Optical Characterization Of Semiconductors With A Free Electron Loser

Abstract

There are few laser sources capable of producing tunable radiation in the mid to far infrared, and none of them is a high-average power or high-peak power source. Free electron lasers (FELs) using a linear accelerator as a source of relativistic electrons do produce a beam with the following general characteristics: tunability from I 1 pm to 2 10 pm, high average power (>1 W), high peak power (2100 kW), and short pulses (-5 ps). There are three such facilities in operation in the United States (a fourth FEL, based on a somewhat different principle, generates cw radiation at wavelengths longer than 50 pm). In this presentation, I report the results of experiments performed on semiconductors which exploit the special features of the EL. The prospects for the use of FELs for routine optical diagnostics of semiconductors are discussed in the light of the experience we have accumulated over the past 5 years. We have performed picosecond pump and probe experiments in thin films of amorphous hydrogenated silicon-germanium alloys. The FEL was tuned around 1.5 pm. The pump beam was obtained by frequency doubling the FEL output and the probe beam was directly obtained from the FEL output. Our results [ 11 show that the photoinjected electrons and holes recombine in less than 10 ps, a result in agreement with earlier results we had obtained on amorphous hydrogenated silicon using conventional short pulse lasers [2]. We are also performing two types of experiments on GAS-A1As quantum wells. In the first experiments in progress, we are frequency doubling the FEL output from 4 pm to 2 pm by taking advantage of the strong resonances due to intersubband transitions in the valence band of very narrow asymmetric p-type wells. In the second experiments also in progress, we are performing pump and probe transmission measurements at 4 pm in similar structures. All these nonlinear optics experiments are strongly facilitated by the high peak power of the free electron laser. Our work using FELs and the few results obtained to date by other groups with FELs indicate that spectroscopic characterization of materials with these complex machines is possible. The model for the four centers that are presently in operation in the United States is that of synchrotron facilities, where users from various disciplines share the use of the machine. I will discuss the improvements that must be made for the FELs to be as useful and reliable as synchrotrons.