- Email: [email protected]
Evaluation of trapping parameters of annealed natural quartz
Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Evaluation of trapping parameters of annealed natural quartz Rui Zhou a,b, Ming-Jian Wei a,⇑, Bo Song a,c, Yan Zhang a,d, Qiu-Yue Zhao e, Bao-Lin Pan a, Teng-Fei Li a a
College of Resources, Environment & Tourism, Capital Normal University, 100048 Beijing, People’s Republic of China Shisanling Seismic Station, Institute of Earthquake Science, CEA, 102200 Beijing, People’s Republic of China c Beijing Jing Yuan School, 100040 Beijing, People’s Republic of China d School of TaiPingqiao, Nan Lu of West Railway Station, 100073 Beijing, People’s Republic of China e Key Laboratory of Tourism and Resources Environment in Universities of Shandong, Taishan University, 271000 Tai’an, People’s Republic of China b
a r t i c l e
i n f o
Article history: Received 28 November 2015 Received in revised form 23 February 2016 Accepted 29 February 2016
Keywords: Natural quartz Thermoluminescence Trap Parameter
a b s t r a c t The thermoluminescence (TL) trapping parameters of annealed quartz have been investigated. The apparent TL peaks observed at temperatures of 133 °C, 211 °C, 266 °C and 405 °C, respectively, were named Peak I, Peak II, Peak III and Peak IV. The Tm Tstop method is applied to investigate the number of peaks and their positions, and to obtain the trap distributions in the quartz. Peak shape (PS), Hoogenstraaten method (Various Heating Rates Method, VHR), and Computerized Glow Curve Deconvolution (CGCD) are used to evaluate the trapping parameters of the annealed quartz. The glow curve can be considered as a superposition of at least nine overlapping peaks. These peaks show up at 133 °C, 211 °C, 266 °C, 308 °C, 333 °C, 384 °C, 441 °C, 466 °C and 484 °C. The PS method can be only used in evaluating the parameters for Peaks I. The VHR method can be used in evaluating the trapping parameters for the first three peaks. CGCD method is complementary to obtaining parameters for the sub-peaks, and the thermal quenching correction with the Urbach’s method is necessary. The Urbach’s coefficient for the quartz is 30.03 kTm. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction The natural quartz plays an important role in retrospective dosimetry, in particular for the dating of Quaternary sediments and archaeological materials. Extensive work on natural quartz, as a result, has been done on different aspects, such as electron spin resonance (ESR), optically stimulated luminescence (OSL) and thermoluminescence (TL) [1]. For thermoluminescence, the TL peaks in alpha-quartz is observed at 110 (150), 220, 325 and 375 °C at a heating of 20 °Cs1. In some literatures, the peak at 110 °C was found to occur between 90 °C and 130 °C, the peak supposedly at 250 °C is observed at 220 °C [2]. The glow curves show a variety of shapes depending on the nature and origin of quartz. Different glow curves result in the diversities of trap types in different types of quartz. However, widely different values of the trap parameters have been reported in the literature, because of the differences in the origin of the quartz, thermal treatment, impurity content and methods used to evaluate the parameters [3–7]. These parameters are the basis of retrospective dosimetry and luminescence dating. One of the critical assumptions in
⇑ Corresponding author. E-mail addresses: [email protected], [email protected] (M.-J. Wei). http://dx.doi.org/10.1016/j.nimb.2016.02.067 0168-583X/Ó 2016 Elsevier B.V. All rights reserved.
thermoluminescence dating is that the trapped electrons used for dating are stable over the age being dated. Thus, it is significant to study the trapping parameters of thermoluminescence traps in quartz for thermoluminescence dating. In fact, the differences between the trapping parameters of different quartz should be more extensively studied. As far as the study nowadays concerned, the trapping parameters of different genetic quartz, or the quartz originated from different geological environments need to be cautiously treated, the trapping parameters and glow curves of these different quartz seem different because of its minimal distinction and is still not enough to be used in thermoluminescence dating based on previous studies, the research on trapping parameters of different types of natural quartz still need more work to do. In the case for quartz, the first-order kinetics has been supported by a majority of literatures, while some other authors believe that some types of quartz exhibiting non-first-order behavior is because of complex overlapping peaks [8]. The presence of the overlapping peaks is the difficulty in determining trapping parameters [9]. Therefore, there is neither an agreement on the published values of the kinetic parameters obtained from different types of quartz nor consistent progression [7–9]. The purpose of this investigation is to determine the trapping parameter of annealed high purity quartz obtained by various methods. The main TL
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
peaks are isolated from others, and their properties are revealed. Three methods are used to deduce the trapping parameters from glow curves. The most significant issue here is to separate the overlapping peaks from the glow curve. Whenever the peaks are clearly separated from each other, the traditional methods are suitable for obtaining these parameters. The number of peaks existing in the glow curve and their positions are identified with the Tm Tstop method. Peak shape (PS), Hoogenstraaten method (Various Heating Rates Method, VHR), and Computerized Glow Curve Deconvolution (CGCD) are used to determine the kinetic parameters of the quartz. The activation energy is a very important parameter for the luminescence dating and retrospective dosimetry. The above-mentioned methods are used to evaluate the peaks in different temperatures of the quartz. 2. Materials and methods 2.1. Sample preparation The sample used in this experiment is a piece of natural quartz which has been manually ground and sieved to 74–100 lm grain size in a dark room. As shown in Table 1, X-ray fluorescence (XRF) spectrometry was used to determine the chemical composition. The measurements were completed in Analysis and Test center of Peking University, Beijing. The powdery sample was put in a crucible gradually heated up to 460 °C for an hour in a furnace. The previously accumulated luminescence was erased before the next irradiation. The powder was placed on several aluminum discs (9.7 mm in diameter), and was fixed with silicon grease. The grains covered about 100% of the disc area. They were irradiated with a calibrated 90Sr/90Y beta source with a dose rate of 8 Gymin1 for quartz samples on discs. 2.2. Measurements The measurements were all made in a flowing nitrogen atmosphere by a RGD-3B manual reader manufactured by DML. The RGD-3B reader had been interfaced with a personal computer [10–12]. The treatments and measurements of the sample were made under red light to avoid the release of the trapped electrons from the semi-stable sites into hole centers (including luminescence centers) due to light sensitivity. A data extraction module based on Excel VBA function was developed to extract the data from RGD-3B TL reader. The lowest heating rate supported by the reader is 2 °Cs1. Without special description, all the measurements were performed with a linear heating rate of 5 °Cs1. After the irradiation, the measurements were carried on at once. 3. Experimental 3.1. Glow curve and dose response curve The glow curves of quartz were obtained after being exposed to various beta radiation doses in a range of 16–1024 Gy, measured at a heating rate of 5 °Cs1. The apparent luminescence peak may be only one genuine peak or consist of several peaks at different temperatures, and can be read directly from the glow-curve. The photon counts of the first apparent peak (Peak I) were overflowed when the discs were irradiated with 512 Gy or 1024 Gy beta doses,
Table 1 Results of X-ray fluorescence (XRF) analysis (Wt%). SiO2 (%)
CaO (%)
Al2O3 (%)
Fe2O3 (%)
Norm. (%)
99.900 0
0.072 9
0.036 1
0.022 1
100.000 0
33
respectively. In order to get the dose response curves, five parallel aliquots were all read out. As is shown in Fig. 1, the glow curves exhibit four apparent thermoluminescence peaks in the temperature range between room temperature to 500 °C. The four apparent peaks clearly emerged at temperatures close to 133 °C, 211 °C, 266 °C and 405 °C (hereafter as Peak I, Peak II, Peak III and Peak IV) are observed in the above-mentioned temperature scanning region. The reason for no apparent peak observed below 100 °C is that the luminescence signal is weak in thermal stability, and the TL signal has completely attenuated before the measurements. To understand the dose response characteristics of apparent Peak III, the baseline is taken into consideration. The thermoluminescence intensity between 250 °C and 270 °C was recorded. The luminescence intensity change with the dose was plotted in Fig. 2(a). The sensitivity for the dose response of each aliquot is quite different, maybe due to the grinding. In the same way as for apparent Peak IV, the luminescence intensity in the range from 400 °C to 420 °C was recorded and plotted in Fig. 2(b). But for highest dose in the figure, the dose of 1024 Gy, the sum of luminescence intensity exceeded the upper limit of statistical software designed by the TL reader, and only three out of five aliquots intensity sums can be recorded in Fig. 2(a) and four aliquots are recorded in Fig. 2(b). The thermoluminescence intensity increases with the rise of doses irradiated with. It has been found that the thermoluminescence signal of this quartz has the requisite TL properties for radiation dosimetry and can be potentially used for geological dating.
