Band 90, Heft 9-11 - Migration 2001

The solubility of amorphous Th(IV) hydroxide at pH 3.0–13.5 and the aqueous speciation at pH < 4 are investigated in 0.5 M NaCl and 25 °C. The laser-induced breakdown detection (LIBD) is used to monitor the initial formation of thorium hydroxide colloids during the coulometric titration of 1.2×10 −2 −1.0×10 −5 M thorium solutions in the pH range of 2.7–4.5. The accurate solubility limit determined by this method is comparable with data measured from undersaturation with an X-ray amorphous solid precipitated at higher pH and dried at room temperature. Based on hydrolysis constants selected from the literature, the solubility product of Th(OH) 4 (am) in 0.5 M NaCl is calculated to be log K′ sp = −44.48 ± 0.24 and log K° sp = −47.8 ± 0.3 (converted to I =0 with the SIT coefficients of the NEA-TDB). In other solubility studies with amorphous Th(IV) hydroxide or hydrous oxide, considerably higher thorium concentrations are measured at pH 3.5–5. Therefore, solutions of comparable H + and thorium concentrations are prepared by careful coulometric titration and examined by ultrafiltration, LIBD and X-ray absorption fine structure (XAFS) spectroscopy. These measurements demonstrate the presence of a large amount of small Th(IV) colloids. The Th L 3 edge EXAFS spectra of these colloidal suspensions are similar to that of the amorphous solid.

The solubility of synthetic becquerelite (Ca(UO 2 ) 6 O 4 (OH) 6 ·8H 2 O) was determined in 0.02, 0.1, and 0.5 M CaCl 2 solutions and at p C H+ values ranging from approximately 4 to 11. The presence of becquerelite in equilibrated samples was confirmed by a combination of techniques involving X-ray diffraction, total chemical composition, and analyses of solubility data. The solubility data were interpreted using Pitzer's aqueous thermodynamic model and the thermodynamic data for U(VI) species available in the literature. The log of the solubility product for becquerelite [Ca(UO 2 ) 6 O 4 (OH) 6 ·8H 2 O + 14H + ⇌ Ca 2+ + 6UO 2 2+ + 18H 2 O] was determined to be 41.4 ± 0.2. This value is similar to the values previously reported for other synthetic becquerelites, but is drastically different from a value reported for a natural sample.

Solubility data for dissolution of plutonium dioxide and tetrahydroxide in laboratory and natural waters are compared with results of modeling calculations. Evaluation shows that selection of metastable Pu(OH) 4 as the solubility-controlling solid results in predicted steady-state plutonium concentrations that are not conservative and oxidation-state distributions that are inconsistent with observation. Although only the presence of Pu(III) and Pu(IV) is predicted at measured solution potentials, Pu(V) and Pu(VI) predominate (>95%) in steady-state solutions at pH 1–9. Behavior is inconsistent with thermodynamic and kinetic constraints of the conceptual chemical model and a strong dependence of experimental equilibrium expressions on the amount of solid hydroxide in the steady-state solution suggests that dissolution of plutonium oxide is not an equilibrium process.

Technetium migration was investigated apatite a possible nuclear wastes conditioning materials. 99m Tc was introduced into fluoroapatite by ion implantation. The analysis of implanted profile evolution under samples treatment allows to identify the migration process as a bulk diffusion. The observed Tc release is correlated to the oxidation of Tc by interaction with air at the sample surface.

The release of natural uranium from Boom Clay was studied to better understand the mechanisms governing the solid-liquid partitioning of uranium. Batch leaching experiments suggested that the portion of natural uranium released from clay is associated with colloids at a low bicarbonate concentration prevailing in Boom Clay. At increased bicarbonate concentrations, uranium was present predominantly as dissolved species indicating a formation of uranium carbonate complexes. The in situ aqueous uranium concentration, i.e ., the concentration in the pore waters collected by piezometers was found to be 2 to 3 orders of magnitudes lower than the one measured by the batch techniques. These results illustrated that the batch techniques may cause a remobilization of uranium containing colloids from clay surfaces into solution when clay is suspended, agitated, and mechanically perturbed. These colloids are attributed to artefacts and are not considered to exist in situ because of the high compaction of Boom Clay. Due to the presence of colloids, a laboratory derived solid-liquid partitioning coefficient is not equivalent to and cannot simply be converted to the distribution coefficient K d currently used in performance assessment calculations.

Coprecipitation may be a significant process in controlling radionuclide (RN) release during spent fuel dissolution/reprecipitation in a breached geological disposal site by water. The objective of this work was to study the coprecipitation of thorium with UO 2 as host phase under reducing conditions (pure N 2 and E h <-100 mV/SHE) as a function of pH and [Th]/[U] initial concentration ratio. For pH < 4, the final concentration of Th was found to be considerably lower than expected if pure phase, Th(OH) 4 (s) was formed. Coprecipitation phenomena and formation of solid solutions (U,Th)O 2+x (s) are suggested to be responsible for the lower solubility concentrations. For pH > 4, the final concentrations seems to be similar to pure phase solubility.

The behavior of rare earth elements (REE) as chemical analogues for actinides during glass corrosion was studied with static long-term batch experiments (7.5 years) at 190 °C. Corrosion tests were carried out using a simulated inactive high level waste (HLW) glass powder. Two different highly concentrated salt solutions (NaCl-rich and MgCl 2 -rich) have been used to simulate the potential corrosion by aqueous solutions in geological salt formations. Samples with surface-to-volume ratios (S/V) of 1000 m −1 and 10000 m −1 were investigated. The chemical composition of the aqueous solution and the formation of secondary phases at the glass/solution interface as well as their elemental composition, especially with respect to the behavior of REE was studied. The glass corrosion behavior and secondary phase formation was significantly different in the two brines. In the MgCl 2 -rich salt solution, with a final pH < 3.5, hydrous Mg-silicates and (Ba,Sr)SO 4 solid solutions with considerable amounts of Sr (up to 15 wt.%) have been observed. Further, REE-sulfates and REE-molybdates with an indication of dissolution (etch pits) as secondary phases could be identified. In the NaCl-rich solution with a final high pH, REE-containing hydrous silicates (up to 35 wt.%) and (Ba,Sr)SO 4 solid solutions with small amounts of Ce (up to 5 wt.%) have been observed. Frequently observed Ca-molybdates (powellite) did not contain detectable amounts of REE.

In the alteration layer of HLW glasses corroded in aqueous solutions, less soluble elements like REE, actinides and other elements are retained by sorption or by formation of secondary alteration products controlling the solution concentrations. In order to quantify the sorption effect on the retention of metal ions, first sorption tests with U, Th and Eu have been performed using precorroded, simulated HLW glass as substrate. To demonstrate that the sorption behaviour of Eu is also representative for homologue trivalent actinides additional sorption tests were performed with Am 241 in deionized water, NaCl and MgCl 2 brine. Distribution ratios R s ( R s =Am sorb / Am sol × V/m) for Am were found independent of the starting concentration in the pH range 2 to 6. At higher pH, a high amount of Am-containing colloidal particles were detected. The R s values of Am show a strong increase with pH, quite similar to the R s values of Eu. In general, the sorption behaviour of Eu and Am is very similar in deionized water and in brines. Clay minerals formed in the gel layer as secondary phases during long-term glass corrosion are assumed to act primilary as strong sorbents. Therefore, the sorption of REE and Am onto synthetic poor crystalline smectite precursors and a "well crystallized" natural smectite is compared with the findings obtained from precorroded glass. The sorption of Am on the precorroded glass is found to be much stronger than the retention on the investigated pure clay phases, indicating a different number of active sites or occurance of a different sorption mechanism.

