Determination of the Molecular Weight of Nylon

J . E. WALTZ, Technical Division, Rayon Department, A h D G1,Y B. TAYLOR, ChPniicnl D e p a r t m e n t ,

E . I . dit P n n t de .%‘emoursand Company, Tilmington, Del.

3Xethods for the determination of the molecular weight of nylons are described.
The active end groups of alcohol-soluble nylona are determined either potentio-
metrically or conductometrically. The amine end groups of alcohol-insoluble,
phenol-soluble nylons may be determined by conductometric titration in the
solvent phenol-alcohol-water. Carboxyl end groups of the latter nylons are de-
termined by dissolting the polymer in benzyl alcohol at 175’ C. and titrating
H ith potassium hydroxide, using phenolphthalein as indicator.

D E T E K M S A T I O S of the n m l r d a r weight of nylon has

proved singularly difficult because of a d w r s e solubility
characteristics, and, in these lahoratorics, only the method of end-
fatty acids only. Strong mineial acids act as solvents hut de-
grade the polymer.
111. Insoluble and infusible.
group determination has given unequivocal results. Sichols (1 ) Those in class I11 are crowlinked or gelled, are of very high or
has made osmotic pressure nieasurement,s on nylon in mixed sol- “infinite” molecular weight, and are not included in this discus-
vents for a number of samples, and has calculated number average sion. Thosein classes I and I1 are heliwed to be xtrirtly linear and
molecular weights ranging from 3000 t o 20,000. Staudinger and \vel1 defined structurally.
.Jordcr ( 7 )were unahlr t o determine the molecular weight of nylon I n order to simplify discussion of polyamides a notational sys-
osmotically because of lon- solubility in organic solvents. They tem is used in which the components are represented by numbers,
measured the intrinsic viscosity of nylon in nr-cresol and estimated the first digit referring to the number of carbon atoms in the di-
the Staudinger viscosity constant, K , from osmotic mrasurements amine and the second to the number of carbon atoms in the di-
of more soluble nylons prepared from srcondary diamines and di- acid. Thus nylon 66 (pronounced six-six) is polyhesamethylene
carboxylic acids. This procedure may be open to question (3) ndipamide.
as may that of Matthes ( 6 ) ,who determined the miilecular weight
of poly-e-caproamide by a viscometric method, using end-group ANALYTICAL METHODS
determination of 1on-t.r homologs for standardization.
I n this work methods are presented for determining t h r r i u i n h ~ i ~ Solubility Class I. The nylons ivhich fall in Class I may b t .
average molecular weight of nylon by end-group titration. prepared by copolymerization of suitable proportions of variouh
Nylon Types. Xylon can be prepared ( 2 ) by condensation diamines, dicarboxylic acids, and aminocarboxylic acids. For
polymerization (1) of diamines and dicarboxylic acids or (2) example, a polymer prepared by the copolymrrization of 40
of aminocarboxylic acids, or by copolymerization of mixturw of parts of 66 salt (hexamrthylenedianinioniuni adipatr), 30 parts
1 and 2. These have the structural arrangenimts: of 610 salt ~hexttmethylenetiianimoniunisebacatel, and 30 parts
HO (OC-R-CO-XH--K ’--SH)
,,H (1, of 6 it-aminocaprolactani) is soluble to the extent of about 15:;
in the l o w r alcohols in the presence of water, and thus is amen-
and able to determination of molecular w i g h t by titration- of tvid
groups. The following procedure may h~~ w d :
HO(OC-It-SH)nH (2)
Samples of 2.00 grams of this nylon are dissolved by refluxing
\vherc n depends upon the molecular weight, and R and R’ may in 100-ml. portions of a 72y0by weight aqueous solution of etha-
he alkylene chains containing difftsrent numbers of methylrms nol. The samples are cooled to room temperature, stirred contin-
groups. If only a n amida- uously a t slow speed, and titrated potentiomet‘rically with 0.1
tion reaction is involved, *V sodium hydroxide or hydrochloric acid solutions, added from
a microburet. p H measurements are made with a Beckman
estimation of the free Model G hydrogen-ion meter with long-style glass and calomel
amino and acidic groups 12 electrodes. Special $1 electrodes for high pH ranges were not
on the ends of the chains used.
will provide a means of
10 T h e influence of the alcohol used iri these experiments on the po-
dekrmining the number
sition of the inflection point’inthe tit’rationof 0.0001S acetic acid
average molecular weight.