3.2. Tm Tstop plot The peak identification means examining the number of peaks from the glow curve. The Tm Tstop method was applied to investigate the number of the peaks and their positions, and was used to obtain the trap distributions in the quartz [13]. All discs prepared for this experiment were annealed in a FJ-427A2 dosimeter (CNNC Beijing Nuclear Instrument Factory), heated from room temperature to 500 °C, followed by irradiation with 120 Gy 90Sr/90Y beta particles. First, a disc was heated to the stop temperature which began with the temperature of 100 °C. Then, the next disc was started from room temperature and ended in a selected stop temperature which was in a small increment of about 5 °C. The thermoluminescence intensity of each disc was recorded. After that, the same disc was measured again with the same heating rate in a temperature range from room temperature to 500 °C, and was sustained at 500 °C for 20 s. The position of the first maximum in the second turn was recorded as Tm. As the measurement processes repeated, some different stop temperatures (Tstop) were emerged, and Tm versus Tstop can be plotted as Fig. 3. The peaks illustrated in Fig. 3 are real peaks which only can be read after peak identification. In the paper, we called it peak or sub-peak of the glow-curve. The peak identification indicates that there are at least nine peaks in the glow curve. These peaks are at 133 °C, 211 °C, 266 °C, 308 °C, 333 °C, 384 °C, 441 °C, 466 °C and 484 °C, respectively. The first three apparent peaks are single peaks at 133, 211, 266 °C. And the last apparent peak (Peak IV, at 405 °C) is a composite peak, consists of six peaks, which are at 308 °C, 333 °C, 384 °C, 441 °C, 466 °C and 484 °C, respectively. Quartz extracted from archaeological and geological materials exhibits either or both of two common TL peaks, the ‘‘325 °C” peak (Rapidly Bleaching Peak, RBP) and the ‘‘375 °C” peak, (Slowly Bleaching Peak, SBP). These peaks are shown in quotes because they apply only to a heating rate of 20 °Cs1. For a lower heating rate, the peaks occur at lower temperatures [14].
34
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
Fig. 1. The grow curves of an annealed quartz irradiated with various doses(5 °Cs1).
Fig. 2. The dose response curves of annealed quartz irradiated with various doses (a. Peak III, b. Peak IV).
is observed at 133 °C at a heating rate of 5 °Cs1. Glow-peak IV centered at 405 °C is a highly overlapping peak at a heating rate of 5 °Cs1, and its sub-peak centered at 308 °C at a heating rate of 5 °Cs1 is ‘‘325 °C” peak (RBP). As is shown in Fig. 1, with peak separation, there are small peaks observed around 350 °C, and the ‘‘375 °C” peak (SBP) consists of sub-peaks at 333 °C and 384 °C. 3.3. Peak separation
Fig. 3. Tm Tstop plot at a heating rate of 5 °Cs1 with 120 Gy beta doses.
The conventionally known 110 °C, 325 °C and 375 °C peaks occur at about 100 °C, 305 °C and 350 °C at heating rate of 5 °Cs1, respectively [15]. For this quartz, Peak I is the ‘‘110 °C” peak, which
The peak separation is of significance to evaluate the trapping parameters for the quartz [16]. Most of traditional peak analysis methods demand that the peak be isolated. The data for Tm Tstop can be used to separate composite TL curves into their constituent components [9,17–20]. The results of peak separation will be used as basis of peak shape method. The first two apparent peaks are single peaks, but the luminescence intensity of apparent Peak III is relatively low, and the fluctuation is relatively strong. Peak IV consists of several smaller peaks, the isolated peak is not a single peak. Thus, the peak separation is only carried on Peak I and Peak II, and the specific method is described below. For instance, the procedures for separating Peak I are as follows. The range of predominant peak is identified.
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
Two groups of data for glow curves at stop temperatures 0 °C and 160 °C are extracted, and named columns I and II, respectively. The isolated apparent Peak I can be obtained from column II subtracted from column I. Then the start and end points of the glow curves should be identified. In this case, it can be seen that the predominant region of this peak is from 93 °C to 160 °C. The suitable temperature range could be selected and plotted as Fig. 4(a). As seen in Fig. 4, apparent Peaks I and II are separated with the abovementioned method. 4. Glow curve analysis 4.1. Peak shape method The first order kinetics is given by Randall and Wilkins [21],
Z E s T E exp I ¼ n0 s exp exp 0 dT 0 kT b T0 kT
ð1Þ
where I is the TL intensity, n0 is the initial concentration, E is activation energy, s is a constant (s1), called ‘‘escape frequency factor”, k is Boltzmann’s constant, and b is linear heating rates, T = T0 + b t, T is in absolute temperature. The second order kinetics is given by Garlick and Gibson, which applies when retrapping and recombination probabilities are equal [22].
2 0 Z T E n0 s E 1þ I ¼ n0 s0 exp exp 0 dT 0 kT b kT T0
ð2Þ
The peak shape method is deduced, respectively from firstorder kinetics and second-order kinetics by Chen [23]. Lushchik obtained,
4.2. Hoogenstraaten method When the measurements are performed at different heating rates, various heating rates (VHR) method can be utilized to evaluate the trapping parameters. The Hoogenstraaten method is a technique that uses various heating rates to obtain the activation energy E in the case of the first-order kinetic. The Hoogenstraaten method is used to estimate the trapping parameters of the quartz at the heating rates of 2 °Cs1, 3 °Cs1, 4 °Cs1, 5 °Cs1 or 6 °Cs1. All discs for this experiment were irradiated with 120 Gy beta doses. Five parallel aliquots were all measured at the same heating rate; total twenty-five aliquots were measured in this experiment. ln(Tm2/b) versus 1/Tm can be graphed in Hoogenstraaten plot, where Tm is the peak temperature. For the first order kinetics, an equation can be expressed as [25–28],
T 2m ln b
kT 2m d
ð3Þ
and this is the equation multiplied by correction factor Cd which was calculated to be 0.976. Chen slightly modified Eq. (3), the modified expression can be written as follows,
E ¼ ½0:976 þ 7:3ðlg 0:42Þ
kT 2m d
! ð4Þ
where d = T2 Tm, lg = (T2 Tm/T2 T1), k is Boltzmann’s constant. T1 and T2 stand for the left and right temperatures at the half maximum intensity [24].
! ¼
E E ; þ ln kT m sk
ð5Þ
where Tm is the peak temperature, E the activation energy, s the frequency factor, b the heating rate, and k the Boltzmann’s constant. 4.3. CGCD method The software for CGCD analysis is restricted to first-order kinetics. The software is used to analyze glow curves on computer. As is shown is Eq. (1), the exponential integral is not solvable analytically. In this case, the exponential integral is approximated by the following expression [29],
Z 0
E ¼ Cd
35
T
Z E E 1 02 E 1 exp 0 dT 0 x expðx0 Þdx0 ¼ E2 ðxÞ; k x k x kT
ð6Þ
E where x ¼ kT ðx0 ¼ kTE 0 Þ, E2(x) is the second exponential integral function which can be evaluated by E2 ðxÞ ¼ aðxÞ expðxÞ., where þa1 xþa2 xþa3 xþa4 x aðxÞ ¼ ba00 þb 1 xþb2 xþb3 xþb4 x
Finally, in terms of maximum condition, the Eq. (1) can be approximated by
E E IðTÞ ¼ Im exp kT m kT E E T E E E exp a exp a kT m Tm Tm kT m kT kT
ð7Þ
where Tm and Im represent maximum temperature and maximum intensity, respectively. E represents the activation energy.
Fig. 4. The isolated peak of quartz separated by the above-mentioned method (a. Peak I, b. Peak II).
