THE MODIFIED UNIFIED INTERACTION MODEL: INCORPORATION OF DOSE-DEPENDENT LOCALISED RECOMBINATION A. Lavon 1 , I. Eliyahu 1,2 , L. Oster 3 and Y. S. Horowitz 1, * 1 Ben Gurion University of the Negev, Beersheva 84105, Israel 2 Soreq Nuclear Research Center, Yavne 81800, Israel 3 Sami Shamoon College of Engineering, Beersheva 84100, Israel *Corresponding author: yigalho@gmail.com Received 6 April 2014; revised 4 May 2014; accepted 9 May 2014 The unified interaction model (UNIM) was developed to simulate thermoluminescence (TL) linear/supralinear dose–response and the dependence of the supralinearity on ionisation density, i.e. particle type and energy. Before the development of the UNIM, this behaviour had eluded all types of TL modelling including conduction band/valence band (CB/VB) kinetic models. The dependence of the supralinearityon photon energy was explained in the UNIM as due to the increasing role of geminate (localised recombination) with decreasing photon/electron energy. Recently, the Ben Gurion Universitygrouphas incorporated the concept of trapping centre/luminescent centre (TC/LC) spatially correlated complexes and localised/delocalised recombin- ation into the CB/VB kinetic modelling of the LiF:Mg,Ti system. Track structure considerations are used to describe the relative population of the TC/LC complexes by an electron–hole or by an electron-only as a function of both photon/electron energy and dose. The latter dependencewas not included in the original UNIM formulation, a significant over-simplification that is herein corrected. The modified version, the M-UNIM, is then applied to the simulation of the linear/supralinear dose–response characteristics of composite peak 5 in the TL glow curve of LiF:Mg,Ti at two representative average photon/electron energies of 500 and 8 keV. INTRODUCTION The unified interaction model (UNIM) was devel- oped at the Ben Gurion University (BGU) of the Negev by Horowitz et al. (1, 2) and Horowitz (3) in the late 1990s in order to simulate the linear/supralinear thermoluminescence (TL) dose –response observed for the glow peaks of LiF:Mg,Ti (TLD-100) and the dependence of the supralinearity on ionisation density, i.e. particle type and energy. TL supralinear- ity is a complex phenomenon dependent on many physical and experimental parameters (4) and no other TL models (5, 6) had previously been capable of mod- elling these characteristics of dose–response. For example, conventional conduction band (CB)/valence band (VB) kinetic models are blind to the effects of non-uniform ionisation density in particle tracks and the effects of overlapping tracks (7) . The UNIM is based on the presence of spatially correlated/coupled trap- ping centres (TCs) and luminescent centres (LCs), which can lead to localised (geminate) recombination. Geminate recombination has been used to describe many luminescent phenomena (6 – 8) . Recently, the BGU group has incorporated the concept of TC/LC spa- tially correlated complexes and localised/delocalised (LDL) recombination in CB /VB kinetic modelling of the LiF:Mg,Ti system (9) . Track structure considera- tions are used to describe the relative concentration of the TC/LC complexes by an electron–hole n e–h , or by an electron-only, n e , as a function of photon energy and dose. The same concept is introduced herein into a modified UNIM (M-UNIM). The dependence of n e–h /n e on dose was not employed in the original for- mulation of the UNIM and was a significant over- simplification. The M-UNIM is a multi-parameter model and many of the parameters (including those governing the relative e–h and e-only population of the TC/LC complex as a function of dose and energy) cannot be estimated from ab initio princi- ples. However, as in the previous applications of the UNIM, an effort is made to restrict the range of allowed values of the parameters using ancillary measurements such as optical absorption, heavy charged particle (HCP) fluence response, reference to other known material characteristics, e.g. dopant levels, etc. The UNIM can also describe fluence response for HCPs (in the framework of the Extended Track Interaction Model) (10) . In fact, two of the variable para- meters in the UNIM applied to photons/electrons are estimated from low-energy alpha particle fluence response studies (11) . The basic idea of the UNIM is that the linear re- sponse at low dose arises from geminate recombin- ation in a localised entity. For gamma rays and electrons, the localised entity is the coupled/spatially correlated TC/LC. For HCPs, the localised entity is # The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com Radiation Protection Dosimetry (2014), pp. 1–11 doi:10.1093/rpd/ncu184 Radiation Protection Dosimetry Advance Access published June 10, 2014 at Ben Gurion University - Aranne Library on June 11, 2014 http://rpd.oxfordjournals.org/ Downloaded from