BETA DECAY HALF-LIVES OF NEUTRON RICH ISOTOPES IN THE

GERMANIUM-BROMINE REGION

A Dissertation

Submitted to the Graduate School

of the University of Notre Dame

in Partial Fulfillment of the Requirements

for the Degree of

Doctor of Philosophy

by

Matthew A. Quinn,

Ani Aprahamian, Director

Graduate Program in Physics

Notre Dame, Indiana

April 2010
 BETA DECAY HALF-LIVES OF NEUTRON RICH ISOTOPES IN THE

GERMANIUM-BROMINE REGION

Abstract

by

Matthew A. Quinn

Understanding the origin of the heaviest elements in the universe is one of the

great outstanding problems in physics. More than half of these elements are thought

to have been created by a process called the rapid neutron capture, or R process.

This process involves many nuclei that are very far from the line of stability. Because

these nuclei have such short beta decay half-lives, they are very difficult to produce

and study. The vast majority of r-process nuclei have not been observed at all.

This work details the measurement of several r-process nuclei that have never

been produced before. These nuclei lie in the r-process path near the Z = 40,

N = 60 region. This region of the chart of the nuclides is also interesting because a

rapid change in the shape of nuclei is seen. With the addition or removal of just a

few nucleons, nuclei go from a spherical shape to a strongly deformed shape.

For both r-process abundance calculations and nuclear deformation calculations,

the beta decay half life and beta-delayed neutron emission ratio Pn values are impor-

tant parameters. This experiment determined these values for the r-process nuclei

in the Germanium-Bromine region. The experimental work was carried out at the

National Superconducting Cyclotron Laboratory at Michigan State University.

90
Twenty-two halflives were determined in our measurments. Three nuclei, Se,
 Matthew A. Quinn

88 89
As, and As, were measured for the first time. In addition, we have confirmed

previously measured half-lives for Y, Sr, Rb, Kr, Br, Se, and As isotopes. The new
92−90 96,94
measurements were used to calculate the Pn values for Br and Rb.

An r-process simulation was performed with the results from this measurement.

Because of the possible change in shape of the Se nuclei in particular, the uncertainty

of r-process production abundances in the A = 90 − 130 region are as large as a

factor of two. This work determined the shape of the Se isotopes in this region to

be deformed, which removed this large uncertainty in the r-process calculations.

 CONTENTS

FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv

TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii

CHAPTER 1: NUCLEOSYNTHESIS . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Big Bang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.1 Big Bang Timeline . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2.2 Big Bang Nucleosynthesis . . . . . . . . . . . . . . . . . . . . 6
1.3 Stellar Nucleosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.1 Charged Particle Nucleosynthesis . . . . . . . . . . . . . . . . 7
1.3.2 S Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3.3 R Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.3.4 R Process Abundance Observations . . . . . . . . . . . . . . . 15
1.3.5 R Process Models . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.3.6 R Process Sites . . . . . . . . . . . . . . . . . . . . . . . . . . 18

CHAPTER 2: NUCLEAR STRUCTURE . . . . . . . . . . . . . . . . . . . . 20

2.1 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.1.1 Vibrational Nuclei . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.2 Deformed Nuclei . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.3 Np Nn Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.1.4 Structure A = 100 . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2 Nuclear Physics Inputs for R-Process Simulations . . . . . . . . . . . 32
2.2.1 R-Process Codes . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.2.2 Masses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.3 Beta Decay Halflives . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

CHAPTER 3: EXPERIMENT . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.1 Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.2 Experimental Efforts . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.3 Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.4 Particle ID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.5 Detector setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

ii
 3.6 Electronics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

CHAPTER 4: DATA ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . 58

4.1 Particle Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.1.1 Energy Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.1.2 Time of Flight . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.2 Gain Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.3 Thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.4 Event Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.4.1 Charge State Contanimants . . . . . . . . . . . . . . . . . . . 64
4.4.2 PIN1 PIN2 Gate . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.4.3 Time-of-Flight Gate . . . . . . . . . . . . . . . . . . . . . . . 66
4.5 Determining Halflives . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.5.1 Fitting Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.5.2 Beta Detection Efficiency from Decay Curve Fits . . . . . . . 72
4.5.3 Maximum Likelihood Method . . . . . . . . . . . . . . . . . . 72
4.5.4 Probability for Observing Decays . . . . . . . . . . . . . . . . 74
4.5.5 Obtaining Background Rates From Zero-Decay Probability . . 75
4.5.6 Likelihood Function . . . . . . . . . . . . . . . . . . . . . . . . 76
4.6 Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.6.1 Statistical Error Contributions . . . . . . . . . . . . . . . . . . 76
4.6.2 Systematic Error Contributions . . . . . . . . . . . . . . . . . 76
4.7 Halflife Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.8 Neutron Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.8.1 Neutron Events . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.8.2 Calculating Pn values . . . . . . . . . . . . . . . . . . . . . . . 87
4.9 Pn Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

CHAPTER 5: DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.1 Half-life Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.1.1 Comparison to Data . . . . . . . . . . . . . . . . . . . . . . . 91
5.2 Deformation Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.3 R-Process Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.3.1 R-Process results . . . . . . . . . . . . . . . . . . . . . . . . . 98

CHAPTER 6: CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

iii
 FIGURES

2.1 Shell model level scheme. [22] . . . . . . . . . . . . . . . . . . . . . . 21

2.2 Nilsson-model calculation showing evolution of single neutron energy
levels as l2 term is decreased. Decreasing this term in the modified
harmonic oscillator Hamiltonian represents more neutron rich nuclei,
therefore the plot shows the evolution of shell structure as nuclei
become more neutron rich. The N = 50, 82, and 126 shell gaps are
all reduced in this calculation because the low j states above the shell
gaps are reduced in energy, while the high j state below the gaps are
raised in energy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3 First 2+ energy levels in neutron rich nuclei around A = 100. A
rapid drop in energy is seen between N = 56 and N = 60. This drop
corresponds to the rapid onset in deformation due to the presence of
intruder states. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.4 Calculation of ground state energy as a function of deformation for
neutron rich Kr, Sr, and Zr isotopes. The onset of a deformed min-
imum for N = 60 in the Zr isotopes can be seen, while deformed
ground states are predicted below N = 60 for the Sr and Kr isotopes.
This suggests that deformation may be seen in the Se and Ge isotopes
with N < 60. From [45] . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.5 Wood Saxon single particle neutron (top) and proton (bottom) energy
levels as a function of deformation parameter β. From [45]. . . . . . . 31
2.6 Difference between nuclear mass from the Semi-empirical mass model
and experimentally measured nuclear masses as a function of neutron
number. The overall RMS fit is good, but large deviations at N =
28, 50, 82, and 126 can be seen. From [37]. . . . . . . . . . . . . . . 34
2.7 SEM vs P value for nuclei between Z = 28-50 and N = 50-82. The
behavior of each element can be seen clearly. By fitting each isotopic
chain, predictions can be made for the masses of unmeasured nuclei
[41]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

3.1 Calculated abundance of neutron-rich seeds produced in an alpha-rich

freezeout following a core collapse supernova. From [26]. . . . . . . . 42

iv
3.2 K1200 Dees and trim coils during installation. From NSCL image
library. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.3 Layout of the NSCL showing the K500 and K1200 cyclotrons, A1900
fragment separator and location of the experimental vault used for
this work. From NSCL image library. . . . . . . . . . . . . . . . . . . 47
3.4 SEGA germanium detectors surrounding the implantation plate. . . . 51
3.5 Schematic drawing of the setup of the SEGA detectors and aluminum
implant plate used for particle id. Pin 0 was used as the dE detector
in this setup, while the same scintillators (not shown) used in the rest
of the experiment provided the particle time-of-flight. . . . . . . . . . 52
3.6 BCS detectors. The four PIN detectors can be seen in addition to
the ribbon cable connected to the DSSD. . . . . . . . . . . . . . . . . 53
3.7 NERO detector. The inner ring of 3 He detectors and the outer rings
of BF3 detectors can be seen. Though not pictured here, the BCS
detectors sit inside central cavity during the experiment. . . . . . . . 55
3.8 Schematic drawing of the time-of-flight, BCS, and NERO detectors
used in this experiment. . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.9 Schematic of the beta-detection electronics setup. . . . . . . . . . . . 57

4.1 PIN1 vs ToF, both uncorrected. . . . . . . . . . . . . . . . . . . . . . 60

4.2 PIN1 vs ToF. PIN1 corrected, Tof uncorrected. . . . . . . . . . . . . 60
4.3 ToF vs focal plane position for Yttrium isotopes. Tof uncorrected.
Lack of counts between the two sets is due to the presence of alu-
minum “finger” to block out unreacted primary beam. . . . . . . . . 62
4.4 ToF vs focal plane position for Yttrium isotopes. Tof corrected for
momentum dependence. . . . . . . . . . . . . . . . . . . . . . . . . . 62
241
4.5 Spectrum from Am source. . . . . . . . . . . . . . . . . . . . . . . 63
90
4.6 Spectrum from Sr source. . . . . . . . . . . . . . . . . . . . . . . . . 63
4.7 PIN1,PIN2,PIN3, and PIN4 energy sum vs focal plane position for
99
Y. Charge state contaminants have a lower total kinetic energy,
and were therefore excluded by the gate shown. . . . . . . . . . . . . 65
4.8 PIN1 energy vs PIN2 energy. Counts outside the gate were excluded. 65
4.9 N-ToF vs S-Tof. Counts outside of the gate shown were excluded. . . 66
4.10 Decay curve for 91 Br with mother, daughter, granddaughter and back-
ground components. . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

v
4.11 Decay curve for 93 Kr with mother, daughter, granddaughter and
background components. . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.12 Decay curve for 94 Kr with mother, daughter, granddaughter and
background components. . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.13 Decay curve for 95 Rb with mother, daughter, granddaughter and
background components. . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.14 Decay curve for 96 Rb with mother, daughter, granddaughter and
background components. . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.15 Decay curve for 96 Sr with mother, daughter, granddaughter and back-
ground components. . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.16 Decay curve for 97 Sr with mother, daughter, granddaughter and back-
ground components. . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.17 Decay curve for 98 Y with mother, daughter, granddaughter and back-
ground components. . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.18 Results for Y isotopes, along with previous measurements [?] and
FRDM-QRPA calculations [50]. Note that FRDM deformed and
spherical resulst overlap in some cases. . . . . . . . . . . . . . . . . . 77
4.19 Results for Sr isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50]. . . . . . . . . . . . . . . . . . . . . . 78
4.20 Results for Rb isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50]. . . . . . . . . . . . . . . . . . . . . . 78
4.21 Results for Kr isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50]. . . . . . . . . . . . . . . . . . . . . . 79
4.22 Results for Br isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50]. . . . . . . . . . . . . . . . . . . . . . 79
4.23 Results for Se isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50]. . . . . . . . . . . . . . . . . . . . . . 80
4.24 Results for As isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50]. . . . . . . . . . . . . . . . . . . . . . 80
4.25 ADC spectra for NERO quadrant A. Y axis is the number of counts in
each detector. X axis is the uncalibrated neutron energy. Detectors
1-4 are 3 He detectors, and show their characteristic spectrum, with a
sharp peak at the Q value of 0.764MeV. Behind that peak are events
where the entire Q value of the reaction was not deposited into the
detector. When the outgoing deuterons strike the tube walls a sharp
drop in the spectra can be seen. Below that, the events for which
the 3 H nuclei strike the tube walls can also be seen. Detectors 5-15
are BF3 detectors, which show a broader peak at the full Q value of
2.792MeV. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

vi
4.26 ADC spectra for NERO quadrant B. . . . . . . . . . . . . . . . . . . 81
4.27 ADC spectra for NERO quadrant C. . . . . . . . . . . . . . . . . . . 81
4.28 ADC spectra for NERO quadrant D. . . . . . . . . . . . . . . . . . . 82
4.29 TDC spectrum from NERO ring 1 detectors. Channels are propor-
tional to time neutron is detected after a beta decay event. Ring 1
is the innermost ring of detectors, and therefore sees more neutrons
shortly after beta decay with relatively few neutrons coming at later
times. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.30 TDC spectrum from NERO ring 2 detectors. Channels are propor-
tional to time neutron is detected after a beta decay event. Ring 2
is the middle ring of detectors, and sees more neutrons at later times
than does ring 1. This is due to the fact that neutrons reaching this
ring have to travel through more moderating material. . . . . . . . . 83
4.31 TDC spectrum from NERO ring 3 detectors. Channels are propor-
tional to time neutron is detected after a beta decay event. Ring 3 is
the outermost ring of detectors, and sees the most neutrons at later
times of all the NERO rings. . . . . . . . . . . . . . . . . . . . . . . . 83
4.32 β background rate versus run number. . . . . . . . . . . . . . . . . . 84
4.33 β background rate versus DSSD front strip number. . . . . . . . . . . 84
4.34 Neutron background rate versus run number. . . . . . . . . . . . . . . 84
4.35 Neutron background rate versus DSSD front strip number. . . . . . . 85
4.36 Background neutrons per β as a function of run number. . . . . . . . 85

5.1 Se isotopes results compared with FRDM predictions. . . . . . . . . . 92

5.2 As isotopes results compared with FRDM predictions. . . . . . . . . . 92
5.3 Measured E(2+ ) values for Se isotopes. . . . . . . . . . . . . . . . . . 93
5.4 Measured E(2+ ) values for Kr isotopes. . . . . . . . . . . . . . . . . . 94
5.5 Measured E(2+ ) values for Sr isotopes. . . . . . . . . . . . . . . . . . 94
5.6 Measured E(2+ ) values for Zr isotopes. . . . . . . . . . . . . . . . . . 96
5.7 Measured E(2+ ) values for N = 56 isotones. . . . . . . . . . . . . . . 96
5.8 Results of R-Process simulation using measured half-life values (black
line) and values predicted for spherical ground state shapes (red line). 99

vii
 TABLES

4.1 PREVIOUS HALFLIFE MEASUREMENTS . . . . . . . . . . . . . . 86

4.2 PREVIOUS Pn MEASUREMENTS . . . . . . . . . . . . . . . . . . . 88

5.1 ASTROPHYSICAL PARAMETERS USED FOR R-PROCESS SIM-

ULATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

viii
 CHAPTER 1

NUCLEOSYNTHESIS

1.1 Introduction

Understanding the elements that make up our universe has been a goal of hu-

mankind since ancient times. The Greeks gave us the concept of atoms (from the

Greek “Atomos” meaning un-cuttable); an idea which states that the universe is

made up of some number of elementary particles. This concept remains with us

today, and while we know that there are particles more fundamental than atoms,

we still can not say why atoms exist in the abundances that we see in nature. The

lightest nuclei, including Hydrogen, Helium, and Lithium are thought to have been

created shortly after the Big Bang. Therefore, elements heavier than hydrogen and

helium must be created inside of stars through thermonuclear reactions [1]. Light

nuclei fuse together to create heavier nuclei until a nucleus with maximal binding

energy is created. Creation of the heaviest elements however, remains a mystery.

Beyond the Iron region, nuclear fusion is unlikely to occur because of the high

electrostatic repulsion between nuclei. This fact seems to indicate that nuclei cap-

turing neutrons create the heaviest elements. Stellar abundance patterns of heavy

nuclei suggest that several neutron capture mechanisms may be responsible. Direct

evidence of any such mechanisms has been elusive. Models created to reproduce the

observed abundance patterns simulate these different processes as a series of neu-

tron captures on seed nuclei. From these models two processes are responsible for

1
most of the production of heavy nuclei. They are called the rapid (r) and slow (s)

neutron capture processes. While a general understanding of the r and s processes

has been around for nearly fifty years [1], the sites where they occur still remain

unknown. The problem of determining how elements heavier than iron are created

remains one of the great unsolved problems in physics [2], and is at the heart of this

project.

We try to understand the r-process through simulations. One example includes

the ngam code by Bradley Meyer. The ingredients of the code can be divided into

astrophysical paramters and nuclear properties. The aim of this dissertation is to

measure the evolution of nuclear structure in the region near A=90. This region is

thought to be a region for the onset of deformation in nuclear structure. Deformation

of a nucleus affects the mass of the nucleus and its beta decay rates. Also, it is a

nearly impossible task to measure all or a significant portion of nuclei that may

occur along the r-process. We do however aim to measure as far as possible from

nuclear stability. The NSCL facility is one such place where we can produce some

of these rare r-process nuclei.

This work seeks to measure the beta-decay halflives and beta-delayed neutron

emission ratios of new nuclei in the r-process path around the region of yttrium to

selenium. These nuclei are interesting because they lie in the r-process path just

past a bottle neck in the process at the nucleus 78 Ni, which can have a great influence

on the abundance peak at A = 130 and higher. The nuclei in this region may even

be the seed nuclei for the r-process. Their properties may reveal how the r-process

starts and how other astrophysical processes end. This work will show the effects

of these new halflives on the solar system abundances via an r-process abundance

network calculation. In addition to their astrophysical significance, these isotopes

are also a few nucleons away from nuclei that exhibit fantastic changes in shape.

