No abstract

Pure polycrystalline A1 (99.998%) was oxidized in dry oxygen and water. The surface products formed were analysed by ESCA. The thicknesses of the amorphous A1,0, formed during oxidadion for 1 h in dry oxygen at 25 OC and 250 "C are 14 and 20 A, respectively. During exposure of the oxidized sample to water a surface layer of amorphous AI,O, is hydrated and AI(OH), is formed by the reaction: Al,O, + 3H,O +2AI(OH), . The hydroxide is not stable in a vacuum. During analysis in the ESCA instrument, AI(OH), is dlecomposed to AI,O, and H,O because the equilibrium pressure of H,O is higher than the partial pressure of watler in the vacuum system. It is found that a 7 hi thick layer of AI,O, is formed on the surface of AI(OH), . The binding energies of the AI3+ (2p) and 02-(ls) signals recorded from AI,O, formed on Al at 22OC and at 25OOC are 7' 5.8 and 532.7 eV, respectively. The binding energies of these peaks are 0.6 eV lower for a sample preoxidized at 25OOC and then exposed to a humid atmosphere and immediately analysed. It is suggested that the change of the binding energy is due to set up of a space charge during exposure to a humid atmosphere. After exposure for a long time to a humid atmosphere the charging decays and the binding energies approach their normal values.

Pure Al (99.998%) and technically pure A1 (99.5%) were oxidized in oxygen at room temperature, 250OC and 500°C. The oxidized samples were exposed to a humid atmosphere (80% RH) at room temperature for times ranging from 1 day to 1 month. The reaction products were analysed by ESCA, scanning Auger microscopy, energy-dispersive x-ray spectroscopy (EDX) and SEM. The thicknesses of the surface oxides formed during oxidation for 5 h at the same temperatures as above are 15, 22 and 65 di, respectively. After exposure of the technically pure A1 material to a humid atmosphere the analyses show localized corrosion and formation of hydroxide at noble precipitates like A1,Fe and a-AlFeSi. The distribution of hydroxide particles on oxidized A1 can be recorded by SEM owing to the fact that the secondary electron yield of A1,0, is larger than the yield of AI(OH), , Exposure of pure Al, preoxidized at room temperature, to a humid atmosphere for 1 week causes a growth of the oxide thickness from 15 to 22 A. The corresponding growth in a dry atmosphere for the same time is only 1 A. It is suggested that the growth in a humid atmosphere is due to hydration of A1,0,. Thereby, the oxide becomes thinner and oxidation of A1 occurs. The AI(OH), formed by hydration is decomposed to A1,0, in the ultrahigh vacuum (UHV) system and is therefore detected as an oxide growth.

No abstract

The effect of the spherical shape of powder on the determination of oxide thickness in ESCA analysis is investigated. A theoretical model, which takes account of the angular dependence of the x-ray flux, the photoelectron intensity and the etch rate, is outlined. The model agrees well with experimental depth profiles obtained for A1 and Fe powders covered with thin oxide layers (14-20 hi). By using an effective escape depth factor, the thickness of thin oxide layers can be estimated from the oxide and metal intensities. The thickness of oxide layers greater than three times the electron mean free path can be estimated from the etch depth where the metal intensity is 65% of its maximum.