The Simmons-Smith Reaction

What the Simmons-Smith reaction does

Take any alkene and you want to staple a single CH₂ across the double bond, sewing the two carbons of the C=C into a strained three-membered cyclopropane ring. That is exactly what the Simmons-Smith reaction delivers — and it does so with a control that most cyclopropanation methods cannot match.

The reagent that performs the surgery is not a free carbene. It is a zinc carbenoid, iodomethylzinc iodide, written ICH₂ZnI. In this species the methylene carbon is bonded on one side to a leaving iodine and on the other to an electropositive zinc. That polarization is the trick: the carbon is electrophilic enough to accept the alkene's π electrons, but it is never released as an unstable, indiscriminate free carbene. The result is a clean, concerted, stereospecific delivery of exactly one CH₂ unit.

        CH₂I₂  +  Zn(Cu)  ──Et₂O, reflux──→  I-CH₂-Zn-I   (the carbenoid)

                    R₂C=CR₂  +  I-CH₂-Zn-I
                              │
                              │  concerted, syn CH₂ transfer
                              ▼
                         R₂C——CR₂          +   ZnI₂
                             \  /
                              CH₂           (cyclopropane)

Cyclopropanes matter far beyond the classroom. The ring's ~60° internal angles store roughly 27 kcal/mol of strain, its C-C bonds have unusual "banana" character with partial π-like density on the ring edges, and the whole motif is metabolically rigid — which is why it turns up in terpenes, pyrethroid insecticides, and dozens of modern drugs (ciprofloxacin, milnacipran, and many kinase inhibitors). The Simmons-Smith reaction is the workhorse for building the simplest one, the unsubstituted CH₂ ring.

The mechanism, step by step

The whole reaction is really two events: make the carbenoid, then transfer the CH₂.

1. Activate the zinc. A zinc-copper couple exposes a fresh, reactive Zn surface. (The copper is a promoter; it is not incorporated into the product.) The zinc undergoes oxidative insertion into one C-I bond of diiodomethane, exactly as magnesium inserts into a C-Br bond when you make a Grignard reagent. Two electrons flow from Zn⁰ into the σ* of the C-I bond, breaking it and giving I-CH₂-Zn-I.
2. Present the alkene. The π electrons of the double bond are the nucleophile. They approach the electrophilic methylene carbon of the carbenoid.
3. Concerted, three-centre transfer ("butterfly" transition state). This is the heart of it. In a single step, three bonds reorganize at once:
   • the alkene π bond becomes two new C-C σ bonds to the incoming CH₂ carbon;
   • the C-I bond of the carbenoid breaks, its electrons collapsing onto zinc;
   • the departing iodide pairs with zinc to leave as ZnI₂.
   The transition state is often drawn as a "butterfly": the CH₂ carbon sits above the plane of the alkene, its two forming bonds fanning down to the two alkene carbons while the C-I bond points away toward the incoming ZnI. Because both new bonds form simultaneously and to the same face, the addition is syn (suprafacial).
4. Release the ring. The alkene carbons are now sp³ and joined by the new CH₂ bridge; zinc leaves as ZnI₂. There is no arenium ion, no carbocation, no radical, and no discrete intermediate to trap. It is a one-flask, one-step methylene transfer.

The electron-arrow logic in the key step: the alkene π electrons attack the carbenoid carbon (arrow from C=C to CH₂), the C-I bonding electrons shift onto zinc (arrow from C-I to Zn), and iodide departs (arrow onto I, leaving as part of ZnI₂). Three arrows, one transition state — that concertedness is why nothing scrambles.

Reagents, catalyst, and real conditions

The exact recipe you pick controls how reproducible and how reactive the carbenoid is:

• Classic Simmons-Smith (1958). CH₂I₂ + Zn(Cu) couple in dry diethyl ether, heated at reflux (~35 °C) for several hours. Cheap and general, but the heterogeneous zinc surface makes batch-to-batch reactivity finicky.
• Furukawa modification (1966). Diethylzinc (Et₂Zn) + CH₂I₂ generates EtZnCH₂I, a homogeneous, more reproducible carbenoid, usable at or below 0 °C in CH₂Cl₂. This is the version most people run today.
• Shi modification. Adds trifluoroacetic acid to Et₂Zn/CH₂I₂ to give (CF₃CO₂)ZnCH₂I, a more electrophilic — and therefore faster — carbenoid that cyclopropanates even unactivated and electron-poor alkenes.
• Charette asymmetric version. Et₂Zn/CH₂I₂ plus a chiral dioxaborolane ligand (made from tartaric acid) delivers cyclopropanes on allylic alcohols in >90% ee — the enantioselective workhorse.

