Installation of synergistic binding sites onto porous organic polymers for efficient removal of perfluorooctanoic acid

Per- and polyfluoroalkyl substances (PFAS) are persistent environmental pollutants with toxicity, prevalence, and bioaccumulation concerns. The U.S. EPA advises combined PFOA/PFOS in drinking water not exceed 70 ppt, motivating advanced technologies for PFAS removal from water. While methods such as oxidative, UV, sonochemical, and electrochemical treatments exist, adsorption is favored for simplicity and efficiency. However, conventional adsorbents (activated carbon, ion exchange resins, minerals, MIPs, biosorbents, CNTs, MOFs, COFs) suffer from low capacities, slow kinetics, weak affinity, poor stability, and low selectivity in the presence of natural organic matter (NOM). The study aims to develop ideal PFAS adsorbents featuring high hydrophilicity and accessible capturing sites, strong interactions to accelerate adsorption, suitable pore sizes to minimize NOM co-adsorption, and excellent stability for regeneration. The authors propose porous organic polymers (POPs), particularly porous aromatic frameworks (PAFs), as platforms to introduce synergistic electrostatic and hydrophobic binding sites to target the anionic headgroup and hydrophobic perfluoroalkyl chain of PFAS for enhanced removal.

Prior PFAS adsorbents include activated carbon, ion-exchange resins, minerals, molecularly imprinted polymers, biosorbents, carbon nanotubes, MOFs, and COFs. Reported limitations include low adsorption capacities (as low as 0.63 × 10⁻³ to 753 mg g⁻¹), long equilibrium times (1–48 h), weak binding affinity, insufficient water/chemical stability, and poor selectivity against NOM. POPs/PAFs offer ultrahigh surface areas, tunable pore sizes, functionalizable pore walls, and robust stability, and have been successfully used in gas separation, proton conduction, catalysis, energy storage, enzyme immobilization, and water treatment, suggesting potential as advanced PFAS adsorbent platforms.

Synthesis and functionalization: PAF-1 was chloromethylated (PAF-1-CH2Cl) and subsequently reacted with tertiary amines N,N-dimethylpropylamine (NDMP), N,N-dimethyl-butylamine (NDMB), or N,N-dimethylhexylamine (NDMH) in ethanol at 90 °C to afford quaternary ammonium-functionalized PAF-1-NDMP, PAF-1-NDMB, and PAF-1-NDMH. Characterization: FT-IR (new bands for CH3–N+, CH2/CH3, and C–N), XPS (N 1s ~401 eV), solid-state 13C NMR (new resonances ~30–32 ppm), elemental analysis (N and Cl contents indicating near-complete CH2Cl conversion), SEM and DLS (spherical particles, 0.75–2.55 µm), EDS mapping (uniform N distribution), N2 adsorption at 77 K (BET areas: PAF-1 3568 m² g⁻¹; PAF-1-NDMP 602; PAF-1-NDMB 108; PAF-1-NDMH 50 m² g⁻¹), pore size distributions (large pores ~12.6–22.8 Å suitable for PFOA), water contact angles (hydrophobic PAF-1 at 116.6° changing to hydrophilic: NDMP 38.2°, NDMB 34.9°, NDMH 36.3°). Adsorption experiments: Kinetics—20 mg adsorbent in 50 mL aqueous PFOA (1000 ppb, pH 6.88) at room temperature; time-resolved sampling and HPLC-MS quantification. Pseudo-second-order kinetic modeling (linear and nonlinear tested; linear chosen with R² ~1). Isotherms—10 mg adsorbent in 100 mL PFOA at various concentrations (20–600 ppm; also reported up to 1600 ppm), 8 h equilibration; Langmuir fits (non-linear regression). Mechanistic probes: Ion exchange assessment via chloride analysis in filtrate and loaded solids; zeta potential vs time during adsorption; FT-IR shifts of C–N; solid-state 19F MAS NMR and 13C NMR on PFOA-loaded samples; controls with mono-functionalized PAF-1-TMA (electrostatic only) and PAF-1-SE (hydrophobic only). Computations: DFT (CAM-B3LYP/6-31G(d,p), Gaussian 09) to optimize functionalized PAF-1 fragments interacting with PFOA and compute electrostatic potential maps and binding motifs. Breakthrough and practical tests: Fixed-bed columns (300 mg adsorbent; 3.3 mm i.d.; ~7.8 cm bed) treating PFOA solutions containing 20 ppm humic acid at either 500 ppb or 400 ppm PFOA; effluent monitored by HPLC-MS to EPA advisory threshold. Real wastewater from Xiaoqing River (initial 166.27 µg L⁻¹ PFOA; pH 6.58), batch adsorption (20 mg in 50 mL) with time-resolved HPLC-MS. Regeneration: Desorption by washing with methanol/saturated NaCl mixture; reuse over at least six adsorption–desorption cycles; porosity checked post-cycling. Cost analysis: Comparative material cost estimation versus DFB-CDP for equivalent PFOA adsorption amounts.

