Colloidal mobilization from soil and transport of uranium in (sub)-surface waters

Assessment of 226Ra and U colloidal transport in a mining environment

The colloidal transport of trace (Fe, Al, Ba, Pb, Sr, U) and ultra-trace (226Ra) elements was studied in a mining environment. An original approach combining 0.45 μm filtered water sampling, the Diffusive Gradient in Thin films (DGT) technique, mineralogical characterization, and geochemical modelling was developed and tested at 17 sampling points. DGT was used for the truly dissolved fraction of the elements of interest, while the 0.45 μm filtration includes both colloidal and truly dissolved fractions (together referred to as total dissolved fraction). Results indicated a colloidal fraction for Al (up to 50%), Ba (up to 86%), and Fe (up to 99%) explained by the presence of submicrometric grains of kaolinite, barite, and ferrihydrite, respectively. Furthermore, the total dissolved 226Ra concentration in the water samples reached up to 10-25 Bq/L (1.2-3.0 10-12 mol/L) at 3 sampling points, while the truly dissolved aqueous 226Ra concentrations were in the mBq/L range. Such high total dissolved concentrations are explained by retention on colloidal barite, accounting for 95% of the total dissolved 226Ra concentration. The distribution of 226Ra between the truly dissolved and colloidal fractions was accurately reproduced using a (Rax,Ba1-x)SO4 solid solution, with values of the Guggenheim parameter a0 close to ideality. 226Ra sorption on ferrihydrite and kaolinite, other minerals well known for their retention properties, could not explain the measured colloidal fractions despite their predominance. This illustrates the key role of barite in such environments. The measured concentrations of total dissolved U were very low at all the sampling points (<4.5 10-10 mol/L) and the colloidal fraction of U accounted for less than 65%. U sorption on ferrihydrite could account for the colloidal fraction. This original approach can be applied to other trace and ultra-trace elements to complement when necessary classical environmental surveys usually performed by filtration on 0.45 μm.

Novel zone elution mode in coiled tube field-flow fractionation for online separation and characterization of environmental submicron particles

Coiled tube field-flow fractionation (CTFFF) is currently applied to environmental and material studies. In the present work, a novel zone elution mode in CTFFF has been proposed and developed. Zone elution mode is based on the separation of particles by stepwise decreasing the flow rate of the carrier fluid and their subsequent elution at a constant flow rate. The fractionation parameters were optimized using a mixture of standard silica submicron particles (150, 390, and 900 nm). Taking samples of volcanic ash as examples, it has been demonstrated that zone elution mode can be successfully used for the fractionation of environmental nano- and submicron particles. For the first time, CTFFF was coupled online with a dynamic light scattering detector for the size characterization of eluted particles. Zone elution in CTFFF can serve for the further development of hyphenated techniques enabling efficient fractionation and size/elemental characterization of environmental particles in nano- and submicrometric size ranges.

Uranium in phosphate rocks and mineral fertilizers applied to agricultural soils in East Africa

Phosphate rock, pre-concentrated phosphate ore, is the primary raw material for the production of mineral phosphate fertilizer. Phosphate rock is among the fifth most mined materials on earth, and it is also mined and processed to fertilizers in East Africa. Phosphate ore can contain relevant heavy metal impurities such as toxic cadmium and radiotoxic uranium. Prolonged use of phosphate rock powder as a fertilizer and application of mineral fertilizers derived from phosphate rock on agricultural soils can lead to an accumulation of heavy metals that can then pose an environmental risk. This work assesses the uranium concentrations in four major phosphate rocks originating from East Africa and four mineral phosphate fertilizers commonly used in the region. The concentration measurements were performed using energy-dispersive X-ray fluorescence spectrometry. The results showed that the uranium concentration in phosphate rock ranged from as low as 10.7 mg kg^-1 (Mrima Hill deposit, Kenya) to as high as 631.6 mg kg^-1 (Matongo deposit, Burundi), while the concentrations in phosphate fertilizers ranged from 107.9 for an imported fertilizer to 281.0 mg kg^-1 for a local fertilizer produced from Minjingu phosphate rock in Tanzania. In this context, it is noteworthy that the naturally occurring concentration of uranium in the earth crust is between 1.4 and 2.7 mg kg^-1 and uranium mines in Namibia commercially process ores with uranium concentrations as low as 100-400 mg kg^-1. This study thus confirms that East African phosphate rock, and as a result the phosphate fertilizer produced from it can contain relatively high uranium concentrations. Options to recover this uranium are discussed, and it is recommended that public-private partnerships are established that could develop economically competitive technologies to recover uranium during phosphate rock processing at the deposits with the highest uranium concentrations.

