of tartaric acid. Make the solution alkaline
with NH 40H and filter. Acidify with Hel and1complete the determination
as directed for nickel steels.
Cast Irons.-Dissolve 5 g. of the iron in 40 l mL of diluted HCr (1 : 1),
carefully add about 15 ml. of diluted HNO a (1 l, 1) to oxidize the iron, and
evaporate to dryness. Drench the hot dried ml\ss with 10 mL of HCl and
then dilute with 75 ml. of hot water. Filter, wash with diluted HCl (sp.gr.
1.12) and evaporate the filtrate to a syrupy consistency. Add 50 mL of diluted
.HCl (1 : 1), transfer to a 200-mL separatory funnel, rinse the beaker with
several small portions of the diluted HCl (1 : 1), add 12'0 ml. of ether, and
complete the determination as directed for carbon steels.
High-Nickel Chromium Alloy Cast Irons (15% nickel, 6% copper, 2%
chromium, etc.).-Transfer a 2.5-g. sample to a 400-mL beaker or flask and
treat with a mixture of 25 ml. of HCl and 25 mL of HNO a• When solution is
. complete add 30 mL of HCI04, 5 to 10 drops of HF, and fume for 10 to 15 min.
after the chromium has been oxidized. Cool somewhat, add 100 ml. of water
and heat to boiling. Filter and wash well with diluted HCI (5 : 95), catching
the filtrate and washings in a 250-ml. volumetric flask.41 Mix the contents,
cool to room temperature, adjust to the mark and mix thoroughly. Pipette
out 50-ml. aliquot portions and proceed by the glyoxime-electrolytic method
as directed for nickel steels. Either dissolve and reprecipitate the nickel gly-
oxime, or determine any occluded copper in the deposit and correct for same.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as with carbon steels.
Nickel Wrought Iron.- Proceed as with 'Dickel steels, but dissolve the first
glyoxime precipitate and again precipitate nickel with dimethylglyoxime as
directed in the procedure with cobalt steels:
. .
1454 CHEMICAL ANALYSIS OF IRO,N AND STEEL
ganate required to impart a pink color to the solution by matching the color
in another beaker containing the same amount of the specially treated diluted
H 2S0 4 at 55° to 60° C. One milliliter of exactly 1 l't_KMn04 is equivalent to
0.01734 g. of chromium. "
Standard Ferrous Sulfate.-Dissolve 32 g. of FeS04lNH 4)2S04'·6H 20 in
a cool mixture of 50 ml. of H 2S0 4 and 950 ml. of water. To obtain the ratio
of the FeS04 solution to the KMn04 solution, take 25 ml. of the former, ~I.ilute
to 350 ml. with cool diluted H 2S04 (5 : 95), add 2 ml. of H aP0 4 and titrate
with KMn04 solution to a faint permanE;nt pink tint. Det~rmine the blank
on the same volume of water and acids, deduct, and calculate the volume of
KMn04 solution which is equivalent to 1 mI. of the FeS04 solution. The
ratio of the FeS04 solution to KMn04 must be determined daily unless the
ferrous 'solution is kept under hydrogen. Stronger f¥llutions of FeS04 and
KMn04 may, of course, be prepared for use ~th high-chromium steels. Ap-
oJ,
45 See p. 1451.
CHEMICAL ANALYSIS OF IRON AND STEEL 1455
of FeS04 solution in & solution of like volume and acidity and containing the
same amounts of the coloring elements iIi their fillal valencies; or (3) by a
second titration of the final solution. The last is: the most convenient 'and
satisfactory in occasional analyses, and may be performed by boiling the
solution which has just been titrated for 10 ,min. in \order to destroy the slight
excess of permanganate, cooling to room temperature, and then titrating with
KMn04 to the color that was originally taken as th~ end point. The sQlution
can be used for the determination of vanadium. 50
Chromium in 2-g. sample, per cent = (A X B) - (C - D) X 50 E
where A=the volume of FeS04 solution,
B=the volume of KMn04 solution which is equivalent to 1 ml. of FeS04
solution,
C =the volume of KMn04 solution,
D = the correction for the end point, and
E =the chromium titer of the KMn04 solution.
If the solutions are equivalent and a 2-g. sample is used, calculate the
chromium as follows:
Chromium in 2-g. sample, per cent = A - ( C - D) X 50 E
High-Chromium High-Nickel Steels.-Transfer 0.5 g. to a 600-mI. beaker
- and treat with 75 m!. of diluted H 2S0 4 (1 : 4). When solution is complete
evaporate to'salts, cool, dilute to 70 mI., and dissolve the iron salts. Oxidize
the iron by the cautious addition of HNO a• Add 3 to 4 drops of HF and again
evaporate just to salts. Cool, dilute to 300 ml. with hot water, add 10 ml. of
the AgNO a solution, 15 g. of (NH4) 2S208, and boil the solution for 10 to 12 min.
Add 5 mI. of diluted HCI (1 : 3) and complete the determination as in chromium
steels. Stronger solutions of FeS04 and KMn04 are more convenient for use
with high-chromium steels.
The visual end point in high-chromium steels can be detected more easily
when oxidation-reduction indicators are used, that is, ortho-phenanthroline
ferrous complex for the FeS04-K 2Cr 207. With o-phenanthroline (1 to 2 drops
of the 0.025 Molar indicator) the end point in the FeS04-KMn04 titration is
indicated by the change in color from pink to clear green (permanent for 60
sec.). With diphenylaminesulfonic acid, the chromic acid is titrated with the
FeS04 solution to a clear green end point. The FeS04 solution "in this case is
standardized either on a standard chromium steel or pure·K 2 Cr 207. Vanadium,
if present, will also be titrated as in the potentiometric method. When
o-phenanthroline is used the vanadium may be determined in the same solution,
after the initial end point is obtained, by reducing the acidity of the solution
with sodium acetate and titrating slowly at 50° C. with the standard KMn04
(0.05 N) to the green end point.
Carbon Steels and Other Steels with less than. 0.15% Chromiilm.-Transfer
10 g. of the sample 51 to a 500-mI. Erlenmeyer flask, add 110 ml. of diluted
60 See the Determination of Vanadium in Chromium-Vanadium Steels by the Ferrous
Sulfate-Potassium Permanganate Method, p. 1459.
51 Larger or smaller samples may be taken, but the volume' of acid shall be varied
accordingly. A good rule to follow is to use the ~quivalent of 1 m!. of concentrated
H 2S0 4 for each 1 g. of steel and then 1 m!. in excess.
CHEMICAL ANALYSIS OF IRON AND STEEL 1457
/
H 2SP4 (exactly 10% by volume). Heat to boiling, boil until reaction is com-
plete and then dilute with 100 m!. of boiling water. Add a solution of NaHCO a
(8%) from a burette until a permanent precipitate appears (approximately
36 m!. with carbon steels) and then 4 m!. in excess. Boil for 1 min., let settle,
filter on a rapid filter, and quickly wash the flask and precipitate two or three
times with hot water. If the precipitation has been properly performed, there
will be no more precipitate than can be conveniently handled on an ll-cm.
paper. The filtrate will become cloudy in the funnel stem and in the receiving
vessel on account of oxidation arid hydrolysis. Ignite the residue in a nickel
or iron crucible (free from chromium) and fuse with ten or twelve times its
volume of Na 202 (free from chromium). Dissolve the cooled melt by immers-
ing it in 100 m!. of cold water, remove the crucible, add 1 g. of Na 202 and boil
for 5 to 10 min. or allow to stand on the steam bath for 30 min. Filter through
an asbestos pad, 52 preferably on a small Buchner funnel, and wash with a cold
solution of NaOH (2%) containing 1% of Na 2S04. Dilute to a measured
volume and compare the color with a standard solution of K 2 Cr 20r containing
approximately the same concentration of chromium and alkali.
To prepare the standard solution, dissolve 0.283 g. of K 2Cr 207 in water and
dilute to exactly 1000 m!. Each 1 m!. shall correspond to 0.1 mg. of chromium.
When these solutions are used, they shall be made alkaline to compare with the
solution under test. Solutions containing from 2 to 10% of NaOH and 1 mg.
of chromium per 100 m!. are suitable for the colorimetric comparisons.
If the color of the unknown solution is too deep for convenient colorimetric
comparison, the solution may be boiled thoroughly to decompose all the per-
oxide, acidified, and one of the titration procedures applied.
Cast Iron (less than 0.15% chromium).-Proceed as with carbon steels.
Cast Iron (over 0.15% chromium).-Transfer 2 g. of the sample to a BOO-m!. .
beaker and add 60 ml. of the sulfuric-phosphoric acid mixture. Heat until
action ceases, add 15 m!. of diluted HNO a (1 : 1) and boil until the oxides of
nitrogen are expelled. Evaporate until salts separate, dilute with about 50 m!.
of warm water and digest until salts are dissolved. Filter and wash the paper
with warm water. If chromium is less than 0.75%, the residue will be practi-
cally free from chromium. To recover any insoluble chromium, ignite the
paper and insoluble matter until all carbon is consumed. Treat the residue with
HF and H 2S0 4 , fuse with Na 2 CO S , and add the solution of the melt to the main
solution. To the latter add 5 ml. of the AgN0 3 , dilute to 300 m!. with boiling
water and complete the determination as directed for chromium steels.
High-Nickel Chromium Alloy Cast Irons (15% nickel, 6% copper, 2%
chromium, etc.).-Ttansfer 2 g. of the sample to a 500-m!. Erlenmeyer flask
and add 20 m!. of a mixture of equal parts of HCI and HNO s. Heat until
action ceases. Then ad.d 15 m!. of HCI0 4 (60%), 5 drops of HF and evaporate
to fumes on a hot plate.' Fume 1 to 2 min. over an open flame and then 10 min.
more on a hot plate. Cool somewhat, add 50 ml. of water and transfer to a
600-m!. beaker. Add 20 m!. of :fi2S04, dilute to 300 m!. and complete the
determination by the persulfate-silver nitrate method.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as with carbon steels.
62 Asbestos is more satisfactory than filter paper. If paper is used, it shall first be
thoroughly washed with a solution of NaOH (5%) in order to remove soluble organic
matter. •
1458 CHEMICAL ANALYSIS OF IRQ:N AND STEEL
about 200 g. of mercury, and the solution electrolyzed using a current density
of approximately 0.16 amp. per sq. cm. as the solution is stirred or agitated.
Continue the electrolysis until iron is absent as indicated by a ferricyanide
test on a small drop of the electrolyte. This should not require more than 45
min. When all the iron has been removed, draw off the electrolyte, and wash
the mercury two or three times with water while the current is continued. lili
In these operations care shall be taken to prevent any amalgam from passing
into the electrolyte. li6 Add 2 to 3 ml. of diluted H 2S0 4 (1 : 1), heat to 70 0
to 80 0 C., and add KMn04 solution until a strong pink color·appears. Heat to
boiling, and pass a current of S02 into the solution until the vanadium.is
reduced (2 to 5 min.),li1 Continue the boiling, and pass a rapid stream of CO 2
(free from O 2) until the solution is free from S02. This may be ascertained by
passing the gas issuing from the flask into 5 m!. of water containing a drop of
diluted H 2S0 4 (1 : 1) and enough KMn04 to give a faint pink tint. Cool the
solution to 60 0 to 80 0 C., and titrate with the KMn04 (0.03 N). Repeat the
reduction and titration until concordant results are obtained. Correct the
titrlttion by a blank determInation (usually amounting to about 0.1 m!.) of
the KMn04 (0.03 N) on a solution of like volume and acidity. To obtain the
percentage of vanadium, multiply the corrected volume by the vanadium titer
of the solution, multiply by 100, and divide by the weight of the sample.
Carbon Steels.-Proceed as with vanadium steels but treat 5 g. of the sample
with 60 m!. of diluted H 2S0 4 (exactly 10% by volume).
Cast lron.- Proceed as with. carbon steels.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as w;ith carbon steels.
/'
CHEMICAL ANALYSIS OF IRON AND STEEL 1461
heat gently until action ceases. (For vanadium steels, use a 2-g. sample and
150 m!. of diluted H 2S0 4 (5 : 95).) Cool the solution to 15° C. and add a
goodly amount of paper pulp. Then add dropwise with constant stirring a
cold, freshly prepared 6% solution of cupferron (ammonium nitrosophenyl-
hydroxylamine, C 6 H 6N· NO· ONH4) until the precipitate just assumes a reddish
brown color. Filter through an ll-cm. paper containing some paper pulp, and
wash 10 to 12 times with cold diluted sulfuric acid (1 : 99). Transfer the
paper and contents to the original beaker, add 20 m!. of HNO a and 10 ml. of
H 2S0 4. Evaporate to fumes of H 2S0 4. Cool, add 10 ml. of HNOa and again
evaporate to copious fumes of H 2S0 4. Cool, dilute to 300 ml. add 3 m!. of
H aP0 4 (85%) and then KMn04 until the solution is pink. From this point
proceed with the addition of FeS04 as described for chromium-vanadium steel.
Cast Iron (less than 0.05% vanadium).-Proceed as with carbon steels.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as with carbon steels .
repeat the heating until the weight remains ~onstant. Treat the ignited residue
with 5 mi. of NH 40H, digest, and filter through a small paper. Wash well
with diluted NH40R (1 : 99). Ignite the paper and contents in the original
crucible, cool, and weigh. The difference in weifhts represents the Moo a
present. •
Tungsten is also precipitated. If present, the a)nmoniacal filtrate shall be
treated as follows: Add 5 mi. of diluted H 2S0 4 (1 : 1) and evaporate to fumes of
H 2S0 4. Cool, dilute to 25 mi. with water, add 1 to 2 mi. of cinchonine solution
(125 g. of cinchonine dissolved in 1000 qJ.I. of diluted HCI (1 : 1)). Digest at
80° to 90° C., preferably overnight. Filter through a tight paper containing
a little paper pulp and wash with cinchonine wash solution (30 mi. 'of the
cinchonine solution diluted to 1000 mI.). Transfer the paper and conttlnts to
a platinum crucible, char the paper and ignite at 750° to 850° C. 'until constant
weight is obtairwd. Cool, weigh, and subtract from the weight of MoO a
obtained previously. In very accurate work any residue obtained here shall
be dissolved and tested for molybdenum by the colorimetric method. 66
With high-molybdenum medium-tungsten steel (8% molybdenulll, 2%
tungsten) the ignited oxides may be weighed, dissolved, and molybdenum
determined by the MoSs-MoOs (after reprecipitation of the sulfide) method.
Tungsten is then obtained by difference. I
If molybdenum is present in very smail amounts (carbon steels), the ignited
a-benzoinoxime precipitate shall be dissolved in NH 40H and the molybdenum
determined colorimetrically. 66 .-
Cast Iron.-Transfer 1- to 5-g. of the sample to a 600-mi. beaker, treat with
100 mi. of diluted H 2S0 4 (1 : 4) and warm. When action ceases, add HNO a
drop wise until rapid effervescence ceases (usually 2 to 5 mi.) and then add 2 to
3 drops in excess. Evaporate the solution to fumes of H 2S0 4, cool somewhat,
and add 100 mi. of water. Warm until salts are dissolved and filter through
a rapid paper. Wash the paper with warm water. Dilute the filtrate to 150
mi. and cool to 25° C. Add sufficient FeS04 to reduce any chromium or
vanadium which may have been oxidized by the' above treatment, cool to
5° C. and complete the determination as directed (or molybdenum steels.
OPen-Hearth Iron.-Proceed as with molybdenum steels.
Wrought Iron.-Proceed as with molybdenum steels.
Filter and wash the sulfur and sulfides with diluted H 2S0 4 (1 : 99) saturated
with H 2S.67 Place the paper and precipitate in the original beaker,68 add
5 mL of H 2S0 4 and 20 ml. of HNO a, cover, and heat to fumes of H 2S0 4 , Cool
somewhat, add 10 rul. of HNO a and again evaporate to fumes. If the solution
is not clear and of a light color, .repeat the treatment with HN 0 3 • •
Cool, dilute to 100 mI., and add a slight excess (10 to 12 drops) of a solution
of NaOH (20%). Heat to boiling and set aside for 5 min. Filter and wash
the paper and residue with hot water. Heat the filtrate to boiling and treat
with H 2S for 10 min. Add diluted H 2S0 4 (1 : 1) until acidified and then a
4 mL excess per 100 mL of solution. Treat with H 2S for 5 min. and digest at
50° to 60° C. for 1 hr. Filter through a tight 9-cm. filter paper, and wash
thoroughly with diluted H 2S0 4 (1 : 99) saturated with H 2S.69 Transfer to a
small porcelain crucible, heat carefully until carbon is destroyed and then at
500° to 525° C. until constant weight is obtained. 70 Test the ignited oxide for
impurities by treating with NH 4 0H. If copper is indicated, determine its
amount colorimetrically, calculate to CuO and deduct. If a residue remains,
6lter., wash with water, ignite, weigh, and deduct from the weight of Mo0 3 .n
The factor for molybdenum in MoO a is 0.667.
Tungsten Steels.-Dissolve 2 to 10 g. of the sample (approximately 0.03 g.
of molybdenum) in 100 mL of diluted HCI (1 : 1), cautiously add 20 mL of
diluted HN0 3 (1 : 1), and then boil gently until.the tungstic acid becomes bright
yellow. Dilute to 150 mI., heat to boiling, filter, and wash the residue with
diluted HCl (1 : 9). Reserve the precipitate. Add 15 mL of H 2S0 4 to the
filtrate, evaporate to fumes of H 2S0 4 , cool, and add 100 ml. of water. Digest
until soluble salts are in solution. . If any ·tungstic acid separates, filter through
a small filter, wash with a little diluted H 2S0 4 (1 : 99), and combine with the
reserved tungstic acid precipitate. Add 5 g. of tartaric acid to the clear filtrate
67 In umpire analyses the filtrate shall be boiled to expel H 2S and its molybdenum
content determined colorimetrically, or the unprecipitated molybdenum (usually not in
excess of 0.5 mg.) shall be recovered as follows: Boil the filtrate to expel H 2S and to
reduce the volume to about 450 ml. Add 20 ml. of the persulfate solution, boil for 8
to 10 min., and then pass in a rapid stream of H 2S for 10 to 15 min. Digest for 1 hr.,
filter, wash, and combine with the main precipitate.
• 68 An alternate method for routine analyses consists in directly igniting the impure'
sulfide at 500° to 525 C., weighing and then correcting the weight of the ignited residue
0
as follows: To the crucible add diluted HCI (1 : 1) and heat until the residue is dissolved.
Transfer the solution to a small beaker, dilute to about 100 mI., and add NaOH in slight
excess (10 drops of a 20% solution). Boil for 1 min., allow the precipitate to settle, and
filter. Dissolve the precipitate in diluted HCI (1 : 1), and precipitate with NaOH as
before. Wash the paper and precipitate thoroughly with water to remove alkali salt;
ignite the paper and precipitate in the original crucible, and weigh. The difference
between the two weights represents MoO a.
69 Precipitation of molybdenum is usually complete, but it is well to test the filtrate
by boiling to expel H 2S, oxidizing with bromine water, boiling to expel bromine, and again
gassing with H 2S.
70 Molybdenum oxide volatilizes at temperatures above 5000 C. but the rate is very
slow at temperatures below 600 0 C. The heating may be done in a muffle with pyrom-
eter, or in a "radiator." This is a 50-mL porcelain crucible containing a disk of
asbestos board, ,4 mm. thick on the bottom and fitted with a nichrome triangle which is
bent to fit the inside of the crucible and supported by bending the end wires over the rim.
The crucible shall be placed so that the bottom is 8 cm. above the top of a Tirrill burner
and heated by a flame 12.5 cm. high.
71 If small amounts of molybdenum are in question, the determination shall be
checked by testing the solution of the oxide as described in the colorimetric method.
1464 CHEMICAL ANALYSIS OF IRON AND STEEL
and neutralize with NH 40H. Add H 2S0 4_u ntil aciqified, then 5 mI. per 100 mI.
in excess, and pass in H 2S as in the absence of tungsten.
Some molybdenum is always carried down by ,the tungstic aaid and shall
be recovered as follows: Dissolve the combined turigstic acid residues in a hot
solution of NaOH (5%), and wash the papers with ~ little water and then with
a little hot diluted H 2S0 4. Add 5 g. of tartaric acid, then H 2S0 4 until the
sblution contains 5 ml. per 100 mI., and precipitate with H 2S as in the absence
of tungsten. Filter, wash, combine with the main sulfide precipitate, and
complete the determination as described for molybdenum steels.
Cast Iron.-Transfer 2- to 5-g. of the sample to a 600-ml. beaker; treat
with 100 mI. of diluted H 2S0 4 (1 : 4), and warm. When action ceases, add
diluted HNO a (1 : 1) dropwise until rapid effervescence ceases (usually 5 to
10 ml.) and then add three to five drops in exces_s. - Evaporate the solution
to fumes of H 2S0 4 • Cool somewhat, add 100 mI. of warm water, stir, and heat
until salts are dissolved. Filter through a rapid filter paper catching the
filtrate in a 600-ml. beaker. Wash the paper well with hot water. To the
filtrate add 5 g. of tartaric acid and neutralize the solution with NH 40H:!
Add 10 ml. of diluted H 2S0 4 (1 : 1) for each 100 ml. of solution, heat to boiling,
and pass in a rapid stream of H 2S for 10 min. Dilu~ with an equal volume of
hot water, and pass in the gas for 5 min. Digest at 50° to 60° C. for 1 hr.,
filter, and complete the determination as described for molybdenum steels.
Carbon Steels.-Proceed as with molybdenum steels. -
OPen-Hearth Iron.-Proceed as with molybdenum steels.
Wrought Iron.-Proceed as with ~olybdenum steels.
present, the ignited cobalt oxide should be dissolved in HCI and cobalt again
precipitated with a-nitroso-{1-naphthoI.85 I
A blank shall be taken through all steps of the determination. it. I-g.
sample of National Bureau of Standards' standard sample 10d of bessemer steel,
or sample 50a of chromium-tungsten-vanadium,ste~l is satisfactory for this
purpose. It is especially important that the same quantity of the a-nitroso-{1-
naphthol be used in the blank run as in the determination. A little macerated
paper added to the blank after the a-nitroso-{1-naphthol 'reagent facilitates
filtration and washing.
Carbon Steels (and Other Steels Containing less than 0.10% Cobalt)..-
Dissolve 10 g. of the sample in HCI and cautiously oxidize with just enough
HNO a• Extract the iron with ether and wash the ether extract once with
diluted HCI. Warm the ether-extracted acid solution to expel residual eth~r,
and oxidize with KCI0 3 • Dilute to 200 mI., and precipitate twice with ~inc
oxide as directed for cobalt steels. In material containing very little cobalt,
it is advantageous to combine the extract acid solution obtained in e~her
separations of a number of separate lO-g. samples. ..
Cast Iron.-Proceed as with carbon steels.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as with carbon, steels.
destroy the carbon of the filter paper. Fuse the contents of the crucible with
1 g. of K 2S 20 7 and dissolve the cooled melt in 25 m!. of diluted H 2S0 4 (10 : 90).
Transfer the solution to a Camp or a Nessler comparison tube, and dilute
to a suitable volume containing 10 m!. of H 2S0 4 per 100 m!. Add H aP0 4
dropwise until any iron color is destroyed. To a second tube, add reagents
K 2S0 4, H aP0 4, etc. in amounts that are present in the unknown, and dilute
with diluted H 2 S0 4 (5 : 95) to the same volume. Cool both to room tempera-
ture. Add 3 m!. of H 2 0 2 (3%) to each and then a standard solution of Ti(S04)z
to the blank run and a similar volume of diluted H 2S0 4 (5 : 95) to the unknown
until the colors match when the solutions are well mixed. The unknown then
contains the amount of' titanium that was added to the blanls: run. If a
colorimeter of the plunger type is used, a suitable standard solution of peroxi-
dized titanium may be prepared and the colors matched by changing the depths
of the columns that are viewed. The amounts of titanium in the two solutions
are then inversely proportional to the heights of the columns.
If the steel contains an appreciable amount of copper, filter the original
H 2SO.4 solution of the sample through a small filter paper containing some
paper pulp. Wash well with hot diluted H 2S0 4 (1 : 9). Cool the filtrate to
15° to 20° C. and precipitate with cupferron as described in the first paragraph.
Transfer the paper containing the acid insoluble material to a 250-m!. beaker,
add 25 m!. of diluted HNO a (3 : 7) and heat until the copper has dissolved.
Add 50 m!. of hot water and a slight excess of NH 4 0H. Heat to boiling, filter,
and wash the paper and precipitate with hot water. Burn off the paper at as
low a temperature as 'possible and add the residue to the ignited cupferron
precipitate. Fuse the combined residues with K 2S 20 7 , dissolve in diluted
H 2S0 4 (1 : 9) and determine calorimetrically as previously described. If the
steel contains vanadium (more especially when small amounts of titanium are
sought), transfer the ignited cupferron precipitate to a 100-m!. platinum dish,
add 5 m!. of HF (48%), 10 ml. of HCl0 4 and evaporate to a volume of 5 m!. or
less. Cool somewhat, dilute to 50 ml. and add an excess of 5 m!. of the NaOH
(10%). Boil for several minutes, let settle, and filter on a tight 9-cm. filter
paper. Wash the paper and precipitate with hot water, ignite, fuse with a
small amount of K 2S 20 7, and complete the determination for titanium as
described above.
Carbon Steels, OPen-Hearth Iron, and Wrought Iron (less than 0.05%
titanium).-Treat 5 g. of the sample with 150 ml. of diluted HCI (1 : 4) and
proceed as described for titanium steels.
Cast Iron (total titanium).-Treat 5 g. of the sample with 100 ml. of diluted
HCI (1 : 2), cover, and warm. When all action has ceased, cool to 10° C.,
add 1 m!. of a solution of cupferron (6%), filter on a paper of close texture, and
wash the insoluble matter with water. Transfer the paper and residue to a
platinum crucible, dry, and ignite under good oxidizing conditions in an un-
covered crucible until all carbon is gone. Treat with 1 to 2 m!. of HF and 1 m!.
of diluted H 2S0 4 (1 : 5), and evaporate to dryness. Fuse the residue with 1 to
2 g. of NazCO g• Dissolve the melt in about 50 m!. of water, digest for 15 min.
at 90° to 95° C., filter, and wash with water. Ignite the residue in platinum,
and fuse with 1 to 3 g. of K ZS Z0 7• Cool, treat the melt with 25 m!. of diluted
H 2S0 4 (1 : 9), transfer to a Camp or Nessler comparison tube, and determine
the titanium calorimetrically as described for titanium steels.
1470 CHEMICAL ANALYSIS OF IRON ANI) STEEL
or more of concentrated nitric acid to oxidize the ferrous ir.on and carbonaceous
matter, and boil to remove nitrous fumes. Any graphite present will do no
harm and will be filtered out later. Dilute to 50 or 75 mI., add 3 m!. of 85%
phosphoric acid (or an equivalent amount of more dilute acid) and 0.3 gram of
sodium or potassium periodate. Boil gently for 15 minutes to oxidize the
manganese, dilute to 150 mI., cool to room temperature, and add slowly, with
constant stirring, 4 or 5 grams of mercuric nitrate (Hg(NO ah·2H 20) dissolved
in a little water. Filter immediately through a fairly large asbestos filter into
an excess of standard ferrous sulfate, wash with cold water four or five times,
and titrate back with standard permanganate.
Procedure for Steel or Iron Containing Not Over 0.1 % of Chromium.-
Dissolve 1 gram in a mixture of 15 mI. of water and 15 mI. of 85% phosphoric
acid. Oxidize the ferrous iron and carbonaceous matter by adding carefully to
the hot solution 1 or 2 mI. of concentrated nitric acid, and boil to remove nitro'us
fumes. Dilute to 100 mI., add 0.3 gram of sodium or potassium periodate, boil
gently 15 minutes to oxidize the manganese, dilute to 150 mI., and cool to
room temperature .. Precipitate the perioQ.ate by adding slowly, with constant
stirring, 2 to 3 grams of mercuric nitrate dissolved in a little water, and filter
immediately through a fairly large a~bestos filter into excess of standard ferrous
sulfate containing 10 mI. of 50% sulfuric acid to prevent precipitation of
mercuric phosphate. Wash with cold water four or five times and titrate back
the filtrate and washings with standard permanganate. The filtration will be
slower than when sulfuric acid is present during precipitation, but the time
required, including washipg, should not be more than 3 minutes.
Procedure for Iron Ore and Other ~)Xide Ores.-Place 1 gram in a 250 mI.
Pyrex beaker, and add 15 mI. of 85% phosphoric acid and a few drops of con- .
centra ted sulfuric acid. Stir until all the ore is free from the bottom of the
beaker, cover, and heat until fumes of suffuric acid are given off, taking care
that it does not froth over or cake on the bottom. If the ore is not dissolved
by this time, keep it hot for a longer time, stirring occasionally, but not allowing
the temperature to rise, because the beaker would be attacked. Cool until the
mass begins to be viscous, then add quickly 100 mI. of water, ~nd heat. Every-
thing should dissolve except some gelatinous silica. From here on the pro-
cedure is the same as for steel containing 0.1 % of chromium.
Procedure for Bronze.-Dissolve 1 gram in a mixture of 3 ml. of concen-
trated nitric acid, 10 mI. of 85% phosphoric acid, and about 7 or 8 mI. of water.
Then dilute to 50 or 75 mI. and proceed as above. Owing to the deep blue color
of the copper salt, the back titration with permanganate is conveniently carried
out electro metrically or by use of alphazurine indicator.
NOTE.-If the amount of manganese is between 15 and 30 mg., the amount of peri-
odate in the above procedures should be increased to 0.5 gram.
INDICATOR METHODS
400 mI. beakers and dissolve in a little water, or measure out that volume of
standard dichromate solution which will require 35 tb 50 ml. of 0.025 N ferrous
sulfate for titration. Add 5 ml. of sulfuric acid (sp.~r. 1.5), 5 ml. of phosphoric
acid (sp.gr. 1.37), and dilute to 200 m!. Then add 0.3 m!. of 0.01 M indicator,
oxidized, and titrate with approximately 0.025 N .f~rrous sulfate. No blank
correction is required in this standardization. 90 ,
Procedure for Vanadium in Tungsten Steels.-Add' 25 to 30 mI. of. wa ter
and 5 ml. of sulfuric acid (sp.gr. 1.83) to a I-gram sample of the steel weighed
into a 400-ml. beaker. Warm gently until the steel is completely decomposed
and the tungsten separates out as-a black powder. Rub loose from the beaker
all tungsten and to the boiling hot solution add 5 ml. of hydrofluoric acid
(48%), then nitric acid (sp.gr. 1.42), at first sl<¢y, using 5 m!. in all., Boil 2
minutes. A clear green solution is obtained. If the solution is not P!'ldectly
clear at this point, as occasionally happens with steels especially difficult to
decompose, boil 2 or 3 minutes longer. Dilute to approximately 100 mI., add
1 gram of ammonium persulfate, and boil 5 minutes. Add 5 ml. of 0.1 N
ferrous sulfate, dilute to between 175 and. 200 ml., and cool to room tempera-
ture. Add to this solution 0.1 N potassium permanganate from a· buret until
a distinct color persists for 2 minutes, to be sure all the vanadium is oxidized.
At this point either procedure A or B may' be used. Procedure A is preferable.
A. Add 0.05 M sodium nitrite slowly from a buret until the solution be-
comes green, then 5 ml. excess of the ,nitrite solution, followed by 2 grams of
urea. Stir the solution thoroughly and allow it to stand for 5 minutes. '
B. Add 5 ml. of 0.1 M sodium azide and boil vigorously (in hood) for 5
minutes to remove all the hydrazoic acid. Cool to room temperature.
To the solution thus obtained add 3 ml. of hydrofluoric acid (48%) and 0.3
ml. of 0.01 M diphenylamine sodium sulfonate which has been oxidized, and
titrate at once with 0.025 N ferrous sulfate standardized as directed below.
The color change of the indicator at the end point is from purple to green. The
correction to be applied for the indicator may be determined from the following
table and should be subtracted from or added to the volume of ferrous sulfate
used in the titration. 90 '
Indicator Corrections in
Vanadium in Steel 0.025 N FeSO.
% ~
O-{}.85 , ... , ...................... '................ ,., -0.10
~:~5::1~ ::::: :'::::::: ::::: ::::::::::::::::: ::::::~>:,; tg:~g
Standardization of Ferrous Sulfate for Chromium Plus Vanadium Deter-
minations.-Place in a 600 ml. beaker that tlmount of potassium dichromate,
either in solid form dissolved in a little water or in the form of a standard
solution, which will require approximately 35 to 50 ml. of 0.05 N ferrous sulfate.
Add 5 m!. of sulfuric acid (sp.gr., 1.5), 5 m!. of phosphoric acid (sp.gr. 1.37) and
dilute to 300 ml. Then add 0.3 mI. of 0.01 M indicator, oxidized, and titrate
with approximately 0.05 N ferrous sulfate. No blank correction for the indi-
cator is required here.
Procedure for Chromium Plus Vanadium in,Tungste'n Steels.-Add 25 to
30 ml. of water and 5 m!. of sulfuric acid (sp.gr.'1.83) to a I-gram sample of the
90 Willard and Young, Ind. Eng. Chern., Anal. Ed., 5, 154 (1933).
CHEMICAL ANALYSIS OF IRON AND STEEL 1475
steel weighed into a 600-ml. beaker. Warm gently until the steel is completely
decomposed and the "tungsten separates out as a black powder. Rub loose
from the beaker all tungsten and to the boiling hot solution add 5 ml. of hydro-
fluoric acid (48%), then nitric acid (sp.gr. 1.42), at first slowly, using 5 ml. in
all. Boil 2 minutes. A clear green solution is obtained. If the solution is
not perfectly clear at this point, as occasionally happens with steels especially
difficult to decompose, boil 2 or 3 minutes longer. Dilute to approximately
300 mI., heat to 60° C. or. higher, add 10 ml. of silver nitrate solution (2.5
grams of silver nitrate per liter), 5 grams of ammonium persulfate, and boil
vigorously for 10 minutes. A few small pieces of broken porcelain in the solu-
tion will prevent bumping. Anyone of the three following procedures may be
used to destroy the permanganic acid formed during the oxidation process.
Procedure A is preferable.
A. Add 0.1 M sodium azide, drop by drop, from a pipette to the boiling
hot solution until all permanganate has been reduced, limiting the excess of
azide used to 1 or 2 drops. Boil for 1 or 2 minutes to remove the hydrazoic acid.
B. Add 0.05 M sodium nitrite, drop by drop, from a pipette to the boiling
hot solution until all permanganate has been reduced, limiting the excess of
nitrite used to 1 or 2 drops. Then add 1 gram of urea at once and stir the solu-
tion thoroughly.
C. Add 5 ml. of 1 to 3 hydrochloric acid to the hot solution and boil vigor-
ously for 10 minutes to reduce the permanganate and remove all chlorine.
To the solution from A, B, or C, cooled to room temperature, add 3 ml.
of hydrofluoric acid (48%) and 0;3 ml. of 0.01 M diphenylamine sodium sul-
fonate, oxidized, and titrate at o,nce with 0.05 N ferrous sulfate standardized
as directed below. Immediately after the addition of the indicator the solution
is a yellowish brown. As the ferrous sulfate is added this yellow color disap-
pears gradually, being replaced first by a purplish red and then by a clear
purple. The color change at the end point from purple to green is very sharp.
An indicator blank -of -0.30 ml. of 0.05 N ferrous sulfate should be applied
to the volume of ferrous sulfate used in the titration.
The usual procedures for determining chromium and vanadium in steels
that do not contain tungsten may be carried out using oxidized diphenylamine
as the indicator.
Preparation of IndicaJor.-A 1/10% solution of diphenylamine is prepared
by dissolving 0.1 gram of the indicator in 10 ml. of sulfuric acid (sp.gr., 1.83)
and diluting this solution with 90 ml. of glacial acetic acid. The method of
preparing portions of oxidized indicator solution from the 0.1 % solution is the
same as used for diphenylamine sulfonic acid, except that 5 m!. of phosphoric
acid (sp.gr., 1.37) instead of water are added to a measured volume of the
indicator in a very small beaker. No sulfuric acid is added, but only 3 or 4
drops of dichromate, followed by very dilute ferrous sulfate.
Procedure for Vanadium in Steels without Tungsten.-Add 30 to 40 m!. of
water to a 4- or 5-gram sample in a 600 ml. beaker. Run in a measured volume
of sulfuric acid (sp.gr., 1.83) from a burette, allowing 1.5 ml. for each gram 9f
steel and 3 ml. in excess. After the steel has been completely decomposed,
boil until a considerable quantity of salts separates out, in order to assist in
decomposing carbides. Dilute with 30 to 40 ml. of water and heat until the
salts have dissolved. To the boiling hot solution add nitric acid esp.gr., 1.42)
1476 CHEMICAL ANALYSIS OF IRON. AND STEEL
to 2 or 3 drops. Then add 1 grain of urea at once and boil the solution for
5 minutes. A reddish color which may develop in the solution during the
boiling period may be disregarded.
To the solution from A or B, cooled to room temperature, add that quantity
of crystallized sodium acetate which is required to react with the excess of
sulfuric acid used in dissolving the steel (1 m!. of concentrated sutfuric acid is
equivalent to 4.8 grams of sodium acetate trihydrate). As soon as this has
dissolved, add 0.5 m!. of 0.1 % diphenylamine, oxidized, and titrate with 0.05 N
ferrous sulfate. The ferrous sulfate should be standardized by the procedure
given under vanadium in steels. without tungsten. That volume of standard
dichromate should be taken which will require about 35 to 50 m!. of ferrous
sulfate.
the excess will require about 23 grams of acetate. If the steel was first dis-
solved in nitric acid, the excess of perchloric acid will be a little greater. The
proper amount of acetate has been added when more would form a permanent
white precipitate 'of ferric phosphate. If this occurs, add dilute sulfuric acid,
with constant stirring, until the solution clears. Unless the acidity is reduced
in this way, the color- change at the end point will be slow. Add 0.6 to 0.8 ml.
of 0.1 % diphenylbenzidine ~olution (in phosphoric or acetic acid), and allow
5 minutes, but not more than 10, for the purple color to develop. It wiIi
appear brownish until most of the chromic acid is reduced. If no color ap-
pears, add a few drops of dilute sulfuric acid. Titrate with ferrous sulfate to a
clear green, approaching the end point carefully. Add a correction of 0.04 m!.
of 0.1 N ferrous sulfate for each 0.5 m!. of indicator.
For the electrometric method add 25 to 30 ml. of sulfuric acid (sp.gr. 1.5)
and titrate chromic plus vanadic acids with ferrous sulfate. If much vanadium
is present, a much sharper end point is obtained by cooling the solution to 5° C.
Whichever method is used, it has already been shown that the ferrous
sulfate should be standardized in the same way and under the same conditions
of volume and acidity at about the time it is used.
Titration of Vanadium.-This may be determined by any of the usual
methods.
Silica has been dehydrated. by the boiling perchloric acid and after the
titration it may be filtered off and determined. 95
green. The end point is very sharp and there is, almost no return of the pink
color for a minute or more. Add sufficient crystallized i5~dium acetate to react
with the free perchloric acid present. The amount of this ~alt requil?d may
be estimated approximately by considering that 5.4 ml. pI 70% perchloric acid
are used to decompose and oxidize each gram of steel ~nd that 1.6 grams of
the acetate will be required for each cubic centimeter of the acid remaining.
If preferred, the acetate may be added in small p'ortions to the solution while
it is being heated, until the addition of further acetate would cause the f0rmation
of a permanent precipitate of ferric phosphate. Heat the solution to 50° C.,
using a thermometer as stirring rod, and titrate the vanadyl at once with the
0.05 N Plilrmanganate. The oxidizing agent should be added slowly, as only
a small volume of it will be required and also because the reaction between
vanadyl and permanganate ions is not instantaneous. The color change at
the end point is the same as in the first titration and the end point is con-
sidered as reached when there is no return of a pale pink shade during an inter-
val of a minute. It is sometimes advisable to add a drop more of indicator
,before titrating the vanadium.
To determine the normality of the ferrous sulfate solution, add sufficient
of it to require 35 to 50 ml. of the permanganate to 250 ml. of water containing
10 to 15 ml. of 70% perchloric acid. Add 2 dI:ops of 0.025 M o-phenanthroline
ferrous complex and titrate at once with the standard permanganate solution.
97 Standard Procedure of the Carpenter Steel Co., R¢aa'i;g Pa., through the Courtesy
of W. J. Boyer, Asst. Chief Chemist.
CHEMICAL ANALYSIS OF IRON AND STEEL 1481
to the original beaker and draw off the ether solution into a c~ean, dry 50 ~IDl.
graduated cylinder. Shake out the acid mixture with Isuccessive 10 ml. por-
tions of ether until all color is removed, if one extraction ~oes not remove same. >
The volume of ether used for the sample should be apprpximately the same as
is required to give a corresponding color with the standard. Transfer the
combined ether solutions to a graduated matching tube.,
Color Comparison.-For this operation a Kennicott-Campbell Hurley
colorimeter may be used. Illumination is supplied by a Wratten Safelight No.
2, made especially for dark-room purposes. This fixture carried a 50 watt bulb,
which gives sufficient light and does not materially affect the atmospheric
temperature and thus cause a rapid deterioration of the standard due to
evaporation of the ether. It is essential that the light be so placed that the
illumination is equal on both mirrors.
Place the matching tUbe containing the unknown in the colorimeter, and
adjust the leveling device until the depth of color in the tube containing the
standard equals that of the unknown. Take the reading in milliliters of the
standard solution in the gradua~ed matching tube. From this value,·the
molybdenum content of the unknown steel canbe calculated.
Standard Soiution.- Weigh two 0.5-gram Ijamples of standard steel. Run
through in the usual manner and make the ether extract up to exactly 50,ml. in
the standard matching tube. Place in the colorimeter and compare. If the
color tints match, one sample may be diluted to exactly 100 ml. as a standard for
low percentage molybdenum, or the two may be united to make a standard for
steels with a higher molybdenum content. Transfer to the leveling tube of
the colorimeter.
The top of the leveling tube should be covered when not in use, in order to
increase the time that a standard can be used. Since the deterioration Ofi a
standard is generally due to evaporation of tlie ether rather than breaking
down of the salt, its life is largely dependent on the room temperature. In this
laboratory the standard is renewed every two hours, two samples of standard
steel being kept fuming slightly on the cooler portion odhe hot plate, to be used
whenever the standard in use is suspected.
Calculation of Results.-If a steel containing 0:23% molybdenum is used
as a standard, a I-gram sample will contain 0.0023 gram of molybdenum; on
dilution to 100 m!. each cubic centimeter will contain 0.000023 gram of molyb-
denum. If 14; ml. of the s~andard are used to match the color of the unknown
steel, the unknown must contain 14 XO.000023, or 0.000322 gram: .
Since the weight of the unknown used is 0.5-gram, it will contajn twice this
amount per gram, or 0.064%.
Ml. standard (14) Xgrams/ml. (0.Q00023)
Grams sample (0.5)
% Mo (0.064).
a short and accurate method for determining the exact amount of these elements
held in the steel. . By electrolysis the interfering elements are deposited while
titanium, zirconium and aluminum, if present are left in the solution.
Apparatus.-The mercury in this method is used as the cathode and the
power is supplied by a l/l6-in. thick platinum wire sealed in a glass tubing.
About Y2 in. of the platinum wire is left outside the lower end of the sealed glass
tube to be dipped in the mercury from the side of the beaker and about 2 in.
above the upper end of the glass tubing to be. connected to the negative pole
of the circuit.
A platinum plate, about 134 in. in diameter, welded to a Ys-in. thick plati-
num wire and sealed in a glass tube, is used as the anode. The anode is con-
nected to the positive pole of the circuit, and lowered about Y2 in. below the
surface of the solution. The plate is provided with a hole about % in. diameter
in the center for the stirring rod.
The mercury is constantly agitated during electrolysis by a " T "-shaped
a
glass rod passing through the hole in the plate and attached to motor driven
stirrer. The" T " end of the rod reaches the bottom of the mercury and is
adjusted to ope,rate without touching the beaker or electrodes.
Procedure.-Transfer one gram sample (two grams if titanium is less than
0.5%) to a 200-ml. beaker, add 20 to 40 ml. of 10% sulfuric acid, and heat slowly
until action is complete. Filter into a 200 m!. tall electrolytic beaker containing
mercury about one inch deep, wash a few times with water, and reserve the
paper. The filtrate and washing should not exceed 75 m!. in volume
Set the beaker containing the solution in a trough of circulating water and
electrolyze at 4 amp. for lY2 to 2 hrs. As soon as electrolysis begins, transfer the
reserve filter paper to a platinum crucible and ignite. Cool, add a little potas-
sium bisulfate and fuse. Add 5 ml. water and 3 drops of sulfuric acid in the
crucible and heat to dissolve the melt. Cool and combine with the main
solution.
At the end of a set period, withdraw a drop of the electrolyte and treat
with a drop of fresh 1 % solution of potassium ferrocyanide; if a blue color does
not develop, raise the electrodes above the surface of the electrolyte and wash
with cold water. Shut off the power and get the beaker out of the water
trough. Filter, wash about 5 times with cold Y2% sulfuric acid and proceed
with the filtrate by either 'of the following two methods.
1. Add enough sulfuric acid to the filtrate to make it 10% acid by volume,
cool and treat slowly with 6% solution of cupferron while stirring, until furthr
addition does not cause a precipitate. Allow this to settle for 5 min. Filter on
suption and wash 5 times with 10% cold hydrochloric acid and 10 times with
cold water. Transfer the paper to a platinum crucible, dry and ignite. Volatil-
ize the silica with hydrofluoric' acid in the presence of sulfuric acid and again
ignite at 10000 C. in an electric muffie for one-half hour. Cool and weigh.
2. Treat the filtrate with excess of ammonia, add a couple of crystals of
ammonium persulfate and boil for 2 min. Add paper pulp in the beaker and
stir. Filter and wash with ammoniacal hot water. Transfer the paper to a
platinum crucible and ignite; Volatilize the silica with hydrofluoric and sul-
furic acid and ignite at 1000 0 C. in an electric muffile for Yz hr. Cool and
weigh.
1484 CHEMICAL ANALYSIS OF IRON AND STEEL
Fuse the residue left in the crucible with a little potassIum bisulfate, cool
the melt and dissolve in 20 m!. of 20% sulfuric acid. Transfer the solution
to a 200-m!. beaker, add 2 grams of citric acid, and stir to .dissolve. Render
the solution decidedly alkaline with ammonia and saturat~ for 5 min. with
H 2S. Filter and wash 10 times with 5% ammonium citrate saturated with
ammonium sulfide. Transfer the paper to a crucible and J.gnite.· Cool and
weigh. The difference in weight is titanium or zirconium oxide. If titanium
is sought multiply by 60, and ·if zirconium is sought multiply the weight of
oxide by 73.9. . .
This method can be applied to all kinds of steel alloyed with titanium or
zirconium.
Determination of aluminum in nitralloy or other aluminum alloy steels is
made by this method, proceeding exactly as described above and precipitating
aluminum by ammonia after iron, chromium and other interfering elements
are deposited.
The determinations of titanium and zirconium in ferrotitanium and ferro-
zirconium are also made by this method, precipitating either one by cupferron
after electrolysis is compJeted.
This procedure works well for uranium in steels, if vanadium is absent.
Mercury does not deposit the vanadium as it does the iron, manganese, chro-
mium, cobalt, nickel, molybdenum, tin and copper.
The mercury can be cleaned and used again after every determination either
by shaking it in a separatory funnel with concentrated hydrochloric acid, or
pressing through a chamois skin.
ALLOYS, III
FERRO-ALLOYS
TENTATIVE METHODS OF CHEMICAL ANALYSIS OF
FERRO-ALLOYS 1
A. S. T. M. DESIGNATION: A 104-36 T 2
FERROSILICON
DEl'ERMINATION OF SILICON BY FUSION WITH SODIUM PEROXIDE AND
DEHYDRATION WITH HYDROCHLORIC ACID
Transfer 0.5000 g. (use 0.4000 g. for 90% silicon) of the finely powdered
sample (passing a No. 150 sieve 3) to a 50-m!. pure iron crucible made from No.
20 gage (0.32 in. thickness) ingot iron, add approximately 10 g. of sodium
peroxide. Mix thoroughly with a platinum or iron rod and carefully clean the
rod of adhering particles by scraping with another rod. Cover the mixture
with a layer of about 2 g. of additional peroxide. Heat the crucible and con-
tents on a hot plate for 5 to 10 minutes to expel any water in the peroxide which
would cause spattering in the subsequent fusion. Carefully fuse over a'low
flame by holding the crucible with a pair of tongs and slowly revolving it
around the outer edge of the flame until the contents have melted down quietly.
When the fusion is molten rotate the crucible carefully to stir up any unattacked
particles on the bottom or sides, the crucible and contents being maintained at
a low red heat. Just before completion of the fusion, which requires only three
or four minutes, increase the temperature to bright redness for a minute. If
the reaction has proceeded too violently, due to too rapid heating, to the use of
insufficient sodium peroxide, or to the lack of thorough mixing, appreciable loss
will occur and the work should be repeated.
Cover the crucible and allow it to cool until the fusion has solidified, and
tap the crucible, while still warm, several times on an iron plate, which will
loosen the fused mass in a solid cake. When the crucible and cake have com-
pletely cooled, transfer the cake to a large (275-ml.) gold or preferably platinum
dish. Cover the dish with a clock glass and cautiously add 50 ml. of cold water.
1 Under the standardization procedure of the Society, these methods are under the
jurisdiction of the A. S. T. M. Committee E-3 on Chemical Analysis of Metals. Repro-
duced through the courtesy of the American Society for Testing Materials, 260 S. Broad
St. Philadelphia, Pa.
2 This is a Tentative Standard and under the Regulations of the Society is subject
to annual revision. Suggestions for revision should be addressed to the Headquarters
of the Society, 260 S. Broad St., Philadelphia, Pa. Issued as Tentative, 1925; Adopted,
1927; Revised, 1934; Reverted to Tentative, 1936.
3 For detailed specifications for this sieve, see Standard Specifications for Sieves for
Testing Purposes (A. S. T. M. Designation: Ell) of the American Society for Testing
Materials, 1936 Book of A. S. T. M. Standards, Part II, p. 1413.
1485
1486 FERRO-ALLOYS,
When ,the reaction ceases, wash the small amount of adhering matter in the
crucible into the dish with a little water. Cool the solution and carefully a,dd
an excess of hydrochloric acid. Evaporate to dryness, preferably on a steam
bath, but do not bake at a temperature above qOo C. If platinurii or gold
dishes are not available for solution of the fused cake, the latter can be dis-
integrated with water in a pure nickel dish and th,e contents then transferred
to a porcelain dish containing sufficient hydrochloric acid to provide an excess
of acid. It is not desirable to dissolve the fusions directly in porcelain, because
of the action of the alkaline solution.
When the residue is dry, allow the dish to cool. Add 10 ml. of hydrochloric
acid (sp.gr. 1.13), cover the dish and heat gently for a few minutes. Dilute,
with 100 ml. of hot water, digest for a few minutes and filter through an 11-cm.
paper. Wash about eight times with hot 5% hydrochloric acid, three times
with warm 5% sulfuric acid and a few times with hot water. Reserve this
paper until the second evaporation has been completed.
Evaporate the filtrate, to dryness and bake the covered casserole or dish
at 110 0 C. for one hour. Heating at a higher temperature is unnecesSRry and
also disadvantageous. Cool, add 10 ml. o~ hydrochloric acid (1 : 1), digest on
the steam bath for ten minutes, add 100 ml. of warm water, filter immediately
and wash about eight times with cold 1 % hydrochloric acid, twice with cold 5%
sulfuric acid and a few times with cold water.
Place the paper and silica from the second dehydration in a large partially
covered platinum crucible and char the paper without inflaming. Add the
paper containing the silica from the first dehydration, char as before and then,
ignite carefully until the carbon has been burned off. Great care should be
exercised in igniting the papers as the current of air produced by a burning
filter is sufficient to carry finely divided silica out of the crucible. When the
carbon of the filter paper has been completely burned, gradually heat the tightly
covered crucible in a muffle at 1100 0 to 11250 C. for 30 minutes. Cool in a
desiccator, weigh and check for constant weight by ten minutes additional
heating at the above temperature. Add sufficient sulfuric acid (1: 1) to
moisten the silica and then 15 to 20 ml. 6f pufe hydrofluoric acid (48%).
Evaporate to dryness, ignite and weigh. Tne loss represents silica which con-
tains 46.72% of silicon. A blank, not omitting the same amount of hydrofluoric
acid, should be carried through all stages of ,'the operation.
N OTE.~The silica remaining in the filtrate after the second. dehydration is practically
negligible, averaging about 0.0006 to 0.0010 g. in hydrochloric dehydration and 0.0010
to 0.0015 g. in sulfuric dehydration.
FERROMANGA~ESE
FERROCHROMIUM
DETERMINATION OF CHROMIUM BY FUSION WITH S.ODIUM PEROXIDE
I
mix the contents of the crucible with a small rod, being careful to clean ·the rod,
which can be done conveniently by scraping with another rod. Cover the
mixture with 1 to 2 g. of sodium peroxide. Carefully fuse the c'ontents of the
covered crucible, preferably in an electric muffle heated to 600 0 to 700 0 C.
Fusion for five minutes at a low red heat after the mass has melted will insure
complete decomposition. When the charge has melted, the crucible should be
given a slight rotary motion to stir up .any unattacked particles.
Place the cooled crucible and cover in a·600-ml. beaker and dissolve the
contents in 200 ml. of water. 'Whim the melt has diss91,ved, remove the crucible
and cover, add 1 g. of sodium peroxide and boil the solution ten minutes to
completely decompose the excess of peroxide., '
Allow the precipitate to settle and filter through a pad of ignited asbestos,
preferably in a Buchner funnel. An asbestos rather than a paper filter should
be used, for the aJkaline solution extracts reducing matter from paper which
would cause low results for chromium. T.horoughly wash the precipitate with
hot water, and at no time allow the precipitate to run dry during the filtering
and washing. If desired, the precipitate can be subsequently dissolved in
dilute sulfuric acid and tested for chromium by adding a slight excess of sgdium
peroxide; the amount of chromium retained is negligible if the precipitate is
filtered and washed as directed.
Cool the filtrate, acidify with 50 m!. ·of sulfuric acid (1 : 1), again cool to
room temperature, transfer to an SOO-m!. beaker, dilute, with cold water to
500 ml. and add a measured excess of ferrous ammonium sulfate. The 'point
at which reduction is complete can be told by changes in the deep green color
which is developed. The ferrou§. salt can be added either as an approximately.
0.12 N solution or in weighed portions from a bottle of the well-mixed salt.
If the solution of ferrous salt is used, approximately 180 m!. \vill be required
for a 70% alloy; if the salt is added, 8 g. will usually be sufficient. Titrate
the excess of ferrous salt with standard O~10 N potassium permanganate. The
first faint permane:r:t darkening of the clear green color is taken as the end point,
which is quite sharp and well delined to the practiced eye. If the ferrous salt
has been standardized against dichromate, subtra;ct the required number of
milliliters of 0.1 N'permanganate from the number of milliliters of 0.1 N ferrous
solution (obtained by multiplying the number of milliliters of the approximately
0.1 N ferrous solution by the, 0.1 N normality factor). The difference repre-
sents the number of milliliters of 0.1 N ferrous solution which multiplied by
q.001733 g. gives the number of grams of chromium present.
,"Vhere the ferrous solution has been standardized against sodium oxahite,
it is necessary to subtract a " blank" from the permanganate titr~tion. This
(/ blank." is obtained as follows: When the end point is reached and the readings
have been noted, add about 5 m!. in excess _of the ferrous solution-and oxidize
the_ excess of ferrous iron by' adding 8 mI. of a 15% solution of ammonium
persulfate. Stir the liquid vigorously for one minute. Carefully add O.l,N
permanganate to the solution with constant stirring until the same end point
is obtained as before. The volume of permanganate required, u~uaJly 10.2 to
0.4 mI., constitutes the H blank" to be deducted from the volume of the per-
manganate titration." For example, ,in. the determination of chrotPium in ,an
0.5000-g. sample, 190.0 m!. of a 0.11 N ferrous solution and 11.0.0 m!. of an
0.1l'{ p€rmanganate (11.40-0.4 m!. blank) were employed.
FERRO-ALLOYS 1493
DETERMINATION OF SILICON
Low-carbon ferrochromium and chromium metal can be decomposed by
means of dilute sulfuric acid, but the high-carbon alloy has to be fused with
sodium peroxide. For low-carbon ferrochromium, transfer 0.9344 g. (a double
factor weight) to a 250-ml. porcelain casserole, add 60 ml. of sulfuric acid
(1 : 4) and warm the covered casserole gently until the reaction is complete.
For high-carbon ferrochromium, transfer 0.9344 g. of the sample passing a No.
100 sieve to a 30-ml. pure iron crucible made from No. 20 gage (0.032 in.
thickness) ingot iron. Add approximately 8 g. of sodium peroxide, mix thor-
oughly and fuse as directed under the determination of chromium. When
the fusion has solidified, tap the covered crucible, while still warm, on an
iron plate to loosen its contents in a solid cake which is then transferred to a
250-ml. casserole. Rinse the crucible with a little water, transfer the solution
to the casserole, cover, and~ carefully add 60 ml. of sulfuric acid (1 : 4) after
which the cover glass is rinsed and removed.
Evaporate the solution obtained by either of the above procedures until
fumes of sulfuric acid are given off. If chromium sulfate solutions are fumed
at too high temperatures or for any great length of time or with too]ittle free
acid present, considerable difficulty will occur in dissolving the salts. The
fuming should, therefore, be made at a temperature not much higher than that
required for evolving the acid. Allow the casserole and contents to partly
cool (if permitted to become entirely cold, difficulty may be experienced in
getting chromium salts to dissolve), add 100 ml. of warm water while stirring
11 See p. 1426.
1494 FERRO-ALLOYS
to prevent the residue from caking on the bottom and boil the solution gentiy
for several minutes to dissolve all salts. Filter, lwashing the paper several
times with cold dilute hydrochloric acid (2%), and then with warm water.
Transfer the filter and contents to a platinum cruciqle, char the paper, carefully
ignite to remove carbon, and then heat the covered crucible over a blast lamp
or in a muffle furnace at 10500 to 1l0J o C. for at least 15 minutes. Cool in a
desiccator and weigh. Add sufficient dilute sulfuri~ acid (1 : 1) to moisten the
silica and then a small amount of hydrofluoric acid. Evaporate to dryness,
ignite and weigh. The difference in weights represents silica which contains
46.72% of silicon.
NOTE.~Crucibles made from pure ingot iron contain only traces of silicon and the
amount of silica present in the sodium peroxide used is usually negligible, but "blanks"
should be run on new lots of crucibles and of peroxide. All nickel crucibles, even the
German "Rein" nickel, contains appreciable amountE! of silicon.
.'
FERROVANADIUM
DETERMINATION OF VANADIUM BY THE .AMMONIUM PERSULFATE METHOD
Next reduce the vanadium in the cold (15 0 C.), solution by ferrous sulfate
added in approximately 1.5 g. excess.. The use 6f~ a 50% solution of the salt
furnishes a convenient way of adding the necessary amount. One gram of
quinquevalent vanadium requires 5.5 g. of FeS0 4 17H 20 for reduction to the
quadrivalent state.\ A 4-g. sample of 30% vanadiub alloy, including an excess
of 1.5 g., would require 8.1 g. of ferrous sulfate. Too large'an excess of ferrous
sulfate should be avoided. The addition of ferrous sulfate will change the
color of the solution to a deep blue. Add 2 drops of a saturated solation of
sulfurous acid, 100 mj.. of molybdate reagent, shake for 10 minutes and allow
to stand at 150 C. for 4 to 6 hours, or preferably overnight. It is w.ell to let
the solution stand in a refrigerator. Filter ~n a paper of close texture, keeping
as much of the precipitate in the flask as possible, and wash the flask and the
precipitate with not more than 50 ml. of cold, dilute nitric acid (1 : 9~). Set
the filtrate and washings aside after thorough mixing and see whether further
separljotion of phospho molybdate occurs. Dissolve the precipitate on the filter
in 20 ml. of dilute ammonium hydroxide (1 : 1) to which have ,been added 2 g.
of citric acid, catch the solution in the original flask containing the bulk of the
yellow precipitate and finally wash the filter several times with dilute hydro-
chloric acid (5 : 95). If the ammoniacal solution of the phospho molybdate is'
not clear at'this point, heat it to boiling, ~lter through the same paper, and
wash the paper with hot wll.ter. In this case, and whenever appreciable amounts
as such elements as titanium or zirc_onium are known to be present In the
original material, preserve the paper and ignite it in a small platinum crucible
with the paper reserved below.
Render the ammoniacal solution acid with hydrochloric acid, add 20 ml.
of magnesia mixture and then ammonium hydroxide (sp.gr. 0.90) slowly until
a crystalline precipitate appears, and then finally an excess of 3 to 5% by
volume. The volume of the solution at this point should not exceed 100 ml.
Allow the solution to stand in a cool place for 4 to 6 hours, or preferably
overnight.
Filter, keeping as much of the precipitate as pq§sible in the flask, and wash
the flask and the' precipitate moderately with dilute ammonium hydroxide
(1 : 20). Set the filtrate and washings aside after thorough mixing and see
whether further precipitation occurs. Dissolve the precipitate on the filter
in 25 ml. of dilute hydrochloric acid (1 : 1), catching the solution in the original
flask containing the bulk of the precipitate. Wash th~ filter thoroughly with
dilute hydrochloric acid (5 : 95). In case the original material contains ap-
preciable amounts of the elements mentioned above (titanium or zirconium),
add the paper to the one reserved above, ignite, fuse any.residue with as little
carbonate as possible (not over 0.5 g.), extra~t the cooled melt with hot water,
filter, and add the water solution to the hydrochloric acid solution. Transfer
the solution to a 200-ml. beaker. In case the amount of arsenic is known to be
very small, add 0.5 to 1 g. of ammonium bron:{ide and boil the hydrochloric
acid solution to a volume of 5 to 10 ml. Where large amounts of arsenic were
originally present, treat with hydrogen sulfide, digest, filter, and wash with
acidulated hydrogen sulfide water. ,Boil to expel hyd~ogen sulfide and to
reduce the volume to 50 to 75 ml. Add 0.1 to 0.2 g. of citric acid, 2 to 3 ml. of
magnesia mixture, make ammoniacal as above', and allow to stand in a cool
place for 4 to 24 hours.
FERRO-ALLOYS 1497
o
Filter, wash with dilute ammonium hydroxide (1 : 20), and ignite the
precipitate carefully and at as Iowa temperature as possible until the carbon
has been destroyed and the residue is white. Finally, ignite at approximately
1000° C. to constant weight. Dissolve the ignited residue in 5 ml. of dilute
nitric acid (1 : 1) and 20 ml. of water. Filter, wash the residue with hot water,
ignite and weigh. Treat with a few drops of hydrofluoric acid, evaporate to
dryness, ignite and re-weigh. Subtract any loss of w~ight from the original
weight of the pyrophosphate, and calculate on the basis of this purified mag-
nesium pyrophosphate. .
DETERMINATION OF SULFUR BY THE OXIDATION METHOD
of the precipitate need not have been removed. Evaporate the solution to
near dryness on a steam-bath, add 10 m!. of strong hydrochloric acid and pass
hydrochloric acid gas into the solution to saturation, add 15 ml. of ether and
continue passing hydrochloric acid gas to saturation. keeping the solution cold
by means of ice water. Let stand several hours, filter on asbestos and wash
with the ether-hydrochloric acid mixture as before.
Dissolve the aluminum chloride in dilute hydrochloric acid (1 : 9) into a
250-m!' beaker, add a few grams of ammonium chloride, make faintly alkaline
with ammonia, boil, filter (using paper pulp), wash with hot water, dry, ignite /
and weigh as AI 20 a.
According to published data, this method gives a complete separation of
aluminum from iron, beryllum, zinc, copper, mercury and bismuth. A little
vanadium is carried down in the first separation, but two chloride precipitations
seem to give a practically complete separation. The proportion of vanadium
retained in the final alumina is hardly appreciable when working on a 35%
vanadium alloy.
The effect of chromium is not known; presumably some of it would be
included in the AhOa. This could be corrected for by fusing the Al 20a with
sodium peroxide and estirnnting the Cr 2 0a by color.
NOTEs.-1. Dilute nitric acid (sp.gr. 1.135) is recommended for solution. While this
requires rather long digestion at steam-bath temperature, solution seems to be quite
complete even when the alloy is high in silicon., In some cases where silicon is very
high it is necessary to use a little hydrofluoric acid with the nitric acid. In this case
solution should be effected in a platinum dish, the solution evaporated to dryness and
the dry residue transferred to a beaker before treatment with hydrochloric acid. A little
dilute hydrochloric acid may be used to loosen any oxides that adhere to the dish.
2. The siliceous residue may retain alumina; it should be ignited in platinum. Silica
should be expelled by HF+H 2SO. and the residue fused with KHSO.. After dissolving
the fusion in water plus a little HCI, the liquid is made faintly alkaline with ammonia
and boiled. Any iron and alumina precipitate is removed by filtration, washed with
water, dissolved in a little hot dilute HCI and added to the solution of the alumina ob-
tained after the first ether-hydrochloric acid separation.
and digest the solution at 80° to 90° C. with occasional stirring, for 30 to 45
minutes.. I
Add some paper pulp to the solution, stir the solution and when the tungsten
precipitate has settled, filter on a 9-cm. paper c,ontaining a little "ashless"
filter pulp. Wash thoroughly with hot cinchonine solution and finally·several
times with warm 1% hydrochloric acid. Gently ignite the filter and residue of
crude WO a in the crucible in which it was originally treated until the carbon is
consumed. Add a few drops of nitric acid and dry in the radiator. Ignite the
covered crucible for 5 minutes in the full heat ot a Bunsen burner, cool, and
weigh. Ignitions may be made at about 750° C. in an electric muffle. As
WO a is slowly but steadily volatilized at temperatures above 750 0 Q.,.serious
error may result from continued ignition of the tungsten trioxide at the tem-
perature that is specified. Add about 5 g. of Na 2CO a and fuse, running the
fusion around the side of the crucible to remove all WO a• Dissolve the fusion
in hot water, add alcohol, heat, filter and wash with hot water. Place the filter
in the crucible and ignite; add a little N a 2 CO a and fuse again. Dissolve the
fusion in water, filter and wash very thoroughly with hot water to remove the
last traces of Na 2CO a ; ignite in the sam~ crucible as at first, cool, anD weigh.
The difference between the weight of crucible plus residue and the weight of
crucible plus impure tungstic oxide is WO a XO.7931 = W.' -
NOTES.-At the best, direct determination of tungsten in high-grade metal is subject
to inherent errors. Accuracy within 0.2% is all that can be expected by ariy method.
DETERMINATION OF CARBON
SILICON·
The residue of tungstic oxide from which the solution for the determination
of sulfur was filtered contains most of the silica. The residue from leaching
the original fusion contains the rest.
Wash the residue from Na 2 CO a fusion from the filter paper with a fine jet
of water into a small casserole, add an excess of HCI and evaporate to dryness
0
on a steam-bath, heat the dry residue at 110° to 120 C. to dehydrate the silica,
drench with hydrochloric acid, dilute to 100 ml. with hot water, heat until
salts are in solution, and filter on a small filter paper (free from ash). Wash
thoroughly with hot dilute HCI (1 : 100) and finally with hot water. Reserve
the filter.
Wash the residue of tungstic oxide, silica and filter pape~ pulp from the
filter into a small casserole, add 5 ml. of HCI and heat for a few minutes, then
filter again through the same paper and wash it thoroughly with hot water.
This second washing is necessary to remove sodium salts completely. (The
original washing sufficed to remove all of the small amount of sulfur present.)
Transfer the well-washed filter and residue to a platinum crucible, add the
small filter containing the little silica recovered from the N a 2CO a fusion residue
and ignite both to constant weight in a platinum crucible. On account of
the presence of filter paper pulp, the ignited residue is porous and friable; if
paper pulp had not been used the WO a would be dense and not easily susceptible
to the subsequent treatment. Cool the ignited residue of WO a and Si0 2 and
weigh it. Add 2 or 3 drops of H 2S0 4 and 5 or 6 ml. of pure HF. Digest at a
gentle heat for some time, and then slowly evaporate off the HF. When
H 2S0 4 fumes are evolved, cool and again add HF, digest as before, evaporate
off the HF, ignite gently to expel H 2S0 4 and heat strongly for ten or fifteen
minutes. Cool and weigh. The loss in weight after expulsion is Si0 2 which
multiplied by 0.4672 equals silicon.
FERROTUNGSTEN
DETERMINATION OF TUNGSTEN
•
Proceed as in the Determination of Tungsten in the analysis of Ferrotungs-
ten and Tungsten MetalY
NOTES.-l. Fuming with perchloric acid can be substituted for fuming. with sulfuric
acid if desired. In this case, add 10 ml. of perchloric acid (60%) after the treatment
with HF and HNO a and evaporate to strong fumes of perchloric acid. Cool, add 50 ml.
of water, transfer the solution to a 400-mI. beaker and proceed as in the case of sulfuric
acid.
2. If the alloy contains molybdenum, the weighed tungsten oxide may contain a
few hundredths of a per cent of molybdenum as MoO a• If extreme accuracy is desired,
the oxide should be tested colorimetrically. If the molybdenum content of the alloy
17 See p. 1499.
1504, FERRO-ALLOYS
is unknown, one of the sodium carbonate'extracts of the fused impure WO s can be treated
with tartaric acid, mineral acid and hydrogen sulfide and then, if appreciable molyb-
denum is indicated, the amount shollid be determined inl the sodium carbonate extract
obtained in the duplicate run. •
3, At the best, direct determinations of tungsten in .p.igh-grade metal is subject to
inherent errors. A_ccuracy within 0.2% is all that can bel expected by any method:
DETERMINATION OF SILICdN
DETERMINATION OF SULFUR
in 80 mI. of hot dilute HCl (1 : 1) and wash the filter with hot water, collecting
the solution (not over 150 m!. in volume) in a 300-ml:. Erlenmeyer flask. Add 1
to 2 g. of tes~ lead and close the flask
with a one-hole rubber stopper carrying
a bent deli\fery tube, as slJ.own· in
Glass Rod Fig. 186. I
Boil gently for 20 minutes. At the
end of this period immerse the end" of
6- the delivery tube in a small beaker"con-
taining about 50 m!. of a strong solution
of NaHCO a and insert the rubber-tipped
tube at C, Fig. 186. Remove the flask
and beaker from the hot plate ant! cool
in a stream of cold wate~. When 'cool,
remove the stopper, add 5 m!. of freshly
prepared starch solution and titrate
FIG. 186. Apparatus for Reduction with the standard solution of iodine to
of Tin.
a permanent blue tint. Subtract the
volume required in a blank run and multiply by the tin titer of the solution.
(e) Determination of Antimony.-Spread the reserved filter paper and pre-
cipitate (B) side down, on a small watch glass and place this over a beaker
containing a few centimeters of HCl and a' drop or two of bromine. Warm the
beaker until decomposition of the sulfide is complete. Rinse cover and paper
with warm dilute HCI (1 : 9), keeping the volume down·to 25 m!. Evaporate
the solution to about 5 m!. on the steam bath and then add 1 g. of pure ferrous
sulfate, 0.5 g. of potassium chlori~, and 40 ml. of HCI. Evaporate the solution
on the steam bath to a volume of about 5 ml. Dilute with 10 ml. of water,
add 1 m!. of HNO a and heat until the iron is oxidized. Dilute to 100 mI. heat
to boiling, and add dilute NH 40H until in definite excess. Let settle, filter,
and wash with warm water. Dissolve the precipitate in 10 to 15 ml. of hot
dilute HCl (1 : 3) containing 0.1 g. of tartaric aci4~ Dilute to 100 mI., heat
and add ammonium bisulfite drop-by-drop to reduce the iron. Boil to expel
the excess of S02, cool and pass in a rapid stream of H 2S for 15 to 30 minutes.
Filter through a weighed Gooch crucible and wash,successively with acidulated
hlYdrogen sulfide water, alcohol, ether and carbon disulfide. Heat in a radiator
and add 'a drop of HNO a, from time to time to oxidize the sulfide. Finally
heat at about 600 0 C. and weigh as Sb 20 4 • '
FERRO MOLYBDENUM
DETERMINATION OF CARBON BY THE DrRECT-COMBUSTION METHOD
DETERMINATION OF MOLYBDENUM
METHOD
'Transfer 0..5 g. of sample to a 15o.-ml. beaker, cover, treat with 10. ml. .of
dilute HNO a (1 : 3), and warm to complete solution. If solution is difficult,
add a drop or two of HF. When reaction is complete, cautiously add 10. m!. of
dilute H 2S0 4 (1 : 1), and evaporate to fumes of the acid. Cool, add 40 m!. of
water and warm. Rinse and remove the cover. Filter and wash the paper
and silica 12 to 15 times with hot water, 3 to 4 times with hot dilute NH 40H
\ 1 : 3), and finally 4 to,,5 times with hot water, the washings being allowed to
run into the main filtrate. Discard the silica. Cool, add NH 40H until it is
difficult to avoid a red tint, and heat the still acid solution to boiling. Pour
the solution slowly and with vigorous stirring into 75 ml. of nearly boiling
dilute NH 40H (1 : 5) contained in a 5o.o.-ml. beaker. Rinse the beaker with
a little water, and then with a little hot dilute NH 40H and add to the main
solution. Add a little paper· pulp, filter into a 5o.O-m!. beaker, and wash the
residue with hot water. Set the filtrate aside. Dissolve~the precipitate in a
slight excess of hot dilute H 2S0 4 (1 : 4), and riearly neutralize and pour into
NH 40H as before. Filter into the reserved filtrate. Dissolve the precipitate
and repeat the operation.
Treat the three combined ammoniacal filtrates ,"-1th 3 g. of powdered
tartaric acid. Stir until dissolved, and saturate the warm solution with H 2S.
If a precipitate appears, filter and wash with a dilute solution of (NH 4 )2S.
22 See p. 1490.
1508 FERRO-ALLOYS
Warm the filtrate, cover the beaker, arid add dilute 'H 2S0 4 (1 : 1) until the
solution contains 10 IIil. in excess for each 100 ml. Oflsolution. Let the solution
stand at the side of the steam bath (about 40°' C.) for 15 minutes, and let Iilettle.
Filter, and wash the sulfide thoroughly with dilut~ H 2S0 4 (1 : 99) containing
10 g. of tartaric acid per liter and saturated with II'2S.23 .
Transfer the MoS a and paper to a 250-ml. beaker,. cover, and treat with (i) ml.
of H 2S0 4 and 10 mL of HNO a• Cautiously heat 'until fumes of H S0 4 are
t
given off. Let cool, add 5 ml. of HNO a, again heat to fumes, and repeat the
treatment until the yellow color due to organic matter has disappeared. Cool,
rinse and remove the cover, rinse the inside of the beaker, and add a solution
of KMn04 (2.5%) very cautiously until ~ permanent red tint is obtained.
Again evaporate to fumes of H 2S0 4. Cool and add 75 ml. of water.' Boil
for a few minutes, add 2 g. of pure zinc shot (0.002% o~ iron or under), and
continue the boiling until any copper has been reduced to metal. Filter
through a filtering crucible and wash with hot water.
If the Jones reductor has been standing idle, pass 100 ml. of warm (40 to
50° C.) dilute H 2S0 4 (5 : 95) through it, then a little water, and discard the
solution. Add 35 ml. of the phosphate solution to the receiver, and then
enough water so that the tip of the reductor dips well beneath the solution
when the receiver is connected with the reductor. Draw the cool or slightly
warm solution of the molybdenum, which should be about 100 ml. in volume
and contain about 5 ml. of H 2S0 4, through the reductor as the receiving flask
is given slight rotary motion. Just before the surface of the liquid reach-es the
zinc, add a 50-ml. portion of water, and finally rinse by adding two more por-
tions, each time just before the surface of the solution reaches the zinc. Close
the stopcock while a portion of the last washing solution remains in the reductor
funnel. Disconnect, and raise the reductor as a little water is allowed to run
through the stem, rinse the outside of the stem, and proceed with the titration
with the standard 0.1 N solution of KMn04. Run a blank, subtract the volume
of 0.1 N solution required, and multiply the remainder by the molybdenum
titer of the solution.
DETERMINATION OF' SULFUR
DETERMINATION OF PHOSPHORUS
DETERMINATION OF COPPER
The methods ordinarily used for examining bituminous substances and their
mixtures may be grouped under four headings, viz.: .
(1) Physical characteristics. (2) Thermal tests.
(3) Solubility tests. (4) Chemical methods of examin~tion.
TABLE I
-
For Adapti- Gauging Purposes ABa
Purposes bility for the Uni- of Criterion
Description of Identi- a Given formity Factory of the
fication Purpose of Supply. Control Quality*
Physical Characteristics:
Fracture ...................... YES 0 .... . ... . ... . ...
Streak ........................ YES Yes . '" 0 ••• ....
Specific gravity ................ YES .... Yes Yes . ...
Viscosity ............ , ......... .... YES Yes Yes . ...
Penetration (Hardness) ......... YES YES YES YES
Ductility ....................... .... YES Yes Yes · '"
Thermal Tests: ,
Fusing or softening-point ....... YES YES YES YES ....
Volatile matter ................ Yes YES r Yes Yes ....
Evaporation test ............... Yes YES Yes Yes ....
Flash-point .................... .... YES Yes Yes . ...
Fixed carbon .................. YES 0:" • . .. , . ...
Distillation test ................ Yes YES Yes Yes YES
Solubility Tests:
Solubility in carbon disulphide ... YES YES YES ... - YES
Carbenes ............. \ ........ Yes ... , Yes Yes YES
Solubility in petroleum naphtha .. YES Yes Yes Yes · ...
Free carbon in tars ............. YES Yes Yes - . '" ....
Chemical Tests: ,
Water ........................ Yes .... 0 ••• . ... YES
Oxygen in non-mineral matter .... YES . .. , . ... .... . ...
Solid paraffins ....... ' j ' • • • • • • • • YES .... i '" . . .. , . ...
Sulfonation residue ............ YES -0- . . · ...
y~~
'"
Saponifiable matter ............ YES .... 0 ••• YES
Diazo reaction ................. YES .... . ... . ... . ...
AnthraqUinone reaction ......... YES .... 0 ••• .... . ...
I
* (a) Purity; (b) Care exerciSed in its preparatipo.; ("1;) _Intrinsic value.
1 By Herbert Abraham. .
1510
-"
EXAMINATION OF BITUMINOUS SUBSTANCES 1511
uotlO'BaU ) 0 0 ) 0 0 0 ) oooco) oooco):Bo) :B!'l:B) ~lS~!o
ouolI!rib~JqtUV
000)
ZZZ
00)
ZZ zz zzz zzzzz zzzzz ~z ~~~ ><><~~z
nQMoualI
" I
'zzz
O."la Izz Izz I zz:z; Izzzzz Izzzzz Izz I III III III I :J!S gs !
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
~~~ ~~><~~
Ii
~
t!l "";= ~
,S! j ,,:B
.&
--a "'a
j
'.::l ,.ij ~ .o~
J~rJ
0
£~
S
:0
~
"
j 15>< .e1 1" i:g ,.!l"" f-<
~ !:.il £~
z~~
0"
~-< -< <2 ,!,
Z"" !:.:;::
Il< &!~ "
0
Q
11 ~1Il
:0-
11 0"
'"" i1:a
d
,J
1"1 f °a
~
1512 EXAMINATION OF BITUMINOUS SUBSTANCES
In general, these tests may have one or more of the following objects in view:
(a) To serve as a means of identification.
(b) To ascertain the value of the substance for a given purpose.
(c) To gauge its uniformity of supply.
(d) As an aid to factory control in its preparation, refining or blending.
(e) As a criterion of its quality.
The most important methods of testing are given in Table I.
Table II gives a list of the principal bituminous substances, together with
such of their physical and chemical characteristics as will enable them to be
distinguished, one from another.
PHYSICAL CHARACTERISTICS
Fracture
This is ascertained upon cleaving the specimen by subjecting it to a sharp
blow, and examining the cleavage surface. Only hard and "brittle" bitum-
inous substances will yield to this test, including the hard asphalts and asphal-
tites. The fracture may either appear conchoidal (rounded and curved like a
shell), or hackly (jagged, irregularly and rough).
Streak
This represents the color of the powder which is left behind on drawing a
piece of the solid bituminous material across the surface of unglazed porcelain.
Hard bituminous materials only will yield to this test. 'The streak may be
classified as white (where no streak is visible), yellowish, yellowish brown,
brown, brownish black and black.
Specific Gravity
Hydrometer Method.-Where speed is essential and great accuracy not
required, the specific gravity of fluid bituminous materials may be determined
with a hydrometer having its scale sub-divided to unity in the third place of
decimals. Usually a series of hydrometers are used, ranging respectively from
0.800 to 0.900, 0.900 to 1.000, 1.000 to 1.080, 1.070 to 1.150, 1.150 to 1.230, and
in such dimensions as to enable them to be used in a 100 ml. cylinder ap-
proximately 300 mm. long having an inside diameter of not less than 32 mm. as
illustrated in Fig. 187. The hydrometer shall conform to the following re-
quirements as to dimensions:
Permissible
Dimension Variation
Length of stem ....... , ....... . 125 mm. 6 mm.
Length of bulb ............... . 105 mm. 5 mm.
Length of scale .............. . 80mm. 4 mm.
Diameter of stem ............. . 6mm. O.5mm.
Diameter of bulb ............. . 22mm. 1 mm.
1514 EXAMINATION OF BITUMINOUS. SUBSTANCES
Most hydrometers are adapted to read at 60° F./60° F., or in other words,
the instruments are calibrated for water at 60° F. taken as unity. The standard
temperature for' testing bituminous materials is 77° F., and they, should ac-
cordingly be brought to this temperature when tested with the hydrometer.
For correcting the reading to water at 77° F., it shoJld be multiplied by 1.002,
as follows: I'
····
.
100·-
101
..
6mm.
-1:- --.
1I ;!
£
I:
102
103
..
..
..
,I
~
104
lOS
. l
.j
I
106 :
107
i
--~-
~
I:
(;)
·
(;)
I
1<\
-{---
3~"
!
I:
_.i.
I
FIG. 187.-Hydrometer for Liquid Bitumir!ous Substances.
0B e 140
aum =Sp.gr. 600/600 F. -130
SpeCl'fi c GravltY=-b-
. c-a
-a
1516 EXAMINATION OF BITUMINOUS SUBSTANCES
When determining the specific gravity of tar and asphalt products which are
too viscous for the method described in the preceding p,aragraph, a small amount
of the material shall be brought to a fluid condition by the gentle application of
heat, care being exercised to prevent loss by evaporation. When sufficiently
fluid, enough is poured into the clean dry pykometer to about half fill it. Pre-
cautions shall be taken to keep the material from touching the sides of the tube
above the final level and to prevent the inclusion of air bubbles. The tube
should be slightly warmed before filling. The pyknometer and contents are
then cooled to room temperature and weighed with the stopper. This weight
is called c. The pyknometer is next removed from the balance, filled with
freshly boiled distilled water, and the stopper firmly inserted. It is then com-
pletely immersed for not less than thirty minutes in a beaker of distilled water
maintained at 25° C. (77° F.) after which it is removed, and all surplus water is
wiped off with a clean cloth. It is immediately weighed. This weight is
called d. The specific gravity of the material shall be calculated from the
formula:
'c-a
Specific Gravity
(b-a)-(d-c)
When making the specific gravity determination it is important that:
(a) Only freshly boiled distilled water f?hall be used.
(b) When weighing the pyknometer completely filled, the temperature of
its contents shall be within 10 C. (1.8° F;) of 25° C. (77° F.).
(c) Precautions shall be taken to prevent expansion and overflow of the
contents from the heat of the hand when wiping the surface of the pyknometer.
(d) The presence of all air bubbles shall be eliminated in filling the pyknom-
eter and inserting the stopper.
(e) Weighings sh~ll be made quickly after filling the pyknometer and shall
be accurate to 1 mg. A number of trial fillings and catch weights may be
necessary to obtain the desired degree of accuracy. .
(f) To prevent breakage of the pyknometer when cleaning it out 'after a
determination has been made upon a very viscous 'or semi-solid material, it
will be found advisable to warm it in an oven a't not over 100° C. until most of
the material may be poured out and then to swab it with a piece of soft cloth
or cotton waste. When cool it may be finally rinsed with carbon dis.ulfide,
benzol or other solvent and wiped clean.
The limit of accuracy of the test is ±0.005 specific gravity.
(B) The specific graVity of creosote oil fractions is determined in a similar
manner by means of a pyknometer, but in this case the results are expressed
at 38° C. (100° F.)/15.5° C. (60° F.).
(0) The pyknometer method may also be used for finding the specific
gravity of hard and brittle bituminous substances, including hard asphalts of
high fusing-point, asphaltites, asphaltic pyrobitumens, non-asphaltic pyro-
bitumens and pyrobituminous shales. Approximately 3.5 grams of the material
ground to 60-mesh are carefully weighed and introduced into a 50-m!. pyknom-
eter, with about 30 m!. of distilled water. A vertical condensing bulb is at-
tached to the pyknometer with a small section of rubber tubing, the open end
being connected with an aspirator to maintain.a partial vacuum. The pyknom-
eter is then boiled on a water-bath to expel all·tIle 'air from the sample. The
EXAMINATION OF BITUMINOUS SUBSTANCES 1517
inside of the" condensing tube is then washed back into the pyknometer, which,
is cooled to the desired temperature, stoppered, filled to the mark with water
a,t the same temperature and weighed. The specific gravity may then be
calculated from the foregoing formula. -
Analytical Balance Method.-This method is adapted for substances which
can be melted and cast in a mold to form a briquette which may be handled on
cooling. It has be_en standardized as follows: .
The determination of specific gravity shall be made with an analytical
balance equipped with a pan-straddle or other stationary s~pport (Fig. 189).
adhering air bubbles being first removed with a fine wire. This weight is
called b and shall be accurate to 0.1 mg. The specific gravity of the material
shall be calculated from the formula:
Specific Gravity = a ~b I
I
The limit of accuracy of the test is ±0.005 specific gravity.
Viscosity
Engler Method.-This method as described on p. 1-718 is adapted for examin-
ing liquid or semi-liquid bituminous substances.
Saybolt Method.-The details of this method will be found on p. 1719.
Float Test.-This instrument is used largely for testing the viscosity or consist-
encyof semi-solid bituminous materials. The range of the float test is limited,
and it cannot be used with very fluid bituminous materials or with hard solids. It
accordingly fills the gap between the Engler viscosimeter on one hand, and the
needle penetrometer on the other. The test is not affected by the presence of
finely divided mineral matter or free carbon. '
The procedure has been standardized as follows:
"~61,-
> ".
'{~1a
I
~n.;C\e
Weigh/"'- ::!:;_<S
, fll2JZI .
UO~()'IO j "'/0'1
mm. Mm.
Colla ..
(Brass)
W,;ght qf Fi.af, :rl.90~ 0.20,.
WeigH qf C4/hlj 9.I/J t ().2()g.
'SfandanJ
Supporl
,p....mti!.)'.
FIG. 190.-Float Test Apparatus.
ExAMINATION OF BITUMINOUS SUBSTANCES 1519
The Hoat (Fig. 190) shall be made of aluminium or aluminium alloy and
shall be in accordance with the following requirements:
The collar shall be made of brass and shall be In accordance with the
following requirements:
The top of the collar shall screw up tightly against the lower side' of the
shoulder.
The assembled Hoat and collar:, with the collar filled flush with the bottom
and weighted to a total weight of 53.2 g., shall Hoat upon water with the rim
8.5 ±1.5 mm. above the surface of the water. Dimensions of the apparatus
additional to 'those required above are given in Fig. 190. The thJrmometer
shall be graduated in either Centigrade or Fahrenheit degrees as specified, the
ranges being _2 0 to +80 0 C. or +30 0 to + 1800 F., respectively. The diam-.
eter of the bath and the depth of water shall be' at least 185 mm.
The brass collar shall be placed with the smaller end on a brass plate which
. has been previously amalgamated with mercury by first rubbing it with a dilute
solution of mercuric chloride or nitrate, and then with mercury.
The sample shall be completely melted at the lowef,t possible temperature
that will bring it to a sufficiently fluid condition for pouring, excepting creosote
oil residues, which shall be mixed and poured at a temperature of 1000 to 1250 C.
It shall be stirred thoroughly until it is homogeneous and free from air bubbles.
The sample shall then be poured into the collar in any convenient manner until
slightly more than level with the top. -
Asphalt and Asphalt Products.-Asphalt and asphalt products shall be
cooled to room temperature, placed in water maintained .at 50 C. for five
minutes, after which the surplus material shall be removed by means of a
sp~tula, or steel knife, which has been slightly heated. The collar and plate
shall then be placed in a tin cup containing ice water maintained at 5° C.,
±1° C., and left in this bath for at least fifteen minutes.
Tar Producis.-Tar products shall be immediately immersed in ice water
maintained at 5c C. for five minutes, after which the surplus material shall be
removed by means of a spatula or steel knife, which has been slightly heated.
1520 EXAMINATION OF BITUMINOUS SUBSTANCES
The collar and plate shall then be placed in a tin cup containing ice water
maintained at 5° C., ± 1° C., and. left in this bath for at least fifteen mihutes.
The bath shall be filled with water and the water heated to the temperature
at which the test is to be made. This temperature shall be accurately main-
tained and shall at no time throughout the test be allowed to vary more than
0.5° C. from the temperature specified. .
After the material to be tested has been kept in the ice water for not less
than fifteen minutes nor more than thirty minutes, the collar with its contents
shall be removed from the plate and screwed into the a:luminium fioat and
immersed in water at 5° C. for one minute. Any water shall then be removed
from the inside of the fioat and the latter immediately floated in the warm batli.
As the plug of material becomes warm and fluid, it is forced upward and out of
the collar until the water gains entrance into the saucer and causes it to sink.
The time in seconds between placing the apparatus on·the water and when
the water breaks through the material shall be determined by means of a' stop
watch, and shall be taken as a measure of the consistency of the material under
examination.
NOTE.-Speciai precautions should be taken to insure the collar fitting tightly into the
float and to see that there is no seepage of water between the collar and float during the
~~ .
Penetration (Hardness)
Penetration is defined as the consistency of a bituminous material, expressed
as the distance that a standard needle vertically penetrates a sample of the
material under known conditions of loading, time and temperature. Where
the conditions of test are not specifically mentioned, the load, time and tem-
perature are understood to be 100 g., 5 seconds, 25° C. (77° F.), respectively,
and the units of penetration to indicate hundredths of a centimeter. •
The container for holding the material to be tested shall be a fiat-bottom,
cylindrical metal or glass dish, 55 mm. (2 136 in.) in diameter and 35 mm. (1%
in.) deep. The needle (Fig. 191) for this test shall ~e made from a cylindrical
I
(1.C: Zoe mm. . s'4q' to 9~""""'~.. . ()J4 ~OJ6m,'
'-.., I ae To' ~
T b" . ' I1<···6.35mm..J
UpplW.
steel rod approximately 50.8 mm. (2 in.) long, and having a diameter of 1.00
to 1.02 mm. This shall be symmetrically tapered at one end to a cO,ne ap-
proximately 6.35 mm. (7i in.) in height and whotle angle shall be within the
range if 8° 40' and 9° 40'. After tapering, the point shall be "blunted" by
grinding off to a truncated cone, the smaller base of which shall be from 0:14
to 0.16 mm. in diameter. The finished needle shall be hardened and highly
polished. .
The water bath shall be maintained at a temperature not varying more than
0.1° C. from 25° C. (77° F.). The volume of _WJl.ter shall be not less than 10
liters and the sample shall be immersed to a depth of not less than 10 cm.
EXAMINATION OF BITUMINOUS SUBSTANCES 1521
(4 in.) and shall be supported on a perforated shelf not less than 5 cm. (2·in.)
from the bottom of the bath. Any apparatus which will allow the needle to
penetrate without appreciable friction, and which is accurately calibrat~d to
yield results in accordance with the definition of penetration, will be acceptable.
The transfer dish for container shall be a small dish or tray of such capacity
as will insure complete immersion of the container during the test. It shall
be provided with some means which will insure a firm bearing and prevent
rocking of the container.
The sample shall be completely melted at the lowest possible temperature
and stirred thoroughly until it is homogeneous and free from air bubbles. It
shall then be poured into the sample container to a depth of not less than 15 mm.
(% in.). The sample shall be protected from dust and allowed to cool in an
atmosphere not lower than 18° C. (65° F.) for one hour. It shall then be
placed in the water bath along with the transfer dish and allowed to remain
one hour.
In making the test, the sample shall be placed in the transfer dish filled
with .water from the water bath of sufficient depth to completely cover the
container. The transfer dish containing the sample shall then be placed upon
the stand of the penetration machine. The needle loaded with specifieli weight
shall be adjusted to make contact with the surface of the sample. This may
be accomplished by making contact of the actual needle-point with its image
reflected by the surface of the sample from a properly placed source of light.
Either the reading of the dial shall then be noted or the needle broyght to zero.
The needle is then released for the specified period of time, after which the
penetration machine is adjusted to,measure the distance penetrated .
.At least three tests shall be made at points on the surface o(the sample not
less than 1 cm. (% in.) from the side of the container and not less than 1 cm.
(% in.) apart. After each test the sample and transfer dish shall be returned
to the water bath and the needle shall be carefully wiped toward its point with
a clean, dry cloth to remove all adhering bitumen. The reported penetration
shall be the average of at least three tests whose values shall not differ more than
four points between maximum and minimum. .
When desirable to vary the temperature, time and weight, and, in order to
provide for a uniform method of reporting results when variations are made, the
samples shall be melted and cooled in air as above directed. They shall then
be immersed in water or brine, as the case may require, for one hour at the tem-
perature desired. The following combination!! are suggested:
At 0° C. (32° F.) 200-g. weight, 60 seconds.
At 46.1° C. (115° F.) 50-g. weight, 5 seconqs.
Ductility
This test has been standardized as follows:
The ductility of a bituminous material is measured by the distance to which
it will elongate before breaking when two ends of a briquet of the material of
the form described, are pulled apart at a specified speed and at a specified
temperature. Unless otherwise specified, the test shall be made at a tempera-
ture of 25° C.±O.5° C. (77° F.±O.9° F.) and with a speed of 5 cm. per minute
(±5.0%).
1522 EXAMINATION OF BITUMINOUS SUBSTANCES
The mold shall be similar in design to that shown in Fig. 192. Dimensions
shalL be as given with the permissible variations indicated. The mold 'shall
be made of brass, the ends, band b' , being known as clips, and the parts, a and
, \
1<-- --.: -1/.9cmtO.lcm. -- - - _", • I .
I
f.
I -.,_-
,
>·.J.IS~
J.25,,,,,
a', as sides of the mold. The dimensions of. the mold shall be such that, when
properly assembled, it will form a briquet having the following dimensions:
Total length ................................... 7.45 to 7.55 cm.
.
Distance between clips ........... :. . . . . . . . . . . . .. 2.97 to 3.03 cm.
, .... Width at mouth of clip ...... " ................. 1.98 to 2.02 cm.
Width at minimum cross-section (halfway between
clips) ..................................... 0.99 to 1.01 cm.
Thickness throughout. . . . . . . . . . . . . . . . . . . . . . . . .. 0.99 to 1.01 cm.
The water bath shall be maintained at the specified test temperature ;varying
not more than 0.10 C. (0.18 0 F.) from this temperature. The volume of water
shall be not less than 16' liters and the sample shall be immersed to a depth of
not le:;;s than 10 cm. and shall be supported on a per~orated shelf not less than
5 cm. from the bottom of the bath. I
For pulling the briquet of bituminous material apart, any apparatus may
be used that is so constructed that the briquet will be continuously immersed in
water as specified, while the two clips are pulled apart at a uniform speed, as
specified, without undue vibration.
The bituminous material to be tested shall be completely melted until
thoroughly fluid by heating it· in an oil bath maintained at- the minimum
temperature needed to properly liquefy the sample. It _shall then be
strained through a No. 50 sieve and, after a thorough stirring, poured into the
mold. The mold shall be assembled on a brass plate and, so as to prevent the
material under test from sticking, the surface of the plate and interior surfaces
of the sides aa' , of the mold shall be thoroughly amalgamated.~ The plate
upon which the mold is placed shall be perfectly flat and level so that the bottom
surface of the mold will touch it throughout. In filling the mold, care shall be
taken not to disarrange the parts and thus distort the briquet. In filling, the
2 The amalgamation may best be effected by immersing the clean mold in a solution
of mercury bisuIphate containing free metallic mercury, .a'p'd 'so as to come in contact with
the latter. Instead of mercury, the metal mold, preferably of stainless steel, may be
moistened with glycerol.
EXAMINATION OF BITUMINOUS SUBSTANCES· 1523
material shall be poured in a thin stream back and forth from end to end of
the mold until it is more than level fiull. It shall be left to cool to room tem-
perature and then placed in the water bath maintained at the specified tem-
perature of test for 30 minutes, after which the excess bitumen shall be cu't off
by means of a hot straight-edged putty knife or spatula so that the mold shall
be just level full.
-
NOTE.-When paving asphalt cements are being tested, the oil bath shall be main-
tained at a temperature of from 125 to 1500 C. (257 to 3020 F.).
The brass plate and mold, with briquet, shall then be placed in the water
bath and kept at the specified temperature for at least 1,Y2 hours, when the'
. briquet shall be removed from the plate, the side pieces detached, and the
·briquet immediately tested. The rings.,at each end of the clips shall be at-
tached to the pins or hooks in the ductility machine and the two clips pulled
apart at a uniform speed as specified until the briquet ruptures. A variation
of ±5% from the speed specified will be allowed. The distance through which
the clips have been pulled to produce rupture shall then be measured in centi-
meters. While the test is being made, the water in the tank of the ductility
machine shal! cover the sample both above and below it by at least 2.5 cm. and
shall be kept continuously at the temperature specified within ±0.5° C.
(±O.9° F.).
A normal test is one in which the material between the two clips pulls out
to a point or thread until rupture occurs at the point where the thre'ad has
practically no cross-sectional area. The average of three normal tests shall be
taken and reported as the ductility of the sample.
If the bituminous material comes in contact with the surface of the water
or the bottom of the bath, the test shall not be considered normal.
NOTE,-When :the specific gravity of the bituminous material to be tested is below
0.98 or above 1.01, the specific gravity of the water bath in the ductility machine shall
be made the same as the material to be tested by the addition of either methyl alcohol
or sodium chloride.
If a normal test is not obtainable on three successive tests, the ductility
shall be reported as being unobtainable under th~ conditions of the test.
THERMAL TESTS
Fusing or Softening-Point
Ring-and-Ball Method.-This has been standardized as follows:
The softening of bituminous materials generally takes place at no definite
moment or temperature. As the temperature rises, they gradually and im-
perceptibly change from a brittle or exceedingly thick and slow-flowing material
to a softer and less viscous liquid. For this reason the determination of the
1524 EXAMINATION OF BITUMINOUS SUBSTANCES
No,13'8'S Gag"
Brass Wire.· ...
-~
"
...,i..IIIIIIL'._.....:..t..J.W......
·r···· f --..~
, 6r••$ Rin9 .nd Wire.
Proper Po.ilion,
Assembly
of B,an.
of Apparatus.
FIG. 193.-Apparatus for Ring-and-Ball 'Method,
(b) A steel ball 9.53 mm. (Ys in.) in' diameter weighing between 3A5 and
3.55 g. .
(c) A glass vessel, capable of being heated, not less than 8.5 cm. (3.34 in.)
in diameter and measuring 10.5 cm. (4.13 in.) in depth from-the bottom of the
flare. (A BOO-m!' beaker, low form, meets this requirement.) "-.
The sample shall be melted at the lowest possible temperature to avoid loss
of volatile constituents and stirred thoroughly, avoiding incorporating air
bubbles in the mass, and then poured into the ring so as to leave an excess on
cooling. The ring, while being filled, should rest on a brass plate which has
been amalgamated to prevent the bituminous material from adhering to it.
After cooling, the excess material shall be cut off cleanly with a slightly heated
knife.
FaT Substances Fusing at 80 0 C. (1760 F.) aT Below.- U se a thermometer which
shall be graduated in ~ither Centigrade or Fahrenheit degrees as may be specified,
the ranges being -2 to +80 0 C., or 30 to 180 0 JT.,..,respectjvely. Fill the glass
vessel to a depth of substantially 8.25 cm. (3.25 in:) with freshly boiled, dis·
EXAMINATION OF BITUMINOUS SUBSTANCES 1525
tilled water at 50 C. (41 0 F.). Suspend the ring containing the sample in the
water so that the lower surface of the filled ring is exactly 2.54 cm. (1 in.)
above the bottom of the glass vessel and its upper surface is 5.08 cm. (2 in.)
below the surface of water. Place the ball in the water but not on the specimen.
Suspend the thermometer so that the bottom of the bulb is level with the
.bottom of the ring and within 0.635 cm. (U in.), but not touching the ring.
Maintain'the temperature of the water at 5° C. (41° F.) for fifteen minutes. 3
With suitable forceps, place the ball in the center of the upper surface of the
bitumen in the ring, thus completing the 'assembly as in Fig. 193. Apply the
heat in such a manner that the temperature of the water is raised 5° C. (9 0 F.)
each minute. The temperature recorded by the thermometer at the instant
the bituminous material touches the bottom of the glass vessel shall be reported
as the softening point. No correction shall be made for emergent stem. The
rate of rise of temperature shall be uniform and shall not be averaged over the
period of the test. The maximum permissible variation for any minute period
after the first three shall be ±0.5° C. (0.9 0 F.). All tests in which the rate of
rise in .temperature exceeds these limits shall be rejected.
FOT Substances Fusing above 80 0 C. (176 0 F.).-The thermometer shall
be graduated in either Centigrade or Fahrenheit degrees as specified, the ranges
being 30 to 1600 C., or 85 to 320 0 F., respectively. The same method as given
above shall be employed, except that U. S. P. glycerin shall be used instead of
water, and the starting point of the glycerin bath shall be 320 C. (89.60 F.).
The bath shall be brought to this temperature and thoroughly agitated, 'then
the apparatus and specimens shall be placed in the bath which shall be main-
tained under agitation at the starting temperature for fifteen minutes, after
. which the assembly shill be completed. by placing the ball on the center of the
specimen and the test carried on as above. In applying the heat, the ring
apparatus shall be placed off the center of the container and the burner placed
midway between the center and edge of the beaker away from the specimen.
Rigid adherence to the prescribed rate of heating is absolutely essential in
order to secure accuracy of results: A sheet of paper placed on the bottom of
the glass vessel and conveniently weighted will prevent the bituminous material
from sticking to the glass vessel, thereby saving considerable time and trouble
in cleaning. The limit of accuracy of the test is ±0.5° C. (0.9 0 F.).
Cube Method.-This method is restricted to testing tar-pitches, and has
been standardized as follows:
The softening of pitch takes place at no definite moment or temperature.
As the temperature rises, pitch gradually and imperceptibly changes from a
brittle or exceedingly thick and slow-flowing material to a softer and less
viscous liquid. For this reason the determination of the softening-point must
be made by a fixed, arbitrary and closely defined method if the results obtained
are to be comparable.
The apparatus shall consist of the following:
(a) A mold suitable for forming a 12.7-mm. (Y2-in.) cube of pitch. (A
recommended,type is shown in Fig. 194).
3 NOTE BY AUTHOR.-It will be found more convenient, and the accuracy of test will
in no way be sacrificed, if the initial temperature of the water is maintained 25° F. below
the softening point of the substance under examination.
1526 EXAMINATION OF BITUMINOUS SUBSTANCES
Slandani
Svpporf..
\.
SIt/e. Front
tflItR"""·---·····
The pitch shall be formed into a 12.7-mm. (Y2'-in.) cube, truly shaped and
with sharp edges, either by melting and pouring, or softe~ing and pressing,
into a mold. In all cases an excess of pitch shall be used and the surplus mate-
rial shall be cut off cleanly with a slightly heated knife. The harder pitches
specified can ordinarily be molded at room temperature, the softer pitches in
water at about 4° C. (39.2° F.). If they ar~_nmlted,. they should first be
thoroughly stirred, avoiding incorporating air bUbbles in the mass, and then
EXAMINATION OF BITUMINOUS SUBSTANCES 1527
poured into the mold so as to leave an excess on cooling. The mold should
rest on a brass plate and the surface of the plate and the interior surfaces of the
mold should be amalgamated to prevent the pitch from adhering to them.
For Pitches Fusing between 110 and 170° F.-Assemble the apparatus as
shown in Fig. 194. Fill the glass vessel to a depth of substantially 9.5 cm.
(3.75 in.) with freshly boiled, distilled water at 15.5° C. (60° F.). Place the
cube of pitch on the wire as shown in Fig. 194 and suspend it in the water so
that its lower edge is exactly 2.54 cm. (1 in.) above the bottom of the glass
vessel and its upper edge is 5.08 cm. (2 in.) below the surface of the water.
Allow it to remain in the water for fifteen minutes before applying heat. Sus-
pend the thermometer so that the bottom of the bulb is level with the bottom
edge of the cube of pitch and within 0.635 cm. Oi in.), but not tou_ching the
cube. . Apply the heat in such a manner that the temperature of the water is
raised 5° C. (9°, F.) each minute. The temperature recorded by the ther-
mometer at the instant the pitch touches the bottom of the glass vessel shall
be reported as the softening-point. No correction shall be made for emergent
stem .. The rate of rise of temperature shall be uniform and shall not be
averaged over the period of the test. The maximum permissible variation
for any minute period after the first three shall be ±0.5° C. (0.9° F.). All
tests in which the rate of rise in temperature exceeds these limits shall be
rejected.
For Pitches Fusing below 110° F.-Use the same method as given above,
except that the water when placed in the glass vessel shall be at a temperature
of 4° C. (39.2° F:). The cube shall be allowed to remain fifteen minutes in
this water before applying the heat.
The use of freshly distilled water is essential, as otherwise air bubbles may
form on the cube and retard its sinking. Rigid
adherence to the prescribed rate of heating is
absolutely essential in order to secure accuracy
of results. A sheet of paper placed on the
bottom of the glass vessel and conveniently
weighted will prevent the pitch from sticking to
the glass vessel, thereby saving considerable
time and trouble in cleaning. The limit of
accuracy of the test is ±0.5° C. (0.9° F.).
For Pitches Fusing above 170° F.-The
heating is performed in an air bath in the appa-
ratus illustrated in Fig. 195. The cube should ~
be suspended in line with the observation win- 'F10. 195. -CUbe-lll- . A:Ir Me th0 d
dows, and the·thermometer bulb brought to the for High Fusing-point Substances.
same level. The temperature is raised 9° F.
per minute, and recorded by the thermometer when the cube drops 1 in.
Volatile Matter
This test is used for identifying various bituminous materials. Thus in the
case of asphalts, the volatilization test will often serve to identify soft native
asphalts, which contain larger percentages of volatile matter than soft residual
or blown petroleum asphalts. Cut-back products also carry a large percentage
of volatile constituents.
1528 EXAMINATION OF BITUMINOUS. SUBSTANCES
This test covers the determination of the loss in 'weight jexclusive of water)
of oil and asphaltic compounds when heated as hereinafter prescribed. The
material under examination shall, therefore, firSt:--he tested for water and if
water is found to be present, it shall be removed by suitable methods of dehy-
EXAMINATION OF BITUMINOUS SUBSTANCES 1529
dration before the material is subjected to the loss on heating test; or another
sample shall be obtained which is free from water.
The oven shall be rectangular iIi form with double walls and heated by
electricity, as illustrated in Fig. 196. Its interior dimensions shall be as follows:
height, exclusive of space occupied by the heating element, not less than
29.26 cm. (11.5 in.), width and depth, each, at least 5.08 cm. (2 in.)' greater than
the diameter .of the revolving shelf. The oven shall be provided with a per-
forated metal circular shelf approximately 24.8 cm. (9.75 in.) in diameter. A
recommended form of aluminum shelf is shown in Fig. 197. This shelf shall
i'-;r.1<
Section A-B.
Top View,
FIG. 197.-Shelf for Volatility Oven.
be placed in the center of the oven, with respect to all dimensions of the interior
of the same, suspended by a vertical shaft and provided with mechanical means
for rotating it at the rate of 5 to 6 revolutions per minute. One side of the
oven shall be hinged and equipped to serve as a tight fitting door which shall
contain a window at least 4 in. square, with double glass, through which a
thermometer, located in front of and level with, the revolving shelf, may be
read without opening the door. One round vent hold shall either be placed in
the top and bottom of the oven, or one only near the top, and one only near the
bottom at each side of the oven. Each hole shall be not less than 1.27 cm.
(0.5 in.) nor more than 1.60 cm. (0.63 in.) in diameter and they shall both be
left open during the operation of the oven.
The thermometer shall be graduated in Centigrade degrees, the range being
155 to 170 0 C.
The container in which the sample is to be tested shall be of metal or glass,
cylindrical in shape, and shall have a flat bottom. Its inside dimensions shall
be substantially as follow,s: diameter, 55 mm. (2.17 in.); de'pth, 35 mm. (1.38 in.).
NOTE.-The American Can Company's 3-oz. Gill style flat-bottom, seamless oint-
ment box, deep pattern, fulfills these requirements.
1530 EXAMINATION OF BITUMINOU~ SUBSTANCES
.
'_ : ., .....# ..
';;'.-.. ~ . ~~ .
//-/~'--'----"
than the ~ontainer. It shall support the container ~ in. above the hot plate
and with at least 3i in. free air space between the sides of the container and of
the air-bath below the opening. A suitable air-bath is shown in Fig. 198.
1532 EXAMINATION OF BITUMINOUS SUBSTANCES
(b) The air-b&th shall be heated upon a suitably mounted hot plate, heated
either electrically qr by. means of a gas flame. The plate shall pe capable of
maintaining the sample continuously at the required temperature, and ap-
paratus necessary to fulfill this requirement, such aE1 ~ rheostat. OJ' gas pressure
regulator, shall bE~ provided.
Tbe tbermometer snall con'iorm to tbe 'iollowing requirementB. 'T'nBBB
specifications cover a special thermometer graduated to either Centigrade or
Fahrenheit degrees as specified, the ranges being -6 to' +400 0 C., or +20 to
+760 0 F., respectively. The sample as received shall be thoroughly stirred
and agitated, to insure a complete mixture before the portion for testing is
removed. I~
One hundred grams (100.00±0.10 g.) of the material to be tested '8hall be
weighed into a tared container, which shall then be placed in tbe air-bath in
position to be heated. The thermometer shall be supported i:n the sample
equidistant from the sides of the container and with the bottoIil of the .bulb
neither more than U in. above nor touching the bottom of the cOrltainer. The
bulb shall be completely immersed in the sample throughout the heating. An
assembly of the apparatus is shown in Fig. 199.
thermometer which may be readily scraped off shall be returned to the con-
tainer, which then shall be removed from the air-bath, cooled and weighed.
The penetration of the residue shall then be determined with the exception that
the 6-oz. container in which the evaporation has been conducted, shall be used
instead of the 3-oz. container specified in the test for" penetration."
It frequently is necessary to make several trials before a residue of the
required penetration is obtained. If it is determined to be greater than that
required, all water shall be removed from the container and the surface of the
sample, and the heating and determination of penetration may be repeated as
before. o.rdinarily a residue shall be considered as satisfactorily obtained when
its penetration is within 15 of that desired, and its percentage by weight of the
original sample shall be calculated. When it is necessary to determine more
'. precisely the percentage of residue having the specified penetration, such a
percentage shall be computed by interpolation between percentages of two
residues, one having a penetration greater and one having a penetration lower
than that specified. The percentage shall be reported as:
P!!rcentage of residue of ........ penetration (determined ........ ) stating, first,
the specified penetration, and second, the penetration actually determined for the sample
tested or calculated by interpolation.
Certain types of road oil will readily form rings of hard asphalt at the side
of the container. Great care should be taken 1{hat this material be completely
fluxed in the sample before the penetration of the residue is determined. Dupli-
cate determinations should not differ by more than 1.0% with the same operator
nor more than 2.5% between different laboratories.
Flash-Point
A number of flash-point testers have been proposed, including the Pensky-
Martens, the Cleveland and the Tag types, which are fully described on p.1744 ff.
Fixed Carbon
The same procedure is followed as for testing coal (" Volatile Combustible
Matter"), as described on p. 1638 and p. 1639.
Distillation Test
The value of this test is to ascertain the adaptability of bituminous mate-
rials for a given use, generally for road treatment; also for gauging the llniform-
ity of supply, for purposes of factory control, and most important of all, as a
criterion of the quality. This test is generally applied to tar products as an
equivalent of the volatility test becomes of value in identifying the kind used
(upon determining the specific gravity of the fractions distilled), as a means of
distinguishing a cut-back tar from a straight-distilled tar (upon determining
the specific gravity of the fractions, their viscosity, alsQ the fusing-point of the
residue), and for detecting the presence of abnormal amounts of naphthalene.
This test has been standardized as follows:
(a) For Road Oils (Asphaltic and Coal-tar Pitch), etc.-The apparatus
consists of a flask, condenser tube, shield, receivers and thermometers as
specified.
1534 'EXAMINATION OF BITUMINOUS SUBSTANCES
Flask: The distillation flask, Fig. 200, shall be It side neck distilling flask,
having the following dimensions: I
Diameter of bulb, outside .................. . 8~ mm. ± 1.5 mm.
Diameter of neck, inside ........ , .......... . 2,2 mm. ± 1.0 mm.
Diameter of tubulature, inside .............. . 10.0 mm. ±0,5 mm.
Height of flask, outside ....... , ............ . 13~ mm.±1.5 mm.
Vertical distance bottom of bulb, outside, to
horizontal tangent at tubulature inside, .. 93 mm. ± 1.5 mm.
Length of tubulature ........ " ., .......... : 220 mm. ±5.0 mm.
Angle of tubulature ....................... . 73 deg. ±2 deg.
Thickness of tubulature wall ............... . 1.0 to 1.5 mm.
. ~ Inside Diameter
~->-I 22 mm.±I.O mm.
I
T-~-
Condenser Tube: The condenser tube shall be a suitable form of tapered glass
tubing of the following dimensions:
Outside diameter of small end .............. . 12.5 mm.±1.5 mm.
Outside diameter of large end .............. . /28.5 mm. ±3.0 mm.
Length ............................... ~ .. . 360.0 mm. ±4.0 mm.
Length of tapered part ................... , .. 100.0 mm. ±5.0 mm.
Shield: A g~lvanized iron shield, lined with Ys-in. asbestos, of the form and
dimensions shown in Fig. 201 shall be used to protect the flask from air currents
and to prevent radiation. The cover (top) may be of transij.f!. board masie in
two parts, or it may be of galvanized iron lined with Ys-in. asbestos.
Receiver: The distillates shall be collected in tared Erlenmeyer flasks having
a capacity of 50 to 100 ml.
Thermometer: The thermometer shall be graduated in either Centigrade or
Fahrenheit degrees as specified, the range being from 0 to 400 0 C. or 30 to
760 0 F., respectively. (
The sample as received, shall be thoroughly stirred and agitated, warming
if necessary, to insure a complete mixture before the portion for analysis is
removed. -
The material may be tested for distillation without dehydration, if water is
present not to exceed 2.0%. If water is prEl_se'iit in excess of 2.0%, the
bituminous material shall. be dehydrated before testing, by distilling 500 ml.
EXAMINATION OF BITUMINOUS SUBSTANCES 1535
I
I
I
r---136 mm.±3 mm.~
, _ r-:o- 45 mm.
~ ±2mm.
, '..:
D!~l
't-
"
'I ' ,I
I,
"I,
I,
",
,-
"
"
'I
I, !, 1
, "'I
I,
:t,
I'
Ill"
I
I,
I,
pelled, the distillate is allowed to settle, the water drawn off and the oils
returned to the residue in the still after the contents have cooled below 212 0 F.
Thermctfftftf
/~l.~.i
Shi'eld·
i?eIorl..
''l:_~r '~;~2~;;:'"
20M"h
~Chimn~'y
The flask shall be supported on a tripod or rings over two sheets of 20-roesh
gauze, 150 mm. square, as shown in Fig. 202. It shaH be connected to the
1536 EXAMINATION OF BITUMINOUS SUBSTANCES
the large end of approximately 18 mm. The outlet end shall be ground to an
angle of 45° ±5° with the inside vertical.
The receivers shall be graduated cylinders, of uniform diameter, with a
pressed or molded base and a lipped top. The over-all height shall be not less
than 24,8 cm. (9% in.) nor ·more than 26.0 cm. (lOX in.). The cylinder" shall
,
be graduated in single milliliters to contain 100 ml., and the graduated portion
shall be not less than 17.78 cm. (7 in.) nor 'more than 20.32 cm. (8 in.) in length.
Each fifth graduation,shall be distinguished by a longer line, and the graduations
shall be numbered from tlie bottom up at intervals of 10 ml. The graduations
shall not be in error by more than 1 ml. at any point on the scale.
The flask shall be supported on a tripod or ring over two sheets of 20-mesh
gauze, 150-mm. square as shown in Fig. 203. It shall be connected to the
condenser tube by a tight cork joint. The thermometer shall be inserted
through a cork in the neck with the bottom of the bulb U in. from the bottom of
the flask. The axis of the flask through the neck shall be vertical. The dis-
tance from the neck of the flask to the outlet end of the condenser tube shall be
not more than 600 nor less than 500 mm. The burner should be protected
from drafts by a suitable shield or chimney.
The adapter shall be adjusted over the end of the condenser tube so as to
conduct the distillate into the receiver, and the top of the receiver shall be
covered closely during' the distillation with a piece of blotting paper or its
equivalent, which shall be cut so as to fit the adapter tightly. The adapter,
shall extend into. the receiver at least 2.54 cm. (1 in.) but not below·the 100-ml.
mark. Unless the laboratory air temperature is between 12.8 and 18.3° C.
(55 and 65° F.) the receiver shall be immersed up to the 100-ml. mark in a
transparent bath maintained between these temperatures. The condenser tube
shall be clean .and dry .
• Two hundred milliliters (calculated from the specific gravity of the material
at 15.5° C. (60° 'F.)) of the sample shall be weighed into the flask, the apparatus
assembled and heat applied so that the first drop comes over in from 5 to 15 min.
1538 EXAMINATION OF BIT-qMINOUS ~UBSTANCES
The distillation shall be conducted at the rat.El of between 50 and 70 drops per
minute. Should the sample foam the distillation rate Will have to be reduced,
but the normal rate shall be resumed as soon as possible. If excess foaming
persists the distillation may be more easily controlleCliby applyi~g the flame
near the edge of the bulb instead of at the center of. samb. The distillate shall
be collected in the specified receivers, and the volume ot distillate at all specified
temperatures recorded. The volume of any separat'ed water shall also be
recorded. When the maximum specified temperatur{l of the test is indicated
by the thermometer, the flame shall be removed and the residue poured imme'-
diately into a 6-oz. tin box placed on its cover to prevent two rapid cooling at
the bottom. Any Qil which may remain in the condenser tube shall be drained
into the last receiver.
As soon as no further vaporization is apparent, the residue shall be stirred
to insure homogeneity, and then poured into the neCessary apparatus for the
required tests. During the progress of the distillation the thermometer shall
remain in its original position. No correction shall be made for the emergent
stem of the thermometer. Temperatures to be observed in the distillation test
shall be corrected for the effect of the altitude of the laboratory in which ~he
test is made.
Tll.p-_r.p.Rl..Uj$ n.f j.be. 11i.'<f..i11.ajjOJJ lm,j..s.b.sll be_rPPfl_rjeD in percentage by vo)u.me
of water-free material. .
(c) For Coal-tar Products.-In testinfl coal~tar products, the foregoing
method shall be modified as follows: The thermometer :Shall be inserted through
a cork in the neck of the flask, with the bottom of its bulb 12 to 13 mm. from
the surface of the oil. The exact location of the th~rmometer bulb shall be
determined by placing a vertical rule graduated in divistons not exceeding 1 mm.
back of the flask, when the latter is in position for Ule test, and sighting the
level of the liquid and the point for the bottom of the thermometer bulb.
The following fractions are usually reported:. I
SOLUBILITY TESTS
Solubility in Carbon Disulfide
This test is useful for purposes of identification, for ascertaining the adapt-
ability of a bituminous substance for a given purpose, for gauging its uniformity
of supply, and as a criterion of its quality (i.e., purity) and consequently its
intrinsic value. Crude bituminous materials are often purchased on the basis
of the percentage soluble in carbon disulfide. The presence of non-mineral
matter insoluble in carbon disulfide may be regarded as just that much inert
material, and in certain cases as an indication that the material has been care-
lessly prepared or overheated in its process of manufacture. The mineral
matter may be regarded as an adulterant. In the case of native asphalts, the
larger the percentage soluble in carbon disulfide, the greater will be their
intrinsic value. 'The percentage and composition of the mineral matter will
often indicate the source of the native asphalts. Asphalts derived from' pe-
troleum are substantially free from mineral constituents, and with the possible
exception of the harder grades, contain little to no non-mineral matter insoluble
in carbon disulfide ..
With a native asphalt containing over 10% of mineral matter, it is necessary
to separate the portion soluble jn carbon disulfide, before ascertaining its
physical characteristics, fusing-point, and sometimes fixed carbon, in which case
the soluble constituents should be recovered as will be described.
Two methods will be consid!'lred, depending upon whether or not the
constituents are to be subjected to a detailed analysis.
. Where the Constituents are Not to be Examined Further.-In this case
the method recommended follows along the lines o£ the one standardized by the
A. S. T. M. which is substantially as follows, differing however somewhat in
phraseology: The sample shall be representative and if it contains more than
2% of water it shall be dehydrated by distillation in a copper still, the water-free
distillate being returned to the residue. If the material is hard and brittle, it
may be ground 4 and dried at a temperature below the temperature of volatiliza-
tion of the material, in a shallow nickel or iron dish.
A Gooch crucible, approximately 4.4 cm. in width at the top, tapering to
3.6 cm. at the bottGm, with a depth of 2.5 cm. shall be set in the filter tube
inserted in the stopper of the filtering flask. The flask shall be connected with
the suction pump. Before suction shall be applied, the crucible shall be filled
'with asbestos 5 suspended in water which shall be allowed to settle partly in the
crucible. A light suction shall be applied to draw off the water, leaving a firm
mat of as.bestos in the crucible. More suspended asbestos shall be added and
the process repeated until a felt is built up that barely transmits light. The
felt shall then be thoroughly washed with water, dried in a drying oven, and
ignited over a Bunsen burner: The crucible shall then be cooled in a desiccator
and weighed. An amount of materiaL which shall contain approximately 1
gram of bitumen shall be weighed into a tared Erlenmeyer flask.
4 Where it is not desirable to crush the rock or sand grains, a lump should be placed in
the drying oven until it is heated through and softened at the lowest possible temperature,
whereupon it may be crushed into a thin layer and dried as described.
6 Asbestos (amphibole) cut in pieces not exceeding 1 cm. in length shredded and
shaken up with water.
1540 EXAMINATION OF BITUMINOUS [SUBSTANCES
flasks shall be washed again with carbon disulfide, the solution decanted and
the process repeated until the washings are practically colorless. The tem-
perature shall be maintained between 20 and 25 0 C.
The crucible and both flasks shall be dried at 100 to 125 0 C. for twenty
minutes, cooled in a desiccator and weighed. The procedure from this'point
on is exactly the same as in the preceding. The per cent constituents soluble in
carbon disulfide, non-mineral constituents insoluble in carbon disulfide, and
mineral constituents shall be calculated.
The author finds that in the preseljlce of large quantities of finely divided
insoluble matter, the method may be materially shortened by adding a weighed
quantity (about twice the weight of bituminous material) of freshly ignited,
long-fibered amphibole to the bituminous substance in the first flask. On
shaking with carbon disulfide, the asbestos serves to dilute the insoluble matter,
preventing the latter from clogging the pores of the filter, and accordingly
reducing the time of filtration.
Where the Constituents are to be Examined Further.-Extract 25 to 50 g.
of the moisture-free bituminous substance with carbon disulfide as previously
described, increasing the quantities of carbon disulfide proportionately. In
certain cases an extractor described on p. 1568 may be conveniently employed.
Residue Insoluble in Carbon Disulfide: This contains the non-mineral con-
stituents insoluble in carbon disulfide (e.g., insoluble bituminous matter, free
carbon and vegetable matter derived from associated soil) together with the
mineral constituents retained by the filter (e.g., clay with 2 molecules of water in
chemical combination, 6 silica, silicates, calcium and magnesium carbonates,
calcium sulfate with half a molecule of water in chemical combination, iron
pyrites, etc.). Dry in an oven at 220 0 F. to constant weight and mix well to
obtain a uniform sample. If the mineral matter is coarse, pulverize sufficient
of the well-mixed material in a mortar to 200 mesh or finer for use in carrying
out the examination described in sections (a) to (e) inclusive.
(a) To determine the 'water in chemical combination with the clay and
calcium sulfate, heat 10 g. to bright redness in a stream of dry hydrogen or
illuminating gas, in an electric combustion apparatus, catching the water in
a tared calcium chloride bulb, which is reweighed at the end of the operation.
The gain in weight represents the water which should be calculated in per-
centage.
(b) To determine the sulfur present in sulfides, weigh 25 g. into a small flask
closed with a stopper carrying a thistle tube filled with hydrochloric acid and
an outlet tube leading into a beaker of bromine water. Introduce the acid and
boil until all the hydrogen sulfide is expelled into the bromine water, which
serves to oxidize it into sulfuric acid. Expel the excess of bromine by boiling
and precipitate with barium chloride in the usual manner. Calculate the
percentage of sulfur present.
(e) To determine the carbon dioxide present in carbonates, weigh 1 g. into a
Schroetter or ;Mohr alkalimeter and treat with hydrochloric acid in the usual
manner. The loss in weight represents the sum of the carbon dioxide (liberated
from the carbonates) and the hydrogen sulfide (liberated from the iron pyrites).
6 The water of hydration in clay is held tenaciously, for none is given up during one
hour's ~eating at 293 C., although it is expelled on heating to redness.
0
1542 EXAMINATION OF BITUMINOUS SUBSTANCES
>'
EXAMINATION OF BITUMINOUS SUBSTANCES 1543
aliquot portion into a tared crucible or dish. Evaporate the solvent over a
steam bath, then incinerate the residue and ignite at white heat to a clean ash,
until no further loss in weight occurs. Moisten with ammonium carbonate
solution, re-ignite at a low heat (not beyond incipient redness) and reweigh.
The ash represents the anhydrous clay, silica, calcium and magnesium car-
bonates, together with the mineral constituents originapy in chemical combina-
tion with the bituminous matters. Calculate the mineral matter associated
with the bituminous constituents in total carbon disulfide extract. Note that
any water of hydration originally combined with the clay is not ascertained
by the foregoing procedure, but the amount present is usually so small that it
may be disregarded without vitiating the results. From the weight of the
residue insoluble in carbon disulfide, calculate the soluble bituminous constit-
uents by difference. Evaporate an aliquot portion of the well-mixed carbon
disulfide extract to exactly the calculated weight.
It has been found that when the solvent is expelled by heating over a water-
bath for one hour, tlie fusing-point is increased 2 to 3° C.; also that the retention
of 1 to 2% of very fine mineral matter does not change the characteristics of the
recovered bituminous constituents to any appreciable extent.
The following variations have been proposed for recovering the bituminous
constituents in their unaltered state:
(a) Evaporation at Atmospheric Pressure.-From the weight of the ex-
tracted mineral matter, calculate the bituminous matter by difference, and
evaporate the carbon disulfide extract to exactly this weight. This may be
conveniently performed by distilling and condensing most of the carbon disul-
fide over an incandescent light or, an electric stove in a large distillation retort
connected with a condenser and receiving flask. The retort shall have a
thermometer inserted, with the bottom of the bulb approximately 1 in. from
the bottom of the retort. The concentrated solution is transferred to a tared
dish, evaporated dry on a steam bath with constant stirring,-which may be
conveniently accomplished with a motor-driven agitator, and the last traces
of solvent expelled in an oven at 105° C. until the residue attains the calculated
weight.
An alternate procedure consists in distilling the filtered extract (containing
a calculated amount of soluble constituents equal to approximately 100 g.) to
a bulk of about 150 mI., then transferring the solution to a spherical flask of
300 mi. capacity. The evaporation is completed on a water-bath which is
heated from 40 0 C. up to the boiling-point, at which it i~ maintained for 5
hours, during which period the contents are stirred at intervals with a tared
glass rod. Finally the flask is heated in an electrically heated oil-bath, from
an initial temperature of 1050 C. to a final temperature of 1300 C., which is
kept constant for a period ranging from 4 to 6 hours, during which the contents
are kept stirred with the tared glass rod. The oil-bath is kept agitated with a
motor driven paddle, which likewise causes the flask to revolve .• The time of
heating in the oil-bath may be determined by dissolving asphalt of known
penetration in carbon disulfide .and ascertaining the time required at 130° C.
to bring its penetration back to its original figure. .
(b) Evaporation under Vacuum.-The solvent is first removed by distilla-
tion as in the foregoing, until the temperature of the residue in the flask reaches
3000 F. The distillation flask is then connected to a filter pump and the
1544 EXAMINATION OF BITUMINOUS SUBSTANCES
distillation continued under a vacuum of 180 mm. mercury, raising the tem-
perature to 500 0 F. at the rate of 5 to 6° F. per minute. If the original pene-
tration is greater than 150 at 77° F., the distillation temperature should not
exceed 425° F. The residue in the flask should thenl be wei~ed. This 'pro-
cedure will give satisfactory results, provided. the bituminous substance does
not contain any appreciable quantity of constituents' {rolatile, at 400° F.
The following modification is claimed to give very accurate results. Dis-
solve the substance in C.P., carbon disulfide previously ,dried by contact with
CaCho The extract should contain 5 to 6% of the soluble constituents, and
not exceed 240 m!. in volume, which may be attained by evaporation of' by
dilution. A shallow, circular brass or nickel plate as illustrated in Fig. 204 i.s
few drops of concentrated HCI. The separated ~lumps are freed from acid by
kneading in several portions of water with a glass rod. The asphalt is then
dissolved in about 240 ml. of carbon disulfide, allowed to stand in a separatory-
funnel until most of the water rises to the surface, whereupon the solution is
drawn off into a flask and shaken repeatedly with granular CaCh at 15 minute
intervals. The carbon disulfide solution is then poured off and treated as
described above to recover the dissolved asphalt.
(c) Evaporation in a Stream of CO 2.-In this case the carbon-disulfide or
benzol extract containing 50 to 75 g. of the asphalt is distilled in a wide-mouthed
250 m!. flask in a form of apparatus shown in Fig. 205. The flask is fitted with
for asphalts having a fusing-point of over 150 0 F. (R. and B.). It is important
that the extracted asphalt solution be distilled on the I same day that it is dis-
solved, otherwise considerable alteration may occur if tlhe solution is allowed to
stand for more than 10 hours, regardless of wh,ether benzol or carbon disulfide
is used, and will result in a change in the physical charadteristics of the extracted
asphalt. I '
A modification of the procedure consists in distilling off the solvent in' a
stream of CO 2 under reduced pressure.
A rapid method for separating asphalt from asphaltic limestones and dGlo-
mites consists of the following: Weigh out 10 g. of the rock asphalt and add
75 m!. of a mixture of 30 ml. concentrated RCI (sp.gr. 1.19) and 45 ml. ethyI
ether. Stir 10 minutes, until the carbonates dissolve~ ,add 75 m!. water and
warm over a water-bath until the ether is expelled. Then filter and wash with
warm water. Dry the filter at 105 0 C. and extract the filter with.. carbon disul-
fide to recover the asphalt. The extract may be treated by any of the pro-
cedures described above.
The soluble bituminous constituents, after being separated in their pure
state, as previously described, may be then examined further for their physical,
and chemical characteristics. Due allowance should be made for the fact that
any bituminous matter insoluble in carbon disulfide will be retained mechani-
cally by the extracted mineral constituents, which with asphaltic products is
usually relatively unimportant, but in the~ case of coal-tar products, will amount
to a considerable item.
Carbenes
The expression" carbenes" has been applied to that portion of bituminous
substances soluble in carbon disulfide but insoluble in carbon tetrachlor.ide.
This test is of value in identifying bituminous substances, gauging their uni-
formity of supply, for purposes of factory control, and as a criterion of their
quality. Certain hard native asphalts and asphaltites; particularly grahamite,
normally contain a percentage of carbenes, whereas petroleum asphalts do not
show carbenes unless they are overheated or over-bI6wn. If more than 0.5%
is present in petroleum asphalts, their quality is: to be regarded as questionable.
Carbenes are found in tars and pitches in varying amounts.
Although carbenes are found in grahamite a:r;J.d certain hard natural asphalts
when tested as such, they disappear upon fluxing to a softer consistency. With
a
petroleum asphalts, tars and pitches, the carbenes are of diffe:rent character,
since they are insoluble in fluxes and do not disappear upon being' so treated.
This test is carried out by following the same procedure as in determining
the solubility in carbon disulfide, but replacing the latter with carbon tetra-
chloride. The carbon tetrachloride must be free from carbon disulfide, which
may be insured by distilling it under a dephlegmatpr, discarding any distillate
below 76° C. The solvent is then filtered througli calcium chloride, and any
free hydrochloric acid removed by blowing dry air through it.
The procedure has been standardized as follows:
4n amount of material which shall contain approximately one gram of
bitumen shall be weighed into a tared. Erlenmeyer, :6~sk. 6ne hundred milli-
liters of chemically pure carbon tetrachloride .shalt be added to the flask in
small portions with continued agitation until all lumps disappear and nothing
EXAMINATION OF BITUMINOUS SUBSTANCES 1547
adheres to the bottom. The flask shall be corked and set aside in subdued
light for at least twelve hours.
The Gooch crucible shall. be set up again with the suction flask and the
carbon tetrachloride solution carefully decanted through the asbestos felt, with
or without light suction as may be found necessary. No sediment shall be
allowed to go onto the filter. A small amount of carbon tetrachloride shall be
used to wash down the sides of the flask and then the precipitate shall be brought
onto the felt and the flask scrubbed with a feather if necessary to remove all
precipitate. The contents of the crucible shall be washed with carbon tetra-
chloride until the washings are colorless. Suction shall be applied to the cruci-
ble to remove the carbon tetrachloride. The crucible shall be dried in the oven
at 100° to 125° C. for twenty minutes, cooled in the desiccator, and weighed.
In case insoluble matter adheres to the flask, it shall be dried and weighed,
and the increase in weight over the original weight shall be added to the weight
of insoluble matter in the crucible.
The proportion of bitumen soluble in carbon tetrachloride shall be reported
on .the 'basis of total bitumen taken as 100:
Proportion of Bitume.n Soluble in} = Bitumen Soluble in Carbon Tetrachloride
Carbon TetrachlorIde. Total Bitumen
The difference between the p~rcentages soluble in carbon disulfide and
carbon tetrachloride respectively, represents the per cent of "carbenes."
.'Filf.rPap"..tIJ
Alundum
Cup
washing by the vapors of the boiling solvent. The filtering medium shall be a
flat-bottom, 30 by 80 mm. RA-98 alundum thimble. The thimble 'shall be
'suspended in the extraction flask either by a wire basket hung from two small
hooks on the under surface of the metal coyer of the flask, or it shall be sup-
ported by making perforations near the upper edge of the thimble and suspend-
ing from the cover by German silver or platinum wire.
An amount of material which shall contain 10.0 g. ±0.1 g. of the substance
shall'be w~ghed into a 100-ml. beaker; 50 ml. of pure benzol shall be adiied and
the solution stirred thoroughly. The solution shall be transferred at once to
the weighed alundum thimble. The beaker shall be rinsed clean with pure
benzol and the washings added to the thimble. The thimble shall then be
covered with a lid of alundum ware and placed immediately in the extraction
apparatus. The extractor shall contain a suitable quantity of pure benzol and
shall be heated sufficiently to boil the solvent. The extraction shall be con-
tinued until the solvent descending fro~ the thimble is colorless. The thimble
shall then be dried at 1050 C.±5° C., cooled in a desiccator, and weighed.
(B) Asbestos Mat Method:-The filtering medium shall consist of a No.3
Gooch, Coors porcelain or equivalent crucible approximately 3.5 cm. in diam-
eter at the top, tapering to 2.2 cm., at the bottom with a depth of 4.0 em., con-
taining a mat of acid-washed medium fiber asbestos approximately 2 mm. in
thickness.
The Gooch crucible shall be placed in the suction apparatus and filled with
acid-washed medium fiber asbestos suspended in water. Gentle suction shall
be applied and more of the suspension, if necessary, shall be added to 'make a
mat approximately 2 mm. in thickness. With· the suction still on, the pad
shall be washed with water until all small particles of asbestos are removed.
1550 EXAMINATION OF BITUMINOU~ SUBSTANCES
The crucible shall be dried at 1050 C.±5° C., placed in a d~siccator, and
weighed. I
An amount of material which shall contain 10.0 g. ±o.i g. of the substance
shall be weighed into a 125-ml. Erlenmeyer flask; 510 ml. .of pure beniol shall
be added and the solution stirred thoroughly and br:ought to a boil. The hot
solution shall be carefully poured into the weighed: prepared Gooch crucible
without suction until the mat is covered. Then gentle suction shall be a:rJpliecl,
to the crucible and the remaining solution added, taking care that the mat is
covered with solution at all times. The Erlenmeyer .flask shall be rinsed
clean with pure benzol and the washings added to the crucible. Benzol shall
be added to the crucible until the descending solvent is colorless. The crucible
shall then be dried at 1050 C.±5° C., cooled in a desiccator, and weighed.
(C) Paper Thimble or Filter-paper Method.-A paper thimble or else two
thicknesses of S': & S. No. 575 or Whatman No.5 hardened filter-paper (15 cm.
in diameter, arranged in cup shape by folding) may be used in connection with
an extractor of the form shown in Fig. 207. In this case the procedure is
~ similar to that outlined in the Porous Thimble Method (A).
The insoluble residue of'''free carbon" in~ludes the mineral ash, which may
be ascertained by incineration. Some tars (e.g. blast-furnace tars) contain
substantial amounts of ash, which would ,otherwise introduce serious errors.
It has been found that the amount of free carbon will increase with the time
the solvent remains in contact with the tar or pitch. The amount of such
increase varies with the particular solvent employed. In the case of carbon
disulfide or chloroform, the increase is less than with benzol or toluol (or mix-
tures of these two) and the amount apparently reaches a maximum III about
120 hours.
A method has been proposed in which 5 g. of the pitch is heated with 200 ml.
tetralin in an autoclave at 240-250 0 C. under a pressure of 12-13 atmospheres
for 2 hours. Upon cooling, the liquid is filtered through a Gooch crucible .and
washed successively with 50 ml. tetralin and 100 ml. benzol. The residue is
dried for 2 hours at 150 0 C. in an atmosphere of CO 2 ¥nd weighed. cAnthracene
oil (freed from crystallizable substances at 15° C.) has also been proposed, in
which case, 5 g. of the substance are digested with 5 g. anthracene oil under a
reflux condenser for 4 to 5 hours, cooled, dilut'ed with 500 ml. benzol, filtered,
washed with benzol, etc., as in the preceding.. Aniline has also been suggested
as extracting medium, involving the digestion of 5 g. of substance with 25 ml.
aniline over a stea'm bath for Y2 hour, ,pouring the liquid 'On a por~)Us clay plate
to absorb the solvent, washing the residue with pyridine, followed by benzol,
drying and weighing. '
EXAMINATION OF BITUMINOUS SUBSTANCES 1551
CHEMICAL TESTS
Water
The estimation of water is made in some cases for purposes of identification,
and in others as a criterion of the quality. Native asphalts and tars are exam-
ined in tliis way to ascertain whether they exist in the crude or the dehydrated
state. This test is also used for dehydrating bituminous. substances to render
them suitable for further examination, where ·the presence of water would in-
terfere.
Substances Distilling at Low Temperatures.-This method is adapted to
crude petroleum, tars, creosote oil and other' fluid bituminous substances
distilling at comparatively low temperatures. The apparatus is set up as
shown in Fig. 208. The copper still is provided with a removable flanged top
and yoke of the form and approximate dimensions shown in Fig. 209•
.-TJ1 t rmommr
Cpnn«f!ng I
1.····...•.... 3875,"' ...•.. ->1
. .. t
) Tubl ''r------I
Separol11ry
Fun"el,~ •....
( C"pac/fy
120cc)
6radvaf~d
inJ cc.-······
I I
~ .. -.•. 3.7·.a2·....~
(9(Om",.1: fi.lmm)
Draw off the water, and return any light oil to the bituminous matter in the still.
The dehydrated materjal should then be used for fuhher tests. .
To prevent frothing and spattering, it has been recommended that the still
be surrounded with a cylindrical vessel, closed at the bottom and open at the
top, of a somewhat greater diameter than the still. 'The vessel is so adjusted
that its upper rim is a little lower than the level of th~ bituminous substance in
the retort, whereupon it is fill~d with water until it overflows. Heat is then.
applied to the shallow layer of tar above the water level, with the ring burner.
The water jacket absorbs heat which would otherwise be transmitted to the
lower level of the tar and cause bumping. The level of water falls gradually by
evaporation, permitting the zone undergoing dehydration to fall slowly. When
the water level has fallen below the bottom of the still, the contents are practi-
cally dehydrated.
A modification of the foregoing method has been suggested which consists in
providing the cover of the copper retort with two openings, one connected with
a receptacle holding 100 m!. of tar and provided with a stop-cock to control the
flow of tar into the retort. A weighed quantity of tar is introduced into the
receptacle and enough run into the retort to cover the bottom, and the re-
mainder is allowed to drip slowly into the retort. This prevents the tar
frothing over. '
Substances Distilling at High Temperatures.-This method is adapted to
asphalts and other bituminous substa]lces compara~ively free from volatile
constituents, and incapable of distilling without suffering decomposition.
Substances Fusing below 300 0 F.-When it is desired to determine the per-
centage of moisture without using the residue for other purposes, a convenient
method consists in distilling it with a solvent. The following procedure has
been standardized:
The apparatus shall consist of a metal still or glass flask, heated by suitable
means and provided with a reflux condenser discharging into a trap connected
to the still or flask. The trap serves to collect and measure the condensed
water and to return the solvent to the still. The t:ype of distilling apparatus
used is not an essential feature of this method, but glass has been generally used
for petroleum products and the metal still for ,road materials and tars.
(a) The metal still (Fig. 210 (a)) shall be a vertical cylindrical vessel, pre-
ferably of copper, having a faced flange at the, top to which the head is tightly
attached by means of a clamp. The head shall be of metal, preferably of brass
or cop~r, and be provided with a tubulation 1 in. in inside diameter.
(b) The glass flask (Fig. 210 (b)) shall be of the short-neck, round-bottom
type, made of well-annealed glass, having an approximate capacity of 500 m!.
The burner used with the metal still shall be,a ring gas burner 4 in. (100 mm.)
in inside diameter. With the glass flask, an ordinary gas burner or electric
heater may be used as the source of heat. The cOIfdenser shall be pf the water-
cooled, reflux, glass-tube type, having a condenser jacket not less than 400 mm.
(15U in.) in length with an inner tube 9.5 to 12.7 mm. (% to 72 in.) in outside
diameter. The end of the condenser to be inserted in the trap shall be ground
off at an angle of 300 from the vertical axis of the condense,r.
The trap shall be made of well-annealed glass .constr,~cted in accordance
with Fig. 210 (c) and shall be graduated as showh-;:--from 0 to 10 m!. in 0.1 m!.
divisions. The error of any indicated capacity shall not be greater than 0.05 ml.
EXAMINATION OF BITUMINOUS SUBSTANCE~ 1553
The outside diameters should be preferably 2.5 to 3.5 mm. Uz to Ys in.) greater
than the inside diameters specified.
l
t··A·...
i f
!
i
i
i
..:1_ -1-
f.
-~--
Courtesy A.S.T.M.
FIG. 21p.-Apparatus for Water Test.
(a) The solvent used when testing petroleum products or bituminous mate-
rials derived from petroleum shall be gasoline, free from water, and shall con-
form to the following distillation requirements: '
5% shall distill at a temperature not below 1940 F. (900 C.) nor above 212 0 F. (1000 C.)
90% shall distill below 410 F. (210° C.)
0
(b) The solvent used when testing bituminous materials derived from coal
tar, water-gas tar, etc., shall be a coal-tar naphtha or a light oil and shall con-
form to the following distillation requirements:
98% shall distill between 248 0 F. (1200 C.) and 482 0 F. (250 0 C.)
The sample shall be thoroughly representative of the material to be tested
and the portion of the sample used for the test shall be thoroughly representative
of the sample itself. Deviation from this requirement shall not be permitted.
N OTE.-The difficulties in obtaining proper representative samples for this determina-
tion are unusually great so that the importance of sampling cannot be too strongly
emphasized.
Procedure:When the sample to be tested contains less than 10% of water,
exactly 100 ml. of the material to be tested shall be placed into the still or flask
and thoroughly mixed with an equal volume of solvent by swirling, proper care
being taken to avoid any loss of material. If the material is measured by vo1-
1554 EXAMINATION OF BITUMINOUS SUBSTANCES
ume, an accurate 100 m!. graduated cylinder sha,ll' be used and, the contents
transferred to the still by rinsing with one 50-ml. I1ortion of solvent, followed
by two successive 25-ml. portions of solvent, the cylihder being allowed to drain
each time. When the sample to be tested contains, more than 10% of water,
the volume of material used shall be decreased to t~at which will yield some-
what less' than 10 m!. of water. '
i •
NOTE.-In special cases where the water content exceeds 10% and it is not desirable
to reduce the size of the sample to that which will yield somewhat less than 10 mI. of
water, a distilling tube receiver graduated from 0 to 25 m!. may be used. This tube
shall be graduated from 0 to 2 ml. in 0.1 mI., from 2 to 5 ml. in 0.2 ml. and from 5 to
25 ml. in 0.5 ml.
The connections between the still or flask, trap and condenser shall be made
by means of tight-fitting corks as shown. The end of the condenser inserted in
the trap shall be adjusted to that position which will allow the end to be sub-
merged to a depth of not more than 1 mm. below the surface of the liquid in
the trap after distillation conditions have been established. When the metal
still is used, a heavy paper gasket moistened with the solvent shall be inserted
between the lid and flange before attaching the clamp. A loose cotton plug
shall be inserted in the top of the condenser tube to prevent condensation of
atmospheric moisture in the condenser tube.
Heat shall then be applied and so regulated that the condensed' distillate
falls from the end of the condenser at the rate Qf from two to five drops per
second. The ring burner used with the metal still should be placed about 3 in.
above the bottom of the still at the beginning of the distillation and gradually
lowered as the distillation proceeds.
The distillation shall be continued at the specified rate until no water is
visible on any part of the apparatus except at the bottom of the trap. This
operation usually requires less than an hour. A persistent ring of condensed
water in the condenser tube shall be removed by increasing the rate of distilla-
tion for a few minutes.
The v:olume of condensed water: measured in thy'trap at room temperature
multiplied by 100 and divided by the volume of the sample used shall be the
percentage of water and shall be reported as " ...... per cent water by volume,
A.S.T.M. method." ,
The accuracy to be expected with this method is that duplicate determina-
tions of water should not differ from each other by more than one division on
the t r a p . , -
. Substances Not Fusing at 300 0 F.-In this case the material "is comminuted
by powdering (to about 60 mesh) or shaving, and a weighed quantity spread in
a thin layer on glass and maintained in an oven at 105 0 C. for one hour, or until
the weight becomes constant. If the substance is oxidizable in air, it should be
heated in an atmosphere of nitrogen or illuminating gas. Cool in a desiccator,
reweigh and calculate the .per cent moisture. I
by all the errors incurred in the other determinations, and especially by the
change in weight of the ash-forming constituents on ignition. Iron pyrites
will absorb oxygen from the air and change to ferric oxide, increasing the weight
of ash, and thereby causing a negative' error in the oxygen, equivalent to three-
eighths of the pyritic sulfur. Any calcium carbonate present will tend to
a!bsorb sulfur combined with the bituminous constituents. On the other hand,
there is always a loss on ignition' of "water of composition" from the clayey
and shaley constituents, also carbon dioxide from carbonates, etc., which tend
to compensate for the absorption of oxygen.
A more correct value is obtained by making the corrections indicated in the
following formula: . .
.
Corrected Oxygen = 100 - [( C.- C') + (H - H') + N + H 2 0+
,
8'+ Corrected
,
Ash]
Where C equals the total carbon; C' the carbon of carbonates; H the total
hydrogen less hydrogen of water; H' hydrogen from water of composition in
clay, shale, etc.; N the nitrogen; H 20 the moisture as found at 1050 C.; 8' the
sulfur' not present as pyrite or sulfate; "Corrected Ash," the mineral constituents
originally present, which for most purposes may be calculated with sufficient
accuracy by adding to the ash as found, five-eighths of the weight of pyritic
sulfur, the CO 2 of carbonates, and the water of composition of clay, shale, etc.
Solid Paraffins
Until recently, it was considered that the presence of paraffin was an indi-
ca,tion of the quality of asphaltic products, and many specifications stipulated
the maximum percentage permissible. It is now generally conceded, however,
that there is no rational relation between the solid paraffins in asphaltic
products and their quality. The determination of paraffin is therefore of
value only for purposes of identification. Traces of solid paraffins are found in
asphaltites, natural asphalts, and in asphalts produced from strictly asphalt-
base petroleums. On the other hand, more or less paraffin is present in as-
phalts derived from non-asphaltic and semi-asphaltic petroleums. It is absent.
in tars and pitches derived from high-temperature distillation processes.
The following method does not give absolute figures, since it merely dis-
closes the paraffin hydrocarbo!lS which are solid at room temperature, without
taking the liquid paraffins into consideration; nevertheless the results are of
value for purposes of comparison. Weigh 50 g. of the material in a tared 6-oz.
glass retort, and slowly distil until nothing but a residue of coke remains. The
distillation should take in the neighborhood of forty-five minutes from the time
the first drop comes over. The distillation is conducted'by heating at the rate
of 40 to 50 0 F. per minute to 450 0 F., followed by a rise of 150 F, per' minute to
5000 F. The distillate is caught in an Erlenmeyer flask, and weighed. Either
5 or 10 g. of the well-mixed distillate, depending upon the quantity of solid
paraffins present in the crude material, are transferred into a large test tube
and dissolved in 25 ml. of absolute ethyl ether and 25 ml. 'of absolute ethyl
alcohol. A similar mixture containing 25 ml. each of ether and alcohol is
made up, and this together with the oil solution is cooled separately to exactly
1556 EXAMINATION OF BITUMINOUS SUBSTANCES
0° F. for one-half hour in a mixture of ice and salt (to which if necessary calcium
chloride may be added). The oil solution is then :r:apidly filtered through a
weighed Gooch crucible similarly maintained at 0° F. by a jacket of ice and salt,
and w~shed with 50 ml. ?f the co?led ether-alcoho~\ ~ixt~re. A sim~le and
convement apparatus consIsts of an Illverted bottle 6 In. III dIameter, havmg,the
bottom cut off, and attached to the same rubber stopper which supports the
funnel holding the Gooch crucible. The space between the bottle, the crucible,
and the supporting funnel is packed with the ice and salt mixture: The Gooch
crucible is then removed, the outside wiped clean, placed on a tared glass and
dried in an oven at 80° C. until the last traces of ether and alcohol are evapo-
rated. The residue is weighed, and the perce~tage of solid paraffins calculated
in the original 50 g. of substance taken for analysis.
Various substitutes have been suggested for the ether-alcohol mixture, in-
cluding methyl-ethyl-ketone (butanone) containing 1.3% water, having a
specific gravity of 0.812 at 20° C.; also nitrobenzene used at 32° C.
It is a mooted question whether the solid paraffins ascertained by the
foregoing method are originally present as such in natural asphalts and asphalt-
ites, or are generated during the distillation process through the destructive
decomposition of other non-paraffinic components.
An alternate method used with success qy the author -consists in dissolving
10 g. of the substance in the smallest amount of benzol, applying heat if neces-
sary, adding 200 ml. of warm 88° naphtha, filtering through dried fuller's earth
and washing with additional warm 88° naphtha until the filtrate is clear. This
removes the dark-colored asphaltic substances, which remain adsorbed by tire
fuller's earth. The filtrate containing the paraffin is distilled to a small bulk,
evaporated to dryness on the water bath under-reduced pressure, the residue
dissolved in 25 ml. of ether and 25 ml. of alcohol, and treated as described
above. - -
Instead of removing the dark-'colored asphaltic constituents by means of
fuller's earth, an alternate procedure consists in extracting them by shakipg the
88° naphtha solution three successive times with 20 illl. concentrated sulfuric
acid in a separatory funnel, followed with N KOB: (in 50% alcohol-water
solution) and finally with water until neutral to phenolphthalein.
Another method for separating paraffin wax from petroleum residues con-
sists in mixing the residue with an excess of full.er's earth, cooling and extracting
either with: (1) cold benzol (which removes both- the oil:y matters and th.e as-
phalt), or (2) cold gasoline (which removes the oily matters only) and then with
cold benzol (which removes the asphalt). The residue of fullei's earth is then
treated with hot water which separates the adsorbed paraffin wax. This
procedure is precisely the reverse of the prec_edi~g, which may however be
explained by the fact that in this case the process is carried on in the cold.
According to Holde, the refractive index of the ,solid paraffins will indicate
whether the original substance was ozokerite or paraffin, or a mixture of the
two. When tested at 90° C. on the Zeiss or Abbe refractometer, ozokerite will
show a refractive index below 15.0, whereas the solid paraffins derived from
petroleum, shale, ligniteJ etc., will test between 15.0 and 30.0. When tested
in the solid state, commercial paraffin waxes show_a refractive index ranging
from 1.500 to 1.540 at temperatures between 31J" and 130° F. Furthermore,
EXAMINATION OF BITUMINOUS SUBSTANCES 1557
Sulphonation Residue
(A) Sulfonation Residue Insoluble in Concentrated Sulfuric Acid.-This
method of test has been proposed for separating saturated and unsaturated
hydrocarbons, as a means of destinguishing various asphaltic products (e.g .•
native asphalts and petroleum asphalts) from tars and pitches, thereby serving
for purposes of identification. Two alternate procedures are available, viz.:
(a) Portion Soluble in 88° Petroleum Naphtha.-The portion soluble in 88°
petroleum naphtha is brought to exactly lOO ml., either by adding more 88°
naphtha or else by evaporation, so that the quantity of substance carried in
solution will be approximately 5 g. This is then shaken in a 500-ml. separatory
funnel at 77° F. for exactly three minutes, with 30 m!. of a mixture of concen-
trated sulfuric acid and fuming sulfuric acid, having a specific gravity of 1.84
at i1° F. The funnel is allowed to stand quietly overnight, whereupon the
acid is drawn off and the oils unacted upon treated with another 30 ml. of the
acid. This time a few hours' standing should effect a sharp separation. If
the. second acid layer is strongly colored, the treatment should be repeated a
third time. The naphtha solution is washed successively with water, a 5%
solution of sodium carbonate and finally with water. The solution is ·evapo-
rated to dryness' over a steam bath and the residue weighed. This is equal
to the saturated hydrocarbons present in the portion soluble in 88° petroleum
J}.aphtha. As a guide in evaporating the last traces of naphtha from the satu-
rated hydrocarbons, a blank test should be run on 100 ml. of the 88° naphtha,
whereupon the portion unacted upon is mixed with 0.75 g. of a non-asphaltic
petroleum residuum and evaporated on the steam bath alongside of the sample
under test, until the former is reduced to exactly its original weight.
The results are expressed as the percentage of saturated hydrocarbons
. present in the portion soluble in carbon disulfide. This is calculated in the
following manner: If a represents the percentage soluble in carbon disulfide, b
the percentage soluble in 88° naphtha and e the percentage of saturated hydro-
car:bons in b; then the saturated hydrocarbons present in the portion soluble in
carbon disulfide will equal (be/a) X 100.
(b) Distillate at 315°-355° C.-This test expresses the percentage of satu-
rated hydrocarbons in the distilla~e between 315° and 3550 C. obtained upon
subjecting the bituminous substance to the flask method of distillation. It is
used to differentiate tars and pitches among themselves as well as from mineral
waxes, asphalts (native and pyrogenous) and asphaltites. The figures for coal-
tar pitches have already been given. The author cites the following additional
figures: wood-tar pitch 0%, saponifiable fatty-acid pitches 0%, un~aponifiable
fatty-acid pitch~s 0%, residual asphalt from Mexican asphaltic petroleum 86%,
wurtzilite asphalt 87% and gilsonite 85%.
The method of determining the sulfonation residue consists in distilling
sufficient of the material under examination by the flask method to obtain at
least 10 m!. of distillate between 315 0 and 3550 C. Exactly lO m!. of this
fraction are measured into a Babcock milk bottle, and 40 m!. of 37 normal
1558 EXAMINATION OF BITUMINOUS SUBSTANCES
sulfuric acid added, 10 m!. at a time. The bottle and its contents are shaken
for two minutes after each addition, and when all the acid has been added, the
bottle is kept at a constant tempe,rature of 98°-lOqo C. for one hour, during
which it is shaken vigorously every ten minutes. At the end of the hour, the
bottle is removed, cooled, filled to the top of the graduations with ordinary
sulfuric acid, and whirled for five minutes in a Babcock separator. The unsul-
Ionated residue multiplied by 2 gives the per cent by volume directly '(each
graduation being equal to 1/200 mI.). '
It is important that the acid should be of the proper strength. A mixture
of fuming sulfuric acid and ordinary concentrated sulfuric acid should be pre-
pared to contain exactly 80.07% of 80a, which constitutes 37 normal acid. If
the sulfonation residue is dark in color, it should be treated with an excess of a
10% sodium hydroxide solution, and if completely soluble in this reagent, the
test is regarded as negative.
(B) Sulfonation Residue Insoluble in Water.-Treat 3 g. of the carbon disul-
fide or benzol soluble portion of the substance, after evaporation of the solvent,
with 6 ml. of' concentrated sulfuric acid (95%) in a stout test-tube and 'l,gitate
continuously for three-quarters of an hour while heating in boiling water. The
contents of the test-tube are then poured into 500 ml. of cold water. Brown-
coal tar and brown-coal-tar pitch are practically insoluble in cold water a11d
partly soluble in hot water. Coal-tar pitch, wood-tar pitch, oil-gas-tar pItch
and naphthol pitch are completely soluJ:.>le in cold water, forming a deep black
solution, which upon filtering leave little to no residue. Lignite-tar pitch, peat-
tar pitch, bone-tar pitch, rosin pitch, fatty-acid pitch, wool-fat pitch and
montan pitch are partly soluble in cold water. Asphalts (both natural, as
well as those derived froni petroleum) and asphaltites are insoluble in cold
water, forming a copious insoluble precipitate. Let the solution stand fO,r two
hours and without agitating the precipitate, decant carefully through a weighed
Gooch crucible of about 7 cm. diameter. Wash the residue with hot water,
and finally transfer the precipitate to the filter. Continue washing with water
until the filtrate no longer gives an acid reacti<;m wi~h methyl orange, then dry
at 105° C. and weigh. Marcusson used this method for determining the per-
centage of coal-tar pitch in admixture with refined Trinidad asphalt. He
found that upon treating the C8 2 soluble portion of this asphalt with sulfuric
acid as outlined above, a small amount of soluble sulfo-derivatives were formed,
so that an average correction of 4% must be made to obtain 'accurate figures.
The foregoing test h~s been modified for the ostensible purpose of yielding
more accurate results, as follows: 2 g. of the substance are weighed in a test-tube
3.5 cm. in diameter, which is immersed 12 cm. in a 30% solution of salt, main-
tained at 107-108° C., whereupon 20 g. of concentrated sulfuric acid (95%) are
added. A mechanical glass stirrer is introduced and caused to revolve at the
rate of 300 turns per minute and the digestion cOhtinued for 2Y2 hours. The
mechanical stirring is claimed to break up the lumps and result in a better
filtration. Carefully introduce water in the test-tube, to a point 12 qm. from
the bottom, and continue the stirring for another 15 minutes, then transfer the
contents of the tube into 1 liter of boiling water. Let stand over night and
carefully filter through two layers of tared filter. paper in'to a 9 cm. porcelain
filter, using suction. Wash the residue with boiling water until the filtrate is
EXAMINATION OF BITUMINOUS SUBSTANCES 1559
clear and no longer reacts acid with methyl orange. The filter containing
the residue is placed on a Petri dish and dried for 2 hours at 105° C., cooled in
a dessicator and weighed, This represents the asphaltic constituents present,
which, however, are contaminated with a certain proportion of insoluble sulpho-
derivatives of the asphalt. An examination of a range of petroleum asphalts,
as well as Trinidad asphalt, indicated that the weight of insoluble asphaltic
constituents with their associated insoluble sulfo-derivatives, as ascertained
by this method, must be corrected by deducting 20%, to obtain the net weight
of asphaltic constituents actually present. Similarly, an examination of a
number of coal-tar pitches showed that in no case was more than 0.1 % of '
insoluble sulfo-derivatives present. No correction need therefore be made for
pitches. Any free carbon present in the pitch (ascertained by the insoluble
matter separated upon originally dissolving the substance in carbon disulfide)
should be taken into account by using the following formula:
o
S
% Asphalt=80%(100-C)'E
where: C=percent free carbon present in the mixture; S=grams residue ob-
tained after sulfonation; E=grams mixture (after having been freed from
insoluble constituents and carbonaceous matter) taken for the sulfonation
process.
Saponifiable Constituents
Certain bituminous substances, such as montan wax, rosin pitch, and fatty-
acid pitch, are often composed largely of saponifiable constituents. Others,
including pine tar, pine-tar pitch, hardwood tar, hardwood-tar pitch, peat tar,
lignite tar, Done tar, bone-tar pitch and other forms of fatty-acid pitches, con-
tain smaller percentages. This test is also used for gauging the uniformity of
supply, and in the case of fatty-acid pitches, as a criterion of the quality.
The following procedure has been devised by the author for specifically examining
bituminous materials or admixtures of bituminous materials with animal or vegetable
oils and fats, since the customary' methods do not adapt themselves especially well,
due to the formation of troublesome emulsions. The bituminous material is first freed
from insoluble constituents, including any mineral matter, by boiling with benzene
under a reflux condenser, cooling and filtering through a Gooch crucible. . The insoluble
constituents are dried at 100° C. and weighed. Sufficient of the bituminous substance
should be taken to yield approximately 5.0 g. of extract. The benzene' solution is
evaporated or distilled to 50 ml. and 50 ml. of the saponifying liquid added from a
pipette. This should consist of a 10% solution of caustic potash, prepared by dissolving
100 g. of anhydrous potash in 500 ml. of 95% ethyl alcohol, and diluting to a litre with
90% benzene. The liquid is allowed to stand overnight to permit any carbonate to settle,
and the clear solution decanted. After the saponifying agent is added, the mixture is
boiled under a reflux condenser, for ~-1 hour, and the contents of the flask while still
warm poured in a separatory funnel containing 150 ml. of boiling water and 25 ml. of a
10% solution of potassium chloride. Add 250 ml. of benzene, agitate vigorously, and
allow the funnel to rest quietly in a warm place until the solvent separates. If an emul-
sion forms which refuses to se:parate on standing, add 200 ml. more benzene and 100 ml.
95% ethyl alcohol and stand III a warm place overnight. This will invariably effect a
more or less complete separation of the solvent. From this point on the method is
illustrated by the following tabular outline:
1560 EXAMINATION OF BITUMINOUS SUBSTANCES
Saponify as described.
Draw off the soap solution as completely as possible.
Decant the benzol layer, leaving the intermediate layer in the separatory funnel.
.
,--=-----Combine---- J
Transpose with dilute hydrochloric acid, warm and exhaust with benzol. Separate
the aqueous solution containing the glycerol and mineral salts. Evaporate the com-
bined benzol extracts to a small bulk, and then complete the evaporation of solvent at
1000 C. Cool and weigh. Weight equals the free acids derived from the saponijiable
constituents.
In the case of bituminous materials that are more or ,less completely saponifiable,
the intermediate layer is apt to be absent. In this, case the process will simplify. itself
considerably. The foregoing procedure will separate the unsaponifiable constituents
in practically an ash-free state.
Diazo Reaction
This test is used for identifying bituminous substal1.ces carrying phenols,
including wood tar and wood-tar pitch, oil-gas- and wate'r-gas-tar.s~ and pitches,
shale tar, peat- and lignite-tars and pitches, bone-tar, bone-tar pitch and the
:various coal-tar pitches.
This reaction is carried out by boiling 2 g. of the bituminous substance with
20 ml. N. aqueous caustic soda, for approximately five minutes. After 'cooling,
the liquid is filtered. If the filtrate is dark colored, it may be lightened by
shaking with finely pulverized sodium chloride and filtering. It is then cooled
in ice to 10° Q., and 'a few drops of freshly prepared diazobenzolchloride solution
added (prepared by dissolving 1 g. aniline chlorhydrate in 10 ml. water and 3 ml.
25% hydrochloric, acid, and then adding drop by drop, a saturated solution of
0.5 g. sodium nitrite in water). To avoid the preparation 'of diazobenzolchlo-
ride (which does not keep), the use of paradiaz6benzol-sulfonic acid (C 6 H 4 · -
N 2 SOa) may be substituted, since it is a more stable chemical and will keep well
EXAMINATION OF BITUMINOUS SUBSTANCES 1561
in' stock. If phenols are present" a more or less fugitive red coloration will
result, sometimes accompanied by a reddish precipitate.
Assuming that the bituminous substance gives the diazo reaction, the
question will often arise whether the product is a straight-distilled pitch, or an
asphalt "cut-back" with a high boiling-point distillate containing phenolic
bodies, derived from coal tar, lignite tar, etc. Marcusson has worked out a
method applicable under these circumstances, which consists in dissolving 10 g.
of the bituminous substance in 15 ml. of benzol, and pouring the solution into
200 ml. of 88° petroleum naphtha. The resulting precipitate is filtered and
washed with petroleum naphtha and dried. It is then boiled for fifteen minutes
with N /2 alcoholic caustic potash under a reflux condenser to extract the '
phenols. The liquid is cooled and filtered, the alcohol evaporated, and the
residue dissolved in water. Sodium chloride is added to clarify 'the liquid and
remove any substances imparting a dark color, the solution is filtered and the
filtrate treated for the diazo test described above. If a straight-distilled pitch
containing phenols is present, a positive reaction will be obtained. If the
origip.al substance gives the diazo test, but the residue treated in the above way
does not, then the admixture of high boiling-point oils containing phenolic
bodies with a substance free from phenols (e.g., asphalts, etc.) is established.
It is claimed that the presence of as little as 10% of pitch containing phenols
may be detected in this manner.
Where bituminous substances contain calcium carbonate, the phenolic
bodies present combine with the lime, forming insoluble calcium phenolate
which yields but a faiht diazo reaction. However, on treating such substances
with a solvent in the presence of hydrochloric acid, the calcium phenolate is
decomposed, and the diazo reaction becomes much more delicate.
A still more sensitive test for ascertaining the presence of phenols (e.g. tars
or pitches) in asphalts consists in the following: A reagent is prepared by dis-
solving 0.2 g. paranitraniline in 20 ml. water and 5 ml. 20% sulfuric acid, where-
upon 0.3 g. sodium nitrite are added. Hard asphalts (3 to 4 g.) are ground in
a mortar with 25 ml. water made alkaline with NaOH; then filtered and the
filtrate acidified with H 2S0 4 and 3 drops of the foregoing reagent added. The
solution is finally made alkaline with NaOH, whereupon the presence of phenols
will develop a more or less intense red coloration. It is claimed that the pres-
ence of 1 % of tar or pitch may thus be detected, inasmuch as a 1 : 25,000
solution of phenol will produce a blood-red color, and a 1 : 250,000 solution a
pinkish-red color. Pure asphalts, on the other hand, yield a light yellow color
or leave the reagent unchanged.
The Dutch Standardization Committee have adopted a combination of
paradiazobenzol-sulfonic acid with Millon's reagent for detecting the presence
of phenols in substances containing colophonium, as follows: 10 g. of the sub-
stance are boiled 20 minutes with 25 ml. N aqueous NaOH and filtered. To
5 ml. of the filtrate add a few drops of paradiazobenzol-sulfonic acid. If no
red coloration results, phenol (tar or pit()h) is absent, but if a coloration occurs,
tar or pitch may be present, but it, may also be caused by colophonium. In
this event, the remainder of the filtrate is almost neutralized with nitric acid
and evaporated to 5 ml. Then add 5 ml. freshly prepared Millon's reagent and
place in a glass beaker with boiling water. If no coloration results within half
an hour, the mixture is free from phenols.
1562 EXAMINATION OF BITUMINOUS SUBSTANCES
Anthraquinone Reaction
The anthraquinone reaction is used for 'detecting dnthracene in ta,.r products
produced at high temperatures, including oil-gas-ta~ and pitch, wd'ter-gas-tar'
and pitch, and the various coal-tar pitches. l'his test is therefore valuable for
purposes of identification. I
The tar or pitch is first subjected to distillation in accordance with the letort
method, the offtake and condensing tube being kept warm to prevent the accum-
ulation of any solid distillate. The distillate passing over between 270 0 and
355 0 C. is caught separately and examined for anthracene in the following
manner. The fraction is heated until it is thoroughly fluid to secure a uniform
sample, and 5 g. weighed out, while hot. After cooling, 10 m!. of abf'olute
ethyl alcohol are added, the solids allowed to crystallize and the liquid decanted.
One to 2 g. of solid substances containing the anthracene are dried on a water-
bath, transferred to a 500-ml. flask connected with a return condenser, 45 m!.
of glacial acetic acid added, and the contents boile<). for two hours. Thefollow-
ing mixture is then added drop by drop through a separatory funnel, viz.: 15 g.
of anhydrous chromic acid dissolved in 10 m!. of glacial acetic acid, and 10 m!.
of water. The boiling is 'continued for another two hours, the flask cooled, and
400 m!. cold water added. This treatment oxidizes the anthracene to anthra-
quinone, which on cooling separates as a s6lid mass. This is filtered, washed
with hot water, then with a hot 1 % solution of caustic soda and again with hot
water. The residue of anthraquinone is then dried and its weight multiplied
by 0.856 to obtain the corresponding weight of anthracene. From 0.25 to
0.75% of anthracene is found in coal tars, and a correspondingly larger percent-
age in coal-tar pitches.
A color reaction for establishing the presence of anthracene consists in boil-
ing for Yz to 1 hour the crystals of anthraquinone (1 part) with zinc dust (2
parts) and 50% NaOH solution (30 parts), whereupon an intense red-colored
solution is obtained, which on filtering in contact with air becomes decolorized.
ExAMINATION OF BITUMINOUS SUBSTANCES 1563
PART II
Resistance to Displacement
This test has been proposed for examining fine-aggregate paving mixtures
(e.g. the wearing-course of sheet-asphalt pavements) and has also been adapted
for testing coarse-aggregate paving mixtures. The procedure when testing
fine-aggregate mixtures consists in weighing the dry sand and filler to 0.1 g.;
heating to 3500 F. and mixing thoroughly. The melted asphaltic binder heated
to 3500 F. is then weighed. to 0.1 g. and thoroughly mixed with the aggregate
by means of a spatula. Approximately 100 g. of the completed mixture are
weighed into a cylindrical forming mold 2-in. in diameter and 4-in. high. The
mold and contents are heated to 250-260 0 F. as determined by the thermometer
used for stirring the mixture. Remove and insert a cylindrical plunger into
I+----r::~.~. of Compression
Machine
(diameter 2-in.)
t===k--- Clamp
.....__----Cirt;ular Washer
(opening 1.7S.in.)
._----Test Specimen
(2.in. diam.·b y I-in high)
FIG. 211.-Apparatus for Testing Resistance to Displacement.
the mold, and then tamp by first giving 60 blows with a tamper weighing 500 g.,
followed by 15--20 blows with a tamper weighing 1400 g. The mold and con-
tents are placed in a compression machine to which is applied a load of 3000-lbs.
1564 EXAMINATION OF BITUMINOUS SUBSTANCES
per sq. in. and the briquette allowed to cool under pr:essure. The compressed
briquette should measure 2-in. in diameter and appfl>ximately I-in. in height.
Let stand at room temperature over night and then, maintain in a water-bath,
at 140° F. for one hour before testing. Two such briquettes should be pre-
pared, each specimen being tested by placing it top lend down in the testing
mold and inserting the plunger as illustrated in Fig. I 211. A circular washer
having a circular opening 1.75 in. in diameter is clamped on the lower ,end of
the mold and the load then applied at a rate which will cause the testing head
to be depressed I-in. in 25 seconds. As the pressure increases, the briquette
will be forced through the orifice of the circular washer. The reading is ob-
served on a spring dyn!Lmometer and the maximum is recorded. The average
of two specimens is taken as a measure of the" stability" of the mixture. This
test indicates the resistance of the pavement to displacement and may be used
to proportion mixtures of sand, filler and asphaltic cement giving optimum
results.
For testing coarse-aggregate paving mixtures a larger mold is recommended
in conjunction with a testing ring % in. thick having a minimum orifice 5:J,i
in. in diameter for a depth of Ys in., after which it tapers out to a diameter of
6 in. at the bottom of the ring.
Samp~es of the finished mixture taken from the testing-plant are warmed to
250°-260° F. and formed into briquettes as described above. Finished pave·
, ments may be tested by cutting out cores 2 in. in diameter and approximately
134 in. high, which are then placed directly in the mold arid tested at 140° F.
Tensile Strength
Any standard testing machine may be used provided it is accurate to w1thin
1 % of the lowest load to be applied. Special specimen holders of steel, as
shown by Fig. 212 (a), shall be used. The specimen holders shall be attached
to the heads of the machine by links held so that the pull is central, to avoid
any transverse strain. For hot-molded materials an~ for plastic cold-molded
materials, the test specimen shall be molded to the fo,rm and dimensions shown
in Fig: 212 (b). .
NOTE.-To permit the use of one mold for all molding compounds the following
variations in dimensions of the test specimen will be permissible:
For hot-molded compounds ................... " ± 5%
For cold-mold~d compounds ................ , .. .,~. ±lO%
All specimens shall be conditioned before testing by drying in an oven at
50° C.±5° C. for 48 hrs., followed by cooling to room temperature in a dessi-
cator. Five specimens shall be tested at room temperature. The speed of
the head of the testing machine shall be such that the load can be accurately
weighed, but shall not exceed 0.050-in. (1.27 mm.) p,er minute with the machine
running idle. All tests shall be made at normal room temperature of about
20° C. (68° F.). .
The report shall include the following:
(a) The breaking load of each specimen in pounds or kilograms;
(b) The thickness and width of each specimen,in. inche's or millimeters as
measured by a micrometer at the center of the- sP6.Cimen, that is, the point of
minimum section;
EXAMINATION OF BITUMINOUS SUBSTANCES 1565
(c) The ultimate tensile strength in pounds per square inch or in kilograms
per square millimeter of each specimen, calculated from the minimum area
measured at the center of the test specimen before the load is applied;
(d) The character of the material tested, with description of the fracture
and its location .wit.h respect to the middle point of the specimen.
(e) The speed in inches or millimeters per minute at which the head of the
testing machine traveled during the test.
Compressive Strength
This method has similarly been standardized as follows:
Any standard testing machine may be used provided it· is accurate to
within 1 % of the lowest load for which it is used. One end of the specimen
1566 EXAMINATION OF BITUMINOUS SUBSTANCES
shall bear upon an accurately centered spherical bearing block, located when-
ever practicable at the top, and the metal bearing plate~ shall be directly in
contact with the ends of the test specimen. J
The test specimen shall be molded in the form of a cylmder having a height
of 1.25 in. (31.8 mm.) and a diameter of 1.125 in. (28.6 mm.) (Fig. 213). The
ends shall be flat and perpendicular to the axis, bein~
ground to this condition if neces.sary. _.
NOTE 1.-To permit the use of one mold for all molding"
compounds the following variations in dimensions of the test
specimen will be permissible: '
For hot-molded compounds .................... , ± 5%
For cold-molded compounds ..................... ±10%
!
' . ' -r," :; , NOTE 2.-Due to the size and shape of this specimen, it is
sometimes difficult to completely cure it within the relatively
short time which would ordinarily be used and with the tem-
. T""""
_:i:..
peratures ordinarily used. This is particularly true when hand
molds, that is, molds heated by conduction from the platens,
Courtesy A.S.T.M. are used. It is, therefore, recommended that special pre!)au-'
tions be taken to insure thoroughly cured parts, experiments
FIG. 213.-Compres- indicating that 20 to 30 minutes may be necessary in some cases
sion Test Specimen. with hand molds.
that may be retained at the top of the p!l;per and to break up any channels
that may have been formed by the carbon disulfide passing through. If the
washings show any color, the thimble shall be put back and extraction con-
tinued until the solvent again becomes colorless. .Itlshall"then be removed,
dried carefully, at a low temperature at first to prevent 'ignition of the absorbed
carbon disulfide, and finally to constant weight at 100° C. (212° F.), cooled
and weighed. ' .. ,
An aliquot portion of the solution in the flask shall be rinsed into a weighed
porcelain or silica evaporating dish or crucible and the solvent burned off under
a hood. The residue shall be ignited over a flame or in a muffle and the ash
weighed and the weight added to tl:).at of the mineral matter in the filter paper.
This is to correct for the fine mineral matter which will be carried through the
paper by the sol:vent. Should there be a consid'erable amount of ash recovered
in this way, and if it is found that the
mineral matter is calcium or other carbon-
ate, it shall be recarbonated by repeated
treatment with ammonium carbonate solu-
tion and finally ignited at a dull red heat.
Sec.tionD-D
Sec.tionC-C.
Ordinarily, however, the mineral matter
going through the paper will be so small in
@)
amount that the difference caused by igni-
tion may be neglected. The corrected loss
in weight on the original sample represents
the percentage of asphalt present.
Section B-~. (b) For Analysis of 50- to 500-g. Sam-
1<.•.• -••.• 6~··-····~ ples.-The apparatus for analysis of sam-
tV
ples containing coarse aggregate shall be .the
large extractor shown in Fig. 214, consisting
of a large brass cylinder, through the bot-
tom of which projects a Hk.candle-power
incandescent carbon filament bulb to supply
heat to the extra!ction apparatus proper,
which is held in the upper portion of the
FIG. 214.-Extraction Apparatus. cylinder. This apparatus is composed of a
cylindrical brass vessel for holding the sol-
vent, a cylindrical wire basket made of 80-mesh wire cloth suspended in the
cylinder, and an inverted 'conical condenser which serves as a top.
A large filter paper, 12 or 13 in. in diameter, shall be fitted inSIde the wire
basket of the extractor by folding once more than in ordinary filtering, or by
wrapping it over a form which fits inside the basket (a cylindrical bottle of
proper size makes a good form) and placing it inside the basket.
The basket with contained filter paper shall bel dried and weighed. The
sample shall be weighed and packed in the filter paper in the basket. Care
should be taken not to pack all coarse particles in one place and th~ fine particles
in another, Qut to have them mixed together in uniform proportions.
The sample shall be covered with a disk of felt or wad of absorbent cotton
to insure even distribution of the dropping solvent,_thus preventing it from
forming a channel through the sample. The baske~shall be suspended in the
extractor and 150 to 200 ml. of carbon disulfide poured over the felt or cotton.
EXAMINATION OF BITUMINOUS SUBSTANCES 1569
•
The condenser shall be placed over the top and water circulated through it.
Current shall be started through the electric lamp underneath the extractor
and the extraction carried on exactly as in the smaller glass extractor, but on a
larger scale. The time for extraction will vary from 3 to 12 or more hours,
depending upon the nature of the sample.
To determine when extraction is complete, the condenser shall be raised
and the basket lifted out to observe if drippings are clear. One or two drops
caught upon white filter paper should leave but a light stain.
The drying and weighing of the basket, burning off of the solution for CCl'-
recti on, and calculation of the weight of mineral matter shall be determined 8S
in the foregoing. Note that any water in chemical combination with the min-
eral constituents is ,not ascertained by this procedure, but the amount present
is usually so small as to be negligible.
Cold Extraction Method.-Weigh out 1000 g. of the well-disintegrated
sample into an 800 m!. beaker and add 500 m!. carbon disulfide. Stir five or
six times during the first 24 hours thoroughly to break up the aggregate, and
then let stand quietly for 48 hours, after which the solution shall be carefully
decanted and filtered through a series of three 100 ml. funnels, using absorbent
cotton as the filtering medium. The aggregate left in the beaker shall be sub-
jected to a second washing, conducted as described. The extracts shall be
combined and placed in a quart jar, which shall be tightly covered and allowed
to stand quietly for 48 to 72 hours, after which the solution shall be filtered
through an asbestos mat in a
Gooch crucible to remove the last
traces of sediment. The mineral
ingredients shall then be dried and
weighed as 'previously described.
In the case of mastic block and
similar bituminized'aggregates con-
taining a high percentage of finely-
divided mineral matter, there is apt
to be a loss of fine material, caused
by· carbon disulfide solution creep-
ing behind the folds of the filter
paper and over its edge, rather than
through the paper. In the case of
mastic the following procedure
is recommended: Two 32-cm. No.
5 Whatman filter papers, after FIG. 215.-Method of Folding Filter-papers
being dried in an oven and cooled for Extraction Test.
in a desiccator, are counterpoised.
These papers are folded together and fastened with a paper-clip in the
manner shown in Fig. 215, .whereupon they are placed in a glass funnel
7-in. in diameter, the top edge of which is ground flat. The funnel is covered
with an 8-in. diameter glass plate, in the center of whioh a %-in. diameter hole
has been drilled. To make certain that the plate completely closes the funnel,
it is advisable to secure the plate to the funnel with gelatine cement, prepared
by dissolving 10 g. of gelatine in 80 m!. water and adding 20 g. glycerin. The
mastic block is warmed in an oven and broken into small pieces. Two 75 g.
1570 EXAMINATION OF BiTUMINOUS SUBSTANCES
represent~tive samples should be taken alld run in- duplicate. One sample is
introduced into the counterpoised filter and- the fUllIl;el covered with the glass
plate as described. Carbon disulfide is introduced directly into the funnel.
through a 1500 ml. separatory funnel as shown in Fip. 216. The addition of
carbon disulfide may be ~ade automatic, so that
when the level of the liquiddn the filter falls below a
predetermined height, a fresl\. supply will be admitted,
and this may continue overnight without attention.
When the filtrate comes through colorless, the carbon
disulfide supply is stopped and the paper allowed
to dry, first in air, and finally in an oven at 1200 C.
for one hour. After cooling in a ~esiccator, the filter
and contents are weighed, the outside pap~r being
placed on the weight-pan as a counterpoise. The
weight of asphalt extracted is ascertained by sub-
tracting the weight of the aggregate from the weight
of the sample taken, and its percentage calculated.
For products other than mastics, the following
weights of material are recommended: surfacings
containing aggregate all passing 10-mesh: two 50 g.
samples; surfacings containing up to 25% U-in. or
%-in. aggregate: two 150 g. samples; surfacings con-
taining up to 25% %-in. aggregate: two 300 g.
samples; surfacings containing up to 50% %-in.
aggregate: two 500 g. samples.
Another variation consists in covering the
coarsely broken sample with carbon disulfide. and
allowing same to stand quietly overnight, whereupon
it will be found that most of the soluble material
has gone into solution. The residue is then w.ashed
with 8 to 10 portions o£ carbon disulfide. Fine
mineral matter is removea. from the combined ex-
tracts by filtering or 'centrifuging. .
Centrifugal Extraction Method.-The most effi-
cient apparatus of this type is illustrated in Figs.
217 and 218. It con'sists of a 1/5 h.p. vertical motor
, a, capable of making 1100 revolutio_ns per minute at
FI~ofl~xt!~b~~a*:s[~r 110 volts, with either direct or alternating current.
Its shaft projects into a cylindrical copper vessel b,
having a concave bottom and draining into the spout- c. A circular brass plate
d, 9Yz in. in diameter supports an inverted iron bowl e, 8Yz in. in diameter by
2 156 in. high, having a 2-in. circular hole at the top. A brass cup f is fastened to
the inner side of the bowl, -having a circle of Ys i~. holes for the 'admission of
solvent, and terminating in a hollow axle which- fits snugly through a hole in
the center 9f the brass plate d. A felt ring g, % in. wide and about 0.090 in.
thick (cut from No. 80 roofing felt) is firmly pressed against the bowl by the
milled nut h for which the hollow axle is suitably threadecr. The axle in turn
fits snugly over the shaft of the motor, to wh~ch:tt -is secured by a slot and
cross-pin.
EXAMINATION OF BITUMINOUS SUBSTANCES 1571
. I
Procedure.-The filter ring shall be dried for 30 minutes in an oven at 212 F.,
0
the bowl and fastened down fairly tight. The bowl shall then be placed in
position in the rotarex machin~ and from 40 to 50 ml. of carbon disulfide po_ured
into the bowl through the funnel opening. After allowing the material to
digest for about 15 minutes, the motor shall be started, slowly at first, and then
the speed increas,ed sufficiently to cause the dissolved bitumen to flow from the
spout in a thin, stream. The dissolved bitumen or filtrate shall be collected in
a beaker or other receptacle. When the first charge has drained, the motor
shall be stopped and a fresh charge of carbon disulfide added and again allowed
to digest for 15 minutes and the machine started as before. This operation
shall be repeated from three to five times or until the solvent runs clear from
the spout showing all bitumen has been dissolved from the sample, this usually
requiring three to four washings.
When the last addition of solvent has drained off, the bowl shall be removed
and placed with the bowl plate and filter ring on a sheet of glazed manila paper
and allowed to dry at room temperature. The aggregate shall then be brushed
,from the bowl onto the paper together with any aggregate adhering to the bowl
plate,and filter ring. The aggregate shall then be transferred to a tared watch
glass or pan and together with the filter ring dried in an oven at 212 0 F.,
cooled in a desiccator and weighed. '
In making the correction for the mineral" matter suspended in the filtrate,
add sufficient carbon disulfide to bring the same to a definite volume and a
multiple of 100 ml. The whole filtrate shall then be thoroughly shaken so that
the mineral matter is uniformly dispersed throughout the filtrate and an aliquot
1572 EXAMINATION OF BITUMINOUE\ SUBSTANCES
part of 100 ml. then taken and transferred into a weighed porcelain or silica
evaporating dish and the solvent burned off under a hood. The' residue shall
then be ignited over a gas flame or in a muffle until it ~s entirely free from carbon,
cooled and weighed and the total ash correction det~rmined as follows:
PART III
Burlap Paper
Light Asbes-
and and
Paper Burlap Duck Cotton Rag-Felt tos
Rag or Light
Fabric Felt
Asbestos Cotton
Felt Fabric
-..,--1----.--1---,- - - - - - -
a a a a a a a a
------1--------------- - - - - - - - - ---
Single Layered:
Saturated only.. WX WX U U TUY TU T
Coated only (one
or two sides). W U U S
Saturated and
coated....... X U U QS ... V QR Q
Laminated
(Bituminated) :
Layers saturated
only. . . . . . . .. .... Z Z Q QT UZ U
Layers saturated
and coated. . . Q Q Q
Layers unsatu-
rated ........ W W W W
One layer un-
saturated and
others satu-
rated. . . . . . .. w w ......... ... ... ... ...... .... .... ... .... ... W W
The finished products falling in this cl~ss include sheet roofings, floor cover-
ings, waterproof membranes, sheathing and -insulating papers, expansion •
joints involving the use of woven or felted fabrics, ~lectrical insulating tap~,
and certain tYpes of wall board. As these are constructed in many different
ways, 'it will obviously be impracticable to describ~ in detail the analytical
methods applicable to each. The ones which follow have been devised spe-
~ifically for examining prepared roofings, but with these as a starting-point,
others may readily be evolved for testing floor coverings, waterproof mem-
branes, sheathing and insulating papers, etc.
This must check within 1% of the average weight per unit area of shipment.
If the sample fails to do this, then additional samples shall be cut from the
same roll, until one is obtained which does, which sample shall be kept for
further examination.
NOTE.-As a referee method or in case any dispute arises regarding the properties
as may be ascertained from the particular sample selected, a 30-in. sample shall be taken
and examined separately from each roll sampled as above.
r--f"T---'-rr---1T
r------, I I I :
L------JGIGrl , l
L n J liZ 1 ,:
: ,------1 1 I ' I!
;s·I---Y3---L1..lGII
1- _____ _' 1 13 I/~' bmml@l1lm
III!
I , ,'I I 1:
r--F4---JGI GI-J I:
r------J 4 151 I !
L ____ _,.J I I I :
L_Y..? __ LL.L ____J....':(.
k············· /2" .•...••.•. l>l
•.•.••., •••........• 18"··········
Each roll shall be unwound and the workmanship and finish of the roofing
observed. Recor!i the weight of any loose surfacing that may fall off in so
doing. ,
The length and width of each roll shall be measured to the nearest >i-in.,
and the square feet of material contained in each roll shall be calculated. From
these values, the average area of the rolls examined ,shall be calculated.
From the above net weight, exclusive of loose surfacing, of roofing per 108
sq. ft. contained in each roll shall be calculated and the minimum weight per
108 sq. ·ft. of the lot.
The average weight per 108 sq. ft. for the rolls examined shall be calculated.
This shall be regarded as the average weight per 108 sq. ft. of the lot .... · CIa)
From the rolls examined, the one whose weight per 108 sq. ft. is nearest the
average weight of the lot shall be selected. The first convolution or two shall
be unrolled, and with a knife and straight edge, the sheet shall be cut across
at right angles to the edges. A section measuring exactly 30 in. in the direction
of the roll's length shall then be removed. The width of this section shall be
measured to the nearest l2 in. and weighed, neglecting any loose s:urfacing.
The weight in pounds per 108 sq. ft. shall then be calculated from the following
formula:
= Weight of 30-in. section in ounces 324 Clb)
P oun d s per 108 sq. f t. W'd h f 30 . t'"
1 to-In. sec IOn m mc es
h X . .
The weight so determined shall check within 1.5% of the average weight per
108 sq. ft. of the lot CIa). If this sample fails to check, then additional samples
shall be cut from the same roll, until one is obtained which does, which sample
shall be kept for further examination.
NOTE.-As a referee method, or in case any dispute arises regarding the properties as
may be ascertained from the particular sample selected, a 30-in. sample shall be taken and
examined separately from each roll sampled.
(c) For Mineral-surfaced Shingles.-Each bundle selected shall be weighed
to the nearest >i lb. The packing material shall be w~ighed to the nearest ~ lb.
and the weight recorded separately for each bundle. The average weight of
packing material per" square" shall be recorded.
The shingles in each bundle, "a," shall be' counted and the workmanship
and finish of the shingles shall be observed. ,Record the weight of any loose
surfacing that may fall off in so doing.
From the above the net weight of each bundle, exclUSIve of loose surfacing,
shall be calculated, "b," The dimensions of two representative shingles shall
be accurately measured and the net area of material in each shingle calculated
in square feet, "c."
The area in square feet of material contained in each bundle (aXc) shall
be calculated. The weight of roofing material per ,108 sq. ft. contained in each
bundle (_b_ X 108) shall be calculated, and the minimum weight per 108
aXc
sq. ft. recorded. This shall be regarded as the minimum weight per 108 sq. ft.
of the lot.
The average weight Cd) per 108 sq. ft. for th~, bundles' examined shall be
calculated. This shall be regarded as the averagE! Feight 'per 108 sq. ft. of the
entire shipment. , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. CIa)
EXAMINATION OF BITUMINOUS SUBSTANCES 1577
.
The average weight per shingle (C1~) shall be calculated ......... (lb)
From the bundles selected, a definite number of representative shingles,
whose weight shall fall within 1.5% of the average weight ascertained in (Ia)
shall be taken for further examination. The number of shingles so selected
shall represent as closely as possible 6 sq. ft. of material, and shall be selected
as far as possible from different bundles.
NOTE.-As a referee method, or in case any dispute arises regarding the properties as
may be ascertained from the particular sample selected, then a similar sample shall be
taken and examined separately from each bundle of shingles sampled.
Thickness
The thickness of eacR specimen shall be measured at 10 equally spaced
points, I in. from a cut edge, with a micrometer gage having flat bearing sur-
faces at both contact points of not less than Vz in. in diameter and preferably
of tl).e spring-and-dial type. The readings for each roll sampled shall be aver-
aged and the minimum, maximum and average thickness calculated to the
nearest 0.001 in.
Resistance to Heat
(a) For Asphalt-saturated Fabrics.-Ftom each specimen, 12 by 6-in. strips
shall be cut at K-1 and K-2 in Fig. 219., care being taken not to disturb any of
the detached surfacing. Each strip shall be weighed and suspended in the
center of an air-oven maintained at 221° F. (105° C.)±5° F. (3° C.) with a thin
wire fastened through holes punctured near one edge.' The thermometer shall
be inserted in the oven to such a depth that its bulb will be in line with the
center of the specimens. The strips shall be kept in the oven for exactly five
hours, then cooled and removed carefully, and each specimen weighed. The
average loss shall be calculated as a percentage and the average percentage of
moisture deducted. The final figure shall represent the average loss on heating,
exclusive of moisture.
