ELECTRONIC POPULATION MANIPULATION IN AN RFQ ION TRAP FOR THE

STUDY OF RARE TRANSITION METAL ISOTOPES

By

Jeremy D. Lantis

A DISSERTATION

Submitted to

Michigan State University

in partial fulfillment of the requirements

for the degree of

Chemistry - Doctor of Philosophy

2020

ABSTRACT

ELECTRONIC POPULATION MANIPULATION IN AN RFQ

ION TRAP FOR THE STUDY OF RARE TRANSITION

METAL ISOTOPES

By

Jeremy D. Lantis

Laser spectroscopy techniques have been extensively used to determine fundamental prop-

erties of the atomic nuclei across broad swaths of the chart of the nuclides, including both

stable and unstable(radioactive) nuclei. However, the transition metal region has proven dif-

ficult to study with laser spectroscopy due to the electronic structure of the transition metals

and short half-lives of radioactive nuclei. Experimental access to the transition metal region

for laser spectroscopy studies would contribute to the understanding of nuclear structure

and provide important data for nuclear stockpile stewardship. In order to gain experimental

access to the transition metal region, an optical pumping technique can be performed to ma-

nipulate the electronic populations of ions. Ions in low-energy electronic states are excited to

a state that preferentially decays to one accessible for precision laser spectroscopy measure-

ments. This work presents the results of commissioning experiments of an optical pumping

system at the BEam COoling and LAser(BECOLA) spectroscopy facility, located at the Na-

tional Superconducting Cylotron Laboratory. An optical pumping simulation was developed

to predict the gain in sensitivity for the first- and second-row transition metals from optical

pumping and guide the commissioning experiments. The results of the simulation show that

most transition metal elements can benefit from the optical pumping technique. New in-

strumentation, including a pulsed laser system and a free-space light transport system, were

installed at BECOLA to perform the optical pumping technique on ions in a Radio-Frequency

Quadrupole(RFQ) cooler/buncher ion trap. The commissioning experiments were performed

with a beam of stable Zr+, produced by an offline plasma discharge sputter ion source and

delivered to the BECOLA RFQ ion trap. Two optical pumping schemes were investigated,

exciting ions from the 4d25s 4F3/2 and 4d25s4F5/2 states. The manipulation of the electronic

populations was probed through changes in the fluorescence signal in subsequent collinear

laser spectroscopy measurements through the excitation of ions in the 4d25s 4F3/2 state.

The electronic populations of ions in the RFQ ion trap were successfully manipulated while

the ions are quickly rethermalized to a room temperature distribution by collisions with a

He buffer gas. The experimental results were in good agreement with the optical pumping

simulation and based on that validation, BECOLA is well situated to study rare isotopes of

essentially all transition metal elements at the Facility for Rare Isotope Beams(FRIB).

ACKNOWLEDGMENTS

I could not have finished this project or dissertation without the help of countless indi-

viduals.

I first want to thank Kei Minamisono, for your mentorship and guidance through my

time in graduate school. You pushed me to do my best and helped me develop as a scientist.

I also want to thank Paul Mantica, for your advice and support through my time in graduate

school, and for introducing me to the world of nuclear chemistry. In addition, I would like

to thank Greg Severin and Oscar Naviliat-Cuncic for serving on my guidance committee.

I also need to acknowledge other members of the BECOLA group, past and present. First,

I would like to recognize David Garand, who started the pulsed laser project at BECOLA and

contributed substantially to the optical pumping simulation. I also want to thank Kristian

K¨onig, Andrew Miller, Robert Powel, and Skyy Pineda for their help with the commissioning

experiments and their friendship. It is always easier to come to the lab knowing friendly

faces are waiting.

Finally, I want to thank my wife and family. Mom and Dad, thank you for your encour-

agement through every level of school. You always supported me, and I wouldn’t be here

without you or your sacrifices. Every year I appreciate more and more the opportunities

that you gave me. Lydia, you are my best friend and I need to thank you for everything

that you have done to support me through my time in graduate school. I could not have

finished this dissertation without you.

iv

TABLE OF CONTENTS

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

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

xi

KEY TO ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxii

Chapter 1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.1 Laser Spectroscopy of Transition Metal Isotopes . . . . . . . . . . . . . . . .
1.1.1 First-Row Transition Metals . . . . . . . . . . . . . . . . . . . . . . .
1.1.2
Second Row Transition Metals . . . . . . . . . . . . . . . . . . . . . .
1.2 Stockpile Stewardship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2.1 Neutron Capture Cross Section . . . . . . . . . . . . . . . . . . . . .
1.2.2
Separation of Nuclear Waste . . . . . . . . . . . . . . . . . . . . . . .
1.3 Laser Spectroscopy of Radioactive Nuclei . . . . . . . . . . . . . . . . . . . .
Issues Facing Laser Spectroscopy of Transition Metals . . . . . . . . .
1.3.1.1 Production of Rare Transition Metal Isotopes . . . . . . . .
1.3.1.2
Inaccessibility of Ground State Electronic Transitions . . . .
1.3.1.3 Charge Exchange Reactions for Laser Spectroscopy . . . . .
1.4 Electronic Population Manipulation . . . . . . . . . . . . . . . . . . . . . . .
1.4.1 Mechanism of Laser Excitation . . . . . . . . . . . . . . . . . . . . .
1.4.2 Optical Pumping for Electronic Population Manipulation . . . . . . .
1.5 Organization of Dissertation . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.3.1

2.1.1

Chapter 2 Hyperfine Interactions . . . . . . . . . . . . . . . . . . . . . . . . .
2.1 Atomic Structure and Nuclear Properties . . . . . . . . . . . . . . . . . . . .
Isotope Shift and Charge Radius
. . . . . . . . . . . . . . . . . . . .
2.1.1.1 Mass Shift . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.1.2 Field Shift . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
2.2 King Plot Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.1 King Plot Analysis of Zr+ . . . . . . . . . . . . . . . . . . . . . . . .

2.1.2 Hyperfine Splitting and Nuclear Moments

3.1.1
3.1.2

Chapter 3 Optical Pumping Simulation . . . . . . . . . . . . . . . . . . . . .
3.1 Design Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3 Level Solution of Rate Equations
. . . . . . . . . . . . . . . . . . .
Simulation Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2 Assumptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.1 Conditions Inside the Ion Trap . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.2 Laser Conditions
3.3 Operation of the Optical Pumping Simulation . . . . . . . . . . . . . . . . .

v

1
1
2
4
6
7
9
10
11
12
12
13
15
15
19
20

22
22
22
23
24
25
27
28

31
31
33
38
39
39
44
45

3.4 Results for Transition Metals

3.3.1 Choice of CLS Step and Output . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
3.4.1 First-Row Transition Metals . . . . . . . . . . . . . . . . . . . . . . .
3.4.2
Second-Row Transition Metals . . . . . . . . . . . . . . . . . . . . . .

45
46
46
48

Chapter 4 Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1 Experimental Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.1 Collinear Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.1.1 Doppler Shift Effect
. . . . . . . . . . . . . . . . . . . . . .
4.1.1.2 Beam Bunching . . . . . . . . . . . . . . . . . . . . . . . . .
4.2 BECOLA Facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.1 Production of Ion Beams . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.2 Radiofrequency Ion Trap . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.3 CLS Beamline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.4 Continuous Wave Laser System . . . . . . . . . . . . . . . . . . . . .
4.3 New Developments to BECOLA Facility . . . . . . . . . . . . . . . . . . . .
4.3.1 Pulsed Laser System . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.2 Free-Space Light Transport System . . . . . . . . . . . . . . . . . . .
4.3.2.1 Pulsed Laser Fiber Optic Cable System . . . . . . . . . . .
4.3.2.2 Free-Space Light Transport
. . . . . . . . . . . . . . . . . .
4.3.2.3 Optical Layout . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.2.4
Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Safety System . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.2.5
4.4 Experimental Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.1 Choice of Zr Ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.2 Transition Schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.2.1 Depopulation of the Ground State
. . . . . . . . . . . . . .
4.4.2.2 Augmentation of the Ground State . . . . . . . . . . . . . .
4.4.3 Measurement Method . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.3.1 Timing Sequence . . . . . . . . . . . . . . . . . . . . . . . .
4.4.3.2 Determination of Optical Pumping Effect

50
50
50
51
57
58
59
62
67
69
72
72
80
81
83
85
89
91
92
92
94
96
98
98
98
. . . . . . . . . . 100

Chapter 5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.1 Properties of Optical Pumping in the RFQ Ion Trap . . . . . . . . . . . . . . 102
5.1.1 Laser Beam Size
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.1.2 Laser Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
5.1.3 Linewidth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.2 Depopulation of the Zr+ Ground State . . . . . . . . . . . . . . . . . . . . . 110
5.3 Augmentation of the Zr+ Ground State . . . . . . . . . . . . . . . . . . . . . 113
5.4 Quenching of Optical Pumping . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.4.1 Extrapolation of the Optical Pumping Effect . . . . . . . . . . . . . . 121

Chapter 6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Improvements to the BECOLA Facility . . . . . . . . . . . . . . . . . . . . . 125
6.1.1 Free-Space Light Transport System . . . . . . . . . . . . . . . . . . . 125

6.1

vi

6.1.2 Characteristics of the Pulsed Laser System . . . . . . . . . . . . . . . 129
6.2 Optical Pumping in the RFQ Ion Trap . . . . . . . . . . . . . . . . . . . . . 131
6.2.1
Ion Bunch Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
6.2.2 Laser Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
6.2.3 Linewidth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
6.3 Manipulation of Electronic Populations . . . . . . . . . . . . . . . . . . . . . 135
6.3.1 Quenching of Optically Pumped Populations . . . . . . . . . . . . . . 136
6.4 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.4.1 Optical Pumping Simulation Accuracy . . . . . . . . . . . . . . . . . 140
6.4.2 FRIB Reach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
6.4.3 Future Laser Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

Chapter 7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

APPENDIX AOptical Pumping Schemes of First and Second-Row

Transition Metals . . . . . . . . . . . . . . . . . . . . . . . 148
APPENDIX BLogic of Shutter Control System . . . . . . . . . . . . . 164

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166

vii

LIST OF TABLES

Table 1.1:

Radioactive isotopes of the first-row transition metal elements that
have had the charge radii determined experimentally. Few studies
have been performed far from stability.
. . . . . . . . . . . . . . . .

Table 1.2:

Radioactive isotopes of the second-row transition metal elements for
which the charge radii have been determined. . . . . . . . . . . . . .

Table 1.3:

The 7 Long-Lived Fission Products of 235U and their respective yields
and half-lives. Methods to separate these isotopes from nuclear waste
are desired to reduce the threat of environmental contamination and
ensure proper disposal.
. . . . . . . . . . . . . . . . . . . . . . . . .

3

6

9

Table 2.1:

Isotope shifts and charge radii of the stable isotopes of Zr . . . . . .

29

Table 3.1:

Typical operating parameters of the BECOLA RFQ ion trap, used
to calculate the effective ion temperature. . . . . . . . . . . . . . . .

42

Table 3.2:

Table 3.3:

Percentage of total electronic population available for CLS measure-
ments of the first-row transition metals, with and without optical
pumping. Elements that require charge exchange reactions are indi-
cated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Percentage of total electronic population available for CLS measure-
ments of the second-row transition metals, with and without optical
pumping. Elements that require charge exchange reactions are indi-
cated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Table 4.1:

Pulsed laser wavelengths and powers tested during installation of the
. . . . . . . . . . . . . . .
Cobra-Stretch laser system at BECOLA.

Table 4.2:

Standard deviation of the relative laser light position at four loca-
tions. The data was collected at four locations, on the laser table,
at the entrance of the steel pipe connecting the BECOLA laser and
experimental areas, at the exit of the steel pipe, and in the lower laser
safety box before injection into the beamline. . . . . . . . . . . . . .

47

49

77

91

Table 4.3:

Natural abundances of the stable isotopes of Zr.

. . . . . . . . . . .

92

viii

Table 4.4:

Decay branches and transition probabilities from the 4d25p 4G5/2
state of Zr+, comparing theoretical calculations from the Kurucz
database [75] with experimental results from [103]. The theoretical
predictions included decays to three states that were not observed
. . . . . . . . . . . . . . . . .
experimentally, and are marked as ’-’.

95

Table 5.1:

Effective absorption linewidth of the 4F3/2 to 4G5/2 transition in Zr+
at different pulsed laser powers.

. . . . . . . . . . . . . . . . . . . . 110

Table 5.2:

Table 5.3:

Results of the depopulation of the ground state of Zr+ using the 4F3/2
to 4G5/2. Results reported as the optical pumping effect.

. . . . . . 113

Time constant of the quenching of the optically pumped electronic
populations by collisional decay with the He buffer gas, across a range
of pressures. The time constants of both the depopulating optical
pumping scheme and populating optical pumping scheme are shown.
If the time constant was not measured at a specific pressure, ’-’ is
shown.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

Table 6.1:

Einstein coefficient A limits based on saturation of the transition
using similar conditions as the commissioning experiments, namely a
laser beam size of 0.27 cm2. The wavelength ranges are those that
were tested during installation of the pulsed laser system.

. . . . . . 133

Table 6.2:

Experimental reach of the first- and second-row transition metals at
BECOLA during the FRIB era. BECOLA will be in a unique position
to study wide chains of virtually every element in the region.

. . . . 142

Table A.1:

Low-energy populated states of Sc+ . . . . . . . . . . . . . . . . . . 149

Table A.2: Optical Pumping Schemes of Sc+ . . . . . . . . . . . . . . . . . . . 150

Table A.3:

Low-energy populated states of Ti+ . . . . . . . . . . . . . . . . . . 151

Table A.4: Optical Pumping Schemes of Ti+ . . . . . . . . . . . . . . . . . . . 152

Table A.5:

Low-energy populated states of V+ . . . . . . . . . . . . . . . . . . 153

Table A.6: Optical Pumping Schemes of V+ . . . . . . . . . . . . . . . . . . . . 153

Table A.7: Optical Pumping Schemes of Cr+ . . . . . . . . . . . . . . . . . . . 154

Table A.8: Optical Pumping Schemes of Mn+ . . . . . . . . . . . . . . . . . . . 155

ix

Table A.9:

Low-energy populated states of Fe+ . . . . . . . . . . . . . . . . . . 155

Table A.10: Optical Pumping Schemes of Fe+ . . . . . . . . . . . . . . . . . . . 156

Table A.11: Low-energy populated states of Co+ . . . . . . . . . . . . . . . . . . 156

Table A.12: Optical Pumping Schemes of Co+ . . . . . . . . . . . . . . . . . . . 157

Table A.13: Low-energy populated states of Y+ . . . . . . . . . . . . . . . . . . 157

Table A.14: Optical Pumping Schemes of Y+ . . . . . . . . . . . . . . . . . . . . 158

Table A.15: Low-energy populated states of Zr+ . . . . . . . . . . . . . . . . . . 159

Table A.16: Optical Pumping Schemes of Zr+ . . . . . . . . . . . . . . . . . . . 159

Table A.17: Low-energy populated states of Nb+ . . . . . . . . . . . . . . . . . . 160

Table A.18: Optical Pumping Schemes of Nb+ . . . . . . . . . . . . . . . . . . . 161

Table A.19: Optical Pumping Schemes of Mo+ . . . . . . . . . . . . . . . . . . . 161

Table A.20: Optical Pumping Schemes of Ru+ . . . . . . . . . . . . . . . . . . . 162

x

LIST OF FIGURES

Figure 1.1: Charge radii of the first-row transition metals as a function of neutron
number. No experimental results have been published to date for
radioactive isotopes of V, Cr, Co. References for each chain can be
found in Table 1.1.
. . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 1.2: Charge radii of the Ca region, modified from [9]. The Ca region
features the absence of a kink signature of the N = 20 shell closure,
as is seen at N = 28. Measurements of Sc and Ti beyond the shell
closure will help illuminate the evolution of the nuclear structure in
the region and how the additional proton in the f7/2 shell impacts
the N = 20 shell closure.
. . . . . . . . . . . . . . . . . . . . . . . .

Figure 1.3: B(E2) values of Cr,Fe,Ge,Se, and Kr isotopes as a function of neutron
number, data taken from [24]. A sharp increase in collectivity is
observed surrounding the N = 40 subshell closure, indicating an onset
of deformation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 1.4: Charge radii of the second-row transition metals as a function of
neutron number. No experimental results have been published to
date for radioactive isotopes of Y,Zr,Nb,Mo, and Cd. References for
each chain can be found in Table 1.2.
. . . . . . . . . . . . . . . . .

Figure 1.5: Neutron capture network of Zr, from [39]. During events with high
neutron flux, a sample rapidly undergoes neutron capture and beta
decay to create many different species. Accurate modeling of the
neutron flux requires the neutron capture cross sections of all species
to be well known, but for many radioactive nuclei no measurements
have been performed. For the Zr network, arrows in purple indicate
. . . . . .
cross sections that have been experimentally determined.

2

4

5

5

8

Figure 1.6:

Laser spectroscopy across the chart of the nuclides, modified from [4].
Nuclides marked in black are stable isotopes and isotopes in red are
radioactive isotopes that have been studied using laser spectroscopy.
The nuclides marked in green have been studied at the BECOLA
facility.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11

xi

Figure 1.7:

Simulated electronic populations after charge exchange neutralization
of a 30 keV Ni+ beam with Na vapor, data taken from [58]. The ver-
tical dashed line indicates the entry energy of the electron. The levels
surrounding the entry energy are initially populated and marked in
red, but quickly decay to low-energy metastable states. The final
population after decay is shown in blue. . . . . . . . . . . . . . . . .

Figure 1.8: Mechanisms for the exchange of electrons between levels. Through
interaction with light, stimulated excitation can excite an electron
from a ground state to a higher-lying state or cause stimulated decay
to the ground state, producing an additional photon. Additionally,
electrons in the excited state can spontaneously decay and emit a
photon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 1.9: Population manipulation of electronic populations to improve the sen-
sitivity of CLS measurements at BECOLA. Pulsed laser light is used
to optically pump electrons from a populated low-lying state to an
excited state that preferentially decays to a state accessible to the
. . . . . . . .
BECOLA CW laser systems for CLS measurements.

14

16

19

Figure 2.1: Hyperfine splitting of the levels associated with the 4F3/2 to 4G5/2
transition of 91Zr. 91Zr has a nuclear spin of I = 5/2 and results in
9 peaks in the hyperfine spectra. The magnetic dipole and electric
quadrupole moments can be extracted from the widths of the splittings. 26

Figure 2.2: Hyperfine spectra of the stable isotopes of Zr as a function of the
isotope shift relative to 90Zr. The spectrum of 91Zr features hyperfine
splitting due to the non-zero spin of the nucleus.
. . . . . . . . . . .

28

Figure 2.3: King Plot of the stable isotopes of Zr. The measured modified iso-
tope shift is plotted as a function of the modified charge radius. The
x-axis is shifted by 1080 u fm2 to reduce the error correlation between
the slope and the y-intercept of the fit. The charge radii of the sta-
ble Zr isotopes are known, and the charge radii of newly measured
radioactive isotopes can be determined through the linear regression
. . . . . . . . . . . . . . . . . . . . . . . .
line, shown as a red line.

Figure 3.1:

Simplification of the electronic states involved in the optical pump-
ing process. As ions from the ground state are optically pumped to
low-lying metastable states, the ions will no longer interact with the
laser light. Spontaneous decay from the low-lying metastable states
is negligible over the time scales involved in the optical pumping pro-
cess, and the states can be pooled together to simplify calculation.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

30

32

xii

Figure 3.2: Proportion of ion population in the ground, metastable, and excited
states as a function of time. The pulsed laser temporal width is 6
ns, marked by the vertical blue bar. While the ions are irradiated
by the laser, the ground state population is optically pumped to the
low-lying metastable states. After the laser pulse, the ions remaining
in the excited state exponentially decay to the ground and metastable
states.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 3.3: Proportion of electronic population in the low-lying states of Zr+ as
a function of temperature. At low temperatures, almost all ions exist
in the ground state, with other low-lying states becoming populated
. . . . . . . . . . . . . . . . . . . . . . . .
as the temperature rises.

Figure 3.4:

Simulated radial distribution of ions in the BECOLA RFQ ion trap,
calculated using the 3DCylPIC software. The green circle describes
the 95% inclusion radius of the radial distribution, approximately 1.4
mm in radius. The orange circle describes the extent of the trapping
region defined by the RFQ rods.
. . . . . . . . . . . . . . . . . . .

Figure 3.5: Results of the optical pumping simulation for the first-row transition
metals compared to traditional CLS measurements. Most elements
will experience a substantial gain in sensitivity through optical pump-
ing, although some elements require very deep UV light that cannot
be produced at BECOLA and therefore have no viable optical pump-
ing schemes.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 3.6: Results of the optical pumping simulation for the second-row tran-
sition metals compared to traditional CLS measurements. Most ele-
ments will experience a significant gain in sensitivity through optical
pumping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 4.1: Magnitude of the mean velocity as a function of time in the ion trap
and He buffer gas pressure. The lines are to guide the eyes. The
He buffer gas serves to cool the beam and limits the velocity of the
ions as they move through the RF field. Doppler broadening of the
absorption linewidth is reduced as the ions slow.
. . . . . . . . . . .

38

40

43

47

48

53

xiii

Figure 4.2: Doppler broadened absorption linewidth from RF motion as a func-
tion of He buffer gas pressure, calculated from the 3DCylPIC software
[80]. The optical pumping process was performed in an RFQ ion trap
and the motion of the ions from the RF field was used to determine
the absorption linewidth. As the He buffer gas pressure increases,
the neutral He atoms serve to limit the velocity of the trapped ions
and the absorption linewidth reaches a plateau. The orange line in-
dicates the expected Doppler broadened linewidth in air at 300 K.
The BECOLA RFQ ion trap is operated at 80 mTorr of He buffer
gas pressure and the Doppler broadened absorption linewidth from
the RF motion is comparable to the expected linewidth in air.
. . .

Figure 4.3: The relationship between kinetic energy and velocity. Three sets
of ions with the same energy spread will have velocity spreads that
become progressively narrower as the ion energy increases. CLS mea-
surements take advantage of the kinetic compression of the linewidth
to remove broadening effects and perform precise measurements of
the transition frequency.
. . . . . . . . . . . . . . . . . . . . . . . .

Figure 4.4: Hyperfine spectra of 90Zr measured using CLS. The width of the
. . . . . . . . . . . . .

hyperfine spectra is approximately 60 MHz.

54

56

57

Figure 4.5: Diagram of BECOLA Beamline

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

59

Figure 4.6: Cross section of the BECOLA PIG ion source. The PIG ion source
can generate ion beams of many metallic elements via plasma sput-
tering, at rates up to 10s of nA.
. . . . . . . . . . . . . . . . . . . .

Figure 4.7: Photograph of the BECOLA PIG ion source. The PIG is attached to
. . . . . . . . . . . . .

the BECOLA beamline with a 6 inch flange.

Figure 4.8: Cathode and Anticathode composed of Zr. The Zr cathodes quickly
eroded over the course of the commissioning experiments. As the
erosion progressed, the hole in the anticathode expanded and the
cathode developed a large hole as well, which prevented operation of
the PIG ion source.
. . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 4.9:

Schematic of the BECOLA RFQ ion trap, taken from [78]. The RFQ
ion trap features three sections separated by differential pumping
barriers. Ions are cooled through collisions with a He buffer gas in
the cooling section, and are trapped in preparation for bunched beam
CLS measurements in the trapping region.
. . . . . . . . . . . . . .

60

61

62

63

xiv

Figure 4.10: Diagram of the crosscut geometry of the BECOLA RFQ rods, taken
from [78]. The view of the rods is longitudinal, with the beam direc-
tion coming out of the page. The crosscut design produces a linear
drag field across the length of the RFQ ion trap with only two power
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
supplies.

Figure 4.11: Diagram of the segmented electrodes of the BECOLA RFQ trapping
region, taken from [78]. The diagram is a crosscut of the RFQ ion
trap, with the beam traveling to the right. Voltages are applied to
the segmented electrodes to produce a potential well to trap the ions
and allow them to be released in bunches for laser spectroscopy mea-
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
surements.

Figure 4.12: Measured time spectra of 90Zr with different numbers of ions trapped.
The time spectrum on the right of the figure featured the fewest
number of ions per bunch and had a FWHM of approximately 1.5 µs.
The time spectrum on the right contained approximately 10 times
more ions per bunch, resulting in a significantly broader FWHM in
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
time.

Figure 4.13: Diagram of the photon detection region in the BECOLA CLS beam-
line, taken from [89].The photon detection region is located immedi-
ately after the charge exchange cell. The new design of the photon
collection system is placed directly after the old design. The photon
detection region serves to collect and focus fluorescence to photomul-
. . . . . . . . . . . . . . . . . .
tiplier tubes for CLS measurements.

Figure 4.14: The BECOLA laser room is located approximately 15 m away from
the BECOLA experimental area. The CW laser light is transported
to the CLS beamline with the use of fiber optic cable. The output of
the fiber optic cable is located in a laser safety box, where the laser
light can be prepared for injection into the CLS beamline. . . . . . .

Figure 4.15: Laser path through the Cobra-Stretch dye laser system from [97].
Laser light from the INDI pump laser excites the lasing medium in
the 20 mm oscillator cell and the 40 mm amplifier cell. The output
wavelength is selected by the grating position, which is amplified in
the amplifier cell, before entering the internal SHG unit where the
frequency of light is doubled. . . . . . . . . . . . . . . . . . . . . . .

Figure 4.16: Amplifier dye cell of the Cobra-Stretch pulsed laser system. The dye
acts as a lasing medium and is circulated through the cell to prevent
laser-induced breakdown of the dye molecules and extend its lifetime.
The amplifier cell is 20 mm in length.
. . . . . . . . . . . . . . . . .

64

65

67

69

71

73

74

xv

Figure 4.17: Conversion efficiency as a function of wavelength for different dyes
across the range of 540-910 nm. Dyes in the range of 540-910 nm are
pumped by second harmonic light at 532 nm produced by the INDI
pulsed laser. Figure taken from Sirah Cobra-Stretch User’s Manual
[97]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 4.18: Conversion efficiency as a function of wavelength for different dyes
across the range of 370-580 nm. Dyes in the range of 370-580 nm are
pumped by third harmonic light at 355 nm produced by the INDI
pulsed laser. Figure taken from Sirah Cobra-Stretch User’s Manual
[97]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 4.19: The width of the temporal pulse produced by the Cobra-Stretch laser
system. The laser light was attenuated using a neutral density filter
before being captured on a photodiode.
. . . . . . . . . . . . . . . .

Figure 4.20: Laser beam profiles of the INDI pump laser(Top) and Cobra-Stretch
dye laser(Bottom) as seen on burn paper. The laser profile of the
INDI system is circular and the power is evenly distributed, while
the laser beam profile of the Cobra-Stretch is elliptical and the power
is not evenly distributed.
. . . . . . . . . . . . . . . . . . . . . . . .

Figure 4.21: Power distribution of pulsed laser light produced by the Cobra-Strech
system. Multiple modes of light are produced by the laser, resulting
in a profile with three local maxima. . . . . . . . . . . . . . . . . . .

Figure 4.22: The BECOLA laser room is located approximately 15 meters away
from the BECOLA experimental area. The free-space light transport
system uses mirrors to deliver light to the experimental area with
minimal loss of power. . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 4.23: New laser safety boxes constructed in the BECOLA experimental
area. Laser light from the BECOLA laser room is sent through a
10-m long steel pipe to the laser boxes. The upper box contains two
breadboards to direct the light through steel pipes to the lower boxes
for preparation before injection into the CLS beamline and RFQ ion
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
trap.

75

76

78

79

80

81

84

Figure 4.24: Optical layout on the pulsed laser table. The yellow light shows the
path of the pulsed laser light. The light is translated from the out-
put of the Cobra-Stretch system to a mirror using a pair of periscope
mirrors. The light passes through two lenses, which are used as a tele-
scoping system to slowly focus the light through the light transport
system. The light is then reflected upward to an inverted breadboard. 86

xvi

Figure 4.25: Optical layout of the inverted breadboard in the BECOLA laser area.
The light is translated to the breadboard by a pair of mirrors used as
a periscope. The light is reflected by two mirrors which can be used
for alignment through a 10 m long pipe to the BECOLA experimental
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
area.

Figure 4.26: Optical layout of the pulsed laser safety box used to prepare the
pulsed laser light for injection into the RFQ ion trap. The light is
translated down to the safety box from the 10 m long pipe using a
pair of mirrors used as a periscope. The light is reflected by two
mirrors which can be used to align the laser light through the RFQ
ion trap. The diagram below shows the laser light path in yellow and
the reflected light path in green, mirrors are blue ovals, the backstop
is black, and the wedged window is a red oval. The red lines indicate
where the optical elements are located in the photograph.
. . . . . .

Figure 4.27: Center of laser beam tracked across 25 minutes at four different loca-
tions. Each set of data is relative to the average position of the laser
light. The data was collected at four locations, on the laser table,
at the entrance of the steel pipe connecting the BECOLA laser and
experimental areas, at the exit of the steel pipe, and in the lower laser
safety box before injection into the beamline. Although movement of
the beam is small in the BECOLA laser room, the laser light moves
on the order of millimeters by the end of the free-space light transport
system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

87

88

90

Figure 4.28: Partial electronic states of Zr+ [76]. The hyperfine levels of each

electronic state are not shown, as the level density of Zr+ is very high. 93

Figure 4.29: Optical pumping scheme for the depopulation of the ground state of
Zr+. Ions from the ground state are excited to the 4d25p 4G5/2 state.
Approximately 32% of ions decay to metastable states that are dark
. . . . .
to subsequent CLS measurements, reducing the sensitivity.

Figure 4.30: Optical pumping scheme for the augmentation of the ground state
of Zr+.
Ions from the first excited state are excited to the 4d25p
4G5/2 state. Approximately 66% of ions decay to the ground state,
increasing the sensitivity of subsequent CLS measurements.
. . . .

Figure 4.31: Optical pumping scheme for the augmentation of the ground state
of Zr+.
Ions from the first excited state are excited to the 4d25p
4G5/2 state. Approximately 66% of ions decay to the ground state,
. . . .
increasing the sensitivity of subsequent CLS measurements.

96

97

99

xvii

Figure 4.32: Example of optical pumping data, pumping ions from the ground
state with the 4F3/2 to 4G5/2 transition of Zr+. As the ion bunch
passes the PMTs at approximately 30 µs, the fluorescent photons are
captured. The photon counts were measured as a function of time
and shutter status. A time gate was applied to remove background.
The time gate is shown as vertical lines around the peak.

. . . . . . 100

Figure 5.1: Optical pumping effect as a function of pulsed laser beam position.
The pulsed laser light was moved off of the trapped ions and scanned
in one direction across the trapped ions to approximate its size. . . . 104

Figure 5.2:

Spectra of fluorescent photon counts as a function of time, with and
without optical pumping. The measurement was performed with a
tightly focused laser beam and high laser power and the optical pump-
ing effect is much larger than expected.

. . . . . . . . . . . . . . . . 105

Figure 5.3: Optical pumping effect as a function of the laser power for the 4F3/2
to 4G5/2 transition in Zr+. The transition quickly saturates, but
experiences a second sharp increase in the optical pumping effect
above 40 mW.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

Figure 5.4: Optical pumping effect as a function of the laser power for the 4F3/2
to 4G5/2 transition in Zr+. A fast transition, the optical pumping
effect quickly saturates.

