Glass Dissolution Model Research Articles

Model validation and interpretation of the interaction between glass and cement in an integrated glass dissolution experiment

This article presents the modeling results from the interaction between simulated nuclear waste glasses and ordinary Portland cement (OPC). In general, the modeling results are in good agreement with the experimental results. A diffusion-based glass dissolution model well describes the release of B and Li from the glass, and a reactive transport model well describes the profiles of dissolved glass elements in the cement. It is found that the effective diffusion coefficient of Li in the cement is much larger than that of B. Sorption of both Li and B onto the cement is required to better fit their measured diffusion profiles. For B profiles in the cement, one spatially uniform distribution coefficient is needed, while for Li, sorption needs to be linked to the gel-like secondary phase at the surface of the cement. The effective diffusion coefficients for both Li and B are relatively low and can be linked to the pore-filling effects of the gel-like secondary phase. Even though it is generally believed that nuclear waste glass will dissolve faster in cementitious environments, the results show that the gel-like secondary phase can both seal the cement and act as a strong sorbent for dissolved glass species, thus inhibiting the migration of dissolved glass species.

Modeling Resumption of Glass Alteration Due to Zeolites Precipitation

Under hyper-alkaline pH or relatively high temperatures, a resumption of glass alteration can occur. This sudden acceleration of the glass corrosion rate is almost always associated with zeolites precipitation. Zeolites precipitation consumes silica and aluminum, key elements of the glass surface layer controlling its dissolution. A better understanding and quantification of this mechanism is crucial for predicting glass long-term behavior. Geochemical modeling applied to experimental data is a useful tool to reach that goal. However, it has never been applied successfully so far for modeling the kinetics of that reaction. Within the framework of the GRAAL glass dissolution model, and with simple zeolite precipitation kinetic assumptions, a simulation of glass resumption of alteration is attempted. Concentrations in solution and element mass balance between amorphous and crystallized phases are reproduced. Key parameters are revealed. Better understanding is now achievable through sensitivity studies on minerals stoichiometry and on the respective weight of dissolution and precipitation kinetics in the control of the whole reaction rate.

The fate of silicon during glass corrosion under alkaline conditions: A mechanistic and kinetic study with the International Simple Glass

International Simple Glass – a six oxide borosilicate glass selected by the international nuclear glass community to improve the understanding of glass corrosion mechanisms and kinetics – was altered at 90°C in a solution initially saturated with respect to amorphous 29SiO2. The pH90°C, was fixed at 9 at the start of the experiment and raised to 11.5 after 209d by the addition of KOH. Isotope sensitive analytical techniques were used to analyze the solution and altered glass samples, helping to understand the driving forces and rate limiting processes controlling long-term glass alteration. At pH 9, the corrosion rate continuously drops and the glass slowly transforms into a uniform, homogeneous amorphous alteration layer. The mechanisms responsible for this transformation are water penetration through the growing alteration layer and ion exchange. We demonstrate that this amorphous alteration layer is not a precipitate resulting from the hydrolysis of the silicate network; it is mostly inherited from the glass structure from which the most weakly bonded cations (Na, Ca and B) have been released. At pH 11.5, the alteration process is very different: the high solubility of glass network formers (Si, Al, Zr) triggers the rapid and complete dissolution of the glass (dissolution becomes congruent) and precipitation of amorphous and crystalline phases. Unlike at pH 9 where glass corrosion rate decreased by 3 orders of magnitude likely due to the retroaction of the alteration layer on water dynamics/reactivity at the reaction front, the rate at pH 11.5 is maintained at a value close to the forward rate due to both the hydrolysis of the silicate network promoted by OH− and the precipitation of CSH and zeolites. This study provides key information for a unified model for glass dissolution.

The Interaction Between Nuclear Waste Glass and Ordinary Portland Cement

The interaction between simulated reference waste glasses SON68 and SM539, and cement has been studied in suspensions of ordinary Portland cement and synthetic cement water with pH 13.5 at 30°C. The cement appears to trigger glass dissolution by consumption of glass matrix components. This leads to fast glass dissolution at a constant rate with formation of a porous gel layer on the glass. A glass dissolution model has been proposed considering that the silicon coming from the glass reacts with portlandite to form C‐S‐H phases. The transformation of C‐S‐H into C‐A‐S‐H phases is a second parallel driving force especially for the Al‐rich SM539 glass. After consumption of the portlandite, the glass alteration rate is expected to decrease.

The BRAG and GM2003 Models for Glass Dissolution

AbstractThe GM2003 model extends the r(t) glass dissolution model with water diffusion through the diffusion layer. Boron and alkali diffusion through the diffusion layer is described by introducing a retention factor Kd,i between boron/alkali and water in the diffusion layer. Introducing a boron/alkali diffusion coefficient, the BRAG model describes boron/alkali diffusion in the diffusion layer as well. It is shown that both models are consistent with each other and an expression is derived for the boron/alkali diffusion coefficient (BRAG) as a function of both parameters of GM2003: the retention factor Kd,i and the water diffusion coefficient DH2O in the diffusion layer. From dissolution data only, it is possible to fit the value for the boron/alkali diffusion coefficient in the diffusion layer but due to correlations the individual values of both parameters Kd,i and DH2O of GM2003 cannot be determined. From theoretical considerations follows that the Kd,i value for boron/alkali should be slightly larger than 0.1 kg/liter. A user friendly code for the BRAG model allows automatic fits of glass dissolution data in water.

