Documente Academic
Documente Profesional
Documente Cultură
RADIATION DETECTORS
AND THEIR USES
Proceedings of the 26 th Workshop
Edited by
S. Sasaki, Y. Kishimoto, M. Hagiwara, T. Sanami
K. Saito, K. Iijima and H. Tawara
High Energy Accelerator Research Organization
FOREWORD
The 26th workshop on "Radiation Detectors and Their Uses" was held on
January 24, 25 and 26, 2012 at the High Energy Accelerator Research Organization
(KEK) in Tsukuba, Ibaraki, Japan. The workshop was hosted by the Radiation Science
Center, KEK under the cooperation of the Society of Radiation Science, the affiliate of
Japan Society of Applied Physics. The workshop offers an outstanding opportunity for
scientists interested in the fields of radiation physics, radiation detector, radiation
measurement, nuclear science, high energy physics and their application to meet and
discuss with colleagues from all over the country.
The number of the participants who registered to the workshop was 102. 48
presentations were given at the workshop. This workshop was the first one after the
accident of Tokai Daiichi Nuclear Plants and there were many presentations related to
measurements of radio-activities released from the nuclear plants as well as general
themes. As the fruits of the workshop, this report was published as the proceedings of
"the 26th Workshop on Radiation Detectors and Their Uses". All papers submitted for
publication in the proceedings received the peer review process by the independent
reviewers. Finally, the 18 original papers were published in the proceedings after the
review process.
The editors would like to express our great appreciation to the authors who
prepared the manuscripts in good earnest and the reviewers who spent their precious
time to check the papers.
November, 2012
Shinichi Sasaki
Workshop Program Chair
High energy Accelerator Research Organization (KEK)
TABLE OF CONTENTS
109
S. Satoh, K. Hirota
ii
1. INTRODUCTION
Management of radiation exposure is very important for astronauts, because the radiation dose
expected in spacecraft is higher by two orders of magnitude than that on the ground. In spacecraft, the
main component contributing to the radiation dose is charged particles and neutrons generated by nuclear
reactions between charged particles and materials composed of spacecraft (1) (2). The radiation effects on
human body in spacecraft are evaluated using dose equivalent H described by the following equation:
H D Q (L) ,
where D is the absorbed dose; L is the Linear Energy Transfer (LET); Q is the quality factor given as a
function of L. The accuracy of H is important for the dosimeters used in spacecraft and the National
Council of Radiological Protection Report No. 142 (NCRP-142) recommends the measurement accuracy
of 30% or better (3).
The TEPC, DOSTEL, RRMD and BBND are known as active dosimeters developed to measure
radiation dose in spacecraft
accuracy for both of charged particles and neutrons. Thus, it is required for spacecraft to equip two or more
kinds of dosimeters for complete measurement of radiation dose. This requirement possibly becomes great
disadvantage because smaller and lighter instruments are more important for on-board equipment of
spacecraft. Therefore, we have been developing a dosimeter called Position-Sensitive Tissue Equivalent
Proportional Chamber (PS-TEPC) to realize a dosimeter which has enough accuracy of measurement
and sensitivity for the both of charged particles and neutrons.
2. INSTRUMENT
The PS-TEPC is a position sensitive gaseous proportional chamber utilizing a -PIC. The schematic
view of the detector is shown as Figure 1. The PS-TEPC consists of a drift plane, shaping frames and a
-PIC. The drift plane and shaping frames form an electric field to drift the electrons generated by
interactions between the gas and incident particles toward the -PIC. The -PIC acts as an electron
multiplier and the schematic view is shown in Figure 2 (8). We use a compact size -PIC with 64 strips
64 strips developed for the PS-TEPC. Owing to the anode strips and the cathode strips, which are
perpendicular to the anode strip, of -PIC, two-dimensional position can be acquired. Meanwhile, the
position along to the drift direction can be obtained by measuring the duration for which primary electrons
reach the -PIC. Thus, the PS-TEPC has an ability to reconstruct three-dimensional tracks and can acquire
the path length of incident particles. Because each pixel of -PIC functions like a proportional counter, the
energy deposit due to incident particles also can be obtained from the amount of collected charge.
Therefore, LET of incident particle can be acquired by one event. The shaping frames are made of tissue
equivalent plastic A-150. Measuring the protons generated from interactions between neutrons and the
plastic, the dose equivalent for neutrons can be estimated.
The basic demonstration of the measurement principle has been done using the prototype version of
PS-TEPC filled with P10 gas (9). Meanwhile, we intend to apply Tissue-Equivalent gas (TE gas) for the
PS-TEPC finally. Therefore, we are now investigating the specification of the PS-TEPC filled with
propane-based TE gas.
3. MEASUREMENT
The particles with the LET from 0.1 to 500 keV/m-water mainly contribute dose equivalent in the
spacecraft (10). Thus, this LET range is the target range of PS-TEPC to measure. At the lowest LET of the
target range, the signal from the detector is faint because the number of primary electrons is small.
Therefore, we performed a beam experiment to irradiate the PS-TEPC with a beam of proton with the
energy of 230 MeV. Its LET corresponds to 0.2 keV/m-water and the beam is suitable to demonstrate the
ability to detect the charged particles with approximately lowest LET in our target range. The experiment
was performed at the Heavy Ion Medical Accelerators in Chiba (HIMAC) and we used the PS-TEPC
filled with propane-based TE gas (C3H8:54.8%, CO2:39.7%, N2:5.5%) at 1atm in this experiment. The
effective size of the -PIC used in this experiment was 2.6 cm 2.6 cm and the gap between the drift
plane and the -PIC was 5.0 cm. A plastic scintillator with 0.2 cm thickness and another one with 1 cm
thickness were placed backward and forward to the PS-TEPC, respectively. The distance from the beam
window to the front scintillator, that from the front scintillator to the detector and that from the detector to
the backward scintillator were about 0.2, 1 and 1.5 m, respectively. The atmosphere was air at 1 atm and
the beam window was 0.1 mm thickness of Al. The data were acquired by the coincidence trigger signal
of the plastic scintillators. The applied voltage to the -PIC and the drift plane were 1010 V and -3300 V,
respectively. In this case, the drift velocity corresponds to 0.77 cm/sec. The diameter of the beam was
about 1 cm and the incident direction was parallel to the cathode strips. This direction seems to minimize
the amount of the charge collected by an anode strip and it is the most difficult situation for detection of the
proton signal. The signal of each strip was summed with every 4 strips and amplified by charge sensitive
pre-amplifier CP3578. Then the 32 output signals (16 anode channels and 16 cathode channels) from it
were fed to the multi-channel digitizer CAEN V1720 and the digitized data were read out by a PC. The
beam irradiation was performed for about 4 minutes and the data of 3000 events were obtained.
Secondary, we reconstructed the track image using the same data and the result is shown in Figure 6.
The tracks of 20 events are superimposed in this figure. Here, the points show the track points calculated
from the drift duration and the position of the hit strips, which the pulse heights of them are higher than the
threshold. The arrow shows the incident position and direction of the beam determined by the geometry of
the experimental setup. The directions of reconstructed tracks align along the beam direction and the
positions of them well correspond to the incident position. Then, we analyzed the length of each track and
investigated its distribution. The result is shown as Figure 7. A very sharp peak around about 2.5 cm is
shown in this histogram and it is consistent with the horizontal size of the fiducial volume determined by
the size of the -PIC.
Finally, we analyzed the amount of collected charge to evaluate the energy resolution. We
converted the pulse height of each anode strip into collected charge using the calibration data obtained
from pre-amplifier calibration and then calculated the total charge by summing them. Figure 8 shows the
distribution of the total charge. We fitted the distribution with a Gaussian and the solid line shows the
best-fit curve. From the best-fit parameters, we obtained the charge resolution of 25.9% (1). Because
amount of charge generated in the gas is proportional to the energy deposition, the resolution of charge can
be considered as energy resolution. Considering energy resolution of proportional counter becomes better
for higher energy deposition and the LET of the irradiating beam approximately corresponds to the lowest
LET in our target range, it is suspected that the PS-TEPC has better energy resolution than that of 30% in
overall target range.
5. SUMMARY
We have been developing a position-sensitive tissue equivalent proportional chamber PS-TEPC as
a dosimeter in spacecraft. We performed a beam experiment at HIMAC. Irradiating the PS-TEPC with
proton beam with the energy of 230 MeV, it is recognized that the signal due to the incident proton has
enough pulse height to be distinguished from that of noise. From the position information obtained by
analyzing the same data, the track image which is consistent with the beam profile was obtained.
Additionally, the distribution of the track length shows a sharp peak corresponding to the distance limited
by the size of -PIC. From the distribution of the collected charge, the energy resolution was estimated to
be 25.9% (1).
It is reported that collected charge in filling gas distorts due to the presence of impurities (11). The
same effect might be occurred for our detector when the beam experiment performed. In this case, better
energy resolution may be obtained if purity of the TE gas in our detector improves. Therefore, we plan to
optimize our vacuum system and baking condition to improve the performance of our detector.
ACKNOWLEDGMENT
This study was carried out as a part of the Research Project with Heavy Ions at NIRS-HIMAC.
REFERENCES
(1) J. E. Keith, et al., Nucl. Tracks Radiat. Meas., 20 (1992) 41.
(2) V. E. Dudkin, et al., Nucl. Tracks Radiat. Meas., 17 (1990) 87.
(3) NCRP Report No.142; National Council on Radiation Protection and Measurements (2002) 56.
(4) P. Kliauga, et al., Raidat. Protec. Dosim., 61 (1995) 309.
(5) R. Beaujean, et al., Radiat. Protec. Dosim., 85 (1999) 223.
(6) T. Doke, et al., Radiat. Meas., 33 (2001) 373.
(7) H. Matsumoto, et al., Radiat. Meas., 33 (2001) 321.
(8) H. Kubo, et al., Nucl. Instr. and Meth., A513 (2003) 94.
(9) T. Nagayoshi, et al., Nucl. Instr. and Meth., A581 (2007) 110.
(10) T. Sakaguchi, et al., Nucl. Instr. and Meth., A437 (1999) 75.
(11) T. Kohno, et al., Nucl. Instr. and Meth., A450 (2000) 456.
Measurement of airborne
airborne radioactivity
radioactivity from
from the
the Fukushima
Fukushimareactor
reactor
accident in Tokushima, Japan
K.Fushimi, S.Nakayama, M.Sakama2 and Y.Sakaguchi3
Institute of Socio Arts and Sciences , The University of Tokushima, 1-1 Minami
Josanjimacho Tokushima city, 770-8502 Tokushima, JAPAN
2 Department of Radiological Science, Division of Biomedical Sciences, Institute of
Health Biosciences, The University of Tokushima, 3-18-15 Kuramotocho Tokushima city,
770-8509 Tokushima, JAPAN
3 Faculty of Integrated Arts and Sciences, The University of Tokushima, 1-1 Minami
Josanjimacho Tokushima city, 770-8502 Tokushima, JAPAN
Abstract
The airborne radioactive isotopes from the Fukushima Daiichi nuclear plant was
measured in Tokushima, western Japan. The continuous monitoring has been carried
out in Tokushima. From March 23, 2011 the ssion product 131 I was observed. The
radioisotopes 134 Cs and 137 Cs were also observed in the beginning of April. However
the densities were extremely smaller than the Japanese regulation of radioisotopes.
Introduction
Serious damage to the Fukushima Daiichi nuclear power plant (141 27E, 37 45N) has
been caused by huge tsunami followed by the huge earthquake on 11 March 2011. The
plants 1,2 and 3 were operating and the plant 4 was stopped before the earthquake. The
plants made emergency stop just after the earthquake, however, all the power plants in
Fukushima Daiichi were seriously damaged by the following big tsunami. All the electric
power got fault and the cooling system was collapsed. From 12 March 2012, a large
amount of radioactive materials was vented to avoid more serious damages. Total amount
of vented radioactive isotopes were estimated as 1.5 1017 Bq for 131 I and 1.2 1016 Bq for
137 Cs [1].
The largest ejection of radioactivity from the plants occurred on 15th March and the
amount of ejected radioactivity decreased after 17 March[1]. During 15 and 16 March,
the wind direction changed from north and south, the wind direction raised the serious
pollution in Iidate village and north Kanto district. After 16 March, the wind direction
changed to west and continued for a few days. The westly wind brought the radioactive
isotopes to the Northern Hemisphere.
In the present paper, the measurement of airborne radioactive isotopes in Tokushima
which is placed in western district of Japan is reported. The arrival date of radioactivity in
the world was analyzed to investigate the behavior of plume exhausted from the reactor.
The precise information for the detection eciency of gamma ray was determined to
analyze the radioactive isotopes. The detection eciency of gamma rays were precisely
estimated by Monte Carlo simulation. The coincidence eect of the detection eciency for
gamma rays which are emitted through cascade transition was appropriately simulated.
The sampling of airborne radioactivity was started on 18 March, 2011, seven days after
the Great East Japan Earthquake. The sampling site was placed at the top of the building
of the University of Tokushima, placed at 134 33E longitude, 34 4N latitude and 15m
altitude, about 700km southwest of Fukushima. The airborne radioactive isotopes were
collected by a high volume air sampler HVC-1000N provided by SIBATA whose sampling
rate was 1m3 /min. The lter for sampling was a commercial glass lter GB-100R provided
by ADVANTEC with the dimension of 203mm254mm. The eciency for retaining particles with a size of 0.3m is 99.88%. The sampling was started at 12:00 and continued
23 hours.
The lter was striped into 1cm width bands and contained into a sample container
made of polycarbonate. A sample container was placed in front of the end cap of a HPGe
detector. The distance between the end cap and the sample container was 0.3cm. Fig.1
shows the Ge detector system, whose shield was opened. The Ge detector and the sample
Figure 1: Ge detector system. The front shield was opened in this picture. A sample
container was put on a foam polystyrene mount.
container was covered with 1cm thick OFHC (Oxygen Free High Conductive) copper plates
and 10cm thick lead bricks. The total gamma ray background was reduced three orders
of magnitude by the shield. The signal from pre-amplier was shaped by the shaping
amplier ORTEC 571. The pulse height was digitized by a multichannel analyzer. The
energy spectrum was stored into a hard disk every two hours and the data taking was
continued for 24 hours after the end of sampling. Since about 12 hours from the end of
sampling, the background events is dominated by the ones from the progeny of Rn, the
data for the present work were taken between 16 hours and 24 hours after the end of
sampling.
The Rn density was used for the check of the sampling eciency of aerosol. The annual
modulation of Rn density in Tokushima has been measured for 17 years. The density of
Rn does not change every year in the same season[2].
The signicant ssion products and activation products were measured from 23 March,
2011. The energy spectrum was shown in Fig.2. After 23 March, clear peaks due to
134 Cs and 137 Cs were observed.
131 I,
COUNTS
900
800
I-131
700
600
Cs-134
500
Cs-137
400
Cs-134
300
200
100
0
300
400
500
600
700
800
ENERGY(keV)
Figure 2: The energy spectra of gamma rays which was emitted by radioactive isotopes
attached to a lter. The red line is the energy spectrum taken on 6 April 2011. The black
line is the one taken before the Great East Japan Earthquake.
The detection eciency for gamma rays from 134 Cs and 131 I must be carefully estimated.
The detection eciency is distorted by coincidence of some gamma rays which are emitted
by cascade decay. For example, the 604keV gamma ray from the excited state of 134 Ba is
accompanied by the other gamma rays. Consequently, the detection eciency is distorted
by the coincidence with other gamma rays. The eciency distortion depends on the
geometrical distribution of the source and the detector. However, the correction of the
distortion is rather dicult because the geometrical arrangement may change by each
measurement.
The Monte Carlo simulation is the good tool to determine the detection eciency for
a complex geometrical arrangement. In the present work, Geant4.9.4.p02 was used to
determine the eciency. Geant4 [3] is the simulation tool kit to simulate the transportation of gamma ray, beta ray and other ionizing radiation particles. The class
G4RadioactiveDecay in Geant4.9.4.p02 generates the radioactive decays of almost all
the nuclei. The properties of unstable nuclei, half life, decay mode, excited state, branching ratio are listed in the class. The cascade emission of gamma rays is properly simulated
by the G4RadioactiveDecay class.
To verify the simulation, a simulation and a practical measurement were performed.
The measurement was performed by using the IAEA-444 standard source [4]. The standard source was contained in a U-8 pack, whose dimension was 50.4mm60.2mm. The
radioactive sources were uniformly mixed into soil. The U-8 pack was put in front of the
end cap of a HPGe detector. The simulation was performed with the same dimension.
The energy dependence of the detection eciency which were derived by experiment and
10
simulation was plotted in Fig.3. The Monte Carlo simulation well agreed with the ex-
Figure 3: The detection eciency of gamma rays. Box: Experimental result. Circle:
Simulated result. Experimental result for 569keV and 563keV could not dissolved so that
the eciency was calculated as 569keV gamma ray.
perimental result. Especially, the detection eciency of 604keV gamma ray was lower
than the energy dependence of eciency which was tted by polynomial function. We
determined the detection eciency for the various shapes of sample containers by Monte
Carlo simulation.
4
4.1
Analysis
Correction of radioactive decay
From the peak yields measured by Ge detector, the density of radioactivity was calculated.
The eect of the radioactive decay is signicant for calculating the proper radioactivity of
short lived nuclei, for example, 131 I. The half life of 131 I is so short as 8.04 day that the
decay between the sampling was not negligible. The amount of attached nuclei on a lter
A is corrected by means of following procedure[6].
First, the total number of attached nuclei AT is expressed as
AT = V TS g,
(1)
where , V , TS and g are the density of the nucleus in air, the volume of sampled air per
unit time, the sampling time and the collection eciency of lter, respectively. During
the sampling, the number of nuclei on the lter A(t) increases as P = V g. The number
of the nuclei decreases by radioactive decay, thus the equation is
dA(t)
= P A(t).
dt
11
(2)
Solving the equation, one gets the actual number of nuclei attached on the lter at the
end of sampling, say AT as,
TS
,
(3)
AT = A(TS )
1 exp(TS )
where, = ln 2/T1/2 is decay constant and T1/2 is half-life.
After sampling shed, the nuclei decays with the decay constant , thus the number
of the nuclei when the beginning of measurement is
A(TC ) = A(TS ) exp(TC ),
(4)
where TC is the time interval between the ending of sampling and the starting of measurement.
The radioactive decay should be considered if the half-life is short. The number of
decay N between the measurement time TM is expressed as
N
= N (TC ) N (TC + TM )
= N (TC ) {1 exp(TM )} .
(5)
(6)
where and a are the detection eciency of gamma ray and the emitting ratio of the
gamma ray.
The important parameters for the present measurement are listed in Table 1.
Table 1: The parameters of the present measurement.
Parameters
Sampling time TS [hour]
Time interval TC [hour]
Measurement time TM [hour]
Decay constant of 131 I [hour1 ]
4.2
Value
23
16
8
3
3.592 10
Eciency of lter
The commercial lter GB-100R provided by ADVANTEC was used for sampling the airborne radioactivity. The lter eciency of aerosol whose diameter is larger than 300nm
is as large as 0.9999 [7]. The capture eciency of airborne radioactive isotope depends on
the chemical property.
4.2.1
Cesium
137 Cs
and 134 Cs are exhausted from a reactor attached on aerosol. The size of aerosol
which attaches cesium was measured as 830860nm [8]. The lter eciency for cesium
was conrmed by measuring the density of 7 Be. The size of aerosol containing 7 Be is the
same as the one containing cesium. The density of 7 Be in Tokushima has been measured
since 2005. The density of 7 Be in Tokushima was 26mBq/m3 , which agrees the results
measured at another sites [9]. The lter eciency of GB-100R for cesium isotopes is
considered as g = 0.999.
