The thermoluminescent properties of natural calcium fluoride for radiation dosimetry

Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651

Contents lists available at ScienceDirect

Nuclear Instruments and Methods in Physics Research B

journal homepage: www.elsevier .com/locate /n imb

The thermoluminescent properties of natural calcium fluoridefor radiation dosimetry

H. Tugay a, Z. Yegingil a,*, T. Dogan b, N. Nur a, N. Yazici c

a Cukurova University, Art and Sciences Faculty, Physics Department, 01330 Adana, Turkeyb Adiyaman University, Vocational School of Besni, Department of Technical Programs, 02300 Adiyaman, Turkeyc Gaziantep University, Engineering Faculty, Physics Engineering Department, Gaziantep, Turkey

a r t i c l e i n f o a b s t r a c t

Article history:Received 26 February 2009Received in revised form 8 September 2009Available online 27 September 2009

Keywords:Thermoluminescence dosimetry (TLD)FluoritesDose responseHeating rateAnnealing of fluoritesRecyclingAkçakent fluorites

0168-583X/$ - see front matter � 2009 Elsevier B.V.doi:10.1016/j.nimb.2009.09.021

* Corresponding author. Tel.: +90 322 338 60 84x2E-mail address: [email protected] (Z. Yegingil).

The characteristics of natural calcium fluoride from Çiçekdagı Massif (Akçakent) in Turkey have beenstudied by analysing its thermoluminescence glow curve structure between 30 and 450 �C for the pur-pose of radiation dosimetry. A variety of thermoluminescence measurement regimes have been exam-ined to determine the most effective and appropriate annealing temperature, heating rate and doserange for the proper and accurate use of this phosphorescent material. After a high temperature anneal-ing as TL readings, optimum values for low temperature annealing and heating rate were obtained as60 �C for 24 h and 1 �C s�1, respectively. In the dose range of 0.5 Gy–1 kGy, the intensity of individualglow peaks and overall glow curve shape changed. The peak intensities of all glow curves located at100 and 120 �C (overlapping considerably), and at 215 �C, at 310, 350 and 410 �C (overlapping) increaselinearly with increasing ionizing radiation over a range of from 0.5 Gy to 10 Gy.

� 2009 Elsevier B.V. All rights reserved.

1. Introduction

Thermoluminescence in solids is the emission of light from aninsulator or semiconductor that takes place during the heating ofa solid, following an earlier absorption of energy from radiation[1,2]. The basis of thermoluminescence is that many vacancies,interstitials and dislocations in materials can act as charge trapsand/or luminescence centers. The subsequent release of thetrapped charge, by increasing the temperature of the specimen ata linear rate, can be recorded as light emission which is measuredby a photomultiplier tube circuit. The application of thermolumi-nescence to radiation dosimetry is possible because, in principle,the luminescence intensity is proportional to the concentrationof traps, as well as to the absorbed radiation dose. Thus, many syn-thetic or natural materials which exhibit thermoluminescenceproperties have become established in luminescence dosimetry[3–8].

CaF2, one of these materials, is available in the mineral form as afluorite with a cubic crystalline structure and exhibits strong radi-ation induced TL, which can be attributed to transition metal dop-ant ions such as Mn and rare-earth ions such as Sm, Ce, Th, Tm andDy. A significant advantage of CaF2 phosphor is that it is an abun-

All rights reserved.

495; fax: +90 322 338 60 70.

dant naturally occurring material and can thus be cheaply ob-tained. Since the first introduction of fluorite as athermoluminescent dosimeter by Schayes et al. [9], the thermolu-minescence properties and emission mechanisms of naturallyoccurring fluorite crystal have been extensively studied experi-mentally [10–14]. Thermoluminescence glow curves were re-corded from the room temperature to 550 �C by Sunta for anumber of natural CaF2 samples [11]. He discovered that the ther-moluminescent spectrum consists of a number of sharp linesacross a spectral range from ultraviolet to infra-red and that almostall of the emission lines can be attributed to one or more of therare-earth impurities. He proposed that the thermoluminescencemechanism in CaF2 involves the release of holes from lattice cen-tres and their recombination at the rare-earth impurity sites. Lumi-nescence spectra are reported for natural fluorite crystals ofdifferent colours by Calderon et al. who discussed apparent differ-ences between analyses; fading studies indicated that the trappingand luminescence sites are not independent but part of large defectcomplexes [14].

Knowledge of the limitations of various luminescence dosime-ters is vital to their employment. Annealing, which is the methodused to deplete the population of trapped electrons or to changethe trap structure to prepare the material for re-use, can affectthe dose measurement. Annealing regimes are therefore specifiedfor some TL materials so that low temperature glow peaks are

H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651 3641

eliminated without affecting the higher temperature peaks [15].Changes in TL intensity and height of a glow peak due to heatingrate have been studied by Sunta [16]. A ‘‘shielded storage experi-ment” for estimating both the ‘‘self dose” and ‘‘fading” of CaF2 (nat-ural) TL phosphors for environmental monitoring applications isdescribed [17]. Ogundare et al. investigated the influence of differ-ent heating rates on the thermoluminescence (TL) glow peak tem-perature, height and ‘‘integrated TL intensities” of the two hightemperature glow peaks of a Nigeria fluorite sample [18].