2
The beta-decay halflives of these isotopes will give the first indication of whether

additional nuclei also change shape rapidly, and will further the understanding of the

mechanisms that govern this and other nuclear properties. This work will explore

the effects of these new measurements on the predictions made for nuclear shapes

in this region.

1.2 Big Bang

In order to understand how elements are made inside of stars, it is instructive

to review how the light elements that make up stars are thought to have been cre-

ated. The Big Bang is a model used to describe the evolution of the universe from a

singularity to its present (and future) state. Monsignor Georges Lemaitre, a physi-

cist and Catholic priest from Belgium, and Russian physicist Alexander Friedmann

separately developed theories describing an expanding universe. Lemaitre called

the idea a “hypothesis of the primeval atom” [3]. Both Lemaitre and Friedmann’s

solutions stemmed from the fact that Einstein’s general theory of relativity did not

permit static solutions for the universe. This meant that instead of being perpet-

ually stable, the universe must have begun at some time in the past. From this

initial point, the universe exploded outward and grew into its present enormous

size. Two years before Lemaitre’s paper, Edwin Hubble had discovered that every

other galaxy is receding from our own at a speed proportional to its distance from

us [4]. He formulated a relation known as Hubble’s Law: v = H0 D where v is the

recessional speed of a particular galaxy, D is its distance from Earth, and H0 is a

constant known as the Hubble constant. That all galaxies in the universe appeared

to be moving away from one another seemed to indicate that Big Bang concept was

correct.

In 1965 the discovery of the Cosmic Microwave Background radiation gave exper-

3
imental support for the Big Bang [5] theory. The CMB is a field of radiation, peaked

at 160.2GHz, that is seen isotropically throughout the universe. Its existence was

predicted as a consequence of energetic photons scattering off of Hydrogen atoms

when the universe was only 379,000 years old. After the Big Bang the universe was

extremely hot, and took that long to cool enough for Hydrogen atoms to be formed.

The presence of the CMB and its extremely isotropic distribution could only be ac-

counted for by the Big Bang hypothesis. The Cosmic Background Explorer (COBE)

experiment, begun in 1989, measured the temperature and anisotropy of the CMB

with great precision. It revealed that the spectrum of the CMB was described nearly

perfectly by a blackbody emitter, as was predicted by the Big Bang theory. In ad-

dition to the spectral shape, COBE also revealed slight anisotropies in the CMB

distribution throught the universe. This anisotropy in the early universe is believed

to be what made it possible for matter condense into galaxies and solar systems

instead of being smeared out evenly throughout the universe. These results were

seen to be strong enough evidence for the Big Bang hypothesis that the Nobel Prize

was awarded to COBE scientists John Mather and George Smoot in 2006.

1.2.1 Big Bang Timeline

The Big Bang theory includes a detailed hypothesis of how the universe evolved

from a very short time after it came into existance [6]. In the first instant after the

beginning of the universe the four fundamental forces: Gravity, Electromagnetism,

Strong and, Weak, were indistinguishable. The universe was extremely hot and

dense, and was expanding outward. This time is known as the Planck Epoch, and

lasted only 10−43 seconds. At this time gravity separated from the other fundamental

forces. The universe had cooled enough so that two particles, the X and Y bosons

could be created. These theoretical particles could have decayed into two quarks

4
or one anti-quark and one anti-lepton. Decaying more favorably into two quarks

could explain why we see far more matter than anti-matter in the universe today.

This period, called the Grand Unification Epoch, lasted until 10−36 seconds after

the beginning of the universe.

After the Grand Unification Epoch, the universe underwent a rapid expansion.

Its size went from approximately 10−50 meters in diameter to about 1 meter in

diameter. This represents an increase in volume by a factor of 1078 in a period of only

about 10−33 seconds. A scalar field settling into a lower energy state caused inflation.

The result of the rapid expansion was that inhomogeneities and anisotropies present

in the universe were smoothed out, which is how we observe the universe today. At

the end of the Inflation Epoch the universe stopped its rapid expansion, and instead

converted the potential energy of the scalar field into radiation and particles such

as quarks, electrons, and neutrinos. The universe went from being largely empty to

being filled with a hot, dense plasma of particles.

The universe continued to expand and cool and eventually, at a time 10−12 sec-

onds after the beginning of the universe, its temperature was cool enough for the

electromagnetic and weak forces to decouple. At this time then all of the four funda-

mental forces existed in the manner in which we observe them today. The universe

was still too hot to allow the combination of quarks into heavier particles. It was

not until 10−6 seconds after the Big Bang that quarks could form heavier parti-

cles known as hadrons, which include protons and neutrons. From this time until

one second after the Big Bang hadrons dominated the universe, being produced in

hadron/anti-hadron pairs until the universe had cooled to the point where these

pairs could no longer be created. After that point, the hadrons and anti-hadrons

collided and annihilated each other, leaving only a small amount of hadrons re-

maining. From one second to three minutes after the Big Bang, particles known

5
as leptons were the dominant matter in the universe. At the three-minute point,

Leptons, which include electrons, stopped being produced as particle/anti-particle

pairs and began to annihilate each other.

1.2.2 Big Bang Nucleosynthesis

Although at three minutes, the universe was largely made up of radiation, the

building blocks of nuclei and atoms are thought to have present. The temperature of

the universe had lowered to the point where protons and neutrons could combine to

form heavier nuclei, and for only about seventeen minutes light nuclei were created.

This process, known as Big Bang Nucleosynthesis or BBN, while creating only a

few isotopes, is thought to have been responsible for the production of 98 percent of

the nuclei we see today. Protons and neutrons combined to form deuterium, which

often captured an additional proton and neutron to form 4 He. As can be seen in

Figure 1.1, small amounts of deuterium, tritium, 3 He, 6 Li, and 7 Li were created via

BBN, while 1 H and 4 He were dominant. Because 8 Be is unstable, the production of

heavier elements from 4 He must take place through the combination of three 4 He

nuclei to form 12 C. This is called the triple α process, and is known to be extremely

slow. Because of the length of time the triple α process takes and because of the

small abundances of other isotopes, heavier nuclei were not created in the Big Bang.

After only seventeen minutes of activity, the universe had cooled to the point that

nuclear fusion was no longer possible. It would not be until 100 million years later

that any new nuclei would be created inside the first stars. But it was from the

nuclei created in the Big Bang that stars themselves would be created, and would

create the elements that we see today.

6
1.3 Stellar Nucleosynthesis
1.3.1 Charged Particle Nucleosynthesis

A look at the present elemental abundances of our solar system reveals clues

about the how those elements were produced in the past, and how they continue

to be produced today. Hydrogen and helium are the most abundant, as they were

produced in the Big Bang. Elements heavier than H, He, and Li are produced in

extremely low abundances. These elements were created only minutes after the Big

Bang, over 13 billion years ago. Changes to these abundances are the result of

nuclear burning and decay processes inside of stars.

The first stage of nuclear burning consumes hydrogen to create helium. This is

done through the p − p chain burning with proton-proton reactions, and through

CNO cycle proton reactions on carbon, nitrogen, and oxygen nuclei. The p −p chain

converts four protons to one 4 He nucleus. The net reaction is:

4p + 4e → 4 He

The process must proceed through intermediary reactions though, because the prob-

ability of four protons reacting at the same time in a stellar environment is essentially

zero. In addition, the reaction:

p + p → 2 He

will not occur because the nucleus 2 He is unstable. This is also true for reactions

that produce 3 Li:

p + p + p → 3 Li
2
He + p → 3 Li

Creating deuterium then, is the solution to converting hydrogen to helium. The

reaction:

7
 p+p→d+e

is governed by the weak interaction and is therefore very slow. The mean lifetime

for a proton to be converted into deuterium in a stellar environment is about 1010

years. This is quite fortunate for us though, as it has allowed our sun to burn long

enough for life to develop here on Earth.

Deuterium burning takes place via several reactions, all of which are governed

by the strong force and proceed much more quickly than hydrogen burning. The

lifetime for a deuterium nucleus in a stellar environment is about 1.6 seconds. The

reactions with the largest cross-section are:

d + p → 3 He + γ

d + 3 He → 4 He + p

Therefore, burning usually proceeds through the 3 He nucleus, rather than the

d + d → 4 He

reaction. This cycle is called the p −pI chain. The process can also proceed through

the chains:

3
He + 4 He → 7 Be

7
Be + e → 7 Li

7
Li + p → 4 He + 4 He

and

3
He + 4 He → 7 Be

8
 7
Be + p → 8 B + γ

8
B + e → 8 Be + ν

8
Be → 4 He + 4 He

These are known as the p−pII and p−pIII chains, respectively. The p−pI chain is

responsible for 86 percent of 4 He production, while the p − pII and p − pIII chains

are responsible for about 14 and 0.02 percent, respectively.

The energy released in these nuclear reactions opposes the gravitational pressure

exerted on the stellar core by the outer layers. When hydrogen burning ceases, the

star will contract due to lack of thermonuclear energy release. The pressure on

the He in the core will increase until burning is possible. As 8 Be is unstable, α-α
12
reactions will not contribute to creating heavier nuclei. Stable C will be created

by the triple α reaction:

3α → 12 C

From the carbon nuclei produced in the triple alpha reaction, oxygen and neon

are then created through the reactions:

12
C + 4 He → 16 O

16
O + 4 He → 20 Ne

After helium burning, carbon burning:

12
C + 12 C → 20 Ne + 4 He

12
C + 12 C → 23 Na + p

12
C + 12 C → 23 Mg + n

neon burning, and oxygen burning occur. The final stage of charged particle reac-

tions is silicon burning. Because of silicon’s high Coulomb barrier, direct burning

9
(28 Si + 28 Si) does not take place. Instead, as the temperature inside the star in-

creases, nuclei are photodisintegrated by the high energy gamma rays present. The
28
result is that Si and 4 He nuclei are available to interact and create 32
S. 32
S will

capture 4 He and form 36

Ar and so forth, allowing the star to build up heavier and
56
heavier elements. Charged particle reactions continue until Fe is created, which

has a higher binding energy per nucleon than heavier nuclei. This means that further

charged particle reactions are unlikely to occur. In order to create heavier elements,

neutron captures will be the dominant mechanism. The two neutron capture pro-

cesses responsible for most of the heavy element production are called the s- (slow)

and r- (rapid) neutron capture processes.

In addition to the s- and r- processes, there are other processes that contribute

to the total heavy element production [1]. This was realized because there are
190 168
proton-rich nuclei such as Pt or Yb that cannot be created by the neutron-

rich r- and s-processes. The p- (proton capture) process creates heavy proton-

rich nuclei by a series of γ-n or γ-α reactions on heavy seed nuclei. Thus, the

p-process is not responsible for creating heavy elements from lighter constituents

as in charged particle reactions or the s- and r-processes, but it does change the

observed heavy element abundances. Because of the high temperatures necessary

for the photodisassociations, possible sites for the p-process include neutron star

mergers and supernova shock fronts [8]. The rp- (rapid proton capture) process

does create heavy nuclei from lighter seeds. It entails making heavy proton rich

nuclei through a series of proton captures on proton-rich seed nuclei followed by β +

decays. Sites for the rp-process are thought to include X-ray bursts on accreting

neutron stars and nova explosions of white dwarfs.

The results of Big Bang and stellar nucleosynthesis can be seen in the elemental

abundances present inside of our own sun. Nuclei created in the Big Bang and

10
in older generations of stars are still present today. Figure 1.2 shows these abun-

dances. It is important to note that these abundances show the effects of nuclear

structure. Especially stable nuclei like 4 He, 12 C, 16 O, and 56 Fe are produced in large

abundances. In addition the r- and s-process peaks occur at the points where these

processes encounter closed neutron shells. Therefore, in order to understand the

abundances of the elements, we must understand the nuclear structure that guides

their creation.

1.3.2 S Process

The first process that dominates production of elements heavier than iron is the

slow neutron capture process (s-process)[9]. The s-process is characterized by lower

neutron fluxes (compared to the r-process) and a path close to stability. For a stable

or slightly unstable nucleus of charge Z and mass A, the capture of a neutron can

lead to a stable or an unstable (A+1) nucleus. If the product nucleus is stable, it

will simply wait to capture an additional neutron (or neutrons) before undergoing

beta decay. If the product nucleus is unstable, it may beta decay to its Z+1 isobar,

or it may capture another neutron before beta decay has a chance to occur. The

neutron flux, beta decay half-lives and neutron capture cross sections determine how

long the process takes, where nuclei will be involved, and what the final abundances

will be. The observed solar abundances show large peaks at masses 90, 135, and

210. These are related to where the s-process path intersects the N = 50, 82, and

126 shell closures.

The s-process can be modeled using a series of coupled differential equations,

which produce time dependent isotopic abundances.

dNA (t)
= Nn (t)NA−1 (t)hσνiA−1 − Nn (t)NA (t)hσνiA − λβ(A) NA (t)
dt

11
 The first term on the right side of the equation describes its production by neu-

tron capture on the lighter A-1 nucleus. The second term describes its destruction

by capturing a neutron to form the A+1 nucleus. The third term describes its

destruction by beta decay. With low capture cross sections and a large enough neu-

tron flux a steady flow will be achieved, with NA × hσνiA = const. The system of

equations can then be solved analytically by making the following substitutions [6]:

λn = 1/tnγ = Nn hσνiA .

One of the following assumptions must be made then: either λ1 << λ2 or

λ1 >> λ2 . For the best fit to the data multiple exponential neutron exposures are

needed.

This model breaks down for cases where λ1 = λ2 . These are the so called s-
134 148 151 154 170
process branching points which include Cs, Pm, Sm, Eu, Yb , and
176
Lu. A network using the temperature dependent beta decay and neutron capture

rates must be used at these points.

Three different components of the s-process have been recognized, in order to

correctly reproduce the solar system s-process abundances. The first is the Main

component. It is responsible for producing most of the nuclei with A 90-208. It takes

place in the He burning shells in low mass (< 4M⊙) asymptotic giant branch (AGB)

stars. The neutron source comes from 13 C “pockets” created in the H-He intershell.

Recurrent thermal instabilities above the He burning shell cause the H-He intershell
13
to become convective during the thermal pulse. Pockets of C are created in the

intershell by protons injected from the hydrogen envelope during each Third Dredge
12
Up (TDU) phase capturing on C. After the thermal pulse, convection mixes the
13 16
He, C and s-process material. The C+α → O + n reaction then produces the

neutrons required for the s-process at a temperature of T ∼ 3 × 108 and a flux of

12
nn = 108 cm3 for a period of about 20000yr.

The second component of the s-process is the weak component. It produces

nuclei with mass up to A ∼ 90 and is thought to occur in massive stars. The

primary neutron source is the 22 Ne + α → 25 Mg + n reaction. This occurs in regions

22
of convective He burning. Ne is produced from the conversion of CNO material
14
into N in the H burning shell. Thus this process is metallicity dependent. The
14 22
N then captures two alpha nuclei in He region to produce Ne.

The third component is the strong component introduced by Clayton and Rass-
208
back in 1967. It is responsible for producing of the Pb abundance, and occurs in

low metallicity Fe/H < −1.5 low mass AGB stars.

These predictions of s-process abundances are bolstered by experimental evi-

dence from grains of stardust found in meteors. These grains contain s-process

elemental abundances that agree closely with predicted s-process abundances [?] [?]

[?]. The large difference in abundances of these isotopes as compared to those in

our solar system is taken as evidence that these grains really do come from outside

the solar system. This material may be reliably called s-process material, and indi-

cates that the s-process abundance predictions are largely reliable. Having reliable

s-process abundnace predictions is important because, as will be described below,

the r-process abundances that this project and other works depend upon will be

derived using s-process abundnace predictions.

1.3.3 R Process

The abundance peaks at masses 80,130, and 190 must be produced by a process

separate from the s-process. The intersection of these masses with the closed neutron

shells indicates that the nuclei involved in this process are much farther from stability

than the s-process nuclei. Because of this, the r- or rapid neutron capture process is

13
defined by beta-decay lifetimes that are longer than the time necessary to capture a

neutron. This is in contrast to the s-process, which is characterized by beta-decay

lifetimes that are much shorter than the neutron capture times. A path far from

stability with lifetimes ∼ 100ms necessitates a large flux of neutrons, which suggests

that the r-process occurs in explosive environments such as supernovae.

