Jost Eikenberg

Rare earth elements and Sr and Nd isotopic compositions of dissolved and suspended loads from small river systems in the Vosges mountains (France), the river Rhine and groundwater

Abstract The aim of this study is to characterize the evolution of the rare earth elements (REE) in non-mature streams from small catchment areas in the Vosges mountains downstream to more mature plain rivers including the river Rhine. The dissolved load REE distribution patterns of the low-pH Vosges streams are very different from those of high-pH plain rivers indicating that different physico-chemical parameters control REE transport in these different water systems. The plain rivers and groundwater show similar REE distribution patterns with a strong negative Ce anomaly and heavy rare earth element (HREE) enrichment. In addition, the river Rhine has a positive Gd anomaly which is of anthropogenic origin. Similar to the worlds major rivers the light rare earth element (LREE) enrichment in the plain rivers is mainly pH controlled. Their Sm/Nd and 143 Nd / 144 Nd isotope ratios are close to average continental crust values. This is not the case for the less evolved, non-mature and low pH Vosges streams. Their high Sm/Nd and 143 Nd / 144 Nd ratios but low 87 Sr / 86 Sr ratios suggest that chemical alteration of accessory middle rare earth element (MREE) enriched minerals such as apatites from rocks in the catchment area control the REE abundances of these waters. A comparison of the dissolved load REE distribution patterns with those of the principal lithologies in the corresponding drainage basins illustrates that especially the Eu anomalies of the Vosges streamlets are strongly lithology dependent. Leaching experiments indicate that the suspended load is isotopically very inhomogeneous. The REE distribution patterns of the suspended load leachates have no similarities with those of the corresponding dissolved load indicating that the leachable reservoir not only contains adsorbed REE but also REE from leachable mineral phases. Their 143 Nd / 144 Nd , Sm/Nd and 87 Sr / 86 Sr ratios support this suggestion being always higher and lower, respectively, than those of their corresponding dissolved loads. They rather point to the presence of relic primary apatite in the leachable portion of the suspended load. The flat PAAS normalized REE distribution patterns, the high 87 Sr / 86 Sr and low 143 Nd / 144 Nd isotopic ratios of the corresponding residues suggest secondary silicate phases such as clay minerals in the residual phase of the suspended load. Leachates and corresponding residual phases define alignments in the Sm/Nd isochron diagram whose slopes define ages ranging between 200 and 390 Ma indicating that the mineral phases in the suspended load retain some memory of their primary precursor minerals in the Hercynian granitic source rocks.

The colloid and radionuclide retardation experiment at the Grimsel Test Site: influence of bentonite colloids on radionuclide migration in a fractured rock

Abstract The colloid and radionuclide retardation (CRR) experiment is dedicated to the study of the in situ migration behaviour of selected actinides and fission products in the absence and presence of bentonite colloids in a water-conducting feature (shear zone) in the Grimsel Test Site (GTS). The technical scenario considers the bentonite backfill/host rock interface as a potential source for colloids. The experiment investigates the migration behaviour of U, Th, Pu, Am, Np, Sr, Cs, I and Tc and the influence of smectitic bentonite colloids by two in situ tracer injections in a well-characterised dipole. The field experiments are supported by an extended laboratory and modelling programme. Colloid breakthrough is determined on-line by a mobile, laser-induced breakdown detection apparatus (LIBD), a mobile photon correlation spectrometer (PCS) and, afterwards in the laboratory, by a single particle counting method (SPC) using a laser light scattering technique. Bentonite colloids generated from bentonite backfill material were found to be stable in the experimental groundwater and the influence of pH and salinity on colloid stability was investigated. The in situ monitored breakthrough of the tri- and tetravalent actinides Am and Pu and of Cs followed the colloid breakthrough indicating some degree of colloid-mediated migration of these radionuclides in the experimental shear zone. But even when no colloids had been added to the tracer cocktail, part of Am(III) and Pu(IV) appears to migrate as colloids. The different colloid detection techniques revealed a colloid recovery between 80 and 90% of the injected bentonite colloids. The CRR experimental results are considered from the perspective of understanding the likely long-term behaviour of a deep geological repository for radioactive waste and as an indicator of the way forward to the next generation of in situ experiments.

Results of the colloid and radionuclide retention experiment (CRR) at the Grimsel Test Site (GTS), Switzerland: impact of reaction kinetics and speciation on radionuclide migration