12
4.2.2
Iodine
The lter eciency g for iodine depends on the chemical structure of iodine. The gaseous
iodine such as I2 , HOI and CH3 I cannot be collected by normal glass lter. Only a small
fraction of iodine ion and chemical compounds of iodine which are attached on aerosol is
cached by our glass lter. The size of aerosol particle which iodine is attached is rather
smaller than the size of 137 Cs, the average size is 590613nm [8]. The lter eciency for
iodine attached on aerosol is treated as 0.999, however, the fraction of the particle iodine
must be considered.
The fraction of iodine forms was investigated by Noguchi [10]. The detail of the fraction
is shown in technical note on the monitoring of airborne radioactivity[11]. The average
fraction of iodine attached on aerosol is 0.20 0.1. The total lter eciency is the product
of 0.999 and the average fraction of particle iodine, resulting g = 0.2 0.1.
Density mBq/m
The daily variation of 131 I densities is shown in Fig. 4. The error of each data is dominated
20
18
16
14
12
10
8
6
4
2
0
M
ar
19
M
ar
24
M
ar
29
Ap
r3
Ap
r8
Ap
r1
3
Ap
r1
8
Ap
r2
3
131 I
Ap
r2 Date
8
in air.
by the error of correction eciency g. The signicant radioactivity was observed after 23
March. The behavior of cesium isotopes was the same as iodine, however, the concentration
of cesium isotopes was one order smaller than the one of iodine.
The maximum of the density was observed on 6 April, about three weeks after the large
vent. This behavior cannot be explained that the radioactive plume came directly from
Fukushima Daiichi Nuclear Plant. The dates which the maximum density was measured
in other sites have a strong correlation. The speed of the transportation of radioactive
plume is explained that westerly wind brought the radioactive plume. Fig.5 shows the
relationship between the dates which the maximum radioactivity was observed in each
cites. The dates of each cites and their longitude have a strong correlation. The speed of
the plume was calculated from the linear tting. The speed was 40km/sec which agreed
the speed of westerly wind.
In western side of American Continent (Seattle), the rst signicant observation was
on 17 March and reached the maximum on 19 March[12]. In eastern side of American
13
Longitude (Fukushima=0)
350
300
250
200
150
100
50
0
10
15
20
25
30
35
40
45
DATE Mar 1 =1
Figure 5: The dates which the maximum density of radioactivity was observed. A:
Seattle (USA)[12] B:Chapel Hill (USA)[13] C:Huebla (Spain)[14] D:Orsay (France)[15]
E:Thessaloniki (Greece)[16] F:Tokushima (Japan, This work)
Continent (Chapel Hill), the rst observation was on 18 March, however, the maximum
was observed on 29 March. This discrepancy came from the rainfall on 20 March [13].
They observed three peaks of density. The rst peak of density was observed on 25 March.
After the rst peak, two peaks were observed on 30 March and 2 April. The density of the
rst peak was reduced by rainfall, so we considered that the plume arrived on 25 March.
The radioactive plume was brought to Europe by westerly wind. In western Spain
(Huevla), the maximum was reported on 28 March[14]. However, the sampling was made
only a few times, 15-17, 21-23 and 28-29 March. After 28-29, they continuously measured
till 15-17 April. In France (Orsay), the continuous measurement was reported by IRSN
(Institut de Radioprotection et de S
urete Nucleaire)[15]. The rst signicant observation
was on 25 March and the maximum was observed on 31 March. In Greece (Thessaloniki),
the rst signicant observation was on 26 March and the maximum was observed on 29[16].
The relationship between the dates of maximum density and the longitude was well tted
by linear function.
To conrm the hypothesis that the radioactive plume which arrived on 6 April went
around the Northern Hemisphere, further analysis was carry out. The isotopic component
in a plume from nuclear plant was dominated by cesium so that the ratio R 131 I/137 Cs
is larger than 10 [11]. The ratio becomes smaller during the long travel by the decay of
131 I and by dissolving into rainwater. The ratio in Seattle was reported as large as 31,
while, the ratios in Europe was between 10 to 4. The decrease in R is more rapid than
the radioactive decay of the isotopes.
The eect on the isotopic ratio was clearly observed by Asian sites. In Taiwan, the
value of R was dropped from 1 to 0.3 between the end of March and the beginning of
April[17]. In Vietnam, the value R decreased exponentially[18] and the ratio was small.
The temporary decrease of R was clearly observed in Tokushima. The daily variation
in the ratio R in Tokushima was analyzed as shown in Fig.6. In the beginning of April
(1st April 2nd April), the value of R was large, which suggests the radioactive plume
came directly from Fukushima. During 3 to 7 in April, the ratio R temporarily decreased
14
Figure 6: The daily variation in the ratio of R 131 I/137 Cs (Upper) and R 131 I/134 Cs
(Lower).
as R 4. From the isotopic component, the biggest peak around 6 April was caused by
the radioactive plume exhausted on 1215 in March and the plume traveled around the
Northern Hemisphere. In Vietnam, the altitude is so small that the radioactive plume did
not pass through there.
The concentration of measured radioactivity was about ve orders of magnitude smaller
than the regulation in Japan. The estimated dose was negligibly low expecting no health
eect in western Japan.
Acknowledgements
The authors thank the University of Tokushima for supporting the continuous measurement.
References
[1] TEPCO Web site. (http://www.tepco.co.jp/nu/fukushima-np/index-e.html)
15
[2] S.Miyamoto et al., Natural Science Research Faculty of Integrated Arts and Sciences,
the University of Tokushima 13 (2000) 1.
[3] Geant4 Web site. (http://geant4.web.cern.ch/geant4/)
[4] Reference sheet of IAEA-444, Gamma emitting radionuclides in soil.
[5] National Nuclear Data Center (http://www.nndc.bnl.gov)
[6] JAEA-Technology 2010-039 (2010). (In Japanese)
[7] ADVANTEC Web site (http://www.advantec.co.jp/english/)
[8] A.Reineking et al., Radiation Protection Dosimetry 19 (1987) 159.
[9] M.Yamamoto et al., J.Env.Rad. 86 (2006) 110.
[10] H.Noguchi, The study on the change of property of radioactive iodine and tritium
in the environment, Thesis, Nagoya University (1991).
[11] Technical sheet on the monitoring of radioactive iodine.,(2003) Ministry of Education,
Culture, Sports, Science and Technology-JAPAN. (In Japanese)
[12] J. Diaz et al., J.Env.Rad. 102 (2011) 1032.
[13] S.MacMullin et al., J.Env.Rad., in press.
[14] R.L.Lozano et al., Environment International, 37 (2011) 1259.
[15] Institut de Radioprotection et de Suret
e Nucl eaire
Web site.
(http://www.irsn.fr/EN/news/Pages/201103_seism-in-japan.aspx)
[16] M.Manolopoulou et al., J.Env.Rad. 102 (2011) 796.
[17] Chin-An Huh et al., Earth & Planet. Sci. Lett. 319-320 (2012) 9.
[18] N.Q.Long et al., J.Env.Rad., in press.
16
17
Cs. In the case of the contamination by the accidents at Fukushima, main contaminants are 131I, 134Cs
and 137Cs. As the life time of 131I is short and has decayed already, we can limit 134Cs and 137Cs as main
contamination sources in soil or food produced on land. The response of dose rate meters to the
gamma-rays from 134Cs and 137Cs of the same radio activity are different. But because the difference of the
dose equivalent rate from the same radioactivity of 134Cs and 137Cs can be calculated and the abundance of
those nuclides are easily estimated, it is relatively easy to convert the correlation line for
134
137
Cs to any
137
111119 mm
0.53.0 (1501250 keV)
2500 cpm/Sv/h (with Cs-137)
8.5% (at 662 keV of Cs-137)
Silicon PIN Photodiode
18
specifications are listed in Table 2. The portable dose rate meter PA-1000 and the simple radioactivity
measurement kit PA-K are shown in Fig. 1.
Table 2: Specifications of PA-K.
Material
Size
Detection Efficiency
Distance from CsI(Tl) Crystal
center to Sample(Minimum)
Distance from CsI(Tl) Crystal
center to Sample(Maximum)
Capacity of bowl
Vinyl Chloride
200200 mm
0.9%
20mm
93mm
1 liter
137
Cs and
134
Radioisotope Association and prepared standard radioactive specimens. The specifications of the standard
nuclide solution are shown in the Table 3.
Table 3. Cesium standard nuclide solution.
Cs-134
Cs-137
We prepared standard radioactive specimens by adding these standard nuclide solutions to water, brown
rice, meat, and shiitake mushroom. The sample mass in 1 liter bowl and the concentration ranges of the
specimens are shown in Table 4. We prepared the specimens of three to five different Cesium
concentrations within these radioactive ranges by mixing certain amounts of solutions and specimens.
Table. 4. Specimen, sample mass in 1 liter bowl and concentration ranges of the standard
radioactive test specimens.
Specimen
Water
Brown rice
Uncooked meat
Shiitake mushroom
1000
43580
930
921010
100500
930
1071071
860
86860
Notes
0.01%KCl solution
0.8%KCl solution
Processed in blender
Processed in blender
Generally the counts of photoelectron absorption peak of the gamma ray in the spectrum are used to
obtain the radioactivity. But PA-1000 counts all the signals including Compton scattering background. As
long as the incident gamma ray spectrum is same, this count including Compton scattering constituent is
19
137
Cs results
The result for the standard water specimen is shown in Fig. 2 and the result for the brown rice specimen
is shown in Fig. 3.
In both of these results, we obtained a linear relationship between the net value in Sv/h and
concentration of radioactive 137Cs in Bq/kg.
We repeated measurements 5 times using the same brown rice specimen to evaluate the measurement
error. The result shows about up to 30% measurement error.
The measurement results of the standard radioactive specimens of water, brown rice, meat and shiitake
mushroom are shown in Fig. 4. We obtained a linear relationship with the same gradient for all the
specimens we measured. Therefore, it can be said that ingredients of the specimen do not have an
significant effect on the linear relationship.
20
0.04
Net value [ Sv/h]
0.3
Net value [ Sv/h]
0.2
0.1
0.03
0.02
0.01
0.0
0.00
0
2000
4000
6000
8000
10000 12000
200
400
600
800
1000
1200
Fig. 2 Results for measurement of standard radioactive water specimen. X-axis shows the concentration
of the standard solution of 137Cs, while Y-axis shows the net value after subtracting the background. The
right graph is with 10 times expanded scale of low concentration area of the left graph.
Standard radioactive brown rice specimen
0.05
0.04
specimen.
X-axis
shows
the
the
137
net
value
standard radioactive
brown rice specimen
0.03
0.02
0.01
after
0.00
0
200
400
600
800
1000
1200
Concentration of Cs-137 standard solution [Bq/kg]
brown rice
uncooked meat
0.04
mushroom
specimen.
water
shiitake mushroom
0.03
0.02
0.01
137
Cs,
0.00
0
21
200
400
600
800
1000
1200
Concentration of Cs-137 standard solution [Bq/kg]
Here, we should discuss the effect of 40K with 1461 keV on the measurement. Many foods contain 40K,
but the energy response of PA-1000 in over 1250 keV is very low. We examined the influence of 40K by
measuring dried kelp and dried shiitake mushroom, which are rich in 40K. Table 5 shows the net value
measured using PA-K, and radioactive concentration of Cesium and 40K measured by the germanium
semiconductor detector. From the result, it can be said that the gamma rays of 1461 keV emitted by 40K
have little effect on the measurement in the condition that 40K concentration is less than about 1600 Bq/kg.
Table. 5 Measurement results of dried kelp and dried shiitake mushroom. Concentration of 134Cs,
137
3.2
134
Cs+137Cs results
Since both 134Cs and 137Cs are contained in the radioactive Cesium emitted from the nuclear power plant
disaster, specimens with these two Cesium types need to be examined. We show here the measurement
results of the standard specimens created using a 1:1 ratio of 134Cs and 137Cs concentration. The results for
the standard water specimen are shown in Fig. 5, and the results for the brown rice specimen are shown in
Fig. 6. They show that the dose rate from the specimens containing both 134Cs and 137Cs is larger than that
from the specimens containing 137Cs only, even if they have same radioactivity concentration.
Standard
Standard radioactive
radioactive water
water specimen
specimen
0.1
0.1
Measured
Measuredvalue(only
value(only Cs-137)
Cs-137)
Measured
Measuredvalue(Cs-134Cs-137=11)
value(Cs-134 Cs-137=1 1)
Net Value[
Value[Sv/h]
Sv/h]
Net
0.08
0.08
0.06
0.06
0.04
0.04
0.02
0.02
00
00
500
500
1000
1000
1500
1500
2000
2000
Concentration
Concentrationof
ofradioactive
radioactiveCs
Cs standard
standardsolution
solution[Bq/kg]
[Bq/kg]
Fig. 5 Measurement results of standard radioactive water specimen. X-axis shows the
concentration of the standard solution of 134Cs+137Cs, while Y-axis shows the net value
after subtracting the background.
22
Standard radioactive
radioactive brown
brown rice
rice specimen
specimen
Standard
0.1
0.1
Measured value(only
value(only Cs-137)
Cs-137)
Measured
Measured value(Cs-134
value(Cs-134Cs-137=11)
Measured
Cs-137=1 1)
Net Value[Sv/h]
0.08
0.08
0.06
0.06
0.04
0.04
0.02
0.02
00
00
200
200
400
400
600
600
800
800
1000
1000
1200
1200
Concentration of
of radioactive
radioactive Cs
Cs standard
standard solution
solution [Bq/kg]
[Bq/kg]
Concentration
137
Cs, by multiplying the ratio 2.69 of these conversion factors and 1.07 of
PA-1000 energy response. Fig. 7 shows the results of calculating the value for a mixture of 134Cs and 137Cs
based on the actual measurement of 137Cs, using these ratio. This figure shows the synthesized line (dot)
agrees very well with the measured dose rates of the water specimens containing both 134Cs and 137Cs.
Using this estimation, we synthesized a correlation line for 134Cs+137Cs using the regression line obtained
from the measurements of the specimens containing
137
134
Cs and
137
Cs
changes with time due to difference in half life, the correlation line can be converted to apply to an
appropriated mixture of 134Cs and 137Cs, from the correlation line made at a different ratio.
23
Cs-134
Cs-137
Measured value(only
value(only Cs-137)
Cs-137)
Measured
Measured value(Cs-134Cs-137=11)
value(Cs-134 Cs-137=1 1)
Measured
Net
NetValue[
Value[Sv/h]
Sv/h]
0.08
0.08
Calculated value(Cs-134Cs-137=11)
value(Cs-134 Cs-137=1 1)
Calculated
0.06
0.06
0.04
0.04
0.02
0.02
00
00
200
200
400
400
600
600
800
800
1000
1000
Concentration of
of radioactive
radioactive Cs
Cs standard
standard solution
solution [Bq/kg]
[Bq/kg]
Concentration
134
134
137
Cs, using the ratio of a 1-cm dose equivalent rate coefficient and the energy response of
134
Cs+137Cs, while
Y-axis shows the value difference after subtracting the average background radiation.
4.2 Detection limit
The detection limit of this method can be estimated from the standard deviation of background. For
example, in Fig. 8 the background during the measurement of the standard water specimens was 0.07
Sv/h and the standard deviation was 0.006. This measurement contains 300 data, each measured for five
times at one-minute intervals. Based on this, the detection limit is calculated to be approximately 220
Bq/kg in 99% of reliability (3), approximately 140 Bq/kg in 97% of reliability (2), and approximately
70 Bq/kg in 68% of reliability (1). The detection limit depends on the variation of the background. And
the variation becomes larger when the background is higher. Therefore to improve the detection limit, it is
necessary to reduce the background by shielding the instruments by lead or other material.
24
Standard radioactive
radioactive water
water specimen
specimen
Standard
0.06
0.06
0.05
0.05
0.04
0.04
0.03
0.03
0.02
0.02
0.01
0.01
00
00
100
100
200
200
300
300
400
400
500
500
600
600
700
700
Concentration of
of radioactive
radioactive Cs
Cs [Bq/kg]
[Bq/kg]
Concentration
Fig. 8 Detection limit of this method. X-axis shows the concentration of the
standard solution of 134Cs+137Cs, while Y-axis shows the net value after subtracting
the background.
4.3 Comparison with a germanium semiconductor detector
We applied this measuring method to six kinds of food and cattle feed from Fukushima. And we also
compared the results with the measurement results by a germanium semiconductor detector. The
correlation line used for this measurement was for the brown rice. The measurement place was in a small
building at Fukushima Prefecture. Background radiation was 0.04 Sv/h there.
The results of the comparison are shown in Fig. 9. The X-axis is the measurement result using the
germanium semiconductor detector, and the Y-axis is the measurement result using this method. This
figure shows that the results of this method agree with the results by the germanium semiconductor
detector.
25
Radioactivity
Radioactivity measured
measured by
by PA-K
PA-K
[Bq/kg]
[Bq/kg]
1000
1000
Ricestraw
straw
Rice
800
800
600
600
400
400
Umeboshi
Umeboshi
Japanesebasil
basilberry
berry
Japanese
200
200
Barley
Barley
Mulberryjam
jam
Mulberry
00
00
200
200
400
400
600
600
800
800
1000
1000
Radioactivitymeasured
measuredby
bygermanium
germaniumsemiconductor
semiconductordetector
detector
Radioactivity
[Bq/kg]
[Bq/kg]
26
1. INTRODUCTION
One of the most important issues on Fukushima Dai-ichi nuclear power plant is the clarification of
the distribution of fuel debris, especially after melting and solidification, in the reactor containment vessel
after the probable meltdown of core, for its dismantling and after treatment.
Muon radiography using cosmic ray muons is the most powerful means for it. It is done by
detecting tracks of muons produced in the atmosphere and analyzing their angular distributions. It has been
(1)
applied to the analysis of inner structure of pyramid by Alvarez 's group with spark chambers , to
(2)
volcanoes by Nagamine's with plastic scintillator hodoscope and so on.
In the case of Fukushima Dai-ichi, high radioactivity and lack of electronic power supplies should
be taken into account. Electronic detectors seriously suffer from these factors. They require power supplies
and huge shielding material to reduce count rate from the background radioactivity to the practical level.
(3)(4)(5)
" as the best muon detector for probing the inner-structure of
We propose "nuclear emulsion
Fukushima Dai-ichi. It is high-resolution photographic film for detecting 3-demensional tracks of charged
particles in its volume with the position resolution of sub-m, which gives us the angular resolution of a
few mrad. As it does not require electronic power, the unit of the detector is compact. Because nuclear
emulsion is accumulative detector, the observation period is limited by integrated background. In the case
of observation under 10 mSv per hour in radioactivity, the maximum of the period with keeping the
practical background level by using a shielding material of 10cm thick lead is about 10 days. Under this
environment and the following conditions: 1, uranium dioxide block, which size is 1m cubic, is placed
behind the 10 m thick concrete wall; 2, the uranium dioxide block is located in the direction of angle of 45
degrees from vertical on the ground and 20 m away from the detector; we need 1 m2 detector area to detect
the uranium dioxide block beyond statistical error of 3 sigma with a position resolution of 40cm, which is
27
equal to an angular resolution of 20 mrad. Moreover, we need larger area to measure in higher radioactive
environment or to perform the tomographic analysis using multiple detectors placed around the target.
Thus, scanning speed of nuclear emulsion is the essential point to realize muon tomography with nuclear
emulsion.
