Retail

6 pages
7 views

Equivalent thermal history ( H E ) of ferruginous sandstones based on the thermal activation characteristics of quartz

of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
The thermal history of four quartz-rich ochre samples from an Upper Palaeolithic site was studied. This work is based on the changes in thermal activation characteristics (TAC) of the 110∘C TL peak of the quartz inclusions. An isothermal study of a
Transcript
  Radiation Measurements 41 (2006) 995–1000www.elsevier.com/locate/radmeas Equivalentthermalhistory (H  E ) offerruginoussandstonesbasedonthethermalactivationcharacteristicsofquartz Christelle Lahaye a , b , ∗ , Dorothy I. Godfrey-Smith c , d , Pierre Guibert a , Françoise Bechtel a a Centre de Recherche en Physique Appliquée à l’Archéologie (CRPAA), IRAMAT Institut de Recherche sur les Archéomatériaux,UMR 5060 CNRS - Université de Bordeaux 3, Maison de l’Archéologie, 33607 Pessac Cedex, France b  Laboratorio di Datazione tramite Luminescenza e di Metodologie Fisiche per i Beni Culturali, Dipartimento di Fisica ed Astronomia dell’Università diCatania, Via Santa Sofia, 6495123 Catania, Italy c  Department of Earth Sciences, Dalhousie University, Halifax, NS, Canada B3H 4J1 d  Radiological Analysis and Defence, Defence Research and Development Canada, 3701 Carling Ave., Ottawa, ON, Canada K1A 0Z4 Received 16 August 2005; received in revised form 11 April 2006; accepted 17 April 2006 Abstract The thermal history of four quartz-rich ochre samples from an Upper Palaeolithic site was studied. This work is based on the changes inthermal activation characteristics (TAC) of the 110 ◦ C TL peak of the quartz inclusions. An isothermal study of a previously unheated samplehas highlighted the importance of the duration of annealing on the sensitization of quartz. In fact, the sensitivity change as a function of theduration of annealing is not monotonic. For that reason it seems necessary to consider the “equivalent thermal history”  H  E  rather than an“equivalent temperature”. Isochronal annealing experiments demonstrate that the initial rise of sensitization overestimates the true  H  E  by about100 ◦ C. Using a geological sample we have thus developed an empirical approach which allows the true  H  E  of artifacts to be determined.Reheating of ochre srcinally heated in antiquity results in desensitization of the TAC.© 2006 Elsevier Ltd. All rights reserved. Keywords:  Thermoluminescence; Thermal activation characteristic (TAC); Ochre; Upper Palaeolithic; Quartz; Thermal history 1. Introduction The thermal treatment received in antiquity by an artifactis important to both archaeologists and chronologists. Recog-nition of the past exposure to heat (for example, heating of flint to improve knapping characteristics; firing of ceramics  . . . )and whether it was deliberate or incidental allows us to makeimportant interpretations regarding the development of firingtechnology in ancient times (type of oven used; hearths). Im-portantly, it also allows the chronologist to verify that the arti-fact has been sufficiently heated in antiquity to be dated todayusing luminescence methods.In order to correctly interpret a sample’s thermolumines-cence (TL) signature, it is necessary to clarify some concepts ∗ Corresponding author. Laboratorio di Datazione tramite Luminescenzae di Metodologie Fisiche per i Beni Culturali, Dipartimento di Fisica edAstronomia dell’Università di Catania, Via Santa Sofia, 64 I-95123 Catania,Italy. Tel.: +390953785258; fax: +33557124550.  