You are on page 1of 26

Production of femtosecond pulses and

micron beam spots for high brightness


electron beam applications

Scott G. Anderson*, D. J. Gibson, F. V. Hartemann,


J. S. Jacob, A. M. Tremaine — LLNL

J. K. Lim, P. Frigola, J. B. Rosenzweig, G. Travish


— UCLA

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, and
by the University of California, Los Angeles under contract numbers DE-FG-98ER45693 and
DE-FG03-92ER40693

*e-mail: anderson131@llnl.gov
Advanced applications require high-
brightness, dense beams
Collision of ultra-relativistic electron with
Scaling of applications to higher focused laser pulse
performance requires shorter bunch
dimensions; examples: advanced
accelerators, next generation light sources.

Hard x-rays (10-80 keV) use only moderate energies:

hc
!w " 0.8 µm, " 1.55 eV, # me c 2 " 20 $ 60 MeV
e!w

Performance depends on intensities of overlapping beams:


dNx Total 2Ne N" #
Flux: (t) = "c[1# $ % k] ''' n& (x,t)ne (x,t) d 3x Nx =
dt dose: $ (2rb2 + w02 )

Source Bx =
Nx X-ray applications need bright
!
brightness: ( 2! )5 /2 " # " x2" x2$ (.1%BW) beams! with short !-functions and
2 sub-ps duration.
! geom
The PLEIADES x-ray source uses
velocity bunching and PMQ focusing
• LLNL-UCLA collaboration Measured x-ray flux vs. energy
108
• Tunable, bright, ICS hard x-ray

X-ray dose (photons)


source
107
• 1.6 cell S-band photo-gun + 4 "
2.5 m SLAC sections gives 20-80 106
MeV, 250 pC, 3 ps (rms), 10
mm-mrad 105

• 800 nm, 250 mJ, 50 fsec,


Ti:Sapphire laser 104
40 60 80 100 120 140
X-ray Energy (keV)

• Compress ps photo-beam
using velocity bunching E-beam compression and PMQ
focusing have significantly
• Use ultra-strong permanent increased source flux and
magnet focusing at ICS brightness.
interaction point
PLEIADES interaction beamline
Compressed Beam
Pulse Length Measurement
1.2 0.4
Autocorrelation Signal (normalized)

Data
1 Fitted Curve E=59.2 MeV
0.8
$%= .2 %

!x (mm)
0.6 0.2

0.4
! < 0.3 psec
t
#n = 9.3 mm-mrad
0.2

0 0
0 5 10 15 20
Delay Arm Position (psec)
2.5 3.5 4.5 5.5
dB/dx (T/m)
Permanent magnet quadrupole Dipole spectrometer
final focus system (energy spread/linac tuning)

e-beam

Linac
laser

Interaction Point Coherent Transition Radiation (CTR) Quadrupole lenses


(OTR to CCD/streak cameras based pulse length diagnostic (emittance measurement and
for spatial/temporal overlap) beam matching)
Magnetic compression of photoinjector
beams gives emittance growth

• Magnetic chicane
compressors produce sub-
picosecond beams
• Collective effects in chicane
cause phase space distortions
– CSR at moderate/high energies
– Space-charge at low energies

• Resulting emittance growth


and energy spread are a
major concern for applications
Transverse phase space
low energy (space-charge)
effect
Velocity compression as an alternative
Longitudinal phase space during compression
• Beam is injected ahead of 30
peak accelerating phase,
and compresses as it slips 25

back in phase.
20

Particle equations of motion


in traveling-wave field: 15

!
10

0
-1.5 -1 -0.5 0 0.5 1 1.5

8.02 " (radians)


80.5

80
Final bunch length: 8.01

8 79.5

7.99 79
!

!
7.98
78.5
7.97
78
7.96
(a) (b)
77.5
7.95

7.94 77
-4 -2 0 2 4 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15
" (degrees) "# (degrees)
Velocity bunching results:
longitudinal dynamics
Simulation Measurements
0.8 4 Interferometer Data
2
0.7 3.5
y = m1+m2*( exp(-(m0-m3)^ 2/...
Value Error
0.6 3 1.8
m1 1.0818 0.085653
m2 0.89985 0.058787
0.5 2.5

"p/p (%)
! (mm)

CTR Signal (normalized)


m3 31.639 0.028427
1.6m4 0.28401 0.02317
m5 5.7725 10.53
0.4 2 Chisq 0.43202 NA
1.4 R 0.93766 NA
z

0.3 1.5

0.2 1 1.2

0.1 0.5
1
0 0
0 5 10 15
z (m) 0.8
Compressing Section ! = 0.28 +/- 0.02 ps
t
S-band rf gun Accelerating Sections
0.6
26 28 30 32 34 36

Final beam parameters: Delay Arm Position (ps)

compressed nominal energy spread

E = 50MeV E = 59MeV
"# "# 0.5%
= 0.5% = 0.2%
# #
" t = 200 fs " t = 3ps
Velocity bunching results:
transverse dynamics
Simulated beam matching • Simulations predict matching critical for
Well matched emittance conservation
0.3 6
• Quad scan measurement agrees!