(b) For Coal-tar Saturated Felt Only.-From the rolls making up the sam-
ple, strips 6 in. wide shall be cut across the sheet. Sufficient of these shall be
taken to make up a sample of 250 g.±5.0 g. These strips then shall be rolled ,
up and placed in the wire basket of the extraction apparatus described on p.1568.
The sample shall then be covered with a disk of soft filter paper to insure an even
distribution of the solvent.
The carbon disulfide extract containing the bitumen shall be transferred to
a 500-mI., short-neck; round-bottom flask. The flask shall be equipped with
a Hempel column, 16 cm. in length, 15 mm. in inside diameter and filled to a
depth of 5 cm. with glass beads. The Hempel column shall be connected with
a water-cooled condenser. A -2 to +80° C. thermometer shall be placed in
the Hempel column so that the top of the bulb shall be opposite the bottom of
the tubulature. Distillation for the removal of carbon disulfide shall be con-
tinued by means of a steam bath until drops cease to fall from the condenser.
The distillation then shall be carried out over a small flame, heating carefully
and observing the thermometer until the distillation ceases and the temperature
of the vapor as indicated by the thermometer shows no further rise on slightly
increasing the flame.
1578 EXAMINATION OF BITUMINOUS SUBSTANCES
I
When the contents of the flask are sufficiently cool to be poured without
appreciable vaporizing, 100 g.±O.l g. shall be distilleli as on p.1533.
Fractions shall be taken at 210 0 C. and 235 0 C. and the distillation stopped
at that point. The per cent of distillate by weight at each teni"perature shall
be calculated on the basis of the weight of sample tar.!en for distill!ition.
(c) For Asphalt Roll-roofings and Shingles.-Two; strips shall be cut from
the sample of roofing or shingles, each approximately ,8 by 8 in. They shall be
weighed and suspended vertically in the same direction as the material would
be applied to the roof, in the center of an air-oven maintained at 1760 F. (80 0 C.)
±5° F. (3 0 C.) with a thin wire fastened through holes punctured near one
edge. The internal dimensions of the oven shall not he less than 12 by 12 by
12 in. An electrically heated oven is recommended. The thermometer shall
be inserted in the oven to such a depth that its bulb will be in line with the
center of the specimens. The strips shall be maintained at the prescribed
temperature for exactly two hours, then cooled in a desiccator and each speci-
men weighed. The average loss shall be calculated as a percentage. Any
change in appearance of the specimen shall be recorded, such as flowing, sagging,
blistering or absorption of the coatings; also' sliding of granular surfacing.
Moisture
From each specimen, 2-in. strips shall be cut, as shown at A-I and A-2 in
Fig. 219. The' 2-in. strips shall be cut into I-in. squares. About '50 g.,
selected at random, shall be accurately weighed and distilled with 100 ml. of a
coal-tar distillate of which 5 to 10% boils below ioO o (ij. (212 0 F.) and of which
at least 90% distills up to 180 0 C. (356 0 F.), in an apparatus of the form and
under thc"'conditions prescribed on p. 1553. The flask containing the solvent
and sample shall be heated in a paraffin bath at a temperature of'302 to 3380 F.
(150 to 170 0 C.) until no further water passes over. The distillate shall be
allowed to stand in the graduated receiver until the water separates, whereupon
the volume of water shall· be measured and its weight calCUlated. From this,
the average percentage of moisture in the fabric, as received, shall be calculated.
From each roll sampled, a 2-in. strip (within l2 in.) shall be cut across the
specimen, as shown at H in Fig. 219. /
Each strip shall be extracted with carbon disulfide in an extractor shown in
Fig. 219, the .extraction being continued for several hours after the drippings
hav:e become- colorless. The desaturated fabric shall be removed and heated in
a ventilated oven at 225 0 F. (107.1 0 C.), cooled in a desiccator, any adhering
comminuted surfacing being brushed off and reta!9-oo,- and weighed as rapidly
as possible. The heating shall be repeated until the weight of the fabric re-
EXAMINATION OF BITUMINOUS SUBSTANCES 1579
this manner shall be accurately weighed, and incinerated i.p. a weighed porcelain
or quartz crucible either over an open flame or in a muffle, until all the carbon is
consumed. A few drops of ammonium carbonate solution shall be added, the
sample ignited gently, and weighed. The percentage pf ash shall be calculated'
on the basis of the moisture-free felt. I
I
NOTE,-As a referee method, or in case greater accuracy is desired, the three portions
taken from each specimen roll shall be weighed and ignited separately. The minimum,
maximum and average of ash present shall be calculated on the basis of the moisture-free
felt.
'....VPE
.~
~C b-
a'lffI.
rnmm71TTmm71nr __
i
-Fine dusfinfj nnisn
--.Topbif~minOlJscoafing
--Btfumlnlzedfelt
-'r&Hom hifuminous coafin~'
. --,eFine dusfing finiSh
The test strips for the separation of prepared roofing and shingles into their
component parts shall be prepared as follows:
From the samples representing the average weight a number of strips shall
be cut, each measuring approximately 2-in. in width by 8 in. in length. These
shall be weighed, the length and width measured to within 1 mm. or lz in., and
the weight calculated per 108 sq. ft. Any of the 2 by 8-in. strips whose weight
'varies more than 1.5% from the average weight of the shipment, shall be
rejecteq, and the remaining strips shall be retained for further examination.
NOTE.-If a strip is cut to exactly 2 by 7.46 in. its weight in grams divided by two
gives directly the weight in pounds per 108 sq .. ft.
Detached Mineral Matter (A-l).-Remove the detached mineral particles
from both sides of the 2-in. strips with a jet of compressed air operating under
a nozzle pressure of 25 lb. per sq. in. and reweigh in grams. Make a correction
to include any loose mineral matter. Calculate the weight in pounds per 108
sq. ft. of the detached mineral matter (A .. l) . ....................... " (2)
Moisture-free Felt (C-l).-Eight of the 2 in. by 8-in. strips shall be ex-
tracted with benzol (boiling-point 80 to 82° C.), or C.P. carbon disulfide in a
suitable extractor or centrifuge, and the extraction continued until the drippings
have become colorless. The desaturated felt shall be removed from the ex-
tractor, being careful not to lose any adhering mineral matter, placed in a tared
weighing bottle provided with a ground-glass stopper, heated in a ventilated
oven at 225° F. (107° C.)±5° F. (3° C.) for thirty minutes~ the stopper in-
serted, then cooled and weighed. The heating shall be repeated until the
weight of the felt remains constant, as determined by two consecutive weighings
taken not less than ten minutes apart, which shall show a further loss of not
more than 0.1 %. Any adhering mineral matter shall be brushed off the felt,
weighed and retained for further examination. The weight of said adhering
mineral matter shall be deducted to obtain the weight of the moisture-free felt.
The extract and accompanying mineral matter shall be retained for further
1582 EXAMINATION OF BITUMINOUS .SUBSTANCES
examination. From the weights so obtained ·and the respective areas of the
strips of desaturated felt" the weight of moisture-free f~lt shall be computed in
pounds per 108 sq. ft., from the following formulas and the results averaged:" (3)
. P d
oun s per
108 it
sq.. A
Weight of moisture-free
f h . .
in grams
.
flit
2212
X.
, rea 0 t e stnps m square centIm~ters
or
Pounds per 108 sq. ft. Weight of moisture-free felt in grams X34 3
Area of the strips in square inches .
The extracted felt shall be retained for examination as to weight, thickness,
strength, etc.
Ash in Moisture-free Felt (A-6).-The strips of desaturated felt obtained
as described shall be cut into .squares about 2 cm. on each side and thoroughly
mixed. Pieces amounting to about 10 g. shall be selected at random and dried
at 225 0 F. (107 0 C.) for thirty minutes. They shall be cooled in a desiccator,
accurately weighed and then incinerated in a weighed crucible either over an
open flame or in a muffle until all the carbon has been consumed. A few drops
of ammonium carbonate solution shall be added, the sample ignited gently,
cooled and reweighed. The percentage of ash. shall be calculated on the basis
of the moisture-free felt. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. (4)
Bituminous Saturation (B-3) in Moisture-free Felt.-From the 2 by 8-in.
strips, cut 40 strips about %: by 2 in. and weigh, this being a convenient size
to handle. These strips shall then be separated into three horizontal sections.
The small arrows a and b, to the left of the various types of roofing iUustrated in
Fig. 220., indicate approximately where the layers should be separated. The
top bituminous coating, with the embedded mineral matter, shall be removed
first.
NOTE.-The purpose shall be to remove the entire coatings with some of the satu-
rated felt adhering and no coatings left on the central section of saturated felt. The
operator may accomplish this in any way that is most conv,enient. Care shall be taken
to preserve the mineral surfacing with the layers to' whicH it belongs if it becomes de-
tached during the operation. A convenient manner to effect the separation consists
in warming the strips in an air-oven at a temperature of approximately 1760 F. (80 0 C.);
then with a knife, the front and back coatings, respectively, shall be peeled off, care being
taken to remove as little as possible of the saturated felt, and at the same time to make
sure that all of the coatings and surfacings are stripped from the felt.
These three horizont~l sections thus obtained shall b~ \veighe.d separately
and the total weight should agree within 1% with the original weight of the
strips which have been separated. The saturated felt (middle section) shall
be extracted with benzol (boiling point 80 0 to 820 C.) or c.p. carbon disulfide,
in a suitable extractor or centrifuge and the extraction continued until the
washings have become colorless. The desaturated felt shall be removed from
the extractor, placed in a tared weighing bottle p~ovided with a ground-glass
stopper, heated in a ventilated oven at 225° F.±5° F. (107° C.±3° C.). for
30 minutes, the stopper inserted, then cooled and weighed. The heating shall
be repeated~ntil the weight of the felt remains constant as,determined by two
consecutive weighings taken with an intervening.:,...heating period of not less
than 10 minutes, which shall show a further lo~ in weight of not more than
0.1 %. Calculate the weight of bituminous saturation by difference. The
EXAMINATION OF BITUMINOUS SUBSTANCES 1583
in an oven at 212 0 F. (1000 C.) for 30 minutes. The coatings shall be scraped
off by holding a dull knife at right angles to the strip of roofing supported on a
firm level surface and drawing the blade sideways under moderate pressure.
Care shall be taken to avoid scraping entirely throu~h the surface coating.
This is important. The scrapings shall then be weighed 'and dissolved in benzol
, (boiling point 80 to 82 0 C.) or c.p. carbon disulfide; and the mineral matter
separated by filtering or centrifuging and washing with successive portions of
solvent. The mineral matter shall then be dried at 2250 F. (107 0 C.), cooled in
a desiccator and weighed. Calculate the percentage of mineral matter l::J{tsed
on the original weight of roofing strips taken, which should check with that
obtained previously. The bituminous matter in the scrapings shall be calcu-
lated by difference and reduced to a percentage basis on the original roofing.
strips used. The bituminous coating shall be recovered by evaporating off the
solvent as described. Both surface coatings should be treated separately in
this manner.
NOTE.-Use the separated bituminous coatings Jor examining their physical and
chemical characteristics.
Calculate the percentage by weight of .fine mineral matter admixed with the
top coating (Type X) ............................................ (l1a)
Calculate the percentage by weight of fine mineral matter admixed with the
bottom coating (Types X and Y) ................................. (lIb)
Calculate the weight in pounds per 108 sq. ft. of fine mineral matter admixed
with the top coating (Type X) i.e. (7 a) X (lla) .................... " (12a)
Calculate the weight in pounds per 108 sq. ft. of fine mineral matter admixed
with the bottom coating (Types X and Y) i.e. (7b) X (lIb) . ........... (12b)
Calculate the weight in pounds per 108 sq. ft. of fine mineral matter em-
bedded in the top coating (Type X) i.e. (8a) - (12a) . . . . . . . . . . . . . . . .. (13a)
Calculate the weight in pounds per 108 sq. ft. of fine mineral matter em-
bedded in the bottom coating (Types X and Y) i.e. (8b) -(12b) . ...... (13b)
Compilation of Results.-The following items shall be reported in pounds
per 108 sq. ft ..:
Type X TypeY
Minera~ Matter:
Fine mineral matter embedded in top coating ........... Item (13a) .0.0 ••••••
PART IV
I
EXAMINATION OF BITUMINOUS-SOLVENT COMPOSITIONS
I
generator; a vapor outlet tube which extends from beneath the rubber stopper
to a point sufficiently above the distilling flask that will permit convenient
connection to the condenser; and with a thermometer. The steam tubing
should be not less than 2 nor more than 4 mm. in internal diameter and the
vapor outlet tube should be not less than 5 mm. in internal diameter.
The condenser shall consist of a 196 -in·. (14.29 mm.) outside diameter No. 20
Stubbs' Gauge seamless brass tube, 22 in. (55.88 cm.) long. It shall be set
at an angle of 750 from the perpendicular and be surrounded with a cooling
bath 15 in. (38.1 cm.) long, approximately 4 in. (10.16 cm.) wide by 6 in. (15.24
cm.) high. The lower end of the condenser tube shall be cu.t off at an acute
angle, and curved downward for a length of 3 in. (7.62 cm.) 'and slightly back-
ward so as to insure contact with the wall of the graduate at a point 1 to lU in.
(2.54 to 3.18 cm.) below·the top of the graduate when it is in position to receive
·the distillate.
A separatory funnel having a capacity of not less than 500 m!. should be
provided. Accessories consist of suitable ring stands for supporting the steam
generator, distilling flask, bath for distilling flask, separatory funnel, and a
thermometer.
Courtesy A.S.T.M.
FIG. 221.-Assembly of Distillation Apparatus.
When the distillation is finished, the water shall be separated from the distiJlate
and the distillate measured and retained for further tests, if required b;} the
specifications. ,
In some cases, the distillate does not separate read~ly from the water, and
this separation can be facilitated by tl;le addition of sodium chloride, which will
result in a sufficient difference in gravity to produce a !clear separation of the'
two layers.
The results shall be reported in per cent by weight or volume as required
by the specifications, based on the weight of the sample taken. '
NOTE.-In the distillation of varnishes the end point differs with the various types
and compositions. In all cases, however, the rate of distillation slows down as the end
point is approached; and, when the end point is reached, acrolein (formed in the decom-
position of the varnish base) is evolved and may be detected by its characteristic odor.
In addition to the evolution of acrolein, the following phenomena are suitable indi- .
cators of end point: .
(1) When the end point of the solvent is considerably lower than the initial boiling
point of the base, the end point of the distillation is indicated by the temperature reaching
a maximum and then starting to fall consistently.
(2) When the end point of the solvent and the initial boiling point of the base are too
close, or overlap, the above temperature drop does not occur. The end point of the
distillation is then indicated by a sudden foaming up of the residual base in the distillation
flask; or, if the distillation has been accompanied by foaming, the end point is indicated
by a breaking of surface tension with a consequent cessation of foaming.
The specific gravity of the 'Volatile matter (solvent) recovered by distillation
shall be determined in accordance with the procedure described below.
The difference between 100% and the percentage of non-volatile matter
determined above shall be calculated and recorded as the percentage of volatile
matter by weight. , , .
The percentage of volatile matter by volume shall be calculated from the
formula:
X=-
AC
B
where
x=the percentage of volatile matter by volume;
A =the specific gravity of the original varnis4;
B = the specific gravity of the volatile matter (solv~nt) recovered by.
distillation,
C = the percentage of volatile matter by weight.
With proper care and attention to detail in making this test, differences
occurring between different laboratories should' not exceed 2% for varnishes
containing solvents which have end points not over 235° C. (455° F.), and
s~ould not exceed 4% when the end point of the so~~ent is as high as 307.2° C.
(585° F.).
To correct the loss of distillate for water-soluble constituents, weigh out
another 100 g. into a 250-ml. flask and distil without steam over an electric
stove. Continue the distillation until the residue ip. the flask reaches a tem-
perature of 200° C. This gives somewhat lowel: resnlts than the first method,
but the distillate should be tested for water-sohible substances to correct the
EXAMINATIGN OF BITUMINOUS SUBSTANCES 1589
Dissolve 50 g. in 150 ml. benzol. Add 10 ml. dilute nitric acid (1 : 1j and boil under a reflux condenser
for one-half hour to decompose any metallic soaps (i.e., driers,.etc.). Add 150 ml. water, boil under reflux
condenser, transfer to a separatory funnel, draw off the aqueous layeri' boil with another 100 rul. water,
and repeat if necessary until all the metals are removed.' I
I
Benzol Solution: I I Aqueous Extrac~;
Distil to 100 ml., add 300 ml. of the saponifying liquid *, boil under reflux condenser Contains the
for one hour, and separate the unsaponilUi.ble and saponifiable constituents as described. metallic bases as
nitrates. Examine
qualitatively and
Unsaponifiable Matter: Saponifiable Matter: Aqueous Layer: then quantitative-
Iy for lead, man-
Examine a small portion. If Separate the fatty and resin Determine per- ganese, cob al t,
higher alcohols are present, sepa- acids as described. centageglycerol. zinc, calcium, and
rate the balance into: Multiply this by magnesium.
10 to estimate (N:B.-Tlie last
pet cent of vege- three used for har-
Hydrocarbons: Higher Alcohols, Fatty Acid.: Resin Acid.: table or animal dening rosin. The
Etc.: oils or fats (tri-
Contain the Include acids Include acids glycerides) pres- mettallic dryers
bituminous sub- Contain cho- derived from derived from en t in the origi- should not ~e
stances (i.e., as- lesterol, etc., de- vegetable and rosin and the nal substance. found by ignitl n,
phalt, coal-tar rived from wool animal oils or fossil resins. since the lead will
pitch, unsap. grease, also the fats, also from be reduced to
matter derived unsaponifiable fatty-acid pitch, metal by the or-
from fatty-acid consti tuen ts ganic matter, and
pit~h, etc.). originally pres"
ent in resins (4
- volatilized.)
to 8%).
* The saponifying liquid shall consist of a 10% solution of KOH dissolved in equal
parts of 95% C2H 50H or 90% benzol. / .
EXAMINATION OF BITUMINOUS SUBSTANCES 1591
PART V
Settlement Test
The following apparatus will be required:
(a) Cylinders.-Two glass cylinders of 500-ml. capacity with pressed or '
molded glass 'base and cork or glass stoppers. The outside diameter shall be
5.0 cm.±0.5 cm. and the cylinders shall be graduated at each 5-ml. interval
to the 500-m!. mark.
(b) Glass Pipette.-A syphon, glass tube pipette, 60-m!' capacity, form
optional.
A 500-m!. sample, :representative of the emulsion, shall be placed in each of
two glass cylinders. The cylinder shall be stoppered air-tight and stood aside
unmolested, at laboratory air temperature for 5 days. After standing for this
5-day period, approximately the first 55 ml. of emulsion shall be removed by
means of the pipette or syphon from the top of each cylinder without disturbing
the balance of their contents. Exactly 50 g. of each of the two samples, after
each has been thoroughly mixed separately, shall be weighed into separate
600-ml. low-form glass beakers and the asphaltic residue determined by
evaporation at 3250 F. (163 0 C.) fQr 3 h:r:. in the apparatus described.
After removal of the first sample, approximately the next 390 ml. shall be
syphoned off from each of the cylinders. The residue remaining in the cylinders
shall be mixed thoroughly and exactly 50 g. shall be weighed out from each of
them and the amount of asphaltic residue (all sediment, if any, included) shall
be determined by evaporation as described above for the two top samples.
The numerical difference between the .average percentage of asphaltic
residue from the two top samples and the average percentage found in the two
bottom samples shall be recorded.
Demulsibility Test
This test indicates the rapidity with which the dispersion "breaks" during
use, and may be ascertained in accordance with the following procedure:
The following apparatus and reagents will be reQ.uired:
(a) Sieves.-Three No. 14 sieves of the U. S. Standard Sieve Series, of iron
wire cloth, unframed, approximately 5-in. square.
(b) Beakers.-Three glass beakers of 600-m!. capacity each.
(c) Glass Rods.-Three glass rods, rounded ends, approximately 156 in. in
diameter.
(d) Burette.-A 50-ml. glass burette graduated in 0.1 ml.
(e) Calcium Chloride~olution (0.02 N).
(f) Calcium Chloride Solution (0.10 N).
The percentage of residue shall be determined by distillation as described
on p. 1593.
1592 EXAMINATION OF BITUMINOUS, SUBSTANCES
The weight of each assembly of beaker, rod and sieve shall be recorded.
Exactly 100 g. of the emulsifi~d asphalt shall be wJighed into each of three
600-ml. tared beakers. Over a period of approximately 2 min., 35 ml. of 0.02 N
CaCh solution (if quick-setting emulsion is being te~ted) or 50 m!. of 0;10 N
CaCh solution (if mixing type emulsion is being testeii) shall be added to each
beaker. from a burette. While adding the solution qf CaCh, the contents of
the beaker shall be stirred continuously and vigorously, kneading lumps against
the sides of the beaker to insure thorough mixing of the reagent with the emul-
sion. This operation shall be performed after bringing the weighed sample of
emulsion and the reagent to the standard temperature of 77° F.±l.O° F.
(25° C.±0.5° C.). .
One of the wire sieves shall be fitted over a beaker or other suitable vessel
and the mixture of emulsion and reagent poured through the sieve. The
beaker, containing the sample and glass rod, shall be rinsed with distilled water.
All lumps shall be kneaded and broken up and the washing of the beaker, ~-od
and sieve shall be continued until there is no longer any appreciable color im-
parted to the wash water. Mter washing as directed, the beaker, rod and sieve
used in each individual t.est sha]} be placed in a drying oven and dried at 325° F.
(163° C.) to constant weight.
The total weight thus obtained less the total tare weight of the beaker, rod
and sieve shall be the weight of the residue by the demulsibility test. The
ratio of the average weight in grams fro:gJ. three tests of each individual sample
of emulsified asphalt, A, to the weight in grams of residue per 100 g. of emulsion,
B, obtained in the test for residue by distillation described in Section 6 -multi-
plied by 100, shall be recorded as the percentage of demulsibility of the sample
tested:
Percentage Demulsibility = ~ X 100
. Effects of Freezing
Approximately 400 g. of the emulsion shall be pla'ced in a clean metal con-
tainer, such as a I-pt. press-top tin. The emulsion in the closed container
shall be exposed to a temperature of 0° F. (-17.7° C.) for twelve consecutive
hours. At the expiration of the freezing period, the emulsion shall be permitted
to thaw by exposure of the container to the temperature of the laboratory.
Mter the first operation of freezing and thawing, the procedure shall be
repeated twice, so that the emulsion will have been subje~ted to. tPree cycles of
freezing and thawing.
After the third cycle, the emulsion may be homogeneous or may have sepa-
rated into distinct layers whi<;h cannot be ren,dered homogeneous by stirring
at laboratory temperature.
The result of this test shall be reported as ,either "Homogeneous" or
"Broken." J'
(B) SEPARATION OF THE DISPERSION INTO ITS
COMPONENT PARTS
I
The following products are likely to be pres_w.t, viz:: water, ammonia,
various chemicals, bituminous matter, animal ami-vegetable oils or fats, other
forms of non-bituminous organic matter and mineral matter.
EXAMINATION OF BITUMINOUS SUBSTANCES 1593
~~
~ _l::,,~
- ~ ~"'~ j!J-
1"
2 I.D.
it-b==- -----------1~J
'I~t_-------:
:t.:..-
-----------
11_ _ _,....
through the small hole in the cover, using a cork stop'per, so that the end of the
bulb is x: in. from the bottom of the still. The ring burner shall be placed
around the still and the heat applied by this means to the top of the still.
Just enough heat from a bunsen burner shall also be\applied to the connecting
tube to prevent condensation of water in this tube. .
After practically all the condensate has been removed from the still and
the temperature of the residue has reached 250 0 F. (121 0 C.), the position of
the heat from the ring burner shall be lowered to midway of the still and held
there until the thermometer reaches 349 F. (176° C.). The burner shall then
0
x:
be rapidly lowered to within in. of the 'bottom of the still and the temperature
increased to and maintained at 5000 F. (260 0 C.) for 15 min. This latter period
of heating is l).ecessary to insure a smooth homogeneous residue in the still.
At the expiration of the heating period at the maximum temperature, the still '
and accessories sha,V'again be weighed and the percentag~ residue ?alculate4, ~nd
reported. The cover shall then be removed from the stIll and sUItable portIOns
of the residue shall be poured immediately into suitable molds and containers
for making the required tests. The'residue in the molds and containers shall
be permitted to cool, uncovered, to laboratory room te'mperature and thereafter
tested for specific gravity, penetration, ductility, fusing-point, solubility in
carbon disulfide and ash.
It should be noted that the distillation residue will carry any non-volatile
dispersing agents present in the emulsiqn.
NOTE.-When it appears impossible to distill an emulsified asphalt in .the still
described above due to excessive foaming of the emulsion, then the modified still shown
in Fig. 222 should be substituted for the still shown in Fig. 223 and the following pro-
cedure followed: Place the 6-in. burner around the larger diameter of the still near its top.
This serves as a support. Place the 4-in. burner immediately beneath the flare and the
2-in. burner not less than 2 in. below the bottom. Distillation is started with only the
2-in. burner lighted. Practically all of the distillate should be over in about 45 min.
When the distillation apparently stops, light the two larger ring burners and adjust to a
low flame. Distillation resames and when it stops again increase the heat by adiusting
the flame of the 2-in. burner. When the temperature cim b~ read upon the thermometer,
increase the rate of heating by raising the flame on botH the 2 and 4-in. burners and
bring the temperature to 500° F. (260° C.). If any evidence is noted of the emulsion
beginning to foam over in the delivery tube, remove the 2-in. burner quickly and raise
a pan of water so as to immerse the still bottom to a depth of about 2 in, for a moment
which will check the foaming. Upon resumption of heating watch delivery tube carefully
and repeat treatment if necessary.
When the residue has reached and remained at 500° F. (260 0 C.) fQr.15 min., proceed
as described for the regular still. While the distillation should be completed in not less •
than 1 hr. nor more than 174: hr. from the first application of heat to the still, the maxi-
mum stated is not mandatory as a longer time may be required in some cases to avoid
foaming of the emulsion into the condenser. If the, residue in the still prior to pouring
the ductility and penetration specimens appears granular or heterogeneous in any way,
stir with a spatula until the material runs from the spatula in strings' instead of drops,
and then pour. Should the residue then appear granuiar and heterogeneous, the test
shall be rejected and the distillation repeated as previously described with the exception
that the still temperature at 500° F. (260° C.) shall be maintained for a longer time than
15 min. until the desired condition of the residue is obtained.
(a) The apparatus shall consist of a metal still or glass flask, heated by
suitable means and provided with a reflux condenser discharging into a trap
connected to the still or flask. The trap serves to collect and measure the
condensed water and to return the solvent to the still.
The type of distilling apparatus used' is not an essential -J, I-<-H Inside Oi.m.
feature of this method.
(b) Metal Still.-The metal still shall be a vertical
cylindrical vessel, preferably of copper, having a faced
flange at the top to which the head is tightly attached
by means of a clamp. The head shall be of metal, pref-
erably of brass or copper, and be provided with a tubula-
tion 1 in. in inside diameter. . '
D
!u
(c) Glass Still.-The glass flask shall be of the short
neck, round-bottom type, made of well-annealed' glass,
having an approximate capacity of 500 ml.
(d) Heat Source.-The burner used with the metal
still shall be a ring gas burner 100 mm. {4 in.) in inside
diameter. With the glass flask, an ordinary gas burner
or electric heater may be used as the source of heat.
(e) Condenser.-The condenser shall be of the water- E 0
cooled, reflux, glass;tube type, having a condenser jacket ....L..
C Inside__"j ~
not less than 400 mm. (15U in.) in length with an inner Oi.m,
tube 9.5 to 12.7 mm. (Ys to Y:2 in.) in outside diameter. A= 45 to 55 mm.
The end of the condenser to be inserted in the trap shall B = 14 to 16 mm.
be ground off at an angle of 300 from the vertical axis of
c= 12 to 16 mm.
0=235 to 255 mm.
the condenser. . E = 23 to 38 mm.
F=186 to 194 mm.
(f) Trap.-The trap shall be made of well-annealed H= 78 to 79 mm.
glass constructed in 'accordance with Fig. 224, and shall FIG. 224.
be graduated from 0 to 25 ml. in O.l-ml. divisions. The
tolerance of the graduations between 0 and 2 ml. shall be ±0.5 ml. and b~
tween 2 and 25 ml. shall be ±O.l ml. The outside diameters should be pref-
erably 2.5 to 3.5 mm. Uz to VB in.)' greater than the inside diameters specified.
The solvent used when testing bituminous emulsions shall be a coal-tar
naphtha or a light oil and shall conform to the following distillation requirements:
98 per cent shall distil between 248 F. (120 C.) and 482 0 F. (250 C.)
0 0
The connections between the still or flask, trap and condenser shall be made
by means of tight-fitting corks as shown in Fig. 21Q (a) and (b). The end of
the condenser inserted in the trap shaH be adjllsted to that position which will
allow the end to be submerged to a depth of not :qJ.pre than 1 mm. (0.04 in.)
below the surface of the liquid in the trap after distillation conditions have been
established. When the metal still is used, a heavy paper gasket moistened
with the solvent shall be inserted between the lid and flange before attaching
the clamp. A loose cotton plug shall be inserted in the top of the condenser
tube to prevent condensation of atmospheric nlOisture in the condenser tube.
Heat shall then be applied and so regulated th"at the conden~ed distillate
falls from the end of the condenser at the rate of from 2 to 5 drops per sec.
The ring burner used with the metal still shall be placed about 3 in. above
the bottom of the still at the beginning of the distillation and gradually lowered,
as the distillation proceeds. The distillation shall be continued at'the specified
rate :Intil no water is visible on any part of the apparatus except at the bottom
of the trap. This operation usually requires l(lss than 1 hr. A 'persistent ring
of condensed water in the condenser tube shall be removed by increasing the
rate of distillation for a few minutes.
The volume of condensed water measured in the trap at room temperature
multiplied by 100 and divided by the weight of sample used, shall be the
percentage of water and shall be reported as ", " . per cent water by weight."
The accuracy to be expected with this lilethod is that duplicate deter-
minations of water should not differ from each other by more than one di:vision
on the trap.
The sum of the percentages of Distillatioll Residue and Water, deducted
from 100 represents the percentage of Volatile Oils.
CEMENT
SPECIFICATIONS AND CHEMICAL ANALYSIS 1
The physical tests ordinarily applied to Portland cement are as follows:
Setting time.
S_oundness.
Tensile strength.2
Che~ical analysis is als~ made, particular attention being paid to the deter-
mination of magnesia, sulfuric anhydride, insoluble residue, and loss on ignition.
As a general rule, however, it may be said that so far as the consumer is con-
cerned, more attention is paid to the physical tests than to chemical analysis.
Standard specifications covering the requirements for cement, both chemical
and physical, have been adopted by the American Society for Testing Materials,
and by the U. S. Government." The former are generally recognized by cement
users as the standard requirements, while the latter are used by the various
branches of the federal government. The two sets of specification may be very
different.
The specifications and methods of making the chemical tests follow.
SPECIFIOATIONS 4
o .
Definition.-Portland cement is the product obtained by finely pulverizing
clinker produced by calcining to incipient fusion an intimate and properly
propprtioned mixture of argillaceous and calcareous materials, with no addi-
tions subsequent to calci~ation excepting water and calcined or uncalcined
gypsum.
Chemical Limits.-The following maximum limits shall not be exceeded by
amounts greater than the respective tolerances indicated as allowable in the
chemical determinations: '
Limits Tolerance
Loss on ignition, %, _..... _. . . . . . . . . . . . . . . . . .. 4.00 0.25
Insoluble residue, % ..... , , , , . . . . . . . . . . . . . . . .. 0.85 0.15
Sulfuric anhydride (S03), % ............ , .. , . " 2.00 0.10
Magnesia (MgO), %........................ " 5.00 0.40
1 By W. C. Hanna, Chief Chemist and Chemical Engineer, California Portland
Cement Co., Colton, California.
2 See A. S. T. M. Method C 77-37 for the Physical Tests. 1937 Supplement to
Book of A. S. T. M. Standards, pp. 61-73.
3 The federal specifications are issued by the Director of Procureme_nt, Washington,
D. C., and sold by the Superintendent of Documents, Government Printing Office,
Washington, D. C. The specifications now in effect covering,cement are as follows:
88-0-158. September 30, 1936. Cements, Hydraulic; General Specifications (Methods
for Sampling, Inspection, and Testing).
S8--C-191a. September 30, 1936. Cement; Portland.
8S-O-201. September 30, 1936. Cement; Portland, High-Early-Strength.
S8--C-206. September 30, 1\}36. Cement; Portland, ModeratecHeat-of-Hardening.
S8-0-181a. May 22,1936. Cement; Masonry.
4 Standard Specifications for Portland Cement. A. S. T. M. Designation: C 9-37.
1597
1598 ' CEMENT
The residue on a standard No. 200 sieve shall not exc~ed 22% by' weight.
A pat o,f neat cement shall remaJn firm and- hard, and show no signs of
distortion, cracking, checking, or disintegration in the steam test for soundness.
The cement sh:}ll not develop initial set in less than ~5 minutes when the
Vicat needle is used or in less than 60 minutes when the Gillmore needle is used.
Final set shall be attained within 10 hours. '
The average tensile strength in pounds per square inch bf not less than three
standard mortar briquettes composed of one part cement and three parts
standard sand, by weight, shall be equal to or higher than the following:
A. S. T. M. DESIGNATION: C 77-37 6
SAMPLING
Size and Number of Samples.-l. Samples for purpose of tests shall weigh
at least 4 lb. •
2. Test samples shall be either individual or composite samples, as may be
ordered, and one test sample shall represent not more than 300 bbl., unless
otherwise specified by the purchaser.
Methods of Sampling.-3. When cement is sampled in cars, trucks, boats,
warehouses, etc;, one sample shall be taken from one sack in each 40 sacks (one
sample per 10 bbl.), and combined to form one test sample. In the case of
truck samples where the cement is being trucked from one mill, it is permissible
to combine the samples from several trucks to form a test sample representing
not more than 300 bbl. When sampling bulk shipments, representative sam-
ples shall be taken from well distributed points.
Sampling Bulk Storage.-4. The cement may be sampled at bulk storage
points by any of the following methods: .
(a) From the Conveyor Delivering to Storage.-At least 4 lb. of cement
shall be taken from at least each 300 bbl. passing over the conveyor. This
may be secured by taking the entire test sample at a single operation, known
as the "Grab Method," or by combining several portions taken at regular
intervals, known as the" Composite Method." When obtaining a composite
sample this shall be secured by combining approximately equal weights taken
at regular intervals, each poftion representing approximately 10 bbl. Auto-
matic samplers may be used in obtaining samples.
5 Under the standardization procedure of the Society these methods are under the
jurisdiction of the A. S. T. M. Committee C-1 on Cement. -
A Manual of Cement Testing to supplement these methods has been prepared and
published as information, not a part of these methods, see Proceedings, Am. Soc. Testing
Mats., Vol. 32, Part I, p. 316 (1932).
6 These methods are issued under the fixed designation C 77; the final number
indicates the year of original adoption as standard or, in the case of revision, the year
of last revision. Adopted, 1904; Issued in Amended Form Under Present Designation,
1930; Revised, 1932, 1937. Reproduced with the permission of the American Society
_ for Testing Materials.
1600 CEMENT
(b) From Storage by Means of Proper Sampling Tubes.-Tubes inserted
vertically may,be used for sampling cement to a maximlJ.ffi depth of 10 ft.
Tubes inserted horizontally may be used where the construction of the storage
permits. Samples shall be taken at points well distributed lover the storage.
(c) From Storage at Points of Discharge.-Sufficient cement shall be drawn
from the discharge openings to obtain samples representative of the cement,
as determined by the appearance at the discharge openings df indicators placed
on the surface of the cement directly above these openings before the drawing
of the cement is started. One 4-lb. sample shall be taken for at least each
300 bbl. .
Preparation of Sample.'-5. The sampling shall be done by or under the
direction of a responsible representative of the purchaser.
6. Samples shall be shipped and stored in moisture-proof, air-tight containers.
Before testing, samples shall be thoroughly mixed and then passed through a
No. 20 sieve in order to' break up lumps and remove foreign materials.
CHEMICAL ANALYSIS 7
and heat at a full red heat (900 0 to 1000 0 C.) for 15 min.; check the loss in
weight by a second heating for 5 min. Take care to wipe off particles of' as-
bestos that may adhere to the crucible when withdrawn from the hole in the
board. Greater neatness and shortening of the time of heating may be secured
by making a hole to fit the crucible in a circular disk of sheet platinum and
placing this disk over a somewhat larger hole in an asbestos board.
(b) Method 2.-Place the crucible in a muffle furnace at any temperature
between 900 0 and 1000 0 C. for 15 min. and check the loss in weight by a second
heating for 5 min.
Insoluble Residue.-9. Add 25 ml. of cold water and 5 ml. of HCl (sp.gr.
1.19) to a I-g. sample of cement. Heat the solution and grind the material with
the flattened end of a glass rod until it is evident that decomposition of the
cement is complete. Dilute the solution to 50 mI. and digest on a steam bath
for 15 min. Filter the residue, wash six times with hot water, and digest the
filter paper and contents in about 30 mI. of a sodium carbonate solution (5%),
the liquid being held at a temperature just short of boiling for 15 min. Filter
the remaining residue, wash twice with hot water, then with a few drops of hot
diluted HCI (1 : 9), and finally eight to ten times with hot water, then ignite
at a red heat and weigh as the insoluble residue. .
Sulfuric Anhydride.-lO. Add 25.mI. of cold water and 5 mI. of HCI (sp.gr.
1.19) to a I-g. sample of cement. Heat the solution and grind the material with
the flattened end of a glass rod until it is evident that decomposition of the
cement is complete. Dilute the solution to 50 mI. and digest on a steam bath
for 15 min., filter, and wash the residue thoroughly with hot water. Dilute the
solution to 250 mI., heat to boiling, add slowly, drop by drop from a pipette, 10
mI. of a hot barium chloride solution (10%), and continue the boiling until the
precipitate is well formed. Then digest. the solution on the steam bath for at
least 3 hr.; preferably overnight. Filter the precipitate, wash and place the
paper and contents in a weighed platinum crucible and slowly char and consume
the paper without flaming. Then ignite, and weigh the barium sulfate. The
weight obtained multiplied by 34.3 gives the percentage of sulfuric anhydride.
The acid filtrate obtained in the determination of the insoluble residue may be
used for the estimation of sulfuric anhydride instead of using a separate sample.
Silica.-l1. (a) Transfer a 0.5-g. sample of the finely powdered cement to an
evaporating dish, preferably of platinum for the sake of celerity in evaporation,
moisten with 10 mI. of cold water to prevent lumping add 5 to 10 ml. of HCI
(sp.gr. 1.19) and digest with the aid of gentle heat and agitation until solution is
complete. Solution may be aided by light pressure with >the flattened end of a
glass rod. Evaporate the ·solution to dryness on a steam bath. Treat the
residue without further heating at first with 5 to 10 mI. of HCl (sp.gr. 1.19) then
diluting to half strength or less, or pour at once upon the residue a larger volume
of acid of half strength. Then cover the dish and digest for 10 min. on the
bath or a hot plate. Dilute the solution with an equal volume of hot water,
immediately filter, and wash the separated ·silica thoroughly with hot water
and reserve. Again evaporate the filtrate to dryness, baking the residue in an
oven 1 hr. at a temperature of 105 0 to 1100 C. Then take up the residue with
10 to 15 mI. of diluted HCI (1 : 1) and heat on the bath or hot plate. Dilute
the solution with an equal volume of hot water and catch and wash the small
amount of silica it contains on another filter paper. Reserve the filtrate and
1602 CEMENT
washings. Transfer this paper and the r~sel'yed paper containing the residues
to a weighed platinum crucible, dry, ignite first over aibunsen burner until the
carbon of the filter paper is completely consumed without flaming and finally
over a burner or in an electric oven at 11000 to 12000 C, ~mtil the weight remains
constant. I •
(b) Treat the silica thus obtained, which will contain small amounts of
impurities, in the crucible with a few drops of water, about 10 m!. of HF and
one drop of H 2S0 4 and evaporate to dryness over a low flame. Finally, heat the
,small residue at 1050 0 to 11000 C. for a minute or two, cool, and weigh. The
difference between this weight and the weight previously obtained represents
the amount of silica. Washing of the silica precipitates can be made more
effective by using a hot solution of diluted HCI (1 : 99), and then completing
the washing with hot water. Consider the weighed residue, remaining after
volatilization of the silica, as combined aluminum and ferri'c oxides and add it
to the result obtained in the determination of these oxides (Section 12).
Ferric Oxide and Alumina.-12. Treat the reserved filtrate (Note 1), which
should have a volume of about 250 mL, with HCI, if necessary, to insure a total
of 10 to 15 mL of the acid, add a few drops 'of methyl red indicator and then
heat to boiling. Then treat with NH 40H (Note 2), drop by drop, until one
drop changes the color of the solution to a ,distinct yellow. Bring to boiling
the solution containing the precipitated iron and aluminum hydroxides, boil
for 1 or 2 min., and then filter. Wash the precipitate onc(l by decantation and
slightly on the filter with a hot ammoni~m nitrate solution (2%). Sef\ aside
the filtrate and transfer the precipitate and filter paper to the same beaker in
which the first precipitation was effected. Then dissolve the iron and aluminum
hydroxides in hot diluted HCI (1 : 3), make up the solution to about 100 mI.,
and reprecipitate the hydroxides as i1,). the first case. Then filter the solution
and wash the precipitate with two lO-mL portions of a hot ammonium nitrate
solution (2%). Combine the filtrate and washings with that set aside and
place the precipitate in a weighed platinum crucible, slowly heat by a bunsen
or Meker burner until the papers are charred, and finally ignite to constant
weight at 1050 0 to 1100 0 C. with, care to pr~vent /reduction, and weigh as
+
Al 2 0 a FezOa. I
NOTES.-l. If a platinum evaporating disn has been used for the dehydration of
silica, iron may have been partially reduced. At thil1 stage, add about 3 ml. of saturated
aqueous solution of bromine to the filtrate and boil the filtrate to eliminate the excess
bromine before adding the methyl red indicator. '-
2. The NH 40H used to precipitate the hydroxides must be free 01 contamination
with carbon dioxide. • _
Ferric Oxide.-This method for determining ferric oxide is not affected by
any titania or vanadium oxide that may be in the cement. It is therefore
understood that these are counted as Al 20 a in this method. If the precipita-
tions of aluminum and iron are made as directed, little if any manganese will'
be retained in the precipitate. '
Solutions Required.-13. The following solutions will be required:
(a) Stannous Chloride.-Dissolve 2 g. of the crystallized salt in four times
its weight of a mixture of three parts of water and one ofiHCI (sp.gr. 1.19).
Add scraps of iron-free granulated tin and boil u_ntil~he solution is clear. Keep
the solution in a closed dropping bottle containing-metallic tin.
CEMENT 1603
(b) Phosphoric Acid Mixture.-Add slowly and with stirring 150 ml. of
orthophosphoric acid (sp.gr. 1.71), and 150 ml. of H 2 S0 4 (sp.gr. 1.84), to 500 ml.
of water and then dilute to 1000 ml. with water and mix.
(c) Diphenylamine Indicator.-Dissolve 1 g. of diphenylamine in 100 ml.
of H 2S0 4 (sp.gr. 1.84).
(d) Standard Potassium Bichromate Solution.-Dissolve 2.457 g. of
K 2 Cr 207 in distilled water and dilute to 1 liter. Standardize against standard
iron wire of known iron content in the same manner as directed for the ferric
oxide determination in cement (Section 14), using a weight of iron wire which'
will yield a titration approximately equal to that required by the cement sample
in question. .
Method.-14. Add 20 m!. of cold water and 10 m!. of HCI (sp.gr. 1.19) to a
I-g. sample of cement. Heat the solution and grind the material with the
flattened end of a glass rod until it is evident that decomposition of the cement
is complete. Dilute the solution to 50 ml. and heat to boiling. Then treat
the solution with 5 m!. of HC:! (sp.gr. 1.19) and then with a solution of stannous
chloride, added drop by drop while stirring until the solution is decolorize<;l,
then add one drop more. Cool the solution to room temperature, rinse the
inside of the vessel with water and add all at once 10 m!. of a cool saturated
solution of mercuric chloride. Then stir the solution vigorously for 1 inin.,
treat with 15 ml. of phosphoric acid mixture and three drops of indicator and
dilute to 150 m!. with cold water. Then titrate with a standard potassium
dichromate solution having a titer no greater than 0.004 g. of Fe 20a. The
end point is taken as the point when a drop causes an intense deep-blue colora-
tion that remains unchanged on the further addition of the standard K 2 Cr 2 07
solution.
Calcium Oxide.-15. Add a few drops of NH 40H to the combined filtrates
obtaided in the alumina and ferric oxide precipitation (Section 12) and bring
the solution to boiling. Add to the boiling solution 25 ml. of a boiling ammo-
nium oxalate solution (5%), the boiling being continued until the precipitated
calcium oxalate assumes a well-defined granular form. Then allow the solution
to stand for about 20 min. or until the precipitate has settled, filter, and moder-
ately wash with an ammonium oxalate solution (0.1 %). Reserve the filtrate
and washings. Place the filter paper containing the precipitate, wet, in a
platinum crucible and burn off the paper over the small flame of a bunsen
burner. Then ignite the precipitate until the oxalate is converted to oxide
(Note), redissolve in HCI and .make up the solution to 100 ml. with. water.
Add NH 40H in slight excess and bring the solution to boiling. If a small
amount of alumina separates at this point, filter it out, weigh, and add the
amount to that found in the original alumina determination (Section 12).
Then reprecipitate the lime by ammonium oxalate, allow to stand until settled,
filter, and wash, taking care to avoid an excess of the washing solution. Com-
bine the filtrate with that already reserved. Weigh the precipitate as calcium
oxide after ignition and blasting to constant weight in a weighed, covered
platinum crucible.
NOTE.-If the ignited precipitate is darkly colored, it may contain an appreciable
amount of manganese. The contamination of the second precipitate of calcium oxalate
by manganese should be as low as possible. The following procedure may be used to
reduce the contamination: add 5 mI. of cold water cautiously to the precipitate and then
1604 CEMENT
cautiously add 10 mI. of diluted HOI (1 : 1). Heat the mixture. If the resulting solution
is colored, digest it on the hot plate gently for 10 min. or longer until the color disappears
or until it is apparent the color will not become lighter on further heating. After a
little practice it is possible to dissolve hard-burnt lime in this way without explosive
spattering. I
If the second precipitate of calcium oxalate is so fine as tol cause difficulty in filtering
or if the precipitate causes bumping, the volume of solution may be increased to 150 or
200 ml., instead of 100 mI. prior to precipitation.
100(a-2b). .
MgO, per cent = 2 X MgO eqUIvalent of 1 m!. standard alkalI,
PHOSPHORUS
of HNO a (sp.gr. 1.42). Boil the mixture until the solution is as complete as
possible. If the cement contains an appreciable ampunt of manganese as
shown by the presence of a red or brown residue, aq<;i a few milliliters of a
I
potassium nitrite solution (5%) to aid in the soltltion. Boil the solution until
all nitrous fumes are completely expelled, taking care w;>t to allow the volume
of the solution to become so small as to cause the precipitation of gelatinous
silica. Filter the solution into a 500-ml. Erlenmeyer flask and wash the re&idue.
The volume of the solution should now be about 150 ml.
NOTE.-The amount of sample and reagents used depend on the content of phos-
phorus in the cement. The minimum rcquirements are sufficient if the cement contains
0.5% P 20 5 or more. The maximum amounts are required if the content of P 20 6 is 0.1 %
or less.
(b) Add 10 g. of ammonium nitrate to the solution and heat until the salt
is dissol ved. Heat the solution to 85 0 C. and add 50 to 100 ml. of the solution
of ammonium molybdate. Stopper the flask and shake it vigorously for 5 min.,
then set it aside for 30 min. Filter the precipitate, rinsing the flask five times
and washing the filter paper eight times with a cold solution of potassium
1\\t'.i:'O,t~ \1%>.
(c) Put the original flask under the funnel containing the precipitate.Pour
25 ml. of diluted NH 40H (1 : 4) over the precipitate and' around the upper
edges of the filter paper un til the preci pita te is entirely dissolved . Wash the
paper three times with cold water, twice with diluted HCI (1 : 1), and finally
three times with cold water. Acidify the solutio!). with HCI and add to it 4 or 5
drops of a saturated solution of citric acid to hold traces of iron, if any, in
solution. Cool the solution to room temperature and transfer it to a 400-ml.
beaker. The volume of the solution should now be about 150 ml. '
(d) To the solution add 25 ml. of magnesia mixture, and then while stirripg
constantly add drop by drop 50 ml. of NH 40H (ep.gr. 0.90) at such a rate that
5 min. are required for the addition. Allow tlle mixture to stand 4 hr. or
longer. Filter the precipitate and wash it a fe'" time~ with diluted NH 4 0H
(1 : 20). Dissolve the precipitate into the original beaker by pouring 25 ml.
of diluted HCI (1 : 1) on the precipitate. Wash the' filter paper thoroughly
with diluted HCI (1 : 20). With the washings jl.nd water, dilute the solution
to 50 to 100 ml. and add to it 1 to 2 ml. of magl].esia mixture. While stirring
constantly, add NH 40H slowly to the solution tl,S before. Allow the mixture
to stand 2 hr. or longer. Filter the precipitate, scrub and.riJIse the beaker, and
wash the precipitate ten 'times with diluted :NH 4 0H (1 :·20}.~ ._Ignite the
precipitate in a weighed crucible at 1100 0 to 1200 0 C. to constant weight. Mul-
tiply the weight of the magnesium pyrophosphate (Mg 2P 20 7) by 0.6379 to get
the weight of the phosphorus in the sample as p.__O ••
(e) Blank Correction.-Make a blank determination, using the same pro-
cedure as outlined above, and using approximately the same. amo~nt of re-
agents. Correct the weight of magnesium pyrophosphate (Mg 2P 20 7) ac-
cordingly.
(j) If a determination of phosphorus is required, the weight of the ferric
oxide and alumina (AI 20 a+Fe 20a) as obtained by the method described in
Section 12 of the Standard Methods of Sampling and ~esting Portland Cement
(A. S. T. M. Designation: C 77) of the American ~(Jt!iety for Testing Materials
shall be corrected for the amount of the P 20. fOllnd.
CEMENT 1611
MANGANESE
diluted HNO a (1 : 33). Titrate the, filtrate immediately with the. standard
solution of As 20 a, The endpoint is reached when a yllllow color is obtained
free of brown or, pur.ple tints and it does not ·change upon
I
further addition of
As 20 3. Calculate the manganese in the cement as Mn2,03. •
(c) Blank Correction.-Make a blank determination, using the same pro-
cedure as outlined above and using approximately the same amounts of reagents.
Correct the titration results accordingly.
(d) If a determination of manganese is required in addition to a determina-
tion of magnesium by Section 16 of Standard Method C 77, manganese shall
be eliminated from the combined filtrates obtained in the alumina and ferric
oxide determination before calcium oxide is p:r;ecipitated (Section 15 of Standard
Method C 77) in the following manner:
Acidify the combined filtrates with HCI and evaporate them to a volume of
about 100 ml. Add 40 ml. of saturated bromine water to the hot solution and
immediately add NH 40H until the solution is distinctly alkaline. Boil the
solution for 5 min. or more, making certain that the solution is at all times
distinctly alkaline. Allow the precipitate to settle, filter and wash with hot
water. Discard the precipitate. Acidify the filtrate' with HCl and boil until
all the bromine is expelled. Make the solution alkaline with NH 40H and
proceed in accorda~ce with Section 15 of Standard Method C 77.
this digestion should be carried on. over nigh,t. Then boil the solution for.,a
minute, allow to settle and pour the clear liquid through a 12.5 cm. paper into
a 600-ml. beaker. If there are any lumps in the casserole, grind them, with an
agate pestle, add about 50 ml. water,. and boil for a few minutes. Allow to
settle, decant the clear liquid through the same paper, add 30 ml. water and
bring to boiling temperature, allow to settle and decant the clear liquid through
the same paper. Repeat this washing by decantation three times and finally'
wash the residue into the filter. Rinse the casserole thoroughly and wash the
residue in filter once with hot water.
Evaporate the filtrate to about 50 mI., add a few grams of solid ammonium
carbonate and boil for a minute. Filter through an 11 cm. paper into a 250~ml.
beaker, washing six times with cold water containing a little concentrated
ammonia and about 10 gms. ammonium carbonate per liter. To the filtrate
add 10 m!. saturated ammonium oxalate solution and 5 ml. 1-1 ammonia
water and evaporate down to about 10 to 20 ml. Filter through a 9 cm. paper
• into a platinum dish, washing five times with hot water. Add 2 or 3 ml.'
concentrated HCl to the solution in the dish, and evaporated to dryness.
Ignite the residue in the dish cautiously over a free flame until the fumes of
ammonium salts are driven off and finally heat at a temperature short of dull
redness. Cool the 'dish, take up the residue in about 50 ml. of hot water, add
1.5 ml. 10% .BaCh solution. Digest on water bath for about 30 minutes,
filter, and wash 5 times with hot water. Evaporate the filtrate to about pO ml.
Add 2 or 3 grams solid ammonium carbonate, boil for a minute, filter, and wash
with ammonia~carbonate wash water .
.To the filtrate add half a gram solid ammonium carbona\,c, 2 ml. saturated
ammonium oxalate solution and a few ml. or 1-1 ammonia water. Evaporate
the solution down to about 10 ml. and filter through a 9 cm. paperinto a clean
platinum dish, washing 5 times with hot water. Add 1 ml. concentrated HCl
and evaporate to dryness, Heat at a temperature short of dull redness until
all ammonium salts are driven off, cool in a dessicator, and weigh. After
weighing, dissolve the salts in about 50 ml. hot water, and filter through a 9 cm.
paper into a 150 ml. beaker. Wash the paper 5 times with hot water, and then
place it in the platinum dish. Ignite until the paper is consumed, cool, and
weigh. The difference between these two weights gives the weight of the
alkalies as sodium and potassium chlorides.
To the solution in the 150-ml. beaker add a few ml. of 1-1 HCI and evap-
orate to 10 ml. Then add 6 ml. of platinum solution containing 5 g. HiPtCI 6 ·-
6H2 0 per 100 ml. for every 0.10 gms. of mixed chlorides weighed, and 3 ml. in
excess. Evaporate on a water bath to a thick syrup. Carry this evaporation
, as far as possible without allowing the solution to go to dryness. The syrupy
solution should solidify on cooling. Then add about 100 ml. 01 85% ethyl
alcohol, stirring the solution thoroughly. The potash in the sample will now
be precipitated as yellow potassium platinic chloride, while the sodium is in
solution.
Allow the precipitate to settle and filter the clear liquid through a weighed
Gooch crucible. Wash the precipitate in the beaker 5 times by decantation
with 80% ethyl alcohol. Transfer the precipitate to the crucible with the
alcohol and scrub the beaker. Wash the crucible and precipitate 6 times with
1614 CEMENT
the alcohol. Dry the crucible for several hQurs at 1050 C., cool, and weigh
as potassium platinic chloride (K 2 PtCI s). j '
Calculate the K 20 and N a 20 in the sample' as follows:
Weight K2PtCI6XO.193S""weight\K20,
Weight K. 2 PtCI 6 XO.3067 = weight KCI,
Weight mixed chlorides-weight KCI=~eight NaCI,
Weight NaCIXO.5303 = weight Na 2 0.
Make a blank determination, using the same procedure as outlined above
and using approximately the same amounts of reagents and correct the deter-
mination of alkalies accordingly. .
the boiling for 5 to 10 minutes and then allow the mixture to stand for 15 to 20
minutes. Filter, receiving the filtrate in a 600-mI. beaker. Wash the pre-
cipitated calcium oxalate eight or ten times with hot water. Reserve the
filtrate and washings for the determination of magnesia. Wash the precipitate
back into the original beaker. Add about 250 mI. of hot water and 20 mI. of
sulfuric acid (1 : 1). Heat the solution just short of boiling and titrate it
with standard potassium permanganate solution, adding the filter paper when
the titration is nearly completed. Calculate the per cent of CaO in the sample.
Potassium Permanganate Solution.-Make up the solution on the basis of
5.634 grams to one liter and standardize it with National Bureau of Standards
Standard Sample of sodium Oxalate or with high grade sodium oxalate specially
made by chemical wanufacturers for this purpose. The latter is more eco-
nomical but should pe checked with a sample from the National Bureau of
Standards. Dissolve 0.7500 gram of sodium oxalate in 250 mI. of hot water
and add 20 mI. of sulfuric acid (1 : 1). Heat the solution just short of boiling
and titrate it with the permanganate solution. If the reading is 62.8 mI., the
standard solution is exactly of the required strength and a reading in analysis
gives directly the pefcentage of calcium oxide in a half gram sample. If the
reading in standardi2lation is not 62.8, it is more satisfactory to use a table of
corrections than to attempt to adjust the strength of the solution because the
solution deteriorates rapidly and forms a sediment of manganese dioxide.
The percentage of calcium oxide is equal to 62.8a+8 where a is the reading in
analysis and 8 is the reading in standardization.
Magnesia.-Acidify the filtrate with hydrochloric acid and boil it down to
about 250 mI. Whep. the solutio~ is cooled to room temperature, add 30 mI.
5% solution of sodium ammonium phosphate and 30 mI. of concentrated
ammonia water. Stir the mixture vigorously until a crystalline precipitate is
formed. Stand the mixture several hours, preferably overnight. Filter the
precipitate and wash it with 5% ammonia water. Ignite it in a tared crucible,
at first slowly until the paper is burned off and then at 1000-1100° C. for 20
to. 45 minutes. Weigh as Mg 2P 207 and calculate the percentage of MgO in
the sample.
Other Components.-Use the A. S. T. M. Standard Method C 77 and
Tentative Method C 114 which may be modified in order to gain speed. 14
NOTE.-This method in 'principle is similar to methods generally used by cement
plant laboratories for routine wor~. The rapid methods ~ive fairly accurate. results
and serve their purpose well, espeClally when they are occasIOnally checked agamst the
official methods of the A. S. T. M. The use of perchloric acid for the dehydration of
silica "was first proposed by Willard and Cake, J. Am. Chern. Soc., 42, 2208 (1920), and
has been improved for tlse on cement by numerous analysts.
14 See also: "DeterrJlination of Sulfuric Anhydride in Portland Cement by Means of
the Wagner Turbidimeter," Robert B. Rudy, J. Res. National Bur. of Standards, 16,
No.6, June 1936. "Determination of Sulfur Occurring as Sulfide in Portland Cement,"
Harry A. Bright, J. Ree. National Bur. of Standards, 16, No.2, February 1937.
1616 CEMENT
_ total Ca as CaCO a •
F ac t or for C a CO a - 60+2h-s
,
A factor thus obtained is valid only if the raw mix in analysis is similar
to the standard sample of raw mix. If the composition of raw ~ix va.ies,
total magnesium and total calcium should be determined from time to time
and the factors be changed accordingly.
If the content of magnesia is consistently much lower than the limit allowed
by specifications, the determination of magnesium carbonate,may be omitted.
Stop the determination of calcium carbonate at the a.,Epearance of the first pink
coloration. x and h will be zero in the formulae given above for ca~cium car-
bonate. When magnesium carbonate is determined, its content decides the
OEMENT 1617
A B C D E F G H
--- --- - - - - - - --- ---
8i0 2 • • • • • • • • • • • • • • • • • • • 17 to 27 21.16' 22.61 20.08 20.54 26.66 24.66 28.91
Al 20 S • • • • • • • • • • • • • • • • • • 3 to 10 5.65 5.30 6.54 4.75 2.86 5.60 10.96
Fe20S .................. o to 7 2.23 5.23 2.40 6.05 1.76 0.38 4.68
CaO ................... 57 to 70 63.36 59.26 65.80 61.50 64.10 61.87 34.64
MgO .................. o to 5 4.20 3.55 1.40 3.89 1.27 0.90 14.82
80 s ................... 1 to 2.5 1.61 1.81 2.50 1.54 1.43 1.51 1.04
Loss on ignition .... : .... 0.5 to 4 0.97 1.21 1.12 0.70 1.60 3.84 4.50
Insoluble residue ........ o to 1 0.09 0.09 0.12 0.09 0.11
A B C D E F G
Si0 2 • • • • . • • . • . • . • . • . . 55.52 65.86 8.55 91.28 0.74 0.54 17.04 1.74 15.37
~l-I
AI 20 s ................. 14.98 9.05 3.08 2.74 0.25 0.11 6.90 0.90 9.13
Fe 20s ............ _... 5.48 3.69 78.67 0.40 0.09 0.07 2.13 0.28 0.25
CaO ................. . 7.30 6.35 4.63 2.10 55.65 32.60 37.53 49.84 25.50
MgO ................. 3.85 2.52 1.08 0.45 0.85 1.54 2.17 1.75 12.35
SOs .................. 0.78 43.29 1.12
Loss on ignition ....... 6.22- • 8.32 3.82 2.39 42.40 21.94 32:88 46.01 35.40
A. S. T. M. DESIGNATION: D 21-16 2
'~
~ 0
0
I:lIJ
>::
~
S
~
Ul
'-
0
w
""...c:
0
-+"
<l)
:g
"2<:(l
11...,
<:(l
Ul
I
to
C'I
C'I
ci
~
z'" ... .
",,,2
--"
~:~~
:!:Cc,,,-
III
..
iii
~
:;;
:c
..'"t:
;:!
i
=~~=
"."......
.G:= ....
a f
i:
COAL AND COKE 1623
Hand Preparation.-The progressive reduc~ion in the weight of the sample
to the qoantities indicated in Table I shall be done by the following methods,
which are illustrated in Fig. 225;
(a) Mixing and Reduction by DiscardiI1g Alternate Shovelfuls.-The
alternate-shovel method of reducing the gross sample shall be repeated until
the sample is reduced to approximately 250 lb., and care shall be observed
before each reduction in quantity that the s/tmple has been crushed to the
fineness prescribed in Table I. .
The crushed coal shall be shoveled into a ccnical pile (Fig. 225) by deposit-
ing each shovelful of coal on top of the precec:iing one, and then formed into
a long pile in the following manner: The saplpler shall take It shovelful of
coal from the conical pile and spread it out in a straight line (Fig. 3A or
8A) having a width equal to the width of the shovel and a length of 5 to 10 ft.
The next shovelful shall be spread directly ov~r the top of the first shovelfui,
but in the opposite direction, and so on back and forth, the pUe being occa-
sionally flattened, until all the coal has been formed into one long pile (Fig.
225~ 4 O[ 9\, The samvler shall thell discard half of this J!ile •. J!roceedin~ as
follows:
Beginning on one side of the pile, at either end, and shoveling from the
bottom of the pile, the sampler shall take one shovelful (shovelful No.1,
Fig. 3B or 8B) and set it aside; advancing aloJlg the side of the pile a distance
equal to the width of the shovel, he shall ta¥e a second shovelful (shovelful
No.2) and discard it; again advancing in the same direction one shovel width,
he shall take a third shovelful (shovelful No.3) and add it to the first. The
fourth (shovelful No:4) shall be taken in a like manner and discardej:l, the fifth
(shovelful No.5) retained, and so on, the sampler advancing always in the
same direction around the pile so that its size will be gradually reduced in a
uniform manner. When the pile is removed, about half of the original quan-
tity of coal should be contained in the new pile formed by the alternate shovel-
fuls which have been retained. (Fig. 5A or lOA show the retained halves,
and Fig. 5B or lOB the rejected halves.)
(b) Mixing and Reduction by Quartering.__,.-After the gross sample has been
reduced by the alternate-shovel method to approximately 250 lb., further
reduction in quantity shall be by the quartering method. Before each quar-
tering, the sample shall be crushed to the fineJless specified in Table I.
Quantities of 125 to 250 lb. shall be mixeo thoroughly by coning and re-
coning (Figs. 12 and 13); quantities less than 125 lb. shall be placed on a
suitable cloth, measuring about 6 by 8 ft., mixed by raising first one end of
the cloth and then the other (Fig. 18, 24 pr ~O), so as to roll the coal back
and forth, and after being mixed thoroughly shall be formed into a conical
pile by gathering together the four corners of the cloth (Fig. 19, 25 or 31).
The quartering of the conical pile shall be done as follows:
The cone shall be flattened, its apex being pressed down vertically with
a shovel, or board, so that after the pile has been quartered, each quarter will
contain the material originally in it. The fl::J.ttened mass, which shall be of
uniform thickness and diameter, shall then be m.arked into quarters (Fig.
14, 20, 26 or 32) by two lines that intersect at nght angles directly under
a point corresponding to the apex of the origir~al cone. The diagonally oppo-
site quarters (BB in Fig. 16, 22, 28 or 34) shall then be shoveled away and
1624 COAL AND COKE
discarded and the space that they occupied brushed Clean. The coal remain-
ing shall be crushed, mixed, coned, and quartered successively until the sample
is reduced to the desired quantity.
(c) The 30-lb. quantity (Fig. 29) shall be crushed\to 136 -in. size or to pass
a 4760-micron (No.4) sieve, mixed, coned, fiattenea. and quartered. The
laboratory samples shall include all of one of the quarters, or all of two oppo-
site quarters (Fig. 34), as may be required. The laboratory sample shall be
placed immediately in a suitable container and sealed in such a manner as to
preclude tampering.
Mechanical Preparation.-Only such mechanical means as will give equally
representative samples shall be used in substitution for the hand method of
preparation herein standardized.
.
MINIMUM WEIGHTS OF GROSS SAMPLES OF COKE
Run-of-oven, Blast-furnace, Foundry, Water-gas, and any coke containing a
range of size of pieces made from uncrushed or co~rsely £rushed coal, b
except coke breeze ................................................ , 500 lb."
Run-of-oven, Blast-furnace, Foundry, Water-gas, and any coke containing a
range of size of pieces made from crushed coal," except coke breeze ...... 250 lb."
Closely-sized coke made from uncrushed or coarsely crushed coal b free of coke
breeze ........................................................... 250 lb."
Closely-sized coke made from crushed coal c free of coke breeze ...... , ...... 125 lb."
Coke breeze (all passing a 72 ()r %;-in. square-hole sieve) ................... 125 lb.
a In case the pulverization of the coal is not known, take quantities designated for
coke made from uncrushed or coarsely crushed coal.
b More than 10% on a 74'-in. square-hole sieve.
c Not less than 90% passing through a U-in. square-hole sieve.
Quantity Represented.-A gross sample shall be taken for each 250 tons or
fraction thereof, or in case of larger tonnages, for such quantities as may be
agreed upon. Each lot of coke arising from a different source or known to be
of different quality or size shall be sampled separately.
Crushing.-The entire gross sample shall be crushed, mixed, and reduced
in quantity to convenient size for transmission to the laboratory. The sample
1626 COAL AND COKE
=tJ is:1~
IJ/
~'!'
.:0'"
.-g
::: ....
, is:l~
_~- is:~f
is:1~
tl\
FIG. 226.-Location of Sampling Points from Exposed Surface of Car.
...;!L
FIG. 227.-Drier for Coarse Samples. The outlet for air at the'top may be con-
nected with a chimney or any other device which will furnish a suitable draft. (Bulletin
No.9, Geological Survey of Ohio, p. 312.)
7 Under the standardization procedure of the Society, these methods are under the
jurisdiction of the A. S. T. M. Committee D·5 on Coal and Coke.
These methods comprise a revision and combinati.on of the former Standard Methods
of Laboratory Sampling and Analysis of Coal (A. S. T. M. Designation: D 22-24) and
of Laboratory Sampling and Analysis of Coke (A. S. T. M. Designation: D 37-24)
which methods were accordingly discontinued in 1927.
8 These methods are issued under the fixed designation D 271; the 'final number
indicates the year of original adoption as standard or, in the case of revision, the year
of last revision. Adopted, 1916, 1918; Revised, 1921, 1924, 1927, 1929, 1930, 1933.
These methods have been approved as American Standard by the American Standards
Association.
9 For det/tils of air-drying oven, see F. M. Stanton and A. C. Fieldner and W. A.
Selvig, "Methods of Analyzing Coal and Coke," U. S. Bureau of Mines Technical Paper
No.8, p. 2 (1929). Reproduced through the courtesy of the American Society for
Testing Ma~rials, 260 S. Broad St. Philadelphia, Pa.
1630 COAL AND COl\:E
(e) Grinder.-A roll crusher or coffee-mill type lof grinder suitable for
reducing the material passing a 4760-micron (No.4) sieve to pass an 840-
micron (No. 20) sieve. To reduce the moisturEj loss while crushing, a coffee-
mill type of grinder should be entirely enclosed and have an enclosed hopper
and receptacle capable of holding 10 lb. of coal.
(f) PUlverizer.-An Abbe ball mill, planeta.ry disk crusher, chrome-steel
bucking board, or any satisfactory form of pulv~rizer for reducing the material
passing an 840-micron (No. 20) sieve to pass a 250-micron (No. 60) sieve.
The porCelain jars for the ball mill should be all proximately 9 in. in diameter
and 10 in. high. The flint pebbles should be SlUooth, hard and well rounded.
(g) Large Rijjie Sampler.-A large riffle sampler with Yz or %-in. divisions
for reducing the coal passing the 4760-micron (No.4) sieve to 10 lb. (see Fig.
228). I
(h) Small Rijjie Sampler.-A small riffle ~ampler'With Kor %-in. divisions
for reducing the material passing the 840-micrOll (No. 20) and the 250-Iilicro~
(No. 60) sieves to a laboratory sample (see Fig. 229).
(i) Sieve.-An 8-in. 250-micron (No. DO) sieve with cover and receiver.
NOTE.~The sieve designations employed are those of the Standard Specifications
for Sieves for Testing Purposes CA. S. T. M. Designation: E 11) of the American Society
for Testing Materials,10 and all sieves shall conform to the detailed requirements of
these specifications. See p. 1329.
should be approximately 9 in. in diameter and 10 in. high. The flint pebbles
should be smooth, hard and well rounded. The roll!! of the hard-steel roll
crusher should revolve at the same speed.
(f) Large Riffle S"ampler.-Same as in Sectibn A (g) f9r reducing coke passing
the 4760-micron (No.4) sieve to 10 lb. (see Fig; 228). ' I
(g) Small Riffle Sampler.-T,he same as Section.A (h) (see Fig. 229).
(h) Sieve.-The same as Section A (i). I •
(C) Ball-mill Method jor Coal Appearing Wet or Dry: (a) This method of
sampling 12 does not require a total moisture sample of the coal passing an
840-micron (No. 20) sieve as do methods A and B. The coal is first air-dried
to bring it to a condition of approximate equilibrium with the air to minimize
moisture change during the preparation of the sample for analysis. After
air-drying all operations are performed with the utmost dispatch to prevent
moisture change. Fine grinding of the sample is done in an air-tight ball mill.
Total moisture is computed from the air-drying loss and the residual ,moisture
in the sample prepared for analysis.
(b) Spread the sample on tared pans, weigh, and air-dry in the special
mositure oven shown in Fig. 227, at 10 to 15° C. above room temperature until
the loss in weight between two successive weighings, made 6 to 12 hours apart,
does not exceed 0.1 % per hour. Record the loss in weight as" air-drying loss."
(c) Immediately after the last weighing, quickly crush the entire sample by
means of a roll-crusher adjusted so the product will pass an 840-micron (No. 20)
sieve. Then, without sieving, quickly reduce the coal on the small riffle sampler
to about 200 g. Put this 200-g. portion at once in the porcelain jar (I-gal.
capacity) of an Abbe ball mill, sealed air-tight by means of a rubber gasket, and
grind to pass a 250-micron (No. 60) sieve. The jar shall be slightly more than
half full of well-rounded flint pebbles, about ~ in. in diameter, and rotated .
at 60 to 75 r.p.m. Bituminous coals require from 10 to 20 minutes and anthra-
cite about 1 hour of rotation to be pulverized to the designated fineness.
(d) When pulverization is complete, pour the contents of the jar on a Yz-in.
square-mesh sieve, and separate the sample from the pebbles by shaking the
sieve over a rubber mixing-cloth. Brush carefully the pebbles and inside of
the jar to remove any adhering pulverized coal. A convenient size of sieve is
one approximately 17 <in. square. Then quickly pass the sample through a
12 This method of sampling is used by the U. S. Bureau of Mines, see F. M. Stanton,
A. C. Fieldner and W. A. Selvig, "Methods of Analyzing Coal and Coke." U. S. Bureau
of Mines Technical Paper No.8, p. 2 (1929). .
1634 COAL AND COKE
250-micron (No. 60) sieve. To minimize moisture change, keep the sieve
covered while the sample is being sieved. 'Usually !lome coarse particles,
amounting to 0.2 to 0.4% of the sample, remain on the sitive. As these particles
are likely to consist largely of ash-forming substances, reduce them on a chro-
mium-steel bucking board to pass the sieve and then add'1{0 what passed through
the sieve. Quickly mix the sample on a rubber mixing-cloth and reduce .to·
• about 50 g. by passing through the small riffle sampler. 'Put this 50-g. portion
which constitutes the laboratory sample, in a rubber-stoppered, wide-mouthl
4-oz. bottle. To minimize moisture change, the mixing and reducing of the
pulverized sample after removal from the ball mill shall be done with utmost
dispatch. The total time elapsing from the opening of the porcelain jar to the
stoppering of the laboratory sample shall not exceed 3 minutes;
Calculation.-As all the analytical determinations are made on the air-dried
sample passing a 250-micron (No. 60) sieve, calculate the analysis to "as-
received" and " dry-coal" bases from the analysis of the air-dried coal in the
following manner:
CALCULATION FROM ANALYSIS OF "AIR-DRIED" COAL TO COAL" AS RECEIVED"
. . (lOO-air-drying loss)
FIxed carbon "as recel~ed"=fixed carbon X 100 •
. (100-air-drying loss)' . .
Hydrogen" as receIved" = hydrogen X 100 -jJ1/9 alr-<irymg loss.
Oxygenin"drycoal"=(oxygen-8/9moisture)X 100. t .
1oo -molS ure
(D) For Coke: (a) Total Moisture Determination.-Dry the entire sample
receiveod at the laboratory, without any preliminary crushing, to constant weight
at a temperature of not less than 104 nor more than 200 0 C.13
Calculate the loss in weight to percentage of moisture, which shall constitute
the total moisture in the coke as received at the laboratory.
The alJowable difference in duplicate determinations by the same analyst is
0.5%.
(b) Reduction of Sample.-Crush the dried sample mechanically with a jaw
or roll crusher, or by hand on a chilled iron or hard-steel plate by impact of a
hard bar or sledge, avoiding all rubbing action, as otherwise the ash content will
be' materially increased by the addition of iron from the sampling apparatus,
even though hardened iron or steel is used. Continue the crushing until all the
sample passes through a 4760-micron (No.4) sieve, mix and reduce on the large
riffle sampler to not less than 51b.; again crush the 5-lb. sample sufficiently fine
to pass an 840-micron (No. 20) sieve; mix and reduce on the small riffle sampler
to 200 g. Transfe! this 200-g. portion to the porcelain jar of an Abbe ball mill
and pulverize to pass a 250-micron (N o. ~O) sieve. When pulverization is
complete, pour the contents of the jar on a %-in. screen and separate the sample
from the pebbles by shaking the screen. Pass all the material through a 250-
micron (No. 60) sieve, pulyerizing any coarse particles by impact in a hard-
steel diamond mortar, and mix with the remainder of the sample. Reduce the
sample through the small riffle sampler to about 50 g. and transfer to a rubber-
stoppered glass bottle.
(c) In case a ball mill is not available for fine grinding, quarter the 5-lb.
sample passing an 840-micron (No. 20) sieve to 200 g. and pulverize to pass a
250-micron (No. 60) sieve by means of a hard steel roll crusher in which the
rolls revolve at the same speed, or by impact hl a hard-steel diamond mortar.
The use of rubbing surfaces such as a <iisk pulverizer or a bucking board is
never permissible for grinding coke.
13 Experiments made at the U. S. Bureau of Mines'have shown that results checking
within 0.5% are obtained between these temperature limits. See A. C. Fieldner and
W. A. Selvig, "The Determination of Moisture in Coke," U. S. Bureau of Mines Techni-
cal Paper No. 148 (1917).
1636 COAL AND COKE
FIG. 230.-Toluene or Glycerin and Water Oven for Determining Moisture (Technical
Paper No.8, Bureau of Mines, p. 9 (1929».
METHODS OF ANALYSIS
DETERMINATION OF MOISTURE
(B) For Coke: Place the capsules containing the dried coke from the mois-
ture determination in a muffle furnace or over a burnell, and heat to redness at
such a rate as to avoid mechanical loss. Finish the ignition to constant weight
(±0.001 g.) at a temperature not exceeding 950 0 C. Cool in a desiccator and
weigh. ' \ "
NOTEs.-Befote replacing the capsules in the muffle for ignition to constant weight
the ash should be stirred with a platinum or nichrome wire. Stirring once or twice
before the first weighing hastens complete ignition.
Test the ash for unburned carbon, by moistening it with alcohol; any carbon re-
maining will show as black particles.
NOTEs.-The cover 'should fit closely enough so thllt the carbon deposit from bi.,
tuminous and lignite coals does not burn away from the under side.
Regulation of temperature to within the prescribed limits is important.
FIG. 232.-Electric Tube Furnace for Determining Volatile Matter. For 1l0-vcIt
alternating current, 60 ft. of Nichrome Wire, No. 17. B. & S. gage will give the required
temperature. :rhe temperature must be controlled by an external resistance. (U. S.
Bureau of Mines Technical Paper No.8, p. 12 (1929».
(B) For Coal.and Coke, Using Meker Burnt!r: Weigh 1 g. of the sam-
ple in a weighed platinum crucible and close with a cover or, in the case
of coke, with another crucible. Place in the flame of a No.4 Meker burner,
having approximately an outside diameter at the top of 25 mm. and
giving a flame not less than 15 cm. high. The temperature should be
1640 COAL AND COKE
DETERMINATION OF SULFUR
capacity, high or low form, or platinum crucible of similar size, and cover with
about 1 g. of Eschka mixture. .
(a) Ignition.-On account of the amount of sulfur contained in artifidal
gas, the crucible shall be heated over an alcohol, .gasoline or natural. gas flame
as described in Paragraph (b), or in a· gas or electrically heated muffle, as de-
scribed in Paragraph (c) for coal and in .Paragraph (d) for coke. The use of
artificial gas for heating the sample and Eschka mixture is permissible only when
the crucibles are heated in a muffle. .. ,
(b) Heat the crucible, placed iIi a slanting position on a ~riangle, over a
very low flame to avoid rapid expulsion of the volatile matter, which tends to
prevent complete absorption of the products of combustion of the sulfur .. Heat'
the crucible slowly for 30 minutes, gradually increasing the temperature and
stirring after all black particles have disappeared, which is an indication of the
completeness of the procedure.
(c) (For Coal).-Place the crucible in a cold muffle and gradually raise the
temperature to 8000 C. ,± 25 0 C. in about 1 hour. Maintain this maximum
temperature for about 1Yz hours.
(d) (For Coke).-Place the crucible in a warm muffle and gradually raise
the temperature to 8000 C. ± 25 0 C. in about 30 minutes. Maintain this
maximum temperature until on stirring all black particles have disappeared.
(e) Subsequent Treatment.-Remove and empty the contents into a 200-
ml. beaker and digest with 100 ml. of hot water for Yz to % hour, with occasional
stirring. Filter and wash the insoluble matter by decantation. After several
washings in this manner, transfer the insoluble matter to the filter and wash
five times, keeping the mixture ~ell agitated. Treat the filtrate, amounting
to about 250 ml., with 10 to 20 ml. of saturated bromine water, make slightly
acid with fICl and boil to expel the liberated bromine. Make just neutral to
methyl orange with N'aOH or Na 2 CO a solution, then add 1 ml. of N HCl.
Boil again and add slowly from a pipette, with constant stirring, 10 ml. of a 10%
solution of BaCI 2 ·2H2 0 .. Continue boiling for 15 minutes and allow to stand
for at least 2 hours, or preferably over night, at a temperature just below boiling.
Filter through an, ashless filter paper and wash with hot distilled water until a
AgNO a solution shows no precipitate with a drop of the filtrate. Place the
wet filter containing the precipitate of BaS04 in a weighed platinum, porcelain,
silica or alundum crucible, allowing a free access of air by folding the paper
over the precipitate loosely to prevent spattering. Smoke the paper off
gradually and at no time allow it to burp with flame. After the paper is prac-
tically consumed, raise the temperature to approximately 9250 C. and heat to
constant weight.
The residue of MgO, etc., after leaching, should be dissolved in HCI and
tested with great care for sulfur. When an appreciable amount is found this
should be determined quantitatively. The amount of sulfur retained is by
no means a negligible quantity.I6
(f) Blanks and Corrections.-In all cases a correction must be applied either
(1) by running a blank exactly as described above, using the same amount of all
reagents that w\lre employed in the regular determination, or more surely (2)
by determining a known amount of sulfate added to a solution of the reagents
16 Journal Am. Chemical Soc., 21, 1125 (1899).
1642 COAL AND COKE
aft~r these have been put through the prescribed series of operations. If this
latter procedure is adopted and carried out, say, once a ~eek or whenever a new
supply of a reagent must be used, and for a series of solutions covering the range
of sulfur content likely to be met with in the samplesj it is only necessary to
add to or subtract from the weight of BaS04 obtained from a sample, whatever
deficiency or excess may have been found in the appropriate" check" in order
to obtain a result that is more certain to be correct than if a " blank" correction
as determined by the former procedure is applied. This is due to the fact that
the solubility error for BaS04, for the amounts of sulfur in question and the .
conditions of precipitation prescribed, is probably the largest one to be con-
sidered. BaS04 is soluble 17 in acids and even in pure water, and the solubility
limit is reached almost immediately on contact with the solvent. Hence, in
the event of using reagents of very superior quality or of exercising more than
ordinary precautions, there may be no apparent" blank," because the solubility
limit of the solution for BaS04 has not been reached or at any rate not exceeded.
(g) Calculate sulfur content as follows:
P t f Sulf _ (Weight of BaS04-blank)X13.74
ercen age 0 ur- Weight of sample
NOTE.-As shown in the preliminary report,l~ the Atkinson and sodium-peroxide
methods give results in close agreement with the Eschka method. Register 19 has
shown that if 5% of nitrogen is present in the gases contained in the bomb calorimeter
the sulfur of a sample is almost completely oxidized to H 2S0 4 and the washings of the
calorimeter may be used for the determination of sulfur. .
tate the sulfur with barium chloride and determine as specified under the
Eschka method. I
(b) Blank Correction.-A blank correction shall be applied by running a
blank on the reagents used. I,
FIG. 233.-Denver Fire-Clay F,urnace and No.3 Melter's Furnace, with Accessories.
The stock design of the No.3 Melter's Furnace should be modified by pro~
viding the upper cylinder with two holes in the side; a 2-in. observation hole
with its center 4 in. from the top of the cylinder (excluding cover plate), and
a I-in. thermocouple hole 90° to the right of the observation hole; the bottoms
of these holes being in the same horizontal plane. The coal-ash fusion furnace
as furnished by the Denver Fire Clay Co. is provided with suitable observation
]646 COAL AND COKE
and thermocouple holes. The interior of the furnaces are cylindrical' and
approximately 7 in. in diameter. Counter-balanced sh~et-iron canopies con-
nected with telescopic flues to an exhaust system are placed over the furnaces
for conducting the hot gases out of the room. Two fufnaces and' accessories
are shown in Fig. 233. I
(b) Refractory Crucibles.-The interior of the fusion furnace as arranged
for making a test is shown in Fig. 234. Crucible a is a Corundite crucible
about 3 in. in inside diameter and about 472 in. in height outside, with a wall
thickness of about 72 in. The crucible is provided with a cover, b, and shall
have two holes in the side, an observation hole 2 in. in diameter and a thermo-
couple hole 1 in. in diameter, 90° to the right of the observation hole. The
bottoms of the two holes shall be in the same horizontal plane. The crucible is
supported by a Corundite support, c, so that the two holes in the crucible are in
line with the correllPonding holes in the furnace wall. The ash cones, d, are
supported on Corundite disk e so that they are visible through the observation
hole of the crucible and furnace wall.
,.
•.•...•~
.'
(e) Brass Cone Mold.-A brass cone mold for making ash cones, % in.
high and U in. at each side of the base, which is an equilateral triangle, is
illustrated in Fig. 235 .
. (f) Pyrometer. 23-The temperature measurements shall be made with a
thermocouple o! platinum and platinum-rhodium used in conjunction with a
high-resistance millivoltmeter or potentiometer; or with an optical pyrometer
of the disappearing filament type, which has been calibrated to be accurate to
10° C. up to 1400° C. and to 15° C. from 1400° to 1600° C. If a theqnocouple
is used it should be protected from the furnace gases by a glazed porcelain tube
and so placed in the furnace that the hot junction is in the immediate vicinity
of the cones. The pyrometer equipment shall be checked frequently by
mounting small pieces of pure gold-and nickel in the same manner as the cones'
With a strong reducing atmosphere, 1452° C. should be obtained for the melting
point of nickel. and 1063° C. for the gold. The pyrometer equipment should
23 For detailed information concerning the use of pyrometers see "Pyrometric Prac-
tice," U. S. Bureau of Standards Technologic Paper No. 170 (1921).
1648 COAL AND COKE
also be standardized from time to time thr.ough the temperature range for
which it is used, by a suitably equipped standardizingJ~boratory, such as that
of the U. S. Bureau of Standards. '
--(g) If temperature measurements are made by sightin'g an optical pyrCimeter
through the glass window of the observation tube the~ follov~ing corrections
shall.be added to the observed temperatures to correct ~or absorption of light
by the glass window:
\
ULTIMATE ANALYSIS
(For Coal and Coke)
CARBON AND HYDROGEN
Method With Gas-Heated Combustion Furnace
The determination of carbon and of hydrogen shall be made with a weighed
quantity of the sample in a 25-burner combustion furnace of the Glaser type.
The products of combustion shall be thoroughly oxidized by passing them over
red-hot CuO and PbCr04, and shall be fixed by absorbing the water in a weighed
Marchand tube filled with granular CaC1 2 and by absorbing the Co. 2 in a Liebig
bulb containing a 30% solution of KOH.
Apparatus.-The apparatus used shall consist of a purifying train in dupli-
cate l .a combustion tube in the furnace, and an absorption train.
(a) The purifying train shall consist of the following purifying reagents
arranged in order of passage of air and oxygen through them: H 2 S0 4, KOH
solution. soda lime, and granular CaCho One of the trains is for air and one
for oxygen. In the H 2 S0 4 and KOH scrubbing bottles the air and the oxygen
1650 COAL AND COKE
shall be made to bubble through about 5 mm. of the purifying reagent. Both
purifying trains shall be connected to the combustion tube by a Y-tube, the
joint being made tight by a rubber stopper. '
(b) The combustion ~ube shall be made of hard Jena glass, or its equivalent.
Its external diameter shall be about 21 mm., and its itotal length 1 meter.
The first 30 cm. of the tube shall be empty; following this empty space is an
asbestos plug (acid-washed and ignited) or in its place a ~oIl of oxidized copper
gauze may be used; the next 40 cm. shall be filled with" wire" Cu 0; a second
asbestos plug separates the copper oxide from 10 cm. of fused PbCr04, whi~h
shall be held in place by another asbestos plug 20 cm. from the end of the tube.
The end of the tube shall be drawn out for rubber-tubing connection with the
absorption train.
(e) The absorption train shall consist of a Marchand tube filled with granular
CaCh to absorb moisture. The CaCh should be saturated with CO 2 before
uxing. The Marchand tube shall be followed by a Liebig bulb containing a
30% KOH solution, in which any possible impurities, as ferrous iron or nitrate:;,
have been oxidized by a little KMn04. A guard tube, containing granular
CaCh and soda linie, shall be attached to the Liebig bulb to absorb any CO 2
escaping the KOH solution and any water evaporating from that solution.
Procedure.-Connect the train to an aspirator which draws the products
of combustion through the entire train. A guard tube of CaCh prevents
moisture from running back into the absorption train. Maintain the suction
constant by a Mariotte flask. The advantage of aspirating the gases through
the train rather than forcing them through by pressure is that the pressure on
the rubber connections is from the outside, so that gas-tight connections are
more easily maintained than if 'the pressure is on the inside of the tube. Make
the connections as tight as possible. The usual test for tightness is to start
aspiration at the rate of about three bubbles of air per second through the
potash bulb, and then to close the inlet for air and oxygen at the opposite end
of the train; if there is no more than one bubble per minute in the potash bulb,
the apparatus shall be considered tight. (
Before starting a determination when the train has been idle some hours,.or
after any changes in chemicals or connections, run a blank by aspirating about
1 liter of air through the train, which is heated in the same manner as if a
determination on the sample were being made. If the Liebig bulb and the tube
containing calcium chloride show a change in weight of less than 0.5 mg. each,
the apparatus is in proper condition for use. ...._
Use a porcelain or platinum boat provided with a glass weighing tube of
suitable size, which is fitted with an accurately ground glass stopper. Weigh
the tube and empty boat. Quickly place in the boat approximately 0.2 g. of
the air-dry sample (passing a 250-micron (No. 60) sieve and finer, or better,
_passing a 149-micron (No. 100) sieve if much free impurity is present). Place
the boat at once in the weighing tube, which is quickly stoppered to prevent
moisture change in the sample while weighing, and transfer to the furnace.
Connect the absorption tubes and transfer the boat and sample from the weigh-
ing tube to the combustion tube, which should be cool for the first 30 cm. The
CuO should be red hot and the PbCr04 at a dull-red,heat. Transfer the boat
from the weighing tube to the combustion tube as-:1apidly as possible.
COAL AND COKE 1651
As soon a.s the boat is in place near the asbestos plug at the beginning of the
copper oxide ins3rt the stopper connecting with the purifying train and start
the aspiration with pure oxygen gas at the rate of three bubbles per second.
Turn on one'burner about 10 cm. back from the boat, and continue the aspira-
tion carefully until practically all the moisture is expelled from the sample.
Then increase the heat very gradually until all the volatile matter has been
driven off. In driving off the volatile matter the heat must be ltpplied gradually
in order to prevent a too rapid evolution of gas and tar, which may either escape
complete combustion 6r may be driven back into the purifying train. Increase
the heat slowly by turning on more burners under the open part of the tube until
the sample is ignited; then the temperature may be increased rapidly, but care,
should be taken not to melt the combustion tube.
Any moisture collecting in the end of the combustion tube or in the rubber
connection joining it to the CaCl 2 tube may be driven over into the CaCl 2 tube
by carefully warming with a piece of hot tile. Continue the aspiration with
, oxygen for 2 minutes after the sample ceases to glo_w, then turn off the heat and
aspirate about 1200 ml. of air. Disconnect the absorption bulbs, wipe with
a clEian cloth, and allow to cool to the balance-room temperature before weighing.
Calculate the percentages of hydrogen and carbon as follows:
P t fh d -11 19 X (increase in weight of CaCb tube)
ercen age 0 y rogen- . Weight of sample
Percentage of carbon =27 27X (increase in weight of KOH bulb)
. Weight of sampl
Weigh the ash in the boat and carefully inspect it for any unburned carbon,
which would destroy the value of the determination.
NITROGEN
. The Kjeldahl-Gunning method is recommended for the determination oi
nitrogen. This method has the advantage over either the simple Kjeldahl or
the Gunning method, in requiring less time for th~ complete oxidation of the
organic matter, and in giving the most uniform results.
COAL AND COKE 1653
OXYGEN
There being no satisfactory direct method of determining oxygen it shall
be computed by subtracting the sum of the percentages of hydrogen, carbon,
nitrogen, sulfur; water and ash from 100. The result so obtained is affected
by all the errors incurred in the other determinations and especially by the
change. in weight of the ash-forming constituents on ignition. In the case of
1654 COAL AND COKE
coal, iron pyrite changes to ferric oxide, increasing the ash and causing a nega-
tive error in the oxygen equivalent to three-eights of \the pyritic sulfur. On
the other hand, there is always a loss on ignition, of water of composition
from the clayey and shaley constituents, CO 2 from carbdnates, etc., whigh tends
to compensate the absorption of oxygen. !
Corrected Oxygen (for Coal).-When a more correct oxygen value is de-
sired, it may be obtained by making the corrections indicated in the following
formula:
Corrected oxygen=100-[(C-C')+(H-H')+N+H,O+S'+corrected ash]
where C = total carbon;
C'=carbon of ca'rbonates;
H = total hydrogen less hydrogen of water;
H' = hydrogen from water of composition in clay, shale, etc., ~
. N=nitrogen;
. H 20 = moisture as found as 105° C.;
S' = sulfur not present as .pyrite or sulfate. This is usually
small. In many types of cdal it may be disregarded.
Corrected ash = mineral constituents originally present in the coal. For
most purposes this can be determined with sufficient accuracy by
adding to the ash, as found, five-eighths of the weight of pyritic
sulfur, the CO 2 of carbonates and the water of composition of,
clay, shale, etc. See alio Determination of Ash, Sections
10 and 11.
CALORIMETRIC DETERMINATION
using, if necessary, a piece of platinum wir"e somewhat heavier than the iron
wire, to make the connection. The platinum and the iron shall both be clean.
About 0.5 m!. of water should be placed in the bottom of the bomb to saturate
with moisture the oxygen used for combustion. Whet} the crucible is put in
place in the bomb, the firing wire should touch the sample or briquet of standard
material. For the combustion of standardizing samples, iron wi!;e is preferable
to platinum.
(c) Filling the Bomb with Oxygen.-Oxygen from the supply tank shall be
admitted slowly to avoid blowing the sample from the crucible without dis-
placing the original air content, and the pressure allowed to reach 20 atmos-
pheres for the larger bombs or about 30 atmospheres for the smaller bombs, so
that the bomb shall contain an amount of oxygen sufficient for complete com-
bustion, namely, at least 5 g. per gram of the sample. This method of filling
should insure sufficient nitrogen for complete oxidation of the sulfur in the coal. 27
(d) Calorimeter Water.-The calorimeter shall be filled with the required
amount of distilled water, depending upon the type of calorimeter. The
amount may be determined either by measurement in stand~rdized flask or by
weighing. The amount shall be kept the same as that used in the standardiza-
tion of the apparatus.
(e) Temperature Adjustments.-The initiil temperature in the calorimeter
snaIl be so adjusted that the final temperature, after the combustion, will not
be more than 1° C., preferably about 0.5° C., above that of the jacket, unGer
which conditions the total correction for heat gained from or lost to the sur-
roundings will be small when the rise of temperature is 2° or 3° C., and the
effect of evaporation will also be small.
" (f) Firing Current.-The eleotric current used for firing the charge shall be
obtained from storage or dry cells having an electromotive force of not more
than 12 volts. A higher voltage is liable to cause an arc between the firing
terminals, introducing additional heat, whIch cannot be measured with cer-
tainty. The circuit should be closed by means of a switch which should remain
closed for not more than 2 seconds. When possible, it is recommended that an
ammeter be used in the firing circuit to indicate when the firing wire has burned
out.
(g) Method of Making an Observation.-The bomb when ready for firing,
shall be placed in the calorimeter, the firing wires connected, the cover put in
place and the stirrer and thermometer so placed as not to be in contact with
the bomb or container. The stirrer shall then be started~and _l!:fter the. ther-
mometer reading has become steady, not less than 2 minutes after'the stirrer is
started, temperatures shall be read at I-minute intervals for 5 minutes and the
charge then fired,. the exact time of firirg being noted. Observations of
temperature shall then be made at intervals depending upon the method to be
used for computing the cooling correction. When the temperature has reached
its maximum and is falling uniformly, a series of thermometer readings shall be
taken at I-minute intervals for 5 minutes to determine the final cooling rate.
(h) Titration.,-Mter a combustion and after allowing the gas to escape the
bomb shall be opened and the inside examined for traces of unburned material
I
27S. H. Regester, "Oxidation of Sulfur Compounds of .Coal, and of Nitrogen in the
Bomb Calorimeter, and the Correction to pe Applied in I>eiermining the Heating Value
of Coal," Industrial and Engineering Chemistry, 6, 812 (1914).
COAL AND COKE 1657
or sooty deposit. If these are found, the observations shall be discarded. If
the combustion appears complete, the bomb shall be rinsed out thoroughly· and
t!J.e washings titrated with a standard alkali solution (1 ml. = 0.02173 g. HNO a =
5 calories) using methyl-orange or methyl-red indicator, to determine the
amount of acid formed. A correction of 230 calories per gram of HNO a should
be subtracted from the total heat observed. An additional correction of 1300
calories per gram of sulfur in the sarn,ple should be made for the excess of dif-
ference in heats of formation of S02 and aquequs H 2S0 4 over the heat of forma-
tion of aqueous HNO a• '
Calculation of Results: The following method 9f calculation is recommended
to take the place of the Pfaundler or other similar formulas for calculating the '
cooling correction (radiation correction) :
Observe (1) the rate of rise, r1, of the calorimeter temperature in degrees per
minute for 5 minutes before firing; (2) the time, a, at which the last temperature
reading is made immediately before firing; (3) the time, b, when the rise of
temperature has reached six-tenths of its total amount (this point can generally
be determined by adding to the temperature observed before firing, 60% of the
expected 2~ temperature rise, and noting the time at which this point is reached) ;
(4) the time, c, of a thermometer reading taken when the temperature change
has become uniform some 5 minutes after firing; (5) the final rate of cooling,
12, in degrees per minute for 5 minutes.
Multiply the rate 11 by the time b - a in minutes and tenths of a minute, and
add to this product (subtracted if the temperature was falling at the time a)
to the thermometer reading taken at the time a. Multiply the rate 12, by the
time c-b and add this product (subtracted if the temperature was rising at the
time c and later) to the thermometer reading taken at the time c. The differ-
ence of the two thermometer readings thus corrected, provided the corrections
from the certificate have already been applied, gives the total rise of tempera-
ture due to the combustion. This multiplied by the water equivalent of the
calorimeter gives the total amount of heat liberated. Divide this result, cor-
rected for the heats of formation of HNO a and H 2S0 4 observed and for the
heat of combustion of the firing wire, when that is included, by the weight of
the charge to find the heat of combustion 'in calories per gram. Calories per
gram mUltiplied by 1.8 give the British thermal units per pound. (See example.)
NOTE.-In practice, the time b-a will be found so nearly constant for a given
calorimeter with the usual amounts of fuel that b need be determined only occasionally.
The results shall be reduced to calories per gram or British thermal units per pound
of dry sample, the moisture being determined upon a sample taken from the bottle at
about the same time as the combustion sample is taken.
28 When the temperature rise is not approximately known beforehand, it is only
necessary to take thermometer readings at 40, 50, 60 seconds (and possibly 70 seconds
with some calorimeters) after firing, and from th~se observations to find when the
temperature rise had reached 60% of the total. Thus, if the temperature at firing was
2.135° C., at 40 seconds 3.05°, at 50 seconds 3.92°, at 60 seconds 4.16°, and the final
temperature was 4.200°, the total rise was 2.07°; 60% of it was 1.24°. The temperature
to be observed was then 2.14°+1.24°=3.38°. Referring to the observations at 40 and
50 seconds, the temperatures were respectively 3.05 and 3.92°. The time corresponding
to the temperature of 3.38° was, therefore: '
3.38-3.05 10 44
40+3.92-3.05X d
=' secon s.
1658 COAL AND COKE
Example.-
OBSERVATIONS.
Water equivalent = 2550 g.
Weight of charge = 1.0535 g. I
Approximate rise of temperature expected '13.2° C.
60% of approximate rise = 1.9° C.,
Thermometer
I
Readings, \ Corrected Temperature
Tim~ Deg. Cent. (or corrected Beckmann thermometer readings)
10-21 15.244 Thermometer corrections taken from .the certificate .
22 15.250
23 15.255
24 15.261
25 15.266
(a) 26 15.272 15.276° C.
Charge fired
(b) 27.2 17.2<>
(c) 31 18.500 18.497° C.
32 18.498
33 18.497
34 18.496
35 18.494
36 18.493
a The initial temperature is 15.27° C.; 60% 6f the expe"ted rise is 1.9° C. The
reading to observe is then 17.2° C.
CALCULATION
The inside of the crucible is lined completely with ignited asbestos in a thin layer
pressed well down into the angles. The sample is then sprinkled evenly over the surface
of the asbestos. Otherwise the procedure is as previously described.
The method of computing the "cooling correction" described in U. S. Bureau of
Mines Technical Paper No.8, pp. 26-30 (1929), may also be used. I
ANALYSIS OF EXPLOS~S 1
The methods d~scribed in this chapter cover all Q( the more common types
of explosives employed in the United States for both commercial and military
purposes. Those of chief commercial importance are black powder, nitro-
glycerin dynamites, including "straight" dynamites, ammonia dynamites,
gelatin dynamites, and low-freezing dynamites, "Permissible" coal mining
explosives and nitrostarch blasting explosives. Military explosives include
smokeless powder, guncotton, trinitrotoluene, picric acid, ammonium picrate,
" Amatol," tetryl and tetranitroaniline. No sharp distinction can, however,
be drawn between commercial and military explosives, as many are utilized
for both purposes.. Many tons of surplus TNT on hand at the end of the
World War were used in commercial work; nitrostarch explosives found impor-
tant application for military use during the war; mercury fulminate, lead azide,
and other detonators and priming compositions are essential in every field of
explosives.
The methods described have largely been used by the writer in practical
explosives testing and analysis in connection with both Government and pri:-
vate work. Most of those applying to commercial explosives have been
approved by. the United States Bureau of Mines for use in its Explosives
Chemical Laboratory.2
BLACK POWDER
The composition of black powder varies to some extent, depending chiefly
on the purpose for which the explosive is to be'used. Black blasting powder
contains sodium nitrate, charcoal and sulfur; black gunpowder),s quite similar
except that potassium nitrate is generally substituted for the soaium nitrate;
black fuse powder is similar to the latter, differing mainly in its granulation.
The same general method of analysis is therefore applicable to all types of
black powder.
Sampling.-From 50 to 100 grams of the original sample is crushed in small
portions in a porcelain mortar and completely p~ssed through an 80-mesh
1 Chapter by C. G. Storm. Published by permission of Chief of Ordnance, U. S. A.
2 See Bureau of Mines Bulletin No. 51, "The Analysis of Black Powder and Dyna-
mite," W. O. Snelling and C. G. Storm, 1913, Bulletin No. 96, "The' Analysis of Per-
missible Explosives," C. G. Storm, 1916, Bulletin No. 219, "Explosives, their Mate-
rials, Constitution and Analysis," C. A. Taylor and W. H.,Rinkenbach, 1923, and Tech.
Paper No. 282, "Analysis of Detonating and Priming..'Mixtures;" C. A. Taylor and
W. H. Rinkenbach" 1922. -
1660
ANALYSIS OF EXPLOSIVES 1661
sieve, ca:r:e being taken to avoid undue exposure to the air. The separate
powdered portions are promptly bottled and the entire sample is finally well
mixed.
Moisture.-The standard method of the Bureau of Mines is to desiccate a
2-grarr. sample on a 3-inch watch glass over sulfuric acid for three days, the
loss of weight being moisture. It has been shown, however, that equally
accurate results can be obtained by drying at 70° C. in a constant temperature
oven to constant weight, for which 2-3 hours is usually sufficient. As much
as 5 hours drying at 70° C. will not cause loss of sulfur. Drying at 100° C.
gives results which are slightly high, due to loss of sulfur.
Nitrates.-About 10 grams of the finely ground sample in a Gooch crucible'
provided with an asbestos mat, is extracted with warm water by means of
suction, the water being added in 15-20 ml. portions and each portion being
allowed to stand in the crucible a short time before suction is applied. About
200 mI. of water is usually sufficient, but the last drops of filtrate should be
tested by evaporation to ensure the absence of nitrates. A blue color on the
addition of sulfuric acid containing a few crystals of diphenylamine will also
indicate the presence of nitrates.
The water extract includes a small amount of water-soluble organic material
from the charcoal in addition to tlie nitrate. It is made up to 250 m!. and an
aliquot portion (50 mI.), evaporated to dryness on the steam bath, treated with
a little nitric acid, again evaporated, heated to slight fusion and weighed.
If allowance for impurities in the nitrate is desired, a direct determination
.of nitrate may be made on a separate portion of the water extract by the
Devarda method or by means of the nitrometer, but for all practical purposes
the evaporation method is sufficient. The usual tests should be made to
determine whether sodium nitrate or potassium nitrate is present.
Some black powders contain baJium nitrate in addition to potassium nitrate.
Barium nitrate, if present, is determined in an aliquot portion of, the water
extract by precipitating with sulfuric acid and' weighing the precipitate of
barium sulfate.
The residue left in the crucible, consisting of sulfur and charcoal, is dried
at about 70° C. to constant weight (for 5 hours or over night if more con-
venient), the loss of weight minus the moisture content being the water-
soluble portion. This result serves as a check on the evaporation result.
Sulfur.-The residue in the crucible is extracted in. a Wiley extractor or
other continuous extraction apparatus with carbon disulfide, until evapora-
tion of a small portion of the solvent passing through the crucible shows
absence of sulfur. The excess of carbon disulfide is then allowed to evaporate
from the crucible in a warm place away from flame, and the residue finally dried
to constant weight at 100° C. The loss of weight is considered as sulfur.
Charcoal.-The dry residue in the crucible should consist only of charcoal.
Ash.-The ash in the charcoal may be determiJ,ed by ignition -over a
Bunsen burner until all of the carbon has been burned off, and weighing.
This ash also contains, of course, any non-volatile matter that may have been
present in the sulfur and nitrate. An ash content in excess of 1% (calculated
on the weight of powder taken for analysis) may be regarded as an indication
of incomplete extraction of nitrate.
1662 ANALYSIS OF EXPLOSIVES
NITROGLYCERIN DYNAMIl'ES
" Straight" Dynamite
So-called "Straight" nitroglycerin dynamite has been manufactured to
only a relatively small extent in this country for a number df years. This was
due in part to wartime developments in tli'e'manufacture of both ammoniilm
nitrate and nitrocompounds, and in part to the high price of glycerin, whixh
reached its maximum about the early part of 1927. During recent years tne
use of ethylene glycol dinitrate (more commonly called nitroglycol) as a sub-
stitute for a considerable part of the nitroglycerin in various types of dynamites
and "permissible explosives" has greatly decreased the amount of straight
dynamites manufactured. "Straight" nitroglycerin dynamites are still used
to some extent, however, particularly where quick-acting dynamites of high
strength are required, as in blasting in hard rock. They consist essentially of
nitroglycerin absorbed in a " dope" composed of It combustible absorbent,
usually wood pulp, and an oxidizing material (sodium nitrate), to which has
been added a small amount of an antacid (calcium caljl)onate, zinc oxide; etc.).
The analysis is best carried out by successive e;xtractions, usually with ether,
water and hydrochloric acid. .
Sampling.-The wrappers are removed from a number of the cartridges,
and from 3 to 5 cm. of the ends of the exposed roll of explosives rejected. The
remainder is thoroughly mixed on a large sheet df paraffined paper or in a lsrge
porcelain dish, and an average sample selected and b()ttleq-usull,lly about one
half pound. The importance of thorough mixing of the sample must not be
overlooked, in view of the fact that there is frequently a decided"tendency for
the nitroglycerin to segregate due to insufficient or unsuitable absorbent, ,so
that this liquid ingredient may not be uniformly distributed throughout the
cartridge. Also if a carefully mixed sample has bee~ allowed to stand for some
days, especially in a warm place, segregation may QCcur in the bottle, so that
it is advisable to mix the sample again before an'alysis.
Qua1itati~e Examination.-Although a qualitative analysis of a sample
known to be straight nitroglycerin dynamite is usually unnecessary, the exact
nature of the sample may be unknown, and a knowl!ldge of the composition of
some of the more complex types of dynamite is n.ecesaary before a quantitative
analysis can be properly conducted. -
ANALYSIS OF EXPLOSIVES 1663
acid and the sulfur almost insoluble. The: sulfur is filtered from the solu-
tion, washed slightly 'with alcohol to remove the acetic acid solution, dried
and weighed.
If a considerable quantity of crystals of sulfur is f6und in the evaporated
ether extract, it is possible that all of the sulfur has ndt been"removed by the
ether, and in this case an extraction is made with carbon disulfide, in exactly
the same manner as the ether extraction. This extraction is made subsequent
to the extraction with water, the sulfur being determined by loss of weight
of the residue or by direct weight after evaporation of the carbon disulfide,
away from free flame. "
Extraction with Water and Determination of Nitrates.-The dried and
weighed residue left in the crucible after extraction with ether, is extracted
with water, using a suction flask fitted with a carbon filter tube in which the
crucible is held by a short length of thin-walled rubber tubing. Cold water is
used for this extraction, as hot water would gelatinize any starch present.
A total of at least 200 ml. of water is passed through the sample, in at least 10
portions, each portion being allowed to stand in contact with the residue foria
few minutes before being sucked into the flask. An evaporation test of a few
drops of the filtrate will determine the completeness of the extraction. When
the extraction is complete, the crucible with its insoluble residue is dried for
5 hours, or overnight; at 95°-100° C., and the loss of weight noted as total
water-soluble material. This includes nit,rates and other soluble salts that may
be present, together with water extract from the wood pulp, flour or other
absorbent. This soluble organic material may amount to as much as 2% of
the total sample, when cereal products are present. Calcium, magnesiul}l,
or'zinc may also be present in solution, resulting from the action of acid decom-
position products of the nitroglycerin on the carbonate or other antacid present.
In routine analyses of ordinary dynamite, the loss of weight on extraction with
water is usually considered as the alkaline nitrate (sodium or potassium), but
where more exact results are desired an aliquot portion of the extract is evap-
orated to dryness with a little nitric acid to oxidize otganic materials, and the
residue weighed as alkaline nitrate. This weight may/be corrected for inorganic
impurities-chlorides, sulfates, iron, aluminum, calcium, etc,-determined
separately by the usual methods. '
Nitrates may be determined by means of the nitrometer, using an aliquot
portion of water extract estimated to contain '.6 to .8 gram of NaNO a or .8
to 1.0 gram of KNO a• This is evaporated on the steam batli"alm,ost to dryness
and transferred with as little water as possible, to the cup of the nitrometer.
This solution is drawn into the generator and 30 to 40 m!. of 95--'96% H 2S0 4
added slowly so as to avoid generating sufficient heat to crack the glass. The
generator is then shaken for a total time of 8-10 minutes in order to be certain
that the generation of gas is complete with the (diluted acid. The gas is
measured and the, % of nitrate calculated as in the case of nitroglycerin.
Extraction with Acid.-When starch is not present in the rejSidue, a simple
extraction of the residue insoluble in water is made with cold dilute HCl
(1 : 10), 100 ml. being drawn through the sample in the crucible in small suc-
cessive portions as described under " Extraction with Wat'er." Several por-
tions of water are then drawn through to wash out-the acid, and the residue in
the crucible dried for 5 hours at 95° to 100° C. The loss of weight is usually
ANALYSIS OF EXPLOSIVES 1667
reported as antacid, but the base dissolved may be determined by the usual
quantitative methods if desired. The acid-soluble materials generally present
are calcium or magnesium carbonate or zinc oxide.
Determination of Starch.-If starch is present in the residue insoluble in
water, it is removed together with the antacid by boiling with dilute acid.
The residue is moistened with water, scraped or washed out of the crucible into
a 500-ml. beaker, the volume brought to about 250 ml. by the addition of water
and 3 ml. of concentrated HCl, and the mixture boiled until a drop of the
solution fails to give a blue color when treated on a spot plate with a drop of a
solution of iodine in KI. This indicates that the starch has been completely
hydrolyzed to dextrin. The mixture is then filtered through a fresh crucible,
washed with water, dried and weighed, correction being made for the weight
of the asbestos mat of the original crucible. ,
The antacid dissolved in the acid filtrate is determined as already described.
The loss of weight by the boiling treatment; minus the antacid found, repre-
sents starch and other dissolved organic materials removed from cereal products
or wood pulp. The insoluble residue includes the wood pulp and the crude
'fibre of the cereal products.
Because of the impracticability of exact separations it is customary to
report all of the soluble organic material included in both. water and acid
extractions as "starch" or "starchy material," and the insoluble organic
residue as " wood. pulp and' crude fibre," or the sum of these organic materials
is often reported as " carbonaceous combustible material."
Insoluble Residue and Ash.-The insoluble residue may contain wood pulp
or sawdust, the crude fibre from various cere!l,l products such as corn meal,
wheat flour, middlings, bran, etc., ground nut shells, vegetable ivory meal,
and more rarely' inorganic material such as infusorial earth (kieselguhr), clay,
etc. These can usually be identified by microscopic examination (see Bureau
of Mines Bulletin 96, Page 74), and a determination of the ash will show
whether inorganic materials are present. A high ash content may also' indicate
incomplete water or acid extractions.
Ammonia Dynamite
So-called ammonia dynamite is essentially" straight" dynamite in which a
large part of the nitroglycerin is replaced by ammonium nitrate. The am-
monium nitrate is frequently protected from moisture by a coating of vaseline
or paraffin and is usually neutralized with zinc oxide. This type of dynamite
generally contains Jess wood pulp than the corresponding grades of " straight"
dynamite, and sulfur and cereal products, such as low grade flour, are usually
present.
The determination of moisture and the various extractions are carried out
as described for "straight" dynamite. An extraction with carbon disulfide
is usually necessary to effect complete removal of the sulfur; this properly
follows the,extraction with water. The analysis of the ether extract may be
conducted as already described. In drying the residue left in the crucible
after extraction with ether, it is important that a temperature of approximately
70° C. be used, because in the presence of ZnO, the loss of ammonium nitrate
is considerable at 100° C. Pure ammonium nitrate is not appreciably affected
1668 ANALYSIS OF EXPLOSIVES
by even 24 hours heating at 100° C., but the presence of the ZnO causes de-
composition at this temperature. "- 1
The water extract contains sodium nitrate and ammonium nitrate together
with practically all of the zinc oxide present, the latt~r ingredient being dis-
solved with the ammonium nitrate, and a small amount of soluble organic
material from the flour or other absorbent. It is analyzed as follows: An
aliquot portion is evaporated to dryness in a platinum o'r silica dish on a steam
bath, the ammonium nitrate volatilized by careful heating over a burner, a
little nitric acid added to re-oxidize any nitrate that may have been reduced to.-
nitrite, and the residue again dried on the steam bath. The zinc oxide is now
in the form of zinc nitrate and may be separated from the sodium nitrate by
either of the following methods:
1. The residue is dried at 110°-120° C. and weighed as NaN0 3 and Zn
(NOah. It is then dissolved in water, the zinc precipitated with Na 2 C0 3,
filtered, ignited and weighed as ZnO, and the NaN0 3 taken by difference; the
total NaN0 3 plus Zn(N0 3 )2 minus (ZnOX2.33) =NaN0 3 •
2. The residue is gently heated over a burner until evolution of oxides of
nitrogen from decomposition of the Zn(N0 3)2 has ceased, and the remaining
residue weighed as NaN0 3 and ZnO. It is then treated with water, the in-
soluble ZnO filte.red on a Gooch crucible, ignited and weighed, the NaN0 3 being
taken by difference.
Ammonium nitrate is determined in a ~eparate portion Qi'the water extract
by the usual method of distillation and titration.
The sum of the amounts of NH 4 N0 3, NaN0 3 , and ZnO found will be some-
what less than the total water extract owing to the presence of water-soluble
organic material from the carbonaceous absorbents.
Gelatin Dynamite
This is a form of nitroglycerin explosive in which the nitroglycerin, instead
of being absorbed in porous materials such as wood pulp, is combined with
n'itrocellulose 'in the form of a gelatinous plastic mas~. As little as 3.5% of
suitable grade of nitrocellulose" containing about 12% nitrogen will, when
heated with nitroglycerin, at about 60° C., form a jelly-like non-fluid mass
when cooled to ordinary temperature. "Blasting gelatin," used to a consider-
able extent where great explosive strength is required, is a stiff_colloid composed
of 90 to 93% nitroglycerin 'and 10 to 7% nitrocellulose.
All blasting explosives containing such colloids of nitroglycerin and nitro-
cellulose combined with an active" dope" or base, consisting of a nitrate and
combustible material, are termed gelatin dynamites. This type of explosive
is also known in some countries as " Gelignite."
Sampling.-Owing to its pasty consistency the sample of gelatin, dynamite
must be prepared by cutting portions of a number of cartridges into thin bits
with an aluminum or platinum spatula. The use of a steel spatula or knife
for this purpose is not to be recommended for reasons of safety.' An ample
quantity of sample thus prepared is well mixed and bottle9,. Owing to its
tendency to again form a solid mass upon standin_g,.:.it- should be analyzed as
soon as possible after being prepared. .
AN AL YSIS OF EXPLOSIVES 1669
Analysis.-The principle ingredients that may be found in the different
types of. gelatin dynamite are nitroglycerin; nitrocellulose; sulfur; rosin;
sodium, potassium or ammonium nitrate; calcium or magnesium carbonate;
wood pulp, cereal products and similar carbonaceous combustible materials.
Low-freezing gelatins may also contain nitrotoluenes or other nitrosubstitution
compounds, etc.
Moisture is determined as described for "straight" dynamite, and the
extraction with ether made in the usual manner except that ether free from
alcohol (distilled over sodium) is used in order to prevent partial solution of
. the nitrocellulose. The latter is readily soluble in a mixture of ether and
alcohol, arid as ordinary U. S. P. ether contains about 4% of alcohol, there is a
possibility that an appreciable part of the 0.5% to 2.0% of nitrocellulose
present in the sample will be dissolved unless pure ether is used. The ether
extract is evaporated and analyzed as already described and the water extrac-
tion made in the usual manner. If more than 1 or 2% of sulfur was present it
will not have been completely removed by the ether, unless the extraction was
continued for a sufficiently long time. In this event, it is necessary to make
an additional extraction with carbon disulfide in the Wiley apparatus sub-
sequent to extraction with water.
NitroceUulose.-After the extractions with ether, water, and CS 2 (if
necessary) have been made, the nitrocellulose is determined, preferably by
extraction with acetone, which is a better solvent for the purpose than a mix-
ture of ether and alcohol. It is advisable to separate the dry residue from the
crucible, leaving the asbestos mat ihtact if possible. The residue is trans-
ferred to a small beaker, covered with acetone and allowed to stand at least
3 or 4 hours with occasional stirring. It is then filtered through the original
crucible, washed with acetone, dried and weighed, the loss of weight being
regarded as nitrocellulose. To correct for small amounts of extract from the
wood pulp or other carbonaceous material, the acetone solution may be evap-
orated to about 20-25 mI., and diluted gradually with a large volume (about
100 ml.) of hot water, which volatilizes the acetone, precipitating the nitro-
cellulose as a white flocculent mass, which is filtered, dried, and weighed.
The remainder of the analysis is conducted ,as for straight dynamite.
It will be found that the results of analysis of a gelatin dynamite do not
agree with its trade markings. For example, the usual "40% strength"
gelatin dynamite actually contains from 30 to 33% of nitroglycerin and about
1% of nitrocellulose. Weight for weight this explosive is considerably weaker
than 40% straight dynamite, which contains 40% of nitroglycerin.
Low-Freezing Dynamite
Low-freezing dynamites vary from the dynamite types already discussed
by containing an ingredient. which reduces the freezing point of the nitro-
glycerin. This ingredient replaces a portion of the nitroglycerin which would
be used in an equal grade of ordinary straight dynamite, ammonia dynamite,
or gelatin dynamite. While straight nitroglycerin dynamite may freeze at
temperatures as high as 8° C. (46° F.), some of the low-freezing dynamites
freeze only at temperatures considerably below 0° C. Many of this type,
'however, cannot be relied upon to resist freezing at temperatures below the
freezing point of water.
1670 ANALYSIS OF EXPLOSIVES
The additions made to nitroglycerin for this purpose. include the nitrotol-
uenes, nitroxylenes, nitrohydrins, nitrosugar, nitropolygl(Ycerin (tetranitro-
diglycerin), and nitroglycol. Any of these substances pres~nt will be found in
the ether extract together with, and in most cases dis-
solved in, the nitroglycerin. \
Moisture.-The determination of moisture may
be carried out as already described for "straight"
nitroglycerin dynamite (p. 1663), except when there
are present such substances as nitro glycol or mono-
nitrotoluene, whose volatility would prevent an accu-'
rate determination. An apprec.iable amount of
nitroglycol, for examplE,l, would be lost in an attempt
to desiccate to even approximately constant weight
over sulfuric acid, even without vacuum. The loss
of weight on desiccation should therefore be checked
against the difference between the total loss on ex-
traction with ether and the direct weight of the
ether extract after evaporation of the ether in a
bell-jar evaporator (p. 1664). .
The following method for· the determination of
moisture especially in dynamites containing nitro-
glycol was developed by the Explosives Chemical
Laboratory of the U. S. Bureau of Mines: 50 grams
of the dynamite is weighed into a 500-ml. round
bottom flask to which is attached a special moisture
collecting tube. To the top of this tube is attached
a small surface condenser. 150 to 200 m!. of carbon
tetrachloride (free from H 2 0) is added through the .i-
moisture collecting tube and the flask heated to
boiling on a water bath for 1 hour. The CC1 4 and
water volatilized are condensed: and fall back into
the measuring tube, the water separating on top of
the CCI 4, while the excess of the latter runs back
into the flask. The measuring tube is graduated
in 0.1 m!. and the moisture is read direct. The
tube should be standardized before use. The
assembled apparatus is shown in Fig:·237.~_
Nitrotoluenes.-Trinitrotoluene is not ~'readily
soluble in nitroglycerin and separates as crystals on
evaporation of the ether from the ether extract, ena-
FIG. 237. Moisture De- bling it to be qualitatively separated and identified.
termination Apparatus for
Dynamites, Oils, Fats, It may be determined by difference, the nitroglycerin
Waxes, Petroleum. being determined by means o£ the nitrometer. Any
dinitrotoluene present may also be dete~mined in
this manner toge_ther with the trinitrotoluene, but if mononitrotoluene is also
6 A very similar method in which a solvent lighter than water is used, preferably
benzene, is recommended for the determination of moisture in-smokeless powders con-
taining nitroglycerin, by F. Becker and P. Dittmar, Z. ges,.JSCliiess_Sprengsto1!wesen.
29, 327, 371 (1934). . -
ANALYSIS OF EXPLOSIVES 1671
" Permissible" explosives are coal mining explosives which have passed
the prescribed tests of the Bureau of Mines and are recommended by the
Bureau for use in gassy and dusty mines. Their important characteristic is a
relatively low flame temperature, which is brought about by modifying the
composition of the usual types of dynamites and other blasting explosives.
The general methods of reducing the flame temperature of explosives 10 are
summarized as follows:
(a) Addition of an excess of carbon-forming less CO 2 and more CO in the
gases of explosion.
(b) Addition of free water or of solids with water ('f crystallization.
(c) Addition of inert materials.
(d) Addition of volatile salts.
10 The thermochemical considerations involved are discussed in Bureau of Mines
Bulletin No. 15, "Investigations ,of Explosives. Used in Coal Mines," 1912, and the
details of analysis in Bureau of Mines Bulletin No. 96, "The Analysis of Permissible
Explosives," C. G. Storm, 1916.
1674 ANALYSIS OF EXPLOSIVES
almost to boiling, and filtering. This reagent is added to the solution con-
taining the gum arabic until no further precipitation pccurs; the mixture is
allowed to stand for several hours, then filtered, washed with absolute alcohol,
dried at 100° O. and weighed. The weight of precipl~ate multiplied by the.
factor 0.4971 (determined experimentally) gives the weight of gum arabic
found. Ohlorides or sulfates, if present, interfere with the determination and.
must be first removed. I .
to suspect that some such material as kieselguhr or clay is present, or that the
extractions with water or acid were not complete.
NITROSTARCH EXPLOSIVES
oration, the residue of ether-soluble materials dried, weighed, and its com-
ponents determined by the methods used for dynamites!.
Extraction with Water: Nitrates, Gums, etc.-The dried and weighed
residue insoluble in petroleum ether is extracted with didtilled water to remove
the nitrates or other water-soluble materials. The in~oluble residue left in
the crucible is dried at a temperature of 80° C. (100° m:a,y cause some decom-
position of the nitrostarch) for several hours, and weighed, the loss of weight
being total water extract and serving as a check against the sum of the com-
ponents separately determined.
Ammonium nitrate, if present, is determined in the water solution by the.
usual method of distillation of the NH3 after adding an, excess of alkali. So-
dium nitrate is determined by evaporating the water extract, volatilizing
ammonium salts, and weighing the residue after re-oxidizing with nitric acid
if charring has indicated the presence of any water-soluble organic materilil
(page 1667).
If the original explosive is granular in form, the presence of a binding or
agglutinating material in the water extract may be suspected. Although
numerous other substances may be used for the purpose, gum arabic is fre-
quently employed as a binding agent in different types of dry explosives. A
qualitative test for gum arabic has been mentioned on page 1675, and its
quantitative determination may be conducted as described on page 1677, by
precipitation with basic lead acetate solution.
Insoluble Residue: Nitrostarch, Charcoal, Cereal Products, etc. Starch.-
A microscopic examination of the weighed insoluble residue will usually serve
to identify its components. Any un-nitrated starch or cereal product is
readily distinguished, from nitrostarch by treating with a drop of KI solution
of iodine and examining under the microscope, when the un-nitrated starch
granules will appear "blue or black and the nitrated starch colorless or yellow.
Charcoal is identified by its color.
Un-nitrated starch, if present in an amount greater than a trace, is deter-
mined by boiling with dilute H 2S0 4 or HCl until iddine solution no longer
colors a drop of the liquid blue, then filtering and w~shing thoroughly. The
residue is dried at 80° C. and weighed, the loss of weight representing starch.
Nitrostarch.-Another portion of the insolu~le residue (the analysis being.
conducted in duplicate) is extracted with acetone in. a Wiley extractor or
other continuous extracti9n apparatus, or by transferring .the residue from the
crucible to a small beaker, digesting in acetone with stirring,. ?-nd filtering
through the same crucible, washing with fresh acetone. This extraction dis-
solves all of the nitrostarch, leaving any charcoal or cereal products that may
be present. The residue is dried at 100° C. al!d weighed, the loss of weight
representing nitrostarch.
The nitrogen content of the nitrostarch may .be determined, if desired,
by precipitating the nitrostarch from the clear acetone solution by the addition
of water and evaporation on a steam bath. A portion of the white, floury
precipitate· is then dried at 70°-80° C., weighed, and its nitrogen content de-
termined in the nitrometer (see page 650, Vol. I). ,
Charcoal.-If charcoal is present its weight may be taken direct, in the
absence of cereal products or other substances._' When the residue contains
cereal products, the material left after hydrolysis of the starch and extraction
ANALYSIS OF EXPLOSIVES 1681
of the nitrostarch will contain the crude fibre of -the cereal together with char-
coal or other insoluble ingredients_ A separation of such components is usually
impracticable.
Trinitrotoluene (TNT)
.
Trinitrotoluene, commonly designated in this country by the abbreviation
TNT, is also known in this and other countries by such names as triton,
trotyl, tolite, trilite, trinol, tritolo, etc. The term trinitrotoluol, which is
probably more commonly used than trinitrotoluene, is incorrect according to
approved chemical nomenclature.
This explosive is of the greatest importance as a high explosive for military
use, being adaptable as a bursting charge for high explosive shell, trench
mortar shell, drop-bombs, grenades, etc., because of its powerful explOSIve
properties, relative- safety in manufacture, handling, etc., its stability, its lack
of hygroscopicity, and absence of any tendency to form sensitive compounds
with metals.
It is classified by the Ordnance Department, U. S. A., into two grades,
according to purity-Grade I with solidification point of at least 80.2° C. and
Grade II with solidification point of at least 76.0° C. Other requirements, the
same for both grades, are as follows: acidity (as H 2 S0 4), not more than 0.01 %;
insoluble in benzene or 95% alcohol, not more than 0.10%; moisture, not more
than 0.10%. Grade, I represents material purified by means of sodium sulfite.
Solidification Point.-The determination of the solidification point or
" setting point" of TNT is the best single test for purity of this compound,
and is preferably carried out as follows:
A sample of about 50 grams of TNT is placed in a I/1X6/1 test tube and
melted by placing the tube in an oven at about 95° C. The tube is then
inserted through a large cork stopper into a larger test tube about lY2/1X7",
which, in turn, is lowered into a wide-mouth liter bottle, so that the rim of the
large tube rests on the neck of the bottle. The inner test tube is provided
with a cork stopper containing 3 openings-one for a standard thermometer
graduated in 1/10° C., one for a short thermometer which is passed just through
the stopper and is used for noting the average temperature of the exposed
mercury column of the standard thermometer, and the third opening being
a small v-shaped notch at the side of the'stopper, through which passes a wire
whose lower end is bent in a loop at right angles to the axis of the tube and
which is used for stirring the molten sample of TNT.
The standard thermometer is so adjusted that its bulb is in the center of
the molten mass, and the stirrer is operated vigorously, the thermometer being
watched carefully as the temperature falls. The temperature will finally
remain constant for an appreciable time and then rise slightly, owing to the
heat of crystallization of the TNT. As this point is reached, readings should
be taken about every 15 seconds until the maximum temperature of the rise
is reached. This temperature will usually remain constant for several minutes
while crystallization is proceeding. The maximum reading, corrected for the
emergent stem of the thermometer, is taken as the solidification point of the
sample.
Ash.-About 5 grams of TNT is moistened with sulfuric acid and burned
in a tared crucible. The residue is again moistened with a few drops of nitric
acid and sulfuric acid and again ignited and the resulting ash weighed.
1682 AN AL YSIS OF EXPLOSIVES
Picric Acid
Ordnance Department, U. S. A., specifications for picric acid prescribe that
it shall have a solidification point of not less than 120° C.; that it shall contain
not more than the following amounts of impurities: .
Moisture-().2% for dry material. .. / ............ " 22.0% for wet
Sulfuric acid (free and combined) . . . . . . . . . . . . . . . .. 0.10%
Ash ........................................... ' 0.2%
Insoluble in water ............................. " 0.2%
Soluble lead... . ................................ 0.0004%
Nitric acid (free) .............................. " none
crucible, washed well with hot water, and the insoluble residue on the filter
dried at 100°, cooled, and weighed.
Soluble Lead.-The presence of soluble lead in picric acid is highly objec-
tionable, because lead picrate is an extremely sensitive explosive and its pres-
ence would greatly increase the dangers involved in handling and loading
picric acid. A weighed sample of about 300 g. is digested in a 2-liter flask
with 100 mI. of a hot saturated solution of barium hydroxide in 65% alcohol.
• 1400 mI. of 95% alcohol is then added and the digestion continued at a tem-
perature below the boiling point (with reflux condenser), until everything
except traces of insoluble matter is in solution. The picric acid is then allowed
to crystallize on cooling, and the solution filtered off, decanting the clear liquid'
from the crystals until 500 mI. of filtrate is obtained. This 500 mI., repre-
senting 100 g. of picric acid, is treated with 5 drops HNO a and 10 mI. of 0.1 %
HgCb solution, and H 2S passed through it for 15 minutes. Allow the pre-
cipitate to settle for 20 minutes, filter and wash with alcohol saturated with
H 2S. Dry and ignite the precipitate, then dissolve the residue in 9 mI. of
HNO a (sp.gr. 1.42) by warming, add warm water to bring the volume to 50
mI., and electrolize at 0.4 ampere and 2.5 volts, temperature 65° C., for 1 hour.
Wash the electrode by replacing the beaker with another one containing dis-
tilled water without interrupting the current.. Dry and weigh the previously
tared anode. The weight of lead peroxide foundXO.8661 gives the percentage
of soluble lead found.
Nitric Acid.-No coloration should result when a water solution of picric
acid is treated with a solution of diphenylamine in sulfuric acid.
Ammonium Picrate
Ammonium picrate, also known in this country as "Explosive D," is of
importance as a military explosive more on account of its insensitiveness to
shock and friction, than because of its explosive strength, which is less than
that of TNT. Its chief use is as a bursting charge in armor-piercing projectiles.
Military specifications require it to be prepared from picric acid of standard
purity, to contain not less than 5.64% ammoniacal nitrogen, and not more
than the following amounts of impurities: . '
Moisture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 0.20%
Insoluble material. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 0.20%
Ash .................................................. 0.20%
Free a~~oni~ (NHa) ................................... 0.025%
Free PICflC aCId ...... " ................................ 0.025%
Moisture.-A ;ample of about 10 grams spread on a tared watch glass is
dried at 95° C. to constant weight (about 2 hours).
Insoluble Material.-A 10-gram sample is boiled with 150 ml. of water for
10 minutes, filtered on a Gooch crucible, the residue washed with hot water,
dried at 100° C. and weighed.
Ash.-A sample of about 1 gram is saturated with melted paraffin and
burned in a tared crucible, the residue ignited to burn off all carbon, and the
ash weighed.
'Free Ammonia or Free Picric Acid.-A sample of about 5 grams is ground in a
porcelain mortar with successive 50 mI. portions of water until the sample is
entirely dissolved, each portion of the water being poured through a filter into
1684 ANALYSIS OF EXPLOSIVES J
Mercury Fulminate
In commercial blasting caps and electric detonators mercury fulminate may
be found intimately mixed with potassium chlorate. It is, however, used
without admixture in certain types of detonators, in the fuzes of high explosive
shell and for other military purposes. It is usually purchased under specifica-
tions which provide that it shall be at least 98% pure, shall be free from acid,
and contain not more than 2% insoluble matter, 1 % free mercury, and 0.05%
chlorine in the form of chlorides.
Preparation of Sample.-Mercury fulminate being packed and handled in a
thoroughly wet condition until dried just before use, it is generally necessary
to dry the sample before testing. This may be done by exposing in a lo,v
temperature oven at not more than 50° C. until practically dry, then in a
desiccator (not a vacuum desiccator)' over sulfuric acid or calcium chloride until
its weight is constant.
Mercury Fulminate Content.-Exactly 0.3 g. is weighed into a wide-
mouthed Erlenmeyer flask of about 250 ml. capaqity, and 30 ml. of a 20%
solution of purest sodium thiosulfate is added quickly and the mixture shaken
for exactly 1 minute. At once titrate with N/lO hydrochloric acid using 3 drops
of methyl red indicator, the titration to be commenced 1 minute after adding
the sodium thiosulfate, and to occupy not more than 1 minute additional time.
The percentage of mercury fulminate is calculated from the volume of
standard acid required, after deducting the volume of acid required for, a blank
determination. Four molecules of HCI are equivalent to 1 mol. of m~rcury
fulminate, or 1 ml. N/I0 HCI equals 0.00711565 g. mercury fulminate. The
< reaction is assumed to be as follows:
HgC2N202+2Na2S203+2H20=HgS406+4NaOH+C2N2.
Acidity.-A 10-g. sample is extracted with 2 successive 25-ml. portions of
boiled distilled water in a Gooch crucible, and 3 drops of methyl orange solution
(1 g. per liter) added to the solution. No red tinge of color should be obtained.
Insoluble Matter.-A 2-g. sample is dissolved in 20% Na 2S203 solution,
filtered through a tare.d Gooch crucible and any insoluble washed several times
with water and dried to constant weight at 60°-70° C.
Free Mercury.-The residue of insoluble matter obtained as described
above is treated with a solution of 3 g. KI and 6 g. Na 2S20a in 50 ml. H 20 by
passing the solution through the Gooch crucible. Any organic mercury com-
pounds are thus converted into mercuric iodide, which is soluble in Na2S20a
solution. The metallic mercury remains behind on the filter, and is washed
with H 20, dried 1 hour at 80°-90° C., and weighed.
Clilorides.-A 5-g. sample of fulminate is extracted in a Gooch crucible
with 2 successive 25-ml. portions of distilled water at 90°-100° C. Three drops
of concentrated HNO a and 10 drops of 10% AgN0 3 solution are added to the
filtrate. If a turbidity results, the AgCI should be determined gravimetrically
or a fresh sample extracted and the filtrate titrated with a standard AgNO a
solution.
Lead Azide
Lead azide, PbN 6, is a primary or detonating explosive. Although it has
been employed in several European countries for many years as an ingredient of
,
1686 ANALYSIS OF EXPLOSIVES I'
detonators and primers for both commercia,! and military purposes, its use in
the United States is comparatively recent. Its prin<iipal application is as a
priming charge in compound detonators, where it is superimposed on a main
charge of tetryl, TNT or other high explosive. N orrfally a crystalline com-
pound, it is extremely sensitive when in large crystals, and hence is prepared for
practical use in the microcrystalline state, or in the form of small rounded
aggregates. Its natural white color may be slightly darkened by exposure to
light. It should be free from acidity. Its lead content may be regarded as an
indication of its purity, provided no other lead compounds are present. Pur~
PbN 6 contains 71.15% Pb. The sample taken for analysis should be dried a~
about 95° C.
Acidity.-A 10-gram sample is extracted in a Gooch crucible with 100 ml.
of previously boiled, cold distilled water in 20 ml. portions. The filtrate is
tested for acidity with methyl orange indicator.
Lead Content.-A weighed sample (approximately 1 g.) is dissolved in 50
ml. of saturated solution of ammonium acetate by warming in a 400-ml. beaker.
The solution is diluted with 200 ml. distilled water, heated to boiling, and 10 ml.
of a 10% solution of K 2 Cr 2 07 added with rapid agitation. After digesting on
the steam bath for 1 hour, the precipitate is collected in a Gooch crucible.
washed with hot water, dried at 100° C .•and weighed. PbCr04XO.64109=Pb.
capsule pressed on top of the charge, must be opened in the manner described
for electric detonators, in order to remove the inner capsule. Detonators of
this type usually contain a main charge of some nitro compound superimposed
\)y a layer of mercury fulminate, mixture of fulminate and chlorate, or lead
azide. Although a clean mechanical separation of the two layers is usually not
possible, portions can be taken from each and identified by qualitative tests
before proceeding with a quantitative examination.
Moisture.-The moisture content of the composition is determined by
desiccating to constant wlJight over sulfuric acid or calcium chloride.
Analysis of Composition Containing Mercury Fulminate and Potassium
Chlorate.-About 2-3 grams of the well-mixed composition is weighed in a
Gooch crucible provided with asbestos mat or disc of filter paper or silk, and
first moistened with a few drops of alcohol, then extracted with 200-250 m!. of
cold water in 15-20 m!. portions, using slight suction after each portion has
remained in the crucible for a few minutes. The residue in the filter is dried
to constant weight at 60°-70° C. (2-3 hours), and weighed.
The water extract contains the potassium chlorate and a portion of the
mercury fulminate, which is slightly soluble in cold water. It is treated with
2 m!. of ammonium hydroxide and H 2S passed to completely precipitate the
dissolved mercury fulminate as HgS. This black precipitate is filtered off,
washed, dried and weighed. Its weight X 1.22 gives the amount of mercury
fulminate dissolved by the water. This weight added to the weight of the
dried residue insoluble in water gives the total weight of mercury fulminate in
the sample. The KCI0 3 is found by subtracting the per cent of mercury ful-
minate+per cent moisture from 100%.
Ana)ysis of Compositions Containing Nitrocompounds.-Trinitrotoluene,
tetryl or picric acid can be identified by melting point test, TNT melting at
about 79°-80° C., tetryl at about 128° C., and picric acid at about 120°-122° C.
They may be extracted from the mixture by means of ethyl ether, in which
mercury fulminate is only very slightly soluble, and the determination of KCIO a
and mercury fulminate then made as described in the preceding paragraph.
If the main charge is an organic nitrate such as nitrated vegetable ivory,
nitrostarch, nitromannite etc., such material will be left with the mercury
fulminate in'the insoluble residue after extraction with water. The mercury
fulminate is then extracted by means of a hot 20% solution of sodium thiosul-
fate, leaving the organic nitrate in the Gooch crucible. These materials in
the detonating composition can be readily identified by microscopic examination.
1688 ANALYSIS OF EXPLOSIVES
In detonators where TNT or tetryl compose the mainj portion of the charge,
a small amount of lead azide, with or without mercury fulminate, may "Qe used
as a priming charge for the purpose of initiating the drtonation of the nitro,
compound. It should be identified in the top portion of the charge, next
to the reinforcing cap, and will in all probability be presertt if mercury fulminate
is not found. It is practically insoluble in water and -in! ether, and will be left
in the insoluble residue. If present, fulminate is destroyed by treating the
residue, in a flask, with 25 mI. of KOH solution. This converts the lead azide
to potassium azide, KN a• A slight excess of H 2S0 4 ill added and the mixture
distilled, the distillll,te, containing HN g, being collected in water. Enough
NaOH is added to the distillate to give an alkaline reaction with litmus, then
a little Pb(NO g)2, wh.en lead azide, PbN 6 will be regenerated as a white precipi:-
tate, which may be filtered off, washed with water, then with alcohol, dried
in the air, and tested by striking a small portion with a hammer.
Primers
Variations in Composition.-Many varieties of composition are used in
primers for small a:t'ms ammunition, and for other military purposes. The
composition must bE:) ignited by the impact of the firing pin, and must give It
flame of sufficient intensity and duration to 'ensure proper ignition of the pro,
pellant or of the detonator, depending on/ the purpose for which the primer is
employed. As primers are used with various kinds and granulation of explo,
sives, a priming composition suitable for one purpose is unsuited for another;
hence there are many types of priming compositions. A few of the types which
were in common usE:) up to comparatively recent years are shown in Table 1
below. Table 2 giV(lS a few examples of the great variety of patented type~ ot
the more modern "non-corrosive" primer compositions, in which potassiuln
chlorate is eliminated because\ the chloride formed as one of the products of
combustion was found to be responsible for corrosion of the gun barrels. "N on,
corrosive" primers have displaced many of the older ,types.
, I
'l~YPES
OF PRIMER COMPOSITIONS
APPlioXIMATE COMPOSITION (PER CENT)
TABLE 1
,
Ingredien ts
_-_-
Mercury fulminate ........ 31 25
- - _-
No, 1 No.2 No.3 No.4 No.5 No.6 No:T 'No~8 No.9 No.l(j
11 - -
--- - -
-
_-
-
_-
-
---
-
--.
28
Potassium chlorate ........ 38 38 53 60 50 51 53 53 47 14
Sulfur ................... - - - 7 , 3 9 - - 22 -
Powdered glass ........... - - - - - 12 - - - 35
Lead sulfocyanate ........ - - - - - .'
- 25 25 - -
Copper sulfocyanate ..... , - - - 3 -' - - - - -
Barium nitrate ........... - 6 - - - - - - - -
TNT ................... - - - - - - 5 - - -
-,
Tetryl .....' ............. - - - - - - - 3 -
Antimony sulfide ......... 31 31 36 30 44 26 17 17- 31 21
Lead oxide (PhO) ........ , - - - - - - L 2 - -
Shellac .... " ............ - - - - ::;;-- 2 - - - 2
Black powder (meal) ...... - - - - 3- - - - - -
ANALYSIS OF EXPLOSIVES 1689
TABLE 2
Nitrocellulose'
Genera1.-The term nitrocellulose, or more correctly cellulose nitrate,
applies to any nitration product of cellulose, ranging from products containing
in the neighborhood of 10-11 % N, which are used in the preparation of lacquers
and other commercial products, to military guncotton ~ith over 13% N. All
of these products are undoubtedly mixtures of the various nitrates of cellulose,
as indicated by the fact that there is always .some material with low nitrogen
content, soluble in ether-alcohol, in high nitrogen' guncotton, and some insoluble
material in the lower nitrated commercial products. It can usually be shown
without great difficulty that any nitrated cottori is a mixture of various nitrates
of cellulose.
The products of military importance are the insoluble guncotton of high
N-content, and the so-called" pyro " or pyrocellulose, soluble in ether-alcohol
and of about 12.60% N-content. In testing these products, the characteristics
of most importance are content of nitrogen, solubility in ether-alcohol, and
stability. Other determinations generally made are solubility in acetone and
ash. ,
Preparation of Sample.-If the sample contains a large excess of water, it
is enclosed in a clean cloth and the excess water removed by means of a press
or wringer. The pressed sample is then rubbed up in the cloth (not with the
bare hand) until lumps are removed, then spread on clean paper trays in an
air bath at .about 35°-40° C. until" air-dry." !.
Samples for stability tests and nitrogen deter;mination are treated as noted
below, the air-dry sample being suitable for determining solubility and ash.
Nitrogen.-About 1 to 1.05 g. of the air-dry sample is roughly weighed in a
tared weighing bottle, dried at 98°-102° C. for 1% hQurs, cooled in a desiccator
and accurately weighed. It is then transferred to the generating bulb of a
nitrometer (Du Pont modification; see p. 650) using a total'Of ~O ml. of 94-95%
c.p. H 2S0 4, The sample must be dissolved in the acid either in the weighing
bottle or in the cup of the generator, before it is drawn into the generating
bulb, and both the weighing bottle and the cup of the generator must be
thoroughly washed out with the 20 ml. of H 2 S0 4, so that none of the sample is
losJ? The determination in the nitrometer is completed in the usual manlier
(p. 650), the result being expressed as per cent N in'the dried sample of nitro-
cellulose. '
Solubility in Ether-Alcohol.-(a) Guncotton: The amount of ether-alcohol
soluble material in guncotton being usually not more than 10-12%, the deter-
mination may be made by evaporating a clear solution. Tw,o grams of air-dry
sample is placed in a clean dry cork-stoppered .J50~ml. cylinder, 67 .ml. of
95% ethyl alcohol added to thoroughly wet tlie guncotton, then 133 ml. of
ANALYSIS OF EXPLOSIVES 1691
ethyl ether (U.S.P. grade, 96%), added arid the mixture well shaken. If the
mixture of 2 parts ether and 1 part alcohol be added at once to the sample, a
gummy mass may result which dissolves with great difficulty, especially if the
solubility is unusually high. .
The cylinder is now allowed to stand at a constant temperature of usually
15.5° C. The solubility of nitrocellulose increases as the temperature is
decreased, hence a constant temperature of digestion is important. During the
digestion, which requires at least 1 hour, the cylinder must be thoroughly
shaken at 5-minute intervals. The cylinder is now allowed to stand for at
least 4 hours, until the insoluble portion of the sample has completely settled
and the supernatant liquid is perfectly clear.
Fifty m!. of the clear solution is now drawn off with a pipette, care being
taken not to disturb the settled pulp, and evaporated in a weighed evaporating
dish on a steam bath, avoiding loss from violent boiling of the ether. When
25-30% of the solution has been evaporated, 10 m!. of distilled water is added
slowly and the evaporation continued to dryness. The effect of the water is to
leav.e the residue in a white, brittle or powdery condition, rather than a tough
film which would lose its solvent with difficulty.
The dish is finally placed in an oven at 95-100° C. for Y2 hour, cooled in a
desiccator, and weighed. The weight of the residue, corrected for the residue
in the 50 m!. of ether-alcohol and 10 m!. H 2 0 used, represents the soluble nitro-
cellulose in 0.5 g. of the guncotton.
(b) Pyrocellulose: The solubility of pyrocellulose may be determined in the
manner described for guncotton, but owing to the much larger amount of
soluble material present, the evaporation of the residue to constant weight
without decomposition involves considerable difficulty. Sufficient water must
be added to precipitate the soluble nitrocellulose from solution in a stringy or
fibrous condition.
The detetmination is usually conducted by either the volumetric method or
the gravimetric method.
In the volumetric method, one gram of the air-dry sample is covered with
75 m!. of 95% ethyl alcohol and allowed to stand for about 2 hours with frequent
stirring, 150 m!. of ethyl ether is then added with stirring and the agitation
continued until solution is complete. The solution is now allowed to stand at
least 4 hours with frequent stirring, during at least 1 hour of which time it is
to be kept at a temperature of 15.5° C. It is then transferred to a " solubility
tube" and allowed to stand for about 16 hours, in order that the insoluble
material may settle completely. The solubility tubes are heavy glass tubes
about 21 inches long and 1.35 inches inside diameter, tapering gradually near
the bottom to a constricted portion about 2 inches long and about 0.3 inch
inside diameter. This narrow bottom portion is graduated in one-tenth ml.
The volume of the settled insoluble material is read in ml. from the scale on
the tube. If the reading is 0.25 ml. or less, the percentage of insoluble is
considered to be 1 % or less. If an exact determination is desired the gravi-
metric method must be used.
In the gravimetric method the solution is prepared and settled in a solu-
bility tube as described above and the clear liquid removed as completely as
possible by means of a narrow siphon tube of glass. Fresh alcohol and ether
are then added as before, the tube shaken and allowed to stand again until the
1692 ANALYSIS OF EXPLOSIVES
insoluble has settled, when the process may be repeated several times, depend~
ing on the amount of insoluble material present. After the last decantation,
the residue is washed from the tube to a beaker, using as small a quantity of
ether-alcohol as possible, and the mixture filtered th~ough a weighed G'C'Joch
crucible with ignited asbestos mat, the insoluble washed, free from soluble nitro-
cellulose with ether-alcohol, dried first at a low temperature and finally at
100° C. to constant weight, cooled and weighed. The crucible is then carefully
ignited over a free flame, cooled and weighed again. The loss of weight on
ignition is considered as insoluble nitrocellulose.
Solubility in Acetone.-A I-gram sample of air-dry pyrocellulose is treated
with about 200 m!. of acetone with frequent stirring until all gelatinous matter
has dissolved. The solution is transferred to a solubility tube (described above),
well shaken and allowed to settle for about 16 hours. If the volumetric reading
is 0.2 m!. or less, the percentage of insoluble is considered to be 0.4% or less.
A gravimetric determination may be made as described in the preceding
paragraph.
Ash.-One gram of air-dry sample is weighed in a tared crucible, moistened
with 10-15 drops of concentrated nitric acid; and digested for 2-3 hours on a
steam bath until converted to a gummy mass. The crucible is theri heated
carefully over a Bunsen. burner until the mass is completely charred, then at a
red heat until its weight is constant. The residue is the ash of the sample.
Stability Test: Heat Test with Potassium Iodide StarchPaper.-The " heat
test" or KI test, as it is commonly dffsignated, is the test most commonly
employed for determining the stability or degree of purification of nitrocellulose,
whether guncotton or pyrocellulose. This test, also referred to as the Abel test,
depends on the action of oxides of nitrogen liberated by the nitrocellulose under
the ir\fluence of heat, the gases in contact with the KI-starch paper liberating
iodine which colors the starch.
The sample is dried with great care to avoid contamination, in a clean paper
tray, at 35° to 40° C., until its moisture is reduced to the amount which will
give the minimum heat test, usually 1.5 to 2%. The/proper amount of mois-
ture is determined as follows: During the progress of the drying, the sample
on the tray is "rubbed up " from time to time, using a piece of clean tissue
paper spread over the back of the hand. When the sample begins to adhere
to the paper, due to static electricity, a sample of 1.3 g. is weighed into a stand-
ard test tube. These tubes are 5Y2 inches long, not less than Y2 inch inside
diameter and not more than %" outside diameter, made ~6f glass about 3/64
inch (1.2 ·mlIi.) thick. As soon as the first sample is weighed, the tray is
replaced in the drying oven for 2-5 minutes, a second sample weighed, and this
process repeated until a series of 5 samples have been taken, the last sample
being completely dry. This serIes of samples, if properly taken, will cover the
range of moisture content giving the minimum heat test. If the sample in the
tray appears to have become too dry during the time the weighings are being
made, it may be placed in a moist atmosphere for not more than 2 hours; the
entire time of drying and making the test must not exceed 8 hours.
The tubes containing the samples are fitted with clean, fresh cork stoppers
through which pass a piece of glass rod into the enq pf which is fused a small
piece of platinum wire ,bent into a hook. The wire--iS-heated in a flame to clean
it, a piece of the standard KI starch test paper, l"X%", attached, taking care
ANALYSIS OF EXPLOSIVES 1693
that' neither wire nor paper are touched with the fingers, and the paper mois-
tened on its upper portion by touching it with a glass rod dipped in a solution
of equal volumes of pure glycerin and water. The stoppers are then inserted
in the tubes and the tubes placed in a constant temperature water bath, so
that they are immersed to a depth of 2.25 inches. The time of placing in the
bath and the time of the appearance of the first faint yellowish discoloration
of the test paper are noted. The minimum test given by the 5 samples is
taken as the result of the test. The discoloration appears at the lower edge
of the moist portion of the paper. The temperature of the heat test bath is
65.5° C. (150° F.) for pyrocellulose, and usually 76.5° C. (170° F~) for gun-
cotton. Pyro is usually required to stand a test of 35 minutes, and guncotton
10 minutes.
A standard test paper is absolutely essential, and is prepared as follows: 16
The paper used in preparing the test paper is Schleicher and SchUll's filter
paper 597. This is cut in strips about 6 by 24 inches, and after b~ing washed
by immersing each strip in distilled water for a short time is hung up to dry
overnight. The cords on which the paper is hung are clean and the room is
free from fumes. The washed and dried paper is dipped in a solution prepared
as follows:
The best quality Gf potassium iodide obtainable is recrystallized three times
from hot absolute alcohol, dried, and 1 gram dissolved in 8 ounces of distilled
water. Cornstarch is well washed by decantation with distilled water, dried
at a low temperature, 3 grams rubbed into a paste with a little cold water,
and poured into 8 ounces of boiling water in a 'flask. After being boiled gently
for 10 minutes, the starch solution is cooled and mixed with the potassium
iodide solution in a glass trough.
Each strip of filter paper is immersed in the above-mentioned mixture for
about 10 seconds and is then hung over a clean cord to dry. The dip=-,ing is
done in a dim light and the paper left overnight to dry in a perfectly dark room.
Every precaution is taken to insure freedom from contamination in preparing
the materials and from laboratory fumes that might cause decomposition.
When dry the paper is cut into pieces about % by 1 inch and is preserved in the
dark in tight glass-stoppered bottles, the edges of the large strips being first
trimmed off about one fourth inch to remove portions that are sometimes
slightly discolored. When properly prepared the finished paper is perfectly
white, any discoloration indicating decomposition due to contamination.
Stability Test at 134.5° C.-In addition to the KI starch test, pyrocellulose
is usually required to stand a test at 134.5° C., made as follows:
The sample is completely dried at about 40° C., and 2.5 grams placed in each
of 2 heavy glass tubes, 290 mm. long, 18 mm. outside diameter and 15 mm.
inside diameter, closed with a cork stopper through which passes a hole 4 mm.
in diameter. A strip of litmus paper or standard normal methyl violet paper,
70 mm. long and 20 mm: wide, is placed in each tube, its lower edge 25 mm.
above the sample, which is pressed down to occupy a depth of 2 inches, the
walls of the tube being wiped clean with a roll of paper. The tubes are then
heated in a constant temperature bath at 134° to 135° C., all but about 6-7 mm.
of the tube being immersed in the bath. They are partially withdrawn for
16 Storm, C. G., Proc. 7th Inter. Congress App!. Chern., 1909; J. Ind. Eng. Chern., 1,
802 (1909).
1694 ANALYSIS OF EXPLOSIVES
examination of the test papers every 5 minutes after .the first 20 minutes of
heating, and replaced at once. The time required for reddening of the litmus
paper or for turning the methyl violet paper to a salmon pink color is noted
as the time of the test. A minimum test of 30 minutes is required with the
methyl violet paper, and heating is then continlied for altotal of 5 hours, during
which time there should be no explosion. I
The standard normal methyl ,violet paper is prepared as follows:
Preparation of Methyl Violet Test Paper.-A solution is prepared contain-
ing the following ingredients: pure rosaniline acetate prepared from 0.2500 g.
pararosaniline, .1680 g. methyl violet (crystal violet), 6 ml. c.p. glycerin, 30 mI.·
water, and sufficient pure 95% ethyl alcohol to m·ake. up to 100 ml. This
solution is placed in the angle of an inclined deep rectangular glass tray, and
large sheets of Schleicher & Schull filter paper (No. 597) cut in four strips are
dipped in it. In dipping, the strip is held by one end and dipped to within Y4"
of this end, withdrawing it slowly up the side of the tray so as to remove surplus
solutidn. The strip is then held horizontally and waved to and fro so as to
prevent the solution from running .and collecting in spots. As soon as the
alcohol has evaporated the strip is suspended yertically to dry, and when dry
is cut in strips 20 X 70 mm. These strips are bottled and kept for use in the
134.5° C. test.
Viscosity.-In the case of commercial nitrocelluloses it is frequently neces-
sary to determine the viscosity of standard solutions. The solutions used for .
this purpose are usually as follows: 17 '
% by Weight
, Formula
A B C
- -- -
Nitrocellulose (dried at 50° C.) .. , ...... , ............ , ... , .. 12.2 20.0 25,0
Ethyl acetate (85--88% grade). , ... , ...................... , 17.5 16.0 15.0
Denatured alcohol, C.D. No.5 (188-190 proof) ......... j ... 22.0 20.0 18.75
Toluol (2° C. range, incl. 110.7°) .... , , ........... 0'. • •• • ••• 48.3 44.0 41.25
I
SMOKELESS POWDER
Nitrocellulose Powders
At the present time the smokeless powder used by all nations is composed
of either colloided nitrocellulose or a mixture of colloided nitrocellulose and
nitroglycerin with or without the addition of other ingredients. All cannon
powder used in this country is of the nitrocellulose type, small-arms powders
being of both types. The. form and size of the grains are of great variety,
depending on the arm in which the propellant is to be employed.
Physical tests made in connection with the examination of smokeless powder
include the compression test, determinations of average measurements of the
grains, specific gravity, gravimetric density, number of grains per pound, and
calculation of burning surface per pound.
Chemical tests include determinations of moisture and volatile solvent,
diphenylamine used as stabilizer, ash, material insoluble in ether-alcohol and
in acetone, and sometimes nitrogen content.
Stability tests include the 134.5° C. test, and the" Surveillance test."
Moisture and Volatiles.-A sample of the powder weighing approximately
1 gram, in the form of thin shavings cut from at least 10 grains, or of whole
grains if the powder is too small to cut conveniently, is placed in a clean, dried
and weighed 250-m!. beaker, 50 m!. of 95% (by volume) alcohol, and 100 m!.
ethyl ether added and the beaker allowed to stand under a cover':jar with
occasional stirring, until the powder is completely dissolved. This usually
requires from 1 to 2 days. Wh;en all gelatinous particles of the powder have
dissolved, the beaker is heated on the steam bath to evaporate a part of the
ether, before precipitation of the nitrocellulose with water. The amount of
ether. to be evaporated is important, since it largely determines the character
of the nitrocellulose precipitate. The presence of too much ether causes a fine
sandy precipitate; too little causes a gummy, gelatinous precipitate. A fine,
flaky, or fibrous precipitate is desirable. The proper amount of evaporation
can be best determined by practice; usually the solution may be evaporated to
about 2/3 its original volume before precipitating. When the proper volume is
obtained, 50 ml. of water is added from a graduate, with continual stirring, in
5 mi. portions. If a thick gummy precipitate forms, add a little ether until it
becomes flaky; then add the remainder of the 50 m!. of water. The heating is
continued with stirring,'until most of the ether has evaporated, and the beaker
is then left on the bath until the precipitate is just dry. It is then placed in
the 100° C. oven for 1 hour, cooled in a desiccator, and weighed as rapidly as
possible. To facilitate weighing the weights .should be placed on the balance
pan before the beaker is removed from the desiccator, so that the exact weight
can be adjusted quickly. If more than 10 seconds are consumed in this weigh-
ing, the error caused by absorption of moisture from the air is an appreciable
one. In any event a check weighing should be made after an additional 30
minutes drying at 100° C.
Absorption of moisture during weighing may be avoided by the use of a
suitable aluminum beaker provided with a tightly fitting cover, instead of an
ordinary glass beaker.
1696 ANALYSIS OF EXPLOSIVES
of the precipitate are transferred frequently from the rod to another beaker of
hot water to prevent the formation of a colloided,mass.1 When 2 g. or more of
the precipitate has been collected and the acetone has peen volatilized by the
hot water, it is removed from the beaker and dried at 35°-40° C. About 1 g.
of this dry precipitate is placed in a tared weighing 'bottle, dried 1 hour at
100° C., weighed, and transferred to the cup of the n~trometer with sulfuric
acid. Part of the acid should be added to the precipitate in the weighing bottle
before transferring to the nitrometer in order to avoid loss of the dry precipitate
in handling. The determination of N is then completed as in the case of
nitrocellulose (page 1690). If the powder contains diphenylamine, a correction
is necessary for the amount of diphenylamine retained by the precipitated
nitrocellulose. This has been found to be an added correction of 0.15% N In
the case of powders containing the usual amount of 0.4% diphenylamine. This
correction compensates for. the nItrogen which becomes combined with t4e
diphenylamine, converting it to nitrodiphenylamines. ..
Instead of correcting for the effect of the diphenylamine, the latter may be
removed from the precipitated nitrocellulose, after air-drying and before final
drying at 100° C., by extraction with pure anhydrous ether. Results are quite
accurate if the determination is conducted with proper precaution.
Muraour 18 describes the following method for direct detel,"mination of
nitrate nitrogen in smokeless powders without preliminary treatment: The
powder sample is dissolyed in acetone and the solution treated with hydrogen
peroxide and sodium hydroxide with gentle warming. Sodium perborate is
then added and the mixture reduced with Devarda alloy in the presence of
additional sodium hydroxide, and finally distilled into a known volume of
standard sulfuric acid. Results are obtained which agree closely with the
theory. The method is applicable in the presence of easily nitrated materials'
such as diphenylamine, Centralite, etc., which interfere with the determination
of nitrate nitrogen by means of the nitrometer.
Typical Dynamite formulas-4D% grades (Bur. Mines, Bull. No. 80, p. 21)
II III IV V VI
---------------1--- ---------------
Nitroglycerin. .. . ...... . ... . . · . 25 15 10 10
TNT ............. .. . . . . .. . .. . . .... 5 5
Ammonium nitrate. ... .. . · . ., . 79 90 94 70
Sodium nitrate. ... . . . . \......... . 34 35
Sodium chloride. .... .. . .. . .. . · . . .. 9
Wood pulp. ........ .. . ........ . 15 12 10 10
Flour. ................................ . 25 17 5
Aluminum powder ....................... 3
Charcoal ............................... 3
Calcium carbonate. ....... .. . . . .. . .. 1 1 1
Zinc oxide. ....... ._ ..... . .. .. . .. . · . · . J
Magnesium sulfate, cryst .. .. ·. 15
FIXED OILS, FATS AND WAXES
PETROLEUM
It will be remembered that the fixed oils are those which leave a permanen~
stain on paper, whereas the essential or volatile oils evaporate. The fixed oils,
if saponifiable, are glycerides of the higher fatty acids, oleic, C 17 H aa COOH,
stearic, C 17H 3S COOH, palmitic, C1sHa1COOH; and if unsaponifiable, hydro-
carbons usually of the paraffin (C n H zn+2) and olefin (CnHzn) series. The fats
differ from ~he oils in having a higher melting-point, caused by a larger percent-
age of stearic and palmitic acid. The waxes differ from the oils or fats in that
the former are esters of monatomic alcohols. The oils are further divided
accotding to their drying power on exposure to the air, into drying, semi-drying
and non-drying oils.
The drying oils contain a large proportion of glycerides. of the unsaturated
acids, particularly linoleic and linolenic, whereas the semi-drying contain a
smaller percentage, and the non-drying little or none qf these esters.
The saponification test, unless mineral or rosin oil be suspected, need rarely
be resorted to ; the reason being that it would show praptically nothing regarding
the nature of the oil. Except in the case of castor (Sapn. No. 181), rape (174),
sperm (135) and cocoanut (260), this characteristic iSj about 193.
Finally, where the importance of the case will warrant, the analyst is advised
to prepare a mixture of oils using the proper proportions indicated by the various
tests, and subject it to the more rapid tests as the specific gravity, viscosity,
Maumene and iodine number. In making out the report of analysis it should-
be borne in mind that, excepting in the case of the special test, the results of
one test cannot be relied upon to determine the nature of an oil, but the evidence
of all the tests here given should be carefully compared and weighed before
rendering a final verdict: in consideration of the fact of the wide variation of the
characteristics of the oils, it is futile to report the quantities of oil found in a
mixture more closely than 1%.
PETROLEUM PRODUCTS
(a) Burning Oils
The tests or determinations to be made are, in the order of their importance,
flash, fire, specific gravity, distillation, sulfur, free acid, sulfuric acid, cioud
test, mineral salts and water. In some cases the color is determined as it is of
commercial importance. .
Flash Test or Point.-By flash point we understantl the lowest temperature
to which an oil must be heated, to give off vapors which when mixed with air
produce an explosive mixture. The results of th~s test will vary according to the
quantity of air over the surface of the oil, and whether this be moving or still;
also according to the distance of the testing flame from the surface of the oil.
Furthermore; the size of this testing flame, the'length of its time of action, its
form and dimensions, and'lastly, the manner of heating the oil, will all influence
the result. 2 • .
is made and noting the point as indicated above. ;The flame is extinguished
by a piece of asbestos board and the heating discon;tinul;ld. In the case of
many illuminating oils this point is from 10° to 20° F. higher than the flashpoint.
In the case of " Mineral Sperm" (300° F. fire test oil) these tests should
be made with the Cleveland open cup. The heating should be oat the rate of
10° F. per minute, and the testing flame first applied at 230° F. and then every
seven degrees until the flashing-point is reached. I
The most satisfactory way of making these tests is to place the watch upon
the desk and read the thermometer at the expiration of every minute, noting
down each reading in the proper column in the laboratory note-book.
Significance of the Tests.-The flash test rather than the fire test measures
the safety of an oil as the vapors producing the flash are evaporated in the fire
test. The writer has had samples of oil of 65° to 70° F. flash which gave a fire
test of 135°-140° F. These, judged by the flash test are unsafe, whereas they
would be passed by the fire test. . .
Specific Gravity: (a) By the Hydrometer.-A hydrometer jar is four-fifths
filled with the oil, a verified Baume hydrometer introduced into it, and the
depth read off to which the instrument sinks into the oil. This may be effected
by placing a strip of white paper back of the jar and noting the point at whiCh
the lower meniscus of the oil touches the scale. The temperature of the oil
is taken at the same time, and in case it be not 60° F. (15.5° C.), subtract 1°
Baume from the hydrometer reading, for every 10° F. it is higher than 60°, and
add 1° Baume for every 10° F. it is lower than 60° F. In practice this can be
done by Tagliabue's " Manual for Inspectors of Petroleum," which gives. the
readings at 60° F. for any gravity
from 10 to 100° Baume, between 20°
and 109° F. The specific gravity may
141.5
be found by the formula 131.5+Be'
Be representing the reading Baume at
15.5° C. /
(b) By tTie Westphal Balance.-
This is a specially constructed instru-
ment, Fig. 239, with a glass plummet
carrying a thermometer counterbal-
anced by a weight. Upon immersing
the plummet in a~·liquiQ. the positions
of the weights, which must be added
to restore the equilibrium, represent
the specific gravity directly. The
largest' weight represents the first
":i;;;?i~ill!!Iiiiiiii~~~.~ decimal place, the next the second,
- and so on.; The instrument is placed
FIG. 239.-Wesphal Balance. upon a level table, and by means of
° _ ° the leveling screw is brought into ad-
justment-i.e., so that the point upon the beam is exactly opposite the point
upon the fixed part. I
The plummet is now placed.in the vial or tJalance jar containing the oil,
cooled to 15.5° C., hung upon the balance, being careful completely to immerse
FIXED OILS, FATS AND WAXES. 1705
it in the oil, weights added to restore the equilibrium, and the specific gravity
read off as above described.
Care should be taken that the 'plummet does not touch the sides of the jar
or vial. For solid fats and some oils the specific gravity is taken at 100° C.,
using a special plummet.
Significance of the Test.-The Specific Gravity or " Gravity" as the oil
man says shows whether the sample is of paraffine or asphalt base, the latter
oils being the heavier.
Distillation Test:
A. S. T. M. METHOD D 86-35.L-KEROSENE
Tolerances
Centimeters Inches
Centimeters Inches
The position of the vapor tube shall be 9 cm.±0.3 cm. (3.55 in.±0.12 in.)
above' the surface of the liquid when the flask contains its charge of 100 m1.
The tube is approximate~y in the middle of the neck and set at an angle of 75°
(tolerance±3°) with the vertical. .
Condenser.-The condenser (Fig. 241) consists of a 196-in. (14.29 mm.)
OD No. 20 Stubbs Gage seamless brass tube, 22 in. (55.88 em.) long. It is set
at an angle of 75° from the perpendicular and is surrounded with a cooling bath
15 in. long (38.1 em.), approximately 4 in. (10.16 em.) wide by 6 in. (15.24 cm.)
high. The lower end of the condenser tube is cut off at an acute angle, and
4 Bureau of Minp.s Technical Paper 323B is an earlier description.
6 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S.·T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
G This method is issued under the fixed designation D 86; the final number indicates
the year of original adoption as standard or, in the case of revision, the yea: of last
revision. Issued as Tentative, 1921; Adopted in Amended Form, 1927; Revised, 1930,
1935. This method has been approved as AMERICAN STANDARD by the American
Standards Association.
1706 . FIXED OILS, FATS AND ~AXES
curved downward for a length of 3 in. (7.62 Qm.) and slightly backward so as to
insure contact with the wall of the graduate at a pointl1 to 1~ in. (2.54 to 3.18
cm.) below the top of the graduate when it is in position to receive the distillate.
Shield.-The shield (Fig. 241) is made of approxi,$ately No. 22 gage sheet
metal and is 19 in. (48.26 cm.) high, 11 in. (27.94 cm.) lbng and 8 in. (20.32 cm.)
wide, with a door on one narrow side, with two open,ings, 1 in. (2.54 cm.). in
diameter, equally spaced, in each of the
-,.- two narrow sides, and with a slot cut
!
in one side for the vapor tube. The
I 1,,"",
;_l;
centers of these four openings are 8%
in. (21.59 cm.) below the top of the
,"v
(b) Electric Heater.-The electric heater, which may be used in place of the
gas flame, shall be capable of bringing over the first drop within .the time speci-
fied below when started cold, and of continuing the distillation at the uniform
rate. The electric heater shall be fitted with an asbestos 'board T
top VB to 74: in.
7 When distilling petroleum products having an en~..point above 470 F. (243.34 C.),
0 0
the hole in the asbestos board shall be 1}1 in. (3.81 em.) in diameter.
FIXED OILS, FATS AND WAXES 1707
(3.18 to 6.35 mm.) thick, having a hole lU in. (3.18 em.) in diameter in the
center. When an electric heater is employed, the portion of the shield above
the asbestos boald shall be the same as with the gas burner but the part below
the board lpay be omitted ..
"BlofH,,'!
Popu
I
.
~
I
I
I
.~.
~~----~------------------~
4.33 in.).
DISTANCE TO 3000 C. OR 572° F. LINE FROM Top OF THERMOMETER: 30 to,45 nun.
(1.18 to 1.77 in.).
FILLING ABOVE MERC~E.Y: Nitrogen gaB.
Top FINISH: Glass ring. .
1708 FIXED OILS, FATS AND WAXES
GR.AD~ATION: AIl lines, figures, and letters clear cut and distinct. The first and each
succeeding 5° C. or 100 F. line to be longer than-the remaining lines. Graduations
to be numbered at each multiple of 10° C. or 20°F. ~ .
IMMERSION: Total.
SPECIAL MARKING: "A. S. T. M. Low Distillation," a: serial number and the manu-
facturer's name or trade mark shall be etched on the stbm. .
SCALE ERROR: The error at any point of the scale when the thermometer is stand-
ardized as provided below shall not exceed 0.5 C. or 1 F.
0 0
(e) The charged flask shall be placed in the l}:4'-in. (3.18-cm.) opening in the
6 by 6-in. (15.24 by 15.24-cm.) asbestos board with th~ vapor outlet tube in-
serted into the condenser tube. A tight connection may be made by means of
a cork through which the vapor tube passes. The positibn of the flask shall be
so adjusted that the vapor tube extends into the condenser1tube not less than 1 in.
(2.54 cm.) nor more than 2 in. (5.08 cm.). I
(f) The graduated cylinder used in measuring the charge shall be placed,
without drying, at the outlet of the condenser tube in such a position that the
condenser tube shall extend into the graduate at least 1 in. (2.54 em.) but not
below the 100-mI. mark. Unless the temperature is between 55° and 65° F.
(12.8° and 18.3° C.) the receiving graduate shall be immersed up to the 100-ml.
mark in a transparent bath maintained between these temperatures. The top
of the graduate shall be covered closely during the distillation with a piece of
blotting paper or its equivalent, cut so as to fit the condenser tube tightly.
When everything is in readiness, heat shall be applied at a uniform rate,
so regulated that the first drop of condensate falls from the condenser in not less
than 5 nor more than 10 minutes. The distillation thermometer shall be read
2 minutes after heat is applied and the indication recorded as the" correction
temperature." This figure is of significance only in cases when there is a ques-
tion as to the accuracy 9 of the initial boiling point, as subsequently determined.
When the first drop falls from the end of the condenser the reading of the dis-
tillation thermometer shall be recorded as tl].e initial boiling point. The receiv-
ing cylinder shall then be moved so that the end of the condenser tube shall
touch the side of the cylinder. The heat shall then be so regulatE!d that the
distillation will proceed at a uniform rate of not less than 4 nor more than 5 mI.
per minute', The volume of distillate collected in the cylinder shall be observed
and recorded, to the nearest 0.5 mI., wh'en the mercury of the thermometer
reaches each point that is a multiple of 10° C. or the Fahrenheit equivalent of
this point (30° C., 40° C., 50° C., 60° C., etc., or 86° F., 104° F., 122° F.,
9 The initial boiling point is one of the observations which it is customary to report
as indicative of the qu~lity of gasoline. It is a point Which' js very difficult to check,
especially when duplicate tests are made at different room temperatures, and it has not
been found practical to devise an accurate method of correcting for this unavoidable
variation in operating conditions. The practical significance of initial boiling point is
not regarded as sufficient to warrant requiring that tests be conducted under regulated
conditions of room temperature. The following directions are included as an expedient
intended to obviate some of the difficulties that develop on accoul)..t of the practice of
including initial boiling point limits in specifications for the purchase and'sale of gasoline:
If the "correction temperature" is below 70 F. (21 Co) or above 80 F. (27 C.),
0 0 0 0
the observed initial boiling point, if it is below 150 F. (66 C.), may be considered
0 0
incorrect by an amount not less than one third the difference between the "correction
temperature" and 75 0 F. (24 0 C.). In case there is a dispute between buyer and seller,
the observed initial boiling point figures obtained by all parties shall be revised accord-
ing to the following formula:
Revised initial boiling point
= {Observed initial} _ Correction Temperature (deg. Fahr.) "':'75
boiling point 3
If all of the revised figures fall. within the specification limit the gasoline in question
shall be considered as passing; otherwise arrangements shall; -if possible, be made to
conduct a 0 distillation test0 ~ith the room temperature maintaIned between the limits of
70 and 80 F. (21 and 27 C.), inclusive.
/'
•
FIXED OILS, FATS AND WAXES 1711
140° F., etc.). If preferred, the reading of the distillation thermometer mlp'
be observed and recorded when the'level of the distillate reaches each lO-ml.
mark on the graduate. In case a product is being tested to ascertain whether
or not it conforms with a given specification, all necessary observations shall
be made and recorded, whether or not they are included in the series ordinarily
employed by the laboratory making the test.
No adjustment of the heat shall be made after the liquid residue in the flask
is approximately 5 ml. unless the time required to bring over the last 5 ml. of
distillate and reach the end point exceeds 5 minutes. The end point is the
maximum temperature observed on the distillation thermometer and is usually
reached after the bottom of the flask has become dry. If the bottom of the
flask is not dry the operator shall record this fact.
In case the time required to bring over the last 5 ml. of distillate and reach
the end point exceeds 5 minutes the test shall be repeated and the heat shall
be adjusted when the liquid residue reaches 5 ml. This adjustment may be
either an increase or a flecrease but must accomplish the purpose of bringing
the'period required to vaporize the last 5 ml. of distillate and reach the end point
within the limits of 3 and 5 minutes.
The total volume of the distillate colle.cted in the receiving graduate shall
be recorded as the recovery.
The cooled residue shall be poured from the flask into a small cylinder
graduated in 0.1 mi., measured whea cool and the volume recorded as residue.
, The difference between 100 ml. and the sum of the recovery and the residue
shall be calculated and recorded as distillation loss.
Accuracy.-With proper care and attention to detail, duplicate results
obtained for initial boiling point and end point, respectively, should not differ
from each other by more than 6° F. (3.3° C.). Duplicate readings of the volume
of distillate colleded in the cylinder when each of the prescribed temperature
points is reached should not differ from each other by more than 2 m!. In case
observations are made on the basis of prescribed percentage points, the dif-
ferences in temperature readings should not exceed the amounts equivalent to
2 m!. of distillate at each point in question.
Correction for Barometric Pressure.-The actual barometric pressure shall
be ascertained and recorded, but no correction shall be made except in case of
dispute. In such cases the temperature points shall be corrected to 760 mm.
(29.92 in.),'by the use of the Sydney Young equation, as follows:
For Centigrade readings:
cc = 0.00012(760 - P) (273 +tc)
sulfur compounds be present, a yellowish layer is formed upon the metal. After
cooling add distilled water drop by drop until the metal is dissolved, and test
for sulfides with sodium nitroprusside.
for the blank. If at this point the solution.in, the absorber should have a pink
color, too much oil has been burned and the determination shall be repeated, but
burning for a shorter time. The'temporary pink color ,that sometimes appears
-in the chimney when rinsing it out should be disregar~ed. Titrate just as,in
the blank, continuing to add the HCl solution until, the color matches that
obtained in the blank. Read and record the volume of HCl solution used.
NOTE 1.-If it is not possible to obtain a good flame free from smoking, as some-
times happens if the oil contains considerable percentages of benzol, dilute the oil with
an equal weight of sulfur-free anhydrous alcohol before putting the sample into the
lamp. If, while maintaining the specified flame height the flame still smokes, a smaller
flame is permissible. In order to get the correct, value when calculating the percentage
of sulfur present in this case, allowance shall be made for the fact that the s"ample as
burned in the lamp is one-half alcohol. This shall be done by doubling the percentage
of sulfur obtained in computation by the formula given below.
NOTE 2.-For accurate work, conditions shall be such that a minimum or zero
blank is obtained. If the titration of the blank requires more than 0.1 ml. of the HCI
solution per gram of sample burned, which corresponds to 0.01 per cent of sulfur, a
high degree of accuracy cannot be expected in the case of oiL'l of low sulfur content.
Calculate the sulfur content of the oil by substituting the proper values in
the following formula: I
_ milliliters of HCl for blank-milliliters of RCI for sample
P ercentage 0 f su If ur -
grams 0 f 01'1 b urne d X 10
This formula is correct only for the standard solutions specified, pp. 1713,
1715, 1 ml. of each being equivalent to 0.001 g. of sulfur. The use.of solutions
of any other strength, such as 0.1 N, involves more complicated calculation and
is not advisable.
As a check on the determination, a sample of known sulfur content shall
be analyzed. Such,a reference sample may be made by adding a pure volatile
sulfur compound, such as carbon bisulfide, to sulfur-free anhydrous alcohol.
This check shall indicate from 85 to 100% of the sulfur present.
Accuracy.-Duplicate determinations by the sarrle operator should not
differ by more than the amounts given in the following tabulation:
Average Deviation
from Mean, i
Sulfur Found, per cent per cent Sulfur
Under. 0.125 ............................ '............... 0.0025
0.125 to 0.25 ..... ',' ..........................•... -" .•". 0.005
0.25 to 0.50. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 0.01
Over 0.50 .............................................. 0.02
For different operators corresponding deviations may be double those given
above.
The barium sulfate can be precipitated and determined for rapid work by
the Jackson Turbidimeter.
The percentage of sulfur in a kerosene should, not exceed 0.05; the Penn-
sylvania oils contain usually 0.02 to 0.03, the Lima 0.04 to 0.05. 13
Detection' of Acidity.-Shake equal quantities of oil and warm water in a
test-tube, pour off the oil, and test the water with litmus paper! If the water be
stron~ly acid, the quantity may be determined B:sjll-"-Free Acid," page 1782.
13 Kissling, Ch., Rev. Fett und Harz. Ind., 14, 157, 1906.
FIXED OILS, FATS AND WAXES 1717
The acid in this case is most probably sulfuric, coming f.rom the refining
process.
Sulfuric Acid Test.-The object of this test is to judge of the degree of
refinement of the oil, a perfectly refined oil giving little or no color when sub-
mitted to the process. One hundred grams of oil and .40 grams of sulfuric acid,
1.73 specific gravity, are shaken together for two minutes in a glass-stoppered
bottle and the color of the acid noticed. For comparative work this color is
matched by solutions of Bismarck brown. 14
Mineral Salts.-Salts of calcium or magnesium when dissolved in the oil
diminish its illuminating power; their action is to form a crust on the wick and
prevent access of air.
Redwood 15 states that 0.02 gram of either of these salts in 1000 grams of oil
diminishes the illuminating power 30 to 40% in eight hours.
They are determined by distilling 100 to 200 m!. of the oil down to about 20
ml., evaporating and igniting this residue, and subsequently treating with hydro-
chloric acid. The calcium and magnesium are then determined in the usual
way'.
Determination of Water.-By rubbing the oil together with a little eosin on
a glass plate the oil will take on a pink color if water be present.
The evaporation method is approximate and applicable only to heavy oils
and greases., Its accuracy even with heavy greases is questionable.
Dilute 16 the oil with an equal volume of benzol, whirl it vigorously in a
centrifuge until the separated layer of water does not appear to increase in
volume. However, as water is somewhat soluble in any diluent used and also
in oils, a portion of the water content will fail to appear, consequently the
,method in which a diluent is used cannot be considered accurate. It.is advis-
able first to agitate the diluent vigofDusly with water and then to separate
with the centrifuge in order to saturate it with water before using. See also
Dean's test under Fuel Oil.
Groschuff 17 states that 100 grams of benzene will dissolve 0.03 gram of
water at 3° C. and 0.337 gram of water at 77° C., whereas petroleum products
(density 0.792) will dissolve from 0.0012 gram at 2° C. to 0.097 gram at 94° C.
While water to the extent even of 3 or 4% is apparently without influence
on the viscosity, 1% extinguishes the flame when making the flash test.
Color.-This test has lost its importance, since oils are sometimes satis-
factory despite their yellow color. The determination is usually, made with
the Saybolt colorimeter,Is in which the depth of the oil is changed in a cylinder
until it matches the color of standard pl;:ttes of uranium glass.
The office of a lubricant is to prevent the attrition of axle and journal box
by interposing itself between them in a thin layer, upon wp.ich, the shaft revolves.
The ideal lubricant is that which has the greatest adhesion to surfaces and the
least cohesion among its own particles, or, as the practical man expresses it,
the most fluid oil that will do the work and stay, in plac~. The determination
of its viscosity or " body" is then of the first importande.
Viscosity is the degree of fluidity of an oil or its iriternal friction. It is
independent of the specific gravity of the oil, although this in the pipette
instruments influences the time of efflux. Within certain limits it may be
taken as a measure of the value of oil as a lubricant, by comparing the viscosity
of the oil under examination with that of other oils which have been found
to yield good results in practice.
The instruments employed for its determination may be divided into two
classes-pipette viscosimeters, giving the time of efflux, as those of Engler,
Saybolt, and others, and torsion viscosimeters, giving the retardation' due to,
the oil, those of Mac Michael and Doolittle.
In expressing viscosity, consequently, it is necessary to give the name of
the instrument with which it is determined. It is sometimes expressed. as
specific viscosity, that is, the time of the oil divided by the time of water; this
is only comparative when done with instruments of the same name, that is,
specific viscosity Engler is not the same figiIre as specific viscosity Saybolt.
Besides this manner of expressing viscosity, it is measured in absolute (C.G.S.)
units or dynes per cm. This unit of visco§ity has received the name of Poise
from Poiseuille, one of the first to express this property in absolute measure.
This is possible when the diameter of the orifice, its length, the quantity and
specific gravity of the oil, its time of efflux and change of head are known.
Where it is impracticable to determine
all these data by direct measurements,
the readings of a viscosimeter may be
changed into dynes by determining the
viscosity in secqnds of standard solu-
tions of glycerine, the viscosity of these
being determined in dynes from tables
of physical constants. Or it may be
done by use of the tables on pages 1726
and 1799.'
Engler Apparahls.·-~De8cription.
The apparatus (Fig. 243) 'consists of a'
flat, brass cylindrical vessel,A, 106 mm.
in diameter and about 62 mm. deep,
holding 240 mI.,' provided with a jet 2.9
mm. in diameter and 20 mm. long.
This vessel is gilt inside and the jet, in
_~-'-_ _ _ _-J._ _~' the standard instruments, is of platinum
FIG. 243.---=-Engler Viscosimeter. -ordinarily it is made of brass; the ves-
sel is surrounded with a bath, B, either
of water or oil, provided with a stirrer and heated by ~ ring b~tner. The jet is
clo!3ed by the woo,den valve, P, passing through the_cOVer, and a thermometer,
, c, ;;hows the temperature of the oil; three studs show-the height to which A is
FIXED OILS, FATS AND WAXES 1719
filled and at the same time when it is level. The oil ordinarily is discharged
into the 200-ml. flask, although in case the oil or time be limited, 100 or 50 m!.
may be used and the time of efflux multiplied by 'a suitable factor. The instru-
ment is standardized with water, 200 m!. of which at 20° C. should run out in
from 50 to 52 seconds.
Manipulation.-The instrument is thoroughly cleaned with alcohol and
ether if necessary ane! dried; any suspended matter is removed from the oil,
which is poured into it up to the level of the studs, stirred until 20° C. is reached
and the bath adjusted to the same temperature. The flask is placed beneath
the orifice, the plug raised and the time required for 200 mI. of oil to flow out is
noted; this is divided by the water value of the instrument and gives then rela-
tive or specific viscosity. If only 50 ml. are allowed to run out the time must
be multiplied by 5, and if 100 ml., by 2.35. If only 50 ml. were put in all.d 40 ml.
allowed to run out, multiply this time by 3.62 to obtain the time for 200 m!.;
if 66 ml. and 50 ml. run out, multiply by 2.79.1 9 If it be desired to express the
viscosity in absolute measure,(C.G.S. units) it can be done by reference to the
table en page 1726. It should be noted that specific viscosity obtained with
a different type of instrument, e.g., the Saybolt, is not the same as with the
Engler.
(a) Oil Tube.-The oil tube (illustrated in Fig. 244) shall be entirely of corro-
sion-resistant metal, with or without plating, and shall conform to the dimen-
sional requirements shown in Table I within the permissible variations pre-
scribed. The lower end of the oil tube shall be I?~ovided with a nut for
locking it in place in the bath and with a cork or other suitable device to prevent
flow until the test is started. For convenience a string or its equivalent nuty
be attached to the cork for rapid removal.
The oil tube shall be standardized, and any correctIon in excess of 0.2%
shall be applied. The time of flow shall be within ±1% (Note 1) of the time
as obtained with the National Bureau of Standards' master Saybolt oil tube
(Note 2).
NOTE I.-For routine testing, oil tubes having corrections as high as ±2% may
be employed, provided, however, that any new oil tubes obtained after the adoption of
this method shall not have corrections higher than ±1 %. For referee testing, oil tubes
having corrections higher than ±1 % shall not be employed. .
The oil may be standardized by anyone of the following methods:
(a) Determination of outflow time for American Petroleum Institute oil standard
and application of appropriate correction to obtain established value for the
oil standard.
(b) Calibration by National Bureau of Standards subsequent to January 1, 193i!.
I (c) Comparison with a tube certified by the ,National Bureau of Standards subse-
quent to January 1, 1935, provided the tube shows no evidence of damage
or injury since certification.
. NOTE 2.-This tube, conforming to the A",'S, T. M. dimensional specifications, has'
outflow times calibrated against oil standards under the maintenance of the American
Petroleum Institute Committee on Viscosity Standards.
TABLE I.-DIMENSIONS OF OIL TUBES
(b) Bath.-The bath shall serve as a support to hold the oil tube in a
vertical position and as a container for the bath liquid. The bath sha.ll be
equipped with a stirring device and with means for heating or cooling. The
source of heat or refrigeration shall be more than Vi in. (3 cm.) from the oil
tube. If an external heater is used it must be mo1e-than 2 in. (5 cm.) from the
oil tube.
FIXED OILS, FATS AND WAXES 1721
Overflow
Rim ....
fa he Drained
before Sfarfing Flow
{em.t O/em.
Irrsicle Diameter
af 6radualiorr Mark
£fevaf/on
of f"/osk.
(c) Receiver.-The receiving flask (see Fig. 245) shall be of glass with a
capacity up to the graduation mark on its neck of 60 ml.±0.05 ml. at 68° F.
(20 0 C.). At the graduation mark, the inside diameter of the neck of the flask
shall be 1 cm.±O.l cm. The cylindrical portion of the neck of the flask shall
1722 FIXED OILS, FATS AND WAXES
extend not less than 0:3 cm. above anq below the gr~duation mark. The
graduation mark shall be 5.8 cm.±1.0 cm. above the bottom of the flask.
The maximum outside diameter shall be less than 5.5 cm.
(d) Oil-Tube Thermomeiers.-The oil-tube therrlOmeters shall conform
to the following requirements. They shall include tWQ sets of six thermometers
each, one set being graduated in Fahrenheit degrees and the other in Centigrade
degrees, the ranges being chosen to include the temper~tures commonly used in
testing. To prevent contact of the thermometer with the orifice in the oil tube
a suitable support shall be attached to the enlargement of the thermometer
stem.
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGES AND SUBDIVISIONS:
Range 66 00 to 800 F., subdivisions in 0.20 F. for tests at 70 0 and 77 0 F.
" 94 to 1080 F., " " " " " " 1000 F.
" 1200 to 1340 F., " . , " " " " 122 0 and 130 0 F.
" 0 0
1340 to 1480 F., " " " " " " 14000 F.
" 1740 to 1880 F . , " " " " "180 F.
" 204 to 218 F., " " " " " " 210 0 F.
or
Range 19 0 to 27°.C., subdivisions in 0.10 O. for tests at 21.1 0 and 25 0 C.
" 340 to 42 0 C., " " . " ' " " , , 37.So C.
" 4900 to 5700 C., " "" " " 50 0 and 54.40 C.
57 0 to 65 0 C., " " " " " " 60 0 C.
" 79 to 87 C., " " / " " " , , 82.20 C.
95 0 to 103 0 C., " "" " " 98.9 0 C.
TOTAL LENGTH: 252 to 256 rom. (9.92 to 10.08 in.).
STEM: Plain front, enamel back, suitable thermometer tubing. Diameter, 6 to 7 mm.
(0:236 to 0.275 in.). The stem shall be made with an enlargement not less than 4
nor more than 7 mm. (0.16 to 0.28 in.) in length, having a diameter 2 to 3 mm.
(0.08 to 0.12 in.) greater than that of the stem, the bottom of the enlargement
being 114 rom. (4.5 in.) above the bottom of the bulb.
BULB: Corning normal or equally suitable thermometric glass.
Length, 25 to 35 mm. (1.0 to 1.4 in.). ,
Diameter, not less than 5 mm. (0.197 in.) and not greater than that of stem.
DISTANCE TO LOWEST SPECIF.IED GRADUATION FROM BOTTOM OF BULB: 135 to 150
mm. (5.3 to 5.9 in.). I
(d) If the test temperature is above that_oJ: the room, the oil shall be heated
to not more than 3° F. (1.7° C.) above the temperature( of test and if the tem-
perature is below that of the room, the oil shall be cooled to not more than 3° F.
(1.7~ C.) below the temperature of test. The oil shall\be poured into the oil I
tube until it ceases to overflow into tlle galleryt The, 6il in the oil tube shall
be kept well stirred with the oil-tube thermometer, care being taken to avoid
hitting the outflow tube. The bath temperature shall' be adjusted until the
oil temperature remains constant. After thermal equilibrium has been attained
no further adjustments shall be made in the bath temperature. The test
results shall be discarded if the indicated bath temperature varies by more
than ±0.05° F. (0.03° C.) in tests at 70°, 77°, 100°, 122° and 130° F. (21.1°,
25°, 37.8°, 50° and 54.4° C.) or by more than ±0.1O° F. (0.06° C.) in tests at
140°, 180° and 210° F. (60~, 82.2° and 98.9° C.).
(e) After the temperature of the oil in the oil tube has remained constant
within 0.02° F. (0.01° C.) of the desired temperature for 1 min. with constant
stirring, the oil-tube thermometer shall be withdrawn and the surplus oil
removed quickly from the gallery by means of the withdrawal tube so that the .-
level of the oil in the gallery"is below the level in the oil tube proper. The tip
of the withdrawal tube shall be inserted at one point in the gallery. The test
shall be started over again if the tip of the withdrawal tube touches the overflow
rim (see Fig. 244). Under no conditions shall, the excess oil be removed by rotat-
ing the withdrawal tube around the gallery.
(f) The receiving flask shall be placed<in position so that the stream of oil
from the outlet tube will strike the neck of the flask. The graduation mark on
. the receiving flask shall be not less than 10 cm. or more than 13 cm. from the
bottom of the bath. The cork shall be snapped from its position and at the
same instant the timer shall be started. The timer shall be stopped when
the bottom of the meniscus of the oil reaches the mark on the neck of the
receiving flask.
