Comparison of naturally accumulated radiation-doses ...svr4.terrapub.co.jp/journals/GJ/pdf/3903/39030201.pdf · In order to evaluate correct accumulated-doses, a sin-gle-aliquot regenerative-dose

201

Geochemical Journal, Vol. 39, pp. 201 to 212, 2005

*Corresponding author (e-mail: [email protected])

Copyright © 2005 by The Geochemical Society of Japan.

Comparison of naturally accumulated radiation-doses between RTL, BTL, OSL,and IRSL using white minerals from burnt archaeological materials

and usefulness of RTL-dating from quartz extracts

TETSUO HASHIMOTO,1* TAKASHI YAWATA2 and MASATO TAKANO1

1Department of Chemistry, Faculty of Science, Niigata University, Niigata 950-2181, Japan2Graduate School of Science and Technology, Niigata University, Niigata 950-2181, Japan

(Received June 18, 2004; Accepted November 2, 2004)

Naturally accumulated doses after pottery manufacture were compared between optically stimulated luminescence(OSL) from quartz grains and infrared-stimulated luminescence (IRSL) from feldspar grains, together with a red-thermoluminescence (RTL) result from quartz grains. Quartz and feldspar extracts from nine Jomon pottery pieces, whichwere manufactured and used 3,500–6,000 years ago, were subjected to three kinds of luminescence measurements using anew automated luminescence measuring system. Since this system included a small X-ray irradiator, a single-aliquotregenerative-dose (SAR) protocol was applicable to each luminescence dosimetry. Naturally accumulated doses fromRTL were higher than OSL- and IRSL-dose results. This is probably due to relatively easy-bleaching effects or unstableluminescence properties related to the OSL in quartz grains and well-known anomalous fading effects of feldspar grains.On the basis of these results, the RTL-luminescence ages, estimated using the accumulated doses and annual doses, arecloser to the predicted archaeological ages. In conclusion, it was also confirmed here that the RTL-dating is most suitabledating-method when quartz extracts from burnt archaeological materials are available.

Keywords: naturally accumulated dose, red thermoluminescence (RTL), optically stimulated luminescence (OSL), infra-red stimulatedluminescence (IRSL), quartz and feldspar, burnt archaeological material

from luminescence intensity. Optically blue-light stimu-lated luminescence (OSL) and infra-red stimulated lumi-nescence (IRSL) have been proposed for the evaluationof accumulated dose using quartz and feldspar grains,respectively. In the OSL and IRSL, the induced-lumines-cence possesses emission properties that follow the anti-Stokes’ law, corresponding to luminescence with shorterwavelengths different from the stimulated light (Aitken,1998).

In order to evaluate correct accumulated-doses, a sin-gle-aliquot regenerative-dose (SAR) technique has beendeveloped mainly for OSL-measurements, in which lu-minescence-measurement and irradiation could be repeat-edly adopted under the same geometrical conditions andaccompanied with the correction of sensitivity changes(Wintle and Murray, 2000). A SAR-RTL method was suc-cessfully applied for volcanic dating, which extended thetime range of luminescence dating to more than 1.2 Ma(Fattahi and Stokes, 2000). Recently, the SAR-protocolwas assured to be attainable using a new automated lu-minescence measuring system equipped with a small X-ray irradiator (Hashimoto et al., 2002a).

Since no-sensitivity changes have been confirmed forboth RTL- and BTL-measurements, a SAR-protocol with-out the sensitivity changes was applied to TL-dating of

INTRODUCTION

The existence of red thermoluminescence (RTL) fromquartz grains was initially discovered from color photo-graphic images from Niigata dune sand in addition to thealready well known blue TL (BTL) (Hashimoto et al.,1986, 1987). Subsequently, both volcanic quartz grainsand artificially burnt quartz samples were also found toexhibit RTL properties (Hashimoto et al., 1991, 1994;Fattahi and Stokes, 2000, 2003). For the volcanic andartificially burnt quartz, RTL measurements are especiallypreferable for dating over a period of 1 Ma because ofthe negligible contribution of supra-linearity and stabil-ity of the related centers and trapped electrons (Hashimotoet al., 1993, 1999; Miallier et al., 1994).

In luminescence dating methods, ages can basicallybe evaluated by dividing the naturally accumulated doseor the paleo-dose by the corresponding annual radiationdose. Practically, the naturally accumulated doses (orequivalent doses) absorbed after zeroing once by sunlightexposure or thermally burnt procedure are measurable

202 T. Hashimoto et al.

the burnt archaeological materials. Both RTL- and BTL-ages were determined from the relationship between theaccumulated doses (or paleo-doses) and the annual doses,which were estimated from the natural radioactivity ofthe ambient soil of the buried pottery pieces and crushedpottery material. As a result, the RTL-ages are closer tothe predicted archaeological ages, giving more reliableresults than the BTL-ages, which usually give diverse agestowards variable but younger ages (Hashimoto et al.,2003a).

In addition to RTL- and BTL-phenomena, other lumi-nescence sources, including OSL of quartz grains andIRSL from feldspar samples, should be induced from therecombination of released electrons with luminescentholes. Principally, these luminescence intensities must bemutually concordant, if the naturally accumulated lumi-nescence sources are sufficiently stable during periods

elapsed after the firing of pottery or zero-setting process.In this case, luminescence dosimetry or dating would cer-tainly offer the possibility of cross-checking within thesame sample, increasing the reliability of final results.