Humic acid (HA) was fractionated by gel permeation chromatography and the complexation of the obtained three fractions with UO 2 2+ was examined by synchronous fluorescence spectroscopy combined with SIMPLISMA, which is one of the mixture analysis methods. Most of the fluorescence of HA resulted from the lowest molecular weight fraction, and the quenching behavior of this fraction was similar to that of unfractionated HA. After the spectral resolution by SIMPLISMA, two components, which were differently quenched with UO 2 2+ , were extracted for both cases. The complexing parameters, the apparent stability constant and the concentration of complexing sites, were evaluated for each component and compared between unfractionated HA and the lowest molecular weight fraction.

Three new uranium(VI) compounds have been prepared with glycine (Hgly), N -(2-mercaptopropionyl)glycine (Hmpg) and 2,2′-dithiobis( N -propionylglycine) (tpg). The crystal structure of [UO 2 (Hgly) 2 Cl 2 ] has been solved, revealing the uranium atom to be seven-co-ordinated. The two zwitterionic glycines are coordinated to uranium atom by the carboxylate groups, one in a monodentate and the other in a bidentate mode. Solid state studies and low temperature 1 H, 13 C, 2D COSY and HMQC solution NMR spectra of the [UO 2 (mpg)(CH 3 COO)], show that mpg coordinates uranium atom by the carboxylate group.

In order to understand the migration behavior of radionuclides in the underground, it is also important to estimate the effect of the competing cations originally present in the groundwater. In this connection, the complexation of Ca(II) with Aldrich humic acid has been examined. For the study at trace concentrations (∼10 −10 M) of Ca(II), the solvent extraction of 45 Ca with TTA and TOPO in cyclohexane has been used. At macro concentrations (10 −4 M) of Ca(II), the measurement of the free Ca 2+ ion concentration with a calcium selective electrode has been conducted. To estimate the polyelectrolyte effect of humic acid separately from its heterogeneous composition effect, polyacrylic acid ([-CH 2 CH(COOH)-] n ) has been selected as a representative of the homogeneous polymeric weak acids and its complexation with Ca(II) has also been examined. The values of logβ α have been obtained at pH 5∼7 in 0.1, 0.4 and 1.0 M NaCl, where β α is the apparent formation constants defined by β α =[ML]/([M][R]). In this definition, [ML] and [M] are the concentrations of bound and free Ca 2+ respectively, [R] is the concentration of dissociated proton exchanging sites. logβ α of humate decreases from 2.19∼2.92 (depending on pH and ionic strength 1.0 < I < 0.4) at pCa= 10 to 1.98∼2.44 at pCa = 4, while the variation of pCa has no appreciable influence on the logβ α of polyacrylate (1.36∼3.24 for I =0.1∼1.0). For both humate and polyacrylate, logβ α decreases linearly with log[Na + ], where [Na + ] is the bulk concentration of sodium ion. Their dependences of logβ α on ionic strength are stronger than those of logβ of monomeric carboxylates such as oxalate and EDTA, indicating the large electrostatic effect of crowded negative charges on the macromolecules of humic or polyacrylic acid.

The stability constants of the Am 3+ complexes of humic, fulvic and citric acids (HA, FA and Cit) were determined as a function of ionic strength (NaCl) using a solvent extraction technique. At a HA degree of ionization of 0.7 for the carboxylate groups, the Am-HA binding constant, logβ 1 , varies from 8.3 at I = 0.2 m to 7.2 at I = 6.0 m. The logβ 1 of Am-Cit varies from 5.9 at I = 0.3 m to 5.10 at I = 5.0 m. Comparison of the binding constants of Am with HA, FA and Cit shows that, at high ionic strength, logβ 1 (AmCit) ≈ logβ 1 (AmFA) ≈ logβ 1 AmHA-1. Comparison of the values of other metals indicate that the logβ 1 values for Cit can serve to estimate HA and FA.

With the intention of estimating the effect of the heterogeneous composition in the humic acid separately from its polyelectrolyte effect, comparative study on the interaction of Eu(III) with humic acid and polyacrylic acid has been carried out by the solvent extraction method with 10 −3 M TTA and 10 −3 M TBP in xylene at various total concentrations of Eu(III) (10 −8 M to 10 −5 M), p cH (5.0 to 5.5) and ionic strengths (0.02 to 1.0 M NaClO 4 ). The interaction has been discussed in terms of the apparent complex formation constants, β α = [ML]/([M][R]), where [M] and [ML] are the concentrations of free and bound Eu 3+ ion and [R] is the concentration of dissociated proton exchanging site. For both humate and polyacrylate, logβ α increased with p cH or the degree of ionization and decreased with ionic strength. The Eu 3+ ion concentration had no appreciable influence on the complex formation of Eu(III)-polyacrylate, and the values of logβ α were obtained to be 7.57 ± 0.05 ( I = 0.1, p cH = 5.53), 7.13 ± 0.04 ( I = 0.1, p cH = 5.20), 6.68 ± 0.04 ( I = 1.0, p cH = 5.54) and 5.61 ± 0.09 ( I = 1.0, p cH = 4.73). On the other hand, logβ α of humate remarkably decreased with the Eu 3+ ion concentration; from 9.32 ([Eu 3+ ] = 10 −11.9 M) to 7.20 ([Eu 3+ ] = 10 −7.5 M) at p cH = 5.53 in 0.02 M NaClO 4 , from 8.63 ([Eu 3+ ] = 10 −12.5 M) to 6.35 ([Eu 3+ ] = 10 −7.2 M) at p cH = 5.53 in 0.1 M and from 7.59 ([Eu 3+ ] = 10 −10.7 M) to 5.20 ([Eu 3+ ] = 10 −7.0 M) at p cH = 5.52 in 1.0 M. The observed dependence of logβ α of humate on the concentration of metal ion strongly suggests the heterogeneity of the complexing sites (the coexistence of weak and strong binding sites) in the humic acid.

X-ray absorption spectroscopy has been used to establish polymer formation of Tc(IV) in aqueous solutions of Na + SO 4 2− and Na + Cl − /SO 4 2− . As the molybdenum chemistry show similarities to that of technetium, we used MoO 2 as a reference to model our technetium species. Fitting of TcO 2 · x H 2 O with this model led to a good correlation with the literature data: (Tc-Tc=2.53 Å, Tc-O=1.87-1.98 Å). In aqueous solution, some polymers are formed regardless to the nature of the media composition: Tc-Tc=2.50 ± 0.02 Å. The general structure is in agreement with a first coordination shell containing 6 O. The modelling shows that, in the first coordination shell, there is no chloride ligand. The observed geometries are close to those found for TcO 2 · x H 2 O, hence the unknown aqueous species must be considered as a precursor of the solid technetium dioxide. Combination of these results with XANES led to attribute Tc n IV O p (4n-2p)+ (H 2 O) q with n>2 to the species.

