Natural variations in the properties of TL and IRSL emissions from metamorphic and volcanic K-feldspars from East Africa: Assessing their reliability for dating

The elevated temperature infrared stimulated luminescence (IRSL) and post-IR IRSL signals of potassium (K)-feldspars have recently garnered attention for their minimal rates of anomalous fading. The post-IR IRSL signal has been used to obtain age estimates for geological deposits, mostly in Europe. Studies on the behaviour of the IRSL and post-IR IRSL signals of K-feldspars from a wider range of geographic regions and depositional contexts are needed, particularly for regions where the OSL signal from quartz is poorly behaved. Discrepancies in the literature regarding the behaviours of the IRSL and TL signals of K-feldspars also highlight the need to characterise the behaviours of samples from a wide variety of contexts. This paper begins to address this problem by characterising and comparing the IRSL signals of a metamorphic and a volcanic K-feldspar sample from two sites in East Africa, a region in which the OSL signal from quartz has generally proven problematic for dating. We demonstrate that the metamorphic and volcanic K-feldspars have substantially different TL glow curves that respond differently to IR stimulation. The sample of metamorphic K-feldspar from Tanzania (MR9) has a peak at 430 °C that is associated with the IRSL signal and an optically less-sensitive peak at 350 °C, while the sample of volcanic K-feldspar from Ethiopia (MB3) exhibits a single broad TL region centred at ～230 °C that responds differently to IR stimulation. Differences in the change of IRSL decay curve shape with stimulation temperature suggest that the processes of IRSL production many vary between the two samples. Using dose recovery tests, we demonstrate that the IRSL (50 °C), IRSL (225 °C) and post-IR IRSL (50 °C, 225 °C) signals of sample MR9 are suitable for dose and age estimation using the single-aliquot regenerative-dose procedure, while those of sample MB3 are less suitable. The post-IR IRSL signal of the latter sample performs poorly in tests of SAR suitability and the three signals exhibit extremely high fading rates over laboratory timescales (g_2days > 19%/decade).     

Assessment of diagnostic tests for evaluating the reliability of SAR De values from polymineral and quartz fine grains

In this study, we applied optically stimulated luminescence (OSL) dating to two fine grain sediment samples collected at Jeongokri, Korea. A single aliquot regenerative dose (SAR) procedure was applied to both polymineral grains and to chemically isolated (H₂SiF₆) quartz grains of 4–11 μm diameter. For polymineral fine grains, the OSL IR depletion ratio and the equivalent dose (D_e) plateau test appear to be equally sensitive indicators of appropriate IR stimulation time for use in the ‘double SAR’ protocol. Additionally, the OSL IR depletion ratio test gives an indication of the relative mineral composition of the samples, hence providing an assessment of the likelihood of obtaining a quartz-dominated [post-IR] OSL signal. Use of higher preheat temperatures would assist in thermally eroding the non-quartz component of the [post-IR] OSL signal from polyminerals. For the quartz fine grains, data from both natural D_e determinations and laboratory dose recovery tests are required to identify the appropriate preheat temperatures for dating, due to problems of thermal transfer. This phenomenon is particularly exaggerated for these samples due to the large D_e values (≥350 Gy) and hence low slope of the dose–response curve. The double SAR method cannot be applied ubiquitously, even after careful and rigorous study of one sample from a section. Quartz OSL dating using a range of preheat temperatures is suggested to be the most suitable method for OSL dating of fine grain sediments.     

Quartz from southern Africa: sensitivity changes as a result of thermal pretreatment

Sensitivity changes are known to occur in quartz during luminescence measurement procedures. Accurate single-aliquot equivalent dose (D_e) evaluations can be made only if the experimental procedure applied causes no sensitivity change, or can correct for it. The quartz single-aliquot regenerative-dose (SAR) protocol developed by Murray and Wintle (Murray, A.S., Wintle, A.G., 2000. Luminescence dating of quartz using an improved single-aliquot regenerative-close protocol. Radiation Measurements 32, 57–73) incorporates a sensitivity correction procedure that was developed using measurements made on a single sample from Australia. This paper presents experimental data designed to test the applicability of this correction procedure to quartz extracted from two southern African sediments. The results indicate that accurate corrections are made using this protocol irrespective of the direction and magnitude of the sensitivity change. The implications for routine D_e evaluation are discussed.     

