Interpretation of very high activation energies and frequency factors in TL as being due to competition between centres

Very high values, considered to be physically impossible, were reported for the activation energy E and the frequency factor s in some thermoluminescence (TL) peaks. In particular, peak 5 of LiF:Mg,Ti was reported to have E = 2.0-2.2 eV and s = 10(20)-10(22) s(-1). If a peak shape method or curve fitting of a TL peak is utilised, there is an inverse relation between the width and the activation energy. Thus, if for a given TL peak, competition causes a narrowing of the peak, the effective activation energy is larger than expected. This, in turn, will cause the effective s to be significantly larger than the original one. In this study a model of one trap and three centres is considered, and the relevant simultaneous differential equations solved. Concentrating on the second peak, which looks like a first order peak and is very narrow, results in E = 2.1 eV and s = 8x10(20) s(-1), for original parameters of E = 1.2 eV and s = 2.5 x 10(11) s(-1).