Burbidge, Burbidge, Fowler and Hoyle (B2FH) in 1957 theorized that the solar

abundances of the elements can be explained via a number of different processes,

however the astrophysical conditions and sites where an r-process may take plance

are still unknown [1]. Their paper describes how the double peaks seen in abundance

plots are due to the superposition of slow and rapid neutron capture processes. It

also shows how these peaks are due to the neutron shell closures that each process

path encounters. They were able to make calculations for the abundances from

these processes, and were able to estimate timescales for the process as well. These

classical models used 56 Fe as a seed for the r-process. Newer models attempt to take

into account the physics of the proposed site in a more realistic way. For models

using core-collapse supernova as the site of the r-process, the tremendous amount

of energy released by the collapse ( 1056 ergs) is enough to disassociate the nuclei

in the core into protons and neutrons. As the temperature drops, the protons and

neutrons combine to form alpha particles. Because of the high density, electrons

in the area are degenerate, and it becomes energetically favorable for protons to

capture electrons and form neutrons. Alpha particles fuse together to form heavier

nuclei, and in the presence of excess neutrons, can form neutron rich nuclei heavier

than iron. This scenario is called an ”alpha rich” freezeout, and provides seeds for

the r-process in a more realistic way than the classical model.

14
1.3.4 R Process Abundance Observations

Determining the amount of r-process material present in a particular star is

not a trivial matter. The elemental abundances present in a given star include

material created in BBN as well as nucleosynthesis from previous generations of

stars. Therefore the presently observed abundances are a superposition of possibly

many p-, s-, and r-processes. For the elements heavier than iron, the solar system

r-process abundances must be obtained from the modeled s-process abundances.

As the s-process path occurs close to the line of stability, beta-decay rates and

neutron capture cross sections can be measured for these nuclei. With this data

and information about the stellar environment, it is possible to generate the solar

abundance due to s-process nuclei. The r-process abundance is then taken as the

difference between the solar observed abundances and the s-process abundances.[10]

Rprocess = Solar − Sprocess

The uncertainties in the r-process abundances due to uncertainties in observations

and s-process calculations have been treated by [11]. The resulting error bars are

small for r-process only elements such as Ba and Ir, while being larger for the

elements created mainly in the s-process.

Recent observations of very old galactic halo stars have shed new light onto

the r-process, while creating new questions at the same time[12]. A class of stars

called Metal-poor halo stars show remarkable agreement with solar system r-process

abundances. These stars are called metal poor because they contain less iron relative

to hydrogen than our sun. Although in astronomy any element past helium can be

referred to as a metal, iron is used to classify the metallicity of stars because it

is produced in high abundance and can easily be detected through spectroscopy.

The log of the ratio of iron to hydrogen is generally referred to when determining

15
metallicity. Stars with values of less than one are considered metal poor (log F e/H <

−1).

Halo stars are thought to contain the remnants from only one or a few r-processes.

This is because they are very old and contain very little heavy element abundance.

These halo stars were produced from the remnants of the first stars in the galaxy and

would contain r-process material only from the nearby stars of that first generation.

Thus, unlike the solar abundances, they contain very pure r-process contributions

instead of a superposition of multiple r-process events from multiple generations of

stars.

For a number of different stars spanning a wide range of metallicities, the r-

process abundances of elements with Z > 55 agree nicely with the solar system

abundances [13]. This suggests that the mechanism for creating the r-process el-

ements above Z = 55 is quite robust, and does not differ very much from event

to event. However for elements with Z < 55, there is a large under-production of

r-process elements in the halo stars. It is unclear then if the r-process outputs can

differ for different stellar conditions, or if there are two (or three) different types of

r-processes that occur with different temperature and neutron density conditions.

Suggestions of a weak r-process [15] or of a Light Element Primary Process [14] have

been made in order to describe these abundance patterns. It is clear, though, that

because Ge-Br isotopes are at the center of this disputed region, measuring their

nuclear properties is critical to understanding the differences between metal-poor

and solar heavy element abundances.

1.3.5 R Process Models

Because the nuclei involved in the r-process are so far from stability, measure-

ments of even their most basic properties has been rare. Early models of the r-

16
process [17] [1] [16] set up a system of differential equations that describe the change

in isotopic abundances. In general this includes neutron captures, beta-decays, beta-

delayed neutron emission, photodisassociation, and fission for all nuclei involved in

the process.

This system of equations is greatly simplified if one assumes that the abundance

of each isotopic chain is contained largely in one nucleus. This is because the

neutron captures in an isotopic chain will occur before any of the nuclei will have

a chance to beta-decay. As more and more neutrons are added the last neutron

will be less strongly bound to the nucleus. At some point the gamma rays present

in this (presumably) violent scenario will photodisassociate the last neutron added,

preventing further neutron captures. The nucleus will then wait at a particular

mass until it beta-decays to the next isotope. This “waiting point” nucleus contains

the majority of the matter abundance for its isotopic chain. The waiting points are

determined by the temperature of the scenario and the binding energy of the last

neutron, which can be obtained from the nuclear mass. The series of differential

equations describing the r-process can then be simplified greatly, using only the

abundance for each isotopic chain. The dependence on the neutron capture rates is

also removed, meaning that the simulation will rely only on the beta-decay rates,

nuclear masses, neutron density, and temperature. The equation for the abundance

of each isotopic chain is given by:

3/2  3/2
2πh2

Y (Z, A + 1) G(Z, A + 1) A+1 Sn (Z,A+1)
= nn e( kT ) .
Y (Z, A) 2G(Z, A) A mu kT
Where Y(Z,A) is the abundance of the nucleus A Z, Y (Z, A + 1) is the abundance of
A+1
the nucleus Z, G(Z, A) and G(Z, A + 1) are the partition functions for A Z and
A+1
Z, nn is the neutron number density, Sn is the neutron separation energy, and

T is the temperature. [19] and [18] showed what neutron density and temperature

17
range the waiting point approximation is valid for, and that this region successfully

reproduces the observed r-process abundances. Generally, temperatures higher than

1GK and neutron densities larger than 1020 /cm3 are required to successfully repro-

duce the r-process.

1.3.6 R Process Sites

The location of the r-process is currently unknown. From an early time [1] it was

recognized that explosive scenarios such as core collapse supernova or neutron star

mergers were good candidates for the site of the r-process. This is because of the

need for high neutron number densities (> 1020 ) and high temperatures (> 1GK).

Core collapse supernova start out as massive stars (M = 8 − 100 M⊙). Through

nuclear burning they reach a point where their cores are composed of iron group

nuclei. Because the iron group nuclei have the highest binding energy per nucleon

further charged particle reactions will be endothermic. The core stops burning

and can no longer support the gravitational pressure from upper layers of the star,

which causes the core to collapse. As the temperature in the core increases, nuclei

are photo-disintegrated to alpha particles and bare nucleons. The density of the

core increases until it reaches nuclear density and a proto-neutron star is formed. A

shock wave rebounds from the now incompressible core and is sent out through the

outer layers of the star.

Unfortunately, in models of the explosion, this shock stalls in the outer layers of

the star and does not produce an explosion that would transfer material throughout

the interstellar medium. Energy released by the proto-neutron star in the form of

neutrinos might be a solution to this shock stall, but the amount of energy deposited

by these neutrinos is still not enough to trigger an explosion. Further attempts in

the form of two- and three- dimensional simulations, with the inclusion of magnetic,

18
acoustic, convective, rotational, and asymmetric effects have been made with no

definitive explanation emerging.

Perhaps the most promising explanation has been the neutrino-driven wind sce-

nario. In it, neutrinos from the cooling center of the proto-neutron star deposit

energy into its surface and ablate material.[20] [21] This high entropy (S > 400kb),

neutron rich material, called the neutrino driven wind, is driven from the proto-

neutron star into the low density region behind the shock. There it recombines

mainly through alpha captures to form neutron rich seed material heavier than

iron. The r-process then occurs with these neutron rich seeds instead of iron group

nuclei, as in the canonical r-process models.

19
 CHAPTER 2

NUCLEAR STRUCTURE

2.1 Structure

The goal of nuclear physics is to understand the properties of a nucleus as a

function of the number of protons and neutrons that make it up. Since the 1940’s

the shell model has been accepted as way to successfully describe the properties of

a nucleus[22]. Its largest success has been the description of the so called magic

numbers. These are the numbers of protons and neutrons for which exciting a

particle to the next excited state takes a large amount of energy. Starting with a

description of a single bound nucleon in a simple square well potential is a natural

first step. It is a mathematically simple description of a central potential that

gives a bound solution with a non-zero ground state energy, and a series of excited

states with large energy gaps between certain numbers of nucleons. These energy

states are generated by solving for the eigenvalues of the Schrodinger equation, using

the square well potential in the Hamiltonian. The levels are filled in order due to

the Pauli principle, which says that two nucleons can not have the same quantum

numbers[23]. Unfortunately, the magic numbers produced are not the same as seen

in experiments, and the energy gaps are identical in energy also unlike what is

seen in experiments. The magic numbers 2,8,16,20,28,50,82, and 126 were able to

be described using a harmonic oscillator plus a term proportional to the square

of the orbital angular momentum and a term proportional to the orbital angular

20
momentum dotted with the spin.

Figure 2.1. Shell model level scheme. [22]

Even though the correct magic numbers are predicted by the shell model, it is

important to note that this is not the full story. The spacing and ordering of the

orbital levels changes as nucleons are added to a system. This is obvious from the

fact that the potential will change with the addition of a proton or neutron. In

addition to the change in the central potential, nucleons also have interactions with

other nucleons. These interactions depend on the overlap of the individual nucleon

orbits and their energy spacing. These interactions include not only proton-proton

and neutron-neutron interactions, but proton-neutron interactions as well. In fact,

the p-n interactions are largely responsible for the changes in the shape of a nucleus,

known as nuclear deformation.

Further from stability the shell structure can change a great deal. Evidence for

21
the breakdown of the N = 8 shell gap has been seen near the neutron drip line in 12 Be

[24]. The emergence of a gap at N = 14 in 22 O has also been seen. Some predictions

have suggested that for heavier nuclei, the familiar magic numbers will be quenched

far from stability and new magic numbers will appear, while others indicate that the

old magic numbers will remain. Using r-process simulations some have shown that

a quenching of the N=50,82 gaps is necessary to correctly reproduce the observed

132
solar system r-process abundances[25]. Experiments in the Sn region have shown

some evidence for this effect[27], but there is still no conclusive picture. Determining

the role of shell closures and magic numbers in the neutron rich region is extremely

important for nuclear structure and nuclear astrophysics. New magic numbers or

the quenching of shell effects test our understanding of the nucleus and will change

the nuclear physics inputs for the r-process.

22
Figure 2.2. Nilsson-model calculation showing evolution of single neutron energy
levels as l2 term is decreased. Decreasing this term in the modified harmonic oscil-
lator Hamiltonian represents more neutron rich nuclei, therefore the plot shows the
evolution of shell structure as nuclei become more neutron rich. The N = 50, 82,
and 126 shell gaps are all reduced in this calculation because the low j states above
the shell gaps are reduced in energy, while the high j state below the gaps are raised
in energy.

23
2.1.1 Vibrational Nuclei

As the number of valence nucleons increases, the number of residual p-n inter-

actions increases[28][23]. This increased number of residual p-n interactions can

overwhelm the single particle shell model structure. The mixing of many nearly

degenerate levels produces one lowered level with a highly coherent wave function.

Such a state will have collective properties for the entire nucleus. Expanding the

residual interaction into multipoles will give monopole, dipole, quadrapole, and

higher terms. The electric quadrapole term plays a strong role in the low lying

states of a nucleus. A quadrapole excitation of a spherical nucleus will cause the

nucleus to oscillate in shape. The situation can be understood by treating the nu-

cleus as a liquid drop and describing the shape of the nucleus with a deformation

on the spherical shape[7].

R = R0 [1 + Σµ αµ · Y2µ (Θ, φ)].

The Hamiltonian is then:

2
1 1 dαλµ
H = Cλ Σµ |αλµ |2 + Dλ Σµ
2 2 dt
where Cλ is a parameter related to the surface and Coulomb energies of the
5
 1/2
nuclear liquid drop, and Dλ = ρRλ0 . Cλ and Dλ are related by ωλ = D
Cλ
λ
, and

~ωλ is the quantum of vibrational energy for the multipole λ.

One phonon quadrapole excitation of a 0+ ground state in an even-even nucleus

will produce a 2+ excited state. A second phonon can produce a triplet of states

with J π = 0+ , 2+ , or 4+ . The two phonon triplet states should be at roughly

twice the energy of the one phonon state. The ratio E4+ /E2+ or the ratio of the

energy of the first 4+ level to the energy of the first 2+ level is used as a guide

24
to identify vibrational collective nuclei. In addition to this vibrational structure,

enhanced collectivity will also lower the first 2+ state in relation to its position in

a closed-shell nucleus. Thus the energy of the first 2+ state can be used as a check

for increasing or decreasing collectivity in a series of isotopes.

2.1.2 Deformed Nuclei

The addition of more nucleons to a nucleus already between shell closures will

produce more residual p-n interactions. These interactions will result in more config-

uration mixing, and a further departure from a spherical shape. Eventually even the

multipole vibrations will not be able to accurately describe the system, and a new

formalism is needed. These nuclei are seen as having a permanent deformed shape

in the ground state and excited states. Deformed nuclei can have quadrapole and

higher deformations, with the quadrapole being the most common at low energies.

The shape of a nucleus can then be described by the equation[7]:

R = R0 [1 + Σµ αµ · Y2µ (Θ, φ)].

Nuclei can either be prolate shaped (football) or oblate shaped (frisbee). Because

the deformed shape breaks angular symmetry the nucleus will be free to rotate. The

rotational Hamiltonian is:

~2 2
H= J .
2I
where I is the moment of inertia, and J is the rotational angular momentum oper-

ator. Generally, I can be obtained by expanding the equation

2
~2

∂H
I=
2 ∂J(J + 1)

25
in powers of J(J +1). For the simple case of considering the nucleus a rigid ellipsoid,

this becomes:

2
I = MR02 (1 + 0.31β)
5
4
p π ∆R
where β = 3 5 Rav

For the ground and low-lying states in even-even nuclei all of the angular mo-

mentum will be due to rotation. The expression for the energies of these states is

then:

h2
E= J(J + 1).
2I
The first few excited states will have energies of:

6h2
E2+ =
2I

20h2
E4+ =
2I

42h2
E6+ =
2I
Therefore, the ratio E4 /E2 will be 20/6 = 3.33. We can get an approximate idea

of the behavior of a nucleus by looking at its E4 /E2 ratio.

Spherical: < 2

Vibrational: 2

Rotational: 3.33

Because the deformation is a result of collective behavior, the first 2+ will also

be lowered compared to spherical nuclei.

26
2.1.3 Np Nn Scheme

A useful way to study collectivity and deformation is through a parameterization

of the valence proton-neutron interactions. This is called the Np Nn scheme because

it looks at the product of the number valence protons and neutrons[30]. It is assumed

that the onset of collectivity and deformation is due solely to the proton-neutron

interactions. The product of the number of protons and neutrons (or holes) counted

from the nearest closed shell is plotted versus the E2+ or E4+ /E2+ values for a series

of nuclei. These plots can show trends across regions that give important clues to

changes in the nuclear structure.

Looking at the average number of proton-neutron interactions per valence nu-

cleon can give good rules of thumb concerning the general nuclear structure. This

is called the ”promiscuity factor” and is given by:

P = Np Nn /Np + Nn .

Plotting the E4+ /E2+ ratio versus P shows us that nuclei are deformed (E4+ /E2+ =

3.33) for P > 5, and may be deformed for P = 4. This means that for a nucleus to

be deformed, both Np and Nn must be greater than or equal to 4.

2.1.4 Structure A = 100

The reordering of levels can have consequences for the ground state properties

as well. Nuclei in the A = 100, N = 60 region show a sudden and severe change in

the ground state E2+ levels, E4+ /E2+ ratios, quadrapole deformations, and half-lives.

This behavior is unlike the gradual change from spherical to vibrational to deformed

properties. The presence of two configurations, one spherical, one deformed account

for the different properties of the nuclei in this region.

A suggested explanation for this behavior is the monopole component of the

27
proton-neutron interaction[31]. The monopole component of the proton-neutron in-

teraction does not depend on the angle between the proton and neutron orbits (as

the P0 (cos θ) Legendre polynomial has no θ dependence). It does depend on the

radial distance between the proton and neutron orbits, so that the proton-neutron

interaction will be strong for orbits with the same orbital angular momentum re-

gardless of what their total angular momentum is. So for instance, the ν1g7/2 (and

ν1h11/2 ) orbit and the ν1g9/2 orbit will have a strong proton-neutron interaction,

which will have the effect of lowering the orbits. As the ν1g7/2 (or ν1h11/2 ) neutron

orbit is filled and the 1g9/2 proton orbit is lowered, the the shell gap at Z = 40 will

be reduced or destroyed.

