PHYS. MED. BIOL. 1966, VOL, 11, NO. 1, 1-14

Neutron Flux and Energy Measurements

J. A. DENNIS, B.SC., A.R.C.S., A.INST.P.

Health Physics and Medical Division, A.E.R.E., Hsrwell, Berks.

Received 2 February, 1965

ABSTRACT. The problem of neutron energy measurement is considered in the energy ranges, thermal ( < 0.5 ev), intermediate (0.5 ev-l0 kev), fast (10 kev-15 MeV) and relativistic ( > 15 Mev), together with the general problem of neutron flux measurement. Sensitivities and resolutions obtainable are given for a variety of methods, It is concluded that for thermal and intermediate, energy measurements in radiological applications it is best to use BFS counters, fission chambers, moderated detectors, together with gold or manganese foils, in conjunction with the Westcott flux convention; at higher energies methods based on proton recoil detectors are preferred for high sensitivity and accuracy, while threshold detectors are suggested if multichannel pulse-height analysers and computers are not available, or if there is a limitation on the size of the detectors.

1. Introduction Although some methods of neutron flux and energy measurement are

applicable over a wide energy range, most methods can only be used in certain ranges of energy. For this reason, and in spite of some artificiality, it is conyenient to discuss the methods in terms of the four main energy groups: (1) Relativistic > 15 Mev, (2) Fast 10 kev-15 Mev, (3) Inter- mediate 0.5 ev-l0 kev, (4) Thermal < 0.5 ev.

With very few exceptions neutrons are initially generated with fast or relativistic energies. They may be produced as mono-energetic, from photo-neutron reactions or by the action of mono-energetic charged particles on thin targets; alternatively they may be emitted with a wide spectrum of energies due either to the spread of energies of the bombarding particles or to the nature of the process by which they are formed, e.g. fission in heavy elements or cascade and evaporation processes induced in light elements by very energetic particles.

If the neutron source is surrounded by a non-caphring, scattering medium the energy of the neutrons is reduced by nuclear collisions and an energy spectrum appears which is characterized by the equation (Hughes 1953):

' ( E ) .$E.u,(E) dE

where E is the neutron energy, +(E) is the neutron flux per unit energy interval a t energy E , S is the strength of the neutron. source, .$ is the

A P.M.B.

2 J . A . Dennis

mean logarithmic energy loss per collision and is a function only of the atomic weight (assuming a constant scattering law), u,(E) is the scattering cross-section a t energy E.

As u,(E) is usually constant below 100 kev the expression leads to the familiar 1/E spectrum of intermediate energy neutrons.

Finally, the energy of the neutrons becomes comparable with the thermal energy of the atoms with which they are colliding and they take up a Maxwellian spectrum.

When moderating material is placed between the source and the measuring point it exerts a profound effect on the neutron flux and spectrum. Apart from the obvious attenuation effect, a mono-energetic spectrum will become contaminated with less energetic neutrons; while on the other hand, a continuous spectrum will become hardened due t o the increasing penetrating power of neutrons with energy above 100 kev.

When trying to decide how to measure the energy and flux of neutrons, we must bear in mind the reasons for measuring these parameters. If we are only interested in the rate of neutron absorption in nuclear inter- actions, in the radioactivity induced in certain materials, or in the mean absorbed doses, then our requirements might be less stringent than if we are investigating the L.E.T. distribution, or the fundamental laws of neutron reactions and neutron moderation.

Although crystal spectrometers and time-of-flight techniques provide the most accurate methods of measuring neutron energy, they are not in general suitable for radiobiological and radiological protection applications, and will not be discussed here. Descriptions of these techniques have been given by Hughes (1953, 1961), Havens (196l), Poole (1960), Neiler and Good (1960), and Rybakov and Sidorov (1960).

2. Thermal neutrons (c 0.5 ev) I n applications related to radiological protection and radiobiology,

one is more likely to have a diffuse source of neutrons than a beam. In this case it is practically impossible to measure the neutron spectrum in any detail. As, however, the only interest is probably in the rate of production of a specific nuclear interaction, the information required can often be obtained by the use of suitable detectors and a convention of interpretation.

