The effect of carbon-black filling on the pyrolysis behavior of natural and synthetic rubber.

Natural rubber and synthetic styrene-butadiene rubber of known formulations with and without carbon filler were analyzed by pyrolysis - Fourier transform-infrared spectroscopy (Py-1R); pyrolysis - gas chromatography/flame ionization detection (Py-GC); and pyrolysis - gas chromatography/mass spectrometry (Py-GC/MS) to determine the effect, if any, of the presence of carbon fillers on the pyrolysis behavior. Carbon black did not affect the qualitative results.

Py-IR and Py-GC are rapid and reliable methods for the identification of the polymer used in commercial carbon-filled elastomers, with no interference from the carbon filler.

Carbon-filled rubbers can be difficult to analyze by spectroscopic or gas chromatographic methods because they are often insoluble and are opaque owing to the presence of the carbon black. Removing the carbon black by extraction is a cumbersome step and is not always completely successful (ref. 1). The usual analytical tools applied to polymers without carbon black cannot always be successfully applied in its presence. For instance, attenuated total reflection spectroscopy (ATR) is of great value in polymer analysis, but is of limited use in the case of carbon-filled rubber analysis because the carbon black causes large absorption bands in the spectrum which often obscure the region of interest.

Py-GC and Py-1R are well-established techniques for analyzing polymers (refs. 2-7), but there are few reports in the literature utilizing them in the presence of large amounts of carbon black. In one study utilizing capillary GC/MS for the study of carbon black filled Neoprene vulcanizate, the mass pyrogram of the filled vulcanizate was found to be identical to the resin, with the exception of the relative intensities of the chloroprene ions, which were three times less in the vulcanizate, presumably due to cross-linking between carbon black and the polymer obtained during the curing process (ref 8). The same study concluded that direct pyrolysis/MS could be used routinely for quantitative analysis of carbon black filled poly(epichlorohydrin-co-ethylene oxide). In another study, carbon black filler was shown to contribute both physical support and increased polymer strength through intermolecular bonding with polybutadiene after vulcanization (ref. 7). These authors found that the ratio of some peak areas changed with increasing vulcanization, but the pyrogram pattern remained the same. In a detailed study of carbon-black elastomer interactions, Ayala et al., demonstrated that the surface of carbon blacks contains numerous readily ionizable fragments such as methyl, [C.sub.2][H.sub.x], [C.sub.3][H.sub.x], etc., including double bonds available for possible interaction with elastomer molecules (ref. 9). Thus, during the heat treatment of vulcanization, it is possible that chemical bonding occurs between the elastomer and the carbon black. The amount of chemical bonding that occurs may depend on the degree of unsaturation of the base polymer; natural rubber (NR) has been found to have less affinity for carbon black than styrene-butadiene rubber (SBR) (ref. 10). This suggests that some types of rubber may have an altered Py-GC chromatogram (pyrogram) or FT-1R fingerprint spectra relative to unfilled material.

To demonstrate the feasibility of using these techniques to determine the polymer composition of carbon-filled elastomeric materials, fully characterized ASTM samples of cured NR and cured SBR, with and without carbon black filler, were analyzed by Py-GC, Py-GC/MS and Py-1R. The presence of carbon black did not affect the results.

Experimental

Pyrolysis/FT-IR

A pyroprobe interfaced to a FT-1R spectrometer equipped with a DTGS detector with a scan speed of 1 scan [s.sup.-1] with a Brill Cell,. with standard 0.635 cm ZnSe windows, was used for these analyses. The Pyroprobe is a filament type pyrolyzer that can be equipped with a platinum coil or platinum ribbon; a coil was used for these experiments. The interface temperature was held at 120[degree] C and the nitrogen purge gas flow was 40 ml [min.sub.-1]. Unless noted otherwise, all pyrolysis was performed at 850[degree] C for 25 s in a quartz boat inserted in the pyroprobe coil. Since the quartz boat absorbs some of the heat, the effective pyrolysis temperature was about 750[degree] C. The sample size was approximately 10 mg; exact sizes are given on the figures.