- PAF-1-NDMB exhibits extremely fast kinetics, removing 99.99% of PFOA from 1000 ppb to 54 ppt within 2 minutes, below the EPA 70 ppt advisory level. - Pseudo-second-order rate constant k₂ for PAF-1-NDMB is 24,000 g mg⁻¹ h⁻¹, the highest reported among PFOA sorbents (compared to DFB-CDP 64.8; MIL-101(Cr) 0.0082; MWNTs 0.00903; FCX4-P 3.8; UiO-67 0.036; AC 4.72 g mg⁻¹ h⁻¹). - Langmuir isotherm fits (R² = 0.996). Maximum PFOA uptake capacity qmax = 2000 mg g⁻¹ (4.8 mmol g⁻¹) at Ce ~600 ppm. This is 32.0× DFB-CDP (62.5 mg g⁻¹) and 24.1× AC (83 mg g⁻¹) under identical conditions; and exceeds other reported materials (e.g., NU-1000 507; MIL-101(Cr) 460; UiO-67 700; AC 52.8 mg g⁻¹; resins 1500 mg g⁻¹). - Ion exchange mechanism evidenced by Cl⁻ release into filtrate during adsorption (e.g., 0.00227 mol g⁻¹ for PAF-1-NDMB) and reduced Cl⁻ content in loaded adsorbents. - Post-adsorption decreases in BET surface area/pore volume and increases in contact angle confirm pore filling by PFOA. - PAF-1-NDMB shows 100% removal efficiency across pH 2–11; also removes PFOS with qmax = 2381 mg g⁻¹ and k₂ = 19,200 g mg⁻¹ h⁻¹. - Selectivity/anti-fouling with NOM: In breakthrough with 500 ppb PFOA + 20 ppm HA, 300 mg PAF-1-NDMB treats 3530 mL to <53 ppt, versus DFB-CDP 242 mL (62 ppt). For 400 ppm PFOA + 20 ppm HA, 1000 mg PAF-1-NDMB treats 9.5 mL to 31 ppt vs DFB-CDP 0.5 mL (41 ppt). - Real wastewater (Xiaoqing River): 99.99% removal within 10 s, reducing PFOA from 166.27 µg L⁻¹ to <70 ng L⁻¹. - Regeneration: Efficient desorption with MeOH/sat. NaCl; at least 6 reuse cycles without observable capacity loss; porosity remains intact. - Cost: Estimated material cost 2.5× lower than DFB-CDP for the same PFOA adsorption amount. - Structure–function insights: Hydrophilicity of particles, suitable pore sizes (~12.6–22.8 Å), and synergistic electrostatic/hydrophobic binding underpin performance; particle size not correlated with performance (0.75–2.55 µm across samples).

The study demonstrates that installing adjacent quaternary ammonium groups (electrostatic sites) and alkyl chains (hydrophobic sites) on PAF-1 creates synergistic binding sites that match the anionic headgroup and perfluoroalkyl tail of PFAS. Multiple lines of evidence support this mechanism: (i) Ion exchange behavior with chloride release confirms electrostatic interactions dominate initial uptake; (ii) FT-IR C–N stretching shifts (≈+7 cm⁻¹) upon PFOA loading indicate strengthened electrostatic interactions; (iii) Zeta potential becomes more negative rapidly during adsorption, consistent with anion binding; (iv) 19F MAS NMR shows chemical shift changes and attenuation for fluorinated segments, and 13C NMR exhibits shifts for carbons adjacent to binding sites, evidencing hydrophobic interactions between PFOA chains and polymer alkyl groups; (v) Controls with only electrostatic (PAF-1-TMA) or only hydrophobic (PAF-1-SE) functionality show significantly reduced kinetics and capacities compared to the dual-functional PAF-1-NDMB, highlighting synergy; electrostatic-only outperforms hydrophobic-only, indicating electrostatics are the primary driver while hydrophobic interactions boost performance; (vi) DFT calculations depict strong Coulombic interactions between [(CH2)2N(CH3)2]+ sites and PFOA carboxylate, with possible hydrogen bonding to F atoms, and suggest binding of two PFOA molecules per functional site, aligning with experimental uptake (~1.75 PFOA per NDMB). The combination of high particle hydrophilicity (reduced contact angles) facilitating mass transfer, appropriate mesopore sizes enabling access, and robust chemistry yields unprecedented capacities and rates, strong NOM tolerance, broad pH applicability, and reusability, directly addressing the shortcomings of prior adsorbents.

This work introduces a versatile strategy to engineer POP-based sorbents with synergistic electrostatic and hydrophobic binding sites for highly efficient PFAS capture. The optimized PAF-1-NDMB achieves record-high PFOA uptake (≈2000 mg g⁻¹) and the fastest reported kinetics (k₂ ≈ 24,000 g mg⁻¹ h⁻¹), rapidly reducing PFOA from 1000 ppb to 54 ppt within 2 minutes. It substantially outperforms benchmark adsorbents (DFB-CDP, activated carbon), maintains performance across pH 2–11, exhibits strong selectivity under NOM background, effectively treats real contaminated water, and can be regenerated for multiple cycles with low material cost. The mechanistic basis—synergistic electrostatic and hydrophobic interactions—suggests this design principle can be generalized to other POPs for broad PFAS remediation. Future work could extend functional group chemistries, optimize pore architectures for different PFAS, and integrate materials into scalable fixed-bed systems for industrial water treatment.

Not explicitly discussed in the paper.