Electron Beam Production and Characterization
for the PLEIADES Thomson X-Ray Source
W. J. Brown1, F. V. Hartemann1, A. M. Tremaine1, P. T. Springer1, G. P.
Le Sage1, C.PJ. Barty1, J. B. Rosenzweig2, J. K. Crane1, R. R. Cross1, D.
N. Fittinghoff1, D. J. Gibson3, D.R. Slaughter1, S. Anderson2
!
Lawrence Livermore National Laboratory, 7000 East Ave., Livermore, CA. 94550
UCLA Department of Physics and Astronomy, 405 Hilgard Ave, Los Angeles, CA. 90095
5
UCD Department of Applied Science, 661 Hertz Hall, Livermore, CA, 94550
2
Abstract. We report on the performance of an S-band RF photocathode electron gun and
accelerator for operation with the PLEIADES Thomson x-ray source at LLNL. Simulations of
beam production, transport, and focus are presented. It is shown that a 1 ps, 500 pC electron
bunch with a normalized emittance of less than 5 Tcmm-mrad can be delivered to the interaction
point. Initial electron measurements are presented. Calculations of expected x-ray flux are also
performed,
demonstrating
an
expected
peak
spectral
brightness
of
1020
photons/s/mm2/mrad2/0.1% bandwidth. Effects of RF phase jitter are also presented, and planned
phase measurements and control methods are discussed.
INTRODUCTION
PLEIADES (Picosecond Laser Electron InterAction for Dynamic Evaluation
of Structures) is a next generation Thomson scattering x-ray source being developed at
Lawrence Livermore National Laboratory (LLNL). Ultra-fast ps x-rays (10-200 keV)
will be generated by colliding an energetic electron beam (20-100 MeV) with a high
intensity, sub-ps, 800 nm laser pulse. Generation of sub-ps pulses of hard x-rays (30
keV) has previously been demonstrated at the LBNL Advanced Light Source injector
linac, with x-ray beam fluxes of 105 photons per pulse [1]. The LLNL source is
expected to achieve fluxes between 107 - 108 photons for pulse durations of 100 fs to
5 ps using interaction geometries ranging from 90° (side-on collision) to 180° (headon collision).
To achieve such a high x-ray flux, a very high brightness electron beam,
capable of being focused to a roughly 20 jim spot size at the interaction point, will be
required. The PLEIADES beamline has been designed to meet these demands by
producing a 1-2 ps, 500 pC bunch with a normalized emittance of less than 5 Timmmrad.
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
830
nteraction Chamber
ending Magnet
FIGURE
FIGURE 1.
1. Interaction
Interaction geometry.
geometry.
EXPERIMENTAL
EXPERIMENTAL LAYOUT
LAYOUT
The
The PLEIADES
PLEIADES facility
facility consists
consists of
of aa Ti-Sapphire
Ti-Sapphire OPCPA
OPCPA laser
laser system
system capable
capable
of
of producing
producing bandwidth
bandwidth limited
limited laser
laser pulses
pulses of
of 50
50 fs
fs with
with up
up to 11 joule of energy at
800
800 nm,
nm, an
an S-band
S-band photo-cathode
photo-cathode RF
RF gun [2],
[2], and a 100
100 MeV
MeV linac
linac consisting
consisting of
of 4,
4,
2.5-meter-long
2.5-meter-long accelerator
accelerator sections.
sections. The
The RF
RF gun
gun is driven by a picosecond, 11 mJ,
ml, UV
laser
laser that
that is
is synchronized
synchronized to
to the
the interaction
interaction drive
drive laser.
laser.
A
A schematic
schematic of
of the
the interaction
interaction region
region is
is shown
shown in
in Figure
Figure 1.
1. To
To maximize
maximize x-ray
flux
flux while
while minimizing
minimizing effects
effects of
of timing
timing jitter,
jitter, the
the laser
laser incidence
incidence angle
angle will
will initially
initially be
be
180
180 degrees
degrees with
with respect
respect to
to the
the electron beam direction, though aa 90 degree interaction
geometry
geometry will
will also
also be
be possible.
possible. The
The focal
focal length
length between
between the
the final
final focus
focus quadrupole
quadrupole
triplet
triplet and
and the
the interaction
interaction region is 10
10 cm
cm to
to allow
allow for
for maximum focus strength and
minimum
minimum electron
electron bunch
bunch spot
spot size.