E-mail address:  christelle.lahaye@ct.infn.it (C. Lahaye).1350-4487/$-see front matter © 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.radmeas.2006.04.021 relating to its potential past heat exposure. The highest-temperature experienced by an artifact in antiquity, its equiv-alent thermal history, and the thermal cycle which yields asecond glow TL curve identical to the first TL glow curve canall be deduced via analysis, but refer to different concepts andso may lead to different interpretations. Let us explore thesethree key concepts: •  An artifact’s thermal history maximum  (T  max )  is the highesttemperature to which it was exposed in the past, even if thisfiring temperature was reached for only few seconds. It cansometimes be determined analytically, for example by X-raydiffraction, which can identify crystalline phases that existin materials only after they had been annealed at a certaintemperature. •  The equivalent thermal history  (H  E )  is the experimentalannealing protocol, consisting of a specific temperature ex-posure held for a specific duration, which has the same ther-mal effect on a sample as its srcinal archaeological thermal  996  C. Lahaye et al. / Radiation Measurements 41 (2006) 995–1000 treatment or treatments. The archaeological thermal treat-ment is also referred to as “first heating” in this paper. Wenote that the first heating is not merely a function of temper-ature alone, but of temperature, duration, and possibly thenumber of thermal cycles at different temperatures and of varying durations. Since some artifacts, for example hearth-stones, can have been fired numerous times at different tem-peratures and for different durations in antiquity, all theseparameters have their importance. •  The thermal cycle that yields second glow curves as similaras possible to the first glow curves. This is exclusively amethodological notion of interest to chronologists: it is thethermal cycle (second heating, in the laboratory) that willleave the sample in a thermodynamic state (regarding thenumber, type and filling state of the TL trapping centers)closest to that it had after the first heating (archaeologicalfiring). The aim is to get more precise dating results (Roque,2001; Roque et al., 2004).This study aimed to determine the equivalent thermal his-tory  H  E  of three ochre artifacts. Others have also exploredthis notion. Valladas (1981) proposed to determine the past fir-ing temperature (really an equivalent thermal cycle) of sand-stones from prehistoric hearths, using a method based on theappearance of the saturation temperature and a comparisonwith a geological equivalent. Sunta and David (1982) pro-posed a method to determine the maximum firing tempera-ture of ceramics based on the study of the 110 ◦ C TL peak.Watson and Aitken (1985) explored this method by testingthe same protocol on five samples (three native quartz andtwo quartz extracted from clay samples) and demonstrated thatit is not generally valid. Valladas (1983) then proposed, forflints heated in the past at temperatures higher than 400 ◦ C, amethod based on the properties of the 380 ◦ C emission. Göksuet al. (1989) developed a way of determining the temperaturereached in the past by studying the sensitivity variations of the 110 ◦ C peak in flints. Godfrey-Smith et al. (2005) also ap- plied this approach to determine that cherts (flints) from anarchaeological site in Canada could not have been heated inantiquity.Our work exploits the sensitization properties of the 110 ◦ CTL peak of quartz, which have been studied extensively in nat-ural quartz (Zimmerman, 1971; Chen et al., 1988; Yang and McKeever, 1990; Benny et al., 2000, 2002;Adamiec, 2005a,b). By convention the TL peak in question is called the “110 ◦ Cpeak” although its precise position can vary slightly accordingto the source and past treatment of the quartz (Yang and McK-ever, 1990; Charitidis et al., 2000). This peak was chosen by us because it is well known to retain the memory of past ther-mal treatments in quartz (Sunta and David, 1982; Watson and Aitken, 1985; Göksu et al., 1989). 