The Nazarov Cyclization

What the Nazarov cyclization does

Start with a divinyl ketone — a carbonyl with a vinyl group on each side (a cross-conjugated dienone, also called a 1,4-dien-3-one). Add an acid, and the molecule zips itself shut into a cyclopentenone: a five-membered ring carrying a ketone and one double bond. One new carbon-carbon bond forms between the two far ends of the vinyl groups, and that single bond turns an open chain into a ring.

The magic is that this is not a nucleophile attacking an electrophile in the usual sense. It is a pericyclic ring closure — the electrons flow around a cyclic array of orbitals in one concerted motion. Specifically it is a 4π-electron electrocyclization, the odd cousin of the Diels-Alder (which is a 6-electron cycloaddition). Because only four electrons are involved and the reaction is thermal, orbital symmetry forces the two ends to rotate in the same direction — a conrotatory closure — and that stereochemical fingerprint is what makes the Nazarov so useful for building rings with defined stereocenters.

        O                              O
        ‖                              ‖
   H₂C═CH   CH═CH₂   ──Lewis acid──→   ⌈‾‾‾‾‾⌉   (2-cyclopentenone)
        \\   /                          |     |
      divinyl ketone                    ⌊_____⌋
   (cross-conjugated dienone)         new C–C bond joins the two ends

The mechanism, arrow by arrow

Four moves take you from open chain to ring. Watch the charge: it starts on oxygen, spreads across a pentadienyl system, collapses into a new bond, and finally leaves as a proton.

1. Activate the carbonyl. A Lewis acid (BF₃, FeCl₃, TiCl₄) coordinates the ketone oxygen, or a Brønsted acid protonates it. Either way, positive charge is placed on the carbonyl, which now demands electron density from the two flanking π bonds. The molecule becomes a 3-oxypentadienyl cation: a chain of five p orbitals (the two C=C π bonds plus the electron-poor central carbon) holding 4 π electrons and a positive charge delocalized over the termini.
2. Conrotatory 4π ring closure. This is the heart of the reaction. The two terminal CH₂ carbons — the ends of the pentadienyl cation — rotate toward each other (in the same rotational sense, conrotatory) and their p-orbital lobes overlap to form the new σ bond. The four π electrons reorganize; the ring snaps shut. What was a linear cation is now a five-membered ring bearing a positive charge delocalized over three carbons and the oxygen — a cyclopentenyl oxyallyl cation (an allyl cation flanking the former carbonyl carbon, with the oxygen as an enol/enolate).
3. β-Proton elimination. A base in solution (the conjugate base of the acid, or the counterion) removes a proton from a carbon adjacent to the cationic allyl system. The electrons of that C-H bond flow in to quench the cation, regenerating a C=C double bond and giving the enol of the product.
4. Tautomerize. The enol tautomerizes to the more stable keto form (see keto-enol tautomerism), and the acid is released, regenerating the catalyst. The final product is the conjugated 2-cyclopentenone.

  divinyl ketone
        │  LA on O  (activation)
        ▼
  [3-oxypentadienyl cation]        ← 4 π electrons, + charge on ends
        │  conrotatory 4π closure  (new C–C σ bond)
        ▼
  [cyclopentenyl oxyallyl cation]  ← ring formed, + charge on allyl/O
        │  lose β-H  (elimination)
        ▼
  cyclopentenone enol
        │  keto ⇌ enol
        ▼
  2-CYCLOPENTENONE  +  regenerated acid

Note the electron bookkeeping never touches oxygen's lone pairs as nucleophiles — the oxygen's job is to hold the positive charge so the two carbon ends become electrophilic-enough termini of a 4π array. The conrotatory ring closure is the only step that makes or breaks the crucial C-C bond.

