Band 94, Heft 9-11 - Migration 2005

The solubility of zirconium(IV) hydrous oxide was measured in the hydrogen ion concentration (pH c ) range from 1 to 13 at 25 °C. The experiment was carried out with I =0.1, 0.5 and 1.0 mol/dm 3 (M) of NaClO 4 solution. The concentration of Zr was measured by ICP-MS after separating the aqueous phase from the solid phase by ultrafiltration methods. Some differences between the obtained values and literature data were observed due to the size distribution of chemical species. In the lower pH c region, the size distribution of filtrate species was investigated by different pore-size filtration in the range of 3 k to 100 kDa NMWL (nominal molecular weight limit) membranes. The solubility values of mononuclear species were obtained by the Flory model, and by using the specific ion interaction theory (SIT) for ionic strength corrections and the known literature value of the hydrolysis constant β 1 °, the tentative value of the solubility product ( K sp °) was determined to be −56.94±0.32.

The solubility of ThO 2 · x H 2 O(am) is studied at I =0.1−4.0 M (NaHCO 3 -Na 2 CO 3 -NaCl) and 22 °C. Several sets of closed system experiments are performed at total carbonate concentrations of C tot =[HCO 3 - ]+[CO 3 2- ]=0.02 and 0.1 M in the range pH c = 8−11. In our recent study at I =0.5 M, the simultaneous evaluation of solubility data at widely varied pH c and carbonate concentrations has shown that Th(OH)(CO 3 ) 4 5- , Th(OH) 2 (CO 3 ) 2 2- and Th(OH) 4 (CO 3 ) 2- are the most important ternary complexes. The present results at I =0.1−4.0 M are used to describe the ionic strength dependence of the equilibrium constants log K s, 1 yz (for the reactions Th(OH) 4 (am)+ z CO 3 2- ↔Th(OH) y (CO 3 ) z 4- y -2 z +(4- y )OH - ) with the SIT. EXAFS measurements in 0.1 and 1.0 M carbonate solutions support the calculated speciation. They show clearly different spectra for the predominant species Th(OH)(CO 3 ) 4 5- and Th(CO 3 ) 5 6- , respectively.

Trace activity coefficients of the Th 4+ ion in dilute to concentrated NaClO 4 , NaNO 3 and NaCl solutions ([H + ]=0.01−0.02 M) have been determined at 22 °C from liquid-liquid phase equilibria with 10−50 vol.% TBP in n -dodecane. Using the specific ion interaction theory (SIT) to describe these equilibria as a function of the medium electrolyte concentration, the following ion interaction coefficients are calculated: ε(Th 4+ , ClO 4 - )=0.70±0.06 kg/mol and ε(Th 4+ , NO 3 - ) = 0.31±0.12 kg/mol. The latter value differs considerably from ε(Th 4+ , NO 3 - ) = 0.11±0.02 kg/mol used in the NEA-TDB. The low distribution coefficients at m NaCl <3 mol/kg do not allow the linear SIT extrapolation to I =0, but the equilibrium constants in 2.5−5.0 m NaCl are compatible with the NEA-TDB value of ε(Th 4+ , Cl - )=0.25±0.03 kg/mol. The SIT coefficients determined for the Th 4+ ion follow the linear correlations between known values of ε(M Z + , NO 3 - ), ε(M Z + , Cl - ) and ε(M Z + , ClO 4 - ) for non-complexed cations M Z+ with Z =1−4, including ε(Pu 4+ , ClO 4 - )=0.82±0.07 kg/mol and ε(Pu 4+ , Cl - )=0.37±0.05 kg/mol calculated from data accepted in the NEA-TDB. The interaction coefficients in the series of the tetravalent actinide ions show a slight systematic dependence on the ionic radius.

Leach resistance of neptunium-doped garnet- and murataite-based ceramics with calculated compositions (in wt.%: 38 Fe 2 O 3 , 21 Gd 2 O 3 , 16 NpO 2 , 15 ZrO 2 , 10 CaO and 55 TiO 2 , 10 Mn 2 O 3 , 10 CaO, 10 NpO 2 , 5 ZrO 2 , 5 Al 2 O 3 , 5 Fe 2 O 3 respectively was studied. The samples were prepared from pre-activated oxide mixtures by sintering/melting in platinum ampules at 1450 °C. The murataite ceramic is composed of predominant murataite structure phases with five- (5C) and eight-fold (8C) elementary fluorite unit cell and minor rutile (≤5% of total bulk). The garnet-based sample is a polyphase ceramic and consists of major garnet phase, interstitial Fe 2 O 3 and Ca-Fe-phase, and rare grains of (Ca,Zr,Np)O 2 -phase (cubic solid solution) included in the larger garnet grains. The edge of the garnet grains is enriched with Np as compared with their rim by approximately 1.5−2 times. Np leach-rates were measured using an ANS_16.1 test. Np leach rate from the murataite-based ceramic reduced significantly from 2.44×10 -6 g/(cm 2 day) within the first 2 hours of leaching to 3.41×10 -9 g/(cm 2 day) at 672 hours of leaching. At the same time for the garnet-based ceramic corresponding values were 1.49×10 -5 g/(cm 2 day) and 1.77×10 -6 g/(cm 2 day), respectively. Higher leach rates at early periods are due to the dissolution of trace phases and Np release from defect surfaces. Leach rate values at later periods of leaching are characteristic of the “equilibrium” leach rate of the element (Np). Much higher Np leaching from the garnet ceramic may be due to 1) higher concentration of Np at the edge of the garnet grains with a relatively low chemical durability; 2) leaching of Np traces from highly soluble minor Np-Zr-rich phase.

Coprecipitation may be a significant process in controlling radionuclide 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 and lanthanum with UO 2+x (s) as host phase under reducing conditions (pure N 2 and E h =−300 mV/SHE) as a function of pH and [Th]/[U+Th] or [La]/[U+La] initial concentration ratios. At pH values between 3 to 9, the final concentrations of dissolved Th and La were found to be considerably lower than expected if pure phases, Th(OH) 4 (s) and La(OH) 3 (s) would have been formed. Coprecipitation phenomena and formation of solid solutions Th y U 1-y O 2+x′ (s) and (U,La)O 2+x′ (s) are suggested to be responsible for the lower solubility.