An obviously desirable property of a TLD detector is that itexhibits a linear relationship between thermoluminescence inten-sity and absorbed dose. However, it is clear that many thermolumi-nescent materials exhibit a non-linear growth of glow intensitywith absorbed dose over certain dose ranges. Both supralinearityand the sublinear growth approaching saturation can lead to prob-lems of under-or overestimation respectively (here saturation re-fers either to the condition where all traps are full or to theonset of appreciable radiation damage). The occurrence of a non-linear region in the dose response curve of a detector does not pre-clude its use in TLD, but it does require careful calibration and cor-rection, from which additional errors may accrue [1]. Thebehaviour of natural CaF2 was studied for doses up to 500 kGy byHasan [19]. Dose responses of individual peaks show that for suffi-ciently low doses the response is linear; at higher values it be-comes supralinear, and at large values it approaches a constantlevel with the same shape glow curve. A batch of naturally occur-ring fluorite (CaF2) from the Middle Benue Valley region of Nigeriawas studied in some detail for its thermoluminescence (TL) prop-erties by Balogun et al. [20]. They obtained TL glow peaks at 119,144 and 224 �C at a heating rate of 10 �C s�1. They observed thatthe TL response increased with increasing dose over the dose rangeexamined and with variations in the decay curves of the variousglow peaks during storage at room temperature. Topaksu and Yaz-ici investigated the TL properties of natural CaF2 after beta irradia-tion at room temperature [21]. Dose responses and fading processwere examined for the determined six peaks of these phosphors.The influence of heating rates on the response of the dosimetricglow peaks of natural CaF2 was studied.

Fig. 1. Energy diagrams for an electron and a hole drawn together [22].

2. Electrons and holes in impure crystals

One of the remarkable developments in recent years has beenthe application of solid state science to phosphorous materialdevelopments in TL dosimetry (TLD) used for radiation dose mea-surements. A group of phosphors in most common use today areCaF2 and CaF2-based materials. These elements crystallize in a cu-bic structure in which the atoms have tetrahedral bonding withtheir four nearest neighbours. If an extra electron is put into theCaF2 crystal (rare-earth (RE) impurity ions) which is at low temper-ature, the electron will be able to wander around in the crystaljumping from one atomic site to the next. We can have a similarsituation if we remove an electron from this neutral insulator. Ifan electron is removed then an electron from a nearby atom canjump over and fill the hole. The electron that jumps over leaves ahole in the atom that it originated from. We can say that the holecan jump from one atom to the next. The mathematics is just thesame for the hole as it was for the extra electron. The hole energylies in a restricted band and near the bottom of the band its energyvaries quadratically with the momentum. The hole also behaveslike a classical positive particle with a certain effective massthrough the crystal. If we put several electrons into a neutral crys-tal, they will move around much like the atoms of a low pressuregas. If there are not too many, their interactions will not be veryimportant. Similarly we could put many holes into a crystal. Theywould roam around at random. One can also have both holes and

electrons together. If there are not too many, they will all go inde-pendently. For obvious reasons, electrons are called the negativecarriers and the holes are called positive carriers. It is also possibleto create an electron–hole pair by taking a bound electron awayfrom one neutral atom and putting it some distance away in thesame crystal. We then have a free electron and a free hole [22].

The energy required to put an electron into a state S – we say tocreate the state S – is the energy E�. It is some energy above E�min.The energy required to create a hole in some state S

0is the energy

E+ which is some energy greater than Eþmin. Now if we create a pairin the states S and S

0, the energy required is just E� + E+. The energy

Epair = E� + E+ required to create a pair with the electron in S andthe hole in S

0is just the vertical distance between S and a S

0as

shown in Fig. 1.Electron–hole pairs can be produced by high energy particles.

When a fast-moving charged particle with an energy of tens orhundreds of MeV goes through a crystal, its electric field will knockelectrons out of their bound states creating electron–hole pairs.Such events occur hundreds of thousands of times per millimetertrack. At any finite temperature there is still another mechanismby which electron hole pairs can be created. The pair energy canbe provided from the thermal energy of the crystal. The thermalvibrations of the crystal can transfer their energy to a pair, givingrise to ‘‘spontaneous” creation [22].

3. Experimental procedure

The aim of the present work is to investigate and report the TLcharacteristics of natural CaF2 using different approaches. Naturalsamples of fluorite from different mines in the central region ofTurkey have different colours of white, violet, green and transpar-

3642 H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651

ent. Green fluorite, obtained from a deposit (Çiçekdagı Massif) inthe Akçakent region, was selected for studying thermolumines-cence properties of this phosphor, in detail. The sample used in thisstudy has been examined in crystalline form. The use of carvingand grinding techniques have been employed to produce smallcrystalline fragments of approximately equal masses (20 mg each)from the main fluorite body; discoloured and imperfect fragmentshave been rejected. After washing the crystal samples, crystal frag-ments have been handled using tweezers. All elements have beenstored in optically opaque bags and in dark room away from sun-light. The experiments have been carried out in the same darkroom.

To give a rough description of the crystalline structure and ele-mental–chemical composition of the sample, XRD and XRF analy-ses were undertaken using an X-ray diffractometer, D8 Advanced(Bruker AXS) and X-ray Fluorescence spectrometer Axios Advanced(Pan Analytical), respectively. The operating voltage of the XRF was

Fig. 2. XRD patterns of th

40 kV and the corresponding current is 40 mA. Mo tube wavelength is 0.70930 Å. The calcium fluoride content in the samplewas found to be 99.67% of the whole sample. Si, Cu, Fe, Mg and Sare appeared as the major trace elements of the sample (Fig. 2) (Ta-ble 1 and 2).

