Citation
Development and comparison of Compton suppression techniques for low-level radio-nuclide analysis

Material Information

Title:
Development and comparison of Compton suppression techniques for low-level radio-nuclide analysis
Creator:
Markwell, Frank Russell, 1948-
Publication Date:
Copyright Date:
1972
Language:
English
Physical Description:
xii, 117 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Coincidence ( jstor )
Crystals ( jstor )
Electronics ( jstor )
Energy ( jstor )
Photons ( jstor )
Photopeaks ( jstor )
Radio spectrum ( jstor )
Radionuclides ( jstor )
Scintillation counters ( jstor )
Water samples ( jstor )
Dissertations, Academic -- Engineering Sciences -- UF
Engineering Sciences thesis Ph. D
Gamma ray spectrometry ( lcsh )
Radioactivity -- Measurement ( lcsh )
Scintillation spectrometry ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 115-116.
Additional Physical Form:
Also available on World Wide Web
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Frank R. Markwell.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
022838594 ( AlephBibNum )
14147955 ( OCLC )
ADB1009 ( NOTIS )

Downloads

This item has the following downloads:


Full Text











DEVELOPMENT AND COMPARISON OF

COMPTON SUPPRESSION TECHNIQUES FOR

LOW-LEVEL RADIONUCLIDE ANALYSIS











By

Frank R. Markwell


A DISS-ETATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE DEGREE 0? DOCTOR OF PHILOSOPHY IN
NUCLEAR ENGINEERING SCIENCES








UNIVERSITY OF FLORIDA


1972








ACKNOWLEDGEMENTS


The author wishes to express his appreciation to Dr. W.E.

Bolch, Dr. M.J. Ohanian, Dr. C.E. Roessler, Dr. W.H. Ellis,

and Dr. H.A. Bevis, members of his supervisory committee, for

their advice and assistance in the preparation of this disserta-

tion. He especially wishes to thank Dr. W.E. Bolch for the

suggestion of the dissertation subject and for his encouragement

and valuable advice. Special thanks are due Dr. C.E. Roessler

for his aid, helpful suggestions and constructive criticism.

Thanks are also expressed to Dr. W.H. Ellis for his helpful

suggestions and encouragement.

Special appreciation is extended to the the staff members

of the Nuclear Engineering Department and the Environmental

Engineering Department in the University of Florida, without

whom this work could not have been performed.

The author acknowledges the support by the Atomic Energy

Commission for the award of a Special Fellowship in Nuclear

Science and Engineering during his graduate studies and the

support by Florida Power Corporation.

In addition, the author wishes to thank Nuclear Physics

for the loan of the Ge(Li) spectrometer.













TABLE OF CONTENTS


Page


ACKNOWLEDGEMENTS ..........

LIST OF FIGURES ...........

LIST OF TABLES ............

ABSTRACT ..................

Chapter


... *. *..#..


* . . .


I. INTRODUCTION ...... ... ..... .....

II. THEORY ..................... ......

A. Ge(Li) Detector ........................

B. Anticoincidence Guarding ...............

C. Active-Sample Guarding .................

III. EXPERIMENTAL SYSTEM ........................

A. Primary Detector .......................

B. Guard Detector ....................

C. Sample Detector ........................


D.

E.

IV. Nal

A.

B.

V. Nal

A.

B.


Counting Room and Shielding ............

Electronics ....................... ..... ...

(Tl) SYSTEM .............................

Configuration ..........................

Electronics ............................

(Tl) PERFORMANCE ........................

Detector Background ....................

Efficiency .............................


iii









TABLE OF CONTENTS -- continued


Page

Chapter -- continued

C. Minimum Detectable Activity ............ 42

D. Peak-to-Compton Ratio .................. 44

E. Complex Spectrum ....................... 44

VI. Ge(Li) SYSTEM PERFORMANCE ................ 51

A. System Background ...................... 51

B. Efficiency ................... ...... ... 55

C. Minimum Detectable Activity ............ 55

D. Peak-to-Compton Ratio .................. 59

E. Complex Spectrum ...................... 59

VII. GUARDED SYSTEM PERFORMANCE ................. 62

A. System Background ...................... 62

B. Guard-Detector Efficiency .............. 64

C. Photopeak Reduction .................... 6g

D. Minimum Detectable Activity ............ 69

E. Peak-to-Compton Ratio .................. 69

F. Complex Spectrum ....................... 73

VIII. ACTIVE SAMPLE SYSTEM ....................... 75

A. Efficiency ............................. 75

B. Minimum Detectable Activity ............ 75

C. Peak-to-Compton Ratio .................. 76

D. Separation of Decay Modes .............. 76

IX. COMPARISON OF SYSTEMS ON ENVIRONMENTAL

SAMPLES .................................... 82








TABLE OF CONTENTS -- continued


Page


Chapter -- continued

A. Standard Samples .......................

B. Air Sample .............................

C. Water Samples ................... ..... ..

D. Solid Samples ..........................

X. EVALUATION OF DETECTION SYSTEMS ............

A. NaI(T1) System ....................... ...

B. Unguarded Ge(Li) System ................

C. Guarded Ge(Li) System ..................

D. Active-Sample System ...................

XI. CONCLUSIONS ...............................


Appendices

A. ELECTRONIC CIRCUITS ...........

B. GUARD-DETECTOR CHARACTERISTICS

C. SPECTRUM ENERGY CALIBRATION ...

D. EXAMPLE OF DIGITAL DATA .......


.............



.............
OOOOOeooooo


LIST OF REFERENCES ................ ....... ........

BIOGRAPHICAL SKETCH ..............................


100

105

110

113


115

117











LIST OF FIGURES


Figure Page

1. Effect of.Compton suppression ................. 7

2. Guarded Ge(Li) system configuration ........... 11

3. Ge(Li) detector side view ................... 12

4. Ge(Li) detector frontal view ................ 13

5. Ge(Li) detector with extension collar ......... 15

6. Guard-detector photomultiplier tube ........... 17

7. Guard-detector with photomultiplier mounted ... 18

8. Sample port mounted in guard detector ......... 19

9. Ge(Li) detector positioned in sample port ..... 20

10. Sample container with photomultiplier mounted 22

11. Sample port for standard technique ............ 23

12. Sample port for active-sample technique ....... 24

13. Counting room and entrance .................... 25

14. Counting room ................................. 26

15. Photomultiplier shielding ..................... 28

16. Guarded Ge(Li) system electronics ............. 29

17. Ge(Li) system analyzer ........................ 30

18. NaI(T1) detector configuration ................ 34

19. NaI(Tl) crystal in lead shield ................ 35

20. NaI(T1) system in counting room ............... 36

21. NaI(T1) system electronics .................... 38

22. Background for NaI(T1) ....................... 40


vi









LIST OF FIGURES -- continued


Figure Page

23. NaI(T1) photopeak efficiency .................. 43

24. Minimum theoretical detectable activity for

NaI(Tl) ....................................... 45

25. Peak-and-Compton spectrum for NaI(T1) ......... 47

26. Peak-to-Compton ratio for NaI(T1) ............. 48

27. Complex NaI(T1) spectrum ...................... 49

28. Counting room background ...................... 52

29. Ge(Li) background with Hg shielding ........... 53

30. Ge(Li) photopeak efficiency ................... 56

31. Minimum theoretical detectable activity for

Ge(Li) ..................................... 57

32. Peak-to-Compton ratio for Ge(Li) .............. 60

33. Unguarded complex Ge(Li) spectrum ............. 61

34. Unguarded Ge(Li) background ................... 63

35. Guarded Ge(Li) background ..................... 65

36. Guard-detector efficiency ..................... 67

37. Minimum theoretical detectable activity for
guarded Ge(Li) ................................ 70

38. Peak-to-Compton ratio for guarded Ge(Li) ...... 72

39. Guarded complex Ge(Li) spectrum ............... 74
40. Active-sample minimum theoretical detectable

activity for Ge(Li) ........................... 77

41. Active-sample peak-to-Compton ratio ........... 79

42. Separation of decay modes .................... 80

vii









LIST OF FIGURES -- continued


Figure Page

43. Zn-65 source spectrum ......................... 83

44. Complex standard-sample spectrum .............. 84

45. Air sample spectrum ........................... 86

46. Water standard spectrum ....................... 87

47. UFTR primary coolant water spectrum ........... 89

48. Soil sample spectrum .......................... 90

49. Oyster sample spectrum ........................ 92

50. Seaweed sample spectrum ....................... 93

51. Sample-detector electronics circuit ........... 101

52. Guard-detector electronics circuit ............ 102

53. Ge(Li) threshold electronics .................. 103

54. System coincidence electronics ................ 104

55. Guard-detector light collection ............... 106

56. Guard-detector background .................... 107

57. Threshold calibration ......................... 108

58. Ge(Li) calibration spectrum ................... 111

59. Ge(Li) calibration curve ...................... 112


viii











LIST OF TABLES


Table Page

1. Background Count Rate for the NaI(Tl) System .. 41

2. Theoretical Minimum Detectable Activities for

the NaI(T1) System ............ .......... 46

3. Background Count Rate for the Ge(Li) System ... 54

4. Theoretical Minimum Detectable Activities for

the Ge(Li) System .......................... .. 58

5. Background Count Rate for the Guarded Ge(Li)

System ................. ......... ......... .. .. 66

6. Theoretical Minimum Detectable Activities for

the Guarded Ge(Li) System ..................... 71

7. Theoretical Minimum Detectable Activities for

the Active-Sample Ge(Li) System ............... 78

8. Digital Data for Guarded Air Sample Spectrum .. 114












Abstract of Dissertation Presented to the Graduate Council
in Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy in Engineering

DEVELOPMENT AND COMPARISON OF COMPTON SUPPRESSION
TECHNIQUES FOR LOW-LEVEL RADIONUCLIDE ANALYSIS


By

Frank R. Markwell


December, 1972


Chairman: Dr. W. E. Bolch
Major Department: Nuclear Engineering Sciences


Optimum analysis of low-level radionuclides requires the

removal of Compton interference from the gamma-ray spectrum.

Significant Compton suppression can be achieved by employing

high-efficiency guard detectors to detect the scattered

gamma photons. However, high detection efficiency prohibits

significant absorption between the primary detector and the

guard detector, thus restricting the sample to ore of small

mass. When large-mass samples are used Compton suppression

is greatly degraded.

To alleviate this problem a new concept has been

developed and a system constructed, and evaluated which, in

effect, makes the sample part of the guard detector. Restricted

to optically clear samples, the system requires the addition

of a liquid scintillator to the counting sample. This










"cocktail" is viewed by a separate photomultiplier tube

and associated electronics which add its detected signal to

the guard signal and both are used in anticoincidence with

the primary detector. This scheme restores the performance

of the system to that of the "small-sample" Compton reduction

system. However, it also removes photopeaks due to simultaneous

(precedingwithin 200 ns) p or decay in the sample which

can also trigger the anticoincidence logic. Alternately, by

requiring coincidence between the sample-scintillator

detector and the primary detector, all electron capture and

metastable states are excluded, leaving only a spectrum of

photopeaks from short lived A* and f- decay daughters but

with partial Compton reduction. In short, this technique

yields two spectra for each liquid sample, a "perfect"

gamma spectrum of some excited states and a "not so perfect"

gamma spectrum of the remaining excited states.

The complete detection system employs a 6% efficient

(re 3x3 inch NaI(T1) ), 2.8 keV resolution Ge(Li) spectro-

meter as the primary detector arranged to extend 12 inches

into the center of a 31 inch diameter by 26 inch long container

of Pilot Chemicalts high efficiency mineral oil scintillator.

This scintillator, which acts as the guard detector, is

viewed by a 15 inch Fairchild K2128 photomultiplier tube

and both are shielded with 2.5 inches of mercury. The sample

container arranged axially around the Ge(Li) crystal and

viewed by a 2 inch diameter RCA 6655 photomultiplier tube,

holds 1.5 litres (607 scintillator). The electronics include








preamplifier, threshold detectors, and logic circuits to give a

timing uncertainty of less than 200 ns.

Comparison of this system with a 4x4 inch NaI(T1)

system and the unguarded Ge(Li) system shows it to have

both superior peak-to-Compton ratios and superior minimum

detectable activities for counting times as short as 40

minutes. In addition to maintaining the point source peak-

to-Compton ratio for large water samples, the active-sample

technique has the advantage of dividing the spectrum into

two spectra according to the decay mode of the radionuclides.


xii












CHAPTER I


INTRODUCTION


Gamma ray spectroscopy has been an important tool in the

analysis of radionuclides for many years. The development of

the NaI(T1) detector opened the field and since then vast

improvements in efficiency and reductions in interradio-

nuclide interference have been made. Improvements in detector

efficiency and resolution have reduced this interference but

not eliminated it. The remaining interference is mostly due

to the Compton continuum associated with photopeaks in the

spectrum.

Since Compton scatter in the detector is only a partial

energy release mechanism, a lower energy gamma photon may also

exit the detector. This characteristic has been used in
(1)(2)
suppressing the Compton continuum. The original systems

for such suppressions employed NaI(T1) as the primary detector

surrounded by a large liquid scintillator which detected the

Compton scattered photons leaving the primary detector and

used them to eliminate such counts from the spectrum. Thus,

this technique is denoted as anticoincidence guarding.

The development of the Ge(Li) detector has resulted in

greatly improved resolution capabilities and reduced effects

from the Compton continuum. The usefulness of such a detector




2

for low level radionuclide analysis has been shown previously
(3)
by Cooper et al. Soon thereafter investigators applied
(4)(5)(6)(7)
anticoincidence guarding to Ge(Li) detectors.

These systems, which suppressed the Compton continuum up

to a factor of 6, employed plastic scintillators as the guard

detector. Current investigators, using NaI(Tl) guards,
137(8)(9)
report a Compton reduction for Cs of up to 9 times,

and those using plastic scintillator guards report a reduction
(10)(11)
of 6 times.

Significant Compton reduction can only be achieved by

employing high-efficiency guard detectors. However, high

detector efficiency prohibits significant absorption between

the primary detector and the guard detector, thus restricting

the sample to one of small mass. When large-mass samples are
(7)
used, Compton suppression is greatly degraded.

To alleviate this problem, a new concept is proposed which,

in effect, makes the sample part of the guard detector.

Restricted to clear water samples, the system would require

the addition of a liquid scintillator to the counting sample.

This "cocktail" would be viewed by a separate photomultiplier

tube and associated electronics which would add its detected

signal to the guard signal and both would be used in anti-

coincidence with the primary detector. This scheme should

restore the performance of the system to that of the "small-

sample" Compton reduction system. However, it would also

remove many photopeaks due to simultaneous (preceding within

200 ns) A or B- decay in the sample which could also trigger

the anticoincidence logic. Alternately, by requiring coinci-





3

dence between the sample-scintillator detector and the

primary detector, all electron capture and metastable states

could be excluded, leaving only a spectrum of photopeaks from

short-lived and 8 decay daughters but with partial Compton

reduction. In short, this technique would yield two spectra

from each liquid sample, a "perfect" gamma spectrum of some

excited states and a "not so perfect" gamma spectrum of the

remaining excited states.

Presented here are the design details of a liquid

scintillator guarded Ge(Li) system with the active-sample

option, as mentioned above, and the evaluation of this system

for environmental radionuclide analysis with respect to their

relative advantages over a 4x4 inch NaI(T1) system.













CHAPTER II


THEORY


In this chapter the theory behind the active sample

anticoincidence guarded Ge(Li) spectrometer is presented.

Included are both the theory associated with the new active

sample technique and the theory of anticoincidence guarding.

The latter is presented since anticoincidence guarding is

still a relatively new technique and not widely known.


A. Ge(Li) Detector


Interaction of a gamma photon with the primary detector

can occur in one of several ways. These are photoelectric

absorption, Compton scatter, and pair production. For

germanium and typical gamma energies (<2MeV), only the first

two are significant. Photoelectric absorption occurs when

the gamma photon releases all of its energy in one interaction

with a bound electron, and has the greatest probability for

photon energies just above the ionization potential energy

of that bound electron. Thus,this method of photon energy

release is most significant for low energies and has
-3.5
approximately a (hO) probability dependance. Photoelectric

absorption transfers all of its energy to the bound electron

which in turn releases this energy by ionization and electronic






5
excitation of atoms in the immediate vicinity of the photon

interaction.

The other interaction of significance is Compton scatter,

which consists of a partial transferral of photon energy to the

bound electron and the subsequent scatter of the photon

(reduction in energy). This reaction has a (hO)-1 probability

dependance and thus is significant for energies higher than

those significant for photoelectric absorption. Again the

liberated electron from the photon interaction produces

ionization and electronic excitation of atoms in its immediate

vincinity.

