Velocity Bunching Experiment at the
Neptune Laboratory
P. Musumeci, RJ. England, M.C. Thompson, R. Yoder, J.B. Rosenzweig
Department of Physics, University of California at Los Angeles, 405 HilgardAve, Los Angeles, CA
90095, USA
Abstract. In this paper we describe the ballistic bunching compression experiment at the
Neptune photoinjector at UCLA. We have compressed the beam by chirping the beam energy
spectrum in a short S-band high gradient standing wave RF cavity and then letting the electrons
undergo velocity compression in the subsequent rectilinear drift. Using a standard Martin Puplett
interferometer for coherent transition radiation measurement, we have observed bunch length as
short as 0.4 ps with compression ratio in excess of 10 for an electron beam of 7 MeV and charge
up to 0.3 nC. We also measured slice transverse emittance via quad scan technique. The
observed emittance growth agrees with the predictions and the simulations. Extension of this
scheme to a future advanced accelerator injector system where solenoidal magnetic field can
compensate the emittance growth is studied.
INTRODUCTION
In recent years electron beam users have increased their demands for high
brightness beam in short sub-ps pulses [1-3]. Applications in the advanced accelerator
community like the injection into short wavelength advanced accelerators, or driving a
plasma wakefield experiment, and in the light source community like driving a short
wavelength SASE Free Electron Laser or Thompson-scattering generation of short Xray pulses, demand high brightness very short electron beam. Recent designs of such
systems include the use of conventional photoinjectors in conjunction with magnetic
compressors [4]. While the magnetic compression scheme has been proved successful
in increasing the beam current, the impact on the beam phase space has been shown to
be quite relevant: performing the compression at low energy [5], space charge forces
are still very significant and their emittance-damaging effect becomes especially
important in bending trajectories, in the case of compression at higher energy [6], one
has to deal with the deleterious effects on the longitudinal as well as the transverse
phase space of Coherent Synchrotron Radiation. Phase space filamentation and in
general emittance growth jeopardize the goal of achieving the high brightness.
An alternative scheme that could preserve the phase space quality has been recently
proposed to supply electron beams with the brightness required by the applications. In
the context of an injector for X-ray Free Electron Laser, Serafini and Ferrario [7]
proposed to use the old idea of RF rectilinear compression. More generally, in every
application in which compression at low energy is required, it seems that velocity
bunching is an efficient alternative to magnetic compression. The idea is based on the
CP647, Advanced Accelerator Concepts: Tenth Workshop, edited by C. E. Clayton and P. Muggli
© 2002 American Institute of Physics 0-7354-0102-0/02/$19.00
858
weak synchrotron motion that the beam undergoes at moderate energies in the RF
wave of the linac accelerating structure. The compression happens in a rectilinear
section so that the damage suffered by going through bending trajectories is avoided.
A main ingredient of this recipe to produce high brightness sub-ps electron beam is to
integrate this compression section in the emittance compensation scheme, by keeping
the transverse beam size under control through solenoidal magnetic field in the region
where the bunch is compressing and the electron density is increasing.
Another possibility is a thin lens version of velocity bunching. Here the synchrotron
motion inside the RF structure is very limited. There is almost no phase advance inside
the longitudinal lens and all the bunching happens in the drift following the linac.
In this paper we experimentally studied this configuration. At the Neptune
photoinjector at UCLA there is a 1.6 cell gun and a PWT standing wave linac that
could be used to test this idea. In the next section we draw the schematics of the
experiment, and show the results. We measured the bunch length by using the
Coherent Transition Radiation technique. After observing a good longitudinal
compression we turn our attention to the transverse dynamics. The big energy spread
on the beam makes it impossible to measure projected emittance so that we had to
concentrate on slice emittance. A 45 degrees dispersing dipole is used to select the
central slice of the beam and as the beam compresses it is clear that the emittance
grows. Simulations agree with this observation. It is important to note that the
beamline at the Neptune photoinjector is not optimized for this experiment, in the
sense that no solenoidal magnetic fields are present to match the increasing spacecharge forces and there is no post acceleration to remove the induced energy spread.
We also studied a system optimized for the ballistic bunching compression, the
proposed injector for the Orion Research facility [8]. Here the solenoids wrapped
around the accelerator should keep the beam under control and the simulations show
the high brightness of the output beam.
NEPTUNE EXPERIMENT
The Neptune facility at UCLA currently operates as an injector for a plasma beatwave advanced accelerator experiment. At the same time the Neptune photoinjector is
being used for pure high brightness beam dynamics studies like emittance growth in
bends [5] and negative R56 compressors [9]. The accelerator can be tune up for
ballistic compression.
