P. 1
Fractional distillation (1903)

Fractional distillation (1903)

|Views: 47|Likes:
Published by Mahmudul Islam-topu
About Fractional distillation
About Fractional distillation

More info:

Published by: Mahmudul Islam-topu on Feb 03, 2013
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less






  • FIQ. 8. Condenser for volatile
  • Vapour pressures of mixtures of alcohols with water
  • apparatus
  • x "PEAR" STILL-HEAD 165
  • Coupler's still



Dr. the process most cases. in great majority of these substances were liquids.PREFACE DURING the past eighteen years I have been engaged in investigations which necessitated the preparation of chemical materials in the purest possible state. for reading the much valuable assistance in compiling the . and that the value and economy of highly efficient still-heads in laboratory work may come to be more widely recognised than is generally the case at present. Brown taken J. and as the of fractional distillation had. E. from which Figures 2 and 34 have been My sincere thanks are due to Professor for the loan of valuable ancient . Campbell works by Libavius and Ulstadius. and colleague. hope that the solution of the which so often occur in carrying out a fractional It is in the difficulties distillation may be rendered easier. F. that this book has been written. to my for proofs. was thus enabled to devise some new methods and forms of apparatus which have been described from time to time in various scientific journals. The difficulties I met with in some of the separations led me and to I make a careful investigation of the whole subject. Francis. to be resorted to for their purification.

S. Lehfeldt. Anyn." Payen's I have made much use of the experimental data of Brown. index and to Professor R. made due acknowledgment.it 1903. Konowaloff. A. . Simmons. Gregory and for Mr. Y. Zawidski and other observers. proofs. as far as possible. Several fractional distillations and numerous experiments have been carried out while the book was being written and the results have in many cases not been published elsewhere.. many useful the arrangement of the MS. but in some of the tables this has not been practicable. and for suggestions regarding the perusal of the in in In the description and illustration of the stills employed commerce I have derived much assistance from articles Thorpe's "Dictionary of Applied Chemistry" and " Precis de Chimie Industrielle. T. BRISTOL. and have.vi PREFACE . A. B.Sc.



18 19 19 (Bredt's) (Brtthl's) 12. 6.LIST FIG. 13. . 4. 17.. maximum vapour pressure 56 Boiling points of mixtures of chlorobenzene and bromo- benzene 18. heated by steam volatile liquids Condenser for 9. 8. 6 61 apparatus 73 . 5. Brown's of liquid and vapour phases. . bad and good forms (a and b). 1. 10. 14. OF ILLUSTRATIONS PAGE 3 3 employed by the Alexandrian Chemists Water condenser described by Libavius (1611) Still 3. 46 47 15. 20 22 Vapour pressure curve Vapour pressures of mixtures Mixtures of Aniline Still of alcohols with water . Composition from straightness (methyl and propyl acetates) . 2. 11.. 16.. 57 58 Apparatus for determining the boiling points of mixtures Deviation of boiling point-molecular composition curve 19. Guard for Bunsen burner Still 4 5 6 8 8 17 7.. 20. Ordinary modern apparatus for distillation Another ordinary form of apparatus Inner tube of condenser. Distillation under reduced pressure Modified receiver for above (Thome's) .

23. liquid and vapour phases. Ulstadius' " still 35. 41. . Linnemann dephlegmator Glinsky 172 . Gahl's apparatus 24. apparatus Composition of liquid and vapour phases. .. modified . 137 and 138 139 Separation of pentanes from petroleum Relation between weight of liquid distilled and composition tillation of distillate Two components. A (ethyl alcohol and benzene) 26. 30. 103 molecular fractional Partial and of total pressures against amount 27.. 37. . original . ( . Total weight of distillate ethyl and propyl acetates. 122 and Fractional distillation of mixture of methyl. apparatus Composition of Lehfeldt's 21. A (ethylene dichloride and benzene) . LIST OF ILLUSTRATIONS PAGK Composition of liquid and vapour phases. plotted against temperature 29. 81 liquid Composition of apparatus Variation of c and vapour phases. Later distillations 146 33 (a to/). 40. 166 166 168 171 38. form (one section) . 109 Fractional distillation of mixture of benzene and toluene. (a . 39. ) . "Pear" "Evaporator" still-heads. First dis- 145 31 and 32. between weight of liquid distilled and composition of distillate Three components Relation . .x FIG. Total weight of distillate plotted against temperature 28. Carveth's 82 with molecular percentage of 25. . b). 147 154 163 165 Modified still-heads 6).. . Relation between weight of liquid distilled and composition of distillate Two components. (a and " Rod aad Disc still-head still-heads 36. 34. 75 Zawidski's 77 22.

. Savalle's CofFey's Section of Pistorius still-head (modified) 201 201 58. 183 distillation . Section of dephlegmator Egrot's . Mansfield's still Coupler's still still 198 50. Specific gravities of distillates from mixtures of ethyl 226 . French column 199 52.. 6. 57.. 51. which 240 contains only 66. 56. 49.LIST OF ILLUSTRATIONS FIG. 45. . 228 229 63. 229 Carbon disulphide and methylal. The " Coffey " still 202 59. Carveth's apparatus 187 Lord Rayleigh's apparatus.. head 46. Young and Thomas dephlegmator " Brown's " regulated temperature still-head Combined dephlegmator and regulated temperature still- PAGE 172 173 181 44. Le Bel-Henninger dephlegmator (. xi 42. Logarithms of relative weights 64. Vapour pressure-molecular composition curve disulphide and methylal Distillation of carbon 227 61. 47. Continuous . 55. c). Composition of liquid and vapour Carbon disulphide and methylal. 200 200 201 201 53 and 54. (a Section of dephlegmator Dubrunfaut's and 6). Relative weights in liquid and vapour Carbon disulphide and methylal. aqueous hydrochloric acid. modicomponents 189 197 fied so as to separate three 48. much less volatile impurities illustrating the distillation of a substance which 243 contains only much more volatile impurities . 62. alcohol and water 60. Relation between c (Brown) and molecular percentage of carbon disulphide 230 Curve Curve illustrating the distillation of a substance 65.. 43. .

Curve Curve Curve illustrating the distillation of a substance which 243 244 68. 248 the hexanes . 69. contains only slightly more volatile impurities illustrating the distillation of a substance which contains both purities 70..xii LIST OF ILLUSTRATIONS PAGE FIG. Fractional distillation of petroleum 257 ... much more and much less volatile im- 245 Curve illustrating the distillation of a substance which 245 contains both slightly impurities 71. more and slightly less volatile .. 67. . Boiling point-molecular composition curve for ethyl alcohol and water 72.. contains bnly slightly less volatile impurities illustrating the distillation of a substance which .. . .

and in many cases the constituents of a mixture of two or more volatile liquids may be separated though frequently at much cost of time and material by means of the simple difficulty in obtaining in a For the fractional apparatus described in this chapter. but. The determinaby ordinary analytical methods. If only one component of a it from a non-volatile substance more frequently. not seldom. The object of distillation is the separation of a volatile liquid different volatility. the separation of two or more liquids of mixture is volatile.FRACTIONAL DISTILLATION CHAPTEE INTBODUCTION I Object Of Distillation. there is no pure state by distillation. in many cases. with considerable accuracy from the results of a single distillation. of the relative quantities of two or more organic compounds in a mixture tion. the composition of the mixture may be ascertained approximately and. or. the improved still-heads described in Chapters X to XII must be employed. distillation of such complex mixtures as petroleum or fusel oil. if a very efficient still-head be B is . Quantitative Analysis by Distillation. often a matter of great difficulty.

less frequently. may be impossible owing to the formation of a mixture of minimum or. but also in interpreting the results obtained. whether it will be possible. indeed. indeed. the boiling which are very close together (6) to the presence more components in relatively very small quantity . of the extent to which the curves would deviate from the normal form. to the formation of mixtures of constant boiling point. such as ethyl alcohol and water. or the vapour pressures. . and therefore to predict the behaviour of a mixture on distillation. in a moderate number of cases. and unless something is known of the form of the curve representing one or other of these relations. In any other case. points of of one or (c) .2 FRACTIONAL DISTILLATION It is CHAP. The subject of fractional distillation is full of interest owing to the fact that difficulties so frequently occur. we must consider the relation between the boiling points. In the common causes : distillation of petroleum. such difficulties are of occurrence and are due to one or other of three (a) to the presence of two substances. from the of these curves The form chemical constitution of the substances. not only in the experimental work. of maximum It is. only in the case of boiling point. substances which are chemically closely related to each other that the statement can be definitely made that the difficulty of separating the components of a mixture diminishes as the difference between their boiling points increases. or benzene and isobutyl alcohol. of con- employed. it is impossible to predict whether the separation will be an easy one or. to form an estimate. hoped that this method may prove siderable value. of mixtures of the substances and their composition. Difficulties Encountered. The separation of two liquids which boil at temperatures even 20 or 30 apart. and it is now possible. relationship of the depends largely on the chemical components.

The oil condenses upper in the wool and can be squeezed out of it. for Aristotle (1)* mentions that pure water may be obtained from sea. very attained. but a great amount of both time and material Fractional distillation is and there may when be saved by the use of a very efficient still-head and the object of the distillation is to ascertain the com. A primitive method of condensation is described by Dioscorides and by Pliny. Ancient Apparatus. B 2 . condenser. who state that an Fio. The Alexandrian chemists added a second vessel.water by evaporation. 2. much greater accuracy is thereby APPABATUS. Ancient still with water head. to the boiler oil may be part of * The numbers in brackets refer to the bibliography at the end of the chapters. 1. in the which is placed some wool. Alexandrian still with Fio. but he does not explain how the condensation of the vapour can be effected. the head or cover. obtained by heating rosin in a vessel. called by the Arabians the alembic. The process of distillation is evidently a very ancient one. position of a mixture. or alembic.ANCIENT APPARATUS is frequently a very tedious process necessarily considerable loss of material by evaporation and by repeated transference from the receivers to the still.

at the thermometer addition. deprived of heat . in FIG. in which the vapour is by a current of cold water the receiver. the head. or which act chemically on cork. The flask or still is fitted with a cork through which passes the still-head. is or and a simple form of apparatus used by them shown in Fig. 1611. similar but. E. D the thermometer. and we may still- The still. through a second cork in the condensing which boil at a high temperature. Fig. Later on. with Liebig's condenser. is taken from Libavius. 1. the Liebig's condenser. a In Fig. 3. was cooled by passing it through a The diagram. Ordinary still. is : the ordinary form of apparatus distinguish the following parts shown. Syntagma Alchymice Arcanorum. CHAP. B . The still is usually heated by means of a Bunsen burner. present time is The apparatus employed in principle. .4 FRACTIONAL DISTILLATION still. the side tube vessel containing water. A . and the side delivery tube from the still-head passes tube. Modern Apparatus. 2. it is more convenient to have the still and still-head in one piece and to elongate the For liquids . c. 3 is used to register the temperature.

otherwise a troublesome and (p. It should not be larger than the is necessary for liquid to amount of distilled. through the Liebig's condenser (Fig. if HEAD necessary. The Still. is often too short. 3. the condensing tube . be too low (p.STILL delivery tube so that it AND may 4). The Condenser. somewhat uncertain correction must be applied the distillation is 12). be long enough for the thermometer to be placed in such a position that not only the mercury in the bulb but also that in the stem is heated by the It should. If a glass flask is used it should be globular. 4. 28). a little air may be carried back from time to time as far as the thermometer bulb and the temperature registered by the thermometer will then fluctuate and will. The still-head. if . possible. because a flat-bottomed flask is liable to crack when heated with a naked flame. as shown if in Fig. Modified form of still with condenser. vapour of the boiling liquid and. on the whole. be The Still-head. be distilled is If the boiling point of the liquid to higher than about 170. plan to seal a short length of wider tubing to the still- head near the bottom. or the thermometer may be cooled slightly below the temperature of the vapour. as supplied by dealers. leaving a sufficient length of the narrower tubing below to pass through the cork in the still. The still -head should not be very narrow. not proceeding quite steadily. STILL pass. It is a good Fio.

5.. . . so may sloping gently downwards. narrow condensing tube for the insertion of the delivery tube from the still-head. It is usual to seal a short wide tube to the long. (b) correct. Wire gauze. If the outer tube is unwieldy. 5 a). should not be cooled by running water for fear of fracture.-. much wider it mm. construction. when filled with water. and. For laboratory purposes an ordinary Bunsen burner is usually employed. sand baths.6 FRACTIONAL DISTILLATION CHAP. be vertical in position instead of Much less space is thus taken receivers are and the more conveniently manipulated. Condensing tube of (a) faulty. up on the laboratory bench. internal diameter of 7 or 8 of 15 mm. and. pool of liquid collects . remedied by heating the wide tube close to the junction with the narrow one until the glass it drawing Fig. the error The fault is easily thus introduced may be serious. somewhat The Source Of Heat. frequently made may be referred to here. in the fractional distillation of a small quantity of liquid. long tube should be used and the cooling effect of the surrounding air will then be sufficient. or water or oil baths . . asbestos cardboard. the distillation little proceeding a . When a Liebig's condenser is used there is no advantage in having either the inner or the outer tube very wide an A . and for the outer tube. is The tubes are often a _ sealed together a in such way that when is FIG. /T7I at the junction (Fig. it is advisable to bend the narrow tube just below that the condenser its junction with the wider one. A mistake that is rather it is inconveniently heavy. is sufficient for the inner. it out very gently until is is soft and then has the form shown in When a long still-head used. 5 b.

The flask should be so placed that the flame actually comes in contact with the bottom of it this is especially necessary when the liquid to be distilled is liable . becomes considerably superheated. when a liquid over lime or phosphorus pentoxide. and a round-bottomed flask. carbon disulphide boils at 25. A water or oil bath need only be used when a solid sub- stance is is distilled present in the flask as. and the best plan or both. as a rule. to be of heat recommended. it may happen that the whole of the carbon disulphide will pass over without any ebullition whatever taking place. the top of the burner liquid.i SOURCE OF HEAT 7 are not. however. below the bottom of the flask. so that the minute flame touches the glass. to "bump. The be placed only about 2 mm. bump more or less violently when the pressure is greatly reduced. which boil quite regularly under the ordinary atmospheric pressure. such as carbon disulphide. however. if properly blown. is so thin walled that there is no danger of fracture when the naked flame is applied. ebullition will take place quietly and regularly. in these circumstances. or a number of small tetrahedra of silver or platinum. If. and if a quantity of it be distilled under this pressure with the flame placed some distance below the bottom of the flask. ing in this way. unless special precautions are taken. It is of ether. because the supply can be much more easily regulated without them. but customary to employ a water-bath for the distillation it is doubtful whether this is necessary or even advisable except in the case of the ethereal solution of a ." Many substances. and if a bubble does form there will be a sudden and perhaps violent rush of the extremely inflammable vapour. There are liquids which cannot be prevented from bumpis then to add a few small fragments of porous porcelain [a clay pipe broken in small pieces answers the purpose very well] or pumice-stone. or when the liquid is liable to decompose when heated with the naked flame. Under a pressure of 361 mm. for instance.

Still with steam jacket. Protection of Flame from Draughts. The most important point is to guard against draughts. Simple flame protector. the still frequently heated by steam under ordinary or increased pressure (Fig. some of the vapour escapes con- densation and comes in contact with a flame in the neighbourhood. and the condensed water is run off at B. is . Steam as Source of Heat. When one that will not bear heating much above an accident occurs it is almost invariably " because.FRACTIONAL DISTILLATION solid substance or CHAP. If a naked flame were used the distillation could be much more easily regulated. On the large scale." or to the distillation being carried on too rapidly. 100. be obtained it In is order that satisfactory results may necessary that B FIG. The steam is introduced through the pipe A. C. to do this. 7. FIG. 6). owing to bumping. 7). generally that below the water-bath. the distillation should proceed with great regularity and the heat supply must therefore not be subject to fluctuations. and there would probably be really less danger than if a water-bath were employed. the ordinary conical flame protector may be used. or a simple and efficient guard may be made from a large beaker by cutting off the bottom and taking a piece out of the side (Fig. and.

(6) The position of the mercury in the stem at any temperature depends on the expansion both of the mercury and the glass. Calibration. notwithstanding its many drawbacks. for two 1. It therefore necessary to compare the readings of a mercurial thermometer with those of an air thermometer. and. The thermometer must be carefully and it would be a great advantage if all thermometers were compared with an air thermometer. the rate of expansion increases with rise of temperature. used in preference to the air or the platinum If accurate results are to be resistance thermometer.I CALIBRATION OF THERMOMETERS The 9 fractional Thermometer. : and incorrect readings (a) In more espe- cially at high ones. it is impossible to obtain an abso- lutely cylindrical capillary tube. (c) Different kinds of glass have different rates of expansion. a number of fixed points may be determined by . which is therefore. 3). so that two thermometers made of different materials even if the capillary tubes were perfectly cylindrical different readings at the same temperature. These requirements are best fulfilled by the ordinary mercurial thermometer. for various reasons the first place. but even when this is done there remain other sources of error. obtained the following points must be attended to. or of another would give is mercurial thermometer which has previously been standardised by means of an air thermometer. but also to take readings of rapidly rising temperatures. and therefore the volume corresponding to a scale division cannot be quite the same in all parts of the tube. for both substances. of glass. will give different at other temperatures. calibrated. not only to read a constant or nearly constant temperature with great accuracy. one must be In carrying out a distillation able. Various methods have been devised for calibrating the stem (2. constructed of different varieties even if correct at 0" and 100. mercurial thermometers. instead of this. Or.

from and the variation of temperature for a differthe normal atmospheric pressure. TABLE 1.10 FRACTIONAL DISTILLATION CHAP. heating the thermometer with the vapours of a series of pure liquids boiling under known pressures. . Table 1 contains a list of suitable substances with their boiling points. Carbon disulphide . . ence of 10 mm. Substance.

not higher than 100" or even 150. apparently. volatilisation takes place. upper part of the tube when. the vapour condensing in the cold. slight fluctuations of temperature of the zero point takes place with much greater rapidity. If. a slight fall of zero point will be observed but after a day or two the greater part of this fall will be recovered. which is rapid at first. and up to. say. rise of more than 20 has several times been observecj in the case of soft German glass thermometers on being subthe jected to prolonged heating at 360. at any given temperature. subsequent heating to lower temperatures has very The best thermometers. and the remainder after a long period. are kept at a high temperature for a long time before being graduated. falling. Volatilisation of Mercury in Stem of Thermometer. contain nitrogen over the mercury. and it seems doubtful whether. which are graduated up to high temperatures. first recommended by Crafts. usually filled with nitrogen. . actual constancy of zero point has ever yet been attained. In all cases the rise. the temperature is really constant it appears to be gradually . such thermo- . The better thermometers. and probably 450. 360. are not . therefore.i ERRORS OF THERMOMETERS 11 are rise the periodical. however. as little effect on the zero point. however. 3. In the cheaper thermometers there is a vacuum above the mercury and. the higher is the final point reached. If. becomes slower and slower. a bulb being blown near the top of the capillary tube to pre- vent too great a rise of pressure by the compression of the gas but thermometers which are only required for moderate temperatures. when the mercury in the stem is strongly heated. higher the temperature the more rapid is the rise A and. at any rate. even after its zero point (b) If a thermometer has been rendered as constant as possible be heated and then cooled very rapidly. a thermometer has been heated for many hours to a given high temperature and then allowed to cool very slowly.

be wrapped round the bulb of the thermometer. and column not heated by the vapour. is somewhat uncertain. on page 13. a little fibrous asbestos. expressed in scale divisions. or even 80. be so placed in the apparatus that not only the mercury in the bulb but also that in the stem is heated by the vapour of the boiling liquid . where T is the observed boiling point. the vapour is liable to be superheated by the flame. or. the thermometer registers a perfectly con5. heated by a naked flame. As already mentioned. and. 2. Correction for Unheated Column of Mercury. N the difference between the mercury and that of glass is very frequently employed. as a rule. better results. after prolonged It heating. which. Superheating of liquid in the still is . 4. if possible. must be applied To the temperature read. would be much better if all thermometers required to were filled with register temperatures higher than 60 nitrogen. thoroughly moist. Table useful. i the temperature of the : the length of the mercury stem above the vapour. but it is found in practice to be too high and Thorpe has shown that the value 0*000143 gives coefficient The O00016 cubical expansion of . and unless the bulb of the thermometer is thoroughly moistened with condensed liquid. otherwise the following correction. add Q-QQQ113(T-t)N. for temperatures above 230. given by Thorpe (4) may be found Vapour. too high a temperature will be registered. the thermometer should. When the amount of very small.12 FRACTIONAL DISTILLATION cat i meters are used for the distillation of liquids boiling at so low a temperature as 100. it remains. a quite perceptible amount of mercury may volatilise and. If a very little cotton wool. with a pure liquid. at the best. errors amounting to 0-2 or 0'3* may occur.

ip eo ep os ep Tt< o o o o o o o b b b b b CO OS CO CO ^* t^ ^* '"H rH lO oooooooooo -H ob eo CO co ^ b 2 g b ^ b I?} t^. p ooooooooo bo ooooooooo bob OS i^.TjH O O <N O O ip O O r~ -H PH 00 -H * PH T <O OS t~* o 1 co ^ eo rH eo CO CO PH rH rH Ol <N r- OS os <N 00 (N bob bob & bob ^ bob 54 co -p CO oo CO T^ o eo <N ip <N & 8 gs -PH 1- CO t~ b b O qo ^ O5 rH PH (N "P <N p rH ^ ^ O O O * OOOOOpHpHpH. rH CO ^ b b b b b b b b 8 2 S .-lrH Ol PH <O co OS CD cp 00 p o o -H o o 8 o o ^ o o o o o CO bob o o o bob OOOOpHpHrHrH-H ^ bob oooo o ? 8 i bob b b O O * < r~4 sy Q-l CQ I ~ rH -H -H o o o o o *P b =? OJ rH * co b t^ r^.

the error is not so great as it would be if the bulb were not protected.). It is impossible to give any accurate and generally applicable formula for correcting the observed boiling point to that under normal pressure (760 mm. as a rule. stant temperature until the last trace of liquid in the bulb has disappeared. for the boiling point of a liquid varies greatly with the pressure. it may be necessary to read it frequently. though. but for water and the alcohols a better result is given by the formula 6 = 0-00010(760 -j>) (273 + t) Crafts (6) has collected together the data for a number of substances. since was published. This correction is applicable. without much error.14 FRACTIONAL DISTILLATION CHAP. to the majority of liquids. T. many additional accurate determinations and vapour pressure have been made.p) (273 + t) are easily obtained. 6. The baro- meter must always be read and corrected to (p. the temperature registered may be too high. it There are a few misprints in the table given by Crafts and. from which the values of c in the formula 6 = c (760 . t. With a water or oil bath the danger of superheating is greater.* Table 3 (p. and p is the barometric is where pressure (5). 269) and. and the cotton wool may become dry at the end of the distillation. 15) contains the boiling points on the absolute scale. In that case. Correction of Boiling Point for Pressure. in a long distillation or in unsettled weather. the values of c and those of dpjdt for some of the substances referred to by Crafts and also for a considerable number of additional ones * (7). "but the following may be taken as approximately correct 6 : = 0-00012(760 -_p)(273 + t) the correction in centigrade degrees to be added to the observed boiling point. of boiling point .

3. .CORRECTIONS OF BOILING POINTS The values marked with an 15 asterisk have been determined indirectly. TABLE Substance. and are not to be regarded as so well established as the others.

Iso-compounds have higher values than their normal isomerides. On the other hand. When replacing one halogen by another no change is usually produced. have very low values. higher. phenols. Examples The normal paraffins methyl and ethyl ethers toluene and meta-xylene acetone and benzophenone benzene. ethyl.Constant 1. and with phenol only 0-000015 lower. the higher the molecular weight the lower is the constant. . but with amyl alcohol it is only 0*000029. neutralise each other more or less completely. 6. the lower esters and perhaps some others. But the influence of the hydroxyl group in lowering the constant seems to diminish as the complexity of the rest of the molecule increases. 2. with methyl. 5. or any series of closely related substances. Thus. except the alcohols. the value is Benzene and a mono-derivative Examples naphthalene and bromonaphthalene. acids. water. acting in opposite directions. .16 FRACTIONAL DISTILLATION Relation of . Isopentane and normal pentane di-isomethyl isobutyrate and methyl butyrate. CHAP. the constant tends in general to be lowered as the molecular complexity increases. : lowered. of there is apparently the constant and the The esters formed from alcohols and acids is have nearly rather low values and here again the constant independent of the molecular weight. Examples: The four halogen deriva- By tives of benzene. All compounds containing a hydroxyl group alcohols. naphthalene and anthracene. in the case of the alcohols at any no relation between the values molecular weight. c to Molecular Weight and ConIn any homologous series. and propyl alcohols the constants must be as much as 0-000035 lower than those of the corresponding hydrocarbons. thus. hydrogen is replaced by a halogen. and if there are two iso-groups the value : is still Examples butyl and w-octane . . and these two factors. . 4. 3. acids. . . phenols. rate. : . stitution.

Modifications of the still-head are of great importance and will be considered later in Chapters X to XIII. there would be danger of fracture if the glass delivery sur- tube were air cooled effect is. very high temperature. 8). Metal vessels are generally t merely a matter of convenience and does not call for special mention. Any alteration in the shape of the Modifications of the Condenser. A. or ordinary boiling point. suffia long tube be emFor volatile very liquids. Condenser liquids. safer to especially when it is very inflammable. the delivery-tube must be cooled by ice or by a freezingmixture (pounded ice and salt. it if it may If a liquid boils at a suffers decomposition at its be necessary to distil it under . by water. of the The cooling rounding cient if ployed. FIQ. or ice and concentrated hydro- chloric acid are convenient for moderately low temperatures). but below about 170. however. the If a liquid boils straight Liebig's condenser is employed. For liquids boiling above the ordinary temperature. For ordinary laboratory purposes a round bottomed glass flask is the most convenient form of still. or " worm. 8. at a higher temperature than 170." tube should be used (Fig.MODIFICATIONS OF CONDENSER 17 Modifications Of the Still. In this case a spiral. it is employ a metal vessel. the drying tube. still is made use of on the large scale. for volatile Condensation of moisture in the receiver is prevented by Modifications of the Receiver. but if a large quantity of liquid has to be distilled.

] Various methods have been devised to allow of the receivers being changed without altering the pressure. Simple apparatus for distillation under reduced pressure. 9 serves to keep the pressure steady and to prevent oscillation of the mercury in the gauge. [The apparatus large globe in Fig.18 FRACTIONAL DISTILLATION CHAP. be collected in separate portions. s pump with the new receiver. The stopcock a is then closed to shut off the still from the pump. and c is turned so as to connect the 1. distillate is to of the receiver Thome's Apparatus. which is then exhausted until . the removal would necessitate admission of air into the and a fresh exhaustion after each change. The stopcock b is closed. 9. A series of stopcocks may be arranged in such a manner that air may be admitted into the receiver and a fresh one put in its place while the distillation bulb remains exhausted (Fig. 9 may be employed. reduced pressure. which is then disconnected and a fresh one is put in its place. of which the following may be mentioned. 10). but if the Fia. For cases of simple distillation the apparatus shown in Fig. and c is turned so as to admit air into the receiver.

several disadvantages . approximately it Fio. are sealed three nar- row tubes c.-itus for distillation under reduced pressure. and at (Fig. there is some risk of leakage when attended by many stopcocks are used even when a three-way stopcock is employed (as in Fig. the Fio. This method. Moreover.-u. liquid through which the condensed flows. right angles to 11). 11. when a and is b are again opened. 10) in place of two simple ones so and this is especially the case because ordinary lubricants cannot as a rule be used for b. Bredt's Hpp. the changing of the receiver. Thome's apparatus for manipulation of the stopcocks. in general use by of receiver may and there are two forms of apparatus means of which this may be done. during which the progress of the distillation cannot be closely watched. though comparatively simple. It is better to employ some method by which the change be effected without admitting air at all. 10. and a fourth c 2 . distillation under reduced pressure. and the exhaustion of the fresh receiver take up some time.MODIFICATIONS OF RECEIVER the pressure falls to 19 the required amount. Bredt's ApparaTo the end of the delivery tube from the still-head is at- tached a round-bot- tomed flask with a long neck to which b. a. 2. tus.

the The bulb three of the flask serves as one receiver. Prevention of is it Leakage. 12) so that any one of the tubes may easily be brought under the end of the This arrangement delivery tube. Briihl's Apparatus. distilled is When PIG. and narrow tubes connected with a The vessel by means of a perforated cork. i2.8 a liquid under reduced pressure.-Brtthi. long-necked flask is first placed with the three receivers in an inverted position. A number of test-tubes are placed in a circular stand which may be rotated within an exhausted vessel (Fig. and ordinary corks are seldom that quite air-tight. Page (8). seems on the whole to be the most ceiver convenient. but they do not in principle from one or forms other of the three described above. as the change of recan be effected with the greatest ease and of rapidity. can be brought vertically below the end of the delivery tube. 3. finds that all leakage . tube which serves to admit air when the is distillation is each of completed. Many other been differ apparatus have devised. under apparatus for reduced necessary all joints in the apparatus should be air-tight. so that the first fraction collects in the flask when a change is to be made. India-rubber stoppers cannot well be used for the still because this substance is attacked or dissolved by so many organic liquids.20 FRACTIONAL DISTILLATION d. however. the neck of the flask is rotated until the drops of distillate fall into one of the cylindrical receivers and each of these in turn cylindrical .

' 3. 2. 145. 159. Weights of Substances and their Specific Gravities when in the liquid state. . Thermodynamical Relations. which is apt to when dry). Young. 112. " Traite pratique de la Thermometrie do precision. 515. the application may be t repeated several times. 6. " Some Ramsay and Young. 217. Crafts. 1887." Trans.. sold crack in bottles (not gum arabic. Kopp. Mag. Berichte." Phil. 86. 8. 20. "Correction of the Boiling Points of Liquids from observed to normal Pressure. 1902.]. 162. Chem. Geschichte der Chemie. The gum may be conveniently applied if with a brush and. 1880. " Methods employed in Calibration of Mercurial Thermometers. " On the Relation between the Molecular Thorpe. 37.' British Association Report for 1882. Soc. "On the Correction of the Boiling Point for Barometric Variations. Chem." JBerl. Soc. 7. p.i PREVENTION OF LEAKAGE 21 may be effectually prevented by first exhausting the apparatus and then covering the cork with the ordinary liquid gum. News. [V. REFERENCES 1." Trans. 777. "Cork versus Rubber. 709. necessary. Beitrage L. 81. 20. Page..~1885. Guillaume. 5..' 4." Chem. 1902.

being then points drawn through the Pressure (Fig. a curve. first method the pressures exerted by the vapour of the liquid at a series of temperatures are ascertained and plotted against the temperatures. of The vapour pressures under atmospheric pressure its boiling point be determined by placing may some of the liquid over the mercury in a barometer tube a liquid at temperatures lower than . the statical By the and the dynamical. but also the boiling points of the liquid under Thus. different pressures. and water boils at 50 under a pressure of 91-98 mm. the vapour pressure of water at 91-98 mm. 13). 13. FIG. Vapour pressure curve. it represents not only the vapour pressures of the liquid at different temperatures..CHAPTEE II THE BOILING POINT OF A PURE LIQUID THERE of a liquid are two methods by which the "boiling point" under a given pressure may be determined. 50 is Evaporation in Absence of Air. This curve has a twofold meaning. the vapour pres- sure curve.

some of this air would be expelled and the measured pressure would be If the sum of the pressures of the vapour and of the air. and. there are certain liquid mixtures which behave in many respects like pure liquids . owing to the time required for diffusion between vapour and gas. but at a different temperature the vapour pressure would no longer be quite independent of the volumes. the liquid is pure and free from air. On the other hand. we could calculate the true pressure of the vapour by subtracting that of the air from the total pressure. the barometer tube contained air. ir THE STATICAL METHOD 23 and heating the tube to different temperatures. at a given temperature the vapour pressure of such a mixture would not depend on the relative volumes of liquid and vapour. The only difference is that evaporation into a vacuum the usually takes place almost instantaneously. under the reduced pressure and at the higher temperature. . after correcting for the expansion of the mercury. and we were able to measure its pressure. This constancy of vapour pressure test of the purity of a liquid. Dalton's law of partial pressures is. whereas in presence of a gas the process is a slow one. if necessary.* is. a very delicate Evaporation in Presence of Air. then. for its vapour pressure. If. otherwise. Kegnault has shown that almost exactly the same amount of evaporation will take place into a vacuous space as into same space containing any gas which does not act chemically on the liquid and is only very slightly soluble in it.CH. the pressure varies only with the temperature and does not depend on the relative volumes of liquid and vapour so long as both are present. in fact. applicable to and in constructing the vapour pressure curve * As will be seen later. this case. gives the vapour pressure of the liquid. It is necessary to take great care that the liquid intro- duced is quite free from dissolved air. The difference between the height of the barometer and that of the column of mercury in the tube. in fact.

Ebullition. The temperature of ebullition depends partly on the gaseous. " but the scientific term " boiling point is not applied to the temperature of a boiling liquid. depends solely on the temperature. in contact with that liquid. and if the walls of the vessel are very smooth and . the temperature of the liquid surrounding the bubble of the pressures of the liquid.24 FRACTIONAL DISTILLATION results. and partly on the cohesive force between its molecules and the it would adhesion of the liquid to the walls of the vessel be very difficult. the formation of bubbles of vapour at the lower surface of the liquid. and therefore. or roughnesses. definition to include all these factors. If a minute bubble of air is formed. by the presence of air dissolved in the liquid or adhering as a film to the glass and by sharp points. in contact with the glass where the heat is received. must be so high that the vapour pressure is equal to the sum atmosphere and of the column of Cause Of " Bumping . if not impossible. must be plotted against the temperatures. on the surface of the glass. partly on the vertical depth of the liquid. from suoh is and not the total pressures. to frame a convenient . or other. it CHAP. it is clear that the vapour must overcome the pressure of the column of liquid above it as well as that of the atmosphere. pressure on its surface. that the partial pressures of the vapour. THE DYNAMICAL METHOD A liquid is said to boil when it is in a state of ebullition. but in order that the bubble may increase in size by evaporation from the liquid is facilitated surrounding it." 1 But if the liquid is very free from air. flask If by means of a we heat a liquid in an ordinary glass Bunsen burner placed below it. under the most favourable conditions. Now the pressure exerted by the vapour of a liquid. it will serve as a nucleus for a larger bubble of vapour.

Point. and the air in the test tube will be expelled by the vapour.it DEFINITION OF TERM "BOILING POINT" formed with much greater difficulty . the temperature will rise." liquid The liability to intermittent ebullition is still greater when is covered by a layer of another lighter but less volatile liquid. when evaporating with a perfectly free surface. water when a bubble is at last formed. saturate the wool with the liquid under examination and suspend the thermometer in a test tube heated in a bath to a temperature at least 20 higher than the boiling point of the liquid. as when water has a layer of oil over it if a globule of well-boiled water is immersed in a bath of oil of the : the liquid same with specific gravity. and at the same time the temperature of the Under these conditions the liquid falls to some extent." By the term boiling point is to be understood the highest temperature attainable by a liquid. under a given pressure of its own vapour. vaporisation will take place. Thus. Definition of the term " Boiling. so that the water is in contact only bubbles is very difficult oil. and the temperature may be raised far above the 100 . and when the heat reaches the surface from without. vapour is evolved and the bubble increases in size with great rapidity. 25 clean. if we cover the bulb of a thermometer with cotton wool or other porous material. the temperature cannot rise above a maximum which is not influenced by that of the bath. bubbles are and higher it is then said to be superheated. so long as the cotton wool remains thoroughly moistened by the liquid. consequently. but depends solely on . the whole of the may be suddenly converted into steam with explosive violence. boils irregularly and is said to " bump. and when a bubble forms. the vapour pressure corresponding to the temperature of the liquid is much greater than the sum of the pressures of the the temperature of the liquid may rise much atmosphere and of the column of liquid. Under these conditions. the formation of indeed.

Condensation also takes place to a slight extent on the thermometer. and some of the vapour condenses. and it is usually more convenient to measure the condensing point of the vapour than the boiling To do this an ordinary distillation bulb point of the liquid. It will be observed that the boiling point of a liquid can only be correctly determined by observing the temperature of the liquid itself under such conditions that ebullition is impossible. is The walls of the vertical generally employed (Fig. the remainder being therefore in contact with condensed liquid. If the liquid. become too hot for vapour to condense on it. for high temperatures. on the atmospheric pressure. the pressure of the vapour and therefore. under these conditions. " This maximum temperature is the true " boiling point of the liquid under a pressure of its vapour equal to that of the atmosphere.) tube give up heat to the surrounding air. since these are equal. The vapour is especially liable to be superheated when a water or oil bath is employed as the source of heat. Determination of Boiling Point. 3. when the vapour is liable to be superheated. and as the thermometer loses heat only very slowly by radiation it may. unless the amount of liquid in the still is very small. playing against the dry walls of the vessel.26 FRACTIONAL DISTILLATION CHAP. is not known to be quite pure. it is advisable to distil a fair it . fibrous asbestos and. 1 The true boiling point of a liquid is identical with the condensing point of its vapour under the same pressure. provided that some liquid Is present and that the vapour is not mixed with an indifferent gas or vapour. owing to the flame or the heated gases from it. in determining a boiling point or carrying out a distillation with the ordinary apparatus. or. it is always advisable to cover the thermometer bulb in this way. the boiling point of which is to be determined. Condensation on the thermometer is promoted by covering with a little cotton wool (1).

not very different and read the in volatility is from the pure liquid itself. on the other boiling point again. hand. of the impurity will be gained by observing the extent of the rise and whether it takes place in the early or late stages of the distillation. but it would be more satisfactory. or whether there is a amount steady rise during the whole period. and some idea of the nature and . is probable. and that the precautions mentioned in the last chapter have been attended to. or to compare it . collect the best portion of the distillate separately. or both. especially in the second case.n . redistil it and to If. Assuming that the liquid is pure. the corrected temperature will give the true boiling point of the liquid under a pressure equal to that of the atmosphere at the time. that the thermometer has been compared with an air thermometer. the presence of one or more substances. If the temperature is steady at but rises rapidly when the distillation is nearly complete. there is a fairly steady rise of temperature throughout. to liquid is present. to compare the boiling point of the liquid with that observed by another experimenter. Reduction of Boiling. it may be assumed that there is a much volatile liquid present. and it impossible to ascertain the boiling point of the pure liquid without carrying out a fractional distillation to remove the impurities. Effect of Impurities on Boiling Point. 1 If there is a rapid rise at first and the temperature afterwards remains steady or nearly so.Point to Normal Pressure. however. It is frequently necessary. the conclusion may be drawn that a much less volatile first more In either case the steady temperature will approximate very closely to the true boiling point. the quite constant during the whole distillation will rise (unless we are dealing with a mixture temperature of constant boiling point). EFFECT OF IMPURITIES 27 quantity of it if it is pure the boiling point should remain if not.

may be admitted through a tube which passes nearly to the bottom of the capillary The latter method is an excellent one for preventing the bumping. or fragments of porous material such as unglazed porcelain or pumice-stone. the Law of Partial Pressures. 14. pressures in fact. and it is therefore convenient what the boiling point would be under normal pressure and. small tetrahedra or platinum. with that to ascertain some other liquid. or 2. (p. applicable to Dalton's law of partial the boiling point of a . that the liquid is liable to boil with bumping 1. the total pressure remains unaltered but the partial pressure of the vapour is diminished. in stating the result for future reference. out it must be remembered that an error is caused by the presence of the air which is introduced. A slow current of air vessel. That this is so may be easily proved by altering the rate at which air enters through the capillary tube the more rapid the introduction of air the lower will be the observed temperature. is. and unless the amount of air is exceedingly small observed boiling point will be sensibly too low. The method of correcting the boiling point of a liquid from observed to normal pressure is given on p. and the observed temperature is lower than the boiling point under the read pressure. is . Prevention Of " Bumping.28 FRACTIONAL DISTILLATION of CHAP. may be added to facilitate the formation of bubbles. the distillation has to be carried out under greatly reduced pressure. If there is air mixed with the vapour. to give this reduced boiling point. The true boiling point of the liquid and the condensing point of the vapour depend on the pressure of the vapour itself and not necessarily on the total gaseous pressure to which the liquid exposed. 24) and two methods have been proposed of silver to prevent this. As already mentioned." especially when It frequently happens.

the temperature fell to . as well as . under a pressure of of sponge soaked in water freer evaporation. was effected by suspending in the bell-jar a thermometer. More rapid diffusion and removal of vapour. Here the partial pressure of the vapour must have been only a very small fraction of the total pressure. for the vapour pressure of the water would be far lower than the total gaseous pressure. the temperature actually fell to . The following experiment (3) affords a still more striking " boiling proof of the correctness of the statement that the point does not necessarily depend on the atmospheric A copper air bath. the whole being placed under a belljar which could be exhausted by an air-pump. The " boiling point of the water therefore fell below the freezing point but in these circumstances.. indeed. at which the boiling point of water under ordinary conditions is 34. Water. " . if air is mixed with the vapour. it is the partial pressure of the vapour and not the total pressure that must be taken into account Many experiments might also be described to the truth of this statement. : Experimental Proofs. through which a current of pressure.. . 40 mm.10. supported on a short tripod on a second clock glass containing strong sulphuric acid. so that. when distilled in the ordinary manner under a pressure of 15 mm. . Here the aqueous vapour was rapidly absorbed by the strong sulphuric acid.ii LAW OF PARTIAL PRESSURES 29 liquid as. the bulb of which was covered with a piece in this case. when the total pressure was 15 mm. the partial pressure of the vapour in contact with the water " must have been only about 4 mm. boils at about 18 but Schrotter (2) observed so long ago as 1853 that when some water was placed in a shallow clock glass. bubbles could not be formed (except possibly air bubbles). the pressure was reduced to 15 mm.3 when. but it will be sufficient to prove mention the following (p. is evident from the fact that in deter- mining a boiling point by the statical method. 23)..

