The migration of arsenic in groundwater is influenced by the heterogeneity of the medium, and the presence of iron minerals adds complexity and uncertainty to this effect. In this study, a stratified heterogeneous sand column with an embedded hematite lens at the coarse-to-medium sand interface was designed. We introduced an arsenic-laden solution and controlled groundwater flow to investigate the spatiotemporal characteristics of arsenic migration and the impact of hematite dissolution. The results showed that the medium structure significantly influenced the arsenic migration and distribution within the lens-containing sand column. The clay layers directed the lateral migration of arsenic, and the arsenic concentrations in deeper layers were up to seven times greater than those on the surface. The extraction experiments of solid-phase arsenic revealed that the main adsorption modes on quartz sand surfaces were the specific adsorption (F2) and adsorption on weakly crystalline iron–aluminum oxides (F3), correlating to the specific and colloidal adsorption modes, respectively. Monitoring the total iron ions (Fe(aq)) revealed rapid increases within the first 14 days, reaching a maximum on day 15, and then gradually declining; these results indicate that hematite did not continuously dissolve. This study can aid in the prevention and control of arsenic contamination in groundwater.

Iron-containing minerals are key factors controlling arsenic (As) transport in groundwater environments. However, few studies have observed the effect of aqueous Fe [Fe(aq)] on As behavior in a water environment. In this study, river sand in the riparian zone was collected for batch experiments to analyze the effect of Fe(aq) on the adsorption of As on river sand, utilizing characterization analyses to identify the reaction mechanism. The results showed that (1) as the concentration of Fe(aq) in the reaction system increased from 0.1 to 20 mg/L, the equilibrium adsorption capacity (Q_e) of river sand for As(III) and As(V) gradually increased. For concentrations of Fe(aq) equal to or greater than 1 mg/L, the Q_e for As(V) exceeds that for As(III), whereas at a Fe(aq) concentration of 0.1 mg/L, the Q_e for As(III) is higher than that for As(V). (2) Compared to the reaction system without added Fe(aq), the adsorption of As(V) onto river sand was inhibited, while the adsorption of As(III) was enhanced under conditions with low concentrations (0.1, 1 mg/L) of Fe(aq). (3) At higher Fe(aq) concentrations (5, 20 mg/L), the adsorption of both As(V) and As(III) by river sand was more effective than in systems without Fe(aq). Characterization tests confirmed this, while Fe(II) reduced As(V), and Fe(aq) adhered to the surface of river sand to form Fe(OH)₃ colloids, thereby facilitating the adsorption of As onto river sand.

Arsenic contamination in polymetallic mining areas is closely linked to surrounding iron-rich manganese minerals. However, conclusive evidence remains limited regarding the retention and migration process of As(V) in naturally manganese-rich manganese ores (especially those with different manganese/iron mass ratios) under dynamic flow conditions. This study investigated As(V) adsorption and transport by four natural manganese minerals (FM1–FM4) through batch/column experiments, characterization, and numerical modeling. Their Mn/Fe mass ratios were 22.7 for FM1, 4.2 for FM2, 3.7 for FM3, and 16.4 for FM4. Batch experiments showed that As(V) adsorption on FM1–FM3 was better described by the Freundlich model, indicating heterogeneous adsorption behavior. Under the tested experimental conditions, the apparent Langmuir q_m values of these minerals decreased from 0.066 to 0.015 mmol·g⁻¹ with decreasing Mn/Fe ratio. However, As(V) adsorption on FM4, which had the lowest Mn and Fe contents, followed the Langmuir model (q_m = 0.012 mmol·g⁻¹), suggesting monolayer adsorption. Column experiments demonstrated rapid As(V) retention for all minerals. In the time domain, increasing the flow rate from 0.5 to 2.0 mL·min⁻¹ generally advanced breakthrough and shortened the desorption tail, although the breakthrough behavior expressed in pore-volume coordinates was not strictly monotonic for all minerals. The Two-Site Kinetic Attachment Model (TSKAM) successfully simulated these dynamics (R² > 0.90, RMSE < 0.05), revealing adsorption controlled by fast and slow kinetic sites, with slow-site contributions diminishing at higher flow rates. Characterization results indicated that adsorbed arsenic on FM1 remained mainly as As(V) and was immobilized primarily through surface complexation involving surface hydroxyl and Fe/Mn–O groups. XRD and SEM-EDS suggested the participation of Fe/Mn-bearing phases, while XPS on FM1 showed pronounced changes in Mn surface species during adsorption. Therefore, As(V) removal by these natural manganese minerals is a coupled physicochemical process influenced by both mineral properties, including Mn/Fe ratio, specific surface area, pore structure, pH_PZC, and Mn surface-state changes, and hydrodynamic conditions in the polymetallic mining areas.