Phương pháp trùng hợp cc (phát hiện hai ctia phát ra gần như đồng thời) đã được sử dụngtrong nghiên cứu cấu trúc hạt nhân 1, và cũng là một điểmsible cách để cải thiện độ nhạy cảm của dấu nhắckỹ thuật phân tích kích hoạt gamma (PGAA) 2.Là một phần của cơ sở cho các neutron gây raquang phổ tia gamma (NIPS) tạiCổng chùm PGAA của Budapest Reac nghiên cứutor một cc coincidence bộ máy đã được phát triển.Trong các bài viết trước của chúng tôi, đồng xu cc khu vựcphương pháp cidence được giới thiệu 2, và meathiết lập bảo đảm đã được mô tả chi tiết 3. Trongbài báo hiện tại chúng tôi trình bày một phép đo clo để minh họa chất lượng của hệ thống, và một phép đo của một loạt các mẫu thủy tinh như làsử dụng thực tế của nó.
Trang 1A practical test of a c–c coincidence measurement setup for PGAA
P.P Ember *, T Belgya, J.L Weil, G.L Moln a ar
Institute of Isotope and Surface Chemistry, Chemical Research Centre of the Hungarian Academy of Sciences,
P.O Box 77, H-1525 Budapest, Hungary
Abstract
A second cold-neutron beam experimental station has been built as part of the renewal of our prompt gamma activation analysis facility at the Budapest Research Reactor This new station has been instrumented for neutron-induced prompt gamma-ray spectroscopy, involving c–c coincidence measurements The c–c coincidence arrangement
is introduced briefly, and its usage is illustrated with a practical example of the inactive tracer method for samples from glass furnace
2003 Elsevier B.V All rights reserved
PACS: 25.40.Lw; 81.70.Jb; 82.80.Ej; 82.80.Jp
Keywords: ðn; cÞ; E n ¼ cold; Prompt gamma activation analysis; Coincidence; Spectral interference
1 Introduction
The c–c coincidence method (detection of two c
rays emitted almost simultaneously) has been used
in nuclear structure studies [1], and is also a
pos-sible way to improve the sensitivity of the prompt
gamma activation analysis (PGAA) technique [2]
As part of the facility for neutron-induced
prompt gamma-ray spectroscopy (NIPS) at the
PGAA beam port of the Budapest Research
Reac-tor a c–c coincidence apparatus has been developed
In our previous articles the regional c–c
coin-cidence method was introduced [2], and the
mea-surement setup was described in detail [3] In our
present article we show a chlorine measurement to
illustrate the quality of the system, and a mea-surements of a series of glass samples as the first practical usage of it
2 Experimental work All measurements were made using the same arrangement, as detailed in our most recent article [3]
2.1 Detector parameters and positioning Two coaxial HPGe detectors were used for the measurements described here Detector HPGe-I had 30% efficiency relative to a 300· 300NaI detector and 1.8 keV resolution at 1332 keV c-ray energy Detector HPGe-II had 15% efficiency and 1.9 keV resolution HPGe-II was placed horizontally at a
*
Corresponding author Fax: +36-1-392-2584/2222.
E-mail address: ember@alpha0.iki.kfki.hu (P.P Ember).
0168-583X/$ - see front matter 2003 Elsevier B.V All rights reserved.
