A cleaning method to minimize contaminant luminescence signalof empty sample carriers using off-the-shelf chemical agents

Nikolaos A. Kazakis a,b,n, George Kitis b, Nestor C. Tsirliganis a

a Department of Archaeometry and Physicochemical Measurements, R.C. Athena, P.O. Box 159, Kimmeria University Campus, 67100 Xanthi, Greeceb Nuclear Physics Laboratory, Physics Department, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece

H I G H L I G H T S

� New and used empty sample carriers suffer from contamination from a fluorite and silica-related source.� A new cleaning method for empty sample carriers is proposed using off-the-shelf chemical agents.� The new method can eliminate any contamination from empty sample holders of various shapes and/or materials.� Contamination signals are reduced to the background level even at relatively high doses (100 Gy).

a r t i c l e i n f o

Article history:Received 2 September 2014Received in revised form21 October 2014Accepted 23 October 2014Available online 1 November 2014

Keywords:Sample carriersSpurious signalThermoluminescenceContaminationCleaning methodDegreaserKerosene

a b s t r a c t

Signals acquired during thermoluminescence or optically stimulated luminescence measurements mustbe completely free of any spurious and/or contamination signals to assure the credibility of the results,especially during exploratory research investigating the luminescence behavior of new materials.Experiments indicate that such unwanted signals may also stem from new (unused) and used emptysample carriers, namely cups and discs, which are widely used for such measurements, probably due tocontamination from a fluorite and/or silica-related source.

Fluorite and/or silicone oil appear to be the most likely sources of contamination, thus, their removal,along with any other possible source that exhibits undesirable luminescence behavior, is necessary.Conventional cleaning methods fail to eliminate such contaminants from empty cups and discs. In thiswork a new cleaning method is proposed incorporating off-the-shelf chemical agents.

Results of thermoluminescence measurements highlight the efficiency of the new cleaning process,since it can completely remove any observed contaminants from both new and used sample carriers, ofvarious shapes and/or materials. Consequently their signal is minimized even at relatively high beta-doses, where it is prominent, resulting in a clean and only sample-attributed signal.

& 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Signals acquired from luminescence measurements, such asthermoluminescence (TL) or optically stimulated luminescence (OSL),must originate only from the sample and be completely devoid of anyother interferences in order to lead to accurate results for the variousapplications (e.g. dating, accidental dosimetry, retrospective dosime-try), especially when investigating the dosimetric/luminescence prop-erties of new materials. However, various unwanted signals may alsobe observed and infect the sample luminescence signal. Some of them

are non-radiation-induced (spurious) and can be attributed to thesample itself or to the measuring conditions. In addition, contamina-tion of the sample with another material may also be possible andthe intensity of its signal (contaminant) depends on its luminescencesensitivity (Aitken, 1985).

Depending of course on the nature of the contaminant, when lowdoses are applied, e.g. during measurements with sensitive materials,the effect of the potential contaminant on the total acquired signalmay be completely obscured and unnoticed. However, when rela-tively high doses are employed (e.g. dating of geological materials,accidental dosimetry or post-sterilization dosimetry) (Kazakis et al.,2014) the unwanted signal is considerably enhanced due to its dose-dependent behavior and it can interfere with the genuine lumine-scence signal from the sample (Aitken, 1985). Since both signals,contaminant and sample-related, may have different dose response,

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/apradiso

Applied Radiation and Isotopes

http://dx.doi.org/10.1016/j.apradiso.2014.10.0210969-8043/& 2014 Elsevier Ltd. All rights reserved.

n Corresponding author at: Department of Archaeometry and PhysicochemicalMeasurements, R.C. Athena, P.O. Box 159, Kimmeria University Campus, 67100Xanthi, Greece. Tel.: þ302541078787, fax.: þ302541063656.

E-mail addresses: nikkazak@ceti.gr, nikkazak@gmail.com (N.A. Kazakis).

