Determination of aluminium, silicon and magnesium in geological matrices by delayed neutron activation analysis based on k0 instrumental neutron activation analysis

Applied Radiation and Isotopes 82 (2013) 152–157

Contents lists available at ScienceDirect

Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso

Determination of aluminium, silicon and magnesium in geological matrices by delayed neutron activation analysis based on k0 instrumental neutron activation analysis I.K. Baidoo a,b,n, S.B. Dampare b, N.S. Opata a, B.J.B. Nyarko a, E.H.K. Akaho c, R.E. Quagraine a a

Nuclear Reactors Research Centre, N.N.R.I., Ghana Atomic Energy Commission, Box LG 80, Legon-Accra, Ghana Graduate School of Nuclear and Allied Sciences, Department of Nuclear sciences and applications, University of Ghana, P.O. Box AE1, Atomic, Accra, Ghana c Graduate School of Nuclear and Allied Sciences, Department of Nuclear engineering and computational science, University of Ghana, P.O. Box AE1, Atomic, Accra, Ghana b

H I G H L I G H T S


In this work, concentrations of silicon, aluminium and magnesium in geological matrices were determined by INAA based on k0-IAEA software.
Silicon concentration is best estimated using 29Si (n,p) 29Al reaction within 5 min activation and 15–20 min delay before counting.
The irradiation scheme overestimates magnesium concentration, so we propose a method able to account for quantitative contribution from other reactions.
The method employed is quick and robust for estimating silicon, aluminium, magnesium and other elements using k0-method.

art ic l e i nf o

a b s t r a c t

Article history: Received 13 September 2012 Received in revised form 12 June 2013 Accepted 23 July 2013 Available online 12 August 2013

In this work, concentrations of silicon, aluminium and magnesium in geological matrices were determined by Neutron Activation Analysis based on k0-IAEA software. The optimum activation and delay times were found to be 5 min and 15–20 min respectively for the determination of Si via 29Si (n,p) 29 Al reaction. The adopted irradiation scheme did not work for the determination of magnesium. Each sample was irradiated under a thermal neutron flux density of 5.0  1011 n cm  2 s  1. Cadmium covered activation indicated that a permanent epithermal irradiation site for research reactors would be very useful for routine determination of silicon in environmental samples. & 2013 Elsevier Ltd. All rights reserved.

Keywords: k0-IAEA software INAA Magnesium Aluminium Silicon Geological matrix

1. Introduction One of the advantages of Instrumental Neutron Activation Analysis (INAA) is its relative simplicity which allows less technical people to be trained on the job to be able to do elemental analysis. However determination of certain elements requires a (high) technical understanding of the activation (nuclear reaction) and/or decay mechanisms in order to obtain very good results. One such difficulty is the determination of aluminium (Al), silicon (Si) and magnesium (Mg). In many instances, the inability or difficulty to produce radioactive n Corresponding author at: Nuclear Reactors Research Centre, N.N.R.I., Ghana Atomic Energy Commission, Box LG 80, Legon-Accra, Ghana. Tel.: +233 246413153. E-mail addresses: [email protected], [email protected] (I.K. Baidoo).

0969-8043/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.apradiso.2013.07.032

products by activation of low atomic number elements such as hydrogen (H), carbon (C), nitrogen (H), oxygen (O), and silicon (Si) had been considered as an advantage because it means less interference which translates to improved sensitivity (Robert et al., 2011). However, silica or silicates are major constituent of materials such as rocks, lavas, slags, refractories, ceramics, glasses, cements and ashes (samples for routine analysis by INAA). For most environmental and/or geological sample analysis, major elemental components like calcium, aluminium, magnesium as well as silicon are of great importance to the researcher. Therefore, there is the need to ensure high degree of accuracy in the determination of these elements by INAA. There are two main kinds of interferences in the calculation of trace element concentration by INAA (Alfassi, 2000). The first is the formation of the same radionuclide from two different elements and the other is gamma interferences during spectrometry measurement

