Localized transitions in the thermoluminescence of LiF : Mg,Ti: potential for nanoscale dosimetry

We describe the effect of nanoscale spatially coupled trapping centre (TC)–luminescent centre (LC) pairs on the thermoluminescence (TL) properties of LiF : Mg,Ti. It is shown that glow peak 5a (a low-temperature satellite of the major glow peak 5) arises from localized electron–hole (e–h) recombination in a TC–LC pair believed to be based on Mg–Livac trimers (the TCs) coupled to Ti(OH)n molecules (the LCs). Due to the localized nature of the e–h pair, two important properties are affected: (i) heavy charged particle (HCP) TL efficiency: the intensity of peak 5a relative to peak 5 following HCP high-ionization density irradiation is greater than that following low ionization density irradiation in a manner somewhat similar to the ionization density dependence of the yield of double-strand breaks (DSBs) induced in DNA. Our experimental measurements in a variety of HCP and fast neutron radiation fields have demonstrated that the ratio of glow peaks 5a/5 is nearly independent of particle species for the protons, deuterons and He ions investigated, is somewhat dependent on HCP energy, and is roughly 2–3 times greater than the peak 5a/5 ratio in low ionization density electron and photon fields. The intensity ratio of peak 5a/5 thus has the potential of estimating the ratio of dose deposited via high ionization density interactions compared to low ionization density interactions in a nanoscale volume without any prior knowledge of the characteristics of the radiation field, (ii) non-linear TL dose response: the relative lack of competitive processes in the localized recombination transitions leading to the TL of composite peak 5 versus the dose-dependent competitive processes in conduction band-mediated delocalized luminescence recombination leads to non-linear dose response (supralinearity) for composite peak 5 and a dependence of the supralinearity on ionization density. This behaviour is modelled in the framework of the unified interaction model (UNIM).