THPCH010

Download as pdf or txt
Download as pdf or txt
You are on page 1of 3

ELECTRON BEAM-LASER INTERACTION NEAR THE CATHODE IN A

HIGH BRIGHTNESS PHOTOINJECTOR


M. Ferrario, G. Gatti, INFN-LNF, Frascati, Italy; L. Serafini, INFN-Mi, Milano, Italy;
J. B. Rosenzweig, UCLA, Los Angeles, USA
Abstract This effect, while physically interesting, is negligible in
The production of high charge short bunches in a high this case. A more interesting effect is produced on the
brightness photoinjector requires laser pulses driving the longitudinal phase space, as will be discussed in the next
cathode with GW range peak power on a mm spot size. sections.
The resulting transverse electric field experienced by the
electron beam near the cathode is of the order of 200-500
HOMDYN SIMULATIONS
MV/m, well in excess of a typical RF accelerating field of The fast running code HOMDYN [2] has been
50-100 MV/m. We present here a preliminary study of the modified in order to include the beam interaction with the
resultant beam dynamics. Simulations including the laser field, here modelled as a simple truncated plane
electron beam-laser interaction have been performed with wave, in addition to the RF and space charge fields
the code HOMDYN taking into account the superposition already implemented in the code, including the important
of incident and reflected laser pulses as well as space effect of the cathode image charges.
charge fields. Under this conditions the emittance
degradation is negligible, as predicted by analytical
methods, but a longitudinal charge modulation occurs on
the scale of the laser wavelength, in case of oblique
incidence, driven by the longitudinal component of the
laser field. Charge modulation is transformed into energy
modulation via the space charge field, which may
produce enhanced microbunching effects when the beam
is further compressed in a magnetic chicane.

INTRODUCTION
Electron beam – laser interaction near the cathode
surface could be important in case of short and strong
laser pulses, when the involved field becomes high and
the electrons have a very low momentum (emission is Figure 1: P-polarized plane wave incident on a lossy
achieved slightly overcoming the cathode work-function medium [3].
to reduce thermal emittance). We consider in this study
the case of a metal photocathode. Other kinds of The expression of the electromagnetic field of a plane
photocathodes, such as high Quantum Efficiency (QE) wave incident on a lossy medium for a p-polarized wave
Cs2Te have no need to be driven at high power, in is the following, see also Fig. 1:
addition their response is slower (some ps), and the Bi =  yˆ ( E i / c)cos(k i .r  t +  i )
considerations that will be made in this framework
require some slight adjustments to fit the alternative case E i = E i [ xˆ cos + ˆz sin  ] cos(ki .r  t +  i )
of semiconductor cathodes. A large laser field at the Br =  yˆ R( E i / c)cos(k r .r  t +  R )
cathode can produce “heating” of the electron beam
through the induced wiggling motion. The degree to E r = E i R[ xˆ cos  + ˆz sin  ] cos(k r .r  t +  R ) .
which the laser field causes emittance growth can be
estimated as [1]: Assuming a cathode reflection coefficient constant
within the laser wavelenght spectrum. The previous
l expressions can be easily extended to the general case of
 n = al  x  Z 0 Pl a real laser pulse, without operating integration over the
2m e c 2
spectrum. It is sufficient to substitute the incoming
Here al is the peak normalized vector potential of the laser envelope expression into the amplitude of the incoming
field, Pl is the peak laser power and l the laser field Ei.
wavelength (note the independence of the beam size x ). Dealing with lossy media (i.e. radiation must be
For example a laser pulse 300 fs long with an energy of absorbed to generate photo-electrons) the p-polarization is
0.2 mJ as required by a blow out experiment [1], able to reflected again as p, but with a phase shift due to the
extract 0.33 nC charge from a 10-5 quantum efficiency complex index of refraction n. Photoemitted electrons
cathode, results in a 0.04 μm induced emittance growth. experience a field that is sum of incoming and reflected
wave. Inside photocathode the field amplitude is the plot, after a time of 1 ps the emittance growth due to
attenuated exponentially in direction normal to the space charge prevails.
surface, and propagates at a slightly different angle, this
effect will be include in a future work.
Both normal and oblique laser incidence (at 70 degrees)
have been taken in consideration in HOMDYN simula-
tions. A laser field amplitude of 190 MV/m is assumed,
the RF field experienced by the electrons injected in the
gun with 26 phase degrees is 53 MV/m, (120 MV/m peak
field). Laser wavelength was fixed at 266 nm, able to
drive Cu and other metal photo-cathodes.
The first case considered here is a 10 ps long laser
pulse at normal incidence and a 1 nC e-bunch charge,
corresponding to the nominal SPARC parameters [4]. An
integration time step of 0.1 fs (corresponding approxi-
mately to 10% of the laser central wavelength period) and
1000 slices 40 nm long (about /7) were used. As was
expected, each slice experiences the fast varying field of
the laser, as a cosequence slice centroids undergoes a fast
oscillation in space. The amplitude of the oscillations
grows with the time, and so this effect is enhanced in the Figure 3: Electron trajectories for different injection
first emitted slices compared to the last ones as shown in phases in a laser field with grazing incidence. (Trans-
Fig. 2. verse displacement [m] versus time [s]).

