Olefin Metathesis

The double-bond switchboard

Most reactions that make a carbon–carbon bond build it up atom by atom, or splice two halves together once and for all. Olefin metathesis does something stranger and more powerful: it breaks existing C=C double bonds and re-forms them with different partners, like cutting two strings of beads at the knot and re-tying them crossed. Take two terminal alkenes — say two molecules of 1-pentene, CH₃CH₂CH₂CH=CH₂. Metathesis snips each at the double bond and recombines the two butylidene halves into a new internal alkene, 4-octene (CH₃CH₂CH₂CH=CHCH₂CH₂CH₃), while the two leftover methylene caps combine and float away as ethylene gas. The carbon skeleton has been re-stitched at the double bond, and nothing was added but a trace of catalyst.

The word metathesis comes from the Greek for "transposition" or "change of position," and that is exactly what happens: the alkylidene fragments on either side of each double bond change places. Because the catalyst is regenerated every cycle, a few mole-percent — sometimes a few hundred parts per million — of metal can turn over thousands of substrate molecules. That combination of skeletal editing and catalytic efficiency is why metathesis reshaped synthetic chemistry across pharmaceuticals, polymers, and petrochemicals.

The Chauvin mechanism: a ring that breaks the other way

For two decades after the first reports of "olefin disproportionation" in the late 1950s, no one knew how the bonds were being swapped. Several wrong mechanisms were proposed — pairwise schemes in which two alkenes met on the metal and traded ends directly. In 1971 Yves Chauvin (with his student Jean-Louis Hérisson) published the answer, and it is the mechanism we still teach today.

The catalytically active species is not a bare metal but a metal carbene, also called a metal alkylidene: a transition metal joined to a carbon by a formal double bond, M=CR₂. The cycle has just two repeating moves:

• [2+2] cycloaddition. An incoming alkene coordinates and adds across the M=C bond, fusing the metal, its carbene carbon, and the two alkene carbons into a strained four-membered ring — the metallacyclobutane. The metal sits at one corner.
• Retro-[2+2] cleavage. The square ring is unstable and pops open. Crucially, it can cleave along the perpendicular diagonal from the one that formed it. When it does, it ejects a brand-new alkene and leaves behind a new metal carbene bearing a different alkylidene group.

If the ring breaks the way it formed, you simply get your starting materials back — an unproductive turn. If it breaks the other way, the alkylidene fragments have been exchanged. The metal then catches the next alkene and repeats. Because both the cycloaddition and its reverse are low-barrier, microscopically reversible steps, the whole network of possible alkenes equilibrates. The chemist's job is to bias that equilibrium toward the one product they want.

One subtle but important point: a terminal alkene (R–CH=CH₂) reacting with a metal methylidene (M=CH₂) produces ethylene. That methylidene is both the least stable and the most reactive carbene in the cycle, and its decomposition is a major route to catalyst death — which is why turnover numbers for terminal-alkene cross metathesis are often lower than for the cleaner ring-opening polymerizations.

The catalysts: Grubbs and Schrock

Chauvin explained how; the practical revolution came when chemists learned to make stable, well-defined carbenes you could weigh out of a bottle. Two families dominate.

Schrock catalysts are high-oxidation-state molybdenum and tungsten alkylidenes, typically of the form (RO)₂(ArN)Mo=CHCMe₂Ph. They are extraordinarily active — they will metathesize hindered and electron-poor alkenes that stall other catalysts — and with chiral alkoxide ligands they enable enantioselective metathesis. The price is fragility: they are exquisitely sensitive to air, moisture, and many functional groups, so they live in a glovebox.

Grubbs catalysts are ruthenium benzylidenes, and their genius is robustness. The 1st-generation catalyst, Cl₂(PCy₃)₂Ru=CHPh, can be handled in air and even tolerates water, alcohols, aldehydes, and carboxylic acids — functional groups that would poison early metathesis systems. The 2nd-generation Grubbs catalyst replaces one bulky tricyclohexylphosphine with an N-heterocyclic carbene (NHC) ligand (an imidazolylidene). The strongly σ-donating NHC accelerates the slow dissociation step that frees a coordination site, sharply raising activity toward sterically demanding and electron-poor alkenes. The Hoveyda–Grubbs catalysts add a chelating ortho-isopropoxybenzylidene that detaches when substrate binds and re-binds the resting metal, making the catalyst more stable and often recoverable.

Catalyst	Metal & key ligand	Activity	Functional-group tolerance	Air/moisture stability
Grubbs 1st gen	Ru, two PCy₃	Moderate	High (alcohols, acids, aldehydes)	Good — bench-stable
Grubbs 2nd gen	Ru, NHC + PCy₃	High	High; handles hindered alkenes	Good
Hoveyda–Grubbs 2nd	Ru, NHC + chelating O	High	High; electron-poor alkenes	Excellent; recoverable
Schrock Mo/W	Mo or W alkylidene	Very high	Lower; sensitive substrates	Poor — glovebox only

This contrast — tunable, sensitive Schrock systems versus rugged, tolerant Grubbs systems — is exactly why the 2005 Nobel Prize was shared three ways. Chauvin supplied the mechanism, Schrock the first crisply defined and most active carbenes, Grubbs the catalysts a working chemist can actually use.

