The U.S. Department of Energy was studying the feasibility of locating a high-level radioactive waste repository in basalt at the Hanford site in south-central Washington. This is a saturated site where ground water transport of radionuclides away from a repository is the mechanism for release to the accessible environment. Sorption as defined herein is the predominant means by which the movement of radionuclides are retarded in a ground water flow system. This work presents preliminary laboratory batch testing results in the context of the entire site system sorption characterization strategy. An anoxic experimental approach was used with basalt flow top core samples recovered from depths of 915 m (3000 ft) below the earth surface. The kinetics of sorption were relatively slow and increased in the order U<Se<Sr, Ra. Proportionate removal increased in the order Se<U<Sr<Ra. Freudlich isotherms were linear for Sr, U and Se over the concentrations tested. The batch results were helpful from a screening standpoint to indicate the direction of further testing, but were unsatisfactory by themselves to explain sorption behavior adequately.

The US Department of Energy is studying the feasibility of locating a high-level radioactive waste repository in basalt at the Hanford Site in south-central Washington. This is a saturated site where groundwater transport of radionuclides away from the repository is the mechanism for release to the accessible environment. Sorption is the predominant means by which the movement of radionuclides is retarded in a groundwater flow system. This report presents results to date from laboratory batch testing using an anoxic (low oxygen) experimental approach with basalt flow top core samples. The kinetics of sorption increased in the order U < Se < Ra. The relatively rapid removal and ease of desorption of selenium was typical of an ion-exchange type process. The removal of uranium from solution was slower and more strongly retained on the solid surfaces. Its behavior was more typical of selective chemisorption or coprecipitation processes. Extremely low solubility of radium was responsible for rapid irreversible removal from solution. Proportionate removal increased in the order Se < U < Ra. Freundlich sorption isotherms were linear for uranium and selenium over the concentrations tested. 30 refs., 17 figs., 6 tabs.

Experiments have been completed in the basalt-Tc-doped groundwater sys- tem to quantify the mobility of Tc in the waste package and near field envion- ment in a repository located in basalt. The experiments were completed using a batch sorption technique in which 8 rushed basalt and Tc-doped groundwater were reacted at 85, 100, 125 and 150 °C for periods of time up to 118 days. The water to rock ratio was 10:1 ml/g. In another set of experiments, basalt crushed under anoxic conditions and Tc-doped groundwater were reacted at 85°C at different water to rock ratios (10:3, 20:3, 40:3 and 80:3 ml/g). The rate and percentage of Tc removed from solution increased with increasing temperature and decreasing water to rock ratio. Also, basalt crushed under anoxic conditions removed >87% of Tc from solution at 85°C in 7 days while no Tc was removed from solution with basalt crushed in air at the same temperature after 118 days. These results are consistent with the hypothesis that (1) ferrous iron in basalt phases preferentially consumes oxygen and will not reduce Tcuntil all free oxygen in the system is consumed. Once oxygen is consumed, reduction of Tc is quite rapid. These data indicate that the potential exists in a basalt hydrothermal environment to satisfactorily immobilize Tc. Reliable application of these data will depend on experimental and theoretical estimates of oxygen consumption with time in the waste package environment following repository closure, subsequent saturation and long term hydrothermal reactions.

Two different colloidal suspensions, representative of those found in waste package interaction tests, were prepared from iron metal and silica powders or sodium-bentonite at 90°C. Aliquots were spiked with 233U, 235Np, 237Pu, or 95mTc at pH ranges from 2 to 12, then shaken for 24 hours followed by a 15Å filtration. Zeta potential measurements were made on unspiked samples. Similar sorptive properties were observed for both colloids. At 25°C both 233U and 237Pu exhibit maximum sorption (50-90%) near pH 6. Sorption drops by about a factor of 5 at pH >8. Slight sorption of 235Np occurs at pH 11 and decreases to zero at lower pH values. 95mTc does not sorb on Fe-silicates and is only slightly sorbed (10%) on smectites.

Sodium selenate and Na275SeO4 were used with a synthetic groundwater and hydrazine to determine selenate-selenium sorption characteristics of crushed Umtanum basalt between 40° and 60°C. Selenium sorption kinetics from both oxidizing and reducing solutions followed an equation of the type:

Sorption of uranium on amorphous ferric oxyhydroxide was investigated at 25° and 60°C from 0.01 M NaCl and 0.01 M NaHCO3 solutions over an initial U concentration range of ∼10−4M to 5 · 10−7M (23,800-93.2 ppb U). Uranium distribution coefficients ranged from more than 2 · 106 ml g−1 from 0.01 M NaCl at 25°C to ∼ 3 · 104 ml g−1 from 0.01 M NaHCO3 at 25°C and fell rapidly with increasing initial U solution concentration. The uranium sorption data fit a Dubinin-Radushkevich sorption isotherm.

