The response of large NaI(Tl) detectors to fast neutrons

Nuclear Instruments and Methods 216 (1983) 141-148 North-Holland Publishing Company

THE RESPONSE

OF LARGE NaI(TI) DETECTORS

141

TO FAST NEUTRONS

J.J. V A N R U Y V E N , Z. S U J K O W S K I *, W . H . A . H E S S E L I N K Vrije Universiteit, De Boelelaan 1081, 1081 H V Amsterdam, The Netherlands

a n d H. V E R H E U L

Received 28 March 1983

The response of NaI(T1) scintillators of three different sizes to neutrons is measured as a function of neutron energy in the 0.17 MeV < E n < 13 MeV range. Various processes contributing to the neutron induced pulse-height spectra are discussed. The importance of multiple inelastic and elastic scattering events is emphasized. The experimental results are presented in the form of nomograms which make it possible to reconstruct the NaI(TI) pulse-height distribution for any given neutron spectrum within the energy range studied. Applications are considered to studies of single photon energy spectra as well as to sum-energy spectra of multiple photon cascades measured in the presence of intense neutron radiation. An example of such applications is described for the case of the 15°Nd(12C, xn'y)162-XDy reaction at E(12C)= 95 MeV.

I. Introduction Large NaI(T1) scintillation detectors are commonly used in the spectroscopy of y-rays of intermediate and high energy. In the domain of nuclear physics these devices are particularly useful for in-beam y-ray spectroscopy with light and heavy ions. They can be used in these applications for measuring the spectral distributions or other individual properties of y-rays from a given reaction (single y-ray spectroscopy) as well as for the spectroscopy of the total energy sum of multiple y-ray cascades (the total y-decay energy spectrocopy). The measured y-rays are often accompanied by intense fast neutron radiation. The detection efficiency and the response of a NaI(T1) crystal of a given size to neutrons are in such case of primary interest in any quantitative measurement of the y-rays. The traditional techniques used in the singles y-ray spectroscopy to discriminate between the detection of a neutron or a gamma-ray are based on time-of-flight a n d / o r pulse-shape considerations. Both techniques, however, have rather severe limitations in practice. The time-of-flight technique requires an appropriate starting signal (such as provided e.g. by a pulsed beam or by simultaneously detected prompt, coincident raditation), and a relatively large flight path. The pulse shape discrimination, on the other hand, is not easily applicable to NaI scintillators. The dominating part of the light output in the case of a neutron interaction is due to

* Present address: K.V.1. der Rijksuniversiteit Groningen, Groningen, The Netherlands; permanent address: Institute of Nuclear Research, Swierk, Poland. 0167-5087/83/0000-0000/$03.00

© 1983 North-ttolland

absorption of 7-rays following inelastic neutron scattering on 23Na and 127I. Thus the pulse shapes due to gamma and neutron interactions with NaI(T1) crystals are very similar [1,2]. These difficulties in the n e u t r o n - g a m m a separation take yet another dimension in the case of using NaI(T1) detectors such as a 4~r sum-spectrometer, especially if the spectrometer consists of only a few modules. The large solid angle and high intrinsic efficiency of such devices for both y-rays and neutrons implies that essentially every nuclear reaction of the type (H.I., x n y ) taking place in the target induces a composite signal in the sum-spectrometer due to y-rays and neutrons. The prompt component due to the y-rays is followed by a delayed component, usually of smaller amplitude, due to the neutrons. The delay times of the neutron induced components are characterized by a large spread of values, typically between a few ns and about 200 ns, caused by the spread in the neutron velocity distribution as well as by the spreading in the flight paths between the target and the interaction point in the large detector. The on-line, event-by-event corrections of the individual pulses being not practically applicable, one is thus left with the necessity to make statistical corrections to the pulse height spectra measured with sumspectrometers. The response of NaI(T1) to neutrons has therefore to be known as a function of the neutron energy and the detector size and shape. The main channels contributing to the energy degradation of a neutron with energies up to about 20 MeV moving in a NaI(T1) crystal are the (n, n), (n, n'y), (n, 2ny), (n, py), (n, a y ) and (n, 7) reactions. The published data on the cross sections for these interactions with 23Na and 127I nuclei are collected in table 1. From

142

J.J. van Ruyven et al. / The response of large NaI(TI) detectors

Table 1 Survey of the measured cross sections of neutron induced reactions on 23Na and 127I. Reaction

1271 Eneutro n ( M e V )

(n, n) (n, n'y)

0.8-2 2.5 0.4-3.5

(n y) (n, 2n) (n, py)

0.8-10 9.5-15 10-20 10 15 10-20

n, a7

23Na cr (mb)

Eexc (MeV)

Ref.