Figure 4: Normalized emittance vs. time for the 1 ps laser,


1 pC case. Blue line: space charge induced emittance Red
line: total emittance including laser induced emittance.
Figure 2: Slice centroids evolution during the interaction
On the other hand since laser contribution is now
with the laser beam. Each dot represents a slice centroid
mainly longitudinal, we expect a some effect on the
located at a distance dz from the bunch central slice.
longitudinal motion. Preliminary simulations show that
This effect occurs on a negligible scale , so that, even if the induced energy spread is negligible, a charge
looking at the macroscopic consequences on the beam, modulation on a scale of the laser wavelength occurs.
neither envelope, nor emittance or energy spread are Different groups of slices emitted within different half
affected in such a way to show a noticeable degradation periods of the laser field will experience in fact
of the beam quality. Emittance on a time scale of 10 ps accelerating or decelerating laser fields components,
reaches a value of 0.05 μm. while the electron velocity changes from a non-relativistic
The case of grazing incidence is more interesting. First to a relativistic regime, due to the superimposed RF
of all, the angle introduces an asymmetry on electron accelerating field. The initial extraction velocity
trajectories, depending on different field phase which =v/c=0.002 in this case is enough to prevent backward
particles experience at different emission times, Fig. 3. acceleration. This periodic shrinking and broadening of
Once again, the transverse dynamics is affected by an the slices, that last until the end of the electron-laser
even weaker effect, than in normal incidence interaction, results in a slice current modulation, as
The computed emittance behaviour for a 1 ps laser shown if Figures 5 and 6. Notice that the current
pulse, 1 pC charge is shown in Fig. 4. As can be seen in modulation can reach a 40% depth. Downstream the
cathode the beam evolution is dominated by space charge because it could produce an enhanced micro-bunching on
effects and will transform this current modulation in the scale of the photo-injector laser wavelength [5].
energy modulation [5].
CONCLUSIONS
Though laser field doesn't affect in a relevant way the
electron beam emittance and energy spread, in the case of
grazing incidence it generates a current modulation inside
the bunch. Current modulation is trasformed into energy
modulation via the space charge field, which may
produce enhanced micro-bunching when the beam is
further compressed in a magnetic chicane. On the other
hand, as a concluding and wishful remark, one can take
profit of such an energy modulation as a possible
alternative to laser heating [6]. By means of a high power
longer wavelength (I.R.) laser conditioned photo-
emission (i.e. illuminating the cathode with an IR laser
Figure 5: Normalised beam current modulation at the end light in addition to the required UV light), one may think
of the photo-emission process. to pre-heating the electron beam by enhancing the current
modulation at the cathode. The space charge induced
energy spread is then tranformed in an uncorrelated
energy spread in a downstream chicane with negative R56
[5]. A more detailed study on this subject is under way to
asses the amplitude of the achievable energy spread.

REFERENCES
[1] J.B. Rosenzweig et al., “Experimental testing of
dynamically optimized photoelectron beams”, Proc.
of the ICFA workshop “The Physics and applications
of high brightness electron beams”, Erice, 2005.
[2] M. Ferrario et al., “HOMDYN study for the LCLS
photoinjector”, SLAC-PUB-8400 (2000).
Figure 6: Normalised beam current modulation at the end [3] T. Plettner, SLAC-PUB-11800 (2006).
of the photo-emission process, blow up of Fig. 5. Notice [4] D. Alesini et al., “Status of the SPARC project”,
the deep current modulation (40%) on the laser these proceedings.
wavelength scale. [5] M. Boscolo et al., “Laser comb: simulations of pre-
modulated e- beams at the photocathode of a high
This effect could be important when a magnetic brightness rf photoinjector”, these proceedings.
compressor is installed in the downstream beam line [6] Z. Huang et al., Phys. Rev. STAB 7, 074401 (2004).

You might also like