PVDF membranes and PFAS: the last word ...
Simon Judd
Why exactly can PVDF membranes be a source of PFAS emissions? And what can be done to minimise them?
Simon Judd of Judd Water & Wastewater Consultants explains the life cycle of PVDF membranes, where PFAS emissions fit in, and which methods and treatment technologies can be used to reduce or offset them.
Simon also takes a look at the alternative materials available and, while the future of PVDF membranes production within the EU may look uncertain, there already exist alternative polymeric membrane materials that can be employed.
Abbreviation Definition CTFE Chlorotrifluoroethylene DMAc N,N-Dimethylacetamide ECHA European Chemicals Agency FS Flat sheet GAC Granular activated carbon HCl Hydrochloric acid HF Hydrofluoric acid (Fig. 1) HF Hollow fibre (Fig. 2) HFPO‑DA Hexafluoropropylene oxide dimer acid IEX Ion exchange NIPS Nonsolvent-Induced Phase Separation NMP N-Methyl-2-pyrrolidone PAN Polyacrylonitrile PE Polyethylene PES Polyethersulphone PFAS Per- and polyfluoroalkyl substances PFCA Perfluoroalkyl carboxylic acids PFECA Per-/polyfluoroalkyl ether carboxylic PTFE Polytetrafluoroethylene PVDF Polyvinylidene fluoride PVP Polyvinyl pyrrolidone PFNA Perfluorononanoic acid PFOA Perfluorooctanoic acid PFOS Perfluorooctane sulphonate PP Polypropylene PVA Polyvinyl alcohol PS Polysulphone RO Reverse osmosis SCWA Supercritical water oxidation VDF Vinylidene fluoride
1. PFAS emissions over the PVDF membrane life cycle
Polyvinylidene fluoride (PVDF) is currently the most common choice of polymer for membrane filtration across the water industry sector, both for water and wastewater purification.
As has been widely reported, the EU is proposing a ban on PFAS manufacture, and the precise status of the ban with respect to PVDF membranes is now becoming clear. The latest European Chemicals Agency (ECHA) proposal is that manufacturers of PVDF membranes should have 6.5 years of exemption from the ban, or 'derogation', to allow transition to new materials from the time the legislation comes into force. So, if the law is finally passed sometime in 2027, then this means that alternatives will have to be available by 2033–34.
So, how did it come to this?
What exactly is the issue with PVDF membranes from an environmental impact perspective?
As pointed out in a previous feature (How risky? PVDF membranes, PFAS emissions and environmental impact), the irony of banning a material widely used to sustain purified freshwater supply and treat wastewater to maintain environmental water quality has not been lost on many.
What's more, there is no question of significant – or, indeed, any – shedding of PFAS (per- and polyfluoroalkyl substances, or 'forever chemicals') from PVDF membranes during their actual use.
The only material that has been shown to leach from the membrane during use is the pore-forming additive polyvinyl pyrrolidone (PVP). This is clearly not great, but the main impact of this degradation is the loss of hydrophilicity of the membrane, rather than any significant health or environmental impact. And, crucially, PVP is not a PFAS.
It also appears likely that any PFAS emission from spent PVDF membranes once they have been disposed to landfill is also of limited significance.
Generally speaking, landfill leachate and, to a lesser extent, landfill gases represent a significant source of PFAS emissions (Hamid et al, 2018; Zhang et al, 2022; Xiang et al, 2025). Mass balance evaluations based on two US landfill sites suggest that possibly 1–3% of all PFAS fed to the site is released (Chen et al, 2024), with almost all of the discharged PFAS (>97%) arising in the leachate.
No study has ever been conducted on the degradation and subsequent emission of PFAS materials from landfilled PVDF membrane modules specifically. But logic dictates that if there is no evidence of loss of PFAS from PVDF membranes during actual use, then it’s unlikely that emissions of any significance will arise following their disposal – at least, not compared with other PFAS sources in landfill discharges. The latter is self-evident from:
- the relative amounts of PVDF membrane and much more prevalent household goods such as food packaging, carpets, and other textiles arising in municipal solid waste, and
- the considerably higher mobility of the PFAS materials used for coatings of these household goods (for example from abrasion of stain-resistant carpets) compared with that associated with the PVDF polymer membrane material.
So, given that PFAS emissions during use or following disposal are insignificant, it follows that the main concern is with the manufacturing stage, i.e. the beginning of the PVDF membrane life cycle. Here, the evidence of a direct link between recorded PFAS emissions and PVDF (and other fluorinated polymers) is indisputable (Song et al, 2018; D’Ambro et al, 2021; Evich et al, 2022; He et al, 2022; Dauchy, 2023, Dalmijn et al, 2024, 2025).
To understand whether it is possible to mitigate the risk of emissions at this initial step, it's first necessary to understand the manufacturing process itself.
