Nuclear Instruments and Methods in Physics Research A 722 (2013) 5–10

Contents lists available at SciVerse ScienceDirect

Nuclear Instruments and Methods inPhysics Research A

0168-90http://d

n CorrE-m

parker.a

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

A thalium-doped sodium iodide well counter for radioactive tracerapplications with naturally-abundant 40K

Andrew J. Parker a,n, Colin Boxall a, Malcolm J. Joyce a, Paul Schotanus b

a Department of Engineering, Lancaster University, Lancaster LA1 4YR, United Kingdomb Scionix Holland B.V., Regelierenring 5, 3981 LA Bunnik, The Netherlands

a r t i c l e i n f o

Article history:Received 15 February 2013Received in revised form3 April 2013Accepted 8 April 2013Available online 30 April 2013

Keywords:Sodium iodide well-counterPotassium 40Non-hazardous radioactive tracerLow-activity counting

02/$ - see front matter & 2013 Elsevier B.V. Ax.doi.org/10.1016/j.nima.2013.04.034

esponding author. Tel.: +44 1524 593014; faxail addresses: a.parker4@lancaster.ac.uk,j1@googlemail.com (A.J. Parker).

a b s t r a c t

The use of a thallium-doped sodium-iodide well-type scintillation detector for the assay of the low-activity radioisotope 40K, in open-source potassium chloride aqueous solutions, is described. The hazards,safety concerns and radiowaste generation associated with using open-source radioactive isotopes canpresent significant difficulties, the use of hot cells and escalated costs in radioanalytical laboratoryresearch. A solution to this is the use of low-hazard alternatives that mimic the migration and dispersioncharacteristics of notable fission products (in this case 137Cs). The use of NaI(Tl) as a detection mediumfor naturally-abundant levels of 40K in a range of media is widespread, but the use of 40K as a radioactivetracer has not been reported. The use of such low-activity sources is often complicated by the ability todetect them efficiently. In this paper a scintillator detector designed to detect the naturally-abundant 40Kpresent in potassium chloride in tracer applications is described. Examples of the use of potassiumchloride as a tracer are given in the context of ion exchange and electrochemical migration studies, andcomparisons in performance are drawn from literature with hyper pure germanium semiconductordetectors, which are more commonly utilised detectors in high-resolution counting applications.

& 2013 Elsevier B.V. All rights reserved.

1. Introduction

Anthropogenic radionuclides provide one of the most challenginghazards associated with remediating environments and industrialfacilities in the nuclear industry. Of these radionuclides the fissionproduct 137Cs is a particularly difficult isotope to remove, as itschemical and physical properties make it highly mobile in aqueousenvironments and its radiological properties render it a significantdanger to human health [1].

The nature of the hazards involved with handling 137Cs extends tolaboratory settings where preliminary remediation experiments anddecommissioning techniques are developed. Extensive risk manage-ment and personal protective equipment (PPE) are required toprotect against the hazard of open-source use of the isotope in thelaboratory, due to the high-energy γ-ray emitted at 662 keV. SuchPPE and risk management strategies are often time consuming,cumbersome and expensive to implement. The development ofprotocols requiring significant quantities of material contaminatedwith 137Cs often requires the use of a hot cell and a remote handlingequipment, escalating the cost of these procedures by several ordersof magnitude. As a result suitable analogues that mimic the char-acteristics of 137Cs on chemical, physical and radiological levels

ll rights reserved.

: +44 1524 381707.

without the safety concerns offer significant advantages in terms ofcost and ease of use for the development of radioanalytical methods.By removing 137Cs experimental risk is controlled at the first stage ofthe Hierarchy of Hazard Control, Elimination [2], where the Hierarchyof Hazard Control is a health and safety philosophy used to reduce,manage and control experimental risk.

One such analogue is the naturally-occurring isotope of potassium,40K. Like caesium, potassium is a Group I element that exhibits verysimilar chemical and physical properties. 40K can also be consideredradiologically similar to 137Cs as it boasts a single, prominent γ-emission in the same energy range as that of 137Cs, i.e. 0.5–2MeV,thus with similar, highly-penetrating characteristics albeit at signifi-cantly lower activity. Indeed, unlike 137Cs its specific activity is so lowthat it can be considered non-radioactive, and therefore safe, despitebeing radiometrically detectable [3]. These features render 40K asuitable safe alternative to 137Cs in remediation method development.

