ACCELERATOR BASED MONOCHROMATIC NEUTRON SOURCES

 

M. Drosg

Institut für Experimentalphysik, University of Vienna, A-1090 Wien, Austria

E-mail: Manfred.Drosg@univie.ac.at

 

 

ABSTRACT

 

Accelerator-based monochromatic neutron sources are reviewed with special emphasis on the requirements for applied purposes. From the 33 neutron source reactions covered by the computer code DROSG-96 [1] only 3 have received recent attention. So new developments are presented for the classical reactions p-T, d-T and for p-7Li, the latter being a main candidate for BNCT (Boron Neutron Cancer Therapy) applications.

1. INTRODUCTION

Three years ago the production of monoenergetic neutrons in the energy range from 100 eV to 200 MeV using accelerators was fully covered in a review paper [2] with special emphasis on low neutron energies. The computer code DROSG-96 [1] presented there can calculate the kinematic properties, differential cross sections and (thick target) neutron yields for 33 mono-energetic neutron production reactions. Since then new data for 3H(p,n)3He, 7Li(p,n0)7Be and 3H(d,n)4He have become available. They are compared with the predictions of the code, and their potential to improve the database of the code is discussed.

2. PROPERTIES OF MONOENERGETIC NEUTRON SOURCES

2.1. Reaction characterization

Two-body reactions with projectiles from accelerators are a convenient source of monoenergetic neutrons. Beams of hydrogen nuclei are superior to other charged particle beams because the energy loss of such beams is minimum and consequently the neutron yield maximum. Thus only reactions of the form kH + A = n + B + Q (k=1,2,3), where the Q-value can be positive or negative, are considered here. Among these reactions three categories can be distinguished:

a. Reactions with positive Q-values: e.g. 2H(d,n)3He, 2H(t,n)4He,3H(d,n)4He

b. (p,n)-reactions: e.g. 3H(p,n)3He, 7Li(p,n)7Be, 11B(p,n)11C

c. Inverted (p,n)-reactions: e.g.1H(t,n)3He,1H(7Li,n)7Be,1H(11B,n)11C, 1H(13C,n)13N, H(15N,n)15O

There is special interest in kinematically collimated neutron beams. Such beams occur when the velocity of the center-of-mass (c.m.) is larger than the c.m. velocity of the neutron (at 0) so that there will be two neutron groups at 0:

- the primary neutrons, corresponding to 0 c.m. emission and

- the secondary, satellite group from 180 c.m. emission.

This happens in (p,n) reactions in a small energy range above the threshold and in all inverted (p,n) reactions over the entire energy range. In such cases all neutrons are emitted into a forward cone enhancing the laboratory cross sections dramatically.

Fig. 1 shows the half- opening angle of the cones of the 7Li(p,n)7Be reaction in dependence on the beam energy minus threshold energy. Very close to threshold the opening angle becomes very small, i.e. the beam is very narrow. This has two consequences:

a) The shape of a thick target yield curve in the double-valued energy region depends on the opening angle of the detector because the narrow beam does not necessarily illuminate the entire detector. This effect is shown in Fig. 2 for the case of 7Li(p,n0)7Be with opening angles of 2 and 5.

b) The shape also depends on effects (angular straggling, elastic proton and neutron scattering) that change the outgoing neutron direction with regard to the incoming beam direction. So a narrow neutron beam can be deflected so much that it does not hit the detector at all. Therefore it is not surprising that the maximum of an actually measured 0 thick-target yield curve [3] is flatter and less pronounced than shown in Fig. 2.

Fig. 1. Half-opening angle versus proton energy

Fig. 2. Thick-target yields near threshold for a 2 opening angle (full curve) and a 5 opening angle (dashed)

2.2. Kinematics, cross sections and yields

For several years the computer code DROSG-96 [1] has been available that allows the calculation of kinematic properties, of cross sections and of yields of 33 neutron producing reactions (see Table 1.). The yield is calculated for isotopically pure targets unless indicated otherwise. The predictions depend on tabulated cross sections taken from the literature and sometimes also on own evaluations like in the cases 3H(p,n)3He,2H(d,n)3He, 3H(d,n)4He, and 7Li(p,n0)7Be.

There are two classes of cross section data:

a) Complete angular distributions of differential cross sections (in some energy range) for: 3H(p,n)3He, 2H(d,n)3He, 3H(d,n)4He, 7Li(p,n0)7Be, 7Li(p,n1)7Be*, 9Be(p,n0)9B, 11B(p,n0)11C, 13C(p,n0)13N and 15N(p,n0)15O.

b) Differential cross sections at 0 and/or 180 (In some cases just isotropic approximation from integrated cross sections): 6Li(p,n0)6Be, 10Be(p,n0)10B, 10B(p,n0)10C, 14C(p,n0)14N, 36Cl(p,n)36Ar, 39Ar(p,n0)39K, 59Co(p,n)59Ni, 7Li(d,n)8Be, 11B(d,n)12C, 13C(d,n)14N, 15N(d,n)16O, 18O(d,n)19F.