Reporting Results.-The,time in seconds as determined by the prescribed
procedure, with the proper .calibration correction, is/the Saybolt Universal
(or Saybolt Furol) viscosity of the oil at the temperature at which the test is
made. Results shall be reported to the nearest 0.1 sec. for viscosity values
-below 200 sec. and to the nearest whole second for values 200 sec. or above.
Accuracy.-With proper attention to details of method of procedure, results
in different laboratories with different operators under referee or standardization
conditions of testing, should not differ by more than 0.5%:-- ,
MacMichael's Viscosimeter.-This is an instrument of the torsion type
in which a disk is suspended in a cup of fluid, the latter being rotated, and the
torsion it produces on the disk noted. The disk is suspended by a gold-plated,
steel wire 10 inches long, held between two grooveu pins at the top of the stand-
ard. The brass cup is oil jacketed, the oil covering the wires 72 inch and heated
electrically, or it may be cooled by ice or brine. A bent thermometer passing
tlirough the cover indicates the temperature, which may be controlled easily
within a fraction of a degree. The graduated dial at the top of the disk is
rotary and may be easily set to 0°. A dash pot filled with engine oil on the
stem of the disk damps the action of the motor. In operating/the instrument.is
levelled, the cup is filled to the mark with the fluid to1>e tested (about 100 mI.),
the temperature adjusted, the cup rotated at about 20R.P.M. and the readings
FIXED OILS, FATS AND WAXES 1725
of the dial noted. These are in degrees of angular deflection, 300 to the cir-
cumference, and noted as ° M. = 1/1000 poise. Water at 20° should read
10° M. The smallest, or " practical unit" is 10M.: by changing the decimal
point, practical units, absolute units or specific viscosity may be obtained at
one reading of the dial, no calculations being required. Three strengths of wire
are supplied, giving water readings of 1°, 10°, and 100°, covering ranges of
viscosity to 5000° M. or 5 poise.
The instrument should be calibrated by solutions of pure granulated sugar
dried over sulfuric acid, the absolute viscosity of which is shown below. The
accuracy of the instrument is well within 0.5%. The instrument is very rapid,
the time required for a reading being very short. It can be applied to colloidal
or lumpy solutions or suspensions, as limes, clay, glues, gums, explosives, paints,
catsups or even jams with fruit seeds in them. The table below gives the wires
applicable to different liquids.
,.
CONSTANTS OF A SET OF TYPICAL WIRES
- .. ,
1.7%
The cloud point of a petroleum oil is the temperature at which paraffin wax
or other solid substances begin to crystallize out or separate from solution when
the oil is chilled under definite prescribed conditions.
The pour point of a petroleum oil is the lowest temperature at which the oil
will pour or flow when it is chilled without disturbance under definite prescribed
conditions.
The test for cloud point shall be used only for oils which are transparent
in layers 172 in. thick.
The test for pour point shall' be used for all other petroleum oils and may
be used for oils on which the test for cloud point is permitted.
Apparatus.-(See Fig. 246.) The test jar, a, shall be of clear glass, cylindri-
cal form, flat bottom, approximately 1 136 to 1/6 in. in inside diameter and'
4Yz to 5 in. high. An ordinary 4-oz. oil sample bottle may be used if it is
within the above specifications, and no test jar is available.
The thermometer, b, shall conform to the requirements of one of the follow-
ing specifications: .
(a) These specifications cover a special thermometer graduated in either
Centigrade or Fahrenheit degrees as specified, the ranges being -38 to +50 0 C.
or -36 0 to +120 0 F., respectively.
TYPE: Etched stem, glass.·
LIQUID: Mercury. .
RANGE AND SUBDIVISION: -38 to +50°C. in 1°C. or -36 to +120°F. in 2°F.
25 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
This method comprises a revision of the Standard Method of Test for Cloud and
Pour Points (A. S. T. M. Designation: D .97-33), 1933 Book of A. S. T. M. Standards,
Part II, p. 761, which method it supersedes.
26 This method is issued under the fixed designation D 97; the final number indicates
the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1921; Adopted in Amended Form, 1928; Revised, 1930,
1933, 1934. This method has been approved as American Standard by the American
Standards Association.
1728 FIXED OILS, FATS AND WAXES
TOTAL LENGTH: 220 to 224 mm. (8.69 to 8.81 in.).
STEM: Plain front, enamel back, suitable thermometer tubilll!. Diameter, 7.0 to 8.0
mm. (0.28 to 0.31 in.). I -
BULB: Corning normal or equally suitable thermometric glass.
Length, not over 9.5 mm. (0.37 in.). ~
Diameter, not greater than stem. ,
DISTANCE TO -38° C. OR -36° F. LINE FROM BOTTOM OF ULB: 120 to 130 mm. (4.73
to 5.12 in.). i ..
DISTANCE TO +49°'C. OR +120° F. LINE FROM Top OF THERMOMETER: 19 to 25 mm.
(0.75 to 0.98 in.).
EXPANSION CHAMBER: To permit heating to 100° C. or 212° F.
FILLING ABOVE MERCURY: Nitrogen gas.
Top FINISH: Plain. .
GRADUATION: Ail lines, figures, and letters clear cut and distinct. Lines at multiples
of 5° C. or 10° F. to be longer than the remaining lines. Graduations to be num-
bered at each multiple of 10° C. or 20° F.
IMMERSION: 108 mm. or 4X: in. The words "10S-mm. immersion" on Centigrade
thermometers or "4X:-in. immersion" on Fahrenheit thermometers and a line
around the stem 108.0 mm. or 4.25 in. above the bottom of the bulb shall be etched
on. the thermometer.
SPECIAL MARKING: "A. S. T. M. Cloud and Pour," a serial number and the manu-
facturer's name or trade mark shall be etched on the stem. ....
SCALE ERROR: The error at any point of the scale, when the thermometer is stand-
ardized as provided below, shall not exceed 0.5° C. or 1° F.
STANDARDIZATION: The thermometer shall be standardized at the ice point and at
intervals of approximately 20° C. or 40° F. for lOS-mm. or 4U-in. immersion and
for an average temperature of 21° C. or 70° F. for the emergent mercury column.
CASE: The thermometer shall be supplied ,in a suitable case -on which shall appear
the marking: "A. S. T. M. Cloud and Pour, -38 to ,+50° C." or "A. S. T. M.
Cloud and Pour, -36 to +120° F." according to the type of thermometer.
NOTE.-For the purpose of interpreting these specifications the following defini-
tions apply: ,
The total length is the over-all length of the finished instrument.
The diameter is that measured with a r.ing gage.
The length of the bulb is the distance from the bottom of the bulb to the beginning
of the enamel backing.
The top of the thermometer is the top of the finished instrument.
(b) These specifications cover a special low ~loud Ind pour test thermometer
graduated 'in either Centigrade or Fahrenheit degrees as specified, the ranges
being -60 0 to +20 0 C. or -70 0 to +70 0 F.
TYPE: Etched stem, glass.
LIQUID: Toluene or other suitable liquid colored red with a permanent dye.
RANGE AND SUBDIVISION: -60° to +20 0 C. in 10 C. or -70"._,to +70 0 F. in 2 0 F.
TOTAL LENGTH: 227 to 231' mm. (8.92 to 9.08 in.). '
STEM: Plain front, enamel back, suitable thermometer tubing. Di~~eter 7.0 to 8.0
mm. (0.28 to 0.31 in.).
BULB: Corning normal or equally suitable thermometric glass.
Length, 8.0 to 9.5 mm. (031 to 0.37 in.). ,
Diameter, 5.0 to 6.5 mm. (0.22 to 0.26 in.).
DISTANCE TO -57° C. OR -70° F. LINE FROM BOTTOM OF BULB: 120 to 130 mm. (4.73
to 5.12 in.). !
DISTANCE TO +20° C. OR +68 0 F. LINE FROM Top OF' THERMOMETER: 35 to 45 mm.
(1.38 to 1.77 in.).
EXPANSION CHAMBER: To permit heating,to +60° C. or +140 0 F.
FILLING ABOVE LIQUID: Gas under pressure.
Top FINISH: Plain. I
GRADUATION: All lines, figures and letters clear cut and distinct, lines at multiples
. of 5° C. or 10° F. to be longer than the remaining lines. Graduations to be num-
bered at each multiple of 10° C. or 20° F.
FIXED OILS, FATS AND WAXES 1729
IMMERSION: 198 mm .. (4~ in.). The words "108-mm. immersion" on Centigrade
thermometers, or "4U-in. immersion" on Fahrenheit thermometers and a line
around the stem 108 mm. (4.25 in.) above the bottom of the bulb. shall be etched
on the thermometer. .
SPECIAL MARKING: "A. S. T. M. Low Cloud and Pour" a serial number and the manu-
facturer's name or trade mark shall be etched on the thermometer.
SCALE ERROR: The error at any point of the scale when the thermometer is stand-
ardized as provided below, shall not exceed 10 C. or 2 0 F.
STANDARDIZATION: The thermometer shall be standardized at the ice point and at
intervals of approximately 20 0 C. or 350 F. for 10S-mm. or 4U-in. immersion, and
for an average temperature of 21 0 C. or 70 0 F. for the emergent liquid column.
CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking" A. S. T. M. Low Cloud and Pour, -600 to +20 0 C." or "A. S. T. M.-
Low Cloud and Pour, -70 0 to +70 0 F.," according to the type of thermometer.
N OTE.-For the purpose of interpreting these
specifications, the following definitions apply:
The total length is the overall length of the
finished instrument.
The diameter is that measured with a ring
gage.
The length of the bulb is the distance from the
bottom of the bulb to tht,' beginning of the
enamel backing.
The top of the thermometer is the top of the
finished instrument.
The cork, c, shall fit the test jar, and- ~
shall be bored centrally to take the test ther-
mometer. The jacket, d, shall be of glass or rl t:-
. .~.
: 1',
d"
_,
metal, shall be water tight, of cylindrical
form, flat bottom, about 4Yz in. deep, with
inside diameter % to Yz in. greater than out-
.f.
side diameter of the test jar.
A disk of cork or felt, e, U in. thick and k~'---ill--.l~
of the same diameter as the inside of the" ~
jacket will be required. IL--JO:~~
The ring gasket, J, shall be about 136 in. ~_.!tiJ
thick, and made to fit snugly around the out- .~
side of the test jar and loosely inside the
jacket. This gasket may be made of cork,
felt or other suitable material, elastic enough to
cling to the test jar and hard enough tohold its FIG. 246.-Apparatus for Cloud
and Pour Test
shape. The purpose of the ring gasket is to (as assembled for Cloud Test).
prevent the test jar-from touching the jacket.
The cooling bath, 1], shall be of a type suitable for obtaining the required
temperatures. The size and shape of the bath are optional but a support,
suitable for holding the jacket firmly in a vertical position, is essential. For
determination of pour points below 50° F., two or more baths should be at hand.
The required bath temperatures may be maintained by refrigeration if available,
otherwise by suitable freezing mixtures.
NOTE.-The freezing mixtl1res commonly used are as follows:
For temperatures down to 50 0 F., ice and water. .-
0
" " " " 10 F., crushed ice and sodium chloride.
" " " " -15 0 F., crushed ice and calcium chloride crystals.
" " " " -70 0 F., solid carbon dioxide and acetone,or gasoline.
1730 FIXED OILS, FATS AND ~4XES
The last named mixture may be made as follows: In a covered metal beaker chill
a suitable amount of acetone or gasoline to lOo-F., or lower, by means of an ice-salt
mixture. Invert a cylinder of liquid carbon dioxide and draw off carefully into a
chamois skin bag the desired amount of carbon dioxide, which through rapid evapora,
tion will quickly become solid. Then add to the chilled ace~one or gasoline enough of
the solid carbon dioxide to give the desired temperature. •I
Procedure.-T,he oil to be tested shall be brought to a temperature at least
25° F. above the approximate cloud point. Moisture, if present, shall be
removed by any suitable method, as by filtration through dry filter paper until
the oil is perfectly clear, but such filtration shall be made at a temperature at
least 25° F. above the approximate cloud point.
The clear oil shall be poured into the test jar, a, to a.height of not less than
2 nor more than 274: in. The test jar may be marked to indicate the proper
level.
The test jar shall be tightly closed by the cork, c, carrying the test ther~
mometer, b, in a vertical position in the center of the jar with the thermometer
bulb resting on the bottom of the jar.
The disk, e, shall be placed in the bottom of the jacket, d, and the test jar--
with the ring gasket, j, 1 in. above the bottom shall be inserted into the jacket
The disk, jacket and inside of jacket .shall be clean and dry.
The temperature of the cooling bath, g, snaIl be maintained at 30 to 35° F.
The jacket, containing the test jar, shall be supported firmly in a vertical
position in the cooling bath so that not more than 1 in. of the jacket projects
out of the cooling medium.
At each test thermometer reading which is a multiple of 2° F., the test jar
shall be removed ffom the jacket, quickly but without disturbing the oil,
inspected for cloud, and replaced in the jacket. This complete operation shall
require not more than three seconds. If the oil does not show a cloud wheI.l it
has been cooled to 50° F., the test jar shall be placed in a second bath main-
tained at a temperature of 0° to +5° F. If the oil does not show a cloud when
it has been cooled to 20° F., the test jar shall be placed in a third bath main-
tained at a temperature of -'30° to -25° F. . /
When such inspection first reveals a distinc~ cloudiness or haze in the oil,
at the bottom of the test jar, the reading of the test thermometer, corrected
for error if necessary, shall be recorded as the cloud point.
Procedure for Pour Point.-The oil shall be poured into the test jar, a, to a
height of not less than 2 nor more than 274: in. The jar-rnay be marked to
indicate the proper level. . When necessary, the oil shall be heated. in a water
bath just sufficiently for pouring into the test jar. _
The test jar shall be tightly closed by the cork, c, carrying the test ther-
mometer, b, in a vertical position in the center of-the jar with the thermometer
bulb immersed so that the beginning of the capillary shall be Ys ih. below the
surface of the oil.
Heat without stirring to a temperature of 115° F. in a bath maintained at
not higher than 118° F. The oil shall then be cooled to 90° F. in air or in a
water bath approximately 77° F. in temperature. Oils on which a pour point
below -30° F. is expected shall be heated as above with the high cloud- and
pour-test thermometer in position, cooled to 60° F ...l.uHFthe low cloud~ and pour-
test thermometer placed in position and the assembly-placed in the jacket. Oils
\
FIXED OILS, FATS AND WAXES 1731
A. S. T. M. DESIGNATION: D 92-33 28
Scope.-The open cup flash and fire test on all products except fuel oils and
those having an open cup flash below 175° F. shall be determined in the
Cleveland Open Cup.
Apparatus.-The Cleveland open cup, a diagram of which appears in Fig.
247, shall include the following major parts:
(a) Flash Cup.-The cup shall be made of Brass and shall conform to 't~
following requirem.ents:
Inches Centimeters
3.2
,
• 0.28 0.32 0.36
TI
6. 0.20 0.24 0.28
_1_
.....
~"
~
+1
Filling Mark 450 ~~
hr~--~--~~~~--~-
--'-___;--!.......::..:.-=---"-~ ~~ I =t~
"RI~
t.!."'>l +1
" ~~ I !~
5p2 Rad.
Approx. 1
~
- _ _ _-1-
l-
~2/06(:!:~n~I
I
FIG. 247.~Cleveland Open Cup.
center of the asbestos board a circular hole just fitting the cup. Heat may be
supplied from any convenient source. The use. of a gas burner, electric heater,
or alcohol lamp is permitted, but under no circumstances are products of com-
bustion or free flame allowed to come up around the cup. The source of heat
shall be centered under the opening i.n the plate and shall be of a type that will
not produce local superheating. If a flame heater is used, it may be protected
from drafts or excessive radiation by any suitable type of shield, that does not
project above the level of the upper surface of the asbestos board.
Thermometer.-The thermometer shall conform to the following require-
ments. These specifications cover a special thermometer graduated in either
Centigrade or Fahrenheit degrees as specified, the ranges being - 60 to +400 0 C.,
or +20 0 to +760° F., respectively.
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGE AND SUBDIVISION: _6° to +400° C. in 2° C. or +20° to +760° F. in 5° F.
TOTAL LENGTH: 303 to 307 mm. (11.92 to 12.08 in.).
STEM: Plain front, enamel back, suitable thermometer tubing.
Diameter, 6.0 to 7.0 mm. (0.24 to 0.28 in.).
BULB: Corning normal or equally suitable thermometric glass.
Length, not over 13 mm. (0.51 in.).
Diameter, not greater than stem.
. .
1734 FIXED OILS, FATS AND WAXES
DISTANCE TO .. _6 0 C. OR 20 0 F. LINE FROM BOTTOM OF BULB: 40 to 50 mm. (1.57 to
1.97 in.). ,
DISTANCE TO 400 0 C. OR 760 0 F. LINE FROM Top OF THERMOMETER: 30 to 45 mm.
(1.18 to 1.77 in.).
FILLING ABOVE MERCD:aY: Nitrogen gas.
Top FINISH: Red glass ring. I
GRADUATION: All lines, figures, and letters clear cut and dis~inct. The 100 C. lines
0
or the first and each succeeding 10 F. line to be longer than the intermediate lines.
Graduations to be numbered at each multiple of 100 C. or 20 0 F.
IMMERSION: 25 mm. or 1 in. The words "25 mm. immersion" on Centigrade ther<-.
mometers or "I-in. immersion" on Fahrenheit thermometers and a line around
the stem 25.4 mm. or 1.00 in. above the bottom of the bulb shall be etched on the
thermometer.
I
I
I
I
I
I ,
?t 1(""-
~I
m&
L1
I:
,.;t t
~ Plale
FIG. 248.-Heatipg Plat\l.
SPECIAL MARKING: "A. S. T. M. Open Flash" a serial number and the manufacturer's
name or trade mark shall be etched on the stem. . --
SCALE .ERROR: The error at any point of the scale when the thermometer is stand-
ardized as provided below shall not exceed PC. up to 372 C. or ~Y20 F. up to
0
700° F.
STANDARDIZATION: The thermometer shall be standardized at the ice point and at
intervals of approximately 100° F. for 25-mm. or I-in. immersion and for the
following temperatures of the emergent mercury column:
(
Average Temperature of Emergent
Thermometer Reading Mercury Column
100'0 C. 20U o F .............................. " 44 0 C. 1100 F.
150 0 C~ 30000 F .............................. " 54~ C. 129 0 F.
200°0 C. 400 F.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 64° C. I 150 0 F.
250 C. 500 0 F . .' ............................ ;. ~_77° C.' 175° F.
300°0 C. 600 0 F ... '" ........ " .......... _. ..-".:-: 91°e. 205 0 F.
350 C. 700 0 F ............................ " ~ .1080 C. 240 0 F.
FIXED OILS, FATS ANP WAXES 1735
TEST FOR PERMANENCY OF RANGE: After being subjected to a temperature between
360 to 370.0 C. or 680 to 700° F. for 24 hours, the accuracy shall be within the
limit specified.
CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking: "A. S. T. M. Open Flash, _6° to +490° C." or "A. S. T. M. Open
Flash, +20° to +760° F.," according to the type of thermometer.
NOTE.-For the purpose of interpreting these specificatiOIis the following defini-
tions apply:
The total length is the over-all length of the finished instrument.
The diameter is that measured with a ring gage.
The length of the bulb is the distance from the bottom of the bulb to the beginning
of the enamel backing.
The top of the thermometer is the top of the finished instrument.' ,
Procedure.-(a) The thermometer shall be suspended or held in a vertical
positi9n by any suitable device. The bottom of the bulb shall be U in.
(0.635 em.) 29 from the bottom of the cup, and above a point half way between
the center and back of the cup.
(b) The cup shall be filled with the oil to be tested in such a manner that the
top.of the meniscus is exactly at the filling line at room temperature. When
asphalt or other solid bituminous material is to be tested, it shall first be heated
to a temperature not less than 300° F. nor more than 350° F., to render it
sufficiently fluid. The cup shall then be filled with the hot fluid at this tem-
perature in the same manner as with oiL The subsequent procedure shall then
be the same for both asphalt and solid bituminous material as with oiL The
surface of the oil shall be free from bubbles. There shall be no oil above the
filling line or on the outside of the apparatus.
(c) The test flame shall be a~proximately l2 in. (0.397 em.) in diameter.
NOTE.-For purposes of comparison it is recommended that a bead of suitable
light colored material be mounted in a convenient position so that the size of the test
flame can be determined. The device for applying the flame may be of any suitable
type but it is suggested that the tip be approximately is in. (0.159 cm.) in diameter at
the end and that the orifice be -h in. (0.079 cm.) in diameter. If the device for oper-
ating the test flame be mounted in such a manner as to permit automatic duplication
of the sweep of the test flame the radius of swing shall be not less than 6 in.
(d) The test flame shall be applied as the temperature read on the ther-
mometer reaches each successive 5° F. mark. The flame shall pass in a straight
line (or on the circumference of a circle having a radius of at least 6 in.) across
the center of the cup and at right angles to the diameter passing through the
thermometer. The test flame shall, while passing across the surface of the oil,
be in the plane of the upper edge of the cup. The time for the passage of the
test flame across the cup shall be approximately one second.
(e) The oil shall be heated at a rate not exceeding 30° F. per minute tem-
perature rise, till a point is reached approximately 100° F. below the probable
flash point of the oil. Thereafter the rate of heating shall be decreased and
for at least the last 50° F. before the flash point is reached, the rate shall be
not less than 9 nor more than 11 ° F. pe~ minute.
The flash point shall be taken as the temperature read on the thermometer
when a flash appears at any 'point on the surface of the oil. The true flasli
must not be confused with a bluish halO that sometimes surrounds the test flame.
29 The immersion line engraved on the thermometer stem will be is in. (0.159 cm.)
below the level of the rim of the cup.
1736 FIXED OILS, FATS AND WAXES
After determining the flash point, the heating shall be continued at the
specified rate of 9 to 11°F. per minute, and applicatiop of the test flame shall
be made at the specified intervals until the oil ignites and continues to burn for
a period of at least 5 seconds. The method of application of the flame shall
be the same as for flash point. The temperature read ~t the time of the flame
application which causes burning for a period of 5 se'conds or more shall be
recorded as the fire point. I '
Precautions.-The flash point and fire point tests shall be made in a room or
compartment free from air drafts. The operator shall avoid breathing over
the surface of the oil. It is desirable that the room or compartment be darkened
sufficiently so that the flash may be readily discernible.
Detection of Soap.-To increase the viscosity of an oil,30 resort is had to the
use of " oil pulp," " oil-thickener," or " white gelatin," usually an ol~ate of
aluminum, though other bases may be present. Its disadvantages are that it
causes the oil to chill more easily and to emulsify, thus increasing the friction.
Furthermore, it is precipitated by contact with water or steam, causing clogging
of the machinery. .;
The test depends upon the fact that the meta phosphates of the earthy and
alkali metals and aluminum are insoluble in absolute alcoho1. 31
The test is applied as follows: five to 10 illl. of the oil to be tested are dis-
solved in about 5 ml. of 86° gasoline or ether, and about 15 drops of the phos-
phoric acid solution (Reagents, p. 1799) add~d, shaken and allowed to stand;
the formation of a flocculent precipitate indicates the presence of soap. An
idea of the kind of soap can be often gained by adding an alcoholic solution of
PtCI 4 • If the precipitate becomes crystalline, it is a potash soap; if it dissolves,
soda, lime, or magnesia; if unchanged, alumina or iron.
For the accurate determination of these compounds a known weight of the
oil must be ignited, the residue determined and quantitatively examined.
Caoutchouc.-Holde 32 states that 1 to 2% of unvulcanized caoutchouc is
sometimes added to oils to increase their viscosity. This may be detected by
adding three parts of alcohol to four parts of the etherlf~l solution, whereby the
rubber material is precipitated and may be dried and weighed.
Test for Fatty Oils.-To detect small quantities of fatty oil (U to 2%)
Lux 33 recommends heating a few ml. of the oil Jor fifteen minutes with some
bits of sodium in a test-tube in an oil bath; a similar test is made with sodium
hydroxide. The temperature employed should be for light oils about 230°,
for dark oils 250°.34 In case fatty oil be present, the contents "of one or both
of the tubes solidify to a jelly of greater or less consistence according to the
amount of fatty oil present.
The quantitative determination of these oils, as for example in cylinder oils,
is effected after the manner of determining the saponification value (p. 1771)
or the detection of unsaponifiable oils in fatty oils (page 1773).
30 In a case which came to the writer's notice the oil would not flow out of the Say.
bolt "A" apparatus at 70°, at 85° required 1167", and at 110 181".
0
,
center of the iron crucible. Apply covers to both the Skidmore and me iron
crucible, the one to the latter fitting loosely to allow free exit to the vapors as
formed.
On a suitable stand or ring, place the bare nichrome wire triangle and on it
the asbestos block or hollow sheet-metal box. Next, centre the sheet-iron
crucible in the block with its bottom r.esting on top of the triangle, and cover
the whole with the sheet-iron hood in order to distribute the heat uniformly
during the process (see Fig. 249). .
Apply heat with a high strong flame from the Meker-type gas burner, so
that the pre-ignition period will be 10 minutes ±IYz minutes (a shorter time
may start the distillation so rapidly as to cause foaming or too high a flame).
When smoke appears above the chimney, immediately move or tilt the burner
so that the gas flame plays on the sides of the crucible for the purpose of igniting
vapors. Then remove the heat temporarily, and before replacing adjust by
screwing down the pinch-cock on the gas tubing so that the ignited vapors
burn uniformly with the flame above the chimney but not above the wire
bridge. Heat may be increased, if necessary, when the flame does not show
above the chimney. The period of burning the vapors shall be 13 minutes
±1 minute.
When the vapors cease to burn and no further blue smoke can be observed,
readjust the burner and hold the heat as at the beginning so as to make the
bottom and lower part of the sheet-iron crucible a cherry red and maintain for
exactly 7 minutes. The total period of heating shall be 30 minutesi=2 minutes
which constitutes an additional limitation on the tolerances for the pre-ignition
and burning periods. There should be no difficulty in carrying out the test
exactly as directed with the gas burner of the type named, using city gas
(about 550 B.t.u.) with the top of the burner about 2 in. below the bottom of
the crucible. The time periods shall be observed with whatever burner and
gas is used.
Remove the burner and allow the apparatus to cool until no smoke appears,
and then remove the cover of the Skidmore crucible (about 15 minutes).
Remove the porcelain or silica crucible with heated tongs, place in the desic-
cator, cool and weigh. Calculate the percentage of carbon residue on the
original sample. .
Special Procedure: SPecial Procedure for Oils Having Carbon Residues in
Excess of 5%, Such as Heavy Crude Oils, Residuums, Heavy Fuel Oils, and
Heavy Gas-Enrichment Oils.-When the carbon residue as obtained by the
procedure described above (using a 10-g. sample) is in excess of 5%, diffi-·
cui ties may be experienced due to boiling over of the sample. Trouble also
may be encountered with samples of heavy products which are difficult to de-
hydrate. In the case, therefore; of samples showing by the usual method of
test a carbon residue in excess of 5%, the test shall be repeated in duplicate
using a 5-g. sample. •
Tolerances.-Weights of oil sample shall be accurate to within 5 mg. Tests
shall be run in duplicate and repeated if necessary until the percentages of
carbon residue differ by not more than 10% from an average.
Gasoline Test.-This shows the presence of tar (still bottoms) or asphaltic'
matters.
1740 FIXED OILS, FATS AND WAXES
Mix 10 ml. of the oil with 90 ml. 86°-88° gasoline (from Pennsylvania crude)'
B. pt. 30°-50° C., allow to stand one hour' at 70° to 8Qo F.; not more than 10%
of flocculent or tarry matter should have settled out. This settling can be
facilitated by whirling in a centrifuge in a. graduated ,ube in which the volume
of the precipitate can be read Gff. If the test b~ applie~ to the oil before making
the flash test and then again after this test it shows t~e extent to which the oil
is changed upon heating. Other things being equal the oil which is changed
the least is the best oil. See page 1761.
Microscopical Test.-Put.a few drops of the well-mixed oil on a slide f!nd
note the nature of the suspended matter-whether carbonaceous specks, flakes
of paraffin which disappear on warming, or foreign matter. A good oil should
be practically free from all these bodies.
Heat Test.-Heat 20-25 ml. of the oil in an Erlenmeyer flask or beaker,
nearly to the flashing point and keep it at this temperature for fifteen minutes;
a satisfactory oil will darken, but remain clear even after standing a day. A
poorly refined oil changes to jet black, and forms a carbon-like precipitate.
This is usually indicative of an oil that has been refined by acid treatment ...
Herschel's Demulsibility Test. 40 Method 3Z0.31.-The test is applicable
to non-emulsifying oils, such as are used with high speed engines and turbines,
and in its present form does not apply to stt;am cylinder or other compounded
oils. It is made as follows: _
Twenty m!. of oil and 40 ml. of distilled water are placed in a 100 m!.
cylinder having an inside diameter of 2.&-cm. and heated to 55°. The oil and
water are' t~en mixed or emulsified by stirring with a paddle for five minutes at a
speed of 1500 R.P.M. The paddle is simply a metal p~ate 89 by 20 by 1.5 mm.
thick, conveniently driven by an electric motor.
The rate at which the oil settles out from the emulsion, when it is allowed
to stand at a temperature of 55° C. is a measure of the emulsibility of the oil.
There is a tendency for the rate of settling out of the oil to increase up to a
maximum and then decrease, and we may take readings every five minutes
(or every minute if the oil separates out rapidly) and ,note the maximum rate.
For example, the following readings were obtained for a certain oil:
9.50 .......
9.55 .......
10.02 .......
0
5
12
, 60
57
0
3
- - 0
36
47 13 65
10.05 ....... 15 43 19 68
10.10 ....... 20 41 ,
17 57
The maximum rate is here 68 ml. per hour, and:this is called the" demul-
sibility." The best transformer, motor, and turbine oils settle out in one
minute or less, so that if we take readings to even minutes and even ml., the
rate would be 1200. Oils are found with rates all the way: down from this
maximum rate to zero. l
LIQUID FUELS
(a) Gasoline. The tests usually made are coIOr,46! doctor test,46 corrosion,
distillation, acidity, and sulfur.
The color, acidity, and sulfur tests are made as with kerosene.
The doctor test (sodium plumbite) is made as follows: 47
Distillation of Gasoline
Apparatus and Procedure.-The apparatus and procedure have been given
in method D 86-35, p. 1705:
Significance· of Test.- R. E. Wilson 50 says, the 10% point is the best
indication of ease of starting, vapor lock (that is stopping of engine by too
volatile gasoline), and evaporation loss. The 50% point is important for
good acceleration. The 90% point is related to the dew point and controls
crank case dilution.
Octane Number is the per cent by volume of iso octane in a mixture of
iso octane and normal heptane required to match the antiknock value of a given
fuel.
48 Under the standardization procedure of the Society, this method is Under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
49 This method is issued under the fixed designation D 130; the final number indi-
cates the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1922; Adopted in Amended Form,. 1930. This methoci
has been approved as AMERICAN STANDARD by the American Standards Association.
50 Chem. Abs. 27, 184.
1744 FIXED OILS, FATS AND WAXES
FUEL OILS
The tests usually made are flash, viscosity,. sulfur, I~ater and sediment.
A. S. T. M. DESIGNATION: D 93-36 52
/
Inches C:;entimeters
Dimensions
Mini- Nor- Maxi- Mini- Nor- Maxi-
mum mal mum mum mal mum
- - - - - -- -
Inside diameter below filling mark ......... 1.950 2.000 2.050 4.953 5.080 5.207
Difference, inside and outside diameters be-
low filling mark ....................... 0.120 0.125 0.130 0.305 0.318 0.330
Inside height ............................ 2.150 2.200 2.250 5.461 5.588 5.715
Thickness of bottom ..................... 0.070 0.095 0.120 0.178 0.241 0.305
Distance from rim to filling mark .......... 0.845 0.860,, 0.875 2.146 2.184 2.223
Distan:ce lower surface flange to bottom of ,
cup .................................. 1.780 1.795 1.810 4.521 4.559 4.597
The insi~e of the cup may be turned to a slightly larger diameter above the
filling mark and the outside may be tapered above t~ flange but the wall
thickness at the upper edge shall be not less than 0.04 in~ '(0,10_~ cm.). The
flange should be approximately 0.5 in. (1.27 cm.) wide and approximately
0.125 in. (0.318 cm.) thick. It. shall be equipped with devices tor locating the
position of the lid on the cup and the cup in ,the stove. A handle, attached
permanently to the flange of the cup, is a desirable accessory.
51 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petr6leum Products .and Lubricants.
Reproduced through. the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
62 This method is issued under the fixed designation D 93; the final number indicates
the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1921; Adopted in Amen.!led Form, 1922; Revised, 1936.
This method has been approved as AME1RICAN STANn'ARD. by -the American Standards
Association.
FIXED OILS, FATS AND WAXES 1745
(b) Lid.-
(1) Stirring Devit;e.-The lid shall be equipped with a stirring device
consisting of a vertical steel shaft, not less than 0.1 in. (0.254 em.) nor more
than 0.125 in. (0.318 em.) in diameter, mounted in the center of the cup, and
carrying two two-bladed brass propellers. The blades of both propellers shall
be approximately 0.313 in. (0.795 em.) wide and shall be set at an angle of
No. 15 and No. 12 B. and S. gage sheet brass. The collars on which the pro-
peller blades are mounted shall have horizontal and vertical dimensions not
greater than 0.4 in. (1.016 cm.). 1 '
The plane of the center of the upper, propeller shall be 0.4 in. (1.016 cm.)
below the level of the rim of the cup. The plane of t~e center of the Iowel\.pro-
peller shall be 2.0 in. (5.08 cm.) below the level of the rim of the cup. The level
of the rim of the cup is in effect the level of the plane~ part of the portion of the
lower surface of the lid inside the rim.
(2) Cover proper.-The cover proper shall be of brass and shall have a rim
projecting downward almost to the flange of the cup and fitting the outside of
the cup closely. The thickness of the cover, measured just inside the rim ,shall
be not less than 0.031 in. (0.079 cm.) nor more than 0.078 in. (0.198 cm.).
There shall be a proper locating device engaging with a corresponding locating
device on the flange of the cup.
There shall be four ope?ings in the cover, as indicated in Fig. 251.
Opening A is an area defined by arcs of two concentric circles and the inter-
sected lengths of two radii. The radius of the outer circle shall be not less than
0.938 in. (2.383- cm.) nor more than 0.969 in. (~4.£n cm.),. The radius of the
inner circle shall be not less than 0.531 in. (1.349 em.) nor more than 0.563 in.
FIXED OILS, FATS AND WAXES 1747
(1.430 cm.). The chord of the arc of the outer circle shall be not less than
0.500 in. (1.270 cm.) nor more than 0.540 in. (1.372 cm.).
Openings Band C are equal areas, each of the same general form as opening
A but of approximately half the (angular) width. The radii of the defining
inner and outer circles shall be within the limits specified for the radii of the
two circles, arcs of which partly define opening A. The chord of the outer arc
for opening B or opening C shall be not less than 0.187 in. (0.475 cm.) nor more
than 0.219 in. (0.556 cm.). The sum of the areas of openings Band C shall
be not less than 75% nor more than 100% of the area of opening A. Openings
Band C shall be equally distant from opening A and radii drawn through each
of their centers shall be at an angle of not less than 1350 nor more than 140°.
Openings A, B, and C need not conform exactly to the shape of geometrical
figures bounded by arcs of two concentric circles and intersected lengths of
radii. Their boundaries must, however, fall on or between the lines indicated
by the limiting values of the dimensional specification of the preceding text
and of Fig. 251.
Opening D is for a thermometer collar. Its center is approximately 0.75 in.
(1.905 cm.) from the center of the lid and on a radius at an angle of not less
than 50 0 nor more than 60° from a radius passing through the center of opening
C. The thermometer collar shall have an inside diameter of approximately
0.5 in. (1.27 cm.). It shall be set at an angle of not less than 10° nor more than
15° from the perpendicular. -
(3) Shutier.-The lid shall be equipped with a brass shutter, approximately
0.094 in. (0.239 cm.) thick operating on the plane of the upper surface of the lid.
The shutter shall be so shaped aI).d mounted that it rotates on the axis of the
horizontal center of the lid between two stops so placed that when in one extreme
position the openings A, B, and C of the lid are completely closed and when in
the. other extreme position these orifices are completely opened.
(4) Flame exposure device.-The flame exposure device shall have a tip
with an opening 0.027 in. (0.069 em.) to 0.031 in. (0.079 cm.) in diameter.
The flame exposure device shall be equipped with an operating mechanism
which, when the shutter is in the " open" position, depresses the tip so that
the center of the orifice is between the planes of the under and upper surfaces
of the lid proper at a point on a radius passing through the center of the larger
opening ~ and apprl1ximately 0.1 in. (0.254 cm.) from the outer edge of the
opening.
NOTE.-A pilot flame for automatic relighting of the exposure flame should be
provided. '
A bead 0.156 in. (0.397 cm.) in diameter, of some suitable material, may be mounted
on the lid so that the size of the test flame can be regulated by comparison.
The mechanism operating the shutter should be of the spring type and constructed
so that when at rest the shutter shall exactly close the three openings. When operated
to the other extreme the three openings in the lid shall be' exactly open and the tip
of the exposure tube shall be fully depressed.
(c) Siove.-Heat shall be supplied to the cup by means of a properly
designed stove which is equivalent to an air bath. This stove shall consist of
(1) an air bath and (2) a top plate on which the flange of the cup rests. '
(1) Air bath.-The air bath shall have a cylindrical interior 1.625 in.
(4.128 cm.) to 1.656 in. (4:206 cm.) deep and a diameter not less than 0.125 in.
1748 FIXED OILS, FATS AND WAXES
(0.317 em.) nor more than 0.156 in. (O.396.cm.) greater than the outside diam-
eter of the cup. The air bath may be either a flame heated metal casting\or
an electric resistance element.
NOTE.-If the heating element is a flame-heated metal c~ting it shall be so designed
and used that the temperature of bottom and walls is approximately the sarrie. On
this account it should be not less than 0.25 in. (0.635 em.) thick. The casting shall be
designed so that products of combustion of the flame cannot pass up and in. contact
with the cup.
If the air bath is of the electric resistance type it shall be constructed so that all
parts of the interior surface are heated equally. This necessitates an even distribution
of'resistance wire over bottom and walls and a method of construction such that heat
is given out from. the whole core of the resistance element rather than directly from
the wire.
(2) Top plate.-The top plate shall be of metal. The total distance from
the upper surface of the plate to the bottom of the air bath shall exceed the
distance from the under surface of the flange to the bottom of the cup by not
less than 0.063 in. (0.160 em.) nor more than 0.125 in. (0.317 em.).
The top plate shall be mounted with an air gap between it and the air bath.
The top plate may be attached to the air bath by means of three screws and
spacing bushings. The spacing bushings should be of proper thickness to
define the air gap which shall be not less than 0.125 in. (0.317 cm.) nor more
than 0.187 in. (0.475 em.). The spacing bushings shall be not more than
0.375 in. (0.952 cm.) in diameter. /
(d) Thermometers.-Two standard thermometers shall be used with the
A. S. T. M. Pensky-Martens tester. The low range, " A. S. T. M. P.M. and
Tag" thermometer shall be used for tests when the indicated reading falls
within the limits 200 to 2000 F. The" A. S. T. M. P.M. high" thermometer
shall be used for tests when the indicated reading falls within the limits 230 0 to
7000 F. For the range 200 0 to 2300 F. either thermometer may be employed,
dependihg on the convenience of the operator. The thermometers shall comply
with the requirements given in Tables I and II. The ,Low Range" A. S. T. M.
P.M. and Tag" t}:lermometer specifications cove., a special thermometer
graduated in either Centigrade or Fahrenheit degrees as specified, the ranges
being _7° to +1100 C. or +20 0 to +230 0 F., respectively. The high range
"A. S. T. M. P.M. high" thermometer specifications cover a special ther-
mometer graduated in ei~her Centigrade or Fahrenheit degrees as specified,
the ranges being 900 to 3700 C. or 2000 to 700 0 F., respecth:ely.
Thermometers shall be mounted so that the bottom of the"bulb is 1.75 in.
(4.445 cm.) below the level of the rim of the cup (which corresponds to the level
of the lower surface of the portion of the lid inside the rim).
TABLE L-8PECIFICATIONS FOR Low RANGE "A. S. T. M. P.M. AND TAG"
THERMOMETER 63
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGE AND SUBDIVISION: _7° to +110° C. in ,0.5° C. or +20° to +230° F. in 1° F.
TOTAL LENGTH: 273 to 277 mm. (10.75 to. 10.92 in.).
63 The low range" A. S. T. M. P.M. and Tag" therm()meter ii?'the same instrument
as that specified for use with the Tag closed tester, see:::the Standard Method of Test
for Flash Point by Means of Tag Closed Tester (A. 8.""1';-1\1. Designation: D 56) of the
American Society for Testing Materials, see 1936 Book of A. S. T. M. Standards, Part II.
FIXED OILS, FATS AND WAXES 1749
STEM: Plain front, enamel back, suitable thermometer tubing. Diameter, 6.0 to 7.0
mm. (0.24 to 0.28 in.).
BULB: Corning normal or equally suitable thermometric glass.
Length, 9 to 13 mm. (0.35 to 0.51 in.).
Diameter, not greater than stem.
DISTANCE TO _7 0 C. OR +20 0 F. LINE FROM BOTTOM OF BULB: 75 to 90 mm. (2.93
to 3.54 in.).
DISTANCE TO 110° C. OR 230° F. LINE FROM Top OF THERMOMETER: 25 to 40 mm.
(0.98 to 1.57 in.).
EXPANSION CHAMBER: To permit heating the thermometer at least 50° C. (90° F.)
above highest temperature on scale.
FILLING ABOVE,MERCURY: Nitrogen gas.
Top FINISH: Glass ring.
GRADUATION: All lines, figures, and letters clear cut and distinct. The whole degree
Centigrade lines or the first and each succeeding. 5° F. line to be longer than the
remaining lines. Graduations to be numbered at each multiple of 5° C. or 10° F.
IMMERSION: 57· mm. or 2.)4 in. The words "57-mm. immersion" on Centigrade or
"2.)4-in. immersion" on Fahrenheit thermometers and a line around the stem
57.0 mm. or 2.25 in. above the bottom of the bulb shall be etched on the ther-
mometer.
SPECIAL MARKING: "A. S. T. M. P.M. and Tag.," a serial number and the manu-
. facturer's name or trade mark, shall be etched on the stem.
SCALE ERROR: The error at any point of the scale when the thermometer is stand-
ardized as provided below shall not exceed 0.5° C. or 1° F., respectively.
STANDARDIZATION: The thermometer shall be standardized at the ice point and at
intervals of approximately 30° C. or 50° F. for 57-mm. or 2.)4-in. immersion and
for the following temperatures of the emergent mercury colUmn:
Average Temperature of Emergent
Thermometer Reading Mercury Column
20°C. 70° F ............................ " .. 20~ C. 70° F.
40°C. 100° F ......... , ...................... 31°C. - 86° F.
70° C. 150° F ................................ 40° C. 104°F.
100° C. 212°.F, ............... " ............... 48° C. 118° F.
CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking: "A. S. T .. M. P.M. and Tag., _7° to + 110° C." or "A. S. T. M.
P.M. and Tag., +20° to +230° F." according to the type of thermometer.
NOTE.-For the purpose of interpreting these specifications the following defini-
tions apply:
The totalleIigth is the over-all length of the finished instrument.
The diameter is that measured with a ring gage.
The length of the bulb is the distance from the bottom of the bulb to the beginning
of the enamel backing.
The top of the thermometer is the top of the finished instrument.
TABLE 1I.-SPECIFICaTIONS FOR HIGH RANGE "A. S. T. M. P.M. HIGH"
THERMOMETER
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGE AND SUBDIVISION: 90° to 370° C. in 2° C. or 200° to 700° F. in 5° F.
TOTaL LENGTH: 273 to 277 mm. (10.75 to 10.92 in.).
STEM: Plain front, enamel back, sui~ble thermometer tubing. Diameter, 6.0 to 7.0
mm. (0.24 to 0.28 in.).
BULB: Corning normal or equally suitable. thermometric glass.
Length, not over 10 mm. (0.39 in.).
Diameter, not greater than stem.
DISTANCE TO 93° C. OR 200° F. LINE FROM BOTTOM OF BULB: 75 to 90 mm. (2.93 to
3.54 in.). .
DISTANCE TO 371°C. OR 700° F. LINE FROM Top OF THERMOMEll'ER: 25 to 40 mm.
(0.98 to 1.57 in.).
1750 FIXED OILS, FATS AND WAXES
FILLING ABOVE MERCURY: Nitrogen gas.
Top FINISH: Glass ring.
GRADUATION: All lines, figures, and letters clear cut and distinct. The first and each
succeeding 10 0 C. or 25 0 F. line to be longer than the remaining lines. Graduations
to be numbered at multiples of 20 0 C. or 50 0 F. I .
IMMERSION: 57 mm. or 274 in. The words :!.57-mm. immersion" ~m Centigrade ther-
mometers, or "274-in. immersion" on Fahrenheit therdlOmeters and a line around
the stem 57.0 mm. or 2.25 in. above the bottom of the ]:lUlb shall be etched on the
thermometer.
_ SPECIAL MARKING: "A. S. T. M. P.M. High," a serial number and the manufacturer's
name or trade mark shall be etched on the stem.
SCALE ERROR: The error at any point of the scale when the thermometer is stand-
ardized as described below shall not exceed 1 0 C. or 2Yz° F.
STANDARDIZATION: The thermometer shall be standardized at intervals of approxi-
mately 50 0 C. or 100 0 F. 'for 57-mm. or 274-in. immersion and for the following
temperatures of the emergent mercury column:
Average Temperature of Emergent
Thermometer Reading Mercury Column
0
1000 C. 200 F ................................ 61 0 C. 1400 F.
150 0 C. 3000 F ................................ 65 0 C. 149 0 F.
200 0 C. 400 0 F ................................. 71 0 C. 1600 F.
250 0 C. 500 0 F ............ " ........•........ , 78 0 C. 1750 F. -
3000 C. 600 0 F ....... , ........ " ....... '.' ...... 87 0 C. 1950 F.
3500 C. 7000 F ...................../........... 990 C. 2200 F.
TEST FOR PERMANENCY OF RANGE: After being subjected to a temperature between
360 0 and 370 0 C. or 680 0 and 700 0 F. for 24 hours, the accuracy shall be within
the limit specified. The test shall be made under the immersion conditions speci-
fied for this thermometer.,
CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking: "A. S. T. M. P.M. High, 90 0 to 3700 C." or "A. S. T. M. P.M. High,
200 0 to 700 0 F." according to the type of thermometer.
thermometer increases not less than 9 nor more than 11°F. per minute. The
stirrer shall be turned at a rate of from 1 to 2 revolutions per second.
(f) Application of the test flame shall be made at each temperature reading
which is a multiple of 2° F. up to 220° F. For the temperature range above
220° F., application 'shall be made at each temperature reading which is a
multiple of 5° F. The first application of the test flame shall be made at a
temperature at least 30° F. below the actual flash point. Application of the
test flame shall be made by operating the device controlling the shutter and
test flame burner so that the flame is lowered in one-half seco:q.d, left in its
lowered position for one second, and quickly raised to its high position. Stirring/
shall be discontinued during the application of the test flame.·
Flash Point.-The flash point is taken as the temperature read on the
thermometer at the time of the flame application that causes a distinct flash in
the interior of the cup. The true flash must not be confused with the bluish
halo that sometimes surrounds the test flame for the applications preceding the
one that causes the actual flash.
Barometric Pressure.-The barometric pressure shall be observed and re-
corded. No corrections shall be made except in case of dispute when the flash
point figure shall be correct"ed according to the following rule:
For each inch (25 mm.) below 29.92 in. (760 mm.) barometric reading add
1.6° F. to the flash point.
For each inch (25 mm.) above 29.92 in. (760 mm.) barometric reading
subtract 1.6° F. from the flash point.
Determination of Flash Point of Cut-Back Asphalts and Other Viscous
Materials and Suspensions of Solids.-The apparatus shall consist of the
following:
(a) Pensky-Martens tester as described above except that the stirrer shall
be mechanically operated to stir in a down{l,ard direction at a speed of 70 to
80 r.p.m ..
(b) Low-range thermometer as specified in Table 1.
Procedure.-The material to be tested and the tester shall be brought to a
temperature 30° F.lower than the estimated flash point. The air space between
the cup and the interior of·the air bath shall be completely filled with water at
the temperature of the tester and material. The temperature shall be raised
throughout the duration of the test at a rate of not less than 4 nor more than
6 F. per min. With the exception of this requirement for rate of heating and
the rate of stirring of 70 to 80 r.p.m., specified in Section 6 (a); the procedure
shall be the same as that prescribed in the Standard MethOd of Test for Flash
Point by Means of the Tag Closed Tester (A. S. T. M. Designation: D 56) of
the American Society for Testing Materials. 54
Accuracy.-Duplicate tests should not differ by more than 5° F.
64 1936 Book of A. S. T. M. Standards,. Part II.
1752 FIXED OILS, FATS AND ~AXES
Inches Centimeters
Dimensions
Milli- Nor- Maxi- Mini- Nor- Maxi-
l11um mal mum mum mal mum
- -- -- - - -- -
Depth of water surface below top of cup .... li.- la". 1';- 2.74 2.78 2.82
Depth of oil surface below top of CUp ....... 1';- II:. III
64 2.90 2.94 2.98
Depth of top of bulb of oil thermometer when
in place below top of CUp ............... 1"" If. III32 3.25 3.33 3.41
Inside diameter of oil cup at top ........... 2l.h- 2Ys 2l'o"li 5.39 5040 5.41
5
Diameter of bead on top of cover .......... -II 32 it 0.36 0040 0.44
Weight of oil cup ........................ - - - - 67 to 6~ g.
"
The plane of underside of cover shall be between the to.p and bottom of the
burner tip when the tip is fully depressed. 'I .
Thermometer.-Two thermometers ar,e required, one for measuring the
temperature of the sample under test, the other for measuring the temperature
of the bath. Each of these thermometers shall conform to the following re-
quirements. These specifications cover a special thermometer graduated in
either Centigrade or Fahrenheit degrees as specified, the ranges being _7°
to +110° C. or +20 0 to +230° F., respectively:
TYPE: Etched stem, glass.
LIQUID: Mercury. -
RANGE AND SUBDIVISION: _7 0 to +110 0 C. in 0.5 0 C. or +20 0 to +230 0 F. in 1 0 F.
TOTAL LENGTH: 273 to 277 IDID. (10.75 to 10.92_in.).,
66 Under the standardization procedure of the Society, this method is under the
joint jurisdiction of A. S. T. M. Committee D-I on Paint, Varnish, Lacquer, and Re-
o lated Products, and Committee D-2 on Petroleum Products and Lubricants. Repro-
duced through the courtesy of the American Society for Testing Materials, 260 S. Broad
St., Philadelphia, Pa.
66 This method is issued under the fixed designation D 56; the final number indicates
the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as,Tentative, 1918; Adopted in Amended:·Form, '1919; Revised, 1921,
1936. . .-F-
67 1936 Book of A. S. T. M. Standards, Part II, p. 897.
FIXED OILS, FATS AND WAXES 1753
in the burner tip, a small quantity of cotton waste placed in the oil chamber to
which the burner tip is attached and the chamber filled with signal, sperm, or
lard oil. An alcohol lamp may be used for heating the water bath as a substi-
tute for gas.
The water bath thermometer shall be place4 in the collar provided for it
and the bath filled. with water until it overflows. The temperature of the
water in the bath shall be such that when testing is started it will be at least
20 0 F. (11 0 C.) below the probable flash point of the oil to be tested.
The oil cup shall be placed in its proper position in the water bath and 50 ml.
of the oil to be tested shall be measured into it, using an accurate graduate or
other measuring device for the purpose. The temperature of the oil shall be '
at least 20° F. (11° C.) below its probable flash point when the t~st is started.
Air bubbles on the surface of the oil shall be destroyed, and the cover with the
flash point thermometer in place shall then be properly attached to the bath
collar. The test flame shall be, lighted, the flame being adjusted to the size
of the small white bead on the cover. '
'Che gas burner or alcohol lamp shall be centrally placed in the base of the
tester and lighted. The flame shall be so adjusted that the temperature of the
oil in the cup rises at the rate of 1.8° F. (1°C.) per min. as closely as possible,
but in any case not faster than 2° F. (1.1 0 C.) nor slower than 1.6° F. (0.9° C.)
per min.
(a) The barometric pressure shall be recorded. If a barometer is not
available, the figure may be obtained from the nearest Weather Bureau Station
and an appropriate correction made -for difference in altitude between such
station and the laboratory.
(b) The initial temperature o(the oil shall be recorded.
(c) When the temperature of the oil is 9° F. (5° C.) below its probable Hash
point, the knob on the cover shall be turned in such a manner as to introduce
the test Hame into the vapor space of the cup, and immediately turned back
again. The time consumed in turning the knob down and back shall be about
one full second, or the time required to pronounce distinctly the words" thous-
and and one."
Cd) The time at which the first introduction of the test flame is made and
the temperature of the oil shall be recorded.
(e) The application of the test flame shall be repeated after each 1° F.
(0.5° C.) rise in temperature of the oil until a distinct flash in the interior of the
cup is observed. The true flash must not be confused with the bluish halo
which sometimes surrounds the test Hame during applications immediately
preceding the actual flash.
(f) The time and the temperature of the oil when the flash point' is reached
shall be recorded.
Repeat Tests.-(a) If the rise in temperature of the oil from the time of
making the first introduction of the test flame to the time at which the flash
point is observed was more rapid than 2° F. (1.1° C.), or slower than 1.6° F.
(0.9° C.) per min., the test shall be repeated, adjusting the gas burner or alcohol
lamp to the proper Fate of heating.
(b) It is not necessary to turn off the test flame with the small regulating
valve on the cover; it may be left adjusted to the proper size of flame ..
(c) After completing the preliminary test to determine the approximate
1756 FIXED OILS, FATS AND WAXES
flash point, the burner or lamp shall be removed, the oil cover lifted and the
thermometer bulb carefully wiped off. The oil cup sha:ll be removed, emptied,
and carefully wiped until dry.
(d) The temperature of the bath shall be lowered py the addition of cold
water until it is 15° F. (8° C.) below the flash point of, the oil as shown by the
preliminary test.
(e) The oil cup shall be replaced and a fresh 50-m!. sample measured into it.
The test procedure, as described above, shall then be repeated, introducing
the test flame for the first time, however, when the oil temperature is 10° F.
(5.5° C.) below the flash point obtained in the preliminary test.
(j) Oil which has once been subjected to the flash test shall be discarded.
(g) If test is to be repeated, a fresh sample shall be used.
Average Value of Flash Point.-If two or more determinations agree within
1° F. (0.5° C.), the average of these results, corrected for barometric pressure:
shall be considered th~ flash point. If two determinations do not check within
1° F. (0.5° C.), a third determination shall be made and, if the maximum varia-
tion of the three tests is not greater than 2° F. (1 ° C.), their average, after cor-
recting for barometric pressure, shall be consid\lred the flash point.
Correction for Barometric Pressure.-Correction for barometric pressure
shall be made only in cases of dispute or when the barometer reading varies
more than 7'2 in. (13 mm.) from the standar.d pressure of 29.92 in. (760 mm.).
When the barometer reading is' below this standard pressure, add to the ther-
mometer reading 1.6° F. (0.9° C.) ,for eacn1 in. (25 mm.) of barometer differ-
ence to obtain the true flash point. When the barometer reading is above the
standard pressure, deduct 1.6° F. (0.9° C.) for each. 1 in. (25 mm.) of barometer
difference to obtain the true flash point. '
quantitative" filter paper and the precipitate washyd with water, first ,by
decantation, then on the filter, till free from chloride. ' The paper and precipi-
tate shall be transferred to a suitable weighed crucible, dried at low heat, till
moisture is evaporated, the paper charred (wit4out flaIhing), and finally ignited
at a good red heat till the precipitate is just burned whi,t~. A satisfactory means
of accomplishing these operations is to place the crucil?le containing the wet
filter paper in a cold electric muffle furnace and to turn on the current. Drying,
charring, and ignition will usually occur at the desired rate. .
(b) After ignition is complete, the crucible shall be allowed to cool to room
temperature and weighed. The use of a desiccator is not recommended.
. (c) From the increase in weight of the crucible the percentage of sulfur
shall be calculated from the formula:
_ grams of BaS04 X 13.734
P ercent age 0 f suIf ur- f'l d .
. grams 0 01 use
Accuracy.-The percentages of sulfur obtained by the same operator with
the same apparatus should not differ by more than 0.02+0.04 A; while the
percentages obtained by different operators should not differ by more' than
0.02+0.08 A (A represents the average of percentages obtained). •
The Jacksqn Turbidimeter can be use,d for rapid work.
Water and Sediment are determined as follows:
"
Water (A. S. T. M. Method D95-30) 60
A_ ......._.•..• ,.•••
45 foSSmm. •
8-(In6{d, Dlam,hr; 14 ., 16m",. -
Ii"'" 21.'
if:""
C-(llI6ldeJJlam,fll')12 " 16mm. _ 1.1.."
2
l'
•
D--··_·······.··- ISO "170mm._ 6 , .. flo
E----···· .......·..--.25" 38mm._ I"" 1£-
F- _····-·_········-100 "'08mm. _ 4" ., 4J'
H---••~--·-- .... -·-. If ,. 19 mm.. _ ~"" .I-
FIG. 253.-Apparatus for Determination of Water.
A. S. T. M. PE:AR-S:S:~PE1> TUBE
A. S. T. M. CONE-SHAPED TUBE
0
Range
0.05 m!.
Limit of Error
0.02 mI.
I. _, _NUmbered
The water or oil bath shall be of sufficient depth for immersing the centri-
fuge tubes in a vertical position to t1e 100-ml. mark.- Means shall be provJa~d
for heating this bath to 120 0 F. (49 0 C.). .-
..
FIXED OILS, FATS 4-ND WAXES 1763
Water by Centrifuge
Apparatus.-The apparatus shall be the same as ill the preceding method.
Procedure.-Exactly 40 ml. of 90% benzol shall be measured into each of
two centrifuge tubes and exactly 20 ml. of the oil to be tested shall then be
added to each. The centrifuge tubes shall be tightly stoppered and shall be
shaken vigorously until the contents are thoroughly mixed.
The t'Wo centrifuge tubes shall then be placed in the centrifuge on opposite
sides and shall be whirled at a rate of 1400 to 1500 r.p.m., or the equivalent;
for 10 minutes. The volume of water at the bottom of each tube shall be
read and recorded, estimating to 0.1 ml. if necessary. The centrifuge tubes
shall t,hen be replaced in the centrifuge, again whirled for 10 minutes as before,
and removed for reading the volume of water as before. This operation shall
1764 FIXED OILS, FATS AND WAXES
be repeated untii the volume of water in each tube 'remains constant for three'
consecutive readings.
The volume of water shall be read on each tube, estimating to 0.1 ml. if
necessar~, and the average perce~tage of water in th~ sample c~lculated.
Calonfic Power. 53-The calorIfic power of ·uncracl.{ed fuel OIl'can be found
by the formula, Cal. Power=17660+(69XA.P.I. g~avity, 0 Baume). Of a
cracked fuel oil, Cal. Power = 17780+(54XA.P.I. gravity, 0 Baume).
The tests most commonly employed for the identification of these oils are as
follows: specific gravity, refractive index, Valenta test, elaidin test, Maumene
test, iodine number, and saponification value.
In addition, certain special and commercial tests are applied, as Bechi test,
Baudouin test, free acid, spontaneous combustion, and drying test.
Specific Gravity.-This is usually determined either by the Westphal
balance (page 1704) or by the picnometer. _
A two-necked flask of 50 ml. capacity, having a thermometer carefully
ground into one neck, the second one being a narrow tube bearing the mark, is
most suitable. This is filled with the oil to be examined, cooled to 15.5 0 (;.,64
the excess of oil removed and weighed. If the weighings be made to 0.5 milli-
gram and a correction applied for the expansion of the glass-1>y the difference in
temperature = 15.5-4= 11.5 0 = -0.025% of the va~ile obtained, the deter-
mination is accurate to 0.00002.65
For the determination of the specific gravity of small quantities of oil,
satisfactory results can be obtained by weighing 1 or 5 ml. of the oil carefully
measured from an accurately calibrated pipette. Or a mixture of alcohol and
water can be made until a drop of oil will stay in any position in it, and its
specific gravity determined. ... ~
Refractive Index.-This is of the same value as the determin~tioh of specific
gravity: it has, however, the advantage that it is more rapid and uses only one
or two drops of the oil.
The apparatus preferably employed is the ADM refractometer, Fig. 256, the
prisms of which are kept at constant temperature, Ui;mally 25 0 C., by cu-culating
water. .'
The illuminating mirror should light the cross h~irs and the telescope should
be sharply focused on them.
63 Faragher, Morrell & Truax, Ind. Eng. Chern., 21, 933 (1929),
64 Allen (Organic Analysis, 33) states that a correction of 0.00064 can be made for
each variation of 1 C.
0
~ '.-
56 Wright, J. Soc. Chern. Ind., 11, 300, 1892.
FIXED OILS, FATS AND WAXES 1765
The double prism is opened by means of the screw heads, and after carefully
cleansing the prisms with cotton and ether, a drop or two of the oil placed on
the horizontal surface of the fixed prism. The prisms are then tightly closed:
The telescope is brought into the position shown and the sector is firmly held
and the alidade (the moving part) moved forward until the field of vision shows
the boundary between li'g ht and shade just intersecting the cross hairs.
By means of the screw on the right
of the instrument this boundary line
should be made as sharp as possible.
The index of refraction is read off directly
from the sector, using a lens if necessary;
the reading is accurate to .0002.
After using, the prisms are again
carefully cleansed and a piece of filter
paper placed between them to pre-
vent them from being scratched. The
instrument is in correct adjustment when
water at 18° gives a mean reading of
1.333. The temperature correction for
oils and fats is 0.0004 for every degree
rise.
Valenta Test.--Although considered
by some to be unreliable, yet as the
indication given by this test may be of
value, it is barely worth the trouble of
execution. It depends upon the solu-
bility of the oil in glacial acetic acid.
Enough oil is poured into a test-tube
to fill it to the depth of about 1 in.,
the exact height being marked by the FIG. 256.-Refractorneter.
thumb; an equal quantity of glacial acetic
acid is poured in, that is, until the acid reaches the point indicated by the thumb.
A light thermometer is placed in the tube, and it is heated until the oil dissolves
-shown by the liquid becoming homogeneous. The tube is now allowed to cool,
and the point noted at which it begins to become thoroughly turbid.
Castor oil is soluble at ordinary temperatures, while rape-seed and other
cruciferous oils are usually insoluble even at the boiling-point of the acid. The
tetnperat ures at which other oils become turbid are given on pages 1795 and
1796. .
Elaidin Test.-Although this is not a quantitative test, yet its ease of ap-
plication and the conclusions which may be drawn from it render it valuable.
It depends upon the change of the liquid olein into its solid isomer ellllidin, and is
especially applicable to olive and lard oils.
M anipulation.- Five grams of the oil are weighed 67-within 2 drops-into a
cordial glass, 7 grams of nitric acid, specific gravity 1.34, are then weighed into
it,67 and two pieces of copper wire (0.6 to 1.0 gram) added. Place the glass in a
pan of cold water at about 12° C., and stir with a short glass rod about 20 to 30
"6Valenta, Dingler Polyt. J., 253, 418; also J. Soc. Chern. Ind., 3, 643, 1884.
67 Not on the analytical balance.
1766 FIXED OILS, FATS AND ,WAXES
turns, not only with a rotary movement, but also with an up-and-down motion,
so as to mix the oil and the evolyed gas thoroughly.1 When the wire has diS"-
solved, add a second piece and stir as before. Th~s second addition should
furnish gas enough if the liquid has been kept cool ~nd the stirring has been
thorough. \ ,
At the end of the first hour, pure lard oil will usually show flakes of a wax-like
appearance, and upon standing without disturbance and at the same tempera-
ture for another hour, the oil will have changed to a solid white cake hard
enough to bear several ounces' weight, or admit of lifting the glass and 'contents
by the glass rod. . .
Most of the fish and seed oils yield a pasty or buttery mass separating from a,
fluid portion, whereas olive, almoIld, peanut, lard, sperm and sometimes neat's-
foot oil, yieid a solid cake. "
Instead of using nitric acid and copper, sulfuric acid of 46° Baume, c(\Il-
taini,ng a little nitric acid and saturated at 0° C. with hitric oxide, may be
employed.
A test should always be made at the same time with an oil of undoubfe'd purity.
, ~ ,
NOTEJs.-If the oil be stirred too much or too' frequently, or is too warm, it has no
opportunity to form a hard cake.
Hubl states that all attempts to make the'test a quantitative one have resulted
in failure.
Mercury can be used instead of copper.
Cailletet's method,68 in which a smaller quantity of oil is used, and sulfuric and
nitric acids allowed to act upon it in a boiling water bath, cannot, in the experience of
the writer, be depended upon to give reliable results. /
Maumene Test. 69-While this, like the'preceding, is not a quantitative t~st~
yet the indications afforded by it are of more value in many cases than those
obtained by, quantitative methods, as, for example, the saponification value.
It depends upon the heat developed by the mixing of the oil with strong sulfuric
acid. This takes place in a small beaker 7Y:; to 9 cm. deep and of 150 ml.
capacity, packed in an agate-ware cup with dry felt br cotton waste packing.
Manipulation."":Fifty grams of the oil are weighed 70 into the beaker to
within 2 drops, and its temperature noted by a thermometer. Ten ml. of
sulfuric acid are now run gradualfy into the oil':__allowing the graduate to drain
five seconds-the mixture being stirred at the same time with the thermometer,
and the stirring continued until no further increase in, temperature is noted.
The highest point at which the thermometer remains "Constant for any ap-
preciable time is observed, and the difference between this and' the initial tem-
perature is the" rise of temperature." This varies with the strength bf the
acid employed, and. to secure uniformity 7'1 the, results should b~ expressed by
dividing the rise of temperature with the oil by the rise of temperature with
water, and multiplying by one hundred. This is c,alled'the " specific tempera-
ture reaction." The rise of temperature with watkt is determined in the same
manner as with oil, using the same vessel. "
~ ?;
68 Milliau, J. Am, Chem. Soc., 15, 156, 1893.
69 80 2Cl gives similar results.
70 Not on the analytical balance. r '_
n Tortelli, J. Soc. Chem. Ind., 23, 668, 1904, is una~e to secure uniformity in ~his
way.
FIXED OI~, FATS AND WAXES 176'7
.. ( NOTEs.-In performing this, test it is important that the oil and acid be of the
same ·temperature, attained by keeping them beside each other. •
The strength of acid should be as far as possible the same; it should be determined
not by specific gravity, but by titration, as 100% and 94.3% acid have the same specific
gravity:' .,
For,concordant results the conditions should be the same, and the same apparatus
sho}lld"p€! used. In case the test is to be applied to a drying oil, it should be diluted
pne-half with a mineral oil, 25° paraffin, for example, thoroughly mixing them. The
"''rise of temperature" is then, the rise of temperature of mixture minus half the rise
of temperature of 50 grams of mineral oil, multiplied by 2.
It is advisable to make a test at the same time with an oil of known purity. Re-
sults should agree within 2%. By the use of the Hubl formula, page 1769, substituting
thermal va,lues, res,ults comparable with those obtained with the iodine value can be
obtained.
Sherman, Danziger, and Kohnstamm 72 have studied this method with the idea
of eliminating the errors. Rather than dilute, the oil with a mineral oil they dilute
the ·acid, using one of 89%. The results obtained are a little lower for vegetable oils
and a little hjgher for animal oils than those usually'found with the strong acid as
employed by Thomson and Ballantyne. Mitchell 7~ uses an inert diluent-carbon
tetrachloride-in a vacuum-jacketed' tube and one-fifth the quantities; all oils are
diluted. He·finds that the;results obtained'are"in close agreement with the bromine
thermal vaJues; further, that the test filll,y b,e of use in determining the degree of oxida-
tion of fats and oils, the figures becoming greatE!r with the age of the oil.
Data upon various oils will be found on pages 1795-1796. '
(t ~, i ,~ , '!-'
"References
·'t Mauril;r:te, Compt.-Re~d.) 35, 572/1852.
Ellis, J! Soc. Chem. Irid.'; 5,'361, 1886.
. Thomsoru:and Ballantyne, J:" Soc. Chem. Ind:; 10, 234, 1891.
",I, Richmpnd, An~y.st, UJ, 58, 1895. "., '
Munr~~I1' Am. Pub~,Hlaltli Alli?)}O, '236, 1884.
Eight drops are approximately 0.2 gram. The beaker is replaced in the watch-
glass and the system again weighed, the difference in ,weight being the amount
of oil in the bottle.
The oil is dissolved in 10 m!. of chloroform, 30 fil. of the iodine solution
(Reagents) added-best from a burette-and allowed: to stand with occasional
shaking for exactly thirty minutes; with oils of an iop.ine number of less than
100, ten minutes suffices; 10 m!. of potassium iodide solution are added; shake
thoroughly, add 100 ml. water, washing down any iodine on the stopper. Ti-
trate the excess of iodine with N /10 sodium thiosulfate, adding the latter
gradually, with constant shaking, until the yellow color of the iodine has almost
disappeared. Add a few drops of starch solution, shake thoroughly to dissolve
out the iodine from the chloroform solution, and titrate to the disappearance
of the blue color. The blue color returns after a few minutes but the end-point
is that of its first disappearance.
At the same time at which the oil is prepared, two " blanks" should be
prepared similarly in every way to the actual tests, except in the addition of
the oil, and treated in every respect like them; the strength of the thiosulfate
solution should also be determined the same day on which this test is carried out.
Standardization of the Thiosulfate Solution.-Ten ml. of potassium iodide
and 100 m!. of water are poured into the' Erlenmeyer flask; 20 m!. of the bi-
chromate solution, equivalent to 0.2 gram of iodine, are now measured in with
a pipette, and to this 5 m!. of concentrated hydrochloric acid added and the
mixture shaken for three minutes. It'is now titrated with the thiosulfate
solution until the yellow color of the iodine has almost disappeared; starch
paste is now added, and the titration continued until the deep-blue color of the
solution changes to a sea-green-due to CrCla,-which is usually brought about
by the addition of a single drop. •
The reactions involved are:
K2Cr207+14HCl=2CrCla+2KCl+7H20+3Clz;
3CI2+6KI=6KCI+3I 2; I .
Ingle 90 has shown that the free acid formed during the process is due to·th.e
action of water upon the iodochlorides. Some of these are reduced by potas-
sium iodide with liberation of iodine and consequent reduction in the iodine
absorption. Iodine chloride is the active agent, and not hypoiodous acid.
For the calculation of the percentage of adulteration of one oil by another,
Hubl gives the following formula: 91
" Let x = percentage of one oil and y = percentage of the other oil, further,
18 J. Soc. Chem. Ind., 2, 435, 1883.
19 Dingler Polyt. J., 253, 281; also J. Soc. Chem. Ind., 3, 641, 1884.
..0 J. Chem. Ind., 11, 183, abstr., 1892.
81 Ibid., 13, 1100, abstr., 18.94.
82 J. Soc. Chem. Ind., 14, 130, 1895.
83 Ibid., 12, 717, abstr., 1893.
84 Chem. Ztg., 19, 1786, 1831, 1895.
85 J. Soc. Chem. Ind., 14, 1031, 1895.
86 J. Soc. Chern. Ind., 12,717, abstr., 1893.
87 Ibid., 14, 1031, 1895. .
88 Holde, Mitt. kgl. Techn. Versuchs., 9, 81, 1891.
89 Dieterich, J. Soc. Chern. Ind., 12,3&1,1893.
90 J. Soc. Chern. Ind., 21, 587, 1902.
91 Dingler Polyt. J., 253, 281, 1884.
1770 FIXED OILS, FATS AND W.AXES I'
m = iodine value of pure oil x, n of pure oil y, and I of the sample under ex-
amination, then . . j
100([-n) "
x=--'-----'-
m-n
\
He further states that the age of the oil, provide~ it be not rancid or thick-
ened, is without influence on the iodine value. Ballantyne 92 'finds that light
and air diminish the iodine number.
As might be expected, the iodine value is inversely proportional to the cold
test.
The method, as will be seen, is a conventional one, and the best results will
be obtained by using measured quantities of reagents and 'carrying through the
process in 'the same manner every time. 93 , ,) ")',. ' ...
The calculation is perhaps most easily made 'as follbws: Subtract the number
of inl. of thiosulfate used for the titration of the oil, from that obtained by
titrating the blank-this gives the thiosulfate equivalent to the iodine absorbed
by the oil. Multiply this number (of ml.) by the value of the thiosulfate in
terms of iodine, and the result is the number of grams of iodine absorbed by the
oil; this divided by.the weight of oil used and multiplied by 100 gives the iodine
number. • '
In case it be desired to recover the iodine used" reference may be had to an
article by Dieterich, abstracted in the Jour. Soc. Chem. Ind., 15, 680, 1896.
Oxidized Oils. Iodine Number of •.L-To find the original iodine number of a..
semi-drying or non-drying oil which has been altered by atmospheric oxidation,
add 0.8 to the iodine number found on the altered sample for each increase of
' h
O•001 10 'fi ' . (k 15.5° C.
t e specI c gravIty ta en at 15.50 .C.
)94
.
;
Bromine Number or Value.-The iodine method just described has, among
oth~rs, the disadvantllge .that it fails to distinguish between addition and sub-
stitution; this is sometimes of importance, and to accomplish it McIlhiney 9.
makes use of the bromine absorption. I
METHOD 98
Petrolic ether may be used with compounded cylinder oils in the quantity
of 50 mI., providing a Soxhlet extraction flask is used to collect this petrolia
ether periodically. The Soxhlet flask should be so adjusted, through the
addition of glass rod ding or beads when necessary, that it willjust overflow
with the full quantity of petrolic ether.
The percentage of fatty oil (or fat) in a compounded petroleum product
can be calculated from the saponification nl!mber of such a product only
when the saponification number of the fatty oil is known. If the saponification
numbers of both fatty oil and compounded oil are known, the following formula
should be used: '
'1
P ercen t age 0 f fa tt y 01 = 100 X saponification number of compounded oil
.
saponification number of fatty 011
For this determination. the following values of saponification number may
be used:
FIXED OILS, FATS AND WAXES 1773
Saponi- Saponi-
fication fication
Fatty Oil Number Fatty Oil Number
Lard oil. .................... 192-198 Soya bean ................... 189-197
Tallow ..................... : 193-198 Peanut ......... , ............ 18{)-197
N eat's-foot. . . . . . . . . . . . . . . . . .. 193-204 Cottonseed ................... 191-197
Fish ......................... 140-193 Blown. rapeseed ............... 195-216
Sperm ....................... 120-140 Blown cottonseed. . . . . . . . . . . .. 210-225
Castor ....................... 176-187 Degras ......................• 110-210
Rapeseed .................... 170-179
NOT:E1s.-Many prefer to cork the flasks tightly and tie down the stoppers, thus
saponifying under pressure; others make use of a return flow condenser, oftentimes
merely a long glass tube. . '
McIlhiney 100 has applied the process to dark-colored substances by making use
of the principle that when ammonium chloride is added to a neutral soap solution,
and the mixture distilled, the amount of ammonia freed is e,quivalent to the quantity
of alkali combined with the fatty acids. As a description of the process is beyond
the scope of the present volume, reference must be had to the original article.
As ordinarily prepared, the alcoholic potash solution turns rapidly reddish brown,
so that it is very difficult to note the end-point. This trouble can be partially avoided
by "adding a drop or two of the'solution to the diluted indicator contained upon a tile
after the manner of the titration of iron by bichromate. As the color is probably
due to the polymerization of the aldehyde formed by the oxidation of the alcohol, it is
more satisfactory to use for the preparation of the potash solution an alcohol which is
practically aldehyde free. Alcoholic potash made up from this, using the so-called
"potash by alcohol," will give a solution which will remain water-white for weeks.
"Thymol blue" may be used as an indicator.
The writer has found, if the stock solution be kept under an atmosphere of hydrogen,
that the coloration by standing is almost entirely prevented.
This method has also been refined by Gill and Simms, J. Ind. & Eng. Chern., 13,
547 (1921) so that a single drop, 0.03 gram, suffices for a test.
and allowed to stand. The gasoline layer should separate clearly and quickly
from the soap solution and the latter is drawn off;' the; gasoline is washed 2 or
3 times with 50% alcohol to extract Any soap, and these washings added to the
soap solution. This latter is extracted, until upon evaporation the gasoline
leaves no stain upon paper, carJfbeing taken to 'wash t~e gasoline extracts each
time with 50% alcohol; three extractions with gasoline are usually sufficient.
The gasoline is distilled from these extracts, the residue heated until the
gasoline odor disappears, and weighed. From the appearance of the residue
some idea of the kind of unsaponifiable matter can be, obtained. This in th~ .
case of sperm oil will be mainly solid "alcohols, probably of the ethylene series.
According to Schicht and Halpern 103 this method is open to the following
errors: incomplete saponification, -incomplete extraction, solubility of soaps'
in the solvent, and the solubility of the unsaponifiable matter in the washing
solution. Their improved. method is as follows: 5 grams of fat 104 with 3
grams of solid caustic potash dissolved in a little water and 25 m!. of absolute
alcohol are boiled half an hour under a reflux condenser. After cobling 25 m!.
of 10% KCI are added and the solution is then shaken four times with 200 ml.
of petroleum eth~r distilling under 60°. The 'petroleum ether is evaporated
and, without washing, the residue is dissolved in 25 ml. absolute alcohol and
the solution made slightly alkaline with normal alkali; 25 ml. of 10% KCI are
added and the shaking witl;l petrole~m ether repeated. The petroleum ether
solution is shaken with 1QO mI.. of 50% alcohol and the wash solution with
100 ml. petroleum of ether, which is afterwards washed with 100 ml. of 50%
alcohol. After combining the extracts the 'J?etroleum ether is driven off and
the residue dried and weighed. .
NOTE.-Care should be taken to use gasoline which leaves no residue on evapora-
tion at 100 C.
0
- r
These may be separated by boiling for two 'hours with an equal q~~~tity of
acetic anhydride; if the substance dissolves and does not precipitate on cooling,
higher alcohols are indicated; if a mass.,of crystals separates out on cooling,
cholesterol and its isomers, or a mixtunt of these with the, higher al~9hols is
indicated; if an oily layer remains on tOPidt;is IUl indication of the presence of
paraffin or ceresenc. For the complete separation and identification of these
reference must 1?e had to Lewkowitsch, ".Analy,sis of Fats, o.ils, and W~~es,"
as .it-is beyond the limits of this chapter.. . J:': ':
h
Test for Animal or Vegetable Oils.~Animal oils contain cholesterol,
C26H 4S OH, while vegetable oils contain the isomeric body phytosterol; hence
t4e isolation and iden~ification of these compounds enables one to say with
certainty as to the presence of one class of oil or the other-for example as-to
the presence of fish oil in linseed. The quantity of these bodies varies from
0.2' to 1%. The method is essentially that of Bomer .106 Fifty grams of the
oil are boiled in a flask 'with a return cooler with 75 ml. of 95% alcohol for five
minutes and the alcoHolic solution separated; this is repeated with another
portion of alcohol. The alcoholic solutions .are mixed with 15 ml. of 30%
Rodium hydroxide and evaporated on a water bath nearly to' dryness in' a
porcelain dish and the residue shaken out with ether. The ether is evap-
orated, the residue taken up with a little ether, filtered, again evaporated,
dissolved in 95% alcohol (by volume), and allowed to crystallize slowly,
DO'mer states that the form of the crystals is more to be relied upon tljan.3.
determination of their melting-point! Cholesterol crystallizes from alcohoL'or,
ether in leaflets or rhomboid tables containing one molecule of water of crys-
tallization. Phytosterol crystallizes also from alcohol with one \llolecule of
water in needles forming stars or bundles. As a further means of identifica-
tion, some of the esters should be made and their melting-points determined.
To this end the crystals above obtained are heated over a low flame in a
small porcelain dish covered with a watch-glass, with 2 or 3 ml. of acetic or
other acid anhydride until it boils: the watch-glass is removed and the excess
of anhydride evaporated on the ·water bath. The contents of the dish are
treated with a small quantity of absolute alcohol to ,prevent crystallization,
more alcohol added and the solution allowed to crystallize. The .crystals are
filtered off through a very small filter, washed with a small quantity of 95%
a:1cohol, dissolved in absolute alcohol, and recrystallized until a constant melt-
. ini\'7oint is obtained. .
The following table shows the corrected melting-points of these alcohols
and their esters:
106 J. Soc. Chern. Ind., 17, 954, 1898; Tolman, J. Am. Chern. Soc., 27, 590, 1905;
Tolman, Bull. 107, U. S. Dept. Agriculture, 1907.
1776 FIXED OILS, FATS AND WAXES
Cholesterol Phytosterol
Alcohol. ......................... 148-150.8° 136-143.8°
Acetate .......................... 113-114° 120-137°
Benzoate. . . . . . . . . . . . . . . . . . . . . . .. 135-151° 142-148°
Propionate. . . . . . . . . . . . . . . . . . . . . . . 97-98° 104-105°
NOTEs.-Some directions state, in isolating the cholesterol or phytosterol, to boil
with the 30% sodium hydroxide until one-fourth of the alcohol is evaporated. As a
result of repeated experiments this has been found to cut down the. yield so much that _
on a large scale 'Practically none of these bodies, particularly phytosterol, was obtained.
This agrees with the observation of Lewkowitsch that by heating cholesterol with normal
alcoholic potash, cholesterin hydrate- is obtained.
The following test will serve to differentiate between cholesterol and phytosteroP07
A very small quantity of cholesterol is warmed with 1.5 ml. absolute alcohol and a
trace of isodulcit or rhamnose (o-dimethylfurfural) added. After cooling,· an equal
volume of concentrated sulfuric acid is added, so as to form a layer below the solution;
whereupon a raspberry-colored ring is produced at the zone of contact of the two liquids.
On mixing the layers while the tube is cooled in a current of cold water, the mixture
becomes intensely colored. With phytosterol the reaction fails or at most a. pink color.
Similar reactions are given by abietic acid. Or mix the pure unsaponifiable matter
with an equal quantity of pure cholesterol; if the melting point be unchanged, the un-
saponifiable matter is cholesterol; lowering of the melting point indicates phytosterol.
As little as 1 % pf cotton-seed has been found in lard, and 4% in any oil has been
detected by this test.
For the means of distinguishing between drying and marine animal oils, see Halphlm,
J. Pharm. Chim., 14,391 (1901), abstracted J. Soc. Chem. Ind., 21, 74, or Chem. Cen-
tralb., 72, ii, 1097 and 1323.
/'
FIXED OILS, FATS AND WAXES 1777
peculiar property. ! . ,
By purifying the acids by the lead salts TO.ftelli and Ruggeri 117 are able to detect
as little as 10% of heated cotton~seed oil.
It is to be noted that pure lard, tung and ~ll've oil are not infrequently met
witllw'hlch give the test, consequently its indication~ cannot be considered as
conclusive. ' ';
Halphen's Test for Cotton~seed OilY5-This depends upon the observation
that this oil contains an unsaturated fatty acid which combines with sulfur,
giving a colored compound_I19 I
Procedure.-Ten ml. of the oil or melted fat are heated, in a 'large test-tube
with a; long' glass condenser tube attached, with an equal volume of amyl
alcohol.and of carbon bisulfide solution of sulfur (Reagents), at first with
ffeq~ent. a~itat~on, in a steam b, ath, and ~hen, after t~e v~~lent boiling has
ceasen; 'iIi 'a Jmne bath (105- ~ 10 0 C.) for ,forty-five' mmutes ,to three hours,
according to .the quantity of ja!iulterant present, the tube being occasionally
removed and shaken. As little as 1 % will give a crimson wine'coloration in
twenty minutes.120 >, ,
_(. . \ ~
Bailey and Baldsiefen, J. 1. and E. Chem., 12, 1189 (1920), have devised a modifica-
tion which is more elaborate and may give more accurate results.
nected to tb_fl. _water bath it can be used indefinitely without refilling and with-
out danger of 'burning out.
Seven grams of ordinary bleached cotton wadding or " absorbent cotton"
are weighed out in a porcelain dish or on a watch-glass, and 14 grams of the oil
to be tested poured upon the cotton and thoroughly worked into it, care being
taken to replace any oil that is lost. The cotton is then placed in the cylinder,
packed about the thermometer so that it occupies the -upper 4~ in. of the
cylinder, and put into the boiling water bath. After·the expiration of an hour,
the bath having been kept in active. ebullition, the temperature is read. Any
oil which shows a temperature exceeding 100 0 C. in one hour, or 200 0 C. in
two hours, should be regarded as a dangerous oil, or liable to produce spon- .
t.aneous combustion. The f,.Qllowing tables show the results obtained in uSIng
this apparatus. 143 '\7)' • cl
:, Temperature 0 C. in
,Oil
1 hr. 1)4 hrs. 172 hrs.
Besides being used for testing oils it can be applied to testing other mate-
,,..ials, oily waste, sawdust, or any mixtures suspected of causing spontaneous
combustion. ,
" The results 144 of the greatest practical value .obtained in the use of this
apparatus have been, first, determining the cause Memes; and, second, deter-
mining the degree of safety of the various oils used in manufacturing. Mineral
oil, as is well known, is not liable to spontaneous combustion; and a certain
percentage of animal or vegetable oil may be addeM to mineral oil without
materially increasing the danger under ordinary circumstances. This percent-
age varies according to the oil; with neat's-foot and first quality lard oil some
50 to 60% may be used, with cotton-seed not over 25% is allowable. The
143 Small percentages (0.5-2.0) of iron, sodium, magnesium or zinc salts 'in "red" or
olive oil may cause a degree of heating which may be dangerous. Gill, Ind. and Eng.
Chern., 16,23, 1924.
144 Richards, Tech. Quarterly, 4, 346,1891.
178~ FIXED OILS, FATS AND WAXES
claim so often made for so-called' safe' oils, said' to hav~ been changed by
special and secret processes of refining so as to be nQ longer dangerous, is easilY
exposed by this test."
Drying Test Upon Glass. 14 2_A few drops of oil are brought upon a glass
plate inclined at about 30 0 from the horizontal. A test of the oil is made from
time to time by touching it with the fingers, the time at which it does not soil
them being noted as the point when it is dry. Good oil should dry in three dtiys.
Archbutt 146 makes this test as follows: A piece of polished plate-glass 7 cm.
square by 4 mm. thick is cleaned and counterpoised on the balance; it is then
heated for an hour at 200 0 C. in an air bath to thoroughly dry it. It is taken
out, laid on a non-conductor, allowed to cool for three or four minutes, and
, the hot glass thinly painted with the oil to be tested by means of a camel's-hair
brush. When the glass is cold it is weighed. and sufficient oil added to make
it up to 0.1 gram. Two glasses are coated with the sample and two with a
standard oil, all placed on a level surface in a . large air bath at 50° C. and
heated for nine hours; one set of plates is withdrawn, cooled, and tested by the
finger. Good raw linseed is tacky, when tested by the finger when cold, in
nine hours and dry in twelve; corn oil is practically dry in fifteen hours, though
slightly tacky; cottonseed, partially dry in eighteen hours and fully dry in
twenty-one. Refined rape oil dried in forty-eight hours, and olive oil was
sticky after thirteen days.
Titer Test.-Under this rather misleading title is expressed the solidification
point of the fatty acids derived from a fat or oil; it has nothing at all to do with
titration, as might be expected. The test is extensively used for the evaluation
of fats, and according to the method provisionally adopted by the Association
of Official Agricultural Chemists is carried out as follows: 147
(a) Standard Thermometer. The thermometer must be graduated in tenth
degrees from 10° to 60°, with a zero mark, and have an auxiliarr reservoir at
the upper end, also one between the zero mark and the 10° mark. The cavity
in the capillary tube between the zero mark and the 10° mark must be at least
1 cm. below the 10° mark, the 10° mark to be aboujt 3 or 4 cm. above the bulb,
the length of the thermometer being about 15 lin. over all. The thermometer is
annealed for 75 hours at 45.0° C., and the bulb is of Jena normal 16 '11 glass,
moderately thin, so that the thermometer will be quick acting. The bulb is
about 3 cm. long and 6 mm. in diameter. The stem of the thermometer is
6 mm. in diameter and made of the best thermometer tubing, with scale etche~,
on the stem, the graduation to be clear-cut and distinct; but quite fine.
(b) Determination. Saponify 75 grams of fat in a metal dish with 60 ml.
of 30% sodium hydroxide (36° Baume) and 75 ml. of 95% (by volume) alcohol
or 120 m!. of water. Boil to dryness, with constant stirring to prevent scorch-
ing, over a very low flame or over an iron or asbestos plate. Dissolve the dry
soap in a liter of boiling water, and if alcohol ~as been used, boil for forty
minutes in order to remove it, adding sufficient I water to replace that lost in
boiling. Add 100 ml. of 30% sulfuric acid (25° Baume) to free the fatty acids,
and boil until they form a clear, transparent layer. Wash with boiling water
until free from sulfuric acid, collect in a small beaker, and place on the steam
J
1<5 Arosel, J. Soc. Chern. Ind., IS, 222, 1896.
1<6 Id., 18, 347, 1899. . ..../ ~
1<7 U. S. Dept. of Agriculture, Bureau of Chemistry Bulletin No. 107, p. 135, 1907.
FIXED OILS, FATS AND WAXES 1785
bath until the water has settled and the fatty acids are clear; then decant them
into a dry beaker, filter, using the hot-water funnel, and dry twenty minutes
at 100° C. When dried, cool the fatty. acids to 15 or 20° C. above the expected
titer and transfer to the titer tube, which is 25 mm. in diameter and 100 mm.
in length (1 by 4 in.) and made of glass about 1 mm. in thickness. Place in a
16-oz. salt-mouth bottle of clear&ass, about 70 mm. in diameter and 150 mm.
high (2.8 by 6 in.), fitted with a Mk, which is perforated so as to hold the tube
rigidly when in position. Suspend the thermometer, graduated to 0.1° C., so
·that it can be used as a stirrer, and stir the mass slowly until the mercur~
remains stationary for thirty seconds. Then allow the thermometer to hang
quietly, with the bulb in the center of the mass, and observe the rise of the'
mercury. The highest point to which it rises is recorded as the' titer of the
fatty acids.
Test the fatty acids for complete saponification as follows:
Place 3 m!. in a test-tube and add 15 m!. of alcohol (95% by volume).
Bring the mixture to a boil and add an equal volume of ammonium hydroxide
(0.96.sp.gr.). A clear solution should result, turbidity indicating unsaponified
fat. The titer must be made at about 20° C. for all fats having a titer above
30° C. and at 10° C. below the titer for all other fats.
REFERENCES
Hefter, G. Technologie der Fette, Oele, und Wachsarten des Pflanzen und Tier-
reichs. 4 volumes, 1906.
Ubbelohde, L. Chemie, Analyse und Gewinnung der Oele, Fette und Wachse.
4 volumes, 1908.
EDmLE FATS
the fat by repeated treatment with petroleum ether and again dry to constant
weight. The loss represents the amount of fat. Ignite the crucible with a low
flame or in a muffie, being careful not to volatilize tl;te salt, until a light-gray ash
is obtained. The loss represents curd and the re~~due ash. By extraction of
the ash with water, and neutralization witli calcium carbonate, the salt can be
titrated with silver nitrate. "
Examination of the Fat.-Butter is adulterated with oleomargarin~, reno-
vated butter, and cocoanut oil. The first may be detected by testing for
'cottonseed or sesame oil either by the color tests or by the index of refraction;
this at 25° is for butter 1.459-1.462, for oleo. 1.465-1.470. Owing to the fact
that, butter. contains a l~rge per cent of volatile fatty acids (butyric, caproic,
caprylic/and capric acids, in aU about 8%), adulterants may be detected by.
determining the aJ,llount of these. The process usually employed is that of
Reichert modifi!;ld by Meissl.
Five grams of the clear fa~; filtered through absorbent cotton, are weighed
into a 250-ml. round-bottomed flask and saponified by 2 ml. potassium hy-
droxide 1 : 1 and 10 ml.iof '95% alcohol, under a return flow condenser for
twenty-five minutes. Tlie'alcohol is rapidly evaporated off on the water bath
until no odor of alcohol is'perceptible. Add slowly 160 ml. of recently boiled
distilled water which has been cooled to 50° or 60°; warm the flask until a
clear solution of the soap is obtained. Cool to about 60° and add 8 ml. sulfuric
acid 1 : 4 to liberate the fatty acids. / Drop into the flask two bits of ~umice
(about the size j of peas) which have been heated and quenched in water, and
tie in a well-fitting cork; warm the flask until the fatty acids have melted and
are floating on the liquid. Cool to about 60° and attach the flask to a con-
denser, using a trap to prevent the sulfuric acid from being mechanically
carried over; 110 ml. are distilled into a graduated flask in as nearly' thirty
minutes as possible. Thoroughly mix the distillate, pour through a dry filter,
and titrate 100 ml. with N/l0 NaOH, using phenolphthalein as an indicator.
Multiply the ml. of alkali by 1.1 and calculate them to 5 grams of fat. The
Reichert-Meissl value for butter is from 24 to 34{ the average is about 28.8;
cocoanut oil gives 6-8 and other fats less than 1.
The procedure is a conventional one and should be followed as exactly as
possible. Cocoanut and other vegetable oils WQuld be shown by the fact that
the unsaponifiable matter would conta~n phytosterol; also by the Polenske
number.14s Renovated butter is best shown by the II'Spoon~ or Foam Test."
This consists in melting a third of a teaspoonful of the sample -in a tablespoon
over a small flame and stirring with a match. Increase the heat until the fat
boils briskly, stirring thoroughly several times. Oleomargarine and renovated
butter boil noisily, sputtering like a mixture 'of grease and water and produce
no foam. Butter boils with less noise and much foam sometimes rising over
the sides of the spoon. The 'pieces of curd in bJtter are much smaller than in
either of the others.
Preservatives, benzoic, boric and salicylic acids, may be examined according
to the procedure given in Woodman and Norton, " Air, Water and Food," pp.
154 and 196. ,- - '
148 Z. Nahr. Gerlussm., 7, 193, 1904, also Gill, liOn-Analysis."
FIXED OILS, FATS Al'fD WAXES 1787
Color may be dE;Jtected according to Allen, " COlI).lllercial Organic Analysis,"
4th Ed., Vol. II, or Leach's" Food Analysis."
Lard is adulterated with water, 25% being added in some cases, with cotton-
seed oil or·stearine and beef stearine.
Water is determined as with butter; cottonseed oil or stearine by the usual
tests. ,It should be borne in mind, however, that hogs fed on cottonseed meal
yield a lard which will give the Halphen test as strongly as if it contained 25%
of the oil. The iodine number and the presence of phytosterol will confirm this
test; the iodine number varies widely according to the source of the fat, but
in general it may be said it should be between 46 and 66.
Beef stearine is very difficult if not impossible of detection. For this, refer-
ence may be had to Lewkowitsch, 5th Ed., Vol. II.
HARDENED OILS
As the name denotes, these are oils which have been changed to more or
less solid fats by the addition of hydrogen, in the presence of a catalyst, usually
a compound of nickel. This betrays their presence and may be tested for as
follows: 149
Ten grams of the fat are heated on the water bath with 10 ml. of hydro-
chloric acid (sp.gr. 1.12) with frequent shaking for two or three hours. The
fat is removed by filtering through a wet filter, receiving the filtrate in a porce-
lain dish; after partial evaporation of the filtrate 2 or 3 ml. of concentrated
nitric acid are added and the evaporation continued to dryness to ensure the
destruction of the organic matter. The residue is dissolved iIi a few m!. of
distilled water, a 'few drops of a 1% solution of dimethylglyoxime in alcohol
added, and a few drops of dilute ammonia. The presence of nickel is shown
by the appearance of the red-colored nickel dimethylglyoxime. The amount
of nickel can be determined colorimetrically by comparison with solutions
containing known quantities.
The quantity of nickel is very minute, not as much as the fats take up when
cooked in nickel or nickeled dishes, and need cause no apprehension.
Hydrogenation destroys all the characteristics, particularly the color tests,
by which the different oils may be sometimes detected.
149 Kerr, J. Ind. and Eng. Chern., 6, 207, 1914.
1788 FIXED OILS, FATS AND WAXES
WAXES
These, as will be remembered, contain no glycerine; the tests applied to
them are the same as to the oils. The characteristics of the more commonly
occurring waxes are given in the table, p. 1797; sperl4 oil, wnich is really a liquid
wax, is included among the oils. '
\
Specific I Flash Viscosity
Oil Gravity, Point, (Saybolt), Cold Test,
Deg, Baume OF. at 70° F. Deg, Fahr,
at 60° F.
/
Black ......... , .......... 29 325 100-120 5--15
Ice machine .............. 26-27 325--360 60-100 Q-4
Crank case ............... 26-27 455 100 -
Transformer .............. - 340-380 400 25
Turbine .................. 30 420 160 -
Spindle ................... 30-35 320-390 58-156 -
Loom .................... 28 360 203 -
Engine ................... 27-30 410 190-210 -
Cylinder ......... _....... 23-25 525 200-300* -
Cylinder ................. 26-28 400-575
I
- -
* At 212 F.
0
Belt Dressings are (1) mixtures of fats, waxes, degras or tallow with castor
or fatty oils; (2) vulcanized corn or cottonseed oil thinned with naphtha;
(3) preparations containing wood tar; or (4) preparations containing rosin,
which is undesirable. Black oils, car oils, well oil or reduced pils are crude oils
from which the naphthas and burning oils have been separated 'by distillation.
Qrank-case oils are pure mineral oils which emulsify but little with water.
Mming-mach~ne or soluble oils are lard, sulfonated oils or mineral oils held in
suspension in water by SO!lpS or alkalies, as borax or soda; the soaps used are.
either ammonium, sodium or potassium with resin, oleic or sulfofatty acids.
Rosin oils are obtained by distilling or " running:" rosin, each distillate being
called a "run" and numbered according to the times it has been distilled.
They oxidize quite rapidly and should not be used as lubricants except as soaps
in lubricating greases. Screw-cutting oils are often mixtures of 27 0 Be. paraffin
and 25% fatty 0:.1, preferably cottonseed, although lard oil' was formerly used.
Stainless oils are spindle or loom oils mixed wjth··fatty oils-lard or neatsfoot.
Transformer ofls should be either pure mineral 'Or rosin oils and as free as
FIXED OILS, FATS AND' WAXES 1789
possible.fl'oII,l water, acid, alkali and sulfur. Turbine oils should be of excellent
quality, free from acid and tendency to resinify, and low in sulfur. Watch
oil is obtained from the porpoise, dolphin, or blackfish, where it exists in cavities
in the jaw and in the brain or " melon" of the fish. Lubricating greases are
mixtures of soaps of palm oil, tallow or rosin oil (with lime or soda as bases)
with various oils or fats such as rosin, tallow or mineral oil. The best are those
made from tallow by saponification with caustic soda. They may also contain
finely powdered talc or graphite. Non-fluid oils are oils or their greases stiff-
ened with" oil pulp" or " dope," i.e., aluminum oleate or palmitate.
" The source, preparation and uses of the various oils "and greases are de-
scribed in Rogers and Aubert's Industrial Chemistry, Chapters XXII, XXIV,
XXV and XXVII.
For the guidance of the analyst, the characteristics of the more frequently
occurring oils are given, the usual figures being given in italic.
The animal'and vegetable oils may be classified into
Drying.-Linseed, Chinese wOQd, poppyseed, sunflower and menhaden.
Semi-drying.-Corn, cottonseed, sesame, rape, black mustard and horse.
Non-drying.-Castor, almond, peanut, olive, cocoanut, palm, seal, cod-liver,
elaine, lard, neatsfoot, tallow, sperm and whale.,
REFERENCES
Lewkowitsch, "Technology and Analysis of Oils, Fats and Waxes." 3 vok
Gill, "A Short Handbook of Oil Analysis."
Battle, "Industrial Oil Engineerin?,"
Lockhart, "American Lubricants.'
1790 FIXED .OILS, FATS AND WAXES
.
,'0
."
:'"
:al
UInllIJXWl'mqdJIlS ."
'.!ol
.C'
:~
mnunx-em
'''''''I uOli"llll·lG
mnunx'B1II
IU"" JOe! 06
~
o
.• .:~
", V
r-~ .lQ
~ ·0
o .~
mmmmw
IlloO jod 9
.",
.",
FIXED OILS, FATS AND WAXES 1791
til
Q)
~t-OOOeo")lt:)i:'<':I
til
0$
Q)
t.....j~e-4c-iet:ie-i
C5 c..
] .....
"g_.g.g.g .g.g
~ .' ...
11
~~~-.c.<t:I".'H~
t; ..... >-i ...... P""OI~'P""I
c..
1792 FIXED OILS, FATS AND WAXES J
~
'"
;:!
.~
g
C,.)
....I
C'I ~u!punodmoo j, :
0>
,....,
,£
,....,
:OOlQlQOOCl.QOC OOOlQOet:l
._C'l'"<1'lQ~_I!'1""':-lQt-OOlQt-o
::q d . . . . . . . . . . ,....j,....ie-:i 0 •
C)
~
~
:;E :1l....: :.......
: : : : : : :.
,
~ :'~~~~~~~~~~
~ :~ : : : :,: : : : :
w
...
......
\
~
rti
~
~ Ul
mUJ~ lad HO:ll
'j]W mnUI!XWI '£l!PF>V : : : : : ~: : : :;~~~~~~~~~~~~~~~
: : : : : : : : : '0000000000000000 :0
:~
...
P :=
0
~
bIl
= 'luaa JOO
.. .. . .. ..
.. ..
··· ..........
. . ..
......
- - -
'.;:I , , , ..... , . .
..,cd · .......... ·_~~oo
p gs P"l nl!P
-.
10100 'W '.1 ,'9 ''1'
...... ';::
Ct
P"lnnpun •10rOO 'W '.1 'S ''1' . :. :.
~
~
:3 r.; mnm!"uw
t-t-t-ooOOt-t-t-oooot-t-t-OO~t-t-t-oooo : :
Z
~ ~
".. mnlII!u!w
w
E-<
Z r.; ummp:uW
~ 1:,
C) ;::
...... umW!U!W
~
I=Q
P
~
~
...
0
W
Z ~ :
0......
E-<
~
C)
...
......
......
C)
~
p.,
W
FIXED OILS, FATS AND WAXES 1793
~
IV
.j
8
~C'I llulPunodmoo
0>
.....
oo~
.....
:o.t:)o.t:)
p;: :~...t(
Q
~ '00 :.~
'00
.",,,,
~
~
A
~.
w
....
...
~ l""l UO!"""OQ
~
w
1794 FIXED OILS, FATS AND W~S
Lard ..................... 35
Linseed ............ , ..... 1.4546 17-24 13-17 179 209.8
/
ci
o
:;l
.~
~
o
U
o
'"~
...
'"'"...
<0
,~ ~ .,'"
"'''' '"<0... '"...
...
'"
<0
""
o
...... '""" I~~
.,o d
.J.'" 60) ~ ........ "o
.... ...a. 0 .,.I"'''''" ;:'" '". U
'"o o '"o o o ''""
00
.'"
'.."
'il
d
11o "o ,,6
<; 6 ~ o 'il
:;;: U ::s
~
U
1796 FIXED OILS, FATS AND WAXES
I I ~I
I I ~I I :I : I II:I I II: II
oi
.~
".."
]
~"
I~
I~
I
I~
U",)
t-
.....I
t-
FIXED OILS, FATS AND WAXES 1797
.
o
:~
o
~
(.)
100
I~
o o
-
C':) t"-
.8; t- co 10>
10> 10>
0>
co
0>
$
0>
20> ~
o o o o
1798 FIXED OILS, FATS AlfD WAXES
DEGREES-ENGLER
I
Time Engler. Redwood, Redwood Adm. Saybolt Universal, Fural, Barbey
600 8001,,000 2 .. 000 3 .. 000 4.000 6,,000 10,,000 20.. 000 aOI_ooo ,~o,ooo 80,,000
Degrees.Engier ,
-4 5 6 7 8 910 200 ~OO 400 600 800 1000 2000 30004000 6000 8000
1I~
\
1/ IA:: 2 5,000
v v
V ~
;
1000 1..1 [,1 0,000
900 9,000
80C a,ooo
70 a ., '(\ 7,000
60 a 6,000
50 a 5,000
0
/'
/,.x.",
'/I/lott <;v
4,000
400 'Q. <le'/
'ti/" /
",'" ' /
aoo ~e' 3,000
0: 1"'- I//.s"~ I_cO-Eo J '
~</<o
~~~~
(j", ..... '/
~ '~20 a _~ ,e 2,000
';;; ~
8°
..... 'V~ ).) ~\~e
:!c}' 1;
SS~100
Q) V ..
~~ ~ l,DDO
i~~ 00 900
\:800
..Q 70 700
« 60 600
II 50
_~ 40
,8
/ J 7V
'/
.'
S 30 "/ // '\ K£Y
.~
/ If 7fg -7 ['1'1. + D.ta 'rom Ar<hbutt & Deeley
.. 20
I
, A Distilled water at 20°C.
• G/~cerol mixtures at 20° C.
~
~ / I
I
I
J,'
I
I
,,
..... o Refined castor 011
/ ,
I II, I
'" Mineraloils
@ Alcohol, 40 %
10
9
8
7
6 ~,
,>
4 ,
, , I
"-
I I , .....
I , I , , .....
2 ,
I
/i ,-, ,,
1
o 20
I
30
I
~F
40
I
5060708090100
lA
200
,
'0
300 400 SPO 8001000
"-
2000
" :-.....
3COO 400060008000
Time Engler. Redwood, Redwood Adm. Saybolt Universal, Furol. Barbey
2 3 4 5 6 7 a 9 10 20 30 40 50 60 80 ~OO 2.00 aoo 400 600 80Q
Degrees-Engler
FIG. 258.-Viscosimeter Conversion Chart 160
.(Directions for Use of Chart),
OBJECT: To find the viscosity reading of a particular oil on any of the standard
viscosimeters when its reading on one of the viscosimeters has been determined by
experiment at the same temperature.
SCALES: There are two sets of scales which must not be confused with each other.
X: The scales at bottom and left hand side go together. They/are the ones regu-
larly used in practice applying to all ordinary oils. ._
Y: The scales at top and on right hand side go ~ogether. They are rarely used,
applying only to very viscous products beyond range of other scales.
150 Power (1922), 377.
FIXED OILS, FATS AND WAXES 1799
NOTE.-There are two lines of scales at the bottom: (a) The lower scale applying
only to degrees Engler. (b) The upper scale applying to all other readings, including
Engler-time. In simply changing from one viscosimeter to another the scales at left
and right are not used. T~e scales at top must never be used in connection with those
::It the bottom.
APPLICATION OF CHARTS: X: For ordinary cases.-GIVEN: Saybolt Uni-
versal Viscosity = 120". PESIRED: Gase (a) Barbey reading. Case (b) Engler-
Degree reading. PROCEDURE: Find point on bottom scales for Saybolt Universal
corresponding to 120" (A). Follow up vertically to curve marked Saybolt Universal
(the known instrument) (B). (a) To get Barbey reading: Follow horizontally to
curve marked Barbey (C). Drop down to bottom scale for Barbey and read (C).
Reading=265 deg. (b). To get Engler-Degree reading: Follow horizontally to curve
marked Degrees-Engler (E). Drop down to lower of the two bottom scales for Engler-
Degrees and read (F). Reading=3.62 deg., i.e., 120" Saybolt=Barbey 265 deg.
= Engler 3.62.
, Y: For very viscous fluids beyond range of other scales. GIVEN: Engler-Degrees
=1100 deg. DESIRED: Saybolt Furol reading. PROCEDURE: Find point on
upper scales under Degrees-Engler corresponding to 1,100 deg. Follow vertically
down to Degrees-Engler, curve. Follow horizontally to curve tfRrked Furol. Rise
vertically .to scales under Furol and read. Reading = 3800", i1!, 1,100 deg. Engler
=,3,800" Furol.
1 The Engler number is the quotient of Engler Time divided by the water value
of the instrument at 20° C. in seconds. '
Reagents
The reagents used in oil analysis are few and easily obtained, A list and
their method of preparation is here given. l5l
Acetic Acid, Glacial. C.P., 99.5% pure. The determination of its strength
should be made by titration and not by specific gravity, as the 9'8% and 80%
161 Proc. Am. Soc. Test. Mat., 15, 1, 288, 1915.
1800 FIXED OILS, FATS AND WAXES
acid have the same specific gravity, 1.067. The determination of the melting-
point gives results equally good with those obtained by titration and requires
less,time. 152 It is made after the manner of the" titer test" (p. 1784), the
tube being half filled, chilled to 10 to 11°C., and further chilled by placing
the outside bottle in ice-water; the temperature of the! super-cooled acid rises
to its melting-point, where it remains sta'tionary for SOme time. The melting-
points of acids of various strengths are as follows:
100%, 16.75° C.; 99.5%, 15.65°; 99%, 14.8°.
For Hanus's solution it must not reduce potassium bichromate and sul-
furic acid.
Acetic Anhydride.
Alcohol: Commercial" Cologne Spirits." For the preparation of alcohol
free from aldehyde for alcoholic potash, cologne spirits are treated with silver
oxide as follows: 1Yz grams of silver nitrate are dissolved in 3 ml. of water,
added to 1 liter of alcohol and thoroughly shaken; 3 grams of potassium hydrate
are dissolved in 15 ml. warm alcohol and, after cooling, added to the alcoholic
silver nitrate and thoroughly shaken again, best in Ii tall bottle or cylinder.
The silver oxide is allowed to settle, the clear liquid siphoned off and distilled;
a few bits of pumice, prepared by igniting it and immediately quenching under
water, being added to prevent bumping. AlcollOl for use in tbe free acid deteI~
mination is prepared by placing 10 to 15 grams of dry sodium carbonate in
the reagent bottle, taking care to filter it before use.
Alcohol, Amyl. /
Bromine. The commercial article; also a N 13 solution, made by dissolving
26.6 grams bromine in 1 liter carbon tetrachloride.
Calcium Chloride. The dry and also the crystallized salt.
Calcium Sulfate. Plaster of Paris.
Carbon Tetrachloride. Baker and Adamson's C.P. or Kahlbaum's " Tetra-
chlorkohlenstoff."
Chloroform. Squibb's, U. S. P.
Copper. Copper turnings or clippings, used for the generation of nitric
oxide. I
Hydrochloric Acid, C.P.-Specific gravity 1.2: For N/~. HCl dilute 39 ml.
of the above acid to 1 liter' and standardize.
Iodine Solution. Fifty grams of iodine to 1 liter of alcohol. For Hanus's
solution dissolve by warming 13.2 grams iodine in 1 liter glacial" acetic acid;
cool and add 3 1111. of bromine. ,
Lead Acetate. One hundred grams of the salt to 1 liter.
Lacmoid. Three grams per liter of dilute alcohol.
Lacmoid Paper. Unsized paper dipped in above isolution.
Litmus Paper. #
Mercuric Chloride. Sixty grams of the salt to 1 liter of alcohol.
Nitric Acid. Specific gravity 1.34. I
Phenolphthalein. One gram of the substance to 500· m!. of alcohol.
_,-'
152 McIlhiney, et al., J. Am. Chern. Soc., 29, 1224, i907:
>'
lfIXED OILS, FATS AND WAXES 1801
#I
M eta-Pho8phoric Acid. A saturated solution of the commercial "stick
phosphoric acid" in absolute alcohol.
Pota88ium Bichromate. Dissolve 3.8633 grams of the C.P. salt in 1 liter of
water; 1 ml._is equivalent to 0.01 gram of iodine. The solution should be
tested against iron wire containing a known percentage of iron.
Pota88ium Hydroxide. N/2: Dissolve 30 grams of "potash by alcohol"
in 1 liter of alcohol. N /6: Dissolve 10 ,grams of "potash by alcohol" in 1 liter
of water and dilute to proper strength. The solution should be protected by
"stick potash" from the carbon dioxide in the air. Ten p~ cent: Dissolve
100 grams of "stick potash" in 1100 ml. of ·alcohol.
Pot.a88ium Iodate. A 2% solution.
Potas8ium Iodide. One hundred fifty grams of the commercial salt are
dissolved in water and made up to 1 liter of water. This should be free from
iodate, shown by yielding no coloration when acidified with strong HCl.
Silver Nitrate. Thirty grams to 1 liter+OA ml. HN0 3 •
Sodium. '
. Sodium Chloride. Ordinary "coarse fine" salt for freezing mixtures.
Sodium Hydrate. 36° Baume. Dissolve 300 grams of caustic soda in
1 liter of water. .
Sodium N itropru88ide. The commercial salt.
Sodium Thi08ulfate. N/lO: Dissolve 26 grams of "sodium hyposulfite,"
in 1 liter of water.
Starch Solution. Rub up in a mortar 1 gram of potato starch, with 10 to
15 ml. of water, pour this into 200 m!. of water which is boiling actively, and
continue the boiling for a few minutes.
Sugar. Ordinarily granulated sugar.
Sulfur.. A 1.5% solution in carbon bisulfide.
Sulfuric Acid, C.P. This should be at least 99.5% pure, and its strength
be determined by titration, as 100% and 94.3% acid have the same specific
gravity.
Dilute. One part acid to ten parts water.
Nitro8ulfuric Acid, for the Elaidin Test. A liter of sulfuric acid of 46°
Baume (1.47 specific gravity) is prepared by diluting 560 ml. commercial sul-
furic acid to 1 liter; a few drops of nitric acid are added and nitric oxide (gen-
erated from copper and nitric acid) passed in until it is saturated. The acid
is then cooled in ice-water and the gas passed in until it is saturated at 0° C.
This is ca1led Roth's liquid.
The author wishes to acknowledge his indebtedness to Mr. Thomas T. Gray
for his careful review of this chapter. Mr. Gray's broad experience in petro-
leum products as Chief Chemist of Tidewater Oil Company, makes his criticism
and suggestions of special value.
1.53 Richmond, J, Soc. Chern, Ind" 9, 479, 1890.
1802 FIXED OILS, FATS AND' WAXES
i'
FIG. 258a.
3. Starting: Remove the lamp again from the bala.nce, fill.the cup L be-
neath the burner heat with spirit and light it.' 'rhen take the hood m an(j
screw the air hose of the air-pump to the valve. When the spirit 9f small cup
is nearly burnt out, press air into the recipient by ;vigorously pushing the air
piston; the air acting upon. the fuel, it will rise into the burner and gasify on
its hot surfa~e. The gas streaming out of the nozzle ignites itself on the spirit
fiame and feeds the ga~ification, even when the spirit is completely burnt out.
If no gas streams out, the nozzle n is obstructe~.i it mfist be cleaned by
means of the special pin included with the appa;.ratUs. An ordinary pin must
not be used as in doing so damage would be done to the nozzle.
FIXED OILS, FATS AND WAXES 1803
4. Regulation: Press as much air into the apparatus that a uniform and
. free burning flame is obtained. (The development of heat should be about
1000 calories per hour, thus about 100 grams of petroleum, benzine or lubri-
cating oil, or 130 grams of spirit must be burnt per hour.)
If the flame is too large, the screw h should be slightly slackened so as to.
let some air escape. When the flame is well regulated, the small wing screw
of the valve is tightened again, the pump is removed and the small hood m is
screwed on the thread. •
5. Measuring: Now introduce the burner' into the combustion chamber
beneath the calorimeter and hang the lamp on one arm of the balance as
indicated but not before having filled the calorimeter with water so that it flows
out at the outlet.
After having introduced the lamp, the thermometer at the water outlet
begins to rise. By means of the regulating cock of the calorimeter the water
supply must be regulated thus, that the difference between the inflowing and
outflowing water is about 10-20°. After having reached the state of perma-
nency, i.e., when the thermometer does not rise any more, put a small weight
on the scale k so that the arm of the balance goes down. The balance will only
reach its state of· equilibrium when a certain amount of fuel has been burnt.
The moment when the pointer of the balance reaches the point of zero,
the hose must be turned over the gauge. At the same time the cup k is charged
with 10-20 grams. After this, begin reading the two thermometers, doing
this preferably at regular intervals. The trapping. of the water and the reading
is interrupted the moment where the lamp again starts rising and the hand
goes again through o. At this point 10-20 grams of fuel have been burnt.
The higher heating value is 'calculated according to
W
H=-T
G '
where H is the heating value in calories per gram,
W the quantity of the water trapped in kilos or liters,
G the quantity of fuel consumed, in grams, ..t\ .
T the difference of temperature between infloMiig and outflowing water.
at the highest. Thus, several oils cannot be used, for instance, neat's foot oil,
rape oil, olive oil. I
Mineral lubricating oils may be used, if they evaporate completely at 250°.
If the preliminary warming is not sufficient, or th~ pressure bf the air is too
• high, the burning is subject to sudden shocks; by sl~ckening the screw hand
warming up again, this inconvenience will be done away with.
When lubricating oil has been burnt, the lamp must be well cleaned with
petroleum followed by benzine . •Then ,he lamp should be soaked with a
quantity of the liquid fuel which is to be tested. Only when this has been
done, the liquid fuel for a new measuring may be poured into it.
There is no danger whatever of explosion, even with benzine, as the flame
will never get back into the recipient, as the pipe and the small nozzle are
always filled with fluid. Naturally benzine should not be poured into the
recipient in the proximity of the burning lamp.
1806
PETROLEUM AND ITS PRODUCTS 1807
.
..L-
1...... 4.0"(IO.2cm.J,·1
I' Outside Diamefer
, FIG. 259.-Distillation Flask.
Tolerances
Centimeters Inches
Centimeters Inches
I
(0.64-cm.) transite or hard asbestos, in the center of which is cut a hole 3Yz in.
(8.9 cm.) in diameter. The board shall not be less than 6 in. (15.24 cm.) in
wiq.th and shall be supported by' an ordinary ring s'tand and ring.
Gas Burner or Electric Heater: (a) Gas Burner.-The burner shall be so
constructed that sufficient heat can be obtl}ined tq distill the product at the
uniform rate specified below. The flame should Inever be so large that it
spreads over a circle of a diameter greater than 5 in. (12.7 cm.) on the under
surface of the asbestos board. A sensitive regulating valve is a necessary
adjunct, as it gives complete control of heating.
:.' (b) Electric Heater.-The electric heater, which may be used in place of the
gas flame, shall be capable of maintaining the distillation at the rate specified in
Procedure (i). The electric heater shall be fitted with an asbestos board top
VB to ~ in. (3.18 to 6.35 mm.) in thickness having a hole 33--2 in. (8.9 cm.rIn
diameter in the center. When an electric heater is employed the portion of
the shield above the asbestos board shall be the same as with the gas burner,
but the part below the board may be omitted. '
Measuring Graduates and Pipettes.-Filling charges and distillation frac-
tions shall be measured with suitable accurate graduated cylinders or other
volumetric glassware.. A 300-ml. pipette with unconstricted outlet, similar
to Fig. 261, is a convenient device for ,measuring charges, but an accurate
graduate may be used if desired. The regular 100-ml. graduate used in gasoline
distillations as ,specified in Section 7
-'--r-
. .. :
/ of the Standard Method of Test for
Distillation of Gasoline, Naphtha, Kero-
:~
• E;-:-
i
• sene, and Similar Petroleum Products
O.S"(1.?7cmJ· • • ~~ ~ : (A. S. T. M. Designation: D 86) of the
OJ-: C\.I :
.~ t! : American Society for Testing Nraterials 1
: ~ : is reasonably satisfactory for measuring
--4
--{;-
~ :
:
distillation fractions, but a graduate
of similar dimensions that can be tightly
JOO ,. I? - -:- stoppered is preferable. One can be
~~ CS made I by '~utting down a standard
m/. 1<"5<3)"'_'~
_.'l-'o",,;.£:: ~-..;:
-r ~ .~ graduate to remove the lip and then
--~~ ,
:,
: C).
(\j •
•
fire p()lishing.
PI;0cedure. - (a) The condenser
:
:......... i
bath shall be filled with cracked ice
'~. ~ : and enough water add~d to cover the
a312S" (0.8 em.)- ~~ : condenser tube. The temperature shall
.(\j
:, ' -
- - L __ .:L cooling
. :
,
be maintained between 32° and 34° F.
(0 and 1.1° C.). Any other convenient
medium may be used pro-
FIG. 261.-Measuring Pipette. vided the~e temperature conditions are
maintaim!d.
(b) The condenser tube shall be swabbed or cleaned inside to remove any
liquid remaining from a previous test.
(c) The specific gravity of the oil to be tested shall first be determined in
-
7 1936 Book of A. S. T. M. Standards, Part II:
_-
accordance with the Standard Method of Test fpr _Gravity of Petroleum and
-
PETROLEUM AND ITS PRODUCTS 1811
The whole procedure snaIl be repeated enough times to yield a t'otal volume
of distillate of not less than 100 pli. j
Gasoline, Naphtha
The specific gravity, ,Baume or A. P. 1., is tak~n with 'a hydrometer (page
1704). This information without other data, SUefl as the distillation curve or
the knowledge of its being a straight run, natural !!;asoline, or blended product,
is of little value in the determination of the quaIit:y- of the gasoline.
The yellow color in a gasoline (test·, page 1742) indicates improper treatment
wit~ aci? or ~odium ~lumbite solution.. A_ dar.k cplor may be caused by con-
tammatIOn WIth heaVier hydrocarbons III pipe lme .or tank cars, and should note
be confused with colors caused' by other methods.
The sweet ethereal odors of gasoline are characteristic of the gasolines from
low sulfur paraffin base crude oil. Sour or " craQked " odor is characteristic.
I
10
1936 Book of A. S. T. M. Standards, Part II. Se&~: 1705.
. -'
PETROLEUM AND ITS PRODUCTS 1813
simultaneously for the same period of time, but not less than 1 hr. in a desic-
cator. If the air in the balance to be used is dried with a desiccant then this
same desiccant shall be used in the desiccator; otherwise no drying agent shall
be used in the desiccator. The test beaker shall then be weighed, using its tare
(Note) on the opposite balanc·e pan. After weighing, the tare beaker shall be
replaced in the desiccator and the test beaker shall be placed in the well of the
bath as shown in Fig. 262. The air flow shall be adjusted to a rate at the point
of measurement of 1 liter per second±15%. The bath shall then be heated
''''5''O
r.1li o.
1~4t.:> To i?efivA' Condel1$er
CoflonorG/oss
Wool Filter
I
I
I
I
I
I
I
I
I
I
I
O)I~ I
_____ 1,/
,rOD.C~pper ,.
Silver Solder
41/ .Joints
TvbingJ/Wall
Flowl1elt't;
I
To Air $vppOi "
I
not drl)wn
t{) 5C(I/e
, (b) The conical adapter shall be removed and 50 ml. of the gasoline poured
into the beaker after which the adapter shall be r()placed on the air outlet so
that it is vertically above the center of the surface of gasoline. After the gaso-
line has evaporated, which should require from 8 to 14 minutes, the beaker shall
be left in place for an additional 15 minutes, the atr1supply being maintained.
During this 15-minute period the tare beaker shall !lJ~o be heated to about 302 0
F. C150° C.). Both test and tare beakers shall then be allowed to cool simul-
taneously and for the same period of time, but not 1ess than 1 hr., in a desiccator
under the same conditions as prescribed in Paragraph (a), and the test beaker
weighed as before the test. The increase in weight of the test beaker repre-
sents the amount of gum in the sample.
Results.-Duplicate determinations shall be made and results shall be
expressed in milligrams of gum per 100 ml. of gasoline.
Accuracy.-Results obtained by different operators in different laboratories
should not differ by more than the following value!$:
For gum contents of 0 to 20 mg. per 100 m!.. . . . . . . . . . . . . . . . . 4 mg.
For gum contents of 20 to 100 mg. per ,100 ml..... , . . . . . . . . . . .. 10 mg.
For gum contents above 100 mg. per 100 ml.. . . . . . . . . . . . . . . .. 20 mg.
Duplicate results obtained by a single operator should not differ by more than
one-naif tne aoove vaiues.
Vapor Pressure determinations give direct information as to the probable
behavior of motor fuels as regards easy starting Itnd "flexibility." It indi-
cates, if the pressure is high, the blending of natural gasoline with a straight
run product. It is important in the shipping of ~asoline in tank cars 13 and
must conform with interstate shipping laws. •
The flash and fire tests (page 1702) must meet the specifications of illumi-
nating and burning oils. A low flash and a high fire indicates a mixture of light
fractions with the heavy fractions, making a poor fuel for the burners. The
flame will not be uniform and the higher fractions will not burn but change into
a vapor in the fire box, causing explosive mixtures. The test determines the fire
hazard and is the real measure of inflammability. Most states have a law
stating the minimum flash and fire points of illuminating oils.
The sulfur determinations (pages 1712, 1756) are of great importance in
illuminating oil as they produce disagreeable odors during the burning period,
and cause a charring of the wicks. The amount should not exceed .1 %.
Fuel oils when atomized with steam form sulfuric acids which cause rapid
deterioration of boilers and pipe stills.
Lubricants
The determination of specific gravity of a lubricating oil (page 1704) may
give information as to the nature of the petroleum from which it is made. A
high Raume gravity and a very high viscosity may indicate a lubricant that has
been mixed with soaps, resins, or oils other than mineral oils.
The color of lubricating oils indicates the source of the oil. A turbid oil
indicates finely divided droplets of water, or if cooled an amorphous wax. A
dark-colored oil should be examined under a microscope for suspended carbon.
When the oil is diluted with naphtha, the foreign matter held in suspension
settles out.
The viscosity test (page 1718) has been regarded as a most important test of
lubricants. The working temperature of the bearing should determine the
temperature at which the test is made. Most lubricants have the same vis-
cosity at 3000 F. A lubricating oil functions only when a continuous film of oil
separates the wearing surfaces and the oil is efficient when its viscosity allows
this film to be continuous without a break in the film. The true "test for a
lubricating oil under working conditions ill the machine for which it is to be uiled.
The cloud test (page 1727) indicates the temperature at which the paraffin
wax separates from the oil solution. It shows whether the process for the
separation of wax has been well done and gives the quality of the wax which is
separated out. If the wax is amorphous in character, the oil must either be
redistilled or centrifuged to remove this wax.
The pour test (page 1730) indicates the temperature under which an oil will
flow. Knowing the working temperatures that the oil will be used will deter-
mine whether it will Qe an efficient lubricant.
The flash and fire test (page 1732) should be above the working tempera-
tures of the bearing. High flash and burning· points indicate good quality
and careful refining. Too low flash and fire test results in a large .loss from
evaporation under working conditions.
The carbon residue test (page 1737) gives information in the laboratory
approximating working conditions with the lubricants, producing more or less
carbon deposits. A large per cent of residue indicates oils. thickened by other
matter or badly refined. .
The demulsibility test (page 1740) determines the value of a lubricant which
is to be used in forced feed lubrication and motor cylinders. A high degree of
1818 PETROLEUM AND ITS PRODUCTS
A. S. T. M. DESIGNATION: P 188-27 T 17
phthalein solution. Boil. If solution turns pink, add 1 nil. of the indicator
and titrate with the sulfuric acid solution until the eixtraction is colorless.
If the addition of 1 drop of phenolphthalein causes no change in color, add
0.1 ml. of methyl orange. If color changes to red lor pink, titrate with the
KOH solution until the red color is just discharged. I /'
The evaporation test (page 1726) is used in testing lubricating oils used in
high vacuum machines, transformer o_ils and air compressor oils. A high
evaporation loss indicates light oils used in blending. It may be used in testing
the dilution of " crank case" oil.
A. S. T. M. DESIGNATION: D 322-35 20
divisions. The error of any indicated capacity shall not be greater than
0.05 ml.
(c) Flask.-The glitilS flask shall be of the short-neck, round-bottom type,
with a nominal capacity of 1 liter. '
(d) Condenser.-The condenser shall be of the water-cooled, reflux, glass-
tube type, having a condenser jacket not less than 400 mm. (15% in.) in length
with an inner tube 9.5 to 12.7 mm. (% to 72 in.) in outside diameter. The
end of the condenser to be inserteu in the trap shall be ground off at an angle
of 60°.
(e) Burner.-The gas burner shall preferably be of the Meker type, 1 in. in
diameter, and the orifice used should be such that a narrow flame coming to a #
point about 4 in. above the top of the burner can be obtained.
Procedure.-The sample shall. be mixed thoroughly, and 25 mI., shall be
1822 PETROLEUM AND ITS PRODUCTS
320 F. and an end point of 400 to 420 F. At least 13.5% of the diluent should be
0 0 0
A. S. T. M. DESIGNATION: D 128-27 22
a separator, allowed to clear, and the aqueous layer (A),24 which contains all
the bases as well as glycerin, shall be drawn into 'another separator. The
METHOD r ~-ETHOO 1
Sample Dec.omposedbyKHSo..and Sample Decomposed by HCLand P.E.
Extracted by PUn 50xhlet Apparatus Filfrt,. thro~gh Gooch
. I I .
I ,
filtrate('Z~yers) Insoluble$
$qxrrrm ASBESTOnALC,~,
I ~,~,6RAPKITt,~
1
Residue
(lno'!J'l'X>
DiWJrrJ.
PE.sJ,n .• (B)
Tot.~~s~~a.I,p'o"
Acid AquJus Sol;',. (A)
T.. co. of FItly Ac;d,P.Q, Chlorid..,
'winrndKC'
6,
Wash JT,inlS wifhWalu
r--L----Watei"
Traces of HCl
Wash TWI" withRE.
It
(fOTes,
Whim'tra Separu/lonof? Imml$Cihft P.E,Washlnqs=(c) Acid Stln,(A)
So/u#OI'lS is Shown by 8randting Lin,s, T .. ""ofFatty Acid, PO. and "CI. ChIQrid'~HC\,Gly",rin.
IIttHll1v,"er(lowerjJo/ufloriIl7SepomfrJ"_ Wash Oneil WI'" Wafttr. R.stuvI (or Gty,,,.,
tunnel Fol/OW5 th. l?'ghl8ronch and I TNt.
,*, LIi/hJIr.)Qlvl<lMFul1oWJ '1M L8_ff
IJranoh P.E!..(C) Wa{er
"rom here, MsModl , Tra,.. 01 rattyAdd,P.o. TrsttsofHCl.
P.E.&:. P.frofetJrrf [thet: iarHllaresams Oiscord.
CtO-Cthyl Cfher. Tdrt1te wrl}, KOH and SepatTTf_.
p a. Petroleum Oil,.
,
I I
1 I , )
P. E. So In = (E) / h f Alcoholic 501n·(D
ht,po,and T..". of Rol.,h Soaps- • Alto ollCWashlngs Pol.sh Soaps,T..u,clfatUo.
Wa,h 3 Tim ej/Wldh50~/U:CohOI "" Trta1ea$n,O: ~O~$· Pla~lnSepolT1for.
_. _ 7" WashOmewilhP.[.
I
p.E.wa~hinCjS
Traces of Fatand P'Q.
AICOhotic'SOI'n.(O)
PE
l
1----<E;------oE-E---I1
So"n.=~E)
Potash Soaps
fvapomt.inB.Qk.r,fI_
in Hoi Wafv,4cidify.Shah
fafandPO. OV-lfw/C4Wlfh[f(1.
Saporllf'y with I(OH,
I
7ifrall! and SepOI'l1M
I ,J
EtO.Soln.D(F)
1
Acid Wa~r
AkoholiJ 50In.v (H) Fatty icid and Traces Het. KC1, HCL
Pol.,h So,ps;Tra«. of PD. Wa.1hTwio wJ/hWanr. Olwzrd.
Water
HCL
DiJrar#.
1,
Acid Water
KCI, He!.
Discard.
petroleum ether layer (B) shall be washed three times with 25 ml. of water
to remove hydrochloric acid, adding the washings to (A)! Solution (S) shall
24 Solutions designated by capital letters in parentheses refer to the solutions given
in Fig. 264.
PETROLEUM AND ITS PRODUCTS 1827
then be run into a 250-ml. separator. The aqueous solution (A) shall then
be washed twice with 20 mi. of petroleum ether (C), whereupon (A) is set
aside for examination for glycerin. Solution (C) shall then be washed
once with 15 mi. of water, which is rejected, and (C) sha,ll be added to (B).
If (B) and (C) are comparatively light colored, 'an approximate determination
of free fatty acids and fatty acids from soap may now be made by titrating the
solution in the separator with '0.5 N alcoholic potassium hydroxide and phenol-
phthalein, ,using 200 as the average neutralization value of the fatty acid (that
is, 1.0 g. of fatty acid requires 200 mg. of absolute KOH for neutralization).
If solutions (B) and (C) are dark, a few drops of phenolphthalein and sufficient
0.5 N alcoholic potassium hydroxide shall be added to make the alcoholic layer "
distinctly alkaline after vigorous shaking. If the solutions have been titrated,
a slight excess of alkali shall be added before separating. The conservative
addition of strong alcohol at this poin.t may aid in securing rapid and sharp
separation. The two solutions shall be allowed to separate sharply and the
lower alcoholic layer (D) drawn off into another separator. The upper pelro-
leum ether layer CE) shall be washed three times, with 30, 25 and 20 mi. of
neutral 50% alcohol respectively, adding these washes to (D). Solution CD)
shall be washed with 25 ml. of petroleum ether, after which CD) is drawn off into
a beaker and the petroleum ether added to CE). Solution CD) shall then be
evaporated to a small volume to remove alcohol, the residue of potassium soap
washed into a separator with hot water, acidified with hydrochloric acid and
shaken out twice with 50 and 25 m!. of ethyl ether (F), respectively. Solution
(F) shall be run into another separator and washed twi~e with 20-ml. portions
of water, which is discarded. Solution CF) shall then be transferred to a weighed
'beaker and evaporated to dryness on the steam bath, blowing with air to remove
. all traces of petroleum ether. The residue consisting of free fatty acid and
fatty acid from soap shall be heated for a short time on a steam bath, adding
and evaporating 5 m!. of absolute alcohol to remove the last traces of water,
and weighed. The exact neutralization value shall then be determined on as
large a sample of these fatty acids as possible. From the total fatty acid found
here, the free acid if any (see Sections on Free Alkali and Free Acid) shall be
deducted and the remainder calculated to percentage of soap, the ash analysis
being used as a guide in distributing the bases, if more than one are present.
The fatty acids may be identified to some exte~t by special tests, such as
odor, crystal form, melting point, iodine number, neutralization value, color
reactions, etc.
If the grease is appreciably oxidized, the fatty acids obtained by the ethyl
ether extraction are likely to be dark in color and hard to identify. For further
study and identification, the neutralized acids may be extracted again (qualita-
tively) with petroleum ether and dilute Hel, thus eliminating the small amount
of more darkly colored matter.
Detection of Glycerin.-In case it is desired to determine whether a grease
has been made from whole fats or from fatty acids, solution (A) shall be neu-
tralized with dry sodium carbonate and sufficient excess added to precipitate
calcium or other metals. The whole mass shall then be evaporated to dryness,
the residue extracted several times with strong alcohol, the combined alcoholic
extracts filtered, and the alcohol evaporated. The residue will then contain
1828 PETROLEUM AND ITS PRODUCTS
most of the glycerin, with a little sodium chloride. The presence or absence
of glycerin in the residue shall be confirmed by suitaple qualitative tests.
Since fats on saponification yield between 10 and 11 % of glycerin, and this
analytical scheme does not include a quantitative determination of glycerin,
the total percentages in greases made from fats should always be 'less than
100. 25 .
Waxes
The melting point of mineral waxes may indicat;lits origin and determines
its classification. A wide range of temperature in th~ melting shows a mixture
of waxes and incomplete refining methods. I .
I
STANDARD METHOD OF TEST FOR MELTING POINT 0F
PARAFFIN WAX 28
A. S. T. M. DESIGNATION: D 87-2229
Water Bath.-Suitable cylinder, 5Ys in. in inside diameter and 6 in. deep.
This water bath shall be provided with a suitable cover and with the guides and
fasteners necessary to hold the air bath firmly in a vEfrtical central position so
that the sides and bottom of the air bath sha~l be ~urrounded by a layer of
.water 1,72 in. thick. The water bath cover shall have a slot for introduction
of a suitable stirrer and shall have an opening for the bath thermometer so
that the latter may be suspended in a vertical position ~ in. from the outside
wall of the water bath. Air bath, water bath, and water bath cover may be
conveniently made of metal in one assembly as shown in-...Fig. 264.
Stirrer in Test Tube.~Brass or copper wire, Ys in. in dianieteJ; and about
12 in. long. A circular loop, ,72 in. in diameter, shall be formed.at one end of
this wire in sU9h a manner that the loop lies in a horizontal plane when the rest
of the wire is in a vertical position. The stirrer thus formed ,shall be passed
through the proper opening in the test-tube cork and the upper end may then
be bent into a shape convenient for holding., •
28 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products 'and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa. J
29 This method is issued under the fixed designation)_87; the final number indi-
cates the year of original adoption as standard or, in We--case of'revision, the year of
last revision. Issued as Tentative, 1921; Adopted, 1922. This method has been
approved as American Standard by the American Standards Association.
PETROLEUM AND ITS PRODUCTS 1831
f,!
-G 1-
'!
t
:
:e:
~I\Z
~-----------~~:----------__'j-
FIG. 264a.-Apparatus for Determination of Melting Point of Paraffin Wax.
up and down throughout the .entire length of the test tube in a steady motion
at the rate of 20 complete cycles per minute. The melting point thermometer
reading, estimated to 0.10 F. shall be observed and recorded every 30 seconds.
The temperature of the wax will fall gradually at first, will then become almost
constant and will then again fall gradually. '
The melting point thermometer reading, estimated to 0.10 F., shall be
observed and recorded every 30 seconds, for at least three minutes after the
temperature again begins to fall after remaining'almost constant. The record
of temperature readings shall then be inspected and the average of the first
four readings that lie within a range of 0.20 F. shall be considered as the uncor-
rected melting point. This temperature shall be corrected if necessary for
error in the thermometer scale and the corrected temperature shall be reported
as the" A. S. T. M. Paraffin Wax Melting Point." .
Accuracy.-Duplicate determinations on the same sample should differ by
not more than 0.20 F.
A. S. T. M. DESIGNATION: D 24of27 32
Apparatus.-Combustion Bombs.-The Atwater, pavis, Emerson, Mahler,
Parr, Peters, Williams or similar bombs may be used. The bomb shall have
an inner surface of platinum, gold, porcelain enamel or other material whi~ is
not attacked by nitric and sulfuric acids, or other products of combustion.
Calorimeter Jacket.-The calorimeter shall be provided with a vacuum
jacket or with a water jacket having a cover to protect the calorimeter from air
currents. In case a water jacket is used, it shall be kept filled with water within
2 or 3° C. of the temperature of the room (except in calorimeters which are
totally submerged, where the jacket temperature is controlled by a thermostat)
and this water should be stirred continuously by some mechanical stirring •
device.
Stirring of the Calorimeter Water.-The water in the calorimeter shall be
stirred sufficiently well to give consistent· thermometer readings while the
temperature is rising rapidly. The speed of 'stirring should be kept constant.
A motor-driven screw or turbine stirrer is r.ecommended and the speed should
not be excessive. This may be determined, by adjusting the temperature of the
calorimeter to equality with that of the jacket and allowing the stirrer to run
co~tinuously for 10 minutes. If the temperature of the calorimeter rises more
than about 0.01° C. in this length of time; the rate of stirring is excessive.
Accurate results cannot be obtained when too much energy is supplied by the
stirring device or when the rate of stirring is irregular. The portion of the
'stirring device immersed in the calorimeter should be separated from that out-
side by non-conducting material, such as hard rubber, to prevent conduction of
heat from the motor or outside air.
Thermometers.-The thermometers used shall have been certified by a
government testing bureau and shall be used with th,e corrections given on the
certificate. This shall also apply to electrical resistance or thermo-electric
thermometers. Correction shall also be made for the temperature of the
emergent stem of all mercurial thermometers, .and for the" setting" of Beck-
mann thermometers. For accurate work, either Beckmann or special calori-
metric thermometers, graduated to 0.01 or 0.02° C., are required. Such ther-
mometers should be tapped lightly just before each re1tdin~o avoid errors
caused by the sticking of the mercury meniscus, particularly. when the tem-
perature is falling. A convenient method is to mount a small-electric ~zer
directly on the top of the thermometer and connect it with a dry cell and avush
button. The button should be pressed for a few seconds immediately before
each reading.
I
3i Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
32 This method is issued under the fixed designation D 240; the final number indi-
cates the year of original adoption as standard or, in tj:le_case o( revision, the year of
last revision. Issued as Tentative, 1926; Adoptedjn-Amended Form, 1927. This
method has been approved as American Standard 'by the American Standards Asso-
ciation.
. PETROLEUM AND ITS PRODUCTS 1835
Oxygen.-The oxygen used for combustions shall be free from combustible
,material. The bomb when filled should contain at least 5% of nitrogen to
insure complete oxidation of the sulfur. The total amount of oxyge:" contained
in the bomb for combustion shall be not less than 5 g. per gram of fuel oi!.
But the combustion must be complete, as shown by the absence of any sooty
deposit on opening the bomb after firing.
Firing Wire.-The fuel oil in the bomb may be ignited by mean!! of either
iron or platinum wire. If iron wire is used, it should be of about No. 34 B. & S.
gage and not more than 10 cm. (preferably 5 cm.) should be used at a time.
A correction of 1600 calories per gram weight of iron wire burned shall be
subtracted from the observed number of calories.
Standardization.-The water equivalent of a calorimeter can best be deter-
mined by the use of the standard combustion samples supplied by the U. S.
Bureau of Standards. The required water equivalent is equal to the weight
of the sample multiplied by its heat of combustion per gram and divided by the
corrected rise in temperature.
The cal~eter shall be standardized by the combustion of standard sam-
ple~ suppliWby the Bureau of Standards, and used according to the directions.
given in the certificates which accompany them. A standardization shall
consist of a series of not less than five combustions of each of two of the standard
~aterials which are, sucrose, naphthalene and benzoic acid. The conditions
as to the amount of water, oxygen, firing wire, method of correcting for radia-
tion, etc., under which these combustions are made, shall be the lame as for
fuel oil combustions. In the case of any disagreement between contracting
parties a check standardization may consist of two or more combustions of
standardizing samples. .
Procedure.-Preparation of Sample.-Approximately enough oil shall be
taken to give a thermometer rise of from 3° to 4° C. After weighing, the sample
should preferably be immediately placed in the bomb and this closed. This
procedure is necessary to avoid sublimation in the use of naphthalene for
standardization .
. Preparation .of the Bomb.-The firing wire, if iron, shall be measured and
coiled in a small spiral and connected between the platinum terminals, using, if
necessary, a piece of platinum wire somewhat heavier than the iron wire to
make the connection. The platinum and the iron shall both be clean. About
0.5 m!. of water should be placed in the bottom of the bomb to saturate with
moisture the oxygen used for combustion. When the crucible is put in place
in the bomb, the firing wire should touch the fuel oil or briquet of standard
material. For the combustion of standardizing samples, iron wire is preferable
to platinum.
Filling the Bomb with Oxygen.-Oxygen from the supply tank shall be
admitted f;llowly, to avoid blowing the oil from the crucible, until a pressure is
reached as indicated by the following table: -
Minimum Gage Pressure,
Capacity of Bomb, mI. Atmospheres
Road Oils
Road oils may be used as dust-laying oils, binders for sand, sandy loam and .
gravel roads. The value of a road oil depends upon its penetration into the soil,
the binding qualities of the oil, and its resistance to weather and wear.
The tests on road oils are important in as much as the importance of the
specifications for the particular duty for which they are intended. Road oils
are often placed on roads without any study as to the adaptability of the oil for
the soil or base on which it is placed.
Gravity is a common test made on road oils hut has little value. The
asphalt in a road oil acts as the binder and weather-resisting element. The
total bitumen (page 1539) is determined to give the value of the oil as a binder.
The penetration of the asphalt into the road bed determines its wearing
qualities. The viscosity should be low enough to enable to penetrate the soil
and to flow with ease and uniformly when applied. A heavy asphalt oil mixed
with a lighter oil will give greater penetration, the lighter oil slowly evaporating,
leaving a layer of high asphaltic content in the soil. The asphalt should contain
little or no paraffin wax. The wax particles combine with the soil and form
little dry particles which soon wear loose and blow away leaving small pits in
the surface which soon wear away forming large holes and ruts in the finished
roads.
1838 PETROLEUM AND ITS PRODUCTS
Asphalts either native or artificial when used for paving should be tested for
total bitumen, penetration (page 1520), viscosity ~page 1518) and carbene
(page 1546) content. If asphalts are made from petroleum residues, the wax
content should be determined. Petroleum with a cltrbene content over .5%
(part insoluble in carbon tetrachloride) indicates th'at the residue had been
overheated or overblown, making a brittle product. .
Shale Oils
• The testing of shale oils is much the :same as the testing of petroleums.
Shale oils are made by the retorting of the oil shales changing the pyrobitumens
into the bitumens. The bitumens formed vary in chemical and physical
properties as the methods of retorting. High temperatures cause" cracking"
with a resulting polymerization. The nitrogen and sulfur content being high
gives a different reaction to heat than petroleums of the same gravity.
The waxes made from shales have a higher melting point than the paraffin
from petroleum. They may be separated into a white flake wax. by washing
the re-run lubricating distillate with acetone. This method can be used as a
quantitative test for waxes in shale oils.
The different'finished products, gasolines, kerosenes, fuel oils, lubricants and
waxes, are tested in the same manner as siniilar petroleum products.
ANALYSIS OF PAINTS
In reporting the results of an examination of a paint, it is advisable to give
all the analytical data as well as a resume showing the probable composition
of the paint. "This is shown in the following example:
RESULTS OF ANALYSIS
Total Pigments or Solids ......................................... 60%
Total Vehicle or J.,iquids .................. : . . . . . . . . . . . . . . . . . . . . . .. 40
Analysis of Pigment Portion
Lead Oxide (PbO) ........................................... . 37.47%
Zinc <?xid~ (Znq) ............................................ . 44.50
AlumIna, Iron, hme ........................................... . 2.90
. Magnesia (MgO) ............................................. . 1.90
Silica (Si02) .................................. '.' ............. . 4.63
Carbon Dioxide (C0 2) ........................................ . 2.50
Sulfuric Anhydride (SOa) ..................................... . 5.02
Water (combined) ........................................ , ... . .73
99.15%
Analysis of Liquid Portion
Vehicle contained 20% volatile matter.
Volatile matter consisted of equal parts of turpentine and mineral spirits.
Non-volatile matter had:
Iodine Number .......................................... : .. 175
Acid Number. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4
Saponification Number ...................................... 188
and contained .02% ash consisting of lead and manganese o~ides.
Probable Composition of Paint
Pigment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 60%
Liquid ......................................................... 40
Pigment
Basic carbonate White lead. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 22%
Rasic sulfate White lead. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 25
Zinc Oxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 43
Asbestine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . .. 10
100%
Liquid
Raw Linseed Oil ................................................. 80%
Mineral Spirits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 10
Turpentine and Drier ................................ : . . . . . . . . . .. 10
100%
Chapter contributed by Henry A. Gardner, John A. Schaeffer and G. G. Sward.
1839 .
1840 ANALYSIS OF PAIN,J;S
its iodine number be very much lowered. :The volatile matter may also be
separated by steam distillation. I
Detection of Resinates.-To determine whether th~ drier in a paint is of
the resinate type or linoleate type, a few drops of tlte oV liquid may be mixed
on a porcelain plate with one or two drops of acetic anhydride, subsequently
adding a drop of sulfuric acid. Upon the addition of the sulfuric acid, a flash
of purple color, turning to dark brown, will be shown ~here rosin is present.
This test is not conclusive and experience is necessary for its proper interpre-
tation. If rosin should be present in the vehicle to a considerable extent, the
oil will have a very high acid number. The approximate percentage of rosin
present may be determined by shaking a portion of the vehicle with 95%
alcohol in a separatory funnel, subsequently separating'the alcoholic extract, .
evaporating and weighing the re'sidue.
Detection of Various Oils.-Tung (Chinese wood oil) may be detected in
the vehicle by mixing the oil with an equal volume of a saturated solution of
iodine in petroleum ether, allowing the mixture to stand in direct sunlight.
Under these conditions, a peculiar, insoluble, spongy polymer of one of the
glycerides of Chinese wood oil is shown. Fish oil can sometimes be detected
by its odor and the dark red color during saponification, but the best method
is probably that of Eisenschiml and Copthorne 2 which depends upon the
formation of insoluble bromine derivatives of the oil. The presence of soy-
bean and other vegetable oils is in some Gases difficult to d,etect. The iodine
numbers of these oils, however, are all lower than that of linseed oil. It must
be remembered, however, that the iodine number of boiled linseed oil is lower
than that of raw oil and that the iodine number of oils extracted from many
paints is usually lower than shown by the original oil. In the presence of
considerable quantities of drier, it is always advisable to extract the fatty
a<!ids from oil and make the iodine number determination upon them. ft1
Detection of Various Thinners.-The distillate from the paint vehicle mllif
consist of turpentine, mineral distillates, benzol and similar solvents. The
presence pf benzol is readily detected by adding a few,' drops of the distillate
to a small quantity of a mixture of concentrated nitric and sulfuric acids.
Upon heating this mixture, the characteristic odor of nitro-benzol will be
recognized if benzol is present. Mineral distillates from petroleum are easily
detected by the polymerization method given under the Examination of
Turpentine. I
oil, whale oil, herring oil, and similar marine animal oils of relatively high
iodine number.
There are given below specifications for, and methods for, the analysis of
linseed oil, in accordance with the latest practice developed by the Federal
Specifications Board of the Procurement Division of the U. S. Treasury De-
partment. These methods may be followed in examining any of the other oils
mentioned above.
Raw Linseed Oil
Maximum Minimum
Foots:
Heated oil, per cent by volume .......................... . 1.0
Chilled oil, per cent by volume .......................... . 4.0
Specific gravity 15.5/15S C ................................ . .936 0.931
Acid number .............................................. . 4.0
Saponification number ..................................... . 195.0 189.0
Unsaponjfiable matter, per cent ............................. . 1.50
Iodine number,a Wijs ...................................... . 177.0
Loss on heating at 105° to 110° C., per cent ........... , ...... . .2
Color .................................................... .
a When high iodine number type of raw linseed oil is specified by the purchaser, the
iodine ilUmber must be not less than 188 and the oil shall conform to all of the other
requirements.
b Not darker than a freshly prepared solution of 1.0 g. potassium bichromate in
100 ml.. pure concentrated sulfuric acid (sp.gr. 1.84).
Maximum Minimum
a When quick process, not kettle-boiled, oil is called for in contract, the minimum
specific gravity shall be 0.931.
Analysis of Linseed Oil
Loss on Heating at 105° to. ~10° C.-Place 10 grams of the oil in an accu-
rately weighed 100-ml. Erlenmeyer flask; weigh. Heat in an oven at a tem-
perature between 105 and 110° C. for 30 minutes; cool and weigh. Calculate
the percentage loss. Make this determination in a current of dry carbon
dioxide gas.
Foots General Test.-With all materials at a temperature between 20° and
27° C., mix, by shaking for exactly one minutli in a graduated tube, 25 ml.
1844 , ANALYSIS OF PAINTS
of the well-shaken sample of oil, 25 mI. of acetone, ,and 10 mI. of acid calcium
chloride solution. Then clamp the tube in an, upri~ht position and allow to
settIe for 24 hours. The temperature during this period should be between
20° and 27° C. The graduated tube shall be of not ~ess than 70 ml. capacity
and shall. have at least 50 mI. graduated in 0.1 mI. The diameter of the tube
shall be such that the 50 mI. graduated portion shalll be not less than 40 cm.
or more than 60 cm. in length. The volume of the stratum lying between the
clear calcium chloride solution and the clear acetone and oil mixture is read
to 0.1 ml. or fraction thereof. This reading multiplied by four expresses the
amount of foots present as percentage by volume of the oil taken.
Heated OiI.-Heat a portion of the oil to 65° C., hold it within 2° C. of
that temperature for 10 minutes, then cool it to room temperature (20° to
27° C.). Promptly make the general foots test as described above.
Chilled Oil.-Heat a portion of the oil to 65° C., hold it within 2° C. of
that temperature for 10 minutes, then place it in a clean, dry bottle, stbpper
tightly, and place in a cracked ice and water mixture (0° C.) for exactly two
hours. At the end of this time place the bottle in a large bath of water at
25° C. and keep it there for 30 minutes, then promptly make the genel"al foots
test as described above.
Specific Gravity.-Determine at 15.5/15.5° C. by any convenient method
that is accurate within two points in the fourth decimal place.
Acid Number.-Transfer 5 to 10 grams of the oil to a 300-ml. Erlenmeyer
flask. Add 50 mI. of a mixture of equal'parts by volume of 95% ethyl alcohol
and chemically pure reagent benzol. (This mixture should be previously
titrated to a very faint pink with dilute alkali solution, using phenolphthalein
as an indicator.) Add phenolphthalein indicator and titrate at once to a faint
permanent pink color with standard sodium or potassium hydroxide solution.
Calculate the acid number (milligrams KOH per gram of oil). . ,., .
Saponification N umber.-Weigh about 2 grams of the oil in a 300Ull.
Erlenmeyer flask. Add 25 ml. alcoholic sodium hydroxide solution. Put a
condenser loop inside the neck of the flask and heat on the steam bath for one
hour. Cool, add phenolphthalein as indicator, amI/titrate with half normal
sulfuric acid. Run two blanks with the alcohOlic sodium hydroxide solution.
These should check within 0.1 ml. N/2 H 2S0 4 : From the difference between
the number of cubic centimeters of N/2 H 2S0 4 required for the blank and for
the determination, calculate the saponification number (milligrams KOH re-
quired for 1 gram of oil)., .~
Unsaponifiable Matter.-Transfer 8 to 10 grams of the oil'to a 250-ml.
long-neck flask. Add 5 cc. of strong solution of sodium hydroxide (equal
weights of NaOH and H 20), and 50 mI. 95% ethyl alcohol. Put a condenser
loop inside the neck of the flask and boil for two hours. Occasionally agitate
the flask to break up the liquid but do not project the liquid onto the sides
of the flask. At the end of two hours remove the condenser and allow the
liquid to boil down to about 25 ml.
Transfer to a 500-ml. glass-stoppered separatory funnel, rinsing with water.
Dilute with water to 250 mI., add 100 mI. redistilled ether. Stopper and shake
for one minute. Let stand until the two layers separat~' sharp and clear.
Draw all but one or two drops of the aqueous 14'er into a second 500-ml.
separatory funne,l and repeat the process using 60 illl: of ether. After thorough
ANALYSIS OF PAINTS 1845
separation draw off the aqueous solution into a 400-ml; beaker, then the ether
solution into the first separatory funnel, rinsing down with a little water.
Return the aqueous solution to the second separatory funnel and shake out
again with 60 m!. of ether'in a similar manner, finally drawing the aqueous
solution i,nta the beaker and rinsing the ether into the first separatory funnel.
Shake the combined ether solution with the accumulated water rinsings and
let the layers separate sharp and clear. Draw off the water and add it to the
main aqueous solution. Shake the ether solution with two portions of water
(about 25 ml. each). Add these to the main water solution.
Swirl the separatory funnel so as to bring the last drops of water down to
the stopcock, and draw off until the ether solution just fills the bore of the'
stopcock. Wipe out the stem of the separatory funnel with a bit of cotton on a
wire. Draw the ether solution (portionwise if necessary) into a 250-ml. flask
and distill off. While still hot, drain the flask into a small weighed beaker,
rinsing with' a little ether. Evaporate this ether, cool and weigh. (The
unsaponifiable oil from adulterated drying oils is volatile and will evaporate on
long heating. Therefore heat the beaker on a warm plate, occasionally blowing
out with a current of dry air. Discontinue heating as soon as the odor of ether is
gone.)
Iodine Number.-Place a small quantity of the sample in a small weighing
burette or beaker. Weigh accurately. Transfer by dropping about 0.12
grams (0.09-0.15 grams) to a 500-ml. bottle having a well-ground glass stopper,
or an Erlenmeyer flask having a specially flanged neck for the iodine test.
Reweigh the burette or beaker and determine the amount of sample used.
Add 10 ml. of chloroform. Whirl the bottle to dissolve the sample. Add
10 ml. of chloroform to each of two empty bottles like that used for the sample.
Add to each bottle 25 ml. of Wijs solution and let stand with occasional shaking
for one hour. Add 10 ml. of the 15% potassium iodide solution and 100 ml.
of water, and titrate with standard sodium thiosulfate using starch as indicator.
The titrations on the two blank tests should agree within 0.1 ml. From
the difference between the average of the blank titration and the titration on
the samples and the iodine value of the thiosulfate solution, calculate the
iodine number of the samples tested. (Iodine number is centigrams of iodine
to 1 gram of sample.)
Ash.-Tare a porcelain crucible or dish. Add 10 to 25 ml. of oil, weighing
the amount added. Place on a stone slab on the floor of a hood. Ignite by
playing the flame of a burner on the surface of the oil and allow to. burn quietly
until most of the oil is burned off; then transfer to a muffle or over a flame
and continue heating at a very low temperature (not over a dull red) until all
carbonaceous matter is consumed. Cool, weigh, and calculate the percentage
of ash.
Driers.-Moisten the ash with a few drops of water and test with litmus
paper. Record whether neutral or alkaline. Wash any ash adhering to the
test paper back into the crucible. Dissolve the ash in dilute nitric acid to
which a little hydrogen peroxide .has been added. After solution is complete
make up the volume to about 50 ml. with nitric acid and water so that the
final volume will contain about 1 volume of concentrated nitric acid and 3
volumes of water. Boil to remove excess of hydrogen peroxide. Determine
manganese by the bismuthate method. See Chapter on ManganesE., Vol. 1.
1846 ANALYSIS OF PAINTS
Ash another portion of the oil and dissolve the ash as above in nitric acid
and hydrogen peroxide. Determine the lead and c\)balt by any convenient
method.
Time of Drying on Glass.-Flow the oil over Ii p¢rfectly clean glass plate
and allow to drain in a vertical position in a well-ventilated room at a,tempera.-
ture between 15° and 39° C. Mter about 2 hours the film is tested at intervals
with the finger at points not less than 2;!1 cm. from the edges. The film will
be considered dry when it adheres no longer to the finger and does not rub up
appreciably when the finger is rubbed lightly acrosS the surface. With boiled
linseed oil this usually occurs in from '5 to 18 hours.
Reagents for Testing.-The following reagents will be required:
Acetone that will pass the specification of the United States Pharmacopoeia.
Acid Calcium Chloride Solution.-Saturate with calcium chloride a mixture
of 90 parts water and 10 parts concentrated hydrochloric acid (specific gravity
1.2). . ,
Standard Sodium Thiosulfate Solulion.-Dissolve pure sodium thiosulfate
in distilled water that has been well boiled to free it from carbon dioxide in the
proportion so that 24.8'3 grams crystallized sodium thiosulfate will be present
in 1000 m!. of the solution. It is best to let this solution stand for about two
weeks before sta~dardizing. Standardize with pure resubiimed iodine. This
solution will be approximately decinormal, and it is best to leave it as it is after
determining its exact iodine value, rather than to attempt to adjust it to
exactly decinormal strength. Preserve in a stock bottle provided with a guard
tube filled with soda lime.
Slarch Solulion,-Stir up 2 to 3 grams of potato starch or 5 grams soluble
starch with 100 ml. of 1 % salicylic acid solution, add 300 to 400 ml. boiling
water, and boil the mixture until the starch is practically dissolved. Dilute
to 1 liter. '
Potassium Iodide Solution.-Dissolve 150 grams of potassium iodidelhee
from iodate in distilled water and dilute to 1000 ml.
, Wijs Solution.-The preparation of the iodine ~bnochloride solution pre-
sents no great difficulty, but it should be done 1Vith care and accuracy in order
to obtain satisfactory results. There shall be in the solution no sensible excess
either of iodine or more particularly of chlorine over that required to form the
mono chloride. This condition is most satisfactorily attained by dissolving in
the whole of the acetic acid to be used the requisite quantity of iodine, using
a gentle heat to assist the solution, if it is found necessary. ~Dissolve iodine
in glacial acetic acid that has a melting point of 14.7° to 15°-C. and is free
from reducing impurities in the proportion so that 13 g. of iodine-will be present
in 1000 m!. of solution. Set aside a small portion of this solution while pure
'and pass dry chlorine into the remainder until the halogen content of the solu-
tion is doubled. Ordinarily it will be found that by passing the chlorine into
the main part of the solution until the characteri§tic color of free iodine has
just been discharged, there will be a slight excess of chlorine which is corrected
by the addition of the requisite amount of the un chlorinated portion until all
free chlorine has been destroyed. A slight excess of iodine does little or no
harm, but excess of chlorine must be avoided. I
Standard Sodium Hydroxide Solution.- PreplJ,rfra sto,ck concentrated solu-
tion of sodium hydroxide by dissolving sodium 'hydroxide in water in the pro-
ANALYSIS OF PAINTS 1847
portion of 200 grams NaOH to 200 ml. water. Allow this solution to cool and
settle in a stoppered bottl~ for several days. Decant the clear liquid from the
precipitate of sodium carbonate into another clean bottle. Add cl~ar barium
hydroxide solution until no further precipitate forms. Again allow to settle
until clear. Draw off about 175 ml. and dilu~ to 10 liters with freshly boiled
distilled water. Preserve in a stock bottle provided with a large guard tube
filled with soda lime. Determine the exact strength by titrating against pure
benzoic acid (C 6H sCOOH) using phenolphthalein as indicator. This solution
will be approxil)lately one-fourth normal, but do not attempt to adjust it to
any exact value. Determine its exact strength and 'make proper corrections ~
in using it.
Alcoholic Sodium Hydroxide Solution.-Dissolve pure sodium hydroxide in
95% ethyl alcohol in the proportion of about 22 grams per 1000 ml. Let
stand in a stoppered bottle. . Decant the clear liquid into another bottle, and
keep well stoppered. This solution should be colorless or only slightly yellow
when used; it will keep 'colorless longer if the alcohol is previously treated with
NaOH (about 80 g. to 1000 mI.), kept at about 50° C. for 15 days, and then
distilled.
Half Normal Sulfuric Acid Solution.-Add about 15 ml. sulfuric acid
(1.84 specific gravity) to distilled water, cool and dilute to 1000 ml. Determine
the exact strength by titrating against freshly standardized sodium hydroxide
or by any other accurate method. Either adjust to exactly half normal
strength or leave as originally made,. applying appropriate correction.
A
B
---c:.~
I
FIG. 265.-Apparatus for Browne Heat Test.
A. 1000-ml. tall-form oeaker, 9 cm. in C. Corks.
diameter by 19 cm. in height. D. Test tubes 15 cm. by 16 rom.
B. Support plate about 6.5 cm. wide, E. Thermometer, 'range 210 to 310 C.
made of Monel metal, aluminum or F. Glass rods 3 rom. in diameter.
stainless steel. G. Guide to prevent cover from slipping.
,
When the bath temperature is 293° C. (560° F.) and very slOwly rising at
this point, place the tube containing 5 ml. of the oil to be tested and the tube
containing 5 ml. of a control sample of known value into the large holes of the
cover. Note the time. Remove the source of heat for about 45 sec. and then
re-apply. Before 2 min. have elapsed the temper~ture of the bath will have
fallen to 282.2° C. (540° F.), at wliich point it sJiould be kept as steady as
possible. When the samples have been in the bath 9 min. raise the glass rods
at intervals of 15 sec. Note the time when each sample becomes firmly set.
At this period the oil will be so stiff that the entire tube may be lifted by aid
of the rod. As setting or jellying takes place :within a few s'econds of fluidity,
a good end point is afforded. Remove the sampJes..:- Heat the bath again to
293° C. (560° F.), and repeat the experiment with fresh portions of the sample.
ANALYSIS OF PAINTS 1849
No stirrer is used in the bath. A screen around the bath enables the tem-
perature to be more easily reached. When the bath oil has become. tarry and
. viscid, it should be renewed; otherwise heating may be irregular.
High quality tung oil should gel in not over 12 minutes by this test. Raw
oiticica oil requires about 18 to 22 minutes.
Worstall Heat Test.-This test consists in heating a quantity of the oil in
an open container. Various refinements have been suggested by investigators,
such as the use of a polished nickel dish, automatic temperature control,
mechanical stirring, etc. However, the following A. S. T. M. method based
on Worstall's test,- and slightly revised in 1935, is probably as rigorous as
needed.
This test requires (1) an ordinary vitreous enamelled iron casserole having
a bottom diameter of 7.5 cm. (3 in.), (2) a wide flange tripod with a 7.5 cm.
(3 in.) opening (the object of the flange is to prevent super-heating of the sides
of the casserole), (3) a total immersion chemical thermometer, range 30-
3000 C. and 14-15 inches in length.
Rut into the casserole 150 g. of the oil and heat so that the temperature
reaches 282.20 C.±I° (540° F.±2°) in 4 minutes, ±30 seconds, stirring with
the thermometer. Turn down the flame and hold the heat as near 282.2° C.
±1° (540° F.±2°) as possible, stirring with the thermometer, until on lifting
the latter, the oil drops with a pronounced string, showing that polymerization
has started. The time required after reaching 282.2° C.±I° (540° F.±2°)
until the string is noted, is the time of the heat test. For pure tung oils .this
will not exceed eight minutes. As soon as the oil strings, remove the lamp
Sp. Gr. Iodine No. Sap on. No. Acid No. Refrac.
Index
and the thermometer, and stir with a stiff. spatula until the oil is solid.. After
stringing, a pure tung oil will require not over 40 seco:r;tds more to become solid.
When solid, allow to stand just one minute, then turn out, upside down, on
clean paper and cut at once with a clean spatula. :pure tung oil gives a gel
that is dry, not adhering to the spatula when cut, that; is firm, crumbling under
pressure of the spatula without sticking, and the cut ;portions should crumble
under pressure like dry bread crumbs. Adulterated tung oil gives a gel that
is soft, sticky, and which will not crumble.
Examination of Turpentine
RECOMMENDE]) METHODS OF THE A. S. T. M.
Turpentine appears on the market under four types, gum spirits tur- ot
pentine, steam-distilled wood, sulfate wood and destructively-distilled wood
turpentine. The physical requirements for the first three types are identical
except for odor. The requirements of the A. S. T. M. specifications for these
types are:
;
Gum Spirits, Destructively
Steam-distilled Wood, Distilled
Sulfate Wood Wood
/
/'
ANALYSIS OF PAINTS 1851
.ANALYSIS OF VARNISH
0'
real value of a varnish. Other tests that give additional information, some-
times a valuable nature, are as follows: Flash point, acid number, ash,
character of solvent, fixed oil and resins. .
Flash Point.-A nickel or iron crucible of 60 mm. diameter and 40 mm.
height is filled with the varnish to within 20 mm. of the top. It is then sup-
ported in a water bath in such a manner as to be about two-thirds immersed
in the water. The water should be from 15° to 20° C. at the start and should
be heated slowly so that the temperature of the varnish, as indicated by a
3 Made by mixing fuming sulfur.ic acid with co~centrated sulfuric acid. For details
sel' specification of the A. S. T. M. or of the Federal Specification Board.
1852 ANALYSIS OF PAINTS.
thermometer suspended in it, will show a rise of about 1 degree per minute.
Test for flash at each half degree, using a very small.flame.
Acid Number.-Ten to 20 grams of the varnish ar:e weighed into a small
Erlenmeyer flask, 50 ml. neutral alcohol added, and a small funnel inserted in
the neck. Heat on the water bath for one-half hour, :With occasional shaking.
Allow to cool somewhat, add two drops of phenolphthalein indicator and titrate
with tenth-normal potassium hydroxide solution. The acid number is the
number of milligrams of KOH required to neutralize each gram of the varnish.
Ash.-Weigh in a porcelain or fused silica crucible several grams of the
varnish. Burn off over a small Bunsen flame, using great caution to avoid
boiling over and spattering. When all combustible matter is destroyed, weigh
the ash and if desired analyze it. ,
Solvent.-Steam distillation of a portion of the varnish will remove~the
solvents, leaving a residue of fixed oils and varnish resins, which may be
weighed after driving off the water. The distillate should be examined as
recommended under Methods for the Examination of Turpentine. The
amount of mineral spirits and turpentine may thus be determined.
Fixed Oils and Resins.-In the above determination, the total amount of
fixed oils and resins is obtained. It is a difficult matter, however, to determine
the exact percentage and character of resins that have been used in the manu-
facture of the varnish. This is due to the fact that during the process of
heating oils in the presence of resins many intricate chemical changes are
brought about, a considerable portion o( the resins being distilled off in the
form of vapors and combinations of the oil brought about that are difficult of
separation. One of the best methods, however, of separating the fixed oils
and varnish resins is carried out in the following manner.
About a half ounce of the varnish resin should be placed in a 300-ml. tared
beaker. There should then be added about 200 ml. of ice-cold petroleum e~r
and the beaker should be covered and allowed to stand, preferably in a dish
containing ice. In an hour's time the resinous ingredients will be found pre-
cipitated at the bottom of the beaker or adhering to the side thereof (With the
exception of rosin, which is largely soluble in petroleum ether). The precipi-
tated resins should be washed with fresh portions of cold petroleum et.her two
or three times, pouring the decanted portions into a large bottle. The com-
bined portions of petroleum ether may then be filtered through a tared filter,
adding by the aid of a stirring rod the resins contained in the beaker. The
filter paper and the beaker with the resins may then be {fried-at 100° C. and
weighed. The combined filtrates may be distilled to obtain the fixed oil which
may be examined for constants. (This fixed oil may contain Tosin.) The
amount of rosin contained in a varnish may be roughly ascertained by thor-
oughly shaking in a separatory funnel a portion of the varnish witlo a large
quantity of absolute alcohol. The rosin may be ol;>tained by evaporation of
the alcoholic extracts. The fixed oils after oxida~ion or polymerization, as
ca,used by the heating of the varnish during manufacture, are not 'readily
soluble in alcohol.
Separation of Polymerized Oils and Resins.-In the making of varnish,
some oils become oxidized or polymerized to a condition resembling resins.
For instance, when a varnish is examined for .re.sill~b-y the above method, it
will often be found that a considerable amount of matter insoluble in petroleum
/'
ANALYSIS OF PAINTS 1853
ether will be obtained even when hard resins are absent. The insoluble sub-
stance is oxidized or polymerized oil. It may be differentiated from varnish
• resins by the fact that it is readily saponified by alcoholic potash. The follow-
ing method by Boughton (Technologic Paper No. 65, U. S. Bureau of Stand-
ards), though involving considerable work, is probably the most accurate
method for the separation of polymerized oils and resins.
To about 4 grams of varnish in a flask add about 25 ml. of water and boil
until the volume is about 10 mI. This removes nearly all of the volatile.
Add 25 mI. half normal alcoholic potash and 25 ml. benzol and boil under a
reflux condenser for one-half hour. Evaporate the solution to about 15 mI.
and add about 10 ml. of alcohol. Transfer completely to a separatory funnel,
washing the flask with water and ether and using a policeman if necessary.
Dilute with water to about 100 mI., add 100 ml: of ether, and shake. Add a
few mI. of alcohol if necessary to make the layers separate. Draw off the
aqueous layer and wash the ether three times with water and transfer to a
tared flask for future use.
TQ the combined soap solution and washings, add an excess of hydrochloric
acid and extract twice with 50 mI. of ether. Discard the aqueous layer, wash
the combined ether extracts with water, transfer to a flask and distill off the
ether. To the dry residue add 20 mI. of absolute alcohol and 20 mI. of a
mixture of 1 volume of sulfuric acid and 4 volumes of absolute alcohol and boil
for two minutes under a reflux condenser. Completely transfer the contents
of the flask to a separatory funnel, washing the flask with water and ether.
Add 100 mI. ether and after agitation add 100 ml. of 10 per cent sodium chlo-
ride solution and again shake. Draw off the aqueous layer, extract it with
50 mI. of ether, combine the ether solutions and wash with water. Add 5Q mI.
of a fifth-normal potassium hydroxide solution and 10 ml. of aicohoi, shake
and draw off the lower layer into a second funnel. Wash the ether layer with
50 ml. of water containing 5 mI. of the potassium hydroxide solution and 5 ml.
of alcohol. .
Extract the combined aqueous portions with two 50 ml. portions of ether
and finally wash the combined ether solutions (containing the ethyl esters of
the fatty acids) with water.
Distill off the ether and boil the residue with 25 mI. of half-normal alcoholic
potash for one-half hour under a reflux condenser. Transfer completely to a
separatory funnel and extract the soap four times with ether. Wash the com-
bined ether solution twice with water and add it to the first ether solution of
unsaponifiable matter obtained.
Unite the solution and washings containing the soaps, add an excess of
hydrochloric acid, and extract twice with ether. Transfer to a tared flask the
combined ether solutions, after washing them with water, distill off the ether,
dry the residue to constant weight at 110° C. and weigh as " fatty acids."
Report the percentage of fatty acids as percentage of oil and calculate the
percentage of resin by difference.
1854 ANALYSIS OF PAIN:rS
, .
OTHER MATERIALS I
For detailed methods other than thpse given in ~his book for the examina-
tion of shellac, resins, bitumens, and other like materials for the vehicle portion
of paints and varnishes, the following references maYi be consulted:
A. S: T. M. Standards on Preservative Coatings for Structural Materials. American
Society for Testing Materials, Philadelphia, Pa.
Physical and Chemical Examination of Paints, Varnishes, Lacquers and Colors, 9th
Edition, by Henry A. Gardner. Henry A. Gardner Laboratory, Inc., 2201 New
York Avenue, N.W., Washington, D. C.
Federal Specifications of the Procurement Division of the U. S. Treasury Depavtment.
Superintendent of Documents, Government Printing Office, Washingtonjl D. C.
The liquid having been extracted from the paint under examination, by
the previously outlined methods, the pigment is left ready for analysis. The
pigment can readily be classified under one of the following heads by its color,
thus shortening any preliminary examination. Many of the colors have a
white base which necessitates a determination of both the colored portion of
the pigment and any white base which may have been. used.
The general analysis of colored pigments is carried out according to the
specific method outlined for the individual colored pigments, together with the
methods for a composite white paint, provided,a qu~litative examination does
not directly reveal the identity of the pigment.!
The pigments used in the manufacture of paints are classified as follows,
in certain instances the trade names being given by which the particular pig-
ments are known.
White Pigments
Lead Pigments
,
Basic Sulfate of Lead-Basic Sulfate White Lead.
Basic Carbonate of Lead-Old Dutch Process White Lead-.:..QuicknProcess
White Lead-Mild Process White Lead.
Zinc Lead-Leaded Zinc.
Zinc Pigments
Zinc Oxide-:-Zinc White.
Lithopone.
Titanated Lithopone.
ANALYSIS OF PAINTS 1855
Other Opaque White Pigments
Titanium Oxide.
Titanium-Barium Pigme~t.
l' Titanium-Calcium Pigment.
Antimony Oxide.
Lead Titanate.
Zirconium Oxide.
Silica Pigments
Silica-Silex.
Asbestine-Talcose.
China Clay-Kaolin.
Calcium Pigments
Whiting-Paris White-Chalk-Alba Whiting-Spanish White.
Gypsum-Plaster of Paris-Terra Alba-Agalite.
Barium Pigments
Barytes-Barite-Blanc Fixe-Barium Sulfate.
Barium Carbonate-Witherite ..
Red and Brown Pigments
Red Lead-Orange Mineral.
. Vermilion.
Ochres-Tuscan Red-Indian Red-Venetia'i! Red.
Umbers-Sienhas.
Blue Pigments
Sublimed Blue Lead.
Ultramarine Blue-Phthalocyanine Blue.
Prussian Blue-Antwerp Blue-Chinese Blue.
CALCULATION OF COMPOSITION
Using the average total of 99.70%, the total lead found and the zinc oxide
content, the composition of this pigment is determined by the following calcu-
lation:
Total percentage of lead compounds present
'=total percentage found of ZnO, PbO and PbS0 4 -percentage of ZnO.
Total percentage of lead compounds present
=99.70% (average total) -:percentage ZnO.
Atomic weight lead ........"................... : .......... 207.2
Molecular weight lead oxide ......................... : . ... 223.2
Molecular weight lead sulfate . ................. '" ........ 303.2
Determining the percentage of lead oxide and lead sulfate present by the
above formulas we find:
303.1 )
( 207.1 X69.00 -95.00
303.1-223.1 per cent PbO = 16.68
223.1
Basic Carbonate White Lead.7 Old Dutch Process White Lead. Quick
Process White Lead
Basic carbonate white lead contains approximately 80% metallic lead and
20% carbonic acid and combined water with ~races sometimes, of silver, anti-
mony and other metals. The formula for this compound is 2PbCOs·Pb(OH)2.
Carbon Dioxide
. . I
Determine by evolution with dilute hydrochloric acid, absorbing the carbon
dioxide in soda-lime or KOH solution. Calculate Cq2 to PbCO a, subtract PbO
equivalent from total PbO and calculate residual Pb? to Pb(OHh
Acetic Acid 9
Place 18 grams of the pigment in a 500-ml. flask, add 40 ml. of sirupy phos-
phoric acid, 18 grams of zinc dust and 50 ml. of water. Connect to a straight
Liebig condenser, apply heat and distill down to a small bulk. Then pass steam
into the flask until it becomes about half full of cond~nsed water, shut' off the
steam and distill down to a small bulk-this operation being conducteJ twice.
To the total distillate which was collected in a larger flask add 1 ml. of sirupy
phosphoric acid, connect to a Liebig condenser, using a spray trap, and distill
to a small volume-about 20 ml. Pass steam through till about 20;) ml. of
water condenses in the distillation flask, shut off steam and continue the dis-
tillation. These operations of direct and steam distillations are cdnducted
until 10 ml. of the distillate require only 1 drop of 0.1 N alkali to give a change
in the presence of phenolphthalein. Then titrate the total di,stillate with'
0.1 N sodium hydroxide and phenolphthalein and calculate the total acidity
as acetic acid. It will be found convenient to titrate each 200-ml. portion of
the distillate as collected.
Metallic Lead
Weigh 50 grams of the sample into a 400-ml. beaker, add a little water and
add slowly 60 ml. of 40% acetic acid and after effervescence has ceased, boil
on hot plate. Fill the beaker with water, let settle, and decant the Clear
solution. To the residue add 100 ml. of a mixture of 360 ml. of conc. NH 4 0H,
1080 ml. of water, 2160 ml. of 80% acetic acid, and boil until all solution is
complete. Fill the beaker with water, let settle and decant the clear solution.
Collect residue on 'a watch-glass, floating off everYthing but metallic lead.
Dry and weigh. Result X2 = percentage of metallic lead in sample.
The following method of A. N. Finn (unpublished) gives total basicity of
a pure white lead: Place 2 grams of pigment in an evolution flask, add a little
CO 2-free water, connect with a separatory funnel and condenser (Knorr type),
add through the funnel, finally washing down, 100 Illl. of N/4 nitric acid,
boil and absorb the CO 2 in a soda lime tube in the usual manner, (having H 2SQ4
and CaCl 2 drying tubes in train) and weigh. To the solution in 'the evolution
flask, add about 20 ml. of neutral sodium sulfate solution and titrate with
N/4 sodium hydroxide solution (carbonate-free), using phenolphthalein. CO 2
is calculated to PbCO a• The amount of N/4' acid corresponding to the CO 2
is calculated and deducted from the total amount of N/4 acid neutralized by
the sample and the difference calculated to combin~d H 20, from ~hich Pb(OH)2
is computed.
9 Thompson's Method, Jour. Soc. Chern. Ind., 24, 487 (1905).
I
ANALYSIS OF PAINTS 1861
Zinc lead and leaded zinc are varying compounds containing zinc oxide
and leJ1d sulfate, the former showing approximately 50% zinc oxide and 50%
lead sulfate, while the latter contains on an average 25% lead sulfate and
75% zinc oxid,e. The most widely used type is probably that with 35% basic
lead sulfate. See also pp. 509, 1060, Vol. 1.
These pigments may be analyzed by the following procedure:
Moisture.-Heat 2 grams at 1050 C. for two hours. Cool and weigh.
Lead and Zinc.-Determine the lead directly by the method given under
Basic Carbonate of Lead, and the zinc by the volumetric ferro cyanide method
as outlined under Sublimed White Lead. See also pages 509 and 1060, Vol. I.
Total Soluble Sulfates 10 (in the absence of BaS04).-Treat 0.5 gram of
the sample with 5 m!. of water, 3 grams of NH 4 CI and 5 m!. of HCI saturated
with bromine; digest (covered) on the steam bath about fifteen minutes, add
25 ml. of H 20, neutralize with dry Na 2CO a and add about 2 grams more. Boil
ten to fifteen minutes; let settle, dilute with hot water, filter and wash with
hot water; redissolve in HCl, reprecipitate as above and wash thoroughly with
• hot water. Acidify the united filtrates with HCl and add a slight excess of
10% BaCl 2 solution. Let stand on steam bath for one hour, filter, wash with
hot water, ignite and weigh the BaS04. Calculate to SOa (includes 80 a formed
from S02).
Total Soluble Sulfate (in the presence of Ba804).-Treat 1 gram in a
600-m!. beaker with 10 ml. of H 20, 10 ml. of strong HCI, saturated with bro-
mine, and 5 grams of NH 4CI, heat on a steam bath in a covered beaker for five
minutes, add hot water to make about 400 ml., boil for five minutes and filter
to separate any insoluble material. (A pure pigment should be completely
dissolved.) Wash with hot water; ignite and weigh the insoluble. Remove
lead with 'N a 2CO a as above, making a double precipitation, acidify, and to the
boiling hot filtrate add slowly, with stirring, 20 m!. of a 10% BaCl 2 solution;
let stand for two hours on the steam bath, filter, wash, ignite, and weigh as
Ba80 4. (Includes 80 a formed from 80 2 .)
Soluble Zinc Sulfate.-Heat nearly to boiling 2 grams of the sample with
150 m!. of water and 50 ml. of alcohol for thirty minutes, filter and wash with
a mixture of alcohol and water (1 : 3). Heat the filtrate to boiling and expel
most of the alcorol; then determine SOa by the usual method of precipitation
with BaCl 2. Calculate to Zn80 4 and to 80 a.
Sulfur Dioxide.-Digest 2 grams of the sample with frequent stirring in
100 ml. of freshly boiled cold water and 5 ml. of concentrated HCl; let stand
ten to fifteen minutes, add an excess of 0.01 normal iodine solution and titrate
back with 0.01 normal sodium thiosulfate solution, using starch indicator.
Report as 80 2 • Run blank on reagents and make corrections.
Calculations.-Report soluble 80 a as ZnS04. Deduct ZnO equivalent of
. ,the Zn80 4 from total ZnO and report residue as ZnO. Deduct soluble 80 a
10 Report of 'Sub-committee VIII of Committee D-l, Proceedings of American
Society for Testing Materials, 14, 271-2, 1914.
1862 ANALYSIS OF PAINTS
ZINC OXIDE
LITHOPONE
/
Lithopone, in the restricted' sense, is a mixture of equimolar quantities of
zinc sulfide and barium sulfate. The commercial pigment contains approxi-
mately 28% of the former ingredient and 72% of the latter. Lithopone is
produced in several modifications, the principal ones being those in which the·
extending portion is calcium .sulfate or magnesigm silicate. ~High strength
litho pone contains from 45 to 60% of zinc sulfitle.
ANALYSIS OF PAINTS 1863
Moistu.e.-Heat 2 grams'f~r two hours at 1050 C.
Barium Sulfate.-Treat 1 gram with 10 ml. conc. HCI and 1 gram of
potassium chlorate, added in small amounts. Evaporate to one-half its vol-
ume, add 100 ml. hot water and a few ml. of dilute H 2S0 4 • Boil, filter, wash
and weigh the insoluble residue, which should show only the presence of barium
sulfate. Examine the residue for silica and alumina.
Zinc Sulfide.-Determine the zinc in the pigment by the method given
under Sublimed White Lead. Calculate to zinc sulfide. From this value
subtract the zinc sulfide equivalent of the zinc oxide as determined below.
Zinc Oxide.-Digest 1 gram of the pigment with 100 ml. of 3% acetic acid
solution for 18 hours. Stir several times during digestion, especially during
the first hour. Filter, wash with 3% acetic acid and determine the zinc by
the usual method.
Soluble Salts.-Digest 2 grams with hot water and examine the filtrate for
soluble salts.
TITANWM PIGMENTS
The titanium pigments used in paints are: titanium dioxide, barium base
pigment, which is a mixture of 25 parts of titanium dioxide and 75 parts of
barium sulfate, calcium base pigment, which is a mixture of 30 parts of titanium
dioxide and 70 parts of calcium sulfate, titanated lithopone, which is a mixture
of 85 parts of lithopone and 15 parts of titanium dioxide, and lead titanate.
A mixture of titanium dioxide and magnesium silicate is sometimes found.
Barium Suifate.-Weigh Y2 gram sample into 250-ml. Pyrex glass beaker;
add 20 ml. concentrated sulfuric acid and 7 or 8 grams ammonium sulfate. Mix
well and heat on hot plate until fumes of sulfuric anhydride are evolved and
then heat directly ov~r flame to boiling for five minutes or until solution is
complete. Traces of silica, if any, remain as an insoluble residue.
Cool, take up with 100 ml. of water, boil and filter off barium sulfate and
silica, washing with 5% sulfuric acid to free residue from titanium. Use the
filtrate for the determination of titanium dioxide.
Calcium Sulfate.-Extract the calcium sulfate QY boiling 0.5 gram of the
pigment in a 250-ml. beaker with 25 ml. of hydrochloric acid (1 : 1) for several
minutes. Dilute to 150 m!. with water, boil and filter. Wash the residue
with hot hydrochloric acid solution (10 m!. of HCI sp.gr. 1.19 and 990 ml. of
water) until the washings show no. calcium, when tested with NH 40H and
(NH 4 )2C 20 4 • Determine the calcium in the filtrate by the oxalate method.
. Titanium Dioxide.-Determine the titanium in the filtrate from the barium
sulfate reducing it in a Jones or similar reductor and titrating the reduced
solution with a standludized ferric ammonium sulfate or potassium perman-
ganate solution.
1864 ANALYSIS OF PAIlITrS
ANTIMONY OXIDE
Antimony trioxide.,(Sb 20 3), a fume pigment that has recently been intro-
duced, may be occasionally found in certain classes of mixed paints. After
being brought into solution, it may be quantitativ~ly estimated by oxidation
with permanganate or iodine. ~
ZIRCONIUM 'PIG~ENTS 11
These pig,ments, while all true :;ilica, pigments, lare widely different in
physical structure. A microscopic examination is of great value.
The particles of quartz or silex are sharp and irregular, those of china clay
are flat, and those of asbestine are rodlike. Diatom~ceous earth may some-
times be recognized in siliceous pigments. The color, fineness, and other such
properties are more important than chemical analysis. However, if desired,
determinations of silica alumina, magnesia, etc'., may be made by recognized
methods. ~
Moisture.-Heat 2 grams at 1050 for two hours.
H 2S0 4 and HF. Any loss is reported as silica. Determine the BaS04 in the
residue by boiling with dilute Hel or by making a potassium bisulfate fusion.
The residue remaining after either of these treatments is reported as barium
sulfate. \ •
Total Lead.-This constituent can best be determined on a separate sample.
To 1 gram add 10 ml. of conc. HNO a, boil, add, after,cooling, conc. H 2S0 4 and
evaporate to strong SOa fumes. Dilute with water, allow to stand several
hours, filter, wash slightly, dissolve and determine the lead volumetrically as
outlined under Sublimed White Lead.
Lead can also be determined on the combined filtrates from the insoluble
residue. Precipitate the lead in an acid solution with H 2S and deter-mine as
directed under Basic Carbonate of Lead. Reserve the filtrate for the deter-
mination of iron, aluminum, zinc, calcium and magnesium.
To determine whether both sublimed white lead and corroded white lead
are present, treat a separate portion of the paint witli boiling acetic acid, filter
and collect the insoluble residue. Determine the lead either in the filtrate or
in the insoluble residue by the volumetric method. The lead soluble in acetic
acid is the basic carbonate of lead and the lead oxide from the sublimed white
lead, while the lead sulfaj:; from the sublimed white lead remains insoluble.
Alumina and Iron Oxide.-Remove the,H 2S from the filtrate from the lead
sulfide, above, by boiling, oxidize the iron with a few drops of HNO a, and
precipitate the hydroxides with ammonia. Separate and determine the iron,
and aluminum in the usual manner. .-'
Zinc.-Acidify the filtrate from the alumina and iron precipitation with
acetic acid, separate the zinc with H 2S and determine as outlined under Sub-
limed White Lead.
Calcium and Magnesium.-Determine the calcium and magnesium in the
filtrate from the precipitation of zinc sulfide in the usual manner, first testing,
however, for the presence of barium.
Sulfate.-Determine as outlined under Zinc Lead and Leaded Zincs.
Sulfide.-Should lithopone be present, separate/.the zinc oxide a'nd zinc
sulfide as outlined under Lithopone.
Carbon Dioxide.-Determine as outlined under Basic Carbonate of Lead.
Calculations.-Silica is reported as silica, except where alumina is present,
showing the presence of China clay. In this case, calculate the alumina to clay
by the following method. '
Weight of A1 20 aX 2.5372 = weight of clay.
Weight of clayX0.4667 = weight of Si0 2 in clay.
Any difference greater than 5% may be considered silica.
Barium sulfate is reported as barium sulfate or as lithopone, if zinc sulfide
is present, according to the given composition 'of lithopone, 70% barium sul-
fate and 30% zinc sulfide.
Lead is reported as Basic Carbonate of Lead on the formula 2PbCO a·
Pb(OH)2, .
Calculate lead soluble in acetic acid, after determining CO 2, to basic lead
carb'onate and any residual lead to lead oxide which, together with the lead
sulfate is reported as Sublimed White Lead. I
Should calcium sulfate be present; the portiQn...sollible in water is examined
for lime or sulfuric acid and calculated to calcium sulfate, any residual lime
ANALYSIS OF PAINTS 1869
being calculated to calcium· carbonate and any residual sulfuric acid being
calculated to lead sulfate.' Any residual CO 2 after calculating ciJJcium carbo-
nate is calculated to white lead and any residual lead is calculated to lead oxide.
Lead oxide should not be reported except in the presence of lead sulfate.
Any large percentage of magnesium indicates the presence of asbestine.
The red lead may also be examined for zinc, carbon dioxide, and soluble
sulfate. . '
A second method for determination of the lead peroxide or true red lead
content is somewhat shorter,u
Treat 1 gram in a beaker with 15 ml. of nitric acid, sp.gr. 1.2 (110 ml. ,nitric
acid, sp.gr. 1.42 to 100 mi. of water). Stir the sample until all trace of red
color has disappeared. Add from a calibrated pipette or burette exactly 10 ml.
of dilute hydrogen dioxide (1 part of 3% hydroge~ dioxide to 3.5 parts of
water). Add about 50 mi. of hot water and stir un'til all the lead dioxide has
passed into solution. In the case of some coarsely/ground oxides the contents
of the beaker may have to be gently heated td effect complete solution. After
the oxide has completely passed into solution, dilute with hot water to about
250 ml. volume and titrate directly with a st!lndard potassium permanganate
solution, having an iron value of 0.005. Titrate to theJaint pink permanga-
nate color. A blank titration on the hydrogen dioxide~ solution must now
be made. _
Into a 'beaker pour 15 ml. of nitric acid of above strength and add exactly
the same amount of hydrogen dioxide (10 mI.). Dilute to 250 ml. with hot
water and titrate with standard potassium perinanganate solution to a faint
pink color. I
The difference between the number of ml. of 'potassium permanganate re-
quired for the blank titration and the number required for the red lead titration
is the amount required for the hydrogen dioxide which was ,reacted on by the
12Dingl. Polyt. Jour., 246, 196. I
13Zeitschrift fUr analytische Chemie, 26, 296. ,::.- - ,
a" Analysis of Lead and Its Compounds," SChaeNer-and White and Calbeck, p. 77.
ANALYsis OF PAINTS 1871
red lead. The difference between the two amounts of potassium permanganate
required multiplied by 3.058 grams gives the percentage of red lead present.
The difference multiplied by 1.067 gives the percentage of Pb0 2 present.
VERMILION
The following portion of Walker's 15 method will suffice for the examination
of this pigment. Should the analyst desire to determine the sulfide of mercury
presep.t or make a more complete examination-reference may be made to the
original method.
True vermilion, or, as it is generally called, English vermilion, is sulfide
of mercury. On account of its cost it is rarely used in paints, and is liable to
gross adulteration. It should show no bleeding on boiling with alcohol and
water and no free sulfur by extraction with carbon disulfide. A small quantity
mixed with five or six times its weight of dry sodium carbonate and heated in
a tube should show globules of mercury on the cooler portion of the tube. The
best test for purity is the ash, which should be not more than one-half of 1%.