Radionuclide distributions in groundwater-colloid-basalt systems were measured using GR-3 groundwater and crushed Umtanum basalt at 60{degree}C. The objective was to estimate the potential for radionuclide transport from a nuclear waste repository in basalt by colloids suspended in groundwater. Three colloids were studied -- a bentonite colloid representing a potential component of packing or backfill materials in the repository, and two hydrated silica colloids which might represent those generated from a glass waste form. The radioelements studied included isotopes of neptunium, uranium, selenium, technetium, and radium. Measurements of radionuclide distributions in experiments with the bentonite colloid showed that uranium was sorbed strongly on the colloid (under both oxidizing and reducing conditions) but was readily transferred to basalt when it was added to the system. Sorption of neptunium, technetium, and selenium on the colloid was greatly enhanced by using reducing conditions. Only small amounts of neptunium and technetium were transferred to the basalt under reducing conditions, but most of the selenium was readily transferred under these conditions.

Radium sorption efficiencies as a function of temperature, Ra concentration, and secondary mineral sorbate were determined in a 0.01 M NaCl solution. Radium sorption on a characterized clinoptilolite, montmorillonite, nontronite, opal, silica gel, illite, kaolinite, and glauconite under comparable experimental conditions allowed determination of Ra sorption efficiency curves for each, through use of Freundlich constants, over the same temperature and initial Ra solution concentration range. Similar sorption data for U on the same secondary minerals over the same temperatures allowed comparison of sorption efficiencies for Ra and U. Clinoptilolite, illite, and nontronite were the most efficient Ra sorbents, while opal and silica gel were the poorest Ra sorbents. Generally, Ra sorption on secondary minerals was much greater than U sorption under the same experimental conditions.

The sorption of U and Ra on finely ground biotite, muscovite, and phlogopite was adequately described by the Freundlich adsorption equation, (x/m) = KCn, at low U and Ra concentrations despite Ra precipitation at the higher temperature. Radium and U sorption-efficiency curves derived from the Freundlich constants generally showed decreased distribution coefficients in response to increasing temperature and increasing Ra or U concentrations. Temperatures investigated were 5°C, 25°C, and 65°C. Solution compositions used were 0.1 M NaCl and 0.01 M NaHCO, for U, and 0.01 M NaCl for Ra. Uranium initial solution concentrations ranged from 1.00 × 10−4 M to 4.00 × 10−7 M; the Ra initial solution concentration range was 6.80 × 10−7 M to 8.60 × 10−9 M. In 0.01 M NaHCO3 solutions, anionic uranyl carbonate complexes were prevalent, and because they are weakly sorbed relative to free uranyl ion and uranyl hydroxy complexes, the result was a relatively low U sorption efficiency on biotite and phlogopite and excellent sorption efficiency on muscovite. Uranyl carbonate complexes decreased in solubility with increasing temperature, so that U sorption efficiency on biotite increased with increasing temperature. Sorption of uranyl ion and uranyl hydroxy cations on biotite decreased with incresaing temperature.

The kinetics of radionuclide sorption and desorption reactions in basalt-groundwater systems were evaluated at 60{degrees}C using a batch equilibration method. It was found that many sorption reactions on surfaces of fresh (unaltered) basalt from the Umtanum and Cohassett flows are slow. Some reactions require more than 50 days to reach a steady state. Sorption of neptunium and uranium in oxidizing (air saturated) groundwater appears to be controlled by slow reduction of these elements by the basalt surfaces. The resulting lower oxidation states are more strongly sorbed. Technetium and selenium, which are anionic under oxidizing conditions, are not measurably sorbed on fresh basalt surfaces, but are slightly sorbed on the altered surfaces of flow top basalt. Under reducing conditions, where the groundwater contains dilute hydrazine, sorption is faster for neptunium, uranium, technetium, selenium, and lead. Plutonium sorption rates were not affected by the groundwater Eh. It was shown that radium was precipitated rather than sorbed under the conditions of these experiments. Umtanum flow top material sorbed radionuclides much faster than fresh basalt surfaces due to its greater surface area and cation exchange capacity. Desorption rates for plutonium, uranium, neptunium, technetium, and selenium were generally much less than sorption rates (especially formore » reducing conditions). These radionuclides are irreversibly sorbed on the basalts. 25 refs., 20 figs., 19 tabs.« less

