Assessing the impact of IR stimulation at increasing temperatures to the OSL signal of contaminated quartz

Radiation Measurements 68 (2014) 14e22

Contents lists available at ScienceDirect

Radiation Measurements journal homepage: www.elsevier.com/locate/radmeas

Assessing the impact of IR stimulation at increasing temperatures to the OSL signal of contaminated quartz Eren S¸ahiner*, Niyazi Meriç, George S. Polymeris Ankara University, Institute of Nuclear Sciences, Bes¸evler-Ankara, 06100, Turkey

h i g h l i g h t s
The temperature effect of IRSL to the resolved OSL signal of quartz is identified.
IRSL above 50  C stimulates both fast and medium quartz OSL components.
SAR equivalent doses with different stimulation modes, yield similar results.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 April 2014 Received in revised form 15 June 2014 Accepted 17 June 2014 Available online 2 July 2014

The purpose of the present study is to identify the effect of the increasing temperature IR stimulation to the component-resolved OSL luminescence signal of mixed quartz-feldspars material. Post IR OSL signals measured at 110  C were analysed via only general order kinetic terms, while IR signals obtained at increasing temperatures were de-convolved using the sum of general order kinetics plus a tunnelling component. By increasing stimulation temperature, it was demonstrated that IRSL at temperatures above 50  C does not only stimulate feldspar but also stimulates both fast and medium quartz OSL components. In the temperature range between 175 and 250  C, the IRSL initial intensity is dominated by the fast OSL component. Estimated equivalent doses using either Post-IR175.OSL110 as well as IRSL175 (with the indices indicating the measurement temperature) are in good agreement between each other, due to both stimulating quartz. Finally, the physical meaningfulness of the fitting parameters for the tunnelling component is also discussed. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Post IR OSL De-convolution Component resolved analysis Tunnelling Fast OSL component

1. Introduction Optically Stimulated Luminescence (OSL) has been established as a promising dosimetric technique in various fields, such as medical, environmental, personal, space and retrospective dosimetry (McKeever, 2001; Bøtter-Jensen et al., 2003; Yukihara and McKeever, 2011). OSL is based on the fact that naturally-occurring minerals like quartz and feldspar act as natural dosimeters and preserve a record of irradiation dose, i.e. energy per unit mass, received through time. This record (namely the dose) could be obtained by illuminating the sample with visible light inside the laboratory. While applying dating protocols, OSL is usually measured during optical stimulation at steady stimulation power at a specific elevated temperature, resulting in a decaying, continuous wave OSL * Corresponding author. Tel.: þ90 (312) 212 03 84; fax: þ90 (312) 215 33 07. E-mail addresses: [email protected], [email protected] (E. S¸ahiner), [email protected] (N. Meriç), [email protected] (G.S. Polymeris). http://dx.doi.org/10.1016/j.radmeas.2014.06.006 1350-4487/© 2014 Elsevier Ltd. All rights reserved.

(CW-OSL) curve. The age is usually determined by using the initial part of the CW-OSL signal minus a background based on the signal level at the end of the stimulation period. Thus, this net initial signal includes contributions from all fast, slow and medium quartz OSL components, according to the terminology earlier adopted by Smith and Rhodes (1994) and Bailey et al. (1997). Therefore, it is desirable to use a well-separated fast OSL component in equivalent dose routines. It has been shown that for the case of quartz, the signal of each component can be separated from others by analytical or instrumental procedures, such as curve fitting and/or linearly modulated OSL (LM-OSL) (Bulur, 1996; Jain et al., 2005; Polymeris et al., 2006; Kitis and Pagonis, 2008). The components of a CW-OSL from quartz can also be separated by using the temporal duration of each component where not only the intensity but also the duration of each component can be assessed (Chithambo and Galloway, 2001). Nevertheless, in some cases these procedures were proved to be time-consuming and model dependent. In the case of contaminated quartz samples or polymineral materials, the situation becomes even more complicated. Besides

