THE DRIVE LASER: EXPERIENCE AT SPARC

1
THE DRIVE LASER EXPERIENCE AT SPARC

• Carlo Vicario
• for
• SPARC collaboration

2
Summary

• SPARC laser system layout and performances
• Laser-to-gun optical transfer line grazing vs
  normal incidence
• Laser-to-RF synchronization measurements
• Longitudinal pulse shaping experience using
  DAZZLER
• Emissive properties of the photocathode

3
SPARC laser layout and systems performances
4
SPARC Laser beam requirements

Laser central wavelength 266.7 nm
Laser pulse lenght FWHM 2-12 ps
Electron charge 1 nC
RMS energy jitter (UV) lt 5 rms
Laser pulse rise time lt 1 ps
Laser pulse longitudinal ripples lt30 ptp
Transverse intensity profile Top hat
Laser spot radius 1.1 (mm)
RMS rf to laser time jitter lt 2ps
Centroid pointing stability 50 µm
Spot ellipticity on cathode (1-a/b) lt10
5
TiSa CPA laser system by Coherent pulse shaper
6
Coherent Laser System
7
Laser layout oscillator
TiSa CW oscillator (Mira) is pumped by 5 W
green laser (Verdi). The oscillator head can be
locked to and external master clock
(synchrolock).
pulse duration 130 fs
Central wavelength 800mn
bandwidth up to 12 nm
rep. rate 79.3 MHz
pulses energy 10 nJ
8
Laser layout time pulse shaper
To obtain the desired square profile a
manipulation of the spectral phase and/or
amplitude has to be applied. The most popular
techniques are the AODPF and the SLM in 4f
configuration. They work at low energy level.
New UV Dazzler S. Coudreu Opt. Lett. 31, (2006),
1899
9
Laser layout CPA
laser pump 1 1KHz, 7 W, 100 ns
laser pump2 10 Hz, 560 mJ, 7 ns
rep. rate 10 Hz
spatial mode Gaussian
output pulses energy, power lt 50 mJ, 0.5 TW
IR amplitude jitter 3
10
Laser layout THG
The third harmonic generator consists of by two
type-I BBO crystals, of 0.5 and 0.3 mm
thickness. The overall efficiency is about 8
and the energy jitter is 5 rms In the THG the
optics can be damaged by the IR high peak power
(self focusing effects).
11
Laser layout UV stretcher
The UV stretcher consists of a pair of parallel
gratings. It introduces a negative GVD
proportional to d, and allows output pulse length
between 2 and 20 ps. Efficiency of the grating
is about 65, the overall energy losses are more
than 80
12
Laser system layout spectral and time diagnostics

• Diagnostics routinely used to monitor
  time/spectral features of SPARC laser
• Ir blue commercial spectrometers resolution gt
  0.3 mn
• ps resolution streak camera
• UV home-built spectrometer with 0.05 nm
  resolution 10 mn bandwidth
• UV home-built multi-shot cross-correlator
  resolution (IR pulse FWFM)

13
UV spectral-temporal measurements
When a large linear chirp a is applied, as in our
case, the spectral profile brings to a direct
reconstruction of the intensity temporal profile
The UV spectrometer as single-shot time profile
diagnostics.
14
Optical transfer line I

• The optical transfer line transports the laser
  beam to the cathode 10 m away.
• The transverse profile is selected by an iris and
  then imaged on the cathode.
• The energy losses are mainly introduced by the
  grating used to compensate the grazing incidence
  distortions.
• Good pointing stability has been observed (50
  µm).

IRIS
laser
15
Laser grazing incidence
The laser beam is injected onto the cathode
surface at grazing incidence angle (72)
Beam exit
Photocathode

• Advantages
• No mirror close to the beam axis for normal
  incidence (no wakefield)
• Higher QE
• Disadvantages
• A circular beam becomes an ellipse on the cathode
  
• Time slew the side closer to the laser entry
  emits earlier than the other side

16
Compensation scheme
A grating with a proper g/mm can be employed to
diffract the beam at 72 and be positioned
parallel to the cathode. A lens is needed to
counterbalance the chromatic dispersion at the
image plane.

• Drawbacks
• High energy losses 65
• Sensitive to lens position (1 mm)
• Difficult to be measured
• Structures in the spot

• Advantages
• Circular beam at cathode (gt98)
• Front tilt compensation (lt 200 fs)
• Work for different spot sizes.

Simulated spot and front at cathode
C. Vicario et al, EPAC06
17
Normal incidence setup

• We change the TL normal incidence to get benefits
  in term of energy budget and spot uniformity.
• With this geometry the cathodes QE is half
  respect to the grazing incidence case.

18
Transverse profile at the virtual cathode

• Transverse spot features
• Sharp edges
• High spatial frequencies
• The beam transverse profile strongly influenced
  the e-beam brightness
• Refractive beam shaper spatial filtering is
  going to be implemented

19
System critical performances

• Reliability
• Laser failures (mainly electronics breaks) cover
  20 time
• Damages on optics especially in THG is not
  improbable
• Laser spot
• Flash lamp pump non-homogeneities worsen the
  TiSA mode
• Laser drifts due to the temperature
• The energy decay with time observed is due
  divergence changing of the flash pumped NdYAG.