66. and the explanation of this fact . instead of remaining however. moving about over the hot surface like a Careful experiments (4) have globule of mercury on a table. the bulb of which was covered with cotton-wool moistened with boiling water. and a thermometer. shown that the temperature of the water under these conditions does not reach 100. both of which remained nearly conlimiting the supply of air and by introducing steam. the temperature of the water fell again to 80.30 FRACTIONAL DISTILLATION CHAP. nothing was altered but the relative pressures of the aqueous vapour and of the air surrounding the water. In this experiment. Lastly. nor does it boil although rapid evaporation takes place. and when the air was almost completely replaced by aqueous vapour. By greater the partial pressure of the vapour. The pressure of the atmosphere was 748*mm. and remained nearly con- bath had fallen slightly. it is clear that the temperature of the water did not depend on that of the air bath or on the atmospheric pressure.. though that of the at nearly 100. the higher was the temperature reached by the water. the temperature of the water rose to about 80. on allowing some of the steam to escape. It is well known that when a drop is allowed to fall on a red-hot sheet of metal. was heated to 205. air could pass freely. the temperature rose very nearly to the ordinary boiling point of water. the temperature of the water rose to 99. such as platinum. fell rapidly to stant at this point. was suspended in the bath through an opening in the top. of water Spheroidal State. the current of air through the bath was retarded by closing the grating in the side. but it assumes the spJieroidal state. and the water on the cotton-wool was in a strongly heated chamber. The stant. it does not touch the metal. and when steam was introduced into the bath so as to replace the air as completely as possible by aqueous vapour. When. yet its temperature.

is Heat as there is . . : received from without towards the surface. why ebullition reasons.ii LAW OF PARTIAL PRESSURES probably that the 31 is aqueous vapour surrounding the spheroidal drop is diluted with air. is platinum 2. 47. the difference temperature which is usually observed between the wet and dry bulb in the ordinary hygrometer depends on the difference between the partial pressure of the aqueous vapour in the air and the maximum pressure possible at the temperature of the dry bulb. Ramsay and Young. on.. Soc. 611. the vapour pressure Non-miseible Liquids." Liebig's Annalen. 188. so that formed. 42. 129. "On the Freezing of Water in Rarefied Air. no contact between the water and the heated to the temperature lower than the atmospheric pressure. "Sublimation. Balfour Stewart's Treatise on Heat. Suhrotter. as will be shown later liquids are distilled together. _'." Trans. is to say. would at once collapse. "On a New Method Vapour Pressures 1885. 6th Edition. Young. so that its partial pressure There are two is less than the total atmospheric pressure. 4. from the whole of which evaporation can take place freely. Chem. 1853. either of which would be sufficient. Wet and Dry Bulb Hygrometer." Thorpe's Dictionary of Applied Chemistry. the fact that Lastly. of Solids of Determining the and Liquids.. each liquid acting like an REFERENCES 1. cannot take place 1. the vapour of indifferent gas towards the other. that of water at that temperature. III. 3. 88. the boiling point is lower than of either component when two non-miscible when distilled alone. The vapour pressure corresponding water if of the a bubble. may similarly be explained by the law of partial pressures. in Again.

The attraction of the like molecules of those of A for each other and of those of B for each other. The mutual attraction of the molecules of A is and B. and. it is not unusual to find a small of contraction attended by slight heat absorption. one should consider 1. when a little water is added to normal propyl but in the case of certain closely related chemical alcohol compounds. such as chlorobenzene and bromobenzene. in the comparison of various pairs of liquids. as the mutual attraction of the unlike increases relatively to that of the like molecules. contraction.CHAPTEE III VAPOUR PRESSURES OF MIXED LIQUIDS Influence of Molecular Attractions on Miscibility of Liquids and on Heat and Volume Changes duringAdmixture. A and B. one may relatively so expect the liquids to be non-miscible or very nearly so. . when mixed together. in many cases. infinite first. Generally. If the attraction of the unlike molecules slight as to be negligible. absorption of heat at finally diminishing to zero and changing to increasing heat evolution and diminishing expansion followed by increasing These various changes do not. and it is reasonable to assume that in the process of mixing there might be slight absorption of heat and slight expansion. as example. amount . 2. one would expect increasing miscibility. In studying the behaviour of two liquids. there for . among liquids which are miscible in all proportions. and run strictly pari passu. With a somewhat greater relative attraction between the unlike molecules there would be miscibility within small limits.

to be two simple cases 1. Konowaloff and other observers. the observed vapour pressure is equal to the sum of those of the two liquids when heated separately to the same temperature. tively so slight that the liquids are practically non-miscible 2. A for A.] There would appear then.CH. Each liquid. in is NON-MISCIBLE LIQUIDS 33 neither any appreciable change of volume nor any measurable evolution or absorption of heat when the liquids are mixed together. the maxipressure may not be reached for a considerable time is unless the heavier liquid by shaking or stirring the surface to facilitate ite evaporation. That in which the attraction represented by a 1-2 : is rela. B for B. If. For such substances it is probable that the different molecular attractions. the upper layer is deep. Magnus. namely that holds good. brought to PARTIALLY MISCIBLE LIQUIDS In the case of two partially miscible liquids the vapour pressure was found to be less than the sum of those of the D . [a l2 represents the attraction of s' the unlike molecules and a x and a2 the respective attractions of the like molecules. behaves quite in fair quantity independently of the other and. The vapour pressures of some mixed liquids and of nonmiscible pairs of liquids have been determined by Eegnault. That of two closely-related and infinitely miscible liquids which show no heat or volume change when mixed together. are very nearly equal. it does not matter what are their relative amounts or what are the vapour. relative volumes of liquid and mum however. in fact. Berthelot (1) and by Galitzine (2). and that the relation suggested by D. NON-MISCIBLE LIQUIDS It was shown by Eegnault that when two non-miscible mercury in a liquids are placed together over the barometer tube. so long as both are present and one is not covered by too deep a layer of the other. and A for B.

relatively very great (formic of the mixture may be It seems reasonable to suppose that. at the same temperature. Guthrie. dep&ads on the relative attraction of the like and the unlike molecules. so with the vapour pressures of the mixtures. respectively.34 FRACTIONAL DISTILLATION CHAP. and . and the question what is the normal behaviour of mixtures has been discussed by several investigators. with normal propyl the vapour pressure. as regards vapour pressure.w)P B '- where P. There can be no doubt that the behaviour of mixtures. when that attraction acid less is and water) the vapour pressure than that of either component. concluded that Normal Behaviour of Mixtures. components A and B. the relation between the vapour pressure. On the other alcohol and water hand. like that of a partially miscible pair of liquids. INFINITELY MISCIBLE LIQUIDS The vapour pressures of many pairs of infinitely miscible liquids have been determined by several experimenters and. the relation between vapour pressure and composition should be a simple one. In other words. but greater than that of either one singly at the same temperature. and w is the percentage by weight of the liquid A. expansion. components. When the mutual attraction of the unlike molecules is not much more than sufficient to cause infinite miscibility for example. when the attractions of the like and unlike molecules are equal or nearly so. may be greater than that of either component at the same temperature. P A and P and of the two B are the vapour pressures of the mixture. the vapour pressures should be expressed by a formula which may be written P = wPA + (100 v . very different results are obtained in different cases. Guthrie (3) if we could find two liquids showing no contraction. as with the changes of volume and of temperature on mixing the liquids. or heat change on mixing.

the critical pressures of which.. m is the molecular percentage of A. or about 7'9 It is probable that. 100 should then hold Van der Waals.c^Guthrie was clearly in error in taking percentages by weight.= s/ajtta does not accurately represent the facts. for there are cases known in which the relation does not hold. although both the liquids have normal molecular weights (for example. the formula per cent.910 mm.raP (100 m) where good. 34. or carbon tetrachloride and benzene). and 33. In order to test the correct- Kohnstamm's Experiments. the maximum deviation from the straight line amounted to about G mm. ness of the conclusion arrived at by Van der Waals. w-hexane and benzene. are nearly equal. Berthelot. Speyers (4) concludes that the relation be- tween vapour pressure and molecular percentage composition is always represented by a straight line when tlfe moleular weight of each substance is normal in both the liquid and + P* gaseous states. Kohnstamm (6) has determined the vapour pressures of various mixtures of carbon tetrachloride and chlorobenzene. The equation. P = . and he finds that the curvature is not very marked. on a total calculated pressure of 76 mm.m NORMAL BEHAVIOUR OF MIXTURES 35 the percentage composition by weight should be represented by a straight line. and Speyer has certainly made his statement too general..180 mm. last-named relation Van is true der Waals (5) considers that the when the critical pressures of the two liquids are equal and the molecular attractions agree with the formula proposed by Galitzine and by D. j 2 = ^Ah. in this case. Speyers. a 12. At the temperature of experiment. D 2 .

toluene and bromobenzene. D. and it is therefore impossible to place complete reliance on the experimental data( The critical pressures of these substances have been determined and those of the components of the mixtures are in no case equal.36 FRACTIONAL DISTILLATION Closely CHAP. in one case at least (carbon tetrachloride : and benzene). From a study of the volume produced by mixing various pairs of together. results. cit.) made such determinations with mixtures of ethyl bromide and ethyl iodide. however. but unfortunately the method. pressures of ethyl bromide and ethyl though it is very probable that they Linebarger (8) has determined the vapour pressures of a few mixtures of each of the following pairs of liquids by benzene drawing a current of air through them (p. . benzene and bromobenzene. Brown (7) concluded that this change would probably be smallest in the case of closely related chemical compounds and he obtained some indirect evidence in favour of this view it seems a fair assumption that for such mixtures this. however. and his results make it probable. gave inaccurate . in nearly every case been carried out with mixtures of liquids which have no very close chemical relationship. Exterminations by Dynamical Method. should be of a simple character. but Guthrie (loc. that the for this equation P = mP* + It ( ~m ^ PB holds good pair of substances. critical is not known. pressures of mixtures may The vapour be determined by the dynamical . though not certain. Direct measurements of vapour pressures by the statical method have. 80) and chlorobenzene. toluene and chlorobenzene. may be. His results are in fair agreement with the above formula. F. and other physical relations. whether the iodide are equal. alteration of liquids Related Compounds. Experimental Determinations by Statical Method.

Chlorobenzene and Bromobenzene. This method has been adopted by Lehfeldt (p. A. 25. known. the logarithms of the at and from the curves so obtained the pressure any required temperature can be read off. pressures or. of . the boiling point determinations for a considerable number of mixtures would have to be carried out through a wide range of pressure. of we may calculate the theoretical vapour pressure each mixture from the formula. 76). In order to obtain the complete vapour pressure curve for two liquids at a given temperature. of the less volatile component. the vapour pressures of mixtures may be deter- mined pressures. 37 method the still is kept at a uniform temperature by means of a suitable bath. and the theoretical vapour pressures at these temperatures were then calculated. The results are given in Table 4. Suppose that we determine the boiling points of mixtures and 75 molecules per cent. 75) and by Zawidski (p. and the latter experimenter finds that mixtures of ethylene di- bromide with propylene dibromide give results in conformity " m)Ps + with the formula P = mI>A The critical ^ . say. however. better. indirectly from their boiling points under a series of The boiling points of each mixture are mapped against the pressures.in DETERMINATIONS BY DYNAMICAL METHOD . 50 peratures. then. If. but a less elaborate investigation is sufficient to ascertain whether or not the above formula is applicable. we know the vapour pressures of each component at the three temcontaining. Again. at a few pressures above and below 760 mm. and the pressures are observed under which ebullition takes place. in order to ascertain the boiling points under normal pressure with accuracy. The boiling points three mixtures of chlorobenzene and bromobenzene were determined in this way (9).. pressures of these substances are not.


the toluene partly aromatic and partly aliphatic. / be drawn from the results that for mixtures of compounds the relation between vapour pressure and molecular composition is represented by a line which is very nearly. and it may be remarked that the chemical relationship is not quite so close in this case. but their critical pressures (Table 5) are widely different (i2). 42]. Volume Change and Vapour It does not appear that change of volume. Comparison of Heat Change. account for the irregular results with w-hexane and ?i-octane. pairs of substances of Influence of Chemical Relationship. and it will be seen from Table 6 (p. or temperature. p.in INFLUENCE OF CHEMICAL RELATIONSHIP 39 Other Mixtures of Closely Related Compounds. 0'57. however. and 0'08 [compare Table 7. For the first three and the last in Table 6.P. 018. the more difficult is it to determine the boiling point of a mixture with accuracy. but temperature are very noticeable with benzene and toluene. 0-12 and 0-02 respectively. and the values of 100(P P)/P 0'58. 0'35. straight. The greater the difference between the boiling points of the two liquids. clusion may closely related Errors of Experiment. The con0'60. it may be said that they are very sniall. the benzene molecule being wholly aromatic. and the greater also to some extent is the This may partly probability of error in the value of P. that it is only when the critical pressures are equal that the formula is applicable. It does not follow. when the liquids are mixed can be relied on as a safe guide to the behaviour of a mixture as regards Pressure. of volume and the expansion and fall of With regard to the mean differences P . 0-20. 40) that the deviations may be exceedingly small when the liquids are closely related. if not quite. the corresponding temperature differences being only 017. the changes temperature are exceedingly small. .


and of toluene. in = P PA +(100-m)PB jnn accurately the components are very closely related. a 2 (Van i* u *u t represented by the formula 3. the rule is not followed. and even when there is both contraction and evolution of heat on mixing the components. When not only is the chemical relationship very close. from methyl to isobutyl alcohol. for the actual pressure is the lower. as wiU be seen from Table 7 (p. related. however. 41 and that. In the case of mixtures of the alcohols with benzene. the observed pressure should be higher than the With benzene calculated. although the value of It will be seen that contraction is frequently by absorption of heat. When we come to consider the behaviour of liquids which are not so closely related. as is the case with the first pair. 1. although there is expansion and fall temperature on mixing. yet the fall in temperature.P)/P. the percentage difference between the observed and calculated pressures When . inconsiderable. and that may P'P (P P)fP diminish as the molecular weights of the alcohols increase. it is remarkable that. Components not Closely Related. as a rule at any rate. we find in many cases much greater changes of volume and temperature.in GENERAL CONCLUSIONS . becomes greater. 42). 2. when there is expansion and fall of temliquids perature. accompanied have a positive value when there is contraction. and the very small volume-changes pass from negative to positive. and also much larger values for 100(F . The following conclusions may : be drawn from the foregoing data When the two components are chemically very closely changes of volume and of temperature on mixing the liquids are. the vapour pressure of the mixture V^ is . but the critical pressures are not equal. the but the critical pressures are equal and a^ = der Waals). General Conclusions.


These alcohols may be regarded as being formed from water by the replacement of one atom of hydrogen by an O H. even no molecular association in tne case of either differences are. 100(P'-P)/P are usually very Importance of Chemical Relationship and Molecular In considering. even when. there is molecular association in the liquid state. the percentage as a rule. When the components are not very closely related and the molecules of one or both of them are associated in the liquid state. Alcohols relationship Water is Paraffins. thus (CnHjn+j) tives of the paraffins. as regards vapour pressure. Association. found that. if the two liquids are not closely related. the values of large indeed. the whether the molecules of either or both of them are question and that. in the homologous series. if of them. The influence of chemical well seen by comparing the properties of mixtures of the monhydric aliphatic alcohols with water on the one hand and with the corresponding paraffins on the other (13). as with methyl and ethyl alcohol. When is there the two liquids are not closely related. the properties of the alcohols recede from those of water and approach those of the This is well seen in Table 8 corresponding paraffins. then. 4. the probable behaviour. thus C n 2n+1 (OH). as the magnitude of the alkyl group increases. associated in the liquid state is also of great importance. 5. of two liquids. or as hydroxyl derivaalkyl group. The alcohols are thus related to. it is (page 44). and their properties are intermediate between those of water and of the paraffins and H . . much greater. it would appear that chemical relationship is the chief point to be considered.in IMPORTANCE OF CHEMICAL RELATIONSHIP 43 100(P'-P)/P is very small.


in fact. and will extract some water . anhydrous copper sulphate is. from strong spirit. as Konowaloff has shown. however. The lowest member of the series. With ethyl alcohol and water the deviation is considerable. we to mixtures of n-propyl alcohol mum find that there is a very well defined maxipressure. 14.the relation those of the pure components. at the boiling points of the alcohols under a pressure of 400 mm. in which the vapour pressures of mixtures of four alcohols with water are given. and more recent and accurate observations have shown that a particular mixture exerts a maximum vapour pressure. Volume and Heat Changes. we come and water. although the liquids are miscible in all . increasing heat absorption in the case of the higher alcohols. Miscibility. according to Forcrand. the vapour pressures of mixtures of these substances are in all cases intermediate between . however. as will be seen from Fig. methyl alcohol. there is. the intermediate ones within limits. When. but it is not a sufficiently powerful dehydrating agent to remove the whole. bears the closest resemblance to water. but the experiments of Konowaloff are not sufficiently numerous to bring out this point.CH 4 may be obtained. while the higher alcohols are practically insoluble in water.in WATER AND ALCOHOLS 45 to a considerable extent. Lastly. quite insoluble in ethyl alcohol. greenish-blue crystals of CuS0 4 . alcohols is also The gradual change in properties of the weU shown by the fact that the lower alcohols are miscible with water in all proportions. and the curve representing between vapour pressure and molecular composition does not deviate very greatly from a straight line. the contraction and heat evolution that take place on mixing the alcohols with water diminish with rise of molecular weight . and. giving a blue solution from which. Vapour Pressures.

isobutyl alcohol nin- and water. proportions.46 FRACTIONAL DISTILLATION CHAP. 80 60 40 Molecular perccntagcotalcoM. 800 Isobutyl alcohol -* 700 20 FIG. of Vapour pressures mixtures of alcohols with water. and the curve shows considerable resemblance to that representing the behaviour of the partially miscible liquids. tion Maximum Vapour Pressure. 14. The queswhether a mixture of maximum vapour pressure will be formed in any given case depends partly on the deviation Mixtures of .

. 15) while. very slightly higher than that of pure qthyl alcohol. Be no pressure (A. for a mixture containing about 89 or 90 molecules 40 60 80 Molecular percentage ofalcohol. partly on the difference between the vapour of of two components. the difference between the vapour pressures at 63 is 229 mm. maximum vapour and the molecular percentage of alcohol in the . If the difference between the vapour pressures of the pure substances at the same temperature were 350 mm. and there is a maximum pressures ethyl of the alcohol vapour pressure.. Thus. Fig. there would . of alcohol (B. in the case and water. 15). the maximum vapour pressure would be far more obvious. Fig. and the deviation of the curve from straightness were the same. if the vapour pressures of alcohol and water were equal and the deviation the same. 15. per cent.Ill MAXIMUM VAPOUR PRESSURE 47 the vapour pressure-molecular composition curve from straightness. FIG.

. and most of them are very difficult to prepare in a pure state. isobutyl and isoamyl alcohols. therefore.] In all Isobutyl alcohol these respects. We cannot well study the miscibility of the alcohols with the corresponding paraffins. the fall of temperature.48 FRACTIONAL DISTILLATION CHAP. and is easily purified. it may be said that the properties ' of the alcohols approach those of as normal hexane. like 20. mixture that exerted it would be something of about 90 (c. Again. Isoamyl alcohol . as the molecular weight increases. may be conveniently prepared by the action of sodium on propyl iodide. normal hexane forms mixtures of maximum vapour pressure with the lower alcohols. but with those of the higher alcohols mixes in all proportions which have been investi- gated ethyl. [Ethyl alcohol . Mixtures of ?*-Hexane with Alcohols. Fig.2-35. 15).2-40.1-85. Lastly. their . 43 112 203 315 mm. w-Propyl alcohol . propyl. increase. however. 14). is by the increasing curvature of the pressure- molecular differences composition lines (Fig. diminishes slightly as the molecular weights of the alcohol rise. Normal hexane. The maximum between the pressures represented by the actual curves and the theoretical straight lines are roughly as follows : Methyl alcohol and water Ethyl n-Propyl Isobutyl . but not with isoamyl alcohol or any other of higher boiling point. because the first four of these hydrocarbons are gaseous at the ordinary temperature. This paraffin is only partially miscible with methyl alcohol at the ordinary temperature. instead The gradual divergence from those indicated of the properties of the alcohols of water.. but molecular weights recede from those of water..2-55. on mixing hexane with the alcohols in equimolecular proportions.

1898. "On Mixtures of Gases. 126. and it is practically certain that no alcohol of higher boiling point would form such a mixture.molecular weights of the alcohols increase. and ethyl alcohol without difficulty by treatment with water. for mixtures of the alcohols with water.. such as normal hexane indeed. which occur on admixture are. n-propyl. 1890. be noted that the relationship of the alcohols to the aromatic hydrocarbon. form a mixture mum vapour pressure. 770. 1703. the variations in the temperature and volume changes." Wied. relatively to that in water. two substances." Compt. though they are much smaller. rend. becomes greater as the . miscible with benzene in all proportions but while methyl. benzene. in all probability. tertiary butyl and isobutyl alcohols form mixtures of maximum vapour pressure with that hydro- carbon. isoamyl alcohol does not.P}jP show a steady rise as the molecular weights of the alcohols increase. is by no means so close as to a paraffin. "On Dalton'a Law. there is expansion and absorption of heat when benzene and hexane are mixed together and the . the values of 100 (P . more . the corresponding values for mixtures with benzene It may also be mentioned that the solubility fall regularly. 41. for while methyl alcohol can be separated from its benzene solution with the greatest ease. . > It will be seen from Table 7 that while. of maxi- REFERENCES 1. Ann.. the extraction of propyl alcohol is more troublesome. On the other hand. and behaves The lower alcohols are in a somewhat similar manner. in the same direction for benzene as for water. isopropyl.in BENZENE AND ALCOHOLS 49 Mixtures of Benzene with Alcohols. Berthelot. of the alcohols in benzene. Benzene is much easily obtained in quantity than hexane. 2. however. D. It should. Galitzine. and that of isobutyl or amyl alcohol by this process is exceedingly tedious. generally. ethyl.

Chem." Trans. Journ." Trans. Vapour Pressures and Boiling Points of Mixed Altschul. 81. " The Young. FRACTIONAL DISTILLATION Guthrie. Chem. 577. . 1902. 81. " On the Vapour Tensions of Mixtures of Volatile 9. and with Benzene and 739. " On the Critical Constants of some Organic Compounds. Liquids. 1895. Kohnstamm." Inaugural Dissertation. Roy. der Waals." Trans. 83... Amsterdam. Chem. der Waals. The Vapour Pressures and Boiling Points of Mixed Liquids. Acad.. 495.]." Phil. 615 and 690. in 4. Chem. "Experimental Investigations based on the Theory of 5. Chem.. 170." Amer. Part II. D. CH. 768. Chem. 55.. Alcohols (1) with Water." Trans. 1901. 3. Van 6. Soc. Sci. " On the Volume of Mixed Brown. 1881. 1902. Amsterdam.. "On some Thermal and Volume Changes attending Admixture. (2) Soc." Journ. 12. 1900. F. Part I. and Water.. Soc. Van 7. Liquids.. 39. 9. Liquids. 1902." Trans. Soc. " Properties of the Pressure Lines for Co-existing Phases of Mixtures. "Some Boiling Point Curves. Compounds 486. Mag. 717 with Benzene. Young. 341. 1889. 13. 11. " On the 11. Speyers.50 3. Vapour Pressures and Specific Volumes of similar 10. IV. 18." Proc. [V. 202.. Chem." Zeitschr. 45. Linebarger. 1900. 8. of Elements in Relation to the Position of those Elements in the Periodic Table. " The Properties of Mixtures of the Lower Young and Fortey.. Soc. physik. Amer. 1884. . " Young and Fortey. Soc.. 17.. 1893.

but also its boiling points under different pressures .CHAPTER IV BOILING POINTS OF MIXED LIQUIDS Statical and Dynamical Methods of Determination. the curve which shows the relation between temperature and pressure representing not only the vapour pressures of the liquid at different temperatures. of the components at the related compounds. this statement applies equally to the boiling points any given mixture under different pressures. for which the formula in all same temperature b. That of non-miscible liquids. That of closely p = mP A + (100 m)PB 100 holds good. for which the vapour pressure of the two (or more) liquids together is equal to of : the sum of the vapour pressures . each vapour behaving as an indiffer- E 2 . it has been shown that thare are two simple cases a. It has been stated that the boiling point of a pure liquid under a given pressure may be determined by either the statical or the dynamical method. which are miscible and of a few other pairs of infinitely proportions. It is in these two cases only that the boiling points can be calculated from the vapour pressures of the components. NON-MISCIBLE LIQUIDS Dalton's law of partial pressures is applicable to the case of non-miscible liquids. As regards the vapour pressures of mixtures. miscible liquids.

is carried out in the ordinary way. and so on. mm. are given for each degree from 89 to 93. Conversely. when the atmospheric pressure is 761-6 mm. ent gas to the others.. until there was scarcely any chlorobenzene visible in the residual liquid. The boiling point is therefore that temperature at which the sum of the vapour pressures of the components is equal to the atmospheric pressure. In Table 9 the vapour pressures of chlorobenzene and of water (1). should be 91. so that the vapours are unmixed with air. equal to that of the atmosphere. when it rose rapidly to nearly . The temperature is necessarily the if same and the total pressure. the observed boiling point. Clilorobenzene. 100 c. Vapour Pressures in Temperature. and also the sum of the vapour pressures at the same temperatures. is the distillation for all the liquids present.c. and the boiling point of each liquid depends on the partial pressure of its own vapour. when chlorobenzene and water are distilled together in an ordinary distillation bulb. 761-6 mm. Thus. barometer tube. In an actual experiment.c. of chlorobenzene and 80 c. when chlorobenzene and water say in a Experimental Verification.52 FRACTIONAL DISTILLATION CHAP... at 91. TABLE 9. two liquids which are practically non-miscible. 89 90 91 92 93 are heated together. and it was found that the temperature varied only between 90-25 and 90-35. the observed vapour pressure at 90* should be 733-8 mm. Calculation from Vapour Pressures (Chlorobenzene and Watep). of water were distilled together when the barometric pressure was 740'2 mm.

will be well to take the case of partially The boiling point of a pair of such substances is higher than that calculated as above for non-miscible liquids. the difference between the observed and calculated temperatures is not serious.iv LIQUIDS MISCIBLE WITHIN LIMITS is 53 : 100. aniline. dissolves only about 3 per about 5 per cent. and the boiling point was found to remain nearly constant at Aniline of cent. and Water aniline 98-75 for some time but afterwards rose slowly to 99 65 while there was still a moderate amount of water visible - with the residual aniline. It may be assumed without - sensible error that the value of dp/dt is constant over this small range of temperature and the boiling point should therefore be 733-8 ) that actually observed.c..c. of aniline and 200 c. when the distillation was recommenced. 90+ ~ = 9Q ' 23 ' which is ( 761-6 very cloafe to LIQUIDS MISCIBLE WITHIN LIMITS Before miscible considering liquids for the which behaviour of those infinitely the relation between vapour is pressure and molecular composition ft represented by a straight line. and Water. but if the miscibility is slight. of water at 12. rose It would therefore appear that the composition of to 98-9. the liquid in the still may alter considerably without producing any great change in the boiling point. of water were distilled together under a pressure of 746-4 mm. though the . and the observed boiling point will be decidedly lower than that of either component. temperature. miscible liquids. The theoretical boiling point calculated as follows The 91 increase of pressure for 1 rise of temperature from 90 to is 761-6 733-8 = 27'8 mm. though the solubility in each case is considerably greater at 100. Fifty c. The distillation was then stopped and a large amount of water was added to the residue the .

Nitrobenzene is reduced by finely divided iron and water with a little hydrochloric acid.54 FRACTIONAL DISTILLATION CHAP. Advantage is taken of the fact that the boiling point of a pair of non-miscible or slightly miscible liquids is lower than that of either pure component. 722-5 Water. or which are mixed boiling points with solid impurities. the commercial preparation of aniline may be described. water is the liquid with which the substance is distilled and the process is commonly spoken of as " steam distillation. so that in the next distillation the aniline is carried over into the still again. Vapour Pressures in Temperature. temperature does not remain so constant as it does with chlorobenzene and water. which contains a little aniline. the products formed being ferrous chloride.- mm. Aniline. is would therefore be ( (97 + 749'5 \ % ^ ^SrlV 722'5/ = 97 9 ' > which 85 lower than the temperature which remained nearly constant for some time and 1'75 lower than that which was observed present. to distil substances which could not be heated to their own without decomposition. ." As an example. Sum. As a rule. The vapour pressures of aniline and water at 97 to 99 are as follows : TABLE 10. The aqueous layer. when there was still a moderate amount of water Praetieal Application. 97 98 99 40-5 42-2 44-0 682-0 707-3 733-3 749 "5 777 '3 The calculated boiling point under a pressure of 746*4 mm. magnetic oxide of iron and The aniline is distilled over with steam and the aniline. is afterwards placed in the boiler which supplies steam to the still. greater part of it separates from the distillate on standing.

16. into shown a 16. we should the vapour pressures of each substance at between their boiling points. mP _ Calculation from Vapour Pressure and Composi- In order to calculate the boiling points of all mixtion. and the curve drawn through the points will give us the required relation between . P= P 100? ~ P *B PA . and B rotating LIQUIDS MISCIBLE IN ALL PBOPOBTIONS of in The boiling point of a mixture two liquids which are miscible all if lated proportions can be calcuthe vapour pressure of the mixture agrees accurately with that given by the formula A +~(100 -m)PB p _ 100 FIG. not so simply related as their vapour pressures to the molecular composition. 760. however. must then be calculated at a series of temperatures between these limits. so calculated. Thus chlorotemperatures benzene boils at 132-0 and bromobenzene at 156-1.IV INFINITELY MISCIBLE LIQUIDS The apparatus is is 55 Fig. A the pipe by which steam is diagrammatically in is introduced the still. tures of two such liquids under normal pressure. say require to know = every two degrees. boiling points of such mixtures are. in this case. The percentage molecular composition of mixtures which would exert a vapour pressure of 760 mm. and we must be able to ascertain the vapour pressures of each substance between 132 and 156. must be mapped against the temperatures. Lastly. the molecular percentages of A. . by means of the formula ra where. Aniline still.

that the ratio of the . For such closely related sub- stances as chlorobenzene and bromobenzene.) of the other pure substance and of a series of mixtures to that of the standard substance at a single pressure.) of any given mixture to that of one of the pure substances at the same pressure is also a at all pressures it would therefore. strictly be sufficient to determine the vapour pressures of speaking. temp.760. PB - PA and 11. boiling point (abs.56 boiling FRACTIONAL DISTILLATION point CHAP. as is also the value of T dpfdt. TABLE . the ratios of the boiling points (abs. the ratio of the boiling points on the absolute scale is a constant at all equal pressures. also the PB . ra. and molecular composition under normal pressure. either substance through a wide range of temperature and constant . The vapour pressures of chlorobenzene and bromobenzene from 132 values of to 156 are given in the table below. in order to be able to calculate the boiling point of any mixture at any pressure. so far as experimental evidence goes. temp. and it appears. Closely Related Liquids.

17. The logarithms of the pressures were plotted against the temperatures. 140 Molrrulv ptnenage tf JSnirifttaune (-m) Pio.iv DETERMINATION BY DYNAMICAL METHOD 57 In the diagram (Fig. The boiling points of three mixtures of chlorobenzene and bromobenzene were determined at a series of pressures between about 690 and 800 mm. (2). Boiling points of mixtures of bromobenzene and chlorobenzene. differ sensibly from that of the original mixture. Determination by Dynamical Method. 17 by circles. 17) the values of ra have been plotted against the temperatures. 25 '01 136'75 50 '00 142-16 73 '64 148'16 in The points representing these values are indicated Fig. Experimental Verification. In determin- ing the boiling points of mixed liquids by the dynamical method. and the curve is drawn through the points which are not themselves indicated. as a rule. because if a relatively large of liquid were converted into vapour the composition of the residual liquid would. and the boiling points under normal pressure were read from the curve with the following results : Molecular percentage of bromobenzene Boiling point . A reflux condenser must be . and it will be seen that they fall very well indeed on the theoretical curve. it is of great importance that the vapour phase amount should be as small as possible.

that of a boiling liquid. in the bulb 1 . through which passes a rather narrower thin-walled tube. provided with a well-fitting cork. The quantity of each mixture placed was such that its volume at Fid. A It consists of suitable apparatus (2) is shown in Fig. and the bottom of the thermometer bulb was about level with that of the tube. volume of vapour in the bulb and vertical tubes was about 75 c. Boiling point the boiling point was about 125 c. for the temperatures of liquid and vapour is practically constant a given mixture. used and it is advisable that the thermometer should be shielded from any possible cooling effect of the returning The temperature of the vapour should be read.c. it is a good plan to read both temperatures at each pressure. to which is sealed a narrow side tube cooled by water to The upper end of the side tube is connected with an exhaust and compression pump and a differential gauge. being higher than " but as the difference between the " boiling point (p. to subtract the average difference between them from the temperature of the liquid. which has a hole blown in it just below the cork.58 FRACTIONAL DISTILLATION CHAP. 24). through which passes the thermometer. the apparatus for mixtures. capacity with a wide vertical tube.c. and to take the mean of the read temperature of the vapour and the reduced temperature of the liquid at each pressure as the true boiling point. This narrower is tube tube is also fitted with a cork. .c. liquid. a bulb of about 155 c. The thin-walled tube was pushed down until the bottom of it was about 3 mm. The wide vertical act as a reflux condenser. even when pure. Apparatus. in a given apparatus with a steady supply of heat. 18. above the surface of the liquid when cold. 18.

17). is exceedingly small. b. ~ m) * B it or. the temperatures being lower than if the line were straight. with the large quantity of liquid present and the small flame that is required. more conveniently. It is found 100 that the curve approximates more and more closely to a straight line as the difference. A.iv BOILING POINT AND COMPOSITION : 59 This arrangement possesses the following advantages a. If the relation between and molecular composition were represented boiling point by a straight line. . the liquid boils up into the thin-walled tube well above the thermometer bulb. It is possible to take readings of the temperature both of the vapour and of the liquid without altering the position of the thermometer. for when the burner is directly below the centre of the bulb. there is no fear of the vapour being super- heated. and the amount of liquid that condenses on the thermometer. The liquid that returns from the reflux condenser cannot come near the thermometer. = tB + *Wi - tB and is useful -LUO to consider the deviations of the theoretical boiling point curves from The boiling point of straightness (3). the boiling point of any mixture could be calculated 1 from the formula t t = (t WA * + ^9 100 ). between the boiling points of the components diminishes. the surface of the liquid immediately below that tube remains undisturbed and the liquid does not come'in contact witfrthe thermometer bulb. on the other hand. and that is the case for any other pair of liquids for which the ~ mPA + formula P = holds good. The relation between the boiling points and the molecular composition of mixtures of chlorobenzene and bromobenzene is represented by a curve (Fig. Calculated Boiling Point-Molecular Composition Curves. but when the burner is moved a little to one side.walled tube. and on the inner walls of the thin.

TABLE t 12. and ages. propyl acetate and tA <B = A = at substances and of methyl acetate 57-15. known.CO FRACTIONAL DISTILLATION is CHAP. * B -^ A the straight line temperatures for these molecular pert' centages calculated from the formula differences = tB + y^. . and 100' the d = t t' . The vapour pressures of both are temperatures between 57 and 102 101-58' 44'43*. m = in Table 12 are given the molecular percentm. of propyl acetate calculated from the formula IO ( P B ~ 76 ) for each fourth degree from 58 to 98.

where \ . -=. but which have different values -. the formula D T~> 2 = K' c ..1 i / .0-18 A)c B 100 V -2 4 ' d' -6 -8 ( . dp used.IV MAXIMUM DEVIATION" For any two of the 61 of four mono-halogen derivatives benzene. must be K= _ (50 / CB . of c i = .c A J.0-000158 and + 0-18A)c A + (50 . A 2 . the values of the D l agree well with those given by formula Z> 2 = Kb? where K is a constant and or T pairs of closely related liquids dt 1 \ . J. . .).

It will be seen from Table 13 that when C A /C B is less than unity and A is small. that D lf it membered when A has a low value. pressures are generally small when C A /C B does not differ greatly from unity they are invariably positive when C A/C B is very small and negative when large.62 FRACTIONAL DISTILLATION CHAT. As regards the maximum deviation. to 0-23. 7tt D1 between the values of m m calculated from the formula = 50 + 0'18 A and those deduced from the vapour . both must be D l and re- D 2 necessarily very small . TABLE Liquids in mixture. the 2 x relative as well as the actual differences should therefore and therefore D - D are be considered. in the case of ethyl alcohol and ?j-hexane. 13. . D l may have a small positive value amounting.

in other cases.. C' . 14. We must now consider whether the actual boiling points of mixed liquids agree with those calculated from the vapour pressures. in particular. A2 .^. when these are not available. and in the majority of cases within 1 whatever the value of C A/CB Comparison of Observed with Calculated Points.0-000168. has a value between 0'95 and 1-05 and probably . We TABLE Liquids in Mixture. and data are also available for four other pairs of closely related liquids (4). agree with those deduced from the vapour pressures (Dj). and are given in Table 14.iv OBSERVED AND CALCULATED BOILING POINTS 63 We may therefore ture deviation which conclude that the maximum temperawould be observed if the formula e calculated with 100 considerable accuracy from the formula 0-000158 . from those calculated from the formula D 2= . whether the observed maximum deviations (Z> 3 ). when when C A /C B the liquids are closely related or. or. c. . and. Boiling- Closely Related Liquids. CA T'. between 0-9 and 1-1. C B have already seen that the agreement is very good in the case of chlorobenzene and bromobenzene.

as a rule. the differences D3 V are certainly within the limits of experimental error in the other cases the differences are very small.64 FRACTIONAL DISTILLATION For mixtures . A 2 C A . may be calculated with fair accuracy from the formula D 0-000158. and with methyl and ethyl alcohol. in the following table (3). though occasionally higher. the boiling point and molecular bracket. are not closely related number of cases that it is 3 D As regards liquids which in a comparatively small only can be accurately ascertained. and these values are given under d and m. which show no change of temperature or of volume when mixed together. -CB 'A and that the maximum temperature deviation. Data.T'. . With chlorobenzene and bromobenzene. but there is evidence that the actual boiling points of such mixtures are. more or less complete. and that the formation of a mixture of constant boiling point is by no means an uncommon occurrence. -D vapour pressures by means of the formula m= * B ~ '. the difference 3 exceedingly small. considerably lower than those calculated from the vapour pressures. which have widely different critical pressures. the number is enclosed in a In a few cases. and it may probably be concluded that the actual boiling points of mixtures of closely related substances may be calculated with very considerable accuracy from the . c'. for eighteen liquids are given "Where the value of 3 can only D be approximately ascertained. of D methyl and ethyl alcohol. experimental data being either insufficient or altogether wanting. and have the same critical pressure.cB Liquids not Closely Related. D.D is also l CHAP. composition of the mixture of constant boiling point only can be ascertained.


of mixtures of The occurrence maximum or minimum . cit. According to Speyers (7) a mixture of minimum boiling point cannot be formed when both constituents have normal molecular weight at all concentrations. Again. hydriodic. a comparatively slight difference and l would be sufficient to account for the between 3 D D formation of a mixture of minimum boiling point. hydrochloric or the liquids are water acid with methyl ether (Friedel) (8) and an acid by chloroform and acetone (A = 4-8) and by chloroform and methyl acetate (A = about 4*3) the first observation has been confirmed by Zawidski and by Kuenen and Eobson (10). when A is very small. and ethyl (6) Zawidski has also shown that the and last of these pairs of liquids form mixtures of maximum vapour pressure. List of Known Mixtures of Constant Boiling: Point. boils O5 lower than ethyl bromide. containing 32 per cent. hydrobromic. and it is clear that. hydrochloric. Mixtures of of mixtures of Maximum Boiling maximum substances of is Point. . The occurrence vapour pressure) is cases one of the boiling point (or minimum comparatively rare . hydrofluoric. nitric or perchloric acid (Eoscoe) (9) but . propionic acid with pyridine (Zawidski) (6). first CHAP. and the second by Miss Fortey (unpublished).66 FRACTIONAL DISTILLATION acetate. It is probable that benzene and w-hexane (A = 11'3) and benzene and carbon tetrachloride (A = 3'44) form such mixtures.) states that a mixture of carbon di sulphide and ethyl bromide (A = about 7'6). cit. Zawidski (loc. and Eyland (loc. Eyland finds that such mixtures are formed formic. acetone and methyl ethyl iodide acetone and ethyl iodide.) finds that methylal and carbon disulphide (A = 4'15) form a mixture of maximum vapour pressure. acetic and . by weight of the sulphide. but this conclusion does not appear to be borne out by the facts. in most of the known base or a compound an acid and the other a basic character formic. acetate.

