Different methods for tritium determination in surface water by lsc
Applied Radiation and Isotopes 71 (2013) 51–56 Contents lists available at SciVerse ScienceDirect Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradisoDifferent methods for tritium determination in surface water by LSCJovana Nikolov a,n, Natasa Todorovic a, Marija Jankovic b, Marija Vostinar a, Istvan Bikit a,Miroslav Veskovic aa University of Novi Sad, Faculty of Sciences, Department of Physics, Serbiab University of Belgrade, Institute Vinca, Radiation and Environmental Protection Department, SerbiaH I G H L I G H T Sc 3 methods of water sample preparation for low-level tritium measurement have been compared.c All prepared samples were measured by LSC Quantulus 1220.c The most novel is that Sample Oxidizer is used as one method for sample preparation.c Paper presents MDA values for each presented method.c For environmental monitoring all three methods give comparable results.a r t i c l e i n f o abstractArticle history: The main aim of this paper was to compare different methods of preparing water samples for tritiumReceived 26 February 2012 analysis by ultra-low-level background liquid scintillation counter, Quantulus 1220. Three methods ofReceived in revised form sample preparation for low-level tritium measurement have been implemented in the Nuclear Physics18 September 2012 Laboratory in Novi Sad: electrolytic enrichment, direct method without electrolytic enrichment andAccepted 20 September 2012 sample Oxidizer 307 method. The examined fresh water samples were rainfall collected duringAvailable online 4 October 2012 ˇ 6 months and water from a stream in the Vinca nuclear research center collected over 3 months.Keywords: The obtained results using these three methods showed satisfying agreement. The appropriateLiquid scintillation counting measuring time by LSC for each sample prepared according to different methods has been determined.Tritium measurement & 2012 Elsevier Ltd. All rights reserved.Sample oxidizer1. Introduction H2O). Tritiated water can replace ordinary water in living cells (approximately 70% of the soft tissue in the human body is water). Tritium is a radioactive isotope of hydrogen with a physical half- Once in living cells, tritium can replace hydrogen in the organiclife of 450078 days (Lucas and Unterweger, 2000) that behaves like molecules in the body. Thus, despite tritium’s low radiotoxicity instable hydrogen and is usually found attached to molecules repla- gaseous form and its tendency to be released from the body rathercing hydrogen (Varlam et al., 2011). This radionuclide occurs in rapidly as water; its effects on health are made more severe by itsnature, originating from natural and anthropogenic sources. Natural property of being chemically identical to hydrogen (Varlam et al.,tritium is produced in the atmosphere from the interaction of 2011).cosmic radiation with atmospheric nitrogen (Madruga et al., 2008). As tritium is a very soft beta-emitter (maximum ener-Anthropogenic production has disturbed the natural levels of tritium gy¼18.6 keV), the common methods for low-level counting ofby nuclear weapons tests and, in addition to that tritium is being tritium are either gas-proportional counting (GPC) or liquid scintil-released into the atmosphere through weapons manufacturing, the lation counting (LSC). For analyzing natural water samples, it isoperation of nuclear power plants and reprocessing of nuclear fuels better to use LSC because in this method the water sample is(Pujol and Sanchez-Cabeza, 1999). directly combined with an appropriate aqueous scintillation cocktail, Tritium most commonly enters the environment in gaseous form the required pre-treatment is minimal and the counting efﬁciency is(T2) or as a replacement for one of the hydrogen atoms in water higher than that of GPC (Noakes and De Filippis, 1988).(HTO, called ‘‘tritiated water,’’ instead of ordinary, non-radioactive The primary objective of all sample preparation methods is to obtain a stable homogeneous solution suitable for analysis by Liquid Scintillation Counting (LSC). There are no absolutes in n Corresponding author. sample preparation; whichever method produces a sample that E-mail address: firstname.lastname@example.org (J. Nikolov). lends itself to accurate and reproducible analysis is acceptable.0969-8043/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.apradiso.2012.09.015
