V O L U M E 23, NO. 3, M A R C H 1 9 5 1 In the case of complete unknowns, the original sample and/or the aqueous extract from the Kappelmeier precipitate must, of course, be examined qualitatively before the method can be applied with confidence. In such cases the complete ultraviolet spectrum of the extract, together with various chemical tests for interfering materials likely to be present, will indicate whether the method is directly applicable or whether suitable modifications should be sought. The authors have found infrared absorption spectroscopic examination of the original sample and the precipitated salts from saponification (as a S u j o l mull) to be very useful in this connection. Stafford et al. ( 5 ) have suggested thtk use of the infrared spectra of the dibenzyl amides of the dibasic acids as a means of determining what dibasic acids are present. Although the method was primarily developed for use in the
445
analysis of surface coatings, it should have useful applications i n other fields. LITERATURE CITED
(1) Am. SOC.Testing Materials, “A.S.T.M. Standards,” Part 4, p. 237, Designation D 563 (1949). (2) Bradley, J. J., Oficial Digest Federation Paint & Varnish Production Clubs, No. 266, 162 (1947). (3) Kappelmeier, C. P. A.. Farben-Ztg., 40, 1141 (1935); 41, 161 (1946); 42, 561 (1937); Paint Oil Chem. Rm.,6. 10 (1937). (4) Marvel, C. S.. and Rands, R. D., Jr., J . Am. Chem. Soc.. 72, 2642 (1950).
(5) Stafford, H. IT., Francel, R. J., and Shay, J. F., 21, 1454 (1949). (6) Swann, M. H., Ihid., 21, 1448 (1949).
;INAL. CHEM.,
RECEIVED September 5, 1950. Presented before the Division of Analytical Chemistry at the 118th Meeting of the AMERICANCHEMICAL SOCIETY. Chicago, Ill.
Residual Monomer in Polystyrene Spectrophotometric Analysis J. E. NEWELL, United States Rubber Co., Passaic, .V. J . A rapid method for the determination of monomeric styrene in polystyrene is described. Styrene absorbs ultraviolet radiation of wave lengths from 250 to 260 mp, 40 to 100 times as intensely as polystyrene. Determination of small amounts of monomer in polymer, 0.1 to 2.09’0, with precision and accurac? suitable for routine analysis, has been accomplished by absorption spectrophotometry. The method is most accurate in the absence of impurities other than styrene monomer. The presence of significant amounts of interfering substances is made evident by poor agreement of calculations for monomer from absorption measurements at three wave lengths.
251, 282, and 291 m p , interference by other materials is likely to be much less a t 251 m y than a t the other wave lengths. In acldition, the low concentration of polymer solution necessary for absorption measurements a t 251 m p is likely to cause less error from light scattering than the more highly concentrated solutions required for absorption measurements a t longer wave lengths. This paper presents a spectrophotometric method of analysis of styrene in polystj-rene by absorption measurements in the 251 mp region. .Is in the methods of Owens ( 7 ) , and McCovern et al. (j),multiple wave length determinations show the presence of interfering substances.
