Review of polarized ion sources (invited).

Abstract

Recent progress in polarized ion sources development is reviewed. New techniques for production of polarized H(-) ion (proton), D(-) (D(+)), and (3)He(++) ion beams are discussed. Feasibility studies of these techniques are in progress at BNL and other laboratories. Polarized deuteron beams will be required for the polarization program at the Dubna Nuclotron and at the deuteron electric dipole moment experiment at BNL. Experiments with polarized (3)He(++) ion beams are a part of the experimental program at the future electron ion collider.

Full text

Page 1

Development of a pepper-pot device to determine the emittance of an ion
beam generated by electron cyclotron resonance ion sourcesa…
M. Strohmeier,1,2,b?J. Y. Benitez,1D. Leitner,1C. M. Lyneis,1D. S. Todd,1and M. Bantel2
1Lawrence Berkeley National Laboratory, One Cyclotron Road, Berkeley, California 94720, USA
2University of Applied Sciences Karlsruhe, Moltkestr. 30, 76133 Karlsruhe, Germany
?Presented 24 September 2009; received 21 September 2009; accepted 1 October 2009;
published online 18 February 2010?
This paper describes the recent development and commissioning of a pepper-pot emittance meter at
the Lawrence Berkeley National Laboratory ?LBNL?. It is based on a potassium bromide ?KBr?
scintillator screen in combination with a charged coupled device camera. Pepper-pot scanners record
the full four-dimensional transverse phase space emittances which are particularly interesting for
electron cyclotron resonance ion sources. The strengths and limitations of evaluating emittances
using optical pepper-pot scanners are described and systematic errors induced by the optical data
acquisition system will be presented. Light yield tests of KBr exposed to different ion species and
first emittance measurement data using ion beams extracted from the 6.4 GHz LBNL electron
cyclotron resonance ion source are presented and discussed. © 2010 American Institute of
Physics. ?doi:10.1063/1.3258024?
I. INTRODUCTION
Electron cyclotron resonance ?ECR? ion sources are
widely used in the particle accelerator community because
they are capable of producing high current beams of highly
charged ions. Their operation relies on a magnetically con-
fined plasma which is heated by microwave radiation. In
order to achieve high charge states, long confinement times
are needed which is achieved by the superposition of an axial
mirror field and a radial multi-pole field, resulting in a mini-
mum B-field configuration. Since the extraction aperture is
located close to the peak of the extraction solenoid coil, par-
ticles are accelerated in the presence of the strong axial field,
which increases their transverse emittance.
In order to design an accelerator with a maximum par-
ticle throughput, the acceptance of the beam line has to be
greater than the beam emittance. To measure the beam emit-
tance, two different types of emittance scanners are currently
in operation at LBNL: a slit scanner1?Allison type? and a
pepper-pot scanner.2,3,6The slit scanner can provide a better
spatial and angular resolution than the pepper-pot, but mea-
surements take in the order of a few minutes. Additionally,
the slit scanners can only extract a one-dimensional data set,
since the intensities are integrated over the whole slit while
stepping through the beam. Scintillator based pepper-pot
scanners capture the image data in just a fraction of a second,
making it less vulnerable to emittance changes caused by
plasma instabilities or other transitions. Furthermore, the
image array of a pepper-pot scanner provides a two-
dimensional data set from which the xx? and yy? as well as
the cross coupled xy? and yx? phase spaces can be extracted.
Since the pepper-pot scanner uses a camera to capture
the light pattern on the scintillator, it is important to investi-
gate the influence of optical parameters on the emittance. For
the scintillator, the absolute light yield, its linearity and the
lattice degradation as a function of the beam exposure time
are crucial parameters.4,5Camera settings influencing the
emittance value are the exposure time of the charge coupled
device ?CCD? chip to the scene, the gain and the brightness.
II. PEPPER-POT HARDWARE CONFIGURATION
In Fig. 1 the prototype design of the pepper-pot device
which was recently commissioned in the ECR beam line at
LBNL is shown. For the initial tests, a 50 ?m thick tantalum
foil with holes of 76 ?m diameter and a regular spacing of
1.45 mm in both x and y direction is used.
The holemask is mounted in an aluminum frame and has
one blocked hole in the center of the foil to provide an ab-
solute spatial reference for the data evaluation. Because of
the low beam power during the initial tests, no additional
cooling of the mask was needed in order to prevent its ther-
mal deformation. A round potassium bromide ?KBr? single
crystal disk ?80 mm diameter with 5 mm thickness? located
25 mm behind the holemask is used as a scintillating surface.
A 45° mirror projects the appearing light pattern toward a
CCD camera looking perpendicular at the beam. The expo-
sure time, gain and brightness of the camera can be adjusted
online via a LABVIEW user interface and a real-time image
a?Contributed paper, published as part of the Proceedings of the 13th Inter-
national Conference on Ion Sources, Gatlinburg, Tennessee, September
2009.
b?Electronic mail: mmstrohmeier@lbl.gov.
FIG. 1. ?Color online? Photograph of the pepper-pot device with the KBr
scintillator behind the holemask.
REVIEW OF SCIENTIFIC INSTRUMENTS 81, 02B710 ?2010?
0034-6748/2010/81?2?/02B710/3/$30.00 © 2010 American Institute of Physics
81, 02B710-1
Author complimentary copy. Redistribution subject to AIP license or copyright, see http://rsi.aip.org/rsi/copyright.jsp

