LUMINESCENCE SENSITISATIONS OF NATURAL QUARTZ USING PRE-EXPOSURE DOSE AND THERMAL ACTIVATION TECHNIQUES EBENEZER OLUBANJI ONIYA UNIVERSITY OF IBADAN LIBRARY Luminescence Sensitisation of Natural Quartz Using Pre-Exposure Dose and Thermal Activation Techniques By Oniya, Ebenezer Olubanji (Matric No: 118223) B. Sc (Hons) Physics (Ado Ekiti) M. Sc. (Ibadan) A thesis in the Department of Physics Submitted to the Faculty of Science In partial fulfillment of the requirements for the Degree of Doctor of Philosophy of the University of Ibadan DECEMBER 2014 ii UNIVERSITY OF IBADAN LIBRARY CERTIFICATION I certify that the work described in this thesis was carried out under my supervision by Oniya Ebenezer Olubanji (118223) in the Department of Physics, University of Ibadan, Nigeria. …………………………… SUPERVISOR I. A. Babalola B.Sc, Ph.D (Ibadan) Professor, Department of Physics, University of Ibadan, Nigeria. iii UNIVERSITY OF IBADAN LIBRARY DEDICATED TO GOD: My Alpha and Omega who always turn impossibilities to possibilities in my life. My Father and Mother: Who made my coming to this world achievable. Dupe: My love, for your patience, prayers, supports and encouragements. Ini, Fise and Temmy: For all those times Daddy was away or too busy. iv UNIVERSITY OF IBADAN LIBRARY ACKNOWLEDGEMENTS Words are inadequate to express my appreciations to God for the grace to complete this work against all the odds. I can say it loud that “weeping may endure for a night but joy cometh in the morning”. The way God has demonstrated HIS sovereignty and supremacy throughout the duration of study made it settled within me that any project that belongs to Almighty God will surely be accomplished. My Father and Redeemer, I thank you. With pleasure, I express my immense gratitude to my supervisor, Prof. I. A. Babalola, and Dr N. N. Jibiri, who was co-opted to see to the completion of the work. Despite their tight schedules they patiently and diligently guided my thoughts. God will protect you for this kind gesture and efforts. I am equally grateful and can not but say a BIG thank you to Dr F. O. Ogundare who initiated this work and introduced me to luminescence world. I will continue to remember you as my mentor and academic tutor. I wish to express my profound gratitude to Prof I. P. Farai, I. R. Obed, Dr A. A. Adetoyinbo, Dr (Mrs.) A. Ademola, Dr Adegoke, Dr Awe, Dr Ogunsola, Dr Otunla and the entire academic staff of Physics Departments University of Ibadan for their cooperation at various times. I say special thanks to Prof. G. Kitis of Nuclear Physics Laboratory, Aristotle University of Thessaloniki, Greece and Dr. G. S. Polymeris of Institute of Nuclear Sciences, Ankara University, Turkey for their technical and academic assistance. I also register my profound appreciation to Dr. N. C. Tsirliganis and the entire member of staff of the Research Center „Athena‟ (Athena R.C.) Greece, for the assistances and hospitality shown to me during my stay in their laboratory where I carried out the bench work. I also express my untold appreciations to Prof I. R. Ajayi, Prof. N. O. Ajayi, Dr R. S. Fayose and all the entire member of staff of Physics and Electronics Department, Adekunle Ajasin University Akungba Akoko for their diverse encouragements and assistances rendered. I sincerely acknowledge my colleagues in the struggle, Dr. Benson Igboin, Dr. E. A. Afe, Dr. Busuyi Mekusi, Dr. Ayuba, Mrs. Igili and Dr. O. Olubosede for their wonderful supports, prayers and encouragements. v UNIVERSITY OF IBADAN LIBRARY I am specially grateful to Adekunle Ajasin University and Education Trust Fund (ETF) Nigeria for the financial support offered in the framework of “2008 Education Trust Fund Academic Staff Training and Development” that was responsible for my sponsorship to Greece where I carried out the experimental work of this study. I wish to acknowledge with due respect, Late Brother and Sister S. A. Adeyemi, Brother and Sister O. Ogunleye and Brother and Sister S. Balogun, Brother Paul Olpha and the leaders of Apostolic Faith, Youth Directorate Development, Ondo District for their great concern and prayers for the success of the research work. This acknowledgment can be complete without appreciating my beloved wife, Dupe Oniya for the patience, tolerance and understanding she displayed during the course of this research. Your prayers and moral supports are acknowledged and appreciated, I love you my love. Lastly, my immense appreciation goes to all that contributed to the success of the study, including you. Thank you all and God bless you. vi UNIVERSITY OF IBADAN LIBRARY TABLE OF CONTENTS CONTENTS Page Fly-Leaf i Title ii Certification iii Dedication iv Acknowledgements v Table of contents vii List of Tables xi List of Figures xii Abstract xvii Notations and Symbols xix CHAPTER ONE: INTRODUCTION 1 1.1 Background 1 1.2 Justification of present study 3 1.3 Aims and objectives of the study 4 1.4 Thesis outlines 5 CHAPTER TWO: LITERATURE REVIEW 7 2.1 Luminescence phenomena 7 2.2 Energy band model 8 2.3 Thermoluminescence 10 2.3.1 TL kinetic expressions 10 2.3.2 Glow curve 13 2.3.3 Emission spectrum 16 2.4 Optically stimulated luminescence 19 2.4.1 CW-OSL 19 2.4.2 LM-OSL 24 2.5 Dose response 25 2.6 Luminescence sensitivity 26 vii UNIVERSITY OF IBADAN LIBRARY 2.6.1 Competitions during irradiation and stimulation 28 2.6.2 Luminescence sensitisation 32 2.6.2.1 Pre-dose sensitisation 32 2.6.2.1.1 Pre-dose model 33 2.6.2.1.2 Radiation quenching 35 2.6.2.1.3 UV reversal 35 2.6.2.2 Thermal sensitisation 36 2.6.2.2.1 Thermal sensitisation model 37 2.6.2.3 Thermal activation curve 39 2.6.2.4 Heating rate effects 41 2.6.2.5 Thermal quenching 41 2.6.2.6 Temperature lags 44 2.6.2.7 Feldspar inclusion 45 2.7 Computerised curves deconvolution. 45 2.8 Retrospective dosimetry 47 2.9. TL/OSL reader 48 2.9.1 The risø automated TL/OSL reader 49 2.9.1.1 Heating system 51 2.9.1.2 Optical stimulation system 51 2.9.1.2.1 Blue LEDs 53 2.9.1.2.2 Infrared LEDs 53 2.9.1.3 Photon detector system 54 2.9.1.3.1 Photomultiplier tube 54 2.9.1.3.2 Detection filters 54 2.9.1.4 Beta irradiator 55 viii UNIVERSITY OF IBADAN LIBRARY CHAPTER THREE: MATERIALS AND METHODS 3.1 Samples collection 59 3.2 Samples preparation 59 3.3 Instrumentation 60 3.4 Feldspar inclusion test 60 3.5 Experimental procedures 60 o 3.5.1 Reproducibility study of pre-dose sensitisation of the 110 C TL peak 63 3.5.1.1 Measurement protocols 63 3.5.1.2 Description of the protocol 63 3.5.2 Study on luminescence sensitisations in unannealed and annealed quartz samples 64 3.5.2.1 Measurement protocol 64 3.5.2.2 Descriptions of the protocol 66 3.6. Computerised curves deconvolution analyses 67 CHAPTER FOUR: RESULTS AND DISCUSSIONS 4.1 Introduction 69 4.2. Reproducibility study of pre-dose sensitisation o of the 110 CTL peak. 69 4.2.1. Sensitisation in unannealed samples 69 4.2.2. Sensitisation in annealed quartz 78 4.2.3. Discussion 79 4.2.4. Implications of results 84 4.3. Study on luminescence sensitisations in unannealed and annealed quartz samples 85 o 4.3.1. Sensitisations of 110 C TL peak and RT-LMOSL of unannealed and annealed samples 85 ix UNIVERSITY OF IBADAN LIBRARY o 4.3.2. Dependence of 110 C TL peak and RT-LMOSL sensitisations on heating rate 100 4.3.3 Dependence of various components of RT-LMOSL sensitisations on heating rate of thermal activation 100 o 4.3.4 Dependence of 110 C TL peak and RT LM-OSL sensitisationson TL activation histories 113 4.3.5. Discussion 122 4.3.6. Implications of results 148 CHAPTER FIVE: CONCLUSIONS AND RECOMMENDATIONS 5.1. Conclusion 149 5.2. Recommendations 150 REFERENCES 151 APPENDIX: PUBLISHED ARTICLES 161 Appendix 1 161 Appendix 2 167 x UNIVERSITY OF IBADAN LIBRARY LIST OF TABLES Table Page 4.1. Coefficient of Variation of sensitization of 10 aliquots of S2 and S4 quartz samples. 77 4.2 Percentage sensitisation signal of each component to the total fitted data 99 4.3 Relative sensitisation factor of aliquots with pre-exposure dose and those without pre-exposure dose 135 xi UNIVERSITY OF IBADAN LIBRARY LIST OF FIGURES Figures Page 2.1 Simple band model illustrating TL and OSL processes 9 2.2 Typical TL glow curve representing the photons released during the recombination at luminescence centres 14 2.3 Schematic comparison of TL glow peaks for first- and second-order kinetics. 15 o 2.4 Emission spectra at 190 – 210 C of quartz extracted from sediments 17 2.5 Three-dimensional emission spectrum 18 2.6 Typical OSL decay curve from a sedimentary quartz 20 2.7 An example of exposing a heated and dosed quartz sample to linearly increasing blue LED 21 2.8 OSL curves of first and second order 23 2.9 Dose response of the TL peaks (P1, P2, P3 and P4) of Brazilian natural quartz 27 2.10 Transitions taking place during excitation stage 29 2.11 Schematic competition of free charges among the electron traps during early state of heating 30 2.12 Schematic competition of free charges among the electron traps during late state of heating 31 2.13 Pre-dose sensitisation schematic band model 34 2.14 Thermal sensitisation schematic band model 38 2.15 TAC of Fleming and Thompson (1970) redrawn‟ 40 o 2.16 Experimental glow-peak shapes of the 110 C TL peak of Norwegian quartz 42 2.17 Schematic diagram of the Riso TL/OSL luminescence reader 50 2.18 Schematic diagram of the combined blue and IR LED OSL unit 52 2.19 The emission spectrum of blue LEDs 56 2.20 Emission spectra of sedimentary quartz and K 57 2.21 Schematic diagram of the cross section of the beta irradiator 58 3.1 Luminescence sensitivity of all the samples to 5Gy dose of radiation 61 3.2 Infra-Red Stimulation Luminescence (IRSL) curves of all the samples to test for feldspar contamination 62 xii UNIVERSITY OF IBADAN LIBRARY 4.1 TL glow curves for unannealed S2 samples 70 4.2 TL glow curves for unannealed S4 samples 71 4.3 Sensitisations resulting from 3 successive irradiations and luminescence readings for 10 aliquots for unannealed S2 sample 72 4.4 Sensitisations resulting from 3 successive irradiations and luminescence readings for 10 aliquots for unannealed S4 sample. 73 4.5 TL test of two runs for S2 samples 75 4.6 TL test of two runs for S4 samples. 76 4.7 TL glow curves for annealed S2 samples. 80 4.8 TL glow curves for annealed S4 samples. 81 4.9 Sensitisations resulting from 3 successive rradiations and luminescence readings for 10 aliquots for annealed S2 sample 82 4.10 Sensitisations resulting from 3 successive irradiations and luminescence readings for 10 aliquots for annealed S4 sample. 83 4.11 Glow curves showing sensitisations resulting from successive irradiations and TL readings of unannealed S2 sample 86 4.12 Glow curves showing sensitisations resulting from successive irradiations and TL readings of annealed S2 sample. 87 4.13 Glow curves showing sensitisations resulting from successive irradiations and TL readings of unannealed S4 sample. 88 4.14 Glow curves showing sensitisations resulting from successive irradiations and TL readings of annealed S4 sample. 89 4.15 OSL curves showing sensitisations resulting from successive irradiations and LMOSL readings of unannealed S2 sample. 90 4.16 OSL curves showing sensitisations resulting from successive irradiations and LMOSL readings of annealed S2 sample. 91 4.17 OSL curves showing sensitisations resulting from successive irradiations and LMOSL readings of unannealed S4 sample. 92 4.18 OSL curves showing sensitisations resulting from successive irradiations and LMOSL readings of annealed S4 sample. 93 4.19 RT-LMOSL curves depicting deconvolution of the LMOSL curves to its respective components unannealed S2 sample 94 4.20 RT-LMOSL curves depicting deconvolution of the LMOSL curves to its respective components unannealed S2 sample 95 xiii UNIVERSITY OF IBADAN LIBRARY 4.21 RT-LMOSL curves depicting deconvolution of the LMOSL curves to its respective components annealed S2 sample 96 4.22 RT-LMOSL curves depicting deconvolution of the LMOSL curves to its respective components annealed S4 sample 97 4.23 Plots of TL sensitisations against heating rates as function of cycle of measurements for unannealed S2 samples. 101 4.24 Plots of RT-LMOSL sensitisations against heating rates as function of cycle of measurements for unannealed S2 samples. 102 4.25 Plots of TL sensitisations against heating rates as function of cycle of measurements for annealed S2 samples. 103 4.26 Plots of RT-LMOSL sensitisations against heating rates as function of cycle of measurements for annealed S2 samples. 104 4.27 Plots of TL sensitisations against heating rates as function of cycle of measurements for unannealed S4 samples. 105 4.28 Plots of RT-LMOSL sensitisations against heating rates as function of cycle of measurements for unannealed S4 samples. 106 4.29 Plots of TL sensitisations against heating rates as function of cycle of measurements for annealed S4 samples. 107 4.30 Plots of RT-LMOSL sensitisations against heating rates as function of cycle of measurements for annealed S4 samples. 108 th 4.31 Plots of 4 sensitised of RT-LMOSL components sensitisations against heating rates as function of cycle of measurements for unannealed S2 samples 109 th 4.32 Plots of 4 sensitised of RT-LMOSL components sensitisations against heating rates as function of cycle of measurements for unannealed S4 samples 110 th 4.33 Plots of 4 sensitised of RT-LMOSL components sensitisations against heating rates as function of cycle of measurements for annealed S2 samples. 111 th 4.34 Plots of 4 sensitised of RT-LMOSL components sensitisations against heating rates as function of cycle of measurements for annealed S4 samples 112 4.35 Comparison of TL sensitisations in aliquots with pre-exposure dose and those with thermal activation for unannealed S2 sample 114 xiv UNIVERSITY OF IBADAN LIBRARY 4.36 Comparison of TL sensitisations in aliquots with pre-exposure dose and those with thermal activation for annealed S2 sample 115 4.37 Comparison of RT-LMOSL sensitisations in aliquots with pre-exposure dose and those with thermal activation for unannealed S2 sample 116 4.38 Comparison of RT-LMOSL sensitisations in aliquots with pre-exposure dose and those with thermal activation for annealed S2 sample 117 4.39 Comparison of TL sensitisations in aliquots with pre-exposure dose and those with thermal activation for unannealed S4 sample. 118 4.40 Comparison of TL sensitisations in aliquots with pre-exposure dose and those with thermal activation for annealed S4 sample 119 4.41 Comparison of RT-LMOSL sensitisations in aliquots with pre-exposure Dose and those with thermal activation for unannealed S4 sample 120 4.42 Comparison of RT-LMOSL sensitisations in aliquots with pre-exposure dose and those with thermal activation for annealed S4 sample 121 o 4.43 Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots with pre-exposure dose for unannealed S2 sample 123 o 4.44 Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots without pre-exposure dose for unannealed S2 sample 124 o 4.45 Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots with pre-exposure dose for annealed S2 sample 125 o 4.46 Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots with pre-exposure dose for annealed S4 sample 126 o 4.47 Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots with pre-exposure dose for unannealed S4 sample 127 o 4.48 Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots without pre-exposure dose for unannealed S4 sample 128 o 4.49 Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots with pre-exposure dose for annealed S4 sample 129 o 4.50 Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots without pre-exposure dose for annealed S4 sample 130 o 4.51 Plots of normalized 110 C TL peak sensitisations against heating rates as function of cycle of measurements for unannealed S2 sample 132 o 4.52 Plots of normalized 110 C TL peak sensitisations against heating rates as function of cycle of measurements for unannealed S4 sample 133 xv UNIVERSITY OF IBADAN LIBRARY 4.53 Comparison of RT-LMOSL component sensitisations in aliquots with pre-exposure dose and those with thermal activation for unannealed S2 sample 136 4.54 Comparison of RT-LMOSL component sensitisations in aliquots with pre-exposure dose and those with thermal activation for unannealed S4 sample 137 4.55 Comparison of RT-LMOSL component sensitisations in aliquots with pre-exposure dose and those with thermal activation for annealed S2 sample 138 4. 56 Comparison of RT-LMOSL component sensitisations in aliquots with pre-exposure dose and those with thermal activation for annealed S4 sample 139 4.57 “OSL glow curve” for unannealed S2 141 4.58 “OSL glow curve” for annealed S2 142 4.59 “OSL glow curve” for unannealed S4 143 4.60 “OSL glow curve” for annealed S4 144 xvi UNIVERSITY OF IBADAN LIBRARY ABSTRACT Luminescence sensitisation is an important stage in the application of quartz in pre- dose retrospective dosimetry and dating. The existing techniques for sensitisation in quartz are Pre-Exposure Dose (PED) and Thermal Activation (TA). Previous works were centred on combined actions of PED and TA with less attention given to the separate contributions of these techniques in pre-dose sensitisation of quartz. This work was undertaken to determine the separate contributions of PED and TA in sensitisations o of 110 C Thermoluminescence (TL) peak and Room Temperature Linearly Modulated Optically Stimulated Luminescence (RT-LMOSL) of quartz. Two sets of quartz samples, one from Oro in Kwara State (S2) and the other from Ijero- Ekiti, Ekiti State (S4) with high sensitisation signals were used. Each of the two sets was divided into two parts; the first was unannealed while the second part was annealed, following standard procedures. Each of the unannealed and annealed samples was further divided into 38 aliquots required for the protocol. Fourteen aliquots each of unannealed and annealed samples were given PED and another set of 14 aliquots were without PED. The TL and RT-LMOSL measurements were carried out on each aliquot using an automated RISØ TL/OSL reader (model-TL/OSL–DA–15). Sensitisation reproducibility of repeated TL measurements on 10 different aliquots of each of the unannealed and annealed samples was quantified using Coefficient of Variation (CV). Data were analysed using descriptive statistics. The sensitisation signals of the aliquots of unannealed samples without PED was higher than that of the aliquots with PED by factor of 76.0 % and 79.0 % for TL and RT-LMOSL respectively for S2 while the corresponding factors obtained for S4 were 45.0 % for TL and 14.0 % for RT-LMOSL. In annealed samples, the sensitisation signal of the aliquots with PED was rather higher than that of the aliquots without PED, by factor of 224.0 % for TL and 201.0 % for RT-LMOSL for S2 and for S4, it was by factor of 245.0 % for TL and 217.0 % for RT-LMOSL. The sensitisations reproducibility of aliquots of unannealed samples were found to be poor with CV of 33.5 % for S2 and 52.0 % for S4. This improved significantly in the annealed samples to CV 6.3 % for S2 and 9.0 % for S4. Luminescence sensitisation by pre-exposure dose was dominant in annealed quartz samples. Therefore, only annealed quartz samples are recommended for pre-dose xvii UNIVERSITY OF IBADAN LIBRARY o retrospective dosimetry and dating. The use of 110 C thermoluminescence peak signal in sensitisation corrections of unannealed quartz is not advisable. Keywords: Quartz, Thermoluminescence, Optically stimulated luminescence, Pre-dose sensitisation, Thermal activation. Word count: 411 xviii UNIVERSITY OF IBADAN LIBRARY NOTATIONS AND SYMBOLS AD Additive-dose CCD Computerized curve deconvolution CW-OSL Continuous-wave optically stimulated luminescence ED Equivalent dose FWHM Full width height maximum Gy Gray,S.I unit of absorbed dose (1J/kg) HF Hydrofluoric Acid HR Heating rates IR Infra-red IRSL Infra-red stimulated luminescence LED Light emitting diode LM-OSL Linear-modulation optically stimulated luminescence MATAC Multiple aliquot thermal activation curves OSL Optically stimulated luminescence PMT Photomultiplier tube POSL Pulsed- optically stimulated luminescence RT Room temperature SAR Single-aliquot regenerative-dose TA Thermal activation TAC Thermal activation curve. TD Test dose TL Thermoluminescence TR-OSL Time Resolved-optically stimulated luminescence UV Ultra-violet xix UNIVERSITY OF IBADAN LIBRARY CHAPTER ONE INTRODUCTION 1.1. Background Quartz is the second most abundant mineral in the continental crust of the Earth after feldspar (Preusser, et al., 2009). It makes up 12.6% by weight of the Earth's crust as crystalline quartz and amorphous silica (Krbetschek, et al., 1997). This mineral exhibits Thermoluminescence (TL) and Optically Stimulated Luminescence (OSL) signals corresponding to amount of prior irradiations it received. Such luminescence signals have been used in retrospective dosimetry in the scope of accident dosimetry, dating geological and archaeological materials over the last 40 years (Wintle, 1973; Forman, et al., 1994; Duller, 1997; Krbetschek, et al., 1997; Prescott and Robertson, 1997; Roberts, 1997; Wintle, 1997; Preusser, et al., 2009). Mechanisms of quartz luminescence are full of complexity. As a result, many studies have been carried out by many workers on quartz in order to have better understanding of its luminescence for improved applications (Wintle, 1997; Bailey 2001; Preusser, et al., 2009; Pagonis and Kitis, 2012; Kitis and Pagonis, 2013; Koul and Polymeris, 2013; Asfora et al., 2014; Ferreira de Souza, et al., 2014; Koul et al., 2014; Sadek, et al., 2014; Topaksu et al., 2014; Polymeris, 2015). However, such reports are very scanty on Nigerian quartz and quartz samples obtain from Africa as a whole. Pioneer works on TL characteristics of quartz from Southwestern Nigerian was reported by Ogundare et al., (2006) on anomalous behaviour of thermoluminescence and its kinetic analysis (Ogundare and Chithambo, 2007a). More studies thereafter on Nigerian quartz using relative new luminescence method of Time Resolved OSL (TR- OSL) have also been reported (Ogundare and Chithambo, 2007b; 2008). Recently, universality study of thermal quenching effect and sensitisation of quartz of various origins that included Nigerian quartz have been reported (Subedi et al. 2011; Appendix 1; Appendix 2). A very important feature of quartz luminescence property is an enhancement in its sensitivity as a result of combined actions of pre-exposure dose of previously received irradiation by the quartz sample and the subsequent thermal activation reading or annealing to an activation temperature. This characteristic termed pre-dose effect o that is customarily allied with 110 C TL peak of quartz has been well studied and 1 UNIVERSITY OF IBADAN LIBRARY documented in the literature (Zimmerman, 1971; Chen 1979; Bailiff, 1994; Koul et al., 1996; Bailey, 2001; Li and Yin, 2001; Adamiec et al 2006; Galli et al., 2006; Koul 2008). The degree of pre-dose sensitisation is directly proportional to the quantity of the pre-exposure dose (Zimmerman, 1971). This forms the main reason behind the use of this technique in retrospective dosimetry. Pre-dose sensitisation is used to monitor o the increase in the sensitivity of 110 C TL peak resulting from pre-exposure dose rather than the accumulation of TL and OSL as in conventional luminescence methods o (Bailif 1994; Koul et al., 2010). Also, the sensitisation of 110 C TL glow-peak is the key instrument that is used in monitoring sensitivity changes of higher TL peaks and OSL signal of quartz as used in Additive-Dose (AD) and Single-Aliquot Regenerative- dose (SAR) methods of dating (Wintle 1997, Murray and Wintle, 2000; Wintle and Murray, 2006). Furthermore, this technique has been effectively utilized in authenticity testing of artefacts and firing temperature measurements of pottery materials (Bailiff, 1994; Koul et al., 1996; Galli et al., 2006; Li and Yin, 2001). Apart from the pre-dose sensitisation, annealing of quartz at temperatures o higher than 200 C has been observed to result in enhancement of the TL sensitivity of o the 110 C glow peak (Yang and McKeever, 1990; Chen et al., 1994; Rendell et al., 1994; Han et al., 2000; Schilles, 2001, Li, 2002; Adamiec, 2005; Koul, et al., 2010). This sensitisation, unlike pre-dose effect, is independent of pre-exposure dose but rather reliant on annealing temperature and duration. Therefore, this type of sensitisation represents the sensitivity change caused by annealing process only. A viable proof of the existence of pure thermal sensitisation was the enhancement in sensitivity after a high temperature annealing of a synthetic quartz that was not expected in view of pre-dose sensitisation (Yang and McKeever 1990; Botter-Jensen et al 1995). This is because synthetic quartz, which is believed to have received insignificant level of irradiations in the past, is not supposed to display enhancement of sensitivity as a result of annealing to high temperature. In addition to their observations on synthetic quartz, Botter-Jensen et al (1995) and Larsen (1997) also established the sole thermal sensitisation of sedimentary quartz samples. The long duration of annealing that was found to result in higher value of sensitisation was linked to a process caused by annealing only by Han et al., (2000). In another direction of thought, Chen and Li, (2000) who reported a similar observation proposed a modified Zimmerman‟s model with multiple R centres (each with different 2 UNIVERSITY OF IBADAN LIBRARY life times leading to a more complex function of time and temperature) to explain their findings. Recently, Koul et al., (2010) tried to identify the possibility of pure thermal o sensitisation in the pre-dose mechanism of the 110 C TL peak of quartz by way of using different annealing temperatures and heating rates (HRs) for thermal activation process. The basis of incorporating different heating rates was on the premise of different time of heating that is associated with each heating rate. Theses authors also confirmed the sole thermal sensitisations in their study that involved three natural quartz samples and one synthetic. In general, sensitivity change of the OSL has also been found to be parallel to o that of 110 C TL peak and that suggested a common mechanism for the two phenomena (Stoneham and Stokes, 1991; Franklin et al., 1995; Bøtter-Jensen et al., 1999; Jain et al., 2003; Koul and Chougaonkar, 2007). The similarity of the OSL and o 110 C TL peak sensitivities has made the latter a reliable monitoring tool for sensitisation of the OSL signal in various dating protocols (Aitken and Smith, 1988; Stoneham and Stokes, 1991; Stokes, 1994; Bøtter-Jensen et al., 1995; Wintle, 1997; Wintle and Murray, 1998; Murray and Roberts 1998; Chen and Li, 2000; Murray and Wintle, 2000; Wintle and Murray, 2006; Kiyak et al., 2007; 2008; Polymeris, et al., 2009). o Another issue of luminescence of quartz is the degree of sensitisation of 110 C TL peak and OSL that changes from sample to sample. The complexity of this peak, just like other quartz luminescence properties, is always attributed to the different crystallization environments during formation of respective sedimentary quartz samples that varies from location to location (Deer et al., 2004; Preusser, et al., 2009). Apart from this, „sample to sample‟ variation in sensitisations, „grain to grain‟ sensitisation variation has also been reported for the case of sedimentary quartz samples (Preusser, et al., 2009), leading to the devising of a „Single grain OSL attachment‟ to TL reader. 