Manipulating Intrapore Energy Barriers in Graphene Oxide Nanochannels for Targeted Removal of Short-Chain PFAS
Summary¶
The work reports β-cyclodextrin–intercalated graphene oxide (GO-βCD) membranes with asymmetric nanochannels that combine high water permeance with strong retention of short-chain perfluorinated acids from multi-component feeds. Filtration experiments (dead-end and cross-flow), diffusion / Arrhenius analyses, and complementary ReaxFF molecular dynamics on PFAS binding to α/β-cyclodextrin support an intrapore energy-barrier picture versus pristine GO and αCD-modified GO.
Broader framing highlights persistent PFAS contamination and the difficulty of removing short-chain species that pass many membranes; modifying GO interlayer chemistry is presented as tuning host–guest binding and activation barriers that set selective rejection under realistic mixed feeds.
Readers should verify numerical values, units, and section references against the peer-reviewed PDF and any Supporting Information, especially when extracts or galley PDFs truncate tables.
Methods¶
- Membrane fabrication (experimental): Shear-alignment printing of GO with added αCD or βCD (heated 90 °C, 2 h stirring), coated on PVDF supports; 80 °C post-bake (1 h). Dead-end tests (Sterlitech HP4750, 14.6 cm\(^2\), 2 bar, 400 rpm stirring); cross-flow long-term trials (CF042 cells, 42 cm\(^2\) active area). PFAS quantification by LC–MS (Agilent workflow described in the article).
- Diffusion / barriers: Temperature-dependent diffusion cells (283–315 K) used to extract activation energies via Arrhenius plots of permeation rates as defined in the paper (Eqs. 1 and 7 in the manuscript).
- ReaxFF binding simulations: Molecular dynamics in Amsterdam Modeling Suite (AMS) using reactive ReaxFF and C/H/O/F parameters from the Liu / Gao / Arkoub fluorinated-organic line; EEM charges trained against HF atomic charges. At least 15 starting poses per PFAS×CD pair, conjugate-gradient relaxation (1 meV/atom), then 500 ps NVT segments at 300 K with a 0.25 fs time step and a canonical thermostat (type per PDF); gas-phase complexes use 3D PBC or large vacuum cells as in the article. Minimum-energy frames support binding energy decomposition (Eq. 8). N/A — NPT barostat / isotropic pressure control in this constant-volume binding pass; N/A — metadynamics; N/A — applied electric field in the protocol quoted here.
Findings¶
- GO-βCD achieves >90% simultaneous retention of PFBA, PFPeA, PFHxA, PFOA from a four-component challenge feed with reported permeance ~21.7 ± 2 L m\(^{-2}\) h\(^{-1}\) bar\(^{-1}\) and large feed up-concentration (~300% in the abstract scenario).
- Simulations report ~20% stronger binding of short-chain PFAS to βCD versus other cyclodextrin models examined, aligning with barrier trends versus GO-αCD and pristine GO (although membrane transport is not fully atomistic, see Limitations).
- Compared to NF270, GO-βCD retains far more short-chain species (e.g. PFBA retention ~35% vs ~89% in the quoted table), with higher water permeance in the experiments summarized—corpus honesty: numbers are as stated in the PDF, not re-derived here.
Limitations¶
ReaxFF binding runs are short gas-phase cyclodextrin models; membrane transport and multicomponent fouling are only partially captured atomistically. Experimental feeds and pH may extend beyond simulated conditions.
Relevance to group¶
Adri C. T. van Duin on ReaxFF support for PFAS–cyclodextrin interactions paired with GO membrane experiments (Majumder group context).