Baeyer-Villiger Oxidation

Why Baeyer-Villiger matters

• Unique C-C oxygen insertion. No other common reaction inserts O between a carbonyl carbon and an alpha-carbon predictably. Hooker/KMnO4 cleaves into fragments; ozonolysis works on alkenes only; Dakin works only on aryl aldehydes with electron-donating groups. BV fills the niche of stereospecific carbon-skeleton oxidation.
• Retention of stereochemistry. The migrating carbon never breaks free of its substituents during the concerted shift, so a stereogenic center stays intact. Mislow's 1957 study of (+)-3-methylcyclohexanone established this — enantiopure ketones from the chiral pool become enantiopure lactones in one step, no racemization.
• Caprolactone industrial process. Cyclohexanone + peracetic acid → ε-caprolactone runs at ~50,000 ton/yr globally as the upstream monomer for polycaprolactone (PCL) — biodegradable medical sutures, drug-delivery matrices, and 3D-printing filament. The Pt-Bi/Si catalytic version (Solvay) avoids the stoichiometric peracid.
• Predictable regiochemistry. Aptitude order H > 3° > 2° ≈ Ar > 1° > Me lets a chemist predict which oxygen-insertion product will dominate before running the reaction. PhCOCH3 always gives phenyl acetate (Ph migrates); cyclohexyl methyl ketone gives the cyclohexyl ester (cyclohexyl migrates).
• Asymmetric BVMO biocatalysts. Cyclohexanone monooxygenase (CHMO) from Acinetobacter calcoaceticus and ~100 other Baeyer-Villiger monooxygenases use NADPH + O2 to do BV with up to >99% ee. Givaudan and Firmenich produce chiral lactone fragrances (γ-decalactone, δ-decalactone) at multi-tonne scale this way.
• Strained ketones react fastest. Cyclobutanones (BV ~10⁴× faster than cyclohexanones) and norbornanones expand to γ-butyrolactones in seconds at 0 °C with mCPBA. This is exploited in ring-expansion strategies — make a small-ring ketone, run BV, get a medium-ring lactone with retention of the substitution pattern.
• Functional-group orthogonality. Esters, amides, ethers, and most heterocycles tolerate mCPBA at 0 °C. The main competitors are alkenes (form epoxides), thioethers (form sulfoxides), and electron-rich arenes (over-oxidize). Buffered NaHCO3 conditions suppress these by keeping the medium near pH 7.

Common misconceptions

• "mCPBA epoxidizes" so it can't BV. Both reactions happen — selectivity depends on relative rates. For substrates with both a ketone and an alkene, the alkene usually wins (epoxidation is faster), so a chemoselective BV requires either no alkenes, electron-poor alkenes, or a substrate where steric/electronic factors favor the carbonyl.
• Methyl groups can migrate if they're the only option. Acetone with mCPBA gives only methyl acetate after long reaction times — methyl can migrate, but the rate is so low that competing peroxide decomposition often dominates. Methyl migration is the slowest in the aptitude scale and rarely useful synthetically.
• The peroxy acid attacks the carbonyl oxygen. No — the terminal oxygen of the peroxy acid (the OH side, deprotonated to OO⁻) attacks the carbonyl carbon. The original carbonyl oxygen ends up as the ester C=O oxygen of the product, while the inserted oxygen is the new alkyl-O-C(=O) ether-type oxygen.
• Migration is anionic. No — the migrating group develops partial positive charge in the transition state, which is why aptitude tracks cation stability (3° > 2° > 1° > Me). If migration were anionic, methyl would be best (small + good orbital overlap), but it is the worst.
• BV always gives the more substituted ester. The more substituted carbon migrates — this puts the more substituted carbon on the alkyl-O side and leaves the less substituted carbon as the C=O side. So for hexyl methyl ketone, hexyl migrates and the product is hexyl acetate (the alcohol-side carbon is hexyl).
• You need anhydrous conditions. BV tolerates traces of water; in fact, aqueous H2O2 with Lewis-acid catalysts (Sn-zeolite, Pt-Bi) is the green industrial standard. Anhydrous conditions matter only when the substrate or product hydrolyzes.

Mechanism

The Baeyer-Villiger mechanism unfolds in two distinct steps. First, the peroxy acid attacks the ketone in an addition resembling hydration: the peroxy acid's terminal nucleophilic oxygen (the OH end) adds to the carbonyl carbon while the carbonyl oxygen picks up a proton. The product is the tetrahedral Criegee intermediate, R(R')C(OH)(O-O-C(=O)R''), where the new C-O bond is to the peroxide oxygen and the original carbonyl O is now an OH. This step is acid-catalyzed and equilibrium-limited; in practice the intermediate is short-lived and rarely observed except for slow substrates at low temperature.

