Comprehensive micropollutant screening using LC-HRMS/MS at three riverbank filtration sites to assess natural attenuation and potential implications for human health.

Riverbank filtration (RBF) is used worldwide to produce high quality drinking water. With river water often contaminated by micropollutants (MPs) from various sources, this study addresses the occurrence and fate of such MPs at three different RBF sites with oxic alluvial sediments and short travel times to the drinking water well down to hours. A broad range of MPs with various physico-chemical properties were analysed with detection limits in the low ng L-1 range using solid phase extraction followed by liquid chromatography coupled to tandem high resolution mass spectrometry. Out of the 526 MPs targeted, a total of 123 different MPs were detected above the limit of quantification at the three different RBF sites. Of the 75-96 MPs detected in each river, 43-59% were attenuated during RBF. The remaining total concentrations of the MPs in the raw drinking water accounted to 0.6-1.6 µgL-1 with only a few compounds exceeding 0.1 µgL-1, an often used threshold value. The attenuation was most pronounced in the first meters of infiltration with a full elimination of 17 compounds at all three sites. However, a mixing with groundwater related to regional groundwater flow complicated the characterisation of natural attenuation potentials along the transects. Additional non-target screening at one site revealed similar trends for further non-target components. Overall, a risk assessment of the target and estimated non-target compound concentrations finally indicated during the sampling period no health risk of the drinking water according to current guidelines. Our results demonstrate that monitoring of contamination sources within a catchment and the affected water quality remains important in such vulnerable systems with partially short residence times.

The thermal impact of subsurface building structures on urban groundwater resources - A paradigmatic example.

Shallow subsurface thermal regimes in urban areas are increasingly impacted by anthropogenic activities, which include infrastructure development like underground traffic lines as well as industrial and residential subsurface buildings. In combination with the progressive use of shallow geothermal energy systems, this results in the so-called subsurface urban heat island effect. This article emphasizes the importance of considering the thermal impact of subsurface structures, which commonly is underestimated due to missing information and of reliable subsurface temperature data. Based on synthetic heat-transport models different settings of the urban environment were investigated, including: (1) hydraulic gradients and conductivities, which result in different groundwater flow velocities; (2) aquifer properties like groundwater thickness to aquitard and depth to water table; and (3) constructional features, such as building depths and thermal properties of building structures. Our results demonstrate that with rising groundwater flow velocities, the heat-load from building structures increase, whereas down-gradient groundwater temperatures decrease. Thermal impacts on subsurface resources therefore have to be related to the permeability of aquifers and hydraulic boundary conditions. In regard to the urban settings of Basel, Switzerland, flow velocities of around 1 md-1 delineate a marker where either down-gradient temperature deviations or heat-loads into the subsurface are more relevant. Furthermore, no direct thermal influence on groundwater resources should be expected for aquifers with groundwater thicknesses larger 10m and when the distance of the building structure to the groundwater table is higher than around 10m. We demonstrate that measuring temperature changes down-gradient of subsurface structures is insufficient overall to assess thermal impacts, particularly in urban areas. Moreover, in areas which are densely urbanized, and where groundwater flow velocities are low, appropriate measures for assessing thermal impacts should specifically include a quantification of heat-loads into the subsurface which result in a more diffuse thermal contamination of urban groundwater resources.

Photo-redox reactions of dicarboxylates and α-hydroxydicarboxylates at the surface of Fe(III)(hydr)oxides followed with in situ ATR-FTIR spectroscopy.

Colloidal mineral-phases play an important role in the adsorption, transport and transformation of organic and inorganic compounds in the atmosphere and in aqueous environments. Artificial UV-light and sunlight can induce electron transfer reactions between metal ions of the solid phases and adsorbed compounds, leading to their transformation and degradation. To investigate different possible photo-induced oxidation pathways of dicarboxylates adsorbed on iron(III)(hydr)oxide surfaces, we followed UV-A induced photoreactions of oxalate, malonate, succinate and their corresponding α-hydroxy analogues tartronate and malate with in situ ATR-FTIR spectroscopy in immersed particle layers of lepidocrocite, goethite, maghemite and hematite at pH 4. UV-A light (365 ± 5 nm) lead to fast degradation of oxalate, tartronate and malate, while malonate and succinate were photo-degraded at much slower rates. Efficient generation of OH-radicals can be excluded, as this would lead to fast and indiscriminate degradation of all tested compounds. Rapid photo-degradation of adsorbed oxalate and the α-hydroxydicarboxylates must be induced by direct ligand-to-metal charge transfer (LMCT) or by selectively oxidizing valence band holes, both processes requiring inner-sphere coordination with direct ligand-to-metal bonds to enable efficient electron-transfer. The slow photo-degradation of malonate and succinate can be explained by low-yield production of OH-radicals at the surface of the iron(III)(hydr)oxides.