5000 2000 250 300 140 300 300 120 100 35 250 0.25-0.20 0-1200 10-40 "small" "small" 0 5

~ 0.058 0.203 0.375 0.418 0.618 0.651 0.745 0.991 > 1.04 > 6.96 al ~) ~ ") a)

11 11 12

Eneutron (MeV) 0.8-10 3.0 4.0 7.0 0.4 3.5

e (mb) 3300-600 750 300 786 (24) 700 100 40 50 120

11 11 11 13 13 11

0.5-5 12.5-20 12 -20 7.20 6 10 10 --20 14.2 12.6

140- 10 0-100 50 100 34 0- 60 160 20 172 159

E~x,. (MeV) ~i 0.440 ~) 0.440 2.08 2.39 2.64 2.98

> 6.82 a) a) a) ,,7 ~') ~') ~,1

Ref. 11 11 14 15 12

11 11 11 17 11 11 18 19

~) Not specified,

these data we can try to estimate the gross structure of a NaI(T1) pulse height spectrum due to interactions of neutrons of energy 0.5 < E~ < 15 MeV with the scintillation material. Due to the (n, n) and (n, n ' 7 ) reactions there will be a low energy c o m p o n e n t of rather high intensity in the spectra. When the neutron energy decreases after a cascade of elastic a n d / o r inelastic scattering processes, the total capture cross section increases. Therefore we can expect a peak or some structure near the end of the NaI pulse height spectrum corresponding to the neutron separation energies in 24Na or in 12sI a d d e d to the kinetic energy of the impinging neutron. The data of table 1 are, however, insufficient to construct a reliable response function for a large NaI(T1) spectrometer in the energy range of interest. The purpose of the present work is to circumvent this difficulty by presenting experimentally determined response-functions of NaI(T1) detectors of three sizes. The effort was undertaken in connection with a study of the 7-deexcitation energy of neutron evaporation residues for heavy ion induced reactions. A large cylindrical sum-spectrometer consisting of six NaI(T1) segments has been used in order to determine the total energy released by ,/-rays in various reactions. A short description of this device will be given in section 2.1. The neutron response has been determined separately for one sum-spectrometer segment [29.5 kg of NaI(T1)], for 5 segments [147.5 kg of NaI(T1)] and for a standard 12.5 cm diameter × 7.5 cm thick NaI(T1) crystal

(3.3 kg). The experimental procedure is described in section 2. In section 3 we present the results in the form of nomograms. With these nomograms one can obtain approximate response functions also for other NaI(TI) geometries than those presently used.

2. The experimental procedure 2.1. The s u m - s p e c t r o m e t e r

As was mentioned in the introduction, the experiments were carried out for a standard 12.5 cm × 7.5 cm cylindrical NaI(T1) detector and for segments of the sum-spectrometer. In view of the rather special geometry of the latter, a brief description of it is given below. The sum-spectrometer consists of six identical, mechanically separable NaI(T1) detectors (segments) which form together a cylinder with a length of 40 cm, an inner radius of 4 cm and an outer radius of 20 cm, in total 48,250 cm 3 of NaI (cf. fig. 1). Each segment is coupled to an EMI-9791 photomultiplier via a 5 cm perspex light guide. A cylindrical scattering chamber can be m o u n t e d inside the axial hole, allowing the accelerated ion beam to pass through. The total solid angle subtended by the scintillation material as seen from the centre of the spectrometer is 3.9,r sr. The y-ray energy resolution per segment is about 10% (fwhm) at 661 keV, and the fwhm of the time spectrum between two segments is about 5 ns at 511 keV. For further details about the spectrometer see ref. 3.