.Perhaps the most widely used convention is that due to Westcott, Walker and Alexander (1958 a), This assumes the existence of an E-l spectrum which terminates abruptly at an energy which is a multiple p of the most probable energy E , of the Maxwellian density spectrum. For a temperature of 20°c, E,= 0.0253 ev. A common assumption is p = 4.3 and therefore E, = 0.12 ev. These values are selected to give the best results for a variety of reactor spectra.

If R(l/v) is the rate of neutron absorption per atom in a material whose cross-section is inversely proportional to the neutron velocity V ,

Neutron Flux and Energy Measurement 3

and uo is the cross-section in barns for neutrons of energy 0.025 ev (velocity W,, = 2200 m sec-1) then the Westcott flux is defined as:

The absorption rate in any other ' one over v ' absorber is then easily obtained from the Westcott flux measured by any other such absorber. The absorption rate R (not liw) in a material whose neutron absorption cross-section is not inversely proportional to the neutron velocity is given by the equation

R (not liw) = 4w U,, ( g + rs) (3) The factor g describes the departure from l l v absorption in the

Maxwellian energy region; it depends on the temperature, or most probable energy, of the spectrum.

The factor S describes the departure from llw absorption in the intermediate-energy region; it is also dependent to some extent on the temperature of the Maxwellian spectrum.

The factor r depends on the ratio of intensities of the E-l and Maxwellian spectra.

0.001 0.01 E " r V 0. l I

Thermal neutron activation cross sections

Fig. 1. Cross-sections of some materials associated with the measurement of the Westcott flux parameters.

Values of g and S have been tabulated for many substances by Westcott (195813). The r factor can be obtained from the ratio of the reaction rates with and without a cadmium cover, which absorbs all neutrons less

4 J . A . Dennis

than about 0.5 ev, depending upon the geometry and thickness. The temperature of the Maxwellian spectrum can be found from the ratio of the reaction rates in two materials, one of which has a strong departure from l /v absorption in the thermal region; alternatively a series of absorbers placed over a detector may be used to produce the same effect (Hughes 1953). A common method is to use a 239Pu fission chamber in conjunction with a loBF3 counter or a 235U fission chamber. Much attention has been focussed on lutecium for this purpose as it has two convenient isotopes (fig. 1). Unfortunately, the cross-sections are not known very accurately and the method is only suitable for high flux levels.

Recently Nisle (1963) has produced a sophisticated modification of the Westcott convention, in that the E-1 spectrum is terminated by a function which more closely corresponds t o reality than the abrupt cut-off used by Westcott.

3. Intermediate neutrons (0.5 ev-l0 kev) Neutrons in this energy region are notoriously difficult t o deal with

because the absorption cross-sections are decreasing rapidly with energy (although scattering cross-sections are constant), and the recoiling atoms from neutron collisions are neither sufficiently energetic nor efficient a t producing ionization to be detected easily, particularly in the presence of y-radiation.

3.1. Absorbers One rough method of obtaining the spectrum is to surround a fission

detector with different thicknesses of boron-10, as first described by Hurst, Harter, Hensley, Mills, Slater and Reinhardt (1956). The sensitivity of such a combination has been measured by a time-of-flight method by Basson and Dennis (1962).

3.2. Moderated detectors A fairly common method is to surround a thermal neutron detector

with a series of moderators of different sizes. Usual combinations are those of polythene spheres with lithium or boron-loaded scintillators. The higher energy neutrons are moderated to lower energies in the polythene and are more easily detected. By selecting a suitable number of spheres of different sizes the spectrum may be deduced. However, users of this method a,re usually forced to interpret their measurements in terms of a combination of fission and E-l spectra (Hankins 1961).

Nachtigall and Rohloff (1964) have described a system of flux measure- ment based on this technique, and using three spheres.

Basson (1963) has shown that a polythene sphere, 6.5 cm diameter, placed around a boron ZnS scintillator gives an almost energy independent sensitivity from 1 ev to 10 kev, and can be used for flux measurements in that energy range.

