Search

A broad range of contaminants has been recorded in sediments of the Loire River over the last century. Among a variety of anthropogenic activities of this nuclearized watershed, extraction of uranium and associated activities during more than 50 years as well as operation of several nuclear power plants led to industrial discharges, which could persist for decades in sedimentary archives of the Loire River. Highlighting and identifying the origin of radionuclides that transited during the last decades and were recorded in the sediments is challenging due to i) the low concentrations which are often close or below the detection limits of routine environmental surveys and ii) the mixing of different sources. The determination of the sources of anthropogenic radioactivity was performed using multi-isotopic fingerprints (

International audience; Uranium (U) isotopic signatures and concentration in sediments are widely used as paleo-redox proxies, as the behavior of U is often controlled by bottom water oxygenation. Here, we investigated the processes controlling U accumulation in the sediments of Lake Nègre (Mediterranean Alps, South-East France) over the past 9200 years. Exceptionally high natural U concentrations (350-1250 μg.g-1) allowed the measurement of U along with other elements by high-resolution X-Ray Fluorescence core-scanning. Weathering and erosion proxies (Ti content, Zr/Al and K/Ti ratios) indicate that sedimentary inputs were controlled by Holocene climatic variations. After a period of low erosion during the Holocene Climatic Optimum, a major regime shift was recorded at 4.2 kyr BP when terrigenous fluxes consistently increased until present with high sensitivity to centennial-scale climatic events. Sedimentary organic matter (OM) inputs were dominated by terrigenous OM from the catchment soils until 2.4 kyr BP, as attested by carbon to nitrogen (C/N) and bromine to organic carbon (Br/TOC) ratios. From 2.4 kyr BP to present, lake primary production and soils equally contributed to sedimentary OM. Uranium fluxes to the sediments were well correlated to terrigenous OM fluxes from 7 kyr BP to present, showing that U supply to the lake was controlled by U scavenging in the soils of the watershed followed by transport of U bound to detrital organic particles. Higher U/OM ratios before 7 kyr BP likely reflect the development of the upstream wetland. The fluctuations of U sedimentary inputs appear to be independent of bottom water oxygenation, as estimated from constant Fe/Mn ratios and δ238U isotopic signatures, and rather controlled by the production, erosion and sedimentation of terrigenous OM. This finding confirms that the use of U (and potentially other metals with high affinity to OM) concentrations alone should be used with caution for paleo-redox reconstructions.

International audience; A geostatistical approach is applied for extracting the geochemical background from gamma dose rate data acquired downstream of a former French uranium mining area. The exploratory data analysis shows that the spatial structure of the gamma dose rate consists of two components: a short isotropic component (10 m-range) that corresponds to the geochemical background; and a long anisotropic component (30–60 m-range) that corresponds to the drainage network features previously fed by the mine discharge water. The gamma dose rate on the whole area of interest was estimated and simulated (to deal with uncertainties) through a kriging of the measured values. The spatial component related to the geochemical background was then extracted through factorial kriging. The proportion of the gamma dose rate explained by the geochemical background according to factorial kriging is consistent within uncertainties with geochemical analyses performed on soil and sediment samples. This study thus highlights the potential of such geostatistical approaches for better exploiting radiometric data.

The long-term fate of uranium-contaminated sediments, especially downstream former mining areas, is a widespread environmental challenge. Essential for their management is the proper understanding of uranium (U) immobilization mechanisms in reducing environments. In particular, the long-term behavior of noncrystalline U(IV) species and their possible evolution to more stable phases in subsurface conditions is poorly documented, which limits our ability to predict U long-term geochemical reactivity. Here, we report direct evidence for the evolution of U speciation over 3,300 y in naturally highly U-enriched sediments (350–760 µg ⋅ g(−1) U) from Lake Nègre (Mercantour Massif, Mediterranean Alps, France) by combining U isotopic data (δ(238)U and ((234)U/(238)U)) with U L(3)-edge X-ray absorption fine structure spectroscopy. Constant isotopic ratios over the entire sediment core indicate stable U sources and accumulation modes, allowing for determination of the impact of aging on U speciation. We demonstrate that, after sediment deposition, mononuclear U(IV) species associated with organic matter transformed into authigenic polymeric U(IV)–silica species that might have partially converted to a nanocrystalline coffinite (U(IV)SiO(4)·nH(2)O)-like phase. This diagenetic transformation occurred in less than 700 y and is consistent with the high silica availability of sediments in which diatoms are abundant. It also yields consistency with laboratory studies that proposed the formation of colloidal polynuclear U(IV)–silica species, as precursors for coffinite formation. However, the incomplete transformation observed here only slightly reduces the potential lability of U, which could have important implications to evaluate the long-term management of U-contaminated sediments and, by extension, of U-bearing wastes in silica-rich subsurface environments.