Secondary minerals have the potential to sequester escaped actinides in the event of a radioactive waste repository failure, but currently, data to define their maximum uptake capacity are generally lacking. To estimate a maximum limit for solid solution in calcite, we took advantage of the behavioural similarities of the 4 f -orbital lanthanides with some of the 5 f -orbital actinides and used rare Earth element (REE) concentration as an analogue. A suite of 65 calcite samples, mostly pure single crystals, was assembled from a range of geological settings, ages and locations and analysed by isotope dilution MC-ICP-MS (multiple-collector inductively-coupled plasma mass spectroscopy). All samples were shown to contain significant lanthanide concentrations. The highest were in calcite formed from hydrothermal solutions and from carbonatite magma. Maximum total mole fraction of REE was 4.72×10 -4 , which represents one substituted atom for about 2000 Ca sites. In comparison, synthetic calcite, precipitated at growth rates slow enough to insure solid solution formation, incorporated 7.5×10 -4 mole fraction Eu(III). For performance assessment, we propose that 7.5 mmole substitution/kg calcite should be considered the upper limit for actinide incorporation in secondary calcite. The largest source of uncertainty in this estimate results from extrapolating lanthanide data to actinides. However, the data offer confidence that for waters in the hydrothermal temperature range, such as in the near-field, or at groundwater temperatures, such as in the far-field, if calcite formation is favoured and actinides are present, those with behaviour like the trivalent lanthanides, especially Am 3+ and Cm 3+ , will be incorporated. REE are abundant and widely distributed, and they have remained in calcite for millions of years. Thus, one can be certain that incorporated actinides will also remain immobilised in calcite formed in fractures and pore spaces, as long as solution conditions do not favour complete dissolution. Partial dissolution only results in resequestration to calcite again.

Thermodynamic properties of the protonation of D -gluconic acid (HGH 4 (aq)) and its complexation with Np(V) have been studied in acidic to nearly neutral solutions at t =25 °C and I =1 M NaClO 4 by potentiometry, spectrophotometry and calorimetry. The protonation constant (log K H ) and enthalpy (Δ H H ) of the carboxylate group are determined to be (3.30±0.10) and −(4.03±0.07) kJ mol -1 , respectively. Gluconate forms two Np(V) complexes in nearly neutral solutions. The formation constants and enthalpies of complexation are: log β 1 = (1.48±0.03) and Δ H 1 = −(7.42±0.13) kJ mol -1 for NpO 2 (GH 4 )(aq), log β 2 =(2.14±0.09) and Δ H 2 = −(12.08±0.45) kJ mol -1 for NpO 2 (GH 4 ) 2 - . The thermodynamic parameters indicate that gluconic acid, like isosaccharinic acid and other α-hydroxycarboxylic acids, is a slightly stronger acid and forms stronger complexes with Np(V) than simple monocarboxylic acids.

The complexation of Th with α-isosaccharinate has been studied using liquid–liquid-extraction techniques. An organic phase consisting of 0.025 M acetylacetone in toluene was used with a 1.0 M NaClO 4 aqueous phase kept at pH 8 in a thermostated AKUFVE unit at 25 °C. The concentration of α-isosaccharinate (ISA - ) in the aqueous phase was incrementally increased and samples of equal volumes were taken from the two phases and analyzed in a liquid scintillation detector. Stability constants, β p t , for the complexation reaction Th 4+ + p ISA - ↔Th(ISA) p (4- p )+ , were determined to be logβ 1 25°C =12.56±5.01, logβ 2 25°C =19.38+0.35/-0.95, and logβ 3 25°C =21.30±0.26.

Potentiometric and microcalorimetric titration techniques were applied to determine the Gibbs free energies and enthalpies of the protonation of some carboxylic acids (acetic, glycolic, malonic and malic acids) and their complexation with Eu(III) in 1.0 M NaClO 4 solution at 25 °C, where Eu(III) was used as a chemical analogue of radiologically important Am(III). By using the values of Δ G determined by potentiometric titrations, the results of calorimetric titrations were analyzed to give the values of Δ H and Δ S . To support the discussion on the obtained thermodynamic quantities, the hydration numbers of Eu(III) in the complexes were obtained by TRLFS measurement, where the luminescence lifetime of Eu(III) excited by 394 nm laser was measured. A few to several kJ/mol of enthalpies were determined for the reactions within the uncertainties of ±0.01 to ±0.4. These enthalpy values indicated that the protonation of these carboxylates were entropy-driven, that is, |− T Δ S |>>|Δ H | in Δ G =Δ H − T Δ S . The complexations of Eu(III) were also shown to be controlled mainly by entropy term. The results of TRLFS measurement revealed that the number of dehydrated water molecule increased with the degree of complexation. A linear relation was found between total entropy change of the system (Δ S T ) and the net number of released molecules from the central Eu(III) ion. However, the values of Δ S T for the hydroxy-carboxylates were smaller than that for simple carboxylates at the same net number of molecules released from Eu(III), suggesting that the dehydration occurred not only in the cation but also in the anion.

The interaction of Cu(II) and Eu(III) ions with humic acid has been studied by potentiometry under normal atmospheric conditions at pH 6, I =0.1 M (NaClO 4 ) at room temperature (23±2 °C). The investigation was performed for the Cu(II) ion directly by means of a Cu 2+ ion selective electrode, whereas for the Eu(III) ion it was indirectly based on competition reactions between the Cu(II) ion and the Eu(III) ion. The evaluation of the conditional formation constants, which was based on the charge neutralization model, resulted in as follows: log β (Cu-HA(II)) = 5.6±0.3 log β (Eu-HA(III)) = 6.2±0.5

To model the influence of humic substances on the migration of tetravalent actinides (Pu, Np, U) in and around nuclear waste repositories we have immobilized humic acid on silica gel and investigated its complexation of Th(IV). It is anticipated that this material might serve as a geochemical model of the humate-coated minerals that are likely to be present in the vicinity of the repositories. The binding of Th(IV) by the immobilized humic acid was examined at pH 4, 5 and 6 in 0.1 mol L -1 NaClO 4 by the batch method. Th(IV)-humate conditional interaction constants have been evaluated from data obtained from these experiments by using non-linear regression of binding isotherms.