. . . . . . . . . . . . . . . . . . . . . . . . 108

Figure 5.5: Effective absorption linewidth of the 4F3/2 to 4G5/2 transition in Zr+
at different laser powers. As the laser power increases, the transition
remains saturated even as the laser moves off resonance, resulting in
a flat-top spectrum.

. . . . . . . . . . . . . . . . . . . . . . . . . . . 109

Figure 5.6: Depopulation of the ground state of Zr+ via optical pumping. The
fluorescent photons are measured using CLS as a function of time.
The ion beam is kept in resonance with the probe laser and alternating
measurements were made with and without the pump laser. . . . . . 111

Figure 5.7: Depopulation of the ground state of Zr+ via optical pumping and

probing the increased electronic population of the first excited state. 112

Figure 5.8: Augmented population of the ground state of Zr+ via optical pump-
ing. The fluorescent photons are measured using CLS as a function
of time. The ion beam is kept in resonance with the probe laser and
alternating measurements were made with and without the pump laser.114

xviii

Figure 5.9: Optical pumping effect of Zr+ as a function of number of laser pulses
used to irradiate the trapped ions. The blue bars show the expected
optical pumping effect based on the optical pumping simulation, with
the orange bars showing the experimental results. The experimental
results show no enhancement of the optical pumping effect, provid-
ing evidence that collisions with the He buffer gas re-thermalize the
trapped ions to a room temperature distribution. . . . . . . . . . . . 115

Figure 5.10: Quenching of the optically pumped electronic populations in Zr+.
The ions held in the RFQ ion trap were excited with a pulsed laser
and held for increasing lengths of time.
Interactions with the He
buffer gas used in the RFQ ion trap leads to cooling of the trapped
ions back to a room temperature Boltzmann distribution.

. . . . . . 116

Figure 5.11: Quenching of the optically pumped electronic populations in the de-

populating pumping scheme of Zr+.

. . . . . . . . . . . . . . . . . . 116

Figure 5.12: Optical pumping effect as a function of the release time of the trapped
ions after irradiation by the pulsed laser light across a range of pres-
sures. The quenching of the optically pumped electronic populations
takes a longer time at lower pressures. . . . . . . . . . . . . . . . . . 118

Figure 5.13: Optical pumping effect as a function of the release time of the trapped
ions after irradiation by the pulsed laser light across a range of pres-
sures. The quenching of the optically pumped electronic populations
takes a longer time at lower pressures. The curves of 50 mTorr, 80
mTorr, and 120 mTorr feature a zoom in of the data.

. . . . . . . . 119

Figure 5.14: Optical pumping effect as a function of release delay of the ions after
they are irradiated by pulsed laser light, comparing corrected and
uncorrected data points for additional ions that enter the trap after
the laser pulse. The overall correction is small.

. . . . . . . . . . . . 120

Figure 5.15: Collisional decay rate between He and the optically pumped electronic
populations of Zr+. The decay rate follows a linear relationship with
the pressure of the He buffer gas. . . . . . . . . . . . . . . . . . . . . 121

Figure 5.16: Comparison of the extrapolated optical pumping effect with the ex-
perimentally observed optical pumping effect. The results of the de-
populating optical pumping are shown on top and the results of the
populating optical pumping scheme are shown on bottom. The ex-
trapolated data is based on the time constant of the quenching of the
optically pumped electronic populations.

. . . . . . . . . . . . . . . 122

xix

Figure 5.17: Extrapolated optical pumping effect of the trapped ions immediately
after irradiation by the pulsed laser, as a function of pressure. The
simulated values are shown as orange and blue bands respectively.
The experimental data agrees with the simulated value of the optical
pumping effect using experimental branching ratios from Ljung et al.
[103]

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Figure 5.18: Extrapolated optical pumping effect of the trapped ions immediately
after irradiation by the pulsed laser, as a function of pressure. The
simulated optical pumping effect using the Kurucz and Ljung branch-
ing ratios are shown as green and orange bands respectively. The
experimental results agree well with the simulated value based on the
Ljung simulation value.

. . . . . . . . . . . . . . . . . . . . . . . . . 124

Figure 6.1: Frequency doubled CW laser light produced by a Sirah WaveTrain
system. The frequency doubled light was measured approximately
2 m from the WaveTrain system. No lenses were used to focus or
collimate the light. The doubled laser light features higher order
modes of light seen as halo structures. The higher order light can
propagate through the free-space light transport system and induce
high stray light background.

. . . . . . . . . . . . . . . . . . . . . . 127

Figure 6.2:

Simulated optical pumping effect as a function of the temporal width
of the laser pulse. Assuming that the transition remains saturated,
the ground state of Zr+ can be almost completely depopulated with
a laser pulse of approximately 100 ns in width.

. . . . . . . . . . . . 130

Figure 6.3: Power broadening of spectral lines. As the laser power increases, the
wings of the spectral profile raise and can excite transitions far from
the center of the center frequency.

. . . . . . . . . . . . . . . . . . . 134

Figure 6.4: Comparison of the experimental optical pumping effect as a function
of pressure with the simulated optical pumping effect with collisional
rethermalization added. The simulation shows good agreement with
the experimental results and demonstrates that the rethermalization
process can accounted for in simulation. . . . . . . . . . . . . . . . . 139

Figure 6.5: Chart of the nuclides zoomed in on the first- and second-row tran-
sition metals. Isotopes in black are stable, isotopes in blue are ra-
dioactive and have previously been studied using laser spectroscopy,
isotopes in green will be accessible for studies at BECOLA, and iso-
topes in pink have previously been studied at BECOLA. . . . . . . . 141

xx

Figure 6.6:

Simulated optical pumping effect of the depopulating optical pumping
scheme in Zr+. The 10 kHz laser system is able to deliver more
laser pulses to the ion bunch, but more laser pulses are required to
depopulate the ground state due to the lower energy density of the
laser light.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

Figure B.1: Schematic diagram of the connectivity of the electronic modules used
to control the physical shutter for measurement of the optical pump-
ing effect.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

xxi

KEY TO ABBREVIATIONS

ˆ AVLIS - Atomic Vapor Laser Ionization Separation

ˆ BECOLA - Beam Cooling and Laser Spectroscopy

ˆ CLS - Collinear Laser Spectroscopy

ˆ CTBT - Comprehensive Test Ban Treaty

ˆ CW - Continuous Wave

ˆ DMSO - Dimethyl Sulfoxide

ˆ FIFO - Fan-in/Fan-out

ˆ FRIB - Facility for Rare Isotope Beams

ˆ FSR - Free Spectral Range

ˆ ISOL - Isotope Separator On-Line

ˆ LLFP - Long-Lived Fission Products

ˆ NIST - National Institute for Standards and Technology

ˆ NSM - Normal Mass Shift

ˆ NSCL - National Superconducting Cyclotron Laboratory

ˆ PIG - Penning Ionization Gauge

ˆ PMT - Photomultiplier Tube

ˆ RFQ - RadioFrequency Quadrupole

ˆ RILIS - Resonant Ionization Laser Ion Source

ˆ SDM - Scale Down Module

ˆ SMS - Specific Mass Shift

ˆ SNR - Signal-to-Noise

xxii

Chapter 1

Introduction

The BEam COoling and LAser (BECOLA) spectroscopy facility [1, 2] at the National Super-

conducting Cyclotron Laboratory(NSCL) at Michigan State University uses precision laser

spectroscopy techniques to study the nuclear structure of rare isotopes. Rare isotopes are

those that do not naturally occur, and are instead produced at particle accelerator facilities.

Although laser spectroscopy is a powerful tool [3, 4], the study of transition metal isotopes

with this technique has remained difficult, as will be discussed in Section 1.3. The goal of

the experiments described here was to develop an optical pumping system to manipulate

the electronic populations of transition metal ions in a Radio-Frequency Quadrupole(RFQ)

ion trap, enabling the study of transition metal elements that were previously inaccessible

to laser spectroscopy measurements. In the following chapter, the motivation to study rare

transition metal isotopes for nuclear structure and stockpile stewardship studies is intro-

duced, as well as the issues that make laser spectroscopy of these elements difficult.

1.1 Laser Spectroscopy of Transition Metal Isotopes

One of the fundamental properties of the atomic nucleus is its size. Laser spectroscopy

measurements can provide information about the charge radius of the nucleus, or the extent

of the proton distribution [5, 6]. Measurements of the charge radius can provide critical

information about the nuclear strong force and serve as a benchmark for nuclear theory

1

[3, 7, 4]. The focus of this work is on the first- and second-row transition metals, and the

following sections will discuss interesting cases for laser spectroscopy studies from the region.

1.1.1 First-Row Transition Metals

Laser spectroscopy measurements have been extended to most elements of the first-row

transition metals, with the exception of V, Cr, and Co, and few measurements have been

performed on radioactive isotopes of the transition metals, especially far from stability [8].

The charge radii trends of the first-row transition metals can be seen in Figure 1.1. The

Figure 1.1: Charge radii of the first-row transition metals as a function of neutron number.
No experimental results have been published to date for radioactive isotopes of V, Cr, Co.
References for each chain can be found in Table 1.1.

charge radii chains in Figure 1.1 include the stable isotopes of each element. The radioactive

isotopes that have been studied for each element are shown in Table 1.1.

From a nuclear structure perspective, the first-row transition metals contain many in-

teresting cases for future study at BECOLA, with two examples to follow. The lighest

2

Table 1.1: Radioactive isotopes of the first-row transition metal elements that have had the
charge radii determined experimentally. Few studies have been performed far from stability.

Element Mass Number Reference

Sc
Ti
Mn
Fe
Ni
Cu

Zn

42-44,46

44

50-54, 56-64

52,53

68

58-62,64,66-75

76-78

62,63,65,69,71-80

[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]

transition metals, Sc and Ti, are of interest towards the N = 20 shell closure. The charge

radii of the region around N = 20 has been extensively measured, as can be seen in Figure

1.2. Typically a kink in the charge radius is observed across a shell closure, as seen at the

N = 28 shell closure, but such a kink is absent or weak for lighter elements such as Ca [17],

K [18, 19], and Ar [20] across the N = 20 shell closure. The absence of the shell closure

signature has been attributed to a balance of a monopole proton-core polarization below

N = 20 and an onset of deformation above N = 20 [18]. 40Ca is a doubly-magic nucleus,

and experimental properties should exhibit trends that indicate the enhanced stability that

it brings. Determination of the charge radii of Sc and Ti across the N = 20 shell closure will

illustrate if the absence of the N = 20 shell closure persists beyond Ca.

Another important example from the first row of transition metals is the case of the

Fe region around the subshell closure at N = 40 [21]. Studies of neutron-rich Fe isotopes

have shown an increase in collectivity surrounding N = 40 manifested by a sharp increase

in the B(E2) [22, 23], shown in Figure 1.3. Typically the B(E2) value reaches a local

minimum at a shell closure, but a local maximum is observed across the potential N =

40 subshell closure, suggesting that the subshell closure is weak or absent. As the B(E2)

3

Figure 1.2: Charge radii of the Ca region, modified from [9]. The Ca region features the
absence of a kink signature of the N = 20 shell closure, as is seen at N = 28. Measurements of
Sc and Ti beyond the shell closure will help illuminate the evolution of the nuclear structure
in the region and how the additional proton in the f7/2 shell impacts the N = 20 shell
closure.

is directly related to deformation of the nucleus, charge radii measurements will provide

complementary information about the static and dynamic deformation of the nucleus and

evolution of the shell structure of this region.

1.1.2 Second Row Transition Metals

In common with the first-row transition metals, few laser spectroscopy studies of radioactive

isotopes of the second row have been performed [8]. The charge radii trends of the second-

row transition metals are shown in Figure 1.4. Although no experimental results have been

4

Figure 1.3: B(E2) values of Cr,Fe,Ge,Se, and Kr isotopes as a function of neutron number,
data taken from [24]. A sharp increase in collectivity is observed surrounding the N = 40
subshell closure, indicating an onset of deformation.

Figure 1.4: Charge radii of the second-row transition metals as a function of neutron number.
No experimental results have been published to date for radioactive isotopes of Y,Zr,Nb,Mo,
and Cd. References for each chain can be found in Table 1.2.

published for Tc, Ru, Rh, or Pd, other elements in the second row have been extensively

studied, complete with measurements far from stability. The radioactive isotopes of the

5

Table 1.2: Radioactive isotopes of the second-row transition metal elements for which the
charge radii have been determined.

Element

Y
Zr

Nb
Mo
Ag

Cd

Mass Number

86-88,90-102

87-89
97-102

90-02,99,101,103

90,91,102-108

97-101
103-105

100-105,107,109,115,117-130

Reference

[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]

second-row transition metals for which the charge radius has been determined are listed in

Table 1.2 with references.

There has been interest in possible shape coexistence in nuclei around 80Zr. 80Zr is

an important isotope for the astrophysical rapid proton (rp) process, as the reaction flow

is controlled by its beta-plus decay [33]. Nuclear structure calculations have predicted the

coexistence of spherical, prolate, and oblate shapes in Zr as well as in surrounding nuclei

and isotopes [34], with calculations predicting different deformations [35]. As 80Zr is near

the proton dripline, experimental data are sparse. Laser spectroscopy measurements of

the region will be possible in the FRIB era, and will provide critical information for the

theoretical description of the region.

1.2 Stockpile Stewardship

The transition metal region is of vital importance for nuclear stockpile stewardship [36, 37,

38, 39]. Since the development of a nuclear weapons program in the 1940’s, the United

States government has been tasked with maintaining a safe and effective nuclear weapon

stockpile. In 1996, the United States became party to the Comprehensive Nuclear Test-Ban

6

Treaty (CTBT), banning all nuclear explosions in all environments. The age of the nuclear

stockpile grows each year, and many weapons have exceeded their designed lifetimes [40]. To

remain in compliance with the CTBT, verification of the readiness of the nuclear stockpile

is left to complex modeling and simulation [40]. The parameters of these codes rely heavily

on properties of the nucleus, especially for nuclei that are far from stability. For accurate

predictions of the conditions within the weapons and in explosions, the properties of nuclei

far from stability need to be known with high precision or predicted with high accuracy.

Laser spectroscopy studies of the transition metal region will benefit the stewardship

program through reinforcement of nuclear theory as well as more direct methods such as

separation of nuclear waste and will be discussed in the following sections.

1.2.1 Neutron Capture Cross Section

One important parameter for the modeling of nuclear explosions is the neutron flux. One

method for measuring the neutron flux is through the use of tracer elements, typically tran-

sition metal isotopes such as 90Zr [38]. During detonation, the nucleus of a tracer element

may undergo a series of multiple neutron captures and decays, allowing for the production

of many different species. The relationship between the possible species that are produced

can be described using what is known as a neutron capture network, as shown in Figure 1.5

from [39]. Accurate calculations of the neutron flux are reliant on knowledge of the neutron

capture cross section for each isotope that can be produced. Unfortunately, the neutron

capture cross section is not known for many radioactive species. It is not possible to produce

targets of short-lived species, preventing direct measurement of the neutron capture cross

section.

There are ongoing efforts to calculate the cross section indirectly, using methods such

7

Figure 1.5: Neutron capture network of Zr, from [39]. During events with high neutron
flux, a sample rapidly undergoes neutron capture and beta decay to create many different
species. Accurate modeling of the neutron flux requires the neutron capture cross sections
of all species to be well known, but for many radioactive nuclei no measurements have been
performed. For the Zr network, arrows in purple indicate cross sections that have been
experimentally determined.

as surrogate nuclear reactions [41] and measurements of the nuclear level density and γ-ray

strength function [42]. Without these methods, the current stockpile stewardship codes rely

on theoretical calculations as inputs for their simulations. However, the uncertainty in the

accuracy of the theoretical results for the less-studied, short-lived isotopes in the neutron

capture network leads to large uncertainties in the neutron flux determination. Through

laser spectroscopy measurements, the nuclear size and shape can be deduced with high

precision, providing important data for constraining the neutron capture cross section and

strengthening the framework behind their calculations.

8

Table 1.3: The 7 Long-Lived Fission Products of 235U and their respective yields and half-
lives. Methods to separate these isotopes from nuclear waste are desired to reduce the threat
of environmental contamination and ensure proper disposal.

Isotope Fission Yield (%)

99Tc
126Sn
79Se
93Zr
135Cs
107Pd
129I

6.13
0.11%
0.04%
5.45%
6.9%
1.25

0.84 %

t1/2 (My)

0.21
0.23
0.32
1.53
2.3
6.5
15.7

1.2.2 Separation of Nuclear Waste

Another crucial field for the US nuclear stockpile stewardship program is the development

of methods for isotope-selective separation. The Atomic Vapor Laser Ionization Separa-

tion(AVLIS) technique [43] has been successfully utilized to efficiently separate isotopes.

The material of interest is first heated to a high temperature (>2000 K) and vaporized to a

neutral beam. Laser light tuned to an optical transition of the desired species is employed to

resonantly ionize the neutral beam and the produced cations can be separated and collected

using an electric field. The AVLIS technique has a high degree of selectivity provided that

the isotope shift between the target and other isotopes is larger than the bandwidth of the

ionization laser. The AVLIS technique has been used to both enrich 235U for nuclear fuel as

well as to separate the long-lived fission product 107Pd [44]. There is particular interest in

separation of the 7 long-lived fission products(LLFP) from 235U, detailed in Table 1.3. The

LLFPs make up a significant proportion of nuclear waste and as their long half-lives make

them a threat to environmental contamination, there is interest in ways to safely separate

[45, 44], or transmute them [46, 47]. Of the LLFPs, three transition metal isotopes are

represented, 99Tc,93Zr, and 107Pd.

9

Due to a lack of knowledge of the hyperfine spectrum of 93Zr and the small isotope

shift between the stable 90Zr and 91Zr isotopes, other laser methods are being developed for

separation [45]. At BECOLA, precision measurements of the hyperfine spectra of 93Zr can

be performed and assist in these efforts.

1.3 Laser Spectroscopy of Radioactive Nuclei

Until the development of the laser in 1960 [48], studies of the isotope shift were restricted

to emission spectra. As laser technology developed, it became possible to perform laser

spectroscopy to measure the isotope shift of atomic beams with high precision [49] using

Collinear Laser Spectroscopy (CLS), which will be given a more in-depth description in

Chapter 3. Since the development of CLS, fundamental properties of the ground and isomeric

states of the nucleus, such as the spin, differential mean-square charge radii, and nuclear

moments, have been deduced for radioactive nuclei across broad swaths of the chart of the

nuclides [4], as seen in Figure 1.6. As particle accelerators for low-energy nuclear physics have

grown more powerful, the reach of laser spectroscopy experiments has also been extended

to isotopes far from stability. At next generation facilities such as the Facility for Rare

Isotope Beams(FRIB), laser spectroscopy studies will even be able to produce isotopes at

the nucleon driplines at sufficient rates for laser spectroscopy studies. The CLS technique

measures the fluorescence of the ion or atom, which occurs on timescales faster than the

decay of the nucleus.

Although FRIB will bring higher beam rates of rare isotopes, there are still technical

challenges to study certain regions of the chart of the nuclides, such as the transition metals.

Key nuclei far from stability in the chart of the nuclides, such as 78Ni with a rate of 7

10

Figure 1.6: Laser spectroscopy across the chart of the nuclides, modified from [4]. Nuclides
marked in black are stable isotopes and isotopes in red are radioactive isotopes that have
been studied using laser spectroscopy. The nuclides marked in green have been studied at
the BECOLA facility.

per second [50] at the ultimate performance (200 MeV, 400 kW) of FRIB, will still be

challenging to measure, making sensitivity of the laser spectroscopic measurement of the

utmost importance.

1.3.1 Issues Facing Laser Spectroscopy of Transition Metals

As previously discussed, there have been relatively few laser spectroscopy measurements of

transition metals far from stability. Both the electronic structure of the transition metals

and the nuclear properties make them difficult to study, and highly sensitive techniques are

necessary.

11

1.3.1.1 Production of Rare Transition Metal Isotopes

One of the significant impediments to the study of rare transition metal isotopes is their

production. A majority of past laser spectroscopy experiments on radioactive nuclei were

performed at Isotope Separation On-Line (ISOL) facilities [51, 52], where a high-energy beam

of protons impinges on a thick production target. The production target is heated to a very

high temperature (>2000 K), producing a vapor of the different species in the target. The

resulting vapor is ionized and mass separated for experimental measurements.

The ISOL production method faces difficulty producing beams of transition metal iso-

topes with short half-lives [53]. Many transition metals are refractory elements with high

melting points, which in combination with chemistry within the target, leads to long release

times. When the release time from the target approaches the nuclear lifetime, it is difficult

to realize production rates sufficient for laser spectroscopy measurements.

In comparison to the ISOL production method, the NSCL and FRIB use an in-flight

production method. The ion beam produced by the accelerator is impinged on a thin target

and the nuclei that result from the projectile-fragmentation reactions are separated in flight

through the A1900 separator [54]. The in-flight production method has a distinct advantage

for refractory elements because the separation process is not dependent on any chemical

properties, and subsequently allows for higher beam rates delivered to experimental systems

for the transition metals. However, in general the production yields are smaller using the

in-flight production method.

1.3.1.2 Inaccessibility of Ground State Electronic Transitions

An additional constraint on the study of transition metal isotopes with laser spectroscopy

comes from their electronic configurations. Electronic transitions can only be probed from

12

populated states, often the ground state. Ground state laser transitions for transition metal

elements typically require deep UV laser light in the range of 100-300 nm. The production

of 3rd and 4th harmonic light is necessary to reach the deep UV regime. Harmonic light

generation is an inefficient, nonlinear optical process and scales with the square of the power

of the fundamental light [55]. Typically, laser spectroscopy studies for the determination of

the charge radius use narrow linewidth, continuous wave(CW) laser systems with relatively

low output power. 2nd harmonic light generation is a well established technique with CW

laser systems, but systems capable of producing 4th harmonic light suitable for isotope shift

measurements have only recently been developed [32].

1.3.1.3 Charge Exchange Reactions for Laser Spectroscopy

If there is no suitable ground-state transition in the ionic state, a beam of neutral atoms

can instead be produced through charge exchange reactions [56, 57]. One method for atomic

charge exchange passes the ion beam of interest through a chamber containing a cloud of

alkali vapor, where the ions are neutralized.

Many transition metal elements have accessible transitions in the neutral state and charge

exchange reactions have allowed the study of radioactive transition metal isotopes with Fe

[12], Ni [13], and Cu [14] to name just a few examples.

Although charge exchange reactions have allowed the study of transition metal isotopes

that would otherwise remain out of the reach of current systems, the atomic beams that are

produced suffer from reduced sensitivity due to fractionation of the electronic population.

The fractionation process has been described in detail before [58], but a short explanation

follows. During the neutralization process, the incoming electrons populate the electronic

levels surrounding the entry energy. The high-energy states quickly decay to low-energy

13

metastable states.

In elements with high level densities, such as the d-block transition

metals, there are typically many different metastable states that can be populated. The

high level density is important to populate many highly-excited electronic states that will

decay to low-energy states, but can also lead to fractionation of the electronic population.

After taking into account the losses due to neutralization efficiency and the fractionation,

only a few percentage of the total ion beam can be probed.

To illustrate, Figure 1.7 (taken from [58]) shows the simulated electronic populations of Ni

after charge exchange with a Na atom. After the charge exchange process, the ground state

of Ni is only populated at 15% of the total atomic population. Typically, the neutralization

efficiency of the charge exchange cell is approximately 50%, limiting the overall efficiency of

converting the ionic beam to an atomic beam to 7.5%.

Figure 1.7: Simulated electronic populations after charge exchange neutralization of a 30
keV Ni+ beam with Na vapor, data taken from [58]. The vertical dashed line indicates the
entry energy of the electron. The levels surrounding the entry energy are initially populated
and marked in red, but quickly decay to low-energy metastable states. The final population
after decay is shown in blue.

14

1.4 Electronic Population Manipulation

One method to overcome the issues caused by the unfavorable electronic structure of the

transition metals is by redistributing the electronic populations to states favorable for laser

spectroscopy. The electronic population manipulation can be achieved using a technique

known as optical pumping. Although a simple concept, the discovery of optical pumping

resulted in the Nobel Prize for Alfred Kastler in 1966 and has been instrumental for impor-

tant scientific developments, such as the laser. Optical pumping has been used to produce

polarized beams of rare nuclei for measurements of nuclear electromagnetic moments us-

ing β-NMR [59] and the angular correlation parameters in β+ decay of 37K [60] with the

included references providing examples of experiments that have been performed with po-

larized beams. A more detailed description of the optical pumping process will be given in

Chapter 3, but a short description follows.

1.4.1 Mechanism of Laser Excitation

It is useful to begin with an ideal two level system to introduce how electrons are excited,

introducing rate equations and saturation of the excited state population.

There are three processes responsible for the movement of electrons between energy levels

during the laser excitation and can be used to determine the rate of change of population

for each state involved. These processes are stimulated excitation, stimulated decay, and

spontaneous decay, and are illustrated in Figure 1.8.

Through stimulated excitation, an electron is excited from a ground state, g, to a higher

energy state, e by absorbing a photon. The time dependent change in the electronic popu-

lation of state g, represented by Ng, due to absorption of a photon with energy hν is given

15

Figure 1.8: Mechanisms for the exchange of electrons between levels. Through interaction
with light, stimulated excitation can excite an electron from a ground state to a higher-
lying state or cause stimulated decay to the ground state, producing an additional photon.
Additionally, electrons in the excited state can spontaneously decay and emit a photon.

as

Ng = −Bgeρ(ν)Ng

d
dt

(1.1)

where Bge is the probability of stimulated excitation and called the Einstein coefficient of

stimulated excitation, and ρ(ν) is the energy density of the light at frequency ν. In addition

to stimulating the excitation of electrons, the applied radiation field can perturb the upper

state e, and cause a stimulated emission from state e to the ground state g, resulting in

a second photon of frequency ν. The rate of change of the population of state e due to

stimulated emission is given as

Ne = −Begρ(ν)Ne

d
dt

(1.2)

where Beg is the probability of the stimulated emission and called the Einstein coefficient of

stimulated emission. The upper and lower B coefficients are related to each other by

Beg =

gg
ge

Bge

16

(1.3)

where gg and ge are the degeneracies of the two levels. The upper B coefficient can be related

to the A Einstein coefficient, which is the Einstein coefficient of spontaneous emission, by

Beg =

πc3
¯hλ3 Aeg

(1.4)

Finally, an electron in state e can spontaneously decay to state g in the absence of any

applied radiation. The rate of change of the excited state e due to spontaneous decay is

given as

Ne = −AegNe

d
dt

(1.5)

where Aeg is the Einstein coefficient of spontaneous emission. Aeg is directly related to the

lifetime of the excited state, τe, by

τe =

1
Aeg

(1.6)

By adding the absorption and emission processes given in the above equations, the total

rate of population change in states g and e can be given by

Ng = Begρ(ν)Ne + AeNe − Bgeρ(ν)Ng

d
dt

and

Ne = −Begρ(ν)Ne − AeNe + Bgeρ(ν)Ng

d
dt

(1.7)

(1.8)

While the external radiation field from the light is applied to the electronic system, the

populations of states g and e will change until the competing transitions reach equilibrium

17

and the rate of change reaches 0.

d
dt

Ng =

d
dt

Ne = 0

(1.9)

Substitutiong the rate equations from Equations 1.2 and 1.5, the ratio of the populations of

the two states at equilibrium can be determined as follows.

Bgeρ(ν)Ng = Begρ(ν)Ne + AeNe

(1.10)

Rearranging Equation 1.10 gives,

Ne
Ng

=

Bgeρ

Begρ + Aeg

At laser power densities where Begρ >> Aeg, Equation 1.11 can be simplified to

Ne
Ng

=

Bge
Beg

=

ge
gg

(1.11)

(1.12)

And the populations at equilibrium will be a ratio of the degeneracies of the state.

In an ideal two state system, the population of the ground state cannot exceed 50% of

the total electronic population due to stimulated emission. Optical pumping occurs when

electrons in the excited state can decay to more than one low-energy electronic state. Elec-

trons that decay to other electronic states become ”dark”, as they are no longer resonant

with the laser light. Electrons in dark states will not experience stimulated emission and

can accumulate until the entire population exists in ”dark” states.

18

1.4.2 Optical Pumping for Electronic Population Manipulation

The goal of this project was to develop an optical pumping system to improve the sensitivity

of laser fluorescence measurements for transition metal isotopes. Taking advantage of the op-

tical pumping phenomenon, electronic populations can be pumped from inaccessible ground

states to low-energy metastable states that are accessible to narrow-linewidth CW lasers for

precision measurements. For some elements that do have accessible ground state transitions,

electronic populations from other low-energy states can be pumped to the ground state to

improve the sensitivity. An example of an optical pumping scheme is shown in Figure 1.9

Figure 1.9: Population manipulation of electronic populations to improve the sensitivity of
CLS measurements at BECOLA. Pulsed laser light is used to optically pump electrons from
a populated low-lying state to an excited state that preferentially decays to a state accessible
to the BECOLA CW laser systems for CLS measurements.

Optical pumping of transition metal isotopes has been performed at other laser spec-

troscopy facilities, for Nb [28], Y [61], and Mn [62] isotopes. The optical pumping technique

is performed in an RFQ cooler/buncher ion trap, where the ions can be held for extended

19

periods of time, allowing for long interaction times between the laser light and the ions, up

to ms. After the manipulation of the electronic populations, the ions can be released and

probed using CLS.

The pumping step is performed using high-powered pulsed laser light, as 3rd and 4th

harmonic light is more readily available due to higher energy density, allowing access to

ground state transitions.

Combining the increase in sensitivity that results from optical pumping with the increase

in beam rates available at FRIB, the BECOLA facility will have the ability to study transition

metal isotopes far from stability.

1.5 Organization of Dissertation

In this dissertation, the development of an optical pumping system to manipulate the elec-

tronic populations of transition metal isotopes at BECOLA is described. The optical pump-

ing technique increases the sensitivity of laser spectroscopy measurements at BECOLA and

will allow for the charge radii and nuclear moments of rare transition metal isotopes to be

determined through the hyperfine spectra. The interactions between the nucleus and elec-

tron that allow laser spectroscopy measurements to determine fundamental properties of the

nucleus are described in Chapter 2. Results of a new King Plot Analysis of Zr+ are described.

An optical pumping simulation was developed to guide experimental efforts and is described

in detail in Chapter 3, as well as results of the simulation and potential for transition metal

isotope experiments. The experimental method of the commissioning experiments of the

optical pumping system are described in Chapter 4. The experimental techniques used to

perform fluorescent laser spectroscopy at BECOLA as well as the techniques used to per-

20

form the electronic population manipulation are described. An introduction to Collinear

Laser Spectroscopy (CLS) is given as well as a description of the optical pumping technique.

Features of the BECOLA facility are discussed as well as two new upgrades, a pulsed laser

system and free-space light transport system. The results of the commissioning experiments

are featured in Chapter 5. Chapter 6 is a discussion of the experimental results. Finally,

Chapter 7 is a summary of the thesis as well as discussion of the outlook for future radioactive

beam experiments using the optical pumping system and beyond.

21

Chapter 2

Hyperfine Interactions

In this chapter, the interactions between the nucleus and electron are described, as well as

how these interactions can be used to determine fundamental properties of the nucleus.

