A large solid angle multiparameter neutron detector

374

Nuclear instruments and Methods in Physics Research A307 (1991) 374-379 North-Holland

A large solid angle multiparameter neutron detector G. Ricco, M. Anghinolfi, P. Corvisiero, E. Durante, S. Maggiolo, P. Prati, A. Rottura and M. Taiuti Dipartimento di Fisica and INFN-Sezione di Genova, via Dodecaneso 33, 16146, Genova, Italy

Received 28 January 1991 and in revised form 6 May 1991

A 4,~ neutron detector has been realized using organic scintillators: the detector is suitable for high efficiency, low background measurements of very low neutron rates in the 0.6-5 MeV energy range. Gamma-neutron discrimination has been performed by pulse shape, energy and neutron lifetime analysis and backgrounds have been reduced by anticoincidence detectors and paraffin-lead shielding. Tests of efficiency, energy resolution and radiation identification have been made with a low intensity Am-Be neutron source.

1. Introduction Recent claims [1] a b o u t the so called " c o l d fusion" have suggested the need for detectors capable of good n e u t r o n identification with reasonable sensitivity. T h e presented system allows a m u l t i p a r a m e t r i c n e u t r o n identification, measuring at the same time pulse shape, energy a n d time span of each detected event. In order to ensure a high detection efficiency a n d sensitivity very close to the values o b t a i n e d in u n d e r g r o u n d laboratories, the detector s u r r o u n d s the source with a 4 4 geometry; the source is placed in a central hole, a n d large anticoincidence plastic scintillators positioned above the detector allow the reduction of cosmic background. A C A M A C interface transmits all the data to a MicroV A X computer, which performs the data analysis.

2. Description of the detector Figs. l a a n d l b show the detector c o m p o s e d of three cylindrical coaxial scintillator shells; each shell is 20 cm long a n d a b o u t 5 cm thick: the inner shell is filled with NE213 liquid scintillator while the outer two are plastic NE102A. The light from the liquid scintillator is collected b y seven photomultipliers (Phillips, mod. 2262B), three o n the u p p e r a n d four on the lower surface; the other two scintillators are coupled each with four p h o t o multipliers (Phillips, mod. 2262B for the m e d i u m detector a n d E M I 9805A for the side detector). C a d m i u m sheets 1 m m thick are interposed between the scintillators a n d a t h i n n e r one (0.5 m m ) is w r a p p e d a r o u n d the central hole (diameter 4 cm) where a 30 c m 3 cylindrical sample can b e inserted (fig. lb). The anticoincidence cosmic ray detector is m a d e up of three plastic N E 1 0 2 A

scintillator slabs: the central one 65 x 30 x 5 cm 3, the two side ones 65 x 20 x 5 cm 3, a r r a n g e d to form a roof o n the detector region, as s h o w n in fig. l b . The whole system is s u r r o u n d e d in all directions by a shielding c o m p o s e d of an internal layer of 20 cm p a r a f f i n (5% enriched in Li 2CO3) followed b y 2 cm electrolytic copper a n d 10 cm lead. In order to reduce electronic noise a n d cosmic ray b a c k g r o u n d the s u m m e d a n o d e signals from the liquid scintillator are accepted only when observed: a) in coincidence between the two groups of u p p e r a n d lower photomultipliers, b) in anticoincidence within a 20 ns interval with the cosmic ray detector. T h e NE213 light pulse has different decay times when p r o d u c e d by g a m m a rays or n e u t r o n s [2]: a pulse shape discriminator (PSD, C a n b e r r a 2160A), operating on the s u m m e d a n o d e pulses of the liquid scintillator, separates g a m m a from n e u t r o n events. Since the s p e c t r o m e t e r d i m e n s i o n s are relatively large with respect to the m e a n free p a t h of a few MeV neutrons, a very high fraction of n e u t r o n s giving a signal in the liquid scintillator thermalizes a n d is c a p t u r e d by c a d m i u m foils. Each event ( n e u t r o n or g a m m a ) releasing energy above the 60 kee (keV electron equivalent, see section 4 a n d fig. 6) threshold in the liquid scintillator gives the start to a t i m e - t o - a m p l i t u d e converter (TAC, C a n b e r r a 2143), that will wait up to 200 ~ts for the stop from g a m m a rays p r o d u c e d by n e u t r o n capture on one c a d m i u m sheet a n d detected by at least one of the three coaxial scintillators. To avoid start a n d stop pulse pairs p r o d u c e d b y the same n e u t r o n slowing d o w n in the scintillators, the start pulse is delayed by 500 ns before the T A C input. A n e u t r o n event is therefore identified by two trigger signals: the first corresponding to a n e u t r o n - t y p e pulse in the liquid scintillator a n d the second to detection of capture