T
HE pronounced effect of residual monomer on the physical properties and molding characteristics of commercial polystyrene makes the determinat,ion of residual monomer in the polymer important. On account of the tenacious retention of styrene by polystyrene, especially by large particles, the usual determination of total volat,ile matter by treatment with heat and vacuum often gives low results. In a single determination, the frozen benzene technique of Lewis and Mayo ( 4 ) removed the styrene quantitatively but left in its place a few tenths of a per cent of benzene. Because of the difficulty in removing the styrene from its polymer in u rt.ate suitable for titration or precipitation, the methods of Kolthoff and Bovey (S),or Bond ( I ) , are not easily applied t o tht. determination of monomer in polymer. Styrene absorbs ultraviolet radiation of wave lengths of 282 and 291 mp, whereas polystyrene has only general absorption in this region of the spectrum. Owens ( 7 ) made use of this difference to analyze monomer in partially polymerized styrene, using a medium quartz spectrograph. McGovern, Grim, and Tfsach (6) adapted this method to a spectrophotometer and invcsstigated variables affecting the analysis. .it a wave length of 251 mp, styrene absorbs ultraviolet. ratli,ition 15 times as intensely as it does a t 282 mp, and 25 times as intensely as a t 291 mp. At this wave length it also has 100 t iriitbs the absorption of polystyrene. Therefore conditions are I‘:ivorable for the determination of small amounts of styrene in polystyrene by absorption measurements in this spectral region. Owing to the relative intensities with which styrene absorbs a t
WAVELENGTH. MU
Figure 1. Ultraviolet Absorption Spectrum of Commercial Polystyrene Curve 1. Before purification Curve 2. After precipitation i n alcohol Curve 3. Difference between curves 1 and 2
The absorption near 250 m p of R niisture of styrene and polystyrene can be resolved by simple mathematics into the absorption of the styrene and that of polystyrene. However, t o give significant results, two conditions must be met. The optical interference of impurities other than styrene must be low enough
ANALYTICAL CHEMISTRY
446
solvent, the optical density, Table I.
Preparation of Polystyrene Samples
Pol ymerization Conditions A Crystallized Heated in vacuo B Distilled Standing a t room temperature C Fractionally distilled Ultraviolet irradiated D Commercial, not purified Emulsion, heat E Commercial Commercial polymer From intrinsic viscosity measurements ( 2 ) .
Sample
llonomer Preparation
OD.of the solution will be
Catalyst None Xone
0.3% biacetyl 0.3% sodium persulfate S o t known
Molecular Weighta 950,000 590,000 540,000 370,000 170,000
multiple absorption readings.
=
100-x ( r W E ,) +
X 100 I.I'Em
(1)
from which
x = 100 (-)EEm --E,Ep
-
that no significant error results. Secondly, the absorption spectrum of polystyrene must be constant-that is, independent of molecular weight, conditions of polymerization, catalyst, or other variables. I n Figure 1 the absorption spectrum of a solution of commercial polystyrene is shown to be a composite of the spectra of pure polystyrene and styrene. Curve 1is the absorption spectrum of a commercial polystyrene solution in chloroform. Curve 2 is the spectrum of an equal concentration of the same polymer, purified by precipitation from chloroform solution with alcohol. Curve 3 is the difference between curves 1 and 2. Within experimental error, curve 3 is the absorption spectrum of the impurities removed in the purification process, and is identical to the absorption spectrum for styrene which is shown in Figure 2. Although there are, no doubt, small quantities of dimer, trimer, ethylbenzene, and other impurities in the polymer, their concentration and absorption are so loiv that the major ultravioletabsorbing impurity is monomer. This is in accord with the reports of other workers ( 5 , 6). Identification of the styrene as the major impurity has been accomplished for several samples of commercial polystyrene from each of three major manufacturers, even for samples with as little as 0.1% residual monomer. If the complete absorption spectrum of the sample solution is not measured, and the calculation of per cent monomer is made from absorbance measurements taken a t a single wave length, all radiation-absorbing impurities are expressed as styrene-an
OD
(2)
where E, Ep and E,fi are the specific extinction coefficients of the sample, monomer-free polymer, and monomer, respectively. EXPERIMEIVTA L
In the investigation to find whether the absorption spectrum of polystyrene was dependent upon any variables of the polymerization reaction, and in the measurement of the coefficient, E,, five samples of polystyrene were used. The methods of preparation of these polymers are summarized in Table I. Each of these polymers was made monomer-free by repeated solution in chloroform and precipitation as a very fine filament by extrusion through a 0.4-mm. capillary into alcohol. In order to avoid loss of the low-molecular-weight fraction of the polymer, the volume of alcohol was always a t least ten times that of the chloroform solution. Each polymer was finally vacuum dried for 16 hours. ilrcurately weighed samples of approximately 0.12 gram of each polymer \\ere dissolved in C.P. chloroform and the solutions were diluted to 250 ml. in volumetric flasks. The optical densities of these solutions a t 250, 255, and 260 mp were measured. A Beckman spectrophotometer was used, with 1.00-cm. silica cells and a slit width of 0.9 mm. at 250 mp and 0.8 mm. a t 255 and 260 mp. The specific extinction coefficients were calculated in the usual way; concentrations were in grams per liter. By calibration with a mercury vapor lamp, the wave-length scale of the spectrophotometer was known to be accurate to * 0.1 mp.
