Clemson University

TigerPrints
All Theses Theses

12-2006

A comparison of the ZnS(Ag) scintillation detector

to the silicon semiconductor detector for
quantification of alpha radioactivity in aqueous
solutions
Jennifer Foster
Clemson University, [email protected]

Follow this and additional works at: https://tigerprints.clemson.edu/all_theses

Part of the Environmental Engineering Commons

Recommended Citation
Foster, Jennifer, "A comparison of the ZnS(Ag) scintillation detector to the silicon semiconductor detector for quantification of alpha
radioactivity in aqueous solutions" (2006). All Theses. 10.
https://tigerprints.clemson.edu/all_theses/10

This Thesis is brought to you for free and open access by the Theses at TigerPrints. It has been accepted for inclusion in All Theses by an authorized
administrator of TigerPrints. For more information, please contact [email protected].
 A COMPARISON OF THE ZnS(Ag) SCINTILLATION DETECTOR TO THE
SILICON SEMICONDUCTOR DETECTOR FOR QUANTIFICATION OF ALPHA
RADIOACTIVITY IN AQUEOUS SOLUTIONS

A Thesis
Presented to
the Graduate School of
Clemson University

In Partial Fulfillment
of the Requirements for the Degree
Master of Science
Environmental Engineering and Science

by
Jennifer E. Foster
December 2006

Accepted by:
Dr. Timothy A. DeVol, Committee Chair
Dr. Robert A. Fjeld
Dr. John T. Coates
 ABSTRACT

A ZnS(Ag) detector was successfully used in the direct detection of alpha

particles from aqueous solutions and the results were compared to the passivated ion

implanted planar silicon (PIPS) continuous air monitor (CAM) detector. The ZnS(Ag)

detector is recommended for on-line detection of gross alpha radioactivity from high-

level liquid waste process streams; however, the detector suffers from limitations due to

variations in detection efficiency with alpha energies. The beta and gamma interference

did not significantly spillover into the alpha region of interest of the ZnS(Ag) detector

which provided absolute detection efficiencies of 7.19 + 0.13%, 5.37 + 0.02%, and 4.21 +

0.03% for 244Cm, 239

Pu, and 234
U/238U solutions, respectively. The detection efficiencies

were found with a 3.5 mg/cm2 ZnS(Ag) scintillation layer, 4.15 cm2 detection surface,

and a 5.2 mm air gap between the liquid surface and the detector. The absolute detection

efficiencies of the PIPS CAM detector for 244Cm, 241Am, 239Pu, and 230Th solutions in the

presence of beta and gamma activity were 4.61 + 0.01%, 3.87 + 0.01%, 2.84 + 0.01%,

and 1.32 + 0.01%, respectively, for a geometrical set-up similar to the ZnS(Ag)

experiments. Two ZnS(Ag) scintillator thicknesses were compared. The thicker

scintillation layer suffered from light absorption and scattering in the ZnS(Ag) layer

resulting in a lower energy signal. Despite the differences, the absolute detection

efficiencies of the two thicknesses were equivalent. The large detection area available to

the ZnS(Ag) detector proved to be valuable for the detection of radioactivity in liquids as

the absolute detection efficiency more than doubled for a surface area of 78.5 cm2 and
238
resulted in a minimum detection concentration of 0.32 Bq/mL for U solution for a
 iii

3600 second count. The detection capabilities and pulse height spectra of the ZnS(Ag)

detector were compared to the PIPS CAM detector as well as computer simulations and

theory.
 DEDICATION

In dedication to my father, who has always been my role model in the sciences.
 ACKNOWLEDGEMENTS

I would like to thank Dr. Timothy DeVol for his guidance throughout my research

project and also for the excellent opportunity to work under him. Mr. Matthew O’Hara

served as my mentor in many safety and laboratory practices and took time out of his

busy day to guide me as I acquired data and conducted experiments. I would like to

thank Dr. Oleg Egorov for his direction, ideas, and the opportunity to work at the Pacific

Northwest National Laboratory. I am grateful for the electrical engineering assistance

which Mr. Joel Simoneau was kind enough to grant. I appreciate both Dr. Robert Fjeld

and Dr. John Coates for being willing and helpful participants on my committee.

I would like to express gratitude to all of my classmates, in particular Ms. Amy

Attardo and Ms. Anna Cornelious, for constructive study groups and for making my time

at Clemson enjoyable. I am grateful for Mr. Thomas Fisher’s patient and knowledgeable

computer assistance and ideas. The most special acknowledgements go to my parents,

Mr. and Mrs. Michael and Joan Foster, for proofreading my thesis, providing many

words of encouragement, and giving their loving support in everything that I do.

This research was supported by the Environmental Management Science

Program, Office of Science (BER) at the U.S. Department of Energy. Grant # DE-FG02-

05ER63971
 TABLE OF CONTENTS

Page

TITLE PAGE ............................................................................................................. i

ABSTRACT............................................................................................................... ii

DEDICATION........................................................................................................... iv

ACKNOWLEDGEMENTS....................................................................................... v

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

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

LIST OF SYMBOLS AND ABBREVIATIONS ...................................................... xv

CHAPTER

1. INTRODUCTION ......................................................................................... 1

2. LITERATURE REVIEW .............................................................................. 3

Defense Waste ......................................................................................... 3

Alpha Particle Theory .............................................................................. 5
Scintillation .............................................................................................. 10
Zinc Sulfide.............................................................................................. 11
Alpha Radiation Monitoring Detector ..................................................... 15
ZnS(Ag) Spectroscopy............................................................................. 24
PIPS CAM Research................................................................................ 26
Computer Modeling ................................................................................. 35

3. RESEARCH OBJECTIVES .......................................................................... 44

4. MATERIALS AND METHODS................................................................... 46

Apparatus ................................................................................................. 46
Reagents and samples .............................................................................. 51
Procedure ................................................................................................. 53
 vii

Table of Contents (Continued)

Page

5. RESULTS AND DISCUSSION .................................................................... 54

Energy Resolution and Spectral Shape .................................................... 54

Absolute Detection Efficiency................................................................. 60
Solution Concentration ............................................................................ 64
Detection Area ......................................................................................... 65
Minimum Detectable Concentration........................................................ 67
Solution Density....................................................................................... 68
Air Gap Separation .................................................................................. 71
Thin versus Thick Scintillation Layer...................................................... 72
Computer Simulation ............................................................................... 77

6. CONCLUSIONS............................................................................................ 90

APPENDICES ........................................................................................................... 92

A. High Voltage Optimization...................................................................... 93

B. Signal Splitter versus Charge-Sensitive Preamplifier.............................. 94
Pulse Pile-up Background.................................................................. 94
Detection Dead Time and Recovery Time Background .................... 96
Instrumental Components .................................................................. 99
Experimental Results ......................................................................... 101
C. Visual Basic Computer Model................................................................. 114

REFERENCES .......................................................................................................... 125

 LIST OF TABLES

Table Page

2.1 Common properties of ZnS(Ag) scintillator [6] ............................................ 14

2.2 The counting sensitivities of 230Th, 235U, 239Pu, 241Am, and 244Cm
solutions using the PIPS CAM detector in direct alpha assay
using two different channel ranges: 200-900 and
550-900 channels [18].............................................................................. 30

2.3 The counting sensitivities for beta and gamma solutions, 99Tc, 60Co,
137
Cs, and 90Sr/90Y for the same channel ranges as Table 2.2 [18].......... 30

2.4 A comparison of alpha activity measurement between the

conventional radio-chemical method and direct alpha
measurement using the PIPS CAM detector [18] .................................... 34

4.1 The experiments which utilized the signal splitter or preamplifier ............... 49

4.2 The active depths and solution effective activities for various
radionuclide energies ............................................................................... 53

5.1 The minimum detectable concentration (MDC) for the various

areas and radionuclides ............................................................................ 68

5.2 A comparison of the computer-simulated stopping power of

water for alpha particles to theoretical values obtained
from the NIST ASTAR database [33] ..................................................... 81

5.3 The absolute detection efficiencies (%) for 239Pu aqueous

solutions for the ZnS(Ag) detector and PIPS CAM
detector experiments and computer simulations...................................... 87

B.1 The absolute detection efficiencies of the ZnS(Ag) detector

connected to both the signal splitter and the preamplifier ....................... 111

C.1 Constants used in the computer simulation ................................................... 118

 LIST OF FIGURES

Figure Page

2.1 The alpha particle absorption curve demonstrates the consistent

count rate and alpha range [5].................................................................. 7

2.2 The Bragg ionization curve for Polonium alpha particles [5]........................ 8

2.3 The stopping power of water [8].................................................................... 10

2.4 The energy band structure of an activated crystalline scintillator [6]............ 11

2.5 The extinction coefficient of ZnS versus photon energy for three
different ZnS thicknesses: 1) 100nm, 2) 200nm, and
3) 400nm [10] .......................................................................................... 13

2.6 The ZnS(Ag) pulse height spectra for 5.5 MeV alpha particles
and thermal neutrons [10] ........................................................................ 14

2.7 The flow through ZnS phosphor detector used to detect

contamination in low-level aqueous solutions [11] ................................. 16

2.8 The geometric diagram used in the derivation of the calculated

depth of alpha particle, X [12] ................................................................. 18

2.9 The flow through detection system set up with scintillation glass
detector and radioactive liquid [13] ......................................................... 19

2.10 The pulse height spectra of alpha and beta solutions using the
cerium doped SiO2 detector [13].............................................................. 20

2.11 The count rate for a 232Th soil sample versus Mylar® absorber
thickness [15] ........................................................................................... 23

2.12 The pulse height spectra for the total events of the ZnS(Ag)
detector background and the coincident events of the
ZnS(Ag) detector and plastic veto detector are shown.
The pulse height spectrum for a 241Am source was
detected with a ZnS(Ag) detector [17]..................................................... 26

2.13 The PIPS CAM detector was placed over 2mL of radioactive
solution in a sample holding cup [18]...................................................... 27
 x

List of Figures (Continued)

Figure Page

2.14 The pulse height spectra from a) alpha emitting solutions and
b) beta and gamma emitting solutions detected with a
PIPS CAM detector [18].......................................................................... 29

2.15 The count rate of 244Cm, 241Am, and 230Th solutions plotted
versus the solution activity [18]............................................................... 31

2.16 The relative efficiency of the PIPS CAM detector in regards to the
effect from the solution density [18]........................................................ 32

2.17 The pulse height spectra obtained for a 241Am solution located
at various distances from the detector [18] .............................................. 33

2.18 The count rate of a 241Am solution, a 241Am solution covered by

Mylar®, and a 230Th solution as a function of stand-off
distance [18]............................................................................................. 33

2.19 The detection resolution as a function of source thickness and

size [30].................................................................................................... 43

4.1 The experimental setup for the ZnS(Ag) detector ......................................... 47

4.2 The experimental set up for the Canberra PIPS CAM detector..................... 47

4.3 The Ludlum Model 43-1 ZnS(Ag) scintillation detector clamped

to a stand with bolts propping it over a PVC sample
holding cup............................................................................................... 48

4.4 The experimental set up for the (a) combined beta source and the
electroplated standard and (b) the electroplated standard only................ 51

5.1 The pulse height spectra of a) 241Am and b) 239Pu samples for
the PIPS CAM detector and ZnS(Ag) detector,
respectively. Spectra corresponding to both aqueous
radioactive sources and electroplated standards are shown........................... 57

5.2 Spectra of the aqueous solutions recorded with the (a) PIPS CAM
detector and (b) ZnS(Ag) detector demonstrated the
spectral shape ........................................................................................... 59
 xi

List of Figures (Continued)

Figure Page

5.3 The absolute detection efficiency for the PIPS CAM detector and
ZnS(Ag) detector for alpha radioactivity in aqueous
solutions ................................................................................................... 60

5.4 Beta particle and gamma ray interference in the ZnS(Ag) detector .............. 62

5.5 The net count rate of the ZnS(Ag) detector plotted against various
concentrations of 241Am and 230Th solution............................................. 65

5.6 The Ludlum Model 43-1 ZnS(Ag) detector was used in the
detection of three detection surface areas of depleted
uranium solution ...................................................................................... 66

5.7 The normalized count rate over concentration of the ZnS(Ag)

detector versus the density of the 241Am solution.................................... 70

5.8 The absolute detection efficiency of the ZnS(Ag) detector is

plotted against the increasing density of the 241Am
solution..................................................................................................... 71

5.9 The absolute detection efficiency decreased as the air gap

distance between the ZnS(Ag) detector and 241Am
and 230Th solutions increased................................................................... 72

5.10 The absolute detection efficiency of the ZnS(Ag) for the thick,
14 mg/cm2, and thin, 3.5 mg/cm2, ZnS(Ag) scintillation
layers for aqueous solutions..................................................................... 74

5.11 The pulse height spectra for the 239Pu blend electroplated
standard detected with the thick, 14 mg/cm2, and thin,
3.5 mg/cm2, ZnS(Ag) scintillation layers for
aqueous solutions ..................................................................................... 75

5.12 The pulse height spectra for the thick, 14 mg/cm2, scintillation
layer on the ZnS(Ag) detector.................................................................. 76

5.13 The flow chart of the Visual Basic computer simulation of the
alpha energy deposition in various absorption layers .............................. 79

5.14 The computer-simulated stopping power of water for alpha

particles .................................................................................................... 81
 xii

List of Figures (Continued)

Figure Page

5.15 The computer-simulated pulse height spectra for the ZnS(Ag)

finite scintillation layer and the infinitely thick detection
surface of the PIPS CAM detector for a 239Pu point source .................... 82

5.16 The computer-simulated pulse height spectra for the ZnS(Ag)

finite scintillation layer and the infinitely thick detection
surface of the PIPS CAM detector for a 239Pu
aqueous solution....................................................................................... 84

5.17 The pulse height spectra for a (a) 239Pu point source and (b) 239Pu
aqueous solution simulated for the thin, 3.5 mg/cm2, and
thick, 14 mg/cm2, ZnS(Ag) scintillation layers ....................................... 89

A.1 The figure of merit for the ZnS(Ag) detector using the
signal splitter............................................................................................ 93

B.1 Tail pile-up occurs whenever a tail or undershoot from a previous

pulse is present. The effect of the pile-up is displayed on the
pulse height spectra to the right as the cross-hatched area [6]................. 95

B.2 Two signal pulses which are close together form into a single
distorted pulse as peak pile-up. The dashed lines
represent possible overlapping scenarios [6] ........................................... 95

B.3 A pulse height spectra of 55Fe demonstrates the effects of peak

and tail pile-up. A low counting rate, where pile-up is
negligible, is represented by the dashed line. The solid
line represents a high counting rate and displays effects
due to both tail and peak pile-up [6] ........................................................ 96

B.4 The electronic signal pulse shape which is signature for dead
time in a Geiger- Mueller counter during high count
rates [6] .................................................................................................... 97

B.5 The recorded events in a detector, paralyzable versus

nonparalyzable models [6]....................................................................... 98

B.6 The true count rate versus the observed count rate for
nonparalyzable and paralyzable detectors [6].......................................... 99

B.7 The schematic diagram of the signal splitter ................................................. 100

 xiii

List of Figures (Continued)

Figure Page

B.8 The schematic diagram of the Canberra preamplifier Model

2006 [33].................................................................................................. 102

B.9 The count rate of the ZnS(Ag) detector in conjunction with both
the charge-sensitive preamplifier and the signal splitter for
the detection of several electroplated standards....................................... 103

B.10 The pulse height spectrum of an aqueous 239Pu solution was

compared to the shape of the 239Pu electroplated
spectrum using the ZnS(Ag) detector and
signal splitter............................................................................................ 104

B.11 The pulse height spectra for 244Cm, 239Pu, and 234U/238U aqueous
solutions using the thick scintillation layer on the ZnS(Ag)
detector and the signal splitter ................................................................. 104

B.12 The pulse height spectra for 1644 Bq and 7151.7 Bq 241Am
electroplated standards and a 490 Bq 238Pu
electroplated standard using a ZnS(Ag)
detector and signal splitter ....................................................................... 106

B.13 The pulse height spectra for 490Bq 238Pu EPS, 1644 Bq 241Am
EPS, and 7151.7 Bq 241Am EPS detected with a ZnS(Ag)
detector and charge-sensitive preamplifier .............................................. 106

B.14 The electronic signal from the ZnS(Ag) detector and (a) signal
splitter and (b) preamplifier demonstrated the pulses
created by a 489.9 Bq 238Pu electroplated standard ................................. 108

B.15 The electronic signal from the ZnS(Ag) detector and (a) signal
splitter and (b) preamplifier demonstrated the pulse
created by a 1644 Bq 241Am electroplated standard ................................ 109

B.16 The electronic signal from the ZnS(Ag) detector and (a) signal
splitter and (b) preamplifier demonstrated the pulse
created by a 7151 Bq 241Am source ......................................................... 109

B.17 The pulse height spectra of the 7151.7 Bq 241Am EPS on top of the
Plexiglas shield with and without the 7.4 x 108 Bq 90Sr(90Y)
source detected with the ZnS(Ag) detector and
signal splitter............................................................................................ 112
 xiv

List of Figures (Continued)

Figure Page

B.18 The pulse height spectra of the 7151.7 Bq 241Am EPS on top of
the Plexiglas shield with and without the 7.4 x 108
Bq 90Sr(90Y) source detected with the ZnS(Ag)
detector and charge-sensitive preamplifier .............................................. 112

B.19 The electronic signal for the ZnS(Ag) detector, amplifier, and (a)
signal splitter and (b) the preamplifier. The 7151.7 Bq 241Am
electroplated standard was set on top of the 0.635 cm
Plexiglas sheet placed above a 7.4 x 108 Bq 90Sr(90Y)
source ....................................................................................................... 113

C.1 The moving spherical coordinate system used in the computer

simulations (Relative thicknesses were not drawn
to scale) .................................................................................................... 115

C.2 The Microsoft Excel worksheet referred to by the Visual Basic

source codes ............................................................................................. 119
 LIST OF SYMBOLS AND ABBREVIATIONS

Greek Symbols

Alpha particle: α

Beta particle: β

Delta (change of): Δ

Gamma ray: γ

Micrometer: µm

Pi: π

Counting Sensitivity: ε

Abbreviations

SRS: Savannah River Site

ZnS(Ag) Silver doped Zinc Sulfide

PIPS: Passivated Ion implanted Planar Silicon

CAM: Continuous Air Monitor

HLW: High-level Waste

PUREX: Plutonium and Uranium Recovery by Extraction

PAM: Personal Alpha Monitors

PNNL Pacific Northwest National Laboratory

EPS: Electroplated Standard

WAC: Waste Acceptance Criteria

PVC: Polyvinlychloride

FWHM: Full Width Half Maximum

 xvi

List of Symbols and Abbreviations (Continued)

SRIM: Stopping and Range of Ions in Matter

TRIM: The Transport of Ions in Matter

GEANT4: Geometry and Tracking: A toolkit for simulation of the passage of

particles through matter.

MCNPX: Monte Carlo N-Particle Transport Code

AASI: Advanced Alpha- spectrometric Simulation

LLD: Lower Level Discriminator

FWTM: Full Width Tenth Maximum

ULD: Upper Level Discriminator

CERN: Conseil European pour la Recherche Nucleaire: European Laboratory for

Particle Physics

NIST: National Institute of Standards and Technology

ASTAR: A Stopping power And Range table for helium atoms

MDA: Minimum Detectable Activity

MDC: Minimum Detectable Concentration

 1

CHAPTER 1

INTRODUCTION

There is a need for the direct detection of alpha radioactivity in aqueous solutions.

There are various detectors which detect alpha particles; however, the short range of

alpha particles originating from a solid or liquid matrix makes detection in these media

problematic. Actinides are present in high-level waste (HLW) within the Department of

Energy complex at concentrations of approximately 8 x 104 Bq/mL for the HLW

supernate and 8 x 108 Bq/g for HLW sludge. The total alpha activity of the liquid waste

is quantified and monitored during the chemical processing of the HLW. At the

Savannah River Site (SRS), the concentration of the treated HLW supernate can not

exceed 970 Bq/mL in order to be accepted for grouting at the Saltstone Facility [1].

Currently, several methods are used at SRS to measure the alpha activity in tank

waste supernate and processed waste. Gross alpha counting is conducted by depositing a

small aliquot of waste onto a planchet. After evaporation, the alpha activity is quantified

with a gas-flow proportional counter. The detection of alpha particles is hindered by the

high salt content of the sample which introduces uncertainty into the measurements. SRS

also periodically uses radiochemical separation techniques to isolate the alpha activity

from the waste matrix prior to quantification by either gas flow proportional counters or

silicon detectors. Inductively coupled plasma mass spectroscopy is also used to identify

the contents of tank waste; however, the mass spectrometer results are sometimes

complicated with isobaric interferences [1].

 2

On-line flow-through alpha radiation detectors are needed at HLW processing

facilities along side beta and gamma detectors for continuous monitoring. An on-line

alpha detector on a process stream would be able to monitor the solution continuously

and nondestructively. Beta and gamma ray radiation are typically easier to detect from

waste due to the longer range and mean free path, respectively, in process solutions. The

ZnS(Ag) detector would simplify the detection of alpha particles from aqueous solutions

if the alpha particles are efficiently detected from the surface of the liquid. An efficient

on-line method of total alpha detection would be useful and practical for chemical

processing applications and numerous other purposes.

 3

CHAPTER 2

LITERATURE REVIEW

Defense Waste

In the 1940’s, the United States embarked on a series of military and civilian

activities which initiated the accumulation of radioactive wastes. The activities began in

earnest with the Manhattan Project and continue today through the use of nuclear power

in the generation of electricity. Various forms of radioactive wastes are produced

through different kinds of activities. Civilian waste consists mainly of spent fuel from

reactors and low-level wastes. Wastes stockpiled from the national defense taskforce

consist of irradiated target materials, spent fuel, transuranic wastes, low-level wastes, and

HLW. In comparison, research laboratories and the nuclear power industry (exclusive of

nuclear power reactor stations) generate mostly low-level wastes and only produce high-

level waste in special cases where nuclear fuel or fissile material is handled for research

[2].