1.2 Justification of present study Despite recent findings in the literature (Koul et al., 2010; Oniya et al., 2012a) coupled with other initial reports on the existence of pure thermal sensitisation of quartz, the conditions leading to thermal type of sensitisation is yet to be established in the literature. Hence, identification of conditions leading to pure thermal sensitisation 3 UNIVERSITY OF IBADAN LIBRARY of quartz is important in the traditional pre-dose dating method. This is because there is possibility of erroneous age estimation in dating if pure thermal sensitisation happens to be the most prevalent mode of sensitisation in a sample to be dated using pre-dose dating technique. Workers have tried to explain the reason for variations in sensitisation of quartz to be as a result of different grain to grain origin, and various irradiations, thermal and optical/bleaching histories that each grain might have possessed during transportation (Deer et al., 2004; Preusser, et al., 2009). Each of these factors is widely known to be highly influential on pre-dose sensitisation. Ordinarily, the same degree of pre-dose sensitisation is expected from all the grains of any crystalline quartz. This assumption results from the fact that all the grains are from a common origin, and possess the same irradiations, thermal and optical/bleaching histories. Therefore, any variation in sensitisation that is exhibited by different grains of a crystalline quartz sample should arise from the intrinsic nature of the crystal and not from prevailing external factors. It is important to investigate this because such findings will definitely shed more light on the complex nature of quartz luminescence characteristics; and subsequently be of high assistance in all area of luminescence applications. Simultaneous sensitisation study of various components of OSL signal and that o of 110 C TL peak of a Nigerian annealed quartz sample has not been conducted based on the available gathered data from the literature. 1.3 Aim and objectives of the study This work was undertaken with the aim to study the pre-dose sensitisations of o 110 C Thermoluminescence (TL) peak and Room Temperature Linearly Modulated Optically Stimulated Luminescence (RT-LMOSL) of unannealed and annealed quartz. The objectives of the study are as follows: (1). Investigation of sensitisation of quartz grain of the same origin. This will reveal more information about the intrinsic nature of quartz luminescence properties and thereby improve luminescence dating techniques, mostly in sensitivity corrections of high TL peak and OSL component. 4 UNIVERSITY OF IBADAN LIBRARY (2). Determination of the individual contributions of pre-exposure dose and thermal activation in pre-dose sensitisation. This will form criteria for selection of appropriate quartz for pre-dose technique and shed more light in luminescence characteristics of quartz generally. o (3). Correlation of pre-dose sensitisation of 110 C TL peak and all components of OSL with a view to identifying which of the components of the OSL is associated with the 110°C TL trap in the two understudied Nigerian quartz samples. This will form a foundation upon which various dating protocols for Nigerian quartz could be based. 1.4 Thesis outlines This thesis is organized in five chapters. Chapter 1 deals with the introduction. This is followed by justification of present study, aim and objectives of the study. Thesis outline concludes this chapter. In Chapter 2, the basic concepts in luminescence are introduced including the band model, thermoluminescence (TL), OSL, dose estimation protocol, luminescence sensitivity and factors affecting sensitivity. Methods of computerised curves deconvolution are also presented. Brief description of retrospective dosimetry is presented in this chapter. Lastly, a description of the automated Riso TL/OSL reader system. Chapter 3 presents the experimental description and preparations of the two quartz samples studied in this work. Followed by these is a Feldspar inclusion test on the two quartz samples studied is investigated. Furthermore, this chapter elucidates a list of TL/OSL protocols employed for the measurements undertaken in this work and their respective descriptions. A description of Computerised curves deconvolution analyses that were performed in this work concluded this chapter. The results of the measurements carried out in this work are presented in chapter 4. The measurements are in two phases: (i) Reproducibility study of pre-dose o sensitisation of the 110 C TL peak and (ii) Modes of luminescence sensitisations in unannealed and annealed quartz samples. Each of these results is discussed. Respective 5 UNIVERSITY OF IBADAN LIBRARY models are proposed and implication of results on retrospective dosimetry and general recommendations for luminescence studies are presented in this chapter. Chapter 5 gives the conclusion and recommendation resulting from the pertinent findings of the study in luminescence dating and retrospective dosimetry generally. 6 UNIVERSITY OF IBADAN LIBRARY CHAPTER TWO LITERATURE REVIEW 2.1 Luminescence phenomena Luminescence techniques that are employed in retrospective dosimetry involve stimulation of light from some insulators (called phosphors) that had been exposed to ionizing radiation earlier (Wintle, 1997). When the stimulation is achieved by application of heat at a linear heating rate, the technique is termed TL while known as OSL when it is accomplished by optical stimulation. The quantity of acquired dose due to the earlier exposure to radiation by the phosphors is estimated by measuring the amount of light emitted during the stimulation. This is based on the assumptions that the acquired dose is proportional to the quantity of the emitted light. Therefore, the measured acquired/equivalent dose is then utilized in retrospective dosimetry either in evaluation of date in archaeological and geological dating or reconstruction of accident dose in accident dosimetry. The most widely used phosphor in retrospective dosimetry is quartz (Preusser et al., 2009). An important reason for this is its abundance in nature. This mineral is one of the more abundant rock-forming continental crust minerals (Ivliev et al, 2006). It is readily found in archaeological artifacts and geological materials like rocks and sediments (Preusser et al., 2009). Furthermore, quartz samples are always present in building bricks or blocks in which accident dose assessment could be based in the case of accident dosimetry. Quartz, either in natural or synthetic form, does not exhibit perfect crystalline nature (Preusser, et al., 2009). This mineral generally contains a vast number of intrinsic and extrinsic point defects. Intrinsic defects are unoccupied sites (vacancies) and occupied sites that in a perfect crystal would be unoccupied (interstitials). On the other hand, extrinsic defects are impurities at sites that in a perfect crystal would be occupied (substitutional impurities) or unoccupied (interstitial impurities). 7 UNIVERSITY OF IBADAN LIBRARY 2.2 Energy band model Theoretically, when electrons are excited in a perfect crystalline material that is free of defects by ionizing radiation, the excited electron will only reside in the -8 conduction band for a moment say less than 10 s before it loses its excitation energy and drops back down to the valence band, where it recombines with a hole. However, in quartz (as it is in other phosphors) the existence of defects results in the creation of allowed energy states in the otherwise forbidden band gap. Therefore, instead of all the excited electrons to recombine directly with free hole at the valence band, electrons are rather trapped at these point defects known as trap centres. The free holes in valence band are also captured at recombination centres. The captured electrons remain trapped at the trap centres until thermal or optical excitation is applied to return the electrons to the conduction band. From where some fraction of the electrons will find their way to the recombination centre and recombine with the trapped holes. However, a fraction of the de-trapped electrons will be re-trapped at the trap centres. The diffusion time is very short and recombination can be regarded as instantaneous (Aitken, 1998). When the recombination is radiative, luminescence signal is emitted and the total luminescence signal emitted is generally a function of the initial absorbed energy of the ionizing source (McKeever and Chen 1997). Meanwhile, a portion of all electron-hole recombination events take place without any light emission (non- radiative). The size of this fraction can vary between 0 and 1, depending on factors such as crystal temperature and type of impurity (Horowitz, 1984). Direct radiative transitions of electrons from conduction band to valence band do not contribute to luminescence signal described above. This is because such transition will give rise to radiation which has quantum energy greater or equal to the energy gap and consequently be absorbed by the material (self-absorption). This mechanism for TL and OSL processes is better portrayed graphically using a simple band model shown in Figure 2.1. The electron and hole pairs created by irradiation with ionising radiation are trapped at trap centres; electrons being trapped at T traps while holes are captured at R-centre, L-centre or K-centre. While excited electrons at conduction band are trapped at T traps, some portion recombined directly with captured holes at L-centre or K-centre during irradiation. The trap Ts represents a 8 UNIVERSITY OF IBADAN LIBRARY Conduction band Ec Ts Trap Tt trap L-Centre K-Centre Tc Trap hvlum R-Centre Ev Valence Band Fig. 2.1. Simple band model illustrating TL and OSL processes 9 UNIVERSITY OF IBADAN LIBRARY o shallow (unstable like 110 C TL peak) trap from where the probability of thermal eviction/detrapping is high. R-centre is also unstable hole-trap. Whereas, Tt is the thermally stable trap in which the probability of thermal eviction (without both thermal and optical external stimulation) is negligible. Tc is extremely stable and deep trap which is thermally/optically disconnect. By stimulating the sample either thermally (TL) or optically (OSL), trapped electrons at Tt and Ts traps may gain sufficient energy to escape and be released into the conduction band. From where some of these released electrons can be retrapped at the any of T traps. The remaining will find their way to either L-centre or K-centre where they recombine with trapped holes. The recombination at L-centre is radiative, that is luminescence is emitted. This luminescent signal is what is employed in luminescence dosimetry. Conversely, the recombination at K-centre is non-radiative in which the energy release may be in form of heat or light with wavelengths that are outside the detecting window of the luminescence equipment. 2.3 Thermoluminescence In case of TL, the stimulation by heat is applied by heating the material with a -1 linear heating rate  (Ks ), resulting in the temperature varying as T = T0 + βt, where T0 is temperature at time t = 0 (K). The mathematical expressions describing TL processes .are presented below. 2.3.1 TL kinetic expressions The rate of thermal release of trapped electrons into the conduction band at temperature  E  T is ns exp  2.1  kT  -3 where n is the number of trapped electron (cm ), E is the trap depth (eV) of the material (phosphor), k is Boltzmann's constant (eV/k) , and s is the frequency factor or -1 attempt to escape frequency factor (sec ). Assuming de-trapping and re-trapping of electron take place in the same trapping state, the intensity of TL signal is given by dm I    Ammnc 2.2 dt -3 in which m is the concentration of recombination centres (hole in centres), (cm ); 10 UNIVERSITY OF IBADAN LIBRARY -3 nc is the concentration of free electrons in the conduction band, (cm ); -3 -1 Am is the recombination probability (cm sec ). From Eq. 2.2 it is evident that the recombination rate is proportional to the number of free electrons, nc, and the number of active recombination centres, m. -3 The equation describing the rate of change of electrons in traps, n (cm ) is given by dn  E   ns exp   n (N  n)A 2.3 c n dt  kT  -3 -1 -3 where An (cm s ) is the retrapping probability and N (cm ) is the total concentration of traps. By using Eq. 2.3, Randall and Wilkins (1945), Garlick and Gibson (1948) and May and Partridge (1964) respectively obtained First, Second, and General order kinetics for TL independently, which are equations governing TL processes. The expressions are: dn  E  First-order kinetics, I    ns exp  2.4 dt  kT  dn s  E  Second-order kinetics, I    n2 e xp  2.5 dt N  kT  dn  E  General-order kinetics, I    nbs 'e xp  2.6 dt  kT  where s'  s / N and b = kinetic order, a parameter with values typically between 1 and 2 However, the general order kinetics is an interpolation between first and second order kinetics. It describes some experimental TL behaviour which does not correspond to any of first and second kinetics orders, but rather to a kinetics order between the two. Similar to this, Chen et al., (1981) suggested another order kinetic theory to describe these cases of intermediate kinetic order. This is called mixed order kinetic. It was given as: dn  E  Mixed order kinetics , I    n(nC)s '.e xp  2.7, dt  kT  where s’ = s/(N+C), C is the concentration of electrons trapped at some kind of deep traps. 11 UNIVERSITY OF IBADAN LIBRARY It follows from Eq. 2.4 that dn  E   ns exp  2.8 dt  kT  If the sample is heated up so that the temperature is set at a linear rate  , so that dT  dt , we can now substitute for dt in Eq. 2.8 and separate the variable to obtain. dn  s   E    exp dT 2.9 n     kT  At the start of the heating when T = To let the number of trapped electrons be equal to no. When the temperature T has been reached there are n electron left in the trap. Now we can integrate Eqs. 2.9 between the limits to give n dn s T  E     exp dT  2.10 no n  To  kT    n  s T  E  ln      exp dT  2.11 T  no   o  kT   by finding the exponential of both side gives  s T  E   n  no exp   exp dT  T  2.12  o  kT    Substitution of Eqs. 2.12 into Eq. 2.4 yields  E   s T  E   I (T )  sn0 exp exp   exp '  dT 2.13  kT    T '  o  kT   which is the expression for the first order kinetics. With similar approach by starting from Eqs. 2.5, 2.6 and 2.7, the following equations are equally obtained for second, general and mixed order kinetics respectively (Pagonis et al., 2006) 2 2n0 s  E   no (b1)s T  E I (T )  exp  1 exp  dT '         2.14 N  kT   N To  kT '   b   E   (b1)s '' T  E   b1 I (T )  s ''n0 exp  1  exp dT '  2.15  kT '    To  kT   12 UNIVERSITY OF IBADAN LIBRARY 2  s 'C  T  E    E s 'C  exp    exp dT ' exp T    o  kT '    kT I (T )  2.16 2   s 'C  T  E    exp   exp dT '      T  o  kT '    −3 where n0 = number of trapped electrons at time t = 0 (m ) s ''  s 'n(b1)o is an empirical parameter acting as an “effective” frequency factor for −1 general-order kinetics (in s ), n   o , but here C only takes positive value. no C 2.3.2 Glow curve During TL measurements, the intensity of the emitted luminescence (TL response) is generally recorded and plotted against temperature; the resulted graph is called a glow-curve. Glow curve is a unique product of the TL measurement of a given phosphor. Since the temperature of measurement is increased linearly with time, the probability of detrapping increases. The intensity initially increases, then it reaches a maximum value at the rate of detrapping peaks, and then it drops to zero again. The drop in luminescent intensity is caused by depletion of the trap. A glow curve may consist of a number of glow peaks having different heights, widths and shapes, each corresponding to different trapping levels. The TL belonging to each peak corresponds to trap centres within the crystal. The positions and shapes of the TL peaks are related to the characteristics of these traps, which are in turn typical of the crystal containing them. Thus, the TL glow curve represents a scan through the various types of trap present in the crystal from which information about the trap can be inferred. The dose imparted on the sample is conventionally estimated by measuring height of desired peak or area under the peak or entire glow curve. A typical example of a glow curve is shown in Figure 2.2. By solving the kinetic equations stated above usually yields glow curves similar to the kinetic equation solved. Figure 2.3 presents glow curves obtained by solving first and second order kinetics Eqs. 2.13 and 2.14. 13 UNIVERSITY OF IBADAN LIBRARY Fig. 2.2. Typical TL glow curve representing the photons released during the recombination at luminescence centres of previously trapped electrons. TL glow curve (heating rate 5 °C s−1) of quartz from Nigeria (GW1, Gumnior and Preusser, 2007) 14 UNIVERSITY OF IBADAN LIBRARY Fig. 2.3. Schematic comparison of TL glow peaks for first- and second-order kinetics. 12 −1 3 −3 The parameters are E = 1 eV, s = 10 s , n0 = N = 10 m . (Pagonis et al., 2006) 15 UNIVERSITY OF IBADAN LIBRARY 2.3.3 Emission spectrum As glow curve was defined as the plot of light intensity against temperature or time, emission spectrum is rather the graph of light intensity against wavelength. Emission spectrum is somewhat similar to TL glow curve. The light emitted during TL possesses a particular colour associated with recombination producing it. However, unlike glow curves, the colour dimension is not unique to TL, but is shared by all types of luminescence. While each TL peak in glow curve gives information about electron trap of the TL material under investigation, each peak in emission spectrum conveys information about recombination centre (hole trap). Particularly, emission spectrum of each glow peak of a glow curve is associated with the recombination/luminous center of that peak. Thus, glow peaks with identical emission spectra have the same recombination center. As a result, emission spectrum analyses provide adequate information about the recombination center that is responsible for each glow peak. Quartz of various origins exhibit broad emission band in the blue, red and even around 990-1000nm. The different geological conditions of formation associated with each origin are apparently responsible for these various natures of glow curves and emission spectra (Jones and Embree 1976; Krbetschek et al., 1997; Preusser et al., 2009). However, the most frequently reported of all are the 380nm and 470nm which 0 0 have their recombination centres linked to [H3O4] and [AlO4] respectively. Optically stimulated luminescence (OSL) is believed to share the same emission band around 380nm. The wavelength emission from this center (380nm) is temperature dependent, while its luminescence is characterized by thermal quenching for all peaks in the family (Yang and McKeever, 1990; Wintle and Murray, 1999). Quartz TL peaks that are associated with the latter do not exhibit thermal quenching and do not bleach, unless the light wavelength is less than about 400 nm. Finally, the 470 nm emission of quartz has been found to remain almost unaffected by the pre-dose effect (Yang and McKeever, 1990; Koul and Chougaonkar, 2007). Examples of emission spectra are shown in Figures 2.4 and 2.5. 16 UNIVERSITY OF IBADAN LIBRARY o Fig. 2.4. Emission spectra at 190 – 210 C of quartz extracted from sediments; ARD 4: Thar deser, India (aeolian), MDL 12: Isrea (aeolian), PTR 4: California (marine- littoral). (after Singhvi and Krbetschek, 1996) 17 UNIVERSITY OF IBADAN LIBRARY Fig. 2.5. Three-dimensional emission spectrum of an Aeolian/fluvial sample (KD 8) from the That desert, India (after Singhvi and Krbetschek, 1996). 18 UNIVERSITY OF IBADAN LIBRARY 2.4 Optically stimulated luminescence Under optical stimulation, the irradiated phosphor is exposed to light (UV, visible or infrared) under a constant temperature and the OSL emission is recorded as function of stimulation time. The integral of the OSL emitted during the stimulation period is a measure of the dose of irradiation absorbed by the sample since it was last exposed to light. Unlike in the case of TL in which heat is applied only at a constant heating rate, there are three popular modes of stimulation in OSL. They are described below: i. Continuous-wave OSL (CW-OSL):- this is a method in which thestimulation light intensity is kept constant and the OSL signal is monitored continuously throughout the stimulation period (Huntley et al., 1985), ii. Linear-modulation OSL (LM-OSL):- in this method, the stimulation intensity is ramped (increased) linearly while the OSL is measured (Bulur 1996), iii. Pulsed-OSL (POSL):- the stimulation source is pulsed and the OSL is monitored only between pulses in this method. This method has been developed into time- resolved OSL (TR-OSL) which provides information about luminescent centres (Bailiff, 2000; Chithambo and Galloway, 2000). The plot of OSL signal recorded as a function of time is called OSL curve. Typical examples of CW-OSL and LM-OSL curves obtained experimentally are presented in Figures 2.6 and 2.7 respectively. More considerations will be devoted on CW-OSL and LM-OSL in this work. 2.4.1 CW-OSL Just like it is with TL, the transitions of charge between energy levels during irradiation and subsequent optical stimulation of a phosphor can be described by a -1 series of non-linear, coupled rate equations. p (s ) which is the rate of stimulation of electrons from the trap is related to the incident photon flux and the photoionisation cross-section  by p(Eo )  (Eo ) 2.17 where Eo is the threshold optical stimulation energy required for charge release and a return of the system to equilibrium. 19 UNIVERSITY OF IBADAN LIBRARY Fig. 2.6. Typical OSL decay curve from a sedimentary quartz samples given a beta dose of 2 Gy obtained using a green light wavelength band of 420 – 550 nm producing 2 16 mW/cm at the sample position (Botter-Jessen, 1997) 20 UNIVERSITY OF IBADAN LIBRARY Fig. 2.7: An example of exposing a heated and dosed quartz sample to linearly increasing blue LED stimulation from 0 to 20 mW/cm2 over 7200 s. (Botter-Jessen et al., 2000). 21 UNIVERSITY OF IBADAN LIBRARY The first-order kinetic describing CW-OSL is given as dm dn IOSL      np 2.18 dt dt with the solution of IOSL  no pexp(tp)  Io exp(t /d ) 2.19 where Io is the initial OSL intensity at t =0 and  d is the CW-OSL decay constant (Botter-Jessen et al., 2003a, McKeever et al., 1997). The second order kinetics is given by Chen and McKeever (1997) as n2 p dn IOSL    2.20 NR dt Where R  A Am after integration it is presented as 2  no pt IOSL  Io 1  2.21  NR  where Io  n 2 o p / NR . For general order kinetics, Eq. 2.21 becomes b  no pt 1bIOSL  Io 1  2.22  NR  with I bo  no p / N Chen and Leung, (2002) later worked on this and considered CW-OSL to be best fitted by a so-called “stretched exponential” of the form:    t   IOSL  Io exp   2.23   d   with 0<  <1. Figure 2.8 shows a typical example of OSL curves of first and second order obtained by solving Eqs 2.19 and 2.21. 22 UNIVERSITY OF IBADAN LIBRARY Fig. 2.8. OSL curves of first and second order. 