J.J. van Ruyven et al. / The response of large Nal(TI) detectors

143 5"x2"

a)

TARGET ~ ~ / _ - " -

m

NE 213

----~

~~

~ - BE AM

150

cm

i,]

Na

I(TI)

i

2x 5"x 2" NE 213

Fig. I. Schematic drawing of the sum-spectrometer. Left." sideview with three of the six photomultipliers on top. Right: front-view with the six segments surrounding an inner axial hole with a diameter of 80 mm. Indicated is also the supporting ring in which the six segments are mounted. The outside diameter corresponding to the dashed line is 400 mm.

b)

TARGET

(](-BEAM 4

150 cm

1

2.2. Principle of the experiment T h e ideal way to measure the response of NaI(TI) detectors is to use a set of intensity calibrated, m o n o e n ergetic n e u t r o n sources. Since this ideal set is not available, we adopted a n alternative a p p r o a c h by using a n e u t r o n source of continous energy distribution a n d we d e t e r m i n e d the detector response function per energy bin. In this case one needs: a) a suitable n e u t r o n source with b r o a d energy range, b) a detection system to measure the n e u t r o n energies, c) an absolute n o r m a l i z a t i o n of the n e u t r o n flux per energy bin. In the present work neutrons were produced by a pulsed 25 MeV c~-particle b e a m ( - 1 5 nA) from the A V F cyclolron of the Free University, impinging on a thick target of nSSn stuck to a small c a r b o n b e a m stop. T h e main reactions were " S S n ( a , 2nT) 12°Te and 12C ( a , x n y p y ) . The n e u t r o n spectrum of the first reaction is expected to be similar to a n e u t r o n evaporation spectrum. F u r t h e r m o r e , the nucleus ~2°Te is stable, a n d in y spectroscopic studies of this nucleus using ( a , 2n) reaction, no isomeric states with a half-life longer than 1 ns were observed. The choice of this final nucleus (12°Te) reduced the off b e a m y-ray background. By using an internal burst suppresion system the b e a m b u r s t repetition rate was decreased to a b o u t 5.5 MHz, with a burst width of effectively 2.0 ns fwhm. The n e u t r o n spectra were measured by the time-of-flight technique using NE213 liquid scintillators. The calculated response [4] of these scintillators was used in order to normalize the relative detection efficiency of NaI(TI) a n d NE213. This calibration m e t h o d is valid, however, only for n e u t r o n energies above a b o u t 400 keV.

2.3. The experimental set-up The experimental set-up is shown schematically in figs. 2 a - c for the three NaI(T1) crystal configurations.

C)

5SUMSPECTROMETER SEGMENTS

@-

PARAFFINWAX

I I,I

SUMSPECTROMETER SEGMENT

150cm

I I'I

Fig. 2. The experimental set-up with the three different NaI(T1) crystal configurations, described in the text: (a) Top-view of the arrangement with the 12.5 x 7.5 cm Nal crystal, and with one 12.5 cm x 5 cm NE213 liquid scintillator. (b) One segment of the sum-spectrometer (top-view). Two liquid scintillators are placed symmetrical with respect to the directions of the a-beam to fulfill the normalization requirements mentioned in the text. (c) Side-view of the set-up with five segments of the sum-spectrometer. The two liquid scintillators are in the plane perpendicular to the plane of the drawing.

All objects on which n e u t r o n s could scatter were rem o v e d as m u c h as possible from the vicinity of the experimental set-up, so that the target a n d the detectors were in the centre of a virtually empty cube of a b o u t 7 x 7 x 7 m 3. The distance between the n e u t r o n production target a n d the detectors was 150 cm. In this way it was possible to measure b y means of the time-of-flight technique the n e u t r o n energies between 30 MeV (an u p p e r b o u n d given by the time resolution of the NaI(T1) detectors) a n d 400 keV (a lower b o u n d given also by the repetition of the cyclotron beam). The rf signal of the cyclotron was used as the time reference. O n e 12.5 cm x 5.1 cm cylindrical liquid scintillator (NE213) was used in the m e a s u r e m e n t s with the stand a r d NaI(T1) crystal to fulfil the normalization requirements. A second similar NE213 scintillator was added in the experiments involving sum-spectrometer seg.ments. The angular distribution effects in the n e u t r o n

144

J.J. van Ruyven et al. / The response of large Nal(TI) detectors

spectra were reduced by placing the liquid scintillator(s) a n d the NaI(T1) symmetrically with respect to the direction of the c~-beam.

5x103!