2.1 Atomic Structure and Nuclear Properties

Various techniques have been established to determine the charge radius of the nucleus, such

as electron scattering [63], charge changing cross-section measurements [64], as well as the

isotope shift of K X-rays [65] and muonic atoms [66]. An additional technique, utilized at

the BECOLA facility, is the measurement of the isotope shift in optical transitions.

The isotope shift in optical transitions arises from the hyperfine interaction between the

nucleus and the valence electrons. From the center-of-gravity isotope shift between two

isotopes of the same element, the differential mean-square charge radius can be determined,

and from the splitting of the hyperfine spectra, the nuclear spin, nuclear magnetic dipole

moment, and electric quadrupole moment can be determined. Measurement of the hyperfine

spectra can reveal a trove of information about the nuclear structure.

2.1.1 Isotope Shift and Charge Radius

Soon after the discovery of the nucleus, experimental observations showed that changes in

the nuclear mass led to shifts in lines of molecular spectra. In the 1930s, the isotope shift

22

phenomenon was finally observed in atomic spectra [67] and although initially believed to

be a result of the change in mass between isotopes, in 1935 experimental evidence showed

that the charge distribution in the nucleus contributed to the shift as well [68]. The isotope

shift, δνA,A(cid:48)
two isotopes of masses A and A(cid:48).

, is the frequency of light necessary to excite an electronic transition between

δνA,A(cid:48)

= νA(cid:48) − νA

(2.1)

There are two contributions to the isotope shift, a mass shift due to the change in the

m , and a shift due to a change in the nuclear volume known as the

mass of the nucleus δνA(cid:48),A
field shift δνA(cid:48),A

f

δνA(cid:48),A = δνA(cid:48),A

m + δνA(cid:48),A

f

(2.2)

2.1.1.1 Mass Shift

The mass shift also has two components, the specific mass shift (SMS) and the normal mass

shift (NMS) which can be represented by constants KSM S and KN M S.

δνA(cid:48),A
m = (KSM S + KN M S)

mA(cid:48) − mA
mA(cid:48)mA

(2.3)

where mA is the mass of an isotope with mass number A and mA(cid:48) is the mass of an isotope
with mass number A(cid:48).

The two contributions to the mass shift are due to different changes in the nuclear motion

as the reduced mass of the electron-nucleus system changes. In systems with more than one

electron, the movements of the electrons are correlated which gives rise to the specific mass

shift. The constant KSM S has no analytical solution and is very difficult to calculate.

23

Currently, ab initio methods have only been able to calculate KSM S for systems up to 5

electrons [69].

In contrast, the NMS contribution is much easier to calculate. When an

electron is excited to a higher state, its kinetic energy and momentum changes and the

nuclear motion must adapt. The change due to the NMS is proportional to the reduced

mass of the system and KN M S is simplifid to meνA, where me is the mass of the electron

and νA is the transition frequency. With both contributions to the mass shift, the constants

KN M S and KSM S can be combined to one constant, K.

2.1.1.2 Field Shift

The second contribution to the isotope shift is the Field shift, also known as the volume shift.

The finite size of the nucleus has an effect on the binding energy of electrons, which have

wavefunctions that have a non-zero probability density within the nuclear volume. The shift

in the energy of an electronic state due to the influence from the nuclear volume, EF N S,

can be described as,

EF N S =

Ze2
6
0

(cid:104)r2
c(cid:105)|ψ(0)|2

(2.4)

which is proportional to the electron density at the nucleus, |ψ(0)|2, and the nuclear mean-
square charge radius (cid:104)r2
c(cid:105). The change in the nuclear volume causes a shift in the energy

of the electronic state due to the change in the electron probability density in the nuclear

volume, producing a shift in transition frequency. A s-electron with angular momentum L

= 0 has a large effect due to its finite probability at the nucleus.

δνA(cid:48),A

F

=

Ze2
6h
0

c(cid:105)(∆|ψ(0)|2)i→f
(cid:104)r2

(2.5)

24

Equation 2.5 can be simplified to show that the relationship between two isotopes with mean
square charge radii (cid:104)r2

c(cid:105)A and (cid:104)r2

c(cid:105)A(cid:48)

is

δνA(cid:48),A

F

=

Ze2
6h
0

(∆|ψ(0)|2)((cid:104)r2

c(cid:105)A(cid:48) − (cid:104)r2

c(cid:105)A) = F δ(cid:104)r2

c(cid:105)A(cid:48),A

(2.6)

where the terms outside of the differential mean-square charge radius has been reduced to

the field shift constant F . With solutions for the field and mass shifts, the total isotope shift

can be expressed as

δνA(cid:48),A = KM + F δ(cid:104)r2
c(cid:105)

where M is the reduced mass of the two isotopes with masses A and A(cid:48),

M =

mAmA(cid:48)
mA + mA(cid:48)

(2.7)

(2.8)

2.1.2 Hyperfine Splitting and Nuclear Moments

The hyperfine interaction between nuclei with non-zero spin and the electron gives rise to a

shift in the energy of an electronic state relative to the fine level given by

∆E =

X
2

Ahf +

3X(X + 1) − 4I(I + 1)J(J + 1)

8I(2I − 1)J(2J − 1)

Bhf

X = F (F + 1) − I(I + 1) − J(J + 1)

(2.9)

where I is the nuclear spin, J is the atomic spin, and F is the quantum number defined by

the sum of the atomic and nuclear spins. In Figure 2.1 it is shown the hyperfine splitting of

the levels associated with the 4F3/2 to 4G5/2 transition of 91Zr.

Measurement of the hyperfine spectrum allows the deduction of the Ahf and Bhf hyper-

25

Figure 2.1: Hyperfine splitting of the levels associated with the 4F3/2 to 4G5/2 transition
of 91Zr. 91Zr has a nuclear spin of I = 5/2 and results in 9 peaks in the hyperfine spectra.
The magnetic dipole and electric quadrupole moments can be extracted from the widths of
the splittings.

fine coupling constants of the electronic state, which contain information about the magnetic-

dipole and electric-quadrupole moments, µ and Q, of the nucleus as given by

Ahf = µ

B0
IJ

Bhf = eQVzz

26

(2.10)

Here B0 and Vzz are the magnetic field and electric field gradients that are produced at the

nucleus by the electrons.

2.2 King Plot Analysis

As previously shown, the isotope shift has two contributions from the mass shift and the

field shift, which contain constants K and F , which are element and transition dependent

and must be known to extract the differential mean-square charge radius. The K and F

factors can be determined through what is known as a King plot analysis [70] provided that

the charge radii of at least three isotopes are known.

The King plot analysis starts by modifying the isotope shifts by dividing each side of the

equation by M from Equation 2.8.

δvAA(cid:48)
modif = M−1δvAA(cid:48)

= F M−1δ(cid:104)r2

c(cid:105)AA(cid:48)

+ K

(2.11)

The modified differential provides a linear relationship between the differential mean-square

charge radius and the modified isotope shift. The data points should lie on a straight line

and through a linear regression the F factor can be determined as the slope of the line and

the K factor through the y-intercept. After deducing the atomic factors, the differential

mean-square charge radius can be extracted from the isotope shift for any isotope of the

particular element, and is indispensable for the study of radioactive isotopes.

27

2.2.1 King Plot Analysis of Zr+

The commissioning experiments of the optical pumping system were performed using a beam

of Zr+. Although previous experiments have been performed to determine the isotope shift

and differential mean-square charge radii of radioactive Zr isotopes [26, 27], these particular

experiments utilized the ground state transitions of d2s4F3/2 to d2p4F5/2 [27] and d2s4F3/2

to d2p4D3/2 [26], whereas the commissioning experiments used the 4F3/2 to 4G5/2 transition.

The4F3/2 to 4G5/2 transition had not been previously used for isotope shift measurements,

and the atomic factors F and K0 have not been determined. The hyperfine spectra of the

stable Zr isotopes were measured using CLS to determine the atomic factors. The hyperfine

spectra of the stable isotopes of Zr are shown in Figure 2.2. The isotope shift is given relative

Figure 2.2: Hyperfine spectra of the stable isotopes of Zr as a function of the isotope shift
relative to 90Zr. The spectrum of 91Zr features hyperfine splitting due to the non-zero spin
of the nucleus.

to 90Zr, the most abundant stable isotope. The isotope shifts relative to 90Zr were extracted

from the hyperfine spectra, and are summarized in Table 2.1. The charge radii of the stable

28

Table 2.1: Isotope shifts and charge radii of the stable isotopes of Zr

0

A δν/MHz
90
91
92
94
96

-199(1)
-509.63
-849.19
-1061.61

δ(cid:104)r2

c(cid:105)90,A/fm2 [71]

(cid:104)r2(cid:105)1/2/fm

0

0.128
0.309
0.534
0.708

4.269
4.284
4.305
4.331
4.351

isotopes of Zr have already been determined and values from Fricke et al. [71] were used.

A linear regression fit was used to extract the atomic factors and resulted in the values

seen in Equations 2.12 and 2.13.

F = −2.24(14) MHz/fm2

K = 749(80) GHzu

(2.12)

(2.13)

The newly determined atomic factors can be used in future experiments involving radioac-

tive Zr beams at BECOLA, but as measurements reach further from stability the impact of

the systematic error of the Atomic Factors will grow progressively larger. The correlation

between the slope, F , and K can be reduced by shifting the isotope shifts by an adjustment

factor α, and results in a significant reduction in the error correlation between the slope

and y-intercept that leads to a smaller uncertainty in the deduced differential mean-squared

charge radius. The procedure to reduce the correlation was described by York et al. [72],

and utilized in King Plot analyses of Cd [32] and Ni [73] isotopes. The adjusted King Plot

is shown in Figure 2.3. An adjustment factor of α = 1080 ufm2 was used and reduced the

correlation between the slope and y-intercept to -0.002. The slope of the regression line

remains unchanged by the correlation reduction procedure, but the Kα is modified to the

29

Figure 2.3: King Plot of the stable isotopes of Zr. The measured modified isotope shift is
plotted as a function of the modified charge radius. The x-axis is shifted by 1080 u fm2 to
reduce the error correlation between the slope and the y-intercept of the fit. The charge
radii of the stable Zr isotopes are known, and the charge radii of newly measured radioactive
isotopes can be determined through the linear regression line, shown as a red line.

value seen in Equation 2.14.

Kα = −1671(4) GHz u

(2.14)

With the adjustment factor, the error in the new y-intercept is reduced by an order of

magnitude by removing the correlation between the errors. The mass shift term loses its

physical meaning, but allows for the extraction of the charge radii of newly-measured isotopes

with reduced error.

30

Chapter 3

Optical Pumping Simulation

In this chapter, the development of an optical pumping simulation is discussed. The optical

pumping simulation was used to guide the commissioning experiments and will be used

to determine potential optical pumping schemes for future experiments at BECOLA. The

design principle of the simulation, the source of the electronic level and transition data used

to generate the pumping schemes, assumptions, and results of the simulation are discussed

here, as well as implications for future experiments in the FRIB era.

3.1 Design Principle

The core of the optical pumping simulation is a solution of the electronic transition rate

equations that can be used to model the resulting electronic populations as a radiation field

is applied to a system of ions or atoms. As discussed in Chapter 1, the solution of a two-level

system is straightforward, but becomes more complex with increasing numbers of electronic

states. As transition metals are d-block elements, they typically have high level densities

and the decay of an excited state may branch to many low-energy metastable states. The

optical pumping simulation uses a three-level solution from Zizek et al [74].

The three-level solution is used to model the optical pumping process due to the long

lifetimes of the metastable states involved, on the order of seconds. As only the ground

and excited states of the electronic system will interact directly with the radiation field, the

31

metastable states are ”dark” and they will not experience stimulated excitation or emis-

sion. The pulsed laser light has a pulse duration of approximately 6 ns, much shorter than

the lifetimes of the low-energy metastable states. As the electronic populations will only

accumulate in the metastable states over the time scale of the simulation, all of the low-

energy metastable states can be pooled together, simplifying the electronic states that need

to be considered, as illustrated in Figure 3.1. The branching ratios of the spontaneous emis-

Figure 3.1: Simplification of the electronic states involved in the optical pumping process. As
ions from the ground state are optically pumped to low-lying metastable states, the ions will
no longer interact with the laser light. Spontaneous decay from the low-lying metastable
states is negligible over the time scales involved in the optical pumping process, and the
states can be pooled together to simplify calculation.

sion to each metastable state are known, allowing the population of the pooled state to be

disentangled, and the final population distribution to be calculated.

32

3.1.1 3 Level Solution of Rate Equations

The optical pumping simulation is based on a solution to the electronic transition rate

equations of a three-level system, with ground state g, excited state e, and metastable state

m [74]. To begin, the rate equations of the populations of the three states can be written as

δN∗
g
δt
δN∗
δt
δN∗
e
δt

m

= N∗
= N∗
= N∗

e Reg + N∗
e Rem + N∗
g Rge + N∗

mRmg − N∗
g Rgm − N∗
mRme − N∗

g (Rgm + Rge)

m(Rmg + Rme)

e (Rem + Reg)

(3.1)

where Rij is the rate from state i to state j, and N∗

i

is the population density in state i.

The values of the rates of change are calculated based on the Einstein coefficients A and B

for the transition, the laser power, and collisional effects. As an example, the rate of change

from state i to state j is

Rij = Bij

E
c

+ kij

(3.2)

where E is the energy of the laser and k is the collisional excitation rate.

The rate equations can be simplified by converting the population density with respect

to the total number density, NT .

(3.3)

Nm = N∗
Ne = N∗

m/NT

e /NT

33

The simplified rate equations are then

δNm

δt
δNe
δt

= Rgm − Nm(Rmg + Rme + Rgm) + Ne(Rem − Rgm)
= Rge + Nm(Rme − Rge) − Ne(Reg + Rem + Rge)

(3.4)

The result is a system of linear equations that can be solved using determinants, where D is

the differential operator.

[(D + Rm)(D + Re) − R∗

meR∗

em]Nm =

[(D + Rm)(D + Re) − R∗

meR∗

em]Ne =

em

Rgm −R∗
D + Rm Rgm

Rge D + Re




(3.5)

−R∗

me Rge

It is assumed that the laser power as a function of time is a step fuction. The rates depend

on the populations of the electronic states, the Einstein coefficients, which are constants,

and the laser power, which is a independent parameter. If the laser power is a step function,

it is then constant as the transition is excited, and all coefficients become constants that can

be solved as

[D2 + (Rm + Re)D + (RmRe − R∗

meR∗

em)]Nm = RgmRe + R∗

emRge

[D2 + (Rm + Re)D + (RmRe − R∗

meR∗

em)]Ne = RgeRm + R∗

meRgm

34

(3.6)

(3.7)

To make things more clear later on, a series of substitutions are performed.

L = Rm + Re
P = RmRe − R∗

me

Qm = RgmRe + R∗
em Qe = RgeRm + RgmR∗
meR∗

emRge

which reduce equations 3.6 and 3.7 to,

(D2 + LD + P )Nm = Qm

(D2 + LD + P )Ne = Qe

The antiderivative of the equations are

Nm(t) = C1e−ξ1t + C2e−ξ2t
Ne(t) = C3e−ξ1t + C4e−ξ2t

where the terms in the exponents are

ξ1 =

ξ2 =

−

+

L
2
L
2

(cid:114)
(cid:114)

− P

− P

L2
4
L2
4

(3.8)

(3.9)

(3.10)

(3.11)

In these equations, the Cx terms are arbitrary constants that are related and will be derived

later. The integrals for Equations 3.6 and 3.7 are

Cm = Qm/(ξ1ξ2)

Ce = Qe/(ξ1ξ2)

35

(3.12)

(3.13)

which allow for complete solutions for the populations of states m and e as a function of

time,

Nm(t) = C1e−ξ1t + C2e−ξ2t +

Ne(t) = C3e−ξ1t + C4e−ξ2t +

Qm
ξ1ξ2

Qe
ξ1ξ2

(3.14)

(3.15)

The derivation of the Cx constants can be performed from the complete solution. By solving

Nm(t) and Ne(t) using the derivatives found in Equation 3.4, we have

C3 = C1

Rm − ξ1
R∗

em

C4 = C2

Rm − ξ2
R∗

em

(3.16)

(3.17)

With the above simplification, only two constants occur in the system of equations and they

can be found with the boundary conditions of the initial populations in states m and e.

From Equations 3.4 and 3.8, the initial populations are

Nmi = Qm/P Nei = Qe/P

(3.18)

By solving the solution with the initial conditions, the constants are found to be

C1 =

Qm/B − Nmi

ξ2 − ξ1

(Rm − ξ1)

Rm − ξ2
R∗

em

− Qe/B − Nei
ξ2 − ξ1

(Rm − ξ2)/R

(3.19)

36

C2 = −Qm/P − Nmi

ξ2 − ξ1

(Rm − ξ1) +

Qe/P − Nei

ξ2 − ξ1

(Rm − ξ2)/R

(3.20)

By providing the following substitutions for ease of reading,

K1 =

K2 =

RR =

(Rm − ξ1)
(Rm − ξ2)

Qm/P − Nmi
ξ2 − ξ1
Qe/P − Nei
ξ2 − ξ1
Rm − ξ2
R∗

em

(3.21)

The final solutions for the electronic populations can be given as

Nm(t) = (K1RR − K2/R)e−ξ1t + (−K1 + K2/R)e−ξ2t + Qm/B

(3.22)

Ne(t) = (K1RR − K2/R)

Rm − ξ1
R∗

em

e−ξ1t + (−K1 + K2/R)

Rm − ξ2
R∗

em

e−ξ2t + Qm/B (3.23)

As Nm and Ne are relative to the total population, the population of the ground state can

be described as

Ng(t) = NT − Nm(t) − Ne(t)

(3.24)

With the function describing the population of each level over time, the simulation is able

to determine the change in population between electronic states during the optical pumping

process. As an example, in Figure 3.2 is shown the proportion of ions in the ground, pooled

metastable, and excited states as a function of time for the 4F3/2 to 4G5/2 transition in Zr+,

with the assumptions that no collisional decay or excitation occurs and that the temporal

37

Figure 3.2: Proportion of ion population in the ground, metastable, and excited states as a
function of time. The pulsed laser temporal width is 6 ns, marked by the vertical blue bar.
While the ions are irradiated by the laser, the ground state population is optically pumped
to the low-lying metastable states. After the laser pulse, the ions remaining in the excited
state exponentially decay to the ground and metastable states.

shape of the laser pulse is a square well. The excited state has a lifetime of 7.1 ns and

the initial ground state population is 72.3%, following a Boltzmann distribution. The laser

light has a temporal width of 6 ns, indicated by the thin blue vertical line. As the ions are

irradiated by the laser light, the transition is quickly saturated. Ions are optically pumped

to the metastable states, and after the laser pulse the remaining ions in the excited state

quickly decay to the ground and metastable states.

3.1.2 Simulation Data

Accurate information on the electronic structure of the transition metals is vital for the op-

tical pumping simulation. The optical pumping simulation relies on information about the

electronic levels and transitions from the Kurucz [75] and National Institutes of Standards

38

and Technology (NIST) [76] spectral databases. The NIST Spectral Database provided ex-

perimentally measured atomic energy levels, while the Kurucz Database provided theoretical

values for electronic transitions and transition probabilities.

3.2 Assumptions

Several assumptions were made in performing the optical pumping simulations, involving

both the optical pumping process and conditions inside the ion trap.

3.2.1 Conditions Inside the Ion Trap

Two key assumptions were made in the optical pumping simulation about the conditions

of the ion trap. The first assumption is that the trapped ions are effectively cooled to

room temperature by collisions with the He buffer gas, which is at room temperature. The

temperature of the ion beam is important as the populations of the low-energy electronic

states are dependent on the temperature.

For a given temperature, the electronic populations of a species is described by the

Boltzmann distribution. Based on the energy of the electronic state and the temperature of

the cloud of ions, the relative populations of each level can be determined. The probability

distribution for state i is given by

pi =

e−Ei/kT

gi
Q

(3.25)

where gi is the degeneracy of the state, Ei is the energy of the electronic state, k is the

39

Boltzmann constant, and Q is the partition function,

n(cid:88)

Q =

gie−Ei/kT

(3.26)

i=1

Division by the partition function is necessary to make sure that all probabilities sum to 1.

At low temperatures, the ground state is predominantly populated for most species. As an

example, the electronic state populations of Zr+ are presented as a function of temperature

in Figure 3.3. At room temperature, the ground state contains approximately 72% of the
total population, the first excited state at 314 cm−1 contains 23% of the total population,
and the second excited state at 763 cm−1 contains approximately 3% of the total population.

As the temperature rises, even higher-energy states are populated.

Figure 3.3: Proportion of electronic population in the low-lying states of Zr+ as a function
of temperature. At low temperatures, almost all ions exist in the ground state, with other
low-lying states becoming populated as the temperature rises.

In laser spectroscopy measurements, the signal is directly proportional to the population

of the state being probed. Species with low level densities, such as Ca+, which has a first

40

excited state at 13650 cm-1, will have almost all ions in the ground state, but a species such

as V+, which has 5 states below 350 cm-1, only 33% of the total population will be in the

ground state, limiting the sensitivity of laser spectroscopy measurements.

As the ion beam passes through the cooling section of the RFQ ion trap, collisions with

the He buffer gas cool the beam to the temperature of the gas itself. The BECOLA RFQ

ion trap [77, 78] was operated at room temperature during the experiments described in this

thesis.

The temperature of the trapped ions in the trap can be calculated using the effective ion

temperature treatment, as developed by Tolmachev et al.

[79]. Through the effective ion

temperature treatment, it is assumed that ions accelerated through the RF field will move

coherently and will eventually acquire the same internal activation energy as ions outside of

any electric field. Instead, the temperature will be dictated by the temperature of the buffer

gas.

The effective temperature of the trapped ions, Tef f , is

Tef f =

mr
mg

TB +

2mr(cid:104)Ki(cid:105)
3kmi

(3.27)

where mr is the reduced mass of the ion-buffer gas system, mg is the mass of the buffer

gas, mi is the mass of the ion, TB is the temperature of the buffer gas, k is the Boltzmann
constant and (cid:104)Ki(cid:105) is the kinetic energy of the ions in the laboratory frame. The kinetic

energy of the ions can be determined through,

(cid:104)Ki(cid:105) =

mgω2r2
0V 2
DC
16000V 2
RF

41

(3.28)

Table 3.1: Typical operating parameters of the BECOLA RFQ ion trap, used to calculate
the effective ion temperature.

Parameter

ω
r0
VDC
VRF

7.5 MHz
3.5 mm

-4 V
72 V

where ω is the angular frequency of the applied RF field, r0 is the radius of the radius of the

trapping region defined by the RFQ rods, VDC is the applied DC voltage field applied to

create the trapping potential, and VRF is the peak of the applied RF voltage field. A more

detailed discussion RFQ ion trap will follow in Chapter 4, and a schematic of the trapping

region is shown in Figure 4.11. The effective temperature of the trapped ions was consistent

with room temperature using the operating parameters of the present Zr+ measurements

shown in Table 3.1.

The second assumption of the trap conditions considered in the optical pumping simula-

tion is that the trapped ions occupy a small volume within the ion trap and that the laser

light completely irradiates the trapped ions. The trapping region of the BECOLA RFQ ion

trap in the transverse direction relative to the ion beam is 7 mm in diameter, but due to

confinement through the RF field and DC potential, the trapped ions are contained within

a smaller volume. The assumption is supported by simulation results performed with the

3DCylPIC software [80], developed at the NSCL by R. Ringle to characterize space charge

and buffer gas effects of ions in different kinds of ion traps. The 3DCylPIC simulation solves

Poisson’s equation in cylindrical coordinates across a grid to calculate both the spatial and

velocity distribution of ions in a trap.

The 3DCylPIC simulation was used to estimate the distribution of ions across the RFQ

42

ion trap, with the results shown in Figure 3.4.

Figure 3.4: Simulated radial distribution of ions in the BECOLA RFQ ion trap, calculated
using the 3DCylPIC software. The green circle describes the 95% inclusion radius of the
radial distribution, approximately 1.4 mm in radius. The orange circle describes the extent
of the trapping region defined by the RFQ rods.

The simulation results suggest that 95% of ions in the trapping region are contained

transversely within a 1.4 mm radius of the center of the trapping region.

The simulation was repeated with the He buffer gas temperature parameter set to 200

K. The trapped ion distribution in the simulation was not sensitive to the temperature of

the He buffer gas, indicating that space charge effects are primarily responsible for the size

of the distribution.

43

The optical pumping process will not be sensitive to the distribution of the ions in the

axial direction.

3.2.2 Laser Conditions

Other assumptions within the optical pumping simulation are made about the qualities of

the laser light used to irradiate the trapped ions and perform the optical pumping. First, it

is assumed that the pulsed laser light fully saturates the electronic transition. The intensity

of light that is required to saturate a transition can be determined from the rate equations

and the Einstein coefficient A of the transition. For the simulations, it is assumed that the

energy density of the laser light is well above the saturation energy density, which can be

determined from the frequency of the transition and the lifetime of the excited state [81].

Esat =

πhc
3λ3τ

(3.29)

where h is the Planck constant, c is the speed of light, λ is the wavelength of the transition,

and τ is the lifetime of the excited state. For a transition at 360 nm and an excited state

with a lifetime of 5 ns, the saturation energy density corresponds to 11.4 mW/cm2. Pulsed

laser systems typically feature pulse widths on the scale of nanoseconds, meaning that the

peak laser intensity is measured in MW/cm2 and saturation should be easily achieved.

The second assumption is that all of the trapped ions are irradiated by the laser beam.

If the laser beam is misaligned, a portion of the ions in the trapping region will not interact

with the light and optical pumping will not occur. A telescope optics system and optical

elements equipped with piezo motors with high positional resolution will allow for control of

the laser beam size and position, and will be discussed later.

44

3.3 Operation of the Optical Pumping Simulation

The following section presents the practical aspects of using the optical pumping simula-

tion. The simulation begins with the user providing a series of constraints for the electronic

transition search process. The element of interest, charge state, minimum and maximum

wavelengths, as well as the minimum transition speed are specified as inputs. Based on the

element and charge state, a text file, which contains information about the electronic states

and their degeneracies from the Kurucz and NIST databases, is read.

The first step of the simulation is the determination of possible first steps for the pump-

ing scheme. As optical pumping must be performed on an electronic state that has some
population, the population distribution among low-energy states below 1000 cm−1 is cal-

culated. Another text file containing the possible transitions for the specific element and

charge state is read and used to identify transitions involving low-lying electronic states that

are populated at room temperature. Based on the constraints used as input, the transitions

are filtered to avoid any that do not fulfill the desired characteristics. After selection of the

possible pumping transitions, the rate equations are solved for each transition and tracks

the electronic population to the low-energy metastable states.

3.3.1 Choice of CLS Step and Output

After tracking the decays from the excited state, the potential transitions from the populated

metastable states are chosen. Any unpopulated metastable state is not considered as a viable

initial state for CLS measurements. Based on constraints to the wavelength and transition

speed, the possible CLS transitions are filtered again.

After selection of possible CLS transitions, the simulation is complete. The possible

45

optical pumping schemes are collected and sorted in order of highest population in the

starting state for CLS measurements. The list of schemes is saved to a text file in a chart

form. The chart first lists the pumping step of the scheme, with the ground and excited

states, the speed of the transition, the wavelength of light required, and the branching ratios

to the ground state and to the metastable state for the CLS step. Next the same information

is listed for the CLS step of the scheme as well as the population of the ground state after

optical pumping.

3.4 Results for Transition Metals

The simulation was used to determine the benefit of optical pumping for the first and second-

row transition metals.

3.4.1 First-Row Transition Metals

The first-row transition metals have many interesting isotopes for nuclear structure studies,

as detailed in Chapter 1. Studies of all elements in the first row except for Sc and Ti require

atomic charge exchange reactions, which severely limits the sensitivity of measurements

due to fractionation of the electronic population (See Section 1.3.1.3). The results of the

simulation for these elements are seen in Figure 3.5, with more detail of the results shown

in Table 3.2. A significant gain in sensitivity is predicted for most isotopes in the first row,

especially for Fe and Co isotopes which feature a 11-14 fold improvement. More detailed

optical pumping schemes are provided in Appendix A.

Although optical pumping will benefit most elements, there are no available optical pump-

ing schemes for ions of Ni, Cu, and Zn. Ground state electronic transitions of these elements

46

Figure 3.5: Results of the optical pumping simulation for the first-row transition metals
compared to traditional CLS measurements. Most elements will experience a substantial
gain in sensitivity through optical pumping, although some elements require very deep UV
light that cannot be produced at BECOLA and therefore have no viable optical pumping
schemes.

Table 3.2: Percentage of total electronic population available for CLS measurements of the
first-row transition metals, with and without optical pumping. Elements that require charge
exchange reactions are indicated.

Element Population Without Charge Exchange Population With
Optical Pumping

Optical Pumping

Sc
Ti
V
Cr
Mn
Fe
Co
Ni
Cu
Zn

37.6
32.9

4
12
12
5
5
15
43
34

No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes

47

60.5
50.3
33.5
58.6
58.4
56.4
69.2
N/A
N/A
N/A

Figure 3.6: Results of the optical pumping simulation for the second-row transition metals
compared to traditional CLS measurements. Most elements will experience a significant gain
in sensitivity through optical pumping.

require light far into the vacuum UV regime (<200 nm). Production of vacuum UV light

is not possible with the BECOLA pulsed laser system, and would require transportation

through vacuum to avoid absorption by molecules in air.

3.4.2 Second-Row Transition Metals

Simulations of the second-row transition metals also show promising results for future studies.

Most elements in the second row require charge exchange, which contributes to the fact that

there have been no published results for radioactive beam studies of isotopes between Mo and

Ag, although a limited number of such experiments have been performed. The simulation

results are shown in Figure 3.6, with more detail in Table 3.3

The results of the simulation show that the second-row transition metals will not benefit

as greatly as elements of the first row, but many species will see an increase in sensitivity.

48

Table 3.3: Percentage of total electronic population available for CLS measurements of the
second-row transition metals, with and without optical pumping. Elements that require
charge exchange reactions are indicated.

Element Population Without Charge Exchange Population With
Optical Pumping

Optical Pumping

Y
Zr
Nb
Mo
Tc
Ru
Rh
Pd
Ag
Cd

94.9
72.3
44.3

4

N/A

7
15
30
70
34

No
No
No
Yes
N/A
Yes
Yes
Yes
Yes
Yes

96.8
83.0
46.7
70.2
N/A
74.6
N/A
N/A
N/A
N/A

The largest increase in sensitivity is seen in Mo and Ru, featuring a 17- and 10-fold increase,

respectively.

The case of Tc must also be discussed. The benefit of optical pumping is unknown for

Tc due to a lack of information about its electronic structure, as there are no stable isotopes

of Tc. To date, few studies have been performed on Tc isotopes, mostly limited to resonance

ionization spectroscopy studies [82, 83]. Future measurements of rare Tc isotopes would

require preparation work in collaboration with other groups capable of performing level

search measurements to identify the location of electronic states and potential transitions.

Also, as no stable isotopes exist, future studies will rely on atomic theory calculations for

the atomic factors used to extract the charge radius from the isotope shift. Tc was not

considered a potential candidate for future experiments in this work.