23 UNIVERSITY OF IBADAN LIBRARY 2.4.2 LM-OSL A linear increase in the intensity (t) of optical stimulation at a fixed wavelength is employed in LM-OSL unlike in the case of CW-OSL in which steady stimulation intensity  is applied. By the adoption of this mode of stimulation we have (t)   t 2.24 0  with   d / dt and  =0 at time t = 0. Therefore, the OSL signal is observed with series of peaks, with each peak corresponding to the optical release of charge from different trap types. Consequently, traps with large photoionisation cross-section at the particular wavelength used in stimulation are emptied first leaving those with small photoionisation cross-section to be emptied at later time of stimulation. This behaviour is well portrayed by the position of each peak in LM-OSL curves. Unlike the case of TL, the depletion of all the traps starts at the same time at the beginning of stimulation of LM-OSL however with different depletion rates. Meaning that all the peaks normally originate from the beginning of the OSL curve (at time t = 0) irrespective of their peak position. This is contrary to the case of TL in which each peak starts from different point depending on the positions of the respective peak in the glow curve. In general, LM-OSL is now becoming more popular than CW-OSL because the former gives more information about the traps involved in OSL measurements than the later (Botter-Jenson et al., 2003a). To illustrate the shape of an LM-OSL curve mathematically, we consider the intensity of optical stimulation is ramped from zero to a maximum value m . Hence  now takes the new form (t)   t for LM-OSL. Therefore p  t 2.25 By substituting Eq. 2.25 into Eq. 2.18 results in the first order kinetics of LM-OSL dn IOSL    tn 2.26 dt from which a Gaussian function 24 UNIVERSITY OF IBADAN LIBRARY   2 n  n exp t  2.27 0  2  is obtained. Hence,   2 IOSL  n0 t exp t  2.28  2  following Whitley and McKeever, 2001, Eq. 2.28 may be written as k   I i 2  2.29 OSL   tn0i i exp t  i1  2  if there are k traps of type-i. Eq. 2.29 describes simple sum of first order LM-OSL curves that represent several traps that are being emptied simultaneously at different rates. 2.5 Dose response An essential element in any dating method is the graph of TL/OSL intensity versus dose known as dose response (McKeever and Chen, 1997). This provides data calibration parameters that are used to estimate natural luminescence signal which is converted into an equivalent dose. The response curve of a dosimeter should be proportional to dose; even ideally linear over dose range of interest. However, dose response of quartz is not always linear with dose because of sensitivity changes. It is rather known to have a non-linear growth curve with, quite often, a faster than linear dependence that is superlinear/supralinear, and slower than linear which is known as sublinear growth. A very important factor that determines the dose response curve is the competitions of electron among the electron traps during „excitation stage‟ and a „readout stage‟ of luminescence (Chen and McKeever, 1997). Chen and McKeever, (1997) presented a linearity index f (D) that can be used to estimate dose linearity of any material as follows: TLi / D f (D)  i 2.30 TL1 / D1 where TLi(D) is the sample TL response corresponding to dose Di, and D1 is the normalization dose in the initial linear region, f(D1) is the sample TL response 25 UNIVERSITY OF IBADAN LIBRARY corresponding to dose D1, (by this definition f(D)>1, f(D)=1, and f(D)<1 respectively imply supralinearity, linearity and sublinearity). Dose response of quartz is associated with saturation at high dose of irradiation. Saturation is normally brought about by absolute filling up of the electron traps from high irradiation dose at which response to subsequent dose is not possible. However, very high dose of irradiation can destroy the electron trap thereby causing decrease of the response. The effect is known as radiation damage. Typical dose responses for four electron traps/peaks (P1, P2, P3 and P4) of Brazilian quartz are shown in Figure 2.9 (Sawakuchi and Okuno, 2004). The peaks exhibit a common pattern up to ~30 kGy, with a fast initial supralinear growth. After that dose the peaks P3 and P4 saturate and the peak P2 decays continuously until 500 kGy and becomes constant. The peak P1 cannot be resolved after ~20 kGy due to the high intensity of the peak P2. However it seems to behave like the peak P2. 2.6 Luminescence sensitivity Luminescence sensitivity of a given sample is defined as the amount of luminescence emitted per unit sample mass in response to a fixed laboratory dose (Furetta, 2003). Ideally, this phenomenon is expected to be constant for quartz but it is not in practice. Quartz of diverse kinds and origins possess different luminescence sensitivity. Better still, change in sensitivity stands as the most widely encountered problem in luminescence dating applications (Murray and Roberts, 1998). This attribute among other features of quartz that vary from „samples to samples‟ are ascribed to different crystallization environments during formation of quartz samples (Deer et al., 2004; Preusser et al., 2009). Apart, there are even some treatments that cause enhancement (sensitisation) or decrease (de-sensitisation) in luminescence sensitivity of quartz of the same kind and origin. Such could arise from prevailing external factors that the quartz sample either naturally or inadvertently exposed to in nature or those that are as a result of artificial or premeditated treatments in the laboratory during or prior to luminescence readouts; like radiaztion exposure, thermal treatments and optical stimulation e.t.c. (Bailey, 2001). The remaining part of this section will be devoted to considering some factors leading to sensitivity changes of quartz. 26 UNIVERSITY OF IBADAN LIBRARY Fig. 2.9. Dose response of the TL peaks (P1, P2, P3 and P4) of Brazilian natural quartz  6o oexposed to rays of Co in the range 1–1000 kGy. Heating rate used: 1 C/s. (Sawakuchi and Okuno, 2004) 27 UNIVERSITY OF IBADAN LIBRARY 2.6.1 Competitions during irradiation and stimulation Various traps usually compete to trap free electrons that are produced during irradiation or stimulation. This process is known as competition. The nature of competitions depends on intrinsic nature of each respective luminescence material and some other external factors that cause alterations in the sensitivity of a given sample like radiation exposure, thermal treatments and optical stimulation (Bailey, 2001). By employing the model of Figure 2.1 as an example, traps Ts, Tt and Tc will contend among themselves for electrons during irradiation stage in which electrons are raised from the valence band into the conduction band. The manner of charge trafficking during irradiation is illustrated in Figure 2.10. The fraction of the total excited electrons to be trapped into each of the electron traps depends on their respective trapping probabilities. The influence of competitions is most apparent during the stage of stimulation like TL readout. Figure 2.11 illustrates charge trafficking during the early temperature o o of heating; say to about 150 C in which 110 C TL peak is evicted. Due to competitions between Ts and the relatively thermal stable traps, it is obvious that only a portion of the electrons released from Ts trap to the conduction band can find their way to the L or K centre where recombination with hole takes place while the remaining fraction are shared among Tt and Tc traps. In this case, traps Tt and Tc are competitors and their degree of competitions depends on the level of their initial fillings. Therefore, there is a reduced competition at higher dose levels of saturation. In that case, the released electrons can only be involved in the recombination which will yield a relative o enhanced 110 C TL signal. The mode of charge trafficking reduces to what is contained o in Figure 2.12 when heating is continued to higher temperature beyond 110 C TL peak. Trap Ts in not taking part here and it is the turn of eviction of electrons in Ts which represent all thermally stable electron traps. Trap Tc is the only competitor with respect to traps Tt. Trap Tc is always potential competitor because it has a very deep depth that makes it to be thermally/optically disconnected always. Disconnected in the sense that once electrons are trapped into Tc trap, they can never be freed or evicted for re- trapping into any of the T traps or recombination with holes at recombination centre. 28 UNIVERSITY OF IBADAN LIBRARY Conduction band Ec Ts Trap Tt trap L-Centre K-Centre Tc Trap hvlum R-Centre Ev Valence Band Fig. 2.10. Transitions taking place during excitation stage in the same energy scheme as that shown in Figure 2.1. 29 UNIVERSITY OF IBADAN LIBRARY Conduction band Ec Ts Trap Tt trap L-Centre K-Centre Tc Trap hvlum R-Centre Ev Valence Band Fig. 2.11. Schematic competition of free charges among the electron traps during early o state of heating (TL) up to ~150 C 30 UNIVERSITY OF IBADAN LIBRARY Conduction band Ec Ts Trap Tt trap L-Centre K-Centre Tc Trap hvlum R-Centre Ev Valence Band Fig. 2.12. Schematic competition of free charges among the electron traps during late o state of heating (TL) from ~150 - 500 C 31 UNIVERSITY OF IBADAN LIBRARY The mechanism of competition explained above for TL is similar for the case of OSL except for the mode of stimulation that differs. This procedure of competition has been used to explain the enhancement of the luminescence sensitivity and then the phenomenon of suparlinearity (Chen et al., 1988) in the framework presented below explanations. During readout, electrons released from, say trap Ts, could be re-trapped in competitor traps Tt or Tc or recombine in L or K centres at low doses of irradiation. This competition thereby causes reduction in luminescence signal. Increase in sensitivity that is paramount to suparlinearity, is expected at higher dose of irradiation. This is because almost or absolute saturation of competitor traps Tt and Tc at higher dose levels will result into a reduced or no competition at all. Consequently, nearly all evicted electrons from Ts trap will be available for recombination at L-centre or K- centre leading to enhanced sensitivity. 2.6.2 Luminescence sensitisation One of the major challenging problems in quartz luminescence world is the issue of sensitivity changes (sensitisation). This is because adequate knowledge of this is required mostly in the regenerative-dose and pre-dose techniques where series of irradiations and thermal reading are employed. Moreover, quartz that is widely used in this task exhibits complex sensitivity nature. As a result, the issue of luminescence sensitivity has been a major point of attraction, both in the past and in the present, to many researchers. Two main causes of sensitisation in quartz are classified as pre-dose and thermal sensitisations. 2.6.2.1 Pre-dose sensitisation Generally, sensitisation is an effect recognized to be the increase in the luminescence intensity of a phosphor to a certain test-dose (TD) of irradiation as a result of some treatment of the sample like heating, irradiation etc. Pre-dose sensitisation of quartz has been widely believed to be an enhancement of sensitivity as a combined result of irradiation and annealing. This implying that sensitivity changes in o the 110 C TL peak result only when the sample has been pre-exposed to a dose of 32 UNIVERSITY OF IBADAN LIBRARY o irradiation and subsequently heated to a given temperature; typically 500 C (Chen and McKeever 1997). Sensitivity change of the OSL also has been found to be parallel to o that of 110 C TL peak and that suggested a common mechanism for the two (Stoneham and Stokes, 1991,). Franklin et al (1995) established that the electrons from the OSL o traps combine with same luminescence centres as those from 110 C TL peak. Many o researchers have confirmed the similarity of the OSL and 110 C TL peak sensitivities (Stoneham and Stokes, 1991, Murray and Roberts 1998, Wintle and Murray, 1998). Pre-dose sensitisation was first discovered by Fleming and Thompson (1970) and the first model to explain the phenomenon was proposed by Zimmerman (1971). Her o model was later amended by Chen (1979) to accommodate superlinearity of 110 C TL peak. 2.6.2.1.1 Pre-dose model The amended model of Zimmerman (1971) by Chen (1979) consists of two electron trapping states, Ts and Tc (the competitor Tc represents both Tt and Tc in Figure 2.1) and two hole states R-centre (the reservoir) and L-centre. During the excitation (see Figure 2.13) by administration of TD of ionizing radiation, electrons are raised from the valence to the conduction band. A fraction of this is trapped at the trapping state Ts and remaining at competitor trap, Tc. The holes go preferably to the reservoir R-centre. However, there is a non-negligible probability of the holes going to L-centre. Thus, the first thermally freed electrons can recombine with them during the o heating (first heating to about 150 C) and this results in the emission of the TL of the o unsensitised material. The annealing (or TL reading), typically to 500 C following the application of a high dose, empties all the electrons from Ts, yielding a rather high o o 110 C TL peak signal. The significance of the thermal treatment to ~500 C is to thermally release holes from the reservoir R, to the luminescence centre L. Zimmerman ascribed the reason for the higher response/signal observed, when a subsequent TD is given, to this increase in the concentration of holes in the luminescence centres by probability of radiative recombination. In attempt to improve on this model, more than one reservoir centres have been introduced to this model and that have been employed to account for most of the TL and OSL experimental observations (Bailey, 2001; Adamiec et al 2006) 33 UNIVERSITY OF IBADAN LIBRARY Conduction band Ec Ts Trap L-Centre Tc Trap hvlum R-Centre Ev Valence Band Fig. 2.13. Pre-dose sensitisation schematic band model. 34 UNIVERSITY OF IBADAN LIBRARY Two major factors that can cause desensitisation of pre-dose effect after thermal activation are (i) radiation or dose quenching and (ii) ultra-violet (UV) reversal. 2.6.2.1.2 Radiation quenching The effect called radiation or dose quenching is a desensitisation phenomenon that is observed in pre-dose effect. This occurs if the sensitivity to a TD following pre- exposure dose and thermal activation of a sample (Baillif 1994) that is expected to increase rather decreases with irradiation dose. This effect is quite different from radiation damage that results from annihilation of electron traps by very high dose of irradiation. Two different models have been used to explain radiation quenching effect. Using the Zimmerman (1971) model, Aitken, (1985) attributed radiation quenching effect to the removal of holes trapped at L-centres as recombination occurs during irradiation leading to diminution in sensitivity of subsequent luminescence measurement. This model is contrary to a model of Bailey, (2001) in which he proposed that recombination is allowed at the R-centre unlike the model of Zimmerman, (1971) and that the concentration of holes trapped at both L- and R- centres increases with dose. Using these, he argued that it is an increase in competition for free charge, during both irradiation (TD) and during heating (TL readout) from the R-centres that produces the quenching effect. However, by considering the sequence of pre-dose technique of dating of o Bailiff, (1994), the recombination that occurs during heating to ~150 C that is meant o for removal of 110 C TL peak which, follows laboratory calibration beta-dose, definitely contributes to radiation quenched sensitivity also. What makes this proposition viable is the level of the depletion of trapped holes at L centre resulting from this which is measurable and quantifiable if the pre-heat TL is recorded. 2.6.2.1.3 UV Reversal UV reversal, as described by Zimmerman (1971), is the substantial sensitivity decrease that is always observed once an irradiated and annealed sample is illuminated by UV light. And a repeated high temperature annealing increases the sensitivity back 35 UNIVERSITY OF IBADAN LIBRARY to nearly the same level as the previous one following the first high-temperature annealing. Zimmerman (1971) suggested that the UV reversal is associated with the transfer of holes from L-traps to R-traps during exposure of sample to UV light. The detail of her model is presented as follows: When the sample is exposed to UV light the electrons in the valence band are excited. Because they acquire too much energy to be trapped by the R traps they are preferentially captured by L traps. The holes transferred to the valence band are then trapped by R traps, decreasing the sensitivity. Another model proposed by McKeever (1991) linked UV reversal with the optical release of electrons from deep traps which then recombine with trapped holes in 0 (H3O4) centres, thereby reducing the concentration of recombination centre. 2.6.2.2 Thermal sensitisation Apart from the pre-dose sensitisation, annealing of quartz at temperatures o higher than 200 C has been observed to result in enhancement of the TL sensitivity of o the 110 C glow peak (Han et al., 2000). The effect of this sensitisation is not limited to annealing temperature only but also to duration of the annealing. Longer duration of heating (annealing time) generally results in higher value of the sensitisation. This sensitisation represents the sensitivity change caused by annealing process only and should not be confused with pre-dose process. This phenomenon is based on some workers findings on thermal sensitisation. This fact was supported by the enhancement in sensitivity after a high temperature annealing of a synthetic quartz that was observed, which was supposed not be in view of pre-dose sensitisation (Yang et al 1990). This is because synthetic quartz, which is believed to have received insiginificant level of irradiations in that past, is not supposed to display enhancement of sensitivity as a result of annealing to high temperature based on pre-dose mechanism. McKeever et al (1983) observed that both TL and radioluminescence (RL) were affected in the same way, indicating that the sensitivity changes were due to alterations to the recombination centres (RL is the luminescence generated in material during exposure to nuclear radiation). Changes to the emission spectra that were also observed by Hashimoto et al (1994) confirmed these alterations to the recombination centres. Botter-Jensen et al (1995), in their study was able to establish the sole thermal sensitisation of sedimentary and synthetic quartz samples also. This was observed in 36 UNIVERSITY OF IBADAN LIBRARY OSL and phototransferred TL (PTTL) of quartz to follow the same pattern of sensitisation after high temperature annealing. They suggested a model to explain thermal sensitisation which is based on that of McKeever et al (1983) (alterations to the recombination centres) and removal of competitors at high temperature annealing. 2.6.2.2.1 Thermal sensitisation model The Botter-Jensen et al (1995) model contain three trapping states. As being o seen in Figure 2.14, a shallow electron trap Ts represents 110 C TL peak and all those shallow trap centres in the real material. The electron tap centres Tt represents all the optically active trap centres that are depopulated during OSL measurements and o thermal stable traps in quartz (e.g 325 C TL peak). Tc represents the thermally deep electron traps that do not empty either optically or thermally during TL or OSL measurement. The model has two recombination centres L-centre and K-centre in which L-centre is radiative and K-centre none-radiative. Based on the assumption that annealing at high temperature alters recombination centres, concentration of L-centre is increased relatively to that of K-centre for annealed case in this model. This is achieved by either increasing L concentration or reducing K-centre concentration. Also, to model the removal of competitors at high temperature annealing, the concentration of Tc was reduced. By applying all these in the model, the following as observed experimentally were simulated (Botter-Jensen et al 1995; Larsen 1997): i. OSL and PTTL sensitivity changed due to annealing temperature, o ii. Temperature shift of the photo-transferred 110 C TL peak, and iii. Dose response of OSL in sedimentary quartz. The agreement between these and experimental results supports the hypothesis that thermal sensitisation observed in quartz is due to alterations to the concentrations of the recombination centres and trap centres and not to pre-dose effect only. These findings also suggested the impact of purely thermal sensitisation in pre-dose and regenerative dose methods. 37 UNIVERSITY OF IBADAN LIBRARY Conduction band Ec Ts Trap Tt trap L-Centre K-Centre Tc Trap hvlum R-Centre Ev Valence Band Fig. 2.14. Thermal sensitisation schematic band model 38 UNIVERSITY OF IBADAN LIBRARY Also, the duration of the annealing has been observed to affect the degree of thermal sensitisation (Han et al., 2000). Longer time of annealing was found to result in higher value of the sensitisation and that was linked to a process caused by annealing only. Nevertheless, Chen and Li, 2000 who reported a similar observation has proposed a modified Zimmerman‟s model with multiple R centres (each with different life times leading to a more complex function of time and temperature) to explain their findings. 2.6.2.3 Thermal activation curve The plot of thermally induced luminescence sensitisation as a function of thermal activation temperature (TAT) is often called thermal activation curve/characteristic (TAC). This is, however, referred to as multiple aliquots thermal activation curve (MATAC) when multiple aliquots are involved. TAC gives room for easy comparison of sensitisation associated with each TAT. An example of TACs is depicted in Figure 2.15. As observed in Figure 2.15, TAC is usually characterized by peak at maximum TAT. The „late activation‟ and „early activation‟ referred to in the figure are due to different heating rates employ for activation. Aitken (1985) points out that the temperature at which the sensitivity maximum of the TAC is reached depends on the time spent at high temperature. If the heating rate is slow, and more particularly, if the maximum temperature is held for, say, a minute before cut off, then the maximum will shift downwards in temperature. The phenomenon behind the usual decrease in value of the sensitivity beyond o the highest TAT in MATAC of quartz 110 C TL peak has been an issue of concern in quartz luminescence research. According to Aitken (1985), this is referred to as thermal deactivation which, he presumed to be due to a direct thermal eviction of holes from L- centres into the valence band. Conversely, this is contrary to the pre-dose model in which L-centre is assumed to be much further from the valence band so that once a hole is captured at L-centre, it cannot be thermally released back to the valence band. Figel and Geodicke (1999) proposed a model which takes care of the possibility of thermal eviction of holes from L-centres into the valence band. These authors argued that recombination of electron trapped at high TL peaks beyond the 39 UNIVERSITY OF IBADAN LIBRARY Fig. 2.15: TAC of Fleming and Thompson (1970) redrawn. Curves (a) and (b) show the „late activation‟ and „early activation‟ phenomena, respectively‟ 40 UNIVERSITY OF IBADAN LIBRARY maximum TAT during further thermal activation will lead to depletion of already enriched trapped holes at L- centres. Consequently, a reduced TL sensitivity will follow. Based on this later model, it is envisaged that the structure of glow curve of thermal activation TL if recorded (which is quantifiable measure of the recombination process) should be related to TAC. Lastly, a model of probable recombination of electron at L-centres during subsequent irradiation after thermal activation was proposed by Chen and Pagonis (2003) to explain this nominal decrease of sensitisation after TAT. This model was experimentally confirmed recently (Appendix 2) 2.6.2.4 Heating rate effects If different heating rates (  ) are used during TL readout, the position of the maximum glow-peak temperature (Tm) is known to shift towards higher temperature with increasing  (Kitis et al., 1993). Furthermore, there is always a drastic reduction of TL signal as the  increases. The later diminution has been attributed to thermal quenching effect. The effect of  is better portrayed in Figure 2.16 Kitis et al., (1993) presented explanation to account for the shifting of Tm with  . At a low heating rate 1 , the time spent by the phosphor at a temperature T1 is long enough so that an amount of thermal release of electrons depending on the half-life at this temperature could take place. As the heating rate increases to  2 > 1 the time spent at the same temperature T1 decreases and therefore the thermal release of electrons is also decreased. So, a higher temperature T2 is needed for the same amount of thermal release to take place at  2 . In this way the whole glow-peak is shifted to higher temperatures as the heating rate increases in a manner depending on the half-life and the time spent at each temperature. 