The Baeyer-Villiger Oxidation

What the reaction does

Take any ketone — two carbon groups joined by a C=O. The Baeyer-Villiger oxidation slides a single oxygen atom into one of those two C-C bonds, right beside the carbonyl. The carbon-oxygen double bond survives; a new C-O-C linkage appears where a C-C bond used to be. The product is an ester: the same carbonyl, but now one side is an oxygen instead of a carbon.

    R-C(=O)-R'   +   R''-CO-O-OH   ──→   R-C(=O)-O-R'   +   R''-COOH
      ketone          peroxyacid            ester          carboxylic acid

    (an oxygen atom is inserted between the carbonyl C and the R' group)

If the ketone is part of a ring, inserting an oxygen makes the ring one atom bigger and turns it into a lactone (a cyclic ester). Cyclohexanone becomes ε-caprolactone — a seven-membered ring. This one-atom ring expansion, done with retention of any stereochemistry, is why the reaction has stayed in the synthetic toolkit for over a century.

The oxidant is a peroxyacid, RCO-O-OH: a carboxylic acid with one extra oxygen bolted onto the OH. That extra oxygen is the one that ends up in the product; the rest of the peroxyacid leaves as an ordinary carboxylic acid. The whole transformation is redox-neutral for the carbon skeleton — you are formally oxidizing the ketone by one oxygen and reducing the peroxide by one.

The mechanism, arrow by arrow

Two chemical steps, one intermediate. The electron bookkeeping is the whole story, so follow the arrows.

1. Addition — form the Criegee intermediate. The peroxyacid is a weak nucleophile at its terminal (OH) oxygen. That oxygen's lone pair attacks the electrophilic carbonyl carbon of the ketone; simultaneously the C=O π electrons collapse onto oxygen to give an alkoxide (or, under the acidic conditions, a hydroxyl). The carbonyl carbon goes from sp² to sp³. It now carries: an OH, the two original R groups, and an -O-O-C(=O)R'' peroxyester arm. This tetrahedral adduct is the Criegee intermediate (Rudolf Criegee, 1948). A proton transfer neutralizes the charges. This step is usually acid-catalyzed — protonating the ketone oxygen makes the carbon a better electrophile.
2. Migration — the concerted 1,2-shift. The weak O-O bond of the peroxyester (bond energy only ≈ 35-45 kcal/mol) is primed to break. As it breaks heterolytically, a good leaving group departs — a carboxylate, R''COO⁻ — and to fill the electron gap, one of the two R groups migrates from carbon to the neighboring oxygen, C-C bond and C-O bond both moving in a single concerted transition state. The oxygen lone pair on the (former) carbonyl OH pushes in to reform a C=O. The net result: an oxygen atom is now wedged between the carbonyl carbon and the migrated group. You have an ester.

  Step 1 (addition):
        O                         OH
        ‖                          |
    R — C — R'   +  HO-O-C(=O)R'' →  R — C — R'      ← Criegee intermediate
                                          |
                                     O-O-C(=O)R''

  Step 2 (migration, concerted):
        OH                              O
        |      R' migrates onto O       ‖
    R — C — R'  ───────────────────→  R — C — O — R'   +   R''COOH
        |      (O-O bond breaks,          the ester
     O-O-C(=O)R''  carboxylate leaves)

The migration step is rate-determining for most ketones, and it is where all the selectivity lives — which of the two R groups moves decides which ester you get.

Migratory aptitude: which group moves?

Only one of the two groups can migrate, and the choice is not random. In the transition state the migrating carbon takes on partial positive character, so the group that best stabilizes that positive charge moves preferentially. Empirically:

   tertiary  >  cyclohexyl  ≈  secondary  >  benzyl  >  phenyl  >  primary  >  methyl
   (best migrating)  ─────────────────────────────────────────────►  (worst migrating)

Consequences you can predict:

• A methyl ketone R-CO-CH₃ almost always inserts the oxygen on the R side, because any R more substituted than methyl outranks it. R migrates onto oxygen and methyl stays bonded to the carbonyl, so the product is the acetate ester CH₃-CO-O-R (the R group becomes an "R-O" alkoxy). Only if R is itself a poor migrator (another methyl, or a strongly deactivated aryl) does the methyl move instead.
• Acetophenone (Ph-CO-CH₃) gives phenyl acetate (Ph-O-CO-CH₃): phenyl outranks methyl, so phenyl migrates.
• Electron-rich aryl groups migrate faster; a p-methoxyphenyl ketone migrates the aryl cleanly, while a p-nitrophenyl ketone (electron-poor) is sluggish and may migrate the other group instead.
• For an unsymmetrical dialkyl ketone, the more-branched carbon migrates. 3-Methyl-2-butanone (methyl isopropyl ketone) migrates the isopropyl group, inserting oxygen between the isopropyl carbon and the carbonyl.