0168-9002/91/$03.50 © 1991 - Elsevier Science Publishers B.V. All rights reserved

375

G. Ricco et al. / A solid angle multiparameter neutron detector TOP View

SIDE View 8

[

•

] g

Cd Foils BOTTOM View

Cd Foils

ANTICOINCIDENCE

I---'1

L)QUIDSCINTILLATOR

PLASTICSCINTILLATOR

O

PMTUBE

~

LIQUIDSCINTILLATOR PLASTICSCINTILLATOR

I--'-I

PMTUBE

ANTICOINCIDENCE

Fig. 1. Spectrometer geometry and composition. (a) Top and bottom view. (b) Side view.

gamma rays within the 200 /ts time window. For each event the linear amplitude of the three scintillator summed anode pulses is also analyzed.

3. Electronics

The convertion and acquisition of those signals (energy signal from the scintillators, pulse shape analysis and n - T time correlation) necessary to select neutrons and their energy, has been obtained with a LeCroy F E R A system composed by a F E R A ADC, a F E R A driver and two 16 Kbyte Fast Memories. The F E R A A D C is dedicated to energy, pulse shape and n - T time correlation acquisition. Two memories are necessary to permit data-storing through the ECL bus during the C A M A C readout procedure. A single event acquisition occurs in two steps separated in time: a) first the liquid scintillator signal and the signal corresponding to the energy released in the whole calorimeter are converted by the F E R A A D C after 100 ns from the event occurrence and then stored in the Fast Memory,

b) then the signals corresponding to the pulse shape and the n - T time correlation are converted and stored 200 p.s later. To perform this task a prompt and a delayed trigger have been realized. The prompt trigger is generated by those events that correspond to a released energy in the liquid scintillator above 60 kee. It is used to generate: i) the A D C gate, 200 ns wide, for the energy acquisition; ii) the start signal for the n - T time correlation; iii) the TAC start signal for the pulse shape analysis; iv) the TAC strobe signal to synchronize the T A C linear outputs; v) the clear signal for the F E R A converter, 12 Ixs delayed. The delayed trigger is generated by the SCA output of the TAC employed in the n - v time correlation and it is used to generate: i) the A D C gate for pulse shape and n--¢ time correlation acquisition, ii) the clear signal for the F E R A converter, 12 ~ts delayed. The TAC for n - T time correlation starts to convert on the prompt trigger and stops with the ORed pulse from the three cylindrical scintillators; since the liquid scintillator participates to generate both, the start signal has been 500 ns delayed. In fig. 2 the trigger logic is presented: in order to minimize the n u m b e r of utilized devices we have used

376

G. Ricco et aL / A solid angle multiparameter neutron detector

]

input PSD [_{strobe ] I - -- --

[

t stop TAC ~ | TAC I' r--] stan sea ~-~

~

LG&S ,

,

I

reset

1

AOC L--t°"I

~-------

I

gate

B

k2_l

cA~CAC I t........-a

7

1

I

II

veto

I I

_

[

valid.__f-'@----]

I

CAMAC

[

Fig. 2. Trigger logic and circuit scheme. Letters on the left-most column indicate the summed anode pulses of: UI - inner upper photomultipliers, LI - inner lower photomultipliers, M - middle photomultipliers, O - outer photomultipliers, UA - upper anticoincidence photomultipliers, RA - right anticoincidence photomultipliers, LA - left anticoincidence photomultipliers. a n octal E G & G 8000 C F D , two fourfold logic units (LeCroy 365 A L a n d E G & G 4010) a n d a Q u a d G a t e a n d Delay G e n e r a t o r Phillips Scientific 794.