Neutron Flux and Energy dleasurenzent 6

3.3. Resonunce detectors A more useful method for investigating the spectrum, although it has

not been used very often, is to make use of resonance reactions, of the type shown in fig. 2 (Hughes 1953, Judd 1963, Adams and Dennis 1963). One can appreciate that for a sandwich of two or three foils of appropriate thickness, made from one of these materials, the variation of activity between the foils yields information about the intensity of the neutron flux near the resonance energies,

3.4. Black ’ detectors Flux measurements in this energy region may be made by ‘ black ’

detectors (Rae and Bowey 1953, Firk 1963). These are typically boron or lithium-loaded glass scintillators in which there is sufficient boron

Fig. Neutron energy - cV

2. Cross-sections of indium, gold and manganese showing resonances at intermediate neutron energies.

or lithium for all neutrons incident upon the detector to be absorbed. They are most frequently used as detectors in conjunction with time-of- flight spectral measurements.

4. Fast neutrons (10 kev-15 MeV) 4.1. Reactions producing charged particles

possible as methods of measuring neutron energy. In this energy range neutron induced charged particle reactions become

6 J . A . Dennis

The reactions are of the type 'A + n = "+1-'B + "C

(4)

The combined energy of the resulting particles is (En &Q), where Q is the energy released in exothermic reactions or absorbed in endothermic reactions. Some such reactions leave the two particles in excited states and this results in some ambiguity in the energy measurements. For this reason reactions with lighter elements are preferred where the possibility of excited states is less likely. Suitable reactions might be

3He+n=lH+sH+764kev (5)

(EnL,?I +-(En _+ &)&C.

6Li+n=4He+3H+4.79Mev

l o B + n = +7Li* + 4He + 2.31 Mev

4He + 2.79 Mev. The positive Q values make discrimination against y-radiation easier.

Because of the two possible modes of energy release in boron, attention is normally concentrated on the first two of these reactions.

As helium is a gas, it is most easily employed by incorporating it into proportional counters and ionization chambers. It can also be used with silicon diode detectors, as explained below.

In helium gas counters, using good techniques it should be possible to measure the energy of the charged particles with a resolution of 2%) so that for very low energy neutrons the instrument would introduce a spread in energy of 14 kev. This sets a lower limit to the neutron energy which may be resolved of about 45 kev with a resolution of 30%. When the neutron energy is greater than 1 MeV the elastic recoils of 3He have energies greater than 0.7 MeV and can be confused with the (n, p ) events produced by low energy neutrons. In an effort to reduce this effect two silicon diodes with helium between them have been used (Lee and Awcock 1963). Coincident events in which the triton and the proton from (n, p ) reaction enter different diodes are used to eliminate the unwanted recoil events which are only detected by one of the diodes. However, a t higher neutron energies the (n, d ) and (n, p ) reactions which take place in the silicon are a complication because the particles generated in one diode are sufficiently energetic to pass into the other.

The reactions in the silicon are 28Si (n, p ) 28AZ - 3.86 Mev 29Si (n, p ) 29AZ -- 2.89 Mev 28Si (n, a) 25Mg - 2.66 Mev 2 9 S i (m, a) 2 6 M g - 0.02 Mev

Although the nuclei are left in excited'states (fig. 3), these reactions can be used as a convenient and elegant method for measuring the energy and energy spread of mono-energetic neutrons (Marcazzan, Merzari and Tonolini 1962). For cross-sections of these reactions see Potenza, Ricano and ,Rubbin0 (1963), and also Konijn and Lmber (1963).

I (8)

0

Neutron Flux and Energy Measurement I

7

l " - 2 , I

Fig. 3. Pulse-height spectra in a silicon diode when irradiated with mono- energetic neutrons after Potenza, Ricano and Rubbino (1963).

The lithium reaction, eqn. (6), may also be used with silicon diodes (Love, Murray, Manning and Todd 1962). The higher Q of the reaction makes it easier to discriminate against unwanted reactions but the sensitivity of the detector is less than that of the helium-3 device. A complication is the reaction

6Li (n, p ) 6He - 2 MeV. (9)

8 J . A . Dennis

Attempts have been made to use the lithium in the form of the scintil- lator Li I(Eu). The system was found by Murray (1958) to have a very poor resolution of 20% a t room temperatures which could be improved to 10% 6y cooling with liquid nitrogen.

4.2. Elastic scattering Perhaps the best way of measuring fast neutron spectra is to make use

of neutron elastic collisions. The most important of these being with hydrogen because of (a ) well-known cross-section ( h 1% or better); ( b ) simple scattering law (isotropic in centre-of-mass space, up to 5 Mev and possibly 10 Mev); ( c ) maximum transfer of energymakes discrimination against y-radiation easier.

