NONDESTRUCTIVE 14 MEV NEUTRON ACTIVATION

ANALYSIS OF JADES

( 玉 之 非 破 壞 性 中 不 活 化 分 析 )

by

Chan Tak Shing

( 陳 德 誠 )

A Thesis submittteCd in Partial Fulfillment

of the Requirements for the Degree of

Master of Philosophy in Physics

The Chinese University of Hong Kong

June 1982

Table of Content

Acknowledgements i

Abstract ii

I. Introduction 1

II. Activation Techinique

2.1 Jade Composition Determinations

2.1.1. Activity Determinations 4

2.1.2. Interference Reactions 6

2.1.3. Optimization of Conditions 11

2.1.4. Compositions 13

2.2 Coincidetuce Counting Applied to the Decay

Scheme of 27Mg

2.2.1. Basic Theory 14

2.2.2. Resolving-Time Determination 17

2.2.3. Accidental Coincidences 18

2.2.4. Coincidence Measurements 24

2.3 Internal Conversion Coefficients 28

III. Experimental

3.1 14 MeV Neutron Generator 31

3.2 Neutron Flux Measurements 32

3.3 The Pneumatic Transfer System 35

3.4 Sample Container 39

3.5 Gamma-Ray Spectrometry

3.5.1. Detectors 42

3.5.2. Electronic Relay Unit 46

36 Standard Samples 47

Page

IV. Results and Discussion

V. Conclusions

Appendixes

A . The Characteristics of Gems

B. Occurrence. of Jades

C. Plates of Jades

References

Page:

52

70

71

75

76

81

1

Acknowledements

I wish to thank my supervl sor, Dr. L. S. Chuang, for

his guidance and encouragement. I also wish to thank

K. S. Sin and S.W. Wong for their technical assistance. The

jades in this project were on loan -through the kindness of

0' Tama Jewellers and the Art Gallery of The Chinese

University of Hong Kong. It is a pleasure to thank

S.Y. Hung for her assistance in taking colour plates of

the jades. Thanks are due to L. Y. Kwok for typing this

thesis. I am also indebted to Dr. T. C. Wong for his

helpful discussions. The financial support of the Physics

Lepartment of the University is grate-ful.ly acknowledged.

Finally, I wish to express my gratitude to my family for

their continued encouragement.

T.S. Chan

June, 1982

11

AbstI',--Lct

Neutrons eneratt.ed from a 14 MeV neutron generator

were used to bombard the various elements of jades, and the

resulting gamma ac. L i vit ices, measured by. a 2-inch x 2-inch

NaI(Tl) scintillation detector coupled to a multichannel

pulse-height analyser, revealed the chemical composition of

the jades. The usefulness and advantages of thi..s analytical

method in the analysis of jades are described.

A simple relay unit which simultaneously initiates

the single-channel analysers, scalers, and multichannel

analyser in coincidence counting was constructed. The

accuracy of this unit is ±0.03 sec.

For comparative measurements of gamma activities,

standards made from chemical mixtures were used. The

accuracy and u.use[ulnc-.ss of these standards are discussed.

A. coincidence technique was introduced to study the

decay scheme of 2 'PvMg, which is a radionuclide resulting from

the nuclear reaction 27Al(n,p)?1M,1g. By measuring the

coincidence between the 0.844 MeV and 1.014 MeV gammas of

27Mg, the time correlation between these two gammas was

investigated.

By an extrapolation method, the internal conversion

coefficients for the gammas of` 7MMMg were found to be

3.63 x 10-' and 4.67 x 10-5 for the 0.84 MeV and the 1.01 MeV

gammas, respectively.

The interference problems encountered in multi-element

determinations, and the solutions to them, are discussed.

1

Chapter 1

Introduction

In recent years, neutron activation analysis has

developed into a powerful analytical method of elemental

analysis, and is widely used in many areas and in various

applications. This technique can frequently offer the

advantages of good sensitivity and selectivity, high

precision and accuracy, and non-destruction of samples.

Coupled with modern techniques such as charged-particle

activation analysis, X-ray fluorescence, and atomic

absorption spectrometry, neutron activation analysis

serves as -one of a series of complementary analytical

methods.

Neutron activation analysis is widely applied in

industrial, environmental, archaeological, and medical-

studies. Perhaps the results of this research will help

to bring the technique to the,commercial field. Strictly,

or mineralogically speaking, jade is divided into two

classes, nephrite and jadeite (Webster, 1975). However,

many greenish tough minerals, which have an appearance

similar to jadeite or nephrite are often, though wrongly,

called jade. From the points of view of both sellers

and customers, it is important and necessary to distinguish

clearly between the green stones, as the price for a

genuine jadeite and an ordinary green stone differs

tremendously. Without going through certain tests, but

instead merely depending upon the external appearance

2

of a given greenish stone, it is di.I:.-i'icult to tell whether

it is really a jade. The tests general l.y employed by

jewellers are the specific gravity test, the hardness test,

the refractive index test, and the absorption spectrum,

test (Liddicoat, 1969). All these tests are nondestructive,

except the hardness test, which may leave a scratch on the

stone. Some minerals, for example smaragdite, chloromelanite,

and the so-called "synthetic jade", may not be distinguished

from jade by the above tests. Nevertheless, if the chemical

compositions of these stones are examined, one may be able

to establish the identities of the stones.

The jades analyzed in this study were each put

near the target of a 14 MeV neutron generator and bombarded

by the fast neutrons. By measuring the gamma rays emitted

from the induced activities, and using the comparator

method, the concentrations of some the elements in the

jades were measured. The gamma-ray spectrum of jadeite

can be easily identified by its two intense 21Mg photopeaks

(formed from aluminum), which 'have energies of 0.844 MeV

and 1.014 MeV, respectively. Besides aluminum, other

elements after being activated can also emit gamma radiation

of about 0.844 MeV (for example, iron) and of about 1.014

MeV (for example, chromium). I f there were a time

coincidence between the gamma rays emitted by 27Mg, the

product nucleus of the 27A1(n, p) 27 Mg reaction, it would

be possible to distinguish the gamma activity of 27Mg

3

from that of other radionuclides that emit gammas of similar

energies. Some workers (Ciuffolotti and Demicheli s, 1 962)

reported that the two 27Mg radiations are in coincidence,

but the present decay scheme (Lederer and Shirley, 1978)

does not show this. With this in mind, we conducted a

reinvestigation of the decay scheme of 27 Mg by studying

the coincidences of the 0.844 MeV and 1.014 MeV radiations,

and their internal conversion coefficients.

In the analysis of the pulse-height spectra,

photopeak areas were obtained by the total peak area

method (Kokta, 1973), which employs a simple baseline

subtraction technique. The counting statistics (Quittner,

1972) obtained for sit. icon, which constitutes about 27%

by weight of jade, were ±1.05% (one standard. deviation),

measuring the 1.779 MeV gamma-ray peak of 2.246-minute

2 6A1.

Chapter II

Activat ion Technique

2.1 J a de Compo sition Peterminations

2.1.1. Activity Peterminat ions

14 MeV neutrons are produced at the target of the

neutron generator (Kaman type 711A) (Karri an, 1973) via the

react ion :

The high-energy neutrons so produced were used to

bombard samples, to produce various radionuclides. The

measured activity of each radionuclide can be related

to the amount of target element in two ways, namely, via

the absolute method, and via the comparator method.

In the absolute method (Garrec, 1969) (Girardi,%

1964), the activity, A, at the end of time tp after an

irradiation for a time duration tj for each radionuclide

is given by:

where N is the total number of atoms of target nuclei (of

a given Z and A) in the sample, a is the cross section of

the target nuclide for a given reaction with neutrons of

5

a given energy, cb is the neutron f lux to which the target

atoms are exposed, f is the total attenuation correction

factor of neutrons and gamma rays, E is the detection

efficiency, and A i s the decay constant of the radionuclide.

The total attenuation correction factor, which is

the product of the neutron attenuation factor and the garnma-

ray attenuation factor, can be calculated by using the

technique developed by Nargolwal l a (1968).

In actual analysis, an accurate absolute neutron

flux value and the attenuation correction factor cannot

be determined easily (Meinke, 1957). By the use of the

comparator method, these difficulties can be overcome.

With it, the mass of the element is determined by comparing

the intensity of the gamma radiation of the sample, as

measured by the area of the appropriate photopeak, with

that of a standard which was irradiated and counted under

the same conditions. Iddings (1964) and Volborth (1963)

have recommended this method as being the more accurate one.

If Au and Ag are, respectively, the activities of

the unknown and of the standard, and if Mu and Ms are the

corresponding masses of the element, then the following

equation holds:

with neutron flux the same for both samples, as well as

all other irradiation and counting conditions. However)

the neutron flux usually varies during the analyses, and so

a neutron m o n i t o r i s u s e d . T h e a b o v e e cj u a t i o n i s t b e n

modified to:

where Rs and R are the neutron monitor counts for the

standard and unknown samp 1 e , respeclively, dur ing 1:he

neutron-irradiation period.

