PIC Simulations of Plasma Beat-Wave Acceleration Experiments at UCLA R. Narang, C.E. Clayton, C.V. Filip, S.Ya. Tochitsky, D.F. Gordon*, C. Joshi, W.B. Mori University of California, Los Angeles, Los Angeles, CA 90095 *Naval Research Laboratory, Washington, D.C. 20375 Abstract. The plasma beat-wave accelerator (PBWA) in the Neptune Laboratory at UCLA utilizes a ~1 terawatt two-wavelength laser pulse to tunnel ionize hydrogen gas at conditions of resonance for driving relativistic plasma waves. This plasma wave is used as an accelerating structure for an externally injected ~11 MeV electron beam from the Neptune Photo-injector. Simulations in 2-D have been done to model this experiment for laser ionized plasmas with mobile ions for two focusing geometries, f/3 and f/18. Simulations have shown that ion motions in the transverse direction for small spot size cases (f/3 case) cannot be neglected, and that the acceleration of electrons is therefore limited by shortening of the effective interaction length due to deviations from the resonant density. In the f/18 case, while ion motions are not as severe as in the f/3 case, ionization induced refraction begins to limit the peak intensity of the laser. In addition, injection of the electron beam into the plasma wave is modeled to determine what acceleration is to be expected in experiments. INTRODUCTION Current efforts to model the PBWA in the Neptune Laboratory [1] utilize the PIC code turbo WAVE [2]. This code allows modeling of the experimental parameters present in the lab. Simulations are done for laser-plasma acceleration experiments in which the plasmas are produced by focusing a two wavelength (10.27 |um and 10.59 |iim), -120 ps laser pulse with f/3 or f/18 optics with intensities of ~1015 W/cm2 and ~2 x 1014 W/cm2, respectively [3]. In these experiments the separation of the laser frequencies is equal to the plasma frequency (A(jo=oop). From the simulations we are able to determine the normalized relativistic plasma wave amplitude, Ex, which allows the estimation of the field gradient present in the plasma. Determination of the wave amplitude allows the estimation of the magnitude of acceleration expected for electrons that are externally injected into the plasma wave. In general for 1-D theory the longitudinal field gradient is ~eno172 V/cm, where no=1016 cm~3 is the resonant density and 8 is the density perturbation. The maximum energy gain is given by: Wmax «E X L[4]. In an actual experiment other factors, such as finite interaction length, will limit the maximum energy gain of accelerated electrons. The propagation of relativistic electron beams inside of a relativistic plasma wave is also simulated for laboratory conditions. The Neptune Photo-injector [5] produces an electron bunch with an energy of 11.4 MeV with approximately 50 pC reaching the CP647, Advanced Accelerator Concepts: Tenth Workshop, edited by C. E. Clayton and P. Muggli © 2002 American Institute of Physics 0-7354-0102-0/02/$19.00 213 interaction point (IP). This beam is focused at the IP to a transverse spot size of Orms=125 |iim and has a bunch length corresponding to ox=1.8 mm (6 ps). Simulations are done with these conditions taken into consideration. A laser rise-time and fall-time of 275COP"1 (-50 ps) is used for all simulations and the frequency ratios are taken to be oo1/(jop=3 1 (10.59 um) and co2/cOp=32 (10.27 um). Table 1 gives the relevant parameters for the simulations presented in this paper, where each run corresponds to a total time of 90 ps (c/cop ~ 53 |um and ai,2 is normalized to ——^). The laser propagates in the x-direction with y as the eE transverse coordinate. The simulation window in the f/18 case is twice the length of the f/3 box, therefore the focus is reached on a later time step (the focus is placed at the center of each simulation window). The Rayleigh range, 2ZR, is -1.5 mm for f/3 focusing with a simulation box that is ~5 mm, while for the f/18 case 2ZR~2.5 cm. Table 1. Beat-wave Simulations for Neptune Parameters_________________________ Run x-cells y-cells Steps ai=a2 WQ(C/COP) dx(c/cop) dy(c/cop) 1024 f/3 1.0 0.1 0.1 0.05 128 10,000 0.3 f/18 4.0 0.05 0.1 0.025 4096 256 20,000 0.1 2-D SIMULATION RESULTS Images of Ex normalized to the cold wave-breaking amplitude, mccop/e, and the charge density, p (where negative and positive values on the color-table mean an excess of electrons or ions respectively), are shown in Figure 1 for the f/3 focusing geometry at t=50.4 ps. 80 100 c/up FIGURE 1. Simulation off/3 focusing, where (a) is snapshot of Ex and (b) is of p, both at t=50.4 ps. 214 After 50.4 ps the laser has blown-out plasma at the focus to less than 50% of the resonant