The chemical shift of the X-ray absorption K -edge of Tc(IV) complexes determined by X-ray Absorption Spectroscopy (XAS) measurement was studied. A correlation between edge position (Δ E ) and partial charge on Tc atoms (δ Tc ) was determined. This correlation was used to determinate δ Tc in Tc n O y (4n-2y)+ . Furthermore, a theoretical relation between δ Tc and the charge of Tc n O y (4n-2y)+ , was also established and the overall charge of this complex was estimated.

This paper describes the effects of aqueous carbonate concentrations on the UO 2 oxidation and dissolution rates under an alpha beam irradiation in a cyclotron. As solid samples, UO 2 colloids were synthesized by nano-particle precipitation. The specific surface area obtained is 85.3 m 2 /g. All aqueous solutions were buffered in the pH range 8<pH<9. The observation of particles under TEM and X-ray diffraction showed well-crystallized UO 2 . Molecular radiolysis products such as H 2 O 2 and carbonate degradation products where analysed. Solution analysis of dissolved uranium and of radiolytically generated H 2 O 2 allowed to deduce dissolution rates of UO 2 as well as H 2 O 2 consumption rates (deduced from the H 2 O 2 concentrations in presence of UO 2 colloids and those in irradiated blank solutions). In absence of carbonate, the dissolution rates of UO 2 are much lower than the H 2 O 2 consumption rates. The difference is explained by the oxidation of uranium(IV) at the UO 2 /solution interface ( e.g. formation of U 3 O 7 ). In contrast, in the presence of carbonate (up to 4×10 -2 mol/L), the dissolution rates of UO 2 are similar to H 2 O 2 consumption rates. In this case the oxidized uranium is complexed with carbonate ions resulting in highly soluble species as indicated by the geochemical model PHREEQC calculations [UO 2 (CO 3 ) 3 4 - ]. The high carbonate concentration did not affect the H 2 O 2 formation/consumption rates. Radiolytic decomposition of carbonate ions and formation of organic radiolytic degradation products were negligible. The results indicate that under the present experimental conditions, UO 2 oxidation by H 2 O 2 and UO 2 dissolution are connected as a sequential reaction: for a given H 2 O 2 concentration, UO 2 oxidation occurs at a constant rate. At low carbonate concentration, the dissolution of oxidized UO 2 is slower than the oxidation rate (hence oxidation is not rate limiting), while at high carbonate concentrations all oxidized UO 2 becomes immediately dissolved and UO 2 oxidation becomes rate limiting for UO 2 dissolution. This mechanism is in agreement with that proposed in the Matrix Alteration Model (MAM).

A glove-box in contact with the lab atmosphere has been used in a test to see whether trivalent plutonium can be kept in this oxidation state at different pH. The box hosted a vessel of Pu(III) solution through which a hydrogen/argon gas mix was bubbled in the presence of a Pd/Pt catalyst. The results of solvent extractions and UV-Vis spectra for this system show that such conditions should not be considered sufficient to keep plutonium trivalent at a pH above two. The solvent extraction method was not found to affect the oxidation state composition of the samples.

We examined the reduction behavior of UO 2 2+ in citrate media at pH 2.0−7.0 by column electrodes. At pH 2.0, UO 2 2+ was reduced to U(IV) through a one-step reduction process, while it was reduced to U(IV) through a two-step reduction process at pH 3.0−5.0. The reduction potential of UO 2 2+ shifted to lower values with an increase in pH from 2.0 to 7.0. At pH 6.0 and 7.0, UO 2 2+ was not reduced to U(IV) completely at the electrode potential above -0.8 V vs . silver/silver chloride electrode. Ultraviolet-visible spectroscopy and speciation calculation of UO 2 2+ in citrate media indicated that uranium existed as a mixture of UO 2 2+ , [UO 2 Cit] - and [(UO 2 ) 2 Cit 2 ] 2- at pH 2−3, and a predominant species at pH 3−5 was [(UO 2 ) 2 Cit 2 ] 2- (H 3 Cit: citric acid). At pH 5−7, polymeric complexes, probably, [(UO 2 ) 3 Cit 3 ] 3- and [(UO 2 ) 6 Cit 6 (OH) 10 ] 16- were present. These findings suggest that the reduction of UO 2 2+ is more difficult by polymerization of UO 2 2+ with citric acid at higher pHs. Absorption spectra of the reduced complexes showed that U(IV) forms soluble complexes with citric acid at pH 2.0−5.0, and presence of U(V) species was not observed during the reduction of UO 2 2+ .

Zirconolite is a candidate host material for conditioning minor tri- and tetra-valent actinides arising from enhanced nuclear spent fuel reprocessing and partitioning, which can be disposed in a geological repository for nuclear waste. Its chemical durability has been studied here under charged particle-induced radiolysis (He 2+ and proton external beams) to identify possible effects on dissolution rates and mechanisms in pure water. Two geometries of experiments have been used to evaluate the influence of the following parameters: solid irradiation, Linear Energy Transfer (LET) at the interface and total deposited energy. Preliminary results on the elemental releases due to the enhanced dissolution of the zirconolite surface during charged particle-induced irradiation are first presented. Then, we focus on H 2 O 2 production which is one of the major molecular species, created under water radiolysis, and likely to interact with the zirconolite surface. In presence of zirconolite, first results indicate an apparent consumption of the radiolytic hydrogen peroxide or its precursors compared to the production in pure water calculated from the primary yield G H 2 O 2 . The measured H 2 O 2 concentration varies linearly with the total deposited energy in water over the irradiation duration (between 1 h and 6 h) and in the conditions of our experiments. Moreover, the H 2 O 2 concentration decreases when the local density of the deposited energy close to the interface increases. Thus, we suggest that the mechanism(s) leading to the consumption of H 2 O 2 or its precursors involve zirconolite surface reactions.

