Đang tải... (xem toàn văn)
1. Giới thiệuGammagamma (gg) là sự trùng hợp ngẫu nhiênkỹ thuật xác định và hoặc định lượng phân rã hạt nhânsự kiện dựa trên sự quan sát của nhiều gray duy nhấtchữ ký. Tính trùng hợp ngẫu nhiên, vì nó thườnggọi, sử dụng ít nhất hai máy dò gray để đocác chùm đồng hồ được phát ra trong mỗi sự kiện phân rã. Trongnguyên tắc, gg đếm sự trùng hợp có thể đạt được một mức cao hơnmức độ phân biệt đối xử hơn không ngẫu nhiên ( single )quang phổ vì nó áp dụng một định nghĩa nghiêm ngặt hơnnhững gì tạo thành một sự kiện hợp lệ, cụ thể là quan sáthai tia g tương quan phân rã trong một khoảng thời gian nhất định.Yêu cầu này rất hữu ích trong việc tách các sự kiện quan tâmtừ số lượng lớn hơn của trở lại không tương quanmặt đất trong một khoảng thời gian tính.Tính trùng lặp được sử dụng thường xuyên bởi hạt nhâncác nhà quang phổ học để làm sáng tỏ các chương trình phân rã phức tạp. Cácáp dụng sự trùng hợp ngẫu nhiên với kích hoạt neutronphân tích (NAA) đã được báo cáo trong tài liệu trên 40 nămtrước 13. Kể từ đó, một số giấy tờ đã đượcxuất bản mô tả việc sử dụng sự trùng hợp ngẫu nhiên trongNAA (sau đây gọi là cNAA) 47, nhưng cNAA chưa trở thànhcông cụ phổ biến của hóa học phân tích hạt nhân. Có thểgiải thích cho sự khác biệt này trong ứng dụng là hạt nhâncác nghiên cứu cấu trúc đòi hỏi một sự xác định sự trùng hợp ngẫu nhiênmối quan hệ giữa các tia g để xây dựng hạt nhâncấp bậc; trong khi quang phổ hạt nhân khi sử dụngtrong NAA phụ thuộc vào kiến thức về nguồn gốc của các nguyên tốvà ở mức độ thấp hơn, cường độ của tia g, nhưng khôngđòi hỏi một sự hiểu biết thân mật về sự trùng hợp graycác mối quan hệ. Vì vậy đã có ít động lực để áp dụngkỹ thuật để đo lường phân tích hơn là cóđã được cho các nghiên cứu cấu trúc hạt nhân.Việc thực hiện cNAA cũng được giới hạn ở mức độbởi cơ cấu hạt nhân. Để một phần tử đượcđược xác định bởi cNAA nó phải có một đồng vị dân cưbằng cách bắt giữ nơtrôn phát ra một thác nước gquang hợp lýcường độ. Nhiều nuclit có gray cascades, nhưng thường làb thức ăn nhánh là yếu hoặc cường độ tuyệt đối của tia gammanhỏ. Đồng thời, có một số đồng vị (ví dụ: 203Hgvà 51Cr) quan trọng đối với NAA phát thải chỉ có một btrì hoãngray. Kết quả ròng là số lượng các yếu tốcó thể được xác định bởi cNAA nhỏ hơn sốcó thể được xác định bằng NAA thông thường
ARTICLE IN PRESS Nuclear Instruments and Methods in Physics Research A 589 (2008) 243–249 www.elsevier.com/locate/nima gg coincidence spectrometer for instrumental neutron-activation analysis B.E TomlinÃ, R Zeisler, R.M Lindstrom Analytical Chemistry Division, National Institute of Standards and Technology, Gaithersburg, MD 20899-8395, USA Received 21 February 2008; accepted 28 February 2008 Available online March 2008 Abstract Neutron-activation analysis (NAA) is an important technique for the accurate and precise determination of trace and ultra-trace elemental compositions The application of gg coincidence counting to NAA in order to enhance specificity was first explored over 40 years ago but has not evolved into a regularly used technique A gg coincidence spectrometer has been constructed at the National Institute of Standards and Technology, using two HPGe g-ray detectors and an all-digital data-acquisition system, for the purpose of exploring coincidence NAA and its value in characterizing reference materials This paper describes the initial evaluation of the quantitative precision of coincidence counting versus singles spectrometry, based upon a sample of neutron-irradiated bovine liver material Published by Elsevier B.V PACS: 82.80.Jp; 29.30.Kv; 82.80.Ej Keywords: Instrumental neutron-activation analysis; Gamma–gamma coincidence; g-Ray spectrometry; Digital data acquisition Introduction Gamma–gamma (gg) coincidence spectrometry is a technique for identifying and/or quantifying nuclear decay events based on the observation of unique multiple g-ray signatures Coincidence counting, as it is commonly referred, employs at least two g-ray detectors to measure the coincident g-rays emitted in each decay event In principle, gg coincidence counting can achieve a higher degree of discrimination than noncoincidence (‘‘singles’’) spectrometry since it applies a more stringent definition of what constitutes a valid event, namely the observation of two decay-correlated g-rays