During this two year project (the original proposal requested 3 years funding) we developed and tested a new design for a mini-bending jig for the hydrothermal atomic force microscope (HAFM) and a modified design for the HAFM itself. These new capabilities now permit study of the connection between stress and mineral dissolution and growth, as well as sub-critical crack growth (SCG). We demonstrated the successful design by imaging SCG of glass in situ, in real time in the HAFM, as a function of changing solution pH. We generated a movie of the SCG process. We successfully accomplished our project objectives through year 2.

Partition coefficients (Kd) of U(VI) and Np(V) on untreated and hydrothermally altered concrete were measured in 0.01 M NaCl and 0.01 M NaHCO3 solutions as functions of concentration of the radionuclides, pH, and time. The partition coefficients for both U(VI) and Np(V) on hydrothermally altered concrete are significantly lower than those on untreated concrete. The partition of both U(VI) and Np(V) are pH dependent, although the pH dependence does not appear to reflect precipitation of U and Np-bearing phases. Both sorption and precipitation are likely processes controlling partitioning of U to concrete; sorption is the most likely process controlling the partitioning of Np to concrete. The presence of 0.01 M carbonate species in solution decreases Kd of U(VI) for both hydrothermally altered and untreated concrete from ≥ 104 mL/g to ~400 to 1000 mL/g indicating a significant impact on U(VI) sorption. In contrast, the presence of carbonate only reduced the Kdof Np(V) by one order of magnitude or less. X-ray absorption spectroscopy analysis of U/concrete mixtures at different pHs and times indicate that uranyl ions are partitioned as monomeric species on untreated concrete, but oligomeric species on hydrothermally altered concrete. Similar analysis of Np/concrete mixtures shows that about half of the partitioned Np(V) is reduced to Np(IV) over a period of 6 months.

In a high level waste repository in which temperatures are elevated due to waste decay, concrete structures will be subjected to hydrothermal conditions that will alter their physical and chemical properties. Virtually no studies have examined the interaction of hydrothermally altered concrete with radionuclides. We present the results of experiments in which soluble and colloid-associated U and Np, were eluted into a fractured, hydrothermally altered concrete core. Although the fluid residence time in the fracture was estimated to be on the order of 1 minute, U and Np in the effluent from the core were below detection (10-9 - 10-8 M), for both soluble and colloid-associated species. Inorganic colloids and latex microspheres were similarly immobilized within the core. Post-test analysis of the core identified the immobilized U and Np at or near the fracture surface, with a spatial distribution similar to that of the latex microspheres. Because hydrothermal alteration followed fracturing, the growth of crystalline calcium silicate hydrate and clay mineral alteration products on, and possibly across the fracture, resulted in a highly reactive fracture that was effective at capturing both soluble and colloidal radionuclides. Comparison of results from batch experiments [1] with these experiments indicate that partitioning of U and Np to the solid phase, and equilibration of the incoming fluid with the concrete, occurs rapidly in the fractured system. Transport of U through the concrete may be solubility and/or sorption limited; transport of Np appears to be limited primarily by sorption.

We have investigated the interaction of U(VI) and Np(V) actinide ions with cementitious materials relevant to nuclear waste repositories using X-Ray Absorption Fine Structure (XAFS) Spectroscopy. The actinide ions were individually loaded onto untreated as well as hydrothermally treated cements. The mixtures were then equilibrated at varying pH’s for periods of 1 month and 6 months. In all cases uranium was observed to remain in the initial UO 2 2+ form in the Near Edge (XANES) spectra. The uranium samples show evidence of inner-sphere interactions on both treated and untreated cements at all pH’s, with the uranyl complexing with the mineral surface via sharing of equatorial oxygens. On treated cement near-neighbor U-U interactions are also observed, indicating the formation of oligomeric surface complexes or surface precipitates. Neptunium was observed to undergo a reduction from the initial NpO 2 + to Np 4+ Calculated % reduction showed ca. 15% of Np(V) is reduced to Np(IV) after a 1 month equilibration time. After 6 months higher % reduction of between 40% and 65% was observed. No Np-Np interactions were observed in the EXAFS spectra, which suggests that surface precipitation of NpO 2 is an unlikely mechanism for sorption.