size. A
A 30-degree
30-degree bend
bend dipole
dipole magnet
magnet will
will be
be used
used to
to
bend
bend the
the electron
electron bunch
bunch out
out of
of the
the x-ray
x-ray beam
beam path
path following
following the
the interaction.
interaction. An
An offoffaxis
axis 1.5
1.5 m
m focal
focal length,
length, parabolic
parabolic mirror
mirror will
will be
be used
used to
to focus
focus the
the laser
laser to
to aa
diffraction-limited
diffraction-limited 15
15 µm
jim FWHM
FWHM spot
spot size
size at
at the
the interaction
interaction point.
point. A
A beryllium
beryllium flat
flat
mirror
mirror placed
placed in
in the
the x-ray
x-ray beam
beam path
path will
will serve
serve as
as the
the final
final steering
steering optic
optic for
for the laser,
while
while being
being transparent
transparent to
to the
the x-ray
x-ray beam.
beam.
ELECTRON
ELECTRON BEAM
BEAM PRODUCTION
PRODUCTION
Initial
Initial experiments
experiments will
will focus
focus on
on the
the generation
generation of
of 30
30 keV
keV x-rays
x-rays produced in
in aa
head on
on collision
collision using
using aa 35
35 MeV
MeV electron
electron beam.
beam. The
The beamline has been fully
head
modeled,
modeled, from
from the
the S-band
S-band photo-cathode
photo-cathode RF
RF gun to the interaction point using the Los
Alamos
Alamos particle
particle dynamics
dynamics code,
code, PARMELA,
PARMELA, and
and the
the electro-static
electro-static and
and
electromagnetic
electromagnetic field
field solvers:
solvers: POISSON
POISSON and
and SUPERFISH.
SUPERFISH. These
These include
include simulations
simulations
of
of emittance
emittance compensation
compensation between
between the
the RF
RF gun
gun and
and the
the first
first linac
linac section,
section, velocity
velocity
831
compression
compression of
of the
the bunch
bunch through
through the
the first
first linac section, and subsequent acceleration
and optimization
optimization of
of beam
beam energy
energy spread and emittance during transport through the
and
subsequent accelerator
accelerator sections.
sections.
subsequent
The two
two primary
primary factors
factors that
that determine
determine the minimum electron bunch spot size
The
the interaction
interaction point
point are
are beam
beam emittance and energy spread. In addition, the x-ray
atat the
flux will
will also
also depend
depend on
on electron
electron bunch charge. However, optimizing emittance and
flux
energy spread
spread places
places aa practical
practical limit
limit on the amount of charge in the bunch. After
energy
finding optimized
optimized cases
cases for
for several
several different
different bunch
bunch charges, 0.5 nC was found to be a
finding
good compromise
compromise between
between high
high bunch
bunch charge and low emittance.
good
The RF
RF gun
gun isis designed
designed to
to accelerate
accelerate an electron bunch to an initial energy of 5
The
MeV. The
The beam
beam will
will then
then be
be transported
transported to the first
first linac section, during which
MeV.
emittance compensation
compensation is
is performed
performed by
by focusing
focusing the beam to a waist with a solenoid
emittance
placed directly
directly after
after the
the RF
RF gun.
gun. In
In the
the first
first linac section, the beam is injected near the
placed
zero
crossing
of
the
accelerating
field
in
zero crossing of the accelerating field in order to provide an energy chirp to the beam.
This results
results in
in velocity
velocity compression
compression of the accelerated bunch. The second accelerator
This
section isis then
then used
used to
to accelerate
accelerate the
the beam
beam from
from about 10 MeV to the final beam
section
energy, about
about 35
35 MeV.
MeV. The
The third
third section can then be used to remove a large portion of
energy,
the energy
energy spread
spread induced
induced by
by the
the velocity
velocity compression process. Throughout the
the
acceleration process,
process, the
the linac
linac solenoids
solenoids are carefully
carefully optimized to minimize emittance
acceleration
growth.
growth.
40
tu
7
'
6
« 6
52 O
5
s»x
f\\.•••-••.*,
'''
/
^
Bunch Duration
—— Bunch
Duration (FWHM)
(FWHM)
dE (%)
- - - dE
(%)
30
30
- - Energy
Energy
25
25 f
^
20
20
8)
15
15 =
^
4
i3
35
oo
f
r/
.\
\
''
3
2
2
1
:
\
^
10
10
11 ^'~ ~ "• ;
00
5
5
1
(0
)
Energy (MeV)
Duration FWHM (ps), dE (%)
8
o
500
500
* * .. . . - - - - - - - - - • •
1000
1000
z
(cm)
z (cm)
1
0
0
1500
1500
FIGURE 2A.
