Fluorspar reacts with sulfuric acid to produce hydrogen fluoride, the precursor used to manufacture LiPF6, the dominant lithium-ion electrolyte salt, plus PVDF binder and fluorinated electrolyte additives [S5]. Every lithium-ion EV battery in commercial production depends on this fluorine chemistry chain running from mined fluorite (CaF2) through anhydrous HF to downstream fluorochemicals.
Inside the cell, however, a second fluorine problem runs in reverse: trace water hydrolyses the LiPF6 salt to generate hydrofluoric acid in situ, a primary failure cause in high-voltage lithium-ion batteries that use carbonate-based electrolytes [S1]. HF in turn attacks the nickel-rich cathode, dissolves transition metals, and corrodes the aluminum current collector, producing irreversible capacity loss in NCM811 cells cycled above 4.4 V [S1].
From CaF2 + H2SO4 to LiPF6: the upstream chain
Fluorite is converted to hydrogen fluoride by reacting CaF2 with sulfuric acid, a process Benchmark Mineral Intelligence describes as the gateway chemistry for both refrigerants and the lithium-ion battery chain [S4]. HF is then used to make LiPF6, the main lithium-ion electrolyte salt, PVDF binders for cathodes, fluorinated thermal-stability additives, and protective coatings that prevent battery degradation [S5].
Benchmark's Fluorspar Market Outlook projects that lithium-ion battery demand alone will push fluorspar consumption past 1.6 million tonnes by 2030 [S4]. That figure is independent of HF demand from semiconductor wafer etching, stainless steel pickling, refrigerant manufacture, and PVDF / PTFE fluoropolymer production, all of which share the same acidspar feedstock [S5].
How LiPF6 generates HF inside the cell
Thermal decomposition of LiPF6 produces PF5, H2O, POF3, and HF, and the resulting HF accelerates further electrolyte breakdown in a positive-feedback cycle that worsens during thermal runaway [S2]. Moisture can enter the cell during electrolyte filling, separator drying, or cell assembly, and even parts-per-million water reacts violently with the PF6 anion to liberate HF [S3].
Solvent choice modulates the rate. DMC shows higher thermal stability than EC, although EC remains essential for solid-electrolyte interphase formation on the graphite anode [S2]. On the cathode side, charging nickel-rich NCM811 to 4.6 V intensifies parasitic reactions with the carbonate solvent, accelerating structural degradation and capacity decay [S1]. Density-functional-theory work has traced the elementary step to H2O reacting with LiPF6 decomposition products at the electrolyte-cathode interface [S6].
Additive and scavenger routes to neutralise HF
Three additive families are now competing for the HF-scavenger slot. Cyano-group additives (trimethylsilyl isocyanate SiNCO, 3-isocyanatopropyltrimethoxysilane PTTSNCO) use isocyanate groups to neutralise acidic species; silane additives (vinyl-triethoxy-silane VTEO, vinyl-trimethyl-silane VTMS) react with F- to form stable Si-F bonds; phosphorus additives (ethoxy(pentafluoro)cyclotriphosphazene PFPN, tripropargyl phosphate TPP) rely on phosphorus atoms, and amide additives such as BSTFA exploit amide groups, all of them limited to a single functional group [S1].
A 2026 dual-functional design, 5-trimethylsilanylthiazole (5-SiTZ), pairs a siloxane group that reacts spontaneously with HF with a thiazolyl moiety that acts as a secondary Lewis base and chelates PF5 to suppress further HF generation, achieving complete HF removal in under 1 second and persisting for over 200 days under open-air exposure [S1]. In commercial graphite||NCM811 pouch cells charged to 4.6 V, 5-SiTZ delivered 92.1% capacity retention after 700 cycles [S1].
Inorganic alternatives are also on offer. SCHOTT's barium-silicate glass powder G018-405 scavenges fluorine ions by chemically binding them to the particle surface, with a quoted HF scavenging power of 40 to 60 mg F- per gram of glass, specific surface area of approximately 5 to 8 m2/g (for d50 of 1 micrometre and 0.7 micrometre respectively), and a quoted maximum operating temperature above 600 degC [S3]. Unlike organic additives, the glass particles remain in place during thermal runaway, separating the electrodes and preventing internal short circuits when coated on or incorporated into the separator [S3].