36
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
The number of sub-peaks and their positions have been deduced with Tm Tstop method [30–31]. Figure of merit (FOM) was a tool to evaluate the quality of fitting [32]. The only acceptable fitting results are the values of FOM between 0% and 3.5%. The Urbach’s coefficient of this quartz is also investigated [33]. 5. Results and discussion 5.1. Results
Table 2 Trapping parameters evaluated by peak shape (PS) method. Peak
T1 (°C)
T2 (°C)
lg
d
E (eV)
Peak I
113
147
0.41
14
0.65
Fig. 5. The glow curve of the annealed quartz with different rates of heating, in °Cs1. Table 3 Trapping parameters evaluated by various heating rates (VHR) method. Heating rates 1
2°Cs 3°Cs1 4°Cs1 5°Cs1 6°Cs1 E (eV)
Peak I
Peak II
Peak III
Peak IV
101 122 129 133 145 0.56
180 186 192 204 212 0.96
235 256 260 266 275 1.13
381 398 400 405 430 2.53
12.5
Peak I Peak II Peak III Peak IV
12.0
11.5 ln(Tm2/β)
In the case of apparent Peak I, for PS method, T1 = 113 °C, T2 = 147 °C, lg = (T2 Tm/T2 T1) = 0.41, d = Tm T1 = 14. Thus, Peak I exhibits the first-order behavior according to PS method. According to formula (4), the activation energy E for Peak I is calculated to be 0.65 eV (Table 2). The PS method is commonly used to obtain trapping parameters for peaks in a certain distance. This method can be applied relatively safely only in Peak I. If the peaks are too close to each other, the thermoluminescence intensity from isolated peaks is quite low. Low intensity causes difficulties in accurately recognizing the factors, such as T1, T2 and d. The isolated peak using PS method can often be directly seen in the glow curve [9]. The smaller peaks identified with Tm Tstop method are difficult to isolate in the glow curve because it may need to irradiate with relatively higher doses. The parameters have been investigated with Hoogenstraaten method. The glow curves are corrected so that each point to correspond to the sample sampling temperature interval. As plotted in Fig. 5, the thermoluminescence peaks moved toward higher temperatures with the rise of linear heating rates. The intensity of luminescence deceases with the increase of heating rates after the correction, but except for at the hating rate of 5 °Cs1. The intensity of luminescence at the heating rate of 5 °Cs1 is roughly equivalent to that at the heating rate of 3 °Cs1. The positions of the peaks in different heating rates are recognized and recorded in Table 3. As shown in Fig. 7, Hoogenstraaten plot is a plot of ln(Tm2/b) against 1/Tm which is often used in the analysis of trapping parameters. The linear segment for the four peaks are linearly fitted, and a straight line of slope E/k with intercept ln(sk/E) was obtained. The parameter of each peak is calculated with VHR method. The activation energy of the four apparent peaks evaluated with VHR are 0.56 eV, 0.96 eV, 1.13 eV and 2.53 eV, respectively. The parameter of Peak I lower than PS may result from the instability of the low temperature peak. And Peak IV is overestimated. The overestimation of high temperature peaks also occurred in literature [7]. With the advantage of VHR method for the first-order kinetic peaks, the data should be taken at the peak maximum, which in the case of a large peak surrounded by smaller ones can be accurately evaluated from the glow curve. The parameter evaluation of highly overlapping peaks in comparable intensities is difficult for Various Heating Rates Method [7]. In general, VHR is suitable for obtaining the parameters when the main peak is relatively higher, and the small sub-peaks beside them are quite low. Whenever the parameters of main peaks need to be evaluated and the smaller sub-peaks can be ignored, VHR method is a better choice. According to Fig. 6, it can be concluded that peaks at high temperature have a better linear coefficient than the ones at low temperatures. The linear coefficient of this quartz is worse than that from previous literatures.
11.0
10.5
10.0 0.0014 0.0016 0.0018 0.0020 0.0022 0.0024 0.0026
1/Tm(K-1) Fig. 6. Hoogenstraaten plot of the annealed quartz (ln(Tm2/b) against 1/Tm).
Combined with identified sub-peaks, the trapping parameters of ten sub-peaks deduced from the glow curves are estimated by a Computerized Glow Curve Deconvolution software called Glow Fit based on the first-order kinetics. The glow curves measured at a heat rate of 5 °Cs1 are fitted. The figure of merit (FOM) of this fitting result is 3.33%. The result of curve fitting is nearly in accordance with the observation in the peak isolating experiment. The trapping parameters for the quartz are attempted to be fitted to other fourteen curves obtained in different heating rates. The activation energy E is allowed to automatically match around to
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
37
Fig. 7. A glow curve fitted by Glow Fit after irradiated with 120 Gy beta doses at room temperature.
Tm at other heating rates corresponding to heating rates 5 °Cs1. The fitting result is shown in Fig. 7 (see Table 4). The parameter of Peak I deduced from CGCD is reliable. The peaks around 181 °C and 238 °C are identified with CGCD. The result indicates that the peaks at 211 °C and 266 °C may be composed by two smaller peaks. The peaks in the range from 300 °C to 400 °C can be found with CGCD method, but the intensity of peaks is quite low. The peaks in higher temperatures are composed by a series of sub-peaks which are nearly continuous. It is a better way to obtain the sub-peaks of Peak IV using CGCD method. The trapping parameters deduced from CGCD are basically consistent with PS, and can provide the details for the last apparent peak. The conclusion can be drawn as follows. For this quartz, the PS method can be only used in evaluating the parameters for Peaks I safely. Glow curve analysis using CGCD method is complementary to obtaining parameters for sub-peaks. VHR method is less useful to the peaks in higher temperatures.
Fig. 8. Hoogenstraaten plot of the annealed quartz (ln(Tm2/b) against 1/Tm) obtained by the same aliquot.
The parameters of apparent Peak III evaluated by VHR method are smaller than that by CGCD method. In the previous experiment, five parallel discs were used in the measurements at different heating rates, and each disc was measured only once. In the discussion, the same aliquot measured in different heating rates has been also involved. Three aliquots in the same heating rate were measured. Linear fitting using the same aliquot is shown in Fig. 8. The activation energy evaluated with VHR method using same aliquot is 0.31 eV, 0.48 eV, 1.25 eV and 2.90 eV, respectively. The sequence of heating rates increases gradually from 2 °Cs1 to 6 °Cs1. The activation energy of low temperature peaks (Peaks I and II) is much lower than the one obtained by several aliquots since discs are repeatedly heated. The integrated TL signal of apparent Peak I, Peak II, Peak III and Peak V as a function of the heating rate is plotted in Fig. 6, respectively. The luminescence intensity of Peak I decreases with the
5.2. Discussion The glow curves of the quartz have been studied in this paper. Apparent TL peaks appeared at temperatures of 133 °C, 211 °C, 266 °C and 405 °C in the glow curve. Compared with previous literatures, the intermediate peak in the range from 300 °C to 400 °C is absence in this case, and it may be caused by low intensity. The peaks around 130 °C, and 200–210 °C are identified in this quartz [2,7]. A quantitative indicator for the first maximum Tm in the second measurement in Tm Tstop method should be established. Lack of the quantitative indicators may result in arbitrariness for peak identification. The duration shouldn’t be set in the end of the first turn measurement in Tm Tstop method. The duration may affect the second turn which will cause nearby peaks being annealed. The dose response curves show that the linear range of glow curve is divided into several parts. The dosimetric properties of this quartz should be further studied.
Table 4 Trapping parameters evaluated by Computer Glow Curve Deconvolution (CGCD). Tm (°C)
E (eV)
Tm (°C)
E (eV)
139.95 181.35 209.35 238.05 272.45
0.790 1.000 1.140 1.160 1.210
309.55 332.85 402.85 442.25 486.15
1.245 1.250 1.480 1.690 1.800
Fig. 9. The behavior of the peak intensity of glow peak represented in Fig. 5 as a function of heating rate.
38
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
Fig. 11. The Urbach’s plot of the annealed quartz.
Fig. 10. The thermal quenching efficiency at different heating rates, Tm at a heating rate of b = 5 °Cs1.
increase of heating rate, but except for the heating rate at 5 °Cs1. As is shown in Fig. 9, the thermal quenching effect is present. The results of activation energy should be carefully discussed. The thermal quenching effect of quartz has been extensively studied among the TL minerals [8]. The expression for the thermal quenching luminescence efficiency can be written as [34],
g¼
AQ 1 ¼ W AUQ 1 þ C ekT
ð8Þ
The thermal quenching efficiencies of the quartz at different heating rates are plotted in Fig. 10. The first three apparent peaks are shown in the figure. The last peak is highly overlapping, and the ending points of that peak at different heating rates are difficult to be identified. The quenching efficiency decreases with the increase of temperature from 400 K to 485 K, and at the heating rate of 3 °Cs1, the efficiency continuously still decreases, while at the other heating rates the efficiency increases the temperature. The thermal quenching efficiencies decreases with the heating rates at the same temperature, but except for the heating rate at 5 °Cs1. The reason why the intensity and efficiency at a heating rate of 5 °Cs1 higher than that of at the heating rate of 4 °Cs1–6 °Cs1 is unknown, but definitely the phenomenon have been observed in the quartz. For quartz, according to criterion from the reference [35], the VHR method can be used is applied to thermally quenched experimental peaks, it can give more reliable results for narrow TL peaks and for g(T) values much larger than 0.5. Moreover, the VHR method can also be applied to thermally quenching narrow first order. Thus, the parameters of Peaks I and II are reliable with VHR method, and the parameter of apparent Peak III is relatively accurate. The overestimation can be seen in apparent Peak IV, because the overlapping sub-peaks. The thermal quenching effect influences the symmetry factor and kinetic order, as a result, influences parameter evaluated by PS method. However, it is possible to evaluate the parameter of Peak I, the influence of thermal quenching on the peak shape methods is negligible when g(T) varies between 1 and 0.7. The trapping parameter of Peak I deduced by PS method is reliable. The CGCD method can be safely used to quenched TL the thermal quenching efficiency g(T) is larger than
0.7, as the PS method, lower than that efficiency, the activation energies may be underestimated. The parameters of Peak I deduced by PS and VHR are lower than CGCD, which may be due to the influence of in instability of the peak at the room temperature. The activation energies of Peak I are 0.65 eV, 0.56 eV and 0.79 eV, respectively by three different methods. As shown in Fig. 11, the Urbach’s plot (E kTm) has also been investigated in this paper. The trapping parameters of peaks in higher temperatures may be underestimated with CGCD method. The linear relationship can be attempted to correct the parameters of the peaks higher than Peak III. The Urbach’s coefficient of the quartz is 30.03 kTm. The Urbach’s coefficient can be used in evaluating the trapping parameters for the annealed quartz.
