Douglas M. Barr

Mass spectrometric methods for assessing the thermal stability of liquid polymers and oils: study of some liquid polyisobutylenes used in the production of crankcase oil additives☆

Abstract The onset temperature of thermal degradation of a liquid polymer or oil, and the characteristics of the pyrolysis at higher temperatures, can be assessed by studying the volatiles produced. However, this approach requires that the true thermal degradation products are assessed, and not any components which are simply evaporated from the sample. It is shown that degradation and evaporation behaviour can be distinguished by utilising the principle that thermal degradation produces components with structures which differ from those of components which are simply evaporated. This approach underlies the following two methods which have been used to study the thermal behaviour of liquid polyisobutylene (PIB). (1) In the temperature-sequence gas chromatography (GC) method, the same sample is subjected to a sequence of 10-s pyrolyses at increasing temperatures, and for each temperature the complete chromatogram of the volatile products is recorded. This method gives a good general picture in which it is demonstrated that at lower temperatures only evaporation occurs, but as the temperature is increased, degradation becomes increasingly important. The method has the potential to provide good characterisation of the volatile products, but its disadvantage is that not enough temperature steps can be taken to allow the onset of degradation and its development with temperature to be defined with acceptable accuracy. (2) In the direct mass spectrometry (MS) of volatiles method, a thin film of the sample is subjected to a continuous (small-stepped) temperature programme, during which helium carrier gas sweeps the volatiles continuously into the source of a MS scanning at the rate of 10 mass spectra per second. Appropriate plotting of selected ion currents of chosen fragment ions against time is used to reveal first the direct evaporation, and then the onset and progress of the degradation. Because of the large number of small temperature steps used, the degradation onset temperature is better defined by this method. The second of these methods has been used to compare the thermal behaviour of two liquid PIB samples which had been produced using different catalysts, and also their corresponding PIB-succinic anhydrides (PIBSA). (The latter are used as starting materials for the production of crankcase oil dispersant additives). The results have been interpreted in terms of general degradation mechanisms, on the basis of which a proposal is made for increasing the stability of the materials.