E. S¸ahiner et al. / Radiation Measurements 68 (2014) 14e22

separating the signal of each component, it is also important to separate the signals from the two ubiquitous natural luminescent dosimeters, namely quartz and feldspar, since they record different doses. This difference in estimated dose could be attributed to a number of reasons, such as (i) the internal dose rate, which is present in the case of feldspathic minerals (ii) different sensitivities, (iii) differences in response to irradiation and/or (iv) the residual dose due to incomplete zeroing of the luminescence signal prior to deposition (Ankjærgaard et al., 2010). Nevertheless, the most important reason lies behind the fact that the dose measured using luminescence is usually under-estimated in feldspars because of the well-known anomalous fading phenomenon (Wintle, 1973; Aitken, 1998; Huntley and Lamothe, 2001). The first instrumental approach towards discriminating between the signals arising from quartz and feldspars was suggested by Hütt et al. (1988). These authors indicated the preferential depletion of the signal arising from feldspar by applying Infrared (IR) stimulation at ambient temperatures. Moreover, GodfreySmith et al. (1988) indicated this latter stimulation mode could not stimulate quartz. Another instrumental procedure towards the isolation of fast OSL component was suggested by Jain et al. (2005). These authors have demonstrated that in the temperature range between 120 and 190  C it is possible to preferentially deplete the fast component only using Infrared stimulated luminescence (IRSL) at 880 nm (FWHM 45 nm). Following these results, Polymeris et al. (2008) have demonstrated that IR stimulation at temperatures above 50  C does not deplete only the fast component in most sedimentary quartz samples studied. According to their approach which also included de-convolution, the net depletion of fast and medium components resulting from IR exposure is sampledependent and occurs faster as the stimulation temperature gets higher (Polymeris et al., 2008). Of course, there is always the option for either physical or chemical isolation between these two, ubiquitously present minerals (Wintle, 1997). The standard protocol towards this aim is time consuming, including extended heavy liquid separation as well as pre-treatments with acids; however, in many cases, physical separation of the minerals is not always possible, due to a variety of reasons such as the presence of feldspar micro-inclusions inside quartz grains. Nevertheless, in many cases the chemical treatments affect also the quartz grain quantity, size and properties as well. Consequently, major effort has been devoted in reducing feldspar contamination, by using mostly three different experimental approaches: (a) the instrumental method for isolating a quartz signal from a mixed quartz-feldspar sample based on pulsed optical stimulation (POSL). This approach relies on the fact that the shapes of the time-resolved OSL (TR-OSL) of quartz and feldspar are very different (Chithambo and Galloway, 2000; Denby et al., 2006), (b) an elevated temperature IR stimulation prior to CW-OSL measurement (Jain and Singhvi, 2001). This approach, which was referred to as IR bleaching, has been applied by a number of authors (Wallinga et al., 2002; Kiyak and Erturaç, 2008), despite the fact that may present difficulties because at high temperatures both the fast and the medium components in quartz OSL are depleted significantly by IR exposure (Jain et al., 2003). Finally, Thomsen et al. (2008) and Ankjærgaard et al., 2010 reported the effective application of a combination of post-IR pulsed blue stimulation with previous IRSL stimulation at ambient temperatures and 175  C, respectively. The present work provides an analytical study of the impact of this IR stimulation at elevated temperatures prior to OSL measurement to the fast OSL component of the natural signal, for a wide range of stimulation temperatures. Our approach includes a combination of both instrumental as well as analytical procedures, since all OSL and IRSL decay curves were de-convolved into their

15

individual components. Implications for dating will also be discussed in the framework of the existing models. 2. Materials and method 2.1. Sample origin The sample subjected to the present study was natural quartz of sedimentary origin that was collected from a fault line in KütahyaSimav, the Aegean Anatolia region, Turkey. The preparation of samples was formed under red light conditions. After sieving, grains of dimensions 90e180 mm were obtained. These grains were treated with HCI (10%), H2O2 (35%), HF (40%, 45e60 min of handling) and a final treatment with HCI (10%) in order to obtain a clean quartz extract. Aliquots with mass of 2 mg each were prepared by mounting the material on stainless-steel disks. All aliquots were checked with infrared (IR) stimulation (880 nm) at ambient temperature to ensure the absence of feldspars. The quartz, however, was proven to be contaminated with feldspars, even after the application of the chemical preparation procedure towards extraction a clean quartz extract. This result was yielded by comparison of the IRSL (at RT) and Blue OSL (at 110  C) signals, which are plotted in Fig. 1. As this Fig. reveals, these two signals provide ratio of IRSL over Blue OSL for the corresponding initial intensities of the order of 1.5. Therefore the contribution of the feldspar contamination, even after chemical procedures is considerably adequate when compared to the blue signal. 2.2. Apparatus and measurement conditions All luminescence measurements were carried out using a Risø TL/OSL reader (model TL/OSL-DA-20), equipped with a 90Sr/90Y beta particle source, delivering a nominal dose rate of 0.130 ± 0.004 Gy/s. A 9635QA photomultiplier tube was used for light detection. The stimulation wavelength is 470 (±20) nm for the case of blue stimulation, delivering at the sample position a maximum power of 40 mW cm2. For IRSL, the stimulation wavelength is 875 (±40) nm and the maximum power of ~135 mW cm2 (Bøtter-Jensen et al., 1999a; Bøtter-Jensen et al., 1999b). The detection optics consisted of a 7.5 mm Hoya U-340 filter (lp ~ 340 nm, FWHM ~ 80 nm). All TL measurements were performed in a nitrogen atmosphere with a low constant heating rate of 2  C/s, in order to avoid significant temperature lag; for the case of TL the samples were heated up to the maximum temperature of

Fig. 1. IR check at room temperature after chemical procedures.