Minimum Boiling Point Binary mixtures.IV 07 boiling point is apt to give rise to mistakes in interpreting the results of a fractional distillation. . list of the cases which have so far been TABLE Mixtures of 16. and it may therefore be useful to give a observed. Substances In mixture.

CHAP. Substances in mixture.68 FRACTIONAL DISTILLATION TABLE 16 (continued). Mixtures of Minimum Boiling Point Binary mixtures (continued). .

69 Maximum Boiling Points Binary mixtures. Substances in mixture.MIXTURES OF MAXIMUM BOILING POINT TABLE 18. .

24. 15. 6. = Ryland. 425. loc. physik. = Linebarger.. Journ. G. Z. 35. phys. Soc. See also Hartman. 81. = Thorpe. 34. 1862.... 1903.. 1902. Th. 176. Soc. Soc. = Jackson and Young. Mag. 1879. iv " On the Composition of Aqueous Acids of Constant Boiling Point. 1883. 23. H. Compt. Chem... 14. Amer. Trans. Chem. 1862. Hoy. J.&W. No.. Soc. [V.. 77 and data not Zawidski. No.." Journ. loc. 146 . Trans. 17 TO 19. = Friedel. Chem.. Soc. 8. cit. = Young and Fortey. Chem." Trans. K. hitherto published. Journ. 1881. 73. 116. = Nernst. 90... 5. Y. 1902. 747. cit. "Observations on Mixtures with Maximum or Minimum Vapour Pressure.. 463.. 493. = Goldschmidt and Constam. Soc.. Trans. cit. Gu. Chem. 83.. 349. 4. Journ. 13." Proc. Ann. No. 1901.]. 23. 1901. Y. Chem. Soc. 9. L. 16. REFERENCES IN TABLES F. 10. loc. & C. &F. 8. 448. Chem. V. " On the First Plait in Van der Waals's Free Energy Surface for Mixtures of Two Substances. Wied. 2976. rend. loc. Ber. 3. = Vincent and Delachanel.. Amer. R. Phil. Ga. 1884.. 1880. Chem. 1861. Kuenen and Robson. 512. 1891. = Guthrie. Chem. Soc. & D. 18. Zeit. Ber. Soc. N. 1899. cit. . = Holley. 5.. 17. 270. 1587. 1902. = Noyes and Warfel. Chem. = Pettitt. = Roscoe. Trans. Ry. " On Perchloric Acid and its Hydrates. = Konowaloff. Ne. 1898. & Y. Chem. 1889.. 1895. No. CH. 129. Journ. FRACTIONAL DISTILLATION Roscoe. Amer." Phil. 717 and 739. .70 9. = Gardner. . P. 615 and 690. 544. phys. 11. = = Young. Mag. VI.

that the composition of the vapour is the the liquid. these components being either the original substances from which the mixture was formed. the composition of the residual liquid differs from that of the vapour. such as that over the mercury in a barometer tube. as a rule. It is only when the liquids form a mixture of maximum or minimum vapour pressure and therefore of constant boiling point and when it is this particular mixture that is introduced into the vacuous space. EXPERIMENTAL DETERMINATIONS Evaporation into Vacuous Space. the residual liquid will be much richer in the less volatile component than the original mixture. . partially miscible or non-miscible placed together in a vacuous space. boiling points of the two components into which the mixture tends to separate are not very close that of and together. or one of these substances and a mixture of the two which has a higher or lower boiling point than that of either of the original constituents. and. the composition of the residual liquid will differ only slightly from the original mixture. liquids When two volatile are miscible. evaporation takes place.CHAPTER V COMPOSITION OP LIQUID AND VAPOUR PHASES. In all other cases the vapour same as that is of richer in the more volatile of the two components into which the mixture tends to separate when distilled. If the volume of vapour is relatively very small. but if it is relatively very if the large.




Methods Employed. The difficulties attending the experimental determination of the composition of liquid and vapour are, in most cases, very considerable and, unless great care be taken, erroneous and misleading results may be obtained. The chief methods which have been employed
are the following

mixture of known composition is introduced into a suitable still; a relatively very small quantity is distilled over, and the composition of the distillate and, in some is determined either (a) from its cases, of the residue also specific gravity (Brown) (1), (&) from its refractive power (Lehfeldt) (2), Zawidski (3), (c) from its boiling point (Carveth) (4), or (d) by quantitative analysis; but for
organic liquids the last





method is not generally suitable. be carried out either in the ordiunder constant pressure, or at constant



A known


of air is passed

through the mixture

at constant temperature (5) ; the total amount of evaporation is ascertained from the loss of weight of the liquid, and

the weight of one component in the vapour



by quantitative analysis. 3. The relation between the boiling points and the composition of mixtures of the two liquids is first determined under constant pressure, and the curve constructed. A distillation is then carried out in such a manner that the boiling points of the liquid and of the distillate can be simultaneously determined (4). The composition of the mixture
of the distillate at different stages of the then ascertained from the curve. 4. A distillation is carried out with a still-head kept at a constant temperature (6). The composition of the distillate is determined in the usual manner; that of the mixture distilled is ascertained from the temperature of the stillhead, the boiling-point-composition curve having been in the still
distillation is


previously constructed.



First Method. Distillations under constant pressure have been carried out by Duclaux (7) and, with great care, by F. D. Brown (1).

Brown's Apparatus.

The apparatus employed by Brown is

" It consists of a copper vessel, s, sharped like an ordinary tin can, but provided with a long neck a. This neck and the upper portion of the vessel are covered

shown in

Fig, 20.

Fio. 20.

Brown's apparatus.

with a copper jacket,
inner vessel by
part of a.

c c c,



which communicates with the some small holes round the upper

This outer jacket

terminated below by a strip

of copper placed obliquely to the axis of the vessel, its lower portion is fitted with a narrow tube, d,

and at which

still with the condenser. The vapour from the liquid in the vessel s, passes through the holes at a, and then descending, passes out at d. The vapour as it rises is thus kept warm, and none of it is con-

serves to connect the






densed until

is of

has entered the outer jacket. Here a slight no influence, as both vapour and liquid

pass together into the receiver. The inclination of the bottom of the jacket serves to prevent the accumulation of any liquid at that part."
After heat has been applied, but before ebullition commences, a good deal of evaporation takes place and the mixture of warm air and vapour passes into the condenser, where most of the vapour is condensed. Brown, therefore,

only made use of the data obtained from the first fraction of the distillate to ascertain the composition of the residue in the still at the moment of change from the first to the

second fraction.

The form




in Fig.

20 was used in

order to avoid evaporation and consequent change in composition of the fractions, and also to allow of the distillation being carried on under reduced pressure. The quantity of liquid placed in the still was usually about 900 or 1,000

grams, and in each experiment about one-fourth of the total

amount was distilled over, and was collected in four fractions. The composition of the residual liquid at the end of the distillation was found in every case to agree satiswith that calculated from the composition of the mixture and of the fractions collected. original This apparatus was employed for mixtures of carbon disulphide and carbon tetrachloride, a preliminary series of determinations of the specific gravity of mixtures of these substances having been made, in order that the composition of any mixture might subsequently be ascertained from
its specific



similar series of distillations


out under reduced pressure (about 430 mm.), the apparatus being connected with a large air In the earlier reservoir and with a pump and gauge. with mixtures of carbon disulphide and experiments





forms of







Lehfeldt's apparatus (2) for disLehfeldt's Apparatus. at constant temperature is shown in Fig. 21.

The still, which is in the form of a large test tube, is provided with a cork perforated with two holes, through one of which passes a thermometer, the bulb of which is covered with a little cotton-wool, which dips just below the surface of the liquid. Through the other hole passes the delivery

FIG. 21.

Lehfeldt's apparatus.

tube, E, E, connected with the condenser F,




vided with a tap, G, for drawing off the distillate, and a tube connected with a pressure gauge, air reservoir and air pump. The bell-jar, j, contains either cold water or a

The still is heated by water in a large beaker placed on a sand bath; the water is constantly stirred and its temperature is registered by a thermometer. In order to prevent back condensation in the vertical part of the delivery tube, an incandescent electric lamp, with
freezing mixture.




the ordinary conical shade, is lowered as close as possible to the water bath, and a cloth is hung round the whole the top of the apparatus is thus kept at least as hot as the



quantities of material

employed by Lehfeldt were

the mixture to be investigated were placed in the stiU and three fractions of about 1 c.c. each

about 30

c.c. of

were usually collected and examined separately by means
of a Pulfrich refractometer.

Steady ebullition was ensured by placing a piece of pumice stone, weighted with copper wire, in the still, and the pressure was adjusted from time to time to keep the
boiling point as nearly constant as possible. Preliminary determinations of the refractive

powers of

mixtures of the liquids investigated were made.
Zawidski's Apparatus.
ratus which



employed an appa-


similar in principle, but more elaborate than is shown in Fig. 22. The still, A, of about

200 c.c. capacity, and containing in each experiment from 100 to 120 c.c. of liquid, is heated by a water bath, o, provided with a stirrer and thermostat. Back condensation is prevented by coiling copper wire round the upper part of the delivery tube, H, and heating this with a small flame. Steady ebullition is brought about by means of a fine piece of platinum wire, P, 0-04 mm. diameter, near the bottom of the still, connected to two thicker platinum leads, which
pass through the side tube opposite the delivery tube and are connected with a battery of three or four accumulators. The fine wire is heated by the current of about 0*4 ampere, and a steady stream of bubbles is thus produced. This
is also recommended by Bigelow (8). Theleceiver, the same form as Lehfeldt's, except that the tube below, instead of being provided with a stopcock, is bent, as shown in the figure, and is connected with a second small


B, is of

receiver, c, into which,

by diminishing the pressure

in the



reservoir, F, the first small portion of distillate, before the

temperature and pressure have become constant, is carried The distillate required for examination (about 1 c.c.) is then collected in B and, after admission of air, c is removed,

and, by slightly raising the pressure, the distillate out of B into a little test tube.



The arrangement of the manometer, the pump and the two air reservoirs, D and F, is shown in Fig. 22. By

Fio. 22.

Zawidski's apparatus.


lowered a

of the various stopcocks, the pressure in F can be little below that in D, and air can be admitted

into the apparatus, either through the stopcock 3 into D, or through the calcium chloride tube into c or B. A series of

determinations was carried out in the following manner About 100 to 120 c.c. of one of the two substances was placed in A, and the pressure under which it boiled at the






small quantity

of the second liquid was then introduced into A, and the first distillation at the same temperature, t, carried out, the

pressure being again noted. At the end of the distillation about 1 c.c. of the residual liquid in A was removed and

The error introduced by the incomplete condensation of the vapour carried over by the air will partially compensate that referred to under number 1. 4. placed in a small test tube for subsequent examination. on. . If a thoroughly dehydrated liquid is exposed to the air even for a short time. If. In case. and these operations were repeated until the mixture in A became rich in the second substance. go distillate will be too rich in the while the distillation is progressing. The series of operations was then repeated. Sources Of Error. may possibly more than counterbalance it. Partial fractionation will. as already mentioned. will The air which is in the still before the liquid is heated become saturated with vapour. starting with the second liquid in A and adding small quantities of the first. partial condensation of vapour will occur and a similar error to the 3. and as it passes through Brown's the condenser. and. and the first portions of more volatile component. if the top of the apparatus is heated before the liquid in the still is boiled. in fact.78 FRACTIONAL DISTILLATION CHAP. and the residual vapour will contain an excessive amount of the more : volatile component. first will be produced. the vapour which is evolved before the temperature of the upper portion of the still has reached the boiling point of the liquid will be partially condensed. the upper part of the still or delivery tube is exposed to the cooling action of the air. this premature con- densation of liquid was considerable in amount. 2. pa#k of this vapour will be condensed. The following sources of error should be noted and guarded against as far as possible 1. When a still is heated from below in the ordinary way. especially if it is poured from one vessel to another so that a large surface of it is so exposed. A further small quantity of the second liquid was then added. and a second distillation was carried out as before.

is probably the safest plan to adopt. and may probably be turbid. Suppose now that a mixture of benzene and carbon a mixture of first is being examined and that a minute amount water has been absorbed during the preparation of the mixture or in pouring it into the still. As a matter of fact. as described. The first portion of distillate will then contain the greater part of this water and tetrachloride of the specific gravity. it is probable that no two liquids are absolutely non-miscible. which are classed as nonabsorbed. series of experi- ments appear to be less accurate than the later ones. and the small portion of distillate will contain the whole or at least the greater part of the water. refractive index. the estimation of the composition of the first fraction. like benzene or the paraffins. containing a minute amount of dissolved water. boiling point and other physical properties of the distillate will be appreciably A considerable error may thus be introduced in altered. appreciable quantities of water.v a certain SOURCES OF ERROR amount It is 79 invariably of moisture will almost be not only liquids which are regarded as hygroscopic which thus absorb moisture. all commonly occurring liquid organic compounds can dissolve miscible with water. the first results in some of his. is heated. minimum boiling point is first formed. and probably of greater influence than is generally recognised. and certainly. Zawidski first rejected the small portion of distillate and avoided the of error. but in spite of this. it is to dry the apparatus as completely as possible. and to keep . but even substances. partially at and second sources any rate. Brown and that first rejected the data derived from the first distillate. Lehfeldt guarded against the first and second sources of error by heating the whole of the upper part of his apparatus by means of an incandescent lamp. When such a liquid as benzene or carbon tetrachloride. by heating the delivery tube in the manner described. It is difficult entirely to avoid the fourth source of error. All that can be done is to dehydrate the liquids.

In Linebarger's experiments a known volume 1 to 4 litres) was passed. through the liquid mixture (40 80 grams) contained in a Mohr's potash apparatus consisting of five small and two This apparatus was completely immersed in a large bulbs. suitable water bath. . and-they number . Second Method. Gahl (10). position of liquid air Determinations of the relative comand vapour by passing a known volume of at constant (9). The total quantity of liquid evaporated. of it in the vapour was estimated by absorption in potash or barium hydrate. One pair of liquids investigated by Linebarger was benzene and carbon tetrachloride the results obtained appear improbable and they differ widely from those of Lehfeldt (2). and of Young and Fortey (11). which are . of Zawidski (3). appear to be more accurate at low temperatures than at higher in the two cases in which such a comparison was made. the temperature of which was kept constant within 0'05. was ascertained by weighing the bulbs containing the mixture before and after the experiment.80 FRACTIONAL DISTILLATION CHAP. and when one component conthe When amount tained sulphur or a halogen. bromobenzene and acetic acid than with others. through a mixture carried out by Winkelmann temperature have been Linebarger (5). Preliminary determinations of the vapour pressures of a of pure liquids were made in order to test the accuracy of the method the results were more satisfactory with the less volatile liquids such as chlorobenzene. at the rate of about 1 litre per hour. the liquids as far as possible out of contact with moist air but in any case it is best always to reject the first small portion . of air (from and others. one of the components of the mixture was an acid. rarely more than 2 grams. of distillate. the process of analysis by means of soda-lime was adopted.

therefore. or that partial condensation occurred before the mixed air and vapour was analysed. The thermometer in A B is so placed that no part of its Q to . Third Method. Through the drawn-out portion of the bulb B is sealed a rather heavy platinum wire is is G facilitated promote ebullition of liquid. apparatus (Fig.CARVETH'S METHOD 81 themselves in good agreement. provided with a reflux condenser and a thermometer. 23) It is possible that the employed by Gahl give better results. a round-bottomed litre-flask.c. After the completion of the boiling point determinations a third hole is made in the cork of the litre-flask and through it tube A B (Fig.c. would Fia 23 . The glass tube E serves to catch the condensed liquid from the reflux condenser and to deliver it through a small opening at the end of the tube into B. ployed by Carveth This (4). being used. copper or platinum in the bulb. must be received with so far as pure The tendency is apparliquids are concerned for the vapour pressures to ently be too low.. about 8 c. The length of this passed a glass about 18 cm. and the capacity of the bulb up to the siie opening. method was devised and em- The boiling points of a series of mixtures of the two liquids under investigation are first determined. the diameter 20 mm. for mixtures of volatile liquids the method cannot be regarded as satisfactory. The quantity of liquid should never be less than 400 c. and this may be further by placing tetrahedra of silver. It is to be feared. D. and even in other cases that the results caution.. and it may be that the air was not completely saturated with vapour. 24)._ Gahrs apparatU8 .

steady ebullition might be ensured by means of electric heating (p. A Carveth points out. 76). is immersed in the liquid in the outer which vessel depends on the difference between the boiling points of the solution and of the condensed vapour if this difference is less than 1 the liquid in B cannot be . 24. that of the residue being read at the same time on the thermometer in the outer bulb. and the boiling point of this distillate is read on the thermometer FIG. made to boil vigorously . It is assumed that the liquid which collects in B has the same composition as the vapour evolved from the liquid in the outer flask . and the composition of liquid and vapour is stated to be thus determined. It was found that the boiling point of the liquid in AB. as Carveth admits. and it is therefore clear that a definite state of equilibrium is attained.82 FRACTIONAL DISTILLATION is bulb the bulb B The depth to as high as the opening D. In an actual -experiment the liquid formed by condensabut. but. in A B. after becoming constant. The composition corresponding to each of these two temperatures is then ascertained from the boiling point-composition curve. where it boils vigorously. as tion of vapour in the re- flux condenser is returned (wholly or in part) to the bulb B. rarely varied 005 in the course of an hour. the condensed . Carveth's apparatus.

Carveth's boiling points are higher (allowing for a want of agreement between the observed boiling points of pure carbon disulphide). The two errors compensate each other to some extent. In the case of mixtures of carbon disulphide and benzene Carveth's boiling point-composition curve is distinctly flatter than Brown's. is half as large the liquid. the boiling pointa of of these two liquids. the volume of vapour minimum amount again as that of of liquid. Carveth made three series of determinations with mixtures In the first. and the liquid formed in the condenser must also be too rich in the more volatile component. As regards the preliminary determination of the boiling points of mixtures of the two substances. more volatile vaporisation does occur. Comparison of Brown's and Carveth's Results. an excess of component must be removed and the if must therefore be raised. is exposed to a higher temperature than its boiling point. but the assumption that this compensation is perfect seems to be rather a bold one.v RESULTS OF BROWN AND CARVETH COMPARED 83 while passing from the condenser to B. the other hand. as the liquid in B is richer in the more volatile component than that in the outer flask. The vapours mix together in the outer flask and the boiling point On mixture passes into the condenser. The vapour as a whole must therefore be somewhat richer in the more volatile component than that evolved from the liquid in the outer flask. the vapour from it must also be richer in that component. the But.c. and it seems difficult to believe that no vaporisation takes place during liquid. 400 c. it may be pointed out with the that. a 2 . the descent. in other words. in the apparatus used. With two substances of widely different volatility the residual liquid would certainly be very appreciably poorer in the more volatile component than the original mixture placed in the flask and the observed boiling point would therefore be too high.

the boiling points of liquid and condensed vapour were determined in the improved form of apparatus. would make In the third series. as will be seen from Table 19. but are much lower than those found by Brown by Method I.84 FRACTIONAL DISTILLATION CHAP. The relation between the composition and the boiling points of mixtures was determined only in series I. the first series. In series II and III the differences between the boiling points of residue and distillate agree very fairly well together. but an earlier form of apparatus was used in which there was more probability of re-evaporation of condensed liquid taking place. except for two observations in II. . In the second series the mixture was actually distilled. It is evident that either Brown's or Carveth's results in favour of Carveth's must be inaccurate. liquid and condensed vapour were determined for a series of mixtures of known composition. determined. the results deduced from series II and III are considerably higher than from series I. where there are differences apparently two mistakes of a whole degree each but these do not agree at all well with those derived from . and the mean being taken as the boiling point of the liquid corresponding to that fraction. which the boiling points of the distillates too high. Carveth admits that there may have been a little loss of carbon disulphide by evaporation. may be men- tioned the fair agreement between the results of the second and third series (which are not given separately in Table 19). condensation in the ascending part of the delivery tube being prevented by means of an electric coil surrounded by The boiling points of the distillate were separately asbestos. Calculating the molecular percentages of carbon disulphide in the vapour corresponding to definite percentages in the liquid. those of the mixture before and after the collection of each fraction being noted.


The composition of the mixtures which would evolve vapour of the same composition as the distillates obtained with the still-head of constant temperature 2. are shown in Table 21. . and also the relations between boiling point and for these From the curves composition of liquid by Method I. Mixtures of carbon disulphide with carbon tetrachloride and with benzene were specially examined and. The actual numbers are given in Table 20. Special experiments further to test the truth of this state- ment were made with mixtures of ethyl alcohol and water. 2. agreed closely with the temperatures of the still-head. liquid (p. The agreement is sufficiently close with both pairs of liquids to allow of the statement being made that the temperature of the still-head is. of the distillate was independent two pairs of liquids. He found that these boiling points. The boiling points of these mixtures. the distiUation was continued until not more than an occasional drop of distillate fell this may probably account for the larger differences between the two . but in this case the comparison was made by determining 1. . and depended only on the temperature of the stillhead. The composition of the distillates when the still-head was kept at 81-8 and 86*5 respectively. these relations he read representing 1. Brown had previously determined the relations between the composition of liquid and vapour. read from the curves. approximately at any rate. observed that the composition of that of the mixture in the still. would give a distillate of the same composition as that actually collected. temperatures. when distilled as in Method I. the same as the boiling point of a mixture which. 181). The composition of the distillate from mixtures which The results obtained boiled at the same two temperatures. In the last four determinations with carbon disulphide and carbon tetrachloride.86 FRACTIONAL DISTILLATION Brown (6) CHAP.


of benzene.]. Chem. 304. Liiiebarger. Phys Chem. the other hand. this method. Soc." Journ. as mixtures of normal hexane and benzene containing." Phil. ( Liquids. Zawidski. 6. 8. and it is may is very satisfactory. 129.. 46." Trans. 39. 547 1898. Part II.." Ami. be used to determine the relation between the composition of liquid and of vapour if the boiling point-composition curve has previously been constructed. 1878. Properties of Liquid Mixtures.. Chim. 5. REFERENCES. [V. 615.. " Lehfeldt. Phys. 1881. The first part of the distillate might be altogether rejected. Chem. for example. . Chem. 3." Ze. 35. like Carveth's. 7. "The Composition of Mixed Vapours. Brown. " On the Vapour Pressures of Binary Mixtures of 4. say.. 1. 1899." ibid. Chem.. Amer. physik. Simplification of Beckmann's Boiling Point Apparatus. 2. Mag. 1900. 37. The method would certainly possess the following advan: tages 1. Vapour given off by a Mixture of Two Liquids.. Carveth.. 14. of Fractional Distillation. Soc. from 1 to 20 per cent. such. 42." ibid. [5]. It would not be necessary to determine the composition of the mixture in the still. "Fractional Distillation with a Still-head of Uniform Temperature. 1881. 1879. 517. 1899. " Tension of the Duclaux. Further experiments are desirable to test its applicability and the accuracy attainable.. 193. F. Brown. 305. " On the Distillation of Mixtures of Carbon Disulphide and Carbon Tetrachloride. D. 3." Journ. and the errors already referred to thus avoided ." Trans.it. "The Vapour Tensions of Mixtures of Volatile Liquids." Amer. Journ. "A Bigelow. 280. " Theory . 2. Soc. 39. would not be suitable for mixtures the boiling points of which vary On very slightly with change of composition.88 It will FRACTIONAL DISTILLATION be seen that the agreement evident that this method CHAP. 35. 17. Chem.. 1880. 49. 1895. " The Comparative Value of different Methods of Fractional Distillation..

the vapour pressure and the boiling point can only be accurately calculated from the vapour pressures of the components if (a) all the liquids are non-miscible. Chem." 10.. 39. "Studies on the Theory of Vapour Pressure. A and pressures D . CONSIDEBED THEOBETICALLY Simplest Cases." Trans. for the composition of the vapour like the vapour pressure and the boiling point is independent of the relative quantities of the components.. Part II. B we shall have. CHAPTEE VI COMPOSITION OF LIQUID AND VAPOUB PHASES. or (&) they are miscible in proportions and show no change of temperature or volume when mixed together. 89 9. this being generally the case when the substances are chemically closely related. A similar state- ment may be made with regard to the composition vapour evolved from the two liquids. It has been seen that when two liquids are placed together in a closed vacuous space. provided that they are both present in sufficient quantity and that evaporation can take place freely the composition of the vapour can be calculated if the vapour pressures and vapour densities of .. Young and Fortey. 1903. 33. 1890. P P . of Liquids. A and Z> B and the vapour Calling the vapour densities at the temperature t. the non-miscible liquids are known. 11. 1900.." Zeit. NON-MISCIBLE LIQUIDS of the The simplest case is that in which the two substances are non-miscible. "The Vapour Mixed Liquids. 179. Ann. 1. 45. " On the Composition physik. Pressures and Boiling Points of Chem. .vi NON-MISCIBLE LIQUIDS Winkelmann. Soc. of the Vapour from Mixtures Wied. 83. Gahl.

Brown (2)]. litre of CHAP. be 37 273 x X =7jx <OU PB respectively. Temperature. As an example we may again consider the case of chlorobenzene and water. . The masses vapour will x I) therefore be D** and * PA and the relative 0-0899 x D* B T . Chlorobenzene and Water. TABLE 22. t in a litre of the 1 litre of A at pressure and 1 of B at t ' and 89 ( PB mm. masses will D A PA [Naumann(l). * and PA mm.90 FRACTIONAL DISTILLATION mixed vapour. the vapour pressures of which at 90 to 92 are given below.

191. when the distillate contained 19-5 per . which This would give 71*4 per cent. of aniline and water was distilled under a pressure of 746'4 mm. while the values observed were 18-7. 20-4. tillation had been collected the aniline was in great a large quantity of water was added and the discontinued. boiling point. calculated The percentage on the assumption that the liquids are non-miscible. of and Water.c. but chlorobenzene flows much more freely. 20-4. value.c. would be 23*6. 20-1. and vapour composition can be calculated in a similar manner when more than two non-miscible liquids are present. PARTIALLY MISCIBLE LIQUIDS two liquids are miscible within limits. the distillate fractions. the observed pressure.vi PARTIALLY MISCIBLE LIQUIDS was collected in five 91 about 3 grams of chlorobenzene and 40 of water remained in the flask. which were found chlorobenzene 1 : to contain the following percentages of 2 3 4 5 72-5 71-5 72-0 70-4 71-7 Mean 71 '6 Calculated 71 '2 Water is apt to adhere in drops to the walls of a glass tube. Aniline 100 c. and it would be fairer to reject it and to take the mean from the other four fractions. chlorobenzene. After the sixth fraction excess . the first fraction is therefore certain to contain too little water. 19-7. boiling point and vapour composition will vapour differ from those calculated in the manner described. of aniline.(3) A mixture of 50 c. of agrees better with the calculated The vapour pressure. but If the the difference will be small if the miscibility is only slight.

as has already been pointed out. or 4-3 higher (Young and Fortey. From Konowaloff s data (4) for the vapour pressures of isobutyl alcohol.92 FRACTIONAL DISTILLATION CHAP. 19'9. of water. is more marked. while that corresponding to the actually 66-8. At 18 one part of alcohol dissolves in 10-5 of water. would be 85 -7 if the liquids were non-miscible. 89-8). mixture in the still. but that. Again. the boiling point of the alcohol and water. and the vapour composition with accuracy from the vapour pressures of the components. that minimum 7'7. when no ap- . > The composition of the distillate is clearly within the somewhat wide limits of experiof that of the mean percentage of aniline. but the solubility diminishes with rise of temperature. the minimum boiling point observed by Konowaloff was actually 90'0. the difference between the observed and bility calculated values. saturated solution of water in the alcohol contains 15*2 per cent. cent. when distilled together under normal pressure. only. but the is 3'7 lower than that When the mutual soluIsobutyl Alcohol and Water. is greater. Thus isobutyl alcohol and At a water are miscible within fairly wide limits. INFINITELY MISCIBLE LIQUIDS In the case it is of liquids which are miscible in all proportions. then. boiling point is heated together we can when two non-miscible liquids are calculate the vapour pressure. and the solubility increases as the tempera- ture rises. the calculated percentage of isobutyl alcohol in the vapour would be 74-5. the calculated values differ from those actually observed. the boiling point. if the liquids are miscible within limits. as regards both temperature and vapour composition. of aniline. reaching a minimum at about 52. independent mental error calculated. the difference being It is clear. the difference increasing with the mutual solubility of the liquids.

Thorpe (7) observed that carbon tetrachloride and methyl alcohol form a mixture of minimum boiling and that. obtained with the formula A better result was . 100 Formula of Wanklyn and Berthelot. Wanklyn (5) and Berthelot (6) arrived independently at the conclusion that the composition of the vapour depends (a) on that of the liquid. In 1863. Calling x l and x 2 the relative masses of the two suband their relative masses in stances in the vapour. The relation between the preciable composition of the mixed liquid and that of the vapour evolved from it appears also to be a simple one only when the vapour pressure of the mixture is accurately expressed by the formula P= **P* + (IM -*)*. 2 The subject was investigated exF. ^1 = #2 W i. D l and D z their vapour densities and Pl and P2 their vapour pressures. for this particular mixture. Brown in 1879-1881 (8). the factor * * t JJ 2 P is approximately equal to unity. on the vapour densities of the components. D. not be accepted as generally true. and he perimentally by found that the Wanklyn-Berthelot formula could certainly Brown's Formula. when point.vi INFINITELY MISCIBLE LIQUIDS 93 change of volume or temperature occurs on admixture that the vapour pressures and the boiling points can be accurately calculated. (b) on the vapour pressures components at the boiling point of the mixture. Many attempts have been made to find a general formula to represent the relation between the composition of liquid and of vapour. l 2 the liquid. the formula would be of the pure (c) W W In 1879.

m)PB 100 ~' . perhaps. for not only does Brown's formula hold good. evidence. it is found that the closer the approximation of the pressure-molecular composition curve to straightness. but the evidence is certainly strong.94 FRACTIONAL DISTILLATION CHAV. Brown's formula. Applicability of Formula. for the two pairs of liquids investigated for which the relation between vapour pressure and molecular composition is represented by a straight line. ethylene and propylene dibromides and benzene and ethylene dichloride. In the following tables are given the observed vapour pressures and those calculated from the formula p_ mPA + (100 . c. but also. which has so far been obtained. in other cases. was substituted for the ratio PJP2 . The number of such cases so far investigated is. The strongest evidence is afforded by two pairs of liquids. investigated by Zawidski (9). in the value of the Zawidski's Results. within the limits of experimental error. the smaller is the variation " constant " c. 100 is ho ia 3 good. points to the conclusion that Brown's law is true for those infinitely miscible liquids for which the relation two which form a mixture The experimental P _P* + (100 -")?. however. and it is obvious that it cannot be true for liquids of constant boiling point. is all not generally applicable to liquids which are miscible in proportions. too small to allow of the statement being definitely made that the law true in such cases. but the agreement between the calculated and observed results was still closer when a constant.

of course. Molecular percentage of ^in liquid. A = Propylene dibromide . . B Temperature = 85 '05 . masses or gram-molecules. c B/PA at 85 = 1'357.VI APPLICABILITY OF BROWN'S FORMULA TIT 95 also the observed molecular percentages of A in the vapour and those calculated from the formula [In this formula x and may. ethylene dibromide .] B XA. represent either TABLE 23. = 1 '31 P . = c W WA ?.

A = Molecular percentage of Temperature =49 '99 Ethylene dichloride . whereas. in this case. the calculated values of c would be low does not at each end of the The value ratio of the of c. p= good wifch considerable accu racy for mixtures of benzene and ethylene dichloride. This series of experiments is of special interest because it is the only one in which the two liquids are very closely related. A in liquid. . differ greatly from the vapour pressures. found that the relation mPA + (100 . and the data for this pair of liquids are therefore given in full. c 1 '134 1-135. A = to A = in the nature of a steady rise or fall 100 per cent. .FRACTIONAL DISTILLATION It will CHAP. 1-357. be seen later that c. = .. TABLE 24. however. it is when there is a real variation in the values of from table.m)P B hdd . 1-31. B= Benzene PK/PA at 50 = . Zawidski.

Molecular . It seems clear that little weight can be attached to the first c has almost exactly the same value. chlorobenzene and benzene. For this pair of liquids the constant 1-135.VI LINEBARGER'S RESULTS 97 agreement With the exception of the first two percentages the is good. for Linebarger found that there was no appreciable heat change on mixing these substances several different proportions. A = Chlorobenzene . . as the ratio of the vapour pressures. in TABLE 25. 7 -16 at 34 8. Linebarger's Results. It is unfortunate that the method employed by Linebarger air through the mixture passing a known volume of (10) at constant temperature did not give more trustworthy results. and in -the case of these two it will be seen that the percentage of A in the liquid is the same. It may be well to give the results obtained with one pair of liquids. t 34 '8 . B= = Benzene c = 7*3 ' . 1 134. for he examined some mixtures of fairly closely related substances. - two observations in this series. while the observed percentages of A in the vapour differ by 1-2.

. of Zawidski. show that the vapour pressures of mixtures of toluene and carbon tetrachloride are slightly lower than those calculated from the formula P= mI>A + 100 ~ m)P s . 3-33. TABLE 26.98 FRACTIONAL DISTILLATION The experiments of Lehfeldt (11) Lehfeldt's Results. B = Carbon tetrachloride c = PB/PA at 50 Molecular percentage of liquid. and of Young and Fortey show that the vapour pressures of mixtures of benzene and carbon tetrachloride are somewhat higher. while those of Lehfeldt. For toluene and carbon tetrachloride a small but distinct variation. c appears to be a is constant. = 2'76 . t = 50'0 . Am . but with benzene and carbon tetrachloride there A = Toluene .

. B= Carbon tetrachloride c = 0'984 + 0-003m t"= 50-0 . 99 A = Benzene . Molecular percentage Vapour pressures. .vi MODIFICATION OF BROWN'S FORMULA TABLE 27. . PB/PA at 50 =1-145.

benzene and . which have been investigated by Brown. not at constant temperature. the latter also by Carveth. Brown and W^W^ observed. the pressure. it is not possible to give tables in the same form as before. constructed on the assumption that the pressure formula holds good. or vary only slightly with. or under atmospheric If the values of c for other substances also are pressure.. however. As the vapour pressures of mixtures carbon disulphide with carbon tetrachloride and with benzene have not been determined at constant temperatures. but an estimate of the relative maximum deviations of the vapour pressures of mixtures of these substances from those Indirect Evidence. of calculated from the formula P= A_ " mP ) may be formed from the boiling point curves. 65) that the maximum deviations. carbon disulphide with carbon tetrachloride and carbon disulphide with benzene. independent of. for the three pairs.100 FRACTIONAL DISTILLATION CHAP. they can be calculated from the results of a distillation carried out either at constant temperature or under constant pressure. from the straight lines given by the formula D t = *__i - _Z L? are. v of the theoretical boiling point curves. but we may at any rate assume that the differences would not be great enough to invalidate the general conclusions deduced from a comparison of the behaviour of different pairs of liquids. It has already been stated (p. but their distillations were carried out in the usual manner under constant pressure. that the same values of c would be obtained in other cases by both methods. It would not be safe to conclude. without further evidence. that the relation between xjx^ was the same for mixtures of carbon disulphide with carbon tetrachloride whether the distillation was carried out under a pressure of 432 mm. calculated from the formula P= mI>A + ~ ^ Ps there are two.

and it follows from this that the vapour pressures of mixtures of benzene and carbon tetrachloride agree most closely with those given lines are differences. . by the formula gence In is P=n for -^. and it will be seen from the table below that the order of the three pairs of as to a. all three cases the values of c may be expressed with accuracy by the formula c = C + am.2-5. while the maximum deviations.vi BROWN'S MODIFIED FORMULA 101 D carbon tetrachloride and carbon tetrachloride. for the three pairs of liquids. s l liquids is the same with regard to fair D D TABLE Substances in mixture. therefore.4-0. and . .8*0 respectively. .0-02. and . . . 28.Dv are. . benzene and carbon disulphide. .O89".4*02 respectively. The D3 . and of the actual boiling point curves from the same straight s carbon .5*8.~ 100 * while the diver- greatest mixtures of benzene with carbon disulphide.0-91.3-34 disulphide. .

B= methylal . t = 35-17.102 FRACTIONAL DISTILLATION TABLE 29. . CHAP. A Molecular Carbon disulphide .

for alcohol. B= Benzene .VI MATHEMATICAL INVESTIGATIONS which have 103 example.0-13 case. Ethyl alcohol . A= Molecular -D= -11-1 TABLE 31. At 50 the vapour pressure of benzene is higher than that of alcohol. Under atmospheric pressure l = . D . D 3 = -11-2. . but at 80 it is the alcohol which has the higher vapour pressure. t = 50. with mixtures of benzene and ethyl also been investigated by Zawidski.

later. q the ratio in the liquid. A and B ponents. In this connection "The Phase Kule. Lehfeldt's Formula.x) their molecular fractional amounts in the liquid mixture. taking the normal molecular weights as correct. and x and (1 . Lehfeldt (14). arrived at by written : The formula Duhem. Lehfeldt gives the equation : logS.log r = __^ . K and r are For r. . may also be consulted.104 FRACTIONAL DISTILLATION CHAP. or where t = the ratio of the masses of the two substances in the vapour.x) where p l and p 2 are the partial pressures of the vapours of the two liquids A and B. "> where TTA and TTB are the vapour pressures of the pure liquids at the temperature of experiment S is the ratio of the number of molecules of the two substances in the vapour . logt = log-BT+ rlogq t = Kg*." by Bancroft (15). may be dlogx dlog(l . but Lehfeldt finds that Kq r the equation holds very well for mixtures which do not . the equation leads to infinite values of dp/dx when x Q or x = l. are the normal molecular weights of the comthat. Formula of Duhem and Margules. . Starting from this equation. which does not agree with the facts. Margules has pointed out t when r < 1. and Zawidski (9). and. Lehfeldt adopts the formula. and constants. by Margules.

log$ [In Tables 32 and 33 the molecular composition instead of the composition by weight. provided that the molecular weights of both substances are in the liquid as well as in the gaseous state.VI LEHFELDT'S FORMULA 105 contain a very small proportion of either component. for = 0-065 + 0-9471og?. we consider the case of mixtures of benzene and carbon which Lehfeldt gives the formula : tetrachloride. may As an example. For the formula does not hold good at all. . Molecular percentage of A (Benzene).] is given TABLE 32. associating liquids normal Benzene and Carbon Tetraehloride.

