EVOLUTION OF THE TRAPPED CHARGE DISTRIBUTION DUE TO TRAP EMPTYING PROCESSES IN A NATURAL ALUMINOSILICATE J. M. Go ´ mez-Ros 1, , V. Correcher 1 , J. Garcı ´a-Guinea 2 and A. Delgado 1 1 CIEMAT, Av. Complutense 22, Madrid 28040, Spain 2 CSIC, Museo Nacional de Ciencias Naturales, C/o Jose ´ Gutierrez Abascal 2, 28006 Madrid, Spain The evolution of the thermoluminescence glow curve of a natural Ca–Be rich aluminosilicate after annealing treatments at different temperatures has been studied in order to evaluate the changes in the trapped charge distribution. The glow curve consists of a single broad peak that continuously shifts toward higher temperatures when the sample is preheated up to increasing temperatures, thus indicating the presence of a continuous trap distribution. The glow curve fitting assuming different distribution functions shows how a gaussian distribution becomes a nearly exponential distribution owing to the thermal leakage of charge carriers from trapping centres. INTRODUCTION Characterisation of the thermoluminescence (TL) emission of natural aluminosilicates is important for dosimetric purposes because of its ubiquitous presence both in nature and in ceramic materials that makes them especially interesting for applica- tions of TL in dose reconstruction and dating. The TL glow curve of natural aluminosilicates exhibit a complex structure commonly associated with a con- tinuous distribution of trapping levels (1–3) . Evidence for the presence of such energy trap distributions may be obtained from different experimental results, as it is the continuous shift in the position of a glow peak when linear preheatings of the sample up to increasing temperatures are performed before readout (4) . A trap distribution can be characterised by the distribution of trapped charges n(E ), assuming that there is no dependence on the frequency factor s, and the shape of the glow peak depends on the distribu- tion function n(E ). In this sense, changes in the trap distribution function from gaussian to exponential shape have been previously reported by other authors (5) in amorphous materials (glasses). These changes have been explained as the consequence of the thermal release of the trapped electrons from the lower energy levels when the sample is subjected to fading or thermal-cleaning processes. In this work, the TL emission of a Ca–Be rich aluminosilicate (Bavenite) has been studied to show that the single broad peak actually arises from a continuous distribution of trapping levels. A glow curve fitting procedure has been used to obtain the kinetic parameters in order to investigate how the distribution function is affected when thermal-cleaning treatments started emptying the shallower energy traps. MATERIALS AND METHODS Material and measurement conditions Bavenite [Ca 4 Be 2 Al 2 Si 9 O 26 (OH) 2 ] is one of the most efficient phosphor with a very high sensitivity to radiation exposure. Bavenite fibrous samples were collected in a large rock cavity hosted in a pegmatite body in the granite massif of Bustarviejo (Madrid, Spain). The crystal was studied under scanning electron microscopy (SEM) performed in a Philips XL20 SEM at accelerating voltages of 20–30 kV. Samples were coated with gold (20 nm) in a Bio- Rad SC515 sputter coating unit. Energy-dispersive X-ray microanalyses were obtained using a Phillips EDAX PV9900 with a light element detector type ECON. Bavenite appears as cotton-like white masses covering these hydrothermal minerals. Crystals show fragile prismatic-needle habits with a random 3-D orientation produced during the process of granite cutting and sample collection. (Figure 1). The TL glow curves were obtained using an auto- mated Risø TL system model TL DA-12 (6) , this reader is provided with an EMI 9635 QA photomul- tiplier and the emission was observed through a blue filter (a FIB002 of the Melles-Griot Company) where the wavelength is peaked at 320–480 nm; FWHM is 80  16 nm and peak transmittance (minimum) is 60%. All the TL measurements were performed using a linear heating rate of 5  Cs 1 from room temperature up to 550  C in a N 2 atmosphere. Four aliquots of 5.0  0.1 mg each of Bavenite were used for each measurement. The sample was carefully powdered with an agate pestle and mortar to avoid triboluminescence (7) . The incandescent background was subtracted from the TL data.  Corresponding author: jm.gomezros@ciemat.es Radiation Protection Dosimetry (2006), Vol. 119, No. 1–4, pp. 93–97 doi:10.1093/rpd/nci522 Advance Access published on May 18, 2006 Ó The Author 2006. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org