Tabletop laser driven accelerator research at LBNL

1
Table-top laser driven accelerator research at
LBNL Wim Leemans LOASIS Group

• Outline
• Introduction into laser accelerators
• Experiments on controlling performance
• Gas jet profile
• Laser pulse shape
• Colliding pulse
• Laser guiding
• Applications

JLAB Seminar November 15, 2002
2
Group, visitors and collaborators
Collaborators ALS J. Byrd M.
Martin CBP J. Corlett D. Li CFI J.
ONeil H. Vanbrocklin EHS R.Donahue
D. Rodgers A.
Smith Engineering M. Hoff
D. Oshatz S. Virostek
N. Ybarazolla U. Colorado/TechX-Corp
D. Bruhwihler J. Cary D.
Dimitrov R. Giacone
Experiment J. Faure C. Geddes (UC Berkeley, Fac.
Adv. Wurtele) W. Leemans C. Toth J. Van Tilborg
(TUE, NL) Theory E. Esarey G. Fubiani (Univ. of
Paris) C. Schroeder B. Shadwick Techs L.
Archambault (retired) M. Dickinson S.
DiMaggio R. Short D. Suyversrud J. Wallig
Visitors D. Auerbach
(Princeton) G. Dugan (Cornell)
B. Marcelis (TUE) A. Misuri
(Univ. Pisa) A. Reitsma (TUE)
R. Trines (TUE)
3
(No Transcript)
4
Non-linear physics with compact CPA lasers

• LOASIS laser
• 1019 - 1020 W/cm2
• 10 Hz
• Table top
• Plasma physics
• Non-linear interaction
• Particle acceleration

From T. Tajima and G. Mourou, PR-STAB 5, 031301
(2002)
5
Laser driven accelerators introduction
We Ez Lint

• Conventional RF accelerators 10-100 MV/m,
  100-1000 m for 10 GeV
• Laser-plasma based 10 - 100 GV/m, 0.1 - 1 m for
  10 GeV

• Optical injection
• Self-modulated laser wakefield
• Colliding pulse injector

Laser 0.8 mm, multi-beam, 10 TW, 10 Hz
gt 40 fs, 100 TW upgrade
6
Laser excitation of plasma waves
Standard/Resonant Laser Wakefield
Vpondmc2ao2/4e where ao eE/wmc
Laser excites plasma wave Ez gt 10 GV/m
7
Self-Modulated-Laser Wakefield Accelerator

• Power gt Pcritical 17 (ncritical/n) GW
  e.g. 1 TW for 3x1019cm-3 and l800 nm
• Laser pulse length gtgtPlasma period

8
On the need for laser guiding

• Standard laser wakefield
• Short pulses not guided
• Requires preformed channel
• Requires external/injected electrons

• Self-modulated laser wakefield
• Relativistic self-focusing
• P gt PcGW 17 (nc/n)
• e.g. 1TW requires 3x1019cm-3
• Requires sufficient density
• Experimentally easy

9
Laser-Plasma Accelerator experiments

• Acceleration of externally injected electrons
• Beat-wave accelerator ('93)
• 0.7 GV/m, few tens of e-s, 9 MeV gain
• Groups UCLA, EP (France), NRC (Canada), ILE-KEK
  (Japan)
• Self-modulated laser wakefield ('95)
• 30 GV/m, few tens of e-'s, 10 MeV
• Groups ILE-KEK
• Standard laser wakefield accelerator ('98)
• 1.5 GV/m, few tens of e-'s, 1.6 MeV gain
• Groups EP, ILE-KEK
• Acceleration of self-trapped electrons
• Single-shot lasers ('95-??)
• gt100 GV/m, nC, up to 100 MeV, collimated beams
• Groups RAL, NRL, U. Michigan
• High rep. rate systems ('99 - ??)
• 10 Hz, 10 TW, Radio-isotopes
• Groups LBNL, LOA, U. Michigan, MPQ Garching,

10
lOASIS Facility

• Test bed for RD concepts towards 1 GeV module
  of a laser accelerator and applications
• Facility includes 10 TW, 50 fs laser system _at_ 10
  Hz (100 TW under development)