. p. TABLE Molecular percentages of 33. examined by Zawidski. Lehfeldt himself gives for mixtures toluene and carbon tetrachloride the formula J = 0-440 + 1-0 log?. we may take carbon disulphide and methylal. the molecular weights of both of which are presumably normal. but which form a mixture of maximum vapour pressure considerably higher than that of either component. or = 2-7552. which agrees almost exactly with that given on ^? 98. A (Carbon disulphide). As an example of two substances. = XA 2-76 2? W A Carbon Disulphide and Methylal. In Table 33 are given the observed molecular percentage of carbon disulphide and those calculated by means of the formula log* = 0-036 + 0'6191ogg.106 FRACTIONAL DISTILLATION of law holds good.

and P2 where a2 and a 3 are constants. for mixtures of liquid oxygen and nitrogen. Lehfeldt's formula gave very good results. equations pressures. from the tangents where x = and x = 1 by means curve at the extreme points . the vanish. the relations of the Zawidski adopts the following between the partial vapour pressures. P p and p z : and molecular fractional components in the mixture. As regards mixtures of ethyl an associating substance. Lehfeldt points out that the relations between log t and log q are far from linear. Baly (16) found that. constants points out that if. = P i/ P2 */(!-*). the values of which can be ascertained from the partial pressure curves or. alcohol. and the amounts. in his formula. substances in the mixture. x and 1 . but Brown's did not. of the equations : _ Zawidski a and /? Relation between Brown's and Zawidski's Formulae. the of the pure components. to l express . and he did not attempt to find an equation for the curves. of the two P2 . Zawidski's Formula. the equations become whence or p l = P^x and p 2 = P2 (l x).x. 2>i/P* = const X /(!-). Associating Substances.vi ZAWIDSKI'S FORMULA 107 Oxygen and Nitrogen. with benzene and *vith toluene. by Marto the total pressure gules's method. pi/P-2 - .

Benzene and ethylene were found to have finite values. Zawidski shows that this simple relation holds for mixtures of ethylene and propylene dibromides and of benzene and ethylene dichloride. Zawidski shows.x) were not found to be applicable the relations between the molecular composition and the pressures. even with benzene and carbon tetrachloride the curvature.x). the constant. though slight. not . Thus. fact.UU B } but also of the formulae p = P^x and p 2 = P2 (l . though the agreement is excellent second pair (0-880 and 0-881). for substances of and normal molecular weight. differs somewhat widely from the ratio -P1 /P2 for the 0'737. 26. A In the case of other mixtures of liquids with normal molecular * weight. simply expresses Brown's law. mistakable. taking the relative number of molecules instead of relative masses for both liquid and vapour. the constants a and ft 100 Fio.108 FRACTIONAL DISTILLATION since. the equation.2 = P2 (l . 26) which A* Total . l as will be seen from the diagram (Fig. and the simple formulae p l = P^ and p. 0-758. is undichloride. the partial pressures are proportional to the number of molecules present. represented by curves. for these two pairs of substances. only the correctness of the relation which is equivalent to P= ^ - ^ ~ m' J. in its final form. in . whether total or partial. but that for the first pair of liquids. are. p is taken from his paper. .

there pressure containing little benzene. so as to compare the calculated values with those obtained by means of Lehfeldt's formula. and there is thus additional evidence (pp. Now the first is the value of the tangents (d-n-/dx) l calculated from constants would be . but for cent. better results might be obtained. For this and other pairs of liquids of normal molecular weight. so that in either case a negative value for the tangent when a. In Table 34 (p. = l.vi ZAWIDSKI'S FORMULA 109 Benzene and Carbon Tetraehloride. a 3 = -0-0168. (dir/dx^ = . Zawidski of included in his substances. From the tangents (d-n-ldx) Q = 90.4-3. that by altering the constants. however. .2 = 0-308. a. For this pair of Zawidski gives the constant a. a. 99 and 105) that these two liquids can form such a mixture. If must be a mixture of maximum vapour there this is correct. From the partial pressure curve. = 0-312. of which the molecules of one component are associated in the liquid state (water). though the difference between the maximum pressure and the vapour pressure of carbon tetrachloride is probably too small to be determined by direct experiment.5-0. The agreement up to 50 molecules per disulphide is excellent. Carbon Disulphide and Methylal. 2 II. though there are occasionally differences amounting to 4 or 5 per cent. or in both the liquid and gaseous states (acetic some pairs . . of carbon mixtures richer in that . component it is not nearly so satisfactory it is probable. 110) are given the results for carbon disulphide and methylal. Associated investigation Substances. a 3 = 0-00733.2 and a 3 calculated by both methods they are liquids I. the agreement between the observed molecular composition of the vapour and that derived from the calculated partial pressures is fairly good.

.110 FRACTIONAL DISTILLATION TABLE 34. CHAP. From (dw/dx) Q A = = + Carbon disulphide B = Methylal. . 578 and (dir/dx^ = .1310 a^ = 2'9 . o3 = - 1 Molecular percentage of Carbon disulphide. .

5.vi GENERAL CONCLUSIONS . even approximately. and it appears to be components necessary to determine it experimentally. 6. Brown's formula is only applicable when the vapour pressure of p = mP when any mixture B> is . given 8 by the formula A + (100-m)P and thig probably neyer the cage there is a marked volume or temperature change on mixing the pure liquids. As regards substances which are not closely related. the values of the constants a 2 and a3 being ascertained from the pressure-composition curve. 111 c. 3. 4. by means of Zawidski's formula unless the average molecular weights of the associating substance under the varying conditions of the experiment are known. x 1 /xz = cWl /W2 The value differ greatly from the ratio of the constant. the relation between vapour pressure and composition must be determined experimentally at the required temperature. does not of the of the vapour pressures at the temperature of experiment. but the data at present available are insufficient to warrant the statement that it is always equal to this ratio. . If the vapour pressures at constant temperature of mixtures of two infinitely miscible liquids the molecular weights of which are normal are not given by the formula P B . The composition of the vapour from any mixture may then be calculated with fair accuracy by means of the formula adopted by Zawidski (p. 107). When the molecules of either liquid are associated. a sufficient number of determinations of the relative composition of liquid and vapour at constant temperature have beentiade to allow of a curve being constructed the molecular percentages of the vapour being mapped against those of the liquid x^x<> against the logarithms of these ratios against each other. . for two miscible liquids. the relation between the composition of the vapour and that of the liquid cannot be ascertained. If.

a modification of Brown's formula . Brown. however. in the case of carbon tetrachloride and carbon disulphide. is not the same at different temperatures. were carried out in the usual manner. PJP^ would be a constant for a distillation at constant temperature. If. x l /x. be used. to in a distillation under constant pressure. however. that when a mixture was boiled the composition of the vapour was independent of the pressure under which ebullition took place. in the case of Brown's distillations. a curve constructed from results obtained at constant temperature could be used to ascertain the vapour composition It is. and this may be used even when the observed vapour pressures differ somewhat considerably from the calculated. or the constants for an interpolation formula may be calculated..112 or the FRACTIONAL DISTILLATION CHAP. the logarithms of the ratios of the masses of the components in the liquid and vapour phases had a linear always distilled whose distillations W . differences are great. vapour pressures of mixtures of the two substances differ but little from those given by the formula If the P= may ~ ^ B . however. the investigations of the relation between the composition of liquid and of vapour have been carried out of Most at constant temperature. these Zawidski should be employed. be noticed that the ratio of the vapour pressures. At any rate Lehfeldt found that. if this were generally true. Better results are. partial pressures of each component separately the molecular fractional amount of one of them against other values may be read from the curve. the formula of however. even of two closely related liquids. log t = K+r log q. but would vary slightly if the distillation were carried out under constant pressure.2 = c'Wl /W2 where c' = c + am. and. and if the relation ic 1 /iC 2 = P1 1 /P2 TF2 is really true for such liquids. but in practice a liquid is almost under constant pressure. generally given by Lehfeldt's formula. found.

]. tion. 1895. of 10. 104. Chem. 35. 9 . with a Still-head of uniform Temperature. Tensions of Mixtures of Volatile Liquids. rend. Vapours emitted by Mixtures of Volatile Sub" Some stances.. [3]." Trav. " On the Distillation of Liquid ^fixtures. "Fractional Distillation 9. . "On the properties of Liquid Mixtures." ibid." Trans.. de FEcole normale Sup. 4. in a Current of Steam. Thorpe. "Distillation. de la faculle de Lille. 1819." Sitzungsber. &c. 615. 544. Soc. Linebarger." ibid. 1863. Soc. 4. " On the 12.. On the Distillation of Mixtures a Contribution to the of Fractional Distillation.. " The Vapour physik. 37.." 1881. 2099. cit. Chem. 1898. "The Comparative Value of different Methods of Frac- tional Distillation... Lehfeldt." Part II. Duhem. Soc. Ann. Vol. " A Contribution to the Theory of Fractional Distilla- 8." Ann. 46. Soc.. Konowaloff. Berichte.. interpolation in these cases. Mag. Chem. 1879.D.. 1887. 2015 . loc. Traite elementaire de mechanique chimique. 6. 1899." Trans. of Volatile Substances. 35..it. Theory : 534. 3. 42. "On a new Method of determining Molecular Weights. 1421. 17." Proc. 1. 7. 13. 35. 2. 39. 1894.vi GENERAL CONCLUSIONS 113 relation. of Fractional Distillation. 57.. tillation of and Baly (16) obtained a similar result with the dismixtures of oxygen and nitrogen under constant Lehfeldt's formula could therefore be used for pressure.. 691. 1863. 5. 517. [V. 547." Trans. 1880. der Wiener Akad. Brown. 304. I . 430. 12. 1889. " Wanklyn. "On the Vapour Pressures of Mixtures of Liquids.. Phil. "Theory 1879. 153 .. 34. Roy. 129. 1879.. 1243. 1895. 1881. " Solutions and Mixtures." Journ." Berl. Toluene. 14. Zawidski. "On the Composition of the Saturated Vapours of Mixtures. 10. 1900. Brown. "On the Distillation of Benzene. Young. Chem." Wied. 39." On the Distillation of Mixtures of Carbon Disulphide and Carbon Tetrachloride. "On the Vapour Pressures of Binary Mixtures Liquids. 49 ." 1881. et Mem. 1877." Compt. Chem. 10. Amer. Margnles. [3]. Naumann. III." Thorpe's Dictionary of Applied Chemistry. Remarks on Mixtures 6. REFERENCES. . Berthelot. Soc." Zf. 11. I.

. pure. CHAP. Mag. is formed. it would be relatively very small indeed. 40. the vapour is richer than the * in the more volatile of the two components into liquid which the mixture tends to separate. 1899. " The Phase Rule. become smaller each time. In order to obtain a fair quantity of both components in a sufficiently pure state.]. if a large number of distillations richer in the more volatile were required. Phil." Phil. 15. in which case the composition of the vapour would be the same as that of the liquid. or nearly The amount of distillate would. then. 284. Bancroft. . "On the Properties of Liquid Mixtures. Unless a mixture of constant boiling point. by repeating the operation several times we might eventually obtain some of the more volatile component in a pure.114 14. 49.]. 397 1898. cit. 47. whether these com- two substances is ponents are the original substances which were mixed together or a mixture of constant boiling point and one of the original substances. however. we distil 'the mixture in the usual manner until. loc. one half of the total quantity has passed over. state. [V. 517- CHAPTEK VII DIRECTIONS FOB CARRYING OUT A FRACTIONAL DISTILLATION MIXTURES SEPARABLE INTO Two COMPONENTS IT has been stated (p. [V. the new distillate would be component. [V. " On the Distillation of Liquid Air and the Composition of Baly.].. we were to redistil the distillate and again still collect the first half. say. 1895." 1897. systematic fractional distillation is necessary. . the Gaseous and Liquid Phases. 71) that when a mixture of heated. 16. * . FRACTIONAL DISTILLATION Lehfeldt. the distillate will be richer than the residue in If the more volatile component. Mag. If. of the same composition as that of the original mixture." 3 Parts. and. 1900. and.

vii COLLECTION OF DISTILLATE IN FRACTIONS 115 Let us suppose that we have 200 grams of a mixture weights of benzene and toluene. but it is better to take two fractions each for the first and last 3. the difference being 30-4 and we might take 10 small flasks to provide for 10 fractions. At any rate the liquid tends to separate on distillation into the original components. the twelve fractions Suitable temperature ranges for the two boiling points. benzene and toluene. xB = c 5. the receivers being changed when the boiling point reaches certain definite temperatures to be arranged beforehand. The boiling point of benzene is 110'6. It is. convenient to have the same number of fractions above and below the middle temperature between and of toluene. Before distilling the mixture we must read the barometer. in the first place. Correction of Temperature. In order to trace the course of the separation as clearly as possible we will. 116). 8 with a range of 3 and 2 with a range of 3*2. because 80*2 and 110-6 are the boiling points of benzene and toluene rei 2 . however. as a rule. The relation between the vapour pressures and the molecular composition of mixtures of these liquids is expressed very nearly by a straight line and it is probable that Brown's law. are given in the second column of (p. not required for the preliminary distillation. and that c does not differ XA "A greatly from 2-47. make the range of temperature nearly the same for most 80-2 of the fractions. of equal W Collection of Distillate in " Fractions. is very nearly true. Table 35 The first three receivers are." ture should be first distilled The mixa and the distillate collected in convenient number of fractions. the mean ratio of the vapour pressures at equal temperatures between 80 and 110. no mixture of constant boiling point being formed.


fractionation f of benzene distillation Systematic Fractional Distillation.vn SYSTEMATIC FRACTIONAL DISTILLATION 11? mixture in the purest state attainable. capacity was used for the first and the bulb of the thermometer was covered with a little cotton. are given in Tables the read temperatures are not stated but only the true temperatures under normal pressure. the height of the barometer must be noted and the thermometer must be read when fractions of suitable bulk have been collected. because the same result will be attained with a smaller number of distillations. 36 and 37 now being described. The results of the and toluene with an ordinary bulb. and to recalculate the corrections if there is any change in the pressure.* The mixture must now be slowly laboratory purposes the drops of distillate should fall at the rate of about 1 per second. 192). 12). flask of distillation A In the first distillation First Fractional Distillation. the above corrections must be made with the greatest possible care and it may not be sufficient to estimate the temperatures to 0-1 but to 0'05 or less it may also be necessary to read the barometer from time to time during the course of the distillation and . t For the results obtained with an improved still-head. time will on the whole be saved. although each distillation takes longer. see Table 52. Unless otherwise stated. the temperature rose almost at once to 86 and the first * When a mixture of unknown composition is to be distilled. Rate of Distillation.c. but on the The large scale a much greater rate would be necessary. (p. all the distillations recorded in this book have been carried slower the distillation the better out at the rate of 1 drop of distillate per second. . In that case. about 270 c. : distilled for is the separation and.wool to prevent super-heating (p. The thermometer readings may be corrected afterwards. . we cannot decide upon the temperature ranges of the fractions beforehand.

the boiling point of toluene. The distillation was therefore stopped. amounting to 10'9 grams consisted . distilled. the apparatus allowed to cool and the residue in the flask weighed This residue. On the other hand the temperature reached 110-6.118 FRACTIONAL DISTILLATION CHAP. before the whole of the liquid had come over. of pure toluene and did not require to be TABLE 36. portion of the distillate was therefore collected in the 4th receiver. .

after the addition of the last fraction. 6 were now added collected to the residue in the in Nos. fraction first was redistilled. 5 for No. 3 when the temperature had risen to 86'2. the Procedure Dependent on Number of Fractions. after the llth fraction from the first distillation had been placed in the still. On recommencing the distillation it became clear that the temperature would not reach 86' 2 for a considerable time and the first portion was added of the distillate was therefore collected The process was continued as before . 5 still and the distillate was and 6 . - no distiUate came over below 107 6. say. In the fourth fractionation. If the range of temperature for each fraction had been larger. No. and also the first portion of the distillate which came over after the contents of receiver No. the first portion of the first distillate. 4 at 89-2 was stopped at 92-2. in receiver flask No. 9 because the temperature rose at once to 104-6 . 3 at 86-2. 4 was substituted for and the distillation No. Fourth Fractionation. and No. carried out in the collected in No. 5 to the residue in the still. 5 instead of 3 for the majority of them. 3. 3 had been added to the residue in the still. 4. then on re- . In other respects the procedure was the same as in the third fractionation. so also. 1. the second being placed in the still when the temperature reached 81 '2. No. no distillate was collected in No.vii FIRST FOUR FRACTIONATIONS 119 soon as the temperature reached 83'2. 4 was put in the place of No. The contents of receiver No. after the cotton-wool on the thermometer had become second fraction from the first distillation dry. the that in No. At 89'2 the flame was removed and. the remainder of the fractionation was carried out in this manner except that. The third fractionation was same manner as the second. being Third Fraetionation.

but that only 9-75 grams of distillate came over below 89-2. the temperature would have risen at once. a large number of receivers there is a greater amount of separation of the components in a complete distillation and. and this is.120 FRACTIONAL DISTILLATION distillation after addition of CHAP. It will also be noticed that the weights of the middle fractions steadily diminished. and receiver No. and in the subsequent distillations the loss was in nearly every case more than The greater loss was due partly evaporation. The total loss after fourteen fractionations was 18. -It will be observed that. time is saved.199-25). It is thus evident that distillation the the separation of the less volatile component is much easier than that of the more volatile. This method is usually adopted. 3 might have been left in position while fraction No. a very appreciable amount. loss of 0'75 On the first distillation there was a gram of material (200-0. and partly to the small amount 1 gram. always found to be the case. 4. or very rapidly. while those of the lowest and . commencing the any fraction to the residue in the still. 2. say. but by taking . while only 31'55 grams of benzene boiling within 1 degree had been obtained. In the second complete distillation the total weight fell from 199-25 to 198-2. in the first distillation. no would have been collected in No. weight of pure toluene recovered was 42-1 grams. due entirely to evaporation. to increased of liquid left adhering each time to the funnel through which the fraction was poured into the still. 10-9 grams of pure toluene were obtained. on the whole. on adding any fraction. 4 was being placed in the still and the distillation was recommenced.1 grams. No. indeed. to the initial temperature of the fraction below distillate thus. Relative Rate of Separation of Components. a temperature 9 above the boiling and at the end of the fourth complete point of benzene . LOSS Of Material.

of liquid are stated to have been collected above 108. 48-3 and 48'8 respectively the percentage was actually 45-4. the it The results of four complete distillations are given in Table 36. while the original mixture contained 50 per cent. the to distil. is distilled . and. It is therefore preferable increase the temperature ranges of the middle fractions. in the second complete distillation tion of a mixture of 25 * An the volume of the corresponding fraction is given as only 3 c. be very nearly equal to that of the more volatile component in the original mixture even in the first distillation. 17c. . Alteration of Temperature Ranges. as will be seen later. the last fraction in the second should have been at least 6 times greater than that stated . but the results given for the second are quite impossible. and will be seen that if the fractionation were continued in the first same manner as become too small before. 121 highest fractions increased this also invariably occurs a mixture of two liquids is repeatedly distilled. in general. when an improved still-head is used. the data for the first distillation are such as might be expected. coming over below the middle temperature between the boiling points of the pure components.c. and it was not until the 8th fractionation that the volume again reached .vii ALTERATION OF TEMPERATURE RANGES . of benzene. was not far below that of the benzene in the original mixture. and the middle fractions would soon first fraction would never gradually to consist of pure benzene. 47'0.* when Weight of Distillate below Middle Temperature.c. for while in the first distillation 17 c. which separates normally and easily into the original com- ponents. and to diminish account has been given by Kreis (1) of the fractional distillagrams of benzene and 25 grams of toluene with an ordinary distillation bulb . it is difficult to understand how the mistake can have arisen. the weight of distillate below the middle temperature should. c. The percentage weight of distillate. Assuming the correctness of the data for the first distillation. Similar results are always obtained when a mixture of two liquids. in the four distillations.

are collected between equal intervals of temperature. together with those of the fourth fractionation. those of the fractions near the boiling points of the pure substances (2). Pio. and it is advisable either to plot the when distillate against the (or total) weight of temperature (Fig. 27. A t. are given in Table 37. So long as the fractions. This was accordingly done in the subsequent fractionations. of course. and to tabulate these ratios as well as the actual weights. by its temperature range. A w. infinitely great . the temperature ranges are gradually altered this becomes less evident. liquid in any fraction. w/A and t. the higher is the ratio A for a pure liquid being. which were carried to their extreme limit.122 FRACTIONAL DISTILLATION CHAP. 27). or most of them. The purer the fraction. The results. Results of fractional distillation of mixture of benzene and toluene. that will it . the weights of distillate indicate clearly but enough the progress of the separation. or to divide the weight of each percentage 20 40 60 80 100 Percentage Weight of distillate. Ratio of Weight of Fraction to Temperature Range.

and the residue taken as pure toluene. and their temperature ranges were reduced to 0'2 and 005 respectively. it was concluded that the first portion of the first distillate in the ninth fractionation would consist of pure benzene it was . and those of the lowest and highest increase much more slowly than before. after No.vii SEPARATION OF PURE BENZENE 123 be seen from Table 37 that in the later fractionations. 2. In the next fractionation. 1. 12 distillate had been added to the residue in the llth receiver was left in position until the still. 6 had been added to the residue in the still. therefore. 12. In the 14th fractionation. the temperature remained a considerable at the boiling point of benzene. the temperature had not risen . while the weights of the middle fractions diminish. lected below 80'45 Separation of Pure Benzene. In the tenth fractionation the weights of the llth and 12th fractions had become very small. first first distillate was taken to be pure Elimination of Fractions. and in the 14th by that of No. the ratios A w/A t continue to change rapidly. the temperature reached 110'6. 8O2. and. and amount of distillate came over below 80'25. the number of fractions was further reduced by the exclusion of No. after No. in the latter. In the 12th and 13th fractionations the last fraction was in each case similarly eliminated. and in each of the subsequent fractionations the portion of the benzene. so that nothing was collected in No. it was found that the temperature rose at once above 81. for a short time. the distillation was then In this stopped. As the range of this fraction was reduced to 0-15. therefore collected separately as there was no necessity to redistil it. way the number of fractions was reduced to eleven by the elimination of No. 5. When the liquid' colin the seventh fractionation was distilled at the beginning of the eighth. and also that when the distillation had been carried as far as possible.





57'1). loss by evaporation.c.p. was 7'8 grams. be most readily isolated and. the fractionation was carried on as before. it is the less volatile component which is the easier to obtain in a pure state.c. of liquid collected in B. The total weight of pure benzene recovered by pushing the series of fractionations to their extreme limit was 81'4 grams out of 100 taken . as a rule. The distillation CHAP. of methyl acetate (b. 7 was then placed in the still. and by transference from flasks was 22-0 grams. MIXTURES SEPARABLE INTO THREE COMPONENTS have seen that when a mixture tends to sepaon distillation into two components. higher than was therefore stopped at this point.128 FRACTIONAL DISTILLATION 914. after cooling. The amount and the to still. temperature reached 100.p. of the other two. of propyl acetate (b. of this is Ethyl and Ppopyl Acetates. and then in No.c. and the residue in the still. and rejected. Final Fraetionations. When a mixture tends to separate into three components the least volatile of them can. Four additional operations were required for the separation of the benzene and two for that of the toluene. Fraction No. 101-55) carried out with a plain vertical still-head one . was placed in a separate receiver R.p. the weight of toluene was 88*8 grams. of ethyl acetate (b. 250 c. The amount little the whole operation was a of time required for over thirty hours. and the distillate collected in 7. fractions below R until the The rest of In the remaining distillations the and above the middle point were treated separately. 77*15) and 200 c. Methyl. An illustration afforded by the fractionation of a mixture of 200 c. the more volatile is rate We the easier to separate. the residue from the last of the lower fractions and the first portion of distillate from the first of the higher ones being collected each time in B.

* It will be noticed that the method adopted differed from that already described in so far that the number of fall gradually increased. when receiver No. : until the temperature rose to 71 '0. 4 until the temperature rose to 63'8. Full details of this fractionation have been published in the Philosophical Magazine (3) and in Table 38 the results of the first eight fractionations and of 12th.vii SEPARATION OF THREE COMPONENTS 129 metre in length. the number corrected of the fraction t. At 84-4 a new receiver. The gas was then turned out and the second fraction from I (No. numbered that they distillations. requires . 8) was added to the residue in the still. was substituted for No. is that at which the distillate began to come over . 11 was put in its place. 20th and 24th are given. to rise to 91'7. and reduced to 760 first pressure. K . 16th. I. the final temperature. 11 and the temperature was allowed fraction from I was then added * p. but in the four fractionations the lowest temperature (in brackets) mm. and the distillation was continued until the temperature reached 77' 1. the weight of each fraction. 5 tion. 5 was substituted for it and the distillation was continued until the temperature reached 71 '0. the fractions are so into their proper places in the later : The data given are as follows . fractions was small at first and was For convenience. A w. when No. I'M. Heat was again applied and the distillate was collected in No. when this receiver was replaced by No. 5) was distilled and the distillate was collected in receiver No. The last to the residue in the flask For the results obtained with an improved still-head see Table 54. 8 until the temperature again rose to 77'1. II. 8. and the ratio First A w/A t. The first fractionano comment the second. for each fraction. F. was carried the first fraction from I out in the following manner (No. 13. 11) was then added to the residue and the distillate was collected in No. and Second Fractionations. The third fraction from I (No. No.





Y. The presence of the middle substance. ethyl acetate. as a large from subsequent fractionations up to the 10th were collected amount of propyl acetate had been removed. Separation of Propyl Acetate. above 101-45) was placed in a new flask. 15 and from 98'4 to 101'5. Separation of Methyl Acetate. D and E. a new fraction. before - the end of the last distillation. which the first portions Accumulation of Ethyl Acetate in Middle Fractions. 14 from 91'7 in No.p. and it was not thought advisable to separate the first portion of the first distillate until the 8th fractionation. and subsequently the residues were placed in w. At the end of the 12th fractionation the temperature rose only to 101'15 and the residue was placed in a third flask. after which to 98'4 fractions were collected in No. being collected at the beginning.134 FRACTIONAL DISTILLATION distillate collected in CHAP. The residues in z. and the No. A. z. The third fractionation III was carried out in a similar manner. subsequent fractionations is clearly shown by the rise in the . No. As the temperature rose to 101 55. when the distillation was stopped and the residue was poured into No. . It was not until the 5th fractionation that the first fraction began to boil at 57'1. after were collected in B. is not clearly indicated until the 4th fractionation. 3. the temperature did not reach 101 '55 and the residue from the llth fractionation (b. 13 up to 91-7. c. v Q. when the value of A w/A t for the 6th distillate (No. x. the boiling point of methyl acetate. as shown in Table 38. 17. 11) is somewhat higher than for those above and below it but the gradual accumulation of the ethyl acetate in the middle fractions in . . and was not redistilled. . . This portion and also the corresponding ones up to the 20th fractionation were collected together in a flask. but after this. the boiling point of propyl acetate. the residue was placed in a separate flask.

and that the remaining fractions B to E and 5 to 10 contained only methyl and ethyl acetates. of the later curves represent the methyl and propyl acetates removed in the first portion of the first distillates and in the residues respectively. The total weight was 490-8 grams. for fractions 9 and 10. 28 (a and &). fraction now containing Graphical Representation of Results. and therefore 86 grams had been lost by evaporation and by transference from flask to still. tions The preliminary series of fractiona- was therefore completed. 9 and 10 was gradually diminished from 0'2 each in the llth fractionation until no rise could value of there was. 11 in the 22nd and 23rd fractionations. Final Fractionations. no perceptible rise of temperature during the collection of No. and the propyl acetate with . after the 10th fractionation.vii GRAPHICAL REPRESENTATION 135 A w/A t for the fractions 8 and 11 and. it is safe to conclude that the fractions Q to z were free from methyl acetate. It was therefore certain that after be detected . The fractions into which the esters had been separated at the end of the 26th fractionation are shown in Table 39. The progress by mapping the temperatures against the percentage weights of distillate collected. and the curves for the first twelve fractionations are shown in The horizontal lines at the extremities Fig. Similarly. but it was necessary to treat the methyl acetate with phosphorus pentoxide to remove moisture. no more than two components. The range of temperature for Nos. indeed. the 26th fractionation the tenth fraction was free from propyl acetate. The final separation of methyl and ethyl acetates and of ethyl and propyl acetates was carried out in the manner described for the later fractionations of mixtures of benzene and toluene. The presence of ethyl acetate is of the separation is well seen clearly indicated in the fourth curve but not in the earlier ones.

Methyl and Ethyl acetates. . TABLE 39. potassium hydrolysis.136 FRACTIONAL DISTILLATION carbonate to remove free CHAP. acid due to it slight The loss was thus much greater than would otherwise have been.

VII COMPLEX MIXTURES COMPLEX MIXTURES 137 to that complex mixtures is similar components. but the number of fractions must be increased and the time adopted is The procedure in the case of when there are three required longer. .

FIG. and propyl acetates (continued). vra 20 tO 60 80 2040 6010100 Percentage Weight of DistUlaJtb. but of a single substance with a boiling point between those of the two which are actually present. and it is only after some progress has been made with the fractionations that the presence of both liquids is clearly shown. afforded .138 FRACTIONAL DISTILLATION CHAP. 36'3) from the light distillate from American petroleum (4). 27'95) and normal pentane (b. Separation of instance of this Pentanes from Petroleum. 28 (b).p. ethyl. the presence of the two liquids. is A striking by the separation of isopentane (b. Results of fractional distillation of mixture of methyl.p.

I. necessary to employ a very efficient still-head (Chapter XII) but the method of arranging the fractions is the same as when an ordinary still-head is used. seems to indicate (5) that a single substance boiling at about 33 is being separated from others boiling considerably lower (butanes) and higher (hexanes) . the two pentanes referred to. The first curve. 182 and 253). 4th. 7th. In order to effect the separation of the pentanes it is . the consists chiefly of butanes and probably a methyl methane. sider the curves showing the results of the 10th and 13th fractionations (Fig. I Weight of Distillate Fio. with very much smaller quantities of peftitamethylene and trimethylethyl methane. 21). in Separation of normal and isopcntauc from American petroleum.VII INTERPRETATION OF RESULTS treatment " 139 After with to a mixture of concentrated nitric little tetra- and sulphuric acids ether remove " petroleum impurities. both of which boil at about 50 there are no substances present except the two pentanes with boiling points between about 10 and 50. Graphical Representation and Interpretation of The fractionations will be referred to more fully later on (pp. but it may be well here to conResults. 29). and hexanes and other hydrocarbons boiling higher than 60. all boiling below 10. 1st.

When. . if that were really the case. is a little over As the curve is more horizontal below than No. This change is a sure sign that there are at least two substances present boiling at temperatures not very far apart. and it is fractionations only by keeping a careful record of the results of the and especially of the values Aw/A.140 FRACTIONAL DISTILLATION CHAP. showing that the higher boiling point 36'0. that the presence of such substances can be detected. Hexamethylene in American Petroleum. it may very easily be overlooked. The upper extremity of the curve XIII is perfectly horizontal at 36'3 and the true boiling point of the less volatile component (normal pentane) is thus established . in a complex mixture. one component is present in relatively very small quantity. approximates to a straight line between about 30 and 37. VII there is evidence that the second component must boil not far from 28. terminating at 36 4. This seems to indicate that the less volatile component boils at a temperature not far from 36 (the Instead of - hexanes having now been almost completely eliminated). Curve VII is fairly straight from about 28'5 to about 35*5 but then becomes much more horizontal. this. say. they become less horizontal at about 33 and the next curve. the curves for subsequent fractionations should become more horizontal (indicating greater purity) at about that temperature. the form of the lower part of the curve indicates that the boiling point of the more volatile component (isopentane) must be very that it close to 28 and further fractionation showed is really 27'95. 25 and 40. and more vertical at. Curve X is distinctly more vertical in the middle it becomes nearly horizontal above and terminates at 36'3. or by plotting the weights of distillate against the temperature. but. . but little or no light is thrown on the actual boiling points. IV.

with a maximum at 67 to 68. and in the 7th fractionation a maximum value of Aw/A was observed at about this temperature in addition to that due to normal hexane near 69. and 7th fractionations between the abovenamed limits of temperature and of the 12th fractionation between 74 and 85 are given in Table 41.HEXAMETHYLENE IN PETROLEUM 141 As an example. in the neighbourhood of 80 was therefore increased. Now. The number of fractions slightly higher instead of lower. the ratio Aw/Ai would become smaller each time at the intermediate temperatures. consider the distillate from American petroleum coming over between 66 and 91 (6). if there were no substance present between the hexanes and heptanes. V. In this case. Above 70 the ratios fell to a minimum at 78 to 84 and then rose again. again. no satisfactory result could be obtained without the use of a very efficient still-head. 6th. and notably at about 80 but it will be seen that in the next fractionation the ratio was . . TABLE 41. The results of the 5th. The fifth fractionation gave high values of Aw/A from 61 '5 to 70.

Mixtures of Constant Boiling.p. when American petroleum is distilled. solution of benzene in hexane boils almost constantly at nearly the same temperature as hexane itself. Hexamethylene in Galieian Petroleum. in this case the aromatic hydrocarbons had been removed by treatment with nitric and sulphuric acids. and that the two substances cannot be of .142 FRACTIONAL DISTILLATION CHAP. 80'85) in a pure state. separated by fractional distillation.Point. Later on. examination of the liquid led to the conclusion that it consisted of pure hexamethylene (8) (9) but it was afterwards found (10) that it could be partially. all that can be done is to .p. though considerable improvement was effected. . and it was necessary to resort to fractional crystallisation to separate the hexamethylene (b. no doubt a heptane. to the formation of mixtures of constant (minimum or maximum) boiling point. C. Fortey (7). as will be seen from No. mostly from about 63 to 68. for a 10 or 20 per cent. a large quantity of Galieian petroleum was fractionated by Miss E. Moreover. Reference has been made in Chapter IV. Further fractionation of the distillates coming over between 72 and 88 showed that there was really a substance present boiling not far from 80. it is not possible to separate both components . who obtained a considerable amount A chemical liquid boiling quite constantly at 80'8. When a liquid contains two components which are capable of forming a mixture of constant boiling point. almost the whole of the benzene comes over below 70. The quantity of liquid was too small to allow of a fraction of quite constant boiling being obtained. but not completely. frozen in an ordinary freezing mixture. It might be supposed that this substance was benzene (b. and. 8O2) but that could not have been the case. It is evident that there is another hydrocarbon present in small quantity. 12.

On the other hand. 8. Hydrocarbons from American Petroleum. 2 per cent. 259. 50 per cent." Liebiy's Annalen.. When a liquid contains three components which are not closely related to each other it may happen that both a ternary and a binary mixture of constant boiling point are formed on distillation. Thomas and Young. 3. " Comparative Investigations on the Methods of Fractional Distillation. Journ. .]. and even this is not possible when the boiling points are very near together. Chem. by less than 0'1. 1897. 224. Sor.. REFERENCES. In that case it is only possible to separate one of the original components in a pure state. and. Normal and Iso-pentane. without title. say. Mendele"eff." Trans. and XV. but Eussian petroleum is much richer in aromatic hydrocarbons. Thus. in all probability. 37. 1884." Phil. American petroleum contains a relatively small amount of benzene. because the boiling points of hexane and of the binary mixture differ. These points will be considered more fully in Chapters XIII. a small quantity of that component could be separated in a pure state from the mixture of constant boiling point (11).. 1883. 2. 1894.. Barrell. paper. Chem. only a portion of it distilling with the hexanes. if the original mixture contained. some of the benzene comes over at its true boiling point.vii MIXTURES OF CONSTANT BOILING POINT 143 that component which is in excess from the mixture of constant boiling point. of benzene. 440. Protok. 189. I. [V. " On the Separation of Three Liquids by Fractional Distillation. and the whole of it comes over with the hexanes . " Some Young and Thomas. say. Phys. 1. of benzene (6). consequently. on the fractional distillation of A Baku Petroleum. Soc. no amount of fractionation with the most perfect separate apparatus would make it possible to separate either pure normal hexane or pure benzene from a mixture containing. Kreis. 71. Mag. 4. Rttss.

x by Brown's formula -? = W . That when this of at the vapour from any mixture formula holds good. 9. That the vapour pressures of mixtures of two closely and rarely of others are represented related compounds with small error by the formula 2. stant c does not differ greatly from the mean ratio of the vapour pressures of the pure substances at temperatures between their respective boiling points. 7. Berichte. 1899. 8. Baeyer. Composition of American Petroleum. presence of Hexanaphthene in Caucasian " On some new conBerl. 19." Trans. Chem. 11. Critical Constants of Hexamethylene. "Specific Gravities and Boiling Points of Mixtures of Benzene and Normal Hexane. 278. Liebig's Annalen. 873. Young and Fortey. the composition is given." Berl. 28. 26. 6. 923. approximately c ?.. Indmtry. Chem. 1898." stituents of Caucasian Naphtha. Soc." Trans. CHAPTEK VIII THEORETICAL EELATIONS BETWEEN THE WEIGHT AND COMPOSITION OF DISTILLATE Application of Brown's Formula. Berichte. Chem. 1898. Experiments on Fractional Distillation." . 974. P = *^A + (100 m)PB 100 . FRACTIONAL DISTILLATION CHAP. 1898. "On 1893." ibid. " Young. 932. "The Vapour Pressures.. Chem. We may conclude. 1893. where the con#A Wj. Jackson and Young. 1072. 88. 229 "On the Reduction products of Benzene. Soc. " On the Markownikoff." Jowrn." Trans. 75. Chem... 1897. Soc. 73. any rate. 73. 10. " Fortey. Specific Volumes and . the Hydro-derivatives of Benzene. 905." Trans. Soc.. 73. 1895. Hexamethylene from American and Galician Petroleum. " Young. Soc. (1) It has been pointed out in previous chapters 1. 577 Naphtha.144 5. 30. 1900.

c~l c)y} c (l B A y) e x) (l the more volatile liquid B in unit weight of the distillate coming over at the instant when x is the quantity of liquid distilled. A. suppose that c = 2 and that In the amounts diagram (Fig. the have been collected are represented as abscissas. of distillate that 30). 02 04 0-6 tion of the distillate alters slowly at first. It will be seen that the composiFio. that Brown's formula can be used without much error for mixtures of two closely related substances. 80. i of mixtures of Mixtures Of Two Components.. and the variation in the composition of the distillate represented graphically. A and B. By means of this equation the changes of composition that take place in the course of a distillation may be traced. M as we (M /L)y{c + (1 = quantity of where y the weights of and obtain by integration c~ l = e . when suitably modified. in the distillate as ordinates. in original mixture.VIII MIXTURES OF TWO COMPONENTS 145 then. are applicable also to mixtures of three or more closely related substances. and the relative quantities of the two liquids. Brown's formula may be written first the case where = residue of liquid A at the of liquid at same instant. then more and more rapidly. and r} = residue originally Taking L and c present and L + M = 1. and it is probable that the two formulae referred to above. To take a very simple case. B any instant. while the distillate first portion of the contains a considerable amount of the less volatile substance. also that Weight of Distillate 5 = 0-5. A =<):. the last portion is very neai'ly free from the . Taking two liquids.

but that the vapour reaches the condenser in the same state as when first evolved from the liquid in the still. 31 and 32. collecting the distillates in six or eight fractions. that two of these fractions. one containing A in the ratio of 9 1 and the other in the ratio 1 9. In making use of this formula it is assumed that no condensation (and therefore no fractionation) goes on in the still-head. it will be seen 02 Fio. X Mixtures of Three Components. now. Again. 32. By using an improved still-head (Chapters to XII) a more rapid separation would be effected. B. These points are fully confirmed by experiment. are distilled separately and completely the results will then B and : : . may be . mixture distilled. If we have a mixture it B and A. Suppose. 31. we shall have a large excess of excess of A in the B in the first fraction and a still larger last. 0-* 06 08 1-0 5 = 0-9. C. of three closely related substances. A = 0-0. be represented by Figs. By fractionating a few times in the ordinary way. A =0-1. 0-4 06 08 1-0 02 Fio.146 FRACTIONAL DISTILLATION lower boiling component. in mixture distilled. and an explanation is afforded of the fact that it is much easier to separate the less volatile than the other component in a pure state. in B = 0-1. B as the last tenth from the second fraction is in A the purification of A still proceeds more rapidly than that of B. that the first tenth of the distillate from the first of these fractions is not so rich in .

ethyl and propyl acetates at the same temperature and the original weights of the three liquids C. B and A as L = = N=b. Distillation of mixtures containing three components. which enable us to follow the course distillation. which is approximately proportional to the vapour pressure of the corresponding liquid. by integration. B. of a Taking the three constants as c = 4. and C. a=1 which are roughly proportional to the vapour pressures of methyl. 33 (a to/). Here again formulae may be obtained 0-2 0-4 06 08 IO 0-2 04 0-6 08 1-0 Fio. Fig 33 a represents the first distillation. each weight being multiplied by a suitable constant.VIII MIXTURES OF THREE COMPONENTS 147 conjectured that the proportion of the three substances in the vapour at any instant is the same as that of the weights of the three substances in the residue in the still. A. M L 2 . b =-- 2.

