PFAS Restrictions and Display Coatings: Where AG, AR and AF Treatments and Bonding Adhesives Stand

An anti-fingerprint topcoat that survived three years of field validation can still become a documentation problem inside one purchasing cycle. The chemistry that keeps a cover glass free of smudges is usually fluorinated, and the PFAS display coating restriction debate has moved into material selection.

This guide maps the fluorinated chemistry in AF topcoats and in some AR stacks, examines how proposals treat polymers differently from small molecules, and sets out the substitution path with its performance cost. It is written for engineers who answer the customer questionnaire.

Which fluorinated chemistry ends up in an AF topcoat

Two silane families account for most AF topcoats today.

Both bond through a hydrolyzable silane head and present a perfluorinated tail at the air interface, and that tail is the part regulators now examine.

A hydrolyzable head group, usually trimethoxy or triethoxy silane, condenses with silanol sites on chemically strengthened glass or on the silica top layer of an AR stack, so the treatment is a covalent surface modification rather than a film. The tail is a perfluorinated chain: a short fluorotelomer in C6 chemistry, a longer C8 segment in older formulations, or a perfluoropolyether backbone where abrasion life matters.

The cured layer is a few nanometres to a few tens of nanometres thick, applied by spray, dip or vacuum evaporation. That thinness is the compliance exposure: the low surface energy comes from fully fluorinated carbon, the defining feature of the PFAS class, and the molecule is small enough that fluoropolymer arguments do not transfer to it.

Where AR stacks hide fluoropolymer content

In low-index binders, doped silica and release films.

Fluoropolymer is rarely the headline material of an AR stack, so a datasheet that lists only layer thicknesses can still describe a coating that contains regulated chemistry.

Inorganic AR is the easier case. Sputtered SiO2 and TiO2 multilayers, or a magnesium fluoride layer, contain no organic fluorinated carbon; the fluorine in MgF2 is bound inside an inorganic crystal and sits outside the PFAS definition as it is usually drafted.

Wet-coated AR is different. A low-index binder can be an amorphous fluoropolymer, chosen because its refractive index sits near 1.34 and it can be applied by spin or slot die instead of vacuum. Fluorinated alkoxysilanes are also added to silica sols, where they hydrophobize the pore structure and lower the effective index at the same time. Anti-glare layers are usually silica particles in an acrylic binder, yet they can carry a fluorosurfactant leveling agent at parts-per-million concentration, and hard-coated PET films laminated above the glass often run on fluorinated release liners.

How restriction proposals treat polymers against small molecules

By class, not by molecular size or use.

The two groups are handled on different timelines: small-molecule PFAS are already restricted substance by substance, while fluoropolymers are still argued as a category of low concern.

Small molecules are controlled today. PFOA and PFOS sit under the Stockholm Convention and, in the EU, under REACH restrictions rather than under the RoHS Directive 2011/65/EU. RoHS restricts ten substances and does not list PFAS as a group, so a part can be fully RoHS compliant and still carry a fluorinated topcoat. The European Commission single market pages track the class-wide proposal, which is a different instrument.

Fluoropolymers sit in a different argument. PTFE, PVDF, FEP and ETFE are high molecular weight, insoluble and not bioavailable, and industry argues that they should be recognised as polymers of low concern with a generic exemption. Authorities have been cautious: production can involve small-molecule PFAS, and exempting one fluorinated polymer invites substitution toward another.

The consequence is narrow. Even if such an exemption were granted, a fluoroalkylsilane topcoat is not a fluoropolymer; it is a small, hydrolyzable molecule needing its own assessment.

Reading a coating datasheet for PFAS it does not state

A coating datasheet is written to sell performance, not to disclose composition. The two purposes are not in conflict, but they answer different questions, and the compliance question is rarely settled on page one.

Start with the vocabulary. Fluoropolymer, fluoroalkylsilane, FAS, PFPE, perfluoro, fluorosilicone and PTFE-filled are all disclosures, even when the word PFAS never appears. A carrier solvent described as a hydrofluoroether belongs to the class as well.

Numbers give hints. A water contact angle above roughly 110 degrees points toward a fluorinated tail, but a silicone surface can reach a similar figure, so treat it as an indicator rather than a test. A hexadecane contact angle is more diagnostic, because hydrocarbon surfaces wet out quickly there and fluorinated ones do not.

Absence is more informative. A datasheet naming no CAS number, no sub-supplier and no formulation revision has not shown that PFAS is absent; it has shown that nobody asked. Close that gap with a full material declaration rather than reading silence as compliance.