The redox stability of PuO 2 2+ was investigated in brine under subsurface conditions. In simulated brines, when no reducing agent was present, 0.1 mM concentrations of plutonium(VI) were stable as regards to reduction for over two years, which was the duration of the experiments performed. In these systems, the plutonyl existed as a carbonate or hydroxy-chloride species. The introduction of reducing agents ( e.g. steel coupons, and aqueous Fe 2+ ) typically present in a subsurface repository, however, led to the destabilization of the plutonium(VI) complexes and the subsequent reduction to Pu(IV) under most conditions investigated. X-ray Absorption Near-Edge Spectroscopy (XANES) confirmed that the final oxidation state in these systems was Pu(IV). This reduction lowered the overall steady state concentration of plutonium in the brine by 3−4 orders of magnitude. These results show the importance of considering repository constituents in evaluating subsurface actinide solubility/mobility and provide further evidence of the effectiveness of reduced iron species in the reduction and immobilization of higher-valent plutonium species.

First results of a periodic and cluster Density Functional Theory (DFT) study of the uranyl ion (UO 2 2+ ) sorption onto the rutile TiO 2 (110) face, based on plane wave and localised basis sets, are presented. A five layers slab with its most internal layer frozen to bulk positions was found to be a good surface model. In a first step and as reference data for the sorption process, the [UO 2 (H 2 O) n ] 2+ systems, with n =4 to 6 were studied. Relative solvation energies confirmed that the uranyl ion adopt a pentacoordinated structure in aqueous solution. From localised approach, an overall 0.91 electron transfer from the first hydration shell to the uranyl ion was calculated. Then, a periodic study of the uranyl sorption on a simplified hydroxylated TiO 2 (110) surface model was investigated. The resulting optimised structural parameters, for the three possible adsorption sites, show that the sorbed uranyl ion first coordination shell (saturated by three water molecules) plays an important role to model the adsorption process. Both methodologies (plane waves and localised atomic orbitals) were also used with a cluster model and gave similar results in agreement with experimental data. This first step in the understanding of the uranyl ion sorption onto the simplified hydroxylated TiO 2 (110) surface shows that hydrogen bonds should be included in the model in order to perform a more accurate description of the uranyl ion sorption process. A study with this surface model is currently performed in order to calculate the relative stabilities between the different uranyl adsorption sites and to compare with the experimental data.

Sorption of trivalent Cm onto γ-Al 2 O 3 is investigated by time-resolved laser fluorescence spectroscopy (TRLFS). The pH dependent sorption of Cm(III) is followed at constant ionic strength of 0.1 M NaClO 4 under argon atmosphere in the pH range between 3.5 and 13.2 at a total Cm(III) concentration of 2.5×10 -7 mol/L. Three different surface-sorbed Cm(III) species appearing at different pH are identified by their emission spectra with peak maxima at 600.6 nm, 602.5 nm, and 605.7 nm using an excitation wavelength of λ ex =396.6 nm. No variation of the emission decay is observed for the three sorbed species (constant lifetimes of 110 μs). The pH dependent species distribution obtained from peak deconvolution of emission spectra is in agreement with earlier Eu/Cm sorption studies and sorbed species are assigned to >Al−O−Cm(OH) x (2- x ) (H 2 O) 5- x inner-sphere complexes. Indirect Cm excitation via energy transfer from the solid to the metal ion at λ ex =355 nm yields emission bands significantly shifted compared to spectra obtained at λ ex =396.6 nm. We interpret this observation as a hint for the existence of Cm inner-sphere complexes bound to different γ-Al 2 O 3 surface sites. The exact nature of those species is presently not known, but they likely involve Cm interaction with Al(OH) 3 phases at the solid surface which may even dominate the solid surface chemistry. Complementary EXAFS measurements are performed with Lu and Gd, taken as chemical homologues of the trivalent actinides. EXAFS for Gd/Lu(III) sorbed onto γ-Al 2 O 3 at pH=6 identifies a single oxygen coordination shell with a high degree of disorder. For Gd this shell consists of 9±1.5 atoms at a 2.45±0.02 Å distance. A coordination number of 7±1.5 and a Lu−O distance of 2.28±0.02 Å is found in the Lu-samples. Those values are quite close to what is known for the respective aquo ions. The large asymmetry in the coordination shell, however, suggests inner-sphere complexation. This is in agreement with TRLFS results showing clearly the existence of inner-sphere surface complexes where the first coordination shell of the trivalent metal ions consists of 5H 2 O/OH - entities. Together with EXAFS data, a metal ion coordination to four oxygen atoms belonging to the γ-alumina surface is concluded. Both spectroscopic studies indicate a certain variability of surface binding sites for trivalent metal ions.

The retention characteristics of the bentonite near-field engineered barrier proposed in most of the concepts for the deep geological disposal of high-level waste and spent fuel are an important component in repository performance assessment studies. Montmorillonite generally constitutes 65 to 90 wt.% of the bentonite. Sorption edge measurements have been performed at trace concentrations for the actinides Am(III), Np(V) and Pa(V) on purified and conditioned SWy-1 montmorillonite under anoxic, carbonate free conditions. To the best of the author´s knowledge, this is the first time a sorption data set has been measured for 233 Pa(V) uptake on montmorillonite. In some earlier work, the 2 site protolysis non-electrostatic surface complexation and cation exchange quasi mechanistic sorption model has been used to model the sorption edges measured for 11 metals with valencies between II and VI. From this study a linear correlation between strong site surface complexation constants and the corresponding hydrolysis constants was established: a Linear Free Energy Relation (LFER). The new measurements for 241 Am(III), 239 Np(V) and 233 Pa(V) fit very well on this previously established LFER. Through the LFER, a chemically consistent picture emerges which explains quantitatively why some metals tend to exhibit very high sorption and others only moderate to weak sorption.

A multi-site surface complexation/ion exchange model for dispersed MX 80 bentonite has been calibrated, considering the dissolution properties of the constituting mineral assemblage, for sorption of a large number of radionuclides, using experimental data from the present study together with well constrained literature data. Emphasis was on tri- and tetravalents actinides and fission products and reducing groundwater compositions.