Applied Radiation and Isotopes 95 (2015) 226–232

the above would be of great importance in the case of materials atwhich dose response is low or the saturation point is reached at verylow radiation doses during luminescence measurements.

Besides the luminescence measurements in the various appli-cations where known dosimeters may be employed, existence andinterference of a contaminant signal is of major importance in thecase of research of exploratory nature, namely during the inves-tigation of the luminescence behavior of new materials. In thiscase, where the dose response (if any) of the under-study materialis unknown, a contaminant signal may be utterly misleading evenwhen low doses are applied. The researcher may not be able todistinguish between the real (stemming from the material) andthe contaminant signal, while the worst scenario would be toregard the contaminant signal as the product of a material whichdo not exhibit luminescence at all.

To the authors' best knowledge, less than a handful of studies havebeen conducted exploring the spurious and/or contaminant lumines-cence signal stemming from the empty sample carriers during TLand/or OSL. Vandenberghe et al. (2008) were the first to observespurious OSL signal in blank stainless steel cups when they weresprayed with silicone oil. They found that silicone oil, which isfrequently used to adhere a single layer of grains onto the carrier,may exhibit spurious signal and its undesirable interference isobserved even in “cleaned” cups which have a history of use withsilicone oil. Few years later, a systematic study on the existence ofparasite luminescence of commonly used aluminum and stainlesssteel sample carriers was conducted by Schmidt et al. (2011).Unwanted signals observed were attributed either to certain chemicalcompounds of the carrier material itself (oxides or carbides) or tosilicone oil residues resisting to conventional cleaning methods.

In the same respect, only recently, Simkins et al. (2013) conducteda similar study measuring empty new and used stainless steel cupswith OSL and proposed two alternative cleaning methods to reducethe intensity of the unwanted signals. In the above studies, research-ers found that the conventional cleaning methods employed formany years in their labs failed to eliminate the contaminant signalsfrom the sample carriers. In addition, alternative cleaning methodsproposed incorporate heavy-duty detergents, developed especiallyfor laboratory use, and/or hazardous chemicals (e.g. HF or methanol),as discussed later, all of relatively high cost.

To this direction, the scope of the present work is twofold. First,to make researchers aware of the contaminant signal coming fromthe sample carriers, which can be very important during explora-tory luminescence measurements, and to bring to their attentionthe need to test the cleaning method employed in their labs.Second, to test the efficiency of a new cleaning procedure, basedon low-cost and less hazardous off-the-shelf chemical agents, onnew and previously used cups and discs to minimize any con-taminant thermoluminescence signal of empty sample carriers,which is more prominent in relatively high doses.

2. Materials and methods

2.1. Instruments and methods

All measurements were conducted using a Riso TL/OSL reader(model TL/OSL-DA-15), equipped with a 90Sr/90Y beta particle sourcecapable of delivering a nominal dose rate of about 3.58 Gy/min atthe time of the measurements. A 9235QA photomultiplier tube witha combination of the appropriate filters, namely a Hoya U-340 withmaximum transmittance at approximately 340 nm and a heat absor-bing Pilkington HA-3 filter, was used for light detection.

Several beta-doses up to 100 Gy, were applied on every emptysample carrier before the TL measurement in order to recorda detectable (if any) contamination signal. All TL measurements

were performed in a nitrogen atmosphere with a constant heatingrate of 2 1C/s up to a maximum temperature of 500 1C, while thesampling frequency was 1 channel/1C, meaning that counts wererecorded every 0.5 s.

2.2. Sample carriers

To explore the luminescence behavior of empty sample carriersand validate the proposed cleaning method two different types ofcarriers were studied, commonly used in many laboratories:

1. Stainless steel cups supplied by RisØ (currently known as DTUNutech) with a diameter of 11.65 mm. Both used and new(unused) cups were studied to shed light on the source of thepotential contamination.

2. Aluminum flat used discs supplied by ELSEC with a diameter of9.7 mm. They were used to investigate the effectiveness of thenew cleaning method on aluminum as well.

