The corrosion rate of copper in laboratory tests is shown to be a sensitive function of relative humidity, sulfur dioxide, hydrogen sulfide, ozone, hydrogen chloride, and chlorine concentrations. Observed indoor corrosion rates obey log normal statistics over the field population of this study. Also, the observed indoor rates correlate reasonably well with the measured reduced sulfur concentrations ( H2S , S8 ). The corrosion rate of silver is shown not to be humidity dependent. Hydrogen sulfide, ozone, chlorine, and hydrogen chloride concentrations substantially influence its corrosion rate. The observed indoor rates obey log normal statistics and correlate well with the reduced sulfur gas concentration. In contrast to copper, where indoor rates are 1% of outdoor values, silver often corrodes faster indoors than outdoors. Its sensitivity to sulfur gases and insensitivity to relative humidity is proposed as a plausible explanation for these findings. It is proposed that metallic silver is stable in polluted acidic atmospheric environments and therefore is the dominant surface species while Cu2O is present on the surface of copper. The thermochemistry and kinetics of these two surfaces will control the stability of silver and copper, respectively, in the presence of pollutants.

The solid-state reaction between Fe,O and MgO is best described as counterdiffusion of the magnesium and iron ions through a relatively rigid oxygen ion lattice of the rock salt structure. A relation between a mutual diffusion coefficient and a controlled cation vacancy concentration was determined by an analysis of the concentration-penetration curve of the Fe,O-MgO solid solution. The diffusion coefficient increased approximately exponentially with a linear increase of the cation vacancy concentration. This may be represented by a linear increase of the entropy of activation with vacancy concentration. The enthalpy of activation for the diffusion reaction, 47.5 kcal per mole, remained independent of vacancy concentration.

Die Korrosionsgeschwindigkeit von Cu ist in komplexer Testatmosphäre eine empfindliche Funktion der relativen Feuchte, während Ag diesbezüglich unempfindlich ist.

N CONNECTION with studies on the diffusion of FeO into singlecrystal MgO, a preliniinary investigation was made into the diffusiou mechanism of FeO and COO into polycrystalline MgO. In particular it was desirous to determine if there was cvideucc of grain-boundary diffusion. The possibility of grainboundary diffusion was suggested by Zaplatynsky.Polycrystallinc MgO of 98'% of theorctical density, 50p in grain size, and 99.99% pure was cut into bars 0.3 by 0.3 by 2.0 cm. I'cO (wiistitc) previously prepared and in fine-powder form was tightly packed around each MgO bar. The specimens were placed in an clectrical resistance furnace a t 1100°C for 25 hours in a controllcd atmosphere of hydrogen and water vapor, ensuring the stability of FeO. After the specimens were removed from the furnace, thcy wcre sectioned, polished, and observed with a reflected-light microscope. A sharp, straight boundary was observed between the diffused and nondiffused part of the MgO bar. There was no evidence of grain-boundary diffusion but only of bulk diffusion similar to the diffusion observed with single I crystals2 and a t a rate comparable with single crystals (see Fig. 1). In both the single and polycrystalline MgO the diffusion process appeared to take place by a counterdiffusion of the cations Fig. 2. Diffusion boundary of COO into "pure" polycrystalline MgO. MgO.Above boundary, solution of COO and MgO; below boundary, "pure" through a relatively rigid oxygen lattice. Because of the nonstoichiometric property of FeO, cation vacancies were present. The diffusiou process appeared to prefer bulk movement by means of vacancies over grain-boundary diffusion.Cobalt oxide was also packed around the polycrystalline MgO in the same manner as,with FeO. The specimens were placed in a gas furnace open to the atmosphere for 25 hours at 1600OC. Results showed a straight boundary indicating only bulk diffusion and no grain-boundary diffusion (Fig. 2). Bars of sintered Fig. 1. Diffusion boundary of FeO into "pure" polycrystalline MgO. Above boundary, solid solution of FeO and MgO; below boundary, "pure" MgO.

No abstract