2.1.2 Interference React ions

M a n y a u t h o r s h a v e u s e d t h e n e u t r o n a. c t i v a t i o n

technique to determine the silicon and oxygen contents of

meteorites (Wing, 1964), and, the aluminium and silicon

contents of rocks (Turner, 1956). In advance of any

activation measurements, it is useful to consider what

chemical elements exist in jades. Since jade comes from

minerals of the Earth, we would expect the elements present

in jades to have some correlation with the abundances of

the elements which constitute the Earth's crust (Weast,

1970). From Table 2.1, it can be seen that some eight

elements have abundances greater than 2% by weight of the

Earth's crust. They are oxygen, silicon, aluminium, iron,

calcium, sodium, potassium and magnesium. All of these

are known to be present in some jades (plus Mn and perhaps

Cr), and can be readily activated, using the technique

of 14 MeV neutron activation analysis (Nargolwalla, 1973).

That isj too closely similar in energy to be resolved from

one another with the type of detector used.iJ L

H owev e r, fr om t h e nu c1e ar d at a for the r e ac tion s of

these elements with 14 MeV neutrons (Nargolwalla, 1973),

(Table 2.2), it can be seen that two types of interference

reactions exist. The first type is that in which two or

more activation products emit gamma rays of the same energy

or nearly the same energy, e.g., the activation of 2Al

and 56Fe with 14 MeV neutrons produces different radionuclides,

but both decay with the emission of gamma rays of about

0.844- MeV energy. The type 2 interference is one in which

two different elements, when activated, yield the same

Oli O V r 11radionuclide, e.g., Na produced from both ;A1 and z Mg.

There are two principal ways to solve such

interference problems, namely, by means of half-life

differences of the radionuclides, or by means of the fact

that one of the elements may give rise to a secondv '

photopeak which is free from interferences. Belonging to

the type 1 interference, 2 7Mg has a half-life of 9.4-5

-minutes, whereas 56Mn has a half-life of 154.9 minutes.

After a waiting time of 45 minutes, for example, the

activity of 27Mg will be reduced to only 0.037 of its

E1 emeut Abundances, in %

0

Si

A1

Fe

Ca

Na

K

Mg

Ti

H

P

M n

S

C

CI

Rb

F

Sr

Ba

Zr

Cr

V

Zn

Ot

46. 60

27 . 72

8 . 13

5.00

3 . 63

2 . 83

2 . 59

2 . 09

0.440

0.140

0. 118

0. 100

0. 052

0.0320

0.0314

0.0310

0.0300

0.0300

0.0250

0.0220

0.0200

0.0150

0.0132

0 . 31

Table 2.1

Chemical Composition of the Earth's Crust;

(Elements present 0.01%, from Mason, 1952

9

initial value, wllcrcas that of 5 6Mn will only be reduced

to 0.818 of its initial value. In the case of the type

2 interference cited as an example, the 1.014 MeV ph.otopeak

of 2 7Mg can be used to ascertain the 27 Al contribution

to the 2 Na activity from that due to 2 1+Mg.

TARGET

NUCLIDI, REACTION

PRODUCT

RADIONUCLIDE

and half-lift

TARGET

% ISOTOPIC

ABUNDANCE

14 MeV

O (nib)

MeV OF MAIN y'S (AND

THEIR % EMISSIONS)

160

2 3Na

2 3Na

2 '4Mg

2 5Mg

2 6Mg

27A1

27a:

2 8 Si

29Si

30Si

39r

UI(

41K

4 2Ca

4 4 Ca

4 4 Ca

5 0 Cr

5 2Cr

5 3 Cr

5 5Mn

5 4Fe

5 ,F e

5 7F e

(n,p)

(n,p)

(n,a)

(n,p)

(n,p

(n, a]

(n,p;

(n,a)

(n,p;

(ri,p)

(n,a)

(n, 2n)

(n,p)

(n,a)

(n,p)

(n,p)

(n,a)

(n, 2n)

(n,p)

(n,p)

(n,a)

(n,2n)

(n,p)

(n,p)

16N5 7.13s

23Ne,37.6s

2 °F, 11.0s

24Na,15.02b

2 5 N a, 6 0 s

2 3 N e, 3 7 . 6 s

27Mg, 9.45 m.

24Na,15.021

2 8A1, 2.246m,

! 29A1, 6.52m

2'Mg, 9.4 5n

38K, 7.6 3n

41Ar, 1.83h

3 8 C1,3 7.2m

42K,12.36b

44K,22m

41Ar, 1.83b

4 9Cr,42.Om

52V, 3.755m

53V, 1.55m

52V, 3.755m

5 3Fe . 8.53m

5 6Mn, 2.5821

:) 1 Mn, 1. 59m

99.76

100.00

100.00

78.99

10.00

11.01

100.00

100.00

92.20

4.70

3.10

93.30

6.70

6.70

0.65

2.08

2.08

4.35

83.79

9.50

100.00

5. 80

91.7

2. 19

39

43

150

190

44

7 7

75

116

230

120

70

3.5

49

39

182

36

35

19

94

40

32

15.5

103

75

6.128 (69%)

0.439 (33%)

1.633 (100%)

.1.369 (100%)

2.754 (100%)

.0.391(12.8%),0.586(13%),

V975(12.8%),1.612(8.8%)

0.439 (33%)

0.844 (72%),

1.014 (28%)

1.369 (100%),

2.754 (100%)

1.779 (100%)

0.511 (200%),

1.273 (91%)

0.844 (72%),

U.014 (28%)

0.511 (200%),

2.167 (100%)

1.294 (99.2%)

.1.642 (32.8%),

2.168 (55%)

1.525 (17.9%)

.1.024(10%),1.127(13.5%).

(1.157(85%),2.150(29%),

2.519 (10%)

1.294 (99.2%)

,0.511 (192%),

1.152 (29.5)

1.434 (100%)

.1.006 (88.7%),

(1.287 (11.3%)

1.434 (100%)

.0.377 (43%),

0.511 (196%)

,0.847(99%),1.811(30%),

2. 113 (15.5%)

.0.122 (9.5%) ,

1.692 (2.7%)

Table 2.2

Values taken from G. Crdtmann, 1976

DATA FOR 14 MeV NAA OF JAPE S

2.1.3 0 p t i m i x a t i o n o f Con d i t i on s

NAA = neutron activation analysis

From the above, it can be seen that proper selection

of the irradiation time is important in multi-elemental

analysis by NAA. The half-lives of the significant

activated nuclides in jades differ greatly, ranging from

7.13 seconds to as long as 901 minutes. Due to economic

considerations, and the need to minimize interference

reactions, the irradiation time for each analysis was

chosen to be 600 seconds. In this time, the fraction of

saturation for 27Mg is 0.520; 28A1, 0.954; for 56Mn, 0.0438;

for 24Na, 0.00766. Shortening the irradiation time to, say,

300 seconds, the fractions of saturation of the short-lived

nuclides would not change much: 27Mg to 0.307 and 28A1 to

0.786. However, for the long-lived nuclides 5 6Mn and 2iiNa,

the fractions of saturation would only reach 0.0221 and

0.00384, respectively, and these activities would then be

too low to be detectable in a long-lived background counting

If the irradiation time were lengthened to 1000 seconds, th.it

would essentially double the activities of 56Mn and 24Na,,

but would correspondingly increase the MCA deadtime loss

and cost of the experiment.

Of equal importance, the decay time for the counting

of the activated samples needs careful selection. Owing

to the great abundance of silicon in jade, the Compton

edge and Compton continuum of the 28A1 (ti = 2.246 minutes)

activity introduce a large background for the detection of

photopeaks of other elements. The only way to remove this

interference is to take advantage of the short half-life ol

28A1, by increasing the decay time. However; there is one

element present in jade, chromium.which is of high interest

5 ?

Too long a decay time will decrease the activity of V,

the product of the 52Cr(n,p)52V reaction, whose half-life

is 3.755 minutes.

In order to measure or attempt to measure the

elements silicon, aluminium, iron, magnesium and chromium,

we divided the counting process into two stages. The first

stage was to measure the short-lived radionuclides, while

the second stage was for the measurement of the longer-

lived radionuclides. For stage 1, a decay time of 20 seconds

was chosen, which still allowed sufficient time for the

1 6sample to be put at the counting position and for N

(ti = 7.13 seconds), the product of the 160(n,p)16N

reaction, to decay out sufficiently. This short decay

time was chosen in order that the gamma activity of 5 2V

might be measured efficiently. ' For stage 2, a decay time

of 45 minutes was chosen, by which time the activities of

28A1 and 27Mg were reduced by factors of 9.30 x 107

and 0 .0369, respectively, and their contributions to the

spectrum were thus made negligible.

The selected counting times for both stages were

2000 seconds, in order to obtain better counting statistics

for the activities of 7Mg and the coincidence countings

for its 0.844 McV ancl 1.014 MeV gammas, and for the

activities of the Iong-1ife nuc 1 ide 2 'fNa. Garrec (1969 )

suggested an irradiation time of 10 minutes, a 130-minute

decay time, followed by a 10-minute counting time. However

this lengthy analysing time was not felt to be appropriate

in this work.

Two runs were made for each sample, the time

interval between the analyses being 1 day for jadeite samples,

and 3 days for other jades, since jadeite produces mainly

short-lived radionuclides, 28A1 and 7Mg, whereas other

jades produce significant amounts of the long-lived

radionuclide, 2t+Na.