A 266 nm 12 ps FWHM long laser pulse hits a single crystal copper cathode inside
a 1.6 cell BNL-SLAC-UCLA RF gun. The photoelectrons generated are then
accelerated by the RF fields and go through the emittance compensation solenoid. At
this point the beam can be energy chirped inside a 6+2 Vi cell S-band PWT RF cavity.
There is the capability of controlling independently the phases of the two accelerating
structures allowing us to test the ballistic bunching idea. Downstream of the linac an
aluminum foil can be inserted and the transition radiation generated is collected by a
parabolic mirror and reflected to a Martin Puplett autocorrelator for pulse length
diagnostic. There are also 4 chicane dipoles along the beamline and they can be turned
on in the 45 degrees dipole mode in order to select a slice of the beam of which
859
measuring
measuring on
on aa YaG
YaG screen
screen the
the beam size changing a quadrupole lens, we can
measure
measure the
the emittance.
emittance.
PWTLinac
PWT
Linac
YaG
Vertically
Ya6
screen
Vertically
Y*Gscreen
$
focusing ^±/
focusing
Quadrupole
Quadrupole^
^ CCD
camera
• ^ ••••
Autocorrelator
CTR foil
Chicane (used
(used as
as 45
45 degrees
degrees
Chicane
dispersing dipole)
dipole)
dispersing
Neptune
Neptune 1.6
1.6cell
cell
gun+solenoid
gun+solenoid for
for
emittance
emittance
compensation
compensation
Transverse
Transverse diagnostics:
diagnostics:
emittance
emittance measurement
measurement
via
via quad
quad scan
scan
Longitudinal
Longitudinal
diagnostics
diagnostics ::bunch
bunch
length
length
Figure 1.
1. Layout
Layout of
of the
the Neptune experiment
Figure
LONGITUDINAL DYNAMICS
LONGITUDINAL
Ballistic bunching
bunching is
is just
just aa thin
thin lens
lens version of the more general velocity bunching.
Ballistic
The phase
phase advance
advance of
of the
the electrons
electrons going through the longitudinal lens (the PWT
The
linac) isis few
few RF
RF degrees
degrees and all the bunching happens in the drift
linac)
drift outside. The
important difference
difference with
with the
the long
long RF-structure slow compression version of the
important
velocity bunching
bunching is
is that
that the
the beam is extracted
extracted still very close to the zero phase of the
velocity
RF
bucket
and
the
RF
non-linearities
RF bucket and the RF non-linearities that usually dominate the final bunch length are
greatly reduced.
reduced.
greatly
One
simple way
way to
to understand the ballistic bunching is to think to the
One simple
the time
time of
of
arrival
difference
for
particles
having
different
velocities.
When
the
time
of
arrival
arrival difference for
different
difference compensates
compensates the
the difference
difference in the longitudinal position, the bunch length
difference
will be
be minimum
minimum and
and that
that is
is the
the maximum compression point. A first
will
first order
approximation to
to describe
describe the
the ballistic bunching can be written as:
approximation
E ⋅ cos(φ ) ⋅ k ⋅ ∆z
∆p
Az
⋅ L = linac
⋅ L, == ∆
z
3
p
(E gun + Elinac ⋅ sin(φ ))3
(1)
where LL isis the
the distance
distance from
from the
the RF
RF structure,
structure, E
Eu
nac is the energy given by the PWT
where
linac
linac and
and EEgun
out of
of the
the gun,
gun, kk is
is the
the RF
RF wavenumber
wavenumber and
and <|φ) is
gun out
is the
the Linac
Linac phase.
phase. This
This
linac
relationship is
is strictly
strictly valid
valid at
at first
first order,
order, ignoring
ignoring space
space charge
charge and
and any
any phase
phase
relationship
advance inside
inside the
the PWT
PWT Linac.
Linac.
advance
860
Measurements
Measurements
We
Radiation technique.
technique. The
The
We measured
measured the
the pulse
pulse length
length by
by Coherent
Coherent Transition
Transition Radiation
electrons
hit
an
aluminum
foil
inserted
in
the
beam
path
and
the
transition
radiation
electrons hit an aluminum foil inserted in the beam path and the transition radiation isis
collected
interferometer with
with two
two
collected and
and reflected
reflected towards
towards aa polarizing
polarizing Martin-Puplett
Martin-Puplett interferometer
Golay
cell
detectors.
The
spectral
content
of
the
radiation
carries
the
information
on
Golay cell detectors. The spectral content of the radiation carries the information on
the
bunch
length.
