Thermooxidative degradation of polyolefins in the solid state—4: Heterogeneous oxidation kinetics

Abstract

Chemical kinetics is an ideal tool to examine reactions in solution. It cannot be applied to the phenomena in solid polymers, especially semi-crystalline polymers, without considerable adjustments. This derives directly from the corresponding polymer model, amorphous domains separated by crystallites. The elementary chemical reactions usually occur in the amorphous domains only. It is the kinetics in these amorphous domains which governs the overall chemistry in the polymer. The elementary reaction kinetics known from solution chemistry need to be adapted to these domains. The main point in this respect is to take into account the fact that for two functional groups on a polymer chain to react with one another, either they must be close enough spatially if they are both part of a high molecular mass fragment or at least one of them must be on a low molecular mass fragment so that it is able to diffuse more or less in the solid matrix. Hence, for example, bimolecular hydroperoxide decomposition can be restricted to a large extent to the associated hydroperoxide groups present in solid polyolefins. The problem of the transposition of the kinetics in the amorphous domains to the kinetics in the observation volume, the only kinetics accessible experimentally, is more complicated. It is usually not possible to transpose directly the experimental macroscopic kinetics to the kinetics of the amorphous domains or vice versa. It is the expression of the kinetic law which determines the possibility or impossibility of direct transposition. Usually, the heterogeneous nature of the polymer has to be taken into account for the passage back and forth. Then there are two main possibilities to be envisaged. The first possibility uses the heterogeneous oxidation model with its mathematical expression. Then the rate law in the amorphous domain can be transformed into the rate law for the observation volume so that the elementary law can be checked. The second possibility envisages the kinetics exclusively in the amorphous domains and calculates it for the observation volume through the use of the spreading equation derived for the heterogeneous oxidation model. This also allows for a check on the agreement with the experimental data and, therefore, the assumptions concerning the elementary kinetics in the amorphous domains. The model permits the interpretation of the experimental kinetics for associated hydroperoxides as well as for free hydroperoxides discussed previously. It also yields a good fit for the accumulation of carbonyl groups on thermal oxidation of PE-LD. Possible improvements are discussed.