Future Prospect for SR, Including 4GLS and Free Electron Lasers

1
Future Prospect for SR, Including 4GLS and Free
Electron Lasers

• Susan Smith
• Accelerator Physics
• ASTeC
• CCLRC Daresbury Laboratory

2
Content

• FELs
• What is a Free-Electron Laser?
• Basic components
• Oscillator FEL
• Going to shorter wavelengths
• Single pass SASE FEL
• Improving pulse quality
• Seeding 4GLS XUV FEL
• Self seeding 4GLS VUV-FEL
• Going to even shorter wavelengths
• HHG High Harmonic Generation
• HAFEL Harmonic Amplifier FEL
• HGHGHigh Gain Harmonic Generation
• ERL Based FELs
• Operational ERLs
• Challenges of future light sources
• ERL Principle
• Why ERL?
• 4GLS challenges

3
Free Electron Lasers
4
What is a Free-Electron Laser?
A beam of relativistic electrons co-propagating
with an optical field through a spatially
periodic magnetic field

• Undulator causes transverse electron oscillations
• Transverse e-velocity couples to E-component
  (transverse) of optical field giving energy
  transfer.
• Interaction between electron beam and optical
  field causes microbunching of electron beam on
  scale of radiation wavelength leading to coherent
  emission

Electron Decelerator
5
FEL Output

• Output is radiation with these properties
• Tunable
• High Peak Power
• Energy source is relativistic electron beam
• Electron KE radiation
• Peak brightness 6 orders of magnitude greater
  than spontaneous undulator source
• Coherent
• Emission from a bunched beam

6
FEL Principle

• relativistic electron beam passes through
• periodic magnetic field - radiates
• mirror feeds spontaneous emission back
• onto the beam
• spontaneous emission enhanced by stimulated
  emission

7
The Oscillator FEL
electron bunch
output pulse
optical pulse
8
Examples of Oscillator FELs
All these FELs use short undulators (low gain)
and rely on high reflectivity optics to reach
saturation
9
Going to Shorter Wavelengths

• If optics not available must use a SINGLE PASS
  through a LONG undulator
• Must also tune undulator to shorter wavelength
• Reduce undulator period
• Technological limits on this
• Reduce undulator parameter K
• Low coupling between electrons and radiation
  longer undulator required
• Increase beam energy
• more space and money for acceleration
• more rigid beam gives less FEL coupling longer
  undulator required

10
High Gain FEL
No Mirrors !
Need for high peak currents - short electron
bunches
11
The SASE FEL
optical pulse
12
Examples of SASE FELs
13
XFEL Exploitation
14
SASE photon output

• SASE output is amplified noise
• Electron beam is unbunched shot noise
• Spontaneous emission generated by electron beam
  in first few gain lengths is the seed which is
  amplified via an exponential process.
• Pulses noisy temporally and spectrally
• No pulse-to-pulse reproducibility
• Pulse length electron bunch length

SIMULATED XFEL OUTPUT
15
Seeding

• Dominate spontaneous noise at start of
  interaction with high quality injected seed
  pulse.
• FEL amplifies coherent seed, rather than noise
• Output pulse is
• smooth
• temporally coherent
• near Fourier transform limited enhanced spectral
  brightness
• reproducible shot-to-shot
• of similar length to seed pulse
• All these qualities offer tremendous advantages
  to users!

16
Example 4GLS XUV-FEL

• Current XUV-FEL Design will deliver
• 1.5 8 GW peak power
• Coherent pulses lt 50fs FWHM
• Variable polarisation
• Tunable 12.4nm 124nm

FWHM 35fs


17
HHG High Harmonic Generation

• Generate harmonics by firing seed laser into a
  gas jet
• Process of ionisation and accelerated
  recombination generates comb of coherent higher
  harmonic pulses
• Seed FEL directly with harmonic pulse

18
RAFEL Regenerative Amplifier FEL

• An alternative to using an external seed is to
  self seed.
• The RAFEL is a HIGH GAIN self seeding FEL
• Optics are required BUT
• gain is very high (10,000) so reflectivity can
  be very low
• radiation is not stored in cavity over many
  passes so tolerances on alignment and thermal
  loading are reduced
• cavity just provides a seed pulse for next pass
• Noise is averaged out giving smooth pulse
  significantly enhanced over SASE