2.1.4. Compositions

Due to the rather poor energy resolution of Nal(Tl)

scintillation detectors, the 1.811 MeV photopeak of 36Mn,

f~ [T gproduced by the nuclear reaction 1 Fe(n,p)3 5Mn, overlaps

with the 1., 779 MeV photopeak of 28A1, produced by the

i

reaction 28 Si(n,p)28A1. The contribution of 2dAl at

1.779 MeV can be found by a method similar to that suggested

by Grieken (1968), that is:

where k„ is the ratio of the activity of 56Mn at 1.779 MeVFe

to that at 0.847 MeV , ancl f rom act ivat ion of pure iron , the

value of kr, was found to be 0.235 ± 0.008.f i n

3 1 2 8The nuclear reaction of phosphorus, P(n,a)~ Al,

also produces the same 1.779 MeV gamma radiation. However

the amount of phosphorus present in meteorites and rocks

is less than 0.3% in all the known cases (Wing, 1964),

and thus phosphorus should not interfere significantly in

the analysis of jades (see Table 2.2 ).

The compositions of the various jade samples were

determined as follows. In stage 1 counting, both aluminium

(27Mg) and iron (56Mn) contribute to the 0.844 MeV photopeak,

r- r- ry Qand both iron (J Mn) and silicon ( Al) contribute to the

1.779 MeV photopeak. In stage 2 counting, only iron (J 6Mn)

plays a role in the 0.844 MeV photopeak, and only magnesium

(2 ''Na) in the 1.369 MeV photopeak. By subtracting the

c o n t rib u tio n o f J GMn (amo un t determine d f rom s t age 2) t o

the 1.779 MeV peak in stage 1, the amount of silicon can

be obtained. The amount of aluminium may be found in%

a similar manner. Chromium, whose 1.434 MeV photopeak

lies on the Compton edge of the 1.779 MeV gamma of 28A1,

by a spectrum-stripping method, using both graphical and

computer techniques (Kowalski and Isenhour, 1968).

2.2 Coincidence Counting Applied to the Decay Scheme of 27Mg 1

2.2.1 Basic Theor'

Besides 27Mg, a number of other fast-neutron

product radionuclides of common elements also emit gamma

rays of about 0.84 MeV andor 1.01 MeV, and hence can

interfere with the determination of the aluminium content

of jade samp 1 es. I f , however , tlie two gamma rays of 27Mg

are emitted in coincidence, it might be possible to

distinguish the ZMg activity from that of, say, 5b.Mn

and 52V - by use of a coincidence technique. Some workers

(Ciuffolotti and Demiche1is, 1962) c1aimed that the two

gammas of 27Mg are emitted in coincidence, but the recent

(Fig. 2.1) decay scheme of 27Mg (Lederer and Shirley, 1978)

(Ophel and Lawergren, 1963), shows that the two are not in

coincidence. Therefore, in the process of conducting these

jade analyses, the decay scheme of 27Mg was also studied.

Suppose a radionuclide source emits two gamma rays

in cascade , and i. t:s d isintegrat ion rate is no. Dectector

1 will have a countine: rate ni . that is:

n i = nob (2 ~ I T)

where £i is the overall efficiency of detector 1,

Similarly for detector 2,

n 2. = no£; (2 - 2)

If these two detectors are connected in coincidence, the

coincidence count rate measured fSieabahn and Bell. 1965)

will be:

ni? — n o £i£ o (2 - 3)

27Mq12 9-45 mi n

6 7 %s

.33 %

1014

J9-844

2713

28AI

13 2-24 6 mi n

100%

1-779

28Si

14

Fiy. 2.1

Decay Schemes for 2 7Mg and 28A1

The accidental coincidence rate ip is given by:ct

(2 - 4)

where t is the resolving time, which is the time period in

which two pulses can be recorded as a coincidence event.

It can be seen from equation (2 - 4) that even a

small increase in no will result in a relatively large

2increase in n , since n increases with no , while n]2

a a

increases with no- The ratio of the true coincidence

counting rate to accidental coincidence counting rate is

thus :

In order that the accidental coincidence rate he

reduced to a minimum with respect to the true coincidence

rate, the resolving time should be kept as short as possible.

2.2.2 Resolving-Time Determination

There are two methods (Bleuler and Goldsmith, 1960)

for the determination of the resolving time of a counting

system. The first method is to plot the number of

coincidences as a function of delay inserted in one of the

channels, as shown in.Fig. 2.2. Since the coincidence

analyser detects only the leading edge of the input signals,

the width of these signals does not affect the resolving

times (Canberra, 1979). Another method is to plot the

rate of coincidences as a function of the product of the

disintegration rates of two unrelated sources (Fig. 2.3).

In this case, two independent sources are counted with the

two detectors, which are well shielded from each other so

that no coincident radiation can reach both detectors.

The coincidences observed, then, are only due to the finite

resolving time and to a background produced by cosmic rays

and natural radiations.

By means of these two methods, the resolving time

of the coincidence circuit was found to be 38 nsec. and

34 nsec., respectively.

2.2.3 Accidenta1 Coinc1dences

For fairly short half-life nuclides, such as 27Mg»

and 28A1, accidental coincidence events can be calculated

as follows. The windows of one single-channel analyser are

set to accept pulses in the 0.84 MeV gamma region, and those

in the other single-channel analyser are set to accept

pulses in the 1.0 MeV gamma region.

300

200

100

= 38 nsec.

2 T

S o u r c e u s e ci: 2 2 N a

0

1 . 72 1. 84 .. 96 2 . 08 2.20 x 0.1 usee.

R e 1 a t i v e D e 1 a y S e 11 i n g

Fig. 2.2

Reso1ving~Time Determination

0d•HdPrO 'O •

oCO 0P COdd oo ca

o P

0 dO Hd0

•HOd

•Ho

o

o00

OOlO

C•rH

+-C0

w

0O£0

HO£

•rHoo

rH0$-P

0To•HOO

300

200

100

0

5 10 15

n - 2T nin 2cl

Slope=34 nsec.

n i x n2

20x106 sec2

Fig. 2.3

Resolving -Time Determination

( i ) Acci dental__coinc i dences duo to pure aluminium

From equations (2 - 1), (2 - 2), and (2 - 3):

where 0.72 and 0.28 are the 2 7Mg branching ratios

(Lederer and. Shirley, 1978) for the 0.844 MeV and

1.014 MeV gammas of 27Mg, respectively. The total

number of accidental coincidences measured in the

counting period is:

2 t ( 0 . 28 ) ( 0 . 72 ) e i c 2 n o 2 dt , ( 2- 5 )

where t is decay time = 20 sec., and tc is counting

time = 2000 sec.

The total counts recorded by the single-channel

analyser set for the 0.844 MeV region is:

(2-6)

Similarly, for the single-channel analyser set for the

1.0 MeV region:

n o e 2 ( 0 . 2 8 ) clt (2 - 7)

Combining (2-5), (2-6), and (2 - 7), we obtain:

N = 2t x 7.25 x 10 4 N2N9a

(ii) Accidental coincidences due to pure silicon

From (2 - 1), (2 - 2), and (2 - 3):

n i = n o e i p i

n 2. - n o e 2 P 2

n = 2 t n ! n 2cl

(2-8)

(2-9)

where pi and p2 are the probabilities that a 1.78 MeV

gamma-ray photon will scatter to an energy of 0.84 MeV

and of 1.01 MeV, respectively.

Total accidental coincidence counts:

1

If n i . 7 b is the coun t ing rate registerecl by a single-channeI

an a1ys er at 1.78 MeV in detecto r 1, wit h counting e f ficien cy

iei.78; then:

i

H I . 7 8 = n o C 1 . 7 8 (2 - 10)

Then, dividing (2 - 8) by (2 - 10), we obtain:

i

Taking Ni.7 8 as the total number of counts registered by

the single-channel analyser at 1.78 MeV in detector 1, we

have :

and

Similarly, if Nj.78 is the total number of counts registered

by the single-channel analyser at 1.78 MeV in detector 2,

2

with counting efficiency e!.7 8,then:

N? N iThe values of (—7 ) and (—7 —) can be found from

N 1 . 7 8 N 1 • 7 8

experimental spectra, and both were found to be approximately

equal to 0.4.

For a typical spectrum of aluminium, the total number of

accidental coincidence counts calculated was 0.23, for a mass

of aluminium = 0.37 g.

For a typical spectrum of silicon, the total number of

accidental coincidence counts calculated is 0.11, for a mass

of silicon = 0.17 g.

2.2.4 Coincidence Measurements

By measuring the coincidences of the two gammas of

27Mg in jadeite samples, it was found that the coincidence

counting rate was directly proportional to the 0.84 MeV

single-channel analyser counting rate, and to the 1.01 MeV

single-channel analyser counting rate, respectively (Fig. 2.4). i

These coincidences may first be considered as true coincidences

Then, when plotted against the product of the two single-

channel analyser counting rates, a. parabolic curve is obtained

o0CO

OOoCnI

C•H

130U20

0

0

£33O

O

0O£0

13•HOt—t

Ho

o

8C

6C

40

20

0

4 8 12 16 20 24 28 32 x 10

Single-Channel Analyser Countingat 0.84 MeV

Fig. 2.4

Coincidence Counts Versus Single-Channel Counts

N1XN2

Fig 2.5

plot of conincidence counting versus product

of Sca 'S Counting

(Fig. 2.5). However, if the gammas are in true con evidence,

the Ni obtained by calculation was much larger than that

found by expertment. Instead of j ade it e , pure a. 1 urnin ium

and silicon were then used for similar coincidence countings,

with the single-channel, analyser windows again at 0.84 MeV

and 1.01 MeV, respectively. An activated sample of silicon

was sandwiched between a pair of thin (1.2 mm thick)

Australian jades before counting, so that the overall

density approached that of aluminium. The result obtained

is indicated in Fig. 2.6. The slopes of the two graphs

are 1.98 x 10 and 1.29 x 10 , respectively. However,

the single-channel analyser countings in Fig. 2.6(a)

include the photopeak area, N, and the Compton continum, B.