.Cor L.

then. and also rich in A but free from C. It is. 1894. " . part of the distillate it is thus clear that there must be far greater difficulty in separating the middle substance than either of the others from a mixture. and the best method of obtaining B in a pure state is to carry out a series of preliminary fractionations in the manner described for methyl. 1. however. Mag. REFERENCE. 8. and all those coming over above that temperature contain only B and A. not difficult to obtain fractions rich iti C but free from A. the amount of advances only from 0'96 to 0'97 in the most favourable . and these separations will already be far advanced." Phil. 0-96 of B and 0'02 of A. ethyl fractions and propyl acetates (p. [V]. On the Separation of Three Liquids by Fractional Distillation.. 128) until all the coming over below the boiling point of B contain only C and B. 37. only to deal with the separation of two components from the mixtures. Thomas and Young. Barrell. We have.viii MIXTURES OF THREE COMPONENTS 149 B containing 0'02 of C.

The molecular percentages of which boil at the temperatures in the first column.CHAPTER IX RELATION BETWEEN THE BOILING POINTS OP RESIDUE AND DISTILLATE Data required. In Table 43 are given benzene in mixtures 1. c. and that Brown's we may safely assume mixtures may be calculated from formula. . The mean ratio of the vapour pressures at temperatures between 80 and 110 is about 2'5.A = 2-5 W . In the case of benzene and toluene. for mixtures of any two substances. the boiling points of mixtures have been found to agree very closely with those read from the theoretical curve. Benzene and Toluene. gives the relation between the composition of liquid and vapour with fair accuracy. We have seen that if the two substances are closely related that the boiling points of -the vapour pressures of the components. but the relation between the composition of liquid and vapour has not been experimentally determined. the boiling point of the distillate may be read from the curve. taking the mean ratio of the vapour pressures for the value of the constant. the curve representing the relations between boiling point and molecular composition has been constructed. and if the relation between the composition of the liquid mixture and of its vapour is known. and we may probably assume that the relation is expressed with sufficient S /WA accuracy by the formula #B/a. If.

CH. TABLE 43. between the boiling points of liquid ' and distillate. . in the 3. IX BENZENE AND TOLUENE 151 2. The corresponding molecular percentages of benzene distillate. Benzene and Toluene. The boiling points of these distillates read from the The differences (A^) curve. 4. calculated by means of Brown's formula.

Thus the fraction that came over from 92-2 to 95'4 was added which was boiling at 92-2. there was some condensation which would lower the temperature to some extent at first. and that the mixture would actually begin to boil at about 87'5. Application to Let us consider the first Distillation first of Benzene and Toluene. 5. mD is 51'5 (p. to 90-5 No. 89-2 to 92-2 No. and the boiling point would therefore fee somewhat higher than if the fraction were distilled alone. No. 86-2 to 89'2. take place in the still-head. 87 - The middle temperature for . On the other hand. with the exception of the first. v being 0-28 when the molecular percentage of normal pentane. The ratio of the vapour pressures at 30 is 1-33. and was therefore richer in toluene. 4. That is to say. but. were not distilled separately but were added to the residues left in the still. The theoretical boiling point curve differs very slightly from a straight line. the middle temperature is 93*8 and from Table 43 we see that the distillate from a mixture boiling at 93'8 would have a boiling point about 6 -2 lower. For fraction 6. four fractionations. details of which are given on p. 59).152 FRACTIONAL DISTILLATION CHAP. D . 5 : . the maximum deviation. and it is clear that a considerable amount of distillate would be collected below 89'2 in receiver No. and this has been taken as the constant in Brown's formula. In the sixth fractionation temperature ranges were No. 4. determined. 4. 5. it would begin to boil at 87 '6. The temperature ranges below 92-2 to 95-4 were as follows No. to the residue from No. the fractions. (p. 6. . 118. In the case of benzene and toluene. or 87' 6. We may perhaps suppose that the two disturbing factors would about counterbalance each other. to 95-4. if this fraction were distilled in such a manner that no condensation could . 124) the corresponding 84-0 to 87'0 . in the actual fractionation. collected from 92-2 to 95-4. as a matter of fact. 90'5. the range of temperature for most of the fractions was 3.

The temperature range would have to be reduced to 0-4 [No. This was. 128. 5. and the boiling point of the distillate would be. The middle temperature would be 88'4. In this case none of the distillate would be collected in No. 5. so that no distillate would be collected in No. and the from this would be 32-2 0'6 31 6. 5. boiling point of the distillate - The middle temperature . 4. and that would mean that the number of fractions would have to be doubled to begin with. 6 would be 32-2. 8 35 - is for the corresponding fractions No. found to be the case. roughly. ' In the thirteenth fractionation the range of No. Let us now Application to Distillation of Pentanes. 6. We should then have normal pentane. 5. 6 would be 92-95. say. 5 might be (and actually was) left in position when No. Very little would therefore be collected even in No. 6 fraction was added to the residue in the still and the mixture was redistilled. No.4-2 = 84-2. and receiver No. 5. for No.ix NORMAL AND ISO-PENTANE 153 No. No. from 81-4 to 95-4. 31-8 32-2 . and that it would not be possible to increase the temperature range of the middle fractions beyond a very small amount. the temperature rising at once above 81. the range of the middle ones would be about O84. consider the behaviour of a mixture of isopentane and The difference between the boiling points and the middle temperature is 32-6. and the distillate from this would begin to boil at about 87 '0 or 5 '95 lower. 31-8 32-6. 88'4. 6 had been increased to 14. 6. 6 might begin to boil at the initial temperature of No. and there would be no object in continuing the fractionation in the same way as before. 32-2 32-6] in order that the distillate from No. The fractions above and below 95'4 were therefore treated separately as described on p. If the distillate were collected in the same number of fractions as in the case of benzene and toluene. in fact. 0'8. 31'0 31'8 .

CHAPTER X MODIFICATIONS OF THE STILL-HEAD Object Of Modifications. was distilled " " When it dilute through is alembic clear that the alcohol collected in the uppermost receiver must have been considerably stronger than that in the lowest. for in a book entitled Philosophorum. by this (Fig. 1553. by Philip Ulstadius.154 It will thus be seen that. there is an illustration of a five-headed still Secretis spirit Fio. even without considering the by evaporation of such volatile liquids. Great progress was made in the improvement of the still- . 34. CHAP. Five-headed still described Ulstadius. and many attempts have been made to modify the still-head such a manner as to bring about a more complete separation in a single operation. 1553. It is evident from what has been stated in Chapter VII that the process of fractional distillation with an ordinary still-head is frequently an exceedingly tedious one. 34). the separation of isopentane from normal pentane with an ordinary still-head would be practically impossible. The idea of improving the still-head is by no means a new one. sen liber de in Naturce. It is only by using a greatly improved still-head that such a excessive loss difficult separation is rendered possible.

the outer parts of the vapour rises up the cylinit come in contact with the condensed liquid flowing down the sides of the tube. The object of the modifications which have been made in the still-head is to bring about a more complete separation of the components of the mixture in a single distillation. a process of fractionation more or less effective according to the form of apparatus and the rate of distillation taking place in the still-head (1). Let us suppose. to facilitate the consideration of what occurs. this liquid is slightly cooled and fresh condensation constantly goes on. and much of it liquid at all pass through the still-head without reaching the the condensed liquid. be convenient to consider first those forms of apparatus which are suitable for laboratory purposes. Of all possible Inefficiency of Plain Vertical Still-Head. for the most part. distillation. that a satisfactory separation of the may . the plain vertical still-head is the least efficient. components of the mixture may take place. It may. Since the tube is constantly losing heat by radiation and by conduction to the surrounding air. however. may be brought about. Brown's Formula. the improved still-heads first employed in the laboratory were. on the other hand. merely adaptations of those already in use on the large scale. so that a state of equilibrium. that we arc dealing with a . In order. as much of the vapour as possible should be brought into intimate contact with the condensed liquid. the outer- most layer without of of vapour probably condensing almost completely of composition.x INEFFICIENCY OF PLAIN STILL-HEAD in 155 of in head alcohol from commerce notably for the better separation weak spirit before any advance was made the laboratory and. indeed. back to the still. flows rapidly down the vertical walls of the tube. as During the drical tube. forms. much change The central portion vapour rises rapidly up the tube and can only reach the liquid by diffusion or by convection currents. as regards composition. however.

same as will not That the composition of the condensed liquid from this. Theoretically Perfect Still-Head. equal weights of the two substances are taken. the total weight of liquid and vapour We may being. and the percentage of c If + is we suppose. to which Brown's TIT" formula XL c^=-2 WA. say. so that Then the composition of the W l = W z . and that the condensed liquid remains in this part of the tube until its weight is. first M small quantity of vapour formed of will be given by the equation will = c #A and the percentage B in the vapour If be c - - + 1 a distillation bulb with exceedingly short still-head be employed. say. as a theoretical case. That the composition of the liquid and vapour (taken in the lower part of a long together) in the still-head will be practically the that of the vapour first formed.* = c x c = c2 . is applicable. its composition will' practically be given by the equation -. that the still-head . differ sensibly or W -B = c. that and theoretically perfect still-head condensation goes on. then assume 1.156 FRACTIONAL DISTILLATION if CHAP. If the residual vapour is brought into thorough contact with the condensed liquid. and 2. in the first place. 1/1000 of that of the mixture taken. Also let us suppose that. 100 times as great as that of the residual uncondensed vapour in the same part of the tube. mixture of two closely related compounds. now. and that x and W refer to the weights of vapour and liquid respectively. Suppose. this will practically be the percentage of B in the first portion of distillate.

and. whe'n the distillate begins to come over. lastly. 44. -5 SR = c n+1 . with equal weights of the components. . the composition of the vapour in the still and in the first nine sections would be as follows If we take c start : TABLE No. of Section. as in the case of benzene and toluene. then we should have. #A and = 2'5. that the condensed liquid in any section has the same composition as the vapour in the section below. for the composition of the vapour in the nih section. and that the distillation goes on in such a manner that perfect equilibrium between vapour and condensed liquid is established in each section.THEORETICALLY PERFECT STILL-HEAD 157 divided into n sections.

That the rate of distillation does greatly influence the distillation separation by Table 45. and is well shown which are recorded the results of three ture. the . Comparison of Still-Heads. An still -head was used. in was pointed out by Brown (2). that shall return as completely as possible to the still. each of 50 grams of the benzene-toluene mixand 120 drops per minute. (c) con- venience in handling. (b) freedom from liability to fracture. the smallest quantity of a substance can be usefully distilled must be considerably greater than that of the liquid and vapour in the stillhead. In comparing the efficiency and usefulness of different stillheads it is necessary always to distil a mixture of the same composition.158 FRACTIONAL DISTILLATION CTIAP. It is also important that. obviously. 1 metre long. It is therefore of great importance to construct the still-head in such a manner that consistently with the quantity of condensed liquid in it at any efficiency moment shall be as small as possible. and the distillate was collected improved distillations. 60. Rate of Distillation. after the removal of the source of heat. in eleven fractions. serves the purpose very well swinging behind the receiver. so that the drops may be easily timed. good plan is to have a seconds pendulum a weight attached to a string 39 -1 inches or. at the rate of 30. say. It is necessary also that the rate of should always be the same. Among other points to be considered are (a) ease of construction. Equal weights of pure benzene and toluene may conveniently be employed. Mixture Distilled. difference) after the temperature had reached point of pure toluene. this liquid and. For convenience. (calculated the twelfth consisting of the residue by the boiling results are stated as percentages. and it is best to A collect the distillate at the rate of one drop per second.

inn Temperature range.RATE OF DISTILLATION TABLE 45. .

. when measured.160 FRACTIONAL DISTILLATION For the very long still-heads. vertical still-head is The Plain Vertical Still. The influence of such alterations (b) is shown in Table 46.Head. sum of the weights of the actual loss by evaporation. the estimated loss by evaporation +0-85 gram was subtracted from the difference between the original weight of the mixture and the fractions. efficient less TABLE 46. yet a certain amount of fractionation does take place in it. varied from O'OO to O65 gram. PLAIN VERTICAL STILL-HEADS. which could not conveniently be weighed. The Although the plain than any other. its internal diameter. and it may be well to consider the effect of altering (a) its length. Vertical height in cm. and was greater for the large than the small still-heads.

Influence of Width. roughly. but while the amount of liquid in the tube is thus diminished. for if the tube is too narrow. 161 weight of liquid and vapour also increases and proportional to the length.. On the other hand. be diminished beyond a certain amount. in making a plain still-head. as will be seen from the results of the last five distillations. secondly. however. a short piece of wider tube is sealed to There must also be a wider piece at the top to admit the thermometer.x INFLUENCE OF WIDTH is. the weight of liquid and vapour in the still-head increases with the diameter of the tube. so also is the efficiency. depending on the nature of the liquid distilled. and when the diameter is rather when it is either increased or diminished. firstly. and it is therefore clearly more advantageous. though the weight of liquid and vapour is only 1-05 as compared with 3'55 grams. however. This blocking is much less liable to occur if the bottom of the narrow one. out that the inefficiency of the plain due to the want of thorough contact and the condensed liquid. especially on its boiling point. length the efficiency less is For tubes smallest rises of approximately equal than 14 mm. It will be noticed also that the narrowest tube gives a better result than the tube of medium diameter nearly twice as long. to use very narrow rather than very wide tubing. the blocking may be prevented by slightly diminishing the amount of condensation. of the It has been pointed vertical still-head is between the vapour to the central portion vapour passing rapidly up the tube and possibly never meeting with the liquid at all. Condensation in the still-head may be diminished by covering it with cotton wool or any other non-conducting material. occasionally useful. to the con- M . owing. for if the liquid in the tube is just on the point of joining up into columns. The device is. The diameter of the tubing cannot. the condensed liquid will unite into columns which will be driven bodily upwards. Modifications of the Still-Head. and.

admixture of down-flow of condensed liquid should therefore increase efficiency. . tion that can be The simplest and most obvious is altera- change the slope of the tube This may be done so as to retard the down-flow of liquid.162 FRACTIONAL DISTILLATION liquid densed Consequently. the tube near the top and bottom so that its ends by bending remain vertical. of Still-head. while the middle part slopes very gently to made upwards. any modification that brings about better the ascending vapour or that retards the its Sloping Still-Head. It will be seen from Table 47 that. by altering a tube in TABLE Nature 47. flowing very rapidly back to the still.

The increase with this apparatus very marked. but when it is it is advisable with and many other forms tilt of still-head to disconnect the condenser and to and still-head. The quantity of liquid is pour in the still-head the and vasame The mm. Still-Head. in efficiency is . the whole of the liquid usually returns to the still.3 described. liquid returns almost completely to the of special importance to avoid all loss. the still M 2 . while hot. as regards the separation of the benzene. the much heads constrictions in outer tube. is slightly diminished. "Rod and Disc" in Fig. manner the quantity of liquid in the still-head Spiral Still-Head. although the down-flow of liquid is retarded. as will be seen from the second part of Table 47 it is easily constructed and is very convenient to handle. The efficiency is further improyed by bending the sloping portion of the tube into the form of a spiral. 35. rents are produced in the ascending vapour. very moving On regreater. and this liquid is protected from the cooling action of the air at the same time eddies and cross cur. 35a the down-flow of part of the In the still-head shown condensed liquid is greatly retarded by the discs on the central glass rod.ROD AND DISC" STILL-HEAD the lfi. but the efficiency Fio. is as that in the plain tube of 19 '4 diameter. If that is done. this still. a notable improvement is effected. still-head is still further reduced. and it is especially remarkable that. probably because a better admixture of the vapour is thus produced by this device the amount of liquid in the . the source of heat. from side to side so as to facilitate the back-flow of liquid.

164 FRACTIONAL DISTILLATION slight A further improvement. Bulb Still-Heads. both in efficiency and as regards the quantity of liquid in the still-head. . the greater the number of bulbs blown on the higher is the efficiency. A vertical tube with a series of bulbs it was recommended by Wurtz (3). Better contact between the vapour and the condensed liquid on the outer tube is thus ensured. For a given diameter of tube and bulbs. but the greater also is the quantity of liquid and vapour in the still-head (Table 48). 48. is effected by constricting the outer tube between the discs (Fig 35&).* TABLE Nature of still-head.

and its velocity of down-flow is no greater in the bottom bulb than in the top one. The " pear " still-head is more efficient than the " rod and " disc tube of the same length and possesses the same advan- tages except that It it is somewhat more difficult to construct. result of this alteration. Greater efficiency. but chiefly as regards the quantity of condensed liquid. It " is probable also that the eddies in the vapour are greater than in the ordinary bulb tube. that collects on a depression (that is to say. after. The liquid. may The be especially recommended for liquids of high boiling point. 36). for a given efficiency. collects on the depression in the bulb below and falls in drops Flo 8tm near the middle of the bulb. but the quantity of liquid and vapour in the still-head was half as large again. mixing with the liquid liquid in condensed in that bulb and flowing down the sides with increasing velocity. instead of spreading itself over the inner surface of the bulb below. is attained by the "evaporator" still-heads. The " Wurtz still-hea/1 may be improved. to some extent in efficiency. by blowing pear-shaped instead of spherical bulbs on the tube As a (Fig. a given vertical height. Still-Heads. on the other hand. flowing past the constriction. and less condensation. The liquid on the '^d* inner surface of each bulb is merely the small amount condensed in that bulb. " Evaporator " for . the condensed any bulb.x "PEAR" STILL-HEAD The tube with thirteen bulbs 165 efficient was somewhat more that the slightly shorter tube with rod and discs. " The "Pear" Still-Head. the total quantity condensed in the part of the still-head above it) is brought well in contact with the ascending vapour in a part of the bulb that is less exposed to the cooling action of the air than any other.

37. of 7'5 to mm. 37. open at each end and widened below into the form of a funnel which rests on the constricted part of A and may be prevented from fitting it too accurately by fusing three or four minute beads of glass to the rim of the funnel. ss. Still-Head. a large amount of condensation takes place. 38. Original The general form of the apparatus. c. . The rises vapour therefore up the inner tube. shown in Fig. then passes down . the vapour first reaches a section from below.166 FRACTIONAL DISTILLATION Form of "Evaporator" in Fig. internal diameter and 60 mm. sides of B near the top 3. CHAP. is by a small funnel. as originally designed. where the inner tube rests on the constriction in the outer one. of 22 to 24 mm. " The ' ' evaporator . PIG. of each 8 An inner thin-walled tube. When one section of 1*" orator'-Ttm-hrad! and the narrow passage. centred and kept in position by the little just glass projections shown in the same figure. like a small inverted test Above the tube c and attached to it three glass legs. (B'). long. The outer tube. connected above and be- low with other sections. is shown Each section consists of three separate parts 1. wider than the depression in the tube A The tubes B and c are above it. An intermediate tube. becomes at once blocked by the condensed liquid. still-head original form. internal diameter and 50 mm. long. 2. D. the length being about 10 or 10'5 cm. c v which must be a little tube. Two large are blown on the . holes. of about 14 mm. internal diameter. B. A. Fio.

owing to superheating of the vapour. and the liquid then spreads itself as before over the . when. falls tube and spreads itself over its surface. vapour that rises through the inner be hotter than that which reaches the section The condensing point of the vapour in the inner tube must. rendered evident by the appearance or disappearance of such drops. falling again in drops from the bottom of this tube and finally flowing through the passage D. In this. " Modified " Evaporator A modification of Still-Head. the condensing point of a mixed vapour becomes lower and lower as the vapour rises through the still-head thus. they may be formed in the lowest section. the spheroidal drops are never seen unless the quantity of liquid in the still is very small. in any section. 39. the "evaporator" still-head is shown in Fig.x "EVAPORATOR" STILL-HEADS 167 between the inner and middle tubes and finally up again between the middle and outer tubes and so into the section above. be higher than the boiling point of the liquid that falls from the little funnel. With a pure liquid. and the progress of a distillation is. bulb and the top of the middle tube c is blown into a flattened c'. and when the two components them there is differ considerably in volatility and neither of in great excess. evaporation of the liquid on the middle tube is may be easily observed. indeed. on which the drops of liquid from the depression above fall and collect into a shallow pool which soon overflows. the tube will above. liquid in any section collects together and in drops from the depression in the section below into the funnel from this it falls on to the top of the middle The condensed . Owing to gradual removal of the less volatile component. indeed. the rather fragile funnel on the three legs is done away with. . Under such conditions a tendency for the drops of liquid from the funnel to assume the spheroidal state.

and for a given height of still-head the effi- ciency is greater. be seen from Table 49 that very good results are obtained with the evaporator still-heads of either form. while. "Hempel" apparatus (5) on the other hand the amount of liquid in the still-head is The "Hempel" Still-head. the the weight liquid is and vapour in still-head usually slightly less. and constricted A short. readily formed in this apparatus. narrower. One section of increase the vertical distance from the the modified "evap" orator still-head. of for a given efficiency. the liquid flowing down the sides of the inverted tube breaks up into separate streams. The still-head consists simply of filled a wide vertical tube. but is Spheroidal drops are not nearly so when rapid separation taking place. surface of the tube. so as to Fio. It will The sections in the modified apparatus are shorter than in the original one. The great advantages of are simplicity and efficiency. to the bottom of the inner tube in the modified still-head the vertical distance from the depression to the flattened top of the inverted tube should be somewhat increased in the lower sections. below to prevent the beads from falling out. bottom of that tube . the excessive. When there are inverted tube should be many sections. and it is therefore unsuitable for the distillation of small quantities of liquid. with glass beads of special construction.168 CHAP. 39. From the following table it will be seen that in efficiency . the made somewhat shorter in the lower sections. vertical tube with side delivery tube is fitted by means of an ordinary cork into the wide tube.

1G9 Nature of still-bead. .HEMPEL" STILL-HEAD TABLE 49.





was a

But the weight of liquid in the 3-section longer. " still-head was considerably less than half, and in 5-section " evaporator" not much more than half as great
as in the
tube. " still-head for length, the modified " evaporator Length would be more efficient, and would contain only about half










Young, "On the Relative Efficiency and Usefiilness of various Forms of Still-head for Fractional Distillation, with a Description of some New Forms possessing Special Advantages," Trans. Chem. Soc., 1899, 75, 679. " The Brown, Comparative Value of Different Methods of Fractional Distillation," Tram. Chem. Soc., 1880, 37, 49. " Memoir on Wurtx, Butyl Alcohol," Ann. Chim. Phys., 1854, III,
42, 129.


"Comparative Investigations on the Methods

of Fractional


Distillation," Liebig's Annalen, 1884, 224, 259. Hempel, "Apparatus for Fractional Distillation,"
Zeitschr. fiir Anal. Chem., 1882, 20, 502.





of the still-heads

example the Coffey

employed on the large scale, (Fig. 58, p. 202) the condensed

liquid is made, by into shallow pools,

means of suitable obstructions, to collect and the ascending vapour has to force its

way through

these pools very good contact is thus brought about at definite intervals between vapour and liquid. The

excess of liquid
pool, tubes. "



finally to the

carried back from pool to still by suitable reflux

It is

convenient to reserve the term
for this particular class of



The "Linnemann" Dephlegrnator.



dephlegmator employed in the laboratory devised by Linnemann (1). number of


cups of platinum gauze,


were placed at

different heights in the vertical tube (Fig. 40) and the liquid collected in these, but, as no

reflux tubes

were provided, the liquid gradually accumulated in the still-head until the quantity became unmanageable, when the distillation had to be discontinued until the liquid Fl(1 40 _ The flowed back to the still. There was thus do much waste of time, and increased loss of material by evaporation, and it was impossible to make an
accurate record of the temperature.




are provided,

In the more recent forms of dephlegmator, reflux tubes and it is on the size and arrangement of these
tubes that the efficiency and usefulness of the still-heads chiefly depends.

The "Glinsky" Dephlegmator. The Glinsky dephlegmator (2) has only one reflux tube, which carries the excess of
liquid from the large bulb to the tube below the lowest obstruction, practically back to the still. In its original form
(Fig. 41), the dephlegmator otherwise resembled the Linnemann still-head very

closely, but, as now constructed, there are bulbs on the vertical tube and spherical






cups, rest on the constrictions between





The "Le Bel-HennmDephlegrnator. In the Le Bel(3)





Henninger apparatus

the obstruction


usually caused by placing platinum cones on the constrictions between bulbs blown on the

each bulb is connected by a one below it (Fig. 42) so that the liquid is carried back from bulb to bulb and not straight to the still. In these dephlegmator s, unlike the Coffey
vertical tube

reflux tube with the

the reflux tubes are external, and the returning liquid is thus exposed to the cooling

Fio. 42.

The " Lc








The dephlegmators of Young and Thomas (5), (6),

J5el-Henninger dephlegmator.

follow the principle of
reflux tubes being


the Coffey still fnore closely, the shorter and being heated by the

ascending vapour.




The "Young- and Thomas" Dephlegmator. The Young and Thomas dephlegmator (6) is shown in Fig. 43 a,




consists of

glass tube of about 17


a long, internal

diameter, with the usual narrow side delivery tube. In the wide


tube are sharp constrictions, on rest concave rings of plati-


by being heated
tubes, T, of the

gauze, B, previously softened to redness, and

these support small, glass reflux



form shown in Fig. The upper and wider part

has an internal diameter of 4-5 mm., the narrow U-shaped part, which serves as a trap, an internal diameter of 3 mm.
of the reflux tube

FIG. 43.

Thomas dephlegmator. however, the number of constrictions exceeds 10 or 12, the traps for the lower reflux tubes should be slightly wider, say, 3'5 mm. The length of the reflux tubes should be about 45 mm., and the distance between two constrictions about 60 to 65 mm. The enlargement, A, on the reflux tube prevents it from slipping through the ring if the tube is inverted, and the reflux tube and ring together are prevented from falling


The " Young and

out of position by the five internal projections (made by heating the glass with a fine blow-pipe flame and pressing it inwards with a carbon pencil) one of which, B, is shown in




horizontal section through the tube at B



in Fig. 43 c.

Comparison of Dephleg-mators.


On comparing

the efficiency of the three dephlegmators each of three it was ftmnd that when a large quantity (400 sections grams) of the mixture of benzene and toluene was distilled,


Le Bel-Henninger

still-head gave slightly better results




than the Young and Thomas, and -both of these distinctly better results than the Glinsky. On distilling 50 grams

Young and Thomas dephlegmator was with the Le Bel-Henninger tube the residual toluene was not quite pure, and with the Glinsky it contained a quite appreciable amount of benzene, although
of the mixture, the

found to be the best


the temperature reached 110'6 in all three cases. With 25 grams of the mixture the differences in efficiency were much accentuated, and it was only with the Young

and Thomas dephlegmator that the temperature reached 110'6, though even in this case the toluene was not quite The highest temperature reached with the Glinsky pure. still-head was 107- 6, and with the Le Bel-Henninger,
107' 35, but the residual toluene was far less pure in the former case, for, on distillation from a small bulb, the Glinsky residue came over between 102-2 and 110-4, and the Le

Bel-Henninger from 105-7 to 110-6, and the Young and Thomas from 110-4 to 110'6. With the "rod and disc," the "pear" and the "evapor" ator (3 and 5 sections) still-heads, nearly as good results were obtained with 25 as with 50 or more grams of the mixture, and in all these cases the residual toluene was
quite pure.
It will

thus be seen that, for small quantities of liquid, the

Young and Thomas dephlegmator gives better results than the Glinsky or Le Bel-Henninger, but that none of them are
so satisfactory as the other forms of still-head. The relative efficiency of the three dephlegmators


400 and 25 grams, respectively,

of the

mixture were

and the

effect of increasing the


of sections

in the case of the in Table 50.

Young and Thomas
results in the last


still-head are shown column were obtained

by taking the
in a

distillation at

this case, 42-6 out of the

only half the usual rate. In 50 grams of toluene were recovered

pure state in the single distillation, and 33'7 grams of benzene were obtained with a temperature range of only 0-5.



^-CM-*pCMC: t-CO^^OCMaOp




35 CO

i-H i-l

O O rl




i I









O 53




M rl CM CO




S li O









O t O CM c
^O Q7^









TC <N 0)




O o O


^ t? s s -





^ H



















Width of Reflux Tubes and Depth of Traps. The than reflux tubes is that part which is necessary also the U-shaped parts. not be made wider in filled with liquid during the distillation to freely carry back the condensed liquid . it follows that. therefore. at the bottom. should be no deeper than is required to prevent the ascending vapour from forcing its way through them. There is thus a tendency. should. for the quantity of liquid in the pools in the lower bulbs to be unnecessarily large. for the liquid to flow past the upper cones without forming a pool. on the one hand. and the platinum cones largest. in order to retard the flow sufficiently for a pool to be formed. consistently with efficiency. &e. however. The results of the experiments which have been made serve to indicate the requirements which should be - fulfilled in order that a dephlegmator may give the best possible results. 1. been pointed out that. increase the Number of Sections. at the top of the tube. In the Young and Thomas apparatus it is advisable to make the upper constrictions somewhat deeper than the lower ones. Additional width or depth simply means waste . the constrictions are frequently made widest. 1 General Remarks on the Construction of Dephleg mators. the amount of condensed liquid in the still-head during distil- lation should be as small as possible. It has 3.176 FRACTIONAL DISTILLATION CHAP. This requirement is best fulfilled by the and the " Young and Thomas " dephlegmators. It should be possible to greatly number of sections without seriously adding to fragility the difficulty of construction or to the of the apparatus. on the other hand. more complete obstruction is necessai y at the top of the tube than In the Le Bel-Henninger dephlegmator. As the amount of condensed liquid flowing back at any level is greatest at the bottom of a still-head and least at the top. Size of Constrictions. acting as traps. "Brown " 2. and.

when formed. Flow of Liquid through the Reflux Tubes. suppose that there were no back flow at all through the reflux tubes and that the traps simply became filled with the first This most volatile liquid portions of condensed liquid. the Young and Thomas is the best in this respect in the Glinsky apparatus. were probably largely due to the excessive width and depth of the reflux tubes. when only 25 grams of the mixture was distilled. are liable to drive out the liquid. and it is for this reason that. slow. To take an extreme case. the flow of liquid was exceedingly . the last fraction might consist of pure toluene. the utmost importance that there should be a rapid flow of condensed liquid through the reflux tubes. when only 25 grams N of . On the other hand. The poor results obtained with the Le Bel-Henninger dephlegmator. It is of 4. the distillation. of a mixture of benzene and toluene. and also to prevent bubbles or columns of vapour from being caught and carried down with the liquid through the traps. the upper part of the reflux tubes may be advantageously made fairly wide so as to facilitate the entrance of the condensed liquid. especially if they are outside the main tube and are not heated by the ascending vapour. on the other hand. but the number very large. Of the three dephlegmators compared. while the residue at the end of the fractionation would be very rich in benzene. the lower traps should be made rather wider than the upper ones. for instance. and the ascending vapour ^may then pass more easily through the traps than through the pools formed by the obstructions.xi REFLUX TUBES if 177 of sections is liquid in the still-head. The weight of liquid and vapour in the still-head was more than twice as great as in those of Glinsky or Young and Thomas. and would then form part of the residue. would thus remain lodged in the traps until the end of the In distillation. Such columns of vapour.

indeed. In this respect. Le Bel-Henninger." the " " " Pear. so that the change in composition may be regular from bottom to top of the That is the case with the dephlegmators dephlegmator. but still. and this only. it is of importance that it should return as completely as possible from the still-head to the still. The weights of liquid actually left in the still-heads after cooling were as follows Glinsky. in order that the quantity of liquid in the pool may not become This is a point that is frequently unnecessarily large. Brown. grams . and Young and Thomas. but it is not nearly so marked. 0'2 gram . is richer in the more volatile component of the mixture than the liquid in the lower pools. 5. Arrangement of the Reflux Tubes. Return of Liquid from Still-Head to Still after the When the residual liquid is Distillation is Completed. quite unsuited for the distillation of very small quantities of liquid. 6. the upper end of the reflux tube should be wide and the junction with the bulb should be low down. / In fractionating tubes of the Glinsky or Le Bel-Henninger type. the residue was so much richer in benzene than the last portions of distillate. overlooked. The same fault is to be observed with the Le Bel-Henninger dephlegmator. the Glinsky dephlegmator gave the best results. which is returned to the Thus the still through the reflux tube. The Glinsky apparatus is. but the amounts of liquid left in the " Bod and Disc. Young and Thomas." and the Evaporator tubes were still smaller. the mixture were distilled. a reflux tube connecting each section with the one below it. there should be To get the best results. Of the various still-heads that have been described.178 FRACTIONAL DISTILLATION CHAP. of Le Bel-Henninger. in the Glinsky apparatus there is only one reflux tube con- necting the top bulb practically with the condensed liquid. valuable. it . 1*4 : 0'55 gram.

Still- The relative efficiency of the different still -heads is well shown by the following comparison of the number of fractional distillations with a plain vertical tube of 30cm. but for great efficiency the evapo" rator still-heads give the best results.EFFICIENCY OF STILL-HEADS 179 efficiency is to be most may be concluded that. required. Comparison of Improved with Plain Vertical Head. height. the when only moderate "Rod and Disc" or "Pear" is " strongly recommended. . which give the same result as a single distillation with the improved apparatus.

a fall of a fraction of a degree would cause complete condensation of the vapour. as has been already pointed out. or if one component is present in large excess. Warren's Still-Head. The spiral tube was heated in a bath of water or oil its length varied from to \ inch. and in the second place. kept at a . When a be thereby greatly increased without unduly adding to the height of the bath. it is better. in either case. while a rise of temperature to a similar extent would prevent any condensation from taking place and there would be no the separation fractionation at all. for. for two reasons. The temperature of the bath. requires very careful regulation if the boiling points of the components are near together. the effective length of the still-head . the surrounding the still-head with water or any other temperature of which is kept as little above the boiling point of the more volatile component as will allow of vapour passing through.CHAPTEE XII MODIFICATIONS OP THE STILL-HEAD (continued) "BEGULATED" OR "CONSTANT TEMPERATURE" STILL-HEADS By liquid. an elongated tempera- was first recommended by Warren (1). a considerable improvement in is effected. The employment of constant or slowly rising spiral still-head. 1 to 10 feet and its internal diameter from . however. ture. the spiral form is more efficient than the may vertical. to bend the tube into the form of a spiral in the first place. regulated temperature still-head is employed.





carried out fractional distillations of petroleum

and other complex mixtures, and observed that, as the fractions became purer, the temperature of the bath had to be
brought nearer to the boiling point of the liquid in the

and required more

careful regulation.

modification of this apparatus, devised with a view to the better control of the temperature

Brown's Still-Head.
the still-head,




scribed by
liquid boiled




be distilled in the vessel


(Fig. 44)

through temperature of which

the vapour rises the coil c, the

that of a liquid boiling in the vessel E. The vapour

from the liquid in B passes through the tube D to the worm condenser F, and the condensed liquid returns by the tube K to the bottom of the vessel E. The pressure under which
the jacketing liquid boils
regulated and measured by a pump and gauge conis

nected with the tube

Fio. 44.


E is heated by the ring burner B. The vapour of the liquid which is being distilled
liquid in

Brown's " regulated temperature"

passes through the side delivery tube G and is condensed and collected in the usual manner. The temperature of the vapour



leaves the still-head is registered by a thermometer at that of the jacketing liquid is regulated by the pressure




and may be read from the vapour pressure curve, or a second thermometer may be placed in the central tube. Brown obtained very good results with his apparatus, and
in distillaarrived at the important conclusion (p. 86) that tions with a still-head maintained at a constant temperature, the composition of the distillate is constant, and is identical

with that of the vapour evolved by a mixture whose boiling
point equals the temperature of the still-head." In Brown's apparatus the temperature of the still-head

can be kept constant for any length of time or
altered rapidly





altering the pressure

can be under
to of

liquid is boiling, but, the fact that the number of observations

which the jacketing

owing perhaps and the amount

attention required during a distillation are greater than when the still-head is merely exposed to the cooling action of the
air, it

has not come into general use.

liquids there can be ture still-head is the

very Volatile Liquids. For very volatile no question that the regulated temperamost suitable, and the principle has been and Dewar to the purification of gases. applied by Ramsay Thus, if a mixture of helium with any other gases is passed
Suitability for

through a spiral tube cooled by hydrogen boiling under reduced pressure, all other known gases would be condensed in the tube and the helium alone would pass through.
at the ordinary pressure


Separation of Pentanes from Petroleum.

A regulated

temperature still-head, combined with a dephlegmator (Fig. 45) has been found very useful for the separation of the lower paraffins from American petroleum (3). The vapour " from the boiling " petroleum ether passed first through a

column Young and Thomas dephlegmator, on leaving which its temperature was read on a thermometer, A. It

then passed upwards through a spiral tube in a large bath, the water in which was either cooled by adding ice, or warmed




by a ring burner below, and was kept constantly stirred by an arrangement similar to that employed by Ostwald, in which a propeller, with four blades of thin sheet copper, was kept rotating by a windmill actuated by a ring of gas

Fio. 45.

Combined " regulated temperature "


and " Young and Thomas




the efficiency of the windmill being greatly increased by the use of a chimney. The temperature of the bath was registered by the thermometer B.
After leaving this part of the apparatus the vapour passed




through a vertical tube, where its temperature was read on the thermometer c, then into a spiral condenser cooled by ice the condensed liquid was collected in ice-cooled flasks. The general course of this separation has already been

referred to

in Fig. (p. 138), and is indicated by the curves Details of a single fractionation, the eleventh, are given in Table 51 the temperatures are the final ones for the


and they are all corrected to 760 mm., and for the thermometric errors. Since the temperature of the vapour is
affected by changes of the barometric pressure, that of the water bath has been altered in each case to the same

The lowest fractions consist chiefly of isopentane the middle ones are mixtures of iso- and normal pentane the highest consist of nearly pure normal pentane.










noticed that the temperature of the vapour before condensawas in all cases nearly the same at that of the bath,
generally very slightly higher. It will also be seen that the fall in temperature of the vapour, during its passage through the

regulated temperature still-head



greatest for the

middle and least pure fractions and smallest for the highest and purest indeed, for the last fraction the three temperatures are almost identical, while for the fourth there is a

difference of 1*2


D and



careful regulation of the temperature of the bath


necessary when

the liquid distilled

was nearly










Liebiffs Annalen, 1865, Suppl., 4, 51. " The 2. Brown, Comparative Value of






Distillation," Trans. Chem. Soc., 1880, 37, 49; Fractional Distillation with a Still-Head of Uniform Tempera-

ture," ibid., 1881, 39, 517.

Young and Thomas,
1897, 71, 440.

