Investigating the Role of Charge Imbalance on the Luminescence Production of Quartz and Feldspar

Abstract

Luminescence dosimetry is now routinely employed in accident dosimetry and geochronological and archaeological studies, using natural minerals such as quartz and feldspar as dosimeters. Quartz and feldspar are the most common minerals in the regolith and can be used to record events related to such diverse phenomena as climate and landscape change, tectonic activity, catastrophic events, and human evolution/migration. Since luminescence dosimetry relies on the relationship of trapped charge and dose, and the recombination of trapped electrons with trapped holes, it is important to know how charge trapping relates to dose, and how trapping and recombination occur in nature and in the laboratory. Dose distributions derived from measurements of luminescence from individual grains are often used to investigate whether a sample was completely reset or zeroed before final burial. However, the resulting dose distributions are almost always wider than can be explained by known measurement uncertainties. Some part of this over-dispersion may arise from differences in irradiation geometry at a scale of individual grains. In particular, there may be grain-to-grain variations in the charge balance. This thesis investigates whether a net excess of electrons or holes exists as a result of natural and/or laboratory irradiation, and whether such a charge imbalance can have an influence on the luminescence production of quartz. Initial results showed that the over-dispersion of single grains increases with an increase in atomic number of the backscattering material. This coincides with an increase in dose rate resulting from an increased contribution from the backscattered spectrum. For a given backscattering material, degrading the incoming beta spectrum by passing it through an attenuator also leads to an increase in the observed over-dispersion. Both results indicate that low energy electrons may play a key role in over-dispersion. Surprisingly, it is also observed that feldspar experiences a ∼15% lower apparent beta source dose rate than quartz. This is despite their very similar composition; modelling shows that the two minerals have indistinguishable dose deposition when exposed to either beta or gamma radiation. It is concluded that the difference in apparent dose rates is most likely to arise from differences in charge trapping and recombination, rather than the actual dose deposition, although the details of this remain unknown. Charge imbalance during natural and laboratory beta and gamma irradiation is then investigated using radiation transport modelling. As far as is known, this is the first time this phenomenon has been considered in luminescence dosimetry. Simulations show that quartz grains acquire a net charge during external irradiations with charged particles; these modelling predictions form the basis for experiments investigating the impact of excess electrons at very large (hundreds of kGy to MGy) doses. As predicted, a depression of both luminescence production and sensitivity is observed with increasing dose from a 200 keV electron beam. This observation is attributed to the recombination by excess electrons of the trapped hole populations giving rise to luminescence. This conclusion is supported by further experiments comparing the effects of irradiating with a lower (100 keV) electron energy. In addition, these electron beam experiments suggest that the excess electron concentrations at high doses increase to the point that they generate electric fields sufficient to repel incoming electrons; this reduces and ultimately prevents further absorption of dose by the sample. Because luminescence is a two stage process (electron detrapping and recombination) it does not give direct information on the behaviour of the trapped electron population alone. To allow investigations to study electron and hole populations separately, a new detector was designed and constructed to measure the electrons which leave the surface of individual grains; these electrons are known to derive from the same traps as the luminescence signals. The design and testing of the instrument is described, and preliminary results presented which show that it is capable of useful measurements of electrons emitted by a single quartz grain. Unfortunately, there was insufficient time to make use of this new detector to investigate over-dispersion, and so these preliminary results are included in an appendix.