SON68 nuclear glass dissolution kinetics: Current state of knowledge and basis of the new GRAAL model P. Frugier a, * , S. Gin a , Y. Minet a , T. Chave a , B. Bonin b , N. Godon a , J.-E. Lartigue a , P. Jollivet a , A. Ayral c , L. De Windt d , G. Santarini e a CEA Marcoule, DTCD/SECM/LCLT, BP 17171, 30207 Bagnols-sur-Cèze cedex, France b CEA Saclay, DEN/DIR/DS, 91191 Gif-sur-Yvette cedex, France c IEM/CNRS-ENSCM Université Montpellier 2, CC 047, Place Eugène Bataillon, 34095 Montpellier cedex 5, France d ENSMP, CG, 35 rue St Honoré, 77305 Fontainebleau cedex, France e CEA Saclay HC/CAB, 91191 Gif-sur-Yvette cedex, France article info Article history: Received 2 April 2008 Accepted 27 June 2008 abstract This article summarizes the present state of knowledge concerning aqueous alteration of R7T7-type nuclear containment glasses, represented mainly by the inactive reference glass designated SON68. Based on this review, we propose to describe the glass alteration kinetics up to and including the final residual rate regime by means of a new mechanistic model known as GRAAL (glass reactivity with allowance for the alteration layer). Phenomenological analysis findings are reviewed for the various glass alteration regimes: interdiffusion, initial rate, rate drop, residual rate and, under very particular circumstances, resumption of alteration. These alteration regimes are associated with predominant mechanisms. Pub- lished work interpreting and modeling these mechanisms was examined in detail. There is a broad con- sensus on the general mechanisms of the initial rate and even the interdiffusion regime, whereas the mechanisms controlling the rate drop remain a subject of dispute not only with regard to nuclear glasses but also for the dissolution of silicate minerals. The reaction affinity responsible for the rate drop is expressed differently by different authors and depending on the underlying theories. The disagreement concerns the nature of the phase (glass or gel) or the activated complex controlling the rate drop, which in turn determines the elements that must be taken into account in the overall affinity term. Progress in recent years, especially in identifying the mechanisms responsible for the residual rate, has shed new light on these issues, allowing us to propose new theoretical foundations for modeling the different kinetic regimes of SON68 nuclear glass dissolution. The GRAAL model considers that water diffusion in the passivating reaction zone (the gel formed under saturation conditions) is a rate-limiting step in the overall glass dissolution kinetics. Moreover, this passivation zone is a soluble phase whose stability is directly dependent on the nature of the secondary phases likely to precipitate and on the solution renewal conditions. Ó 2008 Elsevier B.V. All rights reserved. 1. Introduction The reference option in France for the management of vitrified high-level waste packages is deep geological disposal (Act 2006- 739 dated 28 June 2006). Although the final site has not yet been selected, research has focused on a Callovian–Oxfordian clay layer in the northeastern Paris basin [1]. ANDRA has been operating an underground laboratory in this layer since the early 2000s. The current disposal concept for vitrified wasteforms is based on three containment barriers: the glass package, inside a stainless steel container (primary canister), a carbon steel overpack 55 mm thick, and a layer of argillite. Demonstrating the safety of the concept im- plies assessing the long-term behavior of the glass in contact with groundwater during the thousands of years necessary for decay of the radionuclides incorporated in the glass structure. Over time scales inaccessible to laboratory experimentation, and faced with the chemical complexity of the glass and its environment, model- ing is the principal mean of assessing the source term, i.e. the flow of radionuclides released from the glass over time. Although the source term depends on thermal, hydraulic, mechanical, chemical, radiological or even microbiological phenomena, coupling between chemical and hydraulic phenomena is the key issue for source term prediction. The chemical compositions of the nuclear containment glasses produced at La Hague lie within a range specified by AREVA. The range takes into account the variability of the waste composition in the feed stream, and was determined to meet the technological requirements of glass fabrication (viscosity of the glass melt) and to guarantee the fundamental material properties (homogeneity, 0022-3115/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jnucmat.2008.06.044 * Corresponding author. Tel.: +33 466797724; fax: +33 466796620. E-mail address: pierre.frugier@cea.fr (P. Frugier). Journal of Nuclear Materials 380 (2008) 8–21 Contents lists available at ScienceDirect Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat