Anal. Chem. 1996, 68, 425-430
Compositional and Structural Studies of Vinylidene Chloride/Vinyl Chloride Copolymers by Pyrolysis Gas Chromatography Frank Cheng-Yu Wang* and Patrick B. Smith
Analytical Sciences Laboratory, Michigan Division, The Dow Chemical Company, Midland, Michigan 48667
A pyrolysis gas chromatography approach has been developed to study the composition and structure of vinylidene chloride/vinyl chloride copolymers. The composition and number average sequence length, which reflects the monomer arrangement in the polymer, were calculated using formulas that incorporate the pure trimer peak intensities and hybrid trimer peak intensities. Because of the reactivity difference between vinyl chloride and vinylidene chloride monomers, the structure of the polymer has been further investigated on the basis of the percentage of grouped monomers (i.e., the number average sequence length for vinyl chloride and vinylidene chloride repeat units). For the vinylidene chloride/vinyl chloride copolymer examined in this study, the composition and number average sequence length elucidated from the pyrolysis gas chromatography study are compared to the product composition specification and/or the composition measured by 1H-NMR. The composition and structure of a polymeric system usually can be determined by spectroscopic methods, such as infrared (IR)1,2 or nuclear magnetic resonance (NMR) spectroscopy,3,4 or by a structural degradation method, such as pyrolysis gas chromatography (Py-GC).5 These methods have existed for more than 20 years, and each has its own strengths and weakness in regard to structure elucidation. Since many different polymeric systems have been studied, the choice of which method to use depends mainly on which specific polymeric system is being studied and what kind of information needs to be generated. The vinyl chloride/vinylidene chloride copolymer system has been in use since the 1930s. The most valuable property of poly(vinylidene chloride/vinyl chloride) copolymer is its low permeability to a wide range of gases and vapors. This copolymer also has excellent oil, gas, chemical, and ignition resistance, as well as toughness, flexibility, and heat-sealability. The major applications of this copolymer are in rigid and flexible packaging and coatings to provide a barrier for gas and moisture and for its chemical resistance, inertness, toughness, and clarity. There are (1) Crompton, T. R. Analysis of Polymers: An Introduction; Pergamon Press: New York, 1989; pp 163-183. (2) Keonig, J. L. Spectroscopy of Polymers; American Chemical Society: Washington, DC, 1992; pp 131-134. (3) Randall,. J. C. Polymer Sequence Determination: Carbon-13 NMR Method; Academic Press: New York, 1977; pp 41-69. (4) Davis, E. A.; Ward, I. M. Nuclear Magnetic Resonance in Solid Polymers; Cambridge University Press: Cambridge, UK, 1993; pp 229-31. (5) Irwin, W. J. Analytical Pyrolysis: A Comprehensive Guide; Marcel Dekker, Inc.: 1982; pp 305-308. 0003-2700/96/0368-0425$12.00/0
© 1996 American Chemical Society
several literature reviews about the vinyl chloride/vinylidene chloride copolymer system.6-8 Mechanical and physical properties are an important aspect of the vinyl chloride/vinylidene chloride copolymer system and can be governed not only by the composition but also by the structure of the polymeric molecules and the existence of homopolymer domains.9 Thus, an understanding of the relationship between the copolymer film properties, composition, and structure becomes critically important. The polymer composition and structure have a direct effect on such properties as modulus, glass transition temperature, and chemical resistance. By varying the monomer concentration, monomer feed rates, initiator concentration, and chain transfer agent concentration, such properties as monomer sequence, molecular weight distribution, and particle size distribution can be affected. The ability to calculate the composition and monomer sequence in the final product will be a powerful tool to determine to what degree an experimental parameter affects the polymer structure. The statistical method for calculating the number average sequence length from the trimer distributions has been known for a long time.2 The method is widely used in NMR spectroscopic analysis, because NMR can supply triad molar fractions. In the vinylidene chloride/vinyl chloride copolymer system, if C represents the vinyl chloride monomer and D represents the vinylidene monomer, the number average sequence lengths of both monomers are expressed as
n ˜C )
n ˜D )
NCCC + NCCD+DCC + NDCD (1/2)NCCD+DCC + NDCD NDDD + NCDD+DDC + NCDC (1/2)NCDD+DDC + NCDC
(1)
(2)
where n ˜C and n ˜D are the number average sequence lengths of monomers C and D. NCCC, NCCD+DCC, NDCD, NCDC, NCDD+DDC, and NDDD are the experimentally derived triad molar fractions or numbers of molecules. From the formulas above, if all six triad molar fractions or numbers of molecules can be generated, the number average sequence lengths of monomers C and D can be calculated. (6) Wessling,. R. A. Polyvinylidene Chloride; Gorden and Breach Science Publishers: New York, 1977; Chapter 1. (7) Schildknecht, C. E. Vinyl and Related Polymers; John Wiley & Sons, Inc.: New York, 1952; Chapter 8. (8) Wessling, R. A.; Gibbs, D. S.; DeLassus, P. T.; Howell,. B. A. Encyclopedia of Polymer Science and Technology; John Wiley & Sons, Inc.: New York, 1989; Vol. 17, pp 492-531. (9) Kockott, D. Kolloid-Z. Z. Polym. 1965, 206, 122.
Analytical Chemistry, Vol. 68, No. 3, February 1, 1996 425
Pyrolysis followed by gas chromatographic separation is a mechanism that utilizes thermal energy to break down a polymeric chain to monomers and oligomers and to separate those units for quantitation. Monomers are commonly used to obtain the composition information, and oligomers may be used to obtain structural information. Because of the high temperature limits of common silicone capillary columns, oligomers up to four monomer units of the system studied here can be reliably separated and detected. The major mechanism of producing oligomers with pyrolysis can be attributed to thermal degradation. The intensity of the various oligomer peaks in a pyrolysis gas chromatogram will reflect the monomeric sequence and polymer structure when the formation of pyrolysis products is proportional to their existence in the copolymer. The literature10-18 discusses the determination of the structural sequence from different oligomer peak intensities. Determining composition by pyrolysis normally depends on the monomer production after polymer chain scission. However, vinyl chloride and vinylidene chloride are both gaseous monomers and are not well retained on a capillary gas chromatography column under normal conditions. In addition, other gases that result from pyrolysis of this system, such as hydrogen chloride, butadiene, etc., interfere with the monomer detection. As a result, a composition analysis utilizing the monomers of vinyl chloride/ vinylidene chloride copolymer through pyrolysis is a poor approach. In this study, the trimer peak intensities are used to achieve the composition quantitative analysis as well as the number average sequence determination for the vinylidene chloride/vinyl chloride copolymer system. The unique phenomenon in the pyrolysis of vinylidene chloride/ vinyl chloride copolymer is the trimer formation. Under pyrolysis conditions, the polymer will directly undergo the thermal dehydrochlorination to form a conjugated polyene.19 The polymer will then unzip, followed by a radical cyclization to form benzene, chlorobenzene, dichlorobenzene, and trichlorobenzene. The mechanism can be expressed as
(1) dehydrochlorination s ( CH2sCHCls )fs ( CH d CHs ) s ( CH2sCCl2s )fs ( CH d CCls ) (2) unzipping CCC CCD, DCC, CDC polymer chain f DDC, CDD, DCD DDD (3) cyclization CCC benzene CCD, DCC CDC chlorobenzene f DDC, CDD, DCD dichlorobenzene DDD trichlorobenzene
Because these chlorinated aromatics are so stable, the trimer formation pathway is the major pyrolysis pathway for the chloride/ vinylidene chloride copolymer. 426
Analytical Chemistry, Vol. 68, No. 3, February 1, 1996
Several NMR spectroscopy studies have dealt with determining the polymeric structure of vinylidene chloride/vinyl chloride copolymer.14-18 These studies are based on the same theory of determining the number average sequence length from peak triad intensities. However, the NMR structure determination suffers from poor precision and accuracy due to interferences, especially from common additives. Pyrolysis gas chromatography of vinylidene chloride/vinyl chloride copolymer systems results in very clear and distinguishable trimer peaks. With three assumptions which