16

E. S¸ahiner et al. / Radiation Measurements 68 (2014) 14e22

500  C. Conventional IRSL measurements were performed at various temperatures ranging from RT to 275  C; all blue OSL measurements were performed at 110  C in the continuous wave mode (CW-OSL). 2.3. Experimental protocols The focus of the present work has been concentrated in studying naturally occurring luminescence signal. The experimental protocols applied are presented in tabular form in Table 1. Since the sample subjected to the present study is natural quartz of sedimentary origin, a standard preheat at 260  C was incorporated before all OSL and IRSL measurements. At the beginning, natural TL (NTL) was measured for reference, as well as Blue OSL at 110  C without any previous IRSL measurement, subsequent a standard preheat at 260  C. Each cycle of steps 1e3 and steps 1e4 for protocols 1 and 2 respectively was repeated two times for each IRSL measurement temperature Ti to a fresh aliquot (multiple aliquot procedure). Ti ranged between 100 and 275  C in steps of 25  C, including also IRSL measurements at room temperature (hereafter RT) and 50  C. Protocol 1 is similar to the corresponding protocol applied by Spooner (1994), especially since the measurements were performed in naturally accumulated luminescence signals; nevertheless, instead of blue OSL stimulation, IRSL at various temperatures has been applied. Protocol 2 stands as a modification of protocol 1, incorporating also a post IR, Blue OSL measurement at 110  C. In both protocols, a residual thermoluminescence (rTL hereafter) measurement was received in order to check the bleaching levels of NTL signal. Aliquots with the same mass of 5 mg were attached to stainless steel cups. Each data point reported in this paper was the average of two measurements carried out on two different aliquots. Mass reproducibility was better than 5%. For both cases of IRSL as well as Blue OSL, the stimulation duration incorporated was 100 s; this is the typical stimulation time applied to the majority experimental dating protocols nowadays. 2.4. Method of analysis The Blue OSL curves received at 110  C after IRSL cleaning were de-convolved by using the general order kinetics (GOK hereafter) expression for OSL theory (Bøtter-Jensen et al., 2003; Kitis and Pagonis, 2008):


t b=b1 I ¼ I0 $ 1 þ ðb  1Þ$ ; t

bs1

(1)

where I(t) is the intensity of the luminescence signal as a function of time, t ¼ 1/l (s) is the lifetime and b is the order of kinetics. For the case of general order kinetics the value of kinetic order b was left to vary freely in the range between 1.00001 and 2. In practise, a sum of three components was applied in all cases except the case of the OSL curves received at RT.

Table 1 Tabular form of the experimental protocols applied. Ti attains the according to following values: RT (25), 50, 100, 125, 150, 175, 200, 225, 250, 275  C. Protocol 1 Step 1 Step 2 Step 3 Protocol 2 Step 1 Step 2 Step 3 Step 4

Preheat (260  C, 10 s, 2  C/s) IRSL (Ti  C, 90%, 100 s, 2  C/s) RTL (500  C, 2  C/s) Preheat (260  C, 10 s, 2  C/s) IRSL (Ti  C, 90%, 100 s, 2  C/s) Blue CW-OSL (110  C, 90%, 100 s, 2  C/s) RTL (500  C, 2  C/s)

The anomalous fading yielded for the case of IRSL from feldsparcontaminated quartz is a strong indication that a tunnelling mechanism must be considered as a possible explanation. In this respect, the experimentally obtained IRSL curves of the feldsparcontaminated quartz were fitted with a sum of a tunnelling component plus components of general order kinetics, for the first time in the literature. For the case of the tunnelling component, the following set of equations was applied (Kitis and Pagonis, 2013):

IðtÞ ¼

  3n0 r0 zFðtÞ2 exp  r0 FðtÞ3 t 1 þ zt

(2)


 t F ðt Þ ¼ ln 1 þ z t

(3)

t ¼ ðs$4Þ1

(4)

where n0 the initial concentration of trapped electrons, z a constant, t the lifetime, r’ a dimensionless parameter representing the normalized donoreacceptor density, s is the photo-ionization cross section and 4 the stimulation flux. The abovementioned set of equations was recently suggested by Kitis and Pagonis (2013), who have obtained analytical solutions for the set of differential equations of the model of Jain et al. (2012), by using certain mathematical and physical simplifications. These authors presented a new general kinetic model which quantifies localized electronic recombination of donoreacceptor pairs in luminescent materials. The main physical assumption in the model is the presence of a random distribution of hole traps in the luminescent volume, and an associated range of random nearest-neighbour recombination probabilities. Stimulated recombination takes place only via the excited state of the electron trap, by either optical or thermal stimulation. All curve fittings were performed using the software package Microsoft Excel, with the Solver utility (Afouxenidis et al., 2012) while the goodness of fit was tested using the Figure Of Merit (FOM) of Balian and Eddy (1977).

FOM ¼

  X YExper  YFit  i

A

(5)

where YExper is the experimental glow-curve, YFit is the fitted glowcurve and A is the area of the fitted glow-curve. The FOM values obtained were less than 0.5% in all cases of blue OSL curves, while for the corresponding IRSL curves of the order of 1.5e3%.

