.
‘3?
z: 1

5.51:!
(.‘ilnv :
«VI .. Fl

 

(v.41 . .

Zn.

 

T’TTf‘"' ‘N
Emil-Orb

/‘)

, or
0003

5757’] ? 3 72/

 

 

 

LIBRARY
M'Chlgan State This is to certify that the
UHIVGI’SIty dissertation entitled

 

DEVELOPMENT OF A HIGH-PRESSURE, GAS-FILLED ION
SOURCE FOR A RADIOACTIVE BEAM FACILITY

presented by

PATRICK ANDREW LOFY

has been accepted towards fulﬁllment

 

 

 

 

of the requirements for the
Ph.D. degree in Chemical Physics
ad r P/ofessor’s Sigrgd’re
// 574% ”1/ 0c;
Date

MSU is an Aﬁirmative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DatoDquGS-sz

 

DEVELOPMENT OF A HIGH-PRESSURE, GAS-FILLED ION
SOURCE FOR A RADIOACTIVE BEAM FACILITY

By

Patrick Andrew Lofy

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Chemistry

2003

ABSTRACT

DEVELOPMENT OF A HIGH-PRESSURE, GAS-FILLED ION SOURCE FOR A
RADIOACTIVE BEAM FACILITY

By

Patrick Andrew Lofy

The National Superconducting Cyclotron Laboratory (NSCL) produces radioac-
tive nuclear beams (RN Bs) by in-ﬂight fragmentation of fast ion beams. This method
produces nuclei more exotic (having a more extreme neutron-proton ratio) than meth-
ods at other RNB facilities. However, the resulting beams, by comparison, have other
characteristics, such as very high velocities, which can be less desirable for certain
nuclear science experiments. It was the goal of the research in this work to construct
an ion source which would capture the fast exotic beams produced at the NSCL, and
re-emit them near thermal velocities for use in precision experiments, thus expanding
the capabilities of a fragmentation facility.

We have developed a new high efﬁciency, high pressure, gas ﬁlled ion source for
radioactive ions. The source utilizes a high pressure (~1 bar) helium gas cell which
stops fast (~100 MeV / nucleon) radioactive ions. These exotic nuclei are thermalized
in the helium and extracted with a combination of electrodes and gas ﬂow through a
nozzle. The ions will then be captured in a radiofrequency quadrupole rod structure
in a differentially pumped expansion chamber and then transported downstream for
use in precision nuclear science experiments.

Presented in this volume are the preparations, results and interpretation of the
stopping and extraction efforts. The features and operation of the major components
will be addressed. For radioactive beams spanning a large range of masses (A~40 to
120), the source has shown the ability to stop (typically) 30-70% of the ions introduced

into the gas cell. The combined stopping and extraction efﬁciency has been shown

to be approximately 0.5%. These results demonstrate that this source will be useful
in providing a low-energy, high-quality ion beam to precision experiments for many
new nuclides produced in the N SCL Coupled Cyclotron Facility despite their short
half-life (down to ~100 ms) and low production rate (as low as ~100/s). In addition,
this method of stopping fast beams shows promise for use in future radioactive beam
facilities such as RIA (the Rare Isotope Accelerator). Finally, future work and planned

improvements are presented.

to those who have come before me, who have undergone the fatigue of building the

mods and bridges I now traverse

iv

ACKNOWLEDGMENTS

My thanks starts with the organization known as the NSCL, an exceptional lab
with a dedicated staff. I ﬁrmly believe that I was provided more than the standard
graduate education in chemistry. My degree was completed with the contribution of
more than 100 employees of the lab, everyone from professors to welders to beam
tuners and machinists. The experience of interacting with all these people afforded
me an experience like few other students receive.

The faculty of the NSCL is a group to be admired. The contributions of my
committee to my education will always be remembered. First the greatest thanks go
to my advisor, Professor David J. Morrissey. His value to my education and to the lab
have not gone unnoticed. His mantra on dealing with his group “Everyone’s different”
certainly applied to me, as I sometimes took less-traveled paths in completing my
degree. Professor Paul Mantica was my second reader and sometimes second advisor.
I thank him for making the phone calls which brought me to MSU and for being
the voice ’whispering in the wind’ when I needed additional guidance. To Professor
Thomas Glasmacher I owe a debt for making me an adjunct member of his group
while I was temporarily in need of a project. I recall him often asking (with a mild
German accent) “How would they do that in the Army?” which made me feel more
important....if even for a moment. Finally, my thanks go to Professor Brad Sherrill,
who I can only describe as a man of ‘quiet efficiency.’ He is the ﬁnest example of an
intelligent soul who pauses slightly before giving the most thoughtful answers and
advice.

The next bunch deserving of my thanks are fellow students and group members.
To those students of NSCL Past who made my start at the lab more tolerable: the
wise guys (Jac Caggiano, Jon Kruse, Chris Powell, Erik 'D'yggestad) who continually
reminded me that there was life outside the laboratory walls, and to my academic

family (brother Don Anthony, step-cousins Joann Prisciandaro and Dan Groh) who

helped me along time after time. Don and Joann, to you I owe more than I can repay.
To all of the students of NSCL Present I thank you for your friendship and tolerance,
especially Katie Miller, Jennifer Church, Mark Wallace, Michelle Ouellette, Heather
Olliver, and James Russell Terry. To the LEBIT group members, who have been
instrumental in the continuing success of my project: Professor Georg Bollen, Stefan
Schwarz, Ryan Ringle, Peter Schury and Sun Tao. Finally, I give special thanks to my
group members Leo Weissman and Debbie Davies. Without their many great efforts,
I would still be months away from writing these acknowledgments.

As a last thanks to the lab, I recognize the contributions of the many staff members
who ensured that my work succeeded. Speciﬁc praise is for the welders, the machinists,
the plumbers, the coil-winders, the electricians, the administrators, the cryogenicists,
the designers, the cyclotron operators, and the beam tuners who make science at the
lab possible. Special recognition goes to Mr. Jack Ottarson in design for withstanding
my excessive visits to his oﬂice. Although an older and wiser gentlemen, he somehow
was empathetic in feeling my pain. I thank him for not having the meter running
while I usurped his time.

The last body I wish to thank is in the form of the many inﬂuences on the life
of Patrick A. Lofy. To the greatest inﬂuence in my life, my family, for my fortunate
upbringing. My parents, Francis and Dona (with one ’n’) have always provided for me.
To the church which taught me to respect others and to the Boy Scouts which helped
a young boy grow up. To my high school chemistry teacher, Mr. Brian Molle (and
his strong family) for the inspiration which sent me down the path of chemistry. To
two undergraduate professors who taught me both chemistry and important lessons
of life: Corporal Roy Shaver for the guidance and mentoring of a wise old owl and
Lieutenant (j.g.) Harold Fenrick for his wit and wisdom and enthusiasm for teaching.
Finally, I give thanks to the United States Army for its positive inﬂuences on me and
to the military men and women who serve and have served to grant me the privileges

which I enjoy now.

vi

Contents

1 Introduction
1.1 Motivation for Nuclear Physics Studies .................
1.1.1 Nuclear Structure Motivation ..................
1.1.2 Astrophysical Motivation .....................

1.2 Precision Experiments ..........................
1.2.1 Penning Traps ...........................
1.2.2 Mass Measurements .......................
1.3 The NSCL Gas Stopping Cell ......................
1.3.1 Production and Use of Radioactive Beams ...........
1.3.2 The N SCL .............................
1.3.3 LEBIT Methods and Objectives .................
1.3.4 Similar Efforts ...........................
1.4 Overview of LEBIT / Gas Stopping Station ..............
1.5 Thesis Organization ............................

2 Gas Stopping
2.1 Energy Loss of Ions in Matter ......................

2.1.1 Background ............................
2.1.2 Energy Loss Models .......................
2.2 Introduction to Gas Stopping ......................
2.2.1 Design Considerations ......................
2.2.2 Equipment .............................
2.3 Primary Beam Measurements ......................
2.3.1 Experimental Setup ........................
2.3.2 Primary Beam Stopping Data ..................
2.3.3 Conclusions for Primary Beam Measurements .........
2.4 Secondary (radioactive) Beam Measurements ..............
2.4.1 Range Compression ........................
2.4.2 Equipment for Secondary Beam Measurements .........
2.4.3 Secondary Beam Stopping Data .................
2.4.4 Conclusions for Secondary Beam Measurements ........
2.5 Gas Stopping: Summary and Conclusions ................
3 Gas Flow
3.1 Concept ..................................
3.2 Design Considerations ..........................

vii

mmNOUwMI—‘H

3.2.1 Choosing a Gas ..........................

3.2.2 Stopping Gas Purity .......................
3.2.3 The Gas Handling System ....................
3.3 The Gas Jet ................................
3.3.1 Introduction and Background ..................
3.3.2 Gas Jet Properties ........................
3.3.3 The Nozzle ............................
3.4 Gas Jet Calculations ...........................
3.4.1 Background ............................
3.4.2 Description of the Calculations .................
3.4.3 The VARJET Code ........................
3.5 Gas Jet Measurement Data .......................
3.5.1 Introduction ............................
3.5.2 The Pitot Tube ..........................
3.5.3 Gas Jet Measurements ......................
3.6 Results and Interpretation ........................
3.6.1 Comparison to Calculations ...................
3.6.2 Nozzle Gas Jet Characteristics ..................
3.6.3 Simple Oriﬁce Characterization .................
3.6.4 Gas Jet Measurement Conclusions ................
4 Ion Extraction
4.1 Introduction ................................
4.1.1 Extraction Requirements .....................
4. 1.2 Equipment .............................
4.2 Extraction Measurements ........................
4.2.1 Experimental Setup ........................
4.2.2 Extraction Data ..........................
4.2.3 Interpretation ...........................
5 Summary and Future
5.1 Data Summary and Interpretation ....................
5.1.1 Efficiencies .............................
5.1.2 Mass Measurement Prognosis ..................
5.2 Future Work and Improvements .....................
5.2.1 Improvements ...........................
5.2.2 Related Other Improvements ..................
5.2.3 Gamma Detection after RFQ Transport ............
5.2.4 Laser Studies ...........................
5.2.5 LEBIT Equipment Status ....................
A Beam Transport
A.1 Background ................................
A2 The Gas Stopping Station RFQ .....................
Bibliography

viii

56
57
58
60
60
60

68
70
70
70
71
73

74
74
74
75
80
81
83
85

88
88
88
90
91
91
92
93
93
93

95
95
97

100

List of Figures

1.1 Chart of the N uclides ........................... 2
1.2 Nuclide Chart with some Stellar N ucleosynthesis Processes ...... 4
1.3 Comparison of Mass Models ....................... 6
1.4 Penning 'Ii'ap ............................... 6
1.5 ISOL Method for Radioactive Beam Production ............ 9
1.6 Ptagmentation Production of Radioactive Beams ............ 10
1.7 Beam Production Rates for the Coupled Cyclotron Facility ...... 12
1.8 The A1900 Fragment Separator ..................... 12
1.9 Diagram of Simpliﬁed Gas Stopping Concept .............. 14
1.10 LEBIT Schematic ............................. 16
1.11 Schematic of the Gas Stopping Station ................. 17
2.1 SRIM Calculation for Ions Stopping in Beryllium ........... 22
2.2 SRIM Calculation for Ions Stopping in Beryllium and Helium Gas . . 23
2.3 Schematic of Gas Stopping Beamline and Equipment ......... 25
2.4 Counter-rotation of Glass Degraders ................... 28
2.5 Glass Degrader Ladder .......................... 29
2.6 Silicon Detector Stack .......................... 31
2.7 Example of Raw Data from a Scan of Glass Degrader Angles ..... 34
2.8 Example Stopping Proﬁle ........................ 35
2.9 Energy Spectrum with Fitted ATIMA Curve .............. 35
2.10 Variation of Helium Pressure in Stopping 150 MeV/ A 36Ar ...... 36
2.11 Longitudinal Proﬁles ........................... 38

2.12 Stopping Data for 120 MeV/A l36Xe .................. 40

2.13 Comparison of StOpped Haction for Two Different Beam Momentum

Spreads .................................. 43
2.14 Range Compression Concept ....................... 44
2.15 Particle Identiﬁcation in Secondary Beams ............... 46
2.16 Gas Stopping Station’s Range Compression Optical Elements ..... 47
2.17 Dispersion Measurement Results ..................... 48
2.18 The Eﬂect of Incident Momentum Spread on Stopping Fraction . . . 50
2.19 The Effect of Adding a Mono-energetic Wedge ............. 51
2.20 Mono—energetic Wedge’s Effect on Energy Spectrum .......... 52
2.21 Comparison of Mono-energetic and Homogeneous Wedges ....... 53
3.1 The Gas Handling Manifold ....................... 58

3.2 The Gas Cell, Differentially Pumped Chambers and Gas Handling System 61
3.3 Supersonic Nozzle Schematic ....................... 62
3.4 Example of Calculation Grid for VARJET Code ............ 64
3.5 VARJET Calculation of Free Expanding Jet from Supersonic Nozzle . 65

3.6 Pitot Tube and Nozzle .......................... 67
3.7 Pitot Tube photograph .......................... 67
3.8 Pitot Tube measurements ........................ 68
3.9 Pitot Tube Measurement of Supersonic Nozzle Free Jet ........ 70
3.10 Comparison of Nozzle Calculation and Measurement ......... 71
3.11 Comparison of Oriﬁce Calculation and Measurement ......... 72
4.1 Extraction Scheme ............................ 76
4.2 Gas Cell Drift (Ring) Electrodes ..................... 77
4.3 Gas Cell Spherical Extraction Electrodes ................ 79
4.4 Gas Cell Extraction Simulation ..................... 80
4.5 Extraction Trial Schematic ........................ 81
4.6 Extraction Data: Gamma Singles and Beta Spectra .......... 85

4.7
4.8

5.1
5.2
A.1
A.2
A.3
A.4

Extraction Data: Gamma-Beta Coincidences .............. 86

Extraction Test: Beta Detector Response for Glass Degrader Angle

Change .................................. 86
Mass Measurements: Number of Required Ions ............. 91
Current Status of LEBIT ......................... 94
Basic Radio Frequency Quadrupole (RFQ) ............... 96
Radio Frequency Quadrupole (RFQ) Effective Potential ........ 97
Gas Cell and Radiofrequency Quadrupole (RFQ) Schematic ..... 97
Gas Stopping Station’s Radio Frequency Quadrupole (RFQ) ..... 98

Images in this dissertation are presented in color.

List of Tables

1.1
1.2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.1
3.2
3.3
3.4
4.1
4.2
4.3
4.4
5.1
5.2
A.1

Fragmentation Beams vs. Precision experiment requirements ...... 13
Comparison of Gas Stopping Efforts ................... 15
Summary of Degrader Materials ..................... 26
Glass Degrader Plate Summary ..................... 27
Counter-rotation of Glass Degraders ................... 28
Summary of Primary Beam Experiments ................ 39
Secondary Beam Purities ......................... 45
Dispersion Measurement ......................... 47
Summary of Secondary (Radioactive) Beam Experiments ....... 48
Secondary Beam Parameters ....................... 49
Gas Stopping Summary ......................... 54
Ionization Potentials of Sample Elements ................ 57
Ionization Potentials of Common Compounds ............. 57
Pumping Equipment Characteristics .................. 61
Sample Data from Pitot Tube Measurements .............. 69
Gas Cell Optimized Electrode Voltages ................. 80
Gas Cell Electrode Voltages Used in Experiment ............ 82
Extracted Beams ............................. 83
Extraction Efficiency ........................... 87
Data Summary with Efficiencies ..................... 89
Comparison of Efﬁciencies with other Gas Stopping Efforts ...... 89
RFQ Voltages ............................... 99

xii

Chapter 1

Introduction

1.1 Motivation for Nuclear Physics Studies

The span of pure research in the ﬁeld of nuclear science can largely be summarized as
the further understanding of our universe in two areas: nuclear structure and (astro—
physical) nucleosynthesis. It is our natural curiosity which drives us to comprehend
the details of some of the smallest objects as well as the processes which occur in the
some of the largest objects in the universe that created these tiny objects billions of
years ago. Our learning more about both ﬁelds leads to a ﬁrmer grasp of both the
composition and genesis of our world.

The study of radioactive nuclear species, nuclei not common in our everyday life,
provides insight into the constitution of and the processes which make up matter
and energ in our universe. Radioactive nuclear beam (RNB) research is the focus
of the experimental program at the National Superconducting Cyclotron Laboratory
(NSCL). It is the ﬁrst step in the manipulation of these RNBs, gas stopping, which

is the focus of the present work.

1.1.1 Nuclear Structure Motivation

The landscape of nuclear structure is conveniently represented by the chart of nuclides,
or Segré chart (Figure 1.1). The horizontal axis is the neutron number, while the
vertical axis is the proton number. The dark squares along the central diagonal axis
denote the positions of the less than 300 stable nuclei. Moving away from this line
of stability, the next set of nuclides (in yellow) are those which have been measured
in some form. The remaining nuclei, out to the left (proton drip line) and right
(neutron drip line) extremes are yet unknown, but are predicted by various theories
to be particle stable; that is, each of the nuclei in the green areas should survive long

enough to be studied.

  
   
   
  

\

terrarincogniita l

proton number, Z

——->
neutron number, N

Figure 1.1: Simpliﬁed chart of the nuclides.

It follows then, that one of the ﬁrst questions posed by nuclear science might be
the limits of nuclear existence (including the search for the heaviest at the upper right
of the chart). Once boundaries, or the so—called limits of stability are experimentally
determined (even in part), then the quest for understanding properties of all nuclei
inside these boundaries suggests follows.

From an experimental standpoint, an entire hierarchy of the observable properties
of an unknown nucleus is open to study. These properties range from the mass, half-

life and decay modes, through the nuclear spin and moments, and on to detailed

spectroscopy of its energy levels. At present, the body of information available for a
given nucleus depends largely on the diﬂiculty of producing it in a nuclear reaction.
Fusion / evaporation reactions and ﬁssion have provided the majority of nuclei during
the last half-century of studies. In the last approximately 20 years, a third method,
projectile fragmentation, has developed to the point that it is widely used to create
new nuclei. The NSCL and a few other accelerator facilities were designed speciﬁcally
to be nuclide factories based on the projectile fragmentation technique. These RN B
facilities, distributed worldwide, are capable of delivering radioactive nuclei which
can be studied to provide experimental observables such as nuclear masses (binding

energies), half-lives, excitation energies and production cross sections.

1 . 1 . 2 Astrophysical Motivation

The task to completely characterize the nuclei within the limits of stability takes
on greater signiﬁcance if these nuclei are involved in astrophysical processes, making
their properties important to understanding stellar evolution. In particular, since it is
well accepted that the stars are the factories which build up nuclear mass (i.e., respon-
sible for the creation of all nuclei in the universe beyond lithium), knowledge of the
properties of nuclei are critical to understanding nucleosynthesis and the distribution
of the elements in the universe.

More speciﬁcally, some of the processes in stellar nucleosynthesis are thought to
involve nuclei far from stability. The r-process [1,2] and the rp—process [3] (see Figure
1.2) are the most prominent examples. In addition, energy evolution is a by—product
of the nuclide building processes in stars and the reaction cycles that produce this
energy often encompass radioactive nuclei. The participation of unstable nuclei (some
far from stability as in the r-process) makes the study of RNBs vital to understanding
the composition of our universe.

Nuclear astrophysicists attempt to understand the many processes taking place in

stellar and explosive environments by using reaction rates to predict the distribution

 
   
 

. rp-processl ' ' ’ r-process

, I MOSS known
1 ‘ I Holf-IIfe known
pro ons u nothin known

Figure 1.2: Chart of the nuclides with the approximate paths of the rp— and r- stellar
nucleosynthesis processes.
of all elements in the universe. By using measured reaction conditions (temperatures,
pressures, particle abundances) and measured reaction parameters (cross sections,
decay rates, binding energies) in their network calculations, nuclear astrophysicists
form models of the mass building processes and stellar evolution. Therefore, a pri—
mary contribution of nuclear physics to astrophysics is the measurement of specific
properties of nuclei far from stability, especially those nuclei involved in one of the nu-
cleosynthesis processes. The contribution of RN B3 to astrophysics centers around the
information on masses, half-lives, and cross sections and the reaction energetics and
time scales which they represent. These data enter into calculations which attempt
to model the formation and distribution of elements in the universe.

The need to gather data with high accuracy, achievable in ”precision experiments,”
is crucial to providing greater constraints to both nuclear structure and nucleosyn-
thesis models. The role of a gas stopping cell at a projectile fragmentation facility

will be made clear in subsequent sections.