A paradox in the use of 40K as an analogue is that its low specificactivity can make it difficult to detect at naturally occurringconcentrations with sufficient efficiency to provide easy measure-ments particularly in high background environments, given that 40Kis a main constituent of natural background radiation. The success-ful use of 40K as a radio-analogue of open-source 137Cs is criticallydependent on the following specific spectrometry considerations:

�

extended counting times, � background attenuation, and � the use of large source samples.

A.J. Parker et al. / Nuclear Instruments and Methods in Physics Research A 722 (2013) 5–106

While background attenuation cannot be considered in thedetector design per se, steps can be taken to optimise it to reducesource sample sizes and counting times.

In this paper, we report on the use of a thallium-doped sodium-iodide, NaI(Tl,) well-type scintillation detector, based on optimaldesign choices to improve the detector efficiency to count low-activity sources. The primary purpose of the detector is the radio-metric study of 40K in aqueous potassium chloride solutions as ananalogue for 137Cs in open-source developmental experiments. Sucha device may be used in the future for in situ radiometric measure-ments of radionuclides of concern in the cleanup of nuclear sites.

2. Low activity counting and NaI(Tl) scintillators

Various detection technologies have been developed for low-activity counting applications, most often doped metal halide scin-tillators such as sodium iodide and caesium iodide, whilst theefficiency of these is amongst the best for γ-ray counting, the energyresolution can limit the use of these compounds where a fine degreeof discrimination is needed between lines with very similar energies.Where high-resolution γ-ray spectrometry is desired such as for thecase of 40K where contamination can obscure the 1460.8 keV line, thede facto detector of choice for low-activity counting is typically thehyper pure germanium (HPGe) semiconductor. Numerous examplesexist in the relevant literature where both standard geometry andwell-type HPGe detectors have been utilised to measure low-activitysamples containing 40K, with minimal detectable activities (MDA) inthe range of Bq/m3. For the sake of brevity only a small number ofinstances will be given here; the instances include determiningpotassium content in food produce [3,4], calculating the activity ofnaturally occurring radioactive material (NORM) in building material,of which 40K is a constituent [5,6], and studying the levels of 137Cs insoil and grain samples [7]. Significant drawbacks exist to the use ofHPGe semiconductors, they are expensive to purchase and maintaingiven that they require cooling to 77 K, usually requiring the use ofliquid nitrogen (LN2) dewars [8]. The use of LN2 is often restricted inmany industrial environments due to concerns over potential spillagewhich can be hazardous. Mechanically-cooled HPGe detectors areavailable which provide a means around LN2 restrictions, but againthey can be considerably more expensive [9].

Given the issues surrounding the use of HPGe detectors, well-type NaI(Tl) detectors have found use as a means of counting low-activity samples where the use of HPGe is prohibited and in someinstances demonstrating a lower MDA [10,11], thereby competingwith HPGe detectors on performance as envisaged by Theodorsson[12]. Other scintillation materials exist that, perhaps, offer betterphysical characteristics in the detection of low-activity, theseinclude caesium iodide (CsI), bismuth germinate (BGO) andlutetium oxyorthosilicate (Lu2SiO5) [8]. The higher atomic densityof thallium-doped caesium iodide CsI(Tl), for example, provides ahigher degree of stopping power for energetic γ-rays than NaI(Tl),like the 1460.8 keV emission from 40K [13], and is a competingchoice of material. However, the presence of trace amounts(upwards of 60 mBq/kg) of the fission product 137Cs within theCsI(Tl) crystals provides a significant hindrance to its use as ameans of low-activity counting [14,15]. A variety of exotic, butincreasingly common, scintillator materials are becoming afford-able; however in terms of cost, performance, and ease of manu-facture NaI(Tl) constitutes perhaps the best compromise.

