NEUTRON SPECTROSCOPY OF 22 N AND THE
DISAPPEARANCE OF THE N = 14 SHELL
By
Michael J. Strongman

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Physics
2011

ABSTRACT
NEUTRON SPECTROSCOPY OF 22 N AND THE
DISAPPEARANCE OF THE N = 14 SHELL
By
Michael J. Strongman
The emergence of new “magic” numbers in the light neutron-rich oxygen isotopes has led to
interest in spectroscopy of other neutron-rich elements like nitrogen. One such nucleus is the
N = 15 isotope 22 N. An excited state of 22 N unbound with respect to neutron emission was
observed in a stripping reaction from a 85 MeV/u 26 F beam. The observed decay energy of
650(50) keV places the level, which is interpreted to be the first 3− state, at an excitation
energy of 1.93(22) MeV. Together with the previously measured bound states of 22 N, a
reduction of the N = 14 shell gap compared to less neutron-rich nitrogen isotopes at the
neutron dripline is observed. Based on the magnitude of the reduction of the shell gap for
22 N, a disappearance of the gap and even a level inversion of the ν1s
1/2 and the ν0d5/2
levels in the neutron-unbound nucleus 21 C seems likely.

Dedicated to my brothers, Kevin and Kyle. I wish both of you the fullest success in your
endeavors.

iii

ACKNOWLEDGMENT

Any good Master’s Thesis begins with an advisor, and I had the good fortune to get one
of the best advisors around, Michael Thoennessen. Thanks for taking a chance on me and
encouraging me to see this to the end.
Thank you to my co-advisor, Artemis Spyrou, who has been an excellent guide through
this process. I’d also like to thank the third member of my thesis committee, Filomena
Nunes, who taught me that theory is always important, even for experimentalists.
I would like to thank the members of the MoNA collaboration. Without them, there
would be no MoNA, no experiments, no one to explain how anything works, no captive
audience when I’m trying to explain my work, and no one to take on the all important
graveyard shifts. Within the group, I’d like to specifically acknowledge the contributions
of Warren Rogers, Thomas Baumann, and Paul DeYoung. They managed to teach me
practically everything I know about nuclear instrumentation, as well as the art of figuring
out seemingly unsolvable problems. Warren was also successful in getting me into nuclear
physics in the first place, an achievement that I am immensely grateful for. Calem Hoffman
deserves special mention because he lent me the data for thesis experiment, and allowed me
to use his calibrations and simulations, as well as answering my questions about analysis.
My work would not have been completed without the help of my fellow graduate students. Shea Mosby and Greg Christian patiently answered my numerous questions about
programming and ROOT, even questions I had asked many times already. Special thanks to
Michelle Mosby for being a great colleague as well as an excellent baker; your goodies helped
fuel the late night work during our experiments. Jenna Smith and Jesse Snyder have been
iv

fabulous co-workers and colleagues and I ask them to forgive me for everything I haven’t explained about the data acquisition. Phil Voss, Rhiannon Meharchand, Krista Meierbachtol,
and Greg Pang deserve special mention for being terrific peers in graduate work, and have
all given me good advice and knowledge to navigate the process of defending my thesis.
I would like to extend a special thanks to the families at University Reformed Church
that adopted me as their own: Judy and Gerry Gothro, Steve and Marilyn Herwaldt, Tim
and Sue Herwaldt, Lee and Joella Cogan, Bruce and Jan Jeffries, Billy and Margie Yang,
Dave and Amity Hinkly, and Peeter and Carolyn Lukas. Without you, my mother’s fears
that I would starve probably would have come true, and your willingness to take me into
your family gatherings for the holidays helped stem the tide in the battle of homesickness.
My thanks to my friends in Michigan, Steph, Laura, Brad, Ben, and Eric, who have been
supportive in their prayers, and helped out in the tough parts by making me laugh.
I’ve really appreciated the support of Megan Phillips, Hilary Hayes, and Melissa McLeod.
Megan edited the many drafts of my paper, and provide a safe-haven in Santa Barbara when
I wanted to see the ocean. Hilary has been a devoted listener and caring friend. And a
long distance thanks goes to Melissa, for continuously updating me on her life in China and
challenging me to get my degree done.
Of course the final acknowledgement goes to my parents, Ken and Melissa. Dad, finishing
your law degree and passing the bar exam has been a great source of inspiration for me. Mom,
you have always been my most tireless supporter, my initial cheerleader, and my best critic.
Thanks for listening on the phone when I was discouraged, and for doing every chore that I
should have been doing while I was at home these past few years.

v

TABLE OF CONTENTS

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
List of Figures

viii

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

1 Description of the Problem and Theoretical Foundation
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2 The Nuclear Shell Model . . . . . . . . . . . . . . . . . . .
1.2.1 History of the first “magic” numbers . . . . . . . .
1.2.2 Calculations with the Shell Model . . . . . . . . . .
1.3 Evolution of Magic Numbers in neutron-rich nuclei . . . .
1.4 Extracting shell gaps from single particle states . . . . . .
1.5 Experimental Studies of N = 14 Shell Gap . . . . . . . . .
1.6 Previous Measurements of 22 N . . . . . . . . . . . . . . .

ix

.
.
.
.
.
.
.
.

1
1
2
5
6
8
11
14
16

2 Experimental set-up
2.1 The Sweeper Focal Plane Detectors . . . . . . . . . . . . . . . . . . . . . . .
2.2 The Modular Neutron Array (MoNA) . . . . . . . . . . . . . . . . . . . . . .
2.3 Calibrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

18
19
20
21

3 Analysis of experiment
3.1 Incoming beam identification .
3.2 Element Identification . . . .
3.3 Isotopic Selection . . . . . . .
3.4 Neutron Coincidences . . . . .
3.5 Decay reconstruction . . . . .
3.6 Monte-Carlo Simulations . . .
3.6.1 Fitting to data . . . .

.
.
.
.
.
.
.

24
24
25
27
35
35
37
39

4 Results and Discussion
4.1 Shell model calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2 Disappearance of the N = 14 shell . . . . . . . . . . . . . . . . . . . . . . . .

42
44
46

5 Conclusion and Future prospectives

49

A New Experiment Setup Guide
A.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.2 The Purpose of Experimental Accounts . . . . . . . . . . . . . . . . . . . . .

52
52
52

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

vi

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

A.3 Acquiring the Experimental Account . . .
A.3.1 Passwords . . . . . . . . . . . . . .
A.3.2 Logbooks . . . . . . . . . . . . . .
A.3.3 SSH . . . . . . . . . . . . . . . . .
A.3.4 Windows Password Change . . . .
A.3.5 Connecting to the outside internet
A.4 BASH files . . . . . . . . . . . . . . . . . .
A.4.1 event space . . . . . . . . . . . . .
A.5 Getting the code from monasoft . . . . . .
A.5.1 Readout . . . . . . . . . . . . . . .
A.5.1.1 The experiment directory
A.5.2 Online Analysis Code . . . . . . . .
A.5.3 XLM . . . . . . . . . . . . . . . . .
A.5.4 n2analysis . . . . . . . . . . . . . .
A.5.5 evtMerge . . . . . . . . . . . . . .
A.5.6 Using a script to automate the rest
A.5.7 the current directory . . . . . . . .
A.6 Setting up Readout . . . . . . . . . . . . .
A.6.1 MoNA-LISA Settings files . . . . .
A.6.2 sweeper readout settings files . . .
A.7 Online Analysis Codes . . . . . . . . . . .
A.7.1 ROOTBeer & ROOT . . . . . . . .
A.7.2 SpecTcL . . . . . . . . . . . . . . .
A.8 Utility Codes . . . . . . . . . . . . . . . .
A.8.1 High Voltage Controller . . . . . .
A.8.1.1 Emergency HV control . .
A.8.2 HV fitting . . . . . . . . . . . . . .
A.8.3 NSCLDumper . . . . . . . . . . . .
A.9 Running the experiment . . . . . . . . . .
A.9.1 Readout crashing . . . . . . . . . .
A.9.2 Recovering from a reboot . . . . . .
A.10 Post-experiment . . . . . . . . . . . . . . .
A.10.1 Calibration runs . . . . . . . . . . .
A.10.2 Archive tapes . . . . . . . . . . . .
A.10.3 Analysis Space . . . . . . . . . . .
A.10.4 Executables . . . . . . . . . . . . .
A.10.5 Managing the changes to the major

. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
programs

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

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

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

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

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

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

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

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

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

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

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

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

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii

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

53
54
54
54
56
56
57
58
58
59
60
60
60
61
61
62
63
63
63
63
64
64
65
65
65
67
67
68
68
68
70
71
71
71
73
74
75
77

LIST OF TABLES

1.1

The Hamiltonians used to calculate the level energies in this thesis. . . . . .

7

3.1

The velocities and time-of-flight for the two major components of the incoming
beam to travel between the A1900 focal plane to the target. . . . . . . . . .

24

Table listing the parameters of the corrected time-of-flight from Ref. [12] . .

29

A.1 The spdaq machines and corresponding detector systems and readout codes.

59

A.2 The different high voltage modules and their names and IP adresses. . . . . .

67

3.2

viii

LIST OF FIGURES

1.1

1.2

1.3

1.4

The chart of the nuclides for Z ≤ 13. The red lines denote the traditional
“magic” numbers for this region. (For interpretation of the references to color
in this and all other figures, the reader is referred to the electronic version of
this thesis.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2

A close up of the region in the nuclides containing the “island of inversion”,
highlighted by the blue outline. Isotopes in this region do not have the expected properties of “magic” nuclei for the N = 20 closure. . . . . . . . . . .

3

A close up of the neutron rich region of interest within the chart of the nuclides.
The red lines mark the traditional shell closures in this region: N = 8, N = 20
and Z = 2 and Z = 8. The red dashed lines mark the “new” magic numbers
(N = 14 and N = 16) which emerge in the neutron rich isotopes of this region.

4

The evolution of the sd shell gaps of oxygen isotopes as function of neutron
number using the USD interaction from Ref. [36]. . . . . . . . . . . . . . . .

9

1.5

The evolution of the sd shell gaps of carbon isotopes as a function of neutron
number using the WBT interaction from Ref. [35]. Note the inversion of the
ν1s1/2 and ν0d5/2 shells as the number of neutrons are reduced from 14 to 8. 10

1.6

The shell configuration for the ground state of 22 O (left) and the first excited
states of 22 O (right). Closed circles represent nucleons occupying the orbital
while open circles are positions where a nucleon could occupy. The gaps
between shells are not to scale. . . . . . . . . . . . . . . . . . . . . . . . . . .

12

The “competition” of the 1s1/2 –0p1/2 shells in N = 7 isotones from Ref. [31].
These are the calculated shell gaps for 6 ≥ Z ≥ 4. . . . . . . . . . . . . . . .

13

An inset of the region of nuclei around the N = 14 shell closure. The isotope
of interest in this thesis (22 N) is highlighted in red. . . . . . . . . . . . . . .

14

1.7

1.8

ix

The shell configuration for the ground state of 23 O (left) and the first excited
state of 23 O (right). Closed circles represent nucleons occupying the orbital
while open circles are positions where a nucleon could occupy. The gaps
between shells are not to scale. . . . . . . . . . . . . . . . . . . . . . . . . . .

15

1.10 The shell configuration for the ground state of 22 N (left) and the first excited
states of 22 N (right). Closed circles represent nucleons occupying the orbital
while open circles are positions where a nucleon could occupy. The gaps
between the shells are not to scale. . . . . . . . . . . . . . . . . . . . . . . .

16

1.11 The measured level scheme of 22 N from Ref. [27]. The 1− and 2− are the first
two excited states predicted by the shell model so the 183 KeV and 1017 keV
state are assigned these spin-parity respectively. . . . . . . . . . . . . . . . .

17

1.9

2.1

2.2

3.1

3.2

3.3

3.4

The experimental setup for the MoNA-Sweeper for experiment 05039. MoNA
is the neutron detector array in green (right) and Sweeper dipole magnet is
in blue, located after the reaction target. Figure taken from Ref. [12]. . . . .

19

The Modular Neutron Array (MoNA), a neutron detector array of 144 plastic
scintillating detectors. Figure provided courtesy T. Baumann. . . . . . . . .

22

Energy loss versus time-of-flight for the beam exiting the A1900 focal plane
showing the separated 26 F and 29 Na. . . . . . . . . . . . . . . . . . . . . . .

25

The time-of-flight for beam particles traveling from the end of the A1900 focal
plane and the target position. The intense peak at 297 ns is the 26 F. . . . .

