Fluorine-free battery binder: citric acid and cellulose could hold tomorrow’s lithium-ion batteries together

The binder makes up only a small fraction of a lithium-ion battery’s mass, yet this “glue” determines which solvent is used to make the electrodes, how much energy the drying step consumes, and how difficult the battery will be to recycle at the end of its life. Researchers at the University of California San Diego and the Toyota Research Institute of North America have now introduced a battery binder that contains no fluorine, can be processed with water and, under laboratory conditions, comes off the metal foil after a few minutes of soaking in water. The results were published on 9 October 2026 in Nature Communications, and the university released an accessible summary as well.

In short: The new battery binder is made of polyacrylic acid, carboxymethyl cellulose and citric acid, so it can replace fluorine-containing PVDF and the toxic solvent NMP. Test cells matched conventional ones on initial capacity and showed better cycle stability, while recycling worked by soaking in water, with no detectable fluorinated residue.

What is a binder, and why does it matter so much?

Battery electrodes are not solid metal sheets. On the positive side (cathode), a thick paste – known as a slurry – of tiny lithium iron phosphate (LFP) particles, a little conductive carbon black and a binder is spread onto thin aluminium foil; on the negative side (anode), graphite is coated onto copper foil. Once dried, these layers form the “inside” of the battery, which is then stacked or rolled.

The binder’s job sounds simple: hold the particles together and stick them to the foil. But that bond has to survive thousands of charge cycles, while the particles expand and contract slightly every time the battery is charged and the cell interior is a strongly oxidising environment. In the researchers’ test electrodes the binder made up just 2% by weight – yet its influence runs through the entire production chain and into waste management.

What’s wrong with PVDF and NMP?

The industry-standard cathode binder today is polyvinylidene fluoride, or PVDF. It adheres well and is chemically stable, which is why it has been the default for decades. The problems show up in manufacturing and at end of life.

PVDF does not dissolve in water, so it has to be dissolved in N-methyl-2-pyrrolidone (NMP), a toxic, reproductive-hazard organic solvent that the EU already restricts tightly under REACH. The electrodes must be dried in large ovens, and the evaporated NMP has to be captured and purified with expensive equipment. According to the study’s life-cycle model, making one cell takes about 174 grams of NMP, and even with 95% recovery roughly 8.7 grams per cell are lost.

In recycling, PVDF leaves stubborn residues. Removing them takes high temperatures or aggressive chemicals, and its breakdown can produce lithium fluoride and volatile fluorinated compounds that contaminate the recovered material and, without proper off-gas treatment, pose environmental and occupational risks. PVDF is also a fluoropolymer, which places it in the PFAS family – more on that below.

Citric acid as the “connector”: how the new battery binder is built

Water-processable binders already exist: carboxymethyl cellulose (CMC) is widely used in industry for graphite anodes. On the cathode side, however, they have struggled, because water-based binder films tend to crack as they dry and do not cope well with mechanical and electrochemical stress.

The California team combined two cheap, mass-produced polymers:

• Polyacrylic acid (PAA): forms an even film, adds flexibility and is easy to process.
• Carboxymethyl cellulose (CMC): a plant-derived cellulose derivative used in the food and paper industries; it provides strength and adhesion.

Neither polymer is good enough on its own: dry PAA is brittle and moisture-sensitive, while CMC lacks elasticity and tends to oxidise at high voltage. The solution is citric acid, a three-armed molecule that builds “bridges” between the two polymer chains. The resulting network is held together by three kinds of bonds: covalent ester bonds, ionic bonds mediated by lithium ions, and hydrogen bonds. The researchers call the mixture PCC (PAA–CMC–citric acid).

An important detail is that crosslinking happens under mild conditions – at room temperature, in a nearly neutral (pH ≈ 6) water solution, with the pH adjusted using lithium hydroxide. The best recipe was a 1:1 PAA–CMC blend with 5–8% citric acid by weight. Too much citric acid makes the network too tight, which then hinders the movement of lithium ions.

In laboratory measurements, PCC remained oxidatively stable above 5.0 volts, adhered better than PVDF, and produced smooth, crack-free electrodes after drying. After a tape-peel test, almost no material came off the PCC electrode, while the PVDF electrode lost a significant amount after only a few attempts. The PCC electrode could even be twisted without damage.

How did the test cells perform?