I.K. Baidoo et al. / Applied Radiation and Isotopes 82 (2013) 152–157

due to gamma interactions with the detector medium. The former could be reduced by utilizing neutron interactions (reactions) other than the (n,γ) reaction, mostly produced by thermal neutron. Fortunately, typical neutron source(s) used for activation analysis (especially a research reactor) is constituted mainly of three energy groups. Thermal neutrons, epithermal neutrons and fast or fission neutrons, and their relative abundances depend on the reactor structure. Usually thermal neutrons of a typical research reactor constitute the majority of neutrons (up to about 90–95%). In the usual INAA, the whole reactor neutron energy spectrum is used for irradiation where each energy group may contribute to nuclear reaction of the target isotope in one or more of these interactions (n,γ), (n,n), (n,p), (n,2n) and or (n,α) reactions (Glascock, 2004). However, situations in which part of the neutron spectrum is used are also preferable due to the characteristic differences in the activation cross-sections for the desired and the interfering nuclides in the various parts of the energy spectrum or interfering gamma lines during spectrum acquisition. It is however clear that an advantage can be taken of using the epithermal and fast neutrons in the case where the required element of interest is activated more strongly by them relative to the major elements or to the interfering element (branch activation analysis called Epithermal Neutron Activation Analysis – EPNAA). For example, Mg is normally determined through its less abundant isotope (26Mg – 11.01%) by means of the (n,γ) reaction to form 27Mg radionuclide [26Mg (n,γ) 27Mg]. However, the cross-section for this reaction is low (see Table 1 below), and makes the measurement not very sensitive. More so, it suffers interference from the (n,p) reaction on 27Al to form 27Mg [27Al (n,p) 27Mg] which emits the same gamma energy (γline) as the (n,γ) reaction. Aluminium is also determined by the activity of its radionuclide 28Al via the (n,γ) reaction of 27Al [27Al (n,γ) 28Al]. It can also be produced by the (n,α) reaction on 31P isotope [31P (n,α) 28 Al] and by the (n,p) reaction on 28Si [28Si (n,p) 28Al] (Alfassi and Lavi, 1984). In the case of Si determination, there are three stable isotopes involved: 28Si (92.2%), 29Si (4.7%) and 30Si (3.1%), but the only radionuclide produced by the (n,γ) reaction is 31Si. However, this radioisotope is barely a γ-ray emitter with an intensity of 0.07% at the 1266keV γ-line. From Table 1, one can also observe that 31Si has relatively high crossection and a longer half life (s ¼0.108 b, half life¼2.62 h) for which longer activation in an intense neutron source and longer decay time may prevent interference by the 1st escape peak of 28Al γ-line to 1266 keV γ-line. This may allow the use of the (n,γ) reaction for the determination of Si (Kathryn, 1996). However, for low power research reactors, the activation of Si by higher energy neutrons (EPNAA) leads to the formation of 28Al via both 31P (n,α) 28Al and 28Si (n,p) 28Al reactions and 29Al by the 29Si (n,p) 29Al reaction. In cases where the concentration of Si is high enough (as in sediment, soil and rock matrices), Si can best be determined by the 29Si (n,p) 29Al reaction (Hancock, 1982) though the use of the 29Al radioisotope may also reduce the accuracy of the measurement and limit the minimum amount (high detection limit) of Si that can be determined. The other problem associated with the measurement of 29Al is that its main γline at 1273-keV is interfered by the 1st escape peak (1266 keV) from