In the present study the interaction of Ra(II) with calcium silicate hydrates (C-S-H) as well as with fresh and degraded hardened cement pastes (HCP) has been investigated using batch-type sorption tests. Synthetic C-S-H phases with varying CaO:SiO 2 (C:S)mol ratios, relevant to non-degraded HCP and degraded HCP, were prepared in the absence of alkalis (Na(I), K(I)) and in an alkali-rich artificial cement pore water (ACW). The Ra(II) sorption kinetics, the effect of the solid to liquid ratio and the C:S mol ratio on the Ra(II) uptake were investigated. In addition, the reversibility of Ra(II) sorption was tested using dilution tests. Comparison with data on the uptake of Sr(II) by C-S-H phases obtained from previous in-house studies showed that the same model concept could be applied to interpret the immobilization of Sr(II) and Ra(II) in these systems. Ra(II) sorption on C-S-H in the absence of alkalis could be interpreted in terms of a Ra 2+ -Ca 2+ cation exchange model. The Ra 2+ -Ca 2+ exchange selectivity coefficient, Ra Ca K C , was determined to be (6±4). There are indications that, in the presence of alkalis, the Ra(II) sorption by C-S-H phases can only be described with the help of an ion-exchange model involving Ra 2+ , Ca 2+ and the monovalent alkalis. The cation exchange model developed for C-S-H phases in the absence of alkalis was successfully applied to predict the sorption of Ra(II) on degraded HCP, assuming that the Ra(II) sorbs solely on the C-S-H fraction of the HCP. Model predictions and experimental data were found to agree well in view of the simplified sorption model used.

The sorption of NpO 2 + from aqueous solution on hydroxyapatite was studied as a function of the amount of sorbent, initial NpO 2 + concentration, ionic strength and pcH. The hydroxyapatite was characterized by SEM, EDS, XRD, FT-IR and ICP-MS measurements. At ionic strengths of 0.10 to 5.00 M NaClO 4 , the sorption increased with increased pcH to a maximum between pcH 8−8.5, then decreased as the pcH increased. The kinetics of NpO 2 + sorption on hydroxyapatite followed Lagergren first order kinetics. The temperature dependence of sorption was small in the range of 273−283 K, but increased more sharply at higher temperatures of 298−333 K. The heat of sorption of NpO 2 + was endothermic and the free energy values were exothermic indicating large, positive entropy. The activation energy for the sorption process was calculated to be 29.52±1.2 kJ/mole. The effect of calcium and phosphate on NpO 2 + sorption was studied as a function of concentration and pcH.

The mechanism of uranium(VI) interaction with pyrite was studied by solution chemistry and X-ray Photoelectron Spectroscopy (XPS). Natural pyrite was characterized by X-Ray Diffraction (XRD) and Scanning Electron Microscopy (SEM). After equilibration in 10 -2 mol L -1 NaNO 3 , pyrite was reacted with uranium(VI) by the batch method in an anoxic glove box (P O 2 <1 ppm) at ambient temperature. The reaction products of uranium, iron and sulphur were characterized (oxidation state, chemical environment) by XPS. Quantitative analysis revealed that only a few atomic percent of uranium is retained at the pyrite surface. The U4 f core level binding energies are consistent with the coexistence of an uranium(VI) species and of uranium in a reduced form. No sulphur oxidation products were observed by XPS, but spectral decomposition of the Fe2 p lines revealed the presence of iron(III) oxide or (oxy)hydroxide. These results seem to point to a redox reaction between uranium(VI) and pyrite.

Pyrite and/or galena particles previously conditioned with U(VI) at pH 6 have been characterized by micro-Raman spectrometry, voltammetry, scanning electron microscopy and high resolution X-ray photoelectron spectroscopy. U 3 O 8 is first formed on the sulphide surfaces via a redox process, then, if an excess of U(VI) remains in solution it is sorbed as UO 3 and shoepite. Galvanic phenomena occur when the minerals are in mixture in a medium containing U(VI): galena particles are oxidized whereas U(VI) is reduced at the pyrite particle surface. The maximum amounts of fixed uranium are 0.4 mmol/g for pyrite and 0.1 mmol/g for galena and the 50:50 pyrite:galena mixture.

The sorption of technetium by powdered and polished mineral stibnite Sb 2 S 3 has been investigated in simulated and natural underground waters from the Meuse/Haute–Marne region (France). The sorption by powdered stibnite has been found to be complete under both aerobic and anaerobic conditions in batch experiments. The sorption rate is higher in the absence of oxygen than under aerobic condition. Increasing the temperature from 30 °C to 60 °C results in a rise of the sorption rate by 9.1 and 27 times under anaerobic and aerobic conditions, respectively. The observed differences in sorption kinetics in the presence and in absence of oxygen are explained by the interaction of oxygen with sulfide ion in aerobic conditions and by the reduction of technetium(VII) by iron(II) and by other impurities present in natural water and in the mineral, and by the subsequent sorption of Tc(IV) on stibnite under anaerobic conditions. The sorption on a polished mineral surface resulted in the formation of a technetium film, probably Tc 2 S 7 , with a thickness of 1−3 μg Tc/cm 2 at pH 3−6 and 4−12 μg Tc/cm 2 at 9−12. The simultaneous formation of stibnite colloids with adsorbed technetium occurs at pH 9−12. The study of the technetium film on the mineral by proton induced X-ray emission analysis showed it to be at least one order of magnitude thinner on the SiO 2 impurities than on the main Sb 2 S 3 component and the iron impurities.