For the first time, Diffusive Gradient in Thin-films (DGT) focuses on the inorganic iodine species iodate (IO3) and iodide (I). A silver-doped Cl resin (AgdCl), which is known to selectively accumulate I, was used to make a binding gel. Laboratory investigations were designed to verify the suitability of the AgdClDGT method to measure the total I concentration in environmental waters. Total recovery of I was obtained using an elution solution containing 100 mmol L1 KCN. DGT validation experiments in 10 mmol L1 NaCl showed linear accumulation of I over time, contrary to IO3, thus confirming the selectivity of AgdCl-binding gel. The AgdCl-DGT measurement of total I concentration was independent of pH (4.5e8.8) and was not impacted by the presence of bicarbonate (1e5 mmol L1). Finally, the performance of AgdCl-DGT samplers were tested in two continental waters and a synthetic seawater. The AgdCl-DGT samplers measured 27e33% of the total I concentration in the two continental waters up to 24 h of deployment time, whereas the AgdCl-DGT response retrieved the total I concentration in seawater up to 72 h (106 ± 7%). The difference in DGT response was attributed to the low ionic strength of the two continental waters, limiting the application of AgdCl-DGT method to media with higher ionic strength This work has been supported by the ANR (French National Research Agency), under the “Investissement d’Avenir” framework program [Number ANR-11-RSNR-0002]. Veronique Hénin (Met-rohm France) is gratefully thanked for the loan of the Metrosep Asupp 10 guard and analytical columns, and also Gilles Alcalde for the granulometric measurements of AgdCl-resin beads. All experiments were performed at LUTECE (the SEDRE's experimental platform), and elemental and speciation measurements at PATERSON (the IRSN's mass spectrometry platform). This is PATERSON contribution nº 9. Support from the Spanish Ministry of Science and Innovation is gratefully acknowledged (Project PID2019-107033 GB-C21)

One of the main factors controlling the behavior of uranium (U) in water bodies is bottom water oxygenation, which enables the use of U isotopic ratios and concentration in oceanic sedimentary records as paleo-redox proxies [1]. Here, we investigated the mechanisms governing U accumulation in the sediments of Lake Nègre (Mediterranean Alps, South-East France) over the past 9200 years. These sediments display exceptional natural U concentrations (300–1500 µg.g-1), allowing the use of high-resolution X-Ray Fluorescence core-scanning for the measurement of U. Chemical proxies (Ti content, Zr/K and K/Ti ratios) indicate that erosional sedimentary inputs were controlled by Holocene climatic variations. After a period of low erosion during the Holocene Climatic Optimum, a major regime shift was recorded at 4.2 kyr BP when terrigenous fluxes increased until present with high sensitivity to centennial-scale climatic events. The temporal evolution of carbon to nitrogen (C/N) and bromine to organic carbon (Br/TOC) ratios of sedimentary organic matter (OM) indicate that it was mostly of terrigenous origin from the catchment soils until 2.4 kyr BP. From 2.4 kyr BP to present, lake primary production and soils equally contributed to sedimentary OM. The Fe/Mn core profile points out that the lake bottom water oxygenation remained constant over the past 9200 years. Uranium depositional conditions were also constant over this period as attested by the 238U/235U ratio (expressed as δ238U) core profile. However, U fluxes to the sediments varied substantially and were correlated to terrigenous OM fluxes from 7 kyr BP to present. This correlation highlights that U supply to the lake was controlled by U scavenging in the soils of the watershed followed by transport of U bound to detrital organic particles. The fluctuations of U sedimentary inputs thus appear to be independent of bottom water oxygenation and rather controlled by climate-driven variations in terrigenous OM production and erosion. This finding confirms that the use of U (and potentially other metals with high affinity to OM) concentrations alone should be used with caution for paleo-redox reconstructions.