If the coincidences are plotted against B, then the slope

of the graph becomes 1.33 x 10 3, which is similar to that

of Fig. 2.6(b).

The results thus show that coincidences in activated

a 1 urniniurn arise not from from its ? 7Mg photopeaks , but

rather from gamma-ray seattering. Simi1ar phenomena occur

with jadeite samples, the coincidences recorded being due

to scattered 1.78 MeV gamma-rays of 28A1. This peculiar

coincidence happens as the two detectors are placed very

close to each other and in a face-to-face geometry.

(Fig. 2.7).

2 . 3 Internal Con vers 1 o n C o e ff 1. c i e n t s

To check on the correctness of the coincidence

measurements made in this study of the decay scheme of 27Mg,

the internal conversions of its 0.844 and 1.014 MeV gamma

rays were estimated by extrapo1ation. In returning to the

ground-state configuration, excited states of radionuclides

usually decay either by a radiative process (emitting a

gamma-ray photon) or by a non-radiative process, which in

most cases is the emission of an internal conversion

electron.

The total internal conversion coefficient is

defined as the ratio of the probability for internal

conversion to that for gamma emission. In this work, the

conversion coefficients of the two gamma rays from 2 7Ivlg

were obtained by extrapolating values for various atomic

numbers greater than 30 (Lederer, 1978) to Z=12. By means

of a least-squares fit, plotting the logarithm of the

conversion coefficient against the logarithm of the atomic

number, the internal conversion coefficients for the gammas

i n 2 7 M g were f o u n d 1; o b e

Energy in MeV :iK:Li

aLII Hi

0.844

1. 014

2.97 x 105

3.97 x 105

6.50 x 10 6

7.02 x 106

2 . 20 x 10 '8

1.58 x 108

3.95 x 10 8

1.83 x 10~9

Hence, ar = 3.63 x 10 2 (100% E2) for the 0.844 MeV gamma,

• and = 4.67 x 10 5 (90% Ml + 10% E2) for the 1.014 MeV.

These calculations thus show that internal conversion in

27Mg can be neglected, since the internal conversion

coefficients are extremely small for both energies.

w-P5=1Go

o

CDOGCD

-d•Ho

•Ho

o

50

4(

30

20

10

S1 o p e = 1. 3 3 x 10

against B

1 IN

J3l.

-- a g a i n s t (N+B)

Slope=l.98x10 '

0 2 4 6 8 10 12 14 16 18xl0,(

SCA Counting at 0.84 MeV

( a) A1 urninum Samp 1 e

w+-Gpo

o

ooG0)

•Ho—

oo

5C

4C

3C

2C

1C

0 4 8 12 16 20 24 28x103

Slope=l.29x10

SCA Counting at 0.84 MeV

(b) Silicon S amp1e

TT-i cr 9. I

C oin cidences v ersu s S C A Co unting

i'Yir ( 'A A I nnii mini 1b .9 1 1 i cnn

h v

i nc ident

gamma photon

f ree

e iectron

scat tered

gamma photon

hi;'

scattered

eiectron

Na I (T I)

c rystaI

i

Vsample

Na I ( TI )

crystaI

2

Fig. 2.7

Compton Scat, t e r i n g

Chanter III

Ex per linen Lai

3 • 1 14 MeV Neutron Generator

The neutrons for the present research were produced

by a 14 MeV neutron generator (Kaman type 711A) (Kaman,

1973) via the nuclear reaction:

This generator was provided by the International Atomic

Energy Agency (IAEA) in 1974, and is of the sealed-tube

type, It is installed at the Science Centre of The

Chinese University o f 11 ong Ko n g. As the ins t a11a tio n sit e

is quite close to classrooms, the radiations generated are

heavily shielded to at least a safety factor of 10 (Chuang,

1975), in addition to the allowed safety level suggested

by the manufacturer. Operation of the generator is achieved

through a control console, which is situated in a control

room separated from the generator room by a thick concrete

wall.

Common to all sealed-tube accelerators, this

generator tube consists of an ion source, an accelerating

structure, a target section, and a replenisher. The ion

source is of the Penning Ion Gauge (PIG) type, in which

deuterons and tritons are produced with atomic-to-molecular

ion ratio of about 0.05 (Wood, 1972). Situated behind the

ion source, the resplenisher unit is designed to adjust the

hydrogen gas pressure in the acce 1 erator tube . I t cons:i st s

of two gas occluding elements made of titanium wrapped on

tungsten heating wire, and will absorb or release hydrogen

isotopes depending on the titanium temperature. The target

of the accelerator tube is of the gas-in-me 1:a 1 type , with

copper as the backing plate. Deuterium and tritium gas

are impregnated into a t; itanium matr ix .

As the accelerator employs a high voltage power

supply (160 KV), pressurized sulphur hexafluoride is used

for insulation. In a closed-loop heat-exchange system,

Freon-113 is used to cool the ion source, and in a similar

closed-loop system clean de-ionized, water is used to cool

the target assembly.

The spectrum of the neutrons emitted from the 14 MeV

neutron generator is shown in Fig. 3.1 (Chuang, 1979). The

energy of the neutrons centers at about 14 MeV, and in fact

depends on several factors, including the energy and the

atomicity of the incident ions, whether the reaction is

deuterium-upon-tritiurn ortritium-upon-deuterium , a.nd

the angle of emission (Chuang, 1979), (Stark, 1971),

(Fig. 3.2).

3.2 Neutron Flux Measurements

The absolute 14 MeV neutron output rate from the

generator was estimated by the Texas Convention Method

{Caman, 1969). This technique is based on activation

analysis and the nuclear reaction involved is the

4-I—U.T

ai -~r~c

H—

aP

r~4-K

arr

1C

p

f

2

2 l t 8 1( 12 14 M

ang leof

4 observatior

- c:

- 90(

- 150

Ne111 ron E ri p r q- v ( Mp V

Fie. 3.:

Neutron Energy Spectra at 0°, 90° and 150C

Incident particle energy ~ 0.16 MeV

(1) T + D

(2) D + T

4 He + n

4 He + n

Incident particle energy = 0.08 MeV

(3) T + D

(4) D + T

4 He + n

4 He + n0

bn£0£w

£O£XJ£0

15- 4

15-0

14-6

14-2

13-8

13-4

13-0

(i )

(2 )

(3 )

(4 )

0 20 4 0 60 80 IOC 120 140 160 180

(4)

'3 )(2)

!1)

A n g1e o f Ob s e r v alio n

Fig. 3.2

Variation of Neutron Energy as a Function of the

Ang1e of Observation for Different_ Mode1s of the

Production of the Neutrons

r _ c OCu(n,2n) ' ' C u r o a. c' t i o n . A co p p e r i o i 1 i s i. r r a d 1 a ted a. t

20 cm from the neutron target, and the 0.511 MeV gammas

r ofrom positron annihilation, due to the induced Cu

activity, were counted for one minute. The position of

the foil during co u n ting was 2.7 cm f r om a 3-inc h x 3-inc h

Nal(Tl) scintillation detector. The neutron yield at 2.5 rnA

beam current and 160 KV acceleration voltage was found to

be about 3.3. x 1010 nsec. (Table 3.1).

The Pneumatic Transfer System

For rapid transport of sample to the irradiation

position, and back to the counting station, a pneumatic

system was developed (Wong, 1975). It is a single-tube

system, with a gas inlet connected to the gas supply unit

and an outlet connected to the exhaust gas pipe line of the

building. Transit time is approximately 3 seconds and of

good reproducibility. The transfer gas used in this system

is air, but can be replaced by other gases, e.g., nitrogen

gas .

The transfer system (Fig. 3.3) consists of (i) an

adjustable constant-pressure gas supply unit, which is a

gas cylinder modified to act as a constant-pressure gas

reservoir connected to an air compressor; (ii) a programmable

electronic timer, to control the opening of the various valves

for blowing the sample towards the irradiation station and

backwards, and also the time for irradiation; (iii) an

irradiation station, which consists of a sample rotation

Beam Current, in ti iA Foil Mass . in

i

! Uncorrectey n min t c mi n

Background

C orrect ed

fnn n t~ q rn i

NeutronT ' 1.3 ' ,

l.

1.'

1. '

1.:

2 . ,

1 . 2 3 7 (

1 . 310

1. 240(

1 . 146

1.241!

708'

1055!!

1139!

1171r,

1618!

582!

800!

8 76 (

950

1 o n o

1.49 x 101

1.93 x 101

2.23 x 101

2.62 x 101

q o -i v i n1

Table 3.:

14 MeV Neutron Yield Measurements by the Texas Convention

direction of gas

flow when blowing in

direction of gaso

flow when blowing back

SLl

inletout I et

SL3

i rrad iat ionstat ion

SL 2 SL 4

count i ngsample in

station

SL 5

sample

discharge

Fig. 3.3

Schematic Diagram of the Pneumatic Transfer System

mechanism rotating at a rate of 170 rpm; and. (iv) a sample

loading station from which the sample is sent in for

irradiation .