The
resolution
of
the
interferometer
is
limited
by
the
spectral
the bunch length. The resolution of the interferometer is limited by the spectral
response
does not
not efficiently
efficiently reflect
reflect
response of
of the
the two
two wire
wire grid
grid polarizers.
polarizers. The
The wire
wire grid
grid does
wavelengths
shorter
than
the
wire
separation
distance
that
in
our
case
is
100
µm.
On
wavelengths shorter than the wire separation distance that in our case is 100 jim. On
the
other
side
of
the
frequency
spectrum,
the
analysis
of
the
CTR
interferometer
data
the other side of the frequency spectrum, the analysis of the CTR interferometer data
has
of the
the low
low frequencies
frequencies component
component in
in
has to
to be
be done
done taking
taking into
into consideration
consideration the
the loss
loss of
the
transition
radiation
spectrum
due
to
poor
vacuum
window
transmission
and
more
the transition radiation spectrum due to poor vacuum window transmission and more
importantly
This effect
effect is
is included
included by
by aa one
one
importantly to
to the
the diffraction
diffraction of
of the
the long
long wavelengths.
wavelengths. This
parameter
filtering
function
that
is
fitted
from
the
data
following
previous
work
by
parameter filtering function that is fitted from the data following previous work by
Murokh
et
al.
[10]
Murokhetal. [10]
For
PWT cavity,
cavity, by
by scanning
scanning the
the
For 250
250 pC
pC of
of charge
charge and
and 70
70 degrees
degrees off
off crest
crest in
in the
the PWT
moving
arm
of
the
interferometer,
we
obtain
the
interferogram
shown
in
fig.1.
The
moving arm of the interferometer, we obtain the interferogram shown in fig.l. The
data
ps.
dataanalysis
analysisgives
gives for
for the
the pulse
pulse length
length 0.39
0.39 ps.
ItItisisworth
noticing
the
compressed
beam
than what
what we
we were
were ever
ever able
able to
to
worth noticing the compressed beam is
is shorter
shorter than
get
with
the
magnetic
chicane
for
comparable
beam
charge,
confirming
the
fact
that
in
get with the magnetic chicane for
confirming the fact that in
this
thiscase
caseaamore
more linear
linear part
part of
of the
the RF
RF wave
wave is sampled.
Q = 210 +- 20 pC
0.50
0.50-,
0
0
φ = 70 +- 1
0.45
0.450.40
0.40-
Autocorrelation function
0.35I 0.35
c
?
0.30
0.30-
w
0.25
0.25-
I f
0.200.20
0.100.10
Chi^2 == 0.00043
0.00043
Chi'2
R^2
0.93021
R*2
== 0.93021
cc
rrrsz
rmsz
0.34214
0.34214
-0.32616
-0.32616
5.06834
5.06834
0.39183
0.39183
±0.0024
±0.0024
±0.0125
±0.0125
±0.02002
±0.02002
±0.01888
±0.01868
csi
csi
1.93039
1.93039
±0.13644
±0.13644
AA
V
0.150.15
Model:ctr
ctrautocorrelation
autocorrelation
Model:
t0
to
0.05
0.05
0
2
4
6
8
10
12
ps
ps
FIGURE 2.
2. Autocorrelation results
FIGURE
The predictions
predictions from
from the
the first
first order approximation given in
The
in (1)
(1) have
have been
been
experimentally
verified
by
measuring
the
compression
phase
(|
)
changing
the
energy
experimentally verified by measuring
φ changing the energy
gradient in
in the
the Linac.
Linac. The
The RF-cavity
RF-cavity phase
phase can
can be
gradient
be measured
measured with
with aa very
very small
small error
error by
by
mixing the
the RF
RF fields
fields inside
inside the
the structure
structure with
with aa reference
mixing
reference RF-clock,
RF-clock, at
at the
the same
same time
time
the phase
phase for
for maximum
maximum compression
compression is
is easily
easily determined
the
determined by
by maximizing
maximizing the
the
Coherence Transition
Transition Radiation
Radiation energy
energy on
on the
the bolometer
Coherence
bolometer detectors.
detectors. The
The agreement
agreement
861
∆φ for maximum compression
with
cancellation
withthe
the analytical
analytical formula
formula isis very
very good.
good. Note
Note that
that there
there is
is an
an important
important cancellation
effect.
effect. As
As we
we decrease
decrease the
the energy
energy gradient
gradient in
in the
the Linac
Linac we
we don’t
don't have
have to
to adjust
adjust the
the
phase
beam is
is getting
getting less
less
phasetotomaintain
maintain the
the energy
energy spread
spread because
because at
at the
the same
same time
time the
the beam
energetic
energeticand
andless
lessrigid
rigidto
toaarotation
rotation in
in the
the longitudinal
longitudinal phase
phase space.
space.
