IAEA-TECDOC-465

ISOTOPIC NEUTRON SOURCES FOR NEUTRON ACTIVATION ANALYSIS

USER'S MANUAL PREPARE . HOSTJ Y DB E UNIVERSITY OF GHENT

A TECHNICAL DOCUMENT ISSUED BY THE INTERNATIONAL ATOMIC ENERGY AGENCY, VIENNA, 1988 IAEe Th A doe t normallsno y maintain stock f reportso thin si s series. However, microfiche copies of these reports can be obtained from

INIS Clearinghouse International Atomic Energy Agency Wagramerstrasse5 P.O.Bo0 10 x A-1400 Vienna, Austria

Orders shoul accompaniee db prepaymeny db f Austriao t n Schillings 100, in the form of a cheque or in the form of IAEA microfiche service coupons which may be ordered separately from the INIS Clearinghouse. PLEAS AWARE EB E THAT MISSINE TH AL F LO G PAGE THIN SI S DOCUMENT WERE ORIGINALLY BLANK ISOTOPIC NEUTRON SOURCES FOR NEUTRON ACTIVATION ANALYSIS IAEA, VIENNA, 1988 IAEA-TECDOC-465

Printed by the IAEA in Austria June 1988 FOREWORD

Neutron activation analysis (NAA) is a well established analytical techniqu countrien i e s wher nucleaea r reacto s availablei r . However, manf o y the developing countries have neither a nuclear reactor nor other sources of neutrons, but have still a strong interest in this technique.

Throug Technicas it h l Co-operation Programme Agence ,th bees yha n promoting the use of nuclear analytical techniques such as Neutron Activation Analysi X-rad san y Fluorescence Analysi mann i s y governmenta universitd an l y laboratorie developinn i s g Member States. Instrumentatio s oftei n n provided for such purposes, e.g., gamma spectrometry systems and other radioactivity counting equipment.

mann I y developing majoe countriesth rf o drawback e ,on e lac th f o ks i s appropriate neutron sources, eithe nucleaa r r research reacto neutroa r o r n generator. Wit purpose th h asseso t e feasibilite sth usinf o y g other alternative neutron sources, a Consultants' Meeting on Isotopic Neutron Sources for Neutron Activation Analysis was organized in May 1985. It was concluded that Isotopic Neutron Sources of an appropriate design and characteristics in terms of neutron flux and spectra, could allow small research center universitd san y laboratorie o carr t st interestin ou y d an g meaningful training and research oriented projects based on neutron activation and radiochemistry.

The present User's Manual is an attempt to provide with a series of well thought demonstrative experiment utilizatioe baseA th NA n o dn n i sa f no isotopic neutron source, for teaching and training purposes. In some cases, these ideas can be applied to the solution of practical analytical problems.

The Agency wishes to thank all scientists who participated in the Consultants' Meeting, and particularly Professor Dr. J. Hoste, who undertook tase writo th kt presene eth t manual. EDITORIAL NOTE

In preparing this material for the press, staff of the International Atomic Energy Agency have mounted and paginated the original manuscripts and given some attention to presentation. The views expressed necessarilynot do reflect governmentsthosethe of Memberthe of States or organizations under whose auspices the manuscripts were produced. The use in this book of particular designations of countries or territories does not imply any judgement by the publisher, the IAEA, as to the legal status of such countries or territories, of their authorities and institutions or of the delimitation of their boundaries. The mention of specific companies or of their products or brand names does not imply any endorsement recommendationor IAEA. partthe the of on CONTENTS

INTRODUCTION ...... 9 .

1. ISOTOPIC NEUTRON SOURCES ...... 12

1.1. Alpha (a,n)-sources ...... 12 1.1.1e (a,n)-reactioTh . n ...... 2 1 . 1.1.2. Typ f alpha-sourceo e s ...... 3 1 . 1.1.3. Neutron yield energd an s y ...... 3 1 . 1.1.4. Source choice ...... 7 1 . 1.1.5. Source arrangement ...... 18 1.1.6. Transport container ...... 24 1.1.7. Safety aspects ...... 5 2 . 1.1.8. Source installation ...... 5 2 . 1.1.9. General safety ...... 28 1.2. Spontaneous fission sources ...... 9 2 . 1.3. Photoneutron sources ...... 31 1.4. Neutron multipliers ...... 33

. 2 PRINCIPLE ACTIVATIOF SO N ANALYSIS ...... 5 3 .

2.1. General equation ...... 35 2.2. Activatio isotopin a n ni c neutron source ...... 8 3 . 2.2.1. Fast flux ...... 39 2.2.2. Thermal flux ...... 40 2.2.3. Epithermal neutron flux ...... 1 4 . 2.2.4. Reaction rate for (n,7) reactions ...... 41 2.3. Standardization ...... 42 2.3.1. Absolute method ...... 42 2.3.2. Relative method ...... 42 2.3.3. Single comparator ...... 44 2.3.4. ko-standardization ...... 44 2.4. Source f erroo s r ...... 6 4 . 2.4.1. Flux gradients ...... 46 2.4.2. Neutron shielding ...... 47 2.4.3. Nuclear interference ...... 8 4 . 2.4.3.1. Threshold reactions ...... 8 4 . 2.4.3.2. Fission reactions ...... 48

3. PROMPT GAMMA NEUTRON ACTIVATION ANALYSIS (PGNAA) ...... 0 5 .

4. THERMAL NEUTRON ABSORPTION ANALYSIS ...... 52

4.1. Principle ...... 2 5 . 4.2. Detectio e thermath f no l flux ...... 4 5 . 4.3. Standardization ...... 54 4.4. Other applications ...... 55 . 5 TRANSPORT SYSTEMS ...... 7 5 .

6. MEASURING EQUIPMENT ...... 59

6.1. General ...... 9 5 . 6.2. Sodium iodide detectors ...... 9 5 . 6.3. Germanium semi-conductor detectors ...... 64 6.4. Detector calibration ...... 65 6.4.1. Resolutio d efficiencan n y ...... 6 6 . 6.4.2. Energy calibration ...... 8 6 . 6.4.3. Evaluatio gamme th f no a spectra ...... 9 6 .

7. SAMPLE PREPARATION ...... 72

8. APPLICATIONS ...... 74

8.1. Determination of silver in lead ...... 74 8.1.1. Introduction ...... 4 7 . 8.1.2. Nuclear data ...... 5 7 . 8.1.3. Equipment ...... 76 8.1.4. Sample and standard preparation ...... 76 8.1.5. Irradiation and counting conditions ...... 76 8.2. Determinatio f manganesno pyrolusitn i e ferromanganesd an e e ...... 7 7 . 8.2.1. Introduction ...... 7 7 . 8.2.2. Nuclear data ...... 78 8.2.3. Interferences ...... 79 8.2.4. Neutron and gamma-ray attenuation effects ...... 81 8.2.5. Measuring and irradiation conditions ...... 81 8.2.6. Calculations ...... 83 8.3. Determinatio f siliciuno aluminiud man bauxitn mi e ...... 4 8 . 8.3.1. Introduction ...... 4 8 . 8.3.2. Nuclear data ...... 84 8.3.3. Sample preparation, irradiation and measurement ...... 89 8.4. Determinatio f siliciuno aluminiud man aluminium-siliciun mi m alloys ...... 2 9 . 8.4.1. Introduction ...... 92 8.4.2. Principle ...... 3 9 . 8.4.3. Standardizatio d calculatioan n n ...... 3 9 . 8.5. Determinatio cassiteritn i n ti f no e ...... 5 9 . 8.5.1. Introduction ...... 5 9 . 8.5.2. Nuclear data ...... 97 8.5.3. Sample preparation, irradiatio d measuremennan t ...... 8 9 . 8.5.4. Standardization ...... 0 10 . 8.6. Determination of fluorine in fluorites and ores ...... 100 8.6.1. Introduction ...... 100 8.6.2. Nuclear data ...... 100 8.6.3. Irradiation, counting and calculation ...... 102 8.7. Determination of manganese in plants ...... 104 8.7.1. Introduction ...... 104 8.7.2. Nuclear data ...... 104 8.7.3. Procedure ...... 105 8.8. Determination of cadmium in zinc ore by neutron absorption ...... 106 8.8.1. Introduction ...... 6 10 . 8.8.2. Nuclear data ...... 106 8.8.3. Sample and standard preparation ...... 107 8.8.4. Irradiatio measurind nan g conditions ...... 8 10 . 8.8.5. Interference calculatiod an s resulte th f no s ...... 8 10 . 8.9. Determinatio f borono steen n i neutro y b l n absorption ...... 0 11 . 8.9.1. Determination ...... 0 11 . 8.9.2. Procedur resultd ean s ...... 0 11 . 8.10.Other application ...... S IN f so 2 11 .

REFERENCES ...... 3 11 . INTRODUCTION

Neutron activation analysis (N.A.A.) was proposed half a century e analysith . n LevHevesr H vo fo f dysprosiud y o s . an G y agy b o n rari m e

earths using an isotopic neutron source.

Since this historical experiment neutron activation analysis was

develope a highl s a yd valuable analytical technique especially since

the advent of nuclear reactors. These offer indeed high neutron fluxes

o thas t extremely small quantitie a larg f o se numbe f elemento re b n ca s

determined. Consequently N.A.A. was mainly applied to the determination

of trace element a variet n i sf matrice o y s e.g. high purity materials,

biological, environmental and geological specimens etc.

Apart from the high sensitivity for many elements N.A.A. has the

advantage that it is a purely nuclear method of analysis, i.e. the

results are not influenced by the chemical state of the elements under

investigation. Moreover contaminations after irradiation are not to be

feared, a major advantage in trace analysis.

In its initial stage of development, say up to the begin of the

Fifties, simple techniques were use o measurt d e induceth e d activities,

usually a Geiger-Müller counter. This measuring technique having no

selectivity whatsover, chemical separations were in many cases unavoid-

able. At best identification of a nuclide could be carried out by anal- o completo t no x a ysideca f o s y curve.

A major step forward occured with the introduction of the NaKTl)

scintillator used for gamma spectrometry together with single or

multichannel e pooanalysersth ro t resolutio e Du . f o nthi s typf o e detector, chemical separations were still often required. Truly

instrumental method f o sneutro n activation analysis became really possible with the advent of germanium semi-conductor detector of the

Ge-Li typr o more e recently wite hyperpurth h e germanium, coupler fo d

instanc o 400t e 0 channel analyzer o takt s e full advantag e higth h f o e

resolution. In many cases instrumental analysis thus became possible

for over 40 elements in certain matrices if together with the high

resolution advantag e differencth s takei e f o n f half-lifeo e e th f o s

produced radioactive species, i.e. short- medium- and long-lived

isotopes.

Other source f neutrono s s e thanucleath n r reactor have been

developed namely the so called neutron generator. By bombardment of a

tritium target witw energlo h y deuterons r instancfo , 0 keV15 e .

Practically more energeti V neutronMe 4 1 e cproduce ar s a rat t a ed which

can reach some 10 neutrons per second. These machines are mainly

applied for the determination of oxygen and silicon, as the sensitivity

and selectivity is quite high. Activation analyses have also been

describe y cyclotrob d n produced neutrons, mainl y deuterob y n irradia-

a berylliu tiof o n m target e higTh .h intensity continuous energy

spectru f higo m h energy neutrons induce mainly threshold reactiond an s

allow a sensitive determination of a number of elements which are

difficult to determine by (n,y)reactions induced by thermal neutrons.

Neutron activatio a reacto n i n r t doeallono sn eas d a wan y

sensitive determinatio f ligho n d t an element N r instanc, fo C s a s, 8 e e successfull b . f n Thio Tl heav ca r sd no y an . 0 elementyb P s a s

achieve y b charged d particle analysis with protons, deuteronr o s

hélions with o energiesomt 0 MeV2 p eu .f o s

e Froabovth m e consideration t appeari s s that sensitivn i d an e

many cases instrumental analyses can be achieved for many elements by

applying reactor irradiation, mainly with thermal neutrons, although

epitherma d fasan lt fission neutron e occasionallar s y also appliede th ,

10 most important characteristic being the high neutron flux and conse-

quent high sensitivity.

Neutrons can also be produced from isotopic sources. Usually

however the neutron production is many orders of magnitude smaller than

in reactors. Thus trace analysis is only possible for elements with

adequate nuclear characteristic s i limite a smald an o t sd l numbef o r

w elementsisotopino o t p cU . neutron sources (I.N.S.) havt beeno e n

thoroughly explored, except perhaps in geological prospection, although

they offer interesting analytical perspectives for accurate determina-

tion f mino o majod s an r r constituant n i sores , concentrates. alloys s alsi oetct I obviou. s that r pedagogifo , c purposes. I.N.S e alsar . o

interesting as the basic principles are obviously also valid and not

source intensity dependent. For institutions who cannot for financial

or other reasons dispose of a nuclear reactor, I.N.S. offers interest-

g possibilitiese Beecin D the) p (1 O ky . offeJ y a rb s pointe A . t ou d

numbe f o intrinsir c advantages over other neutron sources namely

extremely small size, relatively low purchase and maintenance costs,

long and short term stability of the neutron output, possibility to use

therma d fasan l t neutron fluxesw cos lo f shieldine o eastd Th .an e n ca g

alse addeb oo thit d s lis f advantageo t d havan s e thu a negligibls e

health hazard.

11 1. ISOTOPIC NEUTRON SOURCES

Three main type f isotopio s c neutron sourcee distinguisheb n ca s d

namely alpha-emitters, which produce neutrons through an (a,n)-reac-

tion, gamma-emitters through a (y, n)-reaction and isotopes of heavy

elements which undergo spontaneous fission. The properties of these

sources have been discussed in detail by K.W. Geiger (2) and M.V.

Blinov (ibidem) at the Debrecen meeting in 1980 of the IAEA Consult-

ants' meeting on neutron source properties. It was also reviewed by

K.W. Geiger at the IAEA Meeting in Vienna in May 1985. This Consultants

Meetins subjeca d ha g t "Isotopic Neutron Source r Neutrofo s n Activation

Analysi. ) 3 ( s

1.1. Alpha (a,n)-sources.

1.1.1. The («,n)-reaction.

Radioactive alpha emitters can be used as neutron sources as they

induce (a,n)-reactions on a number of target nuclides. Usually

beryllium is used as a target material through the reaction

9 12 Be(«,nJ C

a Q-valu whics ha f h5.7o e 0 e MeValphth s .show A r a n Tablfo i n1 1. e 210 , emitteothePo r r nuclide e use b s targe a n d ca s t materia t witbu l h

neutron yields which are substancially lower than with beryllium.

It is thus quite clear that for activation analytical purposes only

beryllium shoul e consideredb d e otheTh .r target material e onlar sy

used for special purposes, e.g. to simulate a fission spectrum.

12 21 0 : TABL(a,n)-Reaction 1 1. n LighEo o P t Nucle f o si

Target Nuclide Q-value Neutron yield

0 1 alpha r pe s

7Li - 2.79 2.6

Be 5.70 80 10. 1 .06 13 11. 0. 16 26 13. 2.22 10 18. - 0.70 29 19. - 1 .95 12

1.1.2. Type of alpha-sources.

A number of alpha emitting nuclides can in principle be chosen as

a possible source. A survey of their characteristics is given in Table

1 .2.

1.1.3. Neutron yield d energyan s .

The neutron spectra can be calculated from the relevant nuclear

data. e.g. alpha energies, stopping powe n i berylliumr , energy level

diagrams etc. These calculations are however somewhat obscured by the

fact tha a absorptiot e alpha-emitte n occuth ca nn i r t rno itsel d an f

only in the infinite thick beryllium target. Thus the continuous alpha-

energy distribution give a broadenins e lineth f o sg correspondino t g e energth y diagrams. Change f energo s y als oa functiooccu s a rf o n

13 U) C , _ ~ r- - ^ - - 4 J- - o - r •^ 2 CtJ i •*£> OOOOOOulO O \ • ••••••• S > - • CU oooooooo cn o V VU V V \t S i ^- O TJ O 4-1 C 1 -H >- S

C o <~» M > 4-1 Q) 3 S CU v-' O^ ~J" vO ^~~ ^o "vf r^> Z O'-Ou^i-

t— \ o o r-^Ln^— ooco^— r*-c^ T oooooooo 13 00 1 — r ^ f* l U" C Up { «- O L N O L C U~T— ] •H 1 « • • • • • • >* ^ r— W ~ ^ "•Cf N C T— • O >i T— •

— , \ 1 1- • \o *^o **O vO '*«o ~vî" r^* r^^ *o o oooooooo r-l T T r T T 0) r- ^"I ~I*""I ~I*~i" T ~I ~^ -.-< 1 OO^DLTiCNrO-<î T— LP| >-, CO • • • • • • • CN *— CN CN T— T- C

t/l n) CO 0) U 13 i-i *- h 1 O c i •- m co CN o oo CN CN O O — * 3 O o Qc ) - r U) > \£ 1 O •H vD i — T — LO r^ W ^ ° v5

C #1 s aCU

0) r*- **o p X~N • • CU î> -3" LO 4-1 (1> o -»î>^T3 K^^T? >-i>^ S-i CT3 •r-l 4J l— 1 CT>OOOOO»— mocN 3 1 OO **O O1 CO • ^O CN CN CU O1 ^—

CU T3 33oeeE«a J^ j

14 solid angle e yiel d Th energ.an d y depends als n sourco o e construction,

e.gs cylindricait . l geometry s compositionit , , presenc f «-emittino e g

daughter isotopes etc.