52 J. Nikolov et al. / Applied Radiation and Isotopes 71 (2013) 51–56However, there will be occasions when more than one method sample spectrum being recorded in the other half of the MCA.will be both suitable and available and the selection of either Quenching was quantiﬁed with the external standard quenchingmethod will depend on other factors (Oxidizer Application Note, parameter, SQP(E), which was used to determine the counting2002). efﬁciency of the system with appropriate calibration curves (Pujol According to the European Commission, the upper limit for and Sanchez-Cabeza, 1999). We determined the counting win-tritium in water is 100 Bq l À 1 (European Commission, 1998). This dows for each of the three methods according to the resultsvalue is not based on health effects relative to its consumption obtained by measuring standard samples.but more as a monitoring value. A tritium activity of 100 Bq l À 1could indicate that leakage or a release occur on a power plantand further analysis are then realized to check if other radio- 3. Sample preparationnuclides are present in water. In considering health concerns, theactivity value that is commonly used as a limit is the one from the 3.1. Method of electrolytic enrichmentWorld Health Organization (10,000 Bq l À 1 for a man of 70 kg whodrinks 2 l of water per day) published in the 4th edition of their After collecting 500 ml of water, samples were distilled toguidelines for drinking water (WHO, 2011). remove any impurities, to reduce quenching and to prevent the Three methods (electrolytic enrichment, direct method and introduction of other radionuclides which might adulterate theoxidizer/LSC method) for tritium concentration determination in results.water samples have been compared in this paper. In order to increase the tritium concentration to an easily The water samples were taken from the Mlaka creek which measurable level, electrolytic enrichment must be appliedpasses through the Vinca Institute of Nuclear Science in Belgrade, (Kaufman and Libby, 1954). When the water is decomposed bySerbia. The Mlaka creek is 5 km long with an average water ﬂow electrolysis, the deuterium and tritium content in the resultantof only 10 l s À 1. The sampling point was situated near the hydrogen is lower than in the electrolyte and, as the electrolysisdischarge from research reactor RA (Miljevic et al., 2000). The proceeds, the tritium concentration in the remaining electrolyteincreases in content can be attributed to a tritium contribution increases (Cameron, 1967). The enrichment process was con-due to liquid release through inﬁltration from the reactor target ducted in an apparatus containing electrolytic cells with co-axialstorage basins into the catchments area. The other sampling stainless steel electrodes. In each cell, 250 ml of previouslylocation was the Zeleno Brdo reference meteorological station distilled water with 1 g of Na2O2 (an electrolyte which forms(441470 N, 201320 E, altitude 243.2 m) for collecting rainfall sam- NaOH in water and yields an alkaline medium favorable toples. Vinca Institute of Nuclear Sciences (VINS) is situated on the electric conduction and hence to the reaction of water electro-Danube, 7 km SE of Belgrade which is potential source of lysis) was electrolyzed. Each electrolysis run comprised 2 refer-environmental dispersal of anthropogenic tritium primarily as ence dilutions of tritiated water and the rest were the samples.HTO from the long-established nuclear facilities. In spite of the Electrolysis was carried out at a current of 5 A (voltage of 48 Vcessation of normal reaction operations at VINS in 1986, this is across each cell) until the volume of electrolyte was reduced topotentially still a source of tritium which will continue to disperse 25–30 ml. All runs were carried out at a temperature of 2–51 C tointo the open environment (Miljevic et al., 2000). The last prevent the loss of tritiated water molecules by evaporation. Afterpublished results were before the year 2000; this paper will also electrolysis the samples were distilled to separate the water fromprovide a new data of activity concentration of tritium in the electrolyte.precipitation and in the water from Mlaka creek and that could An aliquot of 8 ml from the second 10 ml fraction of the distillateindicate a future need for continual monitoring of this area. was placed in a 20 ml polyethylene scintillation vial, and mixed with 12 ml of OptiPhase HiSafe scintillation cocktail from Perkin Elmer. Neither glass vials nor Teﬂon-cooper vials were used because of2. Measuring technique their relatively high background in the