International audience; Determination of uranium isotope ratios is of great expedience for assessing its origin in environmental samples. In particular, the isotope ratio 236U/238U provides a powerful tool to discriminate between the different sources of uranium (uranium ore, typical ambient uranium, or anthropogenic uranium). However, in the environment, this ratio is typically below 10-8. This low abundance of 236U and the presence in large excess of major isotopes (mainly 238U and 235U) complicates the accurate detection of 236U signal by mass spectrometry and highly sensitive analytical instruments providing high abundance sensitvity are required. This work pushes the limits of triple quadrupole-based ICP-MS technology for accurate detection of 236U/238U isotope ratios down to 10-10, which is so far mainly achievable by AMS. Coupled with an efficient desolvating module, N2O was used as reaction gas in the collision reaction cell of the ICP-MS/MS. This configuration allows a significant decrease of the uranium polyatomic interferences (235UH+ ions) and an accurate determination of low 236U/238U. The proposed methodology was successfully validated through measurements of certified reference material from 10-7 to 10-9 and then through comparisons with AMS measurement results for ratios down to 10-10. This is the first time that 236U/238U isotope ratios as low as 10-10 were determined by ICP-MS/MS. Below this ratio, interferences became significant and the actual configuration reaches its limit. The possibility of measuring low 236U/238U isotope ratios open the way to a variety of geochemical studies for the determination of uranium sources in the environment.

International audience; Mineral and glass dissolution is a scientific topic deeply investigated but incompletely understood and of a great interest for the geochemical and materials science communities. If the interfacial dissolution/reprecipitation mechanism seems to be applicable to most of silicate minerals, the debate remains open concerning glass. Here we studied two model glasses, a ternary borosilicate (CJ1) and the same glass doped with 4.1 mol % of Al2O3 (CJ2). The two glasses were altered at 90°C, pH 9, and in conditions far and close to saturation with respect to amorphous silica, to determine the initial and residual rates. Moreover, a specific experiment was conducted for a short duration with a solution highly enriched with 18O and 29Si isotopes to understand how passivating gels form. SEM, TEM and ToF-SIMS characterization, along with Monte Carlo simulations were used to understand the rate limiting reactions at play and infer the role of Al. We show that Al yields a slower matrix dissolution in dilute conditions. However, it slows down the formation and the maturation of the passivating gel and favors alteration by partial hydrolysis of Si and Al entities followed by in-situ reorganization/relaxation into a porous network. Unexpectedly, CJ1 experienced both interfacial dissolution/reprecipitation and partial hydrolysis followed by in situ reorganization of the silicate network during the course of a single experiment. This study offers a unified concept that can pave the way for the future development of a predictive kinetic model based on a detailed description of bond breaking and bond forming as a function of glass composition and alteration conditions.

A new determination of the (229)Th half-life was made based on measurements of the (229)Th massic activity of a high-purity solution for which the (229)Th molality had previously been measured. The (229)Th massic activity was measured by direct comparison with SRM 4328C using 4παβ liquid scintillation counting, NaI counting, and standard addition liquid scintillation counting. The massic activity was confirmed by isotope dilution alpha spectrometry measurements. The calculated (229)Th half-life is (7825 ± 87) years (k = 2), which is shorter than the three most recent half-life determinations but is consistent with these values within uncertainties.

