Theory and Simulation of Plasma Accelerators - Particle Accelerator Conference, 1995., Proceedings of the 1995

We report on some of the recent theoretical and computational results at UCLA and USC on plasma-based accelerator concepts. Topics discussed include beat-wave excitation from short-pulse lasers, self-trapped electron acceleration from self-modulational instabilities and wakefield excitation in preformed channels. INTRODUCTION For some time now plasma structures have been considered as the basis for future accelerators. The plasma serves two purposes: (1) the plasma has no breakdown limits because it is already ionized and (2) the plasma supports large longitdinal waves. In these waves the electrons oscillate back and forth at 0,=(4~ce~n/m)”~ due to the space charge of the immobile ion background irrespective of the wavelength. Therefore by properly phasing these oscillations it is possible to create a wave with V$ C , i.e., a relativistic plasma wave. In such a wave electrons can acquire relativistic energies before they dephase from the wave. The accelerating gradient of a relativistic plasma wave is given by E& V l c m where n is the plasma density in cm-3 and E is the density perturbation of the wave. For a density of 10l6 cm-3 gradients of 10 GeV/m are possible in a plasma. Relativistic plasma waves can be generated by propagating either intense laser beams’ or intense particle’ beams through a plasma. During the past two years there have been several exciting experimental results on laser beam e x c i t a t i ~ n , ~ . ~ , ~ so in this brief report we limit the results to those related to laser-plasma accelerators. In 1979, Tajima and Dawson’ published the seminal paper on laser-plasma accelerators. They showed that relativistic plasma wave wake is excited behind a short pulse with a pulse length matched to half a plasma wavelength, L, = wdo,. This is now called the Laser Wakefield Accelerator (LWFA).6 In addition, Tajima and Dawson suggested two alternative ways to excite a plasma wave: 1) use two longer pulses with a frequency separation equal to w, to resonantly excite a wave or 2) rely on a long pulse to undergo the Raman forward scattering in~tability.~ The first is now called the Plasma Beat Wave Accelerator (PBWA) and the second is related to a scheme named the self-modulated LWFA.’ We will present some results related to each scheme. and less than an ion period. Laser technology now makes it possible to generate pulses which satisfy these constraints for plasma densities above 10l6 cm-3. Based on these considerations, we have investigated the feasibility of a 1 GeV PBWA experiment using existing technology.” The proposed parameters are given in Table I. Thie parameters were modeled Laser wavelengths Plasma Density Plasma Source Laser Pulselength Laser Power Laser Spot Size (20) Rayleigh length (ZR) Plasma Homogeneity Peak Plasma Wave Amplitude Peak Gradient Final Energy Multiphoton Ionization 200 pm 3.1 cm f ‘7% 0.5