Such interactions within the intrinsic region of a

semiconductor produce electron-hole pairs having lifetimes

long enough to allow their collection by an electric field.

The result is a charge collection which is proportional

to the gamma photon energy release within the semiconductor

crystal. A count rate-energy release spectrum (charge

collection) yields a peak corresponding to photoelectric

absorption and a continuum of counts corresponding to the

energy release in Compton scatter. Due to the physics of

Compton scatter this continuum is approximated by a plateau

from zero energy (no scatter) to a maximum energy signicantly

less (typically 100-200 keV) than the photopeak energy

(corresponding to the energy release in photon backscatter).

The ratio of probabilities of the two interactions depends

upon the photon energy and the detector material. For

germanium the photoelectric interaction and Compton scatter





6

are equally probable at 150 keV but at 1 MeV Compton scatter
(7)
is 100 times more probable.


B. Anticoincidence Guarding


The useful information from a Ge(Li) spectrum lies in the

photopeak energy and amplitude since the gamma photon energy

is characteristic of the radionuclide from which it originated

and the amplitude is proportional to the quantity present.

The Compton plateau is a continuum which, because of counting

statistics, is erratic and thus obscures other photopeaks which

may lie within this energy range. Anticoincidence guarding

suppresses the counting of scatter events thereby removing

this continuum and allowing recognition of small amplitude

photopeaks within the energy range (see Figure 1 ).

To achieve this Compton suppression, those gamma photons

which deposit only part of their energy in the primary

detector must be recognized. By placing the primary detector

within a high-efficiency detector, photons which are scattered

in the primary detector can be detected again in the guard

and thus recognized as being Compton scattered. These can

then be prevented from being recorded in the spectrum. To

obtain excellent Compton suppression, the guard detector

must approach 100% detection (not energy collection)

efficiency and photon absorption between the primary detector

and the guard detector must be eliminated. Since good

primary detection requires the radioactive sample to be

placed in the center of the guard detector just adjacent to

the primary detector, the sample itself can act as an























































































aqhli qunoo


0
U-









0


0


0
C,
&<













I

I
I





8

unwanted absorber. Thus excellent Compton suppression in

standard systems can be maintained only with small samples.

In addition, large samples inherently yield a lower peak-to-

Compton ratio due to Compton scatter in the sample before the

photon reaches the primary detector.

An inherent disadvantage of anticoincidence guarding is

the removal from the spectrum of those photopeaks associated

with cascade gamma decay (2 or more simultaneous photons)

and .+ decay whose excited daughter decays simultaneously

(yielding a simultaneous gamma decay photon and annihilation

photons).


C. Active Sample Guarding


To eliminate degradation of Compton suppression and

maintain excellent peak-to-Compton ratios, the sample must be

used as a guard detector also. This allows recognition of

sample-scattered photons which enter the primary detector

and primary detector scattered photons which are absorbed

back in the sample. The combined effects yield a peak-to-

Compton ratio for large samples which approaches that for

point source samples.

An inherent problem in this technique arises when a

radionuclide in the sample decays by emission of a P

particle to its excited daughter and then immediately

emits a gamma photon. In this case gamma photons enter the

primary detector at the same time that the "- particles (poly-

energetic) release their energy in the sample detector. Thus





9

the f particles can not be distinguished from the scattered

gamma photons by the sample detector. 3adionuclides having

excited daughter lifetimes greater than the timing uncertainty

of the detector system, and those radionuclides which decay

by electron capture, are not affected by this phenomenon.

To restore these short-lived excited daughters from f"

decay and remove the long-lived and electron-papture events,

coincidence signals can be required between the sample detector

and the primary detector, indicating #" decay with an

associated prompt gamma photon. This spectrum contains only

photopeaks from these prompt gamma decay states but maintains

only partial Compton suppression.

The active-sample technique, in addition to the normal

guarded system spectrum, yields two additional simplified

spectra- one spectrum of gamma photons from electron capture

and metastable states maintaining total Compton suppression,

and a second spectrum of prompt gamma photons from 83 decay

maintaining partial Compton suppression.












CHAPTER III


EXPERIMENTAL SYSTEM


The complete active-sample experimental system consists

of three detectors, the primary, the guard, and the sample

detector. Figure 2 shows their relationship. This chapter

gives both the component specifications and design parameters

for each detector and establishes this relationship. In

addition, the counting room is described and shielding

considerations explained.


A. Primary Detector


In order to obtain the greatest resolution and thus the

most discriminating and accurate spectrometer, a Ge(Li)

detector was selected as the primary detector. The Ge(Li)

detector yields by far the best photopeak resolution and the

greatest peak-to-Compton ratio of any detector presently
(12)
available. However, the efficiency of the Ge(Li)

detector is typically much less than other available

detectors. Taking both properties into account, the Ge(Li)
(6)
detector still yields a superior spectrum in most cases.

The Ge(Li) detector, produced by Ortec, Inc. has a crystal

volume of 34 cc which yields 6% efficiency (re 3x3 NaI(Tl)),2.9

keV resolution, and 15 to 1 peak-to-Compton ratio for Co.

The detector as shown in Figures 3 and 4 includes a liquid

10



















4-)


Or-4
(D
a,



Ha
Ed


0

Z: H
Q)r-
4.3 0
m 0d

























































Fig. 3 Ge(Li) detector side view.


























































Fig. 4 Ge(Li) detector frontal view.







14

nitrogen container required for cooling of the crystal and a preamp-

lifier mounted on the side.. The standard Ge(Li) detector

configuration prohibits its insertion into a large guard

detector. Therefore to facilitate use of this detector, a

10 inch long by 2.5 inch diameter plexiglass collar having 0.25

inch thick walls was used to raise the detector to a more

accessible position (see Figure 5 ). Due to cooling re-

quirements of the detector, this position could be maintained

for only 24 hours without topping-up the liquid nitrogen.

(In the standard position, the nitrogen container can cool

the detector for approximately three weeks.)


B. Guard Detector


The single requirement of the guard detector is that it

sense nearly 100% of all gamma photons entering it. This

requires that each gamma photon go through at least one

Compton scatter. Since the scatter cross section is typically

much larger than the photoelectric cross section, Compton

scatter dominates the total mass attenuation coefficient
(13)
which is approximately equal (av. 0.1 cm2/g) for all

available detectors within the energy range of interest.

Thus to obtain 95% detection the mass thickness must be
2
30 g/cm2. For NaI(T1) crystals this would require a detector

depth of about 8 cm and for liquid or plastic scintillators

about 30 cm.(Ge(Li) detector configuration limits its position-

ing into a solid guard detector to 30 cm deep.)























































Fig. 5 Ge(Li) detector with extension collar.





16
Liquid scintillator was chosen for the guard detector

because of its faster decay time over NaI(T1) and its larger

output and longer light mean free path than plastic scintilla-
(13)
tor. The former lowers the energy detection limit and

decreases dead time while the latter allows use of a larger

volume detector, and thereby higher detection efficiency.

Pilot Chemical's new high-efficiency (60% re anthracene)

mineral oil scintillator was determined to have one of the

longest mean free paths (5 meters) and the shortest decay
(14)
times (2 ns). The guard detector consists of 80 gallons

of this scintillator in a 31 inch diameter by 26 inch long

cylindrical tank with a Fairchild K2128 15 inch photomulti-

plier (see Figure 6) mounted on one face (see Figure 7) and

a 5 inch diameter by 12 inch long sample port in the oppo-

site face (see Figure 8). The inner surfaces of the tank are

painted with white epoxy paint to increase photomultiplier

light collection.

The Ge(Li) detector can then be positioned in the guard

detector port as shown in Figure 9.


C. Sample Detector

The sample detector was designed in the shape-of a

Marinelli beaker (coaxial around the primary detector) to

allow detection of large-volume low-concentration samples.

Since addition of sample depth beyond 3.5 cm has little

advantage due to scatter and absorption in the sample (50%

attenuation path of 100 keV gamma photons in water), the

sample container size was restricted to 1.5 litres giving






17




































4.



-P







0



0


4-)
C)




0



41-)
a)
Id
rd

$4
cs









lii
'14 "





19































,0


42
C)

4--

0










0


E
cs
Co

0
42


c,
H:
p1

CI




rz
























































Fig. 9 Ge(Li) detector positioned in sample port.





21

about 3.5 cm of depth around the primary detector (see

Figure 10).

The sample container itself is made of 15-mil thick

aluminum to minimize absorption (2% of incident 70 keV gamma

photons), is coated with white reflective epoxy paint and is

viewed on the end by an RCA 6655 2-inch diameter photomulti-

plier tube.

In order to act as a detector, the sample must be mixed

with a liquid scintillator. Several different commercial

scintillators allow addition of up to 40% water. These include

Aquasol by New England Nuclear and Insta Gel by Packard Instru-

ments. In addition, a mixture of p-xylene and Triton N-101
(15)
has been shown to accept 40% water. However, only Aqua-

sol was used due to its superior light transmission.

Two sample ports were constructed for use with and without

the active sample container. Figure 11 shows the sample port

for use without the active sample container and is constructed

of 50 mil thick aluminum. Figure 12 shows the sample port for

use with or without the active sample container and is con-

structed of 125 mil thick plexiglass. The former absorbs 6%

of incident 70 keV gamma photons and the latter 2%.


D. Counting Room and Shielding


The counting room, surrounded by 24 inch thick low-

activity poured concrete, measured 7.5 feet by 18 feet. The

positioning of the guarded detector system in this counting

room can be seen in Figures 13 and 14. The counting room also








23











































rd
Id

z
4-,




(Z





0
4-,
$A



0









rx-





24





































cs


to



cs
0
0)



C)
H














co
C)l







r4-
G('
H




rxl





25






























4d
C0















O
C)







IT-4



d





26













































0
0
0


U
O

r e






r**

OI





27

houses a 4x4 inch NaI(Tl) system.

Gamma exposure as measured on a Nuclear Chicago model

2650 gamma survey meter was 0.01 mR/hr in the counting room

and five times higher outside of the room.

Mercury was used to shield the guarded Ge(Li) system

since it has a higher mass number, greater density, and lower

activity than lead. Five thousand pounds of mercury in 2.5

inches thick double wall tanks (approximately equivalent to 4

inches of lead) surround the guard detector and its photo-

multiplier tube (see Figure 15). With the shield, the

guard detector measures 36 inches diameter by 31 inches long.

Due to inherent activity in construction materials, the

detectors, and the mercury, additional shielding is not

significantly advantageous.

As protective measures against mercury toxicity, a

ventilator system and seamless floor were installed in the

counting room.


E. Electronics


The complete system electronics is block diagramed in

Figure 16. Each of the detectors has its own preamplifier,

linear amplifier, and threshold detectors which drive coincidence

and anticoincidence gating. In addition, the Ge(Li) detector

electronics include an ORTEC model 120-2B preamplifier (supplied

with the detector), a model 485 linear amplifier, and a model 444

biased amplifier. Except for these, the ORTEC high voltage

supply, and a Packard Model 115 400 channel multichannel

analyzer (MCA) (shown in Figure 17) the electronics were











*. .., F-No


It






29













Cd
0



4-)
2. 2






C)

0)













4-)

I) H



Cd)

Ei)





02
cts
PC1





























$-4
:E 0 c


0 cs 0
-r-4 4-



0 cs 0
4-) > 4) Cd4.
a) ) 41












i l






31
custom designed and built in the laboratory.

The signal from the Ge(Li) crystal after amplification

in the 120-2B preamplifier is split, being amplified by the 485

linear amplifier and 444 biased amplifier before entering the MCA,

and being amplified, sensed by a threshold detector, and then

given control of the gate leading to the anticoincidence in-

put of the MCA. Similarly, the guard detector and sample

detector have their own identical preamplifiers, linear amplifiers,

and threshold detectors. In the sample-guard mode, these two

threshold signals are combined by an OR gate which feeds the

Ge(Li) AND gate controlling input to the MCA anticoincidence

input. In the coincidence sample mode, only the guard thres-

hold feeds the Ge(Li) AND gate while the sample threshold con-

trols the coincidence input of the biased amplifier.

The effect of this circuitry for the sample-guard mode

is to send an anticoincidence control pulse to the anticoin-

cidence input of the MCA only when the Ge(Li) detector yields

a signal above its noise level and the guard detector or the

sample detector yields a signal above its respective noise

levels. The effect for the coincidence sample mode is to

allow amplification of the Ge(Li) signal only when the sample

detector yields a signal above its noise level and to send an

anticoincident pulse to the anticoincidence input of the MCA

only when the Ge(Li) detector and guard detector yield signals

above their respective noise levels.

The circuit diagrams for the electrons of the three

detectors, the coincidence units, control pulse generators,




32
and cable drivers are given in Appendix A.












CHAPTER IV


NaI(T1) SYSTEM

In order to better show the advantages of a guarded

Ge(Li) system, comparison is made with a NaI(T1) system. In

this chapter both the NaI(Tl) configuration and its associated

electronics are presented.


A. Configuration


The NaI(Tl) system used as a standard for comparison

consists of a Harshaw 4x4 inch NaI(T1) crystal viewed by a

photomultiplier tube and positioned in the center of a 2 foot

cubic, 2 inch thick lead shield lined with Cu and Cd (gee

Figure 18). The top side of the shield is supported on

rollers to allow access to the detector crystal. (see Figure

19 ). The shield and detector system occupies the rear corner

of the low level counting room as shown in Figure 20.

The NaI(Tl) detector is customarily used to count samples

in both 1.0 litre and 3.5 litre Marinelli beakers. In addition,

samples in 1.0 litre "ice cream" containers and flat samples are

frequently counted on this crystal.


B. Electronics

The electronics consists of a Power Designs Pacific

model HV-1565 high voltage power supply, Nuclear Data models

33

























































Fig. 18.NaI(T1) detector configuration.



















wc~E





37

NTD-180F analog to digital converter, ND-180N 512 channel

memory unit, and ND-180R readout control unit, a Tally

Corporation model 1506 paper punch, a Tektronix model RM503

oscilloscope, a Nuclear Data model ND-316 autofinger mounted

on an IBM electric typewriter, and a Dohrmann Instruments

model 299 chart recorder. The system block diagram is shown

in Figure 21.

The signal after collection by the photomultiplier tube

is amplified, digitized and recorded in the appropriate

memory channel. Upon command, the memory can either be

displayed on the oscilloscope, plotted, or paper punched.

The latter allows computer analysis of the data.




38















o
0
0 0





CO
EO 4) 4*)

















0 r.O
CO0
H" 0 N4

4- H

d -) 0 1p4

0














H Ho
4--

















-4
CO
4-






002
< 0 rz
(_) CH
I ----- I -- I --N












CHAPTER V


NaI(T1) PERFORMANCE


In this chapter the optimum performance of the 4x4 inch

NaI(T1) system in the low-level counting room as described

in Chapter IV is stated. Included are system background,

efficiency, minimum detectable activity, peak-to-Compton

ratios, and ability to analyze complex spectrum.


A. Detector Background


The 4x4 inch NaI(T1) crystal background count-rate was

determined in the absence of a sample and with the shield in

place. Since standard counting procedure in this laboratory uses

40-minute counts, a 40-minute background was accumulated and

is plotted in Figure 22. Only traces of the 0.35 MeV RaB

and the 1.46 MeV 4K photopeaks are visible in the spectrum.

Table 1 lists the background count-rate grouped in various

energy ranges and shows a cumulative background count-rate of

just over 500 cpm for the 70 to 2500 keV range.


B. Efficiency


The absolute efficiency of the 4x4 inch NaI(T1) crystal

was measured for both point sources (1.0 cm from the crystal)

and 3.5 litre sources with calibrated sources obtained from

39











































x0t 91"'T














gu 3c"


TaNNV1H3/SlNnOD


OO
o0
-4






41










4-)
cs
r \- P4-) 104 -.t V 0

0





a,
4'
CO
C




00 0 0

j N N N
-4 O 0 0 0 0
> I I I I
1 0 0 0 0
S 0 0 0





0






0 0
4.1








H H
00
0 0
Oc O










O O 0 0 0
0 a 0 co 0o\ 0










p4 0 0 O O O 0 0 0
-.g C- 0 0 0 0 0 0 0
N 00 co 0 OV 0
t0> 1 1 1 1 1





42

Baird Atomic and the Public Health Service. The plots of

photopeak (full width at tenth maximum) efficiency vs. energy

are shown in Figure 23 and include data from the Public Health
(16) (17)
Service and previous data taken on this system. The

data yields essentially straight lines on log-log paper, showing

increasing efficiency with decreasing energy down to a point

where detector window absorption becomes significant. These

data also show that the 3.5 litre sample efficiency is

typically 1/3 of that for the point source, the difference

being attributed to differences in the average source-to-

detector distance.