An approximate idea of the neutron energies obtained from the 12 9 12 energy levels of C of the reaction Be{a,n) C is given in Fig. 1.1.

Be + a -n V Me 7 5. = Q

9 1 2 Fig. 1.1 : The Be(a,n) C reaction

A line spectru s thui m s expected e froneutroth . m n n o t grou n p o 3 Howeve n energa r y continuu s alsi m o observed experimentally belo5 3. w 1 2 MeV, partly caused by broad levels in C above 9.6 MeV but mainly by

the break-up reaction

* 9 9 8 Be(a,a'e B ) n + e B >

involving levels at 1.67, 2.43 and 3.06 MeV of Be.

n Ia idealizen d source where a-particle e onlar sy slowed down i n

e Be-matrith e distributioth x f neutrono n r unipe s t energy intervar fo l

a-particle s givei y f b E nenergo A s + yE betweed an E n a ft a

15 27/sin Ode AE = cr(9 ) E ) F( (1.1 ) n dE

where e a(9)centeth s f i mas o r s (c.m.) cross sectio r neutrofo n n pro-

e stoppinth ductio € d gan n power cross sectio r berylliufo n. E t a m a The neutron energy in the laboratory system may be expressed as a

functio e c.mth .f o ne for anglth mn i e

6 s co b + a = E n 1.2) and E A irr F(E 1 .3 E (0=0) - E (9=n) e n n

A compariso e calculateth f o n d mesurean d d spectru r Am-Bfo m s showi e n

in Fig. 1.2.

4 - 241 Am-Be proportional counter

stilbene counter

calculated

10 E (MeV) n

: Compariso 2 Fig1. . n with measured spectra.

16 e neutroTh n yield give s calculatei n Tabli n2 1. e d from

Y - 0.95 + 0.152 E ' neutrons per 10 «.particles (1.4j a

In genera e shoulon l d kee n i minp d thae followinth t g causen ca s

defor e energth m y e neutrospectrth d an na yield

1. elastic and inelastic neutron collisions in the source

2. neutron induced fission within the «-emitter

Neutron spectra of other Be(«,n)-source are shown in Fig. 1.3.

242 Cm 244 Cm

2 1 8 4 0

Neutron Energy (MeVJ

Fig. 1.3 : Neutron spectra of isotopic Be(«,n) sources.

9 e Be(n,2n3th . ) reaction

e (r,nB e )4 th . reaction

5. reduction of the effective «-energy within the «-emitting cluster

6. daughter équilibra of «-emitting species.

1.1.4. Source choice.

For activation analytical purposes a number of considerations

determin e e sourcchoicth th ef o e namely price, half-life, neutron

17 yield, required shielding from gamma rays, toxicity and small volume

and of course availability. 2 24 210 o shorto e t ar live From C m d t appeari tabl 2 d 1. an e so P that 239 238 to be of practical value. Pu and Pu are considered too toxic and

o dangerouto e easilb o t y s accepte n mani d y countries. Radium, which

has often been used in the past is difficult to shield due to its high

leve f gammo l a radiatio s stili t n acceptabla lbu n e choic f availabli e e 227 a ver s anyha d long half lifd thuan e s stable neutros i n c outpuA t

an excellent choice as it combines high neutron yield with reasonable

long half s productiolifIt e n from neutron irradiatiof o n

997 Ra(n,y) Ra ———— > Ac has however been stopped and is probably no longer availabl n sufficieni e t quantities. 244 The Cm not being readily available yet, one is left as choice

wit hm whic A onlhy combine24 1 s long half-life with reasonable high

neutro w ngamm lo yiel d a an dradiatio n dose. 244 It is however possible that Cm sources become important in the

futur s a thie s isotop s producei e n copioui d s quantitie n i fass t

reactor fuel.

1.1.5. Source arrangement.

e neutroTh n yields showe calculate ar n Tabli n2 1. e d from equation

d (1.4expressean ) n i dneutron r secondpe s r activatioFo . n analysis

howeve a thermar l neutron flus i usuallx y required although fast

neutrons are also sometimes applied. Consequently the source is

usually installe a moderatin n i d g medium suc s paraffia h n wax, water o r

water expanded polyester. The most convenient is perhaps the former as

a purification system is not needed as in the case with water, where it

is desirable to remove the corrosion products which gradually dissolve

into the water and are activated by the source. Water expanded

polyester can also be considered.

18 One has to keep in mind that the actual flux given by a source

with a yield of 10 n.s will give a thermal neutron flux density two 5 -2-1 order f magnitudo s e lower i.e. wit, appr a hs fas . 0 n.c1 .tm flux

density of approximtely one order of magnitude higher. 5 -2 -1 Considering that 10 n.cm ,s is a minimum to carry out useful

analyse a sourcs e strength correspondins i t s a leasn. 0 o 1 t t g

necessary.

Several types of source arrangements are possible namely a single

cylindrical source n arraa , f sourceo yn annulaa r o s r one.

s Ii obviout s e thaformeth t r will give riz o considerablt e e

lateral flux gradients so that the activation is not Isotropie and lead o seriout s complications, affectin e accurac e th analysisg th f o y .

An array of sources placed symmetrically around the irradiation

placn yielca ea reasonabld y uniform e fluxarrangemenTh . s i t

certainly to be preferred if the source is rather short lived, as is

the cas ef (seC wite hfurther) 252 o thas , t additional sourcee b n ca s

added to compensate for the decay.

For long lived sources a single annular source is the most

advantageous arrangement as the lateral flux is homogeneous. The

complexity of maintenance and wipe test procedures is also greatly

simplified. Alaerts e.a. (4) described the neutron flux distribution 228 measured experimentally from a cylindrical 1g Ra-Be source (height

36 mm. diameter 17 mm). The arrangement and flux distribution is shown

in Fig. 1.4. One can see that a steep axial flux gradient can be

observed.

The same authors (5) studied the flux distributions from an

annular ? ? 7Ac-B e source with a neutron yield of appr. 10R n.s— 1

A drawine sourcth s f showi o eg n Figi n . 1.5.

The thermal, epithermal and fast fluxes were measured by means of

19 m c 0 8

Compressed air

10 cm

Cadmiui fo m ( 1 mm)

50 cm

4. 10 " 10 . 5 6. 10" Counts/Gram Cu.1c se 0

226 : Cylindrica Fig4 1. . l Ra-Be sourc d axiaan e l flux distribution.

20 Ac-Be mixture Monel metal 304L stainless steel

1 _ O *O 7 9 Fig. 1.5 : Annular Ac-Be source of 10 n.s

respectively a gold-aluminum wire, from a manganese activation with and

O Q O Q without cadmium shiel d froe an dth m Si(n.p l e A reaction)th r Fo .

latte e fasth rt flux distributio e excitatioe th sourcth d f an o n e n

function were taken into account.

The results are shown in Fig. 1.6.

It can be seen that the optimal distributions of thermal and fast

flue situatear x t differena d n t thuca heighte optimizeb sd an sr fo d

the thermal or fast neutron activation.

Similar results can of course be expected from the recommended

e IAEA-Meetinsourcth t a e n 1985i g , namel 5 2 curi a y e americium-241-

beryllium annular source with 52 mm inside diameter, 75 mm outside m overalm diamete5 13 ld lengtan r a showh n Figi n . 1.7.

21 4. 10

2.10 CM I e u in G O

3 01 Z

.x50 épi

section of the source

Ac-Be mixture

r adiatioi e th axi nf o stub e

246 Height from bottom of the irradiation tube (cm)

227 Fig. 1.6 : Axial flux distributions of an annular Ac-Be source.

22 a 76.di 0 — * 52.0 dia

! fci

•:••:

o

l ^ \\\N v\\ft\ h v S 1 \\u i i C / Weld

Dimensions mm Material : AISI 316 stainless steel

5 Curi2 : e7 FigAmericium-241. . Berylliu1- m Annular Source.

23 This source is closed at one end with a flange containing a tapped hole

for the handling rod.

1.1.6. Transport container.

2^1 The transportation of a 25 curie Ann-Be- source must conform to

the IAEA regulations, Safety Series N*6, 1973. These specify that the

source encapsulation f speciao mus e b t l ford e thatransporan th m t t

containe f o Typ e b re B(u) e surfacTh . e dose rate mus e lesb t s than e dos0 mRem/th 20 e emeted on ratan ht ra e froe surfacth m e mus e lesb t s

than 10 mRem/h. Such a container is shown in Fig. 1.8.

eal ap Closure sealing plate

hielding plug

Container body

Source retention packing

Shielding plug

Sealing plate

Cap

10 cm

: Transpor 8 Fig1. . t Container

24 1.1.7. Safety aspects.

It is necessary to consider all aspects of safety including

general aspects such as adequate lifting and handling equipment as well

as radiological safety, catering alsr possiblfo o - e ac exposur o t e du e

cidental, negligent or criminal action.

Americium-241 as wel as other possible «-emitters are in the highest

clas f radiotoxio s c nuclides s essentiai t I . l therefore thae sourcth t e

is designe d manufacturean d e highesth o t dt possible standards. Also

that the facility as a whole contributes to this safety and that main-

tenanc s carriei e t conscientiouslou d y during service.

e InternationaTh l Organisatio r Standardisatiofo n n (ISOs ha )

produce a systed f classificatioo m f sealeo n d radioactive sources based n safeto y requirement r typicafo s l usesr activatioFo . n analysis pur-

poses the following requirements should be met :

Temperature : - 75'C (20 min) + 400'C/1h) and thermal shock 400*C

to 20'C

Pressure : 25 k N/m2 - 2MN/m2

Impact : 200 g from/m

Vibration : 30 min 25 to 500 Hz at 5g peak amplitude

Puncture : 10 g from/m

1.1.8. Source installation.

The activation analysis facility should be installed in accordance

with locad internationaan l l regulations int n area o a with controlled

access. There should be barriers with adequate warning signs erected

alone 0.7th g5 mRem/hr isodose line e e sourcwitworkinth th hn i e g

position. This is the maximum permissible dose rate for a

25 non-classified worker or a member of the general public. For

classified personna e maximuth l m dos5 mRem/h2. e rats i r e assumina g

t>0 hour working week. Ttie necessary shielding precautions should be

taken to comply with this requirement. Care should be taken at all

time to ensure that any adjacent rooms or corrid.ors to which people

have access do not have dose rates in excess of those mentioned.

A suitable installation would have the activation analysis

facility supporte a framewor n o d k approximatel a hal d fan metere on y s

above round levele transporTh . t containe n thee ca b placer n d

underneath the facility and the source transferred vertically from the

container inte facilitth o y usin a suitablg e long handlin s showa d nro g

in Fig. 1.9.

Neutron Activalion Facility

Source Handling Rod

Transport Container

241i 2C 5 Am/Be Neutron Source

Support SLand

Fig. 1.9 : Source Transfer Procedure.

26 The bottom screening plu s thei g n replace d lockean d n positioi d n

in the facility. The handling rod is unscrewed and the top screening

plug placed in position until the pneumatic sample transportation tube

is e placereadb o n t positioyi d n insid e annulath e r source.

Once sourcth e s beeha en locate n positioi d a surfacn e contamina-

tion wipe test should be performed of the source and empty transport

container. This wipe test shoul e repeateb d d each time sourcth es i e

reinstalled or be checked at least once every two years by wipe testing

the source.

The neutron and gamma dose rates should be measured around the

facility to ensure the screening integrity. These measurements should

be repeated from time to time to assess the integrity of the screening.

The neutron activation facility shown in Fig. 1.8 has been desig-

neo havt d a totae l surface dose rat f appro e 5 mRem/h2. . . Paraffir o n

expanded polyester is preferred to water on account of possible

leakage.

a „. 1 Alaerts e.a. (I.e.) installe n annulaa d r Ac-Be-sourcs n. 0 1 f o e e sourcth , ecm 0 bein5 directly b g watertana m c placen i 0 y5 df o k

directly neae pneumatith r c tube neae bottomth r e leaTh .d shiel0 1 f o d

cm together with the cadmium foil insures a dose-rate at the surface of

appr. ImRem/hr e excepprolongatioth e n i tpneumatith f o n c tube.

Unacceptable dose rate f neutrono s d associatean s d f o gamma' n ca s

course be avoided by installing a beam catcher on the ceiling. A scheme

e installatioth f o s givei n n Figi n . 1.10.

s obviouIi t s e replacethab e wate th n t ca ry b eithed r paraffine

or water expended polyester. A water purification system is then no

longer required.

27 7.5x7.5 Nal(Tl) detecton i r 6cm lead castle holder Printer

Pneumatic Transport

Photoelectric control of e presence th th f o e irradiation container

Cd shield

Pb shield

?27 Ion exchanger Containerstop Ac-Be source Pump

Fig. 1.1 : Genera0 l schem f irradiatioo e d measurinan n g facility.

1.1.9. General safety.

It is obvious that the general environment should be clean and

fre f greaseo e , dird gritan t ,n adequat a thae floo th s t ha re covering

and that good house keeping is maintained. If compressed air is used

28 for a pneumatic transport system the lines should be kept free from oil and dust by means of properly maintained filters.

Access to the area should be strictly controlled and wherever possibl a elockabl e room shoul e provideb d o prevent d t unlawful entry.

1.2. Spontaneous fission sources.

A number of transuranium elements not only desintegrate by a decay, but also by spontaneous fission releasing several neutrons in 252 the process. Among these only Cf is widely applied. The properties of this isotope is summarized in Table 1.3.

252 Nuclea: TABL 3 1. E r propertief C f o s

Decay Mode

«-emission 96.9 I

spontaneous fission 3.12

Half-life

«-decay 2.731 y

spontaneous fission 85.5 y

effective 2.646 y 1 - 1 - 2 1 Neutron emission rate g 2.34x1. s . n 0

Neutrons emitted per fission 3.75

Average neutron energy 2.14 MeV

Gamma emission rate 1.3.10 photons.s .g

Dose rate at 1 meter in air 3 - 1 ~ 1 neutrons 2.2.10 rem.g . h

gamma s 1.6.102 rad. h'1.g"1

Heat generation 9 W/g"3 1 3 -1 Source volume < 1 cm .g

29 252 A neutron spectrum of Cf is shown in Fig. 1.11.

100 oo ^ 01 C 01

4J •r-l C 3

en C o

01 C

(-1 I01

n) ,—i o)

024 ENERGY (MeV) 252 Fig. 1.11 Cf fission neutron spectrum.

252 From these properties it appears that Cf has a number of advantages

over (a.n)-sources : small size of the source, close similarity of the

neutron spectrum with one obtained from fission of uranium. The main

disadvantage is obviously due to its relatively short effective

half-life of 2.65 y. Thus source replacement has to be planned for.

This can best be accomplished by starting for instance with one source

and increasing the number of source by one unit every one or two years. f courso s ha e e also consideOn tw o e financiath r l implicaton f theso s e 24 1 replacements compared for instance with a single purchase of Am-

Be-source e Consultants th e tim th f o et A . ' meetin n "Isotopio g c Neutron

30 Sources for Neutron Activation Analysis" in May 19fl5 at the IAEA the

1 _ p price of a source of 10 n.s were respectively £ 3.500 and1 £ 26.000

rt f» *J O / 4 i o thas m t A ove 7 replacemenr d an f C fort woul e samth de leao t d

expenditure. Other aspects have of course also to be considered as for

instance the neutron energy spectrum, which might in fact depend on the

considered type of applications.

1.3. Photoneutron sources.

Whee y-energth n y exceed e bindinth s g e neutroenerge th th f n o yi n

nucleus a (y,n ) -reaction can take place. Only beryllium (Q = - 1.665

d deuteriuMeVan - 2.22 )= Q 5( mMeV e )considereb nee o t d d since most

nuclei hav a neutroe n binding energy abov 8 MeVe n I practic. e only

124 Sb is used as a gamma source. This isotope has a half life of 60.2

d, a gamma of 1.691 MeV and 2.091 MeV with intensities of respectively

;9 I and 6 I. 124 n a assemblIn s a representey n Figi d .b wil S 1.1 e lcurion 3 f o e 6 result in an emission of about 3 x 10 neutrons per second. The y-dose

rat s quiti e e high namel f aboun causo y ca e mete0 1 mGy/ton d e t an ra h

) cm 5 0. = Sb-B r ( e

N

0.3 r„ = 1.0 cm r„ m =c 1.5 62. c= m „ r . I 0.2

0. 1

10 20 10 20 10 20

Fig. 1.12 : A Be(y.n) neutron source.

31 1 24 serious handlin s producei gb S problems y neutrob d e Th n. activation

f naturao l antimon a nuclea n i y r reactor a therma n I . .n l 0 1 flu f o x 1 - 2 - e obtainon s 4 Ci/s. gm froc m natural antimony. Sinc e half-lifth e e

is only 60 d. reactivation is required 2 or 3 times a year.

For a single y-energy the spectrum is a near monoenergetic neutron

line at 24 keV although some broadening takes place caused by the fact

thal neutroal t n emission angles occur (see equation 1.2).

A Sb-Be photoneutron facility, specifically designe r neutrofo d n

activation at thermal energies is shown in fig. 1.13.

Fig. 1.1 : Sb-B3 e graphite assembl ) graphite1 : y ) cadmium2 , ,

3) beryllium, 4) container with Sb source, 5) samples,

6) rod.

