- Email: [email protected]
ESR dating of tephra with dose recovery test for impurity centers in quartz
Quaternary International 246 (2011) 118e123
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
ESR dating of tephra with dose recovery test for impurity centers in quartz M. Asagoe a, S. Toyoda a, *, P. Voinchet b, C. Falguères b, H. Tissoux b, T. Suzuki c, D. Banerjee a, d a
Department of Applied Physics, Okayama University of Science, 1-1 Ridai, Okayama 700-0005, Japan Muséum National d’Histoire Naturelle, Paris, France c Tokyo Metropolitan University, Tokyo, Japan d Physical Research Laboratory, Ahmedabad, India b
a r t i c l e i n f o
a b s t r a c t
Article history: Available online 25 June 2011
The multiple-aliquot regenerative-additive method of ESR (electron spin resonance) dating was applied to Numazawa-Shibahara (Nm-Sb) tephra from Fukushima, Japan, with dose recovery tests for impurity centers in quartz. The age obtained from Ti-H center is consistent with the RTL (red thermoluminescence) and Fission Track ages while the ones from Al and Ti-Li center are overestimates. The dose recovery test indicates that the equivalent dose estimate based on the Ti-H center agrees within 6% of the expected dose (370 Gy). The dose recovery test is a useful procedure for choosing the signal appropriate for dating. Ó 2011 Elsevier Ltd and INQUA. All rights reserved.
1. Introduction Determining the age of tephra is important for reconstructing the history of environmental change during the Quaternary epoch. To this end, dating of quartz was attempted using the electron spin resonance (ESR) method. Quartz was first found to be useful for ESR dating of fault gouge (Ikeya et al., 1982), and it was also used for dating of tephra, heated flints (e.g. Porat et al., 1994), and sediments (e.g. Voinchet et al., 2003, 2004). The first investigation pertaining to ESR dating of tephra using quartz was published by Imai et al. (1985) using the Al center (a hole trapped at Al site replacing Si). Subsequently, several other successful results on tephra have been reported (e.g. Imai and Shimokawa, 1988; Imai et al., 1992; Toyoda et al., 1995; Yokoyama et al., 2004). Buhay et al. (1992) reported that the ESR age (45e49 ka) using Al and Ti (Ti-Li) centers of a tephra from New Zealand is consistent with the 14 C age (42e44 ka) within statistical errors. However, in other studies, systematic discrepancies were observed between the ages obtained using the Al center and Ti-Li center (an electron trapped at a Ti atom replacing Si, accompanying a Li ion as a charge compensator). Toyoda et al. (2006) systematically investigated the ESR and RTL (red thermoluminescence) ages of tephra with a known age range of 30e900 ka, and found that ESR dating had problems in obtaining equivalent doses. Using the same dose rate, the RTL ages were consistent with the expected ages while the ESR based results were inconsistent and
* Corresponding author. E-mail address: [email protected] (S. Toyoda). 1040-6182/$ e see front matter Ó 2011 Elsevier Ltd and INQUA. All rights reserved. doi:10.1016/j.quaint.2011.06.027
involved large scatter in data. The scatter in ESR ages was found to increase with age. Toyoda et al. (2009) proposed a new protocol, the multiple-aliquot regenerative-additive dose method, which provides equivalent dose estimates with smaller errors than the traditional additive dose method. This paper indicated that the Al center in quartz could not provide accurate estimates of the equivalent dose by analyzing Nm-Sb tephra in Japan. The present research analyzed the same Nm-Sb tephra to check the reproducibility of dating results, and to test if known doses can be recovered using the multiple-aliquot regenerative-additive dose procedure. 2. Experimental procedures 2.1. Sample preparation The sample analyzed in the present study is a tephra named Numazawa-Shibahara (Nm-Sb), collected at Obazawa, Nishigo Village, Fukushima Prefecture, Japan (37 8.60 N, 140 5.70 E). The sample was taken from the same layer as Toyoda et al. (2009) sampled, but at a different location. Fission-Track ages of 90e130 ka were obtained by Suzuki et al. (1998), and the isothermal RTL age obtained was 116 10 ka (Toyoda et al., 2006). The tephra sample was soaked in 6M HCl overnight and then rinsed in deionized water and dried. It was subsequently sieved to two size fractions, 0.25e0.5 mm and 0.5e1.0 mm. Magnetic minerals were removed by an Nd-B-Fe handy magnet and then by an isodynamic magnetic separator. Sodium polytungstate solution with a density of 2.63 g cm3 was used to remove feldspars with lower density values than the solution. After rinsing and drying, the
M. Asagoe et al. / Quaternary International 246 (2011) 118e123
samples were etched in 20% HF for 2 h to remove the alphairradiated surface and to remove any remaining feldspars. The obtained quartz grains were then gently crushed to 75e250 mm to ensure reproducibility of ESR measurements (Toyoda and Falguères, 2003). 2.2. Multiple-aliquot regenerative-additive dose method The quartz sample was separated into two portions. One portion was heated at 400 C for 30 min. Ten sample aliquots, each with about 100 mg, were prepared from each portion. Gamma ray irradiation was performed at the Japan Atomic Energy Research
119
Institute at Takasaki using a 60Co gamma ray source with a dose rate of 77 Gy h1, from doses of 80 Gy up to 2080 Gy, leaving one aliquot without irradiation. ESR signals of Al, Ti-Li and Ti-H centers (Fig. 1a and b) were recorded for each aliquot at 80e82 K with an X-band ESR spectrometer (JEOL PX-2300), at Okayama University of Science. Analyses used a microwave power of 5 mW, a scan range of 20 mT, a scan time of 30 s, and modulation amplitude of 0.1 mT, with a frequency of 100 kHz. These measurements were repeated five times after rotation of about 30 within the cavity. The peak to peak intensity from g ¼ 2.019 to g ¼ 1.993 was taken as the intensity of the Al center (Fig. 1a), following the method
Fig. 1. (a) ESR spectrum of Al and Ti centers observed in quartz at 82K (a) whole spectrum (b) enlargement of the right wing of Ti centers.