2. Samples and experimental procedures The archaeological material studied here are ferruginoussandstones (ochre) unearthed in the Upper Palaeolithic siteof La Honteyre (Le Tuzan) situated at about 40km south of Bordeaux, France. This site consists of a single Magdale-nian layer which contains, in addition to flints and quartzartifacts, hundreds of yellow and red ferruginous sandstones.Godfrey-Smith and Ilani (2004) recently reported evidencethat red ochre discovered at Qafzeh cave in archaeologicallayers dated to 95ka was formed through heating, in antiquity,of commonly occurring yellow ochre.Oneofthehypothesesproposedtoexplainthepresenceofthenumerous sandstone fragments at La Honteyre is that they mayhave been used as hearthstones. Their composition is 35–65%by mass of an iron oxide matrix which binds quartz inclusions.Four samples were selected: three of them are red (Bdx 7339,Bdx7340andBdx7341),whilethefourthisyellow(Bdx7348).Preliminary TL analysis (Lahaye and Guibert, 2002) concluded that the three red samples were fired in antiquity and that theircolor is due to a hematite matrix. The fourth sample, which hasa goethite matrix, has never been fired in the past. We thereforechose this sample to serve as a geological equivalent for thethree previously heated archaeological samples.The samples were prepared by grinding off the outer 2mm,and crushing plus sieving the interior portion. Successive HCland HF acid treatments allowed disaggregation and removal of the iron oxide matrices. Cleaned quartz grains were sieved toobtain a 75  .  300  m fraction. These grains were mounted on9.8mm diameter stainless-steel discs with silicone oil.All analyses were performed in the TOSL Research Labora-tory at Dalhousie University, Halifax, Canada. Annealing ex-periments utilized an electronically controlled muffle furnace,under a normal atmosphere. TL measurements were carriedout with an automated Ris Z  TL-DA-12 reader equipped witha  90 Sr/  90 Y beta source, an EMI 9325Q bialkali photomulti-plier, and a Corning type 7-59 plus Schott BG39 glass filterswith a blue-violet transmittance band. In order to standard-ize and minimize the irradiation-glowout delay, TL readoutswere recorded immediately after beta irradiations. Each mea-surement was made on at least two aliquots. A high degree of inter-aliquot reproducibility was noted.We based our investigations on differences between thethermal activation characteristics (TAC) of the four samples.Thermal activation is the enhancement or reduction of sen-sitivity  S  of the 110 ◦ C peak of quartz due to the combinedeffects of prior heating and irradiation (Sunta and David,1982; Valladas, 1983; Aitken, 1985; Bailiff, 1994; Adamiec, 2005a). We adopted the following procedure for all TACmeasurements:(1) Beta irradiation with a test dose of 0.04Gy;(2) TL readout up to 160 ◦ C, at 5 ◦ Cs − 1 , to obtain  S  i ;(3) heating to  T  i  (activation temperature) at 5 ◦ Cs − 1 (200 ◦ C  T  i  625 ◦ C, in 25 ◦ C increments), with a holdat  T  i  for 20s;(4) return to step (1) (same aliquot).The sensitivity  S  i  of the 110 ◦ C peak was determined basedon a 70 ◦ integral centered at the maximum peak intensity. Thefirst cycle, prior to any heating, yields  S  o , while subsequentones yield  S  i , corresponding to  T  i  of step 3 above.  C. Lahaye et al. / Radiation Measurements 41 (2006) 995–1000  997 3. Experimental results and discussion The TACs of the three heated artifacts were measured, andtheir thermal history  H  E  deduced by comparison with the TACof the geological equivalent. Using isothermal and isochronalexperiments of the unheated, geological-equivalent ochre, wededuced systematic variations in its TAC. Finally, the effectsof additional laboratory annealing cycles on the TACs of theartifacts were studied. 3.1. Isothermal study of the geological equivalent  In the isothermal experiment, quartz from the yellowgeological-equivalent ochre (Bdx 7348) was heated at 400 ◦ Cfor different durations, ranging from 15min to 4h. The result-ing TAC properties, as well as its TAC prior to any annealing,are shown in Fig. 1.We expect that quartz which has not previously been heatedwill have a strong increase of its  S  i /S  o  for  T  i  > 350 ◦ C. An-nealed quartz, on the other hand, should have an already el-evated  S  o  (due to its prior sensitization during the annealing)and therefore a lower net increase of each  S  i /S  o , thereforeits TAC.The general form of theTAC corresponds to that observed byother authors for natural quartz (see for example Bailiff, 1994;Adamiec, 2005b). The relative sensitivity is unchanged for ac-tivation temperatures  T  i  < 450 ◦ C. However, for  T  i  > 450 ◦ C itincreases strongly up to an activation temperature of   ∼  600 ◦ C,after which it decreases. The sensitivity decrease observed af-ter annealing to 550  .  