Factors determining the thermoluminescence of chronologically significant materials

Thermoluminescence has been successfully applied to the dating of pottery from archaeological sites and the good agreement obtained between the TL dates and the archaeologically known dates suggested that the basic assumptions concerning the build up and release of trapped charges were correct. In this thesis I report the results of a test programme I carried out on recent volcanic lava of known age: the TL ages were all far too young and hence it was apparent that at least one of the assumptions used in the dating of pottery was invalid when applied to lava. The most obvious differences between pottery and lava dating were studied, such as the effect of crushing such a hard material to obtain fine grains and the influence of the known occurrence of radioactive disequilibrium in recent lavas, but the cause of the low ages was found to be the breakdown of a far more basic assumption. The assumption that electrons in deep traps are capable of remaining in the traps over archaeological time is the basic tenet upon which the whole of TL dating is based. In the case of lava it is shown in this thesis that this does not hold; there was a loss of high temperature TL during storage at room temperature for a few hours. This effect was called 'anomalous fading' as it is in disagreement with thermal untrapping of electrons predicted by kinetic theory and widely accepted in the literature. The third part of this thesis describes the experiments carried out in an attempt to elucidate this phenomenon which was found to be exhibited by a variety of minerals that would normally be considered suitable for the application of TL dating. The same phenomenon has since been reported in the study of the TL of lunar samples and similar effects in other materials are also reported in the final chapter. Two models are suggested which are consistent with the experimental evidence accumulated. They both involve the tunnelling of an electron at low temperatures but differ in their explanation of the thermal dependence observed in experiments at room temperature. The implications of these models with regard to testing for anomalous fading are discussed and it is concluded that, when contemplating the application of TL dating to a new material, a test programme should be carried out on samples of known age and, furthermore, the agreement of the archaeological stability of electrons with that predicted by kinetic analysis of the TL peaks should also be demonstrated. Kinetic analysis carried out on one of the high temperature peaks in limestone agreed with the stability suggested by the geological age of the sample. This confirmed the absence of anomalous fading in this mineral as had been suggested by short-term fading tests. In similar studies on quartz the archaeological stability was found to be greater than that predicted by kinetic analysis using the initial rise method. This discrepency is shown to be due to the inapplicability of the initial rise method to samples in which thermal quenching is occurring. In the past the initial rise has been the most frequently used method of trap depth determination because of its independence of the untrapping kinetics; what has been completely overlooked is that in thermoluminescence the signal will also be affected by any temperature dependence of the luminescence centres. Once this had been allowed for the predicted stability again confirmed the absence of anomalous fading in quartz and hence its suitability for dating.