Nuclear Instruments and Methods in Physics Research A 622 (2010) 460–463

Contents lists available at ScienceDirect

Nuclear Instruments and Methods inPhysics Research A

0168-90

doi:10.1

� Corr

E-m

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

Large sample NAA work at BARC: Methodology and applications

R. Acharya a,�, K.K. Swain b, K. Sudarshan a, R. Tripathi a, P.K. Pujari a, A.V.R. Reddy b

a Radiochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, Indiab Analytical Chemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, India

a r t i c l e i n f o

Available online 17 February 2010

Keywords:

Large sample NAA

Internal monostandard method

In situ detection efficiency

Non-standard geometry

Pottery

Uranium ore

Stainless steel

02/$ - see front matter & 2010 Elsevier B.V. A

016/j.nima.2010.02.056

esponding author. Tel.: +91 22 2559 4089; fa

ail address: racharya@barc.gov.in (R. Acharya

a b s t r a c t

Large sample neutron activation analysis (LSNAA) was carried out using thermal column facility of

Apsara reactor at Bhabha Atomic Research Centre, Mumbai, India. The k0-based internal monostandard

NAA (IM-NAA) using in situ detection efficiency was used to analyze large and non-standard geometry

samples of clay pottery, uranium ore and stainless steel. Elemental concentration ratios with respect to

Na as a monostandard were used in the study of pottery and ore samples. For stainless steel sample of

SS 304L, the absolute concentrations were calculated from concentration ratios by mass balance

approach since all the major elements (Fe, Cr, Ni and Mn) were amenable to NAA. Applications of LSNAA

in the above-mentioned three different areas are described in this paper.

& 2010 Elsevier B.V. All rights reserved.

1. Introduction

Large sample analysis is often advantageous for obtainingbetter analytical representativeness instead of replicate sub-sample analysis. It is more so for the samples that are not sohomogeneous at lower level [1,2]. Analysis of large sample isfeasible by neutron activation analysis (NAA). However, it isnecessary to take care of the neutron flux perturbation, if anyduring neutron irradiation, and g-ray attenuation in the largesamples during measurements. The k0 method of NAA has beenused to analyze large size samples, where elaborate proceduresfor accounting the neutron self-absorption/flux perturbationduring sample irradiation in a reactor and g-ray self-attenuationfor determining the efficiency of the detector have been addressed[3,4]. Detection efficiency calibration is carried out using effectivesolid angle concept and g-ray transmission methods usingstandard multi g-ray sources.

Simpler methods for the analysis of large samples undervarying geometrical conditions would greatly enhance theapplicability of NAA for variety of samples like high-purity alloysand metals, archeological objects and samples of biological,geological and environmental origin. A k0-based internal mono-standard method for determination of elements in samples ofnon-standard geometry by prompt gamma ray neutron activationanalysis (PGNAA) was proposed by Sueki et al. [5]. They haveutilized g-rays from the monostandard for the calibration ofrelative detection efficiency. A k0-based internal monostandardNAA (IM-NAA) using in situ relative detection efficiency [2,6],

ll rights reserved.

x: +91 22 2550 5151.

).

developed in our lab, is useful to analyze large and non-standardgeometry samples. In this method, an element present in thesample was used as the monostandard, which takes care ofneutron flux perturbation, if any, inside the sample. Usually, amajor or a minor element present in the sample is chosen as amonostandard, as a better homogeneous distribution of thiselement in the sample is expected compared to the trace elementdistribution. However, a trace element present in the sample canalso be chosen as a monostandard if any of its isotopes hassuitable nuclear properties like high isotopic abundance and (n,g)cross-section, with favorable half-life and high g-ray abundancefor its activation product.