Out of the nation’s radioactive waste, HLW is the top concern as it contains the

bulk of the radioactivity found in nuclear waste. HLW is defined as the highly

radioactive material which typically contains significant concentrations of fission

products and is the consequence of reprocessing spent fuel [2]. Tritium and plutonium,

the materials required for nuclear warheads, were extracted by means of reprocessing

irradiated targets and nuclear fuel rods from the United States’ original fourteen materials

production nuclear reactors. The first materials production reactor went critical in 1944

and the last materials production reactor was shut down in 1988 [3].
 4

High-level waste is a byproduct of plutonium and uranium recovery by extraction

(PUREX) process. The PUREX process is the leading method of reprocessing and was

widely used by the defense department at both the Hanford Site and SRS. The process

requires that the irradiated target be chopped and dissolved in nitric acid. The heavy

elements went into solution while the cladding remained whole. The nitric acid solutions

were processed through solvent extraction which separated the uranium and plutonium

from the fission products and other transuranic elements. The resultant fission products

and transuranic elements are discarded as HLW [2]. The fission products are primarily

responsible for the high activity and external dose rate of the waste while the high-

activity transuranic elements generate considerable thermal energy [3].

After nuclear warfare production declined, the demand for plutonium and highly

enriched uranium decreased and the pace for reprocessing was reduced. The Department

of Defense phased out reprocessing and the production of nuclear weapons and

eventually stopped work in 1992 [3]. The Department of Energy, however, left an

accumulation of HLW, typically stored in large tanks, for the nation to manage. The

original tanks located on the Hanford Site corroded as the single carbon steel walls were

only designed for a useful lifetime of 25 years. Sixty seven Hanford tanks leaked HLW

into the environment causing contamination issues. The three largest leaks topped the

chart at 115,000, 70,000, and 55,000 gallons of HLW liquid into the ground [3]. Twenty

eight new double-walled carbon steel tanks were constructed in 1980 to rectify the

problem. The waste was transferred to the safer containers which were designed to hold

the waste for fifty more years. The Department of Energy is storing approximately 100

million gallons of HLW in 243 underground tanks [3].

 5

The Hanford Site and SRS are currently assessing the existing method of

converting HLW into a more stable form. The sludge, liquid, and solid components of

the aged, HLW on these sites are mixed with molten glass and poured into metal

cylinders. This process is called vitrification, and with the waste in the vitrified, glass

form, it is more stable and less likely to contaminate the environment. The Department

of Energy has already begun vitrification at a plant at SRS and by 2005, 1.64 million

gallons of high-level sludge were successfully vitrified [4]. After vitrification, the

canisters were placed in a specially designed storage facility. Vitrified waste costs less to

store and monitor compared to liquid waste and decreases the threat to security. The

canisters will be kept in the storage facilities until a geologic repository is open then they

will be moved to the repository permanently [3].

Alpha Particle Theory

Alpha particles are heavy, charged particles which are typically emitted from

disintegrating, heavy nuclei. The alpha particle is comprised of two protons and two

neutrons and is equivalent to a doubly ionized helium atom, as it lacks electrons. The

alpha particle is the least penetrating form of radiation due to its large size and positive

charge [5]. The alpha particle interacts with matter through attraction with the orbital

electrons of the absorber atoms via the Coulomb force. The impulse caused by the

passing alpha particle is enough to excite the electron or to ionize the absorber atom—the

primary means of energy transfer. The interaction between the alpha particle and the

nucleus of the absorber atom via Rutherford scattering is infrequent and typically not

significant in the detection of alpha particles. The kinetic energy of the alpha particle

decreases after each interaction which reduces the particle velocity. The alpha particle
 6

loses approximately one five hundredth of its total energy after every interaction [6]. The

alpha particle interacts continuously in the absorber material thus losing energy linearly,

consequently slowing the particle until it stops [6].

The alpha particle intensity passing through an absorber does not decrease until

near the end of the range as shown in Figure 2.1. The energy of the alpha particles

decreases with increasing absorber thickness; however, the number of alpha particles

detected does not decrease until the approximate range is reached and the alpha particles

begin to terminate [7]. The alpha particle range, the distance which particles penetrate, is

much shorter than that of beta particles or the mean free path for gamma rays. The mean

range, the most common reference, refers to the absorber thickness which reduces the

alpha count to half. The extrapolated range is obtained by extrapolating the alpha

absorption curve in Figure 2.1 to zero transmission and represents the largest possible

range [7]. The mean range of alpha particles from common alpha emitting radionuclides

is only a few centimeters through air. The range of alpha particles through liquids and

solids is significantly shorter than the range in air due to the high density of the media

[6]. The energy-range relationship of alpha particles through air is estimated as

0.56 E , E < 4 MeV (1)

R(cm) =
1.24 E − 2.62, 4 MeV < E < 8 MeV

at 1 atm and at 15°C [6]. The alpha particle energy-range relationship through other

media, the Bragg-Kleeman Rule, is found through the relationship

R1 ρ 0 A1 (2)
=
R0 ρ1 A0
 7

where R is the range, ρ is the density of the medium, and A is the atomic mass number of

the medium [7]. If the medium is a mixture of elements, the square root of the effective

atomic mass number is

n1 A1 + n2 A2 + n3 A3 + ... (3)
A=
n1 A1 + n2 A2 + n3 A3 + ...

where n is the atomic fraction of the elements and A is the atomic mass number of each

element [5]. The square root of the effective atomic mass number of air is 3.82 and the

density of air is 1.226 x 10-3 g/cm3 at 1 atm and at 15°C [5].

Figure 2.1 The alpha particle absorption curve demonstrates the consistent count rate
and alpha range [5].

The alpha particle tracks through matter are linear as they are not easily deflected

due to their large momentum. The Coulomb force is not significant in changing the path

of an alpha particle because the alpha particles undergo interactions from all directions as

they travel through matter. Deflections in the alpha particle track do not usually occur

until the end of the range when the alpha particles have decreased energy [6]. Alpha

particles are mostly affected by straggling, the fluctuation in path length of particles with

the same initial velocity, towards the end of the range. The alpha particle energy loss is a

stochastic process and an individual alpha particle range varies by a few percentage

points of the mean range [7].

 8

The ionization of air requires roughly 35 eV per ion pair produced, where the ion

pair is defined as the positive ion and free electron. The exact value for ionization

depends on the velocity of the alpha particle though the medium. Approximately half of

the alpha particle energy lost in air is due to electron excitation as opposed to ionization

[5]. The ionization potentials for O2 and N2 are 13.6 eV and 14.5 eV, respectively. The

specific ionization along the alpha particle path is defined as the number of ion pairs per

unit path length, which equals approximately 2,000 to 6,000 ion pairs per millimeter of

air. Alpha particles ionize the air most heavily near the end of their range as shown in the

Bragg Ionization curve, Figure 2.2, which establishes the relationship between the

average specific ionization and the distance from the source [5].

Figure 2.2 The Bragg ionization curve for Polonium alpha particles [5].

The average linear rate of the energy loss of a heavy, charged particle is called the

stopping power of a medium, S.

dE
S =− (4)
dx

Hans Bethe improved upon Bohr’s classical equation for the average linear rate of energy

loss by incorporating quantum mechanical theory which allowed the equation to be useful

for all heavy, charged particles [8].

 9

dE 4π k02 z 2 e4 n ⎡ 2mc 2 β 2 ⎤
− = ⎢ ln − β 2
⎥ (5)
dx mc 2 β 2 ⎢ I (1 − β 2 ) ⎥⎦
⎣

where:

k0 = 8.99 x 109 N m2 C-2

z = atomic number of heavy particle
e = magnitude of the electron charge
n = number of electrons per unit volume in the medium (e-/m3)
m = electron rest mass
c = speed of light in vacuum
β = v/c = speed of the particle, v, relative to c
I = mean excitation energy of the medium

The number of electrons per unit volume in the medium is

⎛ molecules ⎞ ⎛ g ⎞ ⎛ e− ⎞
6.022 x1023 ⎜ ρ
⎟ ⎜ 3⎟ ⎜
Z ⎟
⎝ moles ⎠ ⎝ m ⎠ ⎝ molecule ⎠ (6)
n=
⎛ g ⎞
A⎜ ⎟
⎝ mole ⎠

where ρ is the density of the medium and the atomic number, Z, represents the number of

electrons in the medium. The mean excitation energies, I, are approximated from

previous experiments for elements of atomic number, Z, using empirical formulas in

equation 7 [8].

⎧ 19.0 eV , Z = 1 ⎫
⎪ ⎪
I ≅ ⎨11.2 + 11.7 z eV , 2 ≤ Z ≤ 13⎬ (7)
⎪ 52.8 + 8.71z eV , Z > 13 ⎪
⎩ ⎭

Equation 5 is simplified greatly by the incorporation of constants into the equation. The

atomic number of the particle is defined as 2, which specializes the equation for alpha

particles in equation 8 [8].

dE 2.032 x 10−30 n ⎡ 1.02 x 106 β 2 ⎤

− = ⎢ ln − β 2
⎥ MeV cm −1 (8)
β I (1 − β )
2 2
dx ⎢⎣ ⎥⎦
 10

The stopping power of a medium for an alpha particle depends greatly on the

energy or velocity of the alpha particle which dictates the only remaining variable, β.

The stopping power of water in MeV/cm for several particles at various energies is

shown in Figure 2.3 [8].

Figure 2.3 The stopping power of water [8].

Scintillation

The energy states are determined by the crystalline lattice of an inorganic material

thus influencing the scintillation mechanism. The lowest band of the valence states

represents electrons which are bound to the lattice sites. The conduction band contains

higher energy electrons which are free to migrate throughout the crystal. The

intermediate region between the valence and the conduction bands, where electrons do
 11

not normally reside, is called the forbidden band. When energy is absorbed from an

alpha particle, an electron is elevated from the valence band into the conduction band

thus leaving a hole in the valence band. If scintillation is favorable then an electron drops

from the conduction band into the valence band resulting in the emission of light or

scintillation photons. An activator is used to dope the pure crystal and to modify the

energy band structures as shown in Figure 2.4. The activator creates energy states in the

forbidden band of the crystalline states of the host crystal which allow electrons to de-

excite with photons in the visible range of the spectrum [6].

Figure 2.4 The energy band structure of an activated crystalline scintillator [6].

Zinc Sulfide

Zinc sulfide, ZnS, and silver activated zinc sulfide, ZnS(Ag), are common

detectors of alpha particles and are among the oldest inorganic scintillators. The ZnS

screen utilized in the famous Rutherford experiment was used to prove the existence of

the atomic nucleus. The scintillation screen allowed for the locations of the alpha

particles to be recorded after interaction with the gold foil. The majority of Personal

Alpha Monitors (PAM) which are used in laboratories for the detection of alpha

contamination are constructed of ZnS(Ag) due to its potential large area and low cost.

ZnS is also commonly used in its un-doped form in “glow-in-the-dark” paints and

products due to its lengthy phosphorescence.

 12

ZnS(Ag) detectors are convenient to use in applications where gross counting is

involved due to the portability of the scintillation detector. The ZnS(Ag) scintillation

crystals are either sprayed onto the light guide, attached with double-sided tape, or

embedded into an epoxy. A photomultiplier tube converts the light from the ZnS(Ag)

into an electronic signal which can be counted with appropriate electronic devices. A

thin layer of Mylar® covers the scintillation layer to create a light-tight atmosphere;

although, the Mylar® layer is delicate and easily punctured or scratched [9].

ZnS(Ag) is only available as a polycrystalline powder and is therefore primarily

limited to thin screens. Thicknesses ranging between 3.5 mg/cm2 and 25 mg/cm2 are

recommended by the Ludlum manufacturers. The ZnS(Ag) layer is unusable at

thicknesses greater than 25 mg/cm2 due to the opacity of the layer to its own

luminescence [6]. At greater thicknesses, ZnS(Ag) crystals are opaque due to the

absorption of light into the material and from light scattering. The extinction coefficient,

a measure of how well a substance absorbs electromagnetic radiation at a given

wavelength, is best used to describe the absorption of the scintillated light. The intensity

of the electromagnetic radiation is dampened as the light propagates through the material.

The light intensity after transition through a material of z thickness is defined as I,

−2ω k
z
I ( z ) = I 0e c
(9)

where k is the extinction coefficient, ω is the angular frequency of the electromagnetic

wave, and c is the speed of light. The extinction coefficient can be determined from

Figure 2.5, a graph of ZnS extinction coefficients at various photon energies [10].
 13

Figure 2.5 The extinction coefficient of ZnS versus photon energy for three different
ZnS thicknesses: 1) 100 nm, 2) 200 nm, and 3) 400 nm [10].

Scintillation decay times for ZnS(Ag) are reported between several hundred

nanoseconds and 10 µs with an accepted decay constant of 110 ns [10]. The decay time

recorded in Knoll is 0.2 µs or 200 ns; although phosphorescence of a longer duration was

noted in ZnS(Ag) scintillation [6]. ZnS(Ag) possesses several other physical properties

listed in Table 2.1: the specific gravity of the scintillator, the wavelength of the emission

[10], and the refractive index [6].

 14

Table 2.1 Common properties of ZnS(Ag) scintillator [6].

Properties of ZnS(Ag) Scintillator

Specific Gravity 4.09

Emission Wavelength (nm)

Peak 450
Minimum 330

Refractive Index 2.36

A pulse height spectrum for the detection of 5.5 MeV alpha particles is shown in

Figure 2.6. ZnS(Ag) is primarily used for the detection of alpha particles; however,

thermal neutrons are detected by incorporating 6Li into the scintillation material, also

shown in Figure 2.6. Fast neutrons are detected with ZnS(Ag) when the scintillation

powder is embedded into a hydrogenous compound. The ZnS(Ag) detects the recoiling

proton from a neutron-proton scattering interaction [10].

Figure 2.6 The ZnS(Ag) pulse height spectra for 5.5 MeV alpha particles and thermal
neutrons [10].
 15

Alpha Radiation Monitoring Detector

A flow-through zinc sulfide phosphor scintillation detector which quantified alpha

radioactivity in low-level radioactive liquids was utilized in 1973 by M.M. Chiles at Oak

Ridge National Laboratory. The scintillation detector was coupled to an alarm, and when

the alpha radiation count rate was above a set concentration discriminator, the alarm

sounded. The system was used to find contamination in the laboratory’s process water.

The detector did not determine the alpha particle energy and therefore was not useful in

the determination of the present radioisotopes. The system was used to distinguish

between those liquids which had alpha emitting radionuclides present and those which

did not. The system, however, had a printed record of the count rate and the time at

which the liquids flowed past the detector [12]. The zinc sulfide phosphor scintillation

layer, 4 or 5 mg/cm2, consisted of a powder which was applied with adhesive to the light

pipe in front of a two inch diameter photomultiplier tube. A thin layer of Mylar® was

placed between the scintillation layer and the low-level liquids in order to not disrupt or

dissolve the scintillation powder. The zinc sulfide phosphor detector was encased in

stainless steel housing and the low-level process waste water was injected into the

system. The injected liquid spread out radially over the Mylar® window as depicted in

Figure 2.7. The contamination of Mylar® was not a concern as the continuous flow of

water had the effect of flushing away any contamination. The alpha background of the

detector quickly returned to normal after the contaminated water had passed through the

detection system. In rare circumstances where the background was elevated because of

cross-contamination, the detector was cleaned with a detergent or the Mylar® was

replaced [12]. The zinc sulfide phosphor detector also boasted a large six-inch diameter
 16

scintillation area, a considerable benefit for the detection of alpha particles from a low-

level liquid source [12].

Figure 2.7 The flow through ZnS phosphor detector used to detect contamination in
low-level aqueous solutions [11].

The relationship between the count rate and solution concentration was found for

the ZnS phosphor detector along with the intrinsic detection efficiency and minimum
238
detectable concentration. The count rate and concentrations of U solution

demonstrated a linear relationship. The intrinsic detection efficiency of the ZnS

233
phosphor detector for U was calculated as 87.7% and the minimum detectable

concentration for the detector was found to be 0.48 Bq/mL for the 238U solution [12].

A mathematical model was created to quantify the number of alpha particles

238
which reached the detector. The model investigated the range of U alpha particles

through water and Mylar® and was used to determine the effective volume of liquid that

an alpha particle could be detected in [12]. The location of the alpha particle when it was

released from the nucleus, solid angle subtended by the detector, and the Mylar®
 17

absorption layer between the solution and detector were accounted for in the model. The

Mylar® window thickness, the only layer between the ZnS phosphor and water, was

converted into water equivalent absorption thickness using absorption data in order to

treat the materials as one. The number of alpha particles, N, per unit time within an

element of thickness which interacted in the detector was calculated as

θ min CARm
N =∫ (1 − cos θ ) sin θ dθ (10)
θ max 2

where C was the specific activity, A was the surface area, Rm was the maximum range of

the alpha particle through solution and Mylar®, and θ was the angle between the

directional vector of the alpha particle and the normal of the solution, Figure 2.8 [12].

The maximum angle of the alpha particle, θmax, occurred when no alpha energy was

deposited in the solution or ZnS phosphor layer. The minimum angle, θmin, represented a

scenario where the alpha particle was emitted perpendicular to the solution and was zero.

The distance the alpha particles traveled through the water was determined by taking the

integration of the thickness of water and the geometry fraction. The geometry fraction in

equation 10, defined as ½(1-cosθ), was simplified for a point source located in the middle

of the surface area. In the experimental homogeneous solution, the alpha particles were

originating throughout the solution which resulted in variations in the x, y, and z

coordinates. Despite the variations in the experimental set up, the mathematical model

only accounted for variations in the x direction. Although the mathematical model did

not precisely portray the physical set up, the calculated activity of 0.83 Bq/mL was only

slightly less than the experimental activity, 0.95 Bq/mL.

 18

Figure 2.8 The geometric diagram used in the derivation of the calculated maximum
distance of alpha particle origin from the detector to assure detection [12].

In 1984, Robert et al., reported using a flow-through alpha detection system

which utilized a cerium-doped SiO2 scintillator sandwiched between the circulation cell

of the liquid and photomultiplier tube, Figure 2.9 [13]. This flow-through alpha detector

was designed for the spent fuel reprocessing industry and for the general use in actinide

laboratories for the detection of alpha particles from aqueous solutions. The liquid

solution came in direct contact with the cerium-doped SiO2 scintillation detector. The

absence of the air gap was an advantage since the alpha particles did not lose energy in

the absorption thickness. The scintillation glass was placed in contact with 5 N nitric

acid for four months without showing signs of deterioration [13]. However, spent fuel

associated with reprocessing is typically 14 M NaOH, a very corrosive liquid which is

able to etch glass. The detector was coupled to a multi-channel analyzer which provided
244
energy spectra of the alpha emitting samples. A spectrum of a Cm solution was

detected using two different scintillation thicknesses: 35 µm and 2 mm. The thinner

scintillator had a greater peak resolution [13]. The cerium-doped SiO2 detector was
233 239 241
tested with three different radionuclide solutions: U, Pu, and Am. For each

radionuclide test solution, the detector showed a linear increase in count rate with the

radionuclide concentration. Beta and gamma emissions, present from fission products in
 19

the waste, were differentiated from alpha isotopes by using the spectrometric properties

of the detector. The spectra of alpha and beta sources, Figure 2.10, showed the alpha

peaks at higher energies than the beta peaks. A lower-level discriminator (LLD) was set

as a means of separating the two types of radiation. The cerium-doped SiO2 scintillator

also possessed a quick decay time of 78 ns, the shortest time for scintillating glass [13].

The fast decay time was an advantage as pulse pile up did not occur in the detector. The

study determined that this detection system was appropriate for use in spent nuclear fuel

reprocessing plants [13].

Figure 2.9 The flow through detection system set up with scintillation glass detector and
radioactive liquid [13].
 20

Figure 2.10 The pulse height spectra of alpha and beta solutions using the cerium doped
SiO2 detector [13].

A ZnS(Ag) detector was developed by McElhaney et al., which possessed a more

rugged surface and better detection efficiency than the conventional ZnS(Ag) detector

[9]. A Mylar® layer of 0.8 mg/cm2 is typically used to achieve light tight properties for

conventional ZnS(Ag) detectors; however, it is easily scratched or punctured as it is not a

robust material. The ZnS(Ag) detector was developed to have a sturdier surface which

consisted of a scintillation layer of ZnS(Ag) crystals and epoxy. The ZnS(Ag) powder

was mixed with an optically transparent, low viscosity epoxy and poured into a mold.

The ZnS(Ag) settled to the bottom of the strong, durable epoxy layer which allowed for a

minimum layer of epoxy in front of the ZnS(Ag) crystals. The viscosity of the epoxy was

lowered further by placing the layer in the oven. The scintillation layer consisted of an

8.0 mg/cm2 ZnS(Ag) layer with an epoxy backing. A 0.087 mg/cm2 aluminum layer was

then spin-coated onto the scintillation layer to ensure an opaque detection window. A

0.02 mg/cm2 cyanoacrylate hardcoat was also applied to the scintillation layer through

spin-coating to increase the ruggedness of the detector. The detection layer was sturdier
 21

than the traditional ZnS(Ag) detector, more resistant to scratches and punctures, and

waterproof. The detection of alpha radioactivity increased with the more rugged

ZnS(Ag) detector due to the thin nature of the protective layers. The count rate of a 239Pu

alpha source with unknown activity at a high voltage of 800 V was 73.4 Bq for the more

rugged detector and 60.6 Bq for the conventional detector [9]. Resistance towards highly

corrosive liquids remains to be determined.

A study of the theoretical response of a ZnS(Ag) scintillation detector, conducted

by Skrable et al., investigated the self absorption of alpha radiation within a sample [14].

A thick solid and a source with minimal weight were studied as dense and light samples

of infinite thickness. The detection of alpha particles from a thick dust deposit on an air

filter proved to be difficult as the thickness resulted in the deposition of the majority of

the alpha energy in the dust particles. The dust deposit resembled an infinitely thick

alpha source as the deposit contained a thickness greater than the range of the alpha

particles in the medium. An 11.4 cm diameter ZnS(Ag) scintillation layer was used to

detect the count rate of the air filter and was then compared to theoretical calculations.

The Mylar® layer, used to keep out light contamination and to keep the scintillation layer

from being contaminated, was included in the calculations, and an air gap was not present

[14]. The weightless sample model consisted of air that contained radon. Alpha particle
222
ranges for various mixtures and compounds were investigated while Rn was used to

model radionuclides in an infinite homogeneous gaseous medium [14].

The total specific alpha activity from soil was directly detected by Phoenix et al.,

using a ZnS(Ag) detector instead of eliminating the sample matrix through radiochemical

procedures [15]. The experimental set up consisted of a Ludlum Model 43-1 ZnS(Ag)
 22

detector coupled to a Ludlum Model 2200 scalar and a printer. The theoretical detector

response count rate equation (cps), y, was

CARs CARs CARs 2

y= − x+ x (11)
4 2 Rw 4 Rw 2

where C was defined as the specific alpha activity, A was the area of the source in contact

with the detector, Rs and Rw were the mean alpha particle ranges in source and window

material, respectively, and x was the window mass density thickness [15]. The three soil

samples investigated included natural uranium, natural thorium (both in secular

equilibrium with their prodigies), and a sample of uranium oxide (99% U3O8). A two-

inch diameter steel planchet was filled with soil and covered with a single piece of 0.297

mg/cm2 Mylar®. The covered planchet was turned on top of a large ZnS(Ag) detector.

