ARTICLE IN PRESS

0168-9002/$ - se

doi:10.1016/j.ni

�Correspondi

fax: +709621 6

E-mail addre

Nuclear Instruments and Methods in Physics Research A 540 (2005) 430–436

www.elsevier.com/locate/nima

Neutron response function for a detector with 3He counters forthe 0.39–1.54 MeV neutron energy range

J. Cernya, Z. Dlouhyb, Z. Dolezala, M.P. Ivanova,c, P. Kubika, E.A. Kuznetsovac,Yu.E. Penionzkevichc, E.A. Sokolc,�, J. Svejdaa, I. Wilhelma

aInstitute of Particle and Nuclear Physics, Faculty of Mathematics and Physics, Charles University, 2180 00 Prague 8, Czech RepublicbNuclear Physics Institute, Academy of Sciences of the Czech Republic, 250 68 Rez near Prague, Czech Republic

cFlerov Laboratory of Nuclear Reactions, Joint Institute for Nuclear Research, 141980 Dubna, Moscow Region, Russian Federation

Received 2 August 2004; received in revised form 14 October 2004; accepted 11 November 2004

Available online 7 January 2005

Abstract

An experimental study was carried out on the neutron response functions of two neutron detector arrays consisting of

39 3He proportional counters with a polyethylene moderator for monoenergetic neutrons within the 0.39–1.54 MeV

neutron energy range. Experimental data on the sensitivity of neutron counting to a change in neutron energy and the

influence of the thickness of polyethylene moderator were obtained. The experimental efficiency curves were compared

with the calculated response functions generated by a neutron transport code.

r 2004 Elsevier B.V. All rights reserved.

PACS: 28.20.Gd; 29.40.Cs; 25.55; 23.40.�s

Keywords: Neutron detector; Delayed neutron emission

1. Introduction

Studies of delayed neutron emission fromneutron-rich nuclei are carried out by usingdifferent types of detectors. These studies requirea neutron detection system of high efficiency, goodenergy and angular resolution and with thenumber of channels sufficient to enable multi-

e front matter r 2004 Elsevier B.V. All rights reserve

ma.2004.11.047

ng author. Tel.: +709621 64843;

5083.

ss: sokol@jinr.ru (E.A. Sokol).

neutron decay measurements. The main advan-tages of a polyethylene moderated multi-counterneutron detector system are a practically zeroenergy threshold, the absence of cross-talk, a lowsensitivity to gamma quanta and the fact that itcan be made of detectors of high neutronsensitivity [1–4]. Systems with 3He proportionalcounters can be used for obtaining the averageenergies of the delayed neutron spectra forchemically and mass-separated sources [5] andthe angular distributions of neutrons [6].

d.

ARTICLE IN PRESS

J. Cerny et al. / Nuclear Instruments and Methods in Physics Research A 540 (2005) 430–436 431

Delayed neutrons from different emitters rangein energy from hundreds of keV up to severalMeV, for example, single delayed neutrons for14Be have energy of 290 keV [7], and the energy ofdelayed neutrons from 15B takes several values, themain ones being 1.77 and 3.2 MeV [8]. This impliesthat the sensitivity of a neutron detector shouldremain constant in a broad energy range or thedetector should provide information on neutronenergy to correct measured results. Differentapproaches to the calibration of neutron detectorsare described in papers: the neutron-sources Am/Li [3,4], RaBe, 124SbBe, 24NaBe, 24NaD2O [5] ofspontaneous fission of 252Cf [5,10,11] or sources ofdelayed neutrons [9,10] were used.

In the present work, we have measured theresponse functions of two neutron detector arrays,with 39 3He proportional counters each, formonoenergetic neutrons within the 0.39–1.54MeVneutron energy range. The main goals of themeasurements were to obtain experimental dataon the sensitivity of neutron counting to varyingneutron energies and the influence of the thicknessof the polyethylene moderator. The experimentaldata will be used for constructing optimal neutrondetector arrays with modules [6] for studies ofthe properties of delayed neutrons from fission

9

Fig. 1. Scheme of 3He neutron detector arrays: (a) with complete mod

hexahedrons are moderator units, black and white circles are counter

fragments within the framework of Dubna Radio-active Ion Beams (DRIBs) [12] and AccelerateurLineaire aupres du Tandem d’Orsay (ALTO) [13]projects.