For the annealing procedure a Nabertherm GmbH model elec-trical oven and a LABART model incubator with digital screen wereused. The oven brought up to the pre-selected temperature whichallowed to stabilise before planchets were inserted. Annealing tookplace in air and contamination affecting the phosphor responsewas avoided by reserving ovens for this purpose only. The temper-ature was controlled to within two degrees Celsius in oven. It hasbeen determined previously by direct measurements that no tem-perature gradients exist within the oven and all planchets in a petridish attained the appropriate annealing temperature regardless oftheir position. Stainless steel planchets were used to reduce ther-mal gradients. The samples were annealed 60 �C for 24 h after

e phosphor sample.

Table 1The results of XRD patterns.

a b c Alpha Beta Gamma Birim HücreHacmi

5.46596 5.46596 5.46596 90 90 90 163.305

Table 2The results of XRF analysis of fluorite sample.

Name Content (%) Error (%)

Ca 82.8585 0.2F 16.7998 0.4Si 0.1477 0.009Cu 0.0716 0.003Fe 0.0637 0.003S 0.0352 0.002Mg 0.0235 0.002

0 50 100 150 2000

1x107

2x107

3x107

4x107

5x107

a

TL In

tens

ity (a

rbitr

ary

units

)

Tempera

natural

0 50 100 150 200

1x10 6

2x10 6

3x10 6

a'

TL In

tens

ity (a

rbitr

ary

units

)

Temperatu

100°C 150°C 200°C 300°C 400°C 500°C

Fig. 3. Natural dose measurements and TL response of three phosphors which were prenatural dose (after 1st readout) (a0) its glow curve after a 0.5 Gy irradiation for different annatural dose (after 1st readout) (b0) its glow curve after a 1.0 Gy irradiation for different annatural dose (after 1st readout) (c0) its glow curve after a 1.0 Gy irradiation for different

H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651 3643

the TL readings to erase natural residual absorbed dose, before thesubsequent irradiation and then rapid cooling of samples wereestablished on an aluminium tray for half an hour to roomtemperature.

All TL measurements on several portions of the annealed sam-ples, were carried out by using the RISO TL/OSL LuminescenceReader Model TL/OSL-DA-20 which allows up to 48 samples tobe individually heated to any temperature between room temper-ature and 700 �C and individually irradiated by radioactive betasource (90Sr/90Yr). The emitted luminescence is measured by a bial-kali EMI 9235QA photomultiplier tube (PMT) which has an ex-tended UV response with maximum detection efficiency between300 and 400 nm. To prevent scattered stimulation light fromreaching the PMT, the reader is equipped with a 7.5 mm Hoya U-340 detection filter which has a peak transmission around340 nm. Thermal stimulation is achieved using the heating ele-ment which heats the sample and lifts the sample into the mea-

250 300 350 400 4500

1x107

2x107

3x107

4x107

5x107

ture (°C)

250 300 350 400 450

1x10 6

2x10 6

3x10 6

0.5 Gy

x10

re (°C)

viously annealed for different temperatures (a) glow curve of sample 1 due to thenealing temperatures between 100 and 500 �C (b) glow curve of sample 2 due to thenealing temperatures between 100 and 450 �C (c) glow curve of sample 3 due to theannealing temperatures between 40 and 100 �C.

0 50 100 150 200 250 300 350 400 450

0

1x107

2x107

3x107

4x107

5x107

0

1x107

2x107

3x107

4x107

5x107

b

TL

Int

ensi

ty (

arbi

trar

y un

its)

Temperature (°C)

natural

0 50 100 150 200 250 300 350 400 4500

1x106

2x106

3x106

4x106

0

1x106

2x106

3x106

4x106

b'

1 Gy

TL

Int

ensi

ty (

arbi

trar

y un

its)

Temperature (°C)

100°C200°C300°C400°C450°C

Fig. 3 (continued)

3644 H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651

surement position. The heating system is able to heat the samplesup to 700�C at linear heating rates from 0.1 to 10 K s�1. The heatingstrip is cooled by a nitrogen flow which also protects the heatingsystem from oxidation at high temperatures. A detachable betairradiator located above the sample carousel accommodates a1.48 GBq (40 m Ci) 90Sr/90Yr beta source which emits beta particleswith a maximum energy of 2.27 MeV. The dose rate in quartz at thesample position is 6.689 Gy/min.

At each luminescence measurements, three CaF2 crystal frag-ments on stainless steel planchets were read out. To investigatethe optimum heating rate for the determined phosphor to be usedfor dosimetric purposes, different readout procedures were per-formed with various heating rates from 1 to 10 �C s�1. The sampleswere irradiated at a high dose (0.1 Gy–1 kGy) to obtain dose re-sponse of the phosphor and the glow curves measured by settingthe linear heating rate of the TL reader at 1 �C s�1. Also, each pieceof CaF2 crystal was read out two or three times and the last readout was considered to be the background of the system (readerand crystal) and subtracted from the first one.