tritium window and their high cost for routine monitoring (Kaihola, 1993). The tritium activity2.1. Liquid scintillation counter Quantulus 1220 of water samples was determined by LSC Quantulus 1220. The counting time was 90 min in 5 cycles for each sample and the All prepared samples were measured by a Quantulus 1220 tritium activity was calculated by averaging the results. The mea-liquid scintillation counter, which is a low-level background surement of the background noise is very important when countingliquid scintillation counter (LSC) manufactured by Perkin Elmer, samples of low activity. As a background sample, distilled raw waterFinland (Quantulus, 1220,2002). This instrument has its own (deep well water) was used. With each set of samples a countingbackground reduction system around the vial chamber, which blank was included. Spiked water of known tritium activity con-consists of both an active and passive shield. The passive shield is centration is used to determine the enrichment factor. The enrich-made of lead, copper and cadmium and the active shield is based ment factor, f, for electrolytically concentrated samples is deﬁned ason a mineral oil scintillator. Low-activity materials were used in the ratio of the concentration after enrichment, SW (spiked water),the construction of the Quantulus, so it is useful for measuring of a standard and the concentration before enrichment (BE) (Sauzaylow-level radiation activity. The system is provided with two and Schell, 1972).pulse analysis circuits that are accessible for the users: a pulse Tritium activity concentration of each enriched sample isshape analysis (PSA) and pulse amplitude comparator (PAC) calculated according to the following expression (Madrugacircuit. There is also a delayed coincidence circuit (DCOS) inside et al., 2008):the Quantulus, which is useful for the correction of chemilumi- Ra ÀRb 1 À1 A Bq l ¼ ð1Þnescence. The Quantulus 1220 has two multichannel analyzers e Â V Â eÀlt ÀV sBE =V sf Áf(MCA), one is used for active shield and the second one is used forspectra record. Those MCA’s are divided in two halves, 1024 where Ra—count rate of the sample (cps), Rb—count rate of thechannels each. The tritium conﬁguration of the MCA’s setting background (cps), e—counting efﬁciency, V—volume of the sam-eliminates the random noise of phototubes, inhibits the coin- ple in the container vial (l), VsBE—volume of the sample beforecidence pulse from the guard and the sample, and monitors the enrichment (ml), Vsf—volume of the sample after enrichmentrandom coincidence by DCOS in a half of the MCA, the whole (ml), f— enrichment factor, l is decay constant for tritium
J. Nikolov et al. / Applied Radiation and Isotopes 71 (2013) 51–56 53(l ¼ ln 2 ; T1=2 ¼ 4500 78 days, half life of tritium) and t is T1=2 (the obtained efﬁciency of our system is 23.25%), F is recoveryelapsed time between sampling and counting (in days). factor (in our measurements it was determined to be 0.9), V is volume of the sample aliquot (8 ml), l is decay constant for3.2. Method for direct measurement of tritium in water samples tritium (l ¼ ln 2 ; T1=2 ¼ 4500 days, half life of tritium) and t is T1=2 elapsed time between sampling and counting (in days). In our laboratory we performed a rapid tritium activity The recovery correction factor is calculated according to ASTMdetermination method by LSC Quantulus 1220. This technique Standard Test Method (ASTM D4107-08, 2006):involves mixing the sample with a proper detection cocktail to be RDWTS ÀRbcounted in a liquid scintillator. F¼ ð4Þ e Â ARWTS As a background sample we used distilled raw water (deepwell water) prepared according to the ASTM Standard Test where RDWTS is the distilled water tritium standard count rateMethod (ASTM D4107-08, 2006). We used the commercial stan- (cps), Rb is the background aliquot count rate, in countsdard of tritium activity from Perkin Elmer, and prepared it with a per second (cps), and ARWTS is the activity of undistilled rawscintillation cocktail. The preparation method of background and water tritium standard, in becquerels.standard should be the same as the preparation of samples. Inevery set of measurements it is important to measure background 3.3. Sample combustion methods and suitabilityand standard, to obtain the same measuring conditions. Eachwater sample was ﬁltered through a slow depth ﬁlter and then The Model 307 PerkinElmer Sample Oxidizer (Fig. 1) is adistilled. After the distillation, the samples were mixed with the sample preparation system for single and dual labeled 3H andOptiphase HiSafe 3 scintillation cocktail. 