From each kind of cups (new or used) and discs, more thantwenty (20) exemplars were measured. It should be noted that usedcups and discs have been recycled for several years, while cleanedbetween measurements with a conventional method as describedbelow. Thus, it was tested whether the new cleaning method (NCM)(described below) was capable of removing any residues afterapplying the conventional cleaning method (CCM). On the otherhand, new cups were first measured without any chemical orphysical treatment (called untreated hereafter) and when a con-taminant signal was detected both cleaning procedures were tested.

2.3. Sample carrier cleaning methods

To test the effect of the cleaning procedure on the contaminantsignals two different methods were employed:Conventional cleaning method (CCM)

Step 1: Rinse with distilled waterStep 2: Wiping of each cup/disc with kitchen paper

wetted with distilled waterStep 3: Rinse with distilled waterStep 4: 40-min Ultrasonic bath with distilled water

(starting at room temperature)Step 5: Rinse with distilled waterStep 6: Wiping of each cup/disc with kitchen paper

wetted with acetoneStep 7: Rinse with distilled waterStep 8: 40-min Ultrasonic bath with acetone (starting at

room temperature)Step 9: Carriers are left to dry in a desiccator

Till present, in certain cases, i.e., when silicone oil was used on adisc, scrubbing of its surface with sandpaper would also be appliedbetween Steps 4 and 5. However, such a mechanical treatment wasavoided in the case of the cups since any contamination or removedmaterial layers would probably be squeezed and trapped in the edgeand side walls of the cup's dent.New cleaning method (NCM)

Step 1: Rinse with distilled waterStep 2: 50-min Ultrasonic bath with household degreaser

(AJAXs Spray degreaser was used in the presentstudy) (starting at room temperature)

Step 3: Rinse with distilled waterStep 4: Scrubbing of each cup/disc with a toothbrush

soaked in kerosene (KERO-SUNs used as fuel inkerosene home/indoors heaters)

N.A. Kazakis et al. / Applied Radiation and Isotopes 95 (2015) 226–232 227

Step 5: Rinse with distilled waterStep 6: 50-min Ultrasonic bath with kerosene (starting at

room temperature)Step 7: Rinse with distilled waterStep 8: Carriers are left to dry in a desiccator

The degreaser employed in the new cleaning method is a house-hold product which is widely used for removing grease or oil stainsfrom surfaces or even clothes. In addition, the kerosene employed isextensively used as fuel in space-heating devices. The above chemicalproducts are much more advantageous than those used or suggestedby other researchers in terms of hazard, accessibility and cost, aselaborated below.

Conventional cleaning procedures for sample carriers employed inother laboratories may incorporate propanol or ethyl methyl ketone(Vandenberghe et al., 2008), ethanol (Simkins et al., 2013) or acetone(Schmidt et al., 2011). In addition, new suggested cleaning methodsmay involve the use of HF and/or methanol. All the above chemicalsare more hazardous than kerosene according to their safety andhealth specifications and their classifications by the Directive 67/548/EEC. More specifically, all are very volatile, with their vapor pressurebeing considerably higher than that of kerosene, while they are alsoextremely flammable, since their flash points are very low (lowerthan 15 1C), contrary to that of kerosene (higher than 40 1C). Inaddition, methanol and HF (though not flammable) are classified ashighly toxic and, thus, extreme caution is required when handlingthem. According to the above, kerosene seems much safer and lesshazardous for indoors use.

At the same time, suggested cleaning methods for samplecarriers (Simkins et al., 2013), along with few of the above chemicals,incorporate special detergents for laboratory use only, which can bepurchased by specialized suppliers only. On the other hand, kero-sene (and the AJAXs household degreaser) can readily be found inthe market at a significantly lower cost compared to the previouslymentioned laboratory solvents.

