2011 IEEE Nuclear Science Symposium Conference Record
NPl.M-193
Characterization and Calibration of Novel Semiconductor Detectors of Thermal Neutrons for ESA Space Applications Zdenek Kohout, Carlos Granj a, IEEE Member, Miloslav Kralik, Alan Owens, Robert Venn, Lukasz Jankowski, Stanislav Pospisil, IEEE Member, Bruno Sopko, Jiri Vacik
Abstract-For the study and detection of neutrons in space
TABLE I. NEUTRON CONVERTERS AND REACTION PRODUCTS
environments such as planetary and earth orbiting missions semiconductor silicon diode detectors have been characterized and calibrated at various thermal neutron sources. Two types of
Isotope
diodes adapted for thermal neutron detection were investigated: 6 silicon MESA planar detectors equipped with thin LiF layers and
silicon heterodiodes with a layer of natural boron or lO enriched B. The response and absolute detection efficiency have been measured. The influence of bias voltage and converter layer thickness
were
studied.
As
neutron
sources
we
used
"6
IOB+n� 1Li+o.
6Li
6Li+n�jH+a
'He
JHe+n�lH+p
[MeV ] (
o !OB+n� 7Li +a+y O.48MeV
2.79
)
2.31
4.78
Branching fraction
0.06
0.94
0.76
Energy of
Cross
secondaries
section
[ bam ]
[MeV ]
Eu �1.01, Ea�1.78
Efj =0.84, Ea =1.47
E," �2. 3, Ea�2.0 H . , p -0,57
7 E O I9 E
5
3842 942
5320
a
homogenous isotropic thermal neutron field by a set of PuBe
II.
radionuclide sources placed in a graphite pile as well as a parallel 5 thermal neutron beam with high Cd ratio (10 ) and suppressed
A.
gamma background. Depending on the converter layer thickness and/or boron layer thickness as well as the choice of the threshold
level, efficiencies of approximately 1 % are obtained for both the 6 silicon diodes with thin LiF and the boron rich silicon detectors. These values guarantee optimal stability of operation in remote and different environments as well as maximum signal-to-noise ratio by enhanced suppression of unwanted signals and gamma background.
I.
Q
Reaction
NEUTRON DETECTORS
Boron Rich Semiconductor Heterodiode
The boron rich semiconductor heterodiodes are manufactured and provided by Cambridge Microfab Ltd. [1]. These devices were prepared on a silicon substrate oriented n type generally 380 /lm thick. The thickness of the boron rich layer can be varied as required. The dimensions of the silicon heterodiodes are 7x7 mm2 (see Fig. 1 and Fig. 2).
INTRODUCTION
d� Li
emiconductor detecting diodes are well established devices for many radiation detection tasks in both fundamental and applied research. Small size and portable devices are desired for applications requiring deployment in remote places which demand integrated signal acquisition including also control and power. The long term goal is to provide self-standing sensitive devices with reasonable efficiency for detection of thermal neutrons. In this work we present the evaluation and calibration of two types of silicon detectors. For neutron detection we employ conventional reactions on 3 He, lOB and 6Li isotopes. The well known characteristics of thermal neutron detection on these materials are presented in Table I. Isotopes JOB, 6Li are usually used as converters with silicon semiconductor detectors.
S
Manuscript received November 14, 2011. This work was supported by the European Space Agency - ESA Contract No. C22 908/09INLlCBi. Z. Kohout, B. Sopko are with the Faculty of Mechanical Engineering, Czech Technical University in Prague, Prague, Czech Republic (telephone: +420 224 352 42 8, e-mail:
[email protected]). C. Granja, S. Pospisil are with the Institute of Experimental and Applied Physics, Czech Technical University in Prague, Prague, Czech Republic. M. Kralik is with the Czech Metrology Institute, Prague, Czech Republic A. Owens is with ESA - ESTEC, Noordwijk, The Netherlands R. Venn, L. Jankowski are with the Cambridge Microfab Ltd., Cambridge CB23 2TA, UK J. Vacik is with Nuclear Physics Institute, Academy of Sciences of the Czech Republic, Rez near Prague, Czech Republic
978-1-4673-0120-6111/$26.00 ©2011 IEEE
�,
Neutron reactive hetcrodiodc region (p type semiconductor layer)
n type silicon substrate
�
Lower electrode structure
9 mm
1!Fl+-_�
)( 9
mm
chip
50 mm2 active area 016.75
mm
Fig. 1. Schematic illustration of the boron rich semiconductor heterodiode device designed for thermal neutron detection (bottom). The principle of operation in a neutron field - neutron reaction products are detected in the reverse biased heterodiode (top).