TrMESPECTRUM

NaI

2.4. The measurements

FWHM 5 ns

2.4.1. The standard N a I detector and one segment of the sum -spectrometer A n example of the time-spectrum measured with the NE213 scintillator is given in fig. 3. The T-rays are suppressed by means of pulse shape discrimination [5,6]. A time-spectrum measured with one NaI segment is presented in fig. 4. The two spectra were measured simultaneously. The absolute n e u t r o n detection efficiency of the NE213 scintillators was calculated on the basis of the data from ref. 4. The electronic threshold was set at 10% of the pulse height of the C o m p t o n edge of 662 keV T-rays from a 13VCs source, which corresponds to the m i n i m u m detectable n e u t r o n energy of a b o u t 400 keV. The loss of valid neutron pulses due to possible malfunctioning of the PSD was determined by the comparison of NE213 time-spectra with and without PSD. Since this loss of valid events was found to be 4 (3)% for the highest n e u t r o n energy bin, and decreased to 1 (2)% at 1 MeV neutron energy, this effect was neglected. The time-spectrum between the pulses of the NaI(T1) a n d the b e a m - b u r s t was divided into 32 slices, 5.5 ns wide, corresponding to a n e u t r o n energy width of 4 M e V at E n = 12.5 MeV a n d 100 keV at E n = 1 MeV. The NaI(TI) pulse height spectra were recorded in a

-~1~ En ( M e V ) tlh

r-..

o,l

~

P'.-

CO

N O

CO

I

0 (D

i

oOr" n"'h :E D Z

S S O ,q'

f

/

i

'

F

'l

~, I

ii, Ii I

'

t '

t

O" o~ v tO F-Z D

0 u LL

©

cn D Z

150

100

50

0

~ 1 ~ TIME ( n s )

Fig. 3. The time-spectrum of the NE213 scintillator signals with respect to the beam burst. The peak at At = 5 ns is due to y-rays not completely rejected by the pulse shape discrimination system. The corresponding neutron energies are indicated at the top of the figure.

=

103

! 102

f 1 ~o L . 150

.

.

.

.

.

i

100

50

~---

TIME ( n s )

0

Fig. 4. The time-spectrum of signals from one sum-spectrometer segment with respect to the beam burst.

(256 x 32 channels) e n e r g y - t i m e matrix. The electric threshold in the NaI(T1) pulse height spectra was set at 350 keV in all runs. For convenience, all NaI(Tl) pulse heights will be converted to equivalent y-ray energies. T h e NaI pulse height spectra are corrected for n o n - b u r s t related b a c k g r o u n d by subtracting from them the spect r u m corresponding to the 19-24 ns time slice (i.e. the time slice immediately after the p r o m p t y-rays a n d before the fastest neutron). Since this b a c k g r o u n d spect r u m has low intensity, only a small error is introduced by this procedure. 2.4.2. Measurement with five segments of the sum-spectrometer. Although all three measurements have been performed as described above, one remark should be made a b o u t the five segments case. To be able to compare this m e a s u r e m e n t with real applications of the sum-spectrometer, i.e. with the target placed in the centre of the spectrometer, all segments except for the central one were shielded from the n e u t r o n p r o d u c t i o n target with a paraffin-wax wall (el. fig. 2c). The n e u t r o n energy spectrum as seen by the central segment is slightly affected by this procedure because of the neutron scattering in the paraffin-wax. However, the changes are to a good a p p r o x i m a t i o n the same for the central segment a n d for the two NE213 liquid scintillators a n d are thus automatically corrected for in our normalization procedure. In this set-up the a m o u n t of N a I as seen from the target was about the same as in the one segment case. The a m o u n t of N a l as seen by the sideways scattered neutrons in the central crystal was increased, however, considerably. The event definition was chosen such that the pulse height spectrum in the appropriate time-bin was updated only if the central segment fired. The pulse-height was defined as the hardware sum of the

J.J. van Ruyven et al. / The response of large Nal(TI) detectors

side, and of a low intensity high energy component. The quantities which are most charactertistic for this spectrum are: the centre of gravity of the spectrum, the total area normalized to the number of neutrons, a characteristic point of the high energy edge of area 1 (e.g. the position of the half-maximum, EHM ). We have determined these quantities for all the NaI spectra from the amplitude vs. time-of-flight matrix, and we normalized the measured areas to the total number of neutrons in the corresponding neutron energy bin. In table 2, we summarize the above mentioned quantities for the 12.5 cm × 7.5 cm crystal and for one segment. The position of the high energy edge of the low energy component of the pulse height spectrum increases with the neutron energy (cf. fig. 5). This effect is shown quantitatively in fig. 7 where we display the y-ray energy corresponding to the position of the half maximum of this edge vs. the neutron energy. Clearly the neutron range in which a linear relationship exists between the EHM and E n increases with the size of the NaI crystal. This is a consequence of the fact that

signals from this segment and from all other segments firing within 35 ns after the central one. The central segment determined, of course, the timing, and thus the neutron energy involved.