49

Chapter 4

Experiment

The experimental techniques used to perform the optical pumping commissioning experi-

ments are described, as well as the features of the BECOLA facility, improvements to the

infrastructure of BECOLA, and how optical pumping was performed are described in this

chapter.

4.1 Experimental Techniques

4.1.1 Collinear Spectroscopy

BECOLA utilizes collinear laser spectroscopy (CLS) [84] to precisely measure the hyperfine

spectra of radioactive nuclei and extract the mean-square charge radii and nuclear moments.

In a CLS system, a beam of ions propagates collinearly with a laser beam, which excites

an electronic transition of the ions, and the fluorescence is detected using photomultiplier

tubes (PMT). Typically, the laser light used to probe the electronic transition is held at

a steady frequency and a small scanning voltage is applied to the photon detection region

to Doppler shift the frequency of light seen by the ion beam.

It is possible to scan the

frequency of the laser light, but such a scan requires re-adjustment of the laser optics and

precise measurement of the change in laser frequency. The stability of the laser also becomes

an issue at frequencies close to the edge of the wavelength range of the laser system. It is

50

easier in comparison to apply a scanning voltage to Doppler shift the absorption frequency.

4.1.1.1 Doppler Shift Effect

The CLS technique is capable of performing high precision measurements by taking advan-

tage of the Doppler effect. The absorption linewidth of a transition is dominated by Doppler

broadening effects, caused by movement of ions relative to the laser light.

The FWHM of the natural absorption linewidth of an electronic transition is related to

its lifetime, such that

∆ν =

1
2πτ

(4.1)

where τ is the lifetime of the excited state of the transition. The excited state of the

transitions studied in this experiment has a lifetime of approximately 7 ns, which corresponds

to a natural absorption linewidth of 22 MHz.

The impact of Doppler broadening can be calculated from the temperature of the ion

cloud and the Doppler shift. The Doppler shift at non-relativistic velocities is,

f = f0(1 +

v
c

)

(4.2)

As the resonant frequency shifts due to the velocity of the ions moving towards or away from

the laser, the distribution of absorption frequencies of the laser light will be,

P (f )df = P (v)

dv
df

df

(4.3)

where P (v) is the distribution of ion velocities, which can be described by the Boltzmann

51

distribution.

(cid:114) m

2πkT

− mv2

2kT dv

e

P (v)dv =

(4.4)

By solving Equation 4.2 for the velocity of the ion and substituting it into Equation 4.3, we

find

P (f )df =

c
f0

Pv(c(

f
f0

− 1))df

(4.5)

Through substitution of Equation 3.7 and simplification, the distribution of frequencies can

be expressed as a Gaussian distribution,

(cid:114)

P (f )df =

− mc2(f−f0)2

2kT f 2
0

f 2
0 e

mc2
2πkT

df

(4.6)

As it is a Gaussian distribution, the full width half max (FWHM) of the distribution i.e. the

Doppler broadened linewidth, is given by

F W HM =

(cid:114)

8kT ln(2)

mc2

f0

(4.7)

The in-air Doppler broadening of the transition used in this experiment increases the ab-

sorption linewidth to 1.1 GHz, much larger than the natural linewidth, but the 3DCylPIC

software [80] used to determine the size of the trapped ion distribution in Chapter 3 can also

be used to estimate the effect of the RF field on the velocity of the ions, and in turn the

Doppler broadening of the absorption linewidth from the RF motion.

The 3DCylPIC software provides a means to solve for the velocity distribution of the

trapped ions, which can be directly related to the Doppler broadening. A series of simulations

were performed to calculate the velocity distribution of the ions as a function of the He buffer

52

gas pressure. The magnitude of the mean velocity as a function of time in the ion trap and

He buffer gas pressure is shown in Figure 4.1. The results show that at lower He buffer gas

Figure 4.1: Magnitude of the mean velocity as a function of time in the ion trap and He buffer
gas pressure. The lines are to guide the eyes. The He buffer gas serves to cool the beam and
limits the velocity of the ions as they move through the RF field. Doppler broadening of the
absorption linewidth is reduced as the ions slow.

pressures, the velocity of the ion bunch increases until it reaches an equilibrium point. At

low pressures, there is not enough He in the RFQ ion trap to effectively cool the beam. The

BECOLA RFQ ion trap is capable of operating at pressures lower than 20 mTorr, but the

simulation suggests that the efficiency will suffer.

The mean velocity from the ion trap simulation was used to determine the FWHM of the

absorption linewidth due to Doppler broadening. The results are shown in Figure 4.2. The

He buffer gas plays a role by slowing the trapped ions as they move in the RF field. Collisions

with the He buffer gas will limit and cool the trapped ions. The BECOLA RFQ ion trap

is typically operated at 80 mTorr of He buffer gas, and the Doppler broadened absorption

linewidth is comparable to the expected linewidth in air. The precision necessary to extract

53

Figure 4.2: Doppler broadened absorption linewidth from RF motion as a function of He
buffer gas pressure, calculated from the 3DCylPIC software [80]. The optical pumping
process was performed in an RFQ ion trap and the motion of the ions from the RF field was
used to determine the absorption linewidth. As the He buffer gas pressure increases, the
neutral He atoms serve to limit the velocity of the trapped ions and the absorption linewidth
reaches a plateau. The orange line indicates the expected Doppler broadened linewidth in
air at 300 K. The BECOLA RFQ ion trap is operated at 80 mTorr of He buffer gas pressure
and the Doppler broadened absorption linewidth from the RF motion is comparable to the
expected linewidth in air.

the isotope shift and hyperfine coupling constants from the HFS is of the order of MHz,

and precision measurements are not possible through spectroscopy of a Doppler broadened

resonance.

The ion beam co-propagates with the laser light, and the Doppler shift of the resonance

frequency can be determined through the velocity of the ion beam. The potential applied

to the ions, V, is the sum of the high voltage of the system, V0, and the scanning voltage

applied at the photon detection region, Vapp.

V = V0 + Vapp

(4.8)

54

An ion beam composed of ions of mass m, charge state q, and accelerated across a potential

difference of V, the ion will have a kinetic energy of

Ekin = qV =

1(cid:112)

1 − β2

mc2 − mc2

(4.9)

where β is the speed of the ion in units of c, the speed of light. By re-arranging the above

equation and solving for β, we find the ion velocity as

(cid:115)

β =

1 −

1

1 + q2(V /m)2 + 2qV /m

(4.10)

As β represents the velocity of the ion, the Doppler shifted laser frequency,fobs, can be

expressed as

fobs = f0

1 − β
1 + β

(4.11)

(cid:115)

Antilinear laser spectroscopy is also possible, where the laser light propagates anti-linearly

relative to the ion beam. The Doppler shifted laser frequency while performing antilinear

laser spectroscopy would be defined as,

fobs,anti = f0

(cid:115)

1 + β
1 − β

(4.12)

The fluorescence is collected as a function of scanning voltage, which is readily converted to

the rest-frame transition frequency according to Equation 4.11.

CLS mitigates Doppler broadening by measuring the resonance frequency of ions in flight

as they travel through the beamline with kinetic energy on the scale of keV, through what

is called kinematic compression [84]. As the kinetic energy is related to the velocity of the

55

ion by

Ekin =

mv2

1
2

(4.13)

The velocity spread of an trapped with a given energy spread is reduced at higher acceler-

ation energies, reducing the spectral width of the absorption linewidth towards the natural

linewidth and allowing for high precision measurements. As shown in Figure 4.3, for a given

Figure 4.3: The relationship between kinetic energy and velocity. Three sets of ions with
the same energy spread will have velocity spreads that become progressively narrower as the
ion energy increases. CLS measurements take advantage of the kinetic compression of the
linewidth to remove broadening effects and perform precise measurements of the transition
frequency.

energy spread, the corresponding velocity spread of the ions will decrease as the acceleration

potential is increased.

A 30 kV acceleration potential is used at BECOLA and the linewidth is dominated by

fluctuations in the high voltage of 30 kV. The typical linewidth observed due to fluctuations

56

in the high voltage is approximately 60 MHz, as can be seen in the HFS spectrum of 90Zr

shown in Figure 4.4.

Figure 4.4: Hyperfine spectra of 90Zr measured using CLS. The width of the hyperfine
spectra is approximately 60 MHz.

A consequence of the CLS technique is a limitation on the speed of transitions that

can be used. Only electronic transitions that can be excited and decay before the ions

pass the photon detection region can be used.

Ions accelerated with a 30 keV potential

pass the photon detection region, which has a length of less than 10 cm, in hundreds of

ns. The lifetime of the excited state must be at most hundreds of ns long, corresponding

to an Einstein coefficient A21 of 107. On the timescale of a CLS measurement carried out

using the existing photon detection system at BECOLA, any electronic state with a lifetime

significantly longer than hundreds of ns can be considered a metastable state.

4.1.1.2 Beam Bunching

A subset of CLS known as Bunched Beam CLS [27, 85, 2] is employed at the BECOLA

facility. The most significant source of background during a CLS measurement is from

57

scattered laser light as the laser light passes the PMTs. The scattered light signal cannot

be filtered from the fluorescence as the wavelengths of light are very similar. The stray light

background prevents measurement of the HFS of radioactive ions with low production rates,

as the signal to noise ratio (SNR) is too low and the peak cannot be adequately resolved.

The stray light background can be suppressed by concentrating the ions, trapping them

in an RFQ cooler/buncher ion trap, and releasing them in short bunches. Typically the

ion bunches reach the photon detection region with a temporal spread of approximately 1

µs. The detected photons are only counted in the window of time as the ion bunch passes

the photon detection region. The background outside of the time window can be neglected

and assuming 100% efficiency in trapping the ions, the signal is concentrated into the ion

bunch and the SNR is dramatically increased.

If the ions are released once per second

with a temporal width of 1 µs, the background will be reduced by 6 orders of magnitude

without reducing the observed signal. The bunched beam spectroscopy technique is critical

for measurement of the HFS of radioactive beams.

4.2 BECOLA Facility

The optical pumping technique described in this thesis was developed at the BECOLA facility

[1, 2] , located at the NSCL. BECOLA is a dedicated end station for laser spectroscopy

studies at the NSCL, and has been used to perform isotope shift measurements of rare

isotopes [18, 12, 17], as well as having produced nuclear polarized beams for β-NMR studies

and β asymmetry experiments. BECOLA can accept thermalized radioactive beams from

the NSCL gas stopper [86] or stable ion beams produced by off-line ion sources [58]. A

diagram of the BECOLA facility can be seen in Figure 4.5 and a description of the major

58

Figure 4.5: BECOLA Facility. Stable off-line beams or radioactive on-line beams are in-
jected into the RFQ ion trap. The trapped beam is then transported to the collinear laser
spectroscopy beamline for fluorescence measurements.

components of the facility follows.

4.2.1 Production of Ion Beams

The BECOLA facility can accept both radioactive ion beams produced by the coupled cy-

clotron facility and off-line stable beams from a variety of ion sources. The ions in the

commissioning experiments were produced by a Penning Ionization Gauge (PIG) ion source

[87]. The PIG ion source is a plasma discharge sputtering ion source, and is primarily used

to produce ion beams of metallic elements and has been used at BECOLA to study Fe [12],

Ca [17], Sc, Ni, and Pd isotopes. PIG ion sources work well to produce beams of metallic

and gaseous elements, producing high ion currents.

A schematic of the BECOLA PIG ion source is shown in Figure 4.6, taken from [58] and

59

a photograph of the BECOLA PIG system is shown in Figure 4.7. The PIG ion source was

Figure 4.6: Cross section of the BECOLA PIG ion source. The PIG ion source can generate
ion beams of many metallic elements via plasma sputtering, at rates up to 10s of nA.

used to produce Zr+ beams for the commissioning experiments. A potential difference of

approximately 400 V was applied between a central aluminum anode held at +150V and a

pair of small anticathode and cathode cylinders situated equidistant from the anode. The

cathode and anticathode are approximately 11 mm in diameter and 3.4 mm thick and are

made of the desired element for an ion beam. A buffer gas, typically an inert element, flows

into the chamber and causes an electric discharge, producing a plasma. Buffer gas cations

collide with the cathode and anticathode and produces sputtered ions of the element of

interest, which are extracted from the chamber through a small hole of approximately 0.08

mm in the cathode by a conical extraction electrode, typically held at -3 kV. A magnetic coil

surrounding the central chamber provides a magnetic field of approximately 500 G, applied

60

along the central axis of the anode and cathode, that confines the electrons and creates a self-

sustaining plasma. A beam current of 2.8 µA was measured with a Faraday cup immediately

Figure 4.7: Photograph of the BECOLA PIG ion source. The PIG is attached to the
BECOLA beamline with a 6 inch flange.

after the ion source, with a primary component of the Ar+ buffer gas. The beam current

after the RFQ ion trap was measured with a Faraday cup at approximately 1.6 nA. The

RFQ also serves as a mass filter and much of the Ar+ component of the beam is removed

from the beam. The beam current after the photon detection region was measured with a

Faraday cup at approximately 1.4 nA.

The PIG ion source was able to successfully produce beams of Zr+, but the operation

of the PIG was significantly different from the conditions observed with other elements.

Sputtered ions can deposit inside the chamber during operation and short the cathodes and

anode. When the short occurs, the PIG ion source must be removed and cleaned. The

leakage current for producing Zr was an order of magnitude higher than what was observed

61

for producing Ni isotopes and the maintenance cycle was substantially reduced to 2-3 days.

The reduced maintenance cycle did not impede these measurements, as it was an offline

measurement that did not have a time limit. In addition, the Zr cathode and anticathode

were substantially corroded through operation. Erosion of the Zr cathode increased the

size of the hole used to extract ions from the central chamber, reducing the pressure of the

plasma chamber and eventually rendering the PIG ion source inoperable. A comparison of

Figure 4.8: Cathode and Anticathode composed of Zr. The Zr cathodes quickly eroded
over the course of the commissioning experiments. As the erosion progressed, the hole in
the anticathode expanded and the cathode developed a large hole as well, which prevented
operation of the PIG ion source.

a Zr cathode after extended period of operation and a new cathode is seen in Figure 4.8.

4.2.2 Radiofrequency Ion Trap

A helium-buffer-gas filled Radio Frequency Quadrupole (RFQ) ion trap is used to cool and

bunch ionic beams at BECOLA. A more detailed description of the RFQ ion trap can be

found in [78, 77] , but a brief description will follow. A schematic of the BECOLA RFQ ion

62

Figure 4.9: Schematic of the BECOLA RFQ ion trap, taken from [78]. The RFQ ion trap
features three sections separated by differential pumping barriers. Ions are cooled through
collisions with a He buffer gas in the cooling section, and are trapped in preparation for
bunched beam CLS measurements in the trapping region.

trap is seen in Figure 4.9.

The RFQ ion trap is separated by differential pumping barriers between three sections,

responsible for collection of ions, cooling via buffer gas collisions, and trapping. The incoming

ion beam is captured and focused into the RFQ ion trap by injection optics consisting

of a ring-shaped hyperboloid and a cone with a central aperture. Voltage applied to the

hyperboloid decelerates and focuses the ion beam, while the cone prevents the RF field from

penetrating into the deceleration field. Two DC drag field regions are used to move the ions

through the RFQ ion trap, and an RF field provides transverse confinement. The first drag

field is 0.8 V/cm, while the second drag field is gentler and is 0.4 V/cm. Instead of using

segmented rods to produce the drag field, the BECOLA RFQ ion trap features a crosscut

design that moves across the length of the electrode. A diagram of the crosscut geometry

from [78] can be seen in Figure 4.10. The geometry of the crosscut allows for the production

of a linear drag field with fewer components and power supplies.

After injection, the ions enter the cooling section of the ion trap. A He buffer gas was

63

Figure 4.10: Diagram of the crosscut geometry of the BECOLA RFQ rods, taken from [78].
The view of the rods is longitudinal, with the beam direction coming out of the page. The
crosscut design produces a linear drag field across the length of the RFQ ion trap with only
two power supplies.

streamed into the cooling section. The pressure of the He buffer gas was measured at the

gas inlet, and there are no pressure gauges within the RFQ ion trap itself. Typically the

RFQ ion trap is operated at 80 mTorr of He buffer gas pressure at the gas inlet. The He

buffer gas is held at room temperature and collisions between the ion beam and the buffer

gas cool the ion beam to room temperature and reduces the emittance.

Another differential pumping barrier separates the cooling region from the trapping re-

gion, where the applied RF field and DC voltage at the segmented electrodes produces

a trapping field. Simulations performed during the construction and installation of the

BECOLA RFQ ion trap suggest that the He buffer gas pressure in the trapping region of

64

the ion trap are approximately 100 times smaller than the cooling section, or approximately

800 µTorr. A detailed drawing of the segmented electrodes can be seen in Figure 4.11, taken

from [78].

Figure 4.11: Diagram of the segmented electrodes of the BECOLA RFQ trapping region,
taken from [78]. The diagram is a crosscut of the RFQ ion trap, with the beam traveling to
the right. Voltages are applied to the segmented electrodes to produce a potential well to
trap the ions and allow them to be released in bunches for laser spectroscopy measurements.

The first two segmented electrodes were grounded, -0.2 V was applied to the third elec-

trode, -4.2 V was applied to the fourth electrode, and 19.8 V was applied to the final electrode

to produce the potential well to trap the ions. The 19.8 V potential is flipped to -15 V for

ejection of the trapped ions. The flipping of the potential of the last segmented electrode

takes approximately 5 µs.

The stability of the ions in the trap can be calculated based on the amplitude and

frequency of the RF field and mass of the ion, and is expressed as what is known as the

65

Mathieu stability parameter [88], q,

q =

4eVRF
0ω2
mr2
RF

(4.14)

where e is the charge of the ion, VRF is the peak amplitude of the RF field, m is the mass of

the ion, r0 is the radius of the trapping region defined by half the distance between the RFQ

rods, and ωRF is the RF angular frequency. An ion is considered stable in the RF field if

q is less than 0.908. During the commissioning experiments, the RF angular frequency was

7.5 MHz, with VRF amplitude of 72.5 V. The mass of the Zr+ is 90 AMU. The radius of

the trapping region is 3.5 mm or 0.0035 m. The Mathieu stability parameter was 0.44 under

these operating conditions.

Ions that are not stable in the RF field will be removed from the beam, allowing the

RFQ ion trap to serve as a mass filter as well. The mass resolution of the BECOLA system,

m/δm, is roughly 2, meaning that conditions stable for an ion of mass 90 will be able to

filter ions with masses more than 45 mass number away.

One limitation of the RFQ ion trap is that it is possible to overfill the trapping region.

The energy spread of the trapped ions increases as the trapping region becomes overfilled

and distorts the beam shape and beam energy, and subsequently the hyperfine spectrum,

due to space charge effects. The temporal width of the ion bunch is significantly lengthened,

serving as an indicator of the overfilling process. An example can be seen in Figure 4.12,

where the time spectra of two filling conditions are compared. In the spectrum on the right

the ion trap is overfilled and has approximately 10 times more ions than the spectrum on

the right. The overfilling results in an almost 3 times increase in the FWHM.

The optics used to focus the beam into the RFQ ion trap can be detuned to decrease the

66

Figure 4.12: Measured time spectra of 90Zr with different numbers of ions trapped. The time
spectrum on the right of the figure featured the fewest number of ions per bunch and had a
FWHM of approximately 1.5 µs. The time spectrum on the right contained approximately
10 times more ions per bunch, resulting in a significantly broader FWHM in time.

overall rate entering the trapping region and mitigate overfilling of the ion trap.

4.2.3 CLS Beamline

The Zr+ ions were transported through the CLS beamline after ejection from the BECOLA

RFQ ion trap. Two kicking electrodes bent the beam 30 degrees in total, overlapping the

ion beam with a continuous wave laser beam injected through a window port in the vacuum

chamber of the bender.

The laser light in the CLS beamline was aligned to the ion beam through two adjustable

apertures located in the beamline. Each aperture can be vertically adjusted to three dif-

ferent diameters, of 3, 5, and 7 mm respectively. The laser light was ”walked” through the

apertures, where the path of the laser beam is adjusted through the use of two mirrors, to

minimize the stray light background. A telescope system of lenses was used to focus the

CW laser light through the beamline. Typically, the best results for the minimization of the

background were seen as the laser light was focused at the end of the beamline, with a beam

67

size of 1.4 mm, although the ion beam was assumed to be approximately 5 mm in diameter.

300 µW of laser light power was used for the CLS measurement with a transmission through

the beamline of 75%.

The ion beam and laser beam propagated collinearly through the beamline to the inter-

action region (photon detection region), where the laser light interacts with the ion beam

to emit fluorescence. A scanning voltage was applied at the photon detection region, which

Doppler shifts the ion beam in and out of resonance with the laser light, as described in

Chapter 2. The laser-resonant photons were captured by three PMTs. One PMT was

placed at the focal point of an ellipsoidal reflector, which is used to collect and focus light

from the second focal point of the ellipsoid. Two other PMTs are part of a new system that

was installed and used in a recent online experiment involving proton-rich isotopes of Ca

[17, 89]. In the new system an elliptical reflector is used instead of an ellipsoidal reflector.

The elliptical reflector also serves to focus the signal, but the focal region is a line instead of

a point and photons can be collected from a longer path length. The fluorescence is focused

to the second focal point of the elliptical reflector outside of the beamline through a vacuum

window. An aperture is used to block stray light photons and the focused signal is sent to a

compound parabolic concentrator. The parabolic concentrator rejects photons that enter at

an angle of more than 20o, and concentrates the signal to the PMT. Stray light photons are

more likely to enter at larger angles, and the background can be further reduced. Although

the design of the reflectors for the PMTs are different, it is important to use both during

experiments, as more collected photons reduces the statistical uncertainty of a measurement.

A diagram of the photon detection region is shown in Figure 4.13.

After the photon detection region the CLS beamline terminates at an anti-reflection

coated glass window. The glass window is used to prevent the laser light from reflecting

68

Figure 4.13: Diagram of the photon detection region in the BECOLA CLS beamline, taken
from [89].The photon detection region is located immediately after the charge exchange cell.
The new design of the photon collection system is placed directly after the old design. The
photon detection region serves to collect and focus fluorescence to photomultiplier tubes for
CLS measurements.

back through the beamline and contributing to the scattered light background. After the

glass window, a laser safety box is used to contain the laser light. A small optical breadboard

is situated in the box, where the laser light that travels the length of the CLS beamline can

be monitored with cameras or power meters during operation.

4.2.4 Continuous Wave Laser System

A CW Sirah Matisse Titanium:Sapphire ring laser [90] was used to perform CLS for the

commissioning experiments. The Matisse system was pumped by a Spectra Physics Millennia

EV CW pump laser [91] at 532 nm with a maximum output power of 20 W and features

a tunable range of 700-1000 nm, and is located in another room away from the BECOLA

69

experimental area.

Although the spectral bandwidth of the Ti:sapphire system is over 300 nm, the output

of the Ti:Sapphire system is made up of over 370,000 modes, or resonances, separated by

160 MHz. The separation between these modes is called the Free Spectral Range(FSR).

Frequency selective optics are used to separate these modes and deliver narrow linewidth

laser beams on the order of 100 kHz.

The wavelength ranges of the CW laser systems can be extended through the use of a

Sirah WaveTrain [92] frequency doubling system. As the CW laser light propagates through

the nonlinear crystal, it causes the electron distribution to polarize. The polarization of

the electron distribution in the crystal generates a second wave with twice the frequency

of the fundamental light. Since this is a nonlinear process, the efficiency of the doubling

process scales with the square of the laser power. CW laser systems have relatively low

energy density, so the WaveTrain system uses a passive resonator ring cavity to produce

constructive interference and increase the total efficiency of the conversion. The WaveTrain

system allows BECOLA to extend the range of wavelengths available for CLS to 350 nm for

the Ti:Sapphire laser system.

The light produced by the CW laser system was stabilized using a wavelength meter,

model HighFinnese WSU-30 [93]. The frequency of the laser light is compared to 633 nm light

produced by a frequency stabilized He-Ne laser [94] using a Fizeau interferometer to compare

the frequencies. A feedback signal from the wavelength meter is used to adjust the piezo

mirror of the Ti:Sapphire laser cavity and adjust the output frequency to match the desired

frequency and controls long term drift of the laser frequency. Short term changes in the laser

frequency are controlled by monitoring the laser light in a Fabry-Perot interferometer. An

electronic signal from a photodiode in the interferometer is used to determine the frequency

70

of light in relation to the resonance frequency of the cavity and a feedback signal is sent to

a PID controller to shift the internal optics of the Ti:Sapphire laser and control the output

frequency.

Two methods of light transport were used for the CW laser system during the experiments

described in this thesis. The first method was through fiber optic cable. The CW laser light

was transported to the BECOLA experimental area using a polarization maintaining single-

mode optical fiber. The relative location of the BECOLA CW laser system and the BECOLA

experimental area is shown in Figure 4.14. The fiber led to a safety box containing optics

Figure 4.14: The BECOLA laser room is located approximately 15 m away from the
BECOLA experimental area. The CW laser light is transported to the CLS beamline with
the use of fiber optic cable. The output of the fiber optic cable is located in a laser safety
box, where the laser light can be prepared for injection into the CLS beamline.

that prepared the CW laser light for injection into the CLS beamline. The safety box features

pressure activated switches that are connected to the BECOLA safety system. Opening the

box while laser light is passing through the system activates physical shutters that block the

laser in two locations to prevent accidental exposure. One shutter is located immediately

after the Ti:Sapphire output, and the other shutter is located after the fiber optic output in

71

the BECOLA experimental area.

The second light transport method was through a new free-space light transport system,

which will be described in detail later in this Chapter. The CW laser light was translated

upwards to an inverted optical breadboard where two mirrors were placed. The mirrors allow

for alignment of the CW laser light through a 10-m long pipe connecting the BECOLA laser

room and experimental areas. The 10-m long pipe terminates in a laser safety box in the

BECOLA experimental area, where the light is translated down into the original laser safety

box, and the light is prepared for injection into the CLS beamline as described for the fiber

optic light transport system.

4.3 New Developments to BECOLA Facility

A new pulsed laser system and method to transport the light from this laser to the BECOLA

experimental area were required to perform the commissioning experiments and are described

here.

4.3.1 Pulsed Laser System

It is desirable to have access to a wide range of wavelengths for optical pumping. As men-

tioned in Chapter 1, some transition metals require deep UV laser light for the pumping

process, which is difficult to produce with CW laser systems. A new pulsed laser system

was procured and installed in the BECOLA laser infrastructure to better enable the optical

pumping experiments. The pulsed laser system is composed of a Cobra-Stretch dye laser [95]

produced by Sirah, pumped by a Quanta-Ray INDI Nd:YAG laser [96]. The Quanta-Ray

INDI laser features a 10 Hz repetition rate and produces fundamental light at 1064 nm that

72

can be converted to 532 nm second harmonic light or 355 nm third harmonic light by a

harmonic generator installed in the pump laser housing. The laser output was measured

during installation. The INDI pulsed laser was able to produce 5 W of light at 1064 nm, 2

W of light at 532 nm, and 1.2 W of light at 355 nm. The pump laser light is split as it enters

the Cobra-Stretch, to pump two dye cells which are used as a gain medium for lasing. The

Figure 4.15: Laser path through the Cobra-Stretch dye laser system from [97]. Laser light
from the INDI pump laser excites the lasing medium in the 20 mm oscillator cell and the
40 mm amplifier cell. The output wavelength is selected by the grating position, which is
amplified in the amplifier cell, before entering the internal SHG unit where the frequency of
light is doubled.

optical layout of the Cobra-Stretch system is shown in Figure 4.15. One cell is known as the

oscillator cell and the other is the amplifier cell. The amplifier cell is shown in Figure 4.16

The pump laser excites the lasing medium in both cells. Lasing mediums in liquid solvents

typically have very broad spectral ranges, and a system of gratings allows for a relatively

73

Figure 4.16: Amplifier dye cell of the Cobra-Stretch pulsed laser system. The dye acts as a
lasing medium and is circulated through the cell to prevent laser-induced breakdown of the
dye molecules and extend its lifetime. The amplifier cell is 20 mm in length.

narrow spectral range to be selected from the output of the oscillator cell. The laser light

from the oscillator cell is sent through the amplifier cell where the total energy is greatly

increased. Pumps circulate the dye through the cells to prevent laser-induced molecular

breakdown and extend the lifetime of the dye.

The wavelength of light produced by the Cobra-Stretch system is determined by the dye,

which for a specific dye can produce light within a range of 30-40 nm with the conversion

efficiency of the fundamental light peaking at the center of the range. The wavelengths of

light produced with different dyes is shown in Figures 4.17 and 4.18 for dyes pumped by 532

nm and 355 nm light, respectively. With a broad selection of dyes, the Cobra-Stretch system

can produce light from 370-900 nm.

In addition to the INDI pump laser and the Cobra-Stretch system, a frequency doubling

74

Figure 4.17: Conversion efficiency as a function of wavelength for different dyes across the
range of 540-910 nm. Dyes in the range of 540-910 nm are pumped by second harmonic light
at 532 nm produced by the INDI pulsed laser. Figure taken from Sirah Cobra-Stretch User’s
Manual [97].

unit was installed that generates second harmonic light using a nonlinear crystal in a manner

similar to the WaveTrain system used for the CW laser systems. However, instead of a ring

cavity, the pulsed laser frequency doubling unit uses only one pass of the pulsed laser to

produce second harmonic light by taking advantage of the high energy density of the pulsed

75

Figure 4.18: Conversion efficiency as a function of wavelength for different dyes across the
range of 370-580 nm. Dyes in the range of 370-580 nm are pumped by third harmonic light
at 355 nm produced by the INDI pulsed laser. Figure taken from Sirah Cobra-Stretch User’s
Manual [97].

laser light. A series of prisms after the doubling crystal are used to separate the fundamental

Cobra-Stretch light from the frequency doubled light. The low end of the wavelength range

76

Table 4.1: Pulsed laser wavelengths and powers tested during installation of the Cobra-
Stretch laser system at BECOLA.

Wavelength (nm) Power (mW)

535
365
270
206
196

350
100
51
27
2

of the pulsed laser system is extended to 200 nm with the doubling unit.

A final unit was added to the pulsed laser system to reach 196 nm, which is not possible

to generate with frequency tripling. This final unit consists of a Raman shifter filled with

dimethyl sulfoxide (DMSO), which utilizes the anti-Stoke shift effect to shift the frequency

of 663 nm light to 555 nm light. A portion of the fundamental light is doubled to 331.5 nm

and mixed with the 555 nm light from Raman shifter to generate 196 nm light. Production

of laser light through mixing is less efficient than doubling using a nonlinear crystal, and

results in a sharp decrease in the laser power generated between 200 nm and 196 nm, but is

sufficient for the use in optical pumping.

The production of a range of wavelengths was tested during installation of the laser

system, with the results summarized in Table 4.1.

The INDI laser produces a laser pulse with a temporal width of 5-7 ns, which was mea-

sured by the manufacturer. The temporal width of the beam produced by the Cobra-Stretch

was measured at the BECOLA facility using attenuated light on a photodiode. The tempo-

ral width of the beam at 360 nm was measured as approximately 6 ns and the oscilloscope

capture is seen in Figure 4.19

77

Figure 4.19: The width of the temporal pulse produced by the Cobra-Stretch laser system.
The laser light was attenuated using a neutral density filter before being captured on a
photodiode.

