A combined experimental and modelling study of the sorption of uranium and technetium on geological materials has been carried out as part of the PNC programme to increase confidence in the performance assessment for a high-level radioactive waste (HLW) repository in Japan. Batch sorption experiments have been performed in order to study the sorption of uranium and technetium onto bentonite, tuff and granodiorite from both equilibrated seawater and de-ionised water under strongly-reducing and non-reducing conditions.

A preliminary study of the sorption of uranium on mineral surfaces in granodiorite has also been undertaken using a nuclear microprobe.

Mathematical modelling using the geochemical speciation program HARPHRQ in conjunction with the HATCHES database has been carried out in order to interpret the results of the sorption experiments.

The pH dependence of the stability constant of Np(V)-humate and the sorption of Np(V) onto soil was studied to clarify the influence of dissolved organic carbon on the migration behavior of Np(V) in soil layer. The stability constant of Np(V)-humate was expressed by logβ1=(0.35 ±0.03)pH+0.04±0.01 in the pH region from 5.3 to 8.7, and the intrinsic stability constant by logβ*1=3.66±0.05.

The sorption of Np(V) increased with pH up to pH 7, and tended to reach maximum, about 95%, at pH 9 in the absence of humic acid. In the presence of humic acid more than 140mg/l, the sorption of Np(V) was lower than that in the absence of humic acid above pH 7, while any influence of humic acid on the sorption of Np(V) onto soil was not observed below pH 7. The sorption behavior of Np(V) onto soil was explained by reversible sorption of NpO2+ and sorption of NpO2CO3- and/or NpO2OH. Further, the influence of the size distribution of Np(V) on the sorption onto soil was examined and discussed.

The absorption of the technetium on rocks and minerals have been studied by using freshly quarried rocks to provide reducing conditions, without the addition of reductive reagents. The initial 99Tc concentration in this study was about 1x10-8 mol/1 (0.62 Bq/ml). Technetium was not absorbed on tuff or granite under oxidizing conditions, but higher Kd values of Tc were obtained for granite under reducing conditions. The Tc absorbed on granite was not desorbed under reducing condition. This indicates the absorption reaction is irreversible.

Sorption experiments have been performed by a batch method, to study the effects of humic acid of different molecular size on the complexing stability with 60Co, 85Sr, 137Cs, 237Np and 241Am, and on the sorption behavior of these radionuclides on a sandy soil.

Equilibrium constants K in the sorption of 137Cs and 237Np onto the soil were not changed at different concentrations of humic acid since 137Cs and 237Np do not interact with humic acid, while those of 60Co and 241Am decreased with increasing humic acid concentration due to forming humic complexes. However, the K of 85Sr was not changed at different humic acid concentrations, despite 85Sr interacts with humic acid. This contradiction was probably caused from that a main binding of 85Sr with humic acid is not based on coordination bond but electrostatic force, due to relatively high concentration of non–radioactive strontium. The theoretical sorption model taking account of the interaction of 60Co and 241Am with humic acid could well reproduce the values of K for each radionuclide at different concentrations of humic acid.

Concentration profiles of the radionuclides in each size fraction of the solution before and after the sorption experiments were examined by ultrafiltration technique. The reduction of concentration of 60Co in the fraction less than 300,000 of cutoff molecular weight (MW) and that of concentration of 241Am in the fraction larger than 100,000MW, respectively, by the sorption onto the soil decreased with increasing humic acid concentration. This decrease resulted in the decrease in the K of 60Co and 241Am with increasing humic acid concentration.

The sorption behavior of neptunium on bentonite was studied with batch type sorption and desorption experiments over a pH range of 2 to 8. A series of parallel studies using Na-smectite, Ca-smectite and admixtures of Na-smectite and calcite quantified the capacity of Ca2+ (which occurs in bentonite as an exchangeable cation of smectite and as a component of calcite) to inhibit the sorption of neptunium. The distribution coefficient (Kd) of neptunium for bentonite was constant from pH 2 to 7, while for pure Na-smectite Kd increased below pH 5 due to “specific sorption” of neptunium on Na-smectite. Specific sorption was defined as occurring when neptunium could be desorbed by a strong acid (1 M HCl) but was stable in the presence of 1 M KCl. It was found that the quantity of neptunium sorbed on Na-smectite was inversely proportional to the concentration of Ca2+ in solution, an effect most pronounced at pH <5. These results show that Ca2+ limits the specific sorption capacity of Na-smectite for neptunium. Similarly, in the mixture of Na-smectite and calcite, sufficient Ca2+ was solubilized to depress neptunium sorption. This investigation demonstrates that Ca2+ contained in bentonite as exchangeable cation and released from calcite reduces the specific sorption of neptunium.

The sorption experiments of carbon-14 on the mortar grain (grain size : 0.50 – 1.0 mm) focused on the chemical form of the carbon-14 were carried out by the batch method. Three kinds of carbon-14 chemical form were used for the experiments : sodium carbonate (Na214CO3) as the inorganic radiocarbon, and sodium acetate (CH314COONa) and acetaldehyde (14CH314CHO) as the organic radiocarbons. 0.30 gram samples of mortar were soaked in the solution with carbon-14 at 15°C for periods of up to 160 days. At the end of each run, carbon-14 concentrations in the supernatants were determined before and after centrifugation (3,500 rpm., 1 hr.).