C. Minimum Detectable Activity


The theoretical minimum detectable activity depends upon

the background count and photopeak efficiency. It was

determined for the 4x4 inch NaI(T1) detector by requiring a 95%

detection certainty. This means that the theoretical minimum

detectable activity equals two times the standard deviation

in net count (gross minus background) under the photopeak

energy spread, corrected for detector efficiency. For a 40-

minute count, the minimum detectable count rate in cpm equals


2 4/ (gross count + background count)
40

which yields a minimum detectable activity in dpm of


Using the measured photopeak efficiency and background


21/2 x background count (under Dhotopeak)
40 (efficiency)










I I I I I


100
90
80
70
60


I I


I I


t~i*


**


4-


21-


SPoint source
* 7 ml sample (16) (17)
S3.5 1. configuration


L. * C I


.2 .3 .4 .5 .6.7 .8 1.0
Energy (MeV)


photopeak efficiency.


0.O


50

40


10
10 -
S9
" 8
U 7
I-

5


2.0


3.0


.I .


I I I I I I I I I I a j I i I


! i


I | I


Fig.23. NaI(T1)




44

data, the theoretical minimum detectable activity was

calculated for nine radionuclides and photopeak energies.

Figure 24 shows a plot of minimum detectable activity in

gamma photopeak dpm vs. energy. In addition, the minimum

detectable activity in pCi and the minimum detectable con-

centration in pCi/1 for several radionuclides of interest

are listed in Table 2.


D. Peak-to-Comoton Ratio


The peak-to-Compton ratio of a detector is important since

it is an indication of the amount of interference between

radionuclides in the complex spectrum. Figure 25 shows the

peak and Compton relationship in a typical point-source

spectrum. Statistical variation in large Compton plateaus

greatly increases the minimum detectable activity in that

energy range. This uncertainty increases with large-volume

samples due to Compton scatter within the sample before

detection. Also in complex spectra addition of Compton

plateaus at the lower energies becomes very significant.

The peak-to-Compton ratio for the 4x4 inch NaI(Tl) de-

tector was measured and plotted in Figure 26 for both point

sources and 3.5-litre sources. The degradation in peak-to-

Compton ratio for the large sample can be readily seen to in-

crease at lower energies.


E. Complex Spectrum


Figure 27 shows a complex spectrum for the 4x4 inch

NaI(T1). The complex spectrum of 226Ra was used in order to
















o o

S0 00

Uh 0 H
EHA
Ol -,-4
* 0
CID H




c sO
**
o4 --



0 0 O
o




*








S 0
0*
*3 ci
T-






















0
c3








*c3












Wda )V2doloHd
~d~ ~F60~0H













H ic--



0 F. 0
p o o - r- as






co p
) ,.-4-) ;-4-



0 o' -Ot
P 4 0 0 4O P0 > 0 \O 00 C.- \.0 rO<




M 0 0 H H H

d >p <

-0 0 C) 0














po >
) P4














S"o op o
S H4-) H


0 rf 4-.) C )
S0 r-0 *
41) 1 0 1 r-H .








0 4 P 'd 0 H
" 0 crH N N C- H \ N x
4'> C- H H
N .I 0 .4-)
4 0 0 -r4 0 0



Ho
H oi ( O Pr \



P 0 ? 0 bD






0O 0



C ) 1 0 o wr




O 4 O LO N JO H ) Ft
4, 0 r 0
ri 4- PA F C













) 0a N H I 0 n H c
PO CC- H >~ H
2 0 w
4- -Qi tO0





e us 0 N 0












. C N H 4 Pt' 4
4 N 0 0 O
ENN H H
\0l m N CO r-4 V -


cd -V m P, z





NNrr



^d

































































SLVE IINnOD




48











0
o
C- *


0 O
















fC3
0 r
cO O







H


















0)
00 o
Jtr'I
-- -- T





































, I I
OIIL'S~Z





























0Do 0O'Z


DOS 9t'T





Dce qc*i


O H ZTIT


DOu C6'0


OUH zz/'O


DVE 19 0


aNNVHiO/LNflOD


D) 4-)


000
0N0
OUO


c0 o-
CM


DetK et*'Z





50

standardize comparison. It can be seen that both the poor

resolution and the addition of Compton continuum raise the

minimum detectable activity in the low-energy range of the

spectrum to approximately the activity of the high-energy

photopeak radionuclides in the spectrum.













CHAPTER VI


Ge(Li) SYSTEM PERFORMANCE


In this chapter the performance of the standard Ge(Li)

system is presented. System background, efficiency, minimum

detectable activity, peak-to-Compton ratios, and complex

spectrum resolution are tabulated.


A. System Background


In order to show both shielding benefits and shielding

problems, a 24-hour unshielded background in the low-level

counting room was taken with the Ge(Li) spectrometer. The

spectrum given in Figure 28 shows a complex collection of

photopeaks from both radon and thoron daughters. Background

with 2.5 inches of mercury shielding, given in Figure 29,

shows a considerable reduction across the spectrum with

fewer photopeaks being discernible. The background still

contains small radon daughter photopeaks mainly due to their

presence in the large air space between the shielding and

the detector. Background with this air space filled with

liquid is discussed in Chapter VIII. Integrated background,

both shielded and unshielded, is tabulated in Table 3 for

several energy ranges.





















0
a
















V- 99
C

































V H 98l'


loqT, Z9*Z



IaOe 9C Z








D01 0120T


(l


Ot(-X 0917 T
oDua 9Ttr'T


1013H ZtZ*1

szeu OZ'1


8z-oV
8ZZ-Vt


966 0
806"0


L&DoL 69L'0


Oeu 609o0


ZC'o
z67z'0
96'0


gTiLL -1 'O


(OX) NO
(t OX) SiNnfo


r 0







































X O9T'I







DUE 03*T"


ov 966*0
ovg 806"0 0






DO: 6o09


tC
o
o


coc
N R
H

00









0 0
bo
00


Of -






C-,
M
fcO



(Q c


o O
0o NNVHO/SLNnfOO
H(




















-- t
lV CO H- H N H
* H







4-)




0
I










a)


C) C) N C'- n V Nq n' H? V' N 0 \O
(Q< <
v,


0)
S0 0 0 C) 0 0 0 0


> I I I I I I I I I I I
0 0 0 0 0 0 0 0 0 0 0 0
00 0 0 0 0 0 0 0
--- 0 0 0 0 0 0 0 c' 0 0 0
N 0 C\ o 0 0 N -- 0
i ~ ~ o~~ ,-I -I ,-





55

B. Efficiency


The efficiency of Ge(Li) detectors is much less than

that of NaI(T1) detectors, ranging up to only 20% as

efficient as 3x3 NaI(Tl) for 6Co. The Ge(Li) detector

used for this system was rated only 6% efficient (re 3x3 inch

NaI(T) 6Co source). To allow activity calibration, an

absolute efficiency for the Ge(Li) detector was determined

for both calibrated point sources and for 1.5-litre sources

(configuration of the active-sample container). Both are

plotted in Figure 30, and essentially are straight lines on

log-log paper with efficiency increasing at decreasing energies.

The 1.5-litre configuration is again about 1/3 as efficient

as that for the point source (1 cm distance) due to the

difference in geometry factors for the two configurations.


C. Minimum Detectable Activity


The theoretical minimum detectable activity was determined

as in Chapter V. For a 40-minute count, it equals in dpm


24 2 x Background count (under nho opeak)
40 x efficiency (of photopeak)


Figure 31 shows the theoretical minimum detectable photo-

peak activity for the 34 cc Ge(Li) detector with 2.5 inches of

mercury shielding and shows a slight increasing sensitivity at

lower energies. In addition, the theoretical minimum detectable

activity in pCi and theoretical minimum detectable concentration

in pCi/1 are tabulated in Table 4 for several radioisotopes of

interest.











I I I I I I I I


6_


1.5 litre
Point source


[I I I I I I


.2 .3 .4 .5 .6.7 .8 1.0
ENERGY (MeV)


2.0 3.0


Fig.30. Ge(Li) photopeak efficiency.


4


3-


.1
.1


I I I I I I I I


I I


I


I I


I I


2k


I


I



















(,
O-4
0
U I

O



>HPo


0 0 O o 0 0

w6a )V dOLOHd


0



o





ci-
o
0






.-






0

c
O




















*
4-



0


c,






* <-






c\





H 58


D r-If




0U .
4 E O N 4 N 0 H 0 4
0 N N \ H H r H C' H

S0


p p crT 0 H
4 -




Loo 0

-H-
O S C VI A N r r H 0
4-) 4 0, 0 0 -- r






) HH

0 0
0*

o aH (\ "0 H N N H c' \0 N
i 0 H HH




H 0
-PP

-i o o H






,0
< 0 .4



H
S ( 0 i 0 n


0 P z H
Ei- ) 0 *
E oS 0 + Nj c^- ( o

>0 C) 0 r \ H c N i c N N l
4 o H un H o f
o5i4-3 H i -Qt




OP4
H CC).r-\ 0 C0 V40


40 4




o 4 0 O 4 ) 0
S0





SN o H r
ce o













0 \0 N C- \ H 4
04 N N 0a 0 COr 4 \0
N N H HI H H
4> C` aD *








*^~c CI +s C (





59

D. Peak-to-Compton Ratio


Due to the high resolution of Ge(Li) detectors, their

peak-to-Compton ratios are typically several times larger

than those for NaI(Tl). The peak-to-Compton ratio for

several radionuclide photopeaks was determined for the 34 cc

Ge(Li) detector and is plotted in Figure 32. Both point-

source data and 1.5-litre data are shown. Again it can be

seen that the peak-to-Compton ratios for the large-volume

sources are much less than for the point sources, being

particularly worse at low energies.


E. Complex Spectra


To demonstrate the usefulness of the Ge(Li) detector

for analysis of complex spectra, the complex spectrum for
226
Ra is shown in Figure 33. It can be seen that the

Compton continuum contributes most to the increase in the

practical minimum detectable activity and not the photopeaks

which are now only 3 keV wide. Thus the minimum detectable

activity at low energies is comparable to the Compton

continuum and can be predicted by the cumulative activity of

all radionuclide photopeaks of higher energies divided by

the average peak-to-Compton ratio.


















0
C.,

H 0



.0


0 *t






*L


OIJIVH


0




A
0














*o 0
m 4-)



*" 0

C-)


0
0





N
*




'- ri
h x3p
s
M


1 I













D 3H W1OZ"


O^H 8T78"I
D-6H 79Z" *T -
DlH E2 *T


-euI 6o0'T




sEH 8gg'T


3UH 0OZ'T


D6uH 6" oo



DuH 69"*0


oDH 609"0


auH XcS'O


I~N~vHD/LLI~mOD


0
o

U)0 0

c'J C

as 4













CHAPTER VII


GUARDED SYSTEM PERFORMANCE


In this chapter the performance of the anticoincidence

guarded Ge(Li) spectrometer is presented. As in Chapter VI,

the system background, peak-to-Compton ratio, and minimum

detectable activity are specified. In addition, factors

determining guard-detector efficiency and reduction in

photopeak efficiency are explained.


A. System Background


In the guarded mode, gamma-ray spectrum background is

reduced due to the additional passive shielding of the liquid

scintillator, the removal of Compton scattered events in the

guard which are detected in the Ge(Li) crystal, and the

removal of Compton scattered events in the Ge(Li) crystal

which are detected in the guard.

Figure 34 shows a 24-hour background spectrum with the

guard off. Comparison of this with Figure 29 (mercury

shielded background) shows a decrease in the background at

high energies but an increase in the background at very low

energies (200 keV) when the liquid scintillator is present.

This difference is due to additional energy removal in

Compton scatter by the low-mass-number liquid,thus reducing the




63










0
C,

O













o










z
60





0 00- o d










-00
I












00 0
ggi Z~'O -












0o0 0-

qsNNVMDA lsNno3





64

energy of the background gamma photons. Figure 35 shows the

effect of anticoincidence guarding on the background spectrum.

A comparison of background is tabulated in Table 5 and in

general shows a guarded background reduction of approximately

6 times over the unguarded for all energies. The photopeaks

remaining in the spectrum are due to the unshielded 3% solid

angle geometry of the sample port and the activity within the
40
detector system. The 4K photopeak in the spectrum corresponds
226
to a point-source activity of 45 pCi and the Ha daughter

photopeaks correspond to an activity of 10 pCi. Thus, this

sets the lower limit on the detectable activity for these

two radionuclides.


B. Guard Detector Efficiency


The main purpose of the guard is to detect Ge(Li)

crystal scattered gamma photons. Interaction in the guard

detector may deposit only part of the photon's energy. Thus

for the highest detection efficiency, the threshold of detection

for the guard should be as low as practical. This was deter-

mined by allowing a photopeak reduction of 10% due to unre-

lated chance pulse counts in the guard detector system and

corresponds to a background count rate of 2,400,000 cpm.

Appendix B gives the background spectrum of the guard system

and shows the threshold corresponding to 2,400,000 cpm to be

deep in the noise region of the photomultiplier tube.

With the guard threshold set at this point (approximately

15 keV) its efficiency was determined and plotted in Figure 36.