2A. PARMELA
PARMELA simulation
simulation showing
showing electron
electron bunch
bunch parameters
parameters (energy,
(energy, energy
FIGURE
energy spread,
spread, and
and
bunch length)
length) versus
versus longitudinal
longitudinal position
position in
in the
the beam
beam line.
line. The
The gun
gun cathode
cathode position
position corresponds
bunch
corresponds to
to zz
cm, while
while the
the interaction
interaction region
region is
is located
located at
at zz == 1700
1700 cm.
cm.
==00cm,
832
6o
55
55
i
44
-^
σx,σy (mm)
4
3
)' ^
CO
(πmm-mrad)
εn(Timm-mrad
6o
— yy emittance
emittance
— yy rms
rms(mm)
(mm)
— xx emittance)
emittance)
— -xx rms
rms (mm)
(mm)
- - Magnetic
Magnetic Field
Field
|
33 ^
^
os
2
V
|\\ \'
11 -
\
\'1V/
\x^^x
;'"
0n
0
V
/•/
!i
^x
Y;
500
500
Xx
:ii
'
v
>
'
Ii
I
! 'i
22
\i
'
ii
- ; -—- -Li —l-^'
1000
1000
zz(cm)
(cm)
^
"i
11
- 0n
1500
1500
FIGURE
FIGURE 2B.
2B. PARMELA
PARMELA simulation
simulation showing
showing electron
electron bunch
bunch parameters
parameters (emittance,
(emittance, and
and spot
spot size.
size,
versus
versus longitudinal
longitudinal position
position in
in the
the beam
beam line.
line. The
The dotted
dotted line
line shows
shows the
the solenoid
solenoid magnetic
magnetic field
field strength
strength
(in
the four
four linac
linac solenoids.
(in kG)
kG) for
for the
the gun
gun solenoid
solenoid and
and the
solenoids.
Figure
beam characteristics
characteristics
Figure 2a
2a and
and 2b
2b show
show the
the evolution
evolution of
of several
several of
of the
the beam
during
during the
the acceleration
acceleration process
process determined
determined from
from PARMELA
PARMELA simulations,
simulations, including
including
emittance,
emittance, energy
energy spread,
spread, and
and bunch
bunch length.
length. Velocity
Velocity compression
compression is
is used
used to
to reduce
reduce
the
bunch
length
from
about
6
ps
at
the
entrance
of
the
accelerator
to
about
1.7
the bunch length from about 6 ps at the entrance of the accelerator to about 1.7 ps
ps at
at its
its
exit.
While
it
is
possible
to
compress
further,
this
bunch
length
was
chosen
exit. While it is possible to compress further, this bunch length was chosen to
to
minimize
minimize emittance
emittance growth
growth resulting
resulting from
from the
the compression
compression process,
process, while
while maximizing
maximizing
x-ray
x-ray yield
yield from
from the
the Thomson
Thomson scattering
scattering interaction.
interaction. By
By not
not compressing
compressing fully,
fully, itit is
is
also
possible
to
remove
more
of
the
energy
spread
in
the
bunch
by
accelerating
also possible to remove more of the energy spread in the bunch by accelerating off
off
crest
crest in
in subsequent
subsequent accelerator
accelerator sections.
sections. This
This is
is performed
performed by
by accelerating
accelerating the
the bunch
bunch
at
the
opposite
zero
crossing
(180
degree
shift)
of
the
zero
crossing
used
at the opposite zero crossing (180 degree shift) of the zero crossing used in
in first
first
accelerator
accelerator section
section to
to compress
compress the
the beam.
beam. At
At the
the exit
exit of
of the
the linac,
linac, the
the bunch
bunch energy
energy
spread
spread has
has been
been reduced
reduced to
to 0.5
0.5 %
% rms,
rms, while
while the
the rms
rms normalized
normalized emittance
emittance is
is 3.5
3.5
πmm-mrad.
Timm-mrad.
Final
Final focus
focus simulations
simulations were
were performed
performed using
using PARMELA
PARMELA and
and Trace-3D.