120
M. Asagoe et al. / Quaternary International 246 (2011) 118e123
proposed by Toyoda and Falguères (2003), the baseline to the position at g ¼ 1.913 was taken as the intensity of the Ti-Li center (Yokoyama et al., 2004) and the baseline to the peak at g ¼ 1.915 as that of the Ti-H center (Tissoux et al., 2008) as shown in Fig. 1b. A shoulder (Fig. 1b) is shown at g ¼ 1.913, which corresponds to the bottom of a peak of Ti-Li center overlapping the Ti-H center signal. The dose response of the heated set of aliquots was fitted to a saturating exponential curve as a function of applied dose. Using this curve as reference, the “apparent” doses were obtained for the non-heated set of aliquots as the doses corresponding to respective signal intensities on the regenerative dose response curve (Fig. 2a). Dose values obtained as “apparent doses”, were then plotted against the actual gamma ray dose given to the nonheated irradiated aliquots (Fig. 2b). This procedure is similar to the SARA (single aliquot regeneration added dose) procedure used in OSL dating (Mejdahl and Bøtter-Jensen, 1994). The slope in this plot indicates the sensitivity change relative to the sensitivity before heating. The slope should be unity where no sensitivity change occurs. By extrapolating the correlation between the two doses to the zero ordinate, the equivalent doses, DE, were obtained. All these calculations were dose by using a computer program SALS (Nakagawa and Oyanagi, 1982), which is available at the Research Institute for Information Technology, Kyushu University (http:www.cc.kyushu-u.ac.jp/scp/system/library/SALS/SALS.html). The errors in age were obtained using this computer program (its
Fig. 3. Regenerative dose response of the Al, Ti-Li, and Ti-H centers signals observed in size fraction 0.5e1.0 mm after gamma ray irradiation following heating at 400 C.
principle being similar to that described in Toyoda et al., 1994) from the errors of the parameters, and the correlation between these were calculated by the SALS program. 2.3. Dose recovery test Another portion of quartz sample was prepared and heated at 400 C for 30 min to erase the ESR signals. It was then separated into two sub-portions. The two groups were subsequently irradiated by a 60Co gamma ray source (dose rate w92 Gy h1) and given doses of 370 Gy and 1430 Gy. These samples were considered to be “natural” samples, and the multiple-aliquot regenerative-additive dose method was applied to obtain the “equivalent doses”. 2.4. Dose rate measurements Uranium, thorium and potassium concentrations were measured by gamma ray spectrometry with a low background pure
Fig. 2. Explanatory diagram for the multiple-aliquot regenerative added dose protocol used in the present study. (a) Using regenerative dose response, apparent doses are obtained as doses corresponding to ESR intensities of natural and dosed non-heated aliquots. (b) The apparent doses are plotted as a function of gamma ray dose give to the non-heated aliquots.
Fig. 4. The apparent doses, obtained as the doses corresponding to the respective signal intensity of non-heated aliquots on the regenerative dose response (Fig. 3), plotted against the actual gamma ray dose. The slopes indicate the change in sensitivity of the signal on heating. Best fitted lines are extrapolated to the zero ordinate to obtain the equivalent doses.
M. Asagoe et al. / Quaternary International 246 (2011) 118e123
121
Table 1 Equivalent doses and calculated ages by multiple-aliquot regenerative-additive dose method. Signal
Al center
Grain size
0.25e0.5 mm
0.5e1 mm
Ti-Li center 0.25e0.5 mm
0.5e1 mm
Ti-H center 0.25e0.5 mm
0.5e1 mm
Equivalent dose (Gy) Slope Age (ka)
163 9 0.812 0.023 259þ15 14
133þ13 12 0.926 0.043 225þ22 20
210þ16 15 1.04 0.048 335þ25 23
224þ29 26 1.11 0.065 377þ48 44
81þ7 6 0.986 0.039 130 10
71 6 1.08 0.043 120þ11 10
germanium semiconductor detector system. Gamma ray peaks were compared with those of a standard sample to deduce those concentrations in the sample. Water content was estimated by weighing the sediment before and after being dried. The dose rate was calculated from the U, Th, and K concentrations by using the conversion factors obtained by Adamiec and Aitken (1998). Corrections were made for water contents (Aitken and Xie, 1990) and beta ray attenuation factors (Mejdahl, 1979). The cosmic dose rate was assumed to be 0.18 mGy y1 obtained from the formula proposed by Prescott and Hutton (1988) where the thickness of the overlying sediment was about 1 m at the sampling site. 3. Results 3.1. Equivalent doses and the ages The regenerative dose responses of Al, Ti-Li, and Ti-H centers are shown in Fig. 3. Using these curves, the apparent doses are obtained for unheated aliquots and plotted as a function of dose as shown in Fig. 4. The equivalent doses (DE) were obtained as shown in Table 1 by extrapolating the lines fitted to these points, to the zero ordinate. The equivalent doses, obtained by the additive dose method and the regeneration method using the same set of present data were essentially the same, within the statistical errors. The dose rate was calculated from the uranium, thorium and potassium concentrations as listed in Table 2 for the two grain size fractions. The difference in dose rate is due to the difference in beta ray attenuation. The ages were obtained by dividing the equivalent doses by the dose rate as shown in Table 1. The obtained ages are consistent within the error between respective grain size fractions. The ages using Ti-H center (120þ11 10 and 130 10 ka, Fig. 5) are consistent with the RTL age of 116 10 ka (Toyoda et al., 2006) and Fission-Track age of 90e130 ka (Suzuki et al., 1998). However, the ages using Ti-Li center (335, 377 ka) and Al center (225, 259 ka) are overestimates.