600 ◦ C may be explained by electronicreorganization due to major lattice changes during the    to   transition of quartz, which takes place at 573 ◦ C. This interpre-tation is supported by the model proposed byAdamiec (2005b),which includes a deep recombination centre which we wouldnot expect to survive the phase transition.We note that the T  i  at which sensitivity increases is a functionof the duration of the 400 ◦ C annealing cycle. For longer an-nealing durations,  T  i  at which the sensitivity begins to increaseis higher than for shorter annealing durations. The maximumsensitivity is also a function of the duration of annealing. Weobserved that the maximum sensitivity increase in the TAC of quartz held for a 4h annealing cycle is more than two timeslower than that observed after a 1h cycle, and is only 20% of that observed for quartz without any prior annealing. The rel-ative sensitivity enhancement thus depends on the duration of the thermal treatment.Averyinterestingresultfromthesedataisthattheincreaseof the sensitivity enhancement with heating duration is not mono-tonic. Whereas Adamiec (2005a), for other natural quartz sam-ples, observed an increase of the relative sensitization followedby a saturation for laboratory heating up to 6000s, we observean increase of the highest sensitization for 400 ◦ C heating up to1h, followed by a decrease. In our sample, the optimal thermalcycle for maximum TAC sensitization is 1h at 400 ◦ C.Given that annealing at one temperature  ( 400 ◦ C )  for differ-ent durations yielded different results, this result demonstratesthe importance of the duration of annealing. It is therefore 05101520300 400 500 600       S       i       /      S      o Activation temperature ( ° C) unheated 15 min30 min 1h2h 4h Fig. 1. TAC of a previously unheated sample without annealing and afteroven annealing at 400 ◦ C for 15, 30min, 1 , 2, and 4h. 051015202530200 300 400 600       S       i       /      S      o 300 ° C 350 ° C 400 ° C450 ° C 500 ° C 550 ° C600 ° C 800 ° C Activation temperature ( ° C) 500 Fig. 2. TAC of the geological equivalent after oven annealing for 1h attemperatures ranging from 300 to 800 ◦ C. imperative to deduce an “equivalent thermal history”  H  E  ratherthan an “equivalent temperature”. 3.2. Isochronal study of the geological equivalent  To mimic archaeological heating, aliquots of quartz extractedfrom the geological equivalent ochre were oven-annealed for1h at temperatures ranging from 300 to 800 ◦ C. Fresh materialwas used for each annealing temperature. The TACs obtainedfollowing these cycles are presented in Fig. 2. As expected, the highest-maximum sensitization  S  max  is observed for the low-est temperature  ( 300 ◦ C )  annealing cycle.  S  max  is progressivelylower for higher-temperature anneals, and stabilizes for anneal-ing temperatures  > 500 ◦ C.Also as expected, the higher the annealing temperature, thehigher the activation temperature at which the earliest sensitiv-ity increase is observed. It is worth noting, however, that thisactivation temperature is not equal to the annealing tempera-ture: the rise overestimates the previous true  H  E  by  ∼  100 ◦ C.Consequently, the method proposed by Sunta and David (1982)does not allow the determination of   H  E  for our ochre artifacts.  998  C. Lahaye et al. / Radiation Measurements 41 (2006) 995–1000 01234567200300400500600    S    i    /   S   o Bdx 7339Bdx 7340Bdx 7341 Activation temperature ( ° C) Fig. 3. TAC of the three archaeological samples, Bdx 7339, 7340 and 7341,without any prior laboratory thermal treatment. 051015202530200300400500600700800900 Firing temperature( ° C)    (   S    i    /   S   o    )  m  a  x Bdx 7339Bdx 7340Bdx 7341 Fig. 4. Determination of the temperature of the equivalent thermal cycle H  E  for the archaeological samples, comparing the maximum sensitization S  max = (S  i /S  o ) max  with a master curve obtained for the geological equivalent,oven-annealed at different temperatures. 