The Mylar® layer on top of the soil and the 0.297 mg/cm2 Mylar® layer included with the

detector were between the sample and detector; however, additional Mylar® layers were

added. The count rate for each respective Mylar® thickness was recorded from the two

original layers (0.594 mg/cm2) to a total of ten layers (2.673 mg/cm2). The plot of count

rate versus absorption thickness showed a quadratic curve, Figure 2.11. The curve was

fit and by using the coefficient on the linear term, the specific activity was calculated. To

calculate the specific activity of the sample, the surface area, mean alpha ranges in the

source and window material, and the window mass density thickness were known. The

total specific alpha activities from the soil samples were determined with a mean percent

deviation of 17%. The alpha activities obtained from the soil samples confirmed the

detection of alpha particles from thick sources using the detector response equation,

however, with limits imposed by stochastic uncertainty and uncertainty from various

parameters. The direct measurement technique was used as a cheap and easy way to
 23

screen soils, ores, and industrial process materials. If the soil was not homogeneous,

however, then the material detected did not represent the entire sample [15].

232
Figure 2.11 The count rate for a Th soil sample versus Mylar® absorber thickness
[15].

Ayaz et al., compared the theoretical and experimental detection efficiencies of

the ZnS(Ag) detector in the detection of liquid samples using equation 11 [16]. The gross

alpha activity was measured from an infinitely thick alpha emitting source; however,

instead of using soil to test the theory, manganese dioxide coated glass fiber filters and

radioactive solutions were used as thick alpha emitting sources. MnO2 coated glass was

used to concentrate radium and uranium from neutral pH water and was then detected

with the ZnS(Ag) detector [16]. In another experiment, aqueous alpha radioactivity was

pumped into a light-tight black box where the radionuclide concentration was quantified

with the aid of a ZnS(Ag) scintillation disc on a photomultiplier tube. The detection

system was designed to be a flow-through alpha particle detector. The only barrier
 24

between the sample and the scintillation layer was air as the Mylar® layer was left off the

detector face for better detection. The experimental and theoretical efficiencies were

within 7% of one another [16].

ZnS(Ag) Spectroscopy

The ZnS(Ag) alpha detector and a plastic veto detector were proposed for the

detection of low activity environmental samples by eliminating background counts in the

ZnS(Ag) detector. Ardid et. al, focused on determining the various sources of ZnS(Ag)

detector background in an attempt to improve the detection of low activity samples [17].

A 3.25 mg/cm3 ZnS(Ag) detector was connected to a plastic veto detector, a detector

typically used in the detection of beta and gamma radioactivity. A multi-channel

analyzer was used for monitoring data collection and the optimization of the operation

conditions [17].

The plastic veto detector and ZnS(Ag) detector, placed face-to-face, were

connected in a coincidence set-up to determine the origin of the ZnS(Ag) detector

background. The plastic veto detector had a higher detection efficiency of beta particles,

gamma rays, and cosmic rays. Although the ZnS(Ag) detection efficiencies for cosmic

rays were low, the parameter was investigated due to increased chance of interaction

from long counting times and large cosmic radiation flux. Coincident events were

detected by the plastic veto detector and ZnS(Ag) detector during a background count. A

coincident count rate of only (8.4 + 1.8) x 10-4 cps was detected while the total alpha

background count rate from the ZnS(Ag) detector without the veto detector was (2.5 +

0.3) x 10-3 cps, a significant fraction. The coincident events were tracked as they

represent potential rejects to be subtracted from the alpha background of the ZnS(Ag)
 25

detector. The coincident events were caused by entities such as cosmic rays, gamma, or

beta rays from the sample, or a gamma ray from the environment. A gamma ray

interaction involving Compton scattering was another coincident event possibility as a

secondary gamma ray could be detected in the plastic veto detector. The subtraction of

the background coincident events from the total alpha background count rate from the

ZnS(Ag) detector demonstrated the decrease in the background of the ZnS(Ag) detector.

The pulse height spectra of the coincident events of the ZnS(Ag) detector and plastic veto

detector were plotted on top of the total ZnS(Ag) background counts in Figure 2.12. The

pulse height spectrum of the 241Am source, detected with only the ZnS(Ag) detector, was

displayed to demonstrate the shape and location of an alpha peak [17].

A lead shield was constructed around the instrumental apparatus in an attempt to

decrease the beta and gamma radiation from the environment. The lead shield decreased

the total counts in the plastic veto detector by a factor of 20; however, it did not reduce

the coincident events in the ZnS(Ag) and plastic veto detectors. Due to the consistency

of the coincident events in the presence of shielding, it was determined that cosmic rays

constituted the main contribution to the background [17].

A 137Cs gamma ray source was detected by the ZnS(Ag) and plastic veto detectors

in order to determine the detection of coincident events due to Compton scattering. The

gamma component of the ZnS(Ag) background was determined to be very small with less

than 10% of coincident events from gamma rays [17].

The effect of radon on the detector background was studied by varying the air gap

distance between the detectors. The total alpha rate increased with increasing air gap

distance; however, the coincident events remained the same. Radon in the environment
 26

was determined to be a significant factor in the detector background, but did not

contribute to the detected coincident events [17].

Figure 2.12 The pulse height spectra for the total events of the ZnS(Ag) detector
background and the coincident events of the ZnS(Ag) detector and plastic veto detector
are shown. The pulse height spectrum for a 241Am source was detected with a ZnS(Ag)
detector [17].

PIPS CAM Research

The usefulness of the passivated ion implanted planar silicon (PIPS) continuous

air monitoring (CAM) detector in the direct measurement of actinides from a liquid was

investigated by Egorov et al., for the application of process waste streams. The study was

aimed at producing an on-line detector for a process waste stream in the chemical

processing of high-level nuclear waste. Although the study was intended for a flow-

through waste stream, the experiments were conducted with 2 mL of solution in a sample

holding cup constructed of polyvinylchloride (PVC), Figure 2.13, with an air gap of 0.5

cm between the detector and solution surface. The detector was connected to a Canberra
 27

alpha spectrometer and the data was recorded with an Aptec analog to digital converter

and multi-channel analyzer [18].

Figure 2.13 The PIPS CAM detector was placed over 2 mL of radioactive solution in a
sample holding cup [18].

Five alpha solutions were used in order to obtain the pulse height spectra in

Figure 2.14(a) using the PIPS CAM detector. The pulse height spectra, an alpha energy

histogram with 1,022 channels or bins, had well-defined leading edges corresponding to

the maximum energy of the alpha particles. The spectral shapes were consistent with the

theoretical process of alpha particles traveling through an infinitely thick source. The

continuous energy distribution was attributed to the decreasing alpha particle energy as

the location of the nucleus which was emitting the alpha particle was at a greater depth.

The leading edge channel numbers, Table 2.2, scaled linearly with the maximum alpha

energy of the radionuclide solution as channel and energy were related. The counting

sensitivities, the net count rate (cps) over activity concentration (Bq/mL), of the PIPS

CAM detector were 0.221 + 0.005, 0.180 + 0.004, 0.160 + 0.005, 0.139 + 0.005, and

0.121 + 0.003% for 244Cm, 241Am, 239Pu, 233U, and 230Th, respectively [18]. The counting

sensitivity of the detector increased with alpha energy, Table 2.2. Despite the changing

detection efficiency for the alpha particle energy, a disadvantage in gross alpha counting,

the solution could be calibrated or the spectra de-convolved.

 28

Beta and gamma interferences are problematic for the detection of alpha radiation.
99 60 137 90
The radionuclides chosen for the study, Tc, Co, Cs, and Sr(90Y) are common

fission and activation products in aged nuclear waste and would interfere in an on-line

detector. The counting sensitivities (cps/Bq/mL) of the beta and gamma solutions were

4.06 x 10-4, 4.69 x 10-6, 1.89 x 10-7, and 8.53 x 10-10% for 90Sr(90Y), 137Cs, 60Co, and 99Tc

as shown in Table 2.3 [18]. Although the beta and gamma interferences were located in

the lower energy region in Figure 2.14(b), when a lower level discriminator was set, the

alpha region of interest was reduced. The beta inference from the 90Sr(90Y) solution was

located at channels as high as 550 while the leading edge of the 230Th pulse height spectra

was at channel 643 in Figure 2.14. The counting sensitivities (cps/Bq/mL) for the

reduced region of interest, 550-900 channels, were 0.113 + 0.004, 0.072 + 0.003, 0.057 +
244 241 239 233 230
0.003, 0.033 + 0.002, and 0.023 + 0.001% for Cm, Am, Pu, U, and Th,

respectively as shown in Table 2.2 [18]. The counting sensitivity of the detector

decreased significantly from that of the larger alpha region of interest, channels 200-900.
 29

(a)

(b)

Figure 2.14 The pulse height spectra from a) alpha emitting solutions and b) beta and
gamma emitting solutions detected with a PIPS CAM detector [18].
 30

Table 2.2 The counting sensitivities of 230Th, 235U, 239Pu, 241Am, and 244Cm solutions
using the PIPS CAM detector in direct alpha assay using two different channel ranges:
200-900 and 550-900 channels [18].

Table 2.3 The counting sensitivities for beta and gamma solutions, 99Tc, 60Co, 137Cs, and
90
Sr(90Y) for the same channel ranges as Table 2.2 [18].

The effects of the solution concentration on the detector count rate of the PIPS

CAM detector were shown in Figure 2.15. The relationship demonstrated linearity

between the count rate and concentration, which showed the absence of pulse pile-up and

dead time at these concentrations. The relationship also signified the capability of PIPS

CAM detection via direct alpha assay [18].

 31

244 241 230

Figure 2.15 The count rate of Cm, Am, and Th solutions plotted versus the
solution activity [18].

The effects of solution density on alpha detection were investigated with the PIPS

CAM detector. The solution density was graphed against the relative efficiency in Figure

2.16. The relative efficiency assumed unity for the detection efficiency of the 1.0 g/mL

density solution. The higher density solutions were associated with lower efficiencies.

The data points followed a general decreasing trend between efficiency and density;

however, the scattered points showed uncertainty in the data. Aged nuclear waste has a

density range between 1 and 1.4 g/mL which resulted in approximately a 20% loss in the

relative counting sensitivity for the PIPS CAM detector [18]. The solution density would

have to be independently quantified in the process waste stream in order for a correction

to be made.
 32

Figure 2.16 The relative efficiency of the PIPS CAM detector in regards to the effect
from the solution density [18].

The distance between the PIPS CAM detector and the solution surface was varied

in order to observe the spectral response and the effect on count rate. Figure 2.17 showed
241
the detector spectral response to a Am solution with increasing distance. The leading

edges of the pulse spectra moved to the lower energy channels as the distance increased.

The alpha particles which reached the detector had a decreased energy due to energy

absorption in the air gap. Fewer alpha particles reached the detector as the air gap

increased due to solid angle effects. Figure 2.18 demonstrated the decrease in count rate

as the stand-off distance increased [18].

 33

241
Figure 2.17 The pulse height spectra obtained for a Am solution located at various
distances from the detector [18].

Figure 2.18 The count rate of a 241Am solution, a 241Am solution covered by Mylar®,
and a 230Th solution as a function of stand-off distance [18].

The minimal detectable concentration (MDC) for the PIPS CAM detector was

found using equation 12:

4.653 b + 2.706
MDC = (12)
f εT
 34

where b is the background counts, T is the counting time, ε is counting sensitivity, and f

is the yield or branching ratio of alpha particles for a particular energy. The branching

ratio was assumed to be one for each radionuclide, the counting sensitivities were taken

from Table 2.2 and a background count rate of 7 x 10-3 cps was used to calculate the
244
MDC for a 10 minute counting time. The minimum detectable concentrations of Cm,
241 239
Am, and Pu were found to be 17.7, 23.0, and 35.0 Bq/mL for the PIPS CAM

detector [18].

Several sample nuclear waste matrices were analyzed using the PIPS CAM

detector in order to evaluate the effectiveness of direct detection of alpha particles from a

liquid surface. The count time was kept short—10 minutes—in order to maintain a

correlation with an on-line waste stream. The 550-900 channel range was used in order

to minimize beta and gamma spillover into the alpha ROI. The detector was calibrated
241
for the detection efficiency of Am and the linear relationship for density was used.

The alpha activities of several solutions were compared in Table 2.4 using direct

measurements from a PIPS CAM detector and a conventional radio-chemical method.

The activities measured by the two measurement methods are in good agreement [18].

Table 2.4 A comparison of alpha activity measurement between the conventional radio-
chemical method and direct alpha measurement using the PIPS CAM detector [18].
 35

Computer Modeling

The scientific model is a simplification of a real system which is used for

prediction and system control. Modeling determines how parameter changes affect the

outcome and the system as a whole. There are many types of models including iconic

models which pictorially or visually represent a system, analog models that employ a set

of properties to represent another set in the desired system, and symbolic models which

require logical or mathematical operations to formulate a solution. Symbolic models are

classified as either dynamic or static. Dynamic models possess a time variable

interaction while static models lack a time variation. Symbolic models are also defined

as either deterministic or stochastic [19]. A deterministic population model makes a

single prediction about the future state of a population while a stochastic model

associates each possible future state with a probability and incorporates at least one

random variable. The models could be applied to the same population or example;

however, they will not necessarily have a matching mean or outcome [20].

Monte Carlo simulation is a popular form of stochastic modeling which is applied

in all areas of science and mathematics [20]. The Monte Carlo method was developed

during the Manhattan Project as a way to integrate otherwise impossible mathematical

functions. One of the first examples of the method was Buffon’s estimation of π as

needles were thrown onto a grid. The Monte Carlo method was used in the examination

of the Boltzmann equation and also in the estimation of the correlation coefficient in

Gosset’s t-distribution along with numerous other applications. The Monte Carlo method

is useful in the modeling of a random sampling from a population or for doing statistical

experiments [21].
 36

Monte Carlo modeling is the simulation of physical and mathematical problems

using random variables and known probability distribution functions. A random variable

is one in which the user does not know the value; although, the estimated values and their

probabilities are known. The simulations are made possible by modern computers and

random number generators. Although a computer generated random number sequence is

impossible, pseudo-random numbers have proved to be acceptable for Monte Carlo

purposes and are assumed to be independent [22]. A pseudo-random number generator

repeatedly simulates variables in a scenario resulting in many possible outcomes with a

mean, variance, and the probability density functions of the output variables [20].

Various Monte Carlo modeling software, such as the SRIM/TRIM package,

GEANT4, and the MCNP codes, are used for simulating alpha particle behavior in a

medium, however, these models were not recommended for spectrometry simulations

[23]. The SRIM package, of which TRIM is a subset, was designed for the calculation of

the range and stopping power of alpha particles or ions through various materials. The

program uses statistical algorithms which allow an averaging of the collision results

between the calculated collisions and gives a full quantum mechanical treatment to the

ion- atom interactions. The TRIM package is capable of calculating damage cascades,

ion distribution, and surface sputtering along with neutron, electron, proton cascades and

the ion energy, angle, and position [24]. GEANT4 is also a simulation toolkit for the

passage of charged and uncharged particles through matter. The GEANT4 package was

created at CERN and is based in the C++ computer language. GEANT4 allows complex

geometries, various materials and fundamental particles, and the responses of sensitive

detector components. The program tracks particles through materials and external
 37

electric fields, explores particle interactions, and generates event data. The program is

particularly useful for nuclear physics, medical physics, accelerator design, and space

engineering [25]. The MCNPX, Monte Carlo N-Particle Transport Code Extended, was

created at Los Alamos National Laboratory for the transport simulation of nearly all

nuclear particles [26]. The Fortran 90 based program allows for three dimensional

geometries and time dependence. The program simulates the transport of 34 various

particle types at different energies, light ion recoil, and fission multiplicity in addition to

many other applications [26].

A Fortran 95 Monte Carlo code, advanced alpha-spectrometric simulation

(AASI), was created for the simulation of the energy spectra of alpha particles. The

modeling program takes into account the decreasing energy of the charged particles and

the Rutherford scattering of the alpha- nucleus interaction. The program also includes the

accessibility of various sample types such as thick samples, aerosol particles, and non-

uniform samples [23]. The AASI Monte Carlo modeling program was developed for the

simulation of energy spectra of alpha radioactivity from an aerosol sample. Previously

published data for aerosol experimentation was used to validate the simulation program

as the experimental and simulated results agree within one standard deviation. The most

significant section of the simulation code, alpha energy loss, was constructed by putting a

thin Mylar® sheet between the source and the detector. The simulation proved to be a

success, as the simulated spectra followed the experimental data almost exactly [27].

The propagation of particles though matter was dictated by two main processes in

the AASI program: scattering and absorption. The emission of a particle from a

randomly selected point in the matrix, the distance it traveled to a scattering point, the
 38

determination of the next direction vector or scattering angle, and the continuous loss of

energy during each step for a charged particle was incorporated into the alpha

spectroscopy simulation code. If the particle energy was below a set cut-off value then

the particle tracking stopped, however, if the energy was above the value, the program

continued to calculate the distances traveled and the energy loss. The AASI program also

took into account the average solid angle which corresponded with the set up geometry in

the calculation of the geometrical detection efficiency. The energy loss of the alpha

particle was calculated using stopping power parameterizations and the energy loss

straggling of alpha particles was approximated by a Gaussian energy distribution [23].

This approximation was not necessarily correct for very thin absorption layers; however,

it gave a reasonable estimate for most cases. The alpha particles were assumed to travel

in a straight line with the exception of scattering on the planchet. The program also took

into account coincidences between particles emitted from the same source. The pulse

summation between the alpha particle and daughter emissions occurred as the excited

state lifetimes were usually shorter than the integration time of the data acquisition

electronics [23].

The AASI computer model contained two limitations: complete energy absorption

in the theoretical detector and a dependency on the response of the detector. The

remaining energy of the alpha particle after passing through the media between the

source and the detector was deposited in the detector. The alpha particles were not

tracked inside of the detector and the energy loss was not calculated as it was assumed

that all of the remaining energy was deposited in the detector. Additionally, several

parameters of the detector were added to the AASI code which eliminated the
 39

independence of the program from the detector such as the detector full width half

maximum (FWHM) and the parameters of the exponential tailing function. The detector

response of a silicon detector to monoenergetic alpha particles was not observed as a

Gaussian fit, but as a double exponential tailing function. The function was entered into

the program with a Gaussian convolution which was then applied to the simulated spectra

[23].

A Monte Carlo simulation was developed by Ferrero et al., to model the scattering

and slowing down of alpha particles in 2π geometry. The alpha particles traveled in a

straight line with the exception of Rutherford scattering with the nucleus and alpha

backscattering as the particles were absorbed into the media. The scattering of the alpha

particle with the nucleus of the absorbing material was considered to be the most

important consideration in both the backscattering and self-absorption correction factors.

The elastic scattering relationship, the scattering of the alpha particle off the nucleus of

the absorbing atom, was derived from classical mechanical theory using center of mass.

Inelastic scattering, the scattering of the alpha particle off the electrons of absorber

atoms, resulted in a loss of energy without significant angular deviations. The energy

loss was calculated from the electronic stopping power data of Ziegler [28]. The

constructed Monte Carlo model was tested against experimental data from a 2π ionization

chamber coupled to a multi-channel analyzer. Various thicknesses of UF4 corresponding

to 4 to 18.8 µg/cm3 were electroplated onto stainless steel planchets with diameters of 1.0

and 2.5 cm, each. The Monte Carlo simulation agreed with equivalent data from

experimental measurements and theoretical calculations within the estimated

 40

uncertainties, in regard to the ratio of counts and activity. The simulated energy spectrum

also had good agreement with the experimental energy spectrum [28].

The simulation by Ferrero et al., successfully calculated the backscattering

coefficients for different material planchets and various alpha emitters in 2π geometry.

The simulation was applied to a semiconductor detector in the investigation of alpha

particles leaving the planchet at extreme angles. Energy resolution depended on the

characteristics of the detector, geometry of the set up, composition and thickness of

deposited source, and the planchet material. The simulation program was used to model

contributing parameters of the peak shape such as the standard deviation, the distance

between the source and detector, the window thickness, and the thickness of the source

itself. The simulated spectra were distributed using a Gaussian function with various

standard deviations and were shown to be symmetric. A standard deviation of 10 keV

was used in the rest of the simulations. The alpha pulse height spectrum broadened as the

distance between the alpha source and detector decreased due to the increase in extreme

alpha angles. The particles which hit the detector at extreme angles lost more energy in

the window and caused an asymmetry in the low-energy peak region. An increase in the

window thickness also caused asymmetry and the FWHM to increase along with the shift

toward the lower energies. As the thickness of the source was increased, the peaks

widened and shifted to the lower energies. Experimental data was taken with a

semiconductor to correspond with the simulated pulse height spectra. The spectra were

in good agreement with the simulation. The backscattering of the alpha particles in the

source and support were tracked and determined to be insignificant to the peak shape

[29].
 41

The determination of particle size, shape, and composition of radioactive aerosols

were attempted by Pickering using alpha energy spectrometry simulations [30]. Monte

Carlo calculations of the energy spectra were conducted using spherical and elliptical

particles as a function of particle size and orientation inside of a volume. Past methods of

particle size determination using radioactivity included autoradiography of filter deposits

and also scintigraphy, when alpha particles interacted with ZnS screens. Both methods

were slow and inconveniently relied on manual track counting. An attempt to automate

the alpha autoradiography by substituting position sensitive semiconductors for the

photographic or polymer sheet detectors proved useful only at low concentrations. The

procedure failed if more than one particle was present on a detector and also if the

particles fell into the space between the detector display. In an attempt to improve upon

the current technologies, a solid-state surface barrier detector was investigated for the

feasibility of using alpha spectroscopy to determine size and composition of radioactive

particles [30]. The dependence of detector resolution on the spectral peak shape was

investigated with samples of various thicknesses. A theoretical monolayer source, an

experimental monolayer source, a semi-infinitely thick source, and a particulate source

resulted in varying pulse height distributions as illustrated in Figure 2.19. The variation

in pulse height distribution demonstrated that peak resolution was a function of the path

length of the alpha particle within the material and also portrayed the difference in peak

shape resulting from the larger and smaller particles in the particulate source. In addition,

the particle size, particle shape, orientation, and the solid angle of the alpha particles were

investigated using Monte Carlo calculations. The orientations of the alpha particles

inside of the particulate were randomly selected along with the direction of the alpha
 42

particles. If the angle between the direction of the alpha particle and y-axis was smaller

than the designated angle than the alpha particle was traveling towards the detector. The

distance to the surface of the particulate was calculated and the energy of the alpha

particle leaving the spheroid was calculated. Negative energies of the alpha particles

meant the alpha energy was completely deposited inside of the particulate and the alpha

particles were never detected. The simulated alpha spectra showed various peak shapes

for particulate size and shape; however, they were difficult to quantify. Although the

elliptical shape of a particle was expected to be determined by detecting the particle from

different angles, the simulated alpha spectra remained hard to quantify in a setting with a

heterogeneous distribution and an unknown density. The method did prove acceptable

for radioactive aerosols produced from the grinding machines in the manufacturing of

nuclear fuels, however. The particles were equiaxial, well characterized, and consisted in

a log-normal distribution; which allowed for a simple experimental method to be derived

which allowed for the particle size to be determined [30].

