When operated at the current densities in excess of 10 mA/cm2 that are typical of efficient solar‐irradiated photovoltaic cells, normaln‐normalCdSe/1MNa2S‐1MS‐1MnormalNaOH/C semiconductor‐liquid junction cells deteriorate with time. In normalCdSe electrodes that have lost activity, formation of a normalCdS‐normalenriched film has been detected by means of Auger spectroscopy and electron beam‐induced luminescence. Such a layer is a barrier to the flow of holes to the surface. The rate of deterioration increases with light flux, usually decreases with stirring, and depends on the crystal face exposed to the solution. Addition of small amounts of elemental selenium to the solution substantially improves the stability by preventing excessive sulfide enrichment of the surface. For example, with 0.5M Se added, the output of <112̅0> face electrodes run at 35 mA/cm2 is essentially unchanged beyond 2×104C/cm2 charge passage.

An electron spin resonance (ESR) technique is employed to determine the free radical distribution in the articulating surfaces of retrieved acetabular cups and knee-joint plateaus (retrieved after more than 6 years of implantation). Similar measurements made on samples prepared from cyclically stressed and unstressed cups, and on samples following oxidations in nitric acid and intralipid solutions provided sufficient data to gain more knowledge about the combined chemical and mechanical effects on PE free radicals during implantation. In UHMWPE free radicals are primarily initiated by gamma-ray sterilization; however, during implantation, peroxy (scission type) free radicals are formed and reach a maximum concentration level (equilibrium state) due to oxidation by chemical (hemoglobin and/or synovial fluids) environment of the joints. Subsequently, due to frictional heating and stress in the loading zones, free radical reaction is accelerated and their number is reduced only in those areas. This is consistent with the observations of a temperature rise in acetabular cups during in vitro frictional wear stress tests and in vivo telemetry observations, as reported by others. Compared with the previously reported SEM micrographs the low-free-radical regions are correlated with high-wear areas and the high-free-radical regions with the low-wear areas.

No abstract

A fundamental pattern of oxide‐substrate reactions has been identified in thermally annealed anodic films on normalGaSb and normalGaAs through the use of Raman scattering and ternary phase diagrams. The relevant reactions on normalGaAs and normalGaSb yield interfacial deposits of As and Sb via AS2O3+2normalGaAs→Ga2O3 +4normalAs and Sb2O3+2normalGaSb→Ga2O3+4normalSb . The failure to observe elemental P generation in annealed anodic films on normalGaP in conjunction with an estimate of the Ga‐P‐O phase diagram suggests that oxide‐oxide reactions occur in the film instead. Formation of either ortho or metaphosphate will depend on the Ga2O3/P2O5 ratio in the film according to P2O5+Ga2O3→2GaPO4 and 3P2O5 +Ga2O3→2normalGafalse(PO3)3 .

Surface optical phonons have been detected on both the polar (100) and nonpolar (110) faces of GaAs, InP, and GaP with the use of high-resolution electron-energy-loss spectroscopy. The observed frequencies (GaAs: 291, InP: 337, GaP: 396 cm ') on all surfaces are in excellent agreement (within 5 cm ') with theoretical predictions and were found to be independent of crystal face, bulk doping level, or method of sample preparation (cleavage in UHV versus sputter-annealing).The insensitivity of the surface-optical-phonon frequencies to the details of surface orientation, reconstruction, and near-surface 0 stoichiometry is a consequence of the effective depth of atomic displacements (-200 A) which contribute to the time-dependent electrostatic potential outside the crystal. Adsorption of atomic hydrogen results in identifiable modes for the Ga-H, As-H, In-H, and P-H stretching vibrations at approximately 1880, 2110, 1700, and 2350 cm, respectively.

The condensed phase portion of the In‐P‐O equilibrium phase diagram has been determined using a combination of thermodynamic calculations and binary mixture reaction experiments. The diagram can be used as a guide for studying the thermal oxidation of normalInP and for predicting patterns for thermally induced reactions between normalInP and its native oxides.

The phases formed from the oxidation of normalGaAs have been considered in the light of an estimated equilibrium Ga‐As‐O condensed phase diagram. Thermal oxidation leads to phases consistent with the phase diagram. Plasma oxidation and anodic oxidation lead to phases inconsistent with the phase diagram. Thermal annealing of the plasma and anodic oxide layers leads to reactions producing phases consistent with the phase diagram. It is concluded that plasma and anodic oxides are produced far from equilibrium and that the estimated phase diagram is a satisfactory approximation to the equilibrium ternary phase diagram.

The identification and distribution of chemical species comprising the native oxides of anodicalty and thermally grown films have been studied using XPS in conjunction with ion milling and chemical etching. Low angle electron diffraction was used to identity crystalline surface (~,150A) layers of GaAsO4 produced thermally from As2OJO2 mixtures at elevated (T > 600~ temperatures. At lower growth temperatures (T ~, 500~ an amorphous modification of this product is indicated. The primary bulk constituents of all films, both anodically and chemically grown, were found to be As2Oz and Ga203. In anodic films the molar ratio of As2OJGa2Oa was approximately unity and uniform from the surface to the oxide/GaAs interface. The anodic interface width was found to be relatively sharp (100-120A). Water rinsing of anodic films dissolves the As20~ from the near surface (~200A) volume. Annealing of 2000A anodic films at 650~ for periods from 1 to 16 hr generated a pentavalent arsenic component which was limited to the surface. A general feature of the thermal oxides was the loss of As203 from the bulk film. In air-grown samples (530~ 4 hr) this loss was nearly complete and the native oxide was primarily Ga20~ with some addition of nonuniformly distributed elemental arsenic and unoxidized GaAs. The use of AS203/O2 mixtures failed to produce bulk oxides with a high, uniform concentration of As203. Studies comparing ion milling and chemical etching demonstrated the existence of artifacts associated with argon beam reduction of As208 in anodie films even though virtually no reduction of crystalline or vitreous As208 control samples was observed.Over the past one to two years, the literature on native oxides on III-V's in general, and GaAs in particular, has grown tremendously. Whereas sporadic individual efforts had been reported on previously, during the recent past, more concerted group efforts in a number of laboratories around the world have been attacking various aspects of the problem. Numerous publications have appeared which describe the results of studies on new and improved oxide growth techniques, electrical and optical characterization of the oxide-semiconductor system, and the fabrication of device structures of the metal-insulatorsemiconductor variety.While the technology and applications have been moving along very rapidly, the scientific understanding of the oxide growth and properties has not kept pace. It is now recognized that much better knowledge of the chemical characteristics (i.e., chemical constitution, uniformity of composition, etc.) is required before fully controllable and useful device structures can be made reproducibly. A good deal of energy is now being expended in this direction and it is here that the full complexity of the problem becomes eminently apparent.Two basic approaches are being followed relative to the surface chemistry involved, namely: (i) analysis of "clean" surfaces and the initial stages of oxygen adsorption and reaction and (ii) analysis of variously oxidized "real" surfaces for o...