within a specified time window This requirement is useful in separating events of interest from the much larger number of uncorrelated ‘‘background’’ events in a counting period Coincidence counting is used routinely by nuclear spectroscopists to elucidate complex decay schemes The application of coincidence counting to neutron-activation analysis (NAA) was reported in the literature over 40 years ago [1–3] Since then, a number of papers have been ÃCorresponding author Tel./fax: +1 301 975 6283 E-mail address: bryan.tomlin@nist.gov (B.E Tomlin) 0168-9002/$ - see front matter Published by Elsevier B.V doi:10.1016/j.nima.2008.02.094 published describing the use of coincidence counting in NAA (hereafter cNAA) [4–7], yet cNAA has not become a common tool of nuclear analytical chemistry One possible explanation for this difference in application is that nuclear structure studies require a determination of coincidence relationships between g-rays in order to construct nuclear level schemes; whereas nuclear spectrometry as employed in NAA depends on a knowledge of the elemental origins and, to a lesser extent, the intensities of g-rays, but does not require an intimate knowledge of g-ray coincidence relationships Thus there has been less motivation to apply the technique to analytical measurements than there has been for nuclear structure studies The implementation of cNAA is also limited to a degree by nuclear structure In order for an element to be determined by cNAA it must have an isotope populated by neutron capture that emits a g-ray cascade of reasonable intensity Many nuclides have g-ray cascades, but often the b-branch feeding is weak or the g-ray absolute intensities are small Also, there are a number of isotopes (e.g 203Hg and 51Cr) important for NAA that emit only one b-delayed g-ray The net result is that the number of elements that can be determined by cNAA is smaller than the number that can be determined by conventional NAA Cooper has ARTICLE IN PRESS 244 B.E Tomlin et al / Nuclear Instruments and Methods in Physics Research A 589 (2008) 243–249 provided a reasonable list of nuclides that may benefit from cNAA [8] Given the aforementioned restrictions, it is clear that cNAA is best described as a niche application Despite the apparent reasons for a lack of cNAA development, this technique does indeed offer worthwhile advantages to the nuclear analytical community The higher level of discrimination that can be obtained with coincidence counting has been demonstrated in the past to yield improved sensitivity and selectivity in the determination of trace elements by NAA [8] This paper describes a gg coincidence spectrometer constructed at the National Institute of Standards and Technology (NIST) using stateof-the-art digital data-acquisition electronics and reports on the initial evaluation of background-suppression characteristics of the spectrometer 6-L LN2 Dewar 63% HPGe Sample Position 31% HPGe Experimental 2.1 Spectrometer characteristics The NIST cNAA spectrometer has been constructed using two coaxial p-type high-purity germanium (HPGe) g-ray detectors with resistive-feedback preamplifiers The detectors have efficiencies of 63% and 31%, respectively, relative to NaI(Tl) at 1.33 MeV, and energy resolutions of 1.8 keV at the same g-ray energy A schematic view of the coincidence spectrometer is shown in Fig The HPGe detectors are oriented at 1801 in axial alignment with their endcaps facing each other; the endcap-to-endcap distance is continuously adjustable from approximately to 200 mm For counting, a radioactive sample is positioned on a phenolic spacer that sits directly on the endcap of the uplooking 31% detector Layered shielding sufficient to cover the entire lengths of both detector capsules has been incorporated Thicknesses of at least 50 mm of Pb and mm of Cu are present at all points in the shielding in order to attenuate environmental background contributions, as well as Pb X-rays generated inside the shielding The data-acquisition system for the NIST cNAA spectrometer utilizes all-digital electronics, based on the XIA LLC Pixie-41 module [9] The Pixie-4 is a four-channel digital pulse-processing module deployed in CompactPCI for Instrumentation (PXI) architecture The waveform of an input signal, taken directly