From these viewpoints, two different optically stimu-lated-luminescence dosimetries were applied to the burntarchaeological pottery-pieces using the SAR-method,which has the correction procedure of sensitivity changes.For the quartz aliquot, the OSL-dosimetry was carriedout using blue LED-light exposure, while the IRSL-dosimetry for the feldspar aliquot was conducted usinginfra-red LED-light excitation. The results were comparedwith the RTL- and BTL-dose results for the same potterypieces. Finally, the accumulated doses from RTL-meas-urement were successfully applied to the age-determina-tion by using the annual doses from both pottery materi-als and ambient soils.

Fig. 1. Sampling locations of Jomon pottery pieces in Okumiomote ruin sites, Niigata, Japan.

Luminescence dosimetry and dating 203

EXPERIMENTAL

Sample pieces of potteryNine pottery pieces from the Okumiomote ruin sites,

situated in the northern district of Niigata, Japan, wereused in this study: Achiyadaira, Numanosawa, Shimozori,Miyasori, and Motoyashiki areas. The sample collectingsites are illustrated in Fig. 1, in which locations and sam-ple codes (cf., Table 2) of nine pottery pieces are indi-cated.

The pottery samples were produced about 3,500–6,000years ago. Covering this time period, so-called Jomonperiod, almost all pottery surfaces were decorated withsome rope-shaped patterns, hence why they are calledJoumon (rope-pattern decorated) pottery. Soil samples ofthe sites, where the pottery pieces were excavated, werealso collected to determine the annual doses from naturalradioactivities. Three typical pieces of the pottery arepresented in Fig. 2.

All procedures for the isolation of coarse quartz andfeldspar grain aliquots have been done under red light toeliminate bleaching effects as much as possible. The sam-ples were gently crushed in an agate mortar. Fine soil orclay constituents were washed out with tap-water leav-ing a residue of quartz and feldspar grains. Then all grainaliquots were treated with 6N HCl solution. After an etch-ing procedure using concentrated HF for 6 hours to re-move surface layer, the surfaces were washed thoroughlywith water. After drying, heavy liquid separation withsodium polytungstate solution was applied to purify thequartz fraction. After mixing well the residue with a heavyliquid solution (specific gravity of 2.63), feldspar grainaliquots were collected as floating fractions on the sur-

face of heavy liquid solution. On the other hand, quartzfractions are isolated as another floating fractions of feld-spar-removing residue by applying a different heavy liq-uid solution (having specific gravity of 2.67). As well-known, feldspar fractions are the major constituent amongwhite minerals and possess more intense luminescenceover quartz grains. Finally, both quartz and feldspar grainswere sieved into 150–250 µm-sized particles.

Thermoluminescence- and IRSL-color imaging and on-line TL-spectrometry

The color characteristics of luminescence from quartzand feldspar extracts were examined with qualitative TLand infra-red (IR) light-stimulated color image (IRSL-CI) techniques, respectively.

The TL color image (TLCI) was taken in the tempera-ture range of 80 to 410°C. The IRSL-CI procedure hasbeen applied to the present study as described in the pre-vious papers (Hashimoto et al., 2002b, 2003b). Since fur-ther quantitative information for the detective wave-lengths is required, on-line TL-spectrometry has beenapplied for the grain samples using a highly sensitivespectrometric system that included an image-intensifierphoto-diode array (IPDA) detector (Hashimoto et al.,1997).

Based on both TLCI and TL-spectrometric results, thequartz extracts from burnt pottery pieces were confirmedto possess the RTL-property (seen in Fig. 6) having abroad peak of 620 nm in wavelength around 330°C.

Luminescence measurementsLuminescence-measuring system All of the luminescencemeasurements were carried out using a new automated

Fig. 2. Three typical Jomon pottery pieces. Upper and bottom photos shows outer surfaces decorated with rope-shape patternsand internal surface, respectively. Predicted ages are indicated in parantheses.


TL/OSL-reader system, which has been developed for theSAR-method (Hashimoto et al., 2002a, c). This systemwas especially focused on the RTL-measurements fromminerals, which have been recently expected to provideboth reliable luminescence dosimetry and dating fromboth quartz and feldspar grains. In the SAR-method, bothevery luminescence measurement and artificial irradia-tion to the same sample aliquot are required to carry outunder the same geometrical conditions (Wintle and

Murray, 2000). For this purpose and high counting effi-ciency, we used a light guide made of a core rod-typeglass pipe (68 mm length, 11 mm diameter, manufacturedby Nissei Denki Co. Ltd.) inserted between a sample ves-sel and a PMT-surface. Artificial irradiation with the de-sirable doses was attained using a small X-ray irradiator(Varian, VF-50J tube with W-target, 50 kV, 1 mA, 50 Wat maximum operation). To eliminate lower X-ray energyconstituents, an Al-absorber of 50 µm thickness was ap-propriately placed between X-ray generator and targetmineral grains (Yawata and Hashimoto, 2004a). The ad-vantages of this X-ray irradiator include the availabilityof variable dose settings by adjusting applied voltages,currents of the tube, irradiation distance, Al-absorberthickness, simple radiation protection without heavyshielding material, and uniformity of irradiated areas con-firmed also by Andersen et al. (2003). In practical use,the applied voltage and current for the X-ray tube were50 kV and 0.15 mA, respectively. Dose-luminescenceresponses of the X-ray irradiator have been verified byusing the Al2O3:C dosimeter and simple range-calcula-tion within quartz combined with RTL-measurements(Andersen et al., 2003; Hashimoto et al., 2002c; Yawataand Hashimoto, 2004a).