are discussed later, the composition in weight percent can be calculated. The number average sequence lengths can then be elucidated on the basis of eqs 1 and 2. These values were compared with product composition specification and 1H-NMR results, when available. EXPERIMENTAL SECTION Sample Preparation. (i) Vinylidene Chloride/Vinyl Chloride Copolymer. All vinylidene chloride/vinyl chloride copolymers were synthesized according to published textbook procedures,20 except for one (sample F, 5%D/95%C), which was purchased from Scientific Polymer Products Inc. (Ontario, NY). All the samples are in the powder form. (ii) Py-GC and Py-GC/MS Conditions. Samples of vinylidene chloride/vinyl chloride copolymer powders (∼2.5mg) were put into a quartz tube. The quartz tube was equilibrated for 5 min in a 180 °C interface connected to the injection port of an HP5890 gas chromatograph equipped with a flame ionization detector (FID). Samples were pyrolyzed (CDS 120 Pyroprobe Pt coil) at a set temperature of 700 °C. The coil was heated to 700 °C at 20 °C/ms and held at 700 °C for 20 s. The pyrolysis products were split in the 250 °C injection port, with 10 psi head pressure, 10:1 split ratio, separated on a fused-silica capillary column (J & W DB-5, 30 m × 0.25 mm i.d., 1.0 µm film) using a linear temperature program (40 °C/4 min, 20 °C/min to 120 °C/ 10 min, then 20 °C/min ramp to 300 °C/23 min), and detected using a FID. In the GC/MS system, the FID was replaced with a VG Trio-1 mass spectrometer. The output from the GC was transferred through a transfer line (280 °C) to the ion source of the mass spectrometer. An electron ionization mass spectrum was obtained every second over the mass range of 29-500 amu. The Py-GC is used for peak intensities measurement, and the Py-GC/MS is used for component identification in this study. The 700 °C pyrolysis temperature was chosen based on the basis of the optimum yield of trimers for both vinyl chloride and vinylidene chloride polymers. (10) Eonmoto, S. J. Polym. Sci. Part A, 1969, 7, 255. (11) Pichot, C.; Guillot, J.; Guyot, A. J. Macromol. Sci., Chem. 1971, A5 (4), 753. (12) Guyot, A.; Pichot, C.; Guillot, J. J. Macromol. Sci., Chem. 1972, 6 (8), 1681. (13) Liu, L.; Weng, Z.; Huang, Z.; Pan, Z. Hecheng Shuzhi Ji Suliao 1993, 10 (2), 19. (14) Chujo, R.; Satoh, S.; Nagai, E. J. Polym. Sci. Part A 1964, 2, 895. (15) McClanahan, J. L.; and Previtera, S. A.; J. Polym. Sci. Part A, 1965, 3, 3919. (16) Enomoto, S.; Satoh, S. Kolloid-Z. Z. Polym. 1967, 219 (1), 12. (17) Yamashita, Y.; Ito, K.; Ishii, H.; Hoshino, S.; Kai, M. Macromolecules 1968, 1 (6), 529. (18) Carman, C. J. Molecular Structure of Vinyl Chloride-Vinylidene Chloride Copolymers by Carbon-13 NMR; ACS Symposium Series 142; American Chemical Society: Washington, DC, 1980; pp 81-91. (19) Wessling, R. A. Polyvinylidene Chloride; Gorden and Breach Science Publishers: New York, 1977; p 139. (20) Wessling. R. A. Polyvinylidene Chloride; Gorden and Breach Science Publishers: New York, 1977; p 31.
Table 1. Average and Relative Standard Deviation of a Number Average Sequence Length Calculation Based on Ten Consecutive Runs of a Vinylidene Chloride/Vinyl Chloride Copolymer normalized/corrected peak intensity sample no.