from an HPGe preamplifier output, is continuously sampled and digitized by a flash ADC at a rate of 7:5  106 samples=s The signal pulse height is determined by a programmable, digital trapezoidal filter implemented in a field-programmable gate array (FPGA) Preamplifier pulse heights are determined to 16-bit resolution Event timing and pulse-pileup inspection is also carried out in the FPGA by a ‘‘fast’’ programmable Certain commercial equipment, instruments, or materials are identified in this paper in order to specify the experimental procedure adequately Such identification is not intended to imply recommendation or endorsement by the NIST, nor is it intended to imply that the materials or equipment identified are necessarily the best available for the purpose Pb Al Cu 30-L LN2 Dewar Fig Cross-sectional schematic of the NIST gg coincidence spectrometer Objects are not drawn to scale trapezoidal filter Events are time-stamped at the full ADC rate of 75 MHz In the present system the Pixie-4 resides in a 3U PXI crate, and a host desktop PC controls the pulseprocessing module and performs data readout via a PCIPXI fiber-optic bridge All operating parameters, including the filter values, are user-adjustable in software on the host PC; the coincidence time window is also set in software with a granularity of 13.33 ns and a lower limit of % 170 ns as presently employed Spectral data are recorded in list mode such that for each event the pulse height, time of detection (to 13.33 ns resolution) and a bit-mask indicating which detectors triggered is stored sequentially in on-board memory and written to the host PC hard disk in periodic blocks This method of data acquisition, as opposed to the ARTICLE IN PRESS B.E Tomlin et al / Nuclear Instruments and Methods in Physics Research A 589 (2008) 243–249 traditional multichannel-analyzer approach, preserves the timing information for each detected g-ray Pulse-height analysis is performed offline A schematic of the dataacquisition system is given in Fig 2a The advantages that digital signal processing offers to g-ray spectrometry have been described by a number of authors [10–13] Perhaps the most significant contribution that digital pulse processing makes is the ability to perform sophisticated pulse-shape analysis on the raw preamplifier signal, either in realtime or offline With regard to the present work, a pulse-shape analysis algorithm is employed in the Pixie-4 in order to execute ballistic deficit correction [11] which can be a serious problem in coaxial HPGe detectors [14], such as those employed in the NIST coincidence spectrometer Additionally, a digitally controlled signal-processing module does not suffer from drift of shaping and timing parameters that often plagues analog modules This is an especially attractive attribute for a device like the NIST cNAA spectrometer that will be used for long individual counts or many short counts over several weeks or months Finally, the high level of integration that is possible in a digital pulse-processing network allows for a device with a compact form factor The simplicity of the present data-acquisition system is especially apparent in contrast to previously reported coincidence counting systems that incorporate many NIM-based analog processing modules (cf Refs [5,8] and Fig 2b) 2.2 Data reduction The NIST coincidence spectrometer data-acquisition system offers the advantage of simultaneous acquisition of singles and coincidence data, allowing for potentially higher sample throughput Also, a dual analysis (conventional singles and coincidence) can be performed on each sample count, thus facilitating the comparison of these two methods under identical counting conditions Data reduction and analysis are accomplished using a combination of in-house software that utilizes the ROOT object-oriented data analysis framework [15], as well as commercial g-ray spectral analysis packages During offline analysis, list-mode events are sorted into one of three categories according to which detectors fired in a given event: detector ‘‘A’’ only (singles), detector ‘‘B’’ only (singles), or detectors ‘‘A’’ and ‘‘B’’ (coincidences) Each singles event is added to a 32k histogram based on the recorded pulse height; a coincidence count is added to a 32k  32k two-dimensional