Pyrolysis-GC/FID

A pyroprobe with a platinum coil and PeakMaster interfaced to a GC was used for Py-GC with a flame ionization detector. The GC capillary column was a 30 m by 0.53 mm SE54. Pyrolysis occurred in the thermal desorber of the PeakMaster, and the resulting volatiles were swept onto a Tenax trap, then desorbed onto the head of the GC column. The PeakMaster program was as follows: valve oven and transfer line, 300[degree] C; desorber temperature, 300[degree] C for 10 min at 40 ml [min.sup.-1] He flow; trap A rest, 30[degree] C; trap A desorb, 300[degree] C for 4 min.; trap A bake, 300[degree] C for 5 min. The GC temperature program was as follows: start 35[degree] C, hold 6 min., then ramp at 6[degree] C [min.sup.-1] to 280[degree] C. All pyrolysis temperatures were 850[degree] C for 10 s. The sample sizes were approximately 0.5 mg.

Pyrolysis - GC/MS

A pyroprobe with a platinum coil was interfaced to a GC/MS. All pyrolysis experiments were conducted on about a 0.5 mg of sample at 900[degree] C for 25 s. The pyrolyzates were subsequently loaded in the splitless mode onto a cooled (20[degree] C) 30 M DB-5 capillary, GC column. The column was temperature programmed from 20[degree] C for 2 min to 300[degree] C for 10 min.

The eluting compounds were detected by the mass spectrometer which was operated in the electron ionization mode (70 eV). The ion source temperature was 200[degree] C and the source pressure was < 1 x [10.sup.-4] Pa. The mass spectrometer was scanned from 10-400 Da in 0.5 s. At least five scans were acquired and averaged per eluting peak. The pyrolyzates were identified by library comparison to archived spectra.

Results and discussion

aNALYSIS OF CHARACTERIZED SAMPLES

To determine the validity of the methods employed, fully characterized samples of cured NR (ASTM D3192) and cured SBR (ASTM D3191) with and without carbon black, were obtained. The exact contents of each formulation are given in table 1. Samples with and without carbon black were analyzed by Py-1R, Py-GC and Py-GC/MS. The results, shown in figures 1-6, show no change in the spectra and pyrograms as a result of the presence of carbon black.

Table 1 - composition of standard rubber

Ingredients                                 Parts by mass
Natural rubber                                        100
Stearic acid                                            3
Zinc oxide                                              5
Sulfur                                                2.5
[MBTS.sup.a]                                          0.6
Carbon black (furnace)                                 50
Total                                               161.1

SBR-1500                                              100
Stearic acid                                            1
Zinc oxide                                              3
Sulfur                                               1.75
Carbon black                                           50
[TBBS.sup.b]                                            1
Total                                              156.75
a Benzothiazyl disulfide
b N-tert-butyl-2-benzothiazole sulfenamide

Figures 1 and 2 show overlays of the Py-IR spectra of NR and SBR with and without carbon black. It is evident that the spectra are very similar. Figures 3 and 4 compare the Py-GC chromatograms of the same samples. Again, the patterns are virtually identical.

Figure 5 presents the Py-GC/MS comparison between NR (top) and carbon-black filled NR (bottom). Notice the similarity between the pyrograms. Only minor qualitative differences, which are within the experimental error of the technique, were observed. The most abundant peaks were identified by comparison to archived electron impact spectra. The largest pyrolysis product was identified as isoprene (peak 1). In addition, toluene (peak 2), xylene (peak 3), cyclohexene, 4-ethenyl-1,4-dimethyl (peak 4) and limonene (peak 5) were the major products detected. These data agree with the results obtained by Hirayanagi et al. who quantitatively analyzed various rubber blends using pyrolysis gas chromatography (ref. 11). We focused our analysis and subsequent conclusions on the most abundant pyrolyzates. The compounding ingredients, MBTS for example, were detected but quantitation was unsuccessful owing to the low abundance. Additional decomposition products were not identified.

