Electron Cloud Effects at Positron/Electron (e+/e~) Machines
and Electron Cloud Diagnostics
K. C. Harkay*, R. A. Rosenberg*, R. J. Macek*, A. Browman*, T.-S. Wang*
^Argonne National Laboratory, Argonne, IL, 60439 USA
*Los Alamos National Laboratory, Los Alamos, NM, 87544 USA
Abstract. Background electrons are ubiquitous in high-intensity particle accelerators. Under certain operating conditions, amplification of the electron cloud can occur. The beam-cloud interaction can seriously degrade the accelerator
performance with effects that range from vacuum degradation to collective beam instabilities. Although electron cloud
effects (ECEs) were first observed 20 years ago in a proton ring, in recent years, they have been widely observed and
intensely studied in e+/e- rings. This paper will focus on describing electron cloud diagnostics, which have led to an
enhanced understanding of ECEs, especially details of beam-induced multipacting and saturation of the cloud. Such
experimental results can be used to provide realistic limits on key input parameters for modeling efforts.
INTRODUCTION
OVERVIEW OF EC EFFECTS
Interactions between high-intensity particle beams
and a background population of low-energy electrons,
known as the electron cloud (EC), have been widely
reported1. The mechanisms giving rise to the cloud
vary1'2, and in many cases, the cloud does not affect
the beam. However, amplification of the cloud can
occur under certain operating conditions, potentially
giving rise to numerous effects that can seriously
degrade accelerator performance. First observed in
proton rings, electron cloud effects (ECEs) have in
recent years been intensely studied in high-energy
positron and electron rings as well. Mitigating ECEs
remains an important concern in present and future
high-intensity, high-energy accelerators.
Many comprehensive review papers can be found
in the proceedings of recent U.S. and European
Particle Accelerator Conferences, giving a summary of
experimental observations and modeling of electron
cloud effects1. Only the briefest of discussions is given
here. We will focus on two phenomena observed in
both APS and PSR: beam-induced multipacting (BIM)
and the build-up and saturation of the cloud. BIM is a
resonance condition that can lead to large amplification of the EC if the secondary electron emission
coefficient (8) of the chamber surface is greater than
unity. BIM can also be responsible for vacuum
degradation due to electron-stimulated gas desorption,
first observed at the CERN proton Intersecting Storage
Ring (ISR)3. Depending on machine conditions, the
cloud density can be observed to either grow or remain
constant over a bunch train (or several turns). In either
case, the cloud density appears to eventually reach
saturation at a level related, in some cases nonlinearly,
to the average beam charge density.
An important tool in an improved understanding of
ECEs has been the development of electron cloud
diagnostics. Experimental data characterizing the EC
properties can be used to provide realistic limits on
key input parameters for modeling and analytical
calculations. Combining EC and standard beam diagnostics allows EC effects to be separated from those
arising from the geometric wake field. Although the
nature of the beam-cloud interaction is rather different
in positron vs. proton machines, there is a potential for
synergy in the techniques applied to study the EC in
both. This will be illustrated using results from the
Advanced Photon Source (APS) at Argonne National
Laboratory (ANL) and the Proton Storage Ring (PSR)
at Los Alamos National Laboratory (LANL).
At APS4, EC amplification and associated pressure
rise due to BIM were clearly observed with positron
beams at certain bunch currents and spacings (the Al
vacuum chambers have a relatively high 8). A more
modest effect is observed with electron beams in
present APS operation. At BIM conditions, the cloud
density saturates after 20-30 bunches. With positron
beams only, a horizontal coupled-bunch instability was
observed at conditions coinciding with the maximum
saturated cloud density. At PSR5, a long observed e-p
instability occurs at a threshold bunch intensity and
buncher voltage. BIM is observed on the trailing edge
(tail) of the bunch, but these prompt electrons appear
to have a less direct relationship to the instability. On
the other hand, electrons that survive the gap are
observed to saturate near the instability threshold.
* Work supported by U.S. Department of Energy, Office of Basic
Energy Sciences under Contract No. W-31-109-ENG-38.
[email protected], [email protected]
* Work conducted at Los Alamos National Laboratory, operated by
the University of California for the U.S. Department of Energy
under Contract No. W-7405-ENG-36.