An additional property of interest is the spectral linewidth of the laser light. The natural

linewidth of the optical pumping transition will be Doppler broadened to the order of GHz.

The spectral width of the pulsed laser should match the Doppler broadened linewidth to

ensure efficient excitation of the transition. The linewidth of the pulsed laser was measured

by the manufacturer at 567 nm. The linewidth of the fundamental laser light after the

Cobra-Stretch dye cells is approximately 0.04 cm-1, which is doubled as the fundamental

laser light passes through the doubling crystal. The end result is that the linewidth of the

laser light is 0.08 cm-1 or 2.4 GHz, which matches the thermal broadened linewidth of a fast

transition.

The frequency of the pulsed laser is monitored using a HighFinesse Angstrom WS6-600

78

wavelength meter [98], which features an absolute accuracy of 600 MHz. A portion of the

discarded fundamental light is sampled and injected into a fiber optic cable, where it is sent

to the wavelength meter. The wavelength meter was used to measure the pulse-to-pulse

frequency stability. Over a period of approximately 1 hour the standard deviation of the

frequency was found to be 0.02 cm-1.

A final property of interest of the pulsed laser system is the quality of the laser beam.

The laser beam profile from the INDI pump laser was examined using burn paper, which is

a special kind of paper that is burnt by the laser light. The profile of the INDI laser light

was circular and the power was evenly distributed. The burn paper results of both the INDI

pump laser and the Cobra-Stretch dye laser are shown in Figure 4.20.

Figure 4.20: Laser beam profiles of the INDI pump laser(Top) and Cobra-Stretch dye
laser(Bottom) as seen on burn paper. The laser profile of the INDI system is circular and
the power is evenly distributed, while the laser beam profile of the Cobra-Stretch is elliptical
and the power is not evenly distributed.

The laser light profile was also measured using a CCD camera, although the power output

79

from the INDI system was too high to safely image. The Cobra-Stretch system produces

light that features more than one mode, resulting in a laser beam with three distinct local

maxima of power. The pulsed laser light was attenuated and reflected to the camera to

prevent damage, and the resulting screen capture is shown in Figure 4.21.

Figure 4.21: Power distribution of pulsed laser light produced by the Cobra-Strech system.
Multiple modes of light are produced by the laser, resulting in a profile with three local
maxima.

The optical pumping simulation assumes that the power is evenly distributed across the

beam, a significant divergence from the experimental conditions and the implications will be

discussed in Chapter 6.

4.3.2 Free-Space Light Transport System

The second major development necessary for the optical pumping system was a free-space

light transport system. The BECOLA laser systems, including the new pulsed laser system

described above, are housed in a climate-controlled room located approximately 15 meters

80

away from the BECOLA experimental area. The layout of the BECOLA laser room and

experimental area can be seen in Figure 4.22.

Figure 4.22: The BECOLA laser room is located approximately 15 meters away from the
BECOLA experimental area. The free-space light transport system uses mirrors to deliver
light to the experimental area with minimal loss of power.

4.3.2.1 Pulsed Laser Fiber Optic Cable System

Although fiber optic cable is capable of transporting CW laser light of sufficient power for

CLS measurements, high energy density pulsed laser light presents unique challenges that

required a new system. The type of fiber optic cable used to transport the CW laser light

was not suitable for pulsed laser light for multiple reasons. The specific fiber optic cable used

at BECOLA has a damage threshold of 250 kW/cm2. Efficient coupling of the laser light to

the fiber optic cable requires the laser light to be focused to a size approximately 70-80% of

the effective mode area of the fiber. The single-mode, polarization-maintaining fiber has a

mode field diameter of 2 µm, limiting the total output power of the CW laser system to 150

mW before risking damage to the fiber optic cable. Typical CLS measurements use less than

1 mW of laser light and the damage threshold is not considered, but the energy density of a

pulsed laser is very high because the energy is distributed over a duration of nanoseconds.

The single mode polarization maintaining fiber has a different damage threshold for pulsed

81

laser light. The provided damage threshold from the fiber manufacturer for a laser pulse

with a temporal width of 10 ns is 5 GW/cm2. The Cobra-Stretch pulsed laser is capable of

producing up to 12 mJ of laser energy per laser pulse at 360 nm, with a pulse duration of

6 ns. The mode field diameter is 2 µm, giving an effective mode area of 25 µm2 or 2.5 10-7

cm2. With 12 mJ of energy concentrated to 2.5 10-7 cm2 over a duration of 6 ns, the peak

energy density approaches 8000 GW/cm2 and is very likely to damage the fiber optic cable.

Additionally, deep UV light causes wavelength and time-dependent damage to fiber op-

tic cables in a process known as solarization. Solarization damage has been observed at

BECOLA in time scales as low as days at less than 1 mW of laser power, and effectively low-

ers the transmission efficiency through the fiber optic cable, and results in a lower-quality

laser beam. Solarization-resistant fibers exist, but allow for the propagation of multiple

different modes of light, making it difficult to manipulate the light after ejection from the

fiber.

Although these challenges were known, a fiber optic light transport system was explored

due to its relative ease of use and enhanced safety in comparison with transporting the

laser light through the air. A 2-m section of UV600-SR solarization-resistant optical fiber

was purchased from Avantes Inc [99]. The solarization-resistant fiber was a multimode fiber

optic cable with a core diameter of 600 µm, compared to 2 µm for the single mode fiber optic

cable used for the CW laser light. The fiber was rated by the manufacturer to withstand

the energy density of the pulsed laser light.

The laser light was successfully coupled to the fiber, but suffered from low transmission

efficiency. In the testing of this fiber optic cable, the laser power after the Cobra-Stretch

system was 110 mW, and the highest-achieved transmission efficiency through the fiber optic

cable was 6%. Additionally, as the fiber optic cable was a multimode fiber, more than one

82

mode of light is able to travel through the fiber. After exiting the fiber, a lens telescoping

system was employed to manipulate the laser light. In the BECOLA experimental area, the

telescoping system would be used to prepare the light for injection into the beamline and

is important for controlling the beam size and divergence. Due to the multimode nature of

the output from the fiber optic cable, the beam waist of the laser light was very large. At a

distance of 4.5 m, the smallest achievable beam size was approximately 7.5 cm in diameter,

compared to the trapping region of the BECOLA RFQ ion trap, which has a diameter of

7 mm. Due to the low transmission and large beam size, the fiber transport system was

determined to not be feasible.

4.3.2.2 Free-Space Light Transport

The alternative to fiber optic light transport for the pulsed laser light is a free-space light

transport system, consisting of optics and infrastructure to safely transport the laser light in

air to the BECOLA experimental area. The free-space light transport system was necessary

to efficiently transport the pulsed laser light to the BECOLA experimental area and also

opens the possibility of simultaneous transportation of CW laser light, delivering the laser

beams to different physical locations in the BECOLA experimental area, where they are

injected into the beamline. The ability to transport deep UV CW laser light will extend the

range of elements that can be studied at BECOLA in the future.

Modifications to existing infrastructure were necessary to construct the light transport

system. An optical breadboard was mounted to a steel support beam located approximately

2.5 m above the floor in the BECOLA laser room. A 10-m long steel pipe was installed

connecting the BECOLA laser room to two new laser safety boxes that were constructed in

the BECOLA experimental area, which are seen in Figure 4.23.

83

Figure 4.23: New laser safety boxes constructed in the BECOLA experimental area. Laser
light from the BECOLA laser room is sent through a 10-m long steel pipe to the laser boxes.
The upper box contains two breadboards to direct the light through steel pipes to the lower
boxes for preparation before injection into the CLS beamline and RFQ ion trap.

The upper laser safety box contains two small breadboards mounted on the side of the

box. Mirrors are attached to the breadboards to direct the laser light down to the lower

safety boxes through shorter steel pipes. Holes in the safety boxes for pipes and cables are

fitted with rubber seals to prevent any light from escaping. The laser safety boxes were

mounted to heavy steel posts, which were aligned to the CLS beamline and bolted to the

ground to improve the stability of the structure.

84

4.3.2.3 Optical Layout

The laser light must travel approximately 15 m from the pulsed laser system to the RFQ ion

trap, and it is important to have good control of the laser beam along the light transport path.

The optical elements to manipulate the laser light through the free-space light transport

system were chosen through simulation using the FRED Optical Engineering software [100].

The optical layout is split into three sections, on the pulsed laser table, the upper section

of the laser room, and the laser safety box in the BECOLA experimental area. The optical

layout of the pulsed laser table is shown in Figure 4.24.

85

Figure 4.24: Optical layout on the pulsed laser table. The yellow light shows the path of
the pulsed laser light. The light is translated from the output of the Cobra-Stretch system
to a mirror using a pair of periscope mirrors. The light passes through two lenses, which are
used as a telescoping system to slowly focus the light through the light transport system.
The light is then reflected upward to an inverted breadboard.

The pulsed laser light exits the Cobra-Stretch system and is immediately translated

downwards using a pair of mirrors that act as a periscope. The light is directed through a

telescoping set of lenses that initially expand the laser beam and focus it over the free-space

86

light transport system, allowing for slow convergence to the BECOLA experimental area.

A concave lens with a focal length of 7.7 cm is the first lens in the telescope and a 50 cm

convex lens is the second lens. Initially, the first lens was a convex lens, but the pulsed laser

light at the focal point of the lens was powerful enough to ionize the air, and was replaced

for safety concerns. A second periscope system is used to translate the laser light up to an

inverted breadboard that is attached to a structural beam. The upper breadboard features

two mirrors that can be used to ”walk” the laser light through the pipe and is critical for

alignment. The optics of the upper area of the BECOLA laser room are shown in Figure

4.25.

Figure 4.25: Optical layout of the inverted breadboard in the BECOLA laser area. The
light is translated to the breadboard by a pair of mirrors used as a periscope. The light is
reflected by two mirrors which can be used for alignment through a 10 m long pipe to the
BECOLA experimental area.

In the BECOLA experimental area, the 10-m long pipe is attached to a pulsed laser

safety box, which features one mirror to direct the laser light down into a second laser safety

87

box. In the lower pulsed laser safety box, the light is prepared for injection to the beamline

and RFQ ion trap. The optical layout of the lower laser safety box is shown in Figure 4.26.

Figure 4.26: Optical layout of the pulsed laser safety box used to prepare the pulsed laser
light for injection into the RFQ ion trap. The light is translated down to the safety box
from the 10 m long pipe using a pair of mirrors used as a periscope. The light is reflected
by two mirrors which can be used to align the laser light through the RFQ ion trap. The
diagram below shows the laser light path in yellow and the reflected light path in green,
mirrors are blue ovals, the backstop is black, and the wedged window is a red oval. The red
lines indicate where the optical elements are located in the photograph.

There are three mirrors in the lower laser safety box, one is part of a periscope system to

translate the laser light from the 10-m long pipe to the lower laser safety box, and two are

used for alignment through the RFQ ion trap. The laser light is injected into the RFQ ion

trap through a 2 degree wedged glass window, which reflects a small fraction of the beam.

88

A backstop was installed in the safety box, situated such that the back reflection from the

wedged window could be picked off by a small 1/2” mirror and sent to the backstop. The

beam path of the back reflection was 2.44 m, a distance equal to the RFQ trapping region

from the wedged window. During operation, it is not possible to observe the beam spot at

the trapping region or test the alignment through the RFQ ion trap. The back reflection

serves as both an alignment tool and way to monitor the beam. During the initial alignment

procedure, a flange with a glass window replaced the PIG ion source and the laser light was

aligned through the full length of the RFQ ion trap. The beam spot on the backstop was

used to mark the correct position for alignment, as well as two irises in the main beam path.

During operation, the laser light path can be adjusted to pass through the irises and to the

marked spot on the beam backstop, ensuring that the laser light is properly irradiating the

trapping region.

4.3.2.4 Stability

Measurements of the stability of the free-space light transport system were performed to

determine if active adjustment systems were necessary to ensure overlap between the pulsed

laser light and the trapping region is maintained. The stability measurements were performed

with laser light from the BECOLA CW laser system and assumes that the stability of the

light transport system is independent of the qualities of the laser light. Movement of the

laser light over a period of time was measured on a CCD camera at different locations in the

free-space light transport system. The centroid of the laser beam was tracked over a period

of 25 minutes, with the data shown in Figure 4.27

89

Figure 4.27: Center of laser beam tracked across 25 minutes at four different locations. Each
set of data is relative to the average position of the laser light. The data was collected at four
locations, on the laser table, at the entrance of the steel pipe connecting the BECOLA laser
and experimental areas, at the exit of the steel pipe, and in the lower laser safety box before
injection into the beamline. Although movement of the beam is small in the BECOLA laser
room, the laser light moves on the order of millimeters by the end of the free-space light
transport system.

The stability of the laser beam was measured at four locations, on the laser table (1.7

m from laser), at the entrance of the steel pipe connecting the BECOLA laser room and

experimental area (4.3 m from laser), at the exit of the steel pipe (13.2 m from laser), and

in the lower laser safety box before injection into the beamline (14.7 m from laser). The

standard deviation of the laser light position is shown in Table 4.2. The results of the

measurements show that the position of the laser light can fluctuate on the order of a few

mm in the distance from the BECOLA laser room to the experimental area. The level of

movement does not impact the overlap of the pulsed laser light and the trapped ions, but is

concerning for the use of CW laser light transport. The stray light background is dependent

on the position of the CW laser beam and the laser beam must pass through two apertures

90

Table 4.2: Standard deviation of the relative laser light position at four locations. The data
was collected at four locations, on the laser table, at the entrance of the steel pipe connecting
the BECOLA laser and experimental areas, at the exit of the steel pipe, and in the lower
laser safety box before injection into the beamline.

Location

Laser Table

Entrance of Aluminum Pipe

Exit of Aluminum Pipe
Lower Laser Safety Box

Distance (m) Standard Deviation (µm)

1.7
4.3
13.2
14.7

6
19
337
337

before the photon detection region. The apertures themselves can be as small as 3 mm

and small movements of the laser beam position can result in the laser light clipping the

alignment apertures and greatly increased the background.

4.3.2.5 Safety System

Pulsed laser light has a high energy density and can be very dangerous. The BECOLA

pulsed laser system is a Class IV laser system. The light from the pulsed laser is dangerous

to the eyes under all conditions, can be dangerous to the skin, and can pose a fire hazard.

It is important that the free-space light transport system features robust safety systems to

prevent unexpected exposure to the pulsed laser light during operation. The new laser safety

boxes were installed with two interlock systems. The weight of the lids of the boxes depresses

a trigger, sending a signal to the safety system that the boxes are closed and laser light can

be safely sent through the system. Screws are used to lock the lids in place, but if the lids

are lifted while the light is being transported, the safety interlock activates and closes two

physical shutters to block the laser light and prevent exposure. The shutters were placed

immediately after the pulsed laser output and in the upper laser safety box of the BECOLA

experimental area, where it can block the light immediately after exiting the steel pipe used

91

Table 4.3: Natural abundances of the stable isotopes of Zr.

Mass number Natural Abundance (%)

90
91
92
94
96

51.45%
11.22%
17.15%
17.38%
2.8%

to contain the laser light between the BECOLA laser and experimental rooms.

The safety system features an override, where a key is used to unlock the system and

allow the physical shutters to be opened while the safety box lids are off, and is used for

alignment. New safety screens consisting of heavy canvas sheets were installed around the

boxes to enclose the area and prevent light from leaving a prescribed area.

4.4 Experimental Method

The goal of the commissioning experiments was to demonstrate that the electronic popu-

lations of ions trapped in the BECOLA RFQ ion trap can be manipulated using optical

pumping as predicted by the optical pumping simulation.

4.4.1 Choice of Zr Ion

Zr+ ions were chosen as the element that would be used to demonstrate the optical pumping

technique. As detailed in Chapter 1, studies of Zr are important from both a nuclear structure

and stockpile stewardship perspective. Additionally, the transitions of interest for Zr are

accessible to both the pulsed and CW laser systems. Zr cathodes were acquired from ACI

Alloys [101] and were made of natural Zr. The natural abundances of the stable Zr isotopes

are detailed in Table 4.3.

92

The PIG ion source was able to readily produce Zr+ beams as discussed in Chapter 4.2.1.

A partial electronic level diagram of Zr+ is shown in Figure 4.28. Zr is a d-block element,

and has a high electronic level density.

Figure 4.28: Partial electronic states of Zr+ [76]. The hyperfine levels of each electronic
state are not shown, as the level density of Zr+ is very high.

At room temperature, only the first electronic level is appreciably populated, with the

level split into four fine lines located at 0, 314,763, and 1322 cm-1 respectively. Only the first

three fine levels are appreciably populated, with approximately 72.3% in the ground state

4F3/2, 23.9% in the first excited state 4F5/2, and 3.7% in the second excited state 4F7/2.

93

4.4.2 Transition Schemes

The optical pumping simulation discussed in Chapter 4 was used to determine the optical

pumping schemes for Zr+. Based on the simulation, two optical pumping schemes were cho-

sen for the commissioning experiments. Both optical pumping schemes utilized the 4d25p

4G5/2 state as the excited state. For the case of Zr+, the branching ratios and transition

probabilities of excited states in the Kurucz database were theoretically calculated. The

calculations are based on calibrations of free-burning arc sources [102]. Additionally, exper-

imental data for the decay of Zr+ was not available in the NIST database. In the literature,

an experimental determination of the lifetime and branching ratios from the 4d25p 4G5/2

was found from astrophysical observations [103]. The excited state has a lifetime of 7.1(3)

ns, compared to the Kurucz prediction of 8.5 ns. The branching ratios to the low-energy

metastable states were significantly different, and a comparison between the experimental

data and Kurucz predictions are shown in Table 4.4.

The difference between the experimental and theoretical branching fractions to the low-

energy metastable states has a significant impact on the expected optical pumping effect and

is discussed later in this Chapter.

94

Table 4.4: Decay branches and transition probabilities from the 4d25p 4G5/2 state of Zr+,
comparing theoretical calculations from the Kurucz database [75] with experimental results
from [103]. The theoretical predictions included decays to three states that were not observed
experimentally, and are marked as ’-’.

Ground

State (cm−1)

Experimental

Branching

0
314
763
2572
2895
4248
4505
5752
7836
14059

Ratio
0.668
0.16
0.004
0.042

-

0.041
0.028
0.051

-
-

Experimental
Coefficient (s−1)

Einstein

9.4x107
2.2x107
5.6x105
5.9x106

-

5.7x106
3.9x106
7.2x106

-
-

Theoretical
Branching

Ratio
0.494
0.248
0.044
0.042
0.015
0.044
0.039
0.065
0.004
0.006

Theoretical

Einstein

Coefficient (s−1)

5.8x107
2.92x107
5.2x106
4.9x106
1.7x106
5.2x106
4.5x106
7.6x106
4.51x105
6.8x105

95

4.4.2.1 Depopulation of the Ground State

Figure 4.29: Optical pumping scheme for the depopulation of the ground state of Zr+. Ions
from the ground state are excited to the 4d25p 4G5/2 state. Approximately 32% of ions
decay to metastable states that are dark to subsequent CLS measurements, reducing the
sensitivity.

The first optical pumping scheme was chosen to depopulate the ground state of Zr+ and can

be seen in Figure 4.29. Ions in the ground state are excited to the 4d25p 4G5/2 state. A

significant proportion of the ions in the excited state will decay to low-energy dark metastable

states and will be inaccessible to the subsequent CLS measurement that probes the ground

state using the same 4d25s 4F3/2 to 4d25p 4G5/2 transition used for the optical pumping

step. The predicted change in electronic population of the ground state is dependent on
branching ratios from the excited state. Using the Kurucz theoretical calculation, (44.1±3)%

decrease in the electronic population is expected, while a (31.9+4−3)% decrease in the electronic

population is expected if the experimental branching ratios are considered.

96

Figure 4.30: Optical pumping scheme for the augmentation of the ground state of Zr+. Ions
from the first excited state are excited to the 4d25p 4G5/2 state. Approximately 66% of ions
decay to the ground state, increasing the sensitivity of subsequent CLS measurements.

A second transition was used to verify the depopulation of the ground state, by probing

the change in the electronic population of the first excited state of Zr+. The pumping step

of the optical pumping scheme remains the 4d25s 4F3/2 to 4d25p 4G5/2 transition, but the

transition used to probe the electronic population manipulation excites ions from the first

excited state at 314 cm−1, using the 4d25s 4F5/2 to 4d25p 4G5/2 transition. The optical
pumping simulation estimates a (65±6)% increase in the population of the first excited state
based on the theoretical branching ratios and a (48±9)% increase in the electronic population

using the experimental branching ratios.

97

4.4.2.2 Augmentation of the Ground State

The second optical pumping scheme was chosen to augment the population in the ground

state, and can be seen in Figure 4.30. Ions in the first excited state 4d25s 4F5/2 at 314 cm−1
are excited to the 4d25p 4G5/2 state. Although the majority of the ions will decay back to

the ground state of Zr+, the total population of the state being pumped is relatively small

and only a modest increase in the electronic population of the ground state is expected.
A (8.3±1.5)% increase in the electronic population of the ground state is expected using
the theoretical branching values from the Kurucz database. A (14.8±1)% increase in the

electronic population of the ground state is expected using the experimental branching ratios.

4.4.3 Measurement Method

In a typical CLS measurement, the laser-resonant photon counts are measured as a function

of the scanning voltage to determine the resonance frequency. The goal of the commissioning

experiments was to determine the ability to manipulate the electronic populations using

optical pumping, and a different strategy was used for efficient and reliable measurements.

4.4.3.1 Timing Sequence

The commissioning experiments were not performed to measure the resonance frequency of

the optical transition for Zr+. Instead, interest was on the overall change in signal with and

without optical pumping. The scanning voltage used to Doppler shift the ion beam was set

to remain in resonance at all times. A physical shutter was used to block the pulsed laser

beam for short increments of time, and the fluorescence signal was measured as a function

of whether the pulsed laser was blocked or not. The measurements with and without the

98

pulsed laser light were performed in alternating, 1 second intervals. This method of data

collection was aimed to control systematic fluctuations in the laser power and ion source

intensity and mitigate any systematic uncertainties.

Figure 4.31: Optical pumping scheme for the augmentation of the ground state of Zr+. Ions
from the first excited state are excited to the 4d25p 4G5/2 state. Approximately 66% of ions
decay to the ground state, increasing the sensitivity of subsequent CLS measurements.

The physical shutter was controlled using a signal from the BECOLA Data Acquisition

(DAQ) system [2]. A trigger signal from the flash lamp of the pulsed laser system was used

to start the DAQ process. A one-second period of time was chosen to alternate between

performing optical pumping and not performing optical pumping. The physical shutter can

reliably open and close in one second increments. It is assumed that changes in the ion source

and laser power intensity occur over many seconds or minutes and can be safely ignored in

the data analysis. As the pulsed laser has a 10 Hz repetition rate with each laser pulse

separated by 100 ms, each interval will consist of ten measurements of the fluorescence with

the pulsed laser irradiating the trapped ions followed by ten measurements of the fluorescence

with the pulsed laser blocked by the shutter. The 10 Hz trigger signal from the pulsed laser

flash lamp was converted to an alternating 0.5 Hz signal to control the shutter. The process

to convert the electronic signal is shown in Appendix B.

99

4.4.3.2 Determination of Optical Pumping Effect

Measurements were made of the number of photons as a function of both time and the

shutter status. An example of the data for one run is seen in Figure 4.32.

Figure 4.32: Example of optical pumping data, pumping ions from the ground state with
the 4F3/2 to 4G5/2 transition of Zr+. As the ion bunch passes the PMTs at approximately
30 µs, the fluorescent photons are captured. The photon counts were measured as a function
of time and shutter status. A time gate was applied to remove background. The time gate
is shown as vertical lines around the peak.

The ions were released in approximately 1 µs bunches in a repetition rate of 10 Hz.

The time-of-flight of the ion bunch to the interaction region was approximately 32 µs and

a time gate was applied to the data to exclude photon counts outside of the arrival of the

ion bunch. Photon counts within the time gate were integrated to obtain the total signal as

the ion bunch passed the interaction region. A background subtraction was performed by

taking an average of the signal in each time bin outside of the time gate.

The results of the optical pumping measurement will be reported henceforth as the optical

pumping effect. The optical pumping effect is the percent change in fluorescent signal of the

100

ion bunches that are irradiated with pulsed laser light relative to the fluorescent signal of ion

bunches of the same measurement that are not irradiated by pulsed laser light. The optical

pumping effect, P is determined by,

P = (

SOP
SN oOP

− 1)100

(4.15)

where SOP is the fluorescent signal with optical pumping and SN oOP is the fluorescent signal

without optical pumping, both with the background signal subtracted. The fluorescent signal

is directly related to the electronic population of the state being probed, and it is assumed

that the proportional change in signal is the same as the proportional change in the electronic

state being probed by CLS.

101

Chapter 5

Results

The commissioning experiments of the optical pumping technique were performed at the

BECOLA facility. A stable beam of Zr+ ions was produced by the BECOLA PIG ion

source. Ions trapped in the BECOLA RFQ ion trap were irradiated with pulsed laser light

and the resulting re-distribution of the electronic population was measured with a subsequent

CLS measurement. Two different optical pumping schemes were explored. The results of

the commissioning experiments are described in this chapter.

5.1 Properties of Optical Pumping in the RFQ Ion

Trap

The optical pumping technique was applied to Zr+ ions captured in the BECOLA RFQ

ion trap.

Initial measurements were performed to identify ideal conditions for pumping

experiments, including laser beam size, laser power, and the overlap of the laser and the

spectral absorption linewidths.

5.1.1 Laser Beam Size

Initial tests were performed with the pulsed laser focused to the trapping region, with a laser

beam of 1 mm diameter. The laser light was scanned across the trapping region of the RFQ

102

ion trap by adjusting the angle of the final mirror in the pulsed laser optics system, using

piezo electronics driven linear actuators produced by Newport [104]. The actuators will be

henceforth referred to as picomotors. Adjustment of the laser position with the picomotors

was performed to ensure good overlap between the trapped ions and laser beam. The mirror

mount with the picomotors has an angular resolution of 0.7 µrad, and as the mirror is located

3.7 m away from the trapping region, the resolution of each step of the picomotors was 2.5

µm in comparison with the expected diameter of the trapped ions of 2.8 mm (See Section

3.2.1).

An attempt was made to measure the size of the trapped ion distribution by scanning

the laser across the trapping region. Hysteresis of the picomotor was observed, where the

number of motor steps to move the laser beam to a position was different from the number of

steps required to return the laser beam to its original position. The laser beam was scanned

in one direction across the trapped ions to try and mitigate the hysteresis effect, and to

approximate the size of the trapped ions by observing the optical pumping effect at each

location. The measurement was performed with the pulsed laser power set to 6 mW. Results

are shown in Figure 5.1.

The optical pumping effect as a function of the pulsed laser light position was fitted
to a Gaussian lineshape and the standard deviation was (1.4±0.1) mm, with 95% of the
distribution within a (2.8±0.2) mm radius. Assuming that the trapped ion distribution is

spherical in shape and that the laser beam profile is circular, the optical pumping effect

will reach zero when the distance between the center of the laser beam and the center of

the trapped ion distribution is equal to the sum of the radii of the laser beam and trapped

ion distribution. Since the radius of the laser beam is known, the size of the trapped ion

distribution can be determined. The radius of the laser beam was 0.5 mm, leading to an

103

Figure 5.1: Optical pumping effect as a function of pulsed laser beam position. The pulsed
laser light was moved off of the trapped ions and scanned in one direction across the trapped
ions to approximate its size.

trapped ion distribution radius of (2.3±0.2) mm. A simulation of the trapped ion conditions

in the BECOLA RFQ ion trap was performed using the 3DCylPIC software [80] and discussed

in Chapter 3. The results of the simulation provided an estimate that 95% of the trapped

ions could be found within a 1.4 mm radius of the center of the trapping region.

An unexpectedly large optical pumping effect of up to -79% was observed while the laser

beam was focused to a small diameter and at high powers, as shown in Figure 5.2. Even if

100% of the ions were excited by the pulsed laser light, based on either the branching ratios

of the excited state from Ljung et al. or the theoretical values from the Kurucz database,

the optical pumping effect should not rise above 51%. It is believed that the high energy

density of the pulsed laser light disrupted the ion trap. Other explanations for this behavior

are that the laser light could have been focused onto one of the RFQ rods and disrupted

the RF field, or it is also possible that the laser light forced ions to be expelled from the

104

Figure 5.2: Spectra of fluorescent photon counts as a function of time, with and without
optical pumping. The measurement was performed with a tightly focused laser beam and
high laser power and the optical pumping effect is much larger than expected.

bunch. The measurement of the optical pumping effect is dependent on the assumption

that the number of ions in the RFQ ion trap does not change due to the laser, and cannot

discriminate between a change in the electronic population and a change in the number of

ions in the ion trap.

The focus of the pulsed laser light was adjusted to increase the size of the laser beam at

the trapping region after measurements of the trapped ion distribution were complete. The

beam spot at the trapping region was elliptical in shape and approximately 7 mm in diameter

along the major axis and 5 mm in diameter along the minor axis. The optical pumping effect

was insensitive to scanning of the picomotor after increasing the laser beam size, indicating

that the trapping region was fully illuminated. The unexpectedly large attenuation of the

fluorescence signal was not observed at the maximum laser power of 120 mW with the

larger laser beam. All subsequent measurements were performed with the larger laser beam

105

diameter.

5.1.2 Laser Power

Saturation of the transition for optical pumping is important to provide the largest possible

improvement in sensitivity for CLS measurements. The Cobra-Stretch laser system is capable

of producing light down to 196 nm with the energy per pulse decreasing with the wavelength.

At 357 nm, the laser system is capable of producing approximately 12 mJ per pulse, or 120

mW as the pulsed laser has a repetition rate of 10 Hz. The optical pumping effect was

measured as a function of the laser power to determine if the optical transition can be fully

saturated by the pulsed laser light.

The saturation curve was measured with laser beams of two diameters, a smaller laser

beam of 1 mm in diameter, and a larger elliptical beam with a major axis diameter of 7

mm and a minor axis diameter of 5 mm (See Section 4.2.2 for more detail). The laser power

was measured immediately after the pulsed laser system. The transmission through the free-

space light transport system was approximately 65%. The saturation curve was fit using the

function,

I = A

P

P + Psat

(5.1)

Where I is the pumping effect, A is a constant, P is the laser power, and Psat is the saturation

power.

The saturation curve measured using the smaller 1 mm diameter laser beam is shown

in Figure 5.3. The optical pumping effect shows a departure from the expected behavior at

high laser powers, where a second plateau is observed above 40 mW of laser power.

The appearance of a second curve and plateau in the optical pumping effect as a function

106

Figure 5.3: Optical pumping effect as a function of the laser power for the 4F3/2 to 4G5/2
transition in Zr+. The transition quickly saturates, but experiences a second sharp increase
in the optical pumping effect above 40 mW.

of laser power above 40 mW was unexpected in addition to observation of optical pumping

effects as large as 79%, it is believed that the pulsed laser light disrupted the RFQ ion trap

at very high energy densities.

