the objective is to demonstrate the feasibility of producing high-efficiency (15 or greater) thin-film gaas solar cells with costs suitable for terrestrial solar electric power generation. the approach is that of growing gaas by organio-metallic chemical vapor deposition on recrystallized germanium (ge) films previously deposited on metal substrates and fabricating amos (antireflecting metal-oxide-semiconductor) solar cells on the gaas. previously it had been determined that a water vapor-grown native oxide (temperature = 25 exp 0 c) was the most useful native oxide for amos cells. a new chemical surface preparation prior to oxide growth led to more uniform oxides and reduced interface contamination, yielding lower reverse saturation current densities, a near-unity diode ideality factor, and better reproducibility. substituting silver (ag) for gold metallization showed no change in starting cell efficiency, but did greatly improve high temperature stability of the amos solar cell. a new study was completed on antireflection coatings on amos gaas solar cells, taking into account the spectral response of the cell and nature of the solar spectra, and the results submitted for publication. xps (x-ray photoelectron spectroscopy) studies had found earlier that the more efficient native oxides had primarily as sub 2 o sub 3 and ga sub 2 o sub 3 with little gaaso sub 4 . a new chemical step etching was developed which can be used to profile the oxide in 5- to 7-a exp 0 steps without modifying the oxide chemistry as does ion sputtering. a new schottky barrier structure is described