Norm. RMS Emittance


0.25 5
RMS Size (cm)

0.2 4

(mm mrad)
0.15 3
Beam Size
0.1 Emittance 2 Measured emittance
40
0.05 1
Horizontal

Normalized Eimttance (mm mrad)


0 0 Verticle
0 100 200 300 400 500 600 700 800
Z (cm) 30

Poorly matched
0.3 7 20
Norm. RMS Emittance

0.25 6
RMS Size (cm)

5
(mm mrad)

0.2
4 10
0.15
3 Uncompressed Emittance
0.1
2
0.05 0
1 4 6 8 10 12 14
0 0
0 100 200 300 400 500 600 700 800 Compressor Solenoid Current (Amps)
Z (cm)
Permanent magnet quadrupoles
achieve ultra-high field gradients
• Chromatic aberration limits
demagnification; need increased B!
p * #p
! 0,opt = " 0 f ; ! min = 2 f "0
#p p

• Idealized, cylindrical PMQ


achieves field gradient: Final spot size vs initial !

#1 1&
B! = 2Br % " (
$ ri ro '
with Br = 1.22T (NdFeB), ri = 2.5mm,
T
and ro = 7.5mm, B! = 640
m
" B! % Segmented
klq = $ l ( 0.5 ) lq * 1 cm
# BR '& q Halbach design Built PMQ
Adjustable PMQ final focus system

+2f -f +f
• Final focus modeled as
simple thin lens triplet
– For PMQ focal length,
fPMQ=BR/B!lq, strongest
triplet focusing is in F-
DD-FF configuration
• System adjusted by Asymmetric triplet model
magnet spacing; L1, L2,
L3
• Trace3D (and
experiments) verify
system tunable to focus
50-90 MeV beams; final
!-functions in 3-5 mm
range
PMQ mover assembly
Final focus performance is enhanced
with PMQ system
OTR image of 70 MeV, 200pC,
• Final focus procedure: 20 µm (rms) final focus
– Perform quad scan with up-stream
magnets
– Use Trace3D to compute EM quad
settings for ~ few meter !0 and PMQ
positions for best focus

• IP spot measured with OTR +


3 µm/pixel video camera
– Measurement problematic;
sensitive to aberrations in
camera lens
– < 20 µm spots directly
measured
• PMQ scan indicates $* = 15
µm
PMQ scan shows !* = 3 mm
X-ray generation using velocity
bunched, PMQ focused beam
• X-ray generation issues:
– Final focus spot size increase
due to increased emittance,
energy spread of compressed
beam
– Shot-to-shot jitter and short term
drift (~minute scale) OTR image of velocity bunched
beam; 30 µm (rms) final focus

X-ray Beam Properties


Parameter Value
Photon energy 65 keV (300 fs),
30-140 keV (3 ps)
Number of photons 106 - 107
Source dimensions 20 µm " 5-10 mrad
Peak Brightness > 1016 &/(s-mm2-mrad2- X-ray CCD image of 300 fs,
0.1%BW) 65 keV, ICS photons
The PLIEADES X-ray source is well modeled

Theoretical Measured
• 3-D time and frequency 20 20

domain code developed to 10 10


model source.

'y (mrad)

'y (mrad)
0 0
– W.J. Brown, et al., Phys. Rev.
ST Accel. Beams, 7 p. 060702 -10 -10

(2004). -20 -20

• The x-rays measured with the -20 -10 0 10 20 -20 -10 0


'x (mrad)
10 20
'x (mrad)
PLEIADES system matched
1
the theoretical flux and 0.9
Theoretical
Measured

Energy Density (a.u.)


profiles very well, once all the 0.8
0.7
electron and laser beam 0.6
0.5
parameters, material 0.4
0.3
transmission, and CCD 0.2

response were taken into 0.1


0
account. -0.02 -0.01
Angle Along x-Axis (rad)
0 0.01 0.02
Spectral–Angular Correlation

0.005” Tantalum Foil


The dependence of x-ray energy
Theoretical Transmission (K-edge 67.46 keV)
on emission angle was observed
using filter absorption edges.

Simulated Spectral Dependence on


Divergence (! x, ! y , F) (! x, ! y , F)
80
Eelectron=55 MeV Eelectron=57 MeV
Ex-ray,peak=73.1 keV Ex-ray,peak=78.5 keV
X-Ray Energy

0.005” Tantalum Foil


(keV)

70
Measured Transmission (K-edge 67.46 keV)

60
0 5 10
Divergence Angle (mrad)

X-rays Ta Foil
E=E(')

CCD Eelectron=54.9 MeV Eelectron=56.9 MeV


Ex-ray,peak=72.8 keV Ex-ray,peak=78.2 keV
Diffraction Measurements
X-Ray Diffraction from InSb showing
Sn K-edge (round aperture)
Saturated
Main Signal

Tin blocks
half the
aperture
CsI Scintillator
InSb Wafer
Pb Aperture

K-edge
energy
Brightness (a.u.)

