DEVELOPMENT AND COMMISSIONING OF A PHOTON
DETECTION SYSTEM FOR COLLINEAR LASER
SPECTROSCOPY AT NSCL
By
Sophia Vinnikova

A THESIS
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
Masters of Science
Chemistry
2011

ABSTRACT
DEVELOPMENT AND COMMISSIONING OF A PHOTON DETECTION
SYSTEM FOR COLLINEAR LASER SPECTROSCOPY AT NSCL
By
Sophia Vinnikova

A photon detection system has been developed for collinear laser spectroscopy
measurements at the Beam COoling and COllinear LAser spectroscopy (BECOLA)
facility at NSCL. The system is based on a design from the University of Mainz, and
relies on an ellipsoidal reﬂector to transport ﬂuorescent light to a photomultiplier
tube. “Blue” and “red” photon detection systems were designed to cover the desired
wavelength ranges. Simulations of the detection system were performed using FRED
optical engineering software, and commissioning measurements were carried out with
a 15 keV Ca+ beam to evaluate performance and to optimize the“blue” photon detection system for the BECOLA facility. Signal to noise measurements with diﬀerent
arrangements of beamline apertures showed that apertures are necessary both upstream and downstream of the reﬂector to reduce scattered laser light. Signal to noise
ratio measurements for various iris sizes at the second focal plane of the reﬂector
showed that the signal to noise ratio increased with iris area less than 12.7 mm in diameter, and then saturated with increasing size. The signal distribution was found to
peak at the second focal point of the ellipsoidal reﬂector, which was well reproduced
by the simulations. The background distribution was uniformly distributed over the
area of the second focal plane that was measured. Overall sensitivity of 1 photon per
128 ions was achieved with the present “blue” photon detection system.

Copyright by
SOPHIA VINNIKOVA
2012

ACKNOWLEDGMENTS

This thesis could not have been completed without the help of Photon Engineering,
who gave me two years of gratis licences for FRED.
This thesis was also made possible with the help of many people. I would like to
thank my adviser, Paul Mantica for the never ending questions during group meeting,
for teaching me how to give presentations, and for the knowledge that he imparted
onto me over the past two and a half years. Secondly I am eternally grateful to the
BECOLA group, thank you Kei Minamisono, Andrew Klose and Anthony Schneider
for the long days (and nights) spent in the experimental area as we searched for a
resonance and for the even longer hours spent on my measurements after the resonance
was observed. Additionally I want to thank Kei for letting me bounce ideas oﬀ him
during the simulation and design processes.
I want to also thank my collaborators from the University of Mainz, especially
Michael Hammen and J¨rg Kr¨mer for answering all of my questions about their
o
a
photon detection systems.
Finally, I want to thank my friends and family for sticking by me, especially during
the past semester. Thank you Mamma, Pappa, Asia and Inessa for the weekly skype
calls, they brightened up my saturdays and got me out of bed before noon. Thank you
Merle, Louisa, Ashley, and Jenn for the support and friendship that you have given
me over the years. Also I want to give thanks to Atos and Konty, whose unconditional
love always made the long drives home worth it.

iv

Contents
1 Introduction
1.1 Motivation . . . . . . . . . . . . . . . . . . . .
1.2 Atomic Spectroscopy . . . . . . . . . . . . . .
1.2.1 Line Broadening . . . . . . . . . . . .
1.3 Hyperﬁne Structure . . . . . . . . . . . . . . .
1.4 Collinear Laser Spectroscopy . . . . . . . . . .
1.5 BECOLA Facility . . . . . . . . . . . . . . . .
1.5.1 Laser System . . . . . . . . . . . . . .
1.5.2 Beam Line Diagnostics and Transport
1.5.3 Oﬄine Ion Source . . . . . . . . . . . .
1.5.4 Photon Detection System . . . . . . .
1.5.5 DAQ and Controls . . . . . . . . . . .
2 Simulation of and Extensions to the Mainz
tem
2.1 Tools . . . . . . . . . . . . . . . . . . . . . .
2.1.1 Geometry Used in FRED . . . . . . .
2.1.2 Inputs and Constraints . . . . . . . .
2.2 Typical Output . . . . . . . . . . . . . . . .
2.3 Optimizing System Geometry . . . . . . . .
2.3.1 Reﬂector Size . . . . . . . . . . . . .
2.3.2 Irises at the Second Focal Plane . . .
2.3.3 Beam line Apertures . . . . . . . . .
2.4 Summary of Results . . . . . . . . . . . . .
3 Design of the BECOLA Photon Detection
3.1 Design . . . . . . . . . . . . . . . . . . . .
3.2 Interaction Region and Light Collection . .
3.2.1 Ellipsoidal Reﬂector Speciﬁcations .
3.2.2 Ellipsoidal Reﬂector Material . . .
3.3 Stray Light Reduction . . . . . . . . . . .
3.3.1 Beam line Apertures . . . . . . . .
3.3.2 Irises for the PMT . . . . . . . . .
3.4 PMT Selection . . . . . . . . . . . . . . .
3.5 “Blue” Light Detection System . . . . . .
3.6 Red Light Detection System . . . . . . . .

v

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

1
1
3
5
6
7
9
11
12
13
15
15

Photon Detection Sys.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

System
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .

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

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

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

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

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

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

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

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

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

.
.
.
.
.
.
.
.
.

17
17
18
20
21
22
22
24
25
27

.
.
.
.
.
.
.
.
.
.

29
29
30
31
31
33
34
35
36
38
40

4 Calcium Ion Beam Spectroscopy and Commissioning of the
Photon Detection System
4.1 Commissioning Introduction . . . . . . . . . . . . . . . . . .
4.1.1 Calcium Ion Spectroscopy . . . . . . . . . . . . . . . .
4.1.2 Commissioning Set-up . . . . . . . . . . . . . . . . . .
4.2 Stray Light Characterization . . . . . . . . . . . . . . . . . . .
4.2.1 Beam Line Preparation Following Simulation Results .
4.3 Commissioning Measurements and Comparison to Simulations
4.3.1 Aperture Conﬁgurations . . . . . . . . . . . . . . . . .
4.3.2 PMT Iris Area . . . . . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

43
43
43
44
45
45
49
49
50

5 Results and Discussion
5.1 Line Shape . . . . . . . . . . . . . . . . . . . . .
5.1.1 Typical Resonances . . . . . . . . . . . . .
5.1.2 Gated Resonance Spectra . . . . . . . . .
5.1.3 S/N Extraction . . . . . . . . . . . . . . .
5.2 Beam Line Aperture Conﬁguration Measurements
5.3 PMT Iris Area Measurements . . . . . . . . . . .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

51
51
51
56
59
60
62

.
.
.
.

.
.
.
.
.
. .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

“Blue”

6 Summary and Outlook

66

Bibliography

68

vi

List of Tables
5.1
5.2

Primary data for extraction of S/N as a function of beamline aperture
conﬁguration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

61

Primary data for extraction of S/N as a function of PMT iris open area. 64

vii

List of Figures
1.1

A three-level energy diagram. Possible transitions from the highest
level are depicted by solid arrows. For interpretation of the references to
color in this and all other ﬁgures, the reader is referred to the electronic
version of this thesis. . . . . . . . . . . . . . . . . . . . . . . . . . . .

4

Typical Lorentzian, Gaussian and Voigt functions used for line shape
analysis, shown as function of frequency. . . . . . . . . . . . . . . . .

6

1.3

Hyperﬁne structure of 43 Ca . . . . . . . . . . . . . . . . . . . . . . .

7

1.4

The schematic drawing of the BECOLA beam line

. . . . . . . . . .

10

1.5

Photograph of the laser optics used to tune the laser light into the
BECOLA beam line. . . . . . . . . . . . . . . . . . . . . . . . . . . .

12

Exploded view of the Colutron Plasma Ion Source components. The
charge holder is not shown in this view. . . . . . . . . . . . . . . . . .

13

Schematic diagram of the modiﬁed version of the Colutron ion source
used at BECOLA . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

14

The ﬁlament of the Colutron ion source. The ﬁlament is kept in place
by bending the contact wires and using ceramic spacers. . . . . . . .

15

1.9

BECOLA DAQ and controls system layout . . . . . . . . . . . . . . .

16

2.1

Engineering drawing of the geometry imported into FRED for ray trace
simulations of the detection system. . . . . . . . . . . . . . . . . . .

19

2.2

Simulated signal and stray distributions at the second focal plane . .

22

2.3

Relative S/N values obtained from the simulation for 397 nm light
collected at the second focal plane for three reﬂector collection lengths. 24

2.4

Simulated relative S/N ratio for a variety of PMT iris sizes . . . . . .

2.5

Inner drift tube aperture conﬁgurations that were simulated using FRED. 26

2.6

Simulated relative S/N ratio for a variety of beam line aperture conﬁgurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.2

1.6
1.7
1.8

viii

25

27

3.1

Geometrical focusing properties of ellipsoidal reﬂector . . . . . . . . .

30

3.2

Geometrical properties of the BECOLA ellipsoidal reﬂector . . . . . .

31

3.3

Reﬂectivity curves of the metals considered for the ellipsoidal reﬂector

32

3.4

Photograph of the ellipsoidal reﬂector after hand polishing. . . . . . .

33

3.5

Schematic view of the placement of a pair of beam line apertures. . .

34

3.6

Photograph of two 7 mm inner diameter beam line apertures after
being painted with graphite paint to be conductive. . . . . . . . . . .

35

Photographs of the 12.7 mm (machined) in diameter circular iris and
the 6.35 × 25.4 mm (hand-made) rectangular iris inside the teﬂon iris
mounting system. The PMT side of the teﬂon holder is shown on the
left and the side adjacent to the inverted viewport window, where the
iris is mounted, is shown on the right. . . . . . . . . . . . . . . . . . .

36

3.8

Mechanical design of the blue photon detection system . . . . . . . .

38

3.9

Photograph of the vacuum chamber with only the inner and outer
drift tubes in place. The beam direction is into the paper through the
aperture inside the inner ground drift tube. . . . . . . . . . . . . . . .

39

3.10 Photograph of the “blue” photon detection system PMT, PMT enclosure, and teﬂon rings used to center the PMT inside the inverted
viewport. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

40

3.11 A mechanical design of the red photon detection system

. . . . . . .

41

3.12 A photograph of the “red” photon detection system PMT enclosure
attached to the photon detection system vacuum chamber. . . . . . .

42

4.1

Electronic structure of 40 Ca . . . . . . . . . . . . . . . . . . . . . . .

44

4.2

Locations where black aluminum foil was placed to reduce stray light
are marked with pink arrows. (a) The reﬂector outer surfaces and drift
tube gaps. (b) The inverted viewport and inverted viewport ﬂange. (c)
The reﬂector ﬂange. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

46

Locations on the (a) Bottom of the reﬂector; (b) outside of reﬂector;
and (c)reﬂector ﬂange where black foil was placed to reduce surface
reﬂectance are shown as hatched areas. . . . . . . . . . . . . . . . . .

46

4.4

Photograph of the ellipsoidal reﬂector covered in black aluminum foil.

47

4.5

The black aluminum foil was placed in the gap between the outer and
inner ground drift tubes on both the upstream and downstream sides
of the reﬂector. Crosshatching shows the locations that were covered
with black foil to decrease reﬂectivity. . . . . . . . . . . . . . . . . . .