153

28

Al of γ-line 1778 keV by virtue of pair production of γ-detector interactions. In the Ghana Research Reactor-1 facility (GHARR-1), the interrelationship between the determination of Al, Si and Mg have not received much attention, as is the case in most of of the literature with few exceptions (Hancock, 1982; Kathryn, 1996; Alfassi and Lavi, 1984). This may be due to one or more of these reasons; GHARR-1 is a research reactor known for its high thermal neutron component (thermal to epithermal ratio (f)) of 18.8 (Akaho and Nyarko, 2002), and all elemental determination has been considered to be by means of thermal neutrons, the contribution from epithermal and fast neutrons are considered negligible except where conscious effort is made to irradiate the sample under a cadmium or boron shield. Hence the determination of Si is not entirely encouraged (not routine). Generally, most research reactors were designed without epithermal activation in mind and hence EPNAA would require conscious effort by the technician to use either boron or cadmium absorbers in order to achieve substantial (n,p) reactions (Hancock, 1982; Alfassi and Lavi, 1984). Also in facilities where the transfer systems are manually operated, this method (EPNAA) becomes radiologically inconvenient, time consuming and may not be employed for routine analysis except in a few cases when the need arises. These seeming problems are compounded by the fact that the INAA quantification method employed in the GHARR-1 facility (and other INAA laboratories) for routine elemental analysis is the relative comparator method and in most cases multi-element analysis is of major interest and convenience. There is, however, no known reference standard which contains silicon that could give better counting statistics (if routine irradiation and counting is applied) in order to be used in quantitative analysis of silicon in samples, if at all, and the peak (1273-keV line) could be statistically appreciable for estimation. However, the current work is motivated by the fact that the k0-method allows for a single comparator (gold monitor) element to be used in concentration evaluation of any identifiable element in the sample. As such the idea for this work would be to improve counting statistics for the 1273-keV γ-line of the sample and quantify the concentration based on a single comparator method (k0 via k0-IAEA software). The flexibility, accuracy, precision and general acceptance of the k0-INAA technique relative to classical NAA (relative comparator method), without the tedium for preparation of synthetic standards from pure substances, have been reported by Matthias and Blaauw (2006). With the k0-Instrumental Neutron Activation, the analysis rather borrows on well characterized reactor neutrons (Baidoo et al., 2013) and constant in-situ thermal neutron flux monitoring (comparator monitor) during the irradiation of samples. Then, the comparator element can be used in routine measurements instead of needing separate standards for each element, based on equation (1) below: (Girardi et al., 1965): ρ¼

ððpa =t c Þ=SCDWÞ 1 ½f þ Q 0 ðαÞau  εpau ððpau =t c Þ=SCDwÞ k0 ½f þ Q 0 ðαÞa  εp ðEy Þ

ð1Þ

Table 1 Relevant nuclear data for Mg, Al and Si [IAEA-TECDOC-564 (1990)]. Target isotope

s (b)

RI

Isotopic abundance (%)

Reaction

Half-life

Energy (keV)

γ-Emission probability (%)

26

0.0372 0.226 0.108

0.024 0.16 0.106

0.1101 1 0.031

26

9.46 m 2.24 m 2.62 h

1014.43, 843.76 1178.99 1266.20

71.4 100 0.07

Half-life (m)

Energy (keV)

γ-Emission probability (%)

2.24 6.56 9.46

1178.99 1273.36 1014.43, 843.76

100 91.30 71.4

Mg Al 30 Si 27

Mg (n,γ) 27Mg Al (n,γ) 28Al 30 Si (n,γ) 31Si 27

Crosssection averaged in a 235U fission neutron spectrum [(n,p) reaction crosssection] Target isotope s (mb) Isotopic abundance (%) Reaction 28

Si Si 27 Al 29

6.4 3.01 4.0

0.9223 0.0467 1

28

Si (n,p) 28Al Si (n,p) 29Al 27 Al (n,p) 27Mg 29

154

I.K. Baidoo et al. / Applied Radiation and Isotopes 82 (2013) 152–157

where, p is the net number of counts in the full-energy peak, au refers to the co-irradiated gold monitor, W is the weight of the sample, w is the weight of the gold monitor, f is the thermal to epithermal neutron flux ratio, Q0 ¼I0/s0 is the resonance integral (I0) to thermal neutron cross-section (s0) ratio, α is the measure for the epithermal neutron flux distribution, approximated by the 1/E1+α dependence with α considered to be independent of neutron energy.ε ¼absolute photo-peak efficiency of the detector, S is given as; ð1eλtir Þ=λ (saturation correction during activation), D is the decay correction (eλtd ), C is the counting correction given as; ð1eλt d Þ while λ¼ decay constant of the activated nuclide, t ir ¼irradiation time, t d ¼decay time and t c ¼counting time. This work seeks to take advantage of the epithermal and fast neutron contributions to the (n,p) reaction for 29Si by extending activation time to improve the 29Si (n,p) 29Al reaction for Si and extend the delay time to ensure decay of 28Al to reduce interference by the escape peak from 28Al. The activation procedure would be repeated under a cadmium cover. The main purpose would be to provide an interim routine procedure for the determination of Si and also to provide pre-assessment for the proposed conversion of the High Enriched Uranium (HEU) core of GHARR-1 to a Low Enriched Uranium (LEU) core to factor into the design a permanent epithermal neutron irradiation site for routine determination of silicon and other elements whose determinations are favoured by ENAA.

