Fundamentals of PFAS Adsorption on Granular Activated Carbon
Granular activated carbon removes PFAS from water primarily through adsorption — the accumulation of dissolved molecules on the solid surface driven by a combination of hydrophobic interactions, electrostatic attraction, and van der Waals forces. The thermodynamic equilibrium between dissolved PFAS concentration and the mass adsorbed per unit weight of carbon is described by the Freundlich isotherm: q = K_F × C^(1/n), where q is the adsorbed-phase concentration (mg/g), C is the liquid-phase equilibrium concentration (mg/L), K_F is the Freundlich capacity parameter, and 1/n is the intensity parameter. For PFAS, K_F increases systematically with perfluorocarbon chain length, reflecting the stronger hydrophobic driving force for longer-chain compounds to partition out of water onto the carbon surface.
Published Freundlich K_F values for PFAS on a representative bituminous coal GAC span roughly two orders of magnitude between PFBA (C4 carboxylate) and PFOA (C8 carboxylate). This means that at equivalent liquid-phase concentrations, the GAC will hold approximately 100 times more PFOA than PFBA per gram of carbon at equilibrium. The practical implication is straightforward: a GAC column that has not yet reached equilibrium capacity for PFOA (still removing 99%+) may simultaneously be at or near capacity for PFBA and beginning to release it into the effluent. This phenomenon — differential breakthrough by chain length — is the central mechanism behind short-chain PFAS failures in certified filter systems.
Kinetic Constraints: Why Equilibrium Capacity Is Not the Full Story
Equilibrium isotherms describe the maximum adsorption capacity at infinite contact time. In an operational filter — whether a small point-of-use cartridge or a large municipal fixed-bed GAC contactor — water flows through the media at a finite velocity, and the contact time between a given water parcel and the carbon is limited. The kinetics of PFAS adsorption on GAC involve two rate-limiting steps: external film diffusion (transport from bulk liquid to the carbon particle surface through a boundary layer) and intraparticle pore diffusion (transport from the particle surface into the microporous and mesoporous interior where most adsorption sites are located).
Short-chain PFAS have lower adsorption affinity, which reduces the thermodynamic driving force for intraparticle diffusion. Under the shorter contact times typical of point-of-use filter cartridges — which may achieve only a few seconds of direct media contact — short-chain compounds have insufficient time to diffuse deeply into carbon micropores even if those sites could eventually adsorb them. The result is that kinetic performance for short-chain PFAS is worse than equilibrium isotherms alone would predict, and the gap between long-chain and short-chain performance is larger in operational systems than in batch isotherm experiments.
Log Koc and Kd Values as Predictors of Field Performance
The organic carbon-water partition coefficient (Koc) and the solid-water distribution coefficient (Kd) are field-applicable parameters for predicting PFAS adsorption behavior. Log Koc for PFAS increases by approximately 0.5-0.6 log units per additional perfluorocarbon chain unit for carboxylates. For sulfonates, the relationship is similar but offset upward due to the sulfonate group’s stronger interaction with positively charged surface sites. Published log Koc values for PFBA range from approximately 1.5-2.0, while PFOA values are in the 3.5-4.5 range — a difference of roughly 2-3 orders of magnitude in partitioning tendency.
These thermodynamic parameters directly predict breakthrough behavior in GAC systems. A compound with log Koc of 2.0 will break through a GAC bed orders of magnitude faster than a compound with log Koc of 4.0 under equivalent hydraulic conditions. In monitoring data from municipal GAC systems treating PFAS-contaminated source water, PFBA and PFBS are consistently among the first compounds to appear in effluent, often while PFOA and PFOS remain non-detect. This differential breakthrough has been documented in full-scale utility data from water systems in Minnesota, Michigan, and New Jersey, providing field validation for the laboratory-derived thermodynamic parameters.
Implications for Filter Design and Service Life
The chain-length dependence of PFAS adsorption has direct design implications for GAC systems. A design based on EBCT and bed volumes to breakthrough for PFOA will significantly underestimate the frequency of replacement needed to maintain performance for short-chain co-contaminants. For a source water containing PFOA, PFOS, PFBS, and PFBA at similar concentrations, the PFBS and PFBA will break through at a fraction of the bed volumes required for PFOA breakthrough — potentially 10-50 times fewer bed volumes depending on the specific carbon type and water quality.
For point-of-use filters, this means that the rated service volume based on PFOA reduction may substantially overestimate the service life for short-chain PFAS removal. A filter rated for 100 gallons of PFOA reduction may exhaust its capacity for PFBS at 20-30 gallons. Users and compliance officers who set replacement schedules based on PFOA-optimized NSF certification data will be operating with an unknown but potentially significant window of non-compliance for short-chain compounds. Compound-specific breakthrough testing, rather than single-compound service volume ratings, is the technically appropriate basis for replacement interval specification.