Sol-gel, silicone and non-fluorinated options compared

Substitution is rarely a one-for-one swap. Each alternative moves the performance balance somewhere else, and the useful question is which balance the application can tolerate.

Read the abrasion column as a relative ranking, not a specification. The result depends on the test method, the load and the substrate, so a supplier figure is comparable only with another figure produced the same way. Ask for the method before comparing numbers.

Performance given up when the fluorinated topcoat goes

Three properties move when the fluorinated topcoat is removed, and only one appears on a typical optical datasheet: oleophobicity, durability at a given contact angle, and compatibility with later processes.

Oleophobicity is the loss users notice. A fingerprint is a mixture of sebum and sweat, and a fluorinated surface presents a low-energy interface that the oil cannot wet, so the mark stays compact and wipes away dry. Silicone and sol-gel surfaces are hydrophobic but less oleophobic, and the same print spreads into a film that needs a solvent.

Durability is the second loss. Topcoats are evaluated with steel wool or eraser rubs under load, and a fresh-sample contact angle says little about the value after a thousand cycles. Non-fluorinated hard coats often start close and fall away faster.

Optical consequences are smaller but real: a change of top layer thickness shifts the AR response, and the optics and backlight guide sets out how those tolerances interact with brightness and contrast budgets.

Bonding adhesive and optical clear resin as the overlooked route

Most PFAS questions stop at the cover glass, which is where the visible treatment sits. The layers underneath carry their own chemistry, and an adhesive or a liner can put fluorinated material into a stack whose coating was already cleared. An audit that asks only about the topcoat leaves that route open.

Optically clear adhesives are usually acrylic, and liquid optically clear resins are acrylic or silicone; neither backbone is a PFAS. The exposure sits in the additives and the consumables. Release liners on adhesive tape are frequently coated with a fluorosilicone, fluorosurfactants serve as wetting and leveling agents, and carrier films can include fluorinated slip agents. Each is present at low concentration, and each is still a PFAS input to a declaration that claims otherwise.

Process chemistry deserves the same question. Plasma pre-treatment improves adhesive wetting, and some plasma recipes use fluorinated process gases that can leave residue on the surface being bonded. A no-PFAS statement that ignores the process is incomplete rather than false.

What can you sign when a buyer demands PFAS-free confirmation?

A scoped declaration, not a blanket PFAS-free claim.

The word free has no agreed threshold for a class of thousands of substances, so an absolute claim usually describes intent rather than something a laboratory can repeat.

What can be signed is a statement with four parts: the class definition, quoted in the fully fluorinated carbon wording and dated; the boundary, meaning the exact part number, revision and layer; the basis, either no intentionally added PFAS or a measured total organic fluorine result with the method and laboratory named; and a change notification commitment.

What should not be signed is an unqualified zero. Detection limits, background contamination and an evolving substance list make an absolute statement indefensible a year later, and a buyer who insists on one is usually asking for the wrong document.

In the United States a separate obligation can attach. California's Proposition 65 list includes some PFAS, so a US-facing statement should be checked against OEHHA's current listing before it is issued rather than after a customer asks.

Vetting a coater's material declaration

A material declaration is only as useful as its scope. Most declarations that fail a review are accurate for a different part, a different revision or a different layer of the same stack. The sequence below catches that mismatch early.

• Declaration level: a full material declaration at CAS level, not a statement of RoHS conformity
• Scope: the part number, the revision and the specific coating stage covered
• Named fluorinated chemistry and its function, whether topcoat, leveling agent, liner or solvent
• Total organic fluorine result with the method, the laboratory and the detection limit stated
• Sub-tier chain: formulator, raw material supplier and any toll coater between them
• Release liner and carrier film specification wherever an adhesive is in the stack
• Change notification clause, plus the issue date and revision of the declaration

Two references frame the answer. Programmes such as UL Solutions show what a defensible declaration looks like, and documentation expectations for displays keep tightening under the Ecodesign regulation for displays (EU) 2019/2021. CDTech's quality and certifications page shows which of these documents are maintained and which come per project.

Changing a coating without requalifying the whole stack

Qualification happens at stack level, not at layer level. A new topcoat measuring the same contact angle as the old one has still changed the surface energy the adhesive sees, the thickness the AR design was tuned around and the interface the environmental tests ran against.

Optical properties need repeating at stack level, because haze, transmission and colour shift are measured through every layer. Adhesion and abrasion need repeating on the new surface. Environmental sequences follow IEC 60068, published through the IEC Webstore, and need a full pass on a control build and on the candidate, side by side.

The same logic applies to a supplier change: a different formulation revision from a second coater is a new material even when the molecule has the same name on the drawing. Write acceptance limits into the drawing, not the datasheet.