No abstract

Photomeasurements on Al–Al2O3–Au thin-film sandwiches revealed a very asymmetric barrier shape, Φ1 and Φ2 being 1.8 and 4 eV, respectively. The image-force lowering of the barrier can be explained by the high-frequency dielectric constant. When photoemission measurements are performed only in the visible region, the Au electrode will not emit photoelectrons because of the large Φ2 value. The Al electrode, however, will emit photoelectrons if the Au electrode is sufficiently positively biased. The existence of only one source for photoelectrons facilitates the interpretation of photomeasurements. Monte Carlo calculations were performed to determine the scattering of hot electrons within these films. Electron-optical phonon scattering was assumed and a mean free path of 10 Å was found. The method to determine scattering by photoemission gives values on scattering within the image-force region.

A pronounced nonuniformity in films of Ta deposited by electron-beam evaporation is reported. This effect is explained by the fact that positive Ta ions are generated during evaporation and, due to reflected electrons, the substrate acquires a high negative potential which accelerates the Ta ions and causes sputtering.

A simple method of separating the opposing photocurrent components which occur in thin-film, metal-insulator-metal sandwich structures at low (or zero) applied bias voltages is presented. The method uses data obtained by illuminating the sample first from one side and then from the other, both electrodes being semitransparent, and calculating the ratio of the resulting photo-yields over a suitable photon energy range. The analysis has been applied to evaporated Al-Al3O3-Al diodes in which the Al2O3 layer was formed by evaporation of sapphire. The experimental results show that the proposed analysis is adequate for its intended purpose.

We report a Monte Carlo study of hole transport in AIXGa, -,As, In, -,Al~,As, and GaAs,Sbr --x. The effects of alloy scattering are significant in all three cases, but mobilities are still high enough to be advantageous in particular device applications. We separately calculate the Hall r factors by a Boltzmann transport method and show that these factors are vitally important when attempting to compare Monte Carlo drift mobilities with experimental Hall data. 0 1995 American Institute of Physics.