Makll the deterniination in a porcelain dish or crucible, using 2 grams of the
sample. Ash in a muffle or in a hood with a. very good draft, as the mercury
fumes are very poisonous. It is seldom necessary to make a determination of
the mercury. .
Genuine vermilion is at the present time little used in paints. Organic
lakes are used for most of the brilliant red, scarlet and vermilion shades. These
organic coloring matters are sometimes precipitated on red lead, orange mineral
or zinc oxide; but as a usual thing the base is barytes, whiting or China clay.
Paranitraniline red, a compound of diazotized paranitraniline and beta-naph-
thol, is largely employed; but a number of colors may be used.
Paranitraniline red is soluble in chloroform. It is also well to try the solvent
action on different reds, of sodium carbonate, etc. The amount of organic
pigment present in such reds is generally very small, and when it cannot be
determined by ignition owing to the presence of lead, zinc or carbonate, it is
best determined by difference.
15 P. H. Walker, Bulletin 109, Revised, Bureau of Chemistry, U. S. Dept, of Agri.,
pp.31-33.
1872 ANALYSIS OF PAINTS
IRON OXIDES
The iron oxides and manganese oxide pigments include the ochres, umbers,
siennas, Venetian red, metallic brown, Indian red aJid Tuscan red.
In analyzing these pigments, the following cons'tituents are sought; mois-
ture, loss on ignition, insoluble residue, iron oxide, xrtanganese dioxide, calcium
and magnesium oxides and sulfur trioxide.
Owing to the similarity of the methods used for the analysis of these pig-
ments to those used in the analysis of iron ores, the analyst is referred to p. 469
on the Analysis of Iron Ores, or to the method of Walker.l6 .
BLUE ,PIGMENTS
The principal blue pigments used in paints are Prussian, ultramarine, blue
lead, and phthalocyanine.
Prussian blue is the double iron and-potassium salt of hydroferrocyanic and
hydroferricyanic acids. .
Ultramarine blue is essentially a silicate and sulfide of sodium and aluminum.
Sublimed blue lead is the fume product resulting from the smelting of'lead
ores. In composition it consists ,of lead sulfate, lead sulfide, lead sulfite, lead
oxide and zinc oxide, with occasional traces of carbon. It is finding its great-
est use as an inhibitive pigment for the protection of iron and steel. Its color
is a bluish gray.
Phthalocyanine blue is a copper compound of pp,thalocyanine and may be
identified by the'presence of' copper. i
ULTRAMARINE BLUE
The pigments under this class all contain chromates, with the exception of
orange mineral, which is analyzed as under Red Lead. Frequently they con-
tain lead sulfate and sometimes lead carbonate. A pure chrome yellow should
contain only lead chromate and insoluble lea~ _~ompoUJlds. Owing to the
18 "The Chemical Analysis of Lead and its ,eompounds," Schaeffer and White
and Calbeck, p. 35.
ANALYSIS OF PAINTS 1875
frequent use of organic colors to brighten up the pigment, it is essential that a
test be made for organic colors as outlined under VermiJion.
The analysis of these pigments is carried out in the following manner:
Moisture.-Heat 2 grams at 1050 C. for two hours.
Insoluble Residue.-Treat 1 gram with 25 ml. of concentrated HCI, boil
and during the boiling add a few drops of alcohol, one at a time. The solution
is diluted to 100 mI., ,the boiling is continued for ten minutes and any insoluble
residue is filtered off, thoroughly washed and examined for silica, barium sul-
fate and alumina.
Lead.-The solution is nearly neutralized with NH40H and the lead is
precipitated as PbS with H 2S. Filter off the precipitate of PbS, dissolve in
HNO a, add H 2S0 4 , boil to strong fumes and determine as outlined under
Sublimed White Lead or weigh as PbS0 4•
Chromium.-The filtrate from the lead precipitation is boiled until all the
H 2S is driven off. The solution is rendered alkaline with NH 40H and the
chromium is precipitated and determined as chromic oxide. Calculate to
chromic anhydride.
If iron or aluminum is present, it will b~ necessary to separate the chromium
by oxidizing it with Na 202 to chromate, and then titrating it with ferrous
sulfate. .
Zinc, Calcium and Magnesiunt.-Precipitate the zinc in the filtrate with
H 2S and determine:as previously outlined.
In the filtrate frop! the zinc precipitation, determine the calcium and mag-
nesium in the usual manner.
If any carbonates are present; determine by the evolution method.
Sulfuric Acid.-Determine the total sulfate as outlined under Zinc Lead
and Leaded Zinc on p. 1861.
Calculations.-Any chromic anhydride is calculated to lead chromate, sul-
furic acid to lead sulfate, if calcium sulfate is absent, and any residual 'lead
is calculated to lead oxide.
GREEN PIGMENTS
Chrome Green
Green pigments are usually mixtures of chrome yellow and Prussian blue,
though organic color is sometimes present, which may be determined by an
extraction with alcohol or chloroform.
A microscopic examination should be made to determine whether the green
is a combined precipitation product, which is of the greater value, or one mixed
after separate precipitation. A good green will show the presence ,of green
1876 ANALYSIS OF PAINTS
and _blue particles, while a poor green will show yellow and 'blue particles mixed
with green. The analysis may be carried out as fol~ows: 19
Moisture.-Heat 2 grams at 1050 C. for two hours.
Insoluble Residue.-Heat 1 gram at a low heat until the blue color has
been decomposed, keeping the temperature sufficiently low so as not to render
~ny of the iron or lead chromate insoluble. Determine the insoluble residue
as outlined under Yellow Pigments.
Lead.-Determine as outlined under Yellow Pigments.
Iron, Alumina and Chromiu:a.-A-ll the H 2S is expelled from th~ filtrate
after the lead precipitation by boiling. Add a few drops of HN0 3, boil a few
minutes and precipitate the aluminum, iron and chromium hydroxides with
NH 4 0H. Filter, wash, dissolve the precipitate in HCI, and make up the solu-
tion to a definite volume.
In one portion the three hydroxides are precipitated together with NH 4 0H
and weighed. Another portion is treated in a flask with an excess of N a 202
until the iron hydroxide has assumed its characteristic reddish-brown color.
Dilute with water, filter, wash and determine the iron in the usual way. Ren-
der an aliquot of the filtrate from the iron precipitation acid with -HN0 3 ,
precipitate the aluminum hy~roxide with N:H 40H and weigh as AI20~.
Chromium is determined :il the filtrate by titration with ferrous ammonium
sulfate as described under Yellow Pigments. Any method for the separation
of the above hydroxides may be used in place of the one. outlined.
Calcium and Magnesium.-These cO"nstituents are determined in the filtrate
from the precipitation of the above hydroxides.
Sulfuric Acid.-One gram after ignition until all the blue has been decom-
posed, is dissolved in 30 m!. of cone. HCI, diluted with water, boiled, filtered,
and washed. The sulfuric acid is determined in the filtrate.
Nitrogen.-Determine as outlin:ed under Prussian Blue.
I Calculation.-The Prussian blue is determined by using the factors given
under Prussian Blue. The sulfate is calculated to lead sulfate and calcium
sulfate, should calcium be present, and the chromiu_{n to lead chromate.
I
BLACK PIGMENTS
The black pigments irtclude those which contain carbon as their essential
constituent. The introduction of many black pigments which are made from
asphaltic and coal-tar mixtures complicates their chemical analysis. For those
pigments which contain coal-tar mixtures, recourse: may be had to works 20
covering this matter thoroughly.
19 "The Analysis of Paints," Gardner and Schaeffer, __ pp.
"36-37. -
20Allen's "Commercial Organic Analysis," 4th:Edition; "'The Analysis of Paints ,-
Gardner and Schaeffer.
ANALYSIS
, OF PAINTS 1877
The analysis of the simple black pigments may be carried out in the follow-
ing way:
0
Moisture.-Dry 2 grams at 105 C. for two hours.
Oil.-Extract 2 grams, with ether in a fat-extraction apparatus.
Carbon.-Determine the carbon by difference after determining the mois-
ture, oil and ash.
Ash.-Ignite 2 grams to a bright red heat until all the carbon is driven off.
If graphite is present, the ignition must be carried out with the aid of oxygen.
Should carbonate be present, mix the ash with a small amount of ammonium
carbonate and again ignite, thus reconverting to carbonate any oxide which
may have been decomposed.
Analysis of Ash.-The ash is boiled with concentrated HCI and the insoluble
residue determined in the usual manner. The filtrate is examined for calcium,
magnesium and phosphoric acid.
Calculate the magnesium to phosphate, any residual phosphoric acid to
calcium phosphate and a~y residual calcium to carbonate.
PAPER AND PAPER-MAKING (MATERIALS ,
I -
STANDARD METHODS OF THE TECHNICAL ASSOCIATION OF
THE PULP AND PAPER INDUSTRY 1
perature of the water in the water jacket should be in the range of 23.5° to
32° C., inclusive. During the first chlorination samples of wood absorb ap-
proximately 230 m!. of Cl gas at room temperature and atmospheric pressure.
This necessitates refilling the burette, which may be done quickly as it is
connected to a Cl tank with a 3-way stop cock. The first chlorination treat-
ment requires from 3 to 4 minutes, after which remove the crucible from the
apparatus and wash the material with approximately 50 ml. of distilled water,
and successively with. 50 m!. of approximately 3% 80 2-water, 50 m!. of water,
then 50 ml. of freshly prepared 2% sodium sulfite solution. Transfer the
material to a 250-m!. Pyrex beaker, using a pointed glass rod. The last traces
of material are washed out of the crucible with 100 m!. of 2% N a 280 3 solution'
with the aid of a rubber policeman drawn gently over the bottom of the crucible,
Stop CI)cfr8
J\ic~=lI-..W4te" Cliff, I'
Thermomtfer
~top CockC
V,X(p-a)
Vo 760(1+0.00367t)
Where V,=volume of gas observed at t deg. C.
Vo=volume under standard conditions.
a=vapor pressure of water vapor over saturated CaCh solution sat-
urated with Cl gas, in mm. mercury at t deg. C.
t=the temperature of chlorination in deg. C.
p=observed barometric pressure in mm; of mercury.
The following table'is of assistance in making these calculations.
tOe. a. 760 X(l +0.00367 t)
23.5 15.9 826.00
24.0 16.6 826.90
24.5 17.3 828.36
25.0 18.0 829.76
26.0 19.4 832.57
26.5 20.0 833.80
27.0 20.8 835.40
28.0 22.0 838.12
29.0 23.5 840.96
30.0 25.0 843.66
31.0 26.4 846.48
32.0 29.0 849.30
The weight of 1 ml. of Cl gas at standard temperature and pressure = 0.003164 gram.
Titrate the combined initial wash waters with standard alkali to determine
the weight of Cl gas that formed HCI. Then determine by difference the
weight of Cl actually consumed by the pulp.
Report.-The data shall be reported as the percentage of Cl consumed,
calculated on the oven-dry weight of the sample.
ture content throughout the material. Samples for moisture content and
alpha cellulose determinations shall be withdrawn a$ needed.
Reagent.-17.5% NaOH Solution.-The sodium hydroxide solution is
prepared by dissolving solid NaOH in an equal weight of water, and allowing
it to stand until the N a 2 C0 3 and other impurities' h~ve settled, leaving a clear
supernatant liquid. The clear liquid is decanted and diluted with CO 2-free
water until its density at 15° C. is 1.197. Such a s61ution contains 17.5±0.1
grams of sodium hydroxide per 100 grams of solution. ..
Procedure.-Weigh accurately in triplicate about 3 grams in a weighing
bottle and transfer to a 250-ml. Pyrex beaker. Add 30 ml. of 17.5% carbon-
ate-free N aOH solution and allow the whole to stand for 5 minutes. Macerate
the pulp for 10 minutes with a short rod which has the end flattened to a disk
1 cm. in diameter. During this maceration add 40 ml. of 17.5% NaOH solu-
tion at 20° C. intermittently in 10-ml. portions. Then ·cover the beaker with
a watch glass. After 30 minutes additional mercerizing action in a watr bath
at 20° C. (total mercerization time 45 minutes) add 75 ml. of distille water
(at 20° C.) to the alkali-cellulose mixture with thorough stirring. Fi ter the
contents of the beaker immediately by means of suction on a 40-ml. Gooch
crucible having a finely perforated bottom, allowing the cellulose to form it\'!
own mat. Pour the filtrate through the mat a second and third time, if
necessary, to catch any fine material. Wash the residue in the Gooch crucible
with exactly 750 ml. of distilled water (at 20° C.) by means of suction. Dis-
connect the suction tube, add 40 ml. ot 10% acetic acid (at 20° C.) and allow
to soak for 5 minutes. Then apply suction to remove the acid. Wash the
alpha cellulose with distilled water until free from acid. Carefully remove it
from the Gooch crucible to a tared, flat, glass-stoppered weighing bottle. Dry
the unstoppered bottle and contents to constant weight at 105° C., making the
first weight at the end of 6 hours and then at I-hour intervals until constant.
Let'the weighing bottle and its contents cool in a desiccator for 30 minutes
and cover before weighing.
I
NOTE.-Drying of the alpha cellulose can be gr!latly expedited if the last washing is
followed by 95% ethanol and then by ether or some other highly volatile non-reactive
liquid. In each case the washing fluid is allowed, to soak into the sample for several
minutes before being removed by suction. In thi~ way the time in the oven at which
constant weight is reached may be reduced to about 2 hours instead of 6 hours or more.
The results of the two drying methods check each other within the limits of experimental
error.
~,
treat with 45 ml. of H 2S0 4, sp.gr. 1.695 (76.76% by weight), at 15° C. and let
the digestion proceed at 25° C. for 16 hours. Transfer the mixture to a 2-liter
Erlenmeyer flask,. dilute with 1570 ml. of water, cover with a watch glass and
boil for 2 hours, maintaining a constant level by the addition of boiling water.
Filter the residu€"through a tared Gooch or sintered glass crucible, wash with
hot distilled water and dry to constant weight at 100-105° C. Correct the
alpha cellulose figure by subtracting the percentage of lignin residue thus
obtained. ~
N OTEs.-(a) 'Time interval for steps in the procedure: An interval of 15 rniAutes should
elapse between the times of starting the alkali treatment of the individual triplicate
samples. Such a procedure allows time before proceeding to the next sample in a tripli-
cate determination for the various steps in the procedure, such as impregnation, macera-
tion, filtration, and partial washing.
If 5 minutes are allowed between the addition cf a part of the alkali and the beginning
of maceration, it will be found that sulfite pulp swells to five or six times its original
volume, and the fibers can be much more easily macerated. The remaining H)-minute
interval, during whicll the remaining 40 ml. portion of the alkali is added, making a '
total of 75 mI. h,Jl;s been found ample time for the maceration of pulps. 'Vith'such a
scheme, for exH.'mple, one would add the alkali to the samples at 9:00 o'clock, 9:15,
and 9:30, respectively. Filtering would begin at 9:45 o'clock; 10:00, and 10:15, re-
spectively.
(b) FiUering lIrfedium: Jena glass filter crucibles (coarse) may be used in place of the
Gooch crucibles specified in the directions.
(c) Testing for acidity: To test the acidity of the wash water on litmus paper, the
last few drops removed after the addition of water should be used.
Report.-Report results on the oven-dry basis to tenths of a per cent.
Precision.-Duplicate samples should check each other to within 0.2.
f I
A-~-l!-_~I
.090 em.
---
Int.diam.
FIG. 268.-Jacket.
1886 PAPER AND PAPER-MAKING ,MATERIALS
The top of the tube is closed with a _rubber stopper, carrying another
capillary which has a ground-glass stopcock. The bottom capillary is closed
with a rubber tube and clamp. Each tube has a wedge-shaped cylinder,
li'ig. 269, the bottdm of which is notched.
l'he cylinder, mad~ from 7i-inch steel or
:M:onel rod, is 2.5 cmt long.
The glass jacket, Fig. 268, supports the
viscosity tube in the thermostat. The 0acket
if! so constructed that it supports the tubf on
FIG. 269.-The Wedge. three glass points at the bottom and is a tlid-
iAg fit at the top.
Calibration of Viscosity Tube.-The volume of the tube is obtained as
follows: Fill the tube, containing the wedge, with water, put the rubber stop-
per 12 tightly in place and force the excess water out through the capillary.
Open the stopcock and drain the contents into a burette, noting the time in
seconds for the meniscus to fall from line A to line C. Read the volume of
the water delivered to the burette and take the average of several determina-
tions as the volume of the tube.
The tube constants C and k are determined from the formula.:
I
(1) f=!!_(t'-~),
C\ t
where V = viscosity in centipoises;
d = specific gravity of the solution;
t = time in seconds for meniscus to fall from line A to 0;
C=tube constant;
and k=gravity constant.
The constant C is fi~st determined with standard glycerine solution of
approximately 1.21 sp.gr. (85% glycerine solution; at 20.0° C. (The sp.gr.
of the glycerine solution must be accurately determined with a pycnometer.)
The viscosity of the glycerine solution may be ob'tained from a curve drawn
from Table I. In this case constant k may ,be disregarded, since the time of
efflux is long. The formula then becomes
dt
(2) C=-v,
from which the tube constant is calculated. The gra~ity constant k is deter-
mined by repeating the operation with water at 20.0° C. and substituting for
C in formula (1) the value found above, as "Well as d = 1 and V = 1. The for-
mula then becomes
(3)
Reagent.-Cuprammonium Solution.-Freshly prepare the cuprammonium
solution at least every 2 months as follows: Place clean copper turnings in a
glass 'tube 24 i;ncbes long ann 4 !ffl'Cb~ i1'. Ui'i>llyet'e;,__:-·un,\)\'ir.i.~ '1I~~u\-, ~\'_,'C. '\'T\-.'C
12 The rubber stopper must be replaced periodiclt"liy-by a fresh one, determining the
new volume each time.
PAPER AND PAPER-MAKING MATERIALS 1887
bottom, until it is two-thirds full. Quickly break up the pulp roll with the
aid of a thin glass stirring rod. Add the remainder of the solution at the top,
scrape the rod 9lean and in'ser:t the stopper. Force th\l excess solution through
the capillary and close the stopcock. Wrap the viscdsity tube in black cloth,
and place it on a wheel rotating 3 or 4 r.p.m., so th~t the metal wedge falls
freely through the solution as the tube rotates. The pulp should be well
broken up so that the metal wedge will fall through the full length of the tube,
before the tube is put on the rotating wheel. Allow 15 hours rotation (over
night) for the complete dispersi9n of the pulp. .
Place the tube in the thermostat at 20° C. till equilibrium is establi hed.
i
Remove the rubber tubing from the capillary and pla,ce the tube in po ition
in the jacket, which remains in the thermostat. Remove the rubber stopper,
allowing the pulp solution to flow through the capillary. Note the time in
seconds (t) for the meniscus to pass between the marks A and C, and· calculate
the viscosity of the solution from the formula:
~
v= C (t- ~)t '
in which d, the sp.gr. of the pulp solution, is·(l.96; C and k are constants found.
Simplifying, the equation becomes:
where
Literature Cited
I
1. Clibbens and Geake, "The Measurement of the Fluidity-of Cotton in Cuprammonium
Solution," J. Textile Inst., 19, 77-92 T (1928).. _,. ~
2. Committee Report of Am. Chern. Soc., Ind. Eng. Chern., Anal. Ed. I, 49-51 (1929).
PAPER AND PAPER-MAKING MATERIALS 1889
has a hole cut in the center sufficiently large to permit heating of onlq the
liquid in the flask. .The distillate coming over betiween 124-130° C. i col-
lected and stored in a glass-stoppered brown bottle. .~
Test Specimen.-The specimen shall consist of a~out 0.5 gram of oven-dry
pulp, accurately weighed. I
Procedure.-The methoxyl groups are determined according to the well-
known method of Zeisel. The substance to be examined is heated together
with hydriodic acid, whereby methyl iodide is formed. The CHaI is swept
from the reaction flask by a current of CO 2 into vessels containing an alcoholic
0.1N AgNO a solution, the CHaI being decomposed 'with the formation of AgI.
WAreR OUr,_cr-
I
J
-,
OIL BATH At 1.1D~
# .... _ _ _ •
Weigh approximately 0.5 gram of oven-dry pulp into the digestion flask
(B), add 15 m!. of redistilled HI (sp.gr. 1.70), and heat the reaction flask in a
glycerine bath at_130° C. ,
1'he temperature of the water in the condenser, used in ,this apparatus,
is maintained at 45-55° C. by means of the preheater (X) through which water
is kept flowing constantly by being connected to a water tap.
Absorb the CHaI in 35 m!. of AgNO a solution placed in the absorption
apparatus (E). Pass a current of dry CO 2 gas ,through the apparatus at a
rate of 60 bubbles per minute. Continue hen:'t:mg until the AgNO a solution,
at first milky, has become perfectly clear. This requires 15 minutes to 2 h0urs.
PAPER AND PAPER~MAKING MATERIALS 1893
Wash the contents of the absorption bulb into a beaker, make up to 200 mI.,
add 10 mI. of conc. HNO a and boil the mixture gently for at least 15 minutes.
After cooling this mixture filter on a tared Gooch filter. Wash the precipitate,
dry for 172 hours at 1050 C., cool in a desiccator and weigh as AgI. Calculate
to methoxyl, CHaO.
Weight AgI
Calculation: Per cent methoxyl = 13.2 X W . h I
eIg t samp e
Report.-The data shall be reported as a percentage of the oven-dry weight
of the sample used.
Additional Information
Zeisel; Monatsh, 6, 989 (1885); 7, 406 (1886).
Manning, R. J. and Nierenstein, M., Ber. 46, 3983 (1913).
Fritz Weishut, Monatsh, 33, 1165-12 (1912).
Klemene, A., Monatsh, 34, 901-12 (1913).
Benedict and Bamberger, Monatsh, 15, 509 (1894).'
Br::y, M. W., Paper Trade J. 87, No. 25; 5!H:l8 (1928).
shipment and the chemists shall be careful to observe that no unusual condi-
tions prevail in the selection of bales. j
b. Details: The samples shall lfe taken by boring into each bale to a dep1lh
of 3 inches, with a special tool which cuts a di.sk about 4 inches in diametir.
Care shall be taken to see that the knife or boring tobl blade does not bend
when taking samples. The disks shall be removed an~ ten of them taken as
a sample, selected as follows:
One disk from second sheet from the wrapper.
Two disks 1 inch deep.
Three disks 2 inches deep.
Four disks 3 inches deep.
The holes to be bored shall be so located that in five successive bales they
will represent a portion extending diagonally across the bale .. Each bale shall
be bored but once, the first bale at the corner, the edges of the cut being at a
distance of 1 inch from the edge of the bale. The second cut shall then be
made half-way between the corner and the center of the bale; the third bale
shall be cut at the center; the fourth bale halfway between the cEffiter and the
corner; and the fifth bale in the opposite corner in a position corresponding
to the first. In case binding wires or straps interfere with the exact location
of cuts, borings shall be made as near the .prescribed location as possible.
(2) Roll Pulp: /
a. Number of Rolls to be Sampled: At least 10% of the lot, but not less
than 30 rolls. Where possible to determine, the same proportion of outside,
rolls to ce~ter rolls should be maintained as coming from the machine.
b. Details: The edge of the first hole shall be 2 inches from one end; holes
in succeeding rolls shall be spaced successively one-fifth of the distance acrosi3
the roll as in bale sampling, traveling toward the opposite end of the roll.
Selection of disks shall be in accordance with bale boring as given above.
I
ASH IN PULPY
Apparatus.-A crucible, such as platinum, alundum,.or ,porcelain, which
has a tightly fitting lid is required. It should have a capacity of30 ml. or more.
An analytical' balance sensitive to 0.1 mg. '
Any suitable desiccator, such as the customary analytical desiccator which
employs sulfuric acid or calcium chloride as the drying agent, may be used.
An electric muffle furnace is recommended for igniting the pulp sample.
A Meker or other gas burner may be employed for this purpose.
A suitable drying oven with temperature control' between 100 0 C. and 105 0 C.
should also be available.
Test Specimen.-Approximately 5 grams of pulp shall be, used for each
determination. J
uniformly maintained at 97° to 100° C. When a new bath is used, the tem-
perature of the material during the treatment· shouid be rechecked. The
type of bath recommended is one which is covered and has holes in the t~
large enough so that beakers may be set down into the. bath until.they are su -
ported by the flared rim. The top of the beaker shoJld be nearly level w th
the cover of the bath. The sides of the beaker are ~ntirely surrounded by
boiling water or steam when this type of bath is used. The height of the water
in the bath should be maintained at a level above that of the liquid in the
beakers.
Tall form 200-ml. Pyrex beakers should be used. Fritted glass. crucibles
(porosity 3) are recommended for filtering the treated pulp, although~lundum
crucibles of a similar porosity are also suitable. -
Caustic Soda Solution.-The 1 % NaOH solution (0.25 normal) is made by
dissolving 10 grams of solid NaOH in water and diluting to 1 liter. Determine
the NaOH content by titrating with standard acid, first using phenolphthalein
as an indicator, then continuing the titration using methyl orange indicator.
The difference between the volume used for the phenolphthalein end-point
and that used for the methyl orange end-point represents that consumed by
half the Na 2 CO a present. By subtracting ,this difference from the volume
used for the phenolphthalein end-point, the volume required to neutralize
the NaOH is obtained, and from this volume the NaOH concentration is
calculated. This shall then be adjusted to between 0.9 ana 1.1%. '
Test Sample.-A sample which is representative of the pulp being tested
shall be air-dried and placed in an air-tight container. The moisture content
shall be determined by the standard procedure. If the pulp is in the form of
machine-dried or pressed sheets which do not absorb liquid readily" it should
first be disintegrated to slush form in water, then filtered into pads which 'are
dried without any pressing. In this way the pulp is obtained in a form which"
readily disintegrates in the alkali solution.
Procedure.-Accurately weigh on an analytical ballwce two samples of the
air-dried pulp each equal to 2 grams of oven-dry pulP' plus or minus 0.1 gram.
To each sample in a tall-form 200-ml. beaker, add 100 ml. of 1 % NaOH solution
measured with a pipette. Stir the pulp until .it is completely disintegrated,
cover the beakers with watch glasses and place:in the, water bath, which.must
be boiling steadily. Leave the beakers in the bath for e~actly 1 hour. Stir
the contents three times, 'at 10, 15, and 25 minutes after heing~p.laced in the
bath. At the end of the 1-hour period, filter the contents of each beaker by
suction on a tared crucible. First wash the pulp briefly with hot water, then
with 50 ml. of 10% acetic acid, and then wash thoroughly with hot water.
Dry the crucible and contents at 105° C., cool in a desiccator, and weigh in a
stoppered weighing bottle. Calculate the loss in( weight of the oven-dried
sample.
Report.-The results shall be reported as percentage of the oven-dry
weight soluble in the NaOH solution. This value may be corrected if desired
for hot-water solubility by subtracting from it the percentage of the oven-dry
material soluble in hot water as determined by the.standard T APPI method
(T 207 m). In this case the result shall be reported as " corrected."
PAPER AND PAPER-MAKING MATERIALS 1899
PARAFFIN IN PAPER 21
Apparatus.-An extraction apparatus with a reflux 'condenser is necessary
for this test. The balance used for weighing shall be sensitive to 0.1 mg.
Test Specimen.-The test specimen shall consist of not less than 5 grams
of air-dry paper obtained by cutting small strips from different portions of the.
test sample in such a way as to be representative of it. The strips shall be cut
into small pieces.about 6 mm. (0.25 inches) square. It is recommended that
the test specimen be ground or shredded. .
Procedure.-Completely extract the paraffin from the paper with chemi-
cally pure carbon tetrachloride, in the extraction apparatus. Evaporate the
extract to dryness, add 25 ml. of 0.5 N alcoholic potash, and again evaporate
to dryness. Take up the dry residue with ether and water, and transfer to a
separatory funnel. The volume of ether in the funnel should be 25 ml. and
that of the water, 150 ml. The water should contain a small amount of sodium
chloride to prevent emulsification. Shake the contents of the flask. thoroughly,
allow separation to take place, and draw off the water into a second separatory
funnel. Repeat the treatment, one or more times as may be found necessary,
with a fresh 25-ml. portion of ether. Combine the ether extracts, and wash
with fresh 100-ml. portions of water until the separated liquids are perfectly
clear. An addition of a strong solution of sodium chloride to the ether extract
may be necessary to thoroughly clarify it. Transfer the ether extract to a
weighed evaporating dish, evaporate to dryness, dry for exactly one hour at
• 100° C. and weigh the dried residue to an accuracy of 0.1 mg. Not less than
two determinations shall be made and the average of the results shall be com-
puted. The percentage results of duplicate determinations of paraffin shall
agree within 0.2.
Report.-The paraffin content shall be expressed as a percenta~e of the
air-dry paper to the nearest 0.1.
some of the mixtures of enzymes used commercially for desizing cotton and
degumming silk have been found to be more rapid in action, less expensive,
and more stable. The preparation Degoma D. L. made by Rohm & Haas
Co., 222 West Washington Square, Philadelphia, Pa., is particularly suitable.
Test Specimen.-The specimen for test shall consist of not less than 25
square inches.
Procedure.-Obtain the air-dry weight of the specimen. Place it in a
suitable container in such a way that the enzyme solution will have free acctss
to the entire surface of the coating. It may be laid in a fiat-bottomed tray
or rolled in cylindrical form and placed in a burette. Pour over the specimen
sufficient enzyme solution to cover it completely. Allow to stand for at least /
1 hour at 50° C., then place the paper on a pane of glass and brush off the
coating with a camel's hair brush, taking care not to dislodge the paper fibers.
Additional heating, and in some cases additional enzyme solution, may be found
necessary for very resistant coatings.
Mter the coating is entirely removed, stand the glass pane at a slight angle
and wash the paper on both sides by means of a wash bottle, holding the paper
on the glass by one corner. Dry the specimen and obtain the air-dry weight.
The difference between this weight and the original weight of the specimen is
the amount of coating material. Not less than two determinations shall be
made and the average of the results shall be reported.
Report.:_The amount of mineral coating shall be reported (1) as a per-
centage of the decoated air-dry paper and (2) as pounds per 500 sheets, 25X40
inches in size. The weight of the decoated air-dry paper, on this same weight
basis, shall also be included in the. report.
Precision.-Owing to the variable nature of coating materials, the precision
is variable. With most ordinary types of coatings, the percentage of coating
material found should be correct to within 0.5 to 2.
Literature Cited
1. Sutermeister and Porter, Tech. 4ssoc. Papers, XIII, No.1; 205 (1930).
RESIN IN PAPER 24
Apparatus.-A suitable extraction apparatus, such as the Soxhlet or the
Underwriters', is required for this determination. In addition, if glue or other
nitrogenous sizing agents are present, two 300-ml. separatory funnels are
necessary.
Reagent.-Acidulated Alcohol.-To 190 ml. of 95% alcohol add 1 ml. of
glacial acetic acid and 19 ml. of water, and mix.
Test Specimen.-The test specimen shall consist of not less than 5 grams
of paper obtained by cutting strips approximat!!ly Yz inch wide froni "the sample
in such a way as to be representative of it. It is recommended that the speci-
mens be shredded or ground.
Procedure.-Obtain the air-dry weight of the specimen. If the paper is
not shredded or ground, fold th~ strips into numerous crosswise folds and place
24 T 408 m-3B. Tentative Standard-Feb. 27,1917. Official Standard-July, 1926.
Corrected-Sept. 15, 1936.
1904 PAPER AND PAPER-MAKING MATERIALS
, '
them in the siphon cup of the extractor. Extract with acidulated alcohol,
siphoning at least 12 times or as !llany more as may' be necessary until the
solvent siphons over colorless.
If nitrogenous sizing agents are present, they must be separated from the
resin as follows, this not being necessary in absence ofi such materials. Wash
the alcoholic extract of resin, which may contain foreign material, into a beaker
and evaporate to a few m!. on a steam bath. Cool, take up in about 25 m!.
of\ther, transfer to a 300-m!. separatory funnel containing about 150 m!. of
water to which has been added a small quantity. of sodium chloride to prevent
emulsification, shake thoroughly and allow to separate. Draw off the water
into a second separatory funnel and repeat the treatment with a fresh 25-ml.
portion of ether. Combine the ether extracts which contain the resin and any
other ether-soluble material, and wash twice, or until the ether layer is per-
fectly clear and the line between the ether and the water is sharp and distinct,
with 100-ml. portions of water to remove salts and foreign matter. Should
any glue, which may be extracted from the paper, interfere by emulsifying with
the ether, it may be readily removed by adding a strong solution of sodium
chloride to the combined ether extracts. Sha,ke thoroughly and draw off the
aqueous solution, repeating if necessary before washing with water.
Transfer the alcoholic or the ether extract to a weighed evaporating dish,
evaporate, to dryness, dry the residue exactly 1 hour at 100° C., and weigh to
an accuracy of 0.1 mg. Not less than t~o determinations shall be made, and
the average of the results computed. The percentage results of duplicate
determinations of resin should agree within 0.2.
Report.-The resin content shall be expressed as a percentage of the air-
dry paper, to the nearest 0.1 per cent.
Additional Information
"Quantitative Determination of Rosin in Paper."-C. F. Sammet, Ind. & Eng. Chern.,
5,732 (1913); Paper 13, No.1, 17 (1913).
NOTE.-8ee also T 416 m, Rosin in Paper (Qualitative).,
MOISTURE IN PAVER 25
Apparatus.-.The special apparatus required for this determination is an
air-tight container in which the specimen is dried and weighed. For the
minimum size specimen designated, a weighing bottle approximately 65 mm.
(2.56 inches) in height and 45 mm. (1.77 inches) in diameter is'suitable. For
larger specimens, proportionately larger containers should be use_d.
The oven used to dry the paper shall be equipped witb. means for ensuring
adequate temperature control and air circulation. and preferably equipped with
means of drying the air entering the, oven.
Test Specimen.-The specimen for test shall consist of not less than 2
grams of paper obtained by cutting small strips from different portions of the
test sample in such a way as to be representative of it.
NOTE.-When the paper under test is not, in 'moisture equilibrium with the sur-
rounding atmosphere, take care to minimize the time of exposure of the test sample to
the a.tJ?osphere as much as possible, gain or loss of mois~uUl being very"rapid under this
condItion. . _
25 T 412 m. Official Standard-Dec., 1926. Corrected-May 15, 1935.
PAPER AND PAPER-MAKING MATERIALS 1905
ASH IN PAPER 26
Apparatus.-A crucible, such as platinum, alundum or porcelain, which
will not change in weight under the ignition conditions used and with a tightly
fitting lid; a balance sensitive to 1 mg.; and a desiccator, are necessary for this
determination. An electric muffle furnace is recommended for burning the
paper.
Test Specimen.-The test specimen shall consist of not less than 1 gram
of air-dry paper obtained by cutting small strips from different portions of the
sample in such a way as to be representative of it.
Procedure.-Weigh the specimen of paper accurately in the previously
weighed crucible and ignite completely. To avoid loss of small particles of
the specimen care must be taken to heat it slowly and to protect the contents
of the crucible at all times from strong drafts. When the paper is completely
burned, as indicated by absence of black particles, remove the crucible to the
desiccator, cover and allow to remain until its temperature has reached equi-
librium with that of the surrounding atmosphere. Then weigh the crucible
and contents and repeat the ignition and weighing until the weight is constant.
All weighings shall be made to an accuracy of 1 mg.
Precision.-The percentage results of duplicate determinations shall agree
within 0.2.
Report.-The amount of ash shall be reported as a percentage of the air-
dry paper to the nearest 0.1.
Additional Information
See T 211 m Ash in Pulp.
NoTE.-As these factors vary with different kinds and grades of material, they should
be determined whenever the nitrogenous material is" available, and whenever possible
the N in the paper before addition of the nitrogenous material should be determined and
subtracted from the total N found.
STARCH IN PAPER 28
QUALITATIVE
QUANTITATIVE
Additional Information
This method is based on that of V. Voorhees and O. Kamm. 29 The original method
was revised in, respect to the Fehling solution (suggested by Miss H. U. Kiely) and the
Benedict solution (suggested by G. P. Genberg), to avoid.use of indicators and to obtain
a sharper end-point.
29 Paper, 24; 1091 (Aug. 27, 1919).
PAPER AND PAPER~MAKING MATERIALS 1909
standard textbooks dealing with mineral analysis. When the ashing of the
paper causes chemical changes in the filler, the actual amount of filler originally
present must be calculated. The more common examples of such changes are:
loss of combined water in clay and crown filler; lossl of CO 2 in carbonates;
reduction of sulfates to sulfides. I
I
Additional Information
Griffin, R. C., "Technical Methods of Analysis." McGraw-Hill, New York.
the first instance. Make a blank titration on 250 ml. of the water heated for
1 hour in the same bath and with the same glassware used for the extractions.
Each test result used in calculating the acidity shall be the average of not less
than two determinations. The results of duplicate determinations of total
acidity shall agree within 0.01.
Report.-Total acidity shall be expressed as a percentage of the moisture-
free paper in terms of sulfuric anhydride, 80a, to the nearest 0.01. It is calcu-
lated as follows:
Let T=ml. NaOH required to neutralize the 3 extracts;
t=3 times the m!. NaOH required to neutralize the blank;
N = normality of the NaOH solution; and
W = weight of test specimen less moisture.
(T-t) XNXO.04X100 - t SO
Then, W - per cen a.
Additional Information
This method is that described by S. Kohler and G. Hall in The Paper Industry, 7,
No.7; (Oct. 1925), with some modifications developed at the U. S. Bureau of Standards.
Studies of tbis method and other acidity methods are reported by Wehmhofl' and by
Wehmer in Technical Association Papers (TAPPI), May, 1930 and May, 1931.
The Dr. Koerner type of grinder is described in U. S. Bureau of Standards Research
Paper No. 295.
solution will then contain 17.5% of NaOH by weight. lhis should be checked
by titration and should be adjustedif necessary within{the limits 17.4-17.6.
Acetic Acid.-Approximately a 20% solution in water.
Test Specimen.-The specimen for test shall be cut. from the test sample in
such a way as to be thoroughly representative of it, ahd shall be completely
disintegrated in the grinder.
Procedure.-Allow the specimen to come to moisture equilibrium with thE
atmosphere of the balance. Weigh (to the nearest 1 mg.) approximately
5 grams of the disintegrated paper. ,Weigh at the same time samples for
moisture and ash determinations, and for determinations of such other non-
cellulose components as may be found necessary for correction of the alpha
cellulose figures (total resins, starch, glue, etc.). These determinations (lh!iJI
be made by the T APPI standard methods. When mineral filler is present, the
weight of it shall be calculated from the ash content. The weight of mineral
coating shall be determined likewise, or according to the standard TAPPI
method.
Transfer the specimens for alpha cellulose. determination to a 400-ml.
beaker, adjust the beaker and contents in a water bath at 20° C., add 50 ml.
of the NaOH solution accurately measured with a pipette, and note the time
immediately after this is added. Macerate the mixture thoroughly with a
flattened glass rod, and add a second 50-ml. portion of the N aOH solution,
this addition being made after about 5 minutes have elapsed from the time
of the first addition. Macerate the mixtU're again and let it stand with occa-
sional stirring for a total time of 30 minutes from the time of first addition of
NaOH. At the end of this period, dilute the mixture with 250 ml. of distilled
water having a temperature of 20° C. and filter it immediately with gentle
suction, using a 7.5-cm. Buchner funnel with a nainsook cloth filter cut to fit
the funnel, the cloth having been previously washed to remove stiffening
material, and dried against glass. Dry the cloth to constant weight at 105° C.
and weigh it in a weighing bottle before using it. Wash as uniformly as
possible with 125 ml. of water at exactly 20° C. and then return the filtrate to
the funnel and allow it to pass through the pad. Then wash the pad with
750 m!. of water having a temperature of exactly 20° C. Occasional packing
of the pad with the flattened glass rod is necessary during the washing as the
pad has a tendency to shrink and permit the wash water to flow around its
edge and not through it. Follow the water with 100 m!. of 20% acetic acid
at about 20° C. and let the pad stand in this condition for'5'miu!ltes, without
application of suction. Finally apply suction and wash the pad with 1 liter
of water at about 95° C.
Transfer the pad and cloth to a Petri dish or watch glass, dry at 70° C.
for 3 hours, transfer to a weighing bottle, dry to constant weight at 105° C.
and weigh in the closed weighing bottle. Determip.ations shall be 'made of
ash and other non-cellulosic components of-the residue as may be fO'und neces-
sary, so that the weight of the residue may be corrected for these. components.
The weight (,)f the residue, after deduction of non-cellulosic constituents,
divided by the weight of the test specimen, likewise corrected, and the result
multiplied by 100, gives the percentage of alpha celh,Ilose in the total cellulosic
content. The weight of the test specimen shall alwtrYi:l be corrected for mois-
ture and both the test specimen and the residue shall always be corrected
PAPER AND PAPER-MAKING MATERIALS 1913
for ash. Both the test specimen and the residue shall be corrected for other
non-cellulosic components whenever they are present in significant amounts.
Not less than two determinations shall be made and the average of the results
shall be reported. Duplicate determinations shall agree within 0.3.
Report.-The amount of alpha cellulose shall be reported as percentage
of the total cellulosic content to the first decimal place.
Additional Information
Cross and Bevan 33 mention the use of NaOH of mercerizing strength in isolating
"normal" cellulose. The method described here is based on that of Jentgen 3' and
modification of his method suggested by -Parsons 35 and Ross 36 and by Burton and'
Rasch. 37
ground paper. Weigh at the same time, samples for moisture and ash deter~
minations, and for determinations of such other nQn~cellulose components
(total resins, starch, glue, etc.) as may be found necessiuy for correction of the
copper number. These determinations shall be made by the TAP PI standard
methods. When mineral filler is present, the 'Weight,' ~f it shall b,e calculated
from the ash content. The weight of mineral coatiJ?g shall' be determined
likewise, or according to the Tappi standard"method. :
Immediately before use add 5.0 m!. of solution A to 95 m!. of B. Bring
the mixture to a boil in 2 minutes, and pour it over 1.5 grams of the 'finely
ground sample in a 125~ml. Erlenmeyer flask. Stir well with a glass rod in
order to distribute the fibers and to remove air bubbles. Fit the flask with
'a loosely fitting glass bulb or stopper and submerge completely in a steam Hath ...
at atmospheric pressure. Occasionally fibers tend to flo;:tt to the surface, 'there~
fore the flask should be shaken from time to time to redistribute them. Re-
move the flask from the steam bath at the end of 3 hours. Filter on an ashless
filter paper in a 7.5-cm. Buchner funnel, using suction. Wash by flooding with
100 m!. of 5% Na 2 C0 3 solution at about 20° C. and then by flooding with
250 m!. or hot water (about 95° C.). Transfer the fibers and filter paper to a
small beaker, add 25 m!. of the molybdate solution and macerate well with a
flattened glass rod. Transfer to a Buchner funnel again and wash thoroughly
with cold water until the blue molybdenum color is removed from the fibers.
Dilute the filtrate with water to approximately 700 m!. .and titrate it with
0.05 N KMn04 to a faint pink. /'
The copper number is defined as the number of grams of metallic copper
in the cuprous oxide reduced from the cupric hydroxide by 100 grams of the
paper fibers. This is calculated as follows: .
6.357Xm!. KMnO,XN
C opper N urn b er= W '
where N is the normality of the KMnO, and W is thlJ weight in grams of the
test specimen after deduction of the weight of the :non-cellulosic materials.
Correction of the weight of the test specimen shall alw'ays be made for moisture
and ash. Correction for other non-cellulosic components shall be made when~
ever they are present in significant amounts. N~t less than two determina-
tions shall be made and the average of the results shall be reported. Duplicate
determinations shall agree within 0.1. "- . .
Report.-The copper humber shi'tll be reported on the basis_of the total
cellulosic content to two decimal places.
Additional Informatio_n
This method is essentially the Braidy 39 modification of the original Schwalbe 40
method, with modifications for its adaption to paper proposed by Scribner ,and Brode,41
and by Burton and Rasch. 42 ,
39 A thorough investigation of the Braidy method was made by Clibbens and Geake,
(J. Text. Inst~, 15, T 31, 1924).
40 Die Chemie der Cellulose, p. 625; 1912. .
41 Technologic Paper No. 354, National Bureau of Sta)l9.,!trds. I
42 Bureau of Standards Jour. of Research, R. P . .,.295, April, 1934. Includes an
illustrated description of the "Dr. Koerner" type of grinder~ .
·PAPER AND PAPER-MAKING MATERIALS 1915
Wires to Potentiometer
A
Kel (sol n )------if--~
draw the solution through the fiber mat with strong ~uction. Disconnect the
suction just before the filtration is completed. As a few fibers may go through
the funnel with the first portion of solution, return the filtrate to the flask
and again pass it through the mat. Pour the filtrate into a clean flask, close
it tightly with a clean rubber stopper, and set aside to Icool to room tempera-
ture. The extract must be protected from the air as much as possible at all
times to avoid CO 2 contamination. I ·
H 0.4526-E .
p = 0.0591 for the qumhydrone electrode.
E-0.2358 . I
pH = 0.0591 for the hydroge~ ~eitrode.
PAPER AND PAPER-MAKING MATERIALS 1919
The voltage obtained with the antimony or glass electrode, depends upon
the individual electrodes used. The pH can be calculated from the voltage
by calibrating the electrodes against buffers of known pH. In some cases
these calibrations are furnished by the manufacturers of the electrodes. The
above equations are for a temperature of 25° C. but they may be used for
measurements made at ordinary room temperature within the degree of ac-
curacy stated below, by changing the denominator plus or minus 1/298 of its
value, respectively, for each ° C. above or below 25. These equations are
valid only when the saturated calomel electrode is used as the reference elec-
trode. ,
The average of the results of at least two determinations shall be reported.'
The results should agree within 0.1 pH.
Report.-Results shall be reported as the pH of the air-dry paper. They
shall be rounded off for the colorimetric method to the nearest 0.1, and for
the electrometric method to the nearest 0.05. The report shall state which of
these methods was used.
Literature Cited
For an illustrated description of a grinder suitable for disintegrating the paper see:
1. Burton and Rasch, U. S. Bur. of Standards Jour. of Research, 6, 605 (Research Paper
295) (1931).
For detailed description of suitable electrometric methods see:
2. Clark, W; M.: "The Determination of Hydrogen Ions,'~ Williams and Wilkens Co.,
Baltimore (1928).
3. Burton, Matheson and Acree, U. S. Bur. of Standards Jour. of Research, 12, 67
(Research Paper 436) (1934). .
4. Leeds and Northrup Company, Philadelphia, Pa.
5. American Instrument Company, Washington, D. C.
For a description of a suitable colorimetric method see:
6. Fawcett and Acree, Jour. of Bacteriology, 17, 171 (1926).
7. Fawcett and Acree, Ind. Eng. Chern., Anal. Edition, 2, 78 (1930).
8. "Modern pH and Chj.orine Control," Bulletin of W. A. Taylor and Company,
Balti.more, Md.
A study of certain of the details of this method is reported in the following articles:
9. Wehmhoff, B. L.: "The Determination of thf pH Values and Total Acidity of
Paper," Tech. Assoc. Papers, 13, p. 231, May, 1930; 14, p. 387, May, 1931.
ARSENIC IN PAPER 45
Apparatus.-The modified Gutzeit apparatus for determination of arsenic
is shown in Fig~ 273.
Two sizes of the apparatus are desirable; No.1 for use when the paper
contains less than 0.01 mg. of arsenic trioxide (As 20 a) and No.2 for use where
the paper contains more than that amount. The dimensions of apparatus
No.1 are listed to the left of the sketch; those of apparatus No.2 are listed
to the right.
Reagents.-Concentrated Nitric ACid, arsenic-free,. having a sp.gr. of at
least 1.4. .
Concentrated Sulfuric Acid, arsenic-free, having a sp.gr. of at least 1.8.
Zinc, -arsenic-free, in the form of sticks (Bureau of Standards Standard
45 T 436 m-36. Tentative Standard-May 15, 1935. Official Standard-8ept. 15,
1936.
1920 PAPER AND PAPER-MAKING M'ATERIALS
Sample No. 43D is recommended). Break the sticks into pieces about 1 cm.
long, rinse with water and treat with HCI until the sJrface of the Zn is clean.
Then wash with water, dry and store in a clean, wide-mouthed, glass-stoppered
bottle. \ .
Mixed Acid.-Dilute 1 volume of As-free H 2S0 4 with 4 volumes of water,
and to this add 10 grams of N aCl for each 100 ml. of ~olution.
Ferric Ammonium Alum Solution.- Dissolve 84 grams of the salt, Fe2
(NH 4)2(S04)4·24H 20, in water, add 10 ml. of the mixed acid and dilute to
1 liter.
Lead Acetate Solution (for wetting glass wool).-Add to a 5% solutiQn'of
lead acetate, Pb(C 2 H a0 2)2·3H 2 0, sufficient. acetic acid to clarify the solutio11l'
Dimensions of Dimensions of
Apparatus No.1 Apparatus No.2
~rB 7.5
9
10 em. long x 4 m. em. long x mm. bo,e.
bore. Constricted 6 em, Constricted 12, em {rom upper end.
from upper end
paper.
15.15A . f . i d f
(b) -s-- = grams I
0 arseruc per square yar 0 paper.
Additional Information
"Standard Methods of Chemical Analysis," 5th Ed., 'vol. I, pp. 101-108.
Griffin, R. C., "Technical Methods of Analysis," 2nd Ed., pp. 52 ahd·56. McGraw-Hill,
New York, N. Y. (1927). -.._,-
Hillebrand & Lundell, "Applied Inorganic Analysis," pp. 218-220. }Viley & Sons,
New York, N. Y. (1929).
Official Methods of The Association of Official Agricultural Chemists, 3d Ed., p. 306.
Williams & Wilkins, Baltimore, Md. (1930). '
I
ZINC PIGMENTS IN PAPER47
The zinc pigments used as paper fillers comprise three main classes: (1) zinc
sulfide alone, (2) high-strength zinc pigment, containing about 50% ZnS and
50% BaS04, and (3) low-strength zinc pigment, containing l~ss than 30% ZnS
_!---
47 T 438 m-36. Tentative Standard-Aug. 1, 1936, - .
PAPER AND PAPER-MAKING MATERIALS 1923
'"
PAPER AND PAPER-MAKING MATERIALS 1925
(1) ZnS-1.l9
(2) "High-Strength" zinc pigment-l.09
(3) II Low-Strength" zinc pigment -1.06
References
Kolthoff and Pearson, flTitrntion of ,Zinc with Potassium Ferrocyanide," Ind. Eng.
Chern., Anal. Ed. 4, 147 (1932).
Steele, F. A., "The Determination of Zinc Sulfide Pigments in Paper/' Paper Trade .J.,
97, No.9, 27 (1933): .
I---DI,;."";·,,, Bur.tt.
6.5 tm,(21 in.Jlang
FIG. 274.
I
r
Apparatus.-The only special apparatus required is a Jones reductor, shown
in Fig. 274. This consists of a glass tube about 2 inches in diameter and 21
I
u T 439 m-36. Tentative Standard-Aug. 1, 1936.
1926 PAPER AND PAPER-MAKING M!A.TERIALS I'
inches long with a glass stopcock at the bottom. (An ungraduated dispe~sing
burette is satisfactory.) It is connected to a liter flas~ as shown in the figure.
A means is also provided for introducing CO 2 into th~ flask through 1J, bubble
trap. The tube is, filled for about one-third of its length with 1000 grams of
a mixture of equal amounts of lO-mesh and 20-mesh Igramilated zinc resting
on a pad of glass wool. The upper portion of the tub~ contains 500 grams of
stick Zn in pieces about 2 inches long. Before use the Zn is amalgamated
as follows: Place, 250 grams of the Zn in a heavy glass flask, add a solution of
11 grams of mercuric chloride in 100 ml. of concentrated HCI, swirl the mixture
for a few minutes, pour off the liquid, and wash the Zn free from acia ..with
water.
NOTE.-When not in use the Jones reductor should be kept filled with water a~
covered to exclude dust and other extraneous matter.
Reagents.-O.l N Potassium Permanganate Solution, standardized against
U. S. Bureau of Standards sodium oxalate.
Standard Ferric Alum Solution, of which 1 ml. is equivalent to about 0.005
gram of Ti0 2 ; standardized by reduction and titration with 0.1 N KMn04.
Dissolve 30 grams of Fe2(NH4h(S04) 4' 24H 20 in 300 ml. of water acidified
with 10 ml. of H 2S0 4 and add KMn04 solution drop by drop as long as the
pink color disappears. Finally dilute the solution to 1 liter. To standardize,
dilute 50 ml. of the solution, measured from a burette, to· about 100 ml. with
dilute H 2S0 4 (1 : 20), reduce in the Jones reduct or and titrltte against the
0.1 N KMn04.
Thiocyanate Indicator.-A saturated solution of NH 4SCN.
Test Specimen.-Sufficient paper shall be ashed according to TAPPI
standard T 413 m to yield the required amount of ash. The oven-dry weight
of the paper shall be determined according to T APPI standard T 412 m.
Procedure.-Qualitative.-A qualitative analysis of the ash should be made,
unless the nature of the loading material is known, sinc,e it will greatly facilitate
the quantitative analysis. The procedures for: quaJitative analysis of fillers
and ash for the commoner loading materials; such as clay (hydrated aluminum
silicate), talc (hydrated magnesium silicate), barium sulfate, barium carbonate,
calcium sulfate, and calcium carbonate, are well known and are included in
the TAPPI Standards. To test for Ti place about 0.5 gram of the ash in a
250-mI. beaker, add 20 ml. of concentrated H 2S0 4 and 10·grams of (NH 4)2S04
and boil for at least 5 minutes. An insoluble residue indicates-Si0 2 or siliceous
matter. Cool the solution, dilute to 100 ml. with water, heat_to boiling, let
settle ,and filter through double Whatman No. 42 paper. Test the filtrate
with H 20 2 • A clear yellow or orange co;lor indicates the presence of Ti.
NOTE.-For a more detailed qualitative procedure, se,e Am. Chern. Soc, Monograph
No. 33 "Titanium," by W. M. Thornton. '
Quantitative.-The determination of Ti involves its reduction and titration
with a ferric salt. The details of the procedure depend upon the other mate-
rials present. The general procedure for the determinatio.n of Ti0 2 is given
in Scheme 1, which should be used when none of th~:Other ash constituents are
to be determined. Where BaS04 also is to be
deterniined, use Scheme 3.
PAPER AND PAPER-MAKING MATERIALS 1927
The procedure for a complete analysis of the ash in the absence of BaS04 is
given in Scheme 2.
Scheme 1 : Determination of Ti0 2 Only.-Weigh out accurately into a 250-ml.
beaker a sample of ash containing not more than 0.25 gram of Ti0 2• (Usually
0.5 gram is suitable.) Add 20--30 ml. or concentrated H 2S0 4 and 10 grams of
ammonium or sodium sulfate. Cover the beaker with a watch glass and care-
fully boil the contents over a flame for at least 15 minutes. Cool and dilute
with water to 125-150 ml. A white precipitate may form, indicating BaS04.
Boil the solution gently and if it contains a precipitate, filter and wash the
precipitate with hot dilute H 2S0 4 (1 : 20) and finally with hot water until
free from Ti, indicated by testing with 3'% H 20 2 solution, which gives a yellow
color with the washings if Ti is present.
N oTE.-Instead of dissolving the ash with concentrated H 2S0 4, it may be fused with
Na 2CO. in a covered Pt crucible over a Meker burner, using 5-10 grams of dry Na 2CO.
for 0.5 gram of ash. Digest the melt with 20--25 ml. of concentrated H 2S0 4in about 1~5
m). of water; boil; and filter to remove any BaS04 and other insoluble matter. Wash
the precipitate free of Ti with hot dilute H 2S0 4 (1 : 20) followed by hot water.
The filtrate and washings containing the Ti salts should have a volume of
about 250 ml. Reduce and titrate as follows: Drain the Jones reductor to the
top of the fine-mesh Zn and add 200 m!. of boiling dilute H 2 S0 4 (1 : 20).
,Slowly drain the reductor and then wash at least four times with boiling water.
Drain the reductor to the top of the fine-mesh Zn, wash the hot Ti solution
into it and allow it to remain for 15--:20 minutes, keeping the reductor covered
with a watch glass. During the whole period of reduction pass a slow stream
of CO 2 through the receiving flask. Slowly drain the reductor into the re-
ceiving flask while CO 2 is continuously passed through it. When the solution
has reached the level of the fine-mesh Zn, add more of the boiling dilute H 2S0 4
and bring the level of the solution to the top of the stick Zn. Drain the re-
ductor to the fine-mesh Zn, then wash 4 times with boiling water in the same
manner as above. Remove the flask, add 7 ml. of the NH 4SCN indicator and
titrate the solution rapidly with the standard ferric alum solution. Calculate
the titration to percentage of Ti0 2 •
NOTES. I.-In titrating the Ti solution, the greatest accuracy is obtained by adding
nearly all of the ferric alum before agitating; then agitate by shaking with a gentle
rotary motion which is continued until the final end-point, a light straw color, is reached.
2. Before using the reductor run a blank, using the same concentration of H 2S0 4 in
water and the identical procedure as used for the Ti determination, the reduced solution
being titrated. This blank should not require more than 0.1 ml. of the standard ferric
alum solution. If more is needed, the Zn should be reamalgamated or another lot of
ferric alum used.
Scheme 2: Analysis of Ash Containing Ti0 2 in Absence of BaS04. (Ti0 2
in mixture with clay, talc, chalk or calcium sUlfate_.)
(a) Silica.-Fuse 0.5 gram of ash with 5-10 grams of Na 2 C0 3 in a covered
Pt crucible over a Meker burner. Treat the melt with about 125 ml. of dilute
HCI (1 : 4), filter, and wash with hot water. Ignite the residue in a tared Pt
crucible at the highest heat of a Meker burner, cool in a desiccator and weigh.
Treat the ignited and weighed residue with about 5-10 ml. of HF and 3-4
drops of concentrated H 2S0 4, evaporate to dryness on a hot plate and finally
ignite to constant weight. If the amount of Si0 2 is considerable, add more
1928 PAPER AND PAPER-MAKING MATERIALS
HF and H 2S0 4 , and repeat the evaporation and ignition. The loss in weight
represents Si0 2 • , .
Evaporate the filtrate from the melt to drynes~ and bake on. the hot nlate
for about 72 hour to insure complete dehydration lof the SiG 2 • Take up the
residue with warm dilute HCl (1 : 4); filter, and 'Yash with water. Dry the
residue, ignite and weigh in a tared Pt crucible. Treat with 5-10 ml. 'of HF
and 3-4 drops of concentrated H 2S0 4 , evaporate to dryness on a hot plate and
finally ignite to constant weight. The loss in weight represents soluble Si0 2•
Add this "result to the previous to give the total Si0 2 •
If residues are present after the Si0 2 has been removed, fuse widl KHS0 4,
dissolve the melt in dilute H 2S0 4 (1 : 4) and add to the filtrate from the Si0 2
determinations.
(b) Titanium Dioxide and Alumina.-Add a slight excess of NH 40H to the
combined filtrates and washings from the Si0 2 determination, boil the solution,
filter, and wash the precipitate with 5% NH 40R. Dissolve the precipitate
in the original beaker with dilute HCl (1 : 1) and reprecipitate with NH 4 0H.
Filter through the original filter, and wash, combining the washings with the
first filtrate and washings. Dry the residue, ignite to constant weight over
a Meker burner and weigh as A1 2 0 a+Ti0 2 • Subtracting the. Ti0 2, as deter-
mined in Scheme 1, gives the A1 20 a. (4.ny Fe present as an impurity will be
weighed as Fe 20a and included with the A1 20 a.)
/
72 hour. Filter and wash the precipitate thoroughly with hot water, ignite
to constant weight oyer a 'blast lamp and weigh as CaO. Calculate to Ca80 4 •
deduct the weight from that of the mixed sulfates and calculate the remaining
Mg80 4 to MgO.
Calculations:
CaO X2.428 = CaS04.
MgS0 4 XO.3349 = MgO.
Sulfur Trioxide.-Pipette 50 or 100 ml. of the original sample into a 300-ml.
Erl,enmeyer flask and completely displace the air above the liquid by a current
of C02, which has first passed through a wash-bottle containing an approxi-
mately 1 % solution of KMn04 to remove any oxidizable sulfur-compoullds.
Then, with the current of CO 2 still passing into the flask, add 10 m!. of con-
centrated HCI and evaporate the solution nearly to dryness to remove S02.
Transfer the solution to a beaker, washing out the flask and adding the wash-
ings to the beaker until the total volume is 100-150 m!. Acidify this solution,
with 2 ml.. of concentrated HCl, heat to boiling, and add 10 m!. of 10% BaCh
solution drop by drop while boiling. Let the precipitate settle for several
hours, preferably overnight, filter, wash thoroughly with hot water, ignite,
cool and weigh as Ba80 4. Calculate to SOa.
Calculation:
BaSO 4 X 0.3430 k 80a.
Total S02.-Determine the total 80 2 by titrating with 0.1 N iodine and
thiosulfate as described in the volumetric method.
Calculations.-Calculate the. SOa to CaS04 and the CaO to Ca(HSO a)2.
'Calculate the MgO to Mg(HSOah Express the excess of 80 2 beyond that
required ..for the bisulfites of calcium and magnesium as free 80 2.
SOaX1.700=Ca80 4 •
CaO X3.606 = Ca(HSOah
MgO X 4.625 = Mg(HSOa) 2.
Ca(HSOa) 2XO.6336 = S02.
Mg(HSO a)2 X 0.6872 = S02.
ANALYSIS OF CASEIN 50
Commercial caseins are often designated according to the method of prepa-
. ration, as acid casein, rennet casein, etc. Acid caseins are again subdivided
according to the acid used, as sulfuric, muriatic, lactic, etc. This designation
was formerly of some value, as the older methods of preparation gave products
of different characteristics according to the acid used. With modern methods
this is not true and caseins of practically the same properties are now pro-
duced with such different acids as lactic and muriatic. The name of a casein,
except for the distinction between acid and rennet caseins, therefore gives no
hint as to its properties.
Casein is a material for which there is at present no adequate system of
testing which can be applied by both producer and consumer. This is due
60 T 607 m. Tentative Standard-May 26, 1932. Official Standard-Oct. 2, 1933.
1932 PAPER AND PAPER-MAKING lY,IATERIALS
to the fact that there is a very wide gap between the information supplied by
strictly chemical tests and that necessary to tell howl the casein wilt work in
practice. This gap can be bridged only by empirical tests made to duplicate,
as nearly as possible, the conditions of actual use, and s'uch tests cannot readily
be standardized. Both types of tests are here describJd.
Sampling.-The usual methods of sampling dried granular material in
packages shall be followed, taking equal portions from 2 to 10% of the pa'ck-
ages according to the size of the shipment. The composite sample shall be
thoroughly mixed, quartered down to suitable size and placed in moisture-
proof containers. :
Appearance.-The sample should be observed for color, odor, cleanliness
and fineness of grinding. There are no exact standards for the first three
properties and only experience can enable one to tell whether.a sample is up
to the desired quality. In general, however, a good casein should be of a
creamy yellow color, free from an excessive amount of dark orange-colored
particles and visible dirt and of a mild, characteristic odor, neither moldy,
musty nor rancid. The fineness of grinding may be determined by screening
on standard screens.
Chemical Tests.-Moisture.-For accurate work a sample of· 3 to 5 grams
shaH be dried for 5 hours at 98° C. in a vacuum oven. For most work it is
sufficiently accurate to dry the sample to cpnstant weight at a tempera.wre of
lOG-105° C. in an ordinary oven. AIl weighings shaIl be made in a- glass-
stoppered weighing bottle, as casein takes on moisture rapidly when dry.
Fat.-A representative portion of the sample shaIl be very finely ground and
then 5 grams extracted with ether in an extractor of the Soxhlet type .. The
extraction shaH run for at least 16 hours or overnight, coIlecting the extract
in a weighed flask. After evaporating off the solvent, the extracted fat shall
be dried to constant weight at not over 100° C.
Routine Method.-A convenient method of moderate accuracy which utilizes the
Babcock milk tester is as follows: f
Soak 2 grams of the casein in 6 mI. of water in a small beaker and when all particles
are thoroughly saturated add, with constant stirring, 9 ml. of concentrated H 2S0 4 •
Pour the solution into a Babcock milk test bottle (I8-gram) and wash out the beaker
with a mixture of 5 m!. of water and 5 mI. of concentrated H 2S0 4, adding the washings to
the bottle. Fill to the base of the neck with dilute H 2S0 4 (5 : 4) and whirl in a centrifuge
for 5 minutes. Then add sufficient hot dilute acid to bring the fat layer completely
within the graduated portion of the neck and whirl for 2 minutes... While still hot read
the amount of fat on the gra(luated neck. As the bottle is designea 'for~a" charge of 18
grams, this reading multiplied by 9 will give the percentage of fat in the casein.
Ash.-Weigh about 2 grams of casein into a large porcelain crucible and
ignite over a low flame until complete carboniMtion has taken place, being
careful that the contents of the crucible do not burst into flame. Then ignite'
over a burner or in a mume until the ash is white qr
very light gray, cool in a
desiccator and ,weigh. The final ignition should be at as Iowa temperature
as is consistent with obtaining a white ash. The procedure should not be
carried out in platinum as this metal is attacked by the phosphorus in the
·casein. '
If oxygen is available the operations may be. ~{ried 0\11, in a Rose crucible
and the final ignition assisted by a stream of oxygen.
PAPER AND PAPER-MAKING MATERIALS 1933
Alkali.__:_Add about 5 m!. of distilled water to the ash and warm. Then
add a drop or two of phenolphthalein indicator and if the ash is alkaline titrate
it with 0.1 N HCI until colorless; then add 2 drops of methyl orange and com-
plete the titration to a pink color. Calculate the total titration to Na 20.
Calculation:
1 m!. 0.1 N HCI=0.0031 gram Na 2 0.
NOTE.-The ash of a pure casein should not be alkaline to phenolphthalein or to
methyl orange. If it shows alkalinity, test it qualitatively for borax, sodium carbonate
and sodium phosphate.
Starch.-Warm, a portion of the sample with water. Cool and add a few
drops of very dilute iodine solution. A blue color indicates starch.
NOTE.-If added alkali has been found in the casein, make the solution slightly acid
before adding the iodine solution..
Total Acidity.-Weigh 20 grams of cas.ein into a beaker and add 100 m!. of
water. After soaking for a short time titrate with 1.0 N or 0.5 N NaOH,
ruhning in a little at a time and warming on the steam bath after each addition
until·the end-point is reached. Litmus paper, moistened and dipped into the
hot solution shall be used as an indicator and the end-point shall be considered
as reached when blue litmus is turned very slightly red and red litmus is turned
very slightly blue. Report the result as percentage of NaOH required to
neutralize the casein.
, Calculation:
1 m!. 1.0 N alkali =0.04 gram NaOH.
Soluble Acidity.-Soak 20 grams of the casein in 200 mI. of water at room
temperature for at least 1 hour, filter and titrate 50 m!. of the filtrate with
0.1 N NaOH, using methyl orange indicator. Report the acidity as the num-
ber of m!. of 0.1 N alkali required to neutralize 1 gram.
Niirogen.-The nitrogen is not usually determined in casein except for
some special purpose. When it is desired the usual Kjeldahl method should
be used.
NOTE.-Authorities differ as to the factor for converting N to casein and it is known
to differ with different lots of casein. According to Browne, NXB.38 will give the per-
centage of pure casein: while Hopfner and Burmeister state that NXB.61 will give the
fat-, ash-, and moisture-free casein present. To convert to commercial, air-dry casein as
received, the factors have been found to vary from 7.69 to 9.10 for different samples.
It is suggested that the factor 8.00 be arbitrarily selected and assumed to apply in all
cases where it is desired to convert nitrogen to commercial air-dry casein.
Empirical Tests.-Insoluble Matter.-Weigh 100 grams of casein into a
liter beaker, add 300 in!. of water and stir thoroughly. After about 10 minutes
add a solution of 15 grams of Na 2 CO a (soda ash) in 100 m!. of water and stir
until thickening takes place. Put the beaker in a water bath at a temperature
of about 60° C. and let it stand, with occasional stirring, for 3 hours. At the
end of this period add 400 m!. of boiling water and stir thoroughly for about
5 minutes. Pour the solution upon a small brass, 200-mesh sieve which has
been weighed to 0.1 mg. and soaked in boiling water. Wash the sieve and
residue thoroughly with boiling water, then dry and weigh. Report the per-
centage of insoluble matter.
1934 PAPER AND PAPER-MAKING MATERIALS
The coated sheets should be dried in the air at room temperature until
fully dry and tested with sealing wax for strength. For this test heat a stick
of sealing wax on one end until it almost drips and press the hot end firmly
down on the surface of the coated paper. When it is fully cold and hard pull
it off with a .quick, vertical pull. When doing this the paper must be held
firmly; this may be done by using a large washer wh~ch will fit over the wax
and hold the paper down entirely around it. If insufficient casein is. used, the
coating will pull off without taking any fiber with it; while if excess casein is
used, the surface of the wax will be entirely covered with coating to which
fibers are adhedng. There is generally an intermediate" critical point" when
the central area of the wax shows fibers and the edges coating only. This/
would be considered just on the line between weak and strong.
The test should be reported as the number of pounds of clay held by one
pound of casein in the coating which gives the critical point. This figure is
obtained by dividing 100 by the number of grams of casein required to give
the critical point.
NOTE.-The results obtained by this test are influenced by the kind of clay used as
well as the kind of paper to which the coating miXture is applied. Of still greater im-
portance is the thickness of coating applied and it has been proved that for accurate work
it is necessary to keep the weight per ream of coating constant within 1 pound on either
side of the desired amount. This cannot be accomplished by spreading the coating
with a brush or even by means of a hand scraper, and the only aatisfactory solution is to
construct a machine which Vlill apply a very uniform and controllable amount.
The principal use of salt cake in the pulp and paper industry is in the manu-
facture of sulfate pulp. Salt cake is an anhydrous sodium sulfate which
contains varying amounts of impurities. The principal impurities usually are
sulfuric acid, common salt, and sulfates of Fe, AI, Ca and Mg. "Chrome Salt
Cake," obtained as a by-product in making chromic acid, has a yellow color
and contains Na 2 Cr 207 (and/or Na 2 Cr04).
A complete analysis of salt cake is somewh·at complicated and involves a
considerable number of determinations, as outlined below. Since some mills
are mainly interested in the acidity and the Na 2S04 content, directions for a
. " short analysis" are also given.
Preparation of Sample.-Mix the sample thoroughly, breaking down or
grinding any lumps; quarter down to about 100 grams; grind the quartered
portion, if necessary, to about 50-mesh or finer; and place in a separate stop-
pered bottle.
COMPLETE . ANALYSIS ,
Chloride.-:--To the solution which has been titrated for acidity add a drop
or two of methyl orange or methyl red indicator, make very slightly acid with
a few drops of dilute H 2S0 4 , and bring back to neutrality by adding about 1
gram of powdered CaCOa and stirring. Set the flask on a white surface or
transfer ·the contents to a porcelain dish; add 10 drops of K 2 Cr04 indicator
solution (10%) and titrate with 0.1 N AgNO a until the brick-red coloration
due to Ag 2Cr04 remains just visible after stirring. Calculate the titration to
percentages of CI and NaCI (and also to the equivalent of Na 20 for use in the
final calculations). '
Calculations: 1 ml. 0.1 ~ AgNO a=0.00355 gram Cl. .
=0.00585 gram NaC!.
= 0.00310 gram N a 20.
Ferrous Iron.-To 5 grams of the original p[epared sample in an Erlen-
meyer flask add 100 ml. of warm water and 5 ml. of concentrated H 2S04 •
Warm on the steam bath a few minutes, just long en~ugh to dissolve all soluble
matter. Cool under a water tap and titrate at once with 0.1 N KMn04.
Calculate the titration to percentages of FeO and FeS04 (and also the equiva-
lent of Fe 20a).
Calculations: 1 ml. 0.1 ~ KMn04=0.0072 gram FeO. I
=0.0152 gra~S04'
= 0.0080 grain Fe20s..
PAPER AND PAPER-MAKING MATERIALS 1937
catch fire. After completing the charring, ignite until white over an ordinary
burner, with the crucible uncovered. Cool and moisten the precipitate with a
few drops of dilute H 2S0 4. Heat very carefully to drive off the acid and finally
ignite to constant weight, cooling between weighings in a desiccator. Calculate
the weight of BaS04 to percentage of S03.
Calculation; BaS04XO.3430=S03'
Final Calculations.-From the total NazO subtract the NazO equivalent
to the NaCI and to the Na ZCr Z07; calculate and report the remaining NazO
as Na ZS04' .
From the total S03 subtract the S03 equivalent to the NaZS04, Al z(S04)3,
FeS04 and FeZ(S04)a; calculate and report any excess of S03 as free HzSO~.~
(The S03 equivalent of each of the sulfates is obtained by subtracting the cor-
responding oxide percentage from the sulfate percentage; e.g., 1.52% FeS04
-0.72% FeO =0.80% S03.)
Form of Report.-It is suggested that the results of a complete analysis to
be reported in the following form:
Per Cent
Moisture, loss at 100° 0 ............................. .
Insoluble in Water ................................. .
Acidity, calculated as H 2S0 4........................ .
Chloride, C1. ...................................... .
Ferrous Oxide, FeO ................................ .
Ferric Oxide, Fe 20 •.................................
Alumina, AlzO a ••••.•••..••••••.•..•...•..••..•..•••
Lime, CaO ......................................... .
Magnesia, MgO ................................... .
Chromium Trioxide, OrO •............... ; ........... .
Sulfur Trioxide, S03 ...... .' ......................... .
Combined as follows:
Sodium Chloride, NaC!. ... , . " .. " ............. ,
Sodium Bichromate, Na2Cr 207 ................... .
Aluminum Sulfate, Al Z(S04)3. ................... .
Ferrous Sulfate, FeSO 4 • • • • • • • • • • • • • • • • • • • • • • • • • •
Ferric Sulfate, Fe2(S04) •..............•..........
Calcium Sulfate, CaS04 ......................... .
Magnesium Sulfate, MgS0 4..... '................ .
Sodium Sulfate, Na 2S04 ........................ .
Free Sulfuric Acid, H 2S0 4....................... .
SHORT ANALYSIS
Rosin is the residue obtained after disti~ling off the volatile matter from
turpentine gum. It is graded according to color. The higher the temperature
to which the distillation is carried, the' darker is the resulting product. The
common commercial grades, in the order of their color, from the palest and
best to the cheapest and darkest, are: X, WW, WG, N, M, K, I, H, G, F, E,
D, and B. Grades G, F, and E are most frequently used in the paper industry;
The acid number is generally considered the best index, in selecting rosin for
making rosin size. Other tests which give valuable data for certain purposes
are saponification number, unsaponifiable matter, toluol-i,nsoluble matter, and
ash. ' /
, Sampling.-Rosin is usually received in wooden oarrels of 500 pounds
gross weight or metal drums containing a gross we~ght of 280 pounds. These
containers are stripped off, leaving a solid block of rosin. The sample for
test should be taken from approximately the center, of the block. Ordinarily it
is sufficient to sample two containers selected at random froII\._,a shipment. In
cases of dispute the number of containers to be sampled shall be,.mutually
agreed upon. The sample shall be reduced to laboratory size by 'crushing to
pass a half-inch sieve, mixing and quartering. The final laboratory sample
shall be prepared immediately before analysis and shall consist of approximately
200 grams. It shall be ground to pass a No. 10 siev~ mixed thoroughly, and
placed in a wide-mouthed bottle of such a size that it will be completely filled
with the powdered rosin. I
NOTE.-As rosin has a marked tendency to oxidize very rapidly, samples should be
kept in lump form as much as possible.
I
Color Grade.-To determine the color grade of rosin' a-set ofstandard cubes
-'-
52 T 621 m. Tentative Standard-Feb. 19, 1934. OffiCial Btandard-ApriI15, 1935.
PAPER AND PAPER-MAKING MATERIALS 1941
1943
1944 POISONS
Classification of Poisons
Nearly all of the co~mon poisons may be placed in lone 'of three groups: I
Group I. Poisons volatilized without decomposition by distillation with ste~m
from an acid solution. I .
Group II. Organic poisons. Most of the poisons in this group can be Mpa-
rated from extraneous matter by treatment with hot alcohol containing
tartaric acid. Alkaloids, glucosides, bitter principles and many organic
compounds come in this group.
Group III. All poisonous metals.
NOTE. Caustic alkalies, mineral acids, and a few special poisons such as oxalic acid
do not adapt themselves to the above scheme and will have t() be treated separately ..
General Procedure
Organs of the human body should first be weighed and then a representative
weighed sample reduced to a finely divided mass (a clElan small meat grinder
answers nicely for this purpose). .
It is advisable always to preserve a portion of the material in its original
form for possible further examination.
About one tenth of the thoroughly mixed materiaJ is used in testing for
each one of the three groups, the balance being reserved. for checking. If only
a small amount of the original material is available, tests lor aIL three groups 01
poisons may be made with the same port~on. In this case, after removal of
volatile poisons, the residue should be divided into two unequal portions, the
larger being used for non-volatile organic poisons and the smaller, togethef
with the residue from the non-volatile organic, for metttllic poisons.
Blank tests should be run upon all reagents used !Iond ·the chemist should
familiarize himself with the various tests before applyin~ them to the unknown.
At least three confirmatory tests, different in character, ehould be applied to the
unknown substance before the presence of a poison can be definitely established.
Whenever possible these confirmatory tests should inch.1de form, color, odor or
taste. /
Group I
Place some distilled water in a clean round-bottomed flask connected with
a Liebig condenser. Distil about 10 ml. and test distillate with silver nitrate
solution. Absence of a precipitate indicates that tlle apparatus is clean.
Next pour out the distilled water and transfer a weighed portion of the finely
divided and thoroughly mi~ed material to the flask, ttdding enough distilled
water for free distillation. Then add tartaric acid solution gradually until the
mixture is acid after thorough shaking. The distillate should be received in a
test tube in which a little distilled water has been,placed. The distillate may
contain: .
Yellow Phosphorus CarboliC Acid
Hydrocyanic Acid Anilirle
Chloroform Chloral Hydrate
Ethyl Alcohol Carbon Disulfide
Methyl Alcohol Benzaldehyde I
Acetone AmylJ{itrite
Iodoform AmY! Alcohol
Nitrobenzene
POISONS 1945
Collect the distillate in several portions. Test the first portion of 5-10 ml.
for hydrocyanic acid, chloroform, ethyl alcohol, methyl alcohol, acetone,
iodoform and nitrobenzene. The other portions of 10-20 ml. will contain the
less volatile substances, carbolic acid, etc. Observe the reaction of the dis-
tillate; if acid, test for hydrocyanic acid .. Observe the odor of the distillate.
A number of the above substances have characteristic odors. When there is
reason to believe that a certain substance is present, confirm the results by
, milking other characteristic tests. It is seldom necessary to test for all members
of the group. Test a portion of the first distillate with silver nitrate T. S.
Faiiure to obtain a precipitate eliminates hydrocyanic acid.
YELLOW PHOSPHORUS
HYDROCYANIC ACID
In testing for hydrocyanic acid examine the brain the contents orAhe
r
stomach and intestines, and organs rich in blood such as the liver, and heart.
Examination should be made as soon as possible, as hydrocyanic acid rapidly
disappears through volatilization and combination wi'th sulfur of !proteins
during putrefaction forming sulfocyanates. I
Note cautiously the odor of the material as well as the odor of the distillate.
Hydrocyanic acid has a characteristic" peach kernel" odor.
Apply the following tests:
1. Schonbein-Pagenstecher Test
In case of suspected cyanide poisoning this test should precede the distilla-
tion. A small quantity (15 grams) of the finely divided material is placed in a
small flask together with enough tartaric acid solution to moisten it thoroughly
and a piece of guaiac-copper" paper" a is suspended in the neck of the flask
in such a ~anner as not to touch the liquid. The flask and contents are then
gently warmed on the water bath. If the paper is not turned blue or bluish·
green, it is positive proof that neither HCN nor a readily decomposed cyanide is
present. A positive test means only that HCN or cyanide may be present as
many substances such as Cl, HCl, NHa and Hij"Oa produce this same change in
color.
2. Prussian Blue Test 4
To a small portion (3 mI.) of the distillate in a test tube add a few drops of
KOH solution, then 1 or 2 drops of a FeS04 solution and 1 drop of a FeCla-
solution. Shake and warm gently .• Acidify with dilute HCI. \ If cyanide is
present in quantity a precipitate of Prussian blue will appear immediately, but
if present only in traces the solution will acquire a blue or bluish green color
which on standing 10 to 24 hours will throw down a flocculent precipitate of
Prussian blue. The maximum sensitivity of this reaction is 1 , 5,000,000.
Since formaldehyde is the principal ingredient of most embalming fluids, the
Prussian blue test cannot be relied upon as a neg~tive lest when working with
embalmed tissues.
3. Sulfocyanate Test
To another portion of the distillate in a small evapor!!ting dish add 2 or 3
drops of KOH solution, enough yellow ammonium sulfide tocolor. ,the solution
yellow and evaporate to dryness on the steam bath. Dissolve in a-lew drops
of water and acidify with a little dilute HCI. Filter through a double filter
paper into another small evaporating dish. Add.2 or 3 drops of dilute FeCla
solution. If HCN is present in the distillate, a
reddish to blood-red color
will appear, which is discharged upon the addition pf a few drops of HgCl 2
solution. The color SQ produced is due to ferric sulfocyanate. The maximum
3 "Guaiac-copper" paper is prepared by saturating strips of filter 'paper with a
freshly prepared 10% alcoholic tincture of resin of guaiac, drying and, just before using,
moistening with a 1 to ·1000 CUS04 solution.
4 The author has found that hydrocyanic acid fails to,r.e!lPond .to this test in the
presence of formaldehyde, due probably to the following.reacl~on: •
HCOH+HCN = HOCH 2CN.
POISONS 1947
sensitivity of this test is 1 : 4,000,000 and is not affected by the presence of
formaldehyde.
4. Silver Nitrate Test
Acidify 1 ml. of the distillate with dilute HN0 3 in a small test tube and add
2 or 3 drops of AgN0 3 sloution. If HCN is present, a white curdy precipitate
resembling that of AgCI is produced which dissolves in NH 40H. This reaction
cannot be due to HCI because when a very dilute solution of HCI is distilled
HCI is not found in the distillate. The maximum sensitivity of this test is
1 : 250,000.
5. Silver Cyanide Crystal Test
Acidify 1 m!. of the distillate with dilute HN0 3 in a small test tube. Place
• a drop of dilute AgN0 3 on a microscope slide and carefully invert the drop over
the center of the tube so as to avoid touching the side of the tube. Allow to
stand for 15 to 20 minutes, then cover with a coverglass and examine under the
microscope, using a power of from 50 to 100 diameters. If HCN is present,
characteristic 'crystals of AgCN will appear under the microscope. (See Fig.
283.)
CHLOROFORM
Phenylisocyanide Test
To a small portion of the distillate add 1-2 drops of aniline oil and ~ few m!.
of KOH solution and heat gently. If chloroform is present, the offensiv,e,
characteristic odor of phenylisocyanide is produced. This test is sensitive to 1
in 5000. This odor is produced by other substances such as chloral and ,chloral
hydrate and for this reason other tests should be applied.
ETHYL ALCOHOL
petection
1. Lieben's Iodoform Test
Warm a small portion of the first distillate to 40° pr 50° C. Add )-2 ml. of
an aqueous iodo-potassium iodide solution and enough KOH solution to give a
distinct yellow to brownish color. If alcohol is present, a yellowish white to
lemon yellow precipitate of iodoform is formed. If the solution is very dilute,
some time may be required to form the precipitate. When fqrmed slowly the
crystals may be identified under the microscope. This test is/very delicate but
is not characteristic of ethyl alcohol ,because othe1"'primary alcohols (except
POISONS 1949
methyl alcohol) and many secondary alcohols produce this reaction: (See
Fig. 281.)
2. Chromic Acid Test
Warm a small portion of the distillate with dilute H 2S0 4 or HCI and add
1-2 drops of a very dilute K 2 Cr 201 solution. The color of the liquid will change
from red to green and give off the odor of acetaldehyde. Many other volatile
organic compounds react in this manner.
4. Vitali's Test
Thoroughly mix a small portion of the distillate in a glass dish with a small
piece of solid KOH and 2 or 3 drops of C8 2 • Let stand a short time without
warming. When most of the CS 2 has evaporated add a drop of ammonium
molybdate solution and then an excess of dilute H 2S0 4. A red color is produced
if alcohol is present. Acetone and acetaldehyde will produce a similar color.
A. 5% solution of alcohol gives this test distinctly.
METHYL ALCOHOL (WOOD ALCOHOL)
intestinal tract and appears in the liver, stomach, blood, kidneys, brain and
urine in a very short time after taking. 1
1. If present in quantity, this acid may be recognized by its odor.
2. Millon's J'est .. \
Heat'1 m!. of the distiUate with a few drops of Millon's Reagent. (See
Special Reagents.) If carbolic acid is present even in minute. quantity, a dis-
tinct red color is produced. The absence of this color is an excellent indication
of the absence of phenol but other substances such as the three cresols and
salicylic acid may produce this color and for this reason other tests should be
applied if this test is positive.
3. Bromine Test
To 1 ml. of the distillate add 1 or 2 drops of freshly prepared saturated
bromine water. If phenol is present in quantity a yellowish white precipitate
is immediately thrown down, but if highly diluted some time may be required.
The precipitate should be examined under the microscope for characteristic
crystals of tribromphenol. (See Fig. 282.)
Phenol in a dilution of 1 : 50,000 yields 01). standing a precipitate containing
some of the characteristic crystals.
Substances like salicylic aldehyde and salicylic acid also produce this pre-
cipitate and other tests should, therefore; be applied.
5. Hypochlorite Test /
Add a little NH 40H to a small portion of the:distillate and then 2-3 drops of
freshly prepared solution of calcium or sodium hypochlorite and warm. If
phenol is present in quantity a blue color is produced, but if very dilute only
green or blue-green color appears. The maximum sensitivity of this reaction
is 1 : 1000.
6. Nitrite Test
To a small portion of the distillate add a few drops of a dilute alcoholic solu~
tiop. of ethyl nitrite and underlay the mixture in a test tube with concentrated
H~S04. If phenol is present, a red zone will appear a.t the point of contact of the
two liquids. The maximum sensitivity of this reaqtion is 1 : 10,000.
CHLORAL HYDR.A'l'E
Group II
ALKALOIDS, GLUCOSIDES AND OTHER ORGANIC COMPOUNDS
Alkaloids form salts with tartaric acid insoluble in ether and chloroform but
soluble in alcohol and water. These alkaloidal salts are decomposed by alkalies
liberating the alkaloid which is insoluble in water, but soluble in ether or chloro-
form. The above is the principle by which alkaloids are separated from
foreign material. .
Meth?d of Analysis
The substance is finely hashed, if solid, and then treated with several vol~
urnes of purified alcohol 5 and made slightly acid with tartaric acid. The
mixture is placed in a flask and heated under a reflux condenser· for about 30
minutes. After cooling it is filtered .and the residue washed with purified
alcohol. The alcoholic extract is evaporated on a water bath to the consistence
of a syrup. From 3 to 4 volumes of purified alcohol are then gradually stirred
in, the solution allowed to stand an hour or more, filtered and carefully eva po-
6 The alcohol used should be purified by treating it with aboot 0.1 % of tartaric acid,
allowing to stand 24 hours and then distilling.
1952 POISONS
rated on a water bath to a thin syrup. This. syrup is now treated with about
50 ml. of distilled water and filtered. I
, NOTE.-If croton oil or nitroglycerine are suspected, they should be looked for in the
residue insoluble in water. ' I
The filtered aqueous acid solution should be placed in a separatory funnel
and extracted with ether, and the extractions repeated until the ether on evapo-
ration gives no residue.
Unite the ether extractions, wash once with a small quantity of water, filter
through dry paper and evaporate. The residue may contain:
Fats and Oils Acetanilid Antipyrine
Ptomaines Phenacetine Salicylic Acid
Picrotoxin Chloretone Benzoic Acid
Ccrlchicine Brometone Veronal
Picric Acid Essential Oils Trianal
Menthol Cantharidin SuIfonal
Camphor Caffeine
Phenol
Cresol
Thymol
Note the general appearance and taste of the residue. If bitter, test for
picrotoxin and colchicine. Veronal and trional have a bitter taste arid' give
characteristic crystals. Sulfonal is tasteless and gives prismatic crystals.
The aqueous liquid is next rendered slightly alkaline with N aOH and ex-
tracted wi_th ether as before.
The extracted substances from which the solvent has been evaporated should
be weighed so as to determine the amount present. If material so recovered
is crystalline, the weight may be taken as representing the actual amount
present. If not crystalline, the extract may be purifi'd in most cases by dis-
solving in dilute acid, making the solution alkalrne as/before and re-extracting
with the volatile solvent.
The residue from the evaporation of the ether:may contain any of the follow-
ing substances:
PtomaineE! Atropine Codeine Antipyrine
Coniine Hyoscyamine Narcotine Quini-ne
Nicotine Scopolamine Pilocarpine Caffeine
Aniline Cocaine Sanguinarine
Toluidine Physostigmine Berberine
Veratrine Brucine
Strychnine Hydrastine
Carefully_ acidify the aqueous alkaline residue remaining from the preceding
ether extraction with dilute HCI and then make alkaline with dilute NH 40H,
using litmus paper for both tests. Extract repeatedly with hot chloroform
containing 10% by volume of alcohol.
POISONS 1953
The residue from the evaporation of the chloroform may contain:
Morphine
Apomorphine
Narceine
ACETANILID
Dose, 2-10 grains.
Lethal dose, 60 grains has proven fatal, less is dangerous to a person with a
weak heart.
Phenylisocyanide Test
Dissolve a small portion of the residue in 5-6 ml. of a boiling solution of
alcoholic potash. Cool, add 2 or 3 drops of chloroform and again heat. If
acetanilid is present, the offensive odor of phenylisocyanide is produced. The
alcoholic potash decomposes acetanilid into aniline and potassium acetate.
The' former with chloroform gives phenylisocyanide.
Dose, 8 to 15 grains.
Lethal dose, 120 grains has proven fatal.
Distribution: Veronal should be looked for in the stomach and bladder
contents.
t 'tIS· 1
..
...
.., ~
, ~ .. .
, t" ~
ftf#
.-.. \
-=>=,_ I
j
,FIG. 279.-Nicotine 1 : 1000. With FIG. 280.-Strychnine 1 : 500. With
Gold Chloride. Ammonium Thiocyanate.
Distribution
Nicotine should be.sought in the stomach, lungs arid liver.
Detection
If present in large quantities, nicotine may be detected by its odor. If nico-
tine is suspected, the ether extract from NaOH solutioh should be evaporated
without heat. If nicotine is present, oily drops of a colorless or yellowish liquid
which reddens phenolphthalein and has a peculiar odor somewhat resembling
stale tobacco smoke will remain in the residue, If exposed to the light, the oil
becomes yellow or brown, thicker and finally resinous. If heated, it volatilizes
completely, giving off white fumes.
If a drop of Hel be placed on a watch glass and inverted over a similar watch
glass containing nicotine alkaloid, a white cloud is produced which is, however,
not so dense as that produced by coniine under like conditions and no crystars
are produced as with coniine.
Place a drop of an aqueous extract of the oily residue on a microscope slide,
add a drop of a solution of gold chloride, let stand for several minutes and
examine under the microscope. If nicotine is present, small characteristic
crystals will appear upon standing.
QUININE
Dose 15 grains.
Detection
Fluorescence Test
Dissolve a portion of the residue in a little dilute sulphuric acid. If quinine
is present, this solution will exhibit a vivid blue fluorescence.
Thalleioquin'Test I
Dissolve a portion of the residue in a few dl,'ops df very dilute acetic acid
and add few drops of saturated chlorine or bromine water and follow with
excess of ammonium hydroxide. The liquid acquires an emerald-green color.
Herapathite Test'
To a small portion of the fesidue add 20 drops of an acRimixt_ure (30 drops
of acetic acid, 20 drops of absolute ethyl alcohol and 1 drop of dilute sulphuric
acid) and heat to boiling. 'Finally add 1 drop of an alcoholic solution of
iodine (1 : 10). When the solution has stood for a few minutes, green leaflets
with a metallic luster will form. '
STRYCHNINE
Bloxam's Test
Add 1-2 drops concentrated HNO a to a portion of the solid residue in a por-
celain dish and gently warm. On adding a small crystal of KOlOa, a scarlet
color is produced if strychnine is present. NH 4 0H changes the color to brown
and a brown precipitate falls. On evaporating slowly to dryness, a dark green
residue is left. This residue is soluble in water, forming a green solution which is
changed to orange-brown by KOH ana green again by HNO a. This reaction
distinguishes strychnine from any of the alkaloids commonly occurring in cases
of poisoning.
Bichromate Test
'PlaJ!e a drop of an acid extraction of the residue on a microscope slide and
add a drop of a dilute solution of K 2 0r 2 07. Let stand for a few minutes and
examine under the microscope. If strychnine is present, characteristic crystals
will be found. Dry the crystals and add a drop of concentrated H 2S0 4, holding
the slide over a white surface. A deep blue color which changes through deep
violet, purplish red, orange to yellow is produced if strychnine is present.
NOTE.-A mixture of hydrastine and morphine causes this same color reaction but
this mixture is not possible if the extractions have been properly carried out.
Sulfocyanate Test
Place a few drops of the acid extraction on a microscope slide and add a drop
of a dilute solution of ammonium sulfocyanate. Strychnine sulfocyanate
produces characteristic needle-shaped crystals.
BRUCINE
Detection
Dissolve a small quantity of the residue in 1 or 2 drops of concentrated
HN0 3 • If brucine is present, a blood red color will appear, which soon changes
to yellowish red and finally to yellow. Add a few drops of a'freshly prepared
dilute solution of stannous chloride to the yellowish red or yellow solution. An
intense violet color will appear. Heat usually changes this color to yellowish
red but more stannous chloride solution restores the violet color.
Dissolve a small portion of the residue in 2 drops of dilute HOI. Place a
drop of this solution on a microscope slide and add a drop of pla~inum chloride
solution. If brucine is present, characteristic rod-like crystals separate and may
be compared under the microscope with those produced in a solution of brucine.
These crystals are formed in solutions as dilute as 1 to 20,000. The best
crystals are obtained from the more dilute solutions.
1958 POISONS
ATRQ~INE
Dose, 1/120 to 1/60 grain.
Lethal dose, 1/20 grain. Much larger doses have, been followed by recovery.
Detection
Vitali's Test
Dissolve a small portion of the dry residue from the ether extraction of solu-
tion made alkaline with NaOH in a few drops of fuming HNO a and evaporate
to dryness on the steam bath. Cool and moisten the yellowish residue with a
few drops of· a 4% solution of KOH in absolute alcohol. If atropine is present,
an evanescent violet color will appear. This color reaction is also obtained with
hyoscyamine, scopolamine, strychnine and veratrine, and therefore other tests
should be applied.
Physiological Test
Dissolve a small portion of the residue in a few drops of very dilute H 2S0 4
and place a drop of this solution in a dog's or cat's eye. If atropine is present,
a noticeable enlargement of the pupil WIll be observed. (Note: Great care
must be exercised if .this test is applied: to the human eye.) One drop of a
1 : 130,000 solution ,of atropine will produce this enlargement.
Dissolve a small portion of the residue in 2 drops of dilute H 2S0 4 • Place a
drop of this solution on a microscope 'Slide and add a drop of Wagner's Reagent.
If atropine is present, an amorphous 'precipitate appears and slowly forms
numbers of small characteristic crystals which may be observed under the
microscope. This test is sensitive to 1 in 8000.
COCAINE
Detectionl, /
The material should be examined for cocaine as soon as possible because
cocaine undergoes changes in the viscera which are said to change it rapidly into
ecgonine,
Dissolve a small portion of the residue in water and_touch it to the tongue.
Cocaine produces a t~mporary anesthesia. .. --.
To a small portion of the residue in a porcelain dish add a few drops of
strong HNO a and evaporate carefully to dryness. Treat the residue with a few
drops of alcoholic potash and warm. If cocaine is present, a characteristic
sweet odor will be given off.
Dissolve a small portion of the residue in 1 dr 2 drops of dilute HCl. Place
a drop on a microscope slide and add a drop of gold chloride solution. Cocaine
produces afteJ; a short time very characteristic crystals, theform depending on
the dilution. This test is sensitive 1 : 20,000.
CODEINE
Dose, Y2 to 1 grain.
Lethal dose, 5 grains. Much larger doses have been recovered from.
POISONS 1959
Detection
Nitric Acid Test
To a small portion of the residue in a porcelain dish add a few drops of a cold
25% solution of nitric acid. Codeine produces a yellow color which soon
changes to red. Concentrated nitric acid produces a reddish brown color.
Froehde's Test
To a small portion of the residue in a porcelain dish add a few' 'drops of
Froehde's Reagent (see Special Reagents) and warm very gently over a very
small flame. Codeine produces a yellowish color which soon changes to green
and finally to blue.
Formaldehyde-Sulfuric Acid Test
To a small portion of the residue in a small porcelain dish add a few drops
of concentrated sulfuric acid to which has been added a trace of formaldehyde.
Codeine produces a reddish violet color which changes. .
to blue-violet.
Marme's_Reagent Test
Dissolve a small portion of the residue in 2 drops of dilute H 2S0 4• Place a
drop of this solution on a microscope slide and add a drop of Marme's Reagent
(see Special Reagents). Characteristic crystals soon begin to form in solution
of codeine as dilute as 1 : 2900 and may be observed under the microscope.
CAFFEINE
Dose 3 grains.
Detection
Murexide Reaction
Treat small portion of the residue in a porcelain dish with a few drops of
strong hydrochloric acid and a minute crystal of potassium chlorate, and
evaporate the liquid to dryness on the water bath. When cold invert the dish
over a vessel containing a few drops of ammonia when a characteristic purple
coloration is produced. •
Theobromine and, xanthine give similar reactions to caffeine with an oxidiz-
ing agent and ammonia. The purple colorations due to caffeine and theo-
bromine are decolorized by adding alkali hydroxide solution, but that due to
uric acid is changed to blue.
NARCOTINE
Dose, 2 to 15 grains.
Dete6tion
Dilute Sulfuric Acid Test
Dissolve a small portion of the residue obtained from the ether extraction of
the NaOH solution in a little dilute H 2S0 4 and evaporate in a small porcelain
dish on the steam bath. If narcotine is present, the residue has a reddish yellow
color which changes on stronger heating to crimson red. As the acid begins to
evaporate, blue-violet streaks radiate from the margin and the entire liquid
finally assumes a dirty red-violet color.
Froehde's Reaction
To a small portion of the residue in a porcelain dish add a few drops of
Froehde's reagent. If narcotine is present, a greenish color is produced which
immediately changes to cherry red. Gentle' heat hastens the appearance of
the cherry red color which is quite persistent.
Froehde's Test
To a small portion of the residue in a porcelain dish add a drop of Froehde's
Reagent. Morphine produces a violet color which changes through blue to
dirty green and faint red. Other alkaloids of the opium group give similar
color reactions.
Iodic Acid Test
Dissolve a small quantity of the residue in dilute H 2 S0 4 , add a few drops of
iodic acid and shake with chloroform. A violet color appearing in the chloro-
form layer may be due to other reducing agents as well as morphine.
Marme's Test
Dissolve a small portion of the residue in 2 drops of dilute HCl. "Place one
drop of this solution on' a microscope slide and add a drop of Marme's Reagent.
If morphine is present in solution to the extent of 1 to 1000, delicate silvery
needle-shaped crystals will form slowly around the edges of the drop and in
more concentrated solutions these crystals gfoup together to form rosettes as
shown in Fig. 284.
OPIUM
Dose, 1 grain.
Lethal dose, 4 grains has produced death. Larger amounts have been recovered
from.
The principal constituents of opium are morphine, codeine, narcotine, nar-
ceine, and meconic acid.
Narcotine can be extracted from an aqueous solution of opium by shaking
out with ether, and evaporating the ether. Meconic acid can be obtained by
adding lead acetate to an aqueous solution of opium and filtering off the precipi-
tate of lead meconate (the filtrate will contain morphine acetate). The pre-
cipitate of lead meconate is decomposed by treating it with dilute sulfuric acid
and gently boiling the mixture, filtering and neutralizing. On adding a drop
of ferric chloride solution, a deep red color, not easily destroyed by a solution
of mercuric chloride or hot dilute sulfuric acid, indicates meconic acid. Mer-
curic chloride instantly bleaches the color caused by sulfocyanates while the
boiling dilute sulfuric acid destroys that due to acetates by decomposing the
acetate and expelling acetic acid.
The solution of morphine acetate should be treated with hydrogen sulfide
to remove excess of lead, the solution filtered and the filtrate evaporated to a
1962 POISONS
soft extract on a water bath. This extract. treated with alcohol, filtered and
evaporated will give morphine acetate to which the I chara9teristic tests for
morphine should be applied.
The presence of morphine, narcotine and meconic !).?id definitely establishes
the presence of opium. .
PTOMAINES
Preparation of Solution
Place from 50 to 100 g. of the finely chopped m~terial in a lOOO-ml. boiling
flask. Dilute with water if necessary until a fluid mixture is obtained and add
pure concentrated HCI and crystalline KCI0 3 in the proportion of 25 ml. HCI
and 2 g. KCI0 3 to every 100 g. Allow the mixture to stand cool for 24 hours
and heat on the steam bath (preferably under a refl-qx.condetiser) with frequent
agitation and the occasional addition of KC10 3 (D,.lj·g. at a time) until the con-
tents of the flask are yellow and fluid except for fatty and white granular matter
POISONS 1903
in suspension and until the liquid does not darken on heating for half an hour
without adding KCIO a. If, during the heating, the addition of KCIO a does not
produee any visible reaction, add more concentrated HCI in the proportion of
10 m!. HCI to every 100 g. material used and continue as directed above. Care
should be used to keep the quantities of HCI and KCIO a as small as possible, as
an excess of HCI may prevent the precipitation of some of the metals by H 2S.
Expel the excess of chlorine by passing a rapid stream of CO 2 through the cool
liquid or by adding small quantities of NaH80 a until a distinct odor of 80 2 is
obtained. Filter the solution through a wet filter and wash the residue
thoroughly with hot water. If a large amount of fatty residue is left, it should
be extracted two or three times with small quantities of hot water acidulated
with HCI and these filtered extracts added to the main filtrate. If silver and
small quantities of lead, thallium, mercury or barium are suspected, this residue
should be examined qualitatively for these metals. If a large excess of HCI is
present, add a concentrated solution of Na 2 COa to such an extent that the re-
action still remains distinctly acid. Place the liquid in a flask fitted with a
'two-hole rubber stopper and tube dipping below ,the liquid and pass pure H 2S
through it at the temperature of the steam bath for 2 hours. Allow to stand
cold for eight or ten hours and filter. A colored precipitate at this stage may
mean nothing, as it may be due to organic thio-compounds, and therefore the
precipitate and filtrate should be tested by any good method of qualitative
analysis for metals.
Detection of Arsenic
Place 10 grams of the finely divided material in a IS0-ml. beaker, add 10 ml.
of concentrated HNO a, cov~r and allow to stand over night. Then add 5 ml.
concentrated H 2SO.{ and warm. When the material starts to turn brown, add
concentrated HNO a 1 ml. at a time until SOa fumes are given off without the
solution turning brown. Cool; carefully add 10 ml. of water and boil down to
80 a fumes. Cool, dilute, add 0.5 g. KI, warm to 90° C., add dilute solution of
SnCh to de colorize, cool and run in the regular small Gutzeit apparatus.
Blood Solution
Saturate a small amount of defibrinated blood with CO. This is done
readily by passing house gas containing CO through blood for a minute or two
or preferably by using air containing 3 to 5% of CO. Prepare a 1 to 20 dilution
of this blood with water. Also prepare a 1 to 20 dilution of CO-free blood.
For standards set up a series of nine test tubes containing 0.0, 0.2, 0.4 . . . 1.6
ml. of CO-blood dilution, each made up to a final volume of 2 ml. with the CO-
free blood dilution. To e!wh of these tubes add 40 mg. of a mixture of equal
parts of tannic and pyrogallic acids. Invert the tube 4 or 5 times to mix.
These standards give a series in which the proportion of saturation with CO
varies from 0 to 80 in steps of 10%. If air is.excluded by a paraffin seal on the
1964 POISONS
surface of the liquid, these standards will be, suffici~ntly permanent for a w~k
or two, but it is little trouble to prepare them fresh for an occasional sampJ.";;.
A sample of air to be examined is obtained in a 250-ml. bottle, using an aspi-
rator bulb and discharging the bulb at least 25 tirhes into the bottom of the
bottle. The bottle is closed with a rubber stopper ot with a well-vaselined glass
stopper if preferred. Such a sample will keep weIl.l When ready for the test,
add to the bottle 2 ml. of the 1-20 CO-free blood dilution, avoiding as much as
possible the escape of the air-gas mixture from the bottle. SIOWlY~Urn the
. bottle so that the blood solution flows over the surface of the inside of t e bottle, .
avoiding violent shaking. After 15 or 20 minutes' rotation, transfer t e blood
solution to a test tube of the same size as the standards, add 40 mg. of the
tannic-pyrogallic mixture, invert for mixing as with the standards, and, after
~tanding 15 minutes, compare with the st~ndards and determine the approxi-
mate saturation of the blood solution. This may be done with an accuracy of
5% by estimating the degree of color as equal to a standard tube or between
two such tubes. The following tables gives the approximate percentage of CO
in the air corresponding to the percentage of saturation of the blood when the
temperature at the time of making the test is not far removed from 20° C. and
the air is otherwise of fairly normal composition, say, containing at least 19%
of oxygen.
Saturation % CO %
o....................... .'................... " 0.00
10 .............................................. 0.007
20 ............ ,.' ........... ' .. , ................. 0.015
30 ........................ '..................... 0.028
40 ............................................. 0.045
50 ................................... : ......... 0.066
60 ........................... , .............. : .. 0.090
70 .. ' ........................................... 0.152
75 ............................ , ................ 0.200
Corrections for variations in temperature and composition may be found in
the original (loc. cit.).
DETERMINATION OF CO IN BLOOD
MARME'S REAGENT
MAYER'S REAGENT
o-(~, _....-/CHs
~Xa"CHa
HC"
I' \:\ CH
H2 C
I
..
I '\. 1 ..
~2 a~ 'Y,q_
: "01"\ 0
.aHa~'o-J'-
This splits up on steam distillation
into laevulinic al1.lehY<ie, laevuli- '
nic acid, and a superoxiae:
Th'e formula indicates that the rubber hydrocarbon is an 8 carbon ring com-
ppund with two double bonds and that there are two methyl groups attached
to the ring. The systematic name for the rubber hydrocarbon is " 1.5 di-
methyl-cyclo-octadiene 1.5." The term octadiene signifies an 8 C-ring with
two ethylene linkages. The 1.5 at the end refers to'the position of the bonds
in the ring whilst the 1.5 at the beginning of the expression refer;; similarly to
the position of the methyl groups. Other formulae for the rubber hydrocarbon
have been proposed, and the formula is still an open questi~m
1 Many investigators prefer the formulae (C.Hs)y; '(CoI!-~r2)" etc.
Chapter by L. E. Salas.
1966
RUBBER, GUTTA-PEReRA, AND BALATA 1967
filter paper. Charge the thimble with the. sample and the tared flask with
60 ml. pure acetone (freshly distilled over anhydrous KiC0 3, using the fraction
56°-57° C.). '
If muslin is used to enclose and interleave the rubb~r it is well to fold' over
the lower end of the roll and fasten it by a few stitches so that there is no
possibility of any of the snippets or detached particles working out betweerif
the folds. .
Distil on water-bath or electric hot-plate for 8 hours. Detach tp.e extrac-
tion flask, distil the solvent and dry the residue at 100° C. to .constant weight.
I--- - - 8 0 - - - - - J
',/1//Q/P1/;,ns/ons in mi//imete~s.
FIG. 285.
"
The flask should be inclined while drying in order to facilitate the removal of
the solvent.
The increase in weight of the tared flask = acetone extract.
The residue can generally be easily detached fr~m the muslin and can be
weighed (as a check) after drying in vacuo over H 2S0 4 •
Figure 285 shows the construction of Cottle's Extraction Apparatus (" Ur,,-
derwriters" design). The advantages of this extractor over the ordinary
Soxhlet are (1) that it is compact, (2) the reflux coIJ.denser is immediately above
the cup holding the sample, (3) the material being-extracted is continuously
RUBBER, GUTTA-PEReRA, AND BALATA 1971
in the vapor of the solvent at its boiling point and thus does not become chil!ed.
(A possible disadvantage is that the size of the sample is somewhat limited.)
Only glass or metal joints should be used in any extraction apparatus (i.e., no
...cork, rubber, or other extractable substance should. come into contact with the
solvent). •
It should be noted that both the extract and the acetone extracted rubber
are liable to oxidation-thefefore the drying should be carried out quickly at
lowest convenient temperature. Any prolonged heating should be avoided.
Weigh immediately.
Brown glass apparatus is sometimes recommended to eliminate the effect/
of light. . .
The percentage of tI Resin" is geI1eraIIy characteristic of the brand-
Para (1.3%), Congo (4-6%), Kassai (4.5%), Borneo (10%), Ceylon Planta-
tion (3.2%), Guayule (average 18%), Pontianak (80%). The character of the
" Resins" is important, as, technically, rubber containing hard resin is pre-
ferred. Part of the acetone extract is saponifiable and part contains unsat-
urated compounds. (Iodine value of rubber" resins "-Para 118, Borneo 31,
Guayule 94.) All rubbers except Para contain optically active" resins" and
this fact has been proposed as a means of detecting rubber other than Para in
a mixture. The optical activity is always dextra-rotatory and it has been
observed that the more optically active, the less. sapon~fiable is the resin.
To Determine the Acid Value of the Acetone Extract.-Take up the weighed
acetone extract with warm neutral ethyl alcohol. Titrate with N/10 alcoholic
KOH, using phenolphthalein as indicator. Calculate the number of milli-
grams required to neutralize 100 g. of the original rubber (=acid number).
Ash.-Heat 3 g. of the sample in a porcelain crucible of about 40 ml.
capacity. In an asbestos card, cut a hole of such a size that when the crucible
is placed in it, the base of the crucible will protrude 1 cm. below the level of
the asbestos. The crucible is placed on a tripod and heated gently at first
with a small Bunsen flame until volatile material ceases to be evolved. During
the initial stages of incineration, care must be exercised that the volatile
decomposition-products do not inflame (or losses may occur rendering the
results unreliable). Finally the temperature is raised to dull red.heat, and
ignition is continued until all the carbon 'is burnt off. The mineral matter
occurring in raw rubber varies' with the source and mode of preparation.
First quality Pale Crepe ranges from 0.10% to 0.5% ash. The color of the
ash should be noted. If white or pale yellow, and infusihle, silica, lime and
magnesia are possible cons'tituents. If highly ferruginous the ash has a brick
red color. Having now determined the moisture, "resin," ash, and washing
loss, an approximate figure for the rubber hydrocarbon may be calculated by
difference. "'
Determination of Copper in Crude Rubber
Ash a 20-g. sample over a Bunsen burn~r and fi~ally in a mufRe furnace.
Do not expose to a higher temperature than a dull red heat. Particles of
carbon enclosed by the ash need not be burned off as any portion of the copper
existing as chlorides may be volatilized at the same time. Add 5 ml. of con..'
..
centra ted HCI and heat the crucible until acid boils. Transfer by means of a
1972 RUBBER, GUTTA-PEReRA, AND BALATA
fine stream of water and rubber policeman to a smali beaker. Boil for a'few
minutes. , I
After the solution has cooled make it alkaline with NH 4 0H. Heaft on a
steam plate until the volume of the solution is not more than 15 mi. (a steam
plate is best for heating as there is often danger .of sp~ttering). By this time
all of the excess ammonia will' have been driven o~. Add a few drops of'
bromine water and heat to boiling. (1) Filter immediately through a Gooch
crucible into a graduated vessel (a graduated test tube which will slide inside
the usual suction filter flask serves the purpose nicely). Dilute the solution
to 30 mI.
Place 10 mi. of the solution, unless the copper is abnormally high (.01 %)
or low (.0005%), in a 50-mi. Nessler tube. If the copper concentration is
high (.01 %) use 5 mi. of solution, and if the copper concentration is low, use
20 mi. of solution. Add 22.5 mi. of 95% pure ethyl alcohol and 2.5 mi. of a
1% solution of sodium diethyldithiocarbamate in 95% pure ethyl alcohol.
Mix thoroughly.
For a standard (2) prepare a similar solution in a Nessler tube using 10 mi.
of water in place of the portion of the sample. Add a solution containing
0.00001 gm. copper per mi. from a pipette graduated in tenths ml. until. an
equivalent color is produced. , .
Calculation: •
The solution now contains all the nitrogen in the form of (NH 4 )2S04. F"it
to the flask a distillation tube containing a " splash buib." Measure into 1m
Erlenmeyer flask 20 m!. (N /10) . H 2S0 4 , or more, according to the amount of N
(NH 3 , protein) present. Allow the end of the distillatidn-tube to dip beneath
the surface of the standard acid. Detach the distillatidn-tube from the Kjel-
dahl flask and add 80 m!. NaO:a; (sp.gr. 1.35) cautiously down the side, keeping
the solutions from mixing until later.
Add a solution containing 2 g. sodium sulfide in 20 m!. water, and then
about 0.5 g. zinc in small pieces.
Now gently whirl the Kjeldahl flask, connect the tube and distil the am-
monia into the standard (N/10) . H 2S0 4 •
When 100 m!. Qf the distillate has passed over, disconnect the tube and
wash the end into the standard acid.
Titrate the excess of (N/10)·H 2S0 4 with (N/10)·NaOH using methyl
orange as indicator.
Calculate the nitrogen and multiply the value by 6.25 to obtain the weight
of protein in the rubber. (The average N content of protein is 16.5%. .
Wilfarth's modification of the Kjeldahl process is also recommended for the
total N in rubber. It consists in digesting 1 g. of rubber with 30 g. concen-
trated H 2S0 4, 7 g. K 2S0 4 and 1 g. anhydrous copper sulfate for four hours Itnd
proceeding as usual.
A blank Kjeldahl should be run on the reagents used.
Instead of distilling the ammonia evolved by the NaOH into N/1O·H 2S0 4
as described above, the gas may be absorbed. by cold distilled water and the
amount determined by Nessler's reagent (See page 2049). •
As stated previously, the protein mate~ial contaimid in rubber is capable of
being separated from the nitrogen-free material by a solvent method (see under
Organic Insoluble).
, Para rubber generally contains about 0.4% N, corresponding to a protein
content of about 2.5%. I
The amount of protein depends upon the sOl,nce tind mode of preparati(m'
of the rubber and may amount to 10% and over.
Determination of the Rubber HydrocarbOil.-As previously stated; f<1C:
technical purposes it is sufficient to subtract the sum of the percentages of
washing loss (moisture and dirt) ash, organic insoluble (protein) and acetone
extract from 100 to obtain the percentage of rubber hydrQcarbon. Such a
method is far from ideal but up to the present no entirely satisfactory direct
method of determining the rubber hydrocarbon has been evolved, and those
suggested are generally somewhat tedious. -
(a) Spence's Method.-Take about 1 g. of ' the acetone extracted rubber
(vacuum dried). Add 100 ml. cold benzene. Allow to swell, and make up to
200 m!. with benzene. Mix thoroughly and filter through a tube plugged with
glass-wool (or allow to settle). '
Take an aliquot part of the solution and evaporate to dryness in a current
of CO 2. Weigh.
(b) Precipitation Method.-Weigh out 3 g. of the material and swell it in
10 m!. of toluene. Make up to 100 ml. with, tolnene. Allow to settle (or
centrifuge) .
RUBBER, GUTTA-PERCRA, AND BALATA 1975
claimed as the only nitrosite method yielding sufficiently exact results. T~ke
0.2-0.5 g. of rubber (dried). j
Extract with acetone for 8 hours. Dissolve the residue in chloroform and
I
pass the nitrogen-oxide gas until the solution is green, Allow the solution to
stand overnight, and filter through a " Gooch" lined with asbestos. Evapo-
. rate the filtrate to dryness. Dissolve the residue in acetone and filter.
Take an aliquot part of the solution in a combustion boat and evaporate to·
dryness. The residue is ignited in a combustion tube and the products are
led through a train of absorption bulbs (eight in all).
The first three contain a mixture of concentrated H 2S0 4 .and K 2 Cr 207, the
fourth contains zinc dust, the fifth is packed with soda-lime, the sixth with
calcium chloride; the seventh bulb contains H 2S0 4 'and K 2Cr 207 while the
eighth holds a dilute solution of palladium chloride. .
The fifth, sixth, and seventh absorption bulbs are weighed before and after
the combustion. The increases in weight give the total CO 2 and H 20 formed,
from which the rubber content of the specimen is calculated .
. (For details see I. R. W., 57, 17 (1917).)
Determinations Required.-
1. Nature and amount of solvent.
2. Percentage of rubber.. '
3. Content of total solids.
The Solvent.-Take an adequate sample and separate the solvent by steam
distillation. Separate the water and the solvent by means of a separatory
funnel. Add a suitable d,ehydrating agent (i.e., fused calci_um chloride) and
allow the solvent to stand over this for a day. . ~
A fractionation test should be undertaken and the nature of. the solvent
investigated.
To Determine the Proportion of Solids to Solvent.-Take a representative
sample of 10 g. in a stoppered weighing bottle. Remove the stopper and place
the bottle and stopper in a clean porcelain basin and evaporate the solvent in
an electrically heated vacuum desiccator. The difference in weight = weight
of solvent.
Investigation of the residue. Cut it into snippets and extract with acetone.
The dried residue from acetone may be dissolved in acetic anhydride by Moraw-
ski's Method and examined for colophony by the.liebermann-Storch reaction
(evanescent violet coloration with concentrated H2S0 4).
RUBBER, GUTTA-PEReHA, AND BALATA 1977
Inorganic fillers are determined as follows: Weigh the sample, dilute with
petrol ether to such degree as. will permit the inorganic fillers to settle. Filter
or centrifuge. Wash the residue with the solvent. Dry and weigh.
The sample having been isolated, it is cut up into snippets with scissots.
Hard rubber (ebonite, vulcanite) is rasped with a coarselclean file and any iron
filings introduced are removed by a magnet. i
Unvulcanized doughs and reclaim may be rolled out into 'a thin sheet on
the laboratory mill. The prepared bulk sample shoJld be ,preserved in a
stoppered wide-mouth bottle preferably of brown glass. I
The scheme of analysis is that of Weber and Henriques, with modifications.' .
It is rarely necessary to go through the whole scheme. The appearance of the
sample will give valuable information. If the rubber is Ilgl,lt or red in color it is
obvious that carbon black, pitch, tar, bitumen, and Pb compounds are absent.
In this case, the chloroform-extraction can be omitted.
Moisture.-Take a 2 g. sample and dry it to constant weight in a vacuum-
desiccator containing concentrated H 2S0 4 • Loss of weight = Moisture. (If
the rubber contains no material likely to be melted out, the sample may be
dried in an air oven at 80° C. or in an electrically-heated vacuum-desiccator.)
Water-Soluble Constituents.-Take a 10 g. sample in a 300-mL flask and
add 100 ml. H 20. Heat to 80° C. and allow the specimen to soak for a few
hours. If present, the following substances will be found in the solution:
Starch (dextrin), gelatine (glue), some aqcelerators (aniline), and part of.
the inorganic fillers (lime, CaSO 4).
Test for alkali; starch by iodine; size (gelatine) by Tannin. Aniline may
be detected by the violet coloration with hypochlorites. Glycerin may be
detected by evaporating a portion of the solution to dryness and heating the
residue with KHS0 4 when the odor of acrolein is obtained.
For other constituents proceed as follows: Take a 2 g, sample and extract
with pure acetone, for 8 hours. .
,
A. Solution B. Residue. Extract with alcoholic KOH (IN)
I
/
.' C. Solution contains D. Residue contains
Sulfurized substitutes I Rubber Hydrocarbon
Chloro-sulfur substitutes Mineral fillers
Oxidized (brown) oils (Sulfur of vulcanization)
Polymerized oils (Chlorine of vulcanization)
....
(Sulfur,and chlorine in above) -. -. , ._
Note.-The mineral fillers are determined on a part of Residue D by ignition, and
the rubber-hydrocarbon is calculated by difference. The Residue D is also used to find
the sulfur and chlorine of vulcanization by method given on pages 1985-1986.
I
Full Scheme for Analysis of Vulcanized Rubber.-If the rubber is black it
may contain one or more of the following substances: Carbon black, pitch, tar,
bitumen, Pb compounds. For a systematic investigation, the following pro-
cedure may be followed: • (This scheme is based on the work of Weber, Henri-
ques and others.) .. ,..
ROBBER, GUTTA-PEReRA, AND BALATA 1979
Take a 5 g. sample and extract with Acetone for 8 hours. (Hard rubber requires
48 hours.)
,
3. Mineral 1. Pitch.
Oils. • 2. Bitumen. E. Extract F. Residue.-Weigh and extract with a
4. Waxes. 3. Tar. rubber solvent. (Petrol,
5. Fatty Oils. may
4. Sulfur in contain: xylene, etc.)
6. Resin Oils. the above. 1.
7. Free Sul- Sulfur
5. M.R. Sub.
fur. (part). 2.
8.·M.R. Chloro- G. Extract H. Residue.-Weigh and
(part). Sub. may extract with boiling
9. Cellulose- 3. Oxidized contain: water.
Esters. Oils. 1. Rubber (Wt. of F-Wt. of
10. N aphtha- 4. Sulfur hydro- H=G.)
lene, and Cl carbon.
aniline. in the 2. Sulfur
11. Accelera- above. and Cl I. Extract K. Residue
tors. in above. may may
12. Antioxi- contain: contain:
dants. 1. Starch. 1. Minerals.
13. Certain 2. Dextrine. 2. Free
organic 3. Gelatine. Carbon.
colors. 4. Albu- ..3. Vege-
. men . table
fibre.
4. Sulfur in
minerals.
(See page (See page (See page (See,page (See page (See page
1980) 1982) 1982) 1982) 1978) 1988)
Note.-The time required for the Acetone Extraction (and its completeness) depend
upon the apparatus and the speed of siphoning; also upon the rubber and its subdivision.
The siphon should function once in about three minutes.
alcoholic solution to 50 nil. Cool the flask and then place it in a freezing mix-
ture capable of producing a temperature of minus. 5° C. The solid paraffins
will separate out. Filter thrOl,lgh an iced funnel and wash with ice-cold alcohol.
Remove the funnel from the ice-pack and dissolve the residue through the
filter with hot cMoroform into a tared flask. Evaporate the solvent and dry to
constant weight (=solid paraffins).
The alcoholic solution remaining after freezing out the solid paraffins is
evaporated to dryness on a water-bath. Dissolve the residue in CC1 4 (15 m!.)
and transfer the solution to a separatory funnel. Add 15 m!. concentrated
H 2S0 4 to the funnel and shake carefully. Separate and repeat till the H 2S0 4
extracts no more colour.
Draw off the acid. To the CC1 4 solution add distilled water and then enough
ether to dilute the CC1 4 to make it float above the water. Shake and repeat the
washing until the" CCl 4-ether solution" is acid-free. Transfer the" CCl 4-ether
solution" to a tared flask and evaporate. Dry constant at 90° C. This'
gives the weight of the liquid paraffins, and mineral oils.
1he Saponifiable Matter in the Acetone Extract comprises:
Fatty acids.
Resin acids.
Added Resins (e.g. Colophony).
Take the aqueous portion from the separation of " Saponifiable" and" Un-
saponifiable." Acidify with HCl. Shake with ether. Wash the ether free
from acid. Filter the ether solution 'and wash with the solvent. Distil the
ether and dry the residue for one hour at 95° C. This gives the weight of the
fatty acids, obtained from the saponifiable oils.
To determine resin (colo phony) use Parry's Method:
Take part of the original acetone-extract or the saponifiable above. Dis-
solve in 20 ml. of 95% alcohol. Add dilute NaOH (made up with 1 of alcohol
+2 of H 2 0) until just pink to phenolphthalein. Heat for 10 minutes. Cool.
Transfer to a lOO-ml. graduated stoppered cylinder. Fill to the mark with
ether, and add 2 g.·AgNO a• Shake and allow to settle. Silver salts of abietic
acid and other organic acids form. The silver salts of the free fatty acids are
insoluble in ether, but those from colo phony (abietic acid) give silver salts
which are soluble. Therefore talfe 50 ml. of the clear solution in a separatory
funnel and shake with 20 ml. dilute HCI. Draw off the ether layer by means of
a pipette and shake twice with ether. Unite ethers and wash. Distil the
ether solution in a tared flask. Residue = Colophony. Dry at 105° C. and
weigh (X by 2).
The Detection and Determination of Accelerators is very difficult and gen-
erally impossible owing to the complex nature of the accelerators themselves,
the small amounts used, and to the great changes they undergo during vul-
canization.
Some disappear"altogether (i.e. isopropyl-xanthates).
Evidence of the use of organic accelerators may be inferred by the enhanced
mechanical properties, the sub-normal sulfur content, and low vulcanization-
coefficient. For further information see Twiss and Martin (Gummi Zeitung,
3S, 1151 (1921».
1982 RUBBER, GUTTA-PEReRA, AND BALATA
AllDw it to. stand until the fillers have settled. Decant and filter thrDugh a
.
GDDCh. Wash with hDt benzene, alcDhDI and ether. Dry the crucible at
90 0 C. to. CDnstant weight.
A centrifuge will save much time.
If an autDclave is available the sample may be heated with xylDI under
pressure.
DeterminatiDn of Sulfur
• The. sulfur in rubber may Dccur in variDus fDrms. 'The determinatiDns
usually required are tDtal sulfur, free sulfur, sulfur Df vulcanizatiDn, sulfur
cDmbined in the fillers (Drganic and inDrganic).
During the. analysis sulfur appears in the variDus extracts and residues.
1. Total Sulfur (J. RDthe's MethDd).-Take 1 g. Df the rubber under test
in a rDund-bDttDm Pyrex flask. Add 30 m!. cDncentrated HNO a anti 1 g. MgO.
Heat upDn a steam-bath until fully Dxidized. Heat mDre strDngly upDn a
sand-bath to. expel brDwn fumes. 5 (If carbDn shDuld remain repeat the Dxida-
tiDn with HNO a and heat again.) Add 5 m!. HNO a and dilute with 30 m!.
H 20, Filter and reserve the residue.
Filtrate.-EvapDrate to. dryness in a pDrcelain dish and bake the residue
for 2 hDurs at 135 0 C. MDisten with HNO a• Add H 2 0 and evapDrate dry.
Add 5 ml. cDncentrated HCl. Heat and dilute to. 100 ml. with H 2 0. Filter
and wash residue. BDil the filtrate and add 10% BaCl 2 sDlutiDn. AllDW to.
stand (lvernight. Filter Dff the BaS04; and wash with distilled water. Dry,
incinerate and weigh. (BaS04 XO.1373 = S.) .
Residue (abDve).-Fuse in.a crucible with abDut 5 g. Df a mixture Df
Na 2CO a and K 2CO a (equal parts). When clear, allDw to. CDDI, leach DUt with
hDt water and bDil the liquid. Filter.
Filtrate.-Acidify with HNO a• EvapDrate to. dryness, bake at 135 0 C. fDr
2 hDurs. Add cDncentrated HCI. Heat and dilute with H 2 0. Filter Dff any
silica Dr antimDny Dxide.
EvapDrate the filtrate to. dryness, add cDncentrated HCl and evapDrate
again. Add dilute HCI and bDil; filter. Add 10% BaCl 2 and calculate sulfur
as stated abDve.
If BaS04 is present, make a fusion and subtract the sulfur in the mineral
fillers from the total sulfur.
Sulfide Sulfur. Steven's Method (Chern. Analyst, 40, 275 (1915».-Arrange
a Kipp Apparatus for C02; and a 250-ml. flask with an inlet tube reaching
nearly to the bottom; a ground-in stopper should be used. The outlet tube
from the flask is connected to two absorption bulbs containing lead-acetate
solution.
Place in the flask 10 ml. concentrated HCI and then 20 ml. ether. Pass a
current of CO 2 through the apparatus until all the air is displaced. Lift the
stopper, and add 1 g. of the powdered sample. Again pass CO 2 (for 30 min-
utes), shaking occasionally. The H 2S evolved from the sulfides is absorbed by
the lead-acetate. Heat the generating flask to ensure' complete decomposition.
Determine the lead-sulfide in the bulbs and calculate to sulfur. If pure nitrogen
is available for sweeping out the apparatus, the procedure may be simplified
by using NaOH in the absorption bulbs in place of lead-acetate. In this case,
the sodium sulfide formed is oxidized to N a ZS04 by means of bromine. Deter-
mine BaS04 and calculate to S.
Sulfate-Sulfur.-Ignite a 1 g. rubber sample in a porcelain boat in a stream
of CO 2 • Boil the residue with Na 2CO a solution. Filter, acidify the filtrate
with HCI, and boil to expel S02 from the sulfites. Precipitate the sulfate as
usual with 10% BaCI 2. Calculate S.
Sulfite-Sulfur.-Proceed as described above for sulfate-sulfur except that
instead of boiling off the S02, it is oxidized by bromine and thus converted to
sulfate.
-Precipitate the total sulfate 'by 10% BaCl z solution and calculate S. Sub-
tract "sulfur as sulfate" from the "sulfur as sulfite+sulfate." Difference
= " sulfur as sulfite." Or t,he S02 boiled off may be absorbed in N aOR solution"
oxidized, and precipitated as BaS04.
The Sulfur of Vulvanization can now be calculated. Subtract the various
" sulfurs" determined from the total sulfur-the remainder = vulcanization
sulfur, i.e., the sulfur actually combined with the rubber-hydrocarbon.
If the rubber is a simple" rubber+sulfur " mixture, the calculation of the
sulfur of vulcanization involves the determination of the total sulfur by one
of the methods described and the determination of the free sulfur (sulfur in the
acetone extract). Subtract the latter from the former.
If the sample is a complicated mixture it is necessary to carry out the follow-
ing extractions in accordance with Weber's scheme of analysis. Take a 2 g.
sample and extract with acetone, chloroform, and alcoholic KOH. This leaves
a residue containing rubber and fillers. Divide this residue into two parts;
on one part find the Total Sulfur; on the other part carry out the solvent
separation of the fillers and determine the sulfur in the latter by a fusion method.
The difference between these values will give the sulfur combined with the
rubber-hydrocarbon.
The sulfur of vulcanization can also be obtained when the nitrosite method
for determining the rubber is used. The nitrosite contains all the sulfur that
was originally chemically combined with the rubber-hydrocarbon.
Take an aliquot portion of the acetone-solution of the nitrosite compound.
Remove the acetone by evaporation and determine the sulfur in the residue
1986. RUBBER, GUTT~-PERCHA, AND BALATA
I
(Carius' Method)
/
Place 0.3 g. of original rubber in a Carius Tube and add 5 m!. fuming HN0 3
and 1 g. of solid AgN0 3 • Seal and heat in the!furnace. Open the tube and
empty contents into a beaker.
The residue consists of a mixture of AgCI and insoluble mineral matter.
(Most rubber compounds contain mineral matter insoluble in HNO a.) Heat
the residue with dilute H 2S0 4 in a Pt crucible and add a lew-pieces of pure zinc.
Allow the reaction to continue until the zinc has dissolved. SWoccasionally.
Filter and wash with hot water. The silver has been reduced by the Zn and
the chlorine is in the filtrate. Determine chlorine .by the usual methods.
(See page 265.) , .
Qualitative Test for Glue (Epstein & Lange's ,Method).-Take 3 g. of the
rubber sample and heat with pure cresol for 1611o~rs at 120 0 C. The rubber
dissolves. Cool and dilute with petroleum-ether. The glue ang other in-
solubles settle out. Allow to stand overnight. Or the liquid may be centri-
fuged. Filter on a prepared "Gooch," wash with warm petroleum-ether,
and then with hot benzene. Dry the crucible and contents at 90° C. Remove'
the contents of the crucible to a beaker and boil.lYltl}i·'H20. Filter and add a
solution of tannic acid. . .
RUBBER, GUTTA-PEROHA, AND BALATA 1987
Determination of Cellulose
Cellulose may be present in two forms.
1. As.the cotton fabric upon which the rubber was spread.
2. As a filler ingredient, in fiber form or cotton watse, etc.
To determine the relation of the tissue to the rubber material and to ascer-
• tain the structure of the fabric as well as the weight of rubber per unit area, a
process is used similar to the solution method for Total Inorganic Fillers- (or
loading).
Take a convenient sample. If it has a thick coating of rubber, the latter
should be removed as far as possible without damaging the fabric; reserve the
rubber. Cut up the remainder of the sample into small pieces and heat in a
flask (use reflux) with petroleum or in an autoclave with benzene or xylene.
Continue the digestion until the rubber dissolves. Filter off the residue, wash
with benzene and afterwards with ether. Dry and weigh.
The material still contains the loading between the fibers.
Heat with NaOH solution. Wash with H 2 0. Heat with acetic acid and
wash with water. Dry to constant weight at 100° C. Weigh.
B'urn off the cellulose material and weigh again.
Loss of weight = cellulose (an"hydrous).
As cellulose matter in the air-dry condition generally contains 7% moisture
add 8% to the result to obtain the weight of the cellulose material as originally
used.
This method of course fails if there are other combustible materials and
volatile inorganic fillers.
Epstein & Moore's Method 6 (B. of S. Tech. Paper No. 154).-Take 0.5 g.
6 See also Standard Methods, page 2009.
1988 RUBBER, GUTTA-PEReRA, AND BALATA I
•
of the sample in a flask and add 25 mi. freshly distilled .cresol; heat for four
hours at 165 0 C. under reflux. Cool and add 200 mi. petrol ether, shake con-
tinuously for a time and then allow to settle. Decant the Clear portion through
a tared "Gooch." Wash the residue in the flask with hbt benzene and then
with acetone. Add hot 10% HCl to the residue in the flask and transfer it
to the" Gooch." Wash with 10% HCI. Wash thoroughly with boiling H 20
(until free from chloride) and then with acetone until the filtrate comes through
clear and colorless. Transfer to a tared weighing bottle. Dry at 105 0 C.
to constant weight (=A). (This gives the weight of the asbestos pad+inor-
ganic fillers and cellulose.)
(a) Transfer the contents of the weighing bottle to a flask and add 15' mi.
acetic anhydride and 0.5 mi. concentrated H 2S0 4 • Digest in the steam bath
for 1 hour. Cool, and add 25 mi. 90% acetic acid. Filter through a tared
"Gooch." Wash with acetic acid and afterwards with acetone. Dry the
crucible at 1050 C. Cool and weigh (=B). Loss of weight (A -B) = cellulose.-
(b) Or the following procedure.
Transfer the contents of the weighing bottle (at A) to a flask a~ add 25 mi.
of " Cu-Oxam" solution. Allow the mixture to stand for 12 hours, shaking
at intervals. Filter through a tared ,I Gooch." Wash first with" Cu-Oxam,"
and then with dilute ammonia until the filtrate runs colorless. Dry the ~rucible
to com;tant weight (=B). Net weight of the residue A -net weight of the
residue B = C (wt. of cellulose dissolved by tlie " Cu-Oxam.")
The weight of the cellulose may be confirmed by precipitating the cellulose
from the blue filtrate by adding sodium chloride and then dilute sulfuric acid.
The precipitated cellulose is filtered, washed with hot H 20 until free from
chlorides (test washings with AgNO a). Dry the cellulose at 105 0 C. and weigh.,
product. -..__
All the inorganic fillers are added in finely powdered form. _The most
commonly used are: Zinc-oxide, litharge, lithopone, pumice powder, chalk,
lime, clay, barium sulfate, magnesia, basic mMnesium carbonate, pentasulfide
of antimony.
The following are less common: Mercury-sulfide, arsenicals, red lead, white
lead (basic lead-carbonate), lead-sulfide, ultramarine, lead-sulfate, zinc dust,
ferric-oxide, aluminium-oxide, kaolin, calcium-sulfate, powdered glass, silica,
asbestos, talc. -
Before a systematic examination of the inorganic fillers can be made the
organic material must be carefully removed. The il!9i.ganic constituents may
then be determined in accordance with systematic' inorganic analysis but the
RUBBER, GUTTA-PERellA, AND BALATA '1989
following will give the general procedure in a rubber laboratory, For the
purpose of separating the fillers, Rothe's Method is recommended, as this
obviates the loss by volatilization of certain ingredients if present (As, Sb, Hg),
(A rough idea of the amount of ash may be obtained by simple ignition in ~
tared porcelain crucible, and this should be carried out as a preliminary.) .
For Rothe's Method t:he amount of rubber taken should be enough to
produce 1 g. of ash. The weighed rubber is placed in a round-bottom Pyrex
flas¥: of about 300 mi. capacity, Add (per gram of rubber) 15 mi. concentrated
HN0 3 and 2.ml. concentrated H 2S0 4 , Heat on a sand bath for 1 hour and
then more strongly to fumes of S03. Finally heat by naked flame to boil the
H 2S0 4. Cool, and add more HN0 3 (10 mI.). . -
Heat until the liquid becomes colorless. If necessary add more HNO a•
Cool, and pour the strong acid solution into 50 ml. H 20. Boil the solution,.and
ru~ .
The residue (A) may contain PbS0 4, CaS04, BaS04, silica, silicates, and
(if Sb is present) antimony oxide and Pb antimoniate, stannic oxide. Pb
may bl:l removed by boiling the residue with several portions of ammonium
acetate (+ NH 40HI solution or the hot ammonium acetate may be passed
through the filter containing the residue. This leaching shou1d be continued
untilltle " washings" fail to contain Pb (test with K 2Cr04 or H 2S).
Wash the residue and reserve it for further treatment (see under Ba).
The "ammonillm acetate-Pb solution" may contain a little calcium. The
calcium in the filtrate from Pb is evaporated to dryness in a porcelain dish.
The ammonium salts are ignited, the residue is dissolved in dilute HCI and
Ca is determined (as on page 210) ..
Barium.-Take the residue after the removal of the PbS0 4 by ammonium-
acetate (above). Ignite with a mixture of sulfur and Na 2C0 3 (equal parts).
Apply cine bunsen to the bottom and another to the top of the crucible and
continue the Iheating until sulfur ceases to burn beneath the cover. Cool.
Leach out with H 20. Filter. The residue may contain BaC0 3 (silica and
silicates) and PbS. The filtrate may contain antimony and tin (if present)-
reserve (B). Take the residue, add hot HCI to dissolve the Ba, Ca, and Pb,
leaving (as insoluble) the silica and silicates. [This insoluble should b!l ignited
and fused with a mixture of K 2 COa+Na 2 C03. Leach with water and .filter.
(If any residue occurs here it .should be tested for bases.) The filtrate from
the alkali-fusion may be discarded as silica is best determined in a separate
portion of the original sample.]
The acid filtrate containing the Ba and Ca is now tested for Pb and, if
foul).d, H 2S is passed through the solution to completely remove it. The
Ba and the Ca are determined in the filtrate by usual methods (see page 122).
The filtrate from residue A may contain Hg, As, Sb, Sn, .Zn, Cr•. Fe, Ca,
and alkali metals, etc.
Hg.-Pass H 2S to precipitate Sb, As, Sn, Hg. Filter and transfer with
H 2S to a beaker. (Reserve, filtrate D.) Boil with dilute NaOH-the Sb, Sn,
As dissolve-and filter the i~soluble HgS: Wash the HgS and dissolve it in
'aqua regia. Boil off the chlorine, dilute, and pass H 2S. Determine the Hg
as HgS in usual manner (see page 577). '.
1990 RUBBER, GUTTA-PEReRA, AND BALATA
As.-Acidify the alkaline. solution containing As" Sb, Sn and pass H 2S.
Filter the precipitate and wash with H 2S water.. Transfer the precipitate to
a beaker and add warm concentrated HCI. The Sb ~nd Sn dissolve. Thor-
oughly wash the As sulfide. Dissolve it in aqua regia and determine the As as
magnesium pyro-arsenate (see page 96). 'I r
Sb.-The solution from the arsenic is now combined with the filtrate B
(above). Add NaOH and bromine (drop by drop) t~ oxidize .all the sulfur
(present as N a2S), Acidify and boil the solution to remove excess of bromine.
While still boiling addferrum-redactum to precipitate the Sb (Sn is not reduced).
Filter through a small filter besprinkled with the iron. Wash with dilute HCI.
Reserve filtrate (C) for Sn. Transfer the Sb to a beaker and dissolve it in HCI
(hot), adding a little bromine. Boil off' the excess bromine. Add KI and
H 2S0 4 to reduce the SbCl 5 to SbCb. Precipitate the Sb as Sb 2S3 and determine
it electrolytically (see page 73). If the amount of Sb is small, determine. it
gravimetrically as Sb 20 4 (see page 72) or colorimetrically by the KMn04
method:
Sn ...,-Take the filtrate C from the" antimony-tin separation." Pass H 2S
gas, filter off the tin-sulfide and determine as Sn02 (see page 959).
Zn.-Take the filtrate D reserved from the precipitation of Hg, Sb, Sn,
and As. Boil off the H 2S. Make alkaline'with NH 40H. Add formic acid.
Pass H 2S, and allow the ZnS to settle out. Filter, and wash with H 2S water
containing a little formic acid. Ignite and determine as ZnO, or by.any other
standard method described under Zinc. (See page 1058.) (Reserve filtrate'
for Fe, AI, Cr. etc.) .
Fe, AI, Cr.-Boil off H 2S. Oxidize and precipitate Fe, AI, Cr. Separate by
standard methods (see page 467).
Ca.-Take the ammoniacal filtrate from Group III (a). Evaporate, acidify
with acetic acid, boil and precipitate the calcium as oxalate. Determine as
CaO (see page 210).
Mg and Alkali Metals.-Take the filtrate from ~a, acidify with H 2S0 4.
Ignite to decompose the NH4 salts. Dissolve, in 1% H 2S0 4. Filter into a
platinum crucible. Ignite to drive off excess of H 2S0 4. The residue may con-
tain MgS0 4, K 2S0 4, Na 2S04.
Weigh" Total residue." Dissolve'in hot water and precipitate the Mg with
ammonium phosphate. Determine Mg as pyrophosphate (see' page 532).
Recalculate the Mg to sulfate and subtract from the" Total· rellidue." Calcu-
late the difference to Na 20. -.....
Silica and Silicates.-(Powdered glass, pum}ce, kaolin, infusorial earth,
etc.) Silica is most conveniently determined alone on the original material. .
Take 1 g. and fuse in a Pt crucible with about 6 g. of a mixture of K 2C0 3 _
and Na 2C0 3 (equal parts). (Before using a platinum crucib.le for the fusion
of rubber compounds it is always advisable to test for Pb. If lead is. present
use a porcelain crucible.)
Cool the melt. Leach with water. Boil and filter. Wash. Acidify
with HCI. Evaporate the filtrate to dryness. ,Bake at 13$)0 C. Add 5 mi.
concentrated HCI. Evaporate dry, ap.d again bake at 135 0 C. for one hour.
Add 5 mi. concentrated HOI and dilute with H~(y' ~F'ilte,r ,and wash.
RUBBER, GUTTA-PEReRA, AND BALATA 1991
: Ignite in a platinum crucible and weigh as Si0 2 (test its purity as on page
801>.
Determine the sulfur in the residue from the chloroform, and subtract the
sulfur in the mineral fillers. The difference gives the sulfur combined with the
rubber itself (i.e. sulfur of vulcanization). The inorganic constituents are .
determined in the ash in a manner similar to that descriHed for soft vulcanized
rubber. ".
The amount of rubber is found by difference.
The resins of gutta percha and ba~ata are a mixture of white and yellow
resins. The white resins (Albanes) are soluble iri hot alcohol (not in cold).
They are crystalline and have a composition approximating to C 17H 260. The'
yellow resins (Fluavils) are amorphous and are solublelin cold alcohol as well
as hot.
This difference in solubility affords a method of separation if required.
Gutta (and Balata) Hydrocarbon.-(Rapid Method.) Boil 1 g. of the sam-
ple with 20 m!. redistilled toluene. Heat to 100° C. on a w.ater-bath until
dissolved. .
Pour the solution into 50 ml. alcohol (95%). Allow the contents of the
vessel to stand for a few hours. Warm and agitate to clear the liquiq. Remove,
redissolve in toluene and reprecipitate by alcohol. Dry the hydrocarbon in
vacuo and weigh in a closed weighing bottle. The hydrocarbon may also be
calculated by difference as for rubber-hydrocarbon.
Dirt.-This is generally under 0.5% for washed and strained material af.ter
the factory cleaning process. The dirt consists of both organic and inorganic
debris. The inorganic matter may be determined by incinerating a 1 g. sample
of the original material.
The" organic" debris consists of fine particles of' bark. If this determina-
tion is required a solvent method must be employed and when the residue is
obtained it should be weighed (organic and inorganic dirt) and ignited to
obtain mineral (weigh). Subtract the latter from the former weight and thus
determine the organic debris. The" dirt" should be examined for ZnO,
litho pone,' antimony sulfide and carbon.
Obach gives the following range as typical of Gutta Percha.
Water .......... " . . . . . . . . . . .. .. . . . . 1-1.5%
Dirt ................................ 3-5%
~~~:ocarbon} .................... " 30-83 %
Alban. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-44%
Fluavil ........................... " /3-21 %
Ash, max........................... ./ 0.5%
After washing:
Balata Hydrocarbon. . . . . . . . . . . . . . . . . . . .. 52%
~:t~:.·" . . . . . . . ::::::::::::::::::::::::: 4r~
Separation of Balata and Rubber, or Gutta Percha and Rubber.-The gutta
(and balata) hydrocarbon is insoluble in petroleum ether below 0° C., whereas
the rubber hydrocarbon and resins remain soluble. This ,fact· (which forms
the basis of the process for producing leaf gutta. at.!rjipetir, Malay) is used as
the means of separation required. -
I) RUBBER, GUTTA-PEReRA, AND BALATA 1995
The rubber containing balata and (or) gutta percha is dissolvedjn warm
petroleum-ether. The solution is then chilled to about minus 160 C. The
balata (gutta.percha) hydrocarbon is precipitated as a white flocculent solid.
Charleson's Method for separating gutta perch a (balata) hydrocarbon from
rubber (~s in compounds for golf ball covers). Mix the sample on the mixing
rolls and then make a very thin sheet. Of this take about 3 g. and dissolve it
in a flask containing 200 m!. of petroleum ether. Warm on a steam bath with
frequent agitation. When the whole is in " solution" allow the flask to cool
to room temperature and then place it in a freezing mixture capable of producing
a temperature of -12 0 C.
The balata (or gutta) hydrocarbon is precipitated, while the rubber and
resins remain in solution.
Filter the contents of the flask through a funnel surrounded by a freezing
mixture and wash with ice-cold petroleum ether. Collect the precipitate and
dry it in an electrically heated vacuum desiccator. Weigh the dried residue.
PHYSICAL TESTS
valuable information regarding the quality of the rubber present. For the
best grades of Hevea rubber, this corrected extract s~lOuld not exceed 5% of
the rubber present. A higher extract may indicate the presence of inferior or
reclaimed rubbers. \
Chloroform Extract.-."The ,chloroform extraction removes a portion of the
bituminous substances and serves as an indication of their presence.
Alcoholic-Alkali Extract•....:....Tlie purp~se of the alcoholic-alkali extraction is
to detect the presence of rubber substitutes.
Free Sulfur.-The free sulfur is defined as that which is removed during
the acetone extraction.
Total Sulfur.-This represents the sulfur that occurs in the compound either
free or chemically combined.
Ash.-The ash is the residue left after ignition, and consists 'principally
of the nonvolatile mineral fillers, together with their reaction products with
sulfur. '
Sulfur in Ash.-The sulfur in ash consists of the sulfur from the mineral
fillers, and also part of the sulfur that was with the rubber, but which during
ignition enters into combination with one or another of the mineral fillers.
Special Determinatioris.-SpeciaI methods are given for the analysis of
compo,upds that contaip. glue, carbon, antimony, and waxy hydrocarbons.
Rubber.-Up to the present time no simple method has been devised for
the direct determination of the amount of rubber presl)nt in a vulcanized
compound. Therefore, an indirect metl(od is proposed which will give satis-
factory results in all cases known to-day, except where there are found to be
'present decomposable fillers, such as carbonates, cellulose, and high percent-
'ages of mineral rubbers. As practically all insulating compounds contain
some of these ingredients, the said compounds when specified shall be analyzed
in accordance with Joint Rubber Insulation Committee's method, as given in
Paragraph 34. ,
When carbonates, talc, and asbestine are present, more accurate results
are obtained by the use of the Joint Rubber Insulatiqn Committee's methods
given in Paragraph 34. I
Reagents
7. The acetone shall be chemically pure and shall be freshly redistilled
over anhydrous sodium carbonate, using the 56° to 57° C. fraction.
8. The alcoholic-alkali' solution shall be of normal strength.' It shall be
made by dissolving the required amount of alkali (either potass[um or sodium
hydroxide) in the smallest quantity of hot, distilled water and adding this to
specially purified alcohol, which shall be prepared as follows:
Dissolve 1.5 g. of silver nitrate in 3 ml. of water, and add it to 1000 ml. of
alcohol. Dissolve,3 g. of alkali in the smallest amount of hot water, cool,
add it to the alcoholic-silver' nitrate solution, and shake thoroughly. Allow
the solution to stand for at least 24 'J:!rs., filter, and distil. (Alcoliol denatured
with 10% by volume of methanol may be used in place of ethyl alcohol.),
9. The nitric acid-bromine reagent shall be prepared by adding a c'onsider::'
able excess of bromine to concentrated nitric acid and shaking thoroughly'. It
can be used immediately.
10. The zinc oxide-nitric acid solution is made by adding 200 g. of zinc
oxide to 1000 ml. of concentrated nitric acid.
11. Barium chlo~ide solution shall be made by dissolving 100 g. of crystal-
lized barium chloride in 1 liter of'distilled water and adding 2 to 3 drops of
concentrated hydrochloric acid. If there is any insoluble matter or cloudiness,
the solution shall be heated over night on the steam bath and filtered. Care
should be taken not to add more acid than th~ amount specified. .
12, Standard 0.1 N potassium permanganate solution shall be made as
follows:
Ditlsolve approximately 3.1 g. of potassium permanganate' in 1 liter 6f
water, and when It is' dissolved filter through an ignited asbestos pad. Weigh
out 0.25 g. of pure metallic antimony and transfer to a 600-mi.' Erlenmeyer
flask. Add 12 to 15 m!. of concentrated sulfuric acid, 10 to 12 g. of potassium
sulfate, and heat until all the antimony is dissolved. Dilute to 250 ml. with
water, add 20 mL of concentrated hydrochloric acid, cool to 10°' to"15° C.,
and titrate with permanganate solution until a faint pink color is obtained.
I f 0 1 N KM '0 _ Wt. of metallic antimony':"
I, m.o. n 4 - MI f t
f . '
-wt.o antimony •
• 0 permangana e
17. Acetone Extract.-The extracting apparatus used here and for other
extractions shall be of the type similar to that shown on page 1970. (See also
Industrial and Engineering Chemistry, 9, 314 (1917).) The flask shall be
heated so that the period of filling an empty siphon cup with acetone and com-
pletely emptying it will be between 2.5 and' 3.5 min.
Place 2 g. of rubber in a thimble made by folding a filter paper so that it
will fit in the extraction cup, which is suspended in a weighed extraction flask.
Extract the sample continuously for 8 hrs., unless the solution.jn ;the thimble
is still colored at the end of that time, when the extraction shall proceed for
a further period of 4 hrs. or longer. For hard rubber the extraction period
shall be a minimum of 72 hrs. Carefully note ail characteristics of the acetone
extract, both when hot and cold. Distil off the acetone on the stearit bath
at as low a temperature as possible. Loss of extract by bumping can be
avoided by means of a gentle current of air or by slightly inclining the flask.
Care must be taken to avoid allowing the flasks to stand on the steam
bath after the solvent has been removed, because/appreciable quant~ties of
free sulfur may be'lost by so doing. Dry the extraction flask and contents
in an air bath for 1 hr. at 70° C., cool, and weigh. Call the residue" Acetone
Extract, Uncorrected."
. Wt. of extract
or
Percentage of acetone extract, uncorrected = Wt. sampJe X 100.
20. Free Sulfur.-Add to the flask containing th,e aC,etone extract, un-
corrected (Paragraph 17). 50 to 60 ml. of distilled water and~ 2 to 3 ml. of bro-
mine, and cover with a watch glass. (If the acetone extract indicated a large
amount of free sulfur, the amount of bromine used may he increased.) Allow
the flask to stand 0.5 hr: on the side of the stea.m bath, then heat cautiously over
the direct steam bath until the solution is practically colorless, filter into a
25Q-ml. beaker, and dilute to, about 175 ml. with distilled water. Precipitate
with barium chloride and determille in the usual manner, taking the proper
precautibns to prevent the reduction of the barium sulfate during the ignition.
Wt. of BaS04XO.1373
Percentage oUree sulfur
o
Wt.0f sampIe X 100.
destroyed. !tis necessary to hold the edge of the crucible with tongs. Toward
_the las.t .half of the operation the flame should be i~(}reased som,e:what" but it
.is _.neyer nec_essary to heat the crucible to redness. With care a cr!1cible can
be used for,at least ten or twelve fusions. \
_Alter a fusion, allow the crucible to cool, place it! in a 400-ml. beaker, add
.suffi<liell;t distilled water to cover the crucible (about 125 ml. are required),
and.digest,on the steam bath for 2 hrs., with occasiotial stirring. If the filtra-
-tion.cannot he made in the same day, do not. add the water, but allow the fusion
.to stand over night'. -' " .
If
j
.
IjI
/
j
-';'
I
'11'
0 I
I
, /'
-/
, j
-
" .
-
5
4 V ,
/ ,
"
2 F
V
/_ -
0 10 200 JOO 400 ,500 ~ W
FIG. 286.
, . - Wt. of BaS04XO.1373
Percentage of total sulfur 4 = W. f 1.' X 100.
-to_ D --samp e .
4 To correct the sulfur for occlusion use Graph -1, lirie marked 2.' .
/
23. Ash.-Wrap a I-g. sample in a filter paper, extract with acetone for 4
~rs., and transfer to a weighed, approximately 50-m!' porcelain crucible. Ash
the sample in a muffle furnace by heating at the following rate:
Time, min ........ 0 5 10 15 70 75 80 85 145
Temp., 0 C........ 0 100 200 300 300 400 500 550 550
Remove crucible from the furnace, cool, and weigh.
If no furnace is available, distil off the rubber over a very small flame, not
• allowing it to catch fire, and ignite gently until burned clean, cool, and weigh .
•
Wt. of ash
Percentage of ash = Wt f I X 100.
. 0 samp e
& To correct the sulfur for occlusion use Graph 1, line marked 1 (figure 286).
2006 RUBBER COMPOUNDS
24. Sulfur in Ash.-Add 3 ml. of nitric acid-bromine mixture to· the ash
(Paragraph 23), cover with a watch glass and heat-for:1 hr., r.emove the cover,'
rinse it with a little distilled water, and evaporate to dryness. Complete the
determination of sulfur as described under Paragraph 21.
\
. Wt. of BaS04XO.1373
Percentage of sulfur III ash 5= W f 'I X 100.
t.o samp e
Save the insoluble residue after filtering the solution of the fusion mixture
in water for testing according to Paragraph 25.
25. Barium SuIfate.-The barium sulfate is calculated from the barium
in the ash, which is determined as follows: Filter off the insoluble matter after
the fusion and extraction in Paragraph 24, wash back into the original beaker
with hot water, dissolve the residue in the beaker and any traces on the filter
paper with hydrochloric acid, and heat the solution on the steam bath. Filter
through the same filter as before and wash thoroughly with hot water. Nearly
neutralize the solution with ammonium hydroxide, leaving it slightly acid.
Saturate the solution with hydrogen sulfide in the cold, and when the lead
sulfide has settled, filter into a 400-ml. beaker and wash thoroughly. The
total volume should be not over 200 ml. Precipitate with 10% sulfuric acid
and determine the barium in the usual mariner. Barium sulfate determined
above is assumed to have been added as such. Obviously, if barium carbonate
is used, it must be determined in order that an undue correction shall not be
made.
. Wt. of BaS04XO.1373
Percentage of sulfur as banum sulfate = W f I X 100.
, t.o samp e
26. Total Antimony.-:--When a qualitative test indicates that antimony is
present, place a 0.5-g. sample, and transfer in a Kjeldahl flask, add 25 ml. of
concentrated sulfuric acid and 10 to 12 g. of potassium sulfate, place a funnel
in the neck of the flask, and heat until the solution gecomes colorless. Coal.
and wash the funnel, dilute the solution to 100 ml. j\vith water, and transfer
to a 400-ml. beaker, dilute to 250 ml. with hot water, and precipitate the anti-
mony with hydrogen sulfide. Filter and transfer the precipitate to a Kjeldahl
flask, add 15 m!. of concentrated sulfuric acid, 10 to 12 g. of potassium sul-
fate, and heat as described above until the solution is colorless. Wash the
funnel, dilute the solution to 100 ml. with water, add 1 to 2..g. of sodium sulfite,
and boil until all the sulfur dioxide is driven out. This is shown when nO"
blue colqr is obtained with starch iodate paper. Add 20 ml. of_concentrated
hydrochloric acid, dilute to 250 to 275 ml. with water, cool to 10° to 15° C.,
and titrate with standard 0.1 N permanganate solution until a faint pink
color is obtained. If iron is found to be absent, it is not necessary to precipi;.
tate the antimony with hydrogen sulfide, and the seMnd heating in. a Kjeldahl
flask can be eliminated.
· 8b (Paragraph 12) XmI. of permanganate 100
•
P ercen t age 0 f antImony = W fIX .
t.o samp e
I
27. Antimony in the Ash.-This is determine<!_.ib.::the ash of a I-g. sample.
Transfer the ash to a 600-ml. Erlenmeyer flask, aad 12 to 15 ml. of concentrated
RUBBER COMPOUNDS 2007
until the first. vigorous frothing ceases, then raise the heat gradually until
the liquid boils. Continue the boiling until the solution becomes clear. Allow
the flask to cool, dilute carefully with 150 ml. of water, and allow to cool.
Add 100 ml. of 50% sodium hydroxide solution, pour~ng it carefully down the
side of the flask so that it does not mix immediately with the acid solution.
Add about 1 g. of granulated zinc to prevent bumping and a piece of paraffin the
size of a pea to diminish frothing. Connect the flask quickly with a condenser,
the delivery tube of which dips into a 500-ml. Erlenmeyer flask, containing
50 ml. of 0.1 N sulfuric acid diluted to about 100 ml. Carefully swirl the
flask to mix the contents and start to heat gently, increase the flame as the
danger of foaming over diminishes, and, finally, boil briskly until about one-
half of the liquid has passed over into the receiver. Add methyl red solution
and titrate the excess acid by means of 0.1 N sodium hydroxide. solution.
A blank determination should be made.
Percentage of nitrogen as glue
_100 (ml. H 2S0 4 Xnormality-ml. NaOHXnormality) (0.014) (6.5)
- Wt. of sample
= (ml. H 2S0 4 X normality - ml. N aO H X normality ) X 4.55. 6
10% hydrochloric acid, and trarisfer the entire contents to the Gooch crucible
with the aid of a "policeman." Continue to treat with hot 10% hydrochloric
acid until the pad has been washed at least ten times. Wash the pad free from
chlorides with boiling water, and run small 'portions of acetone through it until
the filtrate is colorless. Treat with a mixture of equ~l parts of acetone and
carbon disulfide in the same manner. Wash with alc,ohol and dry for 1 hr.
0
and 3,0 min. at 105 C. Remove the pad from the crucible with help of {L pair
of sharp-pointed tweezers, using the underneath portion of the pad as a swa,b
to clean the sides of the crucible, and place all of this material in a tared weigh-
ing bottle. Repl&ce in the drying oven for about 10 min., cool, and weigh.
Weight of weighing bottle, pad, insoluble fillers, and, cellulose
-weight of weighing bottle,
1
= weight of pad, insoluble fillers, and c llulose.
Transfer the contents of the weighing bottle to a 50-ml. beaker and pour
over it 15 m!. of acetic anhyaride and 0.5 m!. of concentrated sulfuric acid,
and allow to digest for at least 1 hr. on the steam bath. After the mixture
has cooled thoroughly, dilute with 25 ml. of 90% acetic acid and filter through a
weighed Gooch crucible. To guard against traces of the material being carried
through, this filtration, as well as the ones to fbllow, must be very slow and only
gentle suction can be used. Wash with hot 90% acetic acid until the filtrate
comes through absolutely colorless, and/then wash about four times more.
Wash with acetone about five times. After having taken care that all of the
material has been washed out of the beaker in which the acetylation took place,
.remove the crucible from the funnel, cleau the outside thoroughly and dry for
2 hrs. at 1500 C. Cool and weigh. Original weight of crucible+weight of pad,
fillers and cellulose - weight of crucible after acetylation = cellulose.
Cellulose
Percentage of cellulose = Wt f I X 100.
• 0 samp e
I
eners is not an exact matter. On the other hand, an approximation is all that
is usually required.
Plantation rubbers yield acetone extracts usually between 2.0% and 3.5%,
while various grades of wild rubbers give extracts up to 10.0%, or even higher
for resinous grades. These extracts are usually of a soft conflistency, and until
the use of added softeners became common, ~ild rubbers were largely used for
softening purposes ..
Reclaimed rubber is variable in composition, depending on the source, but
that made from whole tires usually falls within the following limits:
Specific Gravity .............................. . 1.16 - 1.24
Moisture, max................................ . 1.0%
Alkalinity, max., 4 hr. ext...................... . 0.10%
Acetone Ext .................................. . 7.0% -12.0%
Chloroform Ext ..........................•..... 10.0% .-18.0%
Alcoholic KOH Ext ........................... . 1.00/, - 2.0%
Total S ............................ : ......... . 3.0% - 4.0%
Free Carbon ................................. . 3.5% - 7:5%
Ash ......................................... . 18.0% -35.0%
Added softeners are partially extracted from a cured stock by acetone, and
subsequent extraction with chloroform removes still more, but there is always
an appreciable amount of the softener which is not extracted, depending on the
type of softener, type of compound, cure, etc. Mineral rubbers for instance,
vary greatly, some coloring the acetone extract dark brown, and others not at
all, some carrying a large amount of sulfur and consequently combining with
little additional, while others tend to "hog" sulfur during the cure. Very little
analytical work is available to show the extraction under definite conditions.
The following is an example of the extraction of 3.0% of softener from a well
cured" pure gum" compound carrying 3.5% sulfur.
The latter may be a donsiderable percentage of the total sulfur, as the follow-
ing table of the per cent usually found in some of the pommon ingredients will
show.
Whole Tire Reclaim ......................... : 1. .... .
Reclaimed Friction ................................. .
3.5£)1,
1.0
Mineral Rubber .............................. : .1 ..... '. 1.0- 7.5
Hardwood Pitch ....... : .............. , ... , ......... . nil
Lithopone .......................................... . 9.5
Golden Antimony .................................. . 20.Q-40.0
Ultramarine Blue ................................... . 10.0
Mercaptobenzothiazole .............................. . 35.0
Thiocarbanilid ................................. : .... . 16.0
Diphenylguanidine .................................. . . nil
The only o11.e of these affecting the free sulfur determination is golden
antimony. This comes in a number of grades, the -most common being that
containing 40-43% of coprecipitated CaS04 and 15~17% of free S. Most
accelerators are somewhat soluble in hot acetone, and their S would appear
as free S in an uncured stock, but during the cure these are largely decomposed,
and any residual effect on free S is negligible.
During vulcanization, sulfur combines ilf four ways, (1) with the rubber
hydrocarbon, (2) with the natural resins and proteins, (3) 'with softeners if
present, and (4) with inorganic oxides to form sulfides such as ZnS. This has
been discussed by Kelly in several papers 8 and methods of separation given.
The ordinary figure for combined sulfur (corrected total sulfur-free sulfur) in-
cludes all of these forms, and also the sulfur from various compounding ingre-
dients as noted above. When interpreted in connection with the formula of
a compound and the processing to which it has been subjected, it is an excellent
indication of the state of vulcanization.. The percentage of combined sulfur
(CS) divided by the percentage of rubber present in a compound, gives the
coefficient of vulcanization (YC). Plotting either YC or CS against time of
cure gives a smooth curve for any given temperaturp conditions. It is thus
possible to take a sample from any point in a finished a;rticie, and by determining
its YC, estimate the equivalent cure. It should be remembered however,
that in the case of a tire or other complex article in which several stocks are
cured in contact with each other, migration of sulfur from a high sulfur stock
to one of lower sulfur content (on the rubber) is liable to occur, and also that
sulfur tends to migrate in the direction of the source of heat.
Free sulfur is at times a useful indication of the "non~blOoming" charac-
teristics of a stock, usually less thim 1.0% FS on the rubber bei~ necessary to
prevent sulfur" bloom."
Sulfur determinations have a tendency to .run slightly high, usually not
over 0.10%. Duplicates by extremely careful work can be made to check
within 0.04% S, but on ordinary rubber work, any variation not over 0.10%
is satisfactory checking.
Ash.;---An ash determination is only an approximation of the amount of
inorganic fillers in a stock, as it is affected by several factors. The various
clays, asbestine, lime, etc., carry considerable percentages 9f combined water
which is volatilized over a range of temperature. ,Whiting tends to lose CO 2
. _,-
8 Ind. Eng. Chem. 12, 875; 14, 196; 16, 148.
RUBBER COMPOUNDS 2015
unless very carefully handled. During the ignition, much of the sulfur present
combines with any zinc oxide, lime, magnesia or litharge to form sulfates. \ If
the sample is all()wed to catch fire, zinc oxide will be reduced and volatilized.
In practice, ash serves as a check on the subsequent analysis of the ash, and
sometimes itself gives sufficient information on the leading of a stock.
Analysis of Ash.-In addition to the determinations of sulfur, barium
sulfate and antimony given in the standard methods, it is frequently necessary
to further analyze the ash to secure an idea of the fillers and inorganic accelera-
tors present in the sample. Si0 2 , Fe 20a, Al 20 a, CaO, MgO, PbO and ZnO are
usually determined, but as these occur in many fillers in varying proportions,
some judgment is necessary for the proper interpretation of results, and micro;, •
scopical methods are of considerable aid in determining the fillers present,
Particularly in the field of cheap inert fillers, the compounder's choice is
often affected by the location of the factory and the materials most easily
available, as similar stocks can frequently be made with widely varying com-
positions. •
The following table shows the approximate percentages found in some com-
me'rcial compounding ingredients:
Asbestine-MgO 25%, SiO. 65%, Ah03 3.5%, CaO 1.5%, Ign. Loss 5.0%.
Barytes-BaSO.90-95%.
Dixie Clay-SiO. 46%, R.Oa 39%, CaO trace, MgO trace, Ign. Loss 15%.
Lead Oleate-PbO 25-30%.
Lime-Ca(OH). 91 %.
Lithopone-ZnO 29%, BaSO. 71 %.
Mag. Carb.-MgO 42%, Ign. Loss 50%.
Mag. Usta-MgO 87%, CaO 2.5%, SiO. and R.Oa 4.0%, Ign. Loss 6%.
Whiting (pptd.)-CaO 50%, MgO 3%, SiO. and R.Oa 5%, Ign. Loss 40%.
Free Carbon.-The direct determination of free carbon is quite accurate.
In addition to carbon black" the most used reinforcing filler, small percentages
of free carbon may come from any mineral rubber 01' hardwood pitch present,
but these are usually neglected.
Glue.-The presence of as little as 0.9% glue can be detected qualitatively,
but of late years, the use of glue in rubber compounds has become quite rare.
Calculation of the Probable Composition of a Stock.-The experienced
compounder will draw heavily on his.knowledge of the compounding ingredients
available and suitable to use in a stock giving properties of the sample, in
. calculating its probable composition, but the following will give an idea of
general procedure.
1. From the analysis of the ash, estimate the amounts of fillers and inorganic
accelerators used, allowing 15% ignition loss for clays, and using data such as
general appearance of the stock, state of cure, color, etc., to aid in judging the
amounts of CaO, MgO, and PbO used, if any.
2. Estimate carbon black, glue, and substitutes if any, from their special
analyses. --
3. Estimate rubber hydrocarbon by differenq_e. Subtract 3.0% of this
from the combined sulfur, and see-if the use of reclaim is indicated. The odor
of the stock is also often characteristic in connection with either reclaim or
softeners such as pine tar oil. If much CS is left over, figure it as 3.5% of
reclaim, and check this indicated reclaim content with the other extracts and
ash to see if it seems reasonable.
2016 RUBBER COMPOUNDS
4. Subtract 10% of the estimated reclaim and 3,5% of the estimated rubber
content from the acetone extract, and estimate the kiIld and amount of softeners
used from the remainder, taking into a'Ccount tne character of the extracts,
and correcting the chloroform extract if necessary for the effect of the re6laim
present. .\
5. Figure the sulfur used as total sulfur less 3.5% of the reclaim and 6.0%
of any mineral rubber present. . I
6. Calculate the specific gravity of the stock from the probable composition,
and compare with the actual specific gravity, which should be within 0.02,
except in the case of compounds containing _large amounts of softeners or
sulfur, with which the actual gravity is usually several points higher than the I)
theoretical.
Raw Rubber.-All but a very small pircentage of the rubber used in the
United States is plantation grown, and appears upon the market quite clean
and dry, and graded (by appearance and history) accor'ding to standards es-
tablished by the Rubber Exchange of New York. Except for research pur-
poses, the chemical laboratory is .seldom called on to determine more than
moisture, acetone extract and ash. The factory, however, is interested in the
rate of cure and physical properties with given formulae, which vary WIdely
with rubber of identical appearance. For control purposes, it is common to
blend at least three lots,of rubber together on mixing mills, .and evaluate the
blend by means of physical tests on experimental batches, the formulae used
varying with the factory and the use to which the rubJ;>er is to be put. Methods
for special analytical determinations are given in the previous chapter. For
data on the chemistry and structure of hevea rubber, see Dinsmore, Ind. Eng.
Chem. 18, 1140-5 (1926), and Fisher, Chemical Reviews, VII, 1 (Mar. 1930).
Reclaim.-Reclaimed rubber is classified according to the type of rubber
scrap from which it was prepared, and is subjected when required, to the same
chemical examination as a rubber compound, with particular emphasis on
specific gravity, moistu~e, alkalinity, acetone extract and ash.. These tests
however, are mainly useful in control of material of known sourc_e. and type, and
any new material requires physical testing, both alone and in the type of com-
pound where it is to be used, and also factory,tests in actual production com-
pounds to determine the effect on factory handling conditions.
Accelerators and Antioxidants.-A very considerable number of substances,
mostly organic, has come into use in recent years'as additions to rl!bber com-
pounds to accelerate the ra£e of vulcanization, or to retard the rate of deteriora-
tion after vulcanization. A few of these are of definite composition that per-
mits simple tests for purity. Most, however, are compl\}x, and require ex-
tended physical tests for their evaluation, and al1l'consequently sold chiefly
without test on the strength of their manufacturer's reputation. Data with
RUBBER COMPOUNDS 2017
Formula.- .
Mercaptobenzothiazole
.
a)-s-n
Trade name, "Capta;c."
Moisture, Ash and Fineness, as above.
Purity.-Min. 90.0%. Dissolve one-half gram sample in 25 ml. alcohol
and titrate with 0.1 N alcoholic N aOH, using phenolphthalein indicator. 1 ml.
0.1 N NaOH=.OI68 g. mercaptobenzothiazole.
Softeners
Softeners constitute an important class of rubber compounding ingredients,
as in addition to modifying the properties of the vulcanized compound, and
in most cases reducing the cost, they greatly affect the properties before vul-
canization and often make possible large savings in processing. A great many
materials are used, including petroleum products such as refined lubricating
oils, petrolatums, paraffin, and blown asphalts (mineral rubber), vegetable ojls
such as cottonseed oil, palm oil, and castor oil, wood distillation products such
as pine tar, pine tar oil and hardwood pitch, resins such as rosin, glycerol ester,
cumar, dam mar and shellac, and fatty acids such as stearic, oleic and lauric
and their zinc salts.
21)18 RUBBER COMPOUNDS
The most essential feature from the compounder"s standpoint is the uni-
formity of the material selected, and ordinary tests fo~ melting point, viscosity, '
etc., appropriate to the product are applied, the only features peculiar to rublter
being that copper and manganese must be absent" Ion account of the very
deleterious effect of even minute traces in rubber compoun\ls, and alkalinity
or mineral acidity must be absent on account of the effect on the rate of
vulca~ization. I
The filter containing the calcium oxalate is dropped in a beaker, 150 ml. of
hot water added together with 15 ml. of (1 : 1) H 2S0 4 al1d the oxalic acid titrated
with standard potassium permanganate.
1 ml. of 0.1 N KMn04=0.0028 gram\of CaO.
The Fe value of KMn04 multiplied by 0.5 = CaO value.
NOTEs.-If preferred, iron and aluminum may be precipitated'as hydroxides, and
filtered off, calcium being determined in the filtrate. It is advisable to redissolve the
precipitate to recover any occluded lime and again precipitate the hydroxides with
ammonia, the filtrate being combined with the main filtrate.
In place of titrating the oxalate of lime, it may be ignited and the residue weighed
directly as CaO.
If oxalic acid is added in the form of a fine powder instead of a solution, the calciulJ1
precipitates in a mu~h more granular form and requires less boiling before filtering.
Iron.-(In presence of silica) Half a gram or more of the finely ground
chilled slag in a beaker is treated with 50 ml. of boiling water, the particles
stirred up and kept in suspension and about 25 ml. of concentrated HCI added.
The solution is boiled until clear. (If coke dust is present it will still be evident,
but may be negLected.) Stannous chLoride ,soLution is added, drop by drop,
until the iron is reduced (solution becomes colorless) and 2...:3 drops excess
added. The solution is cooled by placing ~he beaker in cold water. Mercuric
chloride, HgCI 2, solution is added to precipitate the excess of stannous chloride.
The iron is determined by titration ivith standard potassium dichromate,
using- ferricyanide indicator on a spot plate. See subject in the chapter on
Iron. Report as FeO.
1 ml. of 0.1 N K 2 Cr 2 07 = 0.00558 g. of Fe, 0.007i9 g. of FeO,
Fe X 1.2865 = FeO.
NOTE.-The iron may be precipitated as hydroxide, then dissolved in dilute H 2S0 4,
the solution reduced by Qoiling with test lead or reduced in cold solution with zinc and
the solution titrated with standard potassium permanganate., This method is frequently
preferred where previous fusion to decompose the ore has been required. Lime may be
determined in the filtrate from iron. '
Hot water is added to make up to a volume of about 100 m!. The iron is
now precipitated by adding ZnO emulsion until the acid is neutralized and an
excess of ZnO forms on the bottom of the beaker.
The solution is boiled for a few minutes and the manganese titrated while
the solution is still hot (in presence of the precipitate) with standard potassium
permanganate.
The end-point is best seen by allowing the precipitate to settle slightly and
observing the clear upper stratum of the liquid.
1 m!. of 0.1 N KMn04=0.00165 g. of Mn.
The lime value of the KMn04 XO.588 = Mn.
N()TE.-In place of the method given above consult the chapter on Manganese for
the volumetric oxalic acid method.
Zinc.-To 0.5 gram (1 gram if zinc is low) of the sample in a casserole is
added 3 m!. of water, 5 m!. of HCI and 2 ml. of HNO.. When the SiO z is
completely gelatinized, about 4 grams of NH 4 CI is stirred in.
The sample is dehydrated only until the residue crumbles easily as baking
is liable to volatilize some zinc as chloride. About 30 m!. of hot water are
added and the solution brought to boiling, filtered and the residue washed with
hot water. I I
To the filtrate ammonium persulfate and bromine are added, the amount
being governed by the manganese present, i.e., 0.03 g. of ammonium persulfate
and 10 m!. bromine for every 0.01 g. of Mn in solution. Ammonia is added
'in slight excess, the solution boiled about 2 minutes, filtered, and the residue,
Mn02, etc., washed. (It is advisable to redissolve the precipitate in a little
dilute HCI and again precipitate with 'persulfate and bromine, adding the
filtrate to the main filtrate.)
The solution is just neutralized with HCI and 5 m!. excess added, followed
by 2 grams .of ,test _lead. The solution is boiled about 15 minutes.
About 8-10 m!. of HCl are added, the solution heated to about 60 0 C.
and the potassium ferrocyanide added. When the titration is almost com-
pleted, the bluish-white color of the precipitate changes to nearly pure white.
If the ferrocyanide is made up by adding 21.63 grams of potassium ferro-
cyanide per liter of solution, 1 m!. win equal very nearly 0.005 gram of Zn.
Consult chapter on Zinc.
In place of the above method the rapid method given in the chapter on Zinc
may be followed.
Magnesia.-0.5 gram or more of the finely ground chilled slag, placed in
a casserole, is moistened with water, 5 ml. of HCI and a few drops of HN0 3
2024 SLAG ANALYSIS
added. .The acids after reacting with the slag are evaporated off and the
silica dehydrated. The residue is taken up with 15! ml. of HCI and about
30 ml. of water, and boiled.
Silica is filtered off and washed. I
The filtrate is made slightly alkaline with ammorii1t, 8 ml. of (NH 4 )2S and
1 gram of (NH 4)2CO a added and the solution boiled, for a few minutes, the
precipitate then filtered off and washed several times with water containing a
little (NH 4 )2S.
The filtrate is made slightly acid with HCI and boiled down to about
50 ml. The precipitated sulfur is filtered off and the filter washed. If the
solution is cloudy, it is cleared by adding bromine water and boiling.
Ammonia is now added, and a small amount of ammonium oxalate to'
remove any CaO still remaining; after boiling, the oxalate of calcium ("lime'D
is filtered off and washed.
Magnesia is now precipitated in the filtrate by addition of Na 2HP0 4 or
another alkali phosphate, and the magnesium ammonium phosphate filtered
off, washed with 2% ammonium nitrate solution, ignited and weighed as
Mg 2P 20 7 • Consult the chapter on Magnesium.
Mg 2P 2 0 7 XO.3621,= MgO.
Copper: Colorimetric Method with Ammonia.-The percentage of cgpper
normally present in blast furnace slags/should not exceed 0.3%. Converter
slags may contain as much as 2.5% copper. The colorimetric m~thod is
especially adapted for this determination owing to its simplicity and speed.
Slags which have been" chilled" are readily decomposed by HCI. Stirring
constantly during the acid action and diluting as soon as decomposition is
complete prevents the separation of gelatinous silica and hastens the solution
of copper.
Procedure: Color S,tandards.-0.2 gram of pure copper foil is dissolved in
20 m'l. of 1 : 1 HNO a in a covered beaker, warming gf.ntly. After evaporating
down to about 10 mI., the solution diluted to 50,ml. is.,transferred to a graduated
liter flask, 200 mI. of ammonium hydroxide added and the solution diluted to
1000 ml.
5, 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100 ml. portions are taken and
placed in 100-ml. Nessler colorimetric tubes, (In place of the tubes clear
white glass bottles of about 150 mI. capacity may be use9..) The volume of
each solution is diluted to 100 ml. with 1 : 5 ammonium hydroxide.
If a colorimeter is available, a standard is taken that corresponds to that
of the sample as prepared according to directions below and comparisons made
by viewing the solutions through the depth of. solutions adjusted to give the
same intensity of color.
Preparation of the Sample.-Two grams off the finely crushed sample
(100-mesh) is treated in a beaker with 10 ml. of 'water to thoroughly moisten
the powder, 10 ml. of concentrated HCI are added and the mixture stirred,
warming gently until the undissolved residue becomes flocculent, remaining in
suspension. I
About 90 ml. of H 2S water are added or the ~me volume of water and H 2S
gas passed in until the solution is saturated .. The solution is warmed gently
SLAG ANALYSIS 2025
To soaps made by either of the above ,processes may be added various in-
gredients to obtain certain characteristics and qualities desirable from a view-
point of appearance and of detergency. Perfumes, coloring matter, and medi-
cants are added to toilet soaps; transparent soap is a combination of soap,
alcohol, sugar, and glycerin. Some soaps may contain fillers added to impart
special characteristics~ A few of these are colloidal clay, pumice, oatmeal,
bran, sawdust, milk, ground corn, blood albumin, gliadin, cellulose, etc. Medi-
cated soaps may contain carbolic acid, thymol, petroleum, vaseline, camphor,
gelatin, iodides, chlorides, zinc stearate, zinc oxide, and mercuro-compounds.
From this brief description it can be seen that a soap received in a laboratory
may contain any of the various constituents mentio~d. The following meth-
ods include only those commonly encountered inlsoap analysis. When analyses
for fillers are desired other than those described, the usual methods for such
materials are applicable.
In compiFng these methods, in many cases the standard methods for the
sampling and analysis of ,commercial soap and soap prodllct~ 9f the American
Chemical Society 2 and of the American Oil Chemists' Society 3-are used, and
acknowledgment is given. In some cases the A. C. S.-A. O. C. S. methods
are modified or replaced by methods which have been developed for soap works
laboratories where a large number of samples 'are handled daily for control
purposes, so that methods must be rapid as well as accurate. For sampling of
I
1 Compiled by the Chemical Division of the Procter ~nd Gamble Co. 'Through the
courtesy of the Company and C. P. Long. The Methods approved by the A. C. S. and
A. O. C. S. Committees on the Analysis of Soaps and Soap Products have been incor-
porated.
2 Standard Methods for Sampling and Analysis of CommercjAl Soaps and Soap
Products, J. Ind. Eng. Chern., 14, 1159 (1922); Ind. Eng./Chem., Anal. Ed. 9,2 (1937).
3 Standard Methods for Sampling and Analysis. ofpeommercial Soaps and Soap
Products, Oil and Soap, 11,90 (1934). -
2026
ANALYSIS OF SOAPS AND SOAP PRODUCTS 2027
Cakes or Bars
Samples on which an entire analysis is to be made should be run through an
ordinary food chopper. If the sample cannot be easily disintegrated and mixed,
it should be quartered by cutting at right angles in the center and portions from
all freshly cut surfaces shaved off with a knife, portions of these shavings being
used for analysis. Freshly stamped bars obtained in plant control may be
sampled with a cork borer of convenient diameter. Both ends of the sample
plug should be cut off and rejected. The samples should be preserved in air-
tight containers in a cool place. Excessive exposure to the air and ,handling
with the hands should be avoided as much !ls possible.
Liquid Soap
The sample should be mixed carefully with a spatula or spoon. Shaking or
any action to form suds should be avoided. Unused portions should be kept
in air-tight glass containers.
2028 ANALYSIS OF SOAPS AND SOAP PRODUCTS
METHODS OF ANALYSI~
For soaps containing from 5 to 25% of moisture and volatile matter use a
20 gm. (±0.04 gm.) sample. For soaps containing more than 25% moisture
and volatile matter use a 10 gm. (±0.02 gm.) sample! The weighed sample is
carefully transferred to a 500 ml. short-neck round hottom flaSK or an Erlen-
meyer flask. Add approximately 10 gm. of anhydrous, fused sodium acetate
to prevent violent frothing, and then follow with 100 ml. of xylene which has
previously been saturated with water by shaking the xylene with a small quan-
tity of water and distilling., Use the xylene distillate for the determination.
Attach the flask to a graduated distilling tube receiver, 'such .a_s the Bidwell-
Sterling or Stark and Dean apparatus. The receiver should be 'calibrated at
25° C. and normally of 5 to 6 ml. capacity. It is connected to a reflux con-
denser (19" Liebig). Prior to starting the determination, fill the receiver
with saturated xylene by pouring in through the reflux condenser.
~o that the refl~xing ~ill be under better control,! wrap the flask ~nd the tube
leadmg to the receIver WIth asbestos cloth. Apply; heat to the flask by means
of a Argand gas burner or an electric heater and distill slowly. :The rate at the
start should be approximately 100 drops per minute. When the greater part
of the water has apparently distilled over, increase the distillation rate to 200
drops per minute until no more water is collected. rP,)lrge tJ:{e reflux condenser
, A. K. Church and J. H. Wilson, Soap, 7, 35, 193[.'-=-
ANALYSIS OF SOAPS AND SOAP PRODUCTS 2029
during the distillation with 5 ml. portions of xylene to wash down any moisture
adhering to the walls of the ~ondenser. The water in the receiver may be made
to separate from the xylene by using a spiral copper wire. Move the wire up and
down in the condenser occasionally, this causing the water to settle to the bot~
tom of the receiver. Reflux for at least two hours, after which the heat is
turned off. Adjust the water distilled over into the receiver to 25° C. Read
the volume of water and calculate the percentage of moisture' in the soap, as
follows:
Volumeinml.at25°C.XO.997100_01 't .
. h f 1 X - /0 mOlS ure III soap.
Welg t 0 samp e
Total Fatty Acid, Fatty Anhydride, Combined Alkali, and Anhydrous Soap
(a) Soaps Containing a Small Amount of, or No Coconut OiI.-Weigh a
5±.01 gram sample on a tared watch glass, transfer to a 250 ml. beaker, and
dissolve in 50 ml. of hot water. Transfer to a 250 m!. glass~stoppered extraction
cylinder, make up to a volume of 100 mI., and cool. Add 3 drops of methyl
orange and sufficient dilute H 2S0 4 (1 : 4) to give an excess of acid after decom~
posing the soap. Add 40 ml. of ethyl ether and shake gently until the contents
are well mixed. Let the cylinder stand until the ether layer has separated
clear. Draw off the ether layer into a tared 150 m!. Soxhlet flask through a
slender glass siphon by means of suction. Wash the siphon by drawing a few
m!. of ether through it. Repeat the extraction at least 3 more times, using
25 mi. of ether and shaking more vigorously after the first extraction. Silicated
soaps and flakes which show a tendency to dry out should be extracted about
6 times. Combine all extracts in the same Soxhlet flask. If the extract is not
clear transfer to another tared Soxhlet flask through a filter. Evaporate the
ether from the Soxhlet flask by warming on the edge .0J the steam bath. Rapid
evaporation will cause condensation of moisture in the flask. In case of soaps
high (40% or higher) in. coconut oil, evaporate only to about 50 ml. and proceed
as in (b). For other soaps evaporate to dryness, and finally dry for 1 hour
at 90 o ±2° C. Cool and reweigh. Dissolve the extracted fatty acids in about
40 ml. of hot, neutral alcohol, add a few drops of phenolphthalein indicator and
titrate with Nil NaOH to exact neutrality.
Weight of fatty acids X 100 01 t t l f tt .d
Weight of sample /0 0 a a yaCl .
NaOH titrationXO.9
% water equivalent to fatty acids.
Weight of sample
% total fatty acid-% water equivalent = % "Fatty Anhydride."
NaOH titration X 3.1
% combined Na 20.
Weight of sample
NaOH titrationX4.7
% combined K 20.
Weight ?f sample
% fatty anhydride+% combined alkali = % "Real Soap" or Anhydrous Soap.
2030 ANALYSIS OF SOAPS AND SOAP PRODUCTS
(b) Soaps 1;Iigh in Coconut Oil.-If desired, this method may be used for
any soap. It is often called the" soda soap" method. 2 • 3
Proceed with the extraction as for boiled soaps except that the extracts are
collected in a 250 ml. beaker or Soxhlet flask. Partially evaporate the ether
to not less than 50 ml. on the edge of a steam bathi and add 50 ml. of hot,
neutral alcohol redistilled over caustic. Add several'drops of phenolphtlialein
indicator and titrate with Nil NaOH to exact neutrality. Evaporate on the
steam bath to dryness, and while the soap is in the form of a thick paste, swirl
the beaker on its side, thus distributing the soap in a thin layer along the sides
and bottom of the beaker. .
Dry to constant weight in an oven at 105°±2° C. After a preliminary
drying at 105° C. for an hour, drying may be continued at 150° C. On remova1
from the oven cool in a desiccator and weigh as soon as cool.
Correct for neutral salts in t,he caustic solution by determining the amount of
salt per ml. by neutralizing 20 ml. of Nil NaOH with Nil HCI using phenol-
phthalein indicator and drying the residue to constant weight at 105° C, From
the weight of the residue found, subtract the' weight expected if the reagents
had been 100% pure. The difference divided by 20 gives the correction per ml.
for neutral salts. Subtract the product of the total fatty acid titration times
the factor (0.022) plus correction for neutraL salts from the weight of soda soap
to obtain total fatty acids. Calculate as for boiled soap.
If the soap is a cold process soap containing superfat, i.e. un saponified fat,
and total fatty acids are required, it will'be necessary to treat the sample after
dissolving in water with 5 ml. (or more) of 50° Be caustic potash solution.
The solutions should be mixed thoroughly. With cover glass on the beaker,
place over an opening in a steam bath, saponify for one hour with occasional
stirring. Acidify with an excess of dilute H 2S0 4 (1 : 4) and proceed as above.
If the real soap in a soap containing superfat is desired, it will be necessary
to determine the superfat including free fatty acid and unsaponifiable and to
correct the total fatty acid as determined without fur*er saponification.
Chlorides !
Transfer the acid water in the cylinder from the total fatty acid extraction to
a 250 ml. beaker, neutralize with chlorine-free Nil NaOH and back titrate with
Nil H 2S0 4 to make the solution one drop on the acid side of neutral. Add 1
ml. of 10% potassium chromate indicator. Piace over a Wratten Safelight-
Filter No. 00 illuminated by a 200-watt Mazda light at dist"ance·of.12". 'Titrate
slowly with Nl10 AgN0 3 solution, stirring vigorously. The addition of a little
gum arabic may aid in the titration.
Titration XO.585
% NaCI,
Weight of sample
TitrationXO.745 =o/c KCI
Weight of sample 0 •
In case total anhydrous soap is nob. to be determined, it. will be more con-
venient to use the following method. Dissolve 5±.01 grams of soap in 40 ml.
I of hot water, add 10 ml. J : 4 H 2S0 4 and heat on a"steam bath or hot plate until
ANALYSIS OF SOAPS AND SOAP PRODUCTS 2031
the fatty acids rise in a clear layer. Filter through a wet 9 cm. filter paper,
collecting the filtrate in a 250 ml. beaker. Wash the fatty acids and filter paper
with hot water until the total volume in the beaker is about 80 ml. Titrate as
described above with N/l NaOH and N/l H 2S0 4 until the solution is one drop
on the acid side of neutral. Titrate with N/lO AgNO a as described above.
A rapid method is that of Bennett.2 , 3, 5 Dissolve 5±.01 grams of the soap
in 300 ml. of water, boiling if necessary to affect solution. Add an excess of
chlorine-free.magnesium nitrate solution (about 25 ml. of a 20% Mg(NOsh·-
6H 20 solution). Without copling or filtering, titrate with N/lO AgNO a as
described above.
Unsaponifiable Matter
Weigh 5±.01 gms. of the sample into a 200 ml. Erlenmeyer flask. Add 30
ml. of redistilled 95% ethyl alcohol and add 5 ml. of 50% aqueous KOH. Boil
the mixture one hour unde~ a reflux condeiser. Transfer the solution to the
extraction cylinder and wash the flask with 95% redistilled alcohol to the 40 ml.
mark. Complete the transfer first with,warm and then cold water until the
total volume is 80 ml. Then wash the flask with a small quantity of petroleum
ether. Cool the cylinder aIi'd contents to room temperature. Add 50 ml. of
petroleum ether and proceed with the extraction as outlined under "Unsaponi-
fied and Unsaponifiable."
Weigh the residue and correct for fatty acids. Report the result as "Un-
saponifiable Matter." Deduct the unsaponifiable result from that of the
unsaponified and unsaponifiable and report the difference as "Unsaponified
Matter." I •
For soaps containing lanolin, many more extractions for Unsaponified and
Unsaponifiable will be required for complete removal. For any sample, thor-
ough and vigorous shaking is necessary in order to bring the two phases into the
most intimate contact possible.
free fatty acid as oleic acid, or lauric acid if coconut soap. Save the solution
for determination' of carbonates.
(b) When" Builders" Are Determined. a. 2._Weigh a 10±.02 gm. sample
and dissolve in 200 ml. of hot, neutral 95% alcohol. Filter through a 9 cm.
filter paper into a 500 ml. Erlenmeyer flask and wash the residue three times
with hot, neutral 95% alcohol. Protect from CO 2 and other acid fumes during
filtration. If the soap contains borax, or phosphates, it can be washed with
alcohol indefinitely without removing the apparent alkalinity. Cover the flask
with a watch glass and heat to incipient boiling. Titrate with N/4 H 2S0 4 and
report as "free N a 20 (or N aO H) " or "free KzO (or KO H) " as the character of
the soap indicates. If the alcohol is not pink, titrate with N/lO NaOH and
report "free fatty acid" as oleic acid.