2.6.2.5 Thermal quenching Thermal quenching is an effect known to cause decrease in TL and OSL signals as measurement temperature is increased. In other words, it is the loss of luminescence efficiency with increasing measurement temperature. This effect has been reported for both quartz TL (Wintle, 1975) and quartz OSL (Smith et al., 1990; Spooner, 1994) 41 UNIVERSITY OF IBADAN LIBRARY o Fig. 2.16: Experimental glow-peak shapes of the 110 C TL peal of Norwegian quartz o obtained by heating rates in C/s: a = 2, b = 8, c = 20, d = 30, e = 40, f = 50, g = 57, h = 71 42 UNIVERSITY OF IBADAN LIBRARY This effect is only obvious in TL when the measurement is taken with different heating rates. As the heating rate increases the glow-peak is shifted to higher temperatures and the integral of the glow-peak, which measures the luminescence efficiency, decreases. Apart from the reduction in the TL peak (area and peak height) that is caused by this effect, the shape of the peak is distorted (Kitis et al., 1993). This consequently affects the kinetic order since the high temperature side of the glow peak is afflicted more than the lower temperature side. However, in the case of OSL, the effect is noticeable when the OSL is measured at a given elevated temperatures. (Wintle, 1975; Petrov and Bailiff, 1996). Two models have been used to explain the thermal quenching mechanism; Schön-Klasens (McKeever et al., 1997) and Mott-Seitz models (Mott and Gurney, 1948). The reduction of luminescence efficiency with Schön-Klasens model is ascribed to a progressive loss of L-centres due to thermal ionisation of trapped holes into the valence band. Conversely, in Mott-Seitz model an increase in the probability of non- radiative recombination centre relaxation is predicted for higher temperatures, rather than a reduction in the concentration of L-centres. With this, electrons that are captured to the excited states of L-centres from the conduction band are thermally assisted to undergo non-radiative transition to the ground state. Thus, the probability of non- radiative recombination increases with temperatures. The recombination energy that ought to be given out as light is absorbed by the lattice in this case. This model is generally more adopted than the former (Bailiff, 1994; Bailey 2001). The empirical expression for the reduction in radiative intensity due to thermal quenching is of the form (Curie, 1963) 1   2.31 W 1Ce kT where is luminescence efficiency, C is dimensionless quantity, W (in eV) is the activation energy and T is the temperature (K). In view of Mott-Seitz model (Bøtter-Jensen et al., 2003a) W is the energy barrier necessary for an excited state electron to transition non-radiatively to the ground state, with the emission of phonons (Nanjundaswamy et al., 2002). 43 UNIVERSITY OF IBADAN LIBRARY 2.6.2.6 Temperature lags Another effect that has to do with measuring temperature in TL is termed temperature lags or thermal lagging. In practice, the thermocouple that measures the temperature of the sample during TL measurements is usually fixed to the heating element in many of the TL/OSL readers. This meaning that the thermocouple measures the temperature of the heating element rather than that of the sample. However, when physical information from the glow curves is to be extracted, it is essential to know the sample‟s temperature rather than that of the heating element. Differences in the temperature of the heating element and that of the sample that is known as temperature lags, have been studied by many workers (Taylor and Lilley, 1982; Gotlib et al., 1984; Betts and Townsend, 1993; Betts et al., 1993; Piters and Bos, 1994; Facey, 1996;). The major causes of this effect are, non-ideal thermal contact between the heater element and the sample (in case of contact heating TL reader), the temperature gradient across the sample and effects of the inert exchange gas in the chamber that is applied in order to avoid chemiluminescence (light from oxidation) (Piters and Bos, 1994). Theoretical studies in this direction have contributed to the possibility of making necessary correction to this effect. (Gotlib et al 1984; Betts and Townsend 1993; Piters and Bos 1994). Kitis and Tuyn (1998) provided equation that can be used for evaluation of the temperature lag between the heating element and the dosimeter to be:    T ij Ti  c ln   2.32     j  where T is peak maximum temperature at very low heating rate i , T j is the corrected i Tm2 Tm peak maximum temperature at heating rate  j and c  1 . Tm1 and Tm2 are the ln 2 peak maximum temperature for the first two lower heating rates 1 and 2 . These authors demonstrated that Eq. 2.32 can be used for correction of both first and general order kinetics. However, temperature lag correction is mostly necessary when thermal quenching parameters are to be calculated. 44 UNIVERSITY OF IBADAN LIBRARY 2.6.2.7 Feldspar inclusion Inclusion of substantial quantity of feldspar in quartz material to be used for luminescence dating or study is always problematic. This is because feldspar is relatively more sensitive than quartz (Duller, 1997). Therefore, its considerable inclusion in quartz will definitely undermine the overall result of the experiment. Usually, this is checked by exposing aliquots to infrared (IR) at ambient temperatures, in order to check if there is an infrared stimulated luminescence (IRSL) signal. This is possible because feldspar is highly sensitive to IR stimulation and whereas to which quartz is insensitive. Thus, IRSL enables discrimination between the presence of the two. Alternatively, or in addition, a light microscope is normally used to estimate level of infections (Spooner and Questiaux, 1989; Duller, 2003). 2.7 Computerised curves deconvolution Glow curves obtained from TL measurements, of quartz in particular, are usually of several overlapping glow peaks in nature. This consequently makes separation of the composite peaks to their individual glow peak difficult, since this is o required for analyses. In reality, the 110 C TL peak of quartz is always isolated and glaring, but the thermal stable peaks are the major problem of overlapping. On the other hand, a complete composite nature of all the components of LM-OSL curves is always observed in nature. One major contribution to this feature is the fact that the entire components originate from the beginning of the OSL curve, at t = 0. What is desirable in basic research and luminescence applications is the individual peak/component of the glow or OSL curve. This is due to the reason that evaluation of the charge stored in respective trap, which corresponds to each peak/component, is obtained from the area under each peak/component. Furthermore, important information, like trapping parameters, about the traps and subsequently about the crystal in general are deduced from the shape of each glow peak or OSL component. Therefore, it is essential to separate each glow or OSL curve into their individual glow peak or component respectively. The computerized curve deconvolution (CCD) analysis is the general term used for doing this. CCD has proved to be promising and more easily applied than the experimental approach that is used determining the number of peaks in a complex TL 45 UNIVERSITY OF IBADAN LIBRARY glow curve of quartz (Horowitz and Yossian, 1995; Kitis et al., 1998; Kitis, 2001, Afouxenidis et al., 2012). There is no available experimental procedure for isolating the components of OSL curve. Analytical expressions used for CCD are based on the TL and OSL kinetic equations. However, the expressions are transformed to the forms that have some experimental measurable parameters unlike the conventional kinetic equations. n0 and s are replaced with Im and Tm in TL kinetics, whereas, n0 and s replaced with Im and tm in LM-OSL kinetics. Tm is the temperature at glow-peak maximum intensity, Im while tm is the time, t, at component maximum intensity, Im for LM-OSL. General order kinetics for both TL and LM-OSL are often used for CCD. This is because both general and mixed orders reduce to first order form for TL and LM- OSL cases when the order of kinetics b =1.00001. The transformed expressions of general order for TL and LM-OSL are as follows: for TL b  E T T  I (T )  I b b1m exp m    kT T  m  b 2.33   T 2  E T T   b1  b 11  exp m   Z2   m   T  kT Tm m   with 2kT / E, m  2kTm / E, Zm 1 b 1m and the frequency factor s is calculated using 1 E  E  2kT  s  exp  1 b 1 m  2.34 kT 2 kTm  m  E  And for LM-OSL b  2 1bt  b 1 t  b 1  I t  I   m    2.35 t  m 2b  tm  2b  46 UNIVERSITY OF IBADAN LIBRARY The quality of the fit produced by fitting any of these equations to experimental curve is normally tested with figure of merit (FOM) defined by Balian and Eddy (1977) as  I j  I T j  FOM % j100 2.36  I j j where Ij and I(Tj) are the experimental and fit intensities in channel j respectively. The background signal is usually simulated by an equation of the form BKGLM   ct 2.37 where  is the average in the first few seconds of a zero dose LMOSL measurement, and c is a constant. 2.8 Retrospective dosimetry The term „Retrospective Dosimetry‟ generally has to do with determination of the dose of absorbed radiation to environment or locally available material in situations where conventional, synthetic dosimeters were not in place at the time of radiation exposure. Luminescence retrospective dosimetry is applied in dating (of archaeological and geological materials) and accident dosimetry. The quantity of interest in these two areas of applications is absorbed or equivalent dose, ED (in Gy). In dating, the age of the material is evaluated by measuring the ED that the materials received from radiation owing to the natural background, since they were last heated (as in TL) or exposed to sunlight (as in OSL), depending. Thereby, the age is calculated by dividing the ED by the dose rate (in Gy/s) of the natural background. In accident dosimetry, the goal is to reconstruct ED as a consequence of a radiation accident. The techniques used in accident dosimetry and dating applications are identical. Both TL and OSL are routinely used in retrospective dosimetry; even in some cases, the two approaches are used as complementary methods There are three major methods generally employed in calculating ED in luminescence retrospective dosimetry. They are additive-dose, regenerative-dose and pre-dose methods (Wintle, 1997; Bøtter-Jensen et al., 2003a). i. Additive-dose method:- in this, the dose response curve (from which ED is evaluated) is constructed from measurements of luminescence signals due to 47 UNIVERSITY OF IBADAN LIBRARY both naturally/accidentally-received-dose and those from additional artificial doses that are added on the naturally/accidentally-received-dose using multiple aliquots. ii. Regenerative-dose method:- this method is based on giving series of laboratory dose Di on a single aliquot of the sample to be dated following pre-heating and OSL measurements. This procedure is repeated for number of times and by varying the regeneration doses, a dose response curve (also known as a growth curve) showing how the OSL signal grows with radiation dose can be constructed. Interpolating the natural OSL signal onto this growth curve provides way of estimating ED. This method in particular is popular in luminescence dating. It must be noted here that the ED that are calculated in this method and previous method, are from the charges that are trapped at the thermally stable traps of the material. Thermally stable is the sense that the electrons that are captured in them remain trapped for a long 8 o o period of 10 years at 20 C. For quartz, traps that are only depleted at 200 C o and above are considered to be thermally stable. The 325 C TL peak that is often used for both TL and OSL dating in quartz and has a lifetime of about 7 3x10 years (Chen and McKeever 1997). iii. Pre-dose method:- Unlike it is with the first two methods, ED estimation is o achieved by employing thermally unstable trap, 110 C peak in quartz that has half-life of about 2 hours at RT. This is made possible by taking the advantage of changing in sensitivity of this peak due to previous irradiations and thermal treatments. Thus, the sensitivity changes serve as a memory tool that records past irradiations from which ED estimation is made. It must be noted here that apart from its application in retrospective dosimetry, pre- dose technique has been effectively utilized in authenticity testing of artefacts and firing temperature measurements of pottery materials (Bailiff, 1994; Koul et al., 1996; Galli et al., 2006; Li and Yin, 2001). 2.9 TL/OSL reader The most widely used instrument for measuring TL and OSL is “The Risø automated TL/OSL reader”. Both TL/OSL measurement and beta irradiation facilities 48 UNIVERSITY OF IBADAN LIBRARY are included in its compartment. The remaining part of this chapter is dedicated to the full description of this reader. 2.9.1. The Risø automated TL/OSL reader The basic components of the Risø TL/OSL reader are as follows: i. a light detection system ii. a heating system for TL measurements iii. optically stimulation units for OSL measurements and iv. a beta irradiator facility. The reader is a computer-controlled system. It enabled measurement of both TL and OSL. With the system, up to 48 samples can be individually: o i. heated to any temperature between room temperature and 700 C, 90 90 ii. irradiated by a beta source ( Sr/ Y) mounted on the reader and iii. optically stimulated by various light sources in situ. The emitted luminescence is measured by a light detection system which comprised of a photomultiplier tube and suitable detection filters. A schematic diagram of the system is shown in Figure 2.17 Either 9.7 mm diameter aluminium disc or stainless steel cups are used as sample holders for all measurements. While aluminium disc can only be used for TL o measurement up to 500 C, stainless steel cup which can withstand higher temperature o is used for measurements that required heating to 700 C. To ensure that the sample are fastened to the aluminium disc, fine grains samples are mounted on it by deposition through suspension method (Aitken, 1998), while silicone oil/grease/spray is used as a glue for loose grains. 49 UNIVERSITY OF IBADAN LIBRARY Fig. 2.17: Schematic diagram of the Riso TL/OSL luminescence reader used for the measurements 50 UNIVERSITY OF IBADAN LIBRARY Samples in discs or cups are loaded onto an exchangeable sample carousel that could accommodate up to 48 samples. The sample carousel is then placed in the sample chamber. The chamber had a nitrogen atmosphere maintained by a nitrogen flow. The sample carousel rests on a motor driven turntable, which enables rotation (in steps) of the sample carousel. Rotation is computer controlled and position holes drilled though the carousel in close proximity to the sample positions enable the system to keep track of the position of the carousel using optoelectronics and a stepper motor. An infrared light emitting diode (LED) is positioned underneath the turntable, which is switched on during rotation. The measurement is initiated by moving a given sample to the measurement position located directly underneath the light detection system. The sample is then lifted through slots in the sample carousel into the measurement position. In the measurement position the sample can be stimulated thermally and/or optically. Thermal stimulation is obtained by linearly increasing the temperature of the heater strip and optical stimulation is provided by different light sources focused onto the sample position. In both cases, the emitted luminescence is measured by the light detection system during the stimulation. 2.9.1.1 Heating system The heating element that is combined with lift mechanism is located directly underneath the photomultiplier tube. The heating element functions as heater for heating the samples and as well as elevator for lifting the sample into the measurement position. The heater strip is made of Kanthal (a high resistance alloy) which is U- formed to provide good heat transmission to the sample and to lift it securely and reproducibly into the measurement position. Heating is accomplished by feeding a controlled current through the heating element. Feedback control of the temperature employs an Alumel-Cromel thermocouple mounted underneath the heater strip. Heating is provided by a non-switching continuous full sine wave generator operating at 20 o kHz. The heating system is able to heat samples to 700 C at linear heating rates from 0.1 to 30 K/s. The heating strip can be cooled by a nitrogen flow which also protects the heating system from oxidation at high temperatures. 2.9.1.2 Optical stimulation system The reader is equipped with a choice of two stimulation sources as shown in Figure 2.18). They are: 51 UNIVERSITY OF IBADAN LIBRARY Fig. 2.18: Schematic diagram of the combined blue and IR LED OSL unit. The unit contains 28 blue LEDs (in 4 clusters) emitting at 470 nm delivering »40 mW/cm2 at the sample and 21 IR LEDs (in three clusters) emitting at 875 nm delivering »135 mW/cm2 at the sample. 52 UNIVERSITY OF IBADAN LIBRARY i. blue light emitting diodes (LEDs) and ii. infrared (IR) LEDs Stimulation at different intensities is possible with the blue LEDs. This is achieved by varying the stimulation intensity as a function of stimulation time. The array of LEDs is equipped with an optical feedback servo-system to ensure the stability of the stimulation power for the delivered light intensity by the blue LEDs. Stimulation in CW-mode as well as LM-mode is possible. The LEDs are arranged in clusters, which are mounted concentrically in a ring-shaped holder located between the heater element and the photomultiplier tube. The holder is designed in such a way that all individual diodes are focused on the sample. The distance between the diodes and the sample is approximately 20 mm. 2.9.1.2.1 Blue LEDs The blue LEDs (NISHIA type NSPB-500s) are with a peak emission at 470 nm (FWHM = 20 nm). They have an emission angle of 15 degrees and a power output of ≈ 2 cd at 20 mA (Botter-Jensen et al.,1999). The energy fluence rate at a distance of 2cm 2 is 19mW/cm . The blue LEDs are usually arranged in 4 clusters each containing seven individual LEDs. The total power from 28 LEDs is > 40 mW/cm2 at the sample position (Botter-Jensen et al., 2003b). To reduce the amount of directly scattered blue light reaching the light detection system, a green long pass filter (GG-420) is incorporated in front of each blue LED cluster. The filter effectively attenuates the high energy photons from the blue LEDs at the cost of approximately 5% attenuation of the peak centred on 470 nm. Figure 2.19 displays the measured LED emission spectrum compared with the published transmission curve for the GG-420 filter and the U-340 detection filter. 2.9.1.2.2 Infrared LEDs Infrared (IR) stimulation in the region 800-900 nm can stimulate luminescence from most feldspars (but not from quartz at room temperature) probably by a thermal assistance mechanism (Hutt et al., 1988). This has the important advantage that a wider range of wavelengths for the detection window becomes available. The IR LEDs used here emit at 875 nm, which is close to the IR resonance wavelength at 870 nm found in most feldspars (Botter-Jensen et al., 2003b). The IR LEDs are arranged in 3 clusters 53 UNIVERSITY OF IBADAN LIBRARY each containing seven individual LEDs. The maximum power from the 21 IR LEDs is 2 approximately 135mW/cm at the sample position (Botter-Jensen et al., 2003b). 2.9.1.3 Photon detector system A photomultiplier tube (PMT) and suitable detection filters constitute the essential components of the light detection system of the Riso reader. While the PMT detects emitted luminescence, the importance of the detection filters is to define the spectral detection window and to shield the PMT from scattered stimulation light. 2.9.1.3.1 Photomultiplier tube The light sensitive component in the PMT is the cathode. This is coated with a photoemissive substance CsSb and other bialkali compounds are commonly used as substitute for this material. Typically, ten photons in the visible range striking the cathode are converted into one to three electrons. Electrons emitted from the photocathode were accelerated towards a series of dynodes maintained at a positive voltage relative to the photocathode. Electrons with sufficient velocity striking the dynode will eject several secondary electrons from the surface. The PMT attached to the Riso TL/OSL luminescence reader is a bialkali EMI 9235QA PMT, which has maximum detection efficiency at approximately 400 nm, making it suitable for detection of luminescence from both quartz and feldspar. The PMT is operated in “photon counting" mode, where each pulse of charge arising at the anode is counted. As the stimulation sources have to be placed between the sample and the PMT the sample-to-PMT cathode distance in the Riso TL/OSL luminescence reader is 55 mm, giving a detection solid angle of approximately 0.4 steradians. 2.9.1.3.2 Detection filters Detection filters define detection window for the PMT and thereby preventing unwanted light and scattered stimulation light from reaching the PMT. This is important because the spectral stimulation and detection windows must be well separated since the intensity of the stimulation light is ~ 1018 orders of magnitude larger than the emitted luminescence. Quartz has a strong emission centred on 365 nm (near UV) and many types of feldspars have a strong emission centred on 410 nm (violet). In Figure 2.20 emission spectra from several samples of sedimentary quartz and K feldspars are shown. A commonly used detection filter is Hoya U-340 (Figure 2.19), which has a peak transmission around 340 nm (FWHM = 80 nm). 54 UNIVERSITY OF IBADAN LIBRARY 2.9.1.4 Beta irradiator A schematic drawing of the irradiator unit is shown in Figure 2.21. It is a detachable beta irradiator located above the sample carousel. The irradiator normally accommodates a 90Sr/90Y beta source, which emits beta particles with a maximum energy of 2.27 MeV. The source strength is about 40 mCi, which gives a dose rate in quartz at the sample position of approximately 0.075 Gy/s at the time of measurements in this research. The source is mounted into a rotating, stainless steel wheel, which is pneumatically activated. The source-to-sample distance should be as small as possible to provide the highest possible dose rate at the sample, however any spatial variations in dose rate across the source will be accentuated at small source-to-sample is instances, so a compromise is required. The distance between the source and the sample is 5 mm. A 0.125 mm beryllium window is located between the irradiator and the measurement chamber to act as vacuum interface for the measurement chamber. The cross-talk, i.e. the percentage of dose given to an adjacent non-irradiated sample is, on average, 0.1735 2±0.0004% for a 4 mg mono-layer coarse grain quartz sample using the new 48-sample carousel. The 2nd nearest sample only receives 0.0042±0.002% of the primary beta dose. 55 UNIVERSITY OF IBADAN LIBRARY Fig. 2.19: The emission spectrum of blue LEDs. Also shown are the transmission curves for the GG-420 green long pass filter (cut-off filter in front of the blue LEDs) and the Hoya U-340 filter (detection filter in front of the PMT); after Botter-Jensen et al. (1999). 56 UNIVERSITY OF IBADAN LIBRARY Fig. 2.20: Emission spectra of sedimentary quartz and K feldspars (from Huntley et al., 1991). (a) Emission spectra of several sedimentary quartz samples from South Australia obtained for stimulation using the 647 nm line from a Krypton laser. (b) Emission spectra of several sedimentary K feldspars using IR diode stimulation. 57 UNIVERSITY OF IBADAN LIBRARY 90 90 Fig. 2.21: Schematic diagram of the cross section of the beta irradiator. The Sr/ Y source is placed in a rotating stainless steel wheel, which is pneumatically activated. The source is shown in the on (irradiating) position. When the source is off the wheel is o rotated 180 , so that the source points directly at the carbon absorber. 58 UNIVERSITY OF IBADAN LIBRARY CHAPTER THREE MATERIALS AND METHODS 3.1 Sample collection The original quartz samples used were large crystal quartz. This was done in order to guarantee the same radiation, optical and thermal histories for quartz grains obtained from the same large crystal instead of using sedimentary quartz grains that could have different origins. This is because it is likely that grains of various sizes to originate from different source areas and their mixing may be due to diverse transport modes (short and long distance transport, respectively) (Preusser et al., 2009). Individual grains may have experienced a different sedimentary history; some grains may have been transported over long distances and repeatedly been recycled while others may originate more-or-less directly from in-situ weathered bedrock. This complex origin in known to cause differences in the concentrations of point defects among individual quartz grains; this plays a role in the highly variable luminescence characteristics of single grains (Preusser et al., 2009). The quartz samples used were twelve in numbers and were collected from southwestern Nigerian. In the preliminary measurements, only two of the samples that possessed highest sensitivity (Figure. 3.1) and at the same time found to be among those that passed feldspar contamination test (Figure. 3.2) were selected for this study. The first of the samples was from Oro, in Kwara State and named S2, whereas the other was from Ijero-Ekiti in Ekiti State named S4 all from part of Nigerian basement complex. They were large crystals of hydrothermal origins which occurred in veins associated with metamorphic rocks. The two quartz samples were clear rock crystal. 3.2 Sample preparation Grains of dimensions 90-150μm were obtained from each of the natural crystal quartz samples after, smashing in an agate mortal, sieving and rinsed in acetone. Each of the two selected sample types was divided into two. The first sets were unannealed o while the second sets were annealed at 900 C for 1 hour and allowed to cool immediately to the room temperature in the air. This annealing temperature was o selected because it is higher than the temperature of about 870 C, where an irreversible 59 UNIVERSITY OF IBADAN LIBRARY phase change takes place in quartz (Preusser et al., 2009). Therefore, the set of annealed samples represent quartz grains extracted from fired quartz-grained inclusion materials like ceramics, potteries, fired sediments and bricks, metallurgy ovens, and other highly fired objects for dating. On the other hand, the unannealed samples represent set of samples to be dated that have not been artificially heated beyond irreversible phase o change temperature (870 C) like sedimentary quartz grains that are extracted from sediments or overbank deposits. All the aliquots (sub-samples) used for the measurements were of equal mass (about 5mg). 