4. Performanceof the detector C a d m i u m has a high capture cross section for epit h e r m a l a n d t h e r m a l neutrons: almost every n e u t r o n

1000

800

t

~600 "~ ~ 0

400

0

]~Et

,

,

,

i 40

,

,

,

i , 80

,

,

120

,

. . . . 160

200

T [l~S] Fig. 3. Neutron lifetime measurement; neutron capture time distribution measured with an Am-Be source. The source emission rate is 2700 n / s ; the measuring time is 60 s.

capture o n a c a d m i u m sheet is followed b y g a m m a emission (for each capture o n the average three g a m m a rays with a total energy of a b o u t 9 M e V are emitted) which is detected, with a high efficiency (83%), b y at least one of the three scintillators. If we define as n e u t r o n lifetime the time interval between the fast neutron detection in the liquid scintillator (start) a n d its capture (stop), the lifetime d i s t r i b u t i o n follows a double-exponential slope. In fig. 3 is r e p o r t e d the m e a s u r e d T A C spectrum of the A m - B e source: captures in the epithermal region have a time c o n s t a n t of a b o u t 3 ~ts while t h e r m a l absorpitions have a 50 ~ts lifetime: the capture time distribution is quite energy i n d e p e n d e n t , at least in the 0 . 6 - 5 M e V n e u t r o n energy interval, due to the fast energy slowing d o w n to the epithermal region ( a b o u t 200 ns): in fig. 4 simulated (by a M o n t e Carlo code [3]) spectra for a n A m - B e source a n d some m o n o c h r o m a t i c n e u t r o n s are reported. T h e T A C w i n d o w is o p e n for 200 ~ts: in such a time a b o u t 95% of the n e u t r o n s detected b y the liquid scintillators, in the energy range 0 . 6 - 5 MeV, are c a p t u r e d b y c a d m i u m . Longer time windows would increase the system deadtime in a n u n a c c e p t a b l e way with little a d v a n t a g e o n detection efficiency. T h e pulse shape d i s t r i b u t i o n f r o m the pulse shape discriminator (PSD), o b t a i n e d with a n A m - B e source, is s h o w n in fig. 5: the g a m m a a n d n e u t r o n peaks h a v e a

377

G. Ricco et al. / A solid angle multiparameter neutron detector 0.01

.

.

.

.

i

.

.

.

.

f

.

.

.

.

i

.

.

.

2.5

.

. . . .

i

. . . .

i

. . . .

i

. . . .

i

. . . .

C . A m B e

2,45 MeV r-

•

g

5 MeV

0.001

c

1.5

0

-

s

~ - ' ~- . . ~ . . .

g

l

.Q


lff4

1 ~

0.5

0L 1 lff5

,

,

,

,

1 50

,

,

,

,

t 100

.

.

.

.

I 150

.

.

.

. 200

[~s] Fig. 4. Neutron captures time distributions, normalized to 100% detection efficiency, for an Am-Be source and monochromatic neutrons as determined by Monte Carlo simulation. Curve fitting gives: N ( t ) = 0.095XT e-hIt +0.04Xse -xst, X s = 0.30 p,s -1, X~=0.017 ixs -1 for the Am-Be source; N ( t ) = 0.141Xie xIt+0.06)kse xst, ~s=0.26 /,ts-2, ~;=0.017 /is - ] for 2.45 MeV neutrons; N ( t ) = 0.073~ie -x'' d-0.03~kse -hst, As = 0.30 p,s -], ~l = 0.017p, s ] for 5 MeV neutrons.

F W H M of a b o u t 3 ns a n d 11 ns respectively. The P S D is g a t e d by the T A C output; in this way events characterized by a long pulse d u r a t i o n in NE213 b u t with a very short lifetime (typically high energy g a m m a saturated pulses) are rejected. A logic pulse is also p r o d u c e d for each d e a f l y identified g a m m a event to inhibit the T A C conversion a n d reduce the detector d e a d - t i m e (fig. 2); in other words, each start signal not recognized b y the P S D as a n e u t r o n is rejected. Plastic a n d liquid scintillators have a high efficiency for n e u t r o n detection b u t they have a poor energy resolution. Inside the scintillator, the n e u t r o n s release their energy in several collisions mainly with hydrogen nuclei; n e u t r o n s t h a t thermalize in the spectrometer a n d

.

.~

.

.

.

.

.

.

.

.

.

.

.

.

.

t

.

.

.

.

.

.

.

.

i

. . . .

0.8

.m e,.-i

0.6

m

~ 0.4 .Q

< 0.2

0

..i 5

10

15

20

25

30

. , . , , , 35

40

T [ns]

Fig. 5. Pulse shape discrimination: measured crossover time distribution of an Am-Be neutron source. Peak-to-valley ratio is 7.7 and 2.6 for gamma and n peak respectively. Threshold energy was 60 kee.