 43

Figure 2.19 The detection resolution as a function of source thickness and size [30].
 44

CHAPTER 3

RESEARCH OBJECTIVES

An evaluation of the effectiveness of a ZnS(Ag) detector in the quantification of

alpha-emitting radionuclides in aqueous solutions was conducted and the data was

compared and contrasted with the PIPS CAM detector. The experimental results from

the ZnS(Ag) detector and both experimental and previously published data for the PIPS

CAM silicon detector were used to determine which instrument was best equipped for the

direct detection of alpha particles from solution [18].

Several parameters were taken into consideration including solution

concentration, the separation distance between the liquid and detector, solution density,

detection surface area, and beta and gamma interference in the detectors. The absolute

detection efficiency of the alpha particles from the liquid surface and pulse height spectra

characterized with the ZnS(Ag) detector were compared to electroplated standards and

the PIPS CAM detector.

The experimental pulse height spectra of the infinitely thick aqueous sources were

compared to theoretical computer-generated spectra. The spectral shapes of the ZnS(Ag)

and PIPS CAM detectors were discussed in relation to the theoretical expectations. The

theory of alpha particle absorption in matter was explored and the effects of light

propagation and collection were discussed. The effects of thin and thick scintillator

thicknesses on the pulse height spectra were determined for the ZnS(Ag) detector.
 45

Specific tasks conducted to meet the objectives include:

• Determined if the ZnS(Ag) detector was suitable for on-line alpha assay

o Experimental data

• Determined if the PIPS CAM detector was suitable for on-line alpha assay

o Experimental data

o Published data

• Understood and explained the shape of the ZnS(Ag) pulse height spectra

o Experimental data

o Computer simulations

o Theory of alpha particle interactions with matter

 46

CHAPTER FOUR

MATERIALS AND METHODS

Apparatus

The energy spectra of aqueous alpha solutions were achieved by coupling a

ZnS(Ag) detector to a multi-channel analyzer. The detection apparatus was configured

using a Ludlum Model 43-1 ZnS(Ag) scintillation detector (Ludlum Measurements, Inc.,

Sweetwater, TX). The detector had an active face of 78.5 cm2 with a layer of 0.8 mg/cm2

aluminized Mylar® over a 3.5 mg/cm2 layer of ZnS(Ag) scintillation crystals. A pre-

fabricated 14 mg/cm2 scintillation layer was later supplemented onto the detector. Using

the Canberra Model 2006E preamplifier (Canberra Industries, Inc., Meriden, CT) or the

Ludlum Model 297 Signal Splitter, the ZnS(Ag) detector was connected to the Bertan

Associates Model 305 DC high voltage power supply (Bertan Associates, Inc., Syosset,

NY) and the Ortec 572 amplifier (Ortec, Inc., Oak Ridge, TN) as shown in Figure 4.1.

The amplified signal of the ZnS(Ag) detector was read by an Aptec-NRC MCArd analog

to digital converter and multi-channel analyzer which transmitted the 511 channel energy

spectra to the computer. The high voltage was set at 800 V when using the charge-

sensitive preamplifier and 850 V when using the signal splitter. Details of the high

voltage optimization are presented in Appendix A.

 47

Experimental Setup Components of Ludlum Detector

1. Ludlum Model 43-1 ZnS Detector (Ludlum 6. Photomultiplier Tube
Measurements, Inc., Sweetwater, TX) 7. Plastic Light Pipe
2. Ludlum Model 297 Signal Splitter/ 8. 3.5 mg/cm2 ZnS(Ag) Layer
Canberra Model 2006E Preamplifier 9. 0.8 mg/cm2 Mylar® Layer
3. Bertan Model 305 DC Power Supply Additional Apparatus
(Bertan Associates, Inc., Syosset, NY) 10.Spacers composed of
4. Ortec Model 572 Amplifier washers
(Ortec, Inc., Oak Ridge, TN) 11.Sample holding cup
5. Aptec MCArd Multi-channel Analyzer machined from
polyvinylchorine (PVC)

Figure 4.1 The experimental setup for the ZnS(Ag) detector.

A CAM 450AM PIPS detector (Canberra Industries, Inc., Meriden, CT), typically

utilized for continuous air monitoring (CAM), was utilized. The detector had a 0.5 µm

aluminum coating over the 450 mm2 detection surface creating a light proof detector. A

1 µm polymer coating also created a detector surface which was resistant to chemical and

physical properties. The detector was connected to a Canberra Alpha Spectrometer

(Model 7401VR) and the data was recorded with an Aptec-NRC MCArd analog to digital

converter and multi-channel analyzer as shown in Figure 4.2. The bias was set at 24.3V

for the PIPS CAM detector. The original spectra for the PIPS CAM detector had 1024

channels, however, the data was re-binned to 511 channels using Microsoft Excel in

order to directly compare the PIPS CAM data to the data from the ZnS(Ag) detector.

12. Canberra Alpha Spectrometer Model 7401

13. Canberra PIPS CAM 450AM

Figure 4.2 The experimental set up for the Canberra PIPS CAM detector.
 48

Two detection set ups were required: one in the Pacific Northwest National

Laboratory (PNNL) and one at Rich Laboratory of Clemson University. The PIPS CAM

detector data was only acquired in PNNL. The Ludlum Model 43-1 ZnS(Ag) detector,

Figure 4.3, was transported between the two locations which allowed for the same

detector to be used in all measurements. The same models, though different pieces of

equipment, of the Bertan high voltage, Ortec amplifier, and the Aptec-NRC MCArd

analog to digital converter and multi-channel analyzer were used in both locations. The

signal splitter was used in the data taken at Pacific Northwest National Laboratory and

select experiments at Clemson, while the charge-sensitive preamplifier was used only at

Clemson, Table 4.1. Differences in the pulse height spectra of data taken with the

ZnS(Ag) detector in conjunction with the signal splitter and charge-sensitive preamplifier

were explored in Appendix B. A Tektronix oscilloscope TAS 465 was used in

conjunction with the Clemson apparatus in order to observe the electronic signal from the

ZnS(Ag) detector and its associated electronics.

Figure 4.3 The Ludlum Model 43-1 ZnS(Ag) scintillation detector clamped to a stand
with bolts propping it over a PVC sample holding cup.
 49

Table 4.1 The experiments which utilized the signal splitter or charge-sensitive
preamplifier.

Signal Splitter Charge-sensitive preamplifier

Solution Concentration Energy Resolution and Spectral Shape
Solution Density Absolute Detection Efficiency
Air Gap Separation Detection Area
Minimum Detectable Concentration
Thin versus Thick Scintillation Layer

The ZnS(Ag) scintillator was covered by a Mylar® layer designed to make the

instrument light tight; however, a small scratch in the surface of the Mylar® created a

light leak. A black cloth was wrapped around the detector to minimize the effects of the

light leak. The light shield corrected the problem as observed with the oscilloscope. The

light shield was used in all of the experiments conducted at both Pacific Northwest

National Laboratory and Clemson University.

The liquid samples were measured after pipetting the aqueous solutions into the

sample holding cup which was placed at the center of the active face of the detector. The

sample holding cups were made from polyvinylchoride (PVC). The large ZnS(Ag)

detector face was masked to allow a smaller opening, a condition to mimic the

companion experiments conducted with the PIPS detector. After 2 mL of liquid was

placed into the PVC sample holding cup, the resulting thickness of the liquid source was

4.1 mm. The detector was placed on top of spacers that raised the detector face to 5.2

mm above the surface of the liquid as shown in Figure 4.3. The PNNL sample holding

cup had an inner depth of 9.2 mm and an inner diameter of 23.9 mm. The inner diameter

of the sample holing cup corresponded to the active surface of the Canberra PIPS CAM

detector. The Clemson sample holding cup, also made of PVC, was created to mimic the

PNNL sample holding cup with regards to the 5.2 mm air gap distance when filled with 2
 50

mL of solution. The Clemson sample holding cup had an inner diameter of 23 mm and

an inner depth of 9.8 mm. The 4.15 cm2 Clemson sample holding cup was slightly

smaller than the 4.48 cm2 cup used at PNNL. A large diameter PVC cup was constructed

at Clemson University to correspond with the large detection area of the ZnS(Ag)

detector. The large sample holding cup had an inner diameter of 108 mm and inside

depth of 5.6 mm. The arrangement provided an air gap of 5 mm when 30 mL of solution

was present in the cup. The steel rim of the ZnS(Ag) detector was supported by the sides

of the large cup. Large diameter shields, with inner diameters of 40.4 mm and 79.3 mm,

restricted the detection area of the large detection cup for experiments which investigated

the effects of solid angle.

A 6.35 mm Plexiglas shield (749.3 mg/cm2) acted as an absorber between a beta

source and the detectors and also acted as a platform for the alpha electroplated standard

(EPS), Figure 4.4(a). In the absence of the beta source, the shield was used as a stage for

the alpha electroplated standard in an attempt to maintain solid angle, Figure 4.4(b). The

distance between the alpha electroplated standard and the detector was consistent for the

experiments involving the charge-sensitive preamplifier, however, the distance could not

be replicated for the signal splitter experiments.

 51

(a) (b)

Figure 4.4 The experimental set up for the (a) combined beta source and the
electroplated standard and (b) the electroplated standard only.

Reagents and samples

All chemicals used at PNNL were of analytical grade. De-ionized water (18 MΩ-

cm, Barnstead E-Pure, Dubuque, IA) was used in the preparation of the reagents. The
244 241 239
actinide solutions were prepared by diluting stock solutions, Cm, Am, Pu, and
230
Th, with 2% nitric acid and the activities of 2.1 x 103 + 4.2 Bq/mL, 6.9 x 103 + 6.4

Bq/mL, 5.0 x 103 + 7.6 Bq/mL, and 1.8 x 104 + 12.2 Bq/mL, respectively, were

standardized with the Packard 2550 TR/AB liquid scintillation counter assuming 100%
241
detection efficiency. The Am EPS had an activity of 259 + 3.9 Bq. The samples of

various densities were prepared by the addition of sodium nitrate salts to 2% nitric acid

actinide solutions. The resulting density was confirmed gravimetrically and the activity

of each sample was confirmed with the liquid scintillation counter.

The solutions and electroplated standards used at Clemson University included

four actinide solutions, three NIST electroplated standards, and a plutonium electroplated

source made at the Savannah River Site. The Clemson University actinide solutions
 52
244 239 238
consisted of Cm, Pu, U/234U (natural uranium), and 238
U (depleted uranium) at

concentrations of 220 + 3.1 Bq/mL, 16200 + 26.8 Bq/mL, 5137 + 15.6 Bq/mL, and 11.3

+ 0.16 Bq/mL, respectively. The solutions were below pH 2 and the concentrations were

standardized with the Wallac 1415 DSA liquid scintillation counter assuming 100%

detection efficiency. The natural uranium solution was assumed to have an alpha energy

of 4.51 MeV, a calculated average of the 238U and 234U alpha particle energies, while the

total alpha activity was utilized. The NIST electroplated alpha standards included a
241 241 238
7151.7 + 93.0 Bq Am, 1664 + 18.1 Bq Am, and 489.9 + 6.4 Bq Pu. The

plutonium electroplated source was prepared at the Savannah River National Laboratory
238 239 240
using NIST traceable sources. The source consisted of the isotopes Pu, Pu, Pu,
241 242 241
Pu, Pu, and Am with the respective activities, 0.03, 264.07, 2.28, 0.37, 0.09, and

0.46 Bq. A high-activity 90Sr(90Y) source, 7.4 x 108 Bq, and a 6.88 x 105 Bq 137Cs source

were also used to investigate beta and gamma interaction in the ZnS(Ag) detector.

The effective activity, the aqueous radioactivity which contributed to a signal in

the detector, was calculated based on the known activity concentration, the surface area

of the detector, and the active depth. The active depth was an estimation of the solution

thickness which emitted alpha particles that were capable of leaving the solution. The

estimation of active depth allowed for a more realistic detected volume. The active

volume of solution which the detectors were detecting was determined by multiplying the

surface area of the cup, A, by the active depth, x. Equation 13 gives the relationship

between the effective activity (EA), solution concentration (C), and active volume.

⎛ Bq ⎞
EA( Bq ) = C ⎜ ⎟ A ( cm ) x (cm) (13)
2

⎝ ml ⎠
 53

The maximum range of an alpha particle in solution was calculated using equations 1, 2,

and 3 and was taken as the active depth, x, in solution. An alpha particle which

originated below the maximum depth deposited all of the alpha energy into the aqueous

matrix and did not leave the solution. The active depth of the radionuclide solutions, the

maximum depth from which an alpha particle reached the surface, was dependent on

alpha energy as the ranges of alpha particles through the solution was a function of alpha
244
energy. There was a 35% difference in the alpha ranges through solution for Cm, the
238
maximum alpha energy used, and U, the isotope with the least energy used—54.5 µm

and 30.6 µm, respectively. The alpha ranges for the various energies and the solution

effective activities used in these experiments are shown in Table 4.2.

Table 4.2 The active depths and solution effective activities for various radionuclide
energies.

Radionuclide Alpha Active PNNL Solution Clemson Solution

Energy Depth Effective Effective Activity
(MeV) (µm) Activity (Bq) (Bq)
244
Cm 5.81 54.5 52.3 + 0.10 5.36 + 0.076
241
Am 5.49 49.8 155.1 + 0.17
239
Pu 5.14 44.7 99.4 + 0.13 325.3 + 0.539
230
Th 4.69 38.0 302.8 + 0.21
234
U/238U 4.51* 35.4 80.2 + 0.236
238
U 4.18 30.6 0.18+ 0.002
*Average energy of the alpha particle of the two radionuclides

Procedure

Following each sample analysis, the sample holding cups were washed with two

1M nitric acid rinses and two distilled water rinses before being scrubbed dry. Prior to

each experimental set, a background was taken with the detector in order to ensure that

there was no significant contamination on the detector. The background was also

subtracted from the upcoming data set.

 54

CHAPTER 5

RESULTS AND DISCUSSION

The ZnS(Ag) and PIPS CAM detectors were characterized to determine the

detector best equipped to quantify alpha radioactivity from aqueous solutions. The

spectral shape and energy resolution of the pulse height spectra were evaluated and

compared to theory. The absolute detection efficiency of each detector was determined

with and without the presence of beta and gamma interference. The effects of solution

concentration, detection area, solution density, and separation distance on the ZnS(Ag)

detector were investigated and compared with measurements that were previously

conducted with a PIPS CAM detector. Two detection layer absorber thicknesses of

ZnS(Ag) were studied. A computer program was designed to model the experimental set

up and the experimental and theoretical pulse height spectra were compared.

Energy Resolution and Spectral Shape

241 241
The pulse height spectra of an Am electroplated standard and an Am

aqueous solution detected with a PIPS CAM detector are compared in Figure 5.1(a). The

pulse height spectrum of the 241Am electroplated standard measured with the PIPS CAM

detector was relatively narrow and characterized the finite energy distribution of an

electroplated standard. The pulse height spectrum demonstrated a tailing effect due to the

absorption through the air gap and the protective surface layer associated with the CAM.
241
Despite the absorption layers, the alpha peak energy resolution for the Am

electroplated standard was 3.1%. The PIPS CAM detector is renowned for its resolution

due to the solid-state properties of the detector. The PIPS CAM detector allowed for all
 55

the alpha energy to be deposited in the infinitely-thick active detection region. The
241 241
leading edges in the Am aqueous solution and Am electroplated standard pulse

height spectra from the PIPS CAM detector intercepted in Figure 5.1(a) at the maximum

alpha energy that was deposited in the detector. The low energy noise tail present in the

pulse height spectrum for the aqueous solution, Figure 5.1(a), was absent in the pulse

height spectrum of the electroplated standard due to a higher LLD setting. The pulse

height spectrum of the aqueous solution, an infinitely-thick source, illustrated a broad

peak and alpha energy tailing as the alpha particles lost energy in the solution through

inelastic scattering within the medium. An infinitely-thick source possessed a material

thickness greater than the alpha range in that material—a property characteristic of liquid

samples. The alpha particle which originated from the top of the liquid surface deposited

the maximum alpha energy into the detector. Alpha particles located below the surface

lost a significant amount of energy to the aqueous matrix. Alpha particles which

originated deep in the solution never departed from the solution and deposited all of the

alpha energy into the aqueous matrix. The aqueous solution had a high stopping power

and was the leading contributor of alpha energy absorption for aqueous solutions.
239 239
The pulse height spectra of a Pu electroplated standard and a Pu aqueous

solution obtained with the Ludlum Model 43-1 ZnS(Ag) detector in conjunction with the

charge-sensitive preamplifier were compared in Figure 5.1(b). The energy resolution of

239
the pulse height spectrum of the Pu electroplated standard taken with the ZnS(Ag)

detector was 118%—much greater than that of the PIPS CAM detector. The spectral
239
shape of the ZnS(Ag) pulse height spectrum for the Pu (5.147 MeV) electroplated

standard is in good agreement to the pulse height spectrum for a 5.5 MeV alpha source in
 56

Figure 2.6. The poor resolution of the alpha peak was attributed to scintillation properties

of the detector. The ZnS(Ag) scintillation light intensity created was proportional to the

energies deposited in the scintillator. The scattered light was partially absorbed in the

scintillation layer and light pipe which resulted in further degradation of energy
239
resolution. The ZnS(Ag) pulse height spectrum of the Pu aqueous solution, Figure

5.1(b), did not drastically differ from the pulse height spectrum of the electroplated

standard. The poor resolution of the ZnS(Ag) detector created difficulties in

distinguishing alpha peak characteristics. The leading edges of the pulse height spectra

for the electroplated standard and aqueous solution did converge at the maximum energy
239
as expected. The peak maximum of the pulse height spectrum for the Pu aqueous

solution shifted to the lower energies relative to the pulse height spectrum of the

electroplated standard. The pulse height spectrum for the aqueous solution had a greater

number of lower-energy alpha particles as the particles from the solution have a

decreased energy due to energy absorption in the solution matrix.

 57

0.0004 Liquid 0.008

EPS

Effective Activity (Bq) for

Activity (Bq) for EPS

Net Count Rate (cps)/

Net Count Rate (cps)/

0.0003 0.006

Liquids
0.0002 0.004

0.0001 0.002

0 0
0 100 200 300 400 500
Channel Number

(a)

0.0005 0.0015
Activity (Bq) for 239Pu Solution
Net Count Rate (cps)/ Effective

Activity (Bq) for 239Pu EPS

0.0004 0.0012

Net Count Rate (cps)/

EPS

0.0003 0.0009

0.0002 0.0006

Liquid
0.0001 0.0003

0 0
0 100 200 300 400 500
Channel Number

(b)

Figure 5.1 The pulse height spectra of a) 241Am and b) 239Pu samples for the PIPS CAM
detector and ZnS(Ag) detector, respectively. Spectra corresponding to both aqueous
radioactive sources and electroplated standards are shown.
244
The PIPS CAM detector was characterized with four aqueous solutions, Cm,
241 239 230
Am, Pu, and Th, while the ZnS(Ag) detector was characterized with three
244 239 234
aqueous solutions, Cm, Pu, and U/238U. The PIPS CAM pulse height spectra in

Figure 5.2(a) corresponds with the published graph, Figure 2.14(a), but with better
 58

counting statistics. The differences in the resolution of the PIPS CAM and ZnS(Ag)

detectors were observed in the leading edge of the pulse height spectra in Figure 5.2(a)

and (b). The leading edges of the PIPS CAM detector were clear and quickly descended

at the maximum alpha energies. The poor resolution of the ZnS(Ag) detector resulted in

prolonged leading edges. The leading edges of the ZnS(Ag) pulse height spectra were

quantified at the tenth maximum (FWTM) due to the broadening effect of the ZnS(Ag)

detector. The spectral leading edges of the two extreme alpha energies, 5.8 MeV from
244 230
the Cm solution and 4.7 MeV from the Th solution were separated by 95 channels

for the PIPS CAM detector. While the leading edges of the ZnS(Ag) pulse height spectra

were separated by 70 channels for the 5.8 MeV 244Cm solution and a 4.8 MeV 234
U/238U

solution. The energy calibration for the PIPS CAM detector had a larger variation in the

number of channels than the ZnS(Ag) detector.

The published PIPS CAM pulse height spectra, Figure 2.14(a), and the

experimental data in Figure 5.2(a) are essentially the same data but plotted differently.

The only real differences between the data are the count times; in Figure 2.14(a) the

count time was 10 minutes, while the count time for the data presented in Figure 5.2(a)

was 10 hours. The y-axis in Figure 2.14(a) consisted of counts normalized to activity

concentration.

⎛ Bq ⎞
ConcentrationA ⎜ ⎟
NetCountsperChannelB * ⎝ ml ⎠ (14)
⎛ Bq ⎞
ConcentrationB ⎜ ⎟
⎝ ml ⎠

The net counts in each channel from solution B were normalized to the concentration of

solution A, the control solution. The authors of the original work assumed that the units

of concentration (Bq/mL) cancelled when normalized, but the active volumes of the
 59

detected solutions were not the same for the different alpha energies which led to a

misleading comparison. The y-axis in Figure 5.2(a) was created using the net count rate

divided by the effective activity. The effective activity was calculated from the solution

concentration using the active volume of the solution.

