No abstract

higher than given by the equation above, say nl ----2.18 + 0.296 (k/103A --2.305)-1-2. No information is available on what effect surface imperfections may have on the ellipsometric method.The ellipsometric method could detect the small dependence of n, on formation conditions, such as was found to exist to at least statistical significance by the modification of the Abel,s method, only by working at the most sensitive 4o and D. Since data based on thicknesses of anodic oxide films derived from the density of Ta205 8.735g cm -~ of unknown origin given by the Handbook of Physics and Chemistry are still being published (11), it is perhaps worth noting that the thickness obtained from this density will be about 10% too small on the basis of the values of nl here confirmed. The fact that a good fit is obtained for 4o ----70 ~ and for 4o ~ 50 ~ confirms that the optical constants do not depend on the angle of incidence. The fit immersed as well as dry confirms the accuracy of nl and also shows that pores into which electrolyte may enter are absent. A further test is that when points of equal thickness are compared at different angles of incidence (Fig. 6, 7) experimental and theoretical results agree. The increases in thickness for one cycle of the ellipsometer plot ~/2nl cos 41 are consistent when the thickness is obtained from spectrophotometric data with adjustment of nl. ConclusionsThe chief value of the ellipsometric technique in the determination of the thickness of films would seem to lie in its application to films to which the more precise and speedy spectrophotometric location of wavelengths of minimum reflectivity (1, 6) cannot be applied either because the films are too thin (<150A or so for tantalum) or because they must be measured still immersed in the electrolyte or because the optical constants of oxide and substrate are poorly "matched" for sharp interference minima. The method is very sensitive to the film thickness at certain thicknesses and quite insensitive at others, as may be seen by examining the figures. The measurement of intensity of reflection of p-light ( 7) is more precise for the determination of the optical constants of the oxide and metal.

Electrodeposited platinum-tin mixtures show enhanced activity toward the oxidation of methanol, formaldehyde, and formic acid when compared with platinum black. This increased activity is particularly noticeable with formaldehyde, the anode polarization being decreased by 0.4V at a current density of 100 mA/cm 2.The use of platinum-rhenium electrodeposits as an anode catalyst for the oxidation of methanol has been described (1-3). It was suggested (2, 3) that the improvement in performance over a platinum black catalyst was due to the direct reaction of an oxide of rhenium with chemisorbed fuel residues, followed by electrolytic reoxidation of the reduced rhenium oxide. After reviewing the standard oxidation potentials of several metal-metal oxide systems, it seemed likely that other platinum-metal oxide combinations would give enhanced activity toward methanol oxidation. Several elements were codeposited electrolytically with platinum, and the mixed deposits tested, in the first instance, for catalytic activity toward the oxidation of methanol and formaldehyde in 0.5M sulfuric acid. The more active catalysts were examined also for their ability to oxidize formic acid, for, although this is not an economical fuel, it can occur in actual battery systems as a partial oxidation product of formaldehyde (4). Of the combinations tested, platinum-tin was the most promising catalyst.The addition of tin to platinum to improve its performance as an oxidation catalyst has been patented (5, 6), but little detailed information is given as to the characteristics of these catalysts. The present study gives results for the oxidation of methanol, formaldehyde, and formic acid over a range of temperature and fuel concentration.