3.3 Instrumentation The TL and OSL measurements were carried out using the automated Risø TL/OSL reader (model TL/OSL DA-15) systems at Archaeometry Laboratory, Cultural and Educational Technology Institute (C.E.T.I.), R.C. “ATHENA”, Xanthi, Greece. An overall description of the readers was presented in section 2.9. 3.4 Feldspar inclusion test It was considered necessary to confirm weather there was no feldspar inclusion in the two quartz samples used in this study. This was undertaken by conducting Infra- Red Stimulation Luminescence (IRSL) measurements on samples with dose of Natural + 5Gy. Quartz is known not to be sensitive to infra-red stimulation, while feldspar is (Preusser et al., 2009). The result of this is shown in Figure 3.2. It was evidence from this that S2 and S4 were not infected with feldspar while S5, S11 and S12 were contaminated. 3.5 Experimental procedures The TL/OSL measurements carried out in this work were executed in two phases, namely: i. oReproducibility study of pre-dose sensitisation of the 110 C TL peak. ii. Study on luminescence sensitisations in unannealed and annealed quartz samples 60 UNIVERSITY OF IBADAN LIBRARY 6 8x10 6 7x10 6 6x10 6 5x10 6 4x10 6 3x10 6 2x10 6 1x10 0 S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 Samples Fig. 3.1: Luminescence sensitivity of all the samples to 5Gy dose of radiation 61 UNIV TL Integral (a.u)ERSITY OF IBADAN LIBRARY 3 S1 2.5x10 S2 S3 S4 3 2.0x10 S5 S6 S7 3 1.5x10 S8 S9 S10 3 1.0x10 S11 S12 Bakground 2 5.0x10 0.0 0 20 40 60 80 100 Stimulation time (s) Fig. 3.2: Infra-Red Stimulation Luminescence (IRSL) curves of all the samples to test for feldspar contamination. 62 UNIV IRSL (a.u)ERSITY OF IBADAN LIBRARY All measurements were done in a nitrogen atmosphere with different constant heating rates as indicated in the measurement protocols, up to a maximum heat temperature of o 500 C. The test dose and range of doses administered to each of the quartz samples are accordingly indicated in the respective experimental protocols presented in following subsequent sections. o 3.5.1 Reproducibility study of pre-dose sensitisation of the 110 C TL peak Unannealed and annealed sample were used in this section by following the sequence of the experimental protocol presented below. 3.5.1.1 Measurement protocols Step 1: A TD was given to a sample. o Step 2: TL readout up to 180 C. o Step 3: TD and TL readout up to 500 C. o Step 4: TD and TL readout up to 500 C. o Step 5: TD and TL readout up to 500 C. Step 6: Repeat steps 1 to 5 for 10 different aliquots of the same quartz sample. A dose of 2Gy was the TD administered to unannealed S2 and S4 while 1 and 0.5 Gy dose were given to their annealed counterpart respectively. At least 20 aliquots were prepared for the complete run of the above protocol for each of the samples making a total of 40 set of runs. 3.5.1.2 Description of the protocol i. oSteps 1 and 2 measured the sensitivity Sn01 of the TL glow-peak at “110 C” acting additionally as a mass normalization. o ii. Step 3 measured (a) the sensitivity Sn02 of the TL glow-peak at 110 C and o (b) acts as a thermal activation up to 500 C for the sample to ensure that all the stable straps that are used for TL and OSL dating are completely depleted. o iii. Step 4 measured (a) the sensitivity Sn1 of the TL glow-peak at 110 C and (b) acts as a thermal activation for the sample again. iv. oStep 5 measured the further sensitisation Sn2 of the TL glow-peak at 110 C. 63 UNIVERSITY OF IBADAN LIBRARY v. Step 6 allowed sensitisation test among 10 different aliquots of the same quartz kind. 3.5.2 Study on luminescence sensitisations in unannealed and annealed quartz samples o Both unannealed and samples annealed at 900 C were used in this section by following the sequence of the experimental protocol presented below. 3.5.2.1 Measurement protocol Part A Step 1: TD was given to an aliquot. o o Step 2: TL readout up to 180 C at heating rate of HRi ( C/s). Step 3: Another TD was given to the same aliquot o o Step 4: TL (Sn0) readout up to 500 C at heating rate of HRi ( C/s) Step 5: Steps 3 and 4 were repeated four times on the same aliquot leading to signals Sn1, Sn2, Sn3 and Sn4 respectively. Step 6: Another TD was given to the same aliquot o o Step 7: TL (Sn5) readout up to 500 C at heating rate of 2 C/s Step 8: Steps 1-7 were repeated for fresh aliquot and new HRi (Taking HRi= o 0.25, 0.5, 1, 2, 5, and 10 C/s) Part B Step 1: TD was given to a fresh aliquot. o o Step 2: TL readout up to 180 C at heating rate of HRi ( C/s). o o Step 3: TL readout up to 500 C at heating rate of HRi ( C/s) Step 4: Steps 3 was repeated four times more on the same aliquot Step 5: Another TD was given to the same aliquot o o Step 6: TL (Sn1b) readout up to 500 C at heating rate of HRi ( C/s) Step 7: Another TD was given to the same aliquot o o Step 8: TL (Sn2b) up to 500 C at heating rate of 2 C/s 64 UNIVERSITY OF IBADAN LIBRARY Step 9: Steps 1-8 were repeated for fresh aliquot and new HRi (Taking HRi= o 0.25, 0.5, 1, 2, 5, and 10 C/s) Part C Step 1: TD was given on an aliquot. o o Step 2: TL readout up to 180 C at heating rate of HRi ( C/s). Step 3: Another TD was given to the same aliquot Step 4: OSL (Sn0) measurement at RT for 1000s o o Step 5: TL readout up to 500 C at heating rate of HRi ( C/s) Step 6: Steps 3-5 were repeated four times more on the same aliquot leading to OSL signals Sn1, Sn2, Sn3 and Sn4 respectively. Step 7: Steps 1-6 were repeated for fresh aliquot and new HRi (Taking HRi= o 0.25, 0.5, 1, 2, 5, and 10 C/s) Part D Step 1: TD was given to an aliquot. o o Step 2: TL readout up to 180 C at heating rate of HRi ( C/s). Step 3: OSL measurement at RT for 1000s o o Step 4: TL readout up to 500 C at heating rate of HRi ( C/s) Step 5: Step 4 was repeated four times more on the same aliquot Step 6: Another TD was given to the same aliquot Step 7: OSL (Sn1b) measurement at RT for 1000s o o Step 8: TL readout up to 500 C at heating rate of HRi ( C/s) Step 9: Steps 1-8 were repeated for fresh aliquot and new HRi (Taking HRi= o 0.25, 0.5, 1, 2, 5, and 10 C/s) The same TD of 7.5Gy was for both S2 and S4 unannealed samples while 1 and 0.5 Gy were used for S2 and S4 annealed samples respectively. 65 UNIVERSITY OF IBADAN LIBRARY 3.5.2.2 Descriptions of the protocol Part A i. Steps 1and 2 were meant for mass normalization. ii. TD of Steps 1and3 served as pre-exposure dose (these added to the naturally acquired dose for unannealed samples). iii. Step 4 measured the unsensitised TL (named as Sn0) and at the same time o serves as thermal activation to 500 C for the next TL of step 5. iv. Step 5 measured four (4) sensitised successive cycles TL namely; Sn1, Sn2, o Sn3 and Sn4 respectively. Each TL also serves as thermal activation to 500 C for the next successive cycle TL. v. TD of step 6 served as a pre-exposure dose for step 7. vi. Step 7 measured TL (Sn5) sensitisation resulting from successive cycles of irradiations and TL readouts of (i) to (v) and heating rates. Note this step 7 o is always read at 2 C/s heating rate. vii. Step 8 meant to observe the effect of different levels of thermal activation resulting from different heating rates. Part B. The description of Part B is just like that of Part A above. The only difference comes from the fact that intermediate irradiations of Part A step 3 were missing. That was devised to observe pure thermal sensitisation contribution which TL (Sn1b) of step 6 represented while TL (Sn2b) of step 8 is similar to its counterpart of Part A step 7 as described in (vi) above Parts C and D Parts C and D that are for OSL, are similar to Part A and Part B respectively, but instead of TL, RT-LMOSL measurements were employed to measure the sensitivity as indicated in the protocol. 66 UNIVERSITY OF IBADAN LIBRARY 3.6 Computerised curves deconvolution analyses The RT-LMOSL curves were analyzed through a Computer Glowcurve De- convolution Analysis (CGDA) using a general order kinetics expression proposed of Bulur (1996) and later modified by Kitis and Pagonis (2008) into an expression containing only the peak maximum intensity Im and the corresponding time tm. These two variables can be extracted directly from the experimental OSL curves. The modified expression used in the computerized procedure is: b  2 1bt  b 1 t  b 1  I t  I   m  t  2b m  t   m  2b   3.1 where b is the order of kinetics. This latter expression was used providing thus the best test of the first order model assumption correctness for the quartz LM-OSL, since the values of b were selected to be 1.0001. The background signal was simulated by an equation of the form BKGLM   ct 3.2 where  is the average in the first few seconds of a zero dose LMOSL measurement, and c is a constant. The non-zero intensity value at T = 0 or RT in TL and t = 0 in RT LMOSL, was considered to be included in the CGDA analysis in order to accomplish good fit. The value of this now-zero intensity can be observed to be dependent on sample and the entire TL glow curve or OSL curve integral which is a function of applied dose or sensitization factor. See figures 1 to 4 in Kiyak et al, (2007), figure 3 in Polymeris et al., (2009), figures 2 and 3a in Jain et al., (2003) and figures 1 and 2 in Appendix 1). However, this unwanted signal is of insignificant value in TL dating since it can be easily excluded from the desired TL peak generally. Equally, this value is not observed o in LMOSL received at 125 C of one of the present understudied samples which would have made evaluation of fast component used in OSL dating difficult (Oniya et al., 2012a). Hence, the unwanted initial signal is in this work attributed to the signal arising from the decay of extremely shallow traps that decay in RT. Because (i) they decay in RT and do not require optically stimulation (ii) the signal is eliminated by pre-heat to o o 180 C and elevated LMOSL measurement at 125 C in this work. Therefore, in the same direction with Kitis et al., (2010) that this non zero intensity value could be due to 67 UNIVERSITY OF IBADAN LIBRARY the presence of phosphorescence, a first order phosphorescence component of the form, I t I exp  t , was included in the CGDA analysis. The decay rate, , was o numerically adjusted until good fits were achieved for all signals. All curve fittings were performed using the software package Microsoft Excel, with the Solver utility (Afouxenidis et al., 2012), while the goodness of fit of all the curves fitted was tested using the Figure Of Merit (FOM) of Balian and Eddy (1977) given by. YExper YFit FOM  i A 3.3 where YExper is the experimental glow-curve, YFit is the fitted glow-curve and A is the area of the fitted curve. The FOM values obtained were less than 1% for the RT curves. 68 UNIVERSITY OF IBADAN LIBRARY CHAPTER FOUR RESULTS AND DISCUSSIONS 4.1 Introduction Results and discussions of TL/OSL measurements on the two quartz samples investigated in this study, as described in chapter three, are presented in this section. o 4.2 Reproducibility study of pre-dose sensitisation of the 110 CTL peak According to the experimental protocol of section 3.5, the sensitivity Sn01 is the “natural” sensitivity obtained without any treatment. The sensitivity Sn02 is due to the o heating up to 180 C, for which the sensitisation is considered to be negligible. So, 180 o o C is generally adopted as the temperature used to erase the TL glow-peak at 110 C in most of the TL/OSL protocols (Polymeris et al., 2009). The sensitivities Sn1 and Sn2 o are due the first and second thermal activation up to 500 C respectively. 4.2.1 Sensitisation in unannealed samples Figures 4.1 and 4.2 present glow curves of S2 and S4 samples respectively. The most interesting observation could be seen from the different level of sensitisation that took place in various aliquots of the same quartz sample. In these figures, curve (a) represents the glow-curve of natural TL (NTL) plus the test dose (Sn02, step 3, without any thermal activation). Both curves (b) and (c) correspond to the glow-curve after the first thermal activation (Sn1, step 4). However, since the sensitisation level is not the same throughout all the aliquots of each quartz sample, curve (b) corresponds to the o aliquot for which the minimum sensitivity Sn1 of 110 C TL peak occurred, while curve (c) stands for aliquot with maximum Sn1 sensitivity. The degree of scatter of the pre- o dose sensitisation of the 110 C TL peak in 10 aliquots for the two understudied quartz samples is better illustrated in Figures 4.3 and 4.4, for unannealed S2 and S4 samples respectively. To demonstrate that the above scattered is as a result of 69 UNIVERSITY OF IBADAN LIBRARY 5 10 c b 4 10 a 3 10 2 10 0 100 200 300 400 500 o Temperature ( C) Fig. 4.1: TL glow curves for unannealed S2 samples. Curve (a) corresponds to the sensitisation without previous thermal activation (Sn02). Both curves (b) and (c) correspond to the glow-curve after the first thermal activation (Sn1) indicating the minimum and maximum sensitisation respectively. 70 UNIVE TL (a.u)RSITY OF IBADAN LIBRARY c 5 10 b 4 10 3 10 a 2 10 0 100 200 300 400 500 o Temperature ( C) Fig. 4.2: TL glow curves for unannealed S4 samples. Curve (a) corresponds to the sensitisation without previous thermal activation (Sn02). Both curves (b) and (c) correspond to the glow-curve after the first thermal activation (Sn1) indicating the minimum and maximum sensitisation respectively 71 UNIVE TL (a.u)RSITY OF IBADAN LIBRARY 6 6x10 6 5x10 6 aliquot.1 4x10 aliquot.2 aliquot.3 aliquot.4 6 3x10 aliquot.5 aliquot.6 aliquot.7 6 aliquot.8 2x10 aliquot.9 aliquot.10 6 1x10 0 S02 S1 S2 Sensitization Fig. 4.3: The sensitisations resulting from three successive irradiations and luminescence readings for 10 aliquots for unannealed S2 sample 72 UNIVE TL integral (a.u)RSITY OF IBADAN LIBRARY 7 1.5x10 aliquot.1 aliquot.2 aliquot.3 7 1.0x10 aliquot.4 aliquot.5 aliquot.6 aliquot.7 6 aliquot.8 5.0x10 aliquot.9 aliquot.10 0.0 S02 S1 S2 Sensitization Fig. 4.4: The sensitisations resulting from three successive irradiations and luminescence readings for 10 aliquots for unannealed S4 sample. 73 UNIV TL integral (a.u)ERSITY OF IBADAN LIBRARY sensitisation, unsensitised glow curves from two measurements on two aliquots for each of the two understudied samples are presented in Figure 4.5 and 4.6 respectively. It is clear from these curves that the responses were excellent for S2 and S4 samples. The Coefficient of Variation (CV), defined as the ratio of standard deviation to the mean, in all cases were less than 5% Normalisation is routinely employed in luminescence studies and dating for the purpose of attaining good reproducibility (Murray and Roberts, 1998). To observe the influence of normalization on the sensitisation reproducibility, all sensitivities were normalized over the “natural” sensitivity Sn01; this ratio of artificial sensitivities over the “natural” sensitivity (i.e. normalization) will be called sensitisation factor hereafter. The results concerning the reproducibility of the pre-dose sensitisation effects are summarized in Table 4.1. The second column [nS01 (%)] describes CV of sensitisation of 10 aliquots. This column represents the reproducibility of unnormalized and unsensitised signal of the two samples. As seen here, the level of scatter is much in S2 than S4. In the cases of the unannealed quartz the sample reproducibility can be also measured through their NTL signal. This is shown in the third column [NTL (%)]. The excellent reproducibility of fourth column [Sn02/Sn01] in Table 4.1 is a product of normalization that enhances level of reproducibility among different luminescence measurements. The normalization improved the CV from 12.44 to 1.46% and 5.40 to 1.06% for S2 and S4 samples respectively. It should be recalled that since the pre-heat o was only to 180 C up to stage S02, the reproducibility here represents unsensitised signals. Of main interest were the results of column [Sn1/Sn01], which gave the o variation of the sensitisation factor due to first thermal activation up to 500 C. High levels CV up to 33.7 and 51.99% were introduced into reproducibility by sensitisation even after normalization for S2 and S4 samples respectively which indicated that normalization to first thermal activation is not ideal to improve sensitisation reproducibility. Although, slight variations in the sensitisation between different aliquots of the same quartz type for the two quartz types could be expected. Of truth, this variation was well beyond expectation, based on the original state and origin of the samples, as well as, on the good reproducibility resulted from unsensitised values 74 UNIVERSITY OF IBADAN LIBRARY 6 2.5x10 6 2.0x10 6 1.5x10 6 1.0x10 5 5.0x10 aliquot 1 aliquot 2 0.0 -20 0 20 40 60 80 100 120 140 160 180 200 o Temperature ( C) Fig. 4.5: TL test of two runs for S2 samples 75 UNIV TL (a.u)ERSITY OF IBADAN LIBRARY 5 1.6x10 5 1.4x10 5 1.2x10 5 1.0x10 4 8.0x10 4 6.0x10 4 4.0x10 aliquot 1 4 aliquot 2 2.0x10 0.0 4 -2.0x10 -20 0 20 40 60 80 100 120 140 160 180 200 o Temperature ( C) Fig. 4.6: TL test of two runs for S4 samples. 76 UNIV TL (a.u)ERSITY OF IBADAN LIBRARY Table 4.1: Coefficient of Variation (CV) of Reproducibility sensitisation of 10 aliquots of S2 and S4 quartz samples. Quartz Sample Sn01 NTL Sn02/Sn01 Sn1/Sn01 Sn2/Sn01 Sn2/Sn1 (%) (%) (%) (%) (%) (%) S2 Unannealed 12.44 6.60 1.46 33.74 31.51 6.39 S4 Unannealed 5.40 7.5 1.06 51.99 51.31 4.67 S2 Annealed 6.55 -- 0.72 6.29 7.78 3.24 S4 Annealed 12.03 -- 1.23 9.00 9.86 2.00 Keys: Sn01 = “natural” sensitivity, NTL = natural TL signal, st Sn02/Sn01 = normalised sensitivity, Sn1/Sn01 = 1 normalised and sensitised sensitivity, nd Sn2/Sn01 = 2 normalised and sensitised sensitivity, st nd Sn2/Sn1 = ratio of the 1 and 2 sensitised sensitivities. 77 UNIVERSITY OF IBADAN LIBRARY (second to fourth columns, in Table 4.1). The results of column [Sn2/Sn01], which o provides the variation of the sensitisation due to second thermal activation up to 500 C, possesses exact poor reproducibility as it is in column [nS2/Sn01]. Normalization of the second sensitisation to the first sensitisation, rather to the unsensitised signal of Sn02, o shows the effect of TL readout up to 500 C on the sensitisation reproducibility. This is represented in the last column, [Sn2/Sn1], in Table 4.1. It is apparent from this that TL o readout to 500 C improved the CV in sensitisation reproducibility from 31.51 to 6.39% and 51.31 to 4.67% respectively for S2 and S4 samples. 4.2.2 Sensitisation in annealed quartz The results for annealed quartz samples are also shown in Figures 4.7 and 4.8. o The degree of scatter of the pre-dose sensitisation of the 110 C TL peak in 10 aliquots o for the two quartz samples annealed at 900 C are shown in Figures 4.9 and 4.10. As could be observed from theses figures, outstanding enhancement in reproducibility of pre-dose sensitisation by annealing becomes apparent as compared to the case of unannealed samples. The comparison of this sensitisation reproducibility for the two cases of unannealed and annealing is presented in Table 4.1. In this Table 4.1, while annealing increased the sensitisation reproducibility of un-normalized signal of S2 from CV of 12.44 to 6.55%, annealing rather worsen the sensitisation reproducibility of S4 from CV of 5.39 to 12.03% (Column [Sn01 (%)]). However, the sensitisation reproducibility became excellent after normalization for the two samples after annealing as expected. Column [Sn1/Sn01], also here, provides the variation of the sensitisation factor o due to first thermal activation up to 500 C for the case of annealed samples. Notably, o it comes out that annealing at 900 C for an hour surprisingly removed the high levels of CV observed in the case of unannealed samples from 33.7 to 6.29% and 51.99 to 9.00% for S2 and S4 samples respectively. Exact good reproducibility as in column [Sn1/Sn01] is recorded in column, [nS2/Sn01] for the case of sensitisation due to second thermal o activation up to 500 C. Lastly, a good sensitisation reproducibility was demonstrated in o column, [Sn2/Sn1], and it is an indication of the effect of TL readout up to 500 C on the sensitisation reproducibility as anticipated (column [Sn02/Sn01]). 78 UNIVERSITY OF IBADAN LIBRARY 4.2.3 Discussion The sensitisation of a quartz sample is influenced by its thermal and radiation history (Preusser et al., 2009). The original quartz samples were not of sedimentary origin, but instead large crystals of hydrothermal and metamorphic origins which occurred in vein-associated metamorphic rocks. Consequently, the different grains of the same sample were subjected to the same conditions of both heating and irradiation. Therefore, based on their origin, similar pre-dose sensitisation was expected for various aliquots derived from these samples. In order to explain the large variation in the sensitisation observed in case of various aliquots of a sample the following phenomena are proposed. The sensitisation of the same quartz samples after annealing does not yield the same discrepancy as compared to unannealed case, indicating the importance of heating to reduce the CV of the sensitisation. The most probable cause could be the gridding and milling of the samples. In general, any treatment of quartz by mechanical actions such as grinding, milling, crushing, sawing and cutting, may result in the production of defect structures (McKeever, 1985; Ranjbar, et al., 1999; Takeuchi, et al., 2006; Takeuchi and Hashimoto, 2008). These procedures result mostly in the formation of defects such as dislocations in the outer surface of the grains (Ranjbar, et al., 1999). The surface alterations following a prolonged grinding, which results in significant changes in the physical and chemical characteristics of the material, have been reported in the literature (Battaglia, et al., 1993). Formation of free radicals results from the production of large amounts of new fracture surfaces in small quantities of the material. These new surfaces create broken bonds which lay in regions a few nm below the surface. The defect creation is related to the methods and conditions of treatment (Bartnitskaya et al., 1992). The influence of crushing and milling on TL seems to be minimized if HF treatment is applied, since the near-surface layer is usually removed by HF etching (Bartnitskaya, et al., 1992). In the present study, HF etching was not applied because removal of feldspar was not necessary since the samples are not of sedimentary origin. Therefore the creation and presence of these defects affect the TL 79 UNIVERSITY OF IBADAN LIBRARY c b 5 10 4 10 a 3 10 2 10 0 100 200 300 400 500 o Temperature ( C) Fig. 4.7: TL glow curves for annealed S2 samples. Curve (a) corresponds to the sensitisation without previous thermal activation (Sn02). Both curves (b) and (c) correspond to the glow-curve after the first thermal activation (Sn1) indicating the minimum and maximum sensitisation respectively. 80 UNIVE TL (a.u)RSITY OF IBADAN LIBRARY c 5 10 b 4 10 a 3 10 2 10 0 100 200 300 400 500 o Temperature ( C) Fig. 4.8: TL glow curves for annealed S4 samples. Curve (a) corresponds to the sensitisation without previous thermal activation (Sn02). Both curves (b) and (c) correspond to the glow-curve after the first thermal activation (Sn1) indicating the minimum and maximum sensitisation respectively 81 UNIVE TL (a.u)RSITY OF IBADAN LIBRARY 7 2.5x10 7 2.0x10 aliquot.1 aliquot.2 aliquot.3 7 aliquot.4 1.5x10 aliquot.5 aliquot.6 aliquot.7 7 aliquot.8 1.0x10 aliquot.9 aliquot.10 6 5.0x10 0.0 S02 S1 S2 Sensitization Fig. 4.9: The sensitisations resulting from three successive irradiations and luminescence readings for 10 aliquots for annealed S2 sample. 82 UNIV TL integral (a.u)ERSITY OF IBADAN LIBRARY 7 1.2x10 7 1.0x10 aliquot.1 aliquot.2 aliquot.3 6 8.0x10 aliquot.4 aliquot.5 6 aliquot.6 6.0x10 aliquot.7 aliquot.8 6 aliquot.9 4.0x10 aliquot.10 6 2.0x10 0.0 S02 S1 S2 Sensitization Fig. 4.10: The sensitisations resulting from three successive irradiations and luminescence readings for 10 aliquots for annealed S4 samples. 83 UNIV TL integral (a.u)ERSITY OF IBADAN LIBRARY sensitivity of the unannealed samples. However, certain changes occurring in the quartz o lattice with heating at 500 C, such as annealing of these defects, result in dilution of grinding effects. Therefore, in the case of the annealed samples the sensitisation is much more uniform for all aliquots of the same quartz sample. This latter argument is further supported by the column [Sn2/Sn1] of Table 4.1 of the unannealed samples case. Besides the large variation of the sensitisation factor Sn1 due to first thermal activation o up to 500 C in all quartz samples, the ratio Sn2/Sn1 of various aliquots is consistent. o Therefore, it seems that heating up to 500 C, even though instantaneous, could merely anneal the dislocations, resulting thus in much more uniform sensitisation ratio Sn2/Sn1. 