1 24 The Sb is surrounded by beryllium which incorporates channels for

e samplesth e berylliuTh . s surroundei m y b graphitd e servina s a g

further moderato s neutroa d an rn reflector e advantageTh . s wer- de e

scribe s followsa d .

a) Since the energy of the photoneutrons is only 24 keV fewer colli-

sione neede ar r neutros fo d n thermalizatio e mornth e thar enerfo n - 252 getic Cf or Be(a.n) sources.

32 e thermaTh ) b l flux gradien e samplth t a et locatio s smalli n ; this can-

e achieveb t no d with hydrogenous moderators. e absencTh ) f c o eenergeti c neutrons reduce e numbeth s f interferino r g

reactions such as (n,p) and (n,a) reactions. On the other hand acti-

vation analysis by means of these reactions is impossible. A silicon

O Q O Q determination by means of the reaction Si(n.p) Al is not feasi-

ble. 124 ) Since d '"Sth eb withdrawe sourcb n ca e n shieldenow ints it od con-

tainer, samples can be loaded safely.

1.4. Neutron multipliers.

By surrounding a neutron source with a moderator and enriched

uraniu o calles a md neutron multiplie s obtainedi r e assemblTh . s i y 7 -2-1 subcritica e t possibleneutroar bu l0 n.c1 s o n. mt .fluxep u f o s

Type RS-1 multiplie s describewa r ds showi Gambarya d n Figi nan ) . (6 n

1.14.

Fig. 1.1 4: Schemati c drawine subcriticath f o g l uranium assembly 55 RS-1 . (1) Active zone, containing granular UO (U en-

riched to 36 I in 235. U) mixed with polyethylene (660 g 235 U); (2) lead shield; (3) shield of paraffin loaded with

I boro 5 n carbide ) wate(4 ; r shield.

33 o *3 o O *5 C It contains a Pu-Be source and 560 g U (36 per cent enriched

uranium dioxide granular, mixed with polyethylene)9 e powe0. Th s .i r 7 -2-1 watt and the flux 1.3 x 10 n.cm .s . The advantages of the neutron

multiplier are high flux stability, simple operating conditions and the possibilit o irradiatt y a large e numbe f sampleo r s simultaneouslye Th .

neutron multiplier can be operated for a practically unlimited period.

34 2. PRINCIPLES OF ACTIVATION ANALYSIS

2.1. General equation.

Neutron activation analysis is usually carried out with thermal

neutrons through a (n , y)-reaction, i.e. neutron capture by the target

nuclide :

*M (n.r) A*>

a Itargef t s placei a witneutro N m n c i hnucled r npe i > bea

.o-.N (cm~3.s~1) (2.1)

From thie obtainon s e definitioth s e cross-sectioth f o n n

numbe f interactiono r) s . m (c s ————————= ) m (c o- ———————————————— (2.2) «(cm .s ) N(cm )

2 Thue cross-sectioth s s expressei n n area s a a dr mor e.go em .c practi - -24 2 cally in barn (1 b = 10 cm ).

e considerIon f a sampls N f nucleo e i wit a crosh . irradiatea s d

in a constant flux during an irradiation time t. according to the irr reaction :

a X > N ————> N (stable) 1 (n,y)' "2 ' "3

where a is the reaction cross-section and X. the désintégration

s considerei constan N the) f i ns ( d t constant, i.e e burn-u.th s i p

negligible

35 dN 1 N - " dt

Integration between t=o and t. for N = o at t=o gives

«a N f l N, = ____ 1 - expl- X2 t.rr) (2.0

and

= X„N A = O.a.N„ . \ - exp(- >>t) (2.5)

n désintégratio(i wher A e r second e activitpe nth N s i afte )f n o ya r

irradiation time t irr

More complex cases, e.g. several transformations, successive types

of decay, burn-up of N and (or) N should be dealt with by a more

general equation develope y Bateman-Rubinsob d n (7).

If one considers

e isotopith . 1 c e fractioirradiateth f o ö n d element

2. N Avogadro's number, M the relative atomic mass of the element, w A th e irradiatee th mas f o s d element

3. a waiting time t between the end of the irradiation and the coun-

e induceth tin f o gd activity

„ e durindecaN th e f countin4. o yth g t m ti g 2 m then the weight of the element to be determined is given by

A.M. _ . w = o.a.N .s'.e.D.c.e (2'6) A

w NO as N = ———— D = exp(-Xt.) d (2.7 M )1 1-exp(-xt ) = 1-exp(-K S ) t —— - ITT Xtm 36 If gamma spectrometry is applied one has to take into account the peak

detection e detectoefficienc th e absolutth f o d t energa ran e y e E y

t energintensita e gamm . th y ra E f yao y y

If one defines the specific activity as

0.N A = —t?- Q.a.y.e (2.8) p M p s

e numbeth f désintégrations o a r an N d s measured equation e (2.6b n ca ) P writtes a n

N

(2 91 S-.D.cA '. t m sp '

Instea f usino d a singlg e irradiation, waitin d countinan g g timea s

number of successive sequences can be performed in order to enhance the

signal to background ratio and improve the precision on the counting

statistics.

) derive(8 . e cumulativGivenn equatioa dal th t r e sfo n e detector

respons d optimisatioan e n timin h t cyclg e n parametersth e e th r Fo .

activit s givei y y b n

T -* -2X , , T . -(n-l)XT.„ „ . . e + +...= A.( Ae 1+ ) 4 e(2.10 ) n 1

and the cumulative activity is

-XT,, nXT. n n e ( 1 -e ) —————= A -E - ————— = A — (2.11) e i "XT -XT,2 i=1 (1-e ) (1-e )

e activitth whers r i conventiona fo , yA e l activatiod an n

T = cycle time = t + t * t irr w c n - number of cycles

37 For a given total experiment time t.= nT the maximum cumulative detec-

tor response occurs when t. = t with t = o. Although t = o is in w w c irr practic t feasiblno e t e shouls shora s possible a e b optimutd th d an em W conditio f equalito n f irradiatioo y d measurinan n g time does noet

change.

Actually the signal to background ratio should be maximized by

optimization of the number of cycles. The background will be due to a

numbe f isotopeo r s e matripresenth d oftea specifin an i xn ti n c case

to one isotope only. Numerical solution of equation (2.11) allows to

determin e numbeth e f o rcycle s givin a maximug m signa o backgrount l d

ratio.

2.2. Activatio a Isotopi n i n c Neutron Source.

As already described isotopic neutron source e usuallar s y sur-

rounde a moderato y b d o t slor w e dowfasth n t neutron o thermat s l

velocities. Water, paraffine, berylliu d graphit r an minstancfo n ca e e

be used for this purpose.

In a thermal nuclear reactor the neutron flux is divided into

three components : the fast flux, the thermal flux and the epithermal

flux. The type of nuclear reaction and the cross-section are energy 27 r instancfo l dependentA e e casth f neutroo n eI . n captur n givca e e 28 * e compounth riz o t e d nucleu l A whic n desintegratsca h n differeni e t

modes dependine neutroth n o ng energy, each typ f désintégratioo e n

having its own cross-section :

27A1 * n

27m Al + n

* 8 2 y 28+ A1 27 A1I + n ——> Al 27 P + g iA

a N+ a

n 2 26+ A1

38 2.2.1.

Fission neutrons have a continuous energy distribution with a

maximum higher than 10 MeV. They can be described by Watt's empirical

equation

n(E = 0.772) 5 E1/2 exp(- 0.77 ) E 5 (2.12)

e contributioTh f o fasn t neutrons (abov 1 MeVe o (n.yt ) ) reactions i s

usually small and even negligible. They give however rize to threshold

reactions e.g. (n.n'l, (n.p), (n,«l , 2n)(n , . Som f theso e e usear e d

in activation analysi r instance determinatiofo sth r fo e f silicoo n y b n p O p o Si(n,p) Al. Conversely they can cause interferences for instance 27 20 e determinatioth r fo f aluminiuo ne th Al(n.y y b l mreactio A ) n where

the presence of silicon can give rize to positive errors.

The effective cross section in the fast neutron flux is defined as

r )' (E)dE ————— (2.13) CD ( <»' r E)dE

J n

n unperturbea n i Valuea r fo ds fission spectru e founb n n refei dca m -

. renc) 9 ( e

The reaction rate per nucleus is given by

,2.u,

e equivalenth s i whert > e4 fissio n flux. t S T3

The equations and cross sections mentioned above can obviously not be

applied for I.N.S.'s other than 252Cf. For Be(a,n) sources for instance

39 the neutron energy is not continuous as in a reactor but quasi mono-

energetic around 4 MeV. This means that equation (2.14) and the cross

section value s suca s h canno e applied b te determine b t havo bu t , e d

experimentally. Even for a 252Cf it is quite unlikely that the experi-

mental set-up will giv en unperturbea riz o t e d fission flux.

2.2.2. Thermal

Fission neutrons loose their energy through successive collisions

e moderatowitth h r unti n thermali thee ar yl equilibrium. e Thear y

called thermal neutro d folloan n a w Maxwell-Boltzmann distribution.

The most probable neutron velocit t 20* a ys 220 i Cs 0m. correspondin g

n energa to f 0.02o y. DiffereneV 5 t convention e applie- b de n o ca st d

scribe the reaction rate for ( n . Y ) reactions, as for instance developed

by Westcot y Stoughtob d an td Holperinan n e mosTh . t practica r actifo l -

vation analysi s probabli e sproposeon e y th yb Hjlgdah d l valif i d

a(v) - - (2.15) v

e casr almosth fo eThis l nuclidei al st f o interess n i activatiot n

analysis .

The neutron flu d thue reactioan xth s n rat s dividei e d o inttw o

parts usin a Cd-filterg . Conventionall a Cd-cylindey s usei r d (height-

diameter 2/1; wall thicknes ) thamm a 1 cut-ofns = 0.5f energ5 E f o y

eV is accepted. Thus a sample irradiated under Cd is only activated by

epithermal neutrons and the conventional subcadmium flux $ is the sum e conventionath f o l maxwellian flux plu e epithermas th par f o t l flup u x

to an energy of the cadmium cut-off.

40 2.2.3. Epithermal neutron flux.

Neutrons loosing energ y collisionb y s wite moderatoth h t whicbu r h

havt reacheno e d thermal equilibriu e callear m d epithermal neutrons.

e ideath In l e casepithermath e l neutron flur uni f pe xo tenerg y

interva s inverseli l y proportional wite neutroth h n energy.

• (E) - (2. 16)

2.2.4. Reaction rate for (n,y) reactions.

By irradiatin a nuclidg a neutro n i e n flux wher o flutw ex compo-

nent e presenar s t namel e conventionath y l subcadmiue th d an m > flu

conventional epithermal flux e n rate wil e giveb l y b n

« I + * a = R (2.17) e o s o

As already stated equation (2.17 s derivei ) d froe H6gdahth m l conven-

s onli tiod y an nvali f equatioi d n (2.15 d (2.16an ) e validar ) .

If an nuclide for instance 1 97 Au as an aluminium alloy is irradia-

ted d witwithouan h a cadmiut m filte e respectivelth r y reaction rates

wil e giveb l y b n

e o R,C.d = I O

= * s o e

wele ar l know I d nan equatio a f I solvee nb fo dca . an > 7 e s o o Considering eq. (2.17) eq. (2.9) can new be written as

(2.18) t ' N .e.y. (a «• I * JS.D.C.e m At o s o e p

41 e tabulatear I r instanc d fo . de an CortD al Value y t a b e e f o s o o ( 10).

2.3. Standardization.

If the neutron activation facility is kept unchanged i.e. source,

moderator, irradiation geometry etc. it is obvious that the flux will

also be constant and the induced activity will only be a function of

iradiation, measuring and decay time. Only the source strength as a

functio f timo n e wil le takeb hav o t ne into accounte decath s yA .tim e e sourcth s f weli o e l known t shouli , e b possibld o standardizt e e th e

system only once. This can in principle be carried out according to

three methods

2 3.1. Absolute method.

Equation (2.18) allow e determinatioth s e neutroth f o n fluxes

from the irradiation of a suitable monitor with and without cadmium.

Hence equation (2.18) can be solved by an absolute measurement of the

induced activity if the désintégration scheme of the isotope of inter-

est is known and the detection efficiency of the detector is also well

known.

This method canno e considereb t s practicaa d o t man s ya l uncer-

tainties are introduced.

2.3.2. Relative method.

Each sampl s irradiatei e d together wit a knowh n e eleamounth -f o t

ment (s) to be determined. If the neutron activation facility is kept

unchanged co-irradiation of sample and standards is not required as

42 neutron flux variations do not occur as can be the case with a nuclear

reactor. The standard (s) can either be the pure element or a compound

of known stoechiometry. They can be used separately or as a mixture.

If available, standard reference materials (SRM's n als e ca used)b o .

Obviously the result will be given by

N N P w = w P (2.19) X S S. D.C. t € S. D.C. t e m p X m p

Equation (2.19) only holds if the following conditions are fulfilled :

1. Sampl d standaran e d shoul e exposee b samd th eo t dneutro n fluxa f I .

flux gradient occurs vertical and (or) lateral as is usually the

case in an isotopic neutron source facility, a correction might be

necessary if standard and sample do not have the same geometrical

dimensions. If this is not the case a correction should be applied.

Obviously this correctioe determineb o t s ha nd experimentally. e neutroth s 2A . n a flu I.N.Sn i xs i usuall. y quite modest large

sample e oftear s n require o obtait d a reasonabln y high induced

activity. Large samples (or standards) can give rize to neutron self

shielding which can be different in sample and standard. A correc-

tion is difficult to calculate especially for samples with an

irregular geometric shape n generaI . l equations develope o correct d t

for neutron self shielding should be considered as approximations

and allow to estimate if the effect is important enough to be taken

into consideration. In fact the neutron flux can be modified in the

sample by selfshielding, but the presence of a sample and its

container can change the flux in the sample through flux depression

and neutron scatter f appreciablI . e amount n hydrogea f o s r carboo n n

compoun e presenar d n sampli t r containeo e e thermath r o epithert l -

mal and fast flux ratio's can be modified.

n e I.N.Sa sizth A f ss usuallo ei . y quite smale irradiatioth l n

43 sites are quite limited. For an annular source only one sample or

standar e irradiateb n ca da time t a d. Althoug . 2.1eq h 9 correctr fo s

a difference in irradiation, decay and measuring time between sample

an da differenc standarr fo d n measurini an de g geometr s obvii t i -y

ous that the greatest possible similarity between sample and stand-

ard should be aimed at.

2.3.3. Single comparator

I.N.S.'s are quite stable and have a constant neutron output (if

corrected for decay) if the facility is kept unchanged. Once the

specific activitie e elementth f o sf interes o s t have been determinen i d

well established conditions samplee irradiateb n ca s d without stand-

a precautio ardss A . r unforeseefo n n change s gooi t di s practico t e

monitor the neutron flux within a series of experiments to check if the

neutron flu s indeei x d unchanged.

2.3.4 k -standardization. .

Instea f usino e d absolutth g e nuclear datd thermaan a d epian l -

thermal neutro e relativus nn fluxeca ee valueon s : s

9 NA » J.y. I e + —rr= « A s-a p( M (2.20os ) o e p

9 N Q }.7.e p + f ( 9 o o - z e — = o M

where

(2.22)

and

44 I — = Q (2.23) ao

Thus for a sample and a monitor one obtains

„. 8 M a y f + Q € ) — 24 ____2 ( — £ _ °— — A - A sp sp ' * ' ' * • * • * • * i'-'*> € Q + f y a M e M M p o o

where the asterix refers to the monitor. Equation (2.24) can be writ-

s a n te

A = A * . k .k,.k (2.25) sp sp o 1 ?

y a y M 0 k -_____£ (2.26) o * ' * ' * a y M 0 o

f * Q ————= k y (2.27) f + û o

(2.28) € P

Obviousl . e (2.24solvea singlb eq y f n i ca )de monito s usedi r , e.g.a

gold a golfoi r do l alloy (0, gol. 7 n 1aluminiumi d n conditioo ) n thaf t

is known. Thi e flu se th b determine xvaluf i n ca l e al r d fo oncd an e

characteristics remain unchanged. This can be calculated from the Cd-

ratio of a suitable monitor e.g. a gold monitor or from a dual monitor 94 96 for instance the two isotopes of zirconium namely Zr and Zr.

General Conclusion.

e generath f 1I . l a set-uneutro f o p n sourc s kepi e t unchangede th ,

neutron output and hence to neutron flux and its energy spectrum

will remain constant, except obviously for the decay of the source.

45 e neutroTh . 2 n energy spectru e sourcth f o em differ a considerabl o t s e

extend fro a nucleam r reactor spectrum. Cf , e casexcep f th o e n i t 252

Indeed fission neutrons have a continuous energy spectrum, the average

energy being appr. 2 MeV and a most probable energy of appr. 0.8 MeV.

Sourcee l«,nth f )o s type monoenergetiar e t a cabou 4 MeVt , whereas

(y.n) sources produce neutron w energlo t a sy e.g 5 keV2 . .

Hence conventionath e l neutron fluxes describe a nuclea r fo d r

reactor, i.e. thermal- epithermal d fasan - t flux canno e applieb t o t d

I.N.S's. In practice it is however useful to determine those fluxes by

mean f generallo s y accepted techniques.

2.4. Sources of error.

2.4.1. Flux gradients.

As the source size is often rather small important flux gradients

can be expected. These can be determined experimentally by placing a

wire centrall e samplth n i ey holder .r n instancfo ca Piece m f m o 5 es

cue tthear froe n d wirth man measurede e lateraTh . l gradient a t

different heighte measureb n ca sy irradiatin b d g circular foils which

are cut in sectors. The flux gradient should be measured for the

different types of flux e.g thermal, epithermal and fast if these are

applied. One should also keep in mind that the maximal fluxes do not

always coincide.