Fig. 5. Comparison of ages obtained by the multiple-aliquot regenerative-additive procedures used in the present study.
3.2. Dose recovery tests The results of the dose recovery tests are shown in Fig. 6 and Table 3. For the samples irradiated to 370 Gy, the doses obtained using the Ti-H center are consistent with the given dose. The dose obtained using the Ti-Li center for one fraction is consistent with 370 Gy within the error range but the other fraction gives an overestimate by 100 Gy, although it is still within the range of 2s. The values using Al center were both underestimates. For the
Table 2 The concentrations of U, Th, K and calculated dose rate. Grain size
U(ppm)
Th(ppm)
k2o(%)
Water content (%)
Dose rate (mGy/y)
0.25e0.5 mm 0.5e1 mm
0.966
6.36
0.538
113
0.629 0.594
Fig. 6. (a) and (b) Equivalent doses obtained from the dose recovery test by multiplealiquot regenerative-additive procedures for (a) 370 Gy and (b) 1430 Gy.
122
M. Asagoe et al. / Quaternary International 246 (2011) 118e123
Table 3 Equivalent doses and the slopes in regenerative-additive plot obtained by the dose recovery test. Signal
Al center
Grain size
0.25e0.5 mm
0.5e1 mm
0.25e0.5 mm
0.5e1 mm
0.25e0.5 mm
0.5e1 mm
233þ29 25 1.24 0.108
263þ44 37 1.15 0.104
407þ124 89 1.19 0.209
472þ74 62 0.928 0.086
345þ68 54 1.27 0.149
377þ55 46 1.22 0.118
704þ182 133 1.68 0.268
585þ95 75 1.93 0.231
1560þ798 442 0.886 0.257
995þ259 181 1.38 0.248
e e
1240þ530 340 1.69 0.391
Dose recovery test (368 Gy) Equivalent dose (Gy) Slope Dose recovery test (1430 Gy) Equivalent dose (Gy) Slope
Ti-Li center
samples irradiated to 1430 Gy, the obtained doses from Ti-Li center are close while the values from the Al center are under-estimated and the dose was not obtained from Ti-H center for one fraction. 4. Discussion The present results raise a serious question for ESR dating of tephra using quartz, suggesting that the signal may not have retained the correct information on natural accumulated dose. The dose recovery test may be a criterion to check this issue. The test was successful for the Ti-H center in the sample irradiated to 370 Gy. The ESR signal which provides a valid dose recovery test value provided an ESR age estimate consistent with other techniques applied to this tephra. The scatter for Ti-Li and Al center in the dose recovery test is within a factor of two while the ages were overestimated beyond this factor. This would imply that there are other mechanisms which cause the age inconsistencies. There are possible inflection points in dose responses shown in Fig. 2a. These complex dose responses may be one such mechanism. On the other hand, in the other dose recovery test (1430 Gy), although the statistical errors are larger than those for 370 Gy because the fit was not very good in the regenerative dose response. The doses from Ti-Li center are closer than the values from the other signals. For the Ti-H center signal, as shown in Fig. 2, the intensity saturates around 1200 Gy indicating the upper limit of this dating procedure. The value was not obtained for the finer fraction because the intensity exceeded the saturation value. Another serious problem is the reproducibility of the dating results. In the present paper, the same tephra as Toyoda et al. (2009) considered was analyzed, but using a sample taken from another locality. In their investigation, the ages from Ti-Li center were consistent with the RTL age, but the one from the Al center was overestimated. They did not observe a Ti-H center signal in that tephra sample. The dose rate in the present work is about 15% lower than Toyoda et al. (2009) due to higher water content, but the equivalent doses are about double. The present work has shown that even the samples from the same stratigraphy result in different equivalent doses despite originating from the same strata. No obvious xenocrysts were found in both samples, but further mineralogical and/or lithologic study would be necessary to exclude the possibility that some quartz grains of other origins have contaminated the tephra layer. The ESR ages of quartz from tephra deposits is consistent with controls in the age range of about 20 ka (e.g. Yokoyama et al., 2004) but the inconsistency increases with age as shown by Toyoda et al. (2006). The present work has shown that ESR dating of quartz from tephra deposits may give wrong ages with the available methods for estimating the equivalent dose. The dose recovery tests showed that, for the present sample, Ti-H center gives correct equivalent dose for 370 Gy and Ti-Li center for 1480 Gy.