Page 2

histogram allows for the optimal CCD illumination. The
pepper-pot scanner is mounted in a 6-way conflat sealed
cube close to the focal plane of the mass analyzer.AnAllison
type scanner was recently added to the same cube in order to
compare the measurements of both devices. To prevent col-
lisions, the motion of each scanner is cross interlocked in the
LABVIEW control system. A schematic drawing of this setup
is shown in Fig. 2.
III. EVALUATION SOFTWARE
The evaluation software is capable of extracting the full
four-dimensional phase space information from the captured
light pattern. The necessary program steps are described in
the following.
A. Noise treatment
First, pixel defects on the CCD chip are detected by
recording an image without beam exposure and a high cam-
era gain. The value of each damaged pixel is subsequently
replaced by the average of its eight surrounding neighbors.
Next, the noise level of the image is calculated as the average
pixel value, taken over a dark image region multiplied by 1.2
?empirically determined?. After zeroing all pixels below the
threshold value, the data array is flattened with a 3?3
smoothing filter.
B. Spot detection and assignment to a hole
In the next step, the program generates a grid which is
overlaid onto the image. The squared grid is generated by
centering the first cell over the blocked hole which absolute
coordinates have to be determined during the device calibra-
tion. Each grid cell represents the sensitive area for a par-
ticular hole i in the mask where the software searches for
intensity. Since different beam properties cause different
light patterns, the grid can be adjusted in terms of cell size
and cell position. This user interaction assures that cell bor-
ders do not cut through spots which would result in a false
hole-spot assignment. Tests with simulated data sets have
shown that the precise knowledge of the absolute coordinates
of the blocked hole is not crucial. It was found that an un-
certainty of ?0.4 mm introduces a systematic error of less
than ?3% when evaluating the divergence angles. This un-
certainty mainly causes a ?small? shift in the phase space
ellipse relative to the true beam position.
C. Calculation and documentation
The location of the blocked hole regarding the rows and
columns of the hole mask is defined with the index ?0,0?. By
counting the cell indices while stepping through the detected
spots, the program is able to clearly relate a spot i on the
image to its corresponding hole i in the mask, located at
?xi?yi?. As both the x and y coordinates of the hole i are well
known, the x? and y? distribution for every single spot can be
deduced. This is achieved by fixing the image coordinate in
one direction and integrating the intensity over the perpen-
dicular one. The obtained emittance numbers and twiss pa-
rameters are saved in a text file along with the settings for the
beam line and ion source.
IV. CALIBRATION AND ANALYSIS
A. Determination of the blocked hole
As mentioned above, the algorithm needs an absolute
reference on the image to start counting the row and column
indices. In the present case, a blocked hole serves as this
reference and its location is determined by simply focusing
the camera in the plane of the holemask and reading out the
?x?y? pixel coordinates.
B. Angular resolution for several configurations
The angular resolution of the pepper-pot device depends
on the spacing between the holemask and the scintillating
surface. Table I shows the minimum angle xmin
be resolved for a given holemask-scintillator distance d. The
physical pixel size on the chip is 4.6?4.6 ?m2. In the cur-
rent setup this distance can be varied in steps of 5 mm start-
ing from 10 up to 70 mm. Using a CCD camera with a
smaller pixel size increases the angular resolution.
In addition, with increasing d, the absolute spot size in-
creases. Therefore, individual spots are represented by more
pixels and better resolved in the final image. In practice how-
ever, for a given holemask, d mainly depends on the ex-
pected maximum angle xmax
?
that beam particles have. The
scintillator has to be located close enough to the holemask
that spots do not overlap but far enough to ensure a good
spot resolution. Ideally the distance would be remotely ad-
justable to maximize the resolution for each measurement.
?
that can still
C. Linearity of the light yield
Three scintillator materials were tested, potassium bro-
mide ?KBr?, barium fluoride ?BaF2? and quartz. Potassium
bromide had the highest light yield per incident particle and
was chosen as the scintillator for the commissioning. A long-
term exposure test revealed that the light yield for all three
materials decreases when bombarded with beam. This is
most likely due to interactions of the impinging ions with the
scintillator which alter the lattice close to the scintillator sur-
face. Figure 3 shows this degradation for different scintillator
materials, all being exposed to a 10 keV, 100 e? A, O3+
beam. All materials show a drop in their initial light yield by
TABLE I. The angular resolution of the device for a physical pixel size of
4.6 ?m. Greater distances d result in a better angular resolution.
d ?mm?
xmin
?
?mrad?
25
50
75
0.184
0.092
0.046
FIG. 2. ?Color online? Schematic layout of the pepper-pot scanner and the
Allison type scanner, both located at the same z-location.
02B710-2Strohmeier et al.Rev. Sci. Instrum. 81, 02B710 ?2010?
Author complimentary copy. Redistribution subject to AIP license or copyright, see http://rsi.aip.org/rsi/copyright.jsp