"Some Hydrocarbons from American Normal and Iso-pentane," Trans. Chem. Soc.,



Continuous Distillation on Large Scale. Continuous diswort has long been carried out on the large scale by means of the Coffey still (p. 202) but it is only quite recently that any attempt has been made to devise a process of the kind suitable for laboratory purposes.
tillation of

Carveth's Apparatus. Carveth (1) suggests that by maintaining two parts of a system at different temperatures,
corresponding to the boiling points of the two components of a mixture, it should be possible to effect continuous separation, especially if use were made of dephlegmating intercepts. He describes an apparatus used by Derby, in 1900, for It consists (Fig 46) of a mixtures of alcohol and water.


long block-tin tube surrounded at its lower end, A, by the of boiling water, at its upper end, B, by vapour from The mixture of boiling alcohol, and filled with intercepts.

was slowly dropped in at the vapour passing to the condenser at D, and the residue c, through the trap at E. Carveth, however, gives no details regarding the length or diameter of the tubes or the nature
alcohol and water to be distilled
of ;the intercepts. So far as the lower half of the still-head is concerned, there is no objection to the use of steam as a jacket, the



point of

at being to keep the temperature up to the water in order that the alcohol may be

vaporised as completely as possible the great amount of heat evolved by the condensation of steam makes that

it easily becomes superheated and the still-head is not prevented from rising in temperature at all indeed the effect is very similar to that produced by covering the tube with cotton-wool condensation is somewhat diminished and so also is the efficiency. Fio. and it is a mistake to suppose that this " at a part would thus be maintained temperature corresponding to the boiling point of alcohol. if a vapour. but in five cases for which he gives details the percentage varied from 1-5 to 2 '0. Carveth's still for continuous distillation. in order that the vapour of water may be condensed as completely as possible. . such as that of the more volatile component. This can be done by surrounding the still-head with a liquid required which is kept at the required temperature by suitable means. These five distillations were carried out with extreme but is . used as a jacket. 46. Carveth makes the general statement that the percentage of alcohol by weight in the residue was found to average about 0'5 when working carefully. XIII CARVETH'S APPARATUS 187 substance a very efficient heating agent.CH." What is here is to keep the temperature as far as from possibly rising above the boiling point of alcohol. while that of the alcohol passes on to the condenser. But the use of the vapour of alcohol for the upper part of the still-head is wrong in principle.

arranged in two spirals which are mounted in separate pails. never containing more than 0'5 per cent. the average rate being only 7'5 grams of distillate per hour. scribes Lord Eayleigh (2) dean apparatus similar in principle to Carveth's except that the temperatures of both parts of the still-head are regulated by liquids and not by vapours. 40.188 FRACTIONAL DISTILLATION < H. the whole length of tubing sloped gently and uniformly upwards from near the bottom of the lower the mixture could be introduced. the upper one by water maintained at a suitable temperature. were distilled and in all cases the water. of alcohol. to Again. 60 and 75 per cent. a rapid but visible succession of drops and. collected in the lower receiver. of the efficiency of the apparatus Lord Rayleigh's Apparatus. 15 mm. but in the five very slow distillations referred to above the percentages varied from 76'7 to 89'9. The mixture was introduced at such a rate that it fell in pail to the top of the alcohol. The lower end of the tubing was. It does not appear possible to form any very definite idea from the data given. in diameter. was nearly pure. slowness. connected with an air-tight receiver heated to 100. . usually 77. the percentage of alcohol in the distillate is stated have varied from 90'6 to 93'9 in four cases. upper pail. The spirals were connected by a straight glass or brass tube of somewhat greater bore. of alcohol. The distillate varied but little in strength and contained from 89 to 90'3 per cent. For the distillation and water the lower and longer spiral boiling water. when rich in it was previously heated. Mixtures containing 20. if necessary.M-. The apparatus consists of a long length (12 metres) of copper tubing. where it turned downwards and was connected with an ordinary Liebig's condenser. provided with a lateral junction through which of mixtures of alcohol was heated by With the exception of the two extremities. of alcohol.

bath would be kept at 57-1. the middle one at 77-15 and the lowest at 101-55. into the still. . B. It would. where it would be boiled. be possible to devise an arrangement by which three components could be separated in a similar mafaner. 47. 47. 57-1. moreover. which might be heated by a ring burner. Suppose. that the liquids to be separated were methyl. 77-15 and 101-55 respectively). Still for continuous distillation of throe components. p. The propyl acetate would be collected in the receiver D.) diagrammatically in Fig. The mixture would be introduced slowly through the funnel at A.SEPARATION OF THREE COMPONENTS 189 It seems probable that such a continuous process may prove very useful for mixtures which separate into only two components and when the quantity to be distilled is large. for example. the form of apparatus required is shown Fia. Lord Rayleigh's (Adapted from still. ethyl and propyl acetates The highest (b. and the mixed vapour would enter the still -head at c. but it would be necessary to have two coils for the middle substance . Continuous Separation of Three Components.

and two coils in each bath except the lowest and highest. 521. 1. The vapour nearly pure methyl acetate. Mag. and the components of a still more complex mixture should be separable. chiefly in the form of vapour.." Part II. from propyl acetate and the mixture of ethyl and methyl acetates would enter the top of the second coil. 4. Chem." Phil. Vapour Composition.190 FRACTIONAL DISTILLATION CH.. Journ. Rayleigh. if there were as many baths as components kept at the boiling points of those components. Phys. 6. 2. [VI]. " Studies in Carveth. and was collected in E. would be . . 1902. "On the Distillation of Binarj' Mixtures. 1902. It should thus be theoretically possible to separate all three liquids in a pure state. REFERENCES. 253. The condensed liquid which reached the bottom of the coil. would be nearly pure ethyl while the vapour that reached the top of the coil acetate in the highest bath. and was condensed in F. more or less completely. xin in passing upwards through the first coil in the middle bath would be freed.

CHAPTEE XIV FRACTIONAL DISTILLATION WITH AN IMPROVED STILL-HEAD Benzene and Toluene. in height. . was added to the residue in the still. and when fraction No. In the third fractionation 21-9 grams of pure benzene were obtained. When the "evaporator" still-head was used it was not necessary to divide the distillate into more than five fractions. now More than 82 per cent. That much time is saved by the use of an improved still-head is seen by comparing the results of a fractional distillation of 200 grams of a mixture of fied equal weights of benzene and toluene with a modi"evaporator" still. and the lected in receiver No. a little liquid came over at the boiling point of benzene. 2. collected from 81'4 to 95-4. the temperature did not rise the whole of the distillate was therefore coland the small residue (4*95 grams) The apparatus was then dried. but it was not considered The value of Aw/A advisable to collect this separately. 3. In the first distillation (Table 52). In the second fractionation. was continued. and it will be noticed that the values of Aw/A for the middle fractions are relatively very low. was rejected. for the first fraction was more than 300 times greater than for the fourth.head of 5 sections with those obtained with a plain vertical tube 30 cm. more than half the total quantity of toluene was recovered in a pure state. of the toluene was obtained in a pure state. and the distillation above 81 -7 .

. Two more distillations of the lower fractions sufficed to complete the recovery of the benzene.192 FRACTIONAL DISTILLATION CHAP. The distillate was too small to allow of further fraction ation. higher fractions were distilled. TABLE 52. when only 7 '25 grams came over below 110-6. the residue (9*3 grams) consisting of pure toluene. I.

when fraction 3 from II. was. . howture had risen only to 81-5. added to the residue in the still as usual and. a considerable amount was collected below but the quantity that came over between this temperature and 100 was very small. the temperaFraction 8 from I. &c . . in height. . the grams) was temperature had only risen to 65. Methyl.xiv SEPARATION OF THREE ESTERS 193 Comparative Results with Plain and Improved Still-Heads. was also much less actual amount left undistilled was loss of material. Weight of pure benzene recovered . 81 '8 grams hours . 4'0 grams 22 '0 grams. Evaporator still-head Plain still-head 30 cm. had been distilled over as completely as possible. 120 grams of ethyl acetate and 100 grams of propyl acetate was carried out by distillation with a modified evaporator still-head of 8 sections. but fraction 4 from II. ever. though the greater. The great improvement effected by employing an efficient apparatus is clearly shown in Table 53. The chief gain was in the time occupied. Ethyl and Propyl Acetate. had been distilled as completely as possible. to recover as much propyl acetate as possible. but the recovery also was somewhat there better.. on distillation. (5-15 Similarly. About 30 hours. Loss by evaporation. 194). and all that came over below 101-55 rejected. .. The separation of a mixture of 100 grams of methyl acetate. 7 '8 Mixture left undistilled 22 -25 . . in the third fractionation. which was reduced nearly to one-fifth. This last fraction was therefore merely redistilled 89-35. . 7 from I. . and it was clearly not worth while to attempt to recover any more ethyl acetate in the next fractionation from the eighth fraction of II. and in many cases would be much better . of 5 sections. toluene Time required for actual distillation. In the second fractionation. . Details of the fractionations are given in Table 54 (p. TABLE 53. when fraction No. 91'95 6J 81 '4 grams 88'8 .


76 '8 79 '8 94 -2 6 17 Lost by evaporation and transference Lost by treatment with reagents .. and the residue was rejected. . In the long continued fractionation with the plain stillhead. still-head of 8 sections. . Before the fractionations were completed it was necessary to treat the propyl acetate with potassium carbonate and the methyl acetate with phosphorus pentoxide. In IV. . " With the " evaporator apparatus. Left undistilled 21 5'6 17'1 - . 53 '5 72 '5 32 . o 2 . numbered III. was separately distilled. was distilled . and the loss was thus increased. the propyl acetate showed no acid reaction whatever. but in each case consisted of two parts.. when the recovery of propyl acetate was completed.. the first part of the fractionation ended when the temperature reached 58. . .. Time required (1) (2) (3) in hours Percentage weight of material About 70 . were therefore not continuous. .. fraction 6 when the highest was therefore taken to be pure ethyl acetate. TABLE 55. "Evaporator" Long plain still-head. Percentage weight of methyl acetate recovered .. to VI. The Comparative Results with Plain and Improved Still-Head. the distillation was stopped when the temperature had reached 77'7 the still and still-head were dried. . great superiority of the "evaporator" still-head is clearly shown by the results given in Table 55. 48*1 Number ethyl acetate .xiv COMPARATIVE RESULTS 195 The fractionations stopped. and the last fraction from II. propyl acetate of fractionations required . Thus in III. and the methyl acetate was only slightly moist. In this case the tem- perature did not rise above 77*15.. and was recommenced by the . distillation of fraction 4 from III.. .. on the other hand. a considerable amount of moisture was absorbed by the esters and some hydrolysis took place. even fa-action from III. 2 '8 O'O 13'8 .

or glass. The still is usually made of copper or some other metal.. and so on. The pressure in this case was 70 mm. be employed for distillation under REFERENCE. for nitric acid. for the distillation of sulphuric acid. Francis and Young. stoneware. from CHAPTER XV DISTILLATION ON THE MANUFACTURING SCALE The Still. 73. are employed for bromine and iodine. as in the preliminary distillation of coal tar. but the vapours pass directly from the still to the condenser." Trans. Soc. but for special purposes other materials must be employed. cast . 1898. When only a rough separation of the constituents of a complex mixture is required. 136. 920. THE STILL-HEAD. . was used by Francis (1) for the separation of isoheptyl and normal heptyl bromides from the products formed by the action of bromine on the distillate from American petroleum coming over between 93 -5 and 102. dephlegmators or rectifiers are not used. " Separation of Normal Isoheptane American Petroleum.196 FRACTIONAL DISTILLATION CHAP. 1. vessels of glass or platinum are used fron is found to be the best material stills . Fractional Distillation under Reduced Pressure. The specific gravities were again determined and were found to agree well with those given on p. still-heads The im- proved may A 12-column Young and Thomas dereduced pressure. for example. Chem. phlegmator. Thus.

to the dephlegmator. Mansfield's still. A. The vapour. B. (Fig. this device adopted still in Mansfield's 48). E. Coupler's Still. the is opened and vapour then passes directly to the condenser. while that from the other bulbs reaches it successively at higher levels. into a series of bulbs placed in a cistern. D. modified still-heads are employed.MANSFIELD'S STILL 197 For the better separation of the components of a mixture. after- wards the stopcock. 49). containing brine. which is often used. the temperature of is allowed to rise to a suitable extent is . A simple contrivance. A. . which may be warmed by steam from the pipe. a more elaborate cooling arrangement is combined with a dephlegmator. With this still. The liquid condensed in the bulbs returns by the pipes. F. which to cool the still-head with water. Mansfield's is Still. by water which becomes heated surrounded to its boiling point. c. passes first through the rectifier or dephlegmator. F. 48. is The head. B. In Coupier's still (Fig. Mansfield was able to separate benzene in a fairly pure state from coal tar. and then by the pipe. FIG. the less volatile vapours being for the most part condensed in the first bulb and the liquid returning to a low part of the dephlegmator. The vapour from the still. Liquids which boil at temperatures higher than 100 are for the most part condensed in the still-head and return to the still .

E. then passes by the pipe. H. 49. into the tank.198 FRACTIONAL DISTILLATION freed from substances boiling at higher temperatures than that of the water in the tank. A the . the steam-pipe. passes through the and then enters the condenser. G. Here the water receives more heat owing to condensation of vapour in the pipes. series of pipes in the tank. G. Cold water from the cistern. which is divided into sections by . and in its passage upwards is warmed by the condensation of the vapour the warm water passes through the pipe. and the liquid . in passing from the rectifier to cooled by successive stages. J. enters the condenser at the bottom. F. to FIG. is the condenser. the condenser. Coupler's still. Thus the vapour. The liquid to be distilled is heated by a steam-pipe in the still. vapour D. A very similar arrangement is seen in the French column apparatus (Fig. B. so that it is hottest where it leaves the tank by the pipe. rises through the rectifier. c. the vertical partitions shown in the figure. 50). The contents of the still are heated by French Column Apparatus.

a pool of liquid through which the vapour that follows has to DephlegrnatOPS. but its ascent is barred by the dome. The details of the various dephleg- mators differ. 51 to 55. for benzene. then through a surface or multitubular condenser which is provided with a water supply so regulated that its temperature is about that of FIG. Savalle's Still. considerably. as will be seen from Figs. and 140 for the condensed in the regulated temperature still-head returns as usual to the dephlegmator. B. force its way.DEPHLEGMATORS which returns pipe. and the In nearly all the dephlegmators. French column apparatus. separation of toluene and xylene from crude naphtha. 50. .-In the Savalle apparatus for the benzene. part of the vapour is condensed and forms fiers. purified vapour passes on to the cold condenser. Dubrunfaut's method is shown in Fig. Dubrunfaut's Dephlegmator. A. however. is to the 199 much lower part of the rectifier through the richer in the less volatile components than E that which reaches the top by the pipe H'. 51. the va- pours pass first through a dephlegmator or analysing column. about 80 xylenes. and it bubbles through the condensed liquid in the tray. so that very complete contact between vapour and liquid is ensured. H. the boiling point of the the temperatures in this case would be liquid required . The vapour rises through the central pipe. in the tray. rectior analysers. The liquid 110 for toluene.

200 FRACTIONAL DISTILLATION CHAT. 52 a. In many cases the are perforated. Dubrunfaut's dephlegmator. 51. The construction of the dome is indicated in the lowest tray in the figure. The liquid enters the tray by the pipe A. a FIG. c. the lower end of which is trapped by the liquid in the tray below. Fio. an arrangement somewhat similar to the dome is used. the vapour rises through the perforatrays Savalle's . Egret's Dephlegmator. and Coffey's Dephlegmators. 52 b. and follows the course shown by the arrows in Fig. 52. The excess of liquid flows back through the pipe. shown in Fig. 52). but the condensed liquid is made to follow a zigzag course from the circumference to the centre of each tray. In the Egrot dephlegmator (Fig. Egret's dephlegmator. In a large column a number of these domes are arranged in a ring on each tray.

a considerable improvement in the efficiency of this apparatus would be effected by the simple modification The condensed liquid. instead of flowing Fig. Coffey's dephlegmator. The "Pistorius" still. 53. " vapour entering a section the centre to the circumference by the flat dome. 57. 57. Modified 1'istorius still. iO. 53 and 54 show -$ FIG. A it from then FIG. dephlegmator. above it. dephlegmator. Here the in the Pistorius still (Fig. Fio. of the still-head is deflected . 55. the arrangement adopted in two of Savalle's stills the Coffey stilL which is described on p. the excess of which descends through the reflux tubes. is Fig. that in in shown The " head is PistorillS employed A quite different form of stillStill. 202. Savalle's Fio.THE "PISTORIUS" STILL 20J tions and bubbles through the liquid. Figs. shown in down the outer walls of the section would drop from the central tube . Savalle's Fio. 55. . passes back to the centre above the dome and is partially It is probable that condensed by the water in B. 56). 54.

D. CONTINUOUS DISTILLATION. . Various alterations in the still have been made in order to make the process of distillation a continuous one. and passes down the zigzag pipe. The " Coffey " Still. It . where it is heated by the ascending vapour then up the pipe. In this country very strong spirit is K Pio. 166) and the ascending vapour would come into better contact with this hot liquid and would cause some of it to evaporate again. 58. B. into the highest section of the dephlegmator or analyser. C c. the essential parts of which are shown in The wort is pumped from a reservoir. E.202 FRACTIONAL DISTILLATION to CHAl'. A. pipe. The "Coffey" still. obtained in a single operation from wort by means of the Coffey still. on the flat dome as in the " " evaporator still-head (p. 58. up the Fig.

In the upper part of the rectifier there are usually only shelves which compel the vapour to take the same zigzag course as the pipes which convey the wort downwards. G." Dictionary of Applied Chemistry. and is finally allowed to escape through the trapped pipe. however. vapour is. then passes through the pipe r/ to the condenser. 1. Alcohol. F. Articles on "Distillation. so that by the time it has reached is deprived of alcohol. Steam is causes the wort to the bottom it passed into the analyser by the pipe.xv THE COFFEY STILL 203 then descends from section to section through the tubes. H. Vol." "Alcohol. and the descending wort heated. and boil. K. On reaching the top of the column. by each washing. by weight of alcohol. Thorpe. t The ascending vapours pass through the perforations in the plates and bubble through the liquid on them." Pay en. II . 2. where it mixes with the descending wort. a portion of completely the aqueous vapour being thus condensed. the concentrated alcoholic vapour passes through the pipe J into the bottom of plates the rectifier. REFERENCES. containing about 91 per cent. not washed by wort in the rectifier but by the weak spirit formed by partial condensation of the vapour. The weak spirit condensed in the rectifier flows into a reservoir r M. and then ascends through perforated similar to those in the analyser the ascending ." Iitdustrielle. The purified vapour. from which it is pumped into the top of the analyser. " Precis de Chimie " Ben/ene.

The composition of a mixture of liquids which are not difficult to separate may. the temperature midway between the boiling points ' ' ' ' . components. mixture contains more than two. into which the original mixture tends to separate or. it is found that the weight of distillate that comes over below the middle point is. the weights of these components will be very nearly equal respectively to (No. if to separate. 1) the weight of distillate below the . even when the separation is very far from complete. that of a mixture of two liquids. It will from the results the components are more difficult of two or three distillations. to the case of mixtures which tend. as a rule. on separate normally into the original components. By middle point is to be understood in all cases the temperature midway between the boiling points of the two components. first be well to consider distillation. be ascertained with a fair degree of accuracy from the results of a single distillation with an efficient still-head. whether single substances or mixtures of constant boiling point.CHAPTER XVI FKACTIONAL DISTILLATION AS A METHOD OP QUANTITATIVE ANALYSIS Determination of Composition of Mixture. as a rule. almost exactly equal to that of the more volatile component. say n. of any two consecutive If the original fractions of constant boiling point. or. Taking first the simplest case. in the case of more complex mixtures. which separate normally on distillation.

which always makes the weight This loss will be greater of distillate somewhat too low. tilting the tube from side to side to facilitate the flow of liquid back to the still and. It is not. 2 to n . shaking out original form of evaporator any liquid remaining in the funnels. owing increased efficiency. is spherical bulbs " is the best. for a great part of it is caused by the saturation of the air in the flask and some with vapour when the liquid is first heated since vapour is mixed with much air its partial pressure is low.1) the weight of distillate between the successive middle points. gives much better results. the quantity left in the still head is almost inappreciable. however. (No. moreover. Loss by Evaporation. proportional to the amount of liquid distilled. roughly proportional to the the still and still-head. volume of and it is advantageous to use and to employ a still-head of as A plain wide tube or one with " the least satisfactory. and in other cases it may be reduced to a very small amount by disconnecting the apparatus. loss It is obvious that there must be by evaporation.CH xvi first CHOICE OF STILL-HEAD 205 middle point. almost the whole of the condensed liquid returns to the still at the end of the disWith a liquid of low viscosity. Choice of Still-Head. and a large proportion of it escapes] condensation when cooled in the condenser. as small a flask as possible small a capacity as is consistent with efficiency. . while hot. because the capacity is the " evaporator very small relatively to the efficiency. Under otherwise similar conditions still-head this . like one of the tillation. from the condenser. and the amount of condensed liquid in it is smaller than in any other equally efficient still-head. as the initial boiling point of the liquid is lower and as the temperature of the room is higher. the loss air in is. if the " " still-head is used. ri) the weight above the last middle point. therefore. lower paraffins. Of all forms. (No. but the " pear stillto the diminished capacity of the bulbs and the head.

the other with methyl alcohol in excess (these liquids form a mixture of minimum boiling point). The following may be taken as an example of the estimation of loss by evaporation. Mixtures of benzene and methyl alcohol. one with benzene. The following results were obtained : TABLE 56. the distillation being stopped in each case when the middle point was reached. were distilled " " still-head of five sections through an evaporator of the original form.206 FRACTIONAL DISTILLATION Estimation of Loss by Evaporation. .

XVI METHYL ALCOHOL AND WATER TABLE * 207 57. Boiling points I. 82-35. . 64'7 . Methyl alcohol. METHYL ALCOHOL AND WATER. water. 100 point. . middle Mixture taken. Methyl alcohol in large excess.

improved by introducing the correction. TABLE 58. Even without cor- recting for loss but much by evaporation the agreement is fairly good. 132-05 . isoamyl alcohol. by repeating the distillation. 80 '2 . middle point.208 It will FRACTIONAL DISTILLATION was as CHAP. result thus be seen that. . Boiling points Benzene. the satisfactory as that given by a single distillathe alcohol tion when it is was in excess. / Mixture taken. 106-1. ISOAMYL ALCOHOL AND BENZENE.

250 grams. TABLE I. a fairly satisfactory In this result was obtained even with a "pear" still-head. the weight below the middle point rose to 9*0 grams. case the distillate following results was divided into three fractions.xvi ADVANTAGES OF EFFICIENT STILL-HEAD 209 ever. With a larger quantity. 1. 3. . and the were obtained. Below 95-4 95-4 104-7 104-7 110-5 2. 59. Allowing a loss of O3 gram for each distillation the corrected weight would be 9'9 instead 10 grams.

The boiling points of the three esters are 57'1. and 101 '55 respectively. and some details have been given. but the total loss was ascertained in each fractionation.210 vertical still- FRACTIONAL DISTILLATION CHAP. it would not be possible to distribute the loss by evaporation. . The percentage weights of distillate below 67'1. from 67'1 to 89-35. head one metre in length has already been (p. 77-15. between the different fractions. 128). and we may calculate the percentages on the total quantity of material left at the end of each operation instead of on the original quantity referred to taken. and by transference from receiver to still. and above 89-35 were read from the curves (Fig. 28) and are given below : TABLE 60. With so large numbers of fractions and of fractional distillations. and the two middle points are therefore 67'1 and 89-35. Number of fractionation.

. O'l between the two middle points. It was assumed in the still-head was found to be that of this loss. 97 '95 between first and second 120'7 . 30*77 37-75 31-48 100-00 31 '25 37-50 31 '25 100-00 The distillation was continued until the temperature reached the boiling point of propyl acetate. middle points Percentage composition of mixture. 0'7 gram was due to evaporation... . Weight >1 of methy' acetate taken ethyl propyl 100 120 100 320 of distillate of distillate Weight Weight below first middle point. and the p 2 . 98 '45. It will after the first three fractionations. Advantages of Efficient Still-Head. 120-8 Taken. The residue in cold.. and O'l above the second middle the still was weighed when - point. second of these esters.xvi ADVANTAGES OF EFFICIENT STILL-HEAD 211 be seen that the numbers remain nearly constant and that the mean percalculated from the last nine distillations agree centages The actual fairly well with those in the original mixture. due partly to and partly to the minute amount of liquid left evaporation 8 gram. and the loss. loss by evaporation must have been greatest for methyl and least for propyl acetate and the calculated percentages are too low for the first and too high for the acetate... and that O5 gram was lost below the first middle point. With an evaporator still-head of eight sections the following results were obtained (p. . The fractionation of the mixture was continued. Methyl acetate Ethyl Propyl . .. 193). . corrected. Found.

Percentage composition of mixture. p. the percentage coming over between 27'95 and 32-15 became roughly constant after - . Taken. . but the difference between the boiling points of the components is only 8'35 against 20'15 and with the esters 22-4. mean 42 per cent. The variation in this case was greater than (40 46. in 12 fractionations). Complex Mixtures. 138). .. hexanes and 32-15) a very little pentamethylene has been already referred to Taking the weight of distillate between 27*95 (p.. .212 results FRACTIONAL DISTILLATION CHAP... Methyl acetate Ethyl Propyl .. it is impossible to distribute the loss correctly different fractions. The separation of isopentane and normal pentane (b. and 36'3 as 100 in each case. the first distillation. and it is best to calculate the percentages on the total amount of material left at the between the end of the distillation. obtained in the second complete distillation are given below. 31 '36 37-27 31-37 100-00 31 '25 37-50 31-25 100-00 The agreement and is is better than after a single distillation. again. Weight below first middle point Weight between first and second middle points Weight above second middle point Total loss . 27'95 middle point and 36 3 from a mixture containing also butanes... 99 '5 118 '25 99 '55 317-3 2-7 320-0 Here.. Found.. very satisfactory.

90-2. then. For the sake containing three components as a ternary mixture. and a mixture of constant boiling point " " may be estimated by the distillation way as that of a single method in exactly the same substance. point. of normal pentane.xvi MIXTURES OF CONSTANT BOILING POINT. or 36'6 per cent. Mixture taken. Mixtures of Constant Boiling" Point. ISOPROPYL ALCOHOL AND WATEK. At the end of the grams of pure isopentane and 175 grams pure normal pentane were obtained. a mixture of constant boiling point containing two components will be referred to in this chapter simply as " a " binary mixture. 80'37. of the isoparaffin but the loss by evaporation must have been greater. of the two paraffins. about 42 per cent. . consists of isopentane and 58 per cent. 213 It would appear. The quantity of a mixture of constant boiling point of brevity. Two examples may be given to prove this point binary mixtures of isopropyl alcohol and tertiary butyl alcohol respectively with water were mixed with excess of water and distilled with the following results TABLE 61. Boiling points : Binary mixture. The methods of ex. that in American petroleum. and there is never so good a recovery of the more volatile component. 100 '0. middle. water. fractionations 101 of . periment and of calculation are similar in all respects.

79'9 point. Boiling points : Binary mixture. .214 CHAT- TABLE 62. water. . TERTIARY BUTYL ALCOHOL AND WATER. middle Mixture taken. 100'0 . 89-95.

215 I. water. Normal propyl Boiling points alcohol and water. 100 '0 point. 93 "85. . : Binary mixture. 87*72. with the latter in excess.XVI BINARY MIXTURES TABLE 63. middle Mixture taken. .

58-34" point. excess. 61 '5. CHAP. Benzene in excess. Mixture taken. . Boiling points : Binary mixture. middle II. Methyl alcohol in Boiling points : Binary mixture.216 FRACTIONAL DISTILLATION TABLE 64. 64 '7 middle point. I. 58 '34 . methyl alcohol. 69-25. 80 '2 . . . benzene.

the formation of a mixture of three liquids a ternary of minimum boiling point. gives rise. take place in twelve different ways and. the components are represented by the initial letters A. theoretically. : The possible cases are as follows . on Ethyl-Aleohol Benzene Water. For convenience. and W. As an example we may consider mixtures of ethyl alcohol. the separation may. benzene and water. 217 When to on distillation. B. its behaviour of a pure liquid. if the original mixture has the same mixture composition distillation as the ternary would be precisely that mixture. in addition to these.TERNARY MIXTURES Ternary Mixtures.

.. B. The boiling points of all possible components and the percentage composition of the three binary mixtures are given in Table 66. First. less the amount in the binary mixture. 67. .. w. A.W. A. VI. . 66. II. V. B A A...B. 32-36 91-17 67'64 4'43 8'83 The middle points are therefore as follows : TABLE .W.W. (6) the weights of distillate below and between the middle points. A. . as before. B... . 100 95-57 .W. W.. . that the corrected weights of the two distillates are equal to those of the ternary and binary mixtures respectively which would be obtained case if the separation were That being so. W.B. simply that in the original mixture . In Table 68 are given : (a) the actual weight of the components in the mixtures distilled. A.. Percentage composition. . . li.B. Method of Calculation. TABLE Boiling points. in perfect. which the weight of water is given by difference. Second. 100-0 80-2 78-3 78-15 69-25 68-24 64-86 . A.. A. Data Required.B.W. the weight of alcohol is can be calculated that taken.. Fractions.B. the weight of benzene in the ternary mixture is I. I. Middle points.B.B. (c) the calculated percentages of the components in the ternary mixture.W. . . A. A. In calculating the composition of the ternary mixture it is assumed.W. loo 100 . B. . . B. W. .B.218 FRACTIONAL DISTILLATION CHAP. A. IV. .. 71'55 67-05 66-55 67-05 66-55 89-1 84-6 73'3 74'7 74'2 Experimental Results.W.. . A... . A B W W.B. A.

TERNARY MIXTURES TABLE (a) 219 68. Mixtures taken. .

. differ somewhat widely from the mean. are formed. On the other hand. and with equal molecular proportions the rise of temperature is only 1-5 (1). of the components is exceedingly and that the distillation method cannot be relied on for the determination of composition. in fact.220 FRACTIONAL DISTILLATION CHAP. is Cases to which the Distillation Method able. of benzene (2). before the boiling point rises 0-1 above that of pure alcohol. very of more volatile component. rich in the It is. of ethyl alcohol. Inapplic. contains 79 molecules per cent. On the other hand some of the individual values. the boiling point is rapidly lowered the middle temperature between the boiling points of the pure components is. 221. . midway between those of the pure components. The explanation on of the rather large errors in this distillation is given p. found practically to be impossible in either case to separate the mixture of minimum boiling point even from Form of ever. small quantities of alcohol are successively added to water. so that the separation would be that of the mixture of constant boiling point from that component which is in excess. benzene boils only 0'1 higher than normal hexane. notably those of alcohol and water in I. although the difference between In both cases the boilthen. howBoiling Point-Composition Curve. In both cases mixtures of minimum boiling point. and the mixture which has the boiling point 74-6. When water must be added to alcohol until the mixture contains 25 molecules per cent. reached when the mixture contains 6*5 molecules per cent.. and it is in such cases the curve is very flat at either one end or the other is least difficult to separate.boiling points is considerable. the less volatile component. Very similar results are obtained with normal hexane and benzene a mixture containing 16 molecules per cent. ing point-composition curve volatile is very flat where the more component one at when that in large excess.

be safely employed for the determination of the composition of a mixture which separates normally into its components. a second distillation may be necessary. containing from 15 to 25 per cent. and therefore. provided that a very efficient still-head be employed and that the distillation be carried out slowly. in the great majority of cases. water mixtures. it it is must be borne from a mixture. values from 7'6 to 8 instead of 4-43 per cent. W. were obtained. by weight of water. : Alcohol 18-2 Benzene Water 74'3 7-5 100-0 which agrees very well indeed with that observed. General Conclusions. mixture. But a mixture of two liquids in mind that from almost always more difficult to separate the more volatile component than the other. and a large quantity of the original mixture will be required in order to give a sufficient amount of distillate for a second operation. through an 18-column dephlegmator and calculating the percentage of water in the mixture of constant boiling point in the usual way from the weight of distillate below the middle point. as a general . Keferring back to - the calculation of the composition of the ternary ethyl alcohol-benzene-water mixture from the first distillation if we take 7*8 as the percentage of water in the A. 219). if the original mixture contain a relatively very small amount of that component. As regards the separation of three or more components it must be remembered that.xvr GENERAL CONCLUSIONS 221 Thus on distilling ethyl alcoholEthyl Alcohol and Water. the calculated composition of the binary ternary mixture becomes (p. The distillation method may.

Soc. 73. Soc. 81. " Fractional Young and Fortey. 1902. 463.. (Specific Gravities "and Boiling Points of Mixtures of Benzene and Normal Hexane. to separate. the least volatile the easiest. 1901. the composition of the original mixture may be determined if that of the mixture of constant boiling point is known or if the composition of the original mixture is known. ' ' 1898. 923. 23. the mixture of constant boiling point may be determined. 1072. while the intermediate components are the most difficult. 2. Chem. If a binary mixture of constant boiling point is formed. 19. It appears to be only when the separation of the components (either simple substances or mixtures of constant boiling point) by distillation is exceedingly difficult that the method is inapplicable. Soc." Trans." Trans.. REFERENCES. Chem. Jackson and Young.- The method may even be applied to the determination of the composition of a ternary mixture of constant boiling point. "The Boiling Point Curves of Mixtures of Ethyl Alcohol and Water.222 FRACTIONAL DISTILLATION component is CH. Soc. 1. Noyes and Warfel. "Experiments on Fractional Industry. Amer. . of Distillation as a Method Quantitative Analysis. that of . Distillation. xvi rule. Young. Chem." Journ. 752.. 1900." Journ. Chem.

a fairly accurate result may be obtained by . or if one component contains a halogen. This method is which the com- generally. There are several other methods by position of a mixture of constant boiling point may be determined. when a known composition is distilled. 2. By Separation of Pure Mixture. to be recommended. but etc. the amount of that element may be determined this method is not. however.CHAPTEE XVII METHODS BY WHICH THE COMPOSITION OP MIXTURES OF CONSTANT BOILING POINT MAY BE DETERMINED In the last chapter Distillation Method. for example alcohol and benzene.. (6) When one component is easily soluble in water. and . the ordinary methods of volumetric analysis may be conveniently employed. or (d) from its refractive power. (c) from its specific gravity. or a base. shown how the composition of a mixture boiling it of has been constant point distillate that mixture of may be ascertained from the weight of comes over below the middle point. or nearly so. the other insoluble. method The most accu- rate for which applicable. but not universally applicable. (6) by the removal of one component. as a rule. sulphur. only to those mixtures the first method can be employed is to separate the mixture of constant boiling point in a pure state by fractional distillation and to determine its composition either (a) by chemical analysis. (a) If one of the substances is an acid.

to determine the values for a series of mixtures in order to find what correction must be applied. a constant temperature. Such series of determinations of specific gravity have been made by different observers in the case of mixtures of the lower alcohols with water (2) and by Brown (3) for some other pairs of liquids. it is almost always necessary to determine the values for a prepared mixture of about the same composition as that which boils at a constant temperature or. a little of the component that is insoluble in water remains dis- water. and to obtain an accurate result this solution. distillate it would be necessary to distil and to treat the first small portion of the with more water in order to separate the remainder component. but there is inevitably some loss by evaporation and by adhesion to the sides of the separating funnel and to the solid dehydrating agent. better. shaking the mixture with water in a separating funnel and washing the insoluble component once or twice with at a known temperaweight determined.224 FRACTIONAL DISTTLLATK >N CHAP. by successive approximations. Mixtures may be distilled. and. of the insoluble of different composition distils (a) at Method of Successive Approximations. 219). As a rule. or (b) without change of specific gravity. This method was employed for the direct determination of the composition of the ternary mixture of ethyl alcohol. (c and d) As the specific gravities and refractive powers of mixtures are not usually strictly additive properties. . also. The specific refractive powers have been determined for several series of mixtures by Lehfeldt (4) and by Zawidski (5). the benzene being determined in the manner described above and the alcohol from the specific gravity of the aqueous solution (1). The volume of this component its ture may then be ascertained or solved by the aqueous solution of the other constituent. that mixture may finally be made up which 3. benzene and water (p. if this is added.

and the following results were obtained with the two last mixtures. 69. we may proceed and last fraction have the same specific method has been employed in the case of gravity. until the first The last TABLE T. ethyl alcohol and water (2). collecting the distillate each time in three or four fractions. better. (b) If the boiling point of the mixture differs only slightly from that of either of the pure components. . observations of the temperature would be useless.xvii SUCCESSIVE APPROXIMATIONS This method was employed by Kyland 225 (a) (6) to ascertain the approximate composition of the large numher of mixtures of constant boiling point examined by him. but we may find what mixture gives a distillate of the same specific gravity (or refractive power) as itself or.

the percentage of alcohol which has the specific gravity 0'81941 at 0/4 is 95-57. .(molecular) composition curve for carbon * is shown in disulphide and methylal Fig. 1 may be mapped against the percentages of one of the components. The pressure. or the boiling points under constant pressure. If the vapour pressures at constant temperature. in a mixture According to Mendeleeff's tables. \ Spccilit gravity Fio. the composi5. Graphically from Composition of Liquid and Vapour.220 FRACTIONAL DISTILLATION CHAP. values tion applies to the boiling point-composition curve. Graphically from Vapour Pressures or Boiling Points. and the percentage corresponding to the maximum or minimum pressure or temperature can then be read off. Ethyl alcohol and water. If the relative composition of vapour and liquid has been determined for a series of mixtures. but it will be seen that while the maximum pressure can be read with considerable accuracy the corresponding percentage of carbon The same objecdisulphide can only be roughly estimated. 60. these 4. 50. of a series of mixtures of known composition have been determined. * The molecular weights of these two substances aro equal and the molecular percentages are therefore equal to the percentages by weight.

of the two components A and be mapped against the corresponding ratios. p' or m'. of one component in the liquid may be plotted against the percentages. R. B (c) The logarithms of these ratios may be plotted in the same way (4). m. or the molecular percentages. R'. R'. of the weights.XVII GRAPHICAL METHODS of 227 tion the mixture of constant boiling point may be ascertained in various ways. FIG. Carbon disulphidc and methylal. (a) The percentages by weight. of the same component in the vapour (7). m R R As an example of the first method. 60. Whichever method is adopted. and corresponding to equal values of p and p'. Lord Eayleigh's determinations of the composition of liquid and vapour for Q 2 . the composition will be given by that point on the curve at which the ordinate and abscissa have the same value or. or of the number of gram(b) The ratios. in the vapour. or log and log R'. p. in the liquid may molecules. by the point of intersection of the curve with the straight line and m'. in other words.

Carbon disulphide and methylal * take the results obtained disulphide and methylal by Zawidski (Figs.. Fio. 04 08 10 PIQ. mixtures of hydrogen chloride and water (Fig. 80 60 20 40 60 60 100 Percentage by Wetgkt of HCl. 04 04 . Hydrogen chloride and water.228 FRACTIONAL DISTILLATION CHAP. 61) may be mentioned (7) for the second and third methods we may .. carbon . 62. 61. 62 and for mixtures of 63).

XVII 6. Carbon disulpbide and metbylal. . The relative number components and X K> may be calculated from the experimental observations x and the values of -. of molecules (or the relative weights) of the in the liquid. W = ~ WB 1 is that Here. 1-0 05 15 10 25_ 25 10 15 05 10 15 Fio. again. For benzene and ethyl alcohol Results Obtained. to x -X . GRAPHICAL METHODS Graphically by 229 means of Brown's Formula. A and TFB and in the vapour. The percentage corresponding required. we may make use of Zawidski's data for carbon disulphide and methylal (Fig. 230) have been obtained.==^ plotted against the percentage XA WB number of molecules (or percentages by weight) of one . the . results given in Table 70 (p. # A W W component. 64). 03.

Observer. 1. CHAP. . . . 2c. Young and Fortey Ryland .230 FRACTIONAL DISTILLATION TABLE 70. . Method. 26.