1.2 Precision Experiments

The kinetic energy per particle of the radioactive beam fragments produced at the
NSCL is typically of the order of 100 MeV/ nucleon (~0.4c). The energies of these
nuclei are suited to nuclear structure experimental techniques such as intermediate
energy Coulomb excitation [4] and decay studies where the ions are implanted in
a solid [5]. However, other precision studies [6,7] require intensive manipulation of
the radioactive ions and / or low beam energies. These requirements demand further
treatment of the fast beams produced at a fragmentation facility before their use in
such studies. Penning Trap mass measurements and selective ionization by lasers are
two examples of such handling of ions for precision experiments.

Nuclear masses are a well-known example of the importance of precision experi—
ments. Perhaps the most fundamental and essential prOperty of a nucleus that can
be measured is its mass [8]. It is, therefore, the measurement of this property which
is ﬁrst among efforts to study a particular nucleus. Ground state masses reﬂect the
total of all interactions in the nucleus. Nucleon binding or separation energies are
determined directly by mass differences. The separation energies point out the loca-
tion of the boundaries of stability. In addition, mass differences give information on
residual interactions (pairing) and deformations of the nucleus. Since mass diﬂ'erences
are, essentially, the source of energy in stars, knowledge of binding energies constrains
nucleosynthesis calculations.

Continuing with our example of nuclear masses, the need for precision mass mea-
surements can be illuminated with the following illustration (Figure 1.3). The case
for the rubidium (Z237) isotones is presented [7], where the mass differences from
6 different models are plotted as a function of neutron number. In the region where
the masses have been measured, model predictions show excellent agreement. How-
ever, above N264, where masses have not been measured, the models diverge wildly.

Clearly, accurate and precise mass measurements of nuclei far from stability would

provide a signiﬁcant test for these models.

 

 

model differences [Me\/]

-6 '_ iii 1992 _‘

 

 

 

35 40 45 50“ 55 50 ‘55 7o 75 so
N(Z=37)

Figure 1.3: Comparison of mass model calculations for Z=37 isotones [7].

1.2.1 Penning ’Iraps

To date, the most precise mass measurements are being performed with Penning
Traps. Although trap measurements have the capability to contribute in other areas,
their ability to directly determine nuclear masses with extreme precision on just a
few ions has made them the choice for these studies.

By using a static electric quadrupole ﬁeld, provided by a ring electrode and hyper-
bolic caps (Figure 1.4) combined with a strong, static homogeneous magnetic ﬁeld,

the measurement of even a single stable or long-lived ion can yield an accurate mass.

 

Figure 1.4: Simpliﬁed Penning Trap diagram.

 

In principle, Penning Traps act as high precision mass spectrometers by measuring

the cyclotron frequency, lie, of stored ions.

Vc = (Q/m) ' (13/270 (1-1)

where q is the charge of the ion, m is the sought after mass and B is the magnetic
ﬁeld of the trap.

The difficulty with these measurements is they require injection of ions into the
trap at an extremely low and accurately controlled velocity; that is, a kinetic energy

on the order of eV is necessary.

1.2.2 Mass Measurements

The measurement of nuclear masses fall largely into two categories: indirect (such as
reactions and decay measurements) and direct (mass spectrometers, ion traps and
time-of-ﬂight) methods. While the indirect methods have produced mass precisions
below the 100 keV level [9], they are usually not suited to measure masses far away
from stability, as such nuclides’ masses would be dependent on a chain of measure-
ments with diminishing statistics as one studies nuclei further from stability. And,
though a direct time—of-ﬂight (TOP) method [10,11] is suitable for exotic (far from
stability) nuclei, its precision is typically limited to ~100 times worse than Penning
'Ii'ap measurements.

Penning Traps have been used to measure the mass of ions with half-lives as low
as 65 ms [12] or with as few as one stable ion. The resolving power and statistical

accuracy depend on experimental conditions. The resolving power, R, is given as

R = Vc/AVc(FwaM), (1.2)

which depends on an observation time limited by the decay of the exotic nucleus

under study,

AVc(FWHM) z 1/Tobs- (13)

The accuracy of the mass determination is written
rim/m s 1/(R-N1/2), (1.4)

where N is the number of detected ions. Thus, the efﬁcient collection and delivery of
very short-lived ions to the trap is key.

Typical Penning Trap results have an accuracy of 10—7, which translates to a mass
uncertainty of 10 — 20 keV. Increasing the number of ions detected and improving the
emittance of the ion beams could improve this accuracy by a factor of 100. So, while
the current knowledge of masses of nuclei away from stability are often known only to
a few hundred keV (and the masses of nuclei far from stability are mostly unknown),
Penning Trap data could improve on the statistical precision of many of these same

nuclei by factors of 10 to 100.

1.3 The NSCL Gas Stopping Cell

The importance of precision experiments in the realm of nuclear science has been
clearly established. In this section, the case for the construction and Operation of a

gas stopping cell at a fragmentation facility will be presented.

1.3.1 Production and Use of Radioactive Beams

The production of radioactive beams is dominated by two methods: Isotope Separa-
tion on Line (ISOL) and Fragmentation (or in-ﬂight separation). As will be shown,
these methods are complementary rather than competing techniques—each has its own
niche in the scheme of nuclear physics experimental work. The variety of beams, beam

energies and beam qualities that each offers is valuable to nuclear physics.

Isotope Separation On Line (ISOL)

The ISOL facilities generally have peak production for neutron-deﬁcient nuclei near
the valley of stability with large chemical selectivity. The typical ISOL facility makes
use of a production accelerator which accelerates a light beam, often protons, at a
thick production target (see Figure 1.5). The neutral reaction products diffuse from
the target matrix and out through a transfer tube to an ion source where they can
be selectively ionized. The ions leaving the ion source are accelerated to low energies
before they are electromagnetically separated. Post-acceleration is possible, but not

necessary for mass measurements.

Transfer tube

Isotope/
Isobar
Separator

 
 
  
   

 
    

PdeUCflon Tthk, HOI
Beam Target

Figure 1.5: The ISOL (Isotope Separation on Line) method for production of radioac-
tive beams.

While this method suffers from the dependence on the chemistry of the reaction
products as they diffuse from the target, its more severe limitation, when compared
to fragmentation methods, is the effect of the diffusion time on yield through the
half-lives. Usual transfer times are on the order of one second. Although short-lived
nuclides (~100 ms) have been measured with this method in special cases. Neglecting
the types of beams produced, however, ISOLs superior feature is the excellent beam

quality (low emittance and precise energy) which is well suited for precision studies.

Projectile Fragmentation (In-ﬂight Separation)

The forte of fragmentation facilities is the rapid delivery of nuclei far from stability. A
fragmentation facility relies on, most often, the acceleration of a heavy beam of ions
(up to ~0.5c) directed at a thin production target (see Figure 1.6). Heavy projectiles
which encounter target nuclei can be abraded, giving a variety of forward-focused
fragments, which are separated, in ﬂight, by a fragment separator utilizing strong

magnetic ﬁelds in the so—called Bp - AE - Bp technique [13—16].

/

    
    
 
 
   

Fragment
Separator

Secondary
Beam

.4,

Figure 1.6: The fragmentation method for production of radioactive beams.

    

Primary Thin
Beam Production
Target

Comparison of ISOL and Projectile Fragmentation Methods

The nuclear reactions at the heart of the two techniques are very similar. Generally, a
high kinetic energy beam interacts with a target creating a broad range of nuclei and
ﬁssion products. The major difference is that projectile fragments retain their high
velocity while ISOL target fragments are created essentially at rest (in the laboratory
frame). Collecting and separating out high velocity residues is clearly a different task
from sorting out a thermalized mixture of target residues. The projectile fragmenta-
tion process with in-ﬂight separation has been extremely successful in providing fast
beams of the most exotic nuclei.

Compared to the ISOL method, fragmentation provides more exotic beams, but

the beam velocity remains high and the optical properties of the beam are poor due

10

to the effects of the target and wedge speciﬁcally, the fragments retain most of the
velocity of the accelerated heavy ion beam. They can have a large emittance and non-
negligible momentum spread which is proportional to the acceptance of the separator
and the yield. Finally, depending on the desired nuclide, beam purities can vary over
a wide range (1-99%), although the capability to perform beam tracking can aid the
experimenter in dealing with background.

In the case of certain beams, especially those very far from stability with short half-
lives, it would be beneﬁcial to unite the best features of the two production methods,
thereby allowing precision measurement of nuclides previously unavailable. It will be
shown that adding the capabilities of a gas stopping station to a fragmentation facility

(NSCL) is essential to making precision measurements on the most exotic nuclei.

1.3.2 The NSCL

The National Superconducting Laboratory (NSCL) at Michigan State University is
an example of a third generation fragmentation facility [17]. The ﬁrst dedicated device
was the LISE separator at GAN IL [18]. Second generation and improved separators
included the A1200 at the NSCL [19], the FRS at GSI [20], and RIPS at RIKEN [21].
With the recent upgrade completed of the Coupled Cyclotron Facility (CCF), the
NSCL has expanded the explorable regions of the chart of nuclides up to the mid-
mass region (A~120), extending nuclear science experimental possibilities closer to
(or out to) the limits of stability (Figure 1.7).

The A1900 Fragment Separator at the NSCL has the largest acceptance of any
such device and can provide beam with magnetic rigidities up to 6 T-m and with
a momentum spread of :l:2.5% [22]. The A1900 uses a system of superconducting
dipole and quadrupole magnets (see Figure 1.8) to separate the beam fragments by
mass/charge ratio, making a Bp selection in the ﬁrst half, matched energy loss in the
middle and a second Bp selection in the second half.

The beams resulting from the fragmentation and subsequent separation can then

11

 
  
   

Yield #/3

I>1OZ
l> 10
l> 10
l> .015
I>IU

Proton Number

13

E

o

8

(n
,m

‘9

01
o

Neutron Number

Figure 1.7: Beam production rates for the Coupled Cyclotron Facility (CCF).

Intermediate
Focal Plane Image Plane
(Image 41' (Image 2]
‘ ll Production

Target

 
  

. * .
Rare Ion - T" . . -

Beam \_

 

 

- . Primary
Beam

    

Figure 1.8: The A1900 Fragment Separator.

be delivered to experimental endstations. The properties of these beams are highly
disjoint with the requirements for precision experiments (Table 1.1).

Again, many of these required beam qualities are characteristic of beams created at
an ISOL facility. However, the desire to produce radioactive beams far from'stability
(fragmentation facility) and to produce them with these excellent beam qualities
necessary for precision measurements is the greatest motivation for our gas stopping

efforts.

12

Table 1.1: Properties of Fragmentation Production beams compared to Requirements
for Precision Experiments

 

 

Fragmentation production provides: Precision Experiments require:
High energy beams (~100 MeV / nucleon) Extremely low energy (~1eV / nucleon)
Large energy spread Low energy spread
Large emittance Low emittance

DC beam Pulsed beam
Possible contaminants Pure beams

 

 

 

 

1.3.3 LEBIT Methods and Objectives

To have the best of both worlds (very rare beams and good beam quality), we are
developing a gas stopping device. The requirements are stopping the fast radioactive
beams and re-emitting them at lower energies with low angular energy spread. Cre-
ating such an ion source is the leading objective for the collaboration at the NSCL
known as LEBIT (Low Energy Beam and Ion Trap). LEBIT [23] intends to stOp ions
in a buffer gas, strip the gas off and transport the ions to endstations for precision
experiments. The ﬁrst experiment will be mass measurements in a Penning Trap
(already under construction).

The LEBIT method to stop fast radioactive beams is similar to that described by
Geissel et al. [24]. The incoming ion beam will be met in the beamline vacuum by
degraders which will attenuate most of the beam energy (see Figure 1.9). With their
remaining energy, the ions will pass through a thin window acting as the pressure
barrier between vacuum and the ~1 bar helium in the gas cell. The degraders must
be variable as to allow the experimenter to take away the proper amount of energy
from the ions such that the ions will come to rest in the helium gas, with the greatest
stopping efﬁciency. Through a combination of electric potentials and gas ﬂow, the ions
(and the helium) will be extracted through a nozzle. The ions will be captured in an
ion guide while the helium will be skimmed away in a differentially pumped system.
The ion guide will direct the ions downstream to the experimental endstations for

precision studies.

13

Degrader(s) Gas Cell Ion Guide

 

 

0000000000

Beam
—————>

 

0000000000

 

 

 

 

Ring Electrodes Nozzle

Figure 1.9: Diagram of simpliﬁed gas stopping concept.
1.3.4 Similar Efforts

The NSCL is not alone in its attempt to construct and operate a gas catcher. The
expanded opportunities that a gas stopping cell can offer are being introduced at
several high energy facilities: Argonne (Illinois), SHIP and FRS (GSI, Germany), and
RIKEN (Tokyo, Japan).

The ion stopping and collection in a helium buﬂer gas follows on the IGISOL
(Ion-Guide—ISOL) system [25,26] developed originally at Jyvaskyla, Finland, and
implemented in several labs around the world [27, 28]. The name implies the IGISOL
technique is a version of ISOL that relies on thermalization of the products, close
to or even containing the target, in a buffer gas. This technique is now applied to
therrnalize fast reaction products that already have been separated.

As shown in Table 1.2, a variety of cell sizes and pressures are under development.
The RIKEN gas cell extracts ions at right angles to the beam direction from a long
(two meter) gas cell. Although helium appears to be the stopping gas of choice, argon
has been tested.

While it seems clear that the gas cell should have the highest pressure gas with
the highest atomic number in order to have the greatest stopping power for fast ions,
several other factors must be considered. High atomic number gases such as krypton
and xenon are efficient stopping media but they are relatively polarizable and form
adduct compounds with the ions that slow their migration and need to be broken

before the ions are usable.

I4

Table 1.2: Comparison of gas stopping efforts [29].

 

 

 

 

 

 

 

 

Gas Cell Gas Incoming
Length Pressure Beam Energy

Lab, Location [MeV/ A]
AN L, Argonne [30,31] He 20 cm 150 mbar ~5
GSI, Darmstadt (SHIP) [32] He, Ar 18 cm 100 mbar ~5
RIKEN, Japan [29,33] He 200 cm 130 mbar ~50
N SCL Gas Cell He 50 cm 1 bar ~100
GSI, Darmstadt (FRS) [31] He 125 cm 0.5 - 1 bar 100 - 1000

 

The most prominent aspect of the NSCL’s initiative is the incoming beam energy
which the gas cell will meet, highlighted in the last column of Table 1.2. The beam
energies of the N SCL’s Coupled Cyclotrons are typically above 100 MeV/nucleon—
greater than the energies of beams being introduced into the other gas cells (the gas
cell listed last in Table 1.2 is only beginning to come online). This fact makes the
attempt to efﬁciently stop and quickly extract rare, short-lived radioactive beams
even more challenging due to the growth of the residual energy distribution as the
ions slow down. We have chosen to use a high pressure gas cell with static electric drift
ﬁelds. A novel feature of the cell is that a bias will be applied across the extraction
nozzle itself.

Similarly, the beam energies encountered at the NSCL’s gas cell also ﬁgure promi-
nently into the possibility of stopping beams at the next (planned) radioactive beam
facility, RIA (The Rare Isotope Accelerator). Currently, the ~100 MeV / A energies
of the NSCL are nearest the planned beam energies of ~400 MeV/A at RIA. The

prospects of gas stopping at RIA will be greatly inﬂuenced by the N SCL experiments.

1.4 Overview of LEBIT / Gas Stopping Station

Radioactive beams from the A1900 Fragment Separator will be delivered to the N4
Vault at the NSCL. The fast beam will be degraded and brought to rest in the

high-pressure helium in the gas cell of the Gas Stopping Station (see Figure 1.10).

15

A combination of electric potentials and gas ﬂow will extract the helium and ions
through a nozzle. The ions will be captured in a radiofrequency quadrupole (RFQ)
ion guide while the helium will be skimmed off in a differentially pumped system.
The now low energy beam will leave the shielded N4 Vault and the Gas Stopping
Station and travel via the LEBIT beam optics into the accumulator/buncher. From
here the pure, low-energy, low-emittance ion beam will be delivered in pulses to an
endstation for a precision measurement. Currently, a 9.4 Tesla Penning Trap is under

construction to perform precision mass measurements.

I’M

   

6053153122."? COOIer/buncher ”Q '9 Ir, -

9.4 Tesla
% Penning Trap

for high
resolution mass
Beam lines for measurements

 

 

 

. 5-50 keV
Beam from Pulsed beams for
the A1900 shielding other experiments

Figure 1.10: Overhead view of LEBIT (Low Energy Beam and Ion Trap).

Closer examination of the Gas Stopping Station (the focus of this work) in Figure
1.11 illustrates the main features of the gas-stopping scheme. The beam energy will
be attenuated by a pair of variable glass degraders and a mono-energetic wedge, both
of which are discussed in Chapter 2. Ions pass through the thin beryllium window
and are stopped (with maximum efﬁciency) in the high-pressure helium of the gas
cell, a re-entrant chamber in Cross A of the system.

Extraction will be performed by electric potentials (inside the gas cell) and gas
ﬂow through a supersonic nozzle. The ions will then be captured in the RFQ ion
guide while the helium is skimmed off and removed from Crosses A, B and C by
the high-capacity differential pumping system . The ions are then transmitted by the

LEBIT beam optics downstream to the accumulator/buncher for preparation for use

16

Gas

Stopping
Station

 

  
  

Gas
Stopping

Intermediate
energy beam

from the A1900 .\
- — — ->
f '\

degraders wedge A

 

Figure 1.11: Overhead view of the Gas Stopping Station.

in precision studies.

1.5 Thesis Organization

This thesis will be organized into chapters which detail the trials of the two signiﬁcant
challenges of constructing the new ion source: stopping and extraction.

A series of experiments to measure the energy distributions of primary (mono-
energetic) and secondary (up to :E1% in momentum) beams in the helium gas are
described in Chapter 2.

The details of gas ﬂow through the nozzle and an oriﬁce were measured prior to
assembly of the extraction system and are given in Chapter 3.

The results of ion extraction studies with radioactive ions under operating condi—
tions are given in Chapter 4.

A summary and considerations for further work is given in Chapter 5.

17

Chapter 2

Gas Stopping

2.1 Energy Loss of Ions in Matter

2.1 .1 Background

Heavy ions (A34) passing through matter interact largely through the coulomb force
with orbital electrons of the absorber atoms. Although elastic and inelastic interac-
tions of the ions with absorber nuclei are possible, such interactions are rare and
contribute little to the energy loss of the ions. The velocity of the ions are decreased
as a result of their interactions, ionization or excitation, with absorber electrons.
For ions in a given absorber, the linear stopping power, S, is given as the diﬂerential
energy loss for the ion in the absorber divided by the corresponding differential path

length [34]:
dB

The range, or path length, R, can then be deﬁned as:
3° dE
R(Eo) = f —S— (2.2)
o

The range of an ion in an absorber is most often determined experimentally, and so

18

we limit our discussion to stopping power.
The magnitude of -dE/dx along the path length is also named the speciﬁc energy

loss and is given by the Bethe formula:

 

 

dE 4'rre4q2
——a; — movz NZB (2.3)
where
_ 2m0v2 v2 v2

In both Equations (2.3) and (2.4): v and qe are the velocity and charge on the projec-
tile, mo is the electron rest mass, e is the electronic charge, I is the average excitation
and ionization potential of the degrader and N and Z are the number density and

atomic number of the degrader [34].

2.1.2 Energy Loss Models

The Bethe formulation has been known some 70 years. Improvements, in the form
of corrections to this formula, have been added continually since this time, but have
accelerated greatly in the last 20 or so years. Probably the most widely used models
which are pertinent to the energy range of our experiments are the Hubert, Ziegler
and ATIMA codes.

The Hubert model [35] is actually a series of range and stopping-power tables for
2.5 - 500 MeV / A heavy ions stopping in solids. The model centers around a semi-
empirical formula for a heavy ion effective charge parameterization that was derived
from a set of nearly 600 data points. The data covered projectile mass from 16O to 238U
in the 3 - 90 MeV / A range interacting with targets from beryllium to uranium. This
model is focussed on ions stopping in solids, however, so such an empirically parameter
dependent calculation may not be expected to model gas stopping precisely.

Probably the most frequently used predictor for slowing down heavy ions is SRIM
(The Stopping and Range of Ions in Matter) [36] based on the model of Ziegler et

19

al. [37]. The theoretical base of the Ziegler model is grounded in the Bethe-Bloch
concept; however, this tabular collection of stopping powers also includes a number
of free parameters that are continually updated (normalized) to the growing volume
of experimental data.