Ti-H center
The present results give no indication for the reasons for these problems. It is possibly due to some electronic processes within the quartz matrix. Toyoda et al. (2006) noted changes in the shape of the dose response on heating. It would also be possible that alkali ions, which are the charge compensators of Ti centers, move within the crystal structure affecting the signal formation. Further studies are necessary to find appropriate experimental procedures for dating of quartz from tephra deposits. It is also necessary to check if the problems noted in the present work are unique to quartz in tephra, or also occur in quartz in fault gouge or in sediments. Further work is required to test whether the problems are universal for all quartz in tephra deposits. 5. Conclusion The ESR dating technique was applied to a tephra named NmSb. The age obtained from Ti-H center is consistent with the control value while the ones from Al and Ti-Li center are overestimates. The dose recovery test indicates that the Ti-H center obtains correct equivalent doses for the present sample. However, there are still problems in dating of quartz such as reproducibility in obtaining the equivalent doses. Acknowledgements This study was supported by JSPS/CNRS Japan-France Cooperative Science Program and by the Inter-University Program for the Common Use of JAERI Facilities awarded to ST. References Adamiec, G., Aitken, M.J., 1998. Dose-rate conversion factors: update. Ancient Thermoluminescence 16, 37e50. Aitken, M.J., Xie, J., 1990. Moisture correction for annual gamma dose. Ancient TL 8, 6e9. Buhay, W.M., Clifford, P.M., Shcwarcz, H.P., 1992. ESR dating of the Rotoiti breccias in the Taupo volcanic zone, New Zealand. Quaternary Science Reviews 11, 267e271. Ikeya, M., Miki, T., Tanaka, K., 1982. Dating of fault by electron spin resonance on intrafault materials. Science 215, 1392e1393. Imai, N., Shimokawa, K., 1988. ESR dating of quaternary tephra from Mt. Osore-zan using Al and Ti centers in quartz. Quaternary Science Reviews 7, 523e527. Imai, N., Shimokawa, K., Hirota, M., 1985. ESR dating of volcanic ash. Nature 314, 81e83. Imai, N., Shimokawa, K., Sakaguchi, K., Takada, M., 1992. ESR dates and thermal behaviour of Al and Ti centers in quartz for the tephra and welded tuff in Japan. Quaternary Science Reviews 11, 257e265. Mejdahl, V., 1979. Thermoluminescence dating: beta-dose attenuation in quartz grains. Archaeometry 21, 61e72. Mejdahl, V., Bøtter-Jensen, L., 1994. Luminescence dating of archaeological materials using a new technique based on single aliquot measurements. Quaternary Science Reviews 7, 551e554. Nakagawa, T., Oyanagi, T., 1982. Experimental Data Analysis by the Least Squares Method: The Program SALS (in Japanese, the Japanese title translated). University of Tokyo Press, Tokyo. 206pp. Porat, N., Schwarcz, H.P., Valladas, H., Bar-Yosef, O., Vandermeersch, B., 1994. Electron spin resonance dating of burned flint from Kebara cave, Israel. Geoarchaeology 9, 393e407. Prescott, J.R., Hutton, J.T., 1988. Cosmic ray and gamma ray dosimetry for TL and ESR. Nuclear Tracks and Radiation Measurements 14, 223e227.
M. Asagoe et al. / Quaternary International 246 (2011) 118e123 Suzuki, T., Fujiwara, O., Danhara, T., 1998. The fission-track age of quaternary tephras in Kanto and Chubu district, Japan. Journal of the Geological Society of Japan 107, 348e367. Tissoux, H., Toyoda, S., Falguères, C., Voinchet, P., Takada, M., Bahain, J.-J., Desprièe, J., 2008. ESR dating of sedimentary quartz from two pleistocene deposits using Al and Ti-centers. Geochronometria 30, 23e31. Toyoda, S., Falguères, C., 2003. The method to represent the ESR signal intensity of the aluminum hole center in quartz for the purpose of dating. Advances in ESR Applications 20, 7e10. Toyoda, S., Goff, F., Ikeda, S., Ikeya, M., 1995. ESR dating of El Cajete and Battleship Rock Member of Valles Rhyolite, Valles Caldera, New Mexico. Journal of Volcanology and Geothermal Research 67, 29e40. Toyoda, S., Sumitomo, H., Ikeya, M., Ishii, H., 1994. Separation of CO 2 signal from a spectrum of irradiated tooth enamel in ESR dosimetry. In: Assessment of the health and environmental impact from radiation doses due to released radionuclides, NIRS-M-102, pp. 187e205.
123
Toyoda, S., Miura, H., Tissoux, H., 2009. Signal regeneration in ESR dating of tephra with quartz. Radiation Measurements 44, 483e487. Toyoda, S., Tsukamoto, S., Hameau, S., Usui, H., Suzuki, T., 2006. Dating of Japanese quaternary tephras by ESR and luminescence methods. Quaternary Geochronology 1, 320e326. Voinchet, P., Bahain, J.J., Falgueres, C., Laurent, M., Dolo, J.M., Despriee, J., Gagenonnet, R., 2004. ESR dating of quartz extracted from quaternary sediments, application to fluvial terraces system of northern France. Quaternaire 15, 135e141. Voinchet, P., Falguères, C., Laurent, M., Toyoda, S., Bahain, J.J., Dolo, J.M., 2003. Artificial optical bleaching of the aluminium center in quartz. Implications to ESR dating of sediments. Quaternary Geochronology (Quaternary Science Reviews) 22, 1335e1338. Yokoyama, T., Shimada, A., Umemura, T., Toyoda, S., 2004. ESR ages of rhyolitic monogenetic volcanoes in Kozushima, Japan. Bulletin of the Volcanological Society of Japan 49, 23e32 (in Japanese with English abstract).
