The Schmidt Reaction

What the Schmidt reaction does

The Schmidt reaction is a family of acid-catalyzed reactions that all share one trick: hydrazoic acid, HN₃, delivers a nitrogen atom that ends up wedged into the carbon skeleton, and a molecule of N₂ leaves to pay for it. Which product you get depends on the carbonyl you feed it:

• Ketone → amide. A nitrogen inserts between the carbonyl carbon and one of its two neighbors: R-CO-R′ becomes R-CO-NH-R′. This is the nitrogen twin of the Baeyer-Villiger oxidation, which slips an oxygen into the same bond.
• Carboxylic acid → primary amine. R-COOH becomes R-NH₂, and the old carboxyl carbon leaves as CO₂. The amine has one fewer carbon than the acid — a net degradation, mechanistically a cousin of the Curtius and Hofmann rearrangements.
• Aldehyde → nitrile (mostly). R-CHO becomes R-CN, with some N-formyl amine as a minor product.

The common engine underneath all three is a 1,2-migration onto an electron-deficient nitrogen, with N₂ as the world's best leaving group. Get that migration step and you understand the whole family.

The mechanism, arrow by arrow (ketone case)

Take a ketone R-CO-R′. Here is the electron flow, step by step:

1. Protonate the carbonyl. Concentrated H₂SO₄ protonates the carbonyl oxygen, turning the carbon into a much stronger electrophile (a resonance-stabilized oxocarbenium).
2. Azide adds. The terminal nitrogen of HN₃ (the nucleophilic end) attacks the carbonyl carbon. After a proton shuffle you have a tetrahedral azidohydrin: R-C(OH)(N₃)-R′.
3. Lose water. Protonation of the -OH and loss of water gives an iminodiazonium ion — a C=N bearing a diazonium (-N₂⁺) leaving group (i.e. R-C(=N-N₂⁺)-R′, a monocation). This diazonium nitrogen is a spectacular leaving group.
4. Concerted migration + N₂ loss. The R group that sits anti-periplanar to the departing N₂ migrates from carbon to the adjacent nitrogen at the same moment N₂ leaves. This is the committed, product-determining step. Migration happens with retention of configuration at the migrating carbon — the group never becomes a free carbocation.
5. Nitrilium → amide. Migration produces a nitrilium ion, R′-C≡N⁺-R (R has moved onto nitrogen, R′ stays on carbon; or its resonance form). Water attacks the electrophilic carbon, and after tautomerization you get the neutral amide, R′-CO-NH-R.

   R-CO-R'  +  HN₃   ──H₂SO₄──→   R-CO-NH-R'   +   N₂

 1) R-C(=OH⁺)-R'                         (protonated ketone)
 2) + HN₃  →  R-C(OH)(N₃)-R'             (azidohydrin)
 3) -H₂O   →  R-C(=N–N₂⁺)-R'            (iminodiazonium, N₂ ready to leave)
 4)  R migrates ANTI to N₂, N₂ leaves:
              R'-C≡N⁺-R    +   N₂↑        (nitrilium ion)
 5) + H₂O, tautomerize  →  R'-CO-NH-R      (amide)

For a carboxylic acid, steps diverge after the azide adds. The protonated acid is attacked by HN₃ to give, after loss of water, an acyl azide R-CO-N₃. Loss of N₂ with migration of R gives an isocyanate R-N=C=O — exactly the intermediate of the Curtius rearrangement. Water then hydrolyzes the isocyanate to a carbamic acid R-NH-COOH, which spontaneously loses CO₂ to hand you the amine R-NH₂.

   R-COOH  +  HN₃  ──H₂SO₄──→  R-NH₂  +  CO₂  +  N₂

   R-COOH → R-CO-N₃ (acyl azide) → [ -N₂, R migrates ] → R-N=C=O (isocyanate)
          → + H₂O → R-NH-COOH (carbamic acid) → -CO₂ → R-NH₂