In general it is good practice to irradiate samples and (or)

standards of the same size in the same geometry to minimize this type

of errors.

46 2.4.2. Neutron shielding.

As the neutron fluxes are quite small it is advantageous to use

rather large samples to improve sensitivity and count-rate. Non negli-

gible neutron selfshielding effects can thus be important especially if

element e presenar s t with large absorption cross sections. Corrections

are rather difficul o calculatt t d equationan e s which correc r thifo t s

effect are not always reliable.

In fact the neutron flux can not only be modified by selfshielding

but also by the presence of the sample itself or by the use of a

cadmium box.

The total correction for the change in neutron flux as seen by the

e producth o f t threo t e sampledu effects i e : s

s F D F A F = F (2.29)

= selfshieldin . F = flu . F x g; depression ; A D F = flux scatter factor.

The selfshieding factor can be calculated approximately for foils,

cylinders and spheres

E (0.922 d. g -1—5 lo 0. + 8 - ) 1 = (2.30F ) Q € 3 T , A

1 « e a R ; e a R. 4 (2.3 - 3 1 1 = F.A ,c

FA4 , S. = 1 - Rea ; Rea « 1 (2.32)

where d is the foil thickness, R the radius of an infinitely long

cylinder or the sphere diameter, E the macroscopic cross-section.

47 2.4.3. Nuclear interference.

2.4.3.1. Threshold reactions.

As already stated in 2.2. the type of nuclear reaction induced

with neutrons depends on its energy. Fast neutrons can induce so called

threshold reaction. As a (n.p) reaction on element Z + 1 or a ( n , or I -

reaction on element Z + 2 can give the same isotope as a (n,y) reaction

on element Z positive errors can occur. This is for instance the case n aluminiua r fo m determinatio e presencth n i nf silicono e a manga r fo ,-

nese determinatio e presencth n i na f sodiu iroo er o n m determination i n

the presence of aluminium.

27Al(n.y)28Al

, ,. „-, Si(n,pI A )

55Mn(n.y)56Mn

56,-Fe(n,p , n M ),56 U

23Na(n,y)24Na

27Al(n,a)24Na

Usually these effect e onlar sy importan e fas o th thermat tf i t l

flue concentratioth x f ratii e elemens higi d oth an hf o nt giving rize

to the interference is a main component.

This type of intererence can be corrected for experimentally by

irradiatin e purth ge element which give e interferences th riz o t e .

2.4.3.2. Fission reactions.

If uraniu r othe(o m r fissionable material s e i presen)th n i t

sample e fissioth , n product n givca s e rizo t importane t positive

48 errors, espescially for those elements giving rize to high fission a giveyields r nFo isotop .e activit e th fissio th i e f o yn producs i t

given by (for 1 g of uranium)

A (n.f) = J a(n,f) O.N S /Mu (2.33) i l Ai o(n.f) = fission cross section for natural uranium

= atomic Mweigh f uraniuo t m

S' = saturation factor with x. désintégration constant of

isotopi e

J = fission yield of isotope i

The activity by (n.-y) reaction is given by

A (n,y= o(n,r) N S./)e* M (2.34) JL A JL X

so that the ratio of the activities is given by

A (n.f) o(n.f).A . J _i____ = ______1 x (2.35) A (n.f) 233.e.o(n.r)

This ratio has the following values :

U1Ce 0.54

U7Nd 0.20

9°Mo 0.85

Once this interference is determined experimentally an appropriate

correction can be applied.

49 . PROMP3 T GAMMA NEUTRON ACTIVATION ANALYSIS (PGNAA)

Capture of a neutron gives rize to prompt gamma rays from nuclear

excited states. These have half-lives 4 of-1 only 2 11 0 - to 10 seconds.

These gamma's have energy ranges from about 50 keV to about 10 MeV and e usuallar y quite complex s alsi t oI . obvious that nuclides which give

rize to stable isotopes by neutron capture can also be analyzed;

For PGNAA the measured activity under a photopeak of a gamma of energy

s giveEi y y b n

N = NcTtfl .6 .g. t (3.1) P P where

N = number of counts under the photopeak of energy E

N = number of nuclides giving rize to a prompt gamma of

energy E 2 a - neutron cross section (cm ) - 2 - 1 = neutroC>) s , n flum (c x

I = number of prompt gamma's of energy E emitted per neutron

capture

= detectio e n efficienc t energa y E y P y g = geometrical factor depending on the source-sample-detector

arrangement

= totat l experiment time (seconds)

From (3.1) it is obvious that the weight of the analyzed element will

e giveb y b n

N .M .g.t.N p A

Using an INS the method cannot be applied for trace analysis but only

for the determination of major and minor elements in bulk samples as

50 coal, ores, mineral n procesi d an ss stream control. Typical elements

e analyzeb whic n e hydrogenca har d , carbon, nitrogen, sulphur, sodium,

potassium, aluminium, silicon, calcium, titanium, iron arid chlorine.

An outlay for the onstream analysis of coal was described by Herzog et

al. (10) and is represented in Fig. 3.1.

Sourer drive

Sample Conta iner

Detector Sftie/tf/ng

: Irradiatio 1 Fig3. . n Facilit r PGNAAfo y .

252 Usin a sampl a gn o eg Cf-sourc m correspondin2 f o e o apprt g .

120 kg using a Ge detector with an efficiency of 22 7. an analysis is

possible for carbon, hydrogen, nitrogen and sulfur as well as for the

major ash components (silicon, aluminium, titanium, calcium, sodium,

potassium, d irochlorine)an n e correlatioTh . n with independent

chemical analyse e s highlfounb wa so t dy satisfactory.

51 4. THERMAL NEUTRON ABSORPTION ANALYSIS

4.1. Principle.

The neutron absorption of elements with a large neutron cross-sec-

r instancfo tio s a n e cadmiu e determineb d boro an n m ca n n matricei d f o s

low neutron absorption. The attenuation of the neutron flux can be

determined either by direct measurement of the neutron flux for

instanc eF -counterB wit a h r moro , e practicall y measuremenb y e th f o t

activity of a suitable monitor. The latter technique measures the

differenc n activiti e y e monitorinduceth n y i neutrodb , n irradiation,

with and without a layer of the sample under analysis.

In e facfluth t x perturbatio e b divide n ca dn int a fluo x

depression in the vicinity of the sample and a self- shielding effect in

the interior of the sample as shown schematically in Fig. 4.1.

: Schemati 1 Fig4. . c representatio f o flun x perturbation factor: s

/* =flux depression factor, */» =self-shielding factor. /1> -flux

perturbation factor.

The flux 9> at the surface of the sample is lower than the unperturbed

e resultinTh flu. 9 xg flux depressioe selfTh -. « / n » rati s i o O S - 31 0 shieldin a g significan o e t effecinterio th e n du i ts i rt absorptiof o n

the thermal neutrons by the sample As a result the mean flux 0> in the

sample is lower than * at the surface of the sample. The total effect

f fluo x depressio d self-shieldinan n s givei g y <&/b n «.

52 If the unperturbed flux S>o and the attenuated flux c> are measured, a quantitative relationship can be expected between the total perturba-

e fluth tio xf o nexpresse n absorbanca e amounth s a dd f an o g te/

absorbing material. De Norre et al. (11) examined the determination of

cadmium in zinc ores. From the composition of the zinc ores and from

the M rati/ (absorptio o o n cross-section over atomic weightn ca e on )

estimate the effect of neutron absorption of the ma^or and minor compo- e versuor e Cd-concentration e th snentth f o s . Thi s illustratei s n i d

Table 4.1 where the a /M values are given 3

e elements th Lis: TABL f 1 o t 4. E , presene zinth c n i orest , together with some important characteristics (see text)

Element Cone. range in Concentration , /A Concentratio a n zinc ores , 1. in graphite blank. barn equivalent to #9/9 0.005 7. Cd, 'I.

Cd 0.02 - 1 .45 21 .83 Zn 39 68 .5 0.0168 6.5 Fe 2 0. 0 12 .9 0.0457 2.4 S 0 32 .5 0.0162 6.7 Cu 0.1 - 3 .0 0.0593 1 .8 Cl 0 1 3 0.931 0. 12 In 0 0.03 1 .690 0.065 B - <0.01 70 . 21 0.0016 Si 0 1 .6 0.3 0 .00570 19 Al 0 0.6 0.00864 12 .6 Ca 0 1 .8 0 01073 10 .2 Mg 0 0.4 0.05 0.00259 42 Mn 0.07 - 0.4 0.242 0.5 Pb 0.3 - 20 0 00082 133 As 0 0 . 7 0 05739 1 .9 Sn 0 2 .4 0 00531 20 .6

53 ae concentratios th wels a le elemen th 1 givin f n o n i t n absorbanc a g e

equivalen o 0.00t t cadmium. 7 5 .

4.2. Detectio e thermath f o nl flux.

Different techniques of monitoring have been applied. Spenke et

c q al. (12) determined boron in steel using Fe as an internal monitor

whereas Selecki (13) used an indium foil between two layers of the

sample . (I.e.e Norral D .t e e) use a smald l cylindrical vanadium moni-

tor inserted into the cavity of the sample as shown in Fig. 4.2.

Fig. : Irradiatioi< .2 n geometry applie n thii d s wor= vanadiu A k m

detector; B - sample material.

4.3. Standardization.

To determine the imperturbed flux <3> a graphite cylinder contain- o ing the vanadium monitor is irradiated. The induced activity A in the

vanadium detector is proportional to * . To analyze for instance a

zinc ore a mixture of powdered ore and graphite (ratio appr. 1:2) is

pressed int a cylindeo e samth ef o dimensior e activitTh n y inducen i d

the vanadium monitor A proportional to * and is reduced due to the m absorption of thermal neutrons by all the constituants of the sample :

54 A T - 9 10 = E log (4.1) m tot i

S ... . .Fe , Zn , Cd = i

In practice a calibration is necessary for only for zinc, iron and

sulfur. Thi s obtainei s r eac fo f thesdo h e elements from cylinders con-

taining only these elements in graphite. Fig. 4.3a shows the absorb-

ance of the elements as a function of their masses.

They give rize to a linear absorbance and can be expressed as

A 109 - b.g. (4.2) Ag

Thue calibratioth s n linr cadmiue expressefo b e n ca m s a d

A A 0 0 , log — = log _ b.g. (A.3) * E 11 m tot X Cd i

e calibratio wher= slopth . f b eo e n lin f elemeno e i t

g. = amount in gram of element i

Fig. (4.3b) shows a calibration line for cadmium. The detection limit

for cadmium is appr. 0.005 7. whereas a relative precision of better

e achieveb n ca s thacadmiuI i d 1 n 1 percen s preseni me th t a tlevel .

4.4. Other applications.

It is obvious that the same method can be applied to the determi-

natio f otheo n r elements with high absorption cross sections n obviA . -

ous choice is boron, sometimes used for instance in special steels.

n alsca e determineb ou A Othe d an r d g A elementalthoug , In s a sh they

give rize by neutron irradiation to radio-isotopes which can easely be

measured directly or after chemical separation.

55 r lOrafo d C gA / IA g

0.020

10 gXgrams)

Fig. 4.3a : Calibration lines for zinc, iron, sulfur and the absorbance for 10 mg cadmium.

Fig. ^.3 bCalibratio: determinatioe nth linr fo e f cadmiuo n n zini m c ores .

56 5. TRANSPORT SYSTEMS

As already mentioned in 1.1.8 and shown in Fig. 1.10 it is

desirable to install a pneumatic transport system to insure reproduc-

ible irradiation and decay times of the samples, espescially when

dealing with short lived isotopes as is often the case. A suitable size e . pneumatidiameteth mm 0 f 3 o o thas re cm b tm tube sampl0 n 2 ca sf o e

diamete e use b n adequat i n d ca r e containers made from polyethyleer o n

nylon. Suce b controllea systeh n ca my b timind g unit o regulatt s e

irradion, deca d countinan y g time e presence th ,sampleth e th f t o ea s

irradiation site being detected by means of a photocel. A blower and

electromagnetic valves provide the necessary air pressure to transport

the samples to irradiation site and back. The samples can be manually

unloaded and placed on the detector. A distance of ca. 15 m can easely

be covered within 3 seconds. A sample changer and a two way system can

be included to automate loading and unloading of the samples. Automa-

tion of the transfer system can of course also be achieved by means of

a microprocesso r instancfo s a r e describe e NorrD y b ed e.a. (14).

The complete system is controlled by a DEC LSI-11 Microprocessor,

consisting of a central processor unit (CPU), random access memory (24K

bytes MOS) and standard serial and parallel I/O inteifares, connecting

the peripherals to the LSI-11 data and adress bus A schematic repre

sentation of the system is shown in Fig. 5.1.

e pneumatiTh c transport syste d countinan m g chai e connectear n o t d

the LSI-11 bus by a standard DEC parallel interface (DRV-11). This

interface contains three 16-bit registers : DRCSR : control and status

register, DROUTBUF : output register, DRINBUF : input buffer register.

Each operation mode of the transport system and the counting chain

correspond a specifi o t s c numerical contene DRCS d th DROUTBUF an f R o t .

57 Central 12 x K16bi t 50 Hz processor Line clock memory unit i LSI-1 1 bus

Standard Standard Parallel DLV|] DLV11 serial Interface I/Q 1[Uerface I/O interface Modified DRV11 parallel O interfacI/ e Pneumatic transport control

Transport syst. position detectors

: Systemati 1 5. Fig. c representatio e automateth f o n d analysis system.

The DRINBUF register is used for transfer of data from the binary

sealer to the microprocessor memory. The contents of these registers

cae changeb n d dynamicall d conditionallan y y under program control

Thus successiv ee analysistageth f o s s e sequencb executed n ca e ,

skipped or modified automatically by changing parameters in the program

before the start of the analysis run.

A comprehensive program is available to control the pneumatic

sample transfer, irradiation, decay and counting time. A second

program, specific for every analysis problem takes care of the data

reductio d interpretatioan n o providt n e final analysis result n termi s s

f normaliseo d and/or corrected count rate d eventuallan s y concentra-

tions r datFo a. froe singlth m e channel analyser, normalisation, back-

ground subtraction, corrections for differences of the irradiation,

decay and counting periods can be carried out. For the data from the

MCA, peak areas are calculated adapted to the specific problem at hand.

The analyses results can be printed out or saved on disk.

58 6. MEASURING EQUIPMENT

6.1. General.

A large choic f radiatioo e n detector s availabli s e today, although

in N.A.A. the most generally used ones are Nal(Tl) scintillation and

germanium semi conductor detector s a sthe y have good detection effi-

ciencie r gammfo s a radiation s filleGa . d detector s a Geiges r Muller o r

proportional counter f courso n ca se als e usedb o . e mainlThear y y sen-

sitive to beta-radiation, with all the problems inherent to the meas-

urement of this type of radiation. Their use is rather exceptional and e discussedb wilt no l .

6.2. Sodium iodide detectors.

In activation analysis wit a hI.N.S e induceth . d activitiee ar s

rather modest. due to the small fluxes usually available, even when

large samples (several grams) are used and minor or even ma^or

constituent o complee determinedto s i ar st e spectrt no i th x e f ar I a .

therefore preferabl a Nol(Tl)-detecto e us o t e f ratheo r r large siz5 7, e

x 7,5 cm or even 12,5 x 12,5 cm. A well for instance of 30 mm diameter

and 40 mm deep is an additional advantage, as the detection efficiency

s increasei e reproducibilitth s wela ds a l e geometryth f o y. Suca h

configuration has a detection efficiency of at least one order of

magnitude larger than even a "large" germanium detector and obviously

improves counting statistic d sensitivitan s e analysisth f o ye majo Th . r

drawback of Nal(Tl) scintillators is due to its rather poor resolution

which typically is 8 Z for the 0.661 MeV line of Cs.

As is well known a Nal(Tl) scintillator detector is coupled to a

photomultiplier to which a highly stable high voltage has to be

59 applied, so that adequate pulses are obtained which can be fed via a

preamplifie a linea o t r r amplifie a countin d an r g systema e .b Thin ca s

single channel analyzer (S.C.A.) or a multichannel analyser (H.C.A.).

The latter has clearly to be preferred as it gives a complete survey of

the spectrum. In both cases a stable amplification has to be achieved

whic s i onlh y possibl t constana e t room temperature. Even thea n

regular check of possible shifts is advisable. If the counting rate is

high, as might be the case for large samples giving rize to short-lived

specie e e takeb death o st dn ints timha o e account. Each detected

e countinth puls n i e g chain need a finits ee b timprocessed o t e .

During this time called the dead time, a new incoming pulse cannot be

processed and is lost. These losses can be calculated from the equa-

tion :

A = A' - A'Ar (6.1)

A - ' A r -- - ——— - (6.2) A' A ' r 2 ' A - ' A = A (6.3I

A

where

A = registered counting rate

A= rea' l counting rate

T = dead time per registered pulse

T' = dead time per incoming pulse

e countinTh g e determinelosseb n ca s d experimentall y b followiny e th g

activity of a relatively short lived species with well known half-life. f1? l The decay curve of I (half life 25.00 mm) is for instance given in

Fig. 6.1.

