Stir cis-but-2-ene with diiodomethane, CH2I2, and a zinc–copper mixture in ether. A CH2 group bridges the two carbons of the double bond, and the product is cis-1,2-dimethylcyclopropane. Run the same reaction on trans-but-2-ene and you get trans-1,2-dimethylcyclopropane. The two methyls keep the relationship they had in the alkene.
A three-membered ring of carbons is a cyclopropane, and making one from an alkene is cyclopropanation. The piece that adds is a single carbon that forms two new bonds at once, one to each alkene carbon. This section is about that one-carbon piece, the carbene, and the two practical ways to deliver it.
Syn addition, and why the alkene geometry survives
Both new C–C bonds form in one step, on the same face of the flat alkene. That is a syn addition, the same pattern as OsO4 dihydroxylation in Oxidative cleavage & dihydroxylation. No intermediate forms in which a C–C bond could rotate, so the two methyls that started on the same side of the C=C end up on the same side of the ring.
A reaction in which each stereoisomer of the starting material gives its own stereoisomer of the product is stereospecific. Cyclopropanation is stereospecific: a cis alkene gives a cis cyclopropane and a trans alkene a trans one (the labels from Cis/trans and E/Z).
The two products are a useful pair for stereochemistry. cis-1,2-Dimethylcyclopropane has a mirror plane through the CH2 and between the two methyl-bearing carbons, so it is a meso compound, even though it has two stereocenters. trans-1,2-Dimethylcyclopropane has no mirror plane, so it is chiral. The CH2 adds to the top face and the bottom face of the flat alkene equally often, so the trans product forms as a racemic mixture of its two enantiomers.
What a carbene is
A carbene is a neutral carbon with only two bonds and two nonbonding electrons, six valence electrons in all. The simplest is methylene, :CH2. Its formal charge is zero: 4 valence electrons, minus 2 nonbonding, minus 2 bonds. It still lacks an octet, which makes carbenes extremely reactive.
In the form that adds to alkenes in one step, the two nonbonding electrons sit as a pair in an sp2 orbital, and the carbon’s remaining p orbital is empty. That one carbon therefore carries a lone pair, like a nucleophile, and an empty p orbital, like the carbocations of Carbocations. This form, with the two electrons paired, is called a singlet carbene.
Both features are used in the addition. The empty p orbital takes electrons from the alkene’s π bond, and the lone pair forms the second bond to the other alkene carbon, all in a single step. A carbene can also exist with its two electrons unpaired, one in each orbital (a triplet carbene). A triplet adds in two steps and loses the alkene’s geometry, so the stereospecific reactions on this page are the paired, singlet kind.
Dichlorocarbene from chloroform and base
Treat cyclohexene with chloroform, CHCl3, and potassium tert-butoxide. The product is 7,7-dichlorobicyclo[4.1.0]heptane: cyclohexene with a CCl2 bridging what used to be its double bond. The carbene here is dichlorocarbene, :CCl2, made in the flask in two steps.
- The base removes chloroform’s one hydrogen. That C–H is unusually acidic for a C–H bond because the three electronegative chlorines pull electron density away from the carbon and spread out the negative charge left behind (the inductive effect of Factors affecting acidity). The product is the trichloromethyl anion, Cl3C−.
- The anion loses a chloride ion. Its carbon keeps the lone pair and is left with two chlorines: :CCl2.
Both the H and the Cl leave from the same carbon, so this is called an α-elimination. In the E2 reaction they leave from neighboring carbons, which is a β-elimination and makes a C=C instead.
Dichlorocarbene adds in the same single syn step as any singlet carbene, so it is stereospecific too: cis-but-2-ene with CHCl3 and KOC(CH3)3 gives only the cis 1,1-dichloro-2,3-dimethylcyclopropane. Bromoform, CHBr3, with the same base gives dibromocarbene, :CBr2, the same way.
The Simmons–Smith reaction: adding a plain CH2
To add an unsubstituted CH2, the reaction from the opening of this page is the usual choice: CH2I2 with a zinc–copper couple, written Zn(Cu), zinc metal activated with a little copper. It is called the Simmons–Smith reaction. The zinc inserts into one C–I bond and gives iodomethylzinc iodide, ICH2ZnI.
ICH2ZnI is not a free carbene. It is a carbenoid: a reagent that delivers a CH2 group the way a carbene would, without ever releasing one. The CH2 carbon bonds to both alkene carbons in one step while the zinc and the iodide leave as ZnI2. The result is the same syn, stereospecific addition.
Free :CH2 can be made, by shining light on diazomethane, CH2N2, which loses N2. It is a poor reagent, though. Diazomethane is toxic and explosive, and free methylene is so reactive that it also pushes itself into C–H bonds and gives mixtures. The Simmons–Smith carbenoid adds only to the C=C, which is why it is the standard way to make a cyclopropane.
| Reagents | Species that adds | Group put on the ring | Stereochemistry |
|---|---|---|---|
| CH2I2, Zn(Cu) | carbenoid ICH2ZnI | CH2 | syn, stereospecific |
| CHCl3, KOC(CH3)3 | dichlorocarbene, :CCl2 | CCl2 | syn, stereospecific |
| CHBr3, KOC(CH3)3 | dibromocarbene, :CBr2 | CBr2 | syn, stereospecific |
| CH2N2, light | free methylene, :CH2 | CH2 | syn, but C–H insertion competes |
(E)-Pent-2-ene, CH3CH=CHCH2CH3, is treated with CH2I2 and Zn(Cu).
Step 1 — find the C=C and the group that adds. Simmons–Smith adds a CH2 across C2 and C3. The ring is C2, C3 and the new CH2, with a methyl on one ring carbon and an ethyl on the other: 1-ethyl-2-methylcyclopropane.
Step 2 — carry the geometry over. In the E alkene the methyl and the ethyl sit on opposite sides of the C=C. Syn addition keeps that, so they sit on opposite faces of the ring: trans-1-ethyl-2-methylcyclopropane.
Step 3 — count the stereoisomers. C1 and C2 are both stereocenters, and the trans product has no mirror plane, so it is chiral. Attack on the two faces of the alkene is equally likely, so both enantiomers form in equal amounts: a racemic mixture.
Check. The Z alkene would give the cis product instead. Unlike cis-1,2-dimethylcyclopropane, it is chiral, because its two ring substituents differ, so it too forms as a racemate.
A strained ring from a high-energy reagent
Cyclopropane carries about 27.6 kcal/mol of ring strain, the figure from Cyclohexanes, because its C–C–C angles are squeezed to 60° and its C–H bonds are all eclipsed. A reaction can still make such a ring easily, because the carbene or carbenoid starts out even higher in energy. Forming two C–C bonds at once releases more than enough energy to pay for the strain.
What carries forward
Cyclopropanation joins the syn additions to a C=C, beside OsO4 and epoxidation with mCPBA. All three add in one step to one face, so all three keep the alkene’s geometry. When a question asks which product forms from a cis or trans alkene, decide first whether the addition happens in one step on one face. If it does, the alkene’s geometry is carried straight into the product.