19
Example 4GLS VUV FEL

• Shortest wavelength cavity FEL in world
• Will provide
• coherent pulses
• variable polarisation
• MHz repetition rates
• high average power
• Cavity length tuning allows (_at_124nm)
• pulse peak power from 250-1500 MW
• FWHM pulse width from 30-130 fs

upstream mirror
undulator
downstream mirror
20
Example 4GLS VUV-FEL
Pulse Profile_at_124 nm
Spectrum

21
HAFELThe Harmonic Amplifier FEL
Normal planar FEL amplifier including the 3rd
5th harmonics
22
HAFELThe Harmonic Amplifier FEL
Periodic disruption of fundamental FEL
interaction is possible using series of phase
shifters which delay electron bunch wrt FEL pulse

• If shift by one wavelength of third harmonic
• Fundamental interaction is disrupted
• 3rd harmonic interaction SEES NO CHANGE
• Result FEL lases at 3rd harmonic

Collaboration between ASTeC and Uni of
Strathclyde Accepted January 2006 for publication
by Physical Review Letters
Bessy-FEL Phase-shifter
23
HGHG High Gain Harmonic Generation

• Inject electron beam and seed at wavelength ?
• Modulator
• gives energy modulation in e-beam on scale ?
• Dispersive section
• converts energy modulation to spatial modulation
  bunching at ?
• This bunching has a component at ?/n
• Radiator radiates coherently at ?/n
• Repeat above steps in Harmonic Cascade
• Amplifier produces final amplified output at ?/m

Example BESSY LE-FEL proposal
24
Energy Recovery Linac Based FELs
25
ERL Principle
Schematic representation of an ERL. Injected
electron bunches (blue) are accelerated by the RF
field to an energy of 600 MeV. Having passed
round the electron transport loop, the returning
electrons (red) enter the linac 180? out of phase
and give up their energy to the SC cavity and are
decelerated to 10 MeV.
26
Operational ERL-FEL Sources
Three Oscillator FELs
BINP
JLAB
JAEA/JAERI

• Advantage of ERL vs Linacs
• Improvement in efficiency
• Enormous Increase in average current (CW) and FEL
  power
• Reduced dump activation

• Advantage of ERL vs Storage Ring
• Non-equilibrium conditions
• Source characteristic determined by injector
• Shorter bunches more flexible bunch pattern

27
JLab
First high current energy recovery experiment at
JLab FEL, 2000
JLab ERL-based Free Electron Laser
1 MW class electron beam, (100 MeV x 10mA),
comparable to beam power in CEBAF accelerator (1
GeV x 1mA), but supported only by klystrons
capable of accelerating 10-100 kW electron beam.
28
JLab ERL 10 kW IR-FEL Status

• 10 kW average power lasing achieved at 6 microns
• Lased 2 to 7 kilowatts at 1.0, 1.6, 2.8 microns
  with narrowband mirrors
• Lased at 20-100 W from 0.7 to 4.8 microns, tuning
  over the full band in seconds using hole
  outcoupler

• THz and Laser User experiments

Stable, reproducible operation at 115 MeV
29
BINP Status of 1st Stage FEL
The power and relative line width the terahertz
region are record parameters
29
30
JAEA ERL-FEL Status
31
Summary

• Exisiting ERL oscillator ERLs are excellent
  demonstrators of the ERL principle.
• CW Average currents of up to 10 mA (20 mA at
  high emittance)
• High repetition rates 75 MHz
• High efficiency gt 99.97
• Stable user operation
• High average photon power

Producing world leading sources of THz and IR
32
Proposed ERL Light Source Projects projects

• Oscillator FEL
• Kaeri Similar to JAERI FEL
• National High Magnetic Field Laboratory (Florida)
• PK-FEL 30-40 MeV, 1 mA (avg), 5 mm mrad (TESLA
  cavities in Stanford/Rossendorff module c.f.
  ERLP)
• JLab 100 kW IR-FEL
• High Gain FELs
• 4GLS (not in the energy recovery loop)
• BESSY II , X-FEL , LUX (all have mentioned ERL
  options)
• Repetition rates are currently generally low
  enough to make the complexity of recovery
  un-attractive
• Spontaneous Emission
• MARs
• Cornell 5 GeV X-Ray ERL
• KEK 5GeV ERL
• JAEA 6GeV ERL at Naka site
• APS
• ARC-EN-CIEL SACLAY (similar to 4GLS)
• 4GLS Daresbury ...