I
Theory
-Theory
-Measurement
Measurement obtained
obtained mixing
mixing
RF
RFLow
Low Level
Level with
with Linac
Linac Loop
Loop
10
9
8
7
6
5
4
3
2
1
0
-1
-2
-3
-4
-5
-6
-7
-8
-9
-10
5
10
15
20
Average Linac
Linac Energy
Energy Gradient
Gradient (MV/m)
(MV/m)
Average
FIGURE3.3.Phase
Phaseof
ofmaximum
maximumbunching
bunching vs.
vs. Linac
Linac accelerating
accelerating gradient
FIGURE
TRANSVERSE DYNAMICS
DYNAMICS
TRANSVERSE
Asititwas
wasshown
shown in
in the
the last
last section,
section, the
the Neptune experiment confirmed
As
confirmed that
that ballistic
ballistic
bunching
could
be
an
efficient
and
compact
way
of
increasing
many
folds
the current
bunching could be an efficient and compact way of increasing
the beam.
beam. The
The question
question to
to be
be answered
answered becomes if the increase in beam
ofof the
beam current
current
corresponds
to
a
relative
increase
in
brightness,
in
other
words
we
need
to
understand
corresponds to a relative increase in brightness, in other
understand
whathappen
happen to
to the
the transverse
transverse phase
phase space,
space, if the emittance can be preserved through
what
through
the
compression
process.
the compression process.
Becausethe
the beam
beam runs
runs through
through the
the high
high gradient
gradient structure
structure far from
Because
from the
the crest
crest of
of the
the
RFwave
wave to
to chirp,
chirp, the
the energy
energy spread
spread at
at the
the exit
exit of
of the
the Linac
RF
Linac is
is very
very big.
big. For
For example
example
for the
the case
case in
in which
which the
the focus
focus of
of the
the longitudinal
longitudinal lens
lens is
is 33 m
m downstream
downstream on
for
on the
the
beamline, the
the RF
RF phase
phase was
was set
set 70
70 degrees
degrees off
off crest,
crest, resulting
resulting in
in aa energy
energy spectrum
spectrum
beamline,
extending from
from 55 MeV
MeV to
to 99 MeV.
MeV. This
This isis not
not aa problem
problem in
in aa system
system where
where the
the beam
extending
beam
energy can
can be
be boosted
boosted up
up by
by additional
additional accelerating
accelerating cavities
cavities to
to quickly
quickly remove
energy
remove the
the
relative energy
energy spread,
spread, but
but at
at the
the Neptune
Neptune photoinjector
photoinjector there
there is
is no
such capability.
relative
no such
capability.
Thisisisaalimitation
limitation to
to the
the determination
determination of
of the
the transverse
transverse projected
projected emittance
emittance because
This
because
the energy
energy spread
spread translating
translating in
in an
an angle
angle spread
spread will
will appear
appear to
the
to all
all the
the measurement
measurement
technique(that
(thatare
aretrace
tracespace
space measurements)
measurements) as
as unphysical
unphysical transverse
transverse emittance.
emittance.
technique
On
the
other
hand,
the
energy
is
correlated
with
the
longitudinal
On the other hand, the energy is correlated with the longitudinal position
position of
of the
the
beam
and
selecting
a
small
window
of
acceptance
in
energy,
a
longitudinal
slice
beam and selecting a small window of acceptance in energy, a longitudinal slice of
of the
the
beam can
can be
be selected.
selected. Experimentally,
Experimentally, we
we can
can use
use the
the 45
beam
45 degrees
degrees dispersing
dispersing bending
bending
dipoleconfiguration
configuration to
to select
select aa beam
beam slice
slice over
over which
which aa vertical
vertical quad
quad scan
scan emittance
dipole
emittance
measurement
can
be
performed.
In
this
way
the
energy
spread
is
not
measurement can be performed. In this way the energy spread is not aa limit
limit to
to the
the
862
measurement.
beam, the
the slice
slice
measurement. Moreover
Moreover in
in some
some applications
applications for
for high
high brightness
brightness beam,
emittance
vertical phase
phase space
of the
the
emittance isis the
the relative
relative quantity.
quantity. We
We are
are going
going study
study the
the vertical
space of
electron
to the
the beam
beam
electron beam
beam scanning
scanning the
the phase
phase of
of the
the linac
linac to
to understand
understand what
what happen
happen to
as
as itit isis undergoing
undergoing compression.
compression.