3. Results 3.1. Blue OSL de-convolution Fig. 2 presents two typical Blue OSL decay curves, one measured subsequent IRSL measurement at RT (plot A) and one after IRSL at 175  C (plot B), each one being de-convolved to its three individual components, C1, C2 and C3. Residuals of each fitting are also presented in the same Fig., in plots C and D respectively. All Blue OSL decay curves could be perfectly fitted by applying a sum of three individual GOK components. This result is clearly highlighted by all plots of Fig. 2, where both de-convolved OSL as well as fitting residuals are plotted versus stimulation time. Moreover, a great reproducibility of the t values for each one among the three individual components was monitored throughout the IRSL temperature region studied. The mean values along the standard deviation of these values are presented in Table 2. In all cases, the b parameter

E. S¸ahiner et al. / Radiation Measurements 68 (2014) 14e22

17

Fig. 2. Blue OSL decay curves measured after IRSL at RT (plot A) and 175  C (plot B), de-convolved into 3 individual components for each one. Plots C and D present the corresponding fitting residuals.

for the order of kinetic yielded values between 1.03 and 1.07, verifying the first-order kinetics assumption for the case of quartz (Polymeris et al., 2009, 2011). After IRSL at RT, the initial intensity of the Blue OSL decay curve is dominated by the very intense fast OSL component. However, the comparison of the fitting results indicates that after IRSL at 175  C, the Blue OSL signal at the initial stimulation times is not dominated by the fast component, namely the component with the lowest t value. The higher the IRSL stimulation temperature, the less dominant the fast component becomes. Moreover, this is also the case for the integrated signal of each component. Fig. 3 presents the contribution of each component to the Blue OSL signal, in terms of both initial OSL signal (plot A) as well as integrated intensity over the entire duration of stimulation (plot B). For temperatures up to 125  C, the contribution of each component could be considered stable. However, for highest IRSL stimulation temperatures, the most prominent result according to this latter Fig. is related to the decrease of the contribution of the fast component C1 with simultaneous increase of the contribution of the corresponding medium C2 and slow C3 components. The fast component C1 almost to 40e45% of the integrated Blue OSL signal measured after stimulating for 100 s at 110  C after IRSL for the same measurement time at RT. However, this contribution is very fast decreasing with

Table 2 De-convolution parameters of both Blue OSL and IRSL GOK components. Type

Blue OSL GOK components

IRSL GOK components

Parameters

t (s)

b

t (s)

b

C1 C2 C3

1.88 (0.17) 10.98 (1.85) 86.37 (5.85)

1.005 1.007 1.003

1.85 (0.21) 10.66 (1.78) e

1.002 1.005 e

increasing IRSL stimulation temperature, reaching the value of 4% after IRSL at 275  C. As the plot of the contributions of each component to the initial Blue OSL signal reveals, for temperatures higher than 225  C the contribution of the medium Blue OSL component is getting decreased, along with a faint increase of the contribution of the fast Blue OSL. Nevertheless, the contribution of the slow Blue OSL component is monotonically increased throughout the temperature region under study. 3.2. IRSL De-convolution Plot of Fig. 4A presents the IRSL decay curve which was measured at RT, fitted by using the set of Equations (2)e(4) solely. The plot is presented in logarithmic axes in order for the goodness of fitting to be easily assessed. However, as the stimulation temperature of the IRSL is increased from RT and hitherto, the set of Equations (2)e(4) are not adequate enough to provide a good fitting for these IRSL decay curves. Therefore, in this case the experimentally obtained IRSL curves at elevated temperatures were fitted using a sum of a tunnelling component according to Equations (2)e(4) plus GOK components, according to Equation (1). An example of this type of analysis, using a tunnelling plus two GOK components, is given in plot of Fig. 4B, for the case of IRSL stimulation temperature of 175  C. The GOK component required for a good fitting yields t values almost similar to the respective of the fast component C1 that was previously applied for the case of Blue OSL de-convolution. At higher temperatures though, there is also a contribution of the medium OSL component as well. Moreover, the b parameters for the order of kinetic yielded values between 1.03 and 1.05. It should be emphasized that none of the IRSL measurements could be de-convolved by using a sum of three individual GOK components only. As Fig. 4 reveals, at the IRSL decay curve at 175  C, the contributions of the fast OSL component C1 and that of the tunnelling

18

E. S¸ahiner et al. / Radiation Measurements 68 (2014) 14e22

Fig. 3. The contribution of each one among the fast, medium and slow Blue OSL components to the initial Blue OSL signal (plot A) as well as to the integrated Blue OSL signal over 100 s of stimulation (plot B). Each contribution was added by dividing the corresponding component's signal intensity over the total luminescence intensity.