6. Conclusion Kinetic parameters of high purity natural quartz investigated with thermoluminesce are presented in this paper. These parameters are the basis of retrospective dosimetry and luminescence dating. One of the critical assumptions in thermoluminescence dating is that the trapped electrons used for dating are stable over the age being dated [36]. Thus, it is significant to accurately evaluate the trapping parameters of quartz. The trapping parameters of different quartz are widely different values because of the differences between the origin of the quartz, thermal treatment, impurity content and methods used to evaluate the parameters. The suitable traps are potential to breakout the upper limit of luminescence dating. Meanwhile the deeper traps may be difficult to bleach and the residual may be relatively high. Evaluating the trap depths of quartz is a key problem for luminescence dating. It is of certain significance to apply proper methods to evaluate the activation energy in geology materials, and different methods used to evaluate the parameters are needed in specific case study. Further studies should attempt to select a trap in geology samples which can be easily bleached, and has a relatively long lifetime. For retrospective dosimetry, pure quartz can be considered as a reference to study the defect characteristics of doped quartz. (1) The apparent TL peaks at the temperatures of 133 °C, 211 °C, 266 °C and 405 °C can be considered as a superposition of at least nine overlapping peaks. These peaks are at temperatures of 133 °C, 211 °C, 266 °C, 308 °C, 333 °C, 384 °C, 441 °C, 466 °C and 484 °C.
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
(2) The PS method can be only used in evaluating the parameters for Peaks I. The VHR method can be used in evaluating the trapping parameters for the first three peaks. CGCD method is complementary to obtaining parameters for the sub-peaks. The value by VHR method is lower for Peak I than those by PS and CGCD, and higher for Peak IV than CGCD. The trapping parameters of Peak III evaluated by VHR are lower than the other two approaches. Glow curve analysis using CGCD method is complementary to obtaining parameters for sub-peaks in the isolated, highly overlapped peaks above 266 °C. (3) The thermal quenching effect is present in this quartz. The thermal quenching correction with the Urbach’s method is necessary. The Urbach’s coefficient for annealed quartz is 30.03 kTm. The dosimetric property will be precisely investigated in the future. The signal of quartz investigated in this paper will be used in future experiments as a recognition indicator to distinguish the quartz TL signal from poly-mineral materials in luminescence dating with utilization of poly-mineral.
Acknowledgments This work is co-supported by the National Natural Science Foundation of China (NSFC, Grant No. 41471007 and Grant Nos. 41301006, 40871017) and the Beijing Municipal Natural Science Foundation Key Project (B) No. KZ201210028034. The authors are thankful to an anonymous reviewer for his/her constructive comment, which greatly help us improving our manuscript. The authors wish to thank Prof. Huhou Li for instruction in the experiment. Many thanks to Junxin Liu, Tiantian Liu for their laboratory assistance, and Li Zhang for assistance with the XRF analysis.
39
References [1] F. Preusser, M.L. Chithambo, T. Götte, M. Martini, K. Ramseyer, E.J. Sendezera, G.J. Susino, A.G. Wintle, in: Earth Sci. Rev 97 (2009) 184–214. [2] T.M. Brito Farias, S. Watanabe, J. Lumin. 132 (2012) 2684–2692. [3] V. Pagonis, E. Tatsis, G. Kitis, C. Drupieski, Radiat. Prot. Dosim. 100 (2002) 373– 376. [4] G. Kitis, V. Pagonis, H. Carty, E. Tatsis, in: Radiat. Prot. Dosim 100 (2002) 225– 228. [5] N.A. Spooner, D.G. Questiaux, Radiat. Meas. 32 (2000) 659–666. [6] J.M. Gómez-Ros, V. Correcher, J. García-Guinea, A. Delgado, Prot. Dosim. 100 (2002) 399–402. [7] A.N. Yazici, M. Topaksu, J. Phys. D Appl. Phys. 36 (2003) 620–627. [8] A.G. Wintle, in: Geophys. J. R. Astron. Soc. 41 (1975) 107–113. [9] D. Mebhah, D. Imatoukene, F.Z. Abdelazziz, Z. Lounis-Mokrani, Radiat. Meas. 41 (2006) 813–818. [10] K. Tang, H. Cui, Y. Wang, H. Zhu, Q. Fan, Y. Mao, Radiat. Meas. 43 (2008) 1187– 1190. [11] K. Tang, H. Cui, H. Zhu, Z. Liu, H. Fan, Radiat. Meas. 47 (2012) 185–189. [12] K. Tang, H. Cui, H. Zhu, Z. Liu, H. Fan, Radiat. Meas. 48 (2013) 43–46. [13] S.W.S. Mckeever, in: Nucl. Instr. Meth. 175 (1980) 19–20. [14] A.D. Franklin, in: Ancient TL 12 (1994) 5–9. [15] A.D. Franklin, J. Lumin. 75 (1997) 71–76. [16] F.O. Ogundare, M.L. Chithambo, E.O. Oniya, Radiat. Meas. 41 (2006) 549–553. [17] A.N. Yazici, R. Chen, S. Solakand, Z. Yegingi, J. Phys. D Appl. Phys. 35 (2002) 2526–2535. [18] A.N. Yazici, K. Kurt, V.E. Kafadar, S. Gürler, M. Topaksu, in: Nucl. Instr. Meth. B 248 (2006) 133–141. [19] A.N. Yazici, M. Bedir, A.S. Sökücü, in: Nucl. Instr. Meth. B 259 (2007) 955–965. [20] G.C. Taylor, E. Lilley, J. Phys. D Appl. Phys. 11 (1978) 567–581. [21] J.T. Randall, M.H.F. Wilkins, Proc. R. Soc. London Ser. A 184 (1945) 365–389. [22] G.F.J. Garlick, A.F. Gibson, Proc. Phys. Soc. 60 (1948) 574–690. [23] R. Chen, J. Appl. Phys. 41 (1969) 570–585. [24] S.S. Paliwal, O.P. Sinha, Phys. Status Solidi A 38 (1976) 73–76. [25] R. Chen, S.A.A. Winer, J. Appl. Phys. 41 (1970) 5227–5232. [26] S.-H. Li, M.Y.W. Tso, N.W.L. Wong, Radiat. Meas. 27 (1997) 43–47. [27] S.-H. Li, G.-M, J. Lumin. 121 (2006) 51–56. [28] S.-H. Li, M.Y.W. Tso, Radiat. Meas. 27 (1997) 119–121. [29] M. Puchalska, P. Bilski, Radiat. Meas. 41 (2006) 659–664. [30] V. Pagonis, G. Kitis, Radiat. Prot. Dosim. 101 (2002) 93–98. [31] B.A. Sharma, A.N. Singh, S.N. Singh, O.B. Singh, Radiat. Meas. 44 (2009) 32–37. [32] H.G. Balian, N.W. Eddy, Nucl. Instr. Meth. 145 (1977) 389–395. [33] R.K. Gartia, L.L. Singh, Radiat. Meas. 46 (2011) 664–668. [34] J.M. Kalita, G. Wary, J. Lumin. 146 (2014) 508–511. [35] B. Subedi, G. Kitis, V. Pagonis, in: Phys. Status Solidi A 207 (2010) 1216–1226. [36] B. Li, S.-H. Li, J. Lumin. 136 (2013) 5–10.