0.0005
EffectiveActivity (Bq)
Net Count Rate (cps)/

0.0004 244
Cm

0.0003 241
Am

239
0.0002 Pu

230
0.0001 Th

0
0 100 200 300 400 500
Channel Number

(a)

0.0005
244
Cm
EffectiveActivity (Bq)
Net Count Rate (cps)/

0.0004
239
Pu
0.0003
234
U/238U
0.0002

0.0001

0
0 100 200 300 400 500
Channel Number

(b)

Figure 5.2 Spectra of the aqueous solutions recorded with the (a) PIPS CAM detector
and (b) ZnS(Ag) detector demonstrated the spectral shape.
 60

Absolute Detection Efficiency

In the absence of beta and gamma radioactivity, the absolute detection efficiencies

of the PIPS CAM detector were 20% higher than that of the ZnS(Ag) detector, Figure

5.3. To find the absolute detection efficiency (%), the net count rate was divided by the

effective activity of the solution (Bq) and multiplied by 100. The net count rate was

calculated by dividing the counts in the alpha region of interest (ROI) by the live time

and subtracting the background count rate. The calculated absolute detection efficiencies

of the 100-511 ROI pulse height spectra measured with the PIPS CAM detector were
244 241 239
9.01 + 0.02%, 8.63 + 0.02%, 7.95 + 0.01%, and 7.15 + 0.01% for Cm, Am, Pu,
230
and Th solutions, respectively. The absolute detection efficiencies are different from

the counting sensitivities listed in Chapter 2 as the active depths were used in the above

calculations. The absolute detection efficiencies of the ZnS(Ag) pulse height spectra, 20-

511 ROI, were 7.55 + 0.14%, 6.08 + 0.02%, and 4.96 + 0.03% for the 244Cm, 239Pu, and
234
U/238U solutions, respectively.

PIPS CAM (100-511)

10
Absolute Detection Efficiency (%)

y = 1.7052x - 0.826
R2 = 0.9926
8
ZnS(Ag) (20-511)
y = 2.2597x - 5.2793
6
R2 = 0.9757

4
ZnS(Ag) (55-511)
y = 2.255x - 5.9994 PIPS CAM (275-511)
2 y = 2.9789x - 12.58
R2 = 0.9807
R2 = 0.9951
0
4 4.5 5 5.5 6
Alpha Energy (MeV)

Figure 5.3 The absolute detection efficiency for the PIPS CAM and ZnS(Ag) detectors
for alpha radioactivity in aqueous solutions.
 61

Beta and gamma radioactivity, prevalent in high-level waste, interfered with the

alpha region of interest of both detectors despite minimal beta and gamma detection

efficiencies. The absolute detection efficiencies were calculated as (1.0 + 0.1) x 10-2%

and (5.7 + 1.5) x 10-5% for 2 mL of 90

Sr(90Y) beta and 137
Cs gamma solutions,

respectively, for the ZnS(Ag) detector with a region of interest of 5-511. Although the

absolute detection efficiencies for beta and gamma radioactivity were low, they remain

significant due to the high activity levels present in HLW. The absolute detection

efficiency of beta particles were determined to be higher than the absolute detection

efficiency of gamma rays as beta particles lose energy more readily in absorption layers.

Only a fraction of the incident gamma rays interacted in the thin ZnS(Ag) layer. The
90
Sr(90Y) pulse height spectrum, measured with the ZnS(Ag) detector, was located at

channels less than 50 while the 137Cs gamma radioactivity was located at channels lower

than 15, Figure 5.4. The beta and gamma pulse height spectra were located in the low-

energy region of interest which allowed for a beta and gamma upper level discriminator

(ULD) to be set at 55 channels. The beta and gamma pulse height spectra for the PIPS

CAM detector in Figure 2.14(b) more severely spilled over into the alpha region of

interest which resulted in the highest alpha detection efficiency. The beta and gamma

pulse height spectra, found in channels as high as 500 (re-binned channel 250),

significantly reduced the alpha detection region of interest of the PIPS CAM pulse height

spectra, Figure 5.2(a), to channels above 550 (re-binned channel 275). The low beta and

gamma counting sensitivity of the PIPS CAM detector in Table 2.3, 4.06 x 10-4 and 4.69

x 10-6%, for 90Sr(90Y) and 137Cs, respectively, was calculated as having absolute detection

efficiencies of 8.12 x 10-2% and 9.38 x 10-4%, respectively. The absolute detection
 62

efficiencies were found as the entire detection volume, 2 mL, was utilized in both the

PIPS CAM and ZnS(Ag) experiments in the detection of beta and gamma radioactivity.

Despite the absolute detection efficiencies, an alpha region of interest (275-511) was

chosen for the PIPS CAM detector to minimize interference from beta and gamma

radioactivity while maximizing the PIPS CAM alpha detection efficiency.

1.E-08
90
Sr(90Y)
Net Count Rate (cps)/

8.E-09
Activity (Bq)

6.E-09 137
Cs

4.E-09

2.E-09

0.E+00
0 100 200 300 400 500
Channel Number

Figure 5.4 Beta particle and gamma ray interference in the ZnS(Ag) detector.

In the presence of beta and gamma radioactivity, the ZnS(Ag) detector had a 40%

better detection efficiency relative to the PIPS CAM detector. The absolute detection

efficiencies for the PIPS CAM detector of the reduced alpha ROI (channel 275-511) were
244 241 239
4.61 + 0.01%, 3.87 + 0.01%, 2.84 + 0.01%, and 1.32 + 0.01% for Cm, Am, Pu,
230
and Th, respectively, Figure 5.3. The absolute detection efficiencies for the ZnS(Ag)

detector for the reduced region of interest (channel 55- 511) were 7.19 + 0.13%, 5.37 +
244 239 234
0.02%, and 4.21 + 0.03% for Cm, Pu, and U/238U, respectively. The narrowed

alpha regions of interests which minimized the spillover of the beta and gamma events

resulted in lower detection efficiencies for both detectors. The absolute detection
 63

efficiencies of the PIPS CAM detector were reduced more significantly than the ZnS(Ag)

efficiencies because the beta and gamma ULD occurred at a higher channel.

In the absence of beta and gamma radiation, the PIPS CAM detector was better

suited for gross alpha counting than the ZnS(Ag) detector due to decreased detection

efficiency energy dependence. Higher-energy alpha particles were more efficiently

detected with both detectors than lower-energy particles as shown in Figure 5.3. The

slope of the efficiency calibration curve for the PIPS CAM, 1.71, however, was less than

that for the absolute detection efficiency of the ZnS(Ag) detector, 2.26, in Figure 5.3.

The PIPS CAM detector possessed a lower energy dependence than the ZnS(Ag) detector

making it more suitable for the gross detection of alpha particles from a solution which

contained several actinides. The PIPS CAM detector also appeared to be a more

consistent detector as the linear fit fell within uncertainty for the data points and the

regression estimate had a higher correlation coefficient. The increased uncertainty of the

ZnS(Ag) detector was attributed the low-activity natural uranium (234U/238U) solution as

total alpha activity was used in the calculation but the energy was assumed to be the

average of the two main isotopes, 234U and 238U.

In the presence of beta and gamma radioactivity, the energy dependence of the

PIPS CAM absolute detection efficiency increased while the ZnS(Ag) energy dependence

stayed the same. The change of slope for the absolute detection efficiencies of the

ZnS(Ag) detector, 2.26, was less than that for the PIPS CAM absolute detection

efficiencies, 2.98, Figure 5.3. High-energy alpha particles, 244Cm and 241Am, were more

easily detected with the PIPS CAM detector because a greater portion of the pulse height

spectrum was located above the ULD of the beta and gamma region of interest. The
 64
239 230
lower-energy alpha particles, Pu and Th, were more greatly affected as the bulk of

the pulse height spectra was located in the region below the ULD of the beta and gamma

region of interest. The beta and gamma radioactivity were located low enough on the

pulse height spectra of the ZnS(Ag) detector that the alpha region of interests for the

various radionuclides were equally affected.

Solution Concentration
241 230
Am and Th solutions were used to determine the relationship between

activity concentration and the net count rate of the ZnS(Ag) detector in conjunction with

the signal splitter using the 4.48 cm2 sample holding cup. A solution of 241
Am, 13,896

Bq/mL, was diluted to 6,948, 4,627, and 2,084 Bq/mL. A 230Th solution, 35,529 Bq/mL,

was diluted to 17,765, 11,831, and 5,329 Bq/mL. A sample of nitric acid was used as a

blank solution. The non-serial dilutions were conducted by diluting samples of the

original solution to the appropriate concentration. The data in Figure 5.5 suggested a
241 230
linear response of count rate with increased concentration. The Am and Th

solutions have different slopes for net count rate versus solution concentration in Figure
241
5.5 due to differences in detection efficiency. The Am solution possessed a greater
230
slope than the Th solution due to the higher detection efficiency of the high-energy
241 230
alpha particles. The absolute detection efficiencies for the Am and Th solutions,

8.42 and 6.14%, were slightly higher than those previously presented due to differences

between the signal splitter and preamplifier in the experimental set-up. The PIPS CAM

detector demonstrated a similar linear relationship between count rates and the varied

activity concentrations of 244Cm, 241Am, and 230Th as observed in Figure 2.15. The PIPS
 65

CAM detector also demonstrated differences in the absolute detection efficiency through

the variation of slopes in Figure 2.15.

20
Net Count Rate (cps)
16 241
Am
y = 1.8796x - 0.0196 230
Th
12 R2 = 0.9999 y = 1.0447x + 0.0918
8 R2 = 0.9997

0
0 5 10 15 20
Concentration (Bq/µL)

Figure 5.5 The net count rate of the ZnS(Ag) detector plotted against various
concentrations of 241Am and 230Th solutions.

Detection Area

One of the advantages of the Ludlum Model 43-1 ZnS(Ag) detector over the PIPS

CAM detector was the size of the detection area. The ZnS(Ag) detector in conjunction

with the charge-sensitive preamplifier had a detection surface which was previously

constricted for comparison with the PIPS CAM detector. The detection surface was

uncovered to utilize the larger detection surface. The 11.3 + 0.02 Bq/mL depleted

uranium (238U) solution was detected with surface areas of 4.15 cm2, 12.8 cm2, 49.4 cm2,

and 78.5 cm2. The surface area of 4.15 cm2 resulted in a minimum detectable

concentration that was greater than 11.3 Bq/mL. A large sample holding cup and

absorption shields were utilized to keep the surface area of the aqueous solution and the

detection surface consistent. The absolute detection efficiency increased with increasing

detection area as shown in Figure 5.6. The increased surface area of the aqueous solution
 66

was the only way to increase the sample volume due to the small range of alpha particles

in solution. The absolute detection efficiencies for the various detection areas, 12.8 cm2,

49.4 cm2, and 78.5 cm2 were 3.29 + 1.24%, 4.71 + 0.34%, and 6.30 + 0.24%, respectively
238
for U. The absolute detection efficiency of the 78.5 cm2 ZnS(Ag) detection area

almost doubled the absolute detection efficiency of the 12.8 cm2 detection area. The

absolute detection efficiency for the largest detection surface had the least uncertainty as

the associated count rates were higher. The extrapolation of Figure 5.6 also possessed

uncertainty values of more than 1% for the 4.15 cm2 detection area. The increased count

rates resulted in increased absolute detection efficiency and counting statistics while

uncertainty for the count rate was decreased.

8
Absolute Detection Efficiency

6
(%)

0
0 20 40 60 80
2
Detection Area (cm )

Figure 5.6 The Ludlum Model 43-1 ZnS(Ag) detector was used in the detection of three
detection surface areas of depleted uranium solution.

Minimum Detectable Concentration

The minimum detectable concentration of the ZnS(Ag) detector decreased with

increasing absolute detection efficiency and detection area. The minimum detectable
 67

concentrations for the ZnS(Ag) detector in conjunction with the charge-sensitive

244
preamplifier, Table 5.1, were calculated at 33.42, 50.39 and 79.38 Bq/mL for Cm,
239 234
Pu, and U/238U in a 4.15 cm2 sample holding cup, respectively. The minimum

detectable concentrations were calculated using equation 12, however, the ε represented

the effective absolute detection efficiency times the active volume instead of counting

sensitivity. The gross background count for the 20-511 ROI was 54 counts in 600

seconds and the absolute detection efficiencies were 7.55, 6.08, and 4.96% for the 244Cm,
239 234
Pu, and U/238U solutions, respectively. The variations in the minimum detectable

concentration for the radionuclides were due to the changes in the absolute detection

efficiency for the different radionuclides. The minimum detectable concentrations of the

PIPS CAM detector were published as 17.7, 23.0, and 35.0 Bq/mL for 244Cm, 241Am, and
239
Pu, respectively [18]. The PIPS CAM minimum detectable concentrations were lower

than those for the ZnS(Ag) detector due to the higher absolute detection efficiency of the

PIPS CAM detector in the absence of beta and gamma radioactivity and higher

background count rate attributed to suspected 241Am contamination present on the Mylar®

surface of the ZnS(Ag) detector.

The minimum detectable concentration for the maximum surface area of the

ZnS(Ag) detector used in these experiments, 78.5 cm2, for 3600 seconds was the lowest
238
at 0.28 Bq/mL for U, Table 5.1. The minimum detectable concentrations for the

smaller areas, 49.4 cm2 and 12.8 cm2, were 0.60 Bq/mL and 3.31 Bq/mL, respectively.

The projected minimum detectable concentration for the 4.15 cm2 sample holding cup
238
was calculated as 11.86 Bq/mL for U, greater than the solution concentration used.

The minimum detection concentration decreased with increasing detection area.

 68

The largest minimum detectable concentration of the ZnS(Ag) detector was 79.38

Bq/mL, a value over 10 times less then the maximum concentration allowed in the

Saltstone facility at the Savannah River Site [1]. The MDC for the entire detection face

of the Ludlum 43-1 ZnS(Ag) detector was 0.28 Bq/mL, over 1,000 times less than the

maximum concentration. The ZnS(Ag) detector was shown to be capable of detecting

alpha activity from HLW solutions.

Table 5.1 The minimum detectable concentration (MDC) for the various areas and
radionuclides.

Radionuclide Detection Area Time MDC

(cm2) (s) (Bq/mL)
244
Cm 4.15 600 33.42
239
Pu 4.15 600 50.39
234
U/238U 4.15 600 79.38
238
U 4.15 3600 11.86
238
U 12.8 3600 3.31
238
U 49.4 3600 0.60
238
U 78.5 3600 0.28

Solution Density

The density of the 241Am solution was increased with sodium nitrate salts in order

to demonstrate the effects of the high salt content in HLW on the detection capabilities of

the ZnS(Ag) detector in conjunction with the signal splitter. The normalized count rate

over concentration as obtained with the ZnS(Ag) detector decreased as the solution

density increased as shown in Figure 5.7. The normalized count rate over concentration,

or relative efficiency as referred to in Chapter 2, assumed unity at 1.01 g/mL as both the

count rate and concentration were normalized to the original 1.01 g/mL density solution.

The increased solution density decreased the range of alpha particles in solutions

resulting in decreased detection. Both the ZnS(Ag) detector, Figure 5.7, and the PIPS
 69

CAM detector, Figure 2.16, demonstrated decreased count rates with the increased

solution density. The ZnS(Ag) detector and the PIPS CAM detector had similar slopes of

-0.56 and -0.57, respectively, as the relationship was dependent on the density of the

solution and independent of the detector. The data taken with the PIPS CAM detector

had greater uncertainty because several actinide solutions were plotted and the

differences in absolute detection efficiency and alpha range were not taken into account.

The actinide solutions in both experiments were in the absence of beta and gamma

radioactivity.

The absolute detection efficiency of the ZnS(Ag) detector did not decrease despite
241
the increase of the Am solution density. The absolute detection efficiency was

calculated using the effective activity of the solution in the active volume of the sample.

The active depth of the alpha particle, a function of density, decreased as the solution

density increased as demonstrated by equations 1, 2, and 3. The active depth of the alpha

particle in solution decreased from 49.9 µm for the original solution density of 1.01 g/mL

to 39.8 µm for the 1.27 g/mL solution. Alpha particles were more greatly absorbed in the

denser matrix. The decreased active depth reduced the volume and therefore, activity,

exposed to the detector. The absolute detection efficiency did not have a declining

relationship with the solution density because the reduced activities accounted for the
241
density differences. The absolute detection efficiencies for the Am solutions of

various densities were fairly constant as measured with the ZnS(Ag) detector, Figure 5.8.

The direct detection of a solution with the ZnS(Ag) detector provided a straight-

forward relationship between the detector and the chemistry of the solution. Components

of liquid waste, such as salts and organics, would be treated as a single entity as they
 70

effect detection through changes in density. Insoluble particles which float on the top of

the waste would have a negative impact on the method of direct detection along with

non-homogeneous mixtures. Although density effects could be correct for, it is often an

unknown parameter as the density of waste varies. A 30% decrease in count rates existed
241
between the 1.01 g/mL and 1.27 g/mL Am solutions. The decrease in the count rate

could be minimized by using an average density of HLW; however, uncertainty

pertaining to density could not be abolished.

1
Normalized Count Rate/

y = -0.5559x + 1.5544
0.95 R2 = 0.9772
Concentration

0.9

0.85

0.8
1 1.1 1.2 1.3
Solution Density (g/ml)

Figure 5.7 The normalized count rate over concentration of the ZnS(Ag) detector versus
the density of the 241Am solution.
 71

10

Absolute Detection Efficiency

8

6 y = 2.3255x + 5.8093
R2 = 0.9114
(%) 4

0
1 1.1 1.2 1.3
Solution Density (g/ml)

Figure 5.8 The absolute detection efficiency of the ZnS(Ag) detector plotted against the
increasing density of the 241Am solution.

Air Gap Separation

The air gap between the ZnS(Ag) detector in conjunction with the signal splitter

and the liquid surface was varied to determine the optimum distance. The distance
241 230
between the detector and the surface of the Am and Th solutions exhibited an

exponential relationship with the absolute detection efficiency, Figure 5.9. The

monotonic decrease was expected given the variations in solid angle have no bearing on

the calculation of the absolute detection efficiency and the absorber thickness of the air

increased. The active depth of the solution was not calculated as a function of the air

thickness and remained constant with increasing distance. Physically, however, the

maximum depth of the alpha particles in solution decreased with increasing absorption

thickness which resulted in less activity from the 4.48 cm2 sample holding cup.
 72

Absolute Detection Efficiency

16
241
Am
y = 14.403e-0.0983x
12
R2 = 0.9968
230
Th

(%)
8
y = 14.211e-0.1608x
R2 = 0.9981
4

0
0 5 10 15 20 25 30
Separation Distance (mm)

Figure 5.9 The absolute detection efficiency decreased as the air gap distance between
the ZnS(Ag) detector and 241Am and 230Th solutions increased.

Contact between the Mylar® layer and aqueous solution would result in the

highest absolute detection efficiency possible, however, due to the caustic properties of

HLW the absence of an air gap was not practical. An air gap distance of 1.2 mm between

the Mylar® surface and the 241

Am and 230
Th solutions resulted in absolute detection

efficiencies of 12.94 + 0.06% and 10.95 + 0.04%, respectively, and preserved the

integrity of the Mylar® surface. The PIPS CAM detector demonstrated a similar

exponential relationship of count rate as a function of increased air gap distance, Figure

2.18.

Thin versus Thick Scintillation Layer

The ZnS(Ag) detector lacked total alpha energy deposition due to the thin

scintillation layer compared to total energy deposition in the PIPS CAM detector. The

scintillation layer present on the ZnS(Ag) detector did not theoretically absorb the total
239
alpha energy incident on the detector. The maximum range of a Pu alpha particle

through ZnS(Ag) scintillation crystals was estimated as 8.56 mg/cm2 with equations 1, 2,
 73

and 3. The estimated range was greater than the thin scintillation layer of the ZnS(Ag)

detector, 3.5 mg/cm2, however, shorter than a thicker ZnS(Ag) scintillation layer, 14

mg/cm2. The thin detection layer of the ZnS(Ag) detector represented a ∆E detector as

opposed to an E detector, with the remaining energy of the alpha particle deposited on the

light pipe directly behind the ZnS(Ag) layer. The detector effectively detected every

incident alpha particle but did not properly quantify the amount of energy that the alpha

particle possessed. The maximum range of a 239Pu alpha particle in a PIPS CAM detector

was calculated to be 27.5 µm—much less than the 120 µm depletion depth. The PIPS

CAM detector possessed an infinitely thick detection layer relative to the range of an

alpha particle in the media allowing the detector to quantify the total alpha energy

deposited. The range calculations did not incorporate energy absorption in the air,

Mylar®, or aluminized polymer layers.

The absolute detection efficiencies for the thick ZnS(Ag) scintillator layer were

compared to the thin ZnS(Ag) layer in the detection of alpha radioactivity in aqueous

solutions. The absolute detection efficiencies of the 244Cm, 239Pu, and 234U/238U solutions

detected using the two ZnS(Ag) scintillation layers were shown to be equivalent, Figure

5.10. The absolute detection efficiencies were expected to be the same for the two

scintillation layers as the solid angle was maintained. Although the thin scintillation

layer was a ΔE detector and did not quantify the total energy of the alpha particle, the

ZnS(Ag) detector was able to detect each particle which was incident on the detection

surface.
 74

Absolute Detection Efficiency

10
Thick Scintillation Layer
8 y = 1.9953x - 3.9304
R2 = 0.9692
6

(%) 4 Thin Scintillation Layer

y = 2.2597x - 5.2793
2 R2 = 0.9757

0
4 4.5 5 5.5 6
Alpha Energy (MeV)

Figure 5.10 The absolute detection efficiency of the ZnS(Ag) for the thick, 14 mg/cm2,
and thin, 3.5 mg/cm2, ZnS(Ag) scintillation layers for aqueous solutions.
239
The pulse height spectra of a Pu electroplated standard obtained with the thin

and thick ZnS(Ag) scintillation layers were dissimilar. The ZnS(Ag) detector with a 14

mg/cm2 scintillation layer resulted in a pulse height spectrum located in the lower

energies and with greater peak resolution than the 3.5 mg/cm2 scintillation layer

produced, Figure 5.11. The peak maxima of the pulse height spectra were located at 184

and 230 channels for thick and thin scintillation layers, respectively. The peak energy

resolutions for the pulse height spectra of the electroplated standard for the thick and thin

layers were 63 and 118%, respectively.

 75

Net Count Rate (cps)/ Activity (Bq)

0.0025

0.002 Thick

0.0015

Thin
0.001

0.0005

0
0 100 200 300 400 500
Channel Number

Figure 5.11 The pulse height spectra for the 239Pu blend electroplated standard detected
with the thick, 14 mg/cm2, and thin, 3.5 mg/cm2, ZnS(Ag) scintillation layers.