histogram (see Fig 3) according to the pulse height in each detector A one-dimensional coincidence g-ray spectrum is produced by setting a gate on a range of channels encompassing approximately 3s on each side of a given g-ray peak and projecting out the spectrum of all g-rays that were recorded in coincidence with the gross peak counts Thereafter, standard g-ray peak integration techniques are used to obtained net peak counts and background continuum counts fiber-optic/ PCI bus Pixie-4 Module HPGe HPGe preamp preamp Flash ADC Flash ADC SP FPGA 245 Memory Clock DSP (energy, time) host PC COM FPGA SP FPGA PCI I/O SP FPGA = signal processing field-programmable gate array COM FPGA = communication FPGA DSP = digital signal processor PCI = peripheral component interconnect ethernet shaper HPGe preamp gate fast amp CFD delay stop TSCA GDG ADC time TAC HPGe preamp fast amp CFD host PC start gate shaper CFD = constant-fraction discriminator GDG = gate and delay generator ADC energy ADC energy TAC = time-to-amplitude converter TSCA = timing single channel analyzer Fig (a) Schematic of the digital coincidence-counting system as implemented in the present work (b) Representative analog coincidence-counting dataacquisition system (adapted from Ref [5]) ARTICLE IN PRESS B.E Tomlin et al / Nuclear Instruments and Methods in Physics Research A 589 (2008) 243–249 107 Cs (T1/2 = 2.0652 a) 1969 801 103 1400 1167 563 106 569 795 604 604 102 801 keV 795 keV 10 Counts/0.4 keV 134 Counts/(0.4 Vek x 0.4 keV) Detector “B” Energy 246 105 104 1000 100 10 500 1000 1500 Energy [keV] 640 keV 536 keV 569 keV Fig Upper spectrum: singles Lower spectrum: a total projection (i.e., a gate on all energies) Detector “A” Energy Fig A representative portion of a two-dimensional gg coincidence spectrum for a sample containing 134Cs A partial decay scheme highlighting the relevant transitions is provided in the inset The vertical and horizontal lines trailing from each peak are due to random coincidences with Compton-scattered g-rays; the diagonal lines result from backscattered g-ray coincidences 2.3 Performance test A sample of desiccated, jet-milled bovine liver tissue, from part of a reference material certification program, was obtained for use in the evaluation of the coincidence spectrometer The sample, weighing 390.08 mg, was sealed in a high-purity quartz vial and irradiated in the RT-4 pneumatic irradiation channel of the NIST reactor at 3:4  1013 cmÀ2 sÀ1 neutron flux The short-lived and intermediate activities were allowed to decay away over the course of 80 d before counting was commenced The quartz sample vial was placed horizontally in the spectrometer chamber on top of a cm phenolic spacer, such that the vial was approximately centered vertically between the two detector faces with the sample powder evenly distributed along the length of the vial The sample was counted for 109 h at average input count rates of 1960 and 1660 Hz in the lower and upper detectors, respectively, with an average fractional deadtime of 1.5% Discussion A representative singles g-ray spectrum recorded for the irradiated bovine liver sample by the upper detector is shown in Fig Also shown in the figure is the corresponding total projection (i.e., gated on all energies) of coincident g-rays recorded by the upper detector One feature of the coincidence spectrum that is immediately noticeable is the presence of distinctive Compton scattering ‘‘peaks’’ at around 200 and 800 keV These structures are a result of the geometry of the detectors in the spectrometer The face-to-face alignment enhances the probability of recording coincident backscattered g-rays, such that when the intense g-rays above 1000 keV are Compton scattered through % 180 they deposit % 800 keV in one detector and the remainder of their energy (% 200 keV) in the other detector in a true coincidence event The 180 orientation was chosen to permit easy adjustments of the source– detector distances which is a valuable feature when dealing with samples of widely varying activity This orientation also provides a large geometric efficiency when the endcapto-endcap distance is made small A known drawback to this configuration is this detector cross-talk The present configuration does seem to yield a higher continuum for some