Additional
250
THEORY 250
If a sample of polystyrene weighing IY grams containing X% residual monomer is dissolved in 1 liter of optically transparent
260
270
280
290
WAVELENGTH MU
~i~~~~ 2. Ultraviolet Absorption Spectrum of Styrene Monomer
260
270
280
290
WAVELENGTH M U
Figure 3. Ultraviolet Absorption Spectrum of Monomer-Free Polystyrene
V O L U M E 2 3 , NO. 3, M A R C H 1 9 5 1 Table 11.
447
Summary of Extinction Coefficients for Six Polystyrenes Wave Length, M p 250 1.38 1.29 1.34 0.030
Highest value Lowest value Average value Standard deviation
260 2.16 2.08 2.12 0,024
255
1.84 1.77 1.80 0.024
Table 111. Specific Extinction Coefficients of Polystyrene and Styrene Wave Length,
Polystyrene,
EP
Styrene,
Jfp
250.0 265.0 260.0
1.34 1.80 2.12
136 116 87
Table IV.
Per Cent Monomer in Synthetic Mixture of Purified Polystyrene and Stvrene - .
Actual Styrene, 70 0.00 0.14 0.28 Q.56
Table
Em
Wave Length, JIp 250.0 255.0 260.0 0.02 0.02 0.02 0.15 0.16 0.17 0.30 0.29 0.29 0.56 0.57 0.56
Average 0.02 0.16 0.29 0.56 Av.
Error f0.02 +0.02
+0.01 0.00 +O.Ol
\-. Per Cent Rlonomer in Various Polystyrenes
Sample Experimental
-4-1 .4-2 B-1 B-2 B-3 B-4 c-1 c-2
250.0 0.31 0.75 1.04 1.18 0.80 0.79 0.42 0.45 0.16 0.19 0.09 0.20 0.22 1.05 1.58
Wave Length, M p 255.0 260.0 0.34 0.31 0.80 0.77 1.00 0.94 1.13 1.07 0.80 0.79 0.81 0.77 0.42 0.40 0.44 0.42 0.12 0.09 0.15 0.11 0.10 0.18 0.19 1.04 1.57
0.10 0.15 0.17 1.00 1.54
Average 0.32 0.77 0.99 1.13 0.80 0.79 0.41 0.44 0.13 0.15 0.10 0.18 0.19 1.03 1.66
minations of the extinction coefficients a t 250, 255, and 260 mp is given in Table 11. The standard deviations of the estinction coefficients, when expressed as equivalent mmomer content, are 0.02% a t 250 and 255 mp, and 0.03% a t 260 mp. The absorption spectrum of monomer-free polystyrene is shown in Figure 3. The sample of styrene used in the determination of the specific extinction coefficient, E m , was distilled at 50" C. under 20-mm. vacuum. Only the mid-fraction was used. The absorption spectrum of styrene in chloroform solution is given in Figure 2. The specific extinction coefficients of polystyrene and styrene from 250 t o 260 mp are given in Table 111. When the numerical values for E , and Em are substituted in Equation 2, the following equations result: At 250 mp, % styrene = 0.743 (E21o - 1.34) (3) - 1.80) At 255 mp, % styrene = 0.876 (E256 (4) At 260 mp, yo styrene = 1.178 (E2bO - 2.12) (5) The purification process for polystyrene removes quantitatively the monomer and other alcohol-soluble impurities. The spectrophotometric analysis of mixtures of purified polymer and known amounts of monomer does not prove the accuracy of this method in the presence of interfering impurities. However, it does show the precision to be expected, and the accuracy in the absence of interferences. A weighed sample of a purified polymer was dissolved in chloroform, and aliquots were mixed with varying amounts of a 0.00751-gram-per-liter styrene solution. The per cent styrene was calculated on total solute. The results of spectrophotometric analysis of the solutions and calculation of results using Equations 3 to 5 are shown in Table IT.