2. Methodologies 2.1. Sample preparations and packaging Two separate sample preparations were done, one of each for bare and cadmium covered activations. Three replicates of each reference material were weighed (100 mg) onto a polyethylene rubber film and wrapped using laboratory stainless steel forceps. The polyethelene rubber film was heat sealed to a pellet of about 6 mm length and 4 mm thickness. A one millimetre (1 mm) thick cadmium plate was cut and folded into a cylinder of about 15 mm height and 5 mm radius, with an appropriate bottom and top cover. For the cadmium covered activation, each sample pellet was encapsulated in a small polyethylene capsule of 11.7 mm diameter and 10 mm height and then placed in the cadmium cylinder and further encapsulated in an irradiation capsule [16.5 mm diameter and 56.5 mm height (rabbit capsule)] and sealed for irradiation. However, for bare activations each sample pellet was packaged into the irradiation capsule and sealed. Separate monitors were prepared for both bare and cadmium covered activation for which the monitor was affixed to the first sample to be activated in order to map the flux for the other activations.

identification of γ-rays of product radionuclides were identified by their γ-ray energy (ies) via ORTEC MAESTRO-32. The peak reduction and interpretation of the gamma spectrum and elemental quantifications (elemental concentrations) of samples, controls and monitors were done by the k0-IAEA version 4.04 (update dates to February 4, 2009 and is named k0-IAEA version 4.04).

3. Result and discussions Fig. 1 clearly indicates the usefulness of cadmium covered activation to cut off thermal neutrons to reduce activation for certain elements with crossections highly favoured by thermal neutrons and hence reduces interference. Both (Fig. 1) spectra had been delayed enough to ensure that the single escape γ-line for 28Al (1778-keV) had been reduced in order to enhance the 1273-keV γ-line for 29Al produced from the stable 29Si isotope of silicon. The bare activation showed enough interference from the doublet (constituting of both 1266-keV and 1273-keV γ-line) which may still be difficult for most spectrum evaluation software to resolve and properly obtain accurate results. As mentioned earlier in the introduction, amongst the stable isotopes of silicon [28Si (92.2%), 29Si (4.7%) and 30Si (3.1%)], the only radionuclide produced by the (n,γ) reaction is 31Si which is more of a beta emitter than γ-emitter and even the intensity (0.07%) is not favourable for analysis. This makes the determination of silicon by γ-spectroscopy of (n,γ) reaction not only difficult but also limited. However, from Fig. 1, it clearly appears that the best γ-line and reaction to determine silicon in the midst of these interferences, is 1273-keV γ-line from the 29Si (n,p) 29Al reaction. As indicated in Fig. 2, the 1273-keV γ-line becomes very much improved when the counting was delayed for enough time to allow for the decay of 28Al. From Fig. 2, it can be observed that as the activated (bare activation) sample is allowed to decay, interference from the 28Al 1st escape peak reduces to allow proper statistics for the 1273-keV γ-line to be obtained which enables the estimation of Si via 29Si (n,p) 29Al reaction (the adopted irradiation scheme for determination of silicon). This method of schematic irradiation [using the reaction interaction and half-lives of element(s) of interest] is robust and has an added advantage of having been developed from routine samples of interest (geological matrices/ sediment). The method could easily be adopted for routine determination of Si compared to what is shown in the existing literature (Hancock, 1982; Alfassi and Lavi, 1984; Kathryn, 1996) for which Alfassi and Lavi (1984) employed concurrent irradiations of the same sample

