The Unified Interaction Model applied to the gamma-induced supralinearity and sensitisation of peaks 4 and 5 in LiF:Mg,Ti (TLD-100)

The Unified Interaction Model (UNIM), developed to describe thermoluminescence dose response, is based on both radiation absorption stage and recombination stage mechanisms. The UNIM incorporates both the Track Interaction Model (TIM) for heavy charged particles (localised and highly non-uniform ionisation density distributions) and the Defect Interaction Model (DIM) for isotropically ionising gamma rays and electrons, in a unified and self-consistent conceptual and mathematical formalism. The model is applied to explain the unique features of gamma induced supralinearity and sensitisation of peaks 4 and 5 in LiF:Mg,Ti (TLD-100). An entirely new trapping centre model is proposed for peak 4, based on its mainly 'hole trapping' characteristics, which explains its diametrically opposed behaviour in sensitised material, i.e. a decreased sensitivity and supralinear dose response. The UNIM is based on the major premise that these features arise from a localised trapping entity (the track for HCPs), spatially correlated trapping centres and luminescent centres (TC/LC pairs) for gamma rays and electrons, which dominate the dose response at low dose and are not subject to competitive processes, thus leading to linear dose response behaviour. The evidence for spatially correlated TCs/LCs is reviewed and new evidence based on T(m)-T(stop) data and glow curve analysis of LiF:Mg,Ti glow curves as a function of Ti concentration is also presented. At higher dose levels, the decreasing efficiency of the competitive non-radiative processes relative to the luminescence recombination processes, leads to the supralinear behaviour. The decrease of the supralinearity with decreasing gamma ray energy (increasing ionisation density) arises from the increasing probability of the TC/LC pair to capture simultaneously an electron-hole pair, again leading to geminate recombination not subject to competitive processes. The UNIM is shown to be capable of yielding excellent fits to the experimental data with many of the variable parameters of the model strongly constrained by ancillary optical absorption, and sensitisation measurements.