65









0
0.0



T
r o


c








0-0
oo










(. 09'
,co
xrd











~~~('h~ ne d~d 0) I g'fiM

















03 0OT p
cs
0
-^ o













o o o
M -c
S (
_










'0o
o o o
OS 60o -^ ^


(o *A-nb -40 -[do[ ^ ScuO





















C.- _. C -.
C- 0 H H 0 co

C N O 0 0 0 0 0













- I* V '
00 c0 0 0 O r 0


















0 N C- c~ c) H c O 0 O
r-4 N 'V -+ N
H


Q)

00 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
SN 0 -H N N C\N N N NOJ
A M I I I I I I I I I I I I
be.- 0 0 0 0 0 0 0 0 0 0 0 0
C- 0 0 0 0 0 0 0 D'.- 0 0 0
N 0 O 0 0 N 0
H H N H


a
C)








0
0

O













































\O
0-0



v\>>
o

*9Z


0 0 0 0 0
ON co O \

(%) XON3I3IddS





68

Because of the high background count, the efficiency was

calculated by first measuring the photopeak reduction for

one photopeak of a cascaded decay (2 photons) radionuclide
24 60 133 134
(e.g. Na, Co, Ba, and Cs). After deadtime correc-

tion the ratio of the guarded to unguarded count for one

photopeak gives the undetected fraction of the other gamma

photon (133Ba and 134Cs having complex cascade decay required

some approximations).

The plot of efficiency shows a loss of efficiency with

decreasing energy below 500 keV and a loss of efficiency

with increasing energy above 700 keV. Losses in efficiency

can be explained with a 3% energy independent geometry loss

by the sample port, low energy loss due to a 15 keV detection

threshold (loss 2 E(keV)/15, see Appendix C) and the decrease

in attenuation coefficient for high-energy gamma photons.

The reduction in Compton continuum is directly dependent

upon the guard detector efficiency except at low energies where

the Ge(Li) detector mounts and casing absorb some of the crystal

scattered gamma photons.


C. Photopeak Reduction


Since for highest Compton suppression the guard detector

threshold must be set as low as possible (deep in the noise

region), an acceptable photopeak reduction was arbitrarily

chosen as 10%. A 10% reduction while not substantially

affecting counting statistics or efficiency does greatly

increase the allowable guard detector background and thus

allows significant lowering of the threshold setting.




69

The guard detector background count rate which gives 10%

chance coincidence is 40,000 counts per second.


D. Minimum Detectable Activity


The minimum theoretical detectable activity was

determined as in Chapter VI. However, due to reduction in

background count for the guarded spectrum, 40-minute counts

gave backgrounds below 1 count per channel. When this

occurred, Poisson statistics were used in determination of the

minimum detectable activity (3 is the minimum photopeak count
(18)
for 95% confidence in the extreme of zero bkg.) ) Figure

37 gives the calculated minimum theoretical detectable

activity in terms of gamma photopeak decay for both point

sources and 1.5-litre sources as a function of energy. In

addition Table 6 gives the theoretical minimum detectable

activity, and minimum detectable concentration for several

radionuclides for both 40 minute and 24 hour counts.


E. Peak-to-ComDtcn Ratio


The peak-to-Compton ratios for the anticoincidence

guarded 34 cc. Ge(Li) was determined as in Chapter VI for

the unguarded Ge(Li). The guarded peak-to-Compton edge
65
ratio was 9 times higher for Zn and 8 times higher for

137Cs. Figure 38 shows the guarded peak-to-Compton ratios

as a function of energy for both point sources and 1.5-litre

water sources. The peak-to-Compton ratio can be seen to be

quite constant above 200 keV. The low-energy fall-off is


















C U
o4 wa



?ILI
0*





0




HH
co


o


a



ro o
** -p


0


.4







0
S*r














I I I I

0 0 o o
W, MVOIOOHd
Cp







~6~ ~tdO-i







1-A




-



O
O


C3
I-(

0)
0





0
0







0) 0


'4-

C
0















C3
0










0
co




SFP

4- c
O

4-v









0
U)
















B-


T-H

*
0 o0 \1 0\










O O NO\ hi
0 0
*

































MONN


S-- o r4i
N \ 0 N










-H

to N r
N H





# *











-H


*












*
4- N 0 H


N 4 H








N
^f S 3

cs .c


^ G
0 \0 (\1
ut Ps P
(M 0^1


c0

o H H O H 4-0









r-1












Sr-4 O H O CN 0
H (N N H 4 N






H\N \O N\ O N H


* .










4f \O N C\ 0 \O C?










4, \o o \,o \0 co c"
-S r-4











40 O Y0 O i O Oc











0 0 ( o

n CN c \ 0 N

H' o\ o co' \o vn
rH r rH H l






















0


-1
O
0 u




o-i-



O*


0



o
c4





N
I,
CC





>o






*0



00
o








0
cO O
CM -





0
o






0



N
a0
Ij
(^ 0
6;)


I I I - 1





73

due to both loss of efficiency in the guard system and

absorption in the Ge(Li) crystal casing.


F. Complex Spectrum


The usefulness of the anticoincidence guarded Ge(Li)

detector for a complex spectrum is shown in Figure 39.
-226
Radium was used to show both the separation of the many

photopeaks and the small amount of Compton continuum present.

It can be seen that the minimum detectable activity doesn't

significantly increase for complex spectra.


















-2-
x






DUE 822ZT








etU 9t T





DOE 02T*T







DOU 69'*0



DUH 6090 -


WIV 25c'o


0 0
0
0
rI3NNVHD/wNnoo


o


CIO




(!













CHAPTER VIII


ACTIVE-SAMPLE SYSTEM


In this chapter the performance of the anticoincidence

guarded system with active-sample guarding is presented.

Since the active-sample technique uses the same geometry as

the 1.5 litre samples reported in Chapter VII, only differences

in performance are presented. These include efficiency, peak-

to-Compton ratios, minimum detectable activity, and separation

of decay modes.


A. Efficiency


The efficiency of the active sample geometry is the

same as that reported in Figure 30 for the 1.5-litre samples.

Again,as with the standard'anticoincidence guarding, random

coincidence reduces the photopeak count. The guard again is

set for a 40,000 count per second background while the sample

detector is set for a 30,000 count per second background. The

combined guarded background yields a 15% photopeak reduction as

compared to a 10% reduction with guard shielding only.


B. Minimum Detectable Activity


The single radionuclide minimum detectable activity for

the active sample technique is improved only at energies below





76

500 keV and then by only about 15% due to a 30% improvement

in background suppression. The minimum theoretical detectable

activities, determined as in Chapter V, are plotted in Figure

40 as a function of photopeak energy. In addition, the

minimum theoretical detectable activities and concentrations

for several radionuclides are tabulated in Table 7 for

both 40-minute counts and 24-hour counts. The increase in

the minimum detectable concentration is due to the required

dilution of the sample with liquid scintillator.


C. Peak-to-Compton Ratio


The peak-to-Compton ratio for the active-sample technique

is shown in Figure 41 as a function of energy. The peak-to-

Compton ratios are restored to 95% of their point-source

ratios at high energies but only restored to 50% at energies

below 100 keV. However improvement in peak-to-Compton ratio

is still approximately 2-fold across the spectrum.


D. Separation of Decay Modes


In addition to improving peak-to-Compton ratio, the

active-sample technique removes all short-lived f + and ft

decay excited-daughter radiations from the spectrum. By

requiring coincidence between the sample-detector and the

Ge(Li) detector, these short-lived states can be left in the

spectrum while removing all electron capture and metastable

states. Figure 42 shows this division of states; however

to show the effect with available radionuclides. (positron























































































o 0
c'- \o


NdG XV MdOIOHd


-)


o









H
cs



0




4-)

0)
S >











0
Id
o o





4-l
t1)
c' H3







0
.0 0









-I






-4-



C3
^o a Z













(1






<-
+>


r- <


















Si
QU)
SOr-

5 *
4 *
) Ud 0 V O 0 CN 0 N 0 -.
, 0 C co C' O O (- 0 V-) N



3 N ?




CO O
4) 4 -I

4o 0





SE:)

O
0

0
0 0



| C
P o










,4 O
O *
4-,






,v n *i N

) H 4)


0 O
01




a )
0 4 H 4 N \ N NO N H











-P 0 -^ -' -- -
4- 0














) 0 \0 N H 4
4 N n ( ON O 4 O \0 v
N H N H rH H
ft +-> Uoo~o~
Cf E*P Q
0dm kQ =F
0) ^ *N
+1r o- }.- c^> V^ ^} Cs O ^ C

g ^ + C H




















U)




-1



r-l
!o


-1


0.


I I


r( 0




0 0
.14



r-P
o P4

O
o

Oa

O

o
H> 0
~- P



*- a,


o H




0

S4





.-4


-Ir



0 r


0 ILIV


1 _





















21 %1 ^%69
-3N tZ'T


*o'a %66
59-UZ 6TT'T






'-UN oot8'0
7C-ST-sO 080

aTqlqs'es m _-g %96
Zel-SO g99*0


' %001 i7y I- SO
i7IT-SO
UOTqTTqTUUV


CCT-la

CC T --e


09o*0






oo'o
9gW'o
T-19 ,

9C 0
ZOc 17
9U00C


-4


J o
S 1.
C)







o 0
1 ,







--
;
;_ carJ






5^


elT-~a 80* 0


C)e



ISNNVHO/LN nOD


L"

z
La






Z
-J




UJ
>-



C.
(-


*o's %OOT






81

and cascade decay) the guard was disconnected for the

coincidence mode.













CHAPTER IX


COMPARISON OF SYSTEMS ON ENVIRONMENTAL SAMPLES


In this chapter the spectra produced with the various

systems described previously are compared for their ability

to identify radionuclides. Included are prepared samples

and actual environmental samples in various forms.


A. Standard Samples


To show the advantages of the active-sample technique

for water samples, two standards were prepared. The first,
65
containing only Zn, was used to produce the spectra in Figure

43. The upper two spectra are from a guarded and unguarded

point source; the guarded spectrum shows a Compton edge

suppression of a factor of 8. The remaining three spectra

are for a 1.5 litre source. The normal guarded spectrum

shows a Compton tail suppression of a factor of 7 while the

active-sample guarded spectrum shows a suppression of a

factor of 23. This additional Compton tail suppression

would permit recognition of a much smaller photopeak within

this energy region.

The other standard contained 6 radionuclides having

different decay modes. The spectra are shown in Figure 44.

In addition to the suppression of annihilation radiation,




83








Point Source
Unguarded-Guarded o
o
reduction
(x6)
(x6 (x7) (x8\


0


1.5 Litre Source


Unguarded




00
4.








o
Reduction
z
Gu(x7) (x.5) (x
0






Active Sample

O
reduction
(x7) (x84) (x5.3)

o0





000

CHANNEL NUMBER


Fig. 43. Zn-65 source spectrum.




84







9 34ce Ge(Li)
\_ 5Sample -1.5 litre

0 0


e ci EC be

i ,,' o..

I'' 1










Reduction E.C.& Metastable
ACTIVE SAMPLEANTICOINCENCE400
x33




Effect With Available
Standard Sources
.Short Lived
Reduction & M tsal












0 ACTIVE SAMPLE COINCIDENCE 400



CHANNEL NUMBER

Fig. 44. Complex standard-sample spectrum.
Fig. 44. Complex standard-sample spectrum.




85

cascade events, and the Compton tail in the normal guarded

mode, the anticoincidence sample-guarded mode suppressed

the Compton tail another factor of 3. Also the coincidence

sample mode suppressed electron capture and metastable states

while restoring the rest.


B. Air Sample


To demonstrate the usefulness of the normal guarded mode,

an air sample was taken of 45,000 cubic feet of atmospheric

air and allowed to decay for 3 days. Comparison of 40 minute

count spectra for NaI(T1), unguarded Ge(Li), and guarded

Ge(Li) shows no identification of any radionuclide in either

the NaI(T1) or unguarded Ge(Li). However, data analysis on

the guarded Ge(Li) spectrum indicates the presence of 100 pCi

of 7Be, 8 pCi of 103Ru, and 12 pCi of 95Nb (all activities

determined from digital data see Appendix D).


C. Water Samples


To demonstrate the advantages and disadvantages of the

active-sample technique, two environmental water samples were

counted on the various systems. The first sample was an

environmental standard sample prepared by the Public Health

Service and its spectra are shown in Figure 46. The NaI(Tl),

unguarded Ge(Li) and guarded Ge(Li) spectra all show the

presence of 200 pCi of 106Ru and 200 pCi of 144Ce although

their presence in the guarded Ge(Li) is easiest recognized.

The active-sample mode does not show their presence due to

the dilution of the sample with scintillator.

















Nal (Tl)

with background substraction


200


I V ,I


Unguarded Ge(Li)


45,000 cubic feet
3 day decay
40 min. count


~PiJV.AIL&JIL. J. n..


400

u u
00O
o Guarded Ge(Li)
0
I I I
c~- ON
-v\t
0-0- r(N
0'-C 0\OrI1


, LL. ..-. I


00 O
...A L.. L ,


4b0
CHANNEL NUMBER


Fig. 45. Air sample spectrum.


'I JllPlSii ^nnl7~LIEIhnlnyiW


0 \


800


* -d A -in--- warp


800


IT- -J -4 L11 A .. ....... . . ..L


-LlhLYLCLII _L*I_~I-L--~-IL-- ----L


a













NaI (Tl)
with background substractio


3.5 litre


300







0





20


0


1.5 litre


Guarded Ge(Li)


O -I



ScU

0


4o0


1.5 litre


400


400


Active Sample


Water standard
from PHS
40 min. count


0.7 litre


30 reduction


400


400
CHANNEL NUMBER


Fig. 46. Water standard spectrum.


Unguarded Ge(Li)


Z

20

E 0

0 0


U U
o f
.-o-0


,I I II


I


r C1CrrC~p6AF~J ) IrknlL.LhUC


- ----


_ ___-


zWc~ud





88

The other environmental water sample was taken from the

University of Florida Training Reactor coolant water. The

spectra shown in Figure 47 indicates the presence of 20,000
24
pCi of activated Na. The NaI(Tl) spectrum is completely
24
swamped by the Na, however the unguarded Ge(Li) clearly

identifies the presence of 1,400 pCi of 99mTc, The guarded

Ge(Li) spectrum also shows the presence of 700 pCi of 115In

and 115Cd. The active sample mode spectrum, because of the

removal of 24Na ( a short-lived, ) and its Compton continuum,

shows the presence of 13 pCi of 133Ba and 90 pCi of 65Zn

(both contamination in the sample container). It should be
24
noted that the 2Na Compton continuum is less than the back-

ground in the active-sample spectrum and thus does not raise

the minimum detectable activity in that region.


D. Solid Samples


To further show the usefulness of the guarded Ge(Li)

system three environmental solid samples were analyzed. The

first, whose spectra are shown in Figure 48, consists of an

800-gram soil sample taken from undisturbed top soil. All

three spectra for this sample; the NaI(Tl), the unguarded Ge(Li),

and the guarded Ge(Li) show the presence of 1200 pCi of 226Ra and

its daughter products, 500 pCi of 137Cs, and 400 pCi of 40K.

The NaI(T1) spectrum shows considerable interference between

these radionuclides, especially between the 662 keV 137Cs peak

and the 609 keV RaC peak. The guarded Ge(Li) greatly reduces

peak-to-peak interference and Compton-to-peak interference




Full Text

PAGE 1

DEVELOPMENT AND COMPARISON OP COMPTON SUPPRESSION TECHNIQUES FOR LOW -LEVEL RADIONUCLIDE ANALYSIS By Prank R. Markwell A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OP THE UNIVERSITY OP FLORIDA IN PilRTIAL FULFILLySICT OF THE HSQUIREWENTS FOS Tli^ DEGREE 0? DOCTOR OF PHILOSOPHY IK NUCLEAR ENGINEERING SCIENCES UNIVERSITY OF FLORIDA 1972

PAGE 2

ACKNOWLEDGSKENTS The author wishes to express his appreciation to Dr. W.E. Bolch, Dr. M.J. Ohanian, Dr. C.E. Roessler, Dr. VJ.H. Ellis, and Dr. H.A. Bevis, members of his supervisory committee, for their advice and assistance in the preparation of this dissertation. He especially wishes to thank Dr. V.'.E. Bolch for the suggestion of the dissertation subject and for his encoiiragement and valuable advice. Special thanks are due Dr. C.E. Roessler for his aid, helpful suggestions and constructive criticism. Thanks are also expressed to Dr. W.H. Ellis for his helpful suggestions and encouragement. Special appreciation is extended to the the staff members of the Nuclear Engineering Department and the Environmental Engineering Department in the University of Florida, without whom this work could not have been performed. The author acknowledges the support by the Atomic Energy Commission for the award of a Special Fellowship in Nuclear Science and Engineering during his graduate studies and the support by Florida Power Corporation. In addition, the author wishes to thank Nuclear Physics for the loan of the Ge(Ll) spectrometer. 11

PAGE 3

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS j_i LIST OF FIGUHES ,^1 LIST OF TABLES ^^ ABSTRACT ^ Chapter I . lOTRODUCTI ON 1 II. THEORY l^, A. Ge(Li) Detector Ij, B. Anticoincidence Guarding ., 6 C. Active-Sample Guarding 8 III. EXPERIXEr:TAL SYSTEM 10 A. Primary Detector 10 B. Guard Detector 14 C. Sample Detector 15 D. Coxmting Room and Shielding 21 E. Electronics 27 IV. Nal(Tl) SYSTEM 33 A. Configuration 33 B. Electronics 33 V. Nal ( Tl ) PERFORMANCE 39 A. Detector Background 39 B. Efficiency 39 ill

PAGE 4