2.2. Sample activation, counting and interpretations The activations were performed at irradiation siteB2 of the GHARR1 facility at a power of 15 kW with an average neutron flux of 5.0  1011 n cm  2 s  1 (as indicated on the reactor console). Each sample was irradiated for 5 min and allowed delay times of about 15–20 min and data acquisition of the activated sample was done for 600 s. This was to ensure enough decay of 28Al to reduce interference caused by the 1st escape peak (1266-keV) of 1778-keV γ-line which interfere with the 29Si 1273-keV γ-line (as indicated in Fig. 2). Radioactivity measurement of the induced radionuclide was performed by a PC-based γ-ray spectrometry set-up. It consisted of an N-type High purity Germanium detector (HpGe-coaxial type) coupled to a computer based multi-channel analyzer (MCA) via electronic modules. The relative efficiency of the detector is 40%. Its energy resolution is 1.8 keV at a γ-ray energy of 1332 keV of 60Co. The data acquisition and

Fig. 1. Comparison of two gamma SPEC (Spectrum) displayed on the Ortec Maestro under ‘compare file’ mode showing both activation done under a cadmium cover and bare, emphasizing the reduced interference for cadmium covered activation for 1273-keV γ-line by single escape peak (SEP) from 28Al 1778-keV γ-line.

I.K. Baidoo et al. / Applied Radiation and Isotopes 82 (2013) 152–157

155

Fig. 2. Multiplete peaks (1273 keV γ-line and 1266 keV γ-line) as resolved by k0-IAEA software. The multiplet becomes visible after the activated sample had decayed for a minimum of 10 min.

Table 2 Concentration of CRMs obtained in this work compared to the reported concentrations; each activation was done for 5 min and delayed for 10–15 min. GBW07106

Activation type Bare

Element

Concentrations (%)

Al Mg Si

1.87 7 0.204 0.6217 0.739 40.17 7.34

ESSD Cadmium covered

0.052 7 0.11 0.767 0.18 43.46 7 4.48

Activation type Bare

Reported

Element

Concentrations (%)

1.86 7 0.069 0.049 7 0.018 42.22 7 0.09

Al Mg Si

1.97 70.14 0.79 70.10 39.23 77.24

(s) under both reactor neutrons (bare) and epithermal neutrons (under cadmium cover) while Hancock (1982) suggested irradiation of the sample only under cadmium container for the epithermal activation of products 29Si (n,p) 29Al, to be enhanced. However, we have indicated that such method is difficult (especially if the facility is not automated) for routine analysis and sometimes radiologically unfriendly. More so, Alfassi and Lavi's (1984) method led to solving complicated six equations (even to begin with). However, our work was used to develop a routine analytical method (only bare activation) for the determination of Si, Mg and Al while considering their reaction interaction and spectrometric interferences, taking advantage of their half-lives and decay process to improve counting statistics for Si (1273 keV gamma energy for its n,p reaction). Our method is based on the single comparator method (k0) and for that matter does not require like sample or standard for quantification but simultaneous irradiation of gold comparator and the sample(s) (analyte). The method is quite simple and requires only one set of irradiation. Nonetheless, we acknowledge that our work is limited by the fact that it cannot be used to quantify the contribution (interference) of each of these elements (Si, Al and Mg) to one another but it is able to give reliable results for Al and Si as indicated in Tables 3 and 4.

Cadmium covered Reported 0.617 0.06 1.02 7 0.08 42.42 7 6.76

2.2977 0.018 0.388 7 0.009 407 0.16

The cadmium covered activation (epithermal neutron activation analysis) also indicated an enhanced (n,p) reaction while the (n,γ) reaction was inhibited for 30Si and other target nuclide including 27Al. These are very significant and translate into an improved sensitivity for the determination of Si compared with the results of early researchers; (Hancock, 1982) [as low as 20.7% silicon was achieved even with a lower thermal neutron flux 2.5  1011 n cm  2 s  1 compared to the current work]. However, this method (cadmium-coverd or EPNAA activation) failed for the quantitative determination of both Mg and Al, as indicated by the results presented in Table 2. The failure could be associated with the reaction mechanisms of both Al and Mg. From Table 1, it can be observed that when the target 27Al undergoes an (n,p) reaction [27Al (n,p) 27Mg], it ends up forming 27Mg. This is probably the reason why in all the cases (Tables 2 and 3) the concentrations determined for Mg were higher than the values reported on Certificate Reference Materials (CRMs) and the corresponding low concentrations for Al. The inter reaction relationship between these elements (27Al (n,p) 27Mg and 28Si (n,p) 28Al) may be reduced by schematic irradiation and counting. In order to reduce the interrelationship between Al and Mg and at the same time improve the detection