The above suitability of kerosene for laboratory use due to itsadvantages is also recommended by other investigators (e.g. Ajayiet al., 2011). Ajayi et al. (2011) suggest that kerosene, which ischeaper, non-volatile, non-toxic and easily available, can be usedas defattening agent in osteological preparations so as to preventthe bones from looking dirty, attracting dust and also to makebleaching easier, as a substitute for acetone, which is very costlyand not readily available.

3. Results and discussion

3.1. Cups

In the case of the cups, all (twenty) of the studied used cupsgave an appreciable signal after irradiation even with low beta-doses. On the other hand, in the case of the new (untreated) cupsthe percentage of those which exhibited a contaminant signal wasabout 80% (of the 45 new cups in total) after irradiation with beta-dose of 100 Gy, while no signal above background was detected fornon-irradiated cups (and discs as well).

Fig. 1 illustrates eight of the signals acquired in the case of theused cups after irradiation with 100 Gy. It is obvious that the shapeof all glow curves is similar exhibiting three major peaks, indica-tive of the same contamination source as discussed later.

In the same respect, Fig. 2 presents typical TL signals acquiredfrom the various types of empty cups which gave a detectablesignal above the background under different treatment conditionsand irradiation with a beta-dose of 100 Gy. At this point, it must benoted that in all figures, markers on the curves are used for the

sake of clarity, for illustration purposes (as also indicated in thecorresponding legends), and they are not related to the samplingfrequency, i.e., data points recorded during the TL measurements.

A first observation from Fig. 2 is that in the case of the cups(used and new) the TL glow curve in all cases has a similar shape(as previously stated) with three peaks of center temperature inthe range:

� 95–105 1C,� 180–200 1C and� 285–315 1C.

This is a strong indication that the contamination source is thesame in all cases, since no other (of different shape) signal wasobserved for the studied stainless steel cups, while the intensity ofthe signal varies and seems to depend on the history (usage) of

Fig. 1. TL glow curves for eight used cups studied after irradiation with a beta-doseof 100 Gy.

Fig. 2. Typical TL glow curves for the various cups studied after irradiation with abeta-dose of 100 Gy.

N.A. Kazakis et al. / Applied Radiation and Isotopes 95 (2015) 226–232228

each cup. In general, used cups produce considerably higher signalintensities (e.g. a peak amplitude of approximately 226,000 countsfor the third peak after irradiation with 100 Gy) than new cups(untreated or cleaned with CCM). The above is in accordance withthe observations of Schmidt et al. (2011) who found that used andcleaned Al discs produce much higher spurious and regeneratedsignal intensities in the UV and blue detection range than new Alcups. Thus, the source of contamination shows a cumulative trendon the cups' surface and cannot be removed with the CCM.

It should also be mentioned, that besides the intensity, variationsmay also exist in the glow curves of the various cups (new or used)regarding the shape of the peaks. In some cases, peaks may appearwider, while in other cups a shift of the complete glow curve is alsoobserved when compared to the glow curves of other cups. Theabove could possibly be attributed to other contaminants presenton the cups, besides the common one that dominates in all, whosesource might be different from cup to cup.

In addition, the fact that even new untreated cups wouldexhibit an appreciable luminescence behavior is of great interest,since in this case the contamination signal evidently stems from asource present in the manufacturing procedure.

Although it is beyond the scope of the present work, suggestionsregarding the nature of the observed contamination can be madebased on the shape of the observed TL signals (Figs. 1 and 2). Therecorded TL glow curves of all contaminated cups are quite similarto that of fluorite (CaF2) (e.g. Polymeris et al., 2006) in terms ofshape and characteristic temperatures of the observed glow-peaks.

In addition, the characteristic “110 1C” peak of quartz, which inthe case of the natural quartz and its family varies between 89 1Cand 150 1C under the various physical conditions of the specimens(e.g. Yang and McKeever, 1990), is also present in the contaminantsignals. Moreover, Carvalho et al. (2010), who studied the effect ofparticle size in the TL response of natural quartz, observed similarglow curves to the ones of the present study after irradiation with aγ-dose of 500 Gy, namely three peaks near 90, 215 and 325 1C. Theabove indicates that silica may also be involved in the contaminantsignal observed in the present study.