If potassium carbonate is present, the free KOH results will be high due to
the solubility of KzCO a in the alcohol. For liquid soaps containing free KOH
and K 2CO a the following procedure has given satisfactory results: Weigh a 20
gm. sample of the soap solution into a 500 ml. Erlenmeyer flask. Add 10 gms.
of ~eutral dry salt. Mix the salt with the soap thoroughly by shaking. Allow
to stand 10 minutes covered by a watch glass. Add 20 m!. of hot neutral 95%
alcohol. Shake thoroughly and place on the steam bath for 10 to 15 minutes,
shaking occasionally. Filter through a neutral filter paper moistened with
neutral alcohol into another 500 m!. Erlenmeyer flask. Retain most of the
insoluble in the flask, if possible. Wash with 25 m!. of hot alcohol, pouring
around the top of the paper so as to wash the paper at the same tim'e. Give
the paper a slight additional wash with neutral alcohol after the first has drained
through. Allow to drain. Titrate with N/4 or Nj2 acid, and from this
calculate free KzO.
Sodium Carbonate
(a) If No "Builders" Are Present.-Add. from a burette 10 ml. of Nil
H2,S04 to the solution in the flask from the free alkalinity titration (a). Add 50
ml. of distilled water and boil gently for 1 hour, swirling the flask occasionally.
Remove the flask, add 10 m!. Nil NaOH and titrate with N/4 H 2S0 4 to the
colorless endpoint. If this titration is more than 7.5 ml., add another 10 ml.
portion of Nil H 2S0 4 and again boil for 1 hour. Add 10 m!. Nil NaOH and
again titrate with N/4 H 2S0 4• Calculate the per cent carbonate from the acid
titration.
(b) If" Builders" Are Present.-Dissolve the residue upon the filter paper
from the free alkalinity determination (b) with hot water into the original
beaker. If no insoluble matter other than insoluble Si0 2 from the builder is
present, the filter paper should be punctured and the entire contents of the paper
washed through. CQntinue washing until a few drops of phenolphthalein
indicator dropped on the filter paper show only a very faint pink color. Cool
the filtrate'and titrate with Nil H 2S0 4 using methyl orange indicator.' Record
the titration as the "carbonate titration" and in case no silicate is present,
calculate direct to sodium carbonate. If silicate is present, it should be deter-
mined as indicated in the" Note" under" Determination of Silica Present as
Alkaline Silicates." The above titration must then be corrected for the silica
present in calculating the carbonate.
2034 ANALYSIS OF SOAPS AND SOAP ,PRODUCTS
be used. The following method (A. C. S.-A. O. C. 'S.2. 3) has proved satis-
factory.
A 250 ml. Erlenmeyer flask is placed on a gauze over a burnet. The flask is
equipped with a 2-hole rubber stopper, through one opening of which is a lO-inch
reflux condenser and through the other a thistle tube equipped at the outer eIid
with a 3-way stopcock. The lower end of the thistle tube is drawn to a sma)l
point, which is placed very close to the bottom of the flask. Then the straight-
away end of the stopcock is attached to a small funnelfor the introduction of acid
to the flask. The other opening of the stopcock is attached to receive air from
a purifying train consisting of a wash bottle containing concentrated sulfuric
acid, and a second at the outer end of the train containing a 50% solution of
potassium hydroxide. The top of the reflux condenser is attached first to a
drying wash bottle containing concentrated sulfuric acid, and then to a weighed
absorbing train consisting of a suitable potasJ:l/bulb charged with 50% potassium
hydroxide, and a second containing concen,trated sulfuric acid. This train is
attached to a protective U-tube containing calcium chloride. The U-tube is
attached to an aspirator. / .
Procedure.-Set up the apparatus, leaving out the weighed train, and
aspirate with a slow stream of the dry carbon dioxide-free air until the apparatus
is freed of carbon dioxide. Insert the train and continue the aspiration for
Y2 hour. Check the weight of the train to determine if the air is passing
through too fast, or if the system is free from carbon dioxide. The system
must be free from leaks. Weigh out 1 or 2 gms. of the sample into the Erlen-
meyer flask, cover ,vith 20 ml. freshly boiled distilled water, close the apparatus
with the train in place. Add 20 ml. dilute hydrochlori,c acid (1 : 1) thrQugh the
funnel very slowly, with no heat being applied to the flask. The rate of adding
acid should be carefully controlled so that the gas does not pass through the
train too rapidly. As soon as the acid is added, 'start aspiration gently. -When
the absorption begins to stop the gas flow, start heating gently and continue
until the contents of the flask have boiled 15 to:20 minutes. Stop heating and
continue aspirating until,the flask has cooled down. Remov_~ the train and
weigh. Calculate increase of weight as carbon dioxide. Carbon dioxide
multiplied by 2.41 equals sodium carbonate.
All connections should be made glass to glass. Apparatus with ground
glass joints is available and'is preferred. Ascarite-filled absorption bulbs and
scrubbing tower instead of those using KOH are very satisfactory and con-
venient. !
Determination of Silica Present as alkaline Silicates 2. 3
platinum dish at a low temperature. When charred, extract the soluble salts
with water, return the paper and charred residue to the dish and complete the
ignition. Unite the residue in the dish and the water extract, carefully acidify
with hydrochloric acid, finally adding the equivalent of from 5 to 10 ml. con-
ce'ntrated hydrochloric acid in excess. The dish or casserole containing the
,solution should be covered with a watch glass while adding acid so as to avoid
loss by spray.
Evaporate the acidified solution (washing off and removing the cover glass
if used) to dryness on steam bath or hot plate at a temperature not exceeding
120 0 C. Cool, moisten with concentrated hydrochloric acid, let stand 5 to 10
minutes, breaking up all lumps with a stirring rod. Add about 25 m!. of hot
water. Heat a few minutes and filter through a small ashless paper. 'Wash
thoroughly with hot water.
Evaporate the filtrate to dryness and repeat the above treatment, filtering
on a second paper*
Carefully dry the filter papers and residue at 150 0 C. in a tared platinum
crucible, ignite ~t as Iowa temperature 8 as possible until the paper is consumed,
then raise the temperature and heat to constant weight in a muffle or over a
blast lamp. Cool in a desiccator and reweigh. If extreme accuracy is de-
sired, moisten the residue in the crucible with water, add 5 m!. hydrofluoric acid
and 4 drops of concentrated sulfuriC' acid, evaporate to dryness, and ignite as
before for about 2 minutes. Cool in a desiccator and reweigh. The difference
between this weight and the previous weight is the weight of Si0 2• 9
To calculate sodium silicate (1 Na 20 : 3.25 Si0 2) multiply Si0 2 by 1.308.10
. NOTE.-In case carbonates have been determined as under (b), the filtrate may be
used for the silicate determination. In that case, add 5 mI. of concentrated hydrochloric
acid to the solution in the beakllr, evaporate to dryness and proceed as above.
Borax
Qualitative Test.-A qualitative test may be made by wetting a piece of
turmeric test pll:per with a few m!. of the Si0 2 filtrate. If the paper on drying
in air turns a brick red, a borate or perborate is present. To determine if the
coloration is due to perborate, dissolve about 2 grams of the original sample
in about 100 ml. of a 5% potassium iodide solution containing 2 m!. of starch
solution. Add 10 m!. of 1 : 4 sulfuric acid and stir. A blue solution denotes
the presence of an oxidizing agent which with a positive turmeric test can be
assumed as indicating the presence of perborate.
Quantitative Test.~To determine borate quantitatively in case silicates are
not present, weigh 1O±.02 grams of the soap (5±.01 grams if more than 5%
borax is present), transfer to a 250 m!. beaker and break up the SOl\p with 50 m!.
of water and 50 m!. of N /1 H 2S0 4 • If the solution is not. distinctly red to
methyl orange, add N/1 H 2 S0 4 until it is acid. Do not heat longer than
* The filtrate may be collected in.a 250 mI. volumetric flask and used for qualitative
and quantitative phosphate.
S K. A. Krieger and H. S. Lukens, Ind. Eng. Chern., Anal. Ed., 8, 118 (1936) .
.9 W. F. Hillebrand, The Analysis of Silica,te and Carbonate Rocks, U. S. Geological
Survey, Bulletin 700, 102.
10 Report of the Soap Analysis Committee, A. O. C. S., Oil and Soap, 12, 10 (1935).
2036, ANALYSIS OF SOAPS AND SOAP PRODUCTS
necessary. Filter o.ff the fatty acids through a wet paper, wash well with hDt
water, adding all washings to. the filtrate. Neutrali:ze to. methyl Drange with
Nil NaOH and nDte the reading. Add a few drDps o.f phenDlphthalein and
titrate to. the endpDint with Nil NaOH. I
Add abDut 1 gram o.f neutral mannitDl and again titrate very SlDwly. Add 1
gram mDre Df the mannitDl and titrate again and repeat this prDcess until
mannitDl causes no. further actiDn Dn the endpDint. If phDsphate and bDrate
are bDth present and no. silicate, add 1 gm. Df sDdium chlDride with the mannitDl
befDre titrating; Calculate the number Df milliliters Df sDdium hydrDxide
added frDm the methyl Drange endpDint to. the final endpDint Df % Na 2B40 7
by multiplying by 0.503 Dr to. % Na zB 40 7·10H zO multiplying by 0.954.fDr:1O
gm. samples. In case phDsphate is present, a cDrrectiDn must be made since
part Df the phDsphate is titrated in the borate titration. For a 10 gm. sample
the following formula is used:
["Borate Titration"-(% .Na aP0 4 XO.61)]XO.503=% NaZ_B407
Since silicic acid affects the phenolphthalein and methyl orange endpDints in
the mannitol titration, it must be removed by the following method. Weigh
a 5 gm. sample in a tared platinum dish and ignite over a IDW flame until all
of the Drganic matter has been burned Dff, but do not ash the salts. CDDl the
dish and CDntents and place it in a beaker. 'Add abDut 100 ml. Df water and bDil
until the residue is remDved frDm the dish. Wash the contents Df the beaker
into. a 500 ml. distillatiDn flask, add 15 ml. Df cDncentrated hydrochlDric acid,
and distill to. dryness. PDur the distillate back into. the distillatiDn flask and
repeat. CDDI and wash the cDntents Df the flask into. the beaker cDntaining the
distillate. Filter thrDugh a quantitative filter paper and wash well. Stir the
filtrate with a stick Df caustic until nDt quite neutral to. methyl Drange. Neu-
tralize with Nil NaOH and cDntinue titratiDn as abDve with.mannitDl. Cor-
rect calculatiDns fDr phDsphate in the same manner as abDve.
I
I
Perborates I
Weigh a 1 gm. sample and transfer to. a 300 ml. Erlenmeyer flask with cDld
water. Make up to abDut 80 ml. Acidify immediately with 20 ml. Df Nil
sulfuric acid and titrate at o.nce with N 110 Po.tassium permanganate. One ml.
Df pDtassium permanganate is equivalent to. 0.0008 gms. Df o.xygen, 0.007694
gms. Df NaBO a ·4H 2 0, and 0.00409 gms. Df NaBO a• -
Phosphates
A Qualitative Test for Phosphate May Be 'Made on the Filtrate from the
SiO z Determination.-If meta- Dr pyrDphDsphates are present add abDut 10 ml.
of nitric acid and hDld at near bDiling temperature' Dver night to. convert them
to. DrthDphDsphates. To. 20 ml. add 2 ml. Df cDncentrated nitric acid and 5 ml.
Df 30% NH 4 NO a so.lutiDn, phDsphate-free. Heat to. bo.iling and add 20 m!. o.f
ammDnium mDlybdate sDlutiDn. Let the sDlutiDn stand 20 minutes Dn the
steam bath. If a yellDw precipitate (Drms, a phDsphate is ,present.
Quantitative DeterminatiDn ~. 3-Pipette an llliqUDt cDrrespDnding to. 0.50
gm. Dr 1 gm. into. a 250 ml. beaker; add ammDnium hydrDxide in slight excess;
ANALYSIS OF SOAPS AND SOAP 'PRODUCTS 2037
and barely dissolve the precipitate formed with a few drops of nitric acid, stir-
ring vigorously. Add about 15 gms. of dry ammonium nitrate or a solution
containing that amount.
To the hot solution add 70 ml. of the molybdate solution for every decigram
of phosphoric acid (P 2 0s) present. Digest at about· 65° C. for an hour, and
determine if. the phosphoric acid has been completely precipitated by the addi-
tion of more molybdate solution to the clear supernatant liquid. Filter, and
wash With cold water or, preferably, 10% ammonium nitrate solution. Dis-
solve the precipitate on the filter with ammonium hydroxide U : 1) and hot
water, and wash into a beaker to a bulk of not more than 100 ml. Neutralize
with hydrochloric acid, using litmus paper or bromthymol blue as an indicator,
cool, and from a burette add slowly (about 1 drop per second), stirring vigor-
ously, 15 ml. of magnesia mixture * for each decigram of phosphoric acid (P205)
present. After 15 minutes add 12 ml. of ammonium hydroxide (sp.gr. 0.90).
Let stand till the supernatant liquid is clear (2 hrs. is usually enough) filter,
wash the precipitate with dilute ammonium hydroxide (1 : 10) until the wash-
ings are practically free from chlorides; dry, burn first at a low heat and ignite
to constant weight, preferably in an electric furnace, ,at 950°-1000° C.; cool
in a desiccator, and weigh as Mg 2 P 2 0 7 • Calculate and report the result as
• percentage of P 2 0 5 or alkaline phosphate known to be present.
Sulfates
Weigh a 50 gm. sample, dissolve in water, break up the soap with 35 ml. of
concentrated hydrochloric acid and filter off insoluble matter and fatty acids
through a wet filter paper. Wash the filter well with cold water. If silicate is
• present, remove it by evaporation and filtration as in the silicate determination.
In either case make up the silicate-free filtrate to 175 ml. with water, add 5 ml.
of hydrochloric acid and boil. 'To the boiling solution, add 15 ml. to 20 ml.
of a 10% barium chloride solution, drop by drop. Continue boiling until the
precipitate is well formed and keep at a temperature of about 70° C. for one
hour or unt~l the precipitate settles well. Filter through a prepared Gooch
crucible, ignite gently, and weigh as barium sulfate. Calculate to sodium sul-
fate or to sulfate known to be present.
Digest hot a 2 to 10 gm. (±0.01 gm.) sample with 200 ml. of freshly boiled
ethyl alcohol neutral to phenolphthalein (94% or higher). Filter through a
counterpoised filter paper neutral to phenolphthalein, or a weighed Gooch
crucible with suction, protecting the solution during the operation from carbon
dioxide and other acid fumes. Wash the residue on the paper, or in the crucible,
with hot neutral alcohol until free from soap. Dry the filter paper, or crucible,
and residue at 100° to 1050 C. for 3 hours, cool, and weigh the total matter
insoluble in alcohol. (The matter insoluble in alcohol will contain most of the
alkaline salts, such as carbonates, borates, silicates, phosphates and sulfates,
as well as starch, and may be used for the approximate determination of these
* Magnesia mixture-Dissolve 55 gms. of crystallized magnesium chloride (MgCh·-
6H 20) in water, add 140 gms. of ammonium chloride, 130.5 mI. of ammonium hydroxide
(sp.gr. 0.90), and dilute to I liter.
2038 ANALYSIS OF SOAPS AND SOAP PRODUCTS
constituents. These salts are not entirely insoluble in alcohol, so for accurate
determinations separate portions of the soap should qe us.ed.)
This method requires a source of dry, oil-free stearh which is passed through
the sample treated with acid, sufficient to liberate the fatty acids from the soap.
The steam is next passed through strong caustic sqlution to scrub out any
volatile fatty acids while the volatile hydrocarbons I are condensed with the
steam in a suitable arrangement which allows the excess water to flow away,
leaving the volatile hydrocarbon in the measuring burette. The method may
be applied to samples containing substances immiscible with water and volatiie
with steam. For solvents heavier than water a Bidwell-Sterling tube should be
used. "
C
E
Steam
~====~«=======~
~==~
To Sewer
Steam
A B
A - Steam Trap
B - Sample Flask
C- Dropping Funnel J
D-Casu.tic Scrubber-Iron
E - Kieldahl Trap
F -Condenser
G-Siphon H
H-Burette
I - Expansion Bulb
J - Automatic Overflow
FIG. 287
Continue the distillation until there is no increase iIi the volume of the upper
layer for 45 minutes or no small droplets can be note4 in the condensate ..
When distillation is completed, shut off and drain condenser water and allow
the steam to heat up the condenser to drive out the las~ traces of volatile hydro-
carbon. Shut off the steam as soon as vapor begins tolissue from the lower end
of the condenser. Immediately open the stopcock ofl the dropping furinel to
prevent caustic being drawn into the evolution flask., .
Stopper the burette and allow its contents to come to room temperature or
bring them to a definite temperature by placing the burette in a water bath heid
at 25° C. for one to two hours.
Read the volume of the upper layer to the nearest .011)1l. The volume times
the specific gravity equals the weight of the volatile hydrocarbon. The specific
gravity should be determined at the temperature at which the volume is read.
A small Sprengel tube made of 3 mm. glass tubing is convenient for this purpose.
Calculation.-
ml. of volatile hydrocarbon X specific gravity XIOO
Weight of sample
= per cent volatile hydrocarbon.
For some samples the volatile hydrocarbon dontent may be so low that a larger
sample than 50 or 100 gms. is desirable. The size of the evolution flask may
need to be increased if larger samples are used. The amount of water in the
evolution flask and acid used should also be corr<;:spondingly increased.
Titer Test
1. Preparation of Total Fatty Matter (Fatty and Rosin Acids and Un-
saponifiable Matter.-Dissolve 200 gms. of soap * in 200 ml. of water in a liter
Erlenmeyer flask, add 20 ml. of 1 : 3 sulfuric acid and heat until the fatty
matter collects in a clear layer.
The flask should be covered with a watch glass. / Much time is saved by
placing the sample in a liter beaker and using,a motor-driven stirrer to keep
constant agitation. Siphon off the acid water layer and decant the fatty matter
through a dry filter paper into tall form, lipless'lS0 ml. beaker. If droplets of
water can be seen, refilter the fatty acids through a dry filter paper.
2. Determination.-A standard titer therniometer: 14, 15 graduated at zero
and in tenths degrees from 10° to 65° C. is used or in .20 f1'om~2° C. to 66° C. _
It should be certified by the United States Bureau of Standards, or carefully
standardized against a certified thermometer. Special thermometers (gradu-
ated in tenths degree Centigrade) in 20° C. ranges from- 00_i:!0° C. may be used.
After the fatty acids are dry, cool carefully by immersing in and out of water
at about 17° C. until a cloud begins to form, stir,ring continuously with the
thermometer (about 150 r.p.m.) so that no crust forms on the sides or bottom
" * The fatty matter obtained is also used for the iodine value, saponification value,
and may be used for rosin determinations. If the soap contains superfat, it must be
saponified with glycerol-potash as in the titer determination for fats.l0
14 Official and Tentative Methods of Analysis of the Association of Official Agricul-
tural Chemists, 4th edition, 1935, 408. ' -
15 J. Ind. Eng. Chern., 18, 1349 (1926).
,
ANALYSIS OF SOAPS AND SOAP, PRODUCTS 2043
of the beaker. Care must be taken not to cool too rapidly or low results will
be obtained. Hold the beaker in the air and continue stirring more slowly
(about 100 r.p.m,) until the mercury remains stationary for 30 secon{ls or
begins to rise within that time. Wipe the outside of the beaker dry and care-'
fully place the beaker and contents in a suitable container (air-jacket) sub-
merged in water, or in ice, salt, water mixtures when necessary"maintained at a
temperature of 10° C. below the titer. For titers above 30° C. the temperature
should be 20° C. Allow the thermometer to hang without any disturbance
with the bulb in the center of the fatty acids and' follow the thermometer
readings. Record the highest point to which the mercury rises as the titer of ,
the fatty acids.
When not enough soap or fatty matter is available, the official method of the
A. C. S.l5-A. O. C. 8.3 should be used. This method, which differs slightly in
details 'and requires less fatty acids, may also be used on the fatty acids ob-
tained in the above procedure.
Iodine Value 15
The iodine value is determined on the fatty matter obtained in the above
procedure for titer test by the Wijs' method. If rosin is present this method
cannot be used.* The solutions required are Wijs' iodine monochloride solu-
tion,1•. 15 standard Nj10 sodium thiosulfate, 5% potassium iodide solution,
U. S. P. chloroform, and starch indicator. Weigh exactly a sample of such size
that at least 100% to 150% of the amount of iodine absorbed remains in excess,
on a watch glass and transfer to ,a clean, dry 500 ml. glass-stoppered bottle,
washing the watch glass thoroughly with 20-30 ml. of chloroform. Add 25 ml.
of iodine mono chloride and at the same time prepare a blank with the same.
amount of chloroform and iodine mono chloride. Mix the contents gently and
immediately put the bottles in a dark place at a temperature of 75°-90° F. for
exactly 30 mi.nutes. Add 20 ml. of 5% potassim}l iodide to the sample and
blank followed at once by 100 ml. of water. Titrate with Nj10 sodium thio-
sulfate, using starch solution as an indicator. Do not add starch until the
solution·has a pale straw yellow color. Toward the end of the reaction stopper
the bottle and shake violently. The number of cubic centimeters of standard
thiosulfate solution used for the blank minus the amount used for the sample,
gives the thiosulfate equivalent of the iodine absorbed by the amount of sample
used in the determination. Calc!J.late to per cent iodine absorbed. Mechan-
ical agitation using a small variable-speed motor attached to a glass stirrer is
an advantage. It is necessary to use a wide mouthed glass-stoppered bottle.
Saponification Value
Weigh exactly a 5 gm. sample of the fatty matter obtained in the titer deter-
mination and wash into a 300 ml. ErlenqJ.eyer flask with hot, neutral, redistilled
* In case the iodine value must p~ obtained on samples containing rosin, a 0.1 gm.
sample is dissolved in 20 ml. of glacial acetic acid and 20 ml. chloroform in the glass-
stoppered bottle. Immerse the bottle in water at a: temperature of 21.5°-22.5° C. for
30 minutes. Add the Wij,s' solution as above and maintain the temperature at 21.5°-
22.5° C. for one hour. Other, operations are the same as above.
16 P. C. McIlhiney, J. Am. Chem. Soc., 29, 1222 (1907).
2044 ANALYSIS OF SOAPS AND SOAP PRODUCTS
boil on the steam bath and boil exactly four minutes under a reflux condenser.
Remove from the steam bath, add to the liquid about five times its volume of 7
to 10% sodium chloride solution. Transfer to a separatory funnel, washing
the Erlenmeyer flask with small portions of ethyl ether, the ether washing being
added to the aqueous solution in the separatory funnel. Add about 30 ml.
of ethyl ether to the separatory funnel, shake vigorously, and allow to stand
for at least five minutes. Draw off the aqueous solution into another separa-
tory funnel designated as No.2. Add about 30 ml. of ether to funnel No.2,
shake vigorously, allow to stand for at least five minutes and then draw off
the aqueous layer into another separatory funnel designated as No.3. Repeat
the extraction of the aqueous solution three more times. Then discard the
aqueous liQuid and add the combined ether solutions to separatory funnel No. 1.
Add 50 m!. of sodium chloride solution (7-10%) to the ether extract in funnel
No.1, shake vigorously and allow to stand for at least 5 minutes.
Draw off the aqueous solution into another separatory funnel. Repeat this
operation until the washings are neutral to methyl orange, all salt water wash-
ings. being combined. ' Add 50 m!. of ether to the washings in the separatory
funnel. Shake the funnel vigorously, allow to stand for at least five minutes
and then draw off and discard the aqueous salt solution. Add 50 ml. of sodium
chioride solution and wash the ether layer until the. aqueous salt solution is
neutral to methyl orange. 'Dry the ether extract by the addition of about 3
gm. of anhydrous sodium sulfate to the separatory funnel, filter into a 250 m!.
Erlenmeyer flask and evaporate off the ether slowly on the steam bath to
dryness. . . '
Second Esterification.-Coollj,nd dissolve the residue in 20 m!. of absolute
ethyl alcohol and then proceed as above under "First Esterification." Add
30 ml. of neutral alcohol (94% or higher) and titr!J,te rosin or rosin soap as
desired, using phenolphthalein as indicator. (1 ml. N/2 alkali=0.173 gm.
rosin or 0.188 gm. rosin soda soap.) If the true fatty acid soap is desired, sub-
tract the rosin soap from the total anhydrous soap (real soap) obtained under
total fatty acid determination.
The above method gives somewhat high results on all low percentages of
rosin. On the higher percentages, it is approximately correct. In all cases
where the rosin content is found to be less than 5%, the actual presence or
absence of rosin should be checked qualitativefy by the Lieberman-Storch test,
which is as follows:
One to 2 ml. of the sample of fatty acids are shaken in a test tube with 5 to
10 m!. acetic anhydride, reagent grade, with gentle warming on a steam bath.
After cooling, pour 1 to 2 m!. in a white porcelain dish and allow a drop or two
of sulfuric acid (sp.gr. 1.53) to run down the side of the vessel. If rosin is
present, a fugitive violet coloration changing to a brownish tinge is immediately
produced at the margin of contact of the reagents. The test should be checked
with a sample of fatty acids to which a small amount of rosin has been added.
(Sulfuric acid of 1.53 sp.gr. is prepared by diluting 34.7 m!. of concentrated
suifuric acid (sp.gr. 1.84), reag~nt grade, with 35.7 m!. of distilled water.)
WATER ANALYSIS
Probably at no other time has the importance of a ;{rater supply, either for
domestic or industrial purposes, been so great, as in these early years of the
twentieth century. The increasing realization of the effect of contaminating
materials, both organic and inorganic, on a municipal or private drinking supply'·
-and history's record of the devastating nature of epidemics due to water-
borne disease organisms, have led to the careful investigation of water for its
sanitary value and the development of materials and equipment to fight and
eliminate such contamination, and have increased many hundreds of per cent
the factor of safety to the public health. In like manner, and even to a greater
extent, has the value of water for industrial uses been a matter for careful
consideration. It is hard for the public to realize the immense quantities of
water used for industrial purposes, not only for the development of steam and
electric power, but also for purposes of manufacture.
From the standpoint of power developmert we are familiar with the heat
losses and the increased operating expenses due to scale formation in the steam
or locomotive boiler, and also to the continued rapid decrease in value of boiler
prqperty, or of power plant property, due to corrosion or rusting. Another
element of trouble which is noticed not so much in stationary boiler practice as
in locomotive boiler practice, is the element of foaming and priming of a water,
which results in much more rapidly putting the steam raiser out of active service
than either of the other ty'pes of trouble.
From the standpoint of plant deterioration due to rusting and corrosion cine
has only to look to the great mass of works on the corrosion of iron and steel
which are largely results of the growing need for some information as to the
cause and possible prevention of this particular phase pf trouble.
In the world of industry the action of a hard waten/upon soa'p consumption
has been known for centuries, and for a considerable period of time the value of
a water was determined largely by the amount of soap that it would consume
and render insoluble. This same hardness has a noticeable effect in the textile
industry, in bleaching and dyeing, in the canning industry, especially when the
water supply contains such, substances in large quantities. ,;_{n. the photographic
industry the presence of chlorides in water and certain alkalieS-is·a source of
considerable trouble, and in every case, before any intelligent effort can be made
to overcome these troubles, a complete analysis of the water is necessary.
It shall be our purpose in the methods which-follow to give, where it is pos-
sible, first a system of analysis whereby a complete analysis can be made, and t6
follow this up with optional methods which, individually, are equally as good
as those occurring in the system of analysis, and in some cases more satisfactory
where the laboratory has the required equipment, adding any special methq~s
whiLh may be found available. •
Chapter by D. K. French. /
2046
WATER ANALYSIS 2047
SANITARY ANALYSIS
PHYSICAL TESTS
For the physical examination, standards for turbidity and color have been
adopted.
Turbidity.-Turbidity standards are based on parts per ~illion of silica
(Si0 2) suspended in water, and the adopted standard is that of the United States
Geological Survey (A. P. H. A., p. 7). A water with a turbidity of 100 is one
which has 100 p.p.m. of silica (Si0 2) in such a state of fineness that a bright
platinum wire 1 mm. in diameter can just be seen when center of said wire is
100 mm. below the surface of the water and the observer is 1.2 meters above
the wire. The observation must be in open air, not in sunlight, and in the
middle of the day. Standards are preyared with precipitated fuller's earth (to
pass 200':mesh sieve). One gram to one liter of distilled water makes a stock
solution with 1000 turbidity. Standards for comparison are obtained by
dilution:
The Illinois Water Supply Association outlines another method (Proc.
1. W. S. A., 1914, pp. 49-51), whereby a suspension is prepared by shaking silica
2048 WATER ANALYSIS
(Si02) or fuller's earth (ground to pass a 200-mesh sieve), settling for ten hours,
and determining by evaporating and weighing the amqunt of silica (Si0 2) in a
given portion. Standards are then prepared by dilution.
Color.-All suspended matter should be removed b~ filtration. The stand-
ard designated as color 500 is obtained as follows: I
1.246 grams potassium platinic chloride (PtCI 42KCI) \1 containing 0.5 gram of
platinum and 1 gram of crystallized cobalt chloride (CoCI 26H 20) containing
0.25 gram cobalt (Co), are dissolved in water with 100 m!. hydrochloric acid and
made to one liter with distilled water. This solution is diluted with distilleq
water for comparative purposes, but a water with a color greater than 70 should'
be diluted prior to comparison. The standards for observation should be in
100-m!. Nessler tubes with the mark 20-25 cm. above the bottom and should
be viewed vertically downwards to a white reflective surface.
Standard glass disks are used by the United States Geological Survey 2 in
place of the above standard.
Odor.-Observations should be made both on cold and hot samples. Note
should be made immediately on opening containers as some odors are very
transient and rapidly disappear.
Cold.-Shake sample violently in collecting bottle, same to be about half
full. Remove glass stopper and smell at neck of bottle.
Hot.-Use either open beaker, 400 mI.; containing 150 mI. sample well
covered and heated nearly to boiling, or sealed glass stoppered bottle or saponi-
fication flask, heating fifteen minutes just under boiling. Allow to cool slightly,
remove stopper, shake and smell. Designate odor as aromatic, grassy, earthy,
musty, fishy, putrid, disagreeable, peaty, sweetish, etc. The following table
expressing intensity of odor is copied from the American Public Health Asso-
ciation Standard Methods, 1913, p. 12:
Numerical Term.
Value. Approximate Defiyition.
1 Care should be taken that this be the bright yello"l\;. platinic/salt,. and not con-
taminated with the reddish platinous salt. _, -
2 App. made by Builders' Iron Foundry, Providence, R; 1.
WA'fER ANALYSIS 2049
Taste.-May be made on hot and cold samples. A simple statement fol-
lowing largely the terms applied to odor in expressing results, brackish, as-
tringent, salty, sweetish, etc.
CHEMICAL TESTS
Free Ammonia
Apparatus.-The apparatus for this determination should be as far as pos-
sible free from joints or connections that are subject in any way to outside
contamination or to decomposition. The apparatus is composed of a distilla-
tio'n flask 'and a condenser, with possibly a safety tube located somewhere near
the flask to avoid the possible carrying over of impurities in cc;>nnection with the
steam.
Reagents: 1. Ammonia-free Water.
2. Standard Ammonium Chloride Solution.-Dissolve 3.82 grams of ammo-
nium cnloride in 1 liter of distilled water. Dilute 10 ml. of this to 1 liter with
ammonia-free water. 1 ml. =0.00001 gram of nitrogen.
3. Nessler's Solution.-Dissolve 50 grams of potassium iodide in the smallest
possible quantity of cold water .. Add a saturated solution of mercuric chloride
until a faint show of excess is indicated. Add 400 ml. of 50% solution of potas-
sium hydroxide. After same has clarified by sedimentation, make up to 1 liter
with water, allow to settle and decant.
Optional Method.-Dissolve 61.75 grams of potassium iodide in 250 m!. of
redistilled water, and add a cold solution of mercuric chloride which has been
saturated by boiling with excess of salt. Pour in the mercury solution cau-
tiously, and add an amount just sufficient to make the color a permanent bright
red. With a little practice the exact depth of color can be easily duplicated.
It will take a little over 400 m!. of the mercuric chloride solution to reach this
end-point. Dissqlve the red precipitate by adding exactly .75 gram of potas-
sium iodide. Then add 150 grams of potassium hydroxide dissolved in 250 m!.
of water. Make up to 1 liter. Mix thoroughly and allow the precipitate
formed to settle. Pour off the supernatant liquid. Mercuric chloride increases
the sensitiveness and potassium iodide decreases it.
Operation.-Clean apparatus thoroughly as follows:
Fill a flask, which for most satisfactory results should be an 800-ml. Kjeldahl
flask, with 500 ml. of distilled water. Add a pinch of c.p. sodium carbonate and
distill first of all with no running water in the condenser jacket until free steam
blows through the apparatus. Then turn on condenser water and distill off
approximately 250 mI., testing the last 50 ml. with Nessler's solution, and this
portion should not show color in fifteen minutes' time. The flask is then emp-
tied of the remaining water, 500 ml. of the water to be analyzed placed therein, .
and if acid, neutralized with c.p. sodium carbonate. A slight excess hurries the
..tgrlculrn'ttl f,n"" ..... ' Llbr.ary,
205oB apatla. WATER ANALYSIS
ammonia liben1tion but also tends to cause bumping. 'The distillation is then
started, distilling 6 ml. to 10 ml. per minute, and: three (separate portions of 50
ml. each are caught in Nessler jars. After 150 ml. is distilled the flame should
be. removed. !o each 50-ml._portion add 2 ml. Nessler'r solution and. after ten
mmutes' standmg compare with standards from the staudard ammOlllum chlo-
ride solution.
Organic Nitrogen
While this determination is not usually made we give it for the sake of
completeness.
The portion of sample from the free ammonia determination, or a new por::
tion freed from free ammonia by distillation, is acidified with 5 ml. c.p. sulfuric
acid (nitrogen-free) and digested in a hood until colorless and H 2S0 4 fumes are
given off. A little ignited pumice will guard against bumping. a Cool, dilute
with ammonia-free water, neutralize with 10% Na 2COa solution (NHa free),
distill into Nessler tubes and compare as for free and albuminoid ammonia .
.permanent standards 4 can be made by using potassium platinic chloride, .
2 grams dissolved in water, 100 ml. conc. hydrochloric acid and made to 1 liter;
and cobalt solution, 12 grams cob alto us chloride (CoCh' 6H 20) dissolved in
distilled water, 100 ml. conc. hydrochloric acid added and made to 1 liter. The
following table represents the .amounts used, to be made to 50 ml. with distilled
watedn Nessler tubes for comparison, the 150-m!. mark being 20-25 cm. above
thEl bottom, but should be checked against Nesslerized standards and t~e Nessler
solution modified, if necessary, until the standards agree. This is accomplished
by varying the amounts of potassium iodide and mercuric chloride.
Equivalent Volume
of Standard Platinum Solution, Cobalt Solution,
Ammonium Chloride, m!. mI.
m!'
3 KMn04 is sometimes used at this point, but as it may cause loss of N, the practice
is not advisable.
4 Permanent Standards (Jackson, 1'ech. Quart., 13, 320 (1900».
2052 WATER ANALYSIS
Nitrogen as Nitrite
Reagent: 1. Sulfanilic Acid.-Dissolve 8 grams oftHe acid in 1 liter of acetic
acid, specific gravity 1.04. This is practically a saturated solution and keeps
well. I
2. Naphthylamine Acetate.-Dissolve 5 grams of a-naphthylamine in 1 liter
of acetic acid, specific gravity 1.04, and filter through 'absorbent cotton (pre-
viously washed).
NOTE.-A slightly pink color resulting on standing does not materially interfere
with the use of this solution.
3. Sodium Nitrile Solution.-Dissolve 1.1 grams of silver nitrite in nitrite-
free water. Precipitate the silver with sodium or potassium chloride solution
and dilute to 1 liter. Dilute 100 ml. of this solution to 1 liter and then 10 ml. of
this second solution to 1 liter with sterilized nitrite-free water, adding 1 ml. of
chloroform and holding in a sterilized bottle. 1 ml. = 0.0001 milligram' of
nitrogen.
Operation.-Take 100 ml. of the water after filtration and clarificati~n,
preferably with aluminum hydroxide, to remoye possible suspended iron and
material which might interfere with color production. Add 2 ml. each of
solutions No.1 and No.2. After ten minutes' standing compare with standards
made up from the standard sodium nitrite solution (No.3).
PERMANENT STANDARDS
.0 , .0 .OOO~ • . ",
1.1 1.1 .001
3.5 3.0 .003 -
6.0 5.0 .005
12.5 8.0 , .010
The solutions to use for 100 m1. of water are the old ones, as fol~ows: 1 m!.
of hydrochloric acid (1 : 4), 2 m1. of sulfanilic acid (8 grams per liter), and
finally 2 ml. of naphthylamine hydrochloride (8 grams per liter with 10 ml. of
conc. hydrochloric acid), and allow color to develop twenty minutes.
NOTE.-Volume 28, page 742, J. Soc. Chern. Ind., calls attention to the possibility
of a permanent standard composed of a solution of acid..magenta (fuchsine-S, acid
fuchsine according to Weigert). According to this article, 0;2 of a gram of this dye
WATER ANALYSIS 2053
is dissolved in 50 ml. of 2/N HCl and made up to 2000 ml. with distilled water. Of
this solution 200 ml. are mixed with 50 ml. of 2/N HCl and again diluted to 2000 mI.
with distilled water. From this latter solution standard solutions can be prepared
containing various quantities, these standards being made up to 200 ml. with distilled
water after the addition of 5 ml. of 2/N HC}' Considerable work is being done on this,
but the standards have not yet been accepted in this country. However, standards
can be made by matching these solutions against standards prepared in the usual way
and their permanence is much greater than such standards.
N!trogen as Nitrate
ALUMINUM REDUCTION 5
Oxygen Consumed
Reagents: 1. Standard Potassium Permanganaie Solution.-Dissolve 0.4
gram c.p. salt in I liter of distilled water. I ml. is equivalent to 0.1 milligram
available oxygen.
2. Standard Ammonium, Oxalate Solution.-Dissolve 0.888 gram c.p.
ammonium oxalate in lliter of distilled water. 1 ml. is equivalent to 0.1 milli-
gram of oxygen. The standard permanganate solution ,must be standardized
against the ammonium oxalate solution.
5 Univ. of illinois Bull. Water Survey, Series 7, p. 14 (1909); Amer. Jour. Pub.
Hygiene, 19, 536 (1909). .
2054 WATER ANALYSIS
Chlorine as Chlorides
:Reagents: Standard Salt Solution.-16.48 grams fused c.p. sodium chloride
are dissolved in 1 liter of distilled water. 100 Illl. of this solution diluted' to
1 liter gives a standard solution, each mI. of which contains .001 gram of chlorine:
Standard Silver Nitrate Solution.-4.8 grams dried silver nitrate crystals
are dissolved in one liter of distilled water. Each ml. of this solution is equiva-
lent to approximately .001 gram of chlorine, standardized against the Standard
Salt Solution.
,/
INTERPRETATION OF RESULTS
The interpretation of the results of a sanitary water analysis is largely a
matter of experience, and it is impossible to lay down hard and fast rules cover-
ing this one matter. It is, however, possibJe to sum up the meanings of the
various determinations made, as each determination has some bearing upon the
sanitary condition.
In physical data the turbidity refers to insoluble matter in suspension. In
many cases, it is perfectly harmless, although less attractive, and frequently
2056 WATER ANALYSIS
Springs
Lake Streams. lind Deep Drift Deep Rock
Michigan. Shallow Well•. Wells.
Wells.
MINERAL ANALYSIS
Outline of Procedure
. Evaporate 50 ml. of the sample of water to dryness and bake the residue at 270 0
F. in weighed platinum dish. Increased weight Of dish represents
total solid residue. (Can be used for 804 when sample is small.)
Ignite for organic loss.
250 m!. Titrate with N/IO acid or alkali for alkalinity or acidity. (Can be
re-used to make up volume of 500-m!. portion when water sample is
small.) Methyl orange indicator. (With a 50 ml. sample use N/50
acid.)
100,m!. Titrate with NjlO AgN03 for chlorine.
2058 WATER ANALYSIS
100 ml. Acidify, boil, precipitate with BaCh, filter and weigh for total sulfate.
(Use filtrate for Na and K when necessary.) I
100 ml. Add 2 ml. 10% Na 2C0 3, evaporate to dryness, add phenolsulfonic acid,
dilute, then excess of NH 4 0H for total nitrate,
500 ml. Evaporate to dryness (with a few ml. ,concentrated HCI when very
accurate Si0 2 figure is necessary) in No.8. R. B. dish. Bake 30
minutes, cool, add boiling Hel (concentrated); dilute and filter.
Precipitate is Si0 2 Filtrate. Add a few drops of HN0 3 , concentrate to 50 ml., cool;
and silicate impu- add NH.OH, boil and filter.
rities (also BaSO.). 1 - - - - - - - - , - - - - - - - - - - - - - - - -
Unless great accu-
racy is necessary, Precipitate (Fe, AI, Filtrate. Boil and add saturated Am. Ox-
it should be POl') may be re- alate drop by drop, boil and filter.
weighed as such, ported as such or
as Fe and Al after 1----------------
otherwise Si0 2 can
be removed by HFa Qual. test for phos- Prec. Ca as oxalate. Filtrate Mg (and
and correction phate has shown Dissolvein2%H 2- Mn),addlO-25ml.
made. same to be absent. SO.. Titrate with concentrated Sod.
Otherwise both Fe standard KMnO.. Phos. Solution,
and PO.''' should then 50 ml. NH.-
be determined a n d , OH, stir well 2
weight corrected. minutes, or more,
let stand 4 hours,
or more, filter and
wash with. 3%
NH.OH. Dry in
air oven at 50°":'
60° C. Add excess
N/IO H 2SO. to dis-
solve ppt. Titrate
the excess acid with
N/IO NaOH.
NOTE.-For industrial purposes the original addition of HOI is not always necessary
and correction for BaSO., Phos. Mn and separation of Fe andjAI can be dispensed with
unless there is cause to suspect one to be present in material amounts.
acids, with suggested methods for manganese, ammonia, barium, and other
materials which might possibly be present.
Prior to the starting of the analysis, ·the physical characteristics of the water
should be noted, turbid waters should be filtered, the suspended matter analyzed
separately when necessary, and the amount determined either by filtration and
weighing of the separated material (alundum cones are very satisfactory), or
by the difference between two residues, one of which represents the original
water and one the filtered water. The mineral analysis should represent the
filtered supply. This is due to the difficulty of getting uniform samples with
suspended matter at different times. Treatment with HzO z is also sometimes
effective when the color is high.
NOTE.-Due to the tendency of some grades of H 20 2 to ~hange both alkalinity and
chloride figures, the material used should be examined carefully. Merck's Super Oxol
(30% H 20 2) has been found thoroughly satisfactory and has not affected analytical
results.
Silica, Iron, Aluminum, Calcium, Magnesium
Silica
Add 5-10 ml. HCI, evaporate over free flame, then on ~~in. asbestos board,
• to dryness, 500 ml. original water, using a No.8 'porcelain dish. Bake at
110°-130° C. or on asbestos plate over flame for one-half hour. Moisten with
10 ml. concentrated HCl, add 50 m!. of water, boil fifteen to thirty seconds and
filter. Wash with hot water.
NOTE.-For great accuracy, evaporate twice to dryness as above, with the addition,
prior to the sample going to dryness, of 10 mI. HCl, allow to bake as above, following from
there on the usual procedure for filtration.
NOTE.-When the water is highly colored due to. o.rganic matter, a few ml. o.f HNO a
can also. be added. e
The precipitate retained on the filter paper represents the silica or siliceous
matter, including possibly barium sulfate. Ignite and weigh.
NOTE.-If the amo.unt-is over .01 g. per liter, or 10 parts per million, moisten with a
few drops of concentrated sulfuric acid and hydrofluoric acid, expel excess acids, and
reweigh. This must be done in platinum. The loss represents silica, and should be
recorded as such, and the residue represents bases, principally barium, combineu with
sulfuric acid. This will also catch possible calcium sulfate that might be left undis-
solved, due to short bo.iling, to. low ililution, or conditions which would prevent its no.rmal
solubility in the original solutio.n. '
NOTE.-When a large amount of CaS04 is suspected or indicated, extra care must be
taken to remove it completely by solutio.n.
and concentrate to about 25-50 ml. Remove from ho~ plate or flame, add am-
monium hydroxide in slight excess, boil for one or two plinutes, and filter.
The precipitate contains iron, aluminum, and possibly phosphates. Burn
and weigh as oxides of iron and aluminum, plus phosphates, and test 50 ml. of
the original water with treatment in the usual way to determine whether or not
phosphates are present. Where this precipitate of iron Il-nd aluminum oxides is
greater than 0.01 g. per liter or 10 parts per million, or where the separation
of the iron and aluminum is advisable, the precipitate should be fused with
eight or ten times its weight of potassium bisulfate, redissolved in water, the'
iron reduced to the ferrous condition with zinc, and titrated with potassium
permanganate, recording the difference in weight between the original precipi-
tate and the iron determination as aluminum oxide.
NOTE.~Where much water is available and time is an object, an additional 500 ml.
can be carried down to approximately 50 ml. with a few drops of nitric acid, the iron
and aluminum precipitated as above mentioned with ammonia, and the precipitate
before drying redissolved in acid, reduced and titrated with potassium permanganate.
This portion can be started at the same time the original analysis is started, and win.
greatly simplify the determination and save time. .
Calcium
The filtrate from iron, aluminum and phosphate precipitate contains cal-
cium, magnesium, and possibly manganese. Concentrate to about 100 ml.
Add to the hot ammoniacal solution a saturated solution of ammonium oxalate
drop by drop, or add in small portions crystals of ammonium oxalate. Allow
to boil two minutes, stirring, if necessary (on account of heavy precipitate and
tendency to bump), remove, filter and wash. (Eight complete washings are
usually sufficient.)
NOTE.-Where great accuracy is desired, the precipitate on the filter should be
redissolved in a small amount of hot, dilute, hydrochloric acid and reprecipitated with
• ammonium oxalate.
Dissolve in 2% sulfuric acid and titrate with the standard solution of
potassium permanganate. (N /50 KMn04 may be used.)
NOTE.-Where the volumetric method is to be used, eight complete washings are,
as a rule, sufficient to remove the traces of ammonia salts, which are prone to have
considerable influence upon the volumetric results, due to the possibility of traces of
ammonium oxalate still being present.
6 J. Am. Chem. Soc., 23, 96 (1901).
7 J. Ind. Eng. Chem., 5, 301-2 (1913).
2062 WATER ANALYSIS
Optional (Gravimetric)
Instead of redissolving and titrating the calcium I oxalate upon the filter
paper, it can be dried, burned and weighed either as calcium oxide or calcium
carbonate. .
NOTE.-The burning of calcium oxalate to carbonate is not so difficult as it seems, as
an intense heat is necessary to convert it to the oxide, and if the crucible is well watched
and the flame gives just sufficient heat to carbonize and destroy the filter paper, there
will be no chance whatever of any calcium oxide being formed, or any calcium oxalate
being left. Where hypothetical combinations are used, it is very convenient to have the
calcium as carbonate without calculation. Where burned to the complete oxide it is
frequently necessary to use a blast lamp, as large precipitates require a high temPfature
to reduce completely to oxide form. '
Magnesium
The filtrate contains magnesium (and possibly manganese). Acidify with
HCl, concentrate if necessary to 150 mI., add 25 ml. saturated solution of
ammonium sodium hydrogen phosphate (NH4N aRPO,· 4R 20, microcosmic
salt), cool and make alkaline with ammonium hydroxide. Allow to stand at
least four hours. ..
The precipitftte is washed several times by decantation with ammonium
hydroxide solution (1 part NH,OR, sp.gr. 0.90 to 9 parts water), and finally
on the filter. After draining, the filter paper 'is opened out, the moisture
removed as much as possible by means of dry filter paper. The residue may ~e
dried in the room for about forty-five minutes or in the air oven at 50° to 60° C.
for fifteen to twenty minutes. When the filter has dried, ammonia will have
been expelled. Entire paper, etc., is placed in a dry beaker, N/I0 sulfuric acid
added in excess, using the methyl orange indicator, the solution diluted to 100
ml. and the excess of acid titrated with N lID sodiu~/hydro~ide.
One mi. N/lO H 2S0 4 =0.00216 gram MgO.
NOTE.-Accurate results are also obtained with the use of sodium phosphate added
direct to the 'filtrate from the calcium precipitate .'without previously acidifying with
acid, with 25 ml. to 50 m!. of ammonium hydroxide added to make strongly alkaline,
after which the solution should be very thoroughly stirred (for~a_t least two minutes),
using a rubber-ended glass rod. Allow to stand at least four hours. '-.~ _
For very rapid work in either case, if the magnesium solution after the precipitation
is cooled in ice-water, filtration can be frequently made in two hours' time.
For extremely accurate work the precipitate produced in either of the methods above
should be redissolved in a little dilute Hel and the precipitation repeated.
Optional (Gravimetric) l
Optional (Volumetric) I
. used in the general scheme of analysis because it starts with the raw water and
not a concentrated sample. A Mg determination can be made in half an hour
or less.
Solutions: (1) Lime Water.-A saturated solution is made by shaking an
excess of lime with a suitabl'e volume of water. After settling, the clear liquid
is siphoned into a bottle. Changes in temperature may cause slight precipita-
tion in the saturated solution. To avoid this it is a good plan to add 10-20 ml.
of distilled water per liter of solution.
(2) Standard Acid.-N/50 H 2S0 4 •
Procedure.-l00 ml. is measured into a 250 ml. volumetric flask, neutralized
with H 2S0 4 to methyl orange end.point. The liquid is boiled down to about'
50 ml. and to the hot solution 50 ml. of lime water added. The flask is filled
to. the mark with CO 2 free water, plus 5 ml. Place on a steam bath for 10
minutes; cool, allow to settle, pipette out 50 ml. and titrate with N /50 H 2S0 4
using phenolphthalein indicator.
The relation between the lime water and the standard acid is determined by
runl1ing a blank. This figure once obtained can be used until the lime solution
is all used. It is necessary to run blanks on aU freshly made lime solutions.
It is good practice to check the lime figure from time to time.
Blank-5XmI. acid used=ml. used to ppt. Mg.
This figureX2.4=p.p.m. of Mg.
Manganese
Where necessary, manganese should be determined separately in another
portion of the water and correctio~s made. The Knorres Persulfate method is'
the most reliable for large amounts (10 milligrams Mn per liter); the Bismuthate
method for smaller amounts.
KNORRES PERSULFATE METHOD (VOLUMETRIC)
Sulfates
• 100 m!. of the water is slightly acidified with conc. HCI and 5 m!. 10% NH 4CI
solution added, brought to a boil, and if turbid is filtered and washed four or ftve
times with boiling water. The clear or original water is now brought to ll: boil
and 10% barium chloride added drop by drop to the boiling solution in slight
excess. Boil ten minutes, stirring from time to time, if the precipitate is heavy.
Remove and allow to cool prior to filtering. The precipitate consists of barium
sulfate. Wash free from chlorides, testing with AgNOa• Dry, ignite and
weigh.
BaS04 X.411 = S04.
BaS04 X .583 = CaS04.
milligrams BaS04X.340 = CaS04 grains per gallon.
Lithium 10
Use a large quantity of the sample. Obtain the combined chlorides of
sodium, potassium and lithium. Transfer the combined chlorides to a small
Erlenmeyer flask (50 or 100-ml. capacity) and evaporate the solution nearly,
but not quite, to dryness. Add about 30 m!. of redistilled amyl alcohol.
Connect the flask, the stopper of which carries a thermometer, with a con-
denser 11 and boil until the temperature rises approximately to the boiling point
of amyl alcohol (130° C.), showing that all the water has been driven off. Cool
slightly and add a drop of hydrochloric acid to convert small amounts of
lithium hydroxide to lithium chloride. Connect with the condenser and con-
tinue the boiling to drive off again all water and until the temperature reaches
10 See Standard Methods of Water Analysis. A. P. H. A. (1920), page 60.
11 The amyl alcohol may be boiled off without the use of a condenser, but the vapors
are very disagreeable.
2066 WATER ANALYSIS
the boiling point of amyl alcohol. The content of the flask at this time is
usually 15 to 20 ml. Filter through a small paper- or ia Gooch crucible into a
gr~duated cylinder and note exact quantity of filtra,t~, which determines the
subsequent correction. Wash the precipitate with small quantities of dehy-
drated amyl alcohol. Evaporate the filtrate and wasHings in a platinum dish
to dryness on the steam bath, dissolve the residue in wa~er, and add a few drops
of sulfuric acid. Evaporate on a steam bath and expel the excess of sulfuric
acid by ge!ltle heat over a flame. Repeat until carbonaceous matter is cOrnf
pletely burned off. Cool and weigh the dish and contents. Dissolve in a sman
quantity of hot water, filter through a small filter, wash, and return filter to
dish, ignite and weigh. The difference between the original weight of dish and
contents and the weight of tile dish and small amount of residue equals the
weight of impure lithium sulfate. The purity 0:1' the lithium sulfate should be
tested by adding small amounts of ammonium phosphate and ammonium hy-
droxide, which will precipitate any magnesium present with the lithium sulfate.
Any precipitate appearing after standing over night should be collected on a
small filter and weighed as magnesium pyrophosphate, calculated to sulfate, and
subtracted from the weight of impure lithium sulfate. From this weight sub-
tract 0.00113 gram for every 10 ml. of amyl alcohol filtrate exclusive of the I
amyl alcohol used in washing residue because of the slight solubility of solid
mixed chlorides in amyl alcohol. Calculate 'lithium from the corrected weight
of lithium sulfate. Dissolve the mixed chlorides from flask and filter with hot.,
water, evaporate to dryness, ignite gently to remove amyl alcohol, filter and
thoroughly ,wash; concentrate the filtrates and washings to 25 to 50 ml.
To the weight of potassium chloride add 0.00051 gram for every 10 ml.
of amyl alcohol used in the extraction of the lithium chloride, which corrects
for the solubility of the potassium chloride ,in amyl alcohol. Calculate to
potassium.
The weight of sodium chloride is found by subtracting the combined
weights of lithium chloride and potassium chloride (corrected) from the total
weight of the three chlorides. Calculate sodium chloride'to sodium.
Alkalinity
In ordinary cases titrate with N /10 or N /50 H 2S0 4 , using methyl orange as
indicator. Special cases will be considered later.
Reagents.-Sulfuric a,cid, N /10. Methyl orange. PhcIlolphthalein.
Operation.___'::;250 ml. of 'water in 400-ml. beaker or a casserole are titrated
with N /10 H 2S0 4 , using two to five drops of methyl orange indicator (or 50 ml.
can be similarly titrated with·N /50 H 2S0 4).
To colored waters add 5 to '50 m!. H 20 2 to decolorize. Start with 5 mI., and
if this is insufficient, add more, 5 m!. at a time. This totally clears some waters;
in all cases there is a great improvement. In c6rrecting for the acidity of
H 20 2, our laboratory adds 0.1 m!. to the reading for each 5 ml. H 20 2 used.
Calculate for 250-ml. sample.
No. m!.X4X.005=g. per liter CaCOa• I
No. m!. X 4,.X 58.4 X .005 = grs. per g!lJlon CaC03 •
Or X 1.168 = grs. per gallon ·GaGOa•
WATER ANALYSIS . 2067
P=O .............. . M o o
P<!M ............ . M-2P 2P o
P=!M ............ . o 2P o
P>!M ............ . o 2(M-P) 2P-M
P=M ........... ,. .. o o M-
tion of any titration method is complicated by the fact that when salts such
as carbonate, phosphate, aluminate, silicate or chromll-te, as well as some. types
'of organic material, are present, they are titrated to a greater or less extent
along with the hydroxide ion. The standard methqd, employing strontium
chloride, minimizes interference by sulfate, as well as; by carbonate and phos-
phate. A method similar in procedure, but employing barium chloride, is
'outlined in the appendix following the standard method.
STANDARD METHOD 13
..
Evolution Method 14
hydroxide :in flask E shall then be titrated with standard 0.02 N hydrochloric
acid from burette B to the endpoint of phenolphthalein. The procedure out-
lined also serves to calibrate the apparatus against the air in the system.
Indicator~(A) Phenolphthalein indicator solution. 5 g. of phenol-
phthalein cHSsolved in 1 liter of 50% ethyl alcohol. (B) Methyl orange indi-
cator solution. 1 g. of methyl orange dissolved in 1 liter of distilled water.
Procedure; Manipulation.-(A) 50 mi. of N/50 barium hydroxide solution
shall be measured into fll\sk E from burette A, and flask E shall then be
secur~ly attached to the rubber stopper on burette B.
(B) A 200-ml. sample of the water to be tested, containing not more than
100 p.p.m. of total carbon dioxide, shall then be placed in flask D, phenolphthal-
ein indicator added, and the flask attached securely to the rubber stopper of
burette C. The hot plate shall be put into position. The three-way cock
shall then be opened to the atn:lOsphere and the pump startt!d. When bubbles
appear in flask D the three-way cock shall be adjusted to circulate gas through
the closed system comprising flasks D and E.
(CI) With the gas circulating (to stir the solution), N/IO sulfuric acid shall
be added from burette C to the sample in flask D until the phenolphthalein
endpoint is reached. The amount of acid used shall be recorded.
(D) Two drops of methyl orange' solution shall then be added to flask D and
the titration continued to the methyl orange endpoint. The amount of acid
used shall be recorded and then 20 ml. of acid added in excess.
, (E) With the gas still circulating, the solution in flask D shall be heated
until it just reaches the boiling point and maintained at this point for about
5 minutes. The hot plate shall tl~en be turned off and the circulating pump
stopped.
(F) The solution in flask E shall then be titrated with N /50 hydrochloric
acid to the endpoint of phenolphthalein with,constant swirling of the flask.
Calculation.-The results shall be calculated as follows:
(A) The concentration of total carbon dioxide in parts per million of CO 2
shall be calculated from the expression:
total CO 2 = (A - B) X2.2 XI
where A = volume of 0.02 N acid equivalent to volume of N /50 Ba(OH)2 used,
in mi.,
B = volume of N /50 acid used, in ml.,
f = correction factor for N /50 acid.
(B) The concentrations of carbonate and bicarbonate ions in parts per
million of CO a shall be estimated by means of the curves in Fig. 290. To use
these curves, which express the respective fractions of the total carbon dioxide
present as undissociated carbonic acid, as bicarbonate ion, and as carbonate ion,
the pH of the original sample must be known. From the determined value of
pH, the value of the fraction of the total carbon dioxide present as carbonate
ion, as bicarbonate ion, and as undissociated carbonic acid, shall be read re-
spectively from curves a, band c, of Fig. 290. The concentrations of carbonate
ion, bicarbonate ion, and undissociated carbonic acid in terms of parts per
million, shall then be calculated from the expression
2072 WATER ANALYSIS
C03= =(A-B)X3X!Xa
HCO a =(A-B)X3.0·5X!Xb j
H 2 C0 3 = (A - B) X3.1 X!Xc
where a, b, and c are the fractions determined respectiveLy from curves a, b, and
c of Fig. 290 and the other symbols have the meaning defined above.
I
N oTE.--8ince the preceding method of estimating ion concentrations is based upon
data for the ionization of carbonic acid at .25 0 C., the values obtained hold for this'
temperature only. At any other temperature the ion concentrations will have the other
values, which may be calculated if the dissociation constants of carbonic acid at this
temperature are known.
1.0
--- , h
0.9
\ / /
.. 0.8
c\
\ I~ \ IIi)
:!!
II
QO.7
\
b\ 7
c:
-e o~6 \7 . I
U
-;;; 0.5
V J
~
._ 0.4
/\
~\
o
50.3 Y
1:;
..tOo2
\
aJ \
0.1
bf \c / ~b
V , !/
"',,-
\
\
../ <'
o 2 3 " 5 6 7 8 9 10 /I 12 13 JIj.
PH I
FIG. 290.-Fraction of Total Carbon,Dio:X:i.de as pH.
Accuracy.-This method will yield results for total carbon dioxide accurate
to 5% for concentrations as low as 10 p.p.m. of total CO 2 ,
Acidity
For acidity use NjlO Na 2CO. and 250 ro!. water.
Hydrogen-Ion Concentration
It is freqUently desirable to know, particularly in connection with waters
for certain industrial uses and the ·filtration of supplies using alum as a coagu-
lant, the actual reaction of the water in terms of the hydrogen-ion concentration
in addition to the familiar titration with methyl orange which gives the total
available alkalinity of the water and does not indicate the actual dissociation
or pH value. The proper adjustment of the pH value along with treatment
using alum as a coagulant is necessary for economical and efficient operation
of this process. IS In the use of water for steam making and for certain manu-
facturing processes the added data given by this determination are making pos-
sible more accurate adjustments and more advantageous selection of supplies
and treatments.
The most accurate method for making these measurements is by means of
the, hydrogen electrode; 16 however, colorimetric methods have proved satis-
factory for use with waterP The following standard buffer solutions proposed
by Clark and Lubs 18 together with their recommendations for the preparation
and use of indicators make a convenient set of standards for this purpose,
using 10 m!. of the standard buffer solution with five drops of the indicator
compared with the sample used in the same proportions. .
N/20 NaOH per
Indicator pH Range Decigram
Methyl red, ................... , . " 4.4....u.0 7.4
Brom thymol blue .................. 6.0-7.4 3.2
Cresol red .... , ............. , ...... 7.2-8.8 5.3
Thymol blue ............... : ....... 8.6-10.0 4.3
For the preparation of these solutions one decigram of the dry powder is
ground in an agate mortar with the corresponding quantity of N /20 N aOH.
When solution is complete, the thymol blue and brom thymol blue should be
diluted, to a 0.04% solution and the cresol red and methyl red to 0.02% solution.
The American Petroleum Institute 19 recommends the following set of
standards.
Name pH Range Color Change
Brom phenol blue. . . . . . . . . . . . . .. 3.0-4.6 Yellow-blue
Brom cresol green ............... 4.Q.-5.6 Yellow-blue
Chlor phenol red ....... : ........ 5.2....u.8 Yellow-red
Brom thymol blue ............... 6.Q.-7.6 Yellow-blue
Phenol red ..................... 6.8-8.4 Yellow-red
Cresol red ... " ................. 7.2-8.8 Yellow-red
Solutions: Mj 5 Acid, Potassium Phthalate.- Make up a solution of potas-
sium hydroxide dissolving about 60 grams of a high grade sample in about 400
15 J. Ind.'Eng. Chern., 14;395 (1922), 1038: Chern. Met. Eni' May 3,1922, Vol. 26,
No. 18; J. Am. W. W. Ass'n, Vol. 11, No.3, May, 1924.
16 Clark, The Determination of Hydrogen Ions (Williams.an Wilkins, 1920).
17 See also Index for additional discussion on colorimetric pH determination.
18 Clark (see note 2). .
19 Code of Recommended Practice for the Treatment of Oil Field Waters. Report
of Sub-Committee "C." Frank M. Beeson, Chairman. D. & P. E. Bull. No. 204, p. 8.
American Petroleum Institute. Jan. 2, 1930.
2074 WATER ANALYSIS
6.0 50 ml. M/5 KH 2P0 4 5.70 m!. M/5 NaOH Dilute to 200 m!.
6.2 50 ml. " 8.60 m!. " " " " "
6.4 50 ml. " 12.60 m!. " " " " "
6.6 50 m!. " 17.80 ml. I " " " " "
6.8 50 ml. " 23.65 ml. " " " "
"" ... .._,!
7.0 50 m!. " 29.63 m!. " " "
7.2 50 m!. " 35.00 m!. " " " " "
7.4 50ml. " 39.50 m!. " " -Ii
" "
7.6 50 m!. " 42.80 ml. " " " " >
7.8 50 ml. " 45.20 ml. " " " " "
8.0 50 ml. " 46.80 ml. " " " " "
I
BORIC ACID. KCI-NaOH MIX~~RES
7.8 50 m!. M/5 H 3B0 3, M/5 KCI 2.61 m!. M/5 NaOH Dilute to 200 ml.
8.0 50m!."
8.2 50 ml.
"
"
3.97 ml.
5.90 ml.
" ,,'" "" "" ""
8.4 50 ml. " " 8.50 m]. " I"
" " "
8.6 50 ml. " " 12.00 ml. II'
_.- " " " "
20 Clark, The Determination of Hydrogen Ions (WilliamS and Wilkins, 1920).
WATER ANALYSIS 2075
8.8 50 ml. M/5 HaBOa, M/5 ReI 16.30 ml. M/5 NaOH Dilute to 200 m!.
9.0 50 m l . " " 21.30 ml. " " " " "
9.2 50 m!. " " 26.70 ml. " " " "
9.4 50 ml. " " 32.00 m!. "I< . " " "
9.6 50 ml. " " 36.85 m!. " I<
" I<
Chlorine
Titrate 100 ml. of water, using 1 ml. qf 10% potassium chromate as an indica-
tor, with N /10 AgNO a• to first permanent indication of the red silver chromate.
(Acid waters should be neutralized and sulfide waters boiled with a drop or so of
nitric acid and then neutralized for reliable results.)
21 For criticisms of this method see Z. Nahr. Genussm., 24, 429 (1912); also C. A.,
7, 38 (1913). .
2076 WATER ANALYSIS
Nitrates
(Also see Sanitary Analysis)
.
PHENOLSULFONIC ACID METHOD
Hydrogen Sulfide
Due to the fact that hydrogen sulfide is frequently very transient and often
oxidizes to sulfate in transit, it is advisable to collect this sample in a special
container at the point of sampling. Two or three bottles holding exactly 250
m!. of water each, are used, each bottle containing 50 mI. N 1100 iodine solution.
Mter filling, the bottle is sealed. ,The s~mple is titrated with standard N /1~0
sodium thiosulfate upon receipt at laboratory, at which time a blank is run,
using 50 m!. iodine solution made to mark with distilled water. The difference
between the titration of the sample and the blank repreaents hydrogen sulfide
present.
Iodine valueXO.1263X4=sulfur value grams per liter.
When hydrogen suHide is determined on sRmples Rlre8dy collected 8nd sub-
mitted, 100 m!. of the sample is oxidized with bromirte (about 10 m!. is suffi-
cient), HCI is also added, and it is then,boiled until colorless, after which sul-
fates are precipitated with BaCh as already described. The difference between
the oxidized and unoxidized sulfates is calculated to E: 2S.
Oil
Frequently waters from condensing engine, or after passing heaters or oil
separators, still contain oil in small quantities. The following method has
been found most satisfactory:
Reagents: Ferric Chloride Solution.-(10 grams ot iron dissolved in 200 ml.
HCI, oxidized with HNO a and made to one liter.) I
Ammonia c.p. I
taining the oil. Three extractions, as direhted above, will remove all oil from
the water. The solvent is expelled by warming gently and the oil is weighed.
Dissolved Oxygen
Use the Winkler Method 24
Reagents: No. 1. Manganous SUlfate Solution.-48 grams manganous
. sulfate dissolved in 100 m!. distilled water.
No.2. Iodide Solution.-36o. grams NaOH and 100 grams KI dissolved in 1
liter of distilled water ,
No.3. Concentrated hydrochloric acid, or sulfuric acid sp.gr. 1.4 (dilute
1 : 1).
No.4. Sodium Thiosulfate Solution.-Nj100 solution is made as needed
from the Nj10 stock solution.
NOTE.-Not permanent; should be frequently restandardized against N/100 potas-
sium bichromate.
The addition of 5 mI. of chloroform plus 1.5 grams sodium or ammonium carbonate
to Ililch liter of solution on mixing will improve its keeping quality.
N/40 sodium thiosulfate containing 6.2 grams C.P. recrystallized salt per liter may
be preferred to N/100 strength. 1 m!. of this solution is equivalent to 0.2 milligram
oxygen by weight or 0.1395 mI. oxygen by volume, standard conditions.
5. Starch Solution. 25__The starch should first be made into a thin paste with
cold water and about 200 times its weight of boiling water stirred in and ,boiled
for a few minutes. A few drolls o.f chloroform will assist in preserving this
solution.
Collection of Sample.-A small-necked, 250-m!. bottle should be used, etched
or o.therwise marked, with its exact volume previously determined. The collec-
tion should be so arranged to. exclude outside air and result in several co.ntinuous
changes o.f the co.ntents before sto.Ppering, care being taken to exclude air
bubbles.
Operation.-To sample as received add, in both cases by pipette, delivering
below surface of water and away from the air, 2 m!. solution No.}. (manganese
sulfate) and No.2 (NaqH, KI). Restopper and shake thoroughly. After pre-
cipitate has settled add 2 ml. Hel or H 2 S0 4 and again mix by thorough shaking
until precipitate has completely dissolved, transfer 100-m!. to flask, and titrate
with solution No.4 (sodium thiosulfate), using starch as indicator near end as
the color approaches a faint yello.w. . •
N = mI. N /100 thiosulfate solution. .
V = capacity of bottle less 4 mI. (vol. sol. 1 and 2 added).
O=the amount of oxygen in parts per million in water saturated at the same tem-
perature and pressure.
24 Ber., 21, 2843 (1888). Also Z. Ap.a!. Chem., 53, 665-72 (1914); C.F.C.A., 8, 674
(1915). . '
25 Hale gives the following method. "Rub 5 grams of potato starch with cold water
to a thin paste together with 10 milligrams of mercuric iodide. Pour into one liter of
boiling water and boil half an, hour." ·
2080 WATER ANALYSIS
·
(1) Oxygen ni 0.0008N X 1,000,000 81\T .26
p.p.m. 1 0 0 ' ... ~
(2) Oxygen in m!. per liter =.7 oxygen p.p.m.
· Oxygen p.p.m. (observ~d temp. and pres.) X100
(3) 0 xygen per cent sat urat IOn = . ' .
SaturatIOn oxygen p.p.m! (observed temp. and ;pres.)
26
.
ards. This gives copper .
Correcting for displacement for 300-m!. bottle, .8N = .&l1N; for 275-m!. bottle,
.8N=.812N.
No correction for displacement affects result .1 p.p.m._Qxygen:
Twenty-five m!. variation in capacity of bottl!J..a:trects result :01 p.p.m. oxygen.
The above formulre are based upon N /100 thiosulfate, and titratmg 100 mi. volllme.
N = milliliters thiosulfate used.
WATER ANALYSIS 2081
The solution from the above is nearly neutralized with ammonia. It is
then concentrated and 2 grams potassium oxalate and 1.5 grams p<?tassium
sulfate are added and the zinc removed by electrolyzing. (Three hours with
0.3,ampere current.) This gives zinc.
Where copper only is desired it is frequently sufficiently satisfactory to con-
centrate the water from 50 ml. to 75 ml., after which it is acidified with 2 to 5 ml.
concentrated H 2S0 4 , depending upon whether the water is very alkaline with
carbonate of lime, etc., and then the procedure.for copper is followedP
There is no satisfactory method for the quantitative determination of small
quantities of tin. In the above-mentioned procedure, however, in case tin
should be present it would be removed with the ammonia precipitate for the
removal of iron and its presence may be avoided by dissolving the sulfides in the
original
, precipitation
. in HNO a, in which the tin would remain behind insoluble.
HARDNESS
Total Hardness
The most accurate method for total hardness is by calculation of the calcium
and magnesium determined gravimetrically as previously outlined, calculating
the calcium as calcium carbonate and the magnesium to its calcium carbonate
equivalent in terms of parts per million. 28 However, where only the hardness is
desired, gravimetric methods are cumbersome and the following are accepted as
standard.
The standard method for the determination of total hardness, as well as
temporary and permanent, depends upon the action of the lime and magnesia in
solution upon ,!oap, the soap ,added in a very dilute solution in alcohol. Total
hardness represents the total soap acted upon by the water in its original state,
permanent hardness represents the total soap acted upon by the water after the
water in question has been thoroughly boiled and separated from the suspended
matter, and temporary hardness represents the difference between the total
hardness and the permanent hardness, and while it is supposed to represent
combined carbonates of lime and magnesia, and the permanent hardness is
supposed to represent lime and magnesia in other forms than carbonate, this i1l
rarely so due to the fact that a certain material amount of carbonate of lime
and magnesia is soluble in water, even in the absence of carbon dioxide gas.
The hardness 29 is considered entire1y due to Ca, when the amount of soap
used is calculated. The errors due to Mg and Fe present are not .considered; as
a rule the blank on distilled water equivalent to the amount of sample is not
considered.
Preparation of Solutions: The reagents used are standard soap solution and
27Phelps, J. Amer. Chern. Soc., 28, 369 (1906).
C. Bahlrnann, J. Ind. Eng. Chern., 6, 209, 11 (1914).
28 •
29 Errors in the Clark Method for Determining Hardness, George G. Town, J. Am.
Water Works Assoc., 21, 1395 (1929).
2082 WATER ANALYSIS
standard calcium chloride solution, the latter being made under such conditions
that 1 ml. of the sofution is equivalent to 0.2 milligraIl} of calcium carbonate.
Standard Calcium Chloride Solution.-0.2 gram pure calcium carbonate is
dissolved in ~ small amount of dilute I!Cl, taking pai~ to avoid any loss due
to effervescence or spattering. Evaporate the solutIOn to' dryness several
times to remove excess acid. Dissolve in distilled water1and make up to 1 liter.
Standard Soap Solution. is obtained by dissolving approximately 100 grams
dry Castile soap in 1 liter 80% alcohol. This solution should stand several days.
For standardizing this solution should be diluted with alcohol (70% to 800/""), .
until 6.4 ml. when added to 20 ml. of standard calcium solution will produce\.
permanent lather. Usually less than 100 m!. of the original'soap solution will
make 1 liter of standard solution.
For plant control, and to simplify calculations, 58.3 ml. is frequently the
quantity taken, and the soap solution standardized so that 4.5 ml. =20-ml.
standard CaCh; a correction by subtracting 0.5 from all readings is made, a~d
in the resulting figure 1 ml. will be equivalent to 1 mg. CaCOa• If the standard
soap solution is used and the figures to read in grains per gallon are desired,
take 41 ml. of sample and subtract a correction of 0.7 ml. from the reading
obtained. .
For standardizing, use 250-ml. glass-stoppered bottle, add 20 ml. calcium
solution with 30 ml. distilled water. The soap solution should be added' from
a burette, approximately .2 ml. at a time, after which the bottle is shaken
vigorously until the lather formed remaIns unbroken for five minutes after
shaking and after the bottle has been placed upon its side.
NOTE.-Pure potassium oleate may be used in place of soap.30
Operation.---,Fifty ml. of the water in question are measured into a 250-)lll.
bottle, the soap solution added, approximately .2 ml. at a time, and in the same
manner as described for the standardizing of said soap solution. The following
table, copied from p. 33, Standard Methods of Water Analysis, A. P. H. A., 1913,
may be used to obtain ,the total hardness from th\"l reiults so noted:
m!. of Soap 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Solution. m!. Ill!. m!. m!. m!. ml. m!' ml.'"-
-~. ml.
- -- - - -- -- -- -
0.0 ... . ... . ... . ... . ... . . . ... 0.0
'" 0.6 3.2
1.0 4.8 6.3 7.9 9.512.7 14.3
11.1 15.6 l(j~9 18.2
2.0 19:5 20.8 22.1 23.426.0 27.3
24.7 28.6 29.9 31.2
--- -- - ----
-~
3.0 32:5 33.8 35.1 36.4 37.7 38.0 40.3 41.6 42.9 44.3
4.0 45.7 47.1 48.6 50.0 51.4 52.9 54.;3 55.7 57.1, 58.6
5.0 60.0 61.4 62.9 64.3 65.7 67.1 68.6 70.0 71A 72.9
- -- -- -- -- -- -
6.0 74.3 75.7 77.1 78.6 80.0 81.4 82.9 84.3 85.7 87.1
7.0 88.6 90.0 91.4 92.9 94.3 95.7 97.1 98.6 100.0 101.5
.
• 30 C. BIacher, Chern. Ztg., 36, 541 (1912); J .. Soc. C~m;:-Ind.! 31, 555 (1912); C.A.,
7, 1394 (1913); C. BIacher, P. Gruenberg, M. Kissa, Cliern. Ztg., 37, 56-8 (1913); C.A.,
r 7,1938 (1913). L. W. Winkler, Z. Anal. Chern., 53, 409-15 (1914); C.A., 8, 29!2 (1914).
WATER ANALYSIS 2083
It is not desirable to use more than 7 m!. of soap solution for 50 ml. of the
water, and when the figures are higher, the water should be diluted with distilled
water. The reading in the table corresponding to the ml. of soap solution
used is then multiplied by the quotient 50 mil.; x m!. being equal to the amount
xm.
of water taken. In making this determination there is frequently noted a false
end-point sometimes known as the magnesium end-point. To avoid error, it
is advisable, after completing the titration, to read the burette, add 0.5 m!.
more of the soap solution and shake well. If magnesium has been responsible
for the false end-point, after such addition the later will again disappear, and
titration should be continued until a new and true end-point is reached. 'It is
advisable to determine the strength of the soap solution from time to time, as it
is very prone to change upon standing. Results should be recorded in terms of
calcium carbonate, parts per million. There are various other means of report-
ing. The English degree frequently noted as Clark degree, represents grains
calcium carbonate per Imperial gallon and should be multiplied by 14.3 to give
parts per million. Conversely, the result obtained in parts per million divided
by 14.3 will give Clark, or English degrees. French degrees represent parts
per 100,000 calcium carbonate and should be multiplied by 10 to give parts per
million. Conversely, division of the result obtained above by 10 will give
French degrees. German degrees represent parts per 100,000 calcium oxide
and should be multiplied by 17.8 to give parts per million calcium carbonate.
The determination of hardness is not reliable on account of the varying action
of calcium and magnesium salts, find should never be resorted to when possible
to determine these bases direct.
NOTE.-Dr. Hale claims that the soap method for hardness in skilled, hands is
accurate from 10 to 15 parts per million on waters as hard as 300 parts.
For permanent hardness the standard soap solution is used as above stated.
The water, however, is boiled gently for one-half hour, allowed to cool, made to
volume with boiled and cooled distilled water and filtered, after which the above
method is used. The difference between total hardness and permanent hard-
ness is supposed to represent temporary hardness. The alkalinity determina-
tion given on 11 previous page is a much more accurate method of determining
temporary hardness, however, and is also much more easily carried out. When
total hardness and alkalinity are determined, permanent hardness would be
the difference between these two figures. For comparative use as against total
and permanent hardness determined as such, the results would be much differ-
ent, as the alkalinity determination of all the carbonates would give a permanent
hardness representing absolutely non-carbonate hardness; whereas the deter-
mined permanent hardness would contain a material amount of combined
carbonates of lime and magnesia. The American Public Health Ass~ciation,
Committee on Standard Method of _Water Analysis, recommend that the
determination of permanent hardness by the soap solution be discontinued in
connection with softening process, as it is so unsatisfactory in general practice.
Magnesium Chloride
Frequently, when hypothetical combinations are used it is desired to check
up these calculated combinations by some chemical method. Magnesium
2084 WATER ANALYSIS
I
LIME AND SODA VALUE ill
Two very simple methods have been devised for the rapid estimation of the
amount of lime and soda-ash necessary for softening, when water treatment is
considered from the outside softening-plant point of view.
(4a-5b) X3.51 CaO will give milligrams of lime per liter required to soften
the water tested.
Value for Soda
To the neutralized 200 ml. from above titration, add 20 ml. N /10 sodium
carbonate. Heat to boiling. Tran.sfer with CO 2 free distilled water into a 250-
m!. flask to make up to mark with' was}lings from the dish; mix thoroughly and
filter, collecting 200 m!. of the filtrate in a beaker. Titrate with N /10 hydro-
chloric acid for the excess alkali. Designate the number of ml. in this titration
by c.
Formula: (20-b-ic)X33.13 Na 2C0 3 =milligrams of soda per liter neces-
.
sary to soften water in question .
NOTE.-Both formulre are based upon C. P. chemicals, and corrections must be
made for the value of the commercial materials in use. These methods are valuable,
though for actual practice it- is advisable to tryout on a liter of water in question,
u,sing the calculated amounts of lime and soda for experimental purposes.
. Reaction Reaction
Positive Radicals. Coefficients. Negative Radicals. CoeffiCIents.
Using the above table, the parts per million of each radical multiplied by its
reacting coefficient will give its reacting value, and in the formulas which follow
this will be indicated by T prefixed to the chemical symbol of the nidical.
For checking the accuracy of the analysis, the sum of the positive reacting
values should equal the sum of the negative reacting values, and the formula,
g. E h The vaIue 0 I t h'IS
10 OrPos.-rNe N ' t e percentage error 0 I t h e anaiYSIS.
'
r Pos.+r ego •
errOr should never exceed 5 for waters of 100 p.p.m. or morC? dissolved solids, and
should be 2 or less. .
In ordinary analysis, silica, iron and aluminum are present in such small
quantities that they may, for simplicity's sake, be ignored. The following
formulas are given without comment, as full details can be found in Water
Supply Paper No. 274.
Water Soltening
For 1000 Gals. Water.-Pounds lime (90% CaD) required .
=0.26(rFe+rA;+rMg+rH+rHC03+.0454 CO,).
Pounds soda ash i (95% N a,CO,) required
= 0.465(r Fe+r Al +rCa+r Mg+rH -rCO, -r HCO,).
NOTE.-Dr. Hale states the following. If Instead of the extended formulre of the
ionic system, I much prefer my' simple fo.rmulae.
(AlkaiinityX.44+free CO,)X.0106=lbs. CaO per 1000 gals. feed water.
Also (Total hardness-total lime) X.0047 =lbs. CaO per 1000 gals. feed water.
or total magnesia as CaC0 3 •
(Total hardness-alkalinity)X.OO9=lbs. Na 2 CO a per 1000 gals. feed water.
Expressed as Ca003." I t
Corrosion
For Acid Waters
Coefficient of corrosion C=1.008(rH+rAI+rFe+rMg-rCO.-rHCO.).
Hypothetical Combinations
,
The use of hypothetical combinations in the reporting of a mineral water is
frequently of value, in that it gives a more rapid way of placing in simpler terms'
the principal materials present in the water. It should never be assumed from
the hypothetical combinations that the materials so reported are present in the
water in that particu1&1' form, but it is assumed by most, th~it the form in which
such materials are reported will represel).t the condition in which those materials
will combine, when the water.is subjected to increased pressure and iqcreased
temperature. In other words, the hypothetical combinations most generally in
use represent the way materials will appear when combined, due to the law of
mass action under steam-boiler conditions.
! i For such purpose the method which takes care of the insoluble materials or
¥laterials leaying the water first is the most common method in use. This
method combines as follows: Where the sum of the sulfate and carbonate radi-
cals exceeds that of lime and magnesia as bases, the magnesia is first calculated to
carbonate'" the remaining carbonate.is combined with lime, the remaining lime
~'ith sulfate and the remaining sulfate with sodium. This also takes care
of the general condition where the carbonates alone are in excess of the com-
bined carbonates of lime and magnesium, in which case the remaining carbonate
naturally would be calculated to sodium and all the sulfate, as well as chloride,
also, calculated to sodium: Where, however, the sum of the bases is greater
than the sum 'Of the carbonate and thfi suJfate radicals, two possible conditions
ali Can be omitted or ignored unless suspended matter, silica, etc., are present -in
large quantities. .
2088 WATER ANALYSIS
which are slowly clearing up some of the fog which has handicapped the under-
standing of what goes on in the boiler. Research work relative to solubilities
at high temperatures and pressures, to the proper working of mass action in
boiler waters have established certain fairly definite rules the observance of
which can result in a fairly general freedom from boiler water troubles. The
methods from which information as to the water's condition would be gained,
must be methods calling for no special agents such as intense heat, and fall
into the classification of volumetric or turbidimetric methods. In the power
plant the presence of excesses in the boiler water of sodium carbonate, caustic
soda, the phosphate radical, the sulfate radical, and chloride radical have neces-
sitated c\lntrol tests which must be approximately accurate, although they are
not tests that would compare with the methods already given, which should be
considered standard. In the determination of alkalinity, both caustic and
carbonate, there has been no great difference from the standard method. The
use 9f the two indicators, phenolphthalein and methyl orange, is most general,
but the _guantities tested for the most rapid results are 58.3 ml. or a fraction
thereof, and the acid used is N /50. The results read from the burette are
directly in 'grains per gallon. Where other interfering agents, like phosphates,
are present, the first titration using phenolphthalein is frequently modified by
adding a small amount of 10% barium chloride, and then without filtering, and
in the presence of the precipitate, a titration is made to the disappearance
of the phenolphthalein red. For the determination of chlorides the regular
chlorine method is used, but. due to the fact that the sample is fre-
quently alkaline with caustic soda, the sample must be neutralized before
carrying out the titration. 'Plis chloride test gives indirect rather than direct
information. It is used to watch boiler concentration on the assumption that
the chlorides will concentrate uniformly regardless 'of treatment for calcium
and magnesium impurities, and their increase over the amount present in the
original supply will bear some sort of relationship to other constituents in the
boiler water. As a rule additional work must be done through a competent
laboratory so that several chlorine determinations can be compared in their
relationship to total solids and complete boiler water oomposition. From this
information a factor can usually be obtained for use with results of later control
tests. As the alkalinity would be variable the mathematics would be as follows:
Total Chlorine Determination as Sodium Chloride X F + M = Total Solids,
where F represent!? the factor of relationship between chlorides and concentra-
tion, and M represents methyl orange or total alkalinity.
, There is a great deal of publicity and interest in boiler water conditioning
using meta-phosphates, and the presence of slight excesses of phosphate is
important so that rapid methods have been worked out.
There are two, one a modification of a colorimetric method, the other a
modification of the molybdate precipitation, both of which are of interest. Th~
first method is that recommended by F. G. Straub. It is a modification of a
very rapid- and delicate colorimetric method, and aside from requiring five
solutions, one of which is standard, and two of which must be carefully cared
for, is available for field or boiler room control. The method follows;
2090 WATER ANALYSIS
Determination of Phosphate 36
Measure 50 ml. of the filtered sample of boiler water into a 250 ml. vol-
umetric flask, and add 10 ml. of sulfuric acid.! To the flask 5 ml. of molybdic
acid are added,followed by 5 ml. of hydro quinone solution. After five minutes,
15 ml. of the carbonate-sulfite solution are Jtdded, mixed and the flask made up
to volume.
'Measure 10 ml. of standard phosphate solution by a pipette into a 250 mI.
volumetric flask and 20 ml. into a second flask; 40 ml. of distilled water are
added to the first and 30 to the second, followed by the procedure in the
determination of ph9sphate.
Place the flask containing the sample of boiler water between the two flasks
containing the standard solutions and compare colors. If the color is lighter
than the flas1i containing the 10 ml. it contains less than 20 p.p.m. of P0 4.
If it is between the one containing the 10 ml. and the lone containing 20 ml.,
it has between 20 and 40 p.p.m. If it is darker tlian the one containing 20 ml.
it contains over 40 p.p.m. '
Different maximum and minimum values may be established by changing
the amount of standard phosphate solution used.,
The presence of certain organic matter which gives marked color or high
iron content will interfere 'with this phosphate determination. - The color is
not permanent; consequently, comparisons must be made within_15 minutes
after the addition of sodium carbonate solution.
The second method is that recommended by the Hall Laboratories in con-
nection with their system of boiler water conditioning. In this determination,
the temperature of the sample is unimportant; in fact time can be saved by
carrying out the determination on the hot sample.
Reagents: 37 SPecial Ammonium Molybdate Reagent. Solution A.-Place
in a beaker 50 grams of 85% molybdic acid and mix it thoroughly with 120
36 Control of Boiler Water Treatment to Prevent Embrittlement. F. G. Straub.
Mech. Eng., p. 366-7. _.:-:":"- ,
37 Phosphate Control Methods of the Hall System' ot Boiler Water Conditioning.
Hall Laboratories, Pittsburgh, Pa.
WATER ANALYSIS 2091
m!. of distilled water. Add to the mixture slowly and with constant stirring
70 m!. of ammonium hydroxide (NH 4 0H) sp.gr. 0.90. Filter, and then add to
filtrate slowly and with constant stirring 30 ml. of concentrated nitric acid
(HNO a) sp.gr. 1.42.
Solution B.-Add 200 m!. of concentrated nitric acid (HNO a) sp.gr. 1.42 to
480 m!. of distilled water.
When the solutions are cold add Solution A to Solution B slowly and with
constant stirring. Add to the mixture 0.05 gram of ammonium phosphate
. (NH 4 )2HP0 4 dissolved in 10 m!. of distilled water. Shake and let stand for
at least 24 hours, when the precipitate that has formed will have settled out. /
The clear solution is now ready for use.
PotassiuTT} Nitrate.-=-Crystalline c.p. grade. It is important that this
chemical be phosphate free.
Sulfate
The other determination is that involving sulfate and it is important from
sljveral angles. First it is of value in the control of embrittlement, and it is
still one of the inhibiting agents depended upon most strongly and also, where
other methods are employed, it is of value in connection with the use of sodium
sulfite as an absorbent of oxygen for the elimination of gas corrosion. Two
methods are of special interest' in this case, one a turbidimetric method, the
other a colorImetric methqd, which is quite new at this time, but which has
attracted a great deal of favorable attention. In the turbidimetric method
many mechanical devices are used in the obtaining of the ultimate reading
which is based upon the intensity of turbidity necessary to just make impossible
2092 WATER ANALYSIS
.
is recommended:-As the end point is approached, filter off and examine 5 ml. portions,
returning to the main solution the portion each time until the end point is reached .
F. Alkali Reagent-N/IO Solution.-Dissolve 5.3 grams of pure Na 2 CO a
'
and 4.8 grams of stick NaOH in water and dilute to exactly 2000 m!. with
distilled water in a two liter measuring flask. Standardize against the standard
acid solution using alizarine or methyl orange indicator. Record the exact
normality factor.
G. Indicators.
Alizarine.-O~e gram pure alizarine paste mixed with 200 m!. of distilleq· ,
water. One drop is used in the titrations. Color is violet with alkaline solu-
·tion, pale lemon with neutral or acid solutions.
Methyl orange.-(See text.)
Phenolphthal~in.-(See text.)
H. Ammonium Oxalate Solution.-Saturated solution.
I. Permanganate Reagent N/IO.-(See text.) 1 m!. =0.002 g. Ca.
'Methods of Procedure: Chloride Determination.-By means of a 100 ml.
pipette transfer that volume of water to be examined to a porcelain dish. Add a
few drops of 10% potassium chromate indicator solution and titrate with silver
nitrate reagent adding this until a faint but permanent red color (Ag2 Cr04) is
obtained. Report results in- mg. Cl per liter. (Parts per million.)
Carbonate (COs) Bicarbonate (HCO a) and Hydroxyl (OH) Determination.-
Place, by means of a pipette, 100 mI.. of the water in a beaker or porcelain dish;
add a few drops of phenolphthalein indicator. A pink color indicates the
presence of a carbonate. A colorless solution indicates its absence.
If a pink color is obtained titrate with N /10 acid until the pink color dis-
appears; read the burette and record the result as ml. P.
Alkaline Earths .
Calcium and Magnesium Determination.-Calcium and magnesium here
are determined jointly by precipitating with the mixed reagent N a 2CO a and
NaOH (the latter necessary.for the complete precipitation' of Mg). The
excess of the alkali is determined by titration with a standard acid. The two
are reported as calcium. I
Procedure.-Place 100 ml. of the water in an Erl~meyer flask and add 25
ml. of the standard alkali reagent (F); boil, filtel',-'and wash. To the filtrate
containing the excess of the alkali reagent add a drop of the alizarine indicator
WATER ANALYSIS 2097
and titrate (best conducted in an Erlenmeyer flask) the alkali, adding a few
drops of the acid in excess. Boil to expel the CO 2 and determine the excess of
the acid by back titration with the standard alkali to the yellow neutral point,
best obtained by adding the alkali to !l faint pink and back to yellow with a
drop of standard acid.
Calculation.-If A is the total alkali reagent added, i.e., the 25 ml. and the
back titration for acid neutralization; and B is the acid titration (both being
converted to their exact equivalents-ml. in terms of 0.1 N), then A minus B is
the amount of alkali reagent required in the precipitation of Ca and Mg.
One ml. of 0.1 N (i.e., N /10) alkali reagent is equivalent to 0.002 g. Ca. '
Express the results in mg. per liter.
Magnesium Determination.-Concentrate 100 m!. of the water to about a
fourth of its volume. Acidify with HCl and expel the CO 2 by boiling. Cool,
add a few drops of phenolphthalein and neutralize the free acid with NaOH
solution. Now from a burette add a measured amount of standard NaOH
reagent (0.1 N) in amount sufficient to precipitate all the Mg present and a
slight excess. Estimate the total volume present and add twice its volume of
95% alcohol, Mg(OH)2 precipitates. Transfer to a 250 ml. vo~umetric flask,
make up to 250 ml. with 1 : 1 alcohol. Filter off 70-80 m1. and take 50 ml. of
this. Titrate the excess of NaOH with standard acid and multiply by 5.
Calculate the 0.1 N NaOH required by Mg. 0.1 N NaOH will precipitate
per m1.·0.00122 g. Mg.
NOTE.-In the determination of magnesium, elements forming insoluble hydroxides,
under the conditions for magnesium hydroxide precipitation, ilhould be removed previous
to the determination of magnesium. The common interferences are iron, aluminum
and calcium.
Calcium Determination.-Concentrate 100 ml. of the water to about half
its volume if the water is comparatively soft or determine directly if hard.
Details of the oxalate-permanganate method are given in the text.
Alkali Determination.-These are calculated by difference, it being assumed
that the amount, of other radicals in the solution is negligible.
Consult table following. Multiply each radical value by its proper reaction·
~oefficient and b.alance the + and - reacting values; the difference will be the
reacting value of sodium in the solution. Divide this" value by the reaction
coefficient for sodium to obtain the mg. of sodium. C!l.lculate the mg. per liter.
REACTION COEFFICIENTS
Reaction Coefficients
Positive
• r Negative r
IRRIGATING WATERS
Irrigating Waters
Alkali Coefficient
(a) When Na-.65 01 is zero or negative,
(b) When Na-.65 01 is positive, bu~ n9t greater than .48 S04
More than 18. .. Good .. . Have been used successfully fOf many years without special
care to prevent alkali accumulation. \
18 to 6 ........ , Fair .. .. Special care to prevent gradual alk~ accumulation has
generally been found necessary excepton-lQose soils with
free drainage. . - .
5.9 to 1.2 ....... Poor .. .. Care in selection of soils has been found to be imperative and
artificial drainage has frequently been found necessary.
Less than 1.2.... Bad ... . Practically valueless for irrigation.
2099
2100 MICRQSCOPICAL EXAMINATION pF WATER
In New York City's supplies, despite the diversity "of its sources, "the only
organisms which have given offense from odors have been 4sterionella, Tabel-
laria, Volvox, Dictyosp"haerium, Anabaena, Aphanizomenon (with admixtures of
Coelosphaerium, Clathrocystis, and Microcystis), Ceratium, Dinobryon, Perid_i-
nium, Synura, Uroglena, and Crenothrix. .
Asterionella; when present in 500 to 1000 standard units per cubic centi-
meter, produces a slightly aromatic odor. At 1000 units, rarely less, the odor
is distinctly similar to the odor of the geranium. The odor increases in in-
tensity with increase in numbers until several thousand produce a fishy odor.
The fishy odor is also produced when smaller quantities die.
Tabellaria, and similarly Asterionella, in very small amounts produces an
earthy odor (also produced by large amounts of Synedra) , passing through
the aromatic, geranium and fishy stages with about the same relative quan-
tities of organisms as Asterionella. At times the odor of Tabellaria has sug-
gested illuminating gas, no other organisms being' present. In the spring of
1919 extensive complaints of fishy taste in the Catskill supply were occasioned
in New York City by only 700 units of T cibellaria.. This was the result of
chlorination with liquid chlorine at Kensico reservoir. The odor was not
noticeable in the water above the chlorination plant but appeared first just
below the plant. Chlorination killed the' organism, setting free the odorous
principle.
Anabaena and Aphanizomenon, when present in 500 to 1000 units, produce a
faintly grassy odor like freshly-cut grass. With larger numbers the odor
becomes pungent like nasturtium, or even onions. In large numbers, or
when decaying, the odor is of vile, pigpen character. In the spring of 1918
from 500 to 800 units of Aphanizomenon in the Croton supply of New York City
caused numerous complaints of an oily brown scum when bath tubs were
filled with hot water, These complaints came chiefly from high class apartment
houses.
Uroglena produces an oily fishy taste and odor, first noticeable in probably
500 to '1000 units. In larger quantities it is very disagreeable. The flavor is
that Of cod-liver oil. Chlorine has an influence on this organism also.
Synura has caused trouble in as small amount as 50 units. The odor is
vario-qsly described as like cucumber, muskmellon, etc. It leaves a bitter after-
taste. Chlorine intensifies the trouble so that only a few units are noticeable.
Dictyosphaerium, about 700 units, under influence of chlorine has produced
a grassy or pungent nasturtium.odor.
Troublesome organisms occur chiefly in surface waters: Occasionally well
waters containing iron or manganese cause trouble from growths of Crenothrix
and associated forms which clog pipes and cause an unsightly turbid discolored
water. Well strainers become clogged so as to prevent pr()per yield of water.
Microscopic organisms apparently do not affect the health. Possibly the'
taste and odor at times produce nausea or distaste for food. It would take
12,000 units of Asterionella per cubic centimeter to add a milligram of solid
matter to a glass of water.
2102 MICROSCOPICAL ,EXAMINATION OF WATER
I
SEDGWICK-RAFTER METHOD OF EXAMINATION
Apparatus Required _
A microscope (not necessarily high priced). I
Eyepieces-One inch (25 mm.) and one half inch (!g,__-mm.).'
Objective-Two thirds inch (16 mm.) and one sixth-inch (4 mm.).
Ocular micrometer-Fitted to one-inch eyepiece.
MICROSCOPICAL EXAMINATION OF WATER 2103
This is ruled with a square of such size that used with a one-inch eyepiece
and two-thirds inch objective the tube length of microscope may be adjusted
so that the area viewed on the stage will be one square millimeter. The
tube length is adjusted by making the square in the eyepiece coincide with a
stage micrometer. The large square is usually divided into quarters, one
quarter into 25 small squares and one of the latter into 25 tiny squares;
each of the last represents an areal unit of measurement,.400 square microns
(20 microns to the side). A micron equals .001 mm.
A stage micrometer-graduated usually to tep.ths of a millimeter.
Sedgwick-Rafter Counting Cell-a thick slide to which is cemented a brass rim~
with internal dimensions of length 50 mm., width 20 mm., and depth 1 mm.
It has an area of 1,000 'square millimeters' and a capacity of 1 cubic centi-
meter. Its depth only must be exact. Several should be provided.
'Thin cover glasses, 55 mm. by 25 mm., to use as covers for the cells.
One cubic centimeter pipettes (regular bacteria pipettes will do).
Five .cubic centimeter pipettes, volumetric.
A small glass jar of distilled water-for rinsing cells and coverglasses.
Lintless cloth or handkerchiefs-for drying cells and coverglasses.
One-inch test tubes of heavy glass (bacteria dilution tubes) .
.Five-eighths-inch test tubes of heavy glass (bacteria media tubes).
Sedgwick-Rafter funnels. These are cylindrical, 2-inch diameter, sloping
toward the bottom to a neck of i-inch diameter. The distance from top
to slope is 9 inches, length of slope 3 inches, length to bottom 2i inches.
A small glass jar of distilled water-to furnish water for concentrates.
Perforated rubber stoppers-to fit one half inch tube of funnel.
I
Silk bolting cloth No. 15x.
Steel punch-to make i" diameter discs of bolting cloth for rubber stoppers.
White sand, 60 to 120 mesh, i.e., passing a sieve of 60 mesh to the inch and
retained on 120 mesh (Berkshire sand or ground quartz).
Support for battery of funnels.
A small wooden mustard spoon-to measure sand.
Volumetric flasks, 200 m!. capacity.
Test-tube racks.
Record books designed for microscopic examinations.
For field work, which is sometimes of importance due to destruction of
organisms in transit, a folding microscope and a metal sling filter are useful
but not necessary.
Procedure.-The actual procedure of examination is simple. Familiarity
with the various genera is essential.· In the limited space of this chapter it is
impossible to picture the thousands of forms that exist. The best working
book, containing also plates, is "The Microscopy of Drinking Water" by
George C. Whipple. (Revised by Fair and Whipple.) References to other
books showing plates will be found at the end of the chapter.
Moisten the small end of rubber stopper with water so as easily to pick up a
disc of silk bolting cloth from the table and insert tightly into the funnel.
Fill the mustard spoon with prepared sand and drop into funnel making one-
half inch layer. Pour in 200 m!. of the water to be examined from a flask and
2104 MICROSCOPICAL EXAMINATION OF WATER
allow to filter, thus concentrating the organisms uponj the sand. D6 not
allow the sand to stand until drained dry. Slight suction may be carefully
applied to clogged filters. Remove funnel, hold in a slanting position and
carefully remove stopper and insert the end of funnel into ;a one-inch test tube.
Wash sand and contents into the test tube by 5 ml. of dis~illed water, running
the latter quickly around the sides of the funnel. Shake water and sand,
settle a few moments and decant liquid into the smaller (i inch) test tube.
A rack of these concentrates may be prepared at one time for examination.
Rinse three or four Sedgwick-Rafter cells and coverglasses in distilled water
and carefully dry with lintless cloth. Place cover glasses slantingly across
cell so as to leave air space at diagonally opposite ends. Run 1 ml. of concen-
trate into one corner of cell, air leaving at the oppol'ite corner, and slide cover-
glass into position to close cell. Allow cells to stand at least five minutes before
examining so that organisms will have an opportunity to settle completely,
as most of them will. This step is very important. While examination is
being made of one cell, others will be settling. Thus' at least three should be
kept going in rotation when many samples are to be examined. .
Examine under the microscope with one-inch ocular containing micrometer
and j-inch objective and proper tube length as previously determined. Esti-
mate the number of standard units of each organism present in an entire field
within the large square of the eyepiece ane! note on record sheet. Area only
is taken into account. Forms like Asterionella, Diatoma, Synedra, Tabellaria
may be counted and multiplied by a factor to get the number of units-thus
each arm of Asterionella may be estimated as equal to one half unit. Fila-
mentous forms like Melosira, Anabaena, Aphanizomenon, Oscillaria' may be
counted in lengths of five units (one-t.enth of side of large square) and the tot\tl
length divided by the number of filaments laid side by side that it would take
to measure five units. The quotient multiplied by 25 gives the number of units
in th_e field. Synedra in "fine-tooth-comb form," Fragillaria, Uroglena,
Synura and bulky forms may be quickly compared wi~h a 25-unit square and
the number of units in the field estimated. The/above method of estimating
units is quicker than it reads and tends to eliminate largely the personal element
from the count, After the bottom of the slide has been counted, focus should
be changed to the top layer and Cyanophyceae, which float, estimated, that is
Anabaena, Aphanizomenon" Clathrocystis, Microcystis, Coelos.phaerium, etc.
Standard Methods of the American Public Health Association and Am.
Water Works Association require 10 fields of the 1000 to be ex~mined. In
ordinary practice, however, a careful count of five representative fields is
sufficient, choosing fields through the middle of the cell at each end, the center
and half way between. After counting, a search of the cell, particularly the
edges, is made for missed forms, especially Crustacei which swim to the edges
to get air and Protozoa like Uroglena. The latter are frequently rolling around
in the body of the cell. .Anabaena, Uroglena and Dinobryon commonly break
up and disintegrate in the absence of air and must be recognized in such condi-
tion. The process of disintegration frequently may be witnessed under the
microscope. Sometimes disintegration occurs in t.r,ansit. . '
On the record sheet, the amounts of units of each-organism are summed up
for the number of fields examined and multiplied by a factor to find the total
t
MICROSCOPICAL EXAMINATION OF WATER 2105
units as referred to the original water. The latter amounts are sUII).med to
find the total units of organisms per cubic centimeter.
If the quantities recommended in this chapter are used the factor is fQund
as follows:
The ratio of total cell t~ counted fields is 1O~0 = 200.
. . 200 cc. 0
Th e concentratIOn IS -5-- = 4 .
cc.
The final factor is then 2~~ = 5.
•
FIG. 292.-Microscope for Microscopical Examination. FIG. 293.-Ocular Micrometer.
Smallest square is one unit.
Large crustacea found only around the edges and not distributed through
the cell should have their total units divided by the concentration rather than
multiplied 'by the factor in referring the count to the original water.
In addition to the count an examination of the original sample in the bottle
by naked eye is often useful in detecting such forms as Uroglena, Cyclops, etc.,
and. in forming a general estimate of the amounts of organisms present.
Many errors may enter into the quantitative estimation of microscopic
organisms. The sample bottle should be inverted to distribute the forms, as
some float and others settle. Organisms may adhere to the sides of the funnel
or to the sand. Others may pass through the sand and some may disintegrate.
One of the greatest errors is caused by uneven distribution through the cell
when' organisms are not numerous. Whipple states that the total error does
not usually exceed 10%. The author of this chapter believes this figure is too
conservative and unavoidably so owing to the large personal element. How-
ever results are usually comparative at anyone laboratory.
2106 MICROSCOPICAL EXAMINATION OF WATER
Standard
No. of Square
-
2
-
3
-
4
-
5
-
6 7
.
- -
S· 9
-
10
-
Total Units'
Per C. C .•
DIATOMACEAE:
A sterionella. . •••••••••
Cyclotella . .................
r 1
•• • • 2 1 4 20 -
Diatoma ................. '.
Fragilaria . .... ........... .
Melosira .... . , ., ., ......... 3 5 4 10 2 24 120
Navicula ... .. , ...... ' .' " "
Stephanodiscus . ........ ' .... 1 1 1 1 5 25
Synedra .. ................. 20 5 25 125
Tabellaria . ................. 4 4 20
List of Organisms
CHLOROPRYCE.uJ : PROTOZOA: OTHE.R ORGANISMS:
Closterium Cemtium Anguillula
• Conferva Dinobryon .. AC4rina
Eudorina Glenodinium /'
Pandorina Mallomonas
Pediastrum Peridinium TOTAL ORGANISMS:
Protococcus Synum Amorphous Matter
ScenedelW!U8 Trachelomonas Miscellaneous Bodies
Spirogyra Uroglena Sponge Spicules
Staurastrum Vorticella
Cy ANOPHYCEAE: ROTlFE.RA:
Anabaena Anuraea
A phanizomenon Polyarthra
Clathrocystis Synchaeta
'Coelosphaerium
Microcystis C.RUSTACE,A.:
Oscillaria Bosmina
Cyclops
SCHIZOMYCETES, ETC.: Daphnia
I,
Crenothrix
Mold l! yphae
A B
Asterionella before (A) and after (B) treatment with copper sulfate, 1/5 p.p.m.
Note effect upon coloring matter. Magnification 150. Qauses geranium to fishy odor.
Fragilaria, magnificatlo""ll150.
Water concentrate as it appears under the
microscope. 1. Synedra, fine tooth comb.
2. Synedra, needle-form. 3. Asterionella.
4. Melosira.
rubber ball. It is visible to the naked eye and recognizable in the bottle.
Rotifera, Crustacea, etc. (more complicated animal forms) are readily dis-
tinguished from their pictures in most instances. Many are visible to the
.naked eye. Cyclops may readily be recognized in the bottle by its shape and.
swift darting motion. It appears about the size of a pin head shaped some-
what like a tiny fish, but wider in proportion. Bosmina (little cow) and
Acarina (little spider) each has a peculiar characteristic jerky motioh.
DIATOMACEAE:
Asterionella t Aromatic, ger~nium, fishy 0.12*-0.20* 1.0-1.7 0.5*-1.0"
Cyclotella tt Faintly aromatic .... . l.0
Diatoma t Faintly aromatic .... .
Fragilaria t 0.25 2.1
Melosira tH '0.2 1.7 2.0
Meridion Aromatic
Navicula t 0.07 0.6
Nitzchia § 0.50 4.2
Synedraltt Earthy 0.36*-0.50· 3.0-4.2 1.0
Stephanodiscus 0.33 2.8
Tabellaria t Aromatic, geranium, fishy 0.12*-0.50* 1.0-4.2 0.5*-Lo
CHLOROPHYCEAE:
Cladophora t 0.50 4.2
Closterium 0.17 1.4
Coelastrum t 0.05-0.33 0.4-2.8
Conferva t 0.25 2.1
Desmidium 2.00 16.6
Dictyosphaerium t Grassy, nasturtium, fishy 0.5"-1.0
Draparnaldia 0.33 2.8
Eudorina Faintly fishy 2.0-10.00 16.6-83.0
Entomophora 0.50 . 4.2
Gloeocystis . Offensive
H ydrodictyon Very offensive 0.10 0.8
Microspora 0.40 3.3
Palmella 2.00 16.6
Pandorina Faintly fishy 2.0:..10.00 16.6-83.0 1.0
Protococcus
Raphidium 1.00 8.3
Scenedesmus 1.00* 8.3
Spirogyra t 0.12 1.0 0.7-1.5
Staurastrum . Grassy 1.50 12.5
Tetrastrum 1.0
Ulothrix 0.20* 1.7
Volvox t Fishy 0.25 2.1 0.3*-1.0
Zygnema 0.50 4.2
CYANOPHYCEAE:
Anabaena t Moldy, grassy, vile 0.12*-0.48 1.0-4.0 0.5*-1.0
A phanizomenon t Moldy, grassy, vile 0.12*-0.50" 1.0--4.2 0.5*-1.0*
Clathrocystis Sweet grassy, vile 0.12*-0.25* 1.0-2.1 0.5*-1.0
Coelosphaerium t Sweet grassy 0.20*-0.33 1.7-2.8 0.5*-1.0
Cylindrospermum Grassy 0.12 1.0
M icrocystis 0.20 1.7
Oscillaria t 0.20-0.50* 1.7-4.2 1.1
Rivularia Moldy, grassy
The organisms where odors are expressed have been in sufficient quantity at one time
or another to cause the characteristic odor.
* Dosage successful in New York City's supplies. .
t These organisms have caused trouble other than odor.
t These organisms have been affected by chlorine producing characteristic odor and
in many cases controlled by dosage ranging from 0.3 to 3 p.p.m. depending largely on
amount of organisms.
§ Controlled by excess caustic lime, 5 p.p.m. in the case of the diatoms.
NOTE.-Range of dosage largely due to temperature.
2114 MICROSCOPICAL EXAMINATION OF WATER
Chemicals Required for Treatment of Different Genera (Continued)
,
Organisms Odor Parts per Pounds per Chlorim!
Million Mil. Gals. p.p.m.·
I
PROTOZOA: I
Bursaria Irish moss, salt marsh, fishy ..... I . ...
Ceratium tt Fishy, vile (rusty brown 0.24*-0.33 2.0*-2.8 0.3*-1.0
color)
Chlamydomonas 0.50-1.0 4.2-8.3
Cryptomonas Candied violets 0.50 4.2
Dinobryon t Aromatic, violets, fishy 0.18* 1.5 0.3*-T.O
Endamoeba hisloly- 25.0-100.0
lica (cysts)
Euglena 0.50 4.2
Glenodinium Fishy 0.50* 4.2
Mallomonas Aromatic, violets, fishy 0.50 4.2
Peridinium Fishy, like clam-shells 0.50*-2.00 4.2-16.6
Synura t Cucumber, muskmelon, 0.12*-0.25* 1.0-2.1 0.3*-1.0
fishy, bitter taste
Uroglena t Fishy, oily, cod liver oil 0.05*-0.20* 0.4-1.6 0.3*-1.0
CRUSTACEA:
Cyclops tt§ ..... .... 1.0-3.0
Daphnia tt§ 2.0 16.6 1.0-3.0
SCHIZOMYCETES:
Beggiatoa t Very offensive decayed· 5.00 41.5
Cladothrix 0.20 1.7
Crenothrix tt Very offensive decayed, 0.33-0.5* 2.8-4.2 0.5*
medicinal with chlorine
Leptothrix t ..... . ...
Sphaerotilis, natans Very offensive decayed 0.10 3.3
SpirophyUum t ..... .... 0.25*
Thiothrix tt Very offensive decayed ..... . ... 0.5*-1.0
FUNGUS:
Achlya ..... .... 0.6*
Leptomitu8 0.40 3.3
Saprolegnia t 0;18 1.5
MISCELLANEOUS:
Chara t 0.10-0.50 0.8-4.2
N itella, jlexilis t Objectionable 0.10-0.18* 0.8-1.5
Potamogeton t 0.30-0.80 ~ ?5--6.7
Nais t ..... ... '~"- 1.0*
Blood worm, .... .... 15.0-50.0
chironomus t -
.
Blood worm, gnats . ... 3.0
,
.....
,
REFERENCES
• George C. Whipple (Revised by Fair and Whipple), "The Microscopy of Drinking
Water.". Wiley & Sons.
Moore and Kellerman," A Method of Destroying or Preventing the Growth of Algae,
etc." Bulletin 64, Bureau of Plant Industry, U. S. Dept. of Agriculture.
Moore and Kellerman, "Copper as an Algicide and Disinfectant'in Water Supplies."
Bulletin 76, Bureau of Plant Industry, U. S. Depk-q_f"Agriculture.
MICROSCOPICAL EXAMINATION OF WATER 2115
"The Use of Copper Sulfate and Metallic Copper for the Removal of Organisms and
Bacteria from Drinking Water,-A Symposium," Vol. 19, No.4, New England
Water Works Association.
Monfort and Barnes, "Chlorarriine and Crenothrix." Journal of American Water
Works Association, Vol. 6, page 196.
Amsbary, "Treatment of Water to Prevent the Growth of Crenothrix." Jour. Am. W.
'W. Assoc., Vol. 6, page 194. .
Monfort, "Crenothrix Removal." Proe. Convention Indiana Sanitary Water Supply
Assoc., April, 1919.
W. W. Brush, Engineering News-Record, Feb. 16, 1922.
Frank E. Hale, "Taste and Odor in New York City's Supplies," Jour. Am. W. W.
Assoc., Vol. 10 (Sept., 1923), pp. 829-837.
Hudson and Gosse, "The Rotifera." Two volumes and supplement, Longman & Co.
W. Saville Kent, "A Manual of the Infusoria." Three volumes.
Wolle, "Fresh Water Algae of the United States." Two volumes.
Wolle, "Desrriids of the United States." One volume.
Wolle, "Diatomaceae of North America." One volume.
Leidy, "Fresh Water Rhizopods of North America." U. S. Geological Survey of the
Territories, Volume 12.
Ward and Whipple, "Fresh Water Biology." Wiley.
Griffith and Henfrey, "Micrographic Dictionary." London.
Gilbert M. Smith, "The Fresh-water Algae of the United States." McGraw-Hill Book
Company.
BACTERIOLOGICAL EXAMINATIO~ OF WATE~
The routine bacteriological examination is essentialiy simple. It consists
of obtaining empirically the total number of bacteria under certain standard
conditions and of testing for the intestinal group, B. coli. Usually only under
special conditions is search made for specific disease bacteria which may be
carried by water supply, such as Bacillus typhi, Bacillus paratyphi-A and B,
Spirillum cholerae, Bacillus dysenteriae-various types, Bacillus anthracis, etc.
SAMPLIN(j-
bulb immersed in the water, a few minutes before reading. Allow tap samples to
run thoroughly, particularly in warm boiler rooms. Handle sterilized bottles
by the bottom and keep fingers away from the necks. Handle stoppers with
the tinfoil in place. Avoid splashing or over-running bottles. In streams
plunge bottle below the surface and sample by swinging hand IfP-13tream with
mouth of bottle pointed up-stream so as to avoid water washing over the hand
into bottle. Similarly in a reservoir, swing the mouth of the bottle forward
while filling. Do not rinse a sterilized bottle. These rather minute directions
are given because in a long experience they have frequently been found neces-
sary. Experts do not always take the samples though the value of their work
depends upon the care .with which the samples are taken. Men have even
been caught reading thermometers after drawing them up twenty feet from
the surface of a reservoir.
Samples from a depth in a reservoir are best taken if possible from an
effluent pipe when regular draft is from a depth. In other cases sterilized
bottles may be lowered to the required depth and stoppers pulled out by
mea~ of special samplers. In many cases it has been found satisfactory to
take them by the two-bottle method as used for di'lsolved oxygen and described
under "Microscopic Examination of ·Water." In this case the quart bottle
should be thoroughly cleaned, handled as little as possible, and the w8ter
.poured into a bacteria bottle after drawing to the smface. The quart bottle
is pretty thoroughly rinsed during the process of sampling. (Fig. 291, p. 2102.)
Samples should be tested the same day they are collected, preferably within
a few hours. Expressed and iced samples will p;obably give a fairly reliable
index of B. coli content but cannot be relied on for total counts.
tubes for same, capacity in either case four times the amount'ofwater to be
tested.
Metal racks for Smith tubes.
Agateware double boilers for making media.
Agateware kettles for making media.
Agateware funnels, 6 inch, for filtering media.
Cotton flannel for filtering media.
Large beakers for making media.
Earthenware pitchers for filling tubes with media.
Side-neck burettes for filling tubes with media (side ntlck connected to funnel of
media). .
Apparatus for keeping media warm while filtering.
Gas stoves for heating media.
,
FIG. 296.-Bacteriological Apparatus. Fermentation Tubes, Petri Dishes,
Culture Media, etc. '
Petri dishes, 10 cm. diameter and flat bottomed, with both glass covers (20 0 C.
incubation) and porous covers (37 0 C. inCUbation).
Agateware sauce pans for welting media in tubes before plantin.g;
20 0 C. incubator with thermometer, automatic regulator and safety· burner.
37 0 C. incubator with thermometer, automatic regulator and safety burner.
(In many laboratories the incubators are now electrically operated.)
Burner for flaming tubes before planting.
Counting stand with engraved lines to assist in counting.
Engraver's lens for counting, B. & L. 146, magnification 2t diameters, 3t X.
Tallying register for counting. '
Balance, similar to an apothecary's, for quick weighing.
Balance, analytical beam type, for accurate weighing.
Record books for bacteriological results. ,
Microscope with condenser and assortment of len~~ -at least t", t" and I"
eyepieces and ~", i" and 1/12" objectives. "(Satisfactory hanging drops
;u
BACTERIOLOGICAL EXAMINATION OF WATER 2119
may be made with t" eyepiece and t" objective. The 1/12" is used only
in oil immersion with condenser.)
Slides, plain and concaved (latter for hanging d;ops).
Cover slips, round or square.
Platinum needles with holders for transferring cultures.
Cleaning mixture for cleaning glassware.
Distilled water for preparation of all culture media and reagents.
Meat extract; :r5igestive Ferments Company's Bacto Beef Extract is satisfactory.
Peptone: N. Y. Ci~y Water Department Laboratories have found Digestive
Ferments Company's reliable. (Some laboratories have found Armour's
and Fairchild's satisfactory.)
Lactose and other sugars the purest obtainable.
Agar. Frequently needs soaking in water and draining before use because of
salt. Shall be dried t hour at 105° C. before weighing.
Gelatine, of light color, containing not more than a trace of arsenic, copper,
sulfides, free from preservatives, and of such a melting point that" standard
nutrient gelatin" shall not melt below 25° C. Gelatin shall be dried t hour
at 105° C. before weighing.
Litmus-reagent litmus of highest purity (not litmus cubes) or azolitmin
(Kahlbaum's) .
Certified Dyes-Basic fuchsin, Eosin yellowish, Methylene blue, Brilliant green.
General chemicals-chemically pure so far as possible.
Hellige comparator for hydrogen-ion with color discs for brown-thymol-blue,
and cresol red and phenol red.
Adjustment of Reaction.-T,he final reaction for broth, gelatin and agar
shall be between pH 6.4-7.0. For Endo medium the final reaction after the
addition of chemicals should be pH 7.8-8.2. The eosin methylene-blue agar
does not require adju·stment. The increase in the hydrogen-ion concentration
during sterilization will vary slightly with the individual sterilizer in use and
the initial reaction required in order ,to obtain the correct final reaction will
ha ve to be determined.
The simplest method is the use of the Hellige Comparator which employs
discs with colored glasses. Each disc cantains the range of colors for a single
dye, 0.2 pH apart, and the colors are a very good match and accurate. Brom-
thymol-blue has a range 6.0--7.6, Phenol Red 6.8-8.4, and Cresol-red a range
7.2-8.8. The first covers all the ordinary media, the second Brilliant-green bile
diluted, and the latter covers Endo. Procure the dyes from any of the firms
supplying it for pH determination. The preparation of the dye solutions is as
follows according to Clark. Rub up a decigram of brom-thymol blue in 3.2 ml.
of N /20 sodium hydroxide. Dilute to 25 ml. with distilled water, making 0.4%
solution. Dilute again for use to 0.04% solution. Rub up a decigram of
Cresol-red in 5.3 ml. of N /20 sodium hydroxide. Dilute to 25 ml. with dis-
tilled water, making 0.4% solution. Dilute again for use to 0.02% solution.
With Phenol-red 5.7 ml. of N/20 sodium hydroxide is used and final solution is
.01 % strength. In testing bacteriological media, for example agar, add to
8 ml. of distilled water 2 ml. of melted ~ar, add t ml. of the brom-thymol-blue
solution, mix by pouring into one of the comparison tubes. In 'another com-
parison tube pour the similarly diluted media but without the dye and place
this behind the colored glasses. Comparison should be made at once and is
2120 BACTERIOLOGICAL EXAMINATION ,OF WATER
2. Heat slowly on a steam bath to 65° C. until all ingrec,lients are dissolved.
3. Make up lost weight with distilled water and adjust the reaction so that
final-pH shall be 6.4-7.0. .
4. Bring to boil. Filter through cloth and cotton until clear.
5. Distribute in test-tubes, 10 ml. to each tube, or in larger containers as
desired. '
6. Sterilize in the autoclave at 15lbs. (120° C.) for 15 minutes after the
pressure reaches 15 lbs. Cool rapidly as soon as possible.
Nutrient Agar.-To make one liter:
1. Add 3 grams of beef extract, 5 grams of peptone and 12 grams of agar,
dried for one-half hour at 105° C. before weighing, to 1000 ml. of distilled water:
Boil over a wafer bath until all the agar is dissolved.
2. Make up lost weight with hot distilled water and, adjust the reaction so
that final pH shall be 6.4-7.0 .
. 3. Filter through cloth and cotton until clear.
4. Distribute in test-tubes, 10 ml. to each tube, or in larger containers, as
desired.
5. Sterilize in the autoclave at 15 lbs. (120° C.) for 15 minutes after the
pressure reaches 15 lbs.
NOTE.-The Water Department Laboratories of New York City for many years have
used and found the following method of clarifying agar very efficient: The prepared agar
is placed in a suitable container and filtered upward by vacuum in a pressure cooker into
four one-quart jars. The mouth of the jar is covered with a filter made of a thick wad of
absorbent cotton 2 inches thick between two layers of cotton flannel, the filter occupying
thE) neck of the jar. The cotton flannel is tied securely in place on the outside of the neck.
I These jars are inverted in the container holding the agar and placed in the pressure
cooker. The pressure is brought up to 15 pounds and allowed to cool. Upward filtra-·
tion takes place under vacuum. •
Adjust the reaction so that the pH reading after sterilization will be 7.4.
Clarify if desired. Add 10 grams of lactose and dissoh,le.
Place in small flasks or bottles, 100 ml. to each, and ~terilize in autoC'laV'e
at 15 lbs. pressure for 15 minutes after the pressure has, reached 15 lbs. _
2. Preparation of Plates.-Prepare a 3% solution .of cJrtified
I
basic fuchsin
in 95% methyl alcohol. Allow to stand 24 hours and filter.
Melt lactose agar as prepared above and'to each 100 ml. add 1 ml. of the
3% 'basic fl1chsin solution and 0.125 gram of anhydrous sodium sulfite dissolved
in 5 ml. of distilled water. The su)Jite solution must be freshly prepared. Mix
thoroughly, pour plates with usual precautions against contamination and allow
to harden.
The medium should be light pink when hot and almost colorless when cool.
If the medium is too highly colored before incubation the strength of the basic
fuchsin solution may be varied.
FORMULA II
Dipotassium phosphate ...................... " . . .. .. 3.5 grams
Bacto peptone .......................... , . . . . . . . . .. 10.0, "
Agar (washed and dried) ....... ' ......... , . . . . . . . . .. 15.0 "
Water (distilled). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 liter
Lactose (c.p.) ......................... ~ .... " ...... 10.0 grams
To each 100 ml. of the above add:
Sodium sulfite (Anhydrous) ..........--: . . . . . . . . . . . . . .. 0.25 "
Basic fuchsin filtered 10% alcoholic solution. . . . . . . . . . . . 0.5 ml.
NOTE.-It is difficult to make a 10% solution with some basic fuchsin preparations.
Brilliant-Green Lactose Peptone Bile.-The composition for approximately
'one liter is as follows:
Dissolve 10 grams of pepto~e and 10 grams of lactose in not more than 500
ml. of distilled water. ' .'
Add 200 ml. of fresh ox bile or 20 grams of dehydrated ox bile dissolved in
200 ml. of distilled water. The solution of dehydrated ,ox bile shall have a pH
between 7.0 and 7.5. /
Make up with distilled water to approximately 975 ml. and adjust reaction
to 7.4 pH. I
Add 13.3 ml. of a 0.1 % solution of Brilliant-Green in distilled water and add.
sufficient distilled water to make the volume 1000 m!. Filter through cotton,
distribute in fermentation 'tubef:l and sterilize at 15 lbs. pressure for 15 minutes
after pressure reaches 15 lbs. Total exposure to heat should not .exceed one-
half hour. Cool rapidly after removal from autoclave. _,
The reaction after sterilization should be between pH 7.1 and 7.4. 6
Brilliant-Green Lactose Peptone Bile of the above formula may be obtained
mixed in dry form from the, Digestive Ferments Company, Detroit, Mich., and
has the following composition: I
For a water inoculum of 1 m!. or less or for loop inoculations the above dry
material should be used in proportion of 4 grams to 100 ml. media water (dis-
tilled) and tubed in 10 ml. quantities. The concentration of bile in the inocu-
lated tube will then be 2% and of Brilliant-Green 1 part in 75,000. For 10 m!.
water samples a total volume of water plus medium of 30 ml. is used and the
medium is made using 6 grams of the above dry mixture per 100 m!. of media
water, which results in the same final density of ingredients.
For routine presumptive planting on a large scale it is simpler to make up
double strength media as follows:
. Dissolve in 4 liters of hot distilled water, 160 grams of dessicated bile, 80
grams of peptone, and add 15 ml. of N /1 sodium hydroxide. Remove from
flame and add 80 grams of lactose. After solution filter through flannel, make
up to original volume and add 10.6 m!. of 1 % Brilliant-Green. .
For small tubes in which are planted 1 m!. or less of water, dilute 1500 m!.
of the above to 3000 m!. and tube in 10 ml. quantities.
For large tubes in which are planted 10 m!. of water, dilute 2500 ml. to 3750
ml. and tube in 30 ml. quantities.
For testing of 100 m!. quantities of water use the double strength media as
prepared, 100 m!. quantities of media per 100 m!. quantity of water tested.
Hesse Agar.-The composition for approximately, 1 liter is as follows:
Agar (dried). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 grams
-Peptone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. "
Beef extract, Liebig's. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5. "
Salt (sodium chloride). ,'. . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5 "
Distilled water . .................................. . 1000. "
Dissolve 4.5 grams of dry agar in 500 m!. distilled water by heating over a
free flame, making up loss in weight by evaporation. Into another vessel 500
ml. of distilled water is poured and to this is added 10 grams of peptone, 5
grams of Liebig's beef extract, and 8.5 grams of salt. This is heated until
all is dissolved and the loss in weight by evaporation is made up by adding
distilled water. •
Add the two solutions together; boil 30 minutes; make up loss in weight
with distilled water, filter through absorbent cotton held in the funnel by
cotton flannel, passing the filtrate through several times until perfectly clear.
Test the reaction; adjust, if necessary, to pH = 7.0, and tube, using 10 m!. in
each tube. Sterilize for 15 minutes at 15 lbs. pressure in an autoclave. Cool
with running tap water and store in an ice-chest, the air of which is saturated
with moisture.
Russell Media.-To ordinary extract agar adjusted neutral to litmus add
1% of lactose and 0.1% of glucose and sufficient litmus to give a good color.
Tube and slant leaving a generous "butt" at bottom of tube for stab inoculation.
2124 BACTERIOLOGICAL EXAMINATION OF WATER
I
I I
Dulcite + Dulcite -
B. communior B. aerogenes
B. communis B. acidi lactici
I I
I I ' I I
Saccharose + Saccharose - Saccharose + Saccharose -
B. communior B. communis B. aiJrogenes B. acidi lactici
7 AdC'pted by A. P. H. A., 1912, and recommended by the author.
2126 BACTERIOLOGICAL EXAMINATION: OF WATER
-
The "Standard Methods" in testing for the coli-aero genes group require
gas formation in lactose broth plus confirmation in accordance with the fol-
lowing procedure: '
"The following classification is included in the 1936 edition of 'Standard
M'ethods! Its chief usefulness is to divide the group! into two broad types,
B. coli and B. aerogenes. The latter type represents ml\lmbers that can persist
under adverse conditions, hence less sanitary importance may sometimes be
applied to their presence. In England, particularly, more importance is
ascribed to the B. coli test. Attempt has been made to separa:te the aero genes
type into Fecal and Non-fecal grouping, but as all types are found in human
excrement this separation has not generally been found of any great importance.
Coli-Aerogenes Group-Reaction Classification
. Reaction I'ossible Interpretation when Isolated from Water Common .
Combinations by the Standard Methods Source
formation occurs. In order to obtain typical results i,t is essential that plates
be streaked so as to insure presence of some discrete colonies. Transfers
should be made not later than the end of the first 24 hour incubation period if
gas is formed during this time. If at the end of 48 hours gas is formed in tubes
containing less of the sample of water than at 24 hours, transfers should be made
from these tubes. (If larger amounts are not transplanted for confirmation,
they should be recorded as confirmed, even though the smaller portions may
fail subsequently to confirm.)
Incubate the plates at 37° C. for 18 to 24 hours.
If typical colonies develop upon the plates within this period, the confirmed
test may be considered positive. If, however,. no typical colonies develop
within 24 hours, the test cannot be considered definitely negative and' the
completed test should be made.
2. Liquid Confirmatory Media.-Transfer from a lactose broth tube showing
gas to the fermentation tube containing the selected confirmatory· mediulll.
It is recommended that all lactose broth tubes showing gas be confirmed to
comply with this test, but it is allowable to follow the procedure outlined for
solid media. Transfers should be made as soon as gas appears. In routine
testing it is convenient to make observations and transfers at the end of 24
and 48 hours. In making transfers the tuqe showing gas shall first be gently
shaken or mixed by rotating and the transfer shall be made by mMns of a wire
loop not less than 3 mm. in diameter; 9r it is permissible, to use the sterile
capillary tube or sterile pipette when it is considered desirable to transfer larger
amounts:
Incubate the inoculated c~nfirmatory tubes for 48 hours at 37° C.
The formation and presence of gas in any amount in the. inverted vial in
the fermentation tubes at any time within 48 hours constitutes a confirmed test.
Completed Test
Streak one or more Endo or Eosin Methylene-Blue plates from the lactose
broth tubes which show gas from the smallest amomit of water tested or from
the secondary selective medium tubes which ~how gas. If these plates are
streaked from the original lactose broth tubes they should be made at -the
same time that the secondary selective medium tubes are planted. If plates
are used for the confirmed test, completion may be continued after incubation.
Incubate.the plates at 37° C. for 18 to 24 hours. .
Identification.-From the plates fish one or more typical colonies, or if no
typical colonies are present, fish two or more colonies considered.. most likely to
be organisms of the coli aerogenes group, transferring each to an agar slant and
,a lactose broth fermentation tube. The lactose broth fermentation tubes thus
inoculated shall be incubated until gas formation is n,oted, not to exceed 48 hours.
The agar slant shall be incubated at 37° C. for 24 hours, when a microscopic
examination for spores and a Gram stain of at least one culture shall be made,
selecting when possible the one which corresponds to one of the lactose broth
fermentation tubes that has shown gas formation. If none of the lactose broth
fermentation tubes contain gas at the end of 24 hours, all agar slants shall be
examined, and corresponding tubes examined o.n-tlre 'following day.
Results.-The formation of gas in lactose broth and the demonstration
/' .
BACTERIOLOGICAL EXAMINATION OF WATER 2129
Widal Test.-To make the Widal test place a small portion. of the culture
. in a drop of normal salt solution on a cover glass, and invert it over a rubber
ring on a glass slide, using vaseline on the edges so a& to prevent evaporation
and consequent movement by currents in the drop. \
If the bacteria are motile rods resembling B. typkj ~add a drop of equal size
of highly diluted anti-typhoid serum (the latter diluted with physiological salt
solution). If in the course of half an hour the bacteria cease their motions
and agglutinate, the presence of B. typhi is practically established. In rare
cases, some strains of B. coli and allied species may respond to the Widal test,
but almost invariably in dilutions of 1-50 or less. The high-power dry lens is
best for this work.
To determine the point of highest dilution at whIch agglutination take.
place the Widal test is best made in tubes and examined macroscopically.
Place a series of small test tubes in a rack and pour into each varying dilutiom
of anti-typhoid serum. To each tube add an equal amount of 24-hour brQth
culture of B. typhi and incubate at 37° C. for at least three hours. The highest
dilution at which precipitation takes place is the highest point of agglutination.
Russell Media.-As additional rapid means of differentiating between·B.
typhi and B. para-typhi A and B, needle stabs may be made into tubes of Russell
media (see Preparation of Media) and incubated at 37° C. The reactions are
based on the fact that the bacilli growing aerobically, that is on the slant, only
utilize the carbohydrate when present in amounts over 0.1 %, whereas, growing'
anaerobically stabbed in the butt, they must utilize the carbohydrates for their
oxygen supply and therefore ferment the trace of glucose present. The indi-
cations of the Russell media reactions are as follows.:
and saccharose. The type I produces acid in neither, type II produces acid
in maltose but not in saccharose, type III produces acid in saccharose but not
in maltose. The usual final test for all these different types of typhoid and
dysentery is the agglutination with specific serums: The sugar fermentation
tests must be made upon freshly isolated cultures since after artificial cultivation
Park-HJss strains may ferment maltose and Flexner strains saccharose as
demonstrated by Hiss and by Lentz.
The members of the ~yphoid, para-typhoid, dysentery and para-dysentery
groups all form colorless colonies upon Endo media. This is frequently the
starting poiht for isolation. '
REFERENCES
"Standard Methods of Water Analysis." Am. Pub. Health Assoc., Editions 1905,
1912, 1917, 1920, 1923 and 1936. (Contain extensive bibliography.)
Park and Williams, "Patbogenic Microorganisms."
Prescott and Winslow, "Elements of Water Bacteriology."
Chester, "A Manual of Determinative Bacteriology."
Bergey's "Manual of Determinative Bacteriology."
"The Determination of Hydrogen Ions," Clark.
.1
MISCELLANEOUS ANALYSES *
ANALYSIS OF ACETONE 1
PURITY
Reagents.--N/20 Na 2 S 2 0S is made up and standardized carefully against
K 2Cr 2 07. For accurate work standardized burettes must 'be used and burette
corrections and temperature corrections applied.
N/I0 Iodine is made up and is standardized at the time of making a deter-
mina.tion The standardization consists in titrating against the N /20 Na 2S 2 0S
50 ml. of iodine mixed with the same amounts of NaOH and H 2S0 4 as are used
in the determination of acetone. The titrations are made in triplicate and the
average of the three wHh burette and temperature corrections is used as th~
blank in making the calculation.
Normal NaOH is made up and standardizep against 2 N sulfuric acid.
2 N Sulfuric Acid is made up and checked against 1 N NaOH and the
strength adjusted so that 25 ml. of the acid will neutralize 50 ml. of the NaOH.
Starch Solution is prepared according to' standard procedure.
Preparation of Sample.-Approximately 1.4 grams of acetone are carefully
weighed by one of the following methods and made up to a volume of one liter.
(a) Glass ampules %" to ~" in diameter are blown and after filling with
acetone are sealed. The exact weight of sample is determined by the difference
in weight of the empty and filled ampule. Samples varying in weight more than
plus or minus 0.1 gram from 1.4 gram should be rejected. The. glass ampules
are dropped into 500 m!. glass-stoppered bottles containing 200 ml. of boiled
distilled water and broken by shaking. The capillary ,stem is then {)arefully
crushed with a glass rod to allow thorough washing outfof the acetone and the
sample and combined washings made up to one liter at 20° C.
(b) A quicker, easier method is to weigh out 1.75 m!. of acetone from a 2 ml.
pipette graduated in tenths of a ml. into a small gla·ss-stoppered weighing bottle,
conical shape, capacity 12 to 51 ml. (as listed in Elmer and Amend's catalogue.
- This volume will weigh approximately 1.4 grams. The weighing bottle is
inverted and the mouth held under the surface of boiled distilled: 'water in a
liter beaker.' The stopper is ·then removed under water by means of a glass
roa with a hook on the end. The bottle is thoroughly washed out and the water
and acetone are transferred to a one liter flask and made up to the mark at 20° C"
Procedure.-Three 25 m!. aliquots from each of 3 ~asks made up according
to (a) or (b) under Preparation of Sample are pipetteq into 750-ml. Erlenmeyer
flasks containing 50 m!. of the 1 N NaOH solution. A Bureau of Standards
pipette is used for taking the samples from the flasks. Into each flask is run
with constant shaking 50 ml. of the N /10 Iodine. The flasks are then stoppered
and allowed to stand at about 20° C. (cooling necessar_y},!l wanD weather) for at
* Edited by N. H. Furman.
l' Method of Commercial Solvents Corporation.
2134
MISCELLANEOUS ANALYSES 2135
least 20 minutes. At the end of this time 26 ml. of the 2 N sulfuric acid are
poured into the first flask which is then titrated immediately against N /20
Na2S~03. ,The solution must be acid before any thiosulfate is added, otherwise
the determination is ruined. When the color of the iodine has almost faded
out, starch solution is added and the titration continued until the blue color
disappears and the bright 'yellow color of the iodoform suspension is reached.
The mls. of thiosulfate are read, the burette and temperature corrections are
applied and the corrected reading subtnfcted from the average corrected blank.
Percentage of dimethyl ketone-by weight is then calculated from the following
formula:
where
N =Normality of NaOH,
39.7 = Weight of 50 ml. sample eSp.gr .. 795).
Water.-Shake together one volume of the sample and three volumes of
benzene in a test tube. Any water present will be precipitated by the benzene
thus imparting a turbidity to the solution in the tube.
Color.-The sample is compared with a sample of distilled water. It
should be absolutely colorless with no cloudiness or solid matter.
Residue.-Evaporate 25 fil. of the sample to dryness in a tared porcelain
or preferably a tared platinum crucible over a water bath. Dry to constant
weight at 110° C. in an oven.
a
2136 MISCELLANEOUS ANALYSES
ACETONE
If the bottle is not shaken vigorously while adding the iodine solution, the
iodine cannot act completely on the acetone.
ALCOHOL
DETER~nNATION OF ALCOHOL
In carbonated liquids the carbon dioxide should first be expelled by pouring
the liquor back and forth from one beaker to another, or by vigorously shaking
,
2138 MISCELLANEOUS ANAL Y:El.ES
the sample in a large sepatatory funnel and drawing off the lower portion, re-
peating this several times if necessary. After treatm~nt the liquid should be
free from foam.
Distillation Method.-This is the most accurate m~thod for determination
of alcohol. A convenient quantity of the sample is hccurately measured' or
weighed, ~ccording to whether the per cent by weight or measure is desired.
The weight taken varies from 25 to 100 grams, according to the alcohol content.
of the sample.
The sample is placed in a 250 to 400 ml. round-bottom flask and diluted to
150 ml. The flask is connected to a condenser and 90 to 95 ml. of the liquid is
distilled into a narrow-necked flask. The distillate is made to exactly ioo ml.
with distilled water. . .
The specific gravity of the thoro~ghly mixed distillate is taken by means of
the pycnometer or by the Westphal balance, the temperature of the liquid
being exactly 15.6° C. From the specific gravity the corresponding percentage
of alcohol by weight or by volume, or the grams per 100 ml. in the distillate,
is ascertained from the accompanying tables.
The percentage of alcohol by weight in the sample is obtained by multi-
plying the per cent by weight in the distillate by the weight of the distillate,
and dividing by the weight of the sample taken. The per cent by volume irt
obtained by multiplying the per cent by volume in the distillate by 100 and
dividing by the volume of the sample taken.
Evaporation Method.-Should a distillation apparatus not be available the
following method may be used. The specific gravity of the original sample is
determined, the temperature being regulated to 15.6° C. A measured portion
of the liquor (50-100 ml.) is placed in a porcelain dish over a water bath and
three fourths of its volume evaporated off. The concentrate is diluted to
exactly its originllJ volume and its specific gravity again taken at 15.6° C. TQ,
apply the table add 1 to the or.iginal specific gravity and from this total sub-
tract the second specific gravity. The difference is the specific gravity corre-
sponding to the alpohol in the liquor. For example, ,
if'the
I
specific gravity of the'
original solution was 0.9989 and the de-alcoholized sample was 1.0005, then.
1.9989 -1.0005 = .9984 = 1.06% alcohol, by volume.
If the liquor is free from residue its specific gravity may be taken directly
without distillation and the alcohol content determined, provided it is a mixture
of water and alcohol.
dryness is accomplished in a water oven for two and a half hours at 100° C.
The weight of the residue divided by the specific gravity gives per cent of
non-volatile residue.
Acidity Expressed as Acetic Acid.-10 ml. of the alcohol are titrated with
N /10 NaOH in presence of phenolr.hthalein indicator.
1 m!. N/lO NaOH=0.006 gram acetic acid.
Distilled Solution for Determining Aldehydes, Furfural and Esters (A). To
250 ml. of the sample~30 ml. of water are added and the liquor distilled into a
250-m!. volumetric flask. When the distillate has almost reached the mark
the distillation is discontinued and the volume made up to mark by addition of'
water. The distillate contains all the aldehyde, furfural and esters. 1 ml. of
this solution is equivalent to 1 m!. of the original sample.
ALDEHYDES
QUALITATIVE EXAMINATION
(49.04 grams H 2S0 4 mad~ up to 1000 ml. with 95% alcohol) arfd the solution
made to 100 ml. with 95% alcohol. To c6mpensat'i for the (NH 4)2 S04 pre-
cipitated 0.8 ml. alcohol is added. After standing all night the solution is
filtered. 100 ml. of this reagent contains 1 gram of rcetaldehyde. It retains
its strength . I
Reagent for Standard.-Two ml. of the above sqlution diluted with ))0%
alcohol to 100 ml. 1 ml. contains 0.0002 gram acetaldehyde. Make up Mesh
for use.
PROCEDURE FOR ALDEHYDES
Ten ml. of the solution A are diluted to 50 ml. with the aldehyde free alcohol
(diluted to 50% by volume), 25 ml. of fuchsin reagent are added and the mixture
allowed to stand for 15 minutes (all reagents and the sample should be at 15° c.f.
The solution is now compared colorimetrically with standards made up in the
same way. Nessler tubes may be used. Comparison may be made III a
colorimeter. From this the per cent aldehyde is readily calculated.
DETERMINATION OF .,FURFURAL
:Fifty ml. of the distilled solution (A) in an Erlenmeyer flask are exactly
neutralized with N/10 NaOH (phenolphthalein indicator) and about 50 ml.
excess added, the exact amount being recorded. The solution is boiled for an
hour with reflux condenser. After cooling, the excess of alkali is titrated with
N /10 acid. The ml. N /10 N aOH consumed in the sapQnification, multiplied
by 0.0088, gives the gram weight of esters calculated as ,e-thyl,acetate. This
weight divided by the specific gravity of the sample and multiplied by 2 gives
per cent. -
DETERMINATION OF FUSEL OIL
Fifty ml. of the original sample, in an Erlenmeyer flask, are diluted with an
equal volume of water and 20 ml. of N 15 NaOH added. The mixture is now
saponi.fied by boiling for an hour with reflux condenser. The flask is connected
to a condenser and about 90 ml. distilled into a deep beaker. 25 ml. of water
are added to the flask and the distillation continued until the total volume
of the distillate is 115 ml. Finely ground salt (N_.],OI) is ,now added until the
solution is almost saturated and a saturated solt'tion of N aCI added until the
MISCELLANEOUS
, ANALYSES 2141
Fifty ml. of the sample are neutralized with N /10 NaOH (phenolphthalein
indicator) and evaporated to a moist residue. A little distilled water is added
1" m!. of phenoldisulfonic acid, and the mix made alkaline with NH 4 0H.
A bright orange color indicates the presence of nitrates. Run a blank on the
water used. See subject of nitrates in the chapter on Water Analysis.
A hundred ml. of the alcohol are made slightly alkaline with N /10 N aOH
(phenolphthalein) and 5 ml. H 2 0 2 added, preferably in a large platinum dish.
The residue is ignited over an alcohol flame, then taken up with 50 ml. of water,
5 m!. dilute HCI added and 5 ml. 10% BaCI 2 • The precipitat.ed BaS04 is
filtered off, ignited and weighed according to the customary procedure.
BaS04XO.1373=S.
Specific
Gravity at
Alcohol
Specific,
Gravity at
I, Alcohol
60° per cent per cent Grams 60° per cent per {'ent Grams
ti00 F by by per 600 F. by by per
volume. weight,
I 100 ml. volume. weIght, lOQ.ml.
Alcohol Alcohol
Specific Specific
Gravity at Gravity at
6()o per cent per cent Grams 6()0 per cent per cent Grams
000 F. by by per 600 F, by by per
volume. weight. 100 ml. voiunle. weight. 100 rol.
\ \ ..
.98825 .8.60 6.91 6.83 .98359 12.70 10.24 10.07
.98813 8.70 6.99 6.91 .98348 12.80 10.33 10.15
.98801 8.80 7.07 6.99 .98337 12.90 10.41 10.23
.98789 8.90 7.15 7.07 .98326 13.00 10.49 10.31
.98777 9.00 7.23 7.14 .98315 13.10 10.57 10.39
.98765 9.10 7.31 7.22 .98305 13.20 10.65 10.47
.98754 9.20 7.39 7.30 .98294 13.30 10.74 10.55
.98742 9.30 7.48 7.38 .98283 13.40 10.82 10.63
.98730 9.40 7.56 7.46 .98273 13.50 10.90 10.71
.98719 9.50 7.64 7·.54 .98262 13.60 10.98 10.79
.98707 9.60 7.72 7.62 .98251 13.70 11.06 10.87
.98695 9.70 7.80 7.70 .98240 13.80 11.15 10.95
.98683 9.80 7.88 7.78 .98230 13.90 1'1.23 11.03
.98672 9.90 7.96 7.85 .98219 14.00 11.31 11.11
.98660 10.00 8.04 7.93 .98209 14.10 11.39 11.19
.98649 10.10 8.12 8.01 .98198 14.20 .11.47 11.27
.98637 10.20 8.20 8.09 .98188 14.30 11.56 11.35
.98626 10.30 8.29 8.17 .98177 14.40 11.64 11.43
.98614 10.40 8.37 8.25 .98167 14.50 11.72 i11.51
.98603 10.50 8.45 8.33 .98156 14.60 11.80 1l.59
.98592 10.60 8.53 8.41 .98146 14.70 11.88 1l.67
.98580 "'10.70 8.61 8'.49 .98135 14.80 11.97 11.75
.98569 10.80 8.70 8.57 .98125 14.90 12.05 11.82
.98557 10.90 8.78 8.65 .98114 15.00 12.13 11.90
.98546 11.00 8.86 8.73 .98104 15.10 12.21 11.98
.98535 11.10 8.94 8.81 .98093 15.20 12.29 12.06
.98524 11.20 9.02 8.89 .98083 15.30 12.38 12.14
.98073 15.40 12.46 12.22 .97658 19.50 15.84 15.47
.98063 15.50 12.54 12.30 .97648 19.60 15.93 15.55
.9S052 15.60 12.62 12.37 .97638 19.70 16.01 15.63
.93042 15.70 12.70 12.45 .97628 19.80 16.09 15.71
.98032 15.80 12.79 12.53 .97618 19.90 16.18 15.79
.9S021 15.90 12.87 12.61 .97608 20.00 16.26 15.87
.98011 16.00 12.95 12.69 .97598 20.10 16.34 15.95
.98001 16.10 13.03 12.77 .97588 20.20 16.42 16.03
.97991 16.20 13.12 12.85 .97578 20.30 16.51' 16.10
.97980 16.30 13.20 12.93 .97568 20.40 16.59 16.18
.97970 16.40 13:29 13.01 .97558 20.50 16.67 16.26
.97960 16.50 13.37 13.09' .97547 20.60 16.75 Hi. 34
.97950 16.60 13.45 13.17 .97537 20.70 16.84 16.42
.97940 16.70 13.53 13.25 .97527 20.80 16.92 16.50
.97929 16.80 13.62 13.33 .97517 20.90 17.01 16.58
.97919 16.90 13.70 13.41 .97507 21.00 17.09 16.66
.97909 17.00 13.78 13.49 .97497 21.10 17.17 16.74
.97899 17.10 13.86 13.57 .97487 21.20 17.26 16.82
.97889 17.20 13.94 13.65 .97477 21.30 17.34 16.90
.97879 17.30 14.03 13.73 .97467 21.40 ' 17.43 16.98
.97869 17.40 14.11 13.81 .97457 21.50 17.51 l7.06
.97859 17'.50 14.19 13.89 .97446 21.60 17.59 17.14
.97848 17.60 14.27 13.96 .97436 21.70 17.67 17.22
.97838 17.70 14.35 14.04 .97426 21.80 17.76 17.30
.97828 17.80 14.44 14.12 .97416 21.90 17.84 17.38
.97818 17.90 14.20 .97406 22.00 17.92 17.46
14.52
I' I
2144 MISCELLANEOUS ANALYSES
PERCENTAGE OF ALCOHOL BY VOLUME AND BY WEIGHT
(Continued)
Alcohol I Alcohol
Specific Specific
Gravity at Gravity at
60°
600 F.
per cent
by
per cent
by ,G~:~s 60° p~r cent
by
per cent
by
Grams
per
600 F.
volume. weight.
I 100 ml. volume. weight. 100 ml.
Alcohol Alcohol
Specific Specific
Gravity at GravIty at
60°
600 F.
per cent
by I perbycent Grams
per
60°
600 F.
per ('ent
by
per cent
by
Grams
per
volume, weight. 100 mi. volume. weight 100 mi.
- - - -----
.90839 27.40 22 45 21.75 .90300 \
31. 50 25.94 25.00
.96828 27.50 22.54 21.83 .96347 31.60 26 03 25 08
.96816 27.60 \ 22.62 21.90 .96335 I 31.70 26.11 25.16
.96323 3l.80 26 20 25.24 .95787 i
35.90 29.74 28.49 ~
.9u610 25.32'
.96298
31.90
32.00
26.28
26.37 25.40
.9;3773
.95759 II 36.00
36.10
29.83
29.92
28.57
28.65
. "' .96285
.96273
.96260
.96248
32.10
32.20
32.30
26.46
26.54
26.63
25.48
25.56
25.64
.95745
.95731
.95717 I
36.20
36.30
36.40
36 .50
30.00
30.09
30.17
30.26
28.73
28.81
28.88
28.96
32.40 26.71 25.71 .95703
.96235 32.50 26.80 25.79 .95688 36.60 30.35 29.04
.96222 32.60 26.89 25.87 .95674 36.70 30.44 29.12
.96210 32.70 26.97 25.95 .95660 36.80 30.52 29.20
.96197 32.80 27.06 26.03 .95646 36.90 30.61 29.29
.96185 32.90 27.14 26.11 .95632 37.00 30.70 29.36
.90172 33.00 27.23 26.19 .95618 37.10 30.79 29.44
.96159 33.10 27.32 26.27 .95603 37.20 30.88 29.52
.96146 33.20 27.40 26.35 .95589 37.30 30.96 29.60
96133 33.30 27.49 26.43 .95574 37.40 31.05 29.68
.:16120 33.40 27.57 26.51 .95560 37.50 31.14 29.76
.96108 33.50 27.66 26.59 .95545 37.60 31.23 29.84
.96095 33 60 27.75 26.67 .9,'i531 37.70 31.32 29.92
.96082 33.70 27.83 26.75 .95516 37.80 31.40 30.00
.96069 33.80 27.92 26.82 .95502 37.90 31.49 30.08
.96056 33.90 28.00 26.90 .95487 38.00 31.58 30.16
.96043 34.00 28.09 26.98 .95472 38.10 31.67 30.24
.96030 34.10 28.18 27.06 .95457 38.20 31.76 30.32
.96016 34.20 28.26 27.14 .95442 38.30 . 31.85 30.40
.96003 34.30 28.35 27.22 .95427 38.40 31.94 30.48
.95990 34.40 28.43 27.30 .95413 38.50 32.03 30.56
.95977 ~ 34.50 28.52 27.38 .95398 38.60 32.12 30.64
.95963 34.60 28.61 27.46 .95383 38.70 32.20 30.72
.95950 34.70 28.70 27.54 .95368 38.80 32.29 30.79
.95937 34.80 28.78 27.62 .05353 38.90 32.37 30.87
.95923 34.90 . 28,87 27.70 .95338 39.00 32.46 30.95
.95910 35.00 28.96 27.78 .95323 39.10 32.55 3l.03
.95896 35.10 29.05 27.86 .95307 39.20 32.64 31.11
.95883 35.20 29.13 27.94 .9;5292 39.30 32.72 .31.18
.95869 35.30 29.22 28.02 .95277 39.40 32.81 31.26
.95855 35.40 29 30 28.09 .95262 39.50 32.90 3l.34
.95842 35.50 29 38 28.17 .95246 39.60 32.99 31.42
.95828 35.60 29.48 28.25 .. 95231 39.70 33.08 3l.50
.95814 35.70 29.57 28.33 .95216 39.80 33.17 31.58
.95800 35.80 29.65 28.41 .95200 39.90 33.27 31.66
.95185 40.00 33.35 3l. 74 .94519 44.10 37.02 34.99
.95169 40.10 33.44 31.82 .94.502 44.20 37.11 35.07
.95154 40.20 33.53 31.90 .94484 44.30 37.21 35.15
.95138 40.30 33.61 3l.98 .94467 44.40 37.30 35.23
.95122 40.40 33.70 32.06 .94450 44.50 37.39 35.31
.95107 40.50 33.79 32.14 .94433 44 60 37.48 35.39
.95091 40 60 33.88 32.22 .94416 44.70 35.57 35.47
.95075 40.70 33.97 32.30 .94398 44.80 37.66 35.55
.95059 40.80 34.06 32.38 .94381 44.90 37.76 35.63
2146 MISCELLANEOUS ANALYSES
PERCENTAGE OF ALCOHOL BY VOLUME AND BY WEIGHT'"
(Continued)
Specific
Gravity at
Alcohol
Specific
Gravity at
I I Alcohol
60° per cent per cent Grams 60° I pe~ cent per cent Grams
60oF . by by per
I vo~me.
600 F., by per
volume. weight. 100 ml. weight.
I 100 ml.
Then
A - B = Volume of MeOH + H 2 0 in 100 grams, (C)
100 grams-%MeAc by weight;=Weight of MeOH+H 20 in 100 grams CD)
and
~ =Sp.gr. of MeOH+HzO mixture in the sample.
METHYL ALCOHOL
ceased. On a water bath distil the methyl and ethyl iodides formed into about
30 m1. of water. Wash with dilute alkali to eliminate free iodine. Separate
the heavy oily liquid which settles and transfer to a flask containing 5 ml. of
anilin. The flask should be placed in cold water, in ,dase the 'action should be
• violent, or, if necessary, the reactiofi may.be stimulated by
gently warming the flask. After one hour boil the product
with water and add about 20 m1. of a 15% solution of soda;
when the bases rise to the top as an oily layer fill the flask up
to the neQk with water and draw them off with a pipette.
Oxidize 1 m1. of the oily liquid by adding 10 grams of a mix-
ture of 100 parts of clean sand, 2 of common salt, and 3 of
cupric nitrate; mix thoroughly, introduce into a glass tube,
and heat to 90 0 C.foreight or ten hours. Exhaust the product
with warm alcohol, filter, and make up with alcohol to 100 ml.
If the sample of spirits be pure the liquid is of a red tint, but
in the presence of 1 % of methyl alcohol it has a'distinct violet
F !G. 298 . P yc- shade; with 2.5% the shade is very distinct, and still more
nometer. so with 5%. To detect more minute quantities of methyl
alcohol, dilute 5 m1. of the colored liquid to 100 ml. with water;
and dilute 5 ml. of this again to 400 ml. Heat the liquid thus obtained in porce-
lain and iminerse a fragment of white merino wool (free from sulfur) in it for half
an hour. If. the alcohol be pure the wool will remain white, but if methylated
the fiber will become violet, the depth of tint giving a fair approximate indi-
cation of the proportion of methyl alcohol present. .
FORMALDEHYDE
DETECTION OF FORMALDEHYDE 6
DETERMINATION OF FORMALDEHYDE
Hydrogen Peroxide Method. 7-To 50 ml. of normal NaOH in an' Erlen-
meyer flask are added 50 ml. of neutral H 2 0 2 (3% sol.) and 3 ml. of the formalde-
hyde solution under examination, care being taken to add the sample with the
tip of the measuring pipette near the surface of the reagents. A funnel is placed
in the neck of the flask and the mixture heated for five minutes on the steam
bath, with occasional shaking. The solution is now cooled, the funnel rinsed
down, and the excess of N /NaOH titrated by normal H 2 S0 4 , in presence of
purified litmus indicator
1 ml. N /NaOH=0.03003 gram HCHO.
6 Carry out the tests with blanks of pure ethyl alcohol and ethyl alcohol containing a
little methyl alcohol.
.1 J. Assoc. Official Agr. Chemists, Methods of Analysis (1916), page 75.
2150 MISCELLANEOUS ANALYSES
GLYCEROL
There are two recognized methods for .the determination of glycerol:
A. Acetin method, depending upon the conversiOb. of glycerol to triacetin
by means of acetic anhydride and sodium acetate ana a quantitative saponi-
fication of the triacetin. This method is recommended by the International
Committee on glycerol .as giving results nearer the truth and should be em-
ployed for crude and refined glycerines of over 40% strength.
B. Bichromate method, which is based on the fact that glycerol is com-
pletely oxidized to CO 2 and H 2 0 by K 2 Cr 2 07 in presence of H 2S0 4 • The
method is applicable for determination of glycerines in soap lyes. .
The following procedures are taken from the report as recommended. s
Reagents Requir.ed
(A) Best Acetic Anhydride.-This should be carefully selected. A good
sample must not require more than 0.1 ml. n\)l'mal N aOH for. saponification of
the impurities when a blank is run on 7,5 m!. Only a slight color should
develop during digestion of the blank.
The anhydride may be tested for strength by the following method: Into a
weighed stoppered vessel, containing 10 to 20 ml. of water, run about 2 ~l. of
the anhydride, replace the stopper and weigh. Let stand with occasional
shaking, for several hours, to permit the hydrolysis of all the anhydride; then
dilute to about ·200 mI., add phenolphthalein and titrate with N /1 NaOH.
This gives the total acidity due to free acetic acid and acid formed from the
anhydride. It is worthy of note that in the presence of much free anhydride a
compound is formed with phenolphthalein, soluble 'in alkali and acetic ILcid,
but insoluble in neutral solutions. If a turbidity i$ noticed toward the end
of the neutralization it is an indication that: the itnhydride is incompletely
hydrolyzed and inasmuch as the indicator is. withdrawn from the solution,
results may be incorrect. '
Into a stoppered weighing bottle containip.g a known weight of recently
distilled aniline (from 10 to 20 mi.) measure about 2 mi. 0.[ the sample, stopper,
mix, cool and weigh, Wash the contents into about 200' ml:- of cold water,
and titrate the acidity as before. This yields the acidity due to the original,
preformed, acetic acid plus one-half the acid due to anhydride -(the other half
having formed acetanilide); subtract the second result from the first (both
calculated to 100 grams) and double the result, obtaining the ml. N /1 NaOH
per 100 grams of the sample. 1 m!. N /N aOH eql).als 0.0510 anhy.dride.
(B) Pure Fused Sodium Acetate.-The purchased salt is again completely
fused in a platinum,' silica or nickel dish, avoiding charring, powderE;ld quickly
and kept in a stoppered bottle or desiccator. It is most important that the
sodium acetate be anhydrous. I
alkali running into the mixture. When this point is reached the sides of the
flask are washed down with carbon dioxide-free water and the alkali subse-
quontly added drop by drop, mixing after each drop until the desired tint is
obtained. I
N ow run in from a burette 50 ml. or a calculated 'excess pf N /1 N aOH (D)
and note carefully the exact amount. Boil gently for 15 minutes, the flask
being fitted with a glass tube acting as a partial cqndenser. Cool as quickly
as possible and titrate the excess of N aOH with N /1 acid (E) until the pinkish
yellow or chosen enclo-point color just remains.lO A further addition of the
indicator at this point will cause an' increase of the pink color; this must be
neglected, and the first end-point taken.
From the N /1 N aOH consumed calculate the percentage of glycerdl (in-
cluding ~cetylizable impurities) after making the correction for the blank
test described below.
1 ml. N/l NaOH=O.03069 gram glycerol.
I
The coefficient of eXRansion'for normal i2olutions is 0.00033 per ml. for each
degree centigrade. A correction should be made on this account if necessary.
Blank Test.-As the acetic anhydride and sodium acetate may contain
impurities which affect the result, it is necessary to make a blank test, using
the same quantities of acetic anhydride, sodium acetate and water as in the
analysis. It is not necessary to filter ,the solution of the melt in this case,
but sufficient time must be allowed for the hydrolysis of the anhydride before
. proceeding with the neutralization. After neutralization it is not necessary
to add more than 10 ml. of the N /1 alkali CD), as this represents the excess
usually present after the saponification of the average soap lye crude. In
. determining the acid equivalent of the N /1 N aOH, however, the entire amount
taken in the analysis, 50 mI., should be titrated after dilutjon with 300 ml.
water free from carbon dioxide and without boiling. .
Determination of the Glycerol Value of the Acetylizable Impurities.--4The
total residue lJ,t 1600 C. is dissolved in 1 or 2 ml. of w~ter, washed into the acetyl-
izing flask and evaporated to dryness. Then add anhydrous sodium acetate
and acetic ~hydride in the usual amounts and proceed as described in the
regular analysis. After correcting for the blank, calculate the result to glycerol.
Instructions for
.
Calculating. the Actual Glycer.ol- Content
(1) Determine the apparent percentage of glycerol in th~- sample by the
acetin process as described. The result will include acetylizable impurities
if any are present.
(2) Determine the total residue at 160 0 C.' .
·(3) Determine the acetin value of the residue ,at (2) in terms of glycerol.'
(4) Deduct the result found at (3) from the percentage obtained at (1)
and report this corrected figure as glycerol. If volatile acetylizable impurities
are present these are included in this figure.
Trimethylenglycol is more volatile than glycerine and can therefore be con-
centrated by fractional distillation. An approximation lo the quantity can
10 A precipitate at this point is an indication of ·tire presence of iron or alumina, and
high results will be obtained unless a correction is made as described below.
MISCELLANEOUS ANALYSES 2153
be obtained from the spread between tlie acetin and bichromate results on such
distillates. The spread multiplied by 1.736 will give the glycol.
boiling water for two hours and keep protected from dust and organic vapors,
such as alcohol, until the titration is completed. -Add from a weighing bottle a
slight excess of the ferrous ammonium sulfate (C)! making spot tests on a
porcelain plate with the potassium ferricyanide (F)r Titrate back with the
dilute bichromate. From the amount of bichroma~e reduced calculate the
percentage of glycerol.
1 gram glycerol = 7.4564 grams bichromate.
1 gram bichromate =0.13411 gram glycerol.
The percentage of glycerol obtained above includes any mcidizable impurities
present after the purification. A correction for the non-volatile impur.ities
may be made by running a bichromate test on the residue at 160° C.
For complete analysis see Journal of Industrial and Engineering Chemistry,
Sept., 1911, pages 679-683.
NOTEs.-(l) It is important that the concentration of acid in the oxidation mixture
and the time of oxidation should be strictly adhered to.
(2) Before the bichromate is added to the glycerine solution it is essential that the
slight excess of lead be precipitated with sulfuric acid, as stipulated.
(3) For crudes practically free from chlorides the quantity of silver carbonate may
be reduced to one-fifth and the basic lead acetate to 0.5 mI.
(4) It is sometimes advisable to add a little potassium sulfate to insure a clear filtrate.
Apparatus
The absorption requires a O.I-cu. ft. gas mete~ and two activated carbon.
tubes complete. The carbon tubes are cylindrical containers, 1.5 in. in diam-
eter and 12 in. in length over all, fitted with a 20-mesh screen in one end, with
I-in. breasts and screw· caps on both ends. Four connecting/caps consisting of
I-in. screw caps carrying 156 x i-in. copper tubes ~ldered in the centers and
_,
11 By Arthur L. Davis, Standard Oil Company (Indiatui), Wood River, Ill.
MISCELLANEOUS ANALYSES 2155
fitted with I6-in. rubber gaskets are used on each carbon tube during the
. absorption proper.
The distillation of the enriched activated carbon with cresol to remove the
light oil is made, using a 500-ml., round-bottom, side-neck, Pyrex distilling
flask; a 4-in. ring burner; a 200 0 C. thermometer; a "Barrett standard"
condenser; a 200-ml. separatory funnel graduated to 100 mI.; U. S. P. cresol;
and sodium hydro~de solution of specific gravity 1.10 to 1.15.
Procedure
Absorption.-Place 300 ml. of activated carbon in each of the two carbon
tubes. Screw' a connecting cap with rubber gasket to each end of each tube,
and tighten with a pair of combination pliers.
,Connect the inlet of the wet gas meter to the gas sampling tube and allow
gas to pass through for several minutes to insure a thorough displacement of air.
Turn off the gas, set the indicating hands of the meter to the zero position.
(Ca.ution: See that the hands are reasonably loose before turning so that the
driving shafts will not be twisted.) Connect the top of the first carbon tube to
the outlet of the gas meter, the bottom of the first tube to the bottom of the
second tube, and either allow the residual gas coming from the top of the second
tube to escape or be conducted away by an attached exit tube. Keep the tubes
in a vertical position.
Pass an amount of gas containing approximately 30 to 40 ml. of light oil
through the absorption tubes, at the rate of 10 to 15 cu. ft. per hr. At the
completion of the absorption note the volume of gas passed, the temperature
of the gas, and the barometric pressure.
Distillation.-Place the enriched carbon from the first tube in a 500-m!.,
l'ide-neck, Pyrex distilling flask, add 125 ml. of cresol, mix thoroughly by shaking
in the flask, insert the ther~ometer in neck of the flask, and connect to a " Bar-
rett standard" condenser. Heat carefully with the ring burner and distil at a
, moderate rate till the temperature of 1800 C. is reached. Collect the distillate
in the 200-m!' separatory funnel, which is graduated to 100 ml. Add 125 ml. of
sodium hydroxide solution, specific gravity 1.10 to 1.15 (8 to 13%); and agitate
thoroughly. Care should be taken to release the pressure of the gas which is
often built up. Allow to stand for a few minutes, run off the sodium hydroxide
solution, and record the volume of light oil obtained.
Repeat the foregoing procedure with the second carbon tube. Record the
volume of light oil obtained separately.
The sum of the two volumes of light oil obtained above gives the total
volume of light oil from the volume of gas passed at the temperature and
barometric pressure recorded. .
Calculation.-The amount of light oil present in the g~s (volume corrected
to 15.50 C. (760 mm.)) is calculated as follows:
Total volume of light oi1XO.264 = Gal. light oil per 1000 cu. ft. gas.
288.5 Bar. press.
Cu. ft. gasX 273+temp. 0C. X 760 .
Reprinted from J. Ind. Eng. Chern., 15, 689 (1923).
2156 MISCELLANEOUS ANALY'SES
Gal. light oil per 1000 cu. ft. X 12 = Gal. light oil per ton of coal. (One ton of
coal produces approximately 12,000 cu. ft. of gas.) 1
(c) Stir filtrate from above for two seconds ',vith a circular motion, and let
stand thirty seconds. Decant carefully, and compare residue with a standard
sampls of the same grade of starch, treated in the same manner, at the same
time.
Moisture.-Dry approximately 5 grams to constant weight at 100° C.
Loss in weight is calculated as moisture. .
Ash.-Weigh out 1 gram of the sample; transfer to a platinum crucible;
ignite to constant weight.
Acidity.-Weigh out 20 grams on a watchglass. Transfer to a 250-ml.
porcelain dish. Add approximately 200 ml. of water, and stir until starch
forms a smooth cream. Titrate with N /10 caustic soda, using phenolphthalein
as indicator until a faint pink color persists for twenty seconds.
Alkalinity.-Method is same as for acidity, except that sample should be
titrated with N /10 sulfuric acid until the pink coloration just disappears.
Test for the Presence of Alkaline Starch or Alkali.-Add a few drops of
phenolphthalein solution to water in a white porcelain dish; sprinkle in a little
of t;.he dry sample. If some of the grains turn pink, it indicates that the sample
is a mixture of raw starch and alkaline starch or alkali. If all the particles
turn pink, the material is an alkaline finished starch.
Viscosity.-Take 12 grams of the sample in 300 ml. of water, bring to boil,
then boil for ten minutes with constant stirring in an aluminum beaker over
a naked flame. Determine viscosity at 100° C. with 200 ml. of the liquid in
a Scott viscosimeter, standardized as follows: Calibrate to hold 200 ml. of
water at 100° C. Determine viscosity on first 50 ml. withdrawn. Express
result as a ratio. (Starch viscosity) -:- (water viscosity) = viscosity number.
Converted Starches.-Owing to the number and variety of the products
on the market, their classification is a matter of some difficulty. The method
.of conversion, however, may be used as a means, and thus the following classes
can be arranged: (1) Thin boiling starch, mainly used for tub sizing, converted
by treatment with weak acid solutions by the" in suspension" or " drying in "
process. (2) Partly dextrinized starches, and dextrins, made by damping
the starch with small quantities of acid, and converting at a high temperature
by the dry process. (3) Oxidized starches. (4) Acetylated starches. (5)
Alkali-treated starches, including products consisting of raw starch mixed with
certain quantities of various alkalies and alkaline salts. (6) Mixtures of starch
products and mineral fillers. (7) Mixture of starch products with glue and
casein or other sizing agents. (8) Products in solution or paste form.
Sampling, Moisture and Ash are the same as for raw starch.
Acidity and Alkalinity.-First method same as for raw starch.
Total Acidity.-Mix 10 grams of the sample in 200 ml. of water and heat
on a water bath for five minutes after reaching 75° C.; cool and titrate. This
gives the acidity within the starch granule.
Viscosity.-In determining viscosity the starches are divided into three
classes: (a) Low boiling, or laundry starch; (b) medium thin boiling; (c) high
thin boiling, with a decrease in viscosity from (a) to (c). For (a) the viscosity
should be determined in a 10% solution; for (b) in a 20% solution; for (c) in a
30% solution. Otherwise the metbod is the same as for raw starch.
Added Materials.__'!These are tested for by the ordinary methods used in
separating and identifying mineral fillers, metallic salts, glue, casein, and other
2158 MISCELLANEOUS ANALYSES
TANNIC ACID 12
I _~
BLEACHING POWDER
Sampling.-Twenty per cent of the packages in the shipment should be
sampled. The sample shall be taken by boring or cutting a two-inch hole
through the side of the wooden cask or iron drum midway from the ends, or
through the head near the center. The sampler, which consists of a stout iron
scoop about one and a half inches wide and eighteen inches long, shall be in-
serted for two or three inches, withdrawn, and the bleach removed discarded.
The sampler shall be inserted again as far as it will go, removed, the bleach
2160 MISCELLANEOUS ANALYSES
transferred to a glass fruit jar, and the jar closed until. the next cask is sampled.
The portions from the several casks shall all be mixed together to make on!:'
composite sample representing the shipment.
Available Chlorine.-Ten grams of bleaching powCler from the well-mixed
sample is weighed out and triturated in a mortar, itdc\ing small quantities
of distilled water at a time and grinding thoroughlY'8:fter each addition until
the powder is converted to a voluminous paste free from dense particles. The
liquor after each trituration is poured off into a lOOO-ml. graduated volumetric
fiallk, and the paste is finally washed into flask. Fill the flask to the mark
with distilled water, shake well and draw off an aliquot portion by means of a
50-ml. pipette for analysis, observing the precaution that the liquor drawn off
contains its proportion of suspended matter. The available chlorine is detsr-
mined by titration with N /10 sodium arsenite solution, using starch iodide paper .
as an indicator.
NOTE.-The NjlO sodium arsenite solution is made by drying arsenic trioxide, which
answers Krauch's requirements for purity, in a desiccator over night, 4.95 grams being
weighed out and dissolved in 200 m!. distilled water with 20 grams of pure sodium
carbonate, the solution cooled and made up to 1 liter, then standardized with resublimed
iodine.
One gram of potato starch is mixed with distilled water to a cream and thelJ.
poured slowly into 100 ml. of boiling water and the liquid allowed to boil gently
for two minutes when 1 gram of c.p. potassium iodide, dissolved in a little
distilled water, is poured in. The liquicfis then thoroughly mixed, allowed to
settle and strips of unsized paper are soaked in the solution and hung up to dry,
out of contact of fumes. . ,
Chlorates.-The solution in which the available chlorine has been deter-
mined by means of N /10 sodium arsenite is slightly acidified with sulfuric a,cid,
the solution transferred to a narrow-mouth flask, and a quantity of pure 'am-
monium ferrous sulfate is added in excess of that required to reduce all chI orates
to chlorides. The flask is then stoppered with a 2-hole rubber stopper, which
is fitted with a small separatory funnel through one opening, and a small glass
tube ending in a short piece of rubber tubing (which rriay be clamped tight when
needed) is fixed in the other opening. The liquid IS then brought to boiling
and 15 ml. of 1.29 * sulfuric acid is added drop by drop, through the funnel.
When acid is all in, close opening in funnel and. clamp rubber tubing on vent,
cool, and titrate back the excess of ferrous ammonium sulfate with N /10
potassium permanganate.' . -...
Total Chlorides.-The slight pink color produced by the permanganate is
removed by the addition of a drop or two of N /10 ferrous ammonium sulfate,
and an excess of N /10 silver nitrate solution is added.
The silver chloride is filtered off, washed well and the excess silver nitratl)
is determined by means of N /10 potassium thiocyanate, the iron in solution
acting as an indicator. See chapter on Chlorine, Yol. 1.
NOTE.-The chlorine as chlorates plus the available chlorine subtracted from the
total chlorine as chlorides gives the chlorine present originally in the bleach as chlorides.
Determin.ation of Bases and Silica..,-To 2 grams of bleaching powder add
25 ml.1.12* HCI and 50 ml. of 3% H 2 0 2 • Dige:;t.and evaporate on steam bath
* Specific gravity.
MISCELLANEOUS ANALYSES 2161
or hot plate. Dehydrate the silica and determine the silica, alumina, iron,
calcium oxide, and magnesia as in lime analysis.
Free Lime.-To 100 ml. of bleach liquor made up for the available chlorine
determination add 25 ml. hydrogen peroxide, bring to boil and oxidize the
hypochlorite. The solution is then titrated with standard. acid. At the same
time the acidity of the H 2 0 2 used should be determined and a correction maqe
for the amount used in destroying the hypochlorite. The free lime is calculated
from the acid consumption.
Quality of Powder for Producing Bleach Liquor.-Sixty grams of bleaching
powder is weighed out and added to 1000 ml. distilled water in a glass jar 4t
inches in diameter and 5 inches deep. The mixture is stirred vigorously by
means of a glass rod bent so that it just clears the bottom and sides of the jar,
driven by power from electric or water mot,or at 250 r.p.m. for exactly 15
minutes. The jar is then placed on a bench free from vibration and the time
required for the dregs to settle to the bottom noted; 25 ml. of the clear bleach
liqu'or is then measured out and tested in the usual way for available chlorine.
Determination of Wood and Coarse Impurities.-l00 grams of the powder
is triturated in a porcelain mortar, washed through a 60-mesh sieve, residue
is dried and weighed.
Proximate Analysis for Sand and Grit.-To 10 grams of the powder add
75 ml. of 1.12 hydrochloric acid, and warm until all lime salts are dissolved,
dilute to 200 ml. with distilled water, pour off the supernatant liquid, again
dilute to 200 ml. with distilled water, filter, wash, and ignite the residue.
NOTE.-If preferred, a potassium iodide-starch solution preserved with a drop of
CCI. or CS 2 may be used as the indicator in the determination of available chlorine by
testing a drop of the bleach liquor with a drop of the indicator on a pill tile rather than
the use of the pota~sjum iodide-starch paper. If desired, soluble starch, which dissolves
readily in hot water, may be used in the preparation of this indicator.
ANALYSIS OF CLAY 13
Chemical Analysis.-The sample should be finely ground, if necessary a
large quantity, 2 to 10 pounds, should be weighed and dried on steam table,
weighed again and loss in moisture calculated to perp-entage; it may then be
finely ground.
A weighed quantity, 1 to 10 grams, should now be dried at 100 0 C. to
constant weight, this loss calculated to percentage and added to that from the
first drying and is the moisture. A weighed portion, 1 to' 2 grams, of the
moisture-free powder is heated, in a covered crucible, to bright redness for 10
minutes, cooled and reweighed; loss = combined water and organic matter.
Or a weighed quantity is put in a boat which is heated in a tube in a current
)3 By J. M. Wilson, Lefax, 9-196, June, 1919. By courtesy of Lefax, Philadelphia.
2162 MISCELLANEOUS ANALySES
of dry air and drawn through a weighed Ca.C12 tube; gain in weight = combined
moisture. I
Mix 1 gram dry clay with 5 g. Na 2CO a+t g. KNqa in a platinum crucible,
fuse 30 to 40 minutes, cool, detach melt from crucible. Wash crucibliJ with
H 20, then HCI, theJ1 H 20, pouring all washings into ,:1 porcelain dish, and add
the melt. Cov~r dish, add HCI little at a time until in considerable excess.
Evaporate dry, bake 1 hour at 1200 C. Cool, moisten well with HCl. Evap-
orate and bake again. Cool, moisten HCl, take up in H 20, boil 10 minutes,
filter, wash with hot HCl (1 : 1), cold H 2 0, hot HCl, cold H 2 0, then 5 to 6
times hot H 20, ignite in a weighed platinum crucible and hold.
Evaporate the filtrate and washing in a dish, bake as above, moisten with
HCl, evaporate and bake again. Take up HCl+H 20 filter and wash as above.
Burn this filter in same crucible, cool, weigh = crude Si0 2 • . Moisten with few
drops' H 2S0 4 (1 : 4), then H 2 0 to make a thin mud; add enough HF to fill
crucible i full, evaporate at gentle heat; heat cautiously over lamp-to avoid
spattering-raising heat slowly to full heat of lamp. Loss in weight = Si0 2.
If any residue remains in the crucible, fuse with KHS0 4. Leach melt with
H 20 and add to main filtrate.
Boil filtrate from silica, add 1 to 3 mi. HN0 3, boil 10 minutes. Add NH 40H
in slight excess, boil 3 minutes, let settle in Warm place, decant through a filter.
Add hot water to precipitate, stir, let settle, filter, dissolve precipitate in
HCI (1 : 1). Dilute to 250 m!., boil, ~eprecipitate as above, repeat again,
wash thoroughly with hot H 20. Ignite and weigh precipitate = Fe 2 0aAl 20 3-
(P 20 5 ·Ti0 2). '
Fuse with KHS0 4, leach with H 20, reduce in Jones reductor. Titrate F'e
with KMn04. Fe/0.7=Fe 203.
Fuse 2 to 5 grams dry sample with 5 wts. Na 2COa+t wt. KN0 3 • Leach
melt with H 20, filter, wash with H 20+Na 2CO a. Save residue for Ti. Acidify
filtrate with HN0 3, boil off CO 2, add Fe 2 CIs solution (free from Phos.), boil.,
Add NH 40H, filter, take up precipitate in HNO a. Boil well and precipitate
the phosphorus with ammonium molybdate. Deter~ine phosphorus.
PX2.29=P 20 S•
(AI20 a+Fe203+ P 205+Ti0 2) -(Fe20a+P205+Tj02) = AI 2 0 a.
Evaporate united filtrates from A1 2 0 3, etc., to 100 ml. Make ammoniacal,
add Br2 in excess, then 60 to 70 ml. NH 40H, boil 20 minl,!tes, filter, wash with
hot H 20. Ignite and weigh Mn a04. Mn 3 04XO.93006=M"nO;--.
Acidify filtrate from Mn02, evaporate dry, take up in HCI+H 20 (volume
150 mI.), boil, add NH 4 0H till neutral, then ammonium oxalate. Boil 10-
minutes, cool 2 hours, decant close as possible through filter. Dissolve precipi-
tate in HCl, dilute to 100 mI., add ammonium oxalate, boil. Precipitate with
NH 40H, boil 10 minutes, filter after 2 hours, wash, twice with col<,i H 20, then
5 or 6 times with hot H 20. Ignite strongly, weigh while still warm, repeat
ignition and weighing = CaO. Or dissolve precipitate in least quantity HCl
(1 : 1) through filter into an Erlenmeyer flask. Wash with H 20, add 25 ml.
H 2S0 4 (1 : 4), heat to 800 C., titrate with KMn04. Fe value, KMn04/2 = CaO.
Evaporate combined filtrates from CaO to smallrvolume (15 to 20 mI.), add
1 gram ammonium phosphate (NH 4)2HP0 4, ·then .NH40H till solution is
neutral. Cool under tap, stir till precipitate forms, set aside 15 minutes,
MISCELLANEOUS ANALYSES 2163
stirring occasionally. Add! volume NH 40H, stir, let stand '4 hours, filter
but do not wash. Dissolve through filter in HCI into vessel in which precipita_
tion was c.onducted. Wash filter with H 20, reprecipitate with NH 40lI as
before, filter on weighed Gooch felt. Wash with (NH 4NO g, 60 g.; H 2 0, 80 ml.;
NH 4 0H (1 : 3), to 600 mI.), dry at 100° C., ignite cl1refully till a glow passes
over the precipitate, c601 and weigh Mg 2P 20 7•
burette reading will then give the number of milligrams 6f Ti in the clay from
which the percentage may be calculated. I
State of Silica.-A-Combined with bases in the clay, B-Hydrated silica,
C-Sand or silicates. Moisten 2 grams dry powder with H 2 0, add 20 ml.
H 2S0 4 (conc.), cover dish, heat 10 hours, remove cover,1 evaporate dry, cool,
take up in H 2 0, filter, wash with H 20, ignite and weigb.=cA+B+C.
Boil with Na 2 CO a solution, allow to settle, decant through a weighed Gooch
felt or alundum crucible, repeat boiling and decantation until a portion of the
filtrate gives no precipitate with NH 4 CI on standing. Transfer residue to
filter and wash thoroughly with hot H 2 0" acid, then hot H 2 0. Ignite and
weigh = C. A and B are in filtrate.
Boil 4 to 5 grams dry powder with strong solution of Na 2 CO S , filter, wash-
with hot ·H 2 0. Acidify filtrate with HCI, evaporate dry, bake at 120 0 C.;
Gool, moisten with HCI, evaporate dry, bake again. Take up in Hel and H 2 0,
filter, wash, ignite; weigh as Si0 2 =B. (A+ B+ C) -(B+ C) = A.
If A+B+C=total Si0 2 found on analysis, C is quartz. If A+B+C is
greater than total Si0 2 , C is in form of silicate and weight of Si0 2 as silicates =
A+B+C-total Si0 2 found in 1 gram. .
ANALYSIS OF GLASS 14
precipitation is complete. Preserve the filtrate for the estimation of iron, etc. I
Dissolve the lead sulfide in dilute nitric acid (1 : 4)j Filter this solution to
remove the asbestos used in the Gooch. Precipitate lead sulfate from this
solution by adding dilute sulfuric acid with constant stirring. Then add 100%
by volume of 98% ethyl alcohol to the cold 'Solution\. Allow the precipitate
to settle for an hour and filter on a weighed Gooch previously ignited at a red
heat. Wash with 50% alcohol. Heat the residue sl~wly and ignite at a dull
red heat until the residue is white. Cool and weigh as lead sulfate. Calculate
the percentage of PbO by using the proper factors which may be found'in the
table below.
Determination of FezOa and AlzOa(RzOa).-Boil this solution with 2 or
3 illl. of nitric acid to remove the sulfur and oxidize the iron. Cool the solution.
Add about 5 grams of ammonium chloride and then add ammonium hydroxide
until the solution is slightly ammoniacal. Boil cautiously for about 5 minutes.
Allow the precipitate to settle. Filter and wash with hot water or hot 2%
ammonium nitrate solution. Preserve this filtrate for the determination of,
calcium. Ignite the residue in a weighed platinum crucible. Cool and weigh
as FezOa and AlzO a. Fuse this residue with potassium bisulfate and dig~st with
dilute sulfuric acid. Transfer to a 250-ml. Erlenmeyer flask. Pass in hydrogen
sulfide gas to reduce the iron. Then fit a B.unsen valve to the flask and boil to
a small volume to drive off the excess hydrogen sulfide. Cool and add a pinch
of sodium bicarbonate and titrate with stlj.ndard N /200 potassium permanganate
solution. From the amount of standard solution used calculate the percentage
of iron oxide. Deduct the percentage of iron oxide from (FezOa+AlzOa) and
the difference is the per cent AlzO a.
Determination of CaO.-Bring the solution to boiling. While hot add 10 to
20 ml. of hot saturated ammonium oxalate solution with constant stirring.
Then add about 5 m!. of ammonium hydroxide. Allow to stand for several
hours or preferably over night. Filter off the calcium oxalate and wash with
a 2% ammonia water. Preserve the filtrate for the determination of magnesia.
Ignite the residue at full. blast in a previously ignited and weighed platinum
crucible to constant weight. Weigh as calcium oxide.
Determination of MgO.-Concentrate the, solution by boiling and while
hot add 10 to 20 m!. of saturated microcosmic salt solution with constant
stirring. Then add ammonia and stir vigorously, taking care not to strike
the sides of the beaker. Allow the precipitate to settle for several hours or
preferably over night. Filter and wash with a 2% solutiOn of-ammonia water.
Dry the residue and filter paper. Ignite slowly at first and then at full heat
in a weighed porcelain or platinum crucible, until constant weight. Cool and-
weigh up as magnesium pyrophosphate. Calculate the percentage of MgO
by using the factor found in the table below. If large amounts are to be ignited,
it is best to ignite the residue and paper separately in a porcelain crucible.
Determination of the Alkalies.-The method in general use is that of Sulli-
van and Taylor. Weigh out 1 gram of finely powdered glass in a platinum
crucible of about 40 ml. capacity, moisten with water and add 2 grams of oxalic
acid crystals and enough 48% hydrofluoric acid to fill the crucible half full.
This is evaporated to dryness and warmed to a temperatule which is just high
enough to expel the excess oxalic acid. When al,H!le_acid'is expelled, the cruci-
ble is cooled and the evaporation repeated twice more with oxalic acid and water,
MISCELLANEOUS ANALYSES 2167
The amount of oxalic acid used is about 5 grams in all. After the third evapora-
tion when all the excess of oxalic acid is expelled, the remaining oxalates are
taken up with hot water~ allowed to cool and filtered. The residue which
consists of lead oxalate and calcium oxalate is filtee;ed off. The filtrate is
evaporated to dryness in a platinum dish and then heated over a free flame
until oxalates are decomposed to carbonates. The insoluble carbonates are
filtered off and the alkali carbonates are in the filtrate. The filtrate is made
acid with hydrochloric and' evaporated to dryness to remove traces of silica.
The filtrate is evaporated to dryness in a platinum dish, the sodium chloride
and potassium chloride heated to incipient fusion, cooled and weighed. This
gives the weight of the combined alkali chlorides. The chlorides are dissolved '
in about 25 m!. of water and 4-5 m!. of chloroplatinic acid (1 gram Pt in 10 m!.)
added and the solution evaporated to a syrupy consistency. The insoluble
potassium chloroplatinate is filtered on a Gooch. Wash by decantation with
85% ethyl alcohol until the washings are colorless. The Gooch is dried at 130 0
C. to constant weight. ' The potassium chloroplatinate is computed to' potas-
sium. chloride, and the sodium chloride found by difference from the weight of
the combined alkali chlorides. Both potassium chloride and sodium chloride
are computed to their respective oxides.
Determination of B 20 a.-Before proc~eding with any of the determinations,
boric acid must be removed. This is done by adding 25 ml. of methyl alcohol
in both dehydrations of silica. Then proceed as usual. The determination
of B 20 a must be made as follows on a separate sample of glass. Fuse.5 gram
sample with about 4 grams of sodium carbonate. Take up with 20--30 ml.
dilute hydrochloric acid and add a few drops of nitric acid to oxidize the iron.
Place the whole in a 250 ml. round bottom flask. Heat nearly to boiling and
add dry precipitated calcium carbonate in moderate excess. * Connect with a
reflux condenser and boil vigorously for ten minutes to remove CO 2 from the
solution. Filter off precipitate through a small Buchner funnel, washing several
times with hot water. Keep the volume below 100 ml. Return the filtrate
to the flask and add a small amount of calcium carbonate and again heat to
boiling. Connect up to a filter pump, through a splash trap, and continue
suction cautiously, until boiling has nearly ceased, to remove the last traces of
' \ CO 2 in the solution. Cool to room temperature·and filter if necessary. Add
4-5 drops of phenolphthalein indicator and run in slowly N /10 standard sodium
hydroxide solution until the liquid is decidedly pink. Add about 1 gram of
mannite or 10 ml. glycerol and shake. The pin!>: color will disappear. Then
run in more standard NaOH till pink and then more glycerol. Continue
this until the pink is permanent. The glycerol must be neutral. The standard
N aOH should be standardized against a standard sample of B 2 0 a• The percent-
age B 2 0 a can then be computed.
This method is unsatisfactory when zinc and lead are present in the glass as
high results are obtained. Fuse a half-gram sample with 4 grams of sodium
carbonate. Digest with 20--30 ml. of hot water and when the 'melt is entirely
decomposed filter off any insoluble oxides. After washing, transfer the filtrate
to a 250-ml. round bottom flask. A~d about 7 m!. concentrated hydrochloric -
acid. Heat nearly to boiling and add a moderate amount of calcium carbonate
in excess; from here proceed as above from asterisk.*
2168 MISCELLANEOUS ANALYSES
spattering of the residue) to low redness or to the point where the chloridp.s
just begin to fuse. Cool and weigh and note as combined chlorides.
The precipitate in the platinum dish should then be taken up with the
smallest possible amount of water and to it add seven'teen times the weight of
the combined chlorides, of a solution of chloroplatinib acid (1 gram platinum
to 10 mI.). Evaporate this solution to a thick syrup, and take it up with
25 ml. of 95% alcohol and filter through a tared Gooch crucible and wash with
95% alcohol until the washings are colorless. It is then dried at about 160 0 C.,
cooled and weighed. The washings and residues should be preserved for the
recovery of the platinum.
K 2PtCl 6 X 0.307 = KCI.
KCIXO.926XIOO = 01 K CO
0.25 /0 2 3·
When the Sulfates are Over l%.-To the 25 ml. as obtained above and
heated to boiling add barium chloride in slight excess. Let it settle and filter.
Then to the filtrate add ammonium carbonate in excess to precipitate the
excess of barium. Filter. Evaporate this filtrate to dryness and ignite very
carefully to decompose the excess ammonium carbonate and fuse the combined
chlorides. Then proceed as outlined above when the sulfates are not in excess
of 1%.
The potash residues and washings should be s~ved and the platinum re-
covered as follows: The washings and residues are 'combined and dissolved in
hot distilled water. Add sodium formate and sodium hydrate and heat the
solution until the platinum is precipitated. Filter and ignite, cool and weigh.
Dissolve this residue in aqua regia and evaporate with successive additions of
hydrochloric acid until ,all the nitric is expelled. Then pass chlorine through
the solution and evapo'rate to a syrup to decompose the excess hydrochloric
acid. Filter and ignite and weigh and correct for the amount not dissolved.
Make up the solution with distilled water so that 10 ml. contains 1 gram of
platinum. ,
Feldspar, Kaolin and China Clay.-These are the chief sources of alumina
in glass batches and' their frequent examination i~ ~o be encouraged. They all
contain silica, iron and alumina but in different proportions. Kaolin and
China clays are examined for moisture whereas feldspar usually is not. There-
'fore the same methods of analysis may be carried out for these three materials.
Not more than 0.5 gram. sample should be weighed out. (This is done
because the alumina precipitate is so voluminous_ that if larger samples are
taken the precipitate is too bulky to handle weHTnd losses may occur.) This
is mixed thoroughly with about four grams of sodium potassium carbonate in a'
/'
MISCELLANEOUS ANALYSES 2171
platinum crucible and fused. Digest the fusion in dilute hydrochloric acid
and distilled water. Then transfer to a porcelain evaporating dish and evapo-
rate to dryness and bake for about half an hour. Take it up with hot distilled
water and dilute hydrochloric acid and filter. Wash the residue with hot water
until entirely free from chlorides. Preserve the residue. Make a second dehy-
dration on the filtrate and digest as above, filtering and washing. Combine
the residues from the two filtrations and ignite them slowly at first in a clean
platinum crucible. When the residue is snow white and to constant weight, cool
and weigh. Then moisten the residue with dilute sulfuric acid and make the
crucible three quarters full of hydrofluoric acid and evaporate to dryness.