To minimize the influence of black-body radiation athigher temperature beyond 280°C, the small heater areawas made using four pieces of ceramic heater combinedinto one stack (total heating power of 32 W × 4). Thepreheating condition after X-ray irradiation was appliedfor 10 sec at 200°C for RTL-measurements. The mass ofthe measuring aliquot was always fixed at 5.0 mg and

100

80

60

40

20

0

tran

smitt

ance

%T

1000800600400200

wavelength / nm

BTL(B-390)

RTL(R-60 + IR-cut)

OSL(DUG-11)

Blue-LED

GG-420100

80

60

40

20

0

tran

smitt

ance

%T

1000800600400200

wavelength / nm

IRSL(CF-50E)

IR-LED100

80

60

40

20

0

tran

smitt

ance

%T

1000800600400200

wavelength / nm

(a)

(b)

(c)

Fig. 3. Optical transmission-properties of filter combinationsfor the luminescence detection and stimulating light nature(black shapes) from LED-sources. BTL, RTL, OSL, and IRSLmean detection windows of respective luminescence when theoptical filters were employed. (a) RTL- and BTL-detection wave-length regions for quartz extracts, (b) optical properties of OSL-stimulating blue-LED and OSL-detective regions for quartzextracts (GG-420 filter was installed in front of LED-light out-let.), (c) optical properties of IRSL-stimulating LED and lumi-nescence-detection regions for feldspar extracts.

Fig. 4. OSL-stimulation and luminescence-detection compo-nent from quartz aliquot. A light guide, sixteen blue-LEDs, andfilter combinations of GG-420 and DUG-11 (cf., Fig. 3(b)) wereemployed.


three or two aliquots from one sample were subjected tothe same measurements.

The practical RTL-measurements were performed us-ing a photomultiplier tube (PMT, Hamamatsu, R-649S)with a multi-alkali (Na-K-Sb-Cs) photon detection part,associated with a combination of a red-glass filter(Toshiba, R-60) and an infrared-cut filter (Eagle) to re-duce blackbody radiation as shown in Fig. 3(a)(Hashimoto et al., 2003a).

The BTL (ranging of 350–500 nm), which has beenused traditionally for the TL-dating of quartz extractssince early TL-dating work, was also measured for thecomparison of accumulated doses from the same potterypieces (Hashimoto et al., 2003a).OSL- and IRSL-measurements Sixteen blue light emit-ting diodes (blue-LED, Nichia Chemical Industries Ltd.,NSPB-500S) were employed for the stimulation of OSLfrom quartz grains. The diode has an emission peak at470 nm with 20 nm FWHM value. The optical propertiesof blue-LED for excitation and detective OSL-wavelengthregions are illustrated along with filter combinations inFig. 3(b). A schematic view of the OSL-stimulation anddetective parts is indicated in Fig. 4. The stimulation light

from the blue LEDs was completely removed using thecombinations of suitable optical filters prior to the pureOSL-detection with the PMT. For the stimulating LED-light, the short-wavelength regions below 400 nm wereremoved by inserting a GG-420 (Schott) filter in front ofthe blue-LEDs. The rejection of stimulation light into thePMT-detector was ensured by means of a DUG-11(Schott) filter setting just infront of the PMT. In the OSL/SAR (single-aliquot regenerative-dose) method, a 10 sec-ond preheating phase of 250°C and a test dose of 4.2 Gywere applied. The OSL-decay curves were obtained byrecording every 0.1 sec during 100 sec blue light-illumi-nation. The stimulating power of OSL at the sample posi-tion was estimated to be 17.5 mW/cm2.

For the IRSL investigation of feldspar fractions, six-teen IR-LED (infra-red emitting diodes, HamamatsuPhotonics, L2690-02) were used as a stimulating lightsource. The diode has 890 nm emission peak with 50 nmFWHM. The optical property related to IRSL-measure-ment is shown as IR-stimulation and IRSL-detectivewavelength-regions in Fig. 3(c). These IR-LEDs wereincorporated in the LED-holder similar manner to theblue-LEDs as seen in Fig. 4. Stimulating light was re-

Table 1. Radionuclide concentrations in pottery pieces and related soils from γ-ray spectrometryand annual dose results evaluated for each pottery piece

Not determined (n.d.).Annual doses are corrected with water content.In calculation of total annual doses, the effective β and γ contribution are included from the values of pottery piece and soil, respectively.*1Dose of soils is assumed to be the dose of 3.81 mGy/a from Achiyadaira close to these sites.*2Annual doses consist of both β- and γ-ray contribution.*3These annual doses are derived only from γ-ray potion.