We succeeded in developing a high-speed 3-demensional nuclear emulsion read-out system
(6)(7)
S-UTS based on optical microscope with the speed of 72 cm2 per hour through a neutrino oscillation
(8)
experiment, OPERA . This scanning speed was achieved by developing the new contraption of realizing
the scanning speed of 50 views per second with the following conditions of image acquiring: the size of
field of view is about 230x230 m2, imaging resolution is about 0.45 m per pixel, and 16 tomographic
images acquired per view. The technical detail was written in ref (6). The system was also applied to and
(9)
provided the result in muon radiography using nuclear emulsion for volcanoes done by Tanaka's group .
We have conducted the experimental test of muon radiography with nuclear emulsion at the
(10)
experimental fast reactor Joyo, which belongs to Japan Atomic Energy Agency, in order to demonstrate
an imaging method of reactor cores. The result of the demonstration is reported here.
28
Fig. 1 Photographs of the nuclear emulsion OPERA film; (a) before development, (b) a cross
sectional photograph by an electric microscope.
Fig. 2 A photograph, which shows inside of the emulsion layer after development taken by a
high-resolution optical microscope. A signal of the present experiment is a straight track of muon.
Two scattered tracks of electrons and randomly distributing silver grains are background noises of
nuclear emulsion.
2.2 An automated nuclear emulsion scanning system
We have developed an automated nuclear emulsion scanning system S-UTS with scanning
(6)(7)
speed of 72 cm2h-1 to realize the OPERA experiment with a large amount of nuclear emulsion
. The
photograph of S-UTS is shown in Fig. 3. This system with an high-resolution optical microscope takes 16
tomographic images of an emulsion layer by a CMOS image sensor and searches the grains on the straight
line in the tomographic images to recognize them as micro tracks by an online digital processor. The
micro tracks are defined by their positions, angles and pulse heights. The pulse height is the number of
tomographic images, in which contain grains forming the micro track. Pulse height distributions of muons
originated from cosmic ray and fake tracks originated from background noises (see Fig. 2), which pulse
heights are 6 or more, is shown in Fig. 4 (a). On this scanning condition, the micro track recognition
29
efficiency of muon tracks is about 95 %. Although, the recognition efficiency is limited by the fake track
density, because the number of tracks is able to be recognized by the system is limited by the processing
power. Thus, we made the lower threshold of pulse height for scanning, which determines the efficiency.
After recognition of two micro tracks of the top and bottom layer, they are connected across the plastic
base using information of their positions and angles in order to reduce fake tracks produced by background
in emulsion itself, which are randomly distributed silver grains called fog and tracks of very low energy
electron recoiled by environmental gamma rays. The fake tracks due to electrons are not able to cross the
plastic base linearly because the electrons do not have enough energy. Connected tracks are called base
tracks, as shown in Fig. 5. The angular resolution of base tracks of this system is a few mrad in sigma
fitted by Gaussian function, as shown in Fig. 4 (b). Five S-UTS systems are operating now at the nuclear
emulsion scanning laboratory in Nagoya University, Japan.
Fig. 3 A photograph of automated nuclear emulsion scanning system S-UTS based on an optical
microscope.
Fig. 4 The performance of S-UTS; (a) A pulse height distribution of micro tracks. Solid line shows
signal tracks and dashed line shows fake tracks. (b) Angular resolution of base tracks and fitted by
Gaussian function. Data from minimum ionizing particles penetrating OPERA film in the vertical
direction on an emulsion surface are plotted on them.
30
3. DEMONSTRATION IN JOYO
3.1 Experimental overview
We have conducted the demonstration to take the muon tomographic image of the reactor core at
(10)
detectors were placed in the room located just under the vessel of Joyo plant, as shown in Fig. 6. In this
demonstration, 3 OPERA films were assembled into an emulsion detector, as shown in Fig. 7, at Nagoya
University. This structure was designed in order to reject background of base tracks by connecting fake
tracks to each other, as described in the next section. Fig. 8 shows a chamber consisted of 10 emulsion
detectors. The emulsion detector had a very simple structure and was lightweight enough in order to be
carried through narrow passages into the room under the reactor vessel and to be assembled easily and
quickly on the job site. Two chambers were placed with keeping the interval of about 1m. In such
installation condition as shown in Fig. 9, the reactor core was viewed from different directions, which
differed by about 0.1 rad. The observing period was 21 days.
31
Fig. 6 The left figure shows the overall layout of the reactor building. The reactor vessel and the
room are under ground level. The right shows the close-up layout of the structure focused on the
reactor vessel and its surroundings. A is the reactor core, which is a complex structure made of
uranium, plutonium, stainless steel and liquid sodium. B is inside of the reactor vessel filled with
liquid sodium and argon cover gas. C is radiation and thermal shields made of graphite. D is a wall
made of reinforced concrete. Scales are shown at the left bottom of each layout.
Fig. 7 The emulsion detector using OPERA films; (a) a photograph of the detector and (b) a cross
sectional view of the detector. The three OPERA films could be thoroughly bonded each other by
vacuum packaging with an aluminum-coated film. A muon passing thorough three OPERA films
leaves latent tracks in all emulsion layers, as shown in the right figure.
32
Fig. 8 The chamber assembled with 10 nuclear emulsion detectors; (a) a photograph of the chamber
without a rubber plate and lead plates and (b) a side view figure of the structure of the chamber. An
aluminum honeycomb plate gives the flatness of the chamber. The rubber plates are used to
distribute the pressure of the weight of the lead plates, which were used not to shield from radiation
but to fix the detectors on the aluminum honeycomb plate. This chamber was able to be assembled
without screws.
Fig. 9 A photograph (left) and a schematic view (right) of the observation setup. Chamber A and B
were placed at different positions under the reactor vessel. A distance between detectors and the
reactor vessel is about 10 m. The left photograph shows that the nuclear emulsion detectors are very
compact and are able to be taken data without a power supply.
33
34
valley pointed by the arrows in Fig. 10 (b) is about 4 bins, equal to tan = 0.16. This value is almost
consistent with the diameter of the reactor core, which is about 1.5 m, equal to about tan = 0.15 in angle.
On the other hand, two distributions of a region outside of valley give smaller difference in direction. This
is because the region was behind the core. In order to conclude the core effect clearly by reconstructing
three-dimensional distribution of matter density, we need to perform the tomographic analysis using more
detectors placed around the vessel. These radiographs suggest that such compact emulsion detectors
enable us to detect the reactor core and the plant structure.
Fig. 10 Histograms of raw data obtained by track reconstruction processing of nuclear emulsion.
The unit of these histograms is the number of tracks.
4. SUMMARY
An automated high-speed nuclear emulsion scanning technology realized the cosmic ray muon
radiography with nuclear emulsion, which has the advantage of compact, lightweight and no need of a
power supply. We have conducted the basic study of muon radiography of reactor core at Joyo. In this
paper, we described the analysis of two nuclear emulsion detectors placed at different positions. As the
result, we have observed the decrease of muon flux at the direction of the reactor core and the effect of the
35
other plant structure. Also the difference of decreasing by angle of two detectors was consistent with the
difference of the positions of detectors. The result validates the possibility of the imaging of the reactor
core with nuclear emulsion.
REFERENCES
(1) L.W. Alvarez, et al., Science 167 (1970) 832.
(2) K. Nagamine, et al., Nucl. Instr. and Meth. A 356 (1995) 585.
(3) R. Brown, et al., Nature 163 (1949) 47.
(4) T. Nakamura, et al., Nucl. Instrum. Meth. A (2006) 55680.
(5) N. Naganawa and K. Kuwabara, Journal of Instrumentation 5 (2010) P02006
(6) K. Morishima and T. Nakano, Journal of Instrumentation 5 (2010) P04011
(7) K. Morishima and T. Nakano, 2011, Ionizing Radiation, vol.37-1 (2011) -.
(8) M. Guler, et al., An appearance experiment to search for oscillations in the CNGS beam:
experimental proposal , 2000, CERN-SPSC-2000-028.
(9) H.K.M. Tanaka, et al., Earth Planet. Sci. Lett. 263 (2007) 104.
(10) Y. Maeda et al., Nuclear Technology, 150[1], 16-36 (2005).
(11) T. Toshito, et al., Phys. Rev. C 75 (2007) pp.054606-1~8,.
(12) S. Aoki, et al., Nuclear Physics B Volume 196 (2009) 50-53.
(13) K. Morishima, Ionizing Radiation, vol.37, No.3, (2011) 173-178.
36
!
"#$!"!"$
!
%&
"'()$!"$*""+)(+,-($./
% 0$1"
%
"233$1"$*""+)4-(4,$./
&
*"
%
!""5)-)$!"$*""+)'+,+)$./
%0*
$%6
7
*
8")+)+$& 0""$")54--2'./
)9178:81
*" : /6 /
;/
<"
./
9
"
9 =
"
/
$
/
9 8
$ /
/
9
:
/
/
$
/9
!'8 >8!?)@9
$
;
"/
6
9 *
$
/6/
9A
/
$
/
9!6
6
9
=
?
A@$ 6
6
6
9
37
$
6
;
9&
"
$A
6
""
9
///
6
!8 A
/
6
9
$6A6
)3
/
/
+
66/
9
*9):
/A
!8
!865,-D<E(9
!8
$6
6 *9'9A
/
6
6
6
$
6
/
9
38
*9'A
'9'
*(6
1)?/
6
@$1'?!8@$
?1'1)@9
1'
6
!8
$
1)
9
6),
9
:
/A/
-93'F)-),A?(@$6
6
9*
$1'1)
!8
9
*9(
A1)$1'$
*36
?3@$?,@9
6
'9-F)-),A
/
,D<E(9
/
A /
)9)D<E(9<
"
/
/?5@9
39
*93
/6
(9&AC1A&71&
A&8&!8
(9):
/A
!8
/
**93$6
6
A//
/
6
9
$6 / >
/
:
/A6
*9,9 "
$63,
!8
9A
$
66
1)9A1)
!8 " $6A1'
" $
?1)1'@
!89
40
/
9A
$
66
1)
6
!8
9
(9'&/
* 5 6
/
!89
!866
9
$6
?3@$?,@
<
!8
/
9
/
6)(?
/6
),9-/@
65,G9*
$
!8
!8
5,-D<E(,D<E($6/
)92F)-),A$
'9,D<E(?
!8
(',D<E(@<
+9,F)-)5A
*95
!8
**95$
6
/"
9*
$A//
/
6
9A
6 6
$
?/
A D@9D6 $
!8
6
" 69D
'''7
6
" 69
*
$
6
9
3981&81
=
/
/
9 6
9=
A//
41
/
6
9
7*71&
?)@B9 "$.9
C * 7$H925$19)-$8
'--?'@!9"$9%6$%9$19"$%9% $B9
$
.919&9935?'--4@)'
?(@19
$
:
7
6!
1 3-33-4$&D13+232')'+)?'---@$I ./ J
?3@B9B$&9"$7A:8&88C&$3)?)44'@
?,@!9"$9%6$%9=
$%9$19"$&9*"$%%
C $'-)-)-$C'''+
?5@%9=
$9
$.9C * 7$H92-$19)$
. )443
42
F-FDG is mixed in blood and the concentration of radiotracer in blood is measured quantitatively by online
system. The method is simple as well as convenient and immediate results will be very helpful in reducing the
diagnosis and treatment time. As a first step, the conventional detection system used for dose measurements in
nuclear power plant systems (NPPS) is assumed
(1),(2)
published results for comparison and the concentration of radioactive fission fragments in coolant is small,
analogous to our blood vascular source, with difference of the larger dimensions as compared to present case.
Electron Gamma Shower 5 (EGS5) code is used as simulation tool to obtain the detector response along with
considering the detector and collimator dimensions, various geometrical parameters which affect the detector
response. The obtained absolute full energy efficiency curves may be used to estimate the predicted number of
counts from a typical gamma source. These predicted counts may be helpful to decide whether a particular
system is suitable for the measurement of activity, and hence concentration, of given source or not.
2. RADIOACTIVITY MEASUREMENT
(1)
where, abs is detector absolute efficiency, S is net peak area, C is correction factor, Tl is measurement
time, A is source activity, is emission probability of gamma, and kc is decay during measurement. In
our case, the source is not a point source but instead it is cylindrical source which will be modeled by
considering the various parameters which include the blood vascular dimensions, its density, the blood
______________________________________
43
velocity, the medium and distance between the source and detector etc. As a first case, we discretize a
cylindrical source to many point sources and try to determine the suitability of system for the
measurement of activity of a weak source.
3. MONTE CARLO SIMULATIONS AND SYSTEM MODEL
The geometry used in simulation is shown in Fig. 1 and various parameters are mentioned. The
detector is High Purity Germanium (HPGe) semiconductor detector of 10mm thickness and 18mm
diameter shielded by lead collimator of two diameters 10mm and 20mm. In order to obtain a good
statistical certainty, large number of particles was generated and, hence by using source biasing
technique, reasonable counts were obtained for all cases. The particles were transported through the
material and were discarded if they either left the area of interest or their energy was below a certain
cut off energy. In first case, a point source was assumed at distance of 1 m and 106 particles were made
incident on detector front face. The intrinsic fullenergy efficiency, which was determined from
simulation, was used to find the absolute full-energy
efficiency by using analytical relation. In second case,
the results are reported by varying the radius of disk
source at same distance. The solid angle needed to
find absolute efficiency from intrinsic efficiency is
calculated by using various approximated or series
expansion formulas derived from Ruby's formula as
described by Pomme et al. (4). EGS5, used as
simulation tool, is a powerful system which can be
used to produce very complex Monte-Carlo
simulations in various geometries (5).
The absolute full-energy efficiency, very important parameter in the measurement of source
strength, is the ratio of number of detected photoelectron photons of energy E in a full-energy peak
Ndet to the total number of photons emitted Nemit by source with same energy (6).
abs = Ndet / Nemit
(2)
The absolute efficiency for point source can be found by using the relation
Nemit = Ndet.(4/if )
(3)
(6)
where, if is detector full-energy intrinsic efficiency and is the solid angle . In first case a point
source is assumed and the effect of source to detector distance and radial displacement of source is
studied for various values of gamma-ray energies of practical interest. The simulation results for
44
source to detector distance are compared with those obtained by Rodenas et al. by using MCNP
(2)
The normalized efficiency is obtained at various distances from 1 to 2m with step of 25cm and a good
agreement was found. The values of absolute full energy efficiency along with normalized efficiency
at different source to detector distance are given, and hence compared, in Table I. The point source is
then displaced in transversal direction from the axis of detector and absolute full-energy efficiency is
calculated at different radial distance and shown in Fig. 2. The obtained results are compared with
already published results by using GEANT4 (1) and MCNP (2) as well as experimental results (2), and
are summarized in Table II.
TABLE I
(a)
(b)
Fig. 2. Variation of absolute full-energy efficiency with radial distance of point source for collimator
of size (a) 10mm and (b) 20mm
4.2 Disk Source
In second case a disk source is considered at origin as a source detector parallel system. The source
radius is varied from zero (point source) to 10cm and variation in absolute full-energy efficiency with
source radius is shown in Fig. 3. The results are compared with those obtained by using Monte-Carlo
code Geant4 (1) for disc source of diameter 5cm and are shown in Table III. For the disk source, solid
angle is used to convert the intrinsic efficiency to the absolute efficiency. The solid angle is calculated
45
by using approximated and series expansion formulas of Rubys formula as given by Pomme et al. (4).
5. DISCUSSION
The simulation results of a point source and a disk source are presented for two collimator
diameters and five gamma-ray energies of interest. In first part, the effect of source to detector distance
is studied for the comparison of EGS5 simulation results with those obtained by using Monte-Carlo
simulation code MCNP (2). The source to detector distance is varied from 1m to 2m and a good
agreement is found in the calculation of normalized full energy efficiency at different distances. In
second part, the point source is placed at fix distance of 1m and moved in radial direction. As shown
TABLE II
Fig. 3. Variation of efficiency with source radius for collimator size 10mm
46
in Fig. 2, the absolute efficiency remains almost same as long as the radial distance of the point source
is less than the collimator radius and then it starts decreasing. This is due to the reason that the
thickness of lead between source and detector increases with the radial distance and the attenuation of
gamma-rays increases in collimator. The situation can be observed more clearly for the collimator of
20mm diameter as shown in Fig 2(b). The results are compared with those, obtained earlier by
Rodenas et al.(2) and Rehman et al.(1) by using Monte Carlo simulation codes MCNP and GEANT4
respectively as well as experimentally. The results are found in a good agreement as shown in Table II.
In the case of the disk source, it is placed at a distance of 1m and has the center on axis of
cylindrical detector. The radius of the source is varied from zero (point source) to 10cm and again the
absolute full-energy efficiency is measured for 10mm and 20mm collimator diameters for above
mentioned energies. The results are compared with literature values as shown in Table III. The
difference between our results and those obtained earlier is less than 11% except three cases where it is
15.2%(0.834MeV, 10mm collimator, GEANT4), 18%(1.332MeV, 20mm collimator, GEANT4) and
31%(1.332MeV, 20mm collimator, MCNP). The trend of the absolute full-energy efficiency is
decreasing with the increase in radius. The disk source, with negligible thickness, was assumed to
have same total activity uniformly distributed on whole surface, whereas the specific planar activity
decreases with increase in radius.
Various formulas are used to find the average solid angle, later used to find the absolute efficiency
from the intrinsic efficiency, and it was observed that almost similar results are obtained from the all
mentioned formulas. The main reason is that the source-to-detector distance is very large as compared
to the source and detector dimensions and the solid angle is less sensitive to source and detector
dimensions. The absolute full-energy efficiency remains almost constant till the radius of source is
comparable to collimator radius and after that the absolute full-energy efficiency decreases with
increasing the source radius as shown in Fig. 3 for 10mm collimator case. The simulation results
obtained by using Monte Carlo simulation code EGS5 for gamma rays of energy range 0.511 to
1.332MeV and the trends of absolute full-energy efficiency with source-to-detector distance, point
source radial position and disk source radius are consistent with already published ones.
TABLE III
Absolute full-energy efficiency of disc source of diameter 5cm for different energies
47
Various parameters like collimator thickness and diameter, detector radius and thickness, source-todetector distance, source dimensions and radial position could be studied and hence optimized before
the actual experiment by using EGS5 simulation code. The results obtained in this paper will be used
as benchmark for our future work i.e; development of a very fast and convenient online radiotracer
concentration measurement system for nuclear medicine.
6. CONCLUSIONS
In summary, the online measuring system is expected to be very fast and convenient in medical
diagnosis and treatment. The simulation results obtained by using Monte Carlo simulation code EGS5
for gamma rays of energy range 0.511 to 1.332MeV and collimator diameters of 10and 20mm are
found to be in good agreement with those obtained experimentally as well as by using Monte Carlo
codes MCNP and GEANT4. The variations in absolute full-energy efficiency with source-to-detector
distance, point source radial position and disk source radius are consistent with the expected behavior.
Various parameters like collimator thickness and diameter, detector radius and thickness, source-todetector distance, source dimensions and radial position could be studied and hence optimized before
the actual experiment by using EGS5 simulation code. By considering typical values of injected
radioactivity (10mCi) (7) with assumption of uniform distribution throughout the whole volume of
blood , average human blood volume (5L), blood vascular diameter (2.35mm), source length ( 20cm),
source to detector distance (10cm) etc, it may be found that total gamma rays 108 are emitted from
source for counting time of 30min from a point source, as result of discretization of cylindrical source,
and hence the measuring system of absolute efficiency 10-7 (at 1m distance) and 10-5 (at 10cm
distance) is not suitable for the activity measurement of such weak source. However, for strong
sources, the system may be used with the advantage of high energy resolution.