[email protected], [email protected], [email protected]
CP642, High Intensity and High Brightness Hadron Beams: 20th ICFA Advanced Beam Dynamics Workshop on
High Intensity and High Brightness Hadron Beams, edited by W. Chou, Y. Mori, D. Neuffer, and J.-F. Ostiguy
© 2002 American Institute of Physics 0-7354-0097-0/02/$ 19.00
354
and
and PSR,
PSR, RFAs
RFAs were
were installed
installed in regions free from
external
external magnetic
magnetic fields.
fields. At
At CERN
CERN SPS, an RFAbased
based detector
detector was
was installed
installed inside
inside a speciallyconstructed dipole
dipole magnet,
magnet, and multi-strip collector
constructed
electrodes measure
measure the
the transverse
transverse EC spatial density7.
electrodes
ELECTRONCLOUD
CLOUDDIAGNOSTICS
DIAGNOSTICS
ELECTRON
Theplanar
planar retarding-field
retarding-field analyzer
analyzer (RFA),
(RFA), first
first
The
developedand
andimplemented
implementedatatAPS
APS6,6,isisaavery
verysimple
simple
developed
device that
that measures
measures electrons
electrons colliding
colliding with
with the
the
device
chamber
wall.
The
RFA
consists
of
two
grids
and
chamber wall. The RFA consists of two grids and aa
graphite-coatedcollector
collector (to
(tolower
lower its
its 8).
δ). The
The outer
outer
graphite-coated
gridisisgrounded,
grounded,while
whilethe
theinner
innergrid
gridcan
canbe
bebiased
biasedfor
for
grid
electronenergy
energyselection.
selection.The
TheRFA
RFAisis an
an integrating
integrating
electron
device;differentiating
differentiatingthe
thecollector
collectorsignal
signalgives
givesthe
theEC
EC
device;
energydistribution.
distribution.Figure
Figure11shows
showsaaschematic
schematicof
oftwo
two
energy
RFAsmounted
mountedon
onaastandard
standardAPS
APSchamber.
chamber.
RFAs
To detect
detect electrons
electrons in the gap after
after the bunch
To
passage, an
an electron
electron sweeper
sweeper was developed at PSR8.
passage,
curved electrode
electrode subtending
subtending a half-angle of 75º
AAcurved
75Q (~40
(-40
cm long)
long) isis placed
placed opposite
opposite a large-aperture RFA (~8×
cm
(~8x
higher sensitivity
sensitivity than
than the
the original
original design), shown in
higher
Fig. 3.
3. A
A short
short (-20
(~20 ns), -500
~500 V pulse on the electrode
Fig.
selects the
the sampling
sampling time
time in the gap; the fractional
selects
acceptance area
area of
of electrons swept into the RFA is 0.3.
acceptance
Collector
Collector
Repeller
Repeller
Grid
Grid
FIGURE1:1:ANL
ANL schematic
schematic ofof two
two RFAs
RFAs mounted
mounted on
on aa
FIGURE
standard APS chamber, cross-sectional view (42 × 21 mm
standard
APS chamber, cross-sectional view (42 x 21 mm
chamber half-dimensions).
chamber half-dimensions).
Slots &
&
Slots
Screen
Screen
Thebenefit
benefitofofshielding
shieldingthe
thecollector
collector and
andbias
bias grid
grid
The
is that the measurement does not disturb the EC in the
is that the measurement does not disturb the EC in the
chamber. It is very difficult to measure the true wall
chamber. It is very difficult to measure the true wall
flux or energy distribution with a biased beam position
flux or energy distribution with a biased beam position
monitor (BPM), illustrated in Fig. 2. Secondary elecmonitor (BPM), illustrated in Fig. 2. Secondary electron emission from the BPM surface affects the wall
tron emission from the BPM surface affects the wall
flux (the signal can even change sign). Also, the BPM
flux (the signal can even change sign). Also, the BPM
bias changes the collection length (i.e., volume).
Pulsed
Pulsed
Electrode
Electrode
FIGURE 3:
3: LANL
LANL electron
FIGURE
electron sweeper
sweeper (chamber
(chamber diam.
diam.10
10cm).
cm).