doi:10.1016/S0168-583X(03)01663-X
Nuclear Instruments and Methods in Physics Research B 213 (2004) 406–409
www.elsevier.com/locate/nimb
Trang 2distance of 4.3 cm from the target, and HPGe-I
vertically, about 2.6 cm from the target Lead
blocks of 5 cm thickness were used around the
crystals as gamma shields, and plates of 1.5 mm
thick lead were placed between the detectors and
the tube to decrease the number of back-scattered
gamma photons, and to filter out the X-rays
2.2 Test samples
The chlorine sample was a PVC foil of 0.11 g
mass and 1.8· 2.5 cm2 surface area It was
irra-diated in the neutron beam for 23 h to excite the
36Cl nuclei via the 35Cl(n,c) reaction, which has a
large cross section The de-exciting c-rays have
been measured in coincidence mode
The glass samples were borrowed from a series
of industrial measurements carried out with the
Compton-suppressed spectrometer at our PGAA
facility [4,5] Recently a series of experiments were
performed to measure the uniformity of mixing of
input ingredients in a glass furnace with an
inac-tive tracer method This furnace works
continu-ously, and the tracer was injected in an instant
with the ingredients, and the samples were taken at
certain times from the outcoming melted glass
The process is shown in Fig 1 Important aspects
of the tracer selection were: it should not affect the
technology, nor change the parameters of the
produced glass Furthermore it has to have a large
neutron capture cross section Boron was selected
for normal glasses, and gadolinium for the
bo-rosilicate glasses containing approximately 5% of
boron Our samples were taken from a
measure-ment series of borosilicate glasses The original
samples were of several tens of grams, thus we cut
from them about a half gram for our purpose
Table 1 contains the data of the four glass samples: the number of the sample for reference, the mass of the borated glass sample (used for the coincidence measurement), the measuring time (equals with the irradiating time), and the Gd concentration with its uncertainty The concen-tration data were calculated from the Compton-suppressed measurement Sample 0 was taken just before the gadolinium tracer was put into the furnace, so this measurement represents the natu-ral gadolinium concentration of the glass
3 Results and discussion 3.1 Chlorine sample
48 million coincidence events were collected in the chlorine experiment The upper half of Fig 2 shows the total projected time spectrum, which means the histogram of all time values from the list
Fig 1 Draft of the inactive tracing of a glass furnace.
Table 1 Data of the glass samples Sample number 0 1 2 4 Sample mass (g) 0.529 0.536 0.449 0.48 Measuring time (min) 1052 1179 1220 325
Gd concentration (ppm) 0.87 1.4 11.9 21.9 Uncertainty of Gd
con-centration (ppm)
0.1392 0.106 1.547 1.533
Gd concentration was measured by the Compton-suppressed PGAA spectrometer of the NIPS facility.
Fig 2 Time spectrum of 60Co c–c coincidence events Gates were set on the 1173 and 1332 keV c peaks in the HPGe-I and HPGe-II energy spectra respectively.
P.P Ember et al / Nucl Instr and Meth in Phys Res B 213 (2004) 406–409 407
Trang 3file The time spectrum has a peak to background
ratio of 150, and FWHMof 30 ns The structure of
any time spectrum depends on many factors, such
as the detector dimensions, the threshold setting in
the CFD electronics, and the variety of transition
energies and life times of levels involved in the
de-excitation process of the target nuclei In the case
of a broad energy range of transitions, one cannot
obtain as good time resolution as can be obtained
with e.g a 60Co-calibration source [3] and narrow
energy gates, due to the time-walk of the detector
output signals [6] In addition, finite lifetimes cause
centroid shifts or exponential tails that also
broaden the time peak The lower half of Fig 2
illustrates the time walk by showing the time peak
of three full energy peak-to-peak coincidences In
detector HPGe-I a gate on the 517 keV energy
peak was set for all three cases, while in HPGe-II
gates on the 6110, 1951 and 786 keV peaks were
applied respectively The effect of the time walk is
clearly visible; the stop signals of higher energy
(higher amplitude) pulses come earlier, while the
lifetimes of the intermediate levels are negligible
The measured FWHMvalues of the three time
peaks are 13 ns, which is the same as was found for
the60Co time peak [3]
Fig 3 presents the total projection of the
HPGe-I detector energy signals from the35Cl(n; c)
capture reaction As can be seen, a few very strong
lines dominate the spectrum, including the
strongest primary transition of 6110 keV energy
For demonstration of the off-line coincidence
analysis method, we set a gate on the 5086 keV double escape peak, instead of the 6110 keV full energy peak, because the latter and its single es-cape peak form doublets with the single- and double-escape peaks of a weaker, higher energy c ray To estimate the background, we set back-ground regions on both sides of the time and en-ergy peaks The peak and background gate settings are shown as gridded and crosshatched areas, re-spectively, in Fig 3 The two background regions combined cover the same number of channels as the peak gates The analysis program subtracts the projection of the backgrounds automatically from the coincidence projection spectra and also calcu-lates the uncertainties in the gated spectrum
In Fig 4 the spectrum of the c-rays measured
by the HPGe-I detector with the gate settings ex-plained above is shown The strongest coincidence peaks are labeled with their energies As can be seen, the signal to noise ratio is very good The random coincidences have almost completely dis-appeared, as illustrated by the magnified insert of the energy-gate setting area Fig 5 shows the corresponding decay scheme (taken from ENSDF [7]) for the 35Cl(n; c)36Cl reaction, retaining only those transitions which are expected to be in co-incidence with the strongest 6110 keV primary capture c-ray For easier reading the c-ray energies and their absolute intensities in percentage are rounded off The c transitions with intensity less than 1% are drawn with dotted lines
Fig 3 Total projection energy spectrum of the HPGe-I
de-tector from the chlorine experiment.