regions of the spectrum, especially for peaks in the 800–900 keV range Future improvements to the setup may include a 90 orientation with shielding between the detector capsules, or perhaps a Pb sample collimator placed between the detector endcaps In the portion of the coincidence spectrum below 1000 keV, the emergence of several peaks above the Compton continuum demonstrates qualitatively the background reduction which can be obtained even with an unrestricted coincidence requirement It is worth noting that a number of these peaks like the 110m Ag 657 keV line are barely discernible above the level of the continuum in the singles spectrum The 75Se 400 keV g-ray comes from a noncoincident transition and is, therefore, effectively suppressed by the coincidence requirement However, in the case of the 65Zn 1115 keV peak, which is also noncoincident, the suppression is not perfect due to the presence of random coincidences between the abundant 65 Zn decays and all other decay events The coincidence spectrum shown in Fig represents the broadest coincidence requirement, that is any energy in coincidence with any other energy This picture was presented to illustrate some of the general features of the coincidence spectrometer In practice a much higher degree of discrimination is obtained by setting a restrictive energy gate Fig shows a portion of the same singles spectrum given in Fig 4, as well as the coincidence spectrum gated on a narrow range of energies encompassing the 657 keV ARTICLE IN PRESS B.E Tomlin et al / Nuclear Instruments and Methods in Physics Research A 589 (2008) 243–249 peak Three coincident 110m Ag peaks appear along with 134 Cs and 60Co peaks The 134Cs and 60Co g-rays are present as Compton coincidences The gated gross 657 keV peak includes Compton contributions from 1173, 1332, and 1365 keV (134Cs) g-rays, so any g-ray which would occur in coincidence with a full-energy event will also occur in coincidence with these partial-energy scattered events The characteristic of the gated spectrum which should be highlighted is the high-degree of background reduction that has been achieved In fact, it is the thorough suppression of the background continuum that even makes it possible to discern the 884 and 937 keV peaks, which cannot be seen at all in the singles spectrum The example of the 657 keV gate demonstrates qualitatively that the NIST coincidence spectrometer is capable of the significant background suppression that is beneficial in analyzing trace elements, such as Ag in the bovine liver sample 107 Counts/0.4 keV 106 105 104 1000 100 10 800 600 1000 Energy [keV] 1200 1400 Fig Upper spectrum: singles Lower spectrum: a projection gated on the 110m Ag 657 keV line 247 As an adjunct to NAA, coincidence counting is of value if it satisfies any of the following conditions: (i) if it can provide an analytical result with greater precision than conventional singles counting; (ii) if it can resolve the direct interference of two overlapping g-ray peaks; or (iii) if it can reveal the presence of an analyte which is below some minimum detectable activity limit in conventional NAA In the present work, the performance of the NIST coincidence spectrometer was evaluated based on the first criterion, since this arguably may have the widest impact Future analyses will investigate the capabilities with regard to the latter two criteria In instrumental NAA the predominant contribution of uncertainty to a determined concentration is usually the uncertainty in the g-ray peak counting statistics [16], especially for very low analyte concentrations; therefore, the performance of the NIST cNAA spectrometer was evaluated on whether it could produce g-ray peak areas with smaller relative standard deviations than the corresponding singles g-ray peaks Other means of comparing coincidence and singles spectra, such as peak-to-background ratios, which may be interpreted as a form of ‘‘signal-to-noise ratio’’ for g-ray spectra, have been presented by other authors [17]; however, while such measures often reveal significant enhancement of the specified ratio, they not necessarily translate to any improvement in the analytical result This point is illustrated by the values in Table One can see that across the board improvements in the peak-to-background