Comments by Rainer Pitthan on Optical Injection
As Jim said in his e-mail, I have recently been put in charge of examining the possibilities for an X-ray FEL for NLC. An XFEL beyond LCLS, something I am dubbing NCLS. That is were my immediate interest in the topic today comes from. But there might be spin-offs for LCLS.
One thing, which is clear, is that there is no interest in a NCLS, which will have about the same parameters as LCLS. So a preliminary goal for NCLS would be a SASE-XFEL made from a 50GeV beam (arbitrary choice) with a 0.1Åwavelength — 120keV (also arbitrary so far).
The problem to reach shorter wavelength with SASE within a finite length wiggler is the 6 dimensional phase space. Using the NLC-damping ring beams is believed to not work (horizontal emittance is large, the bunch is many mm long). But I don’t know about slice emittance? Using a RF gun, which works for LCLS (barely?)after much tricky bunch compression, would probably not work with a reasonable length wiggler at 50GeV and 0.1Å .
So I have looked for other injection schemes. Maybe we got lucky.
With the recent proliferation in availability of Multi (up to100's!) TeraWatt lasers there are several groups who have produced plasmas in dense gas jets (~1000 bar pressure) and in thin foils. Their goal was to achieve maximum acceleration (GeV) of electrons and injection into Plasma Wake Field Accelerators. The more mundane application of injection into a RF accelerator at lower energies may have been overlooked.
There are active groups in the US at Argonne, Berkeley, Naval Research Lab, and University of Michigan. There may be others. The pictures I will show are courtesy of Don Umstadter of University if Michigan. This group actually seemed to be the most focussed on injection.
There is more I don’t know and understand then I do, but here is what I know:
Basically they focus their TeraWatt lasers onto matter to a micron size spot, and out comes on the other side up to >1010 electrons with energies up to 100MeV range, within a 1deg cone with a pulse length determined by the laser.
In one particular choice of parameters the out coming electron bunch had a length of 30fsec (=10m ), was centered around 30 MeV, and had an invariant emittance of ge»1.2 10-7 m rad (this is calculated from the divergence of 10 measured at 30 MeV).
The physics behind these
effects is electrostatic acceleration
by charge separation through the laser pulse created plasma.
A rule of thumb is: if n is the number of available electrons/cc (~9*10**22 cm-3 in the conductance band of copper) then the plasma field produced is
sqrt(n) V/cm
3*10**7 V/mfor copper (one copper atom contributes one electron to the conductance band).
Hence the 30 MeV.
Naturally these are not off-the-shelve items (yet), but it looks very solid. I would say as solid as our SLAC dye laser 1987.
The topic for discussion: where do we go from here?