11
Self Modulated Laser Wakefield Accelerator Setup
H
Pb
I
M1
CCD
Hc
Laser Compressor
ICT
I
G
Magnet
H
TiAl2O3 laser 400-500 mJ t gt 42 fs 10 Hz 6 mm
spot
e-beam
Jet
PbCu Target (Removable)
OAP
SPEC
Int
SSA
FROG
H
He3 Detector
I
Ion Chamber
Mirror or Splitter
Cd Lined He3 Detector
Hc
Flip-in Mirror
G
GM Counter
Phosphor
12
Divergence reduces to 10 mrad at highest peak
power
E-beam on phosphor screen, 75 cm away
10 mrad
13
Accelerator performance strongly affected by jet
properties
14
Control knobs for laser accelerators Scanning
laser power using compressor system
W.P. Leemans et al., PRL 89, 174802 (2002)

• Laser
• 420 mJ/pulse
• 6 micron spot

• Plasma
• n 2-3x1019 cm-3
• Gaussian (s530 mm)

Peak yield does not occur for shortest pulse
(highest peak power)
15
Grating position changes laser duration and chirp

• Grating separation determines
• Pulse duration
• Correlated frequency chirp

Input beam, ??200 ps
Cam1
Grating 1
Output beam
??50 fs
positive chirp
negative chirp
Grating 2
Pulse duration, FWHM (fs)
Relative grating separation (?m)
experimental bandwidth ??/? 2.6
frequency chirp ? bandwidth
16
Grating pair position changes pulse shape
stretcher-compressor
positive chirp
negative chirp
Higher-order phase dispersive terms in the
laser system Fiorini et al. (94) Backus et al.
(98)
Pulse duration, FWHM (fs)
76 fs
will modify the temporal pulse shape.
Experimental pulse profiles are well-fit to a
skewed Gaussian
Relative grating separation (?m)
17
Asymmetry in electron yield vs. pulse
duration Laser envelope shaping provides
accelerator control knob

• Sharp front larger seed underneath laser pulse
• Self-modulation reaches saturation at earlier
  position in plasma
• Offers new path for enhanced control of wake
  excitation

• Pulse shape depends on compressor position

W.P. Leemans et al., PRL 89, 174802 (2002)
18
Asymmetry in seed depends on plasma density
Low density (Case 1)
High density (Case 2)
19
Electron yield seed and skew asymmetry
Case 2
Fit to plasma density profile ( x 5E19 cm-3)
Bunch Charge nC
Laser pulse duration FWHM fs
Case 2
Case 1
pulse duration
Case 1
Compressor Position microns
Case 1 Laser focused at edge of gas jet (2x1018
cm-3) Case 2 Laser focused near gas jet center
(gt 3x1019 cm-3)
20
Colliding Pulse Injection three-pulse scheme
Esarey et al. (97) Schroeder et al. (99)

• Standard LWFA regime
• Requires 3 laser beams
• drive beam
• 2 colliding beams
• Slow laser beat wave

Normalized potentials
2.0
Trapped Focused Wake Orbit
Trapped Wake Orbit
1.5
plasma wake phase
1.0
0.5
Untrapped Wake Orbit
0
-0.5
Beat Wave Separatrices
-1.0
-1.5
-2
-1
0
2
3
1
21
Colliding pulse injector two-pulse scheme

• Femtosecond bunches
• Few energy spread
• 1-10 pC/bunch
• 40 MeV in 1 mm
• Emittance lt 1 micron

22
Colliding pulse injector beam quality
1010
1010
gvx (m/s)
gvx (m/s)
0
0
550 m
660 m
80 m
y
0
x

• Energy spread (2s) 6 on 13MeV 0.8MeV, but
  should go down with increasing energy
• Beam pulse length of 10.6 fs (2 s)
• Width sy 1.08x10-6m
• Off-axis momentum spy/px 0.02
• Transverse emittance 0.022 p mm-mrad
• g spy sy 0.28 mm-mrad

J. Cary and R. Giacone, CU David Bruhwihler et
al., Tech X
23
Colliding pulse experimental set-up at lOASIS lab
Target chamber
Compressors 12
Compressor 0
Optical diagnostics
24
First phase colliding pulse drive beam 1
non-collinear colliding beam
Top view