International audience; Chang (1998) and Murdock et al. (2001) performed numerous laboratory and field experiments to define the application scope of the AMP-DGT method. Briefly, the response of AMP-DGT was pH independent in synthetic freshwater (pH 2 – 10) and in sea waters (pH 4.5 – 10) for 8 h of deployment time. During this deployment time, no influence of Na+ (50 µmol L-1 – 0.5 mol L-1) and Ca2+ concentrations (5 µmol L-1 – 10 mmol L-1) were observed at pH 5.5 on Cs uptake by DGT from water. Correct DGT performances were obtained up to 1 month deployment time during field application in the Llyn Trawsfydd Lake (pH 6.1 – 8), before being progressively decreased to 81 % of the expected value after 5 months deployment time. The changes in DGT performances over time were possibly linked to the growth of algae / bacteria biofilm, the temperature impact on diffusion coefficient and the chemical degradation of AMP shown by a gradual loss of its characteristic yellow color. Concerning the last point, factors causing AMP binding gel degradation in DGT unit were not discussed, and their consequences on the accumulation of Cs by AMP-DGT samplers were not clearly identified. This study is dealing with an evaluation of AMP-DGT method under unfavourable conditions to maintain stable AMP binding gel in DGT unit, such as moderately basic environmental water (pH 7 – 9). Laboratory experiments were designed: (i) to identify factors controlling AMP binding gel degradation during storage and DGT deployment and (ii) to assess the possible impact of AMP degradation over time on the DGT response in waters with various chemical composition. Conclusions on potential limitations of AMP-DGT methods will finally be discussed for a better use of this technique for field applications.

International audience; A new rapid protocol for 226Ra separation and preconcentration in natural water samples was developed before its determination by Inductively Coupled Plasma Mass Spectrometry (ICP-MS). For this purpose, the commercially available Ra specific resin AnaLig® Ra-01 was used. This resin shows a high selectivity for radium in a large range of acid concentrations and no affinity or possible elution of 226Ra interfering elements. The distribution coefficients of Ra and other elements over a wide range of acid (HCl and HNO3) concentrations were obtained. Due to the high radium selectivity, the new developed protocol uses only 50 mg of dry resin and its performance was evaluated using 100 mL of three natural waters with different ionic strengths, spiked with a known quantity of 226Ra. Radium was successfully separated and preconcentrated-yielding recoveries ranging between 72% and 86%. In parallel with the characterisation of the resin sorption properties, a detailed study of polyatomic interferences was carried out in our ICP-MS allowing to identify the prominent elements favouring interferences at m/z = 226. Furthermore, a 226Ra sensitivity comparison between different ICP-MS instruments and configurations was done in order to determine high sensitivity conditions for radium analysis.

The Fukushima Dai-ichi Nuclear Power Plant (FDNPP) accident in Japan resulted in a major release of radionuclides into the environment. Compared to other radionuclides, few studies have investigated the fate of actinides in the environment. Accordingly, this research investigates the Pu composition in soil samples collected in paddy fields before and after the accident. Furthermore, the vertical distributions of Pu and U isotopic signatures, along with 137Cs activities, were measured in a sediment core collected in the Mano Dam reservoir, in the Fukushima Prefecture. Changes in the relative contributions of the major actinide sources (global fallout or FDNPP derived fallout) were investigated in sediment deposited in the reservoir. The distinct peak observed for all Pu isotope ratios (240Pu/239Pu, 241Pu/239Pu and 242Pu/239Pu) and for 137Cs concentrations in the sediment core was attributed to the Fukushima fallout, and coincided with the maximum atomic contribution of only 4.8 ± 1.0% of Pu from the FDNPP. Furthermore, 236U/238U ratios measured in the sediment core remained close to the global fallout signature indicating there was likely no U from the FDNPP accident detected in the sediment core. More research is required on the environmental dynamics of trace actinides in landscapes closer to the FDNPP where there are likely to be greater abundances of FDNPP-derived Pu and U.

