The Woodward-Hoffmann Rules

What the rules actually predict

Most reactions in organic chemistry go through charged or radical intermediates — a carbocation forms, a nucleophile attacks, a proton leaves. Pericyclic reactions are the strange family that has none of that. Bonds break and form in a single concerted step, moving around a closed loop of overlapping p-orbitals, with no intermediate at all. There is one transition state and the electrons flow in a ring.

Because everything happens in one motion, the outcome is fixed by the geometry and phase of the orbitals in that cyclic array — not by which face is more nucleophilic or which cation is more stable. The Woodward-Hoffmann rules are the bookkeeping that tells you, before you run the reaction:

1. Whether the reaction is allowed at all under a given condition (heat vs. light). An "allowed" reaction has a low, symmetry-permitted barrier; a "forbidden" one must pay a huge orbital-symmetry penalty and effectively will not go by the concerted route.
2. What stereochemistry the product must have. For a ring closure, the rules force a single answer — conrotatory or disrotatory — and that answer dictates whether two substituents end up cis or trans. It is not a selectivity you can push; it is a requirement.

There are three sub-families of pericyclic reaction, and one rule underneath all of them:

• Electrocyclic — an open π chain closes to a ring (or the ring opens). Butadiene ⇌ cyclobutene.
• Cycloaddition — two π systems join into a ring. The Diels-Alder [4+2] and the alkene [2+2].
• Sigmatropic — a σ bond migrates across a π system. [3,3] Cope/Claisen shifts, [1,5]-hydrogen shifts.

The principle: conservation of orbital symmetry

The deep statement, made by Woodward and Hoffmann and put on a rigorous footing by Longuet-Higgins and Abrahamson, is this: in a concerted reaction, the symmetry of every occupied molecular orbital is conserved from start to finish. A bonding orbital in the reactant must correlate smoothly with a bonding orbital in the product; it is not allowed to turn into an antibonding orbital along the way.

You track this with a correlation diagram. Identify a symmetry element that persists through the whole transition state — a mirror plane (σ) for a disrotatory closure, a two-fold rotation axis (C₂) for a conrotatory closure. Classify each MO as symmetric (S) or antisymmetric (A) with respect to that element, then connect reactant orbitals to product orbitals of matching symmetry, in energy order.

• If all the filled reactant orbitals connect to filled bonding product orbitals, the pathway is allowed — the barrier is a normal thermal barrier.
• If a filled bonding orbital in the reactant is forced to correlate with an empty antibonding orbital in the product, the ground states don't connect. That crossing puts an enormous symmetry-imposed barrier in the way — the reaction is forbidden by the concerted route.

The full correlation diagram is the rigorous tool, but for a quick answer you rarely need it. Two shortcuts get you there faster.

The frontier-orbital shortcut

Kenichi Fukui's frontier molecular orbital (FMO) approach reduces the whole thing to looking at one orbital: the highest occupied molecular orbital (HOMO) of the π system (or, for a cycloaddition, the HOMO of one partner and the LUMO of the other). New bonds form only where orbital lobes of the same phase can overlap. Draw the HOMO, look at the sign of the wavefunction at the two ends, and ask which motion brings like-phase lobes together.

    Butadiene, 4 π electrons.  Thermal HOMO = ψ₂.
    Phases at the two termini (C1 and C4) are OPPOSITE:   (+)C1 ··· C4(−)

      To make the new C1–C4 σ bond, like phases must meet.
      → both ends rotate the SAME way  ⇒  CONROTATORY (thermal, 4n)

    Hexatriene, 6 π electrons.  Thermal HOMO = ψ₃.
    Phases at the two termini (C1 and C6) are the SAME:   (+)C1 ··· C6(+)

      Like phases already point the same way.
      → the ends rotate OPPOSITE ways  ⇒  DISROTATORY (thermal, 4n+2)

Photochemistry: absorb a photon, promote one electron HOMO → LUMO. The LUMO has one extra node, so its terminal phases are inverted relative to the old HOMO. Every conclusion flips. Butadiene → cyclobutene becomes disrotatory under light; hexatriene → cyclohexadiene becomes conrotatory. Same molecule, opposite stereochemistry, chosen purely by heat or light.