4.2.4 Implications of results o The fact that 110 C TL peak in quartz is not stable, a number of applications aside pre-dose dating technique, have been proposed by taking advantage of either its sensitivity or sensitisation. Among these, Stokes (1994) suggested the applicability of o the 110 C TL peak as a sensitivity correction in a single-aliquot additive dose protocol. o A linear relationship between sensitivity changes in the 110 C TL peak of quartz and in its OSL emission was reported first by Stoneham and Stokes (1991) providing evidence for a strong link between the two luminescence signals. Consequently, one expects that similar CV of sensitisation is expected also for the OSL emission of the same quartz samples. However, implications of this high CV of sensitisation to OSL are limited due to single aliquot procedures applied. The same insignificant implication applies to the pre-dose dating technique, which is theoretically based on pre-dose sensitisation. Because it also involves single aliquot procedure and of the actuality that the technique was originally established on observations on heated quartz extracted from pottery (Zimmerman, 1971) that is identical to annealed quartz. o Nevertheless, the 110 C TL peak in quartz is used in order to perform mass reproducibility check and correction in many multiple-aliquot protocols, such as the foil technique (Michael, et al., 1997). In the framework of similar protocols, a second TL measurement is performed after the main measurement for each aliquot used. The same o dose is applied to all aliquots. The intensity of the 110 C TL peak of all these second TL measurements is used for both mass reproducibility monitoring and correction. Based on the results presented so far in the present study, this procedure could be successfully applied only in the case of heated quartz, although still a relatively increased error is expected. Unfortunately, special caution should be exercised while 84 UNIVERSITY OF IBADAN LIBRARY applying this specific mass reproducibility correction procedure in the case of geological, unannealed quartz samples, since it could result in erroneous equivalent dose estimation. 4.2 Study on luminescence sensitisations in unannealed and annealed quartz samples Experimental protocols of section 3.3.2 in chapter three were employed to obtain all experimental results presented below. o 4.2.1 Sensitisations of 110 C TL peak and RT-LMOSL of unannealed and annealed samples The representative glow curves from measurements performed on unannealed and annealed S2 and S4 samples using Part A of the experimental protocol are shown for the two samples in Figures 4.11 to 4.14 respectively. Their counterparts of RT- LMOSL curves, using Part C of the protocol, are shown in Figures 4.15 to 4.18. It is o evident from these figures that the two phenomena, 110 C TL peak and RT-LMOSL curves, exhibit nearly identical sensitisation pattern both in unannealed and annealed samples of the two quartz samples. Deconvolution to the respective components of the whole RT-LMOSL curves for the two samples is presented in Figures 4.19 to 4.22. Each of the two unannealed st nd samples has four different RT-LMOSL components namely; 1 component (C1), 2 rd th component (C2), 3 component (C3) and 4 component (C4) while these increased to st nd rd th five, components, 1 component (C1), 2 component (C2), 3 component (C3), 4 th component (C4) and 5 component (C5) in annealed samples. The increase in number of RT-LMOSL components from 4 to 5 following annealing is not a new idea. A similar increase in number of TL glow peaks of some Nigerian quartz after annealing has been reported (Oniya et al., 2012b). Therefore, the number of RT-LMOSL components that are associated with optical release of charge from each different trap types that are also responsible for TL peaks is expected to increase too after 85 UNIVERSITY OF IBADAN LIBRARY Sn4 5 2.1x10 Sn3 5 1.4x10 Sn2 4 7.0x10 Sn1 Sn0 0.0 0 50 100 150 o Temperature ( C) Fig. 4.11: Glow curves showing sensitisations resulting from successive irradiations and TL readings of unannealed S2 sample. Curves Sn0, Sn1, Sn2, Sn3, st nd rd th Sn4 represent unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 86 UNIV TL (a.u)ERSITY OF IBADAN LIBRARY Sn4 6 1.2x10 Sn3 Sn2 5 8.0x10 Sn1 5 4.0x10 Sn0 0.0 0 50 100 150 200 o Temperature ( C) Fig. 4.12: Glow curves showing sensitisations resulting from successive irradiations and TL readings of annealed S2 sample. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 87 UNIV TL (a.u)ERSITY OF IBADAN LIBRARY 5 4.0x10 Sn4 5 Sn3 3.2x10 Sn2 5 2.4x10 Sn1 5 1.6x10 4 8.0x10 Sn0 0.0 0 50 100 150 o Temperature ( C) Fig. 4.13: Glow curves showing sensitisations resulting from successive irradiations and TL readings of unannealed S4 sample. Curves Sn0, Sn1, Sn2, Sn3, Sn4 st nd rd th represent unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 88 UNIV TL (a.u)ERSITY OF IBADAN LIBRARY 5 Sn4 6x10 Sn3 5 5x10 Sn2 5 4x10 Sn1 5 3x10 5 2x10 5 1x10 Sn0 0 0 50 100 150 200 o Temperature ( C) Fig. 4.14: Glow curves showing sensitisations resulting from successive irradiations and TL readings of annealed S4 sample. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 89 UNIVE TL (a.u)RSITY OF IBADAN LIBRARY 4 5x10 Sn4 4 4x10 Sn3 4 3x10 4 2x10 Sn2 4 1x10 Sn1 0 Sn0 0 200 400 600 800 1000 Stimulation time (s) Fig. 4.15: OSL curves showing sensitisations resulting from successive irradiations and LMOSL readings of unannealed S2 sample. Curves Sn0, Sn1, Sn2, Sn3, Sn4 st nd rd th represent unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 90 UNIV LMOSL (a.u)ERSITY OF IBADAN LIBRARY 5 1.2x10 Sn4 Sn3 Sn2 4 8.0x10 Sn1 4 4.0x10 Sn0 0.0 0 200 400 600 800 1000 Stimulation time (s) Fig. 4.16: OSL curves showing sensitisations resulting from successive irradiations and LMOSL readings of annealed S2 sample. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 91 UNIV LMOSL (a.u)ERSITY OF IBADAN LIBRARY 4 8.0x10 Sn4 Sn3 4 6.0x10 Sn2 4 Sn1 4.0x10 4 2.0x10 Sn0 0.0 0 200 400 600 800 1000 Stimulation time (s) Fig. 4.17: OSL curves showing sensitisations resulting from successive irradiations and LMOSL readings of unannealed S4 sample. Curves Sn0, Sn1, Sn2, Sn3, Sn4 st nd rd th represent unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 92 UNIV LMOSL (a.u)ERSITY OF IBADAN LIBRARY 4 4.0x10 Sn4 Sn3 4 3.2x10 Sn2 Sn1 4 2.4x10 4 1.6x10 3 8.0x10 Sn 0 0.0 0 200 400 600 800 1000 Stimulation time (s) Fig. 4.18: OSL curves showing sensitisations resulting from successive irradiations and LMOSL readings of annealed S4 sample. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 93 UNIV LMOSL (a.u)ERSITY OF IBADAN LIBRARY 4 5x10 Expt 4 4x10 I(t) C1 4 C2 3x10 C3 C4 4 Bkg 2x10 Fit 4 1x10 0 10 100 1000 Stimulation time (s) Fig. 4.19: RT-LMOSL curves depicting deconvolution of the LMOSL curves to its st respective components unannealed S2 sample. Curves C1, C2, C3, and C4 represent 1 , nd rd th 2 , 3 , and 4 components respectively. Expt. represents experimental data, I(t) represents Bkg. Represents phosphorescence component background and Fit. represents fitted data. 94 UNIVE OSL (a.u)RSITY OF IBADAN LIBRARY 4 8x10 4 7x10 Expt 4 I(t) 6x10 C1 4 C2 5x10 C3 4 C4 4x10 Bkg 4 Fit 3x10 4 2x10 4 1x10 0 10 100 1000 Stimulation time (s) Fig. 4.20: RT-LMOSL curves depicting deconvolution of the LMOSL curves to its respective components for unannealed S4 sample. Curves C1, C2, C3, and C4 st nd rd th represent 1 , 2 , 3 , and 4 components respectively. Expt. represents experimental data, I (t) represents Bkg. Represents phosphorescence component background and Fit. represents fitted data. 95 UNIVE OSL (a.u)RSITY OF IBADAN LIBRARY 70000 Expt 60000 I(t) C1 50000 C2 C3 C4 40000 C5 Bkg 30000 Fit 20000 10000 0 0 200 400 600 800 1000 Stimulation time (s) Fig. 4.21: RT-LMOSL curves depicting deconvolution of the LMOSL curves to its respective components annealed S2 sample. Curves C1, C2, C3, C4 and C5 st nd rd th th represent 1 , 2 , 3 , 4 , 5 components respectively. Expt. represents experimental data, I(t) represents Bkg. Represents phosphorescence component background and Fit. represents fitted data. 96 UNIVE OSL (a.u)RSITY OF IBADAN LIBRARY 50000 Expt 45000 I(t) C1 40000 C2 35000 C3 C4 30000 C5 Bkg 25000 Fit 20000 15000 10000 5000 0 0 200 400 600 800 1000 Stimulation time (s) Fig. 4.22: RT-LMOSL curves depicting deconvolution of the LMOSL curves to its respective components annealed S4 sample. Curves C1, C2, C3, C4 and C5 st nd rd th th represent 1 , 2 , 3 , 4 , 5 components respectively. Expt. represents experimental data, I(t) represents Bkg. Represents phosphorescence component background and Fit. represents fitted data. 97 UNIV OSL (a.u)ERSITY OF IBADAN LIBRARY annealing. The changes in the nature of the glow-curves and RT-LMOSL components after annealing have been attributed to the alterations made to the recombination pathways and competitions during irradiation and heating. As suggested by Bøtter- Jensen et al. (1995) they are as a result of introduction of more traps and recombination centres. From the objective of this study, it is important to associate each RT-LMOSL o component with their corresponding TL peak; most especially 110 C TL peak. Room o temperature (RT)-LMOSL component associated with 110 C TL peak has always been described to be the broad, intermediate and intense component among all the RT- LMOSL components (Kiyak et al., 2008; Polymeris et al., 2009). Based on this, C3 is o the specific component identified with the 110 C TL peak for both S2 and S4 unannealed samples in this work as could be seen in Figures 4.19 and 4.20. This is, o furthermore, confirmed from Table 4.2 which shows the percentage of 110 C TL peak signal to the total TL response and that of each component of the RT-LMOSL signal to o the overall OSL signal of each sample. From Table 4.2, the signal of 110 C TL peak is 99.01 and 99.00% of the total signal of the whole TL emission for S2 and S4 unannealed samples respectively. It is C3 that follows the exact trend of dominating percentage in RT-LMOSL signal with 96.75 and 95.24% of the total signal of the whole RT-LMOSL emission for S2 and S4 unannealed samples respectively. The first two RT-LM-OSL components (C1 and C2) are recognized to be the two fast components which are known to be associated with the 325◦C peak (Jain et al., 2003; Kiyak et al., 2008; Polymeris et al., 2009) and the fourth component (C4) is considered as the sum of all slower components (Kitis et al 2010). The C4 is only partially emptied in the LM-OSL measurements therefore, not well defined. Different slow components that is believed to makes up the C4 has been confirmed to be associated with TL trap o o slightly above 260 C, another around 400 C and even with very stable TL traps beyond o 600 C (Jain et al., 2003). However, in the case of annealed samples (Figures 4.21 and 4.22), five RT- LMOSL components are recognized. The C1, C2 and C3 stand as the three fast components, while the broad, intermediate and intense RT LM-OSL component C4 is o identified with 110 C TL peak (Kiyak et al., 2008; Polymeris et al., 2009). As earlier observed for C4 in the case of unannealed samples, the behavior of component C5 is 98 UNIVERSITY OF IBADAN LIBRARY Table 4.2: Percentage sensitisation signal of each component to the total fitted data Luminescence S2 S4 Component Unannealed Annealed Unannealed Annealed (% of (% of (% of (% of Experimental Experimental Experimental Experimental Data) Data) Data) Data) o 110 CTL Peak 99.01 90.11 99.00 92.83 C1 0.32 0.39 0.67 0.62 C2 1.22 1.12 2.27 1.57 C3 96.75 3.10 95.24 6.52 C4 1.64 49.62 1.67 45.81 C5 - 45.14 - 44.46 OSL Fitted 99.93 99.93 99.85 99.61 Data 99 UNIVERSITY OF IBADAN LIBRARY not of much interest in the present study (Apendix A1), mainly due to the following reasons: (i) the component has not totally decayed after 1000s of stimulation, (ii) this component is considered as the sum of all slower components. o 4.2.2 Dependence of 110 C TL peak and RT-LMOSL sensitisations on heating rate of thermal activation Having identified C3 in unannealed samples and C4 in annealed samples to o share the same electron trap with 110 C TL peak, the signal of these components was o used as the parallel complementary method to that of 110 C TL peak in this study. With reference to Figures 4.11 to 4.18, both unannealed and annealed S2 and S4 samples exhibited sensitisation after each successive cycle of irradiations and TL/OSL measurements. However, the nature of the glow and OSL curves obtained for the remaining heating rates were similar to this pattern but the only variation observed was from differences in the intensity of the signals. Figures 4.23 to 4.30 better present the dependence of the sensitisation for both TL and RT-LMOSL on heating rate of the preceding TL used for thermal activation for both S2 and S4. As could be seen in Figures 4.23 to 4.30 for the unannealed samples, the level of sensitisation is high for low heating rate and low for the high heating rate. Mostly for unannealed samples, the sensitisation is somewhat decreasing with heating rates generally. However, a careful observation of these figures showed that there are some points in which this heating rate sensitisation dependency trend was distrusted. On the other hand for annealed samples, it is only S2 and TL aspect of S4 that remarkable displayed the above trend of dependence of the sensitisation on heating rates of thermal activation. 4.2.3 Dependence of various components of RT-LMOSL sensitisations on heating rate of thermal activation The sensitisation of various components of RT-LMOSL is nearly identical to C4 described above. This is shown in Figures 4.31 to 4.34 for the two unannealed and annealed samples, which is sensitisation obtained during the measurement of Part C Step 6 (Sn4) of the protocol. The dependence of the sensitisation on the heating rates of thermal activation is more pronounced in unannealed samples than in the annealed for the two samples. As seen in these figures, the degree of the dependence on the 100 UNIVERSITY OF IBADAN LIBRARY 7 Sn4 10 Sn3 Sn2 Sn1 6 10 Sn0 5 10 1 10 o Heating Rate ( C/s) Fig. 4.23: Plots of TL sensitisations against heating rates as function of cycle of measurements for unannealed S2 samples. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively 101 UNIVE TL Integral (a.u)RSITY OF IBADAN LIBRARY Sn4 7 10 Sn3 Sn2 Sn1 6 10 Sn0 5 10 1 10 o Heating Rate ( C/s) 4.24: Plots of RT-LMOSL sensitisations against heating rates as function of cycle of measurements for unannealed S2 samples. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively 102 UNIVER OSL Integral (a.u)SITY OF IBADAN LIBRARY Sn3 Sn4 Sn1 Sn2 7 10 Sn0 6 10 1 10 o Heating Rate ( C/s) Fig. 4.25: Plots of TL sensitisations against heating rates as function of cycle of measurements for annealed S2 samples. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively 103 UNIVE TL Integral (a.u)RSITY OF IBADAN LIBRARY Sn4 Sn1 Sn2 Sn3 7 10 Sn0 6 10 1 o 10 Heating Rate ( C/s) Fig. 4.26: Plots of RT-LMOSL sensitisations against heating rates as function of cycle of measurements for annealed S2 sample. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively 104 UNIVER OSL Integral (a.u)SITY OF IBADAN LIBRARY Sn3 Sn4 7 10 Sn1 Sn2 6 10 Sn0 5 10 1 10 o Heating Rate ( C/s) Fig. 4.27: Plots of TL sensitisations against heating rates as function of cycle of measurements for unannealed S4 samples. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively 105 UNIVE TL Integral (a.u)RSITY OF IBADAN LIBRARY Sn4 Sn3 7 10 Sn1 Sn2 6 10 Sn0 1 o 10 Heating Rate ( C/s) Fig. 4.28: Plots of RT-LMOSL sensitisations against heating rates as function of cycle of measurements for unannealed S4 samples. Curves Sn0, Sn1, Sn2, Sn3, Sn4 st nd rd th represent unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively 106 UNIVER OSL Integral (a.u)SITY OF IBADAN LIBRARY Sn4 7 10 Sn1 Sn2 Sn3 Sn0 6 10 1 10 o Heating Rate ( C/s) Fig. 4.29: Plots of TL sensitisations against heating rates as function of cycle of measurements for annealed S4 samples. Curves Sn0, Sn1, Sn2, Sn3, Sn4 represent st nd rd th unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 107 UNIVE TL Integral (a.u)RSITY OF IBADAN LIBRARY 7 2.5x10 7 2x10 Sn4 7 1.5x10 Sn3 Sn2 7 10 Sn1 6 5x10 Sn0 1 o 10 Heating Rate ( C/s) Fig. 4.30: Plots of RT-LMOSL sensitisations against heating rates as function of cycle of measurements for annealed S4 samples. Curves Sn0, Sn1, Sn2, Sn3, Sn4 st nd rd th represent unsensitised, 1 sensitised, 2 sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively. 108 UNIVE OSL Integral (a.u)RSITY OF IBADAN LIBRARY 8 10 7 10 6 10 5 10 4 10 3 10 C1 C2 2 10 C3 C4 1 10 0 10 1 10 o Heating Rate ( C/s) th Fig. 4.31. Plots of 4 sensitised of RT-LMOSL components sensitisations against heating rates as function of cycle of measurements for unannealed S2 samples. 109 UNIVE OSL Integral (a.u)RSITY OF IBADAN LIBRARY C1 8 C2 10 C3 C4 7 10 6 10 5 10 1 10 o Heating Rate ( C/s) th Fig. 4.32. Plots of 4 sensitised of RT-LMOSL components sensitisations against heating rates as function of cycle of measurements for unannealed S4 samples. 110 UNIVE OSL Integral (a.u)RSITY OF IBADAN LIBRARY C1 C2 C3 8 10 C4 C5 7 10 6 10 5 10 0 1 10 10 o Heating Rate ( C/s) th Fig. 4.33. Plots of 4 sensitised of RT-LMOSL components sensitisations against heating rates as function of cycle of measurements for annealed S2 samples. 111 UNIVE OSL Integral (a.u)RSITY OF IBADAN LIBRARY C1 8 10 C2 C3 C4 C5 7 10 6 10 5 10 1 10 o Heating Rate ( C/s) th Fig. 4.34. Plots of 4 sensitised of RT-LMOSL components sensitisations against heating rates as function of cycle of measurements for annealed S4 samples. 112 UNIVE OSL Integral (u.a)RSITY OF IBADAN LIBRARY heating rates of thermal activation in the unannealed samples is more prominent in the fast components, then followed by C3 and less pronounced in C4. While for annealed samples, similar dependence of the sensitisation of all the components on heating rates of thermal activation was observed. However, C4 and C5 of S4 annealed sample did not show note worthy dependence on heating rates of thermal activation. o 4.2.4 Dependence of 110 C TL peak and RT LM-OSL sensitisations on TL activation histories Identification of the exclusive contributions of pre-exposure of quartz sample to a dose of irradiation and the purely thermal activation by way of experiment was one of the major objectives that motivated this work. There are some underlying principles behind each step of the protocol in attempts to separate the contributions of each of the two factors stated above. Although, the same dose of irradiation was maintained in the measurements irrespective of the protocol parts, however, pre-exposure dose (accumulated dose) was varied with the successive cycle of measurements and irradiations. The comparative sensitisations of the various parts of the protocol with respect to heating rate for both unannealed and annealed S2 and S4 samples are presented in Figures 4.35 to. 4.42. In particular, by looking at TL section of the protocol, the key difference between parts A and B of the protocol is the presence of intermediate irradiations that are sandwiched in between the successive cycles of thermal activation in part A that were absent in Part B. Also, the exact differentiation exits respectively between parts C and D that were meant for OSL of the protocol. Contrarily to what is exhibited in the unannealed samples, a relative higher enhancement in sensitisations of parts A and C over parts B and D respectively was expected based on pre-dose model. This is because the accumulated pre-exposure dose of parts A and C were more than that of parts B and D correspondingly. The two annealed samples excellently displayed this as could be seen in Figures 4.35 to. 4.42. Contrarily, curve Sn4 do not exhibit any clear sensitisation enhancement over curve Sn1b for both TL and RT LMOSL in unannealed samples. The contribution of thermal sensitisation is also reflected in curve Sn1b with respect to curve Sn1 for both TL and RT- LMOSL. The former curve is quantitatively higher than the later. 113 UNIVERSITY OF IBADAN LIBRARY 8 10 Sn1b 7 Sn4 10 Sn1 6 10 1 10 o Heating Rate ( C/s) Fig. 4.35: Comparison of TL sensitisations in aliquots with pre-exposure dose and those with thermal activation against heating rates for unannealed S2 st th sample. Curves Sn1 and Sn4 are respectively 1 and 4 sensitisation of successive cycles with pre-exposure dose. Sn1b is the last sensitisation of successive cycles without pre-exposure dose. 114 UNIVE TL Integral (a.u)RSITY OF IBADAN LIBRARY 7 8x10 7 7x10 7 6x10 Sn4 7 5x10 7 4x10 Sn1 7 3x10 Sn1b 7 2x10 1 10 o Heating Rate ( C/s) Fig.4.36. Comparison of TL sensitisations in aliquots with pre-exposure dose and those with thermal activation against heating rates for annealed S2 sample. st th Curves Sn1 and Sn4 are respectively 1 and 4 sensitisation of successive cycles with pre-exposure dose. Sn1b is the last sensitisation of successive cycles without pre-exposure dose. 115 UNIV TL Integral (a.u)ERSITY OF IBADAN LIBRARY 7 10 Sn4 Sn1b 6 10 Sn1 5 10 1 10 o Heating Rate ( C/s) Fig.4.37. Comparison of RT-LMOSL sensitisations in aliquots with pre- exposure dose and those with thermal activation against heating rates for st th unannealed S2 sample. Curves Sn1 and Sn4 are respectively 1 and 4 sensitisation of successive cycles with pre-exposure dose. Sn1b is the last sensitisation of successive cycles without pre-exposure dose. 116 UNIVER OSL Integral (a.u)SITY OF IBADAN LIBRARY 7 7x10 7 6x10 Sn4 7 5x10 7 4x10 7 3x10 Sn1 7 2x10 Sn1b 1 o 10 Heating Rate ( C/s) Fig.4.38. Comparison of RT-LMOSL sensitisations in aliquots with pre- exposure dose and those with thermal activation against heating rates for st th annealed S2 sample. Curves Sn1 and Sn4 are respectively 1 and 4 sensitisation of successive cycles with pre-exposure dose. Sn1b is the last sensitisation of successive cycles without pre-exposure dose. 117 UNIV OSL Integral (a.u)ERSITY OF IBADAN LIBRARY Sn1b Sn4 7 10 Sn1 1 10 o Heating Rate ( C/s) Fig.4.39. Comparison of TL sensitisations in aliquots with pre-exposure dose and those with thermal activation against heating rates for unannealed S4 st th sample. Curves Sn1 and Sn4 are respectively 1 and 4 sensitisation of successive cycles with pre-exposure dose. Sn1b is the last sensitisation of successive cycles without pre-exposure dose 118 UNIVE TL Integral (a.u)RSITY OF IBADAN LIBRARY 7 2.4x10 Sn4 7 1.6x10 Sn1 Sn1b 6 8x10 1 10 o Heating Rate ( C/s) Fig.4.40. Comparison of TL sensitisations in aliquots with pre-exposure dose and those with thermal activation against heating rates for annealed S4 sample. st th Curves Sn1 and Sn4 are respectively 1 and 4 sensitisation of successive cycles with pre-exposure dose. Sn1b is the last sensitisation of successive cycles without pre-exposure dose. 119 UNIV TL Integral (a.u)ERSITY OF IBADAN LIBRARY Sn1b 7 10 Sn4 Sn1 6 10 1 o 10 Heating Rate ( C/s) Fig. 4.41. Comparison of RT-LMOSL sensitisations in aliquots with pre- exposure dose and those with thermal activation against heating rates for st th unannealed S4 sample. Curves Sn1 and Sn4 are respectively 1 and 4 sensitisation of successive cycles with pre-exposure dose. Sn1b is the last sensitisation of successive cycles without pre-exposure dose. 120 UNIVE OSL Integral (a.u)RSITY OF IBADAN LIBRARY 7 2x10 7 1.8x10 Sn4 7 1.6x10 7 1.4x10 7 1.2x10 7 10 Sn1 6 8x10 6 Sn1b 6x10 1 10 o Heating Rate ( C/s) Fig. 4.42. Comparison of RT-LMOSL sensitisations in aliquots with pre- exposure dose and those with thermal activation against heating rates for st th annealed S4 sample. Curves Sn1 and Sn4 are respectively 1 and 4 sensitisation of successive cycles with pre-exposure dose. Sn1b is the last sensitisation of successive cycles without pre-exposure dose. 121 UNIV OSL Integral (a.u)ERSITY OF IBADAN LIBRARY In order to identify the comparative sensitisations resulting from accumulative successive TL readings and irradiations that each of the aliquots has been subjected to, Part A steps 6 and 7 and Part B step 7 and 8 of the protocol were investigated on each of the aliquots. It should be recalled that steps 7 and 8 respectively in parts A and B of o the protocols were read at heating rate of 2 C/s for all the aliquots. Curves Sn5 and Sn2b in Figures 4.43 to 4.44 resulted from these measurements for parts A and B of the protocol respectively. Virtually identical trend were exhibited by curves Sn4 and Sn1b as compared with Sn5 and Sn2b correspondingly. 4.2.5 Discussion The possible reasons behind the unexpected inconsistencies that were observed in the sensitisation dependence of both TL and RT-LMOSL on heating rates of thermal activation are presented below. The sensitisation reproducibility study of the two samples that were presented in section 4.1, were in line with the recently report that quartz aliquots from the same crystal failed to possess sensitisation reproducibility Appendix 1). The observations of the referred were based on 10 different quartz samples that happened to include the present two understudied quartz samples. It was further observed also that this strong variation is removed by high temperature o annealing as well as heating up to 500 C, involved in the TL measurements. Therefore, since multiple aliquots were used in the present work, the individual sensitisation of each aliquot is believed to be the major influence behind the alteration in the sensitisation pattern of the TL with respect to heating rates that was observed. The structure of curve Sn0 of TL for both unannealed and annealed two samples exhibits a satisfactory trend of sensitisation on heating rate as depicted in Figures 4.23, 4.25, 4.27 and 4.29. With reference to the protocols, it should be born in mind that there was no thermal activation, at all, before TL of Step 4 (part A protocol) that produced o curve Sn0. The TL reading of Step 2 that preceded curve Sn0 was only to 180 C which is below thermal activation temperature. Thus, the absence of contribution of thermal activation, which is prone to sensitisation non-reproducibility, in curve Sn0 is the reason why there is consistency with heating rates in this particular curve for TL part. This adequately revealed the inconsistency that sensitisation non-reproducibility of the multiple aliquots introduced into the sensitisation of curves Sn1-Sn4 that are read after curve Sn0 which served as thermal activation. 