47

3.7

4.3

ix

4.6

Hatched areas show the locations on the (a) Inverted viewport ﬂange.
(b) Inverted viewport tube, window ﬂange, and ﬂange where black foil
was placed to reduce surface reﬂectance. . . . . . . . . . . . . . . . .

48

Photograph of the inverted viewport window ﬂange covered in black
aluminum foil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

48

Schematic represemtation of the inner drift tube aperture conﬁgurations that were tested in BECOLA. (a)-(e) correspond to A-E respectively, (f) is the Mainz system aperture conﬁguration, which was not
tested. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

49

Typical 40 Ca+ resonance and ﬁtted curve used to extract S/N ratio
for evaluation of photon detection system performance. . . . . . . . .

52

Typical 40 Ca+ resonance curves collected with the observed 8 V peak
to peak ripple in the FUG DC supply, with no electronic gates. The
three ﬁgures represent diﬀerent beam optics settings. . . . . . . . . .

54

Snapshots of an oscilloscope trace showing the ripple of the FUG power
supply. The electronic gates that were applied to reduce the impact of
the oscillation on the resonance curves are also shown, as well as the
coincidence relationship established by the DAQ for counting PMT
signals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

55

Typical 40 Ca+ resonance curve collected with an electronic gate on the
upper ripple peak of the 15 kV DC acceleration power supply. . . . .

56

Typical 40 Ca+ resonance curve collected with electronic gate on the
lower ripple peak. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

57

Typical 40 Ca+ resonance curves achieved for photon detection system
tests with electronic gate on upper ripple peak. . . . . . . . . . . . .

58

Typical 40 Ca+ resonance curves achieved with electronic gate on upper
ripple peak with the low-energy tail ignored by the ﬁt. The line is a
Voigt proﬁle ﬁt. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

58

S/N extraction from the 40 Ca+ D1 ﬂuorescence peak. Background (B)
was deﬁned as the non-resonant counts above zero, the signal (S) was
the amplitude of the peak above the background, and the noise (N)
was the ﬂuctuation in the background. . . . . . . . . . . . . . . . . .

59

Experimental S/N as a function of beam line aperture conﬁguration.
S/N is corrected by beam current and laser power. . . . . . . . . . . .

61

5.10 Initial experimental S/N as a function of PMT iris area. S/N is corrected by beam current and average laser power. . . . . . . . . . . . .

62

4.7
4.8

5.1
5.2

5.3

5.4
5.5
5.6
5.7

5.8

5.9

x

5.11 Laser power and ﬂuctuation measured during the second run of the
PMT iris measurements, note the suppressed zero on vertical scale. .

63

5.12 Experimental S/N as a function of PMT iris opening area. The S/N
was corrected by beam current, laser power and laser power ﬂuctuation. 64

xi

Chapter 1
Introduction
1.1

Motivation

The main objectives of the BEam COoler and COllinear LAser spectroscopy (BECOLA)
facility at the National Superconducting Cyclotron Laboratory (NSCL) are two-fold:
(1) to measure hyperﬁne atomic structure of rare isotopes in order to deduce nuclear
properties such as the electric quadrupole moment, the magnetic dipole moment,
ground state nuclear spin, and the change in nuclear charge radius, and (2) to produce polarized nuclei for fundamental symmetry studies.
The hyperﬁne structure of an atom can be measured using atomic spectroscopy,
where the absorption or emission of energy induces transitions between electronic
states. The frequencies of photons absorbed or emitted is related to the diﬀerence in
energy of the initial and ﬁnal electronic states. Atomic ﬂuorescence typically occurs
with relatively little time delay; therefore the system needed to detect photoemission
must be well matched to the source of the photons. The observed ﬂuorescence is typically at the same frequency as the frequency used to induce the transition, for atomic
spectroscopy. A photon detection system must be able to observe the ﬂuorescence
in the presence of a potentially large background. Most atomic spectroscopy experiments use lasers to induce the electronic excitations. Lasers have the advantages of

1

high power, continuous wave or pulsed operation, tunability, and a very collimated
beam.
Lasers have been applied to hyperﬁne structure measurements of unstable isotopes
for some 30 years. The most widely-applied technique is collinear laser spectroscopy.
Such laser spectroscopy experiments have typically been performed at Isotope Separator Online (ISOL) type facilities, where the ion/atom beam is chemically extracted
from the target due to strict requirements on the ion-beam [1]. Chemical extraction
is diﬃcult for some elements, such as transition metals, due to their high melting
points.
The BECOLA facility will make use of isotopes produced by projectile fragmentation at National Superconducting Cyclotron Laboratory (NSCL). Projectile fragmentation coupled with in-ﬂight separation is a “chemistry independent” method
for isotope production that, coupled to beam thermalization, promises to extend the
number of elements available for study by laser spectroscopy. NSCL has pioneered the
thermalization of projectile fragments in helium buﬀer gas to produce high-quality,
low-energy beams [2]. The BECOLA system will be placed downstream of the NSCL
gas stopping area, and the low-energy beams will be available for a wide range of
chemical elements to open new opportunities for laser spectroscopy studies [3]. Experiments with short-lived transition metals and other rare isotopes than have not
been produced at other facilities [3] will be possible.
A key component of BECOLA is the ﬂuorescent photon counting system, that
was designed to meet the following requirements: (1) compatibility with a collinear
geometry of the ion and laser beams; (2) eﬃcient collection of ﬂuorescent light, while
blocking out stray light; (3) applicability over the wavelength ranges available from
the existing BECOLA laser system; (4) ability to physically ﬁt into the existing 8
inch diameter 6-way cross; and (5) applicability to both ion and atom beams.
This thesis describes the development of the photon detection system for blue light
at BECOLA. The remainder of this chapter will provide background information on

2

collinear laser spectroscopy and on the experimental set-up at NSCL. The remaining
chapters will discuss in detail the photon detection system model, design, fabrication,
and commissioning tests.

1.2

Atomic Spectroscopy

Spectroscopy is the study of the absorbed or emitted energy to identify the transitions
between states of a quantum mechanical system. Atomic spectroscopy generally deals
with transitions among the electronic states of an atom. The frequency (ν) at which
energy is absorbed or emitted is related to the diﬀerence in the energies of levels of
the initial and ﬁnal states. In the absence of recoil corrections, the energy diﬀerence
between levels, ∆E, is given by:

∆E = hν = |E2 − E1 |

(1.1)

where E2 and E1 are the energies of the ﬁnal and initial states, respectively, and h is
Plank’s constant.
Atomic states of one electron systems can be described by the quantum numbers
of that electron nljmj , where n is the principal quantum number, l is the orbital
angular momentum quantum number, j is the total angular momentum [with spin1
1
orbit coupling where j = l + s = l ± 2 , with s being the spin of the electron (± 2 )],

and mj is the projection of j onto the quantization axis. Electrons with the same n
belong to the same shell, and electrons with the same n and j belong to the same
subshell.
In many-electron systems, the angular momenta of the electrons have to be combined. In the LS coupling scheme, the total angular momentum, L, and the total
electron spin, S, come from the coupling of the orbital angular momenta (L =

i li )

to the total angular momentum and spins of the electrons are coupled to a total electron spin (S =

i si ) respectively. Total spin J comes from the spin-orbit coupling

3

between S and L (J = L + S). Selection rules for allowed electric dipole (E1) transitions are: ∆l = ±1, ∆L = 0, ±1, ∆J = 0, ±1, and ∆S = 0. ∆l is the change in the
angular momentum of the electron undergoing the transition, and it must change by
one because the photon carries one unit of angular momentum.

(2)

Γ23
Γ21

(3)

Γ31
(1)
Figure 1.1: A three-level energy diagram. Possible transitions from the highest level
are depicted by solid arrows. For interpretation of the references to color in this and
all other ﬁgures, the reader is referred to the electronic version of this thesis.
As a simple example, a three-level energy diagram with an excited state, ground
state, and intermediate state, denoted as (1), (2), and (3), respectively, is depicted in
Fig. 1.1. The decay rates between electronic states are denoted by Γul and are shown
by the solid arrows. The total decay rate from state (2), Γ2 , needs to account for the
decay rates of all possible paths from (2), and is related to the lifetime (τ ) of level
(2) and the natural line width ∆νnat by the expression :
1
= 2π∆νnat .
Γ2 = Γ21 + Γ23 =
τ2

(1.2)

Eq. 1.2 can be rearranged to solve for the natural linewidth [4];

∆νnat =

1
2πτ

(1.3)

The observed line shape for the ﬂuorescence at frequency ν for an induced resonance frequency ν0 is a Lorentzian proﬁle with a full width at half maximum (FWHM)

4

of the natural linewidth, and a centroid of ν0 [4]. The observed linewidth of a transition
is generally broader than the natural linewidth; as a result of various experimental
broadening eﬀects attributed to the velocity distribution of the ion/atom beam as
discussed below.

1.2.1

Line Broadening

The two main types of line broadening are labeled homogeneous and inhomogeneous.
Homogeneous line broadening occurs with an ensemble of atoms with uniform response [5, 6]. Inhomogeneous broadening comes from the transition frequencies in a
set of atoms with non-uniform response.
Doppler broadening is an example of inhomogeneous broadening caused by the
Doppler eﬀect on an ensemble of atoms moving with a random distribution of thermal
velocities [5]. The line shape of inhomogeneous broadening gives a Gaussian line
shape [5, 6].
Time-transit broadening due to the ﬁnite lifetime of the transition and power
broadening due to saturation of the transition are examples of homogeneous line
broadening, which gives a Lorentzian line shape [4, 7].
The experimentally-observed transition line shape can be modeled by a Voigt function, or Voigt proﬁle, which is a convolution of a Gaussian with a Lorentzian. The
linewidths of the Gaussian and Lorentzian components in the Voigt proﬁle permit a
quantitative means to deduce the contributions to the line shape due to homogeneous
and inhomogeneous broadening [5]. The diﬀerences in line shape of Lorentzian, Gaussian, and Voigt proﬁles are depicted in Fig. 1.2. Notice that the Lorentzian function
is more sharply peaked but yet is wider at the base than the Gaussian function.

5

Lorentzian

Gaussian

Voigt

1.0
0.9

Arbitrary Units

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0

20

40

60

80

100

Frequency (MHz)

Figure 1.2: Typical Lorentzian, Gaussian and Voigt functions used for line shape
analysis, shown as function of frequency.

1.3

Hyperﬁne Structure

The electromagnetic interaction of electrons with the multipole moments of the nucleus gives rise to atomic hyperﬁne structure. Fine structure of an atom comes from
spin-orbit coupling between the total electron orbital angular momentum (L) and
the total electron spin (S) and is described by the total electron angular momentum
quantum number, J, where J = L + S. The total angular momentum of the electrons
must be then coupled with the nuclear spin (IN ) to give the hyperﬁne energy levels,
denoted by the total angular momentum of the atom or ion, F , where F = J + IN [8].
The ﬁne and hyperﬁne splitting of 43 Ca is shown in Fig. 1.3 as an example. Fine
structure energy levels are shown in column (B) and hyperﬁne structure levels are
shown in column (C). The two allowed transitions (absorption) between the ﬁne
levels are the D1 (2 S1/2 →2 P1/2 ) and the D2 lines (2 S1/2 →2 P3/2 ). The allowed
transitions between the hyperﬁne energy levels, according to the rules described in

6

Sec. 1.2, in D1 are: F=3 → 3, F=3 → 4, F=4 → 3, and F=4 → 4. Whereas, in D2,
they are: F=3 → 2, 3, and 4, and F=4 → 3, 4, and 5.