A number
number of
of EC
A
EC diagnostics
diagnostics concepts
concepts have
have been
been
proposed
or
are
under
development,
either
proposed or are under development, either9 for
for better
better
energy resolution (Bessel Box at APS) 9, or less
energy
resolution (Bessel Box at APS) , or less
perturbative detection of the electrons in the vacuum
perturbative detection of the electrons in the vacuum
chamber (e.g., by G. Lambertson (LBNL) and S.
chamber (e.g., by G. Lambertson (LBNL) and S.
Heifets (SLAC)). At KEKB, the EC density is inferred
Heifets (SLAC)). At KEKB, the EC density is inferred
indirectly by comparing the bunch-by-bunch spaceindirectly by comparing the bunch-by-bunch spacecharge tune shift with and without the EC.
bias changes the collection length (i.e., volume).
1.5 -
<
(C)
1.0
-- 80eV
— 60 eV
charge tune shift with and without the EC.
0.5
EC MEASUREMENTS AT APS
EC MEASUREMENTS AT APS
0.0
-100
-50
0
In experiments44 where the bunch spacing was
In
where the bunch
spacingcomwas
varied, experiments
the primary (photoelectron)
vs. secondary
varied,
the
primary
(photoelectron)
vs.
secondary
components of the EC production processes could be sepaponents
of thethe
ECRFA
production
processes
be separated. While
only measures
thecould
electron
wall
rated.
While
the
RFA
only
measures
the
electron
wall
flux, analysis of the EC energy spectra gives an indiflux,
analysis
of
the
EC
energy
spectra
gives
an
indirect measure of the electrons near the beam; a highrect
measure
of thefrom
electrons
nearnear
the the
beam;
a highenergy
tail results
electrons
beam
that
energy
tail
results
from
electrons
near
the
beam
that
are accelerated to higher energy. The BIM resonance
are
accelerated
to
higher
energy.
The
BIM
resonance
at 7 λrf and variation in EC distributions can be seen in
at 7 Arf
EC distributions
can be seen
Fig.
4. and
Thevariation
collector incurrent
Ic is differentiated
within
Fig.
4.
The
collector
current
I
is
differentiated
with
c
respect to the bias voltage and normalized to the total
respect
to the Ibias
voltage
and
normalized
to
the
total
beam current
.
In
Fig.
5,
the
measured
EC
build-up
b
beamsaturation
current Ibover
. In positron
Fig. 5, the
measured
build-up
and
bunch
trains EC
at the
BIM
and saturation
positron
bunch trains
at the
BIM
bunch
spacing isover
shown
as a function
of bunch
current.
bunch spacing is shown as a function of bunch current.
50
Energy (eV)
FIGURE 2: Signal produced from a BPM irradiated by 60FIGURE
2: Signal
produced
from a BPM
irradiated
by 60eV and 80-eV
electrons
as a function
of bias
voltage applied
6 electrons as a function of bias voltage applied
eV
andBPM
80-eV
to the
.
to the BPM6.
ANL-type RFA devices have been widely
ANL-type Intense
RFA Pulsed
devicesNeutron
have Source
been (ANL),
widely
implemented:
implemented:
Intense
Pulsed
Neutron
Source
(ANL),
PSR (LANL), Beijing Electron-Positron Collider
PSR
(LANL),
Beijing (KEK),
Electron-Positron
(IHEP),
KEK B-factory
Super ProtonCollider
Synch(IHEP),
KEK B-factory
(KEK), Gradient
Super Proton
Synchrotron (CERN),
and Alternating
Synchrotron
rotron
andPSR,
Alternating
Synchrotron
booster(CERN),
(BNL). At
the RFAGradient
was augmented
with
booster
(BNL).
At PSR,tothe
RFA
was augmented
with
a broadband
amplifier
allow
time-resolved
measureaments.
broadband
allow
time-resolved
Alsoamplifier
at PSR, tothe
effect
of chambermeasuresurface
ments.
Also
at
PSR,
the
effect
of
surface
coatings with lower δ was measuredchamber
in situ. At
APS
coatings with lower 8 was measured in situ. At APS
355
near the peak and in the tail of the bunch. These
< 1 000
<J 1 00
<
10
"-•-....