33
Why ERL Based Light Sources
34
Why ERLS?
35
Beam Size in a Linear Accelerator

• The beam properties are to a very large extend
  determined by the injector system
• The horizontal beam size can be made much
  smaller than in a ring
• While the smallest beams that are possible
  in rings have almost been reached,
• a linear accelerator can take advantage
  of any future improvement in the
• electron source or injector system.

ESRF 6GeV_at_200mA
ERL 5GeV_at_100mA
courtesy Ivan Bazarov
36
Smaller Beams and more Coherence

• Coherent x-ray diffraction imaging

3rd SR

• It would, in principle, allow atomic
  resolution imaging on non-crystalline
  materials.

• This type of experiments is completely limited
  by coherent flux.

ERL
Factor 100 more coherent flux for ERLfor same
x-rays, or provide coherence for harder x-rays
coherent
37
Bunch length in a Linac

• The bunch length can be made much smaller
  than in a ring
• While the shortest bunches possible in rings
  have almost bean reached, a
• linear accelerator can take advantage of
  any future improvement in the source
• source or injector system.

ESRF 5GeV_at_100mA
ERL 5GeV_at_100mA
100fs
2ps
16ps
38

• As compared to a ring, the beam properties are
  largely determined by the injector system
• The bunch length can be made much smaller
  than in a ring
• Smaller emittances
• Higher coherence fraction

ESRF 6GeV_at_200mA
ERL 5GeV_at_100mA
Current of 100mA and energy of 5GeV leads to a
beam power of 0.5GW !!! The energy of the spent
beam has to be recaptured for the new beam.
39
X-ray ERL e.g. Cornell
High flux 100 mA, 2 ps, 77 pC, norm emitt 0.3 mm
mrad 25m undulators, small gap, short period
undulators
High coherence mode 25 mA, 2 ps, 19 pC, norm
emitt 0.08 mm mrad
Ultra fast 1 mA, 50 fs, 1000 pC, norm emitt, 5
mm mrad
40
Challenges of future ERL light sources
Conceptual Layout of 4GLS
41
High average current loop
42
Challenges Generation of low emittance beams

• Photoinjectors
• 77pC, 1.3 GHz (100mA) has never been built before
• Even more demanding laser
• DC version has issues with power supply, high
  voltages
• 100 mA photocathode (short pulses)
• SCRF version has issues with photocathodes
• Other groups are active Cornell, BNL JLab..
• 1nC, 1kHz has never been built before
• Demanding laser
• Thermal problems from RF losses
• Other groups are active BESSY/DESY LBNL .

43
Challenges SC Accelerators

• High input powers (10MeV, 100mA 1MW)
• Three distinct beams in main linac
• Complex pulse trains
• Beam loading
• Need to minimise HOMs extract power at correct
  temperature Beam Break Up
• Phase and amplitude control 0.01?, 0.01 - state
  of the art
• Large scale cryogenics

44
Challenges Electron Beam Transport

• Preservation of small emittance
• Generation of ltltps bunches at the correct
  locations, longitudinal gymnastics
• Minimisation of instabilities (CSR, wakes, )
  long bunches!
• Merge and separation of different beams
• Minimise losses
• Collimation
• 60 MW beam power ILC is 11 MW !
• 1MW Dump
• Diagnostics
• Tuning

45
Challenges FEL design

• Single pass seeded amplifier
• eXtreme Ultra Violet Seed laser (state of the
  art HHG system)
• Undulator tolerances very demanding

1
Low Q cavity (Vacuum Ultra Violet) Mirrors
withstand high peak powers
2
3
High Q cavity (Infra Red)
46
Challenges Combining Sources

• Timing Synchronisation
• All combined sources to have synchronisation
  better than 100fs
• Particular combinations require 10fs
• Many sources of jitter
• Laser
• RF signals
• RF acceleration
• Electron transport
• Photon transport

