Two-Color I-SASE

Two-Color I-SASE
A. Marinelli, J. Wu, C. Pellegrini
LCLS2 Meeting
SLAC 1/30/2013
Diffraction imaging
• It has been suggested that a wide, 1-2% wide X-ray spectrum,
might be an advantage for coherent diffraction imaging.
• A possible alternative is a two colors spectrum with narrow lines
separated by 1-2%.
• A two color spectrum has recently been observed at LCLS while
doing the iSASE studies.
• A dedicated 2-color iSASE experiment is being planned
two colors 1/8/13
Science with 2-Color iSASE
Two proposals already submitted
Use of 2-color iSASE for SAD/MAD Phasing on XPP
Soichi Wakatsuki et al.
Request for separation from
0.1% to 2%
Science with 2-Color iSASE
Two proposals already submitted
Use of 2-color iSASE for SAD/MAD Phasing on XPP
Soichi Wakatsuki et al.
Request for separation from
0.1% to 2%
~within FEL
Proof of Principle Demonstration of
Delay line introduces
longer coherence legnth
Proof of principle demo: use detuned
undulators at LCLS
LCLS iSASE experimental setup
Machine layout:
First 5 undulator sections on-resonant
From 6th on, even number: 6, 8, …, 30, and 32 largely detuned (can either
be random or form a separate spectrum line  two color)
From 6th on, odd number: 7, 9, …, 31, and 33 on resonant
Perform proof-of-principle experiment on LCLS for an
improved SASE (iSASE)
Electron bunch: 150 pC, compressed to 3 kA
8.45 keV FEE HXSSS
13.825 GeV electron energy
two colors 1/8/13
Experimental Results
Observed a reduction of linewidth
by ~ 3.4
Two-Color iSASE
Alternating K value->
One wavelength gains, the other is delayed->
Two colors with narrower bandwidth than SASE
Observed During iSASE Experiment
Observed During iSASE Experiment
Line separation not
consistent with detuning!
1-D Theory
Effect of high-gain FEL on the
collective variables described
by 3x3 matrix
Two Color I-SASE
Define detuning d with respect to average resonant
frequency (w1+w2)/2
Two Color I-SASE
For large detuning:
Two Color I-SASE
For large detuning:
Equivalent to iSASE
transfer matrix…
Undulator dispersion
Undulator delay
Example: Delta = 6
Period = 2 Lg
Sidebands appear at the
frequency given by undulator
Relative intensity of the 4 peaks
strongly dependent on Delta
(weaker dependence for longer
Period = 4 Lg
Period = 8 Lg
Genesis Simulations
Genesis simulations for ideal beam show
same trends. The sideband separation is
consistent with our understanding.
Can we control this structure?
DeltaK = 1.1%
DeltaK = 0.6%
DeltaK = 0.9%
Power VS Time
Gain Curve
Comparison with Alberto’s Experiment
2-color experiment at LCLS worked beautifully…
What does our method do that hasn’t been done yet?
1) Perfect synchronization of two colors (good for imaging) BUT no
tunable delay (actually tunable within slippage length...)
1) Each color has the i-SASE bandwidth (tunability within the SASE
1) Control of sidebands allows extra degree of freedom on the spectral
Similar Power Level (a little higher
for the two undulator method)
A Few Words on LCLS2
-The greater range of tunability of LCLS2 allows even more flexibility in
tayloring the spectrum.
-Mimic ~10% bandwidth with more than two colors?
Open Questions
1) Are we over-complicating the
The chain matrix multiplication
method is direct but does not give
intuitive closed form formulae
Maybe some more intuitive picture is
2) In practice, how much control do we have over the
sidebands for short undulator periodicity?
Can we use them for fine-tuning of the spectrum within the
SASE bandwidth?
Open Questions
What happens with a realistic beam distribution?
Working on GENESIS simuations right now…
Experiment will tell us the answer!
Experimental Plan:
Study spectrum VS K separation looking for:
1) Wavelength separation VS Delta K
2) Sideband structure manipulation changing periodicity
3) XPP spectrometer for overall spectrum
FEE spectrometer for single-shot study of sidebands

similar documents