The ATIMA (ATomic Interaction with MAtter) code [38] is a direct calculator
which incorporates the classical Bethe-Bloch theory. Additional terms are included
which take into account such effects as the dielectric polarization of the stopping
medium (density effect), shell corrections for the binding energy of target electrons
and displacement of the electrons during collisions (Barkas term). Finally, ATIMA
can also incorporate theory by Lindhard and Sorensen (LS) which makes corrections
for high and relativistic energy ions [39].

The software LISE [40,41] has great utility as both a stopping power calculator
and as an estimator of rates of production of secondary beams. For stopping power
predictions, LISE allows the selection of one of three energy-loss models: Hubert,
Ziegler and ATIMA. The code allows the user to include all of the materials in the

path of the beam and the effects of simple ion optics.

2.2 Introduction to Gas Stopping

The goal of our gas stopping initiative was efficient stopping to study rare beams
in a gas which would allow for quick extraction of short-lived ions and further ma-
nipulation of the ion beam. This chapter is devoted to ion stopping experiments
performed on both stable, mono-energetic (primary) beams and on radioactive (sec-
ondary) beams. The primary beam trials helped us to initially characterize and test
our gas stopping system while the secondary beam stopping runs were the ﬁnal ob-

jective of the gas stopping tests.

20

2.2.1 Design Considerations

In order to construct a working gas stopping cell which would be efficient for a wide
range of nuclide atomic numbers and energies, many parameters were ﬁrst carefully
considered and studied. The challenges of bringing ions traveling at nearly half the
speed of light to rest in a gas were many. A gaseous collector is a relatively thin
stopping medium. In addition, energy-loss straggling and angular straggling will add

to the spread of the spatial distribution of ions stopping in the gas.

Ion Stopping in Gases

As mentioned in the motivation, in order to intensely manipulate the ion beams for
making precision measurements, we had to stOp the ions not in a solid, but in a
gas. For our purposes, we knew that we had to attenuate most of the fast radioactive
beam’s energy (~100 MeV / A) with degraders such that a gas cell of reasonable length
could remove the remainder of the energy and ﬁnally stop the ions.

Immediately the disparity between the densities of solids and gases (typically a
factor of 1000) is obvious. The stopping of heavy ions in a solid is well studied.
However, of stopping of ions in a gas, there is less known. One might consider making
a very long (on the order of several meters) gas stopping cell, but this is not a practical
use of laboratory space and, as discussed in the next chapter, would require too much
time to evacuate the interesting short-lived ions. The more promising approach would
degrade the beam’s energy upstream from the gas cell and then have the gas remove
the remainder of the beam’s energy while bringing the ions to rest. The task is to
remove a very high percentage of the beam’s energy with degraders and construct a
relatively short accompanying gas cell. However, an important factor (and challenge)
must be considered—the spread in the longitudinal distribution over which the ions

stop, better known as energy straggling.

21

Energy Straggling Calculations

Energy loss is a statistical or stochastic process, governed by the random interactions
of the projectiles with, primarily, the electrons in the absorber. A spread in the
resultant energies (and therefore ranges) occurs even for a beam of mono-energetic
particles passing through any material. The width of the energy distribution is a
measure of energy straggling [34]. While the amount of straggling encountered when
stopping ions in a solid is small, gas stopping presents a very different situation.
A simple SRIM [36,37] calculation was performed to model a mono-energetic 40Ar
primary beam at typical NSCL energies impinging on a 7 mm thick foil of beryllium

(Figure 2.1). Shown in the ﬁgure is the characteristic narrow width of ions stopping

 

 

 

 

 

 

 

Ion Ranges l
Ion Range = 6.4I mm

«E Straggle = 10.9 pm

0

‘03- 300
E

.9

3

>. 200
(O

E

9..

U)

E inc
.9

i" ‘ ri

41.21 0

49 __Target Depth [mm] 7,0

 

 

Figure 2.1: SRIM calculation resulting showing narrow stopping distribution of 100
MeV/ A 40Ar ions in beryllium metal.

in the solid material. In this case the range is 6.410 mm with a straggling of 0.011
mm, a relative range/straggling value of only 0.0017. The straggling in the SRIM
calculation is precisely deﬁned as the square root of the variance of the distribution.

If we were to pass the same beam through 20 pm less (that is, 6.390 mm) beryllium
and allow the energy attenuated ions to come to rest in 1 bar of helium gas in a cell
sufficiently long to stop them, we would have the results shown in Figure 2.2. Now
194 mm of helium gas at one bar is required to remove the residual energy and the

increased straggling of ions in the gas is very visible. Here the straggling (width) is

22

 

 

Ion Ranges

Ion Range 2 194 mm
Straggle = 83.4 mm

   

(Atoms/cm31/(Atoms/cm2)

 

 

 

 

 

 

Figure 2.2: SRIM calculation showing an expanded stopping distribution of 100
MeV/A 4”Ar ions coming to rest in helium gas after being degraded in beryllium
metal. Note that in this case, the x-axis represents nearly 700 mm of gas.

83.4 mm, and the relative range / straggling becomes 0.43 for the same mono-energetic
beam. For completeness SRIM calculations were performed for the stopping of the
same argon beam in helium gas alone. Not surprisingly, 53.4 m of helium gas would
be required to stop the 100 MeV/A beam, but the straggling of approximately 100
mm is comparable to the straggling in the beryllium/helium shown earlier.

The usual way that target materials are compared is in terms of mass thickness
(or aerial density). Mass thickness has units of mass / area, and is simply the density of
a material multiplied by its actual thickness. This unit relies on the fact that nuclei
are extremely small and that thin pieces of ordinary, randomly ordered materials
will appear to a high energy beam as if all the atoms were in a single layer. In
this way, materials are placed on a basis that factors out their densities. Comparing
the mass thicknesses of materials leaves only their atomic numbers, Z, as the factor
which contributes to their differing stopping powers. Now, returning to the SRIM
calculation in Figure 2.2, we see that the 6.39 mm of beryllium that removes the
bulk of the energy has an equivalent mass thickness of 1180 mg/cmz, while the ﬁnal
volume of helium at 1 bar, 22 degrees Celsius and 50 cm in length would have a mass
thickness of only 9.8 mg/cmz. This illustrates the difﬁculty of attempting to remove

most of the beam’s energy in a relatively thick degrading material before stopping

23

the ions in a ’thin’ amount of gas.

The ﬁrst gas stopping tests at the NSCL were performed by Baumann et al. [42].
These tests used an 36Ar at 100 MeV/A with an aluminum degrader to stop ions
in a volume containing various pressures of helium gas. They determined that the
homogeneity of the energy degrader and the energy spread of the beam were signiﬁcant
parameters in determining the range distribution of ions stopped in the gas.

This result implies the largest challenge for our gas cell—the problem of the ex-
tended ﬁnal distribution of the high energy ions in the gas. Clearly no design or
equipment can reduce this statistical broadening of the stopping range. So, at best,
we could hope to avoid adding to the width of the stopping, thereby keeping the
fraction of ions stopped within the length of our gas cell at a maximum. Keeping in
mind that the SRIM calculations above were performed for exceptional conditions
(a mono-energetic primary beam passing through ideal degrader materials), we took
every pain to minimize the amount of range straggling added by our system in an-
ticipation of performing gas stopping under non-ideal conditions (secondary beams

having a momentum spread passing through imperfect degrader materials).

2.2.2 Equipment

In its fully functional mode, the Gas Stopping Station is designed to stop ions in the
gas cell, extract the ions and then deliver them to the LEBIT cooler/buncher before
their use in precision experiments (previously diagrammed in Figure 1.10). A more
detailed description of the important gas stopping equipment is shown in Figure 2.3.

Before arriving at the Gas Stopping Station, the beam passes through a thin plastic
detector at the Intermediate Image of the A1900 ftagment Separator. This detector
provides a particle count, helping us normalize the number of ions stopped in the many
runs, while still allowing us to measure the magnetic rigidity of the resulting ions in the
second half of the A1900. Further downstream, the ions encounter the ﬁrst degrader

materials, a pair of glass plates. These remotely controlled variable glass degraders

24

 

 

 

 

q doubgfle gas Ge"
thin plastic :6 WOT ”W's Pin
detector screen screen (I
(A1900) °‘°°‘°'

 

 

 

228:", A f \/ V i WM $EWI__
] /\ /\ I [L ll vacugm helium 1:;

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5-45 cm
10cm_ 166 cm 86cm 50cm
dispersive
focus

Figure 2.3: Schematic of gas stopping bearnline and equipment.

are our only mechanism for precisely controlling the amount of energy removed from
the beam, thus allowing the ions to stop in the gas cell with the greatest eﬂiciency. At
the dispersive focus (indicated in Figure 2.3), the (secondary) beam passes through
a mono-kinetic wedge, where its energy distribution is compressed (more on this in
subsection 2.4.1). The ions then are brought to a more parallel beam by the last
superconducting quadrupole doublet magnet in the beam line. After the magnet, a
large-area (50 x 50 x 0.5 mm) Si PIN detector can be moved in and out of the path of
the beam to perform diagnostics and calibration. The beam leaves the vacuum of the
beam line as it passes through the gas cell’s window (beryllium) and enters the high
pressure helium of the gas cell where it comes to rest. For the stopping trials reported
in this chapter, the beam’s path ends in the gas cell. The extraction and transport of
ions in the gas st0pping system will be discussed in later chapters. Finally, a set of
silicon surface barrier detectors were placed in the gas cell to monitor the energies of
the ions in the gas.

Recalling comparisons by mass thickness, a summary of the degrader materials
follows in Table 2.1.

One of our imperatives was to minimize the amount of straggling contributed

to the radioactive beam by our degrader materials. Therefore, exceptional care was

25

Table 2.1: Summary of degrader materials.

 

 

Degrader Mass thickness Typical Mass thickness
[mg/cmz] experimental actual [mg/cmz]
per cm material thickness [cm]
Glass 2500 0.2 500
Beryllium window 1850 0.1 185
Helium (1 bar, 295K) 0.196 50 9.8

 

 

 

 

 

 

afforded to the quality of the materials selected in each of the energy loss media. More

detail on the glass degraders and beryllium window follow.

Variable Glass Degraders

The majority of the beam’s energy loss will occur in the variable degraders. In ad-
dition, the greatest ﬂexibility in controlling energy loss will be due to our ability to
precisely control the amount of material presented to the beam by these principle
degraders. (Consider that the beryllium window is of a ﬁxed thickness and that the
helium gas has comparatively little stopping power.) For these reasons, it is critical
that these degraders can be accurately adjusted and that they are of superior material
which contributes minimally to the straggling of the beam’s energy.

Optical quality glass was chosen as the material for the variable degrader. OPTI-
MAX Systems Incorporated was chosen as the manufacturer due to the speciﬁcations
they were able to meet. BK7 Schott glass was used to construct plates of a rectangular
prism shape. The area of the plates normal to the beam is 30 mm vertical x 90 mm
horizontal. A variety of thicknesses were requisitioned in order to degrade a broad
spectrum of ions and energies.

Four each of glass plates were manufactured and delivered of nominal thicknesses
1.2, 1.5, 2.0, 2.7, 3.6, 5.0 and 7.0 mm. The manufacturer was unable to consistently
fabricate glass plates of a thinner dimension, thus only one plate each of 0.5 and
0.7 mm nominal thickness was delivered. Upon receipt of the glass plates, extensive

measurements were recorded on their dimensions and mass (Table 2.2).

26

Table 2.2: Summary of measurements of glass degrader plates (averages are for that
particular nominal thickness only)

 

 

 

 

Nominal Average measured Average Density

thickness [mm] thickness [mm] Density [g/cm3] error [g/cms]

0.5 0.472 2.51 0.02

0.7 0.672 2.51 0.02

1.2 1.21 2.50 0.02

1.5 1.49 2.51 0.02

2.0 2.02 2.50 0.02

2.7 2.699 2.503 0.021

3.6 3.641 2.517 0.022

5.0 5.031 2.505 0.023

7.0 7.014 2.505 0.024
Errors 0.008

Overall Averages 2.506 0.022

 

 

 

 

 

The speciﬁcations called for a surface roughness on the order of 10 nm, a maximum
thickness deviation of 1 pm and a total homogeneity of better than 10‘4. Our thickness
measurements using a precision micrometer showed surface deviations of less than 3
pm.

The glass degraders were mounted in a system that could be remotely controlled
and had the ﬂexibility to degrade a wide degrade of nuclides and energies without
having to interrupt the experiment. The mechanism had pairs of glass plates opposite
each other on two ladders. The ladders would counter-rotate against each other by
the same angle (up to 45 degrees) perpendicular to the beam in the vertical plane in
order to change the thickness of glass presented to the beam.

Counter-rotation and a pair of plates were incorporated into the design to cancel
the effect of the small angular divergence of the beam, which would be the case for
secondary beams. In this way, all beam particles would encounter the same degrading
material. See Figure 2.4 and Table 2.3.

The ladder mechanism also allowed remote translation of the target ladder in the
vertical dimension in order to select the glass plate pair of appropriate thickness. A

schematic of the mechanism is shown in Figure 2.5.

27

 

 

 

 

 

Figure 2.4: Counter-rotation of glass degraders. In the left portion, a single degrader
plate is shown for clarity. The right portion shows the pair of plates counter-rotated
about their centers.

Table 2.3: Cancelling effect of counter-rotated glass degraders. The approximation
holds well for small beam angles, 45. For example, at a beam angle of 1 degree, the
approximation is good to one part in 5000 for a glass degrader angle, 0, of 30 degrees.

 

 

 

Beam Beam Angle Plate A Plate B Total
Contribution Contribution Thickness
1 —¢ bx/cos(0 + 45) 6x/cos(0 — ¢) ~ mix/coed
2 0 6:1: /c030 (Ste/c030 2693/c030
3 +¢ 6x/cos(0 — (i3) 6x/cos(9 + (b) ~ 26m/c039

 

 

 

 

 

 

Before each experimental run, the rotation control was calibrated such that the
requested glass angle would match the angle actually produced by the target ladder
mechanism. This was accomplished by ﬁrst inserting a ﬁtted rectangular form be-
tween the target ladders to ensure that they were parallel and at 0 degrees rotation.
Next, after aﬂixing a protractor which indicated the angle in degrees under a pointer
attached to the target ladder drive, a series of angles were set by computer control
and then measured. These data were plotted and show a good linearity as well as
excellent agreement between the requested and measured angle. The error in the an-
gle we assigned to be i025 degree. Since the degrader thickness increases with the
cosine of the rotation angle, the combined thickness of the pair of degrader plates
was chosen for each ion to minimize the angle on the glass, thereby minimizing the

uncertainty in the thickness of the glass presented to the beam..

28

   

    

 

 

 

 

Figure 2.5: Mechanical drawing of glass degrader ladder. Each of the 6 plates of glass
are 30 mm high x 90 mm wide and are of varying thicknesses. The drive mechanism
above the ladders allows for remote manipulation of the angle of the glass.

Beryllium Window

The window’s primary task is to act as a pressure barrier between the vacuum of the
beamline and the high pressure (1 bar) of the gas cell. Ideally, the window should
minimally impact the incoming ion beam. It should degrade the beam as little as
possible, thereby allowing us more ﬂexibility to attenuate beam energy using the
upstream variable glass degraders. Also, like the glass degraders, the window should
be of a homogeneity and ﬂatness which would add as little as possible to the energy
straggling of the beam.

To minimize the amount of energy loss caused by the window, the window should
have two features. First, it should be as thin as possible while still able to function as
a pressure barrier. Second, it should be made up of a material of low effective atomic
number, Z, as the amount of energy loss scales with the Z of the absorber.

To minimize the contribution of energy straggling by the window, it should possess
two further characteristics, like those of the glass degraders. First, it should be made of

a very homogeneous material. Second, the surfaces of the window should be parallel,

29

smooth and ﬂat. In addition, the window must not be too thin that it would deﬂect
due to the difference in pressures on its two surfaces, as a window with a curvature
would not present the same thickness of material to all incoming ions.

During preliminary trials on the stopping of primary beam, we used a disc-shaped
stainless steel window, cut from metal sheet, of thickness 0.59 mm. The stainless steel
window proved a thin but effective pressure barrier for these experiments. However,
due to its possible inhomogeneity and poor surface characteristics, it was a potential
signiﬁcant contributor to the energy straggling.

Our next (and current) generation window was formed from beryllium. Initially,
a pair of 1.5 mm thick, disc shaped windows (diameter 53.2 mm) was ordered from
Brush Wellman Incorporated. We speciﬁed a ﬂatness tolerance of 0.0005 inches (13
pm) with a surface ﬁnish of 8—16 micro-inch (0.2 - 0.4 pm). Our measurements of
the windows showed an actual thickness of 1.499 mm and that the thickness range
(ﬂatness) was within 5 pm of this value. Assay of the windows performed by the
manufacturer showed them to be 99.8% beryllium, the balance being trace amounts
of other metals. We used gold o-rings (1 mm thickness, from Scientiﬁc Instrument
Services, Incorporated) to form the vacuum seal which was successfully tested down
to the 10"6 mbar range.

To give us greater versatility using the glass degraders for energy attenuation, we
explored the possibility of using a thinner beryllium window. Before ordering the 1.5
mm thick windows, Jack Ottarson, of the NSCL, performed deﬂection calculations
using a ﬁnite elements analysis for the static displacement. This result showed a
deﬂection of 0.015 mm. A later calculation gave a deﬂection of only 0.050 mm for a
beryllium window thickness of 1 mm. With this knowledge, a second pair of windows
(1 mm thickness) was ordered from Brush Wellman using the same speciﬁcations. On
receipt of the windows, our measurements found their thickness to be 1.022 mm, with

a ﬂatness within 10 pm of the thickness.

30

2.3 Primary Beam Measurements

The ﬁrst attempts at gas stopping in the new system concentrated on stable, pri-
mary beams accelerated by the Coupled Cyclotrons and passed through the A1900.
Such beams are as close to ideal as possible. They are nearly mono-energetic, have
little momentum spread (determined by measuring the size of the beam spot in the
dispersive image of the A1900, and typically on the order of Ap/p ~ 0.07%) and are
relatively easy to produce. By using the simplest possible beams, we could most easily

observe the effects on stopping of small changes made in our gas stopping equipment.

2.3.1 Experimental Setup

The Coupled Cyclotrons accelerated and the A1900 delivered eight different primary
beams of fully-stripped ions to the Gas Stopping Station in the N4 vault at the NSCL.
After energy attenuation in the variable glass degraders and the beryllium window, the
ions entered the gas cell. The helium (assumed to be at room temperature) pressure
in the gas cell was read by a MKS Baratron 722A capacitance manometer (10,000
torr full range, controlled by an MKS 937A Multi-Gauge Controller).

helium Inlet calibrated scale

SI Detector stack

 

Be window Gas Cell
Figure 2.6: Moveable silicon detector stack on the beam axis inside the gas cell.

A stack of four silicon detectors (300 mm2 active area and, in order, 0.1, 0.5, 1.5

31

and 0.5 mm thick) was mounted on a post, moveable along the beam axis, from the
downstream side of the gas cell (Figure 2.6). These detectors were used for particle de
tection and identiﬁcation and, ultimately, to determine the fraction of ions stopping in
the gas cell. Detector pulses were immediately ampliﬁed by a Tennelec TC178 charge-
sensitive preampliﬁer and the channels were checked with a complementary Tennelec
TC 178P pulser. Data acquisition was triggered by constant fraction discriminators
(CFDs).

Normalization of counts in the silicon detectors was aided by the introduction of
a thin plastic scintillator (~28 mg/cmz) at the dispersive plane of the A1900. Using
the counts from this scintillator, variations in the delivered beam intensity could be
factored out.

The primary beam was attenuated to give approximately 500 particles per second
during each of the one to two minute data taking runs. This counting rate kept the
data acquisition dead time low and was well within the operating limits of the silicon
detectors.

Calibration of the silicon detectors was accomplished by removing the degrader
glass and helium gas, thus causing the beam to pass through the entire stack. Com-
paring the resulting AE peaks with stopping power calculations provided the energy
calibration. Several additional calibration points were measured by introducing thin
(known thickness) detectors or degraders into the Ibeam’s path. Uncertainty in the
energy calibration, which depends on calculations [37, 43], is approximately 35%.