A set of heterodiodes with different boron layer thickness ranging from 500 nm up to 4100 run was tested. The converters were prepared from natural boron or highly enriched with JOB.
400
Fig. 2. Boron rich semiconductor heterodiode (right) and assembled unit with preamplifier embedded in ESA test electronics (left).
B.
Fig. 5. Assembly (right) and unit (left) of the MESA neutron detector. III.
Planar Silicon Diodes with 6Li Converter
The second type of studied detectors adapted for neutron detection [2] is based on conventional silicon MESA planar diodes [3] of thickness 300 /lm and size 5x5 mm2 (see Fig. 3). As converter thin 6LiF layers of different thickness ranging from 0.25 to 5.2 mglcm2 were used. Illustration of detector adapted for thermal neutron detection is given in Fig. 4. An assembled device is shown in Fig. 5.
EXPERIMENTAL SETUP AND MEASUREMENTS
The neutron detectors were readout by a standard spectroscopic signal chain consisting of a charge sensitive preamplifier, a spectroscopic amplifier and ADC linked to a PC as illustrated in Fig. 6. The boron rich semiconductor detectors were operated with a preamplifier embedded in ESA tests electronics unit, see Fig. 2. A devoted digital interface [4] to readout the ADC to PC was used.
A
N
A
Fig. 3. Layout of the design (left) and top view photo (right) of the MESA planar silicon detector.
Fig. 6. Spectroscopic signal chain used. A.
Top view
The response and energy calibration of the detectors were measured with alpha particles from 241 Am and from a combined 241 Am + 239pU + 244Cm spectroscopic source (see Fig. 7). The influence of bias voltage and 6LiF and boron layer thickness were studied. Testing and neutron detection absolute efficiency measurements were carried out in a homogenous isotropic thermal neutron field generated by three PuBe radionuclide sources placed in a graphite pile (see Fig. 8). The pile (dimensions 2 m x 2 m x 4 m) was equipped with an irradiation cavity (dimensions 30 cm x 100 cm x 30 cm) which serves as a standard of well defined and stable thermal neutron flux. The thermal neutron flux was equal to 3.26x104 nlcm2/s and was previously determined with activation gold foils.
Side view
.Jf a AI foil ===iii ii i ii ii ii ii ii ii iiiiiiii iil:;O:)iiiiiiiiii r-ii = ii = 6 Li F con verter /f\ t -Imm
CH
""--�;""-------'I
Detector Characterization and Calibration
Si detector
B.
Fig. 4. Schematic layout of the diode with 6LiF (top) and illustration of thermal neutron detection (bottom).
Tests in Parallel Thermal Neutron Beam
Verification and test measurements were carried out with a parallel thermal neutron beam from the research reactor LWR - 15 in Rez near Prague (see Fig. 9). The beam profile is 4 mm x 10 cm and exhibits a high Cd ratio (105) with highly suppressed gamma background.