A subtlety worth knowing: migratory aptitude also depends on conformation. The migrating C-C bond must be antiperiplanar to the breaking O-O bond (and roughly antiperiplanar to a lone pair on the OH oxygen) — the so-called "primary and secondary stereoelectronic effects." When a rigid or cyclic substrate cannot achieve that alignment for the intrinsically better migrator, the electronically-disfavored group can win instead. This is why steroid and terpene Baeyer-Villigers sometimes give the "wrong" regiochemistry.

Stereochemistry: retention at the migrating carbon

Because migration is a concerted 1,2-shift and the migrating carbon never becomes a free carbocation, it keeps its three other bonds intact the whole time. A stereocenter on the migrating group is therefore preserved with full retention of configuration. If you feed in an (R)-configured secondary carbon and it migrates, it comes out (R) in the ester. This is a classic mechanistic probe: chemists ran optically active substrates precisely to confirm the concerted, retentive nature of the shift. It also makes the reaction genuinely useful for chiral synthesis — you can ring-expand a stereodefined ketone into a stereodefined lactone without scrambling.

Reagents, catalysts, and real conditions

Oxidant	Made from / notes	Best for
mCPBA (meta-chloroperoxybenzoic acid)	Shelf-stable solid, ~70-77% assay; used in CH₂Cl₂, 0→25 °C	The default bench reagent; general ketones
Trifluoroperoxyacetic acid (CF₃CO₃H)	Generated in situ from 90% H₂O₂ + (CF₃CO)₂O; very reactive; buffered with Na₂HPO₄	Unreactive / hindered ketones
Peroxyacetic / peroxybenzoic acid	Cheaper, industrial; from acid + H₂O₂	Large-scale, cost-sensitive runs
H₂O₂ + Lewis-acid catalyst	Sn-Beta zeolite, Se or Pt catalysts, methyltrioxorhenium	Green chemistry; water as the only byproduct
H₂O₂ + flavin / O₂ + enzyme	Baeyer-Villiger monooxygenases (BVMOs), NADPH + O₂	Asymmetric, enantioselective ring expansions

• Solvent. Dichloromethane or chloroform for mCPBA; the reaction tolerates most functional groups but competes with alkene epoxidation (see limitations).
• Buffer. The byproduct carboxylic acid (m-chlorobenzoic acid, trifluoroacetic acid) is acidic and can transesterify or open the product lactone. Adding solid NaHCO₃ or a phosphate buffer keeps the medium mild and protects the ester.
• Temperature. Typically 0 °C to room temperature; the O-O bond is thermally sensitive and peroxyacids can decompose exothermically, so scale-up is done cold and dilute.

How it compares to related methods

	Baeyer-Villiger	Beckmann rearrangement	Epoxidation (mCPBA on alkene)
Substrate	Ketone (or aldehyde)	Ketoxime	Alkene
Atom inserted	O (between C and C)	N (between C and C)	O (across C=C)
Product	Ester / lactone	Amide / lactam	Epoxide
Reagent	Peroxyacid (RCO₃H)	Acid (H₂SO₄, PCl₅, etc.)	Peroxyacid (RCO₃H)
Migrating group	More substituted, antiperiplanar to O-O	Group anti to the oxime OH	— (concerted O transfer)
Stereochemistry	Retention at migrating C	Retention at migrating C	Syn addition, retention of alkene geometry
Industrial signature	Caprolactone → polycaprolactone	Cyclohexanone oxime → caprolactam (nylon-6)	Ethylene oxide, propylene oxide feedstocks

The parallel with the Beckmann rearrangement is exact and worth internalizing: Beckmann inserts a nitrogen into a ketoxime to make an amide; Baeyer-Villiger inserts an oxygen into the parent ketone to make an ester. Same 1,2-migration logic, same retention, different heteroatom. Note that mCPBA is a peroxyacid whether it is doing a Baeyer-Villiger or an epoxidation — on a substrate that has both a ketone and an alkene, epoxidation of the alkene is usually faster, and you must protect the double bond if you want to oxidize the carbonyl.