Reagents, catalysts, and conditions

The Nazarov is fundamentally acid-catalyzed. The classic promoters and a rough sense of when to reach for each:

• Brønsted acids. Phosphoric acid, polyphosphoric acid (PPA), and formic acid were Nazarov's own conditions — cheap, but harsh, and often stoichiometric. Modern chiral Brønsted acids (BINOL-phosphoric acids, triflimides) enable asymmetric versions.
• Lewis acids. BF₃·OEt₂, FeCl₃, TiCl₄, SnCl₄, AlCl₃ — strong, oxophilic, and common. Milder triflates — Sc(OTf)₃, Cu(OTf)₂, Zn(OTf)₂ — are often catalytic (2–10 mol%) and tolerate more functionality.
• Solvent and temperature. Typically dichloromethane, 1,2-dichloroethane, or nitromethane, run from −78 °C (for stereocontrol) up to reflux (for sluggish substrates). Low temperature slows the competing side reactions and preserves the conrotatory stereochemistry.
• Substrate polarization. A donor group (alkoxy, amino) on one vinyl and/or an acceptor (ester, ketone) on the other makes the cation far easier to form and steers the regiochemistry. These polarized Nazarov substrates react under much milder conditions than a bare divinyl ketone.
• β-Silicon control. Denmark's silicon-directed Nazarov places a trimethylsilyl group on a β-carbon; because R₃Si⁺ is an excellent cation-stabilizing "traceless" leaving group, elimination goes cleanly toward silicon and fixes the position of the product double bond.

Selectivity and stereochemistry

Two selectivity questions dominate the Nazarov, and both come straight from the mechanism:

• Torquoselectivity (which way the ends rotate). Conrotation can happen in two mirror-image senses. Substituents, chiral catalysts, or a pre-set stereocenter bias one sense over the other, controlling the relative and absolute configuration of the new stereocenters. This is why asymmetric Nazarov reactions with chiral Lewis acids (e.g. chiral N,N′-dioxide-Sc(III) complexes) can reach >90% ee.
• Regioselectivity of elimination (where the double bond ends up). The oxyallyl cation can lose a proton on either flank, giving two possible enone regioisomers. β-Silicon direction, polarization, or ring strain decides which one. Without such bias you get mixtures — the single most common practical headache of the reaction.

Because the ring-closing step is concerted and stereospecific, the Nazarov can build up to two adjacent stereocenters in one operation with predictable relative configuration — a big reason it survives in modern total synthesis despite being an old reaction.

Nazarov vs other cyclopentenone routes

	Nazarov cyclization	Pauson-Khand	Intramolecular aldol / Robinson
Ring made	Cyclopentenone (5-membered)	Cyclopentenone (5-membered)	Cyclohexenone (6-membered)
Bond-forming step	4π conrotatory electrocyclization	[2+2+1] metal-mediated cycloaddition	Enolate addition + dehydration
Key reagent	Lewis/Brønsted acid	Alkyne + alkene + CO / Co₂(CO)₈	Base (aldol) then acid/base (annulation)
Feedstock	Cross-conjugated divinyl ketone	Enyne + carbon monoxide	1,5-diketone / enone + ketone
Stereocontrol handle	Torquoselective conrotation, chiral LA	Chiral ligand on cobalt/rhodium	Enolate geometry, thermodynamics
Typical driving force	Cation quenched by proton loss	CO insertion, strong metal-C bonds	Loss of water, conjugation
Signature intermediate	Oxypentadienyl / oxyallyl cation	Cobalt-alkyne metallacycle	β-hydroxy ketone / enolate

Worked example: a silicon-directed Nazarov

Suppose you want a 2-cyclopentenone with the double bond in one specific place, not a mixture. Denmark's β-silyl divinyl ketone solves it.

    (E)-β-(trimethylsilyl) divinyl ketone
                │
                │  FeCl₃ (1.1 eq), CH₂Cl₂, −78 → 0 °C
                ▼
    3-oxypentadienyl cation  ── conrotatory 4π closure ──► oxyallyl cation
                │
                │  β-elimination TOWARD silicon (loses R₃Si⁺, not H⁺)
                ▼
    single-regioisomer 2-cyclopentenone   (yields commonly 70–90%)

• Reagents. The divinyl ketone bearing a TMS group at the β-position of one vinyl, plus a Lewis acid such as FeCl₃, BF₃·OEt₂, or TiCl₄ (roughly stoichiometric for reliable turnover in this classic version).
• Conditions. Anhydrous dichloromethane, −78 °C addition then slow warm to 0 °C, minutes to a couple of hours.
• Why silicon. After ring closure, the oxyallyl cation is β to the C-Si bond. Silicon hyperconjugatively stabilizes the developing positive charge (the β-silicon effect), so the C-Si bond breaks in preference to any C-H bond. The silyl group leaves as a silyl cation (trapped by the counterion/fluoride), which fixes the alkene position unambiguously.
• Result. One clean regioisomer of the enone instead of the 1:1 mixtures a plain divinyl ketone would give. This trick turned the Nazarov from a temperamental cyclization into a dependable synthetic tool.