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
ESR dating of tephra with dose recovery test for impurity centers in quartz M. Asagoe a, S. Toyoda a, *, P. Voinchet b, C. Falguères b, H. Tissoux b, T. Suzuki c, D. Banerjee a, d a
Department of Applied Physics, Okayama University of Science, 1-1 Ridai, Okayama 700-0005, Japan Muséum National d’Histoire Naturelle, Paris, France c Tokyo Metropolitan University, Tokyo, Japan d Physical Research Laboratory, Ahmedabad, India b
a r t i c l e i n f o
a b s t r a c t
Article history: Available online 25 June 2011
The multiple-aliquot regenerative-additive method of ESR (electron spin resonance) dating was applied to Numazawa-Shibahara (Nm-Sb) tephra from Fukushima, Japan, with dose recovery tests for impurity centers in quartz. The age obtained from Ti-H center is consistent with the RTL (red thermoluminescence) and Fission Track ages while the ones from Al and Ti-Li center are overestimates. The dose recovery test indicates that the equivalent dose estimate based on the Ti-H center agrees within 6% of the expected dose (370 Gy). The dose recovery test is a useful procedure for choosing the signal appropriate for dating. Ó 2011 Elsevier Ltd and INQUA. All rights reserved.
1. Introduction Determining the age of tephra is important for reconstructing the history of environmental change during the Quaternary epoch. To this end, dating of quartz was attempted using the electron spin resonance (ESR) method. Quartz was first found to be useful for ESR dating of fault gouge (Ikeya et al., 1982), and it was also used for dating of tephra, heated flints (e.g. Porat et al., 1994), and sediments (e.g. Voinchet et al., 2003, 2004). The first investigation pertaining to ESR dating of tephra using quartz was published by Imai et al. (1985) using the Al center (a hole trapped at Al site replacing Si). Subsequently, several other successful results on tephra have been reported (e.g. Imai and Shimokawa, 1988; Imai et al., 1992; Toyoda et al., 1995; Yokoyama et al., 2004). Buhay et al. (1992) reported that the ESR age (45e49 ka) using Al and Ti (Ti-Li) centers of a tephra from New Zealand is consistent with the 14 C age (42e44 ka) within statistical errors. However, in other studies, systematic discrepancies were observed between the ages obtained using the Al center and Ti-Li center (an electron trapped at a Ti atom replacing Si, accompanying a Li ion as a charge compensator). Toyoda et al. (2006) systematically investigated the ESR and RTL (red thermoluminescence) ages of tephra with a known age range of 30e900 ka, and found that ESR dating had problems in obtaining equivalent doses. Using the same dose rate, the RTL ages were consistent with the expected ages while the ESR based results were inconsistent and
* Corresponding author. E-mail address: [email protected] (S. Toyoda). 1040-6182/$ e see front matter Ó 2011 Elsevier Ltd and INQUA. All rights reserved. doi:10.1016/j.quaint.2011.06.027
involved large scatter in data. The scatter in ESR ages was found to increase with age. Toyoda et al. (2009) proposed a new protocol, the multiple-aliquot regenerative-additive dose method, which provides equivalent dose estimates with smaller errors than the traditional additive dose method. This paper indicated that the Al center in quartz could not provide accurate estimates of the equivalent dose by analyzing Nm-Sb tephra in Japan. The present research analyzed the same Nm-Sb tephra to check the reproducibility of dating results, and to test if known doses can be recovered using the multiple-aliquot regenerative-additive dose procedure. 2. Experimental procedures 2.1. Sample preparation The sample analyzed in the present study is a tephra named Numazawa-Shibahara (Nm-Sb), collected at Obazawa, Nishigo Village, Fukushima Prefecture, Japan (37 8.60 N, 140 5.70 E). The sample was taken from the same layer as Toyoda et al. (2009) sampled, but at a different location. Fission-Track ages of 90e130 ka were obtained by Suzuki et al. (1998), and the isothermal RTL age obtained was 116 10 ka (Toyoda et al., 2006). The tephra sample was soaked in 6M HCl overnight and then rinsed in deionized water and dried. It was subsequently sieved to two size fractions, 0.25e0.5 mm and 0.5e1.0 mm. Magnetic minerals were removed by an Nd-B-Fe handy magnet and then by an isodynamic magnetic separator. Sodium polytungstate solution with a density of 2.63 g cm3 was used to remove feldspars with lower density values than the solution. After rinsing and drying, the
M. Asagoe et al. / Quaternary International 246 (2011) 118e123
samples were etched in 20% HF for 2 h to remove the alphairradiated surface and to remove any remaining feldspars. The obtained quartz grains were then gently crushed to 75e250 mm to ensure reproducibility of ESR measurements (Toyoda and Falguères, 2003). 2.2. Multiple-aliquot regenerative-additive dose method The quartz sample was separated into two portions. One portion was heated at 400 C for 30 min. Ten sample aliquots, each with about 100 mg, were prepared from each portion. Gamma ray irradiation was performed at the Japan Atomic Energy Research
119
Institute at Takasaki using a 60Co gamma ray source with a dose rate of 77 Gy h1, from doses of 80 Gy up to 2080 Gy, leaving one aliquot without irradiation. ESR signals of Al, Ti-Li and Ti-H centers (Fig. 1a and b) were recorded for each aliquot at 80e82 K with an X-band ESR spectrometer (JEOL PX-2300), at Okayama University of Science. Analyses used a microwave power of 5 mW, a scan range of 20 mT, a scan time of 30 s, and modulation amplitude of 0.1 mT, with a frequency of 100 kHz. These measurements were repeated five times after rotation of about 30 within the cavity. The peak to peak intensity from g ¼ 2.019 to g ¼ 1.993 was taken as the intensity of the Al center (Fig. 1a), following the method
Fig. 1. (a) ESR spectrum of Al and Ti centers observed in quartz at 82K (a) whole spectrum (b) enlargement of the right wing of Ti centers.