(A)

4p

(B)

(C)

2P
5
3/2 D2 2
393.4 nm
4
2P
3
1/2

370 MHz

582 MHz

D1
396.9 nm
4s

2S
1/2

4
3.2 GHz
F=3

7
Figure 1.3: Fine and hyperﬁne structure of 43 Ca, nuclear spin IN = 2 . D1 transitions
are denoted by solid arrows and D2 are denoted by dashed arrows.

Fine energy level splitting is of order a few THz, while hyperﬁne splitting is of order
hundreds of MHz. A precision method is required to resolve the hyperﬁne structure
levels. Linewidth broadening can cause the hyperﬁne levels to overlap and not be
observeable.

1.4

Collinear Laser Spectroscopy

Collinear laser spectroscopy will be applied at NSCL at the new BECOLA facility,
and the photon detection system is the key measurement element for obtaining the
resulting hyperﬁne spectrum. The collinear geometry of the particle beam and the
laser beam will allow scanning of the velocity of the ion beam rather than the laser
frequencyto bring the system into resonance. The velocity of the beam can be adjusted
through application of an acceleration or deceleration voltage in the detection region.
The collinear overlap between the laser and the ion or atom beam (for a distance of
typically 1 m or greater) provides a long interaction time between the beams.

7

Velocity Bunching
Ion post acceleration allows increased sensitivity through velocity bunching [9]. The
initial kinetic energy spread, ∆Ek , of an ion beam remains unchanged after electro1
static acceleration [9, 10]. As the kinetic energy of the beam (Ek = 2 mv 2 ) increases,

the velocity also increases, so that the relative velocity spread (∆v) is smaller at
higher ion velocities after acceleration for a given ∆Ek (∆Ek = mv∆v). Velocity
bunching allows the Doppler broadening to be reduced to the order of the natural
line width, so that the sensitivity and precision of spectroscopy measurements are
improved because the ﬂuorescence peaks will have a smaller FWHM.
Doppler Shift
Atoms and ions traveling at a velocity (v = βc) experience the Doppler-shifted laser
frequency at ν ′ from laser light with a frequency νL as;
1 − β2
δν ′ = νL
,
1±β

(1.4)

where the “+” sign is applied for a collinear beam geometry and the “-” sign for
anti-collinear beams. Here β is the relativistic velocity function given by:

β=

v
=
c

1−

m2 c4
,
(eU + mc2 )2

(1.5)

where c is the speed of light, m is the particle’s rest mass, and U is the accelerating
potential.
The laser frequency νL applied in collinear laser spectroscopy is typically set to a
Doppler-shifted resonance frequency of the particle, ν ′ , so that a transition resonance
frequency ν0 will be absorbed by the atom.

8

Diﬀerential Doppler Shift
The diﬀerential doppler shift (δνdif f ) is the diﬀerential change in the frequency in
the particle’s rest frame per change in the acceleration voltage;

δνdif f =

ν0
mc2 (e +

e(mc2 +eU )
)
eU (2mc2 +eU )

.

(1.6)

The diﬀerential doppler shift gives the relationship between the scanning voltage
and the frequency observed by the particle, which allows extraction of physical quantities such as the frequency of a resonance and the splitting between hyperﬁne energy
levels.

1.5

BECOLA Facility

BECOLA is one of the end stations constructed for experiments with thermalized
beams at NSCL [3]. At BECOLA, ﬂuorescence from the interaction between laser
light and an ion/atom beam will be collected by a photon detection system to measure
hyperﬁne structure of that ion/atom.
A schematic drawing of the BECOLA facility is shown in Fig. 1.4. Laser light is
introduced into the beam line through a laser window in the switchyard, shown on the
left-side of the ﬁgure. A second laser window is located at the end of the beam line.
Light transmitted through the second window is stopped in an external beam dump.
An ion beam with a kinetic energy up to 60 keV can be introduced into the switchyard
using an electrostatic steerer and further transported to the beam interaction region
by a spherical electrostatic bend. Collimator systems and Faraday cups can be placed
in two beam observation boxes (BOBs) to measure the beam current and to ensure
overlap of the laser and ion/atom beams as they pass through the interaction region.
The ion beam can also be neutralized in a charge exchange cell (CEC) to generate

9

an atom beam, if desired [11]. Fluorescence light emitted after the interaction of the
laser light with the ion/atom beam is measured by the photon detection system.

Beam Observation Boxes
Switchyard

Laser In

Ion Beam In

Charge Exchange Cell

Photon Detection
System

Figure 1.4: The schematic of the BECOLA beam line. The laser beam (red line) and
the ion beam (blue line) are collinearly overlapped after the switchyard.
An oﬄine ion source was developed to commission the photon detection system
and other components of BECOLA. A stable beam of singly-charged calcium ions was
produced by the oﬄine ion source and deﬂected into the BECOLA beam line. Electrostatic quadrupole lenses were used to control the divergence of the ion beam after
the switchyard. The ion beam current entering the interaction region was registered in
a Faraday cup that was located in BOB1 and connected to an electrometer. Fluorescence light was collected, as a function of the scanning (i.e. acceleration) voltage, in
the photon detection system, that was developed as the main objective of this work.
The scanning voltage was applied to all of the components of the photon detection
system, since only ion beams were used for the commissioning of the photon detection
system reported herein. A Faraday cup in BOB2 located downstream of the photon
detection system was used to measure the ion beam that was transmitted through
the interaction region. The transmission of laser light through the interaction region
was monitored by a power meter located at the beam dump position immediately
downstream of the second laser window.

10

1.5.1

Laser System

The laser system used to excite the ﬂuorescent levels consists of a continuous wave
Ti:Sapphire (Ti:Saph) ring laser pumped by a 18 W pump laser operated at 532 nm.
The Ti:Saph ring laser has an output power of about 2 W in the wavelength range
of 700 - 1000 nm (“red” light). The infrared light from the laser can be used directly
for spectroscopy experiments, or it can be transported to a frequency doubler, which
produces second harmonic laser light with maximum power of about 200 mW in the
wavelength range of 350 - 500 nm (“blue” light) depending on the input wavelength
and power from the Ti:Saph. Both the ring laser and frequency doubler were purchased from a commercial vendor, Spectra Physics. The laser light intended for the
spectroscopy measurements is passed through a half-wave plate and a polarization
cube to set the polarization axis of the light. The polarization cube can also be used
to adjust the laser power transported to the experiment. A ﬁber optic cable is used
to transport the prepared laser light over a distance of 26 m to the BECOLA beam
line. Fiber optic couplers are used on each end of the ﬁber optic cable to eﬃciently
transfer the laser light between air and the medium of the ﬁber optic cable. An image
of the output coupler and optics table immediately outside the BECOLA switchyard
is shown in Fig. 1.5. The laser light coming out of the ﬁber coupler is focused by a
lens and reﬂected by two mirrors to a second lens mounted on an adjustable support
frame. These two lenses act as a telescoping system and are used to focus and align
the laser beam to the ion beam optical axis. A 50:50 laser beam splitting cube is
placed after the lens system to split half of the beam to a power meter to monitor the
laser power. Two mirrors mounted on ﬁne adjustment mounts are used to align the
laser light so that it passes through the center of two pre-aligned irises. The centers of
the iris openings are set beforehand to correspond to the beam line optic axis using
an optical telescope.

11

Mirrors

Beam Block
Shutter

Fiber Optic
Conduit

Irises
Ion
Beam
Laser
Beam

Telescoping Fiber Optic
Output
Lens System
Coupler

50:50 Laser
Beam Splitting
Cube

Alignment
Mirrors

Figure 1.5: Photograph of the laser optics used to tune the laser light into the
BECOLA beam line.

1.5.2

Beam Line Diagnostics and Transport

As noted above, two BOBs, which contain detectors for laser and ion/atom beam
diagnostics, are located along the BECOLA beam line. One is located upstream (∼1
m) of the photon detection system and the second is downstream (∼1 m) of the photon
detection system. The diagnostics in each BOB included a Faraday Cup and beam
collimator system. The Faraday cups are used to measure the ion current. The beam
collimator system is used to deﬁne both the ion beam size and to ensure alignment
of the laser beam with the ion beam optical axis. The beam collimators in the BOBs
can also be used to deﬁne beam size using ﬁxed openings of diameter: 3, 5 and 7 mm.
Each collimator is mounted on a linear actuator, which allows the transported beams
to be manipulated without breaking the vacuum.

12

1.5.3

Oﬄine Ion Source

A Colutron plasma oﬄine ion source [12] was used to produce stable calcium beams
for the commissioning of the “blue” photon detection system. The ion source was
placed upstream of the switchyard. The major components of the source are depicted
schematically in Fig. 1.6. Calcium ion beams were produced using calcium metal
placed inside a 1.5-inch long, 1/8th inch diameter ceramic tube that served as the
charge holder, not shown. The charge holder was recessed in the boron-nitride chamber while the plasma was conditioned. A voltage was applied to the ﬁlament to start
the plasma. Once the pressure of the source stabilized with the ﬁlament running, the
charge holder was pushed towards the ﬁlament with a push rod, with the goal of
evaporating and then ionizing the calcium metal that was placed inside the charge
holder. The charge holder was inserted at three diﬀerent positions relative to the
boron-nitride chamber, as depicted in Fig. 1.7. The voltage applied to the ﬁlament,
typically 7-9 V, was found to be suﬃcient to evaporate the calcium metal, which was
then ionized in the plasma between the ﬁlament and anode.

Cap Snap
Ring
Anode
Gas Inlet
Tube

Filament

Contact Pins

Cap
Insert Ring
Contact Wire

Boron Nitride
Chamber

Anode Snap
Ring

Figure 1.6: Exploded view of the Colutron Plasma Ion Source components. The charge
holder is not shown in this view.
The ﬁlament used to heat the charge holder and generate the plasma was held
inside the boron-nitride chamber by two ceramic spacers. The contact wires were also
13

Gas Inlet
Tube

Push
Rod

Quartz
Wool

Contact
Pins

Charge
Holder

Anode
Filament

Cap
Calcium Contact Boron Nitride
Metal Wires
Chamber
(a)

(b)

(c)
Figure 1.7: Schematic diagram of the modiﬁed version of the Colutron ion source used
at BECOLA. (a) The ion source with the ceramic charge holder and metal push rod
recessed before entry into the chamber. (b) The ion source with the charge holder
pushed to the beginning of the ﬁlament. (c) The ion source with the charge holder
pushed in to the end of the ﬁlament.
bent to keep the ﬁlament aligned and in place, as depicted in Fig. 1.8.
Calcium ions oﬀered an ideal system for initial commissioning tests of BECOLA
and the “blue” photon detection system . The D1 and D2 lines of Ca+ , with wavelengths of 396.9 and 393.4 nm, respectively, are readily accessible with the present
laser system. In addition, the calcium atom also has ﬁne structure transitions in the
“blue” region of the electromagnetic spectrum, and the same photon detection system
could be applied to perform laser spectroscopy on both calcium ions and atoms.

14

Bent
Contact
Wires

Ceramic
Spacers

Figure 1.8: The ﬁlament of the Colutron ion source. The ﬁlament is kept in place by
bending the contact wires and using ceramic spacers.