The operation of the pneumatic transfer system is

as follows: the gas supply unit is connected to the inlet

of the pneumatic control panel, and the outlet of the

pneumatic control panel is connected to the exhaust gas

pipe line of the building. The outlet of the air compressor

is connected to a gas cylinder, and the air compressor is

? 7turned on until the pressure gauge reading reaches 5.5 Kgcm,

which is the optimum gas pressure for a sample of mass 8 g.

The capsule containing a sample is placed in the

loading station. When a steady production of neutrons is

achieved (a bo u t 3 0 s ec. a f ter t ur nin g o n the n eu11 o n

generation), the capsule can then be sent to the neutron

irradiation station. The distance from the middle of the

capsule to the generator target is 6 cm, and the axis of

the capsule is par• a11e1 to the target surface.

The whole control process is passed to the programm¬

able electronic timer once the start button is pressed.

At first, it activiates the solenoid valves SL2 and SL3,

so that they open for 3 seconds, letting the gas blow the

sample from the loading station to the irradiation station.

Then it starts to count the irradiation time, and when the

preset time is up, the solenoid valves SL1 and SL4 open

for 3 seconds, letting the gas blow the sample from the

irradiation station back to the counting stationdischarging

the sample by opening the solenoid, valve SL5 for 1 second.

3.4 Sample Container

Most of the samples undergoing analysis in this study

are quite expensive, special care had to be taken to prevent

them from being damaged, in the process of transporting them

from one station to another station. Sample containers were

constructed to fulfill this task. They are machined from

low-oxygen polyethene rods. Each capsule is of length

3.0 cm, with a cap which can be screwed tightly onto the

body. The outer diameter of the container is 3.2 cm,

whereas the inner diameter of the pneumatic tube is 3.5 cm.

Soft cylindrical tissue, having the same dimensions as the

container, is split into two halves. The sample to be

analysed is then sandwiched between the tissue before

being put into the container (Fig. 3.4). This is to ensure

that there is no damage to the sample during transportation,

and to fix the sample at the centre of the capsule. After

irradiation, the sample is taken out and put at the assigned

position for counting. This is to prevent unnecessary

contributions from induced activities of the capsule and

tissue,, and to provide the shortest source-to-detector

dist ance.

Low-oxygen polyethene was chosen for capsule making

because it only introduces a low background activity and it

is inexpensive. The transfer capsules can be reused for a

number of times, but it must be ascertained that the cap

screws tightly onto the capsule each time. Otherwise, the cap

may slip off the capsule, resulting in a loss of sample.

i I

s a mt

Fig. 3.b

Schematic Diagram of Detector System

42

3.5 Gamma-Ray Spectrometry

3.5.1 Detectors

Common to many research laboratories, solid sodium

iodide (thallium-activated) scintillation detectors were

used in this study for the detection of the gamma radiations

emitted by the radionuclides formed. In the experiments, the

crystal size employed was 2-inches x 2 inches. The detectors

(Canberra, 802-2) were positioned 2.2 cm apart, without lead

shielding, the arrangement of the detectors being shown in

Fig. 3.5. The scintillation detectors were each connected

to their respective preamplifiers (Canberra, 2005), and each

coupled to a double delay-line amplifier (Canberra, 1411).

A T-connector was used to provide two outputs from the

amplifier. One output was led to a single-channel analyser

(Canberra, 2036A), and the-other to a multichannel analyser

(Canberra, Omega 1, 1024 channels). One single-channel

analyser was adjusted to accept only in the region of 0.84

MeV, and was connected to a logic shaper and delay unit

(Canberra, 2055), where pulses were delayed before they were

delivered to a coincidence analyser (Canberra, 2040). The

other single-channel analyser was adjusted to accept only pulses

in the region of 1.0 N.TeV, and had its output directly led

into the coincidence analyser. This coincidence analyser

gives one count when one pulse from each detector arrives

within the resolving time of the circuit. Two face-to--face

detectors were installed, so that coincidence counting

and gamma-ray activity measurement could be carried out at

the same time.

With fine acljustment of an adjustable voltage gain

at the end of each photomultiplier tube base, the two

scintillation detectors could be made to match each other.

The optimum working voltage for the detector system was

found to be 1100 volts (Fig. 3.6), which was supplied by a

regulated high voltage power supply (Nuclear-Chicago Corp.).

The resolutions of the detector system were measured at the

0.662 MeV photopeak of 1 3 7Cs and at the 1.33 3 MeV photopeak

of 6°Co as 6.1% and 7.6%, respectively. Due to the

instability of the electronic components, three standard

point sources, 13 7Cs (0.6616 MeV), 6 °Co (1.1732 and 1.3325

MeV), and z Na. (0.5110 and 1.2745 MeV) were used in

calibrating the spectrometry system from time to time.

The photopeak counting efficiencies of the system,(Fig. 3.7),

were measured at these energies with standard sources

(Gunnink, 1961), and an absolute appi oach to do tennine the

intrinsic photopeak efficiencies was carried out by the

method of Lazar (1956).

One of the major factors that affect the precisions

attainable in 14 MeV activation analysis is the monitoring

of the neutron output rate, which can be clone by means of

an internal standard (Santos, 1968) or by means of an external

method (Mott, 1965). In our case, a neutron detector of

scintillation type (Nuclear Enterprises Corp., NE 422),

containing lithium in a matrix of zinc sulphide, was used.

As neutrons cannot produce scintillations directly, for

the detection of slow neutrons it is necessary to incorporate

some material, for example, lithium,into the phosphor, in

order to produce ionizing radiation on the passage of

(DPd

Pr~r—r-yPoo

0)

•H;

Co;—1

r-P

510

410

10-'750 850 9 5 0 1050 1150 1250 volts

opt imum

vol tage

A p p 1 led V o 11 a, g e

Fig. 3.6

Opt i mil rri V or king Vo 11 a ge _ Determin a t i_o n

de t e c t or 1

detector 2

G•H

Ocs

•rHo

I•r-1

Cl—w

CuCDQo-Po

£

5-

2

!

05

0-2 0-5 i 2

Gamma Energy, in MeV

P i g 3 . 7

Photopeak Efficiencies of the Two 2-inch x 2-inch

Nal(Tl) Scintil1ation Detectors

neutrons t hrough 111 o mi x111ro . W i t h t h i s dc) t oc tor , nou t runs

are detected in l.lie crystals by seinti 1.1 at ions arising i'rorn

the p r o d u c t s o f 11i e r e a c t i o n :

The signals generated are transmitted through a preamplifier

(Canberra, 2005), an amplifier (Canberra, 2012), and then

to a ratemeter (Ortec, 449), which is connected to a chart

recorder (Hewlett packard, 680). Even rapid changes of the

neutron flux can be read immediately from the chart recorder,

as the time constant of the ratemeter is preset at 0.1 sec.

The counting efficiency per unit volume of the phosphor

is 50% for thermal neutrons (Nuclear Enterprises, 1975).

In some cases, the efficiency of such a crystal may reach

100% (Nicholson and Sue1ling, 1955) (Fig. 3.8).

3.5.2 Electronic Relay Unit

A simple electronic relay was constructed (Motorola,

1976) to initiate the automatic'start of the single-

channel analysers, the scalers, and the multichannel

analyser. Adjustment of delay time (which equals 1.1 x

time constant of the RC circuit) can be done easily through

the change of a resistor andor a capacitor. This is a

monostable type relay, and the single-channel analysers,

scalers, and multichannel analyser will continue counting

once they are started until their preset time is up. When

the irradiation time registered by the programmable electronic

time is up, this relay starts its function. The accuracy

of this relay depends on the electronic stability of the

resistor FL and capacitor Ck , and is about ± 0.03 sec. A

block diagram of the relay circuit is shown in Fig. 3.9,

and the response c 1 iarac terist ics of 111e re 1 ay is ske 1:ched

in Fig. 3.10.

3.6 St andard Samp1es

In these experiments, the major-element composition

of jades was determined by using the comparator method,

which is often employed in activation analysis. In order

to obtain accurate results, the standard should have the

same geometry and have the same matrix as the sample. Under

these conditions, both the standard and the sample will

have the same neutron flux depression and gamma-ray

attenuation, and the same counting efficiencies.

Since standards of the mentioned properties were

not available, we had to prepare the matrices. For this

study, they needed to be of oval shape, of high density(S.G.

2.7 v 3.3), and contribute no interfering effects to the

measurement of the activated elements of interest. Many

materials and chemicals were tried, for example, polvethene,

calcium sulphate, etc. They either had too low a density

or they did not form a hard body that would resist deformation

during analysis. We then tried to use chemical mixtures,

and calcium sulphate mixed with lead oxalate turned outV

T h i c k n e s s o f C r v s t a 1'

F i s;. 3.8V.-1

Calculat eci SIow-Neut ron Detect ion Ef f ic .iency of Lil(Eu)

Crysta.I _as a Function of the Thickness of the

Crystal

- f

r- .

1 f H O A 1Rch If

Y(

I|500

Y()1(

3V rela

to controlswitch

Fig. 3.

M on net ah 1 e Re lav Circuit Diagrai

00 0 00 0 he

rH bD d dO bD 0 0I—I H Co i—!