Given the poor energy resolution of NaI(Tl) relative to alter-natives such as HPGe, well detectors based on this are perhapsbest used for single-isotope detection where the isotope is usuallyknown beforehand. The detection of 225Ac using a 20�20 cm2 NaI(Tl) crystal with a 13 cm deep 2.5 cm diameter well by Pomméet al. is one such example of NaI(Tl) low counting [16]. In this work

225Ac samples were counted to measure the half-life of the isotopeusing six different counting methods. The results obtained fromthe NaI(Tl) well detector were consistent with those from othermethods, namely a HPGe detector, a CsI sandwich spectrometeramongst others, thereby validating NaI(Tl) well counters as a low-activity counting technique against higher-resolution alternatives.

The efficiency of the NaI(Tl) well detectors can be furtherimproved with the addition of a counting ‘lid’ to the detector,commonly found in Positron Emission Tomography (PET). Utilisinga second detector inserted into or over the detector well providesalmost 4π counting geometries. The use of two sodium iodidescintillators, as described by Rieppo, provides for a 4π geometry,though no substantial increase in efficiency is noted over a well-type crystal [17]. The major complicating factor with thesedetector arrangements is often the extra electronics and additionalanalysis of the two spectra that is required [8,17].

The relative ease with which large NaI crystals are manufacturedhas also led to arrays of large NaI(Tl) detectors being used in whole-body counters [18]. Whole-body counters are used throughout thenuclear industry as a means of checking personal exposure toradiation; they have also been used as a means of determining thebodily content of 40K in humans and other animals in nuclearmedicine studies [18–20]. In these studies, 40K is naturally occurringwithin the tissues of the examples being counted. This is in contrastto the intended use of potassium in this work where 40K isspecifically introduced in the form of potassium chloride (KCl).

NaI(Tl) well detectors are perhaps most commonly utilised in thefield of nuclear medicine, where they are utilised as a means ofvalidating radiation levels in PET and Single-Photon Emission Com-puted Tomography (SPECT) analysis [21]. However, the isotopes usedin this analysis are typically low volumes of short half-life isotopes,99mTc being the most common example, with comparatively highactivities upward of 30 MBq [21].The NaI(Tl) well detectors used innuclear medicine are often built into automated multi-sampledetection rigs; examples include the Wallac Wizard and PackardCobra II systems which are unsuitable for our purposes [22].

Whilst the use of NaI(Tl) detectors in well counter systems iswidespread for the assay of contamination in a variety of environ-ments and applications, there have been few if any reports of theiruse specifically for the detection of 40K, specifically where potas-sium has been introduced as a radioactive tracer. In addition to itsuse in whole-body counters, there are numerous instances wheresodium iodide has been used to detect inherent levels of 40K[23,24], but its use for the assay of quantities of introduced 40Ktracer has not been described, often it appears because of the verylow activity of 40K coupled with their often being no restriction onthe use of potassium i.e. due to its chemical significance in a giventracer application per se: thus there are usually higher activityalternatives that need less counting time. The only evidence in theliterature of a potassium isotope being utilised specifically as aradioactive tracer is the use of 38K for medical imaging by Takamiet al. Necessarily, 38K is radiologically very different from its 40Kcousin in that its half-life is only 7.5 min and has a far higherspecific activity [25]; it is these radiological properties that cause itto be selected for this application as opposed to the choice ofelement. In this paper a alternative scenario is reported where thechoice of element is crucial in replicating the chemical propertiesof a given adsorbent, whilst the natural radioactive propertiespresent in the natural isotopic abundance of that element rendersit traceable on a radiological basis.

3. Detector description and experimental setup

The NaI(Tl) well detector, manufactured by Scionix, Bunnik, theNetherlands, used for the measurements performed in this work

A.J. Parker et al. / Nuclear Instruments and Methods in Physics Research A 722 (2013) 5–10 7

was designed and built to be optimised in terms of high detectionefficiency, low intrinsic background, and that it remains portable.The side elevation of the design and the dimensions of thedetector and the machined well are given in Fig. 1.