(the approximate normality of the polymer solution) is shown in
T h e practicability of de-
termining the end groups - 8
G
- Figure 1. T h e inflect’ion point was found to be displaced in the
alcoh’ol solution from the true equivalence point, as shown by
by titration drpends upon
aqueous titration, by a n amount equal to the quantit’y of base
the solubility character- 6 required to bring the solvent alone to the p H of the inflection
istics of the polymer, point. I n all probability, by subtracting the blank titer, the true
which ran be made to equivalence point can be found within the limits of determining
fall into three general 4
the inflection point. I n any case, considering the reproducibility
classes: among repeat runs, the accuracy of molecular weight determina-
tion with this method is believed t o be equal to that of any othcr
I. Soluble in the lower 0.0 0.8 1.6 technique for determining the molecular weight of high polymers.
alcohols, particularly in i r i . of 0.1 A- IiaOH -1typical titration curve for this polymer is shown in Figure 2.
the presence of water. Figure 1. Titration of 0 . O O O l .Y The corrected inflection points correspond to 0.112 niilliequivalrnt
11. Soluble a t room Acetic Acid of carboxyl and 0.044 milliequivalent, of amine ends. The num-
temperature in the phe-
nols, formic acid, and 0 I n water ber of gram equivalents of end groups per gram of polymer are
0 In 72% ethanol
halogen-substituted lower c 72% ethanol alone thereforc 56 X 10-6 COOH and 22 X 10-6 XH,, from which the
448
IULY 1947
nuintitar average molecular w i g h t is 2/(78 X 10-8) or 25,600
grams per mole.
Two objections may be raised to this method of determining
molecular weight-viz., (1) the procedure does not distinguish
hetxeen terminal acidic and basic groups and any other acid or bast.
in the sample, and (2) all molecules are assunicd t o tsrniinate in
aridir or basic groups. Since a t present, no othrr method of sui-
fic.ient accuracy for determining niolecular \wight has becxn &%-
velopecl, t,hese objections cannot' be wholly met. Hoivever, thew
is good reproducibility of titrations among samples from a sin&,
lilend of polymer, and a linear relatioriship was found to exist b ( 3 -
t i w r n molecular weight determined by this method and the log-
arithm of the relative viscosity of the polymer solution.*. (Vis-
cosity determined with a pipet-style viscometer a t a conwntration
of 8.4c; in 9 0 5 aqueous formic acid. Such concentrated solu-
tions were used to take advantage of their grc,atclr sensitivity to
molpcular weight changes.) T t n ~ ~ uthus l d s w m likely that t h t w
IvtAight., .
:I r~ t rui' inolt~t*ulxr
10 0.0 1.0 2.0
M I . of 0.1 A' HC1 A' NaOH
hI1. of 0.1
Figure 2. Titration of Alcohol-Soluble
Nylon in Alcohol-Water
Solvent alone
.in interest,ingfeature of the (wrve sho\vn in Figure 2 is the ap-
pearance of two inflections on the rarhoxyl side of the curve, sug-
gesting the presence of two acids of differing s h n g t h . Although
the carboxyl groups in the polymer originate from acids of differ-
ing strengt,h, after polymerization the carboxyl groups differ only
hy whet,tier they are attached through methylene groups t o the
nitrogen or the, carbon atom of a -COXH- group. Since in-
twnally transmitted polar effects beconie negligibly after propx-
gation through two or three saturatrd carbon atoms ($) this situ-
ation n-ould not cause such a n effrct.
I t was not,ed that when polymers ivtlre prepartd containing cx-
w s s diamine the two breaks n-ertL found on the a m i m s i d e of the
curve, and therefore the hrraks appcwed to be associated wit,h
an excess of one reactant, over the othrr. On the hypothesis t h a t
salt formation was responsible for the two breaks, samples of poly-
iner were titrated for carboxyl ends with and without the addition
o f formaldehyde to react with the amine ends to form methy1tLnr.-
imino groups, as in the Sijrenson method for titrating amino
:wids. The result was thc. curves shown in Figure 3.
IVith amine groups blocked off (curve 13) only one break \vas ob-
rainrd. I t thus appears that if carboxyl groups are present in es-
('ess all of the amine ends present react, with some of the carboxyls
t o form a salt. Cpon addition of sodium hydroxide solution t h e
--C:OOH ends are titrated and this reaction is complete at, about
pH 8.5. Further addition of base suppresses ionization of the
-SH3- ion, and this reaction is complete at pII 11 and stoichio-
449
metrically measures all carboxyl in tht. polymer. Similarly, addi-
tion of hydrochloric acid solution suppressrs ionization of the
--COO- ends, a n d the amine ends are quantitatively determined
a t p H 4.5. If this is true, tht, distance t)titnec,n the two breaks on
the carboxyl side should be a fair nitlasure of the number of amine
ends. Examination of all curvt's showing two wll-defined
hreaks shoived that such \vas the case.