Fig 4 Spectrum of c-rays in coincidence with 5086 keV, the double escape peak for 6110 keV.
408 P.P Ember et al / Nucl Instr and Meth in Phys Res B 213 (2004) 406–409
Trang 4In agreement with our expectations, the gated
energy spectrum of HPGe-II contains only the
peaks expected from the known level scheme of
36Cl The weak 2467 keV line can be easily
iden-tified, but the 511 keV annihilation peak masks the
similarly weak 508 keV line
3.2 Glass samples
The above outlined c–c coincidence technique
was applied for the analysis of the glass samples
The peak areas were normalized for each sample
to 1 g sample mass and 1000 min of measurement
time
Fig 6 shows the calibration curve for the
co-incidence measurement: the calculated and
nor-malized peak areas for the 182 keV gadolinium
peak are plotted versus the gadolinium
concen-tration calculated from the Compton-suppressed measurements Each point has the uncertainty marked in both directions The graph also shows the fitted linear function that was forced to cross the origin The r2value of the fit is also given on the graph As can be seen in Fig 6 the calibration curve is linear, which means the method is appli-cable for concentration determination
4 Conclusions
In this article we have studied the applicability
of c–c coincidence for the determination of con-centrations with PGAA method We found that the coincidence reduces spectral background and interference, but preserves the linear relationship between signal and concentration A possible way
of further improvement is the use of a digital signal analyzer, which would allow larger count rates, and hence better precision for a given measuring time
References
[1] A.H Wapstra, Alpha and Gamma-ray Spectroscopy, Vol I, fifth printing, North Holland Publishing Company, Am-sterdam New York Oxford, 1979, Chapter VIII/C, p 539 [2] P.P Ember, T Belgya, G.L Moln a ar, Appl Radiat Isot 56 (2002) 535.
[3] P.P Ember, T Belgya, J.L Weil, G.L Moln a ar, Appl Radiat Isot 57 (2002) 573.
[4] T Belgya, Zs R eevay, B Fazekas, I H eejja, L Dabolczi, G.
L Moln a ar, Z Kis, J € O Ost€ o or, Gy Kasz a as, in: Proceedings of the Ninth International Symposium on Capture Gamma-Ray Spectroscopy and Related Topics, Vol 2, Budapest, Hungary, 1996, p 826.
[5] G Moln a ar, T Belgya, L Dabolczi, B Fazekas, Zs R eevay,
A
A Veres, I Bikit, Z Kiss, J € O Ost€ o or, J Radioanal Nucl Chem 215 (1997) 111.
[6] G.F Knoll, Radiation Detection and Measurement, third ed., John Wiley & Sons, New York, 2000, Chapter 17/IX,
p 438.
[7] Evaluated Nuclear Structure Data File (ENSDF) produced
by members of the International Nuclear Structure and Decay Data Network, and maintained by the National Nuclear Data Center, BNL, USA Also available online from IAEA Nuclear Data Section Vienna.
Fig 5 Partial decay scheme of 36 Cl Only the lines in
coinci-dence with the 6110 keV peak are shown.
Fig 6 Calibration curve for the 182 keV chlorine peak.
P.P Ember et al / Nucl Instr and Meth in Phys Res B 213 (2004) 406–409 409