The size of the crystal and dimensions of the well machinedinto it were chosen to reflect the size of the samples to bedetected, (up to 50 ml beakers, therefore a well of 50 mm diameterand 50 mm depth, giving a total depth of 4100 cm−3) and tomaximise detector efficiency. Given these dimensions, the wellallows for 93.8% of the isotropically-emitted γ-rays to be detectedfrom a point source placed at the bottom of the well [8]. Thedetector was built of intrinsically-low background materials,namely a body of aluminium [26], and low-background glass usedto construct the photomultiplier tube (PMT). The PMT size pre-sents one of the more difficult design choices, as its size affectsboth the efficiency of the detector and the level of backgrounddetected, almost in inverse proportions. As detailed in Fig. 1, thedetector was built with a 130 mm PMT and photocathode. Thehigher levels of efficiency are potentially offset by the increase inspurious detection events at the optical interface which maycontribute to an increased background count [8].

The detector utilises a 14-pin PMT connexion to connect to anORTEC DigiBase PMT base, with an integrated bias supply, preamplifier

Fig. 1. Side elevation of the NaI(Tl) well counter used in this research, withdimensions and corresponding materials choice of each part of the instrument,courtesy of Scionix (the Netherlands).

Figbloof

and multichannel analyser (MCA). The use of this method of connex-ion increases the simplicity of use and the portability of the detectionsetup, with the detector and PMT base weight o7.7 kg. The detectorspecification is described in detail in [27]. The detector is placed insidea lead collimation setup of 10 cm thickness, a photograph of the fullspectrometer is given in Fig. 2.

The utilisation of the bespoke lead collimation setup aroundthe detector, as seen in Fig. 2, significantly reduced the level ofbackground detected. The 10 cm thick collimator caused thedetected background count to fall to o93% of an un-collimatedcount over a 2 h counting duration, resulting in a count-per-second of 1.28 in the energy range of interest (1461 keV peak).

4. Evaluation and practical application

4.1. Isotopic characterisation

Prior to subsequent low-activity measurements, the energydependence of the detection efficiency and a variety of otherparameters of the NaI(Tl) well detector were characterised using aselection of different γ-ray sources that offer a spread of energies.The results of this experiment are shown in Table 1 in terms ofcounts for each isotope.

The results presented in Table 1 describe the performance ofthe detector with respect to the activity and peak energy from fivedifferent sources. All the sources used were point sources, placedat the bottom of the detector well, with the exception of the 40Ksource. The 40K source was in the form of a 20 g sample of (KCl) ina sealed 40 ml tube placed at the bottom of the well.

. 2. A photograph of the spectrometer used in this research in which the leadcks have been removed from the top of the instrument to provide a better viewthe detector.

Table 1Detector characterisation using different γ-ray sources in terms of absoluteefficiency, and FWHM resolution.

Source 137Cs 137Cs 137Cs 60Co 60Co 241Am 40K

Peak energy (keV) 662 662 662 1173 1332 59 1461Activity (kBq) 34.2 369 369 16.2 16.2 34.1 0.32Source location Well Well Face Well Well Well WellAbsolute efficiency (%) 34.3 27.9 27.8 27.9 27.9 82.3 2.8Resolution (keV) 47 42 49 65 73 11 81Resolution (%) 7.2 6.4 7.4 5.7 5.6 18 5.6

A.J. Parker et al. / Nuclear Instruments and Methods in Physics Research A 722 (2013) 5–108

The activity of the evaluation isotopes is at least 4–5 orders ofmagnitude higher than the activity of samples in the Principal UseStudy (Section 4.3). As such factors such as dead time are likely toaffect the absolute efficiency calculations, giving a lower efficiencythan could be attained for a lower activity source [8]. Theefficiency measurements in Table 1 are further affected by theformation of summed peaks where in a high-rate situation the twodiscrete emissions are erroneously summed to constitute a singleevent; this is especially noticeable for the 60Co spectrum, shown inFig. 3. Again, for the case of the low activity levels the detector hasbeen designed for, this issue can be ignored.

Although the energy resolution is not of primary importancewith regards to detector specification as the isotopes of interestare usually known, the data in Table 1 shows that a satisfactorylevel of resolution is achieved with NaI(Tl). At 1461 keV a resolu-tion of 5.6% (�81 keV) is recorded, which is reasonable consider-ing that 10% loss of energy resolution is expected when a well ismachined into the scintillator crystal [8].