Slicing
Slicing the
the beam
beam with
with the
the 45
45 degrees
degrees dispersing
dispersing dipole
dipole
Because
Because changing
changing the
the linac
linac phase
phase that
that is
is the
the main
main compression
compression knob,
knob, it
it also
also
changes
the
energy
of
the
beam,
it
is
important
to
ensure
that
always
the
same
part
changes the energy of the beam, it is important to ensure that always the same part of
of
the
that
the beam
beam hits
hits on
on the
the small
small acceptance
acceptance YAG
YAG screen
screen (few
(few degrees
degrees of
of bending
bending angle
angle that
isis few
be able
able to
to set
set the
the dipole
dipole
few %
% of
of energy
energy spread).
spread). Experimentally,
Experimentally, we
we need
need to
to be
current
portion of
of the
the beam
beam
current to
to keep
keep at
at the
the 45
45 degrees
degrees bending
bending angle
angle always
always the
the same
same portion
as
as the
the energy
energy of
of this
this portion
portion changes.
changes. This
This is
is accomplished
accomplished first
first by
by measuring
measuring the
the full
full
spectrum
reference
spectrum of
of the
the beam
beam as
as the
the linac
linac phase
phase is
is changed,
changed, then
then individuating
individuating one
one reference
slice
the beam
beam and
and
slice that
that in
in our
our case
case is
is the
the central
central slice
slice or
or the
the maximum
maximum current
current slice
slice in
in the
finally
reference slice.
In
finally compensating
compensating with
with the
the dipole
dipole current
current to
to analyze
analyze always
always the
the reference
slice. In
the
the figure
figure we
we can
can see
see the
the energy
energy change
change of
of the
the central
central reference
reference slice
slice of
of the
the beam
beam as
as
the
the linac
linac phase
phase is
is scanned.
scanned. This
This curve
curve incidentally
incidentally allows
allows us
us an
an independent
independent
determination
determination of
of the
the RF
RF cavities
cavities accelerating
accelerating gradient.
gradient. We
We found
found in
in good
good agreement
agreement
with
the
RF
measurements
the
energy
gain
in
the
1.6
cell
gun
with the RF measurements the energy gain in the 1.6 cell gun to
to be
be 44 MeV
MeV
corresponding
corresponding to
to aa 80
80 MV/m
MV/m gradient
gradient and
and in
in the
the PWT
PWT linac
linac 88 MeV
MeV corresponding
corresponding to a
40
40 MV/m
MV/m accelerating
accelerating gradient.
gradient.
- Maximum
Maximum Faraday
Faraday cup
cupsignal
signal line
line
energy of central slice (MeV)
12
10
8
6
Model:
φ)
Model:Egun+
Egun+Elinac
Elinac **sin(
sinfo)
4
Chi^2
ChiA2 == 0.08008
0.08008
R^2
RA2 == 0.99183
0.99183
2
Elinac
±0.56147
Elinac 7.89671
7.89671
±0.56147
zerophase
74.33696
±3.67494
zerophase
74.33696
±3.67494
period
±0
period 360
360
±0
Egun
±0.63804
Egun 3.78945
3.78945
±0.63804
0
160
180
180
200
200
220
220
240
240
260
PWT
PWT Linac
Linac phase
phase (degrees)
(degrees)
FIGURE.
FIGURE. 4:
4: Energy
Energy of
of central
central slice
slice scanning
scanning the
the phase of the linac
Quad
Quad scan
scan measurement
measurement
On
On the
the central
central reference
reference beam
beam slice
slice we can perform the quad scan. Since we are
dealing
dealing with
with aa beam
beam that
that varies
varies significantly
significantly inside of the quads, we need to go to a
quad
quad scan
scan analysis
analysis that
that takes
takes into
into account
account the thickness of the quad lens.
863
M e a s u re m e n t o n s c re e n 6
ehaicsku le
renms efit
n t oon
n sscreen
c re e n 6(.
