Re-examination of the thermal oxidation reactions of polymers 1. New views of an old reaction

Bimolecular hydroperoxide decomposition in polymers such as polypropylene is re-examined. Initially the reaction has been established and discussed for hydroperoxides in solution. In this work, the reaction is examined in connection with the conditions encountered with polymer melts and polymers in the solid phase. If the reaction between the two hydroperoxide groups is intermolecular, and even more so, if it is intramolecular, the peroxy radical and the alkoxy radical formed will be close and remain close. As a consequence of this proximity, hydrogen abstraction by one of these radicals will also involve hydrogen atoms in proximity to the second radical. If the alkoxy radical abstracts a hydrogen atom in alpha -position to a peroxy radical formed out of an isolated hydroperoxide group, the resulting alpha -alkyl-peroxy radical is likely to give the "ene" reaction. However, if the peroxy radical formed in the bimolecular decomposition of two hydroperoxide groups is situated between two tertiary hydroperoxide groups the "ene" reaction is not possible. Then, the alpha -alkyl-peroxy radical can yield a 1,2-dioxetane group. Decomposition of this unstable group yields a methylketone and an aldehyde with polymers such as polypropylene (PP). Of course, there is simultaneous chain scission. Furthermore, the energy released in the decomposition of the dioxetane can lead to the formation of an electronically excited state accounting for chemiluminescence. The reactions proposed account also for fast termination reactions in spite of the absence of direct interaction between tertiary alkoxy and peroxy radicals. The situation is quite different with polymers such as polyethylene. Even if there were sequences of hydroperoxide groups the "ene" reaction would remain possible. (C) 2001 Elsevier Science Ltd. All rights reserved.