Fluoroantimonic Acid: The Strongest Acid, and What It Does to Methane

Not a Compound: Two Liquids and the Ion Pair They Make

People write the formula H₂FSbF₆, but that is bookkeeping, not a bottle. What exists is a mixture: anhydrous hydrogen fluoride, a liquid only below 19.5 °C, with antimony pentafluoride stirred in. SbF₅ alone is a colourless syrup — roughly 460 cP at 20 °C against water's 1.0 cP — because it is a chain of octahedra sharing fluorine bridges.

One number explains the choice of partner: fluoride-ion affinity, the energy released when a neutral molecule captures a bare F⁻, and the cleanest ranking of neutral Lewis acid strength. SbF₅ has a fluoride-ion affinity of ~490 kJ/mol, versus ~340 kJ/mol for BF₃ — the strongest of the simple binary fluoride Lewis acids, and hungry enough to tear a fluoride off HF itself:

• HF + SbF₅ → H⁺ + SbF₆⁻, the 1:1 stoichiometry written naively.
• 2 HF + 2 SbF₅ → H₂F⁺ + Sb₂F₁₁⁻, what is actually there — the H₂F⁺ cation Mootz and Bartmann pinned down by low-temperature crystallography in 1988.

Sb₂F₁₁⁻ is two SbF₆ octahedra sharing one bridging fluorine, so one negative charge is spread over 11 fluorine atoms on 2 antimony centres. The proton, meanwhile, is never naked: it sits on fluorine as H₂F⁺ in an (HF)_nH⁺ chain and travels by forming one bond as it breaks the next — Grotthuss-style hopping.

Superacidity Is a Pairing, Not a Stronger Bond

The intuition that a strong acid simply has a weak H–X bond is the less important half. Every proton transfer is an equilibrium, and a superacid is one whose reverse is dead: a brutal proton donor paired with a conjugate base too delocalised and too non-nucleophilic to take the proton back or attack the cation it just made.

Sb₂F₁₁⁻ is that base. Its charge is diluted eleven ways across the most electronegative element there is, fluorine's lone pairs are pulled lower still by two Sb(V) centres, and it offers no soft site for a carbocation to find. A cation made here has nothing to react with, which is why George Olah could park carbocations in a magnet and record sharp NMR spectra of species that survive nanoseconds or less in ordinary solvents.

It also explains the obsession with dryness. Water levels everything: in aqueous solution the strongest acid that can exist is H₃O⁺. Water's proton affinity is 691 kJ/mol, so any trace in HF-SbF₅ is irreversibly converted to H₃O⁺Sb₂F₁₁⁻. Superacidity is a property of the medium, and water destroys the medium.

Measuring an Acid With No Water In It

pH is a statement about H₃O⁺ in water; take the water away and it means nothing. The replacement is the Hammett acidity function of 1932: dissolve a very weak base B of known pK_BH+, measure how much is protonated, and define H₀ = pK_BH+ − log([BH⁺]/[B]). In dilute aqueous acid H₀ collapses back to pH. Conant coined superacid in 1927; Gillespie set the threshold still in use, an H₀ below that of 100% sulfuric acid. 1:1 HF-SbF₅ is placed at H₀ = −28, against −12 for 100% sulfuric acid, making it about 10¹⁶ times more protonating.

Worked example. Take an indicator base with pK_BH+ = −10. In 100% sulfuric acid, log([BH⁺]/[B]) = pK_BH+ − H₀ = −10 − (−11.93) = 1.93: about 85 molecules in 86 protonated, a visible colour change. In fluoroantimonic acid the same arithmetic gives −10 − (−28) = 18, a ratio of 10¹⁸:1 — protonation is total and the dye has nothing left to report: an indicator can only read a ratio within roughly two powers of ten of its own pK_BH+. Indicators therefore run out of road long before H₀ = −28. Measured values for HF-SbF₅ saturate near −21 to −22 once the ladder is exhausted, and −28 is the extrapolated figure for the SbF₅-rich end; past that the ranking goes spectroscopic, with Farcasiu and Ghenciu reading acidity off the ¹³C shifts of a protonated ketone probe, mesityl oxide, calibrated against the overlapping indicator ladder.