Distribution, behavior, and erosion of uranium in vineyard soils

Phosphate fertilization contributes to an input of uranium (U) in agricultural soils. Although its accumulation and fate in agricultural soils have been previously studied, its colloidal transport and accumulation along slopes through erosion have been studied to a lesser extent in viticulture soils. To bridge this gap, the contents and potential mobility of U were investigated in vineyard model soils in the Rhineland-Palatinate region, Germany. In addition to elevated U contents, U was expected to associate with colloids and subject to erosion, thus accumulating on slope foots and in soils with fine structure, and reflecting a greater variability. Moreover, another expectation was the favorable erosion/mobility of U in areas with greater carbonate content. This was tested in three regional locations, at different slope positions and through soil horizon depths, with a total of 57 soil samples. The results show that U concentrations (0.48-1.26 ppm) were slightly higher than proximal non-agricultural soils (0.50 ppm), quite homogenous along slope positions, and slightly higher in topsoils. Assuming a homogeneous fertilization, the vertical translocation of U in soil was most probably higher than along the slope by erosion. In addition, carbonate content and soil texture correlated with U concentrations, whereas other parameters such as organic carbon and iron contents did not. The central role of carbonate and soil texture for the prediction of U content was confirmed using decision trees and elastic net, although their limited prediction power suggests that a larger sample size with a larger range of U content is required to improve the accuracy. Overall, we did not observe neither U nor colloids accumulating on slope foots, thus suggesting that soils are aggregate-stable. Lastly, we suggested considering further soil parameters (e.g., Ca²⁺, phosphorus, alkali metals) in future works to improve our modelling approach. Overall, our results suggest U is fortunately immobile in the studied locations.

Cotransport of uranyl carbonate loaded on amorphous colloidal silica and strip-shaped humic acid in saturated porous media: Behavior and mechanism

Uranyl carbonate (UC(VI)) is a stable form of uranyl (U(VI)) that widely coexists with amorphous colloidal silica (ACSi) and humic acid (HA) in carbonate-rich U-contaminated areas. In this context, the cotransport behavior and mechanism of UC(VI) with ACSi (100 mg L^-1) and HA colloids in saturated porous media were systematically investigated. It was found that the ACSi and strip-shaped HA have a strong adsorption capacity for UC(VI), and their adsorption distribution coefficient (K_d) is 4-5 orders of magnitude higher than that of quartz sand (QS). In the ternary system, UC(VI) was mainly existing in the colloid-associated form at low UC(VI) concentration (4.2 × 10^-6 M). Compared with the individual transport of UC(VI), the presence of ACSi and strip-shaped HA in the binary system promotes the transport of low-concentration UC(VI) (4.2 × 10^-6 M) but shows a hindering effect when UC(VI) = 2.1 × 10^-5 M. When ionic strength (IS) increased from 0 to 100 mM, the individual transport of UC(VI) and ACSi was weakened owing to the masking effect and the compression of the electrical double layer, respectively; this weakening effect is more pronounced in the binary (UC(VI)-ACSi) system. Notably, the transport of UC(VI) and ACSi in the ternary system is independent of the changes in IS due to the surface charge homogeneity strengthening the electrostatic repulsion between HA and QS. The Derjaguin-Landau-Verwey-Overbeek theory and retention profiles reveal the co-deposition mechanism of ACSi and UC(VI) in the column under different hydrochemical conditions. The nonequilibrium two-site model and the mathematical colloidal model successfully described the breakthrough data of UC(VI) and ACSi, respectively. These results are helpful for evaluating the pollution caused by UC(VI) migration in an environment rich in HA and formulating corresponding effective control strategies.