4A ER(3fax,

=En - =En for hydrogen (A +

A =atomic weight of struck nucleus, E , =recoil energy. At the lower energies, the hydrogen is most easily used in proportional

counters, such as described by Benjamin (1963b), to measure neutron energy in the range 50 to 500 kev. The upper energy level is governed by the relative magnitudes of the dimensions of the detector and the ranges of the recoiling protons, and the lower level by the amplifier noise and gamma background. End and wall effects and the background of recoils from neutrons with energies greater than 500 kev are important, as is the fact that, for each neutron energy, recoil protons of all energies up to a maximum equal to the neutron energy are equally probable. The labour involved in overcoming these problems is minimized by the use of computer techniques to ' unfold ' the spectra.

For neutron energies greater than 0.5 Mev solid state devices are used to stop and measure the recoiling protons.

The detector may be: (i) nuclear photographic emulsion; (ii) organic or inorganic scintillator; (iii) semiconductor diode.

Each of these has its own characteristics. With all of them a thin hydrogenous radiator may be used as the source of recoil protons, which may then be collimated or selected by methods described by Marion and Fowler (1960, 1963) and Rybakov and Sidorov (1960), to reduce the complexity of the analysis. The background events in the silicon of the semiconductors described in Section 4.1 may be rendered less important by using two electrodes on the diode as described by Dearnaley and Ferguson (1962).

In the case of nuclear emulsions and organic scintillators, the detector, as in the case of the hydrogen filled proportional counter, may be its own source of proton recoils. For nuclear emulsions the analysis is extremely tedious, as fully automatic methods of measuring the proton tracks have not yet been developed, and manual or semi-manual methods must be used (Benjamin 1963a). It is not uncommon for the measurement of a single spectrum to take between three and six months.

Neutron Flux and Energy Measurement 0

I n organic scintillators efforts to reduce the end and wall effects, by increasing the size of the scintillator, introduce the problem of double recoils. There is the addit,ional problem of non-linearity of the photon

0.1 1.0 IO

Neutron e n e r g y - M e V

Fig. 4. Cross-sections of some threshold reactions.

IC

output with proton energy. However, if one uses a computer these problems are not insuperable. Discrimination against y-radiation can be done by using pulse shape discrimination techniques (Wasson 1963).

The great value of the recoil techniques is their high sensitivity.

4.3. Threshold reactions A widely used method of neutron flux and energy measurement in

reactor and accident dosimetry is based on the use of threshold detectors (Hurst et al., 1956). The use of this system for in vivo measurements is described by Mobley, Engel, Parikas and Godden (1964).

The ideal threshold reaction has a constant cross-section above some threshold energy E , but, as demonstrated in fig. 4, this is never found in practice. The cross-sections in the figure are a representative sample of the more commonly used reactions. The secondary particles from the reactions may be detected, particularly so in the fission reactions, but more commonly the radioactivity of the residual nucleus is measured.

10 J . A . Dennis

The method is most useful when some previous knowledge of the spectrum is a,vailable. However, many attempts have been made to elucidate the spectrum by mathematical treatments, In general these are not satisfactory (it is not uncommon for negative values of the neutron flux a t some energies to be obtained), unless they incorporate some assumption, e.g. that the spectrum approximates to a fission spectrum. One might think that the more detectors were employed the greater would be the accuracy of the spectrum derivation; unfortunately, because of the inaccuracies in the measurements of cross-sect'ions and decay schemes, the opposite is the case. The use of detectors with apparently closely similar threshold energies and neutron energy dependence of cross-section can make it very difficult to derive meaningful results. A review of the mathematical techniques for deriving spectra from such measurements has been given by Bresesti, Del Turco, Ostidich, Rota and Segre (1963). The inaccuracies in the cross-sections are well brought out in the Euratom compilation by BeaugB (1963).