Before examining data on primary beam, we should comment on the magnetic
rigidity, or Bp (give in units of Teslaometer), of the magnetic elements in the A1900
separator and the beam line. As mentioned, the momentum distribution for our pri-
mary beam was estimated at 0.07(1)%. The central momentum of the beams was
known to ~0.1% from the values of Bp of the A1900 magnets. The degraded beam
will have a lower Bp than the fully accelerated beam. Therefore, in all stopping trials,

the value of Bp in the last superconducting quadrupole doublet (see Figure 2.3), which

32

is positioned after the glass degraders, was set to the appropriate value given the en-
ergy of the degraded beam. This adjustment ensured the consistency of beam optics.
Calculation of these values was performed with the energy module of the computer

software LISE.

2.3.2 Primary Beam Stopping Data

Measurements were performed on primary, mono-energetic beams over a range of
eight energies and nuclide masses. As a representative measurement, we offer the
data from stopping tests of 150 MeV/A 36Ar. Glass degraders of thicknesses 2.02
and 5.03 mm were paired and installed in the degrader ladder. The detector stack
(silicon telescope) was positioned on axis in the gas cell, with the ﬁrst silicon detector
45.0(2) cm distant from the beryllium window. The number of primary ions detected
was deduced from integrating spectra. Measurements were made for varying degrader
angles, gas pressures and detector distances.

Shown in Figure 2.7 are raw on-line data with several spectra depicting the re-
sponse from the ﬁrst silicon detector in a series of measurements for 36Ar primary
beam The only change from one spectrum to the next is the angle (and, therefore, the
thickness) of the degrader glass. The response of the ﬁrst detector shows a predictable
trend as more energy is removed upstream by the degraders, leaving successively fewer

ions, with a greater range distribution, to be detected.

Stopping Proﬁle

The transmitted fraction can be deﬁned as the normalized fraction of primary beam
ions that were counted in the ﬁrst silicon detector. Now the measurements made for
a ﬁxed detector distance and helium gas pressure could be summarized in a simple
graph of degrader glass angle versus transmitted fraction.

In Figure 2.8, the number of ions observed in the ﬁrst detector begins to descend

steeply as the amount of energy attenuated by the glass degraders enters a range

33

 

 

 

 

 

 

 

 

 

F _ W H___ﬂ_g_.____ _ M Glass angle Glass thickness
- _ A - ~ - [degrees] [mm]
I 0 7.050
7.05 7.104
*3 10.01 7.159
D
o
O
11.97 7.207
14.01 7.266
.._ -~——- —-- —----— 1 Raw data for 150 MeV/A “Ar
ﬁe >- ~~— 810 Torr hellum pressure
Energy (Channels) 2.02 a 5.03 mm glass plates

 

 

 

Figure 2.7: Raw data from ﬁrst silicon detector (AE) over a range of glass degrader
angles. Data is from the stopping of 150 MeV/ A 36Ar in 810 Torr (1.08 bar) helium
with 2.02 and 5.03 mm glass plate degraders.

where ions can stop in the gas. At the largest angles, too much energy is removed by
the glass degraders and the ions are stopped in the beryllium window.

Note that ions with total energy less than approximately 3 MeV were not counted
as they fell below the threshold of the ﬁrst silicon detector. SRIM calculations showed
that, for example, the range of such 36Ar ions would be less than 2 cm in the 1 bar
helium. Therefore, in the case of the detectors at 45 cm distance in the 50 cm gas cell,
the ions would have stopped in the gas cell. The number of ions under the threshold
energy were calculated and corrected by using the calculated energy spectra (see
Figure 2.9).

The correction is signiﬁcant only at large degrader angles where the transmitted

fraction is small. Therefore, the correction minimally affects the transmission curve.

34

 

 

 

 

 

 

 

 

Stopping Profile: 150 MeV/AN-SS In hellum
0.24
3 0.21'«" B m m :
if (my . m .. .. 7% A8101orr
'E in .
L'- o.15 « . , . - . in
5
E 0.12< I
3 009+ "'
.. ] I
s 0.06
E 0.03» m
0 . - . MEL-u n .
o 4 e 12 is 20
Degrader angle [degrees]

 

 

 

Figure 2.8: Stopping proﬁle of 150 MeV/ A 36Ar ions in 810 Torr (1.08 bar) helium.

 

First detector response for 150 MeVIA “Ar

 

Pressure=800 Torr
Glass angle=12.0° ATIMA (Bp=3.654 Tm)

Counts

 

 

 

 

 

 

 

Figure 2.9: Energy spectrum with ﬁtted ATIMA curve for 150 MeV/A 40Ar ions in
800 Torr (1.07 bar) helium for a ﬁxed degrader angle. The Bp is the adjusted magnetic
rigidity value used in the ATIMA calculation.

Pressure Variation and Primary Beam Calculations

Depicted in Figure 2.10 are stopping proﬁles with a ﬁxed detector distance for the
evacuated gas cell and for 800 Torr (1.07 bar) and 1200 Torr (1.60 bar) helium pres-

sures. The data were ﬁtted to a Fermi function of the form:

1

ffx) = m (2.5)

35

where a: is the degrader thickness, 0 is the midpoint of the descending curve and b is
indicative of the slope. The difference between the stopping proﬁle for a given pressure
and the stopping proﬁle for the evacuated gas cell gives a direct measure of the
fraction of the ions stopping in the gas as a function of degrader thickness [This simple
change of x—axis units provides a linear representation of the degrader materials]. This
difference is represented by the two peaked curves of the ﬁgure. One can see that,
at the optimal degrader thickness, approximately 60% of the ions stop in the 1.07
bar helium while 80% stop in the 1.60 bar pressure. Vertical error bars represent the
combined statistical uncertainty in the A1900’s plastic scintillator, the ﬁrst silicon
detector and the error on the low-threshold correction (using a conservative 50% for
this particular correction). Horizontal error bars represent the combined uncertainties

in determining the glass degrader angle and measuring the thickness of the glass.

 

 

 

 

 

 

 

- - “AUSQMeY/A.

g 11 ,,
E a, P=0 Torr ATIMA , __ , P = 0 7°"
8. Calculations ‘
8' 0.6 39:3‘649 "“ ~>\ ,4
g Bp=3-654 — ‘\ / I
B 04 , ,. ’ ,
g ' P -- 750 TOfTIi/ I
502, P=117010ir/ / _ 5
c . l l
g o - _§:://’ .

7.05 7.10 7.15 7.20 7.25 7.30 7.34

 

 

degrader thlcknese [mm]

 

Figure 2.10: 'Itansmitted (solid colored lines) and stopped fraction (dashed colored
lines) of 150 MeV / A 36Ar under three helium pressures ﬁtted with Fermi functions.
Also shown (in black) are the ATIMA calculations (for 0 Torr) with the experimental
Bp and the adjusted Bp.

Also depicted in Figure 2.10 are curves determined by calculation. Data were
compared with well-known stopping power calculations based on three diﬂerent energy
loss models (Hubert, Ziegler and ATIMA) [35,37, 38] as incorporated into LISE. We
found that the calculations using all of the nominal parameters did not agree well

with the data (shown for an ATIMA calculation in Figure 2.10 above in dashed

36

lines). Very good agreement for the ATIMA calculation (solid lines in Figure 2.10)
was obtained after a small adjustment in Bp of 0.14%. The adjustments for the
Hubert and Ziegler calculations were 0.20% and 0.24%, respectively, each producing
similarly good agreement. Therefore, the ATIMA calculation was selected as the most
closely matched model and was used for the remaining calculations. At this point, the
A1900 had been calibrated by a time-of-ﬂight method to provide a 5x 10"4 level of
uncertainty, and so the ﬁelds which ﬁgure into the magnetic rigidities are taken from
Hall probes in the magnets [44]. Our adjustments (0.14%, for example) are slightly

higher than the uncertainty.

Longitudinal Proﬁles

To this point, data has been presented which showed variations in degrader thickness
and gas cell helium pressure. To make our understanding of stopping in the gas
cell more complete, we examined the remaining untested dimension—the longitudinal
stopping proﬁle along the axis of the gas cell. Shown below in Figure 2.11, is a raw
transmitted fraction versus the distance of the ﬁrst silicon detector from the beryllium
window. Comparing the data in the upper panel shows the expected trend on the
transmitted fraction when the gas pressure is varied.

Note that possible undercounting may occur if the lateral spread of the beam was
larger than the silicon detector (300 mm2 area m 19 mm diameter). In the case of
the ﬁxed detector distance, we were concerned that the divergence (or straggling)
perpendicular to the beam’s path through the gas might be larger than detector size.
SRIM calculations before experiments showed the lateral straggling in the gas to be
negligible, discounting that possibility. In the case of the detector being moved along
the beam axis (the longitudinal proﬁle), we had concerns with the optical divergence
of the beam; that is, the beam’s path after the last quadrupole doublet magnet might
not be parallel in the gas cell. Measurements proved this divergence indeed had a

real effect. With the gas cell evacuated of gas, the counting rate was not constant

37

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 1__1111150Me¥[A_i§AL_1 1 _ _-
1 I
P = 0 Torr '
g 0.8 E .
I K” i
r . 1
3 06 l t
I i
'g " P = 800 Torr i
2 0.4 t I I
g P = 1200 Torr ,, a ,
0.2 ” T F .
1f
0 . . . I.’ it .
0 10 20 30 40 50
wlndow to detector distance [cm]
[ .111 1.1.1 1_ 1.. - 11- .11_
1
5 ATIMA calculations
3 08 :3}, for go=3.554 Tm
. .2— '..I
_ I -. ,
§ 0.6 “-1. lee," I
c a .,
B "--:.P = 800 Torr
g 04 En, ': u‘. .
N ‘- ‘~., ’
f 02 P = 1200 Torr "~11? Mug
0 I s .
c ‘r.-. I
o . . . . 1 .
o 10 20 30 4o 50
window to detector distance [cm]

 

 

 

Figure 2.11: Longitudinal proﬁles for stopping of 150 MeV/ A 36Ar. The upper panel
shows the raw transmitted fractions. The lower panel shows data normalized to the
proﬁle of the evacuated gas cell and ATIMA calculations for the adjusted Bp.

when the detector distance was varied (see Figure 2.11). The data given in the lower
half of that ﬁgure have been normalized to the proﬁle of the evacuated cell. Now the
calculations with an adjusted magnetic rigidity are able to reproduce the normalized
data. With this information in hand, however, the beam divergence is not expected
to be a detriment to the stopping and future extraction of ions where the divergence

is much less than the extraction electrode’s inner diameter (on the order of 5 cm).

38

Other Primary Beam Measurements

A summary of measurements made on other primary beams is given below. This
summary includes discussion of the trends observed in measuring various nuclides of
different energies.

Stopping proﬁles were measured for several beams. To maintain beam schedule
ﬂexibility, gas stopping trials were ad hoc in that measurements were made when a

primary beam happened to be on the NSCL’s running schedule. With this constraint

Table 2.4: Summary of primary beam experiments.

 

 

Beam Nominal Energy Bp Correction
[MeV/A] [T-m] to Bp [%]

”mm 150 3.649 0.14
36Ar 150 3.673 0.11
36Ar 125 3.311 0.24
40Ar 100 3.284 0.06
“Ga 110 3.714 0.25
7‘IGe 130 3.960 1.09
86Kr 140 3.342 -1.45
136Xe 120 3.972 -2.27

 

 

 

 

 

 

on beams and energies (see Table 2.4), our intent was to cover the widest possible
range of nuclide masses and beam energies in order to demonstrate the ﬂexibility of
the gas stopping system.

In addition to the beam’s Bp, Table 2.4 lists the Bp correction required to have the
ATIMA calculation agree with the experiment. The adjustments necessary for lighter
beams (Z320) were all much less than 1%, making for accurate stopping predictions.
The heavier nuclides required a wider range (i2%) of corrections.

It can be observed in several of the stopping proﬁles that the ﬂat, upper part of
the proﬁle can sometimes dip prematurely (Figure 2.12) or rise before the descending
part of the curve. We believe this is controlled by the last quadrupole before the gas
cell. After the beam passes through the glass degraders, the attenuated ions have a

greatly reduced Bp (Bpe, as it is the sixth and last section of the beamline before

39

 

, ”‘Xe 120 MeV/A

transmitted/stopped fraction

 

1.22 1.24 1.26 1.28 1.3 1.32 1.34 1.36 1.38 1.4

 

 

degrader thickness [rnrn]

 

Figure 2.12: Transmitted (solid line) and stopped fraction (dashed line) of 120 MeV/A
136Xe under three helium pressures.

entering the gas cell). In early experiments, this magnetic rigidity in the last beamline
section was not carefully set, resulting in a non-ideal (straight, parallel) beam entering
the gas cell.

Although the stopping power was underestimated by AT IMA in the case of the
heavier beam, the general shape and appearance of the stopping curve is very similar
to those of the lighter beams. The only true concern raised by this discrepancy is the
need to have a versatile primary degrading system (covering a wide range of glass
thicknesses) to account for the underestimated stopping power when using heavier

beams.

2.3.3 Conclusions for Primary Beam Measurements

The results of our several measurements on primary beams indicate a functional gas
stopping system operating with good stopping efﬁciency. In addition, predictions of
energy loss calculations seem to largely agree with our ﬁndings.

The major items of equipment to perform gas stopping of primary beams are func-
tioning as required. The variable degraders are performing as expected, although we
have seen a need to have a greater variety of glass thicknesses. The beryllium window

seems robust and appears to be acceptable. In order to prepare for ion extraction

40

tests, ring electrodes were introduced into the gas cell (more discussion of extraction
will take place in Chapter 4). At this point, the detector stack was no longer com-
patible and a single silicon detector was installed between electrode rings. This lone
detector could be moved in and out of the path of the beam; therefore, it was used
for further stopping measurements, but would also serve well later while combination
stopping/ extraction experiments were performed.

The st0pping eﬂiciencies of primary beams was greater than 50% for most ions with
a wide range of masses (A236-126) and energies (100—150 MeV/A). This efﬁciency
has been demonstrated in concert with the central apparatus in our system—the gas
cell. That is, we believe we can perform efﬁcient stopping of ions with our 50 cm gas
cell at 1 bar helium pressure.

The comparison of our data with calculations and theory was rather good, with
a minimum number of adjusted parameters. In our case, the thicknesses, densities
and compositions of all degrader materials as well as the beam’s momentum width
were held constant (at their measured values) in the calculations. Only the magnetic
rigidity was varied. For the lighter primary beams (2520), the adjustment of Bp to
ﬁnd good agreement with calculations was quite small. Even in the heavier beams,
the adjustment was within :I:2%.

The discrepancies between experiment and calculation listed for the 150 MeV/A
36Ar which required adjustments to the magnetic rigidity for the three models (ATIMA:
0.14%, Hubert: 0.20% and Ziegler: 0.24%) are not completely unexpected. Model cal-
culations have been shown to vary from experimental data by as much as 5% [37,43].
However, the sign of the variation can vary over different energy regions; that is, the
model uncertainties can have small overall corrections if the stopping power over a
wide energy range is considered. For the 36Ar data, the three corrections to Bp were
all positive, indicating an overestimation of the energy losses.

With a stronger sense of the operation and performance of the gas stopping station

on primary beams, we then turned our attention to the follow-on objective—a study

41

of the stopping of secondary beams.

2.4 Secondary (radioactive) Beam Measurements

Studying the stopping of secondary beams (that is, radioactive fragments) was one of
the intermediate objectives of the gas cell project. Although the information gained
while studying primary beams was useful, the ﬁnal objective of the Gas Stopping
Station (and LEBIT) is to stop and transport beams of rare, radioactive ions. Some
of the same characteristics which make secondary beams ill-suited for precision exper-
iments, also make them more difficult to efficiently stop. A signiﬁcant energy spread,
non-parallel beams and beam contamination are the forefront issues in dealing with

secondary beams as compared to primary beams.

2.4.1 Range Compression

Our studies of primary beams typically saw the percentage of ions stopping in one bar
of helium gas in the 50 - 70% range. As mentioned earlier, the primary beams typically
had a momentum spread of Ap/ p ~ 0.07%. However the secondary beams produced
at the NSCL have a much greater momentum spread (up to 5%). A quick calculation
examining 100 MeV / A 4°Ar (Figure 2.13) reveals that the expected stopping fraction
of ions having even a small momentum spread of 0.2% would be severely reduced.
Notice that in the more ideal case of the primary beams (upper panel), even at the
optimal degrader glass angle, the range distribution of ions stopping in the gas spans
the entire length of the gas cell, including some beam that is lost; that is, even when
we have the best stopping conditions, some of the beam hits the rear of the gas cell
and some is stopped in the beryllium window. Now, as the momentum distribution
is increased (lower panel), the situation is more severe. The stopping distribution is
so wide that nearly 30% of the beam is lost to each the rear wall and the beryllium

window, leaving just over 40% of the beam to stop in the gas. Also presented (in

42

stopped fraction

 

Degrader angle [degrees]

 

freer wallAﬂe 7 ”Be window]

 

 

     
 

I

stopped fraction

 

Degrader angle [degrees]

Figure 2.13: Comparison of calculated stopping fraction for two different beam mo-
mentum spreads (0.07% and 0.2%.) in stopping 100 MeV/A 40Ar. The solid lines are
calculated for 1 bar pressure while the dashed lines are for 1.5 bar.

dashed lines) is the predicted stopping one can expect by raising the helium gas
pressure to 1.5 bar. In this situation, the fraction of ions one can stop in the gas has
improved. However, as discussed further in chapter 3, the higher gas pressure has
other drawbacks which may preclude its use.

This reduction of stopping capability would be unacceptable when dealing with
most secondary ion beams, so a method to reduce the range of ion momenta was
needed. The idea of range compression has been studied by Weick et al. at GSI
[45,46]. They proposed that by placing a specially shaped mono-energetic degrader at
the dispersive plane following a dispersive element, the momentum spread of reaction

products could be reduced to the 0.2% level (Figure 2.14). Such a reduction would

43

 

Figure 2.14: Range compression concept showing optical elements.

bring the momentum distribution into an acceptable range for collection in a gas cell

of ﬁnite size.

Beam Impurities

Early in the discussion of secondary beam data we must assess the matter of beam
contamination. As noted in the Introduction, one disadvantage of fragmentation beam
production is beam impurities. Certainly impurities were produced and transmitted to
the gas cell along with desired radioactive species. This fact begs an obvious question
about data taking with these impure beams. Were the measured stopping proﬁles
indeed representative of the speciﬁed ion or did some miscounting occur?

The use of the stack (telescope) of silicon detectors provided rates of reaction
products, produced by interaction of the beam with degrader materials (glass and
beryllium window). Reaction products, caused by fragmentation, were typically a
few percent of the primary beam. The contribution of detector counts by reaction
products was corrected in all data presented.

Beam purities were deduced from particle identiﬁcation spectra taken at the focal
plane of the A1900 and are summarized in Table 2.5.

The separation of counts contributed by beam impurities was easily taken into ac-
count by considering the different stopping powers of the ion impurities. This process

is most clearly illuminated in the following diagram (Figure 2.15).

44

Table 2.5: Secondary beam impurities.

 

 

 

 

 

Primary beam Secondary beam Impurities
“Ar 32P 96.2% 33$ 3.4%, 31310.3%
86Kr 8°As 79.7% 81Se 21.3%, 79Ge ~0.1%
400a 380a 56.8% 37K 34.1%, 36A.» 9.1%

 

 

While a clear separation of the impurities is not evident in the typical AE - E
spectra (Figure 2.15a), a plot of AE - TOF (Figure 2.15b) provides a clear separation
of the diﬂerent nuclides. Examining any one of the curves in Figure 2.15b, one can
follow (from right to left) the ions as they ﬁrst leave only a fraction of their energy in
the silicon detector (AE) rise as their energy is attenuated more by the glass degraders.
Then as the amount of energy degrading reaches a critical value, the amount of energy
deposited in the ﬁrst detector drops as the ions are giving up their remaining energy

(total energy) to the silicon.

2.4.2 Equipment for Secondary Beam Measurements

One device was employed to reduce the momentum spread of secondary beams. Mo-
mentum slits in the dispersive plane of the A1900 Fragment Separator can be adjusted
to physically block the outlying beam thereby reducing the momentum acceptance of
the separator. While the full momentum acceptance of the A1900 is 5%, the experi-
menter can close the slits to reduce the momentum spread of the beam to 0.1%.

In order to test the proposal of Weick et al. (Figure 2.14), the beam line connecting
the A1900 to the gas cell in the N4 vault was tuned to provide a dispersive focus in
the horizontal plane at a position just downstream of the glass degraders [47]. This
tune allowed us to insert a mono-energetic wedge in the path of the secondary beam
in addition to the glass degraders (Figure 2.16). The wedge was installed on a drive
which could be remotely controlled.