0 a oFh 0 b

0C)

i—! 0d 0 fcuC

rH 0 0 d(J -H Cb 0I—I bD 0 i—I

•H 0 OQ O

00 O 00 0 CD

CO tfj d dCJ aD 0 0I—I i—i d i—I

0 0. OEH r 0

Or ')

rH (Id 0 h.

C3 0 0 KO H pH I-H bD 0 r-

•H 0 CQQ

0of.

d drH 0

« o

5

5

5

t i m e

t i rn e

t i m e

20

'U-0 Pi sec

t i m e

r:

-1 imr

0 P) sec

Fig. 3.10

response Characteristics of the IC ' s and Re 1 aj

to be the best matrix. Alter adding the desired elements,

for example, si I icon or aluminium powder, and an adequate

amonut of water to them, we could make standards of any

shape we desired. The mixture becomes hard as it dries.

No interference reactions occur, and the specific gravity

of this standard ranges from 2.3 to 2.9, depending on the

mixing ratio of the chemicals and the amount of water added

to it. The mass of the desired element, M, is calculated

f r om the equation:

where M is amount of element added, M is the mass of thee m

matrix, and M is the mass of the standard after drying.

This relation holds when the mixture is homogeneous, and

the water of crystallization of CaSCR is taken into account.

Repeated experiments showed the existence of large

fluctuations in M (greater than 25%), probably due to non-

homogeneity of the mixture.

Shifting to another method, we used pure silicon

wafers, pure aluminium (99.9%), and pure magnesium (99.5%)

metal for the standards. They were made to have a shape

and thickness similar to the jade samples. Since iron was

only available in powder form, special capsules of length

2 cm and diameter 4 mm were made, to hold the iron powder.

This standard was made in such a way that the length of the

iron column inside the capsule did not exceed 5 mm. Results

showed that these standards worked very well.

R(3su 11s and Discuss i on

Over 100 different specimens of jades were analysed,

in this study, via 14 MeV neutron activation.. The results

are tabulated in Tables 4.1, 4.2, 4.3, and 4.4t and

typical Nal(Tl) pulse-height spectra for the various kinds

of jades are shown in Fig. 4.1 to Fig. 4.8. Since all

these jades were on loan from a jewellery, there was just

enough time for each piece of jade to be analysed twice.

No large fluctuations in the percentages of the major

constituents (for example, silicon, aluminium, magnesium,

and iron) for different jades of the same species can be

observed from the data. In total, 24 jadeites were analysed,

resulting in standard deviations of the means for aluminium

and silicon of ± 9.1% and 7.6%, respectively (Table 4.1).

Of course, jades from different places may contain slightly

different concentrations of impurities, and hence exhibit

slightly different chemical compositions. Together with

experimental error, these differences in chemical compositions

lead to the deviations observed. The main experimental

error lies in positioning samples at the designed place

for counting. Repeated tests showed that this variability

ranged from ± 3% to ± 7%.

Pure jadeite crystal has the chemical formula „

4 aA 1 ( S iO 3 ) 2 . Ow i n g to interferences contributed b y the

intense 1.779 MeV photopeak of 28A1 from silicon, the short-

Table 4.1

Results for the Activation Analysis of Jadeites

Sample

I Mass, in 2 iN'O.[Mass of Si . in 2 7 o f S i Mass of A1. in a T of A1

0.9360

0.6483

0.6084

0.3620

0.7087

0.7416

0.5941

0.5705

0,6606

0.7094

A 1

A 2

A 2

A 4

A 5

A 6

A 7

A 8

A 9

A10

0, 216

0. 193

0. 176

0. 123

0. 194

0. 22 3

0. 146

0. 171

0. 175

0. 216

0. 014

0. 011

0. 008

0. 008

0 . 012

0. 013

0. Oil

0. 012

0. 009

0. 009

23 . 04

29. 70

28. 89

OO O QO _) • O O

27. 38

30. 10

24.63

29. 97

26 . 53

30. 41

1. 45

1. 70

1.23

2 . 25

1. 64

1. 74

1.85

2 . 13

1. 40

1. 34

0. 096

0. 086

0. 074

0. 048

0. 078

0. 090

0. 056

0. 074

0. 076

0 . 090

0. 005

0. 004

0. 003

0. 003

0. 004

0. 005

0. 004

0. 005

0. 004

0 n n 0

10.25

13 . 27

12 .21

13.20

10. 96

12 . 10

9 . 45

13 . 03

11 . 53

19 7R

0. 59

0.61

0.43

0. 83

0. 52

0.63

0. 59

0. 86

0.57

0. 38

Table 4.1 (continued)

Sample

Mass, in g No.

Mass of Si, in g % of Si Mass of A1, in cJ ro% o f A1

0.7521

1.0904

1.0561

1.3816

1.0830

1.3702

0.7681

0.8640

0.9667

1.2698

0.8409

1.2164

1.6864

0.8492

All

A 1 Qr-L

A13

A14

A15

A16

A17

A18

A19

A20

A21

A22

A23

A24

0. 187

0. 326

0. 312

0. 397

0. 298

0. 399

0. 234

0. 240

0. 277

0. 334

0. 234

0. 325

0 . 478

0. 240

0. 010

0. 023

0. 018

0. 024

0. 019

0. 027

0. 014

0. 013

0. 02 0

0. 017

0. 014

0. 017

0. 040

0. 013

9 a c q

9 Q C) 9

29. 51

28 . 71

2 7 . 52

29. 11

30 . 52

28 . 29

28. 70

26 . 27

27. 8 7

26 . 71

28 . 35

28 . 29

1 O A_L . o 9

2 1 9

-i f~ —_l . b t

1 . 77

1. 78

2 . 00

1. 86

1 . 57

2 . 08

1. 36

1. 68

1.41

2 . 36

1. 57

0. 071

0 . 12 8

0. 112

0. 159

0 . 121

0. 168

0. 093

0 . 090

0. 112

0. 136

0. 097

0. 132

0. 187

0. 090

0. 003

0. 009

0. 005

0. 008

0 . 0 0r

0. 010

0 . 005

0 . 004

0 . 007

0 . (.J 0 o

1. 005

0 . 0 0 7

0. 012

0 . 004

9.45

11.75

10.59

11 . 47

11. 17

12 . 26

12 .15

10 . 59

11 . 5 7

10. 71

11 . 56

10. 88

11 . 08

10 . 5 9

0 . 36

0. 78

0.48

0 . 58

0.68

0. 70

0. 70

0 . 47

0 . 5

0 . 39

0 . 63

0. 56

0 . 69

0.47

Mean : 27.90 ± 2.12 Mean : 11.49 ± 1.05

Table 4.2

Results for the Activation Analysis of Taiwan Jades (Nephrites)

S anin 1 e M n e o r

S i . 1 id q-1 !class .in si No.

% of SiMass of

Mr, in r% of Mg

Mass of

F e, in r% of Fe

riclS S OI

A1, in g% of A1

0.28210.016

0.21010.011

0.21110.013

0.25110.018

0.217+0.017

0.21410.008

0.20210.012

0.220+0.010

0. 193+0,010

0.19 710.DQn

30.00+1.74

25.3411.34

25.4311.55

30.99+2.22

28.09±2.14

29.77H. 10

27.2111.62

27. 1H1.23

28.30il.47

9 7 9 n +1 9 q

0. 12210.006

0 . 10 310. 0 0 6

0.089-0.006

0.09910.006

0.09910.005

0. 12910.012

0.08210.005

0.10010.006

0.09610.006

0.09110.004

12.9810.65

12,4110.77

10.7U0.75

12. 1810.69

12.8310.63

17.9311.66

11.0510.67

12.3210.74

14.0810.88

12.6110.55

0.01610.010

0.01410.009

0.011+0.010

0.01610.012

0.013+0.010

0.0090.008

0.012+0.009

0.019-0.013

0.009+0.009

0.015+0.011

1.65H. 15

1.6611. 12

1.3011.17

2.0211.43

1.6411.34

1.2411.13

1.6111,22

2.3211.65

1.3811.37

2.0811.42

0.00540.0002

0

0.00210.0005

0.002=0.0003

?

0.00310.0006

0.00210.0003

0,00110.0004

0.00310.0006

0.002=0.0005

0.0510.02

o

0.27=0.06

0,2510.04

0,3710.08

0.2710.05

0,1210.05

0.4410.09

0,2810.07

Mean: | 27.9501.88 Mean: i 12.9112,00 Mean: j 1.6910.35 Mean: ! 0.26=0. 12

Table 4.3

Results for the Activation Analysis of Australian Jades

Sample

Mass, in gNo.