401
8162_83, bias 9V,
-------------
thickness of the boron rich layer 504 nm
3000 �
----"-" - - "'- ut- n--'lT 400 ro - therma lne"' - al ha source Am Pu em 350
2S00
300 2000
250
!l
§o
(J
1500
200 150
1000
100 SOO
so
100
Fig. 7. Experimental vacuum setup for detector calibration with [at lEAP CTU in Prague].
a
200
300
400
500
600
700
800
900
Channel
particles
8158_03, bias 9V,
-----""iT 400 -; 3000 ��------------=------,, - thermal neutron - al ha source Am Pu em 350 2500 thickness of the boron rich layer 2509 nm
300 2000
250
�
§
o (J
200
1500
150
1000
100
,
500
100
200
300
50
400
500
600
700
800
900
1000
Channel
8163_03, bias 9V,
-
thickness of the boron rich layer 4155 nm
00 3oo 0 n-r-------------=----,_ �� �� ln= th e=rma tro� n --rr 4 eu� - al a source Am Pu em 350 2S00
Fig. 8. Isotropic thermal neutron source graphite pile with three Pu-Be radioactive sources [CMI Prague].
�
§o
(J
300
r
2000
2SO 200
1500
1SO
1000
100 SOO
Fig. 9. Irradiating setups with detector holder and positioning systems installed at the parallel thermal neutron beam at the LWR-15 reactor [NPI Rez near Prague].
IV. A.
RESULTS
Boron Rich Semiconductor Heterodiode
Boron rich semiconductor heterodiodes were measured with bias 0 and 9 V. Heterodiodes with different thickness of the boron semiconductor layer (500 nm, 1100 nm, 2500 nm, 3400 nm and 4200 nm ±10%) were studied. The response of three heterodiodes to thermal neutrons and a. particles is given in Fig. 10, where the influence of the thickness of the boron semiconductor layer is apparent. Together with the charged particle pulse height spectra, obtained from thermal neutron detection at the graphite pile, the pulse height spectra from the combined alpha source are presented.
50
100
200
300
400
SOO
600
700
600
900
1000
Channel
Fig. 10. Charged particle pulse high spectra from detection of thermal neutrons by silicon heterodiodes (black curve) and from a three-peak composite 0; source 24tAm + 239pU + 244Cm (red curve). Results are shown for three devices with different boron layer thickness at bias 9 V.
In case of the thin boron semiconductor layer (504 nm) a clear separation of the alpha particle and 7Li peaks is . observed. The alpha peak (at 5. 805 MeV) from 244Cm IS not observed in the heterodiode with the thinnest boron layer (top spectrum) as the thickness of the depleted layer of the heterodiode was smaller than the range of these alpha particles. The peak corresponding to 244Cm is observed when the thickness of the boron semiconductor layer is greater than about 1000 nm. As the thickness of the boron rich semiconductor layer increases, range attenuation in the absorber layer causes a broadening and the convergence of the otherwise separate alpha and 7Li peaks. For this reason, above 1000 nm, the thickness of the boron rich semiconductor layer
402
has only a small effect on the form of the pulse height spectra. The thermal neutron detection efficiencies from thick 1000 nm are also similar and are listed in Table II. A rather high threshold level was chosen for the determination of the efficiency (shown by a green arrow in Fig. 10). This reduces the overall efficiency but maximizes the signal to noise ratio and stability of detector operation in different radiation fields and remote environment conditions. The energy calibration curve for the heterodiode with converter thickness 504 nm is given in Fig. 11.
1000
TT-------=--,====;r 4000 -Ihemla' neu'ron,COfIYerter 5.2mgkm2,,ime fA mea•. Smin
900
3500
-'henna' neu.lOn. converter 2.6mglem2. ';me fA mea•. Smin -,hemm. neu'ron,COf1vefler O.26mgfcm2"ime fA mea•. 60min
800 .
3000
-ufpha&OtJfoeAm+Pu+Cm,ail,dislaIlO86OUrce-deI.lmm
700 C � 0 U �
2500
600 500
2000
400
1500
300
1000
200 100 200
6 > " 5 �4 >. E' 3 :g 2
400
600
y
=
1000
1200
1400
Fig. 12. Charged particle pulse high spectra from detection of thermal neutrons by MESA planar diodes with three different 6LiF converter thickness 1 (black, blue, green curves) and from a three-peak composite a source 24 Am + 2J9pu + 244Cm (red curve).