Selection criteria: which HF-control route to specify
For new high-voltage NCM811 or NCA cells cycled above 4.4 V, the 92.1% retention figure at 4.6 V over 700 cycles with 5-SiTZ sets a current benchmark for silyl-thiazole-class additives, while preserving sub-1 second HF scavenging kinetics that matter for fast-charge and high-temperature operation [S1]. For separator or electrode coating routes, the inorganic glass powder route trades lower scavenging power per gram for inorganic thermal stability, with a quoted maximum operating temperature above 600 degC, and survives thermal runaway as a passive insulating layer [S3].
Cost-driven LFP and mid-nickel NCM cells that run at lower cut-off voltages, typically 3.65 to 4.2 V, do not need an HF scavenger additive, because the parasitic LiPF6 hydrolysis pathway is much slower at those potentials and trace-water control during cell assembly is the dominant lever. For EDLCs, supercapacitors and other energy-storage devices using fluorine-containing organic electrolytes, glass-powder additives are qualified as the more thermally stable option [S3].
When comparing additive routes, four criteria are worth tabulating: HF removal kinetics (sub-1 second for 5-SiTZ, surface-reaction limited for glass powder); long-term persistence (over 200 days for 5-SiTZ, indefinitely for inorganic binding); voltage ceiling (4.6 V demonstrated for 5-SiTZ in graphite||NCM811, electrochemical stability at high voltage also reported for the glass powder); and thermal-runaway behaviour (organic additives decompose, glass particles remain as separators) [S1][S3].
Standards, sourcing and traceability
HF content in lithium secondary battery electrolytes is a controllable spec, not an uncontrolled variable, and a published analytical method (US patent application US20130168264A1, filed 2013) covers the titration route used to measure residual HF in production electrolytes [S8]. Barium-silicate glass powders are characterised for maximum operating temperature and Karl Fischer titration for residual water after drying at 100 degC for 4 hours in air, a useful proxy for the dryness of the additive itself [S3].
On the upstream side, HF is on the strategic-critical-chemical lists of multiple jurisdictions because domestic production depends on acidspar supply, an issue raised in U.S. clean-energy manufacturing policy and reflected in the projected 1.6 million tonne fluorspar demand from the lithium-ion battery sector alone by 2030 [S4][S5]. This places acidspar sourcing, conversion to anhydrous HF, and LiPF6 synthesis on the same supply-security list as lithium, nickel, and cobalt. The supply pressure is visible across the wet vs dry battery separator supply chain and other battery raw-material nodes: see the 2026 wet vs dry battery separator supply analysis for a parallel non-fluorine process trade-off.
Failure modes and unresolved questions
Even with the best additive, LiPF6 hydrolysis remains thermodynamically downhill once water is present, so the feedback cycle of HF formation and accelerated breakdown cannot be eliminated, only rate-limited [S2]. Cathode surface degradation at 4.6 V still drives transition-metal dissolution, oxygen loss, and aluminium current-collector corrosion, and 5-SiTZ addresses the chemical half of the problem but not the mechanical stress on the cathode particle [S1].
On the upstream side, fluorspar remains a single-chokepoint mineral, and Benchmark's projection of 1.6 million tonnes of fluorspar demand from the lithium-ion battery sector by 2030 has to clear the same mines that feed refrigerant, fluoropolymer, and semiconductor-grade HF [S4]. Any disruption in acidspar supply cascades through LiPF6 pricing, which in turn feeds into the lithium carbonate indexation in battery cell contracts.
Trackable signals worth watching: dual-functional additive variants beyond 5-SiTZ entering pilot lines, glass-powder coating adoption in commercial separator SKUs, and any re-rating of acidspar reserves relative to forecast LiPF6 demand. The fluorite-to-HF-to-LiPF6 chain is now a single coupled spec, and treating its three nodes independently is the procurement mistake that gets repeated every quarter.