Zinc is stoichiometric, not catalytic. Each CH₂ delivered consumes one equivalent of the carbenoid and produces one equivalent of ZnI₂; there is no catalytic turnover of a single zinc atom the way palladium turns over in a cross-coupling. Solvent must be aprotic and Lewis-basic-friendly (Et₂O, DME, CH₂Cl₂). Substrates with free hydroxyl groups are a feature, not a liability — the -OH accelerates and directs the reaction (see below) rather than quenching the reagent, because zinc tolerates alkoxides where a Grignard reagent would be destroyed.

Scope, selectivity, and stereochemistry

Three properties make Simmons-Smith the go-to cyclopropanation:

• Stereospecific (retention of alkene geometry). Because the CH₂ adds syn in one concerted step, the relative configuration of the two alkene substituents is preserved. cis-2-butene gives cis-1,2-dimethylcyclopropane; trans-2-butene gives the trans ring. No cis/trans scrambling — the hallmark of a concerted, non-radical, non-cationic process.
• Diastereoselective, directed by oxygen. When an allylic or homoallylic hydroxyl (or ether) sits near the double bond, it coordinates zinc and delivers the CH₂ to the face syn to the oxygen. Diastereomer ratios above 95:5 are routine. This "hydroxyl-directed" delivery is the reaction's superpower and the reason it dominates in total synthesis.
• Chemoselective. The carbenoid ignores isolated C-H bonds (unlike a free carbene, which inserts into them) and tolerates esters, ethers, silyl groups, and even other double bonds if they are less accessible. More electron-rich alkenes react faster, so a trisubstituted enol ether will be cyclopropanated in preference to an isolated terminal olefin.

Rate ordering of alkenes reflects nucleophilicity: electron-rich, more-substituted double bonds react fastest, and a nearby free -OH can accelerate the reaction by two to three orders of magnitude relative to the same alkene with the -OH protected.

Simmons-Smith vs other cyclopropanation methods

	Simmons-Smith (Zn carbenoid)	Free carbene (:CH₂ from CH₂N₂/hν)	Metal carbene (Rh/Cu + diazo)
Active species	ICH₂ZnI carbenoid (C bonded to I and Zn)	Free singlet/triplet :CH₂	M=CHR metallocarbene
Reagent	CH₂I₂ + Zn(Cu) or Et₂Zn/CH₂I₂	CH₂N₂ + light or heat	N₂=CHCO₂R + Rh₂(OAc)₄ or Cu
Stereospecific?	Yes — syn, retains alkene geometry	Singlet yes, triplet no (scrambles)	Usually yes (singlet-like)
C-H insertion side reaction	None — leaves C-H untouched	Severe — inserts everywhere	Common — a designed feature or a nuisance
What CH₂/CHR is installed	Plain CH₂ (or CHR with RCHI₂)	Plain CH₂	CHR bearing ester/aryl (from diazo)
Directed by nearby -OH?	Yes — hydroxyl-directed, face-selective	No	Rarely
Hazard	CH₂I₂ irritant; ZnEt₂ pyrophoric	CH₂N₂ toxic and explosive	Diazo compounds shock-sensitive
Enantioselective version	Yes — Charette dioxaborolane, >90% ee	No	Yes — chiral Rh/Cu, high ee

Worked example: cyclopropanating an allylic alcohol

The textbook demonstration of directed, stereospecific delivery is the cyclopropanation of a cyclic allylic alcohol such as 2-cyclohexen-1-ol.

    2-cyclohexen-1-ol  +  CH₂I₂  ──Zn(Cu), Et₂O, reflux──→  bicyclo[4.1.0]heptan-2-ol
                                                             (CH₂ delivered cis to the -OH)

• Substrate. 2-cyclohexen-1-ol — a ring alkene with a free hydroxyl on the carbon adjacent to the double bond.
• Reagents. CH₂I₂ (2-3 equiv), Zn-Cu couple (2-3 equiv), dry Et₂O, reflux 2-6 h. A trace of I₂ helps initiate the zinc surface.
• What the -OH does. The hydroxyl coordinates zinc and tethers the carbenoid to the same face of the double bond that the -OH occupies. The CH₂ is delivered cis to the hydroxyl with high diastereoselectivity — often >95:5.
• Product. The fused bicyclic cyclopropane (a norcarane-type bicyclo[4.1.0] system) with the new ring on the hydroxyl face.
• Control experiment. Protect the -OH as its methyl ether or acetate and the selectivity collapses and the rate drops sharply — direct proof that it is the free oxygen doing the steering.