The complexation between uranium(VI) and malonate in 1.05 mol kg −1 NaClO 4 was studied at variable temperatures (25, 35, 45, 55 and 70 °C). The formation constants of three successive complexes, UO 2 (OOCCH 2 COO), UO 2 (OOCCH 2 COO) 2 2− and UO 2 (OOCCH 2 COO) 3 4− , and the molar enthalpies of complexation were determined by potentiometry and calorimetry. The heat capacity of the complexation, Δ C o p,m(MLj) , is calculated to be 96 ± 12, 195 ± 15 and 267 ± 22 J K −1 mol −1 for j=1, 2 and 3, respectively. Extended X-ray Absorption Fine Structure Spectroscopy helped to characterize the coordination modes in the complexes in solution. UV/Vis absorption and luminescence spectra at different temperatures provided qualitative information on the temperature effect. The effect of temperature on the complexation between uranium(VI) and malonate is discussed in terms of the electrostatic model and compared with the complexation between uranium(VI) and acetate.

The hydrolysis of uranium(VI) has been investigated in 0.10 mol dm −3 NaClO 4 and KCl and 1.0 mol dm −3 KNO 3 using the potentiometric technique. The results of the data analysis have been combined with data previously obtained from this laboratory and the literature to provide a more definitive description of the hydrolysis behaviour of uranium(VI). The species UO 2 OH + , (UO 2 ) 2 (OH) 2 2+ , (UO 2 ) 3 (OH) 4 2+ , (UO 2 ) 3 (OH) 5 + and (UO 2 ) 4 (OH) 7 + have been identified in nitrate media and all but the last of these species in chloride and perchlorate media. The differences in the complexation behaviour in the media may be due to oxoanion binding by the nitrate, a process which has also been observed for sulfate.

Corrosion and dissolution of UO 2 doped with short-lived α-emitters, the so called α-doped UO 2 , was investigated in leaching experiments combined with X-ray Photoelectron Spectroscopy (XPS) analyses. These experiments are part of an ongoing set of tests aiming at evaluating the effects of α-radiolysis on the dissolution of the UO 2 spent fuel matrix when it will become exposed to groundwater in a deep final repository. Pellets of UO 2 containing ∼10 and ∼0.1 wt.% 238 Pu were used. These compositions simulate the α-radiation field of different types of commercial LWR spent fuel after different storage times. Undoped UO 2 pellets were also used as reference. Static sequential leaching experiments in demineralised water at room temperature under anoxic conditions confirmed that higher fractions of uranium are released in the case of α-doped materials compared with undoped UO 2 . Relatively large amounts of plutonium were found in the leachate of the materials doped with 238 Pu. Significant contributions to the total amount released of uranium were found in the rinse solutions of the leaching vessels. The evolution of the U(VI)/U(IV) ratio on the leached surfaces as a function of contact time was studied by deconvolution of the 4f 7/2 uranium peak obtained from the XPS spectra. In the case of undoped UO 2 , after ∼1000 h of leaching the increase of the U(VI)/U(IV) ratio stabilized to a level corresponding to ∼UO 2.33 . The XPS results for both α-doped materials showed a continuing increase of the U(VI) fraction on the surface with increasing leaching times after the first ∼10 h. The rate of increase depended on the α-activity of the fuel sample, hence on the amount of radiolysis in the water. After ∼1000 h of leaching, essentially only U(VI) was detected on the surface of the high activity α-doped UO 2 . The presence of a yellowish layer of reprecipitated U(VI) phases covering the surface could be observed optically already after ∼100 h of leaching.

The long-term stability of spent nuclear fuel under deep geologic repository conditions will be determined mostly by the influence of α-radiolysis, since the dose-rate for α-radiolysis will exceed that for γ/β-radiolysis beyond a fuel age of ∼100 years and will persist for more than 10000 years. Dissolution rates derived from studies with currently available spent fuel include radiolysis effects from γ/β- as well as α-radiolysis. The use of external α-sources and chemically added H 2 O 2 in UO 2 studies has provided much information on the basic chemistry/electrochemistry of α-radiolysis effects on UO 2 fuel dissolution. Here, alpha-radiolysis effects were measured directly by performing electrochemical experiments with α-doped UO 2 . The 238 Pu-doped UO 2 pellets, produced by milling, pressing and sintering, had activities of approximately 0.01 and 0.1 Ci/g. Discs cut from these pellets were fabricated into electrodes. Corrosion potentials (E corr ) were measured for these electrodes, in glass cells containing de-aerated 0.1 M NaClO 4 (pH 9.5) for periods up to 1600 h. The effects of carbonate and Fe(II) on E corr values were also investigated. Preliminary results showed that E corr values correlated with doping levels, and were slightly higher than E corr values obtained with UO 2 electrodes in the presence of external α-sources. The presence of carbonate reduced E corr values considerably. Addition of Fe(II) caused generally small reductions in E corr values, suggesting that Fe(II) affects the dissolution process, either by scavenging radiolytic oxidants, reducing oxidized surface states, or reducing dissolved U(VI). These effects were often transient due to depletion of the added Fe(II). Accurate dissolution rates could not be determined from measured solution and surface concentrations because of material entering the solutions during the electro-reduction procedure, probably due to the dissolution of fines present on the electrode surface after polishing. Experiments are continuing in order to confirm the preliminary findings, with greater attention being paid to determining accurate dissolution rates.

Technetium(IV) oxide colloids were radiolytically formed by γ irradiation of aqueous solutions of pertechnetate (TcO 4 − ). Pertechnetate solutions (5.5×10 −5 -2.9×10 −4 M) were irradiated with bremsstrahlung from an electron linear accelerator at 40 and 17 °C. The color of irradiated solutions gradually changed to brownish black, suggesting the formation of Tc(IV) oxide colloids (TcO 2 · n H 2 O). A transmission electron microscopy (TEM) analysis showed that the size of colloids distributed around 30 to 130 nm in diameter. The characteristic X-rays from technetium and oxygen were simultaneously detected from colloid particles at the TEM measurements. Round-shaped colloids were produced by irradiation at 40 °C, whereas irregular-shaped colloid particles composed of tiny particles (2 nm in diameter) were produced at 17 °C. The concentration of TcO 4 − in the target solution gradually decreased with an increase of the absorbed dose, corresponding to an increase of the colloid yield. The yield of colloids sharply increased in the solution deaerated by Ar bubbling before irradiation, but strongly suppressed in the solution saturated with oxygen (O 2 ) or nitrous oxide (N 2 O) gas. The fact suggests that hydrated electrons play an important role in the course of the reduction of TcO 4 − and that Tc(IV) oxide colloids were formed via successive disproportionation reactions of Tc(VI) and Tc(V). The formation mechanisms of Tc(IV) oxide colloids are discussed.

