Carboxyl end-group content (CC) of the resins before and after reactive modification was determined according to Pohl's method (40) by titrating a solution of the resin in benzyl alcohol/chloroform with standard NaOH in benzyl alcohol in the presence of phenol
red as indicator. Hydroxyl end group content (HC) was determined by the method described by Robinson et aL (41) and Zimmermann and Tryonadt (42), which involves the reaction of the polyester with a known quantity of 3,5-dinitrobenzoylchloride followed by titration of the excess hydrolyzed acid chloride with standard NaOH in ethylene glycol. An advantage of this tedious and time-consuming method is that the results are not affected by the carboxyl content of the PET since the polyester is filtered before the final titration. The HC method gave good reproducibility (? 2 eq/106 g) for unmodified PET samples; however, for diepoxide or E-GMA modified PETS with higher HC
values, reproducibility was not better than -C 10 eq/ lo6 g, depending on the sample and the particular details of the experimental procedure.