14 C samples. Automatic operation produces consistently repro- In this study we used standard 20 ml polyethylene vials. A ducible, high quality samples for analysis in liquid scintillationtotal of 8 ml of distilled water was transferred to a polyethylene counters (Model 307 PerkinElmer Sample Oxidizer, 2003; Cookvial, mixed with 12 ml of scintillation cocktail and stored for one et al., 2003). The sample is combusted in a continuous ﬂow ofday in the liquid scintillation counter sample holder to allow full oxygen, forming water and carbon dioxide. Oxidation eliminatesdecay of chemiluminescence and photoluminescence. color quenching as well as reduces background counts and It can be noticed that reduced detection efﬁciency may result variation in chemical quenching.from quenching in the sample scintillator mixture. The main Advantages of sample combustion are: sample processing timereasons for quenching are different types of impurities in the is rapid, sample can be wet, dry or freeze-dried, any samplesample, which can inhibit the transfer of energy, or colored containing H and/or C can be combusted, it is ideally suited formaterials, which may absorb some of the emitted light. This is both single and dual label 3H and 14C, sample sizes up to 1.5 g areespecially the case with direct methods, like the method we used, possible, radioactive recovery is excellent ( 497%), memory effectwithout electrolytic enrichment. There are many ways for count- iso0.08%, there is no loss of radioactivity by volatilization, thereing efﬁciency determination, but for the Quantulus the easiest is is no chemiluminescence interference and there is no colorto apply the external standard method—the use of a quench quench interference (Model 307 PerkinElmer Sample Oxidizer,indicating parameter such as SQP(E) via calibration curves and 2003; Cook et al., 2003).internal standard method. According to our test method (ASTM Disadvantages of sample combustion are: initial capital invest-D4107-08, 2006), distillation after alkaline permanganate treat- ment, it is only suitable for 3H and 14C, need a gas supply (oxygenment, eliminates quenching substances, as well as radionuclides and nitrogen), must be operated in a fume hood and reagents arewhich might be present in a volatile chemical form such as corrosive and ﬂammable (Model 307 PerkinElmer Sampleradioiodine or radiocarbon. Oxidizer, 2003; Cook et al., 2003). We calculated the minimum detectable activity for different The combustion system burns the sample in an oxygen atmo-periods of time according to the Standard Test Method (ASTM sphere. The sample is ignited by the ignition basket, which is aD4107-08, 2006). The minimum detectable activity (MDA), platinum coil capable of developing high temperatures. Theachieved was 2.1 Bq l À 1 for a total counting time of 300 min. combustion time is set by the operator and is determined byThe aim was to perform a rapid method, so we try to measure for the sample material and size.total counting time of 90 min for directly prepared sample. The During the combustion process, all isotopes of hydrogen,obtained MDA for that interval of time was 3.5 Bq l À 1. including tritium, are oxidized to ‘‘tritiated water’’. Due to the The counting efﬁciency for each analysis was determined high combustion temperatures, this tritiated water is in the formusing a tritium standard (DPM 3H Spec (Perkin Elmer) with an of steam. The steam is condensed in an air cooled condenser andactivity of 1.48 Â 106 dpm/ml70.82% on 9 July 2008.) using thefollowing equation: RDWTS ÀRbe¼ ð2Þ ADWTSwhere RDWTS is the distilled water tritium standard count rate (cps),Rb is the background aliquot count rate, in counts per second (cps),and ADWTS is the activity of the distilled water tritium standard, inbecquerels. The calculated efﬁciency was (23.2570.05) %. The tritium activity concentrations were calculated accordingto the formula from ASTM Standard Test Method (ASTM D4107-08, 2006): Ra ÀRb À1A Bq l ¼ ð3Þ e Â F Â V Â eÀltwhere Ra is the sample aliquot gross count rate (cps), Rb is thebackground aliquot count rate (cps), e is the detection efﬁciency Fig. 1. Combustion of organic samples by ﬂame oxidation.