ACKNOWLEDGEMENT
The author is thankful to MEXT Japan for providing funding for study at The University of Tokyo
under ID: 103187 and Dr. R. M. Thushara Damayanthi for useful suggestions.
REFERENCES
(1) S. U. Rehman, S. M. Mirza, N. M. Mirza and M. T. Siddique, Annals of Nuclear Energy. 38 (2011)112.
(2) J. Rodenas, A. Martinavarro and V. Rius, Nucl. Instr. and Meth. A. 450(2000)88.
(3) D. Demir, P. Onder and T. Oznuluer, Rad. Phy and Chem. 79(2000)1132.
(4) S. Pomme, Nucl. Instr. and Meth. in Phys. Res. A , 371(2004)535.
(5) H. Hirayama and Y. Namito," The EGS5 code system ", SLAC-R-730, KEK-Tsukuba, Ibaraki, Japan, 2010.
(6) G. F. Knoll, Radiation Detection and Measurement, (John Wiley & Sons, Inc., New York, 2000) p116.
(7) K. Wienhard, Methods. 27(2002) 218.
48
1. INTRODUCTION
After a severe accident on a nuclear power plant, we should monitor and survey some
radioactive nuclides, such as Cs-134, Cs-137 and I-131. The radioactive cesium has relatively
long half-lives such as 2.1 years(1) for Cs-134 and 30 years(2) for Cs-137, and often remains near
the ground surface(3).
They, therefore, cause to rise the ambient dose over many years. The
radioactive I-131 has relatively short half-life of 8 days(1). The radioactive iodine accumulates
at the human thyroid and cause the thyroid cancer. To investigate the effect of radioactive
contamination, the environmental samples should be quantitatively measured.
Under the
situation that the field background level of the radiation is high, the environmental samples,
such as soils, water and aerosols, should be brought back and measured at the facilities, where
the detector can be well shielded.
suffers from background radioactivity, which remains in the air and on any surface.
The
increase of background radioactivity deteriorates the signal to noise ratio or the detection limit.
In the conventional survey techniques, the heavy shieldings, such as lead blocks, are
used to suppress the background.
make the portability low.
(CBC) technique to confine the sensitive area and to suppress background gamma-ray events.
The coincidence and/or anti-coincidence techniques between a primary gamma-ray and a
Compton scattered one are often used to suppress the background events in various low back
ground gamma-ray detectors, such as the multi gamma-ray detector array GEMINI-(4) of
JAEA, and to use for energy calibration in the light element detectors, such as, the plastic
scintillatots(5).
the energy deposition at the primary scattering point and the scattered gamma-ray energy, we
can determine the scattering angle.
The
vertical and horizontal axes correspond to the energy deposition in the two detectors and the z
axis corresponds to the coincidence counts.
energy of an incident gamma-ray is deposited in the two detectors, the event is added at the
point along the line segment between two points corresponding to the 180o backscattering event
in each detector.
Scattered gamma-ray
energy - keV
direction.
Figure 1
300
200
100
0
-180
-120 -60
0
60
120
Scattering angle - degree
180
Fiigure 2
obtained
from
the
The vertical
50
In the CBC technique, a pair of gamma-ray detectors are placed facing each other.
Selecting the events added at the point corresponding to near 180o backscattering in the both
detectors, we can confine the sensitive area into the interspace between the two detectors and
improve the S/N ratio even under the high background condition.
for foreground and background conditions are shown in Fig. 3, when the incident gamma-ray
energy is 364 keV emitted from I-131. For instance, when the scattering angle is 120 degrees,
that is the undesired direction or the background condition, the energies of a recoil electron and
a scattered gamma-ray are 188 keV and 176 keV, respectively.
backscattering occurs, the recoil electron and the scattered gamma-ray energies are 214 keV and
150 keV, respectively.
These events are counted as the foreground events or the desired events.
Figure 4 shows the two dimensional histogram obtained under above conditions.
Figure 3
51
300keV
100keV
300keV
100keV
100keV
300keV
Low
Figure 4
High
Examples of the two dimensional histogram obtained under the foreground and
background conditions.
were spared because the attenuation of the gamma-ray in them is quite low.
point source emitting 364 keV gamma-rays assuming I-131.
Figure 5
52
The detection efficiency distribution along the line with the same distance from the both
detectors is shown in Fig. 6. The horizontal axis indicates the distance from the center axis
connecting the both detector centers.
Coincidence measurements represent the efficiency obtaining the near 180o backscattering
coincidence events.
measurements decrease with decreasing the solid angle subtended by the detector surface.
The
efficiency of the coincidence measurements, however, suddenly decreases outside the detector
diameter.
diameter.
However, when placing the detector far from the center axis, the coincidence count
events, the deposition energies in two detectors can accidentally be the same with the case of the
180o backscattering events.
10-2
Efficiency
10-3
10-4
Single
Coincidence
10-5
10-6
10-7
Fiigure 6
5
10
Distance from center - cm
15
The detection efficiency distribution along the line with the same
distance from the both detectors. The horizontal axis represents
the distance from the center axis connecting the both detector
centers.
53
Figure 7
compared with the area between two detectors. The area between two detectors also
has two-order higher detection efficiency than the surround area.
Using theCBC
technique, we can drastically decrease the detection efficiency outside the sensitive area
sandwiched between a pair of detectors.
Figure 8
54
4. EXPERIMENTAL STUDIES
Through the basic experiments, we demonstrated the feasibility of the CBC technique.
Figure 9 shows the detector configuration in the basic experiments.
LaBr3 scintillators (
1.0x1.0 inch and
1.5x2.0 inch) is 8 cm.
Figure 9
Figure 10 shows the experimental detection efficiency distribution along the line with the
same distance from the both detector surfaces. The horizontal axis represents the distance
from the center axis connecting the both detector centers. The results obtained from the EGS5
simulations are also plotted in this figure.
depending on the solid angle subtended by the detector surface as well as the simulation results.
On the other hand, in the coincidence measurements, the decrease of the efficiency in the
experiments is more gradual and the area with relatively high detection efficiency is broader
than that expected in simulations.
and simulation comes from the finite energy resolution. To put it concretely, the events obtained
from when the source is between the detectors emigrate from the ROI in the two dimensional
histogram and ones when out of the area immigrate. Additionally, because the surrounding
materials are not contained for this simulation, the buildup effect may influence this detection
efficiency curve.
55
Exp.
Sim.
Single
Single
Coincidence
Coincidence
10-2
Efficiency
10-3
10-4
10-5
10-6
10-7
Figure10
2
4
6
8
10
Distance from center- cm
The
Top figure was acquired when the source was placed on the center axis.
Bottom one was acquired when the source was 5 cm far from the center axis.
was on the center axis, the clear peaks were obtained.
far from the axis, the peak width was broadened.
relationship between the scattered gamma-ray energy and the scattering angle as shown in Fig.
1.
Near 180o backscattering, the scattered gamma-ray energy has small change.
On the other
hand, for the scattering angle less than 120 , the slope of the scattered gamma-ray energy
change is relatively high.
Therefore, the peak obtained when the source was far from the
center axis was broader than one obtained when the source was on the axis.
56
184keV
478keV
basic experiments.
coincidence detection efficiency for the area between two detectors was estimated to be higher
than 1x10-5.
On the other hand, when the source position was far from the center axis
connecting the both detector centers, the coincidence detection efficiency was two order lower
than the case when the source position was interspace between two detectors.
From the
experimental results, although the performance confining the sensitive area was deteriorated
because of the finite and poor energy resolution of the detectors, we concluded that the CBC
technique is quite useful to confine the sensitive area and to suppress the influence of the
57
background gamma-rays. But we will evaluate the detail condition which includes the energy
resolution and the geometry and so on.
As future works, we should improve not only the signal to noise ratio but also the
detection efficiency, because the absolute efficiency of the CBC technique is two order lower
than that of the single measurements, in which a single detector is simply used without
coincidence.
The present system for the CBC technique consists of the relatively heavy
scattering and the final absorption for two detectors in order to separately optimize each
function.
On the other
hand, a detector for the final absorption should have high energy resolution, such as a HPGe
detector, in order to suppress the deterioration of the performance confining the sensitive area.
6. REFERENCES
(1) Yu.Khazov and I.Mitropolsky and A.Rodionov, Nuclear Data Sheets, 107 (2006) 2715-2930
(2) A.A. Sonzogni, Nuclear Data Sheets, 103 (2004) 1-182
(3) G.Arapis and E.Petrayev and E.shagalove et al., Journal of Environmental Radioactivity, 34
(1997) 171-185
(4)Y.Hatsukawa and Y.Miyamoto et al., Journal of Radioanalytical and Nuclear Chemistry, 272
(2007) 273-276
(5)L.Swiderski and M.Moszynski et al., Radiation Measurements, 45 (2010) 605-607
(6)H.Hirayama and Y.Namito et al., KEK Report, 2005
58
1. INTRODUCTION
The 3rd generation synchrotron radiation facilities such as SPring-8 produce high brilliance X-ray
beams, where 2-demensional detection equipped with the X-ray photon-counting capability is one of key
technologies for the experimental method. In high energy applications above 20 keV, especially, a high
stopping-power semiconductor material for the sensor and an advanced ASIC (Application Specific
Integrated Circuit) for the readout channel are required. CdTe is a promising semiconductor material for
such high energy X-ray detection, because the high atomic number (ZCd = 48, ZTe = 52) gives a high
quantum efficiency and the large band-gap energy (Egap = 1.44 eV) allows us to operate the detector at
room temperature. Therefore, we have developed CdTe pixel detectors based on the hybrid pixel detector
technology.
The first prototype detector of SP8-01 was designed with the pixel size of 200 m 200 m and the
matrix of 16 16
(1-4)
discriminator and a 20-bits counter in each pixel. The SP8-01 detector successfully operated with a single
X-ray photon-counting mode and fairly good energy linearity performance was achieved in the X-ray
energy region between 15 keV and 120 keV (2-4).
In the next step, we have developed the second prototype of SP8-02, which was redesigned with the
same pixel size of 200 m 200 m but the larger area with 20 50 pixels. In this study, we have
investigated long-term stability of Al-schottky CdTe pixel sensor assembled into the SP8-02 detector at
room temperature and lower temperatures.
59
m in pitch. A guard-ring electrode was placed along the edge. The surface was covered with a passivation
layer having holes 60 m in diameter on each pixel and the guard-ring. The back side (Xray irradiation
side) was processed with a single platinum-electrode. This electrode configuration of Pt/CdTe/Al-pixel has
an advantage that a high Schottky barrier formed on the Al/CdTe interface leads us to operation as an
electron-collecting diode pixel detector (5). Electrons have a larger mobility (electron ~ 1100 cm2/Vs) and a
longer lifetime (electron ~ 3 106 sec) than those of holes (hole ~ 100 cm2/Vs, hole ~ 2 106 sec) in CdTe.
60
Fig. 2 shows a photograph of SP8-02 CdTe detector assembly. The pixel electrodes of CdTe sensor
was bump-bonded to the SP8-02 readout ASIC by the gold-stud bonding technique (6). The ASIC has 55
control pads, which were wire-bonded to the electrical pins on a ceramic package. Negative bias voltage is
applied with wiring connected to the X-ray irradiation side electrode. The detector assembly was mounted
on the detector control board as shown in fig. 3. The readout ASIC equips with a preamplifier, a shaper, a
window-type discriminator and a 20-bits counter in each pixel as described in Fig. 4.
61
Long-term stability of SP8-02 Pt/CdTe/Al-pixel sensor was investigated with 241-Am X-ray source.
The total counts from the whole pixels were measured by performing lower-level discriminator threshold
scans repetitively. The upper-level threshold was set to the voltage of -1000 mV. Each exposure time was
10 seconds. The first scan was started immediately after applying the bias voltage of -300 V to the sensor,
and then repetitive scans were performed until 12 hours later with the 15 minute intervals.
Fig. 6 shows the result at room temperature. The solid line, the dashed line, and dotted lines are
corresponding to the first scan, the last scan, and in-between scans, respectively. The lower voltage side
(left side) corresponds to the high energy side and the edge around -250 mV is caused with 60 keV X-rays
from 241-Am. The rapid increasing between -50 and -100 mV are due to noise contributions. We
62
observed decreasing in the counting rate with the passage of time. This degradation is considered by the
drop of charge-collection efficiency because of the typical polarization effect in CdTe Schottky sensor.
However, the counting rate was still more than 50 % at the threshold voltage of -200 mV even after 12
hours. This degradation is slow enough to be used in most synchrotron radiation experiments.
The detector was placed in the thermostatic chamber and we investigated temperature dependence in
long-term stability. Fig. 7 shows the result at -10 degrees and the stability was dramatically improved. The
counting-rate drop was less than 9 % at the threshold voltage of -200 mV during 12 hours. The
degradation was negligible small to perform standard time-resolved X-ray diffraction experiments in
studying second- or minute-order responses.
Fig. 8 shows the result at 5 degrees. The long-term stability was successfully achieved at the near
room temperature and the counting-rate drop was less than 12 % at the threshold voltage of -200 mV
during 12 hours. This condition leads us to a more simple cooling procedure in designing the final detector
system.
63
64
REFERENCES
(1) H. Toyokawa, et al., Nucl. Instr. and Meth. A, 636 (2011), S218-S221.
(2) H. Toyokawa, et al., Proc. of the 24th Workshop on Radiation Detectors and Their Uses, KEK
Proceedings 2010-10, (2010), 220-225.
(3) T. Hirono, H. Toyokawa, et al., Nucl. Instr. and Meth. A, 650 (2011), 88-91.
(4) T. Hirono, H. Toyokawa, et al., Proc. of the 25th Workshop on Radiation Detectors and Their Uses,
KEK Proceedings 2011-8, (2011), 196-203.
(5) S. Watanabe, et al., Japanese Journal of Applied Physics, 46 (2007) 6043-6045.
(6) T. Takahashi, et al., IEEE Trans. Nucl. Sci., 48, No.3, (2001), 287-291.
65
1. INTRODUCTION
Transmission electron microscopy (TEM) equipped with an X-ray spectrometer is considered to be
one of powerful tools to investigate elemental component of materials. Typical spectrometers performed
with the TEM are energy dispersive X-ray spectrometer (EDS) and/or wavelength dispersive X-ray
spectrometer (WDS). The EDS can analyze almost all elements in one measurement due to high dynamic
range compared with WDS. However, the usual EDS detectors have a disadvantage of relative poor
(e-mail) MM61@kune2a.nucl.kyushu-u.ac.jp
66
energy resolution of approximately 130 eV at 5.9 keV in conventional silicon drifted detector (SDD) while
that of WDS is less than 10 eV.
A microcalorimeter with a superconducting transition edge sensor (TES) has previously
demonstrated an excellent energy resolution(1)(2). We have developed a TES microcalorimeter system
performed on the TEM(3). The TES microcalorimeter is placed outside the TEM column to avoid the
inuence of the magnetic eld generated by the objective lens of the TEM. The solid angle of the EDS
system decreases with inverse square of the length from a TEM-specimen to the absorber surface of the
TES microcalorimeter. Hollerith et al. used a polycapillary optic for increasing the effective solid angle of
the TES microcalorimeter installed in a scanning electron microscope(4). In the TEM-TES EDS system, an
X-ray polycapillary is employed for transporting X-ray emitted by a TEM-specimen to the TES
microcalorimeter. A 3 3 + 1 superconducting transition-edge sensor (TES) microcalorimeter-array has
been devised for the energy dispersive spectrometer performed on a transmission electron microscope. For
increasing sensitive area, a mushroom shaped absorber is deposited on the TES layer in each pixel. A thin
insulating layer is inserted between the overhang layer of the absorber and the TES surface for supporting
the overhanging structure.
In this work, three types of 3 3 + 1 TES microcalorimeter-array were fabricated with different
absorber structure. Response of each TES microcalorimeter-array to X rays was analyzed to characterize
contribution of the inserted insulating layer to detection performance.
2. TES MICROCALORIMETER
A microcalorimeter is a detector that measures the energy of an incident X-ray photon by the
temperature rise of the device with maintained at a very low temperature. It consists of an energy absorber
and a thermometer. The energy of the incident X-ray photon is converted into a temperature rise in the
absorber and the amount of the temperature rise is measured by the thermometer, and the temperature
returns to the heat sink temperature by a heat flow through a thermal link. The TES microcalorimeter
utilizes the strong temperature dependence of the electric resistance of a superconducting thin film in the
phase transition region for a precise measurement of the temperature rise.
In this work, three types of 3 3 + 1 TES microcalorimeter-array were fabricated. Figure 1 shows
schematic drawings of the TES microcalorimeter with different types of an absorber structure in the pixel
array. The TES of Type A, Type B and Type C consist of a bilayer of a Au/Ti 350 m 350 m in area.
Type A is commercialized by SII nanotechnology Inc. The Au absorber with 150 m 150 m and 0.5
m thick is deposited on the thermometer. Type B has a mushroom shaped Au absorber for increasing the
sensitive area. The geometrical dimensions of the absorber are surface area of 330 m 330 m and 0.5
m thick. The absorber stem of 150 m 150 m in the cross sectional area was deposited on the center
of the TES surface. The operational performance of the TES microcalorimeter is affected by deformation
of the overhanging structure of the mushroom shaped absorber. For supporting the overhanging structure
of the mushroom shaped absorber, a Ta2O5 insulating layer of 0.1 m thick is inserted between absorber
67
and thermometer. Contribution of the inserted Ta2O5 insulating layer to the detector response has not been
studied so far. Therefore, the stem of the mushroom shaped absorber is removed in Type C. Figure 2
shows photograph of the fabricated 3 3 + 1 array of TES microcalorimeters.
68
Fig. 2 Photograph of the fabricated 3 3 + 1 array of TES microcalorimeters with a mushroom shaped
absorber
3. EXPERIMENT
Figure 3 shows experimental results of the temperature dependence of electric resistance (R-T curve)
of the fabricated TES microclorimeters of Type A, Type B and Type C. Left panel shows the R-T curve of
Type A, right panel shows R-T curves of Type B and Type C. Experimental values of the transition
temperature, the width of transition region and the normal resistance of each TES are summarized in Table
1. Since transition region and the normal resistance of Type B and Type C are larger and higher compared
with those of Type A, the Ta2O5 insulating layer would cause the degradation of the TES performance.
However a mechanism of the Ta2O5 insulating layer affecting the TES performance is not clear in this
work.
Fig. 3 R-T curves of commercialized TES (left panel) and TES microcalorimeter with mushroom
shaped absorber with and without stem (Right panel).
69
Figure 4 shows relationship between the applied bias current and the SQUID output voltage
corresponding to the TES current (IV curve) of each TES. Left panel shows the I-V curve of Type A,
right panel shows the I-V curves of Type B and Type C. In the I-V curve of Type A, the straight line
section in low bias current region indicates the superconducting state of the TES. The curved line section
implies the TES operating in transition region. The straight line section in high bias current region indicates
the normal conducting state of the TES. For Type B and Type C the normal conducting state region is not
observed in the I-V curve due to broad transition region.
TES microcalorimeters of Type A, Type B and Type C were irradiated by X-ray photons emitted by
55
an Fe source. Operating conditions of each TES microcalorimeter are summarized in Table 2. Figure 5
shows examples of detection signal pulses. In Fig. 5 (a), (b) and (c) show detection signal pulse of Type A,
Type B and Type C, respectively. The electronic noise was not sufficiently reduced in detection signal
pulses of Type B. Different shapes of detection signal pulses with fast and slow components were
observed for Type B and Type C. Difference in the pulse shape implies dependence of X-ray response to
the incident position in the absorber.
Table 1 Property of each TES.