-Mass of Si, in g % of Si Mass of A1, in g % of A1

0.5425

0.6496

0. 6594'

0.6416

0.8377

1.2014

0.4670

0.4059

0.6499

0.8382

C 1

C 2

C 3

C 4

C 5

C 6

C 7

C 8

C 9

C10

0.247± 0.017

0.322± 0.019

0.300± 0.016

0.279± 0.016

0.379± 0.026

0.527± 0.036

0.203± 0.012

0.198± 0.010

0.318± 0.021

0.378± 0.028

45.53± 3.08

49.58± 2.99

45.50± 2.36

43.51 ±2.51

45.25± 3.08

43.83± 3.02

43.49± 2.65

48.82± 2.36

48.93± 3.21

45.10± 3.29

0.002± 0.0002

0.005± 0.0004

0.004± 0.0005

0.002± 0.0002

0.008± 0.001

0.008± 0.0006

0.002± 0.0004

0.005± 0.0006

0.007± 0.0005

0.007± 0.0009

0.33± 0.04

0.77± 0.05

0.65± 0.09

0.37± 0.03

0.93± 0.12

0.68± 0.06

0.38± 0.08

1.63± 0.20

1.02± 0.07

0.79± 0.10

Mean: 45.95± 2. 3z Mean: 0 76± 0 39

Table 4.4

Results for the Act 1vat ion An a1y sis of Agate s

Sample

Mass, in gCD SI (),Photopeak Counts of Si Mass of Si, in g % of Si

0.4641 D 1 66578 ± 3222 3.230 ± 0.014 49.49 ± 3.10

0.3477

0.6396

0.5578

0.4265

D 2

D 3

D 4

D 5

50733 ± 1936

93260 ± 3624

79284 ± 3329

63352 ± 2909

0.175 ± 0.009

0.322 ± 0.017

0.274 ± 0.015

0.219 ± 0.013

50.34 ± 2.65

50.30 ± 2.68

49.04 ± 2.76

51.24 ± 3.09

Mean 50. 08 0.85

' i j !(».I cmj ) , |); M11

Ai (n , p) 2 7Mg

0.84 MeV

1250

1000

750 l

500

250'

0 0-5 10 1-5 -20

0.511 MeV

2nd escape;

of

1-78 MeV

0,75 7 MeV

Al (n,p) Mg

1.01 MeV

- 9 c 2 9 A 1S i (n, p) A1

1st escape of

1.78 MeV

1.28 MeV

' Si(n,p)2 A1

1.78 MeV

Gamma Energy (MeV)

Fig. 4.1

Typical Spectrum of Jadeite with Delay Time 20 sec., Counting

Time 2000 sec.

T—!0Jr~I I't-00r-4

CJ

0)cx

COAJr—3O,

o

1000

750'

500

250f

i—!0Cr-1-HCu

o

r.r-1(1)rrr—-

W4-3r(3o

o

oN 1° 1-5 f 20

G amm a En e r g y (M e V)

Fig . 4,2

Typical Syectrum of Jadeite with Delay Time 45 min.,

C') u n t i n g T1 rn e 2 0 0 0 sec .

24 IJg ( n , p ) 2 4 Nil

2 7 A1 (n , a )2 'fNa

1.46 IvleV

1.37 MeV

28Si(n,p)Z0Al

1.78 MeV

Si (n, p) JA1

1st escape of 1.78 MeV

2nd escape ''Fe (n , p) 5 6Hn.of 1.78 MeVA f

0.77MeV 0.84 MeV 1.28 MeV

1250

1000-

750-

500

250

OL0-5 10 15I v... 2 0'v,

Gamma Energy (MeV)

Fig. _4. 3

Typical Spectrum of Taiwan Jacle with Delay Time 20 sec.,

i—i0)GGdr~]

o

GCD

CO)

r—r iPAMO

o

Countin g Time 2 000 sec._

7;

JJ

CI o mmo 1 ra r~ r» rv 7 ( 1;T r

T7» -4 I

Typical Spec:!, rum o f Taiwan Jade with Delay Time 45 mi

f i - n- Ti O HOH

r

f{

2nd escape oi

1. 78 Me Vi

0.77 MeV

2 9 Si (n, p) 2 9A;

1st escape of

1.78 MeV

1.28 MeV

Si (np) ~ A1

1.78 MeV

1 25C

10CX

750

500

250

0 „n. 10 1-5 _2 0

_ a...

G amm a En e r g y (M e V)

Fig. 4.5

Typical Spec trum of Austra1ian Jade, wit h Pe1ay Time

20 sec., Counting Time 2000 sec.

i—CDr~r»—iGdr~|

o

c.0)o

in+-—',—.»

oo

loooi-

750

500

250 i

ol1-0 1-5 2-0

G amm a En e r g y (M e V)

Fia. 4.6

Typica1 Spectrum of Australian Jade with Delay Time 45 min,

Coun ~l i n g Ti me 2 000 se c_._

1.46 MeV

i—!0£££

O

rrHG)

GO4-r—1r—'

oo

IQOOj-

750ii

5oo(

250

0 0-5L_

2nd escape

of

1.78 MeV

J I

,0.77 MeV

9 9 Si(n,p)9A1

1st escape

of

1.78 MeV

1.28 MeV

Si(n3 p)z'A1

1.78 MeV

1-0 1-5 20

Gamma Energy (MeV)

Fig. 4.7

Ty p i c a 1. Sp e e t r urn o f Ag ate wit h De I ay Time 20 sec.,

Coiin t i ng Time 2 000 sec.

i—I0)Gt—I

cdrCO

UCDp(

W4-r-j2O

u

1000

75'

)1

i25

0 .10

i1-E '2•C

.........-

1. 46 MeV

y vr

G 11 nun a En e r ay (MeV)

Fig. 4.8

Typical Spect rurn of_ Aga te with De1 ay Time 45 mir

Counting Time 2 000 sec.

i—10c—

d

O

U0n

0-Pr—r4—1r—'f—'Q

Lived (It = 11.0 see.) radionucli do , ' F , [X'fduccd hv«

sodium act ival ion, could no I bo del.orm i nod. The average

percent content by weight of aluminium in jadoite was found

to be 11.49% , d 1 f f'or i ng f rom the va 1 ue ca 1 cu.J ated f rom t:he

chemical formula, NaA L ( SiO 3) 2 by 14.0%, whereas the silicon

percentage found differs from the formula value by only

0.6%.

U s 11; 111 y the inone t a 1 y va J, ue o f a j ade i t e s pe c i men

is based upon its size, its transparency, and its green

colour. In this experiment, we also tried to detect one

interesting element, chromium, which is present in jades

as an impurity, and which is believed to be the origin of

the green colour (Savage, 1964), (Switzer, 1979). However

detection difficulties arise, since the photopeak of the

1.434 M e V g a m j 11 a o I 111 e r a d i o n 11 c 1 i d e , 5:' V , p r o d u c e d f r o in

activation of chromium, is situated on the Compton edge

of the 1.779 MeV gami 11 a of 2 8A1, 1;he 1 nduced radioactive

nuclide of the activated silicon. Furthermore, the actual

amount of chromium is extremely small. As a result of

these difficulties, the content of chromium could not be

determined. Methods, including peak stripping of the

spectrum by graphical and computer techniques (Kowalski,

1968), were attempted, but chromium could not be found,

with acceptable statistical accuracy. Nevertheless, from

the experimental spectra, jadeite can easily be .identified

from other jades by its aluminium peaks.

Taiwan jade is a kind o 1' nephrite, and its value

is much lower than thai ot jadeito. Noph r i I e lias the

chemical compos i 1 i on , da 7 (Mg , Fe ) 5 (Oh ) 2 ( Si 4O 1 1 ) 2 i • • , i t

is a silicate ot calcium and magnesium with the presence oi

some iron. In these experiments, 10 Taiwan jades were

analysed (see Table 4.2), an d the composition s foun d were

28.3% for silicon, 13.2% for magnesium, 1.6% for iron, and

0.2% for aluminium, all in percent by weight. The standard

deviations f o r I, he j)e 1 ce 111 con t en t o f t he var ious e 1 erne 111 s

were ± 8.5% for silicon, ± 18.8% for magnesium, ± 17.6% for

iron, and ± 57% for aluminium (Table 4.2).

The gamma-ray spectra o f activated Australian jades

and agates showed that no detectable amounts of other

elements than silicon wore present (Tables 4.3 and 4.4).

These beautifu1 green sto 11 es are usua11y sold under the

name j ade , t he i 11 cost be i 11 g much 1 ower t han j a.de i t e an d

a little higher than Taiwan jades. From the present

experiments, the percent content by weight of silicon in

Australian jade and. agate were found to be 45.95% and

50 .08%, respect i.ve 1 v . A list ing of the ex.perirnent a 1 resu 11s

for compos ition s o f j a deit e an d n e p hrit e b y o t he r wor kers

(Pearce, 1971) is given in Table 4.5 .

The overa 1. L ana. 1.ysing t ime f or one samp 1 e of j ade

was about 90 minutes. The irradiation time was 10 minutes,

so that t he long ha 1 f -1 i f e nuc 1 i des cou 1 d b(3 gener at ed

appropriately. Two decay times, 20 seconds and 45 minutes,

were employed. The counting time was chosen to be 2000

seconds, so that the counting statistics were better for

Table 4,5

Chemi cal An a 1 vs i s of Jacle si

(Percents bv Weight)

Si A1 Fe Mn«r

-g CarIV H

Nephrite

New Zealand

Switzerland

Silesia

C h i n a

A1 a s k a

B cd w e n i t e

J cXcl Gi. ~LG

Switzerland

China

Mexico

26 . 68

26 . 12

26 . 26

27.29

26 . 42

19. 89

27 . 28

27.64

27.63

0.7-

0. 1L

0.8-

0. 24

0.8-

LI. 95

L2 . 14

L3 . 17

2 . 84

4 . 44

4. 20

1. 41

5.84

0 . 9o

1.28

1. 21

0. 86

0.06

0. 66

0. 16

0. 39

13 . 47

13 . 06

14 . 79

15 . 36

12 . 71

25 .66

0.77

0.69

0. 31

9 . 60

8 . 68

5.73

9 . 25

9 . 16

0.46

2 .21

1 . 92

0.69

i

9.45

9.58

9 . 04

| 0.41

0. 26

0.23

0.39

0. 45

0. 03

0. 16

1.46

0. 02

0 . ID

Results taken from Pearce, 1971.

