Ultrafast Laser-Driven Wakefield Accelerators

Oleg Korovyanko 01/12/2009SLAC AARD seminar

Outline

Part 1: Wakefield accelerators:

techniques to generate short

e bunches

Part 2: Production of quality

electron beams,

characterization and

applications

Part 3: Relevant laser

techniques

Part 4: Conclusion and

perspectives

RF vs Plasma

Plasma cavity

100 m

E-field max~ 10 MeV /m

Courtesy of V. Malka

RF cavity1 m

E-field max~ 10 GeV/m

DWADiel. surface field breaks

down @~ 10 GeV/m

                                                                                                   

                                                                                             

Argonne Wakefield Accelerator1.3 GHz, TSA 50

UV Pulse duration after UV stretcher (SUMMARY)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

-25.00 -20.00 -15.00 -10.00 -5.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00

Relative position from main pulse [psec]

Norm

alized

ampli

tude

UV pulse width, SHORTEST of SHORT RANGE (0.87ps)

UV pulse width, LONGEST of SHORT RANGE (11.1ps)

UV pulse width, SHORTEST of LONG RANGE (7.6ps)

UV pulse width, LONGEST of LONG RANGE (18.5ps)

Steve Fournier, PSG RLS

CONFIDENTIAL 06/23/08

Double surface reflection from BS of the IR beam, cause second Xcorrellation pulse. Artificial side post-pulse @ ~18.75ps away from the main pulse.

SHGxtal

> 1.6 W @ 400

nm

THGxtal

D1IR1

IR4

IR2-3

SHG1 SHG2

D2

FS

D3-4

WP

> 900 mW

@ 266nm

ANL AWA1.3 GHz, TSA 50

• Built as DWA: witness, drive bunches

• Two 248nm pulses go to photocathode of RF gun, one or several drive bunches

inter-pulse separation controlled w/ mechanical delay stage 23 cm, ~770ps, or 10.5Lo, Lo=22mm

• A new UV stretcher utilizes thick BBO crystals in series

• Laser mode at photocathode: adjustable iris at 1 m from photocathode

Escalations

Principle

Monoenergetic Beams from Literature

Name Article Lab Energy dE/E Charge Ne IntensityL/Tp

[MeV] [%]

[pC]

Mangles Nature (04) RAL 73 6 22 20 2,5 1,6

Geddes Nature (04) L'OASIS 86 2 320 19 11 2,2

Faure Nature (04) LOA 170 25 500 6 3 0,7

Hidding PRL (2006) JETI 47 9 0,32 40 50 4,6

Hsieh PRL (2006) IAMS 55 336 40 2,6

Hosokai PRE (2006) U. Tokyo 11,5 10 10 80 22 3,0

Miura APL (2005) AIST 7 20 432E-6 130 5 5,1

Hafz PRE (2006) KERI 4,3 93 200 28 1 33,4

Mori ArXiv (06) JAERI 20 24 0,8 50 0,9 4,5

Mangles PRL (2006) Lund LC 150 20 20 5 1,4

x1018 /cm3

x1018 W/cm2]

State-of-art gradient27 GeV/m, SLAC, 27 GeV drive,

Nature’2007

Towards longer interaction length

Diffraction length L~r2/Rayleigh

Dephasing length ~ ap 3/ 2

Pump depletion length a

• Expanding Bubble Injection regime Degrades emittance due to high transverse

field – control trapping

Pre-formed channel injection : plasma “fiber”

Optical injection by colliding pulse

Capillary discharge channel

LOA

Experimental set-up

Injection beam Pump beam

Probebeam

LANEX

B Field

250 mJ, 30 fs fwhm=30 µm I ~ 4×1017 W/cm2

a1=0.4

700 mJ, 30 fs, fwhm=16 µmI ~ 3×1018 W/cm2

a0=1.2

electron spectrometer to shadowgraphy diagnostic

Gasjet

LBNL Group

After 5 Zr / 7.5 mm

0

0.5

1

1.5

2

2.5

800 1200 1600 2000Energy (MeV)

f(E) (a.u.)

w020m 30fs a

04

0.8mP200 TW n

p1.5×1018cm

3

Courtesy of UCLA& Golp groups

Laser plasma injector : GeV electron beams

Parallelpolarization

Crossedpolarization

Monoenergetic bunch comes from

colliding pulses: polarization test

Is it Easy to Build?