4.2. Low-activity counting

Plotted in Fig. 4 is the net count from a 40 ml KCl solution ofconcentration 0.1 mg/ml against the time taken to achieve that netcount, together with the uncertainty in the count of three standarddeviations (3s); where, using the Poisson statistics and errorpropagation, the net count is calculated from a mean backgroundcount subtracted from a gross count and the 3s derived from the3√(gross count+mean background) [8,28]. It thus demonstratesthe sub-linear trend of the square-root dependence of the uncer-tainty in large count experiments, and the approximate region atwhich the uncertainty becomes acceptable.

In Fig. 3 it can be see that it takes over 20 h to satisfy the 3sstatistical requirement to determine the presence of a low-activitysample. Prior to this point, there might be enough statistical

Fig. 3. The energy spectrum of 60Co, showing the 1171 and 1332 keV photopeaks aswell as the summed peak of the two energy releases at 2503 keV.

Fig. 4. Count vs. time taken to collect the count level, for the net count (black line)and 3s (grey line) where s is equal to the square root of the net count for a 40 mlKCl solution of concentration 0.1 mg/ml, equivalent to 1.6 mBq ml, using theoptimised NaI(Tl) detector.

uncertainty to reason that the count is just random backgroundfluctuation as opposed to coming from the sample.

Fig. 4, shown below, is a comparison between a standardgeometry, 50 mm�50 mm thallium-doped caesium iodide scin-tillator (Scionix, Bunnik, the Netherlands) and the NaI(Tl) welldetector of the time required to reach a count greater than itscorresponding 3s error value over a range of potassium chloride(KCl) concentrations. It can be seen from Fig. 5 that the NaI(Tl) welldetector produces the required count greater than 3s in a farshorter time than CsI(Tl); given the larger crystal size and bettergeometric efficiency this is what would be expected.

The comparison shown in Fig. 4 unfortunately does not showequivalent data from a HPGe based detector. Such data are notavailable in the scientific literature specifically for KCl. As suchcomparisons have to be taken from low-activity 40K containingstudies, such evaluation has to be treated cautiously as theexperimental factors can have a significant influence on thedetector performance, e.g. the geometry. Maidana et al. wasconcerned with detecting the levels of potassium in tea infusion[3]. The potassium concentration of the tea infusion in the studywas found to be �0.5 mg ml−1 in tea brews and to achievedetection a counting time of 20 h was employed using a consistentdepth of background shielding (10 cm of lead). In Fig. 5 it can beseen that a similar concentration of potassium chloride can bedetected to a 3s degree of uncertainty in a similar length of time.Other analogous low-activity counting studies using HPGe detec-tors have also employed extended counting times upwards of 20 h,and in some cases as long as 6 days to count [3,29]. From apragmatic experimental view this prolonged duration puts strainon detector cooling and maintenance, e.g. the need for anadequate supply of LN2.

4.3. Principle use studies

The proposed use of NaI(Tl) is to detect the active isotopewithin potassium chloride solutions, where KCl has been used inexperiments to develop a radioanalytical technique as a radio-active tracer. The use of KCl is as a replacement for the hazardous137Cs isotope, which is present in a vast range of nuclear decom-missioning environments. Given below are two examples of radio-analytical experiments used to develop decommissioningtechniques where the physical and chemical similarities of potas-sium to caesium means that KCl can be used as a radiologicalanalogue of 137Cs thus largely removing any radiological hazardsassociated with the experiments.

4.3.1. Ion exchangeThe first example of the use of the NaI(Tl) well detector is the

replication of a decommissioning scenario involving the study ofradioactive caesium adsorption onto a cation exchange resinwithout using a radioactive isotope of caesium. The fundamental

Fig. 5. Time taken to achieve a net count value equal to 3s for a variety of lowconcentrations of potassium chloride with the optimised NaI(Tl) and CsI(Tl)detectors.