156

I.K. Baidoo et al. / Applied Radiation and Isotopes 82 (2013) 152–157

Table 3 An improved determination of Mg by bare irradiation of the samples for 30 s and delayed for 10 min and counted for 10 min. GBW07106

ESSD

Element

Concentration (%)

Reported

Element

Concentration (%)

Reported

Al Mg Si

1.737 0.14 0.059 7 0.02 ND

1.86 70.069 0.049 70.018 42.22 70.09

Al Mg Si

2.42 7 0.24 0.54 7 0.20 ND

2.2977 0.018 0.388 7 0.009 407 0.16

GBW07106: Rock References Materials GBW07106. ESSD: NIST Standard Reference Material 1646a Estuarine Sediment. ND: gamma peak (γ-line) not detected.

Table 4 Elements detected and quantified from the full spectrum after 5 min bare activation. All the concentrations are in mg/kg except if per cent (%) or compound oxide is indicated. GBW07106

Estuarine sediments 1646a

Elements

Concentration

Reported

Elements

Concentration

Reported

Al2O3 CaO Cl K2O MgO Mn Na2O SiO2 Ti V Dy Ba

3.53 7 0.20 0.84 7 0.061 1177 15.68 0.62 7 0.078 1.047 0.074 145.7 7 13.99 0.047 0.003 85.81 7 7.34 17467 188.57 36.117 3.76 11.26 7 2.13 114.17 26.31

3.52 7 0.13 0.30 7 0.05 447 9 0.65 7 0.04 0.082 7 0.03 1557 10 0.0617 0.021 90.36 7 0.20 1580 7 120 337 4 4.17 0.5 1437 22

Al (%) Ca (%) Cl K (%) Mg (%) Mn Na (%) Si (%) Ti (%) V Dy Ba

1.9677 0.18 1.2197 0.11 5650 7 515 1.0637 0.21 1.0197 0.12 236.7 7 21.78 0.69147 0.063 39.23 7 7.24 0.32767 0.051 44.57 74.19 6.03 70.844 ND

2.2977 0.018 0.5197 0.020 NR 0.8647 0.016 0.388 7 009 234.5 7 2.8 0.7417 0.741 407 0.16 0.456 7 0.021 44.84 7 0.76 NR (210)

ND: insufficient counting statistics. NR: not reported. In all cases the source of uncertainties (‘7 ’) associated with each result was calculated and generated by the k0-software which is reported to be those uncertainties as a result of random uncertainties. (Uncertainties associated with the following measurement; peak area, flux, sample mass, dry/wet ratio, sample thickness, detector geometry.)

limit for determining Mg, it was found out that, a reduced activation time of 30 s improves (does not necessary resolve the increased concentration for Mg) the simultaneous determination of Al and Mg. This was because for a short period of activation, (n,p) reaction of 27Al target to 27Mg and that of 28Si target to 28Al radioisotope are likely to reduce and hence yield a more reliable result as presented in Table 3 compared to the high concentration (blotted) results for Mg presented in Table 2. The results presented in Table 2 may also be used to explain the reasons why in most cases the concentration of Al seems to be quite acceptable, even though the target nuclide (28Al) undergoes an (n,p) reaction to form 27Mg (normally the concentration is expected to reduce). However it can also be observed from Table 1 that, when the target nuclide of 28Si undergoes an (n,p) reaction it produces the 28Al radioisotope which emits gamma γ-line of 1779-keV just as when 27Al undergoes the (n,γ) reaction. This situation is a sort of balance; 27Al going to 27Mg and 28Si coming to 28 Al. These interrelations between 28Si (n,p) 28Al and 27Al (n,p) 27 Mg, may explain why the concentration of Al seems to be quite acceptable in Table 2. However the slightly reduced concentration of Al may be understood from the view point that the crossections for (n,p) reactions, the resulting radionuclide, its half lives and their gamma abundance (as seen from Table 1) are not the same for both 27Al (n,p) 27Mg and 28Si (n,p) 28Al and hence might not completely be a balanced exchange for Al and Si. This imbalance could clearly be seen from cadmium covered activation in which Al was in most case not detected due to the dominance of fast and epithermal neutrons, which were the most dominant contributors in such activation analysis (EPNAA). In order to reduce peak interference with respect to the escape peak from1778.9-keV γ-line and improve the ability of the detector