The pulse height spectra of 244Cm, 239Pu, and 234U/238U aqueous solutions detected

with the thicker scintillation layer on the ZnS(Ag) detector in conjunction with the

charge-sensitive preamplifier had a different spectral shape and were located at lower

channels then the spectra recorded with the thin scintillation layer. The leading edges of

the pulse height spectra (FWTM) for the thick ZnS(Ag) scintillation layer, Figure 5.12,
244 239 234
were 288, 246, and 211 channels for Cm, Pu, and U/238U, respectively. The

leading edges of the thin ZnS(Ag) scintillation layer, Figure 5.2(b), for the 244Cm, 239
Pu,
234
and U/238U aqueous solutions were 400, 363, and 330 channels, respectively. The

leading edges of the pulse height spectra were located at lower channels and had more

channels between the leading edges of the various solutions for the thicker scintillation

layer. An increase in the energy dependence was expected for the thicker scintillation

layers since all the alpha energy was deposited in the detector. The spectral shape of the

pulse height spectra of the thick scintillation layer for the aqueous solutions, Figure 5.12,

demonstrated better energy quantification as it had defined leading edges and closer
 76

resembled the continuous energy distribution than that of the thin scintillation layer,

Figure 5.2(b).

Net Count Rate (cps)/ Effective

0.0005
244
Cm
Activity (Bq) 0.0004
239
Pu
0.0003
234
U/238U
0.0002

0.0001

0
0 100 200 300 400 500
Channel Number

Figure 5.12 The pulse height spectra for the thick, 14 mg/cm2, scintillation layer on the
ZnS(Ag) detector.

A light collection phenomenon was responsible for the location of the pulse

height spectra from the thicker ZnS(Ag) scintillation layer in the lower channel region in

comparison to the thin scintillation layer. The ZnS(Ag) crystals produced light intensity

proportional to the incident alpha energy, however, the light was absorbed and scattered

by the ZnS(Ag) crystals. The intensity of the light measured with the photomultiplier

tube was decreased after scattering and absorption. The absorption of the scintillated

light by the ZnS(Ag) was represented by the extinction coefficient. The peak emission

wavelength of ZnS(Ag), 450 nm, or a photon energy of 2.76 eV corresponded to an

estimated extinction coefficient of 0.025 from Figure 2.5. The intensity of the light,

equation 9, transmitted by the ZnS(Ag) crystals was 94.2% for the 3.5 mg/cm2

scintillation layer. The thicker scintillation layer, 14 mg/cm2, transmitted 78.7% of the

original light intensity created by the ZnS(Ag) crystal, significantly less than the thin
 77

layer. The maximum recommended thickness of ZnS(Ag), 25 mg/cm2, only transmitted

65.3% of the scintillated light.

Computer Simulation

A computer simulation was created in Visual Basic to demonstrate the finite

thickness of the 3.5 mg/cm2 ZnS(Ag) scintillation layer and the infinitely thick detection

layer of the PIPS CAM detector. The Monte Carlo computer model simulated the

experimental set-up, but the sample holding cup and detector face were approximated as

a square with the same surface area as the round apparatus. The radionuclides were

randomly distributed throughout the volume in the sample holding cup and the alpha

particle directions were randomly generated. The active volume of the solution simulated

an infinitely thick source by using the maximum range of the alpha particle in solution

according to equations 1, 2, and 3. The computer code differentiated between the

ZnS(Ag) detector and the PIPS CAM detector in physical variables for the differences in

detector thickness and protective coatings over the detection surface. The ZnS(Ag)

detector had 3.5 and 14 mg/cm2 detection layers while the PIPS CAM had an infinitely

thick detection depth. The ZnS(Ag) detector had a 0.8 mg/cm2 Mylar® layer while the

PIPS CAM detector possessed a 0.5 µm aluminum and varnish layer with a total

absorption equivalent to 1.0 µm of silicon, 0.37 mg/cm2, for the PIPS CAM detector.

The protective layers were physical barriers which decreased the alpha particle energy

while protecting the detection surface and excluding light from the apparatus.

Differences between the scintillation and solid state detection properties for the detectors

were not incorporated into the program.

 78

The Visual Basic computer model was created to simulate the loss of alpha

energy in various absorption layers and to display the spectral shape of the theoretical

pulse height spectra. A combination of the Bethe equation, equation 5, and extrapolation

from experimental data [31] were used to calculate the energy loss of the alpha particle

for each layer of the detection system: solution, air, Mylar®, and ZnS(Ag). The alpha

particle energy continuously decreased as it transmitted through the absorption layers.

The stopping power, -dE/dx, was a function of the alpha velocity which was dependent

on alpha energy for the Bethe equation. The experimental extrapolation for stopping

power in water was dependent only upon the alpha energy. The standard-known alpha

energy was initially used to calculate the initial velocity of the alpha particle. A Visual

Basic loop was created in order to calculate the stopping power for each section of the

absorption layers. Each absorption layer was divided into 100 sections and the energy lost

by the alpha particle was calculated over the dx of each absorption layer as shown in the

flowchart, Figure 5.13. The alpha particle velocity was then computed for a given dx

step and input back into the source code. The energy of the alpha particle as it left each

absorption layer was returned to the worksheet. A histogram of the difference between

the alpha energy entering and exiting the ZnS(Ag) layer produced the pulse height

spectrum for the detection system. The PIPS CAM detection layer was programmed to

absorb the total alpha energy. The program processed 19992 Monte Carlo trials for each

run. A more comprehensive explanation of the computer simulation is detailed in

Appendix B.
 79

Figure 5.13 The flow chart of the Visual Basic computer simulation of the alpha energy
deposition in various absorption layers.
 80

The Monte Carlo computer model was benchmarked against published data and

shown to be in good agreement for high alpha energies. The computer-simulated

stopping power of water for alpha particles was plotted in Figure 5.4 from 0.15 to 200

MeV using the Bethe equation. The stopping power below 0.15 MeV was based on the

experimental data of Ziegler from the NIST ASTAR database [31]. Table 5.2 displays

the stopping power of water for higher-energy alpha particles from the computer-

simulated data and values obtained from the NIST ASTAR database [31]. The NIST

ASTAR values were derived using the Bethe equation for alpha values above 2 MeV and

were fit to Ziegler’s experimental data below 2 MeV [31]. The percent difference

between the computer-simulated and NIST ASTAR stopping power of water for alpha

particles was less than 7.86% for alpha energies between 1 and 6 MeV. The computer-

simulated stopping power values for low-energy alpha particles were not benched marked

against the NIST ASTAR database as the data were interconnected. The experimental

data was incorporated into the computer simulation as the stopping power values for

alpha energies below 0.11 MeV, as found with the Bethe equation, were negative as

shown in Figure 5.14.

 81

1.00E+04

Stopping Power in Water

Ziegler
Bethe
1.00E+03 Experimental
Equation

(Mev/cm)
1.00E+02

1.00E+01 Low Energy

Bethe Equation

1.00E+00
0.001 0.1 10 1000
Alpha Energy (MeV)

Figure 5.14 The computer-simulated stopping power of water for alpha particles.

Table 5.2 A comparison of the computer-simulated stopping power of water for alpha
particles to theoretical values obtained from the NIST ASTAR database [31].

Alpha Energy Computer-simulated Theoretical

(MeV) Stopping Power of Water Stopping Power of Water
(MeV/cm) (MeV/cm)
Figure 5.14 [31]
1.0 2.38 x 103 2.193 x 103
4.0 9.73 x 102 1.035 x 103
5.0 8.27 x 102 8.855 x 102
6.0 7.22 x 102 7.777 x 102
200 4.13 x 101 4.952 x 101

The theoretical pulse height spectra for the infinitely thick detection layer of the

PIPS CAM detector and the finite scintillation thickness of the ZnS(Ag) detector were

simulated for a point radioactive source with the computer modeling program. The

theoretical alpha energy deposited in the 3.5 mg/cm2 ZnS(Ag) layer and an infinitely
239
thick PIPS CAM detector are shown in Figure 5.15 for a Pu point source. The pulse

height spectrum for the PIPS CAM detector was a graph of the number of simulated

alpha particles incident to the detection surface divided by the total counts for each
 82

energy bin, Figure 5.15. The leading edge, 4.4 MeV, of the simulated pulse height

spectrum for the PIPS CAM detector had a lower energy than the 5.147 MeV 239Pu alpha
239
particles. The Pu alpha particles had a reduced energy and demonstrated tailing after

the particle passed through 0.5 cm of air and the CAM protective coatings. The
239
simulated pulse height spectrum for the Pu electroplated standard, Figure 5.15,

demonstrated the finite detection ability of the 3.5 mg/cm2 ZnS(Ag) detector as the bulk

of the energy deposited in the scintillation layer was between 2.4 and 3.2 MeV. The

leading edge of the pulse height spectrum was not located at the maximum alpha energy

calculated to reach the detection layer after absorption through the air and Mylar® layer,

4.0 MeV. The alpha energy was not totally absorbed in the ZnS(Ag) layer which allowed

for 4.0 MeV alpha particles to deposit only a fraction of their total energy. The

maximum alpha energy at the detection surface of the ZnS(Ag) detector was lower than

that of the PIPS CAM detector as the Mylar®, 0.8 mg/cm2, was thicker than the aluminum

and varnish coatings, 0.37 mg/cm2.

Incident Counts/ Total Counts

0.05

0.04

0.03
Thin ZnS(Ag)

0.02 PIPS CAM

0.01

0
0 1 2 3 4 5
Alpha Energy (MeV)

Figure 5.15 The computer-simulated pulse height spectra for the ZnS(Ag) finite
scintillation layer and the infinitely thick detection surface of the PIPS CAM detector for
a 239Pu point source.
 83

The pulse height spectra for the infinitely thick detection layer of the PIPS CAM

detector and thin scintillation thickness of the ZnS(Ag) detector were simulated for 239Pu

aqueous solution with the computer model. The computer simulation of the PIPS CAM
239
pulse height spectrum for the aqueous Pu solution was shown in Figure 5.16 and was

consistent with the theoretical continuous energy distribution present at infinitely thick

sources, Figure 2.19. The leading edge of the energy continuum represented the

maximum alpha energy which entered the detection surface—4.4 MeV, a value which
239
corresponded to the leading edge of the simulated pulse height spectrum for the Pu

point source in Figure 5.15. The pulse height spectrum had higher probability of

detection at high energies as most of the detected particles originated at the surface or

directly underneath the liquid surface which has a larger solid angle. The computer-

simulated pulse height spectrum for the 3.5 mg/cm2 layer of ZnS(Ag), Figure 5.16,

showed a narrow peak for the aqueous 239Pu solution. High-energy alpha particles passed

through the ZnS(Ag) layer without losing all of their energy, thus creating a large ΔE

peak. The large, narrowed peak was a summation of alpha particles which deposited

energies greater than 2.4 MeV to the ZnS(Ag) layer. Because the detector was

functioning as a ΔE detector, the amount of energy an alpha particle deposited into a thin

layer of ZnS(Ag) did not have a linear relationship to alpha energy. The exact value of

energy deposited in material was dependent upon the velocity of the alpha particle. The

lack of linearity was demonstrated in the Bragg Curve, the average specific ionization,

Figure 2.2. The stopping power of the water increased with decreasing alpha energy as

shown with the Bethe equation, Figure 5.14. A 4.0 MeV alpha particle, the maximum

energy alpha particle after absorption from air and Mylar®, deposited 2.0 MeV into a 3.5
 84

mg/cm2 ZnS(Ag) layer while a 3.0 MeV alpha particle deposited 2.3 MeV in the ZnS(Ag)

layer as calculated with the Bethe equation. The pulse height spectrum of the 3.5 mg/cm2

ZnS(Ag) detector of Figure 5.16 portrayed the total deposition of low-energy alpha

particles in the ZnS(Ag) layer. The low-energy alpha particles, a property of the

continuous energy distribution from aqueous solutions, have their total energy deposited

in the ZnS(Ag) layer. The pulse height spectrum for the finite detection layer converged

with the low-energy tail of the pulse height spectrum of the PIPS CAM detector in Figure

5.16 as it represented complete detection.

Incident Counts/ Total Counts

0.01

0.008

0.006
Thin ZnS(Ag)

0.004 PIPS CAM

0.002

0
0 1 2 3 4 5
Alpha Energy (MeV)

Figure 5.16 The computer-simulated pulse height spectra for the ZnS(Ag) finite
scintillation layer and the infinitely thick detection surface of the PIPS CAM detector for
a 239Pu aqueous solution.

The computer-simulated pulse height spectra for the PIPS CAM detector of the

point source and aqueous solution, Figures 5.15 and 5.16, were in good agreement with

the experimental spectral shape of Figure 5.1(a). The computer-simulated and

experimental data have similar pulse height spectra for the 239Pu theoretical point source
241
and Am electroplated standard, respectively, as both demonstrated a defined leading
239
edge and tailing. The computer-simulated pulse height spectra of the Pu aqueous
 85

solution detected with the PIPS CAM detector, Figure 5.16, demonstrated a continuous
241
energy distribution also present in the experimental pulse height spectrum of the Am

aqueous solution with PIPS CAM detector, Figure 5.1(a).

The computer-simulated and experimental pulse height spectra for the ZnS(Ag)

detector did not demonstrate the same spectral shape because the scintillation and light

properties were not simulated in the computer model. The narrow peak of the computer-

simulated pulse height spectrum for the 239Pu point source, Figure 5.15, did not resemble

the broad alpha peak of the experimental pulse height spectrum of the 239Pu electroplated

standard detected with the ZnS(Ag) detector, Figure 5.1(b). The computer-simulated and
239
experimental pulse height spectra of the Pu aqueous solution with the ZnS(Ag)

detector as shown in Figure 5.16 and Figure 5.1(b), respectively, did not show similar

spectra shaping. The poor resolution of the ZnS(Ag) detector was created from

propagation of the scintillated light—a property absent in the PIPS CAM detector and not

simulated in the program.

239
The computer-simulated absolute detection efficiencies of the Pu aqueous

solution for the 3.5 and 14 mg/cm2 ZnS(Ag) detector and the PIPS CAM detector were

compared to the experimental data. The ZnS(Ag) and PIPS CAM detectors had

computer-simulated absolute detection efficiencies of 3.10 + 0.1 and 4.87 + 0.2%,

respectively, for the 0.6 – 5.5 MeV ROI in the computer-simulated pulse height spectra
239
of the Pu aqueous solution, Table 5.3. The thick and thin scintillation layers had the

same absolute detection efficiency, a result consistent to the experiments. The

differences in the protective layers of the equipment caused differences of detection

efficiency as the Mylar® on the ZnS(Ag) detector was thicker and absorbed greater alpha
 86

energy. The solid angles of the simulated detection set up were consistent between the

two detectors. The experimentally determined absolute detection efficiencies were 6.08

+ 0.02%, 6.07 + 0.02%, and 7.95 + 0.01% for the thin and thick ZnS(Ag) detectors and

the PIPS CAM detector, respectively. The computer-simulated values of the absolute

detection efficiency were lower than the experimental absolute detection efficiencies for

the ZnS(Ag) and PIPS CAM detectors.

The computer-simulated pulse height spectra of the aqueous solutions revealed a

high-count data point in the lower energy region. The probability of detection of the 0-

0.4 MeV ROI in Figure 5.16 was 2.3 + 0.1%, which was cropped out of the computer-

simulated PIPS CAM pulse height spectrum. The pulse height spectra in the low-energy

region of interest were not accessible in the experimental ZnS(Ag) data as the gain was

too low. The low-energy region of the PIPS CAM spectrum was difficult to examine in

the experimental data due to large noise peaks; however, the excess of counts in the low-

energy region of the computer-simulated spectrum is inconsistent with the theory of

continuous distribution. The low-energy alpha particles were attributed to

inconsistencies of the Bethe equation and Ziegler extrapolation as seen in Figure 5.14 and

also to unknown imperfections in the computer model. The computer-simulated absolute

detection efficiency for the 0- 5.5 MeV ROI for the ZnS(Ag) detector and the PIPS CAM

detector were 3.85 + 0.1 and 7.22 + 0.2%, respectively, and are more consistent with the

experimental values. The differences between the absolute detection efficiencies of the

PIPS CAM detector and ZnS(Ag) detector were expected due to the differences in the

thickness protective layers as demonstrated by Phoenix et al., in Figure 2.11 [15]. The

thinner protective layer over a detector had higher count rates in the study.
 87

Table 5.3 The absolute detection efficiencies (%) for 239Pu aqueous solutions for the
ZnS(Ag) detector and PIPS CAM detector experiments and computer simulations.

Detector Experimental Simulated Absolute Simulated Absolute

Absolute Detection Detection Efficiency Detection Efficiency
Efficiency (ROI 0.6 -5.5 MeV) (ROI 0- 5.5 MeV)
Thin ZnS(Ag) 6.08 + 0.02 3.10 + 0.1 3.85 + 0.1
Thick ZnS(Ag) 6.07 + 0.02 3.10 + 0.1 3.85 + 0.1
PIPS CAM 7.95 + 0.01 4.87 + 0.2 7.22 + 0.2

A channel shift to the higher energies occurred when the ZnS(Ag) scintillation

thicknesses were increased in the computer-simulated pulse height spectra . The pulse
239
height spectra for the Pu point source were simulated using both the 3.5 mg/cm2 and

the 14 mg/cm2 ZnS(Ag) scintillation layers, Figure 5.17(a). In the computer-simulated

spectra for the point source, the pulse height spectrum for the thicker scintillation layer

was located at higher energies than the pulse height spectrum for the thinner layer. The

thicker scintillation layer was capable of greater alpha energy deposition than the thinner

scintillation layer and was able to register higher alpha energies. The peak resolution

improved and the tailing from the air and Mylar® absorption layers was present in the

pulse height spectrum for the point source as detected with the thicker scintillation layer.
239
The simulated pulse height spectrum for the Pu aqueous solution, Figure 5.17(b),

illustrates the same shift in channels. The thicker scintillation layer, 14 mg/cm2, had a

higher energy leading edge and closer resembled the continuous energy distribution. The

leading edge of the pulse height spectra for the thicker scintillation layer was located at

3.6 MeV, lower then the maximum alpha energy, 4.0 MeV, which was incident to the

scintillation layer. The 14 mg/cm2 ZnS(Ag) detector was thicker than the 3.5 mg/cm2

scintillation layer and had better energy detection capabilities. The 14 mg/cm2 ZnS(Ag)

layer did not have infinite detection capabilities in the computer model, despite
 88

expectations based on the range-energy relationship. The lack of complete detection is

believed to lie within imperfections of the model. Alpha particles were transmitted

through the thicker scintillation layer on the ZnS(Ag) detector at 3.27%, a value

inconsistent with theory and believed to be an unknown function of the computer

239
simulation. The range-energy theory for alpha particles determined that no Pu alpha

particle should have enough energy to be transmitted through the 14 mg/cm2 ZnS(Ag)

scintillation layer. The computer-simulated pulse height spectra for the thin versus thick

scintillation layers were different from the experimental pulse height spectra, Figures

5.11 and 5.12, as the computer model did not simulate light propagation and collection

properties.
 89

Incident Counts/ Total Counts

0.04

0.03

Thin ZnS(Ag)
0.02
Thick ZnS(Ag)

0.01

0
0 1 2 3 4 5
Alpha Energy (MeV)

(a)
Incident Counts/ Total Counts

0.006

0.004 Thin ZnS(Ag)

Thick ZnS(Ag)

0.002

0
0 1 2 3 4 5
Alpha Energy (MeV)

(b)

Figure 5.17 The pulse height spectra for a (a) 239Pu point source and (b) 239Pu aqueous
solution simulated for the thin, 3.5 mg/cm2, and thick, 14 mg/cm2, ZnS(Ag) scintillation
layers.
 90

CHAPTER 6

CONCLUSIONS

The ZnS(Ag) detector and PIPS CAM detector can both effectively detect alpha

particles directly from the surface of aqueous solutions. The detectors are both applicable

of detection of alpha radioactivity in high-level waste as the minimum detectable

concentration for each is lower than the concentration expected in process streams. The

ZnS(Ag) detector and the PIPS CAM detector would both make successful on-line alpha

detectors, but under different conditions. The ZnS(Ag) detector had an advantage over

the PIPS CAM detector as beta and gamma interactions were successfully discriminated

against while still maintaining good alpha detection efficiency. The ZnS(Ag) detector

also had the advantage of having a larger detection surface which further increased the

detection efficiency and decreased the minimum detectable concentration. The PIPS

CAM detector proved to be the better detector in the absence of beta and gamma

radiation due to increased detection efficiency and decreased energy dependence.

The ZnS(Ag) and PIPS CAM detectors were affected by experimental systematic

errors in the energy dependence and density effects on detection efficiency. The absolute

detection efficiencies for the different alpha energies varied as much as 41% for the

ZnS(Ag) detector and 71% for the PIPS CAM detector in the presence of beta and

gamma radioactivity. The increased density of the solution, a physical property

independent of the detectors, resulted in a decrease in the absolute detection efficiency.

Density effects contributed an additional 30% variation in absolute detection efficiency

between the 1.01 and 1.27 g/mL solutions. The combination of the effects from absolute
 91

detection efficiency and density created total systematic errors of approximately 70% and

100% for the ZnS(Ag) and PIPS CAM detectors, respectively. The systematic error was

maximized by taking the most extreme values; however, steps could be taken to reduce

the error by calibrating the detection system. The average alpha energies and density of

HLW could be incorporated into the system to improve the accuracy of the detectors.

The pulse height spectra of the ZnS(Ag) detector were obtained for two different

detection thicknesses and compared to the PIPS CAM detector, computer simulations,

and theory. The poor resolution of the ZnS(Ag) detector severely limited the capabilities

of the detector to quantify alpha energy; however, the PIPS CAM detector had excellent

resolution leaving room for possibilities of energy de-convolution based on the leading

edge of the alpha pulse height spectrum. The thin scintillation layer produced a stronger

signal than the thick ZnS(Ag) layer as the light intensity was not diminished from

absorption and scattering in the ZnS(Ag) layer.

 92

APPENDICES
 93

Appendix A
High Voltage Optimization

The optimal high voltage was determined for the ZnS(Ag) detector and signal

splitter which maximized the signal and minimized the instrumental noise. Ludlum Inc.

recommended a voltage range between 500 V and 1200 V for the photomultiplier tube.

The figure of merit, efficiency squared divided by the background, optimized the

detection efficiency and background noise. The figure of merit of the ZnS(Ag) detector
241
was determined using an Am electroplated standard for voltages ranging between 700

V and 1000 V in Figure A.1. The range was selected after the oscilloscope was used to

observe the entire voltage range recommended by Ludlum Inc. The operating voltage

(850V) and the amplifier gain (1000) were chosen in order to decrease the background

noise and to maximize alpha detection when the signal splitter was in use. The high

voltage of 800V was selected as the operating voltage when the charge-sensitive

preamplifier was used and an Ortec amplifier gain of 20 was set.