The anion side checks out too. At 298 K one power of ten costs RT ln 10 = 5.71 kJ/mol, so the 150 kJ/mol fluoride-affinity gap between SbF₅ and BF₃ would buy 26 orders of magnitude if it all became free energy in solution. The real gap — HF-BF₃ near H₀ −17, HF-SbF₅ at −28 — is about 11 orders, the rest lost to solvation and ion pairing.

What It Does to Methane: Protonating a Sigma Bond

Methane is the least cooperative substrate in organic chemistry: no lone pair, no π system, no dipole. Its proton affinity is 543.5 kJ/mol, far below ammonia's 853 or water's 691. Under enough proton pressure it protonates anyway, and where the proton lands is the interesting part.

• Step 1 — the approach. H₂F⁺ meets CH₄. With no lone pair on offer, the proton attacks the electron pair sitting inside a C–H bond.
• Step 2 — the product. The result is CH₅⁺, methanium: a CH₃ tripod plus an H₂ unit slung underneath by one three-centre two-electron bond — the motif that holds diborane's bridging hydrogens in place.
• Step 3 — the electron count. CH₅⁺ has 5 hydrogens but only 4 bonding pairs and 8 valence electrons on carbon. Carbon is five-coordinate, not five-bonded, and not hypervalent: the octet holds, because one pair does the work of two bonds.
• Step 4 — the fluxion. The hydrogen-scrambling barrier is under 1 kcal/mol (~4 kJ/mol), below the zero-point energy, so all 120 permutations interconvert freely. On any spectroscopic timescale the five hydrogens are equivalent and the ion has no fixed structure worth drawing.
• Step 5 — the exit. CH₅⁺ sheds H₂ to leave CH₃⁺, a genuine carbocation and the doorway into Olah's chemistry: it attacks another methane, C–C bonds form, and methane fed into HF-SbF₅ ends up as tert-butyl cation. Run it in DF-SbF₅ and deuterium appears in the recovered methane — proof the protonation is real and reversible.

CH₅⁺ was seen long before it was understood: Tal'rose and Lyubimova found it in a mass spectrometer in 1952, and Takeshi Oka's group recorded its infrared spectrum in 1999 — about a thousand lines with no usable pattern, the spectrum of a molecule scrambling faster than it vibrates.

Hardware: Why Teflon Holds It and Glass Does Not

Fluoroantimonic acid is made and kept in fluoropolymers — PTFE, PFA, FEP, and PCTFE (Kel-F) where clarity is wanted — for a mechanistic reason. In PTFE every carbon is already saturated with fluorine: no C–H bonds to attack, no π system, no lone pairs at an energy worth protonating.

Glass presents a great deal, and the attack is not the proton's doing at all — it is fluoride chemistry: SiO₂ + 4 HF → SiF₄↑ + 2 H₂O, then SiF₄ + 2 HF → H₂SiF₆. Two failures at once: the vessel leaves as a gas, and the water it makes poisons the acid.

Metals are counter-intuitive in the same way. Anhydrous HF, shipped as UN 1052, travels in carbon-steel cylinders because a dry fluoride film passivates the surface; it is wet HF that eats steel, which is why aqueous HF ships in polyethylene and HF piping is Monel 400 or Hastelloy C-276. Fluoroantimonic acid has no UN number of its own: it is made in situ, not shipped.

Isomerising n-butane to isobutane near ambient temperature has been demonstrated on HF-SbF₅, because the equilibrium favours the branched isomer exactly where solid catalysts are too slow. Refineries still do not use it: antimony leaches into the product, feed water destroys the acid, and the plant would need fluoropolymer lining throughout. The rule that matters most is medical — fluoride burns are treated with calcium gluconate because F⁻ strips Ca²⁺ and Mg²⁺ systemically, and a concentrated HF burn over a few percent of body area can kill by hypocalcaemia hours later.

History, and the Incidents That Belong to the HF Half

Conant, Hammett and Gillespie gave the field a ruler; Olah supplied the motive. Working at Dow and then Case Western Reserve, he wanted to see carbocations rather than infer them, and SbF₅-based media were the only solvents inert enough to let one sit still. Magic Acid got its name at Christmas 1966, when Joachim Lukas dropped candle wax into HSO₃F-SbF₅ and pulled a clean tert-butyl cation spectrum out of the dissolved paraffin. The Nobel Prize followed in 1994.