Samples U Th K2O Water content Annual dose(ppm) (ppm) (%) (%) (mGy/a)

AchiyadairaA-5 1.67 +/– 0.09 10.1 +/– 0.36 2.09 +/– 0.04 3.47 +/– 0.11B-8 n.d. 3.25 +/– 0.27 0.43 +/– 0.02 1.80 +/– 0.08B-10 0.92 +/– 0.11 3.87 +/– 0.31 0.61 +/– 0.03 2.18 +/– 0.09B-12 2.48 +/– 0.11 11.6 +/– 0.51 2.19 +/– 0.05 3.69 +/– 0.12Soil 1.88 +/– 0.14 10.9 +/– 0.56 1.86 +/– 0.07 4.22 (2.96 +/– 0.13)*2

(1.08 +/– 0.08)*3

Numanosawa*1

A-8 1.76 +/– 0.08 10.9 +/– 0.31 1.74 +/– 0.03 3.27 +/– 0.11A-14 1.62 +/– 0.11 10.5 +/– 0.41 1.64 +/– 0.04 3.17 +/– 0.11

Shimozori*1

A-18 1.73 +/– 0.09 9.63 +/– 0.41 2.05 +/– 0.04 3.44 +/– 0.11

MiyasoriA-23 1.11 +/– 0.11 8.52 +/– 0.42 1.99 +/– 0.05 3.01 +/– 0.11Soil 0.91 +/– 0.13 9.16 +/– 0.56 2.75 +/– 0.06 4.72 (3.33 +/– 0.12)*2

(1.07 +/– 0.08)*3

MotoyashikiB-18 2.19 +/– 0.12 6.34 +/– 0.48 1.86 +/– 0.05 3.17 +/– 0.13Soil 1.59 +/– 0.14 9.92 +/– 0.67 2.93 +/– 0.07 4.31 (3.70 +/– 0.14)*2

(1.21 +/– 0.09)*3


moved using a visible light filter, CF-50E (Asahi TechnoGlass) in front of the PMT (Hamamatsu, R-585S). In theIRSL/SAR method, a 10 second preheating phase of250°C and test dose of 4.2 Gy were applied. The lumi-nescence from feldspar grains was recorded every 0.1 secduring 100 sec-IR stimulation at 125°C. The stimulationpower of IRSL was estimated to be 28 mW/cm2 at thesample position.

Each luminescence/dose response curve comprisedfive regenerative doses, although two below figures ofthe response curve consist of four experimental points,without the result from 17 Gy irradiation.

Estimation of annual doseThe effects of water content in the ambient soil must

be corrected prior to the evaluation of the annual dose,which is derived from naturally occurring radioactivityof dried-soil material. To estimate water contents, theoriginal soil samples were dried on a hotplate for 48 h at100°C. The dried soil and crushed pottery samples werepacked into a cylindrical plastic container (U-8) andanalyzed by γ-ray spectrometry using a germanium co-axial p-type detector connected with a multi-channelanalyzer (EGPC 120-210-R, EURISYS Measures). Byassuming a homogeneous distribution, one can calculatenatural annual doses from the radionuclide concentrationsin the radioactive equilibrium within uranium and tho-rium decay chains (Aitken, 1985). The potassium, ura-

nium, and thorium concentrations were determined by thefollowing photo-peaks; 1460 keV for 40K and 609 keV of214Bi for the U-series, and 583 keV of 208Tl for the Th-series, respectively. Detection efficiency has been cali-brated by using a standard γ-ray source (Amersham,MXV-01), consisting of mixture of gelatine and 9 kindsof known radionuclides (Hashimoto et al., 2003a). Theconcentrations of radioactive materials are presented inTable 1.

RESULTS AND DISCUSSION

Luminescence color images and TL-spectrometric resultsFour typical TLCIs and corresponding IRSL-CIs are

demonstrated in Fig. 5. It is obvious that all TL fromquartz extracts show red-colored patterns while the IRSL-CIs give blue colored patterns as expected from our pre-vious results (Hashimoto et al., 1994, 1995, 2002b,2003a).

Two typical contour maps of thermoluminescencefrom quartz aliquots are illustrated in Figs. 6(a) and (b).A typical single-RTL peak in the red-spectral region, con-sisting of 620–630 nm in wavelength and around 330°C,could be recognized in both contour maps. All other quartzextracts from the present pottery pieces showed similarRTL-properties, in which an intense peak appears around330–360°C in the red wavelength region of 600–650 nm.These RTL-properties of quartz extracts were concord-

Fig. 5. Thermoluminescence color images (TLCIs) (a) of quartz extracts and IRSL color images (IRSL-CIs) (b) of feldsparextracts from Jomon pottery pieces. Both quartz and feldspar grains were exposed with X-ray of 2.7 kGy. The quartz aliquotswere photographed from room temperature up to 410°C. The IRSL-CIs were obtained under stimulation of IR-LED for 90 sec.


ant with the results of TL color images (TLCI) photo-graphed by a color-sensitive film, although a few quartzextracts showed blue grain components, probably due tofeldspar contamination as seen in Fig. 5(a). The presentresult has also ascertained that quartz extracts from otherburnt relics or pottery pieces gave RTL-properties with-out almost any exception. The RTL-properties of mostquartz grains might be considered to originate from theeffects of archaeological firing. However, there still re-mains some possibility of contamination of the originalRTL quartz grains, which were provided from a volcaniclayer into clay minerals as raw material for the pottery,because the clay products are probably derived from aweathered volcanic layer. These RTL-results are remark-ably concordant with the experimental evidence that thequartz slices fired beyond 900°C can change from BTLinto RTL in addition to the intrinsic RTL-nature of quartzgrains from the volcanic ash layers (Hashimoto et al.,1996).