Contrast this with an unfunctionalized alkene like cyclohexene: it still cyclopropanates to give bicyclo[4.1.0]heptane (norcarane), but with no facial bias because there is no directing group. The lesson chemists exploit constantly: a temporary hydroxyl is a steering wheel for the incoming ring.

Real applications in synthesis

• Pyrethroid insecticides. The gem-dimethylcyclopropane carboxylic acid core of natural pyrethrins and synthetic pyrethroids (permethrin, cypermethrin) is a classic cyclopropanation target; Simmons-Smith-type methylene transfer is one route to such strained rings.
• Terpene and steroid synthesis. Hydroxyl-directed Simmons-Smith installs cyclopropane rings onto ring-fused alcohols with predictable facial selectivity — a staple in steroid and terpenoid total synthesis where a specific ring face must be functionalized.
• Pharmaceutical scaffolds. Cyclopropane rings appear in ciprofloxacin, milnacipran, and many kinase inhibitors because the rigid ring locks conformation and resists metabolism. The asymmetric Charette version builds these as single enantiomers.
• Homologation of alkenes. Because it adds exactly one carbon across a double bond, Simmons-Smith is a controlled one-carbon ring-forming homologation — cleaner than diazomethane, safer than a free carbene.
• Carbenoid probes and D-labelling. Using CD₂I₂ transfers a CD₂ group, a convenient way to place a deuterium-labelled methylene bridge for mechanistic and metabolic studies.

Limitations and side reactions

• Sluggish on electron-poor alkenes. The carbenoid carbon is electrophilic, so it needs a nucleophilic (electron-rich) alkene. Enones, acrylates, and other electron-poor olefins react slowly or not at all under classic conditions; the Shi (CF₃CO₂-modified) carbenoid or a metal-carbene route is better for those.
• Reproducibility of the Zn(Cu) couple. The heterogeneous zinc surface varies with preparation and moisture; a poorly activated couple simply fails to initiate. The Furukawa Et₂Zn/CH₂I₂ homogeneous version was developed precisely to fix this.
• Excess reagent and dihalide cost. CH₂I₂ is dense, expensive, and an irritant; reactions typically use 2-4 equivalents. Diethylzinc is pyrophoric and must be handled under inert atmosphere.
• Competing over-reaction. Substrates with two similarly reactive double bonds can be bis-cyclopropanated; if you want mono-cyclopropanation you rely on the -OH directing group or steric/electronic differences to discriminate.
• Not a C-H functionalization. Unlike a free carbene, the carbenoid does not insert into C-H bonds — usually a virtue, but it means you cannot use Simmons-Smith to make a ring where no alkene exists. You need the double bond first.

Historical discovery

The reaction was reported by Howard E. Simmons and Ronald D. Smith, both chemists at DuPont, in a pair of papers in the Journal of the American Chemical Society in 1958 and 1959. Their key insight was that treating diiodomethane with a zinc-copper couple gave a stable, isolable organozinc species — the carbenoid — that transferred a methylene group to alkenes stereospecifically, without the wild non-selectivity of the free carbenes then being generated from diazomethane. It was one of the first clean demonstrations that a "carbenoid" (a carbene equivalent still bonded to a metal and a leaving group) could behave far more predictably than the free carbene itself.

The method was refined over the following decades: Furukawa introduced the diethylzinc variant in 1966 for reproducibility, and in the 1990s André Charette developed the chiral-dioxaborolane asymmetric version that made enantioselective cyclopropanation of allylic alcohols routine. Today "Simmons-Smith" is used loosely to cover the whole family of zinc-carbenoid cyclopropanations.

Safety and practical notes

• Diiodomethane (CH₂I₂). A dense (2.5 g/mL), light-sensitive liquid; a skin and eye irritant. Store dark and cold; it slowly liberates I₂ (the purple tinge) on standing.
• Diethylzinc (Et₂Zn). Pyrophoric — ignites in air and reacts violently with water. Handle under argon or nitrogen with Schlenk technique; the Furukawa/Charette variants live or die on rigorous exclusion of air and moisture.
• Zinc-copper couple. Freshly prepared and stored under inert atmosphere; deactivates on exposure to air, which is a common cause of "the reaction just won't start."
• Workup. Quench cautiously — residual organozinc and zinc dust are reactive. Pyridine or saturated NH₄Cl is used to precipitate and remove zinc salts; the ZnI₂ byproduct is water-soluble and washes out.