If the nucleons fill the reordered shells from the bottom up as usual, the nucleus

will be spherical. However, if a pair of protons from the 2p1/2 orbit is excited

across the reduced shell gap a deformed configuration will result. This is because

the excitation of a proton pair creates more valence protons, which increase the

collectivity of the system. In the P scheme this would correspond to configurations

100
in Zr with Np = P = 0 for a strong Z = 40 shell gap, and Np = 10, P > 4 for no

shell gap.

Deformed “intruder states” have been seen in other regions of the nuclear land-

scape [32]. In such regions near proton shell closures, the lowest state of the deformed

configuration is a 0+ level that is seen to lower in energy as neutrons are added to-

ward mid-shell. The 0+ state is lowest in energy at the middle of the neutron shell,

and increases in energy as neutrons are added past the middle of the shell. The

difference for nuclei in the A = 100 region is that the deformed 0+ state is lowered

beneath the 0+ spherical ground state. Thus the ground state becomes deformed,

and the ground state properties change dramatically.

The rapid transition to deformed ground states has been seen in Mo isotopes

28
above the Z = 40 shell closure at Zr, and Sr isotopes below. Models predict strong

deformation in the Kr and Se isotopes in this region [45], but measurements of

Kr isotopes do not indicate strong deformation [33]. Because the beta-decay rates

depend on the ground state deformations, measuring the beta-decay half-lives of

nuclei in this region will help to indicate if the nuclei below Zr and Kr around

A = 100 also have sudden transitions to deformed ground states.

Figure 2.3. First 2+ energy levels in neutron rich nuclei around A = 100. A rapid
drop in energy is seen between N = 56 and N = 60. This drop corresponds to the
rapid onset in deformation due to the presence of intruder states.

29
Figure 2.4. Calculation of ground state energy as a function of deformation for
neutron rich Kr, Sr, and Zr isotopes. The onset of a deformed minimum for N = 60
in the Zr isotopes can be seen, while deformed ground states are predicted below N
= 60 for the Sr and Kr isotopes. This suggests that deformation may be seen in the
Se and Ge isotopes with N < 60. From [45]

30
Figure 2.5. Wood Saxon single particle neutron (top) and proton (bottom) energy
levels as a function of deformation parameter β. From [45].

31
2.2 Nuclear Physics Inputs for R-Process Simulations
2.2.1 R-Process Codes

R-process simulations are done by solving a network of coupled differential equa-

tions for the abundances of each isotope. These equations can include terms for

neutron captures, beta-decays, beta-delayed neutron captures, photodisintrigations,

neutrino captures, and fission. The network solves for the abundance of each iso-

tope given initial choices of parameters such as: seed nuclei, neutron number density,
56
temperature, and duration of the simulation. In classical r-process models Fe is

chosen as a seed, while simulations of an alpha-rich freezeout use multiple neutron-

rich seed nuclei. The neutron rich nuclei in the Ge-Br region of interest in this

work are among these seed nuclei. For a canonical r-process simulation considering

neutron capture, beta decay, and beta-delayed neutron emissions each differential

equation will look like [?]:

X X
Ẏ (Z, A) = λZ ′ ,A′ Y (Z ′ , A′ ) + nn hσνiZ ′ ,A′ Y (Z ′ , A′ )Yn
Z ′ ,A′ Z ′ ,A′

Where Y (Z, A) is the abundance of the nucleus A Z, σ is the neutron capture rate for

the nucleus A Z-1. The first term on the right hand side of the equation includes beta-

decays and photodisassociations, while the second term includes neutron induced

reactions.

For large enough neutron number densities and high temperatures (i.e. photon

densities), neutron capture and photodisassociation will occur much more quickly

than beta decay. This situation is known as an (n, γ) ⇔ (γ, n) equilibrium, where

the nucleus with the maximum abundance for each isotopic chain must wait for

the slower beta decay before proceeding to the next isotopic chain. The change in

abundance will then look like:

Ẏ (Z, A) = λA+1 Y (Z, A + 1) − hσνiZ,A Y (Z ′ , A′ )Yn

32
An equilibrium condition ensures that Ẏ (Z, A) = 0 The reaction rates λ between

neutron capture and photodisassociation are related by detailed balance, and are

given by :
  3/2  3/2
2G(Z, A) A mu kT Sn (Z,A+1)
λA+1 = hσνiA e( kT ) .
G(Z, A + 1) (A + 1) 2πh2
where G is the partition function, T is the temperature, and Sn is the neutron

separation energy for the nucleus A+1 Z. Thus, the location of the waiting point nuclei

depends largely on the nuclear mass (via the Sn ), and the temperature. Combining

the abundance equation and rate equation above, the abundance ratios for each

mass of a given Z can then be calculated from the equation:

3/2  3/2
2πh2

Y (Z, A + 1) G(Z, A + 1) A + 1 −Sn (Z,A+1)
= nn e( kT ).
Y (Z, A) 2G(Z, A) A mu kT
In dynamic network codes, for a given time step abundances are calculated in

addition to new values of nn and T. The simulation is ended when the neutron

number density or temperature drop below a certain level. Static network codes

calculate abundances the abundances in each isotopic chain for the full simulation

time, then factor in the contribution to each abundance due to beta decays. The

most important nuclear physics inputs then are the nuclear masses (Sn values),

beta-decay halflives, and the beta-delayed neutron emission ratios. For nuclei far

from stability these values are unknown and must be calculated. There are often

large differences in predicted values, so it is critical that these properties be mea-

sured. Because the astrophysics surrounding the location and mechanism for the

r-process are unknown, accurate nuclear physics information can help to constrain

the astrophysical parameters.

2.2.2 Masses

Because so few r-process nuclei have been measured, performing r-process sim-

ulations requires the use of predictions for nuclear properties. Nuclear masses are

33
perhaps the most important nuclear data needed for r-process calculations. Masses

determine the neutron separation energies, Q values for beta decays, and define the

neutron drip line. With the temperature, the nuclear masses dictate what the r-

process path will be. For a given temperature of 1 − 2 × 109 K, the r-process waiting

points will occur at nuclei with a Sn of 1.4MeV. Because very few masses have been

measured for r-process nuclei, abundance calculations rely heavily on nuclear mass

extrapolations or calculations.

The first mass model was the semiemperical model from von Weizsacker [34]

and Bethe and Bacher [35], which treated the nucleus as liquid drop. It contains

terms for nuclear volume, volume symmetry, surface energy, and coulomb energy

in addition to a surface symmetry term added by Meyers and Swiatecki [36]. Four

coefficients (all terms except Coulomb) are then adjusted to fit the measured mass

values and create a formula to predict unknown masses. This formula produces a

surprisingly good fit for only five terms with an overall rms deviation of 2.97MeV

[37], but fails badly for closed shell nuclei, with an error of around 10MeV.

Figure 2.6. Difference between nuclear mass from the Semi-empirical mass model
and experimentally measured nuclear masses as a function of neutron number. The
overall RMS fit is good, but large deviations at N = 28, 50, 82, and 126 can be seen.
From [37].

Improved mass models attempted to introduce microscopic shell corrections to

34
the liquid drop model. These macroscopic-microscopic models include the Finite

Range Droplet Model (FRDM) and the Extended Thomas Fermi plus Stutinski

Intergral (ETFSI) formulas. The FRDM [38] differs from earlier liquid drop mod-

els in that it takes into account the effect on the surface energy due to the finite

range of the nucleon-nucleon interaction. It also treats the effect of nuclear shape

on the Coulomb field and the diffuseness of the charge distribution, and introduces

a exponential compressibility term that corrects earlier drop models’ tendencies to

overstate the central nuclear density. The microscopic corrections include a poten-

tial for use in the Strutinsky integral in order to introduce shell effects into the

calculation. It is composed of three parts: a spin independent nuclear part, a spin

orbit part and a Coulomb part.

The FRDM also uses a version of the BCS treatment for paring corrections, a

Wigner term to treat T = 0 pairing in nuclei with N ≈ Z, and a charge-asymmetry

term to account for the energy difference between mirror nuclei. The results for

the FRDM are much improved over earlier liquid-drop models, as it produces a rms

error of 0.67MeV and shows shell effects.

The ETFSI mass formula [39] is based on the extended Thomas Fermi approx-

imation. Both the macroscopic and microscopic parts of the calculation center

around a Skyrme-type force. This means that unlike the FRDM calculations, the

microscopic and macroscopic parts of the calculation are done self-consistently. The

results from ETFSI calculation are similar to the purely microscopic Hartree-Fock

calculations, while being much faster to perform.

The latest mass formulas treat the nucleus microscopically, in that they attempt

to solve the Schrodinger equation for some choice of nuclear mean-field force. The

Hartree Fock formula solves the Schrodinger equation for an initial choice of the

mean field (usually either zero-range Skyrme or finite-range Gogny forces) and wave

35
functions. The equation is then solved iteratively in order to achieve self consistency

between the wave functions and the mean field.

~2 2
 
− ∇ + U φi = ǫi φi
2M
Pairing effects are state dependent, and are added by either the BCS or Bo-

golyubov methods. The HFBCS and HFB mass models of the Brussels group [40]

produce an rms error of just over 700keV, but produce worse results for doubly

closed shell nuclei, and nuclei further from stability.

An additional mass model known as the P-Scheme model is based on the re-

lationship between the microscopic portion of the nuclear mass with the average

number of proton-neutron interactions per valence nucleon[41]. The microscopic

mass is obtained by taking the difference between measured nuclear masses and

the macroscopic portion of a FRDM calculation. This quantity is defined as the

Semi-Empirical Mass (SEM)

SEM = Mtot − Mmacro .

The average number of proton-neutron interactions per valence nucleon is defined

as:

P = Np Nn /(Np + Nn ),

where Np is the number of valence protons or proton holes, and Nn is the number

of valence neutrons or neutron holes. Particles are counted when the shell is less

than half full, and holes are counted when the shell is more than half full. The

SEM is plotted against the P values for a series of nuclei in an isotopic chain, and

the resulting curve is fit with a polynomial. With the polynomial calculated from

known nuclei, the SEM value can be obtained for nuclei of unknown mass.

36
Figure 2.7. SEM vs P value for nuclei between Z = 28-50 and N = 50-82. The be-
havior of each element can be seen clearly. By fitting each isotopic chain, predictions
can be made for the masses of unmeasured nuclei [41].

This representation breaks down at or near the closed shells, where the number

of valence particles and holes goes to zero. In these cases, one can use a related

parameterization called F-Spin, which is defined as:

1
F0 = (Nπ Nν ),
2
where Nπ and Nν are the number of proton and neutron valence pairs. This

overcomes the problems of using the P parameterization at closed shells, but is re-

stricted to even-even nuclei, as only pairs of nucleons are considered. Again the SEM

value is plotted, this time against the F0 value for each nucleus, and a polynomial

relationship allows one to calculate mass values for unknown nuclei.

37
2.3 Beta Decay Halflives

Beta decay is a weak nuclear process that converts a neutron in to a proton

(or vice versa), while keeping the nuclear mass number A constant. A beta decay

also emits a beta particle (electron or positron) and a neutrino. There are several

types of beta decay: β + , β − , and electron capture. β + and electron capture occur

in proton-rich nuclei, while β − occurs in neutron-rich nuclei. The general formula

for beta-decay is:

β+ : A A +
Z XN → Z−1 YN +1 + β + ν + Qβ

β− : A A −
Z XN → Z+1 YN −1 − β + ν + Qβ

Electron Capture : A − A
Z XN + e → Z−1 YN +1 + ν + Qβ

Q is the amount of energy released in the decay and corresponds to the difference

in mass between the parent and daughter nuclei. If Q is greater than zero, the decay

has the possibility of occurring.

The beta-decay rate is a function of the transition strengths from the ground

state of the parent nucleus to the ground state and excited states in the daughter

nucleus with the decay rate described by:

m5e c4
λ= f (Z, E0)|ΣµMf hJf Mf ξ|Oλµ (β)|Ji Mi ζi|2
2π 3 ~7
where

ΣµMf hJf Mf ξ|Oλµ (β)|JiMi ζi|2

is the reduced transition probability to the final state, and

f (Z, E0 )

38
is the Fermi function, which describes the effect on the beta particle’s wave function

due to the Coulomb field of the daughter nucleus.

The decay half-life can be related to the decay constant λ through:

d[N]
= −Nλ
dt

N
= eλ·t
N0

ln 2
t1/2 =
λ
where N0 is the number of nuclei originally present in a sample, and N is the

number of nuclei still present at time t. The emitted electron and neutrino both

have intrinsic spin = 1/2. There are two possible alignments of the spins then, 1, and

0. These different types of beta-decays are known as Fermi (spins anti-aligned) and

Gamow-Teller (spins aligned). If the beta particle and neutrino are emitted with

orbital angular momentum ℓ = 0 (the allowed approximation), the total change in

nuclear spin ∆J = 0 for Fermi-type decays, and ∆J = 0, 1 for Gamow-Teller-type

decays. The parity change for both types of decays is ∆π = (−1)ℓ so that for allowed

decays, the parity will be the same in the initial and final states. The change in

isospin ∆T = 0 for Fermi-type decays and ∆T = 0, 1 for Gamow-Teller-type decays.

Therefore, the Fermi-type decays proceed mainly through isobaric analogue states

(IAS) in the daughter nucleus. Because of the large neutron to proton ratio in r-

process nuclei, the parent and daughter generally do not have analogue states near

one another, which means that Fermi type decay is unlikely to occur.

In addition to the allowed (ℓ = 0) decays, there can also be ℓ 6= 0 or forbidden

decays. Decays with ℓ = 1 are known as first forbidden decays and have ∆J = 0, 1

39
for Fermi-type decays, and ∆J = 0, 1, 2 for Gamow-Teller-type decays. The parity

change for both types is ∆π = (−1)ℓ = 1. Because of the higher angular momentum,

first forbidden decays are less probable to occur than allowed decays by a factor of

1000.

As with nuclear masses, the beta-decay rates for most nuclei involved in the r-

process have not been measured. This requires the use of beta-decay rate predictions

for use in r-process simulations. Also like with nuclear masses, beta-decay rates

have been calculated with either global or microscopic formulas. The global Gross

Theory picture treats the beta strength distribution in a statistical manner[?]. It

integrates the strength function over the low-lying excited states in the daughter

nucleus to get the half-life. Microscopic approaches based on the Random Phase

Approximation (RPA) [?] treat low lying excited states in the daughter nucleus

as particle-hole excitations of the ground state. These states can be treated as

spherical or deformed, with the values of the half-life being strongly dependent on

the deformation. Generally, the deformation is determined by a calculation from

a mass model. In our case, we have compared our measured values with a RPA

calculation using the FRDM mass model. Table 4.1 shows the wide range of half-

lives for the region of interest depending on the choice of ground state shape.

40
 CHAPTER 3

EXPERIMENT

3.1 Goals

By performing this experiment we hope to provide information about how the

nuclear orbitals and magic numbers change far from the line of stability, and how

these changes in nuclear structure affect the r-process abundances. The neutron

rich isotopes in the region of Ge-Br approaching N = 60 sit in the middle of both

of these questions. It has been shown for lighter nuclei that the magic numbers

disappear and new magic number appear if enough neutrons are added[24].

As the structure of nuclei change, other nuclear properties also change. This

means that beta-decay half-lives, beta-delayed neutron emission ratios, neutron cap-

ture rates, and masses can all be very different from their values in more stable nu-

clei. Changes in these values will affect the abundances produced by the r-process.

There have been suggestions that the energy gaps of the neutron closed shells at

N = 50, 82 are smaller than previously thought[25] [27].

R-Process calculations done with mass models similar except for their shell

strength have shown that a significant abundance difference develops. The ETFSI

and ETFSI-Q mass models are similar calculations except that the ETFSI-Q model

used quenched shell strength, which has the effect of filling in the abundance troughs

around the A = 80, 130, and 190 peaks. Because the shells are not as strong the

beta-decay half-lives are shorter and material will flow through the region more

41
smoothly, eliminating the abundance troughs. For alpha-rich freezeout scenarios,

neutron rich nuclei in the Ge-Br region may even be the seeds for the r-process.

Figure 3.1. Calculated abundance of neutron-rich seeds produced in an alpha-rich

freezeout following a core collapse supernova. From [26].

In addition to the nuclear structure questions surrounding the A = 100 region,

the presence of different “weak” and “main” r-process abundances for Z < 50 re-

quires additional nuclear data. Information about the neutron rich Ge-Br isotopes

will help to disentangle the contributions from these two different processes, in addi-

tion to any other processes that may contribute to heavy element production across

the universe.

42
3.2 Experimental Efforts

The production and study of nuclei far from stability requires specialized fa-

cilities. Generally, two methods of production are used: the In-Flight and Ion

Separation On-Line (ISOL) techniques. The In-Flight technique uses an energetic

beam of stable nuclei that impinges on and reacts with a target. Exotic nuclei are

produced through reactions such as Fusion-Evaporation, in which a heavier product

nucleus is created; or Projectile Fragmentation, where a lighter product is created.