5. Relativistic neutrons (> 15 MeV) Very little has been done by way of neutron spectroscopy for neutrons

above 15 Mev. Tochilin a,nd Kohler (1958) have used photographic emulsions in conjunction with a proton radiator for energies up to 25 MeV. It should be possible to use high pressure cloud chambers or bubble chambers, together with proton recoils (Stephens and Staub 1960). Above 10 Mev the neutron elastic scattering from protons becomes increasingly non-isotropic in the centre-of-mass system; the total scattering cross-section has been measured up to about 200 Mev (Brolley 1961) and, for energies up t.0 40 MeV, has been discussed by Gammel (1963).

Two threshold reactions have been used in this region

This has a threshold about 20 Mev; I1C is a positron emitter with a half- life of 20 min. The carbon is normally used in the form of an organic scintillator which is part of the detection system (McCaslin 1960).

W ( n , 2n)"C. (11)

The other reaction ZO9Bi (n, fission) (12)

has a threshold about 50 Mev and is used coated on to the electrodes of a giant fission chamber in an effort to increase the sensitivity as described by Hess, Patterson and Wallace (1957). With such a system it is difficult to be sure that the fission events are due solely to neutrons, as other high- energy particles are likely to be present and may either simulate fission events or cause fission of the bismuth on their own account.

6. Flux measurement 6.1. B y cross-section

Given a knowledge of the cross-sections and detection efficiencies, any of the methods used for neutron energy measurement may be applied to the measurement of Rux.

Neutron Flux and Energy Measurement 11

For low-energy neutrons, methods based on the (n, a), reaction in loB or the activation of gold or manganese are the most satisfactory because the cross-sections and the decay schemes for gold and manganese are well known. The so-called ' black ' detectors are best used for inter- mediate-energy neutrons. For the higher energies the methods based on proton recoils as described by Allen and Ferguson (1957) are the most satisfactory because the cross-section for the process is the most accurately known; the others are not known to be better than between & 5 and 10% in most cases.

Detector

______ Thermal 1°BF3 counter (25 cm dia x 18 cm-700 mm Hg) 235U Fission Chamber (100 mgm, 50% Efficiency) 23gPu Fission Chamber (5 mgm, 50% Efficiency)

Intermediate Moderated Thermal-Neutron

Detector loB covered 235 U Fission

Chamber (4 mgm, 50% Efficiency)

- -

Resonance Detectors

- Fast 3He Proportional Counter (5 atmos, 3 cm dia) 3He Semi conductor (1.54 cm2, 0.7 mm gap, 3 atmos) OLi Semi conductor (1.54 cm2, 150 pgm cm-2 6LiF) Reactions in Silicon (1.54 cm2, 400 pgm Si) Proton Recoil Proportional Counter (1 atmos Hz, 3 cm dia) Proton Recoil Scintillator (2-5 cm dia., 1.25 cm thick) Proton Recoil Nuclear Emulsion Threshold Detectors

(a few grams)

Sensitivity counts per neutron

per cm2

5.0 (0.025 ev)

0.075 (0.025 ev)

0.006 (0.025 eV)

0~01-0*10

5 x 10-6

0.0 17 (100 kev) 3 x 10-6 (2.5 MeV)

(2.5 MeV) 3 X (5 Mev) 0.028 (100 kev) 0.10 (4 Mev) Fluence < Z ~ X 10% cm- Fluence > z 3 X 10% cm-2

7 X 10-7

Resolution Yo

30%

30 % R5 10%

I t selected energies

Table 1. Some detector sensitivities and resolutions.

12 J . A . Dennis

6.2. The long counter ’ The most common method of measuring the flux of fast neutrons is the

so called ‘ long counter ’ (Hanson and McKibben 1947), consisting of a BF, proportional counter in a shaped block of paraffin wax. The name derives from the supposed constant sensitivity over a long energy interval as discussed by Allen (1960). Recently de Pangher (1961) has described a version of this device in which the moderator is made of polythene with less chance of the distortion which may occur when paraffin wax is used. A study of the angular resolution of the long counter has been made by Bmresen, Gremeland and Messelt (1962).

7. Conclusions In table 1 an attempt has been made to summarize the detection

efficiencies and resolution capabilities of some of the methods described. I n radiobiological and radiological protection applications, BF, counters, fission chambers, moderated detectors, and gold or manganese foils are probably the most useful instruments for low and intermediate energies with the V‘estcott convention used as the method of interpretation; a t higher energies, those who have the use of multichannel pulse-height analysers and computers should make use of the proton recoil methods because of their high sensitivity and accuracy; other workers must do the best they can with a mixed bag of moderation techniques and threshold detectors. In some cases, particularly for in vivo measurements where size is a limitation, threshold detectors may be preferred, but possible inaccuracies in the cross-section data should be remembered when interpreting such measurements.