(Sometimes it may be necessary to make a second hydrofluoric evaporation.) ,
Ignite, cool and weigh and report the loss in weight as silica. The residue
remaining in the crucible may be fused with potassium bisulfate or sodium potas-
sium carbonate and the fusion added to the filtrate remaining from the silica.
To this solution then cautiously add ammonium hydroxide until it is just
alkaline and a dense voluminous precipitate comes down. Transfer the beaker
to a .flame and boil the solution for a few minutes. Allow the precipitate to
settle and then filter and wash with hot distilled water. Preserve the filtrate.
After the residue is thoroughly washed, place a clean beaker under the funnel
and pierce a hole in the paper and wash the residue through with hot water
and dilute hydrochloric acid. Be sure to wash the paper thoroughly because
any residue remaining and lost will make the results inaccurate. The residue
will be dissolved in dilute hydrochloric acid and reprecipitated with ammonium
hydroxide, boiled, filtered and washed -as described above. The residue is then
ignited in a tared platinum crucible, cooled and weighed and reported as (R 20 a),
4-120a' Fe 20a. This residue should then be fused with potassium bisulfate or
sodium potassium carbonate and the fusion taken up with dilute sulfuric acid.
The iron may then be reduced by a ch~ice of methods and titrated with a stand-
ard solution of potassium permanganate. The filtrate from the second pre-
cipitation of alumina should be combined with the main solution.
The solution is then brought to boiling and, while hot, add about 20 m!. of a
saturated ammonium oxalate solution and stir constantly. Then add a few
m!. of ammonium hydroxide. Allow the precipitate to settle for several hours
and then filter. Wash well with a 2% solution of ammonia water. Ignite the
residue at full blast, in a tared platinum crucible, until the weight is constant.
Cool and weigh as calcium oxide and report it as such.
Concentrate the filtrate by boiling and while hot add a few m!. of micro-
cosmic salt solution and then ammonium hydroxide. Stir vigorously and
constantly. Allow the precipitate to settle for several hours and then filter.
Dry the residue and paper before ignit.ing. Then ignite in a tared platinum
crucible, cool and weigh as magnesium pyrophosphate. Calculate the per-
centage of MgO by using the factor 0.362.
If the alkalies in feldspar are desired by other than the difference method,
a separate sample may be used and the alkali determined as previously de-
scribed.
For moisture in kaolins and clays a five gram sample is weighed out and
placed in a tared porcelain dish and transferred to a hot plate or oven and
heated to about 110° C. for several hours or over night. The loss in weight is
reported as moisture.
2172 MISCELLANEOUS ANALYSES
finally at full· blast. Cool and weigh as calcium oxide. Compute to calcium
carbonate by the factor 1.7847.
Determination of M agnesium.- The magnesium is determined in the filtrate
from the calcium. Boil off the ammonia 'and, while the solution is still hot,
add a slight excess of microcosmic salt with constant stirring. Then make the
solution alkaline with ammonia and stir. Allow the precipitate to settle over
night and filter and wash with a6 dilute ammonia solution. Dry the residue
before igniting. Ignite in a previously ignited and weighed platinum crucible
until white. Cool and weigh as magnesium pyrophosphate. Calculate to mag-
nesium carbonate by using the factor 0.75719.
If only the barium content is desired, a short method may be used. After
the silica is removed, the filtrate is treated with sulfuric acid, added a drop
at a time, and stirred vigorously. The precipitate is allowed to settle and then
filtered on a Gooch crucible which has been ignited and weighed. Wash with
water containing a few drops of sulfuric acid and finally with hot water. Ignite
gently at first and finally at a red heat. Cool and weigh as barium sulfate
and compute to barium carbonate by using the factor 0.8455.
Any residue remaining after the hydrofluoric evaporation of silica is calcu-.
lated as barium sulfate and computed to barium carbonate.
Lime.-Weigh out 0.5 gram sample into a beaker. Cover it with 10 ml.
of distilled water and then 10 ml. of hydrochloric acid and heat until clear.
Filter anc;l wash several times. Transfer the filter paper to a clean platinum
crucible and ignite. Then add about 1 gram of sodium potassium carbonate
and fuse. Take up the fusion in dilute hydrochloric acid and add it to the
main solution which is ..hen transferred to an evaporating dish and evaporated
to dryness. Take up the residue with dilute hydrochloric acid and heat and
filter. Wash with hot water until free from chlorides. , Preserve the residue
and perform a second dehydration as above. Filter and wash and add this
residue to that of the first dehydration and ignite in a previously ignited and
weighed platinum crucible. Cool and weigh as silica. Correction need not
be made on the silica.
To the filtrate from the silica add a small piece of litmus paper and add
ammonium hydroxide until it is just alkaline. Heat to boiling and filter.
Wash the residue with hot water until free from chlorides. Ignite the residue
in a tared platinum crucible. Cool and weigh as (R 20 a) the combined
oxides of iron and alumina. This residue is fused in sodium potassium carbon-
ate or potassium bisulfate, dissolved and the iron reduced and titrated according
to the method given in the Chapter on Iron, Vol. 1.
The filtrate from the iron and alumina precipitation is heated to boiling
and about 20 ml. of a saturated solution of ammonium oxalate is added with
vigorous stirring. Allow the precipitate to settle and then filter and wash with
hot water. Transfer the precipitate and filter paper to the origin;:tl beaker and
dissolve it in 10 ml. of sulfuric acid and 40 ml. of distilled water .. Heat this
solution to 70 0 C. and titrate it with a standard solution of potassium per-
manganate. Compute the results to calcium oxide.
The filtrate from the calcium is concentrated by boiling. To the hot solu-
tion is added a solution of microcosmic saIt with constant stirring. It is then
made strongly alkaline with ammonium hydroxide. Allow the precipitate to
settle for several hours and then filter and wash with a 5% ammonia solution.
?174 MISCELLANEOUS ANALYSES
Dry the precipitate and paper and, when dry, separate them and ignite sepa-
rately. If there is considerable residue, it is advisablEho ignite in a porcelain
. crucible. The residues are cooled and weighed, their weights combined and
computed to magnesium oxide. I
If it is desired to determine the strontium presen't, refer to the method
outlined under the examination of Barium Carbonate. I
nation and evaporated down until the solution fumes strongly. This is cooled,
diluted carefully with about 100 to 150 ml. of water;, and heated on the steam
bath until any basic ferric sulfate which sometimes separates is redissolved.
The precipitate containing the lead sulfate is now filtered off. It is dissolved
in ammonium citrate or acetate, filtered from any in$oluble. matter, the filttate
made strongly acid with H 2S0 4 and the precipitated lead sulfate filtered off
on a Gooch crucible, ignited, and weighed. The insoluble matter from the
ammonium acetate solution should be examined for Ca and Ba.
Another method for the determination of the lead is to neutralize almost all
the acid present with sodium carbonate, saturate the solution with hydrogen'
sulfide, filter off the precipitated lead sulfide, dissolve it in fairly strong nitric
acid, and determine the lead as sulfate by the addition of sulfuric acid as above .•
In this case, solution in ammonium acetate is omitted. The former method
is of advantage where qualitative tests show that there is very little manganese
present, and it is desired to determine only the iron .. The iron is reduced to
the ferrous condition by passing the solution through a Jones reductor, and the
ferrous sulfate titrated with a standard solution of KMn04, which has been
standardized with sodium oxalate.
Iron is separated from manganese and other metals which may be present
by precipitation with ammonia, the precipitate being filtered off, dissolved in
HCI, reprecipitated with ammonia, and again filtered. It is now dissolved in
HCI, sulfuric acid added, and evaporated down until all the HCI is remo:ved;
the solution is diluted and the iron determined as before. Before the additi~n
of the ammonia, if H 2S has been used, the solution should be boiled until it is
removed, and nitric acid added to oxidize the iron to the ferric condition.
H 2S is now passed into the ammoniacal solution from the iron precipitation.
This is allowed to stand over night, and the precipitate, if there is any, is' .
examined for manganese. Should there be much manganese, the sulfide can
be filtered off, and the quantitative determination made by conversion into
the pyrophosphate.
If it is desired to determine the calcium, this can be done after the filtration
from the ammonium sulfide. (If phosphates are' present-as, for instance,
if boneblack is present-a basic acetate separation is required.) In either case
the lead should be separated by H 2S. The filtrate from the manganese sulfide
is heated on the steam bath until the H 2S is removed, ammonia and ammonium
oxalate are added, and the precipitated Ca oxalate is determined either as oxide
or s u l f a t e . . '--
The percentage of ash will be of great assistance in determining the nature
of the pigment. Black oxide of iron is only slightly changed on heating, being
completely oxidized to ferric oxide. Boneblack is composed largely of calcium
phosphate, yielding the greater part of its weight as ash. The presence of
any large amount of phosphoric acid will be sufficient evidence that boneblack
has been used. {
Prussian blue should be tested for qualitatively in the dry pigment. For
this purpose 1 g. of pigment is moistened with 2 or 3 m!. of normal alcoholic
potash, heated on the steam bath until the alcohol is removed, 5 m!. of water
added, and the insoluble matter filtered off. The filtrate' is made acid with
HCI, and filtered again if necessary. When f~tiC chloride is added, a blue
precipitate will be obtained if Prussian blue is present. If blue dye obscures
MISCELLANEOUS ANALYSES 2177
the reaction, the solution is again made alkaline and filtered, made acid with
Hel as before, and copper sulfate added. The precipitate is filtered and washed
thoroughly. In this case we will have r~dish-brown copper ferrocyanide.
It is advisable, in case of doubt, to add a small amount of Pruss ian blue to the
pigment, and make a check test. In the absence of black oxide of iron we may
assume that all of the iron in the filtrate is due to the Prussian blue. The
percentage of Fe 2 0g in the ash, multiplied by 'the percentage of ash in the pig-
ment, multiplied by the factor 1.53, will give, roughly, the amount of Prussian
blue present. The factor 1.53 is obtained by the ratio Fe7(CNhs to Fe 2 0g.
An approximate determination of the dyes can be made by extracting with
alcohol. When the presence' of oxide iron is suspected, 1 g. of pigment is
wrapped in filter paper and the dye extracted with alcohol.. When all of the
dye has been extracted, the paper and contents are dried and the nitrogen is
determined in the residue by the Kjeldahl method, From the nitrogen thus
determined the Prussian blue is calculated, using the factor 3.41. The Fe 20a
present in this amount of Prussian blue is deducted from the total Fe 2 b s found
in the ash. The remainder will be the percentage of iron from the magnetic
oxide, The formula of the latter is theoretically Fe S04, and proper calculation
should be made.
(b) Blue Inks.-A weighed quantity of pigment is ignited as under black
pigments. The ash is analyzed same as before, determining only lead, man-
ganese, and iron if the qualitative tests show that Prussian blue is present. The
lead and manganese are reported as metallic driers, the iron is calculated to
Prussian blue, and the remainder reported as mineral filler. The presence of
ultramarine will be shown by the blue color of the ash. In this case the ash
is reported after deducting the Pb+Mn. Dyes are determined by alcohol
extraction.
(c) Red Inks.-Vermillion (mercuric sulfide) is readily detected by taking
a small quantity of pigment, covering it with 4 or 5 ml. aqua regia, and heating
gently. This is diluted with five volumes of water, filtered, and stannous chlo-
ride added to the filtrate; a grayish precipitate of mercury will be formed' if
even a very small amount is present.
Quantitative method: Dissolve the mercuric sulfide in aqua regia, and after
nearly neutralizing the diluted solution to precipitate the mercuric sulfide with
H 2S, weigh the precipitate on a Gooch crucible, observing all the precautions to
eliminate sulfur which separates during the precipitation.
The following procedure has also been found of value: One g. of the pigment
is treated with a.slight excess of ammonium sulfide. Sodium hydroxide is theri
added, while stirring. The beaker is placed upon the steam bath, adding more
alkali if necessary, until all the mercuric sulfide'has passed into solution. An
excess of alkali should be avoided. The solution is allowed to cool, filtered,
and the residue washed thoroughly. To the filtrate sufficient ammonium ni-
trate to reprecipitate the mercuric sulfide is' added, and it is then boiled to
expel ammonia. The precipitate is allowed to settle, and the supernatant
liquid decanted through a weighed Gooch crucible. The residual mercuric
sulfide is boiled with a little sodium sulfide solution to remove free sulfur, and is
then trltnsferred to the crucible where it is washed with hot water, until it no
longer reacts with slIver nitrate solution. It is dried at 110° and weighed. The
2178 MISCELLANEOUS ANALYSES I
pigment is ignited, and the ash analyzed JQr lead. and manganese. The rest
of the ash is reported as mineraLfiller. I
(d) Green Inks.-The coloring matter may be chrome green, green lake, or
dye. Some of the darker shades are obtained by tl;le addition of lampblack.
The ash of the pigment is determined as usual. Part of this ash is taken and
tested qualitatively for chromium. If present, the ash should be tested for
the following substances: PbCr04, PbS0 4, PbO, BaS04, CaS04, Fe 20a, and
Mn a04. For the determination of sulfur, 0.250 g. of the ash and 5 g. of a
mixture of equal parts of potassium nitrate and sodium carbonate are fused
in a porcelain crucible over a sulfur-free flame. The cooled mass is extracted'
with hot water and filtered. The filtrate is acidified with HCI, heated to
boiling, and 10 ml. 10% barium chloride added. After standing over night, .
the precipitated barium sulfate is filtered off, ignited, and weighed. The
solution should be sufficiently acid to prevent any significant contamination
of the barium sulfate with barium chromate.
For the determination of barium, the insoluble matter is dissolved in HCI,
the solution made nearly neutral with sodium carbonate, and H 2S is passed
into the solution until all the lead is precipitated. The lead sulfide is filtered
off, the filtrate heated to boiling, and 10 ml. of 10% H 2S0 4added. The barium
sulfate is treated as directed under the determination of sulfur.
A fresh portion of ash is mixed with sodium peroxide and fused in a nickel
crucible. The cooled melt is dissolved in hot water and filtered. Carb<tU
dioxide is passed into the filtrate, and the latter heated again on the steam
bath in order to precipitate any lead which may be held up by the caustic •
alkali. Any insoluble matter which may separate is filtered off. The filtrate
- is made strongly acid with HCI, KI added, and the liberated iodine titrated
with a standard Na 2S20a solution. From the amount of thiosulfate used the
amount of CrOa present is calculated.
The two precipitates from the previous determination are combined and
used for the determination of Pb, Fe, Mn, and Ca. They are dissolved off the
filter paper with HCI, the solution is nearly neutrali~ed.with sodium carbonate,
and H 2S passed into the solution. The precipitated lead sulfide is filtered off,
dissolved in nitric acid, and determined as sulfate, as directed under black
pigments.
The filtrate from the lead sulfide is heated until all the H 2S is boiled off.
If any sulfur separates, it is filtered off. Two or three ml. of nitric acid is
added ana the solution again heated. It is then made alkaline with ammonia
and filtered. The precipitate is dissolved in HCI and reprecipitated with
ammonia, the precipitate again filtered off and the filtrate united with that
from the first precipitation. This solution is reserved for the determination
of manganese. The second precipitate of iron is dissolved in HCl, converted
into the sulfate, and the iron determined as under, black pigments.
. The united filtrates from the iron precipitatlon are saturated with H 2S,
allowed to stand over night, and then filtered. If there be a sufficient amount
of mangan~se present, it can be determined quantitatively as pyrophosphate;
otherwise qualitative identification will be sufficient.
It is safe to assume that all chromium was pre_s~nt originally as lead chro-
mate, and it should be so calculated. The irQn..oxide should be calculated to
Prussian blue, provided there is a positive qualitative test. Any barium present
MISCELLANEOUS ANALYSES 2179
should be calculated to sulfate; if' there is any question as to its being originally
present as carbonate, the ash of the pigment is treated with very dilute HCI,
the solution filtered and the filtrate tested for barium. Barytes is difficultly
soluble in cold dilute HCI. In the absence of barytes the sulfur present is
calculated to lead sulfate. The excess of lead over that required for the lead
chromate and sulfate may be considered as drier .• China clay may be present
either as an added part of the chrome green, or as the base of a green lake. In
such cases, the undetermined portion of the ash should be reported as mineral
filler. Green dyes are determined by extraction as usual. In the absence of
chrome green the pigment is ashed, and the ash analyzed for lead and man-
ganese only, the balance being reported as mineral fillers. If h~pblack has
been used to produce a dark shade of green, it can be tested qualitatively by
taking a small portion of the pigment, treating it with strong alkali, and
,filtering through a Gooch crucible, washing first with hot water, and finally
with moderately concentrated HCI. Lampblack will show a black residue,
which will disappear on ignition. It is generally classed with the volatile
con.stituents, which are then reported as aniline dye, lampblack, undissolved
oil, etc. :
The analytical me,thods included the essential constituents of inks: the
determination of iron, tannic acid, gallic acid, total solids, ash, acidity, etc ..
The specific gravity of the ink will indicate the amount of dissolved material
present. F. F. Rupert, in,a paper on the examination of writing inks (J. Ind.
Eng. Chem., 15, 489, 1923), emphasizes the importance of the physical tests
of color, permanence, stability and noncorrosiveness. The following tests
are suggested.
Streak Tests.:-A sheet of a 'good quality of bond paper placed on a pane
of glass and held by clamps serves for taking the streak. The streak is made
by a pipette 25 cm. long, 3.5 mm. in diameter with a file mark 62 mm. from
the tip. Ink filled to the mark (0.6 mi.) is tested by drawing the pipette over
the surface of the paper, at the same time allowing the ink to flow.
Penetration and Fluidity.-The ink should penetrate into the fibres of the
paper but shollld not pass through. A normal ink will give an oval head
(where the streak begins) and the width of the line will be uniform. A poor
ink gives a wide head and the streak narrows rapidly. Stickiness should be
observed.
Temporary Color.-The mark should be dark enough to be easily seen.
A speckled appearance indicates that the ink contains a dye which is not in
solution.
Permanent Color.-The streak is exposed to diffused light for 7 days in
an atmosphere free from chemical fumes and dust. A comparison is made
with a standard ink. The paper is now cut into crosswise strips 1 inch in
diameter. The top and bottom strips are set aside as blanks.
Exposure to Light.-Strips containing the streaks are exposed to sunlight
or ultra violet rays for a period 48 hours to two weeks or longer and the ink
thus exposed compared with the' blanks. ,
Exposure to Weather.-The exposed sample is inspected weekly.
Exposure to Water and Reagents.-Tests in water, in dilute ammonia
(1 : 10), HCI 2% solution, bleaching powder or sodium hypochlorite of 0.005
N available CI.
2180 MISCELLANEOUS ANALYSES
DETERMINATION OF PEROXIDES 17
measured volume of the peroxide from a burette (15-20 ml. of the solution
made by diluting 50 ml. of the commercial peroxide to 500 ml.) with gentle •
agitation of the contents of the bottle. After standing two minutes, 40 m!. of
concentrated Hel are added and the stopper inserted in the bottle. The
contents of the bottle are shaken violently, holding the stopper in place. The
stopper is now cautiously released and the confined gas allowed to escape.
Six to seven ml. of chloroform are added and the unoxidized As 20 S is titrated
with N /5 KIO s solution, shaking the closed bottle after each addition, until
the iodine color in the chloroform has disappeared.
The m!. AS 2 0 3 solution used minus the ml. equivalent of KI0 3 gives the ml.
of arsenous acid required by the hydrogen peroxide.
1 ml. N/5 AS 20 3 = 0.0034016 gram H 20 2.
This method is not influenced by the presence of organic preservatives as
is the permanganate method .which follows .
NOTE I.-The following handling of sample may be used conveniently in many cases,
particularly with moist material: Weigh a large sample, sa:y 50 or 100 grams accurately,
dissolve, and make to 500 or 1000 mI. volume. The "water insoluble" should be run
on separately weighed samples as indicated, the filtrate being discarded. The water
insoluble material should be allowed to settle, on the l~fge made up sample, and an·
aliquot giving 10-12 g. made to 500 m!. for the sulfides, sUlfite, etc., proceeding as out-
lined. Suitable aliquots taken from the strong made-up soluti~n can then be taken for
the Na 2CO a and NaCI determinations. .
NOTE 2.-Methyl red indicator used here is made by heating a small amount of the
dye with water almost to a boil. From one to five mI. of the cool settled solution may
be used. ,
NOTE 3.-Starch solution used here is made by mixing 5 g. soluble starch with 10 mI.
of water, adding 20-30 mI. N/1 NaOH to clear,.tit"rif,ing to phenolphthalein neutral
point with N /1 HCI, and diluting to 500 mI.
MISCELLANEO'US ANALYSES 2183
In many cases too much time will be lost by givirlg the overnight standing as
specified .• For control purposes, the time may be shortened to as smlill a
period as ! hour, introducing a plus error usually undei' 10% of the Na 2C0 3
present. A somewhat longer standing period, sll,y 4 hours, gives the BaS
which forms to some extent, a better chance to decompose.
Calculations:
Thiosulfate-
. (C - D) X .0158 X 100 _ at N (~O' ,
Grams sample - /0 a2"2 3.
Sulfite-
DX.0063X100 _ 01 N S(
- /0 a2 )3.
G ramssamp Ie ,
Sulfide and Acid Sulfide- . -' -
Correct E for carbonate and sulfite by subtracting the sum of the titrations
adjusted for sample size. '
S
g. Sample for B
MISCELLANEOUS ANALYSES·
- .
1 f N CO X T = (N a 2C03+ N a 2S03) mls. on B baSIs-Call T'.
g. amp e or a2 3 .
2185
The following equations and example will clarify the above calculations:
(1) Na 2S+1 2 ~2Nal+S.
B .0039
- N 110 12 = 80.0 ml.
.0028 ..0028
Na 2S + NaHS
E' is less than B - Sample = .5000 g . ..
2186 MISCELLANEOUS ANALYSES
Sulfide as % H 2S-
57.50X.0017X100 =14 0501 H~
.696 . 10. •
Factor N aHS 'to H 2S 34.08/56.07 = .607.
Above NaHS (21.7%) equiv. to 13.16% H 2S.
14.0.5-13.16=.89% net H 2S.
Thiosulfate-
Carbonate-
33.05 X.0053 X100 =21501 N CO
8.174 . 10 a2 3.
Summary--
% H 2S ..........................89
% HaHS ....................... 21.7
% Na 2C0 3 • • • • • • • • • • • • • • • • • • • • • • 2.15
% Na 2S203...................... .86
•
NOTE 5.-Credit for above methods must be given to Sutton's Volumetric Analysis,
DuPont Publication, in J. Ind. Eng. Chem., 17, 176 (1925). Grasselli Chemical Co.,
12-18-29, and various Dow Chemical Company chemists. -
We are indebted to A. W. Beshgetoor, Chief Analytical Chemist, the Dow Chemical
Company, for the section on the complete analysis of sodium sulfide.
PART II
SPECIAL
TECHNIQUES APPLICABLE
TO
DIVERSE SUB STAN CBS
ACIDIMETRY AND ALKALIMETRY 1
The volumetric determination of a free acid or a free base may be accom-
plished with rapidity and accuracy by neutralization with a known quantity of
,tandard base or alkali as the case may require. The point of neutralization or
., end-point" is accertained by means of certain compounds called indicators,
which have a different color in acid solutions than in alkaline solutions, the'
point of transition from one color to the other occurs at the point of neutraJiza-
tion. This end-point may also be recognized by the electro metric method by
measuring the change of potential that occurs with the change of concentration
of the hydrogen ions in the solution. This potential change is usually large
and abrupt at the end-point.
Indicators.-The accuracy of acidimetric and alkalimetric titrations de-
pends, to a considerable extent, on the choice of the indicator used. Indicators
are usually dyestuffs, which have one color in an acid solution and another in
alkaline solutions. The color change should occur as near as possible to the
equilibrium point reached in neutralization. The color change of certain
indicators may be due to, a difference in color of the dissociated from the un-
dissociated compound or to the presence of certain grou'ps of radicals within
the molecule of the compound, called chromophores, which show one color ill
acid solutions and, by internal rearrangement, another color in basic solutions.
This color change depends upon the hydrogen ion concentration of the solution
rather than the absolute amount of acid present. The pH range of color
change differs with different indicators as is shown in the table given on th,e
following page. It is generally the case that indicators whose color change oc-
curs with a hydrogen ion concentration greater than pH 7 are sensitive to
weak bases, while those whose color change occurs with hydrogen ion concen-
tration less than pH 7 are sensatlve to weak acids. Thus methyl red wit}!
color changes between pH 4.2 and 6.3, is sensitive to weak bases, while phen-
olphthalein with a range of pH 8.2 to 10 is sensitive to weak acids.
Indicators may show a color change at entirely different stages of the
chemical reaction, phenolphthalein, for example, in the titration of sodium
carbonate, Na 2 CO a indicates an acid reaction as soon as one of the Na atoms is
neutralized and a drop of excess of the titrating acid liberated CO 2• Methyl
orange at this stage shows alkalinity, NaHCO a being alkaline to methyl orange
and neutral to phenolphthalein. The same difference is shown in the titration
of other polybasic compounds.
Among the indicators commonly employed in acidimetry and alkalimetry
the list given by Kolthoff and Furman 2 is of special interest; tropeolin 00,
pH 1.3 red-3.2 yellow; thymol blue, pH 1.2 red-2.8 yellow; methyl yellow,
transition 2.9-4.0, red to yellow; methyl orange, transition interval 3.0-4.4,
red to orange; brom phenol blue, pH 3.0-4.6, yellow to violet blue; congo red,
3.0-5.2, blue to red; methyl red, 4.4-6.2, red to yellow; brom cresol green, 4.0- .
1 By Wilfred W. Scott.
2 "Vol. Anal.," J. Wiley and Sons.
2191
2192 ACIDIMETRY AND ALKALIMETRY
5.6, yellow to blue; chlorphenol red, 5.0-6.6, yellow to red; brom cresol purple,
5.2-6.S, yellow to purple; sodium alizarin sulfonate, 5.5-6.S, yellow to lilac;
brom thymol blue, 6.0-7.6, yellow to. blue; azolitmin (litmus), 5.D-S.0, red to
blue; pheno.l red, 6.S-S.0, yellow to red; neutral r~d, 6.8-S.0, red to. ye~o.w
orange; rosalic acid, 6.9-S.0, yellow to. red; 'creso.l r~d, 7.2-S.8, yello.w to red;
a naphtho.l phthalein, 7.3-8.7, yellowish rose to green; curcumin, 7.4-S.6,
yellow 'to. red-bro.wn; phenol phthalein, S.2-1O, colorless to red; thymol blue,
S.D-9.6, yellow to. blue; a naphthol benzein, 9.D-11.0, co.lo.rless to blue; thymol
phthalein, 9.3-10.5, colorless to blue; alizarine yellow, '10.1-12.1, yellow to
lilac; tropeolin 0, 11.0-13.0, yellow to o.range brown! nitramine, 10.S-13.0,
colorless to red bro.wn. Certain mixed indicators give a color change more
sharply, for example a mixture of bro.m cresol purple and brom thymo.l blue,
mixed in equal parts. The indicato.r is greenish yellow at pH 6.0 and blue
at pH 6.S. A co.mprehensive table of mixed indicators is given in Koltho.ff and
Furman's work.
A list of the more common o.f the indicators is tabulated belo.w. The pH
transition interval, and the color changes are given.
Solutions of Indicators: Brom Cresol Purp1e.-0.04% alcoholic solution.
Changes color at about pH 6.
Methyl Orange.-0.1 g. per 100 ml. of water. Transition interval pH '3.-
4.4. Apparent pH, of end-point, or pT = 4.
Methyl Red.-The acid is dissolved)n alcohol (0.1 g. per 60 m!.) and illfide
up to 100 m!. with water. If the sodium salt is available, 0.1 g. is dissolved in
• 100,ml. of water. pT = 5.
·Methy1 Yellow.-0.1 g. in 100 m!. of 90% alcohol. pT=4.
Phenolphthalein.-1 g. per 100 mr. of 95% alcohol. If I or 2 drops of this
solution is used per 100 m!. of solution titrated, the end-point is at pH 9. If
about ten times as much of the indicator is added per 100 J!ll. of solution .the
change appears at pH S. Cresol red (0.1 g. per 100 ml. 20% alcohol) changes
color at pH S. .
Thymol Blue (Thymol sulfone phth~lein).-O.l ,~. in 20 ml. of hot alcehol;
dilute to 100 ml. with alcohol. The indicator has two ranges: Changes from
red to yellow at pH 1.2-2.S, pT=2.6; changes from yellow to blue, pH S.-9.6,
pT=9.
Mixed Indicators.-1.25 g. methyl red, 0.S25 g. methylene blue dissolved in
1 liter of 90% alcohol; 0.75 g. methyl red, 0.025 g. guinea green dissolved in
1 liter of 90% alcohol; 5 g. so~ium alizarin sulfonate, 0.625 -g. 'guinea green in
1 liter of water; 5' g. sodium alizarin sulfonate, 1.25 g. indigo car min in 1 liter
of water (A. H. Johnson and Jessie R. Green, Ind. Eng. Chern., Anal. Ed., 2,
2 (1930». '
ACIDIMETRY AND ALKALIMETRY 2193
ULTIMATE STANDARDS
Normal Sulfuric acid contains 49.043 grams of H 2S0 4 per liter of solution.
To make Ii liter of the normal acid the amount of the standard acid required is
100X49.043 .
calculated by the formula t H SO . d d grams standard aCId
per cen 2 4 In stan ar .
necessary. The acid is weighed out in a small beaker, a slight excess being
taken (0.1 gram). The acid is washed into a liter flask and made to volume.
An' aliquot portion is standardized against the standard sodium carbonate.
The solution may now be adjusted to the exact strength",required.
Example.-If 25 m!. of the acid is found to contain 1.25 grams H 2S0 4 we
find the amount of dilution required as follows: 25 m!. of N /1 H 2S0 4 should
contain 1.226075 grams, therefore 1.226075 : 25 :: 1.25 : x, and x = ~~2~:~~~'
Then x minus 25 = the amount of water required for 25 ml. Total dilution
= dilution for 25 multiplied by the volume of. acid remaining in. the flask
divided by 25 =ml. water required to make a normal acid solution.
Fifth normal and tenth normal acids may be prepared by diluting the
normal acid to five or ten volumes as the case requires.
Gravimetric Methods. Precipitation as BaS04.-Sulfuric acid' may be
standardized by precipitating as BaS04 according to the procedure given for
sulfur. BaS04 X0.4202 = H 2S0 4,
Determination as (NH4}2S04.-To 10 ml. of the acid diluted to 50 ml. in a
large platinum dish is added NH 40H until the acid is neutralized and a faint
odor of ammonia is perceptible. The solution is evaporated to dryness on the
water bath and dried a,t 100° C. for half an hour. The residue is weighed as
(NH 4)2S04' (NH 4)2S04XO.7422 = gram H 2S0 4.
acid until it has attained a definite boiling point. The pressure being known
the exact amount of the acid that now distills and that remaining in the flask.
provided the original solution was pure, can be estimated from the following
table of Bonner and Wallace. 4 \
, I
Boiling Point Data of Her I
NOTE.-In a separate flask, with the same amount of solution and in'dicator, obtain
the same color by adding NaOH drop 'by drop. Subtract this blank from the above
titration before calculating normality.
Alkaline solutions that are to be used in presence of CO 2are best standardized against
standard acid solutions prepared by titration against Na 2CO S or against HCI prepared
from the acid of constant boiling point, approximately 20.2% HCl.s
S G. A. Hulett and W. D. Bonner, J. Am. Chern. Soc., 31, 390 (1909), and W. D.
Bonner and R. E. Wallace, J. Am. Chem. Soc., 52, 1750 (1930).
2198 ACIDIMETRY AND ALKALIMETRY
Precautions should be taken against the presence ofr CO 2 in the titrations where
phenolphthalein is used. The distilled water should be freshly boiled, and the air in
the flask should be free of CO 2• ' .
Pure NaOH may be prepared by dissolving 100 grams of the pure stick NaOH in
about 100 ml. of water, transferring to a large test tube ,a;nd allowing to settle. If the
solution is centrifuged for 20-30 minutes the sodium carbonate and other insoluble
material will settle and the clear liquid may be used at once. 7 By'means of a graduated
pipette the required amount of the clear solution is drawn out and made to volume.
40 grams NaOH per liter is a normal solution. Other normalities may be desired.
The NaOH should be kept in bottles lined with paraffine. This paraffining is done by
heating the bottle, pouring in sufficient melted paraffine and rotating the bottle so that
the paraffine flows over the sides and the entire inner surface.
The end-point with methyl orange indicator (3 drops 0.02% solution per 100 ro.l.)
is the color obtained in a solution with pH=4.2.
The corrections apply to pure water. With the reagents indicated below
the correction must be increased by the percent~ges indilJated.
7 Allen and Low, Ind. Eng. Chern., Anal. Ed.,-S;-'1~2 (1933).
8 "Chemisch-Technische Untersuchungsmethodik," 7th Ed., Vol. I, p. 138.
ACIDIMETRY AND ALKALIMETRY 2199
N HNO a 50%, HCl 25%, H 2SO. 45%, NaOH 40%, KOH 40%, H 2C.O. 30%,
Na 2CO a 40%. .
N/IO HNO. 6%, HCI3%, H 2SO. 5%, NaOH 5%, KOH 4%, H 2C.O. 3%,Na.CO a
5%.
STANDARD BURETTES .
Burettes used for this work should be carefully ,checked for accuracy of
delivery. G
For accurate titration of acids
or alkalies it is advisable to have a
titration of 75 to 100 ml. Since the
straight 100-ml. burette if gradu-
ated to twentieths of a ml. would
be too long for convenient handling,
the chamber burette is used. The
chamber located in the upper por-
tion of the apparatus holds 75 mI.,
the lower portion drawn out into
a uniform-bore tube is graduated
in twentieths of a mI. Each tenth
of a ml. has a mark passing entirely
around the tube so that there will
be no error in reading, the eye being
held so that the mark appears to be c
a straight line drawn across the
tube. The burette is enclosed in a
large tube filled with distilled water
Chamber
and carrying a thermometer. The Burette
burette is connected, by means of A
an arm at the base, with a reservoir
of standard acid. The cut, Fig.
300, shows the apparatus con-
nected ready for use. '
If vapor is lost from the stand-
4.' -Wafer
ard reagents and this replaced by Jacket
dry air, as in the common practice, B
the solution gradually changes in
strength. A simple and ingenious
device, designed by H. W. Herig
(Gen. Chern. Co.), is shown at the
top of Fig. 300, whi<;h overcomes
this difficulty. The air drawn into F
the reagent bottle is purified and
saturated with moisture by pass-
ing it through sodium hydroxide.
FIG. 300.
A mercury valve relieves the pres-
sure if expansion of air in the,reagent bottle occurs due to rise of temperature.
2200 ACIDIMETRY AND ALKALIMETRY
NOTE.-The chamber burette shown in Fig. 300 was designed at the Laurel Hill
Laboratory, General Chemi9al Company. I "
. I
TITRATION OF ACIDS AND 4LKALIES
In the acid titration the sample is conveniently titrated in a white porcelain
a
casserole. This gives white background that enables the analyst to see the
end-point. The caustic is run into the acid, to within a few ml. of the end-
point, rapidly and then cautiously to a faint change of color-faint pink with
phenolphthalein or an orange-yellow with methyl-orange. Phenolphthalein
is generally preferred in acid titrations. CO 2-free caustic and '~/ater should
always· be used. J
tube connected with the reservoir of water., The acid in the casserole, upon
washing out the Dely tube, is titrated with standard caustic according to the
procedure for titration of acids.
The tube is dried after washing, with alcohol, followed hy
ether, by heating on an asbestos m'at on a hot plate, dry air
being aspirated through. I
Snake Weighing Tube.-The snake tube is a simple
device that may be easily made by an amateur glass-blower.
It is made out of a glass tube 8-10 ins. long, slightly thinner
than a lead pencil. One end of the tube is drawn out to
capillarity. The tube has a double bend, as shown in tpe
illustration. It is so made that it r~sts on the double bend
with the ends inclined upward to prevent the outflow of the
acid. Fig. 303.
The tube is dried with alcohol, ether and air treatment, as
in case of the Dely tube. After weighing the empty tube, acid
is drawn into it by suction ,through an attached rubber tube.
FIG. 303. The capillary end that has dipped into the sample is wiped
Snake Tube.
dry with tissue paper. The acid and tube are weighed and
the acid estimated by difference.
The acid is run into 150 m}. of water in a casserole, the flow being regulated
by the index finger pressed against the "
larger end of the tube. With careful Horns fOrSuspendip9
regulation of the flow, practically no bur.:efte
bumping occurs. With a small capillary
opening it is not necessary to place the 0 ~ y:/
finger over the larger end of the tube as
the acid flow will be slow. The tube
should be kept in motion to prevent
bumping from overheating anyone por-
tion. Kicking back of the acid indicates
that the capillary end of the tube is too
large. When the contents of the tube a,'
have run out, the tube is rinsed by suck-
ing up some acid from the casserole and
allowing it to run out, repeating sev-
eral times. Suction mlay be applied by - E Capillary
liibe adaplor
means of a rubber bulb attached to the for Fuming
tube. The acid is now titrated with Acids
standard caustic, using phenolphthalein
indicator.
Blay-Burkhard Graduated Weighing
Burette. lO-This apparatus, designed by
V. J~. Blay and W. E. Burkhard, Gen-
eral Chemical Company, is used for
weighing acids or other liquids. The
form for u;eneral use i~ shown in Fig. 304. FIG. !?04. Blay-BurkPard Graduated
THe burette is graduated in half m!. divi- . -' ~ Weighing Burette.
10 A similar burette of larger capacity and small weight was designed by Friedman
and La Mer, Ind. Eng. Chern., Anal. Ed. 2, 54 (1930).
ACIDS 2203
sions, from 0 to 20 ml. An apparatus half this size is used for oleum, where
a 2-ml. sample is sufficient for a determination. For the purpose of running
the sample under water a capillary tub~ (E, Fig. 304), with ground joint, is
attached to the burette. This tube is placed in the solution during titration.
The burette is provided with a glass vented stopper. (A) on the top, and a
glass cap for the tip, both having ground joints, to prevent escape of fumes
from the sample.
W. W. Scott modified the apparatus by replacing the fragile cap (A) by a
tube stopper with capillary vent (see A', Fig. 304). The vent to the air is
opened or closed by a slight turn of this stopper. By means of this tube acid
may be drawn into the burette according to the Lunge-Ray pipette procedure.
With these burettes a man can control his work very accurately and save a
great amount of time, both in weighing and manipulation.
In the analysis of strong oleum, about 50 grams of neutral Glauber salt are
placed in a casserole containing water, and the fuming acid allowed to flow
under the undissolved 'salt. The violent reaction of the acid with water is
thus ~voided. The tube E, Fig. 304, should be made of fused silica.
The glass-bulb method is still used for analysis of strong oleum. The acid
weighed in a sealed tube of known weight is mixed with water by breaking the
bulb in a stoppered bottle containing water, the acid is cooled and titrated as
usual. I
FIG. 305.-Hydrometer.
The Gravity tables given in the following pages have been developed with
considerable care. By means of ,these one is able to obtain quickly and with
fair degree of accuracy the strength of the acid or alkali of which the gravity is
ascertained.
The following precautions should be observed in making hydrometer tests:
1. The hydrometer should be clean and dry, and at the temperature of the
liquid, before immersing to make a reading. '
2. The vessel in which the observation is made should be of clear glass of
suitable size and shape, to allow the hydrometer to float freely (about ~ inch
greater in diameter tllan the hydrometer bulb) and of sufficient lieight to enable,
full reading of the hydrometer (i.e., height greater than the length of the
hydrometer). '
3. The liquid should be thoroughly mixed by means of a stirrer reaching
to the bottom of the vessel. There should be no air bubbles in the liquid or
clinging to the sides of the vessel or the hydrometer.
4. The hydrometer is slowJy immersed in the liquid, slightly beyond the
point where it floats, and is then allowed to float freely.
5. The reading is made with the line of vision horizontal to the plane and
as near as possible to this. The point is taken 'where this surface line cuts the
hydrometer scale.
6. The temperature of the liquid is taken before and after the reading and
allowance made for variation o{ the temperature from standard conditions as
indicated in the tables.
References.----:Cir. 16, 4th edition, Feb. 23, 1916, U. S. Bureau of Standards.
Sulfuric Acid Handbook by Tllos. J. Sullivan, McGrftw-Hill Book Co.
Thorpe Dictionary of Applied Chemistry, pp. 103-114. Longmans, Green and Co.
2206 ACIDS
HYDROCHLORIC ACID
By W. C. FERGUSON j
Degrees
Baumel. Sp. Gr. Degrees
Twaddell.
Per Cent
HCl. I
Degrees
Bau~.
I
Sp. Gr.
1
Degrees Per Cent
Twaddell. HCl.
1.00 1.0069 1.:>8 1.40 14.25 1.1090 21.80 21.68
2.00 1.0140 2.80 2.82 14.50 1.11'll 22.22 22.09
3.00 1.0211 4.22 4.25 14.75 1.1132 22.64 22.50
4.00 1.0284 5.68 5.69 15.00 1.1154 23.08 22.92.
5.00 1.0357 7.14 7.15 15.25 1.1176 23.52 23.33
5.25 1.0375 7.50 7.52 15.50 1.1197 23.94 23.75
5.50 1.0394 7.88 7.89 15.75 1.1219 24.38 24.16
5.75 1.0413 8.26 8.26 16.0 1.1240 24.80 24.57
6.00 1.0432 8.64 8.64 16.1 1.1248 24.96 24.73
6.25 1.0450 9.00 9.02 16.2 1.1256 25.12 24.90
6.50 1.0469 9.38 9.40 16.3 1.1265 25.30 25.06
6.75 1.0488 9.76 9.78 16.4 1.1274 25.48 25.23
7.00 1.0507 10.14 10.17 16.5 1.1283 25.66 25.39
7.25 1.0526 10.52 10.55 16.6 1.1292 25.84 25.56 I
7.50 1.0545 10.90 10.94 16.17 1.1301 26.02 25.72
7.75 1.0564 11.28 11.32 16.8 1.1310 26.20 25.89
8.00 1.0584 11.68 11.71 IG.9 I.l319 26.38 26.05.
8.25 1.0603 12.06 12.09 17.0 1.1328 26.56 26.22
8.50 1.0623 12.46 12.48 17.1 1.1336 26.72 26.39
8.75 1.0642 12.84 12.87 17.2 1.1345 26.90 26.56
9.00 1.0662 13.24 13.26 17.3 1.1354 27.08 26.73
9.25 1.0681 13.62 13.65 17.4 1.1363 27.26 26.90
9.50 1.0701 14.02 14.04 17.5 1.1372 27.44 27.07
9.75 1.0721 14.42 14.43 17.G 1.1381 27.62 27.24
10.00
10.25
1.0741
1.0761
14.82
15.22
14.83
15.22
17.7
17.8
1.1390
1.1399
27.80
27.98
27.41
27.58
•.
10.50 1.0781 15.62 15.62 17.9 1.1408 28.16 27.75
10.75 1.0801 16.02 16.01 18.0 1.1417 28.34 27.92
11.00 1.0821 16.42 16.41 18.1 1.1~26 28.52 28.09
11.25 1.0841 16.82 16.81 18.2 1.1435 28.70 28.26
11.50 1.0861 17.22 17.21 18.3 1.1444 28.88 28.44
11. 75 1.0881 17.62 17.61 18.4 1.1453 29.06 28.61
12.00 1.0902 18.04 18.01 18.5 1.1462 29.24 28.78
12.25 1.0922 18.44 18.41 18.6, 1.1471 29.42 28.95
12.50 1.0943 18.86 18.82 18.7 1.1480 . 29.60 29.13
12.75 1.0964 19.28 19.22 18.8 1.1489 ·29.7S 29 30
13.00 1.0985 19.70 19.63 18.9 1.1498 29.96 29.48
13.25 1.1006 20.12 20.04 19.0 1.1508 30.16 29.65
13.50 1.1027 20.54 20.45 19.1 1.1517 30.34 29.83
13.75 1.1048 20.96 20.86 19.2· 1.1526 30.52 30.00
14.00 1.1069 21.38 21.27 19.3 1.1535 30.70 30.18
I
ACIDS 2207
HYDROCHLORIC ~ID (Continued)
Degrees
Baume.
---
Sp. Gr. Degrees
Twaddell.
Per Cent
HCI.
Degrees
Baumti. S p. Gr. I Degrees
TWaddell.
Per Cent
HCI.
19.4 1.1544 30.88 30.35 22.5 1.!836 36.72 36.16
19.5 1.1554 31.0S 30.53 22.6 1.1846 36.92 36.35
19.6 1.1563 31.26 30.71 22.7 1.1856 37.12 36.54
19.7 1.1572 31.44 30.90 22.8 1.1866 37.32 36.73
19.8 1.1581 31.62 31.08 22.9 1.1875 37.50 36.93
19.9 1.1590 31.80 31.27 23.0 1.1885 37.70 37.14
20.0 1.1600 32.00 31.45 23.1 1.1895 37.90 37.36
20.1 1.1609 32.18 31.64 23.2 1.1904 38.08 37.58
20.2 1.1619 32.38 31.82 23.3 1.1914 38.28 37.80
20.3 1.1628 32.56 32.01 23.4 1.1924 38.48 38.03
20.4 1.1637 32.74 32.19 23.5 1.1934 38.68 38.26
20.5 1.1647 32.94 32.38 23.6 1.1944 38.88 38.49
20.6 1.1656 33.12 32.56 23.7 1.1953 39.06 38.72
20.7 1.1666 33.32 32.75 23.8 1.1963 39.26 38.95
20.8 1.1675 33.50 32.93 23.9 1.1973 39.46 39.18
20.9 1.1684 33.68 33.12 24.0 1.1983 39.66 39.41
2l.0 l.1694 33.88 33.31 24.1 1.1993 39.86 39.64
21.1 1.1703 34.06 33.50 24.2 1.2003 40.06 39.86
21.2 1.1713 34.26 33.69 24.3 1. 2013 40.26 40.09
2l.3 1.1722 34.44 33.88 24.4 1.2023 40.46 40.32
21.4 1.1732 34.64 34.07 24.5 1.2033 40.66 40.55
21.5 1.1741 34.82 34.26 24.6 1.2043 40.86 40.78
21.6 1.1751 35.02 34.45 24.7 1.2053 41.06 41.01
21.7 1.1760 35.20 34.64 24.8 1.2063 41.26 41.24
21.8 1.1770 35.40 34.83 24.9 1.2073 41.46 41.48
21.9 1.1779 35.58 ·35.02 25.0 1.2083 41.66 41.72
22.0 1.1789 35.78 35.21 25.1 1 2093 41.86 41.99
22.1 1.1798 35.96 35.40 25.2 12103 42.06 42.30
22.2 1.1808 36.16 35.59 25.3 1.2114 42.28 42.64
22.3 1.1817 36.34 35.78 25.4 1.2124 42.48 43.01
22.4 1.1827 36.54 35.97 25.5 1.2134 42.68 43.40
Sp. Gr. determinations were made at 60° F., compared with water at 60° F.
From the Specific Gravities, the corresponding degrees Baume were calcu-
lated by the following formula: Baume = 145 -145/Sp. Gr.
Atomic weights from F. W. Clarke's table of 1901. 0 = 16.
ALLOWANCE FOR TEMPERATURE:
10-15° Be. - 1/40° Be. or .0002 Sp. Gr. for 1° F.
15-2ZO Be~ - 1/30° Be. or .0003" " " 1° F.
22-25° Be. - 1/28° Be. or .00035 " " " 1° F.
AUTHORITY - W. C. FERGUSON.
This table has been approved and adopted as a Standard by the Manu_
facturing Chemists' Association of the United States.
W. H. BOWER, JAS. L. MORGAN,
HENRY HOWARD, ARTHUR WYMAN,
A. G.
ROSENGARTEN,
New York, May 14, 1903. Executive Committu.
2208 ACIDS
HYDROCHLORIC ACID
Pressure mm. of
Mercury.
I rer cent of
HCl.
Grams' constant
bOiling distillate
for I mol. HCI.
unless the solution is kept cold-O° C.) The reading of the burette is noted-
total ":/1 fNaOHI A.· The dish is now placed on ~ hot plate ~nd the solution
t.o samp e
0
warmed to about 80 C. and the titration completed with 'the N/l NaOa
solution to a permanent pink. Additional rill. required divided by weigh,t of
sample=B.
,
DETERMINATION OF SULFURIC ACID IN NITRIC ACID
About 10 grams of the acid are evaporated to dryness on the steam ba}h.
The residue is taken up with about 10 ml. of water and the evaporation repeated
until free from nitric fumes, the residue finally diluted 'to 100 ml. and the sulfuric
acid titrated with N 15 NaOH, using phenolphthalein or methyl-orange indi-
cator. Gravimetrically the acid may be precipitated from a hat solution as
BaS04 by addition of barium chloride reagent l1ccording to the method for
determining sulfur. -
One ml. N/5 NI!OH=0.009809 gram H 2S0 4.
BaS04 X0.4202= H 2S0 4. Per cent=100 divided by weight of sample
X H 2S04 obtained. "
NITRIC ACID
From the total acidity expressed as H 2S0 4 is subtracted .the acidity due to
HCI and HN0 2 (lower oxides of nitrogen) expressed in terms of H 2S0 4 • The
remainder is due to nitric afid, in terms of sulfuric acid.
H 2S0 4X 1.285 = HN0 3.
Calculation.
Since N /2 K 2 Cr 207 =0.024517 gram salt, 1 gram K~Cr207= 1.13882 Fe, there-
fore 25 mI.=0.024517X1.13882XO.5643X25=0.3939 gram HNO s equivalent.
0.3939 divided by ml. FeS04 required in the titra~ion=grams.HNOs equiv-
alent per mi. .
B. Reagent to be Used in Determination of HN0 s in H'JAs04 or H S P04. -
Thirty-eight milliliters of K 2 Cr 2 07 solution are titrated with FeS04 according to
directions given in "A." The Fe value multiplied by 0.3762 = gram H:ijOa.
. Calculation.-Thirty-eight milliliters of K 2 Cr 207=0.024517X1.13882X
0:3762 X38 = 0.3991 gram HN0 3 equivalent. 0.3991 divided by ml. FeSj)4
required in the titration = grams H'N0 3 equivalent per ml. I
Factors.
K 2 Cr 2 07 to Fe = 1.13882, reciprocal = 0.8781.
..
2Fe to. HNO s =0.5643, recip. = 1.7722. 3Fe to HN0 3 =0.3762, recip.
=2.6582.
HNO s to 2FeS04·7H 20=8.8235 recip. =0.1133.
HNOs'to 3FeS04·7H20 = 13.2348, recip. =0.07556.
K 2Cr 207 to HNO s =0.6426, recip. = 1.5562. (Titration of A reagent.)
K 2Cr 207 to HNO s = 0.4284, recip. = 2.3342. (Titration of B reagent.) .
of the solution diluted to 500' ml. will give a mixture of the desired strength.
23 m]. are accordingly weighed in a weighing bottle, the acid washed into a
beaker transferred then to the graduated 500-m!' flask and made to volume.
The preliminary run may be made in two or three minutes,
Titration.-A 250-mI. beaker containing 100 mI. of concentrated, nitric free,
H 2S0 4 (93 + %) is placed in a large casserole or deep porcelain dish containing
cold water. 10 m!. of the sample are measured out in an accurately marked
pipette, graduated to contain exactly 10 m!. The solution is' run under the
surface of the sulfuric acid, the delivery tip of the pipette being kept in constant
circular motion to prevent too much local heating. Since the sides of the
beaker are cooled, the tip of the pipette should be kept against the sides jn
the circular sweep during the delivery. By this procedure loss of nitric acid is
reduced to the minimum. '
The ferrous sulfate solution is now added from a burette in a fine stream
until the yellow color that first forms takes on a faint brownish tinge (dirty
yellow). The pipette is now rinsed out by sucking up the mixture and draining
it back into the beaker. The titration is now completed, adding the FeS04
cautiously drop by drop until the yellowish brown color again appears, a drop
in excess producing an appreciable darkening of the solution. A larger excess
produces a brownish red color. With small amounts of HNO a a pink color
will be obtained, instead of the yellowish brown. The end-point 'once recog-
nized is readily duplicated.
Calculation.-The mI. titration minus the blank 0.2 ml. multiplied by the
factor for FeS04=weight of HNO a. HNOax'IOO+wt.=per cent.
Example.-Suppose 10 m!. equivalent to 1/50 of a 42-g. sample weighed,
requires 22 ml. FeS04 whose value = 0.02 g. HNO a per mI., then (0.44 X'lOO) /0.84
=52.4% HNO a.
ing to directions under General Procedure for Nitric Acid. A blank of 0.2 ml. 13
having been deducted, ml. FeS04XHNO a factor for FeS04XlOO divided by
wt. taken = per cent HNO a.
Correction Factor.-In making a number of ruAs with varying amounts of
HNO a, it was found that small quantities of nitric ~cid required a proportion-
ally greater amount of FeS04 than larger quantitieS of HNO a. For example,
0.07392 gram HNO. required 3.9 ml. FeS04, four times the amount of HNO.
required 15 ml. FeS04, in place of 15.6 (3.9X4), and six times 0.07392 gram
HNO a required 22.5 ml. FeS04 in place of 23.4. It was observed that'even
traces of HNO a required a titration of over 0.2 ml. It is·evident that a deduc-
tion of 0.2 ml. makes the titrations multiples of the lowest, e.g., 3.7, 14.1\ and
22.3. Again it· was found that standardization of FeS04 with HNO a checked
the dichromate factor when 0.2 ml. was deducted from the first series of titra-
tions. This led to the conclusion that a constant blank of 0.2 ml. should be
deducted from the ferrous sulfate titrations of nitric acid in presence of 100 ml.
of nitric free sulfuric acid (66 0 Be.).
Comparison of results:
FeS04 value by RN0 3 corrected = 0.02067 gram. Uncorrected = 0.02045' gram
RNO •.
FeSO. value by K 2Cr 207 titration =0.02083 gram RN0 3•
Accuracy of the Ferrous Sulfate Method.-Results obtained by the ferrous
sulfate method agree closely with those obtained by the nitrometer. ['he
following data were obtained by Mr. B. S. Clark, by the FeS04 method, on
nitrated oleums. The figures below the first row are checks obtained on these
samples by purchasers of the acid, the nitrometer methC!d being used.
FeSO. method. 2.40; 2.82; 3.23; 3.35; 3.52; 3.50; 3.48; 3.57; 3.53; 3.56.
Nitrometer method. 2.35; 2.79; 3.26; 3.39; 3.57; 3.53; 3.50; 3.58; 3.5'7; 3.56.
Difference. 0.05; 0.03; 0.03; 0.04; Op5; 0.03; 0.02; 0.01; 0.04; 0.00.
Degrees
Baume.
I Sp. Gr.
60°
6o oF .
Degrees Per Cent
Twaddell. HNO ••
Degrees
Baume.
. Sp, Gr.
~·F.
609
I
Degrees l'er.Cent
Twaddell. HNO••
10 00 1 0741 14.82 12.86 21.25 1.lp8 34 36 28.02
10.25 1 07<31 15.22 13.18 21.50 1,1741 34 82 28.36
10.50 1.0781 15.62 13.49 21.75 1.1765 35.30 28.72
10.75 1.0801 16.02 13.81 22.00 1.1789 35.78 29.07
11.00 1.0821 16.42 14.13 22.25 1.1813 36 26 29.43
11.25 1.0841 16.82 14.44 22.50 1.1837 36.74 29 78
11.50 1.0861 17.22 14.76 22.75 1.1861 37.22 30.14
11.75 1.0881 17.62 1507 23.00 1.1885 37.70 30.49
12.00 1.0902 18.04 15 41 23.25 1.1910 38.20 30 86
12.25 1.0922 18.44 15.72 23.50 1.1934 38.68 31 21
12.50 1.0943 18 86 16.05 23.75 1.1959 39.18 31.58
12.75 1.0964 19 28 16.39 24.00 1.1983 39.66 31.94
1300 1.0985 19 70 16.72 24.25 1.2008 40.16 32.31
13.25 1.1006 ZO 12 17.05 24.f50 1.2033 40 66 32 68
13.50 1 1027 20 54 17 38 21,73 1 2058 41.16 3305
13.75 1.1048 ZO 96 17.71 23 CO 1.2083 41.66 33.42
14.00 1.1069 21.33 1304 25.25 1 2109 42.18 33 80
14.25 I 1090 21.80 18 37 25 . .';0 1 2134 42 68 34 17
14 50 1.1111 22.22 18.70 25.75 1 2160 4320 34.56
14.75 1 1132 22 64 19.02 23.CO 1 2185 43.70 34.94
15.00 1 1154 23.08 19 36 26.25 1.2211 44.22 35.33
15.25 1.1176 23.52 19.70 26.[0 1.2236 44.72 35.70
15.50 1.1197 23.94 20.02 26 75 1.2262 45.24 3609
15.75 1.1219 24.38 20.36 27.CO 1 2288 45.76 36.48
16.00 1.1240 2480 ::0 CD 27 :::3 1.2314 46.28 36 87
16.25 1.1262 25 24 21.03 27.50 1 2340 46 80 37 26
16 50 1 1284 25 68 21 36 27.75 1 2367 47.34 37.67
16.75 1.1306 20.12 21.70 28.00 1.2393 47.86 3806
17 00 1.1328 2G.5B 22 04 28.25 1 2/420 48.40 3846
17.25 1.1350 27.CO 22 38 28.50 1.;2446 48.92 38.85
17.50 1.1373 27.46 22 74 28 75 1 '2473 4946 39.25
17.75 1.1395 27.90 23.08 29.00 1 2500 50.00 39 66
18.00 1.1417 28.34 23 42 29 25 1.2527 50 ..54 ~0.06
18.25 1.1440 28.80 23.77 29.50 1.2554 51.08 <0.47
18.50 1.1462 29.24 24.11 29.75 1.2582 51.64 ~0.89
18.75 1.1485 29.70 24.47 30.CO I.Z00!> 52.18 41 30
19.00 1.1508 30.16 24.82 30.25 1.2637 52 :14 41 72
19.25 1.1531 30.62 25.18 30.£0 1.2664 53 ::3 42 14
19.50 1.1554 31.08 25.53 30 75 1.2692 53.84 42 58
, 19.75 1.1577 31.54 25 88 31 OQ 1.2719 54 28 4300
20.00 1. 16CO 3200 26 24 31.25 1.2747 54.04 43A4
• 20.25 1.1624 32.48 26.61 31.50 ~.2775 55.[0 43 89
.,20.50 1.1647 32.94 26.96 31 75 1.2804 5603 44 34
20.75 1.1671 33.42 27.33 32.00 1 2832 56.64 44.78
21.00 1.1694 33.88 27.67 32.25 1.2861 57.22. 45.24
ACIDS 2221
NITRIC ACID (Continued)
Degre~s
I Sp.
60" Gr. Degrees per Cent Degrees
Sp. Gr.
60° F" Degrees Per Cent
BaulDe. 600 F. Twaddell. HNO,. Baumel. 60° • Twaddell. HNO,.
32.50 ,j 1.2889 57.78 45.68 40.75 1.3909 78.18 63.48
32.75 1.2918 58.36 46.14 41.00 1.3942 78.84 64.20
33.00 1.2946 58.92 46.58 41.25 1.3976 79.52 64.93
33.25 1.2975 59.50 47.04 41.50 1.4010 80.20 65.67
33.50 1.3004 60.08 . 47.49 41.75 1.4044 80.88 66.42
33.75 1.3034 60.68 47.95 42.00 1.4078 81 :56 67.18
34.00 1,3063 61.26 48.42 42.25 1.4112 82.24 67.,95
34.25 1.3093 61.86 48.90 42.50 1.4146 82.92 68.73
34.50 1.3122 62,.44 49.35 42.75 1.4181 83.62 69.52
34.75 1.3152 63"04 49.83 43.00 1.4216 84.32 70.33
35.00 1.3182 63.64 50.32 43.25 1.4251 85.02 71.15
35.25 1.3212 64.24 50.81 43.50 1.4286 85.72 71.98
35.50 1.3242 64.84 51.30 43.75 1.4321 86.42 72.82
35.75 1.3273 65.46 51.80 44.00 1.4356 87.12 73.67
36.00 1.3303 66.06 52.30 44.25 1.4392 87.84 74.53
36.25 1.3334 66.68 52.81 44.50 1.4428 88.56 75.40
36.50 1.3364 67.28 53.32 44.75 1.4464 89.28 76.28
36.75 1.3395 67.90 53.84 45.00 1.4500 90.00 77.17
37.00 1.3426 68.52 54.36 45.25 1.4536 90.72 78.07
37.25 1.3457 69.14 54.89 45.50 1.4573 91.46 79.03
37.50 1.3488 69.76 55.43 45.75 1.4610 92.20 80.04
37.75 1.3520 70.40 55.97 46.00 1.4646 92.92 81.08
38.00 1.3551 71.02 56.52 46.25 1.4684 93.68 82.18
38.25 1.3583 71.66 57.08 46.50 1.4721 94.42 83.33
38.50 1.3615 72.30 '57.65 46.75 1.4758 95.16 84.48
38.'75 1.3647 72.94 58.23 47.00 1.4796 95.92 85.70
39.00 1.3679 73.58 58.82 47.25 1.4834 96.68 86.98
B9.25 1.3712 74.24 59.43 47.50 1.4872 97.44 88.32
39.50 1.3744 74.88 60.06 47.75 1.4910 98.20 89 76
39.75 1.3777 75.54 60.71 48.00 1.4948 • 98.96 91.35
40.00 1.3810 76.20 61.38 48.25 1.4987 99.74 93.13
40.25 1.3843 76.86 62.07 48.50 1.5026 100.52 95.11
40.50 1.3876 77.52 62.77
Specific Gravity determinations were made at 60° F., compare(j with water at 60° F.
Fl'OIn the Specific Gravities, the corresponding degrees Baume were calculated by the
following formula: Baume=145-~.
Sp. Gr.
Baume Hydrometers for use with this table must be graduated 1)y the above formula,
which formula should always be printed on the scale.
Atomic weights from F. W. Clarke's table of 1901. 0= 16.
ALLOWANCE FOR TEMPERATURE:
At 10° - 20° Be. - 1/30" Be. or .00029 Sp. Gr.::::: l' F.
20° - 30° Be. -1/23° Be. or .00044" "::::: 1° F.
30° - 40° Be. - 1/20° Be. or .00060" "::::: 1° F.
40° - 48.5° Be. _1/17° Be. or .00084" "::::: 1° F.
AU1'HORITY- W. C. FERGUSON.
'I.'h!s table has been approved and adopted as a Standard by the Manufacturing Chemists'
.Association of the United States.
W. H. BOWER, .JAS. I,. l\fORGAN.
HENRY HOWARD, ARTHUR WYMAN,
New York, May 14. 1903. A. G. ROSENGARTEN, Executive Committee
2222 ACIDS
NITRIC ACID
LUNGE AND RET
,
Specific
Gravity
I
100 parts by weight
contain
I liter con-
tains grams
Specific
Gravity
100 parts
by weight I liter con-
contain tainsgrams
I iSO I
IS°
in 7
vacuo %
l'I,O. %
DO. N,O.!HNO. 40
in vacuo
%
N,O. I H:O. N,O.!HNO.
-
Specific 100 parts by weight I liter con- Specific JOO parts by weight I liter con-
Gravity contain tains grams Gravity contain tainsgrams
IS° ISO
40
in vacuo
---
%
~N,O. ! R~O. N,O.!RNO. 40
in vacuo
---
%
N,O. ! R~Os N,O, !RNO.
1.3833 53.08 61.92 735 857 1.495 78.52 91.60 1174 1369
1.385 53.35 62.24 739 862 1.500 80.65 94.09 1210 1411
1.390 54.20 63.23 75.1 879 1.501 81.09 94.60 1217 1420
1.395 5:>.07 64.25 768 896 1.502 81.50 95.08 1224 1428
1.400 55.97 65.30 783 914 1.503 81.91 95.55 1231 1436
1.405 56.92 66.40 800 933 1.504 82.29 96.00 1238 1444
1.410 57.86 67.50 816 952 1.505 82.63 96.39 1244 1451
1.415 58.83 68.63 832 971 1.506 82.94 96.76 1249 1457
1.420 59.83 69.80 849 991 1.507 83.26 97.13 1255 1464
1.425 60.84 70.98 867 1011 1.508 83.58 97.50 1260 14/0
1.430 61.86 72.17 885 1032 1.509 83.87 97 84 1265 1476
1.435 62.91 73.39 903 1053 1.510 84.09 98.10 1270 1481
1.440 64.01 74.68 921 1075 1.511 84.28 98.32 1274 1486
1.445 65.13 75.98 941 1098 1.512 84.46 98.53 1277 1490
1.450 66.24 77.28 961 1121 1.513 84.63 98.73 1280 1494
1.455 67.38 78.60 981 1i44 1.514 84.78 08.90 1283 1497
1.460 68.56 79.98 1001 1168 1.515 84.92 9~.07 1287 1501
1.465 69.79 81.42 1023 1193 1.516 85.04 99.21 1289 1504
1.470 71.06 82.90 1045 1219 1.517 85.15 99.34 1292 1507
1.475 72.39 84.45 1068' 1246 1.518 85.26 99.46- 1294 1510
1.480 73.76 86.05 1092 1274 1.519 85.35 99.57 1296 1512
1.485 75.18 87.70 1116 1302 1.520 85.44 99.67 1299 1515
1.490 76.80 89.60 1144 1335
2224 ACIDS
Specific Per Per Cent. Specific Per Specific Per Per cent.
Per Cent.
GrAvity. Cent.
P,O•• H.t: 0 " GrAvity. Cent. H,PO,. GrAvity. Cent. H.PO,.
P,O•• P.O•.
- --
1.809
-
68.0 93.67 1.462
- -
46.0 63.37 1.208 24.0 33.06
1.800 67.5 92.99 1.455 45.5 62.68 1.203 23.5 32.37
1.792 67.0 92.30 1.448 45.0 61.99 1.198 23.0 31.68
1.783 66.5 91.61 1.441 44.5 61.30 1.193 22.5 30.99
1. 775 66.0 90.92 1.435 44.0 60.61 1.188 22.0 30.31
1.766 65.5 90.23 1.428 43.5 59.92 1.183 21.5 29.62
1.758 65.0 89.54 1.422 43.0 59.23 1.178 21.0 28.93
1.750 64.5 88.85 1.415 42.5 58.55 1.174 20.5 28.24
1. 741 64.0 88.16 1.40;) 42.0 57.86 1.169 20.0 27.55
1.733 63.5 87.48 1.402 41.5 57.17 1.164 19.5 26.86
1.725 63.0 86.79 1.396 41.0 56.48 1.159 19.0 26.17
1.717 62.5 86.10 1.389 40.5 55.79 1.155 18.5 25.48
1.703 62.0 85.41 1.383 40.0 55.10 1.150 18.0 24.80
1.·701 61.5 84.72 1.377 39.5 54.41 1.145 17.5 24.11
1.693 61.0 84.03 1.371 39.0 53.72 1.140 17.0 23.42
1.685 60.5 83.34 1.365 38.5 53.04 1".135 16.5 22.73
1.677 60.0 82.65 1.359 38.0 52.35 1.130 16.0 22.04
1.669 59.5 81.97 1.354 37.5 51.66 1.126 15.5 21.35
1.661 59.0 81.28 1.348 37.0 50.97 1.122 15.0 20.66
1.653 58.5 80.59 1.342 30.5 50.28 1.118 14.5 19.97
1.645 58.0 79.90 1.336 30.0 49.59 1.113 14.0 19.28
1.637 57.5 79.21 1.330 35.5 48.90 1.109 13.5 18.60
1.629 57.0 78.52 1.325 35.0 48.21 1.104 13.0 17.91
1.621 56.5 77.83 1.319 34.5 47.52 1.100 12.5 17.22
1.613 55.0 77.14 1.314 34.0 46.84 1.096 12.0 16.53
1.605 55.5 76.45 1.308 33.5 46.15 1.091 11.5 15.84
1.597 55.0 75.77 1.303 3~.0 45.45 1.087 11.0 15.15
1.589 54.5 75.08 1.298 32.5 44.77 1.083 10.5 14.46
1.581 54.0 74.39 1.292 32.0 44.08 1.079 10.0 13.77
1.574 53.5 '73.70 1.287 31.5 43.39 1.074 9.5 13.09
1.566 53.0 73.01 1.281 31.0 • 42.70 1.070 9.0 12.40
1.559 52.5 72.32 1.276 30.5 42.01 1.066 8.5 11.71
1.551 52.0 71.63 1.271 30.0 41.33 1.062 8.0 11.02
1.543 51.5 70.94 1.265 29.5 40.64 1.058 7.5 10.33
1.536 51.0 70.26 1.260 29.0 39.95 1.053 7.0 9.64
1.528 50.5 69.57 1.255 28.5 39.26 1.049 6.5 8.95
1.521 50.0 68.88 1.249 28.0 38.57 1.045 6.0 8.26
1.513 49.5 68.19 1.244 27.5 37.88 1.041 • 5.5 7.57
1.505 49.0 67.50 1.239 27.0 37.19 1.037 5.0 6.89
1.498 48.5 66.81 1.233 26.5 36.50 1.033 4.5 6.20
1.491 48.0 66.12 1.228 26.0 35.82 1.029 4.0 5.51
1.484 47.5 65.43 1.223 25.5 35.13 1.025 3.5 4.82
1.476 47.0 64.75 1.218 25.0 • 34.44 1.021 3.0 4.13
1.469 46.5 64.06 1.213 24.5 33.75 1.017 2.5 3.44
2228 ACIDS
/_
V ~~ V
V ~ "-
9
'\ +
,;..,
.;
!/ 1\ ~
V
""
C)
.... ...
;;:
';J II 1\ ]
""-
If) 1.83 vi / /
I J.837
1.817
D.
on
-/ \I .. 300' 149
80 \ IlL<.
\ ,\I
-100 . .
I
'c50= IZI
FIG. 308. '"
Chart Showing Freezing- and Boilirrg:j;oints of Sulfuric Acid
of Varying Concentration.
ACIDS 2229.
Sulfuric acid readily absorbs SOs so that its acidity may be considerably
over 100%. This acid, commonly known as oleum, fumes when exposed to
the air due to its low vapor tension, the SOs combining with the moisture of the
air with formation of H 2S0 4 mist.
SULFURIC ACID
METHOD OF ANALYSIS 15
50 to 100 grams of the.acid are weighed into a platinum or silica dish and the
acid evaporated (Hood) by heating gently over a direct flame. 'When no more
fumes evolve, the dish is cooled and weighed. The increased weight of the dish
is due to the residue, composed of substances not volatile at red heat and
generally contains iron oxide. Chamber acid may have all of the substances
occurring in sulfur ores present in this residue.
16 Certain details of procedure are published by courtesy of the General Chemical
Company. .
2230 ACIDS
DETERMINATION OF LEAn
DETERMINATION OF IRON
DETERMINATION OF ARSENIC
,
Traces of Arsenic. Amounts less than .OOOp.s%.-Take 50 grams or more
of the acid and distill off the arsenic as As'Cla-according to the distillation
ACIDS 2231
method given on page 91. Determine arsenic in the distillate by the Cut zeit
method. .
Amounts less than .005% may be determined directly by the Gutzeit
metnod. When the arsenic is greater than 0.05% it may be determined by
reducing the arsenic with tartaric acid and titrating with standard iodine solu-
tion, after neutralizing the acid with ammonia and adding sodium bicarbonate.
25-50 grams of the acid are taken and placed in a Kjeldahl flask and! gram of
tartaric acid added together with 2 grams of fused potassium bisulfate. The
acid is heated over the direct flame until the color, first becoming dark, changes
to a straw color. Fuming will cause loss of arsenic and should be avoided ..
The acid is washed into a beaker with 250 ml. of water, then neutralized with
NH 40H, and NaHC0 3 added. The titration with iodine is made in presence
of starch. The end-point is a blue color.
If antimony is present a separation of the arsenic must be made by distil-
lation. 1 ml. NjlO I=.003748 g. As or .004948 g. As 20 a• As 20 a Xl.11616
= As 2 0 •.
DETERMINATION OF ANTIMONY
The arsenic is first removed by distillation as AsCla (page 91, Vol. J). The
antimony is now distilled by adding 20 grams of zinc dissolved in 20 grams of
concentrated HCl. The antimony in the distillate is determined by the modi-
fied Gutzeit method given in the chapter on Antimony, Vol. I.
DETER1~INATION OF ZINC
DETERMINATION OF 'SELENIUM I
Traces.-l0 ml. of the acid diluted with an equal volume of water are placed \
in a test tube and 4 or 5 drops of 1 % KI solution added. ,Heat to boiling to
expel free iodine. The brick red color is compared with standards containing
known amounts of selenium made up in the same way as the sample. If
sufficient selenium in present this is thrown down as a brick red finely divided
preci pita te.
Standard selenium solution is prepared by dissolving .1 g. pure Se in 5 ml.
HNO a and 10 ml. HCl. The solution is evaporated to dryn~ss, then taken up
with water and a little dilute H 2S0 4 and made up to 10001)11. 1 m1.=.OOOl g. Se.
Larger Amounts.-500 grams of the. acid is diluted by pouring into a liter
of water, 100 ml. concentrated HCI added and the solution saturated with S02.
The selenium is allowed to settle 15 to 20 hours, then filtered through a filter of
fine texture. The precipitate is dissoh;ed by placing the filter and' precipitate in
sulfuric acid nitric acid mixture 5 ml. H 2S0 4 +50 mL HNO a. The solution
is evaporated to fumes, adding more HN0 3 if the solution is dark, and again
evaporated to expel HNO a. After diluting with water and filtering, the filtrate
is made up to about 100 mL and gassed boiling hot with H 2S. The precipitated
selenium is settled, filtered into a Gooch, washed with water, then alcohol and
ether and dried with a current of warm air to constant weight.
Sulfur Dioxide Determination in Sulfuric Acid.-1-10 ml. of N /10 iodine
solution are run into 200 ml. of water, starch solution is added and the acid to
be tested is allowed to flow from a burette into this reagent, until the blue color
disappears.
1 mL N /10 iodine = .0032 g. S02.
SULFURIC ACID
Hydrochloric Acid Determination in Sulfuric Acid.-50 to 100 ml. of the
acid is diluted to about 200 ml. by running the acid,'into chlorine-free water.
10 ml. of 1% silver nitrate solution is added and 'the ttirbidity compared against
a standard made by adding standard sodium chloride solution from a burette
into a solution containing an equal amount of sulfuric acid (free from chloride)
diluted with water and containing silver nitrate as in case of the sample. When
the turbidity is the same as the sample the nu~ber of ml. of NaCl solution is
noted. The turbidity is best seen by placing the two solutiollil. side by side on
a black paper. Standard 'chloride solution contains .16 g ..NaCl per liter.
1 ml. = .0001 g. HCl.
If considerable hydrochloric acid is present it may be deterrhineq. by Vol-
hard's method given on page 271, Vol. 1. The sample of acid is diluted to about
20% strength. A measured excess of N /10 AgN9a added and the excess ti-
trated with Nj10 NH 4CNS solution in presence o~ ferric ammonium alum.
1 ml. N/lO AgNO a=.003647 g. HCI.
the combined water in the product decreasing (along with sulfuric acid) with
the increase of free 80 a or sulfuric anhydride. Mixed acid is the technical
name for the mixture of concentrated sulfuric acid and nitric acid or of oleum
and nitric anhydride, the product being commonly used in nitrating glycerine,
cotton and other materials. The analysis includes the determinations of
H 2S0 4, HNO a, N 20 5, N 20a and in the case of oleum, the determination of SOa.
In the presence of the latter, HNO a is assumed to lose its combined water
according to the reaction:
2HN03+S0a=H2S04+N20S.
In absence of nitric acid S02 may be present. It is assumed that if HNO a is
present the S02 is oxidized to 80 3 with formation of H 280 4 and the anhydrides
SOa and N 20a according .to the reaction. .
N205+H20+2S02=N20a+SOa+H2S04.
Some chemists prefer to express the reaction:
2HN0 3 +S0 2 = H 280 4 +N 20 4 •
2234 ACIDS
LOWER OXIDES
Ten m!. of the sample, weighed in a pipette with, capillary delivery tip, are
cautiously run into about 400 mL of cold water, k~~ping the delivery tip well
under the water and in rapid motion to prevent overheating through local.
action. The mixture is titrated with N /10 KMn04 until a pink tint is obtained
that does not fade in three minutes.
1 ml. N/10 KMn04=0.0019 g. N 2 0 a (N 2 0 a to N 2 0 4 fa~tor=2.42.)
1 ml. N/10 KMn04=0.0046 g. N 20 4• - ----
1 ml. Nj10 KMn04=0.00235 g. HNO z. Equivalent H ZS0 4= 0.0049043 g. per mL
1 m!. N/10 KMn04=0.0032035 g. SOz. -
1 mL N /10 KMn04 =0.0041043 g. H 2SO a.
NOTE.--With exactly N/10 KMn04 on a 19-9ram slfmple 1 ml. =0.01 % N 20 a.
oxides and nitric. The evaporation is hastened by blowing a current of hot, dry,
pure air over the sample, see Fig. 309. Instead of a cass'erole, a shallow glass
cell, 3 inches in diameter and 1! inches deep may be used. The air current in
this case is unnecessary. About 5 ml. of water are added and this again
evaporated off. The acid thus obtained is H zS0 4 +SO a.
The acid is taken up with water and titrated with standard caustic, using
CO 2 free water and phenolphthalein indicator.
One ml. Nil NaOH=O.004904 gram H 280 4•
NITRIC ACID
This may be obtained from the above determinations according to calcula-
tions following. It may be obtained by direct titration with standard ferrous
sulfate, by running a weighed amount of mixed acid into 100 ml. of cold sulfuric
acid and titrating to a yellowish red tint. For the detailed procedure see Nitric
Acid: page 2216.
Calculation of Results. A. Nitric Acid and S02 being Absent.-The total
acidity is calculated to 80 3. Reference is made to the table for Oleum from
which the % H 280 4 and free 80 a are obtained.
Example.-8uppose the total acidity in terms of SOa was found to be 84.2.
The acid contains 86% H 2S0 4 and 14% free 80a.
B. Nitric Acid Absent, S02 is Present in the Mixture. Total SO a.-From
total acidity as 80 a subtract 80 2 X 1.25 (i.e., equivalent 80 a).
Combined Water.-100-(total 80 a+S0 2 ) =H 20.
Combined SOa.-HzO X 4.4445 = 80'3 equivalent or combined SOa.
Free SO a.-Total SOa-combined SOa=free SOa.
Sulfuric Acid.-Combined SOa+H 2 0.
Example.-If S~ was found to be 2% and the total acidity in terms of 80 3
=83.5%.
Total 80 3 = 83.5 - (2 X 1.25) = 81 %.
Water=100-(81+2) =17%. Report
Combined 80 a = 17 X 4.4445 = 75.56% 8ulfuric acid = 92.56%.
Free 80 a = 81-75.56 = 5.44%. Free SOa = 5.44.
Sulfuric acid =75.56+17 = 92.56%. 80 2 =2.00. Total 100.
C. Nitric Acid Present and SOz Absent.-Nitric acid in presence of free 80 a
is assumed to be the anhydride N Z05'
N z0 5.-From the total acidity is subtracted the acidity after evaporation,
both being calculated to equivalent 80 a. The difference multiplied by 1.349
=% N Z0 5• •
HzO.-Total SOa (after evaporation)+N 20 5 subtracted from 100=H 2 0.
Combined SOa.-H 2 0 X 4.4445 = SOa equivalent to H 2 0.
Free SO a.-Combined SOa subtracted from total 80 a=free 80 a.
Sulfuric Acid.-Combined SOa+H zO=H zS0 4 •
Example.-If total acidity=84% in terms of 80 a and the total 80 a (after
evaporation) =82%, then the difference 2X1.349=2.698% N 20 5 •
Water = 100-(82+2.698) = 15.302%. Report
2236 ACIDS
A-DZ A-DX
(a) A-B =Y; X-Z=Y; or (b) D-B =Y.
DILUTION OF SOLUTIONS
If a=per cent of solution desired, b=specific gravity of solution to be
diluted and c its percentage then
Volume or ratio of the strong solu!ion is d
a
- - 1 - =d.
bX c
100
. Volume or ratio of wate: required .
a
100--- o r 100-bXd.
c+100
Example.-To make 75% sulfuric acid from 87% solution having a specific
gravity of 1.8. .
Substituting in above formula 7~ 7 = m!. or liters or cubic feet, etc. of
1.8X .8
acid to be taken: i.e. ratio is 47.9 and 100 - 0~:7 = m!. or liters or cubic feet of
water required for dilution, i.e. ratio is 13.79.
47.9 m!. of the strong acid diluted with 13.79 m!. of water will make the
acid 75% strength.
It is evident that the ratio being known large volumes of solution may be
made so that the procedure is appllcable for factory use.
2238 ACIDS
SULFURIC .ACID
By W. C. FERGUSON AND H. P. -tALBOT
SpecIl!c
Per Weight of
, Pounds O. V.
Degrees Gravity Degrees I. Per Cent
BaualC. 60° Twaddell. cent 1Cu. Ft. iu O. V•• in
6~O F. H,SO•. Lbs. A". x Cu bic Foot.
I
--
0 1.0000 0.0 0.00 62.37 0.00 0.00
1 1.0069 1.4 1.02 62.80 1.09 0.68
2 1.0140 2.8 2.08 63.24 2.23 1.41
3 1.0211 4.2 3.13 63.69 3.36 2.14
4 1.0284 5.7 4.21 64.14 4.52 2.90
5 1.0357 7.1 5.28 64.60 5.67 3.66
6 1.0432 8.6 6.37 65.06 6.84 4.45
7 1.0507 10.1 7.45 65.53 7.99 5.24
8 1.0584 11. 7 8.55 66.01 9.17 6.06
9 1.0662 13.2 9.66 66.50 10.37 6.89
10 1.0741 14.8 10.77 66.99 11.56 7.74
11 1.0821 16.4 11.89 67.49 12.76 8.61
12 1.0902 18.0 13.01 68.00 13.96 9.49
13 1.0985 19.7 14.13 68.51 15.16 10.39
14 1.1069 21.4 15.25 69.04 16.36 11.30
/'
-
15 1.1154 23.1 16.38 69.57 17.58 12.23
16 1.1240 24.8 17.53 70.10 18.81 13.19
17 1.1328 26.6 18.71 70.65 20.08 14.18
18 1.1417 28.3 19.89 71.21 21.34 15.20
19 1.1508 30.2 21.07 71.78 22.61 16.23
20 1.1600 32.0 22.25 72.35 23.87 17.27
21 1.1694 33.9 23.43 72.94 25.14" 18.34.
22 1.1789 35.8 24.61 73.53 26.41 19.42
23 1.1885 37.7 25.81 74.13 27.69 20.53
24 1.1983 30.7 27.03 71'1. 74 I 2\).00 21.68
Sp. Gr. determinations were made at GOo F., compared with water at 60° F.
From the Sp. Grs., the corresponding degrees Baume were calculated by
the following formula: Baume = 145-145/Sp. Gr.
Baume Hydrometers for use with this table must be. graduated by the
above formula, which !ormula should always be printed ·the _scale.on
* 66° Baume = Sp. Gr. 1.8354 = Oil of Vitriol (0. V.).
1 cu. ft. wat~r at 60° F. weighs 62.37 lbs. avo -
Atomic weights from F. W. Clarke's table of 1901. 0 = 16.
H 2S0 4 = 100 per 'cent.
% H 2S04 % O. V. I %600
O. V. = 93.19 100.00 =d 119.98
60° = 77.67 83.35 ~ 100.00
50° = 62.18 66.72 = 80.06
ACIDS 2239
SULFURIC ACID (Continued)
-Freezing
Degrees (Melting)
Baume. ~oint.
F. . APPROXIMATE BOILING POINTS
50° B, 295 0 F.
--- 60° " 386° "
0 32.0 61° " 400 0 "
1 31.2 62° " 4150 "
2 30.5 63° " 432° "
3 29.8 64° " 451° "
4 28.9 65° " 4850 "
Acids stronger than 66° Be. should have their percentage compositions
determined by chemical analysis.
* Calculated from Pickering's results, Jour. of Lon. Ch. Soc., vol. 57, p. 363.
AUTHORITIES - W. c. FERGUSON; H. P. TALBOT.
This table has been approved and adopted as a standard by the Manu-
facturing Chemists' Association of the United States.
W. H. BOWER,
HENRY HOWARD,
JAS. L. MORGAN,
ARTHUR WYMAN,
A. G. ROSENGARTEN,
New York, June 23, 1904. Jiucuti've Committe4.
2240 ACIDS
Degrees "(Melting)
Freezing
B"ume. Point.
of.
-- ALLOWANCE FOR TEMPERATURE
25 -23
26 -30 At 10° Be .. 029° Be. or .00023 Sp. Or. = 1°F.
27 -39 " 20° " .036° ."" .00034 " = 1°"
28 -49 " 30° " .035° .00039 " = 1°"
29 -61 " 40° " .031° " .00041 " = 1 "
0
• " 50° " .028° " .00045 " = 1°"
30 -74 " 60° " .026° " .00053 " = 1°"
31 -82 " 63° II" .026° " .00057 " = 1°"
32
33
-96
-97
," 66° .0235° " . .00054 " = 1°"
34 -91
35 -81
36 -70
Per Cent Pounds Per Cent Pounds
37 -60 60v Bauma Baume
500
60° in 50° in
38 -53 Baume. Baume.
I Cubic Foot. I Cubic Foot.
39 -47
----. ---
40 -41 61.93 53.34 77.36 66.63
41 -35 63.69 55.39 79.56 69.19
42 -31 65.50 57.50 81.81 71.83
43 -27 67.28 59.66 84.05 74.53
44 -23 69.09 61.86 86.30 77.27
.. }. ~
~
91.63
93.67
93.11
96.26
114.46
117.00
116.30
120.24
58
59 - "
7 '" 95.74
97.84
99.52
102.89
119.59
122.21
124.31
128.52
,
2242 ACIDS
SULFURIC ACID (Continued)
,
Specific Weight of Pounds O. V.
Degrees Gravity Degrees Per Cent 1 Cu. Ft. in Per Cent in
Baume. 60 0 F. Twaddell. H,SO •. Lb•. Av. O. V. 1 Cubic Foot.
60 0
65!
65!
13.4
.- 1
116.65
118.19
132.70
134.88
145.71
1:i7.63
165.76
168.48
.
66 -29 119.98 137.34 149.87 ~171.56
Per Cent.
Be. Sp. Gr. at 60 0 F. H2S0. Wt. 1 Cu. Ft. (
,
Allowance for Temperature.
Total
SO.
Equivalent
H.BO.
Per Cent
H,SO.
PerCent
Free SO.
Total
SO.
Equivalent
H,SO.
I
Per Cent
H.BO.
I Per Cent
Free SO.
,
81.63 . 100.00 100 0 90.82 111.25 50 50
81.82 100.23 99 1 91.00 111.48 49 51
82.00 100.45 98 2 91.18 111.70 48 \ 52
82.18 100.67 97 3 91.37 111.93 47 \ 53
82.37 100.90 96 4 91.55 112.15 46 54
82.55 101.13 95 5 91.73 112.37 45 55
82.73 101.35 94 6 91.92 112.60 44 56
82.92 101.58 93 7 92.10 112.82 43 57
83.10 101.80 92 8 92.29 113.05 42 58
83.29 102.03 91 9 92.47 113.28 41 59
83.47 102.25 90 1J. 92.65 113.50 40 60
83.65 102.47 89 11 92.84 113.73 39 61
83.84 102.70 88 ' 12 93.02 113.95 38 62
84.02 102.92 87 13 93,20 114.17 37 63
84.20 103.15 86 14 93.39 114.40 36 64
84.39 103.28 85 15 93.57 114.62 35 65
84.57 103.60 84 16 93.76 114,85 34 66
84.75 103.82 83 17 93.94 115.08 33 67
84.94 ]0405 82 18 94.12 115.30 32 68
85.12 104.27 81 19 94.31 115.53 31 69
85.31 104.50 80 20 94.49 115.75 30 70
85.49 104,73 79 21 04.67 115,97 29 71
85.67 104.95 78 22 04.86 116.20 28 72
85.86 105,18 77 23 95.04 116.42 27 73
86.04 105.40 76 24 95.22 116.65 26 74
86.22 105.62 75 25 95.41 116.88 25 75
86.41 105 85 74 26' 95.59 117.10 24 76
86.59 106.07 73 27 95.78 117.33 • 23 77
86.78 106.30 72 28 95.96 117.55 22 78
86.96 106.53 71 29 96.14 117,77 21 79
87.14 106.75 70 30 06.33 118.00 20 80
87.33 106.98 69 31 96.51 118.22 19 81
87.51 107.20 68 32 96.69 118.45 18 82
87.69 107.42 67 33 96.88 118.68 17 83
87.88 107.65 66 34 97.06 118,90 16 84
88.06 107.87 65 35 97.25 119.13 15 85
88.24 108.10 64 36 97.43 119.35 14 86
88.43 108.33 63 37 97,61 119,57 13 87
88,61 108,55 62. 38 97.80 119.80 12 88
88,80 108.78 61 39 97,98 120.03 11 89
88.98 109.00 60 40 98.16 120.25 10 90
89.16 109.22 59 41 98.35 120.48 9 91
89.35 109.45 58 42 98.53 120.70 8 92
89.53 109.67 57 43 98.71 120.92 7 93
89.71 109.90 56 44 98.90 121.15 6 94
89.90 110.13 55 45 99.08 121.37 5 95
90.08 110.35 54 46 99.27 121.60 4 96
90.27 110.58 53 47 99.45 121.83 3 97
90.45 110.80 52 48 99.63 122.05 2 98
90.63 111.02 51 49 99.82 122.28 1 99
100,00 122.50 0 100
Compiled from the table by H. B. Bishop, Van Nostrand's Chemical Annual, 1913;
2244 ACIDS
,Factors:
HCIX1.0978=SOa, HCIX3.1956=SOa· HCland HCI X2.1959 = SOainSOaHCl.
SOaXO.8998=HCI, SOaX 1.4555=SOa·HCI and SOa XO.4554=HClin SOaHC!.
H 20 X 5.4444 = H 2 S0 4 , H 3S0 4 XO.1837 = H 20, H 2S0 4'XO.7436=HCl.
NaCIXO.6238=HCI, S03·HCI=31.29% HCI and 68.71% SOa.
The details of procedure for the determination of boric acid are given in the
chapter on Boron in Volume 1. The method takes advantage of. the fact that
boric acid is a weak acid having an end-point in an alkalimetric titration of
pH 10-11 hence cannot be titrated in presence of indicators having a lower pH
range, such as methyl red or methyl orange, pH 3.1-6.3. It is possible, there-
fore, to liberate boric acid from its combination by means of a mineral acid,
titrate or neutraliz.e the mineral acid with NaOH in presence of methyl red or
18 C. A. Ahlum, The Analyst, 31, 168 (1906).
2246 ACIDS
methyl orange and then titrate boric acid in- presence of phenolphthalein indi-
cator after the addition of a polyhydric alcohol. such as 'glycerol or mannitol.
The borate is first converted to a soluble sodiu~ salt by transposition by
means of sodium carbonate and extraction w~th water pr the boric acid is set
free by boiling with a mineral acid, using a reflux condenser to prevent loss of
boric acid, which is volatile with steam. If sodium carbonate treatment is
used the CO 2 must be expelled by acidifying and heating fot: a few minutes in
a covered beaker avoiding vigorous boiling. The free mineral acid is now
neutralized with NaOH in presence of methyl orange or methyl red indicator.
Neutral glycerol is added in sufficient quantity (25 ml. is sufficient for 0.5 g.
B 2 0 s) and again the solution is neutralized if an acidity is evident from the
glycerol addition (commercial glycerol is apt to contain free acid). Phenol-
phthalein indicator is added and the boric acid titrated with standard NaOH
until a reddish pink color of the combined indicators is obtained.
The solution titrated and the reagents used must be free from CO 2, iron,
aluminum and products other than boric acid, that would react with the alkali
in the titration. . .
1 ml. of 1 N NaOH is equivalent to 0.03482 g. B 20 a•
ORGANIC ACIDS
Titrations of organic acids are made preferably with phenolphthalein.
Methyl orange cannot be used. The wate:.: used for diluting the organic acid
must be free of carbon dioxide.
The more important impurities that are looked for in commercial acetic acid
are formic acid, furfurol, acetoJ;le, sulfmic acid, sulfurous acid, hydrochloric
acid, metals.
In the examination of the acid the physical appearance-turbidity and color
are noted.
FORMIC ACID IN ACETIC
Qualitative.-Ten ml. of t,he acid (glacial diluted 1 : 10) are heated with 1
gram of sodium acetate and 5 ml. of 5% mercuric chloride solution. A turbidity ~
indicates formic acid.
Quantitative.-Five grams of glacial acetic acid or corresponding quantity of
dilute acid are treated with 5 grams of sodium acetate and 40 ml. of mercuric
chloride solution (5%) and 30 ml. of water added. The mixture is heated for
two hours in a flask with a return condenser, the flask being surrounded by
steam.' The precipitated mercurous chloride, HgCI, is filtered off, dried and
weighed. .
Weight of HgC1XO.0977 = formic acid equivalent.
This is best detected by placing in a small flask about 20 m!. of the sample,
adding 5 m!. of strong hydrochloric acid and about 3 grams of zinc and covering ~
with a filter paper saturated with lead acetate. The blackening of the paper
indicates S02 in the sample (e.g., reduced to H 2S by the hydrogen generated by
the zinc).
The sulfurous acid is best titrated with N /10 iodine solution, using starch
indicator. 1 ml. N /10 I2 = .0032 gr. S02.
ACETATES
Two to 5 grams of the material is placed in a KjeJdahl flask connected by
means of a condenser to a receiving flask containing half normal caustic. About
20 m!. of 85% phosphoric acid are added and about 150 m!. of water. Gentle
heat is applied and gradually increased. About 100 m!. of the solution is
distilled into the caustic. Additional hot water is added to the residue in the
Kjeldahl flask and the distillation continued. This is repeated until about 800
m!. of solution has been distilled over. The CO 2 is boiled out of the distillate, a
reflux condenser being used to prevent loss of the acetic acid. If the solution
is alkaline, a known amount of acid is added and the CO 2 boiled out. The
excess acid is now titrated and the amount of acetic acid in the distillate
calculated. •
One ml. N/2 NaOH=O.030015 gram CHaCOOH
ClfaCOOH X 1.3169 = Ca(CH aC0 2)2, or X 1.3663 = CH 2 COONa.
Acetates of the Alkalies and Alkaline Earths.-In absence of other organic
acids, nitrates, etc., a quick method is suggested by Sutton (Vol. Analy., X. Ed.,
p. 91). The salts are converted into carbonates by ignition and the residue
titrated with normal acid.
One m!. N /1 acid = 0.06003 gram CH 3COOH.
2252 ACIDS
ACETIC ACID AT 15°
OUDEMANS
Reaction.-N a 2 CO a + 2H CI = 2N aCI + H 20 + CO 2 •
Procedure.-The solution is titrated with standard HCI, using methyl
orange indicator. Carbonates of barium, strontium, calcium and magnesium
are titrated with an excess of hydrochloric acid and this excess determined with
standard alkali.
1 ml. N/10 HCI=0.0022 g. CO 2 •
CITRIC ACID
The free acid may be titrated with sodium hydroxide, using phenolphthalein·
indicator. One ml. N /1 alkali = 0.07 gram crystallized citric acid. •
About 3 grams of the oxalic acid or its salt are dissolved in 200 m!. of CO 2
free 'water and 50 m!. 2N sulfuric acid added. The solution is heated to
about 70° C. and titrated with standard, normal solution of potassium per-
manganate, to a faint pink, persisting for three minutes.
1 m!. N /KMn04 = 0.04501 g. HlJ 2 0 4 or 0.06302 g. H 2C 2 0 4· 2H 20.
The acid may also be titrated with standard caustic solution. Titration is
made in a hot solution, using phenolphthalein indicator.
TARTARIC ACID
Tartaric acid, cream of tartar, Rochelle salt, tartar ~metic, normal potas~
sium tartrate, iron tartrate and other salts of tartaric arid are obtained from
the residues of wine manufacture. The raw materials-consist of lees, tartars,
calcium tartrate. Tartaric acid is present in these re~idues in the form of
potassium hydrogen tartrate (vitartrate of potassium) br as normal calcium
tartrate. In the examination of the raw material for its evaluation total
tartaric acid and that present as bitartrate are determined.'
Iron and Alumina (alum).-These are determined in the ash of the ignited
prodjlct, the ash being dissolved in hydrochloric acid. P 2 0 5 may also be deter-
mined in this ash. •
Arsenic.-This may be determined by the Gutzeit Method. (See chapter
on Arsenic in Volume 1.)
Lead.-This is determined by the colorimetric prflcedure given on page 517,
Vol. 1.
Free Sulfuric Acid in Tartaric Acid Liquors.-The solution ·10 ml. or more
is treated with ten times its volume of alcohol, and after settling overnight
the mixture is filtered .• Sulfuric acid is determined in the alcohol filtrate in
the usual way. - .
ALKALIES
PROCEDURE
Ten grams of the hydroxide are dissolved in water and the solution made
up to exactly 500 ml. Aliquots of this solution are taken for the following
determinations.
Total Alkali.-FiftY mL of the caustic solution, equivalent to 1 gram of the
solid, are titrated with N /H 2S0 4 in presence of methyl orange indicator, until
the faint pink end-point is obtained.
1 mI.. N/H 2S0 4 =.031 grams of Na 20 (total alkali actual),.
NOTE.-In t4e ~ew York and Liverpool test N /H2S04 value is .032 g. NaOH per mi.
J
2256 ALKALIES
Sodium Hydroxide.-Fifty ml. of the sample is treated with 100 m!. of 10%
BaCh solution and the N aOH then titrated with lV/H 2S0 4 in presence of
phenolphthalein indicator.
I
1 mI. N/HzS04=.04 gram NaOH.
Sodium Carbonate.-Multiply the difference betwe~n the total alkali titra-
tion and the titration for NaOH by .. 053 the result is grams NuzCO s.
Alternative Methods. Sodium Hydroxide and Sodium Carbonate.-Car-
bon dioxide is readily picked up by NaOH so that it is invariably present in
caustic soda forming sodium carbonate. The carbonate and hydrate may be
determined in the same solution as follows .
.Fifty mI. of the sampl~equivalent to one gram of the solid, are titrated with
normal H 2S0 4, in presence of phenolphthalein, until the pink color just dis-
appears. This occurs when all of the sodium hydroxide is neutralized and the
carbonate has been converted to bicarbonate. This titration may be recorded'
as mI. A.
Methyl orange is now added and the titration co;tinued until the yellow
color changes to pink.' This titration is recorded as ml, B. '
Then NaOH=(A-B)XO.04 and Na 2 C0 3=2BXO.053.
DETERMINATION OF IMPURITIES
/'
POTASSIUM HYDROXIDE
,~eagents and Indicators: Water.-The water used in the following methods should
be distilled water, neutral to phenolphthalein and methyl orange indicators.
Hydrochloric Acid (1-1).-Dilute 500 m!. of colorless hydrochloric acid (Hel) of 1.19
specific gravity with an equal volume of distilled water.
Sulfuric Acid (l-J).-Into 3 liters of distilled water cautiously add with stirring 1
liter of colorless sulfuric acid (H 2S0 4) of 1.84 specific gravity.
Nitric Acid (1-4).-Into 4 liters of distilled water add with stirring 1 liter of colorless
·concentrated nitric acid (HNO,).
Ammonium Hydroxide.-The ammonium hydroxide (NH 40H) used should be a
pure, clear, colorless solution of 0.90 specific gravity.
Ammonium Hydroxide (Dilute).-Dilute 100 ml. of ammonium hydroxide (NH 40H)
specific gravity 0.90 with 900 mI. of distilled water.
Ammonium Sulphide Solution.-This solution is prepared by passing hydrogen
sulfide gas (H 2S) through a dilute ammonium hydroxide solution (NH.OH) until the
solution smells strongly of hydrogen sulfide.
Ammonium Oxalate Solution.-Dissolve 40 grams of ·ammonium oxalate, (NH 4)2-
C 20 4 • H 20, in hot distilled water and dilute to one liter. .
Disodium Phosphate Solution.-Dissolve. 100 grams of disodium phosphate (Naz-
HPO •. 12H 20) in distilled water and dilute to one liter.
Starch Solulion.-Dissolve 0.5 gram of soluble starch in 100 ml. of hot water. If
soluble starch is not available, macerate 0.5 gram of ordinary starch to a creamy paste in
a mortar with cold water, wash into 100 ml. of boiling water and boil for several minutes.
19 Ref. J. C. Chiarino, C. A., 18, 1799 (1924).
20 Standard Method of the Solvay Process Company. By courtesy of the Company.
2258 ALKALIES
Allow to stand for a few minutes and then pour off. the supernatant liquor for use. Starch
Starch solutions do not keep well and for best results they· should b~ prepared fresh every
few weeks. .
Sodium Thiosulfate Solution.-Dissolve lOOgrams of sodium thiosulfate (Na 2S 20.·-
5H 20) in distilled water and dilute to one liter. I
Phenolphthalein.-Dissolve 3.5 grams of phenolphthalein 7n 650 m!. of grain alcohol
and make up to 1000 mi. with distilled water.
Potassium Chromate Solution.-Dissolve 50 grams of neutral potassium chromate
(K 2Cr0 4) in 400 mi. of distilled water. Add enough silver nitrate to produce a slight
red precipitate for the removal of chlorides. Filter and make up the filtrate to 1000 mi:
with distilled water.
Methyl Orange.-Dissolve one gram of methyl orange in 1000 mi. of distilled water.
Potassium Sulfocyanate Solution.-Dissolve 50 grams of C. P. potassium sulfo-
cyanate (KCNS) in distilled water and dilute to one liter.
Standard Solutions: Tenth Normal Sulfuric Acid (N/l0 H 2 S0 4 ).-This solution
should be prepared from a stock solution of normal sulfuric acid (N /H2S04) by diluting
exactly 100 mi. to 1000 mi. with distilled water.
The stock solution of normal sulfuric acid is prepared as follows:-This solution
should contain 49 grams per liter of sulfuric acid (H 2S0 4) in distilled water. Mix 30 mi.
of pure sulfuric acid of 1.84 specific gravity with 150 mi. of distilled water in a 1000 mi.
measuring flask and allow to cool to room temperatur!'). Dilute to 1000 m!. with distilled
water and adiust the volume by titrating against 5.3 grams of chemically pure, dry
sodium carbonate (Na 2CO.) until 100 mi. of the acid will just neutralize the 5.3 grams 'of
sodium carbonate, using methyl orange as an indicator.
A watch glass should cover the beaker while' adding the acid to avo,id mechanical
loss. As the end-point is approached care should be taken to stir well, so that the color
change from yellow to pink will be sharp and easily noticeable, the acid being added a
drop at a time. /
Tenth Normal Sodium Hydroxide (N/l0 NaOH).-This solution should be prepared
from a stock solution of normal sodium hydroxide (N /NaOH), by diluting exactly 100
mi. to 1000 mi. with distilled water.
The stock solution of normal sodium hydroxide is prepared as follows:-This solution
should contain 40 grams per liter of sodium hydroxide (NaOH). Dissolve 41 grams
of chemically pure sodium hydroxide in 250 mi. of distilled water in a 1000 ml. measuring
flask. Allow to cool to room temperature and dilute to 1000 mI. Adjust by adding
distilled water until 50 mI. are exactly neutralized by 50 ml. of normal sulfuric acid
(N /H ZS0 4), titrating cold with 5 drops of phenolphthalein indicator. ~
Twentieth Normal Silver Nitrate (N/20 AgNO.).-This I?olution should contain 8,495
grams per liter of silver nitrate (AgNO.). Dissolve 8.5 grams of chemically pure silver
nitrate in distilled water to which one drop of colorless nitric acid (HNO.) of 1.42 specific
gravity has been added and dilute to 1000 mI. with distilled water. This solution is
practically 1/20 normal. It may be standardized as follows:-Weigh out 0.8025 grams
of chemically pure, dry, ammonium chloride (NH.Cl). Dissolve in 250 mI. of distilled
water and titrate 25 mI. of this solution with the silver nitrate solution, using 3 or 4
drops of pota~sium chromatfil (K.Cr04) solution as an indicator,~l).til the red tint of the
silver chromate just appears. If the silver nitrate solution is exactly,.N /20, 30.0 mI.
will be consumed., ,
Tenth Normal Potassium Permanganate (N /10 KMnO.).-This' solution should
contain 3.16 grams per liter of chemically pure potassium permanganate (KMn04).
Dissolve 3.2 grams of potassium permanganate crystals in 500 mI. of distilled water in 'a
1000 mI. measuring flask. Dilute to 1000 ml. Standardize as follows:-Add 0.67
grams of recrystallized, chemically pure sodium oxalate( (NazC 20.) and about 150 mi.
of distilled water to a liter boiling flask. Heat until dissolved, then add about 15 mi.
of (1-3) sulfuric acid (H 2S0 4). Heat nearly to boiling and titrate with the potassium
permanganate solution until a permanent pink color is obtained.
Adjust the solution to the required strength of calculate the proper factor from the
following data: I
100 m!. N/10 KMnO. solution =0.67 grams of sodium-oxalate.
The potassium permanganate solution should be !rea in an amber colored or black
painted bottle.
ALKALIES 2259
One Hundredth Normal Potassium Permanganate (N/IOO KMnO.).- Dilute
exactly 100 ml. of the tenth normal potassium permanganate (N /10 KMnO.) solution to
one liter with distilled water. This solution should be kept in an amber colored or black
painted bottle.
One Hundredth Normal Sodium Thiosulfate (NjIOO Na2S20,).-This solution is
made by diluting exactly 100 mI. of tenth normal sodIUm thiosulfate (N /10 Na 2S 20,( to
1000 mI. with distilled water.
Tenth normal sodium thiosulfate is prepared as follows:-Dissolve 24.83 grams of
chemically pure sodium thiosulfate (Na 2S 20"5H 20) in distilled water and dilute to
one liter. Standardize as follows:-Place 40 mI. of tenth normal potassium perman-
ganate (N /10 KMilO.) solution in a 500 mI. flask, dilute to about 100 mI. with distilled
water and add 15 mI. of )1-3( sulfuric acid (H 2SO.), and 10 mI. of 10% potassium iodide
(Kr) solution (100 grams per liter). Titrate the liberated iodine with the tenth normal '
sodium thiosulfate solution until the color becomes a very light yellow. Then add about
0.5 mI. of fresh starch solution and complete the titration until the blue color just dis-
appears. Adjust the volume of the solution from the following data:
40 mI. N/lO KMnO.=40 mI. N/10 Na2B.O,.
This solution should be standardized once a month. .
Standard Iron Solution.-Dissolve 0.4911 grams of chemically pure, crystallized,
ferrous ammonium sulfate (FeSO.·(NH')2S0,·6H 20) in 50 mI. of distilled water to
whicll20 mI. of (1-3) sulfuric acid (H 2SO.) have been added. Warm the solution slightly
and add tenth normal potassium permanganate (N /10 KMnO.) until the iron is com-
pletely oxidized. Dilute the solution to 1000 mI.
1 mI. of this standard solution =0.00007 grams Fe or=O.OOOl grams Fe.Oa.
Methods of Analysis
Preparation of the Sample.-This operation should be carried out without unneces-
sary exposure to the air, as caustic soda readily takes up moisture and carbon dioxide,
thus lowering the tests. A part of the sample to be tested is wrapped in a heavy cloth
and crushed with a hammer, the remainder being saved for later determinations. Weigh-
ings should be made as quickly as possible. ...
Silica (Si0 2).-Dissolve 50 grams of the prepared sample in 200-300 ml. of distilled
water in a 6-inch porcelain or Jena glass evaporating dish, add 2 drops of methyl orange
and acidify cautiously with concentrated hydrochloric acid (HCl), avoiding loss by
spattering. Evaporate to complete dryness on the steam bath, cover the dish and
dehydrate the silica by heating in an oven at 110° C. for one hour. Cool, drench the
residue with 10 mI. of concentrated hydrochloric acid, taking particular care that each
particle of the residue .is moistened with the acid. Then add 300 mI. of distilled water
and warm on the steam bath to dissolve the soluble salts. Filter and wash the residue
free from chlorides with cold water, adding only the first two washings to the filtrate.
Transfer the filter paper and silica to a weighed platinum crucible, ignite in the open
crucible over a low flame until the filter paper is burned off and then over the blast lamp
for 10 minutes. Cool in a desiccator and weigh.
Weight of Si0 2X2=% Si02.
Iron and Aluminum Oxides (Fe.O,· AJ.Oa)-Dilute the filtrate from the determination
of silica to 500 ml. in a measuring flask and mix thoroughly. Place 250 ml. of this solu-
tion in a 400 ml. beaker, add 1 ml. of nitric acid (HNO,) and about 1 gram of ammonium
chloride (NH.Cl) and boil to oxidize the iron. Add slowly a slight excess of ammonium
hydroxide (NH.OH) and boil until the vapors have only a slight odor of ammonia.
Allow the precipitate to settle, filter and wash the precipitate free from chlorides with
hot water. Place the filter paper and precipitate in a weighed platinum crucible,
carefully burn off the filter paper in the open crucible over a low flame and finally ignite
with free access to the air for 10 minutes oVl\r the Meker burner or blast lamp. Cool in
a desiccator and weigh. •
Weight Fe 20, & Al 20, X 4 = % Fe 20, & Al 20,.
Iron Oxide (Fe20,).-Transfer a 100 ml. aliquot portion of the 500 mI. filtrate from
the determinations of silica to a 250-ml. beaker, add about 1 mI. of nitric acid (HNO,)
and boil to oxidize the iron. Then add a slight excess of ammonium hydroxide (NH.OH)
and continue boiling for a few minutes. Allow the precipitate to settle, filter and wash
the precipitate about 5 times with hot water. Dissolve the precipitate off the filter in
ALKALIES
about 20 mI. of (1-3) sulfuric acid (H 2SO.) and "Wash into a :100 ml. graduated flask.
Dilute to the 100 ml. mark and mix thoroughly. Pipette off· 25 ml. of this solution into
a 50 ml. Nessler tube, add 5 ml. of potassium sulfocyanate (KCNS) and dilute to the
50 ml. mark with distilled water. I
To a second 50-ml. Nessler tube, add 5 ml. of (1-3) sulfuric acid and 5 m!. of potassium
sulfocyanate, dilute nearly to tfie 50 ml. mark with distilled water and titrate with
standard iron solution until the colors in the two Nessler tubes match.
Ml. Standard Iron SolutionXO.004= % Fe20s.
Aluminum Oxide (AlzO s).-% Fe 20s & AI 20 s-% Fe 203= % AI 20 s.
Calcium Carbonate (CaCOs).-Adjust the volume of the filtrate from the determina-
tion of iron and aluminum to about 300 ml. and acidify with (1-1) hydrochloric acid
(HCI), adding about 5 m!. in excess. Add 10 m!. of ammonium oxalate solution,
- (NH')2C 20., heat to boiling, make slightly alkaline by addin~ ammonium hydroxide
(NH.OH) drop by drop and boil for a few minutes. Allow tlie mixture to stand in a
warm place for 1~2 hours, filter and wash the precipitate free from chlorides with small
quantities of hot water. Usually four or five washings suffice. Place the filter paper
containing the precipitate in the beaker in which the precipitation was made and add
50 ml. of distilled water and 15 ml. of (1-3) sulfuric acid (H 2S04). Heat to 70-80° C.
and titrate while hot under constant stirring with N /10 potassium permanganate
(KMnO.) solution until a faint, permanent pink color is obtained.
M!. N/10 KMnO.XO.02=% CaCO•.
Magnesium Carbonate (MgCO.).-Acidify the filtrate from the determination of cal-
cium with (1-1) hydrochloric acid (HCl), adding about 5 ml. in excess. Heat to boiling
and add 10 ml. of disodium phosphate solution (Na2HPO,·12H20). Then add am-
monium hydroxide (NH.OH) drop by drop under constant stirring until the solution
becomes alkaline and a precipitate begins to form. If no precipitate forms the stirring
should be continued .for at least 5 minutes. ,Cool, add 30 ml. of strong ammonium
hydroxide and stir for 5 minutes more if no precipitate forms. Allow the mixture to
stand several hours or over night in a cool place. Filter and wash the precipitate with
dilute ammonium hydroxide until the washings acidified with nitric acid (HNO.) give
no test fot chlorides. Place the filter paper with the precipitate in an inverted position
in a 250-ml. beaker and dry at 60-80° C. When quite dry and free from ammonia, add
25 ml. of distilled water and a measured excess of N /10 sulfuric acid (H 2S0 4), Stir
until the paper is thoroughly diSintegrated and titrate the excess of acid with N'/10
sodium hydroxide (NaOH), using 2 drops of methyl orange as inaicator.
M!. N/10 H~S04-ml. N/IO NaOHXO.0169 = % MgC03.
Manganese (Mn).-Dissolve 40-50 grams of the sample in 300 m!. of distilled water,
add 2 drops of methyl orange indicator, acidify with hydrochloric acid (HCI) and make
slightly alkaline with ammonium hydroxide (NH.OH). Pass hydrogen sulfide gas (H 2S)
into the solution for Y:i hour and allow the sample to stand at about 100° C. until the
precipitate settles. Filter the precipitated sulfides and wash a few times with a weak
solution of ammonium sulfide, (NH.).S. Dry and ignite in a platinum crucible. To
the residue in the crucible add 5--6 times its weight of sodium carbonate (Na 2CO.),
thoroughly mix and fuse over a burner. .
Dissolve the fusion in 10 oml. of concentrated nitric acid (HNO.f and heat for thirty
minutes on the steam bath until all of the nitrous oxide fumes are expelled. Wash into
a 250 ml. beaker with about 15 ml. of water, add 5 ml. of N /20 silver nitrate (AgNO.)
a~d warm to 70-80° C. Introduce 1 gram of ammonium persuIfate (NH.).S 20s, a little
at a time, and shake until a good color is developed. Cool, shake, transfer to a 50 m!.
Nessler tube and dilute to the 50 ml. mark. Should the color be too strong the solution
must be diluted to 100 ml. and an aliquot portion used.
Compare the color with N 1100 potassium permanganate (KMnO.) titrated into 50
m!. of cold (1-4) nitric acid which has been previously boiled for several minutes to expel
the nitrous oxide fumes. The (1-4) nitric acid is'free from nitrous fumes and ready for
use when it is colored by 1 drop of N /10() potassium permanganate.
Ml. N 1100 KMnO.X .011 o/t: M
Weight of Sample 0 n. I .
Copper (Cu).--,Dissolve 100 grams of sample in aboutr400 ml. of distilled wat~r in a
one liter beaker, and acidify slightly with (1-3) suifurio itc1q(H 2SO.). Bring the solution
to a boil and add 20 ml. of strong sodium thiosulfate (Na 2S 20.) solution. Continue the
boiling for Yz hour, filter, using suction and wash several times with hot water. Ignite
ALKALIES 2261
the filter paper in a porcelain crucible, dissolve the residue with 6 to 8 drops of con-
centrated nitric acid (HNO s) and carefully evaporate to dryness over a low flame until
all of the nitric acid is expelled. Dissolve the residue by heating gently with 10 ml.
of N /10 sulfuric acid (H 2S0 4), wash into a 250 ml. Erlenmeyer flask and add a little
sodium carbonate (NaiCO.) to precipitate basic copper carbonate. Acidify with acetic
acid (C.H 40.), cool and add from 2 to 5 grams of potassium iodide (KI) crystals. Titrate
. the liberated iodine with N /100 sodium thiosulfate (N a 2S2 0.) until the color becomes
light yellow. Then add about 0.5 m!. of freshly prepared starch solution and complete
the titration with N /100 sodium thiosulfate until the blue color disappears.
M!. N /100 Na 2S20. X .063 'Ii C .
Weight of Sample 0 opper..
PROCEDURE
Total Alkali, Na 20.-Five grams of the soda ash are dissolved in 50 ml. of
distilled water, preferably in an Erlenmeyer flask, with a funnel, the stem
extending in the neck of the flaSK. 95 ml. of normal sulfuric acid are added
and the solution boiled gently to expel the CO 2, the funnel prevet'lts loss during
boiling. After cooling, methyl orange indicator is added and the titration
completed. The end-point is a faint pink color.
I ml. NjH 2S0 4 =.03I gram Na 2 0.
NOTE.-"New York and Liverpool test" the value of 1 m!. N ·H2S0 4 =.032 g. Na20.
Sodium Bicarbonate, NaHC0 3.-Fiv:e grams of the soda ash are dissolved
in 100 ml. of water and the solution titrated with normal sodium hydroxide
until a drop of the solution on a spot plate produces an immediate dark color
a
with drop of silver nitrate.
I ml. N/NaOH=.084 gram NaHCO a•
Sodium Carbonate.-Deduct the ml. titration, of NaOH for NaHCO. from
the ml. H 2S0 4 titration for N a 2 0, the difference in ml. multiplied by .053
= gram N a 2 CO •.
Sodium Chloride.-Two grams of the ash by the- ammonia-soda process or
5 $!ams by the Leblanc process are dissolved in 50 to 100 ml. of distilled water
and 5 mL of colorless HNO. (sp.gr., 1.42) added. 2 ml. of ferric ammonium
sulfate are used as indicator,21 followed by a few drops of N /20 KCNS solution,
the exact amount of this b~ing noted. The chloride is now titrated with N /20
21 Ferric ammonium sulfate 6% solution made by dissolving the salt in 50 parts of
water to 6 parts of salt and adding an equal volume of colorless nitric acid.
2262 ALKALIES
AgN0 3 solution 22 until the color is just dest:royed, then 1 ml. in excess is added.
The precipitate is filtered off and washed. The filtra;te and washings are now
titrated with N /20 KCNS to a permanent pink color ,23
The total mL KCNS is deducted from the ml. AgNO a added and the differ-
ence multiplied by 0.002923 = gram N aCI. i
NOTE.-8ee Volhard's method for chloride in Vo!' I, page 271.
In addition to the above determin~tions the following may be desired:
Insoluble Matter.-Fifty grams of the soda ash are dissolved in about 500
m!. of water the insoluble matter allowed to settle, the clear solution decanted
through a double filter, which has been weighed, and finally the residue washed
onto the filter. The residue and filters are dried at 100° C. and weighed.
Weight minus tare of filters multiplied by 100 = per cent insoluble matter.
Iron, Alumina, Lime and Magnesia.-These are determined in the insoluble
residue by dissolving out by means of dilute hydrochloric acid. The residue
consists of sand and carbonaceous matter. Iron and alumina are precipitated out
together by addition of ammonia according to the standard procedur~, and
determined. as oxides, Fe 203 and AI 20 s• Iron may be determined in this
residue l5y dissolving in HCI and titrating with stannous chloride according to
the procedure given in Vol. I, page 373. Alumina is obtained by difference.
Lime and magnesia will be found in the filtrates from iron and alumina pre-
cipitates. Calcium is thrown out as an oxalate and magnesium as a phosphate
. and determined as usual. / .
Sodium Sulfate.-Dissolve 5 to 10 grams in dilute HCI and add BaCI 2 •
The precipitate BaS04 is washed and ignited as usual.
BaS04XO.6086 = gram Na 2S04.
Sodium Sulfite.-5 grams are dissolved in water and the solution acidified
with acetic acid. Starch solution is added and the sul~te titrated with N /10
iodine until the blue color appears.
I
1 ml. N/IO 1=0.006303 gram Na 2SOa.
Sodium Silicate.-Ten grams of tJ:e soda ash are treated with an excess of
HCI and the solution evaporated to dryness,. the silica dehydrated at 110° C.
Mter leaching with water the insoluble Si0 2 is determined by filtering off and
igniting by the standar~ procedure. .
Si0 2X 2.0282 = Na2SiO a•
-._
Sodium Sulfide.-This may be estimated by titration with an ammoniacal
solution of silver nitrate (13.81 g. silver per liter = 1 ml.=.005 g. Na 2S). Add
the reagent until no further precipitation occurs of Ag 2S. To get a good end-
point the solution is filtered just before this point is reached and the titration
completed. See also chapter on Sulfur, Volume 'I.
Loss on Ignition.-The sample 10 grams is ignited at a temperature slightly
below 300° C. The loss is due largely to water and organic matter.
22 N /20 silver nitrate contains 8.495 g. AgNO s per. liter , St'andardize against pure
NaC!. . __ - - .
23N /20 potassium thiocyanate, KeNS, contaIns 4.86 grams of the salt per liter.
The solution should be standardized against the silver ~itrate ~olution.
ALKALIES 2263
MODIFIED SODAS
CAUSTICIZED ASH
Total Alkali (Na 20).-Weigh accurately 5 grams df the sample and dissolve
in about. 100 ml. of distilled water in a 250 ml. beaker. Titrate with Nil H 2S0 4,
using 3 drops of methyl orange as indicator. \
(A-B) X 100 .
(1) 4 = % mOIsture.
(2)
(B-C) X2.7097X100 ~% NaHCO .
a
4 ,
25 By courtesy of the Diamond Alkali Compa?~t,biougl]. the kindness of J. D.
Mattern. -
ALKALIES 2265
Dissolve the contents of the platinum crucible in water and titrate with
normal HCI using methyl orange as indicator.
m1.X.053XlOO
4 01
10
N
az· O d as N a 2C 0 3 as f ound'm expreSSIOn
expresse . 2.
% Bicarbonate X84' .
. 53 = % Blcarbonate expressed as Na 2 CO g •
Five grams of the sample are dissolved in CO 2-free water, and the solution
made up to exactly 250 ml. in a measuring flask. Aliquot portions of 25 mI.,
2268 ALKALIES
SODIUM HYDROXIDE SOLUTION ,AT 15°
LUNGE
j
1.4
Na20.
4
-_
NaOH,
De,rees
Sp. Gr.
60°
.
Per Cent Degrees
Sp. Gr.
60°F. Per Cent Degrees
Sp. Gr .
60° Per Cent
Baume. _F. NH,. BaumEl. Baume. _F. NH..
6,° 60° NH3 • 60°
--- --- --- --- ------ - - - _-- - - -
10.00 1.0000 .00 16.50 .9556 11.18 23 00 .9150 23.52
10.25 .9982 .40 16.75 .9540 11.64 23.25 .9135 24.01
10.50 .9964 .80 17.00 .9524 12.10 23.50 .9121 24.50
10.75 .9947 1.21 17.25 .9508 12.56 23.75 .9106 24.99
11.00 .9929 1.62 17.50 .9492 13.02 24.00 .9091 25.48
11.25 .9912 2.04 17.75 .9475 13.49 24.25 .9076 25.97
11.50 .9894 2.46 18.00 .9·159 13.96 24.50 .9061 26.46
11. 75 .9876 2.88 18.25 .9444 14.43 24.75 .9047 26.95
12.00 .9859 3.30 18.50 .9428 14.90 25.00 .9032 27.44
12.25 .9842 3.73 18.75 .9412 15.37 25.25 .9018 27.93
12.50 .9825 4.16 19.00 .9396 15.84 25.50 .9003 28.42
12.75 .9807 4.59 19.25 .9380 16.32 25.75 .8989 28.91
'13.00 .9790 5.02 19.50 .9365 16.80 26.00 .8974 29.40
13.25 .9773 5.45 19.75 .9349 17.28 26.25 .8960 29.89
13.50 .9756 5.88 20.00 .9333 17 ..76 26.50 .8946 30.38
13.75 .9739. 6.31 20.25 .9318 18.24 26.75 .8931 30.87
14.00 .9722 6.74 20.50 .9302 18.72 27.00 .8917 31.36
14.25 .9705 7.17 20.75 .9287 19.20 27.25 8903 31.85
14.50 .9689 7.61 21.00 .9272 19.68 27.50 .8889 32.34
14.75 .9672 8.05 21.25 .9256 20.16 27.75 .8875 32.83
15.00 .9655 8.49 21.50 .9241 20.64 28.00 .8861 33.32
15.25 .9639 8.93 Ii 21. 75· .9226 21.12 28.25 .8847 33.81
15.50 .9622 9.38 .. 22.00 .9211 21.60 28.50 .8833 34.30
15.75 .. 9605 9.83 22.25 .9195 22.08 28.75 .8819 34 79
16.00 .9589 10.28 22.50 .9180 22.56 29.00 .8805 35.28
16.25 .9573 10.73 22.75 .9165 2304
equivalent to 0.5 gram, are taken for analysis. The sample taken is titrated
\vith N /5 hydrochloric acid in presence of phenolphthal~in indicator (2-3 drops).
The ml. titration recorded as " A," represents one half of the sodium carbonate·
present. Methyl orange indicator is now added (2-3 ~rcips) and the titration
with the acid continued until the solution turns faintly pink. The remaining
carbonate and all of the bicarbonate are now titrate<,l. The ml. titration is
recorded as B.
2AXO.0106=Na 2CO a and B-AXO.0168=NaHCO a•
The author desires to acknowledge his indebtedness to Dr. W. B. Hicks,
Chief of the Analytical Department, The Solvay Process Company, for his
review of the section on Alkalies.
Determination of Small Amounts of Acid in Ether.-L. P. Hall, Ind. Eng.
Chem., Anal. Ed., 2,244 (1930), recommends the use of sulfophthalein indicator
such as bromothymol blue for the titration. Care should be exercised against
CO 2 contamination. Color changes from green to blue.
masking by means of a green dye the faint pink color produced by calciu~
carbonate in water on phenolphthalein. The color of the dye does not interfere
in the titration of the alkalinity of calcium hydroxide. The cheapness of the
procedure as compared with the iodine method commends its use in the com-
merciallaboratory. - .
Solutions: Zinc Chloride Solution.-This is made by dissolving 25 grams of
the salt in a liter of water and filtering. 1 m!. of this solution is equivalent to
about 0.01 g. CaO.
Sodium Hydroxide Solution.-The solution is made of such strength that
1 ml. is equivalent to 1 m!. of the zinc chloride solution
Indicator.-5 grams of phenolphthalein are dissolved in 500 mt of 95%
alcohol. This is mixed with a solution containing 2 grams of Alkali Fast
Green E dissolved in 500 m!. of water.
Standardization.-T.he sodium hydroxide solution is standardized against
a standard acid. The zinc chloride solution is now standardized against the
alkali ll,s follows: Into a liter flask c~mtaining 500 rn!. of boiling hot distilled
water 25 to 50 ml. of the NaOH solution are accurately measured from a
burette .. About 0.5-1.0 m!. of the indicator is added and the zinc chloride is
run in until a green color is obtained that persists for half a minute. 10 ml.
excess of the ZnCl 2 solution are now added and this excess, after vigorous
shaking, is titrated with the NaOH solution until a lavender (not pink) color
is obtained.
Determination.-In a dry one-liter flask is placed 1 gram of the sample.
500 ml. of boiling, CO 2-free water are added. As soon as the lime has hydrated,
the solution is titrated with zinc' chloride solution until a green color appears,
an excess of zinc chloride is added and the excess determined with standard
NaOH solution as in case of the standardization above. The end-point is a
lavender color.
NOTEs.-The flask used with the lime determination must be dry, otherwise the
water in small amount reacting with the sample will cause difficulty in the subsequent
~~ ,
The quantity of indicator in standardization and the determination must be the
sam!!, as the quantity influences results. The Alkali Fast Green E dye is made by the
National Aniline Company, Buffalo, N. Y. .
In the case of solids weigh out 0.5 to 2.0 g. and make up with carbon
dioxide free water to 100 ml. Weigh out a second sample for the bicarbonate
determination.
In the case of liquid samples the dilution of 10 ml.j to 100 ml. is convenient.
The bicarbonate determination requires an undiluted sample; It is convenient
to use 2 or 5 mI. I .
MEANING OF pH
Intensity of heat is expressed as degrees on the thermometer scale, and
density as degrees on the Baume scale. Similarly the pH scale is used to denote
intensity of acidity and alkalinity. '"
The numbers from 0 to' 14 are used to express pH values, that-is,jntensity of
acidity and alkalinity. The value pH 7.0, halfway between 0 and 14, is the
neutral point; that i&, a solution having a pH of 7.0 is neither acid nor alkaline.
The numbers below 7.0 denote acidity, intensity of acidity increasing as the
numbers decrease. Thus a solution of pH 6.4 is wiry slightly acid, one of pH 6,0
is more intensely acid and one of 4,6 is still more intensely acid, etc.
On the other hand 'the numbers between 7.0 and 14.0 are used to denote
alkalinity, the intensity of alkalinity increasing as the numbers increase. Thus
a solution of pH 7.2 is very slightly alkaline, one of pH 8.6 is more intensely,
alkaline and one of 10.4 is still more intensely alkaline. These numbers repre-
sent absolutely definite degrees of acidity and alkalinity and therefore. a state-
_J-
d·IspensedWIt
· h· 1 = 10-3, etc.
, I.e. 10 !
3
Since in pure water the H ion and OH ion concentrations must be the same,
if [H+] [OH-] = 10-14 then [H+] = [OH-] = 10-7 • Asolution in which [H+) = lOH-)
= 10-7 is neutral. Whenever [H+] is geater than [OH-] that is greatf'r than
10-7 at 25° C., the solution is acid; whenever [H+] is smaller than [OH-] that is
smaller than 10-7 at 25° C. the solution is alkaline. From the equation.
[H+] [OH-] = Kw it is seen that a simple relation always exists between [H+] ~
and [OH-], thus [H+] = [~~-J. It i~ therefore unnecessary to refer to OH ion
concentration at all and the term hydrogen ion concentration is used even when
referring to very alkaline solutions such as N sodium hydroxide, which contain
a vanishingly small number of hydrogen ions, the predominating ion being the
hydroxyl i;n'":""m
Expre~drogen ion concentrations as 1/10,000,000 or even as 10-7 is
extremely inconvenient. For this and other reasons, such 'as plotting curves,
etc., Sorensen introduced the method of expressing such values in terms of the
negative logarithm to the base 10. He called this the hydrogen exponent and
gave it the symbol pH. Thus pH is the negative logarithm of the hydrogen
ion concentration or the logarithm of the reciprocal of the hydrogen ion con-
centration, i.e. pH = -log [H+] = log [~+]' and [H+] = lO-pH • Accordingly if in
a neutral solution [H+] = 10- 7, the pH will be 7.0.
Similarly pOH is the loga-
1
rithm of the .reciprocal of the hydroxyl ion concentration, pOH = log [OH-]
Several concrete examples may serve to'make clear the exact meaning of pH.
Although a N /10 solution of hydrochloric acid contains 3.65 g. of HCI and 0.1 g.
of ionizable hydrogen per liter, only 84.0% (18° C.) is ionized (active). Recent
developments in the theory of ionization indicate that strong acids are com-
,pletely ionized and that the degree of ionization ordinarily assumed is only
apparent being due to interionic forces. Therefore the normality with respect
to ionized hydrogen is not N /10 but O.~~N or 11~0' Then pH = log 11.90 = 1.08.
. A N /10 solution of acetic acid contains the same amount of ionizable hydro-
gen as does N /10 HCI namely 0.1 g. but only 1.3% (18° C.) is ionized. The
normarIty WIt
. h respect to IOnIZe
. . d h y d rogen IS
. on Iy 0.013
-1-0- N or 769
N an d pH IS·
. Iog
769=2.89.
The apparent degree of dissociation of N /10 NaOH is 92.9% at 25° C. The
. h respect to OH"IOns IS t h erefore 1
· WIt
norma1Ity 0:929
0 N = 10.76
N an d p OH'IS Iog
10.76 or 1.03. pH=14-1.03=12.97.
Dissociation in 1~ ammonium hydroxide proceeds only to the extent of
1.33% at 25° C. Normality with respect to OH ions is therefore only
0.0133 N N
10 or 752 and pOH=2.88. pH=14-2.88=11.12.
When certain salts are dissolved in water not only does ionization occur but
there is a tendency to react with the water as BA+HOH~HA+BOH. With
salts of strong acids and strong bases such as sodium chloride this tendency is
zero and such salts do not affect the pH of water. With salts of strong acids
and weak bases such as aluminum sulfate, ferric chloride, etc. the strong acid
2278 DETERMINATION OF.;pH
resulting dominates the weak base and the solution is acid. The reaction for
aluminum sulfate is Ab(S04)a+3H 2 0!::::j;3H 2S04+ AI(OH)a. With salts of weak
acids and strong bases, the strong base resulting dominates the weak acid
and the solution is' alkaline. The reaction for sodium carbonate is N a 2 CO a
+2H 20!::::j;H 2CO a+2NaOH. This general reaction is termed hydrolysis and
salts subject to it are for practical purposes either ac~ds or bases.
BUFFER ACTION
unionized acetic acid and [H+] increases only. slightly. If a base (sodium
hydroxide) is added H ions are neutralized but more Ij-cetic acid is ionized
producing additional H ions to take the pla.ce of those neutr~lized and [H+J
decreases only slightly. Of course any buffer system has its limits and large
amounts of acid or base will overcome the buffer capacity. Each buffer
system has a rather narrow pH zone over which it functions. However, by
using different systems i.e. different mixtures, practiMlly the entire pH range
can be covered. This is extremely advantageous since it is possible to prepare
solutions of constant pH which serve as standards and it also serves as a means
of holding an industrial solution close to the d'ptimum pH. In the case of
fairly concentrated solutions of strong acids and bases, since the concentra,tiQn
of unionized substance is relatively small, there is no reserve supply of H or
OH ions: However there is such a large concentration of H or OH ions that
for practical purposes such systems function as buffer systems. For example
.when the concentration of either N /10 HCI or N /10 NaOH is reduced to NjlOO,
the pH is changed only about 1 unit.
A
J
II
I V I B
10
9
7 [I
a11 7
___.--I--- V
.-- ~V
/
v [7 ?
8 / (~ ,
Jr J/:
114 12
10 ,l I
ACID r I clc i r &
ALKIALr
i t 2 4
FIG. 311.
The stability of any buffer system will depend on total concentration, and
the ratio of salt to weak acid or base. Reasonable changes in concentration will
not materially change the pH as long as the ratio remains the same. In the
acetate system the ratio [~l:l does not change on moderate dilution and the
change in pH when the system N /10 acetic acid-N /10 sodium acetate is diluted
twenty fold is less than 0.1 pH. Many solutions met with in practice are
buffered. This is an advantage since colorimetric pH measurements can be
made on such solutions even when highly colored or turbid by diluting with
distilled water. The degree of dilution po~sible depends on the concentration
and the specific buffer system but the safest procedure is to dilute only as far
as necessary t? make sati~factory readings_;] . .
Buffer actIon can be Illustrated graphIcally by means of tItratIOn curves
(Fig. 311). Three such curves are shown in which various amounts of M/5
.
HCI and M/5 NaOH are added to 100 m!. samples of (A) distilled water, (B)
,
2280 DETERMINATION OF. pH
a buffer mixture composed of KH 2P0 4 and Na 2 HP()4 in which the total con-
centration is approximately 0.1 M, and (C) a similar buffer mixture in which
the total concentration is one half of B,all samples having an initial pH of 7.0.
The abscissa show the amounts of acid or alkali addea and the ordinates show
the pH of the respective mixtures. I
The extremely steep curve (A) for distilled water shows that this substance is
devoid of buffer action since small amounts of acid or alkali change the pH
markedly. In the pronouncedly acid and alkaline ranges the curve flattens
out because one is then dealjng with relatively concentrated solutions of the
strong acid and base, where pH varies only moderately with concentration.
The curve (B) for the 0.1 M phosphate mixture differs radically from (A).
Addition of 11 mI. of acid only changes the :pH from 7.0 to 5.8, whereas th~
addition of 6 mI. of alkali only changes the pH from 7.0 to 7.8. Beyond these
points the change becomes more pronounced. Between 5.8 and 7.8 the curve
is relatively flat and in tliis zone the phosphate mixture shows strong buffer
action. As the curve indicates, buffer action is st.rongest near the middle of
the relatively fiat portion. It actually reaches a maximum at about 6.8 where
. d H . [l'>ta2 H P0 4]
the concentratlOns of KH 2P0 4 an Na2 P0 4 are the same, I.e. [KH P0 ] = 1.
2 4
The steepness of the curve above and below the zone 5.8-7.8 ·irrdicates little
resistance to change in pH.
Curve (Q) resembles (B) in contour but the relatively fiat portion or zone of
effective buffer action is much shorter, extending only from about 6.5 to 7.5.
This shows the effect of concentration on buffer action.
These curves show that not all buffer mixtures have the same buffer capacity.
Van Slyke (J. BioI. Chern., 52,525 (1922)) .has formulated a method of express-
jng the differences in terms of buffer capacity. A solution has a buffer capacity
of 1 when the'"reaction of a liter of it is changed one pH unit upon the addition of
one equivahmt of acid or base. A knowledge of the buffer capacity of a solution
is frequently of considerable value since it indicates ~ow easily the optimum pH
may be maintained in practice, and also what pfecautions are necessary in
determining the pH. It also provides a practical method of determining the
amount of acid or alkali required tb adjust the pH.
The requirements for buffer action should be clearly kept in mind. Mere
presence of salts does not ensure buffer action. Salts of ~trong acids and strong
bases impart no buffer action to a solution. Solutions ·which. contain only
salts of weak acids and weak bases may not be buffered. To ensure buffer
action a buffer system must be present i.e. both salt and weak acid or weak base
and the pH of the solution must be within the effective zone of the particular
II
system.
Tables showing the method of preparation fO.i" various buffer mixtures can
be found in Clark; The Determination of Hydro'gen Ions, 1928, Williams and
Wilkins, Baltimore, Md.
The following table gives some common buffer systems and the approximate
pH of maximum buffer capacity. The zone of effective buffer action will vary
with concentration but· the general average wi!l:.be ±1.0 pH from the value
given, for concentrations approximately 0.1 molar .
•
COLORIMETRIC 2281
Glycocoll-sodium chloride-hydrochloric acid. . . . . . . . . . . . . . . .. 2.0
Potassium acid phthalate-hydrochloric acid. . . . . . . . . . . . . . . . .. 2.8
Primary potassium citra,te ... " ......... '" . . . . . . . . . . . . . . .. 3.7
Acetic acid-sodium acetate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4.6
Potassium Add Phthalate-Sodium Hydroxide. . . . . . . . . . . . . . .. 5.0
Secondary Sodium Citrate.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5.0
Carbonic Acid-Bicarbonate ...................... :. . . . . . . . . 6.5
Primary Phosphate-Secondary Phosphate. . . . . . . . . . . . . . . . . .. 6.8
Primary Phosphate-Sodium Hydroxide. '.' . . . . . . . . . . . . . . . . . .. 6.8
Boric Acid-Borax.. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . .. 8.5
Borax................ ...... .............................. 9.2
Boric Acid-Sodium Hydroxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 9.2
Bicarbonate-Carbonate .................................... 10.2
Secondary Phosphate-Sodium Hydroxide. . . . . . . . . . . . . . . . . . .. 11.5
1: Name
Acid cresol red ......................... .
Acid meta cresol purple ................. .
pH Range
0.2- 1.8
1.2- 2.8
Color Change
Red-yellow
Red-yellow
Thymol blue ........................... . 1.2- 2.8 Red-yellow
Benzo yellow .......................... . 2.4- 4.0 Red-yellow
Bromphenol blue ....................... . 3.0- 4.6 Yellow-blue
Bromcresol green .........._............. . 3.8- 5.4 Yellow-blue
Methyl red ............................ . 4.4- 6.0 Red-yellow
Chlorphenol red ........................ . 5.2- 6.8 Yellow-red
Bromcresol purple ...................... . 5.2- 6.8 Yellow-purple
Bromthymol blue ...................... . 6.0- 7.6 Yellow-blue
Phenol red ............................. . 6.8- 8.4 Yellow-red
Cresol red ............................. . 7.2- 8.8 Yellow-red
Meta cresol purple ..................... . 7.6- 9.2 Yellow-purple
Thymol blue .............. , ............ . 8.0- 9.6 Yellow-blue
Phthalein red .......................... . 8.6-10.2 Yellow-red
Tolyl red .............................. . 10.0-11.6 Red-yellow
Parazo orange ......................... . 11.0-12.6 Yellow-orange
Acyl blue ...........................•.... 12.0-13.6 Red-blue
•
2282 DETERMINATION OF pH
Indicators are weak acids or bases in which. the ionized form differs in consti-
tution and color from the normal form: An indicator acid will dissociate as
HIn±=}H++In-, HIn being the acid form and In- the alkaline form. The mass
.
Iaw expreSSIOn '11 b [H+] [In-] K Th [In-] . tKr S· K'
WI e [HIn] = I· en [HIn] = [H+]' . mce I IS a con-
stant the ratio i~~~] will depend on the value of [H+]; the hydrogen i(;m eon-
cent ration. This ratio determines the color that the indicator will show. For
example with phenol red the acid form [HIn] is yellow and the alkaline form
[In-] is red, consequently as [In-] increases the color changes from yellow
through orange to red. When HIn is half transformed [HIn] = [In-] and so
[H+j = K I. When the hydrogen ion concentration is the same as the ionization"
constant of the indicator, the concentrations of the acid and alkaline forms of
the indicator are equal. TP.e value of Kr varies for different indicators and
determines the range of hydrogen ion concentration over which each indicator
1
functions. If [H+]=KI then pH=pKI, where pKI=log KI' While this de-
termines the theoretical midpoint of the pH range of each indicator,_ the actual
range used in practice depends on how readil~he colors can be distinguished
by the eye. Consequently the practical range is determined experimentally.
While pKI is_a function of pH, it is also affected to some extent by other factors,
such as salt concentration, proteins and temperature.
A satisfactory indicator for practical pH work must have certain well defined
qualifications. It must give well-defined color changes over a relatively short
range; it must not be unduly affected by substances other than hydrogen or
hydroxyl ions; it must immediately give colors that are stable for a reasonable
length of time, so that there will be no errors due to changes taking place during
the determination; its solution must be a stable reagent. An indicator that
fades rapidly, precipitates, undergoes change in color due to the presence of
salts, colloidal materials, etc., attains a stable color slowly, or is unstable as a
reagent, is certain to prove objectionable in practice. I
The essential requirements in applying the colorimetric method for the
determination of pH are (1) An accurately prepared solution of indicator,
(2) a set of color standards prepared from measured amounts of buffer solutions
having accurately adjusted pH values and containing accurately measured
amounts of the indicator solution. Standards in intervals "Of .0.2 pH are usually
employed. Such a set for phenol red will consist of standards with the pH
values 6.8,7.0,7.2,7.4,7.6,7.8,8.0,8.2 and 8.4. These standards. can be placed
in test tubes or ampoules having uniform bores. Ampoules are more satis-
factory since they can be sealed. If a measured volume of clear sample is
treated with the same amouilt of indicator as is present in the standards, the.
resulting color will depend on the pH of the sample and by comparing this
color with the color standards the pH can be determined. It is essential that
all measurements be accurate and that the bore of the color standards and test
tube used for the sample be practically the same. The intensity of color and
depth of liquid in both cases should be uniform. l
If the materials being tested are absolutely cle.a:r:.-and colorless accurate deter-
minations can be lJlade by comparing the color of tne test sample with those of
COLORIMETRIC 2:.l83
instead of a single wide range indicator, such as the so c,alled universal or utility
indicators. "With the single test wide range indicator there is no way of check-
ing results and marked errors may be made with no means of detec~ing them.
\ \
COLORIMETRIC DETERMINATION OF BUFFERED MATERIALS
Some materials encountered in practice are not very highly buffered and a
few are unbuffered. In such cases care must be exerci,sed in determining the pH
since such solutions are very susceptible to change. Among such materials are
most waters, soil extracts, paper extracts, white 'water, pure sugar liquors, flota-
tion feeds, laundry rinses, clay filtrates, etc.
Distilled water is devoid of buffer action and special precautions must be
observed in testing it. Absolutely pure water has pH 7.0....at 25° C. Ordinary
distilled water is usually 'acid, due to absorption of carbon -dioxide. Distilled
water in equilibrium'with the carbon dioxide of the air will have pH 5.7. It
may be supersaturated with carbon dioxide to give a pH as low as 5.0. Carbon
dioxide may be removed by boiling in Pyrex ve~sels to give water with a pH of
6.6 to 6.8 but superpure water with pH 7.0 can best be secured by means of a
special still. Distilled water should always be kept in Pyrex or tip.ned vessels
and contact with air held at a minimum. In making aqueous solutions or
extractions of slightly buffered materials, proportions of material and water
should be held constant so that results on different samples will be comparable.
In determining the pH of buffered solutions reasonable differences between
the pH of the sample and the indicator solution. will not' affect the results.
_.-
2 McCrumb, Sewage Works .:!ournal, 1, 534 (1929). -
COLORIMETRIC 2285
Effect of Geiatille
'.
It will be noted that the differences encountered near the isoelectric point
are insignificant being within the sensitivity of the colorimetric method em-
ployed. When hydrochloric acid is used the, differences are greater in the acid
range.
The variations between pH measurements by the hydrogen electrode and the
colorimetric method in the precipitation of casein by hydrochloric acid have
been determined by Clark and others 9 using methyl red; and by Miss Benton 10
using bromcresol green.
_EFFECT OF TEMPERATURE
. EFFECT OF CHLORtNE I
The action of chlorine must sometimes be taken intb consideration in water
treatment plants, paper mills, textile mills, etc., whenevE;)r" the dosage of chlorine
is high enough to affect the indicators. When such action occurs the results
secured are influenced by the particular indicators in use, ~he concentration of
indicator employed, the buffer capacity of the solution being tested, the tsm-
perature and, of course, the concentration of available chlorine.
In the procedure employed with the Slide Comparator (Fig. 313) the danger
point with the sulfonephthaleins is about 2.0 parts per million of active chlorini"
Whenever lower concentrations of indicator are employed as in Nessler tubes,
etc., the action of chlorine may cause errors at lower concentrations of chlorine.
There appears to be two types of reaction, chlorination apd oxidation. In
the case of chlorphenol red and phenol red chlorination appears to predominate
at first and the chlorinated dyes which result have lower pH ranges than the
original dyes. This results in an increase in the alkaline color which leads. to
results that are higher than they should be. Of course on longer standing if
the concentration of chlorine is high enough fading will occur due to destruction
of the dye by oxidation. In the case of the other sulfone phthalein dyes the '
·action is largely destructive and the reagings tend to be low or off colors are
secured.
The nitro phenols are more stable in the presence of active chlorine and can
sometimes be used when the concentration of chlorine is as high as 200 parts per
million. 14
One disadvantage of the nitro phenols should be pointed out here. They
are all fairly acidic and their solutions are difficult to stabilize at any definite pH.
Consequently they are more liable to give erroneous results in testing unbuf-
fered samples than are the sulfonephthaleins . .
I
2. To the central tube 0.5 ml. of 'the indicator solution is added, by means of
the pipette and nipple, and the contents are thoroughly 'mixed.
3. The slide containing the color standards is now placed in position on the
base and, holding the instrument toward a window or other source of daylight,
the slide is moved back and forth in front of the test samples until a color match
is obtained. Tlie pH is then read off directly from the yalues on the front of
the slide.
Since all color standard slides are interchangeable, that is theyl can be used
with one ~ase, the range of a single slide Comparator can be increased by pur-
chasing additional slides and a supply of the corresponding indicator solutions,
with pipettes and nipples. For, ranges longer than that of one indicator,
however, the Long Range and Dalite Slide Comparators are recommended.,
All color standards provided for the various models of slide comparators are
guaranteed to retain their accuracy for a period of five years.
THE LONG RANGE SLIDE COMPARATOR
This set is designed to meet the needs of workers who require a portaple
set covering a wider range. It is made in six models covering the range of any
3, 4, 5, 6, 7, or 8 indicators re-
spectively between the limits pH
0.2 and pH 13.6. It is flexible
since its range can be increased
simply by. securing additional
color standard slides.
Each Long Range Slide Com-
parator contains one complete
Slide Comparator; 2 to 7 extra
color standard slides; vials of
the corresponding indicator so-
lutions, ;rith 0.5 ml. pipettes and
nipples;) and 14-5 m!. test tubes.
All models are contained in a pol-
ished wooden case 11;\1 inches
long, 9incheshigh and 5%, inches
wide, with a handle for carrying.
FIG,317.
I
A ~
leqUlpm.en
l' t necessary f or
making tests is contained in the case so that tltese sets can' be-used in field,
laboratory or plant. The total weight of the set varies from 7}'2 lbs. to 9;\1
Ibs. depending on the number of slides.
The method for making determinations is exactly the same as that described
for the single Slide Comparator.
DALITE SLIDE COMPARATdR
FIG. 318.
The pH equipment supplied with this set consists of one complete Slide
Comparator; any 2 t o 7 additional color standard slides; vials of the corre-
sponding indicator solutions with pipettes and nipples; and 10-5 ml. t est tubes.
Each slide contains 9 color standards for any given indicator and these slides
fit in slots in the front of the case as shown in the illustrll,tion. Each vial of
indicator solution is placed directly above the corresponding color standard
slide, thus eliminating error due to the use of the wrong indicator solutions.
A top protects the vials when the set is not in use.
The comparator base is supported by a shelf on the front of the box. Di-
rectly behind the slots in the base is a piece of Dalite glass and in the compart-
ment behind this is a special 25 watt electric light bulb, wit h cord, switch and
socket. In the lower part of the box is a cupboard of sufficient size to hold four
1000 ml. Pyrex bottles. This cupboard serves not only as storage for stock
bottles of indicator solution but also to bring the comparator to a convenient
height for making readings. A rack on the inside of the door holds 10-5 ml.
t est tubes. A test tube brush is also supplied with the set.
In making determinations three of the 5 ml. test tubes are filled to the mark
with the solution to be tested, the proper color standard slide is placed on the
base and 0.5 ml. of the corresponding indicator solut ion is added to the middle
tube by means of the pipette and nipple. The light is then turned on and the
,
2294 DETERMINATION OF pH
slide is moved back and forth until a match is obtained with one of the color,
standards. The pH value is then read off directlyfrOIp. the values on the slide.
FIG. 320.
with indicator, the other to receive the test solution without indicator. This
arrangement provides for color comparison with coloredlor turbid liquids as
well as clear solutions. ~
. In addition to the prism and cells the comparator is sup,l)lied with four round
measuring tubes with graduations from 5 to 10 ml. in 1/1 ml. for ,measuring the
test solution, 1 pipette with graduations 0.20,0.25,0.5 and ;t.O ml. for measuring
the indicator solution, 1 opal glass plate, 50 ml. of indicator solution and 1
color disc.
Each color disc consists of a molded resin circular frame with a numbe.r
(usually 9) of colored glass standards each one representing 0.2 pH so arranged
that they can be revolved in the comparator. The color glass standards are
made from variable layers of colored glass held together to simulate the"
colors resulting when buffer solutions of known pH are treated with the re- '
spective ·indicator. The value of each standard is placed on the disc so that
it is visible in a special opening when that particular standard is in the reading
position. Only one color standard is visible at a time. Separate color discs
are provided for the various indicators.
An illuminator with stand can also be secured to provide the proper artificial
illumination. •
ELECTRO METRIC HYDROGEN
ION MEASUREMENTSl
£Electrometric hydrogen ion (pH) measurements are made with a suitable
electrode system as the detecting unit and a potentiometer as the measuring
ul).it. The electrodes develop a potential (e.m~f.) proportional to the pH of
the solution being measured. The potentiometer measures e.m.f.
POTENTIOMETRIC PRINCIPLE
ELECTRODE SYSTEMS I
A single electrode is not sufficient for pH measurements beca.use there is no
convenient method for measuring its potential, but the potentIal of tw.o elec-
t.rodes, the absolute potential of one of which is known and is unaffected by pH,
can be compared. It is therefore customary to, use a ~~ed electrode as a
reference, and to use as a measuring electrode otteJ)est suited to the measure-
ment at hand.
ELECTRO METRIC 2299'
For {h9se who are not familiar with the theory relating e.mJ. of electrodes
and pH, the rea~er is referred to the treatise on hydrogen-ion measurements by
, Clark, "The Determination of Hydrogen Ions," Williams & Wilkins.
Reference Electrodes: Calomel Electrode.-The calomel electrode is almost
universally used as a reference electrode, due to its constancy of potential and
ease of preparation. The calomel electrode is one in which mercury and calomel
are in contact with a definite concentration of potassium chloride. These are
contained in a vessel which may vary in design accord-
ing to conditions of use. A common type of calomel
electrode is shown in Fig. 322; a more recent design is
shown in Fig. 323. The latter has the distinct advantage
that the electrode, immersed in the test solution, is at
solution temperature.
The potential of the mercury with respect to that of
the solution depends on the concentration of mercurous
ions from the calomel, but this is governed by that of
chlorine ~ns, which are produced in the solution by
potassium}hloride as well as by calomel. The propor-
tion of the total number of chlorine ions that is due to FIG. 322.
the potassium chloride varies with the KCl concentration.
The voltage of the calomel electrode therefore depends on the concentration of
potassium chloride in the electrode solution. Saturated, normal and tenth
normal solutions are used.
The saturated calomel electrode is used more generally than
either of the others. l.t is more easily prepared, and has some par-
ticular advantages.
Speci5l1 precaution must be taken to prevent diffusion between the
saturated solution of a salt bridge with the less concentrated solu-
tions of the normal and tenth normal electrodes, whereas with the
saturated electrode there is no tendency toward diffusion. When
diffusion is prevented the potentials of the normal and tenth nor-
mal electrodes are fairly stable at room temperatures, but at higher
temperatures 'it is diffi.cult to prevent diffusion, with its resulting
error in potential. The saturated electrode shows no change in
potential between 5° and 60° C., other than that due to temperature.
The temperature coefficient is approximately .8 millivolt per degree
C. Another advantage of the saturated calomel electrode is that its
FIG. 323. conductivity is high, which makes for increased sensitivity.
An objection to the saturated electrode is that the solution may creep over
the edge of the containing vessel with the resulting formation of masses of
crystals. This is avoided in the design shown in Fig. 323, if the electrode is im-
mersed in a container of distilled water when not in use.
For accurate and reliable potentials, extremely pure chemicals mWlt be
used in calomel electrodes. Mercury specially purified and calomel produced
by electrolysis for the purpose can be commercially obtained in appropriate
quantities, and their use is recommended. Chemically pure potassium chloride
dissolved in distilled water is a suitable solution for use with them.
The saturated electrode is easily prepared. For a saturated KCI solution
put the salt into warm distilled water in a beaker and stir it 'thoroughly with
2300 HYDROGEN ION
repeated addition until no more of it will dissolve. Then stir in enough calomel
to saturate the solution. A small quantity is sufficient as it is not very soluble.
When the solution is cooled to room temperature, it sho~ld contain undissolved
potassium chloride and calomel. Prepare enough solut~on to fill a reservoir for
flushing the electrode from time to time. Add suffidient mercury (amount
depends upon construction of electrode) to make contact with the platinum
terminal, and sufficient calomel (HgCI) to assure saturation, an excess is
advisable. Moisten with saturated KCl solution, shake vigorously, and then
nearly fill with saturated KCI solution. If the electrode is not of the self-
flushing type, care must be exercised in flushing the salt bridge occasionally.
Silver Chloride Electrode.-Comparable in principle to the calomel electrode
is the silver-silver chloride electrode. It has largely been used in studies on
activities of chlorides in solution, but recently has received increased attention
as a reference in pH measurements. This electrode is not as readily prepared
as is the calomel electrode, but it is highly reproducible, stable over extended
periods of time, and has a small temperature coefficient.
The potential of the silver-silver chloride electrode depends upon the method
of preparing the silver and the chloride. Methods have been described by
Brown,! Harned,2 Mclnnes,3 Noyes and others.4
Measuring Electrodes.-There are a number of electrodes which
have been used as pH measuring electrodes. Those which have
been best established and mQst largely used will be described.
Hydrogen Electrode.-The hydrogen electrode consists of a
noble metal coated with platinum black, immersed in the solu-
tion and under a definite partial pressure of hydrogen gas. Such
an electrode is illustrated in Fig. 324. Means are provided fer
introducing the hydrogen gas, and allowing the gas to escape wjth-
out forcing the solution entirely away from the metal.
Electrodes with large surfaces are generally not advisable as
they are more difficult to saturate with hydrogen and therefore
come to equilibrium less quickly. .'
In work with ordinary chemicalrsolutfons, and when high accu-
racy is not essential, the solution being measured may be contained
FIG. 324. in an open vessel, but the electrode potential will be slightly in-
fluenced by oxygen absorbed from the air. For precise work the
vessel should be stoppered with provisions for carrying the electrodes through
the stopper, and for allOWing the gas to escape. '"' .. _
An electrode is said to be in equilibrium when its potential remains constant
if the solution does not vary. The potential of the hydrogen electrode is not
constant until the platinized surface is saturated with hydrogen, and any
reducing action due to hydrQgen is completed. '
Limitations.-In some solutions, notably biological fluids, the platinum
black may become clogged with substances which p~event absorptibn of hydro-
gen. The hydrogen electrode cannot be used in solutions containing strong
oxidants like ferric ion, dichromates, nitric acid, peroxides and chlorine, nor in
1 J. Am. Chern. Soc., 56,646 (1934).
2 J. Am. Chern. Soc·., 55, 4849 (1933).
3 J. Am. Chern. Soc.; 37, 1445 (1915).
4 J. Am. Chern. Soc., 39, 2532 (1917).
i?1~£
ELECTRO METRIC 2301
those containing strong reductants, such as sulfurous acid (S02) and hydrogen
sulfide. Some organic compounds are very susceptible to reduction, particu-
larly those of the aromatic series, such as azo benzene, aniline dyes and some
unsaturated acids. In a solution containing either oxidizing or reducing com-
pounds, a noticeable drift in the voltage occurs and a balance is difficult to
obtain. The oxidizing or reducing action must be completed before the voltage
is steady. The measurement is not trustworthy, for the pH value of the solution
may be seriously changed in such reactions.
The hydrogen electrode cannot be successfully applied to solutions contain-
ing metal i~s--that fall below hydrogen in the electromotive series of metals
and, also lead, all of which are reduced on the platinum electrode.
P!atinizing the Electrodes.-The film of platinum black is deposited by
electrolysis in 3% platinic chloride solution containing 1/40% lead acetate.
The electrode should be thoroughly cleaned before platinizing. Using two
dry cells in series it usually requires from 30 seconds to one minute to deposit
the platinum black. Experience will determine the amount of plating neces-
. sary.. The electrodes should be washed with distilled water, and immersed in
distilled water when not in use. The electrodes should not be allowed to, dry
thoroughly.
Hydrogen Gas.-A very convenient source of hydrogen is the compressed
gas supplied commercially in metal cylinders. A cylinder containing 100 or 200
cubic feet is commonly used in the laboratory, while small cylinders are available
for portable use.
Compressed hydrogen produced by electrolysis may be sufficiently pure to
. use directly from the cylinder for measurements of ordinary accuracy, but it is
common practice to pass it through solutions of potassium hydroxide and
pyrogallol and then through water to remove traces of oxygen and other im-
purities that would affect the potential of the hydrogen electrode.
Quinhydrone Electrode.-Some of the difficulties encountered in the use of
the hydrogen electrode are avoided by using a quinhydrone electrode. The
basis of this electrode is a piece of platinum or gold the same as used for the
hydrogen electrode, but the surface of the metal is not platinized and it is not
supplied with gaseous hydrogen. Instead, a small quantity of quinhydrone
(benzoquinhydrone) is dissolved in the solution, and in certain conditions the
electrode in the solution acquires a potential that is definitely related to the
hydrogen ion concentration of the solution. The potential is measured against
t.hat of a calomel electrode, and the pH value is found from the measured
voltage in the same manner as with a hydrogen electrode. .
Biilmann 5 was the first to describe the quinhydrone electrode for the pH
measurements. Morgan, Lammert and Campbell 6 have published a series of
papers covering detailed studies of this electrode. The quinhydrone electrode
is quickly prepared, develops its potential rapidly and is not readily poisoned.
Rather than describe the method of preparing quinhydrone, it' 'is suggested
that this material be obtained from chemical manufacturers or suppliers of
high grade chemicals.
Although the quinhydrone electrode can be applied successfully to many
solutions containing oxidizing and reducing substances in which the hydrogen
6 Ann. Chim. Phys., 15, 109 (1921).
6 J. Am. Chern. Soc., 53, 2154 (1931) and 54, 910 (1932).
2302 HYDROGEN ION
electrode would give erroneous results, still it is not applicable in all cases .
. Measurements have been made witll it on dilute nitric :acid, solutions, unsatu-
rated organic acids, and on a variety of oxidizing systems that are of too low
an oxidizing intensity to disturb the electrode potentiall
Limitations of the Quinhydrone Electrode.-The qJinhydrone electrode is
not applicable over the entire pH range. In the acid range, 0 to 7 pH, it is
excellent. In alkaline solutions strongly buffered, results of fair accuracy can
be obtained to about 9 pH. In solutions poorly buffered, the upper limit is
approximately 8.0 pH. In the latter case it is essential not to add an excess ~f
quinhydrone, because it reacts with the alkali and changes the pH value of the
solution. About 6 to 8 drops of a saturated solution of quinhydrone in acetone
is suffici~nt for 50 ml. of test solution. ' .
If the pH value of a solution measured with a hydrogen electrode differs
from that with a quinhydrone electrode, the difference may be due to the" salt
error" in the quinhydrone value. The hydrogen electrode is assumed to have
no salt error. The magnitude and sign of the salt error (positive or negative
divergence from the correct pH value) of the quinhydrone electrode depep.ds
upon the type' and concentrations of the salts. present in the solution. The
salt error in nickel plating solutions of total salt concentration about 2.0 N is
below +0.05 pH. In a 10.Q N ammonium sulfate solution it is approximately
-0.2 pH, while in a 4.0 N sodium chloride solution it is approximately +0.2
pH. Since most solutions are far more dilute than these, the salt error is
generally negligible. /'
In high purity (99%) sugar solutions the quinhydrone electrode gives
values which are approximately .2 pH low, while on low purity products, such
as molasses, the values are high. Over purity ranges of 80% to 97% the
quinhydrone electrode gives good results.
Antimony Electrode.-Of the various metal and metal oxide electrodes that
have been described, the antimony electrode hM proved to be the most satis-
factory. It does not require gas or a catalytic surface as does the hydrogen
electrode; it does not require the addition of a reagent as does the quinhydrone
electrode. The antimony electrode is not readily prepared, but once made up
is useable for very extended periods: It may be obtained commercially ready
for use. It is a rugged electrode and in certain designs is used industrially for
continuous plant service. The inherent accuracy of the antimony electrode is
not as great as is the hydrogen or the quinhydrone electrode.
Its limitations have been described by Perley.7 Notable. among these is
that the e.m.f:-pH relation varies with the degree of buffering of the solution,
and that the degree of saturation with air, pltrticularly below 7 pH, affects the
potential-pH relationship. The antimony electrode is not applicable to solu-
tions which are oxidizing or are strongly reducing, nor in solutions containing
more than a trace of copper, silver, or other metals below antimony in the
electrochemical series of the elements. However, uX;_der known conilitions this
electrode serves a very useful purpose in measuring pH over' the .range of 3 to
12 pH. It is being used quite extensively in sugar solutions (beet, raw cane,
and refinery), paper mills, water treatment, pigments, etc. For measurements
on a given sample extending over long periods of time, this electrode is particu-
larly suitable.
7 Trans. Am. lnst. Chern. Engrs., 29 (1933).
ELECTRO METRIC 2303
, Glass Elecirode.-The work of Hughes,8 Kerridge,9 McInnes and Dole,IO. 11
and others have shown the possibilities of the glass electrode as a pH measuring
electrode. The glass electrode consists, primarily, of a thin membrane of glass,
of a definite composition (72% Si0 2, 22% N a 2 0, 6% CaO), one side of the
membrane being exposed to the solution being measured and the other side
exposed to a standard reference solution. The membrane is made up in the
form of a bulb, with a reference electrode and solution on the inside. The bulb,
along with a reference electrode, is immsered in the solution to be tested. The
glass electrode, which behaves like a hydrogen electrode, is thus interposed
between two fixed electrodes.
A high sensitivit7 galvanometer, or an electrometer potentiometer, ·is
required to me~sure pH with the glass electrode due to its high resistance.
Glass electrodes and suitable measuring devices are commercially obtainable.
The glass electrode is particularly suitable for pH measurements in the
presence of oxidizing and reducing solutions, where other types of electrodes
give erroneous results. It also is the best type of electrode to use in unbuffered
solutions, since the introduction of the glass bulb into an unbuffered solution
does not change the pH of the latter. The electrode is stable over rather long
• periods, the life of the electrodes being largely determined by crystallization
of the glass. The characteristics of the internal reference element and solution
determine how often they must be replaced.
The glass electrode, in some forms, is rather fragile and must be handled
with care. In other forms, with heavier glass wall, the electrode often exhibits
an asymmetry potential; that is, it differs from the Nernst equation by more or
less constant potential. The E.M.F.-pH relationship of the glass electrode is
linear up to pH 9.6, above which it may be used with a special calibration to
12.5 pH. The greatest deviation from linearity is obtained in solutions con-
taining sodium ion concentrations in excess of 0.1 molal.
The electrode can be made in small sizes, permitting of measurements on
minute samples. It has a.further advantage in that it does not contaminate the _ .
solution being measured, should only limited samples be available, and it is
necessary to make other determinations on the same material.
1. Hydrogen-Saturated Calomel.
H- V-0.2648+.00076t
p - 0.0001983T '
at 25° C.,
H= V-.2458.
p .' .0591
2. Quinhydrome-Saturated Calomel.
H_ Y+0.4529+ .00002t
p - 0.00019832T '
8 J. Am. Chern. Soc., 44, 2860 (1922).
9 Biochern. J., 19,611 (1935).
10 Ind. Eng. Chern. Anal. Ed., 1, 57 (1929).
11 J. Am. Chern. Soc., 53, 3315 (1931).
2304 HYDROGEN ION
at 25° C.;
+
H':_ V .4534_.
P - .0591
3. Glass-Saturated Calomel.-When using quinhyd~one-l pH hydrochloric
acid as reference in glass electrode, use formula 'as applied to (2) Quinhydrone
above, up to 9.6 pH. Relation is non-linear above 9.6 pH in the· presence of
sodium ions. '
4. Antimony-Saturated Calomel.
Buffered solutions at 25° C.
Below pH 7, H= V+.0157,
p .0591
V -.0292
Above pH 7, pH= .0527 .
E r = E 0 +0.000198
n T Iog [I on.]
For silver, for example, E Ag =0.798+0.0591 log [Ag+] (at 25° C.). The con-
vention regarding the sign of the electromotive force here used is the opposite
1 N. Howell Furman, Professor of Chemistry, Princeton University, Princeton, N. J.
2305
2306 POTENTIOMETRIC METHODS
, J
of that given by Nernst; we use the sign of the metal 'relative to the solution,
which is the convention adopted by foreign and American electrochemical
societies. . • ,
This electrode potential expression is not limited to substances which can
be formed into solid electrodes. For example, platinized pl(1tinum catalyzes
the reversible exchange of hydro.gen atoms to hydrog~n ions, and a hydrQgen
electrode may be formed by causlhg hydrogen gas to stream through a solution
,into which a lightly platinized platinum, gold, or iridium electrode is dipped.
At 1 atmosphere pressure of hydrogen gas:
EH = 0.0000+0.0591 log [H+] (at 25° C.).
The Eo value of the hydrogen electrode is arbitrarily chosen as the standard of
reference, and is set equal to zero. The Eo value of any other substance is then
found from the measured value of E (with refer~nce to the hydrogen standard)
when the concentration of the ion in question is equal to 1 molar.
If we are dealing with a negative ion, the sign of the second term in the
eleQtrode expression is negative. For the chlorine electrode: ECJ=Eo-0.0591
Jog [CI-] (at 25° C.).
The Eo value is characteristic for each substance and is a measure of the
relative oxidizing or reducing power of the substance relative to hydrogen.
It should be noted that Nernst derived the electrode potential formula on
the basis of ideally dilute solutions, and ~hat the electrod~ potential is really a
function of the activity rather than the concentration of the ion in question.·
Hydrogen ion concentration measurements are standardized at present on
E.M.F. meas!lrements, or in other words on an empirical activity basis, since
the point of reference of the whole system is based on hydrogen-ion concentra-
tion values calculated from conductance measurements. Kolthoff gives the
relation between the exponents of the activity and ion concentration values as:
paH=pH+0.04. 2
As far as the potentiometric determination of ind-points of volumetric
reactions is concerned, activity relations need npt be taken into account.
Oxidation-Reduction Electrodes.-A burnished platinum electrode may,
when dipped into a solution containing an oxidizing agent and its reduction
product, assume a potential which is a measure of the reversible oxidation-
reduction power of the solution. Peters 3 found that for"*he reversible f~mic
ferrous ion system, the relations were expressed by the formula~, _..
[Fe+++] °
E=Eo+0.0591 log [Fe++] ~at 25 C.).
In this case the Eo value is equal to the observed value when [Fe+++] = [FeH ] = 1
(the normal oxidation-reduction potential). .'
In general, for a reversible oxidation-reduction equilibriu~:
Ox+n~Red,
in which Ox represents oxidizing agent, Red its redl!Qtion pr:6duct, F the Fara-
21. M. Kolthoff, .Rec. Trav. chim., 49, 401 (1930), ...1-:- _
3 R. Peters, Z. physik. Chern., 26, 193 (1898).
~
POTENTIOMETRIC METHODS 2307
day, and n the number of Faradays per gram ion of oxidizing agent, We have
if the electrode process is. reversible: '
E =Eo+0.05911
n
[Ox] ( t 25° C )
og [Red] ~ ..
Ke(
..
p:tt~±.=:.::::I'''.4-.. H9~ C~
,--...!.-_ _;r-H~
a.
1. ~.
FIG. 325.-Calomel Electrode Vessels. (1) Modified Ostwald Type. The mercury
is overlain with calomel intimately mixed with mercury; the potassium chloride solution
is introduced through tube, T. The side tube at a ends in a capillary, or a porous plug
to prevent diffusion. Contact with the mercury is made by a platinum wire sealed
through a glass tube, and joined at b to a copper wire. (2) Represents the bottle type
electrode recommended by E. Muller. The siphon tube, H, is made of two glass
T-tubes, joined by rubber tubes, r. The bottom of each tube is plugged at c with agar
jelly made up in potassium chloride solution of the strength needed for the electrode
(0.1 N, N, etc.).
2308 POTENTIOMETRIC METHODS
Other electrodes that are frequently employed are the silver electrode in
contact with silver chloride and a potassium chloride solution of definite con-
centration. When the presence of chloride is undesirable a salt bridge of
sulfate or nitrate may be interposed between the reference electrode and the
solution. A mercury-mercurous sulfate~sulfuric acid electrode is convenient"
for reference purpos,ts in the study of oxidation-reduction processes in the
absence of chloride.-
POTENTIOMETRIC TITRATION'S
.Jf the E.M.F. of a ceil composed of two electrodes dipped ~nto an unknown
s01ution is measured during a titration, the plot of E.M.F. readings as ordinates
against ml. of standard solution as abscissas gives in general an abrupt change
in potential at the analytical end-point of the reactio.n at the end-point. One
electrode serves as an indicator of changes in ion concentration, and the other
electrode is a calomel, or other reference electrode. The indicator electrode
must be appropriate for the particular type of reaction involved. The solution
must be well stirred during the titration, preferably by a motor-driven stirrer.
If the standard solution is added in small equal incremen4l. near the end-j)oint,
say 0.05 ml. at a time, the end-point is in general marked by a maximum change
of E.M.F. per small increment of reagent. That is, the plot- of A E.M.F.
" Am!.
against ml. of reagent gives a maximum at the end-point. In favorable cases
it is not even necessary to record the readings near the end-point, although this
is advisable until some experience has been had witp a particular type of titra-
tion. The outstanding advantages of the potentiometric method are (1)
applicability to colored solutiOIis; (2) precision of determination of end-point;
. (3) the possibility of obtaining a succession of end-points, and hence the deter-
mination of two or more substances by a single titration~ I
A type of assembly which is suitable for gene!].l· study' of potentiometric
titration processes is shown in Fig. 326. If it is~ot desired to determine the
POTEN'PIOMETRIC METHODS 2309
exact voltage range for a given process, the potentiometer may be replaced by a
simple slide wire potentiometer, and the use of the standard cell may be omitted.
Some still simpler forms of apparatus are described on p. 2318. One of the
types of titration cell whicJ;t may be used when air must be excIudci~ showl'l. in
.Fig. 327. The vessel shown in Fig. 328 may be used for hydr6~electrode
titrations if an opening is provided in the stopper for a burette.
N.E.
s.
---~-------I
I
I
0.1 I
I
I
I
I
I
I
I
I 1.4 I
11.31.2" .9 I
I ___________________
L Pot.. 1.0 I ~
acid, ions of metals like Ag, eu which are precipitated, easily reducible organic
compounds, the alkaloids, certain dyestuffs. I
Quinhydrone Electrode. 5-May be used in the titration of acid solutions
which are free from strong oxidizing and reducing, agfnts. May be used !Ol:
alcoholic solutions. .I
FIG. 327.-Titration Vessel. A, inlet for inert gas, also used as electrode holder;
B, reductor, or other form of special apparatus used in introducing sample; C, motor-
driven stirrer; D, burette; E, tube leading to reference electrode; H is a sheltered plati-
num electrode to be inserted in place of E, when the differential method is used. The
stopper, G, is bored to give any desired number of openingS, through which glass tubes
are placed. The rubber tubes at F join the various pieces of apparatus to the glass
tubes carried by the stopper. The gas escapes around the shaft of the stirrer, C.
(Furman and Schoonover, J. Am. Chern. Soc., 53, 2561 (1931)).
I
B,
(.
A
FIG. 328.-Cell for pH Measurement. Pure hydrogen enters a bubbler, B I , at a.
BI is filled with some of the solution whose pH is measured; the latter is put in vessel A,
and is saturated with hydrogen. The gas comes in contact with the lightly platinized
wire or foil b, which is of platinum, and is sealed into a glass tube, where contact is
made with the copper wire, E. The hydrogen escapes through the bubbler, B 2• The
tube, T, is filled with potassium chloride solution, and leads to the calomel electrode;
c is a porous plug, or capillary to prevent diffusion.
The reference electrode may be a calom,el half cell, )Jith a suitable salt bridge
interposed if the presence of chloride is harmful. Theoretically any sufficiently
pH pH
12 Titration Curves Indicator 12
Intervals
11 11
10 10
9 9
8 8
Neutral Point
7 ----------
5 5
4 4
3 3
2 2
1
5 10 15 20 25 30 35 40 o 25 50 75 100
ml Alkali~ % Alkaline Form-~
--
+0.4
+0.3 ,
~€lRt.
"
+0.2
...
+0.1 1\ ,
•
g +0.0 t-- Aec/
...:
~
I -.;;;.::.:,
E-0.1 I
u.j
-0.2 i
I
\
I
I
-0.3
-0.4
\ , 1\
-0.5
5 10 15 20 25 30 35 40 45 50
ml AgN0 3
FIG. 330.-Titration of Mixture of Chloride and Iodide with Silver Nitrate. 25 cc.
of approx. O.l.N KI plus 25 CC. approx. 0.1 N KCI titrated with silver nitrate. The
dotted curve shows the graph obtained when iodide alone is present.
remains in the oxidizing region; as soon ai'! the first, excess of ferro cyanide is
present there is a sharp drop in the E.M.F. of the titra,tion cell.
The titration of zinc with ferrocyanide has been !ltudied by a number of
investigators, and many conflicting statements and conclusions occur.i5 The
solution should be only slightly acid; the titration is.rri6re rapid at 65° than at
room temperature, but there is danger of overstepping the end-point even at the
higher temperature. Kolthoff ~nd Verzijl found that
o C. in the titration of 100 ml. of 0.1 N zinc sulfate the
error was as follows:
Temperature Error in Per Cent of ml. of
O C.
o Substance Added Reagent Required
15 ...... (Neutral sol'n) -1.04
65 .... " (Neutral sol'n) -0.9
15 ...... 0.12 N sulfuric acid -0.77
65 ... ~" 0.12 N H 2S0 4 , 3 g. K 2SO. -0.4
65... . .. 5 g. K 2SO. -0.45
15 ...... 10 g. ammonium. chloride -2.25
65 .... " 10 g. ammonium chloride -0.8
15 ... '" 20 g. ammonium sulfate -0.05
65. . . . .. 20 gI ammonium sulfate -1.15
From Potentiometric, Titrations, 2nd Ed., Kolt~off and
Furman. -
b
Large amounts oj calcium and magnesium ca~e
errors; addition of'"fluoride prevents the interference of
a. ferric ion. Manganese must be removed before the
titration.
FIG. 331. - Oxi- With potassium ferrocyanide the precipitate is
dation-Reduction In-
dicator Electrode. a K2Zna(Fe(CN)6)2, whereas with other alkali metal ferro-
represents about 5" of cyanides the precipitate has a different composition.
No. 28 B. and S. From all of the considerations which have been presented
platinum wire sealed it is evident that good results will be obtained only
through a glass tube
and fused at b to a under carefully standardized experimental conditions.
copper wire, which Cadmium and lead :r;nay be detm-mined by ferro-
leads to the binding cyanide precipitation, but silver and ,copper give erratic
post c. The rubber results. is .
tube, r, supports the
electrode on the bake- Oxidation - Reduction Reactions. - The indicator
lite plate, d, which electrode is of burnished' platinum, either wire, foil or
may be drilled to carry gauze. Fig. 331 represents the conStl'uctio,n of one satis-
a number of electrodes factory type of electrode. A calomel or - a mercury-
of the construction
shown. mercurous sulfate and sulfuric acid reference electrode
may be used.
The reagent may be any sufficiently stable oxidizing or reducing agent,
and each reagent may be use'd to determine one or more substances in a single
(
]5 Knauth, Dissertation, Dresden, 1915; von Bischowsky, Ind. Eng. Chern., 9, 668
(1917); Hedrich, Dissertation, Dresden, 1919; Kolthoff, Rec. Trav. Chim., 41, 425
(1923); Kolthoff and Verzijl, ibid., 43, 380 (1924); Fr. Miiller, Z. anorg. allgem. Chern.,
128, 126 (1923); E. Miiller and Gabler, Z. analyt. Chern., 62, 29 (1923); Treadwell and
Chervet, Helv. Chim. Acta, 5, 633 (1922); 6, 550, 559 (1923). I
16 See Kolthoff and Furman, "Potentiometric Titra~ions," pp. 331-333, 2nd Ed.,
1931, for determination of cerium, lanthanum, gallium;and indium by ferrocyanide
precipitation. -
.
POTENTIOMETRIC METHODS 2315 .
titration. In Fig. 332 is shown the determination of iron, vanadium, and ura-
nium by a single titration of the reduced solution with permanganate. The
uses of potassium permanganate, bichromate, iodate, bromate or ceric sulfate
have been intensively studied, and the potentiometric method' offers the ad-
vantages of accuracy, speed, and usefulness in highly colored solutions. The
• powerful reducing reagents, chroIllous, titanous, or stannous chloride, arsenious
acid, and ferrous sulfate have decided advantages in certain fields of industrial
analysis. We shall describe only those methods which are useful in steel
analysis.
,
V
+0.5
+0.4
V
+0.3
r~ ~VV
~ +0.2
E
L1.i
/ -- II
Fe ---; Felli
+0.1
1
0.0
...J--/
;'-
VI_!! VIV
-0.1
UJ.:', UVI
/~ - 'a I b I~ -c -4
-0.2
2 4 6 8 10 12 14 16 18 20 22 24 26
MI KMn0 4 -
FIG. 332.-8irnultaneous Determination of Reducing Agents. From data of Miiller
and Flath, Z. Elektro_chern., 29, 500 (1923). The cc. of perrnanganate given by (a-c)
correspond to oxidation of DIV to UVl; b cc. are equivalent to the iron, and c ce. to the
vanadium.
solution is diluted to 300 ml. and oxidized with persulfate in the presence of
silver nitrate, etc., exactly as described in the preceding section.
If no vanadium is present, the latter two procedures give chromium alofle.
If both chromium and vanadium are present, the vana~ium alone is determined •
by the bromate oxidation procedure, and hence the amount of chromium can
be determined by difference after the determination of the sum of chromium
and vanadium by the appropriate one of the latter probedures.
20 Cox, J. Am. Chern. Soc., 47, 2138 (1925); MacInnes and Jones, ibid" 48, 2831
(1926); MacInnes and Cowperthwaite, ibid., 53, 555 (1931); MacInnes and Dole, ibid"
51, 1119 (1929); Roth, Z. Elektrochem., 33, 127 (1927); Heckzo, Z. analyt. Chern., 73,
404 (1928); 74, 289 (1928); 75, 183 (1928); Kamienski, Bull. Acad. ·Polonaise des Sci.
et. Lettres, p. 33 (1928); Millier and Kogert, Z. physik. Chern., 136,448 (1928); Hall,
Jensen and Backstrom, J. Am. Chern, Soc" 50, 2217 (l92_8); Cltrke and Wooten, J.
Phys. Chern., 33, 1468 (1929); Shukoff and Gortikoff) Z;-Elektrochem., 35, 853 (1929);
Rabinovich, Z. Elektrochem., 34, 311 (1928). -
,f.;:
POTENTIOMETRIC METHODS 2319
latter (Fig. 334, (2)) made possible the titration of 0.001 N acetic acid with
equivalent baryta solution with a mean deviation of ±0.8%, using quinhydrone
()i.. 1, 2
FIG. 334.-Differential Apparatus. (1) MacInnes and Dole's apparatus. Inert
gas passes through G, and operates the lift pump,. L. The bottom hole of tube A is of
capillary dimensions; opening, H, is a safety outlet which functions when the pump
operates too rapidly. (2) Apparatus of Clarke and Wooten. C is a sheltered electrode.
When tap A is closed, the solution about the sheltered electrode may be pumped into
the main solution by means of the rubber bulb, B. '
oj-
FIG. 335.-A. MUller's Capillary Reference Electrode. So little liquid remains in
the capillary that no correction need be applied in ordinary work. B. Burette electrode
of Willard and Boldyreff. .
2320 POTENTIOMETRIC METHODS
electrodes. MacInnes and Dole have shown that great precision (±0.003%)
may be obtained in the differential method as applied to argentometric
titration. \
Closely allied to the simplified differential method 3;1"e the capillary electrode
of Muller,21 Fig. 335A, and the burette electrode, Fig. 335B, of Willard and
Boldyreff. 22 These devices do not, however, give qifferential readings, I but
rather the cust'omary form of titration graph which has a maximum jump in
potential at the end-point. •
Bimetallic Electrodes. (a) Dissimilar Electrodes.-A number of studies
have been made of the use of pairs of metallic electrodes in neutralizations. 23 A
large number of possible electrode pairs have been studied; the most satis-
factory, according to Holt and Kahlenberg are tungsten-natural graphite,
antimony-natural graphite, and tungsten-silver. These pairs were applieq in
the titration of 0.1 N acids with 0.1 N bases.
The application of dissimilar metallic electrodes in the field of oxidation-
reduction reactions is largely the work of Willard and Fenwick. 24 They found
that a system composed of platinum, or platinum-rhodium, with tungsten gave
excellent results. There is a sharp break in the E.M.F. VB. ml. curve at the
end-point. After th,e break the two electrodes may stay at different potentials,
or come together, depending upon the nature of the reaction. Furman and
Wilson 25 found that the platinum-tungsten ,system if joined in series with a high
resistance and a galvanometer, gave a maximum galvanometer deflection at
, the end-points of many oxidation-reduction reactions (Fig. 336).
Polarized Systems.-If two platinum electrodes, as nearly as possible
identical, are polarized with a very small currel'lt, then at the end-points of
Pt.
TREADWELL (1925)
I-
FIG. 338.-Thermionic Vacuum Tube Circuits. (1) Goode's circuit. Rs serves to
balance out the residual plate current, and resistance R2 serves as a shunt for the gal--
vanometer. (2) Treadwell circuit. A small voltage is put in series with the titration
cell to operate on a favorable portion of the plate current-grid voltage line. The indi-
cating instrument is ~ millivoltmeter of the kind used in the:mocouple work.
end of the titration cell is attached to the gdd of the radio tube. A sensitive
milliammeter, millivoltmeter, or galvanometer in the plate circuit then gives
readings that are proportional to the E.M.F. of the titration cell, if the condi-
tions are such that the plate current is a linear functiqn of the grid voltage. 28
Diagrams of two simple circuits are shown in Fig. 338.1 The circuit described
by Treadwell (Fig. 338 (2)), is easy to construct and ~perate. For industrial
use at intervals over long periods of time it is qesirable to have a rugged switch-
board type of milliammeter, range 0 to 0.5 milliamp. as indicating instrument,
and a "B" battery of small lead storage units (cf. Ehrhardt, loco cit.). Many
commercial vacuum-tube voltmeters are now available. They have the ad-
vantage of compact construction, since batteries are eliminated by the use of a
"power-pack" which enables one to 'plug the instrument into the A.C.lighting
circuit and make continuous-reading titrations. 29
Micro-methods.-The potentiometric method offers certain advantages in
micro determinations, particularly in oxidation-reduction methods, and in 'cer-
tain precipitation processes. 30 •
28 Goode, J. Am. Chern. Soc.;.r44, 26 (1922); 47, 2483 (1925); J. Opt. Sci. Rev. Sci.
lnsts., 17,59 (1928); Calhane and Cushing, Ind. Eng. Chern., 15, 1118 (1923); Treadwell,
Helv. Chim. Acta, 8, 89 (1925); Bienfait, Rec. Trav. Chim., 45, 166 (1926); Ehrhardt,
"Chern. Fabrik," pp. 443, 455,' 463 (1929); Gelbach and Compton, Ind. Eng. Chern.,
Analyt. Ed., 2, 396 (1930). •
29 Some typical circuits 'and further literature references are given by: Partridge,
J. Am. Chern. Soc., 51, 1 (1929); Fr. MUller, :Trans. Electrochem. Soc., 62, 335. (1932);
Compton and Haring, ibid., 62, 195 (1932); Willard and Hager, In,d. Eng. Chern., Anal.
Ed., 8, 144 (1936); Garman and Droz, ibid., 1,341 (1935). .
30 Nussberger, Dissertation, Zurich, 1924, p. 40; Benedetti-Pichler, Z. analyt. Chern.,
73, 200 (1928); Nieder! and Muller, J. Am. Chern. Soc., 51, 1356 (1929); Zintl and Betz,
Z. analyt. Chern., 74, 330 (1928).
CONDUCTOMETRIC METHODS!
1. THE PRINCIPLES OF CONDUCTOMETRIC TITRATIONS
Electrometric titrations can be divided into two groups: potentiometric
methods, discussed in a previous chapter, and conductometric procepures.
The theory underlying. these two groups is el!tirely different. The theory of
potentiometric titrations as has been pointed out is very similar to that 6f
ordinary titrations, insofar as the change of the potential of a suitable electrode
is a linear function of the change of the logarithm of the ion concentration or
the logarithm of the ratio of concentration of oxidant to reductant in the system
to be titrated. A large jump in potential at the equivalence point means a
sharp color change of a suitable indicator; and the titration to a definite poten-
tial is comparable with an ordinary titration in which the reagent is added until
the indicator has assumed a definite color. Therefore, the electrode in a
Rotentiometric titration can be more or less compared to a specific indicator for
the ion or redox system to be titrated:
In conductometric work, on the other hand, all ions present contribute to
the electrical conductivity of the solution. If an electrolyte is added tq a
solution of another electrolyte without changing the volume to any appreciable
extent, the conductivity increases insofar as the electrolytes do not react with
each other. If an ion of one electrolyte unites with an ion of the other to form
a slightly dissociated or slightly soluble substance or if it changes the total ion
concentration by an oxidation or reduction process, then the conductivity of
the solution may change in three different ways, before the equivalence point
has been reached: (1) The conductance decreases. (2) The conductance
remains unchanged. (3) The conductance increases.
The conductance of various ions for the electric current is different. It is
usually expressed by the mobility of the ion. The equivalent conductivity
A of an electrolyte BA is equal to the sum of the mobilities of both ions:
ABA=AB++AA-.
The equivalent conductance A is the conductance in reciprocal ohms of a solu-
tion containing one gram equivalent of solute when plac~d between electrodes
which are one cm. apart; hence, it is equafto the specific conductance divided
by the concentration, the latter being expressed in equivalents per m!.
K
A=C1000 .
with increasing electrolyte content of the solution. The values reach a maxi-
mum at infinite dilution. In dilute solutions the \following relation holds
between the equivalent conductance Ac at a concenttation c and A",:
I
FIG. 341.
and iron wire kept in a small box to protect it from atmospheric disturbances.
If a current passes through the bridge a thermo current is developed and the
2 G. Jander and O. Pfundt, "Die visuelle Leitfahigkeitstitration," 2d Ed., Stuttgart
1934, Verlag Enke, also in "Physikalische Methoden der Aualglischen Chemie," Akad.
Verlagsges. Leipzig, 1936, pp. 1-54.
2328 CONDUCTOMETRIC METHODS
,
galvanometer G gives a deflection. In order to prevent the direct current from
passing through the conductivity cell, the transforme.'r, T, is placed in the cir-
cuit. R is a resistance by means of which
the sensitivity of ~he system can be reg-
ulated. W is a \known resistance and
AB a slide wire u~ed as in the ordinary
method. The apparatus can be used in
different ways; many details have\ to be con-
sidered, for which the reader is ~eferred to
the work of Jander and Pfundt. Simpler
to manipulate is the system in Fig. 342
s where U represents the rectifier. It. may
be mentioned that T. Callan and S. Hor-
VJ
robin 3 a few years ago described a similar
·of the dilution. Extremely dilute solutions of strong acids or strong bases, of
the order of 0.0001 N, can be titrated with the same accuracy as more con-
centrated solutions if care is taken to exclude carbon dioxide.
On the whole, the application of the
conductometric titration in this case has o
only practical significance when the solu-
tion has a dark color so tha·t the indi-
cator metho{1 fails to give results.
Intermediate Weak Acids and Bases.-
The shape of the neutralization curve
depends upon the concentration and the
ionization constant of the acid or base.
By neutralization of the dissociated part
of the acid, the. conductivity will drop
and on account of the formation of' its
salt which behaves as a strong electrolyte,
0.1 N ~ ml NaOH
the cpnductance increases. The practical
neutralization curve is obtained by ad- FIG. 344.
ding the figures of the "acid-depression
curve" to those of the "salt line." The more strongly the acid is ionized,
the more its neutralization curve approaches that of a strong acid and the
less it is ionized the more it behaves like an extremely weak acid; the latter
case will be discussed later.
In Fig. 344 the neutralization
curves are given for the neutraliza-
tion of 0.1 N, 0.001 Nand 0.0001 N
'acetic acid, respectively. From the
above it is evident that in manf'
cases a flat minimum will occur in
the neutralization curve. Its loca-
tion can be calculated from the
ionization constant, the concentra-
b tion and the mobilities of the ions
~o present in the system. The mini-
"C
'C mum itself has no analytical signifi-
U
o cance though it may give us an
indication of the magnitude of the
_ _ ml N~OH
ionization constant of an acid which
may be of importance in unknown
cases. If the acid is relatively
FIG.
highly ionized, the neutralization
ABE-Neutralization Curve. line will give a curve up to the
ED-8odium Hydroxide Curve
OEC-Salt Line. equivalence point and it is hard to
AF-Acid Depression Curve. find· with any degree of accuracy the
point of intersection of the neutrali-
zation curve with the sodium hydroxide line, the latter being the straight line
found after addition of an excess of base.
Such a case occurs, for example, in the neutralization of 0.01 N salicylic acid
as is' shown in Fig. 345. An analysis of the curve shows that the equivalence
2330 CONDUCTOMETRIC METHODS
.
J
point is found at the point of intersection of the salt, line and sodium hydroxide
line. The former can be determined in an experiment!l] way. Suppose 100 mI.,
0.01 N salicylic acid are titrated witli 0.5 N sodium hydroxide so that the sodium
hydroxide line can be drawn. In order to find the sha~e of the salt line a second
determination is made if? which 100 ml. of water instead of dilute acid is taken
'and successive amounts of 0.5 N sodium salicylate of exactly the same strength
as the sodium hydroxide are added. From the latter data, the salt line can be
constructed and the equivalence point can be found with an accuracy of at least
1%. .
If the dissociation of the acid during the neutralization is less than in the
above case, the last part of the neutralization curve, before the equivalence
point will be a straight line. Under these conditions, the point of intersection
can be found by a single titration ..•This is the case, if the ionization constant
of the acid is smaller than about 5 »)10- 4 in the titration of 0.1 N solutions,
smaller than 5 X 10-· for 0.01 N solutIOns and smaller than 5X10-6 for 0.001 N
solutions.
Very Weak Acids and Bases.-The ini~ial, conductivity is very small. and
during the neutralization, the conductivity i~creases according to the salt line.
