No abstract

A high-gain, high-extraction-efficiency, linearly polarized free-electron laser amplifier has been operated at 34.6 GHz. At low signal levels, exponential gain of 13.4 dB/m has been measured. With a 30-kW input signal, saturation was observed with an 80-MW output and a 5% extraction efficiency. The results are in good agreement with linear models at small signal levels and nonlinear models at large signal levels.PACS numbers 42.60.By, 41.70.+t, 42.52. +xThe free-electron laser (FEL) is capable of producing coherent radiation from the ultraviolet to the microwave region of the electromagnetic spectrum. Several recent experiments have demonstrated lowgain, low-efficiency FEL operation in the visible 1 and infrared 2 regions while other experiments have demonstrated high-gain FEL operation in the millimeter-wave regime. 3,4 We have designed an experiment, the Electron Laser Facility (ELF), which can serve as a test of the physical models used to predict high-gain and high-efficiency FEL operation in the visible spectral region. The ELF consists of an amplifier with well-defined initial conditions on the radiation and the electron beam and with no axial magnetic field.

The injector of the Flash X-Ray (FXR) accelerator has a significantly larger than expected beam emittance. A computer modeling effort involving three different injector design codes was undertaken to characterize the FXR injector and determine the cause of the large emittance. There were some variations between the codes, but in general the simulations were consistent and pointed towards a much smaller normalized, rms emittance (36 cm-mr) than what was measured (193 cm-mr) at the exit of the injector using a pepperpot technique. The simulations also indicated that the present diode design was robust with respect to perturbations to the nominal design. Easily detected mechanical alignment/position errors and magnet errors did not lead to appreciable increase in the simulated emittance. The physics of electron emission was not modeled by any of the codes and could be the source of increased emittance. The nominal simulation assumed uniform Child-Langmuir Law emission from the velvet cathode and no shroud emission. Simulations that looked at extreme non-uniform cathode and shroud emission scenarios resulted in doubling of the emittance. An alternative approach was to question the pepperpot measurement. Simulations of the measurement showed that the pepperpot aperture foil could double the emittance with respect to the non-disturbed beam. This leads to a diplomatic explanation of the discrepancy between predicted and measured emittance where the fault is shared. The measured value is too high due to the effect of the diagnostic on the beam and the simulations are too low because of unaccounted cathode and/or shroud emission physics. Fortunately there is a relatively simple experiment that can resolve the emittance discrepancy. If the large measured emittance value is correct, the beam envelope is emittance dominated at modest values of focusing field and beam radius. Measurements of the beam envelope on an imaging foil at the exit of the injector would lead to an accurate value of the emittance. If the emittance was approximately half of the measured value, the beam envelope is slightly space charge dominated, but envelope measurements would set reasonable bounds on the emittance value. For an emittance much less than 100 cm-mr, the envelope measurements would be insensitive to emittance. The outcome of this envelope experiment determines if a redesigned diode is needed or if more sophisticated emittance measurements should be pursued.

This is a preprint of a paper intended for publication in a journal or proceedings. Since changes may be made before publication, this preprint is made available with the understanding that it will not be died or reproduced wtthoat the permission of the author.

No abstract