TES
Transition
temperature (mK)
Transition
region (mK)
Normal
resistance (m)
Type A
Type B
Type C
220
189
212
7
36
16
38
540
666
Fig. 4 I-V curve of commercialized TES (left panel) and TES microcalorimeter with mushroom
shaped absorber with and without stem (Right panel).
70
(b)
(a)
(c)
Fig. 5 Detection signal pulses of TES microcalorimeter Type A (a), TES microcalorimeter Type B (b)
and TES microcalorimeter Type C (c).
Fig. 6 Rise time distribution of commercialized TES (left panel) and TES microcalorimeter with
mushroom shaped absorber with and without stem (Right panel).
Since a production process of detection signal in the TES microcalorimeter contributes to the rise
time component of detection signal pulses, the rise time is evaluated for obtained detection signal pulses.
The rise time is dened as duration of the pulse growing from 10% to 90% of the full pulse height. Figure
6 shows distribution of rise time of the detection signal pulses. Left and right panels show the rise time
distribution of Type A and Type B with Type C, respectively. The rise time distribution of Type A has one
component, while two components are found in the rise time distribution of Type B and Type C. Fast and
71
slow components of the rise time distribution of Type B and Type C appear 2-10 s and 10-30 s,
respectively. For Type B and Type C detection signal pulses with the fast rise time corresponds to events
of X-ray photons absorbed in the TES, while events of X-ray photons absorbed in the absorber create
detection signal pulses with the slow rise time. Pulse height distributions of Type B and Type C were
obtained by eliminating detection signal pulses with the fast rise time. Fig. 7 shows the energy spectra of
X-ray photons emitted by the 55Fe source obtained Type A, Type B and Type C. The FWHM values of
the energy resolution at the full energy peak of 5.9 keV Mn K X ray are obtained to be 54 eV for Type A,
146 eV for Type B and 88 eV for Type C. The poor energy resolution of Type B is caused by the large
electronic noise. In the spectra obtained by Type A and Type C, full energy peaks of Mn K X ray have
asymmetric shape due to the position dependence, while symmetric shape of the full energy peak appears
in the energy spectrum obtained by Type B. The energy resolution of Type B is expected to be improved
by reducing the electronic noise.
(a)
(b)
Mn K
Mn K
(c)
Fig. 7 Energy spectra of X-ray photons emitted by the 55Fe source obtained by Type A (a), Type B (b)
and Type C (c).
4. CONCLUSIONS
Three types of 3 3 + 1 TES microcalorimeter-arrays were fabricated with different absorber
structure. By analyzing the detector response the energy resolution is expected to be improved for the TES
72
microcalorimeter with the mushroom-shaped absorber deposited on the thin insulating layer in an
overhang region. We are conducting further investigations of a mechanism of the Ta2O5 insulating layer
affecting the TES performance.
REFERENCES
(1) S.R. Bandler, et al. J. Low. Temp. Phys. 151 (2008) 400.
(2) W. B. Doriese, et al. Appl. Phys. Lett. 90 (2007) 193508.
(3) T. Hara, et al. J. Electron Microsc. 59, (2010) 17.
(4) C. Hollerith, et al. Nucl. Instrum. Methods A 520, (2004) 606.
73
1. INTRODUCTION
Achieving high resolution PET scanner, the pixel size is an important factor to decide the spatial
resolution. But the small pixel size requires a large number of pixel crystals and corresponding readout
channels. Many readout systems are using analog multiplexing method, such as charge division to
decrease the number of channels and electronics to a reasonable number. But the analog multiplexing
introduces the complexity and the degradation in the count rate of the system. It is also known the analog
multiplexing degrade the final spatial resolution of the PET system, and the individual coupling and
individual readout are ideal. We adopted Time over Threshold (ToT) method to realize one to one coupling
individual readout system because of its simplicity and its low power. The fast digitization in the front-end
provides the benefits of the easier post processing like multiplexing (1). A 48 channel Time over Threshold
based front-end ASIC is newly designed and fabricated for a pixelated Pr:LuAG -APD based PET detector.
The ASIC is fabricated with 0.25 m TSMC CMOS technology and its die size is about 2 mm by 5 mm.
74
The front-end readout board is also fabricated with a ToT-ASIC and threshold controlling DACs. 3 frontend board are used for the readout from 12 by 12 (144 channel) pixelated detectors to form a PET
detector.
2. DESIGN AND FABRICATION OF A PIXELATED PET DETECTOR
2.1 PR:LUAG APD PIXELATED CRYSTAL
The detector is composed of 12 by 12 Pr:LuAG scintillator crystals individually coupled with 12 by
12 pixelated Avalanche Photo Diode. The matrix of pixelated crystals is shown in Fig. 1. The size of each
pixel is 2 by 2 by 10 mm3 and each pixel is divided with a BaSO4 layer of 0.2 mm thickness. Pr:LuAG is a
promising scintillator for PET application with the density of 6.7 g/cm, the fast decay time of 20ns, the
high light output of 20000 photon/MeV and the good energy resolution of 4.6 % at 662 keV (2).
75
76
77
78
Fig. 8 Energy spectrum of 22Na gamma source acquired by ToT-ASIC (APD HV=390 V and ToT was
counted with 4 ns resolution)
79
The coincidence time resolution is measured using two opposing Pr:LuAG APD PET modules with
511 keV annihilating gamma rays from 22Na source (1MBq). The distance between 2 modules are set to be
10.0 mm (Fig. 9). The time difference was measured with a HRMTime TCSPC TDC (SensL). Fig. 10
shows the result of the time resolution, which was 6ns FWHM for coincidence without any energy
window and 4.25 ns for single module. This result is well within the operating specification of our animal
PET system of ~20 ns timing window.
Fig.10 Time resolution for 511keV annihilation gamma rays measured with 2 coincidence modules
3. CONCLUSION AND DISCUSSION
We have newly developed a 144 channel Pr:LuAG APD PET detector for the animal PET
application. The developed detector has the specification of 10 % energy resolution at 511 keV and 4.25 ns
time resolution as a single module, which is well enough for our animal PET system. The Time over
Threshold based individual readout will greatly improve the count rate of the whole system compared with
conventional electronics using the charge division method. We are now developing a animal PET system
with this developed detectors.
80
REFERENCES
(1) Shimazoe, K.; Takahashi, H.; Shi, B.; Furumiya, T.; Ooi, J.; Kumazawa, Y.; Murayama, H.; , "Novel
Front-End Pulse Processing Scheme for PET System Based on Pulse Width Modulation and Pulse Train
Method," Nuclear Science, IEEE Transactions on , vol.57, no.2, pp.782-786, April 2010
(2) Kamada, K.; Yanagida, T.; Tsutsumi, K.; Usuki, Y.; Sato, M.; Ogino, H.; Novoselov, A.; Yoshikawa, A.;
Kobayashi, M.; Sugimoto, S.; Saito, F.; , "Scintillation Properties of 2-Inch-Diameter Pr:Lu3Al5O12
(LuAG) Single Crystal," Nuclear Science, IEEE Transactions on , vol.56, no.3, pp.570-573, June 2009
(3) J. Kataoka, M. Koizumi, S. Tanaka, H. Ishibashi, T. Nakamori, N. Kawai, H. Ikeda, Y. Ishikawa, N.
Kawabata, Y. Matsunaga, S. Kishimoto, H. Kubo, Development of large-area, reverse-type APD-arrays for
high-resolution medical imaging, Nuclear Instruments and Methods in Physics Research Section A,
Volume 604, Issues 1-2, 1 June 2009, Pages 323-326
81
Liquid xenon TPC (LXe-TPC) has been developed for a new type of PET (Positron-emission
tomography) at High Energy Accelerator Research Organization (KEK). The LXe-TPC is a
radiation detector that integrates a time projection chamber (TPC) using dense liquid xenon and
a fast response scintillation detector. The LXe-TPC must have features of Low temperature
operation and Low power because liquid xenon is formed under specific temperature
conditions (at about 165K). Readout circuit needs to locate of near detector because decreasing
input noise. On this request the readout circuit must operate low temperature. In addition, the
TPC need to have a feature of Multi-Channel read out because many of scintillation
detectors can be realized high resolution. Therefore, front-end circuit of the LXe-TPC must
have these features. We have been developing a front-end circuit of the LXe-TPC, the name is
TPCFE ASIC. We created 'TPCFE09' which had these features as the first version of TPCFE
ASIC. We have confirmed that TPCFE09 operated almost as we expected and can be used for
the LXe-TPC. We expect TPCFE ASIC will be used in many applications, and now is planned
to use for experiment of the neutron-TPC in future.
1. Introduction
Aims to apply to a new type of PET (Positron-emission tomography), the liquid xenon TPC
(LXe-TPC) has been developed. The LXe-TPC is a radiation detector that integrates a time
projection chamber using liquid xenon and a fast response scintillation detector. Operating of
the LXe-TPC is as shown in Fig.1. The LXe-TPC can use the two signals of ionize electrons
and scintillation light. Using these signals, the LXe-TPC can decide the three-dimensional
position of the incident radiation at high spatial resolution and high frequency. In addition, using
liquid xenon which has a high density, detection efficiency of a particle which does not have a
charge is higher than the TPC that is using a gas. By these features, we expect the LXe-TPC can
solve the problem of PET that is the low spatial resolution and low frequency.
82
In the development of LXe-TPC, we have created the ASIC for the front-end circuit of PAD
with Open-i . In order to realize high spatial resolution, the ASIC must have features of
Multi-Channel read out, Low temperature operation and Low power. We have created of
TPCFE ASIC, which has these features for the LXe-TPC. We have confirmed that TPCFE09,
the first version of TPCFE ASIC.
83
< 5fC
-25fC25fC
Gain
>10 V/pC
Peaking time
1 usec, 0.5usec
Number of channel
16channel /chip
Power(Power supply)
Ambient Temperature
165K
84
85
Target value
-25fC25fC
-45fC45fC
Gain
>10 V/pC
< 13V/pC
ENC = 650electron
ENC=1600 electron
Peaking time
1usec, 0.5usec
Number of channel
16 channel/chip
16 channel/chip
Power(Power supply)
8.5mW/channel
TPCFE09 had a feature of Low temperature operation (165K), Multi-Channel read out
(16channel /chip), Low Power (8.5mW/channel, 136mW/chip at 165K). In addition,
TPCFE09 achieved target value of front-end circuit of PAD without noise level. But the noise
level (ENC is 1600 electron at 165K) can be detected the signal of PAD. When 5fC Input,
signal-to-noise ratio is 25dB. Looking at Fig 5, noise is considerably smaller than output signal.
Problems seen in the TPCFE09 was variation in baseline (s300mV) in every channel.
Variation in baseline mainly occurred in Shaping Amp block. Next version of TPCFE09, we
86
need to restructure circuit of Shaping Amp block for prevention of variation in baseline.
Although TPCFE09 had the problem, TPCFE09 can be used for front-end circuit of PAD.
Fig6: Making the next version of TPCFE ASIC (Shaping Amp block)
6. CONCLUSION
In the development of the LXe-TPC, we have created the ASIC for the front-end circuit of
PAD. The name of ASIC is TPCFE ASIC, now we created TPCFE09 which is the first version
of TPCFE ASIC. TPCFE09 has a feature of Multi-Channel read out (16channel/chip), Low
temperature operation(165K) and Low power(8.5mW/channel at 165K). Variation in
baseline and noise level seen in the TPCFE09 not achieved our estimation, but we confirmed
that TPCFE09 can be used for front-end circuit of PAD. Next version of TPCFE09, aims to
improve these problems. In the future, we will use TPCFE ASIC for PAD of the LXe-TPC. In
addition, TPCFE ASIC can be widespread application of experiment. Our TPCFE ASIC is
currently planned to use for experiment of the TPC for neutron detection.
ACKNOWLEDGMENT
We appreciate staffs of KEK Detector Technology Project (KEKDTP), Open-i , KEK
Neutron Science Division (KENS) and KEK Meson Science Laboratory (KEK-MSL).
Open-i : Open Source Consortium of Instrumentation, http://openit.kek.jp/
87
REFERENCES
[1] Behzad Razavi, Design of Analog CMOS Integrated, McGraw-Hill Companies (2003)
[2] Glenn F Knoll, Radiation Detection and Measurement, THE NIKKAN KOGYO SHIMBUN
(2001)
[3] M.tanaka, T.Fusayasu, H.Ikeda et al., reference, Advanced Electronics DAQ seminar, 2011
[4] Daisuke kaneko, Master thesis, The University of TOKYO, 2009
[5] T.Higashi, Poster ASIC design and fabrication for front end electronics of liquid xenon
TPC, Tukuba (2009)
[6]T.Tauchi et al., reference "Liquid Xenon TPC for a gamma detector (LXeTPC), 2010
Corresponding author
(e-mail) salvendeneu@gmail.com
88
89
Diamond is a wide band gap semiconductor that is a promising candidate for use in radiation
sensors(12),(13). A high signal-to-noise (S/N) ratio is expected from a diamond detector because of its high
charge collection efficiency, good thermal stability, and excellent radiation hardness(14), (15).A number of
studies have been reported on the development of diamond radiation detectors with thickness ranging from
about 70 m to 400 m(16). Several of them have energy resolution better than 1% and transient time
response of several nanoseconds.
The use of thick diamond detectors as an alternative beam-extraction window has also been explored
for analyzing single hits of high energy ions. A diamond film of about 10 m in thickness is resistant to
mechanical stress. In principle, a thick diamond film can be employed as a transmission radiation sensor.
However, it is not easy to make a high-purity diamond film with electrical properties sufficient for
spectroscopy. Considerable effort has been devoted to developing a thin-film diamond detector for
spectroscopic applications(17). Recent advances in the fabrication of CVD diamond have made available
highly pure CVD diamond films of less than 50 m in thickness(18). In light of these developments, a
transmission high-energy particle detector made of a CVD diamond thin film might be feasible to
incorporate into the design of the beam extraction window of a cyclotron microbeam line in order to
measure the irradiation timing and energy of a single ion when it is immediately above the irradiation
target on the same axis.
In the present research, transmission particle detectors using thin-film diamond were evaluated for
the detection of single hits of high-energy (MeV) ions. This paper describes the preparation of two types of
CVD diamond film detectors. One type used a polycrystalline (PC) diamond film; the other type uses a
single-crystalline (SC) CVD diamond film. Prototype detectors of the two types were fabricated with
ohmic contact electrodes. The detectors were then irradiated with alpha particles and MeV ionized
particles. To obtain the energy response and timing response of the detectors to the ionized particles, single
ion hit experiments was conducted on microbeam lines of tandem accelerators at the Takasaki Advanced
Radiation Research Institute (TARRI), JAEA, and at the Rudjer Boskovic Institute (RBI) in Croatia. In
experiments, the detector response to the single ions was examined through the energy spectrum and
transient pulse signals.
90
A Ti/Pt/Au multilayer of 50 nm in thickness was evaporated onto each side of the SC and PC
detectors to form ohmic contact electrodes. The measured leakage current from the SC detector was less
than 2 10-11 A under bias up to 50 V; that from the PC detector was less than 2 10-11 A under bias up
to 50 V and less than 1 10-10 A under bias up to 100 V.
91
processing system. The IBIC image was constructed by using the two-dimensional position control data
for vertical and horizontal microbeam scanning.
The detectors were also subjected to single ion hits from a 15 MeV O4+ microbeam. Transient current
induced by the single-ion hits of around 10 to 100 cps were measured by using the TIBIC data collection
system at JAEA Takasaki (20). The detector assembly was directly connected to a 3 GHz or 15 GHz digital
storage oscilloscope (DSO; Tektronix, TDS694C) via a high-voltage bias tee, and the transient current
signals were measured in real time by means of a digital signal processing system based on LabVIEW
(National Instruments). Each single-hit event triggers a fast beam switch shutter in the TIBIC collection
system to kick the beam off the axis to the detectors. TIBIC analyzes various characteristics of transient
current in the diamond film.
3. RESULTS AND DISCUSSION
3.1 Response of CVD Diamond Film Detectors to Alpha Particles
Figure 2 shows examples of pulse height spectra for alpha particle from 241Am.Bias voltage of + 50 V
was applied to the both the PC and SC detectors, and in the 5 min measurement, about 5500 counts were
obtained for each detector. In the spectrum, a peak with 2% full width at half-maximum (FWHM) was
obtained by using the SC detector. In contrast, the pulse height distribution for the PC detector does not
correspond to an energy spectrum with a significant peak. The shape of the spectra and count rates were
not significantly different between the SC and PC detectors during the irradiation of alpha particles from
241
from the total charge induced by the charged particle in the diamond film, comparing it with the charge
obtained from the voltage of pulse signals multiplied by the input capacity of the charge sensitive amplifier.
Figure 3 shows pulse height spectra of the SC detector under bias voltages of + 5 V, + 30 V, and + 50 V.
The shape of the spectrum broadened toward the lower channels when the applied voltage was decreased.
This result suggests that the CCE is not perfectly uniform at lower bias voltage and that there might be
some defects distributed inside or on the surface of the SC diamond. Figure 4 show the relations of bias
voltage with CCE and alpha particle count rate for the SC detector. The results show that the detector with
SC diamond is applicable to the analysis of single ions in spectroscopic applications when the bias voltage
is greater than 10 V. The polarity of the bias voltage did not have a significant effect. These results suggest
that the SC detector functioned well in particle detection.
92
Room Temperature(300K)
FWHM:1.8%
SC Diamond
400
300
100
PC Diamond
200
50
100
0
0
256
512
768
-particles
Signal Counts of 241Am -particles
from SC Diamond [counts/channel]
-particles
Signal Counts of 241Am -particles
from PC Diamond [counts/channel]
150
0
1024
Fig. 2 Pulse height distributions of 5.5 MeV alpha particles from 241Am standard radiation source for SC
and PC CVD diamond film detectors.
Fig. 3 Energy spectra of 5.5 MeV alpha particles from 241Am standard radiation source for SC CVD
diamond film detector under different bias voltages.
93
1.5
1.0
0.5
0
0
10
20
30
40
50
7500
5000
Negative Bias Voltage
Positive Bias Voltage
2500
10
20
30
40
50
Fig. 4 Relations of applied bias voltage with charge collection efficiency (top) and count rate of 5.5 MeV
alpha particles from 241Am standard source, as measured with SC CVD diamond film detector.
3.2 Uniformity of CCE during Microbeam Irradiation
The two CVD diamond film detectors were irradiated in an actual microbeam environment.
Microbeams of 1~4.5 MeV H+ generated by a tandem accelerator were focused onto the detectors. The
detector response signal was observed by the IBIC system. Figure 5 shows the pulse height spectra of the
4.5 MeV H+ ion microbeam, as observed with the SC and PC detectors.
The peak in the pulse height spectrum from the SC detector remained steady at 3% FWHM, whereas
that from the PC detector widely fluctuated during measurements. A pulse signal from the PC detector
94
was observed in the start of irradiation, but the pulse height decreased sharply as increasing irradiation time.
Ultimately, the output pulse from the PC detector could not be separated from background noise during
microbeam irradiation (4.5 MeV H+) at 100 cps in a 50 50 m area. The differences between the results
for 5.5 MeV alpha particles (5-10 particles/(mm2 s)) and 4.5 MeV H+ microbeam (104 particles/(mm2 s))
appeared attributable to the difference in fluence. After decreasing with irradiation time, the PC detector
Room
Room Temperature(300K)
Temperature(300K)
-particles
Signal Counts of 241Am -particles
from PC Diamond [counts/channel]
150
400
FWHM:2.7%
SC Diamond
300
PC Diamond
(Gain2)
100
200
50
100
0
0
256
512
768
-particles
Signal Counts of 241Am -particles
from SC Diamond [counts/channel]
pulse signal was easily recovered by briefly inverting the polarity of the bias voltage.