1.5.4

Photon Detection System

The photon detection system that was developed as part of this work was used to
collect ﬂuorescence photons emitted from the 40 Ca+ ion beam after interaction with
laser light. It consisted of an ellipsoidal reﬂector that focused the ﬂuorescence light
from the ﬁrst focal point to a photomultiplier tube (PMT) positioned at the second
focal point (ellipsoid focusing properties will be further discussed in Sec. 3.2). The
PMT then converted the photons to an electrical signal. More details on the photon
detection system are provided in the following chapters.

1.5.5

DAQ and Controls

The BECOLA data acquisition system (DAQ) was used to control the scanning voltage applied to the ellipsoidal reﬂector of the photon detection system, and to read out
the electrical signals from the photon detector. The DAQ consisted of a DC ampliﬁer
(Matsusada AMP) used to apply a scanning voltage of ± 10 kV (maximum) to the
ellipsoidal reﬂector. The electrical signal from the PMT was converted to NIM logic
signal by a NIM to TTL (Transistor-Transistor Logic) converter and counted by a

15

scaler. The layout of the DAQ is depicted schematically in Fig. 1.9. The scalar and
control of scan parameters were accessed through a graphical user interface (GUI)
using the EPICS (Experimental Physics and Industrial Control System) control system [13, 14]. Scan parameters include the scanning voltage range, the step size, the
dwell time, and the number of scans to be accumulated. Small step size allows for
better data resolution in a scan, while larger step size allows the scanning over a
broader voltage range.

AMP

Signal Output

Photon
Detection

PMT

Scalar
DAQ

DAQ Voltage

DAQ/Scalar Control

Computer

EPICS

Onboard
CPU

VME Backbone

Scalar Counts

Laser
Ion/Atom Beam

Velocity
Adjustment
Electrode

VME Bus
Figure 1.9: The basic layout of the DAQ and controls system for BECOLA.

16

Chapter 2
Simulation of and Extensions to
the Mainz Photon Detection
System
2.1

Tools

Optical ray trace simulations of the photon detection system were preformed using
the FRED optical engineering software code version 9.10.0 by Photon engineering [15]
to optimize the system’s performance in BECOLA. The starting point of the simulations was the photon detection system designed by J¨rg Kr¨mer at the University of
o
a
Mainz [16]. The computer-aided design (CAD) ﬁle for the Mainz system was imported
into FRED and the materials, light sources, and analysis surfaces were deﬁned. The
original Mainz conﬁguration ﬁle was then modiﬁed to ﬁnd an optimal geometry for
the ellipsoid, light collection, and photomultiplier tube (PMT). The ﬁgure of merit
used was the signal to noise ratio (S/N), where signal (S) was the number of ﬂuorescence photons simulated to enter the PMT and noise (N) was deﬁned as the square
root of the simulated stray light photon count at the face of the PMT.

17

2.1.1

Geometry Used in FRED

The photon detection system imported from the Mainz CAD ﬁle was centered inside
a 6-way cross vacuum chamber, and two additional vacuum chambers were included
in the model as sources of stray laser light reﬂecting along the beam line.
The signal light sources were generated from a 3-mm diameter cylinder placed
along the beam direction that included the ﬁrst focal point of the ellipsoidal reﬂector.
Two cylinders were created in order to model the ﬂuorescence light from ions or from
atoms. The ﬁrst cylinder started at the upstream inner HV drift tube and ended at
the downstream inner HV drift tube describing the region where the scanning HV
would be applied for an ion beam. The second cylinder started halfway through the
charge exchange cell vacuum chamber to include the region where ﬂuorescence would
start for an atom beam. The atomic ﬂuorescence source light cylinder passed through
the photon detection system chamber and ended at the end of the third chamber (see
Fig. 1.4 for the locations in the conceptual design of BECOLA). The photon detection
system geometry, as developed in FRED, is presented in Fig. 2.1.
Stray light was assumed to have two contributions in the simulations performed
here: laser light reﬂected along the beam line and laser light scattered by the ﬁrst
beam line aperture. The ﬁrst contribution was modeled by light sources on the inner
surfaces of the three vacuum chambers. The second stray light source was modeled
with a ring-like source on the inner edge of the ﬁrst beam line aperture.
Random rays were programmed to be emitted isotropically from the surface of
the light source cylinder to simulate the ﬂuorescence given oﬀ by the particle beam,
and from the inner surfaces of the vacuum chambers to model reﬂection of stray laser
light. A total of 1,000,000 rays were traced for the signal light source, at wavelength
values of 397 nm and 800 nm. Photons were lost as they interacted and were absorbed
by surfaces in the simulation, and they were also reﬂected and transmitted through
materials based on their reﬂective properties.

18

Inverted Viewport
PMT Analysis
Plane
Fluorescence
Light Source

Charge
Exchange Cell
Chamber

Third Vacuum
Chamber

Apertures
Photon Detection
System Chamber

Figure 2.1: Engineering drawing of the geometry imported into FRED for ray trace
simulations of the detection system.
Analysis planes were deﬁned at the second focal point of the ellipsoidal reﬂector
for reference and at the entrance to the PMT to count the deposited photons.

19

2.1.2

Inputs and Constraints

Surfaces in the FRED geometry can be deﬁned in a variety of ways. Lens materials
were already programmed into FRED, as were common optical polymers, air, and
vacuum. Quartz, glass and acrylic were deﬁned using the preset material designations.
Other materials were deﬁned using the real and imaginary refractive indexes for the
wavelength of interest. Fresnel’s equation for reﬂection of a plane wave normal to a
plane surface between vacuum and a material with complex index of refraction is used
within the FRED code to calculate the reﬂectivity, R, from the refractive indices of
a material, with the expression:

R=

(n − 1)2 + k 2
,
(n + 1)2 + k 2

(2.1)

where n is the real component of the refractive index and k is the complex component
of the refractive index for the material.
After a material was deﬁned, coating and ray trace properties were assigned. Surface coating can be used to deﬁne the surface as transmissive, reﬂective or absorbing.
Options for ray tracing included transmitting rays or stopping all rays once they hit
the surface. The “stop all rays” option was helpful for analysis, because it allowed
the photon distributions striking a surface to be examined. Rays could also be set to
change visualization colors once they interacted with a surface, so rays that interacted
in diﬀerent ways with the material could be identiﬁed. This visualization option was
used to ensure that surface materials were correctly assigned in the FRED code.
Analysis surfaces were deﬁned to examine the ray distributions on a plane and were
set to record all rays that were stopped by a surface. Light sources with wavelengths
of 397 nm and 800 nm were used for both the signal (ﬂuorescence) and stray light
simulations. The power from each light source was arbitrarily set to 1 W. Only one
wavelength was traced and analyzed at a time.
The irradiance distribution on the surface of a chosen analysis plane was selected

20

for the present analysis and was used to deduce the number of photons and the S/N.
A distribution of the irradiance (power over area) for the rays stopped on the surface
was provided by the FRED code after each run as visual output. The number of
photons was proportional to the power:

P =

nf E f .

(2.2)

f

Here P is the power, nf is the number of photons with frequency f , and Ef is the
energy of a photon of frequency f . The simulation results for a given wavelength were
compared directly with all other wavelength simulation results, since all simulations
for a given frequency started with the same number of photons.

2.2

Typical Output

The photon distributions at the second focal plane indicated a distinct physical separation of the signal and stray light distributions. The irradiance distributions of the
signal and stray light at the second focal plane for a 76.2-mm ellipsoidal reﬂector,
which was eventually used in the photon detection system for BECOLA, are shown
for example in Fig. 2.2. An iris system at the second focal plane could be used to take
advantage of the physical separation between the distributions and increase the S/N
by blocking the stray light that is concentrated at the outer edge of the focal plane.

21

)

)
(In
is
Ax

Ax

is

(In

lX

(In

)

ca

is

lY

Lo

Irradiance (W/In2)

)
(In
is
Ax

Ax

ca

lX

lY

Lo

ca

ca

Lo

Irradiance (W/In2)

Lo

Figure 2.2: Simulated signal (left) and stray light (right) distributions at the second
focal plane of 397 nm light for the photon detection system that includes a 76.2-mm
ellipsoidal reﬂector.

2.3
2.3.1

Optimizing System Geometry
Reﬂector Size

The Mainz system consisted of an ellipsoidal reﬂector with a 50.8 mm photon collection region (minor ellipsoid axis). The photon collection length was taken to be the
distance that the laser and particle beam interacted to give oﬀ ﬂuorescence inside the
reﬂector. The reﬂector’s photon collection length was varied in the simulations with
FRED to optimize the light collection and S/N of the detection system. A 127-mm
long collection length was selected for use in the simulation because it was the largest
possible reﬂector that could ﬁt into the planned BECOLA beam line. A 76.2 mm collection length system was also created and simulated to get a performance measure
for a reﬂector in-between the two other sized systems.
The goal of the simulations for the diﬀerent reﬂector lengths was to ﬁnd the
optimal light collection geometry, again using S/N as the ﬁgure of merit. Signal and
stray light distributions at the second focal plane analysis surface were analyzed at
397 nm and 1 Watt of power from the signal source for 50.8 mm, 76.2 mm and 127 mm

22

diameter reﬂectors. Stray light was also analyzed at 397 nm with an initial power of
1 Watt from each inner surface of the three vacuum chambers in the simulation. The
three diﬀerent reﬂector geometries were tested with both 100% and 0% reﬂective inner
surfaces of the vacuum chambers to cover both the worst and best case scenarios for
the stray light production. The best case was assumed to be achievable, in practice,
by black plating the inside of the vacuum chambers.
The results of relative S/N for signal and stray light distribution irradiances at the
second focal point for the three simulated reﬂector geometries are shown in Fig. 2.3.
The S/N appears to saturate for the 76.2-mm reﬂector, and S/N can also be increased
when a 25.4 mm diameter iris is placed at the second focal plane. Saturation of the
signal is expected for larger collection lengths, since the signal distribution is well
focused at the second focal point, and the shoulders along the beam axis (due to the
pencil like distribution along the beam path) contribute only a small fraction of the
total light collected at the second reﬂector focal plane. The gain from the shoulders
at larger sizes appears to stop being signiﬁcant at a collection length of 76.2 mm. The
stray light distribution is expected to decrease for larger collection dimensions because
as the surface area inside the reﬂector was increased, the beam path opening size
remains constant. The 76.2 mm reﬂector was employed in all subsequent simulations
described herein, since this collection length was found to have the optimal S/N.

23

no iris

25.4 mm diameter iris

180
160
Relative S/N

140
120
100
80
60
40
20
40

60

80

100

120

140

Reflector Collection Length (mm)

Figure 2.3: Relative S/N values obtained by simulation for 397 nm light collected at
the second focal plane for the three collection lengths speciﬁed in the text. The upper
points include the eﬀect of an iris at the second focal plane.

2.3.2

Irises at the Second Focal Plane

An iris system of variable size was also included in the simulated geometry. The iris
was positioned at the second reﬂector focal plane and the iris opening size was varied
to optimize the S/N for light collection. The simulation results are shown in Fig. 2.4.
The S/N was found to increase as the open iris area was increased for irises having a
circular geometry. Iris systems with a rectangular geometry were also simulated and
the S/N doubled as the area was doubled from a 12.7 x 6.35 mm rectangular iris to
a 25.4 x 6.35 mm rectangular iris.

