Do the CONTIN or the MELT programs accurately reveal the o-Ps lifetime distribution in polymers? Analysis of experimental lifetime spectra of amorphous polymers

High quality positron lifetime spectra (total count of 70 million) were taken for polycarbonate (PC) and polystyrene (PS Hoechst). The spectra were analysed employing the routines CONTIN and MELT which both assume continuous lifetime distributions, as well as the discrete-term analysis routine LIFSPECFIT. The lifetime distributions consist of three peaks, of these, the long-lived one, corresponding to the o-Ps lifetime distribution, splits for low regularization into two sub peaks. No way was found to decide whether these sub peaks, or the smoothed o-Ps lifetime distribution obtained for higher regularization, were due to two (discrete) close neighbour lifetimes, or came from a single, broad o-Ps lifetime distribution. Assuming discrete lifetimes, values of approximately 1700 ps and 2350 ps were estimated for PS (1750 and 2400 ps for PC). Assuming a single but broad o-Ps lifetime distribution, the mass centre tau(3) of that in PS was estimated to 1985 ps (2020 ps in PC). The same analysis was carried out with distributions from simulated lifetime spectra for comparison with experimental data. From this comparison the full width at half maximum (FWHM) of the o-Ps lifetime distribution was estimated to 800-950 ps for both PS and PC. The o-Ps lifetime results may be attributed to two different Ps states or, alternatively, to hole size distributions ranging approximately from r = 0.2-0.32 nm (V = 0.06-0.14 nm(3)). Furthermore. short comments on how to estimate a reliable width for the o-Ps lifetime distribution and on the handling and resolving power of CONTIN and MELT are given. (C) 1998 Elsevier Science B.V. All rights reserved.