3.

Results

and

discussion

3.1. General characteristics o f the response functions

Fig. 5 shows the pulse height spectra corresponding to various time-of-flight bins with the 12.5 cm × 7.5 cm NaI crystal and with one segment of the sum-spectrometer, respectively. The random background was subtracted as described in sect. 2. The indicated neutron energies correspond to bins in the spectrum of fig. 3. To illustrate what kind of information can be extracted from this experiment we consider the schematic structure of the NaI pulse height spectrum for an arbitrarily chosen neutron energy (fig. 6). The spectrum consists mainly of an intense low energy component with a relatively well-defined edge on the high energy

101

E~= 11 8 (2.0)

145

10 3

10 2

1

2

3

4

~- PULSE

5

6

7

8

HEIGHT (equivolent

1

2

3

4

y-reyenergyin

5

6

7

8

9

MeV )

Fig. 5. The NaI(TI) pulse height spectra of the 12.5 x 7.5 cm crystal (left) and one segment of the sum-spectrometer (right). The numbers indicate the centres of the neutron energy intervals (in MeV) to which the spectra correspond.

146

J.J. van Ruy~en et al. / The response of large Nal(Tl) detectors

i

"i tn

Y

lO

3o

2O ~<

?

I[

~.//

/ X ~

._~: 5

0

z

4

I S

~ jx~

8

o

• E.y ( N'IeV )

Fig. 6. Schematic NaI(T1) pulse height spectrum corresponding to mono-energetic (4 MeV) neutrons. The meaning of the symbols is given in the text.

m u l t i p l e scattering o f n e u t r o n s o n 23Na a n d 127I is the m a i n p r o c e s s by w h i c h the n e u t r o n s lose their kinetic e n e r g y m the crystal. O n the basis o f the k n o w n cross sections for inelastic ( - 2000 mb, see table 1) a n d for elastic scattering ( - 4500 mb, see table 1) we can estim a t e that o n the average a 4 M e V n e u t r o n loses 90% o f its kinetic energy in six inelastic a n d a b o u t 30 elastic scattering events, w h i c h can take place within 70 ns. Such a multi-step scattering p r o c e s s can very well occur in the larger NaI(T1) c o n f i g u r a t i o n s . (The time c o n s t a n t o f the pulse height analysing s y s t e m was set at 1 ffs.) T h e s e c o n d characteristic feature o f the s p e c t r a disp l a y e d in figs. 5 a n d 6 is the low intensity high energy b u m p . T h e energy o f this b u m p is close to the n e u t r o n kinetic energy plus the n e u t r o n s e p a r a t i o n energy in either 24Na ( S n = 6.96 MeV, Ocapt ~< 10 m b at E , >_ 1 M e V ) or 128I(S n = 6.82 MeV, Ocapt _S<75 m b at E , > 1 MeV. The origin of the b u m p can be ascribed to the total n e u t r o n a b s o r p t i o n in the crystal in either a single

i

0

i

5

10

15

• ENEUTRO N ( M e V )

Fig. 7. The position of the half-maximum of area 1 (of. fig. 6), in the corresponding 7-ray energy scale, vs. the neutron energy.

step (n, ~,) p r o c e s s or, m o r e likely, in a s e q u e n c e of inelastic scattering events followed by a (n, 7 ) reaction. T h e intensity ratio b e t w e e n the two c o m p o n e n t s in the s p e c t r a given in fig. 5 can be r e a s o n a b l y well e x p l a i n e d b y the ratios o f the cross sections for inelastic a n d elastic scattering given in table 1. It should be n o t e d h e r e that c a p t u r e of t h e r m a l i z e d n e u t r o n s has a m u c h larger cross section, b u t it takes several ms to thermalize a 0.5 M e V n e u t r o n . T h e r m a l n e u t r o n c a p t u r e t h e r e f o r e will a d d only to the r a n d o m b a c k g r o u n d a n d is r e m o v e d f r o m our d a t a by the b a c k g r o u n d s u b t r a c t i o n p r o c e d u r e d e s c r i b e d in section 2.4.1. 3.2. Nomograms T h e r e s p o n s e of N a I to n e u t r o n s with 0.70 (7) ~< 11.5 (1.5) M e V is s u m m a r i z e d in figs. 8a a n d 8b for the 12.5 c m × 7.5 c m crystal a n d for the 5 - s e g m e n t configura-