According to the above observations, it could be concluded thatthe contamination signal of the cups may be attributed to thesingle or joint luminescence activity of fluorite and silica. Con-tamination of used cups with fluorite may be ascribed to pastluminescence experiments in the lab and, since the CCM failed toremove it, this mineral might have been accumulated on the cups.In the case of the new cups, contamination with fluorite could alsobe attributed to the manufacturing process, since this mineral isused in emery wheels, which are employed for the mechanicalcleaning and polishing of metal surfaces (e.g. Chakhaiyar, 2010).Thus, it is possible that tiny shards of such wheels may contam-inate new sample carriers during their production.

In the same respect, the assumption of a silica-related contam-ination corroborates the findings of Vandenberghe et al. (2008)and Schmidt et al. (2011) and supports the theory that a potentialcontamination source may be silicone oil, which is extensivelyemployed during luminescence measurements, justifying the con-tamination of the used cups. In addition, silicone oil is also usedduring the manufacturing of new sample carriers for cooling andlubrication, thus explaining the contamination of the new cups aswell. Silica (SiO2) can be formed from thermal decomposition ofsilicone oil in two ways depending on the environmental conditions(e.g. Shin-Etsu, 2011):

� In an inert gas atmosphere (e.g. N2), combustion occurs around450 1C, thermal decomposition of silicone oil takes place and asilica residue remains.

� In air, oxidation of Si may occur at high temperatures (4150 1C)also leading to the formation of silica.

The former can take place during routine thermoluminescencemeasurements, where the temperature of the sample carrier isincreased up to 500 1C in nitrogen atmosphere, while the latter canoccur during an annealing process in an oven or during thermolumi-nescence measurements up to 200 1C (where N2 flow can be omitted).

In order to gain an insight into the luminescence behavior of thecontamination, the dose response was also investigated for variousbeta-doses up to 100 Gy. The acquired signals for the various dosesapplied are illustrated in Fig. 3 in the case of one of the cups. Itshould be noted that the inset presents the zero-dose signal(background) magnified. As previously mentioned, no signal wasdetected for non-irradiated sample carriers. It is obvious that thereis no shift of the temperature range of the three major peaks as thedose increases, while their area seems to increase in a proportionalmanner with the dose applied. To test the above, the dose responsefor all peaks observed is depicted in Fig. 4. It must be noted that all

Fig. 3. TL glow curves for one of the used cups studied after being irradiated withseveral doses (dose response); the inset illustrates the zero dose (0 Gy) signalmagnified.

Fig. 4. Normalized dose response of the area of the three peaks for the empty cups;values have been normalized with respect to that of the lowest dose studied (10 Gy) foreach peak; the inset illustrates the dose response of the actual area of the three peaks.

N.A. Kazakis et al. / Applied Radiation and Isotopes 95 (2015) 226–232 229

data (peak areas) were normalized with respect to the area of therespective peak of the lowest dose studied (10 Gy), while the insetillustrates the actual (non-normalized) dose response of the area ofeach one of the three peaks.

From Fig. 4, it is evident that all peaks (area) of the contaminantsignal are almost directly proportional to the dose (in the rangestudied) with only slight differences in the rate of increase. In fact,the above dose response can be fitted with a linear function of theform: Normalized peak areaE0.1 �dose (in Gy), with R2 more than0.999 for all three peaks. As a result, due to the above dosedependent behavior of the contaminant signal, in the case of a

material with low dose response or with a low dose saturation point,the ratio of the two signals, i.e., contaminant and material-related,would constantly increase with the dose, leading to erroneous resultsand conclusions. The above constitute further evidence whichenhance the need to completely remove the contamination fromthe cups and eliminate the related signal, since its effect would be ofmajor significance especially when investigating new/unknownmaterials or of low sensitivity.