24

450
400
Relative S/N

350
300
250
200
150
100
50

0

100 200 300 400 500 600 700 800
2

PMT Iris Open Area (mm )

Figure 2.4: S/N ratio, as a function of iris open are at the PMT face, deduced from
simulated signal and stray distribution irradiance at the second focal plane of 397 nm
light for the 76.2-mm reﬂector. The stars denote circular irises while the rectangular
points denote rectangular irises.

2.3.3

Beam line Apertures

The possible placement of beam line apertures was studied in the simulations to
determine the potential impact on S/N. Apertures were placed both upstream and
downstream of the ellipsoidal reﬂector, inside the inner drift tubes (see Fig. 2.5). The
most upstream aperture was designed to cut the laser halo, but scatters the laser
light. The subsequent apertures were designed to block light scattered by the ﬁrst
aperture and to reduce the opening angle into the reﬂector through which stray light
can enter.
Laser light was assumed to only be scattered by the ﬁrst aperture. Backscattering
was not considered in the simulations. Aperture conﬁguration (a) in Fig. 2.5 was not
simulated because there was no aperture to include a stray light source. Aperture
conﬁguration (c) was shown to have the highest S/N of the conﬁgurations simulated.

25

This is reasonable, since a majority of the simulated stray light comes from the ﬁrst
beam line aperture and the additional upstream apertures serve to block more of the
scattered (stray) light.

(a)

(b)

(c)

(d)

Key:
6 mm
7 mm
8 mm
9 mm
10 mm
(e)
Figure 2.5: Conceptual layout of the inner drift tube aperture conﬁgurations that
were simulated in the present work. The laser light passes from left to right through
the center of the apertures and through the ﬁrst focal point of the reﬂector.The sizes
of the apertures are indicated by their color.

26

2250
2000
1750
Relative S/N

1500
1250
1000
750
500
250
0

A

B

C

D

E

Beamline Aperture Configuration

Figure 2.6: The S/N ratio as a function of aperture conﬁguration deduced from simulated signal and stray distribution irradiance at the second focal plane for 397 nm
light with the 76.2-mm reﬂector.

2.4

Summary of Results

The FRED optical engineering software was used to simulate the performance of the
“blue” photon detection system for BECOLA. An evaluation of the S/N as a function
of the length of the ellipsoidal reﬂector showed that optimal performance could be
achieved with a 76.2-mm interaction length. Therefore, a 76.2 mm reﬂector has been
implemented for the BECOLA facility based on the simulation results. The simulation results also revealed a distinct physical separation of signal and stray light at
the second focal plane of the ellipsoidal reﬂector. These results suggest that an iris
system can be used at the second focal plane to optimize the S/N. Such an iris system was simulated with the FRED code and conﬁrmed the improved light collection
performance with restricted opening at the second focal plane. An adjustable iris
system was constructed and placed in front of the PMT at the second focal plane of
the BECOLA photon detection system as a result of these simulations. Simulation
27

results of beam line aperture conﬁgurations indicated, not surprisingly, that beam
line aperture locations have a signiﬁcant eﬀect on S/N. The conﬁgurations labeled
(a)-(e) were tested experimentally to verify the simulations, and the results of the
tests are presented in Chapter 5.

28

Chapter 3
Design of the BECOLA Photon
Detection System
3.1

Design

The design of the photon detection system was based on the results of the simulations
described in the previous chapter, starting from the tested design of the Mainz system
and including new constraints at BECOLA. An added feature is that the BECOLA
laser system consists of a Ti:Sapphire laser and frequency doubler, which can produce
light in the wavelength ranges of 700-1000 nm and 350-500 nm, respectively, and the
two laser wavelength ranges require diﬀerent PMTs for eﬃcient photon detection.
However, the light collection system for the interaction region was designed to be
useful for both wavelength ranges. The photon detection system can work for ion
beams by scanning a high voltage (<10 kV) applied to the interaction region and for
atom beams by applying the high voltage to the CEC. An electrically-conductive material was selected for the reﬂector to permit application of high voltage for scanning
purposes, and it was designed to be isolated from ground.

29

3.2

Interaction Region and Light Collection

The interaction region of the BECOLA facility is the region between the two BOBs
where the ion/atom and laser beams are collinearly overlapped for about 3 meters.
Emission of ﬂuorescence light for atomic beams begins after neutralization, and for
ion beams the ﬂuorescent region begins upon collinearity between the ion and laser
beams. For both atom or ion beams, the light collection region is deﬁned by the
76.2-mm length of the reﬂector.
The geometrical properties of an ellipsoid ensure that all photons emitted at the
ﬁrst focal point are reﬂected to a second focal point [17], as depicted in Fig. 3.1.
The collinearly overlapped ion/atom and laser beams pass from left to right through
the ﬁrst focal point of the ellipsoid. Fluorescence photons (denoted by dashed lines)
are emitted along the beam path, so not all photons are emitted at the ﬁrst focal
point. The ellipsoidal reﬂector is eﬃcient at focusing any ﬂuorescence emitted inside
the collection region to the second focal plane [16, 17]; however, stray light will be
focussed as well [18].

Fluorescence
Photons

Focal
Points

Beams

Figure 3.1: Representation of the geometrical properties of an ellipsoidal reﬂector.
Fluorescence photons emitted at the ﬁrst focal point are reﬂected to the second focal
point. The reﬂector has rotational symmetry around the vertical axis.

30

3.2.1

Ellipsoidal Reﬂector Speciﬁcations

The geometrical properties of the reﬂector constructed for the BECOLA photon detection system are shown in Fig. 3.2. The minor axis of the ellipsoid has a radius of
38.10 mm, a major axis radius of 53.34 mm, and the focal length is 74.68 mm. The
focal length was chosen to allow for an insulation gap of 18.03 mm between the top
of the reﬂector and the inverted viewport window ﬂange. The reﬂector has rotational
symmetry along the vertical axis to increase solid angle coverage.

b = 53.34 mm

f = 74.68
mm

a = 38.10 mm

Figure 3.2: The geometrical properties of the BECOLA ellipsoidal reﬂector. The reﬂector has rotational symmetry around the vertical axis.

3.2.2

Ellipsoidal Reﬂector Material

The ellipsoidal reﬂector must be reﬂective in both the “blue” and “red” system wavelength regions. The reﬂectivity curves in the required wavelength range for the metals
that were considered are shown in Fig. 3.3. The reﬂectivities for all three metals are

31

high in the “red” region but the reﬂectivities of silver and copper drop oﬀ rapidly
as they approach the “blue” wavelength range, thus aluminum was chosen for the
ellipsoidal reﬂector material.

1.0
0.9
0.8

Reflectivity

0.7
0.6
Aluminum
0.5

Silver
Copper

0.4
0.3
0.2
0.1
0.0
300

400

500

600

700

800

900

1000

Wavelength (nm)

Figure 3.3: Reﬂectivity curves of the metals considered for the ellipsoidal reﬂector in
the wavelength regions available at BECOLA. Silver is depicted by diamonds, copper
by circles and aluminum by stars.
Aluminum is a soft metal and is diﬃcult to polish to a mirror-like surface. Machine
polishing was attempted, but deep scratches were found on the inner reﬂector surface.
In the end, the reﬂector was hand polished using micro-mesh polishing pads [19] and
was ﬁnished using a Mother’s Power Ball Mini [20] with Simichrome metal polish [21].
The reﬂectivity was tested using the BECOLA laser system and values of ∼68-72%
were achieved at 794 nm (representative “red” wavelength) and ∼58-62% at 397 nm
(representative “blue” wavelength). A photograph of the ﬁnished reﬂector is shown
in Fig. 3.4.
Aluminum oxidizes in air over time; however, the reﬂector is generally kept under vacuum, which will help slow the oxidative process. Touch-up polishing may be

32

required to remove the layer of aluminum oxide that forms, and also to remove any
alkali metal that leaks into the reﬂector from the CEC.

Figure 3.4: Photograph of the ellipsoidal reﬂector after hand polishing.

3.3

Stray Light Reduction

The sensitivity of the photon detector to stray light was noted by the Mainz group
to be a critical hurdle for experimental work [16, 18]. Increased stray light leads to a
reduced S/N, which is problematic for low counting rates expected for rare isotope
beams. Several features were implemented at the BECOLA beam line to reduce stray
light levels. (1) The inside of the detection system vacuum chamber was black plated.
(2) An outer drift tube was added around the inner drift tubes. The outer and inner
drift tubes were also black plated to minimize scattering of stray light inside the
chamber that houses the BECOLA photon detection system. (3) Apertures were
placed inside the inner drift tubes to cut the laser halo and to block out the resulting
scattered laser light as discussed in Sec. 2.3.3. The apertures were similarly painted
black to decrease their reﬂectivity and reduce stray light. Additional details on the
beam line apertures are given in Sec. 3.3.1.(4) Following the simulation results, the
33

second focal plane was designed to accept an iris of variable opening geometry and
size. The iris is placed immediately in front of the “blue” PMT as described in Sec.
3.3.2.

3.3.1

Beam line Apertures

Apertures placed inside the inner drift tubes were used to cut the laser halo and to
block the resulting scattered laser light. The apertures also reduced the stray light
that entered the open ends of the ellipsoidal reﬂector. The beam line apertures were
fabricated out of aluminum and anodized black. Anodized aluminum is not electrically
conductive, and can charge up when hit by an ion beam. The anodized apertures were
coated with graphite paint to make them conductive to avoid any beam charging and
possible beam deﬂection.
The beam line apertures were designed to be placed in pairs inside the inner drift
tubes, as depicted in Fig. 3.5. The apertures all have an outer diameter of 27.7 mm
and are 3.0 mm thick with a conical opening. The inner diameters of the smaller
opening that were fabricated are 7 mm, 8 mm, 9 mm, and 10 mm, where the angle
is 45◦ between the smaller and larger diameter opening. Two 7 mm inner diameter
apertures are shown in Fig. 3.6 after application of graphite paint.

45°

45°

Figure 3.5: Schematic view of the placement of a pair of beam line apertures.

34

Beam Out

Beam In

Figure 3.6: Photograph of two 7 mm inner diameter beam line apertures after being
painted with graphite paint to be conductive.

3.3.2

Irises for the PMT

An iris can be placed at the second focal plane of the ellipsoidal reﬂector, as discussed
in Sec. 2.3.2. The irises were fabricated from 0.79 mm thick aluminum disks with a
44.1 mm outer diameter. The inner opening size and geometry were varied. Irises
with inner opening diameters of 31.75 mm, 25.4 mm, 19.05 mm, and 12.7 mm were
fabricated. After the machining process, the irises were black anodized to absorb stray
light. Several irises were not machined (6.35 mm in diameter, 6.35 × 12.7 mm, 6.35 ×
25.4 mm) but rather were hand-made from black aluminum foil. A teﬂon holder that
mounts on the PMT was designed to hold the iris in place at the second focal plane
and allowed for eﬃcient switching of irises during experiment. Shown in Fig. 3.7 are
the 12.7 mm diameter and the 6.35 × 25.4 mm rectangular irises, as they were placed
inside the teﬂon holder. The irises fabricated from the black foil were attached to the
teﬂon holder using double-sided tape.