Figure 6 displays the Py-GC/MS comparison between a SBR (top) and its carbon black filled counterpart (bottom). The pyrogram is slightly more complicated than the NR; however, as seen previously, the carbon black had little effect on the distribution of products. The major products were identified by comparison with archived electron impact spectra as toluene (peak 1), vinyl cyclohexene (peak 2), a [C.sub.8][H.sub.10] isomer (a butadiene dimer) (peak 3) and styrene (peak 4). We did not identify all the peaks in the pyrograms; however. the corresponding mass spectra of each peak were identical. We were not able to analyze the intact compounding ingredients since they exceeded the volatility range amenable to our GC. As before, we suggest that the minor variations in quantitative response were within the experimental error of the technique. Thus, we conclude from the Py-GC/MS results that the presence of carbon black had little or no effect upon the pyrolysis products formed from the NR and SBR.

Conclusions

The ease with which carbon-filled rubbers can be analyzed by Py-IR makes it the method of choice in our laboratory,, for an initial determination of the polymer content of these materials. The method requires no sample preparation, and an experienced spectroscopist can determine at a glance after a 30 s scan what polymer is present. The use of archived spectra and a library search simplifies the identification of even completely new polymers, and the ability to determine the degree of match between two spectra further extends the utility of Py-1R . Significant quantitative information can also be obtained by Py-1R. When the Brill Cell is employed, pyrolysis occurs directly in the path of the light beam in the IR so the evolution of pyrolysis products with time can be observed. The change in pyrolysates as the pyrolysis temperature is changed reveals much structural information. In the samples, the maximum carbon black load was 35%. The carbon black present in the samples does not interfere with the pyrolysis products, at least to this level, so the considerable body of information already published in the literature on unfilled polymers utilizing this technique can be consulted in method development and spectral interpretation.

Py-GC is a well-established technique for obtaining detailed structural information about polymers and copolymers, as well as allowing basic identification. Since GC is a very common technique and is frequently used routinely for quality control in plant situations, Py-GC is particularly amenable to use in those environments by individuals who are already accustomed to the GC instrumentation. In addition, it is a powerful research tool for use on many types of sample. The use of mass spectrometric detectors allows positive identifications of each peak as well as a total ion chromatogram, but it is an expensive and complicated approach if complete identification is not necessary.

References

[1.] L.S. Bark and N.S. Allen, Analysis of Polymer Systems, Applied Science Publishers, Essex, 1982. [2.] S. Tsuge, Y. Sugimara and T. Nagaya, J. Anal. Appl. Pyrolysis, 1(1980) 221. [3.] K.V. Alekseeva, J. Anal. Appl. Pyrolysis, 2 (1980) 19. [4.] J.W. Washall and T.P. Wampler, J. Chromatogr. Sci., 27 (1989) 144. [5.] J.W. Washall and T.P. Wampler, Spectroscopy. 6(4) (1989) 38. [6.] P.J. Gale, B.L Bentz and W.L. Harrington, RCA Rev., 47 (1986) 380. [7.] K.G. Hausler, J.L. Stanford and R.F.T Stepto, J. Anal. Appl. Pyrolysis. 13 (1988) 287. [8.] J.M. McGuire and C.C. Bryden, J. Appl. Polym. Sci..35 (1988) 5.37. [9.] J.A. Ayala, W.M. Hess, F.D. Kistler and G.A. Joyce, Rubber Chem. Technol., 64 (1991) 19, [10.] G.R. Cotten and L.J. Murphy, Rubber Chem, Technol., 61 (1988) 609. [11.] S. Hirayanagi. K. Kimura, M. Sato and K. Harata, Nippon Gomu Kyokaishi, 55 (5) (1982) 302-308.