Worked example: cyclohexanone to caprolactone

The textbook case, and a real industrial one. Ring-expand cyclohexanone to ε-caprolactone.

   cyclohexanone  +  mCPBA  ──CH₂Cl₂, NaHCO₃, 0→25 °C, 4 h──→  ε-caprolactone  +  m-ClC₆H₄COOH

• Substrate. Cyclohexanone (1.0 equiv). It is symmetric, so both ring carbons are equivalent — no regiochemistry question, no migratory-aptitude tie to break.
• Oxidant. mCPBA 1.1-1.3 equiv, added portionwise at 0 °C.
• Buffer. Solid NaHCO₃ to neutralize the m-chlorobenzoic acid byproduct and stop it from opening the lactone.
• Mechanism recap. The peroxyacid oxygen adds to the ring carbonyl → Criegee intermediate → one ring C-C bond migrates onto oxygen as m-chlorobenzoate leaves → seven-membered ring lactone.
• Workup. Filter off the m-chlorobenzoic acid, wash with aqueous NaHCO₃ and Na₂SO₃ (to destroy excess peroxide — a safety must), distill.
• Yields. 70-90% caprolactone on the bench. Industrially, peroxyacetic acid is used in continuous processes; caprolactone feeds polycaprolactone (a biodegradable polyester) and can be a route toward caprolactam for nylon-6.

Limitations & side reactions

• Alkenes get epoxidized first. mCPBA is the standard epoxidation reagent too. On an enone or any substrate with an isolated C=C, the double bond usually reacts before the carbonyl. Protect the alkene, or exploit the chemoselectivity deliberately.
• Over-oxidation and Dakin competition on aryl aldehydes. With electron-rich (o-/p-hydroxy or -amino) aryl aldehydes, the aryl migrates and the formate hydrolyzes to a phenol — the Dakin reaction. With ordinary aldehydes, H migrates and you just get the carboxylic acid.
• Acid-sensitive products. The carboxylic-acid byproduct can transesterify, epimerize α-stereocenters, or open lactones. Buffering is essential for sensitive substrates.
• Peroxide hazards. Concentrated H₂O₂ and peroxyacids are shock- and heat-sensitive oxidizers. Trifluoroperoxyacetic acid made from >90% H₂O₂ demands strict temperature control; never distill peroxide residues to dryness. Quench excess oxidant with sulfite before workup.
• Slow on electron-poor ketones. A ketone flanked by electron-withdrawing groups (an α,α-dihalo ketone, a strongly deactivated aryl ketone) has a less nucleophilic carbonyl and a poor migrator; it may need the more powerful trifluoroperoxyacetic acid or fail outright.

Discovery and history

Adolf von Baeyer — the same Baeyer who won the 1905 Nobel Prize for synthesizing indigo, and after whom Baeyer strain theory is named — and his co-worker Victor Villiger reported the reaction in 1899. They treated menthone and other cyclic ketones with Caro's acid (peroxymonosulfuric acid, H₂SO₅) and found, to their surprise, that an oxygen had inserted to give a lactone rather than a simple diol or diketone. They correctly guessed an oxygen-insertion but proposed a dioxirane-like pathway.

The mechanism stayed contested for nearly fifty years. Rudolf Criegee, in 1948, showed by studying the peroxybenzoate of trans-decalin-derived ketones — and by tracking stereochemistry — that the reaction proceeds through the tetrahedral peroxyester adduct that now bears his name, followed by a concerted migration. Later ¹⁸O-labelling experiments (Doering and Dorfman, 1953) nailed down which oxygen ends up where: the carbonyl oxygen of the product comes from the original ketone, and the newly inserted ester oxygen comes from the peroxyacid — exactly what the Criegee mechanism predicts.

Industrial and asymmetric notes

• Caprolactone at scale. Peroxyacetic-acid oxidation of cyclohexanone is run industrially to make ε-caprolactone, the monomer for polycaprolactone (a biodegradable, low-melting polyester used in medical sutures, 3D-printing filament, and drug-delivery matrices).
• Steroid and terpene chemistry. Baeyer-Villiger ring expansions and oxygen insertions are staples in total synthesis, converting cyclic ketones in polycyclic frameworks into lactones with predictable regiochemistry set by conformation.
• Enzymatic, asymmetric Baeyer-Villiger. Baeyer-Villiger monooxygenases (BVMOs), such as cyclohexanone monooxygenase (CHMO), use FAD, NADPH, and molecular O₂ to insert oxygen enantioselectively. They desymmetrize prochiral and racemic ketones — a green, water-based route to chiral lactones that peroxyacids cannot match on selectivity.
• Greener oxidants. Tin-Beta zeolite with aqueous H₂O₂ (Corma and co-workers) performs the Baeyer-Villiger with water as the only byproduct and is chemoselective for the carbonyl over alkenes — the reverse of mCPBA's selectivity — making it valuable for enone substrates.