26

Element identification gated on incoming 26 F via energy loss versus total
kinetic energy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27

Element identification gated on incoming 26 F via energy loss versus energy
loss (two different detector measurements). . . . . . . . . . . . . . . . . . . .

28

3.5

A diagram showing how isotope separation works. Beam fragments leave the
target position with the same energy. Heavier isotopes (blue) are bent less
than the sweeper magnet and thus end up with a positive angle and position
in the focal plane. Lighter isotopes (green) and bent more by the sweeper
magnet and thus end up with negative angles and positions at the focal plane. 29

3.6

A corrected time-of-flight spectrum of oxygen fragments. The peaks correspond to individually separated isotopes (from shortest time-of-flight to
longest) of 22 O, 23 O, and 24 O. . . . . . . . . . . . . . . . . . . . . . . . . .
x

30

3.7

A corrected time-of-flight spectrum of nitrogen isotopes. The peaks lie in the
same time of flight of the oxygen isotopes identified in Fig. 3.6, so the peaks
from left to right are identified as 19 N, 20 N, and 21 N respectively. . . . . . .

31

The angular correlation of oxygen isotopes to time-of-flight. The lighter isotopes are bent more by the sweeper magnet, so the isotopes of oxygen are
identified (from left to right) as 22 O, 23 O, and 24 O. . . . . . . . . . . . . .

32

The angular correlation of nitrogen isotopes to time-of-flight. The lighter
isotopes are bent more by the sweeper magnet, so the isotopes of oxygen are
identified (from left to right) as 19 N, 20 N, and 21 N. . . . . . . . . . . . . . .

33

3.10 The total neutron time-of-flight spectrum. The peak at 68 ns corresponds to
prompt neutrons from reactions on top of random background. . . . . . . . .

34

3.11 The progression of neutron coincidences as more restrictive element and isotope gates are applied. The black (top) spectrum is the total neutron spectrum
from 3.10. The red (middle) spectrum is gated on nitrogen events, and the
blue (bottom) spectrum is gated on 21 N events. This progression shows that
the neutrons accompanying 21 N are likely to be “decay” neutrons from the
break-up of 22 N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

36

3.12 The data points indicate a narrow width Breit-Wgner resonance from a STmona simulation and the solid blue curve indicates the same Breit-Wigner
resonance without any acceptances. The width of the decays were 6 keV with
a energy of 650 keV. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

39

3.13 A contour plot of the χ2 dependence of the Breit-Wigner parameters for the
measured decay energy of 21 N + n. The parameters are the decay energy
(x-axis) and the width of the decay (y-axis) in MeV. The lowest χ2 is located
at 0.65 MeV and 0 keV. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

40

3.14 Neutron-decay energy spectrum of 22 N. The data with statistical error bars
were measured by requiring a coincidence between an identified 21 N fragment
and a neutron. The solid line is the sum of the resonant (dotted) and nonresonant (dashed) contributions of a Monte-Carlo simulation that takes into
account the resolutions and acceptances of the experimental apparatus. The
resonance line-shape is a Breit-Wigner with Edecay = 0.65(5) MeV. . . . .

41

The proposed level scheme of 22 N → 21 N + n decay created from a reaction
of 26 F. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

43

3.8

3.9

4.1

xi

4.2

4.3

The ground state shell configuration of 26 F. Note the lone neutron in the
ν0d3/2 shell. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

44

Level scheme of 22 N. The present experimental data are shown together with
the data of Reference [27] and are compared to WBTM and WBPM shell
model calculations. The neutron separation energy as well as the ground
and first excited state of 21 N are also shown. The shaded area around the
21 N ground state represents the uncertainty due to the neutron separation
energy. The shaded area around the measured 3− state corresponds to the
excitation energy uncertainty which is dominated by the uncertainty of the
neutron separation energy. . . . . . . . . . . . . . . . . . . . . . . . . . . .

45

4.4

The “competition” between the ν1s1/2 and ν0d5/2 levels for the N = 15 isotones of carbon, nitrogen, and oxygen following the prescription of Ref. [31].
Experimental data for the levels of 23 O and 22 N are taken from Refs. [26],[30],[27],
and this work. It should be noted that the states labeled ν1s1/2 are actually
ν0d5/2 -hole states, but these states measure the ν1s1/2 level because a neutron is required to move from the 0d5/2 to the 1s1/2 . The blue box around
21 C indicates the range of expected gaps for the ground state based on the
error bars of the 22 N and 23 O measurements. . . . . . . . . . . . . . . . . . 48

5.1

The “competition” between the ν0d5/2 and ν1s1/2 levels for the N = 14 isotones of carbon, nitrogen, and oxygen following the prescription of Ref. [31].
Experimental data for the levels of 22 O, 21 N, and 20 C are taken from Refs. [30],[29],
and [27]. Higher excited states in 20 C, like the next expected ν1s1/2 -particle
state with J π = 3+ , have not been observed. . . . . . . . . . . . . . . . . . . 50

5.2

The shell configuration of an inverted ground state of 21 C. . . . . . . . . . .

xii

51

Chapter 1
Description of the Problem and
Theoretical Foundation

1.1

Introduction

The earliest studies of the structure of nuclei have revealed that for certain “magic” numbers
of neutrons and protons, nuclei exhibit more stable behavior than neighboring nuclei, as
shown in Fig. 1.1. These magic numbers are 2, 8, 20, 28, 50, 82, and 126. This understanding
has explained, for example, the preponderance of stable isotopes at N = 50 (86 Kr, 87 Rb,
88 Sr, 89 Y, 90 Zr, and 92 Mo) and the spherical nature of such doubly magic nuclei as 48 Ca
and 16 O.
These magic numbers were assumed to be the same for all nuclei, even as the limits of
stability were approached. For example, all N = 20 nuclei were assumed to all have large
gaps between the highest occupied and lowest unoccupied orbital. Experiments in the late
1970’s [32, 37] cast doubt on whether these gaps are universal throughout the chart of the
1

Al

Ne
Z=8

O

N = 20
N = 28
Z=2

He
N=8

Figure 1.1: The chart of the nuclides for Z ≤ 13. The red lines denote the traditional
“magic” numbers for this region. (For interpretation of the references to color in this and
all other figures, the reader is referred to the electronic version of this thesis.)
nuclides by showing that 31 Na and 32 Na do not have the expected mass excess of a magic
nucleus. This so-called “island of inversion” [34], shown in Fig. 1.2, is the first of several
“islands of shell breaking” [6] where the traditional understanding of the shell model breaks
down and new, unexpected properties of nuclei emerge. In respect to this thesis, the light
neutron rich isotopes of oxygen, nitrogen, and carbon, shown in Fig. 1.3, exhibit these new
properties. To investigate the shell model in this region, information on the excited levels of
isotopes need to be measured.

1.2

The Nuclear Shell Model

The main theoretical model that has been utilized to understand the structure of nuclei and
explain the emergence of magic numbers has been the Shell Model. Since describing the
nucleus with the motion of each nucleon is a many body problem and thus non-solvable,
approximations have to be made. The underlying idea behind the Shell Model is that the
2

Al 26 Al 27 Al 28 Al 29 Al 30 Al 31 Al 32 Al 33 Al 34 Al 35 Al 26 Al 37 Al 38 Al 39 Al 40 Al 41 Al 42 Al 43 Al

Mg25 Mg26 Mg27 Mg28 Mg29 Mg30 Mg31 Mg32 Mg33 Mg34 Mg35 Mg36 Mg37 Mg38 Mg

Na 24 Na 25 Na 26 Na 27 Na 28 Na 29 Na 30 Na 31 Na 32 Na 33 Na 34 Na 35 Na

40 Mg

37 Na

Ne 23 Ne 24 Ne 25 Ne 26 Ne 27 Ne 28 Ne 29 Ne 30 Ne 31 Ne 32 Ne 33 Ne 34 Ne

F 22 F 23 F 24 F 25 F 26 F 27 F

29 F

31 F

O 21 O 22 O 23 O 24 O
Z=8

N 20 N 21 N 22 N 23 N

C 19 C 20 C

B

N = 14

19 B

22 C

N = 20

N = 16
Figure 1.2: A close up of the region in the nuclides containing the “island of inversion”,
highlighted by the blue outline. Isotopes in this region do not have the expected properties
of “magic” nuclei for the N = 20 closure.
interactions of a single nucleon with the rest of the individual nucleons inside the nucleus
can be approximated by an overall mean field [16]. Thus, the energy levels of the nuclei can
be calculated using a single central mean potential, just like in atomic physics for electrons.
Within the Shell Model, the orbitals in which nucleons are occupied are labeled in the
similar manner as for atomic orbitals, according to the quantum numbers n, , and J. n
is the principal quantum number.1

is orbital angular momentum. J is total angular

momentum calculated by adding the intrinsic spin of the nucleon to . Therefore, J is either
1
+ 2 or − 1 . Shorthand for labeling the orbitals by quantum number goes as nLJ , where
2
n is the principal quantum number, L is a letter that corresponds to ,2 and J is the total
angular momentum. The occupation number of a particular orbital can be calculated from
1 In this description we choose to start the n at 0.
2 s: = 0, p: = 1, d: = 2, and f : = 3
3

22 Al 23 Al 24 Al 25 Al 26 Al 27 Al 28 Al 29 Al 30 Al 31 Al 32 Al 33 Al 34 Al 35 Al 26 Al

19 Mg20 Mg21 Mg22 Mg23 Mg24 Mg25 Mg26 Mg27 Mg28 Mg29 Mg30 Mg31 Mg32 Mg33 Mg34 Mg35 Mg3

19 Na 20 Na 21 Na 22 Na 23 Na 24 Na 25 Na 26 Na 27 Na 28 Na 29 Na 30 Na 31 Na 32 Na 33 Na 34 Na 3

16 Ne 17 Ne 18 Ne 19 Ne 20 Ne 21 Ne 22 Ne 23 Ne 24 Ne 25 Ne 26 Ne 27 Ne 28 Ne 29 Ne 30 Ne 31 Ne 32 Ne 33 Ne 3
15 F 16 F 17 F 18 F 19 F 20 F 21 F 22 F 23 F 24 F 25 F 26 F 27 F

29 F

12 O 13 O 14 O 15 O 16 O 17 O 18 O 19 O 20 O 21 O 22 O 23 O 24 O
Z=8
10 N 11 N 12 N 13 N 14 N 15 N 16 N 17 N 18 N 19 N 20 N 21 N 22 N 23 N
C

9 C 10 C 11 C 12 C 13 C 14 C 15 C 16 C 17 C 18 C 19 C 20 C

B

8B

9 B 10 B 11 B 12 B 13 B 14 B 15 B

Be 7 Be 8 Be 9 Be 10 Be 11 Be 12 Be

Li 6 Li 7 Li 8 Li 9 Li

He 5 He 6 He

14 Be

17 B

19 B

22 C

N = 20
N = 16

N = 14

11 Li

8 He
Z=2

H

4H

5H
N=8
Figure 1.3: A close up of the neutron rich region of interest within the chart of the nuclides.
The red lines mark the traditional shell closures in this region: N = 8, N = 20 and Z = 2
and Z = 8. The red dashed lines mark the “new” magic numbers (N = 14 and N = 16)
which emerge in the neutron rich isotopes of this region.

4

31 F

2J + 1. Therefore, an orbital with n = 0,

= 2 and J = 5/2 is labeled as 0d5/2 and has an

occupation number of 2 × 5/2 + 1 = 6.
Each orbital has an Effective (spherical) Single Particle Energy (ESPE or SPE). This is
a measurement of the mean effect of other nucleons on the single particle orbit. Two body
matrix elements (TBME) depend on J, T coupled by j1 , j2 interacting nucleons and are
used to calculate the interaction of a nucleon of J = j1 on a nucleon of J = j2 . These two
physical parameters are taken empirically from fits to experimental data.

1.2.1

History of the first “magic” numbers

It was first observed in the late 1940’s that nuclei with neutron or proton numbers of 20,
50, 82, or 126 are particularly stable. This behavior wasn’t explained by the traditional
Wood-Saxon Potential:

V (r) =

V0
1 + exp (r − R)/a

(1.1)

In 1949, a solution was proposed by Mayer [19, 20], Haxel, Jensen, and Suess [11] which
involved adding a correction to the mean field: the spin-orbit force. The spin-orbit force
splits orbits with the same angular momentum ( ) into j-lower (j< ) and j-upper (j> ) total
angular momenta.
The spin orbit potential is:
Vso = −
where

1 dV (r)
·s
r dr

(1.2)

is the orbital angular momentum and s is the intrinsic spin angular momentum of

the nucleon.
5

If we apply the · s term of spin-orbit to a state with total angular momentum of + 1/2:

< ψj= +1/2 | − · s|ψj= +1/2 >= −
2

(1.3)

+1
< ψj= −1/2 | − · s|ψj= −1/2 >= +
2

(1.4)

and − 1/2:

We see that the spin-orbit force splits the degeneracy of levels corresponding to the same
. When added to the traditional mean field of a Wood-Saxon potential, we noticethat the
magic numbers observed emerge. This explains the magic numbers above N = 20.