Absolute Efficiency Squared divided

6
by Background Count Rate

0
650 700 750 800 850 900 950 1000
High Voltage (V)

Figure A.1 The figure of merit for the ZnS(Ag) detector using the signal splitter.
 94

Appendix B
Signal Splitter versus Charge-Sensitive Preamplifier

The signal splitter and charge-sensitive preamplifier were both used to split the

input and output of the electronic signal. The comparisons and differences between the

instruments and results were examined. Pulse height spectra taken with the signal splitter

and charge-sensitive preamplifier were compared to determine the more appropriate

instrument in the direct detection of alpha radioactivity from aqueous solutions.

Pulse Pile-up Background

Pulse pile-up is a phenomenon associated with high count rates when two pulses

are close enough to be received as one pulse by the analysis system. Tail pile-up and

peak pile-up are two types of pulse pile-up. Tail pile-up occurs when pulses are

superimposed upon the long duration tail or undershoot from the preceding pulse. Tail

pile-up creates artificially larger or smaller pulses as shown in Figure B.1. When the tail

is negative, a smaller pulse is created, while pile-up on a positive tail results in a larger

pulse. Detection systems affected by pulse pile-up records only one pulse instead of two,

resulting in a smaller total area in a given spectrum. Pulses, as shown in Figure B.2, are

combined into a pulse with amplitude equal to the summation of the two original pulses.

Tail and peak pile-up affects the 55Fe pulse height spectral peak shape as shown in Figure

B.3. The dash line represents a low-activity scenario where pile-up is absent and the dark

line signifies pile-up due to high-activity [6].

 95

Figure B.1 Tail pile-up occurs whenever a tail or undershoot from a previous pulse is
present. The effect of the pile-up is displayed on the pulse height spectra to the right as
the cross-hatched area [6].

Figure B.2 Two signal pulses which are close together form into a single distorted pulse
as peak pile-up. The dashed lines represent possible overlapping scenarios [6].
 96

Figure B.3 A pulse height spectra of 55Fe demonstrates the effects of peak and tail pile-
up. A low counting rate, where pile-up is negligible, is represented by the dashed line.
The solid line represents a high counting rate and displays effects due to both tail and
peak pile-up [6].

Detection Dead Time and Recovery Time Background

Radiation detectors often fail to properly process incident radioactivity during

large count rates due to the dead time or recovery time of the detector. Electronic pulses

are not formed during the dead time of the detector and do not reach their full magnitude

during the recovery time of the detector. The dead time of a counting system is the

minimum time which separates two events in a detector and allows the events to be

recorded separately. The recovery time of the detector is defined as the time interval

which the detector needs to return to its original state and to produce a pulse of the initial

amplitude. The pulses of a Geiger Mueller tube demonstrated dead time and recovery

time in Figure B.4 [6].

 97

Figure B.4 The electronic signal pulse shape which is signature for dead time in a
Geiger- Mueller counter during high count rates [6].

An event is lost by the detector if it occurs too rapidly in time after the preceding

event as dead time follows each event in the detector. The dead time is set by either the

detector or the associated electronics. Losses in a detector are significant when the

detection system undergoes high counting rates. Two models of dead time behavior are

the paralyzable and nonparalyzable responses [6]. If an event occurs during the dead

time of a paralyzable detector, the count is not recorded and the dead time is prolonged as

shown in Figure B.5. If an event occurs during the dead time of a nonparalyzable

detector, the count is not recorded and the dead time is not prolonged. The top figure in

Figure B.5 demonstrates a paralyzable detector where only three counts are recorded for

the six events. The bottom figure represents a nonparalyzable detector which is able to

record four counts as the dead time is not extended. Although a nonparalyzable detector

is not able to record a count rate equal to the true event rate, the detector detects an

increasing count rate when the interaction rate is increasing. The paralyzable detector is

only capable of detecting a rising interaction rate to the point where the count rate levels
 98

off and then begins to decline, Figure B.6. The dashed line represents when the recorded

count rate, m, is equivalent to the true interaction rate, n, in the ideal detector [6].

Figure B.5 The recorded events in a detector, paralyzable versus nonparalyzable models
[6].

Figure B.6 The true count rate versus the observed count rate for nonparalyzable and
paralyzable detectors [6].

Errors remain high in the counting system even with attempts to correct for dead

time. The value of τ varies and the detector does not necessarily follow the

nonparalyzable or paralyzable model. Dead time losses greater than 30 or 40% have a

large effect on the count rate and a different detection system is recommended [6].
 99

Instrumental Components

The sole contents of the signal splitter were a resistor and capacitor. The circuit,

shown in Figure B.7, was responsible for redirecting the input and output currents. The

signal splitter lacked amplifying capabilities and simply passed the signal to the

amplifier. The (1.5 + 0.3) x 10-3 µF capacitor, used as a filter, allowed alternating current

from the detector to pass into the amplifier; however, it blocked the direct current from

the high voltage supply and detector. The ability of the filter to block DC circuit was due

to the capacitive reactance of the capacitor. The capacitive reactance was the measure of

the opposition or resistance of the capacitor to current. The capacitive reactance naturally

increased at low frequency and decreased for high frequency signals. The steady signal

from DC current had zero frequency which resulted in an infinite capacitive reactance or

total opposition [32].

106 Ω

1.5 µF

Figure B.7 The schematic diagram of the signal splitter.

The Canberra preamplifier Model 2006E was a charge-sensitive preamplifier

which shaped and amplified the output signal from the detector before it was further

amplified by the amplifier. The charge-sensitive preamplifier had several components

including a DC filter, integrator, P/Z differentiator, and buffer as shown in Figure B.8.

The DC filter, a 1.0 x 10-3 µF capacitor, separated the DC high voltage and detector from

the rest of the circuit and signal output. The operational integrator integrated the
 100

incoming charge pulse. The potential difference across the feedback capacitor was

proportional to the accumulated charge from the detector input. The differentiator, a

pole/zero (p/z) cancellation circuit, was responsible for providing a 50 µs tail pulse and

the optimum overload performance. The buffer applied a high voltage gain to the signal

to allow the signal to pass through a long cable without significant pulse degradation. A

gain setting favorable to high signal gain and low noise was applied to the electronic

pulse signal [33].

DC filter

Figure B.8 The schematic diagram of the Canberra preamplifier Model 2006 [33].

The charge-sensitive preamplifier provided gain to the detector signal which

supplied a stronger electronic pulse output than the signal splitter and reduced the gain

needed from the amplifier. The amplifier was capable of amplifying the signal; however,

it also amplified the instrumental noise. In order for the electronic pulse from the signal

splitter to be strong enough to be used by the multi-channel analyzer, the amplifier

needed to boost both the signal and the noise by a factor of 1,000, the maximum setting.

In the presence of the charge-sensitive preamplifier, the instrumental noise levels

remained small as the amplifier was set to 20, the minimum setting.
 101

Experimental Results

The pulse height spectra obtained with the ZnS(Ag) detector connected to the

signal splitter were different from the spectra obtained when the detector was connected

to the charge-sensitive preamplifier while the detection capabilities varied at high count

rates. The absolute detection efficiency of the detector in conjunction with both the

signal splitter and charge-sensitive preamplifier remained consistent for low count rates;

however, characteristics in the pulse height spectra differed. Figure B.9 depicted the

count rates from the ZnS(Ag) detector in conjunction with the charge-sensitive

preamplifier and the signal splitter. The count rates were equivalent at low count rates

and began to differ when high-activity electroplated standards were detected. The

aqueous solutions possessed low activities and were primarily consistent. Despite

similarities in detection, the pulse height spectra taken with the ZnS(Ag) detector and

signal splitter were different from those taken with the ZnS(Ag) detector in conjunction

with the charge-sensitive preamplifier. The differences in pulse height spectra were

present regardless of count rate. The pulse height spectra obtained with the ZnS(Ag)

detector and signal splitter displayed two distinctive properties. The pulse height spectra

possessed overall good peak energy resolution, however, experienced decreased energy

resolution when the spectra shifted to the low-energy region of interest in the presence of

high alpha or beta activity.

 102

Preamplifier Count Rate (cps)

3000 y=x

2000

1000 y = -6E-05x 2 + 1.0231x - 3.2189

R2 = 1

0
0 1000 2000 3000
Signal Splitter Count Rate (cps)

Figure B.9 The count rate of the ZnS(Ag) detector in conjunction with both the charge-
sensitive preamplifier and the signal splitter for the detection of several electroplated
standards.

The pulse height spectra for a 239Pu solution and 239Pu electroplated standard, both

at a 5.2 mm air gap distance and detected with the 14 mg/cm2 ZnS(Ag) detector in

conjunction with the signal splitter, demonstrated similar pulse height spectra in Figure
239
B.10 despite the differences in source medium. The pulse height spectrum for the Pu

aqueous solution was only slightly broader than the pulse height spectra of the

corresponding electroplated standard. The pulse height spectrum obtained with the signal

splitter appeared to offer better energy resolution; however, the pulse height spectrum

lacked characteristics commonly associated with alpha particles. The ZnS(Ag) detector

and signal splitter pulse height spectrum for the electroplated standard in Figure 5.10 did

not resemble the computer-simulated pulse height spectrum for a point source in Figure

5.17(a) as tailing was not present. The pulse height spectrum for the aqueous solution

taken with the ZnS(Ag) detector and signal splitter, Figure B.10, was not consistent with

the theoretical pulse height spectrum for the thick scintillation layer as simulated in
 103

Figure 5.17(b). The experimental pulse height spectrum lacked the continuous energy

distribution of the aqueous solution. The pulse height spectrum did not resemble the

finite scintillation layer in Figure 5.17(b) as the low-energy tail was not present.

ZnS(Ag) was reported to scintillate for alpha energies as low as 0.1 MeV in experiments

conducted at Saint Gobain Crystals, Inc. which suggested the expectation of the low

alpha energy tail [34].

0.002 0.01

Activity (Bq) for 239Pu EPS

Effective Activity (Bq) for

0.0016 0.008
Net Count Rate (cps)/

Net Count Rate (cps)/

Pu Solution

0.0012 EPS 0.006

0.0008 0.004
239

0.0004
Liquid 0.002

0 0
0 100 200 300 400 500
Channel Number

Figure B.10 The pulse height spectrum of an aqueous 239Pu solution was compared to
the shape of the 239Pu electroplated spectrum using the ZnS(Ag) detector and signal
splitter.
244 239
The pulse height spectra for several aqueous solutions, Cm, Pu, and
234
U/238U, taken with the 14 mg/cm2 ZnS(Ag) detector and signal splitter demonstrated a

narrow peak resolution in Figure B.11. The pulse height spectra obtained while using the

signal splitter did not reveal the broad alpha peaks present in Figure 5.12, the pulse height

spectra obtained while using the charge-sensitive preamplifier. The leading edges of the

pulse height spectra obtained with the ZnS(Ag) detector and signal splitter were separated
244 234
by 17 channels for the 5.8 MeV Cm solution and a 4.8 MeV U/238U solution. The
 104

leading edges of the pulse height spectra obtained with the ZnS(Ag) detector and charge-

sensitive preamplifier in Figure 5.12 were separated by 77 channels for the 5.8 MeV
244
Cm solution and 4.8 MeV 234U/238U solution. The differences in the channel numbers

of the various solutions demonstrated that the ZnS(Ag) detector and signal splitter had a

small, non-linear energy dependence.

Net Count Rate (cps)/ Effective

0.002

0.0016
Activity (Bq)

244
0.0012 Cm

239
0.0008 Pu

234
0.0004 U/238U

0
0 100 200 300 400 500
Channel Number

Figure B.11 The pulse height spectra for 244Cm, 239Pu, and 234U/238U aqueous solutions
using the thick scintillation layer on the ZnS(Ag) detector and the signal splitter.

The charge-sensitive preamplifier and the signal splitter were both utilized to

separate the signal for the ZnS(Ag) detector from the high voltage input. The instruments

connected the high voltage input and the signal output to a single electrical port on the

ZnS(Ag) detector probe. The sole purpose of the signal splitter was to split the high

voltage and signal pulses, while the charge-sensitive preamplifier was a more

sophisticated device affecting the electric signal from the ZnS(Ag) detector. Although a

charge-sensitive preamplifier was not required for gross alpha detection of low count

rates, it proved necessary for accurate count rates and pulse height spectra. After
 105

contacting several electronic experts, the properties behind the instruments which caused

the variations in the energy dependence of the pulse height spectra remained unknown.

The pulse height spectra measured with the 3.5 mg/cm2 ZnS(Ag) detector and

signal splitter shifted the pulse height spectrum to the lower energies in the presence of

high activities, an effect not present with the charge-sensitive preamplifier. The

phenomenon was illustrated in Figure B.12 by the shift and broadening of the 241Am and
238
Pu pulse height spectra. The alpha energies for the 241Am electroplated standards were
238
5.479 MeV and Pu had a similar alpha energy of 5.487 MeV. The lowest activity

electroplated standard, 489.9 238Pu, had the best resolution, 9.2%, and the peak maximum

was located in the highest channel, 434. The 1644 Bq 241Am electroplated standard had a

peak resolution of 11.7% and the peak maximum channel number at 424. The highest
241
activity electroplated standard, 7151.7 Bq Am, had the poorest resolution, 28.7%, and

the peak maximum was located in the lowest channel, 365. The alpha peaks were

expected to be located at the same channel since the alpha energies were approximately

the same. A difference of 69 channels for the peak maxima of the pulse height spectra

for the lowest and highest activities indicated that the ZnS(Ag) detector in connection

with the signal splitter did not produce pulse height spectra properly at high count rates.

The pulse height spectra for the 241Am and 238Pu electroplated standards detected with the

charge-sensitive preamplifier did not show a shift in channels or peak broadening in

241
Figure 5.13. The greater peak maximum of the 7151.7 Bq Am pulse height spectrum
241
in Figure 5.13 was attributed to differences in solid angle. The 7151.7 Bq Am

electroplated surface was 0.5 mm closer to the detector than the lower activity 241Am and
238
Pu electroplated standards. At a separation distance of 1.5 mm, the difference in solid
 106

angle was significant. The pulse height spectra of the ZnS(Ag) detector and charge-

sensitive preamplifier were present in the same channels and had the same resolution, as

was expected since the alpha energies were nearly the same and the energy resolution

was poor.

0.01
Net Count Rate (cps)/ Activity

0.008
490 Bq 238Pu
0.006
1644 Bq 241Am
(Bq)

0.004
7152 Bq 241Am
0.002

0
0 100 200 300 400 500
Channel Number

Figure B.12 The pulse height spectra for 1644 Bq and 7151.7 Bq 241Am electroplated
standards and a 490 Bq 238Pu electroplated standard using a ZnS(Ag) detector and signal
splitter.

0.0016 7152 Bq 241Am

Net Count Rate (cps)/ Activity

1644 Bq 241Am
0.0012 490 Bq 238Pu
(Bq)

0.0008

0.0004

0
0 100 200 300 400 500
Channel Number

Figure B.13 The pulse height spectra for 490Bq 238Pu EPS, 1644 Bq 241Am EPS, and
7151.7 Bq 241Am EPS detected with a ZnS(Ag) detector and charge-sensitive
preamplifier.
 107

Recovery time in the electronic signal from the signal splitter was responsible for

the pulse height spectral shift at high count rates. Recovery time was absent in the

electronic pulse from the ZnS(Ag) detector, signal splitter, and amplifier in Figure
238
B.14(a) for the low activity, 489.9 Bq, Pu electroplated standard. The filter capacitor

inside the signal splitter had sufficient time to recover between electronic pulses. The

electronic pulse from the ZnS(Ag) detector, charge-sensitive preamplifier, and amplifier,
238
Figure B.14(b), also demonstrated a similar pulse for the 489.9 Bq Pu electroplated
241
standard which was free of recovery time. The electronic pulse of the 1644 Bq Am

electroplated standard detected by the detector, signal splitter, and amplifier exhibited

recovery time in Figure B.15(a). The electronic pulses to the right of the primary pulse

were smaller and amplified with time. The energy output overlapped with the previous

signal energy if too little time had lapsed. The collection of energy occurred at the

capacitor in the signal splitter. The count rate of the alpha particles, 615.6 + 1.4 cps, was

large enough for the capacitor in the signal splitter to stop functioning properly. The

electronic pulses entered the capacitor at a fast rate, not allowing for the capacitor to

discharge fully after each signal. Although the same electroplated standard was detected

by the detector, charge-sensitive preamplifier, and amplifier in Figure B.15(b), recovery

time was not present in the electronic pulses. Recovery time remained absent in the

electronic pulses of the highest activity 241Am electroplated source, 7151.7 Bq, detected

with the ZnS(Ag) detector, charge-sensitive preamplifier, and amplifier in Figure

B.16(b). The detector, signal splitter, and amplifier electronic signal for the detection of
241
the highest activity, 7151.7 Bq, Am electroplated source was observed in Figure

B.16(a). The electronic signal demonstrated an even more exaggerated recovery time
 108
241
pattern. The recovery time for the 7151.7 Bq Am electroplated source was 120 µs as

compared to only 80 µs for the 1644 Bq 241Am electroplated standard. The higher count

rates overwhelmed the capacitor inside the signal splitter and intensified the problem.

The recovery time did not paralyze the detector as the small electronic pulses were still
241
counted. The tail located on the 7151.7 Bq Am pulse height spectra in Figure B.12

represented the electronic signals which were reduced to lower energies due to the effects

of recovery time. The low-energy electronic pulses caused by the effects of recovery

time were responsible for the shift in the peak maxima of the 241Am pulse height spectra,

Figure B.12. While the ZnS(Ag) detector was connected to the charge-sensitive

preamplifier, recovery time was absent from the electronic signals which resulted in pulse

height spectra which did not shift at high count rates, Figure B.13.

(a) (1V, 20 µs) (b) (2V, 20 µs)

Figure B.14 The electronic signal from the ZnS(Ag) detector and (a) signal splitter and
(b) charge-sensitive preamplifier demonstrated the pulses created by a 489.9 Bq 238Pu
electroplated standard.
 109

Recovery Time

(a) (1V, 20 µs) (b) (2V, 100 µs)

Figure B.15 The electronic signal from the ZnS(Ag) detector and (a) signal splitter and
(b) charge-sensitive preamplifier demonstrated the pulse created by a 1644 Bq 241Am
electroplated standard.

Recovery Time

(a) (1V, 20 µs) (b) (2V, 20 µs)

Figure B.16 The electronic signal from the ZnS(Ag) detector and (a) signal splitter and
(b) charge-sensitive preamplifier demonstrated the pulse created by a 7151 Bq 241Am
source.

The absolute detection efficiency of the ZnS(Ag) detector in connection to the

signal splitter and charge-sensitive preamplifier were consistent for low count rates as

shown in Figure B.1 and Table B.3. The absolute detection efficiencies for the

instruments fell within uncertainty for the 489.9 Bq sample—35.6 + 0.5% and 34.8 +

0.5% for the signal splitter and charge-sensitive preamplifier, respectively. The absolute
 110

detection efficiencies of the ZnS(Ag) detector coupled to the signal splitter and charge-

sensitive preamplifier were 37.4 + 0.4% and 35.9 + 0.4% for the 1644 Bq sample,
238
respectively. The absolute detection efficiencies of the 489.9 Bq Pu electroplated
241
standard and 1644 Bq Am electroplated standard were expected to be similar for the

similar alpha energies, 5.487 MeV and 5.479 MeV, respectively. The recovery time

present in the electrical signal from the signal splitter did not have an effect on the

absolute detection efficiency. The absolute detection efficiencies of the charge-sensitive

238 241
preamplifier for the Pu and Am electroplated standards were close and just fell out

of agreement. The absolute detection efficiencies for the highest count rate detected by

the ZnS(Ag) detector in connection to the signal splitter and charge-sensitive preamplifier

were not consistent with the absolute detection efficiencies for the lower count rates due

to differences in solid angle. The absolute detection efficiency for the 7151.7 Bq 241Am

electroplated standard was high for the ZnS(Ag) detector and signal splitter, 41.3 + 0.5%,

for the 115-511 channel region of interest. The absolute detection efficiency for the

ZnS(Ag) detector in connection with the charge-sensitive preamplifier was artificially

large due to solid angle effects at 39.1 + 0.5% for a region of interest between 20 and 511

channels. The regions of interest varied between the pulse height spectra of the signal

splitter and charge-sensitive preamplifier due to differences in the location of the alpha

region of interest and the noise peak. The noise peak of Figure 5.11 was located in the

higher channels which indicated that a LLD at 115 was necessary for the signal splitter.

The noise peak of Figure 5.2 dictated a LLD of 20 was necessary. The absolute detection

efficiencies of the alpha electroplated standards were summarized in Table B.1.

 111

Table B.1 The absolute detection efficiencies of the ZnS(Ag) detector connected to both
the signal splitter and the charge-sensitive preamplifier.

Count Rate Count Rate Abs. Detection Abs. Detection

Electroplated Activity (cps) (cps) Efficiency (%) Efficiency (%)
Standard (Bq) Signal Splitter Preamplifier Signal Splitter Preamplifier
238
Pu 489.9 + 6.4 174.6 + 0.8 170.7 + 0.8 35.6 + 0.5 34.8 + 0.5
241
Am 1644 + 18.1 615.6 + 1.4 590.4 + 1.4 37.4 + 0.4 35.9 + 0.4
241
Am 7152 + 93.0 2957 + 3.1 2793.7 + 3.1 41.3 + 0.5 39.1 + 0.5

In the presence of high beta activity, the alpha pulse height spectrum shifted to

lower energies for the signal splitter while a shift did not occur for the charge-sensitive
90
preamplifier. The beta particle radiation field was created with a Sr(90Y) source at a
241
distance less than 10 cm from the ZnS(Ag) detector face. The 7151.7 Bq Am

electroplated standard was placed upon the 0.635 cm Plexiglas shield and platform
90
illustrated in Figure 4.4, above the Sr(90Y) source. The peak maximum of a pulse
241
height spectra for the 7151.7 Bq Am electroplated standard in the absence of the beta

radioactivity was located at channel 401 in Figure B.17, when the ZnS(Ag) detector was

connected to the signal splitter. The pulse height spectrum shifted to lower energies in

the presence of the high beta radioactivity which resulted in a peak maximum at channel

101 while the pulse height spectra of the ZnS(Ag) detector and charge-sensitive

preamplifier did not shift in Figure B.18. Although the instrument set up in Figure 4.4

was utilized, the solid angle between the electroplated standard and the ZnS(Ag) detector

connected to the signal splitter did not remain consistent with and without beta

radioactivity due to an increased distance between the detector face and the electroplated

standard. The pulse height spectra of the ZnS(Ag) detector in conjunction with the signal

splitter did not shift despite solid angle or changes in alpha energy in the absence of beta

and gamma radioactivity which permitted the spectra to be compared.