The oxidation of Br − by Cl 2 − is investigated by gamma pulse radiolysis in aqueous solutions of NaCl and NaBr. Depending on the ratio of the concentration of Br − to Cl − , the main product being observed is either Cl 2 − , ClBr − or Br 2 − . The mixed radical anion ClBr − exhibits a broad absorption band at 350 nm with ε 350 =9300 dm 3 mol −1 cm −1 . The rate constants of the equilibrium Cl 2 − + Br −  ClBr − +Cl − are determined to be k f =4×10 9 dm 3 mol −1 s −1 and k b =1.1×10 2 dm 3 mol −1 s −1 . The formation of the Cl 3 − (λ max =220 nm), Cl 2 Br − (λ max =230 nm) and ClBr 2 − (λ max =245 nm) ions in the radiation-chemical oxidation of Cl − and Br − ions in an aqueous solution was observed by pulse radiolysis, and its mechanisms of appearance and the equilibrium constants were determined.

The corrosion of iron and the interaction between corroded iron and U(VI) in anoxic conditions were investigated. The anoxic conditions were obtained by flushing an 99.97% Ar-0.03% CO 2 gas mixture through the test vessel, in which an oxygen trap and six reaction bottles containing synthetic groundwater (10 mM NaCl and 2 mM HCO 3 − .) were placed. The dark-green coloured corrosion product, formed on iron surface after three months corrosion in synthetic groundwater solutions, was identified by powder X-ray diffraction to be carbonate green rust, Fe 4 II Fe 2 III (OH) 12 CO 3 . The iron foil that reacted in a solution (10 ppm U(VI), 10 mM NaCl and 2 mM HCO 3 − ) for three months was analysed by SEM-EDS. The result shows that: (i) an uneven layer of carbonate green rust (1-5 µm thick) formed on the metallic iron; (ii) a thin (0.3 µm) uranium-rich layer deposited on top of the carbonate green rust layer; and (iii) some UO 2 crystals (3-5 µm sized) on the thin uranium layer. The experimental results proved that the U(VI) removal capacity of metal iron is not hindered by formation of a layer of carbonate green rust on the iron. Tests with cast iron and pure iron indicate that they have similar U(VI) removal capacities. At the end of experiment, U concentrations in solution approached the solubility of UO 2 (s), 10 −8 M. The stability of the carbonate green rust at the experimental conditions, pH, E h , [Fe 2+ ] and [HCO 3 − ], is discussed.

Selenium is a redox sensitive element. In reducing conditions its solubility is controlled by the formation of metallic Se and in the presence of Fe 2+ also by the precipitation of FeSe or FeSe 2 . However very few data concerning this species in geochemical reducing environments is found in literature, particularly due to insufficient measuring methods. The assessment to what extent 79 Se is a critical radionuclide for the geological disposal of High-Level Radioactive Waste, depends on its actual speciation in storage conditions. Therefore a new method based on ion chromatography of radiolabelled 75 Se solutions followed by gamma-ray counting was developed to accurately measure selenium species with different degrees of oxidation (selenate (SeO 4 2− ) and selenite (SeO 3 2− )) in solution. This method was then tested in laboratory conditions which mimic the reducing environment in Boom Clay. Different amounts of ground pyrite (< 125 µm) were contacted with Synthetic Boom Clay Water (essentially 10 −2 M NaHCO 3 ) and spiked with different amounts of 75 SeO 3 2− and SeO 4 2− . The batches were allowed to equilibrate over different time periods (up to two months) before analysing. The experiments were carried out in an oxygen-depleted glove box (99.6% N 2 , 0.4% CO 2 ). The kinetics of the redox reactions in the pyrite systems prevented the complete reduction of selenite (SeO 3 2− ) and especially selenate (SeO 4 2− ) on a limited time scale, probably due to the limited redox capacity of the studied systems.

Uptake experiments were carried out with selenate and chromate on fresh and leached Portland and high-alumina cements; and in addition with selenate and selenite on synthetic ettringite. In all experiments with cements, exceptionally high uptake could be observed under conditions where significant amounts of secondary ettringite were formed. Experimental data obtained for pure ettringite corroborated the important role of this mineral phase. However, uptake kinetics show opposite trends in these two systems, which can be viewed as end-members of the same process: Where a fast precipitation of secondary ettringite occurred, initial uptake was high, with K d values in the range of ≍1-5 m 3 /kg for Se(VI), but decreased with time. Uptake by pre-formed (primary) ettringite initially gave lower K d values (≍0.1 m 3 /kg), which increased with time. After prolonged equilibration times, the two systems started to approach each other. Concurrent measurements of sulphate concentrations allowed to extract a mean partition coefficient for the solid-solution formation of selenite and selenate with ettringite. Based on this simple solid-solution model, a pragmatic quantitative relation was developed that permits to estimate K d for Se(VI) on whole cement as a function of the concentrations of sulphate in the cement and in the solution. A test against experimental data shows reasonably good agreement between measurements and calculations. This relation can also be directly applied to estimate K d values of chromate (and perrhenate) by different cements, indicating the same uptake mechanism. The approach may be less well suited for Se(IV), whose uptake on cement appears to be related mainly to minerals other than ettringite.

The uptake of iodine species (I − /IO 3 − ) by HCP (hardened cement paste) and a CSH (calcium silicate hydrate) phase under highly alkaline conditions has been investigated using X-ray absorption spectroscopy (XAS). The study was performed at the I K-edge (33.169 keV) instead of the I L 3 -edge (4.557 keV) to avoid interference with Ca ( K -edge=4.038 keV), a major element in HCP and CSH phases. The XANES (X-rays absorption near-edge structure) spectra revealed no changes in the formal oxidation state of iodide (I(-I) − ) and iodate (I(V)O 3 − ) upon uptake by HCP and CSH. The EXAFS (extended X-ray absorption fine structure) oscillations from I − treated HCP and CSH samples were found to be extremely weak, limiting interpretation of the EXAFS data. The IO 3 − EXAFS spectra showed that the IO 3 − entity consisting of three oxygen atoms with a characteristic I-O distance (∼1.78 Å) is maintained upon uptake by HCP and CSH. XANES further indicated that CSH is not the uptake-controlling phase in HCP.

Extended X-ray absorption fine structure (EXAFS) spectroscopy at the uranium L III -edge was used for determining the structural environment of aqueous uranyl sorbed onto montmorillonite. The study reveals that uranyl uptake at pH ∼5-∼7 and at an initial uranyl concentration of 5×10 −5 M takes place at amphoteric surface hydroxyl sites as inner-sphere complex. The measured bond distances between uranium and the equatorial oxygen atoms are in the range of 2.34 Å and 2.37 Å indicating an inner-sphere coordination. At ∼3.4 Å the presence of a U-Al backscattering pair was determined. This backscattering pair indicates that the binding of the uranyl unit to amphoteric surface hydroxyl sites occurs preferred as a bidentate inner-sphere complex on aluminol groups.