Based on the above, the nature of the silicone oil and thephysical properties (solubility) of fluorite, proper off-the-shelf che-mical agents, i.e., a degreaser and a strong solvent, were recruited for

Table 1Effect of NCM on the area of the three peaks of the TL glow curves illustrated in Fig. 5.

New cup-untreated New cup cleaned with CCM Used cup

Original After NCM Reduction (%) Original After NCM Reduction (%) Original After NCM Reduction (%)

First peak 934,717 2987 99.68 635,669 5920 99.07 6,833,797 12101 99.82Second peak 285,324 2037 99.29 220,785 3935 98.22 2,002,143 4672 99.77Third peak 2,243,475 5051 99.77 1,732,417 11164 99.36 15,883,841 22101 99.86

Values are expressed as the sum of the counts of all channels covering the corresponding temperature range of each peak (1 channel/1C/0.5 s).

Fig. 5. Typical TL glow curves for the various cups studied (after irradiation with 100 Gy) before and after employing the NCM: (a) new cup untreated, (b) new cup treatedwith CCM first, and (c) used cup; insets illustrate the TL glow curve after NCM treatment magnified.

N.A. Kazakis et al. / Applied Radiation and Isotopes 95 (2015) 226–232230

the proposed NCM as previously described. Fig. 5 shows the effect ofthe NCM on the contamination signal for three of the stainless steelcups studied illustrating both the original and the after NCM signal,while the latter is also depicted magnified at the inset. Counts of thearea of the peaks of the selected signals presented in Fig. 5 aresummarized in Table 1 for clarity and comparison purposes regard-ing the cleaning efficiency of the NCM. It should be noted that theobserved trend (related to the peak area reduction after cleaningwith NCM) is typical for all cups studied.

From Fig. 5 and Table 1 it is evident that the NCM can almostcompletely remove the contamination source and eliminate its signalin all cases with most outstanding example that of the used cup. It isobvious that used cups, which have been cleaned for several yearswith the CCM, remain extensively contaminated, pointing out theinsufficiency of this cleaning method. In most cases, the CCM cannot

remove the contamination signal not even in the new cups. Theabove should be expected, since acetone belongs to the solventswhich only partially dissolve silicone fluids (e.g. CLEARCO, 2012;Shin-Etsu, 2011). Similar conventional cleaning methods have alsobeen proven very poor in removing spurious and/or contaminationsignals (Vandenberghe et al., 2008; Schmidt et al., 2011; Simkinset al., 2013), since they also used solvents of low or negligibleefficiency in dissolving silicone oil (e.g. acetone, ethanol, methanol).On the other hand, kerosene, among others, is one of the solventsin which silicone oil is completely soluble (e.g. CLEARCO, 2012; Shin-Etsu, 2011).

However, it was noted that in a limited number of cases (lessthan 3%) the conventional method was capable of removing thesilicone oil and/or fluorite from new cups. Apparently, the tinygrains of fluorite and/or silicone oil used during the manufacturingof these cups lie on the surface without having been extensivelydecomposed or chemically bonded to the carrier material, some-thing that is likely to occur after repeated cycles of radiation-heating (Schmidt et al., 2011). In any case, the NCM seems to bevery effective, since it also removes any residue from the cupswhen the CCM proves inadequate. It should also be noted thateven if the contamination source(s) is(are) different than thesuspected one(s), as previously described, NCM's efficiency cannotbe disputed, but it can further be enhanced, since this would meanthat it can remove any contamination accumulated on the cups'surface after numerous experiments with various materials duringthe previous years in the lab.

The resulting signal, after applying the NCM, in the case ofhighly contaminated cups (e.g. Fig. 5c) although not exactly at thebackground level, it is practically negligible, since one should keepin mind that it refers to a thermoluminescence signal after a beta-dose of 100 Gy. In addition, the percentage reduction of the peakarea of the signal after applying the NCM is stunning (Table 1).Nevertheless, tests have showed that a second treatment withNCM may further reduce the contaminant signal and bring itcloser to the background level.