In the mortar-sodium carbonate system, the retention process of carbon-14 related to reaction on the surface of the mortar was speculated as follows. First, 3CaO-SiO2. and 2CaO-SiO2, of the mortar components contact with water and produce Ca(OH)2. Ca(OH)2 produces Ca2+ and OH− in the solution. Then, calcite forms from Ca2+ and CO32− in the solution. Thus, the sorption ratio of carbon-14 onto mortar will be high until mortar nas been completely carbonated because Ca2+ is rich in the mortar and the solubility of calcite is low.

In the mortar-organic carbon system, the soluble organic carbon-14 is hardly sorbed on the surface of the mortar. Therefore, the cementitious materials may not inhibit the release of organic radiocarbons from the low-level radioactive wastes, contrary to the case of inorganic radiocarbon.

The I– adsorption on a modified mineral, which is prepared by impregnating Ag(l) nitrate complex with thiourea(tu) into montmorillonites, have been studied by batch method. The modified mineral was found to adsorb I– ions selectively even under the coexistence of Cl– ions. The distribution ratio(Kd) is 5000 ml/g under the conditions of initial [I–] = 1.0 × 10−5 M(M = mol/dm3), pH = 6.4, and solution/modified mineral = 100 ml/g. It was found that the [Ag(tu)]I precipitates are formed by mixing Nal solution with Ag(l)-tu complex solution, while any precipitates are not formed in addition of NaCl solution, and that anion exchange property of the modified mineral is low. Furthermore, it was confirmed that the [Ag(tu)]I complexes are present in the modified minerals after adsorbing I– ions. From these results, the I– selective adsorptivity of the modified mineral was considered to be due to the formation of [Ag(tu)]I precipitates within the interlayer region of montmorillonite.

We have carried out a comparative study of sorption and desorption of strontium in groundwater on separated size and magnetic fractions of fracture fillings from deep granite. Complete reversibility of the sorption process was demonstrated by identical Freundlich isotherms, isotopic exchangeability and pH dependence of the distribution coefficient Rd.

The sorption of Sr and Cs has been investigated on Ca-Mg-bentonite preliminary treated by 0.1–1.0 m NaCl solutions at 25, 170–250°C during 0.5–7 months. The bentonite remained stable in these conditions. Moreover, hydrothermal treatment has raised the sorption activity of the bentonite. The treated bentonite has shown a strong selectivity for Sr and less pronounced one for Cs. The sorption of the elements increased with temperature and decreased with NaCl content more significant for Sr, than for Cs. The distribution ratio Rd was nearly independent on ion loading for Cs and strongly increased up to 104.5 mL/g in 10−2.0−10−0.5 meq/g ion loading interval for Sr, dropping sharply at more concentrations.

Results of zeta potential measurements by electroosmosis with site-specific waters and sedimentary rocks from strata overlying the Gorleben salt dome are given and compared with sorption data of radionuclides obtained in batch experiments. It is shown that zeta potentials of sediment-groundwater systems in spite of being lower than the corresponding surface potentials are a sensitive indicator of interface reactions depending on the type of the sediment, ionic strength of waters, pH and temperature. Moreover, it is demonstrated that sorption data and, additionally, the influence of various parameters can often be understood by taking into account variable electrostatic forces between charged surfaces and nuclide species.

The variation of the distribution ratio Rd as a function of aqueous volume to solid mass ratio (V/M) for several sorption systems involving the cations Zn2+, Co2+, Ba2+, the anion I− and different solid matrices has been extensively examined. Empirical relationships for the concentrations of the sorbed species in the aqueous and solid phases as well as for the distribution ratio Rd are proposed. Generally Rd is found to increase with increasing V/M for tracer concentrations. A gradual increase in the initial concentration of the sorbed species eventually leads to a decrease in Rd with V/M.

The possibility to use alkali metals and alkaline earth metals as slightly sorbing tracers in in-situ sorption experiments in high saline groundwaters has been investigated. The cation exchange characteristics of granite and some fracture minerals (chlorite and calcite) have been studied using the proposed cations as tracers. The results show low Kd’s for Na, Ca and Sr (∽0.1 ml/g), while the sorption is higher for the more electropositive cations (Rb, Cs and Ba). A higher contribution of irreversible sorption can also be observed for the latter group of cations. For calcite the sorption of all the tracers, except Ca, is lower compared to the corresponding sorption to granite and chlorite. Differences in selectivity coefficients and cation exchange capacity are obtained when using different size fractions of crushed granite. The difference is even more pronounced when comparing crushed granite to intact granite.

Surface complexation model was used to describe ytterbium sorption on kaolinite using surface acidity constants and surface complexation constants determined on alumina and silica.