35

PMT side Iris side

Figure 3.7: Photographs of the 12.7 mm (machined) in diameter circular iris and the
6.35 × 25.4 mm (hand-made) rectangular iris inside the teﬂon iris mounting system.
The PMT side of the teﬂon holder is shown on the left and the side adjacent to the
inverted viewport window, where the iris is mounted, is shown on the right.

3.4

PMT Selection

The PMT selection was based on a review of the active area, quantum eﬃciency
(QE), dark count levels, and usable wavelength range as listed by the manufacturers.
Dark count levels are due to leakage current between the dynodes and to thermal electrons produced by the photocathode. thermal electron production depends on ambient
temperature, photocathode material, and active area. The usable wavelength range
depends on the photocathode material. The QE is the probability that a conversion
electron will be produced at the photocathode for each photon that hits the active
area and is another property of the photocathode material. The usable wavelength
regions for diﬀerent photocathodes served as the “ﬁrst criterion” for PMT selection,
followed by the active area as well as dark count levels.
The “blue” wavelength region had multiple options for photocathode materials,
and many PMTs are available in photon counting module forms. A photon counting
module has high sensitivity to detect individual photons, and includes built-in elec-

36

tronics to amplify the signal and apply a voltage threshold to minimize the dark count
levels. Simulated signal light distributions at the second focal plane of the ellipsoidal
reﬂector (see Fig. 2.2) showed a peak in the distribution that covered an area of ∼24
mm in diameter around the focal point, with a spread along the beam axis. An active
photon counting area greater than this dimension may lead to an increase in the dark
count rate. Therefore, a PMT with a 25.4-mm diameter active area was thought to
best match the main area of ﬂuorescence collection at the second focal plane, while
discriminating against dark counts. A Hamamatsu H11123 PMT was chosen to detect
photons in the region of 300-500 nm (“blue”). It is a head-on PMT with a 28 mm
diameter, and a quantum eﬃciency of 17.6% at ∼400 nm. At room temperature, the
expected dark count rate is 70 per second.
By comparison, there were fewer photocathode material options for the “red”
wavelength region (700-1000 nm). The “red” photocathode materials also tend to
have higher dark count rates, since they produce more thermal electrons at room
temperature. Therefore, it was imperative to select a PMT with a reduced active area.
Additional simulations were performed with the FRED code to compare available
PMTs for “red” light optimizing QE and active areas. As a result of the simulations,
the PMT selected for detection of red light was the Hamamatsu R10699. The R10699
is a side-on PMT with an active area of 8 x 24 mm. The PMT is designed to have
good eﬃciency over a broad wavelength range (185 nm to 900 nm). The quantum
eﬃciency at 800 nm is reported to be 9%. The measured dark counting rate at room
temperature is below 10,000 counts per second, but depends on the voltage threshold.
The “red” PMT would also be applicable for detection of “blue” light where the
QE increases to ∼ 30% at 400 nm. However, the R10699 has a smaller active area
and higher dark count levels at room temperature than the H11123 PMT chosen for
“blue” light collection; therefore, the R10699 would not be as sensitive to individual
photons in the “blue” region.

37

3.5

“Blue” Light Detection System

The mechanical design of the light detection system for ﬂuorescence photons in the
“blue” region of the electromagnetic spectrum is illustrated in Fig. 3.8.

Inverted Viewport
Hamamatsu H11123
Photomultiplier Tube

Ellipsoidal
Reflector
Inner Ground
Drift Tubes

PMT
Iris

Outer Drift
Tube
Collinear
Beams In
Inner HV
Drift Tubes

Apertures

Figure 3.8: The Mechanical design drawing of the blue photon detection system for
BECOLA.
The light collection region is 76.2 mm long. As indicated above, the reﬂector was
fabricated from aluminum and hand polished to a mirror-like surface. The collinearlyoverlapped laser and atom/ion beams enter the detection system through the inner
ground drift tube, which holds the beam line apertures to cut the laser halo and resulting scattered laser light. The inner ground drift tube is welded to the outer drift
tube as depicted in Fig. 3.9. The beams then pass through the inner high voltage drift
tube, which was welded to the reﬂector. The reﬂector itself is mounted on insulated
legs, which are attached to a ﬂange to allow removal and servicing of the reﬂector and
the beam line apertures. A high voltage feedthrough on the ﬂange is used to connect
a variable scanning voltage to the reﬂector. A thin copper mesh (with 95% trans38

mission for light) is placed on top of the reﬂector to ensure a uniform voltage inside
the reﬂector. Fluorescence emitted along the collection region inside the reﬂector is
reﬂected through a quartz inverted viewport window and to the second focal plane.

Figure 3.9: Photograph of the vacuum chamber with only the inner and outer drift
tubes in place. The beam direction is into the paper through the aperture inside the
inner ground drift tube.
As detailed in the previous section, a Hamamatsu H11123 PMT is used to detect
photons in the region of 300-500 nm at the second focal plane of the ellipsoidal
reﬂector. The PMT is centered inside the inverted viewport tube by two teﬂon rings,
and ambient light leaks from the experimental area are blocked by a PMT enclosure.
The “blue” PMT, enclosure, and teﬂon rings are depicted in Fig. 3.10.
An iris can be placed in front of the “blue” PMT as a means to reduce stray light
from reaching the PMT. Details of the iris system were provided in Sec. 2.3.2 and
3.3.2.

39

Figure 3.10: Photograph of the “blue” photon detection system PMT, PMT enclosure,
and teﬂon rings used to center the PMT inside the inverted viewport.

3.6

Red Light Detection System

The design of the photon detection system for the “red” region of the electromagnetic
spectrum is depicted in Fig. 3.11. The “red” system has essentially the same geometry
as the “blue” system. The iris system for the “red” system is placed in the same
location as the “blue” system; just above the inverted viewport at the second focal
plane of the ellipsoidal reﬂector.
The side-on arrangement of the R10699 PMT is not compatible with the housing
designed for the H11123 head-on PMT or with the inverted viewport tube geometry.
Therefore, the “red” light PMT was placed on top of the viewport tube. A tubular
silver mirror was used to guide photons from the second focal point to the entrance
window of the PMT. The PMT was surrounded by a water-cooled, stainless steel
enclosure to keep it at constant operating temperature, as depicted in Fig. 3.12. The
PMT base is outside of the stainless steel cooling enclosure, since the base generates
signiﬁcant heat and would negatively impact stable operation of the PMT.

40

Inverted Viewport
Hamamatsu R10699
Photomultiplier Tube
Ellipsoidal
Reflector

Silver
M i r r o r Tu b e
Collinear
Beams In
Inner Ground
Drift Tubes

Outer Drift
Tube

Inner HV
Drift Tubes

Apertures

Figure 3.11: The Mechanical design drawing of the red photon detection system for
BECOLA.

41

Figure 3.12: A photograph of the “red” photon detection system PMT enclosure
attached to the photon detection system vacuum chamber.

42

Chapter 4
Calcium Ion Beam Spectroscopy
and Commissioning of the “Blue”
Photon Detection System
4.1

Commissioning Introduction

The “blue” photon detection was commissioned by collecting ﬂuorescence light from
Ca+ ions in the BECOLA system. Speciﬁcally, the ﬁne structure of the 4s → 4p
atomic states in 40 Ca+ was observed. The wavelength of the electronic transition is
397 nm. The ﬁne spectrum was collected by counting ﬂuorescence photons with the
“blue” PMT as a function of the scanning voltage applied to the ellipsoidal reﬂector.

4.1.1

Calcium Ion Spectroscopy

The stable isotopes of calcium are: 40 Ca (96.941%), 42 Ca (0.647%), 43 Ca (0.135 %),
44 Ca (2.086%), 46 Ca (0.004%), and 48 Ca (0.187%). The even-even calcium isotopes

have a ground state nuclear spin of 0 and therefore exhibit no hyperﬁne splitting. The
ﬁne structure of 40 Ca is shown in Fig. 4.1 as an example. The allowed electric dipole

43

transitions between the ﬁne structure levels are the D1 line (4 2 S1/2 to 4 2 P1/2 ) and
the D2 line (4 2 S1/2 to 4 2 P3/2 ).

2P
3/2
4p
2P
1/2
D1
396.9 nm
2S
4s
1/2
Figure 4.1: Schematic representation of the ﬁne (and hyperﬁne) structure of 40 Ca,
which has a ground state nuclear spin IN = 0.
The lifetime of the D1 line of 40 Ca+ is known to be τ = 6.978 ± 0.056 ns, which
corresponds to a natural line width of δνnat = 22.8081 MHz. For the D1 line in 40 Ca+
at a beam energy of 15 keV, the diﬀerential Doppler shift is 22.62 MHz/V.

4.1.2

Commissioning Set-up

BECOLA Beam line
Stable calcium ions were produced using a commercially available Colutron plasma
ion source described in Sec. 1.5.3. Calcium metal was placed in a ceramic charge
holder tube and was pushed inside a boron nitride chamber. A voltage between 7
and 9 V was applied to a ﬁlament to strike and sustain a plasma. The ﬁlament also
heated the holder to evaporate the calcium metal. An anode plate with a ∼0.2 mm
diameter pinhole was used to deﬁne the emittance of the beam. A positive voltage
was applied to the anode for conﬁnement of the plasma. Calcium ions were extracted
from the source and accelerated to an energy of 15 keV and deﬂected by electrostatic
elements in the beam switchyard (refer to Fig. 1.4), into the spectroscopy beam line.
Focusing of the ion beam through the beam interaction region was achieved using
two sets of quadrupole doublets immediately downstream of the switchyard. The ion
44

beam current was measured with Faraday cups located in two BOBs: one before and
one after the detection region of the spectroscopy beam line. Two collimators, again
located in the two BOBs along with the Faraday cups, were used to limit the laser
and ion beam sizes to 5 mm and to ensure overlap between the two beams.
Laser System
The Ti:Saph laser was set to 793.21 nm, and this light was frequency doubled to
a wavelength of 396.605 nm. A quarter wave plate and polarizer cube were used to
restrict the laser power that was coupled into a ﬁber-optic cable to several mW. The
light was transported to the experimental area via ﬁber-optic cable.
The wavelength values were not centered on the Doppler shifted laser frequency
for the D1 line of 40 Ca+ at 15 keV because the present version of the BECOLA
DAQ system can only apply positive scanning voltages to the interaction region. To
observe the resonance within the voltage range currently accessible to BECOLA, the
laser frequency was set to center the resonance at an applied voltage of ∼200 V rather
than zero.

4.2
4.2.1

Stray Light Characterization
Beam Line Preparation Following Simulation Results

Reducing Reﬂectivity of Surfaces of Interaction Region
Black aluminum foil was used to cover the reﬂective surfaces in the photon detection
system chamber, and also to ﬁll in gaps where stray light could enter. Locations where
the foil was applied are indicated in Fig. 4.2.
The outside of the reﬂector was not polished, and could serve as an excellent
reﬂector of stray light around the chamber. The outer surfaces of the reﬂector were
covered with black aluminum foil, as shown in Figs. 4.3 and 4.4.

45

(a)

(b)

(c)

Figure 4.2: Locations where black aluminum foil was placed to reduce stray light are
marked with pink arrows. (a) The reﬂector outer surfaces and drift tube gaps. (b)
The inverted viewport and inverted viewport ﬂange. (c) The reﬂector ﬂange.

