Hybrid graphenic and iron oxide photocatalysts for the decomposition of synthetic chemicals

PFAS are persistent, bioaccumulative organofluorine compounds with exceptionally strong C–F bonds, leading to long environmental half-lives and human health risks. They are frequently detected in aquatic environments and human biomonitoring, with regulatory limits such as the proposed US EPA MCL of 4.0 ng/L for PFOA. Conventional water treatment processes (coagulation, flocculation, sedimentation, filtration) are largely ineffective for PFAS due to high solubility and low volatility, prompting interest in adsorption, membrane separation, and destructive technologies. Photocatalytic degradation has emerged as a promising route to break down PFAS using light-generated reactive species. This work investigates a frugal, scalable approach that immobilizes photoactive iron oxides onto mesoporous graphenic carbon derived from cellulose to combine adsorption (pre-concentration/complexation) with photocatalysis (charge separation and hole transfer), aiming for efficient PFOA degradation at low UV fluence and low catalyst cost.

Adsorptive treatments using activated carbon, ion exchange resins, and modified clays can remove PFAS but require regeneration or disposal. Membrane processes offer separation but generate concentrated waste streams. Advanced destruction methods, including photocatalysis, are being explored to mineralize PFAS. A range of semiconductor oxides (ZnO, CeO2, Ga2O3, TiO2) have been studied under UV irradiation. Desirable photocatalysts feature high activity/selectivity, large surface area, stability, and cost-effectiveness. Iron’s complexation with PFAS and advances in nano-enabled materials motivate hybrid systems coupling adsorption with catalysis. Prior studies frequently report time-based kinetics without standardized UV fluence data, complicating cross-study comparisons; normalization by fluence and catalyst loading is recommended to benchmark performance.

Synthesis: Graphenic carbon (g-C) and Fe-doped g-C were prepared from Chemi-ThermoMechanical pulp (CTMP). Fibers were milled (40 OD mesh), soaked in FeCl3 solutions to achieve target Fe contents (1.6–64 wt%), dried, milled again, and pyrolyzed at 600 °C for 5 min in a muffle furnace under an oxidative atmosphere (heating rate 40 °C min−1). The furnace was turned off at 600 °C to cool to room temperature. Characterization: SEM (Quattro ESEM, 10 kV), TGA (TGA 5500, 25–850 °C, 20 °C/min, N2), powder XRD (Rigaku MiniFlex 6), FTIR (Bruker Invenio, ATR, 4000–400 cm−1, 4 cm−1 resolution, 128 scans), Raman (inVia, 532 nm excitation; D and G band fitting with Lorentzian and BWF), N2 physisorption at 77 K (Micromeritics 3Flex; degassing 150 °C, 24 h; BET, t-plot, DFT pore size distribution), UV–Vis diffuse reflectance (Agilent Cary 5000; 250–2500 nm; Kubelka–Munk function; Tauc analysis assuming indirect transitions), and XPS (Kratos AXIS Nova, monochromatic Al Kα; charge correction to C 1s at 284.8 eV; CasaXPS analysis). Photocatalytic experiments: Conducted in a collimated-beam bench-scale photoreactor with fluence measurement. Typical tests used 50 mg catalyst in 50 mL PFOA solution (1 mg L−1, initial pH 3.5; catalyst dosage 1 g L−1). Suspensions were stirred 30 min in the dark to reach adsorption–desorption equilibrium; UV lamp (254 nm) was preheated 30 min. Reactions at 22 ± 2 °C with samples withdrawn over time. UV fluence rate was 1.42 ± 0.05 mW cm−2; total fluence up to 30.7 J cm−2 in 6 h. Experiments also evaluated catalyst dosage (0.1–1 g L−1), initial PFOA concentration (0.1–5 mg L−1), simulated solar light (AM 1.5G), and dark controls. Reusability was assessed over five consecutive 6 h cycles, replenishing PFOA to ~1 mg L−1 between cycles. Analytical methods: PFOA quantified by UHPLC-MS (Agilent 1200; Waters XTerra MS C18 100×2.1 mm, 3.5 µm; guard column; column at 50 °C; 1 mL min−1; mobile phase A: water + 20 mM ammonium acetate; B: acetonitrile; gradient 50:50 to 10:90 to 50:50; ESI− MRM). Calibration range 0.5 µg L−1 to 2 mg L−1. Iron leaching quantified by ICP-MS (Agilent 7700x) with appropriate calibrations and 1000× dilution as needed. Fluoride mass balance and adsorption assessed via comparison to NaF controls. Fluence-based normalization was used to compare performance with literature.