component Ctun are comparable. Moreover, the initial intensity of the IRSL curves is also dominated by the fast OSL component C1. Fig. 5A presents the dependence of contributions of the fast Blue OSL component to the IRSL signal, in both terms of integrated signal (filled data points) as well as initial signal (open data points) as

yielded by the de-convolution procedure. The corresponding contribution of the medium Blue OSL component to the IRSL signal is presented in plot 5B. For the case of the fast Blue OSL component, a plateau is formed in the temperature range between 200 and 275  C, indicating that almost 45% of the initial signal and 20% of the integrated signal arises from the fast OSL of quartz. For the case of the medium Blue OSL component, a plateau is formed in the temperature range between 150 and 250  C, indicating that almost 35% of the initial signal and 40% of the integrated signal arises from the fast OSL of quartz. Within these temperature regions, a faint contribution of the slow component is also monitored; however, this is only the case of the integrated luminescence signal. Within this localized transition model (as well as in delocalized TL models), the lifetime t in Eq. (4) depends very strongly on the stimulation intensity. Since all IRSL measurements were performed at stable stimulation intensity, the lifetime is expected to be stable. Fig. 5C presents the dependence of both the lifetime t and the parameter r’ on the stimulation temperature. As this Fig. reveals, both parameters attain stable values for stimulation temperatures up to 125  C. Above this stimulation temperature, none of these two parameters was constant, with the value of the parameter r’ being further increased with the stimulation temperature, while the lifetime yielded a decreasing trend with IRSL stimulation intensity. 3.3. Residual TL

Fig. 4. Fitting examples of the IRSL decay curves measured at RT using only a tunnelling component according to Equations (2)e(4) (plot A) as well as at 175  C, using one tunnelling plus two GOK components.

Fig. 6 presents typical residual TL (rTL) glow curves received after applying IRSL at various temperatures and subsequently Blue OSL (step 4 of protocol 2, plot A of Fig. 6) as well as after applying solely IRSL at various temperatures (step 3 of protocol 1, plot B of Fig. 6). As the stimulation temperature is increased, the residual level of TL is further decreased for both protocols. Moreover, the residual TL level after the combined action of IRSL and Blue OSL is lower compared to the one after just IRSL. Fig. 6A yields a prominent decrease for the 325  C TL peak of quartz, indicating its effective bleaching under the combined action of IRSL and Blue OSL. According to Spooner (1994) the OSL signal is directly related with the electron trap, which is responsible for the TL glow-peak at 325  C (depending upon the heating rate). Therefore, the bleaching of this specific peak is expected to be caused by Blue OSL. The TL glow curves recorded following IRSL and Blue OSL stimulation show a photo-transferred TL peak in the temperature range 200e250  C. However, as Fig. 6B indicates, the bleaching of TL glow-peak at 325  C also takes place after solely applying IRSL at temperatures starting from 125  C and upwards. IRSL bleaches the entire glow curve, including the TL 325  C as well. Fig. 6B indicates

E. S¸ahiner et al. / Radiation Measurements 68 (2014) 14e22

that as the stimulation temperature is increased for IRSL, the residual TL indicates a shift in both (a) Tmax as well as (b) initial, rising part of the TL glow curve. This experimental feature stands in good agreement with the main physical assumption in the model, dealing with the presence of a random distribution of hole traps of the same type in the luminescent volume, and an associated range of random nearestneighbour recombination probabilities, verifying thus the use of

19

the set of equations (2)e(4) for the case of IRSL curves. It should be emphasized that a change in the type of hole-traps will change the emission wavelength. However, this is not the case, since only the change in the concentration of hole-traps affects the intensity of luminescence, resulting thus to the monitored shifting. 4. Discussion The experimental results presented in the previous sections clearly demonstrate that IR stimulation at temperatures above 50  C does not deplete only the signal resulting from feldspars. Net depletion of fast and medium component of the Blue OSL signal resulting from quartz also occurs with IR exposure and occurs faster as the stimulation temperature gets values of 125  C and upper. The result is strongly supported by the following experimental features: (a) The stimulation of the entire glow curve of the NTL signal after applying only IR stimulation, (b) The increase of the contribution of the fast Blue OSL component to either net/initial IRSL signal with increasing IRSL stimulation temperature, (c) The decrease of the contribution of the fast Blue OSL component to either net/initial Blue OSL signal with increasing IRSL stimulation temperature, (d) The fitting analysis, which yields that the initial IRSL signal, which is widely used in analysis after routine OSL dating protocols, is dominated by a GOK component corresponding to the fast OSL component of the Blue OSL signal.

Fig. 5. The results of the fitting analysis on the IRSL decay curves according to the set of Equations (1)e(4). Plot A presents the contribution of the fast Blue OSL component to the IRSL signal, while plot B presents the corresponding contribution of the medium Blue OSL component to the IRSL signal. Open data points correspond to initial stimulation intensity while filled data points to contributions to the integrated signal over 100 stimulation seconds. Plot C represents the temperature dependence of the parameters t and r’ of Equations (2)e(4).