benzene
chlorobenzene
dichlorobenzene
molar %
trichlorobenzene
VC(C)
wt %
VDC(D)
VC(C)
VDC(D)
1 2 3 4 5 6 7 8 9 10
0.057 0.057 0.057 0.058 0.057 0.058 0.057 0.057 0.057 0.056
0.136 0.135 0.136 0.135 0.134 0.135 0.135 0.134 0.134 0.134
0.281 0.277 0.281 0.275 0.276 0.279 0.279 0.277 0.279 0.276
0.526 0.531 0.526 0.532 0.533 0.528 0.529 0.532 0.531 0.534
24.1 23.9 24.1 24.0 23.8 24.1 24.0 23.9 23.9 23.8
75.9 76.1 75.9 76.0 76.2 75.9 76.0 76.1 76.1 76.2
17.0 16.8 17.0 16.9 16.8 17.0 16.9 16.8 16.8 16.7
83.0 83.2 83.0 83.1 83.2 83.0 83.1 83.2 83.2 83.3
av SD %SD
0.057 5.4 ×10-4 0.95
0.135 6.9 ×10-4 0.51
0.278 2.1 ×10-3 0.74
0.530 2.7 ×10-3 0.50
24.0 1.3 ×10-3 0.53
76.0 1.3 ×10-3 0.17
16.9 9.7 ×10-4 0.57
83.1 9.7 ×10-4 0.12
Table 2. Time-Dependent Study of Vinylidene Chloride/Vinyl Chloride Copolymer Composition under 180 °C Pyroprobe Interface Temperature normalized/corrected peak intensity
molar %
wt %
time (h)
benzene
chlorobenzene
dichlorobenzene
trichlorobenzene
VC(C)
VDC(D)
VC(C)
VDC(D)
0 1 2 3 4 6 8 10 12 14 16
0.061 0.060 0.059 0.057 0.058 0.058 0.058 0.060 0.061 0.062 0.062
0.137 0.138 0.137 0.137 0.136 0.139 0.140 0.143 0.146 0.150 0.151
0.275 0.277 0.280 0.283 0.285 0.300 0.309 0.320 0.326 0.332 0.338
0.527 0.526 0.524 0.523 0.520 0.502 0.493 0.477 0.468 0.455 0.448
24.4 24.4 24.3 24.3 24.4 25.1 25.4 26.2 26.6 27.3 27.6
75.6 75.6 75.7 75.7 75.6 74.9 74.6 73.8 73.4 72.7 72.4
17.2 17.2 17.2 17.1 17.2 17.8 18.0 18.6 19.0 19.5 19.7
82.8 82.8 82.8 82.9 82.8 82.2 82.0 81.4 81.0 80.5 80.3
Test of Reproducibility. The 700 °C pyrolysis temperature was chosen on the basis of the higher yield of trimer for both vinylidene chloride and vinyl chloride. The reproducibility of pyrolysis data is always a concern when applying the technique to any kind of quantitative study. Table 1 shows normalized trimer peak intensities and the averages and relative standard deviations of a composition calculation based on 10 consecutive runs of a vinylidene chloride/vinyl chloride copolymer. All terms show a relative standard deviation below 1%, which demonstrates the reliability of the pyrolysis method applied to the analysis of vinylidene chloride/vinyl chloride copolymers. Test of Sample Stability in the Pyroprobe before Pyrolysis. Pyrolysis of a vinylidene chloride/vinyl chloride polymer was performed on the dried powder. The powder was heated in the pyrolysis chamber at 180 °C with helium carrier flow for 5 min. A volatility experiment showed there were no detectable materials released during this period. Our studies also showed that vinylidene chloride/vinyl chloride copolymers will start evolving hydrogen chloride above 250 °C as a result of the dehydrochlorination reaction. The rate of hydrogen chloride evolution increases with temperature. Table 2 lists the results of the timedependent study of vinylidene chloride/vinyl chloride copolymer composition at the 180 °C pyroprobe interface temperature. The results show there is no significant effect on the composition for the first 6 h.