The one rule that covers everything

The fastest bookkeeping — the version you memorize for an exam — is the generalized Woodward-Hoffmann selection rule. Break the reaction into components (a "component" is a continuous set of orbitals: a π bond, a σ bond, a whole polyene face). Label each by its electron count and by whether the reaction uses it suprafacially (both new bonds on the same face, subscript s) or antarafacially (bonds on opposite faces, subscript a).

A ground-state (thermal) pericyclic reaction is symmetry-allowed when the total number of (4q+2)_s and (4r)_a components is odd.

For a photochemical (excited-state) reaction, an even count is allowed.

Worked as a checklist for the Diels-Alder: the diene is a 4-electron component used suprafacially — that is a (4r) component but it is suprafacial, so it does not count. The dienophile is a 2-electron component used suprafacially — a (4q+2)_s component, so it counts. Total = 1, which is odd → thermally allowed. Every named rule below is just this single rule applied to a specific component set.

The three families, tabulated

Because heat and light flip every entry, the whole subject collapses into one table. "Total electrons" means the number of electrons flowing in the cyclic transition state.

Reaction type	Total electrons	Thermal (Δ)	Photochemical (hν)
Electrocyclic	4n (e.g. butadiene, 4 e⁻)	Conrotatory	Disrotatory
Electrocyclic	4n+2 (e.g. hexatriene, 6 e⁻)	Disrotatory	Conrotatory
Cycloaddition	4n (e.g. [2+2], 4 e⁻)	supra-antara (forbidden supra-supra)	supra-supra (allowed)
Cycloaddition	4n+2 (e.g. [4+2], 6 e⁻)	supra-supra (allowed)	supra-antara
Sigmatropic [1,j]-H	4n (e.g. [1,3], 4 e⁻)	Antarafacial (H can't reach → forbidden)	Suprafacial (allowed)
Sigmatropic [1,j]-H	4n+2 (e.g. [1,5], 6 e⁻)	Suprafacial (allowed)	Antarafacial
Sigmatropic [3,3]	6 e⁻ (Cope, Claisen)	Suprafacial-suprafacial (allowed, chair TS)	—

Read one line and you have the answer. The pattern is mechanical: for a given family, thermal and photochemical are always opposites, and 4n and 4n+2 are always opposites — so any one entry fixes the other three.

Worked example: the butadiene–cyclobutene ring opening

The cleanest experimental proof of the rules is the thermal ring opening of a substituted cyclobutene. Take cis-3,4-dimethylcyclobutene and trans-3,4-dimethylcyclobutene and heat each to about 175 °C. The C3–C4 σ bond breaks and the ring opens to a 2,4-hexadiene. This is a 4-electron electrocyclic process, so the thermal rule says conrotatory — and the two methyls report exactly which way the termini rotated.

    cis-3,4-dimethylcyclobutene   ──Δ, 175 °C──→   (E,Z)-2,4-hexadiene
        (conrotatory: one CH₃ swings up, the other down)

    trans-3,4-dimethylcyclobutene ──Δ, 175 °C──→   (E,E)-2,4-hexadiene  (major)
        (conrotatory again — the two allowed conrotatory modes are
         degenerate here, and outward rotation of both methyls wins
         on steric grounds, giving the E,E product cleanly)

• The prediction. Conrotation of the cis isomer necessarily sends one methyl up and one down as the p-orbitals turn, producing one E and one Z double bond — the (E,Z)-diene. Disrotation (the forbidden thermal path) would have given the (E,E) or (Z,Z) diene. The (E,Z) product is what you get, so the thermal path is conrotatory, exactly as predicted.
• The photochemical control. Irradiate the same cyclobutene and the 4-electron system now opens disrotatory, delivering the opposite diene geometry. The switch in stereochemistry with no change in substrate is the fingerprint of orbital-symmetry control.
• Why it matters historically. Before 1965 these stereochemical outcomes looked arbitrary. Woodward noticed them while working on the vitamin B₁₂ synthesis, asked Hoffmann to build the orbital picture, and the "arbitrary" results turned out to be dictated by a single symmetry principle.

