In studying the thermal degradation of polymers, experimental problems
such as that of taking the samples rapidly to chosen degradation temp
eratures can be solved by appropriate filament design and the use of s
ufficiently small samples. The pyrolysis-GC technique can then provide
reliable quantitative data which can be used in the following ways to
obtain mechanistic information about degradation processes. (I) Depen
dence of rate constants for volatiles evolution on the initial molecul
ar weight of the sample, and on the pyrolysis temperature, can give sp
ecific information about the mechanisms of initiation and temination i
n chain degradation processes. (II) Direct comparisons of observed rat
e constants for volatiles evolution can indicate whether all observed
products are primary products. (III) Anomalous dependence of product y
ields with decreasing sample thickness can give information about diff
usion control of degradation rates. (IV) Anomalous dependence of produ
ct yields and ratios which increases with increasing sample thickness
can distinguish primary, secondary, and even tertiary products, and ca
n moreover lead to the interpretation of the overall pyrolysis mechani
sms in terms of a reaction scheme involving parallel and consecutive r
eactions. All of the above are illustrated with examples from studies
by the authors on a number of polymeric systems.