54 J. Nikolov et al. / Applied Radiation and Isotopes 71 (2013) 51–56the tritiated water is collected in the tritium counting vial. Anyuncondensed water is collected in the tritium exchange column.In the Model 307 PerkinElmer Sample Oxidizer, the reagentMONOPHASE S is the tritium scintillator. At the end of the cycle,the water in the exchange column is ﬂushed down into thetritium counting vial with the tritium scintilator. After the sampleis completely combusted, steam is injected into combustionchamber as part of the combustion cycle to sweep tritiated waterfrom the combustion ﬂask and condenser into the tritium count-ing vial. The system uses nitrogen or air to dispense the scintilla-tors and absorption chemicals and to ﬂush the combustionproducts from the trapping system into the counting vials(Model 307 PerkinElmer Sample Oxidizer, 2003). Oxidized tritium takes the form of tritiated water and iscollected as water. The water collection is accomplished inmultiple stages. The ﬁrst stage is an air cooled condenser, wherethe steam simply condenses on the walls of the condenser. Afterthe combustion cycle is complete, the water is ﬂushed from thecondenser by a steam and nitrogen/air ﬂush through the ﬂask.Some traces of uncondensed water vapor will be trapped in the Fig. 2. Efﬁciency dependence of SQP(E) parameter.exchange column. The exchange column, containing water,exchanges the water vapor in the long spiral tubing. This water automatically adds the adjusted volume of 13 ml of scintillationvapor is ﬂushed out with a distilled water absorbing chemical into cocktail MONOPHASE S.the collection vial. At the end of the oxidation cycle, the trapped The activity concentrations of 3H were calculated according toand condensed water is ﬂushed from the exchange column with the following formulae (Saxen and Hanste, 2008):the counting solution. This multiple stage tritium collection À1 Ra ÀRbsystem ensures a consistently high percentage of radionuclide A Bq l ¼ ð5Þ c Â V Â e Â eÀltrecovery. In order to optimize sample preparation method, 13 ml on where A—the activity concentration of 3H (Bq l À 1), Ra—the countMONOPHASE S is injected into the 7 tritium collection vials along rate of the sample (cps), Rb—count rate of the background (cps),with the system process water (Model 307 PerkinElmer Sample c—the trapping efﬁciency (97%), V—the volume of the sampleOxidizer, 2003). The following amounts of distilled water added analyzed (l), e—the efﬁciency of the sample measurement. Efﬁ-to the designated vials: in two vials 0 ml (TB and T0); in the other ciencies were determined from the calibration curve for mea-ﬁve vials: 1 ml (T1), 2 ml (T2), 3 ml (T4), 4 ml (T4) and 5 ml (T5). sured SQP(E) parameter for every sample. TB is the tritium background—it is just the MONOPHASE Sscintillation cocktail without added radioactivity. Exactly the 3.4. Minimum detectable activity (MDA)same amount of a known quantity of tritium SPEC-CHECH2833.3 Bq was injected into each of the six vials marked T0–T5. For all three methods mentioned above, minimum detectableAfter recapping, the vials were shaken to ensure a uniform activity (MDA) was evaluated using the Currie formula (Currie,suspension of MONOPHASE S and water. 1968): The vials were places into a liquid scintillation counter in the Ldorder that they were prepared. After dark-adaption, each vial was MDA Bq l À1 ¼ ð6Þcounted for three 20 min measurement cycles in a recommended e Â tb Â Vtritium counting window, recording the quench indicating para- pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃmeter (QIP) and the tritium counts for each vial. The three QIPs and Ld ðcountsÞ ¼ 2:71 þ 4:65 Rb Â t b ð7Þnet tritium counts for each vial (T0–T5) was averaged. The distilled Ld—detection limit, Rb—background count rate (cps), tb—mea-water added into the cocktail vials act like a quenching agent. The suring time of the background sample (s). A factor of 4.65 isobtained results are presented on the calibration curve on Fig. 2. As derived from the statistics and accounts for a 5% probability andone can see in Fig. 2, in the observed region, SQP(E) changes just for radioactivity calculations, 2.71 is often added to the Ld term toa little, but all our samples have the SQP(E) parameter in this account for the zero blank case which corresponds to a 5%region, so this curve is suitable for the efﬁciency determination. probability of a false negative (Cook et al., 2003).SQP(E) parameter is determined by the Quantulus (EasyView soft- The background sample has to be prepared in the same way asware) and it is given without measuring uncertainty. Efﬁciencies the samples of interest, with the same scintillation cocktail and inwere calculated as the ratio of average tritium activity concentration the same proportion. Because