Reagents, catalyst, and conditions

• Nitrogen source. Hydrazoic acid, HN₃. It is almost never bottled and poured; instead it is generated in situ from sodium azide (NaN₃) and the strong acid already in the flask.
• Acid / catalyst. Concentrated sulfuric acid is classic. Other Brønsted or Lewis acids work — trifluoroacetic acid, methanesulfonic acid, TfOH, or TiCl₄/BF₃·OEt₂ for milder alkyl-azide variants. The acid does double duty: it protonates the carbonyl and it liberates HN₃ from azide.
• Temperature. Typically 0 °C to ~40 °C. Ketones often react between room temperature and 50 °C; sluggish substrates may need warming. Keep it cool — HN₃ boils at 37 °C and is explosive.
• Solvent. The concentrated acid can be the medium; chloroform, dichloromethane, or benzene are common co-solvents. Water is the ultimate quench and the source of the amide/amine oxygen back-and-forth.
• Stoichiometry. Roughly 1-1.5 equivalents of azide per carbonyl for ketones; acids and the intramolecular variants have their own optima.

Scope, selectivity, and stereochemistry

The Schmidt reaction is prized for going straight from an ordinary carbonyl to a nitrogen-containing product in one pot, but selectivity is its Achilles' heel:

• Regioselectivity (which group migrates). The group anti to the departing N₂ migrates. Because the iminodiazonium can rotate into either geometry, the more migratory-apt group — usually the larger or more electron-rich one, which prefers the less crowded anti position — tends to migrate. Aryl and tertiary-alkyl groups migrate cleanly; two similar alkyl groups give a mixture of the two regioisomeric amides.
• Stereochemistry. Migration is intramolecular and concerted, so the migrating carbon keeps its configuration (retention). A chiral migrating group arrives at nitrogen with its stereocenter intact.
• Cyclic ketones ring-expand. A cyclic ketone inserts nitrogen into the ring, giving a lactam one atom larger. Cyclohexanone → the seven-membered lactam ε-caprolactam is the textbook example (and the industrial target of the rival Beckmann route).
• Functional-group tolerance. The strongly acidic, oxidizing HN₃/H₂SO₄ medium is harsh: acid-sensitive groups, alkenes prone to protonation, and easily hydrolyzed protecting groups often don't survive.

Schmidt vs the rearrangements it competes with

	Schmidt	Beckmann	Curtius	Baeyer-Villiger
Starting material	Ketone / acid / aldehyde	Oxime (from ketone)	Acyl azide (from acid)	Ketone
Reagent	HN₃ + H₂SO₄	Acid / PCl₅ / TsCl	Heat (Δ) or hν	Peracid (mCPBA)
Atom inserted	N	N	N	O
Ketone product	Amide	Amide	—	Ester
Acid product	Amine + CO₂	—	Amine (via isocyanate)	—
Leaving group	N₂	H₂O (from N-OH)	N₂	Carboxylate
Migration geometry	Anti to N₂	Anti to leaving OH	Anti to N₂	Anti to O-O
Migratory aptitude	More substituted / aryl	Anti group (set by oxime geometry)	Any (single group)	More substituted / aryl
Extra prep step?	No — direct from carbonyl	Yes — must make oxime first	Yes — must make azide first	No
Main hazard	Explosive, toxic HN₃	Corrosive acid	Explosive acyl azides	Peracid / diacyl peroxide

Worked example: benzoic acid → aniline

The acid version is the cleanest demonstration because there is only one group to migrate. Treat benzoic acid with sodium azide in concentrated sulfuric acid:

   Ph-COOH  +  NaN₃  ──conc. H₂SO₄, CHCl₃, 40 °C──→  Ph-NH₂  +  CO₂↑ + N₂↑

• Reagents. Benzoic acid 1.0 equiv, NaN₃ ~1.2 equiv (this generates HN₃ against the sulfuric acid), concentrated H₂SO₄ as the medium, chloroform as co-solvent.
• Mechanistic path. Protonated acid → phenyl acyl azide Ph-CO-N₃ → loss of N₂ with phenyl migration → phenyl isocyanate Ph-N=C=O → hydrolysis to N-phenylcarbamic acid → loss of CO₂.
• Product. Aniline, Ph-NH₂. Note the phenyl ring migrated intact — an aryl group is highly migration-apt — and the carboxyl carbon departed as CO₂, so the product is a C₆ amine from a C₇ acid.
• Compare. Reduce benzoic acid to benzyl alcohol and you keep all seven carbons; run the Schmidt and you trade the carboxyl carbon for a nitrogen. That carbon-losing degradation is exactly the point.