• Fe/g-C hybrid photocatalyst effectively degrades PFOA; UV alone shows negligible removal due to low PFOA molar absorptivity at 254 nm (ε254 = 3.5 ± 0.9 M−1 cm−1).
• Iron content strongly influences activity (1 g L−1 catalyst, 1 mg L−1 PFOA, 6 h, ~30.7 J cm−2): 1.6, 3.2, 4.8, 6.4 wt% Fe achieved 3.7%, 5.8%, 7.2%, 13.4% removal; 16 wt% Fe achieved 66.4%; 32 wt% Fe achieved 89.7% removal. 64 wt% Fe did not further enhance degradation.
• Fast performance: ≥85% PFOA removal in 3 h under UV, and approximately 90% in 6 h, with sustained performance over extended operation.
• Catalyst dosage effect (32 wt% Fe): even 0.1 g L−1 achieved ~79% removal in 6 h; higher dosages increased removal due to more complexation/active sites.
• Initial concentration effect: increasing PFOA from 0.1 to 5 mg L−1 decreased 6 h removal from near-complete to ~86%, indicating strong efficacy across environmentally relevant and concentrated ranges.
• Light source: simulated solar light (AM 1.5G) enhanced removal compared to 254 nm UV at the same conditions, indicating activity beyond monochromatic UV.
• Textural/optical properties: 32 wt% Fe/g-C exhibited BET surface area 427 m2 g−1 and pore diameter 13.4 nm versus g-C at 56 m2 g−1 and 0.92 nm; optical band gaps ~1.28–1.19 eV for hybrids. Raman ID/IG increased linearly with Fe content, indicating increased structural order beneficial for charge transfer.
• Stability/recyclability: Over five consecutive cycles (~30 h total), >90% removal in cycles 1–4 and 88.5% in cycle 5; SEM showed no significant morphological changes; TGA indicated thermal stability in test range.
• Mechanism indicators: Two-step process of adsorption/complexation followed by photocatalytic hole transfer; XPS showed post-reaction fluoride adsorption (F 1s at 648.5 eV) and reduction in oxygen-related functional groups, consistent with electron transfer to g-C and facilitated charge separation.
• Fluoride mass balance: Total fluoride recovery decreased to 42.2% within 1.5 h and plateaued, attributed to significant adsorption of F− onto iron oxide sites; control with NaF showed 47.0% F− adsorption onto 32 wt% Fe/g-C, while pure g-C showed negligible F− adsorption.
• Iron leaching post-photodegradation measured at ~70 ppm (69.948–70.210 ppm).
• Benchmarking: Normalized degradation efficiency (accounting for time, fluence rate, and catalyst dosage/contaminant concentration) indicated Fe/g-C outperformed reported photocatalysts using readily available materials and simple synthesis.

The Fe/g-C hybrid leverages a dual-function mechanism: iron sites complex and pre-concentrate PFOA on the mesoporous graphenic surface, followed by photocatalytic oxidation where holes transfer to the adsorbed PFOA, promoting C–F bond cleavage and stepwise formation of shorter-chain PFCAs. Iron oxide nanoparticles dispersed on conductive, high-surface-area g-C enhance charge separation and transport, as supported by increased Raman ID/IG with Fe content and XPS evidence of electron migration to g-C and reduction of oxygen-containing groups that act as electron acceptors. The mesoporosity (Type IV isotherms with hysteresis) and expanded surface area with Fe doping increase accessible active sites and adsorption kinetics. Enhanced activity under simulated solar suggests broader spectral utilization, improving practicality. The catalyst shows strong stability and recyclability over multiple cycles without structural degradation, supporting scalability. Although fluoride mass balance appears low during irradiation due to adsorption of F− onto iron oxide, total defluorination is consistent with strong F− affinity and does not negate degradation; however, it indicates that downstream management of adsorbed fluoride and potential regeneration strategies should be considered. Overall, the approach addresses the research objective of creating a low-cost, scalable, and effective photocatalyst for PFAS degradation at low fluence, with performance competitive to or exceeding literature benchmarks.

This study introduces a simple, economical synthesis of iron oxide/graphenic carbon (Fe/g-C) hybrids from cellulose-derived precursors, delivering rapid and high-efficiency photocatalytic degradation of PFOA. The optimized 32 wt% Fe/g-C achieves ≥85% removal within 3 hours and ~90% in 6 hours at low fluence rates, maintains performance across cycles, and benefits from high surface area, mesoporosity, and favorable charge-transfer characteristics. The work demonstrates that efficient PFAS degradation can be realized without complex or costly systems, and performance benchmarking using fluence-normalized metrics underscores its competitiveness. Future research should explore: (i) other sustainable graphenic carbon sources and iron oxide phases/architectures, (ii) optimization under solar spectra and in real water matrices, (iii) strategies to manage fluoride adsorption and minimize iron leaching, and (iv) reactor scale-up and continuous-flow operation for practical deployment.

• Fluoride mass balance during experiments was low (~42% after 1.5 h), primarily due to significant fluoride adsorption onto iron oxide sites; while indicative of defluorination, this complicates direct quantification and may necessitate desorption/regeneration steps for complete accounting.
• Measurable iron leaching (~70 ppm) was observed post-photodegradation, which may raise concerns for downstream water quality and catalyst longevity; mitigation strategies are needed.
• Although effective under 254 nm UV and simulated solar light, performance optimization across broader visible spectra and in complex water matrices (natural organic matter, competing anions) warrants further study.
• High Fe loading beyond the optimum (e.g., 64 wt%) did not improve activity, indicating a limited beneficial range of iron content and the need for precise control of composition and dispersion.
• The optical bandgap values of the hybrids (∼1.2 eV) are atypical relative to constituent materials and may reflect amorphous carbon characteristics; while activity is high, deeper understanding of electronic structure–activity relationships is needed.