2

3

4

Neutron energy [MeV]

Fig. 6. Mee (MeV electron equivalent) versus neutron energy calculated in the detector energy range (0.6-5 MeV).

are c a p t u r e d by c a d m i u m distribute their energy in at least one of the three scintillators; the a n o d e pulses from the p h o t o m u l t i p l i e r s of the three scintillators are therefore s u m m e d to o b t a i n the whole energy released in the spectrometer. Since energy released in organic scintillators is calibrated using m o n o c h r o m a t i c g a m m a ray sources, M e V electron equivalent (Mee) is generally assumed as energy unit: 1 M e e represents the light collected after 1 M e V energy has b e e n lost b y a n electron in the scintillator. T o show the lower light response to recoil protons, the Mee scale is plotted in fig. 6 against the c o r r e s p o n d i n g energy loss scale for recoil p r o t o n s in the energy interval of interest for our instrument. T h e p h o t o m u l t i p l i e r ' s voltages a n d the linear amplifications have b e e n optimized to obtain, for every scintillator a n d for a released energy of 2 Mee, a 250 p C signal at the A D C input, the m i n i m u m detectable energy b e i n g 0.06 Mee. A 1 M e e energy releasing in the liquid scintillator produces a 1 V height signal at the P S D i n p u t which is linear up to 5 V. Energy resolution for g a m m a rays has been o b t a i n e d b y fitting the experim e n t a l spectra of different g a m m a sources (in the range 0.06-1.9 Mee) with the results of a M o n t e Carlo code [4] where the intrinsic response f u n c t i o n at each energy has been folded with a G a u s s i a n curve. T h e F W H M values of the Gaussians, which take into a c c o u n t statistical light collection effects, t u r n out to b e a b o u t 13% for the liquid scintillator a n d a b o u t 11% for the sum spectrum at 1 MeV. In fig. 7 the s p e c t r u m of a n A m - B e source, with a n e u t r o n emission rate of 2.7 × 10 3 n / s a n d a c o n t i n u o u s energy spectrum, is c o m p a r e d with the result of a M o n t e C a r l o calculation [3]. T h e disagreement between m e a s u r e d a n d calculated d a t a is mostly due to the coincident g a m m a activity of the A m - B e source (originated by the 9Be(et, n)12C * reaction; g a m m a energy = 4.44 MeV) which has n o t b e e n t a k e n into account b y the simulation. However, the relative b e h a v i o u r is in a g r e e m e n t with the hypothesis of a 4.44 M e V g a m m a ray releasing part of its energy by

G. Ricco et al. / A sofid angle multiparameter neutron detector

378

{

I{{

o

Simulation

•

Measurement

6.4

1.75 1.50

0=

~" 4 . 8

1.25

{t]~

1.00

m ,~ 3.2

tt ~r ~I I I I

•~ 1.6

0,75

[t I I{

t

II]~[II[l]~

I

I[[

{

0.50

'"~,t~,ltl]t|tt 0

,

,

l

I

l

400

,

,

I

l

,

800

,

I 1200

,

,

,

I 1600

= ,

,

~.. - .'.

0.25

~.,::~:. . . .

" i~

, 2000

200

P u l s e Height [kee]

600

800

~:/; ...,~:~':: :.

.

1.75 .

.

.

1.00

..~..,:~z.

"

. -'

..

:.

• .

:.

0.50

,:

.

.

' ?'~!i~ ~i:.,:.. :e~ ,;..~:: ,

•

.:~,~

.

",k~

. ..=

i .-

-:

,."!~.%;L,.

" .... . . . . . .

0.75

.

.

' , . . . ": , ~ . ; t . . , "

"

1.25

4)

• "

.

; ...:- .. :;;~.',"i'--".

.'.~

.

~:'.'?.'L"," : . . . . . .

~ :~'5~.,~": • .7~..?....:~!~.~:..~

" . . . . ""-"

:, ~:ii~:.

~-:~, "

~ ' w . @ = ' : . . -.

.-'~", :'.. ;:,:': i:,."*~ ':."g~'.":::.""

0.25

200

400

600

800

1000

PSD channels

1.75

C ~...

1.50 . . . . . :

1.25

•

... 1.00

Y

"

~ . ; ,..

".

•

-~i,,,-:

0.75 0.50

0,5

1000

PSD channels

Fig. 7. Am-Be neutron source amplitude spectrum of the sum of the anode pulses from the three scintillators compared with a Monte Carlo simulation: the two spectra are normalized to the highest energy point.