Run the same chemistry on a ketone — say acetophenone, Ph-CO-CH₃ — and the phenyl (more migratory-apt) tends to migrate to nitrogen, giving predominantly acetanilide, Ph-NH-CO-CH₃, plus N₂.

Real-world applications

• Ring-expanded lactams. Cyclic ketones give one-carbon-larger lactams in a single step. Cyclohexanone → ε-caprolactam (the nylon-6 monomer) is the showcase, though the shock hazard of HN₃ keeps the tonnage-scale nylon route on Beckmann chemistry instead.
• Total synthesis of alkaloids. The intramolecular Schmidt reaction — using a tethered alkyl azide instead of free HN₃ — is a workhorse in complex-molecule synthesis. Jeffrey Aubé developed azido-ketone cyclizations (with Lewis acids like TiCl₄ or BF₃) that build bridged and fused lactams found in Stemona and Aspidosperma alkaloids and in the dendrobatid (poison-frog) alkaloids such as indolizidine 251F.
• One-step aryl nitriles. Aromatic aldehydes give aryl nitriles directly, a handy shortcut to Ar-CN without a separate dehydration.
• Primary amines by chain shortening. When you specifically want an amine one carbon shorter than an available acid — and can tolerate the hazards — the Schmidt does it in one operation, competing with the Curtius and Hofmann degradations.
• Analytical/degradation tool. Historically used to interconvert acids, ketones, and their nitrogen analogues for structure determination.

Limitations and side reactions

• Regiochemical mixtures. Unsymmetrical dialkyl ketones with two similar groups give both amide regioisomers, because neither group strongly wins the anti competition. Plan around this or pick a substrate where one group is clearly more migratory-apt.
• Tetrazole formation. With excess HN₃ the nitrilium intermediate can be trapped by a second azide before water gets there, cyclizing to a 1,5-disubstituted tetrazole instead of the amide. This is a genuine competing pathway, not a rare curiosity.
• Over-strong acid side reactions. The concentrated H₂SO₄ medium can sulfonate arenes, isomerize alkenes, or cleave acid-labile groups.
• Diazo/nitrogen extrusion hazards. Every run liberates N₂ (and, from the reagent, potentially HN₃ vapor) — pressure and toxicity to manage.
• Sterically hindered or electron-poor carbonyls can be sluggish, needing more forcing (and more dangerous) conditions.

Who discovered it, and when

The reaction is named for Karl Friedrich Schmidt (1887-1971), a German chemist who reported it in the early 1920s (his key papers appeared in 1923-1924). Schmidt found that hydrazoic acid reacts with carbonyl compounds under acid catalysis to insert nitrogen — the ketone-to-amide and acid-to-amine transformations that carry his name. The reaction sits in a cluster of early-twentieth-century nitrogen-migration rearrangements — Hofmann (1881), Curtius (1890), Lossen, and Beckmann (1886) — that together let chemists shuttle between acids, amides, amines, and isocyanates. Its modern renaissance came in the 1990s when Jeffrey Aubé and others developed the intramolecular alkyl-azide version, turning a hazardous curiosity into a precision tool for building complex nitrogen-containing ring systems.

Safety and handling notes

• Hydrazoic acid is a serious hazard. HN₃ is volatile (b.p. 37 °C), acutely toxic (comparable to HCN — it inhibits cytochrome c oxidase), and shock-, friction-, and heat-sensitive enough to detonate. Heavy-metal azides (Cu, Ag, Pb from contact with metal spatulas or drains) are primary explosives.
• Generate in situ, never store neat. Standard practice is to add sodium azide to the acidic reaction mixture so HN₃ forms and reacts immediately, minimizing free HN₃ inventory. Keep temperatures below its boiling point.
• Manage the gas. N₂ evolution is real and continuous; use adequate venting and a blast shield. Never seal the vessel.
• Prefer the intramolecular variant when possible. Tethered alkyl azides with a mild Lewis acid (the Aubé conditions) avoid free HN₃ entirely and are the safer, more selective modern choice for synthesis.
• Why industry uses Beckmann instead. For the caprolactam that becomes nylon-6, the safer (if oxime-requiring) Beckmann rearrangement wins on scale precisely because it sidesteps hydrazoic acid.