In either case magnetic separation is required to remove unreacted primary beam

and uninteresting products from the nuclei of interest. This method can produce a

wide range of proton- and neutron-rich nuclei. It is advantageous because the nu-

clei are produced with velocities near the primary beam velocity, meaning that the

production, separation and detection can be done very quickly. In addition there is

no dependence on the chemistry of the products as in the ISOL method. There are

number of very different facilities that produce nuclei via the In-Flight technique

including: the National Superconducting Cyclotron Facility (NSCL) at Michigan

State University, TWINSOL at the University of Notre Dame, ATLAS at Argonne

National Laboratory, RIKEN in Japan, and the FAIR at GSI in Germany.

Ion Separation On-Line uses a high energy proton beam and a thick target.

Exotic nuclei are produced via spallation in the target and must be extracted from

it. While this technique can not be used for the shortest-lived isotopes, it can

provide high intensities of certain elements, and has the added feature that the

isotopes produced may be easily reaccelerated for use in further reactions. ISOL

facilities include: ISOLDE at CERN, SPIRAL at GANIL in France, HRIBF at Oak

Ridge National Laboratory, and IGISOL in Jyvaskala, Finland.

43
3.3 Setup

This experiment was performed at the National Superconducting Cyclotron Fa-

cility (NSCL) at Michigan State University. The NSCL uses two coupled cyclotrons

to accelerate a primary beam onto a target, producing a secondary beam of neutron-

rich nuclei. Particle accelerators generally work by accelerating a charged particle

through an electric field. To produce a high energy beam either the field must be

very strong, or the particle must be passed through it many times. Cyclotrons work

by using a magnetic field to bend particles into a circular orbit so that they can be

accelerated through an electric field many times.

The particle is injected into the center of the cyclotron and is accelerated in a

spiral shaped path. The electrodes in the center are known as “dees” because they

have the shape of the letter D. The dees are connected to an alternating current

which has a period equal to the orbital period of the particle. This is done so that

the particle will be accelerated through both gaps between the dees on each orbit.

The time for one orbit is the circumference divided by the velocity of the particle.

t = 2πr/v = 2πmv/Bqv = 2πm/Bq

Where B is the strength of the magnetic field, m is the mass of the particle, and

q is its charge. The orbital frequency is:

ω = qB/m.

This value is known as the cyclotron resonant frequency.

The magnets above and below the dees are held at a constant field, so that as

the particle is accelerated its orbital radius increases. The force due to the magnetic

field is:

44
 F = qvxB = mv 2 /r

so that the force is perpendicular to the direction of the velocity and produces

a circular orbit. Therefore the orbital radius is:

r = mv/Bq.

When the (nonrelativistic) particle reaches the edge of the dees it is extracted

with a kinetic energy of:

2
Tmax = 1/2mvmax = (Bqρ)2 /2m = (Bρ)2 /2(q 2 /m)

where ρ is the maximum physical radius of the dees. The term (Bρ)2 /2 is a

parameter that categorizes the capabilities of each particular cyclotron and is known

as the K value for the cyclotron. This is where the names of the two cyclotrons at

NSCL, the K500 and the K1200, get their names. The K value is often given in units

of MeV so that it corresponds to the maximum kinetic energy to which a cyclotron

can accelerate a proton (q = 1, m = 1).

At NSCL the cyclotrons have magnets broken up into high- and low-field sectors.

This is done in order to provide vertical focusing for the beam. Additionally, trim

coils are added to increase the magnetic field, in order to deal with the increase in

mass of the beam particles due to relativistic velocities. The ratio B/m appears in

the equation governing the cyclotron resonant frequency, so for an increase in mass

of:

m0
m= p
(1 − (v/c)2)

45
the magnetic field must be increased by the trim coils in order to maintain the

cyclotron frequency.

Figure 3.2. K1200 Dees and trim coils during installation. From NSCL image
library.

The first cyclotron, the K500, accelerates the primary beam of 136 Xe to an energy

of 10.85MeV/U. The beam is then passed through a stripper foil and injected into

the second cyclotron, the K1200. Next, the second K1200 cyclotron accelerates the

beam to an energy of 120MeV/U. The average intensity of the primary beam was

2pnA.

The primary beam strikes a 240mg/cm2 Be target and fragments into many

lighter neutron-rich nuclei. The fragments are produced with a kinetic energy that

is a large fraction of the primary beam, and proceed in forward direction to the

A1900 fragment separator [48]. Along with the fragments there is also degraded

primary beam that must be removed by the separator. This method of producing a

radioactive beam is known as Projectile Fragmentation. It provides fast production

46
and transport of radioactive nuclei, so that isotopes with very short half-lives may

be studied. Additionally, unlike the slower ion separation on line (ISOL) technique,

Projectile Fragmentation does not depend on the chemistry of the primary beam

or secondary products. While not utilized in this experiment, the radioactive beam

could be used on a secondary target to produce even more exotic tertiary products.

Figure 3.3. Layout of the NSCL showing the K500 and K1200 cyclotrons, A1900
fragment separator and location of the experimental vault used for this work. From
NSCL image library.

In order to separate nuclei of interest from unwanted fragment nuclei and unre-

acted primary beam, a magnetic separator system called the A1900 is used at the

NSCL. It is composed of four superconducting dipole magnets to bend the beam,

and forty quadrapole and higher multipole magnets used to focus the beam. A

dipole magnet with radius (r) and field strength (B) , bends an ion with a given

mass (m) and velocity (v) according to the following equation:

γmv = qBr

Two dipoles will produce a focused beam with a momentum to charge ratio

47
determined by the magnetic rigidity value obtained for a particular choice of the

magnetic field B. The location where this focus occurs is called the intermediate

focal plane. Nuclei with a different momentum to charge ratio are blocked by slits

at the focal plane and will not pass through. Nuclei produced in the fragmentation

reaction have similar velocities, so selection by the dipoles is essentially a selection

of nuclei with particular mass to charge ratios. To separate nuclei with similar m/q

ratios but different Z than the nucleus of interest, a degrader is also placed at the

focal plane. The energy loss for each nucleus in the degrader is given by the Bethe

formula:

dE 4πe4 z 2
− = NA ρB(v)
dx m0 v 2 A
with

2m0 v 2 v2 v2
   
B(v) = Z ln − ln 1 − 2 − 2
I c x
where z is the charge of the nucleus, v is its velocity, ρ is the density of the degrader

material and NA is Avogadro’s number.

Therefore each isotope will have a different momentum after passing through the

degrader. A second set of dipoles will again select a particular momentum, filter out

unwanted nuclei and transmit those that are of interest. For this experiment the

magnetic rigidity of the dipole magnets was tuned to pass 88 Se. A position sensitive

plastic scintillator was placed at the dispersive focal plane to act as a degrader,

and to provide a measure of the horizontal position of each isotope. The scintillator

was also used as the start for time-of-flight measurements. In addition an aluminum

“finger” was placed at the dispersive focal plane in order to block 136 Xe charge states

that were present there. Nuclei that were passed through the A1900 were then sent

to the N3 experimental vault.

48
3.4 Particle ID

To identify individual nuclear species, two methods were used. The first was

to do an identification of one nucleus by measuring a decay from an isomeric level.

Isomers are states in a nucleus that exist for a long time before decaying via gamma

decay. The decay emits a gamma ray of a unique energy that can be used as a

unique signature from one particular nucleus. Once one nucleus was unambiguously

identified, the rest could be identified through a ∆E − T oF plot. This sort of plot

separates nuclei according to A and Z. The energy loss of a nucleus passing through

a detector is also given by the Bethe formula, so that for a nucleus fully stripped of

electrons:

dE ∝ Z 3 /A

The time-of-flight is:

T oF = d/v = dA/p

where d is the distance from the the A1900 focal plane to the N3 experimental vault

and p is the momentum of the nucleus. Thus a ∆E − T oF plot will separate nuclei

by Z and A, which enables gating on a single nucleus in order to study its beta

decay properties.

Isomeric decays are used for the ID because the distance from the target to the

N3 vault is approximately 40 meters. The time of flight for the secondary beam will

then be around 500ns. This time delay ensures that prompt gamma decays have

already occurred, but that isomeric states will still be populated. An isomeric level

of over 1µs was desired to ensure a sufficient counting rate. For this experiment a
99
8.6µs isomeric state in Y was used. With this isotope identified, the remaining

49
nuclei could be identified as well.

In order to measure the 126keV gamma ray from the isomeric state of 99 Y, three

HPGe detectors were used. These detectors come from the Segmented Germanium

Array (SEGA) at MSU. Germanium detectors are semiconductor detectors that

work by transmitting the energy of an incoming photon to the valence band electrons

mainly by Compton scattering. The increased energy of the electrons promotes them

to the conduction band, where they can be collected by an anode at a high potential.

Germanium detectors are preferable because they have a high detection efficiency

and an energy resolution of 3-4keV.

A different implant station was used for this part of the experiment. It consisted

of a thick aluminum plate placed perpendicular to the beam in the N3 vault, with

a Silicon PIN ∆E detector placed in front of the plate. A PIN detector is a silicon

diode made up of three layers: a p-type (hole doped) layer, an intrinsic (no doping)

layer, and an n-type (electron doped) layer. The intrinsic layer acts as a large

depletion region in which electrons and holes are created by the incoming radiation.

A reverse bias is applied to the detector, which causes the electrons to move toward

the p-type material where they are collected by a terminal. These electrons compose

the signal to be amplified and used as the ∆E value.

The three Ge detectors were placed around the beam pipe at the location of the

plate. Nuclei passed through the two plastic scintillators giving a T oF signal, and

through the PIN detector giving a ∆E signal. They were then implanted into the

aluminum plate, without being implanted into the Beta Counting System (BCS).

Once implanted, the isomeric states decayed and were detected by the Ge detectors.

The ∆E−T oF spectrum could be gated on events that produced the 126keV gamma
99
rays. This gated spectrum then showed which events were from Y.

50
Figure 3.4. SEGA germanium detectors surrounding the implantation plate.

3.5 Detector setup

In the N3 vault, nuclei were implanted into the BCS. The BCS consists of four

silicon PIN detectors, one double sided silicon strip detector (DSSD) and one single

sided silicon strip detector (SSSD)[49]. The DSSD and SSSD detectors are similar

to the PIN detectors except that they include multiple terminals that are separated

from one another. Therefore these detectors are able to be read out in individual

strips. Our setup also included a 1000µm thick Ge detector behind the silicon

detectors. The Ge detector was tested for use in identifying gamma rays from

implanted nuclei, but this proved to be ineffective. Instead, the Ge detector was

used to veto light particles (Chapter 4).

51
Figure 3.5. Schematic drawing of the setup of the SEGA detectors and aluminum
implant plate used for particle id. Pin 0 was used as the dE detector in this setup,
while the same scintillators (not shown) used in the rest of the experiment provided
the particle time-of-flight.

The four PIN detectors had thicknesses of 991µm, 303µm, 309µm, and 966µm.

The DSSD had a thickness of 979µm, and the SSSD had a thickness of 988µm.

These thicknesses were chosen so that nuclei were implanted into the DSSD, which

was segmented into forty 1mm strips on the front and forty perpendicular 1mm

strips on the back, giving 1600 pixels. The output from the DSSD was sent to a

dual-gain preamplifier (2V/pC and 0.1V/pC), enabling it to detect both implanted

nuclei and beta decays. Each implant or decay event was recorded with the pixel

location and time stamp in addition to the energy of the event. The SSSD was made

up of sixteen strips and was used to veto light particle implantations not stopped

in the DSSD, as well as detect beta decay events that did not deposit all of their

52
energy in the DSSD. Typically implantation rates were 1 × 10−3 implants/s/pixel.

Figure 3.6. BCS detectors. The four PIN detectors can be seen in addition to the
ribbon cable connected to the DSSD.

In order to detect the beta delayed neutron, the Neutron Emission Ratio Ob-

server (NERO) detector system was placed around the BCS setup. NERO is made

up of 60 proportional neutron counters placed in a polyethylene moderator. There

is one ring of 16 3 He detectors, surrounded by two concentric rings with 20(inner)

and 24(outer) BF3 detectors. The design was chosen to maximize the detection ef-

53
ficiency of the detector. Neutrons are detected via the 3 He(n, d)3 H and 10 B(n, a)7 Li

reactions. The cross section for these reactions is highest in the thermal neutron re-

gions En ∼ 0.0025eV. The neutrons must therefore be moderated, which is done by

the polyethylene block they are seated in. The anode of the detectors measures the

energy from the deuteron and alpha particles respectively. The energy of the neu-

tron is much smaller than the Q value for these reactions, which are 0.764MeV(3 He)

and 2.792MeV(BF3 ), so the energy detected is essentially these Q values. Thus the

energy information the neutrons is lost, but we are concerned with the number of

neutrons detected, not their energy. When a beta decay was detected in the DSSD,

NERO was gated to look for neutrons for 200µs. This amount of time was chosen

to allow sufficient time for the neutrons to be moderated by the polyethylene block.

3.6 Electronics

The experimental logic is shown in Figure 3.11. The master trigger for the

experiment was either a signal in PIN1 from an implant nucleus or a beta decay

event from the DSSD. An additional master trigger was created from the three

germanium detectors timing signals when the SeGA setup was used. The master

trigger live was created by using the master trigger and a NOT busy signal from

the acquisition computer, and was used to trigger the acquisition computer and to

create gates for the acquisition components and logic.

Signals from the PIN detectors were sent to preamplifiers and then to amplifiers

with energy(slow) and timing(fast) outputs. The energy outputs were sent to VME

ADC modules, while the timing outputs were sent to a VME TDC, a scaler module

and a coincidence register.

Energy signals from the SeGA detectors were sent to a spectroscopy amplifier

and then to a CAMAC ADC. The timing signals were sent to a scaler module and

54
Figure 3.7. NERO detector. The inner ring of 3 He detectors and the outer rings of
BF3 detectors can be seen. Though not pictured here, the BCS detectors sit inside
central cavity during the experiment.

a TDC. They were also combined in an OR to create the SeGA trigger.

The signals from the forty front DSSD strips and the forty back DSSD strips

were sent in three groups (1-16, 17-32, 33-40) to Multichannel Systems preamplifiers.

These preamplifiers were able to output signals with two different gains in order to

distinguish between implant events and beta decay events. The low gain signals

were sent directly to VME ADCs, while the high gain signals were sent first to

shaper/discriminator modules and then to VME ADCs and scaler modules. The

high gain signals were also combined in a logical OR and sent to a VME bit register.

This bit was used to readout DSSD strips that had a signal from one strip in their

55
Figure 3.8. Schematic drawing of the time-of-flight, BCS, and NERO detectors used
in this experiment.

group (1-16, 17-32, 33-40).

NERO signals were read out in four quadrants of fifteen detectors each. The

signals were sent through a preamplifier and then to a shaper/discriminator. The

shaper signals were sent to a VME ADC. The fast discriminator signals were sent to

VME TDC and scaler modules. The NERO master gate was created by the master

gate generated by a beta decay event in the high gain DSSD signals. A master gate

live was made by the AND of the master gate, a 200µs gate generated after a master

gate live signal, and a not bust signal from the acquisition computer. This master

gate live was used to set a window to readout the NERO ADCs so that neutrons

only within 200µs of a beta decay were recorded.

56
Figure 3.9. Schematic of the beta-detection electronics setup.

57
 CHAPTER 4

DATA ANALYSIS

4.1 Particle Identification

Particle ID was done with the dE-ToF technique. Both the energy and time-of-

flight signals needed to be momentum corrected because the A1900 spectrometer

had a momentum acceptance of about 4%. The energy loss, when velocity corrected

for the time of flight, gives elemental (Z) separation. The time of flight, when

momentum corrected via the focal plane position, gives isotopic (A) separation.

4.1.1 Energy Loss

The energy deposited in the dE detectors by projectile nuclei is characterized by

the Bethe Formula:

dE 4πe4 z 2
− = NA ρB(v)
dx me v 2 A
With:

2me v 2 v2 v2
   
B(v) = Z ln − ln 1 − 2 − 2
I c x
where e is the electron charge, Z is atomic number of the detector material, me is

the electron mass, v is the velocity of the projectile nucleus, A is the atomic weight

of the detector material, NA is Avogadro’s number, rho is the mass density of the

58
detector material, and I is the average excitation and ionization of the detector

material.

Because the energy of the projectile nuclei has a velocity dependence, a correction

is required to separate nuclei with different Z.

r
dE me v 2 A 4 2
Z= πe z NA ρB(v)
dx 4
In the experiment the time-of-flight was measured, not the projectile velocity, so

the simple relation v = D/t is used to obtain the velocity. The distance D is the

total flight path from the focal plane of the spectrometer to the experimental vault,

which is essentially the same for all nuclei. A linear correction to the energy was

used of the form:

dEcorr = dE − M × T oF + B

where M is the slope and B is the intercept. The raw dE-ToF spectrum is shown

in fig 4.1 while the corrected spectrum is shown in figure 4.2.