0
1024
Fig. 5 Pulse height distributions of 4.5 MeV proton microbeam irradiation from PC CVD diamond film
detector (left) and SC CVD diamond film detector (right).
Figure 6 show the IBIC images of detector response. The images were generated in relation to the
position control data for horizontal (X) and vertical (Y) beam scanning. The IBIC intensity is indicated in
the color bar according to the maximum voltage of the charge signals. The IBIC image from the SC CVD
diamond film showed high, uniform intensity of pulse signals in a 50 m 50 m area. In contrast, the
IBIC image from the PC CVD diamond film had a granular distribution with isolated areas of relatively
high intensity surrounded by areas of low intensity. This indicates the presence of electron-trapping
boundaries in the PC CVD diamond, which cause the decrease in CCE when the ion microbeam is
irradiated onto a very small area of the target. This heterogeneity in the IBIC image disappeared upon
inversion of the bias voltage but quickly reappeared shortly after microbeam irradiation was resumed.
95
50 m
Intensity of IBIC
Max
50 m
50 m
50 m
Min.
Fig. 6 IBIC images of PC CVD diamond film (left) and SC CVD diamond film (right) for a 4.5 MeV
proton microbeam. The beam scanning area was 50 50 m.
IBIC intensity was normalized by the maximum value of the corresponding pulse height distributions.
3.3 Transient Current Observation during Single Ion Hit
The timing response of the CVD diamond film detectors was examined through experiments measuring
the transient current signals recorded by the TIBIC system. A microbeam of 15 MeV O4+ was irradiated
onto the SC CVD diamond film detector at 10 to 100 cps. Figure 7 shows an example of the pulse
response of transient current signals induced in the SC CVD diamond. The transient current was directly
observed by both 3 GHz and 15 GHz DSO connected via a bias tee. Pulse signals of transient current were
obtained at each single ion hit by fast beam switching. A pulse width of around 5 ns was observed for each
single ion hit. Although the detector is large, the transient current signals had a decay time constant of the
order of nanoseconds. Thus, it was confirmed that transient current signals of the CVD diamond film
detector had sufficient time resolution to trigger the signal processing in single-ion-hit irradiation
experiments with the maximum scanning area of the beam extraction window.
In addition to pulse width, many TIBIC parameters including the peak voltage of transient current and
the integral quantity of ion-beam-induced charge are available. In regard to the transient changes in the
detector with the intense irradiation onto the film, such parameters will be effective for evaluating the
material properties of diamond and its radiation resistance under ion microbeam irradiation. Using the
TIBIC system, experiments are ongoing to evaluate the radiation hardness and tolerance of the developed
CVD diamond film detector.
96
10
0
-5
-10
15 GHz
3 GHz
-15
-20
-25
-30
10
15
20
25
time t[ns]
Fig. 7 Pulse shape of TIBIC due to single ion hit of 15MeV O4+.
4. CONCLUSION
Transmission particle detectors based on CVD diamond film were prepared for the detection of single
ion hits of high-energy (MeV) ion microbeams. Two CVD diamond detectors of 50 m in thickness were
characterized by using a standard alpha particle source and ion microbeam facilities. Pulse signals induced
by the irradiation of alpha particles and a single hit of a high-energy ions were successfully obtained by
using the SC and PC CVD diamond film detectors under bias voltage up to 50 V.
The typical responses of the PC and SC detectors to the microbeam irradiation were first examined. The
SC CVD diamond detector gave a uniform response under irradiation of a 4.5 MeV proton microbeam
and in IBIC analysis. The PC detector exhibited completely different responses to alpha particles and
microbeam irradiation. In contrast, the SC detector showed good characteristics with maximum energy
resolution of around 2% in the pulse height spectrum of 241Am alpha particles at room temperature. In
addition, the transient current response of the SC detector was evaluated under microbeam irradiation on
the TIBIC analysis system. Single hits of 15 MeV O4+ produced a good signal. The pulse signal showed a
decay time of around 5 ns, suggesting the applicability of the SC CVD diamond film detector to the
observation of single ion hits under higher-energy microbeam irradiation.
ACKNOWLEDGEMENT
This work was supported under the program "Development of high-energy ion microbeam technology
for novel applications of diamond" by the Strategic Japanese-Croatian Cooperative Program on Materials
Science of Japan Science and Technology Agency (JST) and the Ministry of Science, Education and
Sports of the Republic of Croatia (MSES). The authors thank Dr. Naoya Iwamoto for technical assistance
during TIBIC analysis experiments.
97
REFERENCES
98
1. INTRODUCTION
Dense gaseous and/or liquid xenon have many features suitable for X/
-ray and particle detectors.
Better understanding of basic properties of xenon as a radiation detection medium has led to new and
highly functional detectors, which are being applied to various fields(1).
For further developments of dense xenon detectors, more information on responses to radiation
and/or electron swarm parameters in dense xenon is still required. One of them is the first Townsend
ionization coefficient, which is defined as the average number of electron-ion pairs per unit length
generated by a primary electron through the electron-impact ionization collisions with atoms. This is the
basic quantity characterizing the charge amplification process, which occurs on the basis of the electron
avalanche under sufficiently high electric field.
The charge amplification has been studied in gaseous xenon for a long time and applied for
proportional counters, multi-wire proportional counters, gas electron multipliers, and so on. In particular,
proportional counters filled with xenon of near atmospheric pressure have been widely used for X-ray
detection. In that case, small amount of some kinds of quenching gases are often doped into xenon for the
purpose of stable operation at high charge gains. However, there are only a few cases that these detectors
have been operated with high-pressure xenon. Therefore, charge amplification properties such as the first
Townsend ionization coefficient in high-pressure xenon has not been clarified yet.
In this study, we made further investigation of the charge amplification in high-pressure xenon using
a cylindrical proportional counter. We demonstrate the charge amplification characteristics such as the first
Townsend ionization coefficient and the energy resolution at xenon pressures up to 3.0 MPa.
2. EXPERIMENTAL METHOD
99
2.1 Apparatus
A cylindrical proportional counter as shown in figure 1 was used for this investigation. It was made
of stainless steel tube with an internal diameter of 30 mm, wall thickness of 5 mm, length of 363 mm, and
no entrance windows for radiation in order to be tight enough in high pressure. A gold-plated tungsten
wire was used as an anode, which was soldered on feedthroughs welded at the both ends of the
proportional counter. The diameter of the anode wire was measured to be 10.3 0.16 m using a scanning
electron microscope. The wall of the proportional counter was grounded as a cathode. A radioactive
source of 133Ba was put on the wall of the proportional counter for this study.
The proportional counter and gas filling system were evacuated during baking out at a temperature of
120 C for more than 100 hours. Then, the proportional counter was filled with xenon gas in the pressure
range of 0.5-3.0 MPa, which was measured with a Bourdon gauge. All measurements were carried out at
room temperature of 298 K. Xenon gas was purified continuously for 72 hours in a tower of
barium-titanium getters(2) kept at 600 C before filling. The purity of xenon was checked by the lifetime of
electrons(3). The lifetime was measured using a parallel plate ionization chamber, in which an -particle
source of 241Am was installed. The lifetime was determined by analyzing the shape of pulses induced by
drifting electrons from a charge-sensitive preamplifier. In this investigation, the electron lifetime was
estimated to be longer than 1 ms at 3.0 MPa. As the maximum drift time of electrons in the proportional
counter was estimated to be about 20 s, the electron lifetime was sufficiently long for our measurements.
Thus, electron attachments due to electronegative impurities were negligible.
Signals from the anode of the proportional counter were routed through a charge-sensitive
preamplifier with a decay time of 1 ms, to a shaping amplifier (CLEAR PULSE 4417). The shaping time
of the amplifier was set at 5 s. The output signals from the shaping amplifier were digitized using a
multi-channel analyzer (MCA) (Labo 2100C/MCA). The induced charge at the anode was calibrated with
pulses of known amplitude from a precision pulse generator (ORTEC 419).
,
(1)
100
where a is the anode radius, rc the critical radius at which amplification starts, and r the radial distance
from the central axis. The first Townsend ionization coefficient has been regarded in many studies as a
function of the reduced electric field S, which is defined as the electric field normalized by the number
density of gas molecules E/N. The reduced electric field in cylindrical proportional counters is given by
(2)
where Va is the anode voltage and b the cathode radius. Then the charge amplification factor can be
derived as a function of the reduced electric field from equations (1) and (2)(4),
(3)
where Sa and Sc are the values of S at the anode surface and critical radius, respectively.
The charge amplification factor was determined by the pulse-matching method in this study. The
relations between induced charge and MCA channel can be obtained using the pulses of known amplitude
from the pulse generator. The charge amplification factor can be derived by
(4)
where Q is the total induced charge, e the elementary charge, E
the gamma-ray energy, and W the
W-value, which is defined as the average energy expended per electron-ion pair. The W-value was
assumed to be 21.9 eV(5). When the charge amplification factor is determined by the pulse-maching
method, it needs to be taken into account the difference of pulse rise between signals from the proportional
counter and pulses from the pulse generator. The attenuation of pulse height from cylindrical proportional
counters due to ballistic deficit through pulse-shaping networks has been studied(6)-(8). The total induced
charge corresponding to MCA channel m is obtained by(9)
(5)
where C0 is the input capacitance of the charge-sensitive amplifier from the pulse generator, V0 the
amplitude of pulses from the pulse generator, and m0 the corresponding MCA peak channel. The factors
and 0 are the attenuation coefficients of pulses from the proportinal counter and from the pulse generator,
respectively. The rise time of pulses from cylindrical proportional counters mainly depends on the motion
of positive ions, thus the positive ion mobility is required to obtain . The ion mobility can be expressed
as
(6)
where 0(S) is the reduced mobility corresponding to the standard number density N0 of 2.687 1019 cm-3.
We used the experimental data of 0 for Xe2+/Xe3+ ions in xenon(10) because molecular ions of Xe2+ and
Xe3+ would be instantly formed in xenon from Xe+ ions above atmospheric pressure(11). Although the
mobility of xenon ions has not been studied in high-pressure xenon, we assumed that the relation of
equation (6) is valid in the investigated pressure range.
Although many studies have been carried out experimentally and theoretically, the dependence of
101
on S is still unclear. In this investigation, we used the original Townsend formula as follows,
(7)
where A and B are constants. It was suggested(12) that equation (7) should be used as a standard description
of charge amplification in cylindrical proportional counters. The following expression is derived from
equations (3) and (7),
.
(8)
The assumption of is usually valid. The constants A and B can be derived from data of the
charge amplification factor and equation (8). Then, the first Townsend ionization coefficient is obtained
from equation (7) using those constants.
The original Townsend formula was obtained under assumption of uniform electric field. The
constants A and B in this formula depend only on the gaseous species. However, in non-uniform electric
field conditions such as in cylindrical proportional counters, it was pointed out that the behavior of the
ionization coefficient is different from uniform electric field conditions(13)-(15). In this investigation, the
constants A and B were assumed to be dependent on the pressure for non-uniform electric field
conditions(16).
3. RESULTS AND DISCUSSION
Figure 2 shows a typical pulse-height distribution for 133Ba at a pressure of 1.5 MPa and an anode
voltage of 3.080 kV. The spectral features include the 81.0-keV photopeak, the X-ray escape peak ( 51.3
keV), and Compton continuum mainly due to
rays at an energy of 356 keV emitted by 133Ba. The charge
amplification factor and energy resolution were obtained by fitting the 81.0-keV photopeak to a Gaussian
on a linear background.
133
102
Figure 3: Charge amplification factor M as a function of anode voltage Va in the xenon pressure range
of 0.5-3.0 MPa. The error bars of M and Va are smaller than the symbol size.
The first Townsend ionization coefficient was determined from the obtained charge amplification
factor. Figure 4 shows the reduced first Townsend ionization coefficient /N as a function of reduced
electric field. The uncertainty of S was derived from the errors on anode voltage, anode and cathode radii,
and gaseous density. The uncertainty of /N was derived from the errors on S, constants A and B. The
relatively large uncertainties particularly at 0.5 MPa are mainly due to the error on gaseous density, which
is derived from the error on pressure measurements.
In figure 4, the experimental results by Kruithof(17) and by Sakurai et al.(18) are also shown. The
103
experiment by Kruithof was carried out in a parallel plate chamber in the pressure range of 0.28-165 Torr,
and that by Sakurai et al. was carried out in a cylindrical proportional counter in the pressure range of 1-10
atm. Compared with the previous studies, our results of /N are higher than those by Kruithof and
consistent with those by Sakurai et al. below 1.0 MPa. One possible reason for the difference may be
attributed to the purity of xenon gas using in measurements. We used purified xenon gas as described in
section 2.1 and Sakurai et al. used pure xenon gas less than 1 ppm of contamination, while Kruithof used
xenon gas of up to 0.1 % of contamination. Lower values of /N by Kruithof could be caused by the
electronegative impurities, which cause the signal loss due to electron attachments.
As clearly shown in figure 4, the charge amplification at higher pressure occurs in lower reduced
electric field region. It has been observed that the ionization coefficient is scaled in the reduced electric
field for pressure between 1 and 10 atm(18). However, our results of /N seem to violate S scaling behavior.
Our results change with increasing pressure approximately above 2.0 MPa. In that pressure region, /N
values are larger than the extrapolated values from lower pressures.
Figure 4: Reduced first Townsend ionization coefficient /N as a function of reduced electric field S in
the xenon pressure range of 0.5-3.0 MPa. Previous experimental results by Kruithof(17) and Sakurai et
al.(18) are also shown. Note that S equals Sa for our results and results by Sakurai et al.
3.2 Energy resolution
Figure 5 shows the energy resolution of 81.0-keV photopeak as a function of charge amplification
104
factor. The energy resolution was determined from the full width at half maximum (FWHM) of the
pulse-height distribution of the photopeak.
Figure 5: Energy resolution as a function of charge amplification factor M in the xenon pressure range
of 0.5-3.0 MPa.
In figure 5, the similar behavior in the energy resolution is observed at each pressure with increasing
the charge amplification factor. Namely, the energy resolution becomes better at lower M region, is
approximately constant at intermediate M region, and becomes worse at higher M region. The
deterioration of the energy resolution at lower M region is due to the electronic noise and the fluctuation in
the primary electron-ion recombination. The contribution of the former effect is approximately constant,
and that of the latter effect increases with increasing pressure because the proportional counter with higher
gas pressures is operated at weaker reduced electric field. The recombination of a certain amount of
primary electron-ion pairs particularly generated near the cathode causes eventual fluctuation of induced
charge at the anode. On the other hand, the deterioration of the energy resolution at higher M region is due
to the space charge of positive xenon ions and the large number of photons emitted by excited xenon
molecules. The latter effect is presumably dominant because the larger charge amplification factor has
been achieved with superior energy resolution when suitable species and suitable amount of quenching
gases are doped into xenon of near atmospheric pressure(19), (20).
The pressure dependence of the energy resolution has been measured by several groups(18), (21)-(24) and
105
suggested several causes of deterioration such as electronegative impurities, non-uniformity of anode wires,
and varying the reduced electric field at which proportional counters are operated. The impurity effect was
avoided in our measurements as described in section 2.1. The non-uniformity of an anode wire results in
fluctuations of the charge amplification factor because S depends on a as equation (2). Additionally, the
size of electron swarms reached at amplification region becomes smaller with increasing pressure. This
results from decreases of the range of primary photoelectrons and the transverse diffusion of electrons.
Therefore, the energy resolution would be more sensitive to the anode non-uniformity at higher pressures
because an electron swarm is collected at the narrow area of the anode wire.
It has been pointed out in several works(18), (22), (23) that the excitation collisions of electrons with atoms
would increase at high pressure, which arises from the fact that the working reduced electric field becomes
lower in high-pressure proportional counters. Several processes are possible to occur subsequent to the
excitation collisions such as photon-emitted de-excitation, Penning ionization between excited atoms
and/or excited molecules, and the stepwise ionization, which is the electron-impact ionization with excited
atoms or excited molecules. The Penning ionization includes the associative ionization known as the
Hornbeck-Molnar effect(25). The deterioration of the energy resolution with increasing pressure would be
attributed in part to fluctuation in the above complicated processes occurring during the charge
amplification.
Compared with xenon-based gas mixtures(24), better energy resolutions were obtained for pure xenon
at high-pressure region. The energy resolution at 0.5 MPa was reported to be 9.2% at 59.5 keV for Xe +
CH4 (98:2) gas mixtures. In the case of higher pressures, energy resolutions were reported to be 13.5% at
81.0 keV for Xe + Ar + CH4 (90:8:2) gas mixtures at 1.5 MPa, and 28.2% at 59.5 keV for Xe + C2H4
(98:2) gas mixtures at 2.5 MPa. These energy resolutions are worse than that in pure xenon obtained in this
investigation even taking into account the difference of
-ray energy. Therefore, pure xenon would be
suitable for the charge amplification at high-pressure region, particularly in the charge amplification factors
between 102 and 103.
4. SUMMARY
The characteristics of charge amplification in high-pressure xenon were studied using a cylindrical
proportional counter. The investigated pressure range was from 0.5 to 3.0 MPa, which is higher than
previous studies.
The charge amplification factor was obtained between 15 and 2.4 103 for investigated pressure. The
first Townsend ionization coefficient was determined from the measured charge amplification factors. As
a result, the ratio of /N at higher pressures seems to increase for the same reduced electric field, which
was observed approximately above 2.0 MPa. Consequently, /N cannot apparently be represented as a
function of reduced electric field only.
It was found that a proportional counter is worked at weaker reduced electric field at higher pressures.
106
This fact results in the complexity of electron avalanche processes and the deterioration of the energy
resolution. The non-uniformity of the anode wire would also be a cause of fluctuation of charge
amplification particularly at high pressures. The energy resolution was observed to become worse with
increasing pressure, which is attributed to above reasons. Nevertheless, the energy resolution in pure xenon
is found to be superior to that in xenon-based gas mixtures at high pressure and several hundreds of the
charge amplification factor.
Further considerations about the first Townsend ionization coefficient in high-pressure xenon will be
made in the future, together with the charge amplification in liquid xenon. It may also be interesting to
compare the behavior of charge amplification in cylindrical proportional counters at high pressure with
parallel plate chambers at low pressure at the same reduced electric field. That will be important for better
understanding of the mechanism of charge amplification in cylindrical geometry, or non-uniform electric
field conditions.
REFERENCES
(1) E. Aprile, A. E. Bolotnikov, A. I. Bolozdynya, and T. Doke, Noble Gas Detectors, Wiley-VCH,
Weinheim, 2006, p. 277.
(2) M. Miyajima, K. Masuda, A. Hitachi, T. Doke, T. Takahashi, S. Konno, T. Hamada, S. Kubota, A.
Nakamoto, and E. Shibamura, Nucl. Instrum. Methods 134 (1976) 403.
(3) A. Bolotnikov and B. Ramsey, Nucl. Instrum. Methods Phys. Res. Sect. A 383 (1996) 619.
(4) T. Aoyama, Nucl. Instrum. Methods Phys. Res., Sect. A 234 (1985) 125.
(5) F. I. G. M. Borges and C. A. N. Conde, Nucl. Instrum. Methods Phys. Res. Sect. A 381 (1996) 91.
(6) E. Mathieson and M. W. Charles, Nucl. Instrum. Methods 72 (1969) 155.
(7) R. Gott and M. W. Charles, Nucl. Instrum. Methods 72 (1969) 157.
(8) B. Breyer and M. Cimerman, Nucl. Instrum. Methods 92 (1971) 19.