R ~ S U M I ~

La mesure du flux et de 1’6nergie du neutrons. On Qtudie le probleme de la mesure de 1’6nergie neutronique dans les g a m e s suivantes

de l’bnergie : thermique ( < 0,5 ev), intermediaire (0,5 ev-l0 kev), rapide (10 kev-l5 MeV) et relativiste ( > 15 Mev), en conjonction avec le probleme g6n6ral de la mesure du flux de neutrons. On donne les sensibilitbs et resolutions que l’on peut atteindre avec de methodes diverses.

On conclue que, pour les mesures de 1’6nergie thermique et intermediaire au domaine de radiologie, il convient le mieux d’employer les compteurs t3 BF,, les chambres t3 fission, les dbtecteurs modbrbs, avec les feuilles en or et manganese en conjonction avec les conditions de Westcott, relatives au flux; pour les energies plus hautes on pr6fkre les mbthodes basbes sur les detecteurs au recul de protons, i cause de leur haute sensibilite et exactitude, tandis que les detecteurs ti seuil sont recommand& quand les sblecteurs i canaux multiples e t les machines calculatrices ne sont pas disponibles ou quand les detecteurs ne peuvent pas depasser une limite de dimensions.

ZUSAMMENFASSUNQ Neutronenfluss- und Energiemossungen.

Man betrachtet das Problem der Neutronenenergiemessungen in den Energiebereichen: thermisch ( < 0,5 ev), mittlere Energie (0.5 ev-l0 kev), schnelL (10 kev-15 Mev), und relativistisch (15 > Mev), sowie des allgemeine Problem der Neut.ronenflussmessung. Es werden die Empfindlichkeiten und Auflosungen fur verschiedene Verfahren angegeben.

Neutron Ftux and Energy Measurement 13

Es wird geschlossen, dass (fur Messungen der thermischen und mittleren Energie fur rediologische Zwecke)diebestenErgebnissemitden RF,-Zahlern,Spaltkammern,moderierent Detektoren, zusgmmen mit Gold- und Manganfolien in Verbindung mit der Westcott- Flussbedingung erhalten werden; fur hohere Energien bevorzugt man Methoden, welche auf Riickstossprotonendetektoren basieren, und zwar wegen ihrer Empfindlichkeit und Prlizision; schliesslich werden Schwellendetektoren empf’ohlen in Fallen, wenn Vielkanal- Impulshohenanalysatoren und Rechengerate nicht zur Verfugung stehen oder wenn die Grosse der Detektoren begrenzt wird.

REFERENCES ADAMS, N., and DENNIS, J. A., 1963, Neutron Dosimetry, Vol. I, (I.A.E.A.), 173. ALLEN, W. D., and FEROUSON, A. T. G., 1957, Proc. Phys. Soc., 7OA, 639. ALLEN, W. D., 1960, Fast Neutron Physics, Vol. I (Edit. Marion, J. B., and

BASSON, J. K., and DENNIS, J. A., 1962, Nucl. Inst. and Methods, 16, 321. BASSON, J. K., 1963, Nucl. Inst. and Methods, 22, 339. BEAUG$, R., 1963, Sections Efficaces pour les DBtecteurs de Neutron pur

BENJANIN, P. W., 1963a, Neutron Dosimetry, Vol. 1, (I.A.E.A.), 307. BENJAMIN, P. W., 1963b, Paper t o Symposium on Fast and Epithermal Neutron

BBRRESEN, S., GREMELAND, B., and MESSELT, S., 1962, Nucl. Inst. and Methods,

BRESESTI, M., DEL TURCO, A. M., OSTIDICH, A., ROTA, A., and SEGRE, G.,

BROLLEY, J. E., 1961, Methods of Experimental Physics, Vol. 5A (Edit, Yuan,

DEARNALEY, G., and FERGUSON, A. T. G . , 1962, Nucleonics, 20, (4), 84. FIRK, F. W. K., 1963, Fast Neutron Physics, Vol. I1 (Edit. Marion, J. B. , and