60 to ü 100 150 200 Time (min)

: Countin Fig1 6. . g o dealosset de a du singlstim n i e e channel ana- 1 28 lyzer (decay of I; T - 25.00 min).

One can see that at a count rate of 2.10 counts per minute the

count losses are smaller than 0.1 1 and allow to determine A' by

extrapolation of the straight part of the decay line.

Fig. 6.2 shows the values of T and r' obtained from equations

(6.2) and (6.4). It appears that neither T or T' are constant as a

function of count-rate and hence that A cannot be calculated from

equations (6.5) and (6.6).

(6.5) 1 - Ar

- I / - (6.6) 2r'

61 -o nj OJ

2 -

1.10 2.10 3.10 A.10 5.10

Counting Rate

2 6. . Fig Dead a singltim f o e e channel analyze r registerepe r d pulse

(r) or per occuring pulse (r'l

One can conclude that for a certain counting chain and counting condi-

tions (e.g. integral or differential counting) one has to determine

experimentally the dead time and limit the count rate to an acceptable

level. If moreover standard and sample are similar the errors due to

dead time problems will to a large extend be cancelled.

In case of a Nal(Tl) detector used in conjunction with an M.C.A.

the latter wil f e maicourso lth ne b esourc f deao e d time.e b Thin ca s

split into three parts due to pulse selection, conversion time and

memory cycle time :

= a (6.7)

T - dead time for pulse registered in channel N N a = constant time interval for entering pulse

- memorc y cycle time

62 Aa consequencs e reath e l measuring tim s shortei e r thae clocth n k

time and increases rapidly with increasing count rate. MCA s therefore

offe a choicr e between live-tim d clocan e k time e formeth , r being

correcte e death d r timefo d . This automatic dead time correctiony b ,

increasin e measurinth g g tim s onli e y possible countinth f i eg tims i e

shor n comparisoi t e half-life th n correcca o t n e r thiOn fo t. s effect

accordin o t equationg s develope y Junob d r countind fo (15) a purg e

short lived nuclid e truth e e activit n count(i yr uni f pe stimeo t s i )

givey b n

exp(XDTA ) ———————————A: ' — . — (6.8) - exp(-X.LT1 T D )

where

A' = true counting rate at t - o

= observeA d counting rate

= tota T D l dead tim f countino e A r. = g

LT = total life time

Equations for counting short lived species in the presence of long

lived ones have also been developed.

The complexity of these corrections can however be avoided through

adequate electronic circuits. Bartosek (16) developed a so called dead

time stabilizer which divide e totath s l counting time int a largo e

number of intervals which are short in comparison to the half-life and

thus corrects exactly for the losses during the counting. Moreover

additional dead time is generated to keep the dead time constant during e wholth e measurin s startegi time C n practicI dAD . s everm e 0 th 1 ye

and the additional dead time in each measuring interval is equal to the

chosen dead time. This can be chosen between 10 7. and 60 I in steps of

5 'I. . Obviousl e choseth y n dead time shoul e slightlb d y larger thae th n

real dead time at the beginning of the measurement

63 6.3. Germanium semi- ctou d n r c o detectors.

Germanium semi-conductor detectors have the advantage of a much

higher resolution than Nal(Tl) detectors, typically betterV thake 2 n o (FWHM)C r 1.3 fo V gamme detectioMe 3Th f . o a n efficienc s howevei y r

usually one order of magnitude lower. Lithium drifted germanium detec

tors are now replaced by so called hyperpure germanium detector. They

have advantagth e e tha e tstoreb the n t rooa ca dy m temperatur t musbu e t

also be operated at liquid nitrogen temperature. Germanium detectors

are available in different sizes, configurations and geometries. Large

sizes are usually of the coaxial type as shown in Fig. 6.3.

p-type 0 micron60 s HPGe Radiation

GEM HPGe Crystal

n-type .0.3 microns HPGe

Radiation

GAMMA-X Crystal

Fig. 6.3 : Configurations of Coaxial Germanium Crystals.

Thee availablar y s p-typa e r n-typeo ee advane latteth Th s .- ha r

tag f havino e a thig n sensitive laye w tenta f onlfe o micron r f a o yh ,

so that low energy gamma's and X-rays are detected with high effi-

ciency. The former have usually an insensitive layer of about 500 am

and thus are insensitive to low energy radiation. This can be an

advantag s higa e h intensit w energlo y y radiation easily gives rizo t e

64 high count rates and detector overload. Well-types are also available e Th . mm 0 a dept4 wit d f o an hh m welm 5 l1 sizeo a diametet f p o su f o r

advantag f courso s i n e a ealmos detectiof 4t t n geometre disadth t -bu y

vantage of greater probability of coincidences. The configuration can

be either horizonta r verticalo le preferreb o e t formeTh r .s fo i d r

N.A.A. purposes. It has the advantage of insuring an easier reproduci-

bility of the geometry and allows the use of a Marinelli beaker if

large samples are to be counted (see Fig. 6.4).

Marinelli Beaker

Sample

Ge Detector

Dewar

Schemati: 4 Fig6. . c Diagram Showing Arrangemen f Marinello t i Beaker.

6.4. Detector calibration.

It is important to establish experimentally a number a detector

parameters of both the Nal(Tl) scintillation detector and Ge-semi-

conductor detector e.g. resolution, detection efficiency and energy

calibration. Experimental catalogue f gammo s a spectr f largo a e numbers

f nuclideo e availablar s e e.g. Adam d Daman s s (17) Keath (18).

65 6.4.1. Resolutio d efficiencyan n .

e resolutioTh s i usualln y definee full-widtth s a da tota f o hl

absorption peak, sometimes calle e photopeakth d e maximu th t hal a ,f o fm

count rate (FWHM) r Nal(TlFo . ) detector e resolutioth s s usualli n y 1 37 given in percent for the 0.661 MeV line of Cs :

Resolution CI.} = AE(F?HM) x 100 (6.9)

This resolution is usually about 8 X. Detection efficiencies are avail-

able in compilations such as those from réf. (17) and (19).

For germanium detector e resolutioth s s usualli n t a y V giveke n i n 60 FWHM for the 1332 keV line of Co. It is also usefull to determine

the ful e maximu lth e widttent on f o t a mh (FWTM s thia ) s given a s

indication of the gaussian symmetry of the peak. Resolutions of better e achievedar V thake 2 .n Detection efficiencie e referrear s o t d

Nal(Tl) scintillators of 7.5 cm x 7.5 cm and vary according to the size

(and. '/ price 0 3 d ) an betwee w fe a n

The detection efficiency, is of course a function of energy. The

total efficienc s correlatei e e pea yf th tota o k o t dl absorptioy b n

et = "P7T (6-10)

where

= pea o T totat k P/ l ratio.

T ratia functioP/ s a e o Th f gammo n a energ s givei y y b n

log (P/T) = a log E + b (611) e constantar b d wheran s a edependin e measurementh n o g t conditiond an s

th e) doe t 1 detectortak1 no s. e6 ( int. q E o. account absorbin r o gscat -

66 tering materials surroundin e detectorth g n practicI .s measurei € e d

by means of calibrated point sources with known absolute activity A

(désintégrations per second) and known absolute gamma intensities.

Usefull sourc e givear en Tabl i n e 6.1. together wite halfth h -

lifes, gamma energie d intensitiesan s . Typicaa s la result e r fo s

function of energy are shown in Fig. 6.5.

TABLE 6.1 : Calibration sources for gamma spectrometry.

E (keV) Isotop T e y Isotope T E (keV) y

226 5 1331 . Ba53 y 10.5* .01950 Ra 1600 y 609.32 .4426 61.00 .3904 (continued) 665.45 .0-1492 356.01 .6240 768.36 .04736 806. 17 .01208 Oil Am 458 y 59.5* .3530 839.03 .005577 C "T Co 271 .8 d 122.06 .8554 934.05 .03054 136.48 . 1066 1120.28 .1469 1238. 1 1 .05710 IS? ** u E y 13.5 121.78 .2821 1280.96 .01385 244.7 « .07423 1385.31 .007923 295.94 .004231 1401.51 .01301 344 . 38 .2641 1509 19 .02080 367 76 .008401 1661 .28 .01076 411.3 * .02301 1729 60 .02833 444.0 * .03077 1 .5 1764 . 1509 564.5 « .006133 4 .5 28 11 01177 688.6 « .008493 2204 . 12 04957 778.90 . 1300 2293.36 .002988 867.38 .04161 2441 7 7 015U 964 .0 * .1448 137 * 4 1068 0 6 1 4 66 18 1 . y 2 0 3 s C 8462 5 4 1112.1 08 *4 83 . 135 d 5 2 2 31 n M 9998 1212.94 01390 88 1298 7 4 0 « 8 89 0174 d 3 6 6 10 Y .9340 1408.03 .2071 1836 13 9940 226 *** 60 fia 1600 y 186 02 .0340 y 1 2 1172 83 27 5 o C 9988 241 91 .07082 1332 49 .9998 274 53 .005178 Na 2 604 y 511 0 1 811 7 1 295. .1810 351 90 3510 <£*! 487.08 004342 1274 51 9995

67 i

10

-4 10 E (keV) 0 1 10

Fig. 6.5 : Experimental peak efficiency curve for point sources.

6.4.2. Energy calibration.

Gammaspectrometry is mainly applied with multichannel analysis and

the relation betwee e obtaineth n d poise heigh r correspondino t g channel

number versus gamm e establisheda o energt s ha ye samTh e. sources used

for efficiency calibration can be used to calibrate this relation.

Theoretically this relation a shoullinea e b dr one. Howeveo t e du r

imperfection e pulsth n e i samplificatio n and/or e analoguth ) e digital

converter (ADO this relation is not always perfectly linear. It is

therefore better to carry out a polynomial bit and check the fitting at

regular time intervals (for instance befor a e seriestarth ef f o o st 24 1 measurements) by means of a source of low and high energy, e.g. Am

and Co. A typical polynomial fit is given by equation (6.12)

68 2 3 . . . . k a. + k a„ + k a, + E(keV a = ) (6.12) J i. l o

= V energE ke n i y

= channek l number

a proportionality constants n Typical values for a were found to be

a = - 1.4426; o = , 0.5015a 9

- 0.13315= 0 .1

0.408148.10" -13 - 0.43491 = a 0 1.1

6.4.3. Evaluation of the gamma spectra.

Erdtmann and Petri (20) give a survey of the different steps in

the evaluatio f gammo n y spectrra a s showa a n Figi n . 6.6.

Evaluatio Gamma-raf no y Spectra

Steps of the procedure Data required

Searc peake th r shfo i Determinatio f peano k situations Parameter l energso y calibration curve i Calculatio f gamma-rano y energies Gamma-ray energies found i Parameters of efficiency calibration curve Calculation of peak areas Measuring time I Calculatio f gamma-rano y emission rates Nudide data (Gamma-ray energie abundancesd san . i / half-lives, production modes) Identification of the radionuclides I Decay time Calculation of radionudide decay rates ^ Irradiation time ^-^' __ Neutron flux and cross sections or ^^^^ Results from reference samples Calculatio amountf no elementf so ' sV Fig. 6.6 : Steps for the evaluation of -y-ray spectra and the calcula-

tio f activatioo n n analyses. 69 Commercial M.C.A.'s usually provide Programms for most of the steps

given in Fig. 6.6. In activation analysis with an I.N.S. the spectra

are not as complex as those encountered in reactor activation analysis

and a manual evaluation of the spectra is not too much time consuming.

The peak e omittesearce b peak th e intensn ar f ca shi d e enouge b o t h

identified visually. Calculations of the peak area is best carried out

n integratioba y n procedur s explainea e d from Fig. 6.7. (I.e.)

CD C C « o

ö o o

B, ^1 *-2 W3 ~4 Channel number (c) Fig. 6.7 : Calculation of peak area by integration The peak area is

e use ar o calculatt d B e backgroun th d , an B e B , d F under p r 1 p the peak .

The net peak area F is the area between channels C„ and C_ above the p 23 dashed line and is obtained by summing the counts in channels C + 1 e backgrounTh . -1 s assumee C i lineadb o d t correspondo an t rp d u o t s

the trapezoid B below the dashed line. The height of the left and P

7C right borders of the trapezoid are calculated from the mean over sever-

al adjacent channels.

Thus :

1 B B [W - 4 C c 1 [C1 2 -V

where '2 ; n E = 8 d an n C ' C3 C

e standarTh d deviatio e peath ks n give i o nare y b F na

s 2 B. B n - c r- c ;' 1 r 3 2 ———————— + ———————— 2 2 c2^, •« • (l rW r 1) 1 \ (t rVS- r ^ n\ J 2

71 7. SAMPLE PREPARATION

As already described earlier a annular source or an array of

sources is to be preferred to a single source, namely to enhance the

neutron flux and to reduce flux gradients The type of I N S arrange-

ment limit f courso s e samplth e e size. Thi s obviousli s y alse casth oe

if a pneumatic transport system is installed. In any case it seems

desirable that all the dimensions are chosen in such a way that samples

a diamete f m o diametem e b f irradiateappo 0 2 rn r ca rd thaan d t

source, pneumatic tubd samplan e e holder (rabbit f correspondino e ar ) g

size s alwayA s n neutroi s n activation analysi s gooi t d i s practico t e

use a standard which closely approximates the sample In the case of

powdered samples this can be achieved by mixing carefully the different

component s theia s r oxide r componento s f o wels l known stoechiometry.

As flux gradient d neutroan s n absorption phenomen e importanb n ca ao tw t

alternatives are possible to avoid systematic errors :

1. The powders are homogenized after irradiation e powderTh . e presse2 ar s d into cylindrical pellets under additioa f o n

binding component. This can be a wax or graphite. A proportion of

about 2:1 sample and graphite or wax are sufficient to obtain

mechanically resistan n diametei 0 t1 m m pellet d 0 2 an a rsiz f f o eo s 2 mm height at a pressure of about 400 kg/cm . A manual or hydrolic

e s purposehoweveth ha e use e r b preso t On fo rkeen d . n minca si p d

that it is important to obtain pellets of identical size as this

affects the irradiation as well as the measuring geometry.

Obviously the second alternative is the best one as it allows quite

large samples to be reproduced accurately. If a press is not available

the powder, witr withouo h t bindin ge b presseagen n ca t d e intth o

72 sample holder and spacer can be used to fill up the container

completely so that the sample cannot move during transport.

If the sample is not available as a powder e.g. a metal, it can be machined precisel e desireth o t yd size.

73 8. APPLICATIONS

General Remarks.

s assumeIi t d tha n annulaa t r neutron sourc s availabli e e yielding fi — ? a neutron d fluthaan f approximatelo xts . e sampleb m c n . n ca s0 1 y

irradiated with a diameter of 20 mm. A different neutron fluxes and

(or) sample size will obviously increase resp. decreas e induceth e d

activity. Measurement are carried out with a Nal(Tl) detector of 7.5 x

7.5 cm with a well of 3.3 cm deep and A.3 cm diameter. Other sizes of

detectors wilf o coursl e influenc e detectioth e n efficienc d counan y t

rate. This type of detector can be coupled to a single channel or a

multichannel analyzer s A anothe. r alternativ a germaniue m semi

conductor, preferably hyperpure. can be used, with an efficiency of

about 20 I and a resolution of 2 keV or better, coupled to a 4000

channel analyser.

As already stated earlie e ratheth r r modes ta I.N.S fluf o x. does

only exceptionally allow the determination of trace elements. There-

fore the main applications are in the field of minor or major

components where good precision and accuracy can be obtained. To

insure a reproducible geometry during irradiation and counting,

powdered material, e.g. biological, ores, metals, etc. can be mixed

2 proportio1: wit r o h 1 d pressepowdere1: an n a n di d x graphitwa r o e

into a cylindrical pellet.

6.1. Determinatio f silveo n n leadi r .

8.1.1. Introduction.

Silver presen n orei t s oftei s n concentrate n lea i dy heatin b d n i g

74 the presence of lead oxide and a reductant. This gives rize to metallic

lea n whici de silveth h s concentratedi r e determinatioTh . s usualli n y

carriey cupellationb t ou d r stream ai y ,heatine b lea n i.eTh a .d . n i g

is transformed into oxide and taken up by the cuppel, leaving a

metallic silver bead e weighedb whic n ca h .

8.1.2. Nuclear data.

Irradiatio f silveo n r gives ris o t ethre e radioactive species a s

shown in Table 8.1.

Nuclea: TABL 1 8. E r Dat f Silvero a .

Stable Abundance (n, r) o I T... E (keV) + y 2 1/ o o Isotope product (b) (b) abundance (l)

107 51.8 8 310 37.3 96.0 2.37m 433{ 9. 0.50Z)

618. ( 9 0.26Z)

633. 0 (1. 76Z)

109 8 48.18 140 0 611 0 8 24.665 s (4.5Z)

110m 3.90 69.0 d 8 657249. .8 (94 .5Z)

A number of other element present in lead e.g. copper, zinc, antimony

and tin do not interfere as their concentration is usually very low. b result S n Ia apparenr n o n i su practic C f to n amoun a m e pp f 100o t0

silver content of only 5.2 and 2.3 ppm, whereas zinc and tin are

completely negligibl t theia e r normal level.

75 8.1.3. Equipment.

Well type sodiu mm wit c iodid 5 h 7. Multichanneex détectm c 5 7. a l

analyser, well of 3.5 cm x 4.3 cm.