1.2.2

Calculations with the Shell Model

Because using the whole model space of all possible quantum numbers makes calculations
of energy states cost prohibitive, physicists typically truncate the model space to fit the
particular nucleus that is being calculated. A model space corresponds to all the combined
shells that nucleons are allowed to occupy for a given calculation. For example, calculations
of the neutron rich oxygen isotopes are typically truncated to the sd shell (0d5/2 ,1s1/2 , and
0d3/2 ) and assume an 16 O core (that is, it assumes that the nucleons that fill the 0s and
0p are effectively inert and the contributions of orbitals above the 1s and 0d are negligible).
Each subset of the model space has empirically defined Hamiltonians consisting of different
ESPEs (Effective Single-Particle Energies) and TBMEs (Two-body Matrix Elements).
For neutron rich isotopes with Z ≥ 8, the USD interaction [36] is used to calculate
excitations within the sd only. Excitations to higher orbitals, like the pf shell, are ignored.
This uses a model space of exclusively the sd shell assuming an 16 O core. The interaction
6

Hamiltonian
USD
WBT
WBP

Model Space
sd with 16 O core
spsdpf with no core
spsdpf with no core

Source
[36]
[35]
[35]

Table 1.1: The Hamiltonians used to calculate the level energies in this thesis.

is defined by 63 TBMEs and 3 single particle energies (SPE) for each orbit in the sd space
(0d5/2 ,1s1/2 , and 0d3/2 ). This Hamiltonian dates from 1983 and has been the dominant
interaction used for nuclei with masses A = 18 to A = 38. The TBME and SPE for the
USD are obtained from a least square fit of 447 binding energies and excitation energies
of sd shell nuclei [35]. For the sd shell the TBMEs are scaled as follows: T BM E A =
A 0.3 [35].
T BM E A=18 × 16

In order to calculate carbon and nitrogen nuclei, the WBT and WBP interactions are
needed because a 16 O core is unreasonable and interactions with the π0p1/2 shell are necessary for the protons. These Hamiltonians are active in a full spsdpf model space (which
will include the TBMEs of the USD interaction). The difference between the two interactions is that they use different choices of least-squares fitting of the experimental energies for
the SPE and TBME. The WBT uses a fitting of linear combinations of the effective TBMEs
with some parameters normalized by the G-matrix method [35]. The WBP uses the modified
surface one boson exchange method to achieve similar fits of TBMEs [35]. Both interactions
produce similar results for stable nuclei, and for reasons described in Sec. 4.1, both will be
used.
7

1.3

Evolution of Magic Numbers in neutron-rich nuclei

New magic numbers and shell gaps have arisen for neutron rich nuclei, in particular for the
oxygen isotopes. The first evidence that this might be the case came from investigation
of the systematics of the neutron separation energies (Sn ) as a function of isospin [12].
Experiments investigating the region near N = 16 have confirmed the emergence of a new
magic number. Evidence includes the lack of bound excited states for 23 O and 24 O [29], the
high energy excited states populating the ν0d3/2 shell in 23 O [9], the spherical content of
the 24 O ground state [15], the high first and second unbound excited states of 24 O [14], and
the unbound state energy of 25 O [13]. Furthermore, since the neutron drip line has been
confirmed for Z < 9, the fact that three neighboring elements (carbon, nitrogen, oxygen)
have the same neutron number for the last bound nucleus [22], dramatically points to a solid
shell closure of N = 16.
Another new magic number is N = 14 which was first observed in 22 O. The properties
of 22 O that was the evidence for this included the high excitation energy of the first excited
2+ state [33, 3, 29] and the low B(E2) value of the 2+ state [33].
The N = 14 shell gap develops in the oxygen isotopes due to the attractive monopole
nn
matrix element Vd
as neutrons fill the ν0d5/2 orbit, increasing the binding relative
5/2 d5/2
to the ν1s1/2 orbit [28]. This is exhibited in Fig. 1.4 by the gradual decline in ESPE of
nn
the ν0d5/2 shell from N = 8 to N = 14. Also the Vs
is slightly repulsive, adding
1/2 d3/2
to the effect. However, as neutrons fill the s orbit, between N = 14 and N = 16, a similar
nn
attractive force (Vs
) reduces the ESPE of the ν1s1/2 orbit.
1/2 s1/2
The gap between the ν1s1/2 and ν0d5/2 is predicted to be largest for 22 O and will decrease with increasing distance from 22 O. In the carbon isotopic chain, reducing the number
8

Oxygen Isotopes

0d5/2
1s1/2
0d5/2

2

ESPE (MeV)

0

-2

-4

-6

-8
6

8

10

12
14
16
Neutron Number

18

20

Figure 1.4: The evolution of the sd shell gaps of oxygen isotopes as function of neutron
number using the USD interaction from Ref. [36].

9

Carbon Isotopes

0d5/2
1s1/1
0d5/2

ESPE (MeV)

4

2

0

-2

-4
6

8

10

12
14
16
Neutron Number

18

20

Figure 1.5: The evolution of the sd shell gaps of carbon isotopes as a function of neutron
number using the WBT interaction from Ref. [35]. Note the inversion of the ν1s1/2 and
ν0d5/2 shells as the number of neutrons are reduced from 14 to 8.

10

of neutrons from 20 C ultimately leads to the inversion of the ν1s1/2 and the ν0d5/2 levels
as shown in Fig. 1.5. A similar inversion is possible as neutrons are added to 20 C.
These effects happen for the oxygen isotopes but not for higher Z isotopes because the
Vπj>νj< (the potential between a proton and neutron with the same

but opposite spins)

is attractive so as protons fill the π0d5/2 , the orbit just above ν0d5/2 (the ν1s1/2 ) goes
down in energy.
The question remains as to if the N = 14 shell gap persists for Z < 8. We know from the
level inversion at 15 C [31] that the ν0d5/2 shell starts above the ν1s1/2 for N = 8 as seen
in Fig. 1.5. The same attractive matrix element for ν0d5/2 acts in the carbon isotopes as
in the oxygen isotopes, so the ν0d5/2 is expected to drop in energy, however, the question
remains where the ν0d5/2 will lie in energy relative to the ν1s1/2 .

1.4

Extracting shell gaps from single particle states

The evolution of a shell gap with a neutron number N can be extracted from single particle
or single hole levels in the N − 1 or N + 1 nuclei. The difference in energy between two
orbitals can be determined via the prescription of Lawson and Uretsky [17]. Their center
of gravity theorem displayed in Eqn. 1.5 shows that the weighted average of excited states
that correspond to a configuration reasonably estimates the energy between two orbitals.

E= J

(2J + 1)EJ
2J + 1

(1.5)

J
An example of this would be measuring the gap between the ν1s1/2 and ν0d5/2 in 22 O.
The ground state of the 22 O is a filled ν0d5/2 shell as shown in Fig. 1.6. The first two
11

ν1s1/2

ν1s1/2

ν0d5/2

π0p1/2

π0p1/2

ν0p1/2

π0p3/2

π0p3/2

ν0p3/2

π0s1/2

ν0s1/2

protons

ν0p1/2
ν0p3/2

π0s1/2

neutrons

protons

ν0s1/2
neutrons

22 O∗

22 O

Figure 1.6: The shell configuration for the ground state of 22 O (left) and the first excited
states of 22 O (right). Closed circles represent nucleons occupying the orbital while open
circles are positions where a nucleon could occupy. The gaps between shells are not to scale.

excited states were identified (with guidance of the shell model) to be 2+ at 3.199 MeV and
3+ at 4.582 MeV [29]. Performing a weighted average as follows:

E=

(2 × 2 + 1)3.199 MeV + (2 × 3 + 1)4.582 MeV
= 4.2 MeV
(2 × 2 + 1) + (2 × 3 + 1)

(1.6)

This result agrees well with the predicted gap of 4.3 MeV from the USD [8, 29]. More
complicated configurations, where both the lower shell and higher shell have multiple J
combinations, simply require adding all the relevant states.
This approach for estimating shell gaps from weighted averages of excited states was used
by Talmi and Unna [31] to predict the inversion of the ν1s1/2 shell in 11 Be. Fig. 1.7 shows
the linear extrapolation of the gaps in 13 C and 12 Be to the gap in 11 Be.
12

Figure 1.7: The “competition” of the 1s1/2 –0p1/2 shells in N = 7 isotones from Ref. [31].
These are the calculated shell gaps for 6 ≥ Z ≥ 4.

13

19 Na 20 Na 21 Na 22 Na 23 Na 24 Na 25 Na 26 Na 27 Na 28 Na 29 Na 30 Na 31 Na 32 Na 33 Na 34 Na 3

e 17 Ne 18 Ne 19 Ne 20 Ne 21 Ne 22 Ne 23 Ne 24 Ne 25 Ne 26 Ne 27 Ne 28 Ne 29 Ne 30 Ne 31 Ne 32 Ne 33 Ne 3
16 F

17 F

18 F

19 F

20 F

21 F

22 F

23 F

24 F

25 F

15 O

16 O

17 O

18 O

19 O

20 O

21 O

22 O

23 O

24 O

14 N

15 N

16 N

17 N

18 N

19 N

20 N

21 N

22 N

23 N

13 C

14 C

15 C

16 C

17 C

18 C

19 C

20 C

12 B

13 B

14 B

15 B

17 B

26 F

22 C

27 F

29 F

N = 20

19 B

11 Be 12 Be Figure 1.8: Be inset of the region of nuclei around the N = 14 shell closure. The isotope of
13 Be 14 An
interest in this thesis (22 N) is highlighted in red.
10 Li 11 Li
9 He

1.5

10 He N = 8

Experimental Studies of N = 14 Shell Gap
Z=2

The size of the gap between ν1s1/2 and ν0d5/2 orbitals has been determined for the N = 14
nucleus 22 O [29]. Stanoiu et. al. determined that the gap is 4.2 MeV from measurements of
the 2+ and 3+ excited states of 22 O. The Z = 7 isotone of 22 O, 21 N, was measured using
gamma ray spectroscopy. Because the proton shell is not closed for nitrogen isotopes, more
states are needed to calculate the shell gap. Sohler et. al. measured and identified the four
excited states corresponding to the same orbits as the 2+ and 3+ in 22 O. Using the same
prescription as Lawson and Uretsky [17], Sohler et. al. [27] found an estimated shell gap of
3.03 MeV, a one MeV reduction of N = 14 shell gap, and concluded that there is a partial
survival of the magicness of N = 14 for the nitrogen isotopes.
The size of the gap between ν1s1/2 and ν0d5/2 orbitals has been determined for the
N = 15 nucleus 23 O [26]. The spin and parity of the ground state of 23 O is 1/2+ (a ν1s1/2
particle state) and thus the measurement of the 5/2+ excited state (a ν0d5/2 hole state)
at 2.79(13) MeV directly determines the size of the gap between the ν1s1/2 and ν0d5/2
14

31 F

ν1s1/2

ν1s1/2

ν0d5/2

π0p1/2
π0p3/2

ν0p1/2

π0p3/2

ν0p3/2

π0s1/2
protons

π0p1/2

ν0p1/2

ν0s1/2

ν0p3/2

π0s1/2

neutrons

protons

ν0s1/2
neutrons

23 O∗

23 O

Figure 1.9: The shell configuration for the ground state of 23 O (left) and the first excited
state of 23 O (right). Closed circles represent nucleons occupying the orbital while open
circles are positions where a nucleon could occupy. The gaps between shells are not to scale.

orbitals as shown in Fig. 1.9.
To perform the same analysis for the N = 15 nucleus 22 N, shown in relation to the
isotopes mentioned above in Fig. 1.8, more states are needed. Because the last proton shell
for nitrogen isotopes is not closed, the ν1s1/2 particle and ν0d5/2 hole states are coupled
to a π0p1/2 state (shown in Fig. 1.10). The ground state of 22 N is assigned a spin and
parity of 0− and the first excited state is 1− . Both correspond to the ν1s1/2 particle states,
coupled to π0p3/2 shell for protons. The states that correspond to the ν0d5/2 hole state
are 2− and 3− .3 Measurements of all three excited states (1− , 2− , and 3− ) are necessary
to determine the 1s1/2 –0d5/2 shell gap and study the shell evolution of N = 15 nuclei.