The geological survey of natural barriers plays a critical role in determining the suitability of a site for the final disposal of high-level radwaste. This study focused on the characterization of sorption and the determination of distribution coefficients of several safety relevant radionuclides. Both batch tests and polar-microscopy/autoradiography were employed for analyzing and estimating of the sorption of nuclides on one potential host rock in the Taiwan area, which is composed of mostly granite and little diorite. Nuclides investigated were Co, Sr, Cs, U, I, and Am. De-ionized water and synthetic seawater were applied to study nuclide sorption at various concentrations. The distribution coefficients obtained based on the two-week batch tests indicated that the sorption of Co, Sr, Cs, and U on each rock sample was greater in de-ionized water than in synthetic seawater. In addition, sorption onto a specific host rock depended on microscopic heterogeneity. The photos obtained from polar-microscopy/autoradiography showed that biotite is the principal mineral component that is responsible for the sorption of 137 Cs and 90 Sr onto granite and diorite.

Sorption and desorption experiments of Np on magnetite and hematite under aerobic and anaerobic conditions were carried out to investigate the possibility of reduction of Np(V) to Np(IV) at pH 4 to 8 within 7 days. The amount of sorbed Np on magnetite under anaerobic conditions was about 2 or 3 times greater than that under aerobic conditions. Furthermore, the results of desorption experiments indicated that the dominant sorption behavior of Np on magnetite under anaerobic conditions was quite different from that under aerobic conditions. The oxidation state of Np sorbed on the iron oxides was determined by extraction technique using 0.5 M TTA in xylene and 2.0 M HNO 3 solution from the solid phase after sorption experiment. 90% and 10% of extracted Np was Np(IV) for magnetite system under anaerobic and aerobic conditions, respectively. On the other hand, almost 100% of extracted Np was Np(V) for hematite system under both aerobic and anaerobic conditions. These results indicated that Np(V) was reduced to Np(IV) by Fe(II) in magnetite. Redox reaction between Np(V) and Fe(II) was also studied in homogeneous solution without solid to decide if Fe(II) ions released from magnetite into silution or Fe(II) on the solid cause the reduction. Only 6% of Np(V) was reduced by Fe(II) at pH=4 and 6 even after 7 days. According to this result, it was conjectured that the reduction of Np(V) to Np(IV) takes place not in the liquid phase by Fe(II) ion but on the surface of magnetite.

The FT-IR (Fourier transform infrared spectroscopy) spectrum of IO 3 − adsorbed in the interlayer of NO 3 − hydrotalcite (HTNO 3 ) was determined. The v 3 vibration modes of IO 3 − split into two peaks. All peaks caused by adsorbed IO 3 − were found to be shifted to the lower wavelength region relative to the peaks observed from IO 3 − in the KIO 3 crystal. This suggested that IO 3 − adsorbed in the interlayer of HTNO 3 has asymmetry different from the original C 3v symmetrical structure of aqueous IO 3 − . The structural position of IO 3 − in the interlayer of HTNO 3 was optimized using the CASTEP ( C ambridge s erial t otal e nergy p ackage) code. This code is based on density functional theory (DFT) calculations for crystal structures. It was found that the hydroxyl groups bound to O atoms of IO 3 − are not uniform. This may lead to the structural change of adsorbed IO 3 − .

The sorption behavior of Cs onto C-S-H (calcium silicate hydrated phases) was investigated by evaluating the effect of sorption on the C-S-H structure. The C-S-H have a tobermorite-type layer structure. Each layer consists of central calcium and oxygen atoms sandwiched by silicate tetrahedral chains. Additional calcium and water are located in interlayer zones. The XRD spectra showed that the tobermorite-type layer degraded and that the sorption of Cs made the fragments of layer intergrowths much smaller. The fragmentation of the tobermorite-type layer was observed for C-S-H in contact with CsCl solution or with CsOH solution, but not for the C-S-H in contact with Na solution. Furthermore, 29 Si NMR spectroscopy revealed that the sorption of Cs onto C-S-H induced cleavages of silicate chains in C-S-H. It was also suggested that the OH − ion contributed to breaking the silicate chains.

The effect of α-isosaccharinic acid (ISA) on Th(IV) uptake by hardened cement paste (HCP) has been investigated under alkaline conditions (pH 13.3). Prior to performing the uptake studies the stability of HCP was determined in the presence of ISA. It was observed that the formation of Ca-ISA complexes in solution enhances portlandite solubility. The fraction of portlandite dissolved from the HCP matrix depends on the solid to liquid (S/L) ratio of the system and the ISA concentration in solution. Th(IV) uptake by HCP was found to be reduced above an aqueous ISA concentration of about 0 −4 M. Reduction of Th(IV) uptake can be modelled taking into account the formation of a Th:ISA:Ca=1:2:1 complex in solution. It is indicated that the formation of ternary Th(IV)-ISA complexes may be important in cement systems. The final interpretation of the data however fails due to the large uncertainties in the distribution ratios measured in the absence of ISA.

The complexation of Eu(III) with Polyacrylic acid (PAA) and alumina-bound polyacrylic acid (PAA-Al 2 O 3 ) was studied at pH 5 in 0.1 M NaClO 4 . The experiments were carried out for two polyelectrolytes at 5000 Da and 50000 Da. The quantitative description of the interaction of Eu and PAA showed the formation of a 1:1 Eu:PAA complex and the existence of similar binding modes ("sites") of Eu with both polymers. Interaction studies of Eu(III) with alumina-bound PAA particles were carried out in conditions where their overall surface charge were equal to 0. The description of the data indicated the existence of a similar site in binary and ternary systems for high Eu loading. For low Eu loading, the existence of a new site on PAA-Al 2 O 3 particles resulting from the PAA-Al 2 O 3 interaction was suggested.

Sorption of elements like Cs on clay is one of the principal processes delaying their release from deep repositories of nuclear wastes into the environment. The sorption processes taking place between non-purified natural clay material (bentonite) and synthetic groundwater (containing Ca, Mg, Na, K and carbonates) were therefore studied experimentally and modelled for Cs to determine whether thermodynamic computer codes capable of predicting the behaviour of this element in natural systems might be developed. The model used, based on the properties of a pure montmorillonite phase, incorporates the surface reactions for natural major ions and sorbing cations but does not have any adjustable parameters. The weight of each parameters used in the model is assessed. Surface reactions are classified as either major or minor, and a simplified model of Cs sorption that considers only the major processes is proposed. This simplified model might correspond to the less sophisticated thermodynamic model included in coupled geochemistry-transport models.

The sorption of cesium at the calcium silicate hydrates (CSH) surface was investigated, both through sorption isotherm data and by solid-state NMR experiments. The sorption ability of CSH towards cesium is favored for low solid Ca/Si molar ratios, in agreement with the negative surface charge they develop then. A significant proportion of these sorbed cesium cations remains tightly bound to the surface sites forming, in dehydrated CSH, inner-sphere complexes, which can not be removed by alcohol washing. Chloride seems to present a lower affinity for CSH, even for high Ca/Si molar ratios, where the surface charge becomes positive.