Summary The influence of smectite colloids on the migration behaviour of U(VI), Th(IV), Pu(IV), Am(III), Np(V), Sr(II) and Cs(I) is investigated within the Colloid and Radionuclide Retardation experiment (CRR). Two in situ experiments in a well-characterized granitic fracture zone are carried out in presence and absence of bentonite colloids. Radionuclide retardation observed in the field studies increases in the sequence Np(V)∼U(VI)<Sr(II)<Cs(I), where a small fraction of colloid borne breakthrough is only stated for Cs(I) in presence of bentonite colloids. Am(III) and Th/Pu(IV) mainly migrate as colloids without retardation in the presence and absence of smectitic colloids. The radionuclide migration behaviour is discussed on the basis of results obtained in laboratory batch sorption experiments and spectroscopic studies. Consistent with the field observation, laboratory derived Kd values increase in the order Np(V)∼U(VI)<Sr(II)<Cs(I). Significant kinetic hindrance for the sorption to fault gauge minerals is observed for Sr(II) and Cs(I), but notably for Am(III) and Pu(IV). The slow sorption reaction of tri- and tetravalent actinide ions is explained by their kinetically hindered dissociation from colloidal species. In order to explain the colloidal behaviour of tri- and tetravalent actinides even in absence of bentonite colloids, ultracentrifugation and spectroscopic experiments are performed. It is found that up to 60% of Pu(IV) and Am(III) species can be centrifuged off. Adding Cm(III) (5×10-8 mol L-1) into both injection solutions instead of Am(III) allows for a spectroscopic study by using the time resolved laser fluorescence spectroscopy (TRLFS). Peak position and fluorescence lifetimes (λ=604 nm, τ=110-114 μs) together with the fact that Cm(III) can be widely separated by ultracentrifugation, suggest the existence of inner-sphere surface complexes on groundwater and bentonite colloids. Carbon K-edge XANES analysis of the bentonite colloids reveal the presence of natural organic constituents. They are mainly of aliphatic nature containing high fractions of carboxylate groups. A contribution of these organic species towards the bentonite colloid stability and sorption of actinides is assumed to be likely.

Determination of 228ra, 226Ra and 224Ra in natural water via adsorption on MnO2-coated discs

A fast procedure based on sorption of Ra on MnO2 coated polyamide discs is presented for determination of radium isotopes (i.e. 228Ra, 226Ra, 224Ra) in aqueous samples. The sample discs can be used directly for low-level alpha-spectrometry without the need for further separation and preparation methods to produce planar sample sources. While the activity of alpha-emitting 224Ra and 226Ra can be determined during a first measurement, beta-emitting 228Ra is obtained via ingrowth of the progeny 228Th on the same sample disc after a standing time of about six months. Calculations are presented for optimizing the analytical accuracy as well as for predicting the sorption yield or chemical recovery of radium on the sample disc as a function of exposure time because the sorption uptake proceeds with first-order kinetics. The analyses can be carried out on small samples of 0.5-11 and, for long counting times of one week and use of high-purity silicon surface barrier detectors, a detection limit of 0.15 mBq l-1 is obtained for 226Ra. Since the half-life of 224Ra is only 3.7 d and since 228Th (as a measure for 228Ra) is built up only partially on the sample disc, a slightly higher detection limit of 0.24 mBq l-1 results for the latter isotopes. The procedure is therefore sufficiently sensitive to allow the investigation of Ra isotope relationships in aquifers at typical environmental levels.

Precise two chronometer dating of Pleistocene travertine: The 230Th/234U and 226Raex/226Ra(0) approach

In order to determine the geochemical evolution of a freshwater limestone cave system located in central Switzerland (Hell Grottoes at Baar/Zug,) young postglacial tufaceous limestone and travertine precipitates were investigated using the 230 Th/ 234 U ingrowth system. Additional analyses of further radionuclides within the 238 U decay chain, i.e. 226 Ra and 210 Pb, showed that the Th/ U chronometer started with insignificant inherited 230 Th over the entire formation period of the travertine setting (i.e. 230 Th(0)=0). A contribution from detritalimpurities with 230 Th/ 234 U in secular equilibrium could be precisely subtracted by applying isochron dating of cogenetic phases and recently formed travertine. The resulting precise 230 Th/ 234 U formation ages were found to be consistent with the geological stratigraphy and were furthermore used to demonstrate the applicability of the next geologically important chronometer in the 238 U-decay series, based on decay of excess 226 Ra normalized to the initial, i.e. 226 Raex/ 226 Ra(0). This system is suitable for dating phases younger than 7000 yr when the correction of a detritus component increasingly limits the precision of the 230 Th/ 234 U chronometer. Analytical solutions of the coupled 234 U/ 230 Th/ 226 Ra radionuclide system predicted that the 226 Raex/ 226 Ra(0) chronometer is independent of the actual 230 Th activity build up from decay of 234 U, if the systems starts with zero inherited 230 Th(0). The data set confirmed this hypothesis and showed furthermore that the initially incorporated 226 Ra excess must have remained almost uniform in all limestone over a period of at least 7000 yr, i.e. 4–5 half-lives of 226 Ra. This is concluded because (i) the 226 Raex/ 226 Ra(0) ages agreed well with those derived from 230 Th/ 234 U, (ii) all data plot within uncertainty on the 226 Raex/ 226 Ra(0) decay curve and (iii) the atomic Ba/Ca ratio was found to be constant in the travertine materialindependent of the sample ages. Provided that such boundary conditions hold, 226 Raex/ 226 Ra(0) should be applicable to materials which are suitable for 230 Th/ 234 U dating in sedimentology and oceanography, i.e. travertine, corals, phosphorites, etc., and should strongly support 230 Th/ 234 U for samples that have been formed a few thousand years ago. # 2001 Elsevier Science Ltd. All rights reserved.