A simplified model for glass dissolution in water

Numerical simulations of the water dissolution of a random ternary solid are presented. The three elements represent silica, soluble oxides (alkalis and boron) and quasi-insoluble oxides (Al2O3, ZrO2, Fe2O3,...). The soluble species are dissolved immediately when they are in contact with the solution. Their proportion is kept below the percolation threshold. For the other species, one introduces a model of dissolution-recondensation. It is shown that the dissolution rate constants should be dependent on the bonding environment in order to include surface tension. The condensation fluxes are proportional to the concentration of each species in solution. In the dynamic regime (no recondensation), one observes the congruent dissolution of silica and soluble species, after a short initial phase of selective extraction of the soluble species. The common rate of dissolution decreases with the proportion of insoluble species and increases sharply with that of soluble species. This is mainly due to the formation of a porous hydrated layer whose active surface area increases markedly with the proportion of soluble species. In the static regime (finite solution volume), the equilibrium solubility of silica decreases with the proportion of insoluble species and is practically independent of the proportion of soluble species. The porous hydrated layer is rearranged and almost free of soluble species. The ripening of the surface layer makes it protective and inhibits further extraction of the soluble species. These results are in general agreement with the experimental observations on the dissolution of durable glasses.

Testing the Grambow Glass Dissolution Model by Comparing it With Monte Carlo Simulation Results

AbstractWe have developed a Monte Carlo (simulation) model to describe glass dissolution. This model uses similar parameters to those used in the Grambow model, which has been the main glass dissolution model during the last decade. While the Grambow model is macroscopic, ours is microscopic. This assures that the value of our parameters is time independent and allows changes in the gel structure (not just concentrations) to be described.We relate our parameters to those of the Grambow model and test whether the basic assumptions of the Grambow model are consistent with our simulation results. In these simulations, as the solution evolves towards silica saturation, the silica concentration in solution can become higher than the (final) silica saturation concentration. A possible explanation for this behavior, which is observed experimentally as well, is that the silica saturation concentration is not constant as a function of time. Initially, the silica saturation concentration is the glass saturation concentration. The glass saturation concentration is higher than the final silica saturation concentration, which corresponds to saturation of the gel. In the Grambow model, the initial dissolution rate (at the gel/water interface) and the diffusion coefficient of silica in the gel are not constant either. The existence of a final dissolution rate depends on the silica content of the glass and on the protection provided by the gel layer. In the simulations, a protective gel is mainly formed by the adsorption (precipitation) of dissolved silica particles from the solution in contact with the gel. This makes us expect that (1) the gel would be non protective for dynamic tests (with a low silica content in solution) and (2) the gel would be very protective in static tests with high surface to volume ratios.

A glass dissolution model for the effects of S/V on leachate pH

The observed relationship between the ratio of the glass surface area to the volume of the leachant (S/V) and the leachate pH is discussed in terms of a simple model of the glass dissolution process. Data from leach tests on several nuclear waste glass compositions at different S/V ratios show that the leachate pH increases with time and stabilizes at a nearly constant value beyond about 28 days. This stabilized pH increases systematically with the S/V ratio of the test. The model developed here reproduces the essential features of the data and suggests that a single parameter describing the intrinsic rate of alkali diffusion and ion exchange from the glass is sufficient to represent the major glass composition dependence. Interestingly, the results are essentially independent of the rate constant for matrix dissolution. This study suggests that the diffusion-ion exchange process is central in determining the solution pH and its dependence on S/V. Under static or low-flow-rate conditions, the rate of the matrix hydrolysis reaction slows to a long-term value that is determined by the (diffusional) rate of alkali release.

Enhancement of the Glass Corrosion in the Presence of Clay Minerals: Testing Experimental Results with an Integrated Glass Dissolution Model

ABSTRACTRecent glass dissolution experiments, conducted at 90°C in the presence of potential backfill materials, indicate remarkably faster glass corrosion in the presence of clay, compared to tests where the glass is leached either alone or with alternative backfill materials. This effect correlates with the clay content in the backfill, and may be attributed to the removal of silica from solution. Sorption, or dissolution with reprecipitation of a silica-rich clay, have been proposed as possible mechanisms for the silica consumption.The results of some experiments have been tested against a glass dissolution model, in which a widely used kinetic equation for glass corrosion is coupled with diffusive silica transport through a single porosity, linearly sorbing medium, which represents the backfilling. Because the glass corrosion rates imposed by the kinetic equation are inversely proportional to the silicic acid concentration of the leachant contacting the glass, the model predicts enhanced glass dissolution if silica is sorbed by the porous medium.The experimental data proved to be consistent with the predicted enhancement of the glass dissolution. Moreover, the model-estimated distribution coefficients for silica sorption (Kd) fall within the range of values extracted from available literature data, thus supporting the hypothesis that the observed high corrosion rates are due to sorption of silica on the clay mineral surfaces.

Dissolution of R7T7 Glass in Static and Flowing Conditions: Influence of Si Diffusion Mechanism in the Leached Layer

ABSTRACTLeach tests with R7T7 nuclear waste glass in distilled water were conducted at 50 and 90°C under static and slow flowing conditions, with an SA/V ratio of 50 m−1. A computer model for glass dissolution (LIXIVER) Was used to interpret the experimental data. This mechanistic model is based on a combination of the first-order law governing surface reactions, the silicon mass transport equation for the interstitial solution in the alteration film, and an empirical law for partial silicon retention in the alteration layer. The LIXIVER model satisfactorily accounts for most of the experimental results. The importance of the silicon diffusion mechanism in the diffusion layer is stressed. Values are indicated for the apparent silicon diffusion coefficient, Dsi, which ranged from 10−16 to 10−14 m2·s−1 at 50°C, and from 10−15 to 10−13 m2·s−1 at 90°C.