The in situ relative detection efficiency was obtained usingg-rays of two or more radionuclides produced in neutronactivation that have g-rays covering the energy range of interest.The method was validated using IAEA reference materials in themass range from 50 mg to 5 g [6] as well as by determining addedimpurities in 0.5 kg silica and 0.5 L water [2]. The method wasapplied for the compositional analysis of irregular shape and sizesamples of nuclear fuel cladding materials [2,7–9] namelyZircaloys 2 and 4, stainless steels (SS 316M and D9 alloy) and1S aluminium. In the cases of zircaloy and stainless steel,composition analysis was carried out by standard-less approachsince all major and/or minor elements were amenable to NAA. TheIM-NAA method was applied to wheat grains in the mass rangefrom 50 mg to 1 kg to arrive at the representative sample size. Itwas found from elemental concentration ratios with respect to Nathat the mass ratios are constant for samples having mass 1 g andabove indicating that 1 g is the minimum representative samplesize [9]. This method was also applied to small and large sizepowdered coal samples, where the representative sample sizewas determined as 1 g [10].

R. Acharya et al. / Nuclear Instruments and Methods in Physics Research A 622 (2010) 460–463 461

In the present work, applicability of IM-NAA to large and non-standard geometry samples of ancient and new clay pottery,uranium ore and stainless steel (SS 304L) has been explored.Elemental concentration ratios with respect to Na as a mono-standard were used for the pottery and uranium ore samples. Inthe case of SS 304L, Fe was used as the internal monostandard andits analysis was carried out by a mass balance approach.

5.0x10-4

1.0x10-3

1.5x10-3

2.0x10-3

2.5x10-3

3.0x10-3

3.5x10-3

4.0x10-3

4.5x10-3

Nuclides used: 152mEu, 140La, 56Mn,46Sc, 42K, 72Ga

Insi

tu re

lativ

e de

tect

ion

effic

ienc

y

Energy (keV)0 200 400 600 800 1000 1200 1400 16001800 2000

Fig. 1. In situ relative detection efficiency for large size pottery sample irradiated

using thermal column of Apsara reactor.

2. Experimental

Samples of three ancient potteries and one new pottery (massrange 15–25 g), two different uranium ores (about 55 g) and oneSS 304L (21 g) were irradiated in the thermal column of Apsarareactor for 7 h. For comparison of LSNAA results of one of theuranium ores, four small size homogenized samples (100 mg)were also irradiated in E8 position of Apsara reactor. Thecorresponding thermal neutron fluxes are about 108 and1012 cm�2 s�1. For neutron flux characterization in thermalcolumn (TC), indium (10 mg) was irradiated with and withoutcadmium cover (0.8 mm thick) for 6 h at the nearest position tothe core. Samples were counted using a 40% relative efficiencyHPGe detector coupled to an 8k channel analyzer that hasspectrum analysis software PHAST [11].

The ratio of mass (m) of an element (x) to mass of the internalmonostandard element (y) in the sample by the k0-based IM-INAAmethod is given by the following expression [2]:

mx

my¼ððSDCÞðf þQ0ðaÞÞÞyððSDCÞðf þQ0ðaÞÞÞx

PAx

PAy

ðegÞyðegÞx

k0;AuðyÞ

k0;AuðxÞð1Þ

where PA is the net peak area under the gamma peak of interest, S

the saturation factor, D the decay factor, C the counting factorused for correcting the decay during counting period, f thesubcadmium to epithermal neutron flux ratio, a the epithermalneutron flux shape factor, Q0(a) the ratio of the resonance integral(I0)-to-thermal neutron cross-section (s0) corrected for a, the k0,Au

the literature k0,Au-factors [12] and e the in situ relative detectionefficiency in the present case [2,6]. The in situ relative efficiency isobtained by the following expression:

ln eg ¼ kjþXm

i ¼ 0

aiðln EgÞi

ð2Þ

where ais are the coefficients of the polynomial of order m and kj

is a constant characteristic of the jth nuclide. In the calculations asecond-order polynomial (m=2) was used. As f-value is of theorder of 103 in a highly thermalized irradiation position like inthermal column, Eq. (1) is simplified to Eq. (3) as given below.