(a)

(b)

(c)

Figure 4.3: Locations on the (a) Bottom of the reﬂector; (b) outside of reﬂector; and
(c)reﬂector ﬂange where black foil was placed to reduce surface reﬂectance are shown
as hatched areas.
The inner ground drift tubes were connected to the outer drift tube to center them
in the chamber. The gap between the inner and outer drift tubes is large (∼61-mm
long) and stray light could enter the detection chamber via this gap. Black aluminum
foil was placed as shown in Fig. 4.5 to cover this gap. In simulations, such coverage
of the inner/outer drift tube gap reduced reﬂected laser light by 90%.
The inverted viewport tube, window ﬂange, 8-inch conﬂat inverted viewport ﬂange,
and 8-inch conﬂat reﬂector ﬂange were all fabricated from reﬂective materials. Each
of these components was wrapped with black foil, as shown in Fig. 4.6. In particular,
the viewport window ﬂange, by necessity, overlaps the open top of the reﬂector and

46

Figure 4.4: Photograph of the ellipsoidal reﬂector covered in black aluminum foil.

Figure 4.5: The black aluminum foil was placed in the gap between the outer and
inner ground drift tubes on both the upstream and downstream sides of the reﬂector. Crosshatching shows the locations that were covered with black foil to decrease
reﬂectivity.

47

provides a path for light to scatter back into the reﬂector. Therefore, it was crucial
that the window ﬂange be covered with a non-reﬂective material (see Fig. 4.7).

(a)

(b)

Figure 4.6: Hatched areas show the locations on the (a) Inverted viewport ﬂange.
(b) Inverted viewport tube, window ﬂange, and ﬂange where black foil was placed to
reduce surface reﬂectance.

Figure 4.7: Photograph of the inverted viewport window ﬂange covered in black aluminum foil.

48

4.3

Commissioning Measurements and Comparison to Simulations

4.3.1

Aperture Conﬁgurations

The beam line aperture conﬁgurations in the BECOLA interaction region were varied
and the S/N measured for comparison with results of the simulations of the S/N from
FRED optical engineering software as described in Sec. 3.3.1.

(a)

(b)

(c)

(d)

(e)

(f)

Key:
6 mm
7 mm
8 mm
9 mm
10 mm
Figure 4.8: Schematic represemtation of the inner drift tube aperture conﬁgurations
that were tested in BECOLA. (a)-(e) correspond to A-E respectively, (f) is the Mainz
system aperture conﬁguration, which was not tested.
The photon detection system had to be vented each time to change the aperture
conﬁguration. Caution was taken when changing apertures to ensure that the system
geometry (inner and outer drift tube alignment) did not change. As noted in Sec.
3.3.1, the apertures were painted with graphite paint to make them conductive. Although the painting was completed in advance of the measurement, the pieces were

49

degassed by heating before installation. The beam line aperture measurements were
all completed in a one-day period to minimize systematic changes in the laser beam,
ion beam and/or beam optics. The laser power and ion beam current were sampled
periodically, and used to normalize the observed ﬂuorescence of the 40 Ca+ ions. The
results are detailed and discussed in Chapter 5.

4.3.2

PMT Iris Area

The S/N of the blue photon detection system was characterized with various PMT
irises placed at the second focal plane. The irises used for commissioning tests included
the following: circular irises with diameters of 31.75, 25.4, 19.05, 12.7, and 6.35 mm
in diameter, and two rectangular irises of dimension 6.35 × 25.4 mm, and 6.35 ×
12.7 mm. Recall that the simulation results (Sec. 2.3.2) suggested that the 31.75-mm
diameter iris should give the best S/N.
The PMT iris measurements were also all completed in a one-day period. Alignment of the rectangular irises was performed by placing the iris into the inverted
viewport and rotating it until it was visually parallel to the beam axis. The PMT
was replaced once the iris was aligned. The laser power and ion beam current was
sampled periodically and used to normalize the observed ﬂuorescence of the 40 Ca+
ions. The results are contained in the following chapter.

50

Chapter 5
Results and Discussion
5.1
5.1.1

Line Shape
Typical Resonances

The D1 resonance of 40 Ca+ was used to deduce S/N for the beam aperture and PMT
iris measurements outlined in the previous chapter. An example 40 Ca+ ﬁne spectrum
collected using the “blue” photon detection system of BECOLA is shown in Fig.
5.1. This spectrum was obtained by limiting the data collection time for a given
voltage scan. An oscillation on the isolated main voltage supply was noted early in
the BECOLA commissioning, and electronic gates were established to trigger data
collection within a limited and consistent period to be “in sync” with the observed
AC oscillation (60 Hz), see next section for details.

51

Counts x 10

4

(Arbitrary Units)

5.0
4.5
4.0
3.5

40

+

Ca

D1

482 pA
216

W

0.25 x 0.5 in Iris
Upper Gate

3.0
2.5
2.0
1.5
1.0
0.5
0.0
180

185

190

195

200

205

210

215

Scanning Voltage (V)

Figure 5.1: Typical 40 Ca+ resonance and ﬁtted curve used to extract S/N ratio for
evaluation of photon detection system performance.
Electronic Gates
“D1” ﬁne structure resonance curves for 40 Ca+ collected without discrimination
for time are presented in Fig. 5.2. Initial ﬁne structure measurements of 40 Ca+ at
BECOLA produced resonance curves with multiple peak structures and unexpectedly
large linewidths. The multiple structures were found to vary in intensity and position,
as a function of the ion optics.
The suspected cause of the multiple peaks in the Ca+ resonance spectrum was a
60 Hz ﬂuctuation in the isolated main AC power for the FUG DC acceleration power
supply. The observed oscillation on the 15 kV output voltage of the acceleration
power supply is shown in Fig. 5.3. The AC ripple is about 8 V peak to peak. Two
of the peaks in the Ca+ resonance were believed to correspond to data obtained at
the minima and maxima of the ripple. The source of the third peak, most evident in
Fig. 5.2(b), remains unknown. An electronic gate was used to restrict the counting

52

of signals from the PMT to a 500 microsecond region of the oscillation, which had a
period of ∼16 ms. Gates were set both on the upper and lower portions of the ripple,
as shown in Fig. 5.3. Data were taken for both gate conditions to account for possible
diﬀerences in S/N due to changes in beam optics.
Implementation of the electronics gates signiﬁcantly reduced the data collection
duty factor. However, use of the gate was necessary to realize a resonance line shape
that was more representative of the laser interaction with the ion beam.

53

4.0

40

+

Ca

3.5

D1

(Arbitrary Units)

Counts x 10

5

522 pA
3.0
2.5

202

W

6.35 mm
PMT Iris

2.0

No Gate
1.5
1.0
0.5
0.0
180

185

190

195

200

205

210

215

Scanning Voltage (V)

50

40

+

(Arbitrary Units)

Counts x 10

6

Ca

40

D1

451 pA
W

209

30

31.75 mm

20

No Gate

PMT Iris

10
0
180 185 190 195 200 205 210 215
Scanning Voltage (V)

7.5
40

(Arbitrary Units)

Counts x 10

5

7.0
6.5

+

Ca

D1

362 pA
220

W

12.7 mm
6.0
5.5

PMT Iris
Config B
No Gate

5.0
4.5
4.0
180

185

190

195

200

205

210

215

Scanning Voltage (V)

Figure 5.2: Typical 40 Ca+ resonance curves collected with the observed 8 V peak
to peak ripple in the FUG DC supply, with no electronic gates. The three ﬁgures
represent diﬀerent beam optics settings.
54

PMT Counts

FUG Ripple

Ch 1 5.00 V

Upper Gate

Ch 2 500 V
Ch 3 1.00 V

2.00 ms

(a) Upper Gate
PMT Counts

FUG Ripple

Ch 1 5.00 V

Lower Gate

Ch 2 500 V
Ch 3 1.00 V

2.00 ms

(b) Lower Gate
Figure 5.3: Snapshots of an oscilloscope trace showing the ripple of the FUG power
supply. The electronic gates that were applied to reduce the impact of the oscillation on the resonance curves are also shown, as well as the coincidence relationship
established by the DAQ for counting PMT signals.

55

5.1.2

Gated Resonance Spectra

Resonance spectra based on PMT signals that were gated on a limited region of the
DC voltage were collected for both the beam line aperture conﬁguration measurement
and PMT iris opening measurements. Two distinct resonance structures were observed
in the spectra (the secondary structure appearing as a low energy tail). The source
of the second resonance peak is still an open question, but may be a property of the
plasma conditions inside the ion source.
A typical resonance peak line shape for PMT signals collected using the upper and
lower electronic gates are shown in Figs. 5.4 and 5.5, respectively. The resonance line
shape is slightly asymmetric with a low-energy component. the line shape was ﬁtted
with a Voigt proﬁle that was convoluted with an exponential to phenomologically
model the observed low-energy tail.

Counts x 10

4

(Arbitrary Units)

5.0
4.5
4.0
3.5

40

+

Ca

D1

482 pA
216

W

0.25 x 0.5 in Iris
Upper Gate

3.0
2.5
2.0
1.5
1.0
0.5
0.0
180

185

190

195

200

205

210

215

Scanning Voltage (V)

Figure 5.4: Typical 40 Ca+ resonance curve collected with an electronic gate on the
upper ripple peak of the 15 kV DC acceleration power supply.

56

2.5

40

+

Counts x 10

5

(Arbitrary Units)

Ca

D1

448 pA
2.0

210

W

1.25 in Iris
Lower Gate
1.5

1.0

0.5

0.0
180

185

190

195

200

205

210

215

Scanning Voltage (V)

Figure 5.5: Typical 40 Ca+ resonance curve collected with electronic gate on the lower
ripple peak.
A more extreme example of the low-energy component of the resonance line shape
is shown in Fig. 5.6. The asymmetry of this resonance proved diﬃcult to ﬁt. Various
ﬁtting functions tried were: Voigt, Lorentz, Gauss, Voigt*exponential, Voigt + exponential, and multiple Voigt, but none were successful in reproducing the extreme low
energy tail. When such an extreme asymmetry was encountered, the line shape was
ﬁtted by restricting the ﬁt region to the portion of the line shape well above the tail.
Speciﬁcally, the low energy asymmetric region between 60% of the peak height and
within 1 sigma of the higher energy baseline value was ignored. A single Voigt ﬁt was
then applied to the remainder of the high-energy portion of the line shape as depicted
in Fig. 5.7.

57

Counts x 10

4

(Arbitrary Units)

40

+

Ca

10.0

D1

708 pA
W

170
8.0

0.25 x 1.0 in PMT Iris
Upper Gate

6.0

4.0

2.0

0.0
150

160

170

180

190

200

210

220

230

Scanning Voltage (V)

Figure 5.6: Typical 40 Ca+ resonance curves achieved for photon detection system
tests with electronic gate on upper ripple peak.

Data

Counts x 10

4

(Arbitrary Units)

40

+

Ca

10.0

Voigt Fit

D1

708 pA
170
8.0

W

0.25 x 1.0 in PMT Iris
Upper Gate

6.0

4.0

2.0

0.0
150

160

170

180

190

200

210

220

230

Scanning Voltage (V)

Figure 5.7: Typical 40 Ca+ resonance curves achieved with electronic gate on upper
ripple peak with the low-energy tail ignored by the ﬁt. The line is a Voigt proﬁle ﬁt.