120
M. Asagoe et al. / Quaternary International 246 (2011) 118e123
proposed by Toyoda and Falguères (2003), the baseline to the position at g ¼ 1.913 was taken as the intensity of the Ti-Li center (Yokoyama et al., 2004) and the baseline to the peak at g ¼ 1.915 as that of the Ti-H center (Tissoux et al., 2008) as shown in Fig. 1b. A shoulder (Fig. 1b) is shown at g ¼ 1.913, which corresponds to the bottom of a peak of Ti-Li center overlapping the Ti-H center signal. The dose response of the heated set of aliquots was fitted to a saturating exponential curve as a function of applied dose. Using this curve as reference, the “apparent” doses were obtained for the non-heated set of aliquots as the doses corresponding to respective signal intensities on the regenerative dose response curve (Fig. 2a). Dose values obtained as “apparent doses”, were then plotted against the actual gamma ray dose given to the nonheated irradiated aliquots (Fig. 2b). This procedure is similar to the SARA (single aliquot regeneration added dose) procedure used in OSL dating (Mejdahl and Bøtter-Jensen, 1994). The slope in this plot indicates the sensitivity change relative to the sensitivity before heating. The slope should be unity where no sensitivity change occurs. By extrapolating the correlation between the two doses to the zero ordinate, the equivalent doses, DE, were obtained. All these calculations were dose by using a computer program SALS (Nakagawa and Oyanagi, 1982), which is available at the Research Institute for Information Technology, Kyushu University (http:www.cc.kyushu-u.ac.jp/scp/system/library/SALS/SALS.html). The errors in age were obtained using this computer program (its
Fig. 3. Regenerative dose response of the Al, Ti-Li, and Ti-H centers signals observed in size fraction 0.5e1.0 mm after gamma ray irradiation following heating at 400 C.
principle being similar to that described in Toyoda et al., 1994) from the errors of the parameters, and the correlation between these were calculated by the SALS program. 2.3. Dose recovery test Another portion of quartz sample was prepared and heated at 400 C for 30 min to erase the ESR signals. It was then separated into two sub-portions. The two groups were subsequently irradiated by a 60Co gamma ray source (dose rate w92 Gy h1) and given doses of 370 Gy and 1430 Gy. These samples were considered to be “natural” samples, and the multiple-aliquot regenerative-additive dose method was applied to obtain the “equivalent doses”. 2.4. Dose rate measurements Uranium, thorium and potassium concentrations were measured by gamma ray spectrometry with a low background pure
Fig. 2. Explanatory diagram for the multiple-aliquot regenerative added dose protocol used in the present study. (a) Using regenerative dose response, apparent doses are obtained as doses corresponding to ESR intensities of natural and dosed non-heated aliquots. (b) The apparent doses are plotted as a function of gamma ray dose give to the non-heated aliquots.
Fig. 4. The apparent doses, obtained as the doses corresponding to the respective signal intensity of non-heated aliquots on the regenerative dose response (Fig. 3), plotted against the actual gamma ray dose. The slopes indicate the change in sensitivity of the signal on heating. Best fitted lines are extrapolated to the zero ordinate to obtain the equivalent doses.
M. Asagoe et al. / Quaternary International 246 (2011) 118e123
121
Table 1 Equivalent doses and calculated ages by multiple-aliquot regenerative-additive dose method. Signal
Al center
Grain size
0.25e0.5 mm
0.5e1 mm
Ti-Li center 0.25e0.5 mm
0.5e1 mm
Ti-H center 0.25e0.5 mm
0.5e1 mm
Equivalent dose (Gy) Slope Age (ka)
163 9 0.812 0.023 259þ15 14
133þ13 12 0.926 0.043 225þ22 20
210þ16 15 1.04 0.048 335þ25 23
224þ29 26 1.11 0.065 377þ48 44
81þ7 6 0.986 0.039 130 10
71 6 1.08 0.043 120þ11 10
germanium semiconductor detector system. Gamma ray peaks were compared with those of a standard sample to deduce those concentrations in the sample. Water content was estimated by weighing the sediment before and after being dried. The dose rate was calculated from the U, Th, and K concentrations by using the conversion factors obtained by Adamiec and Aitken (1998). Corrections were made for water contents (Aitken and Xie, 1990) and beta ray attenuation factors (Mejdahl, 1979). The cosmic dose rate was assumed to be 0.18 mGy y1 obtained from the formula proposed by Prescott and Hutton (1988) where the thickness of the overlying sediment was about 1 m at the sampling site. 3. Results 3.1. Equivalent doses and the ages The regenerative dose responses of Al, Ti-Li, and Ti-H centers are shown in Fig. 3. Using these curves, the apparent doses are obtained for unheated aliquots and plotted as a function of dose as shown in Fig. 4. The equivalent doses (DE) were obtained as shown in Table 1 by extrapolating the lines fitted to these points, to the zero ordinate. The equivalent doses, obtained by the additive dose method and the regeneration method using the same set of present data were essentially the same, within the statistical errors. The dose rate was calculated from the uranium, thorium and potassium concentrations as listed in Table 2 for the two grain size fractions. The difference in dose rate is due to the difference in beta ray attenuation. The ages were obtained by dividing the equivalent doses by the dose rate as shown in Table 1. The obtained ages are consistent within the error between respective grain size fractions. The ages using Ti-H center (120þ11 10 and 130 10 ka, Fig. 5) are consistent with the RTL age of 116 10 ka (Toyoda et al., 2006) and Fission-Track age of 90e130 ka (Suzuki et al., 1998). However, the ages using Ti-Li center (335, 377 ka) and Al center (225, 259 ka) are overestimates.