On account of the hydrolysis of the salt formed, the experimental data near the
equivalence point are higher than the con'esponding points on the salt and
sodium hydroxide line; therefore, in the construction of the two lines, points
should be taken at .such a distance from/the equivalence point (at the hydrol-
FIG. 346.
ysis is negligibly small and the points can then be combined to give straight
lines. If the acid is extremely weak (e.g. hydrogen peroxide) the hydrolysis is
so large that the titration does not yield usefui results. For 'the titration
of 0.1 N solutions, the ionization constant should be larger than about 10-1°,
for 0.01 N solutions larger than 10-9 , for 0.001 N solutions larger than 10-8 •
These figures have no exact significance; they only giv~ the order of magnitude
of the ionization constant at which useful results can still be obtained.
From the above we see that the conductoIDet.ric titration can be applied
where the potentiometric or ordinary method ~Qes'not give satisfactm-y.results.
CONDUCTOMETRIC METHODS 2331
In Fig. 346 neutralization curves of 0.1 Nand 0.01 N boric acid with sodium
hydroxide are given. Several applications can be made. Hydroxy benzenes,
like phenol and resorcinol can be titrated very nicely. A systematic study
has shown that resorcinol and hydroquinol behave like dibasic acids, pyro-
catechol on the other hand as a monobasic acid. The trivalent phenols,
pyrogallol and phlorglucinol beha;ve like dibasic acids. The method has been
applied to the titration of nitro phenols, phenolphthalein and other weak acids
whose salts are colored. Phenolphthalein in .50% alcohol behaves like a
dibasic acid, a break occurring after the neutralization of the carboxyl and
phenol groups. The conductometric method should be very useful to organic
chemists in cases where they want to know quantitatively the acid or basic,
character of a colored substance. It can also be recommended for the deter-
'mination of the equivalent weight of amino acids and polypeptides and the acid
or base combining power of proteins. A practical application of the method
can be made in the determination of vanillin in vanillin sugar. The aromatic
substance is extracted with alcohol and titrated as a monobasic acid.
What has been said for very weak acids also holds for very weak bases.
Aniline, hexamethylenetetramine and pyridine can be titrated accurately with
hydrochloric acid.
Neutralization of Weak Acids with Weak Bases.-Though in analytical
work one always uses a strong acid or a strong base as a reagent, the' above
case has practical significance; e.g. when a pure am,monium salt of a weak acid
has to be prepared or when a weak acid must be titrated in the presence of an
ammonium salt or the salt of another weak base.
Figure 347 gives the titration·lines of 0.1 Nand 0.01 N acetic acid, respec-
tively with ammonia. The neutralization curve up to the equivalence point is
____ ml NHa
FIG. 347.
about the same as found with sodium hydroxide since both sodium and am-
monium acetate are strong electrolytes. After the equivalence point has been
reached, an excess of ammonia leaves the conductivity practically unchanged
as the dissociation of the weak base IS depressed by the presence of the am-
monium salt. On account of the hydrolysis, the values .near the equivalence
point are somewhat lower than the corresponding points on the straight lines.
If the acid or Ibase to be titrated is extremely weak (e.g. boric acid with am-
monia), no straight lines can be constructed on account of the strong hydrolysis.
2332 CONDUCTOMETRIC METHODS
-
FIG. 348.
,
ml NaOH
first. neutralized; after the first break, the
conductivity increases and after neutrali-
zation of the phenolic group, a second
break occurs.
The application of the conductometric method to the titration of a mixture
of two weak acids with quite different ionization constants (like acetic and
boric acid) 'does not have much practical significance, for a sharp break never
occurs after the neutralization of the stronger acid. The difference in mobility
of the anions of the two acids determines the sharpness of the break and as this
difference always is very small, the angle between the two salt lines will be very
obtuse. Therefore, as a rule. tlie first equivalence point cannot be found with
a high degree of accuracy. ,
Replacement Titrations.-When a salt of a weak acid is titrated with a
strong acid, the anion of the weak acid is replaced by that of the stronger one
and the weak acid itself is liberated· in the undissociated form. Similarly, the
addition of a strong base replaces the weak base in a salt-of-the latter.
Many important applications of these replacement reactions can be made •
to conductometric titrations; especially those cases are important which do IlOt
yield successful results with the indicator of potentiometric method.
If, for example, hydrochloric acid is added'to a solution of sodium acetate]
the acetate is replaced by the chloride ion. The cpnductivity increases slightly
on aceount of the fact that ;\Cl- is a little greater than ;\acetate. After all the
acetic acid has been liberated, continued addition of hydrochloric acid gives
:rise to a strong increase in conductance. Fig. 349 gives the conductance lines
in the titration of 0.01 N sodium acetate with N hydrochloric acid. On ac-
count of the dissociation of the acetic acid, the \):If]!erimehtal figures near the
equivalence point are somewhat higher than ~htr corresponding ones on the
straight lines. If the ionization constant of the liberated acid is smaller than
CONDUCTOMETRIC METHODS 2333
about 5 X 10-5, 0.01 N solutions of its salt can be titrated accurately. For
0.1 N solutions, the constant may even be as large as 5 X 10-4 • Ethyl alcohol
decreases the ionization of weak acids and in the presence of enough alcohol
even salts of stronger acids than
those indicated can be titrated.
The method furnishes a simple
means for the evaluation of salts
of }veak acids, such as acetates,
benzoates, succinates, etc. Simi-
larly, it can be applied to the ~
determination of ammonia in am- 1;
monium salts by titration ,vith .g
sodium hydroxide. A rapid deter- 8
mination of the ammonium content
of fertilizers can be made in this
way .
. In this review, a condensed
summary has been given Of the
application of conductometric ti- ~ ml H Cl
trations to acid-base reactions. FIG. 349.
For details concerning the com-
bination of different cases and specific cases (carbon dioxide, phosphoric
acid, alkaloids, phenols, salts of heavy metals, etc.) the reader is referred to the
monograph of Kolthoff (1.c., p. 2328).-
From all that has been said, it may be inferred that the significance of
conductometric titrations should be more generally recognized. They often
furnish us information which could otherwise only be obtained by elaborate
work.
~
ing ion, the more accurate will be the result. If silver sal~ is titrated with
lithium chloride, the conductivity decreases during the precipitation ;Z~d
increases after the equivalence point. If hydrochloric acid were used as a
reagent, the conductivity would increase from the peginning of the titration
and the angle between the precipitation and reagent line would be very obtuse.
Therefore, generally it is recommended to titrate qations with lithium salts
and anions with acetates. (2) The larger the mobility of the anion of the
reagent which reacts with the cations to be determined (or vice versa) the more
acute is the angle. For example, it is more advantageous to titrate silver salts
with sodium chloride than sodium nitro prusside, as the mobility of the chloride
ion is larger than that of the nitro prusside. (3) The titration of a slightly
ionized salt does not give good results since the conductivity increases relatively
much from the start of the determination. (4) As has been mentioned in the
beginning of this chapter, the accuracy of a conductometric titration always
suffers from the presence of elec-
trolytes that do not take part in
the reaction. .
• (b) On account of the solu-
bility of the precipitate, -the ex-
perimental figures near the equi-
valence point deviate from the
straight line as is shown in Fig.
350. A E D F C gives the
B experimental curve. If the solu-
bility of the precipitate were
negligibly small, the conductivity
_ ml reagent at the equivalence point would
be equal to B G. By an excess
G
FIG. 350. of the ion which. is precipitated
or an excess of reagent, the solu-
bility is depressed and if the latter is not too large, i1f usually is possible to con-
struct the precipitation and reagent line by joining points on A E and F C.
It can be shown that 0.1 N solutions can still be titrated if the solubility of
the precipitate formed is less than about 0.005 N (uni-univalent electrolyte);
for 0.01 N solutions the solubility should be smaller than about 0.0005 N.
(c) The formation of a microcrystalline precipitate is usually a time reac-
tion. After addition of a small amount of reagent, the-conductance does not
immediately become constant; one must wait for some tir~e: Seeding the •
solution with the precipitate itself has a favorable effect though· in the titration
of very dilute solutions even this does not overcome the difficulty. Quite
generally it is much better to add enough alcohol to make its concentration
about 30-40%. The speed of formation of the precipitate then becomes much
larger and furthermore another advantage is g~ined, the solubility of most
slightly soluble substances is materially reduced by alcohol. Of course, one
should rea:iize that on addition of alcohol to a solution the temperature rises;
therefore, before the titration is started, the mixture should be cooled to room
temperature. I
(d) If the precipitate has strong adsorben~ pr(}perties, the method does not
yield successful results. So, for example, the "titration of heavy metals with
CONDUCTOMETRIC METHODS 2335
SAMPLING
As i~ all analytical processes, the obtaining of a representative s~mple is of
cardinal importance. It is done by the insertion of a tube into the duct or flue
and the withdrawal of the gas into a container. Previous to this, the furnace,
duct or setting must be thoroughly inspected for air leaks and these carefully
stopped. This is all the more to be done if, as often happens, the boiler plant
has been built on filled or " made" land. Finally the fact must not be lost
sight of that the bricks themselves composing the setting are porous. The
search for leaks is conducted using a lighted candle, and the places where the
flame is sucked in are marked and plastered up: the joints in the cleanout doors
are often corroded or rusty and no longer fit tightly. This is also true of the
frames of the doors themselves in the brickwor.k If the bricks and mortar of
the setting leak, it should be given a thin coat 6f size and a couple of coa~s of
thin whitewash to close the pores. Few people have any idea of this loss from
air leakage, and by radiation of heat from the setting. A case cited 2 in a
western city in which the setting was covered with asbestos cement over which
canvas was pasted and the whole heavily painted, showed a saving of 12%.
If the sample be from a boiler it should be taken as soon after leaving it as
practicable-not over the grate where the gases might be dissociated, and not
at the end of a long duct where they may have been air-diluted. Bends in the
ducts should be avoided and a place chosen preferably four or five diameters of
the duct away from the bend, where the gases are flowing smoothly. The tube
should be inserted to a distance of one-fourth to one-third the radius of the duct
from the circumference. Multiple tubes, tubes with varying sized holes, are
not to be recommended, ,as they are of doubtful expediency. Insertion is
preferably made on the grate side of the damper, t9 avoia air leaks, by drilling a
hole in the flue and fitting a screwed nipple into the hole: the sampling tube is
cemented into this nipple with Portland cement in case of a permanent installa-
tion. When the connection is temporary the hole may be puttied or made tight
in ahY way-even with wet cotton waste. A similar hole is made for the
thermometer., "-
a. Tubes.-The best tube is of translucent quartz, about 74 i~ch'in internal
diameter and 3 feet long: this will withstand any temperature or gas likely to be
encountered in practice. Pyrex glass tubes will probabiy be found sufficient
for all practical purposes and may be used with-temperatures up to 600 0 C.
Uncooled metal tubes cannot be used aboye 250 0 as their oxides are reduced by
hydrogen and carbon monoxide even at 200 0 according to the equation Fe20a
+3CO = 2Fe+3C02, the sample will be defective and an utterly false idea
obtained of the process by which it was produced. Dust and smoke can be
removed by plugs of asbestos, or if the temperature permits, rolls of iron wire
gauze or steel wool. These must be replaced when c!o~ged. I
1 Chapter by the late Augustus H. Gill.
2 Electrical WorId.
2336
GAS ANALYSIS 2337
h. Pumps.-Any form of a mechanical pump may be used, of which the
ordinary rubber syringe bulb is the simplest. In using this it should be remem-
bered that the valves do not shut tight, and in sampling it may have to be shut
off. by a pinch-cock. Where a head of water, steam or compressed air is avail-
able, the familiar Richards 3 jet pump (Figs. 351 and 352) may be employed.
This operates on the principle of
YQ\r R the boiler injector-small slugs or
+ pistbns of water filling the jet and
sucking in the gas after them.
Where a fall rather than a head
/.'R w of water is available, the Bunsen
pump (Fig. 353) is used. This con-
a sists of a 7.4:' inch T connected to the
, water reservoir and to
a 12 ft. piece of 74
inch rubber tubing b,
·the stem of the T a
being connected to the
source of water. When
the water flows it sucks
the gas in after it
through c: 12 ft. of
tubing will usually be
found sufficient, 34 ft.
are nerded for the high-
est vacuum. If none of I>
the foregoing be avail-
able, a pair of aspirator
FIG. 351. FIG. 352. bottles described fur-
ther on may be used.
c. Containers.-These should be of glass, or a
metal which is unacted upon by the gases to be
encountered. Rubber bags are unsuitable, for
rubber is by no means the inert substance which
we are in the habit of considering it. Rubber is
oxidized or attacked by various gases and allows
the lighter gases to transfuse through it, and dis-
solves the heavier hydrocarbons, as ethylene. FIG. 353.
\cid gases like sulfur dioxide are dissolved per-
.ceptibly by passing through the usual white laboratory tubing, containing pO
to 60% of zinc oxide, giving low results. In consequence of this, rubber con-
nections exposed to gases in all gas apparatus should be as. short ,as possible,
the glass tubes being butted together, leaving barely room for the pinchcock.
The ends of all glass tubes connected by rubber must be file-rounded, to remove the
sharp cutting edge of the glass tubing-a prolific source of leaks in stoppers as
well as rubber connectors.
The shape of the container should be such as to allow the water in it to
wholly run out; a glass bulb, pear-shaped at the bottom (Fig. 354), is excellent.
3 Am. J. Sci. (3), 8, 412 (1874).
2338 GAS ANALYSIS'
with the gas sought, is the next best, or tap water alone, or strong brine. In
using the latter it should be remembered that it corrodes metal-pinchcocks,
etc., and salt is hygroscopic.
In this connection. the fact should not be overlooked that water does not
dissolve carbon dioxide instantly-but that time and the surf~ce exposed are
important considerations. In dispensing chimney gas to classes from the
sampling apparatus-Fig. 355--under'a pressure of 4 to 6 ft. of water, the writer
has repeatedly observed the fact that the last sample would often show more
carbon dioxide than the first, even after 15 minutes standing. The converse
of the proposition seems to hold true, as the writer has heated samples of water
10° above the saturation point for 24 to 48 hours without the water giving up
the excess oxygen. 4 • 5 • ' /
The foregoing procedure applies to sampling gases from a boiler or flue.
To sample city gas, it should be allowed to burn until the old gas is certainly
out of the piping, or it can be blown out into the outside air.
Df all these instruments, the wet meter and Thomas meter are probably the
most accurate. They, however, like some of the others, cannot be used with
corrosive gases, being constructed of me~al; hence the Pitot tube and rotameter
are the ones to be employed in chemical works.
4 Technology Quarterly,S, 250 (1892) .
. 6 See also Ind, Eng, Chern., AnaL Ed., 7, 37 (1935).
6 Taken partly from Marks' "Mechanical Engineers' Handbook."
7 Said to be accurate.
2340 GAS ANALYSIS
The wet meter consists of a cylindrical drum divided into fom spiral com-
partments, suspended in a bath of water, surrounded !by a tight casing: the
pressure of the gas causes the drum to rotate, emptying a drumful of gas into
the casing and pipes. It must be set level, the water level carefully adjusted
to the mark on the glass gauge, with its top open; as wel~ as the, inlet and outlet
pipes of the meter. The higher the water level the faster the meter. The
dry meter consists of a pair of metal bellows, with sides of leather soaked in oil,
on either side of a diaphragm, and connected with slide valves so that a bel-
lowsful of gas is alternately emptied into the upper part of the meter and piping.
The vibrations of the bellows produced by the gas pressure are transmitted to
clockwork and indexes. It is to be noted that the indexes apparently read-one-
tenth of the actual volume passed: the index must make a complete circuit to
register the amount stamped on the dial. A small index an~ial are usually
present for testing the meter, and a tolerance of 2% is allowed'by law. This
testing is done by meter-provers, carefully calibrated gas holders kept at
constant temperature and the rate determined at different speeds. The meter-
provers, are, in their turn, calibrated by a cubic foot, standardized at the Bureau
of Standards. It should be noted that all this calibration is corrected, not to
standard conditions (0° C. and 760 mm.), but to the cubic foot as fixed by, law-
gas saturated with moisture at 60° F. and 30 inches. 8
The Connersville meter is the pump of that name, or Roots blower, in which
the flow of gas operates the two figure 8 vanes connected with an index.
The Pitot Tube.-Fig. 356. This consists of two glass tubes, D, of about 7i
to 1 inch internal diameter, inserted in the gas stream: one is bent at right
angles and is set so that it receives the
o impact of the gas movement; the other
merely registers the pressure of the gas
in the pipe. The impact tube should
have a straight portion parallel to the
air flow, of a length about 10 times the
internal diameter of the tube. The
open end I'lhould be fire polished. 9 The'
point of insertion of these tubes in the
chimney or duct should be in a long
straight run of pipe, so as to be as free
from eddies as possible. The tube when
FIG. 356. inserted should be preceded by 10 to 20
diameters of straight duef 'Or pipe and
followed by about 5 diameters.9 Davis 10 says that authorities- differ as to
whether the tubes should be a third or one-sixth of the diameter from the cir-
cumference to show the mean speed: he states further that each flue or chimney
is a separate problem and as a result of hundreds of measurements there seems
to be "no settled proportionate distance correspondihg to the mean velocity."
The glass tubes are connected by rubber tubing,' C, through a reverser, B,.
with a U-tube_ A, which is -either set vertically, inclined 30°, or one in ten-this
82° F. makes a difference of about 0.5% in the volume.
9 Marks' Handbook. I
10 "Handbook of Chemical Engineering," 1, 197; also fd~the tables for its use. Also
Walker, Lewis and McAdams "Principles of Chem1cal Engineering" or Marks'
Handbook.
GAS ANALYSIS 2341
carries a vernier reading to hundredths of an inch so that readings to thou~
'sandths are feasible. The U~tube is filled with solutions of different specific
gravity, although ether of 0.74 sp.gr. is the one commonly employed. The
difference between the arms of the U-tube represents the difference between the
kinetic and static pressures of the gas in the flue or chimney.
The formula for calculating the velocity recomlflended by W. W. Scott is
1'=1290
th(1 + .002176t)
BM
where h = vertical differential of gauge reading in terms of inches of water.
! the reading. is taken as the actual reading is double that due to flow pressure
alone. B = barometric pressure of the gas in inches of mercury (29.92/1).
M = specific gravity compared with H = 1. Air = 14.39. t = temperature of
the gas. V = velocity in feet ,per second. The
formula V =42~h gives fairly accurate results.
Davis' formula is
Vo= • I h459+tO X28.55.
'\j 519
I Marks says of the Pitot tubes that they ~' are the
• most satisfactory devices for the measurement of
gas in large volumes." •
The Rotameter, Fig. 357.-This is a German in-
strument depending upon the height to which a float
-.is carried in a glass tube by the velocity of the stream
of gas. A modification of it was used by some of
the gas-lighting companies under the name of the
Thorp Gauge:
It consists of a graduated glass tube fixed upon a
tripod and provided with a plumb linG so that it can
be set vertically: gas passes in at the lower end, raises
the clay or talc float to a certain height and passes
out at the top. The height to which the float is
raised is noted on the graduations of the tube. The
formula for its use is
Vl=V~,
V = volume gas as shown by reading of instru- FIG. 357.
ment, M its specific gravity = 1.0.
V = volume desired, Ml =sp.gr. of gas.
It is made in all capacities from 0.2 cu. ft., per hour up.
The Thomas Electric Gas Meter.-This depends upon the principle that if
the specific heat of the gas be ~nown, and an amount of energy be put into it,
by means of a coil, sufficient to keep a certain difference of temperature between
two thermometerf1, one before and one after the coil, this energy is a direct
measure of the VOlUme of gas flowing. Two electrical thermometers are placed
in the stream of gas with a heating coil between them: 2° difference of tempera-
2342 GAS ANALYSIS
·FIG. 359.
I
tion and its size the quantity of gas passing can be determined:- 'For its use see
Marks' " Handbook" or Walker, Lewis and McAdams, " Principles of Chemi-
cal Engineering."
volume.
Correcting next for temperature (Law of Charles V : V' : : T : T') the for-
mula is Vo : V(oh •• ) : : To : T(ohs.), T being in absolute degrees
273 1
Vo= V(ohs.) X273+To C. 1+0.00367 TO C.'
. 1
then the formula becomes Vo= V(obs.) X 1+0.00367 TO C.
B, Aq and M are in inches or" mercury: 0.51 inch is the t1ension of aqueous vapor
at 60 0 F. In practice this reduction is obtained from Table 9 (Appendix).
Specific Gravity.-The specific gravity of a gas is usually referred to air as
unity: it can be readily calculated by dividing the molecular weight of the gas
by 28.95, the weight in grams of 22.4 liters of air (22.4 liters will be remembered
as the gram-molecular volume).
Specific gravity can be directly determined by Jenkins' apparatus, p. ~7, or
the Edwards gas balance. H , 15 (
This depends upon an indirect weighing of the gases. The apparatus
consists of a closed vessel counterpoised in a water-jacketed cylinder of the gas;
this is connected with a mercury gauge. Dry air is passed into the cylinder
and pressure applied to it until the vessel just balances at a predetermined
point and the gauge reading noted. TIle operation is repeated with the gas
to be tested. The densities of the gases must be the same at these pressures;
or the densities of the two gases at normal presflure are in the inverse ratio of
the pressures, when of equal buoyant foree. Results accurate to 0.2% are
obtained and no calibration is required. / It has given trouble- at times owing
apparently to leaks in the" closed vessel" mentioned.
With the effusion method results are dependent upon the orifice used, the
pressure and the confining medium. They may vary as much as 10% with
different observers. The orifice should be 0.18-0.3 mm. in diameter in a plate
0.04 mm. or less thick: the pressure should not be allowed to fall below 2 inches
of water: results to an accuracy of 2% may be expected.
These are quite varied according to the purpose fof which they are intended.
A very efficient form is the Friedrichs Spiral Gas Washing bottle, Fig. 360;, here
the gas has to pass through a long spiral path. Dennis recommends this for
the absorption of sulfur dioxide. The gas is run through a solution until a
color change takes place. Were the reagent t~ be washed out and titrated it
would not seem so well 'adapted on account of the difllculty _of thoroughly
washing it.
The Varrentrapp and Will bulbs, Fig. 361, the use of which is evident from
the figure, are used for the absorption of ammonia in illuminating gas. The
Wolff absorption tube, Fig. 362, the empty arm 'of which is filled with beads or
broken glass, is used for the absorption of carbon monoxide by bloo.d. The
bulbed tube, provided with a small jet, is generally u'sed for containing standard
solutions of acid, alkali, or potassium permanganate or, in general, a solution
which is to be titrated after absorption. Where the presence of a gas as, for
example, water vapor is to be determined by the increase of weight of the
reagent used, Winkler's spiral, Fig. 363,16 may be employed/
14 Ind. Eng. Chern., 9, 790 (1917). Bur. Stds. Tecl(n:Bull. No. 94.
16 Improvements by Warren, C. A., 30, 2425.
GAS ANALYSIS 2345
FIG. 361.
Gases are sometimes tested by burning them arid investigating the products
of combustion. Often the gas refuses to burn, as in the case of wood gas, as it
contains so much incombustible gas (30% CO 2): ~his is removed with caustic
alkali and the burning test repeated. In performi~g this test, partic~with
electrolytic gases, it must be remembered that they\may contain enough oxygen
or chlorine to support their own combustion like ',' detonating gas." A glass
tube 18 inches long well packed with glass beads, broken glass or porcelain must
always be inserted between the gas holder and the jet: this will extinguish the
flame of the gas wave, should it start, after the manner of the gauze in the Davy
safety lamp, and prevent flame from reaching the explosive mixture in the gas
holder with serious results. Sufficient pressure should of course be maintained
on the holder itself.
DETECTION AND DETERMINATION OF THE VARIOUS GASES
Clemens Winkler 18 divided the gases into seven groups according to their
behavior with various solvents. These were contained in suitable absorption
tubes or vessels and the gases passed through them. His scheme was as
follows:
Gases absorbed by
1. H 2S0 4 1.7 sp.gr.; NHa, (N 20a), N 20 4 .1 9
II. KOH 1.3 sp.gr.;· Ch, HCI, (CN)2, HCN, S02, H 2S, SiF 4, CO 2.
III. AgNOa; PHa, AsH a, SbHa. /
IV. Pyrogallol; 02, (Oa).
V. CuCI; CO.
VI. Acid FeS04 1 : 2; NO.
VII. Unabsorbed; H 2, CH 4 , C 2H 2, C 2H 4, N 20, N 2, CbS, and the noble gases.
The following tables give the specific gravity referred to air, the solubility in
water at 20°, the qualitative tests and quantitative methods of determination
of these gases: additional means will be found mentioned under the several
gases themselves. :
Group r .
GASES ABSORBEP BY H 2S0 4 1.7 SP.GR.
J •
Groupm
GASES ABSORBED BY AgNO a
Group VII'
UNABSOH,BE.D
Methane Ethylene
Acetylene
.
(ethene)
(ethine)
Gravity, air = 1. ..... 0.0694 0.5545 0.9684
0.9056
Qualitative test ...... Dry PdCb gives None Red ppt. w.
HCI am. CuCl
(explosive !)
Quantitative det'n ... By combustion By combustion Absorption w. As C.H" which
or explosion or explosion Br water or see
w. O. w. O. H 2S 20 1 '"
0
Solubility 20 C., 1 ml.
water absorbs ml. .. 0.0183 0.0330 0)'22 1.03
Nitrous oxide Carbon oxysulfide Nitroge~ The noble gases
1.5301 2.075 0.9673 Helium, Neon,
Argon, Krypton,
Xenon
None None None
By explosion with H. Alcoholic KOH By absorption or com~
.._
or combustion w.
CuO
1 : 3 in 66% al-
chol by weight
bustion of all other
gases and measuring
-
the residue which -
also contains the no-
Qle gases
0.670 0.568 .0159
" ("
ANALYSIS OF GASES
The analysis of a gaseous mixture is effected by absorbing the various
constituents and observing the diminution in volume: in case the gas be un-
absorbable, as for example methane (CH 4), it is burned and the carbon dioxide
and water determined. t
(a) Analysis of mixtures for carbon dioxide, oxygen and carbon monoxide
(e.g., chimney gases, producer and blast furnace gas) can be done with any of
the apparatus to be described. The Orsat, or Elliott are the forms usually
employed.
(b) Analysis of mixtures as in (a) and also containing combustible gases as
hydrogen and methane, e.g., city gas.
The Orsat Apparatus. Description.-The apparatus,24 Fig. 364, is enclosed
in a case to permit of transportation from place to place; furthermore, the
measuring tube is jacketed with water
to prevent changes of temperature
affecting the gas volume. The appa-
Ta tu's consists eS'sentially of the level-
ing bottle A, the burette B, the pipettes
pI, pll, pilI, and the connecting tube
T. Pipette P' is filled with potassium
(or sodium) hydroxide solution (see
Reagents) so that when it is <Jrawn
up into the front arm about half an
inch in depth is left in the rear arm.
Pipettes p" and pilI are similarly filled
with potassium (or sodium) pyrogal-
late and cuprous chloride solutions
respectively. These reagents require
\ to be protected from the oxygen of the
air by collapsible rubber bags of good
quality. As the oxygen in the air over
the reagent is absorbed, a diminution
in pressure takes place rendering it
difficult to bring the reagent to the
point on the stem: the obvious remedy FIG. 364.
is to remove the bag temporarily and
adjust the reagent. When the apparatus is first set up, one or two blank
analyses should be madc) to saturate the water aqd reagents with the gases.
For example the potassIum hydroxide absorbs carbon dioxide, it also absorbs
about 3 m!. of oxygen, 2 Ill!. of carbon monoxide and 1.5 m!. of nitrogen, by
virtue of the 100 m!. of water which it contains. A change of temperature of
10 makes a change of 0.36 of the volume of the gas: a change of pressure of 1 mm.
produces 0.13 change in the volume.
Manipulation.-The reagents in the pipettes should be adjusted in the capil-
lary tubes to a point on the stem about midway between the top of the pipette
24 The apparatus shown is one of the simplest and least expensive forms: it has proved
itself well adapted for occasional use. Hayes has constructed it in a very attractive and
compact form in metal.
GAS ANALYSIS
and the rubber connector. This is effected by opening wide the pinchcock
upon the connector, the bottle being on the table, and '[ery gradually lowering
the bottle until the reagent is brought to the point above indicated. Six
inches of the tubing used correspond to but 0.1 mI., so ~hat an error of half an
inch in adjustment of the reagent is without influence I upon the accuracy of
the result. The reagents having been thus adjusted, th!'l burette and connect-
ing tube are completely filled with water by opening d and raising the leveling
bottle. The apparatus is now ready to receive a sample of gas (or air for
practice). In case a flue-gas is to be analyzed d is connected with i, Fig. 354,
A lowered and about 102 mI. of the gas forced over by opening h; or d may be
connected with a T-joint in the gas stream; the burette after filling is allowed
to drain one minute by the sand glass, c snapped upon its rubber tube, and the
bottle A raised to the top of the apparatus. By gradually opening c the water
is allowed to run into the burette until the lower meniscus stands upon the
100 or 0 mark (according to the graduation of the apparatus)'. The gas taken
is thus compressed into the space occupied by 100 mI. and by opening d the
excess escapes. Open c and bring the level of the water in the bottle to the same
level as the water in the burette and take the reading, which should be 100 ml.
Special attention is called to this method of reading: if the bottle be raised, the
gas is compressed; if lowered, it is expanded. ,
Determination of Carbon Dioxide.-The gas to be analyzed is invariably
passed first into pipette pI, containing potassium hydroxide for the absorption
of carbon dioxide, by opening e and raising~. The gas displaces the reagent
in the front part of the pipette, laying bare the tubes contained in it, which
being covered with the reagent present a large absorptive surface to the gas;
the reagent moves into the rear arm of the pipette, displacing the air over it into
the flexible rubber bag "\vhich prevents its diffusion into the air. The gas is
forced in and out of the pipette by raising and lowering A, the reagent finally
brought approximately to its initial point on the stem of the pipette, the burette
allowed to drain one minute and the reading taken. The.difference between
this and the initial reading represents the milliliters of. 'carbon dioxide preslmt
in the gas. To be certain that all the carbon dioxide is 'removed, the gas should
be passed a second time into pI and the reading t~ken as before; these readings·
shouldtagree within 0.1 %, and are called" check readings."
Determination of Oxygen.-The residue frcim the absorption of carbon
dioxide is passed into the second pipette, pll, containing an,~lkaline solution of
potassium pyrogallate, until no further absorption will take'place.. The dif-
ference between the reading obtained and that after the absorption of carbon
dioxide, represents the number of milliliters of oxygen present. -
Determination of Carbon Monoxide.-The residue from the absorption of
oxygen is passed into the third pipette, pili, containing cuprous chloride, until
no further absorption takes place; that is, in this cafje until readings agreeing
exactly (not merely to 0.1) are obtained. The difference between the reading
thus obtained and that after the absorption of oxygen, represents the number
of milliliters of carbon monoxide present.
Determination of Hydrocarbons.-The residue left after all absorptions have
been made may consist, in additon to nitrogen, the:~principal constituent, of
hydrocarbons and hydrogen: see p. 2356. . -' -
GAS ANALYSIS 2351
Accuracy.-The apparatus gives resuits accurate to 0.2 of 1 %, hence figures
obtained by division to 0.01 should not be reported.
Time Required.-About twenty minutes are required for an analysis; two
may be made in twenty-five minutes, using two apparatus.
NOTEs.-The method of adjusting the reagents is the only one which has been found
satisfactory: if the bottle be placed at a lower level. and an attempt made to shut the
pinchcock c upon the connector at thc proper time, it will almost invariably result in
failure.
The process of obtaining 100 m!. of gas is exactly analogous to filling a measure
heaping full of grain and striking off the excess with a straight edge; it saves arithmetical
work, as milliliters read off represent per cent directly.
It often happens when e is opened, c being closed, that the
reagent pi drops, due not to a leak, as is usually supposed, but to
the weight of the column of the reagent expanding the gas.
The object of the rubber bags is to prevent the access of air to
the reagents, those in p" and pili absorbing oxygen with great
avidity, and hence if. freely exposed to the air would soon be-
come useless. These bags if well made will last two years and
am'ply protect the solution: or a "sealing pipette" can be used
for the pyro.
Carbon dioxide is always the first gas to be removed from a
gaseous mixture. In the case of air the percentage present is so
small, 0.08 to 0.1, as scarcely to be seen with this apparatus. It
is important to use the reagents in the order given; if by mistake
the g:;ts be passed into the second pipette, it will a-bsorb not only
oxygen, for which it is intended, but also carbon dioxide; similarly
if the gas be passed into the third pipette, it will absorb not only
carbonic oxide, but also oxygen afl. well.
'The use of pinchcocks and rubber tubes, original with the
author, although recommended by Naef, is considered by Fis-
cher to be inaccurate. The experience of the author, however,
does not support this assertion, as they have been found to be fully
as accurate as glass stopcocks, and very much less troublesome
and expensive.
In case any potassium hydroxide or pyrogallate be sucked
over into the tube T or water in oA, the analysis is not spoiled,
but may be proceeded with by connecting on water at d, opening
this cock, and allowing the water to wash the tubes out thoroughly.
The addition of a little hydrochloric acid to the water in the bottle
A will. neutralize the hydroxide or pyrogallate, and the washing
may be postponed until convenient: Methyl orange indicator ,0
may be added to be sure the water is acid.
After each analysis the number of milliliters of oxygen and
carbonic oxide should be set down upon the ground-glass slip pro-
vided for the purpose. By adding these numbers and subtracting
their sum from the absorption capacity (see Reagents) of each
reagent, the condition of the apparatus is known at any time,
and the reagent can be renewed in season to prevent incorrect
analyses.
Elliott Apparatus. Description.---;The apparatus, 25 Fig.
365, consists of a burette holding 100 ml. graduated in FIG. 365.
tenths of a milliliter and bulbed_ like the Bunte appara-
tus-the bulb holding about.30 ml.; it is connected with a leveling bottle similar
to the Orsat apparatus. The top of the burette ends in a capillary stopcock, the
stem of which is ground square to admit of close connection with the" labora-
20 The simplest form is here shown as being best adapt~d for teaching purposes.
2352 GAS ANALYSIS
tory vessel," an ungraduated tube similar to the burette, except of 125 m!.
capacity. The top of this" vessel" is also closed1with a capillary stopcock,
carrying by a ground-glass joint, or better a rubber stopper, a thistle-tube F,
for the introduction of the reagents. The lower end of this" vessel" is closed
by a rubber stopper carrying a three-way cock 0, and connected with a leveling
bottle D. The burette and vessel are held upon a block of wood-supporter!
by a ring stand-by fine copper wire tightened by violin keys.
Manipulation.-The ground-glass joints are lubricated with stopcock grease.
The leveling bottles are filled with water, the stopcocks opened, and the bottles
raised until the water flows through the stopcocks m and n. m is connected
with the source whence the gas to be analyzed is to be taken, n is closed, D
lowered and rather more than 100 ml. drawn in, and m closed. n is opened,
D raised and E lowered, nearly 100 ml. of gas introduced, and n closed; by
opening m and raising D the remainder of the gas is allowed to escape, the
tubes being filled with water and m closed. n is opened and the water brought
to the reference mark; the burette is allowed to drain one minute, the level of
the water in E is brought to the same level as in the burette, and the reading
taken.
Determination of Carbon Dioxide.-By raising E, opening n, and lowering
D, the gas is passed over into the laboratory vessel; F is filled within half an
inch of the top with potassium hydroxide, ° closed, m opened, and the reagent
allowed slowly to trickle in. A No. 3/evaporating dish is placed under 0, and
this turned to allow the liquid in the laboratory vessel to run into the dish.
At first this is mainly water, and may be thrown away; later it becomes diluted
reagent and may be returned to the thistle-tube. When the depth of the
reagent in the thistle-tube has lowered to half an inch, it should be refilled either
with fresh or the diluted reagent and allowed to run in until the absorption is
judged to be complete, and the gas passed back into the burette for measure-
°
ment. To this end close and then m, raise E, open n, and force some pure
water into the laboratory vessel, thus rinsing out the capillary tube. Now raise
D and lower E, shutting n when the liquid has arrived at the reference mark.
The burette is allowed to drain a minute, the: level' of the water in the bottle E .
brought to the same level as the water in the burette, and the reading taken.
Determination of Oxygen.-The manipulation is the same as in the pre-
ceding determination, potassium pyrogallate being substituted for potassium
hydroxide; the apparatus requiring no washing out. __
Determination of Carbon Monoxide.-Thc.laboratorY vess~l, thistle-tube,
and bottle if necessary, are washed free from potassium pyrogallate and the ab-
sorption made with acid cupr:ous chloride similarly to the 'determination of
carbon dioxide. The white precipitate of cuprous chloride may be dissolved by
hydrochloric acid.
Accuracy and Time Required.-The apparatus is as accurate for absorptions
as that of Orsat; it is stated to be much more rapid-a claim which the writer
cannot substantiate. It is not as portable, is more fragile, and more trouble-
some to manipulate, and as the burette is not jacketed, it is liable to be affected
by changes of temperature. ,
NOTEs.-In case at any time it is desired to stlp.lhe influx of reagent, 0 should be
closed first and then m; the reason being that the !tosorption may be so rapid as to suck
air in through 0, m being closed.
GAS ANALYSIS 2353
The stopcock should be so adjusted as to cause the reagent to spread itself as com-
pletely as possible over the sides of the burette.
By the addition of an explosion tube it is used for the analysis of city gas,'· bromine
being used to absorb the "illuminants," Winkler '7. states that this absorption is incom-
plete; later work by Treadwell and Stokes, and also Korbuly,'s has shown that bromine
water, by a purely physical solution, does absorb the" illuminants" completely; Hempel 2g
states that explosions of hydrocarbons made over W!lter are inaccurate, so that the
apparatus can' be depended upon to give results upon methane and hydrogen only within
about 2%. It is, however, very rapid, a complete analysis of city gas can be made with
it in fifty-five minutes.
Hempel's Apparatus. Description.-The apparatus/a Figs. 366 and 367,
is very similar in principle to that of Orsat; the burette is longer, admitting of
FIG. 366.
the reading of small quantities of gas and the pipettes are separate and mounted
in brass clamps on iron stands. P shows a " simple" pipette 31 provided1with
a rubber bag; this form, after forty years of use, can be said satisfactorily to
take the place of the cumbersome" compound" pipette.
2. Mackintosh, Am. Chern. Jour., 9, 294 (1887).
27 Zeit. Anal. Chern., 28, 286 (1889).
28 Treadwell-Hall's "Quantita~ive Analysis." p. 569.
29 "Gasanalytische Methoden" p. 1O~
30 The simplest forms are here shown.
• 31 Gill. Am. Chern. J., 14, 231 (1892).
2354 GAS ANALYSIS
The pipette for fuming sulfuric acid 32 _is shown at F, and differs from the
ordinary in that vertical tubes after the manner of tqose in the Orsat pipettes
replace the usual glass beads. This prevents the trapping of any gas by the
filling, which was so common with the beads and glas~ wool. E represents the
large explosion pipette,33 of about 250-ml. capacitY,lwith walls half an inch
thick; the explosion wires enter at the top and bottoml to prevent short-circuit-
ing; mercury is the confining liquid. The small explosion pipette holds about
110 m!. and is of glass, the same thickness as the simple pipettes. Water is
here used as the confining liquid, and acidulated with Hel usually in the burette.
FIG. 367.
A " Ford" induction coil capable of giving a fourth-inch spark, with three
dry cells, four " simple" pipettes and a mercury filled burette, complete the
outfit.
The burette should be carefully calibrated and the corrections may very
well be etched upon it opposite the 10-m!. divisions. ..... .
In working with the apparatus the pipettes are placed up;Uthe adjustable
stand S and connection made with the doubly bent capillary tube.
Manipulation.-To acquire facility with the use of the apparatus before
proceeding to the analysis of city gas, it is well to make the following determina-
tions, obtaining" check-readings" in every cas~: 1. Oxygen in air, by (1)
absorption with phosphorus; (2) absorption w;ith potassium (or sodium)
pyrogallate; 34 (3) by explosion with hydrogen.
32 Ibid.,_J. Am. Chem. Soc., 18,67 (1896).
33 Gill, J. Am. Chern. Soc., 17,771 (1895).
34 The writer finds after an experience of more than thirty-five years in the laboratory
with hundreds of students, that sodium pyrogallate can .be~used with practically the same-
results as the potassium compound. The absorption' is complete, as shown by subse-
quent treatment with cuprous chloride.
GAS
. ANALYSIS. 2355
(1) By Phosphorus.-lOO m!. of air are measured out as with the Orsat
apparatus, the burette being allowed to drain two minutes. The rubber con-
nectors upon the burette and pipette are filled with water, the capillary tube
inserted, as far as it will go, by a twisting motion, into the connector upon the
,burette, thus filling the capillary with water; the free end of the capillary is
inserted into the pipette connector, the latter pinched so as to form a channel
for the water contained in it to escape, and the capillary twisted and forced
down to the pinchcock. There should be as little free space as possible be-
tween the capillaries and the pinchcock. Before using a pipette, its connector
(and rubber bag) should be carefully examined for leaks, especially in the
former, and if any found the faulty piece replaced.
The pinchcocks on the burette and pipette are now opened, the air forced
over into the phosphorus, and the pinchcock on the pipette closed; action im-
mediately ensues, shown by the white fumes; after allowing it to stand fifteen
minutes the residue is drawn back into the burette, the latter allowed to drain
and the reading taken. The absorption goes on best at 20° C., not at all below
15" C.; it is very much retarded by small amounts of ethane and ammonia.
It cannot be used to absorb commercial oxygen. No cognizance need be taken
of the fog of oxides of phosphorus.
(2) By Potassium PyrogaUate.-lOO m!. of air 35 are measured out as before,
the carbon dioxide absorbed with potassium hydroxide and the oxygen with
potassium pyrogallate, as with the Orsat apparatus; before setting aside the
pyrogallate pipette, the number of milliliters of oxygen absorbed should be
noted upon the slate s on the'stand. This must never be omitted with any
pipette save possibly that for potassium hydroxide, as failure to do this may
result in the ruin of an important analysis. The reason for the omission in this
case is found in the large absorption capacity-four to five liters of carbon
dioxide-of the reagent.
(3) By Explosion with Hydrogen.-Forty-three m!. of air and 57 m!. of
hydrogen are measured out, passed into the small explosion pipette, the capil-
lary of the pipette filled with water, the pinchcocks and glass stopcock all
closed, a heavy glass or fine wire gauze screen placed between the pipette and
the operator, the spark passed between the spark wires, and the contraction in
volume noted. The screen should never be omitted, as serious accidents may occur
thereby. The oxygen is represented by one-third of the contraction. For
very accurate work the sum of the combustible gases should be but one-sixth
that of the non-combustible gases; otherwise some nitrogen will burn and high
results will be obtained; 36 that is, (H+O) : (N +H) : : 1 : 6.
One hundred m!. of gas are measured from the bottle containing the sample
into the burette.
Determination of Carbon Dioxide.-The burette is connected with the
pipette containing potassium hydroxide and the gas passed into it with shaking
until no further diminution in volume takes place.
36 See Anderson's work, J. Ind. Chern., 7, 587 (1915).
36 This is shown in the work of Gill and Hunt, J. Am. Chern. Soc., 17,987 (1895).
2356 GAS ANALYSIS
From this it is evident that the volume of carbon dioxide is equal to the volume
of methane present; therefore in the above example, in the 56.2 m!. of gas
burned, there were 18.7 m!. methane.
40 Bunsen, "Gasometrische Methoden," 2d ed., 73 (1877),
41 Hempel. "Gas Analytische Methoden," 3d ed., 245 (1901).
2358 GAS ANALYSIS
Hydrogen then requires for its combustion half its volume of oxygen; hence
this 49.3 ml. represents a volume of hydrogen with 72 its volume of oxygen, or
3/2 volumes; hence the volume of hydrogen is 32.9 ml.
(b) Hempel's Method.-Of the 82 mL of gas remaining after the "absorp-
tions, 13.2 mL were used for the explosion; 86.4 mL air introduced, ~iving a
total volume of 99.6 ml.
Residue after explosion. . • . . . . . . . . . ..... . . . . . . . . . . . . . . .. 78.0 ml.
Residue after CO 2 absorption ........ : . . . . . . . . . . . . . . . . .. 73.2 :~
Carbon dioxide formed ....... " . . . . . . . . . . . . . . . . . . . . . . . . 4.8
Contraction ................................ 99.6-78.0=21.6 "
Residue after 0 absorption . ...... ~ ................... .. 70.2 ".
Oxygen in excess ............... " ........... 73.2-70.2= 3.0 "
The carbon dioxide being equal to the methane present, in the 13.2 ml. of
gas burned there were 4.8 ml. of methane. The volume of methane is found
by the proportion 13.2 : 82 : : 4.8 : x, whence x = 29.8 ml. ,
The hydrogen is calculated similarly.
Another method for the estimatiop. of hydrogen is by absorption with
palladium sponge; 4a it, however, must be carefully prepared, and it is the
author's experience that one cannot be sure of its efficacy when it is desired to
make use of it. A still better absorbent of hydrdgen 44 is a i % solution of
palladous chloride at 50° C.; when fresh this WIll absorb 20-50 mL of hydrogen
in ninety minutes. A proportionately longer time is required if more hydrogen
be present or the solution nearly saturated. The methane could then be deter-
mined by explosion or by mixing with air and passing to and fro over a· white-
hot platinum spiral in a tubulated pipette called the grisoumeter 45 (grisou
= methane).
Nitrogen.-There being no direct and convenient method for its estimation
with this apparatus, the percentage is obtained l?y finding the difference between
the sum of all percentages of the gases determined and 100%.
New 46 determined nitrogen in illuminating gas/directly after tl).e method of
Dumas in organic substances; 150 ml. of gas are 'used, the hydrocarbons par-
42 H 20 being as steam at 100° C. At ordinary temperatures this is condensed, giving
rise to "total contraction."
43 Hempel, Berichte, 12, 636 and 1006 (1879).
44 Campbell and Hart, Am. Chern. J., 18, 294 (1896)._
45 Winkler, Fres. Zeit., 28, 269 and 288. ~:__
46 J. Soc. Chern. Ind., 11, 415 (1892).
GAS ANALYSIS 2359
tially absorbed by fuming sulfuric acid and the remainder burned in a com-
bustion tube with copper oxide; the carbon dioxide is a.bsorbed and the residual
nitrogen collected and measured.
Accuracy and Time Required.-For the absorptions the apparatus is
accurate to 0.1 ml.; for explosions by Hinman's method 47 the methane can be
determined within 0.2%, the hydrogen within 0.3%; by Hempel's method
within 1% for the methane and 7.5% for the hydrogen. The time required
by the student for the analysis of city gas is from three to three and one-half
hours; for air, from fifteen to twenty minutes. For very accurate work the ap-
paratus of Burrell should be used.
NOTEs.-The object in filling the capillaries of the explosion pipettes with water or
mercury before the explosion is to prevent the bursting of the rubber connectors on them.
With mercury this is effected by introducing it through the T-joint in the connector.
After testing for oxygen with the pyrogallate a small quantity of dilute acetic acid is
sucked into the burette to neutralize any alkali which by any chance may have been
sucked over into it. The acid is rinsed out with water and this is forced out by mercury
before the burette is used again .
~~
1. The burette must be calibrated.
.
. For results of the highest degree of accuracy the following points must be borne in
Ammonia was studied by Priestley in 1774, who named it " alcaline air."
Ammonia is a pungent, strong smelling gas, lighter than air, specific gravity
about 0.6 50 (air being 1), dissolving readily in water, giving an alkaline solution.
It is the most soluble gas, water dissolving about 700 volumes 50 at or<1inary
temperature (20°). Alcohol dissolves 340 volumes, and ether dissolves 10
volumes at 15°. Liquid ammonia has been an article of commerce for many
years, being used in the refrigeration industry. At ordinary temperature
liquid ammonia exerts a pressure of abbut 125 lbs. per sq. inch (8.5 atmps-
pheres) and it boils at -33.5°. The capacity and sizes of the cylinders in
which it is sold are given in the Appendix, Table 8.
Cylinders containing liquefied gases must be stored in a cool place and
must not be handled roughly. They must be treated like a high explosive-
as some of t)1em are-the writer knowing of several cases in which ~asualties
and deaths even have resulted from thoughtless handling. In handling am-
monia the canister of the gas mask is filled with cupramite. 51 When inhaled
none of the gas is absorbed,s2 the eff~ct being entirely on the surface: 0.0.4% ,.,
irritates the throat and 0.3% is dangerous for short exposure.
Ammonia burns with a yellow flame with pure oxygen, it is explosive with
air, the explosive limits being from 16.5 to 26.8%.53 Accidents in refrigerating
plants have occurred from this cause. 54 From its high hydrogen content it .
behaves like hydrogen, reducing the oxides of the metals. It is decotrlposed
by the electric spark into its elements. The source of ammonia was th.e dis-
tillation of coal,55 but at present it is synthesized, using pressures of abottt 200
atmospheres and a temperature of 450° with a catalyst of iron and traces of
other metals. In nature it is formed by the decQmposition of protein, giving
the "albuminoid ammonia" and "free ammonia" of th~ water analyst.
These in turn are oxidized to nitrites and nitrates, forming the nitrogen cycle.
Qualitative detection is by the white fumes produced by strong hydrochloric
acid, and by the odor.
Quantitative determinliction: by (a) absorption in wl\.~er or standard acid and
subsequent titration, 'cochineal often being used- as an indicator, Table 9.
(b) In the case of tarry or highly colored solutions, an apparatus like the
nitrometer, p. 2418, may be used and the solution treated WIth sodium hypo-
bromite, when a molecule of nitrogen is evolvt:\.d for every 2 mols of free aminonia
present.
2NH3+3NaBrO=N2+3Na~r+3H20.
50 For accurate figures see Appendix, Table 3 . ' \
51 Cupramite,is a mixture of 1.5 parts 8-14 mesh pumice with 1 part CUS04, 5H 20;
45 cu. in. will protect against 2% ammonia for 5 hours.
52 Henderson and Haggard, "Noxious Gases." I
53 Refrigerating Engineering, 17, 65 (1929). J. Gasbeleuchtung, 57, 941 (1914).
54 Power, 72, 128 (1930). __;-:-
55 Ammonia is formed when very dilute nitric. acid acts upon zinc.
GAS ANALYSIS 2361
Or the gas may be strained through cotton, and passed into 10% sulfuric acid
in the familiar gas washing bottles, 5 to 10 cubic feet of gas being used.
The acid is transferred to an ammopia distillation apparatus, as that of
Kjeldahl, the contents made alkaline with sodium hydroxide, distilled into
standard acid and titrated with standard alkali. The distillation is continued
until the steam from the still does not change turmeric paper. (c) By Nessler's
solution. This is a solution of mercuric iodide in potassium iodide, made
strongly alkaline with potassium hydroxide which gives a yellow red coloration
~ith ammonium compounds: this is matched in tubes with colors produced by
standard solutions of ammonium chloride. The process is rapid and very
delicate, one part nitrogen as NHa being detected in 1 billion parts of water.
It is particularly used in water analysis.
Nitrous-Anhydride N 20a according to some authorities does not exist. It is
very unstable and behaves like a mixture of nitrogen dioxide and nitric oxide.
Nitrogen dioxide, Nitrogen peroxide, NO z,. or Nitrogen tetroxide, N 20 4
(cold).-Nitrogen dioxide is a reddish-brown acid gas heavier than air, of about
1.6 sp.gr. It is readily soluble in ,water forming nitric acid and nitric oxide,
H 20+3N0 2=2HNO a+NO (hot) and N204+HzO=HNOa+HN02 (cold). It
is prepared by the heating of lead nitrate 2Pb(NO a)2=4NO z+2PbO+0 2: also
by the action of concentrated nitric acid on copper or tin. At _9° it is a white
crystalline solid; at 26° a bright liquid; at 150° a deep brown gas.
On account of the high content of oxygen, nitrogen dioxide is an energetic
oxidizing agent, carbon monoxide and .phosphorus being oxidized by it. The
dry liquid gas mixed with benzene, C 6H6 or naphthalene C10HS forms a power-
fur explosive when detonated, the pl;oducts being wholly gaseous. It was used I
in depth bombs.
Physiological Action.-" Nitrous fumes" are intensely irritating to the
respiratory tract, and are the most insidious of the irritant gases. After in-
haling the fumes from the burning of dynamite, celluloid 56 or from action of
nitric acid on metals or sawdust, one may be little inconvenienced and appear
perfectly well. Before noon of the next day or even before the morning he
may be dead, from the volume of liquid coming from the dropsy of the lungs;
.011 to .015% is dangerous and twice this quantity rapidly fatal.
Qualitative detection-by means of iodo starch pap_er; or absorb in alkali
and test for nitrates. •
Quantitative determination-by absorption in standard potassium per-
manganate acidified with sulfuric acid.
Allison, Parker and Jones 51 determined small quantities of " nitrous fumes"
in ·the air by treating a known quantity of the gas with caustic soda solution,
evaporating, and treating the residue with phenol sulfonic acid after the pro-
cedure for nitrates in potable waters. ..
56 See Proceedings of a Board of the Chemical Warfare Service, etc. which investi-
gated the disaster at the Cleveland Hospital Clinic, Cleveland, Ohio, May 15, 1929.
67 Bur. Mines Tech. Paper 249 (1921).
2362 GAS ANALYSIS
tinned iron drums; 64 in 1926 Metzger 6y discovered that acid calcium cyanide
Ca(CNk 2HCN, gives off three-fourths of the acid in half a minut~ and the
whole of it in a minute. It is extremely sensitive to moisture: CaH 2(CN)
+2H 2 0= Ca(OHh+4HCN. It is known as "po,~dered hydrocyanic acid."
Physiological Action.-Cyanides or cyanogen artest oxidation in the tissues
and produce asphyxia. The person falls after a few breaths of the polluted
atmosphere and dies in 6 to 8 minutes. 0.012% is dangerous and 0.3% rapidly
fatal. Ammonia and chlorine are the antidotes. Like" nitrous fumes,"
it is evolved when celluloid burns.
Qualitative Detection and Quantitative Deterroination.-By absorp.tion in
water or standard alkali, and testing for a cyanide: or it may be absorbed in
standard silver nitrate, or converted to a ferricyanide. 66
Cyanogen compounds are formed whenever nitrogenous and carbonaceous
compounds cpme together at high temperatures, particularly if alkali be present.
They are ,found in blast furnace, coal and destructor gases.
Hydrogen Sulfide H 2S W5,S carefully investigated by Scheele, who synthe-
sized it from its elements. It occurs in the natural gas of certain Texas locali-
ties to the extent of 20%,67 and in " sulfur' springs." It is a colorless gas of well
known odor, a little heavier than air (sp.gr. 1.18), boiling at -62°: at 20° the
liquid gas exerts a pressure of about 150 lbs. per sq. in. (10 atmospheres). It is
readily decomposed-hence a good reducing agent; it is combustible, explosive
limits 4.3-46.0%, and soluble in water to the extent of 3 m!. per m!. of 'Yater
at 20°,68 alcohol dissolves about 7 volumes. The usual method of preparation
from ferrous sulfide gives a gas containing hydrogen (40% in some cases) from
the excess iron. For the preparation of the pure gas, a pure sulfide, as Stibnite,
should be employed.
Physiological Action.-Sulfuretted hydrogen is nearly as toxic as prussic
acid, and its action may be as rapid. It is a non-cumulative poison, is 'rapidly
oxidized by the blood and its products are non-toxic; 0.01-0.015%, may produce
death in from 8 to 48 hours.59 .05 to .07% is dangerous and 0.1 to 0.3 rapidly
fatal. 70 • /
Qualitative detection by " lead paper," i.e., filter paper moistened with a
solution of lead acetate giving a black or brown coloration.
Quantitative Determination.-It can b~ absorbed in a standard solution
of iodine in potassium iodide. Small quantities can be determined colorimetri-
cally by p-amino dimethyl aniline sulfate. 71
Hydrogen sulfide 'can be separated from carbon-dioxi\!e, hydrogen and
hydrocarbons by powdered manganese dioxide.
Sulfur Dioxide S02-was first prepared in the pure state by Priestley in
1775, by the action of sulfuric acid on mercury, as the result of the breaking of a
thermometer in distilling the acid. It is a colorless gas of suffocating odQr, a
64 Ind. Eng. Chern., 25, 959 (1933).
66 Id., 18, 161 (1926).
66 Rhodes, Ind. Eng. Chern., 4,652 (1912).
67 Bur. Mines Reports of Investigations, No. 3128 (1931).
68 The solution saturated at 0° contains 4.3 rnl. per rnl. of water, and used as a gargle
or spray, is almost a. specific cure for tonsilitis. ,
69 Yant and Sayers, J. Chern. Ed., 4, 613 (1927).-
70 Haggard. Chern. Abstr., 19, 2095 (1925). . - ' -
71 Sheppard and Hudson, Ind. Eng. Chern., Anal.-Ed. 2, 73 (1930).
GAS ANALYSIS 2365
little more than twice as heavy as air (sp.gr. 2.2), readily soluble in water, 1 ml.
dissolving 40 ml. at 20°,72 alcohol dissolves 32%. At this temperature the
liquid exerts a pressure of 48 Ibs. per sq. in. (3}i atmospheres). It boils at
_10°. It is an energetic reducing agent but a very stable gas. It is particu-
larly harmful to vegetation, especially conifers. Three to 4 pll-rts per million
injure the leaves of plants after one hour exposure. 73 Therefore it is recom-
mended that in works "here it is used, or in chemical works in general, that well
"authenticated, clear and dated photographs of the surroundings-particularly
with reference to arboreal or plant damage--be made from time to time.
Physiological Action.-It is said to cause decomposition of haemoglobin,
inflammation, and irritation of the respiratory tract may occur. Acute poison-
ing f~om the gas is rare, on account of the sensitivity of the eyes and throat to
it-it being the gas which is most easily perceptible. .0008% irritates the
'throat and .04% is dangerous.
Qualitative determination-by the bleaching of fuchsin paper or iodo-starch
paper .
.Quantitative determination-in standard iodine solution 74 S02+ 21 + 2H 20
= H 2S0 4 +2HI. This is the method gi ven on p. 2399. Sweeney, Outcault and
Withrow 75 advocate the use of .005 N permanganate; or by absorption in
standard alkali. 76 See also the report of the Selby Smelter, Commission Bur.
Mines. Bull. 98 (1915-).
As has' been stated, it is absorbed by zinc-filled rubber tubing. 77
Silicon fluoride or tetrafluoride was discovered by Scheele in 1771 and
investigated by Berzelius in 1823. It is a colorless gas fuming strongly in the
air, of odor resembling hydrochloric acid: It has a specific gravity of 3.6. It is
r'eadily decomposed by water, 'precipitating silicic hydroxide, 3SiF 4 +4H 20
=Si(OH)I+2H zSiF6 • It has been liquefied-the liquid is said to have a
tension of 50 Ibs. (3.4 ats.) at 20°. It is formed when hydrofluoric acid acts
upon silica or silicates:
heavier than air having a specific gravity of 1.18. It is very slightly soluble in
water, 100 ml. dissolving 2.6 m!. Alcohol dissolves about half its volume, and
ether twice its vqlume. Q'?lalitative detection-by the blackening of silver
nitrate solution in the absence of hydrogen sulfide and the other members of
this group.
Quantitative determination-by converting to phosphoric acid by burning,
after the manner of the determination of sulfur in city gas (p. 2395) and sub-
sequent oxidation with bromine and precipitation of the phosphoric acid as
magnesium phosphate in the usual way. Dennis and O'Brien 84 recommend
the determination in acetylene by absorption in sodium hypochlorite and ulti-
mate precipitation as magnesium phosphate: wt. Mg 2 P 20 7 XO.3059 =wt. PH a or
wt. Mg 2P 20 7 X 201.25 = m!. PHa.
Arsine, AsH a, was discovered by Scheele in 1775, who prepared it by heating
a solution of arsenic acid with zinc. It is evolved when acids act upon arseni-
cal metals. It is a colorless gas of disagreeable-odor, extremely poisonous,85
easily decomposed by heat and readily inflammable. It is 2.7 times as heavy
as ~ir; water dissolves 1/5 its volume of the gas, it is insoluble in alcohol and
ether; it liquefies at - 55°.
When inhaled it produces dizziness, faintness and headache, followed by
nausea and vomiting; it combines readily with hemoglobin, producing anemia.
In acute cases death ensues in from 2 to 6 days.
Qualitative detection-by passing into ammoniacal silver nitrate.
Quantitative determination-by the familiar Marsh test; also by the
Gutzeit 86 test: 2AsHa+12HgC12+3HzO=12HgCl+AszOa+12HC1; the calo-
mel can be weighed, or potassium iodide solution is added sufficient to dissolve
it; 25% excess iodine solution added, and the excess iodine titrated with N /100
thiosulfate: 1 mg. As 20 a =8.08 ml. N7100.thiosulfate. .
Stibine, SbH a, was prepared by Lewis Thompson in 1837; it is formed
whenever " nascent" hydrogen comes in contact with a soluble antimony s~lt.
It is a colorless gas, of odor faintly resembling hydrogen sulfide, inflammable,
easily decomposed by heat or even strong sulfuric acid. It is more than four
times as heavy as air, of boiling point -17°, and is soluble in water to the extent
of 1/5 its volume, 1 m!. alcohol dissolves 15 ml.
Determination-by passing into silver nitrate, forming SbAg a, washing and
dissolving in tartaric acid and determining the antimony.
Group IV-Gases soluble in potassium pyrogallate.
Oxygen, the active constituent of the air, was known to the Chinese in the
eighth century; they even knew it could be made by heating saltpeter: 87
84 Ind. Eng. Chem., 4, 834 (1912).
85 It is stated that Gehlen lost his life in 1815 by the inhalation of a single bubble of
the gas.
86 Barnes and Murray, Ind. Eng. Chem., Anal. Ed., 2, 29 (1930).
87 This fact is made use of periodically, to foist upon an unsuspecting public the
various "coal savers" and "aids to combustion," to "burn up the noxious gases of
combustion, and safeguard the health of the family." They do contain a little Chili
saltpeter, sometimes permanganate, but their chief constituent is usually common salt.
This colors the slightly visible flames of carbon monoxide, so that the combustion ap-
pears to be energetic, and the furnace very hot. Two pounds are sufficient to treat a
ton of coal. 'When it is remembered that about 2 pounds of oxygen are required to burn
a single pound of coal, it will readily be seen how much influence this compound will have
in burning a ton of coal! .
2368 GAS ANALYSIS
Mayow in 1669 found the percentage in the air, but'it remained for Scheele in
1771-2, and in 1774, Priestley, to prepare it in a pure state. They called it the
" vital spirit" or " fire air." It is prepared today,c.ommercially by the dis-
I
tillation of liquid air and the electrolysis of water. It is evolved by green
plants, particularly under the influence of sunlight, with the formation of
cellulose, 6C0 2+5H 20 = CSH1006+602. This is perhaps the most importjint
reaction taking place in the world-the carbonic dioxide evolved by combustion
and animals, being utilized by the plants for their sustenance, and they in turn
contribute oxygen for ours.
Oxygen is 10% heavier than air, and at ordinary temperatures a Ji.ter of
water dissolves 6.4 ml. of oxygen from the air, or 28.4 mI. from pure oxygen. It
dissolves to the extent of 10 volumes in molten silver, which when cooled sud-
denly evolves it, " the spitting of silver," causing losses sometimes in assaying.
It is colorless, odorless and tasteless. It was liquefied in 1877, becoming a steel
blue liquid, imU boiled at -182.5° (nitrogen -195.5°). It comes in commerce
in 110 and 220 cu. ft. cylinders under a pressure of 1800 lbs. per sq. in. at or-
dinary temperature.
If the atmosphere contains les!; than 17% oxygen oil lamps refuse to burn,
at 16% gas does not burn, at 12% respiration becomes deeper, at 10%,respira-
tion becomes more difficult; the lips become blue, 8% the face is leaden color,
and at 5 to 6%, panting and death ensue. 88
Compressed oxygen is one of the co~monest gases of today. The cylinders
are provided with right-hand threads, whereas all the combustible gas cylinders
have left-hand threads, and are served by oil lubricated pumps. "Never put
oxygen into a cylinder with a left-hand thread "-is one of the cardinal principles
to be observed in handling compressed gases. It is a " dangerous or hazardous
material," particularly in contact with oil, even in minute quantities, as ,little
as 0.6 mg. (1/50 drop) causing fire in the presence of high pressure oxygen. 89
Four drops of oil, in a 60 ml. bomb under a pressure of 2300 pounds causes an
explosion. 90 In a case which the writer investigated, 1800 pounds pressure
when exerted nearly 24 hours, blew a 220 ft. cylind~f of compressed oxygen to
pieces, caused the death of two persons, and wreckage of the laboratory; the
floor above was blown up with such violence, as to break the legs of persons
standing upon it; windows were broken two hundred feet away. The amount
of oil of course could not be determined, but other cylinders from the same lot
contained from 50 to lQO mgs. of mineral oil: they were._all intended for com-
bustible gases, but as they were new and supposedly never"used, they were
temporarily used to hold oxygen, a special fitting being made for the purpose.
Cases are known in which people have been burned, and iron fitting& set fire
by incautious oiling, as firewood is kindled by kerosene. Water is the only·
lubricant to be used on an oxygen compressor, with a trace of soap in it for the
fittings. Packings are of lead, asbestos or "nearly incombustible rubber."
Iron or steel should not be used for fittings for high pressure oxygen. A verdict
of $340,000 91 was awarded, due to a fire caused by oil in a steel manifold, even
88 Burrell and Oberiell, Bur. Mines Tech. Paper 122 (1915). I
89 News Edition of Ind. Eng. Chern., 2, 6 (Dec. 20,)924).
90 Hersey, C. A., 17, 3254 (1923). Also Brooks, 0;-"1\:., 18, 905 (1924).
91 News Edition, Ind. Eng. Chern., 2, 6 (Dec. 20. 1924).
GAS ANALYSIS 2369
though it was made in conformity with current practice. Brass or bronze
fittings must be used-these melt but do not fire or burn. 92
Used with compressed hydrocarbon gases or gasoline it finds extensive
employment in cutting and welding iron and steel; this has developed an ex-
tensive technique and literature. It is to be noted that much depends upon
the purity of the oxygen, both as regard the time consumed and the quality
of cut produced. If 99.3% of oxygen or better, be regarded as standard, using
hydrogen, a drop of a single per cent in purity increases the amount used by
25%, and the time by 7%; a drop to 95% purity makes a ragged cut, increases
the time by 40% and the quantity of gas used by 60%.93 The temperatures
attained are 1800° with manufactured or illuminating gas, 2100° with hydrogen
and 2450° with acetylene. 94
Quatitative detection-by the familiar glowing splinter or by the passage of
the gas into nearly colorless pyrogallate of potassium. This can be prepared by
using well boiled water to dissolve the hydroxide and acid and mixing them out
of access of air .
. Quantitative Determination.-In case combustible gases are absent, by
explosion with hydrogen (p. 2355); otherwise absorbents as phosphorus or
potassium or sodium pyrogallate is used; this with much oxygen may evolve
CO. See Reagents. Phosphorus does not absorb pure oxygen; if the con-
centration be reduced to 75% they combine with explosive violence. Burrell
uses a stabilized solution of chromo us chloride called" oxsorbent " with success;
another excellent absorbent when carbon monoxide and acetylene are absent, is
ammoniacal cuprous carbonate or chloride. This is prepared by treating bits
of copper or copper gauze, cleaned with nitric acid, with 500 ml. of a saturated
solution of ammonium carbonat'e mixed with an equal quantity of ammonium
hydroxide, sp.gr. of 0.96. Badger 95 recommends treating the copper with a
solution of equal parts of concentrated ammonia and water, saturated with
ammonium chloride; this absorbs from 50 to 60 times its volume of oxygen.
Oxygen should not contain more than 4% of hydrogen and preferably none
at all, and for respiration be of the highest quality containing not more than 2%
impurities.
For the determination of oxygen dissolved in water the method of L. W.
Winkler 96 is employed; the quantity of dissolved oxygen is usually a measure
of the purity of the water.
Ozone 0 3 , van Marum observed in 1785 that when frictional electric
machines wen~ operated, a peculiar odor was developed at the same time; a
similar odor was observed by the passage of a lightning spark. Little attention
was paid to this until SchOnbein brought it to the attention of chemists in
1840. He and others investigated it and found it not to be an oxide of hydrogen
as was thought, but a form of oxygen. The investigation was a difficult one
as the quantities of it formed were small and being so active it disappeared.
The odor resembles dilute chlorine; it is faintly blue in layers of a meter
thick, and about 60% heavier than air; it is one of the most powerful oxidizing
92 Jakowsky and Butzler, C. A., 17,3920 (1923).
93 Circular International Oxygen Co., 1911; also J. Soc. Chern. Ind., 30, 779 (1911).
94 Drews, "Verdichtete und Verfiuessigte Gase" (1928).
95 Ind. Eng. Chern., 12, 161 (1920).
96 Ind. Eng. Chern., 5, 976 (1913); also C. A., 7, 2266 (1913).
2370 GAS ANALYSIS
agents; 1 ml. of water at 12 dissolves 0.5. ml. ozone., It forms a blue liquid
0
0
boiling at -119 and is an endothermic compound. I
Qualitative Detection.-Ozone browns manganese chloride paper and
liberates iodine from potassium iodide. It does not affect permanganate as do
hydrogen peroxide and" nitrous fumes." It of coursJ does not respond to the
nitrite test with sulfanilic acid and alpha naphthylamin~ as do " nitrous fumes"
(N 20 a and NO), nor is it changed by chromic acid as is hydrogen peroxide . ..It
turns silver foil black which neither of the others affectsY
Quantitative Determination.-By the action upon a neutral solution of
potassium iodide, and titration of iodine with N /10 thiosulfate. 98 2KI +0 3
+H 20=2KOH+2I+0 2• If an acid solution be used, hydrogen peroxide is
formed, which sets free an additional quantity of iodine. For the rapid deter-
mination of ozone in the air of health resorts use may be made of the Wtlrster 99
test, the quantitative oxidation of fresh p-tetramethyl phenylene diamine
which forms a blue color on strips of paper which may be matched with
standards.
Ozone is formed when oils, as turpentine, o,xidize, or during the operation of
blast lamps. It is used industrially for, " killing odors," oxidizing andJ'leach-
ing oils, waxes, ivory, flour and starch; it has been used for purifying water, but
chlorine is said to be cheaper-that, however, produces a taste. Great hopes
were entertained for ozone when it was first made in quantity~ that it could be
used for improving the ventilation of h!_l>lls and schoolhouses; the fact that the
atmosphere was too dry for its efficient use was overlooked, and little effect
resulted other than the production of sore throats for those in the immediate
neighborhood of the machine. lOo In handling ozone, glass or metal apparatus
must be used as it oxidizes all other materials. Flexibility of glass apparatus
can be ensured by spirals or by the use of joints sealed with mercury cup,s.
iron rust, etc., issues with the hydrogen with such velocity through the jet as
to cause electric sparks, which ignite the hydrogen as lit comes in contact with
the air. Pails of water and wet rags are used to extinguish the flames from the
joints in the balloon filling pipes in the Zeppelin hanpar on Lake Constance.
The explosive limits with air are from 4.1 to 75%; if tl1e oxygen content be 5%
it is also explosive. I
Qualitative Detection.-In the absence of olefines (C 2H 4, etc.) and carbQn
monoxide, by the evolution of hydrochloric acid when hydrogen is passed ove~
cold solid palladium chloride. lo9
Quantitative Determination.-If there be no other combustible gases
present, several methods are available:
(a) Combustion with a hot platinum spiral in the apparatus of Dennis and
Hopkins.no .
(b) Combustion with copper oxide at a temperature of 250°-300° according to
Jaeger 111 or in any other similar apparatus.
(c) Explosion with oxygen, the converse of the determination of oxygen in air.
I
Where combustible gases are present the procedure of Campbell and Hart 112
using a 1 % solution of palladous chloride at 900: when fresh, 125 ml. will absorb
25-50 ml. of hydrogen in 90 minutes. Car,bon monoxide and dioxide, unsatu-
, rated hydrocarbons, and oxygen must be absent. The palladium precipitated
can be recovered and a fresh solution ,made therefrom. Platinum black or
palladium sponge or asbestos treated with either has not, in the exp~rience of
the writer, given satisfactory results.
Hydrogen has of late assumed great industrial importance, in the synthesis
of ammonia, the hydrogenation of oils, changing them to fats of any degree of
solidity, the hydrogenation or Berginization of coal and petroleum, and 'prepara-
, tion of hydrocarbons resembling petroleum products from it, synthetic motor
fuels and for balloons: in this latter 'connection a thousand cubic feet of pure
hydrogen will lift 75 pounds, of the commercial gas, about 71 pounds. It is
also used for cooling synchronous condensers, increa~ing their capacity 25%.113
It is made commercially by the electrolysis of water or brine, by passing over
hot iron, the Lane process, by passing water gas over lime at 500°. Hydrogen
is also made by passing natural gas and steam over a catalyst at 870°: the CO
found is removed by another catalyst.1l4 In this connection it should be noted
that attemp~s are occasi9nally made to decompose steam.. ~y the" waste heat
of the boiler furnace" and get ,hydrogen as a fuel, and oxygEm~as,a by-product.
Cast iron retorts are set behind the bridge-wall, steam conductec! through them
and this passes under the grate and up through the coal. There are several
difficulties in the way here: first, at the temperature of 1600 0 no noticeable
(not to say cansiderable) decomposition of steam takes place whereas the
melting point of the cast iron retorts is 1500°, or 100° lower; s~cond, when
109 Phillips, Am. Chern. J., 16, 259 (1894).
110 J. Am. Chern. Soc., 21, 398 (1899).
ill J. Gasbeleuchtung, 41, 764.
112 Am. Chern. J., 18,294 (1896).
113 Power, 68, 995. Science, 72, X (1930). Power, ~,_902 (1930). Easley, Power,
80, 362 (1936). _,-
114 Byrne, Gohr and Haslam, Ind. Eng. Chern., 24, 1129 (1932).
GAS ANALYSIS 2375
the mixture of hydrogen and oxygen (supposing it to have been formed) is
conducted from the retort, if the temperature drops below 1600° they recom-
bine making steam; third, if it were po13sible there would be no gain, the heat
evolved by the combustion of the gases (if formed) having been previously
abstracted from the system to produce any decomposition. A puzzling feature
was the fact that when the apparatus was first operated hydrogen was evolved.
This was of course due to the decomposition of the steam by the red hot
iron 3Fe+4H 20=4H 2 +Fe a04, and was overlooked.
Another alleged source of gain is that obtained in the so-called" water gas
process" of burning coal: Steam is used to produce draft and so cut down smoke,
and of course it is decomposed yielding, with the coal, hydrogen and carbon
monoxide.
Attention is called to the fact tJ:lat these are excellent fuels ready-formed
" made by waste heat," and the fact lost sight of that the reaction by which they
were produced abstracted heat, or cooled the furnace down, and that actually
there is no gain in the process. The thermal efficiency of hydrogen engines is
ve:t:y high (45-50%); they use about 21.2 cu. ft. per horsepower hour.ll5
Methane, CH 4, noted by Pliny, forms the fire damp of the coal mines and
some part of the fuel for the holy fires of the fire worshippers: when marshy
pools are stirred, the gas evolved is largely methane, and it is the chief con-
stituent of natural gas. It too is formed by the decomposition of organic
matter by bacteria, septic tank gas being rich in methane. Volta in 1776
observed that it required twice as much oxygen for its combustion as did
hydrogen, Berthollet showed it contained hydrogen and carbon, but Henry
in 1805 determined its composition.
It has been made from its elements; a common method of preparation is by
heating an acetate with an hydroxide" CH aCOONa+HONa=CH 4+Na 2 CO a ;
this does not always furnish pure methane. Another method is by the action
of a zinc-copper couple on methyl iodide and alcohol;116 care must be taken to
wash and chill out the iodide vapors. Another method of preparation is by
the action of water upon aluminum carbide, Al 4C a +12H 2 0 =4Al(OH)a+3CH 4.
Methane is a colorless, odorless gas, about half as heavy as air (sp.gr. 0.55),
very slightly soluble in water, 1 ml. dissolving 0.035 ml., and being about 10
times as soluble in alcohol. It burns with a slightly luminous flame of about
5-candlepower (burned at the rate of 5 cubic feet per hour). The explosive
limits with air are from 4.9-15.4%. Its boiling point is -164°. Cylinders of
the compressed gas are under about 1800 pounds pressure. It is a very stable
compound both as regards chemical reagents and heat.
Quantitative Determination.-Methane may be determined by 'explosion
with oxygen or air, as is hydrogen, the procedure being given on p. 2356.
Dennis 117 states that while the determination of hydrogen by the Jaeger copper
oxide method is very satisfactory, the determination of methane in the same
tube at a higher temperature cannot be recommended on aC90unt of the time
necessary for the oxidation of· the methane. He recommends therefore com-
bustion in the Dennis combustion pipette or grisoumeter.
Ethylene or Ethene, C 2 H 4 , was discovered by Becher and studied by four
Dutch chemists in 1795 who made it by dehydrating alcohol with sulfuric acid:
115 Power, 77, 90 (1933).
116 Gladstone and Tribe, J. Chern. Soc., 45, 154 (1884).
117 "Gas Analysis," 1929 edn., p. 163.
2376 GAS ANALYSIS
it gave an oily liquid (ethylene chloride) with chlorine and hence received the
mime olefiant (oil making) gas. Henry showed how it differed from methane
and gave it the name of " the heavy hydrocarbon" which still persists, being
applied to that por~ion of coal gas absorbecYby brorriine 'or fuming sulfuric acid
and also known as " illuminants." \
Ethylene is a colorless gas, of peculiar ethereal odor, slightly lighter tha.
air and soluble in 4 times its volume of water at 20°: alcohol dissolves about 10
times as much. Liquid ethylene is found in commerce in cylinders under a
pressure of 620 lbs. (42 atmospheres), and the boiling point is -104.3°. It m
readily inflammable, giving a flame of about 70 candle power (burned at the
rate of 5 cubic feet per hour). Mixed with 3 times its volume of oxygen)' it is
powerfully explosive: the explosive limits with air are from 3.2 to 34%.
The principal source of ethylene at pre~ent is the gases from the stills used
in the cracking of petroleum. During the great war it was made by passing
alcohol over clay heated to 500°-600°. Ethylene is used as an anesthetic,
mixed with carbon dioxide as a stimulant; oxygen was formerly used for this
purpose instead of carbon dioxide, but gave rise to serious explosions in cor-
pore.ll s It is used in a dilution of 1 part per million to color or " age" fruit;
one cubic foot per day introduced into a 5000 cubic foot room produces a
beautiful golden color in oranges-without, in some cases however, imparting
the flavor that goes with it.1l9
Quantitative Determination.-Ethylene is determined by absorption with
fuming sulfuric acid containing about 25% of the trioxide: Worst all 120 states
this may appreciably attack the saturated hydrocarbons if the time of absorp-
tion be more than 15 minutes. Inasmuch as the vapor tension of sulfuric
anhydride is higher than that of water, the gas should be passed into the potash
pipette to remove this vapor as well as any sulfur dioxide which may be formed.
Bromine water may equally well be employed, followed by treatment with
potash. In case of the determination of " illuminants " in ,coal or water gas,
bromine water is almost universally used, as it is more rapid: benzol and its
homologues while theoretically unattacked by bromine water are mechanically
dissolved. I
Acetylene (Ethine), C 2H 2 , was discovered and analyzed by Edmund Davy in
1836, who prepared it from the black residue obtained in making potassium
.from cream tartar and charcoal. Berthelot made a thorough investigation of
it in 1859. It is formed by the incomplete ,combustion of many gases and
volatile liquids: it was the first hydrocarbon to be syntheslzedJr_~ its elements,
by passing an electric arc between carbon poles in an atmosphere of hydrogen;
it may also be made by treating ethylene bromide with alcoholic potash, or as
now commercially prOduced by the action of water upon calcium carbide: 121
118 "Boston Herald," Jan. 24, 1931.
119 C. A., 23, 218 (1929).
120 J. Am. Chern. Soc., 21, 245 (1899). I
121 The early preparation of calcium carbide is interesting: Willson at Spray, N. C.,
was endeavoring to make metallic calcium by heating a mixture of tar and slaked lime
in the electric furnace. Instead of getting a bright beautiful metal, a clinker was ob-
tained, which so disgusted him that he threw it into the rive1'. This treatment was
promptly followed by the boiling up of the water, bubbles of gas ,being evolved, which
burned when ignited, with a brilliantly luminous flame. :..Borne of the clinker was sent
to E. N. Dickerson, a patent lawyer of New York, witlr'the-request to ascertain what it
and the evolved gas were. The answer was returned that the clinker was calcium. carbide
and the gas acetylene. See also Wyatt, Eng. and Min. J., 58, 556 (1894).
GAS ANALYSIS 2377
generators in which the carbide is dropped into the water are the preferable form.
The yield is about 5 cubic feet to.the pound or about 11,000 cubic feet per ton.
Acetyle'ne when pure has a pleasant ethereal odor; as ordinarily generated,
the odor is unpleasant due to compounds of sulfur and phosphorus: air contain-
ing 9% can be breathed without ill effect but this mixture is very explosive;
the poisonous properties which it is said to possess are due to the sulfur, phos-
phorus and arsenic compounds contained in it.122 It is a trifle lighter than air
and dissolves in an equal volume of water at 20°; alcohol dissolves 6 times as
much and it is quite insoluble in saturated brine. It is very soluble in acetone
(1 m!. at 15° dissolves 23 m!. and expands about 5%) and particularly under
pressure at 12 atmospheres (180 Ibs.) it holds 300 volumes. It burns in the air
with an intensely luminous flame of 240-250 candle power, at 5 cubic feet pEr
hour: the light approaches daylight the most closely of any gas light.
At a pressure of 903 Ibs. (63 atmospheres) and 10° acetylene condenses to a
liquid which solidifies at - 81 0, and which it was hoped could be shipped in
cylinders, and thus solve the problem' of the lighting of isolated houses, but
after serious explosions in Paris and Hartford, this was abandoned: the cause
of'the explosions could not be determined for lack of witnesses.l23 It may have
been due to the fact that acetylene being an endothermic compound (47,700
calories being absorbed when it is formed), this same amount of heat becomes
free when it is decomposed and expands the hydrogen left by the decomposition
(which is equal in volume to the acetylene originally present) producing a
pressure, which, coupled with that already existing in the cylinder, is sufficient
to burst it; a shock or the detonation of the acetylide of copper would be suffi-
cient to have initiated the explosion.
The explosive limits of acetylene and air vary, according to different ob-
servers, from 2.5 to 73,124 or 1.5 to 80.5%,125 with the maximum effect at about
7.8. The ignition temperature is 480°, the temperature of the oxyacetylene
flame is said to be the same as the electric arc (over 3000°) and is consequently
extensively employed in welding and cutting metals even under water!
Qualitative Test.-The reddish-brown coloration produced by contact with
ammoniacal cuprous chloride ..
Quantitative Determination.-By absorption in bromine water, or fuming
sulfuric acid, and subsequent treatment with caustic potash: when passed
through ammoniacal cuprous chloride solution a reddish-brown precipitate of
acetylide of copper is precipitated. This is explosive if allowed to dry. Conse-
quently it must always be kept moist. The precipitate is filtered off, washed
free of copper, and dissolved in hydrochloric acid: in case traces of the copper
acetylide remain undissolved, the filter is dried, burned, the ash treated with a
few drops of nitric acid to dissolve' the cODper oxide and added to the acid
solution above; the copper is here determined electrolytically or in any con-
122 Chern. Ind., 20, 53.
123 A cylinder of compressed acetylene at about 15 atmospheres (without acetone)
was not detonated by the fall of a 50D-pound pile driver weight 20 feet, nor by a rifle
bullet shot through it: when put upon a wood fire-as in a burning car wreck-it de-
tonated splendidly, and fairly promptly. When liquefied, it is explosive at 560 lb.
(38 atmospheres) and exerts, when exploded, a pressure of 40 tons per sq. in. (5500
atmospheres). Wolff, Z. angew. Chern., 11, 919 (1898).
124 Burrell and Oberfell, Bur. Mines Tech. Paper, 112.
12S International Critical Tables.
2378 GAS ANALYSIS
Mid-continent and Gulf regions and al~o Canada, to the extent of 1.84%.
Recently a gas containing about 7% has been discqvered. Before the war it
cost $1300 per cubic foot, but at present about $7 per thousand cu. ft.: 135 it is
obtained from the natural gas by compression and cooling. It is present in air
to the extent of 1 part in 250,000, and has been o~tained from liquid air by
fractionation. It is also evolved from an aqueous solution of radium bromide.
It is twice as heavy as hydrogen, much less soluble in water, 1 mI. at 20°
dissolving 0.0138 mI.-the most insoluble gas, and boils at about -268.5°.
It is detected by the spectrum, the double Da line being very characteristic,
along with seven others. It is used as balloon filling, for the yellow vacuum'
signs and for lamp filling. Mixed with 1/5 oxygen it is used on account of its
slight solubility in the blood to enable divers to come back to normal pressure.
Neon, Ne (Neon the new gas), occurs in the air to the extent of 1-2 parts per
100,000, and also in the gases from certain springs. It is prepared iby the dis-
tillation of the crude argon from liquid air. It is 10 times as heavy as hydrogen,
boils at -245.9° and is distinguished by the brilliant red spectrum. This finds
extensive industrial use for signs but the gas leaks out eventually through the
seals of the wires in the tubes. The blue signs are a mixture of mercury vapor
and neon, the green are the same as the blue using an amber tube. "
Argon, A (Argon idle).-Lord Rayleigh, observed that a liter of atmospheric
nitrogen weighed 7 'milligrams more than chemical nitrogen, showing heavy
gases in the air: Cavendish in 1785 had performed a similar experiment with a
like result. Rayleigh and Ramsay repeated the experiment, absorbing oxygen
with not copper, nitrogen with hot magnesium, introducing oxygen, sparking
the residue, absorbing the nitrous fumes in alkali, and removing oxygen, the
gases being circulated for about a week. About 1 % of a gas was obtained,
having a density 20 times that of hydrogen and an orange spectrum: it forms no
compounds, whence its name. The density of the pure gas is 19.0 and the boil-
. ing point -186.1°.
Krypton, Kr (Krypton hidden), occurs in the air to the extent of 1 part in 20
million: its boiling point is -151.7°, density 41.5, a,nd gives a yellowish-green
spectrum.
Xenon, X (Xenon stranger), is the least abundant gas, occurring in the air to
'the extent of 1 part in 170 million: its boiling point is -106.9°, density 65.3 and
gives a blue-green spectrum. These last four gases are obtained by the careful
fractionation of liquid air. I
It is only within comparatively recent times 'that; Gas Analysis has assumed
great importance. The reasons are that the substances with ,vhich it deals are
so intangible, the apparatus is complicated and fragile, and until lately, com-
petition has not compelled manufacturers to seek .every possible source of loss. '
136 Ind. Eng. Chern., t4, 1435 (1932).
GAS ANALYSIS 2381
Here the object is to keep the carbonic acid (C0 2 ) as high as possible, and to
avoid the formation of carbon monoxide: in large plants every additional,
per cent of carbonic acid means the saving of tons of coal. Savings of 20 to 33%
by the use of gas analysis alone have frequently come to the writer's notice.
A satisfactory procedure is to post in the fire-room the percentage of carbonic
a~id obtained by each stoker, and stimulate a rivalry among the men-a bonus
in the pay envelope is also effective. The determinations to be made are:
Analysis of Chimney Gases.-Determination of carbon dioxide, oxygen,
carbon monoxide, nitrogen, and in$ome cases hydrocarbons. For this purpose
the Orsat apparatus is widely employed: the hydrocarbons may be determined
by the Burrell apparatus. These need not be sought unless the carbon mon-
oxide is more than 1 %.
Usually a few determinations of carbonic acid will suffice, but for regular
work the installation of some form of registering carbonic acid indicator is ad-
visable.
Carbonic Acid Indicators. 13 6.-lhese usually depend upon the principle of
collecting 100 ml. of the gas, causing it to pass through a suitable absorber and
collecting the residue in a bell which floats to a greater or less height according
to the residual volume. The fluctuations of this bell are recorded after the usual
manner of self-registering barometers or thermometers: the usual time for this
analysis and record is five minutes.
By modifying this apparatus slightly, it can be applied to the determination
of any absorbable gas as, for example, sulfurous acid or chlorine. It has been
adapted to carbon monoxide absorption, but it is not usual or easy.
Haber 137 employs the refractive index of gases to determine the amount of
carbonic acid in chimney gas; it gives results within half of 1%; 138 it has also
been applied to other gaseous mixtures. The instrument is called the Inter-
ferometer or Gas Refractometer and is made by Zeiss of Jena.139
Another method for the determination of carbonic acid makes use of the
heat cond~ctivity of the'different gases; carbonic acid has but about 60% of
the heat conductivity of air. Consequently a wire will be hotter in it than in
air, all other conditions being the same. This increase in temperature' of the
wire increases its electrical resistance, from which the per cent of carbonic
:lei!! call hI! determined. The descriptioll woujd take UR beyond the lilllit,," of
136 ::lee Bureau of Minell Bulletin No. 91, "Inlltruments for Recording Carbon Dio:dde
in Flue Gases." Also Power, 51,784 and 1058.
137 Z. angew. Chem., 19, 1418 (1906); ibid., 23,.1393 (1910).
138 Mohr, ibid., 25, 1313 (1912).
139 Bureau of Mines Technical Paper 185, "Use of the Interferometer in Gas Analysis,"
1918.
2382 GAS ANA~YSIS
this volume; according to Palmer and Weaver 140 it is by far the best method
for the determination, being accurate, sensit~'ve, easily operated and of moderate
cost.
The " Ranarex" recorder,141 an English instrument, consists of two fans
rotating each in its own chamber by the same electric motor: these fans cause
discs opposite them to rotate, one in air, and the other in the chimney gas.
Different torques are produced, depending upon the density of the gas in which
the discs rotate: this difference is transmitted by levers to a pointer which
indicates the per cent of carbonic acid.
The Determination of Temperature.-This is done by inserting a ther-
mometer, mounted in a metal tube, on the chimney side of the gas sampling
tube. Care should be taken to protect it from radiation to and from the walls
of the flue. They resemble those used for determining steam temperatures or
for "running" varnish. It should register to 360° and, under certain cir-
cumstances, one showing 550° may be desirable. ,A chemical thermometer
with long stem may also be employed; it should never be inserted naked into
the flue-as a sudden hot blast may break it-but always in a tube of cotton-
seed oil or sea sand.142 These thermometers should be tested for accuracy by
comparison with a standard, in a carefully stirred oil bath. The standard
should be kept exclusively for the purpose and be allowed to stay in the bath
until cool. Sudden cooling of a thermometer changes the zero point. The
standard can be certified by the makers or the United States Bureau of
Standards.
Electric pyrometers are also of course available for these measurements.
An error of five degrees (5°) in the reading of the thermometer affects the final
result by about 20 calories.
In case none of these appliances be at hand, the maximum 'temperature can
be determined by utilizing the melting-points of certain pure salts or metals;
as tin 232°, bismuth 270°, cadmium 302°, lead 327°, zinc 419°, cadmium chloride
541°, antimony 630°, etc. These can be suspended in the chimney in srn"all
covered cast-iron boxes. "
Composition of the Coal.-This is determined by the usual methods of
organic combustion and is required only for very accurate work.
Calculation: '
a. Heat passing up chimney;
b. POl,lnds of air per pound of coal.
(a) Heat Passin'g up Chimney.-The accurate- calculation resolves it./Jelf
into finding what 'volume of gas of the compositi.gn determined by analysis
would be produced by a kilo of the coal used, the analysis of which is known.
The temperature of the escaping chimney gases being also known, and their
specific heat, the quantity of. heat they carry off can be calculated: this· divided
by the calorific power of the coal gives' the per cent of heat lost in the chimney
gases. The calculation is rather long and will be found in detail in·the author's
book. l43
140 Bur. Stds. Tech. Paper 249; also Ind. Eng. Cp.em., 12, 359 and 894 (1919).
141 Grant, Industrial Chemist, 2, 435 (1926). . "
142 With rouruled grains, not river sand~as it would make scratches.
143 "Gas and F~el Analysis for Engineers," Wiley.
GAS ANALYSIS 2383
Finally, for very rapid work, Bunte's Fig. 368 may be used. The results
are within 2% for about 12% of carbonic acid. It is used by noting the point
where the diagonal line representing CO 2 cuts the ordinate of temperature--
the abscissa corresponding to this point represents the per cent loss.
144 Power, 30, 1121 (1909).
145 J. f. Gasbeleuchtung, 43, 637 (1900).
2384 GAS ANALYSIS
The following table shows roughly the excess' of air, and per cent of heat lost
in the chimney gases, their temperature being 51~0 F.
I
Per cent CO 2 • •• 2 3 4 5 6 7 8 91 10 11 12 13 14· 15
Vol. air more
than theory
=1.0 ....... 9.5 6.3 4.7 3.8 3.2 2.7 2.4 2.1 1.9 1.7 1.6 1.5 1.4 1.3
Per cent' loss of
heat", ..... 90 60 45 36 30 26 23 20 18 16 15 14 13 lZ
~
-+--+--1300 '<C
50 40 30 20 I 10 o
Per Cent Loss
FIG. 368.
Lockwood 147 has shown how to calculate the excess of' air without the use
of at<?mic weights or other constants, from re.cently de~eloped formulae,
, 146 Already eited. -'-
147 Power, 58, 461 (1923). J. Soc. Automotive Eng., 12, 299 (1923).
GAS ANALYSIS 2385
Theoretical Air per Pound of Combustible.
18.3N 2+34.8CO - 69.60 2
(1)
O.52qN 2+ 5CO +4C0 2- 20 2 •
Substitute per cents of N 2, CO 2, etc., as found by the Orsat apparatus.
This formula as well as those including (9) applies to all fuels, flolid, liquid
and gaseous, except producer and blast furnace gases, where the amounts of
nitrogen and sulfur are negligible, and no free hydrogen or combustible but CO
is in the flue gas.
Actual Air Used per Pound of Combustible .
, .l8.3N 2
(2)
Theoretical Air
Vol. Matter per Pound of
Fuel % Combustible
(b) Equate the expressions for theoretical air from the gas analysis formula
(1) and from the fuel analysis.
(3)
Table B shows the different values for these constants for different fuels.
Table B
CONSTANTS TO CHECK ORSAT ANALYSIS
Fuel n k
\. 4.77N z -9CO-18(C02+02)
Water vapor per pound of combusttble= O.525N 2+5CO+4C02- 202· (6)
72
70
71 %
~
69
68
67
B%
6 56 .Leoz
6 5
63 6
6
9.~
61
6
59 -~
5B lOj'b
57 CO2
56 6
5 4
5 1%
/ CO 2
53 3
6
5 2%
5
51 . , CO2
-{. / 6 "%~
4 3 5 _'c6~
4
47 6
4 / 14r-
46 2 7 5 / _., CO2
45 3 6
c '%--
4
43
4
5 6
O'_'6%-
42 CO 2
2
41
3.
3
• 5 7 6
A 5 6
3 4
~7---(i 2
5
3 --, L 3 ~ st.-
2 4
3' I 3 5
33
32
1
2 • 7
..
3
31 -/ ~
~
.~
1 2 3
:~ 2
3
7 •
2 1 --,I 3
26 ..p 7 /' 3
2
7 --, ,-1 2 7 3 ~
2 3
2-3-CO 1
22 7 ! ~..p _"i' 7
!FI-co'
2
19
/--8
-0,
CO
!
1
2
:rv==
18 1
I
16 ~~z ;~ ....... I 1
,-'l'-
I I 7 1
c 2
2. I
14
13 0,--'-' ~ 1
I
~6 ~-*
1 CQ
,,11 16,2.' :-:r-..
Cor-'! ...l-_l_B;
9
8
7
02 CO
h 9
CO ,_~g~
6
2
r. I
7 8 9 10 11 12 13 14 15- :r6_17 18 19 20
Per Cent COz in Flue Gas
FIG. 369. Brownlie "Combustion," April, 1923.
GAS ANALYSIS. 238 7
. . [ 6(C0 2+CO) ]
Percentage of carbon 1-n combusttble = 0.525N 2+5CO+4C0 -20 X 100. (7)
" . 2 2
Percentage of hydrogen in combustible
= [0.525N2-CO-2(C02+02) ] X 100 (8)
0.525-N 2 +5CO+4C0 2 -20 2 •
. . . 6(C0 2+CO)
Ratto of carbon to hydrogen w combusttble = 0.525N 2_ CO _ 2(C0 + O ) (9)
2 2
o
o
COMBUSTION CHARACTERISTICS
From Gas Analysis
E,,:-Orsa, analysis equals 6.4%CO, and 9.6%0,
Connect CO 2 and O 2 with straight edge, Max. CO 2 equals lL8 %
Nitrogen equals 84", %; then follow down Nittogen line of 84 and
read 76 % of Excess Air. Then pivot straight edge at Max. C02
o
till it touches turning point and read 1100 B.T.U; heating value
of gas being burned, and 9.7 cubic feet of air required •
•
FIG. 370. Marx's Chart.
2388 GAS ANALYSIS
For gas, Marx's Chart, Fig. 370, is usefu1:148 Froin the Orsat Analysis it
shows the maximum CO 2, excess air, the B.t.u. in the gJs and the number of
cubic feet of air required for combustion. I
Smoke.-For the determination of the amount of smoke in the chimney
gases, use may be made of the Ringelmann smoke scale. I This consists 149 of a
series of rectangles %: in. X Y2 in. filled with cross-hatching lines a Isreater or
less distance apart, with which the density of the smoke Can be compared.
Or the Eddy smoke recorder may be employed; this consists of a tube Of stand-
ard length through which the smoke gases are drawn. A standard electric
light is fixed at one end of the tube and viewed through the smoke; its density
is measured by the extent to which the light is obscured and by a photoelectric
cell.
Here the object is the reverse of that in, the chimney gases, to keep the
percentage of carbon monoxide as high as possible and, for gas-engine purposes;
the per cent of hydrogen constant.
The determinations made are the same as in chimney gas-C0 2 , 0, CO/50 N,
and oftentimes hydrogen and hydrocarbons; the quantity of dust is sometimes
important. The heating value is determined as in city gas, p. 2391. The
efficiency of conversion would be found by measuring the number of cubic feet
of gas made per ton of coal gasified; the calorific power of each' (gas and coal)
being known, their quotient represents the efficiency. The heat contained in
the gas due to its sensible heat, found after the manner of calculating the loss
in chimney gases (i.e., volume gas X weight X rise of temperature X specific heat),
is to be added to this for accurate work,
As showing producer gas practice, the following typical analyses are cited:
P. F. Marx, The Refiner and Natural Gas Manufacturer, 10, 121 (1931).
148
Power, 40, 66; also Carter, Power, 74, 678 (1931).
149
150 Tech. Paper 106, Bureau of Mines, "Asphyxiation from Blast Furnace Gas."
151 With 38.7% H 20, 3.2% ash. .
152 From "Gas Producers and Producer Gas Power Plants," R. D. Wood & Co., 1906.
153 Richards, J. W., J. Frank. r_'st., 415 (1900),..J].oofed from V. Ihering, "Gas
Maschinen." . .
GAS ANALYSIS 2389
Gas from Different Kinds of Producers
. Down Up Pressure,
Bla.st Fur- Siemens 1&7
Draft 156 Suction 154 Tay- Mond 154
Draft 156 101" 1£.4, 155 nace 157
Inlet Water
rhermometer
Outret Water
Thermomet ,
To Drain
Wate, Inlet
t 1t
Combustion
_f.~aE~~L
Inlet Water
Valve
FIG. 371.
GAS ANALYSIS 2391
(a) Candle Power.-This can be very satisfactorily found using a 60-in.
open-bar photometer and Leeson contrast disc. The gas should be burned from
a burner commercially obtainable which gives the highest candle power; for
gas from 14 to 21 candle power, Sugg's London argand burner, sizes C to F,
should be used; for richer gases, Sugg's table top or the Bray slit burner. For a
standard of comparison, the sperm candle is
convenient, satisfactory, and very' extensively
used: the Elliott kerosene and Hefner amyl
acetate lamps are also employed.
For accurate work the Lummer-Brodhun
disc and electric standards, or the Hefn.er lamp
should be used. For the determination of can-
dle power, reference may be had to Circular
No. 48 of the Bureau of Standards on "Stand-
ard Methods of Gas Testing," 1914, or Stone,
" Practical Testing of Gas and Gas Meters."
C:j,rburetted water gas shows from 20-28
candle power, coal gas 14-20, oil gas 45--60, oil-
air gas 30-35, gasoline 12-17, acetylene 17(}--200.
Inasmuch as 75% of the gas now made is
used for its heating power (mantle burners), the
candle power is infrequently determined.
(b) Calorific Power. (a) Direct Determi-
nation.-Thisis often determined by the Junkers
calorimeter, although others in use are the
Sargent, Doherty, and in England the Boys and
Simmance-Abady.
The Sargent calorimeter is shown in Fig.
372 and in section in Fig. 371. It consists of a
combustion chamber surrounded by a water
jacket, this being traversed by a great many
tubes. To prevent loss by radiation this water
jacket is surrounded by a closed annular air FIG. 372.
space, in which the air cannot, circulate. The
whole apparatus is made of copper as thin as is compatible with strength.
The water enters the apparatus at the top (water inlet), passes by a ther-
mometer and through the inlet valve, circulates around the apparatus, and
passes out by a thermometer to the measuring or weighing vessel: this can be
shunted in and out by electrical connections at the meter.
There is not only a very large surface of thin copper between the gases and
the water, but the two move in opposite directions, during which process all the
heat generated by the flame is transferred to the water, and the ~aste gases
leave the apparatus approximately at 'atmospheric temperature. The gas to be
burned is first passed, to insure constant pressure, through a pressure-regulator
and then through a meter, Fig. 373. The source of heat in relation to the unit
of heat is thus rendered stationary; and in order to make the absorbing quantity
of heat also stationary, two overflows are provided at the calorimeter, making
2392 GAS ANALYSISi
the head of water and overflow constant. The temperatures of the water enter_
ing and leaving the apparatuEl can be read by the thermometers as shown
before; the quantities of .heat and water passed through the apparatus are
constant. As soon as the flame is lighted, the m~rcury ip the out-going ther-
mometer will rise to a certain point and will remain nearly constant.
Manipulation.-The calorimeter is placed as shown in Fig. 373 so that one
operator can simultaneously observe the two thermometers of the entering
and escaping water, the index of the gas meter; and the mea swing or we~ghing
apparatus. '
No draft of air must be permitted to strike the exhaust. of the spent gas.
The- water-supply tube is connected with the nipple in the center of the
upper container; the other nipple is provided with Ii wa)lte tube to carry away
the .overflow, which latter must be kept running while the readings are taken.
The nipple, through which the heated-'Witer leaves the calorimeter, is
GAS ANALYSIS 2393
connected by a rubber tube with the graduate or a weighed can. The con-
pensed water of the apparatus is caught in a small graduate .
. The thermometers being held in' position by rubber stoppers and the water
turned until it discharges, no water .must issue from ·the drain tube for con-
densate, Fig. 371, as this would indicate a leak in the calorimeter.
The inlet water valve is now set to allow about two liters of water to pass in
a minute and a half, and regulate the gas to flow at the rate of 4 to 7 cu. ft. per
hour as may be found to give the highest result with the gas tested. Adjust
the air so that the..f1ame shows a slight luminous tip, insert the burner as far as
the bracket permits and 09serve the condition of the flame in the mirror. The
excess of air should be as'low as possible, and is controlled by the damper in
the exhaust port: start with t:he damper closed, then open slightly, observing
the outlet thermometer wherrthis reads at a maximum, i.e., when the greatest
heat is given up to the water the damper is in about the correct position.
The flow of water should be regulated so that the temperature of the out-
going water and gas exhaust are the same or differ by no more than 3° C.: the
difference between the in-coming and out-going water should be about 10° C.
The temperature of the inlet water is noted, the reading of the gas meter taken,
and at this same time the outlet tube changed from the funnel to the graduate.
Ten successive readings of the out-flowing water are taken while the graduate
(2-liter) is being filled and the gas shut off.
A better procedure is to collect about 10 liters of water and weigh it. The
thermometer in the outlet can then be read every half minute. For the pro-
cedure of the Bureau of Standards, see pp. 2431-2433.
Calculations.-The volume of gas, V, is determined by the difference in the
m~ter readings: this must be reduced to operating conditions, i.e., 60° F., 30
inches and saturated with moisture. The height of the barometric column
must be reduced to 0°: this is done by observing its temperature and referring
to the proper table in one of the usual" Handbooks of Physical Constants,"
and noting also whether the scale is brass or glass. The certification correction
is also to be applied. The pressure of the gas in inches of water at the meter is
changed to inches of mercury (1 in. =0.07 mm. mercury) and added to the
corrected barometric reading; this represents the total gas pressure: its tem-
perature is given by the meter thermomete.r. Table 9 will then give a means
of reducing the volume.of gas shown by the meter to operating conditions.
To determine the temperature rise, T, subtract the temperature rise of the
inlet water from that of the outlet water: the corrections should have been
applied to these thermometer readings and the correction for stem exposure,
Table 10.'62 The amount of water, W, heated is determined by weighing, the
weights having been verified.
The observed heating value is
WXT
VXF·
F is the reduction factor obtained from Table 9. This heating value is cor-
rected for the effect of atmospheric humidity, Table 11; 163 this is additive to
.162 Bureau of Standards, Circular No. 48.
163 Circular Dept. Public Utilities of Massachusetts, Gas Inspection Division.
2394 GAS ANALYSIS
90% humidity. The heating value is further corrected for the loss of heat
from the calorimeter which under the usual operating conditions is 1 B.t.u.
Where the water temperature is quite different from room temperature, a
correction. must be applied, additive if the room is. warmer, and subtractive if
colder. For calculating total heating value, this' ~orrection amounts to 0.5
B.t.u. per degree difference F,at a room temperature of 50° F. and increases
by 0.1 B.t.u. for each increment of 10° F. room temperature: for calcula.ting the
net heating value it is 0.4 B.t.u. for 50 o-8ao F. room temperature and 0.5
B.t.u. for 90°-100° F. The average of the different runs, which should agree
within 1 %, corrected for humidity and heat loss, is the total heating value of the
gas, and is the one to -be reported. The net heating value is obtained by sub-
tracting the product of the volume of the condensate in ml. X2.3 B.t.u. (0.6° C.)
from the observed heating value and correcting for heat loss.
Page 8d from Circular 48, Bureau of Standards, shows how the record is to
be kept. (See p. 2433.)
Thomas Recording Meter.-In many gas works this apparatus is'in opera-
tion and agrees satisfactorily with the Junkers instrument. 164
NOTEs.-After setting up the apparatus the first thing to be done is to turn on the
water-(not the gas). Similarly, the water should be shut off last. All connections and
the meter should be tested for leaks before each test. The water level in the meter
should be checked daily.' Slight drafts caused by moving suddenly near the apparatus
will vary outlet readings and vitiate the test, The instrument should not be set up near
a window or heating apparatus where radiant heat might affect the readings.
If 0.2 cu. ft. of gas is burned, then an error of 0.1° F. in temperature of water means
an error of 4 B.t.u.; an error of 0.01 lb. water, 0.9 B.t.u.; 1° F. in gas temperature, 1.8
B.t.u.; 0.1 in. (barometer), 2 B.t.u.; 1 in. water pressure of gas, 1.5 B.t.U. 165
The total heating value of the gas is the heat given off when the gas is used for heating
water or in any operation where the products of combustion pass off below 100° C. The
net heating value represents the conditions in which by far the greater quantity qf gas is
consumed, for cooking, heating and gas engines, and is one which should be reported.
It should, howeyer, be corrected,!s6 to the legal cubic foot, that is; measured at 30 ins.
barometric pressure, and 60° F. saturated with moisture.
. The chief sources o~ error a~e,167 in adjusting the meter; in measur~ng the temperature-
nse of the water, and m changmg over the outflo~ wat~r to the welghed vessels.
which is the gross heating value of the dry gas. The correction for the heat
lost is found, as with'liliimney gases, by multiplying the volume of the com-
bustion gases by their weight X specific heat Xrise of temperature.
(c) Sulfur.-Sulfur, being present in gas in so many forms, is determined by
combustion and usually reported in grains of sulfur per hundr~d
cul)ic feet.
One of the most easily portable and satisfactory forms is that of
Hinman and Jenkins described as follows: 169 The upper vessel, Fig.
374, is a " bead glass" 300 mm. long and 60 mm. in diameter; this
is filled with large cut-glass beads, held up by a suitable fluted glass,
giving a large condensing surffl,ce without obstructing the draft.
To this bead glass is attached, by a rubber connector, the adapter,
410 mm. long and 50 mm. lower internal diameter. To the upper
adapter is attached, by means of the" connecting piece," the lower
adapter, 400 mm. long and 40 mm. lower diameter. The connect-
ing piece projects 12 mm. above the top of a rubber stopper, fitting
the. upper adapter, and is surmounted by a watch-glass deflector
carried on platinum wires. An overflow tube carries the conden-
sation to the Erlenmeyer flask hung on the stopper as shown; this
tube is so adjusted that some liquid remains on the stopper to keep
it cool and to absorb some of the ascending gases. The Bunsen
burner is fitted with a lava tip having a 5-mm. hole; surrounding the
burner is a glass tube 20 mm. in diameter, forming the inner wall of'
an annular chamber, of which the outer wall is a glass ring 50 mm.
in diameter. Into this chamber, which serves to contain 10% am- FIG. 374.
,monium hydroxide, the lower adapter dips 10 mm.
The lower adapter is joined to the" connecting piece" by a short cork-lined
metal tube. AlthOugh radically different in form, this apparatus is very similar
to the Referees' in general principle and in ~ethod of use, the principal differ-
ence being the use of ammonium hydroxide instead of dry ammonium carbonate
as a source of ammonia. About 10 m!. of concentrated ammonium hydroxide
is placed in the reservoir about the burner at the beginning of the test and about
5 m!. more added every fifteen or twenty minutes. The gas is consumed at
the rate of 0.4 to 0.6 cu. ft. per hour, and 2.5 to 3 ft. burn~d, if the sulfur is to
be estimated gravimetrically; otherwise 1 cu. ft. is enough. When the run is
coinpleted the apparatus is allowed to cool and is then flushed four times by
pouring 50 ml. portions of water in at the top of the bead tube. To the solutions
and washings are added 2-3 m!. bromine water, and it is evaporated to 30 or
40 m!.; an excess of a hydrochloric acid solution of barium chromate is added
to the hot solution, it is gently boiled, an excess of dilute ammonia added, again
boiled for a minute, filtered and washed. The ammonium chromate in the
filtrate (the chromic acid being equivalent to the sulfuric acid in the original
solution) after being boiled in a stout flask, with a Bunsen valve, to expel the
air, is cooled and titrated directly with stannous chloride (3.25 grams Sn per
Vter) using starch ~nd potassium iodide to accentuate the end-point.
169 Jenkins, J, Am. Chern. Soc" 28, 543 (1906); also Technologic Paper No. 20,
Bureau of Standards, McBride and Weaver, "Determination of Sulfur in Illuminating
Gas," 1913, also Stone, book cited.
2396 GAS ANALYSIS
the gas contains hydrogen sulfide, it must be absorbed by passage over man-
ganese dioxide. It can also be determined by the usual absorption method.
(g) Specific G~avity.177-The readiest method depends upon the time of
I
- -- - - -- -- -- - - -
Coal. ........ 1.6 4.0 0.4 8.5 49.8 29.5 3.2 3.2 16.1 622
Carbo water... 3.0 13.3 0.4 30.4 37.7 10.0 3.2 2.1 22.1 643
Blue water ... 3.4 0.0 0.9 40.9 50.8 0.2 O. 3.5 299
'0' .
Pintsch ...... 0.2 30.0 O. 0.1 13.2 45.0 9.0 1.6 43.0 1276
Blau ......... O. 51.9 O. 0.1 2.7 44.1 O. 1.2 48.2 1704
Oil-water ..... 2.6 7.0 0.2 9.2 39.8 34.6 .. ,- 6.6 19.7 680
OiL ......... 0.3 31.3 O. 2.4 13.5 46.5 3.9 1.1 38.0 1320
Gasoline ..... .. U 18.5 .. , . '" . C6HI4.~ 10.3 69.7 16.0 514
Acetylene ... , .. 96.0 0.8 .... .. .. ., .. .. " 3.2 225.0 13.~0··
H 2S
NaturaL ..... 0.3 0.3 0.3 0.5 2.3 92.6 .... 3.5 ().2 840-1174
0.4 ... . .. .. '" . . ... 87.0 9.6 3.0 .... 1040
efflux of the gas compared ~ith air; sp.gr'=AG2 • G and A represent the times
2 j
of efflux in seconds of gas and air. The apparatus is that of Schilling and is
obtainable from the dealers, or may be constructed according to Jenkins 179, 180
as follows: '\
" It consists, Fig. 376, of two large rubber stoppers, each having a brass tube,
projecting laterally near the large end, and connecting with the hole in the
stopper. A glass piece A in the form of a truncated cone fits tightly over one
stopper; it is 9 ins. long, lYz ins. diameter at the base and 1 in. at the top.
A similarly shaped piece B 9 ins. long by lYz ins. diameter at the lower end fits
over the second stopper; 2 ins. above the latter the tube has a constriction 1 in.
_in diameter, and at its upper part is narrowed to a neck 5/16 in. in diameter
which is ground on the inside to receive the end of a tube 7Yz ins. long aiid ~ in.
in diameter, in the upper end of which is fitted a platinum plate containing the
emission orifice. One and three-fourths inches below this plate is a three-way
glass stopcock, and 3 ins. below the latter a scratch surrounds the tube and
serves as the upper mark in the escape of the gas.
" Fitted into the hole in the stopper is a hollow cylinder of brass to which is
soldered a curved piece of brass wire pointed at the end, which rises lYz ins.
above the surface of the stopper. The tlVo brass tubes projecting from the
outside of the stoppers are joined by a piece of rubber tubing 15 to 18 ins. long.
" In using this instrument the larger tube B is filled with water, of the tem-
perature of the room, nearly to the top, the stopcock being turned so that egress
of air from the smaller tube is prevented. The larger tube is placed on an
elevated surface just high enough so that its bottom is above the level of the
scratch on the narrow outlet tube, the cock is turned so that the air may escape
'through the orifice in the platinum plate, and on the second, when the point of
the bra,'ls wire breaks the surface of the rising water, a stop watch is started.
The latter is stopped when the water exactly reaches the scratch. "
" The large tube is lowered, and the stopcock turned so that at( may enter
through its hollow point. When the water is again alnn the large cylinder, the
cock is turned to connect the small vessel with tlie outside air through the
platinum tip, the 'large cylinder is replaced on the elevation and the operation
repeated. Results should be obtained which ,check within one-fifth second.
" Now connect a rubber tube to the gas supply and to the tip of the stop-
cock, lower the large cylinder and force the water into the latter by means of
the gas pressure. ThOl;oughly saturate the water with-.the.t_gas to be tested;
this may be done by shaking gas and water together and by forcing the water
up and down in the small vessel in contact with the gas. Repeat the operation
with gas in the same manner as described for air. The calculation is made in
accordance with the formula.
" The advantages of this apparatus are its portability, its cheapness, its
rapidity and accuracy. When set up, the cylinders are inclined to be a trifle
unstable; this may be overcome by fastening a lead plate to the base of ea'llh
stopper. Four precautions in connection with its use should be emphasized:
179 Stone, book cited, 261. See also Bureau of Standards, Tech. Papers 89 and 94, by
J. S. Edward. I
Orton, Geol. of Ohio, VI, 137. See article by ~em:p~ Collins and Killin, Ind. Eng.
180
Chem., Anal. Ed. 7, 338 (1935). .
GAS ANALYSIS 2399
(1) The water must be of the'room temperature; (2) the water must be thor~
oughly saturated with the gas; (3) the platinum tip, stopcock, and upper part
of the tube must be kept dry and clean; (4) the large cylinder must always, in
anyone determination, be placed at the same height,"
The use of the Edwards Gas Balance has already been explained on p, 2344.
Another method consists in the use of the Lux gas balance. This consists
of a balanced globe into which the gas previously filtered through cotton, passes
and its specific gravity is read off directly on a scale.
The knowledge of the specific gravity is important; as it is involved in the
formula for the calculation of the flow of gas in pipes; it also enables the gas
manager to ascertain the weight of gas produced from the coal, and to get an-
idea of the nature and amount of impurities in the gas, all these being heavier.
(h) Tar.-For the estimation of tar, Clemens Winkler 181 recommends the
procedure of Tieftrunk: This consists in passing the gas through 25% alcohol
and collecting and weighing the tar on a tared filter.
In this country the so-called" Tar Camera" is used: this consists of a brass
app.aratus 182 into which a 3-inch E and A No. 4970 filter paper is clamped,
through which a measured quantity of gas is passed. By the increase of weight
of the filter, the amount of tar is determined; the volume of gas may vary from
0.1 to 10 cubic feet.
Stone and Prince 183 use an ordinary Gooch asbestos filter for this purpose.
It is weighed after treating with dried and purified gas, subjected to the gas to
be tested and dried and weighed as before.
volume of the aspirated gas. From the quantity of iodine used and the volume
of the gas required to de colorize the solution the pel[ cent of sulfur dioxide i~
calculated.
Should the contact gas contain
80 3, this is: ~bsorbed by passing the
gas through: 50 to 100 ml. of. cone.
H 2S0 4 , to avoid the action of 80 3 on
the rubber tubing of the apparatus.
A rapid current of the gas is passed
through the acid to saturate it with
S02 before making the tests.
The Reich method is more applic-
able for determining small amounts of
S02. A 12-15-liter graduated aspir-
ating bottle is used in works tests of
exit gases for measuring the gas. Since
these volumes are not under standard
conditions (760 mm. and 0° C.), it
will be necessary to convert the vol-
umes obtained in the tests to these
conditions, using the formula .
.......
po-w
·V = V0 760 (1 +0.00367t) '
where VO = measured volume, po = ob-
served barometric pressure, t = terp. per-
ature of the gas, and w = aqueous ~apor
FIG. 377. Portable Reich App~ratus. pressure at temperature of the test.
Volume % SO, " 25 mr. NllO Iodine 10 mi. NII0 Iodine 5 mr. N/lO Iodine
the gali burns, forming a " cap," like the colorless flame above the blue cone in
a Bunsen burner: the length of the flame is a measure df the percentage of gas,
and as little as 0.1 % is visible using the hydrogen flame.
. The ". J. W. Indicator" for the detection of inflar4mable greses or vapors,
depends for its use upon the fact that the resistance of a wire is increased by
the heat of combustion of the gas. The gas or vapor is sucked through a ~ube
containing a platinum wire heated to ,dull redness by a dry-battery current:
the burning of the gas increases the resistance of the wire and from this increase
the quantity of gas or vapor can be determined.
It is safe, easily portable, easily handled by the ordinarily intelligent and
gives direct readings. The Union Carbide Company's indicator is like it in
principle, but the heating element was dropped into the tank itself: this was not
.safe with hydrogen, acetylene or carbon bisulfide. The screens protecting the
element necessitated frequent cleansing. Obtainable from the Bullard Davis
Co., 67 Wall St., N. Y.
Carbon Monoxide.:"_Besides combustible gases or " fire damp," it is' some-
times necessary to get an idea of the amount of carbon monoxide (" white
damp ") in the mine air after an explosion or in the" after damp;" chemical
methods, being too slow, use is made of the behavior of birds and mice when
exposed to such an atmosphere. To this end they are carried in cages by the
rescuing party and their behavior noticed. Canaries show distress in an at-
mosphere containing 0.15% of CO in fi~e to twelve minutes, or with 0.20%
in half this time: Mice are less sensitive,I92 and men may display distress when
carbon monoxide is as little as 0.1 %, wherealil animals may be unaffected. In
case either is overcome by the gas, resuscitation can be effected by' bringing
them out into the open air again. Repeated exposure of the animals to the gas
would seem to be without influence.
VI. ELECTROLYTIC GASES
Gases from electrolytic chlorine, hydrogen and oxygen generators. The
following are to be sought for: / .
(a) Chlorine, (b) oxygen, (l!) carbon dioxide, (d) carbon monoxide, (e) hy-
drogen. '
(a) Chlorine.-HempeI 193 recommends measuring the gas quickly in his
burette over water and then sucking in 5 ml. of 50% potassium iodide solution
through the capillary and shaking; the diminution in volume gives the chlorine.
The other gases are determined in the usual way. l1.8 -phQsphorus' cannot
be used for pure oxygen, a specially prepared potassium pyrogallate, is em-
ployed; cuprous chloride or ammoniacal cuprous carbonate in -the absence of
carbon monoxide is very satisfactory.
VII. ACETYLENE
Commercial acetylene may contain the followiIig gases:
1. Oxygen; 5. Nitrogen;
2. Carbon monoxide; 6. Sulfur-containing gases;
3. Hydrogen; 7. Phosphine.
I
4. Methane;
192 Burrell, Seibert and Robertson, Bureau of Mines-T-;cbnical Paper 62,1914, "Rela-
tive Effects of Carbon Monoxide on Small Animals." Also Tech. Paper 11.
198 "Gas-Analytische Methoden," 4th Ed., 1913, p. 278.
GAS ANALYSIS 2403
(a) Moisture; (b) Carbon dioxide; (c) Ozone; (d) Carbon monoxide; (e)
Bacteria. ,
(a) Moisture 196 by chemical means. The amount of moisture can be
determined by the sling psychrometer, or wet- and dry-bulb thermometer or
by the hair hygrometer. The sling psychrometer is the most rapid and accurate
-the wet- and dry-bulb thermometers are so arranged that they can be rapidly
whirled for fifteen or twenty seconds, stopped and quickly read, the wet bulb
first; this ,is repeated until closely agreeing readings are obtained. The
humidity is determined in the usual manner from the meteorological tables.
The amount of moisture in the air is probably indirectly responsible for our
sensations of comfort or discomfort, rather than the amount of carbon dioxide,
as was formerly thought. This moisture controls the heat loss from the body,
which loss must be normal-neither too high nor too low. The greater the
humidity the less the evaporation, consequently the less the cooling.
For comfort, the higher the temperature the less should be the humidity; the
following shows the relation between the two:
Temp., 0 F. . . . . . . . . . . . . . . . . . . . . . .. 60 68 197 70 80 90
Relative humidity, %.............. 67 40 49 31 16
(b) Carbon Dioxide.-One of the most satisfactory methods is that of
Hesse.l 98 This consists in absorbing the carbon dioxide from a definite volume
of air with standard 'barium hydroxide and determining its loss of strength.
The air is collected in heavy Erlenmeyer flasks of 100-500 or 1000 ml.
capacity, or clear glass bottles; these are stoppered with doubly perforated
rubber stoppers carrying glass plugs and the capacity of the flask or bottle
noted to the depth of the stopper. A IO-ml. pipette and a 15-ml. glass-stop-
19' Book cited.
195 Bureau of Mines Tech. Paper 112, "Explosibility of Acetylene."
196 Benedict, "The Composition of the Atmosphere with Special Reference to its
Oxygen Content," Carnegie Institution of Washington, 1912, Publication No. 166.
197 For indoor work.
198 Eulenberg's Vierteljahrschr. f. ger. Med. u. San., N. F., 31, 2.
2404 GAS ANALYSIS
pered burette graduated in m!'/19 with an 8-cm. tip, complete the outfit; a
solution of 1.7 grams of barium hydroxide and chloripe (20: lrin a liter of
water, B, Fig. 378, and of oxalic acid 5.6325 grams ;per liter (1 ml. = 1 m!.
CO 2), with phenolphthalein 1: 250, are required. Th~ small bottle attached
to the large one is filled with soda lime.
, The bottles are filled with steam by exposure for
three minutes and the vaselined stoppers inserted, or
they may be filled with distilled water and opened in
the place the air of which is to be examined.
In all this work, it should be remembered that the
exhaled breath contains about 400 parts carbon dioxide
per 10,000, consequently care should be taken not to
contaminate the samples, nor should they be warmed
A with the hand.
The 10-m!. pipette is partly filled from the tube A,
Fig. 378, by means of the rubber connector-and sucking
the liquid into the pipette: it is rinsed, filled and inserted
through one of the holes in the rubber stopper of the
FIG. 378. bottle, the other plug being momentarily opened. The
plugs are replaced and the bottles allowed. to rest on
their sides, with occasional rolling, for twenty minutes. Not more th;m one-
fifth of the solution should be used up by the carbon dioxide present.
During this time, the barium hydroxide should be standardized; to this end
a few drops of phenolphthalein and a quantity of the oxalic acid almost sufficient
to neutralize the hydroxide should be run into a 100-ml. Erlenmeyer flask from
the burette; this should be passed through the doubly perforated stopper; 10
ml. of the barium hydroxide solution are run int.o the flask as above described,
and also the oxalic acid until ,a pink color appears.. I
Phenolphthalein is added to the bottles containing the samples, the oxalic
acid burette inserted through the stopper and the excess of barium hydroxide
titrated.
The barometric pressure and temperature in the laboratory are noted and
the volume of the bottle less 10 ml. Ba(OHh calculatei to standard conditions;
the difference in the titer of the barium hydroxide solution gives the volume
of carbon dioxide in the bottle; this is calculated into parts per 10,000.
A more rapid method is that of Cohen and Appleyard.199 This uses a
standard solution of lime, water colored with phenolphthalein to absorb the
carbonic acid: it depends upon the principle that the quantity of air and lime
water remaining the same in each case, the rate of its decolorization will vary
inversely with the amount of carbon dioxide.
Half-liter bo'ttles are used for the purpose and either filled with water, or a
vacuum is produced in them by steaming out and stoppering; the sample is
taken by displacing the water or allowing the air td' fill the vacuum: 10 ml. of
the standard lime water (Reagents) measured with a Vanier overflow pipette
run into the bottle, the time noted, the rubber stopper replaced, and tht bottle
shaken vigorously and continuously, until the pink color disappears when seen
over a white paper, and the time again noted. From the ..time required for
decolorization, the amount of carbon dioxide mll:yjJ~dbund from the table.
m Chern. News, 70, 111 (1894).
GAS ANALYSIS 2405
Time, Seconds Parts CO, per 10,000 Time, Seconds Parts CO, per 10,000
Other methods for this determination are more strictly gasometric, measur-
ing the diminution in volume by absorption: Benedict 200 used Sonden's, and
Anderson 201 recommends a shortened form of the Pettersson-Palmquist
apparatus.
This may be d~scribed as an Orsat apparatus using mercury instead of water
and with a 25-ml. burette the lower part of which is graduated to 0.0025 ml.;
this is connected to a pipette of potassium hydroxide, a delicate manometer
and compensating tube. The apparatus is delicate and rapid, but requires a
skilled operator to manipulate it. Any form of accurate apparatus may be
used.
The amount of carbon dioxide in the outdoor air of the city is about 3.1 parts
per 10,000, in the house, 3.7-3.9; with 6-7 parts in a room, the ventilation may
be considered as excellent, with 10 parts as about the upper limit. In some
theaters which were lighted bY, gas it approached 50 parts.
(c) Ozone.-Probably most conveniently determined by Wurster's method,
p.2370.
Determination of Ozone, Iodide Method of Schonbein. 202 The method
depends upon the reaction
2KI+Oa+H 2 0=I 2 +0 2 +2KOH.
Procedure.-A glass bulb of 300 ml. to 400 ml. capacity (the exact capacity
being determined by weighing empty, then full of water) with two tubes at
opposite sides, closed by stopcocks, one of which was a three-way cock, is
filled with water. The ozone test is introduced into the vessel, displacing the
water. The gas is brought to atmospheric pressure by quickly opening and
closing one of the cocks. A solution of 2 N potassium iodide is introduced in
excess of that required, through the three-way cock by ,means of a second bulb
connected by rubber tube to the vessel after displacing the air in the rubber
connection through the cock. The mixture is well shaken and.,allowed to stand
half an hour. The contents of the bulb are washed out with additional iodide
solution followed by ~istilled water and the liberated iodine titrated with
standard thiosulfate.
1 m!. N /10 N a 2S20a = 0.0024 g. Oa.
(d) Carbon Monoxide.-The qualitative detection is most certainly effected
by the blood test; to this end the gas is drawn through a solution of blood
contained in a Wolff, Fig. 361, or similar absorption tube and examined for its
2ao Book cited, also Haldane's.
2a1 J. Am. Chem. Soc., 35, 162 (1913).
202 Further details of this method may be found in Treadwell and Hall, "Quantitative
Analysis." .
2406 GAS ANALYSIS
Sayers and Yant 204 have worked out a method for the determi'ilation of the
gas by taking advantage of the carmine red c'olor produced by carbon monoxide
blood in tannic acid, and comparing the red produced with standards, as with
" Hoolamite," p. 2373.
For the determination of the gas in vehicular tunnels it is absorbed in
" Hopcalite," a mixture of the oxides of manganese, copper, cobalt and silver.
This produces heat whioh changes the conductivity of a' wire carrying a current;
this is converted into per cent carbon monoxide and automatically recorded;
the same current also turns on the ventilating fans. 20Ii
203 Ind. Eng. Chern., Anal. Ed., 1, 129 (1929).
204 Tech. Paper 373, Bur. Mines, 1925.
206 Bur. Mines Tech. Paper 3551926.
GAS ANALYSIS 2407
, Method of Sampling
Some samples of gas taken by the bureau's investigators are collected by
;tir displacement i.n bottles with ground glass stoppers. At the wells or supply
pipes a stopper".is removed and the gas is allowed to run into a bottle until
air has been entirely displaced. Usually a few minutes is sufficient. While
the bottle is still in place, the stopper is inserted. As a precaution a~ainst
leakage, melted paraffin is poured over the stopper. The bottles are then
well packed in suitable wooden boxes, and mailed to the bureau's Pittsburgh
laboratory. When received at the laboratory the stoppers are removed under
mercury and the gas is transferred to and analyzed in an apparatus that con-
tains mercury as the confining fluid. Some samples have been received in
strong iron cylinders at the full pressure of the well. Other samples have been
collected by a method devised by G. A. Hulett, formerly chief chemist of the
bureau, and at present used for collecting .
some of the bureau's natural gas samples.
The method is as follows:
Glass tubes about 17 cm. long and 6
cm. wide are provided. at" one end with ('.
a glass tube extension about 15 cm. long
and having a 6 mm. bore. The tube is
'left open. Natural gas is directed into.
the sampling tube through the orifice by
means of a long slender brass tube. When
the air originally in the tube has been
displaced, the brass tube is withdrawn
and the glass tube orifice closed by placing
the finger over it. The sampling tube is
then carried to a small alcohol flame and
the glass tube extension sealed off.
• r
found after the combustion to the total volume of. residual gas found after the
combustion; that is, if the total volume after combustion is found to be 70 ml.
and the carbon dioxide is 40 mI., then the partial pressure of the carbon dioxide
. 40x760 I
will be 70 434 mm., and the proper molecular volum~, namely, 0.997,
will be found from the table. I
The foregoing equations have reference to their use at 0°' C. and 760 mm.
pressure. The following table shows the correct molecular volumes for carbon
dioxide at 20° C. and different partial pressures.
MOLECULAR VOLUME OF CARBON DIOXIDE CORRESPONDING TO DIFFERENT PARTIAL
PRESSURES AT 20° C.
Mm. of Mercury Molecular Vol.
100, .. ' ., ............ , .. , ., .... " ' ..... ' .... , ........ , 0.9993
200 ...... " " ., , .... , .... , .. , .. , ... ' ., .. , .. , .' ... , .. .. .9986
300 ... , , .... , ....... ' .... ' ...... , . , . , . , ....... , . , .. , .' .9980
400, ..... , .. , .. , .... " " ., .' " ...... '. , .. , ......... , .. .9972
500 ....... , .. ' ,. , .... , .... , ........ ' " ., ., " '. , .. , '. .. .9965
600 .... " ., ., ., .,' .. " ., .' ..... , ,. , ........ , " , .. ' " " .9958
700 ... , .. " ... , .,' .. " ., ., ., " .. ,. ,. '. , .. , ,. ,. ,. " .. ,. .9951
760 ... ' ....... ' .. , ...... , .. ' , . ' , ......... ' , . " . . . . .. .. .9950
In compiling' this table advantage was taken of the work of Rayleigh,212
Leduc 212 and Chappius 212 having to do with the determination of the specific
gravity and coefficient of expansion of carbon "dioxide. The specific gravity
determina_tions were given by Rayleigh and Leduc for carbon dioxide at"Oo C.
and 760 mm. pressure. Values for 20° C. and 760 mm. pressure were deter-
mined from the coefficient of expansion of carbon dioxide between 0° C. and
200'C. A graph was plotted from two values, the deviation from the gas laws
at 760 mm. pressure and at 380 mm. pressure.
The coefficient of expansion of ethane between 0° and 20° C. has not been
determined, consequently the same molecular volume was used at 20° C., the
laboratory working temperature, as was reported by Bavme and Perrot 213
at 0° C. The error resulting from this usage can be disregarded without
introducing any appreciable error in the analyses, judging from the molecular -
volume of carbon dioxide which at 20° C. is only 0.001 different from the value
at 0° C.
Below are given the molecular volumes to be used in the case of ethane for
different partial pressures.
MOLECULAR VOLUME OF ETHANE CORRESPONDING TO DIFFERENT PARTIAL Pl!.ESSURES
Mm. of Mercury Molecula'r Vol.
0, ........ , ........... ' .......... ' .. , ..... ~, .. , , ..... 1.000
100 .... , .. ' ...... , .., ............. ' .' " .... , ..... ,. .. .. .. .999
~gg ::::::::::::::::::::::': :::::::::::::::::::::: : :::::: :~~~
400 ............... , .' ............ " , ... , ........ ' " " .. .995
500 ............... , ... , .' .... " ... , ., .. " ., ... ' ... , '. .. .993
600 ... , ., : .... , " ... ' ............. , ' .... , .. ' ......... ,. .992
~gg: :::::::::::::~: :::::::::::::::::::::::::::::; ~~ :::: <~~
?12 See Landolt and 13i:irnstein, Phys.-chem. Tabellen, 1912;'i1.14.
213 Sec Landolt and Bornstein, Physikalisch-chemische Tabellen, 1912, p. ).48.
GAS ANA.4 YSIS 2413
Tyndall demonstrated that dust particles are responsible for the visibility
of rays of light, such as the sun beam or the beam of the searchlight. He did
not make clear the best conditions enabling one to see the beam produced by
minute amounts of suspended matter in purified gases, nor did he show that a
gas free of dust may still produce a beam in presence of liquid mist. \The
opinion commonly current regarding the conditions necessary to see the beam
produced by light rays in purified gas, is that the rays s~ould be parallel and
Beam Test for Suspended Matters in Gases
Ughf_
(Inlef
a
FIG. 383.
should be viewed with the visual axis at right angles to this beam. Observing
these conditions the author 214 failed to see the bea'm using direct sunli~ht, bqt
discovered that a cone of light from an arc light gave a-distinct beam. This
led to the discovery that the beam is best viewed by glancing towards the source
of light. Accordingly the apparatlJs shown in Fig. 383 was designed. .
For the examination of a corrosive gas, it is advisable to have the chamber
of the camera made of a material not affected bY-that gas. For example, in
--
214 Contributed by Wilfred W. Scott.
GAS ANALYSIS 2415
FIG. 384. Apparatus for Ascertaining the Pressure of Suspended Material in a Gas by
Means of a Beam of Light. •
The beam caused by liquid mist alone appears as a uniform pencil or band
of light, whereas dust particles produce a sparkling effect 'br irregular flashes
of light: The beam is more intense with the gas in motion. Within certain·
limits the intensity of the beam is directly proportional to amount of suspended'
matter in the gas.
If the gas is dry the beam is completely removed by passing the gas through
a three-inch mat of loosely carded purified asbestos. If the gas, thus purified
of solid matter, is passed through a liquid the beam again becomes apparent.
From this it is evident that the test must be applied to dry gaS'if it is desired to
ascertain the presence of solid matter in suspension.
Fig. 387 shows the apparatus set up for ascertaining the effectiveness of a'
scrubber in a quantitative determination of the solid and liquid suspended
matter in a gas. The apparatus is so arranged that comparisons may be made
of the unfiltered gas, by means of a bypass with that of the filtered or purified
gas. Advantage is taken of the difference in pressures at A and B, so that a
continuous sample may be taken of the gas without the necessity of an aspirator.
A hand arc light shown on the left furnishes the rays of light which enter at " a "
and pass out into the air at "c." The beam is viewed through the eyepiece
"b." The meter measures the vol~me of gas aspirated.
DETERMINATION 'OF SOLID AND LIQUID IMPURITIES IN A GAS 215
, FIG. 385. Apparatus for Determining Solids FIG. 386. Filter for Gases.
and Liquid Mist in Wet Gases.
The funnel is packed first with a layer of glass wool, followed by a-two-inch
layer of well-carded, purified, blue asbestos, and finally a second layer of glass
wool. The glass wool prevents the asbestos plugging the inlet and exit open-
ings. Perforated disks may be used in place of the glas~ wool.
The asbestos is the blue African form. White asbes.tos will not give accu-
rate results as this is attacked slightly by dilute sulfuric acid and decidedly
so by strong sulfuric. The blue form is not appreciably attacked so that the
acid may be extracted as free acid by water. The asbPJltos is purified by digest-
ing with nitric acid, followed by a second digestion ,~it4 hydrochloric acid.
The material is now thoroughly washed to remove the..acids and is dried. The
dry asbestos is placed in wide mouth bottles until required. Before using it is
/'
GAS ANALYSIS 2417
carded or fluffed out by means of a pick and then placed loosely in the funnel
to be used in' the test.
Impurities in Gases
The Test.-The gas is aspirated through the filter at the rate of about five
cubic feet per minute, the gas being measured by a dry gas meter following the
apparatus. When a volume of 10 to 100 cubic feet of gas has been aspirated,
the amount 'depending upon the amount of acid mist present in the gas, the
asbestos filter is, placed in a beaker. (If 80 2 is present in the gas, as is the case
in contact sulfur gases, air is ,first passed through the asbestos to ;remove
the 80 2 .) The tube is rinsed out, the rinsings added to the asbestos. The.
contents of the beaker is now transferred to a two-inch Buechner funnel in
position on a half-liter capacity suction bottle. The ,vater is drawn out of the
The nitrometer, Fig. 388, consists of a gas-generating bklb fitted at the top
with a two-way cock leading to a dissolving cup and a gas-((xit tube, and which
has at the bottom a connection for a rubber tube leading tol a leveling tube, the
whole being filled with mercury to a level just below the uppJr cock; a cylindrical
glass measuring tube graduated from 0-100 m!. connected to a leveling tube
~ Holder for
Gas Generating Bulb
Clamp
~ -{
-a
~
"
.g'"
.
~ Clamp
~" ~C/amp
CIamp
::t
u
.l2£. u 130 '"
-0
"
-a -~
~ go
.~ i
~ "
-.I
.rr"
(
3>~ --
FIG. 388.
cock of the tube is then closed and sealed with melted paraffin. The volume
in this tube may be confirmed at any time by opening the measuring tube and
adjusting the leveling tube so that the level is the same in the three tubes. The
volume of gas in the equilibrium tube is then read and compared with the
calculated volume at the time, a correction in the subsequent gas readings in the
measuring tube being made accordingly.
Nitrogen in Potassium Nitrate.-Approximately 0.4 gram potassium nitrate
is placed in a weighing tube, dried two hours at 110° C., desiccated twenty-four
hours over sulfuric acid and weighed exactly by difference into the cup of the
gas-generating bulb. This is dissolved in 9 m!. 95% sulfuric acid, added through
a siphon thistle tube, entering the top of the dissolving cup through a tight-
fitting rubber stopper. When the salt is dissolved it is drawn into the gas-
generating bulb and followed by two washings 1.5 m!. each of 95% acid. The
bulb is then shaken with a motion normal to its long dimension until the volume
of gas determined by a rough paper s()ale pasted on the leveling tube becomes
constant, this operation taking from three to five minutes.
The gas is passed into the measuring tube, and after standing five minutes
the leveling and measuring tubes are so adjusted that the level in the equilibrium
tube reads 100 ml. and is the same as the level in the measuring tube. The
reading of the latter is then taken. As tJ'le temperature and barometric condi·
tions, insofar as they affect the measured volume of the gas, are automatically
compensated by the equilibrium tube, and as the gas is washed with sulfuric
acid and is, therefore, dry, the percentage of nitrogen may be calculated directly,
correcting only for the calibration of the equilibrium and measuring tubes.
Seventeen determinations made when the room temperature ranged from
20-28° gave 13.71% nitrogen, the theoretical being 13.84%.
The nitrometer gives reliable results within 0.02% in nitrate nitrogen, using
0.4-0.5 gram sample. It is not available for the determination of nitrogen in
celluloid or other substances containing carbon ring compounds which appear
to prevent the complete liberation of nitric oxide in the presence of sulfuric
acid and mercury.
Besides the corrections for calibration and standardization of the nitrometer
in accordance with temperature and barometer, the gas readings should be
corrected for solubility of nitric oxide, which diminishes when the temperature
goes above 28° C., and the formation of other gases by the breaking up of the
'cellulose molecule which increases.
The following table gives the algebraic sum of these two corrections for
temperatures ranging from 20° to 35° C., for nitrocellulose.
explosions, a reducing regulator for any other gas must not be used with oxygen
or nitrous oxide, and the converse of the statement is equally true. Special
fittings to permit the use of apparatus with right-hand threads with left-hand
fittings and vice versa had best not be made or acquired, as this fitting has
caused fatal accidents.
As has been already stated, cylinders of compressed gases should be handled
with care: attention has been called to two fatalities caused by dropping
cylinders of compressed acetylene, a distance of about a foot.
APPENDIX
, REAGENTS AND TABLES
If the percentage of oxygen in the mixture does ndt exceed 28, solution (a)
may be used;222 if this amount be exceeded, (b) must be employed. Oth,rwise
carbon monoxide may be given off even to the extent. bf 6%.
Attention is called to the fact that the use of potassit:m hydroxide purified
by alcohol has given rise to erroneous results. .
Absorption Capacity.-One ml. absorbs 2 ml. O2.
Sodium Hydroxide.-Dissolve the commercial hydroxide in thre~ times its
weight of water. This may be employed in all cases where solution (a) of
potassium hydroxide is used. The chief advantage in its use is its cht;apness.
Sodium pyrogallate is, however, a trifle slower in its actIOn than the cor- .
responding potassium salt.
Table 1 223
CONSTANTS OF THE GASll;S
Sp.Ht.( ./
Sp.Gr. Wt.( Cu.Ft. Press. Explo- Iisoi. (If Gas in 1 Liter
to Liter in at at sivity of Water in ml.
Grazus, B.Pt. i\1.Pt.
Gas Air Atmos. 20° C. % °c. °C.
Calcu- 0°, Press., in Gas in
lated 760mm. 62° F., Lbs. Air
C.P. At 0° At 20° C.
Table 2
TABLE SHOWING THE TENSION OF AQUEOUS VAPOR AND ALSO THE "VEIGHT IN GRAMS
CONTAINED IN A CUBIC METER OF AIR WHEN SATURATED FROM 5° TO 30° C.224
Table 3
SPECIFIC HEATS B.t.u. PER POUND OF GASES AT CONSTANT PRESSURE C.p.
Volumetric 225
Table 4
CALORIFIC POWER OF VARIOUI", GASES m IN BRITISH THERMAL UNITS PER CUBlc FoOT
Name
AT 30 IN. PRESSURE
Symbol
\
Table 5
ENGLISH AND METRIC EQUIVALENTS
1 pound =453.59 grams 1 calorie =3.969 B.t.u.
1 cubic inch = 16.39 ml. 1 cubic meterl =35.31 cu. ft;
1 cubic foot =28.315 liters
1 B. t. u. per cu. ft. = 8.89 calorie per cubic meter
1 calo~e perl~u. meter =0.1124.B.t.u. "
1 in. H 20 pressure = 1.87 mm. Hg pressure 1 mm i Hg pressure = 0.535 in. H 20
pressure
227 H. L. Payne, already cited.
228 Where obtained by cracking oil.
229 The chief constituent of gasoline.
230 Dixon and Coward, Proc. Chem. Soc., 26, 67.
231 Gas Chemists' Handbook.
232 Calculated.
GAS ANALYSIS 2427
Table 6
The following table assembled from compilations by Henderson and Haggard 233,
shows the:
- ~ "'" ] "'«:"
I
·s ·s
'"
0 «:
~ ~ ~
[
. ::;"
0
.!l
~ .
·i .~ 00
~
I'« "
.S is ..:
!f~
'"~ ]
1l "
~ .a 1 ~
"
.a.... ""
0
~
0 0 0
.a....
!!l0
.~
.d
'~
0
13
~
]
~ ~~
«: u u :t: ~ ~ ~ z 'if:
Least detectible odor ....
---
53
- - - - - - - - - - - --
3.5
--
00
3-5
"300
I:Q
233 "Noxious Gases and the Principles of Respiration Influencing their Action."
2428 GAS ANALYSIS
Table 7
VOLUMETRIC FACTORS I
1 ml. of the Different NjlO Solutions Corresponds to the-Following Volumes and 'Veights
of the Different Gases
Table 8
DATA ON CYLINDER>S FOR COl't.fPRESSED GASES
. ... .. c 1<
~'t)'\I:: ...
Diameter incll<'s .
'_'_'
'" .
"""2 ----
J'
- - -"-- -
D
- - --
3X 4)1 • 4)1 4)1 5)1 8>'2 9
Y C
The smaller cylinders from E to AAA have no threads on them to couple to fittings:
they are connected to apparatus by a shnple yoke with a hose connection.
2430 GAS ANALYSIS
Table 10
EMERGENT STEM CORRECTIONS TO READINGS OF OUTLET-WATER THERMOMETERS FOR
DIFFERENT IMMERSIONS OF THERMOMETERS IN CALORIMETER
- - - - - - - - - ~I~ - - -
Water 50° 60° 70° 100°
of.
10 +0.02 +0.03 +0.04 +0.05 +0.05 +0.06
0
Thermometer immersed to 30 F .. 15 +.04 +.05 +.06 +.07 +.09 +.10
.
Thermometer immersed to 40 0 F ..
20
10
15
+.06 +.07 +.09 +.11 +.13 +.15
+.01 +.02 +.03 +.03 +.04 +.05
+.03 +.04 +.05 +.06 +.08 +.09
20 +.04 +.05 +.07 +.09 +.11 +.12
- 10 +.01 +.01 +.02 +.03 +.04 +.05
Thermometer immersed to 50 0 F .. 15 +.02 +.03 +.04 +.05 +.07 +.08
20 +.02 +.04 +.06 +.07 +.09 +.11
10 +.00 +.01 +.02 +.02 +.03 +.04
Thermometer immersed to 60 0 F .. 15 +.00 +.01 +.03 +.04 +.05 +.06
20 +.00 +.02 +.04 +.05 +.07 +.09
This table is not applicable if the emergent portion of the stem includes an enlarge-
ment in the capillary.
Instead of using the above table, it will probably be somewhat more convenient to
make out a stem-correction table for the particular outlet-water thermometer that is to
be used with the calorimeter, the data for this separate stem-correction table being
interpolated from the above table.
Suppose, for example, the outlet-water thermometer to be used was one that was
immersed to the 300 F. mark on the scale, and a stem-correction table were wanted for
an 18° F. rise in.temperature, then from the above table we obtain the following stem-
correction table:
OF. 0
OF. 0
50 0.05 80 0.09
60 .06
. 90 .11
70 .08 100 .13
In the same way a table could be made out for any outlet-water thermometer' by
interpolation in the general table. The table so prepared would apply for the particular
point to which the thermometer was immersed and for the particular rise in temperature
with which the. observer had chosen to work.
I
CHEMICAL MICROSCOPY' 1
"
ADVANTAGES OF MICROSCOPICAL METHODS'
If the term "chemical analysis" is construed in a broad sense, it covers
determinations of properties and behavior as well as of composition. Most
purely chemical data must ultimately be translated into comparisons of purity,
costs, strength, or other aspects of the history' or future performance of the
material in question, but such interpretations are frequently indirect and
inconclusive. The great advantage of microscopical methods in analytical
chemistry lies in the directness of the information which they yield and the
rapidity with which it is obtainable. The ability to deal with minute amounts
of material is in most cases secondary to that of being able to observe the actual
physical condition of the sample and the phenomena which it exhibits, in a
vivid" close-up" rather than by inference.
A further advantage of the application of the mic.roscope to analytical
problems results from the fact that the sample itself is of necessity under direct
examination, so that in most instances unsuspected"and highly significant
additional information, which might other;vise have been missed, is brought
forth.
r
STUDIES OF SINGLE SUBSTANOES
As an aid to ordinary macroscopical analysis or: single substances the micro-
scope is employed in various ways. Single substances can be examined for
purity and uniformity, and their physical state observed. Particle size and
shape of powdered materials may be ascertained, as factors .governing solu-
bility, reactivity or suitability for pigments, fillers, abrasives, etc: Surface
character of coatings, such as paints, glazes, platings, or enamels, may explain
their failure better than data as to their constitution.
Identification of substances is rendered more positive by supplementing the
analytical information by a determination of characteristic structural or optical
properties, for which the microscope is the instrument par excellence. By
utilizing published descriptions or by comparison with authentic "knowns."
conclusive identification is frequently possible without the use of chemical
reactions and without the need of separating the material from its surroundings
or from a mixture. I
1 E. M. Chamot, Professor of Chemistry, Emeritus, 9or~lf University, and C. W.
Mason, Professor, Chemical Microscopy, Cornell University. - .
2434
CHEMICAL MICROSCOPY 2435
standard liquids on an object ~Iide, and c(wcrcd with a cover glass .• \Vith the
diaphragm of the ~ub~tagc conden~r c'o~cd 8~ raf 8!-; prr.::sible, a naTrUW ~right
line (Il llecke line") appears to bound the graill.!:!; this moves toward the sub-
FlO. 389. The Becke Line. Sodium Chloride (n=1.544) in a Liquid of n=1.540.
A-on focusing upward slightly; B-on focusing downward slightly.
1
I
I
'..-"II: ()
~ I'
~""'a
I
I
.~.
;
A.
!
D -.
-FIG. 390. Axial and Oblique Transmitted Illumination. The upper crystal has
a higher re(ractive i.ndex tha.n the mQunting medium; the lower crystal, a,less rerra~tive
index. • -
or a piece of"paper, one side of the field will appear shaded. The boundary of
the particles will also be darker on G~e 8ide; if this shading is on the same side
as the shadow. in the neld, the particle has a higher refractive index·than the
, and !'ire 'versa (f'ee Fig. 390).
surrounding liqiJid,
2438 CHEMICAL MICROSCOPY
Solid particles can best be handled by a glass rod or platinum wire, as men-
tioned above. If suspended in liquid, sedimentation or centrifugal separation
will serve as a means of concentrating tl;tem. Fine capillary pipettes (Fig. 391)
FIG. 391.
made by drawing out glass tubing to about 1 mm. diameter, are useful for col-
lecting sediments or transferrii:J.g small quantities of liquid.
The filtration 10 of small drops may be readily accomplished by the Behrens-
Hemmes method (Fig. 392). A small square of
thick filter paper is placed close to the drop to
be filtered, against the paper is gently pressed a
capillary tube F whose lower end has been ground
true and smooth. The paper with the tube in a
vertical position is slowly pushed forward intolthe
drop. The liquid rises into the tube by capillar-
ity. In general it is best to first attach a rubber
tube R to the tube F and draw up the liquid into
the tube by gentle suction applied by the mouth
to the mouth-piece M. The rubber tube is kept
slightly bent as shown in the figure. At the
m~ment of removal the tube is compressed be- FIG. 392.
8Titus and Gray, Ind. Eng. Chern. (Anal. Ed.), 2, 368-71 (1930).
9"Handbook of Chemical Microscopy," Vol. I, pp. 141-5; Vol. II, p. 407.
10" Handbook of Chemical Microscopy," Vol. II, pp. 13-17.
2440 CHEMICAL MICROSCOPY,
tween the fingers and slightly stretched at the same time; this prevents the
liquid from flowing out of the tube as it is lifted off the fi:lter p~per. , •
When merely the preclPitate and not!
the liquid is required a narrow strip of
filter paper may be: Introduced il1to the
drop and the precillitate pushed into a
little heap.
Decantation 11 will usually accom-
plish a perfectly satisfactory separation
of liquid from a heavy precipitate. The
object slide on which the drop rests
must be scrupulously clean. If the end
of a drawn-out glass rod1or a platinum
wire is introduced into th'e drop near its
circumference and then moved outwards
the liquid will follow the rod and if at the
same time the slide is slightly tipped
the liquid Gim be decanted from the
FIG. 3~3. precipitate.
Simple fractional distillations 12 can
be successfully performed by means of the apparatus shown in Fig; 393, 'the
successive fractions being removed from the receiver end of the distilling tube
by means of a capillary pipette. / '\.
:
MICROSCOPICAL QUALITATIVE ANALYSIS 13
There is usually a specific method for applying the various reagents, depend-
ing upon the relative concentrations required and the nature of the reaction
involved. In any case, however, actual precipitation is the criterion of a
positive test; crystalline products appearing only after considerable evaporation
has taken place are of little or no diagnostic value. Relatively dilute solutions
are preferable, in order that separation of the reagent, the unknown, or various
possible double salts may.not mask the precipitate obtained, and in order that
the latter may not be formed from such a highly supersaturated solution as to
appear "amorphous" or unresolvable by the microscope.
The most important methods of applying reagents are as follows:
a. The drop of r,!lagent is caused to flow into a drop of unknown solution on a micro-
scope slide, by means of a small glass rod or platinum wire. The concentration gradient
in the zone of mixing affords a considerable range of conditions, so that various regions
of the preparation should be studied in order to observe a di~tinctive precipitate.
b. A solid fragment of the reagent is introduced into the test drop, either at the center,
or sometimes at or near the circumference. When introduced at the center, solution is
rapid, and if the reagent fragment is moderately large, excess of the reagent in high con-
centration is made effective. The reagent should be in small fragments-not finely
powdered.
c. A solution of the reagent is drawn in a narrow channel across a dry film of the
material to be tested. The unknown is dissolved in water or acid. A drop of the solu-
tion thus obtained is carefully evaporated to dryness upon a scrupulously clean object
2442 CHEMICAL MICROSCOPY
~
slide, using great care to avoid heating the dry film too hot. A concentrated drop of
the reagent is placed near the film and a narrow channel of. the reagent is drawn across
the dry film, the object slide being slightly tipped at the same time so as to feed reagent,
into the channel. Care must be taken that the reagent does not flood tp1l entire drop.
The reaction product is formed in the zone of contact between rea!?ent and solid unknown.
This method is used when the reaction product sought, is relatirely soluble, and when
high effective concentrations of reagent and of unknown are necessary for its precipitation.
d. The addition of another ion giving a less soluble compbimd can be resorted to
when an isomorphous series exists and the test as first performed has failed to yield a
decisive test. One of the best illustrations is that of testing for lead and (or) copper with
KNO.. If no precipitate is obtained aRb, Cs, or TI salt can be added; the solubility
of the compounds decreasing in the order named. The effect of such a procedure is to
increase the sensitivity of the reaction.
e. The reagent is incorporated into a fiber (viscose, silk, wo<;>l, collodion, etc.) and
the dry treated fiber is introduced into the test drop. This method is of great value in
many cases. One of the best examples is testing for boron with a viscose rayon fiber
impregnated with turmeric (curcumin). '.
f. When HF or fluorides are employed as precipitating reagents they are added to
the test drop by either method a or method b, but special precautions are necessary.
The object slide must be of non-siliceous material such as celluloid, cellophane etc., or an
object slide may be protected by a coating of a hand impervious, transparent, colorless
varnish. All objectives save the one to be used should be removed from the nose-pieces
and this one protected by varnish, a protecting cell or drop of immersion oil and small
cover-glass.
g. In certain specific cases, it is found to be advantageous to cause the reaction tp
take place in capillary tubes. ..
Reactions involving gas evolution (NHa, ,or volatile acids)' are most simply
performed by placing the reacting substances in a tiny crucible (Fig. 394), over
which is placed a glass slide bearing
====== v'-
.
.
e on its under surface a minute
hanging drop of an appropriate acid
or alkaline absorbent; the evolved
gas is then tested for in this drop.
Fusions are carried out in a tiny
platinum cup made from heavy foil;
other high timperature reactions
yielding; sublimates are brought
about in hard glass capillary tubes.
Reagents Used.-Some of the
most useful group reag!'lnts of
microscopical qualU;ative analysis
FIG. 394. are as follows:
The most useful specific identity tests for the common anions and cations
are as follows:
Cations
Na-pale yellow tetrahedra with uranyl acetate; many of the tetrahedra appear to be
black because of total reflection. If the ratio Na : K is less thanl : 5, test usually
fails. Na+zinc uranyl acetate yields pseudo-octahedra, very characteristic;
presence of other alkalis or alkaline earths does not interfere with latter modifica-
tion of test.
K-yellow or orange octahedra with H 2PtCl.; colorless prisms with HCI0 4. NH, must
be absent, as it gives identical reactions. .
NH4-volatilization of NH,. Treat with NaOH in crucible (Fig. 394) covered with an
object slide carrying a hanging drop of very dilute HCI. Warm gently. Test
hanging drop with H 2PtCl.-yellow octahedra of (NH4)2PtCl •.
Ca-fine needles, singly and in sheaves, oblique-ended prisms, and characteristic arrow-
head twins, with dilute H 2S0 4. Prisms with tartaric acid. Very tiny octahedra
with oxalic acid. Sr and Ba interfere if present.
Sr-fine granular precipitate with H 2S0 4, which when recrystallized from hot conc.
H 2S04, yields dagger-shaped crystals. 'Ca and Ba must be absent. With oxalic
acid, .crystals similar to those of Ca. .
Ba-feathery crosses on recrystallization from hot conc. H 2S0 4 • Ca, Sr, Pb, Ag, must
be absent. In concentrated acetic acid K 2CrO, yields a very fine granular or
finely, crystalline precipitate, Ca gives none, and Sr rarely unless the solution
is made alkaline with NHiOH. With oxalic acid, mossy aggregates.
2444 CHEMICAL MICROSCOPY
Tests fall in mixtures of Ca, Sr and Ba. Separate Ca by extraction with boiling,
water, and Sr by dilute HN0 3 • BaSO, remains insoluble; evaporl1tion of aqueous ex-
tract gives crystals of CaSO,·2H 20. Evaporation of HNO a gives residue of SrS04,
which can·be recrystallized from conc. H 2SO,.
Mg-feathery stars developing into prisms, with P0 4--- and thetes~ drop made alkalil'\.e
with NH 40H. Fe, CO, Ni, Mn must be absent. I, ,
Zn-feathery crosses with K 2Hg(SCN),. A very little Cd renders the crosses less
feathery; more Cd leads to the formation of arrow heads. Ni modifies the
crystals in a similar manner. Co colors the crystals blue, Cu lavender or brown,
Fe pink or reddish brown. Oxalic acid forms short prisms and tablets.
Hg-globules of metallic Hg by heating in a closed tube with dry Na<CO a• Acidify a
solution with HCl, boil with a fragment of Cu wire or Cu foil. Hg amalgamates
with Cu. Heat coated Cu in closed tube-globules of Hg. Convert Hg into
HgI2 by iodine vapor-yellow and red crystals.
Hg(ous)-greenish "amorphous" ppt. with KI; white "amorphous" ppt. with HCl,
turning black with NH 40H; dark red granules or stars with K 2Cr.07 in drops
strongly acidified with HNOs.
Hg(ic)-rod HgI2 with KI, soluble in excess; colorless feathery crosses with Zn(C 2H s0 2).
and KSCN.
Al--octahedra with Cs 2SO,. Excess of H 2S0 4 must be avoided.
Sn-convert to SnCI 4; add CsCl-tiny colorless. octahedra. With oxalic acid, stannouS-
salts give oblique ended prisms and skeletal forms; stannic salts give no pre-
cipitate. .
Pb-yellow hexagonal plates with KI. Forked prismEl with HCI.· Brown or black
cubes or tablets with KN0 2 and Cu(C,H S0 2)2 in.acetic acid solution.
Sb-hexagonal plates with CsCl; Bi gives hexagons unless the CsCI is in excess. Evolu-
tion as SbHs and blackening of AgNO s crystal; As must be proved absent by
evolution of AsHs with AI and KOH; H,S, HiP interfere. \
Bi-hexagonal plates with CsCI; rhombic plates with excess. Small prisms with NIl-2S04
and H 2S0 4. .
Cr-faintly purplish octahedra with CS 2S0 4, when Cr is present as Cr.(S04).. Conver-
sion to Cr04-- by evap. with HNO s and KCIO a and subsequent testing With
AgNO a gives pleochroic Ag,Cr04' When Cr is present as a compound of CrO"
treat wi.th NaCl.and conc. H 2S0 4 in crucible with cover slide carrying a hanging
drop of NH 40H. Heat carefully; Cr02Cr. is evolved. Evaporate hanging drop;
test with AgNO s for Cr04--'
Mn-as Mn04- by oxidation with (NH 4)2S208( +AgNO a as catalyst). Or fuse with
Na,CO s and KNO s to give green Na 2Mn04, purple on acidi:(ication.
Mn++-long twinned prisms and six-armed radiates with H 2C.0 4.. Excess of NH4 salts
and mineral acids must be absent. Masked by many other cations.
Fe-blue color and precipitate with K,Fe(CN)6. Red color with KSCN if Fc+++.
Co-blue prisms with K,Hg(SCN)4. Excess of Fe or Ni interferes. When Zn is present,
blue feathery crosses or arrow-heads; organic acids interfere. KN0 2and HC2H sO.
-yellow cubes; Cu, Pb, Ni, Fe interfere.
Ni-fine pink needles with dimethylglyoxime. Tiny brown cubes.... ~ith KNO"
Pb(C,H s0 2). and HC 2H s0 2; Cu, Co must be absent; excess of Fe interferes.
Cu-greenish yellow mossy dendrites with K2Hg(SCN), (lavender or black spherulites if
Zn is present). Tiny black cubes with KN0 2, Pb(C 2H sO,). and HC 2H s0 2
(Ni, Co must be absent). Yellow prisms or orange plates with CsCI.
Ag-" amorphous" precipitate or tiny octahedra with HCl, larger on recrystl111ization
from NH 40H. Red pleochroic acute rhombic plates with K 2Cr 207. Red
granules and hexagonal tablets with AsO,---.
CHEMICAL MICROSCOPY 2445
ANALYSIS BY ESTIMATION
Paper fiber analysis is carried out by an analogous method, after pulping and stain-
ing with Herzberg's reagent. The fibers are tallied in terms of their length, measured
with reference to the diameter of the field. An alternative method ("dot-count") is
based on counts taken as the preparation is moved beneath the microscopEf~~' Every fiber
any portion of which passes beneath a given point (dot or cross haifs) 'In the field is
tallied, the longer ones thus being given proper, emphasis. Counts in different places
and directions on the preparation, dealing with at least 200 fibers on each of three slides,
are customary.
22 Booze and Klein, Jour. Amer. Ceram. Soc., 6, 698 (1!l23).
23 Insley, ibid., 10, 651 (1927). .
:1 Gabriel and Cox, Amer. Mineral, 14,290 (1929).
25 Larsen, Proc. A. S. T. M., 23 I, 236 (1923); Gardner, idem, 26 I, 296 (1926).
26 Coghill and Bonardi, Bur. Mines. Tech. Paper 211 (1919).
2452 CHEMICAL MICROSCOPY
Areal and linear analysis depends on the fact that if a section is taken
through an aggr~gate of heterogeneous material, the ~htercepts of the various
ingredients in the plane of the section, or in a line acros!, it, correspond to their
respective percentages by volume. The specimen should be in compact form,
so as to be surfaced or sectioned easily by cutting or grinding; alloys, rocks,
refractories, etc. are suitable for such preparation. If powdered mixtures,
such as ore concentrates, are to be dealt with they should be embedded in
sealing wax or other cementing material, and the aggregate surfaced for
measurement.
The areas of the constituents intersected by the plane of the specimen are
conveniently measured on a drawing traced on co-ordinate paper by means of a
drawing camera C" camera lucida "). A number of fields must be drawn,
especially if the material is coarse grained or not mixed uniformly. Photo-
graphs (provided they represent enough area) may be cut out along the boun-
daries of the various ingredients, and the respective areas totalled by weighing
the pieces of paper. Estimation or
comparison with a series of known sam-
ples is often feasible as a time-saver.
Linear analysis may be carried
but on drawings or photographs; the
intercepts are easily determined by
marking off those of each ingredient
on a separate strip of paper, the total
lengths finally being measured by a
ruler. Special mechanical stages' are
available, for moving the specimen
across the field of the microscope.
An individu~l screw is used for the
movement 6f each ingredient, the.
corresponding total movements being
shown on separate graduated scales.
The total distance measured should
FIG. 397. Cast Copper, Containing be lOO_:'_ZOO times the diameter of the
about 0.25% OJ.:ygenas Cu- Cu 20 Eutectic average sized graIns. ~..___
(dark). Vertical illumination. The volume percentage composi-
tion furnished by areal or linear anal-
yses will represent the percentage composition by weight only in case the spe-
cific gravities of the ingredients are practically identical. Otherwise, the volume
per cent of each component must be multiplied by its specific gravity, and the
resulting ratios converted to weight per cent. SJch a calculation gives the
relative "amounts of the different solid ingredients, but not necessarily the
chemical analysis of the sample. If these ingr(1dients are of known chemical
composition, the analysis of the mixture may be computed in terms of elements
or radicals if desired, but this is not ordinarily of ,d irect importance in work
where areal analysis is used. . _,-
In the analysis of aJloys it is sometimes desirable to ascertain the per cent
of a given element, such as oxygen or carbon, which is present in combination
CHEMICAL MICROSCOPY 2453
as an oxide, carbide, or eutectic of definite composition. In such a case, the
volume per cent of the compound is multiplied by the per cent of the required
element contained in it. '
Examples of applications of areal methods of analysis are: oxygen in Copper 27
(see Fig. 397), composition of alloys,28 ores,z9 and rocks. ao
BIBLIOGRAPHY
In addition to the works listed below, numerous references to others and to the
original literature are given in the authors' "Handbook of Chemical Microscopy,"
Vols. I and II.
,
General Principles, Methods, and Appsratus
Beck, "The Microscope," (R. and J. Beck, London, 1938).
Chamot and Mason, "Handbook of Chemical Microscopy," Vol. I (John Wiley and Sons,
New York, 1938).
Metzner, "Das Mikroskop" (Fr. Deuticke, Leipzig, 1928).
Barnard and Welch, "Practical Photomicrography" (Arnold, London, 1936).
Hind and Randles, "Handbook of Photomicrography" (G. Routledge and Sons, London,
1927).
Eastinan Kodak Co., "Photomicrography" (1935).
Herzog, "Die Unterscheidung der Flachs- und Hanff'aser" (J. Springer, Berlin, 1926).
von H6hnel, "Mikroskopie der technisch verwendeten£aserstoffe" (A. Hartleben,
Leipzig). . _; _
Kronacher and Lodemann, "Technik der Haar- und Wolleuntersuchung" (Urban &
Schwarzenberg, Berlin, 1930). ,
C.HEMICAL MICROSCOPY 2459
Laurill, "Textile Microscopy" (E. Benn, London, 1928).
Mathews,. "Textile Fibers" (Jehn Wiley & Sons, New York, 1924).
Preston, Modern Textile Microscopy (Emmott, London, 1933).
Schwarz, Textiles and the Microscope (McGraw-Hill Book Co., New York, 1934).
Skinkle, "Elementary Textile Microscopy" (Howes Pub. Co., New York, 1930).
Sutermeister, "Chemistry.of Pulp and Paper Making" (John Wiley & Sons, New York,
1928).
Technical Association of Pulp and Paper Industry, "The Manufacture of Pulp and
Paper," Vol. V (McGraw-Hill Book Co., NeW"'York, 1929).
Metallography
Desch, "Metallography" (Longmans, Green and Co., New York, 1937).
Hoyt, "Metallography," 2 Vols. (McGraw-Hill Book Co., New York, 1920).
Sauveur, "Metallography and Heat Treatment of Iron and Steel" (McGraw-Hill Book
Co., New York, 1936).
Williams and Homerberg, "Principles of Metallography" (McGraw-Hill Book Co., New
York, 1935).
Miscellaneous Technical Applications
Erdmann-Konig, "Warenkunde," 2 Vols. (J. A. Barth, Leipzig, 1925).
Garner, "Industrial Microscopy" (Pitman & Sons, London, 1932).
Hager-Tobler, "Das Mikroskop und seine Anw.endung" (J. Springer, Berlin, 1925).
Hanausek-Winton, "Microscopy of Technical Products" (John Wiley & Sons, New York,
1916).
Lindsley; "Industrial Microscopy" (Wm. Byrd Press, Richmond, 1929).
Molisch, "Mikrochemie der Pflanze" (G. Fischer, Jena, 1921).
Poschl, "Technische Mikroskopie" (Stuttgart, 1927).
Sasserath, "Mikroskopische Warenprtifung" (S. Hirzel, Leipzig, 1910).
Schneider-Zimmermann, "Botanische Mikrotechnik" (G. Fischer, Jena, 1922).
Tunmann-Rosenthaler, "Pflanze mikrochemie" (Borntraeger, Berlin, 1931).
Wallis, "Analytical Microscopy" (Edw. Arnold and Co., London, 1923).
Wilson, "Chemistry of Leather Manufacture" (Chemical Catalog Co., oNew York, 1928;
1929).
QUANTITATIVE MICROCHEMICAL ANALYSIS 1
Introduction.-.l\1icrochemical analysis may be defined as the determiPation
of very small amounts of constituentR when using minute amounts of ID&terial
The weight of the sample used for an analysis ranges from 3 to 15 milli~rams.
\Vhere only small quantities of material are available micromethods mjlst, of
necessity, be used, but even where large 8smples could be used the speei:l with
which a microanalysis can be carried out, without sacrifice of accuracy, }:ias led
many men in the fields of both pure and applied chemistry to discard some of the
time-consuming macro methods. An analysis can be completed in many cases
in less than an hour and very few commme over an hour of the analyst's time.
In many cases, and especially, in organic analysis, it is necessary to coml)ietely
oxidize the material in order to obtain accurate results and the advantage here
of a small sample over a <large one is at once apparent. Anyone whO will
strictly adhere to the established rules and customs of good analytical tecllOique
can acquire within a very short time the skill required in microanalysis. 1rlany
chemists, especially in this country, are not using micromethods becaus~ they
have the false impression that skill can be acquired only after long practi~ and
that the procedures are for the analysis of substances of which only small
quantities arc available. It is hoped that this impression will be overcome
when chemists become familiar with the micromethods and reaHze that these
are authentic and becoming commonly used in all branches of ana~ytical
chemistry. . I
I
I
.~
(
i
I
FIG. 4.00.
The balanoe, Fig. -400, is best mounted on a concrete slab 3 inches thick
which is firmly bracketed-against_a wall at a height of 3 feet. Strips of lead or
rubber are pla._?ed between the bracket and the slab in order to absorb vibration.
Freedom from minor vibration is Dot as important as a uniform temperature in
the room and absence of air em·rents. A small table 30 inches high, 12-18
2462 QUANTITATIVE MICROCHEMICAL , ANALYSIS '
inches wide, and 30 inches long, mounted OIl casters, is used in front of the
balance for holding the notebook, etc. It is made low enough so that it may
be shoved under the balance table. The operator by such an arrangement does
not lean against or jar the table holding the balance. ~t is best to place the
balance in the same room where the microwork is done so that change in
temperature in passing from one room to another is aroided. In order to
shield the balance from dust and air currents, it is best 'placed within a large
plate glass balance case 36 inches long, 15 inches wide, and 18 to 24 inches high.
A light should not be hung close to the balance but indirect lighting used wher-
ever possible or the light placed far enough away so that the heat given off does
not affect the weighing. Side doors are provided through which objects are
placed on, and removed from, the pans without raising the front door of the
balance. This prevents the breath of the operator from causing air currents
and changing the temperature within the balance case. Two hooks are pro-
vided on the left-hand stirrup in order that absorption tubes may be more con-
veniently weighed in a horizontal position.
Tlie beam is shorter than that used in the standard analytical balance and is
divided, beginning on the left-hand side, into 100 equal divisions. When a
5 mg. rider is placed in the first notch, the balance, when released, should be in
equilibrium. Each division corresponds to O.l.'of a milligram, The displace-.
ment of the rider one notch should cause a' ,change of the resting point 10
divisions. By means of the magnifying glass attached to the scale it is possible
to estimate one-tenth of a division so that/the weighing may be made to an
apparent accuracy of 0.001 of a milligram. The change in successive scale
readings of a good microchemical balance is very slight and very regular.
When the rider is moved the operator should make certain that it falls to the
lowest part of the notch and remains in an upright position. A second mag;-
nifying glass extends the length of the beam or moves with the rider-arm t,b
enable the operator to see that the rider is properly seated.
All objects which are frequently weighed such as cruci-
bles, weighing bottles, and absorption tubes, should be
counterpoised, using a small glass flask, Fig. 401, and
glass beads for tare, making a final adjustment by adding
an aluminium wire filed the 'proper amount. For light
objects, such as boats, an' aluminium wire alone may be
used. By this means, most df the weighings may be per-
formed by using only the rider. The 10'IDg., two 20 mg.,
and 50 mg. weights should be calibrated agairu,t the rider
and filed in order that corrections do not have--to be ap-
plied. All weights and counterpoises should be kept in
the balance-case, and forceps and spatulas should be kept
on a glass plate in the large glass balance case. The ..zero
point may be adjusted by mean's of the small vane at
F IG. 401. one end of the beam. The sensitivity of the b,alance
should be tested with various loads in the same way
that the standard analytical balance is tested, except that the rider is moved
0.1 mg. instead of 1 mg. I
The balance should be inspected regularly and th~.nife-edges and supports
kept clean by the use of dry chamois skin. .
QUANTITATIVE MICROCHEMICAL ANALYSIS 2463
As far as possible all objects to be weighed should not be touched by the
hands because of the moisture and grease always present and also to prevent
change in temperature. The object should not be wiped immediately before
weighing because of the removal of the film of moisture which will be again
absorbed while weighing. The objj)ct should be placed on the balance and an
approximate weighing made, .the final exact'reading being taken after the lapse
of three minutes. All ~bjects weighed should be kept free from dust, and the
hands of the operator should be kept free from dirt and body oil by frequent
washing and careful drying. It is the adherence to details, such as these, that
makes for success in microanalysis.
A small block of copper, kept in the desiccator, is used to decrease the time
for the object to attain the temperature for weighing.
It is, of course, not necessar~ for the zero point of the balance to be exactly at the
midpoint 10 or that the balance be adjusted to a sensitivity of exactly 10 divisions for
0.1 mg. However, the sensitivity should be kept between 9 and 11 divisions. It is
important that the balance maintain a zero point, constant to ± 0.5 divisions during
a series of weighings and a properly working balance will do this if the room temperature
is kept within ± 0.5 0 C. Where initial and final weighings involve an interval of
time the zero point should be recorded in each case so that when a shift in zero point
greater than ± 0.5 divisions (0.005 mg.) takes place the appropriate correction can be
made. A typical example of recording and calculating weights is given on page 2464.
The zero point is assuJl1ed to be constant and a tare is used in weighing the boat.
4. A very good description of the various types of micro balances and their use is
given by G. Gorback. 3 Air or magnetic damping is used on some microbalances which
helps to speed up weighing. However, a sufficient period (usually 2-3 minutes) for
air currents to subside must still be allowed before the final reading is taken. Balances
which claim a greater sensitivity than 0.001 mg. are no1Fpractical under ordinary labora-
tory conditions. Such instruments must have every condition rigidly controlled.
2 Roger J. Williams, Ind. and Eng. Chem., Anal. Ed., 8,229 (1936).
3 Mikrochemie, 20, 254-337 (1936); also M. Furter, Ibid., 18, 1-10 (1935).
2464 QUANTITATIVE MICROCHEMICAL ANALYSIS
Example
Sensitivity of balance for 0.1 mg. = 9.2 divisfons or 0.1 division = 0.011 mg.
Beam
Reading Rest
mg. Point
Boat and Sample 5.3 10.9 Difference iiI rest points 2.3 X sensi-
Empty boat 0.2 8.6 tivity 0.011 = 0.025 mg.
Difference 5.1.' 2.3 When the rest point of the sample is
Diff. R.P. 2.3 X 0.011 0.025 greater than the rest point of the boat, the
Weight sample 5.075 weight obtained by multiplying the differ-
ence in rest points by the sensitivity is '
subtracted from the difference of the beam
readings in order to calculate the weight
of sample taken.
Boat and Sample 5.3 6.3 When the rest point of the sample is
Empty boat 0.2 8.6 greater than the rest point of the boat, the
Difference 5.1 2.3 weight obtained by multiplying the differ-
Diff. R.P. 2.3 X 0.011 0.025 ence in rest points by the sensitivity is
Weight sample 5.125 added to the difference of the beam read-
ings in order to calculate the weight of
sample taken.
Boat a~d Sample 5.3 8.6 When the rest points are the same, the
Empty boat 0.2 8.6 weight of the sample is found by taking
Difference 5.1 0.0 the difference of the beam readings. .
Diff. R.P. 0.0 X 0.011 0.000
Weight sample 5.100
seen in Fig. 402. The total length is 30 cm. and the diameter of the vacuum
drying chamber is 3.5 cm. Ten to fifteen samples, in weighed boats and
placed on a glass plate, may be dried at one time and for this reason it is
rgcommended for routine drying. After drying, the samples are placed in the
FIG. 402.
FIG. 403.
taken. The long projections on the weighing bottle are to prevent heating of
the bottle by handling.
Drying and Weighing Hygroscopic Substances.-HygroScopic substances
which take up water slowly may be dried in a boat which is quickly placed in
the weighing bottle, Fig. 403, previously kept in the balance case. In the case
of substances which take up water rapidly and form hydrates, the method of
Milner and Sherman 5 is recommended.
5 Ind. and Eng. Chern., Anal. Ed., 8, 427 (1936).
2466 QUANTITATIVE MICROCHEMICAL ANALYSIS
FIG. 404.
the front legs, Fig. 405. The base, whose edges are curved~t(), fit the drier, is
held fixed within the drier oy another spring clip. The solid glass stopper of
the pig has a small handle with projecting knob which is engaged_pya brass
sleeve comprising a bayonet joint and which enables the stopper to be with-
drawn or firmly seated while the pig is in place. Before the sample is weighed
out, the pig and empty boat are dried and capped in the apparatus. The
sample is weighed and dried and the result recorded. ! Then the pig is opened
and the sample allowed to equilibrate with the moisture in the room. Usually,
after 30 to 60 minutes, equilibrium is so nearly approached that any further gain
in weight while transferring the sample to the combustion tube is negligible.
The pig is then closed and reweighed. The gain in weight upon exposure to
air is, subtracted from the weight of water found ll'iore the percentage of
hydroge~ is calculated. •
QUANTITATIVE MICROCHEMICAL ANALYSIS 2467
The method (see Fig. 402) of heating electrically and having the drying
agent attached closer and with a wider opening to the sample is more convenient
FIG. 405.
than in the Milner and Sherman apparatus. Samples may then be dried at any
temperature by simply changing the rheostat setting.
Weighing Liquids.-When the substance is to be oxidized, as in the deter-
mination of carbon, hydrogen, sulfur or nitrogen, a capillary tube, Fig. 40(\ A-B,
_____
•
~>---------------------<~ B
A
i
-------+--ti§( ;. E
I
I
I
FIG. 406.
is closed and drawn out. A small crystal of potas~ium chlorate, placed in the
capillary, Fig. 406 C, is melted and cooled. A hair capillary is formed by draw-
ing out the open end of lrhe tube, this is broken at a suitable point, Fig. 406 D,
and the tube weighed. The tube is then carefully heated to expel the air but
not sufficiently to melt the potassium chlorate, the open end dipped into the
2468 QUANTITATIVE MICROCHEMICAL ANALYSIS
liquid to be analyzed and allowed to cool. The excess liquid is car~fully re-
moved from the hair capillary by gently heating; after w~ich the tube is sealed
and weighed, Fig. 406 E. .
The handle and tip are broken off and the tube is placed in the combustion
tube on a narrow strip of ignited platinum. It is then quickly placed in position
by the use of a platinum or nichrome wire. I
For halogen determinations, the potassium chlorate is replaced by am-
monium nitrate.
Another method 6 of filling capillary tubes, which is very simple, consists in
placing the open end of the capillary in the liquid contained in a small sample
bottle. Both ar'e placed in a small filter flask equipped with a rubber stopper
and stopcock, Fig. 407. Suction is applied and the flask evacuated, which also
FIG. 407.
evacuates the capillary. This takes only a few seconds. The stopcock is now
slowly opened and the difference in pressure between the outside and inside of
the capillary forces the liquid into it. I
the filter attached to a suctiQn flask by means of a glass tube and rubber con-
nection. The filter is placed in the beaker with the head just above the coagu-
lated precipitate and the liquid withdrawn into the suction flask, Fig. 411. In
the same way the precipitate is wa~hed as prescribed and the wash water re-
moved. The filter is disconnected and remains in the beaker and they are
both dried and weighed as before.
If the precipitate must be ignited then the filter and beaker are made of
platinum, porcelain or quartz and the filtering medium is
platinum sponge, porous porcelain or sintered quartz. A
thin porcelain crucible of 10 ml. capacity serves well as a
beaker. Where a volume of solution greater than 10 ml. must
be handled a micro beaker developed in this laboratory, Fig.
412, with a ground-glass joint, so that the top section may be
removed, is useful. The precipitation is carried out as usual
and then the filtrate removed by the immersion filter. The
bottom section with the filter stick, which has been previ-
ously weighed, is now detached, dried, and reweighed with the
pr_ecipitate. The total weight of the bottom section and filter
stick is 8-10 gms.
Where ignition is required it is best accomplished in an
electric muffle furnace heated to 1100-1200° C. and the beaker
with the filter is placed in a covered unglazed porcelain con- FIG. 412.
tainer. After ignition, cooling takes place in the open air for
~ minutes, then in a desiccator with drying agent, and after 25 minutes they
are weighed.
In all the microgravimetric'determinations to be described the immersion
2472 QUANTITATIVE MICROCHEMICAL ANALYSIS
electric heating for that section of the tube containing the catalyst or oxidizing
agent, but burn the sample with the gas burner.
The electric furnaces, Fig. 414, for the carbon and hydrogen apparatus are
constructed by winding nichrome wire on a 6 Y2 and two 3 inch sections of a
Y2 inch inside diameter porcelain tube. The 3 inch furnace used to heat the
lead dioxide' is wound with No. 30 wire in three layers separated by alundum
cement and asbestos paper. The other two furnaces have two layers of No. 25
wire. The wound cores are hardened by drying at 1500 C. The ends of each
furnace are made of circles of Ysft Transite, 3 inches in diameter. Each furnace
is covered with an aluminum or brass shell. Rheostats are used to adjust the
furnaces to proper temperature. Ammeters may be put in the circuit to see
that the proper amount of current is passing through. Thermocouples may be
embedded in each furnace and the temperature may be automatically regulated
if the voltage varies. It is best that the, lead dioxide furnace be regulated to
±0.2° C. because variation will cause the material to give off or take up water.
The furnaces, Fig. 423, for Dumas nitrogen are of the split type. When
the.lower half,swings down the upper section may then be raised slightly and
the furnace pushed away from the combustion tube. The furnace may then
be closed and thus kept up to the proper temperature for the next determination.
In the drawing the furnace N is shown in the position away from the combustion
tube, and furnace 0 is shown in the position over the combustion tube. The
furnace N is provided with two notches, one of which shows at V, the other at
the same relative position, which fit on the rail X so that this furnace may be
• pushed along the combustion tube.
The furnaces for the halogen-sulfur determination, Fig. 428, are also of the
split type, but cannot be shoved back from the combustion tube. The upper
half swings open and automatically locks in position. The combustion tube
may then be removed and placed on the supports on the front of the apparatus
to cool. By pushing the lever at the back the furnaces may be closed and
thereby kept up to temperature for the next run.
The construction is shown by the drawings. The furnaces are 3 inches and
7 inches in length and 3%, inches in diameter. The ends are made of transite.
The shell is made of aluminum. Fig. 428 with the top section of the small
furnace'removed shows the details of construction of the heating elements. The
heating element of each half consists of two coils of No. 25 nichrome wire sup-
ported on Ys inch alundum rods. This in turn is held loosely in place by
notched porcelain semicircles which are cemented to the alundum refractory.
This refractory consists of a stock alundum tube 1 inch inside diameter, sawed
in two. The top and bottom nichrome heating elements are connected in
series to binding posts on the transite e~ds. The heating elements are easily
replaced at any time. The furnace temperature is measured by a thermocouple
and, if required, suitable resistance is inserted.
For laboratories having a large number of routine organic combustions,
the split type electric furnace unit 12 shown in Fig. 413 can be used to advantage.
The apparatus is fabricated from aluminum and stainless steel. The drawing
shows the small burning furnace, tilted back corresponding to a position away
12Irul. Eng. Chem., Anal. Ed., 10, 110, 111 (1938). This paper gives details of the
application of the apparatus to the determination of carbon and hydrogen, nitroge~,
sulfur, and halogen. .
2474 QUANTITATIVE MICROCHEMICAL· ANALYSIS
from the combustion tube which is not shown, ana the large furnace in a position
over the combustion tube. A small electric motor with 'gear reduction driv{)s
a screw which moves the sample burning furnace along 'the cOII\bustion tube.
When the small furnace is brought in the burning positic;d' over the combustion
tube, it meshes with the screw, the motor starts and a brush contact, attached
to the bottom, passes over it series of insulated metal segments as the furnace
moves along. These segments, connected through suitable resistances, auto-
. matically vary the speed to give slow initial burning and accelerated burning
after the sample has carbonized. The furnace automatically stops when it has
reached a predetermined point and the sample has been burned. Any com-
bination of speeds can be obtained by varying the number, positions: and length
of the segments. The heating elements are the same as in the hand operated
furnace.
FIG. 413.
The method 13 described gives a high degree of accuracy together with speed,
and can therefore be used to advantage in routine work.
The procedure is similar to the combustion method which is known to all
those. who have had occasion to determine carbon and hydrogen in organic
comp0'\lllds. The results obtained are as good as, and in many cases better
than, those obtained by using the macro method. Results averag91 ±O.l %
from theory for compounds containing carbon, hydrogen and oxygen. Those
-containing sulfur, halogen and nitrogen vary ±O.2% from theory.
FIG. 414.
filled with cohcentrated sulfuric acid by dipping the end into a beaker of the
acid and applying gentle suction on the drying' tube U. The excess acid is
then removed from the end by carefully heating over abuf!1er. The drying tube
is charged two-third§! of its length with" Ascarite" (Trade name for sodium
hydroxide on asbestos), fine glass wool being placed a~ bo.th ends (Corning
standard wool No. 008); then a layer of phosphorus pentoxide mixed with
fine glass wool; and at the end another short layer of glas~ wool. The ground-
glass joint is then heated, the two parts coated with a small amount of glass
cement, and after the joint is firmly pressed together it is allowed to cool.
A rubber stopper is placed on the tapered capillary end of this joint. The
pyrex, or preferably quartz combustion tube, 53 cm. in length and 7 mm.
inside diameter, has a capillary opening 1.5 to 2.0 cm. in length. The com-
bustion tube is charged, beginning at the capillary end, with a short wad
of ignited asbestos, then silver wool, 1.5 cm.; powdered lead dioxide mixed
with asbestos or granular lead dioxide, 4.5 cm.; silver wool, 3 cm.; finely broken
copper oxide wire impregnated with lead chromate (by igniting and sprinkling
with lead chromate which fuses on the copper oxide), 15 cm.; silver wool,
3 cm. jn ignited asbestos wad is placed between each of these charges
to prevent shifting. Care must be taken not ,to pack so tightly as to stop
the flow of gas through the combustion tube, Silver removes any halogen
and sulfur present, lead dioxide reacts with the oxides of nitrogen, and lead
chromate also removes oxides of sulfur. Lead dioxide is heated by the electric
furnace G to 1800 C., copper OXIde to 700..::750 0 C. by the furnace F, and the
third furnace E is employed to burn the material in the boat, the temperature of
which if necessary may attain 10000 C. The temperatures of the first two
furnaces are accurately measured by means of a thermocouple, and as each
furnace has a rheostat and a means of reading the current by a small ammeter,
it is possible to regulate the temperature accurately. The control of the tem-
perature of the lead dioxide is very important as fluctuations in temperature
cause it to give up or absorb water vapor. The furnaces are placed on two rods
so that they may be moved along the combustion tl:lbe at will. At the point H
a small electric heater is placed in order to prevent condensation of moisture.
The heat applied should be regulated so as not to damage or destroy the rubber
connection. The combustion tube when freshly charged is kept at full heat for
not less than six hours while a slow stream of oxygen is passed through the tube
in order to burn organic matter and free the tube and filling from absorbed
gases, such as water vapor. '
Absorption Tubes.-Because the oxygen used in the combuStion is not
swept out with air and since phosphorus pentoxide is used for the absorption of
water, the absorption tubes must be capable of being sealed in order to prevent
the escape of oxygen or the entrance of moisture during weighing.
The absorption tube, Fig. 415, is made from a 6 mm. pyrex ground-glass
joint. The main body of the tube is enlarged until1t is 7 cm. in length and
1.3 cm. in diameter with thin walls to keep the weight as small as possible.
The mercury bulbs C at each of the tubes are identical, the internal diameter
being 3-4 mm. The internal' diameter at A and B is 0.2 mm. As an added
precaution bulbs 2.5-3.0 mm. are blown between the points Bland D to catch
any mercury which might be carrIed through. The..stPiiight part of the tube D
has an internal diameter of 1 mm. and for a length-of 1'.5' cm. an external diam-
QUANTITATIVE MICROCHEMICAL ANALYSIS 2477
eter of 3 mm., which fits the special micro tubing used. In Fig. 415, F is solid
glass to strengthen this part of the tube. The.mercury is shown sealing the
capillaries. When the tube is rotated 1800 and shaken slightly, the mercury
f~lls and the tube is open. The ground-glass joint is sealed with glass cement.
c
B
E
Mer,ury
FIG. 415.
The total length of the carbon dioxide absorption tube is about 12 cm. and
that of the water absorption tube slightly less. It is very important that these
tubes be washed in chromic-sulfuric acid mixture, and then with water, fol-
lowed by alcohol and finally ether, so that the main body of the tube is clean
'arid also the capillaries on eit~er side of the mercury chamber.
A modified construction for these tubes has been described.1 4 The mercury
trap and seal are placed within two hollow glass stoppers which are ground to
fit the ends of a Pyrex tube 1 cm. in diameter, Fig. 416. The tube is shown
FIG. 416.
unsealed in the figure. When the tube is rotated 1800 the mercury drops seal
the capillaries.
The water absorption tube K, Fig. 414, is charged first with a wad of fine
glass wool (Pyre:lf.: Standard); then phosphorus pentoxide mixed with a small
amount of fine glass wool; and then a wad of glass wool. The end of the tube is
then heated and sealed with glass cement. A small drop of mercury is then
drawn into each end by J;tleans of gentle suction.
The carbon dioxide absorption tube L is charged first with a short layer of
glass wool; then three-fourths with 20-mesh "Ascarite"; a short layer of fine
glass wool; then a short layer of phosphorus pentoxide; and finally a wad of
fine glass wool. The tube is then sealed and mercury drawn in.
The ends of the tubes in all cases must be glass to glass in order to minimize
the diffusion of gases through the tubing. The charges in both tubes will last
for 30 to 40 determinations. The" Ascarite" changes color from a light brown
to a grayish white upon absorbing carbon dioxiae. A guard tube M, filled
witli phosphorus pentoxide or "Dehydrite," is placed at the end of the carbon
dioxide absorption tube and the guard tube is attached to the Mariotte flask.
The Mariotte flask or aspirator bottle S is used to measure the amount of gas
passed through the combustion tube and aids in keeping the system at atmos-
14 Cornwell, Ind. Eng. Chern., Anal. Ed., 3, 4 (1931).
2478 QUANTITATIVE MICROCHEMICAL ANALYSIS
pheric pressure, and thus minimizes diffusion of 'gases • through the rubber
connections.
By means of a three-way stopcock N between the flask and guard tube the
speed of the oxygen can be regulated. It is also possible by means of stopcock
N to open this line to the atmosphere before removing the absorption tubes
and thus create an outward pressure which prevents the absorption of moistur~
and carbon dioxide from the air. By opening stopcock 0 to the air, and attach-
ing tube P to a ,water tap, it is possible to fill the bottle without removal from
the stand. Tube R runs to the sink and is constricted, or may have ~ stopcock,
so that while the flask is being filled, water will not readily escape..