A simple closed column system was developed to study migration of bicarbonate and carbonate 14C in soil layer. Experimental data for glass beads having less the 14C adsorption indicate that the system can be performed on column experiment keeping mass balance of 14C. Migration behavior of bicarbonate and carbonate 14C in a tuffaceous sandy soil was investigated by this column system in the pH range from 8 to 12. The 14C was delayed to tritiated water, and the retardation became stronger in order, pH 12 < pH 10 < pH 8. This is considered to be due to pH-dependent adsorption of bicarbonate and carbonate ions onto the soil sample.

A five-year cooperative research program has been carried out by China Institute for Radiation Protection and Japan Atomic Energy Research Institute to develop safety assessment methodology for disposal of low level radioactive waste. Migration behavior of radionuclides 3H and 85Sr through field tests, simulation tests in the laboratory and determination of distribution coefficients is discussed and analyzed in this paper. The results show that the retardation coefficients, Rd, from field tests are about 0.08−3.3 times those from simulation tests and Rd from batch tests are 1.1−44 times those from field tests for 85Sr and loess medium. It was observed from field tests that radionuclides moved mainly downward under artificial sprinkling and a part of them moved up besides downward under natural rain condition. In addition, it was discovered that the retardation coefficient, Rd , increases with velocity of unsaturated water flow, u, in the analysis.

Column experiments have been completed on the migration of 85Sr, 134Cs and 60Co through a loess layer in order to examine the migration behavior of radionuclides in an aerated soil layer. Radionuclide concentration distributions between the effluent and the soil layer were measured after the solution containing the radionuclides was introduced into the column from the top of the soil layer and Fifty liters of the underground water were introduced at a constant flow. Results indicate most of the 85Sr, 134Cs and 60Co remained attached to the soil layer, and only a small amount of radionuclide was released from the soil layer. Within the soil layer, the migration depths of three radionuclides are 85Sr > l34Cs = 60Co.

Safety assessments of nuclear waste repositories often require estimation of actinide solubilities as they vary with groundwater composition. Although a considerable amount of research has been done on the solubility and speciation of actinides,1,2 relatively little has been done to unify these data into a model applicable to concentrated brines. Numerous authors report data on the aqueous chemical properties of Np(V) in NaClO4, Na2CO3, and NaCl media, but a consistent thermodynamic model for predicting these properties is not available. To meet this need, a model was developed to describe the solubility of Np(V) in Na-Cl-ClO4-CO3 aqueous systems, based on the Pitzer activity coefficient formalism for concentrated electrolytes. Hydrolysis and/or carbonate complexation are the dominant aqueous reactions with the neptunyl ion in these systems. Literature data for neptunyl ion extraction and solubility are used to parameterize an integrated model for Np(V) solubility in the Np(V)-Na-CO3-HCO3-Cl-ClO4-H-OH-H2O system. The resulting model is tested against additional solubility and extraction data, and compared with Np(V) solubility experiments in complex synthetic brines.

The complexation behavior of Np(V) with humate was investigated by spectrophotometry in the pH range from 5.7 to 10 in 0.01–0.5M NaClO4. The absorption peaks of NpO2+ and Np(V) humate were observed at 979.6nm and 989.9nm, respectively, and the apparent complexation constant of the Np(V) humate was found to be pH-dependent. A two-site complexation model was applied to the interpretation of the observed pH-dependence by considering the participation of the two functional groups of carboxylate and phenolate. The dissociation and complexation constants of both functional groups were determined.

The solubility of UO2 (s) was examined in dilute NaCl solutions at room temperature and in the pH range from 2 to 12. Dissolution equilibrium showed good agreements between the data obtained by oversaturation and undersaturation experiments.

The dissolution reactions that control the solubility of U[IV] are estimated as UO2 (s) + 4H+ → U4+ + 2H2O and UO2 (s) + 2H2O → U(OH)4°. The log K at zero ionic strength are yielded 0.34 ± 0.4 and −8.7 ± 0. 4, respectively. The hydrolysis constant (log β14) is calculated −9.0 ± 0.5 for the hydrolysis reaction of U4+ + 4H2) → U(OH)4° + 4H+. Contribution of the other uranium species to the solubility are of minor importance.

The crystallization of the precipitations of U02 (s) progressed gradually in the oversaturation experiments. Nevertheless the progress of crystallization had little effect on the solubility.

The solubility of Th(IV) and U(IV) hydrous oxide was studied in the aqueous HCO3−-CO32−-OH−-H2O system extending to high concentration. The solubility of the Th(IV) and U(IV) hydrous oxides increases dramatically in both high bicarbonate and carbonate solutions and decreases with the increase in hydroxide at a fixed carbonate concentration. In general, the observed solubility of Th(IV) hydrous oxide at a given total carbonate concentration was three to four orders of magnitude higher than the solubility of U(IV) hydrous oxide. In the studies of the U(IV) system, extreme caution was used to ensure that the dissolved U was present as U(IV). Oxidation state analyses and systematic trends in the U(IV) solubility data similar to those for Th(IV), which is not redox sensitive, indicated that the dissolved U was present as U(IV).