Transcript Slides - Indico

Introduction to Plasma
Physics and Plasma-based
Acceleration
Wakefield acceleration
Various images provided by R. Bingham
Particle acceleration
– Accelerate particles in an electric field
– Hard to sustain constant electric field
over long distance: use wave that
travels with particle at (almost) the
same speed, just below c
Plasma-based acceleration
– Maximum field amplitude in vacuum
accelerator: 10-100 MV/m
– Maximum field amplitude in plasma
accelerator: 10-100 GV/m
– Scale back ultrahigh energy
accelerators from kilometre to
metre size !?
Conventional vs. plasma
42 -> 85 GeV in 1 meter of plasma!
e-
Ionizing
Laser Pulse Li Plasma
(193 nm) ne- 6 · 10 15 cm -3
L- 30 cm
Streak Camera
(1ps resolution) •Cdt
N=1-2 · 10 10
z =0.1 mm
Optical Transition Spectrometer Cerenkov
E=30 GeV
Radiator
Radiators
25 m
Not to scale!
X-Ray
Diagnostic
Dump
Wakefield
Acceleration
Experiments at the
Stanford Linear
Accelerator Centre
How to drive a plasma wave
– Send driver (powerful laser pulse or highenergy particle beam) into plasma
– Driver will blast plasma electrons out of the
way, leading to region with net positive charge
– Behind driver, electrons will rush back and
“overshoot”, leading to region with net
negative charge
– This process is repeated a few times, leading
to a plasma wave that travels at the speed of
the driver
e-beam driven wakefield
Laser-driven wakefield is similar
Types of plasma accelerator
– Beam-driven plasma accelerators
– Electron-electron accelerators
– Electron-positron or positron-positron
accelerators
– Laser-driven plasma accelerators
– Laser beat-wave accelerator
– Self-modulated laser wakefield
accelerator
– “Forced” laser wakefield accelerator
History
1979: T. Tajima and J. Dawson, “Laser Electron Accelerator”, Phys. Rev.
Lett. 43, 267
1985: D. Strickland and G. Mourou develop “chirped pulse amplification”
to generate intense laser pulses
1993: First observations of laser-accelerated electrons at UCLA (led by C.
Joshi)
1991-2002: Discovery of “bubble”-shaped wakefield solutions in
theoretical models
2001-2003: First observations of particle acceleration in beam-driven
wakefields (UCLA-SLAC collaboration)
2004: First observations of laser-accelerated electron bunches with small
energy spread: Dream Beam on the cover of Nature (RAL, LBNL, LOA)
2005: Beam-driven acceleration breaches 1 GeV barrier (UCLA-SLAC)
2006: Laser-driven acceleration breaches 1 GeV barrier (LBNL-Oxford)
2007: Energy doubling in beam-driven acceleration: 42 GeV -> 85 GeV
(UCLA-SLAC)
Laser electron accelerator
– Closely tied to development of high-power
short-pulse lasers
– Early schemes (when laser pulses were “long”):
– beat-wave accelerator (plasma wave driven by beating
of the two frequencies of a CO2-laser)
– self-modulated laser wakefield accelerator (plasma
wave “modulates” long pulse to form train of short
pulses, which then drives plasma wave resonantly)
– Today’s ultra-powerful (50-500 TW), ultra-short
(<50 fs) laser pulses can drive a plasma wave in
one go
Two stages
– Electron injection
– External injection
– All-optical (internal) injection
– Electron acceleration
– Laser pulse needs guiding
– Electron bunch may diverge or perform
transverse oscillations
External injection
– Use photo-cathode or other method
to produce short electron bunch
– Inject this bunch into the plasma
wave
– Timing issues, bunch length issues
– No longer a favoured method
All-optical injection
Used in (almost) all experiments that produce
high-energy low-spread electron bunches
Acceleration
Create sufficiently long plasma column
– Laser usually ionises background gas itself
Keep laser pulse under control
– Laser pulse diverges, needs to be self-guided or
externally guided
– Laser pulse depletes, limiting acceleration length
– Laser pulse may “jump around” from shot to shot
Keep electron bunch under control
– Try to minimise energy spread upon injection
– Acceleration stops when bunch outruns wave: dephasing
– Electron bunch may oscillate transversely: betatron
motion
– Electron bunch may diverge
Main limits to energy gain
Plasma channel: structure for guiding laser and supp
• Diffraction:
Optical diffraction
Ldif  LR   w / 
2
2
0
diff 
Channel

w0
2
z R   w0 / 

order mm!
(but overcome w/ channels or relativistic self-focusing)
n  1
Longitudinal Electric
laser pulse
Goal:
• Dephasing:
c
Vgr
Guide 1018 W/cm2 pulses over many
diffraction lengths
p 2
Ldph 
Approach:
1 Vgr c
order 10 cm
x 1016/no
0
10
20
3
(
Preformed channels production through hydrodynamic shockwave in plas
Dual pulse Ignitor-heater scheme
•Depletion:
For small a0

For a0 >~ 1
>> Ldph
Ldph~ Ldepl
W ch [MeV ] ~ 60 p /w 0  P[TW ]
2
Laser pulse guiding
Self-guiding:
– Electrons oscillate rapidly in laser field, mass
increases, plasma frequency decreases
– Change in refractive index causes laser pulse to
focus
External guiding:
– Laser pulse passes through plasma channel
with depressed on-axis density
– This channel acts like a glass fiber and confines
the laser pulse
Bubble regime
– Ultra-intense driver (laser pulse, particle beam)
blasts almost all plasma electrons aside
– A single “bubble” is formed behind the driver,
containing only ions, with good properties for
acceleration
Images: L.O. Silva, IST Lisbon
Bubble regime
Solves some of our problems:
– Plasma electrons self-injected
– Accelerating field inside bubble mostly
constant -> small energy spread
– Bubble provides focusing field that coincides
with accelerating field -> small transverse
spread
– Bubble is a stable structure: acceleration over
large distances possible
(Virtually) All groundbreaking recent results have
been obtained via such bubbles
Current status: laser wakefield
1 GeV mean energy
with 5% spread
(LBNL-Oxford
collaboration).
W. Leemans et al.,
Nature Physics 2,
696 (2006).
Current status: beam-driven
wakefield
Energy doubling of 42
GeV electrons to reach
85 GeV in 1 meter of
plasma
I. Blumenfeld et al.,
Nature 445, 741
(2007).
But still need to get 42
GeV to start things up.
Near future
Using the future 10 PW upgrade to the
Vulcan laser at RAL, the following results
may be within reach:
– 40nC at 1GeV (large spread)
– 14nC at 4 GeV (medium spread)
– 2nC at 10GeV (small spread)
Energy-wise, laser-wakefield accelerators
are starting to compete
Remaining issues
Energy spread: current best is 1-2 %,
needed is 0.1% or less
Mean energy: always need more, so we
need a bigger laser
Guiding: external guiding yields better
results, self-guiding much easier
Reproducibility: need a more stable laser
that still fires at a high repetition rate and
at high power…
Summary
Introduction to plasma-based acceleration,
wakefield generation, strengths of this
scheme and remaining issues
Much progress has been made, and much
more to be expected in this dynamic
field
Many issues still to be resolved
Much more ground to cover, which will
happen in the coming weeks