The dispersion was measured with a 40Ar primary beam by inserting thin degrader

foils at the A1900 target position. The energy loss and resulting Bp were calculated

45

 

 

 

 

 

 

 

500—
CI

400— )
E 38Co
a) 37K
c300— “Ar
8 s , . s ,[
.C
gzoo~
LIJ
<1

100—

0 l I 1 1 i

1.09. stoppers 40° 56

 

AE [channels]

 

200 300
141 DE.RFTOF
cyclotron rf TOF [channels]

Figure 2.15: Particle identiﬁcation in secondary beams. Data is aggregate of several
experimental runs (for diﬂ'erent glass degrader angles) for the fragments of 140 MeV / A
4”Ca primary beam. Shown in Panel a) is PIN AE vs. TOF. Shown in Panel b) is the
ﬁrst Si detector AE vs. TOF.
with LISE. The position of the beam at the mono-energetic wedge position was ob-
served with a scintillator using a video camera. The results are shown in Table 2.6
and Figure 2.17.

The dispersion was determined to be 0.92 cm / %, in very good agreement with the
expected value of 1 cm/%

With the onset of measuring secondary beams, we had several thin, aluminum

(type 2024) wedge-shaped degraders machined by an EDM (electric discharge ma-

46

   
     
 
 

Glass Degraders
followed by
Monoenergetlc
Wedge

Quadrupole

Beam/e ‘l/1/

from
A1900

Figure 2.16: Beamline upstream of the Gas Stopping Station showing optical elements
performing range compression.

Table 2.6: Dispersion measurement data using aluminum foils with nominal 130
MeV/A 36Ar beam (Bp=3.382).

 

 

Foil thickness AE ABp Beam position Dispersion
[mg/01112] [MeV/AI [‘70] [mm] [Hun/‘76]
78 2.58 1.39 13.3 9.57
65 2.15 1.16 10.4 9.04
52 1.71 0.92 8.5 9.23
39 1.28 0.69 6.7 9.71
26 0.85 0.46 4.4 9.57
13 0.42 0.23 2.2 9.57

 

 

 

 

 

 

 

chining) technique. These approximately 2.5 x 5 cm degraders were wedged in the
horizontal direction (normal to the beam axis). The average thickness and angle on a
wedge was usually a compromise for each speciﬁc beam and energy between thinness
and ability to be precisely manufactured. The parameters of the wedges are included
in the secondary beam summary (Table 2.7, in the following subsection).

The measured proﬁles and wedge’s effect on the gas stopping measurements is

outlined in the next subsection.

2.4.3 Secondary Beam Stopping Data

Our campaign of measuring secondary beams began on 03 January 2003. Whenever

time allowed, both the primary beam and its secondary fragment were measured.

47

 

 

 

 

Dispersion Measurement Y = 9-2446x 1‘ 01165
14
"E" 12
E
-- 10
5
E 8
§ 6
§ 4
2
B
0
o 0.2 0.4 0.6 0.3 1 1.2 1.4 1.6
Delta Brho [$9

 

Figure 2.17: Results of dispersion measurements.

Table 2.7: Summary of secondary (radioactive) beam experiments.

 

 

Date Primary Beam, Secondary Beam, Wedge Parameters:
(Year = 2003) Energy [MeV / A] Energy [MeV / A] middle thickness, angle
[mm]. [mired]
03, 27Jan “Ar, 150 32p, 111 2.28, 10.8
10, 27Mar, 17Apr 86Kr, 140 80As, 93.6 0.736, 2.42
15May 400:1, 140 380a, 91.9 0.736, 2.42
0.975, 11.2

 

 

 

 

 

 

Before the data is presented, we remark on the units used in the presentation of
the data. In the instance of secondary beams, we have an aluminum wedge degrader
in addition to the glass of the variable degraders. Now, using a collection of degrader
materials, using only the units of glass degrader thickness on the x-axis of the stopping
proﬁles is inapprOpriate. Instead, we will need a more all-encompassing unit. If we
recall the Bethe speciﬁc energy loss formula (section 2.1), a more feasible unit can be

readily developed:

dE 41re‘Iq2

Focusing on the terms contributed by the absorber (degrader), the product of N
[atoms / volume] and Z [electrons / atom] gives an electron density [e‘ / volume]. Multi-
plying an electron density [6‘ /cm3] by a real thickness [cm] would give an “electron

aerial density” [e‘ /cm2] much like the mass aerial density [mg/cm2] described car-

48

lier. This new quantity, electron aerial density, like mass thickness, factors out the
absorber density, but goes one step further in factoring out the effective atomic num-
ber (or electron number) of the absorbing material. In our case where we have more
than one degrader material (glass and an aluminum mono-energetic wedge) each be-
ing a composite of many diﬂerent atoms, the electron thickness comparison becomes
a single absorber parameter which can equitably treat any combination of degrader
material compositions and thicknesses.

The parameters describing the measured secondary beams are summarized in
Table 2.8. Bpo is the magnetic rigidity of the primary beam from the K1200 cyclotron,
Bpm is for the section of the A1900 immediately following the production target,
and Bp3,4 is the rigidity value of the ions after the aluminum wedge and count rate
scintillator at Image 2 of the A1900. From these parameters, one can recreate our
experimental conditions in programs such as LISE.

Table 2.8: Secondary beam parameters. Note: a 28 mg/cm2 scintillator was used at
the same position as the wedge (not included in the wedge thickness shown here).

 

Primary Bpo Be Target BP1,2 Al Wedge BP3,4 Fragment
Beam thickness thickness
[ms/CH3] [ms/cmzl
WA: 3.670 658 3.512 300 3.321 32p
86Kr 4.499 285 3.923 300 3.461 80As
4”Ca 3.718 564 2.973 300 2.689 380a

 

 

 

 

 

 

 

 

 

 

The secondary beam data is presented below in blocks which intend to highlight

the usefulness of the methods of reducing the secondary beam momentum spread.

Use of Momentum Slits

Both Image 2 and Image 3 of the A1900 have slits which can be placed in the beam’s
path to physically limit the momentum spread of the beam. By varying the slits at
Image 3, data was taken for 0.3, 1.0 and 2.0% momentum spreads. An example of the
eﬂect of changing the momentum slits on the gas stopping proﬁle is shown in Figure

2.18.

49

 

Comparison of Incident momentum
spreads In stopping 32F
E ,,‘ ~11 EET T 010" 7

‘17' "'75 an.‘ 827 Torr I
.+. “' . ._
0.8 -

‘9." Ap/p = 0.3%

'4":
0.6 Ap/p=l.0%j ‘- , -
E .A

 

 

 

stopped

 

0 . . ' “W ‘L '1 ‘

0.67 0.08 0.69 0.7 0.71 0.72

Electron aerial density to -. mole/cm’]

 

 

 

 

 

 

Figure 2.18: Comparison of the stopping fraction for 3"’P for incident momentum
spreads of 0.3% and 1.0%.

Evident in Figure 2.18 is the improvement in the stopping fraction (difference
between gas and no gas curves). The stopping proﬁle becomes steeper at the slits are
narrowed. At an incident momentum spread of 1.0% only 35% of the beam is stopped,

while at 0.3% momentum spread, 55% of the beam can be stopped.

The Mono-energetic Wedge

Next the mono-energetic wedge, or M wedge, was paired with the proper glass de-
grader thickness and stopping proﬁles were measured. An example of data taken with
the M wedge, where the beam’s momentum spread incident on the glass and wedge
is 2.0%, is shown in Figure 2.19. Using the M wedge, again the fraction stopped in
the gas increases, proving the wedge’s ability to compress the ranges of the secondary
beam ions.

A homogeneous wedge, or H wedge, was also inserted into the beam path. This
wedge has the uniform thickness as the middle thickness of the M wedge; that is, the
H wedge is, more precisely, a ﬂat degrader. It was quickly apparent that the mono-
energetic wedge had the desired energy bunching effect. This could qualitatively be
seen by simply comparing the energy spectra for the M wedge and H wedge produced

50

 

Comparison of wedge and no wedge
In the stopping of 3“Ca

 

 

1.2

1-

 

0.8

0.. i,
0.4 .
glass only I I; If [

0.2 J
o r r r I I L

0.25 0.26 0.27 0.28 0.29 0.3 0.31 0.32
Electron aerial density [9 ' - mole/cm!)

 

 

 

 

 

transmitted/stopped fraction

 

 

 

 

 

 

 

 

 

1.2 - —- —~-~ -1-____-1.-____1- —

g 1
0 Torr

E Z: 827 Torr A
i i1 1
a 02 wedge + glass 7‘ If I
g . . . . . J. $1

0.25 0.26 0.27 0.28 0.29 0.3 0.31 0.32

 

 

Electron aerial density [9 : . molelcmzl

 

Figure 2.19: Range compression demonstrated by comparison of using a mono-
energetic wedge and glass only in the stopping of 38Ca. The momentum spread of
the beam incident on the degrader glass is 2.0%.

under similar conditions (see Figure 2.20). Now if we examine the stopping proﬁles
obtained with the two wedges (Figure 2.21), the M wedge clearly has more of an
effect on the stopping fraction—giving additional credence to the concept of range

compression using a mono-energetic wedge. The M wedge’s stopped fraction is nearly

twice the fraction of beam stopped using the H wedge.

2.4.4 Conclusions for Secondary Beam Measurements

The results of our experiments on secondary beams furthered our understanding of
st0pping fast beams. It was shown that stopping efﬁciencies could be improved with

the use of equipment to shrink the momentum spread of the secondary beams. The

51

 

32P stopping In evacuated gas cell

 

 

 

 

 

120
0.3% momentum spread
100
<— M wedge

90
E
3 60 H ed

W e
40 g \e
20
o 11 a , . 1’
o 20 4o 60 so 100 120
Energy [Channels]

140 o

120 1.0 /o momentum spread
a
c
:I
8

o 20 4o 60 80 100 120
Energy [Channels]

 

 

 

Figure 2.20: Energy spectra in the 1.5 mm Si detector for the M and H wedges with
0.3% and 1% momentum distributions. A zero degree glass angle was used for the
H wedge spectra while a glass angle of ~10 degrees was used to take the M wedge
spectra to account for the slightly different thicknesses of the wedges.

use of slits in the A1900 Ffagment Separator to reduce to beam’s incident momentum
spread provides a deﬁnite improvement on the fraction of ions stopped in the gas cell
(Figure 2.18). In addition, the mono-energetic wedge was shown to indeed compress
the range (and stopping distribution) of the secondary ions resulting in an improved
stopping fraction when compared to glass degraders alone (Figure 2.19) and compared
to using an aluminum wedge of constant thickness (Figure 2.21).

Improvements to stopping secondary beams have relevance to calculations. Com-
parison of data to calculations is complicated by uncertainties in the beam optics. Two
sources are likely contributing to shifts in the stopping proﬁles. One source occurs well
upstream of the gas cell in the A1900—uncertainty in positioning the momentum slits

in conjunction with the beam. In addition, downstream near the gas cell, positioning

52

 

Comparison of M and H wedge stopping 32P

"be ,
08 an 0 Torr

 

 

M Wedge

 

 

g r91 "m” ‘ x 827 Torr
E 0.6 I“ . "13'.
2

 

 

 

0.67 0.66 0.69 0.7 0.71 0.72
Electron aerial density [e - . mole/cm’]

 

 

 

Figure 2.21: Comparison of mono-energetic and homogeneous degraders in the stop-
ping of 32P. The curves at the bottom of the graph (stopping fractions) demonstrate
the mono-energetic wedge’s compression of the range distribution.
of the beam in the center of the mono-energetic wedge also included uncertainty.
Impurities in the beams are manageable (Figure 2.15). We have a method in place
to account for contaminants in the beam when ﬁguring gas stopping efﬁciencies. The
undesired contaminant ions will largely be removed by mass selection of the RFQ ion
guide (Appendix A) and the remainder of the LEBIT beam transport.
A summary of the stopping eﬂiciencies for secondary beams will be included in

the next section.

2.5 Gas Stopping: Summary and Conclusions

The equipment designed and constructed for our gas stOpping efforts is functional
and, save for minor improvements, will be capable of performing gas stopping for the
wide variety of beams produced at the Coupled Cyclotron Facility.

We have shown the ability to stop both stable and radioactive beams over a wide
range of masses and beam energies (summarized in Table 2.9). These efficiencies
look promising; that is, the demonstrated stopping fractions will support the overall

scheme of LEBIT by stopping a high percentage of radioactive beams.

53

Table 2.9: Summary of gas stopping experiments.

 

 

 

 

 

 

Primary beams Secondary beams
Mass range [A] 36-126 32-80
Energy range [MeV/ A] 100—150 92-111
Eﬂiciency range [%] 50-90 35-80

 

Remembering that the gas stopping is an intermediate objective on the way to
providing low-energy radioactive beams, perhaps the most signiﬁcant result of the
gas-stopping tests is the determination of the glass degrader angle which maximizes
the stopped fraction of ions. For example, the in the stopping of 38Ca (Figure 2.19),
the value along the x-axis corresponding to the maximum stopping fraction can be
translated to a degrader glass angle (approximately 24 degrees). This value will be

important in maximizing our overall efﬁciency (discussed further in Chapter 4).

54

Chapter 3

Gas Flow

3.1 Concept

The N SCL’s Gas Stopping Station will function as a high-pressure radioactive ion
source in order to provide low energy beams for precision experiments. The previous
chapter (Gas Stopping) described the process of bringing the fast radioactive ions
to rest in the Gas Cell. These ions must then be quickly and eﬂiciently extracted
from the gas cell through an oriﬁce and into an expansion chamber. This chapter
will be devoted to discussion of the gas ﬂow aspect of the extraction process and the

components of the system involved with gas ﬂow.

3.2 Design Considerations

Gas ﬂow, and the oriﬁce through which the ions would be extracted from the gas
cell was actually the earliest and most central of considerations for the gas stopping
system. Although the overall concept for a gas cell is somewhat new, the stopping of
ions in a gas-ﬁlled chamber and the later guiding of ions through a low pressure gas are
generally well understood. However, the means to perform fast and efficient extraction

of ions with gas ﬂow through an oriﬁce is less well understood and further complicated

55

if the gas stopping chamber is of large dimensions and ﬁlled with (possibly) a high

pressure gas.

3.2.1 Choosing a Gas

Radioactive ions stopped in a gas can be extracted and largely separated from the
stopping medium. In constructing such an ion source, the ﬁrst issue then becomes
which gas to use. The choice of stopping gas ﬁgures greatly into the design and
capabilities of a gas-stopping ion source.

The selection of helium as the stopping medium was arrived at quickly. Although
low mass (A24) helium has the disadvantage of lower stopping power due to low
electron density, several other properties make it quite useful, almost essential, for
the stopping of fast radioactive beams.

l. Noble gas chemistry. Helium will have few to no chemical reactions with ions
under study—no compounds are formed. Argon, with its greater st0pping power could
also be used, but it is signiﬁcantly more polarizable than helium and would lead to
adduct formation [48].

2. Availability and cost. Helium is widely available in the United States. It is
found with natural gas deposits. Only the noble gas argon is cheaper. Local prices for
Grade-A gases (99.999% purity) are typically $60 for 336 scf argon or 291 scf helium
(scf2standard cubic feet: 9514 standard liters and 8240 standard liters, respectively).

3. First ionization potential. Helium has the highest ﬁrst ionization potential of
any element, 24.6 eV. In light of this fact, the gas cell should retain any ions stopped
in the helium at a charge state of 1+.

4. Low mass. Helium’s low mass is very different from ions that will be studied,
an advantage during ion transport. After extraction from the gas cell, the radioactive
ions will undergo several meters of beam transport during which the experimenter can
apply mass selection. Even if helium could be ionized, its charge/ mass ratio would

very different from the ions under study.

56

Helium’s high ﬁrst ionization potential is the most crucial property when it comes
to efficient gas stopping and extraction. The desired charge state of 1+ for the ions
under study is necessary for efficient transport. Ions at a higher charge state will rarely
survive the subsequent ion-guiding process. Although helium’s ionization potential is
best suited for maintaining ions at a 1+ charge state, the possibility for 2+ charge
states exists. That is, elements exist which have a second ionization potential that is

lower than 24.6 eV as well:

Table 3.1: lst and 2nd Ionization Potentials of some sample elements [49].

 

 

 

Element lst IP [eV] 2nd IP [eV]
helium 24.6 54.4
carbon 11.3 24.4
silicon 8.2 16.3
germanium 7.9 15.9
tin 7.3 14.6
neodymium 5.5 10.7

 

 

 

 

Evident is the general trend of decreasing 2nd ionization potentials with increasing
atomic number. In addition, there is a chance for ions to remain triply ionized, 3“.

This is generally only true of the rare earths, including neodymium, Nd, listed above
(3rd ionization potential 22.1 eV).

3.2.2 Stapping Gas Purity

Again, examining the ﬁrst ionization potentials, there are many compounds possibly

present in the gas cell and its vacuum systems which could act as scavengers:

Table 3.2: 1st Ionization Potentials of some common compounds [49].

 

 

Compound lst IP [eV]
water 12.6
nitrogen gas 15.6
oxygen gas 12.1
pump oils ~8-11

 

 

 

 

These undesired compounds, with relatively low ﬁrst ionization potentials, could

57

reduce the eﬂiciency of the gas stopping process by neutralizing radioactive ions.
Therefore, it is important to limit the number of ions neutralized by the impurities
in the stopping gas as we are interested in studying rare radioactive beams which are
produced at low rates (as low as a few per second).

Cleanliness of components and purity of the stopping gas will contribute to a good
eﬂiciency. All components of the gas cell and its vacuum system were cleaned before
installation and later handled with latex gloves. As to the purity of the gas, special
eﬂort and extreme care were given to the selection and handling of components which

make up the gas handling system, detailed in the next sub-section.

3.2.3 The Gas Handling System

To deliver helium for stopping, a gas manifold was designed (Figure 3.1). Its function
is to deliver the ultra-pure helium, providing a constant pressure of helium in the gas

cell. The inner surface of all its 1/4 inch tubing is electropolished. In addition, the

Helium

 

 

Outlet Control Helium
® V3“ ® Iniet1
v12 p, I vs
1 . Helium
‘ ﬂ Inlet 2
Pressure ® ® ® ®
Gauge V9 V7 V5 V4
V11 V10 V8 V6 V2 V1

Vacuum

Figure 3.1: Schematic of the gas handling system that delivers ultra-pure helium to
the gas cell.

valve bodies and stems are electropolished. All the valve-to-tube connections are of
the VCR metal seal type. Great care was taken in its assembly in a clean hood to

maintain the absolutely lowest possible level of contamination.

58

The gas manifold chieﬂy has two modes of operation: helium delivery and baking.

Helium Delivery

The helium delivery mode uses the upper set of valves shown in Figure 3.1. In this
mode, the manifold delivers the ultra-pure helium required for efﬁcient stopping of
the ions. Grade-A, 99.999% purity, or better helium is introduced into the manifold
(at Helium Inlet 1 or 2) where it passes through a N anochem L—2000 puriﬁer. This
puriﬁer uses a proprietary resin in solid black head form, OMX, which contains lithium
hydride and organolithium polymer. The purity level speciﬁed by the manufacturer
for He is <1 ppb each: H20, 02, CO and C02. The puriﬁer can operate at a ﬂow rate
up to 50 slpm N 2. Our operating conditions require ~5 standard liters per minute
(slpm) He (2 slpm N2).

The delivery of this now ultra-pure helium to the gas stopping cell is regulated
by a MKS Type 205E Pressure/ Flow Controller. The controller uses feedback from
a pressure gauge (MKS Baratron Type 121A Absolute Pressure Transducer) located
near the outlet of the manifold in conjunction with a precision valve (MKS Type 148.]
Pressure and Flow Control Valve) to maintain the pressure of helium being delivered

to the gas cell.

Baking

The baking mode largely uses the lower set of valves (shown in Figure 3.1) and the
attached pump system. The vacuum is provided by a Varian Minuteman Turbopump
system. A 70 liter per second turbopump is backed by a diaphragm pump in this dry
(oil-free) system.

The manifold and delivery tubing is lined with resistive heating elements for bak-
ing. When activated, the heating elements cause contaminants to desorb from the
manifold and gas delivery tubing, while the vacuum system evacuates these contam-

inants.