Figinc

Fig. 7. An exploded schematic of the electrokinetic apparatus, (A) 60 mm, (B)150 mm, (C) 140 mm, and (D) 110 mm, with dashed arrows indicating the directionof K+ transport toward the cathode compartment.

A.J. Parker et al. / Nuclear Instruments and Methods in Physics Research A 722 (2013) 5–10 9

design principle of this experiment is that the cation exchangeresins were prewashed with a concentrated potassium chloridesolution, changing the cation on the resin from hydrogen (H+) topotassium (K+). As such the resin could be considered as a K+ formresin; where 0.017% of potassium is the radioactive 40K isotope andtherefore radiometrically detectable [13]. The K+ form resin wasthen mixed with various non-active caesium chloride solutions.

Fig. 6, taken from Parker et al. [30], shows the count detectedfrom 6 g samples of pre- and post-wash resin and 7 ml aliquots ofthe subsequent post-mix solutions. The relationship between theincreased potassium displacement when the caesium chloridesolution is higher is evident.

The 6 g samples of pre- and post-mix resins and 7 ml aliquotsof the subsequent post-mix solutions were subjected to 2-hdetection times, using the NaI(Tl) well counter, to monitor thechange in count as the potassium ions were exchanged off theresin by the caesium ions. Employing a 2 h count for all samplesprovided comprehensive counting statistics such that a detaileddescription of the caesium adsorption could be produced; despitethe low activity of the natural levels of 40K in potassium chlorideused to dose the ion exchange resin. Further method informationand results detailing the adsorption reactions are available in [30].

Fig. 8. The change in K+ concentration of the anodic and cathodic solutions overtime under the application of a 1600 V m−1 potential.

4.3.2. Electrokinetic decontaminationThe second example given is the monitoring of an electro-

chemical migration experiment designed to mimic legacy nuclearfuel storage ponds and the in situ remediation of their contami-nated concrete walls, where the application of an electric field,between two electrodes is used to accelerate the ionic diffusion ofa contaminant through a 10.5 cm diameter, 3.5 cm thick concreteslab, comparable to caesium electrokinetic studies conducted byFrizon et al. and Popov et al. [31,32]. Fig. 6 shows a schematic of anexample experimental rig used for electrokinetic decontamination,where the concrete sample is placed in the central cylindrical tube.

Again, a potassium chloride solution, of initial concentration18 mgml−1, was used as the radiological tracer in the anodiccompartment. Upon application of the electric field, the flux of K+

ions was measured radiometrically by taking 40 ml samples periodi-cally from both the cathode and anode compartments seen in Fig. 7.Fig. 8 shows the change in concentration of the compartments overtime, inferred from the change in detected activity. It is evident fromthe graph the ionic transport through the electrolyte and subse-quently the concrete slab the potassium ions.

. 6. The effect of mixing 6 g samples of K+ form resin with 20 ml solutions ofreasing CsCl concentration.

As a measure of precision, the experiment was repeated using twomeasurement techniques to verify the potassium transport throughthe concrete sample and into the catholyte, (A) radiometric count and(B) conductivity metre (Jenway 4200 Conductivity Metre) to deter-mine the Total Dissolved Solid (TDS). Fig. 9 shows the change inconcentration of the cathode with respect to radiometric count, from40K present, and the TDS. The consistent relationship betweenmeasurement techniques for determining change in potassium con-centration over time is clear. The radiometric method of determinationhas the distinct advantage over the measure of TDS as the onlyradiological material in the system, and therefore the only radio-metrically determinable material in the system, is potassium. TDS, as acomparison, measures all cations in solution.Both the ion exchangeand electrokinetic examples described above, concentrations of potas-sium as low as 1mgml−1 corresponding to an activity of approxi-mately 0.5 Bq were detected with a high degree of accuracy over 2-hcounting times. The MDA over an even longer time will be consider-ably lower, as seen in Fig. 5. The use of potassium as the radio-contaminant in these instances provides a useful analogue to that of137Cs found in the actual nuclear decommissioning scenarios, as theelements share ionic charge and similar diffusion coefficients [33].