Therefore, in some cases only a crude estimate of the cross section is available. For the kinematics nuclear masses from the 1995 mass evaluation [4] are used. The yield calculations depend in addition on energy loss tables (taken mainly from Ziegler [5]).

Table 1. Main menu of the computer code DROSG-96. En0max gives the upper limit of the mono-energetic energy range except for the multiline (d,n) reactions where for Ed=0 MeV the maximum number of neutron lines due to excited levels is given.

ID

REACTION

REMARKS

En0max

CROSS SECTION

 

TYPE

 

 

RANGE [MeV]

1

3H(p,n)3He

isotopically pure target

7.585

1.0191-32.80/318.

101

 

T2O target(ice)

 

 

2

2H(d,n)3He

isotopically pure target

7.706

0.02 - 39.80/85.

201

 

heavy water target

 

 

3

1H(t,n)3He

isotopically pure target

17.639

3.051 - 98.19

301

 

water target(ice)

 

 

302

 

octane target

 

 

4

3H(d,n)4He

isotopically pure target

20.461

0.01 - 40.00/400.

5

2H(t,n)4He

isotopically pure target

23.006

0.015- 59.9/599.

501

 

heavy water target(ice)

 

 

7

7Li(p,n0)7Be

 

0.651

1.8807-7.00/494.

8

1H(7Li,n0)7Be

 

3.846

3.097- 48.745

9

7Li(p,n1)7Be*

(0.43 MeV level)

1.557

2.40 - 7.00

10

1H(7Li,n1)7Be*

(0.43 MeV level)

7.231

16.713- 48.745

11

11B(p,n0)11C

 

2.388

3.020/3.5 - 5.49/26.

12

1H(11B,n0)11C

 

11.880

33.-59.989/284.1

13

13C(p,n0)13N

 

2.278

3.239 - 12.86/30.6

14

1H(13C,n0)13N

 

12.175

41.80/112.28-165.97

15

15N(p,n0)15O

 

5.742

3.94 - 15.62

16

1H(15N,n0)15O

 

25.726

58.659 -232.549

17

6Li(p,n0)6Be

isotropic approximation

2.558

5.95 - 7.874/200.

18

9Be(p,n0)9B

 

1.951

2.06 - 30.0

19

10B(p,n0)10C

isotropic approximation

4.055

4.94 - 8.571/17.1

20

12C(p,n)12N

dummy cross sections

1.678

19.657 - 30.0

21

14C(p,n0)14N

isotropic approximation

2.415

0.6714 - 3.151/20.67

22

1H(14C,n0)14N

isotropic approximation

9.407

9.332 - 43.795/287.2

23

10Be(p,n0)10B

isotropic approximation

1.176

0.251 - 1.040/20.247

24

1H(10Be,n0)10B

isotropic approximation

3.012

2.495 - 10.337/201.2

25

36Cl(p,n)36Ar

isotropic approximation

2.028

0.778 - 20.08

26

1H(36Cl,n)36Ar

isotropic approximation

7.826

27.774 - 716.8

27

39Ar(p,n0)39K

isotropic approximation

2.593

1.225 - 1.300/20.224

28

1H(39Ar,n0)39K

isotropic approximation

10.279

47.367 - 50.28/782.1

29

59Co(p,n)59Ni

isotropic approximation

0.33

1.8897 - 2.240/11.89

30

1H(59Co,n)59Ni

isotropic approximation

4.22

110.534 - 131./695.5

31

7Li(d,n)8Be

isotropic approximation

>=3 lv

0.001 -10.957

32

11B(d,n)12C

isotropic approximation

>=10 lv

0.141 - 2.513/5.564

33

13C(d,n)14N

isotropic approximation

>=5 lv

0.312 - 15.388

34

15N(d,n)16O

isotropic approximation

>=8 lv

2.13E-4 - 5.979/10.1

35

2H(15N,n)16O

isotropic approximation

>=8 lv

0.0016 - 44.53/75.2

36

18O(d,n)19F

isotropic approximation

>10 lv

0.975 - 204/14.696

3. RECENT DEVELOPMENTS

3.1. 3H(p,n)3He

Between 0.288 and 7.59 MeV single line neutron spectra can be obtained. In cases where the background from the triton break-up can be tolerated this reaction has been used even up to 200 MeV. Below about 2 MeV neutron energy (solid) tritide targets rather than gas targets are used for improved resolution and safer handling (at the expense of yield). Recently, necessary corrections when using such solid targets have been investigated thoroughly [6].