3.3. Equivalent thermal history  (H  E )  determination Once it became apparent that our investigations of the ge-ological equivalent ochre demonstrated the inapplicability of Sunta and David’s method, another approach to determine  H  E had to be devised. To that end, the TACs of the three artifactswere measured without any prior laboratory thermal treatments.The results are presented in Fig. 3. The general form of these TACs is very similar to the TAC of the geological equivalent,however the TAC of sample Bdx 7339 shows a maximum sen-sitization much higher than  S  max  of the two other samples. Byreference to the isochronal TAC curves of  Fig. 2, we interpret this as indicative of differences in the past thermal treatmentseach sample experienced in antiquity.In order to compare the maximum sensitization of the ar-chaeological samples with that of the geological equivalent an-nealed at different temperatures (Fig. 2), we have deduced amaster curve for the geological equivalent, showing  S  max , equalto the maximum of the ratio  S  i /S  o , as a function of the oven-annealing temperature (Fig. 4). This curve clearly shows that S  max  initially decreases rapidly as a function of the annealingtemperature, but then stabilizes at a minimum value once theannealing temperature has exceeded 475 ◦ C. BDX 7339 01234567200300400500600    S    i    /   S   o 00.511.523    S    i    /   S   o 00.511.522.5 Activation temperature ( ° C) 200300400500600 Activation temperature ( ° C) 2.5    S    i    /   S   o  Activation temperature ( ° C) 600500400300200 BDX 7341BDX 7340 unheated350˚C 1h400˚C 1hunheated400 ° C 1h450 ° C 1hunheated300 ° C 1h350 ° C 1h400 ° C 1h Fig. 5. TAC of the archaeological samples Bdx 7339, 7340 and 7341, withoutprior heating and after oven annealing for 1h at different temperatures. We then deduced the  H  E  of each archaeological sample byinterpolating its  S  max  on the master curve of  Fig. 4. The results obtained for the three archaeological ochre samples are: •  Bdx 7339:  H  E  ≈  440 ◦ C for 1h; •  Bdx 7340 and Bdx 7341:  H  E  475 ◦ C for 1h.These thermal cycles are compatible with the past thermal treat-ments deduced by Valladas (1981) for hearthstones of the Up-per Palaeolithic. 3.4.  110 ◦ C  peak behavior in the archaeological samples In order to test the effect of a second annealing cycle on oursamples (the first one being firing in antiquity) the TAC of each  C. Lahaye et al. / Radiation Measurements 41 (2006) 995–1000  999 Bdx 7339 0.0E+003.0E+076.0E+070100200300400500600    1   1   0             °    C  p  e  a   k   T   L   i  n   t  e  n  s   i   t  y   /  m  g not laboratory annealed350 ° C 1h400 ° C 1h Activation temperature ( ° C) Fig. 6. TL intensity per unit mass as a function of the activation temperaturefor sample Bdx 7339 unannealed, annealed at 350 ◦ C for 1h, and annealedat 400 ◦ C for 1h. of the three archaeological samples was measured both withoutany further treatment and after annealing for 1h at  T  n , where300 ◦ C  T  n  450 ◦ C. The results are presented in Fig. 5. Foras-recovered samples (unheated in the laboratory) the initialriseinrelativesensitivitywithincreasingactivationtemperatureis defined by a “critical temperature”,  T  c , which varies fromsampletosample.Weexpectthatannealingat T <T  c  wouldnotalter the TAC curve, and that a 1h annealing at  T >T  c  shouldresult in a decrease of the relative sensitivity (and therefore of  S  max ) as a function of the activation temperature. Based on thepost-annealing changes in the TACs of the three samples, wededuce that their “critical temperatures”  T  c  are: •  BDX 7339: 350 ◦ C <T  c < 400 ◦ C; •  BDX 7340: 400 ◦ C <T  c < 450 ◦ C; •  BDX 7341:  T  c < 300 ◦ C.We note that  T  c  is different from  H  E .The behavior of the archaeological samples, heated for thesecond time (the first heating being the one in antiquity), at tem-peratures higher than  T  c , the temperature we have defined as“critical”, has not yet been reported, at least not under the sameexperimental conditions. We