TEMPERATURE DISTRIBUTION IN THERMOLUMINESCENCE EXPERIMENTS .2. SOME CALCULATIONAL MODELS

This is the second of a linked pair of papers on the subject of thermoluminescence experiments; it is primarily about calculational models, while the first (preceding) paper describes experimental results. In a typical thermoluminescence experiment, a sample rests on a metallic strip which is heated in a controlled fashion so that the strip temperature rises linearly with time. Thermal contact is improved by the use of an inert exchange gas, usually argon. With this procedure, samples of interest emit light spectra of low intensity as electrons escape from traps. The technique is applied, for example, to dating of artefacts or geological materials, to radioactive dosimetry, and to the characterization of optical materials. In this paper we consider some situations for which exact solutions of the heat conduction equation can be obtained. Horizontal temperature distributions in the heating strip are dealt with, as is, importantly, the time-dependence of the temperature within, or at the top of, a sample of finite thickness resting on the heater strip. We show that, beyond a short initial transient, the temperature variation caused at a point in the sample by a linear ramp of the strip is of the form beta(p)t - delta(p) where delta(p) represents a constant temperature lag and beta(p) is a rate which may be smaller (but not greater) than the ramping rate of the strip. Delta(p) and beta(p) are calculated, for a plausible model, in terms of heat transfer parameters defined in the paper, and these can be measured. These effects are important where data from different laboratories are compared and we conclude that practitioners should routinely take note of the following points. Firstly, ramping rates should be as low as possible consistently with the optical sensitivity available (there is a trade-off between the two, but temperature differences are likely to be important where ramping rates exceed 5 K s-1). Secondly, it is advisable to use helium as exchange gas rather than the more usual argon, because it is a much better conductor of heat. Thirdly, users should do a few basic thermal experiments with their apparatus so that, with the aid of formulae given in this paper, they can make corrections where these are important.