Analysis of rubber materials by pyrolysis GC.

Because few analytical samples are simply pure materials, chromatographic techniques have become widely applied to their analysis, used to resolve the various components before identification. Gas chromatography, especially coupled with a mass spectrometer, provides a powerful tool for the identification of various components in a complex mixture or solution. Samples are traditionally injected in a solution, and must consist of analyses capable of vaporizing at temperatures below about 325 [degrees] C to pass through the column. The utility of GC has prompted analysts to devise means to introduce samples by means other than syringe to meet the needs of specific applications, including vapor-phase sample loops, heated headspace injections and thermal desorption of compounds from a solid matrix.

What GC would seem incapable of analyzing are solid materials which cannot be volatilized at the operating temperatures of the instrument, materials such as polymers, paints, rubbers and goods manufactured from them. Pyrolysis of these samples, however, generates volatile compounds from them in a way which is both reproducible and characteristic of the original macromolecule. When heated to pyrolysis temperatures (generally from 400 [degrees] to 800 [degrees] C), the large molecules in polymer samples fragment into smaller compounds which retain much of the chemical structure and information about the polymer. When these volatile compounds are analyzed by gas chromatography they produce a pyrolysis-chromatograrn or "pyrogram."

Many polymers generate significant amounts of monomer, sometimes almost exclusively, sometimes with higher oligomers. Polymethyl methacrylate, for example, "unzips" almost entirely to methyl methacrylate monomer, and produces a chromatogram like that shown in figure 1. At the other extreme, polyethylene generates a wide distribution of oligomers, producing a complex but repeating pattern of peaks in the pyrogram, which always resembles the chromatogram shown in figure 2. Other polymers generate pyrograms just as characteristic. Polyethyl methacrylate, polybutyl methacrylate and so on, produce simple chromatograms of mostly monomer, while polypropylene, polybutene and the other polyolefins produce patterns of many oligomers.

[Figure 1-2 ILLUSTRATION OMITTED]

In general, if two polymers, or polymeric systems are different from each other, they will produce pyrograms which are different from each other in ways which help identify those differences. Consequently, pyrolysis-gas chromatography (Py-GC) and Py-GC/ MS have been used for decades to elucidate the structures of polymers, quantitate the monomer ratios in copolymers, differentiate among various polymer formulations, and identify complex items used in everyday life. Forensic applications include the analysis of adhesives used to make bombs (ref. 1), automotive paint (ref. 2) and photocopier toners (ref. 3). Smith et al. (ref. 4) have analyzed streptococci by Py-GC/ MS, while direct pyrolysis-MS is employed by Scisson in similar investigations, including Legionnaires Disease (ref. 5). Direct PY-MS is applied to a variety of synthetic polymers by Quan et al. (ref. 6). PY-GC has been employed to study polyesters (ref. 7), vinyl polymers (ref. 8) edible oils (ref. 9), materials in art and archeology (ref. 10) and a wide array of other materials, including rubbers, as will be discussed here.

Experimental

Pyrolysis for all samples was carried out with a setpoint of 750 [degrees] C for 10 seconds, using a CDS Analytical Pyroprobe 2000, and autosampler. The autosampler pneumatics oven was set to 300 [degrees] C, and the pyrolysis chamber was cleaned to vent at 1,000 [degrees] C between runs. Samples of approximately 100 [Micro]g were placed into quartz tubes and held in place with quartz wool. The Pyroprobe autosampler was interfaced to a Hewlett Packard 6890 equipped with a HP-5 column (30 m, 0.25 mm, 1 [Micro]m film) and an MSD. The GC was programmed from 40 [degrees] for two minutes to 295 [degrees] at 6 [degrees] C/minute, using an inlet pressure of 6 psi and a 75: 1 split ratio.