Table 2 Properties of neutron induced pulse height spectra for the 12.5 x 7.5 cm NaI crystal and a sum-spectrometer segment. Neutron energy (MeV)

Centre of gravity (MeV)

Half-maximum (MeV)

12.5 cm X 7.5 cm

1 segment

12.5 cm × 7.5 cm

1 segment

12.5 cm × 7.5 cm

1 segment

13.8 9.9 9.9 -7.4 7.4 -5.81 5.81 -4.65 4.65 -3.81 3.81 -3.17 3.17 -2.69 2.69 -2.30 2.30 -2.00 2.00 1.75 1.75 -1.54 1.54 -1.37 1.37 -1.23 0.830-0.650

2.31 (8) 2.03 (8) 1.87 (6) 1.68 (6) 1.51 (6) 1.44 (5) 1.39 (6) 1.33 (4) 1.26 (4) 1.20 (4) 1.30 (4) 1.30 (6) 1.38 (8) 2.30 (15)

3.76 (7) 3.95 (5) 3.01 (4) 2.40 (4) 2.13 (3) 1.90 (3) 1.71 (3) 1.59 (4) 1.42 (2) 1.25 (3) 1.15 (3) 1.04 (4) 1.01 (4) 1.82 (10)

3.68 (10) 3.51 (10) 3.15 (9) 2.76 (9) 2.51 (9) 2.48 (8) 2.30 (8) 2.33 (8) 1.97 (9) 1.81 (10) 1.57 (14) 1.37 (15) -

6.08 5.56 4.78 4.19 3.70 3.21 2.92 2.58 2.30 2.02 1.68 1.47 1.37 0.65

0.224 (35) 0.176 (20) 0.170(14) 0.149(11) 0.122 (11) 0.109(10) 0.099 (8) 0.078 (8) 0.071 (9) 0.065 (9) 0.043 (8) 0.050 (8) 0.043 (8) 0.068 (11)

0.375 (39) 0.300(30) 0.290(30) 0.263 (28) 0.230 (25) 0.224 (23) 0.209 (18) 0.177(15) 0.177 (15) 0.184(14) 0.152 (14) 0.158 (13) 0.120(15) 0.073 (18)

Number of Nal pulses/neutron

(10) (8) (8) (7) (8) (7) (6) (6) (5) (5) (3) (3) (4) (5)

J.J. van Ruyven et al. / The response of large NaI(TI) detectors

147

l°° a)

b)

TOTAL

TOTAL 5

10-I i 5"x

CRYSTAL

/

o 10 -2

10_3

z

3"

SEGMENTS

Q

035

-

135

a

Q35

-

1.35

b

135

-

235

b

1 35

-

235

-

335

c

2.35

-

335

4.35

d,

335

-

435

c

2.35

a.

3.35

e.

4.35-

535

e

4 35

-

535

f

5.35

-

63,5

f

535

-

635

g

6.35

-

735

g

635

-

735

h.

7.35-

835

h

735

-

835

9.35

i

835

-

935

835

[D Z

-

g

I' 10 -4

llliHl

i

I

J

i

101

i

i i LLIL

I

[

10 2

E n(MeV)