(9) P. J. B. M. Rachinhas, T. H. V. T. Dias, A. D. Stauffer, F. P. Santos, and C. A. N. Conde, IEEE Trans.
Nucl. Sci. NS-43 (1996) 2399.
(10) P. H. Larsen and M. T. Elford, J. Phys. B: At. Mol. Phys. 19 (1986) 449.
(11) H. Helm, Phys. Rev. A 14 (1976) 680.
(12) A. Zastawny, Nucl. Instrum. Methods Phys. Res., Sect. A 385 (1997) 239.
(13) H. Date, K. Kondo, M. Shimozuma, and H. Tagashira, Nucl. Instrum. Methods Phys. Res. A 451
(2000) 588.
(14) H. Date and M. Shimozuma, Phys. Rev. E 64 (2001) 066410.
(15) K. Mitev, P. Sgur, A. Alkaa, M. C. Bordage, C. Furstoss, C. Khamphan, L. de Nardo, V. Conte, and
P. Colautti, Nucl. Instrum. Methods Phys. Res. A 538 (2005) 672.
(16) I. K. Broni
and B. Grosswendt, J. Appl. Phys. 88 (2000) 6192.
107
108
1. INTRODUCTION
Position sensitive neutron detectors are used for neutron scattering, neutron imaging, and neutron
radiography. Developments in neutron detectors are mainly focused on spatial resolution and high
counting rate in these fields. After the Fukushima nuclear power plant disaster, many simple radiation
counters are now marketed in Japan and are useful for estimating the radiation level. However, no simple,
equivalent device exists for neutron measurements.
We have previously developed position sensitive neutron detectors using a ZnS/6LiF scintillator(1).
These detectors are used as high spatial resolution detectors for focusing SANS experiments(2) and beam
profile monitors(3,4). One of the authors of this paper, Satoh, along with his co-workers, has developed a
neutron network (NEUNET) system(5) as a readout system for 3He position sensitive detectors (PSDs).
Many NEUNET systems are used for over half of the instruments(6,7) at the Materials and Life Science
Experimental Facility (MLF) in the Japan Proton Accelerator Research Complex (J-PARC). Thus, reliable
neutron counters that are also easy to handle under different experimental conditions are needed for
detector development.
In this work, we have developed simple neutron counters (LiM counter and HeM counter) with the
following features. In the LiM counter, a 6Li glass scintillator is employed as a neutron converter for a
large dynamic range. A silicon photon MPPC (multi-pixel photon counter)(8) detector is used separately for
photon counting, thus reducing the size of the device. In the HeM counter, a 3He neutron detector is
employed. Both counters employ a pulse-height analysis function for ensuring reliable data and display the
pulse-height distribution on a graphical liquid crystal display (G-LCD). The LiM counter can be used for
about 6 h using a battery, operating at 264 mA and 5.1 V.
109
2. LiM counter
2.1 Neutron sensor
The LiM counter is shown in Fig. 1.
It has a 128 64 pixel G-LCD in the center,
with adjustable resistors and switches
located around the center. The head of the
small board of the counter has a neutron
scintillator and an MPPC, wrapped with
black tape. The scintillator uses Bicron
GS20 6Li glass with dimensions 2.3 2.3
1 mm3. The MPPC, which has a 3 3 mm2
detection
area,
was
obtained
from
Li glass neutron scintillators are sometimes used as neutron detectors. Since light can travel
relatively far in glass, there is no thickness limit. For this reason, detection efficiency is usually higher than
that for ZnS/6LiF. For example, the absorption efficiency of 1 mm 6Li glass and 0.4 nm wavelength
neutrons is more than 95%. Gamma sensitivity is, however, higher than 3He or ZnS/6LiF and depends on
the thickness.
This LiM counter consists of a 6Li glass scintillator and an MPPC with small size and high counting
rate. Although the 6Li glass scintillator has gamma sensitivity, the LiM counter is able to distinguish a
neutron from the noise signal using a pulse height distribution function.
2.2 Design of electronics and its functions
A schematic diagram of the LiM counter is shown in Fig. 2. The 6Li glass scintillator is used as a
110
neutron detector and generates light by reacting to a neutron. The MPPC converts light to an electric signal.
The signal is amplified by the amplifier and converted to a pulse height value by the ADC1 (analog digital
converter).
T he a mplif ier
consists of a charge
sensitive amplifier and
a pu ls e s ha p ing
amplifier. In radiation
measurement, a charge
sensitive amplifier is
ver y effective for
reducing noise, because
it is an integration
circuit. However, the charge sensitive amplifier correspondingly has a low counting rate. A high counting
rate is achieved by adjusting the decay time constant to about 2 s. For reference, see the oscilloscope
waveforms in Fig. 5. The time constant of the pulse shaping amplifier is also set to 0.1 s.
The ADC1 obtains the pulse height with 2 timing signals from the comparators. One is a lower level
discrimination (LLD) signal, and the other is a peak search signal consisting of just a capacitor and a
resistor. Because the peak search circuit is weakened by high frequency noise, the ADC1 needs the LLD
signal for pulse height analysis.
Temperature compensation is required for the MPPC due to a change in gain with temperature. Since
gain is also controlled using supply voltage, we controlled the supply voltage using a DAC (digital analog
converter). The DAC compensates the voltage using a temperature sensing IC chip after approximately
every 5 s. The compensation coefficient is 56 mV/C, as reported by Hamamatsu photonics KK. The
supply voltage is measured by ADC2.
The whole system is controlled by a field
programmable gate array (FPGA). The FPGA is
XC6SLX9 of the Xilinx Spartan-6 series and has a
32 bit CPU. Because the CPU (MicroBlaze, Xilinx)
has a serial communication port and can be
programmed with the C language, the LiM counter
can easily communicate with a control PC.
However, most PCs do not have a serial
communication port, and so, the XPort(9) module
from Lantronix is used for data storage over a
TCP/IP network.
The FPGA, programmed by the VHDL language, is also used to accumulate the pulse height and
111
time analysis distribution functions, which requires high-speed processing. Distribution data are
accumulated in 512 word (2 kB) memories.
Pulse height analysis is the most important function of the LiM counter (Fig. 3). ADC121S1012(10)
(National Semiconductor) is used for converting signals into the pulse height distribution and has a 12 bit
ADC with a high speed serial interface consisting of 3 digital lines. Activating the chip select line (CS)
converts analog input that is held at this time to digital data. A clock line (SCLK) activates the digital
output (SDATA) for conversion. Therefore, the FPGA only reads the ADC immediately after the peak
search signal and obtains the pulse height value. The pulse height analysis of the FPGA is thus very simple
and suitable for the LiM counter. The maximum counting rate is less than 1 Mcps with the operation time
being 1 s per pulse read.
2.3 Details of electronic parts of the LiM counter
The circuitry of the LiM counter is shown in Fig. 4, consisting of a main board (128 79 mm2) and a
detection board (128 20 mm2).
When the mode switch is pushed for
more than 3 s, all data and the
display are cleared. When the LiM
counter is battery operated, a buzzer
sounds after approximately 5 min.
Pushing the switch stops the buzzer
for another 5 min. The power switch
is used to select the AC adapter or
battery
operation.
The
battery
112
The voltage adjustment changes the supply voltage of the MPPC. Automatic temperature correction
is accomplished by adjusting this voltage. When automatic temperature correction is being performed, a
* is displayed once in the upper right corner on the first line of the G-LCD approximately every 5 s. The
default correction coefficient is 56 mV/C as mentioned in 2.2, but it can be changed through the network
connection.
The T0 input is a trigger signal, indicating the start time of a pulsed neutron source. This input is used
for a time analysis distribution with 512 channels of 0.1 ms time width. Time analysis distribution data is
only accepted through the network connection. The analog output signal comes out from a charge
sensitive amplifier and a pulse shaping amplifier. The analog output can be connected to an oscilloscope or
another high input impedance
voltmeter for analyzing the raw
analog signal.
2.4 Data measurement using
Fig. 5(b) G-LCD graph and waveforms at 100 kcps
113
height is constant as long as the counting rate is lower than 10 kcps in Fig. 5(a). If the counting rate
increases, pulse height decreases in Fig.5(b) and Fig.5(c) due to the pileup effect of the charge sensitive
amplifier. However, this device can handle a test pulse of 500 kcps, as shown in Fig. 5(c).
2.5 Neutron measurements
Fig. 6 shows a screenshot of the LiM counter display measuring a 252Cf neutron checking source with
moderator. The horizontal axis is proportional to the pulse height distribution, and the vertical axis
expresses the neutron number, normalized by the maximum data value. The shoulder on the left side of the
graph corresponds to noise from gamma rays. The peak on the right side is attributed to neutrons. Neutron
data can easily be discriminated from gamma ray
noise by setting the LLD at the valley in the middle
of the graph.
The efficiency of the LiM counter was
obtained at Kyoto University Research Reactor
Institute (KUR). Using a 1 mm2 square neutron
beam for 100 s, the LiM counter registered 15,798
counts where as the 3He neutron tube had
respectively, at the elapsed time of 8 ms using the LiM counter. These rates were consistent with previous
data of the BL10.
2.6 Application to a HeM counter
A 3He gas neutron detector is indispensable for validation of neutron detector systems, due to its
reliability and stability in neutron measurement. Therefore, we developed the HeM counter with a 3He gas
detector. Its design is similar to that of the LiM counter. Fig. 8 shows the HeM counter. The 6974 V
voltage generator in Fig. 2 was replaced by a 10001800 V voltage generator.
Fig. 9 shows the pulse height distribution of the HeM counter using a 252Cf source. The large peak on
the right corresponds to the reaction energy of a neutron with 3He (pulse height 400, 765 keV). The
114
Fig. 8 HeM counter.
shoulder in the center (pulse height: 150350) is the distribution due to the wall effect in the gas tube. The
small peak on the far left corresponds to gamma radiation and electrical noise.
Fig. 10 shows individual waveforms of the ADC1 for pulse height distribution. The input of the
amplifier is fed an 80 mV rectangular wave through a 1 pF capacitance. Due to the slow response of the
He neutron gas detector, the charge sensitive and shaping amplifier time constants should be 50 s and 2
s, respectively. The pulse height analysis time is the same as that in Fig. 3 at about 1 s.
Generally, a 3He neutron gas detector is comparatively large, with a 1 inch diameter and 30 cm length.
of neutrons.
Moreover, as the detector efficiency is high, the natural backgroundof neutrons that are considered to be of
cosmic ray origin can be detected. Fig. 11 shows the pulse height distribution of the natural background of
neutrons, which is very similar to the plot of Fig. 9. Because the counting rate is 24 counts per minute
(cpm), a distribution similar to Fig. 9 is obtained in a few hours and can be used for the calibration of the
HeM counter.
3.
Summary
3.Summary
A simple neutron counter, the LiM counter, was developed for a large dynamic range, high
performance, and compactness. A 6Li glass scintillator and MPPC are the best choices for miniaturizing
115
the LiM counter. The graph on a G-LCD helps distinguish between neutron signal and noise. The pulse
height analysis is simply performed using a high speed serial ADC. The LiM counter was shown to be
useful for the estimation of beam line intensity.
The HeM counter was also developed and found to be useful for evaluating neutron detector systems.
The large size and high efficiency of the HeM counter enabled the measurement of the natural background
of neutrons.
Acknowledgements
We appreciate M. Hino, M. Kitaguchi, and other staff members at KUR for their help in the
irradiation test at E2 and CN3 in KUR. We also appreciate M. Harada, K. Oikawa, N. Yamada, and other
staff members at J-PARC for their help in the irradiation test at BL10 and BL16 in MLF, J-PARC. These
measurements were performed under the Common-Use Facility Program of KUR and J-PARC.
REFERENCES
(1) K. Hirota, et.al., Phys. Chem. Chem. Phys. 7(2005) 1836-1838.
(2) H. Iwase, H. Endo, M. Katagiri, M. Shibayama, J of Appl. Crys. 44 (2011) 558-568.
(3) T. Shinohara, et.al., Nucl. Instr. and Meth. Phys. Res., Sect. A 651 (2011) 121.
(4) T. Ino, et.al., Physica B 404 (2009) 2667-2669.
(5) S. Satoh, et al., Nucl. Instr. and Meth. Phys. Res., Sect. A 600 (2009) 103-106.
(6) K. Nakayoshi, et.al., Nucl. Instr. and Meth. Phys. Res., Sect. A 623(2010) 537-539.
(7) S. Yano, et.al., Nucl. Instr. and Meth. Phys. Res., Sect. A 654 (2011) 421.
(8) http://www.hamamatsu.com.
(9)
116
H. Tomita
Facility for Nuclear Materials, Nagoya University
Nagoya, 464-8063, Japan
M. Isobe
National Institute for Fusion Science
Oroshi-cho, Toki, 509-5292, Japan
1. INTRODUCTION
In deuterium plasmas of an fusion experimental reactor, neutrons around energy of 2.5
MeV (DD neutron) are emitted as accompanying products of DD fusion reaction. The
measurement of DD neutron emission profile plays an important role for diagnostics of energetic
deuterium ion in DD plasma because the neutrons are mainly produced by beam-plasma
interactions during neutral beam injection heating. The DD fusion reaction produces a tritium, so
the second reaction stands for deuterium-tritium and involves the combination of a deuterium
nucleus and a tritium nucleus to produce a helium-4 nucleus and a neutron with a kinetic energy
of approximately 14.1 MeV. The DT reaction is used more than the DD reaction because the
yield of the DT reaction is 50100 times higher than that of the DD reaction.
As a method of imaging for DD neutron and DT neutron, we propose a compact neutron
pinhole camera based on nuclear emulsion. Here, a nuclear emulsion used in early is a solid track
(e-mail)minato.haruna@h.mbox.nagoya-u.ac.jp
117
detector that records tracks of charged particle depositing energy in the emulsion within AgBr. A
nuclear emulsion have almost never been tried to be applied in engineered field because it has
had a lot of trouble to count the number of tracks by an operator. However it is possible to count
tracks automatically by progress of image processing technique, and possibility of new
engineered application of a nuclear emulsion is becoming widespread(1). We have reported that
spatial resolution was estimated to be about 10 cm in the case of combination with a collimator
made of Tungsten and a nuclear emulsion for DD neutron emission profile measurement.
However it is necessary to improve detection efficiency by several orders of magnitude(2). In this
paper, we show basic performance of fast neutron imaging with emulsion cloud chamber based
on Monte-Carlo simulation.
118
collimators and scintillation detectors because the emulsion has high spatial resolution and single
pinhole system covers for wide viewing angle.
4. Conclusion
We proposed neutron pinhole camera with nuclear cloud chamber for measurement of DD
and DT neutron emission profile. We estimated energy resolution and re-constructed 2D image
based on Monte-Carlo simulation with stacked emulsion. In the result, in the case of ten-stacked
119
film, the energy resolution was 29% (0.71/2.45) in FWHM and the spatial resolution was 9 cm.
We plan to demonstrate DD and DT neutron imaging using neutron pinhole camera with
emulsion cloud chamber by experiment for the next step.
REFERENCES
(1) T. Nakano: Proceedings of International Europhysics Conference on High Energy Physics, Budapest,
Hungary, 1218 July 2001.
(2) Y. Nomura et al., Plasma Fusion Research, in press.(2011)
(3) K.Morishima, Ionizing Radiation,37, No.3, (2011).
120
200
Count
collimater
Tungsten block
150
FWHM=0.71(MeV)
100
Count
100
50
0
0
2
3
Energy[MeV]
0
0
2
3
Energy[MeV]
one layer
ten layer
Count
2000
1000
0
0
2
3
Energy[Mev]
Fig. 5 The calculated energy spectrum of incident neutron which was compared one layer with ten layer
FWHM=9.0[cm]
150
Count
100
50
-4
-2
0
X[cm]
Fig. 6 Reconstructed image of neutron point source in projected plane at 4 m from center of collimator
121
INTRODUCTION
The dark matter is of great interest in the eld of particle astrophysics and nuclear physics.
Components of the universe have been studied by many cosmological observations. The
cosmological parameters were derived from the WMAP data combined with the distance
measurements from the type Ia supernovae and the baryon acoustic oscillations in the distribution of galaxies [1]. The density parameters of the baryon, the cold dark matter and
2
the cosmological constant are derived as b h2 = 0.02267+0.00058
0.00059 , c h = 0.1131 0.0034
and = 0.726 0.015, respectively, where, h = H0 /100 is Hubble constant [1]. The
dark matter distribution in the Universe has been studied by means of the micro lensing
observation. The three-dimensional distribution map was reported by COSMOS collaboration [2]. From many observations described above, cold dark matter is the promising
candidate of the main ingredient of the dark matter. WIMPs (Weakly Interacting Massive
Particles) are the most promising candidate of cold dark matter.
The nucleus-WIMPs interaction is dominated by the weak interaction [3]. The nucleusWIMPs interaction is classied into three types; Spin-independent elastic scattering (SI),
spin-dependent elastic scattering (SD) and inelastic excitation of nucleus (EX) [4]. The
processes for nucleus-WIMPs interaction are schematically illustrated in Fig.1. The properties of each interaction are described below.
1. Spin-independent elastic scattering (SI) is mainly scalar interaction. The cross section is proportional to the scalar matrix element gS2 = f 2 A2 , where A is the mass
number of the target nucleus. The cross section is enhanced for heavy nuclei, however, the cross section is reduced signicantly by small form factor.
2. Spin-dependent elastic scattering is mainly axial-vector interaction. The cross sec2 = C 2 2 J(J + 1).
tion is proportional to the nuclear spin matrix element gA
1
2
Corresponding author
Present address: SEIKO Instruments Inc.
122
(SI)
(SD)
gA
g s
(EX)
gA J
1
2
2 | A|M 1|A | ,
gM
123
2. In the case of the inelastic excitation of nucleus or the atomic electron, the low energy
-ray or X ray is emitted [4, 8]. If the -ray or X ray escapes from the segment and
it is detected in the neighboring segment (S2 in Fig.2). The inelastic signal makes
coincidence event. The signals S1 and S2 are observed simultaneously.
3. The background event is mainly due to Compton scattering of high energy -rays.
Since the scattered -ray is detected by another segments, the background events is
eciently identied and rejected.
EXPERIMENT
124
Figure 3: A module of the PICO-LON detector. Black ame was made of acrylic. A silver
square surface is an aluminum sheet glued on the ame to avoid humidity.
RESULTS
COUNTS
COUNTS
The pulse height spectrum was obtained by simply summing all the ADC values. The
energy spectra obtained by the rst module of PICO-LON test module are shown in
Fig.4. In the energy spectrum of 241 Am, a 59.5 keV -ray peak was tted by Gaussian
2200
2000
1800
3000
2500
1600
1400
2000
1200
1500
1000
800
1000
600
400
500
200
0
0
20
40
60
80
100
120
140
ENERGY(keV)
0
0
20
40
60
80
100
241 Am
120
140
ENERGY(keV)
source. (Right)The
function to calculate the energy resolution. Two peaks, which were close to each other,
were due to L-X ray of neptunium at around 17keV and K-X ray escape peak of iodine at
around 28keV. The energy threshold was as low as 2keV.
In the energy spectrum of 133 Ba, a low energy peak which was L-X ray of cesium was
clearly observed at around 4keV. The low energy threshold and the linearity of scintillation
response in the low energy region was conrmed by the present test measurement.
The good energy resolution was obtained as 25% at 60keV in FWHM. The energy
threshold was determined to be about 2 keV, which is the most important property for
the dark matter search.