Paper I: The use of measurement-time dependent single-aliquot equivalent-dose estimates from quartz in the identification of incomplete signal resetting

The identification of incomplete signal re-setting of optically stimulated luminescence signals in sedimentary quartz is a vitally important step in the continued improvement of optical dating. It is shown that narrow spectrum (blue-green) laboratory partial bleaching of aliquots of natural sedimentary quartz has a significant effect on equivalent dose (measured using a single aliquot procedure) calculated as a function of measurement time (D_e(t)). The blue-green stimulation spectrum mimics that found underwater and the results suggest that incomplete re-setting of waterlain sediments may be possible using the D_e(t) method.     

Comparison of the properties of various optically stimulated luminescence signals from potassium feldspar

Various optically stimulated luminescence signals from K-feldspar have been used to determine the equivalent doses of sediment samples. Understanding the properties of these optical signals is critical to evaluate their applicability and limitations to optical dating. In this paper, some properties of IRSL, post-IR OSL and post-IR IRSL signals (detected in the UV region using U-340 filters) from a museum sample of K-feldspar were investigated by analyzing the relationships between optical and TL signals, and the effect of optical bleaching and heating on optical signals. The trap parameters of the different optical signals were calculated using the pulse annealing method. The results show that this sample exhibits two regenerated TL peaks at ～140 and ～330 °C. Corresponding to the low temperature TL peak, the OSL and post-IR OSL signals appear to be more associated with lower temperature TL than the IRSL signal measured at 50 °C. Corresponding to the high temperature TL peak, the post-IR IRSL signals mainly originate from the more thermally stable traps associated with the high temperature TL, compared with the IRSL and post-IR OSL signals. However, the post-IR IRSL_225 °C signal is shown to be hard to be bleached by blue light and simulated sunlight, compared with the IRSL_50 °C and low temperature post-IR IRSL signals. The implication for optical dating is that the elevated temperature post-IR IRSL signals can be preferentially applied over other signals from K-feldspar, but it is desirable that the effectiveness of the pre-depositional zeroing of these signals is assessed.     

Effects of thermally transferred signals in the post-IR IRSL SAR protocol

The recently developed post-IR IRSL SAR protocol is promising to isolate more stable IRSL signals of feldspars. However, the high temperature thermal treatments used will inevitably induce thermally transferred post-IR IRSL₂₉₅ (TT-post-IR IRSL₂₉₅) signal, which would contribute to the measured post-IR IRSL₂₉₅ signal of the test dose and may lead to inaccurate sensitivity correction. In this study, the effects of TT-post-IR IRSL₂₉₅ signal in the post-IR IRSL₂₉₅ SAR protocol are investigated using medium polyminerals from a loess section at Caoxian, northwestern Loess Plateau in China. The sensitivity changes of the IRSL₅₀ and the post-IR IRSL₂₉₅ signals are different from natural cycle to regenerative cycles, which can be attributed to the interference of the TT-post-IR IRSL₂₉₅ signal. Such difference is dependent on the test dose and the post-IR IR stimulation time. The TT-post-IR IRSL₂₉₅ signal is also shown to affect the post-IR IRSL₂₉₅ D_o and D_e values and to cause overestimation of the fading rates. Our study therefore highlights the need of serious consideration on the effects of TT-post-IR IRSL signal when the post-IR IRSL SAR protocol is employed for dating.     