TABLE OF CONTE^JTS — continued Page Chapter -continued C. Minimum Detectable Activity kZ D. Peak-to-Compton Ratio JWE. Complex Spectrum kk VI . Ge ( Li ) SYSTEM! PERFORMANCE 51 A. System Background 51 B. Efficiency 55 C. Minim-uiT! Detectable Activity 55 D. Peak-to-Compton Ratio 59 S. Complex Spectrum 59 VII . GUARDED SYSTEM PERFORMANCE 62 A. System Background 62 B. Guard -Detector Efficiency 64 C. Photopeak Reduction 6g D. Minimum Detectable Activity 69 E. Peak-to-Compton Ratio 69 F. Complex Spectrum 73 VIII . ACTIVE SAMPLE SYSTEM 75 A. Efficiency 75 B. Minimum Detectable Activity 75 C. Peak-to-Compton Ratio 7S D. Separation of Decay Modes 76 IX. COMPARISON OF SYSTEr!ENTAL SAMPLES 82 iv

PAGE 5

TABLE OF CONTENTS — continued Page Chapter -continued A. Standard Samples 82 E. Air Sample 85 C. Water Samples 85 D. Solid Samples 88 X. EVALUATION OF DETECTION SYSTEMS 94 A. Kal(Tl) System 92+ B. Unguarded Ge(Ll) System 95 C. Guarded Ge(Ll) System 95 D. Active-Sample System 96 XI. CONCLUSIONS 97 Appendices A. ELECTRONIC CIRCUITS 100 B. GUARD-DETECTOR CHAxRACTERISTICS 105 C. SPECTRUM EI:ERGY CALIBRATION 110 D. EXAJ^PLE OF DIGITAL DATA II3 LIST OF REFERENCES II5 BIOGRAPHICAL SKETCH II7

PAGE 6

LIST OF FIGURES Figure Page 1. Effect of Compton suppression 7 2. Guarded Ge(Li) system configuration 11 3. Ge(Li) detector side view ., 12 ^. Ge(Li) detector frontal view 1^ 5. Ge(Li) detector with extension collar 15 6. Guard-detector photomultiplier tube 17 7. Guard-detector with photomultiplier mounted ... 18 8. Sample port mounted in guard detector 19 9. Ge(Li) detector positioned in sample port 20 10. Sample container with photomultiplier mounted . 22 11. Sample port for standard technique 23 12. Sample port for active-sample technique 2^ 13 . Counting room and entrance 25 1^. Counting room 26 15, Photomultiplier shielding 28 16, Guarded Ge(Li) system electronics 29 17. Ge(Li) system analyzer 30 18, Nal(Tl) detector configuration 3^ 19. Nal(Tl) crystal in lead shield 35 20. Nal(Tl) system in counting room 3^ 21. Nal(Tl) system electronics 38 22, Backgrouna for Kal(Tl) ^0 vi

PAGE 7

LIST OF FIGURES — continued Figiire Page 23. Nal(Tl) photopeak efficiency ^3 24. Klnlmum theoretical detectable activity for Nal ( Tl ) 45 25. Peak-and-Compton spectrum for Nal(Tl) k? 26. Peak-to-Compton ratio for Nal(Tl) 48 27. Complex Nal(Tl) spectrum 4-9 28. Counting room background 52 29. Ge(Li) background with Hg shielding 53 30. Ge(Li) photopeak efficiency 56 31. Minimum theoretical detectable activity for Ge(Li) 57 32. Peak-to-Corapton ratio for Ge(Li) 60 33. Unguarded complex Ge(Li) spectrum 6I 34. Unguarded Ge(Ll) background 63 35. Guarded Ge(Li) background 65 36. Guard-detector efficiency 67 37. Minimum theoretical detectable activity for guarded Ge(Ll) 70 38. Peak-to-Compton ratio for guarded Ge(Li) 72 39. Guarded complex Ge(Li) spectrum 74 40. Active-ssimple minimum theoretical detectable activity for Ge( Li) 77 41. Active-sample peak-to-Compton ratio 79 42. Separation of decay modes , 80 vil

PAGE 8

LIST OF FIGURES — continued Figure Page ^3. Zn-65 source spectrum 83 ^. Complex standardsample spectrum 8^ ^5. Air sample spectrum 86 ^6. Water standard spectrvim 87 ^7. UFTR primary coolant water spectrum 89 ^8. Soil sample spectrum 90 ^9. Oyster sample spectrum 92 50. Seaweed sample spectrum ,, 93 51. Sample-detector electronics circuit 101 52. Guard-detector electronics circuit 102 53. Ge(Li) threshold electronics 103 5^. System coincidence electronics 104 55. Guarddetect or light collection IO6 56, Guarddetector background , 10? 57' Threshold calibration IO8 58. Ge(Li) calibration spectrum Ill 59. Ge(Li) calibration curve 112 viii

PAGE 9

LIST OF TABLES Table Page 1. Background Count Rate for the Nal(Tl) System .. ^1 2. Theoretical Minimum Detectable Activities for the NaT ( Tl ) Sys tern k6 3. Backgroiind Count Rate for the Ge(Ll) System ... 5^ ^. Theoretical Minimum Detectable Activities for the Ge(Li) System 58 5. Background Count Rate for the Guarded Ge(Li) System 66 6. Theoretical Minimum Detectable Activities for the Guarded Ge(Li) System 7I 7. Theoretical Minimiom Detectable Activities for the Active-Sample Ge(Li) System 78 5. Digital Data for Guarded Air Sample Spectrum .. 114 ix

PAGE 10

Abstract of Dissertation Presented to the Graduate Council in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in Engineering DEVELOPMENT Aim COMPARISON OP COMPTON SUPPRESSION TECHNIQUES FOR LOW -LEVEL RADIONUCLIDE ANALYSIS By Frank R. Karkwell December, 1972 Chairman: Dr. W. E. Bolch Major Department: Nuclear Engineering Sciences Optimum analysis of low-level radionuclides requires the removal of Compton interference from the gamma-ray spectruun. Significant Compton suppression can be achieved by employing high-efficiency guard detectors to detect the scattered gamma photons. However, high detection efficiency prohibits significant absorption between the primary detector and the guard detector, thus restricting the sample to one of small mass. When large-mass samples are used Compton suppression is greatly degraded. To alleviate this problem a new concept has been developed and a system constructed, and evaluated which. In effect, makes the sample part of the guard detector. Restricted to optically clear samples, the system requires the addition of a liquid scintillator to the counting sample. This

PAGE 11

"cocktail" is viewed by a separate photomultipller tube and associated electronics which add its detected signal to the guard signal and both are used in anticoincidence with the primary detector. This scheme restores the performance of the system to that of the "small-sample" Compton reduction system. However, it also removes photopeaks due to simultaneous (preceding within 200 ns ) fi or fi decay in the sample which can also trigger the anticoincidence logic. Alternately, by requiring coincidence between the sample-scintillator detector and the primary detector, all electron capture and metas table states are excluded, leaving only a spectrum of photopeaks from short lived and fi' decay daughters but with partial Compton reduction. In short, this technique yields two spectra for each liquid sample, a "perfect" gamma spectrum of some excited states and a "not so perfect" gamma spectrum of the remaining excited states. The complete detection system employs a Si efficient (re 3x3 inch Nal(Tl) ), 2.8 keV resolution Ge(Li) spectrometer as the primary detector arranged to extend 12 Inches into the center of a 31 inch diameter by 26 inch long container of Pilot Chemical's high efficiency mineral oil scintillator. This scintillator, which acts as the guard detector, is viewed by a 15 Inch Fairchild K2128 photomultlplier tube and both are shielded with 2.5 inches of mercury. The sample container arranged axially around the Ge(Li) crystal and viewed by a 2 inch diameter RCA 6655 photomultlplier tube, holds 1.5 litres (60t scintillator). The electronics include

PAGE 12

preamplifier, threshold detectors, and logic circuits to give a timing uncertainty of less than 200 ns . Comparison of this system with a i*-x^ inch Xal(Tl) system and the unguarded Ge(Li) system shows it to have both superior peak-to-Compton ratios and superior minimum detectable activities for counting times as short as ^0 minutes. In addition to maintaining the point source peaJcto-Compton ratio for large water samples, the active-sample technique has the advantage of dividing the spectrum into two spectra according to the decay mode of the radionuclides. xii

PAGE 13

CHAPTER I INTRODUCTION Gamma ray spectroscopy has been an important tool in the analysis of radionuclides for many years. The development of the Nal(Tl) detector opened the field and since then vast improvements in efficiency and reductions in interradlonuclide interference have been made. Improvements in detector efficiency and resolution have reduced this interference but not eliminated it. The remaining interference is mostly due to the Compton continuum associated with photopeaks in the spectriim. Since Compton scatter in the detector is only a partial energy release mechainism, a lower energy gamma photon may also exit the detector. This characteristic has been used in (1)(2) suppressing the Compton continuum. The original systems for such suppressions employed Nal(Tl) as the primary detector surroumded by a large liquid scintillator which detected the Compton scattered photons leaving the primary detector and used them to eliminate such counts from the spectrum. Thus, this technique is denoted as anticoincidence guarding. The development of the Ge(Ll) detector has resulted in greatly improved resolution capabilities and reduced effects from the Compton continuum. The usefulness of such a detector

PAGE 14

for low level radionuclide analysis has been shown previously (3) by Cooper et al . Soon thereafter investigators applied (^)(5)(6)(7) anticoincidence guarding to Ge(Ll) detectors. These systems, which suppressed the Compton continuum up to a factor of 6, employed plastic scintillators as the guard detector. Current investigators, usinp: Nal(Tl) gusirds , 137 (8)(9) report a Compton reduction for Cs of up to 9 times, and those using plastic scintillator guards report a reduction. (10)(11) of 6 times. Significant Compton reduction can only be achieved by employing high-efficiency guard detectors. However, high detector efficiency prohibits significant absorption between the primary detector and the giiard detector, thus restricting the sample to one of small mass. When large-mass samples are (7) used, Compton suppression is greatly degraded. To alleviate this problem, a new concept is proposed v:hich, in effect, makes the sample part of the guard detector. Restricted to clear water samples, the system would require the addition of a liquid scintillator to the counting sample. This "cocktail" would be viewed by a separate photomultiplier tube and associated electronics which would add its detected signal to the guard signal and both would be used in anticoincidence with the primary detector. This scheme should restore the performance of the system to that of the "smallsample" Compton reduction system. Hov;ever, it would also remove many photopeaks due to simultaneous (preceding xvithin 200 ns ) ^ or ^ decay in the sample which could also trigger the anticoincidence logic. Alternately, by requiring coinci-

PAGE 15

3 dence between the sample-sclntillator detector and the primary detector, all electron capture and metastable states could be excluded, leaving only a spectrum of photopeaks from short-lived p and fi decay daughters but with partial Compton reduction. In short, this technique would yield two spectra from each liquid sample, a "perfect" gamma spectrum of some excited states and a "not so perfect" gamma spectrum of the remaining excited states. Presented here are the design details of a liquid scintillator guarded Ge(Li) system with the active-sample option, as mentioned above, and the evaluation of this system for environmental radionuclide analysis with respect to their relative advantages over a 4x^ inch Kal(Tl) system.

PAGE 16

CHAPTER II THEORY In this chapter the theory behind the active sample anticoincidence guarded Ge(Ll) spectrometer is presented. Included are both the theory associated with the new active sample technique and the theory of anticoincidence guarding. The latter is presented since anticoincidence guarding is still a relatively nev; technique and not widely known. A. Ge(Li) Detector Interaction of a gamma photon with the primary detector can occur in one of several ways. These are photoelectric absorption, Compton scatter, and pair production. For germanium and typical gamma energies (<2MeV), only the first two are significant. Photoelectric absorption occurs when the gamma photon releases all of its energy in one interaction with a bound electron, and has the greatest probability for photon energies just above the ionization potential energy of that bound electron. Thus^thls method of photon energy release is most significant for lov: energies and has approximately a (hO) * probability dependance. Photoelectric absorption transfers all of its energj' to the bound electron which in tm'n releases this energy by ionization and electronic

PAGE 17

5 excitation of atoms in the immediate vicinity of the photon interaction. The other interaction of significance is Compton scatter, which consists of a partial transferral of photon energy to the bound electron and the subsequent scatter of the photon (reduction in energy). This reaction has a (hO)"-^ probability dependance and thus is significant for energies higher than those significant for photoelectric absorption. Again the liberated electron from the photon interaction produces ionization and electronic excitation of atoms in its Immediate vincinity . Such interactions within the intrinsic region of a semiconductor produce electron-hole pairs having lifetimes long enough to allow their collection by an electric field. The result is a charge collection which is proportional to the gamma photon energy release within the semiconductor crystal. A count rate-energy release spectrum (charge collection) yields a peak corresponding to photoelectric absorption and a continuum of counts corresponding to the energy release in Compton scatter. Due to the physics of Compton scatter this continuum is approximated by a plateau from zero energy (no scatter) to a maximum energy signicantly less (typically 100-200 keV) than the photopeak energy (corresponding to the energy release in photon backscatter) . The ratio of probabilities of the two interactions depends upon the photon energy and the detector material. For germanium the photoelectric interaction and Compton scatter

PAGE 18

are equally probable at I50 keV but at 1 HeV Compton scatter (7) is 100 times more probable. B, Anticoincidence Guardint^ The useful information from a Ge(Li) spectrum lies in the photopeak energy and amplitude since the gamma photon energy is characteristic of the radionuclide from vxhich it originated and the amplitude is proportional to the quantity present. The CoTnpton plateau is a continuur. vrhich, because of counting statistics, is erratic and thus obscures other photopeaks which may lie within this energy range. Anticoincidence guarding suppresses the counting of scatter events thereby removing this continuum and allowing recognition of small amplitude photopeaks within the energy range (see Figure 1 ), To achieve this Compton suppression, those gamma photons which deposit only part of their energy in the primary detector must be recognized. By placing the primary detector within a high-efficiency detector, photons which are scattered in the primary detector can be detected again in the guard and thus recognized as being Compton scattered. These can then be prevented from being recorded in the spectrum. To obtain excellent Compton suppression, the guard detector must approach 100^ detection (not energy collection) efficiency and photon absorption het\-;een the primary detector and the guard detector must be eliminated. Since good primary detection requires the radioactive sample to be placed in the center of the gixard detector just adjacent to the primary detector, the sample itself can act as an

PAGE 19

o m CO 0) u p. p. ca C o p p. s o o o p o a> bO 8q.BH ^unoQ

PAGE 20

8 unwanted absorber. Thus excellent Compton suppression in standard systems can be maintained only with small samples. In addition, large samples inherently yield a lovxer peak-toCompton ratio due to Compton scatter in the sample before the photon reaches the primary detector. An inherent disadvantage of anticoincidence guarding is the removal from the spectrum of those photopeaks associated with cascade gamma decay (2 or more simultaneous photons) and ^* decay whose excited daughter decays simultaneously (yielding a simultaneous gamma decay photon and annihilation photons ) . C. Active Sample Guarding To elimirate degradation of Compton suppression and maintain excellent peak-to-Compton ratios, the sample must be used as a guard detector also. This allows recognition of sample-scattered photons which enter the primsiry detector and primary detector scattered photons which are absorbed back in the sample. The combined effects yield a peak-toCompton ratio for large samples which approaches that for point source samples. An Inherent problem in this technique arises when a radionuclide in the sample decays by emission of a ^ particle to its excited daughter and then Immediately emits a gamma photon. In this case gamma photons enter the primary detector at the same time that the fi' particles (polyenergetic) release their energy in the sample detector. Thus

PAGE 21

9 the ^ particles can not be distinguished from the scattered gamma photons by the ssimple detector. Radionuclides having excited daughter lifetimes greater than the timing uncertainty of the detector system, and those radionuclides which decay by electron capture, are not affected by this phenomenon. To restore these short-lived excited daughters from fi' decay and remove the long-lived and electron capture events, coincidence signals can be required between the sample detector and the primary detector, indicating fi' decay with an associated prompt gamma photon. This spectrum contains only photopeaks from these prompt gamma decay states but maintains only partial Compton suppression. The active -sample technique, in addition to the normal guarded system spectrum, yields two additional simplified spectra ••one spectrum of gamma photons from electron capture and metastable states maintaining total Compton suppression, and a second spectrum of prompt gamma photons from fi" decay maintaining partial Compton suppression.