to resolve the 1266-keV γ-line and 1273-keV γ-line, the activated sample was allowed to decay for 10–15 min before counting which saw gradual dominance of the 1273.36 γ-lines as the 28Al decays (see Fig. 2). This made it less difficult for the spectrum evaluation software (k0-IAEA software) to resolve the multiplet (1266-keV γ-line and 1273-keV γ-line; see Fig. 2). In this work we found that the optimum time for counting after 5 min of bare and cadmium covered activation was 10–15 min. Tables 4 and 5 show all elements identified and quantified in the full gamma spectrum after the sample had been irradiated for 5 min. Table 5 presents the results for a routine test analysis on two different like matrices from ISE – International Soil-Analytical Exchange [Wageningen Evaluating Programs for Analytical Laboratories (WEPAL)] and standard reference materials GBW07107 (note: the code name for the ISE samples as presented in Table 5 are those adopted for our laboratory). The other relatively long lived elements like sodium (Na), barium (Ba) and potassium (K) were determined after the same radioactive sample had been further recounted for a decay time of 25–30 min. Simultaneous determination of these elements is necessary for reducing the analytical turnaround time (within an hour), and also the longer activation time improves the counting statistics which eventually improve sensitivity and accuracy of these relatively short lived radionuclides. It must also be noted that Tables 4 and 5 does not represent acceptance criteria for all the detected elements on the basis of our laboratory practice but an indication of what must be expected when the irradiation scheme employed in this work is used elsewhere. It would therefore be necessary for the technician or the technologist to accept or reject some of the results on the basis of quality control and assurance of his/her laboratory.

I.K. Baidoo et al. / Applied Radiation and Isotopes 82 (2013) 152–157

157

Table 5 Routine test analysis on ISE – International Soil-Analytical Exchange (Wageningen Evaluating Programs for Analytical Laboratories – WEPAL) and standard reference materials GBW07107. All the concentrations are in mg/kg except if per cent (%) or compound oxide is indicated. Sample

114 ISE 2 This work

Element Al Ca Cl K Mg Mn Na Si% Ti V Dy Ba

45040 7 4369 147207 1413 ND 14450 7 1690 16570 71822 11417 111.8 60107 546.9 36.50 77.99 2569 7 488.1 91.737 12.93 8.18 71.13 ND

103 ISE 3 Reported

This work

NDA

NDA st. dev.

46220 12230 NR 14755 9853 1057 707.1 33.90 3042 80.88 NR 236.2

1900 1010 491 533 62 484 0.82 204 7.00 16.9

16590 7 1708 49870 7 4637 392.30 7 37.66 77127 948.6 45550 7 4692 195.80 7 1801 21917 201.6 39.517 11.13 1546 7 649.32 30.647 6.74 5.45 7 0.6 153.30 7 33.73

GBW07107 Reported NDA

NDA st. dev.

18020 69520 NR 7967 8936 202.3 2362 33.39 1815 38.05 NR 246.2

1120 6700 402 644 37.8 468 1.05 301 10.47 15.8

Elements

This work

Reported

Al2O3 CaO Cl K2O MgO Mn Na2O SiO2 Ti V Dy Ba

17.8 7 1.71 0.95497 1976 68.757 7.93 4.114 75101 3.353470.55 159.9 7 15.51 0.27627 0.0268 ***ND 4009 7436.98 104.8 711.42 4.5617 0.76 386.2 753.73

18.82 7 0.22 0.60 7 0.06 (37) 4.167 0.15 2.017 0.07 173 717 0.3570.03 59.23 7 0.23 3950 7 190 877 6 5.17 0.5 450 7 45

NDA – Normal distribution Approximation, NDA st. dev – NDA Standard Deviation, (37) – Non certified result, ND – insufficient counting statistics, and ISE – International SoilAnalytical Exchange [Wageningen Evaluating Programs for Analytical Laboratories (WEPAL)]. In all cases the source of uncertainties (‘7 ’) associated with each result was calculated and generated by the k0-software which is reported to be those uncertainties as a result of random uncertainties. (Uncertainties associated with the following measurement; peak area, flux, sample mass, dry/wet ratio, sample thickness, detector geometry.)