C o m p t o n scattering in coincidence with a neutron therrealization. This effect could take into account the unpredicted low energy tail in the experimental spectrum; nevertheless in "cold fusion" measurements where no gammas are produced (if the supposed reaction takes place), this problem does not affect the detector reliability. In fig. 8 a simulated spectrum of 2.45 M e V m o n o chromatic neutrons is displayed; the spectrum has a Gaussian shape centered at about 0.4 M e e and a total variance of 0.15 Mee. P S D and energy spectra have been joined in the bidimensional plot shown in figs. 9 a - 9 c for a 6°Co g a m m a source, for an A m - B e source and for background: discrimination is definitely improved by the bidimensional analysis and the two gamma and neutron regions can be well separated by an exponential curve.

400

0.25

0.4 200

400

600

800

1000

0.3

PSD Channels ~

0.2

Fig. 9. Correlation between pulse amplitude in the liquid scintillator and PSD output in a two-dimensional plot. (a) 6°Co gamma source. (b) A m - B e neutron source. (c) Background.

0.1

0

0.2

0.4

E

0.6

I1

0.8

[Mee]

Fig. 8. Simulated spectrum of monochromatic neutrons E = 2.45 M e V ) normalized to 100% detection efficiency. The spectrum is fitted by f ( E n ) = 0.224/(2~o2) 1/2 exp((E n -

0.41)2/(2o2)); o = 0.15 Mee.

The neutron detection efficiency has been measured using the calibrated low intensity A m - B e source. The source emits also a great quantity of g a m m a rays from the decay of excited C levels; in order to avoid effi-

G. Ricco et al. / A solid angle multipararneter neutron detector

ciency losses, the anticoincidence detectors were disactivated. In the T A C spectrum the random contribution of start and stop pairs inside 200 ~ts must be taken into account: this can be done by subtracting from the double-exponential slope (fig. 3) the statistical (random) contribution which reaches asymptotically a constant value. The measured efficiency 7/ is 10 + 0.5% for the A m - B e neutron source. The coincident gamma activity of A m - B e increases the detection efficiency (neutrons below the energy threshold may be detected if coincident in the central detector with a low energy gamma pulse produced by a C o m p t o n scattering of a 4 MeV g a m m a ray from 12C deexcitation): this effect was estimated by calculating the 4 MeV g a m m a detection efficiency (20%) and the fraction of neutrons emitted with energy below 0.06 Mee; even assuming that each of these neutrons were coincident with a gamma, the increase in detection efficiency is less than 0.5%. Efficiency values (10.5% for A m - B e ) very close to the experimental ones were obtained with the Monte Carlo simulation. The background level (B), using the shielding quoted in section 2 and averaged on the whole A m - B e energy interval, has been measured to be of the order of 0.02 cps. We may define the detector sensitivity as the lowest detectable neutron rate R n as: R , = 3 ~rff/ T 11

,

where T is the measuring time; with the quoted ~/ and B values for A m - B e neutrons and T = 1 h, R n is of the order of 0.07 n / s . For 2.45 MeV monochromatic neu-

379

trons, B = 0.01 cps, ~/= 12.6% (computed by the Monte Carlo code) and the sensitivity turns out to be 2.4(T) -1/2 n / s .

5. Conclusions A high sensitivity neutron detector has been realized. Neutrons in the 0.6-5 MeV energy range are identified by a multiparametric analysis, performed by energy and pulse shape discrimination and a T A C with a time window of up to 200 ~ts. Background is reduced down to 0.02 cps in the 0.6-5 MeV neutron energy range by a composite passive shielding and anticoincidence detectors. The spectrometer is suitable for the detection of neutrons emitted by a source placed in its center with sensitivity up to 10 -2 n / s within a few hours measuring time.

References [1] S.E. Jones, E.P. Palmer, J.B. Czirr, D.L. Decker, G.L. Jensen, J.M. Thorne, S.F. Taylor and J. Rafelski, Nature 338 (1989) 737. [2] F.T. Kuchnir and F.J. Lynch, IEEE Trans. Nucl. Sci. NS-15(9) (1968) 107. [3] M. Anghinolfi, G. Ricco, P. Corvisiero and F. Masulli, Nucl. Instr. and Meth. 165 (1979) 217. [4] M. Taiuti, M. Anghinolfi, P. Corvisiero, G. Ricco and A. Zucchiani, Nucl. Instr. and Meth. 211 (1983) 135.