One of the current options for the final disposal of high level radioactive wastes is the deep geological repository (DGR). The metal canisters represent the first physical barrier to radionuclide migration towards the geosphere and their corrosion products may play a significant role in sequestering radionuclides by sorption and/or (co)precipitation phenomena. In this work, the sorption of 137 Cs on pure nanocrystalline magnetite has been experimentally studied in a wide range of pH, ionic strength, and radionuclide and sorbent concentration. The magnetite was synthesized in laboratory under controlled conditions and its main physico-chemical properties as microstructure, surface area and surface charge were analysed previous to sorption experiments. Sorption was negligible up to pH 8.5 and maximum sorption values were reached around pH 12. A significant increase in Cs sorption was observed when decreasing the ionic strength. Linear sorption isotherms were observed within the range of Cs concentration used (up to 10 -6 M). As expected, Cs showed very small sorption onto the oxide, but the whole experimental results could be satisfactorily fit with a simple model. In addition, the model developed in the oxide-electrolyte system was able to reproduce fairly well the sorption of Cs onto magnetite in two more complex waters (synthetic bentonite and cement porewaters). All the possible effects on sorption that could increase the uncertainties on the (small) sorption values were analysed in depth. In addition, the possible influence of mineral trace impurities on the sorption (for example silicates from experimental vessels), already discussed in the literature, was avoided.

This study focused on the determination of distribution coefficients ( K d ) of 137 Cs under aerobic (∼21% O 2 ) and anaerobic conditions (1−10 ppm O 2 ) which might be found in the deep geologic environment. The anaerobic conditions were simulated in a glove box filled with highly pure argon (99.9%). Both batch tests and solid-phase analyses, including X-ray diffraction, N 2 -BET surface area, and polar-microscopy/autoradiography were employed for analyzing the sorption of 137 Cs on granite, which is the potential host rock for waste disposal in Taiwan. Before starting batch tests, the rock samples were placed in the glove box for one week, and the E h values of experimental liquids were periodically measured until the E h value decreased below 0 mV. The sorption of 137 Cs was studied at various concentrations in synthetic groundwater. The Freundlich sorption isotherm seems to be adequate to quantitatively describe the sorption of Cs for the concentration ranges ( i.e. , 10 -3 −10 -6 M) conducted in both conditions. The distribution coefficient ( K ) obtained from the two-week batch tests indicated no obvious difference in two systems and the values of K and n are similar. In addition, sorption onto a specific host rock was found to be dependent on its mineralogical components. The photos obtained from polar-microscopy/autoradiography analyses showed that biotite is the principal mineral component responsible for the sorption of 137 Cs onto granite.

For establishing a reliable safety assessment methodology for high-level radioactive waste (HLW) geological disposal, it is essential to clarify the applicability of sorption data determined with a conventional batch sorption method to an actual repository environment. In the present study, sorption data represented as distribution coefficients ( K d ) were compared experimentally by applying a batch method with crushed sedimentary rock samples and a dynamic circulation method with intact sedimentary rock samples. The K d values for cesium (Cs) determined with the dynamic circulation method are around one order of magnitude lower than those obtained using the batch method, depending on Cs concentration at sorption equilibrium. This disagreement was discussed considering permeability, specific surface area, solid/liquid ratio (S/L) and solution chemistry. The K d values determined with the dynamic circulation method were corrected for different specific surface areas. As a consequence, the K d values determined with the two methods are then comparable within an order of magnitude. The K d values become even more comparable when taking into account that they decrease with increasing the S/L ratio as it is the case when going from batch to column experiments. The results basically suggest that the K d determined with the batch method are reasonably applicable to describe sorption in the actual environment in a given nuclear waste repository after appropriate corrections.

The Institut für Nukleare Entsorgung (INE) at the Forschungszentrum Karlsruhe (FZK), Germany, has constructed and is operating a beamline at the synchrotron source ANKA dedicated to actinide speciation investigations related to nuclear waste disposal as well as applied and basic actinide research. Experiments on nuclides not suited as nuclear fuel with activities up to 10 6 times the limit of exemption inside a safe and flexible containment concept are now possible. The design is for a multi-purpose beamline, i.e. , a number of methods (XAFS, surface sensitive and spatially resolved techniques) are envisaged on one and the same sample, with X-ray energies from 2472 eV (S K edge) to 23220 eV (Rh K edge). The commissioning of the INE-Beamline was officially completed in September 2005 and meanwhile regular user operation has been started. Here we present a description of this new facility and examples for X-ray absorption studies on actinides and homologues performed during the beamline commissioning phase.

Laboratory radial in- and out-diffusion experiments were performed to investigate the reversibility of the iodide (I - ) uptake by argillaceous rocks from the Tournemire site (France). At first, the suitability of the method was demonstrated by means of deuterium depleted water (DDW) diffusion experiments. The values for the DDW effective diffusion coefficient (1.7 to 2.7×10 -11 m 2 s -1 ) are indeed very close to those obtained from previous through-diffusion experiments carried out on Tournemire samples with tritiated water. The diffusion of chloride and bromide led to the determination of halide-accessible porosities, which are necessary to calculate the retardation factor ( R ) and the distribution ratio ( R D ). The calculated values for the halide-accessible porosity (2 to 5%) clearly indicated the effect of anionic exclusion and are consistent with previous data. On the contrary, the in-diffusion experiments performed with iodide clearly showed its uptake by argillite, with rock capacity factor values ranging from 14% to 25%. The corresponding values of R D (0.035 to 0.08 L kg -1 ) are one order of magnitude lower than those previously derived from batch methods. At last, the experiments of iodide out-diffusion revealed that only iodide located in the halide-accessible porosity diffused out of the rock samples, suggesting that the uptake of iodide by argillite would not be reversible or that the kinetics of desorption would be low (>70 days).

In a nuclear reactor, isotopes such as 35 Cl present as impurities in the nuclear fuel are activated by thermal neutron capture. During interim storage or geological disposal of nuclear fuel, the activation products such as 36 Cl may be released from the fuel to the geo/biosphere and contribute to the “instant release fraction” as they are likely to migrate in defects and grain boundaries. In order to differentiate diffusion mechanisms due to “athermal” processes during irradiation from thermally activated diffusion, both irradiation and thermal effects must be assessed. This work concerns the measurement of the thermal diffusion coefficient of chlorine in UO 2 . 37 Cl was implanted at a 10 13 at/cm 2 fluence in depleted UO 2 samples which were then annealed in the 900−1200 °C temperature range and finally analyzed by Secondary Ion Mass Spectrometry (SIMS) to obtain 37 Cl depth profiles. The migration process appears to be rather complex, involving mechanisms such as atomic, grain boundary, directed diffusion along preferential patterns as well as trapping into sinks before successive effusion. However, using a diffusion model based on general equation of transport, apparent diffusion coefficients could be calculated for 1000 and 1100 °C and a mean activation energy of 4.3 eV is proposed. This value is one of the lowest values compared to those found in literature for other radionuclides pointing out a great ability of chlorine to migrate in UO 2 at relatively low temperatures. In order to unequivocally determine the diffusion behaviour of both implanted and pristine chlorine before and after thermal annealing, the structural environment of chlorine in UO 2 was examined using micro X-ray fluorescence (micro-XRF) and micro X-ray absorption spectroscopy (micro-XAS).

