Solid materials can be classified according to a variety of criteria. Among the more significant of these is the description of a solid as being either crystalline or amorphous. The solid state physics community has tended during the period from the mid-1940's to the late 1960's to concentrate a much larger effort on crystalline solids than on the less tractable amorphous ones.

An amorphous solid exhibits a considerable degree of short range order in its nearest-neighbor bonds, but not the long range order of a periodic atomic lattice; examples include randomly polymerized plastics, carbon blacks, allotropic forms of elements such as selenium and antimony, and glasses. A glass may alternatively be thought of as a supercooled liquid in which the viscosity is too large to permit atomic rearrangement towards a more ordered form. Since the degree of ordering of an amorphous solid depends so much on the conditions of its preparation, it is perhaps not inappropriate to suggest that the preparation and study of amorphous solids has owed rather less to science and rather more to art than the study of crystalline materials. Intense study since the 1960s on glassy solids such as amorphous silicon (of interest for its electronic properties) is likely to create a more nearly quantitative basis for interpreting both electronic and structural features of noncrystalline materials.

In the basic theory of the solid state, it is a common practice to start with models of single crystals of complete perfection and infinite size.

Elementary concepts

The generally accepted picture of thermoluminescence outlined briefly in chapter 1 has its origins in the energy band theory of solids from which a simple explanation of the observed luminescence properties of various types of material can be obtained. It is, therefore, worthwhile to examine, in a rudimentary fashion, the energy band model of solids with a view to highlighting those characteristics which give rise to luminescence in general, and thermoluminescence in particular. Much of the modern-day theory on trapping, recombination and luminescence phenomena has remained unchanged since the original treatments provided by pioneers such as Shockley, Rose, Williams and others from the late 1940s to early 1960s. For purposes of explanation, I shall refer to several of these original works. More recent applications and confirmations of the theory will be treated in later sections and chapters of the book when specific examples are being discussed.

Energy bands and localized levels: crystalline materials

The solution of the Schrödinger equation for electrons subjected to a periodically varying potential reveals that the allowed energies for the electrons lie only in ‘allowed zones’. Other possible energy values constitute ‘forbidden zones’ or ‘band gaps’. This is the situation which arises in a solid where each atom is subject to a periodic array of potential wells, as in the Krönig-Penney model of a crystal.

In this chapter several other effects which are often observed in thermoluminescence experiments are discussed. Correct interpretation of these effects is important for a more complete understanding of the thermoluminescence data. To begin with some explanations of the phenomenon of supralinearity are examined as alternatives to the competing trap model detailed in the previous chapter.

Further discussions of supralinearity

Multi-stage reaction models

The supralinear growth curves described in figures 3.26 and 3.27 are characterized by a linear region, followed by a region of supralinear growth. In some cases, the opposite effect is seen in which the supralinear region precedes the linear part. Such behaviour is generally less common than that shown in these figures, but has been observed in several classes of material. An often-quoted example is the thermoluminescence from quartz extracts from pottery, for example figure 4.1 in which the low-dose supralinear region presents difficulties when using thermoluminescence as a method of age determination (see chapter 7). In general, supralinear growth can manifest itself in a growth curve in which the thermoluminescence intensity, I, increases as a function of Rl, where l is not necessarily one, nor is it necessarily an integer, and R is the absorbed dose.

Previous books dealing with thermoluminescence have tended to treat only specialist aspects of its usage – e.g., dating, dosimetry, geology or analysis of glow-curves. This is a reflection of the way in which thermoluminescence research has developed over the last decade; its study has become fragmented, each fragment being tailored to suit only the discipline to which it is being applied. As a result, the various disciplines have become somewhat insular, the advances in one arena not necessarily being transmitted to another. (For instance, one might find the occasional inappropiate application of different methods of analysis of glow-curves to problems associated with, say, geology or archaeology; or a lack of appreciation of the solid-state defect reactions which can take place in phosphors when thermoluminescence is applied to measure radiation dose.)