density, the wave is disrupted, and the amplitude of the wave is reduced near focus. At the end of the f/3 simulation (t=90 ps) the focal region has nearly all of the plasma blown-out, with an excess of ions at the edge of the plasma. The peak wake amplitude achieved in the f/3 focusing is -0.1 mccop/e near focus, while it is -0.3 mccop/e away from focus. From this amplitude an average accelerating gradient of 2 GeV/m (calculated from the expression for the longitudinal field gradient) over a distance of approximately 3 mm can be expected, which implies a net acceleration of -6 MeV that could be achieved. Figure 2 illustrates that blow-out of plasma is not significant for f/18 focusing, where WQ is -200 um. In the f/18 simulation, where the spot size is now 4.0 c/cop, ion motions do not affect the plasma wave structure and the results are similar when immobile ions are used. The peak wave amplitude achieved is -0.1 mccop/e which corresponds to an accelerating gradient of 1 GeV/m (for a 1 cm interaction length we should expect -10 MeV acceleration). Each laser line has an amplitude of vosc/c=0.1 which corresponds to an intensity of -2 x 1014 W/cm2, and since the threshold for ionizing hydrogen is 1.47 x 1014 W/cm2 [6], the peak wave amplitude may be limited by the laser field amplitude. The position of the best focus appears to move -20 c/cop upstream as the laser beam ionizes the hydrogen, which is indicative of the onset of ionization induced refraction and this may also limit the peak wave amplitude. 50 100 150 200 50 100 c/up 150 200 20 10 FIGURE 2. Simulation results for f/18 focusing where (a) is an image EX and (b) is an image of p, both at t=70.2 ps. The plasma wave structures are quite different for these two simulations and can be attributed to the differences in the transverse fields. Figure 3 shows that the amplitude of Ey is 10 times larger for f/3 than for f/18 focusing. For the f/3 case, since Ey is 215 -0.3 mcoop/e, the blow-out of electrons can be attributed to the transverse ponderomotive force produced by this field. For the f/3 case the parameter kpR is ~1 (where R is the transverse dimension) while it is approximately 4 in the f/18 case which, according to Fedele et al.[7], implies that the ratio of Ey/Ex would be approximately 4 times larger for the f/3 case. This is consistent with the f/18 simulation, which gives the electron plasma wave structure for Ey that has been observed in previous 2-D simulations [8]. FIGURE 3. (a) is an image and lineout of Ey for the f/3 simulation at t=50.4 and (b) is for the f/18 run at t=70.2 ps. ACCELERATED ELECTRONS To get an estimate of the electron acceleration an electron beam is injected on axis into each of the accelerating structures with y=22.3 (11.4 MeV), neglecting the emittance of the beam. This injected electron bunch has a transverse spot size of oy=1.25 c/GOp and a length corresponding to ox=36 c/cop (6 ps). The transverse size of the beam is taken to be smaller than the experimental value due to limitations in the transverse size of the f/3 simulation box, and to limit the interaction region to the center of the plasma wave structure. The electrons are injected at the same time for each simulation (t^36 ps) such that the bunch interacts with the wave before and after it reaches its maximum amplitude. The maximum wave amplitude is reached at the center of the simulation window at t~50 ps for the f/3 case and t~70 ps for the f/18 run. In Figure 4 snap shots of the relativistic electron phase space are shown for the f/3 simulation (for y>10). At the time when the beam exits the simulation window y is approximately 34 (~17 MeV) for the most energetic electrons. This is in reasonable agreement with the estimate of ~6 MeV calculated from the normalized longitudinal wave amplitude obtained from the simulation, which corresponds to a field gradient of ~2 GeV/m. 216 (a) (a) (b) (b) c/ω c/ω c/co c/ωpp c/ωnpp FIGURE 4. 4. (a) (a) is is the the relativistic relativistic electron electron phase phase space space for for the f/3 simulation simulationproduced produced att=36.0 t=36.0 ps(b) (b) FIGURE FIGURE 4. (a) is the relativistic electron phase space for the f/3 f/3 produced at at t=36.0 ps ps (b) is at t=72.0 t=72.0 ps. ps. isisatatt=72.0 ps. InFigure Figure 555 the the electron electron energies energies produced produced for for the the f/18 f/18 simulation simulation are are shown shown (for (for In Figure the electron energies In f/18 simulation are shown (for γ>10). This simulation gives a maximum energy of approximately 18 MeV, which is y>10). This This simulation simulation gives gives aa maximum maximum energy of approximately 18 MeV, which γ>10). which