" Zeitschr.. 717. Properties of Mixtures of the Lower Alcohols with Water. 547 May. 134." Trans. 2. 3." ibid. " The Fortey.INDIRECT METHOD OF SEPARATION REFERENCES. of Zawidski. 521. 1902. Jour-n.." Part II. May. owing to the formation of a mixture of constant boiling point. 42.. Lehfeldt. 81. Chem. Disulphide and Carbon Tetrachloride. Formation of Binary Mixture of Minimum BoilingSuppose. 35. " On the Vapour Pressures Binary Mixtures of 6. The components cannot be separated by fractional distillation because a mixture of minimum boil- . Ryland." Amer. 81. 304. D. "Liquid Mixtures of Constant Boiling Point. however. many cases in which the two components of a mixture cannot be separated by fractional distillation. 46. " On the Distillation of Rayleigh. CHAPTEE XVIII INDIRECT METHOD OF SEPARATING THE COMPONENTS OF A MIXTURE OF CONSTANT BOILING POINT. Distillation after the addition of It has been pointed out that there are a third substance. by weight of the alcohol.. 1902. ' F. "The Properties of Mixtures of the Lower Alcohols with Benzene and with Benzene and Water. 1899. Chem. Liquids. [V]. Binary Mixtures... 1898. for example. 4. Point. "On the Distillation of Mixtures of Carbon .. 1881. physik." Trans." Phil. sometimes possible to eliminate one of them by adding a third substance and then distilling the mixture." Trans. Chem. "On the Properties of Liquid Mixtures. Soc. 7. 1902. 1900. Brown. 1879. "Theory of Fractional Distillation. Young and 4. 35. 5.. [VI]. Fortey. 1. 22. that we have a mixture of isobutyl alcohol and benzene containing 10 per cent. Soc. 739. It is. Chem. 231 Young and Chem. Phil. Soc. 384. 39.

of isobutyl alcohol. If. and (6) pure isobutyl alcohol Here the difference between the boiling (b. 108-05). formed. it 82 -55 . for example. boils at 89'8. of - which alcohol. boiling at 69'25 and containing 91'2 per cent. 69'25). The mixture will then tend to separate into (a) the binary benzenewater mixture (b. This alcohol is a crystalline solid. is considerable. the binary alcohol-water mixture distilled.232 FRACTIONAL DISTILLATION c HU ing point (79'93). owing to the fact that the solubility of isobutyl alcohol in benzene is greater than in water. Formation of Ternary Mixture of Minimum Boiling" Point. only a benzene would be carried over with it and the remainder would form the binary benzene-alcohol mixture the residue in either case would consist of pure isobutyl . If. The substance added frequently forms a ternary mixture of constant boiling point with the two components of the original mixture. Take. points If too much water were added we should have.p. and the separation is an easy one. the case of tertiary butyl alcohol and water. too little water were added. however. a little water be added and the mixture be is the first fraction will consist of the binary mixture benzene and water. containing 9*3 per cent. all that is necessary is to add water to the mixture in the ratio of 8 7 to 100 parts by weight and to distil with an efficient still-head. Neither can they be satisfactorily separated by treatment with water. as an intermediate fraction. when liquefied. of the part of the on the other hand. alcohol. is melting at 25'53 and boiling at miscible with water in all pro- . of benzene. and contains 66'8 per cent. In order to remove the benzene.p. but the relative weights of these components in the ternary mixture differ from those in the binary mixture of constant boiling point which they themselves form. Tertiary Butyl Alcohol and Water with Benzene.

fact. the highest melting point of the distilled alcohol observed being 25'25. Eventually. It was not. . Fractional crystallisation gave a better result. both the melting and boiling points (given above) of the residual alcohol being higher than those of the product Table 71 gives the boiling purified by recrystallisation. it was found that the last traces of water could best be removed by distillation with benzene. and the boiling point 82-45. however. or rather of the binary mixture. from the alcohol by distillation is difficult. TABLE 71. the melting point of the recrystallised alcohol being 25-43. of the alcohol. points and the composition of the binary and ternary mixtures and the boiling points of the single components.xvin WATER FROM TERTIARY BUTYL ALCOHOL 233 portions and forms with it a mixture of constant boiling point (79'91) containing 88-24 per cent. in found possible to obtain the alcohol quite pure by this method. The difference between the boiling point of the binary mixture and of the alcohol is only 2-64 and the separation of the last traces of water.

It is better to add a larger amount of benzene. 145 grams of the ternary mixture and 57 grams of alcohol would be obtained. and if a perfect separation could be effected by a single distillation of the mixture. in it. grams of the alcohol-water - "evaporator" still-head of the original form. Theoretical composition.. start with 100 grams of this mixture .234 FRACTIONAL DISTILLATION CHAP. the water . in the first place. the second and third fractions still containing a little water. say 125 grams. we should get a rather smaller amount of the ternary mixture. 72.-.. Not only can we remove the last traces of water from the nearly pure alcohol by means of benzene but it is also possible to obtain the pure alcohol from the binary alcoholwater mixture. when the quantity of the second fraction (alcohol-benzene) will be increased and the smaller amount of residual alcohol will be obtained free from water in a single operation. TABLE Temperature range. . In practice. Suppose that we . and the following fractions were collected. a little binary alcohol-benzene mixture and a residue of alcohol. 11 -7o grams. . would require 11-76x70-5 8'1 = 102 grams of benzene for conversion into the ternary mixture. however. In an actual experiment. 117*5 mixture (containing 103 7 grams of tertiary butyl alcohol and 13'8 grams of water) with 145 grams of benzene were distilled through a 5-section Results of Distillation.

and the second rather more benzene and a very little water. the upper one of benzene containing some ^alcohol and a little water. to add about 35 grams more. the weight of this mixture would be 41-0 grams and to remove the water 41'8 grams of benzene would be required. the lower one consisting chiefly of alcohol and water with a little benzene. but the first fraction must really have contained rather more alcohol and less benzene and water. the second fraction and the additional benzene we should have a liquid of approximately the following composition. But the second fraction contains about 25 grams of benzene and it would be necessary. By adding more water in a separating funnel." The first fraction separated two layers. The third fraction must have contained some benzene with the dry alcohol. and the distillation stopped when the . the first two fractions collected as before. running off the aqueous alcohol and washing the benzene repeatedly with small quantities of water to extract the remaining alcohol. On mixing together the recovered alcohol-water mixture. and by fractional distillation almost the whole of the alcohol could be recovered in the form of the binary alcohol-water mixture of constant boiling dilute point. Theoretically. a into aqueous solution of the alcohol could be obtained almost free from benzene. the fractions would have had the composition stated in Table 72.xvin TREATMENT OF FRACTIONS 235 If the separation had been complete. in practice. Alcohol 50-8 GO'S 4-8 Benzene Water 115-9 The mixture should now be distilled. Treatment of " Fractions. making altogether about 60 grams.

CHAP. Temperature range. TABLE 73. The results would. Theoretical composition.23G FRACTIONAL DISTILLATION . first still distillation second middle point is reached fraction 3 from the should then be added to the residue in the and the cally. . theoreti- be those given in Table 73. distillation continued.

On the other hand the methyl alcohol may readily be mixture with benzene by shaking with water. and its solubility in that substance may perhaps be said to be greater than in benzene. as the com. because. be possible to recover about 215 out of the 264-7 grams of alcohol in the pure state. extracted from its The greater the molecular weight of an alcohol.xviii ALCOHOLS WATER BENZENE 237 yield. the more . in fact. no separation can be effected by adding water and distilling. the composition curve is nowhere nearly boiling point-molecular of minimum boiling point horizontal. Methyl alcohol and water may be separated from each other without difficulty by distillation with an efficient still-head. say. as distillation. the properties of the alcohols recede from those of water and approach those of the paraffins or of benzene. an additional 20 grams of pure alcohol. it resembles water much more closely. plexity of the alkyl group increases. A ternary mixture does not come over. as alkyl derivatives of water and. as . in the case of isomers. on the one hand. the higher the boiling point. and. although the alcohol is miscible in all proportions both with benzene and with water.hydroxyl derivatives of the paraffins and that. because their properties are so similar. on the other. Methyl alcohol and benzene cannot be separated from each other by and. is much lower than that of the alcohol mixture of constant boiling point. indeed. Alcohols Water Benzene. and their boiling points so far apart that a mixture is not formed. would. but the first fraction still consists of the benzene- minimum boiling point is formed. because their properties are so dissimilar that a mixture of its boiling point binary benzene-water mixture. or. It has been pointed out that the monhydric aliphatic alcohols may be regarded. Methyl Alcohol. with very little would It actual loss.

Isobutyl Alcohol. the extraction of the alcohol from a solution in . normal propyl and tertiary butyl alcohol all form binary mixtures of minimum boiling point both with water and with benzene. the properties of the two substances. isopropyl. and to extract benzene from the higher alcohols by adding water and distilling. Higher Alcohols. Ethyl. latter substance may be removed from the strong alcohols by distillation with benzene. the first portion as has been stated. only with one dehydrating agent freshly ignited lime that the the similarity of last traces of water can be removed. and when a mixture of these three of the distillate consists. Isobutyl alcohol. Isopropyl. and dehydrating it is only under very special conditions and. however. with isobutyl alcohol the Normal Propyl and Tertiary Butyl Alcohol.238 difficult is FRACTIONAL DISTILLATION CHAP. . by distillation because its boiling point is so very little higher than that of the alcohol-water mixture and. of the benzene-water mixture which boils at 69-25. Pure ethyl alcohol cannot be obtained. however. on account of . agents act in a very similar manner on them. apparently. alcohol or General Statement. be removed from any other of higher boiling point by sufficient water to form the binary benzene-water adding mixture. it is possible to extract water from the lower alcohols except methyl alcohol by adding benzene and distilling. in fact. isobutyl Benzene may. does not form a ternary mixture of "minimum boiling point with benzene and water. Thus. even from its very strong aqueous solution. alcohols. Ethyl. and distilling with an efficient still-head. forms boiling point with benzene Each and the of the four a ternary mixture of minimum and water. liquids is distilled. benzene by means of water process is very slow indeed.

2. Cham. it is . Purposes for which Fractional Distillation quired. Young. 1902. The separation of the components from a mixture of known 3. Fractional distillation of is re- may be employed for various are which the following among the more important. or the determination of the quantitative composition of a mixture of constant boiling point. qualitative composition. Young and Fortey. CHAPTEK XIX GENERAL REMARKS PURPOSES FOR WHICH FRACTIONAL DISTILLATION is EEQUIRED. 717. Soc. 1. CHOICE OF STILL-HEAD. "The Properties of Mixtures of the lower Alcohols with Benzene and with Benzene and Water.. The isolation of a single substance from a mixture with the smallest possible loss. purposes. &c." Trans. or when it is definitely not known known what that certain substances are present but others there may be. 81. NUMBER OF FRACTIONS.xvi IT It is GENERAL REMARKS 239 probable that other cases will be discovered in which distillation one component of a mixture may be removed by with an additional substance. "The Preparation of Absolute Alcohol from Strong Spirit." ibid. INTERPRETATION OF EXPERIMENTAL KESULTS. 4. 81. REFERENCES. A general study of the qualitative and quantitative composition of a complex mixture when only the genera nature of the chief components is known. of the quantitative composition of a mixture of which the qualitative composition is already The determination known. 707. 1902..

be taken as most probably pure and will not require to be re- . Much time will be saved if an improved still-head is employed. the "pear. 65. In the second and subsequent fractionations the greater part of the distillate from the first fraction may. FIO. In any case a careful record should be made of the temperature range and weight of each fraction. Interpretation of Results. for liquids of high boiling point. we may is conclude that there I present a rela- tively small quantity of some impurity of much higher boiling point. have a perfectly cdnstant and we may conclude that it most probably consist of the pure substance required. the first fraction will boiling point will when redistilled. the procedure will be somewhat simplified if the true boiling point of the chief component is already known. THE SEPAKATION OF A SINGLE SUBSTANCE FROM WITH THE SMALLEST POSSIBLE Loss MIXTURE 1. 65. FRACTIONAL DISTILLATION A CHAP. 122) or plot the total weights of distillate against the final temperatures of the fractions. but that impurities more volatile than the chief VeighL'of Distillate component are absent." " and for volatile liquids the " evaporator apparatus may be especially recommended.240. If the results obtained by the distillation of the liquid under examination give a curve similar in form to Fig. When a pure substance is to be isolated from a mixture. and we may then either calculate the values of Aw/A (p. As the temperature shows no appreciable rise until towards the end of the distillation. First Case. in each case.

the refractive index. the problem will be much simplified. pure. and that we were unacquainted with the previous history of the specimen there would be the possibility of the presence of some benzene. of the aromatic hydrocarbon boils at almost exactly the same temperature as pure . or some other physical constant may be the distillate : if gravity of the distillate agree determined for comparison. by fractional distillation. if of the pure substance is known. If any doubt exists and if the specific gravity of the pure liquid is accurately known. however. the melting point (if the liquid can be easily solidified). such as members of a homologous series. Proofs of Purity. Even. and therefore the presence of the benzene could not be detected. original liquid consisted chiefly of hexane. we cannot be absolutely sure of* the purity of the liquid unless we know from the method of preparation that there is no possibility of the formation of a mixture which distils without change of composition at If we practically the same temperature as the pure liquid.xix distilled. can be sure that the original mixture contains only substances which are very closely related to each other. the vapour density. and if the true boiling point the observed boiling point agrees with it. and the impurity could not be removed. FIRST CASE The separation 241 of the liquid of constant boiling point will in such a case as this be an easy one. for it may then be concluded with certainty that no mixture of constant boiling point can be formed. or a chemical analysis of the liquid may be made. we may conclude with confidence that the distillate is really Instead of the specific gravity. Hexane and Benzene. it is best to determine that of both the boiling point and the specific with those of the pure substance. But a mixture of normal hexane and benzene containing as much as 10 per cent. The high specific gravity of the distillate B . that the normal hexane. for example. Suppose.

on the other hand. and also that the binary mixture can be separated in a pure state from water. Ethyl alcohol containing. 15 or 20 per cent.242 FRACTIONAL DISTILLATION CHAP. in this case. say. of alcohol is extremely difficult to separate from water. In this case. for the mixture of constant boiling point which contains 95'57 per cent. however. it were known that the hexane had been prepared synthetically from pure propyl iodide by the action of sodium. there would be no possibility of the presence of benzene. the boiling point of the binary alcoholwater mixture is so slightly lower than that of pure alcohol that the reading of the thermometer unless very accurate would hardly be sufficient to distinguish with certainty between the two. If. but. Second Case. but a mixture containing at least 5 and probably as much as 7 or 8 per cent. show that some other substance besides normal hexane was present. 65). similar results would be obtained except that the difference between the boiling points of the pure alcohol and of the binary mixture is sufficiently great for the two to be distinguished without difficulty by the observed temperature. Isopropyl Alcohol and Water. nor even the pure mixture of constant boiling point. we should not have pure alcohol in the first part of the distillate. liquid is of Figure 66 represents the separation of a constant boiling point from a mixture which contains only much more volatile impurities. With isopropyl alcohol and water. of water. even with the most efficient still-head. of water would behave in the manner indicated by the same curve (Fig. although there is a wide difference between the two boiling points. Ethyl Alcohol and Water. again. would. it an easier one than the last since is The separation the least volatile .

in such a case would be much more and several fractionations would be required before FIO 67. the boiling point of not Weight of DislilLaU much which was higher than that difficult. that we might have a mixture of minimum \ ! boiling point containing neither substance in excess but contaminated with a more volatile impurity.XIX SECOND AND THIRD CASES 243 component that If the temperature remains is to be isolated. for mixtures of maxirpum boiling point are not often . the last portion need not be redistilled. When the boiling points of the components of a mixture are R 2 . though unlikely. it is possible we may be dealing with Wcight ofDlsmilte i a mixture of maximum boil- F IO . It is also just possible. 67. A curve like that in Fig. again. would result from the distillation of a liquid containing an impurity. for example. and in subsequent fractionations the distillation may the maximum may has pure. constant for a considerable time and shows no rise whatever at the end of the distillation. be stopped as soon as temperature been reached and the be residue taken as that Here. but it is less probable. met with. in the separation of the binary mixture of chloroform and methyl acetate from a slight excess of the ester. component. co. ing point as. of the chief The separation the boiling point of the first fraction became quite constant. Third Case.

67 as Fig. chloride and benzene. p. 64-5) with chloroform (b. * Disttuau 67. if present in large excess in the original mixture. p. and it would be practically impossible to separate either pure tetrachloride or the mixture of constant boiling point from any mixture of the two substances. the highest fraction might consist of a mixture maximum boiling point such as chloroform-methyl acetate (b. 08. again. less probably. 68. the . Fifth and Sixth Cases. but in which the difin large excess in chief ference between the boiling points of the components was small. In the case of Fig 69. for example. the isopropyl alcohol-water mixture (b. . A curve such as that in Fig. 60'5) in excess or. of a mixture of minimum boiling point containing a more volatile impurity. very near together and the chemical relationship is not very mixtures of constant boiling point are not unlikely to be formed that is the case. Fourth Case. chief The separation of the al- component would by most invariably be easier than in cases represented Weight Fig. although benzene. would represent the behaviour on distillation of any mixture such as that last-named with benzene which the was the less component volatile. p. Here. p. the same relationship to Fig.244 FRACTIONAL DISTILLATION CUM-. as. Indeed. for example. 69 and 70 represent the distillation of liquids which contain impurities of both greater and less volatility. 66 to Fig. . 80'35) with a little of . 78-3). ethyl alcohol (b. could be separated in a pure state by repeated fractionation. with carbon tetraclose. The curves in Figs. 68 bears FIO. Fig. 65.

for the middle substance is always more difficult to used (p. Here the boiling point of the chief component. When. frequently to be carried out. the process is tedious if an ordinary 128). If the boiling points of the components are very uear . 12 bulbs. benzene acids. in the separation of ethyl acetate from methyl and propyl apparatus is acetate. Weight of Distillate Fio. in the case of Fig. When. and even with a 5-column "evaporator" still-head several fractionations are required (p. Weight of Distillate Fio. 69. however. while a certain amount of dinitrobenzene is formed. the collection of pure nitrobenzene may be commenced after the second or third and far fractionation. say. the boiling points are not so far apart. is very much higher than that of benzene below that of dinitrobenzene. separate than the others. and the mixture is distilled through a "pear" preparation still-head with. 193). as. 70. for example. is nitrated by treatment with some of the benzene is usually unacted on. nitric and sulphuric nitrobenzene.XIX FIFTH AND SIXTH CASES 245 that boiling points of these impurities are far removed from of the substance to be separated . and the separation is If pure benzene is used for the therefore an easy one. for example. 70 Such separations as these have very they are near it.

On the actual boiling points will depend the nature of the still. IV) there is no possibility of the formation of mixtures of constant boiling point.head that can be used. while the less volatile will only be eliminated with difficulty and the distillation will correspond to those represented by the curve in Fig. 68 rather than in may Fig. if it can be ascertained. Boiling Points of Components. ponents and the difference between them. . for example a liquid may contain only impurities of greater volatility but the boiling points of some of these be near. The most volatile impurities in such a case will be easy to remove. those of others far below that of the chief component. Other Cases. 70) the separation of the middle substance is extremely troublesome. The chief points to be considered are (a) the boiling points of the pure components.246 FRACTIONAL DISTILLATION CHAP. together (Fig. the more easily can the separation be effected. . the greater the difference between them. By far the simplest cases are those in which the substances present are chemically closely related to each other. THE SEPARATION OP THE COMPONENTS OF A MIXTURE OF KNOWN QUALITATIVE COMPOSITION Two Components. Other cases than those referred to may be met with. 66. and the difference between them (b) the chemical relationship of . the components (c) the form of the boiling point-molecular composition curve. for the form of the boiling pointmolecular composition curve must then be normal or nearly so (Chap. Closely related Substances. and there is greater probability that mixtures of constant boiling point may be formed than in the cases previously considered. and the only points to be considered are the actual boiling points of the com.

it is practically impossible to separate that component in a pure state. if the boiling point of the mixture is very near that of one of the two components. But. propyl alcohol and water. 220) that when the curve is difficult. pointed out of this form it is very (p. 67) may be consulted. generally. and its boiling point is far below that of the more volatile component. as in the case of ethyl alcohol and water (Fig. to separate that component which is in excess where the curve is nearly horizontal. usually that corresponding It has been to the lowest temperature. at any rate. mixtures of constant boiling point may be formed and the list of such mixtures which are at present known (p. such mixtures are most frequently met with when the molecular weight of one of the components is abnormal in the liquid state. It will be sufficient to remark here that water forms mixtures of minimum boiling point with the majority of organic compounds that. and it may be impossible to separate the mixture of constant boiling point even when the component which boils at a widely different . and may be impossible. . to make the curve nearly horizontal over a large part of its course starting from one extremity. it is usually possible to separate in a pure state both the mixture of constant boiling point and that component which is in excess. and that this is usually the case with compounds which contain a hydroxyl group. It should also be remembered that when the boiling points of any two substances are near together. as in the case of normal Boiling Point. or far above that of the less volatile component as with nitric acid and water.ta SUBSTANCES NOT CLOSELY RELATED 247 Substances not Closely Related. or of methyl alcohol and benzene. or of normal hexane and benzene. or. a small deviation from the normal boiling point-molecular composition curve is sufficient to give rise to the formation of a mixture of constant boiling point. 71). Mixtures of Constant If the substances are not closely related to each other. When a mixture which distils without change of composition is formed.

Thus.248 FRACTIONAL DISTILLATION CHAP. the binary IUU 95 90' 85' 80 75 ( . in the cases referred to. whatever the composition of the original mixture. it is impossible to obtain normal hexane. ethyl alcohol. temperature is in excess.

Amount of each Component Present. It would probably be best to collect all the distillate that came over below 100'0 in one fraction and that from 110'0 to 110'6 in a The large residue would consist of pure toluene. Suppose. for a given mixture. that it was desired to separate the benzene as completely as possible from a mixture consisting of 30 grams of that hydrocarbon with 270 grams of toluene. . The best plan would be to employ for the first distillation the most efficient still-head available. . The approximate quantity of each component must be taken into consideration not only in deciding upon the number of fractious but also in choosing the still-head to be used. : (a) (&) (c) The efficiency of the still-head employed The approximate quantity of each component The difference between the boiling points of the com. The same dephlegmator might be used for the second distillation and the distillate might be collected in three or four . the more efficient the still-head the smaller is the number of fractions that will be required. ponents (d) The form of the boiling point-molecular composition curve. second. Efficiency of Still-Head. A very slow distillation would be advantageous. The most important of these are the following Number Head. and if the recovery of the toluene need not be regarded as of special importance. so many points have to be taken into consideration. and a dephlegmator of many sections might be used. the complete return of the liquid from the still-head to the still at the end of the distillation would not be essential. for example.xix NUMBER OF FRACTIONS REQUIRED It is difficult to 249 of Fractions Required and Choice of Stillformulate any definite rules regarding the number of fractions in which the distillate should be collected. It may be stated quite generally that.

would not be worth redistilling. A distillation the residue it would would require to be redistilled. and above 95-6. would be important that the amount condensed liquid in the still-head should be as small as possible and an evaporator still-head of 5 sections would probably be the most convenient. it would again be best to employ a very efficient stillhead. In the second fractiona- tion. fractions might be collected below 80'5. The re- small and an evaporator still-head of not more than 5 sections would have to be used now be remaining distillations. For mixtures of closely related substances. temperature ranges of which would depend on the efficiency of the still-head. the distillate from the first fraction of I. or others which behave normally on distillation. consist of pure toluene . and as. if not. distilled through a very still-head. If the original mixture contained 270 grams of benzene and 30 grams of toluene. would consist of nearly pure benzene maining fractions would for the and might be rejected. or even more. an evaporator still-head of many sections would be the best for the purpose. the greater the difference between the boiling . efficient mixture of this composition. The would now be small. After rejecting this. Boiling Points of Components. the FRACTIONAL DISTILLATION CHAP. and it would be necessary to employ a still-head of It fewer sections for the remaining of distillations. would give a considerable amount of nearly pure benzene so far as the recovery of toluene is concerned. from 80'5 to 95'6. in this case. a most essential point would be that the liquid should return very completely from the still head to the still at the end of the distillation. If the temperature reached 110*6 before the end of the .250 fractions. the first 100 grams. With an 18 section Young and Thomas dephlegmator it is probable that nearly 20 grams of fractions distillate would come over below 81 '2. and it was desired to recover the toluene.

be somewhat smaller. p. the number of fractions in the low temperature region must be considerably increased while those above the middle point may be diminished in number. When the form of the boiling point-molecular composition curve is normal it is steeper in the region of high temperature than of low. distillation. Boiling Point-Composition Curve. the least volatile component will be the easiest to separate and the substance of intermediate boiling point the most difficult. however. 36'3 spectively) would require a much larger number of fractions than one of benzene and toluene if a still-head of the same efficiency were employed for both. If tionation has made some Three Components. has been shown 153) that a mixture and 27'95 renormal and isopentane (b. the components are not closely related to each other. Thus. however.xix BOILING POINT-COMPOSITION CURVE more readily can the it 251 points the of components be separated by (p. If. construction of the boiling point-molecular curve are wanting. we cannot decide on the number of fractions that may ultimately be required until after the fracprogress. In most cases. When the actual boiling points of mixtures of the two substances are lower than those given by the normal curve. and in general the temperature ranges above the middle point may be somewhat greater than below in other words. unless] we can judge composition from the nature of the substances in the mixture whether data for the the deviation of the actual from the normal curve to is likely be large or not. there are three substances present and the mixture behaves normally on distillation. the number . this difference above and below the middle point becomes accentuated and when the curve is very flat near its lower extremity. of fractions may . one or more mixtures of constant boiling point may be formed and the problem becomes more complicated. and. .

(b) one of the three possible binary mixtures of constant boiling point.2). . for example.. 217). benzene. with a mixture of ethyl alcohol (AJ. 95'15). . First fraction... There are. as would be the case. or the fraction a alone. in separation may take 12 different ways in which place.2 A detailed description the separation of the three . .. It may.. happen that the quantities 'in the original mixture are such that we have only the two fractions a and b. Second fraction. with boiling point. but under no conditions can the mixture distil without change of composition..252 FRACTIONAL DISTILLATION CHAP. A. First fraction. . unchanged (p. Again.. 69-25) or of water and amyl alcohol (b. and isoamyl alcohol (A. benzene and water is distilled it tends to separate into (a) the ternary mixture of constant boiling point.. . when a mixture of ethyl alcohol... the substances present may be capable of forming one or two binary mixtures but no ternary mixture of constant That would be the case. therefore five different ways in which separation may occur.. or a and c. A2 A2 A. p. A. benzene and water W separations.. for the only mixtures isoamyl of constant boiling point that. alcohol. There are. . however. the following three separations would be possible. If only a single binary mixture of constant boiling point can be formed. (c) that pure component which is ia excess. or the mixture may distil fact. B and we have the following possible alcohol.. Thus. for example. for the components Employing the initial letters A.. Third fraction. A B of l . can be formed are those of benzene and water (b. p. benzene and water. . A^B A^B AtB . W. Third fraction.

ethyl acetate and propyl acetate (both with an ordinary and an improved still- head) 3.at nearly the same // there are isomeric members of a homologous series present in the. the qualitative composition of distilled which is known. TITATIVE COMPOSITION OF A OF THE QUALITATIVE AND Rough Estimate of Composition. An instance been referred to (1) in the case of the of this light distillate has already from . or (c) two or more of the substances form mixtures of constant boiling point. Two Components Pentanes in Petroleum. a rough estimate of the boiling points and of the amounts of the components may frequently be obtained from the quantities of distillate in the first case. But confusion is apt to arise when there are two substances present with boiling points very near together. or from the temperature ranges in the second. OR WHEN IT IS DEFINITELY KNOWN THAT CERTAIN SUBSTANCES ARE PRESENT. is through an efficient still-head only partly and the distillate is collected in fractions of either equal temperature range or of approximately equal weight. (a) Temperature. (6) one of the components is present in relatively very small amount. BUT IT is NOT KNOWN WHAT OTHERS THERE MAY BE. or. The use of FRACTIONAL DISTILLATION as a METHOD OF QUANTITATIVE ANALYSIS has been fully discussed in ' Chapter XVI. Boiling. methyl acetate. generally. When a complex mixture. "mixture. is given in Chapters VII and XIV. A GENERAL STUDY QUANCOMPLEX MIXTURE WHEN ONLY THE GENERAL NATURE OF THE CHIEF COMPONENTS IS KNOWN. from the values of Causes of Confusion. 4. it is very likely tliat their boiling points may in some cases be very near togctJier. a.xix CAUSES OF CONFUSION 253 closely related liquids.

45 (p.254 FRACTIONAL DISTILLATION American petroleum consisting chiefly of butanes. repeated fractionation that the presence of both normal and isopentane (boiling points 36'3 and 27'95) is The fact. in the first fractionation. in which the results of . 183). the results of the first distillation seem to indicate the presence of only a single substance between and it is only the butanes and hexanes. while for those below and above them they increase. fractions the values of Aw/A diminish. have the highest values are Nos. . the fractions that 7. and in the third 74. after This the may be seen from Table 74. in the second. 5 and they are Nos. boiling at about 33 . is a clear indication that we are not dealing with a single substance. pentanes and hexanes. the highest values of Aw/A are first those for fractions 5 and 6 . Even with the very efficient still-head shown in Fig. however. that for the middle clearly shown. three fraction ations are given it will be seen that. TABLE I. 4 and 7.

and benzene was removed before the fractionation was commenced.xix HEXANES IN PETROLEUM is. It does not appear. or. that of the Galician with the combined . through an ordinary or fairly efficient still-head appears to indicate the presence of a single substance boiling at about 66. The American petroleum at first for the most part between 28 and 95. 139). is taking place. and after a few fractionations with a very efficient still-head it is seen that a further separation. Aromatic hydrocarbons were not removed before the fractionation by treatment with nitric and sulphuric acids. the Galician petroleum. distillation for the hexanes A preliminary example. however. The fractionation of the American petroleum was carried out with a 12-section Young and Thomas dephlcgmator. Both the pentanes may be obtained in a pure state by distillation only. p. The weight of the From Galician petroleum was about 5 times as great as that of the American. that any hydrocarbon that boils at a higher temperature than normal pentane can be so separated. Hexanes in Petroleum. the hydrocarbons boiling below 40 and above 72 had previously been for the most part separated by distillation. came over Material Employed. 255 The whole course in the of the separation however. at any rate. as in the case of the pentanes. (2). were carried out. after removal of small quantities of impurity by means of a mixture of nitric and sulphuric acids. Let us consider. 29. apparently into two components. Still-heads Used. richest in hexanes the weight of distillate between 56 and 74 was about 800 grams. and was . but after long continued fractionation it is found that the Two series of fractionations separation does not end there. better seen diagram (Fig. one of American the other of Galician petroleum.

Description of Results. and it will be seen that at this early stage the fraction from 65 to 66 has the highest value. and the still-head employed was more efficient. and had the lowest value. TABLE IV. The Galician petroleum was thus freer from pentanes and heptanes to begin with than the American. . 45).256 FRACTIONAL DISTILLATION CHAT. 75. In Table 75 are given the temperature ranges and the values of Aw/A for the fourth fractionation of the American petroleum. dephlegmator and regulated temperature still-head which was employed for the separation of the pentanes (Fig. but in the subsequent fractionations the ratio for the corresponding fraction (above 65) steadily after the tenth fractionation this fraction fell.

only one curve has been drawn for each of these two pairs. the last portion of liquid boiled quite constantly at 36-3. 72. the values of Aw/A for the lowest and highest fractions show a steady rise. the experimental results for American petroleum being indicated in these and other cases by crosses. With the hexanes. . the curves became quite horizontal at these two temperatures. . Fio. in the thirteenth fractionation. the maximum value of Aw/A (at about 69'1) showed very little .XIX GRAPHICAL REPRESENTATION 257 tionations of American petroleum correspond very closely with the fourth and seventh respectively of the Galician. The referred to as IV A fractionations of American petroleum #re and of Galician as IV VII G A . in the eighteenth. in the case of American petroleum. after the sixteenth fractionation. in other words. .) After the fifth fractionation of the pentanes. and. Highest Fractions. no fraction was obtained with perfectly constant boiling point even though. . until.95 . X . the first portion of distillate came over quite constantly at 27. on the other hand. the liquid which came over above 66 was fractionated thirtyone times. . . Indeed. (Fractions from 58 to 72. Results obtained by fractional distillation of American and of Galician petroleum. and for Galician by circles.

Later on there would appear to be a further separation into two components with the approximate boiling points 61 and 69 '5 corresponding to the two pentanes which boil at 27'95 and 36-3 respectively. the last fractionation. the value of With the Galician petroleum. the maximum only 626 for the fraction coming over between 60'55 was and 60'75. These facts may be explained by assuming the presence of at least four substances in both American and Galician petroleum boiling at temperatures between. say. (189 for the fraction coming over between 60'85 and 61'0). no further improvement in Aw/AJ for the fraction a little above 69 was noticeable after the eleventh fractionation. the highest value had fallen to 141 from 60*75 to 60-9. two of them with boiling points not far from 61 and two not far from 69'5. it the twenty-first fractionation diminished slightly. but in this case also the values for the higher and lower fractions appeared to be increasing slightly. Again. for the fraction Aw/A With Galician petroleum the highest individual value of a^ * ne lower temperatures was reached in the eleventh In fractionation (790 for the fraction from 60'6 to 60'85). 59 and 72. but. the values for the fractions above and below these temperatures showed a tendency to rise. in the case of American petroleum. just as with the pentanes we have in the first place apparently the separation of a single component boiling at about 33. Lowest Fractions. on the other hand. the highest individual value of Aw/Ai was reached in the twelfth fractionation. state. In the sixteenth fractionation. We should then have at first the apparent separation of a single substance boiling at about 66.258 further rise. But while the two pentanes can be separated in a pure the two components from the hexane fractions . at the lower temperatures. FRACTIONAL DISTILLATION and after CHAP. the seventeenth. Interpretation of Results.

it is extremely doubtful whether. Normal Hexane and another Component. separations must necessarily be far But each of the final more difficult than the components of earlier ones. or is just possible that the pentamethylene about 50-5. gr. Di-isopropyl. of a little di-isopropyl. however. and although a partial separation may be effected. 0/4). sp. accuracy of the determinations of their boiling points and which have been made up to the present to warrant the formation of any estimate as to the time. and. either of these paraffins has yet been prepared in a perfectly pure state. and this may be due to the presence (CH 3 ) 2 CH CH(CH 3 ) 2 which boils . whether their physical properties. the fractions boiling near to 69 -5. 0-7506 at 20'5/4.). however. may not have been completely removed. Isohexane and Diethyl-methyl Methane. at 58-1 and has a someit what higher specific gravity (0-67948 at (b. with the most efficient still-head. we have the beginning of a further subdivision. p. for the boiling points of the each pair are much closer together than the apparent boiling points of the pairs themselves. With we have the regard to following s 2 . in fractionations. As regards the components boiling not far from 61. It is noticeable. which is certainly present in petroleum. perature ranges or specific gravities of the fractions.xix INTERPRETATION OF RESULTS the later 259 cannot. there is the possibility of the presence of the two isomers with the formulae (CH 3 ) 2 CH CH 2 CH 2 CH 3 and (C 2 5 ) 2 CH CH 3 which would certainly resemble each other very closely indeed in It is doubtful. any one of the components could be separated in a pure state by fractional distillation alone. that the observed for the fraction specific gravities (minimum 0*6728 at 0/4 from 60' 55 to 60'75) are slightly higher than the most probable value for isohexane (about 0-6721 at /4). relative quantities of the two hydrocarbons from the temspecific gravities. and sufficient reliance cannot be placed on the H .

petroleum was fractionated thirty-one times the fractions were then treated several times with mixed nitric and separately sulphuric acids. facts to guide us. TABLE Temperature range. but above 69 ing the specific gravity rose rapidly. The boiling point of pure normal is 0-67697 is 68' 95 and its specific gravity at 0/4 hexane but in . In both cases the specific gravities of the fractions above 66 were higher than those of pure normal hexane. The less volatile portion of the American . The best specimen that could be obtained boiled at 69-05 and had the specific gravity 0-67813 at 0/4. . the last fractionation of the Galician petroleum the fraction from 69-12 to 69-20 had the highest value of Aw/Af and its . and were subsequently fractionated many times with the object of separating pure normal hexane. specific gravity at 0/4 was about 0-685. in the course the fractionations. 66-4 67-85 68-85 68-6 . 69-2569-4 . 68-9569-03 69-0 69-1 . For the fractions from 66 to 69 the change was slight.260 FRACTIONAL DISTILLATION CHAP. The followof determinations were made 76.

whereas methyl pentamethylene is acted on rapidly at the ordinary temperature. methylene. 0-7968 at 0/4) and methyl . p. which boils at 80-85. Preparation of Pure Normal Hexane. for example. 80'85 sp. Further light has been thrown on the problem by the researches of Markownikoff (3).xix METHYL PENTAMETHYLENE the fractions given in 261 Evidence afforded by Specific Gravities. sp. gr. p. about 71'5 both of which substances have also been prepared 0/4). and this seems also to show that the boiling point of the second substance cannot be far from that of normal hexane. Tfte two hydrocarbons differ widely in their synthetically. with evolution of much heat. and it may therefore be concluded that the substance present in American and Galician petroleum with a boiling point not far above that of normal hexane is methyl penta. it has been shown (p. Methyl Pentamethylene. about 0-766 at pentamethylene (b. the substance boiled at a temperature as high as 80 the rise would certainly be more gradual at first. much heat being evolved and a large amount of acetic acid formed. and acetic acid is the chief product of oxidation. can be separated from the hexanes by fractional distillation. and of Aschan (5) on Eussian petroleum. which has been found to contain large quantities of hexamethylene (b. for hexamethylene is only attacked slowly even when heated. If. Moreover. of Zelinsky (4). specific gravities of Again if the Table 76 are plotted against the mean temperatures it will be seen that there is a very sudden rise of specific gravity above 69. Not only methyl pentamethylene. 140) that hexamethylene. a large amount of adipic acid being formed. gr. behaviour with fuming nitric acid. Now the fractions for the first three or four degrees above 69 are attacked in the cold by fuming nitric acid. but also the isohexanes and other hydrocarbons which contain a group may be removed by >CH .

paraffin appeared to be perfectly pure. It was found possible to separate a liquid of quite constant boiling point and as derivatives of hexamethylene could be prepared from it (6). and the boiling points are so close together that even with an oil exceedingly difficult. it was found . and it was found that when the fractions from Galician petroleum which came over between 66 and 69'2 were subjected to prolonged heating with the fuming acid. Another example of the presence. long continued fractionation would be necessary even to indicate the presence of two isomeric amyl alcohols. the liquid which at first seems to be a single substance boiling at about 66 proves to be a mixture of four substances. Distilled through an ordinary stillhead. the isomeric amyl alcohols which are present in fusel may be mentioned. The first pair of substances are very closely related to each other. and were afterwards distilled two or three times. however. of two substances which boil at nearly the same temperature is afforded by the distillation of the portion of American or Galician petroleum that comes over between 75 and 80. and normal hexane and methyl boiling" points 68-95 and (about) 71'5 respectively. As another example of a pair of very closely related liquids which are frequently met with.262 FRACTIONAL DISTILLATION CHAP. heating with fuming nitric acid. efficient still-head the separation is very Hexamethylene and a Volatile Heptane. Later on. of this It has thus been shown that in the of American or Galician petroleum. Amyl Alcohols in Fusel Oil. it was concluded that that substance had been separated in a pure state. in a complex mixture. Summary distillation of Results. the second pair are not. two isomeric hexanes boiling at nearly the portion pentamethylene with the same temperature. 61. almost pure normal hexane was obtained indeed from the fractions between 66 and 68'95 the normal .

Such Pentamethylene have been described are of common occurrence in the distillation of petroleum. Pentamethylene in Petroleum. It in crystallisation. in Small If one or more components of a complex mixture are present in relatively very small amount. it was found that there is a small quantity of pentamethylene . VII in The manner small of (p. American. and Trimethyl-ethyl-methane. For example. and of notably higher specific gravity (7). and it is only by keeping a careful record b. they are apt to be altogether overlooked.xix PENTAMETHYLENE 263 that the liquid could be partially but not completely frozen an ordinary freezing mixture and eventually. by fractional pure hexamethylene was obtained with practically the same boiling point as before but with a definite melting point. of the weights which the presence of a relatively very quantity hexamethylene was recognised in American petroleum has been described in Chap. Galician and Eussian petroleum all contain a certain amount of pentamethylene which boils at about 50 but there is difficulties as . and it appears to be impossible nearly to separate these hydrocarbons by fractional distillation. 140). One or more Components Present Quantity. and temperature ranges of the fractions. was evident that the substance separated by fractional distillation only was a mixture of two hydrocarbons. that. or else the two substances form a mixture of constant boiling point almost identical with that of hexamethylene. of which the former was present in much the larger quantity. and by calculating the values of Aw/A or mapping the total weights of distillate against the temperatures. nexamethylene and a heptane. and that either the boiling points are almost identical. also present a hexane. the existence of these components can be detected. trimethyl-ethyl-methane. In a similar manner. boiling at the same temperature.

and collected between about 37 and 60. gave no indication of the presence of any substance boiling in the neighbourhood of 50. TABLE I. 77. .264 FRACTIONAL DISTILLATION also (and trirnethyl-ethyl-methane) in American petrol- eum (2). as will be seen by Table 77. The first fractionation of all the available distillates from petroleum ether obtained with the combined dephlegmator and regulated temperature still-head.