58

5.1.3

S/N Extraction

The S/N was extracted from the resonance spectrum as detailed in Fig. 5.8. The
background (B) was the baseline of the resonance, which depended on stray laser
light levels. The noise (N) was the ﬂuctuation in the background, and the signal
(S) was the amplitude of the resonance peak above the background. The S/N was
normalized to both the beam current (II ) and laser power (IL ) to compare the results
of several independent measurements.

40Ca+
400-500pA
~150 uW
Config. D
Voigt*exp fit

S

N
B

Figure 5.8: S/N extraction from the 40 Ca+ D1 ﬂuorescence peak. Background (B) was
deﬁned as the non-resonant counts above zero, the signal (S) was the amplitude of the
peak above the background, and the noise (N) was the ﬂuctuation in the background.
S was normalized to give normalized signal (SN ) by:

SN =

S
.
II IL

Normalized background (BN ) was similarly obtained as:

59

(5.1)

BN =

B
.
II IL

(5.2)

Normalized noise (NN ) was obtained as:

NN = √

N
.
II IL

(5.3)

S/N was obtained from the normalized values as:

S/N =

5.2

SN
.
NN

(5.4)

Beam Line Aperture Conﬁguration Measurements

Five diﬀerent conﬁgurations of the inner drift tube beam line apertures were employed to characterize the S/N response and select an optimum aperture setting for
the beam/laser interaction region. Table 5.1 contains the primary data used to deduce S/N. Weighted averages of the relative S/N extracted for each conﬁguration are
shown in Fig. 5.9. Recall that the conﬁgurations A-C only had apertures upstream of
the reﬂector. Conﬁgurations D and E had apertures placed both upstream and downstream of the reﬂector (refer to Fig. 4.8 for conﬁgurations). Placement of apertures
downstream of the reﬂector increased S/N by at least one order of magnitude.
Backscattering of laser light from the end of the beam line, most likely from the
laser window, was found to be a signiﬁcant source of stray light. Black plating of
the inner surfaces of the chambers downstream of the photon detection system and
placing the laser window at a non-normal angle to the beam are both possible ways
to mitigate the scattering of laser light back into the reﬂector that can be tested in
the future.

60

Table 5.1: Primary data for extraction of S/N as a function of beamline aperture
conﬁguration.
A
A
B
B
C
C
D
D
E
E

S
31000
44000
4200
3600
12000
10000
37000
29000
22600
23400

δS
9000
9000
400
600
3000
2000
3000
1300
700
610

B
49600
12930
5050
4857
12380
13510
770
820
2220
2640

N
100
90
17
18
40
11
17
11
14
10

II
433
461
330
395
305.0
311.8
413
376
526
561

δII
3
2
2
1.2
0.9
0.8
4
3
4
3

IL
212
218.1
194
187.6
184
181
208
204
188
194

δIL
5
0.7
1.1
0.6
2
1.2
1.0
4
1.8
1.8

S/N
1.1
1.6
0.91
0.8
1.2
0.8
7.4
9.2
5.3
6.9

δS/N
0.3
0.3
0.08
0.13
0.3
0.19
0.6
0.5
0.18
0.19

Relative S/N /pA/

W

9
8
7
6
5
4
3
2
1
0

A

B

C

D

E

Beamline Aperture Configuration

Figure 5.9: Experimental S/N as a function of beam line aperture conﬁguration. S/N
is corrected by beam current and laser power.

61

5.3

PMT Iris Area Measurements

The PMT iris opening area at the second focal plane of the ellipsoidal reﬂector was
varied and relative S/N was deduced from the observed 40 Ca+ resonance curves, again
with the goal of optimizing the photon detection system performance. The relative
S/N values for the various PMT iris opening areas are presented in Fig. 5.10. The
ﬁlled circles denote results obtained with circular irises and are labeled by diameters.
The S/N results for the rectangular irises are denoted by ﬁlled square symbols. The
two rectangular iris data points are labeled by the length and width of the rectangular
irises, in millimeters, where the longer dimension is along the beam axis. Systematic
errors due to laser power ﬂuctuation were not included in the error bars presented in
Fig. 5.10.

Relative S/N /pA /

W

7

31.75

6

6.35
x
25.4

5

6.35
x

4
3

6.35

12.7

12.7

25.4
19.05

2

0 100 200 300 400 500 600 700 800 900
PMT Iris Area (mm

2

)

Figure 5.10: Initial experimental S/N as a function of PMT iris area. S/N is corrected
by beam current and average laser power.
A second set of measurements with the diﬀerent irises was completed where the
laser power was monitored by splitting the laser beam and monitoring any ﬂuctuations

62

with a power meter. The average laser power and the one sigma ﬂuctuation in laser
power for each run are shown in Fig. 5.11.

Laser Power (

W)

130
129

6.35
x
12.7
25.4
6.35
x

128

25.4
19.05

127
126
125

31.75

12.7

6.35

0 100 200 300 400 500 600 700 800 900
PMT Iris Area (mm

2

)

Figure 5.11: Laser power and ﬂuctuation measured during the second run of the PMT
iris measurements, note the suppressed zero on vertical scale.
The S/N was deduced for each iris opening using the primary data listed in Table
5.2. The S/N results are shown in Fig. 5.12. The experimental data are presented as
in Fig. 5.10, and the simulation results (see Fig.2.4) are denoted by the dashed line.
The solid line in Fig. 5.12 is a phenomological ﬁt to the experimental results. The
additional normalization of the S/N by the laser power ﬂuctuations reduced the systematic error in the measurements and allowed better comparison of the experimental
data to the simulation results.
It was reasoned that the ﬂuorescence light has a distribution that is well focused
at the second focal point of the ellipsoidal reﬂector based on the trends observed in
the experimental data. The largest of the PMT iris openings was too small to observe
the expected stray light distribution obtained from simulation (see Fig. 2.2) due to
constraints introduced by the inverted viewport window. The stray light distribution
63

Table 5.2: Primary data for extraction of S/N as a function of PMT iris open area.
Iris Area
(mm2 )
31.67
31.67
80.64
126.68
161.29
285.02
506.71
791.73

S

δS

B

N

II

δII

IL

δIL

S/N

δS/N

27000
27000
165000
250000
249000
87000
64000
94000

1400
1500
8000
13000
13000
6000
3000
6000

1550
1510
15700
26500
33900
5750
5710
8580

40
40
120
160
180
80
80
90

1093
1080
1050
990
910
820
756
708

4
10
5
14
15
14
7
5

128.1
126.2
127.4
129.1
129.0
126
128.3
128.5

0.9
0.8
0.9
0.9
0.8
1
0.7
0.7

1.9
1.9
2.7
3.6
3.4
3.6
4.2
3.9

0.3
0.3
0.4
0.5
0.5
0.4
0.5
0.5

Experiment

Fit

Simulation

Relative S/N /

W /pA

5.0
4.5

25.4

4.0
31.75

12.7

3.5

6.35

3.0

x

6.35

2.5

19.05

12.7

25.4

x

2.0

6.35

1.5
1.0
0

100 200 300 400 500 600 700 800 900

PMT Iris Area (mm

2

)

Figure 5.12: Experimental S/N as a function of PMT iris opening area. The S/N was
corrected by beam current, laser power and laser power ﬂuctuation.

64

was assumed to be uniform over the measured second focal plane region. The linear
increase in S/N with an increase in the PMT iris opening area due to the drop-oﬀ in
S/N for larger iris openings after the signal had saturated, and also due to the trend
observed in the background levels as a function of PMT iris opening area.
The PMT iris openings with diameters between 12.7 to 31.75 mm gave the same
S/N, within the reported error. The function ﬁtted to the data peaks at ∼20 mm
for a circular iris opening. Therefore the 19.05 mm diameter circular iris was selected
for general use in the “blue” photon detection system. Experimental results matched
expectations from simulations.

65

Chapter 6
Summary and Outlook
A photon detection system was developed to detect ﬂuorescence photons from collinear
laser spectroscopy experiments performed at BECOLA. Two systems for “red” and
“blue” light were developed based on an initial design from the University of Mainz.
The two systems were optimized for use in BECOLA via ray trace simulations. Commissioning measurements of the “blue” system were performed with Ca+ to verify
simulation and optimize sensitivity of the photon detection system.
The 40 Ca+ D1 resonance lineshape was measured for a variety of beam line aperture conﬁgurations and PMT iris openings. S/N ratios were deduced for optimization
of the “blue” photon detection system. The optimum beam line aperture conﬁguration was found to be “D” in Fig. 2.5, which had 7 and 8 mm apertures in the inner
ground drift tubes both upstream and downstream of the ellipsoidal reﬂector. The
optimum PMT iris opening size was found to be the 19.05 mm diameter circular iris.
The “blue” photon detection system, at the optimum setting, has a sensitivity of 1
photon per 128 ions when corrected by the electronics gating time, which corresponds
to ion beam rates of about 1 × 105 ions per second.
Commissioning measurements for the “red” photon detection system will be carried out in the near-term with a 39 K atomic beam.

66

BIBLIOGRAPHY

67

Bibliography
[1] H.-J. Kluge and Wilfried N¨rtersh¨user. Spectrochimica Acta B, 58:1031, 2003.
o
a
[2] L. Weissman et al. Nucl. Instr. and Meth. Phys. Res. A, 540:245, 2005.
[3] K. Minamisono et al. WSPC Conf. Proc., April 2008.
[4] Erst W. Otten. Nuclear radii and moments of unstable isotopes. In Treatise on
Heavy-ion Physics, volume 8. Plenum Press, New York, 1987.
[5] K. Shimoda. High-resolution laser spectroscopy. In Topics in Applied Physics,
volume 13. Springer-Verlag, Heidelberg, 1976.
[6] Gordon W. F. Drake (Ed.). Handbook of Atomic, Molecular and Optical Physics.
Springer, New York, 2005.
[7] Christopher J. Foot. Atomic Physics. Oxford University Press, New York, 2005.
[8] C. Schwartz. Phys. Rev., 97(2):380, 1947.
[9] S.L. Kaufman. Optics Communications, 17(3):309, 1976.
[10] William H. Wing et al. Phys. Rev. Lett., 36(25):1488, 1976.
[11] N. Bendali et al. J. Phys. B, 19:233, 1986.
[12] Colutron research corporation. http://www.colutron.com/products/kit/kit.html.
[13] Ron
Fox.
Integrating
controls
and
http://docs.nscl.msu.edu/daq/appnotes/Controls.pdf.

event

data.

[14] Epics. http://www.epics.org/.
[15] Photon engineering. http://www.photonengr.com.
[16] J. Kr¨mer. Construction and Commissioning of a Collinear Laser Spectroscopy
a
Setup at TRIGA Mainz. PhD thesis, Johannes Gutenberg-Universit¨t Mainz,
a
2010.
[17] N.G. Mikhailus. Zhurnal Priikladnoi Spekrtoskopii, 11(3):552, 1969.
[18] M. Hammen. Development, construction and test of an eﬃcient ﬂuorescence
detection region for collinear laser spectroscopy at triga-mainz. Master’s thesis,
Johannes Gutenberg-Universit¨t Mainz, 2010.
a
68

[19] Micro-mesh. http://www.sisweb.com/micromesh/.
[20] Mother’s power ball mini. http://www.mothers.com/.
[21] Simichrome. http://www.competitionchemicals.com/simichrome/.

69