2. Neutron detector arrays

The detection assemblies are shown in Fig. 1,each array consisting of 39 modules [6] arranged insix layers. In the first assembly, all the moduleswere identical and each of them consisted of ahexagonal polyethylene unit (distance betweenparallel planes—50 mm) with a cylindrical cavity(diameter 32 mm) for a counter of neutrons. Thelength of the counter and the polyethylenemoderator was 50 cm. The counters were filledwith 3He up to a pressure of 7 atm. Every counterof neutrons was supplied by a pre-amplifier whichwas hardly fastened directly on a counter in acylindrical case; the cases of pre-amplifier wereused for fixing of counters of neutrons inassembling.

In the second assembly, the polyethylene mod-erator was removed from the modules of the 2ndand 4th rows.

9

erator and (b) without moderator in the 2nd and 4th rows. Gray

s.

ARTICLE IN PRESS

J. Cerny et al. / Nuclear Instruments and Methods in Physics Research A 540 (2005) 430–436432

The detection assemblies were surrounded by ashield from boron-loaded polyethylene with awindow in the front side; this window was coveredby 1mm cadmium foil.

Fig. 2. Two-dimensional spectrum of total charge versus

charge in the tail of a pulse obtained on a stilben crystal for

neutron with the energy En ¼ 0:93MeV:

3. Energies of neutrons and intensity monitoring

The experiments described here were performedwith the Van de Graaf electrostatic acceleratorHV2500AN at Charles University, Prague, CzechRepublic.

Monoenergetic neutrons were produced assecondary particles from a 3H(p,n)3He reactionon a Ti–T target (0.2 mg/cm2) with a molybdenumbacking. The target was mounted at an angle of451 to the incident beam of protons, so the realthickness of the target was about 0.28 mg/cm2. Theenergies of protons we used and the correspondingneutron energies are shown in Table 1.

The energy spread of protons due to the energyloss in the target and the corresponding energyspread of neutrons were calculated to be not morethan 30 keV.

The proton energy spread for the beam of theaccelerator was not more than 2 keV. Eachneutron detector array was mounted at a distanceof 3 m from the tritium target and the center of thearray was placed at 01 to the beam axis, neutroncounters were located perpendicularly to this axisand were seen at an angle of 761 from the target.The neutron energy spread due to the angularspread was about 10 keV.

Measurements of the energy of neutrons andrelative intensity monitoring were made with astilben crystal detector of diameter 3.0 cm andthickness 2.2 cm connected to a Philips XP-2020photomultiplier tube and placed in front ofneutron detector arrays. Using a stilben scintilla-tion crystal gives an efficient way to discriminatebetween incident gamma rays and neutrons by

Table 1

Neutron energies En vs proton energies Ep

Ep (MeV) 1.2 1.3 1.4 1.7

En (MeV) 0.39 0.50 0.61 0.9

means of pulse shape discrimination (PSD). Weapplied a PSD method, which is based on chargeintegration of the pulse current over two differenttime intervals using a charge-integrating QDC.The method is described in Ref. [14] and the sameelectronics was used in the current experiment.

Plotting the charge in the tail of a pulse againstthe total charge gives the two-dimensional spectrain Fig. 2 for incident proton energy of 1.7 MeVand a respective neutron energy of 0.93 MeV. Onecan see a separation between neutrons (topbranch) and g-rays (bottom branch) in the figure.Making a contour around the neutron branch andprojecting it on the x-axes, we obtain a one-dimensional energy spectrum of neutrons. In otherwords, we determine the experimental neutronresponse function and can calculate the neutroncounts for the respective branch. The correspond-ing one-dimensional spectrum is shown in Fig. 3.

The process of neutron scattering on protons atneutron energy of up to 10 MeV is isotropic in acenter-of-mass coordinate system. So the responsefunction of a detector based on the simple

1.8 2.0 2.2 2.3

3 1.03 1.23 1.44 1.54

ARTICLE IN PRESS

Fig. 3. One-dimensional spectra obtained on a stilben crystal

for neutron with the energy En ¼ 0:93MeV:

Fig. 4. Plot of relative extrapolated Enextr neutron energies

versus proton energies.

J. Cerny et al. / Nuclear Instruments and Methods in Physics Research A 540 (2005) 430–436 433

hydrogen scattering of monoenergetic neutrons inthis energy range should have a rectangular shape,extending from zero to the full incident neutronenergy [15]. But due to different distortionmechanisms [16], we can see that the one-dimen-sional spectrum has a significant deviation from apure rectangular shape. The relative neutronenergies determined using the crossing of theextrapolated right part of the spectra with the x-axes are shown in Fig. 4. One can see that the

accuracy of determination of neutron energieswith this method is quite good and we really havemonoenergetic neutrons with an energy spread ofnot more than about 35 keV.