Study of the relationship between infrared stimulated luminescence and blue light stimulated luminescence for potassium-feldspar from sediments

In luminescence measurements of potassium-feldspar (K-feldspar), both infrared (IR) and blue light (BL) can be used as stimulation sources. Component analysis suggests that the blue light stimulated luminescence (BLSL) measured at 60 °C from K-feldspar can be fitted using three components, namely fast, medium and slow. In order to explore the relationship between the origin of the infrared stimulated luminescence (IRSL) signal and the different components of the BLSL, five sets of experiments were conducted, namely post-IR BLSL (pIR-BLSL), post-BL IRSL (pBL-IRSL), pulse annealing tests, dose response and laboratory fading rate tests. It is observed that most of the IRSL signal can be bleached by BL, while the BLSL signal can only be partially bleached by the IR. The sources for IRSL are mainly associated with the fast and medium components of the BLSL signal.     

Observations of thermal transfer OSL signals in glacigenic quartz

The conclusions of Rhodes and Pownall (Rhodes, E.J., Pownall, L., 1994. Zeroing of the OSL signal in quartz from young glaciofluvial sediments. Radiation Measurements 23, 329–333) were somewhat discouraging for the prospects of using quartz OSL for dating glacigenic sediments, while the more detailed study of Rhodes and Bailey (Rhodes, E.J., Bailey, R.M., 1997. The effect of thermal transfer on the zeroing of the luminescence of quartz from recent glaciofluvial sediments. Quaternary Science Reviews (Quaternary Geochronology) 16, 291–298) provided more encouragement. Specifically, the latter authors were able to account for the relatively high D_e values observed for recent glacigenic (dominantly glaciofluvial) sediments in terms of an anomalously high thermal transfer effect, rather than simply insufficient bleaching prior to deposition. In other locations, the OSL of quartz from glacigenic material appears to provide reliable age estimates, and does not suffer from these effects (Owen, L.A., Richards, B., Rhodes, E.J., Cunningham, W.D., Windley, B.F., Badamgarav, J., Dorjnamjaa, D., 1998. Relic permofrost structures in the Gobi of Mongolia: age and significance. Journal of Quaternary Science 13 (16), 539–548; Richards, B.W., Owen, L.A., Rhodes, E.J., 2000. Timing of Late Quaternary glaciations in the Himalayas of northern Pakistan. Journal of Quaternary Science 15, 283–297). In this paper, laboratory bleached samples from both the above studies are remeasured, using an experimental design to correct for OSL sensitivity changes, which has some similarities to the single aliquot regenerative dose (SAR) protocol of Murray and Wintle (Murray A.S., Wintle A.G., 2000. Luminescence dating of quartz using an improved single-aliquot regenerative-dose protocol. Radiation Measurements 32, 57–73). Clear evidence of thermal transfer OSL signals, generated during the preheating procedure, is presented. Further measurements, using the full SAR protocol, demonstrate that the source of this charge is related to natural dosing. For one sample, the apparent effect is subtracted effectively by the SAR protocol. Another sample studied in detail shows a striking relationship between the thermally transferred OSL signal and the total TL observed during the ramping of the preceding preheat treatment.     

Investigations of the post-IR IRSL protocol applied to single K-feldspar grains from fluvial sediment samples

The post-IR IRSL protocol with single K-feldspar grains was applied to three samples taken from a fluvial sedimentary sequence at the archaeological site of the Dali Man, Shaanxi Province, China. K-feldspar coarse grains were extracted for measurement. Approximately 30–40% of the grains were sufficiently bright to measure, and after application of rejection criteria based on signal strength, recuperation, recycling ratio and saturation dose, ～10–15% of the grains were used for D_e calculation. The relationship of signal decay rate and form of D_e(t) with the recovery dose were investigated. The dose recovery ratios of the samples after initial bleaching with the four different light sources were within uncertainties of unity. No anomalous fading was observed. The over-dispersion of the recovered dose and D_e values were similar, suggesting neither incomplete resetting of the post-IR IRSL signals nor spatially heterogeneous dose rates significantly affected the natural dose estimates. The values of D_e obtained with the single K-feldspar grain post-IR IRSL protocol were in the range ～400–490 Gy. Combining all of the measured single-grain signals for each of the individual samples (into a ‘synthetic single aliquot’) increased the D_e estimates to the range ～700–900 Gy, suggesting that the grains screened-out by the rejection criteria may have the potential to cause palaeodose over-estimation, although this finding requires a more extensive investigation. Thermally transferred signals were found in the single K-feldspar grains post-IR IRSL protocol, and the proportion of thermally transferred signal to test-dose OSL signal (stimulation at 290 °C) from the natural dose was higher than from regenerative doses, and the proportion was grain- and dose-dependent. As such, TT-post-IR IRSL signals at 290 °C have the potential to cause dose underestimation, although this may be reduced by using larger test-dose irradiations. Our study demonstrates considerable potential in the post-IR IRSL method in providing chronological control in studies relevant to human evolution in the later-Pleistocene.     

