VOLUME 82, NUMBER 8 PHYSICAL REVIEW LETTERS 22 FEBRUARY 1999 A Laser-Accelerator Injector Based on Laser Ionization and Ponderomotive Acceleration of Electrons C. I. Moore, A. Ting, S. J. McNaught,* J. Qiu, H. R. Burris, and P. Sprangle Beam Physics Branch, Plasma Physics Division, Naval Research Laboratory, Washington, D.C. 20375 (Received 15 October 1998) A new laser-accelerator injector based on the high-energy electrons ejected from a laser focus following the ionization of high-charge states of gases has been investigated. This method can generate electron pulses with MeV energies and pulse widths substantially less than the laser pulse width using terawatt lasers. Experiments at intensities of 3 3 10 18 Wycm 2 show that two highly directional electron beams with energies up to 340 keV are produced in the laser polarization direction. These highly directional beams are inconsistent with simulation results using Ammosov-Delone-Krainov (ADK) tunneling theory to model the ionization process. It is shown that this is likely due to the release of electrons with larger canonical momentum than predicted by ADK theory. [S0031-9007(99)08533-6] PACS numbers: 41.75.Lx, 32.80.Fb, 32.80.Rm Significant progress has been made on laser-based par- ticle acceleration [1] in the past decade: Background plasma electrons have been accelerated to approximately 100 MeV [2]; extended focused propagation distances have been generated [3]; acceleration in plasma [2,4]; vac- uum [5], and structures [6] has been observed; and clean sinusoidal accelerating plasma waves have been measured [7]. Despite these advances, many challenges remain. One of these challenges is a suitable electron injector. An ideal laser accelerator injector should satisfy the following requirements: (a) electron bunch lengths much less than the period of the accelerating field, (b) femtosecond tim- ing between the injector pulse and the accelerating struc- ture, (c) simplicity of operation and reasonable tolerance on alignment, and (d) a good emittance source. This in- jector will allow efficient coupling of the injector to the accelerator, and precise reproducible phasing and uniform acceleration of the electrons in the accelerating buckets. Recently the LILAC [8] and the colliding pulse laser injector [9] schemes have been proposed. These concepts are attractive since they are predicted to produce small emittance electron pulses a few femtoseconds long with excellent synchronization between the injector and accel- erator. However, they require extremely tight tolerances on alignment and are unsuitable for nonplasma-based accelerators. We have performed experimental and numerical stud- ies to investigate a new injector based on laser ioniza- tion and ponderomotive acceleration (LIPA) of electrons as an electron source. This scheme uses the elec- trons ejected from a high-intensity laser focus following ionization of high-charge states of gases [10,11] as an electron source. These electrons are ejected due to a com- bination of ponderomotive acceleration [12] and conser- vation of canonical momentum [13]. By appropriately choosing the gas, laser polarization, focusing geometry, and intensity, very short pulse electron bunches st , 5 fsd with high energy sE , 2.8 MeVd and small emittance s´ , 0.8 mm ? mradd are achievable with current laser systems. In the preliminary experiments described here, we have observed well-directed electron beams with ener- gies up to 340 keV. The experiment used a linearly polarized 2.5 TW, 400 fs laser pulse from the NRL T 3 laser sys- tem sl › 1.054 mmd focused to a peak intensity of 3 3 10 18 Wycm 2 in a vacuum chamber backfilled with 1 Torr of krypton gas. Ejected electrons were measured by wrapping direct exposure x-ray film (DEF) in a 3-cm-diameter cylinder around the laser focus with the laser axis concomitant with the axis of the cylinder of film. Electrons exposed the film allowing measurement of the ejected electrons’ angular distribution. One or three layers of 6 mm thick aluminum foil covered the film to prevent exposure from stray light and low-energy electrons. Thomson scattered light at 1054 nm was imaged through an interference filter onto a charge-coupled device at 90 ± to the laser axis to determine the effects of atomic self-focusing (ASF) and ionization induced defocusing (IID) [14]. At background pressures of 10 Torr and below, no evidence for ASF or IID was observed — diffraction limited propagation was observed. At higher pressures, both IID (at 15 Torr and above) and ASF (at 50 Torr and above) were evident. Figure 1 shows the observed electron distributions. Both images show the film “unrolled” with the horizon- tal axis as the azimuthal angle f around the laser axis and the vertical axis as the z position along the laser axis (the focus is at z › 0). The small marks at z › 23 mm and f › 90 ± are punch marks for locating the laser focus po- sition. The polarization direction was in the f › 90 ± and f › 270 ± directions. Figure 1(a) shows the higher en- ergy electron distribution, where three layers of aluminum foil completely blocked electrons below about 70 keV and significantly attenuated electrons up to approximately 140 keV [15]. For this image, nine laser shots were taken to expose the film with an optimal contrast. Figure 1(b) shows the lower energy electron component, where one 1688 0031-9007y 99 y 82(8) y 1688(4)$15.00 © 1999 The American Physical Society