PAGE 22

CHAPTER III EXPERIMENTAL SYSTEM The complete active-sample experimental system consists of three detectors, the primary, the sruard, and the sample detector. Figure 2 shows their relationship. This chapter gives both the component specifications and design parameters for each detector and establishes this relationship. In addition, the coijinting room is described and shielding considerations explained, A, PriTT.ary Detector In order to obtain the greatest resolution and thus the most discriminating and accurate spectrometer, a Ge(Li) detector was selected as the primary detector. The Ge(Li) detector yields by far the best photopeak resolution and the greatest peak-to-Compton ratio of any detector presently (12) available. However, the efficiency of the Ge(Li) detector is typically much less than other available detectors. Taking both properties into account, the Ge(Ll) (6) detector still yields a superior spectrum in most cases. The Ge(Li) detector, produced by Ortec, Inc. has a crystal volume of 34 cc which yields 6^ efficiency (re 3x3 KaI(Tl]l,2.9 keV resolution, and 15 to 1 peak-to-Gompton ratio for Co. The detector as shovm in Figures 3 and 4 includes a liquid 10

PAGE 23

11 >5 o HlW (X

PAGE 24

12 Fig, 3 Ge(Li) detector side view.

PAGE 25

13 Fig. 4 Ge(Li) detector frontal view.

PAGE 26

Itf. nitrogen container required for cooling of the crystal and a preamplifier moiinted on the side. The standard Ge(Li) detector configuration prohibits its insertion into a large guard detector. Therefore to facilitate use of this detector, a 10 inch long by 2.5 inch diameter plexiglass collar having 0.25 inch thick walls was used to raise the detector to a more accessible position (see Figure 5 )• Due to cooling requirements of the detector, this position could be maintained for only 2i^ hours without topping-up the liquid nitrogen. (In the standard position, the nitrogen container can cool the detector for approximately three weeks.) B. Guard Detector The single requirement of the guard detector is that it sense nearly 100^ of all gamma photons entering it. This requires that each gamma photon go through at least one Compton scatter. Since the scatter cross section is typically much larger than the photoelectric cross section, Compton scatter dominates the total mass attenuation coefficient P (13) which is approximately equal (av. 0.1 cm /g) for all available detectors within the energy range of interest. Thus^ to obtain 95'^ detection the mass thickness must be 2 30 g/cm . For Nal(Tl) crystals this would require a detector depth of about 8 cm and for liquid or plastic scintillators about 30 cm.(Ge(Li) detector configuration limits its positioning into a solid guard detector to 30 cm deep.)

PAGE 27

15 Fig. 5 Ge(Ll) detector with extension collar.

PAGE 28

16 Liquid scintillator was chosen for the guard detector because of its faster decay time over Nal(Tl) and its larger output and longer light mean free path than plastic scintilla(13) tor. The former lowers the energy detection limit and decreases dead time while the latter allows use of a larger volume detector, and thereby higher detection efficiency. Pilot Chemical's new high-efficiency {GQ% re anthracene) mineral oil scintillator was determined to have one of the longest mean free paths (5 meters) and the shortest decay (14) times (2 ns). The guard detector consists of 80 gallons of this scintillator in a 31 inch diameter by 26 inch long cylindrical tank with a Pairchlld K2128 15 inch photomultipller (see Figure 6) mounted on one face (see Figure?) and a 5 inch diameter by 12 inch long sample port in the opposite face (see Figure 8). The inner surfaces of the tank are painted with white epoxy paint to increase photoraultiplier light collection. The Ge(Ll) detector can then be positioned in the guard detector port as shown in Figure 9» C, Sample Detector The sample detector was designed in the shape of a Marinelll beaker (coaxial around the primary detector) to allow detection of large-volume low-concentration samples. Since addition of sample depth beyond 3.5 cm has little advantage due to scatter and absorption in the sample {50% attenuation path of 100 keV gamma photons in v;ater ) , the sample container size was restricted to 1.5 litres giving

PAGE 29

17 0) -p U O •P o ft >H o -p o 0)
PAGE 30

18 0) C pi O E 0) p. -p H O -P o -p o o
PAGE 31

19 ^1 o -p o 0) -p 0) xi n u OS p c :i o e u o p. 0) H P, i en 00

PAGE 32

20 Fig. 9 Ge(Li) detector positioned in sample port.

PAGE 33

21 about 3.5 err. of depth aroiind the primary detector (see Figure 10). The sample container Itself is inade of 15-niil thick aluminum to minimize absorption (2^ of incident 70 keV gamma photons), is coated with white reflective epoxy paint and is viewed on the end by an RCA 6655 2 -inch diameter photomultiplier tube. In order to act as a detector, the sample must be mixed with a liquid scintillator. Several different commercial scintillators allow addition of up to ^0^ water. These include Aquas ol by New England Nuclear and Insta Gel by Packard Instruments. In addition, a mixture of p-xylene and Triton N-101 (15) has been shown to accept kOfo water. However, only Aquasol was used due to its superior light transmission. Two sample ports were constructed for use with and without the active sample container. Figure 11 shows the sair.ple port for use without the active sample container and is constructed of 50 mil thick aluminum. Figure 12 shows the sample port for use with or without the active sample container and is constructed of 125 mil thick plexiglass. The former absorbs 6% of incident 70 keV gamma photons and the latter 2%, D. Counting Room and Shielding The Counting room, surrounded by 2k inch thick lowactivity poured concrete, measured 7>5 feet by 18 feet. The positioning of the guarded detector system in this counting room can be seen in Figures I3 and 14. The co\inting room also

PAGE 34

22 (0 o •H H Pi :3 o -p o s: p, s: c «H Oti P o o 0) H ft E d CO to

PAGE 35

23 0) a* •r-\ c s: o a> p OS § CO ^4 O «M -P U o Pi d) H ft a 08 CO bO

PAGE 36

24 C o 0) -p +i o a o o p. 0) rH (X B 03 CO H

PAGE 37

25 0) o c 00 u -p c
PAGE 38

26 a o o u to o

PAGE 39

27 houses a 4x4 Inch Nal(Tl) system. Ganma exposure as measured on a Nuclear Chicago model 2650 gamma survey meter was 0.01 mR/hr in the counting room and five times higher outside of the room. Mercury was used to shield the guarded Ge(Li) system since it has a higher mass number, greater density, and lower activity than lead. Five thousand pounds of mercury in 2.5 inches thick double wall tanks (approximately equivalent to k inches of lead) surround the guard detector and its photomultiplier tube (see Figure 15). With the shield, the guard detector measures 36 inches diameter by 31 inches long. Due to inherent activity in construction materials, the detectors, and the mercury, additional shielding is not significantly advantageous. As protective measures against mercury toxicity, a ventilator system and seamless floor were installed in the coimting room. E. Electronics The complete system electronics is block diagramed in Figure I6. Each of the detectors has its own preamplifier, linear amplifier, and threshold detectors which drive coincidence and anticoincidence gating. In addition, the Ge(Ll) detector electronics include an ORTEC model 120-23 preamplifier (supplied with the detector), a model 485 linear amplifier, and a model 444 biased amplifier. Except for these, the ORTEC high voltage supply, and a Packard Model 115 ^00 channel rr.ulti channel analyzer (KCA) (shovzn in Figure 1?) the electronics were

PAGE 40

28 c •H -d iH 0) CO U ^ rH ft O -P o A< H

PAGE 41

29

PAGE 42

30 ^^ m H (A § a
PAGE 43

31 custon! designed and built in the laboratory. The signal from the Ge(Li) crystal after amplification in the 120-2B preamplifier is split, being amplified by the ^85 linear amplifier and 44^ biased amplifier before entering the ViCA, and being amplified, sensed by a threshold detector, and then given control of the gate leading to the anticoincidence input of the MCA. Similarly, the guard detector and sample detector have their own identical preamplifiers, linear amplifiers, and threshold detectors. m the sample-guard mode, these two threshold signals are combined by an OR gate which feeds the Ge(Li) Ara gate controlling input to the MCA anticoincidence input. In the coincidence sample mode, only the guard threshold feeds the Ge(Li) AND gate while the sample threshold controls the coincidence input of the biased amplifier. The effect of this circuitry for the sample-guard mode is to send an anticoincidence control pulse to the anticoincidence input of the MCA only when the Ge(Li) detector yields a signal above its noise level and the guard detector or the sample detector yields a signal above its respective noise levels. The effect for the coincidence sample mode is to allow amplification of the Ge(Li) signal only when the sample detector yields a signal above its noise level and to send an anticolncident pulse to the anticoincidence input of the MCA only when the Ge(Li) detector and guard detector yield signals above their respective noise levels. The circuit diagrams for the electrons of the three detectors, the coincidence units, control pulse generators.

PAGE 44

32 and cable drivers are given in Appendix A.

PAGE 45

CHAPTER IV Nal(Tl) SYSTEM In order to better show the advantages of a guarded Ge(Li) system, comparison is made with a NaI{Tl) system. In this chapter both the Nal(Tl) configuration and its associated electronics are presented. A. Confi'^uration The Nal(Tl) system used as a standard for comparison consists of a Harshaw ^x4 inch Nal(Tl) crystal viewed by a photomultiplier tube and positioned in the center of a 2 foot cubic, 2 inch thick lead shield lined with Cu and Cd (^ee Figure 18). The top side of the shield is supported on rollers to allow access to the detector crystal (see Figure 19 ). The shield and detector system occupies the rear corner of the low level counting room as shown in Figure 20. The Nal(Tl) detector is customarily used to coiint samples in both 1.0 litre and 3.5 litre Harinelli beakers, in addition, samples in 1.0 litre "ice cream" containers and flat samples are frequently counted on this crystal. B. Electronics The electronics consists of a Power Designs Pacific model Kv-1565 high voltage power supply. Nuclear Data models 33

PAGE 46

3^ Cd and Cu ^^•' Lining Lead Shield 23"x23"x25" 2" thick ,-NaI(Tl) 1 .P M "Tube A Fig. 18.Nal(Tl) detector configuration.

PAGE 47

35 H m d en 0) OS p (0 o OS tH Oh

PAGE 48

36 o o u bD i o e
PAGE 49

37 KD-180F smalog to digital converter, ND-180M 512 chemnel memory unit, and ND-180R readout control unit, a TallyCorporation model 1506 paper punch, a Tektronix model RM503 oscilloscope, a Nuclear Data model ND-316 autofinger mounted on an IBM selectric typewriter, and a Dohrmann Instruments model 299 chart recorder. The system block diagram is shown In Figure 21. The signal after collection by the photonmltiplier tube is amplified, digitized and recorded in the appropriate memory channel. Upon command, the memory can either be displayed on the oscilloscope, plotted, or paper punched. The latter allows computer analysis of the data.

PAGE 50

38 0)

PAGE 51

CHAPTER V Nal(Tl) PERFORMANCE In this chapter the optimum performance of the kxh inch Nal(Tl) system in the lov:-level counting room as described in Chapter IV is stated. Included are system background, efficiency, minimum detectable activity, peak-to-Compton ratios, and ability to analyze complex spectrum, A. Detector Back^oiind The H-xH' inch Nal(Tl) crystal background count-rate was determined in the absence of a sample and with the shield in place. Since standard counting procedure in this laboratory uses it'O-minute counts, a 40-minute background was accumulated and is plotted in Figure 22. Only traces of the 0.35 MeV RaB and the 1.46 MeV K photopeal^s are visible in the spectrum. Table 1 lists the background count-rate grouped in various energy ranges and shows a cu-nulative background count-rate of just over 500 cpm for the 70 to 2500 keV range. B. Efficiency The absolute efficiency of the 4x4 inch Nal(Tl) crystal was measured for both point sources (1.0 cm from the crystal) and 3.5 litre sources with calibrated sources obtained from 39

PAGE 52

i^o o xi c o u o m CM CM o o ca laNKVHO/SuLNnOO

PAGE 53

^1 0) •p 4J ft ^^ o o o a o -p >i CO t & o H I I o u P o o > s p s •p ft C o o o I o o o o I o o 00 o o CM I o o 00 00 o o
PAGE 54

42 Baird Atomic and the Public Health Service. The plots of photopeak (full width at tenth maximum) efficiency vs. energy are shown in Figure 23 and include data from the Public Health (16) (17) Service and previous data taken on this system. The data yields essentially straight lines on log-log paper, showing increasing efficiency with decreasing energy down to a point where detector window absorption becomes significant. These data also show that the 3.5 litre sample efficiency is typically 1/3 of that for the point sovirce, the difference being attributed to differences in the average source-todetector distance. C. Minimum Detectable Activity The theoretical minim-urn detectable activity depends upon the background count and photopeak efficiency. It was determined for the 4x4 inch Nal(Tl) detector by requiring a 95^ detection certainty. This means that the theoretical minimum detectable activity equals two times the standard deviation in net count (gross minus backgroiind) under the photopeak energy spread, corrected for detector efficiency. For a 40minute count, the minimum detectable count rate in cpm equals 2 V (gross coimt -f background count) 40 which yields a minimum detectable activity in dpm of 21/2 X back-round count (under photopeak) ^0 (efficiency) Using the measured photopeak efficiency and background

PAGE 55

^3 90

PAGE 56

data, the theoretical minini"um detectable activity was calculated for nine radionuclides and photopeak energies. Figure 24 shows a plot of miniintuii detectable activity in gamna photopeak dpm vs. energy. In addition, the minimum detectable activity in pCi and the minimum detectable concentration in pGi/1 for several radionuclides of interest are listed in Table 2. D. Peak-to-Com-pton Ratio The peak-to-Compton ratio of a detector is important since it is an indication of the amount of interference between radionuclides in the complex spectrum. Figure 25 shows the peak and Compton relationship in a typical point-sotirce spectrum. Statistical variation in large Compton plateaus greatly increases the minimum detectable activity in that energy range. This uncertainty increases with large-volume samples due to Compton scatter within the sample before detection. Also in complex spectra • addition of Compton plateaus at the lov;er energies becomes very significajit. The peak-to-Compton ratio for the ^x^ inch Nal(Tl) detector was measured and plotted in Figure 26 for both point sources and 3.5-litre sources. The degradation in peak-toCompton ratio for the large sample can be readily seen to Increase at lower energies, S. Complex Spectrum Figure 27 shows a complex spectrum for the kxH^ inch Nal(Tl), The complex spectrum of ^^°'Aa. vias used in order to

PAGE 57

^5

PAGE 58

»A k6 H •s EH

PAGE 59

^7 QS is u o «« a ^ -p o a, CO o -p ft s o o § >!! OS 4) Pi aJ/VH iNAOD

PAGE 60

48 vo

PAGE 61

ii-9
PAGE 62

50 standardize comparison. It can be seen that both the poor resolution and the addition of Compton continuum raise the minimum detectable activity in the low-energy range of the spectrum to approximately the activity of the high-energy photopeak radionuclides in the spectrum.

PAGE 63

CHAPTER VI Ge(Li) SYSTEM PERFORMANCE In this chapter the performance of the standard Ge(Ll) system is presented. System background, efficiency, minimum detectable activity, peak-to-Compton ratios, and complex spectrum resolution are tabulated. A. System Background In order to show both shielding benefits and shielding problems, a 2if-hour unshielded backgroimd in the low-level counting room was taken with the Ge(Ll) spectrometer. The spectrum given in Figure 28 shows a complex collection of photopeaks from both radon and thoron daughters. Background with 2.5 inches of mercury shielding, given in Figure 29, shows a considerable reduction across the spectrum with fewer photopeaks being discernible. The background still contains small radon daughter photopeaks mainly due to their presence in the large air space between the shielding and the detector. Background with this air space filled with liquid is discussed in Chapter VIII. Integrated background, both shielded and unshielded, is tabulated in Table 3 for several energy ranges. 51