The primary rationale for studying the transport behavior of radionuclides through the Engineered Barrier system / Near Field Environment (EBS/NFE) is to ascertain whether the material properties of the introduced and altered host rock can significantly affect the transport of radionuclides from the waste container to the far field. The intent of this report is to present data and modeling results that can be used to assess the importance of canister corrosion products and cementitious materials to transport of radionuclides to the far field.

Geochemical modeling was used to test the hypothesis that cation exchange withfracture-lining clays during fluid mixing in the sp Hard Rock Laboratory can significantlyaffect major element chemistry. Conservative mixing models do not adequatelyaccount for changes in Na, Ca and Mg concentrations during mixing. Mixingbetween relatively dilute shallow waters (e.g. HBH02) and more concentrated watersat depth (e.g. KA0483A) along fracture zones was modeled using the EQ3/6geochemical modeling...

The observed variability of fluid chemistry at the Aespoe Hard Rock Laboratory is not fully described by conservative fluid mixing models. Ion exchange may account for some of the observed discrepancies. It is also possible that variably charged solids such as oxyhydroxides of Fe can serve as sources and sinks of anions and cations through surface complexation. Surface complexation reactions on hydrous ferric oxides involve sorption of both cations and anions. Geochemical modeling of the surface chemistry of hydrous ferric oxides (HFOs) in equilibrium with shallow HBH02 and deep KA0483A waters shows that HFOs can serve as significant, pH-sensitive sources and sinks for cations and anions. Carbonate sorption is favored especially at below-neutral pH. A greater mass of carbonate is sorbed onto HFO surfaces than is contained in the fluid when 10 g goethite, used as a proxy for HFOs, is in contact with 1 kg H{sub 2}O. The masses of sorbent required to significantly impact fluid chemistry through sorption/desorption reactions seem to be reasonable when compared to the occurrences of HFOs at Aespoe. Thus, it is possible that small changes in fluid chemistry can cause significant releases of cations or anions from HFOs into the fluid phase or, alternately, result in uptake of aqueous species onto HFO surfaces. Simulations of the mixing of shallow HBH02 and native KA0483A waters in the presence of a fixed mass of goethite show that surface complexation does not cause the concentrations of Ca, Sr, and SO{sub 4} to deviate from those that are predicted using conservative mixing models. Results for HCO{sub 3} are more difficult to interpret and cannot be addressed adequately at this time.

Geochemical modeling was used to simulate the mixing of dilute shallow groundwater with deeper more saline groundwater in the fractured granite of the Redox Zone at the Aespoe underground Hard Rock Laboratory (HRL). Fluid mixing simulations were designed to assess the role that cation exchange plays in controlling the composition of fluids entering the HRL via fracture flow. Mixing simulations included provision for the effects of mineral precipitation and cation exchange on fluid composition. Because the predominant clay mineral observed in fractures in the Redox Zone has been identified as illite or mixed layer illite smectite, an exchanger with the properties of illite was used to simulate cation exchange. Cation exchange on illite was modeled using three exchange sites, a planar or basal plane site with properties similar to smectite, and two edge sites that have very high affinities for K, Rb, and Cs. Each site was assumed to obey an ideal Vanselow exchange model, and exchange energies for each site were taken from the literature. The predicted behaviors of Na, Ca, and Mg during mixing were similar to those reported in a previous study in which smectite was used as the model for the exchanger. The trace elements Cs and Rb were predicted to be strongly associated with the illite exchanger, and the predicted concentrations of Cs in fracture fill were in reasonable agreement with reported chemical analyses of exchangeable Cs in fracture fill. The results of the geochemical modeling suggest that Na, Ca, and Sr concentrations in the fluid phase may be controlled by cation exchange reactions that occur during mixing, but that Mg appears to behave conservatively. There is currently not enough data to make conclusions regarding the behavior of Cs and Rb.