Diffraction of the x-ray beam has been observed off


several crystal materials, including graphite (C), gold (Au),
and indium antimonide (InSb). Because of the small
source size, the divergent x-ray beam that reflects off the
crystal has a strong correlation between the position on the
CCD detector and the energy of the x-ray photons. Placing
X-ray Energy (keV) a foil with a K-edge in the range of the energy of the
The x-ray spectrum was measured by diffracted electrons creates a notch in the diffracted image.
looking at the diffraction intensity as a
function of angle of incidence upon an It is the dark region produced this way that can allow for
InSb (111) crystal. observation of dynamic diffraction effects.
Summary

• Velocity bunching is a very effective compression


technique at moderate energies.
– " 300 fsec, 250 pC bunches produced; stability sufficient for
PLEIADES experiment
– Emittance control possible with solenoid focusing, but not
perfect; needs more study with low emittance system.
• A Halbach type PMQ final focus system has been
developed [J. K. Lim, et al., Phys. Rev. ST Accel. Beams, 8, p.
072401 (2005).]
– Ultra-high field gradient — 560 T/m
– " 20 µm spots, 3-5 mm !-functions routinely generated

Both techniques enhance Beam density in 1015 cm-3 range;


PLEIADES performance ~104 higher than at gun exit
Compton x-ray production

E-beam issues for x-ray generation:


• Focused spot size
• Timing stability

Final focus spots Timing jitter


30

Time Difference (psec)


Laser/Electron Arrival
600 µm 600 µm 20

10

-10

uncompressed compressed -20


$=27 µm $=43 µm
-30
2 4 6 8 10
Measurement Number
X-rays generated

X-ray CCD Images

Compressed Uncompressed
electrons electrons
N"
X-ray brightness: Bx = 52
(2# ) $ t$ x2$ x2% (0.1%BW )
N" & # t $
Brightness increase 70%
!
# x% &
PLEIADES experimental beamline

E=59.2 MeV
$%= .2 %

Energy spread measurement

e-beam

Linac
laser

Dipole spectrometer
(energy spread/linac tuning)
PLEIADES experimental beamline
Compressed Beam
Pulse Length Measurement
1.2

Autocorrelation Signal (normalized)


1

0.8
Bunch length measurement
0.6

0.4
! < 0.3 psec
t
0.2

0
0 5 10 15 20
Delay Arm Position (psec)

e-beam

Linac
laser

Coherent Transition Radiation (CTR)


based pulse length diagnostic
PLEIADES experimental beamline

560 T/m field gradient!

PMQ mover system Single PMQ

e-beam

Linac
laser

Permanent magnet quadrupole


final focus system
3D simulations quantified design
tolerances
• RADIA — 3D magnetostatic field solver simulate design
parameters: ID = 5 mm, OD = 15 mm, L = 10 mm, and Br = 1.22
T

Vector field plot Linearity good to r ~ 2 mm


leff = 10.4 mm
in bore region B! = 573 T/m

• RADIA + ELEGANT error studies performed to find manufacturing


tolerances
– ± 50 µm bore radius error ( ± 3% B! variation
– 2% wedge shape and easy axis orientation allowable
– 10 mrad rotation (skew) error produces significant emittance growth
Measurements of built PMQs agree
with RADIA simulations

• Manufacturing process
ensures consistency
between PMQs, minimizes
skew errors.
PMQ construction steps

Pulsed-wire scan verifies field linearity


to r ~ 2 mm. Magnetic and
mechanical centers within 25 µm Hall probe measurement
gives B! = 560 T/m
Adjustable PMQ final focus system

+2f -f +f
• Final focus modeled as
simple thin lens triplet
– For PMQ focal length,
fPMQ=BR/B!lq, strongest
triplet focusing is in F-DD-
FF configuration
• System adjusted by magnet
spacing; L1, L2, L3 Asymmetric triplet model
• Trace3D (and experiments)
verify system tunable to
focus 50-90 MeV beams;
final !-functions in 3-5 mm
range

Trace3D tunes at 70 & 90 MeV


The PMQ mover system meets
experimental requirements
• CNC machined “PMQ holders”
constrained by rail system
– < 1 mil PMQ to system center-line
throughout range of motion
• Push-rods + stepper motors +
LabVIEW for on-line, < 50 µm
resolution longitudinal positioning
PMQ mover assembly

• Alignment verified optically


(theodolite) in PLEIADES
beamline

PLEIADES PMQ final focus


Advanced applications require high-
brightness, dense beams
• Scaling of applications to higher performance requires
shorter bunch dimensions; examples: advanced
accelerators, next generation light sources.
• Plasma wake-field accelerator (PWFA) scaling:
Wave-breaking field
PWFA longitudinal field
EWB = me c 2 k p / e

Efficiently excited wake amplitude

E ! k p2 N b if k p! z < 1

Ultra-strong focusing

! eq = 2" k p#1

You might also like