Thermal transfer and apparent-dose distributions in poorly bleached mortar samples: Results from single grains and small aliquots of quartz

In the assessment of doses received from a nuclear accident, considerable attention has been paid to retrospective dosimetry using the optically stimulated luminescence (OSL) of heated materials such as bricks and tiles. Quartz extracted from these artefacts was heated during manufacture;, this process releases all the prior trapped charge and simultaneously sensitises the quartz. Unfortunately unheated materials such as mortar and concrete are more common in industrial sites and particularly in nuclear installations. These materials are usually exposed to daylight during quarrying and construction, but in general this exposure is insufficient to completely empty (bleach) any geological trapped charge. This leads to a distribution of apparent doses in the sample at the time of construction with only some (if any) grains exposed to sufficient light to be considered well bleached for OSL dosimetry. The challenge in using such materials as retrospective dosemeters, is in identifying these well-bleached grains when an accident dose has been superimposed on the original dose distribution.

We investigate here, using OSL, the background dose in three different mortar samples: render, whitewash and inner wall plaster from a building built in 1964. These samples are found to be both poorly bleached and weakly sensitive (only 0.3% of grains giving a detectable dose response). We study thermal transfer in single grains of quartz, investigate the grain-size dependence of bleaching in the size range 90-300 mum and compare the dose-distributions obtained from small aliquots and single-grain procedures. A comparison of three different methods viz. (a) first 5%, (b) probability plot and (c) comparison of internal and external uncertainties, is made for equivalent dose estimation. The results have implications for accident dosimetry, archaeological studies and dating of poorly bleached sediments. (C) 2003 Elsevier Ltd. All fights reserved.