These facts do not appear to have imposed a limit on the rate of publication of papers dealing with thermoluminescence; several hundred articles on this topic appear each year. Considering this, perhaps the time is now appropriate for a pause in the accelerated use of thermoluminescence in order to reflect upon its capabilities and to become more aware of its limitations.

This book is intended as a step towards this goal, and towards the unification of the insular approaches by presenting the topic of thermoluminescence as a single subject. Extensive referencing of published works was felt to be an important requirement and as a result over 1000 references are listed.

General

As noted in chapter 1, the use of thermoluminescence in geology was one of the earliest applications of the phenomenon and both terrestrial and extraterrestrial (meteorite) rocks were examined for thermoluminescence before the beginning of the twentieth century. The natural thermoluminescence from a geological specimen depends, for its character and intensity, not just upon the nature of the phosphors involved, but also upon the dose received in nature, the period over which this dose is delivered, and the temperatures) experienced by the specimen during this time. Thus, the natural thermoluminescence contains locked within it information relating to the detailed thermal and radiation histories of the samples and unlocking this information has been the main goal of research into the natural glow from geological samples. Clearly the three parameters alluded to above, namely, dose, time and temperature, cannot be determined from a single measurement and the requisite detailed studies are not trivial. Information on kinetics, emission spectra, mineral types, radioactivity, dose response, sensitization, etc. has to be gathered if premature judgements are to be avoided.

A different ‘family’ of measurements which are often reported concerns the ‘artificial’ thermoluminescence, namely, that induced by a known dose given in a controlled environment in the laboratory. Measurements of this nature are usually carried out to derive data on the basic properties of the luminescent phosphor. Such data can be used to infer information about the source origin of the mineral, and, for example, its degree of metamorphism, etc.

General introduction

Among the list of uses for thermoluminescence outlined in chapter 1, the study of the defect structure of materials using this technique is perhaps the most difficult, in that interpretation of the results is not straightforward. In fact, on its own, thermoluminescence is only of limited value in arriving at a true characterization of the defect structure of the solid under study. Greatest benefit can be attained when the technique is used in conjunction with other experimental methods (dielectric loss, electrical conductivity, ITC/TSPC, optical absorption, electron spin resonance, etc.) in which case the information gained can be extremely useful. It should be realized, however, that these different experimental techniques may be sensitive to different types or collections of defects, so that conclusions derived from one method should be applied cautiously to the results obtained from another.

Some areas requiring special caution ought to be noted. The techniques mentioned above are normally used to give information about specific defects. If correlations can be found between various defect concentrations and certain thermoluminescence peaks, then such results are normally used to argue in favour of a particular defect being responsible for a particular peak. However, it is important to remember that thermoluminescence requires two principal defect types – a trap and a luminescent centre. In the techniques listed above it is not known which one, if either, of these defects is being monitored.

The application of the phenomenon of thermoluminescence to the measurement of absorbed radiation dose has progressed a great deal since the initial work by Daniels and colleagues referred to in chapter 1. Several thermoluminescent phosphors are now used routinely in many dosimetric applications. This chapter deals with the characteristics of thermoluminescence dosimetry (TLD) materials and several of their uses (environmental monitoring, personal dosimetry and medical applications). To begin with, it is useful to examine what properties are looked for when assessing the quality of a new TLD material.

General requirements for TLD materials

The selection of a phosphor for a thermoluminescence dosemeter requires a precise knowledge of the particular application being considered. Broadly, the applications may be conveniently listed as personal dosimetry (dose estimate in body tissue) and environmental Monitoring (dose estimate in air), with special considerations for medical applications and reactor dosimetry. Clearly there are large areas of overlap between each usage. Modern dosemeters are manufactured with specific applications in mind, for instance, measurement of the low-energy X-ray or β dose absorbed by the epidermis at a depth of 5 mg cm−2 (~ 50 μm) below the surface of the skin, or assessment of the dose-equivalent from a mixed neutron field absorbed by body tissue at a depth of 1000 mg cm−2. To make these measurements within an acceptable uncertainty limit, correct choice of dosemeter is essential.

