No abstract

This work presents millimeter wave performance achieved by ion-implanted InGaAs/GaAs metal semiconductor field-effect transistor devices. A current gain cutoff frequency ft of 126 GHz and maximum frequency of oscillation fmax of 232 GHz have been measured for 0.20 μm gate length devices. The ft and low-field Hall mobility data, measured at 300 and 112 K, lead us to conclude that the average electron velocity under the gate is mainly due to the high-field velocity rather than low-field electron mobility.

No abstract

Degradation in optical and electrical properties has been observed for high-purity and high-mobility p-type GaAs layers which contain significant concentrations of an unidentified shallow acceptor-like defect, labeled “A”, that is frequently incorporated in crystals grown by molecular beam epitaxy. Low-temperature photoluminescence and variable temperature Hall-effect measurements were employed to monitor the aging process in samples stored for about one year at room temperature. Profound changes in the exciton recombination spectra, indicative of increasing concentration of the “A” defect, have been accompanied by a decrease in hole mobility and an increase in carrier concentration. These results are discussed in the context of the acceptor-pair defect model, originally proposed by Eaves and Halliday [J. Phys. C: Solid State Phys. 17, L705 (1984)].

Surface-related effects in undoped and Cr-doped semi-insulating (SI) liquid-encapsulated Czochralski (LEC) GaAs, as seen in Hall effect and resistivity measurements, are shown to be caused by mixed conduction due to a p-type surface space-charge layer, as predicted by theoretical calculations employing a surface-interface state model recently developed to explain surface-potential changes in epitaxial GaAs. Wet chemical treatments with ammonium hydroxide were used to induce especially large surface effects and are shown to form a substantial hole concentration at the surface. The excellent agreement between the theoretical results and experimental data provides strong support for the surface-interface state model used and demonstrates its utility as the first ‘‘working’’ mathematical model of surface-interface states in GaAs. In addition, wet chemical treatment in hydrogen peroxide is shown to minimize the surface layer and thereby allow accurate as well as stable and repeatable electrical characterization of the n-type bulk.