Results and discussion

Although many rubber products are based on copolymers of two or more different monomers, it is helpful to look at pyrograms of homopolymers before examining more complex systems. Natural rubber is polyisoprene, and when pyrolyzed yields large amounts of the monomer (isoprene), the dimer (dipentene) and higher oligomers. In figure 3 isoprene is the first large peak, dipentene elutes at about 11 minutes, and the trimers, tetramers and pentamers can be seen as bunches of peaks at about 20, 30 and 37 minutes respectively. The pyrogram of polybutadiene resembles that of polyisoprene, with the first peak monomer, the dimer (vinyl cyclohexene) at six minutes, then groups of peaks representing the trimers, tetramers and pentamers of butadiene. Polystyrene is converted extensively to monomer, which elutes at about seven minutes under these conditions, but also produces dimer and trimer, as marked in figure 5. Finally, polyacrylonitrile is shown in figure 6, with the monomer and dimer marked.

[Figure 3 - 6 ILLUSTRATION OMITTED]

Although the overall pyrograms make distinguishing among these different polymers fairly obvious, it should be pointed out that the analysis and identification of the specific compounds in each pyrogram reveals much about the detailed nature of the specific polymer. Stereo chemistry may be elucidated, as for polystyrene by Nonobe et al. (ref. 11), and defect structures in polyisoprenes (head-to-head and tail-to-tail structures) were studied by Hackathom and Brock in 1971 (ref. 12). In addition, the effects of temperature and sample dimensions on the monomer and dimer production from natural rubber have been documented by Groves et al. (ref. 13).

Copolymers of these four important monomers - butadiene, isoprene, styrene and acrylonitrile - will also generate (among other things, of course) peaks for the monomers, so a qualitative identification of the materials used in a typical rubber formulation is a fairly straightforward task. In fact, as with the homopolymers, a great deal of information is to be found in the analysis of specific peaks, relating to microstructure (ref. 14) and sequencing of monomer units in block and random copolymers (ref. 15), but even a general analysis can reveal the monomers involved, identifying at least the type of polymer or copolymer system being assayed. Looking at a copolymer of styrene and butadiene then, as in figure 7, the two largest peaks are the styrene and butadiene monomers. Polystyrene/acrylonitrile (figure 8) shows a similar result, with acrylonitrile and styrene again prominent in the pyrogram. Further, if the rubber were formulated from acrylonitrile, butadiene and styrene, each monomer would be present.

[Figure 7-8 ILLUSTRATION OMITTED]

Analysis of rubber products

Based on the analysis of known rubbers, both homopolymers and copolymers, what can be shown for typical rubber goods which people encounter in everyday life? Many things are identified as rubber or latex, some indicating that they are pure, natural rubber, others that they are a rubber product, formulated with other things as well. Figure 9 shows the pyrogram of a sample taken from a rubber glove. The package containing the gloves indicated that they were made of natural latex, and in fact, the pyrogram shown only isoprene monomer and dimer, and the entire pattern closely resembles that of the natural rubber shown in figure 3. Most rubber bands and dried rubber cement produce pyrograms which look much the same.

[Figure 9 ILLUSTRATION OMITTED]

The effect of filler materials, especially carbon black on the pyrolysis products from rubber goods has been studied (ref. 16), and in general produces little interference. This is mostly because pyrolysis degradation mechanisms are largely intramolecular free radical reactions, which would take place in the rubber sections of the product, but have very little interaction with the carbon black particles. (This also explains why random and block copolymers may be differentiated via pyrolysis). Figure 10 is the pyrogram of a natural rubber product, with carbon black filler, after curing. The presence of isoprene does indicate natural rubber, but the product clearly contains butadiene in addition. Both monomers are confirmed by the presence of the respective dimers as well. The ability of pyrolysis-GC to analyze products formulated with blends of natural rubber, synthetic rubber and carbon black filler makes it a good choice for the investigation of automobile tires. Figure 11 compares analyses of samples taken from tires manufactured by three different sources. Each can be seen to contain isoprene monomer and dimer as well as butadiene monomer and dimer. Only the first 15 minutes of the pyrograms are shown, expanded to allow comparison. An easy indication of the relative monomer content can be seen by comparing the sizes of the two dimer peaks.

[Figure 10 - 11 ILLUSTRATION OMITTED]

Figure 12 shows a good example of how PyGC can indicate quickly whether material is actually what it is thought to be. This product was labeled "ABS," so the pyrogram should show peaks for acrylonitrile, butadiene and styrene monomers at a minimum. Instead, the analysis shows no acrylonitrile or butadiene, but that the material is a styrene/isoprene copolymer. Again, the presence of isoprene is confirmed by the dimer, dipentene. In another example, the butadiene elastomer shown in figure 13 contains not only butadiene and styrene, but also methyl methacrylate.