With regard to data accuracy there has been a breakthrough at lower energies. A measurement of the integrated cross sections up to 4.5 MeV [7] agrees below 1.55 MeV perfectly in the shape with an independent R-matrix analysis [8], which incorporated recent polarization data [9]. The required scale adjustment of 2.0% is well within the 5 % scale error of the new data. This new work excels by its careful energy determination, which is crucial in this energy range, as has been pointed out before [10]. Above 2.9 MeV the shape agrees perfectly with the tabulated data of DROSG-96. So in version 2.0 of DROSG-96 only data below 2.9 MeV had to be modified. The new database uses Hale's differential cross section data [9] up to 1.55 MeV and the shape of Brune's integrated data [8] further up to 2.9 MeV. This new solution is shown in Fig. 3 by the full line, the old one by short dashes.

Fig. 3. Reaction cross sections of 3H(p,n)3He. Open circles are data of Brune [8] multiplied by 1.031, full circles are old data by this author. See text for discussion.

 

3.2. 7Li(p,n)7Be

This is the most common source at lower energies. Only between primary neutron energies of 0.121 and 0.650 MeV there are single line spectra at 0. Below 0.121 MeV and above 0.650 MeV (due to the excitation of the 0.43 MeV level of 7Be) double (or triple) line spectra are present. The intensity of the secondary line (from the excitation of 7Be) is typically less than 10% [11] so that it can be tolerated in several applications. The advantages of this source are [12]:

a. Small kinematic energy spread

b. Reasonable neutron intensity

c. Relatively high projectile energy which gives e.g. better time resolution in time-of-flight experiments

d. Simple target production (one would think, see [13]).

A paper in the year 1996 [13] suggested that the integrated thick target yields below 1.99 MeV as calculated by the code WHIYIE of DROSG-96 are high by about 50 %. However, this paper has an internal inconsistency of 46.6% if the yield of metal targets is compared with that of LiF and Li2O. After raising the measured data by that amount a rather good agreement is obtained with the calculated prediction (see Fig. 4). The reason for this discrepancy was traced to inadequate handling of the metal target so that hydroxide was present [14].

Fig. 4. Total neutron yield from the p-7Li reaction using lithium metal as a target. Full line: prediction of DROSG-96, full dots: data of [13] multiplied with 1.466.

In another experiment the 0 differential cross sections between 2.7 and 4.0 MeV proton energy [15] were measured. These data agree within their uncertainties so well with the predict-ions of the code that they do not suggest any change of the database in this energy range.

 

3.3. 3H(d,n)4He

There are new differential cross section data [16], which extend the energy range up to 19.5 MeV. The measured zero-degree excitation function agrees perfectly with that of the FENDL-evaluation [17] but differs noticeable from a later prediction [2] using charge symmetric 3He(d,p)4He data.

Fig. 5. Differential cross-sections of d-T at 19.5 MeV. Diamonds are from [16], stars from [18] and crosses from [19].

Fig. 5 shows the new differential cross section data at 19.5 MeV together with older data at 19 MeV [18,19]. Obviously, these new data with a distinctly lower third maximum will contribute to a better prediction of the d-T cross sections above 20 MeV.

At such high deuteron energies d-T is a useful "monoenergetic" neutron source under special circumstances only. As the neutron background from the deuteron break up on tritium is strongly forward peaked the signal to background ratio is much more favorable around 45 (at the second maximum or the shoulder of the differential cross section) than at 0. This advantage of the 45 position is offset by the fact that the meager monoenergetic yield of this reaction is reduced even further (by a factor of about 5).

4. CONCLUSIONS

There has been no evidence up to now that the cross section data bases used in the code DROSG-96 do not suffice today's applied needs (e.g. for BNCT). This is true specifically for the four "working horses" 3H(p,n)3He, 7Li(p,n)7Be, 2H(d,n)3He and 3H(d,n)4He.

Excellent new integrated data together with an independent R-matrix analysis of the 4He-system allowed to improve the data base of 3H(p,n)3He below 2.9 MeV.

The recent 3H(d,n)4He data near 20 MeV will be a valuable input for an evaluation at higher energies improving the data base near and above 20 MeV.

However, attempts to improve the lower energy data base of the reaction 7Li(p,n)7Be which is of special interest in connection with BNCT failed insofar as the quality of the new data is inferior to those used as data base in the code. In particular, it appears that in these cases the attempts to measure cross sections more accurately than in the (even far) past are hampered by the fact that much experimental know-how, e.g. the preparation of a metal lithium target, has not been transferred to this generation of physicists, so that they have still to learn from their own mistakes.