The amounts of clay minerals in geologic samples from the Australian arid zone were measured using the SIROQUANT program, which quantifies mineral abundances from X-ray diffraction patterns using the Rietveld method. The sorption of trace cesium on the same samples was investigated in batch experiments. The statistical relationships between the Cs distribution coefficients (K d values) and the smectite, illite and kaolinite contents were assessed. Other characterisation data such as the BET surface area and total clay content were included in the statistical analysis. The correlation of the illite content with the K d for trace Cs was stronger than any other correlation that was tested. This can be attributed to the presence of frayed edge sites on illite which are highly selective for the adsorption of trace Cs. Sorption experiments were also carried out with higher Cs concentrations, which were expected to saturate the selective Cs sorption sites on illite. Under these conditions, there was no significant correlation between the illite content and the cesium K d . However, the K d was related to the smectite content of the materials. There was no correlation between the kaolinite content and the K d for Cs sorption under any of the experimental conditions. The sorption of Cs on the samples was influenced by the BET surface areas, but normalisation of the K d data by surface area did not fully explain the sorption results. Only the quantitative XRD data were able to explain the shapes of Cs sorption isotherms. The isotherm was distinctly different for an illite-rich sample and did not conform to the Freundlich isotherm exhibited by other samples. This study demonstrates that quantitative XRD measurements are valuable in predicting Cs sorption properties of complex geologic samples, particularly sorption data obtained with trace Cs levels where bulk measurements such as the BET surface area are less useful.

An optical cell system for spectroscopic speciation of f -elements in hydrothermal solutions was developed and applied to study the luminescence properties of lanthanide[Ln](III) ions by time-resolved laser-induced fluorescence spectroscopy. The apparatus to maintain the hydrothermal conditions consists of a HPLC pump, an optical cell with three sapphire windows, an electric furnace, a back pressure regulator, etc . Temperature and pressure of sample solutions can be controlled independently in the range of ambient conditions to 723 K and 40 MPa, respectively. Emission spectra and lifetimes of Ln(III) in HClO 4 solutions were measured as a function of temperature and pressure. The pressure effect on the emission spectra and lifetimes was not observed in the range of 0.1 to 40 MPa at a constant temperature. From the temperature dependence of the luminescence properties at 40 MPa, it was found that Ln(III) exist as a hydrated ion up to ca . 500 K and that the variation of the luminescence properties at higher temperature is mainly due to the hydrolysis of Ln(III). Thermodynamic parameters and isotope effects for the quenching of excited Ln(III) in H 2 O and D 2 O solutions were also estimated from the temperature dependence of the luminescence lifetimes.

For the determination of plutonium isotopes in environmental samples, chemical separation of plutonium together with a suitable sample preparation method for measurement is necessary. In this work, a combined method for the determination of 238 Pu, 239 Pu, 240 Pu and 241 Pu based on alpha spectrometry, Liquid Scintillation Counting (LSC) and Inductively Coupled Plasma Mass Spectrometry (ICP-MS) is presented. For the alpha spectrometric determination of 238 Pu and 239(40) Pu, sample preparation by coprecipitation of Pu with NdF 3 on a membrane filter after separation by anion exchange, using 242 Pu as tracer, is widely used. This sample preparation method was also applied as a preconcentration step for the determination of 240 Pu/ 239 Pu ICP-MS after wet ashing of a part of the membrane filter. 241 Pu was measured with LSC after dissolving a part of the same membrane filter in a suitable cocktail and using the total alpha activity of the filter for determination of the chemical yield. The use of this approach gives the advantage of combining the strengths of both radiometric and mass spectrometric methods for plutonium determination after chemical separation from environmental samples. Results from the measurement of reference materials and from sediment samples from the Irish Sea are presented and the applied methods are compared with respect to their detection limits for the investigated nuclides.

Humic colloids in deep groundwater are characterized in order to ascertain how heavy metal ions of chemical homologue to actinides are associated in different size fractions. The colloid size fractionation is made by two different methods: flow field flow fractionation (FFFF) and size exclusion chromatography (SEC), which is followed by analysis of chemical composition using UV spectroscopy for organic components and ICP-MS for inorganic components. Relative number density of humic colloids following the size fractionation is determined by laser-induced breakdown detection (LIBD). For the appraisal of colloid size change upon metal ion complexation, purified Aldrich humic acid is loaded with the Eu 3+ ion on increasing the concentration and the size change is then determined by LIBD. Humic colloids are interacted with radioactive tracers, 155 Eu(III) and 228 Th(IV), to appreciate their sorption behaviour onto different colloid size fractions and thus to compare with natural humic colloid-borne elements of M(III) and M(IV). Whereas the organic component of humic colloids is found in the size range about 3 nm, the inorganic components of actinides homologues, M(III) and M(IV), are observed in the fractionated size range from 10 nm to 35 nm. This observation leads us to presume that the inorganic composites, composed of heavy metal elements, are peptised by humic/fulvic acid. The laboratory traced elements, 155 Eu(III) and 228 Th(IV), are quantitatively sorbed only onto the organic component. This fact infers the complexation of traced metal ions with humic components of colloids without visible interaction with the inorganic components. Once purified humic acid is complexed with Eu(III) in an excess amount, the size increases over 200 nm and the resulting products undergo flocculation or precipitation. The results of the present experiment indicate that the natural humic colloid-borne heavy metal ions appear in different chemical states as compared to the same but laboratory traced metal ions. As a result, the laboratory simulation entails a better understanding of the underlying molecular level chemical reactions, which are important to the assessment of the colloid-borne actinide migration.

This paper describes a study of technetium solubility and migration under chemical conditions representative of those prevailing in a Boom Clay environment. Laboratory and in situ measurements yielded similar aqueous concentrations of technetium, of about 1×10 −8 mol dm −3 , close to the concentrations measured for hydrated technetium(IV) oxide TcO 2 ·1.6H 2 O in the solubility studies. From fitting the curves of the Tc concentrations as function of time, distribution coefficient (K d ) values were estimated to lie between 0.8 cm 3 g −1 and 1.8 cm 3 g −1 . Exposure of the system at 80 °C and to γ-radiation dose rates of several hundred Gy h −1 resulted in only minor differences in behaviour.

Under the geochemical conditions prevailing in situ in the Boom Clay Formation (pH, E h , …), calculations predict that U(OH) 4 is the dominant uranium species present in the interstitial water and the concentration is solubility limited. However the boundary of the domain where the non solubility limited UO 2 (CO 3 ) 3 4− species dominates is very close. It is therefore of prime interest to know the correct speciation of uranium during the migration process. Electromigration was used as technique with the advantage that it can provide information on the speciation because the movement of the species towards the electrodes depends on its charge and speciation. Electromigration experiments have been performed with preconditioned 233 UO 2 (CO 3 ) 3 4− sources, starting from the hypothesis that this species should migrate without retardation towards the anode. Despite relatively long electromigration times, sufficient to displace strong retarded tracers, no displacement of the migration profile towards the anode was observed. All 233 U remained near the source position, but the electropherograms clearly showed the presence of species moving towards the cathode. This indicates the presence of neutral or positively charged uranium species. These electropherograms are interpreted as a change in uranium valence state: reduction of UO 2 (CO 3 ) 3 4− and precipitation of U(IV) oxy-hydroxides near the source position. The solubility limited species, U(OH) 4 (aq), are carried with the pore water towards the cathode. The electromigration experiments indicate, in support of the speciation calculations, that the dominant migrating U-species is probably the solubility limited U(OH) 4 .

