Frustrated Lewis Pairs

How a frustrated Lewis pair works

An ordinary Lewis acid and Lewis base combine by donating the base's lone pair into the acid's empty orbital, forming a stable dative (donor–acceptor) adduct — think H₃N·BF₃. In a frustrated Lewis pair, both partners carry large substituents (tert-butyl or mesityl groups on phosphorus/nitrogen; three electron-poor pentafluorophenyl rings on boron) so that the acid and base cannot approach closely enough to bond. The lone pair and the empty p-orbital are left “frustrated”: chemically unquenched and pre-organized to act in concert.

Instead of bonding to each other, the two centers cooperatively clamp onto a small molecule that sits between them. With dihydrogen the base's lone pair pushes into the H–H σ* orbital while the boron's empty orbital pulls electron density out of the H–H σ bond. The bond breaks heterolytically — not into two H atoms but into a proton and a hydride — giving a phosphonium (or ammonium) cation and a borohydride anion, e.g. [t-Bu₃PH]⁺[HB(C₆F₅)₃]⁻. This is exactly the split that transition metals achieve via oxidative addition, but here it is done entirely with main-group elements.

Reagents, conditions and the H2 split

The defining FLP reagent is tris(pentafluorophenyl)borane, B(C₆F₅)₃ (“BCF”, first made by Massey and Park in 1963). Its three fluorinated aryl rings make boron extremely electrophilic yet very bulky, so it resists forming adducts with hindered donors. Typical bases are tri-tert-butylphosphine (P(t-Bu)₃), dimesitylphosphine, or hindered amines such as 2,2,6,6-tetramethylpiperidine (TMP) and 1,8-diazabicycloundec-7-ene (DBU).

• Solvent: non-coordinating aromatics or halogenated solvents — toluene, bromobenzene, dichloromethane — so the solvent does not itself quench the acid.
• Pressure: H₂ at roughly 1–100 bar; many benchtop reactions run at 1–4 bar.
• Temperature: H₂ splitting is often fast at 25 °C; Stephan's original linked phosphine-borane releases H₂ above ~100 °C and re-adds it at 25 °C, making the process reversible.

Beyond H₂, FLPs sequester CO₂ (the base attacks the carbon while boron binds an oxygen, giving a boryl-carboxylate that can be reduced toward methanol-level products), trap N₂O, SO₂, and NO, and add across alkenes and alkynes in 1,2-fashion.

Metal-free catalytic hydrogenation

The most important application is hydrogenation without any transition metal. After an FLP splits H₂ into [base-H]⁺ and [H–borate]⁻, the hydride is delivered to an electrophilic substrate and the proton follows, regenerating the free acid and base to close a catalytic cycle. Stephan and Erker developed the first practical examples around 2007–2008.

• Imines and enamines: polarized C=N bonds are ideal; the hydride reduces the carbon and the proton lands on nitrogen. Catalyst loadings of a few mol% B(C₆F₅)₃ give amines in high yield.
• Silyl enol ethers, aromatic aldehydes, and N-heterocycles such as quinolines and pyridines can be reduced under similar conditions.
• Bulky-substrate advantage: because FLP catalysis is metal-free, it tolerates sulfur- and phosphorus-containing groups that poison Pd or Pt catalysts.

Milder borane acids such as HB(C₆F₅)₂ (Piers' borane) and fluorinated aryl boranes with fewer or different fluorines broaden the substrate window and lower the pressures needed.

Asymmetric FLP catalysis and selectivity

Making FLP hydrogenation enantioselective requires chirality in the acid, the base, or the counterion. Chiral boranes and chiral phosphines have delivered good results: hydrogenation of prochiral imines and enamines has reached enantiomeric excesses in the 80–99% range in the best systems reported by the Repo, Du, and Klankermayer groups. Because the proton and hydride are transferred in a tight ion pair, the chiral environment around the [base-H]⁺···[H-borate]⁻ pair controls which prochiral face is reduced.

Selectivity also shows up as chemoselectivity: FLPs favor polarized unsaturations (C=N, C=O activated substrates) over unactivated C=C bonds, and they can reduce one functional group in the presence of others that would react on a metal surface. This complementary selectivity is a key reason FLPs are studied alongside, not merely as replacements for, classical catalytic hydrogenation.

Scope, limitations and current directions

The scope has expanded from boron/phosphorus to many main-group and even carbon-based partners: aluminum, gallium, silylium, and borenium cations serve as acids, while carbenes, pyridines, and hindered anilines serve as bases. Intramolecular FLPs — where acid and base sit on the same molecule with a rigid tether — give faster, more reversible H₂ uptake because the two sites are held at an ideal distance.

• Moisture and oxygen sensitivity: B(C₆F₅)₃ reacts with water and many donors, so FLP chemistry is typically run under inert atmosphere (glovebox or Schlenk line).
• Strong donors quench the acid: substrates or byproducts that bind boron too tightly can shut down the cycle, limiting some functional-group scope.
• Emerging uses: CO₂ reduction toward formate and methoxide-level products, C–H functionalization, dehydrogenation of amine-boranes for hydrogen storage, and “FLP-quenching” thermally activated delayed fluorescence and polymer applications.

A short history

The seeds were planted long before the name existed. In 1942 H. C. Brown noted that 2,6-lutidine binds BF₃ but not the bulkier BMe₃ — an early hint that steric mismatch can prevent adduct formation. G. Wittig and E. Frankland-era observations of sterically thwarted acid–base pairs followed sporadically.

The modern field was born in 2006, when Douglas W. Stephan's group reported reversible, metal-free H₂ activation in Science, and Stephan and Gerhard Erker coined the term frustrated Lewis pair shortly after. Within a few years FLPs went from a curiosity to a general strategy for small-molecule activation, and the concept now appears across catalysis, materials science, and CO₂ utilization research.

Classical Lewis adduct vs. frustrated Lewis pair

Property	Classical Lewis adduct	Frustrated Lewis pair
Acid–base interaction	Strong dative A←B bond	Sterically blocked, no bond
Example	H₃N·BF₃, Me₃P·B(C₆F₅)₃	t-Bu₃P + B(C₆F₅)₃
Reactivity toward H₂	Inert (sites saturated)	Heterolytic H₂ cleavage
Empty/lone-pair sites	Both consumed	Both remain accessible
Catalytic use	Rare	Metal-free hydrogenation & more