3 Guided by the shell model
15

ν1s1/2

ν1s1/2

ν0d5/2

π0p1/2
π0p3/2

ν0p1/2

π0p3/2

ν0p3/2

π0s1/2
protons

π0p1/2

ν0p1/2

ν0s1/2

ν0p3/2

π0s1/2

neutrons

protons

ν0s1/2
neutrons

22 N∗

22 N

Figure 1.10: The shell configuration for the ground state of 22 N (left) and the first excited
states of 22 N (right). Closed circles represent nucleons occupying the orbital while open
circles are positions where a nucleon could occupy. The gaps between the shells are not to
scale.

1.6

Previous Measurements of

22

N

A neutron knockout experiment of 22 N was carried out at the Fragment Separator at GSI
Darmstadt. Rodriguez et. al. [24] measured the momentum distributions of neutron knockouts from several neutron-rich nuclei, including 16−22 N, 19−23 O and 21−26 F. Of interest
here are their results of 22 N. The experimenters deduced from the larger cross section and
small width of the longitudial-momentum for 22 N that there is a change from a 0d5/2 to
1s1/2 valance neutron. This is an effect already observed for 23 O and 24 F, and expected
for isotopes above N = 14. This is consistent to the shell model (see Fig 1.4 and Fig 1.5)
that the ν1s1/2 is above the ν0d5/2 . Therefore, the assignment of the ground state of 22 N
to a spin-parity of 0− is justified and higher excited states are expected from the spinflip
pair of 0− (1− ) and ν0d5/2 -hole states (2− and 3− ) respectively.
16

Figure 1.11: The measured level scheme of 22 N from Ref. [27]. The 1− and 2− are the
first two excited states predicted by the shell model so the 183 KeV and 1017 keV state are
assigned these spin-parity respectively.
Solher et al. [27] performed gamma ray spectroscopy of the neutron-rich nitrogen isotopes
including 22 N at GANIL. Since the gamma rays of 183 keV and 834 keV of energy were in
coincidence,they attributed this to a gamma cascade from a 1.017 MeV state to a 183 keV
state. With guidance from the shell model, the first and second excited states were identified
as 1− and 2− respectively. Figure 1.11 shows the level scheme of 22 N. Since Sohler did not
observe any other gamma rays, any higher excited states are assumed to be neutron unbound
and higher than the neutron separation energy of 1.28(22) MeV. Therefore, it is necessary
to perform neutron spectroscopy experiments to measure any higher lying states.

17

Chapter 2
Experimental set-up
Neutron spectroscopy experiments involve the reconstruction of a neutron unbound state
from its decay products. The unbound states are populated from various reactions, such
as one-proton knockout and multiple-neutron evaporation from a rare isotope beam. One
neutron spectroscopy experiment was performed in the summer of 2005 involving a 26 F rare
isotope beam, which was designed to measure the neutron unbound ground state of 25 O,
the results of which have already been published in [13] and [12]. Decays of 22 N were also
observed in this experiment.
The experiment was carried out with the MoNA-Sweeper set-up in the N4/N6 vault
at the National Superconducting Cyclotron Laboratory (NSCL). A primary beam of 48 Ca
accelerated to 140 MeV/nucleon impinged on a 987 mg/cm2 Beryllium target. Isotopic
selection of 26 F was achieved with the A1900 fragment separator [21]. The beam energy was
85 MeV/nucleon and contained other constituents including a factor of 100 more 29 Na. A
5 mm thick plastic timing scintillator was located at the end of the A1900 fragment separator
to allow the constituents of the beam to be separated by time-of-flight.
18

Figure 2.1: The experimental setup for the MoNA-Sweeper for experiment 05039. MoNA is
the neutron detector array in green (right) and Sweeper dipole magnet is in blue, located
after the reaction target. Figure taken from Ref. [12].
The beamline between the A1900 and the target position had two position sensitive
parallel-plate avalanche chambers (PPACs), located 3.97 m upstream of the target, as well
as a focusing triplet magnet. The PPACs measured the spatial characteristics of the incoming
beam. 6.25 cm before the target position was a 0.254 mm plastic timing scintillator that
provided timing information of the beam. The target consisted of 470 mg/cm2 of Beryllium.

2.1

The Sweeper Focal Plane Detectors

After the neutron unbound nucleus decays, the energy and momenta of the neutron and the
leftover fragment are measured, because they are required in order to reconstruct the decay
energy.
The expelled charged fragments from the reaction are sent through a large gap dipole
magnet (Sweeper) [4] that deflected these fragments 43◦ into the focal plane charged-particle
detectors. Any neutrons produced in the reaction are not deflected, pass through the large
19

gap in the dipole, and are detected by the Modular Neutron Array (MoNA) [2] at zero
degrees 8.2 meters downstream of the target position. Figure 2.1 shows the positions of the
Sweeper and MoNA relative the target positions.
After the sweeper magnet, there were two cathode-readout drift chambers (CRDCs) to
detect the position of charge fragments leaving the focal plane. These CRDCs, located
1.88 m apart, had 128 pads in the dispersive direction, where a Gaussian position of the
charge distribution determines the position. The position in the non-dispersive direction is
determined by the time taken to drift to the anode.
Downstream the CRDCs, there is an ion chamber to detect the energy loss of charged
fragments. Behind the ion chamber were two plastic scintillators 0.5 cm and 15 cm thick
respectively. Each scintillator has two photo-multiplier tubes (PMTs) attached above and
below the device. The energy signals of all four PMTs are gain-matched and geometrically
averaged to determine the energy of the signal. The thinner scintillator was the trigger for the
entire system and served as a secondary measurement of the energy loss of the fragments,
here to after known as the dE scintillator. The average of the timing signals for the dE
scintillator also served as the absolute timing of fragments exiting the focal plane. The other
scintillator measure the total kinetic energy (TKE) of the fragments.

2.2

The Modular Neutron Array (MoNA)

The Modular Neutron Array, shown in Fig. 2.2, consists of 144 10 × 10 × 200 cm3 BC-408
plastic scintillator bars [2]. They are arranged in a 16 × 9 configuration 8.20(5) m from the
target and aligned at 0◦ relative to the beam direction. The size and position of the array
allow for an angular coverage of neutrons from ±7.0◦ in the horizontal and ±5.6◦ in the
20

vertical.
Each bar has photomultiplier tubes (PMTs) attached to each end with an accompanying
light guide. Neutrons interact with the particles in the plastic (hydrogen or carbon nuclei)
and those charged particles become excited and emit photons of light to de-excite. These
photons are collected by the PMTs at the ends of each bar which determine the time and
charge collected. The time difference between PMT signals at each end is used to calculate
the location of the event in the horizontal position within a bar. Vertical position and
distance from the target of the interaction is calculated via the position of the bar within
the array that detects the event with an uncertainty of ±5.0 m. The time measurement of
the event was calculated from the mean time of the PMTs of the bar with the particular hit.

2.3

Calibrations

It should be noted that this experiment was designed to investigate the ground state decay
of 25 O. The principal investigator for this experiment, Calem Hoffman, performed all the
necessary detector calibrations for this experiment.
Calibrations for the Modular Neutron Array were performed by using data collected from
cosmic ray muons. Since the fast moving muons scatter off the plastic inside the MoNA bars
in the same manner as a neutron, depositing the average electron equivalent of 22 MeV
energy in each bar, cosmic data can be used to perform high voltage gain matching for
each PMT and energy calibration for each charge-to-digital (QDC) module that interprets
the raw PMT charge collected. Time information from the PMTs is collected by time-todigital (TDC) modules which are calibrated by a pulser device that gives a series of pulses
at a known interval. Converting the time difference into a position on the bar requires a
21

Figure 2.2: The Modular Neutron Array (MoNA), a neutron detector array of 144 plastic
scintillating detectors. Figure provided courtesy T. Baumann.

22

calibration of positions. Because cosmic muons illuminate the bars uniformly, analysis of the
time difference spectra for each bar during a cosmic run yields a slope and offset to convert
time to a position on the bar.
Calibrations for the charged particle detectors in the beamline and the sweeper focal
plane involve gain matching the response of PMTs relative to each other, correcting for the
position dependence of the energy deposition on the thin and thick detectors, and doing
relative energy calibrations. The position sensitive CRDCs and PPACs are calibrated by
placing masks of holes in known absolute positions in front of them as the dispersed beam is
sent through the focal plane. Identifying the holes in the position spectra allows for absolute
position determination.
A detailed description of these calibrations can be found in Reference [12].

23

Chapter 3
Analysis of experiment

3.1

Incoming beam identification

The first step of the analysis is to identify the incoming beam. As the incoming beam was
not pure but composed of 26 F and 29 Na as well as some low Z constituents, it is necessary
to tag events with the proper incoming beam. The magnetic rigidity of the beam (26 F
and other constituents) at the end of the A1900 spectrometer was 3.9301 Tm. Because the
components of the beam had the same rigidity and different masses, they have different
velocities as shown in Table 3.1.
The beam passes through 35.68 m of beam-line between the A1900 scintillator and the
timing scintillator, so the 26 F is separated by time-of-flight from the 29 Na. No target runs
Isotope
26 F
29 Na

Velocity (cm/ns)

time-of-flight (ns)

12.0164
12.967

296.9
275.2

Table 3.1: The velocities and time-of-flight for the two major components of the incoming
beam to travel between the A1900 focal plane to the target.
24

with the sweeper set to accept the 26 F and 26 Na were used to identify the relative time-offlight of the beam. As Fig. 3.1 shows, the 26 F is clearly separated from the rest of the beam
by time-of-flight. Therefore, by gating on time-of-flight between the A1900 scintillator and
the target scintillator (in Fig. 3.2), we isolate the incoming 26 F beam.

Figure 3.1: Energy loss versus time-of-flight for the beam exiting the A1900 focal plane
showing the separated 26 F and 29 Na.

3.2

Element Identification

Element identification is achieved through measuring the energy loss and total kinetic energy
of fragments that travel through the sweeper. Energy loss is measured by the ion chamber
and the dE scintillator. The ion chamber has better resolution than the thin scintillator.
Total kinetic energy (TKE) is measured by the TKE scintillator.
25

counts

104
103
102
10
1
220 230 240 250 260 270 280 290 300 310 320

time-of-flight (ns)
Figure 3.2: The time-of-flight for beam particles traveling from the end of the A1900 focal
plane and the target position. The intense peak at 297 ns is the 26 F.

Plotting the energy losses in the ion chamber and the dE scintillator gives distinct element
identification. The intense section in Fig. 3.4 is the unreacted fluorine beam, therefore,
counting down we can identify oxygen and nitrogen fragments. We can also plot energyloss (ion chamber) versus TKE to identify elements, shown in Fig. 3.3. Both methods were
combined to make a clean nitrogen gate.

There were 6556 events of Oxygen and 1961 events of Nitrogen in the data. Adding
neutron time-of-flight gate (see Sec. 3.4) brought the total number of fragment-neutron
coincidence events to 4758 of oxygen and 1120 of nitrogen.
26

Total Kinetic Energy (arbitrary units)

40
14
35
12
30
10
25
8
20
6

15

4

10

2

5

0
0

500

1000

1500
2000
2500
Energy loss (arbitrary units)

0

Figure 3.3: Element identification gated on incoming 26 F via energy loss versus total kinetic
energy.

3.3

Isotopic Selection

Isotopic selection is traditionally achieved by separating the fragments of nitrogen by timeof-flight. However, since the flight paths through the Sweeper magnet are not uniform, mass
selection directly from time-of-flight is not possible. As Fig. 3.5 illustrates, because the
different isotopes have a different mass to charge ratios (A/Q) the direction of the flight
paths exiting the sweeper are correlated with fragment mass. Therefore, we can introduce
first and second order corrections for angle, position, and other parameters to correct the
27

Energy loss (arbitrary units)

40

14

35

12

30
10
25
8
20
6
15
4

10

2
0
0

5
50

100 150 200 250 300 350 400 450 500
Energy loss (arbitrary units)

0

Figure 3.4: Element identification gated on incoming 26 F via energy loss versus energy loss
(two different detector measurements).

time-of-flight. Dr. Hoffman found corrections that allowed separation of the oxygen isotopes,
shown in Table 3.3. As shown in Fig. 3.8, the different isotopes of oxygen are correlated with
angle. Because higher mass fragments of the same charge are deflected less than their lower
mass counterparts, we know that higher mass is correlated with larger angle. Because 25 O
is unbound and the sweeper was selected to let A/Q fragments of 3/1 through the center of
the focal plane, the slowest isotope isotope in Fig. 3.6 is identified as 24 O.
Consider a similar plot for the nitrogen isotopes in Fig. 3.9. The same 3/1 ratio applies,
28

magnetic field

CRDC

+x

high mass
medium mass
–x

low mass

target

Figure 3.5: A diagram showing how isotope separation works. Beam fragments leave the
target position with the same energy. Heavier isotopes (blue) are bent less than the sweeper
magnet and thus end up with a positive angle and position in the focal plane. Lighter
isotopes (green) and bent more by the sweeper magnet and thus end up with negative angles
and positions at the focal plane.