Field-flow fractionation for nanoparticle characterization

This review presents field-flow fractionation: The elements of theory enable the link between the retention and the characteristics of the nanometer-sized analytes to be highlighted. In particular, the nature of force and its way of being applied are discussed. Four types of forces which determine four types of techniques were considered: hydrodynamic, sedimentation, thermal, and electrical; this is to show the importance of the choice of technique in relation to the characterization objectives. Then the separation performance is presented and compared with other separation techniques: field-flow fractionation has the greatest intrinsic separation capability. The characterization strategies are presented and discussed; on the one hand with respect to the characteristics needed for the description of nanoparticles; on the other hand in connection with the choice of the nature of the force, and also of the detectors used, online or offline. The discussion is based on a selection of published study examples. Finally, current needs and challenges are addressed, and as response, trends and possible characterization solutions.

Characterization of volcanic ash nanoparticles and study of their fate in aqueous medium by asymmetric flow field-flow fractionation-multi-detection

Dimensional and elemental characterization of environmental nanoparticles is a challenging task that requires the use of a set of complementary analytical methods. Asymmetric flow field-flow fractionation coupled with UV-Vis, multi-angle laser light scattering and ICP-MS detection was applied to study the nanoparticle fraction of a volcanic ash sample, in a Milli-Q water suspension at pH 6.8. It has been shown that the separated by sedimentation nanoparticle fraction of the Klyuchevskoy volcano ash suspension contains 3 polydisperse populations for which size ranges (expressed in gyration radius, rG), hydrodynamic behaviours (evaluated via shape index) and elemental compositions are different. These 3 populations did not dissolve over the 72-h study but aggregated and settled out differently. Thus, the population of particles with gyration radii <140 nm (P1), which contained 6% Al₂O₃ and represented approximately 20% by mass of the nanoparticle fraction, remained in suspension without observable aggregation. The populations P2 and P3, which represented 67% and 13% by mass in the initial suspension, covered the rG range 25-250 nm and contained 17% and 15% Al₂O₃, respectively. Over time, populations P2 and P3 aggregated and their concentration in suspension at 72 h decreased by approximately 40% compared with the initial suspension. The decrease of these nanoparticle populations occurred either from the beginning of the temporal monitoring (P2) or after 30 h (P3). Aggregation generated a new population (P4) in suspension with rG up to 300 nm and mostly consisting of P2. This population represented only up to 6 to 7% of the nanoparticle fraction and decreased beyond 50 h. As a result, the trace elements present in the nanoparticle fraction and monitored (Cu and La) were also no longer found in the suspension. The results obtained can offer additional insights into the fate of volcanic ash nanoparticles in the environment.

Co-transport of uranyl carbonate and silica colloids in saturated quartz sand under different hydrochemical conditions