4. Results and discussion

Occurrence and temperature of thermoluminescence glowpeaks of natural fluorite exhibit a complex range of order and usu-ally they depend on the source of the mineral. For these reasons,we decided to perform a detailed study of Çiçekdagı Massif fluor-ites before a probable use of it for routine radiation monitoringpurposes. We examined the general thermoluminescence proper-ties of the CaF2: natural dosimeter material to investigate: (1) anannealing procedure which is usually being utilized to improvethe performance of TL material before irradiation, (2) dependenceof TL on radiation dose, (3) effect of heating rate on TL behaviour,(4) reusability of this TLD material.

4.1. Thermal annealing

It is worth mentioning that the reliable re-use of TLD materialsoften requires the use of strict thermal annealing procedures.Thermal annealing normally consists of a particular recipe of

0 50 100 150 200 250 300 350 400 4500

1x107

2x107

3x107

4x107

5x107

0

1x107

2x107

3x107

4x107

5x107

TL

Int

ensi

ty (

arbi

trar

y un

its)

Temperature (°C)

natural

c

0 50 100 150 200 250 300 350 400 450

0

1x106

2x106

3x106

4x106

5x106

6x106

7x106

0

1x106

2x106

3x106

4x106

5x106

6x106

7x106

c'

1 Gy

TL

Int

ensi

ty (

arbi

trar

y un

its)

Temperature (°C)

40°C50°C60°C70°C80°C90°C100°C

Fig. 3 (continued)

H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651 3645

pre-irradiation annealing (i.e. a given temperature for a givenperiod). A post-irradiation annealing recipe is also recommendedbefore reusing of the dosimeters. The effect produced by this lowtemperature anneal is a redistribution of the trapping centres tofavour the high temperature ‘‘dosimetric” traps [16]. Generally, thepurpose of pre-irradiation annealing is to re-establish the thermo-dynamic defect equilibrium which existed in the material beforeirradiation and readout. This is achieved by reversing the thermallydriven diffusion reactions which sometimes occur during the TLprocess. These reactions may themselves be catalysed by the trap-ping of non equilibrium charge at the defects during the irradiation[23,24]. It is worth mentioning that since the TL sensitivity of agiven peak is affected by the presence of deep, thermally discon-nected, competing traps, it will be necessary to reset the occupancyof these more stable centres, if they exist, by high temperatureannealing. Incorrect choice of annealing temperatures and times,however, can have the opposite effect to that desired. Exactly theright equilibrium state must be achieved each time if the materialis reliably re-used [7]. The annealing protocols which have beenestablished are believed to achieve these aims.

Earlier experiments on fluorite indicate that the CaF2: natural,similar to LiF, shows an enhancement of sensitivity after being irra-diated to a gamma exposure. This increase in sensitivity is quitedependent on the temperature and on the duration of annealingafter the sensitizing irradiation. The ‘‘standard” post-irradiationannealing for CaF2: natural to increase sensitivity is the annealingfor 30 min at 450 �C [4].

In our study, to find the most suitable annealing conditionsfor routine dosimetric use, three portions of the CaF2 crystal,each is about 20 mg were used for annealing experiments. Thedosage and time period of annealing applications are adjustedto prevent crystal damage. Natural TL readings by heating spec-imens from 30 to 450 �C at a heating rate of 1 �C s�1 in a N2

atmosphere up to erasing any residual information before subse-quent irradiation have been performed. It was found that firstreadout annealing is insufficient to completely erase the accu-mulated exposure and empty the deep traps in CaF2 and thereadings have been repeated until the complete erase of the sig-nal. The natural glow curves of these three samples are shown inFig. 3a–c.

3646 H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651

Pre-irradiation annealing was performed on two samplesthrough annealing temperatures of 100, 150, 200, 300, 400, 450and 500 �C for 15 min. The third specimen was heat-treated be-tween 40 and 100 �C with the subsequent temperatures of 10 �Cdifference for 24 h. All annealed samples were cooled down onan aluminium tray for about half an hour to room temperature.

After each annealing process, the first and second fluorite sam-ples were further exposed to b irradiation doses of 0.5 and 1.0 Gyfrom the 90Sr/90Y source, respectively. The TL responses of the irra-diated phosphors were read soon, up to 450 �C, using the slowheating rate (1 �C s�1). We found that the absolute and relativeheights of the peaks can be changed by this annealing for differenttemperatures. It is worth mentioning that since the annealingtakes place before irradiation, these changes are not caused bythe release of electrons from traps but must represent either achange in the number of traps related to a given peak (the trappingefficiency) or a change in the conversion efficiency from trappedelectrons to photons [4]. Fig. 3a0 and b0 show the changes in theglow curves as a function of TL reading temperature for differentannealing temperatures of 100, 200, 300, 400, 450 and 500 �C. Theyexhibited different complex glow curves mostly with five verynoticeable peaks of different intensities at approximately 100and 120 (overlapping), 210, 300 and 350 �C (overlapping) and400 �C The very high temperature peaks, as illustrated inFig. 3a0–c0, are difficult to measure without permanently damagingthe phosphor by the heat treatment. Fig. 3a0 and b0 show the effecton the glow curves of annealing at high temperatures. After 15 minpre-irradiation annealing from 100 to 450 �C, all of the peaks of theglow curves of both of the samples were reduced and lower peaks(peaks 1 and 2) were appeared. With increasing annealing temper-ature, TL intensity of the peaks, located at 100 and 120 �C (overlap-ping considerably), 210, 300 and 350 �C (overlapping) and 400 �Cwere significantly reduced (Fig. 3a0 and b0). Because various atomicsubstitutions take place easily in this phosphor’s crystal structurepeaks 250 and 280 �C are the peaks that are most useful for dosim-etry purposes as they are located between 250 and 300 �C [4,25].Thereby, the glow peaks which were not susceptible to risingannealing temperature could produce maximum peak intensityat lower annealing temperatures.