Parameter
Afp
Afp2
Xfp
Xfp2
Xta
Yta
Df light

Coefficient
60
-0.25
-3.5
-0.6
40
20
0.25

Description
1st order dispersive angle
2nd order dispersive angle
1st order dispersive position
2nd order dispersive position
1st order target x position
1st order target y position
flight path in focal plane

Table 3.2: Table listing the parameters of the corrected time-of-flight from Ref. [12]

29

counts

200
180
160
140
120
100
80
60
40
20
0
32

33

34

35

36

37
38
39
40
Corrected ToF (arb. units)

Figure 3.6: A corrected time-of-flight spectrum of oxygen fragments. The peaks correspond
to individually separated isotopes (from shortest time-of-flight to longest) of 22 O, 23 O, and
24 O.

so the heaviest isotope, which is centered in the focal plane, is identified as 21 N (which has
an A/Q ratio of 3/1). The subsequent lighter isotopes are 20 N and 19 N, which are clearly
separated by the corrected time-of-flight in Fig. 3.7.
30

counts

35
30
25
20
15
10
5
0
32

33

34

35

36

37
38
39
40
Corrected ToF (arb. units)

Figure 3.7: A corrected time-of-flight spectrum of nitrogen isotopes. The peaks lie in the
same time of flight of the oxygen isotopes identified in Fig. 3.6, so the peaks from left to
right are identified as 19 N, 20 N, and 21 N respectively.

31

X-Angle (mrad)

80

16

60

14

40

12

20

10

0
8
-20
6

-40

4

-60

2

-80
-100
32

33

34

35

36
37
38
39
40
Corrected ToF (arbitrary units)

0

Figure 3.8: The angular correlation of oxygen isotopes to time-of-flight. The lighter isotopes
are bent more by the sweeper magnet, so the isotopes of oxygen are identified (from left to
right) as 22 O, 23 O, and 24 O.

32

X-Position (mm)

80

4

60

3.5

40

3

20

2.5

0
2
-20
1.5

-40

1

-60

0.5

-80
-100
32

33

34

35

36
37
38
39
40
Corrected ToF (arbitrary units)

0

Figure 3.9: The angular correlation of nitrogen isotopes to time-of-flight. The lighter isotopes
are bent more by the sweeper magnet, so the isotopes of oxygen are identified (from left to
right) as 19 N, 20 N, and 21 N.

33

counts

900
800
700
600
500
400
300
200
100
0

-40

-20

0

20

40

60

80

100 120 140
time-of-flight (ns)

Figure 3.10: The total neutron time-of-flight spectrum. The peak at 68 ns corresponds to
prompt neutrons from reactions on top of random background.

34

3.4

Neutron Coincidences

After identifying events that contain the isotope of interest, it is necessary to tag events that
are coincident with a neutron. MoNA (the Modular Neutron array) is designed specifically
to detect fast neutrons and the total neutron time-of-flight spectrum from MoNA is shown
in Fig. 3.10.
Figure 3.11 shows the progression of adding more restrictive fragment gates to the total
neutron time-of-flight spectrum. As fragment and isotope gates are added, the neutrons left
are those that have the time-of-flight expected of decay neutrons at the same velocity as the
beam. The expected time-of-flight is 68 ns, because that is the time it takes for 85 MeV/u
neutrons to traverse 8.2 m.

3.5

Decay reconstruction

Invariant mass spectroscopy was used to reconstruct the decay energy (Edecay ). Equation
3.1 shows the formula used for calculating the decay energy.

Edecay =

m2 + m2 + 2(Ef ∗ En − pf ∗ pn ∗ cos(θopen )) − mf − mn
n
f

(3.1)

The equation takes into account the masses, energies, and momenta of the neutrons
and the fragments as well as the opening angle between the two particles. The masses are
taken from the most up-to-date atomic mass tables [1], while the rest of the parameters are
measured observables.
For the neutrons, the angle is calculated from the interaction point in MoNA and the
35

counts

103

102

10

1
-40

-20

0

20

40

60

80

100 120 140
time-of-flight (ns)

Figure 3.11: The progression of neutron coincidences as more restrictive element and isotope
gates are applied. The black (top) spectrum is the total neutron spectrum from 3.10. The
red (middle) spectrum is gated on nitrogen events, and the blue (bottom) spectrum is gated
on 21 N events. This progression shows that the neutrons accompanying 21 N are likely to
be “decay” neutrons from the break-up of 22 N.

36

energy and momentum is calculated from the time-of-flight and flight distance. For the
charged fragments, the angles are reconstructed from a partial inverse matrix produced
using COSY INFINITY [18], which also calculates the energy and momenta.

3.6

Monte-Carlo Simulations

In order to extract the energy of the resonance observed, Monte-Carlo simulations were
performed. The software package ST-mona is designed to take into account the experimental
resolutions and acceptances of the MoNA-Sweeper set-up in a Monte-Carlo simulation. These
decay events are analyzed in the same manner as experimental events so they can be directly
compared to experiment. Details of the simulation are given in Ref. [25, 23, 10, 12]. The
geometrical acceptances and resolutions were taken from Ref. [12].
The ST-mona simulation procedure is as follows: experimentally determined input parameters of the secondary beam on target were used to calculate the positions and angles
of the 2p-2n removal reaction inside the target. After the 2p-2n removal, the decay of the
parent nucleus into a fragment plus the neutron is nearly instantaneous. Two different distributions were used for this analysis; a single symmetric Breit-Wigner for the resonance and
a Maxwellian distribution for the non-resonance contribution.
After the decay the fragments are forward tracked through the Sweeper magnet using
a forward ion-optical matrix created by COSY [18]. Neutrons are forward tracked to the
position of MoNA. This allows for both neutron and fragment data to be directly compared
to experimental data. Ref. [12] describes and shows how the experimental data matched the
simulation for 24 O + n and 23 O + n events.
Since the simulation is produced in the same manner as the experimental data, the same
37

invariant mass method can be used to make decay spectra that can be compared directly
to data. The simulated decay spectra are fit to the experimental decay spectra to extract
physical properties of the decay. This was carried out by varying the parameters of the
distributions and adopting a goodness-of-fit minimum chi-squared (χ2 ) value.
A single Breit-Wigner was used as the input for the resonance contribution of the decay
while a Maxwellian distribution was used for the non-resonance contribution. The choice
of a Maxwellian was an arbitrary choice because the low statistics do not justify extracting
any physics from the non-resonant distribution. In fact, even the choice of the energy was
arbitrary because the shape of the background was insensitive to the fit of the simulation to
data.
The input for a Maxwellian distribution is:

fE = 2

E
πkT 3

exp −

E
kT

The free parameter of the distribution for fitting is kT , usually measured in MeV.
The basic equation for a resonance comes from Breit Wigner as shown in equation 3.2.

fE =

Γ
2
(E − Er )2 + Γ
4

(3.2)

Two physics parameters are extracted; the energy of the resonance (Er in the equation)
and the width of the resonance (Γ in the equation).
It should be noted that for these kinds of experiments with MoNA-sweeper, the actual
width of the resonance in the decay spectrum is dominated by the resolution of the setup as
shown in Fig. 3.12. The Breit-Wigner function plotted has a width of 6 keV while the input
38

counts

2500
2000
1500
1000
500
0
0

0.2

0.4

0.6

0.8

1

1.2

decay energy (MeV)
Figure 3.12: The data points indicate a narrow width Breit-Wgner resonance from a STmona simulation and the solid blue curve indicates the same Breit-Wigner resonance without
any acceptances. The width of the decays were 6 keV with a energy of 650 keV.
of the simulation was a Breit-Wigner of zero width.

3.6.1

Fitting to data

Simulations with the resonance and the non-resonance decays were fitted via a minimization
of χ2 allowing the amplitudes of the contributions to vary freely as well as the energy and
width of the resonance. The temperature of the thermal was insensitive to the fit, so an
arbitrary choice of 6.5 MeV was taken to reduce the number of free parameters. A χ2 fit
of a Breit-Wigner resonance on top of a Maxwellian background yielded a decay energy of
39

Decay width (MeV)

0.12
0.1
0.08
0.06
0.04
0.02
0

0.55

0.6

0.65

0.7

0.75

Decay energy (MeV)
Figure 3.13: A contour plot of the χ2 dependence of the Breit-Wigner parameters for the
measured decay energy of 21 N + n. The parameters are the decay energy (x-axis) and the
width of the decay (y-axis) in MeV. The lowest χ2 is located at 0.65 MeV and 0 keV.
650 keV ± 50 keV with zero width. Figure 3.13 shows the 1-σ and 2-σ confidence levels of
the fits. The width of the resonance was dominated by the experimental resolution of the
experimental set-up and only an upper limit of 60 keV could be established from a 1-σ limit.
Figure 3.14 shows the final fit to the data.

40

C o u n ts / 1 5 0 k e V

3 0

re s o n a n t
n o n -re s o n a n t
s u m
D a ta

2 0

1 0

0
0

1

2

3

E n e rg y (M e V )
Figure 3.14: Neutron-decay energy spectrum of 22 N. The data with statistical error bars
were measured by requiring a coincidence between an identified 21 N fragment and a neutron.
The solid line is the sum of the resonant (dotted) and non-resonant (dashed) contributions
of a Monte-Carlo simulation that takes into account the resolutions and acceptances of
the experimental apparatus. The resonance line-shape is a Breit-Wigner with Edecay =
0.65(5) MeV.

41

Chapter 4
Results and Discussion
In order to extract the excitation energy of the resonance observed, the binding energy has
to be added to the decay energy. The neutron separation energy of 22 N is 1.28(21) MeV [1].
Adding this to the resonance of 0.65(5) MeV results in an excited state energy of 1.93(22) MeV.
It is interesting to note that although the statistics for the determination of decay is rather
limited, the uncertainty of the excitation energy is still dominated by the uncertainty of the
mass measurement.
The underlying assumption is that the observed decay goes to the ground state of 21 N.
While we cannot rule out the possibility that the resonance decays to an excited state of 21 N,
an examination of the level scheme in Fig. 4.1 shows that this case is unlikely. The lowest
measured excited state of 21 N (a 3/2− state) is located at 1.177 MeV [27], which would
place the resonance above 3 MeV in 22 N. This is unlikely, because none of the commonly
used shell model interactions predict the first unbound excited state of 22 N at such a high
excitation energy.
The spin and parity assignment of the observed resonance is guided by shell model cal42

MeV

26 F

3

2p-2n reaction
3/2−

2
1/2−
1

3−

neutron emission

650 keV

21 N

2−

1.28 MeV
0

22 N

1−
0−

Figure 4.1: The proposed level scheme of 22 N → 21 N + n decay created from a reaction of
26 F.
culations. This is the common approach to analyze states in nuclei far from stability where
angular distribution measurements are difficult. The ground state of 22 N has been assigned
a spin and parity of 0− . Sohler et al. [27] identified the first two excited states of 22 N to be
1− and 2− respectively, guided the shell model. The next predicted state is a 3− , thus we
attributed the decay observed to be the 3− .
B. A. Brown [5] calculated the single particle width for a 0d5/2 -wave decay at 650 keV to
be 55 keV. With the spectroscopic factor for a 3− to 1/2− decay equal to 0.1059 (calculated
via the shell model [7]), the observed width would be 5.8 keV which is within the upper limit
of the width observed at 60 keV.
The second 2− state of 22 N, which would decay in a 0d3/2 -wave decay is unlikely to be the
state observed by this experiment because the expected width of such a state (∼ 500 keV [5])
is an order of magnitude larger than the maximum possible width observed. Also, since we
populate this state from a beam of 26 F, there is only one neutron in the ν0d3/2 shell, as
shown in Fig. 4.2. Since this is the outermost shell, the cross section to strip out this nucleon
43

ν0d3/2
ν1s1/2
ν0d5/2
π0d5/2
π0p1/2

ν0p1/2

π0p3/2

ν0p3/2

π0s1/2

ν0s1/2

Figure 4.2: The ground state shell configuration of 26 F. Note the lone neutron in the ν0d3/2
shell.
in a multiple nucleon stripping reaction is quite high.