5. ACKNOWLEDGEMENTS

I have received private communications regarding the new data from C. Brune, D. DeSimone, R. Walter and Y. Harker. G. M. Hale, T-2 of LANL, provided me with his new R-matrix results. All these contributions to this work are thankfully recognized.

REFERENCES

[1] M. DROSG," DROSG-96, PC database for 33 neutron source reactions", documented in the IAEA report IAEA-NDS-87 Rev. 4 (March 1997).

[2] M. DROSG: "Monoenergetic neutrons in the energy range from 100 eV to 200 MeV from two-body reactions with the hydrogen nuclei." Proc. 5th International Conf. on Applications of Nucl. Techniques "Neutrons in Research and Industry", Sissi, Crete, June 1996, Proc. SPIE 2867, 490 (1997).

[3] W. I. KONONOV, E. D. POLETAEV and B. D. JURLOV, At. Energija 43, 303 (1977).

[4] G.AUDI and A.H.WAPSTRA, "The 1995 update to the atomic mass evaluation", Nuclear Physics A595, 409 (1995).

[5]. J. F. ZIEGLER, ed., "The Stopping and Ranges of Ions in Matter", Vol. 3 "HYDROGEN, Stopping Powers and Ranges in All Elements", Vol. 5, "Heavy Ions, Stopping Powers and Ranges", Pergamon, 1977, 1980.

[6] S. P. SIMAKOV, G. N. LOVCHIKOVA, A. M. TRUFANOV, V. A. VINOGRADOV and M. G. KOBOZEZ, "Nonmonoenergety of neutron source based on the solid tritium target", XIV Int. Workshop on Nucl. Fission Physics, Obninsk (Russia) 12 - 15 Oct 1998.

[7] C. R. BRUNE, K. I. HAHN, R. W. KAVANAGH and P. R. WREAN, "Total cross section of the 3H(p,n)3He reaction from threshold to 4.5 MeV", priv. communication from Brune by way of a preprint (1999).

[8] G. M. HALE, priv. communication (1999).

[9] W. S. WILBURN, P. R. HUFFMAN, N. R. ROBERSON, W. TORNOW, C. R. GOULD, C. D. KEITH and G. M. HALE, Few Body Systems 24, 27 (1998).

[10] M. DROSG: "The 3H(p,n)3He Differential Cross Sections Below 5 MeV and the n-3He Cross Sections", LA-8215-MS, LASL (1980).

[11] H. LISKIEN and A. PAULSEN, At. Data Nucl. Data Tables 15, 57 (1975).

[12] M. DROSG, Nucl.Sci.Eng. 106, 279 (1990).

[13] Y. D. HARKER, F. HARMON, J. SEAMANS,jr., S. SERRANO, W. TRAMMELL, L. YOST, X.-L. ZHOU and R. HAMM, "Accelerator neutron sources for neutron capture therapy using near threshold charged particle reactions", Proc. 5th International Conf. on Applications of Nucl. Techniques "Neutrons in Research and Industry", Sissi, Crete, June 1996, Proc. SPIE 2867, 80 (1997).

[14] Y. D. HARKER, priv. communication (1999).

[15] D. J. DESIMONE, G. H. R. KEGEL, J. J. EGAN, P. BERTONE and P. STAPLES, Nucl.Inst.Meth.Phys.Res. A388, 443 (1997).

[16] H. TANG, Z. ZHOU, B. QI, C. ZHOU, Y. DU, H. XIA, R. L. WALTER, C. R. HOWELL, W. TORNOW, R. T. BRAUN, ZE. CHEN, ZH. CHEN and Y. CHEN, "Excitation function and angular distributions of the 3H(d,n)4He reaction in the energy range 6 to 23 MeV", INPC '95: International nuclear physics conference, Beijing (China) 21 - 26 Aug 1995, and priv. commun. (1999).

[17] G. HALE and M. DROSG, "d-t reaction evaluation", Charged-particle library for FENDL/C-2.0, Version 1.0 March 1997.

[18] N. A. VLASOV, G. F. BOGDANOV, S. P. KALININ, B. V. RYBAKOV and V. A. SIDOROV, " Investigations of reactions of 18-MeV deuterons with light nuclei by the time-of-flight method", Int. Conf. Neutron Interactions with the Nucleus, Columbia Univ., N.Y., USA, report TID-7547 (1957), W. W. Havens, jr., Editor.

[19] J. E. SIMMONS and J. J. MALANIFY, Bull. Am. Phys. Soc. 13, 564 (1968); and priv. communication from J. J. MALANIFY (1977).