ratios (A=b) are obtained in the coincidence-gated spectra relative to the singles spectra Examination of the relative standard deviations, however, reveals that in only two cases, i.e., Ag and Sc, is any enhancement obtained by the coincidence method Table Counting statistics and calculated parameters for relevant coincidence peaks in bovine liver sample Line (keV) Singles A 657 (110m Ag) 884 (110m Ag) Coincidence B b A=b 3436 4277 238 475 464 862 26 497.2 21 130.1 1173 (60Co) 1332 (60Co) 870 129 784 729 22 597 15 710 836.9 654.6 604 (134Cs) 795 (134Cs) 84 757 59 103 317 184 367 007 26 432.0 33 364.3 413 361 370 003 032 122 489 549 247 271 206 599 907 182 712 424 192 (59Fe) 1099 (59Fe) 74 706 537 278 889 (46Sc) 1120 (46Sc) 8351 5243 121 136 264 279 (75Se) (75Se) (75Se) (75Se) 0.13 0.20 1039.7 1198.8 RSTD (%) B=A 25.7 29.9 69.4 109 A B b A=b RSTD (%) 84 102 10 0.71 0.23 117.6 442.0 13.7 10.6 22 705 22 534 200 252 8.70 10.08 2611.1 2235.5 0.7 0.7 0.1 0.1 0.026 0.020 3.21 1.77 1.4 2.0 3.74 6.21 2999 3024 16 55 1.00 3.44 2999.0 879.7 1.9 1.9 249 696.8 220 659.9 75 598.5 64 765.8 1.66 6.21 13.7 7.56 0.7 0.2 0.2 0.4 5.44 1.61 0.88 1.46 34 531 139 536 141 726 35 655 1746 5427 3093 953 134.31 387.64 193.31 59.56 257.1 360.0 733.1 598.6 0.6 0.3 0.3 0.6 890 447 140 343 127 206.7 8255.5 0.59 65.1 2.1 0.2 11.9 0.26 2388 2696 246 75 35.14 3.75 68.0 718.9 2.3 2.1 558 367 51 568 50 760.6 5729.8 0.16 0.92 22.2 8.1 66.9 9.84 176 207 0.69 0.23 254.2 897.0 8.4 7.2 Note: A ¼ net peak counts; B ¼ integral continuum counts; b ¼ avg continuum counts; and RSTD ¼ relative std dev of net peak counts ARTICLE IN PRESS 248 B.E Tomlin et al / Nuclear Instruments and Methods in Physics Research A 589 (2008) 243–249 The failure of an increased peak-to-background ratio to correlate with an improved peak area uncertainty is explained by the significantly reduced peak counts in the coincidence spectrum As the net area of a given peak is reduced, eventually a point is reached where the uncertainty due to the net counts is more significant than the uncertainty contribution due to the continuum counts Since coincidence counting efficiency is roughly equal to the product of the efficiencies of each individual detector, then it stands to reason that a coincidence spectrum will always contain fewer counts than the corresponding singles spectrum Thus, the application of coincidence counting to NAA is only of value in situations where spectral background can be reduced without too great a loss of net peak counts In other words, coincidence detection efficiency is as important as background reduction In the examples of Table 1, Sc and Ag started out with rather low peak-to-background ratios and the background reduction was more than enough to compensate for the smaller coincidence peaks; the Co peaks, on the other hand, started with much larger peak-to-background ratios, so even if the background in the coincidence spectrum were reduced to zero, the loss of peak counts would increase the relative peak area uncertainties One of the goals in evaluating the behavior of the cNAA spectrometer was to establish, at least empirically, a means of deciding when it is of value to use coincidence gating Since peak-to-background increases not necessarily correlate with improved analytical results, an alternative parameter was sought to predict the effectiveness of coincidence counting for analytical purposes Using the relation given in Eq (1) as the basis of a quantitative benchmark, the expression in Eq (2) was derived: ðsA =AÞcoincidence o1 ðsA =AÞsingles (1) where sA is the standard deviation of the net peak counts A, and Bs Ac þ Á 41 (2) 2As As where As and Ac are net peak counts for singles and coincidence spectra, respectively, and Bs is the integrated continuum counts under the given singles peak Eq (2) illustrates the contributions of both background reduction (in the B=A term) and coincidence efficiency (in the Ac =As term) towards enhanced precision, in agreement with the general conclusion of Galloway [18] For the detectors and source geometry presently employed, the quantity Ac =As has been empirically determined to fall within the range of 0.03–0.05 Thus, it may be conjectured that, in the singles spectrum, a background-to-peak ratio (Bs =As ) of % 40 or more will yield a