Radiation Measurements 133 (2020) 106281 Available online 24 February 2020 1350-4487/© 2020 Elsevier Ltd. All rights reserved. A new method to predict the response of thermoluminescent detectors exposed at different positions within a clinical proton beam Alessio Parisi a, * , Pawel Olko b , Jan Swako� n b , Tomasz Horwacik b , Hubert Jabło� nski b , Leszek Malinowski b , Tomasz Nowak b , Lara Struelens a , Filip Vanhavere a a Belgian Nuclear Research Centre SCK CEN, Mol, Belgium b Institute of Nuclear Physics, Polish Academy of Sciences (IFJ PAN), Krakow, Poland A R T I C L E INFO Keywords: Proton therapy Thermoluminescent detectors Microdosimetric d(z) model ABSTRACT The proton depth dose profile measured by luminescent detectors differs from the one measured with reference dosimeters (i.e. ionization chambers) because of several effects including efficiency quenching and changes in the attenuation of the emitted light in case of partial detector irradiation. Using the Microdosimetric d(z) Model in combination with the Monte Carlo radiation transport code PHITS, a methodology was developed to tackle all these factors and calculate the response of luminescent when exposed at different positions along a proton Bragg peak. The results were compared against experimental data gathered with 7 LiF:Mg,Ti (MTS-7) and 7 LiF:Mg,Cu,P (MCP-7) thermoluminescent detectors, showing a very good agreement (average relative deviation ~ 3% for both detector types, smaller than the combined experimental uncertainty). 1. Introduction It is well known that the response of luminescent detectors depends strongly on the type of detector, the amount of dose imparted and the radiation quality to be measured. Due to a saturation of the luminescent trapping centers, the efficiency of these detectors is characterized by a general decrease with the increase of the LET of the incident particle. Additionally, it was noticed that the efficiency in measuring two different particles with the same LET is different, with higher values for the heavier particle (Berger and Hajek, 2008). This happens because of their different microscopic pattern of energy deposition, being more energetic the δ-ray spectrum liberated by the heavier particle because of its higher speed (Olko, 2007). To this regard, the Microdosimetric d(z) Model (Parisi, 2018; Parisi et al., 2018b) is a recently developed model able to describe and predict the relative luminescence efficiency of luminescent detectors exposed to different radiation qualities by analyzing their stochastic energy depo- sition at the nanoscale. The model was applied to assess the efficiency in case of LiF:Mg,Ti (MTS) and LiF:Mg,Cu,P (MCP) thermoluminescent detectors for monoenergetic charged particles from 1 H to 132 Xe (simu- lated energy range: 3–1000 MeV/u, Parisi et al., 2017c, 2017d, 2018b) and photons (simulated energy range ¼ 10–1250 keV, Parisi et al., 2019b). For the whole particle and energy ranges, a very good agreement with experimental data was found in case of model calcula- tions performed in a site size of 40 nm. Furthermore, using the Microdosimetric d(z) Model, a novel method to assess average LET quantities and relative biological effectiveness (RBE) in proton therapy beams was developed and successfully bench- marked against Monte Carlo computer simulations (Parisi et al., 2019a , 2019c) and an in vitro colonial survival study (Parisi et al., 2019a). Also in this case, efficiency-vs-LET calibration curves were a priori deter- mined by coupling the Microdosimetric d(z) Model with specific energy density distributions simulated with the radiation transport code PHITS (Sato et al., 2018) for monoenergetic beams. However, the radiation fields characterizing the hadron therapy and space environments are composed by a large variety of different parti- cles with broad energy distributions. Consequently, efficiency calcula- tions using the Microdosimetric d(z) Model should consider this increased complexity by taking into account both the primary and the secondary particle-energy spectra and their interaction with a realistic model of the detectors. To this aim, in order to investigate the applica- bility of the Microdosimetric d(z) Model also to not-monochromatic and multi-particle exposure conditions, in this paper a new methodology to predict the response of 7 LiF:Mg,Ti (MTS-7) and 7 LiF:Mg,Cu,P (MCP-7) thermoluminescent detectors exposed at different positions along a proton Bragg peak is presented and validated against experimental * Corresponding author. E-mail address: alessio.parisi@sckcen.be (A. Parisi). Contents lists available at ScienceDirect Radiation Measurements journal homepage: http://www.elsevier.com/locate/radmeas https://doi.org/10.1016/j.radmeas.2020.106281 Received 28 October 2019; Received in revised form 6 January 2020; Accepted 18 February 2020