mx

my¼ðSDCÞyðSDCÞx

PAx

PAy

ey

ex

k0;AuðyÞ

k0;AuðxÞð3Þ

For the samples in which all the major and the minor elementsare amenable to NAA, the absolute concentrations can bedetermined using a mass balance procedure. This procedure isused in those cases where the major and minor elements accountfor nearly 100% of the sample mass. The sum of relative elementalconcentration ratios with respect to an internal monostandard (y)can be written as

Xn

i ¼ 1

mi=my ¼W=my ð4Þ

wherePn

i ¼ 1 mi ¼W , where W is the mass of the sample. Finally,the wt.% of the element is obtained by

mið%Þ ¼mi=my

W=my� 100 ð5Þ

Details of calculations by IM-NAA method using mass balanceprocedure can be found in our earlier publications [8,9].

3. Results and discussion

3.1. Characterization of irradiation sites

The f-value of thermal column was found to be(6.070.4)�103 (499.9% thermal neutron component) and thecorresponding thermal equivalent neutron flux is 1.2�108 cm�2

s�1 [10]. The cadmium ratio method using multi-monitors (Au, Zr,Zn, Mo and In) and Au was used for determination of a and f in E8position of Apsara reactor and the corresponding values are0.03570.003 and 50.071.5 [13], respectively.

3.2. Application of IM-NAA to potteries

Chemical analysis is an important tool for provenance study ofarchaeological artifacts like pottery, brick and stone. In thepresent work, samples of three ancient potteries and a newpottery were analyzed for provenance study using elementalconcentration ratios. Concentrations of 16 elements namely Na, K,Sc, Cr, Fe, Co, Zn, Ga, As, Br, La, Ce, Sm, Eu, Yb and Th, weredetermined in small and large samples. A typical in situ relativedetection plot for large size pottery sample (23 g) is given in Fig. 1,which covers the energy range of 122–1811 keV.

The elemental concentration ratios of 8 elements with respectto Na and La/Ce values are given in Table 1. The uncertainties onthe concentration ratios are 2–7%, which are due to countingstatistics only. From the La/Ce values, it was observed that thesefour pottery samples broadly fall into two major groups: group I(OP1, OP2 and OP3) and group II (NP1), though La/Ce value of OP3(0.20) is slightly different from OP1 and OP2 (0.17). The elemental

Table 1Elemental concentration ratios of clay pottery samples determined by IM-NAA.

Element ratio OP1 OP2 OP3 NP1

K/Na 5.870.2 6.470.2 2.0470.05 13.370.3

Sc/Na (2.870.2)E-03 (2.470.1)E-03 (2.0270.05)E-03 (14.070.8)E-03

Ga/Na (3.570.2)E-03 (3.770.2)E-03 (1.8970.06)E-03 (16.070.5)E-03

Fe/Na 6.770.4 10.670.7 6.370.2 44.873.3

Yb/Na (1.1070.08)E-03 (1.4070.07)E-03 (0.7270.05)E-03 (1.7070.08)E-03

As/Na (1.2170.06)E-03 (1.1070.05)E-03 (0.4070.01)E-03 (9.770.6)E-03

La/Na (1.2070.04)E-02 (1.2170.03)E-02 (0.5070.01)E-02 (3.7070.07)E-02

Ce/Na (6.970.3)E-02 (6.870.3)E-02 (2.670.1)E-02 (6.770.2)E-02

La/Ce 0.17 0.17 0.20 0.55

OP: old pottery; NP: new pottery.

Table 2Relative and absolute elemental concentrations in uranium ore samples.