V
"*'"-v---v..^^
>-f
1
X
7
'"""'•••""•••••--........ :
•s:w:^^^^5ia,;
1 2
1
f
o
°-
i
0.01
1
50
1 00
1 50
200
250
SUMMARY
SUMMARY
Electron cloud diagnostics based on the planar
300
electron energy (eV)
FIGURE 4: EC energy distribution vs. bunch spacing (units
of
rf wavelength,
A^); 2 mA/bunch,
bunches.
FIGURE
4: EC energy
distribution 10
vs. positron
bunch spacing
(units
of rf wavelength, λrf); 2 mA/bunch, 10 positron bunches.
current per bunch
2mA
1 0
<
1 .5 mA
1
mA
O
E
0.1
0
1 0 20 30 40 50 60
number of bunches in train, N.
FIGURE 5: Measured EC build-up and saturation over
FIGURE
5: Measured
ECBEVI
build-up
and saturation
over
positron
bunch
trains at the
bunch spacing.
A horizontal
positron
bunch
trains
at
the
BIM
bunch
spacing.
A
horizontal
coupled-bunch instability occurs for trains longer than ~20
coupled-bunch instability occurs for trains longer than ~20
bunches with a threshold current of about 2 mA/bunch.
bunches with a threshold current of about 2 mA/bunch.
EC
AT PSR
PSR
EC MEASUREMENTS
MEASUREMENTS AT
Extensive
measurements
have been
been carried
carried out
out atat
Extensive
measurements have
5
5 to study the EC properties and correlation with
PSR
PSR to study the EC properties and correlation with
the
electron sweeper
sweeper data
dataare
areparticparticthee-p
e-p instability.
instability. The
The electron
ularly
interesting;
a
typical
result
is
shown
in
Fig.
ularly interesting; a typical result is shown in Fig. 6.6.
The
are: 7.7
7.7 µC/pulse
|iC/pulseand
and280-ns
280-ns
The machine
machine conditions
conditions are:
bunch
on the
the trailing
trailing edge
edge of
of the
the beam
beam
bunch length.
length. BIM
BIM on
accelerates
electrons that
that strike
strikethe
the wall
wall
accelerates and
and amplifies
amplifies electrons
1.5
Beam Pulse
1
Amplitude(V)
0.5
HV Pulse
0
-0.5
-1
-1.5
Swept Electron Signal
-2
Prompt Electron Signal
-2.5
100
200
300
400
prompt
for populatnear
the electrons
peak andare
in believed
the tail responsible
of the bunch.
These
ing theelectrons
gap butarethe
relationship
is nonlinear.
The
prompt
believed
responsible
for populatelectrons
swept
from
the
gap
are
more
nearly
linear
ing the gap but the relationship is nonlinear. The
with beam
current
for gap
intensities
in nearly
the region
electrons
swept
from the
are more
linear of
greatest
interest
(>
4
jiC/pulse)
and
saturate
at 1-2%
with beam current for intensities in the region
of
averageinterest
beam neutralization
nearand
thesaturate
instability
greatest
(> 4 µC/pulse)
at onset.
1-2%
average beam neutralization near the instability onset.
500
600
Time(ns)
Time(ns)
FIGURE 6:
6: Typical
Typical PSR
FIGURE
PSR electron
electron sweeper
sweeper data
data (sweeper
(sweeper
500 V, RFA
RFA bias
bias voltage
voltage –25
500V,
-25 V,
V,32
32macropulses
macropulsesaveraged).
averaged).
RFA
have cloud
been successfully
and broadly
Electron
diagnostics based
on theapplied
planar to
RFA
successfullyand
anddynamics
broadly applied
to
studyhave
the been
EC properties
in positron,
study
the and
EC proton
properties
and
in distribution
positron,
electron,
rings.
Thedynamics
EC energy
electron,
rings.
EC energy
distribution
(for fluxand
at proton
the wall)
andThe
estimate
of density
can be
(for
flux
at
the
wall)
and
estimate
of
density
be
extracted from the RFA spectra - this is notcan
possible
extracted
fromBPMs
the RFA
spectra –plates.
this is EC
not diagnostics
possible
with biased
or collecting
with
biased
collecting
plates. EC
diagnostics
to date
areBPMs
mostlyorlimited
to electrons
collected
at the
towall
dateorareswept
mostly
to electrons
collected
at the to
to limited
the wall.