coincidence-gated peak with a lower relative standard deviation Inspection of the B=A values given in Table reveals that this correlation is generally borne out—small improvements in the relative standard deviations are observed for B=A around 10 and significant improvements when B=A is above 60 The ratio B=A provides a convenient means for evaluating the utility of coincidence gating, since a singles spectrum is acquired simultaneously with a coincidence spectrum and can be quickly analyzed before further analysis Based on the analysis of the bovine liver material, it may be concluded that in certain cases, as demonstrated by Ag and Sc, quantitative gg coincidence spectrometry as employed in the NIST coincidence spectrometer can be an effective means of obtaining greater precision in NAA elemental determinations While greater precision is not the only reason to apply coincidence counting to NAA, it does potentially have a bearing on the greatest number of elements For the NIST spectrometer, the primary condition which indicates that a given peak could benefit from coincidence counting is the presence of a very high background continuum as indicated by background-topeak ratios determined from the corresponding singles spectrum In activated materials, primary contributors to the g-ray spectrum continuum include Bremsstrahlung and Compton scattering Sample matrices, such as biological materials, which are rich in P and Zn, tend to display high background levels due to relatively long-lived 32P Bremsstrahlung and 65Zn Compton events; therefore, cNAA may prove useful in the analysis of such materials Conclusions A gg coincidence spectrometer has been constructed at NIST, using two HPGe g-ray detectors and an all-digital data-acquisition system, for the purpose of exploring coincidence NAA and its value in characterizing reference materials The present work demonstrates the successful substitution of advanced digital signal processing for conventional analog electronics in quantitative gg coincidence spectrometry An initial evaluation of quantitative coincidence counting in comparison to singles spectrometry, based upon a sample of neutron-irradiated bovine liver material, corroborated the feasibility of this approach It has been determined that in this case only those peaks in the singles spectrum that have very large background-topeak ratios display any enhancement in peak area uncertainties by means of coincidence gating A simple empirical criterion for identifying these g-ray peaks was deduced The value of coincidence counting in resolving direct g-ray peak interferences or in determining elements which fall below minimum detectable activity levels in conventional spectrometry will be considered in concert with future work on the characterization of reference materials by gg coincidence NAA Disclaimer Contribution of the National Institute of Standards and Technology, not subject to copyright in the United States ARTICLE IN PRESS B.E Tomlin et al / Nuclear Instruments and Methods in Physics Research A 589 (2008) 243–249 References [1] [2] [3] [4] [5] [6] [7] J.I Kim, A Speecke, J Hoste, Anal Chim Acta 33 (1965) 123 E.T Bramlitt, Anal Chem 38 (1966) 1669 W.D Ehmann, D.M McKown, Anal Lett (1969) 49 M Vobecky´, et al., Anal Chim Acta 386 (1999) 181 J Jaku˚bek, et al., Nucl Instr and Meth A 414 (1998) 261 Y Hatsukawa, et al., J Radioanal Nucl Chem 272 (2007) 273 H Huber, Ch Koeberl, I McDonald, W.U Reimold, J Radioanal Nucl Chem 244 (2000) 603 [8] J.A Cooper, Anal Chem 43 (1971) 838 [9] W Hennig, et al., Nucl Instr and Meth B 263 (2007) 175 [10] W.K Warburton, M Momayezi, B Hubbard-Nelson, W Skulski, Appl Radiat Isot 53 (2000) 913 249 [11] W Skulski, M Momayezi, B Hubbard-Nelson, P Grudberg, J Harris, W Warburton, Acta Phys Pol B 31 (2000) 47 [12] M Bolic´, V Drndarevic´, Nucl Instr and Meth A 482 (2002) 761 [13] B Hubbard-Nelson, M Momayezi, W.K Warburton, Nucl Instr and Meth A 422 (1999) 411 [14] G Knoll, Radiation Detection and Measurement, third ed., Wiley, New York, 2000 [15] R Brun, F Rademakers, Nucl Instr and Meth A 389 (1997) 81 [16] E.G Moreira, M.B.A Vasconcellos, M Saiki, J Radioanal Nucl Chem 269 (2006) 377 [17] P.P Ember, T Belgya, G.L Molna´r, Appl Radiat Isot 56 (2002) 535 [18] B.B Galloway, Nucl Instr and Meth 55 (1967) 29