 112

0.01

7152 Bq 241Am on Plexiglas shield

Net Count Rate (cps)/ 0.008
Am Activity (Bq)
in the 90Sr(90Y) beta field

0.006

7152 Bq 241Am on
0.004 Plexiglas shield
241

0.002

0
0 100 200 300 400 500
Channel Number

Figure B.17 The pulse height spectra of the 7151.7 Bq 241Am EPS on top of the
Plexiglas shield with and without the 7.4 x 108 Bq 90Sr(90Y) source detected with the
ZnS(Ag) detector and signal splitter.

0.002
Net Count Rate (cps)/
Am Activity (Bq)

0.0016
7152 Bq 241Am on Plexiglas shield
0.0012 in the 90Sr(90Y) beta field

0.0008
241

0.0004 7152 Bq 241Am on

Plexiglas shield
0
0 100 200 300 400 500
Channel Number

Figure B.18 The pulse height spectra of the 7151.7 Bq 241Am EPS on top of the
Plexiglas shield with and without the 7.4 x 108 Bq 90Sr(90Y) source detected with the
ZnS(Ag) detector and charge-sensitive preamplifier.

Combined alpha and beta radiation demonstrated greater increased noise in the

electronic signals in the ZnS(Ag) detector and signal splitter than the charge-sensitive

preamplifier. The oscilloscope trace of the ZnS(Ag) detector in connection with the

signal splitter and amplifier, Figure B.19(a), and the ZnS(Ag) detector in connection with

the charge-sensitive preamplifier and amplifier, Figure B.19(b) revealed differences in

 113

the noise baseline. The baseline of the signal splitter oscilloscope trace, Figure B.19(a)

demonstrated an enlarged baseline greater than one volt due to noise and lower energy

deposition in the ZnS(Ag) detector from beta interactions. The electronic signal from the

noise and beta particles overwhelmed the capacitor inside of the signal splitter which

caused recovery time problems. The electronic signal from the detected alpha particles

did not properly charge the capacitor. The electronic pulses reduced the amplitude of the

pulses during combined alpha and beta detection. The decreased pulse amplitude and

therefore energy explained the shift of the pulse height spectrum of Figure B.17. The

beta signal interference was absent in the electronic signal of combined alpha and beta

radioactivity, Figure B.19(b), for the ZnS(Ag) detector and charge-sensitive preamplifier

detection system. The charge-sensitive preamplifier was able to maintain a flat baseline

in the presence of beta radioactivity. The absence of recovery time in the electronic

signal of the ZnS(Ag) detector in connection with the charge-sensitive preamplifier

explained the consistency of the location of the pulse height spectra with and without beta

interference in Figure B.18.

Recovery Time

(a) (1V, 20 µs) (b) (1V, 20 µs)

Figure B.19 The electronic signal for the ZnS(Ag) detector, amplifier, and (a) signal
splitter and (b) the preamplifier. The 7151.7 Bq 241Am electroplated standard was set on
top of the 0.635 cm Plexiglas sheet placed above a 7.4 x 108 Bq 90Sr(90Y) source.
 114

Appendix C
Visual Basic Computer Model

A computer model was designed in Microsoft Excel using Visual Basic to predict

the pulse height spectra for the detection of alpha particles from an infinitely-thick source

with a finite thickness detector. The infinitely thick radioactive source, the aqueous

solution in a sample holding cup, was simulated by randomly sampling a point (x,y,z)

inside of the theoretical box, Figure C.1. The cylindrical sample holding cup and circular

detection surface from the experimental set up were simplified to a square cup and square

detection surface with a surface area equivalent to that of the experimental apparatus.

The coordinate system was placed at one of the corners of the square at a distance, z1,

from the top of the liquid surface. The distance, z1, was selected as the active depth of

the liquid which determined the volume of liquid exposed to the detector. The active

depth, z1, was calculated using equations 1, 2, and 3 and was an important parameter in

the computer simulation. If the distance was too large, complete absorption for the

majority of the simulated points resulted in poor counting statistics for the pulse height

spectrum.
 115

Figure C.1 The moving spherical coordinate system used in the computer simulations.
(Relative thicknesses were not drawn to scale.)

All random numbers were generated using the Microsoft Excel random number

generator. The simulation consisted of 19,992 generated alpha particles. The randomly

generated origin of each alpha particle, (xo, yo, zo), represented a spherical moving

coordinate as shown in Figure C.1. The x and y components, the origin of the alpha

particle, were randomly generated between 0 and 2.04 cm. The z component was
 116

generated between 0 and the active depth, the maximum range the alpha particle traveled

in solution.

0 < x0 < 2.04 cm

0 < y0 < 2.04 cm (14)
0 < z0 < z1 (cm)

The direction of the alpha particle was randomly generated in all directions. Spherical

coordinates, θ and φ, were generated as

0 < θ < 2π
(15)
0 <ϕ <π

The spherical coordinate of r represented the distance the alpha particle traveled and was

a function of the vector z and φ. The standard spherical coordinate equations

x = r sin ϕ cos θ
y = r sin ϕ sin θ (16)
z = r cos ϕ

were utilized where x, y, z, and r were variables [35]. The vertical distance, z1-zo, and the

known φ were used to determine the total distance the alpha particle travels through the

water absorption layer, r1. The variables, r2, r3, and r4, were calculated for each

absorption layer. The distance of the vector, r1, between the surface of the liquid, z1, and

the original point, zo,

z1 − z0
r1 = (17)
cos ϕ

was found. The distance to the surface of the liquid from the origin of axes at the corner

of the box in Figure C.1 was always equal to the active depth, z1. The z components of

the box, the thickness of the air, Mylar®, and ZnS(Ag) layers were known. The

calculated r and randomly generated φ were then used to determine the x and y at the
 117

Mylar® surface using equation 16. For example, the distance, r2, from the original point,

zo, to the surface of Mylar®, z2, was determined when φ and θ were known as shown in

equation 17. The coordinates, x2 and y2, were the intersection points at the air-Mylar®

surface while xo and yo were the coordinates for the original randomly generated point,

equation 18. The edges of the sample holding cup collimated the alpha radiation onto the

surface detection area.

x2 = r2 sin ϕ cos θ + x0
(18)
y2 = r2 sin ϕ sin θ + y0

The directional alpha particle vectors were generated in 4π and a method to

eliminate the alpha particles which were never incident on the detection surface was

created. A series of “If Statements” were designed in Microsoft Excel using binary

systems. A zero was placed in the column if the statement was false and a one was input

into the spreadsheet if the statement was true. The “If Statements” defined the

parameters of the detection geometry. If a false, or zero, was present for a generated

point, the point was cancelled and was determined to never be incident with the detection

surface. A false was output if the randomly generated φ was above π/2 meaning that the

alpha particle was generated into the negative direction. The distance the alpha particle

traveled in water was compared to the maximum range of alpha particle in solution. If

the alpha particle traveled a distance that was greater then the maximum range, a false

was recorded. If the computed values of x2 + xo or y2 + yo were greater than 2.04 cm,

then the alpha particle left the box and a false was recorded. If x2 + xo or y2 + yo were

negative then the alpha particle left the box and never reached the theoretical detection

surface. The original coordinates, x and y, were added to the intersection values to take

into account the moving coordinate system.

 118

The simplified Hans Bethe equation,

⎡ 2 ⎤
⎢ ⎛v⎞ ⎥
C2 ⎜ ⎟ 2
dE C1n ⎢ ⎝ c⎠ ⎛v⎞ ⎥
− = ⎢ ln ⎛ − ⎜ ⎟ ⎥ MeV cm −1 (19)
dx ⎛ v ⎞2 2
⎞
⎢ I ⎜1 − ⎛ v ⎞ ⎟ ⎝ ⎠ ⎥
c
⎜ ⎟ ⎜
⎜ ⎝c⎠ ⎟ ⎟
⎝c⎠ ⎢ ⎝ ⎠ ⎥
⎣ ⎦

was used to simulate the energy lost by the alpha particles above 0.15 MeV in the various

materials. The extrapolated equation from experimental data conducted by Ziegler,

equation 20, was used to calculate the stopping power for alpha particle energies between

0.001 MeV and 0.15 MeV where E was in MeV [31]. Alpha energies less then 0.001

MeV were assumed to be 0 MeV.

dE
= −29096 E 2 + 11329 E + 320.86 (20)
dx

The constants in Table C.1 were called upon in the Microsoft Excel worksheet,

Figure C.2, by the Visual Basic source code in the assessment of equation 19. C1 and C2

were constants present in equation 8 while n and I were found using equations 6 and 7,

respectively. The value for the speed of light in vacuum, 3 x 108 m/s, was a known

constant.

Table C.1 Constants used in the computer simulation.

Constants HNO3 Air Mylar® ZnS(Ag) Al Si

Solution
C1 2.03x10-30
C2 1.02x106
n 3.00x1029 3.92x1026 4.08x1029 1.15x1030 7.84x1029 6.99x1029
I 92.6 98.1 69.2 258.0 163.3 174.7
c 3 x 108
 119

Figure C.2 The Microsoft Excel worksheet referred to by the Visual Basic source codes.

The velocity of the alpha particle was not a constant and was changing through

the simulation. The original velocity of the alpha particle was dependent upon the alpha

particle energy of the radionuclide. The particle energy was converted to joules from

MeV. The original velocity of the alpha particle was found using the non-relativistic

kinetic energy equation as the alpha mass was too large to be relativistic. The original

velocity was used as a starting point for equation 19. The alpha particle continuously lost

energy as it traveled through the material which resulted in a constantly decreasing

velocity. A loop was formed to recalculate the velocity of the alpha particle 100 times in

each absorption thickness as demonstrated in Source Code C.1. The loop was only

employed if the length of the distance traveled, r, was above zero. The length of the

alpha particle was zero if the point was cancelled with “If Statements,” and the alpha

particle was determined not to reach the detection surface.

 120

The code ran for each of the four absorber thicknesses: liquid, air, Mylar®, and

ZnS(Ag). Source Code C.1 modeled the loss of alpha particle energy in the solution, and

output the energy which the alpha particle had as it left the solution. The alpha particle

energy was then input into Source Code C.2 as the initial energy of the alpha particle as it

began to transport through air. Source Code C.2 simulated the loss of alpha energy as it

traveled through air. Source Code C.3 and Source Code C.4 modeled the Mylar® layer

and ZnS(Ag) layer, respectively. Source Code C.4 output the final energy of the alpha

particle as it exited the ZnS(Ag) layer into the Microsoft Excel worksheet. The change of

energy, energy deposited in the scintillation layer, was found by subtracting the final

energy of the particle after having exited the ZnS(Ag) layer from the energy of the alpha

particle after it left the Mylar® layer.

The histogram function in Microsoft Excel was used to determine the number of

counts in each energy bin. The efficiency or “incident counts over total counts” was

found by dividing the counts in each bin by the total number of counts. Using the

computer model, the pulse height spectra or counts per energy bin versus energy bin,

were simulated.
 121

Source Code C.1. The computer code for the simulation of the energy of the alpha
particle as it left the aqueous solution.

Sub Loopysolutionwater()
For tmp = 15 To 20006
L = Cells(tmp, 1)
EmeV = Cells(5, 1)
EJ = Cells(7, 1)
Cells(7, 3) = EJ
dL = L / 100
M = Cells(8, 1)
c = Cells(6, 4)
Cells(7, 4) = c
n = Cells(1, 3)
I = Cells(1, 6)
C1 = Cells(9, 1)
C2 = Cells(10, 1)

If L = 0 Then
Cells(tmp, 5) = EmeV
End If

If L > 0 Then
For Intervals = 1 To 100
If EmeV > 0.15 Then
v = Sqr(Abs(2 * EJ / M))
B = v / c 'unit-less
term1 = ((C1 * n) / (B ^ 2))
term2 = (C2 * B ^ 2) / (I * (1 - B ^ 2))
term3 = Log(term2) - B ^ 2
dEdx = term1 * term3
dE = dEdx * dL
EmeV = EmeV - dE
EJ = EmeV * 1000000 * 1.60217646E-19
ElseIf (0.001 < EmeV) And (EmeV <= 0.15) Then
dEdx = -29096 * EmeV ^ 2 + 11329 * EmeV + 320.86
dE = dEdx * dL
EmeV = EmeV - dE
ElseIf (EmeV <= 0.001) Then
EmeV = 0
End If
Next Intervals
Cells(tmp, 5) = EmeV
End If
Next tmp
End Sub
 122

Source Code C.2. The computer code for the simulation of the energy of the alpha
particle as it left the air absorber layer.

Sub Loopysolutionair()
For tmp = 15 To 20006
L = Cells(tmp, 2)
EmeV = Cells(tmp, 5)
EJ = EmeV * 1000000 * 1.60217646E-19
dL = L / 100
M = Cells(8, 1)
c = Cells(6, 4)
n = Cells(2, 3)
I = Cells(2, 6)
C1 = Cells(9, 1)
C2 = Cells(10, 1)

If L = 0 Then
Cells(tmp, 6) = EmeV
End If

If L > 0 Then
For Intervals = 1 To 100
If EmeV > 0.15 Then
v = Sqr(Abs(2 * EJ / M))
B = v / c 'unit-less
term1 = ((C1 * n) / (B ^ 2))
term2 = (C2 * B ^ 2) / (I * (1 - B ^ 2))
term3 = Log(term2) - B ^ 2
dEdx = term1 * term3
dE = dEdx * dL
EmeV = EmeV - dE
EJ = EmeV * 1000000 * 1.60217646E-19
ElseIf (0.001 < EmeV) And (EmeV <= 0.15) Then
dEdx = -29096 * EmeV ^ 2 + 11329 * EmeV + 320.86
dE = dEdx * dL
EmeV = EmeV - dE
ElseIf (EmeV <= 0.001) Then
EmeV = 0
End If
Next Intervals
Cells(tmp, 6) = EmeV
End If
Next tmp
End Sub
 123

Source Code C.3. The computer code for the simulation of the energy of the alpha
particle as it left the Mylar® absorber layer.

Sub LoopysolutionMylar()
For tmp = 15 To 20006
L = Cells(tmp, 3)
EmeV = Cells(tmp, 6)
EJ = EmeV * 1000000 * 1.60217646E-19
dL = L / 100
M = Cells(8, 1)
c = Cells(6, 4)
n = Cells(3, 3)
I = Cells(3, 6)
C1 = Cells(9, 1)
C2 = Cells(10, 1)

If L = 0 Then
Cells(tmp, 7) = EmeV
End If

If L > 0 Then
For Intervals = 1 To 100
If EmeV > 0.15 Then
v = Sqr(Abs(2 * EJ / M))
B = v / c 'unit-less
term1 = ((C1 * n) / (B ^ 2))
term2 = (C2 * B ^ 2) / (I * (1 - B ^ 2))
term3 = Log(term2) - B ^ 2
dEdx = term1 * term3
dE = dEdx * dL
EmeV = EmeV - dE
EJ = EmeV * 1000000 * 1.60217646E-19
ElseIf (0.001 < EmeV) And (EmeV <= 0.15) Then
dEdx = -29096 * EmeV ^ 2 + 11329 * EmeV + 320.86
dE = dEdx * dL
EmeV = EmeV - dE
ElseIf (EmeV <= 0.001) Then
EmeV = 0
End If
Next Intervals
Cells(tmp, 7) = EmeV
End If
Next tmp
End Sub
 124

Source Code C.4. The computer code for the simulation of the energy of the alpha
particle as it left the ZnS(Ag) scintillation layer.

Sub LoopysolutionZnS()
For tmp = 15 To 20006
L = Cells(tmp, 4)
EmeV = Cells(tmp, 7)
EJ = EmeV * 1000000 * 1.60217646E-19
dL = L / 100
M = Cells(8, 1)
c = Cells(6, 4)
n = Cells(4, 3)
I = Cells(4, 6)
C1 = Cells(9, 1)
C2 = Cells(10, 1)

If L = 0 Then
Cells(tmp, 8) = EmeV
End If

If L > 0 Then
For Intervals = 1 To 100
If EmeV > 0.15 Then
v = Sqr(Abs(2 * EJ / M))
B = v / c 'unit-less
term1 = ((C1 * n) / (B ^ 2))
term2 = (C2 * B ^ 2) / (I * (1 - B ^ 2))
term3 = Log(term2) - B ^ 2
dEdx = term1 * term3
dE = dEdx * dL
EmeV = EmeV - dE
EJ = EmeV * 1000000 * 1.60217646E-19
ElseIf (0.001 < EmeV) And (EmeV <= 0.15) Then
dEdx = -29096 * EmeV ^ 2 + 11329 * EmeV + 320.86
dE = dEdx * dL
EmeV = EmeV - dE
ElseIf (EmeV <= 0.001) Then
EmeV = 0
End If
Next Intervals
Cells(tmp, 8) = EmeV
End If
Next tmp
End Sub
 125

REFERENCES

[1] Correspondence with David Hobbs, Savannah River National Laboratory.

[2] Cochran, Robert and Nicholas Tsoulfanidis. The Nuclear Fuel Cycle; Analysis
and Management. LaGrange Park, Illinois: American Nuclear Society, 1990.

[3] Closing the Circle on the Splitting of the Atom; The Environmental Legacy of
Nuclear Weapons Production in the United States and What the Department of
Energy is Doing About it. Washington D.C.: The U.S. Department of Energy
Office of Environmental Management, 1996.

[4] Savannah River Site. United States Department of Energy. March 6, 2006.
http://www.srs.gov/

[5] Evans, Robley. The Atomic Nucleus. Malabar, FL: Krieger, 1987.

[6] Knoll, Glenn. Radiation Detection and Measurement. Hoboken, NJ: Wiley &
Sons, 2000.

[7] Cember, Herman. Introduction to Health Physics. New York: McGraw-Hill,

1996.

[8] Turner, James. Atoms, Radiation, and Radiation Protection. New York: Wiley
& Sons, 1995.

[9] S.A. McElhaney, J.A. Ramsey, M.L. Bauer, and M.M. Chiles. “A more rugged
ZnS(Ag) alpha scintillation detector,” IEEE Transactions on Nuclear Science,
vol. 37, pp. 868-872, 1990.

[10] ZnS(Ag) Product Data Sheets, Saint-Gobain Crystals.

www.detectors.saint-gobain.com/data/element/node

[11] M.Y. Nadeem and Waqas Ahmed. “Optical properties of ZnS thin films,” Turk.
J. Phys., vol. 24, pp. 651-659, 2000.

[12] M. M. Chiles, “Low-level detector for alpha contamination monitor in

continuously flowing process waste water,” Nucl. Instrum. Methods, vol. 109, pp.
541-546, 1973.

[13] A. Robert, C. Sella, and R. Heindl, “Measurement system for alpha emitters in
solution,” Nuc. Instrum. Methods Phys. Res., vol. 225, pp. 179-184, 1984.
 126

[14] K. W. Skrable, K. A. Phoenix, G. E. Chabot, C. S. French, M. Jo, and G. A. Falo,

“Theoretical response of ZnS(Ag) scintillating detector to alpha-emitting sources
and suggested applications,” Health Phys., vol. 60, pp. 381-392, Mar. 1991.

[15] K. A. Phoenix, K. W. Skrable, G. E. Chabot, C. S. French, M. Jo, and G. A Falo,

“Determinating the specific alpha activity of thick sources using a large- area Zinc
Sulfide detector,” Health Phys., vol. 64, pp.64-69, Jan. 1993.

[16] B. Ayaz and T.A. DeVol, “Experimental- theoretical response of a ZnS(Ag)

scintillating disc for gross alpha measurements of aqueous radioactivity.” IEEE
Transactions on Nuclear Science, vol 51, no. 4, Aug. 2004.

[17] M. Ardid, J.L. Ferrero, and A. Herrero, “Study of the background on a ZnS(Ag)
alpha counter with a plastic veto detector.” Nucl. Instrum. Methods Phys. Res. A,
vol. 557, pp. 510-515, 2005.

[18] O.B. Egorov, R. S. Addleman, M. J. O’Hara, T. Marks, and J.W. Grate, “Direct
measurement of alpha emitters in liquids using passivated ion implanted planar
silicon (PIPS) detectors,” Nucl. Instrum. Methods Phys. Res. A, vol. 537, pp. 600-
609, 2005.

[19] Rubinstein, Reuven. Simulation and the Monte Carlo Method. New York; Wiley
& Sons, 1981.

[20] Jakeman, Beck, and McAleer. Modeling Change in Environmental Systems.

West Sussex, England; Wiley & Sons, 1993.

[21] Vose, David. Risk Analysis; A Quantitative Guide. West Sussex, England;
Wiley & Sons, 2000.

[22] Sobol, I.M. The Monte Carlo Method. Chicago; Univ. of Chicago, 1974.

[23] T. Siiskonen and R. Pollanen, “Advanced simulation code for alpha

spectrometry,” Nucl. Instrum. Methods Phys. Res. A., vol. 550, pp. 425-434, Sept.
2005.

[24] SRIM- The Stopping and Range of Ions in Matter; Instruction Manual. June
2006. www.srim.org.

[25] S. Agostinelli et al., “GEANT4 – A simulation toolkit.” Nucl. Instrum. Methods

Phys. Res. A. vol. 506, pp.250-303, July 2003.

[26] MCNPX Capabilities. April 2006. http://mcnpx.lanl.gov/

[27] T. Siiskonen and R. Pollanen, “Simulation of alpha particle spectra from aerosol
samples,” Appl. Rad. Isotopes, Vol. 60, pp. 947-953, June 2004.
 127

[28] J.L. Ferrero, C. Roldan, M.L. Acena, and E. Garcia-Torano. “Backscattering and
self-absorption corrections in the measurement of alpha-emitters in 2π geometry.”
Nucl. Instrum. Methods Phys. Res. A., vol. 286, pp. 384, 1990.

[29] C. Roldan, J.L Ferrero, F. Sanchez, E. Navarro, and M.J. Rodriguez, “Monte
Carlo simulations of alpha spectra in low- geometry measurements.” Nucl.
Instrum. Methods Phys. Res. A., vol. 338, pp. 506, 1994.

[30] S. Pickering, “The interpretation of alpha energy spectra from particulate

sources.” J. Aerosol Sci., vol. 15, pp.533, 1984.

[31] ASTAR: Stopping Power and Range Tables for Helium Atoms. National
Institute of Standards and Technology. Physics Laboratory.
http://physics.nist.gov/PhysRefData/Star/Text/ASTAR.html

[32] Hammond and Gehmlich. Electrical Engineering. New York; McGraw- Hill,
1970.

[33] Model 2006 Proportional Counter Preamplifier User Manuel. Canberra,

Industries, Inc. Meriden, CT, 2003.

[34] Correspondence with Saint Gobain Crystal, Inc.

[35] Stewart, James. Calculus. Pacific Grove, CA; Brooks/ Cole. 1995.