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Evaluation of trapping parameters of annealed natural quartz Rui Zhou a,b, Ming-Jian Wei a,⇑, Bo Song a,c, Yan Zhang a,d, Qiu-Yue Zhao e, Bao-Lin Pan a, Teng-Fei Li a a
College of Resources, Environment & Tourism, Capital Normal University, 100048 Beijing, People’s Republic of China Shisanling Seismic Station, Institute of Earthquake Science, CEA, 102200 Beijing, People’s Republic of China c Beijing Jing Yuan School, 100040 Beijing, People’s Republic of China d School of TaiPingqiao, Nan Lu of West Railway Station, 100073 Beijing, People’s Republic of China e Key Laboratory of Tourism and Resources Environment in Universities of Shandong, Taishan University, 271000 Tai’an, People’s Republic of China b
a r t i c l e
i n f o
Article history: Received 28 November 2015 Received in revised form 23 February 2016 Accepted 29 February 2016
Keywords: Natural quartz Thermoluminescence Trap Parameter
a b s t r a c t The thermoluminescence (TL) trapping parameters of annealed quartz have been investigated. The apparent TL peaks observed at temperatures of 133 °C, 211 °C, 266 °C and 405 °C, respectively, were named Peak I, Peak II, Peak III and Peak IV. The Tm Tstop method is applied to investigate the number of peaks and their positions, and to obtain the trap distributions in the quartz. Peak shape (PS), Hoogenstraaten method (Various Heating Rates Method, VHR), and Computerized Glow Curve Deconvolution (CGCD) are used to evaluate the trapping parameters of the annealed quartz. The glow curve can be considered as a superposition of at least nine overlapping peaks. These peaks show up at 133 °C, 211 °C, 266 °C, 308 °C, 333 °C, 384 °C, 441 °C, 466 °C and 484 °C. The PS method can be only used in evaluating the parameters for Peaks I. The VHR method can be used in evaluating the trapping parameters for the first three peaks. CGCD method is complementary to obtaining parameters for the sub-peaks, and the thermal quenching correction with the Urbach’s method is necessary. The Urbach’s coefficient for the quartz is 30.03 kTm. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction The natural quartz plays an important role in retrospective dosimetry, in particular for the dating of Quaternary sediments and archaeological materials. Extensive work on natural quartz, as a result, has been done on different aspects, such as electron spin resonance (ESR), optically stimulated luminescence (OSL) and thermoluminescence (TL) [1]. For thermoluminescence, the TL peaks in alpha-quartz is observed at 110 (150), 220, 325 and 375 °C at a heating of 20 °Cs1. In some literatures, the peak at 110 °C was found to occur between 90 °C and 130 °C, the peak supposedly at 250 °C is observed at 220 °C [2]. The glow curves show a variety of shapes depending on the nature and origin of quartz. Different glow curves result in the diversities of trap types in different types of quartz. However, widely different values of the trap parameters have been reported in the literature, because of the differences in the origin of the quartz, thermal treatment, impurity content and methods used to evaluate the parameters [3–7]. These parameters are the basis of retrospective dosimetry and luminescence dating. One of the critical assumptions in
⇑ Corresponding author. E-mail addresses: [email protected], [email protected] (M.-J. Wei). http://dx.doi.org/10.1016/j.nimb.2016.02.067 0168-583X/Ó 2016 Elsevier B.V. All rights reserved.
thermoluminescence dating is that the trapped electrons used for dating are stable over the age being dated. Thus, it is significant to study the trapping parameters of thermoluminescence traps in quartz for thermoluminescence dating. In fact, the differences between the trapping parameters of different quartz should be more extensively studied. As far as the study nowadays concerned, the trapping parameters of different genetic quartz, or the quartz originated from different geological environments need to be cautiously treated, the trapping parameters and glow curves of these different quartz seem different because of its minimal distinction and is still not enough to be used in thermoluminescence dating based on previous studies, the research on trapping parameters of different types of natural quartz still need more work to do. In the case for quartz, the first-order kinetics has been supported by a majority of literatures, while some other authors believe that some types of quartz exhibiting non-first-order behavior is because of complex overlapping peaks [8]. The presence of the overlapping peaks is the difficulty in determining trapping parameters [9]. Therefore, there is neither an agreement on the published values of the kinetic parameters obtained from different types of quartz nor consistent progression [7–9]. The purpose of this investigation is to determine the trapping parameter of annealed high purity quartz obtained by various methods. The main TL
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
peaks are isolated from others, and their properties are revealed. Three methods are used to deduce the trapping parameters from glow curves. The most significant issue here is to separate the overlapping peaks from the glow curve. Whenever the peaks are clearly separated from each other, the traditional methods are suitable for obtaining these parameters. The number of peaks existing in the glow curve and their positions are identified with the Tm Tstop method. Peak shape (PS), Hoogenstraaten method (Various Heating Rates Method, VHR), and Computerized Glow Curve Deconvolution (CGCD) are used to determine the kinetic parameters of the quartz. The activation energy is a very important parameter for the luminescence dating and retrospective dosimetry. The above-mentioned methods are used to evaluate the peaks in different temperatures of the quartz. 2. Materials and methods 2.1. Sample preparation The sample used in this experiment is a piece of natural quartz which has been manually ground and sieved to 74–100 lm grain size in a dark room. As shown in Table 1, X-ray fluorescence (XRF) spectrometry was used to determine the chemical composition. The measurements were completed in Analysis and Test center of Peking University, Beijing. The powdery sample was put in a crucible gradually heated up to 460 °C for an hour in a furnace. The previously accumulated luminescence was erased before the next irradiation. The powder was placed on several aluminum discs (9.7 mm in diameter), and was fixed with silicon grease. The grains covered about 100% of the disc area. They were irradiated with a calibrated 90Sr/90Y beta source with a dose rate of 8 Gymin1 for quartz samples on discs. 2.2. Measurements The measurements were all made in a flowing nitrogen atmosphere by a RGD-3B manual reader manufactured by DML. The RGD-3B reader had been interfaced with a personal computer [10–12]. The treatments and measurements of the sample were made under red light to avoid the release of the trapped electrons from the semi-stable sites into hole centers (including luminescence centers) due to light sensitivity. A data extraction module based on Excel VBA function was developed to extract the data from RGD-3B TL reader. The lowest heating rate supported by the reader is 2 °Cs1. Without special description, all the measurements were performed with a linear heating rate of 5 °Cs1. After the irradiation, the measurements were carried on at once. 3. Experimental 3.1. Glow curve and dose response curve The glow curves of quartz were obtained after being exposed to various beta radiation doses in a range of 16–1024 Gy, measured at a heating rate of 5 °Cs1. The apparent luminescence peak may be only one genuine peak or consist of several peaks at different temperatures, and can be read directly from the glow-curve. The photon counts of the first apparent peak (Peak I) were overflowed when the discs were irradiated with 512 Gy or 1024 Gy beta doses,
Table 1 Results of X-ray fluorescence (XRF) analysis (Wt%). SiO2 (%)
CaO (%)
Al2O3 (%)
Fe2O3 (%)
Norm. (%)
99.900 0
0.072 9
0.036 1
0.022 1
100.000 0
33
respectively. In order to get the dose response curves, five parallel aliquots were all read out. As is shown in Fig. 1, the glow curves exhibit four apparent thermoluminescence peaks in the temperature range between room temperature to 500 °C. The four apparent peaks clearly emerged at temperatures close to 133 °C, 211 °C, 266 °C and 405 °C (hereafter as Peak I, Peak II, Peak III and Peak IV) are observed in the above-mentioned temperature scanning region. The reason for no apparent peak observed below 100 °C is that the luminescence signal is weak in thermal stability, and the TL signal has completely attenuated before the measurements. To understand the dose response characteristics of apparent Peak III, the baseline is taken into consideration. The thermoluminescence intensity between 250 °C and 270 °C was recorded. The luminescence intensity change with the dose was plotted in Fig. 2(a). The sensitivity for the dose response of each aliquot is quite different, maybe due to the grinding. In the same way as for apparent Peak IV, the luminescence intensity in the range from 400 °C to 420 °C was recorded and plotted in Fig. 2(b). But for highest dose in the figure, the dose of 1024 Gy, the sum of luminescence intensity exceeded the upper limit of statistical software designed by the TL reader, and only three out of five aliquots intensity sums can be recorded in Fig. 2(a) and four aliquots are recorded in Fig. 2(b). The thermoluminescence intensity increases with the rise of doses irradiated with. It has been found that the thermoluminescence signal of this quartz has the requisite TL properties for radiation dosimetry and can be potentially used for geological dating.