The disaster record belongs to the hydrogen fluoride half: HF is handled by the tonne, fluoroantimonic acid is not. The 1986 Goldfish field trials in Nevada released pressurised anhydrous HF deliberately and killed the assumption that it rains out as a puddle: it flashes into a dense, ground-hugging aerosol that travels. A year later a crane at Marathon's Texas City refinery ruptured an HF line, releasing tens of tonnes and sending around a thousand people for treatment. In 2012 an HF tanker transfer in Gumi, South Korea, killed five workers; in 2015 an explosion at ExxonMobil's Torrance refinery threw debris within metres of the HF alkylation settler. None of this is superacidity — it is the fluoride chemistry that dissolves glass and bone.

Failure Modes and the Look-Alikes It Gets Confused With

Water is the failure mode. There is no slow degradation: moisture converts the working ion pair into H₃O⁺Sb₂F₁₁⁻ and hydrolyses SbF₅ to antimony oxides, and the acidity falls off a cliff.

It is not an oxidiser. Acid strength and oxidising power are unrelated axes. Chlorine trifluoride sets sand alight because it is a ferocious oxidiser; aqua regia dissolves gold because nitric acid oxidises the metal and chloride carries it off as AuCl₄⁻. Fluoroantimonic acid only protonates, which is why it destroys paraffin but sits happily in a Teflon bottle.

Hydrofluoric acid is the classic confusion. They share a component and the glass-eating habit, yet sit at opposite ends of the scale: hydrofluoric acid in water is a weak acid, pKa 3.17, held back by hydrogen bonding and the stability of HF₂⁻. It is dangerous for a reason unrelated to acidity: the neutral molecule crosses skin and the fluoride poisons calcium metabolism from within.

And "the strongest acid" is ambiguous. This is the strongest common acidic medium, gauged by H₀. As isolated compounds the carborane acids such as H(CHB₁₁Cl₁₁) are stronger Brønsted acids — gas-phase deprotonation enthalpy near ~240 kcal/mol against sulfuric acid's ~302. Strength and violence are not the same measurement.

Where fluoroantimonic acid sits on the acidity ladder, and what its neighbours are actually used for

Acid or system	Hammett H<sub>0</sub>	Conjugate base	What it does
100% sulfuric acid, H<sub>2</sub>SO<sub>4</sub>	−11.93, the −12 benchmark	HSO<sub>4</sub><sup>−</sup>	Gillespie's dividing line: below this you are a superacid. Protonates ketones and alcohols; ignores alkanes.
Triflic acid, CF<sub>3</sub>SO<sub>3</sub>H	about −14.1	CF<sub>3</sub>SO<sub>3</sub><sup>−</sup>	The practical bench superacid. Hygroscopic but air-stable, miscible with water, sold in glass bottles.
Anhydrous HF	about −15.1	F<sup>−</sup> and HF<sub>2</sub><sup>−</sup>	Superacidic only when bone dry. Diluted into water it becomes a weak acid, pKa 3.17, yet still etches glass.
Magic acid, HSO<sub>3</sub>F-SbF<sub>5</sub>	about −20 to −23, composition-dependent	[SbF<sub>5</sub>(OSO<sub>2</sub>F)]<sup>−</sup> and oligomers	Olah's workhorse for long-lived carbocations. Dissolved a Christmas candle at room temperature in 1966.
Fluoroantimonic acid, 1:1 HF-SbF<sub>5</sub>	−28	Sb<sub>2</sub>F<sub>11</sub><sup>−</sup>	About 10<sup>16</sup> times more protonating than 100% sulfuric acid. Protonates methane to CH<sub>5</sub><sup>+</sup>.
Carborane acid, H(CHB<sub>11</sub>Cl<sub>11</sub>)	no H<sub>0</sub> of its own — it is a solid, not a medium	CHB<sub>11</sub>Cl<sub>11</sub><sup>−</sup>	Among the strongest isolable single-compound Brønsted acids — its fluorinated cousin H(CHB<sub>11</sub>F<sub>11</sub>) is stronger still — and gentle enough that protonated-benzene salts can be bottled.