Electrothermal mineralization of per- and polyfluoroalkyl substances for soil remediation

PFAS are widely used anthropogenic chemicals that persist in the environment, bioaccumulate, and pose risks to humans and wildlife. Their strong C–F bonds (−485 kJ mol−1) and long half-lives (>100 years in soils) make natural or microbiological degradation ineffective. Existing soil remediation methods—stabilization with sorbents, chemical oxidation with strong oxidants, and thermal treatments—have drawbacks such as merely immobilizing PFAS, generating large volumes of wastewater and secondary pollution, long processing times at high temperatures, and emission of toxic short-chain fluorocarbons. Thermal mineralization that converts PFAS to metal fluorides (e.g., CaF2) is promising but typically requires hours of furnace heating, additional calcium reagents, high energy inputs, and often achieves <80% mineralization. The study proposes rapid electrothermal mineralization (REM), leveraging direct electric heating with biochar additives for ultrafast heating/cooling and potential in-place treatment, to efficiently mineralize PFAS in soil while preserving soil properties.

Prior remediation strategies include: (1) Stabilization by mixing sorbents (activated carbon, clays) to reduce PFAS mobility and bioavailability, which does not destroy PFAS and poses long-term risks. (2) Chemical oxidation using strong oxidants, which requires extensive post-washing leading to water consumption and potential secondary pollution. (3) Thermal treatment via furnace heating (400–1100 °C for hours) for desorption/degradation, which can emit toxic short-chain fluorocarbons (CF4, C2F6, C2F4) due to incomplete C–F bond decomposition and degrades soil properties. Thermal mineralization of PFAS to metal fluorides (especially with Ca2+) has been reported but typically needs added Ca reagents, long durations, high energy, and achieves <80% mineralization. Emerging direct electric heating offers rapid heating/cooling and ultralow energy consumption, suggesting a route to overcome these limitations.

The REM process mixes PFAS-contaminated soil with a conductive carbon additive (primarily biochar) to achieve sufficient electrical conductivity. In bench-scale tests, ~200 mg soil mixed with ~100 mg carbon additive was loaded into a quartz tube reactor sealed with O-rings and contacted by graphite separators and brass electrodes connected to a capacitor bank. A direct-current pulse (typically 100 V for 1 s; resistance ~3.5 Ω) produced peak currents ~140 A and temperatures up to ~1370 °C with heating and cooling rates ~10^3 °C s−1. REM temperature was tunable (300–2500 °C) by adjusting input voltage (40–150 V). Sealed operation minimized emission of volatile fluorocarbons. PFAS targets included PFOA, PFOS, PFHS, and PFBS (spiked in clean Rice University soil at ~100 ppm). Quantification: residual PFAS by HPLC-DAD and triple quadrupole LC-MS (detection down to sub-ppb for PFOA), mineralized fluoride (F−) by ion chromatography, total fluorine by combustion ion chromatography, and speciation by 19F NMR (D2O extraction). Evolved gases were analyzed by GC-MS. Soil property assessments included particle size (laser diffraction), XRD, XRF, BET surface area, water infiltration rate tests, pH, cation exchange capacity, soil carbon and organic fractions (humic/fulvic), and exchangeable nutrients (P, Ca, K, Mg, Mn, Fe, nitrate). Arthropod microcosm assays (isopods and springtails) evaluated ecological viability. Mechanistic studies compared mineralization efficacy of added carbonates (CaCO3, MgCO3, Na2CO3), characterized products by XRD/XPS/IR, and performed thermodynamic (ΔG via HSC Chemistry) and DFT/MD simulations (VASP; CINEB; MD at 1500–2500 K) to probe C–F cleavage and Ca-assisted mineralization pathways. Scale-up employed a higher-capacitance system (C=0.624 F) for ~7 g batches to 1700 °C within ~6 s, and a third-generation AC-sourced system treating 2 kg soil with 500 g metcoke in a 25.4-cm clay pot using four graphite rod electrodes (steady current ~18 A; ~1000 °C). Sampling across radial/axial positions quantified uniformity. Finite-element simulations modeled current density distribution for pot- and 1 m^3-scale systems, including electrode geometry effects. LCA (cradle-to-gate) compared REM to thermal treatment, chemical oxidation, and ball milling; metrics included cumulative energy demand, GHG emissions, water use, and waste generation. TEA estimated operating expenses per tonne of soil treated.