of that, there are some speciﬁc(for every prepared sample T0–T5) and injected activity of tritium factors which have to be used in some particular preparationstandard (Model 307 PerkinElmer Sample Oxidizer, 2003). The methods. For electrolytic enrichment the minimum detectableuncertainties for efﬁciencies were very small because of the high activity (MDA) was calculated according to the formula (Villa andactivities of the prepared set of samples, the counting statistic is Manjoli, 2004):better and the uncertainties are not visible on the graph. For everyother sample efﬁciency and appropriate uncertainty could be 2:71 þ4:65pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Rb Â t b 1 À1 MDA Bq l ¼ ÂÀ Áf ð8Þcalculated according to the ﬁtting formula given on Fig. 2. e Â tb Â V V sBE =V sf All the samples were prepared according to operating proce- À Áfdures (Model 307 PerkinElmer Sample Oxidizer, 2003). Volume of where 1= V sBE =V sf introduce enrichment part in the equation0.8 ml of the examined water was injected into combustion cone, and has the same notation as in Eq. (1).after that a 0.4 ml of Combustaid by Perkin Elmer was added to For direct method according to the ASTM Standard Testhelp moderate the burn process. In all samples Sample Oxidizer Method (ASTM D4107-08, 2006), there is a recovery factor F and
J. Nikolov et al. / Applied Radiation and Isotopes 71 (2013) 51–56 55the formula is given by the expression: Table 2 rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Results of measurements of water samples prepared using three different ta methods. Tritium activity concentrations represents the mean value 7 standard 2:71 þ 3:29 Rb Â t a 1 þ tb error for a given measuring conditions for every method of preparation. À1MDA Bq l ¼ ð9Þ e Â ta Â V Â F Â eÀlt 3 Sample H activity concentrations [Bq l À 1] In this case, we measured just background sample, so ta ¼tband e À lt E1, so the previous equation can be expressed as Direct Electrolyticaly Samples measurement, enriched, prepared 2:71 þ 4:65pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Rb Â t b 300 min 90 min by sample À1MDA Bq l ¼ ð10Þ oxidizer, e Â tb Â V Â F 300 min And for the Sample Oxidizer method, Eq. (6) was used, withaddition of the trapping efﬁciency factor c: 1. Zeleno brdo, April 2010 o MDA 1.37 7 0.20 1.11 7 0.15 2. Zeleno brdo, May 2010 2.19 7 0.20 2.53 7 0.13 1.91 7 0.27 2:71 þ 4:65pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Rb Â t b 3. Zeleno brdo, July 2010 2.63 7 0.24 2.59 7 0.20 2.67 0.4 À1MDA Bq l ¼ : ð11Þ 4. Zeleno brdo, Avgust 2010 3.7 7 0.3 3.97 0.4 4.37 0.7 e Â tb Â V Â c 5. Zeleno brdo, October 2010 2.83 7 0.26 2.93 7 0.24 2.87 0.4 6. Zeleno brdo, November 2010 2.82 7 0.26 3.56 7 0.27 3.37 0.5 7. Mlaka, April 2010 26.7 7 2.4 26.5 7 1.1 277 44. Results and discusion 8. Mlaka, July 2010 24.9 7 2.3 26.07 1.1 287 4 9. Mlaka, October 2010 44 74 42.8 7 1.1 447 6 The obtained MDA values for all three methods are presentedin Table 1. For the direct method there our result for MDA is in agreement nuclear reactors (in spite of the cessation of normal reactionwith (Palomo et al., 2007) they got a MDA value 0.6 Bq l À 1 for a ˇ operations at the Vinca Institute of Nuclear Sciences in 1986, thismeasuring time of 500 min, and (Pujol and Sanchez-Cabeza, 2000) is potentially still a source of tritium which will continue togot an MDA value of 2.6 Bq l À 1 for measuring a time of 360 min. disperse into the open environment). According to the last For the electrolytic enrichment, some of the reported MDA published results for this area, reference (Miljevic et al., 2000)values are 0.8 Bq l À 1 for a measuring time of 300 min (Madruga tritium content ranged from 0.9 Bq l À 1 (October 1997) toet al., 2008) and 0.03–0.05 Bq l À 1 for a total counting time of 11.5 Bq l À 1 (July 1993) at Zeleno Brdo following normal seasonal300–500 min (Baresic et al., 2010). variations. In the precipitation from the year 2010, shown in There are no data in the literature for MDA values for Table 2, the maximum value was in August 2010. The samepreparation of water samples by Sample Oxidizer 307, Perkin reference (Miljevic et al., 2000) state that the concentrations inElmer equipment. The similar equipment for preparation of water the Mlaka creek had changed substantially from 38.2 Bq l À 1 insamples which uses combustion as one of the steps in preparation May 1990 to 133.8 Bq l À 1 in January 1993, indicating an erraticand Quantulus for measurement gives the minimum detectable distribution at the