Where it shows up

• Terpenoid and steroid synthesis. The cyclopentanone ring is everywhere in natural products — the D-ring of steroids, the cores of many sesquiterpenes and prostaglandins. Nazarov cyclizations (and their interrupted variants) have been used in syntheses of guanacastepene, roseophilin, silphinene, and numerous cyclopentanoid terpenes.
• Prostaglandin-type building blocks. 2-Cyclopentenones with defined substitution are the workhorse intermediates for prostaglandins; the conrotatory stereocontrol lets chemists set the ring stereochemistry early.
• Iso-Nazarov and aza/oxa variants. Replacing a vinyl carbon with nitrogen or oxygen, or the ketone with an imine, gives access to cyclopentene-fused heterocycles and pyrrolinones by the same 4π logic.
• Complexity in one pot (interrupted Nazarov). Trapping the oxyallyl cation with an arene, halide, enol ether, or a tethered alkene builds a second ring or installs a substituent immediately — Tius' allene-based and West's cation-trapping Nazarov variants are staples of modern fragment assembly.

Limitations and side reactions

• Regiochemical scrambling. The oxyallyl cation's two possible eliminations give enone mixtures unless the substrate is polarized or silicon-directed. This is the classic reason "textbook" Nazarov reactions look messy.
• Slow, forcing conditions on unactivated dienones. A plain, symmetrical divinyl ketone lacks the cation stabilization to cyclize easily; you need strong acid, heat, or long times, which invite polymerization and decomposition.
• Wagner-Meerwein and retro-Nazarov pathways. The intermediate cations can undergo 1,2-shifts, [1,2]- and [1,4]-hydride migrations, or reopen (retro-electrocyclization), eroding yield and stereochemistry.
• Competing Friedel-Crafts and cation trapping. If an arene or nucleophile is present (even adventitiously), the electrophilic oxyallyl cation can be intercepted before elimination — a bug in a simple Nazarov, but the very feature exploited in the interrupted Nazarov.
• Acid-sensitive functionality. Strong Lewis/Brønsted acids are incompatible with acetals, some protecting groups, and base-labile stereocenters; mild triflate catalysts and polarized substrates mitigate this.

Discovery and the orbital-symmetry epilogue

Ivan Nikolaevich Nazarov (1906-1957), a chemist at the Zelinsky Institute of Organic Chemistry in Moscow, discovered the reaction in the 1940s (his key papers span 1941-1949). He was mining the rich chemistry of acetylene-derived divinyl and allyl vinyl ketones, and he noticed that these cross-conjugated dienones cyclized to cyclopentanones under acidic conditions. Nazarov correctly identified the products but had no way to explain why a five-membered ring formed with the observed stereochemistry — the concept of orbital symmetry did not yet exist.

The explanation arrived two decades later. In 1965 Robert Burns Woodward and Roald Hoffmann published the rules of orbital-symmetry conservation, recasting the Nazarov as a textbook 4π-electron electrocyclization. Because the pentadienyl cation carries four π electrons and the reaction is thermal, the Woodward-Hoffmann rules demand a conrotatory closure — precisely the stereochemistry Nazarov had observed empirically. The reaction thus became one of the cleanest real-world illustrations of the Woodward-Hoffmann rules, and interest in it revived sharply once its mechanism was understood.

Practical and safety notes

• Handle Lewis acids dry. BF₃·OEt₂, TiCl₄, and SnCl₄ fume and hydrolyze violently, releasing HF or HCl. Work under inert atmosphere with dry solvents; quench slowly into cold aqueous bicarbonate.
• Divinyl ketones can polymerize. Cross-conjugated enones are Michael acceptors and prone to acid- or base-induced oligomerization; store cold, use promptly, and add acid slowly at low temperature.
• Low-temperature control pays off. Running at −78 °C not only improves enantioselectivity with chiral catalysts but also suppresses cation rearrangements and over-reaction.
• Catalyst loading. Triflate Lewis acids are often genuinely catalytic (a few mol%), but strong halide Lewis acids and the product's oxygen can sequester the acid, so classic protocols use near-stoichiometric loadings — budget accordingly.