• REM rapidly heats soil-biochar mixtures to >1000 °C within seconds, achieving PFAS conversion to CaF2 using inherent Ca in soil and biochar, with negligible harmful fluorocarbon emissions under sealed operation.
• Bench-scale PFOA: >99% removal; optimal fluorine mineralization ratio of ~94% at 100 V, 1 s. Repetitive pulses reduced PFOA below New Jersey residential soil standard (130 ppb) within 2 pulses and to ~1.1 ppb after 4 pulses. 19F NMR shifted from PFOA multiplets to a single hydrated F− peak at −128 ppm.
• Generality: PFOS, PFHS, PFBS each showed >99% removal and >90% fluorine mineralization with a single pulse; 19F NMR confirmed only hydrated F− post-REM.
• Sealed REM vs furnace calcination: sealed REM achieved mineralization ratio ~94% with >99% removal, whereas furnace calcination reached only ~0.34% mineralization despite high removal, highlighting mitigation of short-chain fluorocarbon emissions.
• Mechanism: Ca2+ is the most effective counterion for mineralization among Na+, Mg2+, Ca2+ (PFOA removals: Ca ~99.7%, Na ~98.5%, Mg ~94.2%). XRD/XPS/IR confirmed disappearance of C–F features and formation of CaF2. Thermodynamics showed ΔG for perfluoro species degradation becomes favorable with Ca2+ across a broad temperature range, while requiring >1500 °C without Ca2+. DFT/MD indicated reduced C–F cleavage barriers (0.67 eV) and strong driving force (−1.24 eV) with Ca; >90% C–F bond cleavage with Ca vs ~20% without. Higher Ca:F ratios further increased mineralization.
• Soil properties: REM preserved particle size distribution, crystalline phases (quartz dominant), and near-baseline water infiltration (~34 cm h−1 vs ~28 cm h−1 raw), pH (7.58 vs 7.19 raw), and CEC (15.45 vs 15.25 cmol kg−1). Calcination caused severe aggregation, very high infiltration (~455 cm h−1), elevated pH (10.63), and reduced CEC (4.08 cmol kg−1).
• Nutrients and biodiversity: REM increased most exchangeable nutrients by 10–102% (small ~5% decrease for Fe); residual biochar contributed beneficial ion exchange. Arthropod assays showed REM soil had survival comparable to raw soil, whereas PFAS-contaminated soil caused rapid mortality and calcined soil reduced survival.
• Scale-up: 2 kg batch treatment achieved ~97% average PFOA removal with good radial/axial uniformity; simulations indicated uniform current density for pot- and 1 m^3-scale systems, with electrode area influencing achievable temperatures. Wet soil (~30 wt% moisture) yielded mineralization comparable to pre-dried soil.
• Sustainability and economics: Estimated energy consumption ~420 kWh tonne−1. LCA showed cumulative energy demand 3053 MJ tonne−1 (31–33% lower than thermal treatment and ball milling; comparable to chemical oxidation), 40–65% lower GHG emissions, 47–67% lower water consumption, and no chemical waste. TEA operating expense ~$130 tonne−1 (vs thermal ~$117, ball milling ~$411, chemical oxidation ~$473). REM achieves >99% removal within seconds, outperforming reported methods on efficiency-time tradeoff.

The results demonstrate that ultrafast electrothermal treatment using conductive biochar enables in-situ mineralization of PFAS in soil to benign CaF2 by leveraging naturally occurring Ca species. Sealed, pulsed Joule heating achieves high removal and mineralization while suppressing emissions of toxic short-chain fluorocarbons, overcoming a key limitation of conventional furnace-based thermal treatments. Mechanistic experiments and simulations show Ca2+ lowers the thermodynamic and kinetic barriers for C–F bond cleavage and drives formation of the most stable fluoride (CaF2), rationalizing the observed high mineralization ratios without added Ca reagents. REM maintains soil structure, texture, infiltration, and exchange capacity, and even enhances exchangeable nutrient pools, supporting ecological function and biodiversity as validated by arthropod survival comparable to clean soil. Scaling studies and simulations indicate uniform heating and effective remediation at kilogram scale with potential extension to field-scale volumes, including moist soils. The LCA and TEA confirm that REM is environmentally favorable and cost-competitive, with significant reductions in energy, GHG emissions, and water use relative to current practices.

This study introduces rapid electrothermal mineralization (REM) as an efficient, general, and scalable approach for remediating PFAS-contaminated soils. Using biochar-enabled direct electric heating, REM achieves >99% PFAS removal and >90% fluorine mineralization within seconds by converting PFAS to CaF2 using inherent soil/biochar calcium, while minimizing harmful emissions and preserving soil properties. Mechanistic insights establish the central role of Ca2+ in driving mineralization thermodynamically and kinetically. REM scales to kilogram batches with uniform performance and shows strong sustainability and economic advantages over thermal desorption, ball milling, and chemical oxidation. Future work suggested by the study includes field-scale demonstrations with optimized electrode configurations, on-site implementations for large soil volumes (including wet soils), long-term post-remediation monitoring of soil health and biodiversity, and further process integration to streamline biochar handling and recycling.

• At higher input voltages (e.g., 150 V), a slight decrease in measured mineralization ratio and total fluorine recovery was observed, attributed to deposition of insoluble F-containing compounds on reactor walls (quartz tube).
• Achieved removal in the 2 kg scale-up averaged ~97%, slightly below bench-scale >99%, indicating room for optimization in large-scale operation and electrode configuration.
• The process requires mixing conductive additives (e.g., biochar or metcoke) into soil and sealed operation to prevent emission of volatile fluorocarbons.
• REM soil showed a small residual biochar content (darker color) and a minor ~5% decrease in exchangeable Fe.
• Experiments used spiked clean soils; while wet soil tests and simulations support practical applicability, broader field validations across diverse soil types and contaminant mixtures were not included in this study.