Trace-SD. The
The
electrons
electrons are
are transported
transported 33 meters
meters from
from the
the accelerator
accelerator exit
exit and
and focused
focused with
with aa
quadrupole
quadrupole triplet.
triplet. The
The triplet
triplet is
is about
about 50
50 cm
cm long,
long, and
and the
the focal
focal length
length is
is about
about 10
10 cm.
cm.
The
The maximum
maximum field
field gradient
gradient in
in the
the quadrupole
quadrupole magnets
magnets is
is 15
15 T/m.
T/m. The
The rms
rms
convergence
convergence angle
angle of
of the
the focus
focus is
is about
about 33 mrad.
mrad. In
In the
the simulation,
simulation, the
the beam
beam is
is initially
initially
defocused
y, leading
defocused in
in y,
leading to
to aa larger
larger size
size in
in this
this dimension
dimension going
going into
into the
the second
second
quadrupole.
quadrupole. This
This results
results in
in more
more significant
significant chromatic
chromatic aberrations
aberrations in
in the
the yy dimension
dimension
than
than in
in x,
jc, and
and is
is manifested
manifested by
by aa slightly
slightly larger
larger emittance
emittance and
and spot
spot size
size in
in yy at
at the
the
interaction
point,
as
can
be
seen
in
PARMELA
simulations
of
the
final
focus.
Figure
interaction point, as can be seen in PARMELA simulations of the final focus. Figure 33
833
shows the transverse profile of the electron bunch at
at the focus
focus as
as determined
determined by
by
PARMELA. The electron beam focus parameters are
are summarized
summarized in
in Table
Table 1.1.
0.015
0.01 in
0.010
0.010-
y (cm)
0.005
0.0050.000
o.ooo-0.005 -0.005
-0.010
-0.010
-0.015
-O.i
-0.015
.D15
-0.010
-0.010
-0.005
-0.005
0.000
0.000
xx(cm)
(cm)
0.005
0.005
0.010
0.010
0.015
0.015
FIGURE 3.
FIGURE
3. Profile
Profile of
of the
the electron
electron beam
beam at
at focus.
focus.
TABLE 1.
TABLE
1. Electron
Electron Beam
Beam Parameters
Parameters at
at Focus.
Focus.
Parameter
Value
Parameter
Value
3.4
εfxn
xn (mm-mrad)
3.4
(mm-mrad)
5.0
ε£yn
yn (mm-mrad)
5.0
(mm-mrad)
12
σ
12
C7xX (µm)
(urn)
20
σ
(µm)
y
20
<jy (Jim)
0.74
σ
0.74
&tt (ps)
(ps)
Bunch Charge
0.5
Bunch
Charge (nC)
(nC)
0.5
CALCULATION
CALCULATION OF
OF X-RAY
X-RAY PRODUCTION
PRODUCTION
The expected
expected x-ray
The
x-ray production
production was
was calculated
calculated and
and the
the effects
effects of
of RF
RF phase
phase and
and
timing
jitter
were
determined
by
integrating
the
emission
probability
per
unit
timing jitter were determined by integrating the emission probability per unit time,
time,
dNx/dt,
/dt, given
given by
by
dN
x
dN
dN x ( t ) = σc nγ ( x ,t ) ne ( x ,t ) d 3 x,
∫∫∫
dt
dt
JC / ^ \ _
834
(1)
(1)
where
density,
where NNx x isis the
the total
total number
number of
of x-rays
x-rays produced,
produced, «/x,/)
nγ(x,t) is
is the
the laser
laser photon
photon density,
nnee(x,i)
section. The
The calculations
calculations
(x,t) is
is the
the electron
electron density,
density, and
and a
σ is
is total
total Thomson
Thomson cross
cross section.
were
with the
the PARMELA
PARMELA
were performed
performed for
for aa 300
300 ml,
mJ, 300
300 fs
fs laser
laser pulse
pulse in
in conjunction
conjunction with
output
profile,
(x,t). In
In this
this case,
case, «/x,/)
nγ(x,t) was
was assumed
assumed to
to have
have aa Gaussian
Gaussian profile,
output in
in place
place of
of nnee(xj).
given
given by
by


  t − z / c 2 
ηγ
2r 2
-exp
(2)
,
nγ ( r , z , t ) =
exp
2
exp
−
−
 2
 
 
2
2 
1 + ( z / z0 )
  ∆t  
 w0 1 + ( z / z0 )  
where
where zz00 isis the
the laser
laser Rayleigh
Rayleigh length,
length, At
∆t is
is the
the pulse
pulse duration, T|
ηyγ is the peak photon
density,
and
w
is
the
minimum
laser
spot
size.