was accomplished by comparing retention times with those of standard compounds, as well as identification by Py-GC/MS in the electron ionization (EI) mode. Benzene, chlorobenzene, dichlorobenzene, and trichlorobenzene are four major products formed in the pyrolysis of vinylidene chloride/vinyl chloride copolymer. To make the composition calculation, the first assumption is that all trimer peak intensities generated from the pyrolysis gas chromatography after correction for pyrolysis efficiency and detection efficiency accurately represent the triad distribution of the vinylidene chloride/vinyl chloride copolymer. If a close relationship exists between the triad distribution in the polymer chain and the production of trimers in pyrolysis, the composition and number average sequence length can be calculated on the basis of the trimer production in the pyrolysis. Two major factors dominate the relationship between triad distribution and trimer production. The first is the pyrolysis efficiency, which represents the probability/efficiency of breakdown of a specific triad configuration to produce the corresponding trimer. The second is the detection efficiency, which results in variable FID responses for the trimers. These two factors cannot be separated in the vinylidene chloride/vinyl chloride copolymer composition and structure determination case. The relationship between trimer production and triad distribution can be expressed as
RESULTS AND DISCUSSION Figure 1 shows the typical pyrogram of a vinylidene chloride/ vinyl chloride copolymer. The identification of all four trimers
exptl trimer peak intensity × Kn f triad distribution in the polymer Analytical Chemistry, Vol. 68, No. 3, February 1, 1996
427
Figure 1. Typical pyrogram of a vinylidene chloride/vinyl chloride copolymer.
where Kn is the combination of pyrolysis efficiency and detection efficiency. Because the trimers CCD, DCC, and CDC all form chlorobenzene, the second assumption needs to be made that Kn is the same for CCD, DCC, and CDC. Additionally, the same assumption needs to be made for dichlorobenzene with the trimers of DDC, CDD, and DCD. The triad distribution in the polymer and the trimer peak intensities from pyrolysis can be written as
benzene peak intensity × K1 ) CCC distribution in the polymer chlorobenzene peak intensity × K2 ) CCD, DCC, CDC distribution in the polymer dichlorobenzene peak intensity × K3 ) CDD, DDC, DCD distribution in the polymer trichlorobenzene peak intensity × K4 ) DDD distribution in the polymer
The K1, K2, K3, and K4 values can be calculated by pyrolyzing four different known compositions of vinylidene chloride/vinyl chloride copolymer standards. The composition calculation from the trimers will be as follows:
mol % of C ) normalized/corrected benzene peak intensity + (2/3)normalized/corrected chlorobenzene peak intensity + (1/3)normalized/ corrected dichlorobenzene peak intensity 428
Analytical Chemistry, Vol. 68, No. 3, February 1, 1996
mol % of D ) normalized corrected benzene peak intensity + (2/3)normalized/corrected dichlorebenzene peak intensity + (1/3)normalized/ corrected chlorobenzene peak intensity Table 3 shows the composition results calculated from pyrolysis peak intensities compared with the 1H-NMR results for six different compositions of vinylidene chloride/vinyl chloride copolymers (samples A-F). The Kn values used in this calculation are K1 ) 1.0, K2 ) 1.0, K3 ) 1.0, and K4 ) 2.8. The determination of number average sequence lengths for the vinyl chloride and vinylidene chloride is challenging owing to the fact that six of eight trimers are not resolved by pyrolysis gas chromatography. As mentioned previously, there is no way to know how much chlorobenzene peak intensity is contributed from triad CCD and DCC or from CDC. The same situation exists for the dichlorobenzene peak from the triad of DDC, CDD, and DCD. To utilize eqs 1 and 2, a third assumption needs to be made to obtain all six terms of triad intensities. The third assumption results from the known reactivity difference between vinyl chloride and vinylidene chloride monomers in the copolymerization process. In the polymer molecules formed in the beginning of the polymerization, vinyl chloride exists in the polymer chain as a single unit among many vinylidene chloride units. In the polymer molecules formed in the latter part of the polymerization, vinylidene chloride exists in the polymerization chain as a single unit among many vinyl chloride units. Because of the relative reactivity ratio (r1, r2) of vinylidene chloride and vinyl chloride monomers, the probability of forming an alternating monomeric unit in the copolymer molecules is minimal. With this assumption, the polymer molecules have a vinylidene chloride/vinyl chloride distribution as follows:
Table 3. Composition Results Calculated from Pyrolysis Peak Intensities Compared with the 1H-NMR Results of Five Different Compositions of Vinylidene Chloride/Vinyl Chloride Copolymer sample
corrected normalized peak
b
intensitya
Cl0B Cl1B Cl2B Cl3B
A
B
C
D
E
0.019 0.078 0.278 0.626
0.024 0.087 0.281 0.607
0.034 0.104 0.286 0.576
0.062 0.136 0.275 0.527
0.366 0.227 0.265 0.142
F 0.963 0.002 0.005 0.029
pyrolysis molar %
VC(C) VDC(D)
16 84
18 82
20 80
24 76
61 39
97 3
pyrolysis wt %
VC(C) VDC(D)
11 89
12 88
14 86
17 83
50 50
95 5
1H-NMR
VC(C) VDC(D)
11 89
12 88
14 86
17 83
48 52
95b 5b
wt %
a Where Cl0B represents benzene, Cl1B represents chlorobenzene, Cl2B represents dichlorobenzene, and Cl3B represents trichlorobenzene. Weight percentage data from commericial product specification.