Page 3

a factor of 2 after being exposed to the beam for 20 min.
Therefore, the exposure time of the scintillator to the beam
has to be minimized during a measurement. The original
light yield can be recovered if the crystal is repolished.
Another important factor for accurate emittance mea-
surements is the linear dependence of the generated light on
the incident beam current. If an intense region of the beam
cross section generates proportionally less light due to scin-
tillator saturation effects, the contribution of weaker beam
regions to the emittance would be overestimated and would
have to be compensated. For the KBr scintillator, this depen-
dency was investigated with an oxygen and argon beam. For
both ion species, Fig. 4 shows a linear increase in the total
generated light with the incident ion beam current. Further-
more, comparing the light yield for different charge states of
one ion species, it can be seen that the light generation in-
creases with the charge state since the ions have more kinetic
energy. Within the measured intensity range no saturation
effects were found.
V. EMITTANCE MEASUREMENTS
First emittance measurements for the xx?-plane were per-
formed with the pepper-pot scanner in the ECR beam line.
For an extraction voltage of 10 kV, Fig. 5 shows the depen-
dence of the 1 rms norm. emittance after a 5–10% intensity
cut. As measured by other groups,1,7the emittance for
heavier ions is smaller compared to lighter ions for the same
m/q-ratio. From this measurement, one might conclude that
the ion temperature for oxygen is higher than for argon.
Other groups also measured a decreasing emittance for
higher charge states of argon. This could be explained with
the model that highly charged ions are created and extracted
closer to the source axis.1Extracting these highly charged
ions from a region close to the axis keeps their transverse
momentum growth small. However it is not clear why oxy-
gen does not show the same trend.
VI. OUTLOOK
Future work on the pepper-pot scanner will focus on
improving the alignment and mechanical design of the de-
vice. Systematic measurements on the ECR ?6.4 GHz? and
AECR-U ?14 GHz? are planned to investigate the influence
of the magnetic confinement field on the emittance.
ACKNOWLEDGMENTS
This research was conducted at LBNL and was sup-
ported by the Director, Office of Energy Research, Office of
High Energy and Nuclear Physics, Nuclear Physics Division
of the U.S. Department of Energy under Contract No. DE
AC03-76SF00098. The emittance data presented in this
paper was taken together with students attending the Eighth
Summer School on Exotic Beam Physics 2009 in Berkeley,
CA.
1D. Wutte, M. A. Leitner, and C. M. Lyneis, Phys. Scr. T92, 247 ?2001?.
2A. Pikin, A. Kponou, J. Ritter, and V. Zajic, “Pepper pot emittance meter,”
BNL Report No. C-A/AP/#244, July 2006.
3S. A. Kondrashev, A. Barcikowski, B. Mustapha, P. N. Ostroumov, R. H.
Scott, and S. I. Sharamentov, Proceedings of the PAC07, Albuquerque,
New Mexico, 2007 ?unpublished?.
4M. Vogel, D. F. A. Winters, H. Ernst, H. Zimmermann, and O. Kester,
Nucl. Instrum. Methods Phys. Res. B 263, 518 ?2007?.
5S. A. Kondrashev, A. Barcikowski, B. Mustapha, P. N. Ostroumov, and N.
Vinogradov, Nucl. Instrum. Methods Phys. Res. A 606, 296 ?2009?.
6H. R. Kremers, J. P. M Beijers, and S. Brandenburg, A Versatile Emittance
Meter and Profile Monitor ?KVI, Groningen, 2007?.
7T. Thuillier, J. C. Curdy, T. Lamy, A. Lachaize, A. Ponton, P. Sole, P.
Sortais, and J. L. Vieux-Rochaz, Proceedings of the ECRIS04 Berkeley,
2005.
FIG. 3. ?Color online? Degradation of different scintillator material over
time. The data is normalized to the respective starting intensity.
(a) oxygen
(b) argon
FIG. 4. ?Color online? The light yield for a 10 keV ?a? oxygen and ?b? argon
beam depends linearly on the incident particle current. Higher charge state
ions generate more light since the particles carry more kinetic energy.
FIG. 5. ?Color online? The initially increasing emittance for oxygen at a
small m/q-ratio can be an optical effect of the image acquisition.
02B710-3Strohmeier et al.Rev. Sci. Instrum. 81, 02B710 ?2010?
Author complimentary copy. Redistribution subject to AIP license or copyright, see http://rsi.aip.org/rsi/copyright.jsp