The flavors of metathesis

The same two-step cycle, pointed at different substrates, produces a whole toolbox of named transformations:

• Cross metathesis (CM). Two different alkenes are stitched into a new internal alkene; ethylene is usually the byproduct. Selectivity follows Grubbs's "type" classification of how readily each alkene homodimerizes.
• Ring-closing metathesis (RCM). A single molecule bearing two terminal alkenes (a diene) folds back on itself; the catalyst joins the two ends into a ring and expels ethylene. This is the standard way to build five-, six-, and larger rings, including strained medium rings and macrocycles.
• Ring-opening metathesis (ROM) and its polymer cousin ROMP. A strained cyclic alkene is opened; in ROMP the opening repeats to make a long unsaturated polymer chain. The reaction is "living" with good catalysts, giving narrow molecular-weight distributions.
• Acyclic diene metathesis (ADMET). A step-growth polymerization of dienes that builds linear unsaturated polymers, again venting ethylene.
• Ene–yne metathesis. An alkene and an alkyne combine to give a conjugated 1,3-diene — an atom-economical entry to dienes for later Diels–Alder chemistry.

What drives it: equilibrium and ring strain

A puzzling feature of metathesis is that, on paper, it looks thermoneutral. You break one carbon–carbon double bond and make another of similar bond strength (a C=C π bond is worth roughly 65–75 kcal/mol of the ~146 kcal/mol total C=C bond energy), so the equilibrium constant is often near 1. Selectivity therefore comes not from a big thermodynamic gradient but from shifting the equilibrium:

• Venting ethylene. In CM, RCM, and ADMET, the small alkene byproduct is gaseous ethylene (boiling point −104 °C). It bubbles out of solution and is swept away, and by Le Chatelier's principle its removal drags the reaction toward the cyclized or cross product. Running the reaction dilute also favors intramolecular ring closure over intermolecular oligomerization.
• Relieving ring strain. ROMP is the exception that is strongly downhill. Opening a strained ring releases its strain energy — about 27 kcal/mol for norbornene and around 7 kcal/mol for cyclooctene — which makes polymerization essentially irreversible for the most strained monomers. Cyclohexene, with almost no strain, refuses to undergo ROMP at all.

Because the cycle is reversible, metathesis can also be run "backward" for analysis or recycling: feed it a polymer and ethylene and it depolymerizes; feed it a mixture of internal alkenes and it equilibrates them. That same reversibility is the engine behind dynamic combinatorial libraries built on alkene exchange.

Where it shows up: drugs, plastics, and oil

In pharmaceutical synthesis, ring-closing metathesis is the standard tool for forging macrocycles that older cyclization methods made only in poor yield. It is a key step in the manufacture of the hepatitis-C protease inhibitors grazoprevir and the macrocyclic ring of related HCV drugs, and it has been used to assemble dozens of macrocyclic natural products (epothilones, the antifungal target structures, and more). Because Grubbs catalysts tolerate the polar functional groups common in complex molecules, RCM can be performed late in a synthesis without protecting everything in sight.

In materials, ROMP is a commercial workhorse. Polydicyclopentadiene (sold as Metton and Telene) is reaction-injection-molded into tough body panels and large parts; polyoctenamer (Vestenamer) is a rubber-processing additive; and ROMP's living character lets chemists make precisely defined block copolymers and functional materials. In petrochemicals, the Shell Higher Olefin Process (SHOP) and "olefin conversion technology" use metathesis at megaton scale to interconvert ethylene, propylene, and longer linear alkenes, balancing supply to market demand. More recently, cross metathesis of plant-oil-derived alkenes (commercialized by Materia and Elevance) turns renewable fats into specialty chemicals and surfactants.

Variant	What it makes	Driving force	Representative use
Cross metathesis (CM)	New internal alkene	Ethylene loss	Plant-oil upgrading, fragment coupling
Ring-closing (RCM)	Carbocyclic / macrocyclic ring	Ethylene loss + dilution	Grazoprevir, macrocyclic drugs
ROMP	Unsaturated polymer	Ring-strain relief	Polydicyclopentadiene, polynorbornene
ADMET	Linear unsaturated polymer	Ethylene loss	Precision polyethylene-like materials
Ene–yne	Conjugated 1,3-diene	Atom-economical addition	Diene synthesis for Diels–Alder

Selectivity, E/Z control, and limits

Two practical questions dominate real metathesis chemistry. The first is chemoselectivity: which alkenes react and which combine. Grubbs's empirical "type I–IV" ranking of alkenes by how fast they homodimerize lets chemists predict, for instance, that an electron-rich terminal alkene will cross cleanly with an electron-poor acrylate rather than dimerizing. The second is stereoselectivity: ordinary metathesis tends to give the thermodynamically favored E (trans) alkene, but many targets need the Z (cis) isomer. A major advance of the 2010s was the design of stereoretentive and Z-selective catalysts — specific Schrock Mo/W complexes and dithiolate-modified Grubbs-type Ru catalysts — that hold the new double bond cis by controlling the geometry of the metallacyclobutane intermediate.

The limits are real, too. Metathesis is poisoned by strongly coordinating groups such as basic amines, nitriles, and thiols if they out-compete the alkene for the metal, though protonation or protecting groups can sidestep this. Sterically congested tetrasubstituted alkenes are sluggish even for the best catalysts. And because terminal-alkene metathesis routes through the fragile methylidene, catalyst decomposition sets a ceiling on turnover for the most demanding cross-metathesis reactions. None of this dimmed the impact: by giving chemists reversible, catalytic control over the C=C double bond — the most common functional handle in organic chemistry — olefin metathesis became one of the defining reactions of modern synthesis.