The researchers used a lithium iron phosphate (LFP) cathode as their model material – a chemistry increasingly common in lower-cost electric cars and stationary storage. The key results:

• Initial capacity: the PCC cathode delivered 160.2 mAh/g versus 160.5 mAh/g for PVDF – practically identical. First-cycle efficiency was 96.8% compared with 95.7% for PVDF.
• Lifetime: cycling at a roughly three-hour charge-discharge rate (C/3), the PCC cell still delivered 114.9 mAh/g after 1,000 cycles, a capacity retention of 75.7%. The PVDF cell faded faster, especially after cycle 400.
• Conductivity: the electronic conductivity of LFP powder mixed with binder was 41 mS/cm with PCC and 24 mS/cm with PVDF.
• Many electrode types: the binder also worked with manganese-enriched LMFP cathodes (just 0.01% capacity loss per cycle over 590 cycles, and 91.2% retention after 180 cycles at 42 °C), nickel-rich NCM811 cathodes, graphite anodes and silicon anodes (the latter still above 3,000 mAh/g after 100 cycles).
• More realistic cells: a single-layer graphite–LFP pouch cell retained 76.4% of its initial capacity after 1,600 cycles. A multi-layer pouch cell stood at 74.5% after 700 cycles, against 83.5% for the single-layer cell at the same point – a gap the researchers attribute to uneven electrolyte and current distribution between layers.

Soaking in water: recycling made simpler

From a waste-management point of view, the most important part of the study may be end-of-life treatment. After 100 cycles, the researchers took the cells apart and recovered the active material in two ways:

• PVDF electrodes required the established industrial approach: 2 hours of stirring in NMP, followed by an ethanol wash.
• PCC electrodes needed just 5 minutes of stirring in deionised water: the binder dissolved, the coating separated cleanly from the aluminium foil, and washing was done with water as well.

The resulting “black mass” was dried, topped up with lithium carbonate to replace lost lithium, and heat-treated under argon at 540 and 650 °C. This is known as direct recycling: rather than breaking the cathode material down into its elements, it “repairs” the crystal structure so the material can be used again.

The outcome: X-ray photoelectron spectroscopy clearly detected fluorine on the surface of LFP recovered from the PVDF battery, but none on material recovered from the PCC battery. The regenerated LFP’s crystal structure was fully restored, and cells built from it showed better capacity retention and rate performance than cells made from material recovered from PVDF batteries. Graphite could be recovered with the same simple water soak: after heat treatment at 1,000 °C it delivered around 345 mAh/g over 420 cycles, close to graphite’s theoretical maximum.

What does the life-cycle assessment show?

Using models from the US Argonne National Laboratory (GREET 2024, BatPaC, EverBatt 2023), the researchers compared the full manufacturing and recycling process per 0.25 kWh pouch cell.

At the raw-material level the gap is large: producing PVDF emits around 55.8 kg CO₂-equivalent and requires about 90 MJ of energy per kilogram, compared with 1.5–2.7 kg CO₂-equivalent and 32–39 MJ for PAA and CMC. At whole-cell level the difference is more modest, since the binder is only a small share of the mass:

• Manufacturing: the PCC cell accounts for 11.9 kg CO₂-equivalent, 6.9% less, and its energy demand is 9.3% lower. Water use (about 41 litres per cell) and cost (USD 17.7 per cell) also fall slightly, mainly because NMP recovery is no longer needed.
• Recycling: energy use drops by 10.0%, water use by 2.9% and greenhouse gas emissions by 7.2%, while modelled profit rises by 7.1%.
• Fluorine: in a conventional graphite–LFP cell, about 6.5 g of fluorine comes from PVDF and about 15 g from the electrolyte salt lithium hexafluorophosphate (LiPF₆). Dropping PVDF therefore cuts total fluorine input by roughly 30%.

Where caution is needed

The results are promising, but the limitations should be clear:

• This is laboratory scale. The pouch-cell electrodes measured 3.3 × 3.3 cm and stack pressure was not controlled. Industrial-scale, high-speed coating and drying still have to be demonstrated. The researchers also note that they did not measure mass-based recovery yield in recycling.
• The cell is not fully fluorine-free. The electrolyte still contains LiPF₆, which can also generate lithium fluoride and volatile fluorinated compounds (such as hydrogen fluoride) at end of life. “Fluorine-free” refers to the binder, not the whole battery.
• The life-cycle assessment is scenario-based. The authors themselves stress that they modelled a US industrial context, ran no sensitivity analysis, and that the results indicate trends rather than absolute industrial values. The two cathodes compared also differed slightly in composition (96% vs 97% LFP).
• “Renewable” needs nuance. CMC is indeed made from cellulose, but commercial polyacrylic acid is typically made from petroleum-derived acrylic acid. The environmental benefit comes more from eliminating fluorine and NMP than from bio-based origin.
• Commercial interest. The research was funded by a Toyota fellowship, one corresponding author works for Toyota’s research institute, and two authors have filed a patent application on the technology. This is common in applied research but worth knowing.
• Direct recycling is not yet an industry standard. Most of today’s battery recyclers rely on shredding and hydrometallurgy, so the full benefit of a water soak depends on direct regeneration becoming widespread.