The position information was derived from the ratio of PMT outputs. The three PMT
outputs on each side (north, east, south and west) were used for the position analysis by
125
EW
E+W
N S
,
N +S
(1)
(2)
where N ,E,S and W are the sums of three PMT outputs from northern, eastern, southern
and western sides, respectively. The result of position resolution for No1 module was shown
in Fig. 5. They are transformed to real scale. The position resolution of X-axis and Y-axis
were 30% and 34% in FWHM. The position resolution in the wider side (15cm15cm)
was as good as about 5 cm in FWHM.
DISCUSSION
Three thin and wide area NaI(Tl) scintillator was developed to search for the dark matter.
The energy resolution in FWHM, the energy threshold and the position resolution were
measured by low energy -rays and X rays. The energy resolution and the energy threshold
of thin NaI(Tl) scintillators were good enough to search for WIMPs dark matter.
126
Goal
< 25
<5
< 30
< 30
No1
25
2
30
34
No2
26
10
33
33
No3
24
2
31
34
The thin NaI(Tl) has showed that it should be applied for future dark matter search.
To enlarge the background reduction eciency, an active shield with 4 solid angle will
be applied. The KamLAND detector system is the good candidate to install PICOLON system. The expected sensitivity reaches to WIMPs interaction cross section of
105 106 pb.
ACKNOWLEDGMENT
The present work was supported by JSPS Grant-in-Aid for Scientic Research (C) 20540294.
References
[1] E. Komatsu et al, The Astrophysical Journal Suppl. Ser. 180(2009)330.
[2] R. Massey et al, Nature 445(2007)286.
[3] J.D.Lewin et al., Astroparticle Physics 6 (1996) 87.
[4] H.Ejiri, K.Fushimi and H.Ohsumi, Physics Letters B317 (1993) 14.
[5] K.Fushimi et al., Journal of Physical Society of Japan 75 (2006) 064201.
[6] K.Fushimi et al. Journal of Physical Society of Japan 74 (2005) 3117.
[7] K.Fushimi et al. Journal of Physics: Conference Series 120 (2008) 042024.
[8] H.Ejiri, Ch. C. Moustakids and J.D. Vergados, Physics Letters B634 (2006) 218.
127
Akira Yoshikawa
Institute of Multidisciplinary, Research for Advanced Materials, Tohoku University, 2-1-1,
Katahira, Aoba-ku, Sendai-shi, 980-8577, Japan
1. INTRODUCTION
Various inorganic crystals have been developed for high performance ray
scintillation detectors with the high sensitivity, the high light yield and the fast response. Yb3+containing compounds exhibit the charge transfer (CT) transition(1), (2), (3) which is promising
luminescence process for high performance scintillators. Since the CT luminescence of Yb3+ is
strongly affected by thermal quenching, it is necessary to characterize scintillation properties of
YAP:Yb at low temperatures. Several groups have reported scintillation properties of YAP:Yb
measured by photoluminescence technique in the temperature range from 10 K to room
temperature(4), (5) and the light yield of YAP:Yb for 662 keV rays at a temperature of 100 K(6).
In this work, scintillation response of the YAP:Yb crystal to 662 keV rays was
experimentally obtained by employing an avalanche photodiode (APD) in the temperature range
from 50 K to 175 K.
2. EXPERIMENTAL PROCEDURE
A 5 x 5 x 1 mm3 YAP:Yb30% crystal was used in this work. The crystal was
transparent and had no visible crack. Fig. 1 shows a circuit diagram for measurement of the
scintillation response of the YAP:Yb crystal to ray in low temperatures. In previous work, an
Corresponding author; (tel)092-802-3484, (fax) 092-802-3483, (e-mail) yasumune@kune2a.nucl.kyushu-u.ac.jp
128
Fig. 1
APD of Hamamatsu S8664 30K was demonstrated to be operated at low temperatures below
100 K(7). Since the Hamamatsu S8664 30K type APD has a circular sensitive area of 3 mm in
diameter which is smaller than the YAP:Yb crystal, the YAP:Yb crystal was coupled to a
HAMAMATSU S8664-8099(X) type APD with the sensitive area of 10 x 10 mm2 for detecting
scintillation photons. The HAMAMATSU S8664-8099(X) type APD was found to exhibit same
operational performance as the Hamamatsu S8664 30K type APD at low temperatures below
100 K. The APD with the YAP:Yb crystal was cooled on the copper holder attached to the cold
stage of the Gifford-McMahon refrigerator. The YAP:Yb crystal was irradiated with rays
emitted by 137Cs source placed in the refrigerator. The APD was connected to a charge-sensitive
preamplifier (CANBERRA 2003BT) placed outside of the refrigerator by a signal transmission
line consisted of manganin wires. Signal charges of the APD were converted into voltage signal
pulses by the charge-sensitive preamplifier. Fig. 2 shows examples of voltage signal output of
the charge sensitive preamplifier in irradiation of the YAP:Yb crystal with rays emitted by a
137Cs
source at a temperature of 50 K. As shown in Fig.2, the rise time constant of the voltage
129
Fig. 2
YAP:Yb crystal. Logic pulse generated by the CFD was fed into a gate generator that provides
the trigger for the MCA. The other pulse of the preamplifier was transformed into Gaussianshaped pulses by a shaping amplifier (ORTEC 572) and fed into a multichannel analyzer (MCA;
AMPTEK MCA8000A) for recording pulse hight distributions as a scintillation response of the
YAP:Yb crystal to rays.
For evaluating the scintillating light yield of the YAP:Yb crystal, response of the APD
to 5.9 keV X rays emitted by an 55Fe source was obtained by operating the APD with adjusting
the temperature, the reverse bias voltage and settings of the shaping amplifier to those in the
measurement of the scintillation response of the YAP:Yb crystal to 662 keV -rays emitted by
the
137Cs
source with using the circuit of Fig.1. In an obtained pulse height distribution, a
position of the full energy peak corresponds to the production of 1,640 electron-hole pairs in the
APD for the signal charge in the detection of 5.9 keV X-ray photon. The number of electron
hole pairs produced in the APD by scintillating photons generated in the YAP:Yb crystal by
absorbing 662 keV -rays was estimated by comparing a position of the photoelectric absorption
peak for 662 keV -rays in the the scintillation response of the YAP:Yb crystal with that of the
full energy peak of directly detected 5.9 keV X-rays from an
55Fe
scintillation light yield was calculated from the number of electronhole pairs and the quantum
efficiency of the APD.
130
(a)
(b)
Fig. 3
(a) Scintillation response of the YAP:Yb crystal to 662 keV -rays emitted by
the 137Cs source at a temperature of 50 K. (b) response of the APD to 5.9 keV
X rays emitted by an 55Fe source, which was obtained by operating the APD
with the same temperature, the reverse bias voltage and settings of the shaping
amplifier as those in the measurement of Fig. 3 (a).
Fig. 3 (a) shows an experimental result of the scintillation response of the YAP:Yb
crystal to 662 keV -rays emitted by the
137Cs
photoelectric absorption peak of 662 keV -rays is observed at around 1200 channels. The
photoelectric absorption peak of 662 keV -rays is expanded in the inset of Fig.3 (a). The full
width at half maximum of the peak is estimated to be 13%. Fig. 3 (b) shows a response of the
APD to 5.9 keV X rays emitted by an
55Fe
with the same temperature, the reverse bias voltage and settings of the shaping amplifier as
those in the measurement of Fig. 3 (a). The full energy peak of 5.9 keV X-rays appears at
around 400 channels. By comparing a position of the photoelectric absorption peak in Fig 3 (a)
with that of the full energy peak in Fig. 3 (b), the number of electron-hole pairs produced in the
APD is evaluated to be 4,920 for detecting scintillating photons of the YAP:Yb crystal
131
175
690 160
160
1490 280
150
2110 220
130
2570 280
115
3240 330
100
3430 320
77
4040 340
50
4170 370
generated by absorbing 662 keV -rays at 50 K. Table 1 summarizes experimental results of the
number of electron-hole pairs produced in the APD by detecting scintillating photons of the
YAP:Yb crystal generated by absorbing 662 keV -rays. Values of the number of electron-hole
pairs listed in Table 1 were obtained by averaging experimental results of two different
measurements with same conditions. Belogurov et al. reported that 3,000 electron-hole pairs
were produced in the large area APD by detecting scintillating photons of a 2 mm thick YAP:Yb
crystal generated by absorbing 1MeV -rays at a temperature of 100 K, while a value of 5181
Fig. 4
Relationship between the light yield of the YAP:Yb crystal and temperature.
132
4 CONCLUSION
4.
CUNCLUSION
Scintillation response of the YAP:Yb crystal to 662 keV rays was experimentally
obtained by employing the APD in the temperature range from 50 K to 175 K. The light yield of
the YAP:Yb crystal was found to increase with decreasing temperatures and reaches to 10,000
photons/MeV at 50 K.
REFERENCE
133
1. INTRODUCTION
The ionization chambers filled with rare gases have been used in wide field of scientific
research. Although the energy resolutions experimentally obtained in these ionization
chambers are practically good in many applications, they are still worse than those theoretically
estimated. Attempts have been made to explain the discrepancy but are not successful. In
this report, comparison is made between the energy resolution theoretically expected and that
experimentally obtained in rare-gas filled ionization chambers and some discussions are made
on this discrepancy.
(tel)03-3925-2045, (e-mail)shibamura-eido@spu.ac.jp
134
is Ni, the number of produced electron-ion pair follows Poisson distribution whose average is Ni.
In this case we obtain energy resolution RU as the upper limit by putting F = 1 in Eq. (1).
Table 1 summarizes the EP, W, F, R and RU for each liquid rare gas.
The energy resolution R is very good although it does not include the electronic noise and
other experimental factors. In particular, R in liquid xenon is comparable to that in germanium
semiconductor detector for gamma rays.
Table 1. Theoretically expected energy resolution R in FWHM and related quantities in liquid
rare gases.
135
Fig. 1.
field.
Collected charge for 976 keV electrons in liquid argon as a function of applied electric
Fig. 2. Noise-subtracted energy resolution for 976 keV electrons in liquid argon as a function
of applied electric field.
3.2. Experimental energy resolution in liquid xenon
Aprile et al.9) studied the energy resolution in liquid xenon ionization chamber. Fig. 3
shows the collected charge for gamma-rays with energy EP = 570 keV as a function of applied
electric field. The solid and broken line shows relation (2) and its modification fitted to the
experimental data9). Fig. 4 shows the noise-subtracted energy resolution for gamma-rays with
EP = 570 keV as a function of applied electric field. The resolution was 39 keV at the electric
field of 9 kV/cm. In Fig. 4 comparison is made of noise-subtracted energy resolutions
obtained for various particles by other investigators. Agreement of the resolutions among
136
many experiments is fairly good but the resolution of 39 keV is far worse than R = 1.6 keV
theoretically expected and worse even than RU = 7.0 keV for Poison fluctuation. They
suggested that the recombination of the electron and ion was the cause of deterioration of the
energy resolution.
Fig. 3. Collected charge for 570 keV gamma rays in liquid xenon as a function of applied
electric field.
Fig. 4. Noise-subtracted energy resolution for 570 keV gamma rays in liquid xenon as a
function of applied electric field.
4. DISCUSSION
137
where QO is the charge produced by electron with energy EP, [0 and [1 are parameters relating
ion pair density at the end of delta-ray and at the minimum ionizing part of the electron track, a
and b are functions of the minimum and the maximum energies of the delta electron. In liquid
argon eqs. (3) and (4) are in fairly good agreement with experimental data 10). In liquid xenon,
however, the agreement of eq. (4) to the experimental data is rather poor10).
4.2 Test of the recombination model by doping liquid rare gases with photo-ionizing
molecules.
If Thomass model10) is a fact, appreciable part of produced ion-electron pairs recombines
at the end of delta-rays under the experimentally applicable electric field. It is considered that
photo-ionizing molecules doped in liquid rare gases will produce additional electron-ion pairs
by absorbing photons emitted from recombined electron-ion pairs at the end of delta-rays and
will improve the energy resolution to that theoretically expected.
Figs. 5. and 6. show collected charge and energy resolution for 976 keV electrons,
respectively in pure and 10 ppm allen-doped liquid argon ionization chambers11). The solid
lines are fitting result of Eqs. (3) and Eq. (4) to the experimental data in pure liquid argon. Eqs.
(3) and (4) are applied to liquid rare-gases doped with photo-ionizing molecules and the
following equations are derived:
where r = Nex/Ni is the ratio of number of exited atoms Nex to the number of ionized atoms Ni, K
a fraction for emitted photons from the recombined electron-ion pairs to be absorbed by
photo-ionizing molecules, I the quantum efficiency of photo-ionizing molecules to be ionized
138
by absorbing photons. The dotted line shows fitting result of modified Eq. (5) and (6) to the
experimental data in allen-doped liquid argon.
The fitting results and experimental data are in good agreement for collected charge,
which indicates that at the highest electric field most of the initially produced charge is collected
and that the charge additionally obtained by the doping is due to emitted photons from initially
excited atoms. Moreover at a high electric field, the energy resolution is in agreement with (4)
rather than with (6), that is, the energy resolution is not appreciably improved by the doping as
expected by Thomas model.
Fig. 5. Collected charge for 976 keV electrons in pure liquid argon and in liquid argon doped
with 10 ppm allene.
Fig. 6. Noise-subtracted energy resolution for 976 keV electrons in pure liquid argon and in
liquid argon doped with 10 ppm allene.
139
Figs. 7. and 8. show collected charge and energy resolution for 570 keV electrons in pure
and 20 ppm TEA-doped liquid xenon ionization chambers, respectively11). The solid lines are
fitting result of Eq. (3) and Eq. (4) to the experimental data in pure liquid xenon and dotted line
fitting result of Eq. (5) and (6) to the experimental data in TEA-doped liquid xenon. The
fitting results and experimental data are in fairly good agreement for collected charge but poor
for energy resolution. At a high electric field the energy resolution is not appreciably
improved by the doping as expected by Thomas model.
Fig. 7. Collected charge for570 keV gamma rays in pure liquid xenon and in liquid xenon
doped with 20 ppm TEA.
Fig. 8. Noise-subtracted energy resolution for570 keV gamma rays in pure liquid xenon and in
liquid xenon doped with 20 ppm TEA.
140
It is concluded that the recombination is not the main cause of poor energy resolution in
liquid rare-gas ionization chambers. We might consider that liquid rare gas is composed of
cluster of rare gas atoms with different size, that the cluster size is fluctuating from single-atom
size to whole detector size, and that the properties such as W-value are different with the size of
the cluster. If solid rare-gas filled ionization chamber is applicable, we might obtain a key to
solve the energy resolution problem .
.
5.5. SUMMARY
SUMMARY
In rare-gas filled ionization chambers the energy resolution is far worse than that
theoretically estimated. The recombination of produced electron-ion pairs are not the main
cause of the poor resolution experimentally obtained. More investigation is needed to solve
the energy resolution problem in liquid rare-gas filled ionization chambers.
REFERENCES
1) M. Miyajima et al, Phys. Rev. A9, (1974) 1438.
2) T. Takahashi et al, Phys. Rev. A12, (1975) 1771.
3) T. Takahashi et al, J. Phys. C17, (1974) 230.
4) T. Doke et al, Nucl. Instr. and Methods 134 (1976) 353
5) R. L. Platzman, Int. J. Appl. Rad. Isotopes 10 (1961) 116.
6) W. Shockley, Czech. J. Phys. Bll (1961) 81.
7) G. D. Alkhazov, A.P. Komar and A.A. Vorob'ev, Nucl. Instr. and Meth. 48 (1967) 1.
8) E. Aprile et al., Nucl. Instr. and Meth. A261 (1987) 519.
9) E. Aprile et al., Nucl. Instr. and Meth. A302 (1991) 177.
10) J. Thomas et al., Phys Rev. A38 (1988) 5793.
11) E. Shibamura et al., Jpn. J. Appl. Phys. 34 (1995) 1897.
141
LIST OF PARTICIPANTS
T. Ohama (KEK)
M. Ohno (The University of Tokyo)
N. Ishihara (KEK)
H. Takahashi (The University of Tokyo)
H. Hirayama (KEK)
W. Kada (JAEA)
K. Takahashi (KEK)
S. Komatsu (Seiko EG&G Co.,Ltd)
S. Ban (KEK)
T. Nagayoshi (Rigaku Corporation)
T. Taguchi (Rigaku Corporation)
F. Tokanai (Yamagata University)
T. Torii (JAEA)
M. Numajiri (KEK)
T. Okunoyama (Shizuoka University)
T. Koike (IUHW)
M. Kumada (NIRS)
K. Bessho (KEK)
I. Murayama (Yokohama National University)
Y. Unno (AIST)
E. Matsumoto (HORIBA, Ltd.)
Y. Ogata (Nagoya University)
T. Ito (Shizuoka University)
A. Toyoda (KEK)
T. Takagi (Nagoya University)
H. Ichiki (Kyushu Sangyo University)
H. Toyokawa (Japan Synchrotron Radiation Research Institute)
I. H.Suzuki (KEK)
H. Nakamura (KEK)
J. Kawarabayashi (Nagoya University)
M. Nakayama (Nagoya University)
H. Monjushiro (KEK)
K. Ishihara (Nagoya University)
C. Sakamoto (Nagoya University)
K. Saito (KEK)
H. Minato (Nagoya University)
K. Kawai (Nagoya University)
142
Y. Arai (KEK)
H. Matsumura (KEK)
H. Kitaguchi (Hitachi, Ltd.)
K. Takahashi (KEK)
A. Yamaji (IMR)
A. Komiya (KEK)
Y. Ito (Yamagata University)
T. Mukumoto (Shizuoka University)
A. Yoshikawa (Institute for Materials Research)
K. Masumoto (KEK)
K. Fushimi (The University of Tokushima)
M. Taira (KEK)
H. Shimazaki (Kyoto University)
K. Shima (Kyoto University)
Y. Hoto (Kyoto University)
S. Sasaki (KEK)
Y. Takagi (Yokohama National University)
S. Nakamura (Yokohama National University)
T. Michikawa (KEK)
H. Onabe (Raytech Corporation)
H. Iwase (KEK)
M. Takeyama (Yamagata University)
T. Moriya (Yamagata University)
T. Sanami (KEK)
A. H. Malik (The University of Tokyo)
T. Onodera (Tohoku Institute of Technology)
M. Hagiwara (KEK)
K. Maehata (Kyushu University)
R. Maeda (Kyushu University)
H. Tawara (KEK)
K. Harada (The University of Tokushima)
Y. Yamashita (Kyoto University)
K. Hozumi (KEK)
K. Watanabe (Nagoya University)
N. Dewi (The University of Tokyo)
Y. Kishimoto (KEK)
K. Shimazoe (The University of Tokushima)
K. Okada (Hitachi, Ltd.)
Y. Namito (KEK)
H. Sakai (Toshiba Corporation)
143
N. Toyoshima (KEK)
F. Yamana (Keirex Technology Inc.)
K. Mori (JAXA)
K. Muromachi (KEK)
I. Murakami (Yokohama National University)
K. Fujiwara (The University of Tokyo)
M. Sawabe (KEK)
K. Morishima (Nagoya University)
H. Kusano (Waseda University)
I. Knno (Kyoto University)
T. Miura (KEK)
S. Sato (KEK)
H. Maruyama (Nagoya University)
Y. Ito (Yamagata University)
M. Sato (KEK)
S. Kurosawa (Tohoku University)
Y. Fujimoto (Tohoku University)
S. Kishimoto (KEK)
N. Fujioka (Seiko EG&G Co.,Ltd)
S. Takahara (KEK)
E. Shibamura (Saitama Prefectural University)
M. Kuroyanagi (Tokyo Nuclear Services Co.,Ltd)
A. Takahashi (KEK)
K. Terasawa (Keio University / JAXA)
K. Iijima (KEK)
144