Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Schoessow 1999 0158

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

Proceedings of the 1999 Particle Accelerator Conference, New York, 1999

SLAB SYMMETRIC DIELECTRIC MICRON SCALE STRUCTURES FOR


HIGH GRADIENT ELECTRON ACCELERATION*

P. V. Schoessow,
Argonne National Laboratory, Argonne, IL 60439
J. B. Rosenzweig,
UCLA Department of Physics and Astronomy,
405 Hilgard Ave., Los Angeles, CA 90095

Abstract relatively low Q of these devices (100-1000) and


correspondingly short fill times (~0.5 ps) allows gradients
A class of planar microstructures is proposed which of 1 GeV/m to be obtained before breakdown becomes
provide high accelerating gradients when excited by an problematic[9]. The beam aspect ratio is highly asym-
infrared laser pulse. These structures consist of parallel metric; the ribbon beam-planar structure configuration is
dielectric slabs separated by a vacuum gap; the dielectric advantageous in that dipole deflecting modes are sup-
or the outer surface coating are spatially modulated at the pressed[6], analogous to the vanishing of the transverse
laser wavelength along the beam direction so as to deflection for TM0n modes in conventional structures.
support a standing wave accelerating field. We have
developed numerical and analytic models of the Laser propagation direction
accelerating mode fields in the structure. We show an
Mirror transmission y
optimized coupling scheme such that this mode is excited modulated in z
resonantly with a large quality factor. The status of
planned experiments on fabricating and measuring these y=0 b a x
z
planar structures will be described. Beam propagation and laser
polarization direction in z
Structure invariant in x
1 INTRODUCTION
Metal Dielectric

Advances in the technology of lasers have led to Figure 1. Geometry of slab-symmetric, laser-excited,
increased interest in their potential applications for dielectric-loaded resonant accelerator structure.
accelerating particles. Plasma beat wave[1] and laser
wakefield[2] acceleration methods use laser radiation to A number of different methods have been investigated
drive a plasma wave with luminal phase velocity. The to introduce the necessary periodic variation in the
longitudinal electric fields in this plasma wave are in turn structure geometry. The original paper in this area[5]
used to accelerate an electron beam. Inverse suggested the use of a longitudinally modulated
Cherenkov[3] and inverse free electron laser [4] permittivity in the dielectric. While this was attractive
acceleration techniques use light optics or magnetic fields from the point of view of construction and analysis (one
respectively to achieve an electric field component could concievably generate the modulation by “writing”
parallel to the beam direction from the transverse electric an interference pattern in a photorefractive dielectric with
field in the laser pulse. We have proposed a class of a second laser) it proved to be difficult to obtain good
resonant dielectric loaded planar structures[5] capable of coupling of the laser energy to the accelerating field.
producing GeV/m accelerating gradients which are driven More recent analyses have concentrated on using a
by laser radiation much as a conventional rf cavity is spatially unvarying dielectric medium and modulating the
driven by microwave power from a klystron. transmittivity of the coating [7]. This approach was found
to improve the laser coupling to the structure significantly
The basic idea is the use of a dielectric microstructure, but was found to yield uncomfortably large surface fields
analogous to a Fabry-Perot resonator, consisting of a two on the structure.
parallel dielectric planes separated by a vacuum gap and
with a partially transmissive coating on the exterior (see We discuss here some recent progress in the
Figure 1). Some parameters of the structure are assumed understanding of slab structure design. We have
to vary periodically in the beam direction at the investigated the use of finite thickness conductive
wavelength of the illuminating laser. A standing wave cladding on the structure, with a single coupling slot per
with an appropriate phase velocity forward component is period. With this geometry good coupling of laser energy
induced in this periodic structure by the laser pulse, which to the accelerating fields can be obtained, while at the
in turn accelerates the beam. The laser pulse is swept same time reducing the surface field/accelerating field
along the surface of the structure such that the cavity is ratio significantly over our previous results. The shape of
filled only in the neighborhood of the beam. The the coupling slot is also seen to be important, in that the