Why does this matter for Europe and Hungary?

The EU Battery Regulation sets increasingly strict recycling-efficiency and material-recovery targets, and from 2031 new industrial and EV batteries must contain minimum shares of recycled cobalt, lithium and nickel. Meeting these targets requires technologies that recover materials cleanly, with little energy and without contamination.

Regulatory pressure is also growing on the fluorine side. In 2026 the European Chemicals Agency (ECHA) is in the final stage of its group restriction proposal for PFAS substances, and fluoropolymers – including PVDF – are at the centre of the debate. Whatever the final decision, battery makers would do well to prepare fluorine-free alternatives.

Hungary has become one of Europe’s battery manufacturing hubs, and the Hungarian Battery Association expects battery-industry waste in the country to multiply in the coming years. The siting of battery recycling plants has triggered serious local opposition in several places in recent years, including near Miskolc. A technology that works without toxic solvents and fluorinated decomposition products could reduce risk at the most sensitive points of waste treatment.

This matters especially for LFP batteries. They contain no expensive cobalt or nickel, so conventional pyrometallurgical and hydrometallurgical recycling barely pays off, if at all. Direct regeneration – and a water-soluble binder that makes it easier – could turn that equation around.

The bottom line from a waste-management perspective

The battery binder is a tiny component, but it is a good example of how the circular economy starts at the design stage. If a product is designed from the outset to be taken apart with water at the end of its life, the waste processor no longer has to deal with toxic solvents, high temperatures and fluorinated flue gases. As a next step, the California researchers are working on scaling up production and plan to commercialise the binder with industry partners. Whether this becomes a real industry shift depends on industrial-scale testing – but the direction is clear.

Frequently asked questions about the fluorine-free battery binder

What is the new fluorine-free battery binder?

The new battery binder is a blend of polyacrylic acid and carboxymethyl cellulose, crosslinked into a strong network by citric acid. It was developed by researchers at UC San Diego and Toyota. It contains no fluorine and can be processed with water, so it can replace today’s standard PVDF and the toxic solvent NMP in lithium-ion electrode manufacturing.

Why is PVDF binder a problem in batteries?

PVDF can only be processed in the toxic solvent NMP, which requires energy-intensive drying and costly solvent recovery. In recycling it leaves stubborn residues that need high temperatures or aggressive chemicals to remove, potentially producing lithium fluoride and volatile fluorinated compounds. As a fluoropolymer, PVDF also belongs to the PFAS family of substances.

Does the new binder perform as well as the conventional one?

Yes, in laboratory tests it performed at least as well. The LFP cathode delivered an initial capacity of 160.2 mAh/g versus 160.5 mAh/g with PVDF, and retained 75.7% of its capacity after 1,000 cycles, while the PVDF cell aged faster. It also worked well with graphite, silicon and nickel-rich cathodes.

How does it make recycling easier?

Spent electrodes only need to soak in water for five minutes: the binder dissolves and the active material separates cleanly from the metal foil. PVDF electrodes need two hours of stirring in NMP. LFP recovered with water showed no detectable fluorine and, after direct regeneration, worked again as fully functional cathode material.

Does this make the battery completely fluorine-free?

No, only the binder becomes fluorine-free. The electrolyte still contains the salt lithium hexafluorophosphate (LiPF₆), which can also produce lithium fluoride and volatile fluorinated compounds at end of life. Removing PVDF therefore does not eliminate fluorine from the cell, but it reduces total fluorine input by roughly 30%.

How much lower is its environmental impact?

According to the life-cycle assessment, the savings per cell are moderate but noticeable. Manufacturing greenhouse gas emissions fall by 6.9% and energy demand by 9.3%, while recycling cuts emissions by 7.2% and energy by 10.0%. The analysis is scenario-based and models a US industrial context, so it shows trends rather than absolute values.

When could the new binder reach industry?

There is no firm date yet, because the results are still at laboratory scale. The researchers are working on scaling up binder production and plan to commercialise it with industry partners. First, though, they need to prove it performs just as well in high-speed industrial coating and in large-format cells.

Sources:

• Lin, J., Mu, A., … Muldoon, J. & Chen, Z.: Fluorine-free polymer binder enables circular manufacturing and recycling of lithium-ion batteries. Nature Communications, 2026. (Open access, CC BY 4.0 licence)
• UC San Diego Today: Battery Manufacturing and Recycling Could Go Green With This ‘Glue’, 9 October 2026.