The sorption behavior of selected radionuclides on the Columbia River basalts has been investigated. Radionuclide distribution coefficients, using a batch-equilibrium technique, have been determined for iodine, selenium, technetium, strontium, cesium, uranium, radium, plutonium, americium, and neptunium. Since the distribution coefficient value is an empirical value, the effects of temperature, pressure, groundwater composition, and Eh conditions on the distribution coefficient value for these isotopes have been investigated. In addition, sorption isotherms, describing the dependence of radionuclide sorption on radionuclide concentration for cesium, strontium, radium, plutonium, and uranium (under both oxidizing and reducing conditions) have been determined. Based on these sorption data, it appears that, under the expected ambient repository conditions (e.g., reducing, alkaline conditions), the Columbia River basalts are capable of strongly retarding cesium, strontium, radium, and neptunium migration and moderately retarding uranium, technetium, and plutonium migration. The basalts are not capable of significantly retarding the migration of iodine and selenium.

The Basalt Waste Isolation Project (Rockwell Hanford Operations) is conducting a safety assessment of nuclear waste storage in a repository on the Hanford Site. Pacific Northwest Laboratory, in support of the assessment effort, is generating radionuclide distribution coefficient data between simulated groundwaters and basalts and their secondary mineral products under the range of physicochemical conditions expected in a repository in basalt. Experimental radionuclide distribution coefficients were determined for crushed Pomona, Flow E, and Umtanum basalts at 23°, 60°, 150°, and 300°C at both normal oxygen partial pressure (~0.2 atm) and lower oxygen partial pressure (~10{sup -7} atm), using a static technique. Little or no changes in distribution coefficients were noted for selenium, uranium, technetium, neptunium, or plutonium over the oxygen partial pressure range noted above. Sodium dithionite and hydrazine are now under study as system additives to lower Eh to -0.3 to -0.5 V, the conditions expected to prevail in the closed repository in basalt. Radium, strontium, cesium, and americium are not expected to change oxidation states under repository conditions, while iodine remains an anion in either oxidation state. Lowering the system Eh to the -0.3 to -0.5 V expected in a repository in basalt should result in an oxidation state change and enhanced removal from solution for selenium, uranium, technetium, neptunium, and plutonium. Sorption of iodine was not affected by the Eh changes. Temperature change effects on most radionuclide distribution coefficient (Kd) values over the 23° to 300°C range were major with the exception of iodine and technetium, neither of which were appreciably sorbed at normal to ~10{sup -7} atm oxygen partial pressure. Uranium Kd values increased with an increase in temperature. In addition, uranium Kd values at 23°C decrease by an order of magnitude in response to added CO{sub 3}{sup 2-} in the solution. Cesium basalt Kd values decreased from 23° to 150°C and increased from 150° to 300°C. Americium and plutonium Kd values increased from 100 to 200 ml/g at 23°C to several thousand ml/g at 150°C. Strontium Kd values reacted to temperature increases in an individualistic response that apparently depended upon the basalt contacted. Flow E basalt showed little strontium Kd change between 23° and 150°C, while an order of magnitude strontium Kd increase was noted for Umtanum basalt over the same temperature range. Selenium Kd values increased from ~5 ml/g at 23°C to 100 ml/g at 150°C. The effect of radionuclide concentration on the Kd value was shown graphically for cesium and strontium over a range of from 1 x 10{sup -10} or 10{sup -12} to 1 x 10{sup -4}M. Molarity was plotted versus Kd on log scales. The Kd values remained linear with increasing cesium or strontium concentration until ~12 x 10{sup -7} concentration were attained. Above 1 x 10{sup -7}, the Kd values decreased. When the natural log of equilibrium solution concentration (moles/liter) was plotted versus the natural log of equilibrium solids loading (moles/gram) the strontium sorption data were linearized in a Freundlich plot. The cesium sorption data, on the other hand, were linearized by the Dubinin-Radushkevich relationship. The use of {sup 99}Tc to study technetium concentration effects on sorption was unsuccessful. Technetium-95m will be utilized for this purpose in the future. Initial work was begun on Kd values obtained under controlled Eh and pH conditions to simulate specific oxygen partial pressure and pH conditions expected to occur in the repository environment. Eh values from +0.60 to -0.55 V and pH values of ~6 to 10 are expected over a period of time in the repository. Hydrazine and sodium dithionite are under investigation for use in Eh control at the lower end of the Eh range and quinhydrone for intermediate Eh range. The upper end of the pH range also can be duplicated with hydrazine and the lower end with sodium dithionite.