A natural analogue of high pH cement pore waters from the Maqarin area of northern Jordan. II: results of predictive geochemical calculations

Abstract Chemical thermodynamic models (geochemical codes) are extensively used in the performance assessment of radioactive waste repositories. The applicability of such models (and their databases) can be tested by comparison of predictions of elemental solubility and speciation with observations in nature. This is particularly difficult for the hyperalkaline conditions found within cement-filled repositories but the rather anomalous geochemistry of the Maqarin area provides an ideal opportunity for such comparisons. This paper presents the first model predictions of the solubility and speciation of U, Th, Ra, Pb, Se, Sn and Ni which were made in advance of the sampling exercise presented in a companion paper (Khoury et al., this issue). Initial results indicate the significance of defining the major Ni, Se and Sn phases in detail and suggest that such work is of little use for U due to large uncertainties in the solute speciation databases.

Electrodeposition of actinides for alpha-spectrometry

The electrodeposition of the actinides U, Th, Pu, Am, and Cm on stainless steel disks was studied. A sodium hydrogen sulphate — sodium sulphate buffer 0.63M in total sulphate and a ratio sodium hydrogen sulphate to total sulphate ranging from 0.16 (pH 1.8) to 0.48 (pH 1.3) allows the quantitative electrodeposition of the given actinides. No actual pH measurement is needed. Moreover, sodium hydrogen sulphate is an excellent substrate for the wet ashing of organic matter prior to electrodeposition. The method is very robust.

Availability of Zinc and the Ligands Citrate and Histidine to Wheat: Does Uptake of Entire Complexes Play a Role?

Organic ligands in soils affect the availability of trace metals such as Zn to plants. This study investigated the effects of two of these ligands, citrate and histidine, on Zn uptake by wheat under hydroponic conditions. Uptake of (65)Zn in the presence of these ligands was compared to uptake in the presence of EDTA at the same free Zn concentration (Zn(2+) ~ 50 nM). In the presence of citrate Zn root uptake was enhanced ~3.5 times and in the presence of histidine, by a factor of ~9, compared to the EDTA treatments. Citrate uptake was slightly reduced in the treatment containing ligands and Zn compared to the treatment containing the same ligand concentration but no Zn. In addition, a higher uptake of Zn than of citrate was observed. This suggests that the enhanced Zn uptake was primarily due to increased supply of Zn(2+) by diffusion and dissociation of Zn-citrate complexes at the root surface. Histidine uptake was much higher than citrate uptake and not influenced by the presence of Zn. As histidine forms stronger complexes with Zn than citrate, the results suggest that the enhancement of Zn uptake in the presence of histidine was in part due to the uptake of undissociated Zn-histidine complexes.

228Ra/226Ra/224Ra and 87Sr/86Sr isotope relationships for determining interactions between ground and river water in the upper Rhine valley

Ground and river waters of the upper Rhine valley (Alsace, France) were investigated for chemical composition of the major elements, Sr isotopes and radionuclides from the U and Th series. In particular, the isotope ratios and concentrations of Ra and Sr were used as geochemical tracers to distinguish between different types of water and their interactions. The bulk chemical analyses suggest that the surface waters in the Rhine valley can be described as mixtures between Ca-Na-HCO3-rich ground water and less mineralized slightly acidic river waters which have migrated through crystalline (mainly granitic) basement rocks of the Vosges mountains. Mixing of these waters yields positive correlation between bulk Sr, U, Ca and HCO3, indicating that carbonate-rich sediments are the main source of U and (non-radiogenic) Sr in the Rhine valley aquifers. The combination of the Ra and Sr isotope systems (228Ra/226Ra, 87Sr/86Sr) shows, however, that probably three sources contribute to the surface river waters in the upper Rhine valley, i.e. (i) a highly radiogenic crystalline component, (ii) a ground water source and, (iii) a third component from infiltrating Rhine water along the flow path of the parallel running river Ill in the northerly direction. The Sr and Ra isotope systems were also used to calculate small-scale mixing fractions of tributaries along the flow path of the Ill. Mixing ratios of non-pure end-member waters were determined using three isotope diagrams (i.e. 224Ra/226Ra vs. 228Ra/226Ra) and the results obtained with the Ra isotope system were found to be consistent with the data using Sr isotope relationships (i.e. 87Sr/86Sr vs. 1/Sr).

Testing models of trace element geochemistry at Poços de Caldas

Abstract In order to test the geochemical models used in repository performance assessment, modelling groups were provided with selected major element analyses of Pocos de Caldas groundwaters and asked to predict “blind” the solubility, speciation and limiting solid for a number of trace elements. This paper documents these predictions and compares them to field analyses. These tests illustrate particular stengths and weaknesses in current models/databases and allow recommendations for amendments/improvements to be made.