beginning part of polymerization - - -DDDDDDDCDDDDDD- - - - -DDDDCDDDDCDDDD- - - - -DDDCDDDCDDDCDD- - - - -DDCDDCDDCDDCDD- - l - - -CCDCCDCCDCCDCC- - - - -CCCDCCCDCCCDCC- - - - -CCCCDCCCCDCCCC- - latter part of the polymerization - - -CCCCCCCDCCCCCC- - -
different physical properties or performance. To explore that difference in the structure, it is important to determine the fraction of two or more of the same type of monomers bonded together. Such parameters are termed the percent of grouped monomers and the number average sequence length of grouped monomers. For monomer C in the CD copolymer, these two terms are expressed as follows:
% grouped monomer C ) (NCCC + NCCD+DCC) × 100% The triad distribution can be expressed as follows:
NCCC ) normalized/corrected benzene peak intensity NCCD + NDCC ) (2/3)normalized/corrected chlorobenzene peak intensity NCDC ) (1/3) normalized/corrected chlorobenzene peak intensity NCDD + NDDC ) (2/3)normalized/corrected dichlorobenzene peak intensity NDCD ) (1/3)normalized/corrected dichlorobenzene peak intensity NDDD ) normalized/corrected trichlorobenzene peak intensity These terms are subsequently used for the number average sequence length determination. Table 4 shows a comparison of Py-GC and NMR values for the number average sequence length calculation of vinylidene chloride/vinyl chloride copolymer samples A-F. The number average sequence lengths for vinyl chloride, N(C), and vinylidene chloride, N(D), of samples C, D, and E are all within 1 unit difference (1.53 vs 1.40, 1.78 vs 1.50, and 3.69 vs 3.00 for vinyl chloride, 6.17 vs 5.70, 5.51 vs 5.10, and 2.40 vs 2.20 for vinylidene chloride). These values indicate that the results from Py-GC have a very good agreement with 1H-NMR results. Because of the reactivity difference between the vinyl chloride and vinylidene chloride monomers, a slight shift in the polymerization temperature, or the monomer feed ratio and feed sequence, will likely produce copolymers with different structures. This will be reflected in copolymers having the same composition yet
or
(1 - NDCD) × 100% (3)
number average sequence length of grouped monomers ) NCCC + NCCD+DCC (4) (1/2)NCCD+DCC Table 5 lists composition and structural calculation for five vinylidene chloride/vinyl chloride copolymer samples G-K. The first six rows give the normalized trimer peak intensities and the composition calculated from them. The second six rows give the six different normalized trimer peak intensities obtained experimentally with the assumptions made previously. The number average sequence length of each monomer was calculated from them. The last four rows give the molar percentages of grouped monomers and the number average sequence lengths of grouped monomers calculated using on the formulae described above. Among these five samples, three had well-controlled synthesis conditions, and temperature varied beyond the desired limits in the synthesis of the other two samples. The compositions of the five samples are very similar for vinyl chloride (13%-14%). Checking the number average sequence length (NASL) data (as in Table 5, NASL N(C) and N(D) rows), it is clear that samples G, H, and I have similar composition/structure and samples J and K are slightly different. Upon examining the number average sequence lengths of the grouped monomers (GNASL) monomer data (as in Table 5, GNASL N(GC) and N(GD) rows), sample K was the most significantly different in preparation conditions and correspondingly showed the greatest variation in the sequence data. With this example, the determination of the percent of grouped monomers and number average sequence length of grouped monomers has allowed us to explore the copolymer structural differences which were not observed in bulk composition or conventional number average sequence length measurements. Analytical Chemistry, Vol. 68, No. 3, February 1, 1996