Geochemical modeling was used to test the hypothesis that cation exchange with fracture-lining clays during fluid mixing in the Aespoe Hard Rock Laboratory can significantly affect major element chemistry. Conservative mixing models do not adequately account for changes in Na, Ca and Mg concentrations during mixing. Mixing between relatively dilute shallow waters and more concentrated waters at depth along fracture zones was modeled using the EQ3/6 geochemical modeling package. A cation exchange model was added to the code to describe simultaneously aqueous speciation, mineral precipitation/dissolution, and equilibration between a fluid and a cation exchanger. Fluid chemistries predicted to result from mixing were compared with those monitored from boreholes intersecting the fracture zone. Modeling results suggest that less than 0.1 equivalent of a smectite exchanger per liter of groundwater is necessary to account for discrepancies between predictions from a conservative mixing model and measured Na and Ca concentrations. This quantity of exchanger equates to an effective fracture coating thickness of 20 {micro}m or less given a fracture aperture width of 1,000 {micro}m or less. Trends in cation ratios in the fluid cannot be used to predict trends in cation ratios on the exchanger because of the influence of ionic strength on heterovalent exchange equilibrium. It is expected that Na for Ca exchange will dominate when shallow waters such as HBHO2 are mixed with deeper waters. In contrast, Na for Mg exchange will dominate mixing between deeper waters.

Construction-induced groundwater flow has resulted in the mixing of relatively dilute shallow groundwater with more concentrated groundwater at depth in the underground Hard Rock Laboratory (HRL) at Aespoe, Sweden. The observed compositional variation of the mixed groundwater cannot be explained using a conservative mixing model. The geochemical modeling package EQ3/6, to which a cation-exchange model was added, was used to simulate mixing between the two fluids. The results of modeling simulations suggest that cation exchange between groundwater and fracture-lining clays can explain the major element fluid chemistry observed in the HRL. The quantity of exchanger required to match simulated with observed fluid chemistry is reasonable and is consistent with the observed fracture mineralogy. This preliminary study establishes cation exchange as a viable mechanism for controlling the chemical evolution of groundwaters in a fracture-dominated dynamic flow system. This modeling study also strengthens their confidence in the ability to model the potential effects of fracture-lining minerals on the transport of radionuclides in a high level nuclear waste repository.

Predicting the chemical changes that result from excavating a repository below the groundwater table in granitic terrain is a major focus of the SKB geochemistry program. The modeling study presented here demonstrates that cation exchange can play a major role in controlling the fluid chemistry that results when groundwaters of differing composition mix due to flow induced by excavation of the HRL tunnel. The major goal of this study was to assess whether an equilibrium cation exchange model could explain the composition of groundwater sampled from boreholes in the HRL tunnel. Given the consistency of the cation exchange hypothesis with observations, geochemical modeling was used to assess whether the quantity of exchanger necessary to match model results and observation was physically reasonable. The impact of mineral dissolution and precipitation on fluid chemistry was also evaluated. Finally, the compositions of exchanger phases expected to be in equilibrium with various Aespoe groundwaters were predicted.

A program has been designed to support the prediction of cement degradation and the chemistry of water in contact with cement, over extended periods of time (e.g., 10,000 y). This multidisciplinary experimental and computer modeling investigation is intended to characterize the structural and thermodynamic properties of crystalline phases found in cement at elevated temperature. Many of these crystalline phases my be hydrated. The hydration state of these phases must be known to the interpret experimentally obtained thermodynamic data, to evaluate the stability of hydrated phases and to estimate long-term water availability, such as would be required for prediction of the radioactive-waste repository`s lifetime. The parts of the program associated with assessing and predicting dehydration/ rehydration behavior are described in this paper. (1) identification of phases present in standardized grout mixtures exposed to elevated temperatures; (2) mechanistic and thermodynamic analysis of the hydration/ dehydration behavior of hydrated calcium silicates as a function of temperature, pressure, and relative humidity; and (3) measurements of thermodynamic data for hydrated calcium silicates.

As part of the task of developing conceptual, physical and chemical, and performance models of the dynamics of water and dissolved and colloidal radionuclide constituents in the near-field, measurements of flow and transport properties of repository rocks and other near- field components are required. An apparatus has been fabricated that will be used to collect pertinent flow and transport data at temperatures expected in the near-field. Description of the core-flow apparatus (CFA) and preliminary testing results are included in this letter report. The apparatus was designed to study the flow of radionuclide-bearing solutions through fractured or unfractured cylindrical samples of rock and other competent materials (e.g. concrete). Because near-field transport was the goal, the CFA was designed to be operated at elevated temperatures.