Another experimental verification of the aforementioned result is revealed by Fig. 7, where both IRSL and Blue OSL curves were normalised over their initial intensity at time t ¼ 0 and presented in plots 7A and 7B respectively. As it becomes obvious from Fig. 7B, the Blue OSL curves change their shape with increasing IRSL stimulation temperature. This change in shape would have been expected if each Blue OSL curve were measured in different temperatures. However, all Blue OSL measurements were measured at 110  C. Therefore, this change of shape is attributed to the differential contribution of each component, with the slow component contributing more as the IRSL stimulation temperature is increased. In terms, this experimental feature results from the depletion of fast Blue OSL component to pre Blue OSL IRSL curves. This result stands in good agreement with the suggestion of Polymeris et al. (2008), who have indicated that IRSL at ambient temperatures stimulates only the feldspars, while at higher temperatures it also stimulates quartz. Moreover, these authors indicated that with increasing temperatures, the fast Blue OSL component is depleted as faster as higher the stimulation temperature gets. However, at the same time, the intense depletion of the medium Blue OSL component contradicts the aforementioned citation, which reports that slight depletion of the medium component takes place for stimulation temperatures above 150  C. Moreover, from a quantitative point of view, net percentage depletion of the fast component solely agree with the values reported by both Polymeris et al. (2008) as well as Jain et al. (2005). These latter authors reported that IR stimulation for 1500 s (115 mW cm2) resulted in 10% depletion of the integrated intensity of the fast component at 160  C and 22% at 190  C, while no depletion of either the medium or the slow components was detected. Nevertheless, Fig. 7A indicates that normalised IRSL curves almost coincide, independent on the stimulation temperatures, similarly to the case of prompt isothermal decay (PID) of Durango Apatite at various temperatures (Sfampa et al., 2014). Moreover, according to Fig. 5B, stability is yielded for both the lifetime t of the

20

E. S¸ahiner et al. / Radiation Measurements 68 (2014) 14e22

Fig. 6. Residual TL glow curves after both IRSL and subsequent Blue OSL (plot A) and after solely IRSL (plot B).

tunnelling component as well as the corresponding r’ parameter for stimulation temperatures up to 125  C. This stability could be considered as an experimental argument towards a continuum of energy states being available near the excited state. The possibility of such a continuum of states within the energy gap in feldspars has been considered in detail in the work by Poolton et al. (2002a,b) and in the recent extensive work by Jain and Ankjrgaard (2011). These authors proposed that all feldspar IRSL signals are derived from a range of electronehole recombination routes that relate to electrons from the same electron trap. In the present work, the shape of the rTL glow curves, especially at the low temperature part, after solely IRSL stimulation, strongly supports the possibility of such a continuum. According to the model of Jain et al. (2012), the value of the lifetime yields a strong dependence on the trapcenter distances. As discussed in the theoretical work by Poolton et al. (2002a,b), one can expect significant disorder in the structure of these materials, due to the presence of a continuum distribution of depths and widths in adjacent quantum wells. Such a distribution of distances and energies could in principle correspond to a continuum states within the energy gap, which in terms could be attributed to enhanced phonon vibration modes as the temperature of stimulation is gradually increased. However, in order to explain the decrease of the yielded lifetime for larger stimulation temperatures, the right part of Equation (4) should also include one term describing the thermal contribution at elevated temperatures, namely a term of the form:

ttherm ¼ s1 eE=kTIRSL

(6)

From a physical point of view, the stimulation is neither purely thermal nor optical, but the sum of those two, i.e.:

ptot ¼ popt þ ptherm

(7)

where pi stands for the corresponding probability to free an electron. Consequently:

1 1 1 ¼ þ ttot topt ttherm

(8)

where ttot is the lifetime estimated according to the fitting procedure, topt the term described by Equation (4) and ttherm the term of equation (6). For low temperatures the term ptherm is negligible compared to the term popt, therefore the term 1/topt dominates. However, since the IRSL takes place at steady stimulation intensity, the lifetime is constant. For high IRSL stimulation temperatures, the term ptherm becomes dominant, therefore the ttot decreases gradually due to increase of ptherm and corresponding decrease of ttherm.

The increase of the parameter r’ on the further increase of the stimulation temperature is consistent with the decrease of the lifetime. Since high temperature IRSL stimulation also depletes the 325  C of quartz, it seems that after 125e150  C, the contribution from the quartz TL trap becomes substantial. Therefore, the enhanced number of electrons results in increasing the effective donoreacceptor distances and in turn to an increase of the tunnelling recombination probability, a fact which is mathematically expressed in the framework of the model as an increased value of the parameter r’. To be more precise, increasing the number of electrons does not directly increase the donoreacceptor distances. However, it decreases the holes in the more closely spaced donor acceptor distances, resulting in a residual of available holes in larger distances.

5. Implications on dating As Figs. 3 and 5 are going to further reveal, for IRSL stimulation at 175  C, almost 50% of the initial fast Blue OSL signal is depleted during IRSL measurements while the other 50% while stimulating using Blue light. Therefore, it becomes obvious that these two stimulation modes would probably yield similar equivalent dose results for the case of mixed samples. In order to study this possible implication of the IRSL cleaning at elevated temperatures to the signal of quartz, the following equivalent dose estimation protocols were applied to the sample of the present work: 1. SAR protocol using IRRT (both un-corrected and corrected for anomalous fading) (Aitken, 1998) 2. SAR protocol using IR175 3. SAR protocol using Post-IR175.OSL110 (Wallinga et al., 2002) 4. SAR protocol using Post-IRRT.OSL110 (Banerjee et al., 2001) It should be highlighted that even though protocols 1, 3 and 4 have been extensively used in the literature, protocol 2 is applied for the first time in the literature. Other approaches in the related literature reported the use of the Post-IR175OSL110 solely, but not the IRSL175 (Wallinga et al., 2002; Kiyak and Erturaç, 2008). Results are presented at Table 3. Indices indicate the temperature of the corresponding measurements. As seen at Table 3, all stimulation modes give similar equivalent doses within errors. These results are very much important since they provide an alternative experimental way for measuring fast and partly medium OSL components while stimulating by using IRSL. Moreover, negligible anomalous fading was detected for case of the IRSL signal measured at 175  C. In other words, chemical separation might be not so much important, if similar results could

