Magic acid

Magic acid (FSO₃H·SbF₅) is a superacid made by mixing fluorosulfuric acid (HSO₃F) and antimony pentafluoride (SbF₅), usually in equal amounts. This combination of a Brønsted acid and a Lewis acid was created in the 1960s by Ronald Gillespie and his group at McMaster University. George Olah later used it to keep carbocations and hypercoordinated carbonium ions stable in liquid form. Magic acid and similar superacids help speed up the rearrangement of saturated hydrocarbons and can donate protons to very weak bases, such as methane, xenon, halogens, and hydrogen gas.

The term "superacid" first appeared in 1927, when James Bryant Conant discovered that perchloric acid could turn ketones and aldehydes into salts in nonaqueous solutions. Gillespie later coined the word "superacid" after Conant mixed sulfuric acid with fluorosulfuric acid and found the blend was millions of times more acidic than sulfuric acid by itself. The magic acid system came about in the 1960s, designed by Gillespie to study stable carbocations. He also used it to create electron-deficient inorganic cations. The name "magic acid" came from a Christmas party in 1966, when someone in Olah’s lab dropped a paraffin candle into the acid, which dissolved quickly. A ¹H-NMR scan of the solution showed a tert-butyl cation, meaning the wax’s paraffin chain had broken apart and rearranged into a stable tertiary carbocation. The name later appeared in a paper from Olah’s lab.

In terms of structure, a 1:1 mix of HSO₃F and SbF₅ works best for making carbonium ions, but other ratios have been studied. When the SbF₅-to-HSO₃F ratio is below 0.2, two main equilibria dominate the solution, as seen by ¹⁹F NMR. The first equilibrium accounts for about 80% of the NMR data, and the second for about 20%. As the ratio rises from 0.4 to 1.4, new NMR signals appear and grow stronger with more SbF₅, though the signals become less clear because the liquid gets thicker.

Any acid stronger than pure sulfuric acid is considered a superacid, measured by the Hammett acidity function. Sulfuric acid has an H₀ of −12, perchloric acid has −13, and the 1:1 magic acid system reaches −23. Fluoroantimonic acid, the strongest known superacid, is thought to have an extrapolated H₀ as low as −28.

Magic acid’s low nucleophilicity helps carbocations stay stable in solution. Classical trivalent carbocations, which are planar and sp²-hybridized, can be seen in this acid. Because the carbon has only six valence electrons, it is very electron-poor and reactive. These ions are easy to draw with Lewis structures since they only have two-electron, single bonds. Many tertiary cycloalkyl cations, like the 1-methyl-1-cyclopentyl cation, also form in superacidic solutions. This particular cation can come from either cyclopentane or cyclohexane; from cyclohexane, a secondary carbocation rearranges into the more stable tertiary form. Cyclopropylcarbenium ions, alkenyl cations, and arenium cations have also been observed.

As magic acid became more common, higher-coordinate carbocations were found. Penta-coordinate carbocations, or nonclassical ions, cannot be shown with just two-electron, two-center bonds. Instead, they use two-electron, three-center (or more) bonding, where two electrons spread across more than two atoms. This makes the bonds so electron-deficient that saturated alkanes can take part in electrophilic reactions. The discovery of these hypercoordinated ions fueled the nonclassical ion debate in the 1950s and 1960s. Because ¹H-NMR is slow, rapidly shifting positive charges on hydrogen atoms might be missed, but IR spectroscopy, Raman spectroscopy, and ¹³C NMR have been used to study bridged carbocation systems. One controversial example, the norbornyl cation, has been seen in several media, including magic acid. Its bridging methylene carbon is pentacoordinated, with three two-electron, two-center bonds and one two-electron, three-center bond. Quantum calculations show that the classical model is not an energy minimum.

Magic acid can also protonate alkanes. Methane, for instance, reacts to form the CH₅⁺ ion at 140 °C and normal pressure, though some heavier hydrocarbon ions appear as byproducts, along with hydrogen gas. When FSO₃D is used instead of FSO₃H, methane swaps hydrogen for deuterium, and HD gas is released instead of H₂. This shows that methane acts as a base, accepting a proton from the acid to form CH₅⁺. That ion then either loses a proton (explaining the hydrogen exchange) or loses a hydrogen molecule.