Scope, selectivity, and stereochemistry

The rules are unusually powerful because they predict stereochemistry with essentially no exceptions when the reaction is genuinely concerted. A few practical consequences:

• Diels-Alder is suprafacial on both partners. The 6-electron [4+2] is supra-supra allowed thermally, which locks in the two hallmark stereochemical results: the cis principle (substituents that are cis on the dienophile stay cis in the product) and the endo rule (the endo transition state is favored by secondary orbital overlap). Both are downstream of the supra-supra requirement.
• Thermal [2+2] essentially doesn't happen concertedly. Two alkenes cannot both add antarafacially, so the allowed thermal path is inaccessible and the supra-supra path is forbidden. Thermal cyclobutane formation therefore goes by a stepwise biradical instead — or waits for light, which makes supra-supra allowed (the basis of [2+2] photocycloadditions and DNA thymine-dimer damage from UV).
• Ketenes cheat the [2+2]. A ketene (R₂C=C=O) can add to an alkene thermally in a genuine concerted [2+2] because its orthogonal second π system lets it act as the antarafacial component — the supra-antara path the rule allows. Its perpendicular geometry is the loophole.
• [1,5]-H shifts are ubiquitous; [1,3]-H shifts are not. Cyclopentadiene scrambles its ring hydrogens by facile thermal [1,5]-H shifts (6-electron, suprafacial, allowed). The analogous [1,3]-H shift is thermally forbidden suprafacially, and the allowed antarafacial route is geometrically impossible for a hydrogen 1s orbital, so 1,3-shifts of H are vanishingly rare.
• Carbon can migrate with inversion. In a [1,3]-carbon shift, the migrating carbon can use both lobes of its p-orbital, so it can travel suprafacially with inversion at the migrating center — the allowed thermal path. Retention would be forbidden. Berson's deuterium-labeled bicyclic systems confirmed exactly this inversion.

Woodward-Hoffmann vs. Hückel's rule vs. FMO theory

These three ideas all use the same 4n / 4n+2 counting and are easy to confuse. They answer different questions.

	Woodward-Hoffmann rules	Hückel's aromaticity rule	Frontier MO (FMO) theory
Question answered	Is a pericyclic reaction allowed, and with what stereochemistry?	Is a cyclic π molecule aromatic (stable) or antiaromatic?	Where and how strongly do two molecules react?
Author & year	Woodward & Hoffmann, 1965	Erich Hückel, 1931	Kenichi Fukui, 1952
What the count means	4n+2 electrons in the transition state → supra-supra / disrotatory (thermal)	4n+2 π electrons in a ground-state ring → aromatic	Not a count — looks at HOMO–LUMO energy gaps and coefficients
Applies to	Transition states of concerted reactions	Stable ground-state molecules	Any two reacting orbitals
Role of light	Central — hν inverts every outcome	None (ground-state property)	Explains why hν flips W-H via HOMO→LUMO
Relationship	An aromatic (4n+2) transition state is the "allowed" one (Zimmerman-Dewar view)	Supplies the 4n+2 concept W-H borrows	The practical shortcut for applying W-H
Nobel Prize	Hoffmann (shared 1981)	No (died 1980, unrecognized)	Fukui (shared 1981)

The unifying insight — the Zimmerman-Dewar aromatic transition state model — is that an "allowed" thermal pericyclic transition state is simply an aromatic one: a Hückel array with 4n+2 electrons, or a Möbius array (one phase inversion) with 4n electrons. Woodward-Hoffmann and Hückel are two faces of the same 4n+2 idea, one for transition states and one for molecules.