n
(xj)
is
replaced
with a delta-function
ee(x,t)
density, and w00 is the minimum laser spot
representing
the
position
and
charge
of
each
PARMELA
macro-particle
as a function
representing the position and charge of
of
of time.
time. Figure
Figure 44 shows
shows the
the calculated
calculated x-ray
x-ray pulse
pulse profile
profile in time. The average photon
19
energy
energy isis 30
30 kV,
kV, and
and the
the peak
peak x-ray
x-ray flux
flux is about 6 x 10 photons/s with an integrated
88
photon
photon yield
yield of
of about
about 10 .. A
A 3-D
3-D frequency
frequency domain calculation [3] of the x-ray output
has
has also
also been
been performed,
performed, showing
showing the
the spectral bandwidth to be about 10%. The peak
20
2
2
spectral
photons/s/0.1% bandwidth/mm /mrad ..
spectral brightness
brightness is
is calculated
calculated to
to be
be 1020 photons/s/0.1%
Effects
Effects of
of phase
phase jitter
jitter were simulated by varying the relative phase of the linac
sections
sections with
with respect
respect to
to the
the RF
RF gun
gun in
in the PARMELA simulations, and using the
resulting
resulting electron
electron beam
beam parameters
parameters in the calculation of the x-ray production. Figure 55
shows
shows the
the expected
expected x-ray
x-ray yield
yield versus
versus phase
phase jitter
jitter for the cases where the electron
beam
beam isis compressed
compressed in
in the
the first
first section
section (accelerated near the zero crossing), and not
compressed
compressed (accelerated
(accelerated on
on crest).
crest).
Flux (Nxx/s)
/s) as
as aa function
function of time (ps
Flux
19
6 .10
6-10
4.8 .10
4.8-10
19
19
3.6 .10
3.6-10
2.4 .10
2.4-10
19
19
1.2 .10
1.2-10
0
3.75 -2.5
2.5 -1.25
1.25
~55 -3.75
0
1.25
2.5
3.75
5
8
FIGURE4.
4. X-ray
X-ray flux
flux (photon/s)
(photon/s) vs.
vs. time
time (ps).
(ps). The
The total
total number of photons in the pulse is about 10
FIGURE
108,
andthe
the average
average photon
photon energy
energy is
is 30kV.
30kV.
and
835
Integerated X-ray Dose (x107)
25
20
15
10
5
0
-5
-3
-
3
-1
-
1
11
33
5
5
Phase
Phase Shift
Shift (degrees)
(degrees)
FIGURE5.5.Simulated
Simulatedx-ray
x-rayyield
yieldversus
versusphase
phaseshift
shift (Solid:
(Solid: compressed
compressed beam.
beam. Dotted: uncompressed).
FIGURE
seenthat
that the
the phase
phase jitter
jitter requirement
requirement for
for the compressed beam is less than
ItItisisseen
than one
one
degree(or
(orabout
about11ps).
ps).Thus,
Thus, picosecond
picosecond phase
phase and
and timing
timing control
control will be required to
degree
maintainstability
stabilityof
ofthe
thex-ray
x-ray source
source when
when velocity
velocity compression
compression is implemented.
maintain
ELECTRON BEAM
BEAM CHARACTERIZATION
CHARACTERIZATION
ELECTRON
Todate,
date,electron
electronbunches
buncheswith
with up
up to
to 700
700 pC
pC of
of charge
charge have
have been
been produced with up to
To
100µJjoJofofUV
UVlaser
laser energy
energy incident
incident on
on the
the gun
gun photo-cathode.
photo-cathode. The beam has been
100
been
transportedthrough
throughthe
thelinac
linac and
and accelerated
accelerated up
up to
to 60
60 MeV.
MeV. Emittance
Emittance measurements
transported
havebeen
beenperformed
performed using
using aa quad
quad scan.
scan. Figure
Figure 66 shows
shows quad
quad scan
scan data
data for
for aa 300
have
300 pC,
pC,
60 MeV
MeV bunch.
bunch. The
The measurements
measurements were
were performed
performed with
with aa 10
10 cm
cm effective
effective length
60
length
quadrupolemagnet
magnet placed
placed after
after the
the final
final linac
linac section.