59
Figure 4.1. PIN1 vs ToF, both uncorrected.

Figure 4.2. PIN1 vs ToF. PIN1 corrected, Tof uncorrected.

60
4.1.2 Time of Flight

The time-of-flight measurement is used to distinguish between different masses

of each isotope. From the following equation we can see the mass and momentum

dependence on the time-of-flight.

T oF = D/v = Dm/p

The momentum of nuclei is not directly measured in our experiment, but the

magnet settings of the spectrometer define the momentum to be accepted through

the equation:

Bρ = p/q

so that

mD
T oF =
Bρq
The 4% acceptance of the spectrometer means that there is a distribution in

time-of-flight for each mass. Using a correction for the momentum gives a unique

ToF for each mass m. We used a linear correction for the momentum dependence

of the time-of-flight. This is based on the horizontal position of each nucleus at the

image 2 focal plane of the spectrometer.

61
Figure 4.3. ToF vs focal plane position for Yttrium isotopes. Tof uncorrected. Lack
of counts between the two sets is due to the presence of aluminum “finger” to block
out unreacted primary beam.

Figure 4.4. ToF vs focal plane position for Yttrium isotopes. Tof corrected for
momentum dependence.

62
4.2 Gain Matching

The high gain outputs of the DSSD and SSSD were gain matched using a 241 Am

source before the experiment. The source was placed just in front of the DSSD and

SSSD.

241
Figure 4.5. Spectrum from Am source.

4.3 Thresholds

Low energy thresholds for the high gain outputs of the DSSD and SSSD were
90
set using a Sr source. Figure 4.6 shows the spectrum from a typical DSSD strip.

90
Figure 4.6. Spectrum from Sr source.

4.4 Event Definitions

Implant events were defined as events that had a signal in PIN1, and low gain

signal in the front and back of the DSSD. Events that produced a low gain signal in

63
the SSSD were rejected. For events with a signal in more than one strip on the front

or back sides, the strip with the maximum energy was recorded as the strip in which

the event took place. The pixel location (front and back strip numbers) and the time

of implantation were recorded for each event. Decay events were defined as events

that did not have a signal in PIN1 or PIN2, and produced a low gain signal in the

front and back of the DSSD. Events that produced an overflow signal in the small

Ge detector behind the SSSD were rejected. This was done to reject light particles

coming with the beam that could be confused with beta decay events from implant

events. For decay events, like with implants, when a signal in more than one strip on

the front or back sides occurred, the strip with the maximum energy was recorded

as the strip in which the event took place. To be considered as decays correlated

with implantations, decay events had to come within a ten seconds of an implant.

Decay events to one pixel away from the implant were considered valid. Up to three

decay events (ie. mother, daughter, granddaughter) within the ten second window

were recorded. The decay time for each decay event was the difference between the

implantation event time and the decay event time. A window from forty seconds

after an implant event in a particular pixel until the next implantation in the same

pixel was used to determine background. Any decay event that occurred in this

window was considered to be a background event.

4.4.1 Charge State Contanimants

To distinguish between fully stripped nuclei of interest and contamination from

hydrogen-like nuclei with mass A+2,3 the total kinetic energy of each nucleus was

looked at. In our experiment the pre-amplifiers for the DSSD were saturated for

many implant nuclei, thus the DSSD could not be used to determine the total kinetic

energy. However,the sum of the energy in the four PIN detectors was sufficient to

64
distinguish between fully stripped nuclei and charge state contaminants. Gates were

drawn in the image 2 focal plane position versus the PIN energy sum plot in order

to exclude charge state nuclei.

Figure 4.7. PIN1,PIN2,PIN3, and PIN4 energy sum vs focal plane position for 99 Y.
Charge state contaminants have a lower total kinetic energy, and were therefore
excluded by the gate shown.

4.4.2 PIN1 PIN2 Gate

In order to reject light particle and primary-beam contaminants, a gate was

made on the PIN1 vs PIN2 spectrum. Figure 4.8 shows this plot, and the gate used.

Figure 4.8. PIN1 energy vs PIN2 energy. Counts outside the gate were excluded.

65
4.4.3 Time-of-Flight Gate

Two photo tubes, one on each side of the image 2 scintillator were used to

measure the time-of-flight. Because of a high primary beam rate at the image 2

focal plane, the image 2 scintillator began to fail during the experiment. This was

manifest through failing to produce a signal for some particles. This resulted in

some events with spurious ToF signals. These events were rejected by creating a

gate in the N-ToF vs S-ToF spectrum. This spectrum is the time-of-flight from one

photo tube (N-ToF) versus the time-of-flight from the other photo tube (S-ToF).

Figure 4.9 shows this plot with the gate used.

Figure 4.9. N-ToF vs S-Tof. Counts outside of the gate shown were excluded.

4.5 Determining Halflives

4.5.1 Fitting Curves

For the isotopes with high statistics, the beta-decay half life could be obtained

through a fit of the decay curve. The decay times for each isotope were put into

66
bins of 100-400 ms to form decay curves. The curves were then fit using exponential

parent, daughter, granddaughter, and constant background components. These

equations give the activity at a time t, and are known as Bateman equations.

n
X
An = N0 ci e−λi t
i=1
where
Qn
λi
i=1
cm = Qn
i=1 ′(λi − λm )

The expression for the total number of decays seen at time t becomes:

N = C1 e−λp t + D1 e−λp t + D2 e−λd t + G1 e−λp t + G2 e−λd t + G3 e−λg t + B

with

C1 = N0

λd N0
D1 =
λd − λp
λd N0
D2 =
λp − λd

λd λg N0
G1 =
(λd − λp )(λg − λp )
λd λg N0
G2 =
(λp − λd )(λg − λd )
λd λg N0
G3 =
(λp − λg )(λd − λg )
where λp is the decay constant of the parent nucleus, λd is the decay constant of the

daughter nucleus, and λg is the decay constant of the granddaughter. B is the value

of the background, which is assumed to be constant. The halflife was determined by

minimizing the chi-squared of the fit, with the parent halflife, initial parent scaling,

and background being free parameters. The daughter and granddaughter halflives

were treated as fixed parameters.

67
 Halflife of Br-91 h2
Entries 49
Mean 3133

Counts
RMS 2893
120

100

80

60

40

20

0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Time (ms)

Figure 4.10. Decay curve for 91 Br with mother, daughter, granddaughter and back-
ground components.

Halflife of Kr93 h2
Entries 49
Mean 3655
Counts

RMS 2797
200
180
160
140
120
100
80
60
40
20
0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Time (ms)

Figure 4.11. Decay curve for 93 Kr with mother, daughter, granddaughter and back-
ground components.

68
 Halflife of Kr94 h2
Entries 99
Mean 1364

Counts
RMS 1381
70

60

50

40

30

20

10

0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Time (ms)

Figure 4.12. Decay curve for 94 Kr with mother, daughter, granddaughter and back-
ground components.

Halflife of Rb95 h2
Entries 99
Mean 1310
Counts

RMS 1353

600

500

400

300

200

100

0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Time (ms)

Figure 4.13. Decay curve for 95 Rb with mother, daughter, granddaughter and back-
ground components.

69
 Halflife of Rb96 h2
Entries 99
Mean 1416

Counts
220 RMS 1418
200
180
160
140
120
100
80
60
40
20
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Time (ms)

Figure 4.14. Decay curve for 96 Rb with mother, daughter, granddaughter and back-
ground components.

Halflife of Sr96 h2
Entries 28
1800 Mean 3450
Counts

RMS 2815
1600

1400

1200

1000

800

600

400

200

0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Time (ms)

96
Figure 4.15. Decay curve for Sr with mother, daughter, granddaughter and back-
ground components.

70
 Halflife of Sr97 h2
Entries 99
Mean 1546

Counts
RMS 1401
700

600

500

400

300

200

100

0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Time (ms)

97
Figure 4.16. Decay curve for Sr with mother, daughter, granddaughter and back-
ground components.

Halflife of Y98 h2
Entries 49
1200 Mean 1599
Counts

RMS 1403

1000

800

600

400

200

0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Time (ms)

98
Figure 4.17. Decay curve for Y with mother, daughter, granddaughter and back-
ground components.

71
4.5.2 Beta Detection Efficiency from Decay Curve Fits

The DSSD beta detection efficiency could be determined from the parameters of

the decay curve fits. This is an input parameter for the Maximum Likelihood fit.

The efficiency is taken as the number of parent beta decays detected divided by the

number of parent implants detected. The number of implants can be simply counted

from the particle-id spectrum, but because the number of beta decays detected for

each isotope is a combination of parent, daughter, and granddaughter decays, the

total number of decays for each isotope gate can not be used for calculating the

DSSD efficiency. The number of parent decays can be obtained from the parent

activity calculated in the above Bateman equation:

(Ao )f it
NDparent =
λf it
therefore

NDparent
ǫβ = 100 ×
NI

4.5.3 Maximum Likelihood Method

For the isotopes with few events, fitting a decay curve is not possible, so another

method of determining the half-lives was needed. We used the Maximum Likelihood

Method (MLH), which has been used before for cases of poor statistics[?][?]. This

method uses the Bateman equations to determine the probability that beta decays

with a given half-life will occur at the times seen in the experiment. The half-life

of the parent is varied so as to maximize the probability. That maximizing value is

assigned as the half-life of the parent nucleus.

More specifically, the likelihood is joint probability of observing parent, daugh-

ter, granddaughter, and background events for a particular value of the beta decay

72
constant. The beta detection efficiency, beta decay background, daughter and grand-

daughter half-lives, beta delayed neutron emission ratios (Pn ), and delayed neutron

emission daughter and granddaughter half-lives need to be known.

The likelihood function to be maximized is made up of the probability density

function for the observance of zero, one, two, or three decays at a given time t. The

probability density for a parent decay at a time t, given the decay constant λ1 is:

f1 (λ1 , t) = λi e−λ1 t

The probability for a parent decay to occur within a time t is :

Z t
F1 = f1 (λ1 , t′ ) dt′ = 1 − e−λ1 t
0

The probability density for observing a daughter decay with a decay constant

λ2 from a parent with decay constant λ1 is :

λ1 λ2
f2 (λ1 , λ2 , t) = (e−λ1 t − e−λ2 t )
λ2 − λ1
The probability for that daughter decay to occur within a time t is:

λ1 λ2 1 −λ1 t 1
F2 (λ1 , λ2 , t) = 1 − ( e − e−λ2 t )
λ2 − λ1 λ1 λ2
The probability density for observing a granddaughter decay with a decay con-

stant λ3 from a granddaughter with decay constant λ2 and a parent with decay

constant λ1 is :

λ1 λ2 λ3
f3 (λ1 , λ2 , λ3 , t) = ×
(λ2 − λ1 )(λ3 − λ1 )(λ3 − λ2 )
(λ3 − λ2 )e−λ1 t − (λ3 − λ1 )e−λ2 t + (λ2 − λ1 )e−λ3 t
 

The probability for that granddaughter decay to occur within a time t is:

73
 λ1 λ2 λ3
F3 (λ1 , λ2 , λ3 , t) = 1 − ×
(λ2 − λ1 )(λ3 − λ1 )(λ3 − λ2 )
 
(λ3 − λ2 ) −λ1 t (λ3 − λ1 ) −λ2 t (λ2 − λ1 ) −λ3 t
e − e + e
λ1 λ2 λ3
Additionally, the probability for observing exactly r background events in a time

t, given a background rate b is :

(btc )r e−btc
Br =
r!

4.5.4 Probability for Observing Decays

To perform the MLH analysis, the probability of actually seeing a decay must be

calculated. This includes the probability for decay events to occur as shown above,

as well as incorporating the efficiency of detecting the beta decays. This can be

shown simply for the detection of no beta decays within a correlation time tc . The

equations for one, two, or three observed beta decays follow in the same manner but

are too lengthy to be listed here. To determine the probability for detecting a beta

decay, we must look at the probability that a decay from a particular generation i

happens Di , and the probability that it is observed Oi . The efficiency for detection

of each generation is given by ǫi . The notation F = 1 − F ǫ = 1 − ǫ is used, with Bo

as he probability that no background events are seen in the correlation window.

The total probability to see no parent, daughter, granddaughter or background

event comes from four terms: 1) that a parent does not decay, 2) that a parent decays

but is not observed and the daughter does not decay, 3) that the parent decays and

is not observed, the daughter decays and is not observed, and the granddaughter

does not decay and 4) the parent, daughter, and granddaughter decay but are not

observed.

74
 P0 = (D 1 + D1 O1 D2 + D1 O 1 D2 O2 D 3 + D1 O1 D2 D 2 D3 O3 ) × B0

Plugging in the probability functions F from above and the detection efficiencies,

the equation becomes:

P0 (λ1 , λ2 , λ3 , t) = [1 − F1 (λ1 , t)ǫ1 − F2 (λ1 , λ2 , t)ǫ1 ǫ2 − F3 (λ1 , λ2 , λ3 , t)ǫ1 ǫ2 ǫ3 ] × B0

By including the possibility that one, two, or three neutrons may be emitted

after beta decay the equation for the probability becomes:

P0 (λ1 , λ2 , λ3 , λ2n , λ3n , λ3nn , t) = [1 − F1 (λ1 )ǫ1

−((1 − Pn )F2 (λ1 , λ2 )

+Pn F2 (λ1 , λ2n ))ǫ1 ǫ2

−((1 − Pn )(1 − Pnn )F3 (λ1 , λ2 , λ3 )

+(1 − Pn )Pnn F3 (λ1 , λ2 , λ3n )

+Pn (1 − Pn2n )F3 (λ1 , λ2 , λ3n )

+Pn Pn2n F3 (λ1 , λ2n , λ3nn ))ǫ1 ǫ2 ǫ3 ] × B0

4.5.5 Obtaining Background Rates From Zero-Decay Probability

For each isotope it is possible to calculate the background from the number of

events where no decay was observed N0 , relative to the number of events where at

least one decay was observed N123 .

N0 P0
=
N123 1 − P0
N0 and N123 are both measured in the experiment, N0 is the number of implants

for which no decay event is seen, while N123 is the number of implant events for

which at least one decay is seen. The background rate for r = 0 is then:

75
 − ln B0
b=
tc
Using the equation for P0 , one can calculate the background rate from the num-

ber of implants seen:

N0
B0 = × (1 − F1 ǫ1 − F2 ǫ1 ǫ2 − F3 ǫ1 ǫ2 ǫ3 )−1
N0 + N123

4.5.6 Likelihood Function

The total likelihood function for decay events is the sum of the probability func-

tions for seeing one, two, or three decays:

N
Y 123

L123 (λ1 ) = (δ(ni − 1)p1 + δ(ni − 2)p2 + δ(ni − 3)p3 )

i=1

The decay constant λ1 the solution for the maximization equation:

4.6 Errors
4.6.1 Statistical Error Contributions

The reported errors include both statistical and systematic error contributions.

The statistical error is obtained directly from the Maximum Likelihood analysis,

which has been used before in cases of poor statistics. Because the likelihood prob-

ability density is not symmetric about the maximum, an interval was chosen such

that the interval distance was the minimum to contain 68% of the area under the

curve.

4.6.2 Systematic Error Contributions

The uncertainties in the half-lives and Pn values of the daughter and granddaugh-

ter as well as the N-1 daughter and granddaughter contribute to the systematic error

76
for the parent half-life calculation. In addition, the background and beta-detection

efficiency contribute as well.

The contribution of these factors to the overall error was obtained by varying

each of these parameters within their own uncertainty limits. The plus and minus

one sigma values for each were used for each parameter, while the nominal values

were used to determine the quoted half-life values.

4.7 Halflife Results

Y isotopes

Present values
1500 Previous values
FRDM deformed
FRDM spherical
Halflife (ms)

1000

500

0
98Y 99Y 100Y 101Y
Nucleus

Figure 4.18. Results for Y isotopes, along with previous measurements [?] and
FRDM-QRPA calculations [50]. Note that FRDM deformed and spherical resulst
overlap in some cases.

77
 Sr isotopes
1500
Present values
Previous values
FRDM deformed
FRDM spherical

Halflife (ms)
1000

500

0
96Sr 97Sr 98Sr 99Sr
Nucleus

Figure 4.19. Results for Sr isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50].

Rb isotopes
3000
2700 Present values
Previous values
FRDM deformed
2400 FRDM spherical
Halflife (ms)

2100
1800
1500
1200
900
600
300
0
94Rb 95Rb 96Rb 97Rb
Nucleus

Figure 4.20. Results for Rb isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50].