Solubility Class 11. The application of a similar method to
the determination of the molrcular weight of the nylons in Class
I1 deprnded upon finding a suitable solvent. Aitypical nylon
in Class I1 is 66 (polyhexamethylene adipamide), which may have
added to it some monofunctional group, such as acetic acid, as a
inolwular w i g h t stabilizer. Since no evidence of loss of acetic
acid during polymerization has been found, it is assumed that
the number of acetylated amine rnds in the polymer can be esti-
mated satisfactorily by the amount of acid addc,d before poly-
merization. Thus only terminal amino a n d carboxyl groups
need be determined by titration.
Several solvents for 66 nylon arr acidic in reaction and might
permit, a sudden increase of hydrogen-ion roncrntration a t the
equivalence point of the amine end-group titration, but the most
satisfactory solvent which is not highly acidic is phenol. It was
desired, although not rrquirrd, to dilutc. the solvent phenol con-
siderably before titration, SO as t o provide less chance for am-
'
biguity in t,he interpretation of the titration c u r v c s T o this end,
it was found that if 2 grams of polymer ~ v e r etiissolvrd in 50 ml. of
885, phenol, the polymer was retained in a single-phase solution
upon additiop of 25 ml. of 957( ethanol and 25 ml. of water. Com-
mercially available phenol contains a n acid impurity which can
be removed by treating with potassium cxrhonatt. anti sutisequent
distillation.
Curve B in Figure 4 shows the titration of 100 nil. of a purified
phenol-alcohol-water solution with 0.1 .Y hydrochloric acid solu-
t,ion. Curve A in Figure 4 shows the same titration when the
solution contains 2.00 grams of a 66 nylon polymcr. T h e equiva-
li
00 10 0.0 1 IJ 2.0
M I . of 0 1 S S a O H
Figure 3. Titration of Carboxyl Groups of an
Alcohol-Soluble Nylon
Formaldehyde added to sample B
lence point, C, is found by subtractingfroni curve h the amount of
titrant required to bring the solvent alone t o the p H of curve .1at
the inflection point. I t was demonstrated hy titration of a knoa-n
standard (66 salt) that this method accurately determined amine
end groups, but for routine analysis a some^ hat simpler method
ryas desired. Preliminary mrawrements of the conductance of
the solution during titration showed a pronounced increase in
conductance after the equivalence point was passed, and the fol-
lowing procedure was therefore adopted:
 450 VOLUME 19, NO. 7
h 2.00-gram sample of dry nylon 66 is disolved in 50 ml. of
purified phenol by shaking, 25 ml. of 95YC ethanol and 25 ml. of
water are added, and, with constant slow stirring, the solution is
titrated conductometrically with 0.1 S hydrochloric acid solution
added from a microburet. The conductance cell consists of two
platinum electrodes about 3 sq. cm. in area and 1 cm. apart im-
mer:ed in the solution. Resistance is measured by a Leeds &
Soitliruii re i-tance bridge using 110-volt, 60-cycle alteriiating
curre..t and ail alternating current galvanometer as null-poi it in-
strume..t. The re istance is measured after successive 0.2-ml.
additions of the acid, until t,lie equivaleiice point is passed 1)y 0.G
to 0.8 ml. The reciprocal of resistance is plotted versus the milli-
liters of titrant, and the points lie o n two straight lines, the inter-
section of which estab!ishes the equivalence point.
some typical titration curves are s h o m in Figure 5. Curre A
is tlie titration of 0.05 millimole of 66 salt, curve 13 the titration of
2.00 grams of an alcohol-soluble polymer shown by potentiomet-
ric titration t o contain 36 x 10-6 gram equivalents of amine end
groups per gram of polymer, :tnd curve C shows a typical titratioii
of 66 nylon polymer. Tile agreement of curves .i and B n-ith t!ie 0 1 2 0 1 2 0 1 2
known co:iceiltr:ition of amine is escc1lc:it. 11' of 0 1 S HC1
Figure 5 . Conductometric Titration for Amine Dissol\ed
The theory behind the titration is simple and sn:iighlforivard. i n Phenol- ilcohol-B'ater
.ipparently the niobility of t,he polymer molecules is so lon. that
they need not be considered. Until the equivalence point is
reached the addition of hydrochloric acid in solution adds the
modwately mobile chloride ion t,o the solution, and tllc conduct- tained tiy heating alcohol alone for the same length of time as the
aric~,increases slightly in a linear fashion. Xftcr the cquivnlence solution, is roquired for all titrations. The hot solvent is a satis-
, point has been passed, addit,ion of hydrochloric :acid prouitlcn factory titration mcdium for carboxyl end groups, as is shown by
both the chloride ion :tnd the highly mobile hydrogen ion t o the! the data in Tablc' I . T h e titrant was 0.1 S potassium hydroxide
solution, so the conductance increases sharply in a lincvir fashion. solution in 10TCniethanol-90yc benzyl alcohol and phenolphthal-
The int,ersection of the t,wo straight lines, arising from :I plot of ein \\-as uqcd as indicator.
