A study of the pyrolysis of poly(ethylene terephthalate) by means of model systems

The increasing importance of poly(ethylene terephthalate) terylene has made it necessary to have a fuller understanding of the manner in which this polyester undergoes thermal breakdown. Several investigations into this have already been reported, but in no case has it been possible to produce sufficient evidence from which the actual mechanism of breakdown might be deduced. Furthermore, there is even lacking a proper understanding of the manner of breakdown of simpler esters, particularly those that might be postulated as holding an Intermediate position in the general plan of polyester decomposition. It was clear, therefore, that any further advance in this subject must be restricted to hypothesis until a fuller investigation had been undertaken. It was also clear that this investigation would have to be concerned with simpler esters, which were related to the polyester, and which contained the same groupings. By way of providing this basic knowledge, a series of model compounds has been prepared, and subjected to pyrolysis under varying conditions. These compounds have been chosen because of the relationship they bear to the polyester itself and, in several instances, to products that might be postulated as coming from the polyester breakdown. It has also been within the scope of this work to study the pyrolysis ix of several other model compounds which were not directly related to the polyester. These were commands which contained alight structural modification to the original model compounds, thereby illustrating whether the behaviour observed in the original model was general or not. As a result of this work it has been possible to observe and explain several reactions in the thermal breakdown of eaters which had not hitherto been reported. These findings have included (1) the thermal rearrangement of vinyl esters to a p-ketoaldehyd , which then undergoes decarbonylation to a methyl ketone, (2) the reversibility of thermal rearrangement of alkenyl esters, (3) the thermal disproportionation of diesters, (4) the decarboxylation of esters, and (5) the orientation of proton elimination in alkylene diesters. Several of these results can now be applied successfully to explain the reported findings from earlier studies of the breakdown of the polyester, many of which at that time appeared to be quite anomalous