Paper II: The interpretation of measurement-time-dependent single-aliquot equivalent-dose estimates using predictions from a simple empirical model

Narrow spectrum (blue-green) laboratory partial bleaching of aliquots of natural sedimentary quartz has been shown to have a significant effect on equivalent dose (measured using a single aliquot procedure) calculated as a function of measurement time (D_e(t)). A simple model, based on empirical data, is presented in which the effects of partial bleaching on D_e(t) are predicted, taking into account the influences of pre- and post-burial doses. The model is applied to the case of heterogeneous populations of partially bleached single grains with various ranges of residual doses. Modelling shows that under realistic conditions, D_e histogram plots are unable to distinguish between samples having age overestimates and those with correct ages, whereas the proposed D_e–Z plot is able to make this distinction. Furthermore, modelling shows that D_e–Z plots can identify sub-populations of grains/aliquots that show most evidence of full bleaching, which can then be used to estimate a correct mean D_e value.     

Optically stimulated luminescence production in the single-aliquot regenerative dose protocol

When using a single-aliquot regenerative dose (SAR) protocol for luminescence dating of sedimentary quartz grains, the fundamental assumption is that the sensitivity of the optically stimulated luminescence (OSL) produced by the regenerative doses can be monitored by their following test dose OSL response. Using well-bleached coarse quartz grains, OSL production in a SAR protocol was studied in detail when dose response curves were constructed using both single and multiple aliquot regenerative dose procedures. During application of the SAR protocol, two preheats are applied, each ahead of an OSL measurement that is produced by the regenerative dose and test dose, respectively. It is shown that sensitivity changes caused by heating were well corrected for using the OSL response to the test dose. However, these preheats are shown to result in thermally stimulated OSL signals that contribute both to the OSL response from the regenerative doses used to construct the dose response curve and to the OSL response from the test dose used to monitor OSL sensitivity changes as the quartz grains are repeatedly measured. A simple test is proposed to identify the contribution of the latter signal.     

Yellow stimulated luminescence from potassium feldspar: Observations on its suitability for dating

Yellow stimulated luminescence (Y-OSL) is the light detected from potassium-rich feldspars at 410 nm under stimulation by a yellow light source emitting 590 nm. The investigation of this study aimed at understanding basic luminescence physics of Y-OSL in order to assess the suitability of the technique for dating. The Y-OSL signal properties tested were signal intensity, thermal assistance, thermal stability, sensitivity to daylight and the suitability of a single aliquot regenerative (SAR) protocol to be employed for equivalent dose (D_e) estimation. D_e measurements were conducted on samples of Holocene, last glacial and Tertiary age. The tests were undertaken on sedimentary feldspar separates extracted from aeolian, fluvial and coastal deposits.Results from experiments show that the signal intensity increases by measuring Y-OSL at elevated temperature suggesting thermal assistance characteristics similar to infrared stimulated luminescence (IRSL). The yellow stimulated signal remains unaffected by preheat temperatures up to ～200 °C suggesting higher thermal stability than the IRSL signal. The Y-OSL signal is less light sensitive than the IRSL signal and D_e residuals obtained from modern samples are up to 7 Gy indicating suitability of the technique for ‘older’ and well-bleached sediments. The dose recovery tests successfully recovered the given dose if the specific light sensitivity of Y-OSL is taken into account. For every sample Y-OSL D_e values obtained by a single aliquot regenerative dose protocol (SAR) are higher than those obtained by an IRSL SAR approach. From these results we infer high thermal stability and a minimal anomalous fading of the Y-OSL signal. We conclude that Y-OSL has a high potential to date Quaternary sediments that were sufficiently bleached in nature.     

