A Framework for Comparing Technologies
Selecting a PFAS treatment technology requires understanding not just what each technology can do, but how its performance varies across the structural diversity of the PFAS family. Chain length is the single most predictive structural variable for performance on adsorption-based systems, and its effects cascade through the design parameters of every major PFAS treatment approach. A rigorous treatment selection process should map available removal efficiency data across chain lengths for each technology under consideration.
The following comparison covers the four primary PFAS removal technologies — granular activated carbon (GAC), strong base anion exchange (SBA-IX), reverse osmosis (RO), and nanofiltration (NF) — against the chain-length spectrum from C3 (TFA) to C8+ (PFOA, PFOS). Performance ranges are drawn from published peer-reviewed studies and are presented as approximate ranges rather than fixed values, because real-world performance depends on water matrix, operating conditions, and competing co-contaminants.
Granular Activated Carbon: Strong at C8+, Weak Below C6
GAC performance is highly chain-length-dependent. For C8+ PFCAs (PFOA, PFNA, PFDA) and C6+ PFSAs (PFOS, PFHxS), GAC achieves 90–99% removal under typical drinking water treatment conditions with adequate empty bed contact time (EBCT ≥ 10 minutes). Performance for C7 PFCA (PFHpA) is similar. For C6 PFCA (PFHxA), removal drops to 20–50%. For C5 PFCA (PFPeA) and C4 sulfonate (PFBS), removal is typically 10–30%. For C4 PFCA (PFBA) and C3 PFCA (TFA), effective removal approaches 0%.
Carbon type matters: bituminous coal-based GAC generally outperforms wood-based and coconut shell-based carbons for short-chain PFAS due to higher mesopore volume, which provides better access for the shorter, more hydrophilic molecules. But even optimized bituminous GAC shows poor short-chain performance compared to its performance against long-chain PFAS.
Strong Base Anion Exchange: Better Short-Chain Performance, but Not Complete
SBA-IX resins consistently outperform GAC for short-chain anionic PFAS. Single-use PFAS-selective ion exchange (PFAS-SIX) resins with engineered hydrophobic exchange sites show particularly strong performance: reported removal rates of 60–85% for PFBS and 50–75% for PFHxA under conditions where GAC achieves 10–30%. For long-chain PFAS, SBA-IX performance matches or exceeds GAC: >95% removal for PFOA, PFOS, and their C8+ analogs is routinely reported.
The chain-length dependency is less steep for ion exchange than for GAC, because the electrostatic component of adsorption does not decline with chain length in the same way. However, short-chain PFAS still show reduced resin affinity compared to long-chain compounds, because the hydrophobic contribution — which reinforces the electrostatic interaction — is smaller. For TFA and other C3 compounds, even specialized ion exchange resins show minimal removal.
Reverse Osmosis and Nanofiltration: Most Chain-Length-Agnostic
Pressure-driven membrane processes offer the most chain-length-uniform PFAS removal performance. RO membranes achieve >90% rejection for C6+ PFAS under standard operating conditions and typically >85% for C4–C5 compounds. Performance for TFA is lower — 50–80% depending on membrane type and operating pressure — but still substantially better than adsorption or ion exchange alternatives. NF performance is roughly 10–20 percentage points lower than RO across the chain-length spectrum.
The relative chain-length independence of membrane rejection reflects the dominance of size exclusion and Donnan exclusion mechanisms over compound-specific adsorption. However, membrane processes are not without limitations: energy consumption, brine disposal, and susceptibility to neutral PFAS and precursor compounds (which lack the anionic character that drives Donnan exclusion) all require consideration in a comprehensive system design.
Combining Technologies for Broad-Spectrum Coverage
No single technology achieves comprehensive removal across the full chain-length spectrum, particularly when precursor compounds and neutral PFAS are included. Systems designed for broad-spectrum PFAS removal increasingly use treatment trains: for example, GAC or IX as a primary polisher for long-chain compounds, followed by membrane treatment for short-chain and neutral PFAS. This layered approach increases capital and operating costs but is necessary when the compound profile includes short-chain or replacement PFAS at concentrations of concern.