Arsenic mobility in multispecies flow-through systems: Model-based analysis of electrostatic and chemical interactions with mineral assemblages

Arsenic groundwater contamination is a problem of enormous scale affecting the health of millions of people around the world. Extensive research over the last decades have allowed identifying the main processes controlling the As occurrence and mobility in groundwater. Despite these tremendous efforts, the prediction and interpretation of the occurrence and mobility of arsenic for mitigation and remediation of As contaminated aquifers remains a formidable challenge.
The mobility and fate of arsenic in groundwater is governed by the interplay of multiple co-occurring processes that depend on the local hydrochemical, geochemical and hydrodynamic conditions. However, process-based characterization of arsenic mobility has generally been performed focusing on geochemical processes in isolation and have often been limited to batch systems involving a limited number of aqueous species and single synthetic minerals, in particular Fe and Al oxides. Furthermore, due to the complex coupling of physical, chemical and biological processes in natural systems, it is generally difficult to distinguish the effects of specific mechanisms on the mobility of arsenic from the sole analysis of groundwater chemistry. To this end, experiments with an intermediate level of complexity are required in order to identify and characterize the various processes affecting the mobility of arsenic. Moreover, numerical reactive transport models, integrating fundamental scientific understanding of individual processes, are instrumental for the quantitative interpretation and prediction of the occurrence and fate of arsenic in natural systems.
This thesis explores the mobility of arsenic in semi-controlled flow-through systems involving co-occurring processes and approaching natural systems from the hydrochemical and mineral perspectives through the combination of experimental and numerical methods. The specific focus was the investigation of the interactions between arsenic and solid mineral phases in multicomponent systems and mineral assemblage. A key aspect of this research was the use of extensive experimental datasets collected under different hydrochemical conditions and their model-based interpretation with reactive transport simulations coupling physical and geochemical processes and accounting for the local aqueous chemical conditions. This approach enabled to illuminate the interplay between processes and the specific effects of aqueous charged species on As mobility.
The first part (Chapter 2 of this thesis) focuses on the release of arsenic from well-characterized iron oxides (i.e., goethite and ferrihydrite) during in-situ exposure of synthetic mineral phases to natural groundwater conditions. In particular, the synergic and competitive effects of aqueous charged species on transient desorption of arsenic from goethite were investigated using surface complexation models (SCMs). The capabilities of two SCMs broadly applied in the field of reactive transport models (i.e., DDL and CD-MUSIC models) were compared as these surface complexation descriptions include different descriptions of the electrostatic interactions taking place at the surface-solution interface. The simulation outcomes demonstrate the importance to rigorously account for both the chemical and electrostatic interactions taking place at the surface interface in multicomponent systems in order to describe the macroscopic sorption behavior of arsenic. Whereas some species can directly compete with arsenic for sorption sites (e.g., phosphate), others, in particular major cations (e.g., calcium), can significantly affect the surface charge behavior of iron oxides through electrostatic interactions and therefore impact the sorption affinity of arsenic. This interplay leads to non-linear surface complexation behavior of arsenic which indicates that the sorption of arsenic in natural systems strongly depends on the local chemical composition and the hydrodynamic conditions. Furthermore, a model was developed to simulate the observed As temporal release during in-situ reductive transformation of ferrihydrite observed at different spatial locations in the field. The model considered the coupling between abiotic and biotic kinetic reductive dissolution/transformation of ferrihydrite, the sorption of arsenic and aqueous species naturally present in the groundwater as well as the arsenic sequestration into newly-formed secondary iron mineral phases. The model-based interpretation shows that ferrihydrite reductive dissolution is the primary driver for the mobilization of arsenic. However, secondary abiotic mineral transformations triggered by the dissolution of ferrihydrite and dependent on the local hydrochemical conditions and on the mineral composition have also fundamental implications for the mobility of arsenic.
The second part (Chapter 3 of this thesis) addresses the transport of arsenic in natural mineral assemblage that was studied under well-controlled aqueous chemical conditions in flow-through laboratory experiments. This work focused on naturally-coated quartz sand since it represents the dominant solid phase in various aquifers contaminated by arsenic. However, detailed investigation of As mobility in naturally-coated quartz sand has hardly been studied. This research was conducted in two successive parts. First, the propagation of pH fronts and the protonation behavior of the mineral assemblage were characterized. Second, the effects of the silica porous media and aqueous charged species on arsenic mobility were investigated. A series of column experiments was performed and consisted in injecting As solutions with different pH and background electrolyte concentrations in order to explore the interactions between arsenic, the aqueous charged species and the surface of the mineral assemblages. The experiments were repeated using different types of silica porous media in order to compare their effects on the transport of aqueous solutes. The flow-through experiments were combined with the characterization of the natural sand properties and with reactive transport modeling. In particular, the interactions between the aqueous charged species and the surface of the mineral assemblage were described with surface complexation models using a component additive approach in order to account for the distinct contribution of quartz and metal oxides present in the natural sand coatings. Moreover, the quartz surface was described with a bimodal acidity behavior based on recent insights gained from interfacial molecular studies. These multiple lines of evidences were used to distinguish the geochemical processes governing the mobility of the protons, major ions and arsenic. The outcomes showed that the silica porous media substantially interacted with the aqueous solutes and revealed an interesting sequential sorption mechanism. This mechanism consisted in a strong release of protons from the quartz surface upon interaction with the background electrolytes which, subsequently, led to a profound impact on the arsenic affinity for the natural sand coatings.
In conclusion, this PhD thesis has contributed to advance the knowledge on arsenic mobility towards a comprehensive description of arsenic transport in complex natural flow-through systems. It has been shown that interpretation and prediction of arsenic behavior in the subsurface requires consideration of coupled physical, chemical, electrostatic and biological processes and of the specific effects of individual aqueous solutes and mineral phases. In this thesis, the synergic combination of experimental investigation and reactive transport modeling allowed the quantitative understanding of the controlling effects of natural hydrochemical conditions and mineral assemblages on the fate of arsenic in groundwater.