Optically stimulated luminescence (OSL) study of synthetic stishovite

Optically stimulated luminescence (OSL) of synthetic stishovite was investigated for a future dating technique of meteor impact craters. Luminescence around 330 nm was measured on the γ-ray irradiated stishovite under two stimulating light sources of infrared laser (830 nm) and blue light emitting diode set (470 nm). Thermoluminescence (TL) studies before and after the OSL measurements showed the intensities around 100–200°C and 220–350°C to increase and those around 350–450°C to decrease. This indicates that a part of deep-trapped charges excited during the OSL measurements were retrapped by shallower traps. The infrared stimulated luminescence (IRSL) after the TL measurement up to 450°C could not be detected, while the blue light stimulated luminescence (BLSL) after TL had about one-tenth of the intensity before TL. This indicates that a part of the charges in shallower traps were detrapped thermally and returned to the deeper traps which were related to BLSL. The result implies that some of the BLSL-related traps are quite stable at room temperature and could be used for geological dating. In addition, two paramagnetic centers produced by sudden release of high pressure in synthesis process were found in the unirradiated stishovite by electron spin resonance (ESR). Their g-factors are g=2.00181 and g=2.00062 for an axial signal and g=2.00305 for the other isotropic signal. These signals could be used for an evidence of impacts if those signals could be stored in geological time.     

Circumventing possible inaccuracies of the single aliquot regeneration method for the optical dating of quartz

A simple method of assessing the reliability of the single aliquot regeneration method is presented, based on the repeated measurement of identical regeneration doses. It is shown that for the majority of the samples measured the requirement of a constant repeat point in each cycle is not met. The ability to maintain a constant repeat point is shown to be dependent upon preheating conditions, the requirements for which were found to vary at both inter-sample and inter-aliquot levels. The expected effect on estimated values of D_e is calculated and shown to be significant. Three possible means of circumventing these inaccuracies are presented: (i) adjustment of the measurement procedure (specifically the preheat conditions); (ii) modelling and correcting for the effect, and (iii) stabilising the response prior to construction of the regenerated growth curve. It is suggested that much of the variation in D_e observed between individual aliquots and between individual grains from well-bleached sediments may possibly be attributed to differences in preheating requirements, and that such differences may also account for some of the spread in D_e for partially bleached sediments.     

The Dependence of IRSL Intensity on Radiation Dose for Samples of Chlorides Contained in Feldspars

The intensity of the luminescence generally increases with radiation dose and measurement of these phenomena can be used to characterise the degree of dependence on beta doses. In this study, in order to test whether this is a significant problem on the optically stimulated luminescence (OSL) studies, the radiation dose response of the OSL signal from samples of chlorides contained in feldspars have been investigated by irradiating the samples with beta doses. The infrared-emitting diodes were used with a wavelength of (880±80) nm, and an IRSL (infrared stimulated luminescence) intensity parametres, m, was described and found m = 1±0.03.     

Determining the K-content of single-grains of feldspar for luminescence dating

Feldspars form a solid-solution series whereby the K-content may range from 0 to 14%. LA-ICP-MS measurements for density-separated single-grains of feldspar yielded realistic concentrations of K within the range of those naturally occurring, and also highlighted the difficulty in isolating the pure end members during density-separation. No direct relationship was found between the thermal stability of the infrared-stimulated luminescence (IRSL) signal and measured K-content of individual grains. However, the brightest IRSL and post-IR IRSL signals originated from grains with ～12% K-content. All grains giving a measurable signal had K-content between 6 and 13%, therefore it is suggested that an internal K-content of 10 ± 2% can be assumed for routine single-grain dating of density-separated K-feldspars.     