-MTMeasurement
TThick
h ic k lelens
n s fit
fit
0 .2 5
0 .2 5
M o d e l: th ic k le n s
MModel
o d e l: ththicklens
ic k le n s
C h i^2 = 0 .0 0 0 0 7
CR
h^2
i^2 ==
.0.909030073
Chi«2
= 000.00007
RR
^2A 2
== 0 0.99303
.9 9 3 0 3
P1
0 .7 5 4 0 4
P
2
4
.1
P P1
1
0 0.75404
.7 51400347
PPP2
23
42.131.6053677 1
P3
2 3 .6 5 6 7 1
2
2
2
sigma
) 2)
sigma(mm
(mm
0 .2 0
0 .2 0
0 .1 5
0 .1 5
± 0 .0 3 1 4 1
±00.0.139144014
±±0.03141
±01.1.091490247
±±0.19404
± 1 .0 1 9 2 7
0 .1 0
0 .1 0
0 .0 5
0 .0 5
0 .0 0
0 .0 0
9 .5
9 .5
1 0 .0
1 0 .0
1 0 .5
1 0 .5
1 1 .0
1 1 .0
1 1 .5
1 -1
1 .5
1 2 .0
1 2 .0
s q rt(K ) (m-11 )
(m- ))
ssqrt(K)
q rt(K ) (m
1 2 .5
1 2 .5
1 3 .0
1 3 .0
FIGURE 5: Thick lens treatment for the quad scan
FIGURE5:5:Thick
Thicklens
lenstreatment
treatment for
for the quad scan
FIGURE
The
parameterization
of
the
square
of
the
measured beam
beam size with respect to the quad
quad
Theparameterization
parameterizationof
ofthe
thesquare
squareof
of the
the measured
measured
The
beam size with respect to the
the quad
strength
(K)
is
strength(K)
(K)isis
strength
[[ (( ))
(( ))]]
2
2 [sin ( K l ) + l cos( K l )][cos( K l ) −
-=
K [sin ( K l ) + l cos( K l )][cos( K l ) −
K
 sin ( K l )

sin ( K l ) + l cos( K l ) σ
+ l cos( K l ) σ

σ 2 (K ) = cos K l − K l sin K l 2 σ +
σ 2 (K ) = cos K l qq − K l dd sin K l qq σ 1111 +
2
q
q


K
K
d
d
q
q
q
q
((
))]]
K l sin K l σ 12 +
K l dd sin K l qq σ 12
+
(2)
(2)
(2)
2
2
q
q
d
d
q
q
22
22


where
are the quad and drift lengths, respectively. Here we take into account the
wherellIq,d
q4 are the quad and drift lengths, respectively. Here we take into account the
where
q,d are the quad and drift lengths, respectively. Here we take into account the
full
thick
lens
matrix
instead
of
more
simple
analysis
in which
which the
the betatron
betatron phase
phase
fullthick
thicklens
lensmatrix
matrixinstead
instead of
of aaa more
more simple
simple analysis
analysis in
in
full
which the
betatron
phase
advance
inside
the
quad
is
negligible
and
the
thin
lens
approximation
can
be
used.
advanceinside
insidethe
thequad
quadisisnegligible
negligibleand
andthe
the thin
thin lens
lens approximation
approximation can
can be
advance
be used.
used.
In
the
figure
is
shown
the
observed
emittance
growth.
In
the
figure
is
shown
the
observed
emittance
growth.
In the figure is shown the observed emittance growth.
- Quad
Quad scan
scan data
data
TREDI
simulation
Quad
scan
data results
-TREDI
simulation
results
15
Emittance
(mm-mrad)
Emittance
(mm-mrad)
15
14
TREDI simulation results
14
13
13
12
12
11
11
10
10 9
98
87
76
65
54
4
100
100
100
120
120
120
140
140
160
160
Linac
Linac140phase
phase160
Linac phase
180
180
180
FIGURE
results and
and simulations
simulations
FIGURE 6:
6: Emittance
Emittance growth
growth during compression, experimental results
FIGURE 6: Emittance growth during compression, experimental results and simulations
864
Simulations
Simulations
At
necessary to
to compare
compare the
the experimental
experimental results
results with
withthe
thepredictions
predictions
At this
this point
point it
it is
is necessary
from
the
theory.
The
system
is
not
optimized
to
maintain
the
transverse
phase
space
from the theory. The system is not optimized to maintain the transverse phase space
quality,
but
it
is
interesting
to
check
our
simulation
tools
against
this
problem.
The
quality, but it is interesting to check our simulation tools against this problem. The
first
task
of
the
simulations
should
be
to
fully
understand
the
systematic
of
the
first task of the simulations should be to fully understand the systematic of the
measurement.
In
fact
the
first
question
to
ask
is
what
happen
to
the
beam
as
the
linac
measurement. In fact the first question to ask is what happen to the beam as the linac
phase
we are
are measuring
measuring the
the emittance.
emittance.
phase is
is scanned
scanned before
before we
70 degrees off crest in the Linac
30 degrees off crest in the Linac
FIGURE 7. Longitudinal crossover. The particles
FIGURE
particles are
are color
color mapped
mapped by
bythe
theinitial
initiallongitudinal
longitudinalposition.
position.