[Figure 12-13 ILLUSTRATION OMITTED]

The term rubber is also applied to consumer goods which are soft or flexible, but are chemically quite different from polymers of olefins like isoprene and butadiene. Foam rubber is almost always polyurethane foam, and the polyurethanes are very frequently made using toluene diisocyanate (TD1). When pyrolyzed, these polymers regenerate a substantial amount of TDL which elutes at about 19 minutes under the conditions used here, as shown in figure 14. Further, polyurethanes are generally classified as polyester or polyether types, which may also be differentiated using Py-GC. Finally, silicone rubbers represent another polymeric material encountered as caulking and sealants. When silicone rubber (polydimethylsiloxane) is pyrolyzed, it produces cyclic oligomers and a pyrogram showing a repeating pattern of individual peaks, as in figure 15.

[Figure 14 - 15 ILLUSTRATION OMITTED]

Conclusions

The range of polymers, copolymers and blended systems termed rubbers is quite extensive, including natural rubber, synthetic polyolefins, polyurethanes and silicone polymers. Finished goods like automobile tires, bumpers, insulation and rubberized clothing articles will include not only the polymers or copolymer systems, but fillers, fibers, colorants and so on as well. Pyrolysis-GC/MS permits the identification of the specific polymers used in such materials, regardless of the presence of such additives, and may be used to demonstrate the relative amounts of various monomers used. This makes the analysis of such materials a relatively straightforward task, easily indicating the presence of additional materials such as acrylics. Further analysis reveals information about the structure of the polymers themselves, including molecule structure, defects and stereochemistry.

References

(1.) N.L Barkowski, E.C. Bender and T.O. Munson, J. Anal. Appl. Pyrolysis, 8, 483 (1985).

(2.) K Fukuda, Forensic. Sci. Internat, 29, 227 (1985).

(3.) E.J. Levy and T.P. Wampler, J. Forensic Sci., 31, 258 (1986).

(4.) C.S. Smith, S.L. Morgan, C.D. Parks, A. Fox and D.G. Pritchard, Anal. Chem., 59, 1410 (1987).

(5.) P.R. Scisson, R. Freeman, N.F. Lightfoot and I.R. Richardson, Epidemiol. Infect 107, 127 (1991).

(6.) K. Qian, W.E. Killinger, M. Casey and G.R. Nicol, Anal. Chem., 68, 6, 1019 (1996).

(7.) H. Ohtani, T. Kimura and S. Tsuge, Anal. Sci., 2, 179, (1986).

(8.) R.P. Lattimer, W.J. Kroenke and R.G. Getts, J. Appl. Poly. Sci., 29, 3783 (1984).

(9.) J.M. Nazer and C.T. Young, J. Food Sci., 49, 662 (1984).

(10.) A.M. Shedrinsky, T. P. Wampler, N. Indictor, N. S. Baer, J. Anal. Appl. Pyrolysis, 15, 393 (1989).

(11.) T. Nonobe, H. Ohtani, T. Usami, T. Mori, H. Fukumori, Y. Hirata and S. Tsuge, J. Anal. Appl. Pyrolysis, 33, 121, (1995).

(12.) M. J. Hackathom and M. J. Brock, Polymer Letters, 8, 617 (1971).

(13.) S.A. Groves, R.S. Lehrle, M. Blazso and T. Szekely, J. Anal. Appl. Pyrolysis, 19, 301 (1991).

(14.) V.G. Zaikin, R-G. Mardanov, V.A. Yakovlev and N.A. Plate, J. AnaL Appl. Pyrolysis, 23, 33 (1992).

(15.) S. Tsuge, Y. Sugimura and T. Nagaya, J. Anal. Appl. Pyrolysis, 1, 221 (1980).

(16.) M. J. Matheson, T.P. Wampler and W. J. Simonsick, J. Anal. Appl. Pyrolysis, 29, 129 (1994).