Limitations and where the rules stop applying

• Only concerted reactions obey them. The rules govern the single-step, no-intermediate pathway. If a reaction can drop into a biradical or zwitterionic intermediate, the symmetry constraint evaporates — a "forbidden" transformation just takes the stepwise road. Thermal [2+2] via a biradical is the classic example.
• "Forbidden" means high barrier, not impossible. A symmetry-forbidden concerted path has an extra barrier of roughly 40–60 kJ/mol from the avoided crossing. Given enough temperature, or a substrate that relieves strain on reaction, a forbidden path can still be observed — it is disfavored, not banned by physics.
• Substituents can subvert the geometry. Torquoselectivity — whether a given substituent rotates inward or outward during a conrotatory opening — is set by donor/acceptor effects that the bare symmetry rule does not capture. The rule tells you conrotatory; it does not, by itself, tell you which of the two conrotatory modes wins.
• Metals rewrite the rulebook. Transition-metal d-orbitals can supply the extra symmetry the organic system lacks, so metal catalysis routinely enables "forbidden" transformations — a formally forbidden [2+2] or a [2+2+2] cyclotrimerization runs smoothly on cobalt or rhodium because the metal orbitals carry the reaction around the symmetry barrier.
• They predict feasibility, not rate or yield. An allowed reaction can still be slow if the geometric approach is strained; the rules set the symmetry permission, and ordinary kinetics and sterics do the rest.

History: a synthesis puzzle that became a Nobel Prize

The rules were born from a practical frustration. In the early 1960s Robert Burns Woodward, mid-way through the monumental total synthesis of vitamin B₁₂ at Harvard, kept seeing electrocyclic ring closures give a single, "unexpected" stereochemistry that ordinary steric arguments could not explain. In 1965 he took the problem to a young theoretician down the hall, Roald Hoffmann, who applied extended Hückel molecular-orbital calculations and found that the terminal-orbital phases dictated the rotation.

Their conclusions appeared as a rapid series of communications in the Journal of the American Chemical Society in 1965 — on electrocyclic reactions, cycloadditions, sigmatropic shifts, and group transfers — and were gathered in the 1969 monograph The Conservation of Orbital Symmetry. Independently, Kenichi Fukui in Japan had introduced the frontier-orbital idea back in 1952, providing the practical shortcut, and H. C. Longuet-Higgins and E. W. Abrahamson gave the formal correlation-diagram proof.

The 1981 Nobel Prize in Chemistry went jointly to Roald Hoffmann and Kenichi Fukui "for their theories, developed independently, concerning the course of chemical reactions." Woodward, who by convention could have shared it, had died in 1979 — and Nobel Prizes are not awarded posthumously — so his name lives on in the rules but not on that particular medal (he had already won the 1965 Prize for synthesis). Chemists still call the whole framework the Woodward-Hoffmann rules.

Why chemists still reach for them daily

• Retrosynthetic planning. When you need a six-membered ring with defined stereochemistry, an allowed thermal Diels-Alder is often the first disconnection a chemist draws — precisely because the rules guarantee the relative configuration of up to four new stereocenters in one step.
• Predicting stereochemistry for free. Once you know a step is a 4n conrotatory electrocyclic closure, the cis/trans relationship of the product is fixed before you touch a flask. That is rare in organic chemistry, where stereochemistry usually has to be coaxed.
• Explaining natural product biosynthesis. Vitamin D₃ forms in your skin by a 6-electron conrotatory (photochemical) electrocyclic ring opening of 7-dehydrocholesterol followed by a thermal [1,7]-H shift — a textbook Woodward-Hoffmann sequence running in living tissue. The endiandric acids and many terpenoid frameworks assemble by cascades of allowed electrocyclizations and cycloadditions.
• Understanding photochemical damage. UV light drives the otherwise-forbidden [2+2] cyclodimerization of adjacent thymine bases in DNA, forming the cyclobutane pyrimidine dimers responsible for sunburn and skin-cancer-initiating mutations. The rules explain why the damage needs light and cannot happen in the dark.