Procedure.-:-The furnaces are turned on and also the small heater H; the
absorption tubes are wiped with a damp cloth, dried, and after unsealing (by
rotating 180 0 and tapping lightly), are attached to the combustibn tube. 'l'he
oxygen is turned on and the pressure in the ThIariotte flask regulated so that the
gas passes at the- rate 9f 10 ml. per minute, and is allowed to pass while a 3 to 7
mg. sample is being weighed in a cleaned and ignited porcelain or platinum
boat, suitable precautions, being taken if the substance is hygroscopic Or a
liquid. The heater II is turned off 5 minutes before the absorption tubes are
removed. The stopcock N on the oxygen line T of the ThIariotte flask is turned
so that the oxygen flows into the air. The.guard tube is removed while oxygen
continues to flow and then the" Ascarite" tube L is removed. The tube is
rotated 180 0 and is tapped lightly causing the mercury to seal the capillaries.
The ph.osphorus pentoxide tube is handled in a similar manner. In remoVing
the tube only the ends and not the main body are touched, and these are CEt,re-
fully wiped each time. It is only necessary to wipe the main body of the
absorption tube at the beginning of the day's run to remove accumulated dust.
The phosphorus pentoxide tube is placed on the balance, ·its approximate
weight obtained, and is allowed to stand 10 minutes before the final reading
is taken. The other tube is ,then weighed in the same manner, after remaining
3 minutes in the balance. "\Vhile the first tube is on the balance the boat Con-
taining the weighed sample is placed in the combustion tube by means of
platinum-tipped forceps and a clean platinum or nichrome wire. The flow of
oxygen now sweeps out any moisture and carbon dioxide which has entered.
The absorption tubes are attached, the flow of oxygen is regulated,_to 5-6 m!. per
minute by the stopcock N, the combustion furnace is turned on, and the sample
is burned by slowly moving the furnace over the boat. The small heating t::oil
at the capillary end of the combusti.on tube is turned on to prevent the Con-
densation of :~Yater vap9r._ The actual burning or the sample takes about 5--10
minutes and', another 20-30 minutes is allowed to sweep .out the products of
combust job., 100 ml: of oxygen during the first 15-20 minutes, another 200 ml.
in 20 minut~s. In most cases a total of 200 ml. of oxygen is sufficient. 1:'he
correct amount sho!;Ild be checked by running standard samples. The absorp_
tion tubes are then removed, sealed, and weighed as' before.
}Vhile these are coming to equilibrium in the balance, the duplicate sample
is placed in the tube and after the tubes are weighed and attached, another
analysis is made in the same manner. From the weights of carbon dioxide B.nd
water the percentage .of c'arbon and of hydrogen in the sample is calculated.
I
QUANTITATIVE MICROCHEMICAL ANALYSIS 2479
NOTEs.-(l) The sample should not be oxidized so rapidly that the gases formed are
forced into the tube D. A capillary is placed in the tube to help prevent this. .
(2) The current of oxygen should be carefully regulated so that the gases formed may
pass over the several fillings at such a rate that they are completely oxidized and
absorbed.
(3) The systeI;;,\ should be tested for leaks by turning off stopcock N and observing
the bubble counter C.
, (4) The blank test is very important. This is carried out in exactly the same manner
as in an ordinary combustion except that no sample is placed in the boat. If the absorp-
tion tubes show an increase of 0.1 mg. or IJ;lore, the apparatus should be inspected for the
following; leaks, el'cess glycerol on the rubber connections,
insufficient time allowed for burning and conditioning a
new tube filling, or oxygen containing impurities. Before
beginning a series of determinations a standard should
be run, preferably containing those elements present in the
samples to be analyzed.
(5) The absorption tubes and drying tube D are refilled
by heating the ground glass joint, which melts the glass
cement and allows the tubes to be emptied and refilled. a
(6) Because phosphorus pentoxide and" Ascarite " are
used, oxygen may be passed through the apparatus at a
more rapid rate than that recommended by Pregl. For
compounds containing much nitrogen and sulfur the rate
should be slower, not more than 5 ml. per minute, so that
the oxides of nitrogen and sulfur will be reacted upon by
the reagents in the tube.
(7) The absorption tube next to the furnace should
always be placed with the ground glass joint away from the
furnace to avoid softening of the cement.
(8) The oxygen may contain organic impurities in
which case a preheater is used consisting of a combustion
chamber filled with platinized asbestos or CUO heated to
650-700° C. by an electric furnace or burner. The gas
is usually cooled by passing through a glass coil immersed b
in water so that it will not decompose the rubber con-
nections. The gas is then passed through" Ascarite " and
phosphorus pentoxide before entering the combustion tube.
A common type I. is shown in Fig. 417. Another simpler'
form made of Pyrex ,and without a ground_glass joint was FIG. 417.
desigiled by White and Wright.I. "The apparatus is
constructed, Fig. 418, from 10 mm. Pyrex tubing, into which two 4 mm. tubes, AB
and BC, are joined by ring seals at B. The gas inlet and outlet being attached, the
bottom of the tube is sealed off, and, in operation, may be immersed in a beaker of water
to cool the burned gases. Into the annulus around BA is packed 1.5 em. lengths of
coarse fibre asbestos, and then fluffy platinized asbestos. The top of BA is then loosely
packed with coarse fibre asbestos and the top of the 10 mm. tube is sealed off. An elec-
trical heating unit surrounds the top of the preheater. It should not fit too tightly
and should be closed at the top with a stopper of asbestos paper which can be removed
• in order that the correct (dull red) temperature may be ascertained."
(9) Where compounds contain arsenic, antimony, tin, bismuth, or phosphorus, the
following procedure 17 is recommended which is a modification of the Falkov and Raiziss
procedure. '8 A portion of the copper oo:ide-Iead chromate filling of an ordinary Pregl
Universal filled combustion tube is removed and replaced by a 3 cm. cylinder made of
200 mesh platinum wire gauze filled with granulated lead and a 1 cm. plug made of
I. Nieder! and Roth, Ind. Eng. Chern., Anal. Ed., 6, 273 (1934).
16 Can. Jour. Res., 14, Section B, 427 (1936).
17 F. C. Silbert and W. R. Kirner, Ind. and Eng. Chern., Anal. Ed., 8, 353-5 (1936).
18 J. Am. Chern. Soc., 45, 998-1003 (1923).
2480 QUANTITATIVE MICROCHEMICAL ANALYSIS
platinum wire gauze. The red lead is prepared by igniting Merck's granular lead per-
oxide in a stream of oxygen in a microcombustion tube in an ele'ctric combustion furnace
at the normal combustion temperature. To prevent any red lead dust from sifting
through the platinum gauze a thin layer of asbestos lines the entire inner surface of the
, cylinder, including the ~nds. This snug fitting
- - - - - - -....- cylinder enSUres more intimate contact of the
combustion gases with the red lead than if the
latter is merely placed in a boat as recommended
by Falkor and Raiziss. The details of the filling
are shown in Fig. 419.
(10) Analysis' of Coal for Carbon and
Hydrogen.
E No special technique is required, but the
E
sample must be finely ground (about 100 mesh)
in an agate mortar. For accurate analysis, it is
necessary, of course, to dry the sample, weigh in
a weighing bottle, and then transfer it quickly to
the combustion tube before it takes up moisture.
Moisture cannot be determined by micro
methods because this figure should represent a
large mine fun sample. The sampling of the
coal is, likeWise, done in the usual macro way.
The final sample which is submitted to the
analyst is then taken and prepared for analysis .
. Kirner 19 gives the following example: 50 gms.·of
German bituminous coal in fairly .large pieces
were crudely ground in a mortar, sample A.
Five grams of sample A were pulverized for 5
minutes, sample B. Sample B did not give con-
cordant results but on taking 1 gram of sample B
and pulverizing in an agate mortar for 5 minutes,
it gave satisfactory checks. Lucas and Grassner 20
take the final 125 gm. sample, grind to 100 mesh,
and take 10 mg. of this for a determination.
(11) Solid compounds which are explosive
when heated are mixed in the boat with fine
copper oxide or some inert material such as
ignited finely ground ;iilica.21
(12) Wherel compounds are run which con-
tain carbon, hydrogen and oxygen, the tube 'fill-
ing can be much simplified. Only platinized
asbestos, or better, platinum contacts are re-
quired to attain complete combustion. This
filling is heated by the furnace which heats the
lead chromate-copper oxide ·mixture. Because
nitrogen is absent, no lead dioXide is required.
c (13) If the apparatus has been standing for
a week or so, it is best to burn in the combus-
FIG. 418. tion tube a few crystals of sugar or, preferably,
a few milligrams of a compound containing the
elements present in the compounds to be analyzed. It.sometimes happens that the
first run is low in hydrogen due to the lead dioxide a\isorbing moisture. By first
burning an organic compound, the water-lead dioxide equilibrium is made constant.
(14) A side inlet combustion tube is often preferred, especially for ~arbon and hy-
drogen work. _ A solid rubber stopper is then used.
(15) Analysis of Gases for Carbon and Hydrogen.
The method essentially consists of measuring the gas accurately in a gas burette,
19 Ind. Eng. Chern., Anal. Ed., S, 363-9 (1933). . _ . r -
20 Mikrochem., 6, 116-32 (1928). -
,21 M. Furter and J. L'Orange, Mikrochemie, 17,38-42 (1935).
'\
mixing it with oxygen, 'and then passing it into the combustion tube. A special tube
with side inlet, Fig. 420, filled with CO 2 absorbent, such as " Ascarite," is placed between
the pressure regulator and the drying tube with bubble counter. The measured gas
sample, plus an excess of oxygen contained in a t!;as burette, is attached to the inlet A
by a tight glass to glass connection of capillary tubing. The apparatus is swept out with
air or oxygen. The oxygen from the pressure regulator is cut off. The gas sample
FIG. 419.
mixed with oxygen is now admitted from the gas burette at the same rate as in a regular
combustion. When all the gas is out, the burette cock is closed and the method of
sweeping out is the same as with the standard procedure.
(i6) Many absorption tubes have been designed which can be sealed. Those
which have a ground-glass joint, which must be turned and lubricated, may give diffi-
culty due to loss of lubricant when wiped. The use of copper wires in the capillaries of
. FIG. 420.
the Pregl tube, Fig. 421, to help prevent entrance of moisture and carbon dioxide have
been suggested by Boetius 22 and used by Niederl and Roth.23 The tubes however are
fi!led with air when weighed.
The objection raised to a sealable tube is that changes in temperature and pressure
affe'<,t the constancy. The author has found, however, that during 45 minutes, which
FIG. 421.
is the time taken to make a carbon and hydrogen determination, it is rare that there is
so great a change in pressure as to affect the accuracy of the determination. The tem-
perature of the laboratory'does not vary enough to make it a factor. For routine work,
a sealable tube which can be weighed while filled with oxygen is much more convenient.
The wiping at the end of each run of the main body of the absorption tube has not been
22 Uber die Fehlerquellen bei mikroanalytischen Bestimmung des Kohlen-und
Wasserstoffes nach der methode von Fritz Pregl, p. 27, verlag chemie Berlin 1927.
• , 23 Ind. Eng. Chem., Anal. Ed., 6, 272 (1934).
2482 QUANTITATIVE MICROCHEMICAL ANALYSIS
found necessary. References to the different types of sealable absorption tubes are
given by A. Friedrich. 24 •
(17) The water absorption tube may be filled with a desiccant such as " Anhydrone "
(anhydrous magnesium perchlorate) and then ,finally a 72 inch layer of phosphorous
pentoxide, t? indicate when the tube m~t be refilled. Indi~ating" Drierite" (an-
hydrous calcIUm sulfate) may also be used m the water absorptlOn tube. It is impreg-
nated with a cobalt salt which changes to a reddish color on taking up water. It can
be regenerated at least 12 times by heating to 235-250° C. for t to 2 hours.
(18) Good results have been obtained by replacing a part of the lead chromate-copper
oxide mixture with a 4 cm. section of platinum contact foil or a roll of platinum 80 mesh
gauze.
(19) The lead dioxide furnace should not be shut off at the end of a series of runs.
By always keeping the filling up to temperature the lead dioxide water equilibrium is
more readily maintained.
(20) One combustion tube filling will last for 200 determinations. If halogens are
frequently run the renewal of the silver wool nearest the boat after 20-30 such deter-
minations is desirable.
(21) Compounds which are salts of alkalies or alkaline earths must be mixed with
dry pulverized potassium dichromate in a porcelain boat in order to prevent the forma-
tion of carbonates.
The carbonate is decomposed with hydrochloric acid and the evolved carbon
dioxide absorbed and weighed. The method 25 is accurate when samples con-
taining not less than 1.5-2.0 mg. of carbon dioxide are used; with smaller
amounts of carbon dioxide, the error may reach 2%. The values for carbonate
carbon vary ±0.3-0.4% from t h e o r y . ' ,
Apparatus.-The apparatus, Fig. 422, is construcled of Pyrex glass. A
rubber stopper is used in the 50 m!. extraction flask A, so that flasks may be
interchanged. The water condenser is approximately 20 cm. long with the
central tube beveled on the lower end B, so that it will not be closed by drops
of condensed water. The trap P containing sulfuric acid, is sealed between the
condenser and drying tube,R, which latter contains the solid. drying agents to
absorb most of the water before it enters R. This trap also indicates. the rate of
flow of gas through the apparatus. It can be readily emptied by _Pleans of the
stopcock at the bottom and refilled through the same by applying ,suction to
the apparatus. The drying tube, R, 26 cm. long"is made of a 1.7 cm. ground-
glass joint, which is sealed with glass cement.
The drying tube is filled with magnesium perchlorate trihydrate, "Dehy-
drite" or anhydrous magnesium perchlorate, " Anhydrone.': This is powdered
so that about half passes through a 20-mesh sieve, the two portions being kept
separate. The tube is charged first with a wad of fine glass wool, then a 1 cm
l~yer of the coarser "Dehydrite"; fine glass wool; 1 cm. layer of anhydrous
24 Mikrochemie, 19, 23--37 (1935). . -' ::.. -
25 Kemmerer and Hallett, Ind. Eng. Chem., 19, 1352 (1927).
QUANTITATIVE MICROCHEMICAL ANALYSIS 2483
copper sulfate to show the. presence otwater; fin~ glass wool and alternate 1 cm.
layers of coarse and fine" Dehydrite" or "AnhydronEi." If sulfites 26 are pres-
ent the carbonate is decomposed with an excess of potassium dichromate 5%
solution, boiled, and toward the end of the procedure, a small amount of dilute
sulfuric acid is added. •
The acid bulb, W, supplied with an Ascarite-filled guard tube, T, to prevent
absorption of carbon dioxide from the air, is attached to the flask A through a
FIG. 422.
three-way stopcock. The air passed through the apI)aratus to remove carbon
dioxide is first drawn through a sulfuric acid trap S, which is charged with acid
through the funnel and may be emptied by disconnecting it at the rubber con-
nection and blowing out through L. The air then Passes through the tube D
which is fil~ed with "Ascarite" to absorb any carbcm dioxide present. The
constriction N retards the diffusion of the carbon dio)(ide liberated in A. The
whole apparatus is easily attached to a single ringstand.
26 E. R. Marle, J. Chern. Soc., 95, 1491 (1909).
248.4 QUANTITATIVE MICROCHEMICAL ANALYSIS
() 10
~
SCALE·CM.
FIG. 423.
gas contain only 0.001 ml. which is not absorbed by a 50% potassium hydroxide
solution. In order to keep the acid out of contact with air the upper and middle
chambers are connected with glass tubing containing ground glass joints. The
acid used is made by mixing equal parts of freshly boiled distilled water and
concentrated hydrochloric acid. The marble chips used are etched in dilute
hydrochloric acid and then washed with water before they are placed in the
generator, in order to remove absorbed gases.
When the generator is assembled and charged, the upper chamber is freed
* For an automatic combustion apparatus, see p. 2473 and ref. 12.
28 Trautz, Mikrochemie, 9, 300 (1931); Trautz and Niederl, Ind.Eng. Chern., Anal.
Ed., 3, 151 (1931).
2486 QUANTITATIVE MICROCHEMICAL ANALYSIS
from air by adding a few lumps of marble, the U tube A is put in place, and the
upper B and middle chamber C connected by a glass to ~lass rubber connection
• at E. Carbon dioxide is allowed to flow from the middle C to the upper chamber
until there is a positive pressure shown by the mercury in the U tube which
seals the upper chamber from the atmosphere. An excess pressure should
exist within the upper chamber at all times and the U t',ube charged with me'i--
cury allows the pressure or partial vacuum to be read. A column of mercury
5 cm. high in one arm is sufficient to take care of the variations in pressure.
Before an analysis and at other times when necessary, carbon dioxide is passed
from the middle to the upper chamber by turning stopcock F until gas escapes
through the valve. By removing the U tube spent acid may be siphoned off
and fresh acid added.
A gasometer K containing mercury is placed between the generator and
combustion tube in order to measure the volume of carbon dioxide used during
a determination. Specially treated and aged rubber tubing previously de-
scribed (page 2469) is used for glass to glass rubber connections at the points
H, P and R. The diameter of the glass tubing at these points is 4 mm. The
gasometer is connected at M to the combustion tube by a capillary tube,
tapered at the end to accommodate the rubber stopper which fits into the
combustion tube.
Combustion Tube.-A Supremax or preferably quartz combustion tube,
50 cm. in length, and 7 mIll. inside diameter has a'l to 2 mm. capillary opening
1.5 to 2.0 cm. in length. This tube is the same as that used' for the determina-
tion of carbon and hydrogen.
The material used to charge the combustion tube is first ignited. The
tube is permanently charged, beginning at the capillary end, with a short wad
of asbestos, then broken copper oxide wire, 9 cm.; a short wad of asbestos, then
4 cm. of freshly prepared broken copper wire. This is made by passing hydro-
gen (washed with a solution of potassium permanganate acidified with sulfuric
acid) over heated broken copper oxide wire and cooling in a current of hydrogen.
A wad of asbestos is added, then 9 cm. of broken copper oxide wire. This is
held in place with a wad of asbestos. Before using, the charged tube is pre-
heated in a current of carbon dioxide and then allowed to cool.
Nitrometer.-The precision micronitrometer :S, Fig. 423, has a capacity of
1.5 rol. and is read to the nearest 0.001 ml. It is charged by adding mercury
until it covers the gas inlet. Fifty per cmit caustic potash solution made by
adding 100 gm. of potassium hydroxide to 100 gm. of water, -is tlten added by
the funnel to fill the measuring tube.
Procedure.-A sample which will yield 0.3 to 0.5 ml. of nitrogen is placed
in a dry, weighed porcelain boat, dried, and then reweighed. If the material is
in large crystals, it should be ground in an agate mortar. Fine freshly ignited
copper oxide (prepared by grinding broken copper oxide wire until it passes a
60 mesh sieve) is added to the combustion tube until;a layer of 2 cm. is formed,
then the boat containing the sample is added, and the tube gently shaken and
rotated until the sample is mixed with a fine copper oxide. Anoth~r 2 cm. of
fine copper oxide is added and then a 4 cm. layer of coarse copper oxide. The
end of the tube is wiped with absorbent cotton to remove any' fine copper oxide
which would prevent the rubber stopper from making 3; tight fit.
QUANTITATIVE MICROCHEMICAL ANALYSIS 2487
A blank determination using pure cane sugar should be made when the
procedure or apparatus is modified, in order to determine the volume of un-
absorbable gases which enter during each run. A stand~rd sample should be
run before making a" series of determinations in order to see that the apparatus
is working properly. Liquid materials are weighed in a sealed capillary, the
tip broken, and after adding 2 cm. of fine copper oxide, the capillary is dropped
into the tube (brol>:en tIP first). It is then covered with another 2 cm. of fine
copper oxide and finally with 4 cm. of coarse copper oxide.
. After introducing the sample, the combustion tube is attached to the
apparatus and the air is eliminated by passing through the system about 100 ml.
of carbon dioxide, while cold. At the end of this operation, which requires one
minute, the nitrometer is attached and small micro bubbles should appear in
the nitrometer. ,Bubbles of proper size are recognized by their uncertain rise
so that if the gas flow is sp~eded up the streaJ7l of bubbles which are no larger
than pin points have a tendency to turn over and go toward the bottom of the
nitrometer. If the bubbles are not of micro size, it shows that the air has not
been eliminated or that a leak in the system is present. In the 'next operation,
the actual eombustion stopcock L of the gasometer is closed, both furnaces are
brought over the tube, and the combustion tube is heated to red heat (700-
750 0 C.). Care must be taken that furnace N does not come too close to the
sample and cause premature combustion. After the bubbles have ceased to
come over, due to the heating of the tube, any gas which has collected in the
nitrometer is now released. The burning of the sample is started by gradually
moving furnace N nearer and flearer to the furnace O. The rate at which the
bubbles enter the nitrometer must not exceed 2 bubbles in 3 seconds and this
rate is carefully regulated by the proper movement of the furnace N.
During the combustion the gasometer is refilled with carbon dioxide. When
no more bubbles rise in the nitrometer, the combustion is at an end and the
gaseous combustion products are conveyed into the nitrometer by a current
of carbon dioxide; stopcock L is carefully op'ened and adjusted to give a rate
of one bubble per second for five minutes. The rate of carbon dioxide flow is
next increased to three bubbles per second and when the bubbles begin to
approach microsize both electric furnaces are opened and pushed away, and
the tube allowed to cool. Furnace N is shown in position away from the com-
bustion tube while 0 is in the position over the combustion tube. Carbon
dioxide is passed through the system until micro bubbles of the same size as
at the beginning of the run are obtained. The levelling bulb is raised, attached
to a holder, and the nitrometer disconnected.
After 15 minutes the readings are taken of the nitrogen volume in the nitrom-
. eter with the aid of a magnifying lense to within 0.001 ml. with the potassium
hydroxide in the levelling bulb at the height of the nitrometer meniscus, and of
its temperature to within 0.10 C. with the bulb of the thermometer touching
the wall of the measuring capillary, and of the barometric pressure to within
0.1 mm. The time required from the filling of the combustion tube with the
sample and fine copper oxide to the completion of the run is 30 to 45 minutes.
Calculation According to Pregl.-A number of investigators have found
that a deduction of 2% from the total volume will take care of any unabsorbable
gases liberated from tube filling during the combustion, the vapor pressure of
potassium hydroxide and the film of liquid on the side of the nitrometer.
2488 QUANTITATIVE MICROCHEMICAL ANALYSIS
Example.-
Acetanilide sample, mg ........ : ................... 1. ........ . 4.950
Volume of nitrogen collected mI. at 29.5° C. and 753.3 mm.1ressure .. . 0.471
Calibration. correction for burette, mI.... ',' ............ 1 ••• 0.001
2% correctlOn, m!. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... 0.009
Total correction ................................... ". ~......... . 0.010
Net volume of nitrogen to be reduced to standard conditions, mI.... . 0.461
Nitrogen found, % ............................................ . 10.40
Nitrogen calcd., % ............................. : .............. . 10.37
I F
FIG. 424.
I-liter, ro~d-bottomed, Pyrex flasks, respectively. A is charged with 500 grams of acid
potassium carbonate dissolved in 1200 ml. of water, and B with 170 mI. of water fol-
lowed by 17c1mll. of concentl'ated sulfuric acid. The heat of dilution raises the tempera-
ture of the acid almost to the boiling point and thus contributes to the subsequent
elimination of contained air, D and E are sealed, and the desired amount of water is
introduced into the bubble ()ounter G. Then the T-shaped, three-way stopcock J ~th
30 Edgar Poth, Ind. Eng, Chern., Anal. Ed., 3, 202 (1931).
31 Milner and Sherman, lnd. Eng. Chern" Anal. Ed., 8, <331 (1936).
2490 QUANTITATIVE MICROCHEMICAL ANALYSIS
its .perpendicular arm K, filled with and extending into mercUry, is temporarily con-
nected through rubber tubing at I, and the entire system is eva;cuated with an oil pump
simultaneously through Hand L. The pumping is continuedfdr about 15 minutes when
the appropriate amount of mercury is run into F. Both during the pumping and for
spme time afterward, the bicarbonate solution in A will evolve. ¢arbon dioxide. Besides
being a great help in sweeping out the last traces of dissolved ail.' from the solutions, this
property is used in starting off the generator. After closing the, T-shaped stopcock, it is
disconnected from the suction pump at L. The evolution of carbon dioxide in A drives
the gas into B to build up sufficient pressure to force acid into A upon evacuation through
H. If H is closed immediately following the introduction of the acid, the gas generated
will force more carbon dioxide into B. By repeating this procedure, sufficient pressure
can be built up in B to force the mercury seal F. During this process, the T-shaped
stopcock is cautiously opened from time to time to ascertain when the pressure in B is
equal to atmospheric pressure plus the height of the mercury column in F. When this is
attained, the T-shaped stopcock can be removed, and the generator will deliver a stream
of practically pure carbon dioxide whenever H is opened. As the apparatus cools, or
after it has stood for long periods of time, it may be necessary to apply suction at Hand
again force acid into A to reestablish the desired pressure in the system.
Where a supply of dry ice is available, the method of Hershberg and Wellwood 32
provides an excellent source of carbon dioxide and should be used in place of the Kipp
generator method if possible. A two-liter vacuum bottle fitted with a mercury U-shaped
safety valve and charged with dry ice will generate carbon dioxide for·2-3 weeks. When
freshly filled, it should be connected to the apparatus and the carbon dioxide allowed to
. displace the air in the bottle. This usually requires about 12 hours but, in some cases,
it may require a slightly longer period before gas of sufficient purity is obtained.
(8) Ground-glass joints on the combustion tube instead of tubber connections· have
been used. They are satisfactory but not necessary for accurate work.
(9) Some analysts evacuate the combustion tube after it has been connected for a
run in order to eliminate occluded gases. The apparatus requires a three-way stopcock
at Land P, Fig. 423. This procedure cuts down the blank, but where the procedure has
been standardized as to the correction to be applied, the evacuation procedure is super-
fluous when time is a factor.
(10) If the sample is difficult to oxidize, thus giving low results, a small amount, 30
mg., of finely divided potassium perchlorate is sometimes added to facilitate combustion.
The excess oxygen is removed by the metallic copper.
(11) Many analysts use a three-way stopcock between the nitrometer and combus-
tion tube. The presence of a stopcock at this point is not required and often develops
leaks. The regulation of the flow of carbon dioxide with the stopcock between the
combustion tube and the gasometer is much more satisfactory.
(12) The heating of the sample twice, as advocatE)d by Pregl and others, has not
been found necessary.
(13) It is best to use two nitrometers so that while one is cofning to constant tem-
perature the other can be used for the next run. I
(14) The results of Milner: and Sherman and E. P. Clark for th~ vapor pressure of
50% potassium hydroxide show that the subtracting of 0.3% of the volume!?f gas is to.o
small. While these results are interesting and should be taken into consideration when
standardizing a new apparatus, it will be found that for ordinary routine work the
subtracting of a fixed percentage will prove more accurate. The blank is essentially a
constant for a particular apparatus and its method of operation, and thus will vary
according to the individual operator. The following table gives the results of Milner
and Sherman 33 together with those of Clark 34 and those from the international critical
tables. ' . :
32 Ind. and Eng. Chern., Anal. Ed., 9, 303 (1937).
33 Ind. Eng. Chern., Anal. Ed., 8, 331 (1936).
34 Jr. &soc. Off. Agric. Chem., 16, 575 (1933).
QUANTITATIVE MICROCHEMICAL ANALYSIS 2491
(15) Compounds which cannot be analyzed by the Dumas method even with the'
use of potassium chlorate have been noted by A. Friederich,35 and Milner and Sherman.'"
No general rule so far has been formulated for those compounds which cannot be analyzed
by the Dumas method but there are relatively few such compounds reported.
Hayman and Adler 36 report success in the case of certain pyrimidines which gave
low resplts by mixing about 5 parts by weight of copper acetate to one part of sample
and carefully mixing with this about 6-7 grams of fine copper oxide. It was found that
decomposition was only complete when a very hot flame was used.
Spies and Harris 37report a modification of the perchlorate procedure. To overcome
the objectionable low decomposition point of the perchlorate, it is decomposed after the
combustion of the organic compound while the site of the unburned residue is heated to
about 600 0 C., thereby subjecting the nitrogenous residue to an atmosphere of oxygen
at high temperature. The potassium chlorate, 100--125 mg. contained in a porcelain
boat, is placed about 40 mm. from the regular filling which contains the sample. A 15
mm. section of copper is addetl, separated from the regular 40 mm. section of copper by a
3 mm. wad of asbestos. The 15 mm. of copper is removed after each analysis. They
report accurate results for compounds which were derivatives of pyrimidine and purine.
(16) For samples that contain only small amounts of nitrogen, less than 5%, and
·for blank runs, a nitrometer of 0.2 ml. capacity graduated in 0.001 ml. is necessary.
Such nitrometers must bE,\made to order.
(17) In routine work the fine copper oxide takes care of the oxidation of the sample
so that the main filling of coarse copper oxide is but little reduced. Thus only the fine
coppe;r oxide is rejected after each DID. A filling normally lasts for several hundred
determinations.
(18) It sometimes happens that the liquid seeps from the funnel into the upper part
of the nitrometer, thus pushing the column of gas down and preventing a proper reading
of the volume. This column of liquid may be pushed back by gently opening the stop-
cock and raising the levelling bulb above the funnel. With a little practice the liquid
can easily be allowed to escape, and when the gas reaches the top of the nitrometer the
stopcock is quickly closed before any escapes.
Micro-Dumas Estimation of Nitrogen in Liquids Low in Nitrogen-Milk. 38_The
usual nllcro Kjeldahl method for estimating nitrogen in small quantities of a liquid such
as milk may be satisfactorily-replaced by a micro-Dumas process. A small copper bQat
9f approximately 0.25 ml. capacity enclosed in a small thin-walled glass dish covered
with a glass plate is accurately weighed, milk added and reweighed. The boat is placed
in a small vacuum desiccator and kept at 30--350 C. until the water has evaporated; the
boat and its contents are introduced into the combustion tube and covered with fine
copper oxide, and the combustion conducted in the standard manner.
3. "Die Praxis der Quantitativen Organischen Mikroanalyse," Leipzig and Vienna,
Franz Deuticke, 1933.
36 Ind. and Eng. Chem., Anal: Ed., 9, 197 (1937).
37 Ind. and Eng. Chem., Anal. Ed., 9, 304 (1937).
38 G. Kraemer, J. Prakt. Chem., 97, 59 (1918).
2492 QUANTITATIVE MICROCHEMICAL ANALYSIS
s
FIG. 425.
steam to pass into E, then M to the distilling flask N. The lower bulb N is
6 cm. in diameter and the total height from lower to top of upper bulb is 22 cm.
The use of a vacuum jacket around the bulb N and partially coverjng the upper
bulb is recommended for routine distillation as it eliminates the burner at this
point. The condenser T is supplied with running water and burners are placed
under A and N. The steam generated in A passes to N, liberates and carries
ammonia and steam through the tube H, and on passing through T is con-
densed and collected in a 50 ml. graduated flask containing 25 ml. of ammonia-
free water. The condenser tube is placed below the surface of the liquid and
25 ml. of distillate are collected.
The efficiency of the apparatus is shown by an experiment jn which a sample
of ammonium chloride containing 0.070 mg. nit~o5ej}.. was introduced and the
39 Kemmerer and Hallett, Ind.Eng. Chern., 19, 1295 (1927).
QUANTITATIVE MICROCHEMICAL ANALYSIS 2493
West and Brandon •• report that rapid oxidation is obtaiIled by the addition of 0.2%
selenium to the micro Kjeldahl digestion mixture. Unlike mercury-it does not have to
be precipitated before distillation. They give two micro Kjeldahl digestion mixtures.
For biological fluids, add,l mI. of selenium oxychloride to 500 mI. of 1 to '1 sulfuric acid;
usually 1 ml. is added to the sample as the digestion mixture. IFor organic compounds,
mix one volume of concentrated sulfuric acid to one volume bf a saturated solution of
potassium sulfate and add 1 mI. of selenium oxychloride. Use 11 m!. of this mixture for
digesting the sample. . I
(2) The distilling of ammonia into boric acid and titrating directly with mineral
acid has been used successfully in several laboratories and was first proposed by Winkler.""
This eliminates one standard solution. The boric acid retains the ammonia but does
not interfere with its titration with acid. A definite quantity of boric acid should be
used each time because the color of the indicator varies with the cOl1Centration of boric
acid. Stover and Sandin 46 used 4 m!. of a 2% solution which is sufficient to retain 0.1
to 21.1 mg. of nitrogen. Distillation is continued until a. volume of 50 ml. is obtained.
A freshly distilled sample may be titrated directly, but if CO 2 is present the solution must
be heated to drive off the carbon dioxide and cooled before titrating. The following
indicators may be used: 1. Methyl red (concentration 0.1 gm./100 ml. in 95% alcohol)
(use 2 drops). 2. Methyl red and tetrabromphenol blue, made by mixing equal volumes
of solutions of the separate indicators. Concentration of methyl red is given above.
Tetrabromphenol blue has a concentration of 0.1 gm. in 20 m!. of warm alcohol and
diluted to 100 mI. The mixture of methyl red and tetrabromphenol blue changes from
green to a gray color when almost at the end-point, and to pale gray-violet with slightly
more acid. (Use 2--4 drops.) ,
FIG. 426.
spiral. The spirals have a pitch of 5 mm. and through them runs a glass rod
which is sealed to the spiral at each end. The combustion tube is drawn out
to·a capillary end 1.5-2.0 cm. ,in length with an opening of 1 mm.. A large test
tube about 25 mm. in diameter is carefully cleaned with sulfuric chromic acid
mixtu~e and water, and charged with a mixture of 2 ml. ()f pure saturated sodium
carbonate solution and three drops of sodium bisulfite solution. This solution
is aspirated into that' portion of the combustion tube contain-
ing the spiral. Then two freshly ignited platinum contacts are
introduced. If the contacts become" poisoned" they are etched
with hot aqua regia. These contacts are made of platinum foil
0.05 mm. thi~k. Each contact is 75 mm. in length and is formed
in the shape of the letter Z. Contacts having greater surface,
shaped with six fins (Fig. 427) are on the market and are very
satisfactory, or a 15 cm. roll of 80 mesh platinum gauze may be
used. The contacts are placed so that they are 8-9 cm. distant
from the glass spiral. .
.Procedure.-5-8 mg. of the substance to be analyzed are
weighed in a platinum boat and introduced into the combus-
tion tube. (When iodine is determined gravimetrically 5-10·
mg. are used.) Oxygen from a cylinder ,enters the tube at a rate
of 2 bubbles per second, through a capillary tube inserted in a
rubber stopper. The platinum contacts are heated to 700-
750 0 C. by the furnace A, Fig. 428, and then the material in the
boat is slowly burneq by furnace B. (When iodine is determined
it sometimes condenses beyond the heated part and must be
driven by heating into the spiral filled portion.) When the com-
bustion is finished, both furnaces are opened and the tube re-
moved to the supports;1t C and D on the front of the apparatus
where it is allowed to cool in a current of oxygen. The tube and
spiral should be washed three times with water free from halogen, FIG. 427.
the total volume of solution being 30 m!. Two more drops
of bisulfite are added and the solution is filtered to remove any 'suspended
matter. In the determination of chlorine and bromine 2 drops of "Perhydrol"
are added to the solution, and it is then warmed on the water bath 3-5 minutes.
When iodine is present in the solution, heating causes the formation of iodate.
This is avoided by adding 4-5 drops of "Perhydrol" and allowing the solution
to stand at room temperature for ten minutes. After cooling, 1 ml. of con-
centrated nitric acid is added and double the quantity of silver nitrate necessary
to precipitate the halide. This solution is heated on the water bath 10-15
minutes to coagulate the precipitate, is cooled, and then filtered. The filter
tube prepared and dried as previously described under general directions, page
2470, is hung on the balance pan by suspending it on the hook, and weighed.
The tube is placed in a filtering flask shown in Fig. 409 and the precipitate is
transferred by means of a syphon tube not more than 4 mm. in diameter. The
syphon tube is placed just ab'bve the precipitate, suction is applied, and the
sup'ernatant liquid filtered off. The precipitate is washed by adding a few m!.
of nitric acid 1 : 100, then the long arm of the syphon is lowered, and the pre-
, tube by suction. The test tube is rinsed twice,
cipitate transferred to the filter
2496 QUANTITATIVE MICROCHEMICAL ANALYSIS
alternately with dilute nitric acid and alcohol. The syphon is removed, the
precipitate washed once with alcohol, the tube wiped and dried at 1250 C.,
cooled, and weighed. The percentage of halogen prese* in the sample ,is
calculated: t
Weighed Sought Factor
AgCl. ... , .............. Cl 0.2474
AgBr ................... Br 0.4255
AgI .................... I 0.5406
TOP REMOVED
/
o I , r f
10
SCALE· eM.
FIG. 428.
NOTES.-(1) When substances contain halogen free from sulfur and other elements
which combine with metallic silver, the simple method of Dennstedt· 7 is often used.
The combustion tube without a bead mling is charged first with a platinum boat con-
taining the weighed sample, then two platinum contacts·and finally a boat containing a
known weight of molecular f;ilver. The contacts are heated to redness and the molecular
silver to a temperature of 500-600° C. The sample is slo:rly burned in a current of
oxygen and the.products of combustion passed over the platinum contacts, the halogen
being absorbed by the metallic silver. The percentage of halogen in the sample is
calculated from the increase in weight of the boat containing the silver.
(2) Burning -the organic material in a platinum lined micro Parr bomb has been done
successfully.48 A stainless steel bomb is not satisfactory due to corrosion and subsequent
erratic results .. Beamish .9 in a macro procedure reports th~t a nickel'bomb gives good
.7 C. Funk and Kon, J. Am. Chern. Soc., 48, 1667 (1926),:-
48 Elek and Hill, ibid., 55, 2550-4 (1933).
49 Ind. Eng. Chem., Anal. Ed., 6,352 (1934).
QUANTITAT~VE MICR9CHEMICAL ANALYSIS 2497
results even ~ith prolonged heating with bromine compounds, but a stainless steel bomb
gives low results under such conditions. The explosion mixture consists of 1.5 gm. of
pure sodium peroxide, 300 mg. of a mixture of potassium nitrate and sugar in a ratio of
3 : 1. The lid is clamped on, the mixture well shaken, and exploded by heating the
bomb with a small, flame. The fusion is dissolved in 1 : 1 nitric acid and filtered, and
the halogen determined gravimetrically as silveFhalide.
(3) Occasionally, a compound will be found which will give low results by the com-
bustion method due to incomplete burning. In such a case the micro Carius method
should be used which is, of course, the same as the macroprocedure except for the
quantities of material used.
A tube closed at one end 20 cm. long, bore 8 mm., wall 2 mm., is charged with
approximately 0.5 ml. fuming nitric acid, 15-25 mg. of silver nitrate, and the sample,
which is contained in a capillary if a liquid, or in a glass capsule, if a solid. The tube
is sealed, placed in a bomb furnace, and heated at 3000 C. for 5-8 hours, cooled (usually
overnight), and then after carefully breaking the tube, the silver halide is washed on
to a filter, dried, and weighed.
NOTEs.-(l) The volume of excess acid (0.1 mI.) added to the bromine solution
may be exceeded by 0.1 to 0.2 ml. but more than 'this is. inadvIsable. .
(2) The phenol should not be added dropwise but in one portion. The dropwise
addition of phenol often yields a turbid solution consisting of tri-bromophenol and tri-
bromophenol bromide. The latter compound liberates iodinelfrom KI and thus high
results are obtained. / 'I
(3) The distilled water used should be freshly boiled to ren;tove oxygen.
(4) For micro work, N/100 thio~ulfate is often recommended, but at this dilution
rapid deterioration of the solution takes place and the time lost in frequent restandardiza-
tion makes its use in routine work unsatisfactory. N /50 thiosulfate has a constant titer
for at least one week. For routine work it is best to make up every week the required
amount of thiosulfate by diluting the N /10 solution. The addition of 1 % of amyl
alcohol helps to prevent the deterioration of thiosulfate solutions. 51 Another method
used is by adjusting the pH between 9-9.5 by the addition of borax (3.8 gm. per liter of
thiosulfate solution). It is recommended that one of these two methods be used for
preserving the N /10 solution.
of the end-point. Ilowever, it is best to filter or ceptrifuge off the precipitate for ac-
curate work. .
(3) The presencll of sulfur dioxide in the solution also tends to proc1uce a fading
end-point and shoul(i be removed by heating.
1itrating B.a1og~n.l> \5,?':..n.g k\1l>~ipt':..~n. In.\1':..t.at~"1l>.-\','r.t\I)n.u~, !;:ll'VW,:'-;I.", VI
iodide may be titrated dIrectly with silver nitrate using dichlorofluorescein.·
For micro work t}1e halide solution must be neutral. This may be adjusted
by adding acid or alkali until the characteristic green fluorescence of the
indicator appears. Near the end-point a brownish tinge appears; at the end-
point the precipitate assumes a reddish-orange color. As the precipitate is
very finely divided in dilute solution, the solution itself takes on tIlis color.
NOTEs.-(I) The Volhard method for bromide is satisfactory because the precipitate
of silver bromide need not be removed, and the acid concentration is not critical. When
titrating, using dichlorofluorescein, the solution must be carefully neutralized; thus, time
is not saved by using this method for bromine. With chlorine, where centrifuging is
required, the use of the dichlorofluorescein titration may save some time. However,
the end-point is not strikingly sharp with 0.01 N silver nitrate. Bulloc)!: and Kirk 56
found the only way to use the method was by adding an equal volume of acetone to
the solution which sharpened the end-point. They also suggest titrating in a partially
darkened room by the light of a daylight lamp placed at the side of the observer. A
black background is preferred to a white one. Back titrating with sta]"ldard 0.01 N
chloride is possible SO that the end-point may be checked. This procedure is also sug-
gested. for dilute solutions by Kolthoff, Lauer, and Sunde 57 but the reverse titration
must be made immediately, as the indicator reacts slowly if allowed to stand. The
correct amount of illdicator is 2-4 drops of 0.1 % per 50 ml. of solutio]"l· If smaller
volumes are Ui?ed, it is better to employ a concentration of 0.01 %, 2 drops beirig sufficient
for 4.-5 ml. Under the best conditions, the titration error is 0.5 to 1 %.
Calculation.-Vsing 0.020 N silver nitrate.
. _1._5_98__:_,X:...,m;;cI:=-:. .:::si::..lv:;-e:_r_:n=i~tr..::a.:::te:....:X....:...:.lO:..:0:!,
PerceIl.t age b romme=
Wt. of sample
' 0.70914Xml. silver nitrateXI00.
Percell t age chi onne= W f I
t.o samp e
the room, and poured in a thin stream with stirring into the nitric acid solution
of ammonium sulfate. The resultant liquid is diluted tb 1 liter, allowed to
stand for two days, filtered and kept in a well~stoppered bottle of brown glass
in a dark, cool place. l
2. Nitric Acid Containing Sulfuric Acid.-30 ml. of sulfuric acid Csp.gr.
1.84) are poured into 1 liter of nitric acid Csp.gr. 1.19 to 1.21). The latter acid
is obtained by mixing 357 ml. of nitric acid Csp.gr. 1.40) with 500 ml. of water.
3. A 2% Aqueous Solution of Pure Ammonium Nitrate.-The solution is
acidified with a few drops of nitric acid.
Methods of Oxidizing the Organic Material.-I. According to H. Lieb:
The material, 2 to 5 mg., is weighed into a small platinum boat. An excess of
finely powdered pure calcined sodium carbonate and sodium nitrate mixed in
equal portions is added, the whole is mixed with a short looped platinum wire
which is finally laid on the boat, and then carefully covered with the oxidizing
mixture. The boat is placed in a pyrex combustion tube about 15 cm. long,
one end of which is drawn out to a capillary and bent upwards at right angles.
The material is burned in a slow stream of oxygen. As soon as the reaction is
finished the boat is heated for some minutes to the full temperature of a Bunsen
burner and is then allowed to cool in a stream of oxygen. The material in the
boat is then boiled out with dilute nitric acid in a small test tube and poured
through a filter into a precipitation vessel which consists of a large thick walled
test tube which has previously been washed in sulfuric-chromic acid mixture.
If any of the melt has passed into the combustion tube during t~e reaction this
is washed out with dilute nitric acid. To the clear filtrate, 2 ml. of nitric acid
containing sulfuric acid are added, and also water if necessary to dilute the
total volume to 15 ml.
II. According to H. Lieb and O. Wintersteiner: 59 The following .method
requires less attention and is simpler than the other, and can be used also to
determine arsenic as well. It is reliable in most cases. 60 The sample, 3-6 mg.,
is digested in a micro Kjeldahl flask with 0.5 ml. of concl3ntrated sulfuric acid
and 0.5 ml. of perhydrol until fumes of sulfur trioxide /tre evolved, cooled, a
few drops of perhydrol added and the evaporati<;m is repeated until a clear
solution is obtained. To this are added 2 ml. of nitric acid containing sulfuric
acid, and water if necessary to make a total volume of 15 ml.
NOTE.-(l) Elek and Hill 61 employ a Parr microbomb using sodium peroxide to
destroy the organic matter. . -- _...__
Treatment with Molybdate of Solution Obtained Above.-According to
H. Lieb: The liquid, 15 ml., is warmed on the boiling water bath. The hot
solution is well shaken, 15 ml. of the ~ulfate-moly-bdate reagent added. The
vessel then is allowed to stand for three minutes, and again shaken for half a
minute. The precipitate is allowed to settle for 30-,60 minutes, and filtered
with the aid of suction. The filter tube, or filter crucible, is washed WIth water,'
hot dilute nitric 'acid, and again with water, finally displaci.\lg the water by
alcohol and ether or by acetone. The filter tube is then wiped with a dry
59 Mikrochemie, 2, 78 (1924). -
~o A. Elek, J. Am. Chern. Soc., 50, 1213 (1928), reports..a low result when thymus
neuclic acid was analyzed by this method. . -"
61 J. Am. Chern. Soc., 55, 3479 (1933).
QUANTITATIVE MICfWCHEMICAL ANALYSIS 2501
chamois leather, and left for at least half an hour in an evacuated desiccator-
containing no drying agent. It is removed from the desiccator immediately
before use, placed on the balance, and the time carefully noted which has
elapsed between removal from the desiccator and the final weighing, which
may conveniently be 5 minutes. These conditions must be exactly repeated
when the' precipitate is weighed.
The precipitate is drawn into the filter tube by means of the same automatic
arrangement that is used for the silver halide precipitate, page 2470. After the
supernatant liquid has been syphoned from the precipitate it is washed in the
test tube with 2% ammonium nitrate solution, and afterwards transferred to
the filter; in order to remove the last traces of precipitate from the walls of the
precipitating vessel, 95% alcohol and ammonium nitrate solution are used
alt~nately. The tube is finally filled oD,ce with alcohol, and then twice with
ether or acetone. After drying half an hour in vacuo the tube is weighed
under exactly the same conditions as those of the initial weighing. The factors
are:
Phosphorus pentoxide. . . . . . . . . . . . . . .. 0.33260
Phosphoric acid. . . . . . . . . . . . . . . . . . . . .. 0.03326
Phosphorus ......................... 0.014524
. Volumetric Method. 62-Instead of weighing the yellow precipitate it may
be transferred to the original beaker and approximately twice the amount of
N /50 sodium hydroxide necessary for dissolving the precipitate added, and the
solution heated to boiling for 5 minutes to drive off the ammonia formed when
the caustic is added. The volume of solution should be reduced to 15-25 m!.
To the solution after cooling,' 5 drops of Y2% phenolphthalein indicator is
added and 3 to 5 ml. of N /50 hydrochloric acid in excess. The solution is
boiled for ten to fifteen seconds, cooled and titrated back with N /50 sodium
hydroxide. From the amount of alkali neutralized, the weight .of phosphorus
can be calculated. According to the equation for the reaction, 28 equivalents
of alkali are required per atom of phosphorus:
1 m!. N /50 sodium hydroxide= 0.02214 mg. phosphorus.
Accurate results are obtained with quantities of phosphorus 0.01-10 mg.
Procedure when Arsenic is Present (Kuhn, loco cit.).-Phosphorus may be
determined in material also containing arsenic by the addition to 5 ml. of the
phosphorus-arsenic solution of 10 ml. concentrated hydrochloric acid) 0.2 g.
potassium bromide and 0.3 g. hydrazine sulfate and then.. distilling. The
phosphorlls is then determined nephelometrically or otherwise after previous
destructitin of the excess hydrazine with 0.5-1.0 m!. concentrated nitric acid.
Colorimetric Determination of Phosphorus and Arsenic.-When the amount
of ammoniUI!l phosphomolybdate is so small that it cannot be weighed or
titrated with accuracy the following Deniges 63 colorimetric method is used.
If ferric iron is present in concent:ration greater than 4 to 6 parts per million,
it interferes with the determination. If phosphorus and arsenic are both
62 Richard Kuhn, Z. Physiol. Chern., 129, 64 (1923); R. H. A. Plirnrner, Biochern. J.,
27, 181'0--13 (1933).
63 Deniges, Cornpt. rend., 171, 802 (1920); Troug and Meyer, Ind. Eng. Chern.,
Anal. Ed., 1, 136 (1929).
2502 QUANTITATIVE MICROCHEMICAL ANALYSIS
present the two are first determined together. In a second sample the arsenic
is reduced with hydrogen sulfide or a little sodium sulfide in acid solution.
After boiling to drive off the excess of hydrogen sulfide, free sulfur is removed
by adding pure filter paper pulp, shaking vigorously and ~ltering. The color
due to the phosphorus alone is then developed. By difference the amount' ~f
arsenic may be obtained.
Preparation of Reagents: 1. Ammonium Molybdate-Sulfuric Acid Solution.
-Dissolve 25 grams of ammonium molybdate in 200 m!. of water heated to
60° C. and filter. Dilute 280 ml. of arsenic and phosphorus-free concentrated
sulfuric acid to 800 mI. After both solutions have cooled, the ammonium
molybdate solution is added slowly, with shaking, to the sulfuric acid solution.
After the combined solution has cooled to room temperature it is diluted with
water to exactly 1000 ml. This is a 10 N sulfuric acid solution containing 2.5
grams.of ammonium molybdate per 100 mI.
2. Stannous Chloride Solution.-25 grams of. stannous chloride crystals
are dissolved in 1000 mI. of hydrochloric acid (sp.gr. 1.05). Filter if necessary.-
The solution should be protected from the air by floa.ting a layer of white mineral
oil about 5 mm. thick over the surface.
3. Standard Phosphate Solution.-Dissolve 1).2195 gram of potassium di-
hydrogen phosphate and dilute to 1000 ml. A stock solution is made by taking
50 ml. of this and diluting to 500 ml. This stock solution contains 5 p.p.m.
and is used for making the standard solution for comparison. To make this
standard solution 5 m!. of the stock solution are taken, diluted to 95 m!. with
distilled water, 4 ml. of the ammo~ium molybdate-sulfuric acid solution are
added and thoroughly mixed by shaking in an Erlenmeyer flask. Six drops of
stannous chloride are added and shaken. It is diluted to exactly 100 mI.,
shaken, and the solution is then ready for use. It contains 0.25 p.p.m. of
phosphorus. For very dilute solutions 2 ml. of the stock solution are used, but
the same amount of reagents, giving a standard which contains 0.1 p.p.m. of
phosphorus. After standing 10 to 12 minutes, the standll'rd will start to fade,
and a drop more of stannous chloride should then, be arTded to bring the full
color back, which will again be p'ermanent for 10 to 12 minutes.
Procedure.-In the analysis of water, water ,extracts of soils, minerals,
fertilizers, etc., it is permissable to add the reagents directly to these unless they
are colored, turbid, or decidedly acid or alkaline. I Turbidity and color should
be removed by appropriate. means. A decidedly acid or ~alkaline reaction
should'be neutralized before adding the reagents. Organic mater'ials may be
ignited with magnesium nitrate. In this case it is best to keep the 10 N sulfuric
acid and 2.5% ammonium molybdate as separate solutions, in order that the
ignited residue may be dissolved with the sulfuric acid and the ammonium
molybdate added after proper dilution. In all cases the reagents should be
present after final dilution, in the proportion of 4 mI. of'lO N sulfuric acid, 4 mI.'
of 2.5% ammonium molybdate, and 6 drops of stannous chloride per 100 m!.
The reagents should be thoroughly mixed with the test solution before the
stannous chloride is added, after which thorough mixing should again be
accomplished. Comparison with the standard should be clade within 10
minutes after adding the stannous chloride. Blan..Ltests should be run.
QUANTITATIVE MICROCHEMICAL ANALYSIS 2503
NOTEs.-In work where maximum accuracy is required, the presence of much ferric
iron will interfere with the results by decreasing the color intensity and giving trouble-
some greenish tints. By reducing the ferric iron in a Jones reductor, using metallic
cadmium, these troubles are eliminated. The reduction should be made just previous
to the addition of the ammonium molybdate.
Aluminum and manganous salts may be present in considerable quantity without
influencing the color. Nitrate, as potassium nitrate, equivalent to 100 p.p.m. of nitrogen
does not affect the color, while 200 p.p.m. reduces the intensity about 10%. Calcium
oxide or magnesium oxide may be present up to 1000 p.p.m. without influence on the
color.
The substance is oxidized with sulfuric acid and "Perhydrol" and the
arsenic precipitated with magnesia mixture.
Procedure.-5-10 mgm. of organic substance are placed in a micro-Kjeldahl
flask with 4-5 drops of dilute sulfuric acid, heated, then cooled, 5 ml. of "Perhyd-
rol" added, and again heateq until fumes of sulfur trioxide are evolved. The
oxidation is continued in the same way until Mter evaporation a clear solution
is obtained. The solution is transferred to a glass dish of 30 to 40 ml. capacity
and placed on a water bath, evaporated to dryness, then the residue is dissolved
'in 3 to 4 mI. of ammonia and treated while hot with 1 ml. of magnesia mixture
(this is prepared by dissolving 5.5 gm. magnesium chloride crystals and 10.5
gm. ammonium chloride in 100 ml. of water). The magnesium ammonium
arsenate which is precipitated is allowed to stand for 6 or preferably 12 hours
at room temperature until it becomes crystalline.
The precipitate is then filtered through an ignited and weighed micro
Neubauer or porcelain crucible, and alternately washed with a 3% solution
of ammonia and alcohol with the aid of gentle suction. The crucible is re-
moved and ignited. The residue of magnesium pyroarsenate still occludes
magnesium salts, and must therefore again be washed repeatedly with a very
dilute solution of ammonium hydroxide. After igniting strongly, the crucible
is placed on a copper block and may be weighed after 10 minutes. The per-
centage of arsenic in the sample is calculated.
2As
Factor MAO 0.4826.
g2 S2 7
gives reliable results when halogens are absent in. the organic substance and is
particularly serviceable in elementary organic analysis. With halogens absent
the average deviatIon from theory is ±0.2%, when titrating 2 to .4 mg. of
. I
arsemc. I
all the arsine. The scrubber cotton should be saturated with a 15% solution
of cuprous chloride in 1 to 1 hydrochloric acid to remove hydrogen sulfide,
phosphine, and stibine, if present. 6. In, case these gases have not been effi-
ciently removed, the yellow stain imparted to the disk is not changed to brown
when imm@rsed in the" cadmium iodide solution. When using large samples of
some products that contain considerable reducible sulfur or phosphorus com-
pounds, the scrubber cotton may not remove all impurities, and a preliminary
treatment is necessary. The sample is heated a short time with an alkaline
solution, after which it is slightly acidified with hydrochloric acid and the excess
sulfide removed by boiling. The remaining sulfur and reduced phosphorus
compounds are oxidized with bromine water, and then the bromine is removed
by boiling before adding the sample to the reaction flask.
is then swept out for another 25 minutes while the boat is kept at red heat (700-
750 0 C.). The hydrogen sulfide is absorbed in 4 nil. of sbdium hydroxide and,
at the completion of the run, this is washed with boiled water into a 125 ml.
. Erlenmeyer flask containing 6-7 ml. of iodine and 4 mq of hydrochloric acid,
and diluted with 10 mI. of water. The total volume of the solution should ·not
be over 30 to 40 nil. The excess iodine is titrated with 0.02 N thiosulfate and
the percentage of sulfur is calculated. .
1 ml. of 0.02 N.iodine= 0.3206 mg. sulfur.
The iodine solution is standardized against the thiosulfate.
NOTEs.-(l) The sulfur in most organic compounds is quantitatively reduced to
hydrogen sulfide with platinized asbestos, but occasionally some, according to ter Meulen,
require a catalyst of platinum foil or a 6 inch roll of 80 mesh platinum gauze for rapid
conversion. If platinized asbestos is used with such compounds, the sample must be
burned very slowly and with extreme care. H If metallic platinum is used, the presence
of nitrogen in the sample causes the formation of hydrogen cyanide which acts the same as
hydrogen sulfide thus giving high results. In order, therefore, to prevent the absorption
of hydrogen cyanide, a solution of dilute zinc sulfate, approximately 0.05 N, to which is
added sodium acetate, apprmdmately 0.05 N, and a· drop of acetic acid must be used.
As before, iodine together with dilute hydrochloric acid, is added, and when all the zinc
sulfide has disappeared, the excess iodine is titrated with thiosulfate.
(2) ter Meulen also states that platinum black (prepared by reducing at a low tem-
perature ammonium chloroplatinate with hydrogen) must be mixed with petroleum in
order to get complete conversion of sulfur to hydrogen sulfide. This catalyst must be
freshly prepared, I1S heating in air causes a diminishing of its activity.
(3) When a zinc sulfate solution is used and zinc sulfide is precipitated, the spiral
type of absorber shown in Fig. 430 cannot be used because of the" difficulty in washing
out the precipitate. In such a case it is best to omit the spiral and make the bore· of the
stopcock at the bottom sufficiently large, so that the precipitate can be easily rinsed out.
The Pregl type of absorption tube used with the methoxyl ethoxy apparatus, Fig. 438, is
also satisfactory.
(4) After 5-6 runs, oxygen must be passed through the tube in order to burn out
the organic matter and regenerate the catalyst. To prevent the danger of an explosion,
carbon dioxide should be passed through first to sweep out tJle hydrogen.
-
,I
MICROELECTROANALYSIS
The Microelectrolytic Cell of Clarke and Hermance. 75_A cell for micro
electroanalysis has recently been constructed wh,ich permits more effective
agitation of the solution during electrolysis than can be accomplished by the
Pregl method of boiling, or allowing the solution to cool during the period of
electrolysis.. The cell is constructeq so that the solution is agitated by a jet of .
air and the solution maintained at a suitable temperature by a water jacket
heated with a micro burner. While the cell is more difficult to construct than
74 Rec. Trav. Chim., 53, 121 (1934).
75 J. Am. Chem. Soc., 54, 877 (1932).
QUANTITATIVE MICROCHEMICAL ANALYSIS 2511
that of Pregl,• it yields accurate results even when the deposit on the electrode
weighs 1 mg. The average results reported are 0.1 % hightlr than theory. The
individual determinations had an average deviation of ±0.2-0.3% from the
amount recovered.
Apparatus.-The large outer vessel (Fig. 431) sealed to the electrode com-
partment at its base is a water-jacket which can be heated by a micro-flame.
Figure 432-1 shows the glass frame which in the assembled cell fits inside the
electrode vessel and serves to support both the electrodes and pumping device.
FIG. 431.
Its lower end is flared to forin a conical bell having a diameter about 4 mm. less
than that of the container. While the two electrodes are held very close
togother, they cannot make contact or become fouled by touching the glass wall
of the container. The support permits a free circulation of the electrolyte.
This circulation is accomplished by the tube shown in Fig. 432-II, which is
held erect in the axis of the cell. A stream of bubbles of air or inert gas is
introduced by means of this tube into the space under the bell and causes a
€ontinuous overflow of liquid over the top of the tube A that supports the irner .
2512 QUANTITATIVE MICROCHEMICA~ ANALYSIS
electrode. At the same time, the suction caused by the removal of the liquid
from under the bell further assists circulation in this direction by drawing upon
that portion of the electrolyte in the space between the outer electrode and the
wall of the containing vessel. A rapid agitation is thu1s maintained whereby
the whole of the electrolyte is forced through the inter-~lectrode space several
times a minute.
. "The electrodes are two platinum gauze cylinders 15 mm. high, the one
having an inside diameter of 9 mm., the other 18 mm., weighing 0.8 g., and
1.5 g., respectively."
II III
AJR TUBE /
o 1 I
I 2 3 eft
FIG. 432.
FIG. 433.
the upper and three to the lower edge of the gauze to prevent contact with the
sides of the cell during withdrawal.
The anode is a platinum wire 130 mm. in length (Fig. 435), to which two
glass extensions are sealed. so that it may be more easily kept in the center of
the cell.
The cell consists of a test tube, 16 mm. in diameter, and 105 mm. in length.
The terminals of the electrodes are bent
so as to make contact by dipping into
mercury (Fig. 434).
Procedure.-The gauze cathode is suc-
cessively dipped into hot concentrated
nitric acid, water, alcohol, and ether, then
dried over the Bunsen flame or in an elec-
tric oven, cooled, and weighed. '
Determination· of Metals Using the
Pregl Cell: Determination of Copper 'in
Alloys.8L-The material is brought into
solution with concentrated nitric acid.
The amount of copper present for each
determination should be about 10 mg. and
the volume of the solution 5 to 6 ml. If
the lUoy i~ high in copper, an aliquot of the
original soluj;ion is taken for electrolysis.
The cell is assembled, the solution
added, and heated to boiling. The cur-
rent is turned on, and electrolysis takes
place with an e.m.f. of 2.7 to 3.1 volts.
The flame is removed and the solution
allowed to cool during electrolysis. If the
c~ncentration of nitric acid is too high,
FIG. 435.
deposition will not take place, and in this
case a crystal or so of potassium sulfate is added; and if this is not sufficient,
ammonia solution is added drop by drop until deposition begins. A drop
of alcohol added to the solution at the commencement of the determination
will prevent spraying due to evolution of gas at the anode. A few minutes
after electrolysis begins, the walls of the cell should be rinsed free of any traces
of electrolyte with a few drops of 1% nitric acid solution.
After 25 minutes the deposition should be complete, and without interrupt-
ing the current, remove first the anode, then the cathode, and wash the copper
deposit with distilled water. If lead has been present it will be deposited on the
anode and may be removed with nitric and oxalic acid. Benedetti-Pichler
recommends that only water should be used to wash the deposited copper if it
appears spongy, as alcohol tends to cause a flaking off. It is then dried at
90° C., cooled and weighed.
NOTEs.-(I) In order to keep the volume of the solution constant during electrolysis,
water must be added or a water condenser (A in Fig. 434) used.
(2) Where electrolysis is carried out in sulfuric acid solution, the time is shortened
.. by 10 to 15 minutes and in many determinations there is nothing to prevent the use of
82 Benedetti-Pichler, Z. Anal. Chern., 62, 321 (1923).
2518 QUANTITATIVE MICROCHEMICAL ANf\._LYSIS
sulfuric acid. One drop of concentrated acid in 5 ml. of solution is sufficient and an
e.m.f. of 2 volts is maintained. Where organic matter is 'present i~ the electrolyte, the
metal is deposited in the regular manner, but in order to free the deppsit from impurities,
the electrodes are removed, fresh electrolyte placed in the cell, the 'electrodes again as-
sembled, and the current reversed so that the metal goes ba<;k into solution. The current
is again changed and the metal redeposited. \.
The Determination of Mercury.83-The method will determine 0.1 to 0.9
mg. of mercury with an average deviation from theory of ±0.003 mg.
Dissolve the metal in 0.5 to 2 ml. of concentrated nitric acid, transfer to a
measuring flask, and take an aliquot of the diluted solution for electroly'sis.
The cell is a small narrow vessel larger than the Pregl cell and holds 8 to 20 ml.
The cathode is a gold wire 8 cm. in length and 0.3 mm. in diameter and the
anode a platinum wire formed in a spiral. Conditions for electrolysis are:
acidity 0.1 N; current 0.015-0.08 amperes; e.mJ. 3 volts; time 15-20 hours;
temperature 20° C. The electrodes are removed without interrupting the
current, and the cathode carefully washed with water, dried over phosphorus
pentoxide for 3 hours, and then weighed. To remove the mercury, the cathode
is heated carefully over a small flame .
. NOTEs.-Verdino 84 uses the Pregl celI and a gold plated platinum gauze cathode pre-
pared as follows: 50 mg. of gold are dissolved in aqua regia and the solution evaporated
to dryness several times with water to remove the excess acid. The residue is finally
dissolved in 5 ml. of water containing 0.65 g. of pure potassium cyanide. The gold is
deposited at an e.m.£. of 3.5 volts; temperature 55° C.; time 2 hours.
The organic material containing 2 to 5 mg. of mercury is placed in a small hard glass
tube, and 10 drops of fuming nitric acid added. The tube is sealed and placed in a bomb
for 2 hours at 270-280° C. After cooling, the tube is broken and the contents rinsed
into the electrolytic cell with 5 ml. of water. The conditions for electrolysis are: e.m.f.
3.5 volts; temperature 40° C.; time 40 minutes. Acidity in the ratio of 1 ml. concen-
trated nitric acid to 9 ml. of water-or 5 ml. of electrolyte should contain 0.5 m!. con-
centrated nitric acid. Before removing the electrodes the cell is cooled for 5 minutes
with cold water. The cathode is washed with water, alcohol, and ether. It is dried
by waving in the air and weighed. If the deposit is contaminated with organic matter,
it should be redeposited in fresh electrolyte.
The Determination of Goldy'-It is possible to obtlj-in exlellent results with
a solution containing less than 0.5 g. gold per liter, using a volume as small as
0.5 ml. and electrolyzing from a potassium cyanide solution which is the same
procedure as given for gold plating the platinum cathode in the determination
of mercury outlined above.
Fuchs suggests covering the solution with a little toluene to...protect the
anode from possible attack by' chlorine. The temperature of thesolution
should be close to boiling and a current of 3 amperes at 3-3.5 volts is sui~able.
The Determination of Silver. 8lL_The total volume of aqueous solution is
7 m!. to which are added a few drops of concentrated sulfuric acid, and 1 ml. of
20% tartaric acid. If a wet oxidation to recover the silver has been carried out,
the acid concentration should be reduced by evaporation to less than 1 ml.
Conditions for electrolysis are: 1.3-1.4 volts for 15 minutes, then 1.8 volts for
30 minutes. A sample containing 1-4 mg. of silver is best, but samples con-
taining as low as 32 'Yare reported as being accurately analyzed.
83 F. Patat, Mikrochemie, New Series, 5, 16 (1932).
84 Mikrochemie, 6, 5 (1928).
85 Fuchs, Mikrochemie, 1, 86 (1923). ._
86 Friederick and Rap6port, Mikrochemie, 18, 227-34 (1935).
QUANTITATIVE MICROCHEMICAL ANALYSIS 2519
minutes on a gently boiling water bath the mixture is filtered warm, and the
precipitate is washed three times with wash liquid (0.4 m!. of 6.5% nitric acid
and 15 ml. of acetone diluted to 20 ml. with water), aoo dried for an hour at
110-1150 C. The method is suitable for amounts up to' 1 mg. of copper
(C,II.Br,ON),Cu contains 9.53% copper.
DET~RMINATION OF ACETYL
,,!,. I I,' I I
oeM. 5 10
FIG. 436.
sulfonic acid reagent are then added thorough the funnel and the closed stopcock
is sealed with 2 to 3 drops of distilled water added through the funnel. In
the analysis of halogen-containing compounds, a J~w milligrams of silver
sulfate should be added. . ..,...-
"The side arm of the flask is connected to the c-ondenser by means of the
QUANTITAT~VE. MICROCHEMICAL ANALYSIS 2525
freshly rinsed short section of rubber tubing, a glass-to-glass joint being easily
effected. The receiving flask, previously charged with 5 ml. of the 0.01 N
iodine solution with 1 to 1.5 ,grams of potassium iodide (conveniently measured
out as a powder by means of a small marked test tube) dissolved therein, is
attached to the rubber stopper on the lower end of the condenser and adjusted
so that the sintered plate is aboqt 1 cm. above the surface of the liquid. The
side arm of the receiver is capped with a well-washed rubber nipple. The
receiver, up to the side arm, is cooled by immersion in a beaker of finely cracked
ice, renewed from time to time during the hydrolysis after removal of the water
formed by the melting. A stream of cold water, is circulated through the
jacket of the condenser. The flask is tapped sufficiently to mix the sample
intimately with the acid.
"An asbestos board is placed as a shield between the burner and the
condenser. The mixture is heated by bringing the water in the surrounding
beaker to boiling. The level of water in the beaker is maintained during the
period of heating by using an inverted bottle containing water, held well
above the apparatus, and having a 1 cm. tube reaching just to the desired level
in the beaker.
"During the course of the heating, the flask is vigorously tapped at intervals
in order to agitate its contents. Hydrolysis is continued for 1 hour for o-acetyl
and 2.5 to 3 hours for n-acetyl compounds. Where the nature of the linkage is
unknown the longer period is, of course, to be employed. At the end of the
hydrolysis, the hot water is siphoned out of the beaker and replaced by ice cold
water. Mter leaving the apparatus to cool for about 5 minutes, an aspirator is
connected through a manometer and stopcock to the side arm of the receiver
and the system is evacuated to 'a pressure of 50 to 60 mm. The receiver is
lifted, together with its cooling beaker, until the sintered plate is about 2 mm.
from the bottom of the receiver and the water surrounding the hydrolyzing
flask is 'heated. At first a few very small bubbles are seen to pass through
the iodine solution, but the bubbling stops in a few seconds as the pressure is
~qualized throughout the system. When the fluid in the flask has completely
distilled over, aided by occasional tapping, about 1.5 mI. of water are run in
through the funnel without either breaking the vacuum or stopping the heating.
After the contents of the flask have become dry, a second, and finally a third
portion of water is introduced in a similar manner. Heating is then continued
until the contents of the flask are completely dry and for 5 to 10 minutes
thereafter. The receiver is then lowered, using a gentle rotatory motion, until
the sintered plate is 2 to 3 cm. above the fluid. The stopcock between the
receiver and the gage is closed; the connection between the stopcock and the
gage is broken, and then by slowly opening the stopcock the system is brought
back to atmospheric pressure, and the burner is extinguished.
"The condenser is disconnected from the flask and washed through three
times with small volumes of distilled water, flow across the sintered plate being
aided by the application of suction through the side arm of the receiver. The
outside of the lower part of the condenser-is washed into the receiver with
distilled water; the total volume should be about half its capacity. The solution
in the receiver whose side arm is capped with the rubber nipple, is titrated with
the standard thiosulfate, using'2 drops of the starch indicator added toward
the end of the titration. To the now colorless solution about 2 ml. of the
2526 QUANTITATIVE MICROCHEMICAL ANALYSIS
potassium iodate solution are added, and the receiver is immediately stoppered
with a soda-lime tube, placed in a beaker of water at about 35" C., and allowed
to .stand for 20 ~nutes. At the ~nd of ~hat time the soluti<)n is titrated with the
thIOsulfate solutIOn, the end-pomt bemg taken where the blue color which
develops during the titration just disappears. I
"The value from the initial titration is to be subtra~ted from the value
obtained by titration of 5 m!. of the iodine solution in which 1 to 1.5 grams of
the potassium iodide are dissolved and 2 to 3 drops of dilute acetic acid added.
(It is especially important when measuring the iodine solution ,from a micro-
buret to wait 1 to 2 minutes before reading the buret; the drainage error may
otherwise be considerable.) This latter titration should not be made until the
mixture has stood for about 5 minutes at room temperature. The difference
so calculated is a correction for the sulfur dioxide and varies from 0.00 to 0.05
m!. Where it exceeds the upper limit the analysis should be rejected. The
difference, doubled to account for the dibasicity of the sulfuric acid formed, is
deducted from the final titration as a correction for the acid introduced as
sulfurdioxide."
Calculation: The acetyl content is given by the formula
Hooks to be on each
side of tube and flask
for tension springs
14mm.
FIG. 437.
,
THE DETERMIN:ATION OF MOLECULAR ~IGHT
--._
For Solids-By Rise of the Boiling Point.-The following is a micro modi-
fication of the method developed by Menzies and Wright.103 There are two
forms of the apparatus, the chief difference being that in one the solvent is
boned by a small coil of platinum wire placed directly in the solvent 104 while in
the other a small gas or alcohol flame is used.I°. Fi~. 440 shows the bottorp
of the boiling tube with platinum coil, and Fig. 441 shows the Smith and Milner
apparatus. B9th are the same' except for the lower section. By using a
102. A. Friederich, Mikrochemie, 7, 185 (1929). •
l02b Willstatter and Utzinger, Ann. Chem., 382, 148 (1911). Kjister and Maag,
Z. Physiol. Chem., 127, 190 (1923). Phillips and Goss, J. A~.o. A. C., 20, 292 (1937).
103 J. Am. Chem. Soc., 43, 2314 (1921). .J-
10; A. F. Colson, Analyst, 57, 757 (1932).
105 Smith and Milner, Mikrochemie, 9, 117 (1931).
QUANTITATIVE MICROCHEMICAL ANALYSIS 2531
is then returned to its original position and the .liquid boiled briskly enough
to pump it in an almost continuous stream over th~ lower bulb of the
thermometer. Then at intervals of 1 or 2 minutes, readings ,are made on
the millimeter scales etched on the thermometer: Whe? the readings have
become constant, they are recorded. The sample, 10-15 mg., made into the
form of a pellet in a pellet press, is accurately weighed and is then introduced
into the boiling tube. Again the thermometer liquid is allowed to become
constant and the reading recorded. This may require 5-10 minutes. Addi-
tional pellets may then be introduced and the corresponding records made.
Calculation.-For the conversion table of the differential thermometer the
original paper should be consulted. lo6 The ebullioscopic constant, K 1 , is for
one mole of solute in 100 gm. of solvent. This requires also knowing the
density of the solvent at its boiling point. The ebullioscopic constant, K 2 ,
is for one mole of solute in 100 ml. ot'"solvent at its boiling point.
Kl = Rise B.P. Xmolecular wt. of solute Xwt. of solvent,
100Xwt. of solute
Rise B.P. X molecular wt. of solute Xml. of solvent
100 Xwt. of solute
Constants Useful in EbuJ.lioscopy
J. Am. Chern. Soc., 43,2320 (1921)
Acetone .. ' .... .' ....... 57.0 26.4 0.7522 0.0011 17.2 22.9
Benzene ............... 80.1 23.4 0.8149 0.0010 26.1 32.0
Carbon disulfide ........ 46.3 24.7 1.2223 0.0016 23.7 19.4
Carbon tetrachloride .... 76.8 23.3 1.4801 0.0021 48.0 32.4
Chloroform ............ 61.2 25.2 1.4101 0.0018 39.0 27.7
Ethyl acetate .......... 77.2 25.1 0.8302 0.0010 27.9 33.6
Ethyl alcohol. ......... 78.4 30.3 0.7389 o.ooil 11.7 15.8
E-thyl ether ............ 34.6 26.9 0.6968 0.0011 21.6 31.1
Methyl acetate ....... , . 57.1 26.8 0.8802 0.0014 20.6 23.4
Methyl alcohol ......... 64.7 29.6 0.7468 0.0010 8.8 11.8
apparatus. The melting point is then determined with the aid of a lens. The
capillary is cooled and a second reading taken to check the first.
Calculation:
1000 S.K.
At
Lfl
S= weight of substance,
l( = moleculitr freezing point depression of the camphor used,
L= weight Of camphor in mg.,
Ll = observed depressing of the freezing poin.t.
NOTES.-(l) It o~en happens that a larger sample may be used a~s a iarger
capillary Can be employed. It will be found much easier to charge a large rather than
small one !Such as is described above. It is also much easi~r to add camphor to a larger
capillary. '
(2) For those who find that they have a great number of determinations to.carry out
by tIns method, the apparatus described by Carl Tiedcke lO~ may be of interest. It gives
greater cOhvenience in holding the thermometer and capillary, together with a novel
container for the sulfuric acid. It consists of a 250 ml: fia~k with ground-glass top into
which is inserted a thermometer. The flask is provided with a stirrer and pla.tinum hook
to hold th~ capillary.
(3) JO~eph Pirsch uo suggests Bornylamine (m.p. 1640 C.; mol. m.p. depression
E=40.6) :).nd Camphene dibromide (E=115), as also suitable for the determination of
molecular weight.
manometer is etched just below the seal F. Before use the inner jacker A is
filled with water to the reference mark to determine its volume. N is an as-
l1tstos board with a hole~to admit the jacket B. A is supported in the outer
jacket B by a split stopper. H is a 300 mm. reference scale. Determinations
may be carried out under reduced pressure by attaching th'e rubber connection
K to A and applying suction.
Determination.-The liquid, 2-10 mg., is weighed in the' usu'al manner into
a capillary of such diameter that it will readily pass through the bore of the
stopcock D. The form is shown in Fig. 444-A. The upper end of the constric-
FIG. 443.
tion is closed and a bead formed, Fig. 444-B, which is larger than the bore of the
stopcock. A liquid with a boiling point 15-20° C. above that of the sample i8
poured into B to a depth 10 mm. below the end of the inner jacket A. The
manometer is filled with mercury and A tilted to fill trap, G. It is then placed
inside of B, stopcock D is closed, and the burner under B is lighted. The
liquid is boiled just vigorously enough so that B is filled with vapor at all times.
The level of the mercury in the right hand manometer is brought to the refer-
ence mark and D is closed. When no change in mercury level is noted when D
is closed; the determination is begun.
Stopcock D is opened and the capillary ip.serted, the bead of which rests on
the bore of the stopcock. The mercury is brought to the reference mark and
the level of the mercury in I is read on the scale H to 0.2 mm. D is closed, thus
breaking the capillary and it falls to the bottom of A. The levelling bulb is
gradually raised and when no further inc~ease in pressure in A is noted, the
2536 QUANTITATIVE MICROCHEMICAL ANALYSIS
mercury is adjusted to the reference mark and the level read. The difference
in the initial and final readings gives the change in pressure in A. I is lowered,
the plug from the broken capillary r~mqve.d from D and a
long capillary inserted into A, and the :vapor removed by
suction. By making the long capillary slightly smaller than
the sample capillary the latter may be retained on the end of
the long capillary, when suction is applied, and thus is re-
moved from the apparatus.
Those compounds which tend to decompose when heated
to their boiling points at atmospheric can be vaporized at
reduced pressure as follows: The sample tube is inserted in
D and L is connected to D by the rubber pressure tubing
K. L is closed, and M is connected to a vacuum line. I is
lowered so that the mercury at F falls to a point just above
G. L is slowly opened and if necessary I is again lowered.
L is then closed and the mercury brought to the reference
mark. If the mercury level remains constant, the level
of the mercury is recorded in mm., the sample tube broken
by turning D and the 'determination completed as previ-
ously described. :
Calculation:
A B
MW =22410X760XTXW
.. lOOOX273XVXLlp'
where T = absolute temperature of the vapor bath,
W = weight of sample taken in·mg.,
FIG. 444. V = volume of inner jacket in mI.,
Llp = change of pressure within A in mm.
Collecting Constants.
TXW
M.W. = KXt;:p' f
I
The constant K for any apparatus can be calculated when the volume of the
inner jacket is known.
N OTES.-(l) The following minor changes will be found to make the apparatus more
convenient: '
(a) The tube 'I should be ~ade of tubing 6-7 mm. inside diaxneter with a funnel
made of tubing 15 mm. X 100 mm. sealed on. It is easier to read the mer.cury' on the
scale when it is contained in a smaller tube. .
(b) The tube beginning at F should not have as steep an angle, as there is consider-
able strain on this tube 'with the weight of mercury.
(c) A vacuum jacket beginning at the point C and extending up to the bottom of the
water condenser will prevent excessive condensation and maintain a more uniform
temperature in A.
(d) If the apparatus is in constant use, a thermometer welt is useful so that the tube A
need not be removed to get the temperature of the boiling liquid .
(2) If air causes decomposition of the sample, jacket A may be filled with an inert
gas. .
(3) Bath liquids which can be used are: water, B.P. 1000 C., p-cymerie, B.P. 1800 C.,
alpha-naphthyl methyl ether, B.P. 2690 C., benzyl benzoate,_ B_:P. 3199 C.
,
QUANTITATIVE MICROCHEMICAL ANALYSIS 2537
GAS ANALYSIS n
= ~-.. -M
The apparatus described is designed for
carrying out a gas analysis using 25 to 100 cu.
mm. The degree of precision is of the same order
of magnitude as that ordinarily attained in macro
gas analysis. ll4. = --, L
Apparatus.-Fig. 445 is a diagram (Blacet
et al.) drawn to scale of the apparatus. A is 0.. rlt.... -- J
a water jacketed microburette. This is a capil- FJIf:-: ----H
lary tube of approximately 0.5' mm. in diameter ••---- . -" G
and graduated to read in cu. mm. with the zero
mark at the top of the scale. The total capacity
is about 112 cu. mm. The upper end is bent as
shown and ends in a capillary of small external
diameter to facilitate the intake and discharge of
--·....-c
gases. The tip of this capillary is ground and fire
polished to a radius of curvature somewhat less
than that of the top of the, gas holders Hand J.
The lower end of the burette is sealed to a larger
tube containing a trap to insure the exclusion
from the capillary of impurities and chance solid -- --A
material which may be introduced in the sealing
process. The device represented by P, R, S, and
B is used to move the mercury and gas sample up
and down the microburette. 'p is an iron
cylinder into which the bell-shaped lower end of
the burette is cemented by the sealing wax N.
The wax is also used here to close the water --...... D
jacket of the burette. R represents an elastic
rubber cap. The lower flange of this cap is tyt;,
firmly secured against the lower edge of P in a u -- ...... E
-~ ~
manner which may be deduced from a study of
the drawing. S is a metal plunger which is
attached by a swivel to the screw B. ThtJ space
above R is filled with mercury. By manipulating
B the plunger is forced either up or down causing ~ -. N
a flow of mercury or of gas through the capillary
burette A. Since P is the only one of the metal ,... -_._p
parts which comes in contact with the mercury,
it is the only one which must be made of some' , -'. R
non-analgamating material. By means of tele-
scoped brass tubes C and D and the machined
screw E, everytliing on the table F can be moved
up or down with perfect control. G is a mercury
rut:':
reservoir of 7 cm. diameter. Hand J are the FIG. 445.
containers for the gas sa~ples and are also used
114 Blacet et aI., Ind. Eng. Chern., Anal. Ed., 3, 266 (1931); 5, 272 (1933); 6, 334
(1934); 9, 44 (1937). Swearingen, Gerbes and Ellis, Ind. Eng. Chern., Anal. Ed., 5,
369 (1933); Seevers and Stormont, Ind. Eng. Chern., Anal. Ed., 9,39 (1937).
2538 QUANTITATIV~ MICROCHEMICAL ANALYSIS
FIG. 448.
and sealed into place after warming the metal parts, effects a mercury~tight
union of .both the end plate and the diaphragm with the reservoir. In as':
sembling, tlie capillary tube is inserted through the unit· and spaces VV are
filled with sealing wax over a rubber gasket. Mercury is introduced into the
apparatus through the opening at the top of the manipulator reservoir and
sealed with a'tight-fitting needle valve. As indicated by the arrows in the
diagram, water from a constant-temperature bath circulates by gravity feed
between the capillary and th'e reading buret and returns to the outlet through
the space between the reading buret and the outer jacket. It empties into a
reservoir from which it is elevated to the bath by a water pump. Forced
pumping through the apparatus results in sO'muchbubbling as to make reading
difficult.
The relation of the microburet to the absorption reservoir is shown in
Fig. 449.
By rotation, of the buret unit around its horizontal axis, the buret tip may be
introduced into any capsule of either the sample carrier N or the capsule holder
Q, since the latter two have not only a vertical axis of rotation, but are free to
describe an arc around their respective supporting rods.
2540 QUANTITATIVE MICROCHEMICAL A,~ALYSIS
FIG. 449.
removal. Steel forceps made from a sponge clamp are used to handle the
capsules, which are filled by slowly inverting under mercury before insertion
into the holder. Care must be exercised to prevent .small air bubbles from
adhering to the walls of the capsule during this procedure.
The interchangeable bead holders are of glass tubing fastened into the metal
shank with wax, or preferably an adjustable rod of stainless steel fastened to
the bead-holder arm by means of a set screw. The platinum till is of sufficient
length that the glass portion of the holder does not ,enler the'sample. The
bead holders are of uniform size and shape and hel<fin fixed position by two
pins upon the cross arm T (Fig. 449) which is raised or lowered by the rack and
QlJANTITATIVE MICROCHEMICAL ANALYSIS 2541
pIllIOn. It is thu~ possible to insert four beads into the absorption capsules
without any bead coming in contact with the inner surface of tlle capsules.
As reported repeatedly, such an occurrence ruins the analysis. The level of the
mercury in the reservoir· may be changed by raising or lowering the Bakelite
block K. This provision is necessary in order to maintain a constant level of
mercury in the reservoir, 'and is shown by the ivory tipped indicator at M.
Preparation of Samples.-All samples must be dried before analysis and the
burette must also be kept free from moisture. It has been found that the
normal vapor pressure of water does not exist in such small gas volumes. The
dry wall of the burette absorbs moisture and erratic results are thus obtained
in the analysis of samples containing moisture.
The Preparation of Beads.-(l) Beads of solid reagent are prepared unless
otherwise stated, by melting the material and taking it up in a small loop of
platinum wire. The platinum wire is sealed into a glass holder as shown in
Fig. 445.
(2) Beads for liquid reagents are made by taking soft glass 100-150 mesh
and heating it carefully until it just begins to fuse. It is shaped to a sphere of
approximately 2 mm. diameter. This is then attached to a straight platinum
wire by heating the wire and letting the end of it become embedded in the
bead. The pores of the bead are filled when it is dipped into a liquid reagent. I
Swearingen, Gerbes and Ellis made a bead for liquid reagents by taking a
mixture of 70% grog and 30% kaolin. The grog, ground porous earthenware,
is sifted to a uniform size, passing 120 and retained by 140 mesh screen. The
kaolin should pass-200 mesh. The consistency of the mixture Ilhould be such
as to allow it to be dipped up witj:J. the platinum wire loop in order to get a
smooth surface. The bead is baked in a flame for a few minutes and is then
ready for use.
Operation of Apparatus.-The operation of the apparatus, Fig. 445, as
described by the authors is given below taking the determination of oxygen in
dry air as an example.
Snecific Determinations: Oxygen.-First the containers Hand J are filled
with mercury. This is done by placing a glass tube, which has been drawn out
to a capillary and properly bent, so that its end will go down into the reservoir
G and up to the top on the inside of the containers. If the containers are clean
,and have a top of smooth curvature all the air can be drawn out with ease. The
hi~ surface tension of mercury facilitates this. Several hundred cubic milli-
meters of the air are introduced into H. This constitutes the sample for analy-
sis upon which several determinations presumably are to be made. If a
• Toepler' pump is used for the purpose of introducing samples, its outlet is bent
in the same manner as the tip of the buret and placed permanently in position
in the reservoir and at the same height as the buret tip. The rotating table
M is mounted on ll.ij, adjustable arm so that the containers can not only rotate
completely around the central shaft but may be moved to any desired position
within the re~ervoir. Accordingly, the outlet for the Toepler pump may be at
the back side..of the reservoir and a container can be moved over its tip and
filled.
The next step is to obtain a known volume for analysis. By means of the
screw B all of the air in the buret is replaced by mercury. The tip must be
.
under the surface of the mercury in G while this is done, otherwise when it is
2542 QUANTITA~rrVE MICROCHEMICA~ ANALYSIS
submerged it will invariably carry a small aIr 'bubble w.ith it. H is 'now placed
directly over the tip, and by lowering the table F, the tip is brought into the
gas sample. By means of B a portion of the sample IS drawn into t~e buret.
The volume ordinarily taken varies from 25 to 100 cu. :&tm. depending upon the
total amount available and the subsequent treatment to which it is to be sub
jected. The table is then raised so that. the level of the1mercury in the reservoi~
is on the level with the uppermost calibration mark in the water-jacketed part
of the buret. By further unscrewing B, the gas, followed by a thread of mer-
cury, is brought into the calibrated portion for measurement of its volume.
When its upper level is at the same height as the surface of the mercury in the
reservoir, it is then theoretically under atmospheric pressure. However, the
apparent pressure in the buret may be con~iderably in error, owing to the fact
that mercury has a tendency to stick to clean dry glass. If no precautions are
taken two subsequent volumes on the same gas sample may be obtained which
vary as much as 1% from one another. To reduce this source of errorJ.wo ex-
pedients are used. First; the buret is gent~y tapped with the finger as the gas
is being brought to the desired position. Second, five independent readings
are always recorded and the average taken in calculating the volume. These
. precautions reduce the probable error in all/cases below 0.2%.
After the volume readings for the run have been recorded along with t.emper-
ature and barometric readings, container, J is brought over 'the buret tip and
the gas in the buret expelled into it. Tlie gas is now ready to have the oxytoo
absorber introduced. The removable holder L has a platinum loop sealed in
its tip. In th\s loop is fused a small bead of yellow phosphorus. The fusion
may ordinarily be done by placing a fresh piece of phosphorus in the loop by
means of forceps and then holding it over a warm resistance coil or even in. the
sunlight. By quickly removing when fusion occurs and placing it under the
mercury, spontaneous combustion may be avoided. In any case it should not
be allowed to remain in the air any length of time, for the moisture of the air
along with oxygen causes a film of phosphoric acid to coat the bead, thus making
it a much slower absorber. As shown in Fig./445,A;he holder is placed in the
guide K. The container J is brought into position and the phosphorus intro-
duced into it. In this operation care must be taken not to allow the glass of
the holder L to touch the walls of the container. It has been found that a cold
tube introduced into mercury. in this way is coated' with a film of air, part of
which will be imparted to the walls of J if contact is made-between the two ~lass
surfaces. On this account the guide is necessary for the introduction of the
absorber up the center of the container. The height of K is-,also arranged so
that only the phosphorus bead and a little platinum wire enters the gas bubble.
Fifteen minutes has been found sufficient time to allow for complete absorp-
tion in practically all cases. However, absorption to constant volume is
practiced. After absorption has occurred and th'e absorbent removed, the -tip'
of the buret is brought to the top of the container (more accurately, the top of
the container is brought to the buret tip) and the residual gas taken into the
buret. With the proper manipulation of B, E and M, this operation presents
no practical difficulties. If the entire gas bubble is not drawn into the buret
intact the first time, it need only be driven b~ck:-into the container and the
process repeated. The experimenter soon acquires skill in this regard. The
volume of the residual gas is now measured and the temperature and the
QUANTITATIVE MICROCHEMICAL' ANALYSIS 2543
pared over a glowing electric coil. A flame should not be used because of the
moisture' generated by it. .
Carbon Monoxide.-Specially prepared dry silver oxide is used. It is pre-
pared by means of a strong base and thoroughly washed by decantation.
W~ile still somewhat moist the solid is compressed into pellets by applying 'a
pressure of 6006 lbs. per square inch. The pellets are allowed to dry at room
temperature: They should not be unduly exposed to air as they take up
carbon dioxide. The tip of a straight platinum wire is dipped in a concentrated
solution of sodium silicate and then touched to a piece of regular shaped silver
oxide 1.5 mm. in diameter. The cement is dry in 10 minutes. Absorption
of the carbon fIlonoxide is complete in 10 minutes. No volatile products are
formed, so no follow-up absorbent is required.
'Ethyl en e.-The sintered glass bead is dipped slowly into fuming sulfuric
acid. The space in the bead becomes filled with liquid and the air is forced out.
The bead is wiped with a piece of filter paper to remove excess acid from the
exterior surface, and then introduced in the usual way into the gas sample.
After 2 minutes the acid is removed and a slightly moist potassium hydroxide
bead is added to take up the sulfur trioxide vapor. This absorption is also
very rapid. After another 2 minutes the bead is removed and the decrease
in volume of the sample due to the presence of unsaturated hydrocarbons is
measured.
Acetylene.-A stiff paste ip made by moistening cuprous chloride with a
dilute potassium hydroxide solution. This is molded in the platinum loop and
heated gently until dry, care being taken not to heat the solid until it turns
dark in color. This bead 'gives complete absorption of the acetylene contained
in an average sample in 5 minutes. The copper acetylide formed becomes an
integral part of the .solid bead and does not contaminate the mercury as it is
being removed from the reacting system. The potassium hydroxide takes up
all of the hydrogen chloride or water vapot formed, so that the decrease in
volume gives directly the amount of acetylene which is-present.
Hydrogen Chloride, Sulfur Dioxide, Sulfur Trioiide, Hydrogen Cyanide,
Carbon Dioxide.-(l) May b~ absorbed by using a bead of fused potassium
hydroxide -'which, before use, is allowed to absorb moisture from the air.
(2) Or, by using a sintered glass bead and a concentrated solution of
potassium hydroxide. The water vapor pressure over this absorbent is negli-
gible. 2 minutes are required for either method. '
115 Ind. Eng. Chem., Anal. Ed., ~, 369-70 (1933).
2544 QUANTITATIVE MICROCHEMICAU ANALYSIS
2. The Heyrovsky Polarograph. The principle is not new but the design of the
apparatus and method of analysis was developed by J. Heyrovsky! of Charles University,
Prague.
By using a dropping mercury cathode and by constantly increasing the voltage to a
sample in solution, a point is reached, characteristic for each reducible compound, where
the intensity of the current suddenly increases. The intensity varies with the concen-
tration of the substance. Only a few drops of solution are, required and the sample is
not destroyed by the analysis. While the method is not recommended as a routine
procedure of general application it has been found useful, by workers in this cOlrntry and
abroad, in the study of special problems dealing with the quantitative analysis of traces
of metals, and some inorganic and organic radicals. The usual limit of sensitivity is
0.0001 mg. A study of the method has been made at the University of Wisconsin, and
numerous references to the method appear in the literature. A booklet, "Chemische
Analysen mit demo Polaragraphen" by Dr. Hans Hohn, published by Julius Springer,
Berlin (1937) deals with the polarograph and its practical applications in a very thorough
manner. •
3. The Use of Fluorescent Analysis in Michrochemistry. Max Haitinger, Mikro-
chemie, 16, 321 (1934-35). '
Physical Methods'
1. Microcalorimetry. Brian Whipp, Phil. Mag., 18, 745-59 (1934). Determina-
-tion of Heat of Combustion with the Microbomb. J. J. Vrigling, Chern. Weekblad., 32,
20-2 (1935). Researches on Calorimetry. Generalization of the Method of Electrical.
Compensation. Microcalorimetry. A. Tian, J. chim. phys., 30, 665-708 (1933). A
Semimicrocalorimeter for Measuring Heat Capacities at Low Temperatures. Daniel
R. Stull, J. Am. Chern. Soc., 59, 2726-33 (1937).
2. Density. Micropycnometer Method for Density Determinations. G. R. Clemo'
and A. McQuillen, J. Chern. Soc., 1220 (1935). Micro Density Deter,mination of Solids
and Liquids. Eugene W. Blank and Mary L. Willard, J. Chern. Ed., 10, 109-12 (1933).
3. Distillation,. Microanalytical notes: 1. Some Improvements in Methods of
Dealing with Small Quantities of Liquids and Precipitates. B. L. Clarke and H. W.
Hermance, Mikrochemie, 18, 289-298 (1935). This deals with improved methods of
filtration, evaporation and distillation. Fractional Distillation of Extremely Small
Volumes of Liquids. Benedetti-Pichler and Rachele, Mikrochemie, 19, 1-5 (1935).
Separates 10 cu. mm. into fractions of 2 cu. mm. Microdistillation of Liquids. A
Microdistillation Apparatus. Lyman C. Craig, Ind. Eng. Chern., Anal. Ed., 8, 219-20
(1936). Isolation from Human Tissues of Easily Volatile Organic Liquids and their
Identification. A. O. Gettler and H. Siegel, Arch. Path., 19, 208-1.2--(.lQ_35). Micro-
fractionating Column for Liquids Having Low Heat of Vaporization. PaillEl-. Weston,
Ind. Eng. Chern., Anal. Ed., 5, 179 (1933).
4. Extraction. Micro Fat Extraction. Microextractor. L. Titus and V. W.
Meloche, Ind. Eng. Chern., Anal. Ed., 5, 286-88 (1933). '
5. Microrefractometer. A microrefractometer (for liquids) and its use in chemical
microscopy. E. E. Jelley, J. Roy. Microscop. Soc., 54, 23H5 (1934).
6. Viscosity. Microviscometer. F. M. Lidstone, J. Soc.' Chern. Ind., 54, 189-90 T
(1935). Viscosity Determinations with Small Amounts of Material. Otto Merz,
Farben. Ztg., 37, 1192-3 (1932). A Microsaybolt-type Viscometer. Stuart M. Rogers
and Linden R. Adkins, Ind. Eng. Chern., 20, 742 (1928). /
Detennination of Organic Radicals' - -
~.-
1. Carbonyl Group. Microdetermination of Carbonyl Groups. 'F. Frhr. v. Falken-
hausen, Z. Anal. Chern., 99, 241-57 (1935).
QUANTITATIVE MICROCHEMICAL ANALYSIS 2547
THE MICROSCOPE
FIG. 451.
PHOTOG~HIC MATE~S
EQUILmRIUM DIAGRAMS
FIG. 453.
II
Composition
FIG. 455.
I
The portions at each end of the diagram appear similar to that 'shown in
Fig. 453 of alloys whose component metals are completely soluble in the solid
state whereas the middle portions show the diagram typical of those alloys
whose component m~tals form a eutectic. A photomicrogra,.vh taken at C
would appear as in Fig. 456,11 Here the metal in exceEjs_Qf the ,e utectic ratio has
_r-
11 After. Sauveur.
METALLOGRAPHY 2555
solidified first, forming the major portion of the alloy with the matrix composed
of the eutectic 12 of the two metals which solidifies along the line AB. At E
I
B
Composition
FIG. 459.
12 EUTECTIC (from the Greek EV'f'1K'fO$, easily melting), the alloy of a certain series
possessing the lowest temperature of solidification. Their structures are made up of
alternate plates or "lamellae" of the different constituents.
2556 METALLOGRAPHY
1100
~ . ,/
~
1000 .
.... ~~~d s
. . . .Its
............. i'..
900 -
1\ ~~ ~Zn3
-~
.. 800
."
l!!
.~
c
LJ 700
\\ 1\ P I \\'Y+~
, ,
~
1!
g
\ a +11\ fJ+'Y
I
.{,a + Li
"'"
.'\
600
- ., - '
500
\ 'Y ,,
\ ~ \
\ , r\
,,--- ..... i~.
VHf .. \€+ ,, \
400
a+'Y 'Y+£ £ £+'1] 'I]
300
-
"Io Z n. 10 20 30 40 50 60 70 80 90 100
% Cu. 90 80 70 60 50 40 30 20 10
FIG. 460.
COPPER-ZINC ALLOYS
FIG. 461. PhotOmicrograph of Brass Cold Rolled and Annealed.. Magnified 100
diame~ers. Etched with ammonia and hydrogen peroxide.
·Here there are three solid solutions, alpha, beta and gamma. The color
of the alpha brasses varies from a pale yellow to a yellowish red, the color
becoming redder as the copper content increases.' These brasses have moderate
tensile strength, high ductility and are suitable for cold rolling. Figure 461 is
a photomicrograph of a section of alpha brass which has been cold rolled and
annealed. Alloys containing the beta solution are somewhat harder and when
cold are much less malleable \nd ductile. Beta brass has a greenish-red color
and is darkened more quickly by the etching reagents than the alpha variety.
Brasses containing the gamma solution being very brittel are without mechan-
ical value and are only employed for castings of a purely ornamental character.
By regulating the amounts of copper and zinc and the temperature to
which the alloy is heated and the rate of cooling from that temperature, the
most advantageous properties of each of the constituents can be imparted
to the ;metal. Mechanical working also improves the 'physical properties.
Other metals are often added to copper-zinc alloys to make the alloy more
.suitable for the purpose for which it is to be used, e.g., iron, manganese, and
aluminum are added in small quantities and increase the tenacity of the
alloys considerably.
14 Mter Hofman.
2558 METALLOGRAPHY
IRON-CARBON ALLOYS
500
9(,Co 1.0 1.7 2.0 3.0 4.0 .3 5.0 6.0 6.67
%Fe3C 0 1.5' 25.5 30 45 bo 64.5 7~ 90 100
I
I
FIG. 462. I
diately after solidification has taken place, the steel/? consist of a solid solution
of carbon in gamma ironP Gamma iron, then, ,vill dissolve carbon until a
total content of approximately 1.7% is reached at which point the.sQlid t3olution
becomes saturated and further increase in the amount of carbon res:D:lts in the
precipitation of the carbon in the form of the aarbide of iron (Fe 3 G) (called
cementite) which forms a eutectic with the saturated austenite. At the pqint
E the alloys solidify as 100% eutectic. With more than 4.3% carbon the
constituent in excess of that necessary to form the eutectfc is the carbide of iron ...
I
15 After Sauveur. I
16 The division between the steel series and the cast iron series is a purely arbitrary
one, but the fact that there are very few commercial products containfng between
1.7% 'and 2.5% carbon makes it a particularly convenient one. .
. 17 Many authorities define austenite as a solid solution of FeaC jn gamma iron.
Recent research, however, on crystal structure founded on X-ray examinations shows
that independent molecules of FeaC do not exist in austEfrl'ite._ Z.· Jeffries .and R. S.
Archer, Chern. and Met. Eng., 24, 1057 (1921); 26, 249 (1922).
METALLOGRAPHY 2559
0/0 Carbon 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
0/0 Fe3C 0 3 6 9 12 15 18 21 24 27 30
FIG. 463.
points merge forming a single point, Ara.2, which in turn merges at 0.85%
carbon content with. the Arl point forming the Ara.2.1 point. From 0.85%
carbon to about 2.0% carbon there are two points-the new point being
known as the Arcm point.
It may be seen that the shape of these curves closely resembles that shown
in the diagram illustrating the general type of curve representing .the formation
18 There is much controversy as to the presence of more than two allotropic forms.
Some authorities contend that beta iron does not exist. Recently another allotropic
modification has been noted as occurring at 1400° C. (2552 0 F.), denoting the presence
of delta iron.
19 The "r" in "Ar," etc. is the first letter of the French word "refroidissement."
The points obtained on heating are "Aca" etc., the "e" being the first letter of the
French word "chauffage." Due to hysteresis the transformation points occur at
somewhat IO'Yer temperatures on cooling than on heating.
2560 METALLOGRAPHY
FIG. 464. Photomicrograph of 0.40% Carbon Steel (Cast). Magnified 100 diameters.
. Etched in a 4% alcoholicsolution of nitric aCid. .,
The four photomicrographs shown in Figs. 464, 465, 466, and 4'67 illustrate
the structure of ~ast steel slowly cooled containing varying amounts of carbon.
The black constituent in the above photomicrograph is the eutectoid of
iron and the carbide of iron called "pearlite" because of its resemblance in
color with the naked eye to mother-of-pearl; the white constituent is made
FIG. 465. Photomicrograph of 0.85% Carbon Steel (Cast). Magnified 100 diametern.
Etched in a 4% alcoholic solution of nitric acid.
up of the crystals of iron (in excess of the eutectoid ratio) a1)d is called "ferrite."
Pearlite is relatively strong and hard, whereas ~!'.rite is )V'eak and soft.
Figure 465 shows the structure of slowly cooled· steel which is made up of
the single constituent pearlite, and Fig. 466 snows the same steel under a higher
METALLOGRAPHY 2561
FIG. 467. Photomicrograph of 1.25% Carbon Steel (Cast). Magnified 100 diameters.
, Etched in a 4% alcoholic solution of nitric acid.
The carbide of iron is an unstable compound and in the case of the cast
irons at high temperatures is .readily decomposed into graphite and iron.
The result of this decomposition is that when the iron has cooled to atmospheric
temperature, part of the carbon is present as FeaC, and part as free graphite.
Certain impurities as well as the rate of cooling influence the breaking up of
the carbide, e.g., silicon and slow cooling promote the formation of graphite,
and manganese and fast cooling tend to keep the carbon in the combined
J
2562 METALLOGRAPHY
condition. Figure 468 shows the structure of cast iron anB explains why this
metal is weak, lacks ductility and cannot be forged or hammered. The
black plates of graphite break up the continuity of, the mAss rendering it soft
and weak. . I
FIG. 468. Magnified 100 diameters. FIG. ,469. Magnified 100 diameters.
Etched wjth a 4% alcoholic solution of Etched, with a 4% alcoholic solution of
nitric acid. nitric acid.
/'
FIG. 470. Magnified 100 diameters. FIG. 471. Magnified 100 diameters.
Etched with a 4% alcoholic solution of Etched with a 4% alcoholic solution of
njtric acid. nitric acid.
high temperature for forty~eight hours or more, ca3-~iDg the .(arbide of iron to
decompose into iron and graphite and a small amount of carbon to be removed
from the surface. The graphite particles resulting from this decomposition
due to their size and shape are less d'e trimental to the physical properties than
I \ ' . .
METALLOGRAPHY 2563
t}J.e plates present in gray cast iron. This may readily be seen in Figs. 469
and 471. The G~!'k constituent in Fig. 469 is pearlite and the white is the car-
bide of iron. in Fig. 471 the black constituent is the graphite resulting from the
breaking up of the Fe 3 C and the white the ferrite.
Figure 470 shows the process only partly completed. The black areas are
graphite, the lighter areas pearlite, ,and the white areas ferrite.
+
+
+
+
+
+
+
+
+
~s
Very coarse Coarse Less Fine
.>tructure structure coarse strl)cture
t FIG. 472.
As the steel solidifies at A (or is reheated to TV after the steel has cooled)
the crystalline grains of the solid solution increase in size as the temperature
falls to the critical range. In order to reduce this growth to a minimum, work
should be stopped just above the critical range. If the work is stopped at
for 1', a larger grain growth will resulUhan if the work is stopped atf". Work
performed below the critical range when the metal is composed of an aggregate
distorts the grains into elqngated masses producing harmful internal strains.
In :Figs. 473, 474, 475 are shown the structure resulting from hot working
steels of different carbon content.
20 The term ,. Hot Work" as used here denotes work performed above the critical
range and "Cold Work" that performed below the critical range.
21 After Sauveur.
2564 METALLOGRAPHY
Cold work is performed in wire drawing, rolling sheJt steel, etc. It in-
creases the elastic limit, tensile strength and and decreases the
ductility as represented by the elongation and reduction
one another if a force is exerted which disturbs them. The slip bands appear
on polished and etched surfaces after the metal has been deformed as small
hair lines (actually differences in level) for the most part running in one direc-
tion in the same grain, but in different directions in different grains. In
some metals such as brass (see Fig. 461) annealing after cold work results in the
FIG. 477. Photomicrograph Showing Small Strip of Steel Slightly Bent after Pol-
ishing and Etching. Slip bands appear as black lines in those grains where the strain
was greatest. Magnified 100 diameters. Etched with a 4% alcoholic solution of
nitric acid.
FIG. 478. Magnified 100 diameters. FIG. 479. Magnified 100 diameters.
Etched with a 4% alcoholic solution of Etched with a 4% alcoholic solution of
nitric acid. nitric acid.
and etching whereas slip bands disappear. Their exact nature is not known
but they are probably caused by a suddenly produced deformation, which
mayor may not have been accompanied by fracture.
! I
!566 METALLOGRAPHY
Steel containing 12% manganese or 25% nickel, for example, will remain
austenitic even after slow cooling. Austenite is mineralogically softer than
martensite, but is more resistant to abrasion .. Under the microscope it has
the general characteristics of .a solid solution. Troostite is often present
along with martensite in hardened steel and possesses similar physical prop-
erties, but in a lesser degree. Figure 480 shows the appearance of martensite
and troostite.
Maximum strength combined with maximum ductility is obtained by
producing the constituent sotbite. This can be done by cooling through the
range at the proper speed, but more exactly by hardening followed by heating
to a temperature between 400. and 600 0 C. according to the composition of
the steel and the properties desired. Tro(')stite may also be formed in such a
llJII!flner if a lower reheating or "drawing" temperature is used. The appear-
ance of sorbite when examined under the microscope is shown in Fig. 481.
CASE HARDENING
The core and case may then be heat treated to further improve the physical
properties. Oase-h!1rdened objects appear under the microscope as in Fig. 482.
2568 METALLOGRAPHY
Manganese has a greater affinity for sulfur than . ironl and, therefore, if
sulfur is present, manganese sulfide. is formed. lJnder the microscope this
appears in castings as small round areas, pale gray in coI6r,' and in forgings
as elongated streak~ of the Bame ~l)ll)r. If there is any mangane'ile remaining
after the sulfide has formed, the ~arbide I)f manganese (MnaC) is produced,
which forms a solid solution with the iron.
Sulfur, when present in sll).all amounts (e.g., less than 0.08%), combines
wjth the manganese as explained, but if in excess of this amount, the sulfide
of iron (FeS) is produced which forms around the grains. Steel containing
iron sulfide is weak and brittle at high temperatures. FeS appears a reddish
brown under the microscope.
Phosphorus _is present in solid solution as the phosphide of iron (Fea P ).
In cast irons, due to the large amount of carbon present, a eutectic of iron and
the phosphide of iron is formed. Iron phosphide makes the steel brittle,
especially when cold.
Silicon is present as FeSi, ',\'hich enters into solid solution ,with the iron.
When in small am.ounts, silicon produces no marked influence on the-physical
properties.
METALLOGRAPHY 2569
It will be noted that in one case (Fig. 487)' the bottles containing the solu~
tions are stationary and the liquid in each and in the constant temperature
bath is kept in motion by means of revolving stirrers. This form of apparatus
was used by Moody and Leyson (1908) for the determination of the solubility
of lime in water and is particularly adapted for relatively slightly soluble COm-
2572 DETERMINATION OF SOLUBILITY I
pounds for which rather large quantities of the' satura~ed solution are needed
for accurate analysis. There is also shown in the figure~ the provision for with-
drawing the saturated solution through a filter within th,e inverted ,thistle tube.
The stirrers in the bottles are fitted with mercury seals to prevent access of
air containing carbon dioxide. Other features of the a:pparatus will be readily
understood from the drawing. I
Flexible
Solution
('==r==J::=-=t~~II::
Gas
Inlet
•
Salt
~]r
FIG..489. Berkeley's Apparatus for Determining Solubility.
determinations at temperatures between 30° and 90° and the range of vari-
ations from the set temperature of the bath was, for 2-3 hour periods, within
about 0.2°. For the inner vessel containing the salt, the range was about
0.05°. At each temperature two determinations of density and solubility were
made; one on the solution obtained by stirring a supersaturated solution in
contact with solid salt, and the other on the solution obtained by stirring an
unsaturated solution in contact with an excess of salt.
In the case of determinations at the boiling point a special apparatus was
required. Two forms, described by the Earl of Berkeley (1904), are shown in
2574 DETERMINATION OF SOLUBILITY
Figs. 490 and 491. The first was used for the less soluble salts and consisted of
an outer tube A containing water and an inner tube B'containing salt and solu-
tion: By boiling the water vigorously and closing the .side tube C, steam
passing through the tube '[) stirred the solution thorollghly arl.d the temperature
rose to the boiling point of the saturated solution and remaihed C'Onstant when
saturation was attained. The second form of apparatus (Fig. 491) was devised
Thermometers
/
Filter
Salt
Stirrer
solution. When the temperature of the oil bath was below the boiling point,
salt dissolved; when above, salt was thrown out of solution. Considerable
difficulty was experienced in filling the pycnometer with the saturated solution
without introducing errors due to steam bubbles causea by the suction which
was applied.
panying Fig. 492. A glass cylinder A is closed at each- end jwith large one-DOle
rubber stoppers. The mixture of salt and solution is contained In this cylinder
and is stirred by the rotation of the tube E which is provided with an enlarge-
I
ment at its lower end in which there are two small holes at H and I. The
stirrer rotates in the bearing formed by the hollow woode!l cylinder J. The
glass rod K carries a rubber stopper L which closes the filtering tube M, in
which a platinum cone N supports an asbestos filter O. The siphon P con-
nects the filtering tube with the flask R which is provided with an outlet through
the small tube S. The apparatus is immersed in a constant temperature water
bath W, to about the level shown. After stirring the mixture of salt and
.the temperature of the bath the stirrer was stopped, the end of the small tube B,
which lYas covered with a piece of closely-woven musHn, was dipped below the
surfac~ of'the solution and the liquid drawn into A l)y applying suction at C.
The tube A was then removed, weighed and the contents analyzed.
An apparatus which was used by Donnan and W}lite (1911), for the deter-
mination of equilibrium in the system palmitic acid and sodium palmitate is
shown in Fig. 494. The stirring in this case was accomplished by means of a
current of dry air, free of carbon dioxide. The appar!l-tus consists of two parts,
namely, an ip.ner chamber E, where equilibrium walS attained, and an outer
case A, designed for isothermal filtration. The whole was immersed in a
thermostat to the level W. A side tube B permitte(i connection with a filter
pump. C is a weighing bottle to receive the filtered saturated solution and D
a Gooch crucible provided with a paper filter. The cork, closing A, was cov-
ered with a plastic layer to render it airtight. The tllbe at the lower end of E
was closed with a ground glass plug F, the stem of which was enlarged to a
small bulb at G and then drawn out to ,pass easily through H, leaving an air
free outlet' around it. The small cork I was used to &upport the stopper when
lifted to allow the contents of E to flow down for filtration. The dry air .by
which the mixture was stirred was drawn through K 1)y applying suction ~t H.
'tb..~ \lr~b..~atit\.'b Qf_ tb..i" air wa" a~el)m\llis.b..~d b-s drawin'b it thrullJbQ a.. thin.
spiral immersed in the thermostat. The connection between the equilibrium
apparatus and preheater was made through a mercury seal, which permitted
lifting the apparatus easily without damage to the fragile preheater permanently
mounted in the bath. This apparatus provided for the recovery, separately, of
the saturated solution and undissolved, solid. These authors also des.cribe an
improved l!1ectrically heated and controlled constant temperature bath.
Determinations' at lower temperatures that can l>e constantly maintained
with the aid of a water bath require special forms of apparatus which permit
of temperature control under more or less restricted conditions. ' An apparatus .
of this type, which was used by Cohen and Inouye (1910), for determination
of the solubility of phosphorus in carbon disulfide, is shown in Fig. 495, and is
intended for the range of temperature between -10" and 10°. The satur-+
ating vessel D consists of a glass cylinder to the upper end of wh~ch is cemented
a steel collar E, containing a deep cha9nel. A mixture of litharge and glycerol
was used as the cementing material for this purpose. The inverted steel
cover F fits into the channel of this collar and the seltl of the joint is effected,
in the usual way, by means of a layer of mercury. The cover F is provided
with a brass tube K, to which the pulley M is attached, and is also pierced by
the tightly cemented-in glass tube I. The glass rod (J, containing on its lower
end the three stirring wings H H H, is cemented into the brass tube K. The
saturating vessel is, for stability, tightly fastened in It hole in a block of lead,
S, contained in the Dewar cylinder A. An atmosphere of CO 2 in the saturating
vesse~ is provided by introducing CO 2 under pressure through I and allowing
the excess to escape through the mercury seal in E. After charging the'
apparatus, I is closed with a rubber tube and plug and the stirrers H H H set
in motion. A Witt stirrer, 0, keeps the contents of the bath in rapid circula-
tion, Water is used in the bath for temperatures about 0°, and alcohol for
those below 0°. The regulation of the'temperature is Itccomplished by addition
2578 DETERMINATION OF SOLUBILITY
of ice or solid CO 2 as found necessary and, therefore, requjres very close atten-
tion on the part of the experimenter.
A novel and simple form of apparatus, which was u~ed by Ba,hr (1911),
for the determination of the solubility of thallium hydroxide at temperatures
up to 40° is shown in Fig. 496. As will be seen, this consist.s of a gas washing
flask to the arms of which a Y tube provided with two stop-cocks is sealed:"
The .inside walls of the apparatus were coated with hard paraffin and the
FIG. 495. Cohen and Inouye Apparatus FIG. 496. Bahr's Apparatus for De-
for Solubilities. 'termining Solubilities.
To Pump
R1 ",, I
,
I
I
I I I
f I I I
~I tI \I \I
I I 11
: ; : I I
I, ~... ) J
\ .....~:::.,~,.
decrease of volume or the two layers, as estimated from the change in position
of the meniscus, the reciprocal solubility of the two liquids is calculated. It is
to be noted, however, that the solubility of liquids is in practically all cases
reciprocal, and without an analysis of the two layers the true solubility can
not usually be deduced.
Titration Method.-A special case of the reciprocal solubility of liquids is
that representing equilibrium in ternary systems yielding two liquid layers.
Such equilibria are usually determined by relatively simple titration pro-
cedures, but for the interpretation and description of the results, special terms
have been developed and these require more or less detailed explanation.
When a third liquid is added to a mixture of two others which are miscible
to only a slight extent, the added liquid, if soluble in each of the others, will
distribute itself between the two and an equilibrium will be reached. If the
two layers are then analyzed and the results plotted on coordinate paper, two
points, corresponding to the two layers, will be obtained. If more of the third
liquid is added, equilibrium will again be established after a short period of
shaking and the analysis of the two layers, to which the designation conjugate
layers has been given, will fix two more points when plotted on the coordinate
paper. The process may be repeated until a considerable number of points
have been obtained. When this has been done, it will always be found that
these points are the locus of a smooth curve, to which the designation binodal
curve has been given. If the pairs of points corresponding to the conjugate
layers are. connected, the lines so obtained are defined as tie lines. Since it is
evident that with the continued addition of the third or consolute liquid, a
point .\Dust finally be reached at ,which the resulting mixture will no longer
s'eparate into two conjugate layers, the tie lines successively determined as
above described will become shorter and shorter until finally the last one is
reduced to the point corresponding to the homogeneous mixture of the three
components. To this is given the name plait point.
Although for the above example a ternary system made up of three liquids
has been taken, there are a large number of salts and other solid compounds
which, when dissolved in mixtures of liquids of certain concentrations, cause
the latter to separate into conjugate liquid layers. These systems have aroused
much interest from time to time and considerable data for them are given in
the literature.
Since~t is usually difficult and frequently impossible to analyze directly a
homogeneous mixture of liquids, and thus determine the points on a binodal
curve, a simple titration method for this purpose has come into general use.
By means of this a homogeneous mixture of known amounts of two of the
components is titrated with the third just to the point of initial separation of
the second layer, which is usually very sharply indicated by the appearance of
clouding or opalescence. The procedure may also be reversed and the con-
solute liquid added just to tile point of clearing of the clQudy mixture of the
other two. By this plan the synthetically derived composition of one of the
two conjugate layers and thus of one point on the binodal curve is known.
The determination of the tie line and therefore, the identification of the cor-
responding point on the curve for the conjugate liquid, requires an additional
experiment for its location. Several procedures for this purpose have been
developed. They usually depend upon the determination of one or more
2586 DETERMINATION OF SOLUBILITY
shown in the figure. Connection with the air pump is made at D. During the
boiling the lower opening of the inlet tube E is above the surface of the liquid
in C, the stopcock B being closed. When the air.has been completely expelled,
the screw pinchcock F is closed while the air pump is still in operation. The
flask C is now raised until the lower end of E reaches nearly to the bottom of the
flask. The air pump is now connected at G and the cock A opened so as to
make connection with the pipet. B is now opened .and the inflow of air
through D regulated by gradually opening F in such a manner that the liquid
is very slowly forced up into the pipet. In this manner the liquid never comes
into contact with the air under full atmospheric pressure but only under greatly
diminished pressure. The absorption of air under these conditions can only
be inappreciable, especially since the liquid in the flask remains perfectly quiet,
and only the lower portion is used."
2590 DETERMINATION OF SOLUBILITY
,
Having filled the pipet B, Fig. 499, with the air-free solvent as just described,
"T is connected with the source of gas supply and the cocks C and D are turned
in such a way as to allow the gas to sweep out th~ air from the capillary, M.
The buret is then filled in the usual manner by lo\yering the leveling tube F,
the cock D having been turned so as to connect T \with E. Care is taken to
keep the entering gas under a slight pressure by keening the mercury level in F
slightly above that in A. This prevents air from ehterink through any leaks
in the train connecting the gas generator with the buret." The gas must be
completely saturated with the vapor of the solvent and this, with other than
aqueous solvents, may require, in addition to drawing it through some of the
solvent in H, that a thin layer be placed in the buret and time allowed for it to
saturate the gas sample.
'; After again allowing the current of gas to flow through the capillary M
for a short time the buret and pipet are connected with each other by turning
the three-way cocks D and C in the proper direction. The determination of
the amount of absorption is then made as follows: A portion of the- gas is
passed into the pipet by raising F and opening G, the displaced liquid being,
caught in a graduated cylinder. The cock q .is closed and the gas and liquid
in the pipet brought into intimate contact with each other by shaking the whole
apparatus. C is now opened to allow gas .to enter from the buret to replace
that absorbed. This process is repeated until, on ope~ing ·c, there is no further
decrease in the volume of gas in A. The volume absorbed is found by sub-
tracting from the original volume of gas, the volume remaining in the buret
plus the volume in the pipet. The volume of gas in the pipet is equal to the
volume of liquid drawn off. The volume of liquid remaining is easily calcu-
lated from·th!l known volume of the pipet. The absorption coefficient or
'solubility' is the ratio of the volume of gas absorbed, measured at th~ tempera-
ture of the experiment, to the volume of the saturated liquid. It may be
reduced to the coefficient used by Bunsen by dividing by (l+dt)."
In the case of the majority of investigators who have used this method,
particularly for determinations at high or low temp~ratures, the absorption
pipet has been kept at the temperature of the experim~nt and the gas measuring
buret at room temperature, the two being conhected by mean.s of a flexible.
capillary which permits the absorption pipet to be independently ·shaken.
This arrangement makes it necessary, in calculating the absorption coefficients,
to apply the usual corrections for temperatu~e and vapor pressure to the
volume pf gas in the buret~ This is a complication which in some cases causes
uncertainties in regard to the accuracy of the results as finally calculated.
An apparatus designed for determinatio~s at very high presirures, using a
Caillet compression tube, is described by Sander. (1911-12). It was used for
determination of the solubility of carbon dioxide in water, alcohols, and other
organic solvents. The principle involved is that the pure gas is ·first com-
pressed above mercury in a graduated tube an~ the ·volumes correspondiNg to
given pressures noted. Similar readings are then taken for the same gas after a
i:\maU accurately measured amount of solv~nt has been introduced into the
graduated tube. The difference between the two'volumes at the same tem-
perature and pressure, reduced to 1 kg. per sq. cm:_...aRd-1 ml: of liquid, repre-
sents the solubility of the gas in the given solvent. -
DETERMINATION OF SOLUBILITY 2591
Finally, attention should be called to the method of determination of gas
solubility based on the principle that, for volatile solutes which obey the laws
of Dalton and Henry, the amount which is carried away by an inert gas when
known volumes are bubbled through solutions of known strength of volatile
solute, can be used to measure the comparative solubilities in solvents of differ-
ent concentrations. An example of this method is the determination of the
solubility of ammonia in aqueous salt solutions by Abegg and Riesenfeld (1902).
The very ingenious apparatus consists of a generator for developing a stream
of H 2+0 2 from aqueous NaOH, by means of an electric current measured with
the aid of a copper voltmeter, and the volume of gas thus determined. This
was passed through a spiral in the vessel containing the ammonia solution or
known concentration. The mixed gasses passing out of thIs were received in a
third vessel containing 5 m!. of 0.01 n HC!. Electrodes were provided in this
vessel and, by means of conductivity measurements, the point determined at
which all of the HCI became saturated with NH a. Since the volume of the
+
H 2 O2 required for this purpose was known, the partial pressure of the NHa
in the mixture could be directly ascertained. Comparative determinations of
the'vapor pressure of the ammonia in water and a series of salt solutions made
in this way were calculated to ammonia solubilities on the basis of the relation
that, for two solutions of equal ammonia content, the ammonia pressure is
reciprocally proportional to the solubility of the ammonia in them.
I
SPECTROGRAPHIC ANALYSIS
, 1 ,
..
FIG.5010 Energy Level Diagram for Potassium (Numbers on lines between
levels indicate the wavelength of the emitted energy.)
quartz spectrograph is 1,lsed the dispersion is nearly linear for the spectral region
when plotted on a frequency scale so that the accuracy of plotting or reading
is approximately the same in all regions of the spectrum. With a grating
spectrograph the data is observed on a linear or nearly linear dispersion based
on the wavelength scale. For all spectrographic data involving line or emission
spectra the data are recorded in wavelength in angstrom units, and for nearly
all types of qualitative and quantitative analysis this is the only system' with
which one need be familiar.
APPARATUS
Spectrographic analysis involves rather complicated apparatus and requires
a considerable capital outlay with regard to the cost of the equipment. The
upkeep, however, is slight and once the apparatus has been properly adjusted
and calibrated the technical knowledge for the preparing of the photographic
data and the interpretation of the data is such that much of the work can bEt1
done by technical assistallce in the laboratory., The .type of spectrograph
_ necessary for analytical work w~ll depend on the (lomposition of the samples to
be analyzed, the degree of accuracy expected and the speed with ;Which the
analysis must be made. For many of the non ferrous metals giving simple
spectra, such as Zn, Cd, Hg"Ag, Au and the like a high dispersionspectrograph
is not necessary. The usual type of Cornu prism instrument (Fig. 502) with
a dispersion of about 10 inches from 7000 to 2000 A is quite satjsfactory, and in
some cas!,!s a slightly lower dispersion such as given .!>l'--a small model Littrow
or the Cornu instruments may be sufficient for puqfoses of analysis (Figs. 503
and 504).
SPECTROGRAPHIC ANALYSIS 2595
For the analysis of ferrous metals, W, Mo, U, Pt and elements giving large
numbers of lines it may be necessary to use an instrument capable of giving
sufficient dispersion to separate the spectrum lines. For this purpose one
usually employs a grating instrument or a prism instrument of the Littrow
. typ~ (Figs. 505a, band 506).
Both of these instruments (C~rnu and Littrow) employ the same optical
principles as indicated in the illustrating figures (Figs. 502 and 505). The
prism spectrograph or spectroscope utilizes the optical principles of refraction
and dispersion, i.e., the bending of a light beam by a prism and the fact that
FIG. 503. Bauscl;t and Lomb Spectrograph Medium Size (4X10 in spectrum plate).
ove'r the entire spectral region to which the phot"ographic plate is sensitive.
The'light which is focused on the slit of the instrument is collected by a collimat-
ing lens so as to project a parallel beam through the i~rism, After refraction
by the prisms and the dispersion of the beam the light is focused by a second
lens on the plate or eyepiece, It should be noted that in the Lit.t row instru-
ment a single lens serves both as focusing and collimating, and the prism 1s only
30° with a mirro~ back so that the light beam travels. half 'lay through !"hat
would be a 60° prism if it were continued, and then is reflected 'b'ck so that the
beam passes through the prism twice, In view of.ltlle economy of .quartz
pieces, as well as size of the instrument, the Littrow construction is particularly
" well suited for .the larger instruments with high dispersion.
SPECTROGRAPHIC ANALYSIS '2597
The illumination of the slit of the spectrograph may be varied to suit the
conditions of the analysis. While the common practice is to focus the image of
the source on the slit, and so diaphragm the slit with a wedge or Hartmann
diaphragm (Fig. 507) as to cut off undesirable illumination from certain portions
to diaphragm part of the source one.may use screen absorbers near the focusing
lens to reduce the total intensity. For certain types of ~nalysis it is desir'\ble
to have the slit illuminated uniformly along its entire length; this is especially
true in the use of the logarithmic sector method or the step slit, method of
analysis. This can be easily accomplished by the 'use of a.long focu::, tens placed
as near the slit as IS po5'sible and placing
the light source irl. such a position...that
the image of the source is focused on
the collimating lens of the spectrograph
(L1-Fig. 502). If one attaches a rec-
FIG. 507. Wedge and Hartmann Dia- tan gular diaphragm to this lens it is
phragm for Spectrograph Slit. possible to eliillinate,the undesira,ble por-
,tions of the spectrum. This system
of illumination does not make full use of the resolving power of the instrument,
but in most cases this is not necessary and the reduction in the light intensity
is not a serious handicap in most analytical work since the increased e~posure
provides for a more thorough sampling of the material to be analysed. >
LIGHT SO,tJRCES
Flame.-Flame sources are useful for rough qualitative tests which may
be applied in addition to visual color tests of the flame, for, such elements av
Cu, Na, Li, K, Ca, Sr;Ba, etc. The temperature a.nd ,constancy of the flame
are not very satisfactory for the spectrographic analysis and with the exception
of the alkali metals, those elements which give colored flames, usually do so as
compounds with certain anions (i.e. QaCI+, CuCI"", etc.), the color being caused
by the band spectrum of the diatomic or triatomic molecule rather than by the
simple element' A quantitative method has been described, however, for the
estimation of boron by the color of the green flame produced-by the burning of
methyl borate, by which relatively small amounts of boron may b;aetected and
estimated quantitatively.
Arc.-The two most important methods, of atomic excitatiqn for spectro-
graphic analysis are by means of the electric arc and the electric spark. It is
difficult,to draw a sharp dividing line between these methods especially when
high voltage arcs of 2 to 3 thousand volts are comtnonly used. In general,
however, the arc sources are confined to the use of current densities of 3 to 5
amps. and 50 to 200 volts (Fig. 508). The arc constancy can be improved by
the inclusion in the circuit of a reactance (such as the primary coil of a 1 KVA
transformer in which the secondary circuit has been shorted). While the
current source may be from 120 to 220 volts (DCl.,tli.e-actmtl drop across the
arc gap may be only 40 to 60 volts, due to the use of a resistance in the circuit.
SPECTROGRAPHIC ANALYSIS 25~9
One usually uses rods of the sample about t inch in diameter. Metals are
usually worked by using pairs of rods of the material. For non-conducting
substances and small quantities, as well as solutions, carbon rods are usually
used. In such cases the,lower carbon is usually cupped so as to take the sample.
The lower carbon may be the positive or negative electrode, depending on the
substance for which the analysis is being made and the spectral characteristics
desired. The commercially available carbon corries in several grades, of which
the spectrographic and high purity types are suitable for most analyses. The
highest purity material is usually free from nearly all impurities with the excep-
~
L
~
Spark
D.C. R
Supply Arc
o_ _ _ _ ~~~_ _ _ _~~t
.[
FIG. 508. Electrical Diagram for Spark and Arc Sources. T=Transformer 10,000
to 20,000 V.; L = Self Inductance 0.25 millihenry; C = Capacity, 0.005 microfarads:
D.C. Source 110-220 volts; R=Resistance; .Arc current 2 to 5 amps., 40-70 volts;
I = Reactance. .
tion of a trace of Si. Less expensive rods usually contain minute traces of
Fe, AI, Ca, Cu, Ti, Mg, and V. These may not interfere with the analysis
or they can be removed by purification. One method of purification involves
the soaking of the carbon in HCI and HNO a• A recently described method of
purification which yields excellent results is effected by heating the carbon as a
resistance rod to a temperature of about 2400° in a vacuum furnace. This
treatment seems to be effective in the removal of nearly all of the impurities
from carbon rods containing a considerable amount of impurity. In some cases
it is pQssible to use pure metal electrodes, such as copper as the holder for the
substance to be examined. Duffendack and Wolfe have recommended the use
of a high voltage arc using 2.5 amps. at '2500 volts. It is claimed,that this pro-
cedure reduces if not entirely eliminates background and that in the examina-
tion of salts the thin layer of salt solution on the carbon rods, formed by the
evaporation of a drop of solution on the rods is sufficient to prevent the produc-
tion of the carbon-cyanogen bands. The temperature of the electrodes remains
very low so that low-melting alloys may be conveniently analysed by this
method. .
Spark.-The change in electrical excitation from arc to spark results in
an alteration in the spectrum with regard to intensities of lin(Os, production of
SPECTROGRAPHIC ANALYSIS
new lines and the elimination of arc lines. :rhe usual electrical system is
indicated in Fig. 508. The use of a self inductance changes the spark charader-
istics with a reduction in the intensity of air lines (from 0 and N) and in some
c.ases may so alter the characteri~tics of the spark so thaH~t. may appear. I?-0re
hke an arc spectrum. The self mductance does help stabIhze the condItIOns,
and it is to be recommended in spectrographic analysis,) especially where it
may be desirable to have reproducible conditions. There appears to be a satis-
factory medium between the condensed and oscillatory spark since the factor
determining the number of oscillations is a function of tIre product of the
capacity of the condenser system and the inductance of the self inductance.
Since the equipment available in each laboratory will differ in some detail
and the' conditions of excitation must be vaHed for different elemental analysis,
it is not practical to give in this limited discussion any more detailed descrip-
tions of arc and spark procedures. Neariy all articles on spectrographic
analysis indicate with some detail the method of excitation and in attempting
to duplicate such methods it is advisable to follow th~ excitation proced)lre as
closely as possible. In the spectrographic analysis of solutions the sample
may be applied to a carbon electrode and arced or sparked or a spark d~scharge
may be effected by using the solutio.n as one electrode and a carbon, gold' 01:
copper rod as the other electrode.
Other Light Sources.-For the analysis of &ases one usually uses a gas
discharge tube with a high frequency high voltage current. The presence of
impure gases may often interfere with, or eve~ completely eliminate, the lines
of certain substances so that the method may not be entirely satisfactory.
For the purpose of providing a b,Jtckground of nearly continuous radiation for
the observation of absorption spectra one may use ,the incandescent tungsten
filament for the visible and near ultraviolet. The ultraviolet range may be
extended a little by the use of a special bulb of correx glass or a quartz window
in the glass bulb. For an extension in the ultraviolet to the limit ofthe gelatin
emulsion one may use an under water spark from a tesla coil or a hydrogen
discharge, both of which give a continuous spectrum, without spectral lines over
nearly the entire photographic range from 2000 to 8000 A: For some types of
absorption work, especially where the bands are broad, satisfactory results can
be obtained by using an arc or spark between tungsteiJ., iron or nickel electrodes.
Spectrographic Standards.-In the definition of dispersion measurement
terms it was indicated that rather than defining wav~length. in angstrom units
as a ten thousand millionth of a meter it was defined as 1/6438.4696-part of the
wavelength of the cadmium red line, and this line is known as the primary
standard.' As further guides in the measurement of the wavelength of un-
known lines a number of secondary standards have been designated and their
wavelength .accurately measured. Tliese secondary'standards include some
217 iron, 9 copper, 3 silicon, 2 nickel and 25 neon lines. I The almost universal
use of the iron spectrum as a means of identification and comparison of other
spectra has resulted in the accurate determination of the wavelength values
of" some three hundred and eighty-four additional lines which are known as
tertiary standards. The wavelength values of unknown lines can he accurately
determined by measurement of the spacing between the Jlnknowti line and se-
lected standard lines and the application of the Hartmfnn formula to these
data •.
SPECTROGRAPHIC ANALYS!S 2601
TABLE I. Continued
,PI 1787.5
2136.8
Ru I 3436.7
13499.0
Sn I
SrI
4524.7
14607.3
VI 3183.4
3184.0
2149.8 3596.2 4832.1 3184.0
2536.4 Ru II 2678.7 4872.5 13185.4
2554.0 2692.1 4962.3 VII 13093.1
Pb I 12170.0 2712.4 Sr II 14077.7 3102.3
2833.1 2945.7 4215.5 3110.7
C> 3639.6 2965.6 Ta I 13311.1 3118.4
3683.5 2976.6 3318.9 3125.3
4057.8 Sa I 4390.9 3406.7 WI 14008.8
1682.4 4424.4 Tb I 3509.2 4294.6
PI) II 12203.6 4434.3 3561.8 4302.1
Pd I 13404.6 Sb I 2068.4 3848.8 WI! 2397.1
3421.2 2175.9 3874.2 2589.2
3517.0 2311.5 Te I 2142.8 3613.8
3609.6 2528.5 2383.3 YI 4643.7
3634.7 2598.1 2385.8 14674.8
Pd II 2488.9 3232.5 2530.7 Yt I! 13710.3
2498.8 3267.5 2769.7 3774.3
2505.7 Se I 3907.5 Th I 3538.8 3788.7
2658.7 13911.8 3601.1 Yb I 3289.4
2854.6 Se II 13613.8 4019.1 3694.2
Pr I 4062.8 3630.8 Th II 3290.6 3988.0
4179.4 3642.8 Til 3635.5 Zn I 12138.5
4189.5 Se I 1960.2 3642.7 3282.3
4225.3 2039.7 13653.5 3302.6
Pt I 2659.4 2062.6 4981.7 3344.5
2830.3' 14730.9 4991.1 Zn II 12025.5
2929.8 4739.1 4999.5 2061.9
2998.0 4742.3 5007.2 Zr I 13519.6
i3064.7 Si I 2506.9 5014.3 3547.7
Ra I 14825.9 2516.1 Till 13349.0 3601.2
Ra II 13814.4 2528.5 3361.2 4687.8
4682.2 12881.6 3372.8 4710.1
Rb I 4201.8 3905.5 3383.8 4739.5
4215.6 Si II 1526.8 TIl 3775.7 4772.3
17800.3 11533.6 15350.5 4815.6
7947.6 Sn I 2840.0 Tu I 3462.2 Zr II 13392.0
Rh I 3323.1 2863.3 3761.3 3438.2
3396.8 3009.1 3761.9 3496.2
134.34.9 3034.1 UI 355'2.2 357'2.5
3658.0 3175.0 3672.6
3692.4 3263.3 4241.7
I
FIG. 509. Arc Spectra of Copper and 13ras&.Samples•
. _r-
SPECTROdRAPHIC ANALYSIS 2609
ableait the present time is the Eder and Valenta atlas, which includes arc, spark
and flame spectra of nearly all of the available elements, taken with glass prism,
quartz prism and grating spe<;trographs. The dispersion, h~wever, is not
satisfactory for the accurate identification of many elements, especially where
the spectrum is complicated. Adam Hilger have published several photo-
graphic tables, one of which gives the persistent lines of a large number of
elements. They also have published iron, copper and neon spectra in which
the dispersion is high and a large number of the lines identified. To further
aid in the identification of other elements, tables have been prepared in which,
in addition to the iron spectrum, there is indicated the position at which a
strong line of some other element would appear if it were present. Tables of
this type have been published by Bardet, Crook and others. From the tables
in this' chapter and with the use of an accurately calibrated scale on the instru-
ment or with an iron spectrum in contact with the unknown spectrum and a
photographic identification of the iron spectrum, one can satisfy nearly all the
requirements of the qualitative analysis of elements 'by the spectrographic
metho.d.
QUANTITATIVE ANALYSIS
de Gramont has proposed that the.ultimate lines or the Raies Ultimes of the
elements are to a certain d~gree the basic lines of the element and as such
their intensity will be proportional to the amount of the element present. The
first practical application of this theory to quantitative analysis was made by
Meggers Kiess and Stimson in the analysis of platinum metals. In all of the
suggested' methods of quantitative spectrographic analysis the results are
dependent on the known chemical composition of comparative standards.
One might subdivide the me_thods into those in which the comparative standard
is observed under similar conditions to that of the unknown observation, and
those in which the standard substance is added to the unknown so that both
photographs are made simultaneously. .
Comparison with Standards.-tJertain of the tentative standard and
recommended methods of analysis of materials by the spectrographic method
involve the comparison of line intensities between standard photographs and
unknown photographs taken under similar conditions. If the conditions of
exposure, development, slit, current density, and type of plate are maintained
it may not be necessary to repeat the photograph of the known standard bu~'
merely to compare with a previously prepared standard negative. A certai:'l
amount of possible error can be avoided by the taking of a photograph of the
standard sample adjacent to the unknown sample on the same negative. In
many cases this method of comparative analysis is satisfactory. On account
of the large number of variable factors which exist it is not possible to supply
2610' SPECTROGRAPHIC ANAL Y;SIS
-3.0
-5,0
c
Vl
I
co
L9G SECTOR METHOD
-7.0
%
SJ· Log % 0.50 0.75 -0.00 1.00
-1.50 -1.00 -0.50
FIG. 510. Working Curve for the Analysis of Bismuth, Using Tin as the In
ternal Standard. Determinations made by the Log Sector Method. Bi=2989.04 A
Sn=3009.14 A.
~o
-.04
gjl..,08
,_J
PHOTOMETRIC METHOD
-:12
TABLE III
REPORT ON WORKING CURVE DATA
The choice of line pairs for comparison will of course determine the concentra-
tion range over which the data will best apply. Over the major portion of the
working curve an accuracy to within about 5% of the amount of 'unknown
element can be determined. For high concentration the. method may be used,
although this same percentage error will still hold and is usually much greater
than that obtained by other analytical methods. For extremely small amounts,
however, an error of 5 or even 10% of the amount present may be quite per-
missible.
Measurement of Intensities of Lines.-The determination of the blackening
or line densities on the. photographic plate may be photometrically measured
by visual or electrical means. A number of recording and semi-automatic
densitometers for the measurement of plate and line density are available for
this purpose. These instruments may have certain constant errors such as the
failure to obtain a linear response from the photo cell over a range of varied
light intensities. A combination of densitometer and spectrograph can be
effected by placing directly in front' of the spectrograph slit a rotating sector:
cut either smoothly or in steps and usually cut in a logarithm spiral or in
logarithmic steps, or by the use of a step slit such as the Hansen or Duffendack
type. This procedure produces a spectrogram in which the line tapers in its
degree of blackening until it fades completely out, or is reduced by steps to
2612 SPECTROGRAPHIC ANALYSIS
FIG. 512. Typical Spectra by Log Sector and Ordinary Spectrum Photographs
_ Used in the Preparation of Working Curves.
termination of the line may be eliminated in part by the }l§§ of a reversed line
. in the eyepiece and matching the two lines to a uniform,crensity.
, In methods where comparison is to be made between spectrograms, especi-
ally where the spectrograms are on different plates, it is somewhat important
to determine the plate characteristics and many workers feel that this procedure
should be carried out when the data are all on the)lame plate. The variation
SPECTROGRAPHIC ANALYSIS 2613
of the plate characteristics (i.e. the d/log I curve) with different spectral regions,
nature of exposure, development time and temperature, etc., may require that
for accurate work the plate characteristics should be determined through the
use of a standard lamp for illumination of the spectrograph and that correction
be made so as to indicate the absolute intensities of the lines. In routine
analysis where the same plate treatment is used, the same type of plates and
photometric methods, it is possible to prepare working curves which will give
accurate quantitative results.
Control of Conditions.-An important working condition in quantitative
spectrographic analysis which must be carefully controlled involves the light
source. From a study of the cause and nature of spectrum lines it will be
noted that there are successive stages of ionization of the element, which are
dependent on the applied voltage and temperature. The spectrum produced
under 10':' io:t;lization conditions may be quite different in the lines present and
their intensities as compared with the spectrum from the highly ionized
element. This change in relative intensities of the emission lines with a change
in exciting conditions requires a careful duplication of conditions when making
a series of observations, and requires a rather exact definition of the exciting
conditions if one is to apply the working curve or analytical data prepared and
published by others. Gerlach has attempted to define the working conditions
by which repeated observations could be made, through the selection of lines
of the known, element from different states of ionization, which at the given
conditions of observation are equal in intensity. When they are equal the
excitation can be said to he at the standard or specified condition. Gerlach
has then varied the amount of "unknown" dement with respect to the internal
standard (the internal standard may be the element present in largest amount
or an added standard of known amount) and selected for each concentration
pairs of lines of the standard known and unknown which are of equal intensity.
These pairs are known as homologous pairs. In view of their equal intensity
it can be seen that these results will not be appreciably influenced by plate
characteristics and other errors which have been mentioned. If the internal
standard is to be some element other than the substance present in largest
amount (the pure substance containing the impurity for which the analysis
is being made) it will be necessary to control the amount of internal standard
which is added. This is accomplished by choosing a pair of lines, one from the
pure or base substance and the other from the internal standard, 'which are
equal in intensity under the excitation and concentration conditions which are
used for the observation. This method may be better understood from Table
IV whieh is a portion of an analyses table published by Scheibe.
TABLE IV
Quantitative analysis for Titanium in an Aluminum sample. Control of excitation
conditions Al II 2616.18=AI I 2575.11 A. Internal standard control or coupling,
pair Al I 2652.48=Cu 2618.38 A.
Wavelength Wavelength Percent of Ti in Sample when
Titanium Copper Lines Are of Equal Density
Ti II 3361.22 .............. 3307.95 0.26
Ti II 3361.22 .............. 3337.85 0.204
Ti II 3335.19 ..... , ........ 3349.26 0.188
Ti II 3361.22 ..... , ....... .4022.70 0.166
Ti II 3361.22 ............. :3365.36 0.109
2614 SPECTROGRAPHIC ANALYSIS
It would seem that while this method may have eliminated a number of the
possible errors of other methods it involves the setting up ofidefi:tlite conditions
of excitation, internal standard concentration and line intensity mealffirement,
which in the end are required of nearly all methodFl whichl give reproducible
results. In the method described by Gerlach the spectrum of the so-called
internal standard may be taken separately, either on top lof or adjacent to
the sample to be analysed and need not be that of a mixture of the sample and
internal standard, just so long as the homologous and coupling or control pairs
are equivalent in density. ,
In order to control more carefully and standardize the spectrographic data
in an analysis, Duffendack and his coworkers have developed the use of plate
calibration through the use of a step slit so as to remove possible errors which
might 'result in the failure of the negative to yield a linear d-Iog I curve and
of the same slope for all wavelength values, .
It would seem that while the method of quantitative spectrographic analysis
has been rather carefully studied and applied to a large number of prQblems
with considerable success, it is not possible to give such specific data as is avail-
able for qualitative spectrographic analysis in the w!1velength tables. Because
of.the nature of the method it is necessary in nearly all cases to prepare standard
negatives, mixtures or working curves which apply to the particular spectro-
graph, photographic plate, photometer or other influencing factors. The value
of published work in the field of quantitative spectrographic methods lies,
largely in the development of technique and the experirnental determination of )
satisfactory groups or pairs of lines for specific analytical purposes. If
Plates.-For the usual type of qualitative or quantitative spectrographic
analysis a medium speed photographic plate such as the Eastman 33 or 40 is
quite satisfactory. Development can be by elon-hydroquinone (metol-hydro-
quinone) or other standardized developer. For spectral regions 'beyond 5000 A
panchromatic or special red sensitive negatives are available to about 1,4,000 A.
,
ABSORPTION SPECTRA
When an obj~ct or solution absorbs a portion of the homogenous spectral
radiation which passes through it we describe the J;Ilaterial as being colored.
The absorption of homogenous spectral radiation by molecules is usually quite
selective and limited to definite spectral regions, or bands. These absorption
bands are usually diffuse, with partial absorption ove~ a considerable region
and a gradual increase in absorption to a maximum of absorption near the
center of the absorption band. The intensity and position of the absorption
maximum, as well as the rate of increase in absorption on eag}l side of the
maximum are definite properties of the compound being_exammed and may
serve to identify or characterize the substance. WithIn. the visible spectral
'region one might indicate the production of color by absorption bands as follows:
SPECTROGRAPHIC ANALYSIS 2615
TABLE V
'RELATION BETWEEN ABSORPTION BAND ,POSITION AND OBSERVED COLOR
Position of Absorption Band Color'Absorbed Color Transmitted
'450 filt. . . . . . • . . , .. Blue Yellow
500 ............... Blue-Green Orange
550 ........ , , , , . , . Blue-Green-Yellow Red
600 ..... , , . , . , , ... Green-Yellow Red and Blue
650 ..... , .. , .. , . , . Yellow-Orange Blue
700 .... ,., .... ,., . Red Blue-Green
400 and 750 ... , ... Blue and Red Green
Since absorption bands are usually not well defined like emission lines it is
the practice to record wave length values in millimicrons (mtL=metersXlO-9),
For work in the ultraviolet region the data is often recorded in frequency or
wavenumber, since absorption bands represent in many cases harmonic
vibrations which give regularly spaced bands on a frequency scale.
The intensity of absorption is an important value for the quantitative
estimation of the absorbing material. Where one is interested in a description
of the transmitted color one usually records the per cent of light transmitted
~"Ot variobs wavelength values. If one is interested in the amount or character
of the absorbing substance one usually records the extinction coefficient. The
extinction coefficient may be defined as the Log of the incident intensity divided
by the transmitted intensity (log 1 0 /1). By recording in extinction coefficient
values, the height of the absorption band is directly proportional to the con- •
centration or cell thickness (assuming that Beer's law holds for the material).
,From the extinction coefficient of the maximum of the absorption band of a
substance one may obtain a much more accurate value for the concentration
... than through a colorimetric measure of the transmitted light. In addition, the
position and shape of the curve may enable one to identify the molecule pro-
ducing the absorption. The production of absorption is dependent on ,the
I presence in the molecule of an absorbing or chromophoric group. Not all
'substances absorb radiation within the visible or photographic ultraviolet
region, and hence it is possible to detect and identify traces of absorbing
materials in otherwise transparent materials. For example, it is possible to
detect and quantitatively estimate the amount of benzene present in ethyl
alcohol, with an accuracy of about 10% of the amount present, when the
concentration may be as low as 0.001 %. The spectrophotometer has been
used in the quantitative estimation of such colored inorganic substances as
cobalt, iodine, copper, and the rare earths (Fig. 513). For the purpose of
estimation of impuriti~s and·their detection one may use the'method of sub-
tractive absorption in which one subtracts the absorption band of the known
I pure ,substance from the observed band, the difference representing the impurity
absorption (Fig. 514). Commercial dyestuffs are standardized and their
concentration determined by absorption spectra methods which give rapid and
accurate quantitative results. Much of the synthetical work in the fields of
vitamins, hormones and natural pigments is based on the absorption spectra
data obtained by the study of these and related compounds.
The data in the field of absorption spectra measurements is so varied that
it is not possible to give methods of procedure and lnterpretation of data. For
a more detailed discussion of these fields one should consult the references in the
bibliography which follows.
2616 SPECTROGRAPHIC ANALYSIS
~
::: 1.00 !.,-----+---,_,J-+--i-_.,.
j
700 800
Wavl> j&ngth, m p. ,
FIG. 513. The Absorption Spectra of Cobaltous Halides in their Corresponding
Halogen Acids - - - CoCh in HCI; ----- CoBr2 in HBr; - - - CoI2 in HI.
(Brode,.J. Am. Chern. Soc., 53, 2457 (1931).) /
000
-_ -_ /.
1 .00
,:/ l7
0.90
2 o.so
o.20 0.70
"
",
0.60
/
o.40 " ,f 0.50
0.40
'f ...
1/ 0.30
o.60 <7
.so 3 // I 0.20
I 0.15 ~
.00 1. I/ 0.100 E
.~
/ ~:g~g ~
~\
I
1.20 0.070 ..:
/' 0.060
1.40
\ \ __ .I 1/ ~
0.050
0;040_
.60 \J / 0.030
0.020
.
1.S0
0.015
•00
0.010
400 420 440 460 480 500 520 540 560 580 600 620 640 660 ~80 700 :720 740 760
Wave Length
References
Comprehensive discussions of Spectrographic AnalysiS with an extended list of
Bibliographical references are to be found in the following:
1. "Chemical Spectroscopy," W. R. Brode, John Wiley and Sons, New York (1939).
2. "Spektroskopische und Radiometrische Analyse," Scheibe, Akad. Verlag., Leipzig
(1933).
3. "Metallurgical Analysis by the Spectrograph," D. M. Smith, British Non-Ferrous
Metals Research Association, Monograph No.2. (1933).
4. "Die Chemische Emissions Spektralanalyse," Gerlach et al. Verlag Voss, Leipzig,
3 Vol. (1929--1933-1936), English Translations of Vol. 1 and 2 published by Adam
Hilger, London.
Analysis Tables:
Kayser, "Handbuch Der Spectroscopie," 8 Vol., Hirzel, Leipzig.
"International Critical Tables," McGraw Hill.
Kayser, "Hauptlinien der Linienspektra aller Elemente," Springer, Berlin (1926).
Smith, "Visual Lines," Hilger, London (1928).
Twyman and Smith, "Wavelength Tables for Spectrum Analysis," Hilger (1931).
~General:
Bardet, "Atlas de Spectres d'arc," Doin, Paris (1921). •
Baly, "Spectroscopy," 3 VoL, Longmans Green (1927-29).
Lewis, "Spectroscopy in Science and Industry," Biackie, London (1933).
Lundegardh, "Die Quantitative Spektralanalyse der Elemente," 2 Vol., Fischer,
Jena (1929--1934).
Twyman and Allsopp, "Spectrophotometry," Hilger (1934).
Crook, "Metallurgical Spectrum Analysis," StanfoI:d Univ. Press (1934) .
.Spectrographic catalogs from Adam Hilger, 24 Eochester Place, Camden Road,
London, N.W. 1, England; The Bausch and Lomb Optical Company,
Rochester, N. Y.
INDEX TO AUTHORS *
A Ansbacher, S., 2520
Abady, J., 2391 Appelyard, G., 2404
Abbe, B., 17M, 17\)5 Archbutt, L., 1726, 1780, 1784
Abderhalden, E., 2503 Archer, R. S., 2558, 2564
Abegg, R., 2591 Aristotle, 574
Abraham, H. A., iii, v, 1510 Arkel, A. E. van, 1104
Abria, J. J. B., 1186 Arnold, C., 632
Acker, C. E., 961, 963 Arnold, F. W., 1216
Acree, S. F., 1919, 2285 Arnold, J. 0., 52
_<\.dams, R., 952, 1102 Arrhenius, S., 1191
Addie, R. H., 874 Association of Official and Agricultural
Adkins, L. R., 2546 Chemists, v
Adler, S., 2491 Astin, S., 1218
Adolph~ W. H., 409 Atack, F. Vi., 2, 311, 617, 618
Agricola, G., 399 Atkinson, A. J., 1642
4gulhon, H., 163, 185 Atwater, C. G., 1654, 1834
Ahlum, C. A., 2245 Aubert, A. B., 1789
Aisinman, S., 1737 Auerbach, F., 2310
Alexander, P., 1975 Austen, L., 435
Alexejeff, W., 2582 Austin, M., 560
Allart, A., 966 Autenrieth, W., 1945, 1950
Alldredge, S. M., 368, 503, 504, 527 Avogadro, A., 1191
Allen, A. H., 656, 1764, 1771, 1777, 1781,
,1787, 1876, 2147 B
t Allen, l!l. T., 179, 180, 905, 917 Bach, 0., 1781
Allen, r. C., 1717 Backstrom, S. A., 2318
Allen, N., 4Q6, 2198 Badger, W. L., 2369
Allen, W. S., v, 640, 909 Baggesgaard-Rasmussen, H., 2310
Alling, H. L., 2453 Bahlmann, C., 2081
Allison, V. C., 2361 Bahr, F., 2578
Allsopp, C. B., 2617 Bailar, J. C., 213, 2270
Alsberg, C. L., 226, 635 Bailey, A. J., 2526 .
Alten, F., 1217 Bailey, E. G., 1318
Aluminum Company of America, iii, 1 Bailey, H. S., 1780
American Brass Company, iv Bailey, R. K., 872
American Chemical Society, v Bainbridge, 1186
American Cyanamid COIJ1pany, iv Baker, H. A., 957, 964, 971
American Society for Testing Materials, Baldsiefen, W. D., 1780
iii, iv, v, 1011, 1071, 1421, 1485, 1597, Ball, T. R., 984
1620, 1705, 1713, 1719, 1727, 1732, Ballantyne, H., 1767, 1769
1737, 1743, 1750, 1761, 1807, 1814, Baly, E. C. C., 2617
, 1818, 1823, 1830, 1839, 1850 Bamberger, M., 1893
Amsbary, F. C., 2115 Banks, R. M., 97-
Amsel, H., 1784 Barbieri, G., 320
Anderson, R. P., 686, 1818, 2345, 2355, Bardet, J., 259, 2609, 2617
2405 Bardwell, E. S., 2453
Andrew, R. L., 460 ·Bardy, C., 2147
Andrews, L. W., 194, 269, 455 Barker, J. F., 242
Andrews, T. M., 1185, 1880 Barlow, H. C., xiii, 305
Angstrom, A., 2592 Barnard, J. E., 2457
Anneler, E., 2370 Barneby, O. L., 978
* Page numbers gretlter than 1300 are in Volume II.
Small Roman numerals in italics refer to the list of authors in Volume II.
1
2 INDEX TO AUTHORS
Cunningham, T. R., iv, v, xiii, 285, 296, Dieterich, Eo, 1769, 1770
·304, 563, 567, 568, 597, 610, 613, 706, Dietrich, E., 244
808, 983, 1005, 1011, 1016, 1018; 1053 Dietz, H., 454
Curtis, H. L., 1998 Dingley, W. F., 174
Di Nola, E., 432, 819
D Dinsmore, R. P., 2016, 2017
Dahlbergj A. 0o, 2287 Dische, Z., 2547
Dains, Fo Bo, 266 Ditmar, R., 1997
Dalton, J" 1191, 2S91 Ditt, M., 203
Daman, 1318 Dittmar, Po, 1760
Dammer, 0o, 744 Dittrich, Mo, 289, 980
Danziger, Jo Lo, 1761 Dixon, B. E., 148
Darroch, Jo, 957 Dixon, Ho B., 2426
Da Silva, see Silva Dobbins, J. T., 320, 882
Davies, H., 862, 1874 Dodge, B. Fo, 2342
Davies, J., 486 Dodson, R. Wo, 465
Davies, J. Lo, 374 Doherty, H. L., 2391
Davies, W., 862, 1874 Dole, Mo, 2303, 2313, 2318, 2319
Davis, 435 Dole, R. B., 2085, 2088
D<1vis, 1654, 1834 Donath, E., 958
Davis, A. L., 2154 Donau, J., 2457
Davis, O. C., 1805, 1996, 1998 Donnan, F. G., 2576, 2577
Davis, C. H., iv Dow Chemical Company, iv, 2181
Davis,,JJ. P., iv Dowd, J. E., 2106
Davis, G. E., 2340 Dowsett, 439
Davisson, B. So, 646, 647 Drawe, P., 2084
Davy, E., 2376 Drechsel, E., 195
Davy, H., 117, 163, 528, 899 Dressler, R. Go, 179
Day, T. G., 2515 Drew, H. D. K., 2541
Dean, E. W., 1340, 1342, 1345, 1717, 1841, Drew, S. K., 2369
1889 Drinkwater, 2142
Deeley, R. M., 1741 Droz, M. E., 2322
Deering, E. C., 981 Drnshel, W. Ao, 874, 936
De Gramont, A., 2609 Dnbosc, A., 1998
Delano, P. Ho, 2515 Dubsky, J. V., 1214
Del Rio, M., 1030 Duffendack, O. S., 2599, 2611, 2614, 2617
De Luce, R., 435 Dufty, L., 330
Dely, J. G., 687, 2201 Dulong, P. L., 1185, 1186, 1187, 1191
Demorest, D. Jo, 85, 374, 1384 Dumas, J. B. A., 634, 2358, 2474, 2485,
Demorest, N. Bo, 355 2491
Deniges, G., 118, 188, 193, 2501 Duncan, J. B., 74
Dennis, L. M., 430, 2344, 2345, 2359, Dunlap, F. G., 1771
2367, 2374, 2375, 2396,2423 Dunnington, F. P., 988, 992
Dennison, C. Ho, 1778 Dupre, A., 272
Dennsb3dt, M., 490, 2496 Dupre, F., 456
Derby, W. Go, 398, 441, 834 DuPont, G., 1190
Desains, P., 1189 DuPont, J., 1777 °
Kilpatrick, M. L., 990 Kolthoff, I. M., iv, v, vi, 8, 12, 143, 193,
Kimberly, A. E., 1916 273, 296, 310, 350, 351, 423, 456, 457,
Kimura, K., 1104 479, 529, 531, 535, 559, 571, 582, 640,
Kimura, W., 2498 644, 662, 819, 878, 879, 882, 1027,
King, E • .T., 803 1065, 1203, :].206, 1214, 1471, 1478,
King, .T. F., 481 1925, 2191;r2287, 2306, 2307, 2308,
, King, J. T., 838, 839 2310, 2311, 2312, 2314, 2323, 2328,
King, V. L., 7,468 2333, 2335, 2499
INDEX TO AUTHORS 11
• Kon, S., 2496, 2508 La Tourette, H., 2418
Konen, H. M., 2601 Laufis, A., 1782
Koninck, L. L. de, 532, 533 Laurie, L. G., 2459
Koppeschaar, W., 1950 Lauro, M. F., 633, 249,3
Korbuly, M., 2353 Lavoisier, A. L., 442, 629, 1192, 2365
Koref, F., 1188 Lea, C., 738
Koremnann, 1. M., 64 Leach, A. E., 1787, 2149, 2158
Kraemer, E. 0., 521 Leaver, E. S., 393
Kraemer, G., 2491 LeBlanc, M., 922
Krasnitz, ~, 634 LeChatelier, H., 1140, 2552
Kra.tz, G. D., 1973 Ledue, A., 1189, 2411, 2412
Krauskopf, F. C., 892 Lee, M. F., 8
Kraut, K., 246 Leech, W. D., iv, 460
Kraybill, H. R., 2472 Leeds and Northrup Co., iii
Kreis, H., 1782 Leidy, J., 2115
Kretzschmar, M. J., 7 Leighton, P. A., 2537
Krieger, K. A., 2035 Leitch, I., 460
Kriesel, F. W., 426 Lenher, V., v, xiii, 434, 438, 779, 781,
Kronacher, C., 2458 782, '783, 793
Kroupa, E., 2545 Lenssen, E., 964
Krueger, A. C., 2320 Leopoldi, G., 521, 820
Krass, G., 486 Lepper, W." 945
Krug, W. H., 3 Lerch, W., 1618
Krumholz, P., 1~14 Leroy, A., 2545
Kubelkova, A. 0., 2523 Levenson, H., 1699
KiVm, W. E., 2398 Levol, A., 96
Kuenen, J. P., 1189 Levy, L., 979 •
Kaster, W., 2530 Levy, L. H., 374, 1190
Kuhn, R., 2501, 2526 Lewis, E. J., 626
Kukolich, S. 1., .2290 Lewis, H. F., 2290
Kundert, A., 582 Lewis, S. J., 2617
Kuras, M., 203 Lewis, W. H., 2340 .
~rtenacker, A., 285 Lewkowitsch, .T., 1773, 1775, 1777, 1779,
Kurtz, F. E., 634 1787, 1789
Kyle, T. D., 157 Lichtin, J. J., 296
Liddell, D. M., 833, 2389
L Lidstone, F. M., 2546
Lachele, C. E., 108, 2505 Lieb, H., 2500
'Ladd, R. M., 655 . Lieben, A., 1948
• Lakshminarayan, S., 2354 Liebermann, C., 1774) 1781) 1979
Lamar, M. 0., iv, 5, 32, 181, 538, 813 Liebich, C., 2313
Lamb, A. B., 219 Liebig, J. von, 221, 277, 661, 1650
Lamb, F. W., 521 Lind, S. C., 759, 167
LaMer, V. If:-., 2202, 2310 Linder, J., 2468
Lammert, O. M., 2301 Lindgren, J. M., 622
Landholt, H., 2411, 2412 Lindo, D., 871
Landrum, R. W., 329, 330 Lindow, W. C., 2519
Lang, R.,' 455, 1210 Lindsey, A. J., 2515, 2547
Langbein, G., 824 Lindsley, L. C., 2459
Lange, E., 2313 Lindsly, H., 2508
Lange, N. A., 2601 Lingane, J. J., 882, 1027, 1471
Lange, W. E., 1986, 1987 Little, H. V. F., 246
Langley,IJ. W., 656 Littrow, M. de, 2594, 2595, 2596, 2597
Langmuir, A. C., 534 Lochte, H. L., 2534
Laplace, P. S., 1192 Lockemann, G., 108
Larison, E. L., 360 Lockhart, L. B., 1789
Larsen, E. S., 2435, 2451 Lockwood, E. H., 2384, 2386
Larson, W. D., 582 Lockyer, N., 1789
Lassieur, A., 958, 961, 2HS Lodemann, G., 2458
Lassone, J. F. M., 2370 Losch, J., 252
Lathrop, E. C., 1178 Liiwe, F., 2601
12 INDEX TO AUTHORS
Lowe, J., 154 McGill, A., 237
Logan, W. N., 2453 McGill, W. J., 2310 /
Long, C. P., iii, vi, 1340, 2026 McIlheny, P. G., i767, 1770, 1773, 1800,
L'Orange, J., 2480 2043
Lord, N. W., 525 McInerney, T. J., ~289
Lovejoy, D. R., 2379 MacInnes,. D. A., I ~300, 2313, 2318
Low, A. H., 65, 75, 91, 375, 402, 405, MacIntosh, J. B., 2353
509, 535, 580, 957, 1022, 1064, 1202, MacKay, W., 225 \
1856 McKenzie, A. E., 647
Low, G. W. Jr., 2198 MacKey, W. McD., 1782
Lowenthal, R., 21G8 McKinney, D. C., 2312
Lowrey, T. M., 311 McMaster, L., 2370
Lowry, H. H., 1998 MacMichael, R. F., 1724
Lubs, H. A., 20n McNabb, W. M., 77·, 696
Lucas, C. C., 803 McNeill, C. L., 379
Lucas, R., 2480, 2515 McNeill, T. R., 285, 296
Lucko"iy, C., 73 McPherson, A~ T., 1998
Lucker, F., 2499 McTrusty, J. ·W., 2401
Luff, B. D., 1998 Maczowski, E. E., 1614
Luiginin, W. F., 1190 Maes, P., 225
Lukens, H. S., 2035 . Mahin, E. H., 'v
Lummer, 0., 2391 Mahler, .P., 1187, 1654, 1834
Lundegardh, H., 2617 Mahr, C., .156, 157
Lundell, G. E. F., 5, 8, 14, 30, 405, 408, Maier, C, G., 396
531, 534, 538, 575, 576, 626, 955, 980, Makower, W., 767
981, 983, 984, 997, 1027, 1094, 1097, Malatesta, G., 432, 819
1100, 1102, 1216, 1421, 1922 Manning, R. J., 1893
Lunge, G., 244, 486, 935, 956, 967, 1144, March, F., 1777
1147, 1158, 1167, 2147, 2198, 2201, Marchand, R. F., 578, 1650
2222, 2268, 2378 Marchlewski, L., 244, 1144
Lungwitz, 1736, 1769 Marcusson, J., 1343, 1345, 1548
Luttringer, A., 1998 Margosches, B. M., 273, 454
Lux, F., 1736 Marignac, C., 332
Lyons, E., 487 Mariotte, E., 1650, 1652
Lyons, M. A., 2449 Marks, L. S., 2339, 2340, 2389
Lyson, L. T., 2571 Marks, M. E., 226
MarIe, E. R., 2483
M Marme, M., 1959, 1p61, 1965
Maag, O. L., 606, 613, 1481 Marr~an, G. ,F., 2qii7
Maag, W., 2530 Marnan, P.: M., 2547
Maas, A. R., vi, 1943 Marsh, J., 87, 109, 2367
Maas, H., 1215· Marsh, W. -T., 620
Maasen, G., 628 Martin, 222
McAdams, W. A., 2340 Martin, G., 11981, 2017
McBride, R. S., 1209, 2395 Marvin, G. G., 1095, 1481
McCabe, C. R., 994 Marx, P. F., 2387, 2388"-- _____ _
McCandless, J. M., 532, 696 Mascazzini, A., 73
McCarthy, E. S., 2396 Mason, C. W., iv, vi, 2434, 2435, 2428,
McCay, L. W., 71, 78, 95, 590, 862, 958 2439, 2440, ~447, 2453, 2454, 2457
McClendon, J. F., 460 Mason, H. M., 2450
McCollam, C. H., 606 Mason, S. N.~ 1774
McCoy, H. N., 255 Massol, G., 1189
McCroskey, C. R., 1481 Matheson, H., 1919
McCrumb, F. R., iv, vi, 2274, 2284, 2285, Mathews, J. H.,' 2459
2287 Mathewson, C. H., 2453
McDaniel, A. S., 2588, 2589 Mathy, E. L., 2378
McDonald, R. G., 1884, 1910, 1913, 1916 Mattern, J. D., 2264
McDonnell, C. C., 91 Maumene, E., 1701, 1766, 1.1.813
McDowell, J., 660 Maxymowicz, W..!..:_152 /
Mace, J., 2458 Mayer, J. L.; f'965 -
,McFarland, D. F., 2379 Mayow, J., 2368
INDEX TO AUTHORS 13
Mayr, c., 2498 Moraht, H., 486
Mayrhofer, A., 2435, 2458 Morley, E. W., 437, 1347, 2408
Mazzotto, D., 1188 Morgan, J. L., 1143, 1146, 1151, 2207,
Muzzuchelli, A., 753 2221, 2339
Meade, R. K., v, 374, 1384, 1619 Morgan, J. L. R., 2301
Mecklenberg, W., 1094 Morgan, S., 1998
Meggers, W. F., 2609 Morrell, J. C., 1764
Meiklejohn, C., 957 Morris, G. W., 2458
:lVIeiklejohn, R. M., 1161, 1219 Morse, H. N., 1223
Meineke, K., 1441 Moser, L., 152, 427, 445, 446, 896, 897,
Melaven, A. D., 47, 1458 943
Mellor, J. W., 148, 244, 246, 410, 721, Moskovitz, B., 644
729, 733, 956, 968, 1036, 1141 Muehlberg, W. F., 1095
Meloche, Y .. W., 2546 Muhlsteph, W., 108
Memmler, K., 1997, 1998 Muelken, Mrs. A. Z., v
Mendeleef, D., 429, 1192 Mueller, C. W., 2394
Mene, C., 966 Muller, E., 2307, 2313, 2314, 2315, 2318,
Mmmick, H., 276 2319, 2320
Menzies, A. W. C., 2530, 2532 Muller, F., 988
Merwin, H. E., 410, 413, 988 Muller, Fr., 2314, 2322
Merz, 0., 2546 Muller, R. H., 2322, 2472
Messinger, J., 2136 Muer, H. F., 2065
Metzel, A., 72 Muggenthaler, H., 1179
Metzger, F. J., 254, 2364 Muir, M. M., 157
Metzner, P., 2457 Muller, J. H., 429, 430
Meulen, H. Ter, 225, 226, 227, 635, 682, Mulligan, J. J., xiii, 159, 160
2509, 2510 Munroe, C. R, 1767
Meyer, A. H., 2501 Munson, L. S., 1768, 2039
Meyer, Y., 2423, 2534, 2575 Muraour, H., 1698
Michaelis, L., 2289 l"furgelescu, I. G., 2335
Middleton, A. R., 2 Murphy, G. M., 444 .
Mika, J., 2464 Murray, O. W., 108, 2367
Miller, A. L., 5 Murray, W. M. Jr., 596
Miller, C. 0., 00 Myers, F. B., 1028
Miller, E. J., 521 Mylius, F., 434, 753
Miller, L. F., 749, 751
Miller, S. P., 435 N
/
Milliau, E., 1766 National Lead Company, iii, iv, 995,
Milligan, L. H., 2507 1387 ff.
Millon, N. A. E., 1950, 1964 Nauss, A. 0., 2363
Mills, E. J., 1769 Nayak, U. M., 2534
Mills, P. L., 108 Neal, J. L., 56
Milner, H. W., 2530 Neller, J. R., 108
Milner, R. T., 2465, 2489, 2490, 2491 N elsmann, H., 966
Miner, H. S., 948, 950 Nemst, W., 2305, 2306
Mitchell, C. A., 1767 Nessler, J., 630, 656, 657, 658, 1974,
Mitscherlich, A., 646 2019
Moberg, E. G., 2547 Neubauer, H., 2469, 2499
Moeller, J., 2458 Neuberg, C., 1776
Mohler, E., 1185, 1186 Neuman, C., 1182
Mohr, F., 75, 76, 238, 271, 272, 456, 1541 Neumann, B., 2565
Moir; .T., 439 Neumann, E. VV., 905, 908
Moisaan, H., 399 New, C. H., 2358
Molisch, H., 2459 Newhall, C. A., 934
Mollow, G., 823 New Jersey Zinc Co. (of Pa.) iii, 1971,
Monfort, W. F., 2111, 2115 1202
Montequi, R., 156 Newton, H. D., 984
Moody, L. T., 2571 Nichols, M. S., 461
Moore, G. T., 2111, 2112, 2114 Nicklay, A. J., iv, 527
Moore, H., 2552 Nicloux, M., 1948
Moore, R. L., 1987 Nicolardot, P., 1102
14 INDEX TO AUTHORS
Niederl, .T. B., 2322, 2479, 2481, 2485, Papish, .T., 429, 430, 894, 896
2488, 2534 Paracelsus, 149, 1054
Nierenstein, M., 1893 Park, B., 372; 626 '
Nikolow, C., 320 Park, W. F., 2132, 2133
Nissensohn, H., 74 Park, W. H., 2133, \
Nitchie, C. C., 1089 Parker, J. C., 2164, 2169
Nola, see Di Nola, E. Parker, W. L., 2361
Noponen, G. E., 640, 1214 Parr, S. W., v, 622, 1643, 1654, 1834,
Nordlander, B. W., 2423 2508
Norris, J. F., 785 Parrodi, G., 71 "
Norris, W. V., 410 Parry, E. J., 1873, 1981
Norton Company, iv Parsons, C' L., 140, 148 .
Norton, J., 1786 Parsons, J. L., 1913
Norton, R. H., 229, 230, 231 Parsons, T.. R., 2289, 2310
Noyes, A. A., 208, 429, 430, 434, 465, Partridge, H. M., 2322
893, 979, 2300, 2572 Pass, A., 1217
Nussberger, 2322 Patat, F., 2518
Nusselt, W., 2373 Patterson, W. L., 2549
Nydegger, 0., 921, 2064 Pattinson, H. S., 95
Pattinson, J., 95, 560
o Paul, .T., 72
Obach, E., 1993, 1994 , Payne, H. ,L., 2425, 2426
Oberfell, G. G., 2368, 2377, 2403 Peachey, S. J., 1977
O'Brien, W. J., 2367 Pearce, :EJ. V., 100
Ode, W. H., 803 Pearce, S. J., 2547
Oersted, H. C., 1 Pearson, E. A., 1925
Oesper, R. E., 455, 1205, 1206 PearsOli, H. C., 1998
Offerman, H., 409 . Pease; V. A., 2450
Ogburn, S. C., 738, 749, 751 Pechard, E., 1004
Ohm, G. S., 1192 Peligot, M., 1021
Oilar, R. D., 1778, 1779 Pemberton, H. V., 2039
Okac, A., 1214, 2516 Penfield, S. L., 409, 1,340
Okuma, K., 529 Perkin, F. ~., 436 •
O'Leary, W. J., 894, 896 Perley, G. A., 23(l2
Olsen, J. C., 401, 1136 Perrot, F. L., 2411
Olson, H. W., 180 Perrot, G. St. J., 2454, 2456
Opfermann, E., 1884 Person, G. G., 1188, 1189, 1190
Orsatt, M. H., 2342, 2345, 2349, 2385, Peters, 1654, 1834/
2388 Peters, F. P., 627
Orton, E., 2398 Peters, R., 2306
Osborn, R. A., 634, 2493 Petit, T. A., '1191
Ostwald, Wm., 2307, 2587, 2588 Petkow, N., 1778
Ostwald, W 0., 1977 Petterson, 0:, 2405
Otto, C., 2 Petterssoli, 0., 1188
Oudemans, A. C., 1156, 2252 Peyau, H., 1778 -....:
Oulman, 2254 Pfiitzer, G., 120 --.....
Outcault, H. S., 2365 PfUlldt, 0., 2327, 2328, 2335
Outerbridge, A. E., 1701 Phelps, E. B., 2081
Owen, F., 2266 Phillips, F. C., 2407
Phillips, M., .;l530
p Physick, P. C., 2366
Pacific Coast Borax Co. iv, 174 Pickering, P. S. V., 1190
Page, R. T., 2343 Pickering, S. W j 11g'1, 2239
Pagel, H. A., 249 Pickles, S. S., 2u12
Pagenstecher, A., 1946 Pied, H., 340, 343
Palkin, S., 890 Pierce, A. W., 785 I
Pallet, K., 966 Pierle, C. A., 1005, 1021
Palmer, C. S., 952, 1102 Pierson, R. H., v, 1804 I
Palmer, I. A., v, xiii, 835, 2025 Pinkus, A., 96I_,':" ~
Palmquist, A., 2405 Pirsch, .T., 25-34, 2469
.
J>aneth, F., 767 Pisani, F., 832
INDEX TO AUTHORS 15
Pitkin, Lucius Co., v Raschig, F., 647, 920, 926
Pitman, J. R., 650 Rast, K., 2533
Pitot, 1173, 2340 Rathgen, F., 3
Plato, W., 69, 1189, 1190 Rathsburg, H., 78
Plechner, W. W., 989 Rauchswerger, D., 1776
Plent1, A., 2534 Rawlins, n. J. B., 966
Plimmer, R. n. A., 2501 Rawson, 0., 2158
Pliny, 954, 2375 Ray, P. R~ 1217, 1218 •
Polenske, E., 1786 Rayleigh, Lord (J. Strutt) 2380, 2411
Pollak, J., 2521 Rebiere, G., 832
Pollak, L., 582 Redmond, J. C., 1607
Pollitzer, F., 1188 Redwood, B., 1741, 2401
Polushkin, E. P., 2453 Reed, R. M., 1345
Polyakov, A., 116 Reedy, J. H., 892
Ponomarev, V. D., 108 Regester, S. n., 1642, 1656·
Pope, G. S., 1318 Reglade, A., 1102
Popoff, S., v, 905, 908 Regnault, H. V., 1160, 1188, 1189
Popp, G., 166 Reich, F., 2399, 2400, 2401
Porter, L. W., 1903 Reicher, L. T., 2575
Poschl, V., 2459 Reid, E. E., 1986
Poth, .E., 2489 Reif, W., 943, 1218
Powell, A. R., 333, 338, 339, 346, 445, Reiff, O. M., 990
773, 942, 1004, 1099 Reinhardt, C., 299, 476, 482
Pozzi-Escot, E., 433 Reiss, M. A., von, 73
Prandtl, W., 252, 1104 Reissner, R., 1218, 2521, 2547
Pratt, L. A., 252 Reith, J. F., 460
Prausnitz, P. n., 2389 Remington, R. E., 460
Prenle, W.O., 436 ' Renard, A., 1780
Pregl, F., 2481, 2487, 2490, 2494, 2510, Rethberg, P. B., 2498
2511, 2516, 2517, 2518, 2526, 2528, Rey, n., 1147, 2201, 2222
2547 Rhodes, F. n., 2364 .
Prescott, A. B., 262, 743 Rhodin, J. G. A., 50
Prescott, S. 0., 2133 Rice, A. C., 952, 1102, 1216
Preston, J. M., 2459 Rice, F. 0., 990
Price, T. S., 360 Rich, C. E., 2493
Price, W. B., 355, 374, 1384 Richards, J. W., 1188, 1783, 2388, 2389
Priestly, J., 2360, 2362, 2364, 2368, Richards, T. W., 360, 831, 832 2337
" 2373, 2378 Richardson, 0., 1619
Prince, N. F., 2399 Riche, A., 2147
Pringsheim, P., 267 Richmond, H. D., 165, 1767, 1801
Prister, "A., 437 Rieche, A., 2533
Procter and Gamble 00., iii Riesenfeld, E. n., 120, 2591
Proctor, H. R., 2158 Rigg, W. L., xiii, 305
Pulman, O. S., 593, 1027 Riley, H. L., 1218
~ Q Ringelmann, 2388
Rinkenbach, W. R., 1660, 1686
Quill, L., 978 Risdale, N. D., 997 .
R Ritschel, E., 896, 897
Rab.inovich, A. I., 2318 Rittenberg, G. S. C., 372
Rachele, J. R., 2546 Roark, R. 0., 91
Rafter, G. W., 2102, 2103, 2104 Robbs, O. E., 157
Raikow, P. N., 1778 Roberts, E. J., 252
Raiziss~ G. W., 2479 Roberts, H. S., 2320
Rajmann, E., 1214 Roberts, L. D., iii, xiv, 682, 758
Rammelsberg, 0., 889 Roberts, R. T., 392
Ramsay, W., 683, 1190, 2380 Robertson, D. W., 1864
Randall, D. L., 594 Robertson, 1. W., 2402
Randles, W. B., 2457 Robertson, P. W., 1188, 1190
Rapoport, S., 129, 2518 Robinson, R. A., 2310
Rasch, R. S., 1913, 1914, 1916 Robinson, R. J., 2526
16 INDEX '1'0 AUTHORS
Robitschek, J., 531 Sauveur, A., 2457, ·2552,2553, 2554, 2558,
Robson, W. G., 1189 2563,2564 _
Roe, R.. B., 1880 Savell, W. h, 6281
Rohre, K., 69 Saybolt, G. M., 1~18, 1717, 17,19
Rossler, L., 432 Sayers, R. R., 1963, 2364, 2406
Roger, R., 225 Schaeffer, ,J. A., vi, 525, 527, 1839, 1870,
Rogers, A., 1789 1874, 1876 I
Hogers, L. A., 1343, 1345 Scheele, C. W., 2611 554, 689, 1001, 2362,
Hogers, S. M., 2546 2363, 2364, 2365, 2368
Roscoe, H. E., 730, 742, 744, 748, 1036 Scheibe, G., 2601, 2617
Rose, H., 122, 156, 314, 403, 1036 Scheibler, C., 244, 2366
Hose, T. K., 436, 440 Scherrer, J. A., 70
Roseman, R., 479, 989 Schicht, H., 1774
Rosengarten, A. G., 1143, 1146, 1151, Schidrowitz, P., 1996, 1998
2207, 2221, 2239 Schiff, H., 1198, 2148
Rosimthaler, L., 2459 Schiodt, E., 2287
Ross, E. M., 2092 Schloklitsch, K., 2545
Ross" J. F., 872, 884 Schmidt, G., 1973
Ross, J. H., 1913 Schmitz, W., 82
Ross, W. H., 824 Schneider, A., 2458
Rossem, A. van, 1998 Schneider, F., 2445, 2458
Roth, H., 918, 2526, 2547 Schneidewind, R., 917
Roth, R. T., 2479, 2481 Schoeller, W. R., iii, v, xiv, 331, 332, 33:
Roth, W. A., 1189, 2318 334, 338, 339, 340, 34-1, 342, 343, 34'
Rothaug, C., 286 345, 3;16, 348, 426, 445, 773, 793, 94i
Rothe, J., 1983 981, 1004, 1099
Rothe, J. W., 312,317 Schonbein, C. F., 1946, 2369, 2405
Rother, E., 2335 Scho}l, C. Eo, 1026
Roux, L., 1185 School, A •. T., 210 ,
Rowe, F. M., 529 Schoonover, I. C., 820, 882, 1027, 1214
Rowell, H. W., 74 Schopper, L., 1995
Rudberg, F., 1188 Schor lemmer, C., 730, 742, 744, 748
Ruder, W. E., 811 Schotz, S. P., 1998
Rudorff, Fr., 1156, 2397 Schou, S. A., 2310 -
Rudy, R. B., 1615 Schreiber, H., 1737
Ruggeri, R., 1778 Schreinemakers, 2581
Ruigh, W. L., 529 Schrenk, H. H., 103, 2547
Rupert, F. F., 2179 Schrenk, W. T., 80~, 2515
Russ, F., 346 Schroeder, W. C., 129, 919
Russell, W., 225, 226 Schroetter, A. R. :Von, 238, 1541
Russell, W. J., 2123, 2132 Schuchardt, T., 2370
Rutherford, D., 2378 . Schiitzenberger, R., ~76
Rutherford, E., 767 Schulek, E., 180, 662
Schultz, G., p29 '"
S Schulze, F., 50, 2310
Saint-Rat, L. de, 1245 Schum, E., 3, 53 __
Saint-Sernin, A., 216 Schumb, W. C., 1095, 1101, 14:81-
Salas, L. E., iv, vi, 1966 Schupp, O. E., 481
Sand, H. J. S., 71 SchwackhOfer, 1185
Sandell, E. B., 8, 296, 310, 531, 559, 571, Schwaibold, J. von, 460
644, 1471, 1478 Schwalbe, C. 'G., 1891, 1914
Sander, W., 2590 Sc~wappach, A., 2527
Sanders, J. McC., 931, 932 Schwarz, E. R., ~459
Sanders, M. T., 1677 Schwartz, 1948 ,
Sanders, R. B., 2493 Schwartz, E., 2313
Sandin, R. B., 2493 Schwartz, R., 976
Sandstedt, R. M., 2493 Schweitzer, E., 2617
Sargent, G. W., 2390, 2391, Schweitzer, H., 1736, 17691
Sarker, P. B., 2509 Scott, A. H., 199,8__ I
Sarrau, E., 1185 Scott, A. W.,.121~ _
Sarver, L. A., 249, 319, 474, 628, 1206 Scott, J., 2458 -
,
INDEX TO AUTHORS 17
Scott, W. W., iii, iv, v, vi, viii, ix, xiv, vi, Skinkle, J. H., 2459
16,97,180,185, 233,238,368,413,478, Skinner, L. B., 11, 482
488, 503, 821, 824, 827, 913, 970, 1024; Skinner, W. W., 194, 459, 891
1421, 2020, 2180, 2191, 2203, 2214, Skowronski, S., v, xiv, 383, 384, 386, 398,
2414, 2415 727
Scribner, B. W., 1914 Skraup, Z. H., 531
Seamon, W. H., 580, 917 SIegel, H., 2546
Sears, G. W., 978 Slossen, 1185
Sedgwick, 2102, 2103, 2104 Slotta, K. H., 2529
Seeker, A. F., 517 Slottman, G. Y., 2378
Seeman, L., 432 Slyke, D. D. van, 2280
Seevers, M. H., 2537, 2538, 2543 Smith, A. W., 1189
Seibert, F. M., 2402, 2407 Smith, E. A., 437, 440, 831
Seidell, A., vi, 2570 Smith, E. F., 202, 720
Selenyi, P., 953 Smith, E. H., 371
Selfstrom, N. G., 1030 Smith, D. M., 2617
Sellars, W. S., 966, 967, 968 Smith, D. P., .779
Seltz, H., 2312 Smith, F. L., 2nd, 521
Selvig, W. f>._., 1628, '1629, 1633, 1635, Smith, G. Frederick, v, xiv, 861, 872, 884,
1642, 1643 1204, 1205, 1335
S'elwoQd, P. W., 255, 259 Smith, G. McPhail, 984
Sendroy, J., 2287 Smith, Gilbert B. L., 434
Sharp, R. S., 1984 Smith, Gilbert, M., 2115
Sharp, T. F., 2289 Smith, J. H., 2224, 2225
Sharwood, W. J., 578, 583 Smith, J. H. C., 2530
Shattuck, G. A., 2343 Smith, J. H. F., 203
Shaw, E. H., 1986 Smith, J. Lawrence, 6, 36, 344, 534, 799,
Shaw, H. G., 1179 881, 882, 1612
Shaw, T., 2401 Smith, J. M., 7
I Sheen, R. T., 919, 2092 Smith, M. M., 1096, 1101
Shelberg, E. F., 67 Smith, W. H., 2174
Shelton, W., q8 Smolczyk, E., 2310
Shennan, R. J., 203 Smoot, A. M., 857, 1046
Shepherd, M., 2345 Snell, C. T., 989
Sheppard, J. R., iv, xiv, 527 Snell, F. D., 989
Sheppard, S. E., 2364 Snelling, W.O., 1660
Sherer, .T., 1945 Snodgrass, J., 1769
Sherman, H. C., 1767, 1770 Sobotka, H., 2493
Sherman, M. S., 2465, 2489, 2490 Soddy, F., 767
'Shields, J.; 2383, 2384 Sorensen, S. P. L., 2276
Shimer, E. B., 985 Soloveva, N. A., 634
Shimer, G., 231 Soltsien, P., 1778, 1779
Shimer, P. W., 985 Soltys, A., 2526, 2547 ,
Short, M. N., 2458 Solvay Process Company, iii, 240, 2257
Shrewsbury, C. L., 2472' Someya, K., 255
Shukoff, J. J., 2318 Sonden, K., 2405
Sickman, D. Y., 879 Sottery, C. T., 2, 1213
Siedler, P., 74 Soxlet, F., 196(l, 2029
Siegmann, F., 445, 446 Spacu, G., l-57, 203, 320, 582
Silberberg, B. H., 2450 Speakman, T., 2366
Silber mann, J. T., 1185, 1186 Spear, E. B., 465
Silberstein, J., 958, 968 Spence, D., 1973, 1974, 1975, 1997
Silbert, F. C., 2479 Spies, J. R., 2491
Silvn, A. J. F. da, 1782 Spikes~ W. F., 2457
Silverman, L., 1481 Spitz, G., 1773
Simmance, J. F., 2391 Spring, W., 1188
Simms, H. S., 1768, 1773 Squibb, 2142
Simpson, E. S., 331, 345 Stabler, H., 2085, 2098
. Simpson, M. E., 647 Stabler, 1i. J., v
Simpson, S. G., 1095, 1101· Stahler, A" 2470
Singleton, W., 953 Staniewski, M., 1188
18 INDEX TO AUTHORS
Stansby, M. E., 423 Swift, E. H., 427, 455, 465, 980
Stanton, F. M., 1629, 1633 Szebelleay, 'L., 474 .1
Stapleton, P. L., 811 Szendrii, P., 2504
Starck, G., 405
Stark, D. D., 1340, 1341, 1345, 1841, 1889 T
Starke, K., 634 Tabern, D. L., 67 '\
State, H. M., 582, 697, 1205 Taufel, K., 634 L
Stead, J. E., 91 Talbot, H. P., 422, Lt50, 1151, 2238, 223&
Steele, F. A., 1925 Tananaev, N. A., 108
Steiger, B., 1218 Tang, N. G., 11
Steiger, G., 410 Taylor, B. S., v
Steinbeck, D., 374 Taylor, C. A., 1660, 1686
Stelling, E., 963, 970 Taylor, G. B., 1671
Stenger, V. A., 193 Taylor, L. V., 2493
Stephanov, A., 266 Taylor, W. A;, iv, vi, 1919, 2274, 2291
Sterba·Biihm, J., 588 Taylor, W. C., 2166
. Sterling, C., iv, xiv, 619, 625, 628 Technical Association of Pulp and Paper
Sterling, W. F., 1340, 1342, 1345 Industries, v, 1?78
Stetser, J. B., 221, 229, 230, 231 Tennant, J., 2493
Stevens, H. P., 1973, 1983, 1985, 1998 Ter Meulen, see Meulen
Stevens, W. H., 1968 Thaler, H., 634
Stief, F. A., 114, 1384 Thenard, ,L., 163
Stimson, F. J., 2609 Thiollet, R., 2017
Stoddard, J. T., 360 Thomas, -T. S., ,2394
Stoddard, W. B., 51 Thompson, A., 486
Stohman, F. C. A., 1185, 1186 Thompson, C. L., 2453
Stokes, H. N., 2353 Thompson, G. W., 85, 957, 1381, 1860
Stone, C. H., 2395, 2398, 2399 Thompson, H. W., 2025 •
Stoppel, A. F., 591 Thompson, J. J., 1470
Storch, L., 1774, 1781, 1979 Thompson, L., 2367
Storm, C. G., iv, vi, 1660, 1664, 1671, Thompson, R. T., 165
1675,,1677, 1693 Thomsen, J., 1167, 1185, 1186, 1190
Stormont, R. T., 2537, 2538, 2543 Thomson, R., 1767
Stortenbeker, W., 1190 'Thornton, N. C., 2366
Stover, N. M., 1215, 2494 Thornton, W. M., Jr., iv, v, 479, 481, 707,
Strafford, N., 2017 975, 980, 981, 982, 983, 989, 992, 996
Straub, F. G., 2089, 2090 Thorp, G., v, 494
Straus, J. V., 1053 Thorpe, T ..E., 691, /17'3, 781, 2205
Strebinger, R., 682, 2521, 2548 Thresh, T. C., 159 I
Stromeyer, F., 122, 197 Thurston, R. V., 395
Strutt, R. J., 2379 Tickell, F. G., 2093
Stuart, A., 2435 Tieftrunk, F:, 2399
Stull, A., 2493 Tiemann, H. 'P., .2566
Stull, D. R., 2547 Titanium Alloy Manufacturing Company,
Sucharda, E., 2533 991
Suggs, 2391 Titus, L., 2546
Suitsu, K., 529 Titus, R., 2435
Sullivan, E. C., 2166 Tobler, F., 2459
Sullivan, T. J., 2205 Tolloczko, S., 1189
Sullivan, V. R., 1204 Tomicek, 0., _?65, 479, 628, 2313
Sun, C. H., 225 Topf, G., 1870
Sunde, C. J., 273, 2499 Topf, W., 524
Sutermeister, E., 1902, 1903, 2459 Tortelli, M., 1766; 1778
Sutton, F. A., 272, 374, 437, 456, 918, Tougarinoff, B., 1215, 1218
926, 2187 Touzalin, L. A., ,2389
Swanger, W. H., 754 . Town, G. G., 2081
Sward, G. G., 1839 Trautz, O. R., 2485, 2488, 2534
Swearingen, J. S., 2537, 2538, 2543 Treadwell, F. P., 307, 403/454, 6,,5, 639,
Sweeney, O. R., 2365 889, 900, 908, 919,921, 926, 1021', 1036,
Sweetham, 1737 2353, 2370 .,:.J __
INDEX TO AUTHORS 19
Accn. ~
NO.S..Z ..
_.-t.-
Call ....fJ!2
__
..........~.... .
Tile book should be, returned on 0; oewre-
the last date stamped below. Otherwise fines a~
per the library rules will be charged.