0-7803-5573-3/99/$10.00@1999 IEEE. 3624


Proceedings of the 1999 Particle Accelerator Conference, New York, 1999

introduction of a taper at the slot opening provides In reference[7], the fields of a similar structure having an
improved coupling over the case of a rectangular slot infinitesimally thick, sinusoidally modulated transmit-
profile. tivity outer cladding were computed. While reasonable
coupling was achieved, it was also pointed out that the
surface electric field on the mirror-dielectric interface was
2 NUMERICAL RESULTS undesirably large, roughly twice the maximum field in the
vacuum gap. The case of a finite thickness mirror with a
In the study of these structures we have found it useful to
rectangular aperture was studied for this paper. For a 2
rely on numerical solution of the Maxwell equations in
slab geometry. We use a custom finite difference time µm slot width the corresponding peak surface field is 2.0
domain code to inject and propagate the laser fields into Statvolt/cm for a gap field amplitude of 1.06 Statvolt/cm.
the structure geomerty under study. The program The largest surface fields in this device occurred at the
implements periodic boundary conditions in z (the beam exterior corners of the slot; this might be further improved
by rounding the corners. (The present version of the
propagation direction) and absorbing boundary conditions
simulation code cannot handle curved boundaries except
in y (the laser propagation direction) to handle any
via a stairstep approximation.) An increase in the
reflected laser energy from the structure. The structure is coupling slot width to 3 µm produced slightly worse
assumed to be of infinite extent in the x-view. For a given results with a surface field/gap field ratio of 1.9/0.87.
structure geometry, the integration is continued until a
steady state condition is reached. The structure is “tuned” The effect of tapering the aperture of the coupling slot
numerically by adjusting one of its parameters (typically is shown in Figure 2. The surface field/gap field ratio is
the dielectric constant) until the asymptotic stored energy relatively small, 1.56/1.23, while the gap field strength is
is maximized. (In the laboratory one would tune the also a maximum for all similar device geometries. Stored
structure by changing a geometrical parameter like the energy and gap field strengths vs time are shown in
vacuum gap size.) Figure 3. From the energy history we can read off a fill
time of 0.57 ps, corresponding to Q=102. As we have
We have analyzed a structure with dielectric thickness discussed previously[7], the fields in the vacuum gap are
(b-a) equal to the vacuum gap (a =1.6 µm), period of a superposition of Fabry-Perot-like (zero phase shift per
10.6 µm (corresponding to a commone CO 2 laser line), period, providing no net acceleration) and accelerating
and dielectric constant ε ≅ 3.7 . The conductive cladding (forward wave componet resonant with an ultrarelativistic
thickness in the simulation is 0.3 µm. The laser coupling particle) modes. The Fabry-Perot mode amplitude has
and field strengths in the structure were studied as a been diminished to 20% of the accelerating mode
function of the shape of the coupling slot. While not amplitude in the present case, solely through through the
exhaustive, these calculations indicate a promising periodicity of the coupling. This proportion seems to be
approach to the problem of coupling optimization. The about the most favorable we have been able to obtain.
following results are normalized to a peak laser electric
field of 0.25 Statvolt/cm = 75 kV/m.

Figure 3. Fill curves for the tapered aperture structure.


The fill time is 0.56 ps, corresponding to Q=102. The
Figure 2. Pseudocolor contour map of E z in the comput- stored energy is normalized to its maximum value, while
ational area for the tapered aperture structure. Laser radia- the accelerating and Fabry-Perot modes are shown
tion is incident from the top of the figure (y=0). The area relative to the accelerating field amplitude.
depicted in solid black is the conductive cladding, and the
thin black line indicates the dielectric-vacuum boundary.

3625
Proceedings of the 1999 Particle Accelerator Conference, New York, 1999

3 CONCLUSIONS

We have studied a class of dielectric loaded structures for


laser acceleration which can potentially produce GeV/m
gradients. We have given some consideration to questions
of structure breakdown and preservation of beam quality.
Resonant planar structures showing good coupling of
laser radiation to the desired accelerating mode have been
demonstrated numerically, with reasonable surface
field/vacuum field strengths and quality factors
commensurate with the requirements of a practical
accelerator. These properties imply that the structure is a
good candidate for further development as an accelerator,
as it can be coupled well (it can be fully impedance
matched upon filling, just as a standing wave linac
cavity). This development will probably proceed at 10.6
micron design wavelength. Choice of this wavelengthe is
based both on availability (e.g. at the UCLA Neptune
laboratory[9]), and on mitigation of the experimental
challenges one faces on moving orders of magnitude
down in accelerator wavelength.

4 REFERENCES

[1] C. Clayton et al., Phys. Rev. Lett. 70 37 (1993)


[2] P. Sprangle, E. Esarey, A. Ting and G. Joyce, Appl. Phys. Lett. 53,
2146 (1988).
[3] W.D. Kimura, et al., Phys. Rev. Lett.. 74, (1995).
[4] Y. Liu, et al., Phys. Rev. Lett. 80, 4418 (1998)
[5] J. Rosenzweig, A. Murokh, C. Pellegrini, Phys. Rev. Lett. 74 2467
(1995).
[6] A. Tremaine, J. Rosenzweig, P. Schoessow, Phys. Rev. E 56 7204
(1997).
[7] J. Rosenzweig, P. Schoessow, Proc. 1998 Advanced Accelerator
Concepts, (in press)
[8] D. Du et al., Appl. Phys. Lett. 64 3073 (1994)
[9] C. Clayton, et al., ., Nucl. Instr. Meth. A 410, 235 (1998), J.B.
Rosenzweig, et al., Nucl. Instr. Meth. A 410, 437 (1998).

_____________________
*
Work supported by US DoE Contracts W-31-109-ENG-38, DE-FG03-
92ER40693 and DE-FG03-98ER45693.
#
Email: rosenzweig@physics.ucla.edu

3626

You might also like