these points, is the equivalence point of the reaction. -
I n direct contrast to the potentiometric method in which much
weight is placed on points near the equivalence point, in thib ti-
tration such points are given no weight, ( 5 ) . I n general, seven or Table I. Titration of icids in Benzyl Alcohol at 17.5' C.
eight measurements suffice to determine t,hetwo lines. l l i l l i e q u i \ alent of
Carboxyl
T h e conductometric method was so simple and accurate t h a t it \\eight, Found
l\laterial Grams Calculated (corr )
was subsequently applied with excellent results t o the titration of
both amine and carboxyl groups cf alcohol-drtble polymers. Adipic acid 0.0182 0,250 0 249
66 salt 0 0365 0 2i8 0 274
Phenol was, of course, not satisfactor a solvent for the de- 66 s a l t 0 0830 0 683 0.635
Aniino-stearic acid 0 060 0,200 0 198
terminat,iori of carboxyl ends in 66 nylon for, after t h r tquivalence Alcohol-soluble nylon 3 00 0.159u 0.160
point was passed, reaction of the h a w v i t h phthnol pic-vented any 0 Frorii t i t r a t i o n in ethaiiul a-titer
sudden increase in hy-
drosyl-ion conccntra-
tion. Efforts to detect
the appearance of phrno- Tlic, ugrccmc~ntof obscrved va1uc.s with calculated i: good. For
late ions by conduct- these tests and in subsequent work benzyl alcohol was kept at
ance nieasuremcnts about 175' C. by keeping the solvent beaker in contact with the
were likewise unfruit- v:tpor of boiling p-cymene.
ful. S o other room It, is interesting to note t,hat in this system also titrations may
temperature solvents 4 be made conductometrically. The conductance is considerably
were found which per- L lo\wr than for the phenol-alcohol-water solvent, and the break
mitted unambiguous 3 not so sharp, but good agreement of end point with phenolphthal-
titration of carboxyl ein color change was found.
groups. Thesearch was
then directed toward the 2 - SUMMARY
higher alcohols, some '
Met,hods are described for determining the number average
of which were known to
molecular weight of alcohol-soluble nylons and alcohol-insoluble,
dissolve nylon a t high phenol-soluble nylons.
temperature. 0.0 0.b l.(i
MI. of 0.1 N HC1
The most promising LITERATURE CITED
solvent uncovered was Figure 4. Titration of 66 Nylon
benzyl alcohol, which in Phenol-Alcohol-Water (1) Alexander, J., "Colloid Chemistry," Yol. 6, Chap. 60 (by J. B.
Nichols), New York, Keinhold Publishing Corp., 1946.
dissolves 66 nylon to B. Titration of s o l v e n t alone (2) Carothers, W. H. ( t o E. I. du Pont de Nemours and Co., Inc.),
the extent, of about U.S.Patents 2,071,250 and 2,071,253 (Feb. 16, 1937).
10% a t 175" C. Al- (3) Flory, P. J., J . Am. Chenl. Soc., 65, 372 (1943).
though benzyl alcohol distills readily a t its boiling point with- (4) Gane, R . , and Ingold, C . K., J . Chem. SOC..1931,2153.
(51 MacInnrs, D A , . "Principles of Electrochemistry," p. 384, New
out decomposition, no method has bcen found for preparing the York, Keinhold Publibhing Corp., 1039.
completely neutral alcohol. Repcated distillations have given (6) Matthes, h., J . p r a k t . Chem., 162, 245 (1943).
slightly acid products. I n addition the alcohol develops acidity on (7) Staudinger, H.. and Jiirder. H . , Kunsfseide u. Zellwolle, 24, 88
heating a t 175' C . at the rate of 0.005 t o 0.02 milliequivalent' per (1942). EERBISHURW3M W,
hour per 100 ml. This m w n i that a corrcction, which is ob- COSTRIBCTION 211;