4. Neutron intensity monitoring with a 3He counter

We also used 3He proportional counters withouta moderator for relative monitoring of intensity ofincident neutrons at detector arrays. A 3Heproportional counter was placed in front ofneutron detector arrays along the line connectingthe center of the array with the target. To estimatethe neutron intensity, we integrated the amplitudespectra of the pulses from the counter. Thisintegral is mainly connected with the thermalneutrons originating in the array, floor and wallsin the experimental room because of the very smallsensitivity of 3He-counter to neutrons of higherenergies. But these thermal neutrons were origi-nated from neutrons from reaction on tritiumtarget, and so the intensity of these thermalneutrons reflects the intensity of primary neutrons.We obtained good agreement in neutron intensitymonitoring for both types of detectors used.

5. Results

The counting rates of the separate groups ofneutron counters arranged in the central part ofeach layer of a detection assembly and a monitordetector were measured at different energies ofincident neutrons. In Fig. 1(a,b), these groups ofneutron counters are marked with a black tone.The distributions of the counting rates (recalcu-lated per one counter and normalized to themonitor counting rate) are shown in Fig. 5(a,b)for neutron detector arrays with a completemoderator and without a moderator in the 2ndand 4th rows, respectively. The error of everypoint of the response function is less than the pointsize. The good sensitivity of both arrays to thechanging of neutron energies is clearly seen, so thevalid correlation of count rate ratios for differentrows of neutron counters with neutron energieswas obtained. If the detection sensitivity versus

ARTICLE IN PRESS

1 2 3 4 5 6

0.02

0.1L

og

of

Inte

nci

ty,a

.u.

row number

En610En930En1230En1540En1440

Array 1: with moderator in each row

1 2 3 4 5 6

0.02

0.1

row number

En610En930En1230En1540En390

Array 2: without moderator in 2 and 4 rows

Fig. 5. Neutron response functions (distributions of the counting rates recalculated per one counter and normalized to the monitor

counting rates) for a detector array (a) with complete moderator and (b) without moderator in the 2nd and 4th rows at different

energies of neutrons.

J. Cerny et al. / Nuclear Instruments and Methods in Physics Research A 540 (2005) 430–436434

neutron energy curve for one row of counters isquite different from that for another row ofcounters, the ratio of counts for the two rows willbe a function of neutron energies.

It is seen from a comparison of the two responsefunctions (Fig. 5a,b) that removing part of themoderator leads to the displacement of themaximum of the rate to the deeper layers ofthe counters and the total detector sensitivitydrops with the decrease of the volume of themoderator.

Plots of the sensitivity of neutron detectionversus incident neutron energy (in recalculationper one counter) independent of the number ofdetecting rows are shown in Fig. 6. The sensitivityof the first array with all its counters is taken as100% at 1540 keV of incident neutron energy. Asis visible, for detection assembly with a largeramount of moderator the sensitivity of registrationis systematically higher for all neutron energies.

As neutron energy increases, neutron registra-tion sensitivity increases. This can be explained bythe fact that at low energies, a considerablefraction of neutrons is reflected by the front layersof a detector and does not penetrate into theassembly, neutrons of higher energies penetratedeeper into the assembly and their losses due toreflection are smaller. 4p detectors will registerneutrons of small energies more effectively, be-cause they do not abandon a detector due to thereflections in other internal layers, but neutrons ofhigh energies will leave the assembly.

The experimental sensitivity curves were com-pared with calculated response functions generatedusing the MCNP code [17] version B. Duringcalculations good agreement with experimentalinformation was achieved, calculations with alarger amount of moderator in a detector wereconducted, the distance between the parallel planesof polyethylene hexahedrons being changed from 5to 6, 7 and 8 cm. In Fig. 7, the calculateddependences of the sensitivity of neutron registra-tion on the thickness of the moderator for twoneutron energies (0.61 and 1.54 MeV) are pre-sented. As is obvious from the graphs, the distancebetween the parallel planes of polyethylene hex-ahedrons about 6.5 sm is optimal. Taking intoaccount that in experiments with 4p detectors,neutrons get into a detector isotropically, theeffective thickness of moderator in our modules isnear optimal. It is also necessary to notice that anincrease in the thickness of the moderator resultsin an increase in the neutron lifetime for adetector, and so experiments on studying delayedneutrons in coincidence with b-particles or g-raysor in multiple neutron measurements result in anincrease of the background from random coin-cidences.