35). 6090 and 90120. too low acid amounting to 16*3 grams. however. 3060. yield glutaric therefore contained pentamethylene. As the calculated vapour density of . c. 39 '2. much larger in Eussian than in American petroleum. was too small to allow of the continued with the combined dephlegmaseparation being tor and regulated temperature still-head. and The specific gravity. and then treat each fraction with a mixture of strong nitric and sulphuric acids. Benzene in American Petroleum. 0'751 at 15/15 vap.Point.xix in the MIXTURES OF CONSTANT BOILING POINT 265 subsequent operations. when the value of Aw/A rose to 163 '0 for the fraction from 49-45 to 49'55. we shall find that a considerable amount of dinitrobenzene will be obtained from the fraction from 60 90 and of dinitrotoluene from that between 90 and . who showed that both these hydrocarbons are present in Eussian petroleum. rising from 12-9 in the fourth to 34-9 in the seventh and 48'0 in the ninth. it would appear that the distillate consisted of a mixture of pentamethylene and trimethyl. being. was found to on oxidation with nitric acid. 0'7035 at /4 was. The results are quite in accordance with those obtained by Markownikoff (8). but six additional fractionations were carried out with a 12 column dephlegmaThe quantity of material tor.ethyl-methane in approximately equal molecular proportions. if we petroleum to distillation. the quantity of pentamethylene and other naphthenes as compared with that of the paraffins. too high for pure pentamethylene (sp. den. The distillate. gr. hexane is 43. subject a quantity of American collecting fractions from For example. or more of the components form mixtures of constant boiling point the difficulty in interpreting the experimental results is greatly increased. and the vapour density. Mixtures of Constant When two Boiling. however.

but it is practically certain that the second is the correct one. it would then be found. In order to explain these facts two assumptions may be made either (a) it is not benzene and toluene that are present in American petroleum. the fact that when American petroleum is distilled. stated long ago to be the case by Schorlemmer. we were to separate it into nine with a range of 10 degrees each. instead of collecting the distillate in three large fractions. as already stated. Possible Explanations of Results. say. p. is distilled. although the boiling points of benzene and toluene are 80 2 - and 110'6 respectively. which dinitrobenzene and dinitrotoluene respectively are formed by the action of nitric and sulphuric acids or (b) benzene and toluene. this was. The obvious conclusion from these results is that benzene and toluene are present in American peti*oleum and If. when mixed with the paraffins present . 68-95) with. however. the benzene cannot be separated but comes over with the hexane at 68' 95. that on treating each fraction with the mixed acids. after repeated distillation. Lastly. of benzene.266 FRACTIONAL DISTILLATION CHAP. . was actually made by one observer. dinitrobenzene would be obtained almost exclusively from the distillate that came over between 60 and 70 and dinitrotoluene from that between 90 and 100. that when a mixture of normal hexane (b. the greater part of the benzene comes over below 69 (mostly at 65 66) is explained by the presence of isomeric hexanes which boil at lower temperatures than normal hexane and also carry down the benzene with them. 10 per cent. in petroleum come over chiefly at temperatures 14 below their true boiling points. in fact. so far as the conversion of a hydrocarbon other than benzene into dinitrobenzene is concerned. or 15 The first assumption. 120. for it has been found. but some other more volatile compounds from .

Hecht (9) found that on distilling a mixture of acetic. the middle portion contained chiefly butyric acid and the last portion . and if the amount of acetic acid was large. 1. it tends to separate into three or more of the following components. 107'1 5. but it may be pointed out that. the whole of the cenanthylic acid came over in the first portion of the distillate. 3. and in that case the highest fraction would consist of the formic acid. and if the amount of the acid was relatively small it might all be carried down with the water at temperatures below 107. first of the remainder of the benzene in the second while the third fraction or residue consists of the fraction. when a mixture of formic. most volatile of the original components. Butyric acid-water (mixture of minimum boiling point) . water. yet it is very difficult to separate the acid from its dilute aqueous solution by distillation. 2. 6. The acids would then come over in the reversed order of their boiling points. is distilled through a very efficient still-head. benzene and water has already been referred to (p. 267 of Ethyl Alcohol The behaviour mix- tures of ethyl alcohol.xix ALIPHATIC ACIDS AND WATER Benzene Water. minimum boiling point with water. 217) . Again. comes over in ^he the three fractions into which the distillate tends to separate. 119'0 163'8 With a large amount of water. when alcohol a mixture of equal weights of benzene and of 95 per cent. butyric and oenanthylic acids with much water. Boiling point. acetic and butyric acids with water is distilled. Water Formic acid Formic acid-water (mixture of Acetic acid Butyric acid 99*2 lOO'O 1007 maximum boiling point).water mixture of maximum boiling point. 4. the substance of highest boiling point. the last fraction would consist of that acid but although acetic acid does not form a mixture of . Aliphatic Acids and Water. the whole of the butyric acid would come over in the lowest fraction. alcohol.

I. " 5. Markownikoff. 1881. 73. All the alcohols with the exception methyl form mixtures of minimum boiling point with water. "Composition 1898. 73. Berichtc. 1898. Chem." Berl. Vapour Pressures. 1. 1222. from American Chem. follow the same order as those of the alcohols themselves. 103. 154. from j8-Hexyl Iodide. 31. Markownikoff. 440. " The Volumes and and Young Fortey. MX contained acetic acid nearly free from the other two. 1897. 3. Hecht. American Petroleum. 387. 301.. 905. 8. 209. 1897. 75." Liebiy's . 1899. and cenanthylic acid soluble in water. 1897 X 30. 9." Liebiy's Annalen. Specific Critical Constants of Hexamethylene. " Some Hydrocarbons Normal and Isopentane. the possibility of their formation must always be borne in mind in interpreting the results of a fractional distillation. "On the Presence of Methyl Pentamethylene in Caucasian Petroleum Ether. Soc. Berichte. Trans." Fortey. 1898. "On Isoheptoic Acid Annalen. up to amyl alcohol at any rate." Berl. is nearly in- of Alcohols and Water. Young and Petroleum. Berichte. Chem. Hecht points out that acetic acid is miscible with water in all proportions with considerable heat evolution butyric acid is also miscible with water in all proportions but very little . " Hexamethylene from American and Galician Petroleum." of Trans.. Soc. 71. but as the boiling points of the binary mixtures. Young. REFERENCES. "Researches in the Hexamethylene Group." Trans. Aschan. 4. Soc." Trans. Soc. 7. Thomas.. 30. Zelinsky.. 2. "On Cyclic Compounds. 1898. As mixtures of constant boiling point are of frequent occurrence among liquids which are chemically not closely related to each other. Chem." Berl. heat change is observable. 873. 321. 1803. 6. such reversals as are observed with the acids do not occur. On Methyl Cyclo-Pentane from different Sources and some of its Derivatives.268 FRACTIONAL DISTILLATION CH.

H. or corrected to. degree from 10 In Table 78 the values of 0-000172 Ht are given for each to 30 and for pressures at intervals of 10 mm. = HH 0-000172 Ht if the scale is etched on the glass 2.APPENDIX Correction of Height of Barometer for Tempera- The atmospheric pressure is usually expressed in millimetres of mercury read at. In practice the height. say. . t. 800 mm. or Q H = 0-000161 Ht if a brass scale is used. 0. at any temperature between 10 and 30 ascertained from the diagram. calculated : H . temperature of the room. The correction for any pressure up to 800 mm. may then be easily . of the barometer is read at the ture. and the value at by means of one of the two following equations 1. from 720 to 780 for the mm. the readings are smaller than 720 mm.. for alternate degrees from 10 to 30 on curve paper.. and in Table 79 the values of 0-000161 If same temperatures and pressures. it is convenient Ht to plot the corrections against the pressures from to.






65. SO American petroleum. 6. 238 intermediate between paraffins and water. 83. 101. 213 binary. 102. 261. SSS. water and. petroleum Amyl alcohols in fusel oil. by distillation 1. on hydro-derivatives of-benzene. 211 efficient still-head. 217 ethyl alcohol. 104. 230. The Phase Rule. 268 pour. 11U BANCROFT. alcohols. 65 composition of liquid and vapour. 247 and benzene. ancient. mixtures of three components. 43 Alembic. 220 general conclusions. ACETONE and carbon disulphide. 45 benzene. on iniscibility vapour pressures. on Russian petroleum. the more important are indicated by the numbers being printed in thick type. iniscibility.. and water. water and. separation of three liquids by dist illutii >n. 206 : and heat and volume 35. 1UU BALY. 42 Analysis. uiiscibility. 105. 205 mixtures of two components. 42 boiling points. influence of molecular. 3 ALTS< inn.INDEX Ifhen there are several references. 204. 3 modern. 209 mixtures of of maximum vapour t>7 complex mixtures. 221 is inap- carbon totrachloride. quantitative. 43 on composition of liquid and va67. lilt Barometer. and chloroform. 42. 45 on vapour pressures. abstraction of water from Apparatus for distillation. 4 ARISTOTLE. 267 42 Aniline and water. 206 choice of still-head. 212 mixtures of constant boiling point. 11$ theoretical relations between weight and composition of distillate. 4. 97 and T 2 . 3 ASCII AN. use of water or oil. influence of molecular. 43. 99. 3 Alexandrian still. r. critical pressures. Names of authors are printed in small capitals. 44 higher. boiling points. Ili9 advantages of 200. 48 and water. pressure. 237 and alcohols. 269 BARRELL. 49. 237 compared with water. 53. 67 and hexane. 7 Benzene. 205 estimation of loss. 128. 268 Association. 109. 262 Amyl (iso) alcohol and benzene. 45 volume and heat changes. still used for preparing strong. 214 ternary. 42. Numbers printed in italici refer to the bibliographies at the ends of the chapters. 101 and chlorobenzene. on distillation. 253 loss by evaporation. heat and volume changes. 42. 109 and carbon disulphide. 49 minimum boiling point. oxygen and nitrogen. 202 Alcohols. still. 78 Alcohol. THOMAS. BAEYER. 38. by liquids. benzene and water. correction for temperature. 42. absorption of moisture from. 144. on formation of mixtures of mini- lower. critical constants. 102 Acids. 32 on vapour pressures. 65. M7 cases to which method plicable. 107. 238 mum boiling point. aliphatic. 41 changes. 111 Attractions. and benzene. 42. Bath. and YOUNG. 107. 91 55 Air. 38 vapour pressures.

on distillation. 55 components closely related. 9. 41. 191 ethyl alcohol. 56. for pressure. 241. 246 distillation and water. 48 volume changes. 42. 7'. 81. for 266 alcohol.'. 51. 143. 58 calculation of. 173 of improved with plain still-heads. 90. 63 -molecular composition curves. 36. 100. p BIOELOW. 102 COFFEY. 41. 110. 21. 217. 232. 42. 41. 63. 96. 246 choice of still-head. cause of. 93. 35. calculated curves. 42 and benzene. 28. 172. 61 comparison of calculated and observed. 228 boiling and ethylene dichloride. 35. 26 of mixed liquids. mcthylal Carbon tetrachloride and benzene. 240 of pure liquid. x'f. 72. 100. 25 depends on partial pressure of vapour.63 of still-heads. Comparison sure. 38. 2. M 231 and water. 42. benzene v. . Boiling point. composition of liquid and vapour. 63 and carbon tetrachloride. 37. 158. 185 specific gravity of mixtures. to normal pressure. under reduced pres37. separation of. 106 CARVETH. 98. 19 Bromobenzene and ehlorobenzere. 251 mixtures of on on on on on heat changes. 27. 73. 67. 98. correction of. 179 regulated temperature still-head. 181. and hexane. 45. benzene and water. 87. and chlorobenzene. 113 BERTHELOT. 190 Chemical relationship. by dynamical method. 152 separation of. 50 changes on mixing composition of liquid and vapour. 93. 42. from vapour pressure. 113 influence of rate of distillation on 90. BREST. 105. 103 mixture of constant point. 237 14. 56 not closely related. 107. 65. 101 and carbon tetrachloride. 43. 72. 209 . determination of. c Butyl (tertiary) alcohol. 40. 179. 72. 152 components of mixture. 63 boiling points of residue and distillate. 202 52. 21 Carbon disulphide and acetone. 24 Bulb still-heads. influence of. 60 observed and calculated values. 97 and bromobeuzene. 231 toluene. molecular attractions.27 Boiling points of residue and distillate. 53 of infinitely miscible liquids. S8 formula} for non-miscible liquids. S8 continuous distillation.276 Benzene and ethyl sures. 52 of partially miscible liquids. 142. 85. 65. 95. 76. 59 maximum deviation of curve and methylal. 170 dephlegmator. INDEX vapour pres- BROWN. 64 composition of liquid and vapour. 28 Butyl (iso) alcohol and benzene. 31. F. 231 BROHL. 24. 101 100. 20 Bubbles. 237 in petroleum. water and. 36. 164 Bumping. 39.. S3. 55 and molecular composition. volume liquids. 115. 65. composition of liquid and vapour. 24 prevention of. vide mix- tures on separation of components. 42 of bearing on distillation. 101 65. 200 still. 247 reduction of. 224. influence on separation. 45. distillation under reduced pressure.. ISO. and isobutyl and separation. 48. 158 of dephlegmators. on vapour pressures. 90 Chloroform and acetone. 22. 56. formation of. 61. 14. 45. 57. 150. 108 70. 92 electrical heating. 51 CALIBRATION of thermometers. 68. 27. dephlegmator. 49 boiling points. 39. 56. 31 effect of impurities on. 68 and toluene. 7. 45 miscibility. 110 constant. 35. 42. r. 21. 39. D. 87. 42 alcohol. D. 150 Chlorobeuzene and benzene. 42. and carbon disulphide. 81. determination by statical method. 220. 246 on formation of mixtures of constant boiling point. 65. 63. 116 definition of term. 46. 232. composition of liquid and vapour. 113 infinitely miscible liquids. 247. 220. 42. influence of form of. 49 from straight line. 65 of non-miscible liquids. 83. water and. 265 BERTHELOT. 42 and ethyl acetate. 101. 267 tertiary butyl alcohol.

Linebarger. purification of gases. determination of. 224 graphically from vapour pressures or boiling points. 199 Egret's. remarks on construction influence of numbers of. 73 Lehfeldt. 23. 171 Glinsky's. 1 on manufacturing rate of. Carveth's apparatus. 239. 89. 67. 209 complex mixtures. 150 theoretical relation and composition Distillation. 170 flow of liquid through reflux tubes in. 188 modified for three components. continuous distillation. 213. 176 Correction for imheated cury. 189 temperature. 175 Brown's. non-miscible liquids. 172. 54 I'. 226 by Brown's formula. by fractional distillation. to still.trir readings. 91 infinitely miscible liquids. Carveth. 228. by successive approximations. 172. IS6. 144 Condenser. 25 COUPIEK. in.n. 206 . 93 Brown. 18 Bredt's apparatus. 173 . Linnemann's. 14. 212. 116 of tlu-rmo]ii. 38 of partial pressures. 10. 206 object of. 12. 206 throe components. 18 Thome's apparatus.It Margules. 14. Brown. criterion of purity of. 180 Construction of dcphlegmators. V. 90. 230 of liquid and vapour phases. 197 CRAFTS. prevention of leakage. 26 CONSTAM. 110 DIOSCORIDES. 144 determination of. 174. 120. 226 from composition of liquid and vapour. 4. 99. 214 ternary. 228 results obtained. Liebig's.INDEX Composition. and weight." 204 Dehydrating agents. 6 correction of temperature 17 Condensing point of vapour. 76 by second method. law Dephlegmator. loss of material in. v. 174. 269 Ixiiling point for pressure. 20 with improved still-head. 115 between weight of. 178 DMBBT. 178 return of liquid from. 75 200 Dcphlegmators. 20 witli steam. 116 continuous. 93 of width of reflux tubes and depth of traps. 175 of size of constrictions. 11 correction of boiling point for pressure. 176 Zawidski. 158 scale. 21. 85 theoretical. moditi cations of. GOLDSCHMIDT Constant. still. c(=dt/Tdj)). of distillate. on large scale. 27. 217 cases to which the method is inapplicable. 175 Le Bel-Uenninger's. 104 Zawidski. zero point of thermometer. 61 values of. 92 formula of Wanklyn and Berthelot. 175 Dubruufaut's. '." -'"> " middle point. 199 comparison of. 202 in laboratory. 223 by separation of pure mixture. 176 of sections in. determination by distillation method. 241 relation between boiling point of residue and. 225. boiling point under reduced pressure. l'J6 under constant pressure and at con- mixtures of. remarks on. 28 Definition of term " boiling point.. 205. 112 Constant temperature still-heads. 182 Duhem and modified. Brown. 104 Lehfeldt. 200 Cofifey's. 14. 221 of mixtures of constant boiling point.80 by third method. 2-23 DALTON. 14. 21 Critical pressures of various liquids. 186 Deviation of boiling point-molecular composition curve from straight line. --'3'J. 204. 253 mixtures of constant boiling point. barometer for temperature. (30 DKWAR. 16 'boiling point. 117. 81 by fourth method. 179 Young and Thomas's. 44 mixtures of two components. 213 binary. 174. 200 Savalle's. 253 277 Cotton wool on thermometer bulb. action on alcohols and on water. 171. condensation of vapour. 174. 107 general conclusions. 172. 110 partially uiisciblc liquids. 19 stant temperature. 177 arrangement of reflux tubes in. 71 by first method. :i of <>f column of mer- Briihl's apparatus. 173. 15 related to molecular weight and constitution. 180 of weight of distillate below middle Lord Kayleigh's method. 3 Distillate. 220 general conclusions.




r. 2,


quantitative analysis,


experienced in,


Ethyl alcohol and benzene, v. benzene and methyl alcohol, 38, 40, 64 and water, vapour pressures, 42, 40
boiling points, 65, 67,
221, 238, 242, 247

263, 265
directions for carrying out, 114 purposes for which, required, 23l isolation of single substance from mixture by, 239, 240 separation of components by


behaviour on distillation,





point, 65, 67, 237


determination of quantitative composition by, 204, 239, 253 general study of complex mixture by, 239, 253 systematic fractional : 117 alteration of temperature ranges,

composition of, 221, 225 benzene, water and, 217, 237, 267 composition of ternary mixture,

still used for preparing strong, 202 benzene and toluene, 38, 40, 63 bromide and iodide, 36 Ethylene dibromide and propylene di-

procedure dependent 011
fractions, 119 separation of two

number of


bromide, 37, 95 dichloride and benzene, 42, 96, 108 Evaporation in absence of air, 22, 71 in presence of air, 23
loss of material by, 120, 205


with improved

still-head, 191




separation of three components, 128, 251 with improved still-head, 193 of complex mixtures, 137, 212, 239,

Evaporator still-heads, 165

graphical representation of results,
v. graphical interpretation of results, v. interpretation hexamethylene in petroleum, 140, 262 hexaues in petroleum, 255, 259 pentanes in petroleum, 138, 182, 185, 253 relative rate of separation of components, 120, 134, 221 relation between boiling point of residue and distillate, 150 between relation theoretical

FIXED points on thermometer, 9, 10 Flame, protection of, from draughts, 8 FORCRAND, copper sulphate and methyl


alcohol, 45 for correction of

barometer for

temperature, 269 of boiling point for pressure, 14

unheated column of mercury,


vapour pressures of mixed liquids

for molecular attractions (Galitzine and D. Berthelot), 33 for maximum deviation of theoretical boiling point-molecular com-

weight and composition of
tillate, 144


position curve from straight line, 63
for composition of vapour from iionmiscible liquids, 89, 90 for composition of liquid and vapour


composition of liquid and vapour, 73, 88

DUHEM, formula for composition of liquid and vapour, 103, 104, 113 Dynamical method for vapour pressures,
for boiling points of

phases, 93

Wanklyn and



93, 94

pure substance

of mixtures, 58

modified, 99 Margules, 104 Lehfeldt, 104

Duhem and

and Brown compared,


Zawidski, 107

intermittent, 25


and Brown compared, 107 relation between weight and

EOROT, dephlegmator, 200
Errors in boiling point, 12, 25 of mixtures, 39 in composition of liquid and vapour phases, 78 in interpreting results of fractional * distillation, 266 of thermometers, 9 Ethyl acetate and propionate, 38, 40, 63, and carbon tetrachloride^ 42, 68

composition of distillate, 145 FORTET, chloroform and methyl acetate

hexamethylene in Galician petroleum, 142, lltlt and YOUNG, vapour pressures and
boiling points of mixed liquids, 42, 50, 80, 98 alcohols and water, 42, 50, 67, 92 alcohols, benzene and water, 42, 50, 67, 230

FORTEY and YOUNO, fractional distillation as a method of quantitative
analysis, 222


hexamethylene, 268
Fractional distillation, v. distillation Fractions, collection of distillate in, 115

Hexane (normal) and alcohols, 48 and benzene, 42, 65, C8, 70, 142, 241, 247, 266 and n-octane, 38, 40, 63
"preparation of pure, 201




required, 115, 119, 153,

FRANCIS, fractional distillation under reduced pressure, 196 and YOUNG, heptanes from petrol-

eum, 196 French column apparatus, 198 FRIEDEL, hydrochloric acid and methyl
ether, 66, 69

255, 259, 262 still-heads for liquids of, 165, 240 HOLLY, mixtures of minimum boiling point, 67, 68, 70 t Hydrogen chloride and water, 228 Hydroxyl group, influence of, on constant c, 16 on formation of mixtures of miniboiling point, 65, 247 Hygrometer, wet and dry bulb, 31

Hexanes in petroleum, High boiling-points,





alcohols in, 262

of liquid and vapour, 89 Galician petroleum, v. petroleum GALITZINE, molecular attractions, 33, 35,

GAHL, composition
80, 81,


influence point, 27




on temperature-weight of distillate
curve, 240, 242, 243, 244 Infinitely miscible liquids, v. liquids Interpretation of results of fractional distillation 139, 240, 258, 266 causes of confusion, 2, 253. 263.

GARDNER, mixtures
ing point,
69, 70




GLINSKY dephlegmator, 172,


of boiling point, 67, 70 of results of Graphical representation fractional distillation : benzene and toluene, 122 methyl, ethyl and propyl acetates, 135, 137, 138 pentanes from petroleum, 139 impure liquids, 240, 243, 244, 245 hexanes, &c., from petroleum, 256 GUILLAUME, exact thennometry, 21 Gum, for prevention of leakage, 20 GUTIIRIE, vapour pressure and composition of mixtures, 34, 50 ethyl bromide and iodide, 36

265 two substances

of nearly



boiling point, 253

pentanes in petroleum, 138, 253 hexanes, in petroleum, 255


alcohols in fusel



heptane, 262



253, 263

and trimethylethyl-methane. 263

one component in small amount,

pentamethylene in petroleum,

hexamcthyleue in petroleum, '
140, 262

68 70






boiling point,


point, 253, 265 benzene in American eum, 265

HARTMAN, mixtures
70 42 alcohols


two substances,
liquids, 32, 39,

Heat changes on mixing

and water, 45

and n-hexane,

and benzene, 42, 49
closely related substances, 40 substances not closely related, 41 source of, 6 steam as, S

alcohol-benzeue-water, 217, 267 aliphatic acids and water, 267 errors in, 266 v. amyl Isoamyl, Isobutyl, v. butyl Isopropyl, v. propyl


JACKSON and YOUNG, benzene and n-hexaiie, 42, 68,


HKCHT, aliphatic acids and water,


1 1 I


still-head, 168, 170 and Le Bel, dephlegmator,

KOIINSTAMM, vapour pressures of carbon

and chlorobenzene,


172, 173, 179
in petroleum, 262 i'l K:\.innjtliylene in petroleum, 140, 142,
italic (volatile)



of KONOWALOFF, vapour pressures alcohols and water, 42, 45, 67, 92, 113 KOPP, distillation by ancients, - /

KKKIS, on fractional


MABKOWNIKOFF, hexamcthyleiie
sian petroleum,

KUENEN and ROBSON, mixture


mum boiling point,

of still-heads, 104, 170 of maxi66, 70


in Rus261 in Russian pentamethyleiie


of partial pressures, Dalton's, 23,

petroleum, 261, 368 and pentaniethyleno trimethylethyl-methane in Russian petroleum, 265, 268 Maximum deviation of boiling pointmolecular composition curve from
straight line, 60

Leakage, prevention


LE BEL and HENNINGEK, dephlogmator,
72, 173, 179
37, 42

MENDELtEFF, alteration
ranges, 121, 11,3 specific gravities of alcohol, 226



LEHFELDT, vapour pressures of mixtures,
composition of liquid and vapour phases, apparatus for, 75, 88 determinations of, 98 formula for, 1 04, Ilk of mixtures of constant boiling
point, 230

aqueous ethyl-

Methyl alcohol and benzene, 42, 237 and ethyl alcohol, 38, 40, 64

and water, 42, 46, 65, 207, 237 and propyl acetates, fraetiunation
of, 128,


Methylal and carbon disulphide, vapour
pressures, 42

rttm, 4

Syntagma alchemlai arcano4, 6,

composition of liquid and vapour,
102, 106, 109, 110

LIEBIO, condenser,
tures, 36, 50 composition of


pressures of


composition of mixture of mini-

mum boiling point,

227, 229, 230

liquid and vapour phases, 80, 97, llii LINNEMANN, dcphloginator, 171, 179 Liquid and vapour phases, v. composition return of residual, to still, 158, 178 Liquids, mixed, boiling points of, v. boiling points vapour pressures of, v. vapour pressures iiou-miscible, boiling points of, 31,

Methyl pentamethylene in petroleum, 261 Middle point, definition of term, 204 weight of distillate below, 121, 204
Miscibility, influence of molecular attraction, on, 32 of alcohols with water, 45

with n-hexane, 48 with benzene, 49 Mixtures of constant boiling point,
142, 213, 265
list of,



vapour pressures of, 33, 52 composition of vapour of, 89, 90,
distillation of, 31

determination of composition by distillation, 213, 223 by separation of pure mixture, 223

by method of successive approxivapour pressures
mations, 224 graphically from vapour pressures or boiling points, 226 from composition of liquid and vapour, 226 by Brown's formula, 228 indirect method of separating com-

partially miscible,

33 boiling points
of, of,



composition of liquid and vapour


infinitely miscible,

vapour pressures of, v. vapour pressures normal behaviour of, 34

boiling points of, v. boiling points proofs of purity of, 241 still-heads suitable for high boiling, 240 for volatile, 182 Loss of material in distillation, 120, 135, 193, 195, 205 estimation of, 206

ponents of, 231, 238 influence of chemical relationship on formation of, 65, 247 of minimum boiling point, 65, 214 list of binary, 67 list of ternary, 68 of maximum boiling point, 66 list of, 69

maximum and minimum


pressure, 46

apparatus for determining boiling

v. distilla-

MAGNUS, vapour pressures



fractional distillation of,

MANSFIELD, still, 197 MABCHIS, rise of zero point, 10 MAKQULES, formula for partial pressures

and composition of mixtures,
107, 113



tion separation of components of, r. separation boiling points of, v. boiling points vapour pressures of, v. vapour pressures

Modifications of condenser, 17 of receiver, 17


Petroleum, American, trimethyl-ethyl-



still-head, v. still-head Moisture, absorption of, by liqviids


air, 7S

Molecular association, influence
pour, 102, 107, 100, 111

on vapour pressures, 48 on composition of liquid and vaof,

methane in, 263 separation of pentanes from, 138, 182, 185, 253 Galician, heptane in, 262 hexanes in, 255 pentamethylene in, 263 of separation hexamethylene from, 142, 262 Russian, benzene in, 143
hexamcthylcne in, 261 pentamethylene in, 263
PETTITT, mixtures of point, 67, 58, 70

of mixtures of minimum boiling point, 65, 247 Molecular attractions, influence of, on miscibility and on heat and volume changes, 32

on formation



on vapour pressures,

35, 41

Molecular composition and boiling points of mixtures, v. boiling points

PISTORIUS still, 201 PLINY, condensation of vapour, 3 Pressure, correction of boiling point
14, 21, 27 critical, of various liquids,



NAUMANN, non-miscible liquids, 90, 113 NERNST, benzene and acetic acid, 67, 70 Nitric acid and sulphuric acid, for purification of petroleum, 130, 200 fuming, action of, on hydrocarbons,

curve of vapour, 22 reduced, apparatus under, 18

for distillation

Thome's, 18
Bredt's, 19
Brilhl's, 20

Nitrogen and oxygen,



in thermometers, 11 Non-miscible liquids, v. liquids Normal pressure, reduction of boiling point to, 14. ?/, 27, 116

distillation under, IS with improved still-head, 196 prevention of leakage, 20 bumping during, 7, 28

NOYKS and WARFEL, ethyl water, 67, 70, 232, 248



OCTANE (normal) and n-hexane,



OSTWALD, stirring arrangement, 183 Oxygon and nitrogen, distillation

vapour, v. vapour pressure. Pressures, law of partial, 23, 28 Prevention of bumping, 7, 28 of leakage, 20 Propyl (normal) alcohol and benzene, 42, 237 and water, 42, 46, 65, 215, 237 benzene, water and, 237 (iso) alcohol and water, 242 benzene, water and, 237


Propylcne dibromide and ethylene bromide, 37, 95 Protection of flame from draughts, 8 Purity of liquids, proofs of, 241


Him to prevent leakage, 20, 21 I'arafVms compared with alcohols, 43 Partial pressures, law of, 23, 28
Partially miscible liquids, v. liquids. PAYKX, Precis de Chimie Industrielle, SOS Pear still-head, 165 Pentamethylcne in petroleum, 263 Pentano, normal and iso-, relation between boiling point of residue and distillate, 151, 153

QUALITATIVE and quantitative study complex mixtures, 253


Quantitative analysis by fractional distillation, v. analysis


from petroleum,


RAMSAY, purification of gases, 182 RAMSAY and YOUNO, some thermodynavapour
mical relations, SI pressures of
liquids, 31

182, 185, 253

Petroleum, American, comparison with




of, ///

Ranges, temperature, for fractions, 115
alteration of, 121

difficulties in fractional distillation


of, 2, 253, 203, in, 143,

265 265


In-] italic in,


hexamathylene in, 140, 262 hexanos In, 255, 'J.v.i methyl pentamethylene in, 261 pen tarn ethylene ii

of distillation, 117, 158 relative, of separation of components, 120, 134, 221 RAYLEIGH, LORD, continuous distilla-

tion, 188, 190

composition of mixture
boiling point, 227, SSI



Receiver, 4 modifications

SPEYERS, mixtures of



of, 17 for reduced pressures, 18 ; 18


point, 66, 69 Spheroidal state, 30 Spiral still-head, 163 Statical method of determining boiling points, 22, 51

Bredt's, 19
Brtthl's, 20 Reduced pressure, v. pressure Reflux tubes in dephlegmators, 171, 172, 173 width of and depth of traps in, 176 flow of liquid through, 177 arrangement of, 178 RKGNAULT, evaporation in presence of air, 23 vapour pressures of liquids, 33 Regulated temperature still-heads, 180 Relation between boiling point of residue

vapour pressures, 36 Steam, as source of heat, 8
distillation, 54

Alexandrian, 3

and distillate, 72, 81, 150 between Brown's and Lehfeldt's
formulae, 105

'ordinary, 4, 5 modifications of, 17 for aniline, 55 for composition of liquid and vapour, Brown's, 73 Lehfeldt's, 75 Zawidski's, 76, 77 Carveth's, 81, 82
for continuous distillation, 186, 188,


and Zawidski's formulae, 107 between weight and composition
distillate, 144

manufacturing purposes, 196


Mansfield's, 197 Coupler's, 197

French column, 198 weight

of constant c to molecular

and constitution, 16
Relationship, chemical,

ROBSON and KUENEN, mixtures

Rod and

mum boiling point,
point, 66, 69, 70

chemical of maxi-

66, 70 disc still-head, 163 mixtures of



RYLAND, mixtures

of constant boiling

Savalle's, 199 Pistorius', 201 Coffey's, 202 return of residual liquid to, 158, 178 Still-head, 4, 5 modifications of, 154, 161, 171, 180 theoretically perfect, 156 characteristics of good, 157 weight of vapour and liquid in, 159

point, 65, 66, 69
of, 225, 230,


plain vertical, 160 inefficiency of, 155 influence of length on efficiency of, 160

SAVALLE, dephlegmator, 200

width on

efficiency of, 161


SCHROTTER, freezing of water in
air 29 31 Separation,

by fractional two components,

distillation, of 114, 191, 246

separation of benzene and toluene with, 115 of three esters with, 128 sloping, 162 spiral, 163

of three components, 128, 193, 251 of single substance from mixture,

"rod and

disc," 163

bulb (Wurtz), 164 "
pear," 165

pentanes from petroleum,


182, 185, 253 continuous, by distillation, of components, 186, 188


"evaporator," original, 166 modified, 167

Hempel, 168 dephlegmator type,



of three components, 189 indirect, of components of mixture of constant boiling point, 231,


benzene from alcohols, 237 from isobutyl alcohol, 231 of water from alcohols, 237 from tertiary butyl alcohol, 232 of components, relative rate of, 120,
134, 221 Sloping still-head, 162

" " regulated or constant temperature," 180 Warren's, 180 Brown's, 181, 185

combined with dephlegmator, 183 advantages of efficient, 193, 195, 209,
211 choice of, 179, 205, 249 suitable for liquids of high boiling point, 165, 240 for volatile liquids, 182 Still-heads, comparison of, 158, 179 " evaporator," 165 improved, can be used for reduced pressures, 196

Source of heat,

6, 8 Specific gravities, evidence afforded by, 241, 260, 261

SPEVERS, vapour pressure and molecular composition, 35, 50

Still-heads, improved, comparison plain, 179. 193,195 fractional distillation with,



Thermometer, correction column in, 12


determination of fixed points on,

193 Successive approximations, method of, 224 Sulphuric and nitric acids, for purification of petroleum, 139, 260
Tables for use in laboratory work

with nitrogen,


normal, 10 redetermination of zero point
volatilisation of 11


mercury in stem

THOMAS and YOUNG, pentanes from* petNo.

Tahle of boiling points for ther-

mometer corrections, 10 for un2. Table of corrections heated column in thermometer, 13 3. Table of values of T,dp/dt and c, 15 78 and 79. Tables for correction of
height of barometer for temperature, 269, 270, 271
16. List of

roleum, US, 268 dephlegmator, 173, 179 combined with regulated temperature still-head, 183, 185

BARRELL and YOUNU, separation of three liquids by fractional distillation,


THORNE, receiver for reduced pi cssures,18 THORPE, correction for unheateu column in thermometer, 12, 21 mixture of minimum boiling point,
67, 70, 93,

mixtures of



boiling point (binary), 67 17. List of mixtures of minimum boiling point (ternary), 68 18. List of mixtures of maximum boiling point (binary), 69
Tables illustrating systematic fractional

Dictionary of Applied Chemistry, 203

Toluene and benzene, vapour pressures and boiling points, 38, 40, 63
boiling points of residue and distillate, 150, 152 separation of, by distillation, 115, 191 Toluene and carbon tetrachloride, 42, 98, 106 and ethyl benzene, 38, 40, 63 Trimethyl-ethyl-methaue in petroleum, 263


Arrangement of temperature ranges of fractions, 116 36 and 37. Fractional distillation of





ordinary still-head, 118, 124 52. Fractional distillation of benz-

ene and toluene with improved still-head, 192 38 and 39. Fractional distillation of three esters with plain vertical still-head, 130, 136


PHILIP, Philosophorum, seu de secretis Natune, 154

VAPOUR, superheating



Fractional distillation of three






head, 194 Detection of hexamethylene in American petroleum, 141

51. Illustration


with regulated temperature
still-head, 184

in petroleum, first three fractionations, 254 Detection of pentamethylene in petroleum, 264 Temperature, correction of, in distilla74.


tion, 115

liquid phases, v. composition pressure, mixtures of maximum and minimum, 34, 46 pressures, curve of, 22 of pure liquids, 23 of non-miscible liquids, 33, 52 chlorobenzene and water, 52 of partially miscible liquids, 33 aniline and water, 54 of infinitely miscible liquids, 34 of mixed liquids, 32 closely related, 36, 38, 40, 41 not closely related, 41, 42 relation between molecular composition and, 35

height of barometer for, 269 middle, weight of distillate below,

determination by statical method, 36



ranges of fractions, 115
alteration of, 121 range, ratio of weight of fraction to,

122, 141, 240, 253, 254, 256, 258, 263, 264

by dynamical method, 36 comparison of heat and volume changes with, 39 chlorobenzene and bromobeuzene, 56 alcohols and benzene, 42, 49 and water, 42, 45




calibration of, 9, 21

141. 42. 49 and water. BREAD ST. 254. 69. 32. 267 tertiary butyl alcohol. paraffins and. AND BUNOAY. 231 power of mixtures. "pear. 255. 264. lUt. 52. 65. 218 chlorobenzene. 41. regulated temperature stillhead. 222. E. Jackson." 163 . v.. VAN DEB. 68. 263. 268 " rod and disc. 40. 37. 93." 165. Francis. in rarefied air. Barrell. 69 mixtures of constant boiling point. lUt composition of American petroleum. 89 still-head. HILL. 67. 110 formula for. 182 suitable for reduced pressures. Thomas Young and Thomas dephlegmator. 38. 44 46. 76. 185 Water. 98. 42. 258. 237 ethyl alcohol. LTD. 53. 41 substances not closely related. 67. 139. 139. 72. still-head VINCF. 68. 170 compared with alcohols. 121." 165. 46. 267 aniline. 237 alcohols. 90 hydrogen chloride. 214 aliphatic acids. 50. 37. freezing of. 107. 69 boiling points of non-miscible and partially miscible liquids. ZELINSKY. .NT and DELACHANEL. composition of liquid and vapour phases. v. 113 ture. correction Volume (and heat) changes on mixing liquids. 196 and and and and and and and ZAWIDSKI. 70 Volatile liquids. 10 R. SUFFOLK. 65. 204 of fraction. 70. alcohols and methyl alcohol. 70 42 alcohols alcohols and benzene. 91. benzene and. 179 quantitative analysis by fractional distillation. 43. still-heads suitable for. "evaporator. 41. 42. 180. bulb 80. 43 Weight and composition of distillate. 11 and WURTZ. 42 of constant boiling point. 122. 38. 29. 261. to temperature range. 232. 94. 14. 42. 179. mixtures of minimum boiling point. 58. 45 WAALS. of boiling point for pressure. 46. 90. 111.C. 104. 29. 42 closely related substances. 31 composition of vapour from nonmiscible and partially miscrble liquids. 253. 106.. 222 preparation of ethyl alcohol from strong spirit. 59. 239 v. vapour pressures of mixed mixtures liquids. 113 WARFEL and NOTES. 42. 30. 102. and alcohols. ethyl alcohol and isopropyl alcohol. -U vapour pressures and boiling points of mixed liquids. 53. method and apparatus. 217. 173. 228 alcohols. 88 experimental results. 182 Volatilisation of mercury in thermometer. 69 composition of liquid and vapour phases. 17 YOUNG. benzene and. ethyl alcohol and water. composition of Russian petroleum. 248 WARREN. benzene and. composition vapour phases. dephlegmators. of liquid Vertical plain still-head. 215. Fortey. 175. 92 tertiary butyl alcohol. 50 WANKLYN. 237 isobutyl alcohols. 42. 113 experiments on fractional distilla- tion. 264 specific refractive 224. 103. 66. vapour pressure and molecular composition. ratio of. still-heads. chemical relationship of. 67. 52. 35. 91 benzene. 37. 256. 144 of distillate below middle tempera- Ramsay. v. CLAY AND SONS. 240.284 INDEX WINKELMANN. 237 and ethyl alcohol. 207. 268 Zero point of thermometers. 242 and n-propyl alcohol. 21 boiling point dependent on partial pressure of vapour.

> .


}9>pOJ SllJ} to o to UJOJJ PJBD p <D CO: 9ip 9AOUI3J ^OU 8 H OQ 01MOH01 JO A1ISH3AIM .

You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->