4.1

Shell model calculations

Shell model calculations using the code NuShell@MSU [7] were carried out for 22 N levels.
Because the USD TMBE used in the WBT interaction reproduce the level energies for
oxygen isotopes successfully, but not the carbon isotopes, Staniou et al. [30] proposed a
25% reduction of the neutron-neutron TBME to make a better empirical agreement for
the neutron rich isotopes of carbon. Sohler et al. [27] observed the same compression for
nitrogen isotopes and thus a 12.5% reduction (half of 25%) was proposed. These reductions
were justified by comparing the calculation to measured bound excited states of several
neutron rich nitrogen (19−22 N) and carbon (17−20 C). We reproduced these calculations
with the code NuShell@MSU. We applied the same 12.5% reduction to the WBT and WBP
interactions (called WBTM and WBPM respectively).
44

Energy (MeV)

3

2
3/2

3

–

2

3
1/2

1

–
–

3

–

2

N+n

2
1
0

–

–

–

2

–

1

–

21

2

–

–

1

–

–

–

22

N

WBTM

WBPM

Figure 4.3: Level scheme of 22 N. The present experimental data are shown together with the
data of Reference [27] and are compared to WBTM and WBPM shell model calculations.
The neutron separation energy as well as the ground and first excited state of 21 N are also
shown. The shaded area around the 21 N ground state represents the uncertainty due to
the neutron separation energy. The shaded area around the measured 3− state corresponds
to the excitation energy uncertainty which is dominated by the uncertainty of the neutron
separation energy.
Figure 4.3 shows the results of NuShell calculations for 22 N excited states. As can be seen,
the calculated levels with the WBTM are consistently at higher excitation energies compared
to measured levels, especially the 1- state. Calculations using the WBP Hamiltonian, which
uses a different choice in the fitting of psd levels, showed an improvement in energy for the
1− state. We applied the same 12.5% decrease of the neutron-neutron interaction strength to
the WBP interaction (label as WBPM interaction in Fig. 4.3) which resulted in an improved
overall agreement with the data. The excitation energy of the 3− calculated with the WBPM
interaction is 2.12 MeV which is within the uncertainty of the measured resonance. It should
be noted that the modified theoretical calculation of the 2− state shows that it is within the
uncertainty of the neutron separation energy of 22 N.
Another way to look at what would be predicted for the 3− state of 22 N is to examine
45

the measured results for 20 N. The first two states for 20 N are a (π0p1/2 )×(ν0d5/2 ) configuration, so by the triangle condition the spin-parity is 2− or 3− . The ground state is 2−
and the first excited state is 3− . The shell model was successful in predicting the energy of
the excited state energy of 20 N. Now the 22 N has the same configuration as 20 N with the
exception being a filled ν1s1/2 shell. So the energy difference between the 2− and 3− in
22 N should be similar to the 2− and 3− in 20 N. Adding in the excited energy of 20 N to the
measured energy of the 2− state of 22 N gains:

E2− + E3− = 1.09 + 1.02 = 2.11 MeV

(4.1)

This compares favorably with our measured 3− energy of 1.93(23) MeV.

4.2

Disappearance of the N = 14 shell

As described earlier in Sec. 1.4, performing a weighted average of states corresponding to
particular shell orbits will yield a gap in energy between the shells. This allows experimentalists to predict and confirm a level inversion between two orbits. Shown in Fig 1.7, Talmi
and Unna [31] used this method to confirm a level inversion between the 0s1/2 and p1/2
shells in 11 Be. Similarly, Talmi and Unna [31] used averaging of states in the N = 9 isotones
to predict a level inversion of the ν0d5/2 and ν1s1/2 shells. This result is confirmed in shell
model calculations that show that the ν1s1/2 goes down in energy as protons are removed
from 17 O. Comparing the levels for N = 9 for oxygen (Fig. 1.4) and for carbon (Fig. 1.5)
bears this conclusion out. As neutrons are added to the N = 9 isotopes the orbits return to
their normal ordering because the 0d5/2 drops in energy as it is filled with neutrons.
46

We can take a similar approach in analyzing the N = 14 shell gap for the N = 15
isotones. The N = 14 shell gap is the gap in energy between the ν1s1/2 and the ν0d5/2
shells. The shell gap for 23 O is 2.79 MeV as evidenced by the measured excited state energy
of the 5/2+ state [26]. (Only one excited state is necessary for 23 O because the 5/2+ is the
only possible configuration as shown in Fig. 1.9). With the addition of the 3− state of 22 N
to the known states of 22 N, we can estimate the shell gap for 22 N. The 2J + 1 summing
average of the 0− and 1− states corresponds to the ν1s1/2 level and the average of the 2−
and 3− corresponds to the ν0d5/2 level. The resulting gap of 1.41(17) MeV, significantly
smaller than the N = 14 shell gap of 2.79(13) MeV deduced for 23 O. This reduction of
1.38(26) MeV in the N = 15 isotones is consistent with the reduction of 1 MeV reported
between 22 O and 21 N for the N = 14 isotones [27]. The continuation of the reduction of
the size of the shell gap and perhaps the emergence of the level inversion can be calculated
using a linear extrapolation first introduced by Talmi and Unna [31]. Figure 4.2 shows
the measured difference between the ν1s1/2 and ν0d5/2 levels for the N = 15 oxygen and
nitrogen isotones with extrapolation to 21 C. The extrapolation suggests that the levels are
essentially degenerate (0.03 MeV) in 21 C. Within the experimental uncertainty, the levels
could be again inverted similar to the level inversion of 15 C.

47

Energy (MeV)

4

3
ν1s1/2

2.79

2
3−

1.93

1
1.02

1−

0.19
0

C

2−

0

N

0−

O

ν0d5/2

Figure 4.4: The “competition” between the ν1s1/2 and ν0d5/2 levels for the N = 15 isotones
of carbon, nitrogen, and oxygen following the prescription of Ref. [31]. Experimental data
for the levels of 23 O and 22 N are taken from Refs. [26],[30],[27], and this work. It should
be noted that the states labeled ν1s1/2 are actually ν0d5/2 -hole states, but these states
measure the ν1s1/2 level because a neutron is required to move from the 0d5/2 to the
1s1/2 . The blue box around 21 C indicates the range of expected gaps for the ground state
based on the error bars of the 22 N and 23 O measurements.

48

Chapter 5
Conclusion and Future prospectives
We report here the measurement of a resonance of 650(50) keV, from the neutron decay of
an excited state in 22 N. This state was induced from reactions of a 85 MeV/u 26 F beam on
a beryllium target. The decay fragments and neutrons were detected by the MoNA-Sweeper
set-up and reconstructed to achieve a decay energy. We attributed this to the neutron decay
of the first 3− state of 22 N with an excitation energy of 1.93(22) MeV.
Combining this result with previous work on excited states of 22 N by Sohler et. al. [27],
we calculated a N = 14 shell gap of 1.41(17) MeV for 22 N. Having established the N = 14
gap for the N = 15 nucleus 22 N from a weighted average of the ν1s1/2 -particle and ν0d5/2 hole states of 22 N, we combined these results with previous measurements of the N = 14
shell gap for 23 O, 22 O and 21 N. Extrapolation of the energy between the ν1s1/2 and ν0d5/2
for the the N = 15 isotones point to a possible inversion in the orbitals for 21 C. This would
be a similar effect to the one observed by Talmi and Unna [31] of the inversion of the ν1s1/2
and ν0d5/2 orbitals in the N = 9 isotones.
This extrapolation leads to the conclusion that the assignment of the angular momentum
49

3+

4.58
4.20

7/2−

3.32
3+

5/2−
3.20

Energy (MeV)

4

3

??

ν1s1/2

2+

5/2−

2.41
2
2+

1.59

3/2−

1.18
1

0

C

N

O

ν0d5/2

Figure 5.1: The “competition” between the ν0d5/2 and ν1s1/2 levels for the N = 14 isotones
of carbon, nitrogen, and oxygen following the prescription of Ref. [31]. Experimental data
for the levels of 22 O, 21 N, and 20 C are taken from Refs. [30],[29], and [27]. Higher excited
states in 20 C, like the next expected ν1s1/2 -particle state with J π = 3+ , have not been
observed.
of the ground state 21 C may not be trivial. The possibility exists that the 21 C ground state
is 5/2+ and not 1/2+ (the shell configuration is shown in Fig. 5.2). It would be interesting
to measure the ground state energy of 21 C as well as investigate the J of 21 C.
There have also been incomplete studies of the N = 14 gap for the N = 14 nucleus 20 C.
Fig. 5 shows the same analysis of 1s1/2 -0d5/2 gap for the N = 14 isotones. As the linear
extrapolation suggests a reduction of the gap for N = 14 isotones, however, measurements
for the N = 14 carbon isotope 20C is incomplete. Stanoiu et. al. [30] measured the excited
states of carbon isotopes and found only one bound excited state for 20 C, a 2+ state at
50

ν0d5/2
ν1s1/2

π0p1/2
ν0p1/2
π0p3/2

ν0p3/2

π0s1/2

ν0s1/2

Figure 5.2: The shell configuration of an inverted ground state of 21 C.
1.588 MeV. Measuring the 3+ excited state of 20 C would complete the levels needed to
calculate the N = 14 shell gap. Since it was not observed in gamma spectroscopy, neutron
decay spectroscopy is needed to measure this state.
While we have found the dripline up to neon, the neutron drip line above neon remains
a mystery. With new experiments exploring the upper limits of stability, physicists have
found isotopes that are bound more than we ever thought before. For example, a recent
experiment at the NSCL discovered two neutron rich isotopes of 40 Mg and 42 Al [2] provide
evidence that the dripline is further slanted to neutron rich nuclei than we have thought.
The emergence of new magic numbers and closed shells is likely to continue.

51

Appendix A
New Experiment Setup Guide

A.1

Introduction

The purpose of this document is to describe how to create and manage a new MoNA-LISASweeper experimental account. Most of the following procedures can be done from a data U
machine by logging in with the experiment user account.
While the original guides attempted to do a step by step procedures for everything, the
adoption of ROOT for analysis (with its vast customizability and extensibility) has made such
procedures moot because an experimenter might have different choices for online-analysis or
daq needs. Every attempt will be taken to make this coherent and that newbies can read
this guide, but the following rule applies: There is more than one way to skin a cat.

A.2

The Purpose of Experimental Accounts

The entire purpose of an experimental account is provide the following for the experimenter:
• Provide space to temporarily store data from the experiment in a write protected disk
52

space, called event space.
• Provide a location for customized programs that run the experiment. Some examples
include: the data acquisition, online data analysis, high voltage controls, and trigger
logic.

There are four kinds of Linux computers that can access the experimental account. These
are the data acquisition computers (DAQ), called spdaq-machines, the data U machines, and
the element (32-bit) and the fishtank (64-bit) machines. The spdaq and data U machines
have the same architecture, however, the spdaq machines (located in the vault) are dedicated
for use in data acquisition. The data U machines can access the spdaq machines (more on
that later) and can be used to do special online analysis without taking up CPU of the
spdaq machines. Both spdaq and data U machines has access to the event space (\events\).
While the element and fishtank machines have the same user area as the data U and spdaq
machines, they do not access to event space. They can be used for offline analysis after the
data has been copied to evtdata space (a non-write protected disk space for data).
All the current codes and files for running a MoNA-LISA-Sweeper experiment are located
in the monasoft account. They can be accessed and copied directly from the /user/monasoft
account or via cvs 1 .

A.3

Acquiring the Experimental Account

A couple weeks before the experiment is scheduled to run you should request the computer
department to create a user account. The user account name is set to the experiment number
1 cvs is a version control system used manage changes in code for important programs. It
requires “checking out” code from a repository in the monasoft account.
53

(#####) issued by the PAC with an e in front (like e03038 or e05124). If several different accounts for the same experiment are needed (like for time-stamped experiments) the
different accounts will have a lowercase letter appended on the end (e.g. e10023a, e10023b).
When the account is made, the computer department will give you a randomized password
for the account.

A.3.1

Passwords

The Linux (or DAQ) password must be changed to a custom one when you login the first
time by using the yppasswd command. It will ask you to enter the original password and
then your new password twice.

A.3.2

Logbooks

Log books are provided by the NSCL and one should be taken from the main-level copy
room. Put the date and the experiment number on the cover. It is customary to put on the
first page a list of collaborators and the password.