The quality of the results, and precision of the measurementtechnique, from these studies show that it is possible to replicateradioactive environments and study them through the use ofthe specialised detector in what could be considered a completelynon-active environment, where there are no radiowaste generationor personnel safety issues associated with using a open-sourceradioactive material.

Fig. 9. The change in potassium concentration in a catholyte solution using two methods, (A) radiometric determination, and (B) conductivity metre.

A.J. Parker et al. / Nuclear Instruments and Methods in Physics Research A 722 (2013) 5–1010

5. Conclusion

We have described the use of a thallium-doped sodium iodidewell counter that has been designed for the detection of low-activity samples and specifically the detection of 40K radiologicaltracers. The use of NaI(Tl) detectors is common in the detection ofnatural levels of potassium in organic matter, however the detec-tion of an intentionally-introduced potassium chloride radiologicaltracer has not been previously reported. The detector has beenshown to compete with literature examples of HPGe semiconduc-tor detectors in terms of low-activity performance, and itssuccessful use in laboratory-based nuclear decommissioningremediation development experiments has been demonstratedwith the use of potassium chloride as the radioanalytical tracer.Thus, the combined use of the detector and 40K tracer provides anelegant solution to the control of the radiological experimentalrisk on the Heirarchy of Hazard Control by eliminating most of thehazards associated with the use of anthropogenic contamination.

Acknowledgements

The authors would like to thank John Caunt Scientific Ltd. andScionix Holland B.V. for their help in manufacturing the detector;along with Sellafield Ltd. for the support and direction. We wouldlike to thank the Nuclear Decommissioning Authority (NDA) forfunding and the National Nuclear Laboratory (NNL) for advice.Colin Boxall is supported by the Lloyd's Register EducationalFoundation (LREF), an independent charity working to achieveadvances in transportation, science, engineering and technologyeducation, training and research worldwide for the benefit of all.

References

[1] Argonne National Laboratory, Cesium, Human Health Fact Sheet, August, 2005.[2] Health and Safety Executive, Leadership and Worker Involvement Toolkit

Management of Risk when Planning Work: The Right Priorities, Leadershipand Worker Engagement Forum, 2011.

[3] N.L. Maidana, V.R. Vanin, C.L. Horii, F.A. Ferreira, M.U. Rajput, Food Chemistry116 (2) (2009) 555.

[4] G. Espinosa, I. Hernández-Ibinarriaga, J.-I. Golzarri, Journal of Radioanalyticaland Nuclear Chemistry 282 (2) (2009) 401.

[5] M.B. Cooper, Naturally Occurring Radioactive Materials (NORM) in AustralianIndustries—Review of Current Inventories and Future Generation, September,2005.

[6] N. Lavi, F. Groppi, Z.B. Alfassi, Radiation Measurements 38 (2) (2004) 139.[7] N.H.M. Kamel, Journal of Environmental Radioactivity 101 (4) (2010) 297.[8] G. Knoll, Radiation Detection and Measurement, fourth ed., John Wiley & Sons

Ltd., 2010, p. 860.

[9] K.-P. Ziock, C. Boehnen, H. Jason, A.C. Raffo-Caiado, A mechanically cooled,highly portable, HPGe-based, coded-aperture gamma-ray imager, In: Proceed-ings of the Institute of Nuclear Materials Management, 2010.

[10] A. Perez-Andujar, L. Pibida, Applied Radiation and Isotopes 60 (1) (2004) 41.[11] C. Tsabaris, C. Bagatelas, T. Dakladas, C.T. Papadopoulos, R. Vlastou, G.

T. Chronis, Applied Radiation and Isotopes 66 (10) (2008) 1419.[12] P. Theodorsson, Nuclear Instruments and Methods in Physics Research Section

A: Accelerators, Spectrometers, Detectors and Associated Equipment 506(2003) 143.

[13] Argonne National Laboratory, Potassium-40, Human Health Fact Sheet,August, 2005.