78
 Kr isotopes
2000
1800 Present values
Previous values
FRDM deformed
1600 FRDM spherical

Halflife (ms)
1400
1200
1000
800
600
400
200
0
92Kr 93Kr 94Kr 95Kr
Nucleus

Figure 4.21. Results for Kr isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50].

Br isotopes
2000
1800 Present values
Previous values
FRDM deformed
1600 FRDM spherical
1400
Halflife (ms)

1200
1000
800
600
400
200
0
90Br 91Br 92Br
Nucleus

Figure 4.22. Results for Br isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50].

79
 Se isotopes
1800
1600 Present values
Previous values
FRDM deformed
1400 FRDM spherical

Halflife (ms)
1200
1000
800
600
400
200
0
88Se 89Se 90Se 91Se
Nucleus

Figure 4.23. Results for Se isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50].

As isotopes
1400 Present values
Previous values
FRDM deformed
1200 FRDM spherical

1000
Halflife (ms)

800

600

400

200

87As 88As 89As

Nucleus

Figure 4.24. Results for As isotopes, along with previous measurements [?] and
FRDM-QRPA calculations[50].

80
Figure 4.25. ADC spectra for NERO quadrant A. Y axis is the number of counts
in each detector. X axis is the uncalibrated neutron energy. Detectors 1-4 are 3 He
detectors, and show their characteristic spectrum, with a sharp peak at the Q value
of 0.764MeV. Behind that peak are events where the entire Q value of the reaction
was not deposited into the detector. When the outgoing deuterons strike the tube
walls a sharp drop in the spectra can be seen. Below that, the events for which the
3
H nuclei strike the tube walls can also be seen. Detectors 5-15 are BF3 detectors,
which show a broader peak at the full Q value of 2.792MeV.

Figure 4.26. ADC spectra for NERO quadrant B.

Figure 4.27. ADC spectra for NERO quadrant C.

81
Figure 4.28. ADC spectra for NERO quadrant D.

Figure 4.29. TDC spectrum from NERO ring 1 detectors. Channels are proportional
to time neutron is detected after a beta decay event. Ring 1 is the innermost ring of
detectors, and therefore sees more neutrons shortly after beta decay with relatively
few neutrons coming at later times.

82
Figure 4.30. TDC spectrum from NERO ring 2 detectors. Channels are proportional
to time neutron is detected after a beta decay event. Ring 2 is the middle ring of
detectors, and sees more neutrons at later times than does ring 1. This is due to
the fact that neutrons reaching this ring have to travel through more moderating
material.

Figure 4.31. TDC spectrum from NERO ring 3 detectors. Channels are proportional
to time neutron is detected after a beta decay event. Ring 3 is the outermost ring
of detectors, and sees the most neutrons at later times of all the NERO rings.

83
 Beta background (counts per second)
0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
0 50 100
Run

Figure 4.32. β background rate versus run number.

Beta background (counts per second)

0.05

0.04

0.03

0.02

0.01

0
0 10 20 30 40
Strip

Figure 4.33. β background rate versus DSSD front strip number.

Neutron background rate (counts per second)

0.003

0.0025

0.002

0.0015

0.001

0.0005

0
0 50 100
Run

Figure 4.34. Neutron background rate versus run number.

84
 Neutron Background Rate per DSSD Strip

Neutron Background rate (counts per second)

0.005

0.0045

0.004

0.0035

0.003

0.0025

0.002

0.0015

0.001

0.0005

0
0 10 20 30 40
Strip

Figure 4.35. Neutron background rate versus DSSD front strip number.

0.08
Neutrons per beta decay

0.06

0.04

0.02

0
0 50 100
Run

Figure 4.36. Background neutrons per β as a function of run number.

85
 TABLE 4.1

PREVIOUS HALFLIFE MEASUREMENTS

Present Work Previous Theory

Nucleus T1/2 (ms) NNDCT1/2 (ms) FRDM Def(ms) FRDM Sph(ms)[50]
87 0.157
As 0.3980.104 0.5600(0.08) 0.269 2.458
88
As 0.2810.419
0.155 0.061 2.263
89
As 0.1931.151
0.069 0.066 0.374
88
Se 1.1030.280
0.227 1.5300(0.06) 0.403 12.312
89
Se 0.3850.071
0.061 0.4100(0.04) 0.114 9.050
90 0.128
Se 0.2840.087 0.134 1.127
91
Br 0.5600.042
0.039 0.5410(0.005) 0.055 0.762
92
Br 0.3110.062
0.051 0.3430(0.015) 0.036 0.054
93 0.117
Kr 1.3570.117 1.2860(0.01) 0.516 0.078
94 0.021
Kr 0.2400.019 0.2120(0.005) 0.548 0.559
95
Kr 0.1670.050
0.039 0.1140(0.003) 0.373 0.061
95 0.024
Rb 0.4170.023 0.3775(0.0008) 0.078 0.386
96 0.013
Rb 0.2160.012 0.2028(0.0033) 0.071 0.057
97
Rb 0.2660.065
0.053 0.1699(0.0007) 0.046 0.046
96 0.062
Sr 0.9750.062 1.0700(0.01) 1.079 0.817
97 0.025
Sr 0.4570.025 0.4290(0.005) 1.179 0.119
98
Sr 0.6510.032
0.032 0.6530(0.002) 0.724 0.724
99 0.106
Sr 0.3410.101 0.2690(0.001) 0.359 0.359
98
Y 0.6260.014
0.014 0.5480(0.002) 0.305 0.302
99 0.069
Y 1.4110.069 1.4700(0.007) 0.167 0.167
100 0.047
Y 0.7960.045 0.7350(0.007) 0.318 0.318
101
Y 0.4890.264
0.155 0.4500(0.02) 0.149 0.149

4.8 Neutron Analysis

4.8.1 Neutron Events

NERO was allowed to count neutrons after a high-gain (beta) DSSD event. This

was done by opening a gate for 200µs after any such event. The neutron energy

signal was recorded via an ADC, and the neutron time was recorded by a TDC.

The ADC only recorded the first neutron event during the 200µs window, while the

86
TDC recorded all events. The individual ADC spectra and TDC spectra for each

ring are shown below.

4.8.2 Calculating Pn values

Obtaining the Pn value for a particular nucleus begins with the equation:

Nβ−n
Pn = 100 ×
Nβparent

This equation states that the Pn value (in %) is the number of beta-delayed neutrons

detected divided by the number of parent nucleus beta decays. Because the beta de-

cays detected could be from the parent, daughter, granddaughter, or background, we

must calculate the number of parent beta decays that should be detected. Nβparent

is the number of parent nucleus beta decays, and is calculated from the total number

of beta decays for a particular nucleus by:

Nβparent = Nimp (1 − e−λtcorr )

where λ is the decay constant of the parent nucleus, and tcorr is the correlation time

for beta decay events (either 5 or 10 seconds). Nimp is the total number of implants

for that particular nucleus.

The number of beta-delayed neutrons detected Nβn−detected must be corrected

for the efficiency of NERO and the BCS, as well as for the neutron background

observed. Thus, Nβ−n is given by:

Nβn−detected − Nβn−background
Nβ−n =
ǫβn

where Nβn−detected is the raw number of beta-delayed neutrons detected, Nβn−background

is the number of background beta-delayed neutrons detected in the correlation win-

dow, and ǫβn is the total efficiency for detecting beta-delayed neutrons. ǫβn is given

by: ǫβn = ǫN ERO ǫβ where ǫN ERO is the detection efficiency of NERO, and ǫβ is the

87
detection efficiency of the BCS. The number of beta-delayed neutrons detected in

the correlation window after an implant Nβn−background is calculated by the equation:

Nβn−background = Rateβn × tcorr × Nimp

where tcorr is the correlation window of 5 or 10 seconds, Nimp is the number of

implants of a particular nucleus, and Rateβn is the per pixel neutron background

averaged over each run.

4.9 Pn Results

TABLE 4.2

PREVIOUS Pn MEASUREMENTS

Present Work Previous Theory

Nucleus Pn(%) Previous value(%) FRDM Def(%) FRDM Sph(%)[50]
90Br 23.6(1.7) 25.2(9.0) 1.818 99.800
91Br 14.6(2.0) 20(3) 3.319 73.365
92Br 40.9(5.0) 33.1(21.0) 13.032 8.458
94Rb 7.3(0.4) 10.01(0.23) 1.414 1.628
96Rb 6.5(2.2) 13.4(0.4) 10.634 10.546

88
 CHAPTER 5

DISCUSSION

5.1 Half-life Calculations

The evolution of shell structure away from the line of stability is a major open

question in nuclear physics. The results of shell gaps disappearing or new gaps

appearing, affect our understanding of the nucleus and of stellar processes like the

r-process. While not a direct measure of the shape of a nucleus, the beta decay

halflife can give insight into the nuclear shape, and the evolution of shape in a

region of the chart of the nuclides.

The present results have been compared to QRPA halflife calculations using

the FRDM mass model. QRPA calculations are able to provide information on a

wide range of nuclei far from the line of stability, and have been used previously

in r-process calculations. The calculation begins by computing the ground state

nuclear masses and deformations with the FRDM. The FRDM, or Finite Range

Droplet Model, is a macroscopic-microscopic type model that calculates the total

potential energy of the nucleus. It treats the nucleus as a liquid drop, but also

takes into account the effect on the surface energy due to the finite range of the

nucleon-nucleon interaction. Additionally, it treats the effect of nuclear shape on

the Coulomb field and the diffusness of the charge distribution, and introduces

a exponential compressability term that corrects earlier drop models’ tendencies

to overstate the central nuclear density. Microscopically it includes a three part

89
potential for use in the Strutinski integral in order to introduce shell corrections,

and the Lipkin-Nogami version of the BCS method for pairing corrections.

The potential energy surface of the nucleus is calculated as a function of the

quadrupole deformation ǫ2 and octapole deformation ǫ4 . The ground state defor-

mation of a nucleus is determined by finding the minimum in this potential energy

surface. With the deformations determined, values that depend on the shape of the

nucleus such as the mass and fission barriers can be calculated.

With the ground state shapes and masses determined, a quasi random phase

approximation (QRPA) calculation is then done to obtain the beta decay halflives

and Pn values. This is done by adding a Gamow-Teller force of the form:

VGT = 2χGT β 1− β 1+

to the folded-Yukawa Hamiltonian, where:

β 1± = Σi σi t±
i

The beta decay halflife is obtained from the beta strength function, which is the

square of the nuclear matrix element. Therefore:

Sβ ∝ hφf |VGT |φi i2

The halflife is then:

Qβ
1 X
= Sβ (Ei ) × f (Z, Qβ − Ei )
T1/2 Ei

where Qβ is the Q value of the beta decay, Ei = Qβ − E0 is the excitation energy in

the daughter nucleus, and f (Z, Qβ −Ei ) is the Fermi function, which is proportional

to (Qβ − Ei )5 .

90
 The beta delayed neutron emission ratio can be calculated by looking at the

ratio of the beta strength above the neutron separation energy Sn to the total beta

strength.

PQβ
Sβ (Ei ) × f (Z, Qβ − Ei )
Pn = PSQnβ
0 Sβ (Ei ) × f (Z, Qβ − Ei )

5.1.1 Comparison to Data

The results of this experiment are compared with QRPA calculations and with

previous results in Figs 5.1 and 5.2. There is generally good agreement with previ-
98 100
ously measured values, with the possible exception of Y and Y. These nuclei

have low-lying states from which beta decay is possible. Because we can not separate

these states, the observed halflives are superpositions of these two states.

For the Se and As nuclei, there is good agreement with the QRPA deformed

calculations. The half-lives are slightly larger than the predicted values, but the

overall trend is consistent, and the predicted values for spherical shapes are much

higher.

91
 Se isotopes
1800
1600 Present values
Previous values
FRDM deformed
1400 FRDM spherical

Halflife (ms)
1200
1000
800
600
400
200
0
88Se 89Se 90Se 91Se
Nucleus

Figure 5.1. Se isotopes results compared with FRDM predictions.

As isotopes
1400 Present values
Previous values
FRDM deformed
1200 FRDM spherical

1000
Halflife (ms)

800

600

400

200

87As 88As 89As

Nucleus

Figure 5.2. As isotopes results compared with FRDM predictions.

92
5.2 Deformation Discussion

The agreement of the measured beta decay halflife with the FRDM-QRPA pre-

dictions for a deformed shape, as opposed to a spherical shape agrees with predic-

tions for the ground state deformation[44],[45]. The deformed ground state shape is

predicted because of a shell gap at Z = 34 due to the splitting of the 1f 5/2 orbitals.

The E(2+ ) systematics for the Se isotopes are unclear as N approaches 56, as can

be seen in Fig 5.3. Generally, the first 2+ state in a spherical nucleus will be at a

higher energy than for deformed nuclei. Looking at Fig 5.7, one can pick out the

high E(2+ ) for the N = 50 closed shell. The trend past N = 50 looks as though

the E(2+ ) at N = 56 could either increase to over 1MeV, which would indicate a

spherical shape; or may decrease to below 800 keV, which would indicate a more

deformed shape.

Figure 5.3. Measured E(2+ ) values for Se isotopes.

93
Figure 5.4. Measured E(2+ ) values for Kr isotopes.

Figure 5.5. Measured E(2+ ) values for Sr isotopes.

94
 The suggestion of N = 56 as a spherical sub shell closure comes largely from
96
the case of Zr, which has a large E(2+ ) value. The trend of the N = 56 isotones

suggests shell closures at Z = 40 and 50 but not away from those nuclei(Fig 5.7).

The suggested means for creating the N = 56 subshell closure is by the lowering of
94
the ν2d5/2 level, which can hold six neutrons. Sr nucleus (Fig 5.5).

Based on the N = 56 systematics, a spherical shape for the ground state of

94
Sr would require that Z = 34 be shell closure as well. There are predictions of

a N = 34 subshell closure around Z = 20 [?], however experimental evidence [?]

does not support such a closure. The proposed mechanism for the emergence of

this subshell closure is the lowering of the 2p1/2 below the 1f 5/2 due to a weaker

monopole interaction[46]. Even if a similar closure were to exist for Z = 34, it

seems unlikely that the π2p1/2 and π2p3/2 states would strongly interact with the

ν2d5/2 state necessary to create the N = 56 closure because of the large difference

in spin-orbit coupling ( 12 vs 52 ).

95
Figure 5.6. Measured E(2+ ) values for Zr isotopes.

Figure 5.7. Measured E(2+ ) values for N = 56 isotones.

96
5.3 R-Process Calculations

In order to investigate the impact of these measurements on the r-process, the

measured halflives and Pn values were put into a classical r-process simulation code.
56
This code starts with a Fe seed and uses the assumption of a

(n, γ) ⇋ (γ, n)

equilibrium to calculate the isotopic abundances for each element. The isotopic

abundances are determined by the Saha equation, so that the neutron number den-

sity Nn , the neutron separation energy Sn , an the temperature T largely determine

the waiting point nuclei.

 3/2
Y (Z, A + 1) G(Z, A + 1) A + 1
= Nn
Y (Z, A) G(Z, A) A
3/2
2π~2
 
Sn (Z,A+1)
× e kT
mu kT
Where G are the partition functions. The abundance for each element is then:

X
Y (Z) = Y (Z, A)
A

. Once the abundances for each isotopic chain are calculated, the flow from one

element to the next can be calculated via:

dY (Z, A) (Z−1,A)
= Y (Z − 1, A)(1 − Pn(Z−1,A) )λβ
dt
(Z−1,A+1) (Z,A)
+Y (Z − 1, A + 1)Pn(Z−1,A+1)λβ − Y (Z, A)λβ

After a process time t freeze-out occurs, and only β decays are considered, until the

nuclei have decayed back to stability. After freeze-out occurs, beta-delayed neutron

emission may occur and change the final abundance distribution.

97
5.3.1 R-Process results

A simulation using the present experimental results is shown in Fig 5.8 along

with the observed solar system r-process abundances. Also shown are the r-process

abundances for the FRDM-QRPA predictions for spherical Se nuclei. The simulation

was done with three neutron exposures of:

TABLE 5.1

ASTROPHYSICAL PARAMETERS USED FOR R-PROCESS SIMULATION

Nn Temperature (GK) Time (s) Weighting

1 × 1020 1.35 2000 9.2
5 × 1022 1.35 1800 1.0
1.5 × 1025 1.35 1650 0.75

The results show that the halflife of 90 Se has a strong effect on the abundances of

the nuclei up to the A = 130 peak. The difference between a spherical and deformed

shape for this nucleus changes the abundances by a factor of two. The measured
90
halflife of 284ms for Se produces more material for A = 90 -130 than does the

spherical prediction of 1127ms. These results agree with previous predictions of

more r-process material due to shorter beta-decay halflives[47].