A.3.3

SSH

Many DAQ codes are called from the Data U machine but actually run on another computer
in the experimental vault. Since the codes are trying to communicate with the spdaq machine, it will ask for the password before continuing. This pause in the action will usually
cause the desired program call to fail. By setting up a proper .ssh directory you can then call
these programs without providing a password. In the /user/monasoft account is a script
that will crate the .ssh directory automatically. Copy auto ssh.sh from /user/monasoft
54

and run it to do this. There are also directions in the Data U for setting up the .ssh directory
to make secure connections to DAQ machines of the same account without passwords.

auto_ssh.sh:
#!/bin/bash

### This script will make it such that a password is not necessary
### to ssh into other NSCL machines.
### -Written by WAP
### -Transcribed by GC
### -Quality control by MJS and SMM

cd ~/
chmod 0750 .
rm -rf .ssh
mkdir .ssh
cd .ssh
chmod 0700 .
ssh-keygen -t dsa -P "" -f id_dsa
mv id_dsa.pub authorized_keys
mv id_dsa identity
chmod 0600 identity
chmod 0640 authorized_keys
cd ~/
55

After setting up the .ssh, delete the auto ssh.sh from the home directory to keep your
home directory clean.

A.3.4

Windows Password Change

The current password for the Windows machine (u5pc1) is written on the back page of the
MoNA moving log Logbook. The password needs to be changed periodically (every couple
of months or so), because of the security settings of the operating system. If prompted to
change the password, do so, and write the new password in the MoNA moving log’s back
page.

A.3.5

Connecting to the outside internet

The data U machines are only allowed to have an outside connection to the internet (to use
www.google.com, etc) with a special proxy. Directions to do this are found on the NSCL
Computer Department website: https://intra.nscl.msu.edu/hd_drupal/faq/internet/proxy.
The basic steps are:
• Open Firefox.
• Go to Edit and then Preferences.
• Click on the Advanced tab.
• Go to the Network tab and click on Settings.
• Click on Manual Proxy Settings and fill in the HTTP Proxy: webproxy.nscl.msu.edu
and Port: 3128
56

• Check the box that says Use this proxy server for all protocols. Click OK and close
the preferences window and you are good to use Firefox to search the web.

A.4

BASH files

Similar to the .ssh directory, some of the programs refer to the standard directory structure
(outlined in the next chapter) and are given environment variables within the .bashrc file.
The standard bash files can be copied from the /user/monasoft/Dot files directory. After
logging in with the experiment account, follow these commands. For the first three files, the
terminal is going to ask if you want to overwrite; the answer will be yes.

> cd ~/
> cp -r /user/monasoft/Dot_files/ .
> cp ~/Dot_files/.bash* .
> cp ~/Dot_files/.g* .
> cp ~/Dot_files/.v* .
> cp ~/Dot_files/.rootrc .
> cp -r ~/Dot_files/.ipython ./
> source ~/.bashrc
> rm -rf ~/Dot_files

The important files are the BASH files. There are:

• .bash aliases – for useful aliases of shell commands. Feel free to add any more aliases
you would like to this file.
57

• .bash env – environment variables for BASH...used to source proper directories in
spectcl, ROOT, and readout and to use cvs
• .bashrc – Configures your BASH environment; sources .bash aliases and .bash env on
login and sets the BASH prompt.
• .gvimrc, .viminfo, .vimrc – useful files to customize VIM if you so desire
• .rootrc – this customizes ROOT so you can use the classes from n2analysis
• .ipython – a directory that is add configuration files that add some nice utils for
ipython, including some functions that are used by ROOTBeer

A.4.1

event space

When the account was created the computer group should have also issued space on an
event disk for recording event files. Create a link called stagearea to this issued directory.
> cd ~/
> mkdir /events/e08016/stagearea
> ln -s /events/e08016/stagearea stagearea
Replace 08016 with the proper assigned values issued by the computer group.

A.5

Getting the code from monasoft

There are four classes of codes you need to make an experiment work: readout, an online
analysis code (SpecTcl or ROOTBeer), utility codes and settings files, and the trigger logic
(xlm).
58

Computer System
readout name
command
spdaq34
sweeper readout-sweeper caesar
godaq
spdaq40
MoNA
readout mona
gomonareadout
spdaq42
LISA
readout lisa
golisareadout
Table A.1: The spdaq machines and corresponding detector systems and readout codes.

A.5.1

Readout

Readout is the code that reads out the data from the respective VMEs and writes this data
to disk in a known structure that can be reinterpreted later by an an analysis code.
Because the spdaq machines can handle only so many VMEs being read out to them,
MoNA-LISA-Sweeper separates the daq process onto three separate machines necessitating
three readout codes and merge the data post-experiment.
A handy (yet incomplete) guide to Ron Fox’s Readout code is located in
http://docs.nscl.msu.edu/daq/bluebook/html/.
Check out all three codes by typing the following commands and entering the password
for the monasoft account when prompted2 :

>cd ~/
>cvs checkout readout_mona
>cvs checkout readout_lisa
>cvs checkout readout-sweeper_caesar
To make the executables execute the script make all.sh from every readout ****/skel
directory. Another task the scripts will do is put scripts in the $HOME/bin directory that
start each readout without needing to be in a specific directory. These scripts (also known
as commands) are (for MoNA, LISA, and sweeper) gomonareadout, golisareadout, and
2 all cvs commands require the monasoft password
59

godaq, respectively.

A.5.1.1

The experiment directory

The readout codes need an experiment folder to run properly. Create this by starting either
mona or lisa readout. When the ReadoutGui starts up make sure that the Record button is
unchecked and then start and stop a run. With this test of Readout, it will automatically
make a linked experiment directory as well as the subdirectories of stagearea: current,
orphans, and staged. When Readout runs during the experiment, a folder is made after
each run is completed called run####.

A.5.2

Online Analysis Code

The NSCL supports the use of SpecTcl for analyzing data. Our group has created a program
that uses SpecTcl called spectcl. (Unique, right?) This code has all that is necessary to
unpack online data and event files for sweeper, MoNA, and LISA (both separately and
merged). If you so choose, this spectcl can be used as the online analysis code.
>cd ~/
>cvs checkout spectcl
>cd spectcl/skel
>./make_all.sh
You can start spectcl by typing gospec in the terminal.

A.5.3

XLM

The XLM controls the trigger logic for MoNA, LISA, and for Level 3.
60

To check out the XLM

>cd ~/
>cvs checkout tools

and give the proper password when prompted.
Before the XLM can be used properly, check with Thomas Baumman what the correct and
most current bitfiles that are needed. They should be put in fpga folder and the .bash env
file should have the correct environment variables of bitfiles pointing to the specific ones.

A.5.4

n2analysis

The n2analysis package contains C++ classes and macros to use ROOT for analysis of
MoNA-LISA-Sweeper data. It also has the code that unpacks .evt files into root files.
To check out this package from the repository, type:

>cd ~/
>cvs checkout n2analysis

Further instructions on how to make ROOT work are in Sec. A.7.1.

A.5.5

evtMerge

The code that merges event files from timestamping can be checked out by typing:

>cd ~/
>cvs checkout evtMerge
61

A.5.6

Using a script to automate the rest

A useful script called make homedir.sh, automates the copying of files from monasoft to
the home account. Execute this script by typing:

>cd ~/
>cp /user/monasoft/make_homedir.sh .
>export THEHOME=$HOME
>./make_homedir.sh

The code should be understood, so that if it becomes necessary, someone can diagnose
or do by-hand what the script does. I would suggest opening the script to see that you
understand what the script does3 .
The script does in order:

• Creates the bin directory including sub-bins for 32-bit and 64-bit executables.

• Copies the SpecTclInit.tcl and Xamine files for SpecTcl.

• Copies from monasoft the following: fpga, noncvstools, and win.

• Checks if you have a stagearea.

• Create a sweeper settings folder in experiment/current and links the settings files in
the sweeper account so readout-caesar sweeper can work.

• Create a mona settings folder in experiment/current and links settings files to it.
3 It is also a handy way to learn how BASH scripts work.
62

A.5.7

the current directory

The contents of the experiment/current directory is copied into each run#### folder. All
important setting files for the experiment should be linked to from the experiment/current
directory. Since the sweeper readout cannot run without the sweeper settings files to be
linked, there are two folders created in the current mona settings and sweeper settings.

A.6

Setting up Readout

If you haven’t made any changes to the experimental set-up, then Readout is properly set-up
and you need not do anything else.

A.6.1

MoNA-LISA Settings files

Both MoNA and LISA readout are scripted. That means that TcL scripts configure the
hardware that the program readouts out. The critical script to check if anything needs to
change in the readout are:
• **** config setup.tcl – Defines the names and locations for the QDC and TDC
hardware.
• **** hardware run.tcl – Sets the packet strucutre and initializes the hard ware.
• **** readout run.tcl – Prints the configuration of the Readout.

A.6.2

sweeper readout settings files

Daniel Bazin puts his settings files for all the hardware for the sweeper set-up in the
/user/sweeper/stagearea/current/sweeper settings folder. Since these settings files
63

needed to be sourced by readout for every Begin Run, we make links to these files in
stagearea/current. The script that automates the rest of setting up the account should
have made these links already, but you should check that these are correct and if need fix or
add these links.

A.7
A.7.1

Online Analysis Codes
ROOTBeer & ROOT

To configure ROOT and the custom classes that allow ROOT to interpret MoNA-LISA data,
please do the following:
• Edit the .rootrc file so that all references to the account name are correct
• Type in the terminal:

>cd ~/n2analysis/src
>./compile_lib.sh detector hh
>./compile_lib.sh sweeper cc
>./compile_lib.sh caesar cc
>./compile_lib.sh mona cc
>./compile_lib.sh utils cc
>./compile_lib.sh unpacker cc

• Then do: > make clean
• > make
64

And now when ROOT starts up is should have the classes defined to be able to read root
files of MoNA-LISA data and the evtunpacker should be to unpack .evt files.

A.7.2

SpecTcL

Starting spectcl requires that you type >gospec. The window files for Xamine are located
in win.

A.8
A.8.1

Utility Codes
High Voltage Controller

Located in noncvstools/high volts.
things you should be able to do with the controller
List of important commands
• array loadconfig *.txt - loads an array configuration file. Needs to be done every time
the hv controller is started.
• array connect - open TCP connections to the power supplies. If you get 0x0, then the
connection was successful.
• array on/off - turn on/off the whole array
• array importHV fname - Imports HV voltage settings file of new high voltages from
gain matching.
• array exportHV fname - Exports the Current HV voltages to a text file for gain matching.
65

• tube set name volts - Set the tube named name to an integer volts.
• tube on/off name - Turn on/off the named tube.
• tube readV/readI/readP name - Read the current voltage/current/power state of name.
1. To open to the HV controller program, type HVnewtext.
2. In the HV controller prompt, type:
MoNA HV> array loadconfig MoNA_hv_defs_1007.txt
MoNA HV> array connect
MoNA HV> array on
MoNA HV> array exportHV ../hv_files/Vout_current.tcl
MoNA HV> end
These commands connect the HVcontroller program to the HVmodules, turns on the
voltages (if they weren’t on already), and exports the current voltages to a file that is
readable by the FORTRAN HV fitting code.
1. Now start the HV controller program and connect to the HV modules typing:
> HVnewtext
MoNA HV> array loadconfig MoNA_hv_defs_1007.txt
MoNA HV> array connect
2. Now source the new voltage values into the HV channels by typing into the HV controller:
MoNA HV> array importHV ../hv_fitting/new_hv_values.txt
MoNA HV> end
66

HV Module
MoNA right
MoNA left
LISA right
LISA left

$HVmodule
caenhv03
caenhv02
lisacaenhv01
lisacaenhv00

IP address
35.9.56.157
35.9.56.157
35.9.56.178
35.9.56.177

Table A.2: The different high voltage modules and their names and IP adresses.
A.8.1.1

Emergency HV control

If the above program does not work, in a pinch you can telnet directly to the computers of
each high voltage controller.
Type > telnet $HVmodule 1527, where $HVmodule is one of the modules listed in Table
A.8.1.1.