The second step is the rate-limiting concerted migration. The C-C bond between the carbonyl carbon and one alpha-substituent (R or R') breaks; that alpha-substituent migrates to the adjacent peroxide oxygen; the O-O bond breaks heterolytically; the carboxylate (R''CO2⁻) leaves; and the OH on the carbonyl carbon loses its proton to become the new ester C=O. All five bond changes occur in a single transition state, with antiperiplanar geometry (the migrating C-R bond and the breaking O-O bond aligned at ~180°). This concerted geometry is what enforces retention of stereochemistry at the migrating carbon — the C-R bond breaks and C-O bond forms on the same face.

The migratory-aptitude order (H > 3° > 2° ≈ Ar > 1° > Me) reflects the partial positive charge that develops on the migrating carbon during the shift. Tertiary alkyl groups stabilize this charge through hyperconjugation and inductive donation, so they migrate ~10⁴ to 10⁶ times faster than methyl. Aryl groups migrate competitively with secondary alkyls — electron-rich aryls (p-OMe) faster, electron-poor (p-NO2) slower. Hydrogen migration is fastest of all and converts aldehydes to formate esters or carboxylic acids. The aptitude order, established through hundreds of substrate-comparison studies in the 1950s-1970s, is the chemist's predictive tool.

Migratory aptitude — quantitative comparison

Migrating group	Relative rate (vs methyl = 1)	Cation stability	Notable example	Typical conditions	Product
Hydrogen (H)	~10⁹ (overwhelming)	Hydride is highly stabilized in TS	Aldehyde → formate ester (Dakin-like)	mCPBA, 0 °C; or H2O2/base	HC(=O)OR ester
tert-Butyl / cyclohexyl 3°	~10⁵	Tertiary carbocation strongly stabilized	Adamantanone → adamantanyl ester	mCPBA, 0 °C, 30 min	tert-alkyl ester
Cyclopentyl / sec-butyl 2°	~10³	Secondary carbocation moderately stabilized	Cyclohexanone → ε-caprolactone	mCPBA or peracetic acid, 0–25 °C	2°-alkyl ester / lactone
Phenyl (Ar electron-neutral)	~10² to 10³	Aryl migration via phenonium-like TS	PhCOMe → PhO-C(=O)Me phenyl acetate	mCPBA or CF3CO3H, 0 °C	Aryl ester / lactone
p-Methoxyphenyl (Ar electron-rich)	~10⁴	Resonance stabilizes positive charge	PMP-CO-Me → PMP-O-Ac	mCPBA, 0 °C, fast	Aryl acetate
p-Nitrophenyl (Ar electron-poor)	~10	Resonance destabilizes — competes with methyl	p-O2N-Ph-CO-Me → ambiguous; alkyl wins	Often gives mixture	Mixed regiochemistry
Primary alkyl (n-propyl, n-butyl)	~10	Primary cation poorly stabilized	Hexyl methyl ketone → hexyl acetate	mCPBA, 25 °C, longer	1°-alkyl ester
Methyl (Me)	1 (reference)	Methyl cation least stabilized	Acetone → methyl acetate (slow)	mCPBA, 25 °C, hours	Methyl ester (only when no other choice)

Famous syntheses and applications

• Polycaprolactone industrial process. Solvay and Perstorp produce ε-caprolactone from cyclohexanone + peracetic acid (in situ from H2O2 + AcOH) at ~50,000 ton/yr globally. The lactone is ring-opening polymerized to polycaprolactone (PCL), used in biodegradable sutures, ~$300M/yr drug-eluting stents, and Mendable 3D-printing filament that softens at 60 °C.
• BVMO chiral lactones (Givaudan, Firmenich). γ-Decalactone (peach/coconut aroma) and δ-decalactone (creamy) are produced enzymatically using cyclohexanone monooxygenase variants from microbes. Multi-tonne scale per year for the food/fragrance industry; the enzymatic route gives >98% ee, which the chemical route cannot match without an expensive chiral peracid.
• Estrone steroid synthesis (Schering, Bayer). A regiochemically clean BV ring-expansion of a key cyclopentanone D-ring intermediate to a δ-valerolactone has been used in the industrial estrone synthesis since the 1970s. The retention of stereochemistry is critical — racemization would destroy the steroid's biological activity.
• Bryostatin and pumiliotoxin total syntheses. K. C. Nicolaou, David Evans, and Larry Overman have all used BV for late-stage oxygen insertion in complex polyketide natural products. The pumiliotoxin synthesis uses BV to install the lactone with retention of two stereogenic centers in a single step.
• Caprolactam/nylon-6 hybrid routes. Although the dominant industrial route to caprolactam (~5 million ton/yr) is Beckmann rearrangement of cyclohexanone oxime, several Asian plants use BV-derived caprolactone followed by amination — a cleaner waste profile but less mature catalyst technology. The combined nylon-6 BV-route capacity is ~200,000 ton/yr as of 2024.