59

3.3 The Gas Jet

3.3.1 Introduction and Background

Although the fundamentals of stopping ions in a gas are understood, the extraction
of these ions from a high pressure gas cell would require considerable attention. It
was clear that the ions would be passed from the stopping region of high pressure
(stagnation pressure) gas into a region of lower pressure through an oriﬁce. The use
of electric ﬁelds to maneuver the ions through the gas-stopping cell to the immediate
region of the oriﬁce will be discussed in the next chapter. The remainder of this
chapter will be devoted to the characterization of the helium gas ﬂow as it carries

ions through the oriﬁce and expands as a free jet into the region of lower pressure.

3.3.2 Gas Jet Properties

The distribution of the helium gas after the oriﬁce must be known for modeling
the capture and transport of the ions by the ion guide. Practical constraints from
hardware placed an upper limit on the length of the gas cell at ~100 centimeters
(not so important to the jet, but more a gas cell pressure versus gas cell length
consideration). In addition, the length in the ﬁrst lower pressure expansion chamber
was limited to approximately 50 centimeters.

As a result, an accurate description of the free jet was necessary to complete the
design of the gas cell-expansion chamber combination (including a skimmer) and to
optimize our ability to stop and extract short-lived radioactive ions. The gas cell
pressure would ideally be high to aid stopping, but the pressure must be low enough
such that (1) the expanding gas jet length is conﬁned to the length of the expansion
chamber and does not interact with the skimmer that follows and (2) the gas pressure
does not prevent prompt extraction of ions from the gas cell.

The oriﬁce should be of a shape such that transverse expansion of the jet is

minimized and of a size that allows for a reasonable background pressure in the

60

expansion chamber (given the available pumping speed).

Gas Handling System
0 O O 3

5 2» Puriﬁer 3
m L Booster
‘3l pump
n

”*EJAEFE

j
£4 {LE/4c <7

 

 

 

 

 

/
Ll
I

  
   

(505 Cell Inlet

 

Beam ------ * 3—

l

   
 

 

 

P F
Skimmer U! 5331
p ‘— Turbopumps -+ II)
\ g / g
\/

Figure 3.2: Schematic of the gas handling system, the gas cell and the differentially
pumped chambers of the Gas Stopping Station.

The typical throughput for the oriﬁce is approximately 10 Watts. [The throughput,
Q, is the conductance, C (ma/s), times the pressure difference, AP, (N /m2).] The
differential pumping is accomplished by Varian vacuum equipment (see Figure 3.2
and Table 3.3).

Table 3.3: Gas stopping pumps and speeds. All pumps are manufactured by Varian.

 

 

 

 

 

 

Pump Type Model Pumping Speed
Booster Roots Blower VB—5400 2550 L/s, 5400 cfm
Backing (Booster) Mechanical Piston KT-1350 368 L/s, 779 cfm
Turbopumps ’I‘urbomolecular 2000HT 2000 L/s

Backing (Turbo) 2-stage Rotary Vane DSlGOZ, DSlOOZ 1380, 900 L/m

 

 

61

3.3.3 The Nozzle

The oriﬁce separating the high pressure gas cell and the expansion chamber would take
the form of a nozzle. The nozzle shape and size expressly determine the characteristics
of the gas jet formed when the helium carrier gas expands from the high-pressure
chamber through the nozzle to a region of chamber pressure. Calculations (discussed
in the next section) and experimental measurements strongly suggested a horn-shape

nozzle (Figure 3.3).

 

 

Gas Cell

u
.°3§\

90 —> gas jet

 

 

 

Figure 3.3: Schematic of the (calculated) optimized supersonic nozzle shape.

It was soon thereafter considered a supersonic nozzle, as the typical velocities of
the gas in the expansion chamber (immediately downstream of the nozzle) would
exceed the speed of sound, C, in a gas:

0 = (gRT/MW)‘/2 (3.1)

where g is the adiabatic constant, R is the gas constant, T is the temperature and

MW is the molecular weight of the gas.

62

3.4 Gas Jet Calculations

3.4.1 Background

In addition to understanding of the shape of the expanding free gas jet, a simulation,
or preview, of the jet was required to complete the design of the gas cell-expansion
chamber combination. In this way, certain parameters of the gas stopping station
(i.e., gas cell pressure, nozzle shape and size, and geometries) could be identiﬁed to
optimize our ability to stOp and transport ions given the physical constraints of the
chambers and pumps that already made up the Gas Stopping Station.

Our understanding of the gas jet was greatly enhanced by collaboration with Dr.
Victor Varentsov from St. Petersburg, Russia. He contributed experimental knowledge
and perform detailed calculations using his own computer code to provide a thorough
description of several gas jets based on varying input parameters. Some of the results

of these calculations are described in the sub-section which immediately follows.

3.4.2 Description of the Calculations

Complete calculations describing the viscous ﬂow and subsequent free gas jet were
undertaken using the VARJ ET code [50] based on a solution of full time-dependent
system of Navier-Stokes equations for multi-component gas mixtures including device
geometry, type of gas or mixture, and background pressure and temperature. The
pressure in the gas cell is high enough that the mean free path of the atoms is small
compared to the opening in the nozzle, so that the ﬂow is viscous [51].

In short, the Navier—Stokes Equations are partial differential equations which de-
scribe the laminar ﬂow (motion) of compressible or incompressible, non-turbulent
ﬂuids. Such equations are made from a set which expresses conservation of mass,
linear momentum and energy for general motion. The conservation laws are applied
to individual cells in a grid, one by one. The entire grid is iterated until a steady

state is achieved. The calculations are performed in two dimensions, with cylindrical

63

symmetry assumed (and satisﬁed in the Gas Stopping Station) to complete the third

dimension.

3.4.3 The VARJET Code

Before calculation could begin, Dr. Varentsov carefully divided the volume under
consideration into cells (see Figure 3.4). The size of these cells was determined by the
volume under consideration. As show in Figure 3.4, the cells could be coarser in the
areas which were expected to see relatively small gradients in the gas ﬂow parameters
(pressure, density, velocity). The time steps between calculations were typically of
the order of microseconds, but depended on the size of the cell and, again, the gas

ﬂow parameters.

calculation grid nozzle region
cross section

 

 

 

 

   
    
 

second

.—. spherical
E. '3 / electrode
,_ 0.01
Nozzle
0.005

 

z [m]

Figure 3.4: On the left, an example of a grid which represents the two-dimensional
extraction geometry used for calculation under the VARJET code. The right portion
shows a cross section of the nozzle region including spherical electrodes.

The last requirement before calculation was the input of the expected experimental
parameters: stagnation (or static background) pressure in the gas cell, background

pressure (determined by the oriﬁce size and pumping speed), temperature and the

64

physical sizes of the devices under study.

The output of the calculations could be translated into jet temperatures and
velocities. However, the most important result of the calculation was the pressure
proﬁle of the gas jet; i.e., the density of the carrier gas (and therefore, a density of

the ions). An example pressure proﬁle is shown in Figure 3.5.

1 VARJET calculation

pressure [mbar]

 

 

vertical distance [mm]

    

horizontal distance [mm]

Figure 3.5: VARJ ET calculation showing pressure proﬁle of free expanding jet for 1
bar helium gas pressure in the gas cell.

The location of the Mach disk (high pressure region where the jet reconverges) at
approximately 32 mm nearly agrees with the prediction (39 mm) of Beylich’s imperial
formula [52]:

x = 2/3 - D(Po/Pa)1/2 (3.2)

where X is the distance from the nozzle to the Mach disk, D is the nozzle throat
diameter, P0 is the stagnation pressure and Pa is the ambient pressure.

From these calculations, we could focus our search for the proper parameters
(stagnation pressure, nozzle shape and size) and focus our measurements on conditions

displaying these optimal parameters.

65

3.5 Gas Jet Measurement Data

3.5.1 Introduction

A series of measurements of the pressure downstream from nozzles with two different
shapes were performed as a function of helium pressure. These measurements were
carried out with a Pitot tube mounted on a mechanical translator. The results are

described immediately hereafter.

3.5.2 The Pitot Tube

Background

A Pitot tube can measure a ﬂuid ﬂow’s static pressure (”overpressure”) by changing
the ﬂow’s kinetic energy into potential energy. The transformation takes place at the
stagnation point, at the Pitot tube entrance. The kinetic to potential energy change
results in a pressure greater than the free jet (dynamic pressure). A manometer can
then be used to measure the static pressure. A Pitot tube typically is the device
used to measure air speeds in airplanes. A simple conversion of the measured static

pressure can give the relative ﬂuid velocity.

N SCL Gas Jet Measurement Equipment

In the earliest iterations of the gas jet measurements, a simple re—entrant cylinder
was crafted to function as the gas cell. On its upstream side, it had an inlet to
allow introduction of helium from a gas bottle. Its downstream stainless steel ﬂange
contained the nozzle. The supersonic nozzle, shown Figure 3.6, was machined at the
NSCL.

A mechanical translator was designed, built and installed into Cross A (the ex-
pansion chamber) to facilitate the gas jet measurements. This device served to ﬁrmly

hold the Pitot tube and to precisely position the Pitot tube during its transverse

66

 

 

 

 

0000000000000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TUBE SIZES

$2.1m 0D x 31.6m [0
$1.6m OD x $1.1m ID
$1.1m OD x @0.8m ID

 

 

 

Figure 3.6: Nozzle and Pitot tube diagram with dimensions.

’4.

‘(C
I!

A.

i. . 1<L N 5&1} .

.L 3""; as" I

Figure 3.7: Pitot tube photograph.

and longitudinal movement through the gas jet. Assuming the circular nozzle cross
section would produce a cylindrically symmetric gas jet, the transverse movement of
the Pitot tube was limited to a vertical plane. Limiting the tube’s movement to two
dimensions also greatly reduced the complexity of the movement mechanism.

The vertical movement of the Pitot tube was driven by a linear motion feedthrough
positioned above Cross A. This arrangement allowed extremely ﬁne adjustment from
outside the chamber and knowledge of the tubes vertical position (within 0.1 mm).

The vertical range of the feedthrough was 25 mm. Therefore, the Pitot tube could

67

Linear motion feed-through
\

Pitot tube gauge
I ./

L

     

  
 
      
        

 

Gas Cell gauge

 

5%

\\\\\\\\\

\\\\\\\\\\\\\\\\\\\\\\

 

 

\

 

. .
lllllll

 

I
cg
C
3.
3
a
E.
7/%
s\\\\\\\\\

 
 
 
 
 

7///4

Figure 3.8: Layout for Pitot tube measurements.

travel about 12.5 mm above and below the center of the oriﬁce. The horizontal, or
longitudinal, movement of the tube was accomplished by manual movement of the
Pitot tube’s ﬁxing mechanism along a perforated bar. Here the adjustment could not
be made from the chambers exterior. The horizontal movement was constrained to
one cm increments and had a range along the gas jet’s axis of 15 cm.

Finally, plastic tubing transmitted the static pressure measured by the Pitot tube

to a gauge attached to the outside of Cross A.

3.5.3 Gas Jet Measurements

An MKS Baratron (capacitance manometer) transducer (range up to 100 torr) con-
trolled by an MKS 937A Multi—gauge controller determined the static pressure. An
MKS 317 Convection Enhanced Pirani gauge determined the background pressure in

the expansion chamber with the same multi-gauge controller interpreting the Pirani

68

transducer. The typical stagnation pressure in the gas cell was 1 bar (~750 torr),

while the background pressure in the expansion chamber was ~0.6 mbar (~05 torr).
An example of the data obtained during a Pitot tube measurement is summarized

in Table 3.4.

Table 3.4: Representative data taken from Pitot tube measurements. The data is

measured on the supersonic nozzle with 760 Torr (1.01 bar) helium pressure in the

gas cell. The Pitot tube is in the ﬁrst horizontal position, placing its tip 3 mm distant
from the downstream nozzle face.

 

 

Vertical Pitot tube Background
position [mm] pressure [Torr] pressure [Torr]
20.0 0.11 0.050
19.0 0.11 0.050
18.0 0.12 0.051
17.0 0.18 0.051
16.5 0.25 0.051
16.0 0.36 0.051
15.5 0.70 0.051
15.0 2.0 0.051
14.5 6.9 0.051
14.0 16 0.051
13.5 15 0.051
13.0 16 0.051
12.5 16 0.051
12.0 20 0.051
11.5 18 0.051
11.0 14 0.051
10.0 1.8 0.051
9.0 0.39 0.051
8.0 0.19 0.051

 

 

 

 

 

Approximately 20 measurements were made at each of 15 horizontal points for
ﬁve different stagnation pressures with two different nozzle conﬁgurations. In all,
about 3000 separate measurements were performed. The results from a complete

two-dimensional measurement of the nozzle at one pressure is shown in Figure 3.9.

69

Pitot tube measurement pressure [mbar]

 

so 100 150
horizontal distance [mm]

vertical distance [mm]

 

Figure 3.9: Pitot tube measurement showing pressure proﬁle of free expanding jet for
1 bar helium pressure in the gas cell. Compare with calculation made under the same
conditions (Figure 3.5).

3.6 Results and Interpretation

3.6.1 Comparison to Calculations

The measurements show good agreement with calculations (see Figure 3.10). The
features of the gas jet show excellent qualitative agreement: (1) the size (length and
width) agree, (2) the structure is quite similar: a convergence point followed by region
of higher pressure.

One comment is offered regarding the apparent slant in the jet observed in the
nozzle measurement. The results show an approximate a 3 mm vertical deviation over
a run of 100 mm (~1.7 degrees). To further explore this, the nozzle was rotated by 90
degrees and again showed a slant. This is attributable to imprecise machining; that
is, the throat of the nozzle was not exactly normal to the ﬂange surface. The ﬁxing of
the nozzle ﬂange to the gas cell was checked by micrometer to ensure that the ﬂange

was uniformly tightened to the gas cell.

3.6.2 Nozzle Gas Jet Characteristics

The general qualities of the gas jet expanding through the nozzle seem favorable. The

axial size of the jet is small; that is, the ions are largely conﬁned to the beam axis and

70

Comparison of nozzle calculation and measurement

   

calculation

 

 

 

 

_. 0
E
g
8
c: -10 - -
:3
‘0
(T3 measurement
‘E 20
9 ,
10 .l
0 so 100 150

horizontal distance [mm]

Figure 3.10: Comparison of nozzle calculation and measurement under the same con-
ditions (1 bar helium pressure in the gas cell).

the acceptance of the ion guide. In addition, the length of the jet is useable; that is,
the jets horizontal footprint will be largely conﬁned to the expansion chamber (Cross
A).

3.6.3 Simple Oriﬁce Characterization

As a check on the measurement and the calculation, we replaced the nozzle with
a simple oriﬁce (1 mm diameter) and repeated the measurements. The results are
shown in Figure 3.11. Notice that the measured pressure map is not slanted.

Again, the measurement agrees quite well with calculations. In this case the mea-
surement was performed before the calculation. After the Pitot tube measurements,

we transmitted only the measurement parameters (nozzle shape, stagnation and back-

71

Comparison of orifice calculation and measurement

‘0 calculation
I l
I ° e

-10 ’

 

 

 

 

 

 

E
g
a)
o
c
*2 pressure [mbar]
'6
6 measurement
0 20
E
>

I 10

t '
0 .
o 50 100

horizontal distance [mm]

Figure 3.11: Comparison of 1 mm oriﬁce calculation and measurement under the same
conditions (1 bar helium pressure in the gas cell).
ground pressures) to Dr. Varentsov who undertook the calculations and replied with
the diagram in Figure 3.11 (oriﬁce calculation).

The oriﬁce is clearly not suitable for our needs. To transport the ions downstream,
we will use an RFQ ion guide. This RFQ will have an inner clearance diameter of 14
mm. That is, 7 mm above and 7 mm below the nozzle opening. In the previous ﬁgure
(oriﬁce), the vertical scale is in mm. Using this ’nozzle’ shape, one can see that the
jet ’blows up’ quickly upon expanding through the oriﬁce. That is, such a jet shape
would not be compatible with our RFQ—the ions would not be efﬁciently captured

and conﬁned to the beam axis.

72

3.6.4 Gas Jet Measurement Conclusions

As a result of the calculations and experiments involving gas ﬂow, we decided upon
the optimal conditions for our gas stopping system. With some knowledge of the
extraction times offered by reasonable electric ﬁelds at given pressures, a pressure
of one bar (and length of 50 centimeters) was chosen for the gas stopping cell. The
supersonic (horn-shaped) nozzle, with a 38 degree angle on the low pressure side was

selected for the oriﬁce.

73

Chapter 4

Ion Extraction

4. 1 Introduction

Having studied the stopping of ions in helium and the ﬂow of the carrier gas through
the supersonic nozzle, we turned our attention to the extraction of ions from the gas
cell and to the transporting of these ions to LEBIT’s low energy buncher / accumulator.
The fast and efﬁcient extraction and transport of rare, short-lived ions is overall goal
for the project. This work covers experiments in extracting ions from the gas cell, but
contains no data on ion transport. Such tests are currently underway, so only a brief
description of the ion transport system will be provided for reference in this chapter

and Appendix A.

4. 1 . 1 Extraction Requirements

The ultimate goal is to measure short-lived ions that are at the limits of nuclear
stability and thus appear with short half-lives. As such, speed and efﬁciency in the
extraction and transport process are essential to facilitate the study of the most
interesting radioactive ions. Our goal is to make measurements on ions with half-lives
on the order of tens of milliseconds. A quick order of magnitude calculation shows

that for our gas-ﬂow system with 10 Watts throughput and 1 bar helium gas cell

74

pressure, the conductance through the nozzle will be ~ 10‘4m3/s. For our cylindrical
gas cell with a length of 50 cm and a diameter of 21 cm, this would result in an
evacuation time constant, T, of approximately 170 seconds with gas ﬂow alone. In
this amount of time, all the short-lived ions would be lost to decay. It is easy to see
that the extraction of ions from the gas cell is the rate-determining step in the entire
process. The time between fragment production and gas stopping is on the order of 1
us. Once ions have cleared the supersonic nozzle, the time required to transport ions
to the LEBIT low-energy experimental stations is on the order of a few ms. Clearly
then, the evacuation of ions from the gas cell requires additional means to hasten the
travel of ions to the nozzle. A system to quickly extract the ions with the greatest

efficiency would then be complete.

4.1.2 Equipment

We developed an extraction and transport scheme which would use DC potentials to
extract ions from the gas cell and a combination of DC and rf potentials in the ion
guide system to capture the ions on the beam axis and transport them downstream.
Figure 4.1 illustrates the extraction and transport concept, highlighting the pressures
and qualitative features of the electric potentials the ion will encounter along their
path. After coming to rest in the 1 bar helium of the gas cell, drift (ring) electrodes
will form a DC gradient to push ions toward the nozzle. Spherical electrodes (not
shown) inside the gas cell will focus the ions toward the nozzle with their funnel-
shaped DC ﬁeld. The ions will see a potential difference between the upstream side
and the downstream side of the nozzle, while being swept forward as part of gas jet.
The jet will be captured in the ﬁrst RFQ section (A) which is segmented to provide
both an rf ﬁeld to keep ions on axis and a DC gradient to push them downstream. In
this region (A) most of the helium will be removed by the large pumping system. The
ions then pass through a skimmer into the B and C sections of the RFQ. These RFQ

sections are also segmented, but currently are placed at the same DC potential so

75

that the ions feel only the rf ﬁeld. Finally, after leaving the Gas Stopping Station, the
ions encounter accelerating electrodes which drive them through the remainder of the

LEBIT beam transport system and towards the accumulator / buncher and Penning
Trap.

 

 

 

 

 

 

 

 

 

To the

  
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

beam , trap
—+l|||_ll||Im. —*
Pressure 1 bar ~0.1 mbar ~10“ mbar < 10‘6 mbar

Neale Skimmer Oriﬁce

.u _ f

ﬁg 0 L .
a: E _\
g 8 k x

+RF +RF +RF

Figure 4.1: The gas cell extraction scheme with pressures in the differentially pumped
system and potentials along the length of the beam axis. All three RFQ sections are
segmented, but only the ﬁrst section (Cross A) is shown this way. This is to allow a
DC gradient to be applied across its length. Currently, RFQ sections B and C have
the same DC potential down their entire lengths.