section. The
The beam
beam profile
profile was
was imaged
quadrupole
imaged
withaaYAG
YAGscintillator
scintillator placed
placed 1.42
1.42 meters
meters from
from the
the quadrupole.
quadrupole. The
with
The rms
rms normalized
normalized
emittanceisis14.6
14.6πmm-mrad,
Timm-mrad,and
andthe
the Twiss
Twiss parameters
parameters are
are given
given by
emittance
by aα =
= -1.15
-1.15 and
and (3
β ==
7.84mm/mrad.
mm/mrad.Improvements
Improvementsin
inemittance
emittance are
are expected
expected with
with improvements
improvements in
7.84
in the
the UV
UV
drivelaser
laseruniformity
uniformity and
andoptimization
optimization of
of the
the electron
electron beam
beam transport.
transport.
drive
Phase
jitter
between
the
drive
laser
oscillator
and
the
RF phase
Phase jitter between the drive laser oscillator and the RF
phase in
in the
the gun
gun has
has
been
measured
by
mixing
a
2.8
GHz
signal
generated
by
the
laser
oscillator
with aa
been measured by mixing a 2.8 GHz signal generated by the laser oscillator with
signalproduced
produced by
by aa probe
probe inside
inside the
the RF
RF gun.
gun. This
This has
has shown
shown short
signal
short time
time scale
scale (<
(< 11
minute)
phase
stability
of
±
1
degree.
Longer
term
phase
drifts
will
be
eliminated
minute) phase stability of ± 1 degree. Longer term phase drifts will be eliminated by
by aa
computer
controlled
feed
back
loop
and
voltage
controlled
phase
shifter.
Plans
are
computer controlled feed back loop and voltage controlled phase shifter. Plans are also
also
underwaytoto implement
implement aa direct
direct measurement
measurement of
of the
the UV
UV drive
drive laser
laser arrival
underway
arrival phase
phase in
in
theRF
RFgun.
gun.The
Thedirect
directphase
phase measurement
measurement will
will employ
employ aa 20
20 GHz
GHz bandwidth,
bandwidth, 800
800 nm
nm
the
opticalfiber
fiber switch
switchmodulated
modulated by
by microwaves
microwaves sampled
sampled from
from the
the RF
RF gun
gun to
optical
to provide
provide aa
836
nonlinear correlation between the gun fields and the IR laser pulse used to produce the
UV photocathode drive laser.
0.8
_
0.6 -
E.
«r 0.4
'
0.2
0
0.5
1
1.5
2
2.5
3
dB/dx (T/m)
FIGURE 6. Quad scan measurement of 60 MeV electron beam. en = 14.6 Tcmm-mrad.
CONCLUSIONS
The PLEIADES Thomson x-ray source facility will provide high brightness (>
107 x-rays/pulse), picosecond pulses for dynamic measurements in matter, including
radiography, dynamic diffraction, and spectroscopy. The PLEIADES electron
beamline will be capable of producing the high brightness electron beam needed to
drive this source. Simulations have shown that a 1 ps, 500 pC bunch with an rms
normalized emittance of less than 5 Timm-mrad should be achievable, and will allow
for the attainment of a 20-30 jim spot size at the interaction point. Initial electron
beam measurements have been performed, and expected performance parameters and
planned jitter measurements and control methods have been presented.
ACKNOWLEDGMENTS
This work was performed under the auspices of the U.S. Department of Energy by the
University of California, Lawrence Livermore National Laboratory under contract No.
W-7405-Eng-48.
837
REFERENCES
1. W. Schoelein, et. al., "Femtosecond x-ray pulses at 0.4 Angstrom generated by 90 degrees
Thomson scattering: A tool for probing structural dynamics of materials," Science 274, 236
(1996).
2. S.G. Anderson, J.B. Rosenzweig, G.P. LeSage, J. Crane, "Emittance measurements of the
space charge dominated Thomson source photoinjector," Proceedings of the 2001 Particle
Accelerator Conference (Cat. No.01CH37268), Piscataway, NJ, IEEE, p.2260-2 (2001).
3. F.V. Hartemann, H.A. Baldis, A.K. Kerman, A. Le Foil, N.C. Luhmann, B. Rupp, "Threedimensional theory of emittance in Compton scattering and X-ray protein crystallography,"
Physical Review E, vol.64, (no.l), p.016501/1-26. (2001).
838