Photoreductive dissolution of iron(III) (hydr)oxides in the absence and presence of organic ligands: experimental studies and kinetic modeling.

This study investigated the kinetics of the photoreductive dissolution of various iron(III)(hydr)oxide phases, lepidocrocite (gamma-FeOOH), ferrihydrite, and hydrous ferric oxide, in the absence of organic ligands as a function of pH in deaerated and aerated suspensions. Photoreductive dissolution of lepidocrocite and ferrihydrite only occurred below pH 6. Under oxic conditions, we observed both the formation of aqueous Fe(II) and H2O2 during photoreductive dissolution of lepidocrocite and ferrihydrite at pH 3. These experimental findings are consistent with the light-induced reduction of surface Fe(III) at the (hydr)oxide surface and the concomitant oxidation of surface-coordinated water or hydroxyl groups, leading to surface Fe(II) and *OH radicals and subsequently to H2O2. The formation of *OH radicals atthe surface was confirmed by photodissolution experiments conducted in the presence of *OH radical scavengers. Kinetic modeling of the experimental data suggests that the relevant pathway for the formation of H2O2 is the reoxidation of surface lattice Fe(II) by molecular oxygen. This study furthermore shows that in the presence of strong iron binding ligands such as siderophores, specifically desferrioxamine B, the photoreductive dissolution of lepidocrocite, ferrihydrite, and to a lesser extent hydrous ferric oxide may also proceed at seawater pH.

Wavelength-dependence of photoreductive dissolution of lepidocrocite (gamma-FeOOH) in the absence and presence of the siderophore DFOB.

Photoreductive dissolution of lepidocrocite (gamma-FeOOH) in the presence/absence of the siderophore desferrioxamine B (DFOB) was investigated at different wavelengths. At pH 3 in the absence of DFOB, Fe(II) formation rates normalized to the photon flux increased with decreasing wavelengths below 515 nm, consistent with enhanced Fe(II) formation at lower wavelengths by photolysis of surface Fe(III)-hydroxo groups or by surface scavenging of photoelectrons generated in the semiconducting bulk. In the presence of DFOB at pH 3, photoreductive dissolution rates, normalized to the photon flux, increased more strongly with decreasing wavelengths below 440 nm. We hypothesize that acid-catalyzed hydrolysis of DFOB generates degradation products that form photoreactive surface complexes leading to an increase in photodissolution rates at low pH. At pH 8 in the presence of DFOB, normalized photodissolution rates had a maximum in the spectral window 395-435 nm and were significantly smaller at lower wavelengths, suggesting that adsorbed DFOB is directly involved in the reduction of surface Fe(III) by a light-induced ligand-to-metal charge-transfer reaction within the surface Fe(III)-DFOB complex. The strong response in the visible light suggests that photoreductive dissolution of iron (hydr)oxides promoted by siderophores with hydroxamic acid groups may occur deep into in the euphotic zone of oceans.

Iron isotope fractionation during proton-promoted, ligand-controlled, and reductive dissolution of Goethite.

Iron isotope fractionation during dissolution of goethite (alpha-FeOOH) was studied in laboratory batch experiments. Proton-promoted (HCl), ligand-controlled (oxalate dark), and reductive (oxalate light) dissolution mechanisms were compared in order to understand the behavior of iron isotopes during natural weathering reactions. Multicollector ICP-MS was used to measure iron isotope ratios of dissolved iron in solution. The influence of kinetic and equilibrium isotope fractionation during different time scales of dissolution was investigated. Proton-promoted dissolution did not cause iron isotope fractionation, concurrently demonstrating the isotopic homogeneity of the goethite substrate. In contrast, both ligand-controlled and reductive dissolution of goethite resulted in significant iron isotope fractionation. The kinetic isotope effect, which caused an enrichment of light isotopes in the early dissolved fractions, was modeled with an enrichment factor for the 57Fe/ 54Fe ratio of -2.6 per thousandth between reactive surface sites and solution. Later dissolved fractions of the ligand-controlled experiments exhibit a reverse trend with a depletion of light isotopes of approximately 0.5 per thousandth in solution. We interpret this as an equilibrium isotope effect between Fe(III)-oxalate complexes in solution and the goethite surface. In conclusion, different dissolution mechanisms cause diverse iron isotope fractionation effects and likely influence the iron isotope signature of natural soil and weathering environments.