Evaluation of activation energies in the semi-localized transition model of thermoluminescence

Recently a semi-localized transition (SLT) kinetic model was developed for thermoluminescence (TL), which is believed to be applicable to important dosimetric materials like LiF: Mg,Ti. This model contains characteristics of both a localized transition model and a single trap model and is characterized by two distinct activation energy levels. This paper describes the simulation of several standard methods of analysis for the TL peaks calculated using the SLT model in an effort to extract the two activation energy parameters of the model. The methods of analysis are applied to both possible types of transitions within the model, namely the direct recombination of the hole-electron pairs as well as the delocalized transitions involving the conduction band. In the former case of direct recombination, the methods of analysis give consistent results for the activation energy E. In the latter case of transitions involving the conduction band, it was found that extra caution must be exercised when applying standard methods of analysis to the SLT model because of the possibility of strongly overlapping TL peaks. Specifically the peak shape methods consistently fail to yield the correct value of E, while careful application of the fractional glow, thermal cleaning and variable heating rate methods can yield the correct energy values when no retrapping is present within the localized energy levels. A possible explanation is given for the previously reported failure of the peak shape methods to yield the correct activation energies within the SLT model. The heating rate methods of analysis consistently yield the correct activation energies E with an accuracy of a few per cent.