122 UNIVERSITY OF IBADAN LIBRARY Sn5 7 10 Sn4 6 10 1 10 o Heating Rate ( C/s) o Fig. 4.43: Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots with pre-exposure dose for unannealed S2 th th sample. Curves Sn4 and Sn5 are respectively 4 and 5 sensitisations of o successive cycles. It should be recalled that Sn5 was read at HR of 2 C/s 123 UNIVE TL Integral (a.u)RSITY OF IBADAN LIBRARY 8 10 7 Sn1b 10 Sn2b 6 10 1 10 o Heating Rate ( C/s) o Fig. 4.44: Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots without pre-exposure dose for unannealed S2 th th sample. Curves Sn1b and Sn2b are respectively the 4 and 5 sensitisation of o successive cycles. It should be recalled that Sn2b was read at HR of 2 C/s 124 UNIVE TL Integral (a.u)RSITY OF IBADAN LIBRARY 7 7x10 7 6x10 Sn5 7 5x10 Sn4 7 4x10 1 10 o Heating Rate ( C/s) o Fig. 4.45: Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots with pre-exposure dose for annealed S2 th th sample. Curves Sn4 and Sn5 are respectively 4 and 5 sensitisations of o successive cycles. It should be recalled that Sn5 was read at HR of 2 C/s 125 UNIV TL Integral (a.u)ERSITY OF IBADAN LIBRARY 7 3.5x10 Sn2b 7 2.8x10 Sn1b 7 2.1x10 1 10 o Heating Rate ( C/s) o Fig. 4.46: Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots without pre-exposure dose for annealed S2 th th sample. Curves Sn1b and Sn2b are respectively the 4 and 5 sensitisation of o successive cycles. It should be recalled that Sn2b was read at HR of 2 C/s 126 UNIV TL Integral (a.u)ERSITY OF IBADAN LIBRARY 7 4x10 7 3x10 Sn4 7 2x10 Sn5 7 10 1 10 o Heating Rate ( C/s) o Fig. 4.47. Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots with pre-exposure dose for unannealed S4 th th sample. Curves Sn4 and Sn5 are respectively 4 and 5 sensitisations of o successive cycles. It should be recalled that Sn5 was read at HR of 2 C/s 127 UNIV TL Integral (a.u)ERSITY OF IBADAN LIBRARY Sn2b Sn1b 7 10 1 10 o Heating Rate ( C/s) o Fig. 4.48: Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots without pre-exposure dose for unannealed S4 th th sample. Curves Sn1b and Sn2b are respectively the 4 and 5 sensitisation of o successive cycles. It should be recalled that Sn2b was read at HR of 2 C/s 128 UNIVE TL Integral (a.u)RSITY OF IBADAN LIBRARY 7 2.4x10 Sn5 7 1.8x10 Sn4 7 1.2x10 1 10 o Heating Rate ( C/s) o Fig. 4.49: Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots with pre-exposure dose for annealed S4 th th sample. Curves Sn4 and Sn5 are respectively 4 and 5 sensitisations of o successive cycles. It should be recalled that Sn5 was read at HR of 2 C/s 129 UNIV TL Integral (a.u)ERSITY OF IBADAN LIBRARY 7 1.2x10 Sn2b 6 9x10 Sn1b 1 10 o Heating Rate ( C/s) o Fig. 4.50: Comparison of 110 C TL peak sensitisations resulting from measurement histories in aliquots without pre-exposure dose for annealed S4 th th sample. Curves Sn1b and Sn2b are respectively the 4 and 5 sensitisation of o successive cycles. It should be recalled that Sn2b was read at HR of 2 C/s 130 UNIV TL Integral (a.u)ERSITY OF IBADAN LIBRARY In addition, the sensitisation curves of Figures 4.25 and 4.26 for annealed sample satisfactorily demonstrate the dependence of sensitisation on heating rate that is envisaged. Hence, the curves of figure for annealed S2 demonstrate the true picture of the dependency of sensitisation on heating rate since the level of sensitisation non- reproducibility in annealed samples have been observed to be generally insignificant (Appendix 1). It was consequently anticipated that normalization of each curve of Figures 4.23 to 4.30 for S2 and S4 unannealed samples to the first sensitised curve, Sn1, rather than Sn0 should present the true dependence of sensitisation on heating rate. The outcome of this normalization is presented in Figures 4.51 and 4.52 that are the true sensitisation dependence on heating rate in which the sensitisation inconsistency has satisfactorily disappeared for the two unannealed samples. The only point where this trend is not o observed is after 5 C/s in S4 samples which seems to possess relatively less sensitisation feature as compared with S2. As it was afore mentioned under observations, it is apparent on the average that there was no special sensitisation contribution made by accumulated pre-exposure dose in parts A and C (curve Sn4) over parts B and D (curve Sn1b) of both TL and RT- LMOSL in unannealed samples (see Figures 4.35, 4.37, 4.39 and 4.41). It should be recalled that Sn4 and Sn1 respectively received accumulated pre-exposure dose that are by factors of 3 and 1 TD higher than the one received by Sn1b for both TL and RT LMOSL. In the same way, Sn4 and Sn1b received the similar successive cycles of TA that is factor of 3 higher than the one received by Sn1. According to the pre-dose model, sensitisation of Sn4 was expected to be higher than that of Sn1b. Although, sensitisation non-reproducibility that is more pronounced in unannealed sample has certainly contributed to the obvious demarcation of curve Sn4 sensitisation with respect to curve Sn1b that is missing in unannealed samples for both TL and RT-LMOSL. Nevertheless, it can be seen from Figures 4.35, 4.37, 4.39 and 4.41 that 131 UNIVERSITY OF IBADAN LIBRARY 7 6 c 5 b 4 3 a 2 1 10 o Heating Rate ( C/s) o Fig. 4.51. Plots of normalized 110 C TL peak sensitisations against heating rates as function of cycle of measurements for unannealed S2. (a) Sn2/Sn1, (b) st nd Sn3/Sn1, (c) Sn4/Sn1 with Sn1, Sn2, Sn3, Sn4 representing 1 sensitised, 2 rd th sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively 132 UNIVER Normalized TL IntegralSITY OF IBADAN LIBRARY 3.0 2.5 2.0 c b 1.5 a 1 10 o Heating Rate ( C/s) o Fig. 4.52. Plots of normalized 110 C TL peak sensitisations against heating rates as function of cycle of measurements for unannealed S4. (a) Sn2/Sn1, (b) st nd Sn3/Sn1, (c) Sn4/Sn1 with Sn1, Sn2, Sn3, Sn4 representing 1 sensitised, 2 rd th sensitised, 3 sensitised, 4 sensitised, luminescence readings respectively 133 UNIVE Normalized TL IntegralRSITY OF IBADAN LIBRARY curves Sn4 and Sn1b overlap on the average. Even better still, Sn1b is always higher for all o points of the heating rate 0.25 C/s for the two samples. As presented in Table 4.3, the sensitisation signals at the highest thermal activation of the aliquots of unannealed samples without pre-exposure dose was higher than that of the aliquots with pre- exposure dose by factor of 76.0 % and 79.0 % for TL and RT-LMOSL respectively for S2 while the corresponding factors obtained for S4 were 45.0 % for TL and 14.0 % for RT-LMOSL. In annealed samples, the sensitisation signal of the aliquots with pre- exposure dose was rather higher than that of the aliquots without pre-exposure dose, by factor of 224.0 % for TL and 201.0 % for RT-LMOSL for S2 and for S4, it was by factor of 245.0 % for TL and 217.0 % for RT-LMOSL. The exact sensitisation trend as observed above are displayed by the fast components of S2 and S4 unannealed and annealed samples (Figures 4.53 to 4.56) Yet, curve Sn1b is quantitatively higher than Sn1 in Figures 4.35, 4.37, 4.39 and 4.41 for unannealed samples. Based on pre-exposure dose received, these observations were contrary to pre-dose model. Thus, the identical sensitisation that both Sn4 and Sn1b share can only be explained on the ground of the similar level of thermal activation the two received as mentioned earlier. Equally, the degree of thermal activation of Sn1b that is above that of Sn1 clarifies why higher sensitisation is in favour of Sn1b. The sensitisations of S2 and S4 annealed samples Figures 4.36, 4.38, 4.40 and 4.42 truly and satisfactorily obey pre-dose model in its case contrarily to the case of unannealed samples. Therefore, the inference that could be drawn from these is that contribution of thermal sensitisation and pre-exposure dose are opposite in unannealed and annealed samples. This means that thermal sensitisation is the key contribution of sensitisation in unannealed sample while pre-exposure dose played the chief role in annealed samples sensitisation. A support to this conclusion is the dependence of the observed sensitisation on heating rates of thermal activation that is more pronounced in unannealed samples as compared with S2 and S4 corresponding annealed samples (Figures 4.23 to 4.30). This is because the dependence of sensitisation on heating rates of thermal activation has been associated with the pure thermal sensitisation (Koul et al., 2010). Although, a o common activation temperature of 500 C was used all through, it was the various heating rates that accounted for the different activation heating times. The TL 134 UNIVERSITY OF IBADAN LIBRARY Table 4.3: Relative sensitisation factor of aliquots with pre-exposure dose and those without pre-exposure dose Samples TL OSL Sensitisation Sensitisation Sensitisation Sensitisation factor without factor with PED factor without factor with PED PED for for annealed PED for for annealed unannealed sample (%) unannealed sample (%) sample (%) sample (%) S2 76.0 224.0 79.0 201.0 S4 45.0 245.0 14.0 217.0 Key: PED = pre-exposure dose 135 UNIVERSITY OF IBADAN LIBRARY 6 10 C1Sn1d C1 Sn4 C2Sn1d C2Sn4 5 10 4 10 1 10 o Heating Rate ( C/s) Fig.4.53: Comparison of RT-LMOSL sensitisations in aliquots with pre- exposure dose and those with thermal activation against heating rates for th unannealed S2 sample. Curves C1Sn4 and C2Sn4 are respectively 4 sensitisation of successive cycles with pre-exposure dose for components C1 and C2 while C1Sn1b and C2Sn1b are respectively the last sensitisation of successive cycles without pre-exposure dose for components C1 and C2. 136 UNIVE OSL Integral (a.u)RSITY OF IBADAN LIBRARY C1Sn1d C1Sn4 6 10 C2Sn1d C2Sn4 5 10 1 10 o Heating Rate ( C/s) Fig. 4.54: Comparison of RT-LMOSL sensitisations in aliquots with pre- exposure dose and those with thermal activation against heating rates for th unannealed S4 sample. Curves C1Sn4 and C2Sn4 are respectively 4 sensitisation of successive cycles with pre-exposure dose for components C1 and C2 while C1Sn1b and C2Sn1b are respectively the last sensitisation of successive cycles without pre-exposure dose for components C1 and C2. 137 UNIVE OSL Integral (a.u)RSITY OF IBADAN LIBRARY C1Sn1d 7 10 C1Sn4 C2Sn1d C2Sn4 C3Sn1d C3Sn4 6 10 5 10 1 10 Number of cycle Fig. 4.55: Comparison of RT-LMOSL sensitisations in aliquots with pre- exposure dose and those with thermal activation against heating rates for th annealed S2 sample. Curves C1Sn4, C2Sn4 and C3Sn4 are respectively 4 sensitisation of successive cycles with pre-exposure dose for components C1, C2 and C3 while C1Sn1b, C2Sn1b and C3Sn1b are respectively the last sensitisation of successive cycles without pre-exposure dose for components C1, C2 and C3. 138 UNIVE OSL Integral (a.u)RSITY OF IBADAN LIBRARY C1Sn1d C1Sn4 7 10 C2Sn1d C2Sn4 C3Sn1d C3Sn4 6 10 5 10 1 10 Number of cycle Fig. 4. 56: Comparison of RT-LMOSL sensitisations in aliquots with pre- exposure dose and those with thermal activation against heating rates for th annealed S4 sample. Curves C1Sn4, C2Sn4 and C3Sn4 are respectively 4 sensitisation of successive cycles with pre-exposure dose for components C1, C2 and C3 while C1Sn1b, C2Sn1b and C3Sn1b are respectively the last sensitisation of successive cycles without pre-exposure dose for components C1, C2 and C3. 139 UNIVE OSL Integral (a.u)RSITY OF IBADAN LIBRARY o readings at 0.25 and 10 C/s heating rates will result into 2000 and 50s heating time of o the aliquots respectively. As indicated by this, an aliquot that was read out at 0.25 C/s o would have received heating of 40 times more than the one that was read at 10 C/s heating rate. In real sense, sensitisation of S4 does not practically show any dependence on heating rates of thermal activation while slight dependence on heating rate can be seen in S2 in these figures for annealed samples. Another argument, which supports this inference is an abrupt change on the sensitivity of all samples between the first (Sn0) and the second (Sn1) TL and RT- LMOSL measurement that is a clear product of pre-dose sensitisation effect (Oniya et al., 2012b). In this same vain, this abruption of sensitisation was more conspicuous in S2 and S4 annealed samples than their unannealed counterpart (Figures 4.11 to 4.18). This was unnoticeable at all in S2 unannealed sample. These observations hereby identify and present distinctive factor that is mostly responsible for sensitisation of each of unannealed and annealed samples of quartz. The enormous increase in the percentage emission from 1.64 and 1.65% of C4 respectively in S2 and S4 unannealed samples to 45.14 and 44.46% of C5 of their annealed counterparts is informative (Table 4.2). Jain et al., 2003 reported a similar observation for a Korean sample where its OSL is dominated by slower components after annealing. More light is shed on the observations in question by matching each RT-LMOSL component with their corresponding glow peak in curve c Figures 4. 57 to 4. 60. These figures depict the “OSL glow curve” that is the difference between TL glow curve obtained before OSL measurement (Un-bleached TL) and the residual TL (Bleached TL) after OSL stimulation. As a result, Curve c in Figures 4. 57 and 4. 60 corresponds to the total signal of the whole RT-LMOSL emission recorded during the bleaching. Percentage signal of C3 and C4 respectively in unannealed and annealed o samples, adequately and quantitatively commensurates with the signal of 110 C TL glow peak in curve c. The same relationship is maintained between the fast components o and 325 C TL peak of curve c. But unlike what is observed in unannealed case, the percentage signals, 49.62 and 45.81%, of C4 in respective annealed S2 and S4 samples o do not commensurate with those of 110 C TL peaks (Table 4.2). Therefore, as the o signal of 110 C TL peak almost completely dominates the entire signal of the TL response, it can be inferred that the electron trap(s) that was 140 UNIVERSITY OF IBADAN LIBRARY 6 10 5 10 4 10 3 a 10 b 2 10 c 1 10 0 100 200 300 400 500 o Temperature ( C) Fig. 4.57. For unannealed S2 represents glow curves (a) C4 (Un-bleached TL) obtained from step 5 of Part A, (b) bleached TL from step 5 of Part C and (c) the “OSL glow curve” which is the different between (a) and (b) . 141 UNIVE TL (a.u)RSITY OF IBADAN LIBRARY 6 10 5 10 a 4 10 3 10 b 2 10 c 1 10 0 100 200 300 400 500 o Temperature ( C) Fig. 4.58. For annealed S2 represents glow curves (a) C4 (Un-bleached TL) obtained from step 5 of Part A, (b) bleached TL from step 5 of Part C and (c) the “OSL glow curve” which is the different between (a) and (b) . 142 UNIVE TL (a.u)RSITY OF IBADAN LIBRARY 6 10 5 10 4 10 a 3 10 2 10 b 1 10 c 0 10 0 100 200 300 400 500 o Temperature ( C) Fig. 4.59. For unannealed S4 represents glow curves (a) C4 (Un-bleached TL) obtained from step 5 of Part A, (b) bleached TL from step 5 of Part C and (c) the “OSL glow curve” which is the different between (a) and (b) . 143 UNIVE TL (a.u)RSITY OF IBADAN LIBRARY 5 10 4 10 a 3 10 a 2 10 c 1 10 0 100 200 300 400 500 o Temperature ( C) Fig. 4.60. For annealed S4 represents glow curves (a) C4 (Un-bleached TL) obtained from step 5 of Part A, (b) bleached TL from step 5 of Part C and (c) the “OSL glow curve” which is the different between (a) and (b) . 144 UNIVE TL (a.u)RSITY OF IBADAN LIBRARY responsible for some of the components that make up the C5 in annealed sample is not depleted during the TL measurement. Hence this further confirmed that some of the o components of C5 are associated with electron traps around 600 C as earlier claimed. It could be proposed here that the trap responsible for the slower component in question is activated or created during annealing process. This therefore explains why its response is unnoticed in unannealed samples. It was important to identify the effect of thermal quenching on the observations. By considering curve Sn0 in Figures 4.23, 4.25, 4.27 and 4.29 in particular, it shows that these curves were purely determined by thermal quenching effect since there was no sensitisation up to this level. If that is untrue, the structure of curve Sn0 for RT-LMOSL in Figures 4.24, 4.26, 4.28 and 4.30 also would have shown dependence on heating rates. That curve Sn0 for RT-LMOSL in Figures 4.24, 4.26, 4.28 and 4.30 do not show significant dependence on heating rates of thermal activation is not far from expectation since all the RT-LMOSL were not read in varied temperature, which would have guaranteed obviously seen thermal quenching effect. Thus, the curve Sn0 structure of TL is a product of thermal quenching. Also, a close look at curves Sn1-Sn4 in the same Figures 4.23 to 4.30 shows these curves exhibit nearly identical curve nature (exclusive intensity) with Sn0 that suffers only from thermal quenching. Hence, the cause behind the nature of curves Sn1-Sn4, which has been claimed to be due to sensitisation as a function of heating rate all along in this work, is prone to be mistaken for thermal quenching. It is therefore obvious that, the comparative sensitisations resulting from accumulative successive TL readings and irradiations that each of the aliquots were subjected to, clarify the possible mix-up. The pattern of curves Sn5 and Sn2b, which were o read using the same heating rate (2 C/s) on all the aliquots earlier used for curves Sn0- Sn4 and Sn1b respectively (Figures 4.43 to 4.50) confirmed this. For if the trend of curves Sn1-Sn4 in these figures are truly caused by thermal quenching effect, curve Sn5 is not supposed to follow any pattern with heating rates. Therefore curve Sn5 that shows increase of sensitisation with decreasing heating rates solely reflects the dependence of sensitisation on TL reading histories of each aliquot. The same observation was observed for part B that involved no intermediate successive irradiations (see curves Sn2b and Sn1b in the same Figures 4.43 to 4.50). 145 UNIVERSITY OF IBADAN LIBRARY Another support for this is seen from RT-LMOSL sensitisation that showed dependence on heating rates exactly like TL counterpart. This is because, all the RT- LMOSL were read in room temperatute and could not have suffered from thermal quenching. In conclusion, the practically indistinguishable trend that is exhibited by curves Sn5 and Sn1b as compared with Sn4 and Sn2b respectively, indicates curves Sn1-Sn4 and Sn1b to be the actual sensitisation pattern resulting from the successive cycles of TL luminescence measurements and irradiations. However, this claim does not completely rule out the contribution of thermal quenching in the curves but rather asserting that it is minimal. The little contribution of thermal quenching effect is only obvious in S4 unannealed samples as seen in Figures 4.47 and 4.48. Given that curve Sn5 is the next successive TL reading immediately after curve Sn4, curve Sn5 would be expected to always be higher than Sn4 for all the heating rates. Contrarily, curve Sn4 happens to be higher than Sn5 for 0.25 and 0.5C/s heating o rates points, while the two curves are almost the same for 1 C/s heating rate, curve Sn5 takes the lead over Sn4 for the remaining heating rates points. A similar relation is observed in the same referred figure for part B of the protocol where there were no intermediate irradiations. The above behaviour is the reflection of thermal quenching o effect. This is because TL of curve Sn5 was read using 2 C/s heating rate for all the o aliquots and 2 C/s will experience relatively more thermal quenching than 0.25, 0.5 and o 1 C/s heating rates. That accounts for the reason why curve Sn4 takes the lead over Sn5 for the three low heating rates. In addition, the sensitisation of RT LM-OSL of annealed S4 does not show any dependence on heating rates of thermal activation (Figure 4.30). Therefore, the decrease in sensitisation with heating rate that its TL counterpart o displayed after heating rates of 2 C/s (Figure 4.29) is definitely caused by thermal quenching. Another observation that merits further consideration is the enormous sensitisation that curve Sn2b exhibited over Sn1b as seen in annealed S2 and S4 samples (Figures 4.46 and 4.50). This was neither seen in unannealed samples (Figures 4.45 and 4.49) even in part A protocol (Figures 4.43, 4.44, 4.47, and 4.48) of the two annealed samples. The original Zimmerman, 1971 model was able to adequately proffer a possible explanation for this. According to Zimmerman, 1971 model, the successive TL readings of the protocol will cause depletion of the R-centers. The irradiations that were sandwiched in between successive cycles of thermal activation of part A of the protocol will replenish the depleted R-centre after each successive TL readings. Conversely, R- 146 UNIVERSITY OF IBADAN LIBRARY center will be enormously depleted at the end of successive TL readings for part B protocol of which the successive TL readings were not sandwiched with irradiations. This will consequently result in high competition for holes between L-centre and R- centre during the following administration of TD (Step 5 Part B protocol) prior to the o next TL readout to 500 C of Step 6 that produced Sn1b. Most of the holes produced by this TD are therefore captured by R-centre owing to its heavy depletion and the preferential hole trapping nature (competition) that it possesses over L-centre (according to pre-dose model). The consequential effect of this will lead to a relatively intense reduction in TL signal of Sn1b as compared with the following pre-dose sensitisation of Sn2b that resulted from combined effect of TD of step 5 and TA of TL of step 6 of part B of the protocol. This comparatively reduced signal of Sn1b is responsible for the questioned vast sensitisation that curve Sn2b exhibited over Sn1b of annealed S2 and S4 samples in Figures 4.46 and 4.50 respectively. The absence of this vast depletion accounts for why part A behaved differently from this. The fact that the above observation, which is based on pre dose model, only pertains to annealed samples is another support to the initial inference that two different mechanisms are responsible for sensitisation in unannealed and annealed samples. The overall sensitisations as observed in this study could be explained under the framework of the existing models as follows. Since the unannealed S2 and S4 samples have not received thermal activations prior to the measurements, the process of cycles o of TL readings to 500 C resulted into the pure thermal sensitisation for all the o unannealed samples. Moreover, the annealing of the unannealed samples to 900 C for 1 hour is believed to cause the alteration made to the recombination centre and very deep electron traps to attain the peak. This accounts for the relative increase of sensitivity after annealing by factors of 116 and 27 that is demonstrated by S2 and S4 respectively in this study. This therefore makes the contributions of pure thermal sensitisation to become less significant after the annealing (as the case of the annealed S2 and S4 samples) and consequently offers the pre-dose sensitisation the prominent role it played in annealed samples. To buttress this, the relative lesser thermal sensitisation during o successive cycles of TL readings to 500 C that was observed in annealed samples could be recalled. The dependence of sensitisation on heating rates observed in this work, which on its own was an evidence of purely thermal sensitisation, also supports the above inferences. This is because it is outstanding in unannealed samples comparably to its annealed samples counterpart. 147 UNIVERSITY OF IBADAN LIBRARY 4.2.6 Implications of results Thermal sensitisation was observed to be the major mode of sensitisation in unannealed sample while pre-exposure dose played the chief role in annealed samples sensitisation. Therefore, pre-dose technique is not appropriate for unheated sample but rather ideal for fired quartz grains inclusion materials like ceramics, fired sediments and bricks, metallurgy ovens, and other highly fired objects. Furthermore, the protocol used in this study can be employed for authenticity testing to determine which samples have been previously heated. Another insinuation on dating that could be inferred from this work is the observed dependence of the luminescence sensitisation on heating rate used for thermal activation, irradiation dose and TL readout histories. This is important o since integrated light output of the 110 C TL peak is sometime employed for mass normalization in TL/OSL dating procedures and luminescence study. Thus, such mass normalization will be erroneous if the concerned aliquots have been subjected to various thermal activations through different heating rates, TL readout histories and irradiation dose. It is therefore suggested that this type of mass normalization be done at the beginning of measurement. However, if that is not possible, these factors as revealed in this work should be taken into consideration. Lastly, due to sensitisation resulting from TL readout histories, the practice of taking background TL measurement after each measurement should be taken into consideration in respect of changes in sensitivity of the concerned aliquot that such treatments could introduce. The parallel pattern of the pre-dose sensitisation that the TL and RT-LMOSL exhibited in this work lend a solid support to the earlier suggestion of Polymeris et al., 2009 to extrapolate the TL pre-dose methodology to the OSL pre-dose effect using only the appropriate component of RT-LMOSL, instead of the initial part of the CW-OSL signal. Therefore, it is here by proposed on the basis of the identical dose response (Polymeris et al., 2009) and the similar pre-dose sensitisation reported in this work for the two methods can serve as complementary dating methods with the motive to use it to authenticate the result obtain with either of them or as an alternative method. It should be noted also that the RT-LMOSL guarantee the avoidance of heating during the luminescence measurement. This in return rules out the possibility of chemiluminescence which makes OSL measurements not necessarily mandatory in nitrogen atmosphere and the total absence of thermal quenching effect. 