On the Need for Deconvolution Analysis of Experimental and Simulated Thermoluminescence Glow Curves

Abstract

Simulation studies of thermoluminescence (TL) and other stimulated luminescence phenomena are a rapidly growing area of research. The presence of competition effects between luminescence pathways leads to the complex nature of luminescence signals, and therefore, it is necessary to investigate and validate the various methods of signal analysis by using simulations. The present study shows that in simulations of luminescence signals originating from multilevel phenomenological models, it is not possible to extract mathematically the individual information for each peak in the signal. It is further shown that computerized curve deconvolution analysis is the only reliable tool for extracting the various kinetic parameters. Simulation studies aim to explain experimental results, and therefore, it is necessary to validate simulation results by comparing with experiments. In this paper, testing of simulation results is performed using two methods. In the first method, the influence of competition effects is tested by comparing the input model parameters with the output values from the deconvolution analysis. In the second method, the agreement with experimental results is tested using the properties of well-known glow peaks with very high repeatability among TL laboratories, such as the 110 °C glow peak of quartz.

Keywords: competition between levels; computerized glow curve deconvolution; kinetic parameters; stimulated luminescence; superposition principle; thermoluminescence.

Figure 4

CGCD analysis results of the four REFERENCE glow curves used for simulation. (a) REFERENCE-01, (b) REFERENCE-02, (c) REFERENCE-03, (d) REFERENCE-04.

Figure 1

Traditional methods to extract the TL intensity from experimental glow curves either by selecting the peak height at some point of the glow curves (perpendicular lines) or by integrating the signal between two temperatures (region within the box).

Figure 2

Energy band model used for simulation. (Upper): Irradiation stage. (Middle): Relaxation stage. (Down): Heating stage.

Figure 3

TL peak shape in the case of REFERENCE-01 glow curve, as evaluated from the numerical solutions of the model in Section 2.1. (a) Shapes of peaks 1, 2 and 3, (b) shape of peak 4, (c) shape of peak (5) and (d) total glow curve shape. It is obvious that except peak 1, all other peaks, although they must be of single-peak shape, look composite, because the electron distribution within the conduction band contributes to their numerical evaluation.

Figure 5

(a) Histogram for the position of peak maximum temperature, (b) histogram for the values of the activation energy and (c) histogram of the logarithm of the frequency factor.