8.I.A. Sampl d standaran e d preparation.

m diametem Cylindrica 1 0. + r m 0 heighm 2 1 ld 0. an t sample + 5 f o s

are machined from available lead ingots. This correspond a weigh o t s t

f 17,o 5 gram. Standard e preparear s y smeltinb d a quart n i g z crucible

high purity silver foil (99.9 Z) cut into fine bits with lead grains.

l smeltingAl a hydroge e n carriear i st ou n datmospher o avoie t e th d

formation of lead dioxide. After heating for 30 minutes under a

hydrogen stream on a bunsen burner at appr. 550 "C the liquid smelt is

pourea graphit n i d e mould preheate a temperaturd e just above th e

smelting poin f leao t d (327 *C) e graphit.Th e moul s immersei d n i d

cold water to avoid silver concentration gradients. Standards

containing between 250 and 4000 ppm can be prepared in this way.

8.1.5. Irradiation and counting conditions.

e leaTh d sample r standardo s e introducear s a suitabl n i d e poly-

ethylene containee irradiatear d an rd counte dan durin, s d 0 6 during g e spectruTh . s aftes f Figo m 0 0 a deca6 1 1 showr. 8. f n o intensya s e

Pb-X-ray fluorescence th f o a photopea em pea8 su d keV 65 an e k t .Th a k

count s takei s s a nshow n Figi n . e give8.1th r n .Fo condition t ne a s

activit f abouo y t 5000 count f r r silve100o minutpe pe sm s d 0i pp ran e

obtained with a background of about 1000 counts per minute. For silver

concentration at the 1000 ppm level the precision is approximately 1 Ï.

16 3000

0) C Pb X-Ray mC .c 72 keV o

c o CJ 2000

110 Backscatter 1000

50 100 150 200 Channel number

Fig. 8.1 : Spectrum of Ag.

8.2. Determination of manganese in pyrolusite and ferromanganese.

8.2.1. Introduction.

Pyrolusit n importana manganeseZ s i e0 e containin5 or o tt Ï . 0 3 g

It is used as such in the iron and steel industry or is transformed into ferromanganese wit a manganesh. Manganes I 0 8 o t e e I conten 5 7 f o t determinatio n i nthes e materials requir a hige h precisiow fe a f o n

tenta percent f o h A titrimetri. c procedur s usualli e y e applieth r fo d purpose namely the pyrofosfate or bismuthate methods. Typical pyrolusi- te and ferromanganese compositions are shown in Tables 8.2 and 8.3.

77 Chemica: TABL 2 8. E l compositio f somo n e pyrolusites.

O Si e F n M N A12°3 P2°5 Ca° Mg° S

(1) 48.46 5.36 2.88 5.86 0.237 0.00 0.45 0.009

(2) 5*. 20 9.71 5.93 0.235 0.119 3.61 0.42 0.021

1 . 12 4.8 9 .1 01 5 ) (3 0.37 0.13 2.90 0.40

(4) 37.68 16.5 1.75 0.2

TABLE 8.3 : Chemical composition of some ferromanganese.

i S C e F Mn

(1) 78.48 13.38 6.87 0.279 0.155 0.008

(2) 77.83 13.45 7.00 0.148 0.141 0.021

(3) 75.84 16.10 6.88 0.229 0.216 0.034

8.2.2. Nuclear Data.

The relevant nuclea e major th dat f ro acomponent f o spyrolusit e

and ferromanganes e givear en Tabl i n e 8.4.

78 TABLE 8.4 : Nuclear Data.

o l T... E (keV) + Stable Abundance (n.y) or o o 1/2 y ) Isotop(Z (n.pe 2 abundanc) e ) (b ) (b

product

55 56 Mn 100 Mn 13.3 14.0 2.5785h 846.8 (98.9Z) 27 28 AI 100 Al 0.226 0.16 2.24 m 0 1778.9 (100Z)

1810.7 (27.2Z)

2113.1 (14.3Z) 49 Ca 0.0019 Ca 1.12 0.50 8.719 m 92.1 (0.8Z)

3°Si 3.10 31Si 0.107 0.710 2.62 h 2 1266.2 (0.07Z) 26 27 Mg 11.01 Mg 0.0372 0.024 9.458 m 170.7 (0.84Z) 51. 52. 99.75 4.79 2.63 3.75 m 1434.0 (100Z) 56 56 Fe 91 .7 Mn 0.0009 28 28 Si 92.21 Al 0.02

8.2.3. Interferences.

From the data of table 8.4 it is possible to calculate the induced

activities for instance in pyrolusite 2. The activation by epithermal

neutrons was neglected, considering the high thermal to epithermal flux

ratio, wherea a stherma d fasan l t 0 s assumed1 n.cfluwa f o x s m. .

Th eg sampl 1 result e a represente ar er fo s n Figi d . 8.2.

It appears that the most important interference is due to 28.Al. at the

end of the irradiation. The interference due to Si is completely

negligible as the gamma intensity is only 0.07 Z. The other elements

P als d o an induc Cag M , e negligible activitie f gammi s a spectrometrs i y

applied wit a Nal(Tlh ) detector. Some pyrolusites contain tracef o s

vanadium. Even at a 0.1 Z level the interference due to the 51V(n.y)

79 10- 56, Mn(from n,y)

10 to 4J C o Ü

IQ- U Ü

28 Al

31 10 Si 48 Ça 56 Mn(from n,p) 27 Mg -1 10

40 80 120 160 Irradiation cime (min)

: Calculate 2 Fig8. . d activit n pyrolusiti y e 0970/67a s a 7

function of irradiation time (A - t ). thermal fast

52 V reaction is entirely negligible. Some pyrolusites show a small

natural radioactivity mainly due to the presence of members of the

? *\ ? ? T ft U series r To .h This activit n reacca yn integraa h l count rate

f abou o r 0 count100pe 25 tr 0d gra pe ssecondsan m t introduceI . a s

positive error of only 0.08 I. For the analysis parameters given

elements with large absorption cross-sections (e.g. B, Cd, rare earths)

presen n somi t e pyrolusites r instancfo , n thosi B ee from South African

origin, can cause additional neutron shadowing effects in large sam- e samplth plesf I e .siz s limiteg i (see 3 eo t furthed r sample size)

80 this effec s alsi t o negligible e activitTh . y inducee thresholth y b d d

reaction Fe(n.p s alsi n oM s )negligibli fiv et i orders a ef o s

c c c c magnitude lower than the one induced by the Mn(n,r) Mn reaction even

whee Mn-Fe-ratith n s 5:1i o .

8.2.4. Neutro d gamma-raan n y attenuation effects.

e largth Duo et e sampl r pyrolusitfo e g siz 3 ( e eg wachs wit7 1. h ; g ferromanganes 8 9. d 0.9 an ee larg 6g wachsth d e an )therma l neutron

cross-section of manganese, considerable self-shadowing effects do

occur as well as gamma-ray attenuation. If samples with constant

amounts of pyrolusite or ferromanganese are irradiated and measured the

resulting coun n i fact rats ti e influence y bote b neutrod th h d an n gamma attenuation effects. Moreover flux attenuatio n occuca nr within

the individual particles. To avoid this effect the particle size

should be kept smaller than 40 tim.

e measurinIth f g condition e kepar st constant (see further 8.2.5)

as wel s sampla l e d sizsamplan e en establisca densitt e ne on a hy

neutron attenuation effect as shown in Fig. 8.3. The specific

activitie e obtainear s d from samples with known manganese contente On .

can conclude thae calibratioth t n line a e manganeslineaar s r fo r e

content between 1 g and 2,5 g.

8.2.5. Measuring and irradiation conditions.

As already mentioned abov a wele l type Nal(Tl) detecto s i ruse d

couple a singl o t d e channe r multichanneo l l analyser e formeth f s I i r .

o includt t usee e bacbasse th th ede ke b scattelinn ca e r peak, whereas

the upper limit should include the 2113 keV line. If a MCA is used

channel summation shoul e applieb d o includt d e samth e e energy region.

81 160 1 5 x _ -3

o 0) 01 Oo o y=-(10847+340)x +(174509+590) 160

155 y=-(11033+280)x +(174202+530) .

150 O ferromanganese

• pyrolusite

J_ _L 1.0 2.0 Gram Mn Fig. 8.3 Net-neutronattenuation curve r pyrolusitfo s d an e

ferromanganese.

fl If a source of 10 n.s is availabl e followinth e g sequencn ca e

be applie: d

Irradiation time : 200 s

s 0 Deca80 : y time

Measuring time : 800 s

82 6 This results in a total count of appr. 10 for samples with the

following compositions :

3 g pyrolusite + 1.7 g wachs

9 g ferromanganese + 1 g wachs

After mixing carefully they are pressed into pellets of 20 mm diameter 2 m heighm a an pressur7 d0 f t othe kg/cI a t*0 . f rmo e size e usear s d

the calibration curves will obviously be different.

8.2.6. Calculations.

From a calibration line as shown in Fig. 8.3 the manganese content

cae calculateb n y successivb d e approximatio: n

Mn , x A A

""... --TT—-rr1 . st st,1 :

Mn = weight manganese in sample, first approximation

= weighMn t manganes n standari e d

A = number of counts in sample S = numbe f Acounto r n standari s d

S = specific activity of standard

e e b correcteneutroth o t r s fo nha d . attenuatio, ^ S n difference st, 1

n M d an t . betwees ^ n M n 1 , st

R x ) . n (8.2M - . )n (M + . . S = , . S st, 2 st.1 st,1 st

e self-shieldine slopth th f o es i wher R e g calibration lin f o eFig .

8.3. Usuall e seconth y d approximatio e thirth f d o s withiwiti Z n 1 h 0. n 8 -1 a source strength of 10 n.s and the experimental conditions given. A

precision of better than 0.1 1 can be achieved for a triplicate

analysis.

83 8.3. Determinatio f siliciuo n d aluminiuan m n bauxitei m .

8.3.1. Introduction.

n importanBauxita s i w materiaee aluminiura tth r fo l m industry

Henc a rapied accurat an d e analysi f greao s i ts importance th n i e

bauxite exploitatio e aluminiun th sit n s i a wele s a ml industry itself.

The most precise analytical method is still the gravimetric one but is

very time consuming. Instrumental methods as X-ray fluorescence,

atomic absorption, atomic emission a.o. are also time consuming and

often laborious sample preparation and (or) dissolution is required.

Through neutron activation analysis wit n I.N.Sa h . fast instrumental

analysis can be achieved within less than 15 minutes with a precision

of better than 1 1. .

8.3.2. Nuclear data.

o p 7 0 Aluminiue b determine n ca m d throug e reactioth h n Al(n,rl A )

with thermal neutrons, whereas siliciu s determinei m e reactioth a vi dn

p n p D Si(n,p l A wit) h fast neutrons .e achieveb n Thica sd througa h

double irradiation techniqu r instancfo e e d withouwitan h a cadmiut m

filter. Typical compositions of bauxite are given in Table 8 1.

From the composition given in Table 8.5 it appears that a number of

. P d an V , Ti , correctio Mn e presenc, th e requireFe b o t f y o e ma ne du d

Pertinent nuclear data are given in Table 8 2.

C R C R r c e correctioe Th n reactioninduceth M y b r d fo ns d Fe(n,pan n M )

Mn(n y} vary between 0.1 to 1.0 I. depending of course on the Fe and

e correctea measuremen b n M y n contentb ca r d fo dan st afte e decath r y

p p p / c c e presencTh a N interfere . f o Al e s f o howeven M r e witth h

84 Z 0 CT> O in co r^ 0, 1 CO fO m OO CN — O 0 0 X • . . • 1 1 1 II 1 pa ^ r^ i — CN CN o 0 0 0 0 in *~ CNl vO r- O O O O in *— ^o *_ cr, CNl f~^ O T— X 00 • . « 11-11 1 1 1 1 1 pa

vD CN co oo o — i 0 CJ* ,— 0 CN oo cn o o X

o CO OO cr\ m OO CM

o oo CN o CN X r- ^j- CO 0 pa CM . . • 1 1 1 1 1 1 1 1 1 1 v£> 0

CTi in r^ r^ m i^ O m X co vO -a- -a- CM

_ ^ CN u-l O CN vD (^ v_ tn cr. CO CO O X CN . » 1 1 • 1 1 1 1 1 1 1 pa CN 0 oo CN ^3- CN O O CM "* *~

^ 0 CO 0 vD in vD LO n r^ CO 0 X CT> • • * 11-11 1 1 1 1 1 ^J- ro o-i CNl CN CN O m CM

tn o m CN o % X co CO i^^. ^j- MD CTi ^o O 0) pa — . • • • 11-11 1 1 1 1 1 CN \JT) « — v£5 CN o O m i CN

CN m -T r-^ m CN in ^J- x — - • • • O "Z. M3 U"l ^o o^ m CN o O O O O O o o o o o U"l CN i- v v v o o

1 U) 1/1 co u) !-i 0 OJ •— 1 oo f~\ G CO CO O CsJ •H C O o CN n i n i CM o W CN o o o CN O o O O O CM O O O CN O co •J — , i •i-4 CO C -H QU J N f .,-4 M C n U MC« CN O pa < w Ma p iJ ) Inc_ > H u c o : N & K z: s:

85 measurement so that a correction is necessary, estimated from the

approximate aluminium content.

Typical decay curves are shown in Fig. 8.4.

28 10 - Al (2.24 rain) pos 1 28 Al (2.24 min) pos 2

27Mg (9.45 min) u CD 24Na (15.02 h) CO O O

0 O O 10

10

0 15 0 12 0 9 0 6 0 3 time (min)

Fig. 8.* : Decay curves of pure Al (99.99 Z) in position 1(«)

ann positioi d n 2(A).

86 These decay curves were obtained by irradiating at position 1 and

2 correspondin w fas lo o thermat a higtd o an ht g l neutron flu s showa x n

in Fig. 8.5.

12

l •r- O. •e. 0>

I en a •e-

Û 2468024

Height from bottom of irradiation place a s a , . .. 0 ,/ . Rati: 0. Fig5 d 0 o8. an . j./*,., . —-——— fast thermal thermal epithermal •function of height from the bottom of irradiation

place.

The interference due to Ti is negligible for a Ti-concentration

of _+ 3 I in bauxite. The interference due to the reaction 31 Pin,a)2 8 Al

is also negligible for fosfors concentrations below 0.2 I.

As show n Tabli 6 nvanadiu 8. e n occuca mn concentrationi r p u f o s 52 V activit e s th shori y s tA o 0.1livet . 21 d (3.7 ) wit5m a gammah - o o ray at H34 keV it can not be separated from Al. Only the knowledge

87 TABLE 8.6 : Nuclear data for neutron activation of bauxite.

Reaction Isotopic Cross section Half-life ^-energy

abundance 1 (mb) (keV)

28SKn,p)28Al 92.21 2.0 2.24m 0 1778.9

29 29 Si(n,p) Al 4.7 0.6 6.52 m 1273.3

3 27 °Si(n,a> Mg 3.09 0. 1 m 8 9.45 843.8; 1014.4

2?Al(n.y)28Al 100 0.231.10' 2.24m 0 1778.9

27Al(n.p)27Mg 100 4.3 9.45 m 843.8; 1014.4

AKn.aa N ) 100 0.65 14.95h 9 1368.6;2754 .0

56Fe(n.p)56Mn 91 .7 0.9 2.5785h 846.8; 1810.7;

2113 1

56Mn(n,r)56Mn 100 13 .3. 10J 2.5785h 846. 8; 1810;

2113 1

5°Ti(n.y)51Ti 5.4 0.171 5.752 m 320. 1 ;928.6

51V(n,r)52V 99.75 4.88.10' 3.75 m 1434 .0

31P(n,«)28Al 100 1 .3 2.24m 0 1778.9

of the approximate vanadium concentration allows for a correction. In

the given irradiatio d countinan n g conditions (see further t amounti ) s

for aluminium oxide to a positive error of appr. 1 I absolute for 0.1 '/

vanadium pentoxide and to a negative error on silica of appr. 0.6 I.

One should also note that certain bauxites contain a non negligible

amoun f naturao t l n radioactivit(thoriu4 4 e th + o n mt 4 232e d du yan )

series (uranium 238) n som.I e cases this amount o t 120s 0 countr pe s

300 seconds and causes a positive error on alumina of 1.4 1. A measure

e samploth f e before activation allows obviousl o t correcy r thifo t s

effect.

88 8.3.3. Sample preparation, irradiation and measurement.

Finely grinded bauxite (particle sizs mixei e) d am belowit 0 4 hw

wachs (ca. 4.5 g bauxite and 0.9 g wachs) and pressed into a pellet of 2 8.5 mm height and 20 mm diameter at a pressure of 400 kg/cm . The

pelle s measure i te samth n ei d condition s a aftes r irradiatioo t n

determin e naturath e l radioactivity e pelleOn . s firsi t t irradiaten i d

position 1 (optimal fast flux) and one in position 2 (optimal thermal

flux)t feasibl no f I this .i e irradiateson e a cadmiu n i sx (walbo m l

thicknes ) insteamm 1 f spositio o de irradiatio Th . 1 n s 0 n 20 tim s i e

A 4 for a source strength of appr. 10 n.s . After a waiting time of 40s a

first measuremen s e detectoi madt Th ea wel. s s durinli r 0 typ30 ge

crysta s a lalread y described. Integral countin s i applieg d wita h

a S.C.Athreshol t a . d lower e thabackscatteth n r peak A .secon d

counting of 300 s is taken after a 100 second decay i.e. after complete

no o f decay of Al and Mg. A possible scheme of analysis of a triplicate

sampl s showi e n Figi n . 8.6.

e aluminTh d silican a a contene calculateb n ca t s followa d : s

, W b + . w a = . , A (8.3) A 1 , net si Al

A . = c w . + b w , (8.4) l A i s t 2,ne 4

where

(A - BG) - (A - BG) x k (8.5) ,net w.b

(A - BG) - (A - BG) x k —:————————= . , A -————— (8.6) 2,net w.