Fig. 5. Comparison of ages obtained by the multiple-aliquot regenerative-additive procedures used in the present study.
3.2. Dose recovery tests The results of the dose recovery tests are shown in Fig. 6 and Table 3. For the samples irradiated to 370 Gy, the doses obtained using the Ti-H center are consistent with the given dose. The dose obtained using the Ti-Li center for one fraction is consistent with 370 Gy within the error range but the other fraction gives an overestimate by 100 Gy, although it is still within the range of 2s. The values using Al center were both underestimates. For the
Table 2 The concentrations of U, Th, K and calculated dose rate. Grain size
U(ppm)
Th(ppm)
k2o(%)
Water content (%)
Dose rate (mGy/y)
0.25e0.5 mm 0.5e1 mm
0.966
6.36
0.538
113
0.629 0.594
Fig. 6. (a) and (b) Equivalent doses obtained from the dose recovery test by multiplealiquot regenerative-additive procedures for (a) 370 Gy and (b) 1430 Gy.
122
M. Asagoe et al. / Quaternary International 246 (2011) 118e123
Table 3 Equivalent doses and the slopes in regenerative-additive plot obtained by the dose recovery test. Signal
Al center
Grain size
0.25e0.5 mm
0.5e1 mm
0.25e0.5 mm
0.5e1 mm
0.25e0.5 mm
0.5e1 mm
233þ29 25 1.24 0.108
263þ44 37 1.15 0.104
407þ124 89 1.19 0.209
472þ74 62 0.928 0.086
345þ68 54 1.27 0.149
377þ55 46 1.22 0.118
704þ182 133 1.68 0.268
585þ95 75 1.93 0.231
1560þ798 442 0.886 0.257
995þ259 181 1.38 0.248
e e
1240þ530 340 1.69 0.391
Dose recovery test (368 Gy) Equivalent dose (Gy) Slope Dose recovery test (1430 Gy) Equivalent dose (Gy) Slope
Ti-Li center
samples irradiated to 1430 Gy, the obtained doses from Ti-Li center are close while the values from the Al center are under-estimated and the dose was not obtained from Ti-H center for one fraction. 4. Discussion The present results raise a serious question for ESR dating of tephra using quartz, suggesting that the signal may not have retained the correct information on natural accumulated dose. The dose recovery test may be a criterion to check this issue. The test was successful for the Ti-H center in the sample irradiated to 370 Gy. The ESR signal which provides a valid dose recovery test value provided an ESR age estimate consistent with other techniques applied to this tephra. The scatter for Ti-Li and Al center in the dose recovery test is within a factor of two while the ages were overestimated beyond this factor. This would imply that there are other mechanisms which cause the age inconsistencies. There are possible inflection points in dose responses shown in Fig. 2a. These complex dose responses may be one such mechanism. On the other hand, in the other dose recovery test (1430 Gy), although the statistical errors are larger than those for 370 Gy because the fit was not very good in the regenerative dose response. The doses from Ti-Li center are closer than the values from the other signals. For the Ti-H center signal, as shown in Fig. 2, the intensity saturates around 1200 Gy indicating the upper limit of this dating procedure. The value was not obtained for the finer fraction because the intensity exceeded the saturation value. Another serious problem is the reproducibility of the dating results. In the present paper, the same tephra as Toyoda et al. (2009) considered was analyzed, but using a sample taken from another locality. In their investigation, the ages from Ti-Li center were consistent with the RTL age, but the one from the Al center was overestimated. They did not observe a Ti-H center signal in that tephra sample. The dose rate in the present work is about 15% lower than Toyoda et al. (2009) due to higher water content, but the equivalent doses are about double. The present work has shown that even the samples from the same stratigraphy result in different equivalent doses despite originating from the same strata. No obvious xenocrysts were found in both samples, but further mineralogical and/or lithologic study would be necessary to exclude the possibility that some quartz grains of other origins have contaminated the tephra layer. The ESR ages of quartz from tephra deposits is consistent with controls in the age range of about 20 ka (e.g. Yokoyama et al., 2004) but the inconsistency increases with age as shown by Toyoda et al. (2006). The present work has shown that ESR dating of quartz from tephra deposits may give wrong ages with the available methods for estimating the equivalent dose. The dose recovery tests showed that, for the present sample, Ti-H center gives correct equivalent dose for 370 Gy and Ti-Li center for 1480 Gy.