3.2. Discs

The NCM was also tested in empty aluminum used discs.However, in this case the most dominant signal observed in themajority of the discs (80%) studied consists of merely the firstpeak, namely the characteristic “110 1C” peak of quartz. The rest ofthe discs exhibited contamination signals similar to that of thecups, as previously discussed. This silica-related behavior is illu-strated in Fig. 6, which presents the dose response of the useddiscs. As already mentioned, no signal above background wasdetected for non-irradiated discs.

Although the contamination signal may differ in general fromthose reported earlier in the case of the cups, due to the presence ofaluminum oxide, however, an indication of quartz residues is stillpresent. The fact that the other two peaks (as in the case of thecups) are not present in the thermoluminescence glow curves of thediscs indicates a fluorite-free and a more silica-related contamina-tion. This may be attributed to the better cleaning of the discscompared to the cups, though employing the CCM, due to their flatand dent-free shape. Their shape facilitates the easier and bettercleaning of their surface, since even wiping of it with paper allowsthe removal of a great portion of any material, as they cannot betrapped in the walls as in the case of the cups. Consequently, in thecase of the discs, removal of the silicone oil seems to be the greatestchallenge, since its liquid state and nature makes it more adhesive,and thus harder to be removed, to the disc surface. Moreover, sincethe discs are made of aluminum, another potential contaminationwould be Al2O3, since layers of it may be formed on the disc surfacefrom an oxidation process.

Fig. 6. TL glow curves for one of the used discs studied after being irradiated withseveral doses (dose response); the inset illustrates the zero dose (0 Gy) signalmagnified.

Fig. 7. Normalized dose response of the area of the “110 1C” peak for the emptydiscs; values have been normalized with respect to that of the lowest dose studied(10 Gy); the inset illustrates the dose response of the area of the peak.

N.A. Kazakis et al. / Applied Radiation and Isotopes 95 (2015) 226–232 231

Fig. 7 illustrates the normalized dose response for the observedpeak normalized with respect to its area at the lowest dose studied(10 Gy), while the inset illustrates the actual (non-normalized) doseresponse of its area. According to Fig. 6, the area of the peak seems toincrease with the applied dose. However, Fig. 7 implies that the abovedose response is not linear as in the case of the cups. In this case, thedose response can perfectly be fitted by an exponential function (I¼a[1�e�D/b], where I and D are the TL intensities in counts/0.5 s andthe dose applied in Gy, respectively, while a and b are constants) forthe dose range studied, while the increase rate of the resulting peakarea is much lower than the one observed in the cups (for the samepeak of the 110 1C). In addition, it seems that the saturation level isnear the 100 Gy, although further measurements are required tocertify this statement. In any case, the different behaviors of thecontaminant signal observed in the discs could be an indication of adifferent contamination sources than in the cups. An assumptionwould be that the cups suffer from the joint action of fluorite andsilicone oil, while discs only from the single action of silicone oil.

The cleaning efficiency of the NCM on the empty Al discs isillustrated in Fig. 8. Fig. 8a illustrates the acquired signal of an emptyAl disc after irradiation with 100 Gy, in which the CCM has beenrepetitively employed for several years. The TL glow curve of thesame disc after application of the NCM is also depicted in the samefigure (Fig. 8a). Once again the NCM demonstrates a high cleaningefficiency eliminating almost entirely the contamination signal.

In the same respect, the CCM was also applied in Al discsfollowed by scrubbing of their surface with sandpaper to mechani-cally remove any residues and Al2O3 layers. A recorded TL signalafter irradiation with 100 Gy is given in Fig. 8b. In order tohighlight the efficiency of the NCM a contaminant signal in whichall three peaks are detected has been selected.