20
Laser to RF phase noise measurements
21
Motivations
Laser phase stability is mandatory for stable
machine operation. For SPARC phase 1 is requires
lt 2ps rms, other application demands for more
challenging level of synchronization.
Coherent Synchrolock
22
Laser to RF phase-noise measurements
23
Phase noise at oscillator level
Statistics on the laser relative phase
Stdev0.35 deg
FFT of the relative phase
24
RF to Laser synchronization measurements on 10
Hz UV pulses
On time scale of few minutes the phase jitter is
within sRMS0.48 RF deg. Investigation of the
causes of the slow drift (temperature?) and
active RF phase shift compensation.
25
Longitudinal pulse shaping experience using
DAZZLER
26
Dazzler experience I experiment at Politecnico
in Milan
Input spectrum
The dazzler was studied as a stand-alone system.
The time profile was measured with a SH
cross-correlator. The shaped profile was imposed
by producing a square spectrum and add even terms
polynomial phase.
Phase applied
Amplitude filter
Two passed in the AO crystal
Single pass in the AO crystal 60 cm SF56
efficiency 0.5
efficiency 0.25
C. Vicario et al, EPAC04
27
Dazzler experience at SDL-BNL

• The experiment was in the framework of a
  INFN/LCLS/SDL-BNL collaboration.
• The motivations were
• Study the effects of CPA on the shaped pulse
• red shift, saturation effect, gain function of
  wavelength
• Study the effects of shaped pulse on the CPA
• Quantify the distortion introduced by the
  Harmonic conversion
• Eventually e-beam characterization

28
Dazzler experience at SDL-BNL
Reduction of the e-beam transverse emittance
could be observed due to this shaping of the
laser.
H. Loos et al, PAC05
29
DAZZLER experience at SPARC short amplified IR
pulse
The UV spectral shape as function of the input IR
pulse length
Measured (solid) and simulated (dashed)
harmonics spectra
IR pulse length ps
C. Vicario et al, Opt. Lett, 31,2006, 2885
A large enough pulse width (0.6 ps) is needed to
preserve the square spectrum throughout the third
harmonic generation
30
Equations for SH with vs linear chirp
Equation for the SH generation for the complex
fields Ai,j
In the frequency domain we can integrate A2 and
obtain the output intensity Hp Phase matching,
not depletion regime and negligible velocity
dispersion
The output spectrum is the convolution product
Similar consideration can be extended to the THG
31
Effects of non-linear crystal tilt
If the non-linear crystal is tilted by an angle ?
from the phase matching condition, the output
spectra are distorted
Simulated and measured SH spectra vs the tilt of
the crystal.
SH crystal tilt ? mrad
The crystal tilt act as a frequency shift and
therefore it introduces an asymmetry in the
output spectrum.
32
The UV temporal and spectral profile

• Using a chirped IR pulse (with 0.5 ps duration)
  and a square-like infrared spectral intensity we
  obtained a square-like UV shape.
• The measured UV rise time appears to be too long,
  2.5-3 ps.

33
Simulated UV intensity profile

• Ingredients to achieve this profile
• 1 Perfect square IR spectrum 12 nm
• Limitation form Dazzler resolution
• and amplifier distortions
• 2 Long IR input 10 (ps)
• Harmonics efficiency prop I(t)
• 3 140 um thick SHG crystal instead of 500um
• 40 um thick THG crystal instead of 300 um
• Harmonic efficiency prop. L2
• 4 Perfect alignment and time overlap

1 ps rise time
We can obtain more sharp edges clipping the
spectrum tails where it is spatially dispersed!
34
Modified UV stretcher to obtain sharper rise time
M. Danailov et al, FEL06
35
Preliminary measuremnts time and spectral
intensity
UV cross-correlation
UV spectrum converted in time (blue)
Calculated cross-correlation between the
measured IR pulse length and the UV (red)
36
Modified stretcher considerations

• The spectral measurements indicate rise time less
  than 1 ps can be obtained. New diagnostics is
  required to measure such feature directly in
  time.
• The energy losses due to the filtering is about
  20.
• The alignment is quite long and tedious.
• Distortions of the transverse profile and
  aberrations have been observed. Investigations
  are going on.

37
Cathode laser cleaning
Laser cleaning of the single crystal copper
cathode was operated moving the laser across the
surface step 100 µm . The optical energy was 10
µJ focused over 100 µm diameter, at 72 deg
incidence. The cleaning ware performed in
presence the moderate field 40 MV/m. Improvement
in term of beam brightness due to more
transversely uniform e-beam.
QE map before and after laser cleaning at low
field
Vacuum during the cleaning
38
Conclusive remarks

• SPARC laser performances are satisfying but the
  system requires constant maintenance
• Critical points flash-pumped NdYAG, high peak
  power
• Normal incidence is advisable in particular for
  large bandwidth lasers
• Synchronization level can be improved
• Uniform transverse laser intensity and constant
  QE is critical for e-beam quality
• Pulse shaping research is still facing the rise
  time problem. Balance between uniform transverse
  profile and flat top pulse in time and is still
  an open issue
• Cathode laser cleaning proved to be reliable
  technique