4. Conclusions/recommendation

References

This method is a fast procedure developed to estimate the concentrations of Si, Al and Mg and other elements in geological samples. Optimum activation and counting times were found to be 5 min with a decay time between 15–20 min and 10 min to accumulate the spectra. With this irradiation scheme, enough activation is done for the 29Si (n,p) 29Al reaction with a delay between 15 and 20 min allowing for enough decay of 28Al γ-line to reduce spectrum interference of the single escape peak (SEP) of 28Al (1266 keV γ-line to 1273 keV γ-line of 29Al). However, a 5 min irradiation over estimates the concentration of Mg but a reduced irradiation time of 30 s strongly improves the results by reducing the contribution of the (n,p) reaction from Al. It was also found that cadmium covered activation (epithermal neutron or fast neutrons) is favoured for the determination of Si. It is recommended that pure Mg and Al should be used to investigate the level of interrelationship (how much (n,p) reaction of 27Al is contributed to the concentration of Mg per Al concentration for a particular reactor facility) between Mg and Al during activation in order to quantify the level of Al contribution to Mg. We also recommend that the committee and the expert partners undertake the conversion of the HEU core of GHARR-1 to an LEU core to factor into the design, a permanent epithermal neutron irradiation site in order to facilitate the routine determination of silicon and other equally high favoured epithermal activation.

Akaho, E.H.K., Nyarko, B.J.B., 2002. Characterization of neutron flux spectra in irradiation channels of MNSR reactor, using Westcott-formalism for the k0 neutron activation analysis. Appl. Radiat. Isot., 265–273. Alfassi, Z.B., 2000. Instrumental Neutron Activation Analysis. In: Meyer, R.A. (Ed.), Encyclopedia of Analytical Chemistry. John Wiley and Sons Ltd., Chichestre, pp. 12497–12526. Alfassi, Z.B., Lavi, N., 1984. Simultaneous determination of Na, Mg, Al, Si and P by instrumental neutron activation analysis under reactor and epithermal neutrons. The Analyst 109, 959–962. Baidoo, I.K., Nyarko, B.J.B., Akaho, E.H.K., Dampare, S.B., Sogbadji, R.B.M., Poku, L.O., 2013. Characterization of low power research reactor neutrons for the validationof k0-INAA standardization based on k0-IAEA software. Appl. Radiat. Isot. http://dx.doi.org/10.1016/j.apradiso.2013.05.005. Glascock, M.D., 2004. An overview of neutron activation analysis. 〈www.missouri. edu/  glascock/archlab.htm, retrieved August 14 2012〉. Girardi, F., Guzzi, G., Pauly, J., 1965. Reactor neutron activation analysis by the single comparator method. Anal. Chem. 37, 1085–1092. Hancock, R.G.V., 1982. On the determination of silicon in pottery. J. Radioanal. Nucl. Chem. 69 (1–2), 313–328. IAEA-TECDOC-564, 1990. Practical aspects of operating NAA laboratory. Vienna, pp. 197–237. Kathryn, R.W., 1996. Neutron activation analysis of zeolite catalysts. J. Radioanal. Nucl. Chem. Lett. 212 (5), 361–371. Matthias, R., Blaauw, M., 2006. Progress report in the k0-IAEA program. Nucl. Instrum. Methods Phys. Res. A 564, 698–701. Robert R. Greenberg, Peter Bode, Elisabete A., De Nadai Fernandes, 2011. Neutron activation analysis: a primary method of measurement. Spectrochimica Acta. Part B, Atomic Spectroscopy.