PAGE 64

52 a faO p o o lOUi 29*2 •.O^H 9C*Z Om, 019*1 O^H Z9Z'I otrx og-f?*! .O^H OZI'I 822-oV 966*0 822-oV 806*0 Oqi 02^*0 O^H 609*0 SBH ^62*0 g^H 2-t72*0 am 5ii*o 2^XSI-I ^21-0 C P( o to o a o o u bO o o • CM • 60 P4 (^OIX) SiNllOO

PAGE 65

53 0) B o o o o OMi 018*1 V "'''' OHH 021*1 oVq^-, 966*0 cvj O^H 609 '0 SBH 2-f7Z*0 -*o o o

PAGE 66

e o p m >s CO faO 5^ c^ CO H VPi (^ CM r^ t o H r-i (M VPi u^ H H (D w O CM CVJ 0^ CM CO c^ vn H

PAGE 67

55 B. Efficiency The efficiency of Ge(Li) detectors is much less than that of Nal(Tl) detectors, ranging up to only 20;^ as efficient as 3x3 Nal(Tl) for °Co. The Ge(Ll) detector used for this system was rated only 6% efficient (re 3x3 Inch Nal(Tl) Co source). To allow activity calibration, an absolute efficiency for the Ge(Li) detector was determined for both calibrated point sources and for 1.5-litre sources (configuration of the active-sample container). Both are plotted in Figiire 30, and essentially are straight lines on log-log paper with efficiency Increasing at decreasing energies. The 1.5-litre configriration is again about 1/3 as efficient as that for the point soirrce (1 cm distance) due to the difference in geometry factors for the two configurations. C. mnimum Detectable Activity The theoretical minimum detectable activity was determined as in Chapter V. For a ^0 -minute count, it equals in dpm 2H 2 X 3ackgroi;.nd count (under -phox^opeak) "kO X efficiency (of photopeak) Figure 31 shows the theoretical minimum detectable photopeak activity for the Jk cc Ge(Ll) detector with 2.5 inches of mercury shielding and shows a slight increasing sensitivity at lower energies. In addition, the theoretical minimum detectable activity in pCi and theoretical minimum detectable concentration in pCi/1 are tabulated in Table ^ for several radioisotopes of interest.

PAGE 68

56 J. V 9

PAGE 69

51 o

PAGE 70

58

PAGE 71

59 D. Feak-to-Coinpton Ratio Due to the high resolution of Ge(Li) detectors, their peak-to-Compton ratios are typically several ti^.es larger than those for Nal(Tl). The peak-to-Compton ratio for several radionuclide photopeaks was determined for the 3^ cc Ge(Li) detector and is plotted in Figure 32. Both pointsource data and 1.5-litre data are shown. Again it can be seen that the peak-to-Compton ratios for the large-volume sources are much less than for the point sources, being particularly worse at low energies. E. Complex Spectra To demonstrate the usefulness of the Ge(Li) detector for analysis of complex spectra, the complex spectrum for Ra is shown in Figure 33. It can be seen that the Compton continuum contributes most to the increase in the practical minimum detectable activity and not the photopeaks which are now only 3 keV wide. Thus the minimim detectable activity at low energies is comparable to the Compton Continuum and can be predicted by the cumulative activity of all radionuclide photopeaks of higher energies divided by the average peak-to-Compton ratio.

PAGE 72

60 OliVH

PAGE 73

61 0)

PAGE 74

CHAPTER VII GUARDED SYSTEM PERFOFiMANCE In this chapter the performance of the anticoincidence guarded Ge(Li) spectrometer is presented. As in Chapter VI, the system background, peak-to-Compton ratio, and minimum detectable activity are specified. In addition, factors determining guard-detector efficiency and reduction in photopeak efficiency are explained. A, System Background In the guarded mode, gamma-ray spectrum background is reduced due to the additional passive shielding of the liquid scintillator, the removal of Compton scattered events In the guard which are detected in the Ge(Li) crystal, and the removal of Compton scattered events in the Ge(Li) crystal which are detected in the guard. Figure 3^ shows a 2i^-hou^ background spectrum with the guard off. Comparison of this with Figure 29 (mercury shielded background) shows a decrease in the background at high energies but an increase in the background at very low energies (200 keV) when the liquid scintillator Is present. This difference is due to additional enercry removal in Compton scatter by the low-mass-number liquid, thus reducing the 62

PAGE 75

63 1-1 2i O o ^1 w o as ti
PAGE 76

6k energy of the backc-round gamma photons. Figure 35 shows the effect of anticoincidence cruardlng on the background spectrum. A comparison of background Is tabulated In Table 5 and in general shows a guarded background reduction of approximately 6 times over the unguarded for all energies. The photopeaks remaining in the spectrum are due to the unshielded J% solid angle geometry of the sample port and the activity within the detector system. The K photopeak in the spectrum corresponds to a point-source activity of k^ pCi and the Ra daughter photopeaks correspond to an activity of 10 pCl . Thus^ this sets the lower limit on the detectable activity for these two radionuclides. B. Guard Detector Efficiency The main purpose of the guard Is to detect Ge(Ll) crystal scattered gamma photons. Interaction in the guard detector may deposit only part of the photon's energy. Thus for the highest detection efficiency, the threshold of detection for the guard should be as low as practical. This was determined by allowing a photopeak reduction of 10^ due to unrelated chance pulse counts in the guard detector system and corresponds to a background count rate of 2,^1-00,000 cpm. Appendix B gives the background spectrum of the guard system and shows the threshold corresponding to 2,400,000 cpm to be deep in the noise region of the photomultiplier tube. With the guard threshold set at this point (approximately 15 keV) its efficiency was determined and plotted in Figure 36.

PAGE 77

65 4J c O O (•ATuba 'c^d ^od ^+7) o^-:« 09+7*1 lOBK 021*1 O^H 609*0 (•A-fnba 'c^d tod 01) a^K Z^£'0 9«a 2+?2*0 r o .0 CVJ CM L o : ,0 -o 00 A9^ 02 PI0t{S9J:L{,T pasno o o 00 o o o o CQ w o c o f-l o a 0) o -d
PAGE 78

66 n u o \^^ 0^ r-i o csj CM 00

PAGE 79

(^1 o

PAGE 80

68 Because of the hip-h background count, the efficiency was calculated by first measuring the photopeak reduction for one photopeak of a cascaded decay (2 photons) radionuclide 24 60 133 IJh (e.g. Na, Co, Ba, and Cs ) . After dead:;ime correction the ratio of the guarded to unguarded count for one photopeak gives the undetected fraction of the other gamma 133 13^ photon ( ^-^Ba and Cs having complex cascade decay required some approximations). The plot of efficiency shows a loss of efficiency with decreasing energy below 500 keV and a loss of efficiency with increasing energy above 700 keV. Losses in efficiency can be explained with a 3^ energy independent geometry loss by the sample port, low energy loss due to a 15 keV detection threshold (loss E(keV)/15, see Appendix C) and the decrease in attenuation coefficient for high-energy gamma photons. The reduction in Compton continuum is directly dependent upon the guard detector efficiency except at low energies where the Ge(Ll) detector mounts and casing absorb some of the crystal scattered gamma photons, C. Photopeak Reduction Since for highest Compton suppression the guard detector threshold must be set as low as possible (deep in the noise region), an acceptable photopeak reduction was arbitrarily chosen as 10^. A 10^ reduction while not substantially affecting counting statistics or efficiency does greatly increase the allowable guard detector backgroimd and thus allows significant lowering of the threshold setting.

PAGE 81

69 The guard detector background count rate which gives 10^ chance coincidence is ^0,000 coiints per second, D. ]yininum Detectable Activity The minimum theoretical detectable activity was determined as in Chapter VI. However, due to reduction in backgroiind count for the guarded spectrum, ^0-minute counts gave backgrounds below 1 count per channel. When this occurred, Poisson statistics were used in determination of the minimum detectable activity (3 is the minimum photopeak count / -1 C N for 95'^ confidence in the extreme of zero bkg. ), Figure 37 gives the calculated minim\im theoretical detectable activity in terms of gamma photopeak decay for both point sources and 1.5-litre sources as a function of energy. In addition Table 6 gives the theoretical minimum detectable activity, and minimum detectable concentration for several radionuclides for both ^0 minute and 24 hour counts. E. Peak-to-Comptcn Ratio The peak-to-Compton ratios for the anticoincidence guarded 3^ cc. Ge(Ll) was determined as in Chapter VI for the unpruarded Ge(Li). The guarded peak-to-Compton edge 65 ratio was 9 times higher for Zn and 8 times higher for 137 ^'Cs. Figure 38 shows the guarded peak-to-Corapton ratios as a fimction of enerc-y for both point sources and 1.5-lltre water sources. The peak-to-Compton ratio can be seen to be quite constant above 200 keV. The low-energy fall-off is

PAGE 82

70 1

PAGE 83

71

PAGE 84

72
PAGE 85

7^ due to both loss of efficiency in the guard system and absorption in the Ge(Li) crystal casing, F, Conplex Spectrum The usefulness of the anticoincidence guarded Ge(Li) detector for a complex spectrum is shown in Figure 39. -226 Radium was used to show both the separation of the many photopeaks and the small amount of Compton continuuTi present, It can be seen that the minimum detectable activity doesn't significantly increase for complex spectra.

PAGE 86

7^ (D -P O C ^'— P p ^ o o u o CO P ^ rH C C\J • ^ 03 ^ ff! O^H -f/OZ ^ ri o O^H 8^8*1 0«H 8E-i*T -^ CO O^H C^^'x E pi o ft (0 SBH 25C*0 ai3H ?62'0 O^K 021*1 OBH 69^*0 O^H 609*0 Z3 a^H ^^7^*o — 922--8H 981'0' \ CQ 1:3 J o o o H o o la^^vHo/iNnoo o >< 0) ft s o o ^

PAGE 87

CHAPTER VIII ACTIVE -SAMPLE SYSTEM In this chapter the perforn:ar.ce of the anticoincidence gruarded system with active-sample guarding is presented. Since the active-sample technique uses the same geometry as the 1.5 litre samples reported in Chapter VII, only differences in performance are presented. These include efficiency, peakto-Compton ratios, minimum detectable activity, and separation of decay modes. A. Efficiency The efficiency of the active sample geometry is the same as that reported in Figure 30 for the 1.5"lltre samples. Again, as with the standard' anticoincidence guarding, random coincidence reduces the photopeak count. The guard again is set for a ^0,000 count per second background while the sajnple detector is set for a 30,000 count per second background. The combined guarded background yields a 13% photopeak reduction as compared to a 10^ reduction with guard shielding only, B. Nininium Detectable Activity The single radionuclide minim\im detectable activity for the active sample technique is improved only at energies below 75

PAGE 88

76 500 keV and then by only about 15^ due to a 30^ improvement in background suppression. The minimum theoretical detectable activities, determined as in Chapter V, are plotted in Figure ^0 as a function of photopeak energy. In addition, the minimum theoretical detectable activities and concentrations for several radionuclides are tabulated in Table 7 for both ^0-minute counts and 2i»— hour counts. The Increase in the minimum detectable concentration is due to the required dilution of the sample with liquid scintillator, C. Peak-to-Corapton Ratio The peak-to-Compton ratio for the active-sample technique is shovm in Fig\Are ^1 as a function of energy. The peak-toCompton ratios are restored to 95% of their point-source ratios at high energies but only restored to ^0% at energies below 100 keV, Hov;ever improvement in peak-to-Con:pton ratio is still approximately 2-fold across the spectrum, D, Separation of Decay Modes In addition to improving peak-to-Compton ratio, the active-sample technique removes all short-lived /S"*" and fi decay excited-daughter radiations from the spectrum. By requiring coincidence between the sample-detector and the Ge(Ll) detector, these short-lived states can be left in the spectrum while removing all electron capture and metas table states. Figure k2 shows this division of states; however to show the effect X'llth available radionuclides (positron

PAGE 89

11 O U o Vi >» p -p o OS H -P O < o -aXT, o v> c WJQ >IV3dOiOHci

PAGE 90

78

PAGE 91

79 0) iH E CB
PAGE 92

80 •0*3 ^66 ?9-uz 6lI'I •D*a %ooi ^ I7CI-SO 509 uoT:;Bi-[qTUUV 1X5 '0*3 ^001 GCi--ea OQC'O OOG'O 9^2*0 CCI--BH sgo CO 0) -d o e o 0) -d o o OS ^1 08 ft
PAGE 93

81 and cascade decay ) the guard v/as disconnected for the coincidence mode.

PAGE 94

CHAPTER IX COMPARISON OF SYSTEMS ON EInTVIRONMENTAL SAMPLES In this chapter the spectra produced with the various systems described previously are compared for their ability to identify radionuclides. Included are prepared samples and actual environmental samples in various forms, A. Standard Samples To show the advantages of the active-sample technique for water samTDles, two standards were prepared. The first, containing only Zn, was used to produce the spectra in Figure ^3. The upper two spectra are from a guarded and rmguarded point source; the guarded spectrtun shows a Compton edge suppression of a factor of 8. The remaining three spectra are for a 1.5 litre source. The normal guarded spectrum shows a Compton tail suppression of a factor of 7 while the active-sample guarded spectrum shows a suppression of a factor of 23. This additional Compton tall suppression would permit recognition of a much smaller photopeak within this energy region. The other standard contained 6 radionuclides having different decay modes. The spectra are shown in Figure kk» In addition to the suppression of annihilation radiation. 82

PAGE 95

83 reduction Point Source Unguarded-Guarded (x7) (x8; OS 0) o •p o s: p. \ hvMr>-*>'^^^'-*t^ o o o reduction (x7) 1.5 Litre Source Unguarded > +^ 0) as S rH 13 a> p4 o -p o xi Pi V^vvVv'K'JuAv•^fVw^^^^^^^ Guarded (x4.5) (xi^) uJ^ reduction (x23) Active Sample (x8) l' l> .4 . iMJi I / l ^. iia J^ .ia l l CHANNEL NUMBER (x5.3) Q) ft O P o p< 0) p. o p o ft ^00 Fig. ^3. Zn-65 sourcsj spectrum.

PAGE 96

8i+ 34cc Ge(Li) Sample-1.5 litre

PAGE 97

85 cascade events, and the Compton tail in the normal guarded mode, the anticoincidence sa-ple-guarded mode suppressed the Compton tail another factor of 3* Also the coincidence sample mode suppressed electron capture and metas table states while restoring the rest. B. Air Sample To demonstrate the usefulness of the normal guarded mode, an air sample was taken of it-5,000 cubic feet of atmospheric air and allowed to decay for 3 days. Comparison of kO minute count spectra for Kal(Tl), \mguarded Ge(Li), and guarded Ge(Li) shows no identification of any radionuclide in either the Nal(Tl) or unguarded Ge(Li). However, data analysis on the guarded Ge(Li) spectrum indicates the presence of 100 pCl of "^Be, 8 pCi of ^°^Ru, and 12 pCi of "^^Hh (all activities determined from digital data see Appendix D). C. Water Samples To demonstrate the advantages and disadvantages of the active-sample technique, two environmental water samples were counted on the various systems. The first sample was an environmental standard sample prepared by the Public Health Service and its spectra are shown in Firnre 46. The Kal(Tl), unguarded Ge(Li) and guarded Ge(Li) spectra all show the presence of 200 pCi of "^ Ru and 200 pCi of Ce although their presence in the guarded Ge(Li) is easiest recognized. The active-sample mode does not show their presence due to the dilution of the sample with scintillator.

PAGE 98

86 40 20 Nal ( Tl ) with background subs traction 200 Unguarded Ge(Li) ij-5,000 cubic feet 3 day decay 40 min. count ^\ikUtMi^4J.u^, Mt-Ai AViVt>i>xA,*««>^<.<-i^L,^*J^ 400 o o p, p, o cc 2 ^Guarded Ge(Li) O vPi O-H ON II I — CD :3 ,o ^ PQ cd so Ph rON NO H 800 fy-t '/lMlrHlrt
PAGE 99

87 300 Nal ( Tl ) with background substractio o 20 3.5 litre yi^vqf^^ Unguarded Ge(Li) 1.5 litre ^'^^^'^''"^^V>>'*w^^->>T^'^r'V-^f^ O ^'^^'•^A^'*"*V ^00 13 O 0) o "^ P^ Guarded Ge(Li) cc 0, ^ o cv o 1.5 litre r-\ o O »\ 400 20 Active Sample \. 30'^ reduction Water standard from PHS kO mln. count 0.7 litre 400 CHANNEL NUMBER 400 Fi^o 46. Water standard spectrum.

PAGE 100

88 The other environmental water saniple was taken from the University of Florida Training Heactor coolant water. The spectra shown in Figure H-7 indicates the presence of 20,000 pCi of activated Na. The Nal(Tl) spectriim is completely 2k swamped by the Na, however the unguarded Ge(Li) clearly Identifies the presence of 1,400 pCi of ^^°^Tc, The guarded Ge(Ll) spectrum also shows the presence of 700 pCi of ^In and ^Cd. The active sample mode spectrum, because of the oh, removal of Na ( a short-lived/^ ) and its Compton contlnuTjim, shows the presence of 13 pCi of -^^^ga. and 90 pCi of °^Zn (both contamination in the sample container). It should be Oh. noted that the Na Compton continuum is less than the background in the active-sample spectrum and thus does not raise the minimum detectable activity in that region. D. Solid Samples To further show the usefulness of the guarded Ge(Ll) system three environmental solid samples were analyzed. The first, whose spectra are shown in Figure 48, consists of an 800-gram soil sample taken from undisturbed top soil. All three spectra for this sample; the Nal(Tl), the unguarded Ge(Li), P ^ A and the guarded Ge(Li) show the presence of 1200 pCi of Raand its daughter products^ 500 pCi of Cs , and 400 pCi of ^K. The Nal(Tl) spectrum shows considerable interference between these radionuclides, especially between the 662 keV -^ ' Cs peak and the 609 keV RaC peak. The guarded Ge(Li) greatly reduces peak-to-peak interference and Compton-to-peak interference