l

I Illl

i

i llpltl

L

I ~

J

lltllll

i

101 E n (M~V)

J

t Ji1~ll

l

t

L tttt

102

Fig. 8. (a) The response of the 12.5 ×7.5 cm NaI(TI) crystal to neutrons in the energy range 0.7 < E n ~<13 MeV. The curves labeled a-i correspond to energy-bins in the Nal pulse height spectra, with a width of 1 MeV. (b) Idem, for 5 segments of the sum-spectrometer. The uncertainties of the experimental points include both the individual statistical errors and the systematical one. The main contribution to the latter are due to the spread in the flight path caused by the finite detection thickness and to the uncertainty in the NE 213 detection efficiency. tion, respectively. In these plots we give the number of N a I pulses per incoming neutron in given pulse height intervals as functions of the neutron energy. These nomograms make it possible to construct the pulse height spectrum resulting from an irradiation of a NaI(T1) crystal with neutrons having an arbitrary known spectrum in the range 0.7 _< E , < 13 MeV. The constructed pulse-height spectrum will consist of 9 channels with 1 MeV resolution, starting at 350 keV. It should be noted here that the threshold in our NaI(T1) pulse-height spectra, equivalent to 350 keV y-ray energy, implies that the number of interactions is underestimated. Two effects have to be considered. In the first place, pulses due to a low number of elastic scattering events on 23Na or inelastic scattering to the first excited state in 127I generally do not pass the threshold. This effect contributes to a systematic error for all three set-ups. The second effect is of importance for the case of the 5-segment configuration and for the resulting corrections for the sum spectra. Neutrons can be scattered in the central segment with the resulting pulse height below the threshold. The scattered neutrons might enter in one of the neighbouring segments and the following interaction might result in a large total sum pulse. We have estimated the corrections due to both effects, using the experimental differential neutron cross-sections of table 1. At low neutron energy ( E , - 1 MeV) the second effect gives the largest contribution to the systematic error in the response function for the 5-segment config-

uration, whereas at high neutron energy ( E n - 10 MeV) both effects are of equal importance. The estimated systematic errors are 6% ( E . = 1 MeV) and 3% ( E . = 10 MeV), respectively.

3.2. An example of application As an illustration of how the nomograms of fig. 8 can be used for correcting experimental data for the contributions due to neutrons we consider the sumspectrometer spectra measured for the reaction 15°Nd(lZC, xny)162-XDy at E ~ 2 c = 9 5 MeV. At this projectile energy the 6n and 7n channels are dominant. The angle integrated neutron energy spectra for both reaction channels have been measured previously [8]. In a recent experiment (cf. ref. 9) we measured the total y-decay energy for the different reaction channels with the sum-spectrometer in coincidence with a Ge (Li) detector placed at 180 ° with respect to the beam direction. The sum spectra for different reaction channels (cf. fig. 9) were produced by setting off-line energy gates in the Ge(Li) spectra on peaks due to known transitions in the final nuclei. F r o m the neutron spectra given by Hageman [8] and the data from fig. 8b we calculate that the chance that a neutron interacts with the NaI is 18.7 (9)% for neutrons in both the 6n channel and the 7n channel, with an average energy deposited per interacting neutron in the sum-spectrometer of 1.6 (2) MeV and 1.5 (2) MeV, respectively. The correction for the average

148 tt) p-

Z 73 0 U L 0 or" W m

J.J. v a n R u y v e n et al. /

560

17.4

T h e response o f large N a l ( T I ) detectors

20.2

400 240

7n//~

\ \ \ 6n

80 Z

10 20 30 I~ ENERGY ( M e V )

40

Fig. 9. The experimental sum spectra for the reactions ]5°NaI(12C, 7ny) 155Dy (drawn line) and 15°Na(t2C,6ny]56Dy (dashed line). The arrows labeled 17.4 and 20.2 indicate the centres of gravity of the experimental spectra (in MeV); the arrows labeled G(7n) and G(6n) indicate calculated centres of gravity of the entry populations (Grogi2). total y-decay energy measured with the sum-spectrometer is given by: ....

= 7, . . . . . .

- g2nkff~,

where ~;u . . . . r is the measured centre of gravity of the sum-spectrometer pulse height distribution, I2 is the solid angle s u b t e n d e d by the sum-spectrometer, n is the n u m b e r of n e u t r o n s per reaction, k is the probability that a n e u t r o n interacts with the NaI, and /Y is the average energy deposited in the sum-spectrometer per interacting neutron. The total correction is thus for the 6n channel 1.8 (2) MeV, a n d for the 7n channel 2.0 (2) MeV. The centres of gravity of the total y-decay distrib u t i o n become 18.5 (1.2) MeV a n d 15.5 (1.2) MeV, respectively. For comparison we give the results of a statistical model calculation for this reaction (Grogi2 [10]), yielding 2: = 16.3 MeV and 2: = 12.7 MeV for the 6n a n d 7n reactions, respectively. The calculations were performed using the s t a n d a r d set of Grogi input parameters, without any specific parameter fitting.