2. PVDF manufacture and the risk of PFAS emissions
PVDF membrane manufacture is a multi-stage process (Table 1). It starts with the production of the vinylidene fluoride (VDF) monomer and ends with integrity-testing of the finished membrane module, before the module is packed up ready to send out to the user.
The complete process can be divided into two sets of stages:
- Stages 1–6: Monomer synthesis, through to fibre hydrophilisation. These first six stages generally lie within the realm of the PVDF material and membrane manufacturers, and
- Stages 7–10: Fibre cutting, through to module testing. These tasks are generally undertaken by the membrane module fabricators, where a membrane module is fabricated from a finished membrane starting material.
There are some membrane module fabricators who also produce membranes from PVDF polymer (Stages 3-6), but polymer production (Stages 1-2) is always a separate process.
| Stage | Process |
|---|---|
| 1 | VDF monomer production |
| 2 | Polymerisation to PVDF |
| 3 | Dope preparation |
| 4 | Hollow fibre spinning (NIPS) |
| 5 | Washing & post-treatment |
| 6 | Hydrophilisation |
| 7 | Fibre cutting & bundling |
| 8 | Potting |
| 9 | Module assembly |
| 10 | Integrity testing |
During Stages 1–6, there is a relatively increased risk of significant PFAS emissions because of the gaseous and liquid effluent streams generated. Historically, it has been the polymerisation step (Stage 2), and to a much lesser extent the subsequent washing stage (Stage 5), that have led to the discharge of the most significant quantities of PFAS. And within the polymerisation stage, it has been the use of PFAS fluorosurfactants that has been the cause of greatest concern.
From Stage 3 onwards – and especially the membrane module fabrication stages (7 onwards) – the material is in polymerised form; it is an essentially stable 3D matrix with only a very small risk of PFAS emissions.
3. Role of PFAS fluorosurfactants in PVDF production
Emulsion polymerisation of fluoropolymers like PVDF has until recently required fluorosurfactants to emulsify and stabilise aqueous dispersions, with long‑chain perfluoroalkyl carboxylates such as PFOA and PFNA (perfluorooctanoic acid and perfluorononanoic acid) being used. A wealth of evidence has emerged to suggest that the mobility of these specific PFAS chemicals, along with other processing aids, has caused them to be emitted from PVDF manufacturing facilities in significant quantities (Table 2).
| Finding | Relevance to PVDF / fluorosurfactants | Reference |
|---|---|---|
PFCA perfluoroalkyl carboxylic acids; HFPO‑DA hexafluoropropylene oxide dimer acid); PFECA per-/polyfluoroalkyl ether carboxylic acids | ||
| Very high long chain PFCA levels in soils/dust downwind of a PVDF/fluoroelastomer plant; profiles match processing aids | Direct evidence that PVDF processing aids are emitted and deposited locally | Dauchy (2023) |
| HFPO DA and other processing aids in air downwind of a fluoropolymer plant; emissions drop after abatement | Shows airborne losses of contemporary fluorosurfactants and effectiveness of controls | Dalmijn et al (2025) |
| European inventory: fluorosurfactant emissions reduced, but other PFAS (by products, solvents) remain substantial | Indicates fluorosurfactants are only part of a broader PFAS emission mixture from fluoropolymer plants | Dalmijn et al (2024) |
| Major Chinese fluoropolymer facility: diverse PFECAs and HFPO oligomers in river/sediment linked to polymerisation aids | Demonstrates aqueous releases of emerging PFAS from processing aids/impurities | Song et al, (2018) |
| Global review: fugitive PFAS releases during manufacturing cause widespread contamination | Identifies PVDF/fluoropolymer plants as important point sources in a global PFAS cycle | Evich et al, (2022) |
A number of studies have demonstrated that:
- the post‑polymerisation steps (washing, drying, dispersion handling) are key points where fluorosurfactants leave the product stream and can be emitted to air, wastewater, and solid wastes (Dalmijn et al, 2024, 25)
- only a few per cent of emitted PFAS deposit near facilities, with most of the materials being transported regionally or globally to contribute to widespread contamination (Evich et al, 2022; D’Ambro et al, 2021), and
- thermal processing of fluoropolymer dispersions (e.g. coating/sintering on fabrics) emits volatile PFAS (e.g. HFPO‑DA‑derived ethers, tetrafluoroethene) and thermal transformation products.
So, in the context of PVDF manufacturing, the use of PFAS fluorosurfactants creates a direct emission source of persistent, mobile PFAS. This applies to both legacy PFCA (perfluoroalkyl carboxylic acids)-based processes and the newer PFECA (per-/polyfluoroalkyl ether carboxylic acids) processes (Dauchy, 2023; Dalmijn et al, 2024, 25; Song et al, 2018).