International audience; Stable Pb isotope ratios were measured and compared to U distributions in three soil cores located in a wetland highly impacted by water discharge of a former U-mine. Pb isotope ratios showed notable alignments in binary mixing plots demonstrating the dissemination of radioactive-enriched material from the U-mine. Thanks to theses alignments and to the measurement of 204Pb isotope, a precise characterization of the Pb isotope composition of the U-ore was performed without the use of U-ore samples. The well-defined end-members with the help of a reevaluated isotope mixing model allowed the accurate determination of the radiogenic Pb percentages in the cores that were overall found to be >50%. Non-correlated distributions of radiogenic 206Pb and U are observed in several of the wetland soil samples studied. They reveal post-depositional U redistribution explained by major U speciation changes due to redox cycling in the wetland. On the contrary, the radiogenic 206Pb showed no or small post-depositional mobility and thus can be considered to be a memory tracer of the dissemination of U-rich radioactive material: even after an important U loss, the radiogenic 206Pb is able to reveal past contamination by U-rich materials.

International audience; Uranium mining and milling activities raise environmental concerns due to the release of radioactive and other toxic elements. Their long-term management thus requires a knowledge of past events coupled with a good understanding of the geochemical mechanisms regulating the mobility of residual radionuclides. This article presents the results on the traces of anthropic activity linked to previous uranium (U) mining activities in the vicinity of the Rophin tailings storage site (Puy de Dôme, France). Several complementary approaches were developed based on a study of the site's history and records, as well as on a radiological and chemical characterization of soil cores and a dendrochronology. Gamma survey measurements of the wetland downstream of the Rophin site revealed a level of 1050 nSv.h−1. Soil cores extracted in the wetland showed U concentrations of up to 1855 mg.kg−1, which appears to be associated with the presence of a whitish silt loam (WSL) soil layer located below an organic topsoil layer. Records, corroborated by prior aerial photographs and analyses of 137Cs and 14C activities, suggest the discharge of U mineral particles while the site was being operated. Moreover, lead isotope ratios indicate that contamination in the WSL layer can be discriminated by a larger contribution of radiogenic lead to total lead. The dendroanalysis correlate U emissions from Rophin with the site's history. Oak tree rings located downstream of the site contain uranium concentrations ten times higher than values measured on unaffected trees. Moreover, the highest U concentrations were recorded not only for the operating period, but more surprisingly for the recent site renovations as well. This integrated approach corroborates that U mineral particles were initially transported as mineral particles in Rophin's watershed and that a majority of the deposited uranium appears to have been trapped in the topsoil layer, with high organic matter content.

This study investigates the analogy between basaltic and borosilicate glasses of nuclear interest, by focusing on mechanisms controlling glass dissolution under silica saturation conditions. These conditions are representative of a non- or slowly renewed contacting solution, favouring the formation of a potentially passivating silica rich gel layer and secondary phases. Laboratory batch experiments were performed with synthetic basaltic glass altered at 90 °C, at pH 7 in a saturated 29Si-doped aqueous solution for more than 600 days. Using elemental and isotopic solution analysis and solid characterizations by SEM, TEM and ToF-SIMS, we show that basaltic glass corrodes at an unexpectedly high and constant dissolution rate of 4 × 10−3 g m−2 d−1 associated with the absence of passivating gel. Our results highlight the fact that the dissolution rate is controlled by the hydrolysis of the glassy network, sustained by the precipitation of clay-type minerals and amorphous silica. When tested in similar conditions, the International Simple Glass (ISG), a six oxide borosilicate glasses of nuclear interest displays a much lower rate limited by water diffusion through a passivating layer. The different behavior of the two glasses is explained by their ability to form secondary crystalline phases at the expense of an amorphous passivating film.

A new reference material was characterized for (229)Th molality and thorium isotope amount ratios. This reference material is intended for use in nuclear forensic analyses as an isotope dilution mass spectrometry spike. The reference material value and expanded uncertainty (k = 2) for the (229)Th molality is (1.1498 ± 0.0016) × 10(−10) mol g(−1) solution. The value and expanded uncertainty (k = 2) for the n((230)Th)/n((229)Th) ratio is (5.18 ± 0.26) × 10(−5) and the n((232)Th)/n((229)Th) ratio is (3.815 ± 0.092) × 10(−4).