A.8.2

HV fitting

The HV fitting folder has the following files:
• class.C – a class file that defines the qdc histograms
• class.h – a class file that has the root files being sourced
• save hists.C – uses the class to output the histograms of qdcs to a macro file.C
• hvfit.C – functions for fitting — max bin, peak, etc
• run hvfit.C – macro that runs the fitting routine on file.C for gain matching
• run qdcfit.C – macro that runs fitting routine on file.C for qdc thresholds
• Edit copy cosmicruns2root.sh – to copy the evtfiles from stagearea to evtdata and to
use the evtunpacker to unpack the evt files to root files.
Consult the Calibrations guide to learn how to use these files.
67

A.8.3

NSCLDumper

The NSCLDumper is a NSCL supported program that allows for examination of events from
online and from event files. To start the nscl buffer dumper go to /noncvstools/NSCLDumper
and type >wish NSCLDumper.tcl &.
Within the same directory is a TcL script that can calculate the total scalers for an event
file. To start it, type wish scalersum.tcl &.

A.9
A.9.1

Running the experiment
Readout crashing

If a readout or a spdaq machine crashes during a run, the automatic copying and transferring
of settings and event files will not happen properly. A properly recorded run creates a run
folder, /experiment/run####, that contains the two settings folders and a link to the actual
.evt file in /events/e#####/complete.
To get the links and the Readout program back to working order after a crash first check
to see if the spdaq machine needs to be rebooted.
One can reboot the computer remotely when logged in to spdaq from data U by typing
sudo reboot.
> ping spdaq##
Type [CTRL]^C to quit pinging. If there is no response you must go into the N2/N3 vault
and reboot the spdaq machine (directions for rebooting are on orange tape on the machine).
If the pinging worked or after rebooting spdaq16 make sure there are no Readout codes
running. The following command lists the current programs containing ”eadout”.
68

> ps axuww|grep eadout
Then kill any programs still running by using the procedure number listed:
> kill ####
Rebooting the spdaq machine will also kill stray procedures, but requires the user to restart
many control programs after.
Follow this check list before continuing with the next run.
1. Restart Readout and manually increase the run number.
2. Start and then stop a run without recording.
3. Use the >ls -la command in the /experiment directory to look for a run folder
(run####) made for the interrupted run.
4. Look for copied mona settings and sweeper settings folders within the the run
folder.
5. If needed, copy the settings folders into the interrupted run folder:

> cd ~/experiment
> cp -rp /current/mona_settings/ run####/
> cp -rp /current/sweeper_settings run####/
6. Now look for a linked .evt file.
Within the run folder, look for a linked .evt file pointing to the actual .evt file in
one of the stagearea subdirectories.
Look for a linked .evt file in the experiment/orphans or /current directories.
69

7. Find the actual .evt file in stagearea/complete or /current or /orphan directories.
8. Move the actual .evt file for the interrupted run into stagearea/complete.
9. Fix links to the .evt file for the interrupted run within the experiment/run#### folder
to link to the actual file now located in the stagearea/complete directory.

> cd ~/experiment/run####
> ln -s ~/stagearea/complete/run####-4096.evt .

10. If needed, delete old links from the experiment/orphans or /current directories.

If all goes well, the next run can now be recorded. Be sure to make a note in the log
book during which run the Readout program halted.

A.9.2

Recovering from a reboot

To recover from the reboot, do the following:
1. Restart the CFDs.
2. Restart the XLM. Check XLM.
3. Restart Readout.
4. Restart scalers.
5. Restart online analysis. Attach to spectrodaq.
6. Start run!
70

A.10

Post-experiment

As the experiment time is completed there a few prudent procedures to perform to ensure
the data is capable of being analyzed properly.

A.10.1

Calibration runs

Before turning MoNA-LISA high voltages off, take a couple cosmic background runs for
future calibration checks. One long run ( 6 hours) with MoNA/LISA self-triggered and selfstopped and multiplicity set to ”1-fold” and then another long run, again self triggered and
stopped, with multiplicity set to ”2-fold”. The first one can be used to double check the
QDC and Xpos calibration. The second run is used to set the T-mean independent offsets
for each bar.

A.10.2

Archive tapes

When all the experiment runs and the additional calibration runs are completed, the event
files must be copied to tape and then read off the tape into the evtdata disk space. The
NSCL provides a nice manual for copying data onto tapes called ”Preparing to Close your
Experimental Account”. I will not cover all the directions contained in that manual, so
it should be read as well. I will go through the most common steps and address common
problems that have arisen in the past.
1. Acquire two LTO or similar digital tapes from the Computer help room. LTO tapes
are named by their size in 100 GB (i.e. LTO1 holds 100 GB, etc.). One of the tapes
only needs to hold a GB or so, so choose the smallest size available for that one; the
other should be large enough to hold all of your event files.
71

2. Double check that all run folders within the experiment directory contain proper links
to actual .evt files in the stagearea/complete directory.
3. Login to the “tapehost” machine (currently in the Data U. 2 cubicle) with the experimental account.
4. Change directories to the experiment directory. > cd

/experiment

5. Identify the tape drive you are using. Maybe /dev/st2 or /dev/st1. It is used as the
third argument in steps 6 and 9.
6. Record entire experiment directory onto tape. > tar -cvhf /dev/st1 .
7. Wait for entire directory to be recorded onto the tape (it may take a couple hours).
The terminal window should display all items written.
8. Change directories to the evtdata/e##### space provided by the computer group. >
cd /evtdata/e#####
9. Read the files from the tape. > tar -xvf /dev/st1 .
10. When the reading is completed, check that the entire experiment directory was copied,
including actual .evt files (not links).
11. Eject tape and label it with the date, experiment number, and runs included.

If all the /experiment run folders were repaired properly if Readout crashed the actual
.evt files should have been copied correctly. If they were not copied correctly, find the run
folders for the missing .evt files and rewrite the link to the actual .evt file. Then delete the
files you just wrote to the evtdata/e##### space and start over from step 4.
72

Since this process copies each .evt file inside their respective experiment/run#### folders,
it is not possible to attach multiple files when analyzing the data with SpecTcl and/or ROOT.
A useful trick is to make a new folder within the /evtdata/e##### space that contains links
to all the .evt files.

> cd /evtdata/e#####
> mkdir evt_links
> cd evt_links
> for i in ../run*/*.evt ; do ln -s $i . ; done

After you are done, check that all .evt files have a link in this folder.

A.10.3

Analysis Space

Along with an /evtdata space the computer group also should have provided a /projects
space to copy the analysis files. The /projects area can be accessed from the high-speed
element and fishtank machines, and should be used for all off-line analysis. You will need to
transfer the contents of your experimental account’s home directory (i.e. /user/e#####) to
the /projects area. It is a good idea to make a tape backup of the home directory area as
well; this should be done on a tape separate from the one containing your event files. To do
this, use a procedure similar to that outlined in Section A.10.2, except that you will need to
specifically exclude the /experiment and /stagearea directories (since you don’t want
to copy the event files twice):

1. Change directories to the experimental home directory: > cd
73

2. Record the non event data portion of the home directory onto tape:
> tar -cvf /dev/st1 .

--exclude=./experiment --exclude=./stagearea

3. Change directories to the projects/e##### space provided by the computer group:
> cd /projects/e#####
4. Read the files from the tape: > tar -xvf /dev/st1 .
5. When the reading is completed, check that the entire /user/e##### directory was
copied, except for the event files and links to them.

6. Eject tape and label it with the date, experiment number, and directory copied.

Now you have your links pointing to the proper place, and SpecTcl will still be able to
read the files once the experimental directory has been deleted. There may still be other links,
however, that need to be updated. It is a good idea to scan through all of the subdirectories
in your projects space to look for any links still pointing to the experimental directory. If
you should find any, update them following the same procedure outlined above. Since the
subdirectory structure stays the same when you copy all of the files from the experimental
home directory to projects, you should be able to update the links by simply replacing the
/user/e##### portion of the file name with /projects/e#####.

A.10.4

Executables

Decide where you want the executable for spectcl to go, and make sure to add it to your
path. ROOT does not need this, and since all other programs are unusable without a daq,
there is no need to change the environment variables.
74

A.10.5

Managing the changes to the major programs

Before you close the account and tell the computer department that is is okay the delete the
account, it is necessary to check to see if any changes to the major codes (readout, spectcl,
ROOT) deserved to be put into the cvs repository for use in future experiments.
For each major code (readout mona, readout lisa, readout-sweeper caesar, n2analysis,
spectcl, evtMerge, and tools), you will need to go into the directory that houses the code
and query the status of the files within the repository. For example, to do this for spectcl,
type:

>cd ~/spectcl
>cvs status > test.txt
This will put the output of the cvs status command in the file test.txt (which will be a
temporary file you will delete afterward). Any files that are shown to be Locally Modified (or
even perhaps unknown to the repository) will need to evaluated indiviually if these changes
need to commited to the repository. Since this is a complex and detail oriented process, I
will omit directions on how to manage change with cvs. Consult any guide on using cvs and
consult the groups experts on the program projects before committing any changes.
You will also need to examine if any files that are not in repository (some configuration or
utility files) need to be reused in future experiments. Copy any of these into an appropriate
folder in monasoft that will be copied into the future experiments.

75

BIBLIOGRAPHY

76

BIBLIOGRAPHY

[1] G. Audi et al. Nucl. Phys. A, 729:3, 2003.
[2] T. Baumann et al. Nucl. Instr. Meth. A, 543:517–527, 2005.
[3] M. Belleguic et al. Nucl. Phys. A, 682:136c – 142c, 2001.
[4] M. D. Bird et al. IEEE Trans. Appl. Supercond., 15:1252, 2005.
[5] B. A. Brown. private communication, 2009.
[6] B. A. Brown. Physics, 3(104), 2010.
[7] B. A. Brown and W. D. M. Rae. NuShell@MSU, 2007.
[8] B. A. Brown and B. H. Wildenthal. Ann. Rev. Part. Nucl. Sci., 38:29, 1988.
[9] Z. Elekes et al. Phys. Rev. Lett., 98:102502, 2007.
[10] N. H. Frank. Spectroscopy of Neutron Unbound States In Neutron Rich Oxygen Isotopes.
PhD thesis, Michigan State University, 2006.
[11] O. Haxel, J. H. D. Jensen, and H. E. Suess. Phys. Rev., 75:1766, 1949.
[12] C. R. Hoffman. investigation of the Neutron-rich Oxygen Isotopes at the Drip Line.
PhD thesis, Florida State University, 2009.
[13] C.R. Hoffman et al. Phys. Rev. Lett., 100:152502, 2008.
77

[14] C.R. Hoffman et al. Phys. Lett. B, 672:17–21, 2009.
[15] R. Kanungo et al. Phys. Rev. Lett., 102:152501, 2009.
[16] R. Kr¨cken. Introduction to shell stucture in exotic nuclei. Contemporary Physics,
u
52(2):101–120, 2010.
[17] R. D. Lawson and J. L. Uretsky. Phys. Rev., 108:1300, 1957.
[18] K. Makino and M. Berz. Nucl. Instr. Meth. A, 558:346, 2005.
[19] Maria G. Mayer. Phys. Rev., 74(3):235–239, 1948.
[20] Maria G. Mayer. Phys. Rev., 75:1969–1970, 1949.
[21] D. J. Morrissey et al. Nucl. Instr. Meth. B, 204:90, 2003.
[22] Alex C. Mueller and Bradley M. Sherrill. Annu. Rev. Nucl. Part. Sci., 43:529–583, 1993.
[23] W. A. Peters. Study of Neutron Unbound States Using the Modular Neutron Array
(MONA). PhD thesis, Michigan State University, 2007.
[24] C. Rodr´
ıguez-Tajes et al. Phys. Rev. C, 82(024305), 2010.
[25] H. Scheit. Simple track for MoNA, 2006.
[26] A. Schiller et al. Phys. Rev. Lett., 99:112501, 2007.
[27] D. Sohler et al. Phys. Rev. C, 77:044303, 2008.
[28] O. Sorlin and M.-G. Porquet. Prog. Part. Nuc. Phys., 61:602–673, 2008.
[29] M. Stanoiu et al. Phys. Rev. C, 69:034312, 2004.
[30] M. Stanoiu et al. Phys. Rev. C, 78:034315, 2008.
[31] I. Talmi and I. Unna. Phys. Rev. Lett., 4:469, 1960.
[32] C. Thibault et al. Phys. Rev. C, 12(2):644–657, 1975.
78

[33] P. G. Thirolf et al. Phys. Lett. B, 485:16, 2000.
[34] E. K. Warburton, J. A. Becker, and B. A. Brown. Phys. Rev. C, 41(3):1147–1166, 1990.
[35] E. K. Warburton and B. A. Brown. Phys. Rev. C, 46(3):923–944, 1992.
[36] B. H. Wildenthal. Prog. Part. Nuc. Phys., 11:5, 1984.
[37] B. H. Wildenthal and W. Chung. Phys. Rev. C, 22(5):2260–2262, 1980.

79