3.2. Tm Tstop plot The peak identification means examining the number of peaks from the glow curve. The Tm Tstop method was applied to investigate the number of the peaks and their positions, and was used to obtain the trap distributions in the quartz [13]. All discs prepared for this experiment were annealed in a FJ-427A2 dosimeter (CNNC Beijing Nuclear Instrument Factory), heated from room temperature to 500 °C, followed by irradiation with 120 Gy 90Sr/90Y beta particles. First, a disc was heated to the stop temperature which began with the temperature of 100 °C. Then, the next disc was started from room temperature and ended in a selected stop temperature which was in a small increment of about 5 °C. The thermoluminescence intensity of each disc was recorded. After that, the same disc was measured again with the same heating rate in a temperature range from room temperature to 500 °C, and was sustained at 500 °C for 20 s. The position of the first maximum in the second turn was recorded as Tm. As the measurement processes repeated, some different stop temperatures (Tstop) were emerged, and Tm versus Tstop can be plotted as Fig. 3. The peaks illustrated in Fig. 3 are real peaks which only can be read after peak identification. In the paper, we called it peak or sub-peak of the glow-curve. The peak identification indicates that there are at least nine peaks in the glow curve. These peaks are at 133 °C, 211 °C, 266 °C, 308 °C, 333 °C, 384 °C, 441 °C, 466 °C and 484 °C, respectively. The first three apparent peaks are single peaks at 133, 211, 266 °C. And the last apparent peak (Peak IV, at 405 °C) is a composite peak, consists of six peaks, which are at 308 °C, 333 °C, 384 °C, 441 °C, 466 °C and 484 °C, respectively. Quartz extracted from archaeological and geological materials exhibits either or both of two common TL peaks, the ‘‘325 °C” peak (Rapidly Bleaching Peak, RBP) and the ‘‘375 °C” peak, (Slowly Bleaching Peak, SBP). These peaks are shown in quotes because they apply only to a heating rate of 20 °Cs1. For a lower heating rate, the peaks occur at lower temperatures [14].
34
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
Fig. 1. The grow curves of an annealed quartz irradiated with various doses(5 °Cs1).
Fig. 2. The dose response curves of annealed quartz irradiated with various doses (a. Peak III, b. Peak IV).
is observed at 133 °C at a heating rate of 5 °Cs1. Glow-peak IV centered at 405 °C is a highly overlapping peak at a heating rate of 5 °Cs1, and its sub-peak centered at 308 °C at a heating rate of 5 °Cs1 is ‘‘325 °C” peak (RBP). As is shown in Fig. 1, with peak separation, there are small peaks observed around 350 °C, and the ‘‘375 °C” peak (SBP) consists of sub-peaks at 333 °C and 384 °C. 3.3. Peak separation
Fig. 3. Tm Tstop plot at a heating rate of 5 °Cs1 with 120 Gy beta doses.
The conventionally known 110 °C, 325 °C and 375 °C peaks occur at about 100 °C, 305 °C and 350 °C at heating rate of 5 °Cs1, respectively [15]. For this quartz, Peak I is the ‘‘110 °C” peak, which
The peak separation is of significance to evaluate the trapping parameters for the quartz [16]. Most of traditional peak analysis methods demand that the peak be isolated. The data for Tm Tstop can be used to separate composite TL curves into their constituent components [9,17–20]. The results of peak separation will be used as basis of peak shape method. The first two apparent peaks are single peaks, but the luminescence intensity of apparent Peak III is relatively low, and the fluctuation is relatively strong. Peak IV consists of several smaller peaks, the isolated peak is not a single peak. Thus, the peak separation is only carried on Peak I and Peak II, and the specific method is described below. For instance, the procedures for separating Peak I are as follows. The range of predominant peak is identified.
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
Two groups of data for glow curves at stop temperatures 0 °C and 160 °C are extracted, and named columns I and II, respectively. The isolated apparent Peak I can be obtained from column II subtracted from column I. Then the start and end points of the glow curves should be identified. In this case, it can be seen that the predominant region of this peak is from 93 °C to 160 °C. The suitable temperature range could be selected and plotted as Fig. 4(a). As seen in Fig. 4, apparent Peaks I and II are separated with the abovementioned method. 4. Glow curve analysis 4.1. Peak shape method The first order kinetics is given by Randall and Wilkins [21],
Z E s T E exp I ¼ n0 s exp exp 0 dT 0 kT b T0 kT
ð1Þ
where I is the TL intensity, n0 is the initial concentration, E is activation energy, s is a constant (s1), called ‘‘escape frequency factor”, k is Boltzmann’s constant, and b is linear heating rates, T = T0 + b t, T is in absolute temperature. The second order kinetics is given by Garlick and Gibson, which applies when retrapping and recombination probabilities are equal [22].
2 0 Z T E n0 s E 1þ I ¼ n0 s0 exp exp 0 dT 0 kT b kT T0
ð2Þ
The peak shape method is deduced, respectively from firstorder kinetics and second-order kinetics by Chen [23]. Lushchik obtained,
4.2. Hoogenstraaten method When the measurements are performed at different heating rates, various heating rates (VHR) method can be utilized to evaluate the trapping parameters. The Hoogenstraaten method is a technique that uses various heating rates to obtain the activation energy E in the case of the first-order kinetic. The Hoogenstraaten method is used to estimate the trapping parameters of the quartz at the heating rates of 2 °Cs1, 3 °Cs1, 4 °Cs1, 5 °Cs1 or 6 °Cs1. All discs for this experiment were irradiated with 120 Gy beta doses. Five parallel aliquots were all measured at the same heating rate; total twenty-five aliquots were measured in this experiment. ln(Tm2/b) versus 1/Tm can be graphed in Hoogenstraaten plot, where Tm is the peak temperature. For the first order kinetics, an equation can be expressed as [25–28],
T 2m ln b
kT 2m d
ð3Þ
and this is the equation multiplied by correction factor Cd which was calculated to be 0.976. Chen slightly modified Eq. (3), the modified expression can be written as follows,
E ¼ ½0:976 þ 7:3ðlg 0:42Þ
kT 2m d
! ð4Þ
where d = T2 Tm, lg = (T2 Tm/T2 T1), k is Boltzmann’s constant. T1 and T2 stand for the left and right temperatures at the half maximum intensity [24].
! ¼
E E ; þ ln kT m sk
ð5Þ
where Tm is the peak temperature, E the activation energy, s the frequency factor, b the heating rate, and k the Boltzmann’s constant. 4.3. CGCD method The software for CGCD analysis is restricted to first-order kinetics. The software is used to analyze glow curves on computer. As is shown is Eq. (1), the exponential integral is not solvable analytically. In this case, the exponential integral is approximated by the following expression [29],
Z 0
E ¼ Cd
35
T
Z E E 1 02 E 1 exp 0 dT 0 x expðx0 Þdx0 ¼ E2 ðxÞ; k x k x kT
ð6Þ
E where x ¼ kT ðx0 ¼ kTE 0 Þ, E2(x) is the second exponential integral function which can be evaluated by E2 ðxÞ ¼ aðxÞ expðxÞ., where þa1 xþa2 xþa3 xþa4 x aðxÞ ¼ ba00 þb 1 xþb2 xþb3 xþb4 x
Finally, in terms of maximum condition, the Eq. (1) can be approximated by
E E IðTÞ ¼ Im exp kT m kT E E T E E E exp a exp a kT m Tm Tm kT m kT kT
ð7Þ
where Tm and Im represent maximum temperature and maximum intensity, respectively. E represents the activation energy.
Fig. 4. The isolated peak of quartz separated by the above-mentioned method (a. Peak I, b. Peak II).