Element (X) UO-1 (100 mg) N=4 UO-1 (55 g) N=1 UO-2 (56 g) N=1

X/Na7SD X/Na Conc.7Unc. (mg kg�1) X/Na Conc.7Unc. (mg kg�1)

U (9.0570.42)E-02 9.23E-02 35874 1.02E-1 494711

K 2.5470.09 2.57 99877110 1.88 91217151

Mn NA 1.54 5984759 6.15E-1 2980760

Cr (6.170.4)E-02 5.8E-02 225712 6.0E-2 291717

Yb (2.5870.15)E-02 2.71E-02 10575 1.78E-2 8674

Sc (2.0270.08)E-03 1.90E-03 7.370.4 1.53E-3 7.470.5

NA: not available; SD: standard deviation.

R. Acharya et al. / Nuclear Instruments and Methods in Physics Research A 622 (2010) 460–463462

concentration ratios of 8 elements (Table 1) appear to suggest thatOP3 is clearly different from OP1 and OP2. Based on theseobservations, four potteries under investigation can becategorized into three different groups [group I: (OP1, OP2);group II (OP3) and group III (NP1)], which matched with theirsample history and collection location. These results showpromise to the LSNAA in the provenance studies ofarchaeological artifacts of non-standard geometries suggestingthat elemental concentration ratios obtained without usingconcentration of an external or internal standards are adequatein the provenance studies.

3.3. Analysis of uranium ore

The IM-NAA method was used for analysis of two uranium oresamples (UO-1 and UO-2), where both large size (55 g of UO-1 and56 g of UO-2) and four small size (100 mg of UO-1) samples weretaken for experiment. Using in situ relative detection efficiency,elemental concentration ratios with respect to Na were deter-mined by IM-NAA. To examine the homogeneity of samples, sixsamples of varying masses (50 mg–56 g) of UO-2 were analysed.The U/Na concentration ratios obtained by IM-NAA were found tobe almost constant from the sample mass of 2 g and above, whichindicated representative sample size of 2 g. Hence large samplesof both sets were analyzed for other elements and the relative andabsolute concentrations are given in Table 2. The relativeconcentrations with respect to Na for homogenized four sub-samples of UO-1 are also given in Table 2 along with unweightedstandard deviations. The concentration ratios of the large sampleof UO-1 were found to be in good agreement with that of foursub-samples (Table 2), indicating quality of the large sampleresults.

Concentrations of Na were obtained in 2 g uranium oresamples using relative method. The uranium concentrations inUO-1 and UO-2 were found to be 35874 and 494711 mg kg�1

(Table 2), respectively. The uncertainties quoted for large size

samples are due to counting statistics and peak fitting errors,whereas for small samples of UO-1 the uncertainties were arrivedfrom standard deviations of four replicates. Method for uraniumdetermination was validated by determining uranium concentra-tion in a 2 g sample of RGU-1. The uranium concentration wasfound to be 41777 mg kg�1 with respect to the certified value of40072 mg kg�1. Results show that large samples can be used formaterials where inhomogeneity is observed or is likely tooccur.

The element sodium, which is a major element in bothpotteries and uranium ores, was used as the monostandard.Though it has a low neutron (n,g) cross-section (0.53 b), it hasmany favorable nuclear properties like it is monoisotopic and itsactivation product has a half-life of 14.96 h and emits two high-energy g-rays (1368.6 and 2754.0 keV) with about 100% abun-dance.

3.4. Composition analysis of stainless steel

Stainless steel SS 304L, which is an austenitic SS, is used as astructure material for the Indian prototype fast breeder reactor(PFBR). For nuclear quality control, it is important to determinethe elemental composition accurately in the SS samples to ensuretheir conformity to specification. The elemental concentrationratios were determined using Fe as monostandard in IM-NAA.Since all the major elements (Fe, Cr, Ni, Mn, etc.) in SS 304L areamenable to NAA and they amount to more than 99%, thecomposition was arrived at by the mass balance method asdescribed in Eqs. (4) and (5) [8]. The elemental concentrationsdetermined are given in Table 3. The uncertainties quoted are dueto propagated counting statistics and peak fitting errors ofelement of interest to the monostandard and they are in therange 3–10%. The concentrations of major elements (Cr, Ni andMn) are found to be within specification limits. Along with majorelements, concentration values of W, Co and As present at tracelevels could be determined.