89 transpor f samplo t detectoo t e r measuremens 0 30 t (decay = 40 s) n correctioM for n

200 s irradiation measuremens 0 30 t ,-WJ o. D D D D D , 28 . 27 D decay of AiA 1and MMg D D D ,D

40 80 120 160 200 240 Time (min)

>•' i 8 •_ 8. b : Analysis scheme for simultaneous determination of Al.,0- and SiO„ in three bauxite ores wit 6 standardizeh d pellet eacr fo sh ore. A ;A = total count of 1st measurement in position 1 resp. 2

A' ;A' = total count of 2nd measurement in position 1 resp. 2

BG = background including natural radioactivity

k = correction factor for decay of Mn between 1st and 2nd

measurement i.e.

, .0.693 x 108.

a.c = silica coefficients i.e. specific activity due to silica

after irradiation in position 1 resp. 2

b,d = alumina coefficients i.e. specific activity due to alumina

after irradiatio n positioi 2 n d an 1 n

w = Sl°9 weight per gram bauxite O l C w., = Al„0^ weight per gram bauxite Al 2 3 = bauxite weighw n samplei t ,

By solving equation (8.3) and (8.*) one obtains a first approxima-

, fro4 e resultw th d m an f equationo s . w tiof so n(8.5 l d A (8.6)an ) S.i These results do not take into account the activity of 24 Na in the 56Mn

measurement. It can be calculated as follows

(A - BG)-[(A - BGJ-tA, x w..)] x k

Wb

(A - BG)-[(A - BG)-(A x w )] x k 2 ———————J - = —" ?A ——————2, ne —————t£ — — -—— (8.8) Wb

where 24 A" and A" = Na activity during the second measurement in

irradiation position 1 and 2 as measured from dedecay curve from pure

Al 0 pellet irradiated and measured in the same conditions as a

sample .

91 e irradiatioth Value r fo d smeasurin an n g conditions sa give e ar n

follows :

a = 35.710 b = 49.479

c = 6.092 d = 41.994

1 A ^ = 110 A"2 = 25

It is apparent from this data that the 24 Na correction is quite small e neglecteb n an ca dr man fo d y practical purposes.

It appears that wite methoth h d described abov n absoluta e e precision

of better than 0.5 I can be achieved for both alumina and silica.

8.4. Determination of silicium and aluminium in aluminium-silicium

alloys.

8.4.1. Introduction.

The silicium content of silicium-aluminium alloys varies between

0.5 Z and 13 Z. The composition of typical alloys is given in Table

8.7.

Concentratio: TABL 7 8. E f elemento ) (% n s presen Si-Al-alloyn i t s

Si Fe Cu Mn Mg Zn Ti Cr Ni V Ca

13.4 0.91 0. 10 0.57 0.75 0. 156 0.60 0.53 0.12 13.0 0.15 0.01 0.015 0.01 0.05 0.02 10.8 0.35 0.02 0.35 0.006 0.07 0.01 10.0 0.29 0.035 0.22 0.22 0.054 0.007 0.009 0.035 7.95 0.14 0.010 0.054 0.45 0,097 0.15 6.3 0.67 0. 11 0.65 0.80 0.125 0.10 0.07 0.045 4.65 0.42 0.30 0.39 0.44 0.032 0.025 0.975 0.23 0.005 0.31 0.80 0.091 0.99 0.62 4.91 1.25 1.64 0.096 0.094 0.094 0.10 0.50 0.54 0.040 0.038 0.033 0.079 0.048 0.036 0 040 0.046

92 As described further a single irradiation of the alloy under cadmium

cover allow a srapi d determinatio f siliciuo n m wit a precisioh n better

than 0.1 Z absolute.

8.4.2. Principle.

From Table 8.3 it appears that by using an irradiation under

cadmium and a discriminator setting at 1400 keV interferences in the 29 29 silicium determination due to the reactions Si(n,p) Al,

Si(n.a) Mg and Al(n,p) Mg are eliminated whereas those due to

97 9 Z. Rfi RC *} L. Al(n.a) Na and Fe(n.p) Mn are greatly reduced. The residual Na

activity can be taken into account, whereas the Mn-activity is

completely negligible, due to the small iron and manganese content of

the alloys.

8.4.3. Standardizatio d calculationan n .

e specifiTh c activitie f o saluminiu d siliciuan m e determinear m d

by irradiating pure silicium and aluminium having the same dimensions e sampleath s s i.em diameterm .0 m heigh2 m cylinder d s 6 A an .t f o s

already mentioned they are irradiated in a cadmium box with wall

thickness of 1 mm. For an irradiation at the site of maximal fast flux

of 200s and measured during 300s after a decay of 40s in the well-

type Nal(Tl) detector a specific activity of approx. 21.000 counts is

obtaine r grape md siliciu d appr 0 an m count80 . r grape s m aluminium 24 including about 90 counts Na.

The silicium content is now calculated by successive approxima- tion :

93 (8-9)

where

: specifiSA c activit f siliciuo ye sampleth n A i mS , beine th g 28 24 first approximation a ,actiN i.e -d .an l includinA e th g

vity

: specifiSA c activit f siliciuo ye standardth n i m .

The first apprimation for aluminium is given by

Si.- C° A1.10 ,* 1 (8'10)

By comparing the specific activity of the aluminium standard one

obtains a first approximation of the activity of the aluminium in the

sample :

s,2 s.- i A • -TOII T - SA = SA SA (8 11)

where

SA : specific activity of the pure aluminium standard Al

The new value of SA can be introduced in equation (8.9) and the j , t sequence repeated until

) A S - ISA ——iiü————— S|< 0.00""1 1 (8.12) SA s.n

This is the case for n = 4.

In the above procedure it is assumed that the sum of aluminium and e casth et equatio no f thi I s siliciui s. Ï n0 10 s (8.10equao i m t o t s li )

be replaced by

94 . C_ - = (10C,) D _ - 0 (8.13) Al.n Si.n

where D is the sum of all the elements present except Al and Si.

e abovTh e procedure give n accuraca s f o ybette Z r 2 tha0. n

absolute compared to a chemical determination of Si. It is however

obvious that the relative precision will be dependend on the Si content e alloyth f o . Thi s illustratei s n Figi d . 8.7.

c o c o oc • f-l O •J o> (0 S-i d a. •H

>J 01 4-1 0) n! i—i 0)

0

: Experimenta 7 Fig8. . l precisio e siliciuth n o n m determination i n aluminium-silicium.

8.5. Determinatio n cassiteritei n ti f o n .

8.5.1. Introduction.

Cassiterite is by far the most important tin ore (SnO ) occuring

in granitic rocks or after erosion as a sandy ore. Typical compositions

are given in Table. 8.8.

The chemical standardization of tin in cassitirite is a rather

difficult problem. A gravimetric procedure can be used by precipitating

tin as SnO from a nitric acid solution. The coprecipitated impurities

are determine y differencb d e afte. I r Sn removin s a n ti g

95 ^ 00 O CM ~3-00--— cMOu"!*— OO'— Pd D 4-^ 1 " r^ - r^ ooocMoco

CM O"! CO •— — • O O O r-» CM OO vO

CM O v r*^« O f o m CM W r-. ^•^^•Illlll! <}• ^— o o o o o o

\D »- en ^^ CO m in 0 W t-» 1 1 1 1 1 1 t ! 1 1 1 1 vO CM *— o

m oo i o~ r^ o c oo M C O pa co CM O") O ^O •^ C^ CO W r- ... 1 1 1 1 1 1 1 1 vO CM o CM m T- o o o

^ ^ OO vO < 0 O O> o W VÛ 1 1 1 1 1 1 1 1 1 1 1 1 xO vD O o

to s 01 T— r^ 0 * ^- V-i «U o 0 t- 0 o fjj vu • • 1 1 1 1 1 1 1 1 1 1 1 1 \O in o 0 HI r-^ 4J V 1 1- • i-i 0) t— - v r < r < 4J O C7> f"^ o r •r-l < 0 cr. oo CM CM o m 00 . .. fr"1 ) ^£ >1 1 1 1 l 1 1 1 (0 vD 00 CM *- 000 m ^O o i—i 03 Ü cn CM OO ^

Ow W LO 1 1 1 1 1 1 1 1 1 1 1 1 o CO CM 0 U

0 en oo »- vO CO oo t-O *«^" r**^ pa cr. ^O — • ^O M rC o r < - a U-l w m ... i i i i i i i i O **D -^- r-> T- O ^D •* O m .— G O •H 4J ^ o o D v n M C — • O ON WJ M M C o~> O ^ O \ O

, 00 • ^_l o c n i m 00 0) C/ ^

96 The other alternative is a iodemetric procedure by oxidizing

Sn(II o Sn(IVt ) ~ I0 )r o wit I h

8.5.2. Nuclear data.

The nuclear data of some elements present in cassiterites were already

given previously. Table 8.9. includes only those not given previously and

presen n cassiteritesi t .

. W d Nuclea: an TABL a 9 T 8. E. r Sn dat f o a

Isotope Occurence Isot.prod. I /o T y-energy (Z) o o 1/24/0

122Sn 4.63 123m Sn 5.40 40.08m 60.2(84Z); 332 . 0(97Z )

12Sn 5.79 125m Sn 60.1 9.525m 332.0(972)

181Ta 99.988 182m Ta 33.3 15.8m 17 1 . 6UOZ ) ; K6. 8 UOZ ) ; m.9(20Z) ;318.4(5Z)

186 M 28.6 187 W 13.7 23.9 h 1 34 . 2(9 . 5 ) ; 479 . 6(23 . 4 ) 685.7(29.2)

Froe compositioth m f cassiteriteo n d froe nucleaan sth m r datt i a

appears that activation analysis wit a I.N.Sh s possibli . f i eepi -

thermal activation is applied, combined with gamma spectrometry with a

Nal(Tl) detector for the reaction Sn(n,-y) Sn on the gamma at 332

e activateth l keVAl . d isotopes havo n gamma'e s coincidin n energi g y

V linf ke 125mo e 2 Sn33 wit.e th hexcep e 318.th t4 lin f 182mo e Ta.

However even at a concentration of 11. the interference is only 0.2 I.

91 8.5.3. Sample preparation, irradiatio d measurementan n .

16.5 g of cassiterite is mixed with 3.3 g wax and pressed into a

cylindrical pelle f 17.o t m diametem 0 d 26.an r m heighm 0 a pressur t a t e 2 e 0 kg/cpelleTh 40 s + . i mirradiate tf o a cadmiu n i x d(wal bo m l

thickness 1 mm) during 1000s at a place of optimum epithermal flux.

Afte e sampla waitinrth , s s countei e 0 g r 10 120tim fo df o e. 0 s After

a second waiting time of 2000 s, a second count is started in the

negative counting mode e obtaineTh . d spectru s givei m n Figi n . 8.8.

40-5 ._ It appears that e 160.smala n half-lifS th V t ) 2lke mm pea 0 f 4 o ek

1 9 f\m s i well separate n pea S t a 332.k d t 0 peafro e ne keVth m ke areth , a

8000-

332.0 keV

ac) f. 6000 u en ue. o u : 4000

123m ' Scn 1 160. V ke 2 ! JU / 2000 A V^V^vV ^~ ———— ^ ——— «J L 1 . r 0 30 0 020 100

Channel Number

t Nal(TlNe : Fig8 ) 8. spectru. a cassiterit f o m e sample.

98 being appr. 2.10 counts. This are s calculatei a d accordine th o t g

Wasson metho: d

-k

A = E n. - (k+1/2) . u

s illustratei t I n Figi d . 8.9.

k=0

8000

u g 6000 o

4000

2000

0 -2 0 -4 0 -6 20 60

Channel Number (k) : Pea Fig9 k8. . surface determination according Wasson.

99 8.5.4. Standardization.

It appears from Tabl . tha8 e concentratioeth t n cassitentei n ti f o n s

varies between 50 1 and 76 Z, whereas a numer of other elements as Fe.

Mn, Ti, Nb and W also vary between considerable limits. It can be shown

by a Monte Carlo method of calculation that although the absorption of

the 332.V gamme differenc ke s 0th i quit e a) th 2 e0 r 3 fo ehig+ ( h

different compositions is below 0.15 I and can thus be neglected. Since

most cassiterites contai , nanalytica O morSn e T tha 0 l9 n gradO Sn e

can be used as a standard, by pressing pellets in the same way as de-

scribed for the samples. For standards with a tin content between 55 1

and 79 Z a linear calibration line is obtained. The error on a single

analysi e estimateb n ca s. I t aboua d5 0. t

8.6. Determinatio f fluorino n n fluoritei e d oresan s .

8.6.1. Introduction

Althoug e availabilitth h f o specifiy c fluoride electrodes ha s

greatly increase e reliabilitth d f o fluoriny e determinations this

technique requires a dissolution of the sample and can give rize to

losses and systematic errors. Activation analysis is therefore an

interesting alternative as it is non destructive, rapid and precise.

8.6.2. Nuclear data.

1 9 Fluorin a mon s i eo isotopi n givca ed crizan elemenF e wit s a th

thermal, epitherma d fasan l t neutro a numbe o t nf nuclea o r r reactions

give n Tabli n e 8.10.

100 1 9 TABLE 8.10: Nuclear data for neutron reactions with F and interfering

reactions.

Reaction Thermal Resonance Fast Threshold Half- Ma in Cross Integral Cross (MeV) life •y-energies section section (keV) ) (b (mb(b) )

-3 19F(n.r)2°F 2.6.10-3 21 . 10 11.41s 1633.1

19F(n.p,190 1 .35 4.25 28.91s 1974;1375.6

19F(n.2n)- 18F 7.3x10~3 10.98 6.59.101 351 s

19F(n,«)15N 7.85 1.60 7.14s 6128-.71117

160(n.p»16N 0.019 10.2 7.14 6128:7117

4 Q 4 f% 4 Q Methods based on the reaction F(n,2n) F are not specific as F is a

pure ß -emitter. Thus a chemical separation is required. Activation 9 1 0 2 with thermal neutrons in the presence of fast neutrons yields F, 0.

These isotopes cannot easily be measured instrumentally as the gamma's

are superimpose e Comptoth n o nd radiatio e higth h f o energn y onese Th .

reaction 1 Q F(n,a) 1 RN has a reasonably high cross-section. a low 1 6 threshol n isotopa d yieldan , N de wits h short half-lif d higan e h

energy gamma radiation. If measured with a Nal(Tl) with threshold set 1 6 e measurementh V e onlMe Th ay4 t . s specifiN i interferenct r fo c n ca e

be due to the presence of oxygen. The 0(n,p) N has however a very

high threshold energ t a y10. 2 MeV. Consequentl e activitth N y f o y

induced by the latter by means of an (a.n)-source is lower by a factor

of approximately 200. If fluorine is present as a major or minor

element this correctio s usualli n y negligible.

101 8.6.3. Irradiation, counting and calculation.

A powdered sample of fluorite of 14.5 g is mixed with 1.45 g of

wachs and pressed into a pellet of 2.00 cm diameter and 1.75 cm height.

It is irradiated for 20s at the site of maximal fast flux, transferred

into the well of the Nal(Tl) detector and measured during 20s in the

energy region of 4 MeV to 7.5 MeV. The number of counts collected is

approx. 5000 counts per gram of fluorine. Usually the CaF content is 7

to 92 I CaF and a correction for oxygen is not required. Teflon (with

a fluorine conten a e standarusef b 61.29o ts a n d ca )d fro a cylindem r

e samth e samples f th o siz s a e . Fluorin n als e ca eb determineo r fo d

instanc n zini e c ores where fluorinth e e content varies approximately

. 7- 1 d an betwee . 1 1 0. n

A spectrum obtained from the teflon standard (A) and a zinc ore

containin Ï fluorin 5 0. g s showi e n Figi n . 8.10.

The fluorine content is of course obtained from the ratio's of the

activitie n sampli s d tefloan e n standards.

Remarks :

e halth fs A lif . 1s quiti e e shor e t sampleremoveth no t e ar sd froe th m

irradiation container. A blank correction can therefore be required

instea a pur f o ed background.

2. For a low fluorine content and a high oxygen content the correction

for the 0(n,p) N reaction has to be applied. Therefore an oxalic

acid e irradiatepelleb n ca t o t determind e exacth e t correction.

Obviously the oxygen content has to be known or can be deduced from

the compositio e oreth . f o n

102 16.

600 -

1-1 01 o.

c: o CJ

400 ~

200 -

200 300 400 Channel number

Fig. 8.10 : Nal(Tl) spectru a 16. f mo 5 g teflo n sample (spectrua d an ) mA 14. g zine 5sampl or c e (spectru ; irradiatiomB) n; s tim 0 2 e cooling ; tims countin 4 e g . s Spectrutim 0 2 e s i scale mB p u d ba yfacto f teno r .

103 8.7. Determination of manganese in plants.

