Hydride nucleation and formation of hydride growth centers on oxidized metallic surfaces—kinetic theory

Abstract

The initial stage of hydrogen–metal reactions leading to the formation of metallic hydrides, frequently proceeds (on massive metallic samples) by the appearance of hydride “spots” (or “patches”) on the surface of the metal. These spots (henceforth denoted as growth centers, GC’s) grow at a certain velocity, overlap, and finally cover the whole surface by a hydride layer (which continues to thicken into the bulk of the metal). In certain systems it has experimentally been verified that the total available number of such GC’s is limited to a certain fixed number (typical to the given metallic surface and the reaction conditions). The kinetics of formation of these GC’s is characterized by a rate function which depends on the reaction conditions, i.e., temperature and pressure, and on the metal surface characteristics. Qualitatively, these rate functions initially accelerate, reach a maximum rate, and finally decay to zero. In the present article, a kinetic theory, which accounts for that behavior, is proposed. The model assumes that the initiation of the growth process can take place only at certain sites, with a limited number of such sites available on the surface. Also, the probability for the start-up of the growth process increases with increasing supersaturation of the hydrogen concentration at these sites. The interplay between these two factors (i.e., available GC’s sites and local hydrogen concentrations) reproduces the observed kinetics of formation of the GC’s. Analytical expressions for these rate functions are derived and compared with the experimental kinetic data reported for some metal–hydrogen systems. Universal forms of the rate function and its parameters provide a simple procedure for the evaluation of specific kinetic parameters from experimentally available data.

UD3formation on uranium: evidence for grain boundary precipitation

“…The initial development of hydrides on metallic surfaces is observed to occur in isolated zones or 'spots' [6], [10], [11], [12], [13], [14], indicating that certain areas favour the onset of hydride forming reactions. It follows that these zones experience shorter induction times than elsewhere on the sample surface, which may be attributed to the effects of defects or cracks in the oxide layer, localised thinning, regions of enhanced hydrogen species permeability or regions of enhanced hydrogen surface dissociation (although the exact nature of the hydrogen species which transport through the surface oxide film is currently unknown) [15].…”

The Hydrogen Corrosion of Uranium: Identification of Underlying Causes and Proposed Mitigation Strategies

“…Researchers at AWE and NRC-Negev regard the metal-borne defects to be of secondary importance to corrosion initiation, and instead identify the properties of the superficial oxide layer -namely, thickness and mechanical integrity -as the determinants of initiation site. 36,37 In this point-of-view, the localized and spatially heterogenous pattern of uranium hydride formation reflects a corresponding pattern of environmental hydrogen penetration through the imperfectly protective oxide to the underlying uranium. The original model of Glascott (AWE) is based on purely diffusive hydrogen penetration of a defect-free oxide layer (i.e., oxide lattice diffusion) to the metal, and an ensuing corrosion reaction at the oxide-metal interface; 38 similar Fickian transport models, describing subsequent hydrogen penetration into and reaction with the uranium bulk, were previously presented by and are regarded as the definitive theory on the hydride corrosion kinetics of massive uranium.…”

“…In the case of NRC-Negev, the oxide is assigned the central moderating role in hydride corrosion through its inhibitory effect on developing nucleation sites; 37 yet, even when they described how defects endow neighboring nucleation sites with the ability to develop into full-blown corrosion sites (i.e., "growth centers"), they chose to focus on the associated oxide breach rather than the metal defect which gave rise to the breach. Clearly, within the model construct that they themselves have defined, fully developed hydride corrosion at "growth centers" would not occur in the absence of metal defects.…”

The crystallographic structure of the air-grown oxide on depleted uranium metal

“…In terms of the Glascott [8] and Ben-Eliyahu [7] models discussed in the introduction, both of these approaches consider the diffusion of hydrogen through a surface oxide layer to the underlying metal in terms of a single diffusion coefficient. In both cases neither the structure of the oxide nor the exact hydrogen diffusion paths are defined.…”

“…The model put forward by Y Ben Eliyahu and co-workers [7] assumes that hydride reaction sites are limited to a maximum concentration for a given uranium surface. In contrast, Glascott proposed a model in which there is a distribution of oxide layer thickness across the uranium surface which leads to a variation in the observed number of hydride reaction sites with time [8].…”