E. S¸ahiner et al. / Radiation Measurements 68 (2014) 14e22

Fig. 7. Blue OSL decay curves (plot A) and IRSL decay curves (plot B), all normalized over the initial intensity of each curve.

Table 3 Equivalent doses estimated according to SAR with different stimulation modes. Stimulation mode

Equivalent dose (Gy)

IRSLRT a IRSLRT Post-IRRT.OSL110 IRSL175 Post-IR175.OSL110

98.8 111.1 109.3 114.4 107.5

± ± ± ± ±

4.2 9.2 7.5 3.6 7.7

a Corrected by g-value estimated as1.27 ± 0.13 according to Huntley and Lamothe (2001).

be obtain for a large number of either contaminated quartz or polymineral/mixed samples. Further work is required in order to establish whether this high temperature IR bleaching property is prevalent or sample dependent. 6. Concluding remarks Based on a series of experimental evidence, it is clearly demonstrated that IR stimulation at temperatures above 50  C does not only stimulate feldspar but also stimulates quartz. These evidence include the shape of the glow curve of the NTL signal after applying only IR stimulation, the increase of the contribution of the fast Blue OSL component to either net/initial IRSL signal with increasing IRSL stimulation temperature as well as the decrease of the contribution of the fast Blue OSL component to either net/initial Blue OSL signal with increasing IRSL stimulation temperature. The latter two results were indicated according to the fitting analysis, which yields that the initial IRSL signal, which is widely used in routine OSL dating protocols, is dominated by a GOK component corresponding to the fast OSL component of the Blue OSL signal. Estimated equivalent doses using either Post-IR175$OSL110 as well as IRSL175 are in good agreement because due to both stimulating quartz. Finally, a slight modification is suggested for the equations of Kitis and Pagonis (2013), who have obtained analytical solutions for the set of differential equations of the model of Jain et al. (2012). References Afouxenidis, D., Polymeris, G.S., Tsirliganis, N.C., Kitis, G., 2012. Computerised curve deconvolution of TL/OSL curves using a popular spreadsheet program. Radiat. Prot. Dosim. 149 (3), 363e370.