First row:
row: evolution of beam going 70 degrees
First
degrees off
off crest,
crest, the
the particles
particles in
in the
the head
headof
ofthe
thebunch
bunchend
endup
upinin
the tail,
tail, there is longitudinal crossover. Second row:
the
row: beam
beam going
going 30
30 degrees
degrees off
off crest
crest inin the
the Linac,
Linac,inin
this case the energy spread is smaller and there is non longitudinal
this
longitudinal crossover.
crossover.
If we look in the configuration space it is clear
If
clear that
that for
for aa big
big energy
energy spread
spread beam
beam
longitudinal cross over takes place as
as soon
soon as
as the
the beam
beam enters
enters the
the dipole.
dipole. The
The more
more
energetic particles that are in the tail reach the less energetic
energetic ones
ones that
that bend
bend more
more in
in
the dipole field.
field. If the energy spread is not too big (<10%)
the
(<10%) the
the longitudinal
longitudinal crossover
crossover
doesn’t happen and the beam just bends inside the
doesn't
the dipole.
dipole. This
This isis ultimately
ultimately the
thereason
reason
that we observe significant emittance growth
that
growth at
at phases
phases for
for which
which the
the compression
compression
should not
not be
be so
should
so severe
severe in
in aa rectilinear
rectilinear drift.
drift. Whenever
Whenever there
there isislongitudinal
longitudinalcrossover
crossover
the
quality
of
the
beam
is
dramatically
affected.
We
have
the quality of the beam is dramatically affected. We have performed
performed fully
fully
3dimensional simulations
Sdimensional
simulations with
with the
the Lienard
Lienard Wiechert
Wiechert potential
potential code
code TREDI
TREDI [11]
[11] and
and
found good
good agreement
agreement with
found
with the
the data.
data.
865
ORION PROPOSAL
The Neptune
Neptune configuration
configuration is aa very particular one. It is not optimized for ballistic
The
bunching
compression
experiment. The goal of our experiment was to explore as
bunching compression experiment.
much
as
possible
of
the
scheme and compare with theory and simulations. The
much as possible of the new scheme
idea behind
behind this
this approach
approach is that once we understand what is going on in the Neptune
idea
experiment we
we would
would be ready to design a system in which velocity bunching actually
experiment
increases the
the brightness
brightness of
of the
the electron
electron beam.
increases
In general,
general, especially
especially when
when the compression has to be done at low energy when
In
space charge
charge forces
forces are
are very strong,
strong, bending trajectories and magnetic compression
space
are not
not aa possibility and velocity bunching is a competitive solution. For example in
are
the proposed Orion injector, the injector is an S-band 1-6 cell gun and the booster
the
accelerating cavities
cavities are
are two x-band structures. To limit the energy spread, the beam
accelerating
has to
to be
be compressed
compressed before being injected into the shorter wavelength cavities. One
has
competitive proposal is
is to
to use the ballistic bunching. A short high gradient standing
competitive
wave
cavity
will
chirp
the
beam and the compression will happen in the following
wave cavity will chirp
following
drift
before
the
x-band
cavity.
drift before the x-band cavity.
0.2
0.15
0.1
z
σ (mm)
I
0.05
S-band RF
RF gun
gi
S-band
0
0
I 22
11
33
44
S-band PWT
PWT Buncher
Buncher
(iTl)
S-band
zZ (m)
5
6
7
X-band travelling
travelling wave
wave linacs
linacs
X-band
FIGURE 8:
8: Orion
Orion Ballistic
Ballistic bunching.
bunching. Longitudinal
Longitudinal beam
beam size
size
FIGURE
To maintain
maintain the
the transverse
transverse beam
beam quality
quality the
the beam
beam remains
remains slightly
slightly
To
undercompressed in
in order
order to
to stay
stay away
away from
from the
the deleterious
deleterious effect
effect of
of the
the longitudinal
longitudinal
undercompressed
crossovers.
crossovers.
Another important
important element
element in
in aa velocity
velocity bunching
bunching injector
injector is
is the
the solenoidal
solenoidal
Another
magnetic field
field to
to keep
keep the
the beam
beam under
under control
control while
while it
it is
is compressing.
compressing. Because
Because the
the
magnetic
beam is
is compressing
compressing and
and getting
getting denser
denser the
the plasma
plasma frequency
frequency of
of the
the transverse
beam
transverse
oscillation is
is increasing.
increasing. Solenoid
Solenoid magnetic
magnetic field
field keep
keep the
the beam
beam focused
focused to
to control
control the
oscillation
the
beam
size
and
the
emittance
oscillation.
