No abstract

The porosity of the oxide scales formed on a number of metals during oxidation in dry oxygen has been investigated by measurements of specific surface and of density. The oxides formed during the linear oxidation of the eight metals: calcium, cerium, lead, magnesium, niobium, thorium, tungsten, and uranium are porous whereas those formed during the parabolic oxidation of copper and cobalt are impervious to oxygen gas.The oxidative behavior of a metal in oxygen depends to a marked extent on the physicochemical properties of the oxide layer formed on its surface. If the layer of reaction product remains continuous during oxidation, reaction can proceed only by the diffusion of either one or both of the reactants in the ionized form through the oxide lattice. In some cases, as with aluminum (1, 2) at 600~ the rate of oxidation falls off to a negligibly small value as time proceeds, while in others, as for example copper (3) at 1000~ the oxidation follows a parabolic rate law and the rate of attack is such that a sheet of material is oxidized completely in a few hours.On the other hand, some metals, usually above a certain characteristic temperature, oxidize at a constant rate which is maintained until the whole of the metal is consumed, a behavior exemplified by magnesium (4) at 500~This "linear oxidation" usually is interpreted in terms of the formation of an oxide scale which is porous in the sense that oxygen gas is able to pass freely through its thickness; the rate of oxidation is believed to be controlled by the rate of diffusion either of metal ions or oxygen anions through a thin barrier film of oxide adhering to the metal.The present study was undertaken in order to provide experimental evidence for the postulated porosity of the oxide layer formed during linear, or almost linear, oxidation and to show that the oxide layer formed during parabolic oxidation is, in fact, impervious to oxygen gas. The metals whose oxides were investigated were calcium, cerium, lead, magnesium, niobium, thorium, tungsten, and uranium, which oxidize at a linear rate, and cobalt and copper which oxidize at a parabolic rate. Two distinct experimental techniques were used to investigate each of the oxide products: the specific surface was measured by the method of krypton sorption (5) at --195~ and the density (6) by immersion in mercury and in carbon tetrachloride, respectively. ExperimentalEach oxide was prepared by oxidizing the metal in dry oxygen at atmospheric pressure in a glass or silica reaction chamber on a thermal balance. The experimental arrangement was the same as that previously described (4) except that the oxygen was further dried by passing it through a column packed with Linde molecular sieves (7).The percentage purity of the metals was: Ca, ~99.4; Ce, ~99.99; Co, 99.999; Cu, 99.999; Pb, 99.998; Mg, 99.95; Nb, ~99.8; Th, ~99.6; W, >99.95; and U, ~99.9. Except for calcium, cerium, and lead, the metals were in the form of sheet, 1/32 in. thick for niobium and tungsten and 1/16 in. thick for the others; a sample in ...

The kinetics of the oxidation of niobium in dry oxygen at 1 atm pressure have been measured at temperatures in the range 350°–750°C with some additional measurements at 0.1 atm. At 350°C, in tests lasting 270 hr, the oxidation is protective, but at 400°C and above the metal oxidizes, apart from an initial period, at a constant rate with a second breakaway reaction (rate transition) at 450° and 500°C but not at higher temperatures. The anomalous temperature coefficient of the linear rate has been confirmed. The specific surface of the oxide scale has been measured and, apart from an anomaly at 400°C, is shown to decrease with increasing temperature of oxidation and this is ascribed to sintering. The oxidation kinetics of a purer batch of niobium were also investigated. The effect of moisture is to decrease the rate at 400° and 450°C, whereas at 600°C the rate is unchanged.

No abstract

The accepted mechanism of paralinear oxidation that an inner, compact layer of oxide forms at a parabolic rate and transforms to an outer, porous layer at a linear rate has been tested in the case of tungsten. Samples were oxidized at 750° and 800°C for increasing times, and the amount of porous oxide present on each sample was estimated by the method of krypton sorption. The rate of porous oxide formation is not constant, but instead decreases with time to reach an almost constant value so that the results are not in accord with the paralinear model. A similar conclusion is reached by metallographic examination of the partially oxidized samples. Results for the oxidation at 600°C are used to illustrate the difficulty of testing the paralinear rate equation by an analysis of kinetic data alone.