International audience; High-level, long-lived nuclear waste arising from spent fuel reprocessing is vitrified in silicate glasses for final disposal in deep geologic formations. In order to better understand the mechanisms driving glass dissolution, glass alteration studies, based on silicon isotope ratio monitoring of 29Si-doped aqueous solutions, were carried out in laboratories. This work explores the capabilities of the new type of quadrupole-based ICP-MS, the Agilent 8800 tandem quadrupole ICP-MS/MS, for accurate silicon isotope ratio determination for alteration studies of nuclear waste glasses. In order to avoid silicon polyatomic interferences, a new analytical method was developed using O2 as the reaction gas in the Octopole Reaction System (ORS), and silicon isotopes were measured in mass-shift mode. A careful analysis of the potential polyatomic interferences on SiO+ and SiO2+ ion species was performed, and we found that SiO+ ion species suffer from important polyatomic interferences coming from the matrix of sample and standard solutions (0.5M HNO3). For SiO2+, no interferences were detected, and thus, these ion species were chosen for silicon isotope ratio determination. A number of key settings for accurate isotope ratio analysis like, detector dead time, integration time, number of sweeps, wait time offset, memory blank and instrumental mass fractionation, were considered and optimized. Particular attention was paid to the optimization of abundance sensitivity of the quadrupole mass filter before the ORS. We showed that poor abundance sensitivity leads to a significant shift of the data away from the Exponential Mass Fractionation Law (EMFL) due to the spectral overlaps of silicon isotopes combined with different oxygen isotopes (i.e. 28Si16O18O+, 30Si16O16O+). The developed method was validated by measuring a series of reference solutions with different 29Si enrichment. Isotope ratio trueness, uncertainty and repeatability were found to be andlt;0.2%, andlt;0.5% and andlt;0.6%, respectively. These performances meet the requirements of the studies of nuclear glasses alteration and open up possibilities to use this method for precise determination of silicon content in natural samples by Isotope Dilution.

Thermal ionization mass spectrometry (TIMS) is a well established instrumental technique for providing accurate and precise isotope ratio measurements for elements with low first ionization potential. In the nuclear domain, particularly for the study of spent nuclear fuel sample solutions, the reduction of quantities is an important issue in order to decrease the analyst exposition to the radioactive samples. This paper presents experimental results obtained for picogram quantities of samples using electron multipliers (Channeltrons) implanted on the new generation of TIMS associated with the total evaporation (TE) acquisition method. TE method presents some advantages for the measurement of low quantities of samples in relation to the whole sample consumption and the integration of the whole signal. For this study, a Thermal Ionization Mass Spectrometer from IsotopX (Isoprobe-T) equipped with a multi-collector system including Faraday cups and four movable Channeltrons is used. This work is focused on lanthanide elements, and particularly europium, in relation to the importance of their characterization in the spent fuel to validate neutronic calculation codes. In a first approach, europium isotopic ratios by total evaporation method using Faraday cups are obtained for quantities down to 250 pg. The results obtained by TE (Faraday cups only) on such low quantities confirm the potentialities of the method. A reproducibility of 2.5‰ (k = 2), was obtained on the 153Eu/151Eu ratio (n = 11), for natural samples with an accuracy around 1‰. In a second approach the same procedure is tested on Channeltron detectors by reducing the amounts of samples (from 50 pg to 1 pg). In order to obtain accurate isotope results, intrinsic parameters such as applied voltages, dead time and noise were determined. A Channeltron specific intercalibration gain method was also developed, as this parameter was found to be the major source of uncertainties using this type of detectors. The reproducibility on the 153Eu/151Eu ratio was respectively found around 3% (k = 2). These results show the potentialities of isotopic measurements of sub-nanograms quantities and the different sources of uncertainties using total evaporation and Channeltrons on the Isoprobe-T.