The record of the global atmospheric fallout could be found in the lake sediments. A mass balance for fallout radionuclides in Blelham Tarn and its catchment is established. The sediment activity contribution is coming from direct atmospheric fallout and from indirect atmospheric fallout via the catchment. The catchement activity is conveyed to the sediment by the rivers and the direct streaming. A comparison of the fallout and the sediment inventory allows the activity estimation of these different contributions and to understand the mobility of these elements on the catchment and in the sediments. The study of activity profile in sediment core allows to characterise the different radioactive events occurred in the past. For the lake Blelham, the results show two cesium activity peaks and only one peak for transuranic activities. The deepest peaks correspond to the atmospheric nuclear test fallout in the sixties (1963) and the second peak to the Chernobyl accident (1986). The activity ratio 239-240 Pu/ 137 Cs allows estimating the ratio between cesium activities in sediments coming from these two events. Plutonium and cesium diffusion coefficients are calculated with a simple analytical model.

The diffusion of radionuclides in water-saturated porous media, such as compacted bentonite, is traditionally modeled assuming diffusion in the pore water and immobilization by adsorption on the clay surface. In reality there are several sorption mechanisms acting in the clay-water system. We have therefore carried out a careful diffusion study of the cations Na + , Sr 2+ , Cs + , which are predominantly sorbed by electrostatic interaction, and Co 2+ , for which the dominant sorption mechanism at pH > 7 is inner-sphere complexation and/or surface precipitation. The study has been performed at different densities of clay compaction (1.8, 1.6, 1.2, 0.8, 0.4 g cm −3 ) using a through diffusion technique. In each experiment, the apparent diffusivity and transport K d were obtained by computer simulation of the experimentally measured cumulative flux through a plug of compacted bentonite and the concentration profile within the bentonite. Batch sorption experiments with Na + , Sr 2+ , Cs + and Co 2+ were also carried out. Measured batch K d values were compared to diffusion-derived K d values using the computer code ANADIFF. The obtained diffusivities of Cs + and Sr 2+ decreased as the density of bentonite increased. The diffusivity of Cs + decreased from 1.5×10 −7 to 1.1×10 −8 cm 2 /s as the clay dry bulk density increased from 0.4 to 1.8 g/cm 3 and the diffusivity for Sr 2+ varied from 4.3×10 −7 to 9.0×10 −8 cm 2 /s. The diffusivities of Co 2+ also decreased from 3.0×10 −9 to 1.0×10 −10 cm 2 /s in the same range of clay density. An interpretation of the diffusion data is proposed based on differences in porosity and tortuosity of the bentonite compacted to different densities, and on differences in binding energy, hydration energy, and hydrolysis constants for the cations studied.

The mineralogy of the supergene-weathered Bangombé natural fission reactor (RZB) and surrounding uranium deposit has been analyzed and compared with the thermodynamically predicted minerals based on groundwater compositions. The primary U-minerals are uraninite UO 2+x and minor coffinite (U[SiO 4 ]· nH 2 O) ; sometimes with phosphorous. The U 6+ minerals include fourmarierite Pb 1−x [(UO 2 ) 4 O 3−2x (OH) 4+2x ]·4(H 2 O), bassetite (Fe 1−x 2+ Fe x 3+ [(UO 2 )(PO 4 )] 2 (OH) x ·8− x H 2 O); possibly associated with U(HPO 4 ) 2 ·2H 2 O and/or chernikovite ((H 3 O) 2 [(UO 2 )(PO 4 )] 2 ·6H 2 O); torbernite (Cu[(UO 2 )(PO 4 )] 2 · 8-12H 2 O), Ce-françoisite-(Nd) (REE(UO 2 ) 3 O(OH)(PO 4 ) 2 · 6H 2 O), and uranopilite ((UO 2 ) 6 (SO 4 )O 2 (OH) 6 (H 2 O) 6 ·8H 2 O). Autunite (Ca[(UO 2 )(PO 4 )] 2 ·10-12H 2 O) has also been reported. Thermodynamic equilibrium modeling was completed using Geochemist's Workbench¯ with an expanded data base and the groundwater composition (−112 mV ≤ E h ≤ 143 mV; pH = 5.96) at the base of RZB in drill-hole BAX03. The new ΔG 0 f,298.15 data were obtained from the literature or estimated using the polyhedral contribution method. Based on the updated database, E h -pH diagrams predict that coffinite, U(HPO 4 ) 2 ·H 2 O and UOF 2 ·H 2 O are the only stable U 4+ phases and that uranopilite, torbernite and bassetite will become stable during oxidative alteration. Except from UOF 2 ·H 2 O, this is in accord with mineralogical observations. The role of Cu was predicted from log a Cu -pH diagrams, which predicts that torbernite is stable at log a Cu =3.98×10 −14 and pH≥2.2 at E h =143 mV in BAX03 groundwaters. At E h = −112 mV, torbernite is stable at pH>5.5. Soddyite ((UO 2 ) 2 SiO 4 ·2H 2 O) was predicted to form at the expense of coffinite, but soddyite has not been identified at Bangombé. Previous blind prediction modeling, often omitting P and S, had only predicted soddyite and haiweeite (Ca(UO 2 ) 2 [Si 5 O 12 (OH) 2 ]·4-5H 2 O) and hence, failed to predict the U 6+ minerals observed at Bangombé. The results stress the importance of SO 4 2− and PO 4 3− resulting from dissolution of accessory apatite, monazite and sulfides in the retardation of U owing to the formation of U 6+ sulfates and phosphates.

To study the interaction between Tc and subsurface bacteria, we conducted batch experiments with soil and groundwater or sterilized deionized water. The system water/soil was amended with lactate and phosphate for bacterial growth. Nitrate and sulfate were added to stimulate the growth of indigenous denitrifying and sulfate-reducing bacteria. During denitrification Tc-concentration did not change with time. In the presence of sulfate-reducing bacteria, Tc-concentrations decreased in reacted waters which could be attributed to Tc(VII) reduction and precipitation of TcO 2 and/or TcS 2 . Coprecipitation with newly formed iron sulfide is expected to contribute to Tc removal. Additional experiments with U and Tc showed that these elements were simultaneously reduced by sulfate-reducing bacteria. This work shows that 1) subsurface mixed cultures of denitrifying bacteria do not remove Tc from solution, this is different from uranium and 2) sulfate-reducing bacteria reduce and remove Tc from aqueous solutions and thus in situ bioremediation of subsurface waters and soils may be possible with such ubiquitous bacteria.