Luminescence measurementsRTL-measurements The growth tendency of both RTL-glowcurves by applying the SAR-method are shown inFig. 7(a) for quartz extracts from the pottery piece ofShimozori. In the RTL-glowcurves, there are two intensepeaks, around 230°C and 350°C, although some otherpeaks appear between them. The low temperature-peakswere induced by artificial X-ray irradiation, because natu-ral TL (NTL)-glowcurve never show prominent peaks inthis region. On the other hand, the natural RTL-glowcurveoffered a broad peak in the high temperature region be-yond 300°C. The integrating temperature-ranges werefixed here from 300°C to 410°C by using plateau tests asseen in Fig. 7(b); the ratios of ATL (artificial TL)- to NTL-glowcurve are plotted against heating temperatures. TheRTL-values integrated over plateau regions as a functionof regenerative doses brought on the corresponding re-sponse curves as shown in Fig. 7(c). It should be empha-

400

350

300

250

200

150

100

700650600550500

Tem

pera

ture

/ o C

Wavelength / nm

Wavelength / nm

Tem

pera

ture

/°C

(a)

(b)

400

350

300

250

200

150

100

700650600550500

Fig. 6 . Contour maps of artificially induced TL from quartzextracts from pottery pieces. The quartz extracts of about 10mg were exposed to X-ray dose of 2.7 kGy: (a) A-18 potterypiece from Shimozori site, (b) B-8 pottery piece fromAchiyadaira site. The blue and violet regions (350–500 nm)are excluded because of contamination of feldspar constituentsin quartz aliquots (Hashimoto et al., 2003a).

Fig. 7. Experimental RTL-results for quartz extract from A-18pottery piece from Shimozori site. (a) Glowcurve changes byapplying regenerative doses with X-ray irradiator. (b) Plateaucurves for the determination of the integrating regions. (c) Dose-response curve using RTL integrated of 300–410°C regions ofglowcurves.

3500

3000

2500

2000

1500

1000

500

0

45040035030025020015010050

33.6 Gy16.8 Gynatural8.39 Gy

RT

Lin

tens

ity /

coun

ts

300x10 3

250

200

150

100

50

0

4035302520151050Inte

grat

edR

TL

inte

nsity

/ co

unts

AT

L/ N

TL

Temperature / °C

Temperature / °C

Regenerative dose /Gy

(a)

(b)

(c)

NTLATL

33.6 Gy

16.8 Gy

8.39 Gy

Integrating range4

3

2

1

0

420400380360340320300


sized that each response curve of RTL from SAR-methodgave different gradients from aliquot to aliquot. Thismeans that the RTL-sensitivity properties reflect differ-ently in each quartz grain as well as in each aliquot. Indose-response curves, the curves often give supra-lineartendency over regenerative doses of 10–15 Gy in manycases, so that almost all growth curves were fitted usinga polynomial curve.

The naturally accumulated dose was estimated frominterpolation of the naturally originating RTL-value intothe dose-response curve.OSL- and IRSL-measurements The OSL-decay curves ofthe same quartz extract from sample A-18 are demon-

strated in Fig. 8(a). This OSL-decay behavior was indi-cated for only 10 sec, although the illumination of theLED continued for 102 sec. As seen in Fig. 8(a), the OSL-decays show the strongest luminescence just after the startof illumination, and then they follow an exponential-likedecay. Since OSL-intensities at the starting points areknown to be directly proportional to the absorbed doses,the OSL-intensities were obtained from integration ofdecay curve for the first 1 sec-portion, corresponding tothe rapid decay component. Each OSL-intensity was ob-tained by subtracting a background contribution assign-able to the last 1 sec-part of the decay-curve. The OSL-intensities greatly reflect the sensitivity changes duringOSL-measurements and irradiation cycles, so that the

60x10 3

50

40

30

20

10

0

14121086420

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

4321

Measurement cycle

6000

5000

4000

3000

2000

1000

0

109876543210

Stimulated time / sec

B-O

SLin

tens

ity /

coun

tsIn

tegr

ated

B-O

SL in

tens

ity /

coun

tsC

orre

ctio

n fa

ctor



(a)

(b)

(c) Measurement cycle

12.6 Gy8.39 Gynatural4.20 Gy

naturalexperimental datacorrected data

140

120

100

80

60

40

20

0

50454035302520151050

IRSL

inte

nsity

/ co

unts

Inte

grat

ed I

RSL

inte

nsity

/ co

unts

Cor

rect

ion

fact

or



(a)

(b)

(c)Measurement cycle

naturalexperimental datacorrected data

12.6 Gy8.39 Gy4.20 Gynatural

2.0

1.5

1.0

0.5

0.0

4321

1600

1400

1200

1000

800

600

400

200

0

14121086420

Fig. 8. Experimental OSL-results for quartz extract from A-18pottery piece from Shimozori site. (a) Decay curves accompa-nied with stimulation of blue light emitting diodes (LEDs) toquartz extract. (b) Correction factors (repeated OSL/first OSLafter test dose irradiations) based on sensitivity changes withexperimental cycles of measurement and irradiation. (c) SARdose-response curve of OSL-intensities integrated for 1 sec fromillumination start using decay curves.

Fig. 9. Experimental IRSL-results for quartz extract from A-18pottery piece from Shimozori site. (a) Decay curves accompa-nied with stimulation of IR on feldspar extract. (b) Changes ofcorrection factors with experimental cycles of measurement andirradiation. (c) SAR dose-response curve of IRSL-intensitiesintegrated for 1 sec from illumination start using decay curves.