No significance

....)()()()( 30

200 dcba

Quadratic dispersion (glass etc.)

Spectral Phase

Cubic dispersion (gratings etc.)

Water radiolysis

D.A. Oulianov et al JAPS’ (2007).

• How to control injection?

-inject electron beam from LINAC (SLAC, Nature’07)

ANL LINAC Chuck Jonah, 198821 MeV; 7 ps; 4nC; plasma density 4-7x1010 cm-3

-use laser-based ionizationDWA : “chirped” bunches, break down due

to CCR multiphoton ionization • *control of laser PW, wavelength

• How to control acceleration?

-plasma density-channel guiding (LBNL)-colliding pulse (LOA)

2/1

22

0

0 )2ln4(1

GDD

0

0.005

0.01

0.015

0.02

0 5 10 15 20

SH

out

put (

arb.

uni

ts)

Generation number

Acousto-optic shaping

Dazzler - from Fastlite

No need for zero dispersion stretcher

Controls different dispersion orders

RGA pulse optimization test w/ SPIDER&Dazzler

-30

-20

-10

0

10

20

30

40

790 795 800 805 810 815 820 825

phas

e (r

ad)

w avelength (nm)

-10

0

10

20

30

40

50

60

800 808 816 824 832

11 41

21

311

Phas

e (ra

d)

wavelength (nm)

Generation #

0

0.005

0.01

0.015

0.02

0 5 10 15 20

SH

out

put (

arb.

uni

ts)

Generation number

0

0.2

0.4

0.6

0.8

1

-1000 -500 0 500 1000

11

21

41

1

Ampl

itude

, ar

b.un

its

t, fs

Generation #

Injection assisted by laser ionization

• Laser-assisted ionization of atoms or ions

• Two types: multiphoton and Frank-Keldysh tunneling

• 13.6 eV vs 1.5 eV

• DFG: Reducing laser frequency increases ponderomotive potential ~

• HE TOPAS ~100 J @ ~9 m

Laser techniques

• Multi-bunch generation w/ DWA• Pulse shaping• DFG due to detuned from 800

synchronized Regen pulses

• Atto-second science: CEP

Applications, Conclusions and Perspectives

W should be 7.2 GeV with laser parameters

(100TW, aL~3.8cm)

• THz source CCR

• Hard X-ray fs source

• X-ray free electron lasers

• Radiology, biophysics around water window

• Early stage of proton acceleration

• 1TeV is a goal for HE physics is too far

32 kJ of laser energy (100 lasers of 300J)

• Optical Parametric CPA

• Efficiency

• Emittance

• Charge

• Atto-second-ESASE

Thank you

Background. Parametric interaction

p = s + i

phase matching conditions in a uniaxial

x-stal such as BBO

kp = ks + ki

Non-collinear

Each photon in idler beam generated

together with a photon in signal

beam

S PI II P

PW

2/1

22

0

0 )2ln4(1

GDD

No significance

....)()()()( 30

200 dcba

Quadratic dispersion (glass etc.)

Spectral PhaseCubic dispersion

(gratings etc.)

FROGs• Frequency Resolved Optical Gating (Kane and Trebino’ Opt

Lett’ 1993)• Suitable for single-shot detection• Not an interferometric technique, just 2D spectrogram of

cross-correlation function• Not easy to reconstruct E(w,t): iterative algorithm, t-direction

ambiguity• Slight modification (Masalov et al, JOSA 2001) makes use of

spatial interferometry

wavelength

tDoubling x-stal

2nd harmonic

slit

Images

SPIDER

• Spectral phase interferometry for direct electric-field reconstruction (Iaconis and Walmsley, Opt Lett. ‘ 1998)

• Spectral interferogram of two frequency-shifted up-converted pulses; no reference needed

• Non-iterative reconstruction algorithm; 1D data set

2

2 ps

t

APE design

Pulse tilt

TFPA- pulse front inversion

KERI