36
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
The number of sub-peaks and their positions have been deduced with Tm Tstop method [30–31]. Figure of merit (FOM) was a tool to evaluate the quality of fitting [32]. The only acceptable fitting results are the values of FOM between 0% and 3.5%. The Urbach’s coefficient of this quartz is also investigated [33]. 5. Results and discussion 5.1. Results
Table 2 Trapping parameters evaluated by peak shape (PS) method. Peak
T1 (°C)
T2 (°C)
lg
d
E (eV)
Peak I
113
147
0.41
14
0.65
Fig. 5. The glow curve of the annealed quartz with different rates of heating, in °Cs1. Table 3 Trapping parameters evaluated by various heating rates (VHR) method. Heating rates 1
2°Cs 3°Cs1 4°Cs1 5°Cs1 6°Cs1 E (eV)
Peak I
Peak II
Peak III
Peak IV
101 122 129 133 145 0.56
180 186 192 204 212 0.96
235 256 260 266 275 1.13
381 398 400 405 430 2.53
12.5
Peak I Peak II Peak III Peak IV
12.0
11.5 ln(Tm2/β)
In the case of apparent Peak I, for PS method, T1 = 113 °C, T2 = 147 °C, lg = (T2 Tm/T2 T1) = 0.41, d = Tm T1 = 14. Thus, Peak I exhibits the first-order behavior according to PS method. According to formula (4), the activation energy E for Peak I is calculated to be 0.65 eV (Table 2). The PS method is commonly used to obtain trapping parameters for peaks in a certain distance. This method can be applied relatively safely only in Peak I. If the peaks are too close to each other, the thermoluminescence intensity from isolated peaks is quite low. Low intensity causes difficulties in accurately recognizing the factors, such as T1, T2 and d. The isolated peak using PS method can often be directly seen in the glow curve [9]. The smaller peaks identified with Tm Tstop method are difficult to isolate in the glow curve because it may need to irradiate with relatively higher doses. The parameters have been investigated with Hoogenstraaten method. The glow curves are corrected so that each point to correspond to the sample sampling temperature interval. As plotted in Fig. 5, the thermoluminescence peaks moved toward higher temperatures with the rise of linear heating rates. The intensity of luminescence deceases with the increase of heating rates after the correction, but except for at the hating rate of 5 °Cs1. The intensity of luminescence at the heating rate of 5 °Cs1 is roughly equivalent to that at the heating rate of 3 °Cs1. The positions of the peaks in different heating rates are recognized and recorded in Table 3. As shown in Fig. 7, Hoogenstraaten plot is a plot of ln(Tm2/b) against 1/Tm which is often used in the analysis of trapping parameters. The linear segment for the four peaks are linearly fitted, and a straight line of slope E/k with intercept ln(sk/E) was obtained. The parameter of each peak is calculated with VHR method. The activation energy of the four apparent peaks evaluated with VHR are 0.56 eV, 0.96 eV, 1.13 eV and 2.53 eV, respectively. The parameter of Peak I lower than PS may result from the instability of the low temperature peak. And Peak IV is overestimated. The overestimation of high temperature peaks also occurred in literature [7]. With the advantage of VHR method for the first-order kinetic peaks, the data should be taken at the peak maximum, which in the case of a large peak surrounded by smaller ones can be accurately evaluated from the glow curve. The parameter evaluation of highly overlapping peaks in comparable intensities is difficult for Various Heating Rates Method [7]. In general, VHR is suitable for obtaining the parameters when the main peak is relatively higher, and the small sub-peaks beside them are quite low. Whenever the parameters of main peaks need to be evaluated and the smaller sub-peaks can be ignored, VHR method is a better choice. According to Fig. 6, it can be concluded that peaks at high temperature have a better linear coefficient than the ones at low temperatures. The linear coefficient of this quartz is worse than that from previous literatures.
11.0
10.5
10.0 0.0014 0.0016 0.0018 0.0020 0.0022 0.0024 0.0026
1/Tm(K-1) Fig. 6. Hoogenstraaten plot of the annealed quartz (ln(Tm2/b) against 1/Tm).
Combined with identified sub-peaks, the trapping parameters of ten sub-peaks deduced from the glow curves are estimated by a Computerized Glow Curve Deconvolution software called Glow Fit based on the first-order kinetics. The glow curves measured at a heat rate of 5 °Cs1 are fitted. The figure of merit (FOM) of this fitting result is 3.33%. The result of curve fitting is nearly in accordance with the observation in the peak isolating experiment. The trapping parameters for the quartz are attempted to be fitted to other fourteen curves obtained in different heating rates. The activation energy E is allowed to automatically match around to
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
37
Fig. 7. A glow curve fitted by Glow Fit after irradiated with 120 Gy beta doses at room temperature.
Tm at other heating rates corresponding to heating rates 5 °Cs1. The fitting result is shown in Fig. 7 (see Table 4). The parameter of Peak I deduced from CGCD is reliable. The peaks around 181 °C and 238 °C are identified with CGCD. The result indicates that the peaks at 211 °C and 266 °C may be composed by two smaller peaks. The peaks in the range from 300 °C to 400 °C can be found with CGCD method, but the intensity of peaks is quite low. The peaks in higher temperatures are composed by a series of sub-peaks which are nearly continuous. It is a better way to obtain the sub-peaks of Peak IV using CGCD method. The trapping parameters deduced from CGCD are basically consistent with PS, and can provide the details for the last apparent peak. The conclusion can be drawn as follows. For this quartz, the PS method can be only used in evaluating the parameters for Peaks I safely. Glow curve analysis using CGCD method is complementary to obtaining parameters for sub-peaks. VHR method is less useful to the peaks in higher temperatures.
Fig. 8. Hoogenstraaten plot of the annealed quartz (ln(Tm2/b) against 1/Tm) obtained by the same aliquot.
The parameters of apparent Peak III evaluated by VHR method are smaller than that by CGCD method. In the previous experiment, five parallel discs were used in the measurements at different heating rates, and each disc was measured only once. In the discussion, the same aliquot measured in different heating rates has been also involved. Three aliquots in the same heating rate were measured. Linear fitting using the same aliquot is shown in Fig. 8. The activation energy evaluated with VHR method using same aliquot is 0.31 eV, 0.48 eV, 1.25 eV and 2.90 eV, respectively. The sequence of heating rates increases gradually from 2 °Cs1 to 6 °Cs1. The activation energy of low temperature peaks (Peaks I and II) is much lower than the one obtained by several aliquots since discs are repeatedly heated. The integrated TL signal of apparent Peak I, Peak II, Peak III and Peak V as a function of the heating rate is plotted in Fig. 6, respectively. The luminescence intensity of Peak I decreases with the
5.2. Discussion The glow curves of the quartz have been studied in this paper. Apparent TL peaks appeared at temperatures of 133 °C, 211 °C, 266 °C and 405 °C in the glow curve. Compared with previous literatures, the intermediate peak in the range from 300 °C to 400 °C is absence in this case, and it may be caused by low intensity. The peaks around 130 °C, and 200–210 °C are identified in this quartz [2,7]. A quantitative indicator for the first maximum Tm in the second measurement in Tm Tstop method should be established. Lack of the quantitative indicators may result in arbitrariness for peak identification. The duration shouldn’t be set in the end of the first turn measurement in Tm Tstop method. The duration may affect the second turn which will cause nearby peaks being annealed. The dose response curves show that the linear range of glow curve is divided into several parts. The dosimetric properties of this quartz should be further studied.
Table 4 Trapping parameters evaluated by Computer Glow Curve Deconvolution (CGCD). Tm (°C)
E (eV)
Tm (°C)
E (eV)
139.95 181.35 209.35 238.05 272.45
0.790 1.000 1.140 1.160 1.210
309.55 332.85 402.85 442.25 486.15
1.245 1.250 1.480 1.690 1.800
Fig. 9. The behavior of the peak intensity of glow peak represented in Fig. 5 as a function of heating rate.
38
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
Fig. 11. The Urbach’s plot of the annealed quartz.
Fig. 10. The thermal quenching efficiency at different heating rates, Tm at a heating rate of b = 5 °Cs1.
increase of heating rate, but except for the heating rate at 5 °Cs1. As is shown in Fig. 9, the thermal quenching effect is present. The results of activation energy should be carefully discussed. The thermal quenching effect of quartz has been extensively studied among the TL minerals [8]. The expression for the thermal quenching luminescence efficiency can be written as [34],
g¼
AQ 1 ¼ W AUQ 1 þ C ekT
ð8Þ
The thermal quenching efficiencies of the quartz at different heating rates are plotted in Fig. 10. The first three apparent peaks are shown in the figure. The last peak is highly overlapping, and the ending points of that peak at different heating rates are difficult to be identified. The quenching efficiency decreases with the increase of temperature from 400 K to 485 K, and at the heating rate of 3 °Cs1, the efficiency continuously still decreases, while at the other heating rates the efficiency increases the temperature. The thermal quenching efficiencies decreases with the heating rates at the same temperature, but except for the heating rate at 5 °Cs1. The reason why the intensity and efficiency at a heating rate of 5 °Cs1 higher than that of at the heating rate of 4 °Cs1–6 °Cs1 is unknown, but definitely the phenomenon have been observed in the quartz. For quartz, according to criterion from the reference [35], the VHR method can be used is applied to thermally quenched experimental peaks, it can give more reliable results for narrow TL peaks and for g(T) values much larger than 0.5. Moreover, the VHR method can also be applied to thermally quenching narrow first order. Thus, the parameters of Peaks I and II are reliable with VHR method, and the parameter of apparent Peak III is relatively accurate. The overestimation can be seen in apparent Peak IV, because the overlapping sub-peaks. The thermal quenching effect influences the symmetry factor and kinetic order, as a result, influences parameter evaluated by PS method. However, it is possible to evaluate the parameter of Peak I, the influence of thermal quenching on the peak shape methods is negligible when g(T) varies between 1 and 0.7. The trapping parameter of Peak I deduced by PS method is reliable. The CGCD method can be safely used to quenched TL the thermal quenching efficiency g(T) is larger than
0.7, as the PS method, lower than that efficiency, the activation energies may be underestimated. The parameters of Peak I deduced by PS and VHR are lower than CGCD, which may be due to the influence of in instability of the peak at the room temperature. The activation energies of Peak I are 0.65 eV, 0.56 eV and 0.79 eV, respectively by three different methods. As shown in Fig. 11, the Urbach’s plot (E kTm) has also been investigated in this paper. The trapping parameters of peaks in higher temperatures may be underestimated with CGCD method. The linear relationship can be attempted to correct the parameters of the peaks higher than Peak III. The Urbach’s coefficient of the quartz is 30.03 kTm. The Urbach’s coefficient can be used in evaluating the trapping parameters for the annealed quartz.