21

Aitken, M.J., 1998. An Introduction to Optical Dating. Oxford University Press, Oxford. Ankjærgaard, C., Jain, M., Thomsen, K.J., Murray, A.S., 2010. Optimising the separation of quartz and feldspar optically stimulated luminescence using pulsed excitation. Radiat. Meas. 45, 778e785. Bailey, R.M., Smith, B.W., Rhodes, E.J., 1997. Partial bleaching and the decay form characteristics of quartz OSL. Radiat. Meas. 27, 123e136. 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. Nucl. Instrum. Meth. 145, 389e395. Banerjee, D., Murray, A.S., Bøtter - Jensen, L., Lang, A., 2001. Equivalent dose estimation using a single aliquot of polymineral fine grains. Radiat. Meas. 33, 73e94. Bøtter - Jensen, L., McKeever, S.W.S., Wintle, A.G., 2003. Optically Stimulated Luminescence Dosimetry. Elsevier, Amsterdam. Bøtter-Jensen, L., Duller, G.A.T., Murray, A.S., Banerjee, D., 1999a. Blue light emitting diodes for optical stimulation of quartz in retrospective dosimetry and dating. Radiat. Prot. Dosim. 84, 335e340. Bøtter-Jensen, L., Mejdahl, V., Murray, A.S., 1999b. New light on OSL. Quat. Sci. Rev. 18 (2), 303e309. Bulur, E., 1996. An alternative technique for optically stimulated luminescence (OSL) experiment. Radiat. Meas. 26, 701e709. Chithambo, M.L., Galloway, R.B., 2000. A pulsed light-emitting-diode system for stimulation of luminescence. Meas. Sci. Technol. 11, 418e424. Chithambo, M.L., Galloway, R.B., 2001. On the slow component luminescence stimulated from quartz by pulsed light-emitting-diodes. Nucl. Instrum. Meth. B 183, 358e368. Denby, P.M., Bøtter-Jensen, L., Murray, A.S., Thomsen, K.J., Moska, P., 2006. Application of pulsed OSL to the separation of the luminescence components from a mixed quartz/feldspar sample. Radiat. Meas. 41, 774e779. Godfrey-Smith, D.I., Huntley, D.J., Chen, W.H., 1988. Optical dating studies of quartz and feldspar sediment extracts. Quat. Sci. Rev. 7, 373e380. Huntley, D.J., Lamothe, M., 2001. Ubiquity of anomalous fading in K-feldspar and the measurement and correction for it in optical dating. Can. J. Earth Sci. 38, 1093e1106. Hütt, G., Jaek, I., Tchonla, J., 1988. Optical Dating: K-feldspars optical response stimulation spectra. Quat. Sci. Rev. 7 (3e4), 381e386. Jain, M., Murray, A.S., Bøtter-Jensen, L., 2003. Characterisation of blue light stimulated luminescence components in different quartz samples: implications for dose measurement. Radiat. Meas. 37 (4e5), 441e449. Jain, M., Murray, A.S., Bøtter-Jensen, L., Wintle, A.G., 2005. A single aliquot regenerative-dose method based on IR (1.49 eV) bleaching of the fast OSL component in quartz. Radiat. Meas. 39 (3), 309e318. Jain, M., Ankjrgaard, C., 2011. Towards a non-fading signal in feldspar: insight into charge transport and tunnelling from time-resolved optically stimulated luminescence. Radiat. Meas. 46, 292e309. Jain, M., Guralnik, B., Andersen, M.T., 2012. Stimulated luminescence emission from localized recombination in randomly distributed defects. J. Phys. Condens. Matter 24, 385e402. Jain, M., Singhvi, A.K., 2001. Limits to depletion of blue-green light stimulated luminescence in feldspars: implications for quartz dating. Radiat. Meas. 33, 883e892. Kitis, G., Pagonis, V., 2008. Computerized curve deconvolution analysis for LM-OSL. Rad. Meas. 43, 737e741. Kitis, G., Pagonis, V., 2013. Analytical solutions for stimulated luminescence emission from tunneling recombination in random distributions of defects. J. Lumin 137, 109e115. Kiyak, N.G., Erturaç, M.K., 2008. Luminescence ages of feldspar contaminated quartz from fluvial terrace sediments. Geochronometria 30, 55e60. McKeever, S.W.S., 2001. Optically stimulated luminescence dosimetry. Nucl. Instrum. Methods Phys. Res. B 184, 29e54. Polymeris, G., Kiyak, N.G., Kitis, G., 2008. Component resolved IR bleaching study of the Blue LM-OSL signal of various quartz samples. Geochronometria 32, 79e85. Polymeris, G.S., Afouxenidis, D., Raptis, S., Liritzis, I., Tsirliganis, N.C., Kitis, G., 2011. Relative response of TL and component-resolved OSL to alpha and beta radiations in annealed sedimentary quartz. Radiat. Meas. 46, 1055e1064. Polymeris, G.S., Afouxenidis, D., Tsirliganis, N.C., Kitis, G., 2009. The TL and room temperature OSL properties of the glow peak at 110  C in natural milky quartz: a case study. Radiat. Meas. 44, 23e31. Polymeris, G.S., Tsirliganis, N.C., Loukou, Z., Kitis, G., 2006. A comparative study of the anomalous fading effects of TL and OSL signals of Durango apatite. Phys. Status Solidi. A 203 (3), 578e590. Poolton, N.R.J., Ozanyan, K.B., Wallinga, J., Murray, A.S., Bøtter-Jensen, L., 2002b. Electrons in Feldspar II: a consideration of the influence of conduction band-tail states on luminescence processes. Phys. Chem. Min. 29, 217e225. Poolton, N.R.J., Wallinga, J., Murray, A.S., Bulur, E., Bøtter-Jensen, L., 2002a. Electrons in feldspar I: on the wavefunction of electrons trapped at simple lattice defects. Phys. Chem. Min. 29, 210e216. Sfampa, I.K., Polymeris, G.S., Tsirliganis, N.C., Pagonis, V., Kitis, G., 2014. Prompt isothermal decay of thermoluminescence in an apatite exhibiting strong anomalous fading. Nucl. Instrum. Methods Phys. Res. B 320, 57e63. Smith, B.W., Rhodes, E.J., 1994. Charge movements in quartz and their relevance to optical dating. Radiat. Meas. 23, 329e333. Spooner, N.A., 1994. On the optical dating signal from quartz. Radiat. Meas. 23, 593e600.

22

E. S¸ahiner et al. / Radiation Measurements 68 (2014) 14e22

Thomsen, K.J., Jain, M., Murray, A.S., Denby, P.M., Roy, N., Bøtter-Jensen, L., 2008. Minimizing feldspar OSL contamination in quartz UV-OSL using pulsed blue stimulation. Radiat. Meas. 43, 752e757. Wallinga, J., Murray, A.S., Bøtter-Jensen, L., 2002. Measurement of the dose in quartz in the prescence of feldspar contamination. Radiat. Prot. Dosim. 101, 367e370.

Wintle, A.G., 1973. Anomalous fading of thermo luminescence in mineral samples. Nature 245, 143e144. Wintle, A.G., 1997. Luminescence dating: laboratory procedures and protocols. Radiat. Meas. 27, 769e817. Yukihara, E., McKeever, S.W.S., 2011. Optically Stimulated Luminescence: Fundamentals and Applications. John Wiley & Sons Ltd, UK.