sampling point situated near the dischargeactivity of approximately 10 Bq l À 1 for measuring time of from RA. At the same sampling location we got the highest result360 min (Yankovich et al., 2011). in October 2010 of around 44 Bq l À 1. Besides the measuring time, the time needed for sample The research reactor RA in the Vinca Institute of Nuclearpreparation can also be important in choosing the right method. Sciences worked from 1959 to 1986. The RA reactor was placedIn our case, water sample preparation by method of electrolytic in ‘‘long term shutdown’’ condition in April 1986 in order toenrichment lasted for 8 days. The same water sample could be reconstruct and improve practically all the vital reactor systems.prepared by the direct method in a couple of hours (5–6 h) and if Since then, during the regular inspection of the fuel channels, thewe use the automatic Sample Oxidizer, the sample could be release of tritium in HTO form into the environment has beenprepared in 5–10 min (not taking into account the initial and possible, increasing concentration in Mlaka river (Jankovic et al.,ﬁnal cycle which are necessary every time when the Sample 2008). According to the reference (Miljevic et al., 2000), annualOxidizer is turned on or turned off (Model 307 PerkinElmer mean tritium concentrations in precipitation ranged from 2.2 toSample Oxidizer, 2003)). 35.4 Bq l À 1, decreasing with distance from the nuclear facilities of The results of measurements of nine water samples prepared Vinca institute.using three different methods and measured by LSC have beenpresented in Table 2. The obtained results from three different methods showed 5. Conclusionthat all the measured 3H activity is comparable within theuncertainties, even for low active samples. Three different methods of preparing water samples for The obtained values for the activity concentration of tritium in tritium analysis by ultra-low-level background liquid scintillation ˇMlaka collected in the Vinca Institute of Nuclear Sciences indicate counter, Quantulus 1220 have been presented in this paper. Theincreased activity compared to the precipitation collected at obtained results demonstrate that, for environmental monitoring,Zeleno Brdo (reference meteorological station where the concen- direct counting, electrolytic enrichment and Sample Oxidizertration of tritium in precipitation has a low value), which is methods give comparable results and acceptable limits of detec-attributed to the inﬂuence of local contamination by heavy water tion. Electrolytic enrichment followed by liquid scintillation counting will give much lower limits of detection than the otherTable 1 two techniques, and is the preferred method for mineral drinkingMDA values for all three methods of sample preparation. water studies. But if we want a rapid method, Sample Oxidizer Method Time of measurement [min] MDA [Bq l À 1] could be an excellent choice. Otherwise, direct method is also applicable for the activity concentrations of tritium above Electrolytic enrichment 90 0.1 2.1 Bq l À 1, the counting time has to be longer but the time Direct method 300 2.1 Sample oxidizer 300 1.1 needed for sample preparation is negligible comparing to the electrolytic enrichment.
56 J. Nikolov et al. / Applied Radiation and Isotopes 71 (2013) 51–56 Although electrolytic enrichment and direct method for sam- ´ ´ Jankovic, M., Miljevic, N., 2008. Tritium content in precipitation and atmosphericple preparation for tritium determination in water samples are ˇ water vapor of the reactor hall in the Vinca Institute of Nuclear Sciences. In: Proceedings of the 9th International Conference on Fundamental and Appliedwell known technique, Sample Oxidizer is relatively new techni- Aspects of Physical Chemistry, September 24–26, Society of Physical Chemistsque. This are the ﬁrst results of water samples preparation by of Serbia, Belgrade, pp. 653–655.Sample Oxidizer for measuring tritium content by LSC Quantulus Kaihola, L., 1993. Glass vial background reduction in liquid scintillation counting. Sci. Total Environ. 130–131, 297–304.1220. According to the obtained values we can conclude that this Kaufman, S., Libby, W.F., 1954. The natural distribution of tritium. Phys. Rev. 93,method is also suitable for laboratory use. 1337–1344. Until we reach the MDA for the direct method, this method is Lucas, L.L., Unterweger, M.P., 2000. Comprehensive review and critical evaluation of the half-life of tritium. J. Res. Natl. Inst. Stand. Technol. 