The
increasing
magnetic
field
to
match the
beam size and the emittance oscillation. The increasing magnetic field to match
the
beam plasma
plasma frequency
frequency is
is given
given in
in the
the Orion
Orion case
case by
by properly
properly tailoring
tailoring the
the solenoids
solenoids
beam
wrapped
around
the
x-band
linacs.
wrapped around the x-band linacs.
866
3500
3000
2500
B (G)
2000
1500
1500
CO
1000
1000
500
0
0
50
100
100
150
150
200
200
250
300
350
350
400
400
zz(cm)
(cm)
4
3.5
2
1.5
n,x
ε
(mm-mrad)
3
2.5
1
0.5
0
0
100
200
300
400
500
zz(cm)
(cm)
FIGURE
FIGURE9:
9: Magnetic
Magnetic field
field and
and emittance
emittance compensation
compensation in
in Orion
Orion case
CONCLUSION
CONCLUSION
The Neptune
Neptune ballistic
ballistic bunching
bunching experiment
experiment demonstrated
demonstrated the efficiency
efficiency of the
The
rectilinear RF
RF compression.
compression. A
A compression
compression ratio in excess of 10
10 was achieved due to
rectilinear
the fact
fact that
that RF-non
RF-non linearities
linearities are
are strongly
strongly suppressed
suppressed in this configuration.
the
Experimental investigation
investigation on
on the
the transverse
transverse phase space quality showed the
Experimental
deleterious effect
effect of
of having
having aa longitudinal
longitudinal crossover
crossover anywhere
anywhere along the beamline.
deleterious
Future experiments
experiments are
are needed
needed to
to investigate
investigate the full potential of this method for
Future
increasing the
the brightness
brightness of
of photoinjector
photoinjector beams,
beams, and
and the use of the magnetic
increasing
solenoids to
to keep
keep the
the beam
beam under
under control.
control. One important point to be addressed is to
solenoids
investigate the
the difference
difference between
between the
the thin
thin lens
lens version ‘ballistic’
'ballistic' bunching and the
investigate
long version
version of
of the
the rectilinear
rectilinear compressor.
compressor. UCLA
UCLA will
will be involved also in experiments
long
on this
this last
last configuration
configuration both
both at
at the
the Pleiades
Pleiades Thomson
Thomson source and at the INFN
on
SPARCinjector.
injector.
SPARC
ACKNOWLEDGEMENTS
ACKNOWLEDGEMENTS
The authors
authors would
would like
like to
to thank
thank X.J.
XJ. Wang,
Wang, Luca
Luca Serafini
Serafini and Massimo Ferrario for
The
useful discussions.
discussions. This
This work
work isis supported
supported by
by U.S.
U.S.Department of Energy, grant No.
useful
DE-FG03-92ER40693.
DE-FG03-92ER40693.
867
REFERENCES
1.
2.
3.
4.
5.
Linac Coherent Light Source (LCLS) Design Report, Tech, Pub,, SLAC-R-521 (1998).
Clayton C.E., Serafini L., IEEE Trans, On Plasma Science, 24, 400 (1996)
Hartemann F., "Pleiades experiment", these proceedings
Rosenzweig J.B., Barov N., Colby E., IEEE Trans, On Plasma Science, 24, 409 (1996)
Anderson S. et al. "Commisioning of the Neptune Photo-injector" in Proceedings of the 2001
Particle Accelerator Conference 2001, Chicago, p. 89 (2001)
6. Graves W.S. et al. "Ultrashort electron bunch length measurement at DUVFEL" in Proc, Of Particle
Accelerator Conference 2001, Chicago, p. 2224 (2001)
7. Serafini L., Ferrario M., "Velocity Bunching in Photo-Injectors" in Physics of, and science with, the
X-Ray Free-Electron Laser, edited by S.Chattopadhyay et al., AIP Conference Proceedings 581, 19th
Advanced ICFA Beam Dynamics Workshop, Arcidosso, pp. 87 (2001)
8. "Orion Research Facility: Technical Design Study" SLAC Tech. Rep. (2002)
9. England RJ. et al., "Negative R56 compressor" these proceedings
10.Murokh A. , Rosenzweig J.B., Tremaine A., "Coherent Transition Radiation based diagnosis of
electron beam pulse shape" in Proc, Of Advanced Accelerator Concepts 1998, Baltimore, AIP
Conference Proceedings 472, 38 (1999)
11.Giannessi L. et al., Nucl, Instr, Meth,, 393, 434, 1997
868