[14] Y.D. Kim, T.Y. Kim, D.H. Choi, J.M. Choi, I.S. Hahn, M.J. Hwang, R.K. Jain, W.G. Kang, H.J. Kim, S.C. Kim, S.K. Kim, Y.J. Kwon, E.K. Lee, H.S. Lee, J.H. Lee, J.I. Lee, M.H. Lee, S.E. Lee, D.S. Lim, S.H. Noh, H. Park, I.H. Park, E.S. Seo, E. Won,H.Y. Yang, M.S. Yang, I. Yu, J. Zhu, Journal of the Korean Physical Society 43(2003) 40.

[15] Y.D. Kim, U.K. Kang, S.E. Lee, H. Park, I.H. Park, E. Won, H.S. Won, H.Y. Yang, M.S. Yang, Journal of the Korean Physical Society 40 (3) (2002) 520.

[16] S. Pommé, M. Marouli, G. Suliman, H. Dikmen, R. Van Ammel, V. Jobbágy,A. Dirican, H. Stroh, J. Paepen, F. Bruchertseifer, C. Apostolidis, A. Morgenstern,Applied Radiation and Isotopes: Including Data, Instrumentation and Methodsfor Use in Agriculture, Industry and Medicine 70 (11) (2012) 2608.

[17] R. Rieppo, International Journal of Applied Radiation and Isotopes 27 (1976)457.

[18] M.D. Jenson, J.S. Braun, R.J. Vetter, M.H. Marsh, Mayo Clinic Proceedings 63 (9)(1988) 864.

[19] O. Selberg, J. Bo, M. Pirlich, E. Henkel, M. Peter, M.J. Mu, Hepatology Research16, (1999) 36.

[20] J.T. McVey, J. Hunziker, J.F. Holson, J.F. Young, Journal of PharmacologicalMethods 5 (1) (1981) 36.

[21] P.J. Markiewicz, G.I. Angelis, F. Kotasidis, M. Green, W.R. Lionheart, A.J. Reader,J.C. Matthews, Physics in Medicine and Biology 56 (21) (2011) N247.

[22] J. Hattink, J.J.M. De Goeij, H.T. Wolterbeek, Environmental and ExperimentalBotany 44 (2000) 9.

[23] N. Kumar, S.S. Chaturvedi, S.K. Jha, Journal of Radioanalytical and NuclearChemistry 290 (1) (2011) 95.

[24] T. Martinez, M. Navarrete, L. Cabrera, F. Juárez, A. Ramos, K. Vazquez, Journal ofRadioanalytical and Nuclear Chemistry 273 (3) (2007) 569.

[25] R.S. Tilbury, W.G. Myers, R. Chandra, J.R. Dahi, R. Lee, Journal of NuclearMedicine 21 (9) (1980) 867.

[26] P. Collon, M. Bowers, F. Calaprice, C. Galbiati, D. Henderson, T. Hohman, C.L. Jiang, W. Kutschera, H.Y. Lee, B. Loer, R.C. Pardo, M. Paul, E. Rehm,D. Robertson, C. Schmitt, R. Scott, R. Vondrasek, Nuclear Instruments andMethods in Physics Research Section B: Beam Interactions with Materials andAtoms 283 (2012) 77.

[27] A.J. Parker, C. Boxall, M.J. Joyce, P. Schotanus, A NaI(Tl) scintillator for in situenvironmental studies and laboratory detection measurements of aqueouspotassium chloride, In: Proceedings for the NSS/MIC Fall Meeting 2012,Anaheim, CA, 2012.

[28] P. Bevington, D.K. Robinson, Data Reduction and Error Analysis for the PhysicalSciences, 3rd ed., McGraw-Hill, 2003.

[29] M. Inoue, K. Komura, Journal of Radioanalytical and Nuclear Chemistry 273 (1)(2006) 177.

[30] A.J. Parker, C. Boxall, M.J. Joyce, Journal of Radioanalytical and NuclearChemistry (2013).

[31] F. Frizon, S. Lorente, C. Auzuech, Cement and Concrete Research 35 (10) (2005)2018.

[32] K. Popov, I. Glazkova, V. Yachmenev, A. Nikolayev, Environmental Pollution(Barking, Essex: 1987) 153 (1) (2008) 22.

[33] A.M. Friedman, W. Kennedy, Journal of American Chemical Society 446 (1952)(1955) 1.