Drift Electrodes

The system of drift electrodes consists of 21 polished aluminum (type 6061) rings
and a support structure ﬁxed to the same ﬂange which holds the beryllium window
(see Figure 4.2). The rings have a 5.6 cm inner diameter, a thickness of 4 mm and a
center-to—center spacing of 2.2 cm. They run the length of the gas cell, with the ﬁrst
ring 6 cm downstream of the beryllium window and the last ring 1.5 cm from the ﬁrst
(largest) spherical electrode (Figure 4.3). The rings are ﬁxed in place by three support
rods which are attached to the window ﬂange. The entire assembly is inserted into the

gas cell from the upstream side. The support rods each consist of a threaded stainless

76

steel rod inside a Macor tube with Macor spacers between the rings. Macor is a low
outgassing, high temperature machineable glass ceramic. DC voltages are applied to
the rings to form the gradient along the length of the gas cell. These are applied to
the rings via ﬁve feedthroughs which are connected to the lst, 2nd, 6th, 16th and
2lst rings. The remainder of the rings have their voltages delivered from these ﬁve
connections by resistive chains. These chains use 1 M9 RX-IM resistors (inside the
gas cell) from Ohmite Manufacturing Company which are hermetically glass sealed

for use in ultra-high vacuum applications.

 

Beryllium window

  

-'\' "{{mrms \

 

 

silicon
. detector

Figure 4.2: Cutaway mechanical drawing (above) and photograph (below) of the drift
(or ring) electrodes. The 21 rings’ inner diameters are 5.6 cm and their center-to-
center spacing is 2.2 cm. Also shown in the photograph is the silicon detector mount
which allows the detector to be inserted into or removed from the beam axis by means
of a rotary feedthrough.

77

Spherical Electrodes and Nozzle

The drift electrodes form a ﬁeld to push the ions toward the nozzle along trajectories
parallel to the beam axis. However, in order to focus the ions toward the small throat
of the nozzle, spherical electrodes were added inside the gas cell (Figure 4.3). These
four electrodes form a funnel-shaped ﬁeld which concentrate the ions towards the
nozzle. Once the ions approach within a few millimeters of the nozzle’s throat, gas
ﬂow takes over, sweeping the ions out of the gas cell. Stefan Schwarz developed a
design of the spherical electrodes to form the appropriate ﬁeld while maximizing the
number of ions which would reach the nozzle. The electrodes are made of aluminum
(type 6061) and have Macor separators acting as insulators. The diameters of the
electrodes’ sharp tips (furthest upstream) are 45.8, 28.2 15.2 and 9.0 mm. The entire
package of electrodes, insulators and nozzle are secured, with the addition of an o-ring
seal, to the ﬂange which makes the downstream boundary of the gas cell. Biases to
the spherical electrodes are provided by feedthroughs which enter at the top of Cross
A, pass through the nozzle ﬂange, and make contact with the electrodes in the 1 bar
of helium. The design of the extraction electrode system was complicated by the fact
that the entire system must hold a pressure of greater than 1 bar of helium (in the
gas cell) with vacuum on either side. Signiﬁcant effort was necessary to design the
seals and the high voltage feedthroughs.

The nozzle was machined entirely from the semiconductor germanium (by Lattice
Materials Corporation) with the intent of placing different potentials on its upstream
and downstream side (-166 and -190 Volts, respectively). Holding a large potential
difference across the germanium nozzle proved difﬁcult, and it will soon be replaced

by a silicon version.

Simulations

In order to minimize the evacuation time and maximize the extraction efﬁciency, the

geometry and voltages of the drift and spherical electrodes required optimization.

78

 

 

 

 

 

 

Figure 4.3: Cutaway (left) and cross section (center) mechanical drawings of the spher-
ical electrodes and supersonic nozzle. For comparison, the (red, triangular) drift elec-
trode’s inner diameter is 5.6 cm. On the right is a photograph of the electrodes and
their mounting ﬂange.

Peter Schury of the LEBIT group performed SIMION (3D version 7.00) [53] beam
transport simulations with the constraint that the highest potential in the gas cell
would be approximately +2000 Volts. Spark breakdown tests showed that the high
voltage feedthroughs would tolerate, at most, roughly 2 kV in helium. It is known
that the breakdown voltage in helium gas is a factor of 5-6 times lower than that in
air.

SIMION simulations using a viscous drag force showed that, for a stopping dist-
ribution centered along the length of the gas cell at 1 bar, more than half the ions
should be extracted in less than 60 ms. The voltages applied to the electrodes for the
simulation are given in Table 4.1. Shown in Figure 4.4 are the ﬁeld lines created by the
electrode potentials and the time required for extraction depending on the position
where the ion comes to rest. It is also worth noting that the simulation includes the
dependence of extraction time on gas cell pressure (given that the electric potentials
remain constant). While higher gas cell pressures can stop a larger fraction of the in-

coming ions, more short-lived ions could be lost due to longer extraction times. Data

79

Table 4.1: Optimized voltages for the drift (ring) and spherical electrodes in the gas
cell.

 

Ring [V] Ring [V] Ring [V] Spherical [V]

 

1136 8 1075 15 915 1 565
1130 9 1055 16 885 2 420
1125 10 1045 17 855 3 225
1115 11 1015 18 815 4 -160

1 105 12 995 19 775
1095 13 975 20 670
1085 14 945 21 590

\lOzmAwNi—I

 

 

 

 

 

 

 

 

 

 

on the extraction times will be available when tests with a laser ablation system are

performed in the future (see Chapter 5).

 

 

 

beMIun drift Wei
[window electrodes eleclrodes
__—1"1'![{30"."fi']"lill['["Ix
[Iii -3 x) ‘1 I'- [ I'I“3i‘1".r’rff
400 3'..i.m.s‘ril: rrrizrrrrrrrrrnritiaf
,_,350 +1bar
£300 +1.5bar
£250 +2bar
a
:200
o
3150
3100
50
0
O 100 200 300 400 500
Position along the gas cel's axle [mm]

 

 

 

Figure 4.4: Simulation results for gas cell extraction produced by the software
SIMION. The ﬁeld lines created by the electric potentials are shown in the inset
at the top of the ﬁgure. The extraction times for various gas cell pressures is also
shown. Note that ions stopping in the ﬁrst 10 cm of the gas cell will not be extracted.

4.2 Extraction Measurements

We decided to make measurements of the extraction efﬁciency with a collection elec-
trode just downstream of the nozzle. The extraction efﬁciency is the most important

unknown quantity in our system and these tests did not require the operation of the

80

RFQ ion guide.

4.2.1 Experimental Setup

We chose to measure the extraction efﬁciency by detecting the gamma rays of ex—
tracted radioactive beam. This would provide unambiguous evidence that a speciﬁc
nuclide had exited the gas cell. Careful consideration determined which nuclide would
be measured based on the available primary beam, secondary beam purities and nu-
clide half-lives and decay energies. To perform the gamma ray detection, the layout
of Cross A of the Gas Stopping Station was altered (Figure 4.5) to include an ion

collection electrode on a 0 detector and a nearby germanium detector.

Extraction schematic [
. GOS Cell , , ion catcher
LE 3

  
  

i ,, 1m.

  

Beam [1' UUUUUUUUUUUUUUUUUUUUU»‘
.............. . [
1 000000000000000000000‘
En .
m
r7 ,7 7,- #M
Bewindow [I ‘ F
i .
. spherical
ring electrodes electrodes "

Ge detector
crystal

Figure 4.5: Schematic of the extraction test. Ions were captured on the catcher by
applying a negative potential. Upon decay, beta particles could be registered in the
plastic scintillator detector and gamma rays could be observed by the germanium
detector.

A re—entrant cylinder provided close geometry for an Ortec germanium detector

and lead shielding (to reduce the number of detected gamma rays originating from

other than the catcher—most notably the spherical electrode region). The ion catcher

81

(a 2.5 cm diameter disc), with a plastic scintillator detector behind it, was mounted
on a rod, moveable along the beam axis. The longitudinal placement of the ion catcher
was determined by the position of a high pressure in the gas jet, at approximately 9
cm, and the center of the germanium detector. In order to hold a bias, the catcher
was electrically connected to the metal mount (and rod) behind the plastic detector.
Positive and negative high voltage was applied via a feedthrough in cross A connected
to the metal rod. The concept was, for detecting positive ions, that ions (and their
decay gamma rays) should be observed for a negative bias on the catcher while few ions
should be seen for a positive bias. The plastic scintillator detector could count beta
particles and allow for beta-gamma coincidences to be observed. The detectors were
calibrated by gamma/ beta and gamma sources placed at the catcher. The standard
sources covered a range of energies from approximately 300 to 1800 keV. The plastic
detector (1 cm thick and 2.5 cm in diameter) had an efﬁciency of 0.30. The germanium
detector was an Ortec (120%), and we determined its efﬁciency as 0.007 for 1568 keV
gamma rays in its given geometry.

A slight deviation from the optimal electrode voltages was used. Because the
resistivity of the germanium nozzle could not hold the required potential difference
across its two ends, the voltage scheme required adjustment. The potentials used for

this extraction are given in Table 4.2.

Table 4.2: Voltages for the drift (ring) and spherical electrodes used in the gas cell
extraction experiments.

[V] Ring [V] Ring [V] Spherical [V]
1 135 8 1055 15 966 1 565
1 130 9 1038 16 885 2 420
1 125 10 1023 17 826 3 225
1118 11 1009 18 767 4 18
11 1 0 12 997 19 708
1095 13 985 20 649
1074 14 975 2 1 590

 

 

0Q

 

 

 

 

 

 

 

 

 

 

82

4.2.2 Extraction Data

As previously mentioned, in determining a candidate radioactive ion for extraction
tests, many factors came to bear: available primary beam, secondary beam rate and
purity, and nuclide half—life and gamma-ray energy. Table 4.3 summarizes the ions we

considered for extraction studies.

Table 4.3: Summary of extracted secondary beams.

 

 

Primary Beam, Secondary Beam, T1 /2
Energy [MeV / A] Energy [MeV/ A]
36Air, 130 32p, 111 14.3 d
86Kr, 140 80As, 93.6 15.2 8
40Ca, 140 ”Ca, 91.9 440 ms

 

 

 

 

 

Among the three nuclides we studied for extraction, the most ideal candidate was
38Ca, with its 440 ms half-life and strong gamma ray at 1568 keV (branching ratio of
21%). The short half-life reduced the contribution of build-up in the gamma spectra
and the 440 ms was more closely matched with those we expect to measure in the
future. The 1568 keV gamma ray energy was well beyond the background gamma
rays and then continuum at 511 keV from positron decay. In addition, our detection
technique continually improved with experience, so only the data on 38Ca is presented
here.

The conditions for the extraction test of 38Ca follows. Table 2.8 lists the beam
parameters for this secondary beam of approximately 92 MeV / A which was delivered
to the Gas Stopping Station. A single 1.49 mm glass degrader plate turned to 24.1
degrees (the angle determined for greatest stopping efficiency, as discussed in Chapter
2) and the M1 mono-energetic wedge (0.736 mm middle thickness and 2.4 mrad angle)
degraded most of the beam energy before the ions passed through a 1.0 mm thick
beryllium window into the gas cell containing a helium pressure of 780 Torr (1.03 bar).
The collector voltage was set, in alternating 10 minute runs, to either +350 or -350

Volts. The cyclotron beam was cycled on (3 seconds) and off (3 seconds) to reduce

83

the contribution of prompt gamma rays. The voltages for the electrodes were set as
previously described (Table 4.1) with some exceptions. Due to a problem with the
nozzle electrode system, the nozzle voltage was set to ground and the ﬁnal (smallest)
spherical electrode was set to 18 Volts. The adjusted voltages are summarized in
Table 4.2. The responses of the plastic scintillator and the germanium detector were
observed to compare the eﬂect of the diﬂerences of the positive and negative catcher
voltage. An extraction rate and eﬂiciency could then be deduced from this comparison.

Shown in Figure 4.6 are the data from the response of our two detectors. The
germanium detector shows ~20% effect between the positive and negative catcher
voltages for the 1568 keV gamma rays of 38Ca. This relatively low difference can be
attributed to detection of gamma rays emitted from nuclei not on the catcher. As
seen in Figure 4.5, the available space and geometry which only allows for limited
shielding of gamma rays coming from nuclei stOpped in the region of the nozzle and
spherical electrodes. The beta spectra show a larger effect between the positive and
negative voltage (Figure 4.6 lower panel) due to its higher efficiency and a sensitivity
to the contaminant secondary ion 37K (the only other signiﬁcant contaminant was
the stable nuclide ”Ar).

A more stringent test of ion collection is shown in Figure 4.7 where a gamma-beta
coincidence is required of the same data set as presented in the gamma ray and beta
singles spectra. The 1568 keV line is a factor of 16 larger for a negative voltage (albeit
with low statistics).

As another test of the ion extraction, the glass degrader angle was varied about
the optimal stopping angle (24.1 degrees) while the response from the beta detector
was observed (gamma-beta coincidence statistics were too low). The data from the
beta detector is shown in Figure 4.8. The beta detector was sensitive to both the 38Ca
and the 37K, so the stopping curve is similar but wider than that shown in Figure
2.19 in Chapter 2. Fewer counts were observed in the plastic as the glass degrader

angle was moved away from the optimal angle in either direction, showing that indeed

84

 

Gamma singles

catcher voltage

 

 

 

 

 

 

 

 

 

500 — -350 I
400 —+350
g 300
0 200
100
0
1500 1520 1540 1560 1580 1600
Energy [keV]
Beta spec“? catcher voltage
1000 - -- - “— ' "— " ” ' _-35ovf
800 1 —+350 V”

 

 

Counts

600

400] - --

200 \

0 50 100 150 200 250 300 350 400 450
Channels

 

 

 

0

 

 

 

 

Figure 4.6: Gamma singles (upper panel) and beta spectra (lower panel) for a pair
(-350 and +350 catcher voltages) of 10 minute runs at a ﬁxed glass angle and 780
Torr (1.04 bar) helium pressure in the extraction of 380a from the gas cell.

fewer ions were being stopped and then extracted.

4.2.3 Interpretation

The most important number derived from our extraction data is the extraction efﬁ-
ciency. We arrived at an efﬁciency by two routes: one using the beta counting data
and one using the beta gated gamma data. The errors were dominated by the un-
certainty in the detection efﬁciencies, and in the coincidence case by low statistics.
In both cases the observable from the experiment is actually the combined stopping
and extraction efﬁciency. The previously determined stopping fraction will then be
factored out to arrive at the extraction efﬁciency.

In using the beta particle data, we make the assumption that both the 38Ca and
the main radioactive contaminant, 37K had an equal chance of extraction given that

they were both stopped in the gas cell; that is, we assume that the calcium and the

85

 

Gamma-beta coincidences

istcher volts 6
—-350V
—+350V

O 500 1000 1500 2000 2500
Energy [(koVI

 

     

.
r

 

 

 

 

 

 

Figure 4.7: Gamma-beta coincidence spectrum using the same experimental data as
seen in Figure 4.6, demonstrating the effect of switching collector voltage between
-350 and +350 Volts. In this spectrum a simultaneous gamma ray in the germanium
detector and beta particle in the plastic scintillator detector is required. The ”Ca
gamma ray is at 1568 keV.

 

 

 

 

 

 

 

Beta spectrum net response for
changing glass degrader angle
_ 0.010 1 , , , - ,
2% 0.008 °
3 g 0.000 . .
8 - 0.004 . I
i g 0.002 +7 ,, . 747 ﬁ_.____ ,7 ° _1
0.000 ° . f T 3 - m
19 21 23 25 27 29
Door-d" anal-[doom]

 

 

 

Figure 4.8: Graph of beta detector net response for a changing degrader glass angle.
As the glass angle varies from the optimal angle (24.1 degrees) the beta detector
response clearly is affected, showing that fewer ions are being stopped and extracted.

potassium has an equal chance to be extracted as an atomic or molecular ion (recall
the particle identiﬁcation presented in Figure 2.15). Now, the efﬁciency for extraction,

53x“; can be obtained from:

Np = €3T0p€Exm€cha+K (4-1)

where N5 is the number of counts in the beta detector, NCa+K is the total num-

ber of beam particles delivered to the gas cell, and the remaining three terms de-

86

scribe the stopping, extraction and beta counting efﬁciencies. Using this method (with
25,171,296 38Ca and 14,791,380 37K ions counted and 43,524 beta events), a combined
stopping and extraction efﬁciency of 0.36(7)% was determined, with the error domi-
nated by the beta counting uncertainty (20%). The main error for the beta efﬁciency
comes from uncertainty in beta counting, with the beta efficiency obtained from sev-
eral sources, conservatively given as 30(5)%.

As a cross check, we also determined the extraction efﬁciency by using the gamma-

beta coincidence data, though it had low statistics. In this case:

N3, = NCa€STOP€EXTR€ﬁle5688R1568 (42)

where, in considering the number of gamma-beta coincidences, N37, the efficiency
for 1568 keV gamma rays being detected (0.007) and the branching ration (BR) to
the 1568 keV level (21%) must also be considered. Using this method, the stopping
and extraction efﬁciency of 0.29(8)% was determined with the error dominated by
contributions from the beta counting uncertainty (20%) and the low statistics of the
gamma-beta data (18%).

When the previous two values for the total efﬁciencies, 0.36(7)% and 0.29(8)% are
combined, we determine a total efﬁciency for our gas cell of 0.33(5)% (see also Table
4.4). This value obviously has potential for improvement, but is sufficient to allow the
mass measurement of many new nuclei at the NSCL [54]. Recall that the stopping

0.13

fraction (efficiency) at this degrader angle (Figure 2.19) was found to be 072(1020),

so that the extraction efﬁciency is 0.46(fgj]§)%.

Table 4.4: The extraction efﬁciency for 38Ca derived from the combined stopping-
extraction value.

 

 

 

CSTOP ° €EXTR CSTOP CEXTR

value 0.0033 0.72 0.0046
uncertainty 0.0005 +0.13 +0.0011
-0.20 -0.0015

 

 

 

 

 

87

Chapter 5

Summary and Future

5.1 Data Summary and Interpretation

5.1 .1 Efficiencies

A summary of the efﬁciencies and time required for various steps of the overall process
of mass measurements are given in Table 5.1. For various secondary beams, the ﬂight
time of the ions in each section will be approximately the same on this gross scale.
The pre-stopping efﬁciency value includes the transmission through the A1900 and
to the N4 vault. The collection and transport eﬂiciency in this pre—stopping phase is
very dependent on the secondary ion. For example, the eﬂiciency can be as high as
100% if the momentum distribution is narrower than 2%. Thus, a typical value of 1/2
is given. The stopping and extraction efﬁciencies were discussed in Chapters 2 and 4,
respectively. Simulations of the beam Optics indicate that ion losses of no more than
15% for transferring and capturing ions in the Penning 'Irap should be expected and
typical Penning Trap detector eﬂiciencies are known to be around 30% [54]. Thus,
if we combine all of these factors, we expect an overall efficiency of ~ 3 - 10‘4 from
production to detection in the Penning Trap. This number is sufﬁciently large to allow
a number of studies of exotic nuclei.

For perspective, the efficiencies of the NSCL’s gas stopping system is compared

88

Table 5.1: Summary of data and eﬂiciencies for the different systems of the Gas Stop-
ping Station and LEBIT taken for a mid-mass nuclide with half-live of 100 ms. The
mass measurement alone usually requires approximately 2-T1/2.

 

 

 

 

 

 

 

Event Efﬁciency Required time
Pre—stopping ~0.5 ~0.5 us
Stopping 0.5 - 0.7 ~0.05 [LS
Extraction 0.005 60 ms
Transport & cooling (predicted) [54] 0.85 10 ms
Mass measurement (predicted) [54] 0.3 200 ms
Overall (3.2-4.4)-10‘4 ~270 ms

 

(Table 5.2) with the most recent data from other gas stopping efforts (recall Table
1.2). The overall efficiency of the NSCL system and the RIKEN systems are similar
although the geometries and the collection electrodes are rather different. The factor
of two difference between these results is easily understandable. The efficiency of the
ANL device is signiﬁcantly higher than any other results, including all of the earlier
IGISOL studies [27,28]. The difference comes down to an extraction efﬁciency that is
about two orders of magnitude larger than other similar efforts. The relatively small
amount of information on the details of the ANL extraction studies [31] precludes a
detailed discussion of the differences.

Table 5.2: Comparison of efﬁciencies with other gas stopping efforts (mass measure-

ment efﬁciency factored out of NSCL ﬁgures from Table 5.1). The Xs mark the events
which factor into the given overall efficiency.