Several experimental polymers and eight commercial polymers have been analyzed with the spectrophotomet'er. The results of some of these analyses are shown in Table V. Sample weights were approximately 0.12 gram, and solution volume was 250 ml. The commercial samples were from three manufacturers (letters A, B, and C in the sample designation), and were in the form of beads, clear cubes, or opaque chips. DISCUSSION OF RESULTS
The results in Table IV indicate that in the absence of interfering impurities the absolute accuracy of the method is approximately * 0.03%, and the absolute precision is about * 0.02%. The analytical results for some commercial samples are less accurate and precise. I n Table V, the results for the duplicate determinations of monomer in samples A-2 and C-2 are in poor agreement. I n these two cases the polymer sample was in the form of coarse chips or cubes, and the variation in monomer in the particles was great enough that the 0.12-gram sample did not represent the whole material under examination. S o doubt better agreement of duplicate determinations could have been achieved by dissolving much larger amounts of polymer, and diluting aliquots to suitable concentrations. On the other hand, by using smaller sample weights (minimum 1 mg.) and correspondingly small solution volumes (minimum 3 ml.), it is possible to determine the styrene content in individual polymer chips when the distribut,ion of monomer among the different particles is of interest. The results for monomer content for the experimental polymer, and for A-1, B-1, B-2, and B-4 agree EO well at the threewave lengths that it is likely that the values in the average column are accurate. On the other hand, the results for A-2, B-3, C-1, and C-2 indicate the presence of interfering material. In these cases, the percentages of monomer are high. These results may still be significant when it is considered that the lowest per cent shown for each represents its maximum possible monomer content-for example, samples B-3 and C-1 have monomer contents of 0.11% or less, and 0.17% or less, respectively. This information may be sufficient for some purposes in quality control. SUMiMARY
The spectrophotoniet'ric method of analysis of polystyrene for residual monomer has several advantages, and some limitations. No treatment of sample other than solution is required. The method is rapid, and has precision and accuracy suitable for quality control. However, the method is limited to polymers which have no pigments or dyes. Agreement of results a t multiple wave lengths shows when interfering materials are present. I n the presence of int,erference, only maximum monomer content can be determined. ACKNOWLEDGMENT
The author is indebted to several members of the General Laboratories staff for suggestions and assistance, and particularly to John Burke, who determined several of the absorption spectra. LITERATURE CITED
(1) B o n d , G. R., A s a ~ CHEY., . 19, 390-2 (1947). ( 2 ) Eaart, R. H., and Tingey, H. C., Abstracts of 111th Meeting, il~ CHEM. . Soc., Division of Rubber Chemistry, Atlantic City. N. J., April 1947. (3) Kolthoff, I. &I., and Bovey, F. A , ; ~ X A L .CHEY.,19, 498 (1947). ENG.CHEY.,ANAL.ED.,17, (4) Lewis, F. M., and h I a y o , F. R., IXD. 134 (194,5). --, ( 5 ) hIcGovern, J. J., Grim, J. >I., a n d Teach, Wr.C., A N ~ LCHEM., . 20, 312-4 (1948). (6) Meehan, E. J., J . PoZu/mer Sei.. 1, 175-82 11946): reminted i n Rubber Chem. and Technol., 19, 1077 (1946). (7) Ow'2n8, J. S., IND.ENG.CHEY.,AXAL.ED.,11, 643 (1939). ~~
~
\ -
RECEIVED April 21, 1950. Presented before the Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., February 1950. Contribution 102, General Laboratories, United States Rubber C o .