69

the 0.84 MeV and 1.37 MeV peaks. In fact, if the attempted

determination of chromium is omitted, the total time of

analysis can be shortened to about 30 minutes (10 minutes

irradiation, 1. minute decay, and 10 minutes counting) per

jadeite sample.

The analysed jades were chosen to have approximately

oval shapes arid similar sizes, so that the counting

efficiencies for the p ahotopeaks being measured were

essentially constant. Also, each jade sample had a small

thickness( 3 mm 0. 5 mm), and since the absorption

coefficients for the various jades do not vary much, the

total attenuation correction factor can be considered to be

5/10 for all the jades. For aluminium, the intensity of the

0. 844 MeV peak was found to be 2.4 times larger than that

of the 1. 014 MeV peak, hence only the 0. 844 MeV photopeak

was used in evaIuatitry' the alull]iniuIll coneentrations if

jades of shapes other than oval shapes are used, the

corresponding standards can be made by using the same

calcium sri].phaLe/load oxalate mixture metthod. The

accuracy of this method should be better for larger sample

sizes, as the mixture would then be more homogeneous than

for smaller samples.

Chapter V

Conclusions

From the experiments, we see that the two gamma

radiations of 2;Mg are not simultaneous events, which

agrees with that deduced from the recent decay scheme of

27Mg. The internal conversions of these gamma radiations,

0.844 and 1.014 MeV, are extremely small: 3.65 x 10 5

and 4.67 x 10 5, respectively. The accuracies of these

values depend on the accuracy of the published data used.

The ratio of the intensity of the 0.844 MeV photopeak of

27Mg to that of its 1.014 MeV photopeak was found to be

2.41 ± 0.17, which differs from the accepted value of

2.33 by 3.4%. With the information provided in this paper,

we hope that we have established a nondestructive test

for the authenticity of jades. To our regrets, however,

the differences in the greenish colour, which is considered

to be the most valuable characteristic of jade, could not

be distinguished using this method.

71

Appendix ( A )

The Characteristics of Gems

In Tables 1 to 4, the gem minerals are classified

according to crystal. system, hardness, specific gravity,

and optical PropeltieS.

Table 1

Crstal System

diamonds, gold, garnetcubic system

zircon, rutiletetragonal system

corundum, quartz, tourmaline, berylhexagonal system

topaz, enstatile, chrysoberylorthorhombic system

nephrite, jadeite, malachite, serpentine.monoclinic system

diopside

turquois, cyanite, rhodonitetriclinic system

opal, amberamorphous

72

Table 2

Hardness

HardnessHardness MineralsMinerals

1.0 6.5nephritediamond

6turquois9corundum

5.5-6.5opal8.5chrysoberyl

5-6rhodonite8topaz

5-6diopside7.5-8beryl

3.5coral7. 5zircon

3.5malachite7-7.5tourmaline

2.5-3.57 pearlquartz

2.5-4serpentine6.5-7.5qarnet

2.5-36.5-7 goldjadeit

2-2.5amber6-6.5rutile

73

Tab e 3

Spccitic Gravity

S. G,MineralsS. G.Minerals

3.2diposide4.0-4.8zircon

3.1tourmaline4.2rutile

3.0nephrite4..0corundum

2.7beryl4.0malachite

2.73. quartzchrysoberyl.

2.7turquois3.5diamond

2.6serpentine3.5topaz

2.6pearl3.4epidote

1.9-2.3opal3.3jadeite

74

Table 4

Optical Properties

BiOptical

Character relringenceIndices of RefractionMinerals

n2.417diamond

0,287£2.903w2. 616rutile

0.057El.988wl.931zircon

0.254al.655131.875malachite yl.909

0O08El.760wl.768corundum

0.051yl.780a1.729epidote 131.763

0.010al.747chrysoberyl 131.748 Y1.757

0.007al.716 y1.723epidote 131.719

yl.726 0.024a1.702diopside 131.708

0.013y1.667al.65401.659jadeite

0.016£1.670wl.654tourmaline

-yl.627 0.008al.619topaz 131.620

y1.650 0.040(1.610tourquois 131.620

yl.632 0.026(1.606nephrite 131.620

0. 008E1.590wl.598beryl

0.011Y1.557a1.546r31.550serpentine

0.009El.553w1. 544quartz

nl.460opal

75

Appendix (B)

occurrence of Jades

jadeite

nephrite

76

Appendix (C)

Colour plates of jades

A6A2

A21

A 20

AlA 13

A 22A 5

Plate 1

jadeite

77

A 24 A 23

A 17 A 18 A12 A 14

A 16A 19

A 9

A 11

A 7

A 8

A 15

A 10

A 4 A 3

Plate 2

jadeite

78

B5

B3

B4 B2

B1

B 10

B6

B8

B7 B9

Plate 3

Taiwan jade

Cl

C5

C4C2

C3

C 10

C6

C9

C8C7

Plate 4

Australian jade

D2

D5D4

D3D1

Plate 5

agate

81

References

1.Bleuler, E. and G. J. Goldsmith, 1960, Experimental

Nucleonics, Chapter 22 (Holt, New York).

2.Canberra Industries, Inc., 1979, Instruction Manual for

Coincidence Analyzer 2040.

3.Chuang, L.S., et al., 1975, Journal of the Chinese

University of Hong Kong, 3, No. 2, 459.

4.Chuang, L.S., et al., 1979, Nucl. Instr. and Meth., 160,

487.

5.Ciuffolotti, L. and F. Demichelis, 1962, Nuclear Phys.,

39, 252.

6.Crdtmann, G., 1976, Neutron Activation Tables (Verlag,

New York).

7.Gijbels, R., et al., 1968, Anal. Chim. Acta., 43, 190.

8.Garrec, J.P., 1969, C.R. Acad. Sc. Paris (Series D), 3021.

9.Girardi, F., et al., 1964, Anal. Chem., 36, 1588.

10.Grieken, R.V., et al., 1968, Anal. Chian. Acta, 43, 199.

11.Gunnink, R. and A.W. Stoner, 1961, Anal. Chem., 33, 1311.

12.Iddings, F., 1964, Anal. Chim. Acta., 31, 207.

13.Kaman Nuclear, A-711 Neutron Generator Instruction Mannual.

14.Kaman Nuclear, 1969, Technical Bulletin No. 109.

15.Kokta, L., 1973, Nucl. Inst. and Meth., 112, 245.

16.Kowalski, B.R. and T.L. Isenhour, 1968, Anal. Chem., 40,1186.,

17.Lazar, N.H. et al., 1956, Nucleonics, 14, 452.

18.Lederer, C. M. et al., 1978, Table of Isotopes, 7th Ed.)

(John Wiley, New York).

19.Liddicoat, R.T., 1969, Handbook of Gem Identification

(Gemological Institute of America, Los Angels).

82

20. Mason, B., 1952, Principles of Geochomistry (Wiley and

Sons, New York).

21. Meinke, W.W., and R. 11. Maddock, 1957, Anal. Chem. 29,1171.

22. Motorola Semiconductor Products Inc., 1976, Semiconductor

Data Library, Series B, Vol. 6, Chapter 8.

23. Mott, W. E., 1965, Anal. Chem., 37, 1338.

24. Nargolwalla, S.S., 1968, Anal. Chem., 40, 666.

25. Nargolwalla, SS. and E. P. Przybylowicz, 1973, Activation

Analysis with Neutron Generators.

26 Nicholson, K.P. and G.F. Snelling, 1955, Brit. J. Appl.

Phys., 6, 1-04.

27. Ophel, T.R. and B.T. Lawergren, 1963, Physics Letters,6,230.

28. Pearce, G.L., 1971, The Story of New Zealand Jade.

29. Quittner, P., 1972, Gamma-Ray Spectroscope (IIi1_ger, London).

30. Sakikawa, N., 1968, Jade (Japan Publications,Tokyo).

31. Santos, G.G. and R.E. Wainerdi, 1968, J. Radioanal. Chem.,

1, 509.

32, Savage, G., 1964, Chinese Jade (Adams and Mackay, London).

33, Siegbahn, K. and R.E. Bell, 1965, Alpha, Beta and Gamma

Spectroscopy, Vol. 1 (North-Holland :Publishing Co.,Arnsterdam).

34. Stark, D.S., 1971, Services Electronics Research Laboratory

(England) Technical Report No. M296.

35. Switzer, G.S., 1979, Gemology (Wiley, New York).

36. Turner, S.E., 1956, Anal. Chem., 28, 1457.

37. Volborth, A., and II.E. Banta, 1963, Anal. Chem. 35, 2203.

38. Weast, R. C., 1970, Handbook of Physics and Chemistry

(The Chemical Rubber Co., Cleveland).

83

39. Webster, R., 1975, Gems (Butterworth„, London)

40. Wing, J., 1964, Anal. Chem., 36, 559.

41. Wong, K.C., 1975, Thesis Chinese University of Hong Kong.

42. Wood, D. E., 1972 Advances in Activation Analysis, Vol. 2,

273.