429
Table 4. Composition Results Calculated from Pyrolysis Peak Intensities Compared with the 1H-NMR Results of Five Different Compositions of Vinylidene Chloride/Vinyl Chloride Copolymera sample
corrected normalized peak intensity
Py-GC NASL
1H-NMR
a
NASL
CCC CCD DCD DDD DDC CDC N(C) N(D) wt % C wt % D N(C) N(D) wt % C wt % D
A
B
C
D
E
F
0.115 0.317 0.568 0.748 0.222 0.031 1.38 7.06 11.16 88.84
0.138 0.330 0.532 0.737 0.228 0.035 1.43 6.70 12.12 87.88
0.172 0.349 0.479 0.719 0.238 0.043 1.53 6.17 13.79 86.21 1.40 5.70 13.66 86.34
0.253 0.371 0.375 0.697 0.243 0.060 1.78 5.51 17.25 82.75 1.50 5.50 14.95 85.05
0.604 0.250 0.146 0.360 0.448 0.192 3.69 2.40 49.73 50.27 3.00 2.20 46.77 53.23
0.997 0.002 0.002 0.876 0.100 0.024 392.90 13.53 94.93 5.07
Where NASL represents number average sequence length.
Table 5. Composition Results Calculated from Pyrolysis Peak Intensities Compared with the 1H-NMR Results of Five Different Compositions of Vinylidene Chloride/Vinyl Chloride Copolymer sample G
H
I
J
K
0.064 0.080 0.233 0.623
0.068 0.088 0.230 0.614
0.067 0.079 0.231 0.623
0.033 0.100 0.274 0.593
0.095 0.047 0.213 0.645
corrected normalized peak intensitya
Cl0B Cl1B Cl2B Cl3B
pyrolysis wt %
VC(C) VDC(D)
corrected normalized peak intensity
CCC CCD DCD DDD DDC CDC
0.328 0.274 0.398 0.774 0.193 0.033
0.334 0.288 0.378 0.771 0.193 0.037
0.341 0.268 0.391 0.775 0.192 0.033
0.172 0.349 0.479 0.733 0.226 0.041
0.479 0.160 0.361 0.803 0.177 0.020
Py-GC NASLb
N(C) N(D)
1.87 7.72
1.92 7.51
1.91 7.77
1.53 6.48
2.27 9.24
pyrolysis molar %
%G(C)d %G(D)e
60 97
62 96
61 97
pyrolysis GNASLc
N(GC) N(GD)
4.40 10.03
4.33 10.00
4.54 10.09
13 87
14 86
14 86
13 87
52 96 2.99 8.49
14 86
64 98 8.02 11.07
a Where Cl0B represents benzene, Cl1B represents chlorobenzene. Cl2B represents dichlorobenzene. Cl3B represents trichlorobenzene. b Number average sequence length. c Grouped number average sequence length. d Percent grouped monomers of the vinyl chloride monomer. e Percent grouped monomers of the vinylidene chloride monomer.
CONCLUSIONS The determination of the composition and structure of vinylidene chloride/vinyl chloride copolymer is made possible through the detection of the pyrolysis trimers and the application of three critical assumptions for the polymer system. The monomer number average sequence distribution and composition of all polymers tested are in excellent agreement with 1H-NMR measurement and the product composition specification (sample F). To distinguish the structural differences of polymers with the same composition, the structure of a copolymer of two monomeric types can be quantified by further deriving the percent of grouped monomers and the number average length of grouped monomers.
430 Analytical Chemistry, Vol. 68, No. 3, February 1, 1996
This method could be extended to any copolymer system, as long as all six trimer peaks can be identified and the peak intensities obtained represent the polymer compositions. This method extends the capabilities of pyrolysis not only in the quantitative study of monomer composition but also in the realm of polymer structure investigation. Received for review September 29, 1995. November 20, 1995.X
Accepted
AC950988T X
Abstract published in Advance ACS Abstracts, January 1, 1996.