Post-IR IRSL production in perthitic feldspar

A museum sample of perthitic feldspar was used to study the production of post-IR IRSL signals. It was found that traps responsible for low temperature (∼230 °C) TL peaks play an unexpectedly important role in post-IR IRSL production. During the production of the IRSL signal during low temperature IR stimulation (100 °C), electrons are optically transferred from IRSL traps into these TL traps which have been emptied by the preceding preheat at 320 °C. Subsequent heating to 300 °C causes thermal transfer of these electrons from these traps back into previously emptied IRSL traps which are related to the high temperature TL peaks. IR stimulation of these electrons results in post-IR IRSL. Thus the initial source of the post-IR IRSL signal is the same as the IRSL signal, with a role being played by intermediate traps that give rise to TL signals between 200 and 250 °C, and the final source is similar to that of the IRSL signal. Therefore the post-IR IRSL signal is a by-product of the production of the IRSL signal. It was also found that post-IR IRSL production with high post-IR IR stimulation temperatures (e.g. >230 °C) additionally includes a small contribution from the post-IR isothermal decay of high temperature TL peaks that are not sensitive to IR stimulation at low stimulation temperatures.     

Infrared (IR) stimulated luminescence from modern bricks in retrospective dosimetry applications

It has frequently been observed that certain roof tiles and bricks, especially from relatively modern European buildings, do not contain enough quartz grains in a suitable grain size range to permit dose reconstruction using thermoluminescence (TL) or optically stimulated luminescence (OSL) methods. In this paper the feasibility of using infrared-stimulated luminescence (IRSL) on the feldspar fraction of such bricks and tiles has been investigated. Appropriate preheating treatments were employed in order to select the most stable signals, and procedures were developed to enhance the signal to noise ratio. The possible effect of anomalous fading under application of these procedures was tested. In the dose range above 100 mGy, it has been demonstrated that using IRSL on the feldspar fraction of such material provides a feasible alternative to the use of green-light-stimulated luminescence (GLSL) on the quartz fraction, for the purposes of retrospective dosimetry. Furthermore, since the use of IRSL as described in this paper involves the measurement of polymineral fine grain fractions of bricks, a technique for the calibration of the built-in β source against the γ source in Secondary Standard Dosimetry facilities for routine use of the technique is described.     

Testing Post-IR IRSL protocols for minimising fading in feldspars,using Alaskan loess with independent chronological control

Concern over anomalous fading has been the biggest single factor responsible for deterring the widespread use of the infra-red stimulated luminescence (IRSL) or thermoluminescence (TL) signal from feldspars for luminescence dating. There has therefore been great interest in the use of the recently proposed Post-IR IRSL signal, because it has been shown to significantly reduce the degree of anomalous fading observed in feldspars and therefore potentially provides a means of circumventing the issue. This study undertakes a systematic investigation into various preheat and Post-IR IRSL measurement conditions proposed in the literature, by using two samples from the Halfway House loess section in Alaska which bracket the Old Crow tephra which has been dated using fission track methods. Preheat plateau tests show a dramatic change in equivalent dose with Post-IR IRSL measurement conditions, and further tests reveal that these changes are driven by preheat temperature rather than Post-IR IR stimulation temperature. Dose recovery tests on laboratory-bleached material mimic the findings of the natural preheat plateau test data, and sensitivity change between the first and second Single Aliquot Regenerative dose (SAR) measurement cycle is found to be responsible. Comparison of the Post-IR IRSL ages with the independent age control shows that, for the samples in this study, the Post-IR IR signal stimulated at 290 °C is inappropriate for dating. However, use of lower preheat (250–300 °C) and Post-IR IR stimulation temperatures from 225 to 270 °C gave rise to ages which were in agreement with the independent age control.