OSL-sensitivity changes were corrected by the interpo-lating OSL-measurements after test-dose irradiation of4.20 Gy between increments of regenerative doses. Typi-cal sensitivity-changes are shown in Fig. 8(b). In this case,increasing trends of OSL-sensitivity were recognized withincrements of repeated cycles, because the correction fac-tors were derived from luminescence ratios of first cycleto recycled measurements.

The integrated OSL-intensities were plotted as a func-tion of regenerative doses as demonstrated in Fig. 8(c).The OSL-response curves were obtained from the opencircles by correcting the sensitivity changes. The naturalaccumulated doses were evaluated by inserting the natu-rally accumulated OSL-intensity to the response curve.

In the IRSL-measurements, all OSL experimental pro-cedures were adhered to with the exception of the use ofthe feldspar aliquot and different stimulation and detec-tion wavelength regions. Typical decay curves are seenin Fig. 9(a). It is evident that the intensities of IRSL areappreciably lower than the OSL-intensities (cf., Fig. 8(a))at the same adsorbed doses, although there are apparentdifferences between luminescent minerals; the OSL isderived from quartz grains while the IRSL occurs fromfeldspar grains. Concerning the sensitivity changes, thereappears no remarkable difference as seen in Fig. 9(b). Theintegration region of the IRSL was the first 1 sec-compo-nent.

As a result, the IRSL-response curve to doses is de-picted in Fig. 9(c), in which naturally occurring IRSL wasinterpolated to this line for the sake of accumulated doseestimation.

Comparison of evaluated dose-resultsAll accumulated dose-results from RTL, OSL and

IRSL, in addition to BTL, are summarized together withRTL-dating and related annual doses in Table 2.

The experimental errors of equivalent doses are de-rived from two to three aliquot analyses. Of the accumu-lated doses, all of the RTL-values are above 11.6 Gy andup to 26.0 Gy. In principal, if there is no escaping of lu-minescence sources during buried period, the accumu-lated dose evaluations from quartz and feldspar extractsshould agree with each other because they are subjectedto the same firing-process in each burnt pottery. How-ever, the estimated OSL- and IRSL-accumulated dosesusually give lower values than the RTL-ones. These re-sults mean that the optically stimulated luminescencesources are liable to suffer some loss during storage pe-riod. Specifically, the lower accumulated doses of IRSL-intensities could be explained by well-known anomalousfading phenomena which occur in the case of the feld-spar grains (Lamothe et al., 2003). For the annual-dose,the feldspar IRSL value should become higher than thatof the associated quartz. In other words, the internal ex-

posure should be considered due to the existence of ra-dioactive elements in feldspar grains themselves, whereasthe external exposure alone would affect the quartz frac-tion which contains negligible small amounts of radioac-tive elements.

In the case of the OSL-measurements, the accumu-lated doses also gave lower values down to undetectedlevels than those from the RTL-measurement, despitebeing measured on the same quartz extracts. These lowerdoses could be attributed to relatively easy-bleaching orunstable luminescence sources related to the OSL duringthe buried period of pottery pieces.

Comparing the BTL-evaluation, the higher accumu-lated values from RTL have been already described in anearlier paper (Hashimoto et al., 2003a). As mentioned inthat article, the BTL-glowcurves always showed broadpeaks below 300°C, which are probably attributed to con-comitant feldspar grains. It is now well-known that thepurification of the quartz fraction from feldspar grains isextremely difficult and the feldspar fraction has a higherBTL-sensitivity in comparison with quartz grains(Montret et al., 1992). It should be emphasized here thatthe quartz fractions from burnt samples such as potterypieces tend to show the RTL-property, probably owing tothe heating process used during the manufacture of suchmaterial. Therefore, the RTL measurements are consid-ered to be preferable for accumulate dose-evaluation.

Evaluation of RTL-agesThe annual doses were evaluated from contents of ra-

dioactive nuclide contents in both crushed pottery piecesand surrounding soil by correcting for the water content(Aitken, 1985). The data related to the annual dose evalu-ation are summarized in Table 1. Since the thickness ofpottery pieces are less than 20 mm, the radioactivitiesfrom crushed pottery were utilized for the dose-evalua-tion due to γ-ray contributions and those from ambientsoil for β-ray contributions. The contribution of cosmicrays was assumed to be 0.18 mGy/a. The annual dosesevaluated thus were utilized to calculate the RTL-ages.

As seen in Table 2, the RTL-dating results are in goodagreement between RTL-ages and the ones predicted fromthe decorated rope-modes of pottery surface within ex-perimental errors. In an advanced RTL-application, theauthors are intending to improve the RTL-measuring con-ditions aiming towards the reduction of experimental er-rors, using the application of a single grain method(Yawata and Hashimoto, 2004b).

In comparison with the RTL-evaluation, it is empha-sized here that lower accumulated doses from OSL andIRSL, as well as BTL, could cause the estimation of er-roneously younger ages for burnt archaeological materi-als. The dominant presence of RTL-quartz has been rec-ognized not only in archaeologically burnt pottery, but


Not

det

erm

i ned

(n.

d.).

*1 B.P

.: b

efor

e pr

esen

t (2

000)

.*2 H

ashi

mot

o et

al .

(20

03a)

.

Tabl

e 2.