Several recommended standards have been suggested for dosemeter performance, depending upon the application.

What is thermoluminescence?

Volume 84 of the Physics Abstracts Subject Index for 1981 lists over 300 separate entries for papers concerning the topic of thermoluminescence. Bräunlich (1979) states that there are over 500 articles published per year on phenomena relating to thermoluminescence (generally termed thermally stimulated relaxations). Newcomers to this field and experienced research workers alike may find such a high rate of publication surprising. This response may be reinforced when they realize that, as an experimental technique, thermoluminescence finds favour in such diverse scientific disciplines as archaeology, geology, medicine, solid-state physics, biology and organic chemistry, to name just some of the mainstream areas of study. The reader may justifiably enquire what it is about thermoluminescence that makes it such a well-used experimental method, enjoying widespread popularity and displaying enormous versatility. The answer to this enquiry is the essential theme addressed in this book. To answer the question fully it will be necessary to illustrate how the thermoluminescence characteristics of a material relate directly to the material's solid-state properties and how these solid-state properties are being utilized in the diverse fields mentioned above. However, before delving into the detail necessary to answer the question in full an initial approach to the problem must be made by stating what thermoluminescence is.

Thermoluminescence is the emission of light from an insulator or semiconductor when it is heated. This is not to be confused with the light spontaneously emitted from a substance when it is heated to incandescence.

Introduction

One of the prime objectives of a thermoluminescence experiment is to extract data from an experimental glow-curve, or a series of glow-curves, and to use these data to calculate values for the various parameters associated with the charge transfer process in the material under study. These parameters include the trap depths (E), the frequency factors (s), the capture cross-sections and the densities of the various traps and recombination centres taking part in the thermoluminescence emission. Of course, arriving at values for these parameters does not necessarily mean that we fully understand them, or that we are knowledgeable about the defect model with which they are associated. (Indeed, having calculated, say, E and s, there is often a temptation to ask ‘so what?’.) Nevertheless, calculations of this sort are an important step in arriving at an acceptable level of understanding of the underlying processes and a great deal of effort has been directed towards the development of a reliable method of analysis. Unfortunately, this development is proving to be less straightforward than was at first imagined.

The most popular procedure begins by selecting the rate equations appropriate to a particular model (cf. chapter 2) and continues by introducing simplifying assumptions into these equations in order to arrive at an analytical expression which describes the variation in thermoluminescence intensity with temperature, in terms of the desired parameters. From these equations even simpler expressions are produced which relate the parameters directly to the data.

Introduction

Modern thermoluminescence recording equipment can vary from the very simple to the extremely sophisticated. The nucleus of all the various designs is a light detection system, a sample heater and a temperature control unit but the designs of each of these components are many and varied. Enhanced sophistication in design is introduced if glow-curves below room temperature are required, if there is a need to record emission spectra at different glow-curve temperatures, or if simultaneous thermally stimulated current measurements are to be recorded. Extensive use of computer-controlled apparatus is becoming more and more popular and many research groups develop their own computer-based operating system. With the large numbers of microcomputers commercially available, an experimenter is faced with a wide choice of options. A typical schematic arrangement is shown in figure 9.1.

In the sections that follow, a general description of some of the necessary components for thermoluminescence detection will be given (although the exact details of the apparatus design will depend on the experimenter's individual requirements). In Appendix B, a list is given of the addresses of suppliers of commercial thermoluminescence equipment. Reference to published papers will be limited to those wherein the technique or apparatus described is particularly useful or novel, although the reader may wish to refer to two articles which deal with instrumentation for thermoluminescence in some depth – these are by Manche (1979) and by Julius (1981).