PFECAs are a subset of PFAS developed as replacements for legacy long-chain PFAS such as PFOA – which is now subject to EU restrictions. The manufacturing, placing on the market, and use of fluorosurfactants based on PFOA-type chemistry has not been legally permitted in the EU since 2020. However, this does not equate to a ban on PVDF specifically.
Even where fluorosurfactant emissions are reduced, other PFAS from the fabrication process remain significant, meaning the overall PFAS footprint of PVDF production can still be relatively large (Dalmijn et al, 2024; He et al, 2022).
4. Mitigation
Modern PVDF production facilities seek to address the PFAS emission risk through such enhancements and mitigation measures as (Fig. 1):
- enclosed process design
- elimination of fluorinated surfactant
- thermal destruction of off-gases, and
- advanced wastewater treatment.
PFOA and PFNA have largely been replaced by perfluoroether carboxylic acids (PFECAs) and HFPO‑DA (hexafluoropropylene oxide dimer acid), though these can evidently still lead to PFAS emissions. In some newer ‘fluorosurfactant‑free’ processes, monomers or chain‑transfer agents are used to generate fluorinated surfactants in situ, which can nonetheless still create PFAS by‑products. These emerging PFAS‑free or 'no added PFAS' PVDF‑CTFE (chlorotrifluoroethylene) processes can achieve comparable polymer performance, suggesting a pathway to reduce these emissions at source (Lundberg et al, 2022).
The transitioning away from PFAS‑based processing aids – the primary origin of PFAS contamination from PVDF and broader fluoropolymer manufacturing in the past – has been coupled with significant tightening of emission controls in the EU. There are additionally regulatory requirements, such as compliance with the Industrial Emissions Directive (IED) and Best Available Technology (BAT)-led determinations for fluorochemical manufacture. These are coupled with strict monitoring requirements for total organic fluorine (TOF) and target PFAS compounds, both for liquid effluent and stack emissions. This tightening of processes and protocols seems to have had the desired effect: it has been demonstrated in at least one study (Dalmijn et al, 2025) that abatement technologies implemented at a manufacturing site reduced emissions by around a factor of eight.
5. Prognosis
So, what can be made of all this? The fluoropolymer manufacturing sector (Steps 1–6) will argue that it’s got its house in order by implementing best practices and some very advanced and aggressive thermal or electro-oxidation technologies. While this is certainly the case, this is unlikely to make any much difference to the implementation of the EU legislation for two main reasons.
The first is, in effect, it's all too little, too late. It cannot be argued that advanced oxidation technologies such as supercritical water oxidation (SCWO) have only just been commercialised. At least five SWCO plants – a candidate technology for PFAS destruction (Fig. 1) – were implemented at full scale for industrial or hazardous wastewater between 1995 and 2005 (though, admittedly, none of them stood the test of time). There has unquestionably been the means of addressing PFAS pollution for some time. The fluoropolymer manufacturing sector has perhaps not been sufficiently encouraged to take remedial action sooner, and it things may have been different had the relevant EU legislation been implemented sooner.
What's more, despite significant reductions in PFAS emissions from fluoropolymer manufacturing plants, the legacy of past practices means that PFAS contamination is still a huge concern – and will remain so for some time.
There are parallels with the nuclear industry. Present-day management of spent fuel, process safety and effluent treatment is subject to extremely close scrutiny. Radioactive pollutants captured by the treatment process (ion exchange/adsorption, RO, evaporation) are solidified, encapsulated in stable solid matrices like concrete or bitumen, and safely disposed of in controlled underground repositories. Sadly, such diligence has not always existed within the sector, as evidenced by the significant levels of radioactive contamination of the environment at sites such as Hanford, Marcoule, Mayak, Savannah River and Sellafield/Windscale from the 1950s through to the 1980s.
The second, and perhaps most important, reason is the existence of alternative membrane materials. Currently, perhaps over 70% of all commercial membrane products used in the water industry for freshwater supply and wastewater treatment are PVDF-based. But this wasn’t always the case. Based on a brief review article from sixteen years ago (Santos et al, 2010), almost 40% (19 out of 48) of the individual commercial products at that time were non-fluorinated polymers (Fig. 2). And this figure excludes the subsequent development of ceramic flat sheet membrane materials, whose implementation has steadily increased over the past seven years or so.
Proposed EU legislation to ban the production of PFAS materials is likely to include PVDF. The proposed legislation has not yet come into effect, and is currently limited to the EU.
Any ban will directly impact EU-based PVDF polymer manufacturers – who, it seems, may in any case have 6.5 years grace. The module fabricators producing modules from a membrane starting material should be able to adapt their process to produce module products using existing alternative non-PFAS membrane polymeric materials.
It's encouraging to see that module fabricators and suppliers have already started the process of proactively identifying and employing alternative materials in anticipation of the new EU legislation. This is likely to become a continuing trend.