50 100 150 2000.0

5.0x101

1.0x102

1.5x102

2.0x102

2.5x102

3.0x102

3.5x102

P3P2

P1

Peak No Ea (eV) ln(s)(s-1) b

1 0.8393 23.44 1 2 0.664 16.62 1

3 1.178 25.11 1

4 1.766 32.30 1 5 0.9848 14.64 1

6 0.8718 11.15 1

TL

Int

ensi

ty (

arbi

trar

y un

its)

Temp

Fig. 4. Analysis of the glow curve from natural CaF2 by using the CGCD method after 10 GThe light curves are fitted individual peaks and the bold dashed curve is sum of these l

It is known from the general research on phosphor characteris-tics that the low temperature trap structure can be rearranged (thedipole trap structure becomes three associated dipoles) by anneal-ing at 80 �C for 24 h [24,7]. For the examined phosphor, we con-ducted the same experiment by using the third portion, with anexposure radiation of 1 Gy, but with lower annealing temperatures(less than 100 �C) and an increased duration to ensure the bestannealing range. The TL was initially recorded without any anneal-ing process to clarify the difference between the unannealed andannealed samples dose responses. The sample was annealed for24 h duration at 10 �C intervals from 40 to 100 �C. After eachannealing, the sample was irradiated with a 1 Gy beta dose andTL measurements were performed repeatedly with a heating rateof 1 �C s�1 up to 450 �C under a N2 atmosphere. The build up ofTL for peak 4 which is most useful for dosimetry purposes wasmuch stronger in 60 �C annealing. Comparison of these datashowed that the most reproducible glow and dose response curvesfor dosimetric usage of natural fluorite can be obtained by anneal-ing the sample at 60 �C for 24 h (Fig. 3c0).

4.2. Dose response

Different dose rates (0.1 Gy–1 kGy) were used for the purposeof dose response study in the natural CaF2 crystals using 90Sr/90Yrbeta source with the dose rate of 6.689 Gy/min at that point of irra-diation. First, we performed TL reader annealing repeatedly for ashort time at a high temperature (from 30 to 450 �C) up to erasingany residual information before subsequent irradiation in CaF2. Thereadouts of the fluoride samples were carried out on the TL readerwith a heating rate of 1 �C s�1

. A pre-irradiation (and/or re-use)annealing of 60 �C for 24 h to arrange low temperature trap struc-ture was applied to all the samples tested after TL readerannealing.

The computerized glow curve deconvolution (CGCD) methodwas used to analyse the thermoluminescence glow peak after10 Gy beta irradiation (Fig. 4). We determined six peaks at 100,120, 215, 310, 350 and 400 �C. The 100 �C peaks (peak 1) and120 �C peaks (peak 2) were not present in the natural glow curvesof all fluorite samples because of their short lifetime at room

250 300 350 400 450

P6

P5

P4

erature (oC)

y dose level. The main signal is made up of two component peaks of 310 and 350 �C.ines.

1 10 100 1000105

104

103

105

106

107

108

TL

Int

ensi

ty (

arbi

trar

y un

its)

Dose (Gy)

a

peak 3

peak 4

20 40 60 80 100100

150

200

250

300

350

400

100

150

200

250

300

350

400

Tem

pera

ture

(°C

)

Dose (Gy)

peak 3

peak 4

b

100 200 300 400 500 600 700 800 900 1000100

150

200

250

300

350

400

100

150

200

250

300

350

400

Tem

pera

ture

(°C

)

Dose (Gy)

c

peak 3

peak 4

Fig. 5. (a) The TL dose response curve of dosimetric glow peaks in the glow curves of the phosphor CaF2 sample determined by the peak height and dose value, (b and c)Shifting of glow peak temperatures with rising beta dose level between 0.5–100 Gy and 0.5 Gy–1 kGy, at a heating rate of 10 �C s�1, respectively.

H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651 3647

temperature; they faded very quickly. These peaks will appearwhen the samples are irradiated; the lower temperature peak isof no use in dosimetry. In the present study, after irradiation, thepeaks at 215 (peak 3) and 310 �C (peak 4) were found to be stable

and hence the dose response (0.5 Gy–1 kGy) for these two glowpeaks are examined graphically (Fig. 5a–c).

It is worth mentioning that we define the normalised dose re-sponse function (or supralinearity index) f (D) such that:

3648 H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651

f ðDÞ ¼FðDÞ

D

� �

FðD0 ÞD0

� �

where F(D) is the dose response at a dose D and D0 is a low dose atwhich the dose response is linear. Thus, our ideal dosimeter wouldsatisfy f (D) = 1 over a wide dose range. Mostly, it is found thatf (D) = 1 only over a narrow dose range, up to a few Gy, in many TLDmaterials. Supralinearity, defined as f (D) > 1, is commonly observed,while sublinearity (f (D) < 1) is most often observed during the ap-proach to saturation. Eventually, at very high doses, the material sat-urates. Such a material has f (D) = constant over the entire dose rangeand the response of the dosimeter is the same [7].