The French Atomic Energy Commission has carried out several experiments including the mini-INCA (INcineration of Actinides) project for the study of minor-actinide transmutation processes in high intensity thermal neutron fluxes, in view of proposing solutions to reduce the radiotoxicity of long-lived nuclear wastes. In this context, a Cm sample enriched in 248Cm (∼97%) was irradiated in thermal neutron flux at the High Flux Reactor (HFR) of the Laue-Langevin Institute (ILL). This work describes a quadrupole ICP-MS (ICP-QMS) analytical procedure for precise and accurate isotopic composition determination of Cm before sample irradiation and of Cm and Cf after sample irradiation. The factors that affect the accuracy and reproducibility of isotopic ratio measurements by ICP-QMS, such as peak centre correction, detector dead time, mass bias, abundance sensitivity and hydrides formation, instrumental background, and memory blank were carefully evaluated and corrected. Uncertainties of the isotopic ratios, taking into account internal precision of isotope ratio measurements, peak tailing, and hydrides’ formations ranged from 0.3% to 1.3%. This uncertainties’ range is quite acceptable for the nuclear data to be used in transmutation studies.

International audience; The French Atomic Energy Commission has carried out several experiments for the study of minor-actinide transmutation processes in high intensity thermal neutron flux. In this context a Cm sample enriched in $^{248}$Cm (∼97%) was irradiated in a thermal neutron flux at the High Flux Reactor (HFR) of the Laue–Langevin Institute (ILL). The precise and accurate determination of Cf isotope ratios and of $^{249}$Bk/$^{248}$Cm and $^{249}$Cf/$^{248}$Cm elemental ratios in the 248Cm irradiated sample is crucial for the calculation of actinide neutron capture cross-sections. This work describes an analytical procedure for the separation and the isotope ratio measurement of Bk and Cf in the irradiated sample. The Bk and Cf separation is based on a lanthanides separation protocol previously developed by the laboratory. Well-defined retention times for Bk and Cf were obtained by coupling the Ionic Chromatography (IC) with an ICP-QMS. All conditions of element separation by IC and the different steps of the analytical protocol in order to obtain the isotopic and elemental ratios are presented. Relative uncertainties of Cf isotopic ratios range from 0.3% to 0.5% and the uncertainty of the $^{249}$Bk/$^{248}$Cm and $^{249}$Cf/$^{248}$Cm elemental ratios are respectively 6.1% and 3.2%. This level of uncertainty for both isotopic and elemental ratios is in perfect agreement with the requirement for transmutation studies.

Among the late Quaternary Mediterranean sapropels, the S5 (125 ka) is one of the best preserved due to its high organic carbon content that has limited postdepositional oxidation. The high uranium content in this sapropel, >40 dpm g−1, makes this layer interesting for studying uranium series disequilibrium in organic-rich sediments. For this reason, the present work provides isotopic measurements of the U decay series in a S5 sapropel by applying more precise mass spectrometric methods, TIMS/MC-ICPMS, and gamma spectrometry. Assuming that U in the sapropel mostly originated from seawater the (234U/238U), (230Th/238U), (226Ra/230Th) and (231Pa/235U) activity ratios show systematic deviations from the theoretical values for a closed-system evolution of the U series over the 125 ka since sapropel formation. The radiogenic 234Urad and 226Ra show clear evidence of migration in the sapropel with modeled diffusion coefficients of (7.1 ± 1.1) × 10−12 cm2 s−1 and (1.6 ± 0.2) × 10−10 cm2 s−1, respectively. The diffusion of 234Urad cannot explain the high (230Th/238U) and (231Pa/235U) activity ratios observed in the sapropel. Two possible mechanisms or a combination of both are proposed for explaining the irregular (230Th/238U) and (231Pa/235U) activity profiles in sapropel S5. The first one is an enhanced export flux of 230Thxs and 231Paxs excesses exceeding the production rate in seawater, during the time of sapropel formation, and the second one is diffusion of authigenic Uauth in the sapropel. However, the ambiguous determination of 230Thxs and 231Paxs in the sapropel and the poorly understood processes that might lead to Uauth migration in anoxic sediments still limit a final explanation for the deviation of (230Th/238U) and (231Pa/235U) activity ratios from their expected theoretical values.