C

ompa

riso

n of

nat

ural

ly a

c cum

ulat

e d d

ose s

usi

ng R

TL

, B

TL

, an

d O

SL f

rom

qua

rtz

e xtr

acts

and

IR

SL f

rom

fe l

dspa

r e x

trac

tsc o

ntai

ned

in a

rcha

e olo

gic a

lly

burn

t po

tte r

y pi

e ce s

, an

d th

e ir

RT

L-a

ges

Pot

tery

pie

ces

Mod

e of

Jom

on p

otte

ryor

Jom

on s

tage

Acc

umul

ated

dos

es (

Gy)

Ann

ual

dose

RT

L a

ge*2

Pre

dict

ed a

ge

RT

LB

TL

OSL

IRSL

(mG

y/a )

(yea

rs B

.P.*

1 )(y

ears

B.P

.)

Ac h

iya d

a ira

A-5

Min

amis

anju

inab

a13

.1 +

/− 4

.06.

3 +

/− 1

.411

.3 +

/− 2

.00.

7 +

/− 0

.23.

47 +

/− 0

.11

3800

+/−

115

040

00

B-8

late

r pe

riod

of

earl

y st

a ge

11.6

+/−

3.3

4.4

+/−

2.2

5.6

+/−

2.2

n.d.

1.80

+/−

0.0

864

00 +

/− 1

860

6000

B-1

0m

iddl

e pe

riod

of

earl

y st

age

14.8

+/−

3.5

9.3

+/−

0.8

6.3

+/−

0.8

0.2

+/−

0.1

2.18

+/−

0.0

968

00 +

/− 1

640

6000

B-1

2ea

rly

peri

od o

f ea

rly

sta g

e26

.0 +

/− 2

.2n.

d.n.

d.3.

9 +

/− 0

.13.

69 +

/− 0

.12

7000

+/−

630

6000

Num

anos

a wa

A-8

Shin

bo17

.3n.

d.15

.20.

923.

27 +

/− 0

.11

5300

+/−

180

5000

A-1

4K

asor

iB3

14.7

+/−

2.1

13.3

n.d.

2.8

+/−

0.4

3.17

+/−

0.1

146

00 +

/− 6

7035

00

Shim

ozor

iA

-18

late

r pe

riod

of

late

sta

ge15

.0 +

/− 2

.41.

476.

9 +

/− 2

.02.

0 +

/− 0

.23.

44 +

/− 0

.11

4400

+/−

710

5000

Mot

oyas

hiki

B-1

8la

ter

peri

od o

f la

te s

tage

12.2

+/−

1.0

n.d.

10.5

+/−

2.5

0.13

3.17

+/−

0.1

338

00 +

/− 3

6035

00

Miy

asor

iA

-23

Han

adum

ikas

ou22

.7 +

/− 2

.9n.

d.11

.8 +

/ − 0

.97.

1 +

/ − 1

.73.

01 +

/ − 0

.11

7500

+/ −

100

060

00


also in white minerals of widespread volcanic-related lay-ers. Consequently, the RTL-measurements from quartzgrains have been verified to become potentially useful inpaleo-dose evaluation or retrospective dosimetry. In otherwords, they are preferable for the dating of thermal re-setting events, such as archaeological burning of pottery,porcelain and burnt stone and soil as well as volcanic erup-tions (Montret et al., 1992).

CONCLUSIONS

Four kinds of luminescence measurements involvingRTL, BTL, and OSL for quartz grains and IRSL for feld-spar grains are simultaneously applied to the estimationof the naturally accumulated doses for the quartz extractsfrom archaeologically burnt pottery pieces by applyingthe SAR-method. The natural doses accumulated after thelast firing of pottery manufacture showed significant disa-greement among three kinds of luminescence measure-ments. The highest accumulated doses of all samples wereevaluated from the RTL-measurements of quartz extractsby applying the SAR-method. On the other hand, the OSLand IRSL-measurements, including BTL-ones, gave loweraccumulated doses regardless of the correction for sensi-tivity changes during repeated measurement, X-ray irra-diation, and preheating; this correction has been done bynormalizing regenerative-dosed luminescence intensity toone at a fixed irradiation. From these results, the RTL-measurement of quartz fraction was found to be prefer-able for the dosimetry of either naturally accumulatedradiation or retrospective dose.

The ages evaluated from RTL-dose and annual doseresults are in excellent agreement with the expected ages.Therefore, the RTL-dating will be recommended forquartz extracts from archaeologically burnt pottery. Inconclusion, the accurate TL-dating of archaeologicallyburnt materials should be carried out after judging whetherTL-properties of quartz extracts are RTL or BTL.

Acknowledgments—The authors would like to express theirappreciation to Mr. Yasuo Takahashi, a chief director of theAsahi-machi Culture Research Center, Niigata, Japan, for pro-viding archaeological pottery pieces and for valuable discus-sions. The present study was supported by a foundation fromTokyo Electric Power Co. Ltd. and a grant-in-aid for Founda-tion Research from the Ministry of Education, Science, Cul-ture, and Sport, Japan (No. 14340231). They wish to thank Dr.Fattahi, Oxford University, for critical review of the manuscript.

REFERENCES

Aitken, M. J. (1985) Thermoluminescence Dating. AcademicPress, London.

Aitken, M. J. (1998) An Introduction to Optical Dating: TheDating of Quaternary Sediments by the Use of Photo-Stimu-

lated Luminescence. Oxford University Press, Oxford.Andersen C. E., Boetter-Jensen L. and Murray A. S. (2003) A

mini X-ray generator as an alternative to a 90Sr/90Y betasource in luminescence dating. Radiat. Meas. 37, 557–561.