It is obvious that the intensity of the three peaks is practicallynegligible indicating that scrubbing may potentially contribute toa great extent to the elimination of the contamination signal, asalso observed by Schmidt et al. (2011). Yet, such a mechanicaltreatment cannot easily be applied in the non-flat cups andactually it is not recommended due to their dented shape, sincescrap metal along with any contaminants may be squeezed andeventually end up embedded in the walls.

Despite the effectiveness of the scrubbing of the flat disc withsandpaper, the disc was also subjected to the NCM. The acquired

signal is also illustrated in Fig. 8b. It is evident that the NCM isefficient enough to reduce even more the TL contamination signaland diminish its intensity to that of the background level.

4. Conclusions

In the present study, a new cleaning method based on off-the-shelf chemical agents is proposed for the cleaning of empty samplecarriers which can suffer from spurious and/or contamination signals.The method combines high efficiency with low cost employing ahousehold degreaser and kerosene. The suggested method proves tobe of extremely high efficacy for various shapes and/or materials ofthe carrier, while it can eliminate contamination signals of emptycarriers, where conventional methods fail, especially when highirradiation doses are employed in the measurement ensuring thatthe acquired signal is free of any interference from the carrier itself.

References

Aitken, M.J., 1985. Thermoluminescence Dating. Academic Press, London.Ajayi, T., Ijisan, O., Huthman, I.O., Olaseinde, O.F., Adenowo, T.K., Adefule, A.K.,

Huthman, A.S., 2011. Kerosene: a substitute for acetone in bone preparations.Internet J. Biol. Anthropol. 4 (2).

Carvalho, A.B., Guzzo, P.L., Sullasi, H.L., Khoury, H.J., 2010. Effect of particle size inthe TL response of natural quartz sensitized by high dose of gamma radiationand heat-treatments. Mater. Res. 13 (2), 265–271.

Chakhaiyar, H., 2010. Periwinkle Environmental Education Part X, Third editionJeevandeep Prakashan Pvt. Ltd., Mumbai.

CLEARCO, 2012. Solubility of Silicone Fluids. Available from: ⟨http://www.clearcoproducts.com/pdf/library/Solubility1.pdf⟩. (accessed 10.08.14).

Kazakis, N.A., Tsirliganis, N.C., Kitis, G., 2014. Preliminary thermoluminescence andoptically stimulated luminescence investigation of commercial pharmaceuticalpreparations towards the drug sterilization dosimetry. Appl. Radiat. Isot. 91,79–91. http://dx.doi.org/10.1016/j.apradiso.2014.05.012.

Polymeris, G.S., Kitis, G., Tsirliganis, N.C., 2006. Correlation between TL and OSLproperties of CaF2:N. Nucl. Instrum. Methods Phys. Res. B 251, 133–142.

Schmidt, C., Kreutzer, S., Fattahi, M., Bailey, R., Zander, A., Zöller, L., 2011. On theluminescence signals of empty sample carriers. Anc. TL 29 (2), 65–74.

Shin-Etsu, 2011. Silicone Fluid, Technical data. Available from: ⟨www.shinetsusilicones.com/files/dmf_et.pdf⟩. (accessed 12.05.14).

Simkins, L.M., DeWitt, R., Simms, A., 2013. Methods to reduce sample carriercontamination for luminescence measurements. Anc. TL 31 (1), 19–28.

Vandenberghe, D.A.G., Jain, M., Murray, A.S., 2008. A note on spurious lumines-cence from silicone oil. Anc. TL 26 (2), 29–32.

Yang, X.H., McKeever, S.W.S., 1990. Point defects and the pre-dose effect in quartz.Radiat. Prot. Dosim. 33 (1–4), 27–30.

Fig. 8. Typical TL glow curves for the Al used discs studied (after irradiation with 100 Gy) before and after employing the NCM: (a) used disc treated with CCM first; insetillustrates the TL glow curve after NCM treatment magnified and (b) used disc treated with CCM followed by scrubbing with sandpaper first.

N.A. Kazakis et al. / Applied Radiation and Isotopes 95 (2015) 226–232232