Table 3Elemental concentrations (in wt.% unless mg kg�1 is indicated) of SS 304L.

Element Eg (keV) Present work Specifications

Fe 1099.3 69.973.3 Balance

Cr 320.1 19.270.6 18.00–20.00

Ni 1481.8 9.571.0 8.00–12.00

Mn 846.8 1.2670.04 2.0 max

Wa 685.7 399713 NA

Coa 1332.5 193712 500 max

Asa 559.1 68.672.9 NA

a in mg kg�1; NA: not available.

R. Acharya et al. / Nuclear Instruments and Methods in Physics Research A 622 (2010) 460–463 463

Concentration values of Fe, Cr, Ni, Mo, Mn and As weredetermined in the austenitic stainless steel CRM, BCS 466 (using asample of 5 g mass) by IM-NAA method using Fe as monostan-dard. The nominal composition of BCS 466 is almost similar tothat of SS 304L. The results obtained are in good agreement(within 75%) with its certified values. The uncertainties on thedetermined values are within 3% except for Ni (10%), which arealso due to counting statistics and peak fitting errors.

Acknowledgements

Authors thank Dr. V. Venugopal, Director, Radiochemistry andIsotope Group and Dr. V.K. Manchanda, Head, RadiochemistryDivision, BARC for their support and encouragement. Authorsthank Mr. K.B. Dasari and Mr. A. Shinde for their help during

experiments. The authors are thankful to the personnel of Apsarareactor for their cooperation during the irradiation of samples.This work was carried out under IAEA Coordinated ResearchProject (CRP Code: F2.30.27).

References

[1] P. Bode, R.M.W. Overwater, J.J.M. De Goeij, J. Radioanal. Nucl. Chem. 216(1997) 5.

[2] A.G.C. Nair, R. Acharya, K. Sudarshan, S. Gangotra, A.V.R. Reddy, S.B. Manohar,A. Goswami, Anal. Chem. 75 (2003) 4868.

[3] R.M.W. Overwater, P. Bode, J.J.M. de Goeij, Nucl. Instr. and Meth. A 324 (1993)209.

[4] R.M.W. Overwater, P. Bode, J.J.M. de Goeij, J.E. Hoogenboom, Anal. Chem. 68(1996) 341.

[5] K. Sueki, K. Kobayashi, W. Sato, H. Nakahara, T. Tomizawa, Anal. Chem. 68(1996) 2203.

[6] K. Sudarshan, A.G.C. Nair, A. Goswami, J. Radioanal. Nucl. Chem. 256 (2003)93.

[7] R. Acharya, A.G.C. Nair, A.V.R. Reddy, A. Goswami, J. Nucl. Mater. 326 (2004)80.

[8] R. Acharya, A.G.C. Nair, A.V.R. Reddy, A. Goswami, Anal. Chim. Acta 522 (2004)127.

[9] R. Acharya, A.G.C. Nair, K. Sudarshan, A.V.R. Reddy, A. Goswami, Appl. Radiat.Isot. 65 (2007) 164.

[10] R. Acharya, K.K. Swain, T.N. Newton, R. Tripathi, K. Sudarshan, P.K. Pujari,A.V.R. Reddy, Explor. Res. At. Miner. 19 (2009) 294.

[11] P.K. Mukhopadhyay, Proceedings of the Symposium on Intelligent NuclearInstrumentation (INIT-2001), Bhabha Atomic Research Centre, Mumbai,2001.

[12] F. De Corte, A. Simonits, Atom. Data Nucl. Data Tables 85 (2003) 47.[13] R. Acharya, J.S.B. Rao, K.C. Jagadeesan, C.R. Venkata Subramani, A.V.R. Reddy,

Proceedings of the Nuclear and Radiochemistry Symposium (NUCAR 2009),Mithibai College, Mumbai, 2009.