Fattahi, M. and Stokes, S. (2000) Extending the time range ofluminescence dating using red TL (RTL) from volcanicquartz. Radiat. Meas. 32, 479–485.

Fattahi, M. and Stokes, S. (2003) Dating volcanic and relatedsediments by luminescence methods: a Review. Earth-Sci.Rev. 1280, 1–36.

Hashimoto, T., Koyanagi, A., Yokosaka, K. and Sotobayashi,Y. (1986) Thermoluminescence color images from quartzof beach sands. Geochem. J. 20, 111–118.

Hashimoto, T., Yokosaka, K. and Habuki, H. (1987) Emissionproperties of thermoluminescence from natural quartz-blueand red TL response to absorbed dose. Nucl. Tracks Radiat.Meas. 13, 57–66.

Hashimoto, T., Sakai, T., Shirai, N., Sakaue, S. and Kojima, M.(1991) Thermoluminescent spectrum changes of naturalquartz dependent on annealing treatment and aluminum-contents. Anal. Sci. 7, 687–690.

Hashimoto, T., Kojima, M., Shirai, N. and Ichino, M. (1993)Activation energies from blue and red-thermoluminescence(TL) of quartz grains and mean lives of trapped electronsrelated to natural red-TL. Nucl. Tracks Radiat. Meas. 21,217–223.

Hashimoto, T., Sakaue, S., Aoki, H. and Ichino, M. (1994) De-pendence of TL-property changes of natural quartzes onaluminium contents accompanied by thermal annealing.Radiat. Meas. 23, 293–299.

Hashimoto, T., Notoya, S., Ojima, T. and Hoteida, M. (1995)Optically stimulated luminescence (OSL) and some otherluminescence images from granite samples exposed withradiations. Radiat. Meas. 24, 227–237.

Hashimoto, T., Notoya, S., Arimura, T. and Konishi, M. (1996)Changes in luminescence colour images from quartz sliceswith thermal annealing treatments. Radiat. Meas. 26, 233–242.

Hashimoto, T., Sugai, N., Sakaue, H., Yasuda, K. and Shirai,N. (1997) Thermoluminescence (TL) spectra from quartzgrains using on-line TL-spectrometric system. Geochem. J.31, 189–201.

Hashimoto, T., Fujita, H. and Yasuda, K. (1999) Measn-lifeevaluation of naturally trapped electrons associated withred- and blue-thermoluminescent quartz-grains from dunesand. Radioisotopes 48, 673–682 (in Japanese).

Hashimoto, T., Nakagawa, T., Hong, D. G. and Takano, M.(2002a) An automated system for both red/bluethermoluminescence and optically stimulated luminescencemeasurement. J. Nucl. Sci. Technol. 39, 108–109.

Hashimoto, T, Usuda, H. and Yawata, T. (2002b) Color Imagesof Infrared Stimulated Luminescence (IRSL) from GraniteSlices Exposed with Radiations. Ancient TL 20, 19–23.

Hashimoto, T., Nakagawa, T., Usuda, H. and Yawata, T. (2002c)Development of an automated system equpped with a smallX-ray irradiator for red/blue thermoluminescence and opti-cally stimulated luminescence measurement from naturalminerals. BUNSEKI KAGAKU 51, 625–632 (in Japanese).

Hashimoto, T., Yawata, T. and Takano, M., (2003a). Preferable


use of red-thermoluminescence (RTL)-dating for quartzextracts from archaeologically burnt pottery—comparisonof RTL and BTL (blue-TL) measurements using single-aliquot regenerative-dose (SAR) method. Ancient TL 21, 85–91.

Hashimoto, T., Usuda, H., Mitamura, N. and Yawata, T. (2003b)Imaging and measurement of red-infrared stimulated lumi-nescence (R-IRSL) from feldspar samples. Ancient TL 21,1–6.

Lamothe, M., Auclar, M., Hamzaouuui, C. and Huot, S. (2003)Towards a prediction of long-term anomalous fading of feld-spar IRSL. Radiat. Meas. 37, 493–498.

Miallier, D., Fain, J., Sanzelle, S., Pilleyre, Th., Montret, M.,Soumana, S. and Falgueres, C. (1994) Attempts at datingpumice deposits around 580 ka by use of red TL and ESR

of xenolithic quartz inclusion. Radiat. Meas. 23, 399–404.Montret, M., Miallier, D., Sanzelle, S., Fain, J., Pilleyre, Th.

and Soumana, S. (1992) Tl dating in the Holocene usingred TL from quartz. Ancient TL 10, 33–36.

Wintle, A. G. and Murray, A. S. (2000) Quartz OSL: Effects ofthermal treatment and their relevance to laboratory datingprocedures. Radiat. Meas. 32, 387–400.

Yawata, T. and Hashimoto, T. (2004a) Availability of a smallX-ray tube combined with Al-absorber as a new irradiatorfor luminescence researches. Radioisotopes 53, 207–212 (inJapanese).

Yawata, T. and Hashimoto, T. (2004b) Identification of the vol-canic quartz origins from dune sand using a single grainRTL-measurement. Quart. Sci. Rev. 23, 1183–1186.

Documents

Comparison of naturally accumulated radiation-doses ...svr4.terrapub.co.jp/journals/GJ/pdf/3903/39030201.pdf · In order to evaluate correct accumulated-doses, a sin-gle-aliquot regenerative-dose