The size of the laser beam was increased to try and mitigate this issue. The laser beam

size was increased by adjusting the position of one of the lens of the telescope used to focus

the light to the trapping region. The resulting laser beam was elliptical in shape and had

a major axis diameter of 7 mm and a minor axis diameter of 5 mm. The saturation curve

was measured with this larger beam size and is shown in Figure 5.4. The saturation data
were fitted and the saturation power was found to be (3.2 ±0.3) mW. The beam area is 110
mm2, or 1.1 cm2, and the saturation power density was (2.9±0.3) mW/cm2. The pulsed

laser has a temporal width of 6 ns, and with the fitted saturation power density, the peak

power density is approximately 48 MW/cm2

107

Figure 5.4: Optical pumping effect as a function of the laser power for the 4F3/2 to 4G5/2
transition in Zr+. A fast transition, the optical pumping effect quickly saturates.

The theoretical saturation power density calculated from the Einstein coefficient of the

optical transition is approximately 8 mW/cm2 (See Section 3.2.2). The experimental results

show that the BECOLA pulsed laser system can readily saturate the transition.

5.1.3 Linewidth

As previously discussed, a major strength of CLS measurements is the ability to reduce the

absorption linewidth of the optical transition to approach the natural linewidth, on the order

of 10s of MHz. In the RFQ ion trap, the thermal movement of the ions results in a Doppler

broadened linewidth, which is typically of the order GHz. The 4F3/2 to 4G5/2 transition of

Zr+ has a natural linewidth of 22 MHz, which is broadened at room temperature to 1.1 GHz.

Efficient excitation of the pumping transition requires good overlap between the absorption

and pulsed laser spectral linewidths.

The absorption linewidth of the pumping transition was measured by shifting the fre-

108

Figure 5.5: Effective absorption linewidth of the 4F3/2 to 4G5/2 transition in Zr+ at different
laser powers. As the laser power increases, the transition remains saturated even as the laser
moves off resonance, resulting in a flat-top spectrum.

quency of the pulsed laser light off resonance measuring the optical pumping effect via CLS.

The measurement was repeated at different laser powers to observe power broadening effects,

and the results are shown in Figure 5.5. The flat top feature of the spectra measured above

4 mW indicates that the transition is fully saturated even as the frequency of the pulsed

laser is moved far from the resonance frequency.

A saturated Gaussian function was determined to be better suited to the data, as sig-

nificant power broadening was observed. A saturated Gaussian function is often used to

fit power broadened spectral profiles [105, 106, 107]. The saturated gaussian function is

109

Table 5.1: Effective absorption linewidth of the 4F3/2 to 4G5/2 transition in Zr+ at different
pulsed laser powers.

Laser

Power (mW) Linewidth (cm−1)

Absorption
0.44±0.02
0.49±0.02
0.68±0.02
0.69±0.02
0.68±0.02

1
2
4
8
20

P = A

expressed as,

− (ν−ν0)2
2w2
− (ν−ν0)2

S ∗ e
S ∗ e

2w2 + 1

(5.2)

where A is an amplitude parameter, ν is the frequency of the pulsed laser light, ν0 is the

resonance frequency, w is the FWHM of the spectral profile, and S is a saturation param-

eter. Significant power broadening was observed at only 4 mW of laser power. The overall

magnitude of the optical pumping effect and the linewidth of the spectral profile increased

with the laser power. The linewidths obtained from the fit are detailed in Table 5.1.

5.2 Depopulation of the Zr+ Ground State

Using the measurement method outlined in Chapter 4, commissioning experiments were

performed to demonstrate the ability to manipulate the electronic state population of Zr+

in the BECOLA RFQ ion trap.

Measurements were first performed to depopulate the ground state of Zr+ using the 4F3/2
to 4G5/2 transition. A (-31.9+4−3)% reduction in signal was expected from the optical pumping
simulation based on experimental branching ratios and a (-44.1±3)% based on the theoret-

ical values from the Kurucz database. The depopulation of the ground state was observed

110

through the fluorescence signal. An example of the collected spectra is shown in Figure
5.6. The observed reduction in signal for the depopulating transition was (-24.6±1.6)%.

Figure 5.6: Depopulation of the ground state of Zr+ via optical pumping. The fluorescent
photons are measured using CLS as a function of time. The ion beam is kept in resonance
with the probe laser and alternating measurements were made with and without the pump
laser.

The uncertainty is dominated by systematic error. The ion beam remains in resonance with

the CW laser light during the CLS measurement, allowing for fast accumulation of photons

and reducing the statistical error. The overall statistical error is not totally captured by a

single measurement. The internal consistency of the measurements was used to estimate the

overall statistical error by comparing the optical pumping effect observed by measurements

taken under similar conditions, namely with the He buffer gas pressure set to 80 mTorr,

with the laser power set to its maximum of 120 mW, and with a large laser beam. The

measurement was repeated 26 times under these conditions. The overall statistical error

from the counting statistics from these measurements was 0.07%. The standard deviation

of the 26 measurements was used to estimate the contribution of the internal consistency to

111

the statistical error, and was evaluated as 1.6%.

A set of measurements was also performed that depopulated the ground state of Zr+,

but probing the change in population of the 1st excited state of Zr+ located at 314 cm-1.

An example of the collected spectra is shown in Figure 5.7 The 1st excited state has a room

Figure 5.7: Depopulation of the ground state of Zr+ via optical pumping and probing the
increased electronic population of the first excited state.

temperature population of 23.9%. Based on the experimental results of probing the ground
state, it is expected that the 1st excited state population would grow by (36±2)%. The

measurement probing the 1st excited state resulted in an observed increase in the electronic
state population of (32.5±4.1)%, agreeing well with the expected value. A summary of the

optical pumping effects from this section is shown in Table 5.2.

112

Table 5.2: Results of the depopulation of the ground state of Zr+ using the 4F3/2 to 4G5/2.
Results reported as the optical pumping effect.

Probing State (cm−1) Experiment
-24.6±1.6

0
314

36±2

Expected from
4F3/2 Results

-

32.5±4.1

Simulation, Ljung

-31.9+4−3
47.7±6

Simulation,

Kurucz
-44.1±3
65.6±6

5.3 Augmentation of the Zr+ Ground State

The second optical pumping scheme that was studied involved exciting ions in the first ex-

cited state, 4F5/2 located at 314 cm-1, which were optically pumped to the ground state.

The overall magnitude of the optical pumping effect is smaller than the depopulating opti-

cal pumping scheme because the first excited state is populated by only 23.9% of the total
electronic population. The optical pumping simulation predicts a (8.3±1.5)% increase in the

population of the ground state using the theoretical values of the branching ratios from the
Kurucz database, and a (14.8±1)% increase in the ground state population if the experimen-

tal values of the branching ratios from the excited state are used. An example of the spectra

captured for the populating optical pumping scheme is shown in Figure 5.8. The observed

increase in signal for the populating transition was 12.6(2.7)%. The overall statistical error

of the measurement was evaluated in a similar manner, by comparing the observed optical

pumping effect for many measurements performed under the same conditions. 32 measure-

ments were compared where the He buffer gas pressure was set to 80 mTorr, with a pulsed

laser power of 120 mW, and a large laser beam, and the standard deviation of the optical

pumping effect was used to determine the magnitude of the error.

113

Figure 5.8: Augmented population of the ground state of Zr+ via optical pumping. The
fluorescent photons are measured using CLS as a function of time. The ion beam is kept in
resonance with the probe laser and alternating measurements were made with and without
the pump laser.

5.4 Quenching of Optical Pumping

The BECOLA RFQ ion trap is capable of holding ions for up to one second with minimal

loss of efficiency. The BECOLA optical pumping simulation was used to estimate the optical

pumping effect after irradiating the trapped ions with multiple laser pulses. A measurement

of the optical pumping effect after irradiation of the trapped ions by multiple pulsed laser

pulses was attempted, but results indicated that the optical pumping effect was not enhanced

by multiple laser pulses. A comparison of the simulated optical pumping effect and the

experimental results is shown in Figure 5.9.

Instead, it was observed that holding the

trapped ions after irradiation by the pulsed laser resulted in quenching of the optically

pumped electronic populations due to collisional decay with the He buffer gas in the RFQ

ion trap. The quenching continued until the electronic state populations returned to thermal

114

Figure 5.9: Optical pumping effect of Zr+ as a function of number of laser pulses used to
irradiate the trapped ions. The blue bars show the expected optical pumping effect based on
the optical pumping simulation, with the orange bars showing the experimental results. The
experimental results show no enhancement of the optical pumping effect, providing evidence
that collisions with the He buffer gas re-thermalize the trapped ions to a room temperature
distribution.

equilibrium with the He buffer gas.

The optical pumping effect as a function of the delay time for the release of the trapped

ions is shown in Figures 5.10 and Figure 5.11, for the populating and depopulating optical

pumping schemes respectively.

The quenching of the optical pumping effect was fitted to an exponentially decaying

function,

P = P0e−t/τ

(5.3)

where P0 is the maximum pumping effect observed at time 0 and τ is the time constant

of the decay. At 80 mTorr, the time constant of the quenching of the depopulating optical
pumping scheme was found to be (2.25±0.1) ms and the time constant of the quenching of

115

Figure 5.10: Quenching of the optically pumped electronic populations in Zr+. The ions
held in the RFQ ion trap were excited with a pulsed laser and held for increasing lengths of
time. Interactions with the He buffer gas used in the RFQ ion trap leads to cooling of the
trapped ions back to a room temperature Boltzmann distribution.

Figure 5.11: Quenching of the optically pumped electronic populations in the depopulating
pumping scheme of Zr+.

116

Table 5.3: Time constant of the quenching of the optically pumped electronic populations
by collisional decay with the He buffer gas, across a range of pressures. The time constants
of both the depopulating optical pumping scheme and populating optical pumping scheme
are shown. If the time constant was not measured at a specific pressure, ’-’ is shown.

Pressure (mTorr) Depopulating
Scheme Time
Scheme Time
Constant (ms) Constant (ms)

Populating

7.3±1.2
4.4±0.6

10
20
30
40
50
60
80
100
120
140

-

-

8.6±0.5
5.1±0.5
4.6±0.4
2.7±0.3
2.2±0.2
1.9±0.2
2.4±0.3
1.0±0.2

-
-

-

1.3±0.2
1.0±0.3
0.5±0.2

-

-

the populating optical pumping scheme was determined to be (0.76±0.1) ms.

The quenching process involves collisions with the He buffer gas, and is thus sensitive

to the pressure. A measurement was performed where the quenching process was observed

across a range of pressures, and the results were compared. The optical pumping effect as a

function of the release delay is shown for the different pressures that were tested in Figures

5.12 and 5.13 for the depopulating and populating optical pumping schemes respectively.

More detailed information about the time constant of the relaxation process across the

pressures is shown in Table 5.3. The minimum He buffer gas pressure that was tested was 10

mTorr at the He gas inlet. The trapping efficiency of the RFQ ion trap is drastically affected

at lower pressures. The upper limit of the He buffer gas pressure that could be used was

140 mTorr. High buffer gas pressures can cause more frequent high voltage sparks, which

disrupt measurements and can damage equipment. A correction factor was applied to the

measured data based on the observation of ions entering into the ion trap after irradiation

117

Figure 5.12: Optical pumping effect as a function of the release time of the trapped ions
after irradiation by the pulsed laser light across a range of pressures. The quenching of the
optically pumped electronic populations takes a longer time at lower pressures.

of the pulsed laser. The ion rate from the PIG ion source was high enough to overfill the

trap. The ion beam was deflected by the application of 1 kV to a quadrupole immediately

after the PIG ion source. The deflecting voltage was applied for 98 ms, then turned off for

2 ms to allow the ion trap to be filled. It was discovered that the deflecting voltage was not

118

Figure 5.13: Optical pumping effect as a function of the release time of the trapped ions
after irradiation by the pulsed laser light across a range of pressures. The quenching of the
optically pumped electronic populations takes a longer time at lower pressures. The curves
of 50 mTorr, 80 mTorr, and 120 mTorr feature a zoom in of the data.

applied for the measurement of the relaxation time across the range of pressures, and the

ion beam was continuously entering the ion trap. As ions enter the trap after irradiation by

the pulsed laser light the total fluorescent signal will increase and dilute the observed optical

pumping effect.

The correction factor was applied assuming that the ion rate was continuous. The signal

increase from the added ions should be the same between the ions that are optically pumped

and not optically pumped, and will be proportional to the additional time spent in the ion

trap. The ions are held for at least 100 ms, so ions that are held for 10 additional ms will

119

gain 10% more signal from the added ions. The correction factor subtracted photon counts

that are assumed to have been caused by the additional beam, and the optical pumping

effect was calculated based on the corrected counts. The change in the optical pumping was

small, and the corrected data points are compared to uncorrected data in Figure 5.14.

Figure 5.14: Optical pumping effect as a function of release delay of the ions after they
are irradiated by pulsed laser light, comparing corrected and uncorrected data points for
additional ions that enter the trap after the laser pulse. The overall correction is small.

The decay rate due to collisional decay can be determined from the time constant, simply

as the inverse of the time constant. The decay rates of the populating and depopulating

optical pumping schemes as a function of pressure is shown in Figure 5.15. The decay rate

follows a linear relationship with the He buffer gas pressure. The slope of the fit gives the
dependence of the pressure on the collisional quenching rate, and was found to be (5.4±0.2)
Hz/mTorr and (12.6±0.9) Hz/mTorr for the depopulating and populating optical pumping

schemes respectively.

120

Figure 5.15: Collisional decay rate between He and the optically pumped electronic popula-
tions of Zr+. The decay rate follows a linear relationship with the pressure of the He buffer
gas.

5.4.1 Extrapolation of the Optical Pumping Effect

The BECOLA DAQ system features a delay of 350 µs, meaning that even at the fastest

release times there is a delay between irradiation and the release. The 350 µs delay provides

an opportunity for the He buffer gas to significantly cool the trapped ions. The optical

pumping effect immediately after irradiation by the pulsed laser light can be extrapolated

from the relaxation times. It is expected that the extrapolated optical pumping effect would

converge across pressures.

The extrapolation was performed for both optical pumping schemes. The extrapolated

values of the depopulation of the ground state are shown in Figure 5.16, compared to the

experimentally observed values. The experimentally-observed data generally trend towards

zero as the pressure increases. The error bars of the extrapolated data are larger based on

uncertainty in the time constant of the quenching process and the statistical uncertainty of

121

Figure 5.16: Comparison of the extrapolated optical pumping effect with the experimentally
observed optical pumping effect. The results of the depopulating optical pumping are shown
on top and the results of the populating optical pumping scheme are shown on bottom. The
extrapolated data is based on the time constant of the quenching of the optically pumped
electronic populations.

the experimental data. Collisions between the He buffer gas and the trapped ions are more

likely, and the quenching process occurs faster. The difference between the extrapolated

values and the experimental data is relatively small in the depopulating optical pumping

122

scheme, but becomes significant in the populating optical pumping scheme because the rate

of the re-thermalization process is faster.

The extrapolated values were also compared to the results of the optical pumping simu-

lation. The extrapolated values of the depopulation of the ground state are shown in Figure

5.17, compared to the simulated values. In general, the extrapolated values converge across

Figure 5.17: Extrapolated optical pumping effect of the trapped ions immediately after
irradiation by the pulsed laser, as a function of pressure. The simulated values are shown as
orange and blue bands respectively. The experimental data agrees with the simulated value
of the optical pumping effect using experimental branching ratios from Ljung et al. [103]

pressures, with the exception of an outlier at 20 mTorr. The weighted average of the opti-
cal pumping effect is (-27.1±3)%. The experimental data agree with the simulated optical

pumping effect using the experimental branching ratios from Ljung et al. [103].

The extrapolation of the optical pumping effect was repeated for the optical pumping

scheme that increased the population of the ground state, and results are shown in Figure

5.18. The extrapolated values of the optical pumping effect converge across the measured
pressures. The weighted average of the extrapolated optical pumping effect is (16.8±2.9)%.

123

Figure 5.18: Extrapolated optical pumping effect of the trapped ions immediately after
irradiation by the pulsed laser, as a function of pressure. The simulated optical pumping
effect using the Kurucz and Ljung branching ratios are shown as green and orange bands
respectively. The experimental results agree well with the simulated value based on the
Ljung simulation value.

The weighted average value agrees well with the simulated value derived from the experi-

mental branching ratios of Ljung et al. [103].

124

Chapter 6

Discussion

6.1

Improvements to the BECOLA Facility

The infrastructure improvements to the BECOLA facility to enable the optical pumping

technique, the pulsed laser system and the free-space light transport system, were new to

BECOLA and resulted in different lessons that were learned that will assist in guiding future

improvements.

6.1.1 Free-Space Light Transport System

The free-space light transport system was constructed to deliver laser light to the BECOLA

experimental area for both optical pumping and CLS experiments. As the longest arm of the

system is over 10 m, the system was designed to minimize vibrations that would translate

into large movements of the pulsed laser beam position at the target region.

The new free-space light transport system was able to successfully deliver pulsed laser

light to the RFQ ion trap. Movement of the laser light due to vibrations did not have a

measurable effect on the experiments performed for this thesis, even when the beam was

tightened to a small radius.

On the other hand, the performance of the light transport system for the CLS measure-

ment was significantly worse. The CW laser light was kept at approximately 300 µW. The

125

background due to stray light in the CLS beamline was over 300 kHz, as opposed to the less

than 50 kHz background that is possible with fiber optic light transport, where light from

the BECOLA laser room is coupled to a fiber optic cable and delivered to the experimental

area. The increase in the stray light background is thought to have three main contributions.

The first contribution is due to the size of the laser light through the CLS beamline. The

focus of the laser light was controlled via a telescope system in the BECOLA laser room,

which slowly focuses the light to the CLS interaction region. The beam waist was larger

with the free-space light transport system. The measured beam waist was 2 mm in diameter,

whereas with the fiber optic transport system a beam waist of 1.3 mm is possible. The larger

laser beam can be clipped by the aperture system inside the CLS beamline, and stray light

scattered from the aperture will increase the background.

The second contribution is the propagation of higher-order modes of light through the

light transport system, which reflect through the beamline as stray light. Although the

Ti:Sapphire produces a Gaussian TEM00 beam, the WaveTrain system produces light that

interferes with itself as it propagates, producing halos around the main laser beam, as seen

in Figure 6.1. These halos are typically blocked during injection into the fiber optic system,

but can propagate freely with the main laser beam through the free-space light transport

system. The higher-order modes of light will focus at different distances after passing through

a lens. The free-space light transport system relies on a lens telescope system to collimate

the laser beam, and any slight change in the focal length for each mode of light results in

spatial broadening of the higher-order modes of light. As the higher-order modes of light

diverge through the free-space light transport system, they can clip on components such as

the metal pipe used to safely contain the laser light between rooms. Some proportion of the

reflections from the light clipping on a surface are able to transverse the entire free-space

126

Figure 6.1: Frequency doubled CW laser light produced by a Sirah WaveTrain system. The
frequency doubled light was measured approximately 2 m from the WaveTrain system. No
lenses were used to focus or collimate the light. The doubled laser light features higher order
modes of light seen as halo structures. The higher order light can propagate through the
free-space light transport system and induce high stray light background.

127

light transport system and are injected into the CLS beamline, where they contribute to a

higher background. The background produced by the higher-order modes of light can be

partially mitigated by using a iris to block the halos while leaving the center untouched,

but reflections of the higher-order modes from the iris is possible and contributes to the

background. Another possible future solution is to coat the inside of the 10-m long pipe

with anti-reflection paint that can reduce the scattered light.

The third contribution is movement of the laser beam due to vibrations. Although not an

issue for the pulsed laser light injected into the RFQ ion trap, the CW laser light is carefully

aligned through an aperture system that blocks stray light. The laser light is reflected from

two mirrors on a breadboard that is secured to a building support beam in the BECOLA

laser room, and these optics are responsible for the most significant sources of vibration in

the free-space light transport system. If the air conditioning unit is turned on, or the door

to the laser room opens, or heavy foot traffic passes through the hallway separating the

BECOLA laser and experimental rooms, these optics will vibrate. Although the standard

deviation of the movement is 337 µm, spikes in the movement can cause the laser beam to

clip the apertures and induce spikes in the background.

The background rates observed with the free-space light transport system do not inhibit

the study of isotopes produced by off-line ion sources since the fluorescence signal is large,

but would severely limit the range of radioactive isotopes that can be studied. There are

possible ways to mitigate the vibration and reduce the background. Active stabilization

systems can be installed to monitor the position of the laser light through the free-space

light transport system and counteract movements of the laser beam due to vibrations or any

other source. An active stabilization technique has been employed at GSI to transport laser

light over 80 m [108].

128

In combination with the active stabilization system, a mode filter could be used in the

BECOLA experimental area to remove higher order modes of light from the beam. The mode

filter works by focusing the laser light through a small hole on the order of µm. The higher

order modes of light will not focus to a tight spot, and the lowest order Gaussian mode can

be selected and cleaned. The mode filter technique was attempted for the commissioning

experiments of the optical pumping system, but the transmission of the laser light through

the mode filter was very sensitive to the movement of the laser light and the power could

not be stabilized for spectroscopy measurements.

6.1.2 Characteristics of the Pulsed Laser System

The Cobra-Stretch pulsed laser system was able to successfully excite and saturate the optical

pumping transition in the RFQ ion trap and many additional findings were observed that

will benefit future studies at BECOLA. Initial assumptions made during development of the

optical pumping simulation led us to underestimate the sensitivity of the temporal width of

the pulsed laser on the optical pumping effect.

The Cobra-Stretch pulsed laser delivers pulses with a temporal width of approximately

6 ns. The optical pumping effect becomes very sensitive to the width of the temporal laser

pulse as the lifetime of the excited state approaches the temporal laser width. At these time

scales, the ions can be excited multiple times by the pulsed laser, magnifying the pumping
effect. The 4G5/2 state of Zr+ has a lifetime of (7.1±0.3) ns and the optical pumping effect

as a function of the temporal width of the laser pulse is shown in 6.2. It was assumed that the

optical transition remained saturated. The magnitude of the optical pumping effect grows

very quickly and the ground state is almost completely depopulated with a laser pulse of

100 ns.

129

Figure 6.2: Simulated optical pumping effect as a function of the temporal width of the
laser pulse. Assuming that the transition remains saturated, the ground state of Zr+ can be
almost completely depopulated with a laser pulse of approximately 100 ns in width.

A 10 Hz pulsed laser was used for the commissioning experiments, but the use of higher

repetition rate lasers or CW laser would provide the largest change in the electronic popu-

lation. A limitation of the use of CW lasers is the energy density of the light that is used

to perform optical pumping. The low energy density will limit the wavelength range of the

CW laser system, and subsequently the optical pumping schemes that can be explored. The

current BECOLA CW laser system has a lower limit of 350 nm and of the first- and second-

row transition metals, the optical pumping schemes of all elements except Sc, Ti, and Zr

require laser light below 350 nm.

There are currently efforts towards the development of 4th harmonic CW laser light at

BECOLA, but currently only a few hundred µW of laser power has been produced at 230 nm,

and is not enough power for efficient optical pumping. Additionally, if 4th harmonic CW laser

was produced at BECOLA, it would likely be experimentally simpler to use that laser system

130

and transition for CLS measurements, as the CW laser light could access the populated

low-lying states and would not rely on favorable decays which could reduce the sensitivity.

Further development will open opportunities for laser spectroscopy measurements, and with

the optical pumping system, will provide access to multiple techniques, allowing the choice

of one that is best-suited for a given element or experiment.

6.2 Optical Pumping in the RFQ Ion Trap

The results obtained from measurements of the effective linewidth and saturation curve

indicate that the pulsed laser light efficiently performed optical pumping.

6.2.1 Ion Bunch Size

It was assumed that the spatial distribution of the ion bunch in the RFQ ion trap had

a radius of approximately 1.4 mm, based on results from simulations performed with the

3DCylPIC software [80]. A rough measurement of the spatial distribution of the trapped

ions in the X-Y plane of the RFQ ion trap was performed and the 95% inclusion radius was
determined to be (2.3±0.2) mm (See Section 5.1.1).

The results confirm that the ion bunch is small, although the measurement was hampered

by hysteresis of the picomotors used to scan the pulsed laser beam across the ion bunch.

The size of the laser beam was increased to roughly the same size of the trapping region

defined by the RFQ rods, with a diameter of 7 mm. The large size of the laser beam ensures

that the entire ion bunch is irradiated by the laser light and mitigates small movement of

the laser light due to vibrations in the free-space light transport system.

The trapped ion distribution in the simulation was not sensitive to the He buffer gas

131

temperature parameter. The BECOLA RFQ ion trap is designed to operate at liquid N2

temperatures, and further work should be done to determine if the trapped ion distribution

will change at low temperature.

6.2.2 Laser Power

The optical pumping effect as a function of the pulsed laser power was performed (See

Section 5.1.2) The results demonstrated that the BECOLA pulsed laser system was capable
of saturating the optical transition, with a saturation power of (2.9±0.3) mW/cm+. The
theoretical saturation power of (8±0.2) mW/cm2 was much lower than the total available

power from the Cobra-Stretch laser system, and can be used to estimate the limits of optical

pumping schemes that can be explored at BECOLA.

The saturation power is inversely proportional to the lifetime of the excited state [81],

such that

Isat =

πhc
3λ3τ

(6.1)

The Cobra-Stretch laser system is capable of producing 120 mW of laser light at 360 nm for

a power density of approximately 109 mW/cm2, 37 times greater than the saturation power
density. The lifetime of the excited state in the commissioning experiments was (7.1±0.3)

ns and saturation can be achieved with transitions that are 37 times slower, or with lifetime

on the order 265 ns. A lifetime of 265 ns corresponds to an Einstein coefficient A of 3.6 106,

while a lifetime of 7.1 ns corresponds to 1.4 107. In Table 6.1 are presented the minimum

Einstein coefficients for future optical pumping experiments using similar conditions as the

commissioning experiments described here and based on the available laser powers at different

wavelengths. Light at higher wavelengths does not need to undergo the doubling process

132

Table 6.1: Einstein coefficient A limits based on saturation of the transition using similar
conditions as the commissioning experiments, namely a laser beam size of 0.27 cm2. The
wavelength ranges are those that were tested during installation of the pulsed laser system.

Wavelength (nm)

Laser Power

Density

(mW/cm2)

Minimum Einstein
Coefficient A (s-1)

196

202-212
230-270
360-365

535

7
98
185
435
1270

2.2 108
1.6 107
8.7 106
3.7 106
1.3 106

after being produced in the Cobra-Stretch dye laser, and will thus have much higher energy

per pulse and allow access to slower transitions.

6.2.3 Linewidth

Good spectral overlap between the absorption linewidth of the electronic transition and the

linewidth of the pulsed laser light is necessary for efficient optical pumping. If the linewidth

of the pulsed laser is smaller than the absorption linewidth, a proportion of the ions will not

be excited. If the linewidth of the pulsed laser is too large, a proportion of the laser power

will be distributed off-resonance with the absorption linewidth and is effectively wasted. In

Chapter 4, it was shown that at room temperature the absorption linewidth of an electronic

transition is dominated by Doppler broadening effects, and results in an absorption linewidth
of approximately 1.1 GHz, or approximately 0.04 cm−1. The linewidth of the fundamental
light of the pulsed laser system is also 0.04 cm−1, but is doubled to 0.08 cm−1 when 2nd

harmonic light is produced, significantly larger than the expected absorption linewidth. The

effective absorption linewidth was determined by measuring the optical pumping effect as a

function of the pulsed laser frequency. The measurement was repeated at different pulsed

133

Figure 6.3: Power broadening of spectral lines. As the laser power increases, the wings of the
spectral profile raise and can excite transitions far from the center of the center frequency.

laser powers and the resonance linewidth was extracted. The absorption linewidth was

observed to significantly grow at laser powers as low as 4 mW. At higher laser powers a

flat top profile was observed as the electronic transition was saturated by laser light even

though the pulsed laser frequency was centered off resonance. The increase in the effective

absorption linewidth is due to what is known as power broadening. The power broadening

phenomenon is illustrated in Figure 6.3. If we assume that the spectral line profile of the

pulsed laser follows a Lorentzian distribution, as the laser power is increased the wings of

the profile will raise and can excite a transition even if the peak of the spectral profile is off

resonance. The observed power broadening and flat-top behavior confirmed that the pulsed

laser is capable of completely exciting the electronic transition for all ions in the RFQ ion

trap.

134

6.3 Manipulation of Electronic Populations

The primary goal of the commissioning experiments was to demonstrate that the pulsed laser

system is capable of manipulating the electronic populations of ions in the BECOLA RFQ

ion trap. The population manipulation was performed in the RFQ ion trap and probed via

measurement of the fluorescence signal with CLS. The results of the commissioning experi-

ments clearly demonstrated that the population manipulation was successful. A statistically-

significant change was observed in the fluorescence signal, and subsequently in the electronic

population of the electronic state being probed. The observed optical pumping effects of
(-24.6±1.6)% for the depopulation optical pumping scheme and (12.6±2.7)% were in better

agreement with the results of the optical pumping simulation based on the experimental

branching ratios provided by Ljung et al.[103].

The branching ratios from the Kurucz database are based on calculations from emission

spectra from free-burning arc sources [102]. It was assumed that the oscillator strengths can

be described by a linear relationship between the wavelength of the transition, the energy

of the excited state, and the intensity of the measured peak. The difference between the

experimental results and the Kurucz calculations may be due to both experimental error in

the measurement of the emission spectra, and a non-linear dependence between the oscillator

strength and the parameters used to calculate the oscillator strength. The experimental

branching ratios from Ljung et al. are based on direct measurement of Zr+ emission spectra

and should be more reliable.

The difference between the results of the simulation and the experimental observations

can be explained by the occurrence of collisional excitation and decay that redistributed the

electronic populations of ions in the RFQ trap back to thermal equilibrium.

135

6.3.1 Quenching of Optically Pumped Populations

The observed optical pumping effect for both the depopulating and populating transitions

were smaller than expected based on the developed optical pumping simulation, for both the

simulated values using the Kurucz theoretical branching ratios [75] and the Ljung experi-

mental branching ratios [103]. The discrepancy between the observed optical pumping effect

and the simulated value can be explained by the observation of collisional decay inside the

RFQ ion trap. The He buffer gas quenched the optically pumped electronic populations and

re-thermalized the ion bunch back to a room temperature Boltzmann distribution. Although

the BECOLA DAQ system was set to release the ion bunch immediately after irradiation by

the pulsed laser light, there is a 350 µs delay between the laser pulse and the release of the

ion bunch. In the 350 µs time frame, the He buffer gas began to quench the optical pumping

effect. Measurement of the time constant of the decay provided a method to extrapolate the

optical pumping effect immediately after irradiation by the pulsed laser.

The extrapolated optical pumping effect of the depopulation of the ground state of Zr+
was (-27.1±3)%. The optical pumping simulation predicted a value of (-31.9+4−3)% based on
the experimental branching ratios found in Ljung et al., and (44.1±3)% using the theoretical

branching ratios from the Kurucz database. The extrapolated optical pumping effect of
the population of the ground state of Zr+ was (12.6±2.7)%, while the simulated values
were (14.8±1)% and (8.3±1.5)% using the experimental and theoretical branching ratios

respectively. The extrapolated data of both optical pumping schemes agree well with the

simulated optical pumping effect using the experimental branching ratios from Ljung et al..

The theoretical branching ratios from the Kurucz database significantly overestimate the

fraction of excited state ions that decay to the metastable dark states.

136

The effect of collisional excitation and decay on the optical pumping effect was measured

across a range of He buffer gas pressures, from 10-140 mTorr measured at the gas inlet.

