Characterisation of scintillator-based gamma spectrometers for determination of sample dose rate in OSL dating applications

A recent inter-comparison (Murray et al., 2015) has helped to highlight variability in the measurement of dose rate between luminescence laboratories. Part of this variability probably reflects the difficulties of homogenising and dissolving samples so that the < 500 mg used in e.g. ICP-MS and NAA is representative. High resolution gamma spectrometry is the obvious alternative because it can measure samples 100-1000 times larger, but the instrumentation is low-throughput, high capital and running cost, and requires skilled personnel to maintain operation over many years. Here we investigate the potential of traditional low-cost, low maintenance alternatives based on a 3 '' x 3 '' NaI(Tl) scintillation crystal. The temperature stability is investigated, and a linear (with intercept) correction for spectra drift based on the 1.46 MeV peak from K-40 and the similar to 100 keV composite X-ray peak from uranium and thorium is shown to minimise this problem. Using a calibration based on wax impregnated standards, the minimum detection limits (MDL) are 25 Bq/kg (K-40), 4.8 Bq/kg (U-238), 2.5 Bq/kg (Th-232) for 250-300 g of sample; systematic deviations around the expected values are also shown to be acceptable as the MDL is approached. Finally, we compare the activity concentrations and resulting dry dose rates derived from our NaI-based system with those from routine high resolution gamma spectrometry, and conclude that the new analytical facility is very suitable for accurate and precise dose rate determination.