Uranyl carbonate (UC) and silica colloids (cSiO₂) are widely distributed in carbonate-rich subsurface environments associated with uranium pollution. Mobile colloids such as cSiO₂ can affect uranium's transport efficiency in the groundwater environment. Therefore, elucidating the mechanism of UC and cSiO₂ co-transport in a saturated porous medium with different ionic strength (IS), pH, and UC concentration is essential for the prevention and control of groundwater radioactive pollution. At low UC concentrations (<2.1 × 10^-5 M), cSiO₂ is more prone to be deposited on the surfaces of quartz sand (QS) than UC, resulting in cSiO₂ preventing UC transport. Compared to pH 7 and 9, at pH 5 the adsorption of uranium [in the form of 81.5% UO₂CO₃(aq), 8.6% UO₂²⁺, and 5.2% UO₂OH⁺] on cSiO₂ renders cSiO₂ more prone to aggregate, causing smaller amounts of cSiO₂ (86.6%) and UC (55.8%) to be recovered. Mechanisms responsible for the evolution of the pH and zeta potential in effluents have been proposed. Chemical reactions (ligand-exchange reactions and deprotonation) that occur in the QS column between UC and cSiO₂/QS cause the pH of the suspension to varying, which in turn causes changes in the zeta potential and particle size of cSiO₂. Eventually, the recovery rates of cSiO₂ and UC are changed, depending upon the colloid particle size. Changes in ionic strength can seriously affect the stability of cSiO₂ particles, and that effect is more significant when UC is present. Moreover, colloidal filtration theory, a non-equilibrium two-site model, and the Derjaguin-Landau-Verwey-Overbeek theory successfully describe the individual-transport and co-transport of cSiO₂ and UC in the column. This study provides a strong basis for investigating UC pollution control in porous media.

Centrifugal ultrafiltration preconcentration for studying the colloidal phase of a uranium-containing soil suspension

The objective of this work was to explore centrifugal ultrafiltration (UF) to separate and / or preconcentrate natural colloidal particles for their characterization. A soil suspension obtained by batch leaching was used as a laboratory reference sample. It was preconcentrated with concentration factors (CF) varying from 10 to 450. The dimensional analysis of the colloidal phase was carried out by Asymmetric Flow Field-Flow Fractionation (AF4)-multidetection. The colloidal masses were estimated by mass balance of the initial suspension, its concentrates and filtrates. The size-dependent distribution (expressed in gyration radius) and total colloidal mass (especially recovery), as well as chemical composition and concentration (including species partitioning between dissolved and colloidal phases) were determined to assess the effects of UF preconcentration on colloidal particles. The gyration radius of the colloidal particles recovered in these concentrated suspensions ranged from about 20 nm to over 150 nm. Neither de-agglomeration nor agglomeration was observed. However, only (64 ± 4) % (CF = 10) of the colloidal particles initially in the soil suspension were found in the recovered concentrated suspensions, and this percentage decreased as CF increased. The filter membrane trapped all other particles, mainly the larger ones. Whatever the CF, the centrifugal UF did not appear to change the dissolved-colloidal partitioning of certain species (Al, organic carbon); whereas it led to an enrichment of the colloidal phase for others (Fe, U). The enrichment rate was specific to each species (15% for Fe; 100% for U). By fitting the observed trends (i.e. conservation, depletion or enrichment of the colloidal phase in the concentrate) as a function of CF, the colloidal concentrations (total and species) were assessed without bias. This methodology offers a new perspective for determining physicochemical speciation in natural waters, with a methodology applicable for environmental survey or site remediation studies.

Colloidal Size and Redox State of Uranium Species in the Porewater of a Pristine Mountain Wetland

Uranium (U) speciation was investigated in anoxically preserved porewater samples of a natural mountain wetland in Gola di Lago, Ticino, Switzerland. U porewater concentrations ranged from less than 1 μg/L to tens of μg/L, challenging the available analytical approaches for U speciation in natural samples. Asymmetrical flow field-flow fractionation coupled with inductively coupled plasma mass spectrometry allowed the characterization of colloid populations and the determination of the size distribution of U species in the porewater. Most of the U was associated with three fractions: <0.3 kDa, likely including dissolved U and very small U colloids; a 1-3 kDa fraction containing humic-like organic compounds, dispersed Fe, and, to a small extent, Fe nanoparticles; and a third fraction (5-50 nm), containing a higher amount of Fe and a lower amount of organic matter and U relative to the 1-3 kDa fraction. The proportion of U associated with the 1-3 kDa colloids varied spatially and seasonally. Using anion exchange resins, we also found that a significant proportion of U occurs in its reduced form, U(IV). Tetravalent U was interpreted as occurring within the colloidal pool of U. This study suggests that U(IV) can occur as small (1-3 kDa), organic-rich, and thus potentially mobile colloidal species in naturally reducing wetland environments.