We studied the effect of ionic strength on the interactions of Europium(III) and Curium(III) with a Gram-negative bacterium Paracoccus denitrificans . Bacterial cells grown in 0.5-, 3.5-, and 5.0% NaCl were used in adsorption experiments and laser experiments that were performed at the same ionic strengths as those in the original growth media. The distribution ratio (log K d ) for Eu(III) and Cm(III) was determined at pHs 3−5. To elucidate the coordination environment of Eu(III) adsorbed on P. denitrificans , we estimated the number of water molecules in the inner sphere and strength of the ligand field by time-resolved laser-induced fluorescence spectroscopy (TRLFS) at pHs 4−6. The log K d of Eu(III) and Cm(III) increased with an increase of pH at all ionic strengths because there was less competition for ligands in cells with H + at higher pHs, wherein less H + was present in solution: cation adsorption generally occurs through an exchange with H + on the functional groups of coordination sites. No significant differences were observed in the log K d of Eu(III) and Cm(III) at each pH in 0.5-, 3.5-, and 5.0% NaCl solutions, though competition for ligands with Na + would be expected to increase at higher NaCl concentrations. The log K d of Eu(III) was almost equivalent to that of Cm(III) under all the experimental conditions. TRLFS showed that the coordination environments of Eu(III) did not differ from each other at 0.5-, 3.5-, and 5.0% NaCl at pHs 4−6. TRLFS also showed that the characteristic of the coordination environment of Eu(III) on P. denitrificans was similar to that on a halophile, Nesterenkonia halobia , while it significantly differed from that on a non-halophile, Pseudomonas putida . These findings indicate that the number of coordination sites for Eu(III) on P. denitrificans , whose cell surface may have similar structures to that of halophiles, increased with increasing ionic strength, though their structure remained unchanged.

This paper summarizes the effect of pH on the speciation and cellular localization of uranium bound by bacterial strains isolated from the S15 deep-well montoring site, located at the Siberian radioactive subsurface depository Tomsk-7, Russia. Microbiological methods in combination with extended X-ray absorption fine structure (EXAFS) spectroscopy, transmission electron microscopy (TEM), and energy dispersive X-ray spectroscopy (EDX) were applied. EXAFS analysis showed that the cells of the two isolates, Microbacterium oxydans S15-M2 and Sphingomonas sp. S15-S1, precipitate U(VI) as m -autunite-like phase at pH 4.5, probably due to the release of inorganic phosphate from the cells as result of the microbial metabolism. However, at pH 2 uranium formed complexes with organically bound phosphates of the cell surface. The results of the EXAFS studies corroborate those found using TEM and EDX analysis. Hypotheses explaining the different coordination chemistry of uranium to bacteria as a function of pH of uranium solution in terms of solubility of m -autunite and/or microbial activity are discussed.

The presence of organic ligands in radioactive wastes is a major concern because of their potential for increasing the transport of radionuclides from disposal sites. Biotransformation of radionuclides complexed with organic ligands should precipitate the radionuclides and retard their migration. We investigated the biotransformation of Pu(IV) (10 -8 to 10 -5 M), by Pseudomonas fluorescens in the presence of excess citric acid. Analysis of 242 Pu-citrate by electrospray ionization-mass spectrometry (ESI-MS) indicated the presence of biligand Pucit 2 as the predominant complex. XANES and EXAFS analyses showed that Pu was in the +4 oxidation state and associated with citric acid as a mononuclear complex. Citric acid was metabolized by P. fluorescens at a rate of 4.9 μM/h, but in the presence of 10 -8 and 10 -6 M Pu, this rate decreased to 4.0 and 3.8 μM/h, respectively. An increase in the ionic strength of the medium from 0.18 M to 0.9 M lowered citrate metabolism by ∼65%. Pu added to the growth medium in the absence of bacteria remained as Pu(IV) in solution as a complex with citric acid. However, solvent extraction by thenoyltrifluoroacetone (TTA) and microfiltration (0.03 μm) of the medium containing bacteria after citrate biodegradation revealed the presence of polymeric Pu. The extent of formation of the Pu polymer depended on the Pu:citrate ratio, the extent of citrate metabolism, and the ionic strength of the medium.

The iodination of Gorleben and synthetic humic substances (HS) was studied complementarily by UV-visible and X-ray Photoelectron Spectroscopy (XPS). The I 2 (aq) consumption kinetics could neither be linearly correlated to [H + ] nor to HS concentration as already observed in the literature. Nevertheless, the electrophilic substitution mechanism was further evidenced by the production of both I 3 - and I - , leading to a covalent bonding. The XPS analysis of the iodinated samples confirmed the covalent bonding between iodine and carbon of HS for all the samples.