GAMMEL, J. L., 1963, Fast Neutron Physics, Vol. I1 (Edit. Marion, J. B., and

HANKINS, D. E., 1961, I D 0 16655. HANSON, A. O., and MCKIBBEN, J. L., 1947, Phys. Rev., 72, 673. HAVENS, W. W., 1961, Methods of Experimental Physics, Vol. 5A (Edit. Yuan,

HESS, W. N., PATTERSON, H. W., and WALLACE, R., 1957, Nucleonics, 15, (3)) 74. HUGHES, D. J., 1953, Pile Neutron Research (Addison-Wesley). HUOHES, D. J., 1961, Methods of Experimental Physics, Vol. 5A (Edit. Yuan,

HURST, G. S., HARTER, J. A., HENSLEY, P. M., MILLS, W. A., SLATER, M., and

Fowler, J. L., Interscience Publishers), 179.

Activation, RecommandQes par le Group de Dosimktrie d’Euratom.

Spectra in Reactors, Harwell, (E.A.E.S.).

16, 135.

1963, Neutron Dosimetry, Vol. I, (I.A.E.A.), 27.

L. C. L. and Wu, C., Academic Press), 461.

Fowler, J. L., Interscience Publishers), 2237.

Fowler, J. L., Interscience Publishers), 2185.

L. C. L., and Wu, C., Academic Press), 495.

L. C. L., and Wu, C., Academic Press), 566.

REINHARDT, P. W., 1956, Rev. Sci. Inst., 27, 153.

14 Neutron Flux and Energy Measurement

JUDD, A. Al., 1963, Nucl. Inst. and Methods, 23, 29. KONIJN, J., and LAUBER, A., 1963, Nucl. Phys., 48, 191. LEE, M. E., and AWCOCK, M. L., 1963, Neutron Dosimetry, Vol. I (I.A.E.A.), 441. LOVE, T. A., MURRAY, R. B., MANNING and J . J., TODD, H. A., 1962,

RICCASLIN, J. B., 1960, Health Physics, 2, 399. MARCAZZAN, M. G., MERZARI, F. and TONOLINI, F., 1962, I.R.E. Trans., NS-0,

234. MARION, J. B., and FOWLER, J. L., 1960, Fast Neutron Physics, Vol. I (Inter-

science Publishers), 1963, Fast Neutron Physics, Vol. 11, (Interscience Publishers).

Nuclear Electronics, Vol. I (I.A.E.A.), 415.

MURRAY, R. B., 1958, h’ucl. Ins t . , 2, 237. XOBLEY, I. S., ENGEL, R. E., PARIKAS, V., and GODDEN, W. R., 1964, Biological

XACHTIGALL, D., and ROHLOFF, F., 1964, Nukleonik, 6, 330. NEILER, J. H., and GOOD, W . M., 1960, Fast Neutron Physics, Vol. I , (edit.

Marion, J. B., and Fowler, J. L., Interscience Publishers), 509. NISLE, R. G., 1963, Neutron Dosimetry, Vol. I (I.A.E.A.), 111.

POOLE: M. J., 1960, Contemporary Phys. , 2, 15. POTENZA, R., RICASO: R., and RUBBINO, A., 1963, Nucl. Phys., 41, 298. RAE, E. R., and BOWEY, E. 31.: 1953, Proc. Phys. Soc., A66, 1073. RYBAKOV, B. V . , and SIDOROV, V. A., 1960, Fast Neutron Spectroscopy (Chapman

STEPHENS, W. E., and STAUB, H., 1960, Fast Neutron Physics, Vol. I (Edit.

TOCHILIE, E., and KOHLER, G. D., 1958, Health Physics, 1, 332. WASSOK, M,, McR., 1963, AERE-R4269.

Effects of Neutron and Proton Irradiations, Vol. I, (I.A.E.A.), 103.

DE PANGHER, 1961, HW-SA-2140.

and Hall).

Marion, J. B.: and Fowler, J.-L., Interscience Publishers), 335.

WESTCOTT, C. H., w7ALKER, W. H., and ALEXANDER, T. K., 1958a, A/Conf. 15jPp02.

WESTCOTT, C. H., 195813, CRRP-787.