It is clear from Fig. 5a that the 215 �C glow peak (peak 3) and310 �C glow peak (peak 4) exhibited a linear relationship betweenthermoluminescence intensity and absorbed dose up to 10 Gy. Ifwe look at the TL response for the peaks 3 and 4, we see an exampleof the TL signals which have the well known linear–supralinear–sub-linear-saturation behaviour. In the dose response graphs, supralin-earity–sublinearity is very weak for peak 3 but it is very clear forpeak 4. The phosphor material has a linear response to the dose upto a limit,�10 Gy, and then has a greater than linear response, calledsupralinearity, up to 50 Gy after which the curve has sublinearitycharacter up to 100 Gy. The onset of this supralinearity is dependenton another impurity, namely the hydroxyl ion [4]. After 100 Gy doseof exposure, up to 1 kGy, the dose response curve saturated since allthe available sites within the track volume became saturated.

Fig. 5b and c show the change in the temperatures of the peakmaximums versus recorded dose values of 0.5–100 Gy and0.5 Gy–1 kGy, for peaks 3 and 4, up to maximum temperature of450 �C at a heating rate of 1 �C s�1, respectively. Changes in doseexposure have several pronounced effects on the glow curve.Increasing the dose from 0.5 to 10 Gy did not change the maximumpeak temperatures of peaks 3 and 4 while from 10 Gy to 1 kGy,changed through shifting to the lower temperatures.

0 50 100 150 200

0.0

5.0x10 5

1.0x10 6

1.5x10 6

2.0x10 6

2.5x10 6

3.0x10 6

3.5x10 6

4.0x10 6

TL

Int

ensi

ty (

arbi

trar

y un

its)

Temperat

1°C/s

3°C/s

5°C/s

7°C/s

10°C/s

Fig. 6. The glow curves after the beta dose exp

4.3. Heating rate

To check the heating rate effects on peak positions and TL inten-sities of glow curves, five different heating rates of 1, 3, 5, 7 and10 �C s�1 and five different exposure doses of 0.5, 1.0, 5, 10 and50 Gy were employed. The thermoluminescence intensity of eachglow peak for all glow curves and doses decreased at differentquantities with increasing heating rate. At the heating rate of1 �C s�1, the intensities of all peaks are relatively high comparedwith the peaks recorded at higher heating rates. In addition, themaximum temperatures of all glow peaks were shifted to highertemperatures with greater heating rates at all exposure doses.The TL glow curves from the phosphor sample, which were irradi-ated with the beta dose of 1.0 Gy and read using different heatingrates, show changing maximum peak heights and peak tempera-tures without affecting the shape of the glow curves (Fig. 6). Inaddition to the decrease in height for peaks 3 and 4 of the exam-ined phosphor samples (Fig. 7a), the positions of these two peaksshifted to the right, to higher temperatures (Fig. 7b) with increas-ing heating rate.

Also, different amounts of radiation produced different glowcurves for various heating rates. Fig. 8a and b compares differentb exposures for the examined heating rates for peak 4. In Fig. 8a,the relative exposed doses of 0.5, 1.0, 5, 10 and 50 Gy were plottedas a function of thermoluminescence intensity for heating rates of1, 3, 5, 7 and 10 �C s�1, for peak 4. This tends to suppress the low-ering value of TL intensity of peak 4, while raising the value ofheating rate. A heating rate of 1 �C s�1, when compared to a heatingrate of 10 �C s�1, resulted in greater enhanced peak 4 intensitieswhich are 3�, 5�, and 10� higher than that of for doses of 5, 10,and 50 Gy, respectively.

From the same glow curve data, it is also evident that the max-imum peak temperatures of peak 4 shifted to greater values withincreasing heating rates (Fig. 8b). We observed that the minimumshift toward the maximum temperatures for different heating rateshas appeared for relatively high 50 Gy b dose exposure. The maximum

250 300 350 400 450

0.0

5.0x10 5

1.0x10 6

1.5x10 6

2.0x10 6

2.5x10 6

3.0x10 6

3.5x10 6

4.0x10 6

1 Gy

ure (°C)

osure of 1 Gy with different heating rates.

0 1 2 3 4 5 6 7 8 9 10 11 12

5.0x105

1.0x106

1.5x106

2.0x106

2.5x106

3.0x106

5.0x10 5

1.0x10 6

1.5x10 6

2.0x10 6

2.5x10 6

3.0x10 6

TL

Int

ensi

ty (

arbi

trar

y un

its)

Heating rate (°C/s)

peak 3 peak 4

a

0 1 2 3 4 5 6 7 8 9 10 11 12200

220

240

260

280

300

320

340

360

380

4000 1 2 3 4 5 6 7 8 9 10 11 12

200

220

240

260

280

300

320

340

360

380

400

Tem

pera

ture

(°C

)

Heating rate (°C/s)

peak 3 peak 4

b

Fig. 7. Changes in glow peak characteristics by changing heating rate: (a) decrease in the TL intensity of 215 �C (peak 3) and 310 �C (peak 4) glow peaks, with increasingheating rate, (b) shifting of the maximum peak positions of peaks 3 and 4 to the right with increasing heating rate.