observe, in Fig. 6, a gradual de-sensitization of the TAC with increasing temperature of the sec-ond oven-annealing. For high-temperature annealing (400 ◦ Cin Fig. 6), the TAC is seen to decrease for  T  i  > 215 ◦ C. A sim-ilar effect was noted by Godfrey-Smith and Ilani (2004). They attributed it to aliquot darkening, as their samples could notbe etched, due to very small sample masses. In our case, thisexplanation cannot be proposed, as we were successful in ex-tracting pure quartz.We therefore propose that this is a physicalphenomenon of quartz.We are currently seeking an appropriatemodel to explain it; one of the possibilities may be competition. 4. Conclusion Our study has shown that both the duration and tempera-ture parameters of firing interact to determine the sensitivitychanges of the 110 ◦ C peak. It is thus necessary to refer to the“equivalent hermal history”  H  E , rather than to an equivalenttemperature.For the ochre artifacts of La Honteyre the method of Suntaand David, based on the rise of the TAC sensitivity, couldnot be applied. Equivalent thermal history can nevertheless bedetermined by studying the behavior of the 110 ◦ C peak andchanges in its activation characteristics following annealing,and by comparing the TAC of an archaeological sample to thatof a geological equivalent.Finally, the study of the effects of a second annealing onthe archaeological samples has resulted in an early decrease of sensitization, for which we are still seeking an explanation. Acknowledgements This work was supported by CNRS, University of Bordeaux3, Région Aquitaine and Service Régional de l’Archéologied’Aquitaine.Support of the TOSL laboratory by the Natural Sciencesand Engineering Council (NSERC) and the Social Sciencesand Humanities Research Council (SSHRC), both of Canada,through research grants to DIGS, is gratefully acknowledged. References Adamiec, G., 2005a. Properties of the 360 and 550nm TL emissions of the“110 ◦ C peak” in fired quartz. Radiat. Meas. 39, 105–110.Adamiec, G., 2005b. Investigation of a numerical model of the pre-dosemechanism in quartz. Radiat. Meas. 39, 175–189.Aitken, M.J., 1985. Thermoluminescence Dating. Academic Press, London.359pp.Bailiff, I.K., 1994. The pre-dose technique. Radiat. Meas. 23, 471–479.Benny, P.G., Sanjeev, N., Gundu Rao, T.K., Bhatt, B.C., 2000. Gamma rayinduced sensitization of 110 ◦ C TL peak in quartz separated from sand.Radiat. Meas. 32, 247–252.Benny, P.G., Gundu Rao, T.K., Bhatt, B.C., 2002. The E’1-centre and its rolein TL sensitization in quartz. Radiat. Meas. 35, 369–373.Charitidis, C., Kitis, G., Furetta, C., Charalambous, S., 2000. Superlinearityof synthetic quartz: dependence on the firing temperature. Nucl. Instrum.Methods Phys. Res. B 168, 404–410.Chen, R.,Yang, X.H., McKeever, S.W.S., 1988. The strongly superlinear dosedependence of thermoluminescence in synthetic quartz. J. Phys. D: Appl.Phys. 1452-1457.Godfrey-Smith, D.I., Ilani, S., 2004. Past thermal history of goethite andhematite fragments from Qafzeh Cave deduced from thermal activationcharacteristics of the 110 ◦ C TL peak of enclosed quartz grains. RevArchéométrie 28, 185–190.Godfrey-Smith, D.I., Bouchet-Bert, L., von Bitter, P.H., Storck, P.L., 2005.Thermal activation characteristics and thermoluminescence of chert fromthe Red Wing, Ontario region, and its putative heat treatment in prehistory.Geochronometria 24, 13–20.Göksu, H.Y., Weiser, A., Regulla, D.F., 1989. 110 ◦ C TL peak records theancient heat treatment of flint. Ancient TL 7, 15–17.Lahaye, C., Guibert, P., 2002. Etude par thermoluminescence de l’état dechauffage de fragments de grès ferrugineux rubéfiés provenant du sitepaléolithique supérieur de La Honteyre, Le Tuzan, Gironde. Rapport derecherche auprès de la DRAC, Septembre 2002, 6pp.Roque, C., 2001. Datation en Archéologie : des recherches méthodologiquesen TL et en OSL aux référentiels chronologiques du Solutréen(France), du Néolithique (Grèce) et de la culture préhispanique Moche(Pérou). Thèse de doctorat de troisième cycle, Université Bordeaux III,406pp.
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x