Principle
of Electron Microbunch Formation
The
source of electrons in a linear accelerator (linac) is typically
either from a thermionic cathode, where the electrons “boil” off
a hot surface, or a laser-driven photocathode, where laser light
causes the electrons to emit off a non-heated surface. For a number of reasons, laser-driven photocathodes
are preferred in laser-driven electron acceleration (laser acceleration)
and this is the type of electron emitter used by the BNL Accelerator
Test Facility (ATF) linac.
In
the ATF linac, the electrons are emitted from the photocathode with
a typical pulse length of approximately 3 ps (1s). These electrons
are then accelerated using conventional microwave-driven cavities
to an energy of 45 MeV for the STELLA experiment. In the experiment, these electrons interact
with the ATF CO2
laser light, which has a pulse length of »180 ps, inside the inverse free electron laser (IFEL) undulators.
Since the 3-ps electron beam (e-beam) pulse length
is much shorter than the laser pulse length, the electrons experience
essentially a uniform electric field from the laser light.
The
electric field in the laser pulse consists of a train of oscillating
waves with a period equal to the laser wavelength, i.e., 10.6 mm. The 3-ps e-beam pulse corresponds to
900 mm
in length. Hence, there
are roughly 85 wave periods within the 3-ps e-beam pulse. This means some of the electrons will experience an accelerating
electric field, some a decelerating field, and some little field
along the length of the e-beam pulse.
Graphically, this can be represented by creating an energy-phase
plot, in which the energy shift of each electron from the mean e-beam
energy is plotted as a function of laser field phase within one
wavelength period. Phase
is equivalent to position along a laser wavelength where 2p phase equals 10.6 mm. Figure 1 shows the energy-phase plot for the
electrons initially entering the STELLA experiment from the ATF
linac. This plot is created by a computer simulation
of the experiment and contains 5,000 particles representing the
electrons distributed initially uniformly over all phases of the
laser light. (Recall the e-beam pulse is much longer
than the wavelength of the laser light.) Thus, except for an intrinsic energy spread or width about the mean
energy, all the electrons are near zero energy shift.

Figure
1. Initial energy distribution of electrons
entering STELLA experiment. |

Figure
2. Energy distribution of electrons after interacting
with laser light in first IFEL. |
Figure
2 shows what happens after the electrons interact with the laser
beam in the first IFEL. Within
each of the wavelength periods covered by the e-beam pulse,
there are accelerated and decelerated electrons.
(The distribution shown Fig. 2 repeats itself 85 times over
the length of the e-beam pulse.)
As
these electrons continue traveling downstream of the first IFEL,
the faster electrons catch up in phase or longitudinal position
with the slower ones as depicted in Fig. 3.
This process is called bunching because now the electrons
become grouped together into a microbunch as shown in Fig. 4, which
is a plot of the electron density distribution of the energy-phase
plot given in Fig. 3. (Note,
the abscissa of Fig. 4 is plotted in units of microns rather than
radians.)

Figure
3. Energy distribution after the electron energy
is modulated as shown in Fig. 2 and the electrons are allowed
to bunch together. |

Figure
4. Longitudinal density distribution of electrons
shown in Fig. 3. |
The
microbunch shown in Fig. 4 is approximately 1 mm long, which is equivalent to 3 fs in time duration.
Since Fig. 4 represents only one of 85 wavelengths within
the e-beam pulse envelope, this means a train of 85 1-mm-long
microbunches is formed separated by 10.6 mm.
A
shortcut for causing the electrons to bunch sooner after the e-beam
energy is modulated within the IFEL undulator is to replace the
drift space after the undulator with a magnetic device called a
chicane. The chicane used during the STELLA experiment (see Fig. 5) has both
a fixed magnetic field (i.e., permanent magnets) and a variable
magnetic field (i.e., electromagnets).
The fixed field forces the
electrons to travel through a “V-shaped” trajectory in which the
faster electrons generated by the IFEL traverse a shorter path than
the slower ones, thereby causing the electrons to bunch together.
This is trajectory is depicted in Fig. 6 for an electron
at the mean e-beam energy, i.e., 45 MeV.

Figure
5. Photograph of STELLA chicane.
|

Figure 6. Trajectory of electron at mean e-beam energy
traveling through chicane. |
Usage of the chicane makes the entire system more
compact with a total length of 1.2 m from the entrance of the
first IFEL to the exit of the second IFEL.
The second IFEL is used to trap and accelerate the microbunches. Control of the phase delay between the microbunches
and the laser field in the second IFEL is achieved by adjusting
the variable field of the chicane.
This enables resynchronizing the microbunches with the
accelerating phase of the laser field by slightly delaying when
the electrons exit the chicane.
|