105 (4), 541–549.always the best method for preparation, but for underground and Madruga, M.J., Sequeria, M.M., Gomes, A.R., 2008. Determination of Tritium inmineral water electrolytic enrichment should be used, and in the Waters by Liquid Scintillation Counting, LSC 2008, Advanes in Liquidcase of emergency if we need a quick response for possible Scintillation Spectrometry. Eikenberg, J., Jaggi, M., Beer, H., Baehrle, H. (Eds.), pp. 353–359.increased tritium content we will use Sample Oxidizer method, ˇ ˇ ´ ´ ˇ Miljevic, N., Sipka, V., Zujic, A., Golobocanin, D., 2000. Tritium around the Vinca ˇbecause it is the fastest way of sample preparation. Institute of Nuclear Sciences. J. Environ. Radioact. 48, 303–315. Model 307 PerkinElmer Sample Oxidizer, 2003. Operation Manual, Publication No.1694057 Rev. L, PerkinElmer Inc.Funding Noakes, J.E., De Filippis, S., 1988. Tritium monitoring of nuclear power plants by liquid scintillation counting. In: Proceedings of the Second International Seminar for Liquid Scintillation Analysis, Tokyo, Japan, pp. 123–138. The authors acknowledge the ﬁnancial support of the Ministry Oxidizer Application Note, 2002. A Comparison of Sample Oxidation and Solubi-of Education, Science and Technological Development of Serbia, lization Techniques, By Jock Thomson, PerkinElmer Life Sciences, Inc. Palomo, M., Penalver, A., Aguilar, C., Borrull, F., 2007. Tritium activity levels inwithin the projects Nuclear Methods Investigations of Rare environmental water samples from different origins. Appl. Radiat. Isot. 65 (9),Processes and Cosmic Rays no. 171002 and Biosensing Technol- 1048–1056.ogies and Global System for Continuous Research and Integrated Pujol, L.J., Sanchez-Cabeza, J.A., 1999. Optimization of liquid scintillation counting conditions for rapid tritium determination in aqueous samples. J. Radioanal.Management no. 43002. Nucl. Chem. 242 (2), 391–398. Pujol, L.L., Sanchez-Cabeza, J.A., 2000. Natural and artiﬁcial radioactivity in surfaceReferences waters of the Ebro river basin (northeast Spain). J. Environ. Radioact. 51, 181–210. Quantulus 1220, 2002. Instrument Manual, Ultra Low Level Liquid Scintillation Spectrometer, PerkinElmer, 1220-931-06.ASTM D4107-08, 2006. Standard Test Method for Tritium in Drinking Water. ASTM Sauzay, G., Schell, R., 1972. Analysis of low level tritium concentrations by International, West Conshohocken, PAhttp://dx.doi.org/10.1520/D4107-08. electrolytic enrichment and liquid scintillation counting. Int. J. Appl. Radiat.Baresic, J., Horvatincic, N., Krajcar-Bronic, I., Obelic, B., 2010. Comparison of two Isot. 23, 25–33. techniques for low-level tritium measurements—gas proportional and liquid Saxen, R., Hanste, U.M., 2008. An Oxidizer/LSC Method for the Determination of 14 scintillation counting. In: Proceedings of the Third European IRPA Congress, C in Foodstuff Samples, LSC 2008, Advanes in Liquid Scintillation Spectro- Full Papers of Poster Presentations, Helsinki, Finland. IRPA 2010, pp. P12-21- metry. Eikenberg, J., Jaggi, M., Beer, H., Baehrle, H. (Eds.), pp. 279–285. 1–P12-21-5. Varlam, C., Stefanescu, I., Faurescu, I., Vagner, I., Faurescu, D., Bogdan, D., 2011.Cameron, J.F., 1967. A survey of Systems for Concentration and Low Background Establishing routine procedure for environmental tritium concentration at Counting of Tritium in Water, Radioactive Dating and Methods of Low-Level ICIT. Rom. J. Phys. 56 (1–2), 233–239. Counting. IAEA, Vienna, pp. 543–573. Villa, M., Manjoli, G., 2004. Low-level measurements of tritium in water. Appl.Cook, G.T., Passo, C.J., Carter, B., 2003. Environmental liquid scintillation analysis. Radiat. Isot. 61, 319–323. In: Michale, F. L’Annunziata (Ed.), Handbook of Radioactivity Analysis. Aca- WHO, 2011. Guidelines for Drinking-water Quality, fourth edition. World Health demic Press, Elsevier, USA, pp. 537–607. Organization, WHO Press, Geneva, Switzerland, ISBN: 978 92 4 154815 1.Currie, L.L.A., 1968. Limits for qualitative detection and quantitative determina- Yankovich, T.L., Kim, S.B., Baumgartner, F., Galeriu, D., Melintescu, A., Miyamoto, tion. Application to radiochemistry. Anal. Chem. 40, 586–593. K., Saito, M., Siclet, F., Davis, P., 2011. Measured and modeled tritiumEuropean Commission, 1998. European Drinking Water Directive 98/83/EC of concentrations in freshwater Barnes mussels (Elliptio complanata) exposed 3 November 1998 on the Quality of Water Intended for Human Consumption. to an abrupt increase in ambient tritium levels. J. Environ. Radioact. 102, Ofﬁcial Journal Legislation 330. 26–34.