8.7.1. Introduction.

Manganes n essentiaa s i e l elemen r plantsfo t s functioit , n being

relate o t oxidation-reductiod n processes s i als t oI . par f somo t e

enzymes. Its content varies widely according to the species and

accordin e soi th n whico lo t gh thee grownar ye orde Th f magnitud.o r e

is usually between 10 and 5000 ppm. A number of SRM's of the National

Bureau of Standards, Washington DC, U.S.A. have been certified for

manganese as SRM/575 Pine Needles (675 jug/g Mn ) ; SRM/573 Tomato Leaves

(238 ng/g MnI ; SRM/572 Citrus Leaves 23 ng/g Mn); SRM/570 Spinach (165

Atg/g Mn). Certified Reference Materials are also available from the

Community Bureau of Reference (B.C.R.) : BCR No60 Aquatic Plant (1759

1 6 Aquati/ig/g o N Mn) cR BC Mos; s (377 2 Oliv6 1 o N

(57 /ig/g). These reference materials can of course be used to check the

accurac e methodth f o y.

8.7.2. Nuclear data.

As already give n Tabli n e 8.4. manganes a quit s ha e large thermal

neutron cross-sectio f 13.o n a favourabl3 bard an n e . half-lifh 6 2. f o e 56 It is a monoisotopic element as Mn-55. Mn is a -y-emitter with an

energy of 846.9 keV. Thus its nuclear properties are quite favourable

and good sensitivities can be expected. Plant materials contain

however large concentration of potassium and sodium and also trace

elements as bromine, zinc, copper, etc., so that a purely instrumental

metho y gamma-spectrometrb d y wit a germaniuh m semi-conductor cannoe b t

applied and a chemical separation is usually required. It is also

apparent that plant materials contain important amount f irono ss A .

104 a positiv o t n give eca nz e errot thi es ri I 91.abundanc e 7F f o e

through the reaction Fe(n.p) Mn (cross section 0.9 mb.) with the

fast neutron flux. Even whe a n fas o t thermat l flux ratio equao t l

unity is assumed and an iron to manganese ratio of 100. this error is

completely negligible.

8.7.3. Procedure.

Weigh 1g (or more if required) of dried and pulverised plant

material into an irradiation capsule and irradiate at the optimal

thermal flux site durin . h Transfe 2 g e irradiateth r d material inta o a mixtur f o l em d 0 2 nitrian d l ) m ad beakeN c d 0 4 an aci25 r(1 d

perchlori a 1:1 n 0i proportion ) Î c 0 aci(7 d . Heat tho gentl a n o y

plate. Add more nitric acid if required. Fume down to perchloric acid

fumes N perchlori 1 Leav 0. o t d cool edissolvan m l5 c 2 acid n i e .

Filter off any residue due to SiO or KC10 . Collect the filtrate into

a 250 ml separatory funnel. Add 10 ml of saturated hydrazine sulfate

solution to reduce the manganese to the bivalent state. After 5 mm.

add 20 ml of a 0.05 mol/liter potassium biphalate buffer and adjust to

pH=2 with 14N ammonia Extract the copper with 2 successive 20 ml

portion5 mol/lite0 0 f o s r oxin n chloroformi e . Thereafter adjuso t t

pH=8 with KN ammonia. Extract twice with 20 ml 0.05 mol/liter oxine

in chloroform Wase organith h c phase twice wit n aqueoua h s solution 24 of pH = 8 to remove traces of Na contaminating the organic phase.

Transfe e organith r c phase int a countino g viad placan l n o e3"x3 "

Nal(Tl) detector couple n M.C.a o A t d Determin t activitne e th e y under

the photopeak at 846 9 keV

The manganese conten s i calculatet de th e froratith f mo o

activities of a standard and the sample. As a standard a known amount

105 of manganese (e.g a .manganes e sulfate solution soaked into cellulose

puld driedan p s irradiatei ) e samth n ei d geometr d treatee an yth n i d

same way as the sample.

8.8. Determinatio f cadmiuo n y neutrob n zine i m or cn absorption.

8.8.1. Introduction.

Cadmium occur n mosi s t zinc ored zinan sc concentrate n conceni s -

trations varying between 0.1 I and 1 1. as shown in Table 8.11.

These ores and concentrates are an important source of cadmium.

The cadmium is usually determined by atomic absorption spectrometry

after dissolution of the zinc ore in sulfuric acid. Cadmium is usually

removed from these solutions by deposition on zinc powder, as the

forme s mori r e electro-positive Cadmiu a hig s hha m absorption cross-

section for thermal neutrons (2440 barn) so that a direct determination

on the ore is possible by the neutron absorption technique.

8.8.2. Nuclear

s alreadA y describe a f smali >, n i dl metallic vanadium cylinde6 ( r

m m heightm m diameter a measur e use e neutrob 6 s th a ,d n f o eca n)

absorption. The nuclear characteristics of vanadium are indeed quite

favourabl : shore t half-life (3.755 mm), large thermal neutron

cross-section (4.79 barn d largan ) V isotope abundance th e f o e 52 s inducei V o d tha(99.75s e witth t) 1 h0 good yields,a even i n

modest flux.

106 TABLE 8.11 : Concentration (Z) o-f the most important elements in zinc

ored concentratesan s .

Zn Zn Zn_ Zn 2 Zn Zni 3 4 D D

Zn 54.21 59.64 5 51.7 47.67 61.37 51.56 Cd 2 0. 0.1 3 0. 0.41 0.74 0.96

BCR1 BCR2 BCR3

Zn 53-55 46-48 51-53 Fe 7 15 95 Cd 0.1 0.15 1-1 . 1 S 31 29 30 Cl 0.01 0.06 0.005

sio2 3 0.3 0.8 Al„0„ 1 0.05 0. 1 2 3 CaO 0.3 0.4 0.3 HgO 0.05 0.05 0.2 0.4 0.5 0.2 Mn3°4 Pb 1 0.7 10 Cu 0.07 0.8 1 .5 As 0.05 0. 1 0.3 Co 0.015 - 0.005 BaO 0.05 - 0.001 Ag 0.007 0.012 0.015 Sn 0.001 0.001 0.001 Sb 0.005 0.005 0.01 Ni 0.001 0.001 0.001 Ge - - 0.02 F - 0.02 -

8.8.3. Sample and standard preparation.

s i mixe e d or wit e Th h1 ove wach a 2 proportior n i s d pressean n d

into two pellets of 10 mm thickness and 20 mm diameter corresponding

107 a weigheac o e uppef t ho Th 6.2 t e rpres. th g 5 sids shapei f so e o s d

m tham m diametedee m 6 a hol ts obtained3 i pf o ed an r A pressur. f o e 2 about 400 kg/cm is required. The two half samples are then fitted

with the vanadium detector.

Standards are prepared by mixing analytical grade ZnO, elemental

sulfur, carbonyl iron and cadmium oxide with wachs. This standards

(not including the wachs) contain 55 I Zn. 31 l S, 7 7. Fe and cadmium

. I 5 1. o t froI 0 m

8.8.* 2rradlatlon and measuring conditions.

e pelletTh s containin e vanadiuth g m detecto e irradiatear r e th t a d

optimal thermal flux site during 200 s. The vanadium detector is

removed froe pelleth m d counte" Nal(Tl3 an t x " d3 ) wite welth hl type

detecto e integrath n i r l mode (i.e. fro n energa m n correspondino y o t g 6 2-1 40 keV) during 400 s. At a thermal flux of appr. 10 n.cm .s the

52V-activity corresponds to appr. 900.000 counts. The same vanadium

detector can be re-used after a delay of 15 minutes.

8.8.5. Interference d calculatioan s e resultsth f o n .

As the neutron absorption technique is non-specific, other

"neutron poisons" will interfere. This could be the case especially

for boron as this element has a sensitivity which is 3.3. times higher

than cadmium o N boro. n could howeve e detecteb r n zini d c d orean s

concentratesZ 7 d . an MoreoveS n calculat Z ca 0 e 3 on r, eZn Ï tha 5 5 t

e givF e respectivel n a yabsorptio rizo t e n correspondin o 0.03t g . T 6

y variatioAn d 0.01an 0.02 . I Cd n concentratio2i n5 Z f theso n e

elements as occurring in ores and concentrates can only cause a maximal

erro f 0.0o r 1Z cadmium .

108 The results are calculated from the relation

log — = a + bx + ex (8.15)

where

52 V activit t n ne blani y= Ac o standard 52 A = net V activity in sample

x = mg Cd in sample.

Typical values for a. b and c in eq. (8.15) obtained by least fitting are shown in Fig. 8.11.

0.20 -

00 o

0.15 -

-0.249x10~= lo i g 2+ 28.7x10~ 485.87x10~- x 6x2 A 0.10 -

0.05 -

80 120 mg Cd

Fig. 8.11 : Thermal neutron absorption as a function of cadmium content

109 8.9. Determinatio f boroo n n steei n y neutrob l n absorption.

8.9.1. Introduction.

In steel containin I silico3 g n boro s sometimei n se addeth t a d

m leve10pp 0o t limprov e magnetith e c properties. Boro s usualli n y

determined by spectrophotometry after dissolution of the sample,

distillation of the boron as its trimethylester and addition of

curcumine to form a red complex. As boron has a large absorption

cross-section (76w atomilo 5a barncd weightan ) e sensitivitth , s i y

3.3 larger than for cadmium. Moreover the density of steel being higher

e ba achieveda sensitivitfactob y 7 n f ca o r B . f m abouo ypp 1 t

8.9.2. Procedure and results.

e procedurTh e describe8 on 8. s exactl e i e$ th n e sami dth s ya e

except thae sampleth t e machinear s o t obtaid n m m cylinder 0 2 f o s

0 iro1 diameten heightf o d an r, containin e centeth a m gm holn i 3 r e 52 deep and 6 mm diameter. It appears that the net V activity yields a

linear relation as a function of the boron content as shown in Fig.

8.12.

The precision is 1.5 wg/g at the 100 ppm level.

110 Fig. 8.12 : Thermal neutron absorption as a function of Boron content.

Ill 8.10. Other applications of INS

It is also evident that a I.N.S. can also be used to study the

chemical effects of neutron capture reactions, which is part of the so

called hot-atom chemistry. Szilar d Chalmeran d s already discoveren i d

1934 that if ethyliodide is irradiated with neutrons, part of the

radioiodin o longen s i r e presene ethyliodideth n i t . This procesf o s

chemical bond rupture due to recoil is therefore called the Szilard-

Chalmers process. It is perhaps usefull to mention briefly a number of

hot-atom applications where I.N.S. can applied either for research or

education. Reactions with organic halides can be studied, not only to

determine the free radiohalogens after neutron irradiation but also to

identify the numerous organic products formed. Of special interest can 0 8 m 80 be the separation of isomenc states e.g. Br from Br. The role of

organic scavengers can also be studied in this context.

Irradiations of oxycompounds are also of great importance in hot-

atom chemistry r instancfo s a , e perchlorates, periodates, chlorates,

bromates, iodates, permanganates etc t onl.no y The e ar interestiny g

from ,.ta hot-atom chemistry point of view, but also allow the prepara-

tions of radioactive isotopes of high specific activity. Thermal

annealing studief courso n ca se als e studiedb o . Finally self labelling

of organic compounds with tritium by the Li(n,a) H can also be

attempted.

Summarizin n concludca e on g e tha t a onlI.N.St no y s i .interestin g

as a tool for neutron activation analysis but can also find numerous

applications in hot-atom chemistry. The latter should however be dealt

with in a separate monography.

J12 REFERENCES

e BeecTreatis: d . p (1J O k ) n Analyticao e l Chemistry. Second Ed.. part I, Vol. H. p. 739. John Wiley 8, Sons Inc., New York, N.Y., U.S.A. 1986. ) Geige(2 r K.W e I.A.E.A .Proc: th f o . . Consultants' Meetinn o g Neutron Source Properties edited by K. Okamoto, June 1980, pages 43 to 78. Blinov M.V. : ibidem, pages 79 to 106. (3) Proceedings on the Consultants' Meeting on Isotopic Neutron Sources for Neutron Activation Analysis. I.A.E.A. Vienna, Austria, 21-24 May 1985; Edited by H. Vera Ruiz. (4) Alaerts L., Op de Beeck J., Hoste J. : J.R.A.C., 15(1973)601. e Beec) d Host, A.C.A.: Alaert(5 p J. O k. J e, L. ,s 69(1974)1. (6) Ganayev I.H. : Neutron multiplier type Co-1 for activation analy- sis, Vses. Ob;]. Isotop, Atomizdat, Moscow 1970, p. 145. (7) Rubinson W. : J. Chem. Phys.. 17(1949)542. (8) Givens W.W., Mills W.R., Caldwell R.L. : Nucl. Instrum. Methods. 80(1970)95. (9) Handbook on Activation Data, I.A.E.A., Technical Reports Series n' 273, May 1987. Herzo) , Wild(10 W. g e H.R .Kraftwerkstechnik: , 64(1984)348. e Beecd , Host : J.R.A.C.p J. e O kNorr. D J , e) L. e (11 , 78(1983)137. ) Spenk , (12 Cles , H. Karlie Radiochemica: T. s . B k l Method f Analyo s - sis. Vol. I, I.A.E.A., Vienna 1965. p. 245. (13) Selecki A., Nowakowski Z., : Radiochem. and Radioanal. Letters, 1 ( 1969)247. . Radioanale BeecJ d , Host : p J. e O Norrk. D J e, ) L. e. (14 Chem., 59(1980)453. ) JunoRappor: . (15 E d t CEA-R 2980, Paris 1966. ) Bartose(16 , Adam , J. HostkNucl: F. s. J e. Intr d Meth.an . , 103(1972) 43. Applie: d Dam ) Adam. an R s . (17 F ds Gamma-Ray Spectrometry, Pergamon Press. Oxford, G.B., 1970. (18) Heath R. : Gamma-Ray Spectrum Catalogue, Ge(Li) and Si(Li) Spec- trometr. ReporEd d t 3r yANCR-1000-2 , USAEC, Idaho Falls, Idaho, U.S.A., 1974. (19) Grosjean C.C., Bossaert W. : Table of Absolute Detection Effi- ciencies, Univ f Ghento . , Belgium (1965).

113 HOW TO ORDER IAEA PUBLICATIONS

An exclusive sales agent for IAEA publications, to whom all orders and inquiries shoul addressede db bees ha ,n appointed followine inth g country:

UNITED STATES OF AMERICA UNIPUB, 4611-F Assembly Drive, Lanham, MD 20706-4391

followine th n I g countries IAEA publication purchasee b y sma d froe mth sales agents or booksellers listed or through major local booksellers. Payment can be made in local currenc witr yo h UNESCO coupons.

ARGENTINA Comisiôn Nacional de Energi'a Atômica, Avenida del Libertador 8250, RA-1429 Buenos Aires AUSTRALIA Hunter Publications GippA 8 5 , s Street, Collingwood, Victoria 3066 BELGIUM Service Courrier UNESCO, 202, Avenue du Roi,B-1060 Brussels CHILE Comisiôn Chilena de Energi'a Nuclear, Venta de Publicaciones, Amunategui 95, Casilla 188-D, Santiago CHINA IAEA Publication Chinesen si : China Nuclear Energy Industry Corporation, Translation Section, P.O. Box 2103, Beijing IAEA Publications other than in Chinese: China National Publications Import & Export Corporation, Deutsche Abteilung , Beijin, 88 P.O x g.Bo CZECHOSLOVAKIA S.N.T.L., Mikulandska4,CS-11686 Prague1 Alfa, Publishers, Hurbanovo nämesti , CS-8158e3 9 Bratislava FRANCE Office International de Documentation et Librairie,48, rue Gay-Lussac, F-75240 Paris Cede5 x0 HUNGARY Kultura, Hungarian Foreign Trading Company, P.O149x Bo . , H-1389 Budapes2 6 t INDIA Oxford Book and Stationery Co., 17,Park Street,Calcutta-700016 Oxford Book and Stationery Co.,Scindia House, New Delhi-110001 ISRAEL Heiliger & Co. Ltd. Kere3 2 n Hayesod Street, Jerusalem 94188 ITALY Libreria Scientifica, Doit. Luci Biasie od o "aeiou", Via Meravigli 16, 1-20123 Milan JAPAN Maruzen Company, Ltd, P.O. Box 5050,100-31 Tokyo International PAKISTAN Mirza Book Agency, 65, Shahrah Quaid-e-Azam, P.O. Box 729, Lahore 3 POLAND Ars Polona-Ruch, Centrala Handlu Zagranicznego, Krakowskie Przedmiesci , PL-00-06e7 8 Warsaw ROMANIA 136-137x lleximBo . O ,P Bucharest SOUTH AFRICA Van Schaik Bookstore (Pty) Ltd, P.O724x .Bo , Pretoria 0001 SPAIN Diaz de Santos, Lagasca 95, E-28006 Madrid Diaz de Santos, Balmes417, E-08022 Barcelona SWEDEN AB Fritzes Kungl. Hovbokhandel, Fredsgatan 2, P.O. Box 16356, S-10327 Stockholm UNITED KINGDOM r Majesty'He s Stationery Office, Publications Centre, Agency Section, 51 Nine Elms Lane, London SW85DR USSR Mezhdunarodnaya Kniga,Smolenskaya-Sennaya 32-34, Moscow G-200 YUGOSLAVIA Jugoslovenska Knjiga,Terazije 27, P.O. Box 36, YU-11001 Belgrade

Orders from countries where sales agents have not yet been appointed and request r informatiosfo n shoul addressee db d directl: yto Divisio f Publicationno s International Atomic Energy Agency Wagramerstrass , P.O100e5 x .Bo , A-1400 Vienna, Austria