PAGE 101

89 Nal ( Tl ) 8000 2000 IQOO 1000 :3 o o } rH I iH O rH P, I Xi O o o ^MUiHiiJUr 100 Unguarded Ge(Ll) -— «

PAGE 102

90 Nal ( Tl ) with bkcr. substraction 100 50 100 o o o i^O ^0 Ge ( Li ) Unguarded Soil sample 800 gram 40 min, count '^>^Hvwv^ iWi L 400 ^ Guarded o 03 CEl O cv H O o I o OD O ^^ O rH H rH — Jl i>» I !. .. .L A . .».-. ..^. 1 ..>, iJiOO 800 CHANNEL NUI-:BER Fig. 48. Soil sample spectrum.

PAGE 103

91 making the identification of these radionuclides simpler. The second sample consists of a ^70-grani oyster sample taken from Crystal River. Its spectra, given in Figure 49, shov: no radionuclides visible in either the Nal(Tl) spectrum or the xinguarded Ge(Li) spectrum. However the guarded Ge(Li) spectrum shows the presence of 200 pCl of 40 „ m 226 K, 18 pCi of -^^-^I and I50 pCi of Ra. The third solid sample consists of a 720-gram sample of seaweed taken from Crystal River. Its spectra shovm In Figure 50^ show the presence of 700 pCi of "^K recognizable in all three spectra, and 320 pCi of "^Be, 80 pCi of ^-^^Th and 250 pCi of '^'^°Ra recognizable only in the guarded spectrum.

PAGE 104

92 Nal ( Tl ) 100 v^/Y^nr^/\Jlv^ n^ ^ ^^ ^ /^r^^ ^^^ ' ^-f^ Q Oyster sample 470 gram ^0 min. count o En O o 60 40 20 40 20 JGe(Ll) Unguarded 400 800 O CQ(H -H p. aJ I O KM P< o vr>CJ^ 00 H vtnO H Ge(Ll) Guarded ,nM'kUfcLM»-w>»«i»^.n ....t..-^ ^ -H I o tri ft o o vO O X »• m mt — t J !>*> ><» » 4i *>'«>i «H <400 CHANNEL NUKBER 800 Fig. 49. Oyster sample spectrum.

PAGE 105

93 Nal ( Tl ) with bkg. substraction 100 -0^ 100 Seaweed sample 720 gram 40 mln. count 2 o Eh o o Ge(Ll) Unguarded }|fJ^^1t^iAh ^ )^^ ^ ^Mi *k» ^> ^ ^^*if^Hml^A ,, ! ..». -i.i\.> Ge(Li) Guarded * j i i f i^ji h ,}\> imHo » . 4*» CM -H x: o H PL, 00 O O OD • o 800 I o o o vO O ll I » I llMIII lll>t»« » 400 CH/iMNEL NUMBER "4 00 Fig. 50. Seaweed sample spectrum.

PAGE 106

CHAPTER X EVALUATION OF DETECTION SYSTEMS In this chapter the advantages of each system are enumerated. In addition, the relative usefulness of each system is explained. A. Nal(Tl) System The ^x^ inch Nal(Tl) system has the advantage of high detection efficiency (approaching the efficiency due to geometry factor alone below 400 keV) which allows less statistical error for short counting times due to the higher counting rate. The 4x4 inch Kal(Tl) system therefore has a lower single radionuclide theoretical minim\im detectable activity than the unguarded, relatively low efficiency 34 cc. Ge(Li) system. However, due to the background reduction in the guarded 34 cc. Ge(Li) detector, the 4x4 inch Nal(Tl) system is superior for the detection of only a few radionuclides of low activity concentration (two adjacent radionuclide associated photopeaks appear as one in the Nal(Tl) and the Nal(Tl) system accepts larger samples). The poor resolution of the Nal(Tl) detector lends itself to misldentification of photopeaks and the masking of photopeaks by the large Compton continuum of others. Thus, minimum detectable activity in the presence of a large amount 94

PAGE 107

95 of high-energy cacnia decay radionuclides is preatly increased. Therefore, in complex spectra the ^x4 inch Nal(Tl) detector may not have a lovzer minimtun detectable activity than the tmguarded 3^ cc. Ge(Li). B. Unguarded Ge(Li) System The -unguarded Ge(Li) system although having a higher minimum detectable activity than the Nal(Tl) allows the pinpoint identification of photopeak energies to vrithln 2 keV. This^ along with its higher peak-to-Compton ratio^ allows more reliable identification of radionuclides than the I"aI(Tl), especially in complex spectra. C. G\iarded Ge(Li) System The guarded 3^ cc. Ge(Li) system has the advantages of a higher peak-to-Compton ratio (8 times better than the unguarded Ge(Li) and l6 times better than the Nal(Tl)) and a reduction of background (a factor of 6 over the unguarded Ge(Li)). This large background reduction gives a lower minimum detectable concentration by a factor of 2 and a lower minimum point-soiorce minimum detectable activity by a factor of ^ over that for the 4x4 inch Nal(Tl) for long co tinting times, but approximately the same m.inimu'r detectable activity for 40 minute counts (due to the required change to Foisson statistics for low counts). In addition, the exceptionally high peaLk-to-Compton ratios allo'.:s the maintaining of this good minimum detectable activity in cor:plex spectra.

PAGE 108

96 HoT-rever, due to the nature of anticoincidence guarding, cascade decay events and Z?"*" decay events are suppressed. Thus the usefulness for identification of such decay events is greatly reduced (50-fold for 0"^ and. 7-fold for cascade emitters) D. Active-3aTT:ple System The active-sample system restores the point-source peakto-Compton ratio for large water samples, in complex spectra with the presence of large amounts of radionuclides, this additional Compton suppression improves the minimuni detectable activity below 200 keV by a factor of 2 in spite of the required 2-fold dilution of the sample. In addition, the technique produces two simplified spectra which can aid in the identification of the radionuclides or selectively suppress Interference due to either of the two classes of decay mentioned previously.

PAGE 109

CHAPTER XI CONCLUSIONS Anticoincidence gruarding of a 3^ cc. Ge(Ll) crystal has shown itself superior in its minimuin detectable activity and minimum detectable concentration over an unguarded kxk inch Nal(Tl) detector for counting times as short as kO minutes. For 2^-hour counts, its minimum detectable point-source activity is roughly 4 times better than the unguarded kxUinch Nal(Tl). In addition the excellent peak-to-Compton ratio of the guarded Ge(Li) practically eliminates radionuclide spectral interference. The active-sample technique, while showing only slight advantage in minimum detectable activity at low energies for typical environmental samples, shows particular promise in the selective removal of spectral counts of large quantities of certain radionuclides from the spectrum. . This removal process could be described as gamma-beta anticoincidence counting and gamma-beta coincidence counting. Several improvements can be made in the guarded Ge(Li) system to further reduce its minimum detectable activity. First an increase in Ge (Li) detector efficiency would Improve the minimum detectable activity by the square root of its relative improvement as would an Increase in guard efficiency. 97

PAGE 110

98 By employing a 2.^% efficient Ge(Li) detector, using beryllium instead of aluminum in the detector and sample port, reducing the sample port diameter to tliree inches, and using bi -alkali photocathodes in the guard detector photomultiplier tube (or cooling to 0°C to reduce phototube noise), an improvement in minimum detectable activity by a factor of h and an improvement in the peak-to-Compton ratio of the same factor could be achieved. Thus peak-to-Compton ratios could approach 800 and minimum detectable activities approach the 0.1 pCi level (for ^'Cs point source 2^ hour count).

PAGE 111

APPENDICES

PAGE 112

APPEIJDIX A ELECTRONIC CIRCUITS The active-sample anticoincidence guarded Ge(Li) spectrometer system required the design and construction of additional electronics including preamplifiers, threshold detectors, coincidence gate circuits and cable drivers. This appendix contains the circuit diagrams for the above required electronics. Figure 51 gives the circuit diagram for the sample-detector preamplifier, cable driver, threshold detector, and control pulse shaper. Figxxre 52 gives the circuit diagram for the guard-detector preamplifier, threshold detector, and control pulse shaper. Figxire 53 gives the circuit diagram for the Ge(Li) detector timing preamplifier, threshold detector, and control pulse shaper. The coincidence and anticoincidence con'ferol circuits along with the pulse generators to drive the biased amplifier coincidence gate and the multichannel analyzer anticoincidence gate are given in Figure 5^» 100

PAGE 113

101 o w o •H C o ^1 +J o 0) iH (D U O +^ O
PAGE 114

102 0) o c (D o c o o o HI' -p

PAGE 115

c o O -H o o 103 CO o c o u 4-1 O 0) rH
PAGE 116

10^ !> CM r-i + rii'
PAGE 117

AFPEMDIX B GUARD DETECTOR CHARACTERISTICS In this appendix the data leading to the determination of the guard detector energy threshold are presented. In addition the procedures used to determine this threshold are outlined. Figure 55 shows two guard-detector gamma spectra taken 137 with a Cs point source. Since one spectrum vms taken with the source positioned 8 inches from the phototube-end of the cylindrical guard detector and the other with the source positioned the same distance from the opposite end of the guard, the geometry factors are approximately the same and thus differences in the spectra are due to light absorption, reflection and collection. The similarity of the two spectra indicate". very adequate light collection and very little position dependence throughout the guard. Figure 56 shows the background spectrvim of the guard with the guard detector threshold so marked. By requiring the integration of all channels above the threshold to yield ^0,000 counts per second, the threshold channel could be determined. Figure 57 shows an attempt to energy-calibrate the en threshold with a point soujrce -^'Co standard. By extrapolating the leading edge of the spectrum, an approximate channel 105

PAGE 118

106 p"[oqs8j:qq. jlqzAi'^uy—

PAGE 119

107 > o o CM II H <0 OS s: o PIOLis8aL[:^ aortoac^ap pj:Bn3 a^^i 91 PIoi{ssai{c^ j.qzJ:it3ut3 as>[ 17 — -

PAGE 120

108 o o

PAGE 121

109 calibration of the photopeak energy is deduced. Comparison of the integration below the threshold to the total spectrum integration gives the fractional loss for ^'Co. The effective energy threshold of the guard detector is 15 keV and the effective fractional count loss is '^/E(keV).

PAGE 122

APFEIIDIX C SPECTRW! ENERGY CALIBRATION In this appendix the data used for the energy calibration of the analyzer display are presented. Included are the spectra of the standard radioisotopes used for calibration (see Figure 58) and the energy versus channel nimber calibration curve (see Figure 59). The slope of this calibration curve shows a 1.85 keV per channel energy division with no zero channel offset. The biased amplifier was used to produce an 800 channel spectrum in two runs with a ^00 channel MCA, 110

PAGE 123

Ill (OSC) 09-00 2CC*I (^82) Z2-^N ?-i2*I (^Cz) 09-00 CZ.1'1 (I^) 17?-"W ^C8*0 (161) g^H Zii'O B 3 fH P O Q) P. CO 5: O •H -P flfl ^1 03 o -H ISNNVHO/tLNnOO

PAGE 124

112 1600t lU-W: 128CH 1120 100 200 300 100 CHANNEL NUMBER 200 300 ^00 i''ig. 59 Ge(Li) calibration curve.

PAGE 125

APPENDIX D EXAMPLE OF DIGITAL DATA Because of noise in the plotter mechanism, some spectrum plots do not clearly show the photopeaks present. In this appendix a sample of the digital data is presented to show that this problem is indeed in the plotter. Table 8 gives the digital data for the guarded air sample spectrum (Figure ^5) between ^00 keV and 800 keV. Three peaks can be easily distlnguised the 0.^77 Be-7, the 0.-^97 Ru-103, and the O.765 Kb-95. 113

PAGE 126

11^ CVJOOOOrHOOiHOHOrHOOOOCMO OOOOOOOOOOOOOOOOOOO OrHOHOrHOrHOOJOrHCNJOOOOrHO OOOOOOOOOOOOOOOOOOO

PAGE 127

LIST OF REFERE?TCE3 1. Conally, R. E. "Two-Crystal Gamma-Hay Scintillation Spectrometer," Review of Scientif ic Instruments 2k, ^58 (1953). 2. Evans, A. E. , B. Brovm and J. B. Marion. "Anticoincidence Shielded Gamma-Ray Spectrometer for Nuclear Reaction studies," Revie:-: of Scientific Instruments (August, 1966). 3. Cooper, J. A., N. A. V/agman, H. E. Palmer, and R. W. Perkins. "The Application of Solid State Detectors to Environmental and Biological Problan," Battelle >!emorial Institute Report, BN^.-JL-SA-llO-^ (I967). 4. Hill, M. W, "An Anticoincidence-Shielded Ge(Li) GammaRay Spectrometer," Nuclear Instruments an d Methods, 21, 350-352 (1965). 5. Cooper, J. A. "Anticoincidence-Shielded Ge(Li) GammaRay Spectrometer for High Sensitivity Counting," IEEE Transactions on Nuclear Science . NS-15 , No. 3, 4-0?^12 (June, 1968). 6. Cooper, J. A., L. A. Rancitelli, R. W. Perkins, W. A. Haller, and A. L. Jackson. "An Anticoincidence Shielded Ge(Li) Gamjna-Ray Spectrometer and Its Application to Neutron Activation Analysis," Battelle Memorial Institute Report, BN'>VL-SA-2009 (I968) 7. Phelps, P. L., K. 0. Hamby, B. Shore, and G. D. Potter. "A Ge(Li) Gamma-Ray Spectrometer of High Sensitivity and Resolution for Biological and Environmoi tal Counting," In Radionuclides in the Environment, Advances in Chemistry . Series 21/202-230, (I97O). 8. Latner, N. and C.G. Sarderso. "The HASL Ge(Li )-NaI(Tl) Low Level CountinSystem," I ES5 Transactions on Nuclear Science . NS-I9 . No. 3, 141-150 (February, 1972), 9. Zimmer, I-.'. H. "An Anticoincidence Guarded Ge(Li) Detector for Environmental Level Gamma Energy Analysis," I SEE Tr a nsactions on Nuclear Science, NS-19 . No. 3, 151-154 (February, 1972). 115

PAGE 128

116 10. Phelps, P. L. and K. 0. Hamby. "Experience in the Use of an Anticoincidence Shielded Ge(Li) GamzTia-Hay Snectrometer for Low Level Enrironniental Radionuclide Analysis," ISSS Transactions on Nuclear Science . NS-19 . No. 3, 155-165 (February, 1972). 11. Chlwiller, R. W. "The KcClellan Central Laboratory's Anticoincidence Ge{Li) Spectrometers for Low Level Applications," lESE Transactions on Nuclear Science . NS-19 , No. 3. 166-171 (February, 1972). 12. Miller, G. L. "A Brief Review of Recent Advances in Compound Semiconductors for Radiation Detectors," IEEE Transactions on Nuclear Science , NS-19, No. 1, 251-259 (February, 1972). 13. Price, W. J. Nuclear Radiation Detection . New York: f'^cGraw-Hi 11, 196^. ~ 14. Webb, R. C, M. G. Kauser and R.E. Mischka. "Response of a Mineral Oil Eased Liquid Scintillator to Heavily Ionizing Particles," PPAH, 26 (June, 1970). 15. Porter, C. R. "Annual Report of the Eastern Environmental Radiation Laboratory, " EFA Report SERL 71-4, 17-18 (December, 1970). 16. Douglas, G. S. Radioassay Procedures fo r Environmental Samples , Public Health Service Publication No. 999RH-27 (January, I967). 17. Bolch, V,'. E. "Environmental Surveillance for Radioactivity in the Vicinity of Crystal River Nuclear Power Plant: An Ecological Approach," quarterly proEress report November 1^ I970 January 31, 1971. University of Florida. 18. Currie, L. A. "The Meas\irement of Environmental Levels of Rare Gas Nuclides and the Treatment of very LowLevel Counting Data," IEEE Transactions on Nuclear Science , N3-19 . No. 1, 119-126 (February, 1972).

PAGE 129

BIOGMPHICAL SKETCH Frank Russell Karkwell was born in Miami, Florida, on November 7, 19^8. In August, 1969f he received the degree of Bachelor of Science in Nuclear Engineering Sciences at the University of Florida. In September, I969. he enrolled in the Graduate School of the University of Florida and in March, 1971i received the degree Master of Science in Engineering. Since that time he has continued his work In nuclear engineering toward the degree of Doctor of Philosophy in Engineering. During his graduate study, he has held first, intermediate, and terminal Special Fellowships in Nuclear Science and Engineering granted by the United States Atomic Energy. Commission, During his study at the University of Florida, he has been elected to membership in the societies, Tau Beta Pi, Sigma Tau, and Phi Kappa Phi. In addition, he is a member of the American Nuclear Society and the Institute of Electrical and Electronics Engineers. He presently holds a Second Lieutenant's commission in the United States Army Ordinance Branch, 117

PAGE 130

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. William E. Bolch, Ph.D. Chairman, Assoc. Prof, of Environ. Engineering Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope , and quality, as a dissertation for the degree of , Docto;?^6j| Philosophy. Mihir^nVtr. "^effan^Lan, Ph.D. Prof, and Chairman of Nuclear Engineering Sciences I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Charles E. Roes-^ler, Ph . D . Asst. Prof, of Environmental Engineering Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. :^^^^ .am H. Ellis, Ph.D. Assoc. Prof, of Nuclear Engineering Sciences

PAGE 131

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. iLU G. /^. jUhU Herbert A. Bevis, Ph.D. Assoc. Prof, of Environmental Engineering Science This dissertation was submitted to the Dean of the College of Engineering and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December, 1972 Dean, College of Engineering Dean, Graduate School

PAGE 132

/^ kJ v .V \ 9> ^ %•