H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651 3649

peak temperature for peak 4 followed a comparatively stable trenddue to the rest of the heating rates for the value of 1 �C s�1.

4.4. Reusability

For some TL materials, their repeated use might place changeinto the deep, competing traps, and this change accumulates witheach re-use. Therefore, repeated use of a TL sample (i.e. cycles ofirradiation and readout) means that the sensitivity would evolveand be non-constant over the usage period. To confirm that theTL material has precisely the same properties during each use, re-peated cycles of thermal annealing plus irradiation exposure fol-lowed by TL readouts should be conducted in the determinationof TL characteristics for any type of phosphor material. To checkdose response of the fluorite sample and its reusability and alsoto investigate the range of applicability of ‘‘unannealed” CaF2: nat-ural TLDs, a cycle involving irradiation, annealing (at 60 �C for 24 h)

and readout (up to 450 �C, heating rate 1 �C s�1) was repeated ninetimes. The TL responses of five different samples, nearly 20 mg ofeach, were determined and mean value of maximum intensitiesof peak 4’s for each glow curve have been calculated for every cy-cle. A dramatic decrease was observed both for annealed and unan-nealed samples for the low dose values of 0.5 and 1.0 Gy just afterthe first cycle (Fig. 9). Than, up to 9th cycle, in peak 4 intensitiesand between the samples themselves, no difference has been ob-served for the subsequent TL readings. After absorbing the 10 Gydose, both of the annealed and unannealed phosphors showedthe same kind of behaviour during the recycling process. We haveobserved approximately 50% sudden decrease in the TL intensity ofpeak 4 in the second cycle for both of the phosphors and then asimilar action was followed. The decrease of the maximum peakheight of peak 4 in the 9th cycle compared with 2nd cycle was�20% for both of the phosphors (Fig. 9). Vice versa, considerablevariability was observed between their relative glow peak heights

5 10 15 20 25 30 35 40 45 50 55

1x107

2x107

3x107

4x107

5x107

1x10 7

2x10 7

3x10 7

4x10 7

5x10 7

peak 4

TL

Int

ensi

ty (

arbi

trar

y un

its)

Dose (Gy)

1°C/s3°C/s5°C/s7°C/s10°C/s

5 10 15 20 25 30 35 40 45 50 55250

300

350

400

450

500

250

300

350

400

450

500

peak 4

Tem

pera

ture

(°C

)

Dose (Gy)

1°C/(s3°C/(s5°C/s7°C/s10°C/s

a

b

Fig. 8. (a) Heating rate change effects the TL intensity of peaks 3 and 4 (b) temperature dependence of the same peaks for the same data shown in (a).

3650 H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651

in the recycling process of the annealed and ununnealed phosphorsfor 50 Gy dose. The TL intensity has changed�20% just after the 1stcycle for the unannealed phosphor whereas no alteration has beenobserved for the annealed sample. The expected change has beenobserved in a lesser degree after the second cycle for the latestone. The effect of annealing has been recorded clearly to reusabilityof the phosphors for 50 Gy dose just after the first cycle. As it hasbeen seen in Fig. 9, the paths for the annealed and unannealedsamples passing through the maximum peak values of peak 4 inthe graph, for different cycles, treated as parallel to each other be-tween the 2nd and 9th cycles.

5. Conclusion

Since the choice of a suitable phosphor for a particular study de-pends on the nature of the application [4] we anticipated to revealthe most important TL parameters of this useful TLD phosphorCaF2: natural.

The study has pointed out that a CaF2: natural TLD can be usedin multiple uses as a dosimetric material because it exhibits en-hanced dose response curves and relatively simple glow curvestructure after being annealed and irradiated to a beta dose expo-sure. This increase in dose response was quite dependent on thetemperature and on the duration of annealing before and after irra-diation. The pre-irradiation annealing to increase the expecteddose response was determined as for 24 h at 60�C. In CaF2: naturalsamples the low level dose detection was high and the dose re-sponse curve was linear up to approximately 10 Gy. The TL re-sponse for the 215 �C glow peak (peak 3) and 310 �C glow peak(peak 4) exhibited expected linear–supralinear–sublinear behav-iour. Changes in the heating rate had several effects on the glowcurves of CaF2: natural sample. Temperatures of the glow peaksas a function of heating rate were changed. Also variations in thepeak heights and peak areas have been observed. In this research,for the measurements of the TL by peak heights, the annealing con-ditions of at 60 �C for 24 h and a relatively slow heating rate of1 �C s�1 have been chosen for intercomparison analysis of glow

0 1 2 3 4 5 6 7 8 9 10

50

100

150

50

100

150

Nor

mal

ized

TL

Int

ensi

ty (

arbi

trar

y un

its)

Experimental Cycle

60°C 0.5 Gy60°C 1 Gy60°C 10 Gy60°C 50 GyNo annealing 0.5 GyNo annealing 1 GyNo annealing 10 GyNo annealing 50 Gy

Fig. 9. Recycling process for the natural fluorite.

H. Tugay et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 3640–3651 3651

curves. The recycling treatments of the annealed phosphors pro-duced the same results as the unannealed phosphors for relativelylower dose exposures. We deduced the effect of annealing from thereusability cycles of the phosphors for a high amount of dose of50 Gy.