A number
of concepts
wall
or swept
to thenondestructively
wall. A numberwithin
of concepts
to
measure
the cloud
the vacuum
measure
the
cloud
nondestructively
within
the
vacuum
chamber have been proposed but not yet implemented.
chamber have been proposed but not yet implemented.
Despite the obvious differences in the beam-cloud
Despite thethere
obvious
differences
beam-cloud
interaction,
is likely
to bein the
synergy
between
interaction,
there
is
likely
to
be
synergy
betweenthe
proton and positron rings in understanding
proton
and positron
rings inandunderstanding
mechanisms
of EC saturation
the correlationthe
with
mechanisms
of
EC
saturation
and
the
correlation
with
cloud-induced instability thresholds.
More analyses
cloud-induced
instability
thresholds. More analyses
and experiments
are planned.
and experiments are planned.
ACKNOWLEDGMENTS
ACKNOWLEDGMENTS
The authors would like to thank all who contributed
The authors would like to thank all who contributed
to
EC studies at APS and PSR.
to EC studies at APS and PSR.
REFERENCES
REFERENCES
Seelinks
linkstotoworkshop
workshop
proceedings
review
talks:
http:
1.1. See
proceedings
andand
review
talks:
http:
//www.aps.anl.gov/asd/physics/ecloud/papers_gen.html
//www.aps.anl.gov/asd/physics/ecloud/papers_gen.html
M.A.A.Furman
Furmanand
andM.M.T.T.F. F.Pivi,
Pivi,Report
Report
LBNL2.2. M.
No.No.
LBNL49711/CBP
Note-415
(June
2002).
49711/CBP Note-415 (June 2002).
Grobner,Proc.
Proc.10th
10thIntl.
Ml.Conf.
Conf.ononHigh
HighEnergy
Energy
3.3. O.O.Gröbner,
Accel.,Protvino,
Protvino,Russia,
Russia,
277
(1977).
Accel.,
277
(1977).
K.C.Harkay
Harkayand
andR.A.
R.A.Rosenberg,
Rosenberg,Proc.
Proc.
1999
PAC,
4.4. K.C.
of of
1999
PAC,
1641(1999)
(1999)and
andK.C.
K.C.Harkay
Harkay
Report
CERN1641
et et
al.,al.,
Report
No.No.
CERN2002-001(2002).
(2002).
2002-001
R.J.Macek
Maceketetal.,al.,Proc.
Proc.of of2001
2001
PAC,
(2001)
5.5. R.J.
PAC,
688688
(2001)
andand
referencestherein,
therein,and
andR.R.Macek
Maceket etal.,al.,Report
ReportNo.No.
references
CERN-2002-001(2002).
(2002).
CERN-2002-001
R.A. Rosenberg
Rosenberg and
andK.C.
K.C.Harkay,
Harkay,Nucl.
Nucl.Instrum.
Instrum.
6.6. R.A.
Methods
MethodsA453,
A453,507
507(2000).
(2000).
7.7. G.G.Arduini
thethe
Electron
Cloud
Arduinietetal.,al.,“Measurement
"Measurementof of
Electron
Cloud
Properties
Means
ofof
a Multi-strip
Detector
at the
CERN
Propertiesbyby
Means
a Multi-strip
Detector
at the
CERN
SPS,”
of of
2002
EPAC
(2002).
SPS,"totobebepublished
publishedininProc.
Proc.
2002
EPAC
(2002).
8.8. T.T.Wang,
andand
R. R.
Macek,
“The
Wang,D.D.Barlow,
Barlow,A.A.Browman,
Browman,
Macek,
"The
Static
in in
a Beam
StaticElectric
ElectricField
Fieldofofa aCurved
CurvedElectrode
Electrode
a Beam
Pipe,”
Note
01-003
(May
2001).
Pipe,"PSR
PSRTechnical
Technical
Note
01-003
(May
2001).
9.9. R.A.
of of
2001
PAC,
R.A.Rosenberg
Rosenbergand
andK.C.
K.C.Harkay,
Harkay,Proc.
Proc.
2001
PAC,
2069
2069(2001).
(2001).
356