4. Conclusions In this study we performed experiments to determine the response of large NaI(T1) crystals to neutrons with energies typical for heavy ion induced reactions (0.7 _< E , _< i3 MeV). W e investigated the pulse height distrib u t i o n s and yields for three different N a I crystals irradiated with n e u t r o n s as functions of the n e u t r o n energy and the crystal size. The effect of the crystal size was studied by comparing the n e u t r o n response of a cylindrical (12.5 cm diameter x 7.5 cm thick) N a I crystal and an oblong (40 cm length, 8 cm width, 16 cm thick) crystal (cf. figs. 1 a n d 2). A third experiment was performed on the same oblong crystal, sandwiched between similar large N a l crystals. The n u m b e r of N a I pulses per incoming neutron resulting in pulse-height not smaller t h a n that equivalent to a 350 keV y-ray determined as 0.10 (1) for

the 12.5 cm × 7.5 cm crystal, and 0.21 (3) for both the 1-segment a n d the 5-segment set-up. The average energy per detected n e u t r o n was 1.63 (6) MeV for the standard detector, 2.14 (6) MeV for one segment, and 2.50 (5) MeV for the five segments. These average n u m b e r s are, however, strongly d e p e n d e n t on the neutron energy spectrum. One could argue that the differences in average detected energy are due to differences in y-ray efficiency of the three crystal configurations. The shapes of the N a l pulse height spectra indicate, however, that inelastic scattering of neutrons to levels in 25Na and 12vl is the main interaction process, and that an imp o r t a n t fraction of the neutrons lose their kinetic energy via multiple scattering effects. This explains very well the increase in the average deposited energy with increasing a m o u n t of N a | . The shapes of the pulse height spectra are strongly d e p e n d e n t on the neutron energies. In fig. 8 we give therefore the n u m b e r of N a I pulses per incoming neutron in one out of nine pulse height intervals as a function of the n e u t r o n energy. O n the basis of these n o m o g r a m s a 9-channel NaI spectrum can be constructed for the N a l crystals described in the text. A crude interpolation procedure allows for use of these d a t a also for other N a I crystals of dimensions comparable to those used in the present study.

References [1] C.M. Bartle, Nucl. Instr. and Meth. 124 (1975) 547. [2] G.H. Share, J.D. Kurfers and R.B. Theus, Nucl. Instr. and Meth. 148 (1978) 531. [3] J.J. van Ruyven, Thesis, Free University, Amsterdam, Holland. [4] M. Drosg, Nucl. Instr. and Meth. 105 (1972) 573. [5] P. Sperr, H. Spieler, M.Q. Maier and D. Evers, Nucl. Instr. and Meth. 116 (1974) 55. [6] J..Bialkowski and J. Szczepankowski, Nucl. Instr. and Meth. 153 (1978) 589. [7] D. Witmore and P.E. Hodgson, Nucl. Phys. 55 (1964) 673. [8] D. Hageman, Thesis, University of Groningen, Groningen, Holland (1981). [9] F. Schuling, K.V.I. internal report, University of Groningen, Groningen, Holland (1981). [10] J. Gilat, Grogi2, BNL 50246 (T-280), Brookhaven National Laboratory, New York (1970). [11] J.R. Stehn, M.D. Goldberg, B. Majurno and R.W. Chasman, Report BNL-325, Brookhaven National Laboratory, New York (1964). [12] D.A. Lind and R.B. Day, Ann. Phys. (N.Y.) 12 (1961) 485. [13] R. Bars, W. Kessel and G. Majoni, Nucl. Instr. and Meth. 30 (1964) 237. [14] E.N. Shipley, G.E. Owen and L. Madansky, Phys. Rev. 115 (1959) 122. [15] J.H. Towle and R.O. Owens, Nucl. Phys. AI00 (1967) 257. [16] J.H. Towle and W.B. Gilbay, Nucl. Phys. 39 (1962) 300. [17] C.F. Williamson, Phys. Rev. 122 (1961) 1877. [18] J. Picard and C.F. Williamson, Nucl. Phys. 63 (1965) 673. [19] G. W61fer and M. Bormann, Z. Physik 194 (1966) 75.