Acknowledgements

This study was carried out at the Cukurova University (CU),Department of Physics. We are grateful to TUBITAK (Turkish Scien-tific and Technology Research Council) for its financial support un-der the Contract No. 105Y349 to purchase RISO TL/OSL DA-20equipment. We would like to acknowledge the CU Rectorate Re-search Unit for providing the financial support for this research un-der the Contract No. FEF2007YL22.

References

[1] S.W.S. McKeever, Thermoluminescence of Solids, Cambridge University Press,USA, 1985, p. 374.

[2] R. Chen, S.W.S. McKeever, Theory of Thermoluminescence and RelatedPhenomena, Word Scientific Publishing Co. Pvt. Ltd., 1997.

[3] J.H. Schulman, W.D. Compton, Colour Centers in Solids, Pergamon Press, NewYork, 1962, p. 368.

[4] J.R. Cameron, N. Suntharalingam, G.N. Kenney, Thermoluminescent Dosimetry,University of Wisconsin Press, Madison, WI, 1968.

[5] L.A. DeWerd, R.P. White, R.G. Stang, T.G. Stoebe, The relation betweendeformation and thermoluminescent defect centers in LiF (TLD 100), J. Appl.Phys. 47 (1976) 4231.

[6] K. Mahesh, P.S. Weng, C. Furetta, Thermoluminescence in Solids and ItsApplications, Nuclear Technology Publishing, Ashford, Kent, England, 1989, p.306.

[7] S.W.S. McKeever, M. Moscovitch, P.D. Townsend, ThermoluminescenceDosimetry Materials: Properties and Uses, Nuclear Technology Publishing,Ashford, Kent, England, 1995, p. 204.

[8] F.O. Ogundare, F.A. Balogun, L.A. Hussein, Evaluation of kinetic parameters oftraps in thermoluminescence phosphors, Radiat. Meas. 41 (2006) 892–896.

[9] R. Schayes, C. Brooke, I. Kozlowıtz, M. Lheureux, in: Proceedings of the FirstInternational Conference on Luminescence Dosimetry, Stanford. CONF 650637,AEC Symposium Series 8, NTIS, Springfield, VA, 1967, p. 138.

[10] S. Watanabe, E. Okuno, Thermoluminescence response of natural Brazilianfluorite to 137Cs gamma rays, in: Mejdahl (Ed.), Natural Fluorite Crystals V.Proceedings of the Third International Conference on Luminescence DosimetryHeld at Riso, 1, 1971, pp. 380–391.

[11] C.M. Sunta, Thermoluminescence of natural CaF2 and its applications, in:Proceedings of the Third International Conference on LuminescenceDosimetry, Held at Danish AEC Research Establishment Riso, 1971.

[12] M.A. El-Kolaly, C.M. Sunta, A.K. Ganguli, Effect of temperature on X-ray exitedluminescence of natural CaF2 and its implications in thermoluminescence, J.Lumin. 23 (1981) 423–443.

[13] A.M. Eid, M.I. El-Gohary, S.M. Kamal, Effect of heat treatments onthermoluminescence of naturally occurring materials, Radiat. Phys. Chem. 26(1985) 663–667.

[14] T. Calderon, M.R. Khanlary, H.M. Rendell, P.D. Townsend, Luminescence fromnatural fluorite crystals, Int. J. Rad. Appl. Inst. Part D Nucl. Track Rad. Meas. 20(1992) 475–485.

[15] L.A. DeWerd, L.K. Wagner, Characteristics of radiation detectors for adiagnostic radiology, Appl. Radiat. Isot. 50 (1999) 125–136.

[16] C.M. Sunta, A review of thermoluminescence of calcium fluoride, calciumsulphate and calcium carbonate, Radiat. Prot. Dosim. 8 (1984) 25–44.

[17] K.S.V. Nambi, Self-dose and fading of CaSO4 (Dy/Tm) and CaF2 (natural) TLphosphors in environmental radiation monitoring, Nucl. Instrum. Methods197 (1982) 453–457.

[18] F.O. Ogundare, F.A. Balogun, L.A. Hussain, Heating rate effects on thethermoluminescence of fluorite, Radiat. Meas. 40 (2005) 60–64.

[19] F.A. Hasan, Thermoluminescence as a function of dose in natural calciumfluoride: a proposed mathematical model, Nucl. Instrum. Methods Phys. Res.,Sect. B 12 (1985) 175–180.

[20] F.A. Balogun, J.O. Ojo, F.O. Ogundare, M.K. Fasasi, L.A. Hussein, TL response ofnatural fluorite, Radiat. Meas. 30 (1999) 759–763.

[21] M. Topaksu, A.N. Yazici, The thermoluminescence properties of natural CaF2

after beta irradiation, Nucl. Instrum. Methods Phys. Res. 264 (2007) 293–301.[22] R.P. Feynman, R.B. Leighton, M. Sands, The Feynman Lectures on Physics,

Addison-Wesley, 1964, pp. 142–144.[23] R. Muccillo, J. Rolfe, J. Phys. Status Solidi B 61 (1974) 579.[24] K.J. Kao, M.M. Perlman, Phys. Rev. B 19 (1979) 1197.[25] S. Ganguly, I.K. Kaul, Analysis of thermoluminescence glow peaks from natural

calcium fluoride, Mod. Geol. 8 (1984) 155–162.