Ti-H center
The present results give no indication for the reasons for these problems. It is possibly due to some electronic processes within the quartz matrix. Toyoda et al. (2006) noted changes in the shape of the dose response on heating. It would also be possible that alkali ions, which are the charge compensators of Ti centers, move within the crystal structure affecting the signal formation. Further studies are necessary to find appropriate experimental procedures for dating of quartz from tephra deposits. It is also necessary to check if the problems noted in the present work are unique to quartz in tephra, or also occur in quartz in fault gouge or in sediments. Further work is required to test whether the problems are universal for all quartz in tephra deposits. 5. Conclusion The ESR dating technique was applied to a tephra named NmSb. The age obtained from Ti-H center is consistent with the control value while the ones from Al and Ti-Li center are overestimates. The dose recovery test indicates that the Ti-H center obtains correct equivalent doses for the present sample. However, there are still problems in dating of quartz such as reproducibility in obtaining the equivalent doses. Acknowledgements This study was supported by JSPS/CNRS Japan-France Cooperative Science Program and by the Inter-University Program for the Common Use of JAERI Facilities awarded to ST. References Adamiec, G., Aitken, M.J., 1998. Dose-rate conversion factors: update. Ancient Thermoluminescence 16, 37e50. Aitken, M.J., Xie, J., 1990. Moisture correction for annual gamma dose. Ancient TL 8, 6e9. Buhay, W.M., Clifford, P.M., Shcwarcz, H.P., 1992. ESR dating of the Rotoiti breccias in the Taupo volcanic zone, New Zealand. Quaternary Science Reviews 11, 267e271. Ikeya, M., Miki, T., Tanaka, K., 1982. Dating of fault by electron spin resonance on intrafault materials. Science 215, 1392e1393. Imai, N., Shimokawa, K., 1988. ESR dating of quaternary tephra from Mt. Osore-zan using Al and Ti centers in quartz. Quaternary Science Reviews 7, 523e527. Imai, N., Shimokawa, K., Hirota, M., 1985. ESR dating of volcanic ash. Nature 314, 81e83. Imai, N., Shimokawa, K., Sakaguchi, K., Takada, M., 1992. ESR dates and thermal behaviour of Al and Ti centers in quartz for the tephra and welded tuff in Japan. Quaternary Science Reviews 11, 257e265. Mejdahl, V., 1979. Thermoluminescence dating: beta-dose attenuation in quartz grains. Archaeometry 21, 61e72. Mejdahl, V., Bøtter-Jensen, L., 1994. Luminescence dating of archaeological materials using a new technique based on single aliquot measurements. Quaternary Science Reviews 7, 551e554. Nakagawa, T., Oyanagi, T., 1982. Experimental Data Analysis by the Least Squares Method: The Program SALS (in Japanese, the Japanese title translated). University of Tokyo Press, Tokyo. 206pp. Porat, N., Schwarcz, H.P., Valladas, H., Bar-Yosef, O., Vandermeersch, B., 1994. Electron spin resonance dating of burned flint from Kebara cave, Israel. Geoarchaeology 9, 393e407. Prescott, J.R., Hutton, J.T., 1988. Cosmic ray and gamma ray dosimetry for TL and ESR. Nuclear Tracks and Radiation Measurements 14, 223e227.
M. Asagoe et al. / Quaternary International 246 (2011) 118e123 Suzuki, T., Fujiwara, O., Danhara, T., 1998. The fission-track age of quaternary tephras in Kanto and Chubu district, Japan. Journal of the Geological Society of Japan 107, 348e367. Tissoux, H., Toyoda, S., Falguères, C., Voinchet, P., Takada, M., Bahain, J.-J., Desprièe, J., 2008. ESR dating of sedimentary quartz from two pleistocene deposits using Al and Ti-centers. Geochronometria 30, 23e31. Toyoda, S., Falguères, C., 2003. The method to represent the ESR signal intensity of the aluminum hole center in quartz for the purpose of dating. Advances in ESR Applications 20, 7e10. Toyoda, S., Goff, F., Ikeda, S., Ikeya, M., 1995. ESR dating of El Cajete and Battleship Rock Member of Valles Rhyolite, Valles Caldera, New Mexico. Journal of Volcanology and Geothermal Research 67, 29e40. Toyoda, S., Sumitomo, H., Ikeya, M., Ishii, H., 1994. Separation of CO 2 signal from a spectrum of irradiated tooth enamel in ESR dosimetry. In: Assessment of the health and environmental impact from radiation doses due to released radionuclides, NIRS-M-102, pp. 187e205.
123
Toyoda, S., Miura, H., Tissoux, H., 2009. Signal regeneration in ESR dating of tephra with quartz. Radiation Measurements 44, 483e487. Toyoda, S., Tsukamoto, S., Hameau, S., Usui, H., Suzuki, T., 2006. Dating of Japanese quaternary tephras by ESR and luminescence methods. Quaternary Geochronology 1, 320e326. Voinchet, P., Bahain, J.J., Falgueres, C., Laurent, M., Dolo, J.M., Despriee, J., Gagenonnet, R., 2004. ESR dating of quartz extracted from quaternary sediments, application to fluvial terraces system of northern France. Quaternaire 15, 135e141. Voinchet, P., Falguères, C., Laurent, M., Toyoda, S., Bahain, J.J., Dolo, J.M., 2003. Artificial optical bleaching of the aluminium center in quartz. Implications to ESR dating of sediments. Quaternary Geochronology (Quaternary Science Reviews) 22, 1335e1338. Yokoyama, T., Shimada, A., Umemura, T., Toyoda, S., 2004. ESR ages of rhyolitic monogenetic volcanoes in Kozushima, Japan. Bulletin of the Volcanological Society of Japan 49, 23e32 (in Japanese with English abstract).