The rethermalization time decreased with the He buffer gas pressure, as the probability of

a collision between a Zr ion and a He atom became smaller, and a difference in the rether-

malization time was observed between the depopulating and populating optical pumping
schemes. The rates were fitted to a linear regression line, with slopes of (5.4±0.2) Hz/mTorr
and (12.6±0.9) Hz/mTorr for the depopulating and populating optical pumping schemes,

respectively. The difference in the slopes between the two optical pumping schemes can be

explained by the difference in probability between a collisional excitation and a collisional

decay. In Chapter 3, it was shown that the thermal populations of the electronic states can

be described by a Boltzmann distribution. The ions in a system populate the low-energy

states through collisional excitation and decay, which will eventually reach equilibrium, such

that the relative population Ni is

Ni =

2J + 1

Q

e−Ei/kT .

(6.2)

In a two-level system at equilibrium with levels i and j, the rate of change in the population

of state i can be given by the rate equation,

δNi = NjKji − NiKij = 0

(6.3)

where Kij is the collisional excitation or decay rate from state i to state j. By substituting

137

Eq. 6.2 into Eq. 6.3, it can be shown that,

2Ji + 1

Q

e−Ei/kT Kij =

2Jj + 1

Q

−Ej /kT
e

(6.4)

The degeneracy of the states, g, is given by the 2J + 1 term, and by simplifying the equation

we find,

gie−Ei/kT Kij = gje

−Ej /kT Kji

(6.5)

Finally, Equation 6.5 can be further simplified to show that the rates of collisional excitation

and decay are related by

Kij =

gj
gi

−(Ej−Ei)/kT Kji

e

(6.6)

The two optical pumping schemes involved optically pumping ions from the ground or

first excited states of Zr+. Assuming that the ground and first excited states are responsible

for most of the collisional decay and excitation during the rethermalization process, it can

be shown that the rate of the rethermalization process should differ by a factor of 2.7 with

a gi of 1.5 and gj of 2.5. The slopes of the rethermalization process for the populating and
depopulating transition differ by a factor of (2.3±0.2), agreeing well with the expected value

and demonstrates why the difference occurs.

The fit of the collisional rate was also used as an input to the optical pumping simulation

to incorporate the collisional rethermalization and check the experimental results. The

simulated optical pumping effect is shown as a function of He buffer gas pressure in Figure

6.4, compared with the experimental data. The simulated optical pumping effect is based

on the experimentally-determined branching ratios from Ljung et al.. The uncertainty of

the simulated values results from the experimental error of the branching ratios from the

138

Figure 6.4: Comparison of the experimental optical pumping effect as a function of pressure
with the simulated optical pumping effect with collisional rethermalization added. The
simulation shows good agreement with the experimental results and demonstrates that the
rethermalization process can accounted for in simulation.

excited state as well as the uncertainty of the fit of the collisional rethermalization rate. The

simulated values agree well with the experimental data, demonstrating that the collisional

rethermalization can be accounted for in future experiments through simulation. Future

preparatory experiments should test the collisional rethermalization as a function of time to

improve the simulation, as it is unclear if the collisional rate will be element-dependent.

6.4 Outlook

The results of the optical pumping simulation for the first- and second-row transition metals

shows promise for future studies of many elements. During the FRIB era, BECOLA will

be in a unique position to gain access to rare isotopes of the transition metals at very high

rates, but future experiments will need to verify optical pumping schemes during preparatory

139

experiments.

6.4.1 Optical Pumping Simulation Accuracy

The discrepancy between experimental results and the predictions based on the Kurucz

database present an issue for future experiments using the optical pumping system. As

a tool for planning future experiments, it is important that the estimates of the increase

in electronic population given by the optical pumping simulation be accurate. The results

shown in Chapter 3 for the first- and second-row transition metals are based on the NIST

and Kurucz databases. Currently, the optical pumping schemes of V, Nb, Mo, and Ru are at

least partially based on theoretical values of the transition scheme. The NIST line database

does not include the Einstein coefficients of these elements, and preparatory experiments

should test multiple optical pumping schemes to determine the most efficient scheme. Fu-

ture use of the simulation for planning experiments should investigate the most promising

optical pumping schemes to verify the predictions from the optical pumping simulation. The

preparation can be performed with stable ion beams and off-line ion sources, and multiple

schemes can be tested to find the most suitable. Literature searches can also be performed

to find the most up-to-date data behind the transitions that would be used.

6.4.2 FRIB Reach

Based on the statistics of a previous experiment performed at BECOLA to study the iso-

tope shifts of 52,53Fe [12], an estimate of BECOLA’s reach in the FRIB was performed.

The estimate was based on the statistical precision and length of measurement time of the

neutron-deficient Fe isotopes. The measurement of 52Fe required 13 hours of collection time

140

Figure 6.5: Chart of the nuclides zoomed in on the first- and second-row transition metals.
Isotopes in black are stable, isotopes in blue are radioactive and have previously been studied
using laser spectroscopy, isotopes in green will be accessible for studies at BECOLA, and
isotopes in pink have previously been studied at BECOLA.

for a statistical error of 3 MHz, or approximately 0.2% of the overall isotope shift of 1.836

GHz. At the 3 MHz level of statistical error, the systematic error becomes a limiting factor

for precision. The average rate of the 52Fe beam over the 13 hour collection time was 4.4 105

particles/s, for a total of 2.06 1010 particles for the 13 hour collection time. The BECOLA

charge exchange cell was used to neutralize the beam at an efficiency of 50%, and only 5% of

the neutralized beam populated the electronic state used for the CLS measurement, leading

to an estimate of 5.2 108 total ions required for a 3 MHz uncertainty.

Using estimates of the final beam rates for the first- and second-row transition metals

provided by FRIB and based on the LISE++ software [50], the experimental time necessary

to achieve a 3 MHz uncertainty was calculated for each isotope. A cutoff of 5 days was

applied, as that is the typical length of an online experiment. The results of the estimate are

shown in Figure 6.5 with more detail in Table 6.2. During the FRIB era, BECOLA will be

able to study essentially every element in the transition metal region far from stability. One

141

Table 6.2: Experimental reach of the first- and second-row transition metals at BECOLA
during the FRIB era. BECOLA will be in a unique position to study wide chains of virtually
every element in the region.

Element Mass Number

Sc
Ti
V
Cr
Mn
Fe
Co
Ni
Cu
Zn
Y
Zr
Nb
Mo
Tc
Ru
Rh
Pd
Ag
Cd

39-56
40-58
43-60
44-63
46-66
48-70
50-72
53-74
56-77
58-78
77-104
79-107
82-109
85-112
N/A
88-118
91-119
94-122
96-124
98-128

exception is the case of Tc, where studies are limited by sparse knowledge of the electronic

structure and not by beam rate or technological limitations at BECOLA. Due to the high

beam rates, even elements that require charge exchange reactions for CLS measurements

will be accessible.

6.4.3 Future Laser Systems

Future optical pumping experiments will be able to take advantage of three new pulsed laser

systems that will be installed at BECOLA. The pulsed laser systems are 10 kHz repetition

rate Ti:Sapphire lasers that were previously used at the Holifield Radioactive Beam Facility

142

Figure 6.6: Simulated optical pumping effect of the depopulating optical pumping scheme in
Zr+. The 10 kHz laser system is able to deliver more laser pulses to the ion bunch, but more
laser pulses are required to depopulate the ground state due to the lower energy density of
the laser light.

[109] located at Oak Ridge National Lab prior to being decommissioned. The pulsed lasers

were used as part of a Resonant Ionization Laser Ion Source (RILIS) system and were

described in Liu et. al [110]. Two of the three laser systems are equipped with units to

generate 2nd, 3rd, and 4th harmonic light and have a wavelength range of 217-960 nm. The

systems can produce up to 30 mW of laser light at 217 nm, 120 mW of light at 270 nm,

800 mW at 410 nm, and 2.5 W at 820 nm, which are the peak powers for the fundamental

and harmonic light respectively. Although the overall power of the system is comparable to

the Cobra-Stretch laser system, the laser power is distributed over 1000 times more laser

pulses. Assuming that the overall laser power and temporal width of the two systems are

the same and that the difference that the energy per pulse of the 10 kHz laser system is a

factor of 1000 lower, the optical pumping effect was simulated and is shown in Figure 6.6.

Although the 10 kHz laser system would be able to irradiate the ion bunch many times,

143

the low energy density of the laser light prevents saturation of the transition. The optical

pumping effect after 100 ms is -42.9%, which is a significant improvement over the simulated

and experimentally observed optical pumping effect using the 10 Hz Cobra-Stretch laser

system.

Another benefit of using the 10 kHz laser system is an increase in flexibility regarding

the amount of time that ion bunches can be held in the RFQ ion trap. Currently, the 10 Hz

laser system requires that the ions be held for at least 100 ms to irradiate each ion bunch. A

10 kHz pulsed laser system would allow for ions to be held for shorter periods of time, and

periods of time that are not multiples of 100 ms.

Finally, the 10 kHz pulsed laser system could potentially mitigate the effect of collisional

decay on the observed optical pumping effect. Each laser pulse is separated by approximately

100 µs, compared to the 100 ms separation time between laser pulses with the Cobra-Stretch

pulsed laser system. The collisional rethermalization process was observed to occur over the

time span of a few ms, and the 10 kHz laser system will be able to continuously excite the

laser transition and prevent the ion bunch from returning to thermal equilibrium.

144

Chapter 7

Conclusion

Laser spectroscopy studies of the first- and second-row transition elements can provide im-

portant information for nuclear structure and stockpile stewardship. However, radioactive

isotopes of the transition metals have remained difficult to study. An optical pumping system

was commissioned at the BECOLA facility to enable the study of rare transition metal iso-

topes. Through optical pumping, the electronic populations of ions held in an RFQ ion trap

were manipulated and improved the sensitivity of subsequent collinear laser spectroscopy

measurements.

An optical pumping simulation was developed to guide the experiments performed in

this thesis, as well as future experiments. The simulation solves the rate equations of the

electronic states to determine the most efficient laser transitions for optical pumping studies.

The results of the simulation indicate that most transition metals will see an improvement

in measurement sensitivity through optical pumping.

A new pulsed laser system and a free-space light transport system were installed at the

BECOLA facility to deliver high-power pulsed laser light to the BECOLA RFQ ion trap and

perform the optical pumping. Commissioning experiments involving a stable Zr+ indicate

that the electronic populations of ions held in the RFQ ion trap can be successfully manip-

ulated. The observed optical pumping effect was in good agreement with the simulation,

bolstering confidence that the simulation can be used for future experiments.

An unexpected contribution of collisional decay was observed, quenching the optical

145

pumping effect as collisions between the optically pumped Zr+ ions and the He buffer gas

rethermalized the electronic populations back to room temperature. The rethermalization

process was measured over a range of pressures and the results were used to both extrapolate

the experimental results to determine the initial optical pumping effect, and to add the

collisional process to the optical pumping simulation. The extrapolated data converged

across the tested range of pressures and was in good agreement with the optical pumping

simulation. The experimental data also agreed well with the results of the optical pumping

simulation with the collisional process added. Through the commissioning experiments, the

collisional decay effect is well understood and can be accounted for in future experiments.

In the FRIB era, BECOLA is well situated to study essentially all elements in the transi-

tion metal region, and with the addition of three new pulsed laser systems, there is potential

for further development of the optical pumping program.

146

APPENDICES

147

APPENDIX A

Optical Pumping Schemes of First

and Second-Row Transition Metals

This appendix contains the potential optical pumping schemes for the first- and second-row

transition metals, as calculated using the optical pumping simulation discussed in Chapter

4. Each table is split into sections with a potential pumping step, featuring the ground

and excited states and properties of the corresponding transition, and different possible

CLS transitions that can be used. The optical pumping schemes are in order of decreasing

electronic population of the state being probed for the CLS measurement, and within each

scheme the possible transitions for CLS are sorted in order of the speed of the transition.

It was assumed that no collisional decay will occur for the calculation of the change in the

electronic population, and should be seen as the maximum possible optical pumping effect.

A.1 Scandium

Promising optical pumping schemes of Sc are shown in Figure A.2. The first three states of

Sc are populated, with the relative populations shown in Table A.1.

148

Table A.1: Low-energy populated states of Sc+

State (cm−1) Population

0

67.72

177

0.312

0.376

0.311

The first three states of Sc+ are almost evenly populated, but the first excited state at

67 cm-1 is the most populated and the most promising optical pumping schemes move the

electronic population to that state.

149

Table A.2: Optical Pumping Schemes of Sc+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

CLS Ground State

Population in

CLS Excited State

(cm−1)

State

(cm−1)

A (s−1)

177

27602

1.43E7

362.3

0

0

39115

7.11E7

255.6

29742

3.19E7

336.2

67.72

27443

2.51E7

365.4

67.72

67.72

67.72

67.72

0

0

67.72

67.72

67.72

67.72

67.72

177

177

67.72

67.72

67.72

67.72

67.72

177

177

0

0

0

177

177

0.605

0.605

0.605

0.605

0.505

0.505

0.581

0.581

0.581

0.581

0.581

0.312

0.312

0.53

0.53

0.53

0.53

0.53

0.312

0.312

0.505

0.505

0.505

0.312

0.312

29742

28021

29742

27917

27443

27917

30815

27602

28021

29742

27917

27841

28161

27602

28021

29742

27917

29823

27841

28161

27443

27917

29736

27841

28161

1.58E8

1.41E8

1E8

5.09E7

1.52E8

1.48E8

3.33E6

1.58E8

1.41E8

1E8

5.09E7

1.88E8

1.83E8

1.58E8

1.41E8

1E8

5.09E7

2.01E7

1.88E8

1.83E8

1.52E8

1.48E8

1.38E8

1.88E8

1.83E8

A.2 Titanium

The promising optical pumping schemes of Ti+ are shown in Table A.4, with the distribution

of the electronic population shown in Table A.3.

150

Table A.3: Low-energy populated states of Ti+

State (cm−1) Population

0

94

225

393

908

983

0.312

0.376

0.311

0.223

0.004

0.004

Although Ti+ features 6 states below 1000 cm-1, only four are appreciably populated.

The study of Ti isotopes with CLS does not require charge exchange, but optical pumping

can provide a 52% improvement in the sensitivity of laser spectroscopy measurements.

151

Table A.4: Optical Pumping Schemes of Ti+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

CLS Ground State

Population in

CLS Excited State

(cm−1)

State

(cm−1)

A (s−1)

94.1

29544

2.48E7

339.5

225.7

29544

2.48E7

341.1

0

32602

3.3E7

306.7

94.1

30836

3.38E7

325.3

225.73

29734

2.18E7

338.9

0

0

0

0

0

0

0

0

94.1

94.1

94.1

94.1

94.1

0

0

0

0

94.1

94.1

94.1

94.1

94.1

0.503

0.503

0.503

0.503

0.488

0.488

0.488

0.488

0.472

0.472

0.472

0.472

0.472

0.471

0.471

0.471

0.471

0.434

0.434

0.434

0.434

0.434

32532

30836

29544

32602

32532

30836

29544

30958

32602

29734

30958

31113

32025

32532

30836

29544

32602

32602

29734

30958

31113

32025

1.54E8

1.16E8

1.09E8

3.3E7

1.54E8

1.16E8

1.09E8

2.87E7

1.12E8

1.11e8

3.38E7

2.66E7

8.21E6

1.54E8

1.16E8

1.09E8

3.3E7

1.12E8

1.11e8

3.38E7

2.66E7

8.21E6

A.3 Vanadium

The promising optical pumping schemes of V+ are shown in Table A.6 and the low-energy

populated states are shown in Table A.5.

152

Table A.5: Low-energy populated states of V+

State (cm−1) Population

0

36

106

208

339

0.092

0.233

0.276

0.237

0.163

There are many low-energy states of V+ and the electronic population is already highly

fragmented at room temperature. Additionally, the study of V typically requires charge

exchange reactions, which result in only 4% of the total ion beam populating the state used

for CLS. Considering a charge exchange efficiency of 50%, only 2% of the total ion beam is

available for study. For each optical pumping step, there are relatively few transitions that

can be used for CLS measurements.

Table A.6: Optical Pumping Schemes of V+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

106.6

106.6

339

106

106

46754

37369

47181

46690

46740

6.47E7

214.3

1.98E7

268.3

1.24E7

213.5

1.08E7

214.7

1.08E7

214.4

A (s−1)

3.47E6

3.47E6

1E6

1.45E6

1E6

CLS Ground State

Population in

CLS Excited State

State

0.335

0.292

0.154

0.124

0.115

(cm−1)

36489

36489

37369

37259

37520

(cm−1)

36.1

36.1

13594

13511

13594

153

A.4 Chromium

The promising optical pumping schemes of Cr+ are shown in Table A.7. All ions are found in

the ground state at room temperature, requiring that all optical pumping schemes start from

that state. Traditional CLS measurements of Cr require atomic charge exchange reactions to

neutralize the beam, with 12% of the total population in the electronic state that is probed.

A large gain in sensitivity is expected through optical pumping of Cr.

Table A.7: Optical Pumping Schemes of Cr+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

0

0

0

0

48632

2.09E8

205.6

48491

1.63E8

206.2

49005

1.06E7

204.1

49351

7.65E6

202.6

CLS Ground State

Population in

CLS Excited State

(cm−1)

12496

12496

12147

12303

12303

12147

12147

19797

19797

19797

20024

20024

19797

19797

State

0.586

0.586

0.293

0.405

0.405

0.218

0.218

0.329

0.329

0.329

0.282

0.282

0.134

0.134

(cm−1)

48632

47751

47040

1.63E8

47227

47040

48398

49005

49564

49351

1.38E8

49838

49005

49544

A (s−1)

2.22E8

1.99E8

1.19E8

6.27E7

1.19E8

9.81E7

1.39E8

1.1E8

4.5E7

1.98E6

1.39E8

1.1E8

A.5 Manganese

Promising optical pumping schemes for Mn+ are shown in Table A.8. The entire electronic

population is found in the ground state of Mn+. Typically atomic charge exchange reactions

are required for CLS measurements, with 12% of the total ion population in the state being

probed. There are relatively few possible optical pumping schemes, with only transition is

154

suitable for the pumping step of the scheme.

Table A.8: Optical Pumping Schemes of Mn+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

0

43370

1.55E6

230.5

CLS Ground State

Population in

CLS Excited State

(cm−1)

9472

9472

9472

14325

State

0.584

0.584

0.584

0.136

(cm−1)

43557

43484

43370

43370

A (s−1)

1.96E8

1.86E8

1.86E8

4.32E7

A.6

Iron

The promising optical pumping schemes of Fe+ are shown in Table A.10, with the electronic

population of the low-energy states shown in Table A.9.

Table A.9: Low-energy populated states of Fe+

State (cm−1) Population

0

384

667

862

977

0.863

0.109

0.021

0.006

0.002

The vast majority of the electronic population is in the ground state of Fe+. Typical

CLS measurements require atomic charge exchange reactions, where only 5% of the total

ion beam is in the electronic state being probed. There are only two transitions from the

low-energy states that can be used for optical pumping. Both transitions pump ions into

the electronic state located at 7955 cm+, from which there are only a handful of transitions

155

suitable for laser spectroscopy measurements. A very large gain in sensitivity is expected

with optical pumping.

Table A.10: Optical Pumping Schemes of Fe+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

0

0

44232

4.85E6

226.1

44446

4.14E6

224.9

CLS Ground State

Population in

CLS Excited State

(cm−1)

7955

7955

7955

7955

State

0.564

0.564

0.444

0.444

(cm−1)

44232

44446

44232

44446

A (s−1)

2.11E8

1.13E8

2.11E8

1.13E8

A.7 Cobalt

The promising optical pumping schemes of Co+ are shown in Table A.12, with the populated

low-energy electronic states shown in Table A.11.

Table A.11: Low-energy populated states of Co+

State (cm−1) Population

0

950

0.998

0.002

Essentially all of the electronic population is found in the ground state. CLS measure-

ments typically require atomic charge exchange reactions, resulting in only 5% of the atoms

in the state being probed. Many optical pumping schemes of Co+ require very deep UV light

but it remains a very promising case for the optical pumping system, as the gain in sensitivity

is very large in the schemes that are accessible to the BECOLA pulsed laser system.

156

Table A.12: Optical Pumping Schemes of Co+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

0

0

0

48556

1.67E7

205.9

49697

49348

7.68E7

201.2

9.83E7

202.6

CLS Ground State

Population in

CLS Excited State

(cm−1)

9812

9812

10708

9812

9812

State

0.692

0.692

0.644

0.589

0.589

(cm−1)

45197

45378

45972

45197

45378

A (s−1)

1.7E6

1.39E6

1.01E6

1.7E6

1.39E6

A.8 Nickel, Copper, and Zinc

No suitable optical pumping schemes were identified for Ni, Cu, and Zn. The most suitable

ground state transitions of Ni and Cu require 150 nm light, and will not be experimentally

possible at BECOLA for the foreseeable future. There is a ground state transition of Zn+,

exciting the 3d104s2S1/2 to 3d1042

pP o

3/2 transition and requiring 202.5 nm, but it is a cyclical

transition and does not feature branches from the excited state to metastable states accessible

to the BECOLA CW laser system.

A.9 Yttrium

The promising optical pumping schemes of Y+ are shown in Table A.14 with the room

temperature electronic population distribution shown in Table A.13.

Table A.13: Low-energy populated states of Y+

State (cm−1) Population

0

840

0.949

0.051

157

Almost all of the electronic population is found in the ground state of Y+. There are

multiple transitions suitable for CLS from the ground state, and atomic charge exchange

reactions are not necessary. Optical pumping can benefit the study of Y+ because the ground

state has electronic angular momentum J = 0 and the HFS does not feature splitting. The

nuclear moments cannot be deduced without the hyperfine splitting, but by pumping the

electronic population to another state, the hyperfine splitting can be observed.

Table A.14: Optical Pumping Schemes of Y+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

0

0

0

28595

3.46E7

349.7

32124

1.32E6

311.3

44568

1.22E8

224.3

CLS Ground State

Population in

CLS Excited State

(cm−1)

840

840

8003

8003

1045

1045

1045

840

840

3296

3296

14098

14098

State

0.697

0.697

0.224

0.224

0.498

0.498

0.498

0.252

0.252

0.438

0.438

0.375

0.375

(cm−1)

32048

28595

28595

27227

32124

28730

27532

32048

28595

33336

26147

44568

32283

A (s−1)

2.77E8

1.13E8

3.93E7

1.33E7

2.05E8

1.32E8

1.08E8

2.77E8

1.13E8

1.16E8

1.01E8

2.32E8

1.16E7

A.10 Zirconium

The promising optical pumping schemes of Zr+ are shown in Table tab:Zr and the electronic

population distribution is shown in Table A.15. In contrast to the other elements in this

work, data about the transitions is not based on the Kurucz database. The experimental

branching ratios from Ljung et al.

[103] were used instead, as the experimental data was

found to better describe the results of the commissioning experiments.

158

Table A.15: Low-energy populated states of Zr+

State (cm−1) Population

0

314

767

0.723

0.239

0.037

There are suitable ground state transitions in Zr+ without optical pumping, but a modest

gain in sensitivity can be gained through optical pumping. All optical pumping schemes

involve pumping ions from the first excited state located at 314 cm-1 to the ground state,

from which CLS measurements can be performed.

Table A.16: Optical Pumping Schemes of Zr+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

314

314

314

27983

2.92E7

361.4

30435

7.67E6

331.9

29505

4.38E6

342.6

CLS Ground State

Population in

CLS Excited State

(cm−1)

0

0

0

0

0

0

State

0.830

0.830

0.801

0.801

0.771

0.771

(cm−1)

27983

36451

27983

36451

27983

36451

A (s−1)

1.34E8

1.8E8

1.34E8

1.8E8

1.34E8

1.8E8

A.11 Niobium

The promising optical pumping schemes of Nb+ are shown in Table A.18 and the electronic

population distribution is shown in Table A.17.

159

Table A.17: Low-energy populated states of Nb+

State (cm−1) Population

0

158

438

801

0.316

0.443

0.193

0.047

The electronic population is spread among four low-energy states, from which there are

limited options for optical transitions. Available transitions have weak oscillator strengths,

leading to low detection efficiency Through optical pumping the electronic population can

be moved to a state from which a strong transition can be probed. The first demonstration

of optical pumping for electronic population manipulation for CLS measurements was used

to study rare Nb isotopes [28] and the optical pumping simulation identifies the optical

pumping scheme used in that experiment as the most promising. Although the low-energy

states feature weak oscillator strengths, two optical pumping schemes are included that

probe the first excited state of Nb+, as a consideration due to its relatively high electronic

population.

160

Table A.18: Optical Pumping Schemes of Nb+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

438

0

158

37797

3.09E7

267.6

36731

9.49E6

272.2

37298

1.27E8

269.2

158

38985

1.17E7

257.5

CLS Ground State

Population in

CLS Excited State

(cm−1)

158

158

158

158

2356

2356

2356

12805

12805

State

0.467

0.467

0.465

0.465

0.275

0.275

0.275

0.182

0.182

(cm−1)

35520

34886

35520

34886

36731

37298

33351

38985

44638

A (s−1)

2.52E7

1.09E7

2.52E7

1.09E7

1.2E8

2.05E8

1.2E8

4.73E7

3.51E7

A.12 Molybdenum

The promising optical pumping schemes of Mo+ are shown in Table A.19. There are no

low-energy states, and all of the ions can be found in the ground state at room temperature.

Typical CLS measurements of Mo+ requires atomic charge exchange reactions, where only

4% of the atomic population is in the state being probed. Mo is a very promising case for the

optical pumping system, as the optical pumping simulation predicts electronic populations

of up to 70% for CLS measurements.

Table A.19: Optical Pumping Schemes of Mo+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

0

0

0

49481

2.57E7

202.1

48860

1.84E7

204.6

49608

1.57E7

201.5

A (s−1)

2.27E8

5.63E7

8.76E7

1.6E8

1.28E8

8.39E7

CLS Ground State

Population in

CLS Excited State

State

0.702

0.702

0.544

0.544

0.34

0.34

(cm−1)

48959

47999

47232

47999

46148

46614

(cm−1)

13460

13960

12900

12900

12034

12034

161

A.13 Technetium

Technetium was not considered for this study due to a dearth of knowledge about its elec-

tronic structure (See Section 3.4.2).

A.14 Ruthenium

The promising optical pumping schemes of Ru+ are shown in Table A.20. All of the electronic

population is found in the ground state of Ru+. Typical CLS measurements of Ru+ require

atomic charge exchange reactions, and only 7% of the atoms will populate the state being

probed. There is only one optical pumping step that is available for use at BECOLA, exciting

ions in the ground state to an excited state located at 50845 cm-1 and requiring 196 nm light.

This is the edge of the wavelength range of the BECOLA pulsed laser system. The excited

state strongly decays to a metastable state at 10150 cm-1, from which CLS measurements

can be performed.

Table A.20: Optical Pumping Schemes of Ru+

Pumping Ground

Pumping Excited

State (cm−1)

State (cm−1)

A (s−1) λ (nm)

0

50845

1.72E8

196.6

CLS Ground State

Population in

CLS Excited State

(cm−1)

10150

10150

19378

19378

State

0.746

0.746

0.153

0.153

(cm−1)

46711

46420

55694

50845

A (s−1)

9.27E8

1.67E8

7.87E8

4.21E8

A.15 Rhodium, Palladium, Silver, and Cadmium

There are no suitable optical schemes for the cases of Rh (160 nm), Pd (137 nm), Ag (120

nm) , and Cd (226 nm). Rh, Pd, and Ag require vacuum UV laser light, down to 120 nm

162

in the case of Ag, and will not be possible in the foreseeable future at BECOLA. Cd does

feature a transition at 226 nm, exciting the 4d105s2S1/2 to 4d105p2P o

1/2 transition, but it

is almost completely cyclical and cannot be used for optical pumping. The next available

transition requires 105 nm and will not be possible.

163

APPENDIX B

Logic of Shutter Control System

A physical shutter was used to block and unblock the pulsed laser light during the commis-

sioning experiments. The shutter allowed measurements of the fluorescence signal with and

without the pulsed laser light in quick succession and mitigated any systematic changes in

the ion source intensity or laser power. The shutter was controlled by the BECOLA DAQ

system. A schematic diagram of the electronics modules can be seen in Figure B.1. A trigger

signal from the pulsed laser is used to start the DAQ system, and the physical shutter is

opened and closed in one second intervals to control for systematic fluctuations. Approx-

imately 350 µs after the trigger signal arrives from the pulsed laser, the trapped ions are

released from the trap and a signal is sent to a level adapter to convert the TTL signal to

NIM. The NIM signal is sent to a fan-in/fan-out(FIFO) electronic module. From the FIFO

module, a signal is sent to a Gate Generator module to delay and broaden the signal. From

the gate generator, the signal is split and sent to two separate scale down modules(SDM). In

one module, after 10 ion bunch releases, a signal is sent to one input of a coincidence module

and to the second SDM. In a second SDM, the signal from the first SDM is scaled down to

one pulse per two pulses, and two signals are created and used. One signal from the second

SDM is sent to the same coincidence unit as the first SDM. When the signals from the SDM

coincide, a signal is sent to a gate generator with start/stop operation. The gate generator

provides a 5 V signal that opens the shutter. The second output from the second SDM is

sent to the gate generator with start/stop operation, and is used to stop the 5 V signal that

164

Figure B.1: Schematic diagram of the connectivity of the electronic modules used to control
the physical shutter for measurement of the optical pumping effect.

it generates. The result is that for every 10 ion bunches, a signal is generated to either open

or close the shutter.

165

REFERENCES

166

REFERENCES

[1] K. Minamisono, P.F. Mantica, A. Klose, S. Vinnikova, A. Schneider, B. Johnson,
and B.R. Barquest. Commissioning of the collinear laser spectroscopy system in the
BECOLA facility at NSCL. Nucl. Instrum. Methods Phys. Res. A, 709:85–94, 2013.

[2] D.M. Rossi, K. Minamisono, B.R. Barquest, G. Bollen, K. Cooper, M. Davis, K. Ham-
merton, M. Hughes, P.F. Mantica, D.J. Morrissey, R. RIngle, J.A. Rodriguez, C.A.
Ryder, S. Schwarz, R. Strum, C. Sumithrarachi, D. Tarazona, and S. Zhao. A field
programmable gate array-based time-resolved scaler for collinear laser spectroscopy
with bunched radioactive potassium beams. Rev. Sci. Instrum., 85:093503, 2014.

[3] B. Cheal and K.T. Flanagan. Progress in laser spectroscopy at radioactive ion beam

facilities. J. Phys. G Nucl. Part. Phys., 37:113101, 2010.

[4] P. Campbell, I.D. Moore, and M.R. Pearson. Laser spectroscopy for nuclear structure

physics. Prog. Part. Nucl. Phys., 86:127–180, 2016.

[5] E.W. Otten. Nuclear radii and moments of unstable isotopes. Springer US, Boston,

MA, 1989.

[6] H.-J¨urgen Kluge and W. N¨ortersh¨auser. Lasers for nuclear physics. Spectrochim. Acta

A, 58:1031–1045, 2003.

[7] K. Blaum, J. Dilling, and W. N¨ortersh¨auser. Precision atomic physics techniques for

nuclear physics with radioactive beams. Phys. Scr., T152:014017, 2013.