We studied the interaction of U(VI) with vegetative cells, heat killed cells, spores, and decomposed cells of Bacillus sphaericus . The characterization of the formed complexes was performed by time-resolved laser fluorescence spectroscopy (TRLFS) and extended X-ray absorption fine structure spectroscopy (EXAFS). We observed no significant differences in the sorption behavior of vegetative and heat killed cells, whereas the spores showed a higher sorption of U(VI) (related to their dry weight). Regardless of the higher relative sorption of the spores of B. sphaericus , the fluorescence and EXAFS spectra of the vegetative cells, heat killed cells and spores were almost identical. Analysis of the data proved that U(VI) forms inner sphere complexes with organic bound phosphate groups on the cell surface. We observed no significant differences in the coordination numbers and the distances of the oxygen and phosphorus atoms in the inner coordination sphere. After eight weeks, the vegetative cells of B. sphaericus were completely decomposed. Lysing of the cell walls and activity of enzymes led to a release of various decomposition products. We found that large amounts of H 2 PO 4 − were released which caused a quantitative precipitation of bacterial U(VI) as UO 2 (H 2 PO 4 ) 2 . The H 2 PO 4 − was detected by Raman spectroscopy. The decomposed bacterial suspension showed the same fluorescence spectrum as UO 2 (H 2 PO 4 ) 2 which differed significantly from those of the bacterial U(VI) surface complexes.

The bio-sorption of neptunyl (NpO 2 + ) by Pseudomonas fluorescens was investigated. The overall goals of this research are to identify key interactions between neptunium and soil bacteria and to model these effects under subsurface-related conditions. Neptunyl, which is generally thought to be non-sorptive, was significantly sorbed under all conditions studied. At initial neptunyl concentrations of 4.75 µM and pH = 7, as much as 85% of the neptunium was sorbed under aerobic conditions. Kinetic studies show that neptunyl was sorbed rapidly within the first 15 minutes. The extent of sorption also increased with pH. In all cases, the sorbed neptunium was shown to be NpO 2 + by X-ray absorption near edged spectroscopy (XANES) analysis, confirming that there was no reduction to Np(IV) under the conditions of our experiment. The sorption data were modeled using Langmuir and Freundlich isotherms. A comparison of the two approaches showed a significantly better fit for the Freundlich isotherm, and the Freundlich parameter values suggest interactions between sorbed NpO 2 + molecules. These data show that bio-sorption, even for neptunyl, has a significant role in defining the speciation of neptunium and, hence, its overall mobility in the subsurface.

Clostridia are strict anaerobic, spore-forming, fermentative bacteria commonly present in soils, sediments, and wastes; and, they play a major role in the decomposition of a wide variety of organic compounds. They also are involved in the reduction of iron, manganese, and uranium, thereby affecting their solubility. However, little is known of the ability of Clostridia to reduce technetium (Tc). We investigated the reduction and precipitation of pertechnetate by Clostridium sphenoides able to metabolize citrate as its sole carbon source, and Clostridium sp. capable of fermenting glucose but not citric acid. Both species reduced Tc(VII) to Tc(IV), although C. sphenoides did so at a greater rate and extent than Clostridium sp. The reduced Tc was predominantly associated with the cell biomass. It also was present in solution complexed with bacterial metabolic products (MW>5000). Adding diethylenetriaminepentaacetic acid (DTPA) to Clostridium sp. resulted in the formation of a soluble Tc(IV)-DTPA complex, whereas with C. sphenoides only a small amount of Tc was present in solution, indicating that insoluble Tc species were formed. These results suggest that Clostridia may play a major role in regulating the mobility of Tc under anaerobic conditions in wastes and subsurface environments.

Thermodynamic data for Zr(IV) have been reviewed in the context of a recent update of the Nagra/PSI thermochemical database, which is primarily used in safety assessment calculations for radioactive waste repositories in Switzerland. In spite of the important role of Zr in nuclear reactor technology and as fission product in radioactive waste, the thermodynamic properties of this element in aqueous solutions are poorly constrained and published data are frequently contradictory. Specifically, hydrolysis constants, which are critical to any geochemical calculation of Zr speciation in aqueous solutions, may vary by several orders of magnitude. Published solubility and hydrolysis data for Zr oxides were reviewed and evaluated. Recent solubility measurements on crystalline ZrO 2 , obtained in our laboratory, were integrated in this analysis. These new data are the only ZrO 2 solubility measurements obtained in a pH region relevant for environmental aqueous systems, where equilibrium concentrations decrease to nanomolar levels. All solubility data of (hydrous) Zr oxides can be reproduced satisfactorily by a single set of hydrolysis constants, provided that only data measured from undersaturation are considered. This set includes the species Zr 4+ , ZrOH 3+ , Zr(OH) 4 0 and Zr(OH) 5 − . However, present experimental data are still not sufficient to establish a fully reliable database.

The Nagra/PSI Chemical Thermodynamic Data Base has been updated to support performance assessments of the planned Swiss repositories for radioactive waste. The update from version 05/92 to 01/01 involved major revisions for most of the actinides and fission products. Altogether, more than 70% of the database contents have been revised. Data for U, Np, Pu, Am and Tc recommended by the NEA TDB project were considered in the update. Thermodynamic data for Th, Sn, Eu, Pd, Al, and solubility and metal complexation of sulphides and silicates were extensively reviewed. Data for Zr, Ni and Se were examined less rigorously as these elements are currently being reviewed in phase II of the NEA TDB project. Our experiences from this two year team effort can be summarised as follows. Detailed in-house reviews and critical appraisal of NEA recommendations greatly improved the chemical consistency and quality of the selected data. On the other hand, we could discern major gaps in the data, especially missing carbonate complexes. In some systems, e.g . Th(IV)-H 2 O and U(IV)-H 2 O, experimental data could not be described by a unique set of thermodynamic constants. There, a pragmatic approach based on solubility data was chosen for application to performance assessment. The electronic version of our database and information concerning its full documentation is available on our PSI web site (http://www.psi.ch/les).

Radionuclide sorption on mineral-water interfaces can be thermodynamically modelled, similar to solid-solution aqueous-solution systems (only in chemical elemental stoichiometry), if definitions of the standard and reference states, surface activity terms (SAT), and elemental stoichiometries of surface-bound species are unequivocally established. A pre-requisite is that a unique common value of the reference (site) density (Γ o ) must be part of the definitions of standard and reference states, while the sample- and surface-specific maximum density parameters (Γ max ) be included into the SAT corrections to reproduce saturation of physically available adsorption sites instead of introducing the additional mass balance constraints. Subsequently, the standard partial molal properties of surface complexes at multiple surface types on different sorbents can be found, comparable with the standard molar properties of solids, gases and aqueous species. Using surface complexation models (SCM) of U(VI) adsorption on quartz and amorphous silica (SiO 2 ) phases in GEM (Gibbs energy minimization) implementation, a feasible way is shown how to construct a uniform, internally consistent thermodynamic dataset for surface species of radionuclides; to use standard partial molal Gibbs energies G o 298 of surface species in sorption modelling; how to convert G o 298 values from/to log K or intrinsic adsorption constants log K int to use in the law-of-mass action (LMA) speciation codes, and finally, into “smart K d ” values for the applications related to waste repository performance assessment.