Atom-Probe Tomography, TEM and ToF-SIMS study of borosilicate glass alteration rim: A multiscale approach to investigating rate-limiting mechanisms

摘要

Significant efforts have been made into understanding the dissolution of silicate glasses and minerals, but there is still debate about the formation processes and the properties of surface layers. Here, we investigate glass coupons of ISG glass - a 6 oxide borosilicate glass of nuclear interest - altered at 90 C-omicron in conditions close to saturation and for durations ranging from 1 to 875 days. Altered glass coupons were characterized from atomic to macroscopic levels to better understand how surface layers become protective. With this approach, it was shown that a rough interface, whose physical characteristics have been modeled, formed in a few days and then propagated into the pristine material at a rate controlled by the reactive transport of water within the growing alteration layer. Several observations such as stiff interfacial B, Na, and Ca profiles and damped profiles within the rest of the alteration layer are not consistent with the classical inter-diffusion model, or with the interfacial dissolution-precipitation model. A new paradigm is proposed to explain these features. Inter-diffusion, a process based on water ingress into the glass and ion-exchange, may only explain the formation of the rough interface in the early stage of glass corrosion. A thin layer of altered glass is formed by this process, and as the layer grows, the accessibility of water to the reactive interface becomes rate-limiting. As a consequence, only the most easily accessible species are dissolved. The others remain undissolved in the alteration layer, probably fixed in highly hydrolysis resistant clusters. A new estimation of water diffusivity in the glass when covered by the passivating layer was determined from the shift between B and H profiles, and was 10(-23) m(2). s(-1), i.e. approximately 3 orders of magnitude lower than water diffusivity in the pristine material. Overall, in the absence of secondary crystalline phases that could consume the major components of the alteration layer (Si, Al), it is assumed that the glass dissolution rate continuously decreases due to the growth of the transport limiting alteration layer, in good agreement with residual rates reported in the literature for this glass. According to our results it can be expected that new kinetic models should emerge from an accurate time dependent budget of water within the nanoporous alteration layer. (C) 2016 Elsevier Ltd. All rights reserved.