[8] I. Angeli and K.P. Marinova. Table of experimental nuclear ground state charge radii:

An update. At. Data Nucl. Data Tables, 99:69–95, 2013.

[9] M. Avgoulea, Yu. P. Gangrsky, K.P. Marinova, S.G. Zemlyanoi, S. Fritzsche, D. Iablon-
skyi, C. Barbieri, E.C. Simpson, P.D. Stevenson, J. Billowes, P. Campbell, B. Cheal,
B. Tordoff, M.L. Bissell, D.H. Forest, M.D. Gardner, G. Tungate, J. Huikari, A. Niem-
inen, H. Penttil¨a, and J. ¨Ayst¨o. Nuclear charge radii and electromagnetic moments of
radioactive scandium isotopes and isomers. J. Phys. G Nucl. Part. Phys., 38:025104,
2011.

167

[10] Yu. P. Gangrsky, K.P. Marinova, S.G. Zemlyanoi, I.D. Moore, J. Billowes, P. Camp-
bell, K.T. Flanagan, D.H. Forest, J.A.R. Griffith, J. Huikari, R. Moore, A. Nieminen,
H. Thayer, G. Tungate, and J. ¨Ayst¨o. Nuclear charge radii of neutron deficient titanium
isotopes 44Ti and 45Ti. J. Phys. G Nucl. Part. Phys., 30:1089–1098, 2004.

[11] H. Heylen, C. Babcock, R. Beerwerth, J. Billowes, M.L. Bissell, K. Blaum, J. Bonnard,
P. Campbell, B. Cheal, T. Day Goodacre, D. Fedorov, S. Fritzsche, R.F. Garcia Ruiz,
W. Geithner, Ch. Geppert, W. Gins, L. K. Grob, M. Kowalska, K. Kreim, S.M. Lenzi,
I.D. Moore, B. Maass, S. Malbrunot-Ettenauer, B. Marsh, R. Neugart, G. Neyens,
W. N¨ortersh¨auer, T. Otsuka, J. Papuga, R. Rossel, S. Rothe, R. S´anchez, Y. Tsunoda,
C. Wraith, L. Xie, X.F. Yang, and D.T. Yordanov. Changes in nuclear structure along
the Mn isotopic chain studied via charge radii. Phys. Rev. C, 94:054321, 2016.

[12] K. Minamisono, D.M. Rossi, R. Beerwerth, S. Fritzsche, D. Garand, A. Klose, Y. Liu,
B. Maaß, P.F. Mantica, A.J. Miller, P. M¨uller, W. Nazarewicz, W. N¨ortersh¨auser,
E. Olsen, M.R. Pearson, P.-G. Reinhard, E.E. Saperstein, C. Sumithrarachchi, and
S.V. Tolokonnikov. Charge radii of neutron deficient 52,53Fe produced by projectile
fragmentation. Phys. Rev. Lett., 117:252501, 2016.

[13] S. Kaufmann, J. Simonis, S. Bacca, J. Billowes, M.L. Bissell, K. Blaum, B. Cheal,
R.F. Garcia Ruiz, W. Gins, C. Gorges, G. Hagen, H. Heylen, A. Kanellakopou-
los, S. Malbrunot-Ettenauer, M. Miorelli, R. Neugart, G. Neyens, W. N¨ortersh¨auer,
R. S´anchez, S. Sailer, A. Schwenk, T. Ratajczyk, L.V. Rodr´ıguez, L. Wehner,
C. Wraith, L. Xie, Z.Y. Xu, X.F. Yang, and D.T. Yordanov. Charge radius of the short-
lived 68Ni and correlation with the dipole polarizability. Phys. Rev. Lett., 124:132502,
2020.

[14] M.L. Bissell, T. Carette, K.T. Flanagan, P. Vingerhoets, J. Billowes, K. Blaum,
B. Cheal, S. Fritzsche, M. Godefroid, M. Kowalska, J. Kr¨amer, R.Neugart, G. Neyens,
W. N¨ortersh¨auser, and D.T. Yordanov. Cu charge radii reveal a weak sub-shell effect
at N = 40. Phys. Rev. C, 93:064318, 2016.

[15] R.P. de Groote, J. Billowes, C.L. Binnersley, M.L. Bissell, T.E. Cocolios, T. Day
Goodacre, G.J. Farooq-Smith, D.V. Fedorov, K.T. Flanagan, S. Franchoo, R.F. Gar-
cia Ruiz, W. Gins, J.D. Holt, A. Koszor´us, K.M. Lynch, T. Miyagi, W. Nazarewicz,
G. Neyens, P.-G. Reinhard, S. Rothe, H.H. Stroke, A.R. Vernon, K.D.A. Wendt, S.G.
Wilkins, Z.Y. Xu, and X.F. Yang. Measurement and microscopic description of odd-
even staggering of charge radii of exotic copper isotopes. Nat. Phys., 16:620–624,
2020.

[16] L. Xie, X.F. Yang, C. Wraith, C. Babcock, J. Biero´n, J. Billowes, M.L. Bissell,
K. Blaum, B. Cheal, L. Filippin, K.T. Flanagan, R.F. Garcia Ruiz, W. Gins,

168

G. Gaigalas, M. Godefroid, C. Gorges, L.K. Grob, H. Heylen, P. J¨onsson, S. Kauf-
mann, M. Kowalska, J. Kr¨amer, S. Malbrunot-Ettenauer, R. Neugart, G. Neyens,
W. N¨ortersh¨auser, T. Otsuka, J. Papuga, R. S´anchez, Y. Tsunoda, and D.T. Yor-
danov. Nuclear charge radii of 62-80Zn and their dependence on cross-shell proton
excitations. Phys. Lett. B, 797:134805, 2019.

[17] A.J. Miller, K. Minamisono, A.Klose, D.Garand, C.Kujawa, J.D. Lantis, Y. Liu,
B. Maaß, P.F. Mantica, W. Nazarewicz, W.N¨ortersh¨auser, S.V. Pineda, P.-G. Rein-
hard, D.M. Rossi, F. Sommer, C. Sumithrarachchi, A. Teigelh¨ofer, and J. Watkins.
Proton superfluidity and charge radii in proton-rich calcium isotopes. Nat. Phys.,
15:432–436, 2019.

[18] D.M. Rossi, K. Minamisono, H.B. Asberry, G. Bollen, B.A. Brown, K. Cooper, B. Isher-
wood, P.F. Mantica, A. Miller, D.J. Morrissey, R. Ringle, J.A. Rodriguez, C. A. Ryder,
A. Smith, R. Strum, and C. Sumithrarachchi. Charge radii of neutron-deficient36K and
37K. Phys. Rev. C, 92:014305, 2015.

[19] F. Touchard, P. Guimbal, S. Buttgenbach, R. Klapisch, M. De Saint Simon, J.M. Serre,
and C. Thibault. Isotope shifts and hyperfine structure of 38-47K by laser spectroscopy.
Phys. Lett., 108B:169–171, 1982.

[20] A.Klein, B.A. Brown, U. Georg, M. Keim, P. Lievens, R. Neugart, M. Neuroth, R.E.
Silverans, L. Vermeeren, and ISOLDE Collaboration. Moments and mean square
charge radii of short-lived argon isotopes. Nuc. Phys. A, 607:1–22, 1996.

[21] R. Broda, B. Fornal, W. Kr´olas, T. Pawlat, D. Bazzacco, S. Lunardi, C. Rossi-Alvarez,
R. Menegazzo, G. de Angelis, P. Bednarczyk, J. Rico, D. De Acu˜na, P.J. Daly, R.H.
Mayer, M. Sferrazza, H. Grawe, K.H. Maier, and R. Schubart. N = 40 neutron subshell
closure in the 68Ni nucleus. Phys. Rev. Lett., 74:868–871, 1995.

[22] W. Rother, A. Dewald, H. Iwasaki, S.M. Lenzi, K. Starosta, D. Bazin, T. Baugher,
B.A. Brown, H.L. Crawford, C. Fransen, A. Gade, T.N. Ginter, T. Glasmacher, G.F.
Grinyer, M. Hackstein, G. Illie, J. Jolie, S. McDaniel, D. Miller, P. Petkov, Th. Pissulla,
A. Ratkiewicz, C.A. Ur, P. Voss, K.A. Walsh, D. Weisshaar, and K.-O Zell. Enhanced
quadrupole collectivity at N = 40: The case of neutron-rich Fe isotopes. Phys. Rev.
Lett., 106:022502, 2011.

[23] A. Gade, R.V.F. Janssens, T. Baugher, D. Bazin, B.A. Brown, M.P. Carpenter, C.J.
Chiara, A.N. Deacon, S.J. Freeman, G.F. Grinyer, C.R. Hoffman, B.P. Kay, F.G.
Kondev, T. lauritsen, S. Mcdaniel, K. Meierbachtol, A. Katkiewicz, S.R. Stroberg,
K.A. Walsh, D. Weisshaar, R. Winkler, and S. Zhu. Collectivity at N = 40 in neutron-
rich 64Cr. Phys. Rev. C, 81:051304, 2010.

169

[24] B. Pritychenko, M. Birch, B. Singh, and M. Horoi. Tables of the e2 transition proba-
bilitys from the first 2+ states in even-even nuclei. At. Data Nucl. Data Tables, 107:1,
2016.

[25] B. Cheal, M.D. Gardner, M. Avgoulea, J. Billowes, M.L. Bissell, P. Campbell, T. Ero-
nen, K.T. Flanagan, D.H. Forest, J. Huikari, A. Jokinen, B.A. Marsh, I.D. Moore,
A. Nieminen, H. Penttil a, S. Rinta-Antila, B. Tordoff, G. Tungate, and J. ¨Ayst¨o.
The shape transition in the neutron-rich yttrium isotopes and isomers. Phys. Lett. B,
645:133–137, 2007.

[26] D.H. Forest, J. Billowes, P. Campbell, P. Dendooven, K.T. Flanagan, J.A.R. Griffith,
J. Huikari, A. Jokinen, R. Moore, A. Nieminen, H.L. Thayer, G. Tungate, S. Zemlyanoi,
and J. ¨Ayst¨o. Laser spectroscopy of neutron deficient zirconium isotopes. J. Phys. G
Nucl. Part. Phys., 28:L63–L68, 2002.

[27] P. Campbell, H.L. Thayer, J. Billowes, P. Dendooven, K.T. Flanagan, D.H. For-
est, J.A.R. Griffith, J. Huikari, A. Jokinen, R. Moore, A. Nieminen, G. Tungate,
S. Zemlyanoi, and J. ¨Ayst¨o. Laser spectroscopy of cooled zirconium fission fragments.
Phys. Rev. Lett., 89:082501, 2002.

[28] B. Cheal, K. Baczynska, J. Billowes, P. Campbell, F.C. Charlwood, T. Eronen, D.H.
Forest, A. Jokinen, T. Kessler, I.D. Moore, M. Reponen, S. Rothe, M. R¨uffer, A. Saas-
tamoinen, G. Tungate, and J. ¨Ayst¨o. Laser spectroscopy of niobium fission frag-
ments: First use of optical pumping in an ion beam cooler buncher. Phys. Rev. Lett.,
102:222501, 2009.

[29] F.C. Charlwood, K. Baczynska, J. Billowes, P. Campbell, B. Cheal, T. Eronen, D.H.
Forest, A. Jokinen, T. Kessler, I.D. Moore, H. Penttil¨a, R. Powis, M. R¨uffer, A. Saas-
tamoinen, G. Tungate, and J. ¨Ayst¨o. Nuclear charge radii of molybdenum fission
fragments. Phys. Lett. B, 674:23–27, 2009.

[30] R. Ferrer, N. Bree, T.E. Cocolios, I.G. Darby, H. De Witte, W. Dexters, J. Diriken,
J. Elseviers, S. Franchoo, M. Huyse, N. Kesteloot, Yu. Kudryavtsev, D. Pauwels,
D. Radulov, T. Roger, H. Savajols, P. Van Duppen, and M. Venahrt. In-gas-cell laser
ionization spectroscopy in the vicinity of 100Sn: Magnetic moments and mean-square
charge radii of N = 50-54 Ag. Phys. Lett. B, 728:191–197, 2014.

[31] U. Dinger, J. Eberz, G. Huber, and R. Menges. Nuclear moments and change in the

charge radii of neutron-deficient silver isotopes. Nucl. Phys., A503:331–348, 1989.

[32] M. Hammen, W. N¨ortersh¨auser, D.L. Balabanski, M.L. Bissell, K. Blaum,
I. Budinˇcevi´c, B. Cheal, K.T. Flanagan, N. Fr¨ommgen, G. Georgiev, Ch. Geppert,

170

M. Kowalska, K. Kreim, A. Krieger, W.Nazarewicz, R. Neugart, G. Neyens, J. Papuga,
P.-G Reinhard, M.M. Rajabali, S. Schmidt, and D.T. Yordanov. From calcium to cad-
mium:testing the pairing functional through charge radii measurements of 100-130Cd.
Phys. Rev. Lett., 121:102501, 2018.

[33] H. Schatz, A. Aprahamian, J. G orres, M. Wiescher, T. Rauscher, J.F. Rembges, F.-K.
Thielemann, B. Pfeiffer, P. M oller, K.L. Kratz, H. Herndl, B.A. Brown, and H. Rebel.
rp-process nucleosynthesis at extreme temperature and density conditions. Phys. Rep.,
294:167–263, 1998.

[34] S. J. Zheng, F. R. Xu, S.F. Shen, H.L. Liu, R. Wyss, and Y.P Yan. Shape coexistence

and triaxiality in nuclei near 80Zr. Phys. Rev. C, 90:064309, 2014.

[35] M. Akbari and A. Kardan. Shape evolution and coexistence in neutron-deficient Kr,

Rb, Sr, and Zr nuclei. Nucl. Phys. A, 990:109–117, 2019.

[36] F.N. Mortensen, J.M. Scott, and Stirling A. Colgate. How archival test data contribute

to certification. Los Alamos Science, 28:38–46, 2003.

[37] L. Ahle, M. Hausmann, R. Relfarth, K. Roberts, M. Roeben, B. Rusnak, and D. Vieira.
Realizing the opportunies of neutron cross-section measurements at RIA. AIP Conf.
Proc., 769:796–799, 2005.

[38] M.N. Kreisler. National nuclear security and other applications of rare isotopes. In
Proceedings of Particle Accelerator Conference(PAC07), pages 124–126. Institute of
Electrical and Electronics Engineers, 2007.

[39] E. Hartouni. Role of radioactive beams in stockpile stewardship science, 2010. presen-

tation.

[40] R. Jeanloz. Science-based stockpile stewardship. Physics Today, 53:44–50, 2000.

[41] J.E. Escher, J.T. Burke, F.S. Dietrich, N.D. Scielzo, I.J. Thompson, and W. Younes.
Compound-nuclear reaction cross sections from surrogate measurements. Rev. Mod.
Phys., 84:353–397, 2012.

[42] A. Spyrou, S.N. Liddick, A.C. Larsen, M. Guttormsen, K. Cooper, A.C. Dombos,
D.J. Morrissey, F. Naqvi, G. Perdikakis, S.J. Quinn, T. Renstrøm, J.A. Rodriguez,
A. Simon, C.S. Sumithrarachchi, and R.G.T. Zegers. Novel technique for constraining
r -process (n,γ) reaction rates. Phys. Rev. Lett., 113:232502, 2014.

171

[43] J.A. Paisner. Atomic vapor laser isotope separation. Appl. Phys. B, 46:253–260, 1988.

[44] C.R. Locke, T. Kobayashi, T. Fujiwara, and K. Midorikawa. Selective photoionization
of palladium isotopes using a two-step excitation scheme. Appl. Phys. B, 123:240,
2017.

[45] T. Kobayashi T. Fujiwara and K. Midorikawa. Selective resonance photoionization of
odd mass zirconium isotopes towards efficient separation of radioactive waste”. Sci.
Rep., 9:1754, 2019.

[46] W.S. Yang, Y. Kim, R.N. Hill, T.A. Taiwo, and H.S. Khalil. Long-lived fission product

transmutation studies. Nucl. Sci. Eng., 146:291–318, 2004.

[47] S. Chiba, T. Wakabayashi, Y. Tachi, N. Takaki, A. Terashima, S. Okumura, and
T. Yoshida. Method to reduce long-lived fission products by nuclear transmutations
with fast spectrum reactors. Sci. Rep., 7:13961, 2017.

[48] T.H. Maiman. Stimulated optical radiation in ruby. Nature, 187:493–494, 1960.

[49] K.-R. Anton, S.L. Kaufman, W. Klempt, G. Moruzzi, R. Neugart, E.-W Otten, and
B. Schinzler. Collinear laser spectroscopy on fast atomic beams. Phys. Rev. Lett.,
40:642–645, 1978.

[50] O.B. Tarasov and D. Bazin. Lise++: Radioactive beam production with in-flight

separators. Nucl. Instrum. Methods Phys. Res. B, 266:4657–4664, 2008.

[51] M.J.G. Borge and Bj¨orn Jonson. ISOLDE past, present and future. J. Phys. G Nucl.

Part. Phys., 44:044011, 2017.

[52] A. Voss, T.J. Procter, O. Shelbaya, P. Amaudruz, F. Buchinger, J.E. Crawford,
S. Daviel, E. Man´e, M.R. Pearson, and W. Al Tamimi. The collinear fast beam
laser spectroscopy(CF BS) experiment at TRIUMF. Nucl. Instrum. Meth. Phys. Res.
A, 811:57–69, 2016.

[53] U. K¨oster, P. Carbonez, A. Dorsival, J. Dvorak, R. Eichler, S. Fernandes, H. Fr˚anberg,
(im-)possible ISOL

J. Neuhausen, Z. Novackova, R. Wilfinger, and A. Yakushev.
beams. Eur. Phys. J Spec. Top., 150:285–291, 2007.

[54] D.J. Morrissey, B.M. Sherrill, M. Steiner, A. Stolz, and I. Wiedenhoever. Commis-
sioning the A1900 projectile fragment separator. Nucl. Instrum. Methods Phys. Res.
B, 204:90–96, 2003.

172

[55] R.W. Boyd. Nonlinear Optics. Academic Press, Orlando, Florida, 2008.

[56] N. Bendali, H.T. Duong, P. Juncar, J.M. Saint Jalm, and J.L. Vialle. Na+-Na charge
exchange processes studied by collinear laser spectroscopy. J. Phys. B At. Mol. Phys.,
19:233–238, 1986.

[57] A. Klose, K. Minamisono, Ch. Geppert, N. Fr¨ommgen, M. Hammen, J. Kr¨amer,
A. Krieger, C.D.P. Levy, P.F. Mantica, W. N¨ortersh¨auser, and S. Vinnikova. Tests
of atomic charge-exchange cells for collinear laser spectroscopy. Nucl. Instrum. Meth-
ods Phys. Res. A, 678:114–121, 2012.

[58] C.A. Ryder, K. Minamisono, H.B. Asberry, B. Isherwood, P.F. Mantica, A. Miller,
D.M. Rossi, and R. Strum. Population distribution subsequent to charge exchange of
29.85 kev Ni+ on sodium vapor. Spectrochim. Acta. B, 113:16–21, 2015.

[59] M. Keim, U. Georg, A. Klein, R. Neugart, M. Neuroth, S. Wilbert, P. Lievens, L. Ver-
meeren, B.A. Brown, and the ISOLDE Collaboration. Measurement of the electric
quadrupole moments of 26-29Na. Eur. Phys. J. A, 8:31–40, 2000.

[60] B. Fenker, J.A. Behr, D. Melconian, R.M.A. Anderson, M. Anholm, D. Ashery, R.S.
Behling, I. Cohen, I. Craiciu, J.M. Donohue, C. Farfan, D. Friesen, A. Gorelov, J. Mc-
Neil, M. Mehlman, H. Norton, K. Olchanski, S. Smale, O. Theriault, A.N. Vantyghem,
and C.L. Warner. Precision measurement of the nuclear polarization in laser-cooled
optically pumped 37K. New J. Phys., 18:073028, 2016.

[61] K. Baczynska, J. Billowes, P. Campbell, F.C. Charlwood, B. Cheal, T. Eronen, D.H.
Forest, A. Jokinen, T. Kessler, I.D. Moore, M. R¨uffer, G. Tungate, and J. ¨Ayst¨o. Nu-
clear spin determination of 100m Y by collinear laser spectroscopy of optically pumped
ions. J. Phys. G: Nucl. Part. Phys., 37:105103, 2010.

[62] F.C. Charlwood, J. Billowes, P. Campbell, B. Cheal, T. Eronen, D.H. Forest,
S. Fritzsche, M. Honma, A. Jokinen, I.D. Moore, H. Penttil¨a, R. Powis, A. Saasta-
moinen, G. Tungate, and J. ¨Ayst¨o. Ground state properties of manganese isotopes
across the N =28 shell closure. Phys. Lett. B, 690:346–351, 2010.

[63] H. de Vries, C.W. de Jager, and C. de Vries. Nuclear charge-density-distribution
parameters from elastic electron scattering. At. Data Nucl. Data Tables, 36:495–536,
1987.

[64] A. Estrad´e, R. Kanungo, W. Horiuchi, F. Ameil, J. Atkinson, Y. Ayyad, D. Cortina-
Gil, I. Dillmann, A. Evdokimov, F. Farinon, H. Geissel, G. Guastalla, R. Janik,
M. Kimura, R. Kn¨obel, J. Kurcewicz, Yu. A. Litvinov, M. Marta, M. Mostazo,

173

I. Mukha, C. Nociforo, H.J. Ong, S. Pietri, A. Prochazka, C. Scheidenberger, B. Sitar,
P. Strmen, Y. Suzuki, M. Takechi, J. Tanaka, I. Tanihata, S. Terashima, J. Vargas,
H. Weick, and J.S. Winfield. Proton radii of 12-17B define a thick neutron surface in
17B. Phys. Rev. Lett., 113:132501, 2014.

[65] F. Boehm and P.L. Lee. Changes of mean-square nuclear charge radii from isotope

shifts of electronic K α x-rays. At. Data Nucl. Data Tables, 14:605–611, 1974.

[66] R. Engfer, H. Schneuwly, J.L. Vuilleumier, H.K. Walter, and A. Zehnder. Charge-
distribution parameters, isotope shifts, isomer shifts, and magnetic hyperfine constants
from muonic atoms. At. Data Nucl. Data Tables, 14:509–597, 1974.

[67] H. Sch uler and J.E. Keyston. Uber einen isotopenverschiebungseffekt der hyperfein-

strukturerme von thallium. Z. Phys., 70:1–9, 1931.

[68] H. Sch uler and Th. Schmidt. Uber abweichungen des atomkerns von der kugelsym-

metrie. Z. Phys., 94:457–468, 1935.

[69] B. Maaß, T. H uther, K. K onig, J org Kr amer, J. Krause, A. Lovato, P. M uller,
Krzysztof Pachucki, M. Puchalski, R. Roth, R. S´anchez, F. Sommer, R.B. Wiringa,
and W. N¨ortersh¨auser. Nuclear charge radii of 10,11B. Phys. Rev. Lett., 122:182501,
2019.

[70] W.H. King. Isotope Shifts in Atomic Spectra. Springer Science and Business Media,

New York, 1984.

[71] G. Fricke and K. Heilig. Nuclear Charge Radii. Springer-Verlag Berlin, Heidelberg,

2004.

[72] D. York, N.M. Evensen, M. L´opez Mart´ınez, and J. De Basabe Delgado. Unifed
equations for the slope,intercept, and standard errors of the best straight line. Am. J.
Phys., 72:367–375, 2004.

[73] S. Kaufmann. Laser spectroscopy of nickel isotopes with a new data acquisition system

at ISOLDE. PhD thesis, Technische Universit¨at Darmstadt, 2019.

[74] G. Zizak, J.D. Bradshaw, and J.D. Winefordner. Rate equation solution for the tem-

poral behavior of a three-level system. Appl. Opt., 19:3631–3639, 1980.

[75] R.L. Kurucz. The Kurucz Smithsonian Atomic and Molecular Database. In Astro-
physical Applicaations of Powerful New Databases, pages 205–210. International As-
tronomical Union, 1995.

174

[76] A. Kramida, Yu. Ralchenko, J. Reader, and NIST ASD Team. NIST Atomic Spectra
Database (Version 5.7.1. [Online]https://www.nist.gov/pml/atomic-spectra-database,
2019.

[77] B.R. Barquest, G. Bollen, P.F. Mantica, K. Minamisono, R. Ringle, S. Schwarz, and
C.S. Sumithrarachchi. RFQ beam cooler and buncher for collinear laser spectroscopy
of rare isotopes. Nucl. Instrum. Methods Phys. Res. A, 866:18–28, 2017.

[78] B. Barquest. An advanced ion guide for beam cooling and bunching for collinear laser

spectroscopy of rare isotopes. PhD thesis, Michigan State University, 2014.

[79] A.V. Tolmachev, A.N. Vilkov, B. Bogdanov, L. Pasa-Tolic, C.D. Masselon, and R.D.
Smith. Collisional activation of ions in RF ion traps and ion guides: The effective ion
temperature treatment. J. Am. Soc. Mass Spectrom., 15:1616–1628, 2004.

[80] R. Ringle. 3dCylPIC-a 3D particle-in-cell code in cylindrical coordinates for space
Int. J. Mass Spectrom.,

charge simulations of ion trap and ion transport devices.
303:42–50, 2011.

[81] P.W. Milonni and J.H. Eberly. Laser Physics. John Wiley and Sons, 2010.

[82] T. Kron, R. Beerwerth, S. Raeder, S. Fritzsche, R. Heinke, P. Sch¨onberg, M. Tr¨umper,
and K. Wendt. Hyperfine structure study of 97,98,99Tc in a new laser ion source for
high-resolution laser spectroscopy. Phys. Rev. C, 102:034307, 2020.

[83] S. Raeder, T. Kron, R. Heinke, J.L. henares, N. Lecesne, P. Sch¨onberg, M. Tr¨umper,
and K. Wendt. High resolution spectroscopy of the hyperfine structure splitting in
97,99Tc. Hyperfine Interact, 238:15, 2017.

[84] S.L. Kaufman. High-resolution laser spectroscopy in fast beams. Opt. Commun.,

17:309–312, 1976.

[85] A. Nieminen, P. Campbell, J. Billowes, D.H. Forest, J.A.R. Griffith, J. Huikari, A. Joki-
nen, I.D. Moore, R. Moore, G. Tungate, and J. ¨Ayst¨o. Cooling and bunching of ion
beams for collinear laser spectroscopy. Nucl. Instrum. Meth. Phys. Res. B, 204:563–
569, 2003.

[86] C.S. Sumithrarachchiand D.J. Morrissey, S. Schwarz, K. Lund, G. Bollen, R. Ringle,
G. Savard, and A.C.C. Villari. Beam thermalization in a large gas catcher. Nucl.
Instrum. Methods Phys. Res., 463:305–309, 2020.

175

[87] Z. Nouri, R. Li, R.A. Holt, and S.D. Rosner. A penning sputter ion source with very

low energy spread. Nucl. Instrum. Methods Phys. Res. A, 614:174–178, 2010.

[88] W. Paul. Electromagnetic traps for charged and neutral particles (Nobel Lecture).

Angew. Chem. Int. Ed. Engl., 29:739–748, 1990.

[89] A.J. Miller. Measurements of Charge Radii of Neutron-Deficient Calcium Using
Collinear Laser Spectroscopy at BECOLA. PhD thesis, Michigan State University,
2019.

[90] Matisse 2 TS. http://www.sirah.com/laser/cw-ring-lasers/matisse-ts.

[91] Spectra-Physics Millenia

eV.

https://www.spectra-physics.com/products/cw-

lasers/millennia-ev.

[92] Spectra-Physics WaveTrain.

https://www.spectra-physics.com/products/tunable-

lasers/wavetrain.

[93] Highfinesse WSU-30 Wavemeter. https://www.highfinesse.com/en/wavelengthmeter.

[94] R-32734 stabilized HeNe laser. http://www.newport.com/p/R-32734.

[95] Cobra-Stretch Pulsed Laser. http://www.sirah.com/laser/pulsed-lasers/cobra-stretch.

[96] Quanta-Ray INDI Pulsed Nd:YAG Laser User’s Manual.

http://www.Spectra-

Physics.com.

[97] Pulsed Dye Laser’s Manual version 2.0.4. http://www.Spectra-Physics.com.

[98] Highfinesse

Angstrom

WS6-600

wavelength

meter.

https://www.highfinesse.com/en/wavelengthmeter.

[99] Avantes Inc. http://www.avantes.com.

[100] Fred version 19.40.2. http://www.photonengr.com.

[101] Acl Alloys Inc. http://www.aclalloys.com.

[102] C.R. Cowley and C.H. Corliss. Moderately accurate oscillator strengths from nbs

intensities. Mon. Not. R. astr. Soc., 203:651–659, 1983.

176

[103] G. Ljung, H. Nilsson, M. Asplund, and S. Johansson. New and improved experimental
oscillator strengths in zr ii and the solar abundance of zirconium. Astron. Astrophys.,
456:1181–1185, 2006.

[104] Picomotor piezo linear actuators.

https://www.newport.com/f/picomotor-piezo-

linear-actuators.

[105] T. Kessler, K. Bru¨uck, C. Baktash, J.R. Beene, Ch. Geppert, C.C. Havener, H.F.
Krause, Y. Liu, D.R. Schultz, D.W. Stracener, C.R. Vane, and K. Wendt. Three-
step resonant photoionization spectroscopy of Ni and Ge:
ionization potential and
odd-parity Rydberg levels. J. Phys. B Mol. Opt. Phys., 40:4413–4432, 2007.

[106] S. Raeder, V. Sonnenschein, T. Gottwald, I.D. Moore, M. Reponen, S. Rothe, N. Traut-
mann, and K. Wendt. Resonance ionization spectroscopy of thorium isotopes-towards
a laser spectroscopic identification of the low-lying 7.6 ev isomer of 229Th. J. Phys. B
At. Mol. Opt. Phys., 44:165005, 2011.

[107] Y. Liu, T. Gottwald, C. Mattolat, and K. Wendt. Three-photon resonance ionization
of atomic Mn in a hot-cavity laser ion source using Ti:Sapphire lasers. J. Phys. B Mol.
Opt. Phys., 48:115006, 2015.

[108] S. Schmidt, Ch. Geppert, and Z. Andelkovic. Laser spectroscopy methods for probing

highly charged ions at GSI. Hyperfine Interact., 227:29–43, 2014.

[109] J. Beene, D.W¿ Bardayan, A. Galindo Uribarri, C.J. Gross, K.L. Jones, J.F. Liang,
W. Nazarewicz, D.W. Stracener, B.A. Tatum, and R.L. Varner. ISOL science at the
Holifield Radioactive In Beam facility. J. Phys. G Nucl. Part. Phys., 38:024002, 2011.

[110] Y. Liu, C.U. Jost, A.J. Mendez II, D.W. Stracener, C.L. Williams, C.J. Gross, R.K.
Grzywacz, M. Madurga, K. Miernik, D. Miller, S. Padgett, S.V. Paulauskas, K.P.
Rykaczewski, and M. Wolinska-Cichocka. On-line commissioning of the HRIBF reso-
nant ionization laser ion source. Nucl. Instrum. Meth. Phys. Res. B, 298:5–12, 2013.

177