In spent nuclear fuel (SNF), the micrometer- to nanometer-sized epsilon phase (Mo-Ru-Pd-Tc-Rh) is an important host of 99 Tc which has a long half life (2.13×10 5 years) and can be an important contributor to dose in safety assessments of nuclear waste repositories. In order to examine the occurrence and the fate of the epsilon phase during the corrosion of SNF over long time periods, samples of uraninite from the Oklo natural reactors (∼ 2.0 Ga) have been investigated using transmission electron microscopy (TEM). Because essentially all of the 99 Tc has decayed to 99 Ru, this study focuses on 4 d -elements of the epsilon phase. Samples were obtained from the research collection at University of Michigan representing reactor zone (RZ) 10 (836, 819, 687) and from RZ 13 (864, 910). Several phases with 4 d -metals have been identified within UO 2 matrix at the scale of 50−700 nm; froodite, PdBi 2 , with trace amounts of As, Fe, and Te, and palladodymite or rhodarsenide, (Pd,Rh) 2 As. The most abundant 4 d -metal phase is ruthenarsenite, (Ru,Ni)As, which has a representative composition: As, 59.9; Co, 2.5; Ni, 5.2; Ru, 18.6; Rh, 8.4; Pd, 3.1; Sb, 2.4 in atomic %. Ruthenarsenite nanoparticles are typically surrounded by Pb-rich domains, galena in most cases; whereas, some particles reveal a complexly zoned composition within the grain, such as a Pb-rich domain at the core and enrichment of Ni, Co, and As at the rim. Some ruthenarsenites and Rh-Bi-particles are embedded in surrounding alteration products, e.g. , chlorite, adjacent to uraninite (no further than ∼5 μm). A few of those particles are still coated by a Pb-rich layer. Based on these results, the history that epsilon phases have experienced can be described as follows: (i) The original epsilon phase was changed to, in most cases, ruthenarsenite, by As-rich fluids with other trace metals. Dissolution and a simultaneous precipitation may be responsible for the phase change. (ii) All Mo and most of the Tc were released from the epsilon phase. Galena precipitated surrounding the 4 d -metal phases. (iii) Once the uraninite matrix has dissolved, the epsilon nanoparticles were released and “captured” within alteration phases that are immediately adjacent to the uraninite.

Rock analysis and study of groundwater data was combined with a short duration column experiment in order to better understand the behavior of rhenium (Re) in the site of Maqarin (Jordan). Results show that the combustion metamorphism having affected rock masses in Maqarin (Jordan), and led to the formation of “natural cements”, has strongly modified the spatial distribution of several elements, among which Re, creating a specific source term for this element leached by groundwaters. Unaltered cements are shown to contain up to 1.2 μg/g Re to be compared to 0.15 μg/g in the initial biomicrite. Re is rapidly leached from unaltered cements and can be used as a tracer of recent cement alteration episodes. Preliminary data suggest that Re, a chemical analogue of Tc, could be partially retained in fracture minerals.

Computer-aided surface complexation models (SCMs) are widely used in describing mineral-water interface reactions such as sorption of actinides onto (hydr)oxides and clays. Most SCMs are built upon balances of surface binding sites. In fact, such models can reproduce only competitive Langmuir isotherms, plus optional electrostatic corrections. Hence, it may be difficult to extend such SCMs to some real-surface phenomena (site heterogeneity, lateral interactions, surface precipitation) described in isotherms other than Langmuir. In this contribution, a new approach to overcome this difficulty is presented which does not use surface site balances, is not restricted only to Langmuir isotherms, and can be implemented within the Gibbs energy minimization solvers of chemical (adsorption) equilibria. It is suggested to define the ideal behavior of a surface species using a linear isotherm only. Conversely, an adsorption isotherm equation of choice can be split into a linear term (involving the standard-state density and the equilibrium constant referenced to infinite dilution) and a non-linear surface activity coefficient term (SACT, involving the site density parameter). A simple method is proposed for deriving a SACT for the monodentate surface binding from the Langmuir isotherm. This method is then applied to obtain SACTs for mono-, bi-, tri- and tetra-dentate binding from the “quasi-chemical approximation” isotherm, as well as a SACT accounting for non-ideal lateral interactions between surface species from the Frumkin isotherm. The new approach is illustrated by modeling U VI adsorption on goethite at ambient conditions using a GEM SCM. It is shown that SACTs are useful at moderate- to high surface coverage, but only when the site density parameter is fixed from independent (crystallographic) data and not adjusted to fit the titration data. Ways to enhance SCMs with BET, Freundlich and other isotherms of interest for retention of actinides and fission products on mineral surfaces are also discussed.

For the safe final disposal and/or long-term storage of radioactive wastes, deep or near-surface underground repositories are being considered world-wide. A central safety feature is the prevention, or sufficient retardation, of radionuclide (RN) migration to the biosphere. To this end, radionuclide sorption is one of the most important processes. Decreasing the uncertainty in radionuclide sorption may contribute significantly to reducing the overall uncertainty of a performance assessment (PA). For PA, sorption is typically characterised by distribution coefficients ( K d values). The conditional nature of K d requires different estimates of this parameter for each set of geochemical conditions of potential relevance in a RN´s migration pathway. As it is not feasible to measure sorption for every set of conditions, the derivation of K d for PA must rely on data derived from representative model systems. As a result, uncertainty in K d is largely caused by the need to derive values for conditions not explicitly addressed in experiments. The recently concluded NEA Sorption Project [1] showed that thermodynamic sorption models (TSMs) are uniquely suited to derive K d as a function of conditions, because they allow a direct coupling of sorption with variable solution chemistry and mineralogy in a thermodynamic framework. The results of the project enable assessment of the suitability of various TSM approaches for PA-relevant applications as well as of the potential and limitations of TSMs to model RN sorption in complex systems.

A physico-chemical model developed for spent fuel alteration was integrated in a global reactive transport model of a spent fuel disposal system, considering both decaying and stable isotopes, corroded steel canisters, bentonite backfills and a clayey host-rock. Fuel evolution took into account radiolytic-enhanced corrosion and long-term solubility-controlled dissolution as well as instantaneous release fractions. The calculations show that spent-fuel dissolution has no significant alteration effect on the near-field components except an oxidizing plume in the vicinity of the waste packages. The dissolved uranyl species, partly precipitate as schoepite on the fuel pellets, and partly diffuse in the near-field where magnetite and pyrite reduce U(VI) to yield uraninite precipitation. Under disposal conditions, preliminary calculations indicate that steel corrosion may generate sufficient dissolved hydrogen as to react with radiolytic oxidants and inhibit fuel dissolution. The formation of a protective schoepite layer could also reduce the alteration of fuel pellets. Radionuclides migration (Am, Cs, I) in the near-field is discussed in a second stage discriminating between sorption, precipitation and radioactive decay processes. The migration of Cs is translated in terms of cumulative activity profiles useful for integrated performance assessment.