Longer half-lives in a region generally result in more r-process abundance for

that region. Longer halflives at the waiting points is the reason that the nuclei at or

near closed shells (such as N = 82, 126) have larger r-process abundances. Results

were qualitatively similar for several different mass models.

From the simulations, it is clear that the Se isotopes have a larger effect on

the r-process at low neutron number densities. The exposure of Nn = 1020 was

the only one to be greatly effected by the variation in Se halflives. Because of this

98
 R-Process Abundances
Spherical vs. Deformed Se Halflives
10
Solar Abundances
Experimental Values
Spherical Values

Abundance
0.1

0.01

100
A

Figure 5.8. Results of R-Process simulation using measured half-life values (black
line) and values predicted for spherical ground state shapes (red line).

importance to lower values of Nn , the question of a weak r-process or other neutron

capture process may be of concern. It was seen that neutron number densities of

between 1018 and 1021 can be affected by the halflives in this region. At Nn = 1016

though, the change in halflives did not strongly affect the final abundances.

99
 CHAPTER 6

CONCLUSION

The goal of this work was to provide a better understanding of the mechanisms

that are responsible for creating the matter that makes up our solar system, and

the world around us. This knowledge has been sought by civilizations across the

world for thousands of years. However, it has only been within the last century

that we have approached a scientific understanding of the origin of the elements.

The connection between nuclear fusion and stellar fuel was first made by Arthur

Eddington in 1920 and extended by George Gamow, and Hans Bethe (among many

others). They realized that the energy provided by the fusion of hydrogen into

helium could provide enough energy to counteract the gravitational contraction of

stellar material. This link was the first step to explaining how heavier elements are

made.

We know that after the Big Bang mostly hydrogen and helium nuclei were

present. It was not until the first generation of stars began to convert hydrogen

into helium that the nuclear abundances changed. The p-p chains and CNO cycle

account for conversion of hydrogen into helium. Helium is converted into carbon

through the triple-alpha reaction, while oxygen and neon are created by the burn-

ing of helium and carbon. Further nuclear burning can produce nuclei as heavy as

iron. Carbon, oxygen, neon, and silicon burn until iron group nuclei are produced.

Because these nuclei have the highest binding energies per nucleon, heavier nuclei

100
are generally not produced by charged particle reactions. In some cases though, the

rapid proton capture process can create elements heavier than iron.

To make nuclei heavier than iron, reactions involving uncharged neutrons are

required. Nuclei of any Z can capture neutrons, so we know that elements beyond

iron must be created through neutron reactions. The shape of the elemental abun-

dance curve provides some clues about how different processes contribute to the

total abundances. The abundance curve is dominated by three sets of double peaks

past the iron region. These double peaks seem to indicate two separate processes

driven by similar physics. These two processes are known as the slow (s) and rapid

(r) neutron capture processes.

Currently, the reason for two sets of peaks is thought to be that they are the

s and r processes reaching neutron shell closures at different points. The s-process

moves along closer to the line of nuclear stability, and therefore has a greater number

of protons at each neutron shell closure. The r-process path is further from stability

which means that it will have fewer protons at each neutron shell closure. A nucleus

in the r-process will capture several neutrons, until the point where any additional

neutrons will be emitted as soon as they are captured. This nucleus, known as

a waiting point nucleus will beta decay to its daughter (with Z+1). Thus the β

decay rate is extremely important in determining the final abundances from the

r-process. Longer β decay halflives at particular waiting points will result in greater

nuclear abundance around those waiting points, with reduced abundance beyond.

Shorter halflives will result in a quicker r-process and greater abundance of the

heaviest nuclei. In addition, the beta-delayed neutron emission ratio Pn is also very

important to determining the final nuclear abundances from the r-process. After the

process ends and nuclei begin to beta decay back to stability, beta- delayed neutron

emission will change the mass A of nuclei.

101
 Because of the importance of β decay halflives and Pn values to the r-process,

it was the aim of this work to measure the β decay halflives Pn values of several

r-process nuclei. The isotopes chosen were just below Z = 40 near N = 60 and

include: Y , Sr, Rb, Kr, Br, Se, and As isotopes. These nuclei are interesting
100
because they lie in the r-process path and are near Zr, which is characterized by

its unexpected ground state deformation. With the addition of only a few neutrons,

Zr goes from a spherical shape to a deformed shape. This change in shape also

produces a sharp increase in the beta-decay halflife. The nuclei we sought to measure

might also have such a rapid change in shape, but a corresponding change in the
90
halflife and P n value would have consequences for the r-process. The nucleus Se
96
in particular is interesting because it has N = 56 like the spherical nucleus Zr.

There are predictions that it too might have a spherical shape and a longer half-life.

A computer simulation shows that show that if the neutron rich Se isotopes were

spherical in shape, that the r-process abundances between A = 90 − 130 would be

lowered by a factor of two.

In this work, the beta decay halflives were measured and Pn values were cal-

culated for neutron rich isotopes below Z = 40 produced at the NSCL coupled

cyclotron facility at Michigan State University. Because these nuclei are so far from

stability, they are extremely difficult to produce. The NSCL is one of only a few

facilities around the world that can produce nuclei so far from stability. To measure

the properties of these rare nuclei a system of silicon, germanium, and scintillator

detectors was needed. Determining the halflives was done by a χ2 minimization

for the nuclei with enough counts to allow binning. For the nuclei with too few

counts for this procedure to work, the Maximum Likelihood Method was used. This

method has been used in other low count experiments, including similar experiments

at the NSCL.

102
 90
Twenty-two halflives were determined in our measurments. Three nuclei, Se,
88 89
As, and As, were measured for the first time. In addition, we have confirmed

previously measured half-lives for Y, Sr, Rb, Kr, Br, Se, and As isotopes. The new

measurements were used to calculate the Pn values for 92−90 Br and 96,94 Rb shown in

Table 4-2.

Beta decay halflives are a good first indication of nuclear structure. As stated
90
above, Se is a good testing ground for the strength of the N = 56 subshell. Based

on FRDM-QRPA calculations for the halflives depending on the nuclear shape, our
90
results suggest that Se is deformed in shape. This agrees with predictions for

the ground state shape based on a Z = 34 deformed shell gap. However, this

refutes the concept of an N = 56 subshell closure below Z = 40. Ultimately, more

measurements are needed to confirm this indication.

56
Using a Fe, three different neutron exposures, and with the assumption of a

(n, γ) ⇋ (γ, n) equilibrium, r-process abundances have been calculated. As stated

above, the results show that if the neutron rich Se isotopes were spherical in shape,

that the r-process abundances between A = 90-130 would be lowered by a factor of

two. Because of their importance in neutron exposures with Nn ≤ 1021 , the isotopes

in this region may be important for the weak r-process and other proposed neutron

captures such as the LEPP. With the measurement of 90 Se in particular, it is possible

to set new limits on the end points of processes such as the alpha-rich freezeout or
90
the LEPP. The two possible shapes of Se added a factor of two uncertainty in the

production values in this region.

Many of the nuclei that were measured in this work have been measured previ-

ously, often with greater precision. The reason for this has to do with the methods

of production in this experiment versus the previous experiments. In previous work,

the method known as ion separation on line (ISOL) was used. This method pro-

103
duces nuclei far from stability as fission products from a thick target. The products

are then extracted from the target and reaccelerates them to be stopped in a detec-

tor system. ISOL works very well for some nuclei, as can be seen in the previous

measurements. This method has two large drawbacks though. The first is that

the nuclei of interest must be fission products of the target. The fragment mass

distributions for most fissioning nuclei are peaked at A = 90 and A = 140. Away

from these masses the fission yeilds drop off sharply. This means that attempting to

produce nuclei with masses much different from these values is extremely difficult.

In addition to the limits of production, ISOL must also extract the product nuclei

from the target and reaccelerate them. Because of the chemistry involved, some

isotopes are difficult to extract at all. More importantly, this process takes time,

which means that very short lived nuclei can not be extracted and reaccelerated

before they decay.

These limitations mean that, especially for r-process studies, fragment separation

is the method of the future. As can be seen from this work’s region of interest, many

of the nuclei were first produced twenty or thirty years ago, with little progress since

then. This work was the first study in this area using the method that will eventually

take us beyond even the r-process nuclei. Despite having production rates well below

what were estimated, this experiment was able to produce new nuclei beyond what

the more mature ISOL techinque has been able to. There is no inherant reason

that fragmentation cannot deliver nuclei in this region and beyond in much greater

quantities. Using the knowledge gained from this experiment, future experiments

can optimize magnet settings and make even greater discoveries. The history of

r-process experiments has been a slow and steady one. This work has delivered new

results, and has laid the foundation for future discoveries.

While we have a general idea of how the heavy elements are made, we do not

104
know the specific details in either the nuclear structure side or the astrophysics side.

There have been very few r-process nuclei produced and measured. We currently rely

on models of nuclear properties to make predictions about the r-process abundances.

Unfortunately it is not easy to extrapolate nuclear structure properties to nuclei far

from stability. New shell closures can open, while known shell closures can disappear

as extra neutrons are added. The fact that the r-process exists far from stability

makes it exciting to study. However the extreme neutron excesses of the r-process

means that it is difficult to produce and measure the properties of r-process nuclei.

The future will provide many more answers to the questions of heavy element

creation. Today, only about ten percent of the nuclei thought possible to exist

have been studied at all. Most of these unknown nuclei are extremely neutron rich.

This work has extended the understanding of a section of the neutron rich nuclei

involved in the r-process. Further study at current facilities will provide additional

understanding of these nuclei, however an upgrade in accelerator facilities is needed

to obtain experimental information about all r-process nuclei. The current plan for

the Facility for Rare Isotope Beams (FRIB) will be able to produce neutron rich

nuclei via projectile fragmentation with sufficient intensities to study nearly all the

r-process nuclei. While the future is bright for the study of the r-process, it is critical

that new facilities are built, in order to continue to answer the questions about how

the elements around us are created.

105
 BIBLIOGRAPHY

[1] W.A. Fowler, E.M. Burbidge, G. R. Burbidge, and F. Hoyle Synthesis of the
Elements in Stars. Rev. Mod. Phys., 29(4):547–650, 1957.
[2] National Research Council of the National Academies Connecting Quarks with
the Cosmos: Eleven Science Questions for the New Century The National
Academies Press, Washington, D.C., 2003.
[3] G. Lemaitre Expansion of the Universe. Monthly Notices of the Royal Astro-
nomical Society, 91:490-501, 1931.
[4] E. Hubble A Relation Between Distance and Radial Velocity Among Extra-
galactic Nebulae. PNAS, 15(3):168-173, 1929.
[5] A. A. Penzias and R. W. Wilson A Measurement of Excess Antenna Temperature
at 4080 Mc/s. ApJ., 142(1):419, 1965.
[6] Claus E. Rolfs and William S. Rodney Cauldrons in the Cosmos. The University
of Chicago Press, Chicago, Illinois, 1988.
[7] Aage Bohr and Ben R. Mottelson Nuclear Structure W.A. Benjamin, Reading,
Massachusetts, 1975.
[8] M. Arnould and S. Goriely The P-Process of Stellar Nucleosynthesis. Phys.
Rep., 384(1):1-84, 2003.
[9] M. Busso, R. Gallino, and G.J. Wasserburg Nucleosynthesis in Asymptotic Giant
Branch Stars. Ann. Rev. Astron. Astrophys., 37(1):239-309, 1999.
[10] F Kappeler, H Beer, and K Wisshak S-process Nucleosynthesis-Nuclear Physics
and the Classical Model. Rep. Prog. Phys., 52: 945-1013, 1989.
[11] S. Goriely Uncertainties in the Solar System R-abundance Distribution. Astron.
Astophys., 342:881-891, 1999.
[12] S.G. Ryan, J.E. Norris, and T.C. Beers Extremely Metal-poor Stars. II. El-
emental Abundances and the Early Chemical Enrichment of the Galaxy. ApJ.,
471(1):254-278, 1996.
[13] C. Sneden et. al The Extremely Metal-poor, Neutron Capture α rich Star CS
22892-052: A Comprehensive Abundance Analysis. ApJ., 591(2):936-953, 2003.
[14] F. Montes, et. al Nucleosynthesis in the Early Galaxy. ApJ., 671(2):1685-1695,
2007.

106
[15] G.J. Wasserburg, M. Busso, and R. Gallino Abundances of Actinides and
Short-lived Nonactinides in the Interstellar Medium: Diverse Supernova Sources
for the r-Processes. ApJ., 466(2):109-113, 1996.
[16] D.D. Clayton, W.A Fowler, T.E Hull, and B.A Zimmerman Neutron Capture
Chains in Heavy Element Synthesis. Annals of Physics, 12(3):331-408, 1961.
[17] P.A. Seeger, W.A. Fowler, and D.D. Clayton Nucleosynthesis of Heavy Ele-
ments by Neutron Capture. ApJS., 11(1):121, 1965.
[18] K.L. Kratz Nucear Physics Constraints to Bring the Astrophysical R-Process
to the Waiting Point. Reviews in Modern Astronomy, 1(1):184-209, 1988.
[19] A.G.W. Cameron, J.J. Cowan, H.V. Klapdor, J. Metzinger, T. Oda, and J.W.
Truran Steady Flow Approximations to the Helium R-Process. ApSS., 91(1):221-
234, 1982.
[20] S.E. Woosley, and E. Baron The Collapse of White Dwarfs to Neutron Stars.
ApJ., 391(1):228-235, 1992.
[21] B.S. Meyer et al. R-Process Nucleosynthesis in the High-Entropy Supernova
Bubble. ApJ., 399(2):656-664, 1992.
[22] Maria Goeppert Mayer and J. Hans Jensen Elementary Theory of Nuclear Shell
Structure John Wiley and Sons, New York, 1955.
[23] R.F. Casten Nuclear Structure From a Simple Perspective Oxford University
Press, New York, 2000.
[24] A. Navin et al. Phys. Rev. Lett., 85:266, 2000.
[25] B. Pfeiffer, K.-L. Kratz, F.-K. Thielemann, Z. Phys., A357:235, 1997.
[26] K. Farouqi et al. Astrophysical conditions for an r-process in the high-entropy
wind scenario of type II supernovae. Nucl. Phys. A, 758:631-634, 2005.
[27] I. Dillmann et al. Phys.Rev.Lett., 91:162503, 2003.
[28] K. Heyde Nucl. Phys. A, 507:149-154, 1990.
[29] K.S. Krane Introductory Nuclear Physics Wiley, New York, 1988.
[30] R.F. Casten Nucl. Phys. A, 443:1-28, 1985.
[31] P. Federman and S. Pittel Phys. Rev. C, 20:820, 1979.
[32] A. Aprahamian et al. Phys. Lett., 140B:22, 1984.
[33] U.C. Bergmann et al. Nucl. Phys. A, 714:21, 2003.
[34] CF Von Weizscker Theory of Nuclear Masses Z. Phys., 1935.
[35] H. Bethe and R. Bacher Stationary States of Nuclei Rev. Mod. Phys., 8:165,
1936.

107
[36] W.D.Myers and W.J.Swiatecki Nuclear Masses and Deformations Nucl. Phys.,
81:1, 1966.
[37] D. Lunney, J.M. Pearson, and C. Thibault Recent Trends in the Determination
of Nuclear Masses Rev. Mod. Phys., 75, 2003.
[38] P. Moller et al. Nuclear Ground-State Masses and Deformations ADNDT,
59:185, 1995.
[39] A.K. Dutta et al. Nucl. Phys. A, 458:77,1986.
[40] J.M. Pearson and S. Goriely Nuclear Mass Formulas for Astrophysics Nucl.
Phys. A, 777:623,2006.
[41] A. Aprahamian et al. Hyperfine Interactions, 132:417424, 2001.
[42] T. Takahashi, M. Yamada and T. Kondoh ADNDT, 12:101,1973.
[43] J. Halbleib and R. Sorensen Nucl. Phys. A, 98:542,1967.
[44] P. Moller, J.R. Nix, and K.L.Kratz ADNDT, 66:131, 1997.
[45] J. Skalski, S. Mizutori and W. Nazarewicz Nuclear Physics A, 617:282, 1997.
[46] R.V.F. Janssens et al. Phys. Lett. B, 546:55, 2002.
[47] P.T. Hosmer et al. Phys. Rev. Lett., 94:112501, 2005.
[48] A1900 Nucl. Instrum. Meth. B, 204:90, 2003.
[49] J.I. Prisciandaro, A.C. Morton and P.F. Mantica Nucl. Instrum. Meth. A,
505:140, 2002.
[50] B.Pfeiffer, K.L. Kratz, and P.Moller Status of delayed neutron precursor data:
half-lives and neutron emission probabilities Prog. Nucl. Energy, 41:39-69, 2002.
Cowan J.J. Cowan, F.K. Thielemann, and J.W. Truran Physics Reports, 208:267,
1991.

108