6. Conclusion Kinetic parameters of high purity natural quartz investigated with thermoluminesce are presented in this paper. These parameters are the basis of retrospective dosimetry and luminescence dating. One of the critical assumptions in thermoluminescence dating is that the trapped electrons used for dating are stable over the age being dated [36]. Thus, it is significant to accurately evaluate the trapping parameters of quartz. The trapping parameters of different quartz are widely different values because of the differences between the origin of the quartz, thermal treatment, impurity content and methods used to evaluate the parameters. The suitable traps are potential to breakout the upper limit of luminescence dating. Meanwhile the deeper traps may be difficult to bleach and the residual may be relatively high. Evaluating the trap depths of quartz is a key problem for luminescence dating. It is of certain significance to apply proper methods to evaluate the activation energy in geology materials, and different methods used to evaluate the parameters are needed in specific case study. Further studies should attempt to select a trap in geology samples which can be easily bleached, and has a relatively long lifetime. For retrospective dosimetry, pure quartz can be considered as a reference to study the defect characteristics of doped quartz. (1) The apparent TL peaks at the temperatures of 133 °C, 211 °C, 266 °C and 405 °C can be considered as a superposition of at least nine overlapping peaks. These peaks are at temperatures of 133 °C, 211 °C, 266 °C, 308 °C, 333 °C, 384 °C, 441 °C, 466 °C and 484 °C.
R. Zhou et al. / Nuclear Instruments and Methods in Physics Research B 375 (2016) 32–39
(2) The PS method can be only used in evaluating the parameters for Peaks I. The VHR method can be used in evaluating the trapping parameters for the first three peaks. CGCD method is complementary to obtaining parameters for the sub-peaks. The value by VHR method is lower for Peak I than those by PS and CGCD, and higher for Peak IV than CGCD. The trapping parameters of Peak III evaluated by VHR are lower than the other two approaches. Glow curve analysis using CGCD method is complementary to obtaining parameters for sub-peaks in the isolated, highly overlapped peaks above 266 °C. (3) The thermal quenching effect is present in this quartz. The thermal quenching correction with the Urbach’s method is necessary. The Urbach’s coefficient for annealed quartz is 30.03 kTm. The dosimetric property will be precisely investigated in the future. The signal of quartz investigated in this paper will be used in future experiments as a recognition indicator to distinguish the quartz TL signal from poly-mineral materials in luminescence dating with utilization of poly-mineral.
Acknowledgments This work is co-supported by the National Natural Science Foundation of China (NSFC, Grant No. 41471007 and Grant Nos. 41301006, 40871017) and the Beijing Municipal Natural Science Foundation Key Project (B) No. KZ201210028034. The authors are thankful to an anonymous reviewer for his/her constructive comment, which greatly help us improving our manuscript. The authors wish to thank Prof. Huhou Li for instruction in the experiment. Many thanks to Junxin Liu, Tiantian Liu for their laboratory assistance, and Li Zhang for assistance with the XRF analysis.
39
References [1] F. Preusser, M.L. Chithambo, T. Götte, M. Martini, K. Ramseyer, E.J. Sendezera, G.J. Susino, A.G. Wintle, in: Earth Sci. Rev 97 (2009) 184–214. [2] T.M. Brito Farias, S. Watanabe, J. Lumin. 132 (2012) 2684–2692. [3] V. Pagonis, E. Tatsis, G. Kitis, C. Drupieski, Radiat. Prot. Dosim. 100 (2002) 373– 376. [4] G. Kitis, V. Pagonis, H. Carty, E. Tatsis, in: Radiat. Prot. Dosim 100 (2002) 225– 228. [5] N.A. Spooner, D.G. Questiaux, Radiat. Meas. 32 (2000) 659–666. [6] J.M. Gómez-Ros, V. Correcher, J. García-Guinea, A. Delgado, Prot. Dosim. 100 (2002) 399–402. [7] A.N. Yazici, M. Topaksu, J. Phys. D Appl. Phys. 36 (2003) 620–627. [8] A.G. Wintle, in: Geophys. J. R. Astron. Soc. 41 (1975) 107–113. [9] D. Mebhah, D. Imatoukene, F.Z. Abdelazziz, Z. Lounis-Mokrani, Radiat. Meas. 41 (2006) 813–818. [10] K. Tang, H. Cui, Y. Wang, H. Zhu, Q. Fan, Y. Mao, Radiat. Meas. 43 (2008) 1187– 1190. [11] K. Tang, H. Cui, H. Zhu, Z. Liu, H. Fan, Radiat. Meas. 47 (2012) 185–189. [12] K. Tang, H. Cui, H. Zhu, Z. Liu, H. Fan, Radiat. Meas. 48 (2013) 43–46. [13] S.W.S. Mckeever, in: Nucl. Instr. Meth. 175 (1980) 19–20. [14] A.D. Franklin, in: Ancient TL 12 (1994) 5–9. [15] A.D. Franklin, J. Lumin. 75 (1997) 71–76. [16] F.O. Ogundare, M.L. Chithambo, E.O. Oniya, Radiat. Meas. 41 (2006) 549–553. [17] A.N. Yazici, R. Chen, S. Solakand, Z. Yegingi, J. Phys. D Appl. Phys. 35 (2002) 2526–2535. [18] A.N. Yazici, K. Kurt, V.E. Kafadar, S. Gürler, M. Topaksu, in: Nucl. Instr. Meth. B 248 (2006) 133–141. [19] A.N. Yazici, M. Bedir, A.S. Sökücü, in: Nucl. Instr. Meth. B 259 (2007) 955–965. [20] G.C. Taylor, E. Lilley, J. Phys. D Appl. Phys. 11 (1978) 567–581. [21] J.T. Randall, M.H.F. Wilkins, Proc. R. Soc. London Ser. A 184 (1945) 365–389. [22] G.F.J. Garlick, A.F. Gibson, Proc. Phys. Soc. 60 (1948) 574–690. [23] R. Chen, J. Appl. Phys. 41 (1969) 570–585. [24] S.S. Paliwal, O.P. Sinha, Phys. Status Solidi A 38 (1976) 73–76. [25] R. Chen, S.A.A. Winer, J. Appl. Phys. 41 (1970) 5227–5232. [26] S.-H. Li, M.Y.W. Tso, N.W.L. Wong, Radiat. Meas. 27 (1997) 43–47. [27] S.-H. Li, G.-M, J. Lumin. 121 (2006) 51–56. [28] S.-H. Li, M.Y.W. Tso, Radiat. Meas. 27 (1997) 119–121. [29] M. Puchalska, P. Bilski, Radiat. Meas. 41 (2006) 659–664. [30] V. Pagonis, G. Kitis, Radiat. Prot. Dosim. 101 (2002) 93–98. [31] B.A. Sharma, A.N. Singh, S.N. Singh, O.B. Singh, Radiat. Meas. 44 (2009) 32–37. [32] H.G. Balian, N.W. Eddy, Nucl. Instr. Meth. 145 (1977) 389–395. [33] R.K. Gartia, L.L. Singh, Radiat. Meas. 46 (2011) 664–668. [34] J.M. Kalita, G. Wary, J. Lumin. 146 (2014) 508–511. [35] B. Subedi, G. Kitis, V. Pagonis, in: Phys. Status Solidi A 207 (2010) 1216–1226. [36] B. Li, S.-H. Li, J. Lumin. 136 (2013) 5–10.