47
www.4gls.ac.uk
48
Cornell ERL Injector Prototype Project
Gun
Dump with quadrupole
optic
Buncher
Injector
Main linac
High brightness _at_ High average current
49
DC Photoinjector
Simulations normalized r.m.s. emittances lt 0.1
mm-mrad ,77 pc/bunch, I 100 mA.
(Bazarov et al., Phys. Rev. ST-AB 8, 034202
(2005) 
Yb fiber laser 100 nJ/micro pulse 750 KeV 100 mA
supply in Autumn NEA GaAs and GaAsP cathodes gt17
QE
The photocathode load lock and preparation
system, with translation mechanisms
50
Two cell SRF cavity
gt120 kW, e2v klystrons, efficiency gt50, RF
testing soon Two ports, coupler power 50 kW, two
tested soon 1st successful vertical test of 2
cell cavity Fabrication of five cavities (tests
by next year) Horizontal test of complete
assemblies early next year
75 kW beam dump constructed 575 kW required for
full injector test Beam diagnostics challenging
design are in progress.
51
KEK-JAEA Test Facility
Preliminary plan of ERL test facility at KEK
Tentative parameters of the ERL test facility
Close collaboration with Superconducting Test
Facility (STF) team at KEK.
52
Development of a DC gun
high-voltage test was completed (245kV).
high-voltage terminal
2m
Photocathode Test Bench (Cathode holder)
SF6 gas tank
ceramic tube
under designing
electron bunch
under designing
load-locked cathode preparation
drive laser
pump
53
Energy Recovery Linac Prototype (ERLP)
35 MeV
54
Photoinjector Laser

• Wavelength 1.05?m, multiplied to 0.53?m/0.26?m
  (NdYvanadate)
• Pulse energy 80nJ on target
• Pulse duration 10ps FWHM
• Pulse repetition rate 81 MHz
• Macropulse duration 20 ms
• Duty cycle 0.2
• Timing jitter lt1ps
• Spatial profile circular (top hat) on
  photocathode

Laser system commissioned 2005.
55
ERLP Gun
electrons
JLab GA
Gun ceramic major source of delay at
Daresbury (1 year late)
transverse emittance 3 mm mrad
56
Superconducting Booster Module

• 2 x Stanford/Rossendorf cryomodules 1 Booster
  and 1 Main LINAC.
• Booster module
• 4 MV/m gradient
• 32 kW RF power
• Main LINAC module
• 14 MV/m gradient
• 16 kW RF power

Delivery April/July 2006 (7 months late)
JLab HOM coupler design adopted for the LINAC
module
PKU-FEL are using the same modules for their
IR-FEL project
57
Status
Laser system ready RF powers upply under test on
site Gun ceramic major source of delay at
Daresbury (1 year late) Accelerator modules
arrive April/July 2006 (6 months late) 4 K
commissioning May 06 Gun commissioning
August-October 06 2K commissioning November
06 Complete machine ready December 06 Energy
Recovery Spring 07 Exploitation 2007
Coming soon ERLP A World Class ERL based
Facility for the Development of
Accelerator/Photon Science and Technology...
58
ERLP photon science NWSF 2.9M, 3 years
X-rays Time resolved X-ray diffraction studies
probing shock compression of matter on sub
picosecond timescales.
90º focus mirror
X-rays
Probe
Pump
IR
180º focus mirror
CBSInteractionPoint
THz
THz Ultrahigh intensity, broadband THz radiation
to be utilised for the study of live tissues.
Dedicated holethrough concreteshield wall
25TWLaser
Laser-SR synergy Pump-probe expts with
table-top laser and SR
59
4GLS Status Milestones

• Science case Dec 2001
• April 2003 Prototype design/build and CDR funded
• ERLP Commissioning (2006-2007)
• Exploitation of ERLP (2007)
• Accelerator science
• Photon science THz X-rays (CBS)
• CDR April 2006 www.4gls.ac.uk (baseline
  design)
• TDR 2007-8
• Costing
• High priority technical work
• Prototyping (SC RF, Photoinjector) 3-5 years
• Bid for funding late 2007
• Construction 2008 20012/13

60
RF Deflection Scheme
61
RF Deflection Scheme
62
Alpha-X
GeV/ cm !!!
63
The End!