148 UNIVERSITY OF IBADAN LIBRARY CHAPTER FIVE CONCLUSIONS AND RECOMMENDATIONS 5.1 Conclusion Study of Thermoluminescence (TL) and Optically Stimulated Luminescence (OSL) sensitisations of unannealed and annealed quartz samples using two quartz samples collected from southwestern Nigeria have been carried out. From the results obtained, the following conclusions were drawn: 1. Reproducibility study of the pre-dose sensitisation effect within the same quartz crystal revealed that there is a very good reproducibility between the initial sensitivities in both cases of fired and unannealed samples based on the standard deviation from 10 aliquots of the o normalised sensitivity. Heating up to 180 C does not influence the sensitivity of the samples. The reproducibility of the pre-dose sensitisation, observed within various aliquots of the same sample, was observed to be much better in annealed than unannealed samples. 2. oThe contributions of pre-exposure dose and thermal activation on 110 C TL peak and TR-LMOSL sensitivities of two quartz samples showed o sensitisation pattern exhibited by 110 C TL peak and RT-LMOSL to be nearly identical. This therefore, supports the model of these two mechanisms to share a common electron trap and recombination centres. The luminescence sensitisation was observed to somewhat decrease with heating rates generally. Furthermore, thermal sensitisation was observed to be the key contribution of sensitisation in unannealed sample while pre-exposure dose played the chief role in sensitisation of annealed samples. Thus, this clarifies the usual mixed up of the two mechanisms. 3. Luminescence sensitisation was observed to decrease with heating rate used for thermal activation mostly in unannealed samples but increase with quantity of pre-exposure dose in annealed samples. However, the sensitisation increased with the number of TL readout for both annealed and unannealed samples. 149 UNIVERSITY OF IBADAN LIBRARY 4. Both the pre-dose sensitisation of TL and RT-LMOSL nearly followed o identical pattern. 110 C TL peak were observed to share the same electron traps with component 3 (C3) and component 4 (C4) in unannealed and annealed samples respectively. 5.2 Recommendations Following the observations in this study and for future luminescence dating and studies, it is recommended that. 1. Pre-dose method of dating is only appropriate for heated quartz samples and not feasible for unannealed samples. 2. Special caution should be exercised while applying specific mass reproducibility correction procedure in the case of geological, unannealed quartz samples, since it could result in erroneous equivalent dose estimation. Generally, the systematic understanding of effects of each step of o luminescence protocol on mechanisms describing 110 C TL peak characteristics must be taken into consideration in order to chose appropriate point for sensitivity test and correction for unannealed and annealed samples. 3. Due to sensitisation resulting from TL readout histories, changes in sensitivity of the concerned aliquot should be taken into consideration in the practice of TL background measurements after each TL readings in order to avert introduction of errors in results. 4. The protocol used in this study can be employed for authenticity testing to determine which samples has been previously fired or unannealed and in determination of thermal histories of quartz generally. 5. Heating rate of 1C/s is suggested to be used for the TL reading that serves as thermal activation in pre-dose technique. 6. The pre-dose sensitisation of RT LMOSL can serve as a complementary o dating methods to the traditional 110 C TL peak. 150 UNIVERSITY OF IBADAN LIBRARY REFERENCES o Adamiec, G., 2005. Properties of the 360 and 550 nm TL emissions of the „110 C peak‟ in fired quartz. Radiation Measurements. 39: 105 – 110. Adamiec, G., Bluszcz, A., Bailey, R., Garcia-Talavera, M., 2006. Finding model parameters. Genetic algorithms and the numerical modelling of quartz luminescence. Radiation Measurements 41: 897 – 902. Afouxenidis, D., Polymeris, G. S., Tsirliganis, N. C., Kitis, G., 2012. Computerised curve deconvolution of TL/OSL curves using a popular spreadsheet program. Radiation Protection Dosimetry. 149 (4): 363–370. Aitken, M. J., 1985. Thermoluminescence Dating. Academic Press, London. Aitken, M. J., Smith, B. W., 1988. Optical dating. recuperation after bleaching Quaternary Science Reviews. 7: 387–393. Aitken, M. J., 1998. An Introduction to Optical Dating. Oxford University Press, Oxford. Asfora, V. K., Guzzo, P, L., Pessis, A-M., Barros, V. S. M., Watanabe, S., Khoury, H. J., 2014. Characterization of the burning conditions of archaeological pebbles o using the thermal sensitization of the 110 C TL peak of quartz. Radiation Measurements, http://dx.doi.org/10.1016/j.radmeas.2014.04.022. Bailey, R. M., 2001. Towards a general kinetic model for optically and thermally stimulated luminescence of quartz, Radiation Measurements. 33: 17-45. Bailiff, I. K., 1994. The pre-dose technique, Radiation Measurements. 23 (2/3): 471- 479. Bailiff, I. K., 2000. Characteristics of time-resolved luminescence in quartz. Radiation Measurements. 32: 401–405. Balian, H. G., Eddy, N. W., 1977. Figure-of-merit (FOM). an improved criterion over the normalized Chi-squared test for assessing goodness-of-fit of gamma-ray spectral peaks. Nuclear Instruments and Methods. 145: 389–395. Bartnitskaya, T. S., Vlasova, M. V., Zelyavskii, V. B., Kakazei, N. G., Kurdyumov, A. V., Sukhikh, L. L., Chistyakov, V. I., 1992. Defect structure formation in silicon on grinding, Soviet Powder Metallurgy and Ceramics. 31: 903-907. Battaglia, S., Fanzini, M., Leoni, L., 1993. Influence of grinding methods on the 101 X- ray powder diffraction line of a-quartz. in methods and practices in X-ray powder diffraction, ICDD, USA. 151 UNIVERSITY OF IBADAN LIBRARY Betts, D. S., Townsend, P. D., 1993. Temperature distribution in thermoluminescence experiments. II. Some calculational models. Journal of Physics D: Applied Physics. 26: 849–857. Betts, D. S., Couturier, L., Khayarat, A. H,, Luff, B. J., Townsend. 1993. Temperature distribution in thermoluminescence experiments. I. Experimental results. Journal of Physics D. Applied Physics. 26: 843–848. Botter-Jensen, L., 1997. Luminescence techniques. instrumentation and methods. Radiation Measurements, 27(5/6): 749-768. Botter-Jensen, L., Duller, G. A. T., Murray, A. S., Banerjee, D. 1999. Blue light emitting diodes for optical stimulation of quartz in retrospective dosimetry and dating. Radiation Protection Dosimetry, 84: 335-340. Botter-Jensen, L., Bulur, E., Duller, G. A. T., Murray, A. S., 2000. Advances in luminescence instrument systems. Radiation Measurements. 32: 523-528. Botter-Jessen, L., McKeever, S. W. S., Wintle, A. G., 2003a. Optically stimulated luminescence dosimetry. Elsevier Science B.V. Amsterdam. Botter-Jensen, L., Andersen, C. E., Duller, G. A. T., and Murray, A. S. 2003b. Developments in radiation, stimulation and observation facilities in luminescence measurements. Radiation Measurements. 37: 535-541. Bulur, E., 1996. An alternative technique for optically stimulated luminescence (OSL) experiment. Radiation Measurements. 26: 701–709. o Chen, R., 1979. Saturation of sensitisation of the 110 C TL peak in quartz and its potential application in pre-dose technique, Eur. PACT 3: 325-335. Chen, R., Kristianpoller, N., Davidson, Z., Visocekas, R., 1981. Mixed-order and second order kinetics in thermally stimulated processes. Journal of Luminescence. 23 (3,4): 293–303. Chen, R., Yang, X. H., McKeever, S. W. S., 1988. Strongly superlinear dose dependence of thermoluminescence in synthetic quartz. Journal of Physics D: Applied Physics 21: 1452-1457. Chen, R., Fogel, G., Kristianpoller, N., 1994. Theoretical account of the sensitisation and de-sensitisation in quartz. Radiation Measurements. 23: 277-279. Chen, R., Mckeever, S. W. S., 1997. Theory of Thermoluminescence and Related Phenomena. World Scientific Publishing, Singapore. 152 UNIVERSITY OF IBADAN LIBRARY Chen, G., Li. S. -H., 2000. Studies of quartz 110 °C thermoluminescence peak sensitivity change and its relevance to optically stimulated luminescence dating. Journal of Physics D: Applied Physics. 33: 437-443. o Chen, G., Li, S. -H., Murray, A. S., 2000. Study of the 110 C TL peak sensitivity in optical dating of quartz. Radiation Measurements. 32: 641–645. Chen, R., Leung, P. L., 2002. The decay of OSL signals as stretched exponential functions. Radiation measurements. 37: 519 – 526. Chen, R, Pagonis, V., 2003. Modelling thermal activation characteristics of the sensitisation of thermoluminescence in quartz. Journal of Physics D: Applied Physics. 36: 1–6. Chithambo, M. L., Galloway, R. B., 2000. A pulsed light-emitting-diode system for stimulation of luminescence. Measurement Science Technology 11: 418–424 Curie, D., 1963. Luminescence in Crystals. Wiley, New York. Deer, W. A., Howie, R. A., Wise, W. S., Zussman, J., 2004. Rock-Forming Minerals, nd V4b. Framework Silicates. Silica Minerals, Feldspathoids and the Zeolites. 2 edition. Geological society of London, Xv pp +982. Duller A. T., 1997. Behavioural studies of stimulated luminescence from feldspar. Radiation Measurements. 27 (5/6): 663-694. Duller, G. A. T., 2003. Distinguishing quartz and feldspar in single grain luminescence measurements. Radiation Measurements 37: 161–165. Facey, R., A., 1996. Heating-rate effects in glow peak measurements for thermoluminescent dosimetry, Health Physics. 12: 717-720. Ferreira de Souza, L. B., Guzzo, P. L., Khoury, H. J., 2014. OSL and photo-transferred TL of quartz single crystals sensitised by high-dose of gamma-radiation and moderate heat-treatments. Applied Radiation and Isotopes. 94: 93–100. o Figel, M., Goedicke, C., 1999. Simulation of the pre-dose effect of the 100 C TL peak in quartz. Radiation Protection Dosimetry. 84: 433–438. Fleming, S. J., Thompson, J., 1970. Quartz as a heat resistant dosimeter. Health Physics. 18: 567–568. Forman, S. L., Pierson, J., Smith. R. P., Hackett, W. R., Valentine, G., 1994. Assessing the accuracy of thermoluminescence for dating backed sediments beneath late Quaternary lava flows, Snake River plain, Idaho. Journal of Geophysical Research. 99 B (8): 15569-15576. 153 UNIVERSITY OF IBADAN LIBRARY Franklin, A. D., Prescott, J. R., Scholefield, R. B., 1995. The mechanism of thermoluminescence in an Australian sedimentary quartz. Journal of Luminescence. 63: 317–326. Furetta, C., 2003. Handbook of Thermoluminescence. World Scientific publishing Co. Pte.Ltd., Singapore. Galli, A., Martini, M., Montanari, C., Panzeri, L., Sibila, E., 2006. TL of fine-grain samples from quartz-rich archaeological ceramics. dosimetry using the 110 and o 210 C TL peaks. Radiation Measurements. 41: 1009-1014. Garlick, G. F. J., Gibson, A. F., 1948. The electron trap mechanism of luminescence in sulphide and silicate phosphors. Proc. Roy. Soc. London A60: 574-590. Gotlib, V. I., Kantorovich, L. N., Grebenshicov, V. L., Bichev, V. R., Nemiro, E. A., 1984. The study of thermoluminescence using the contact method of sample heating. Journal of Physics D: Applied Physics. 17: 2097–2114. Gumnior, M., Preusser, F., 2007. Late Quaternary river development in the southwest Chad Basin. OSL dating of sediment from the Komadugu palaeofloodplain (northeast Nigeria). Journal of Quaternary Science. 22: 709–719. Han, Z. Y., Li, S. H., Tso, M. Y. W., 2000. Effect of annealing on TL sensitivity of granitic quartz, Radiation Measurements. 32: 227-231. Hashimoto, T., Sakaue, S., Ichino, M., 1994. Dependence of TL-property changes of natural quartzes on aluminum contents accompanied by thermal annealing treatment. Radiation Measurements. 23: 293-299. Horowitz, Y. S., 1984. Thermoluminescence and Thermoluminescent dosimetry, volume Vol.1. CRC Press. Horowitz, Y. S. Yossian, D., 1995. Computerized glow curve deconvolution . application to thermoluminescence dosimetry. Radiation Protection Dosimetry. Spec. Issue, 60. Huntley, D. J., Godfrey-Smith, D. I., Thewalt, M. L. W., 1985. Optical dating of sediments. Nature. 313: 105–107. Huntley, D. J., Godfrey-Smith, D. I., Haskell, E. H., 1991. Light-induced emission spectra from some quartz and feldspars. Nuclear Tracks and Radiation Measurements. 18: 127-131. Hutt, G., Jaek, I., and Tchonka, J., 1988. Optical dating. K-feldspars optical response stimulation spectra. Quaternary Science Reviews, 7: 381-385. 154 UNIVERSITY OF IBADAN LIBRARY Ivliev, A. I., Kashkarov, L. L., Kalinina, G. V., 2006. Comparative Thermoluminescence characteristics of the different origin natural quartz. Electronic Scientific Information Journal. Herald of the Department of Earth Science RAS. 1(24): 1-2. Jain, M., Murray, A.S., Bøtter-Jensen, L., 2003. Characterisation of blue-light stimulated luminescence components in different quartz samples. implications for dose measurements. Radiation Measurements. 37: 441–449. Jones, C. E., Embree, D., 1976. Correlations of the 4.77-4.28eV luminescence band in silicon dioxide with oxygen vacancy. Journal of Applied Physics. 47: 5365- 5371. Kitis, G., Spiropulu, M.. Papadopoulos, J., Charalambous, S., 1993. Heating rate effects on the TL glow-peaks of three thermoluminescent phosphors. Nuclear Instruments and Methods in Physics Research. 73: 367-372. Kitis, G. Gomez-Ros, J. M. Tuyn, J. W. N., 1998 . Thermoluminescence glowcurve deconvolution functions for first, second and general orders of kinetics. Journal of Physics D. Applied Physics. 31: 2636-2641. Kitis, G., Tuyn, J. W. N., 1998. A simple method to correct for the temperature lag in TL glow-curve measurements. Journal of Physics D: Applied Physics. 31: 2065–2073. Kitis, G., 2001. TL glow-curve deconvoluion functions for various kinetic orders and continuous trap distribution. Acceptance criteria for E and s values. Journal of Radioanalytical and Nuclear Chemistry. 247 (3): 697-703. Kitis, G., Pagonis, V., 2008. Computerized curve deconvolution analysis for LM-OSL. Radiation Measurements. 43(2/6): 737 – 741. Kitis, G., Kiyak, N. G., Polymeris, G. S. Tsirliganis, N. C., 2010. The correlation of fast o OSL component with the TL peak at 325 C in quartz of various origins. Journal of Luminescence. 130: 298–303. Kitis, G., Pagonis, V., 2013. Analytical solutions for stimulated luminescence emission from tunneling recombination in random distributions of defects. Journal of Luminescence. 137 : 109–115. Kiyak, N. G., Polymeris, G. S., Kitis, G., 2007. Component resolved OSL dose response and sensitisation of various sedimentary quartz samples. Radiation Measurements. 42: 144–155. 155 UNIVERSITY OF IBADAN LIBRARY Kiyak, N. G., Polymeris, G. S., Kitis, G., 2008. LM–OSL thermal activation curves of o quartz. relevance to the thermal activation of the 110 C TL glow peak. Radiation Measurements 43: 263–268. Koul, D. K., Nambi, K. S. V., Singhvi, A. K., Bhat, C. L., Gupta, P. K., 1996. o Feasibility of estimating firing temperature using 110 C TL peak of quartz, Applied Radiation and Isotope. 47: 191-196. Koul, D. K., Chougaonkar, M. P., 2007. The pre-dose phenomenon in the OSL signal of quartz. Radiation Measurements 42: 1265–1272. o Koul, D. K., 2008. 110 C TL glow peak of quartz – a brief review. Pramana. 71: 1209- 1229. Koul, D. K., Polymeris, G. S., Tsirliganis, N. C., Kitis, G., 2010. Possibility of pure o thermal sensitisation in the pre-dose mechanism of the 110 C TL peak of quartz. Nuclear Instruments and Methods in Physics Research B. 268: 493–498. Koul, D. K., Polymeris, G. S., 2013. Impact of firing on the optically stimulated luminescence of geological quartz. AIP Conference Proceedings. 1512: 916– 917. Koul, D. K., Patil, P. G., Oniya, E. O., Polymeris, G. S., 2014. Investigating the thermally transferred optically stimulated luminescence source trap in fired geological quartz. Radiation Measurement. 62(2014), 60–70. Krbetschek, M. R., Gotze, J., Dietrich, A., Trautmann, T., 1997. Spectral information from minerals relevant for luminescence dating. Radiation Measurements. 27(5/6): 695-748. Lersen, N. A., 1997. Dosimetry based on thermally and optically stimulated luminescence. Dissertation submitted June 1997 for the Ph.D degree at the Niels Bohr Institute, University of Copenhagen. Li, S.-H., Yin, G.-M., 2001. Luminescence dating of young volcanic activity in China, Quaternary Science and Reviews. 20: 865-868. Li, S.-H., 2002. Luminescence sensitivity changes of quartz by bleaching, annealing and uv exposure. Radiation Effects and Defects in Solids. 157: 357–364. May, C. E., Partridge, J. A., 1964. Thermoluminescence kinetics of alpha irradiated alkali halides. J. Chem. Phys. 40. 1401-1409. McKeever, S. W. S., Strain, J. A., Townsend, P. D., Udval, P., 1983. Effects of thermal cycling on the thermoluminescence and radioluminescence of quartz. PACT 9: 123-132. 156 UNIVERSITY OF IBADAN LIBRARY McKeever, S. W. S., 1991. Mechanisms of thermoluminescence production, some problems and a few answers. Nuclear Tracks and Radiation Measurements 18: 5-12. McKeever, S. W. S., Chen, R., 1997. Luminescence models. Radiation Measurements. 27(5/6): 625-661. McKeever, S. W. S., Bùtter-Jensen, L., Agersnap Larsen, N., Duller, G. A. T., 1997. Temperature dependence of OSL decay curves. experimental and theoretical aspects. Radiation Measurements. 27(2): 161-170. Michael, C. T., Zacharias, N., Maniatis, Y., Dimotikali, D., 1997. A new technique (foil technique) for measuring the natural dose in TL dating and its application in the dating of a mortar containing ceramic fragments. Ancient TL. 15(2/3): 36–42. Mott, N. F., Gurney, R. W., 1948. Electronic Processes in Ionic Crystals, 2nd ed. Oxford University Press, London. Murray, A. S., Roberts, R. G., 1998. Measurement of the equivalent dose in quartz using a regenerative-dose single aliquot protocol. Radiation Measurements 29(5): 503 – 515. Murray, A. S., Wintle, A. G., 2000. Luminescence dating of quartz using an improved single-aliquot regenerative-dose protocol. Radiation Measurements. 32: 57–73. Nanjundaswamy, R., Lepper, K., McKeever, S. W. S., 2002. Thermal quenching of thermoluminescence in natural quartz. Radiation Protection Dosimetry. 100: 305-308. Ogundare, F. O., Chithambo, M. L., Oniya, E. O., 2006. Anomalous behaviour of thermoluminescence from quartz. A case of glow peaks from a Nigerian quartz. Radiation Measurements. 41: 549-553. Ogundare, F. O., Chithambo, M. L., 2007a. Thermoluminescence kinetic analysis of quartz with a glow peak that shifts in an unusual manner with irradiation dose. Journal of Physics D. Applied Physics. 40: 247–253. Ogundare, F.O., Chithambo, M.L., 2007b. Time resolved luminescence of quartz from Nigeria. Optical Materials. 29: 1844–1851. Ogundare, F. O., Chithambo, M. L., 2008. The influence of optical bleaching on lifetimes and luminescence intensity in the slow component of optically stimulated luminescence of natural quartz from Nigeria. Journal of Luminescence. 128: 1561–1569. 157 UNIVERSITY OF IBADAN LIBRARY Oniya, E. O., Polymeris, G. S., Tsirliganis, N. C., Kitis, G., 2012a. On the pre-dose sensitization of the various components of the LM-OSL signal of annealed o quartz; Comparison with the case of 110 C TL peak. Radiation Measurements. 47: 864-869. Oniya, E. O., Polymeris, G. S., Tsirliganis, N. C., Kitis, G., 2012b. Sensitization of high temperature thermoluminescence glow-curve peaks in various quartz samples. Geochronometria. 39(3): 212-220. Pagonis, V., Kitis, G., Furetta, C., 2006. Numerical and Practical Exercises in Thermoluminescence. Springer Science+Business Media, Inc. New York. Pagonis, V., Kitis, G., 2012. Prevalence of first-order kinetics in thermoluminescence materials: An explanation based on multiple competition processes. Physica Status Solidi B. 249(8): 1590–1601. Petrov, S. A., Bailiff, I. K., 1996. Thermal quenching and the Initial Rise technique of trap depth evaluation. Journal of Luminescence. 65: 289-291. Piters, T. M., Bos, A. J. J., 1994. Effects of nonideal heat transfer on the glow curve in thermoluminescence experiments. Journal of Physics D: Applied Physics. 27: 1747–1756. Polymeris, G. S., Afouxenidis, D., Tsirliganis, N. C., Kitis, G., 2009. The TL and room o temperature OSL properties of the glow peak at 110 C in natural milky quartz. A case study. Radiation Measurements. 44: 23–31. Polymeris, G. S., 2015. OSL at elevated temperatures: Towards the simultaneous thermal and optical stimulation. Radiation Physics and Chemistry. 106: 184– 192. Prescott, J. R., Robertson. G. B., 1997. Sediment dating by luminescence. a review. Radiation Measurements. 27: 893-922. Preusser, F., Chithambo, M. L., Götte, T., Martini, M., Ramseyer, K., Sendezera, E. J., Susino, G. J., Wintle, A. G., 2009. Quartz as a natural luminescence dosimeter. Earth-Science Reviews. 97: 184–214. Randall, J. T., Wilkins, M. H. F., 1945. Phosphorescence and electron traps II- The interpretation of long period phosphorescence. Proct. R. Soc. London Ser. A, 184: 390-407. Ranjbar, A. H., Durrani, S. A., Randle, K., 1999. Electron spin resonance and thermoluminescence in powder form of clear fused quartz; Effects of grinding, Radiation Measurements. 30: 73-81. 158 UNIVERSITY OF IBADAN LIBRARY Rendell, H. M, Townsend, P. D., Wood, R. A., Luff, B. J., 1994. Thermal treatment and emission spectra of TL from quartz, Radiation Measurements. 23: 441-449. Roberts, R. G., 1997. Luminescence dating in archaeology from origins to optical. Radiation Measurements. 27: 819-892. Sadek, A. M., Eissa, H. M., Basha, A. M., Kitis, G., 2014. Resolving the limitation of the peak fitting and peak shape methods in the determination of the activation energy of thermoluminescence glow peaks. Journal of Luminescence. 146: 418– 423. Sawakuchi, G. O., Okuno, E. 2004. Effect of high gamma doses in quartz. Nuclear Instruments and Methods in Physics Research. B. 218: 217-221. Schilles, T. Poolton, N. R. J., Bulur E., Bøtter-Jensen L., Murray A. S., Smith G. M., Riedi P C., Wagner G. A., 2001. A multi-spectroscopic study of luminescence sensitivity changes in natural quartz induced by high-temperature annealing. Journal of Physics D. Applied Physics. 34: 722–731. Singhvi, A. K., Krbetschek, M. R., 1996. Luminescence dating of arid zone sediments. A review and perspective. Ann. Arid Zone. 35 (3): 249 – 279. Smith, B. W., Rhodes, E. J., Stokes, S., Spooner, N. A., 1990. The optical dating of sediments using quartz. Radiation Protection Dosimetry. 34: 75-78. Spooner, N. A., Questiaux, D. G., 1989. Optical dating - Achenheim beyond the Eemian using green and infrared stimulation. Proceedings of Workshop on Long and Short Range Limits in Luminescence Dating. In. Occasional Publication 9, Research Laboratory for Archaeology and the History of Art, Oxford. Spooner, N. A., 1994. On the optical dating signal from quartz. Radiation Measurements. 23: 593-600. Stokes, S., 1994. The timing of OSL sensitivity changes in a natural quartz. Radiation Measurements. 23: 593–600. o Stoneham, D., Stokes, S., 1991. An investigation of the relationship between the 110 C TL peak and optically stimulated luminescence in sedimentary quartz. Nuclear Tracks and Radiation Measurements 18: 119–123. Subedi, B., Oniya, E., Polymeris, G.S., Afouxenidis, D., Tsirliganis, N. C., Kitis, G., 2011. Thermal quenching of thermoluminescence in quartz samples of various origin. Nuclear Instruments and Methods in Physics Research B. 269: 572–581. 159 UNIVERSITY OF IBADAN LIBRARY Takeuchi, A., Nagahama, H., Hashimoto, T., 2006. Surface resetting of thermoluminescence in milled quartz grains, Radiation Measurements 41(7/8): 826-830. Takeuchi, A., Hashimoto, T., 2008. Milling-induced reset of thermoluminescence and deformation of hydroxyl species in the near-surface layers of quartz grains, Geochronometria. 32: 61-68. Taylor, G. C., Lilley, E., 1982. Effect of clustering and precipitation on thermoluminescene in LiF (TLD-100) crystals. Journal of Physics D: Applied Physics. 15: 1253-1263. Topaksu, M., Dogan, T., Yüksel, M., Kurt, K., Topak, Y., Yegingil, Z., 2014. Comparative study of the thermoluminescence properties of natural metamorphic quartz belonging to Turkey and Spain. Radiation Physics and Chemistry. 96: 223–228. Wintle, A. G., 1973. Anomalous fading of thermoluminescence in mineral samples. Nature. 245: 143–144. Wintle, A. G.,1975. Thermal Quenching of Thermoluminescence in Quartz. Geophys. J. Roy. Astronom. Sot. 41: 107-113. Wintle, A. G., 1997. Luminescence dating. laboratory procedures and protocols. Radiation Measurements. 27(5/6): 769-817. Wintle, A. G., Murray, A. S., 1998. Towards the development of a preheat procedure for OSL dating of quartz. Radiation Measurements. 29: 81–94. Wintle, A. G., Murray, A. S., 1999. Luminescence sensitivity changes in quartz. Radiation Measurements. 30: 107–118. Wintle, A. G., Murray. A. S., 2006. A review of quartz optically stimulated luminescence characteristics and their relevance in single-aliquot regeneration dating protocols. Radiation Measurements. 41: 369 – 391. Yang, X. H., McKeever, S.W.S., 1990. The pre-dose effect in crystalline quartz, Journal of Physics D: Applied Physics. 23: 237-244. o Zimmerman, J., 1971. The radiation-induced increase of the 110 C thermoluminescence sensitivity of fired quartz, Journal of Physics. C: Solid State Physics. 4: 3265-3276. 160 UNIVERSITY OF IBADAN LIBRARY APPENDIX 1 161 UNIVERSITY OF IBADAN LIBRARY 162 UNIVERSITY OF IBADAN LIBRARY 163 UNIVERSITY OF IBADAN LIBRARY 164 UNIVERSITY OF IBADAN LIBRARY 165 UNIVERSITY OF IBADAN LIBRARY 166 UNIVERSITY OF IBADAN LIBRARY APPENDIX 2 167 UNIVERSITY OF IBADAN LIBRARY 168 UNIVERSITY OF IBADAN LIBRARY 169 UNIVERSITY OF IBADAN LIBRARY 170 UNIVERSITY OF IBADAN LIBRARY 171 UNIVERSITY OF IBADAN LIBRARY 172 UNIVERSITY OF IBADAN LIBRARY 173 UNIVERSITY OF IBADAN LIBRARY 174 UNIVERSITY OF IBADAN LIBRARY