5.70E-03x + 5.19E-Ol
R'
=
1.00E+00
W1 o +------.---� 400
200
Channel
800
600
1000
Fig. II. Energy calibration curve of heterodiode with the thin lOB rich semiconductor layer. Data shown for one detector (see spectrum on top Fig. 10).
� R'
TABLE II. THERMAL NEUTRON DETECTION EFFICIENCY OF BORON RICH SEMICONDUCTOR HETERODIODES. detector
B.
800
Channel
thickness of edge boron layer [nm] [channell
alpha counts [-I
time of neutron nux measurement [n/s/cmJ [5]
504
125
62810
900
3.26E+04
0.43%
8161_03
1106
132
165385
900
3.26E+04
1.13%
8158_03
2509
136
148385
900
3.26E+04
1.01%
8165_E4
3358
140
145485
900
3.26E+04
0.99%
8163 03
4155
135
148454
900
3.26E+04
1.01%
600
Channel
800
9.99E-01
1000
1200
1400
Fig. 13. Energy calibration curve of a MESA planar diode with 0.26 mg/cm2 thick 6LiF converter.
[:1.]
8162_83
400
200
efficiency
3X-2 .34E-01
=
TABLE III. THERMAL NEUTRON DETECTION EFFICIENCY FOR THE MESA PLANAR SILICON DIODES WITH 6LiF CONVERTER. detector
Planar Silicon Diodes with 6Li Converter
The response of MESA planar diodes was measured with bias 30 V for different thickness of the 6LiF converter. In the case of the thin converter (black curve) the efficiency of thermal neutron is very small so the time of measurement is 12 times longer. The pulse height spectra from a three-peak composite ex source (241 Am + 239pU + 244Cm) together with the response to thermal neutrons detection measured at the graphite pile are shown in Fig. 12. The energy calibration with alpha particles was done under the same conditions as with the neutron converter (i.e., in air, same detector and same alpha source distance 1 mm - see Fig. 3). The pulse height spectrum is also influenced by the small window in the centre of the MESA planar detector's top electrode. The energy calibration curve for one diode with converter thickness 0.26 mg/cm2 is given in Fig. 13. The efficiencies of the diodes with converters to thermal neutrons are presented in Table III. The threshold was set in the valley between the two peaks of the charged particle products, i.e. at around channel 340 in the spectrum in Fig. 12.
converter
neutron counts
efficiency
[-I
time of measurement [ 5]
neutron flux
[mg/cm2]
[n/slcm2]
[%]
0.26
33349
1800
3.26E+04
0.23%
2.6
16628
300
3.26E+04
0.68%
5.2
24289
300
3.26E+04
0.99%
V.
CONCLUSIONS
A number of compact light weight solid state devices intended for thermal neutron detection in remote and different radiation environments have been investigated. The two types of semiconductor diodes studied were characterized and absolutely calibrated. Depending of the converter layer thickness and/or boron layer thickness, efficiencies of about 1% are obtained for the silicon diodes with thin 6LiF and around 1 % for the boron rich silicon detectors. These values are given for a stable and reliable operation of the devices under different conditions and harsh environment. ACKNOWLEDGMENTS
This work was supported by the European Space Agency ESA Contract No. C2290S/09!NLlCBi. REFERENCES [I] [2]
[3]
403
R. Venn, Private communication. S. Pospisil, B. Sopko, E. Havrankova, Z. Janout, J. Konicek, I. Macha, J. Pavlu, "Si Diode as a Small Detector of Slow Neutrons", Radiation Protection Dosimetry, vol. 46, no. 2, pp. 115-118, 1 9 93. D. Chren, M. Juneau, Z. Kohout, C. Lebel, C. Leroy, V. Linhart, S. Pospisil, P. Roy, A. Saintonge, B. Sopko, "Study of the characteristics of
[4]
silicon MESA radiation detectors", Nuc!. Instrum. and Methods A, vol. 460, pp. 146-15 8, 2001. P. Masek, V. Linhart, T. S1avicek, "RIO device based on USB 1. 0 for spectroscopy DAQ", Proceedings ofCTU in Prague Workshop, pp. 222223, 2008.
404