Consequently, to ensure that neutron registra-tion sensitivity depends weakly on the energy ofincident neutrons up to an energy of 1.5 MeV, adetector of neutrons must consist of no less than4–5 layers of modules, consisting of neutrondetectors (7 atm 3He, 32mm in diameter) and

ARTICLE IN PRESS

0.2

0.3

0.4

0.5

Effi

cien

cy

Array 2 1 row

Array 1 1 row

Array 2 2 rows

Array 1 2 rows

0.5

0.6

0.7

0.8

Effi

cien

cy

Array 2 3 rows

Array 2 4 rows

Array1 3 rows

Array 1 4 rows

600 900 1200 15000.7

0.8

0.9

1.0

Effi

cien

cy

neutron energy, keV

Array 2 5 rows

Array 2 6 rows

Array 1 6 rows

Array 1 5 rows

Fig. 6. Efficiency of neutron detection as a function of the

energy of incident neutrons.

Fig. 7. Calculated efficiency of neutron detection as a function

of the thickness of moderator.

J. Cerny et al. / Nuclear Instruments and Methods in Physics Research A 540 (2005) 430–436 435

moderator surrounding one detector by a layerapproximately 1 cm thick.

Neutrons of higher energies—from 3 to5MeV—can be obtained in a (d,d) reaction onthe same accelerator and experiments on determi-nation of the response function of an assembly ofneutron counters for this range of energies will beconducted.

Acknowledgements

The authors express their gratitude to J. Brozand Yu.A. Usov for their help. The financialsupport from the Grant Agency of the CzechRepublic under Grant Nrs. 202/00/0899 and 202/03/0831 is gratefully acknowledged.

References

[1] J. Shergur, B.A. Brown, V. Fedoseyev, et al., Phys. Rev. C

65 (2002) 034313.

[2] A. Wohr, B. Pfeiffer, K.-L. Kratz, et al., Eur. Phys. J. 13

(2002) 281.

[3] U.C. Bergmann, L. Axelsson, M.J.G. Borge, et al., Nucl.

Phys. A 658 (1999) 129.

[4] D.J. Loaiza, Nucl. Instr. and Meth. A 422 (1999) 43.

[5] P.L. Reeder, J.F. Wright, L.J. Alquist, Phys. Rev. C 15

(1977) 2098.

[6] E.A. Sokol, V.I. Smirnov, S.M. Lukyanov, Yu.E.

Penionzkevich, Nucl. Instr. and Meth. A 400 (1997) 96.

[7] N. Aoi, K. Yoneda, T. Nakamura, et al., Z. Phys. A 358

(1997) 253.

[8] R. Harkevich, D.J. Morrissey, B.A. Brown, et al., Phys.

Rev. C 44 (1991) 2366.

[9] A.T. Reed, O. Tarasov, R.D. Page, et al., Phys. Rev. C 60

(1999) 024311.

[10] J.C. Wang, P. Dendoovev, M. Hannawald, et al., Phys.

Lett. B 454 (1999) 1.

[11] O. Sorlin, D. Guillermaud-Mueller, A.C. Mueller, et al.,

Phys. Rev. C 47 (1993) 2941.

ARTICLE IN PRESS

J. Cerny et al. / Nuclear Instruments and Methods in Physics Research A 540 (2005) 430–436436

[12] Yu.Ts. Oganessian,’’DRIBs’’: The Dubna Project for

Radioactive Ion Beams, http://159.93.28.88/flnr/in-

dex.html

Yu.Ts. Oganessian, in: Fourth International Conference

on Dynamical Aspects of Nuclear Fission, Casta-Papier-

nicka, Slovak Republic, 19–23 October 1998.

[13] S. Essabaa, J. Arianer, P. Ausset, et al., Nucl. Instr. and

Meth. B 204 (2003) 780.

[14] J. Cerny, et al., Nucl. Instr. and Meth. A 527 (2004)

512.

[15] H.A. Bethe, P. Morrison, Elementary Nuclear Theory,

Wiley, New York, 1956.

[16] V.V. Verbinski, et al., Nucl. Instr. and Meth. 65

(1968) 8.

[17] J.F. Briesmeister (Ed.), Program MCNP, Version 4A, Los

Alamos, Report LA-12625-M, 1993.