• Focusing spatial overlapping of beams
  dumb-bellinterferometer
• Femtosecond synchronization

Side view
25
Enhanced electron yield seen with colliding
beam on
Enhanced electron yield seen with colliding
beam on
Y and Timing scans do not show Sensistive
dependence Further allignment needed
Charge,nC
Both beams
0.2
0.1
Main beam
0.0
.08
.04
T, hours
26
Two-stage laser wakefield accelerator
stage 1
stage 2
(channel-guided) LWFA
SM-LWFA
e- beam
SM-LWFA Plasma density ne gt 1019 cm-3 High
charge yield up to 8 nC Broad energy
spectrum Boltzmann distribution
(temperature few MeV) High energy electrons
collimated ( 10 mrad divergence)
SM-LWFA (channel-guided) LWFA Plasma density ne
1018 cm-3 (Lchannel3 mm) Capture 40 of
injected bunch charge Increased average energy
50 MeV Fractional energy spread reduced to
60

• Ref A. Reitsma, W. P. Leemans, E. Esarey et al.,
  PRST-AB,5, 051301 (2002)

27
1 GeV module SM-LWFA injector channel with 100
TW laser
28
First data from Ignitor - Heater on 10TW system
Channel side Interferograms
First ignitor heater sparks produced on 10TW
system
Ignitor Only - 20mJ
Heater Only - 300mJ
Diagnostics Interferometry Side/front/back
scatter Mode image
Ignitor Heater
29
Integrating injector and channel
experiments increasing mean energy

• Photo-nuclear reaction sensitive to g-ray energy
• Target design
• Pb e-beam dump -gt g's
• Cu 63Cu -gt 62Cu,61Cu
• Pb and Cu bracket 8-30 MeV
• Observed gt 2.5 mCi activity in 3 hr
• Na24 from Al27(n,a) i.e. gt6 MeV neutrons
• Monte-Carlo simulations for Cu62 yield agree
  with experiment

30
Simulation of an electron bunch with large energy
spread (use of a distribution function uniform in
position and angle over the spheroids)
Space charge blowout occurs at a very early stage
(50-100 mm) then the motion is dominantly
ballistic. We clearly see a low energy tail
formation and a highly non linear interaction
between particles.
31
Ion acceleration using multi-terawatt lasers

• Experiments at LLNL, RAL-VULCAN, U. Michigan,
  LULI, LOA, MPQ, ILE-Osaka
• LLNL, Petawatt few 100 J in 0.5-5 ps, few
  shots/day, Au and CH targets (50-150 mm), 3x1013
  protons (30 J), several MeV, ions (lt55 MeV)

• Impact
• Compact hadron injector
• Isotope production
• Particle probe or ignitor for fusion
• lOASIS upgrade 100 TW laser system, 10 Hz
• Activation of positron emitters using (g,n) and
  (p,n) reactions

32
Summary

• Active RD around world
• LBNL, LOA, LULI, MPQ, MBI, NRL, RAL, UCLA, U
  Maryland, UMichigan, UT Austin,
• Ultra-high gradient gt 100 GV/m
• Up to 200 MeV electrons observed (LOA), but 100
  energy spread
• Multi-nC bunches, femtosecond, micron source
  size, few mrads divergence, 10 Hz
• Electron yield control/enhancement using
• Shaped/skewed pulses
• Plasma profile and location of laser focus
• Additional laser pulses (colliding pulse, LILAC)
• Electron energy
• Channel guiding to extend distance
• Optimized stage design plasma tapering or
  operation in pump depletion regime

33
Conclusion possibilities and challenges

• Progress in laser driven accelerators intimately
  tied to
• Laser performance
• Stability laser power (-level), pointing,
• Average power 100 TW, 10 Hz 30-40 W system
• Compactness
• Efficiency low but diode pumped systems being
  RDd
• Plasma source
• Gas jets, gas cells
• Channel production over extended distance
• Reproducibility and controllability
• Development of bunch diagnostics with fs, micron
  resolution
• Possibilities (next 5 years)
• Electron and ion injectors
• Compact 1 GeV accelerator 1 GeV in 10 cm, 100
  pC/bunch
• Coherent mid to far - infrared source
• Radio-isotope production using (g,n) or (x,n)
  where xp,d,a,