 

 

 

NSCL RIKEN [29] ANL [31]

92 MeV/A 38Ca 70 MeV/A 8Li ~5 MeV/A 120Cs
Event
Pre—stopping X X
Stopping X X X
Extraction X X X
Transport X X X
Overall 0.001 0.002 0.43

 

 

 

 

 

 

We believe the extraction efﬁciency of our gas cell can immediately be boosted
by an order of magnitude with improvements in the nozzle potentials. As previously

mentioned (Chapter 4), our germanium nozzle could not hold the required (optimal)

89

voltage between its upstream and downstream face. As a result, for our extraction
studies, the potential for the entire nozzle was set to zero and the voltage for the
innermost spherical electrode was greatly affected. We now have acquired a silicon
nozzle which has a greater resistivity and is expected to hold a much greater voltage
across its length. Phrther SIMION [53] calculations show that while the transport of
ions through a nozzle with zero Volts across its length is, at best, a few percent, a
nozzle which can hold 60 Volts (achievable by our silicon version) should transport

50% of the ions that it receives.

5.1.2 Mass Measurement Prognosis

Given the efficiencies that we have attained and the performance of the NSCL’s
coupled cyclotrons, an estimate of the nuclides which can be measured (and the
corresponding accuracies) can be determined. Shown in Figure 5.1 is the required
number of detected ions as a function of half-life for a variety of nuclide masses
needed to achieve a 10‘7 statistical accuracy in the mass measurement. Therefore the
required number of secondary ions produced by the coupled cyclotrons and A1900 can
be estimated by simply dividing the y-axis in Figure 5.1 by the overall efficiency of
~ 3 - 10", or multiplying by 3000.

The data taken on the Gas Stopping System to date validate the assumptions of
the ﬁrst Penning Trap measurements proposed at the NSCL. An overall efﬁciency of
10‘4 (as compared to the ~ 3 - 10““ obtained thus far) was suggested as a target to
allow measurements of several nuclides with accuracies of 10‘7 or better. When the
LEBIT system is fully functional, N~Z nuclei involved with the rp—process will be
measured: 61‘62Ga, 64Ge, 65As, 67““3""’Se, 70Br, and 72Kr. In addition nuclear structure

beyond N228 will be studied in: 40‘42P, “-443, and 43’45Cl.

90

 

 

 

Required Number or Detected ions

 

 

 

 

 

1' v' w I ’ at
0,025 0.05 00750.1 0.25 0.5 0.75 r
HalHitels]

Figure 5.1: Required number of detected ions for obtaining a statistical accuracy of
rim/m = 10”. Typically, a minimum of 100 ions are required to obtain a cyclotron
resonance curve during mass measurements in the Penning 'Ifap [54].

5.2 Future Work and Improvements

5.2.1 Improvements

The present studies were performed without the optimal voltages on the extraction
electrodes and the nozzle. Up to the present, the nozzle was left at ground potential
and the electrode potentials were scaled down to the values given in Table 4.2. We
have ordered a new nozzle made from higher resistivity silicon to allow us to reach the
design potentials. This change should provide a substantial increase in the extraction
efficiency. All of the simulations indicate that the thermalized ions will be transported
to the upstream face of the nozzle so that the electric ﬁeld at the nozzle throat is
critical for ion extraction.

A second improvement in the system would be to increase the volume of the gas
cell. In particular, an increase in the drift electrode diameter (and a corresponding
increase in the gas cell diameter) to match the 10 cm diameter of the beam line would

allow for complete coverage of the degraded beam from the range-compression ion-

91

optical system. The drift time of the ions is only slightly dependent on their radial
distribution. A larger diameter system would also allow more space for the bulky high
voltage feedthroughs.

We are encouraged that we have been able to attain signiﬁcant extraction efﬁ-
ciencies with the gas cell operating at 1 bar helium pressure. Future studies of ion
collection at even higher pressures should be carried out because at higher pressures
a larger fraction of the incident ions are stopped in the gas. The pumping system
that we have developed should be able to run at pressures up to 5 bar, but the ion
transport through the RFQ and skimmer under these conditions will require further

investigation.

5.2.2 Related Other Improvements

Two projects are underway to further the work of gas stopping and beam transport.
The ﬁrst idea involves monitoring the beam current inside the gas cell to conﬁrm the
stopping fraction inside the gas cell. A logarithmic ampliﬁer, able to accept a wide
range of negative current will be connected to the ﬁrst (furthest upstream) drift ring
in the gas cell. Ions stopping in the gas will ionize the helium as they come to rest.
While positive ions will be driven downstream by the electric potential, electrons
will travel upstream and be collected on a sparse grid strung across the ﬁrst ring.
The current on this ring is directly proportional to the rate of ions stopping in the
helium. Given that the energy and effective ionization potential for helium is known,
the ionization rate can be monitored in a non-destructive way.

In addition, tests of ion transport through the RFQ are being undertaken. The
LEBIT design included a test ion source (see Figure 5.2) which can direct a test
beam steered by a four-way quadrupole deﬂector (located on the beamline near the
ion source) either downstream or upstream. This plasma ion source is providing a
beam of argon ions which are directed upstream (backwards) through the beam optics

and through the RFQ towards the gas cell. Current at the gas cell nozzle is being

92

monitored and early results show at least 35-40% of the beam being transmitted.

5.2.3 Gamma Detection after RFQ Transport

Once the operation of the RFQ is fully tested, the original tests in which the extracted
radioactive beam was collected just downstream of the nozzle by an electrode at
negative voltage can be repeated near the end of the RFQ. A design for collection on
a wire electrode to make such a measurement has been developed. This measurement
will give us a qualitative test for ions undergoing stopping, extraction and transport
through the RFQ. The design calls for a pivotable wire that will sample ions in the
center of RFQ section C, remove them from the RFQ axis and position them in
front of a combination of plastic and germanium detectors similar to that previously

described in Chapter 4.

5.2.4 Laser Studies

A project using laser ablation to study extraction times and efficiencies is also under-
way. Instead of using beam from the cyclotrons, we will create ions in the gas cell by
using a Nd-Yag green laser pulse to ablate a target. With knowledge of the laser pulse
start time and energy, the ion start time and number can be known for a given target.
By measuring the beam current and arrival time of the extracted beam at a point

downstream of the nozzle, the extraction times and eﬂ‘iciencies can be determined.

5.2.5 LEBIT Equipment Status

After describing the future work and improvements, we would like to comment on
the progress of the LEBIT project and the gas stopping initiative. The current status
of LEBIT is shown via schematic in Figure 5.2. Previous chapters have detailed the
progress of installation and testing of hardware in the Gas Stopping Station. The last

item of equipment for collection and transport of the ions after the nozzle is the RFQ

93

that terminates in Cross C, which is described in Appendix A.

After the beam traverses the length of the RFQ, it will encounter accelerating
and steering (correction) electrodes just beyond Cross C. Next, an einzel lens will
transport the beam to the 10 degree electrostatic bender, after which the beam will
be on the proper trajectory towards the Penning Trap. A second einzel lens will
transport the beam through the shielding wall, past a diagnostic section, and then
into the accumulator/buncher. All the aforementioned equipment (between the RFQ
and the buncher) are in place and have been tested. The accumulator/buncher is
under construction and nearly complete. When operational, this section of LEBIT
will provide ion pulses to the experiments. A cooler section (not completed) will
follow and serve to improve the emittance of the beam. Finally, after passing through
a completed (but untested) beam observation and einzel lens section, the beam will
be transported to the Penning Trap for mass measurement. Although the 9.4 Tesla
solenoid is in place, the beamline bore tube for the magnet is being machined and

the Penning Trap itself is undergoing its last iteration of design.

 

[ LE BIT Room
Penning
beam observation Trap
and transport

Gas Stopping Station

   
    

_.---::". .. ’j ’ 3 9.4 T solenoid

 

accumulator/
7 ﬂ Wbuncher

 

test ion source

 

Figure 5.2: Top view of the LEBIT project layout. Various items in grey have yet
to be installed. Those systems labeled in bold print are in place but have yet to be
tested.

94

Appendix A

Beam Transport

A radiofrequency quadrupole (RFQ) ion guide will be used to transport ions from
the nozzle and gas jet to the LEBIT beam transport section. Such a device will
complement our differentially pumped system; that is, ions captured by the RFQ
would be conﬁned to the beam axis while the neutral helium carrier gas will be easily

pumped away by the vacuum systems.

A.1 Background

A radiofrequency quadrupole (RFQ) is a device which focuses beams of charged par-
ticles onto its central axis. The RFQ consists of four rods. The rods opposite each
other form a pair which are connected to be at the same electrical potential. These
two pairs are then placed at equal but opposite potentials driven by a radiofrequency
ampliﬁer (Figure A.1). The electric ﬁeld that an RFQ produces has the net effect
of conﬁning ions to the longitudinal axis. In this way, ions introduced to one end of
the RFQ will be efﬁciently guided or transported along its length. RFQs have been
developed for a number of applications including ion guiding, mass spectrometry and

buffer gas ion cooling.

95

 

Figure A.1: Simple RFQ depiction. Alternating current is applied to the rods, with
opposite rods connected so that they are at the same potential. Ions passing through
the RFQ’s length experience a conﬁning ﬁeld which keep them on the beam axis.

The electric potential, <1), and the electric ﬁeld, E, of the RFQ are given by:

x2

 

(p:

‘2 ’2 [U - Vcos(f2t)] (A.1)
r0

—x 2
E —— ( y )g‘U — Vcos(f2t)] (A.2)

where the spatial coordinates x and y are perpendicular to the beam axis (z), r0 is
the characteristic nearest half-distance between opposite rods, U is the applied DC
voltage, V is the zero to peak radiofrequency voltage, and Q/21r is the radiofrequency
[55]. If we consider the effective potential, D, for a radiofrequency voltage only (where
U=0) [56]:

e 1 2 e V2 r2

D=smi °=EZﬁ§f§

(A.3)

As an example for mass = 100u, V = 50 Volts, Q/er = 1 MHz and r0: 7 mm,
the potential, D, at r = r0 would be approximately 1 eV. From Equation A.3, one
can see that by choosing the appropriate rf voltage, the RFQ can also provide mass

selection.

96

t = 0 t = T/2 effective potential

         
    
 

Ht

i’No

.\ \‘t 3 0' "4 I I

. \&&t:€i?n7l,vlll,’//‘

< 6301\1‘3‘35'ﬁ'f‘75’3!”

Hi). it. 2.37874]
1.0.”,

 
    

\ u“

. v .

\ \ \ "

esters
WWW“
"\\\\\‘\‘

     

-
'. O.
- - ~

”o...
0.-....-
... O. -
........

      
 

X

Figure A.2: Snapshots of the RFQ potential for two instances. The sum of the poten-
tials (shown at time zero and at time equal to one half the period) gives an effective
ﬁeld which constrains ions on both the x and y axes.

A.2 The Gas Stopping Station RFQ

The Gas Stopping Station’s RFQ was based on a similar ion guide design used at
CERN for ISOLTRAP [57]. Each pole has a diameter of 18 mm and the characteristic
radius, r0, is 7 mm. The total length of each pole is approximately 145 cm, made up

of short individual segments (see Figure A.3).

rf circuits and power

 

 

 

 

 

 

 

feedthroughs
it: “”3
(E I
l......... ..... I ..... . 3 HLLLI-I—‘L—rm
)3 II
r—‘i r...

.[7 H
Gas Cell l l “j n

 

>

Skimmer Oriﬁce

Figure A.3: Side view of the Gas Cell and the Gas Stopping Station’s RFQ. The
power feedthroughs and circuits sit atop the crosses for each section of the RFQ.

Each of the four poles is assembled by adding stainless steel segments separated

by ceramic spacers to a central 6 mm stainless steel rod (threaded at each end). A

97

stainless steel nut at each of the rod compresses the segments and spacers. Segments
are 10 mm, 20.5 mm and 41.5 mm in length and the ceramic spacers provide a 0.5 mm
separation between segments. The number and length of the segments are chosen to ﬁll
the available spaces in Crosses A, B and C. A support system of aluminum octagons
and ceramic stand-offs provides rigidity and orientation to the RFQ. Two large (~6
inch diameter) ceramic discs (Figure A.4) form the boundaries between Crosses (A / B
and B / C) and are incorporated into the length of the RFQ poles. This ﬁrst boundary
(A/B) disc also holds the skimmer. The skimmer has an oriﬁce of 3 mm diameter,
chosen to allow maximum passage of ions while allowing the desired pressures to be
achieved by the differential pumping. The second boundary (B / C) disc contains an

opening (oriﬁce) which is simply the clearance inside the RFQ poles.

 

Cross B

ceramic
deers

 

 

 

Figure A.4: Mechanical drawing (above) and photo (below) of the RF Q and its sup-
port structure, showing the different sections of the ion guide (only a portion of the
RFQ in Cross C is displayed). The copper bands connect opposite rods, placing them
at the same potential. The yardstick accompanying the photo shows the scale.

98

Direct current is provided by bipolar power supplies capable of delivering :I:400
Volts and the radqiofrequency is provided by a square wave driver / ampliﬁer built
by Peter Schury. The DC and the rf potentials are fed to resonant circuits which
rest outside Crosses A, B and C. Multi-pin power feedthroughs carry these voltages
inside the vacuum system. Cabling from these feedthroughs to the RFQ segments uses
KAP2 from Caburn MDC—Limited, a silver plated c0pper conductor coated by Type
F Kapton. We chose to take advantage of the segmentation in Cross A so that a DC
gradient could be placed on RF Q section A, helping to drive the ions downstream.
Though RFQ sections B and C are also segmented, we currently have all segments at

the same potential. The voltage scheme is:

Table A.1: Planned DC voltages for the RFQ segments of Cross A, based on the
voltages placed across the nozzle (with the ﬁrst segment placed at the same potential
as the downstream face of the nozzle). The remainder of the segments in Crosses B
and C will be at the same voltage as the skimmer.

Segment 1 2 3 4 5 6 7 8 skimmer
Voltage 0 -5 -15 -25 -35 -45 -55 -56 -60

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Initial trials with the RFQ show that it transmits 35—40% of the beam. Because
of available resources, these trials were performed by running the beam backward
through the RFQ from an ion source upstream near the accumulator/buncher. When
the beam travels in the expected direction (with the skimmer oriented properly), this

percentage is expected to rise, conservatively, to 50%.

99

Bibliography

[1] E. M. Burbidge, G. R. Burbidge, W. A. Fowler, and F. Hoyle. Rev. Mod. Phys.,
29:547, 1957.

[2] A. G. W. Cameron. PASP, 69:201, 1957.
[3] R. K. Wallace and S. E. Woosley. ApJS, 45:389, 1981.
[4] T. Glasmacher. Annu. Rev. Nucl. Part. Sci, 48:1, 1998.
[5] P. F. Mantica et al. Phys. Rev. C, 67:014311, 2003.
[6] Yu. Kudryavtsev. Nucl. Instr. and Meth. B, 179:412, 2001.
[7] D. Lunney and G. Bollen. Hyperﬁne Int, 129:249, 2000.
[8] R. F. Casten and B. M. Sherrill. Prog. Part. Nucl. Phys, 45:8171, 2000.
[9] C. J. Barton et al. Phys. Rev. C, 67:034310, 2003.
[10] G. F. Lima et al. Phys. Rev. C, 65:044618, 2002.
[11] A. S. Lalleman. Hyperﬁne Int, 132:315, 2001.
[12] F. Herfurth et al. Eur. Phys. J. A, 15:17, 2002.
[13] J. P. Dufour et al. Nucl. Instr. and Meth. A, 248:267, 1986.
[14] R. Anne et al. Nucl. Instr. and Meth. A, 257:215, 1987.
[15] K. H. Schmidt et al. Nucl. Instr. and Meth. A, 260:287, 1987.
[16] H. Geissel et al. Nucl. Instr. and Meth. A, 282:247, 1989.
[17] D. J. Morrissey and B. M. Sherrill. Phil. Trans. R. Soc. London, 356:1985, 1998.
[18] A. C. Mueller and R. Anne. Nucl. Instr. and Meth. B, 56/57z559, 1991.

[19] B. M. Sherrill, D. J. Morrissey, J. A. Nolen, N. Orr, and J. A. Winger. Nucl.
Instr. and Meth. B, 70:298, 1992.

[20] H. Geissel et al. Nucl. Instr. and Meth. B, 70:286, 1992.

100

[21] T. Kubo et al. Proc. First Intl. Conf. on Radioactive Nuclear Beams, Berkeley,
W. D. Meyers, J .M. Nitschke, and E. Norman, eds., World Scientiﬁc, Singapore,
p. 563, 1990.

[22] D. J. Morrissey, B. M. Sherrill, M. Steiner, A. Stolz, and I. Wiedenhover. Nucl.
Instr. and Meth. B, 204:90, 2003.

[23] S. Schwarz, G. Bollen, D. Lawton, P. Lofy, D. J. Morrissey, J. Ottarson, R. Ringle,
P. Schury, T. Sun, V. Varentsov, and L. Weissman. Nucl. Instr. and Meth. B,
204:507, 2003.

[24] H. Geissel et al. Nucl. Instr. and Meth. B, 195:3, 2002.

[25] J. Arje and K. Valli. Nucl. Instr. and Meth. A, 179:533, 1981.
[26] J. Arje et al. Nucl. Instr. and Meth. A, 247:431, 1986.

[27] P. Dendooven. Nucl. Instr. and Meth. B, 126:182, 1997.

[28] J. Aysto. Nucl. Phys. A, 693:477, 2001.

[29] M. Wada et al. Nucl. Instr. and Meth. B, 204:570, 2003.

[30] G. Savard et al. Nucl. Phys. A, 701:292c, 2002.

[31] G. Savard et al. Nucl. Instr. and Meth. B, 204:582, 2003.

[32] J. Dilling et al. Hyperﬁne Int, 127 (l /4):491, 2000.

[33] M. Wada et al. In Proceedings of Non-Neutral Plasma Physics IV, vol. 606, page
625. American Institute of Physics, 2002.

[34] G. F. Knoll. Radiation Detection and Measurement. John Wiley & Sons, Inc.,
New York, 3rd edition, 2000.

[35] F. Hubert, R. Bimbot, and H. Gauvin. At. Data and Nucl. Data Tables, 46:1,
1990.

[36] J. F. Ziegler and J .B. Biersack. computer program SRIM 2000/01/03. Available
from <http://www.srim.org>.

[37] J. F. Ziegler, J. P. Biersack, and U. Littmark. The Stopping and Range of Ions
in Solids, volume 1. Pergamon, New York, 1985.

[38] C. Scheidenberger and H. Geissel. Nucl. Instr. and Meth. B, 135:25, 1998.
[39] J. Lindhard and A. H. Sorensen. Nucl. Instr. and Meth. B, 135:25, 1998.

[40] D. Bazin, O. Tarasov, M Lewitowicz, and O. Sorlin. Nucl. Instr. and Meth. A,
482:314, 2002.

[41] O. B. Tarasov and D. Bazin. Nucl. Instr. and Meth. B, 204:174, 2003.

101

[42] T. Baumann et al. Nucl. Phys. A, 701:282c, 2002.

[43] H. Paul and A. Schinner. Nucl. Instr. and Meth. B, 179:299, 2001.
[44] A. Stolz, 2003. Private communication.

[45] H. Weick et al. Nucl. Instr. and Meth. B, 164-165zl68, 2000.

[46] C. Scheidenberger et al. Nucl. Instr. and Meth. B, 204:119, 2003.
[47] J. Stetson, 2003. Private communication.

[48] Yu. Kudryavtsev. Rev. Sci. Instr., 69:738, 1998.

[49] D. R. Lide (ed.). Handbook of Chemistry and Physics. CRC Press, Washington
D.C., 2002.

[50] V. L. Varentsov and A. A. Ignatiev. Nucl. Instr. and Meth. A, 413:447, 1998.

[51] A. Roth. Vacuum Technology, page 69. Elsevier, The Netherlands, 3rd edition,
1996.

[52] A.E. Beylich. Z. ﬂugwiss. Weltraumforsch. 3, Heft 1, 1979.
[53] LLC Bechtel BWXT Idaho. SIMION (3D version 7 .00) software, 1999.

[54] G. Bollen. ”LEBIT—Pilot Mass Measurement Program,” NSCL proposal for
experiment, 2002.

[55] D. Gerlach. Ion-Molecule Reaction Dynamics, Part 1, Advances in Chemical
Physics Series, volume 87. John Wiley & Sons, Inc., New York, 1992.

[56] S. Schwarz, 2003. Private communication.

[57] F. Herfurth et al. Nucl. Instr. and Meth. A, 469:254, 2001.

102

 

    

   

        
  

G

[[1]]

3

SI

’]
H
:[l
n

9

.
[Will]:
3 024

[1

i]
[
1