is is in close agreement with the estimate obtained for a uniform wave that is 1 cm long. close agreement agreement with with the the estimate estimate obtained obtained for a uniform inin close uniform wave that is is 1 cm cm long. long. Although the the simulation simulation box box is is only only 111 cm cm in in length, length, for for aaa Rayleigh Rayleigh range range of of Although the simulation box is only cm in length, Although for Rayleigh range of 2Z ≈2.5 cm, the main contribution to energy gain is from the central 1 cm region with r 2Zr r≈2.5 ~2.5cm, cm, the the main main contribution contribution to to energy energy gain gain is is from from the 2Z the central central 11 cm cm region region with with eitheredge edgeof ofthe theplasma plasmacontributing contributing minimally minimally to to the the energy energy gain. gain. For Forthe thef/18 f/18run run either edge of the plasma contributing minimally to either the energy gain. For the f/18 run the final energy is not significantly different than for the f/3 case (see Figures 4b and the final final energy energy isis not not significantly significantly different different than than for for the f/3 case the the f/3 case (see (see Figures Figures 4b 4b and and 5b), which which may may not not be be the the case case for for larger larger aσyy since since the transverse transverse fieldswould would affect 5b), 5b), which may not be the case for larger σy since the the transverse fields fields would affect affect injected electrons. electrons. For For the the f/18 f/18 run run a longer longer interaction interaction length length would would produceaalarger larger injected injected electrons. For the f/18 run aa longer interaction length would produce produce a larger energy gain given that the wave amplitude is maintained for that distance. energygain gain given given that that the the wave wave amplitude amplitude is is maintained maintained for for that energy that distance. distance. (a) (a) (b) (b) 200 226 c/ωp c/ωp c/co c/ωp c/ωpn relativistic electron electron phase phase space space for for the the f/18 f/18 simulation simulation atatt=36.0 t=36.0psps(b) (b)isisatat FIGURE 5. (a) is the relativistic FIGURE t=81.0 ps. 5. (a) is the relativistic electron phase space for the f/18 simulation at t=36.0 ps (b) is at t=81.0ps. t=81.0 ps. CONCLUSIONS CONCLUSIONS CONCLUSIONS the PBWA PBWA in in the the Neptune Neptune Laboratory Laboratory atat UCLA UCLA has hasbeen beendone doneusing using Modeling of the Modeling of the PBWA in the Neptune Laboratory at UCLA has been done using PIC code code turboWAVE. turboWAVE. Both Both f/3 f/3 and and f/18 f/18 focusing focusinggeometries geometriesthat thatare areused usedininthe the the PIC the PIC code turboWAVE. Both f/3f/3 andcase f/18the focusing geometries that are used in the simulated. For For the the transverse fieldsblow-out blow-out theplasma plasma experiment are simulated. f/3 case the transverse fields the experiment are simulated. For the f/3 case the transverse fields blow-out the plasma 217 which has been shown to be one of the limiting factors in the production of large amplitude plasma waves. For f/18 focusing the intensity of the laser and ionization induced refraction of the laser beam may be the limiting factors in producing large amplitude waves for electron acceleration experiments. The relativistic plasma wave amplitudes are consistent with the phase space plots for accelerated electrons produced by injecting an electron bunch. For experimental parameters an accelerating gradient of ~2 GeV/m and ~1 GeV/m could be achieved for f/3 and f/18 configurations, respectively. Clearly the f/18 focusing geometry produces a more uniform accelerating structure for the PBWA than f/3 focusing. ACKNOWLEDGMENTS Simulations were done at NERSC on the IBM SP RS/6000. This work is supported by U.S. DOE Grant number DE-FG03-92ER40727. REFERENCES 1. Clayton, C.E., et al., Nuclear Instruments & Methods in Physics Research A, Amsterdam: Elsevier Science, 1998, pp. 378-387. 2. Gordon, D.F., et al., IEEE Transactions on Plasma Science, 28, No. 4, 1135-1143 (2000). 3. Tochitsky, S.Ya., et al., Optics Letters, 26, No. 11, 813-815 (2001). 4. Tajima, T. and Dawson, J.M., Physical Review Letters, 43, No. 4, 267-270 (1979). 5. Anderson, S.G., et al., Advanced Accelerator Concepts, AIP Conference Proceedings 569, New York: American Institute of Physics, 2000, pp. 487-499. 6. Augst, S., et al., Physical Review Letters, 63, No. 20, 2212-2215 (1989). 7. Fedele, R., et al., Physical Review A, 33, No. 6, 4412-4414 (1986). 8. Decker, C.D., et al., IEEE Transactions on Plasma Science, 24, No. 2, 379-392 (1996). 218
© Copyright 2025 Paperzz