The Hofmann Rearrangement

What the Hofmann rearrangement does

Take a primary amide — a molecule ending in –C(=O)–NH₂. Treat it with bromine and a strong base, and it comes out the other side as a primary amine –NH₂ that has lost the carbonyl carbon entirely. A butanamide (4 carbons) becomes propylamine (3 carbons). Benzamide (7 carbons) becomes aniline (6 carbons). The missing carbon leaves as carbon dioxide.

The overall transformation, balanced:

    R-C(=O)-NH₂  +  Br₂  +  4 NaOH  ──→  R-NH₂  +  2 NaBr  +  Na₂CO₃  +  2 H₂O

The reason chemists reach for it: it is one of the very few clean ways to remove exactly one carbon while turning a carboxylic-acid derivative into an amine. And because the carbon skeleton migrates as an intact group rather than detaching, a stereocenter next to the carbonyl survives the trip with its configuration untouched — something almost no other amine synthesis can promise.

The mechanism, arrow by arrow

Five distinct events, and the whole reaction turns on step 4, where the skeleton rearranges. Follow the nitrogen and the migrating group R.

1. Deprotonation of the amide. Hydroxide removes one N–H proton (amide N–H, pK_a ≈ 17–20) to give a resonance-stabilized amidate anion, R-C(=O)-NH⁻. The lone pair on nitrogen is now nucleophilic.
2. N-bromination. That nitrogen lone pair attacks Br₂, displacing bromide, to give an N-bromoamide, R-C(=O)-NHBr. The nitrogen now bears a good leaving group.
3. Second deprotonation. The remaining N–H is far more acidic now (the adjacent electron-withdrawing bromine and carbonyl), so hydroxide plucks it off, giving the N-bromoamide anion (a bromamide anion), R-C(=O)-N⁻-Br.
4. Concerted migration — the rearrangement. This is the heart of the reaction. As bromide departs from nitrogen (making nitrogen electron-deficient), the C–R bond does not wait: the R group migrates from the carbonyl carbon directly onto the nitrogen in the same step. Bromide leaving and R migrating are concerted — a free acylnitrene is never released. The product is an isocyanate, R-N=C=O. Because R never fully detaches, the migrating carbon keeps its stereochemistry (retention of configuration).
5. Hydrolysis and decarboxylation. Hydroxide adds to the electrophilic central carbon of the isocyanate to give a carbamate/carbamic acid, R-NH-C(=O)-O⁻. Carbamic acids are unstable: the C–N bond stays, the C–O carbon leaves as CO₂ (captured by base as carbonate), and the amine R-NH₂ is released.

  step 1:  R-C(=O)-NH₂        + OH⁻  →  R-C(=O)-NH⁻            (amidate anion)
  step 2:  R-C(=O)-NH⁻        + Br₂  →  R-C(=O)-NHBr  + Br⁻    (N-bromoamide)
  step 3:  R-C(=O)-NHBr       + OH⁻  →  R-C(=O)-N⁻-Br          (bromamide anion)
  step 4:  R-C(=O)-N⁻-Br             →  R-N=C=O       + Br⁻    (CONCERTED migration → isocyanate)
  step 5:  R-N=C=O  + OH⁻/H₂O        →  R-NH-COO⁻  →  R-NH₂ + CO₂  (hydrolysis + decarboxylation)

Notice where the carbons go. The carbonyl carbon (the "C" of C=O) becomes the central carbon of the isocyanate and then leaves as CO₂. The R group's carbon — the one bonded to nitrogen in the product — is the carbon that migrated. The chain is genuinely one carbon shorter.

Reagents, conditions, and why each one is there

• Bromine (Br₂), 1 equivalent. The source of the electrophilic bromine that ends up on nitrogen. Chlorine works too (Hofmann's original work used both); the mechanism is identical with an N-chloroamide.
• Strong base — NaOH or KOH, ~4 equivalents. The base does triple duty: it deprotonates the amide twice (steps 1 and 3), it drives the concerted elimination in step 4, and it hydrolyzes the isocyanate and neutralizes the CO₂ and HBr byproducts. Under-basing stalls the reaction at the N-bromoamide.
• Water, warm (often 60–80 °C). Aqueous conditions are standard; gentle heating pushes the rearrangement and hydrolysis. For hindered or acid-sensitive substrates the isocyanate can be intercepted before hydrolysis (see below).
• Sometimes an alcohol instead of water. If you run the reaction in methanol/sodium methoxide, the isocyanate is trapped by methoxide to give a stable methyl carbamate (R-NH-CO-OCH₃) rather than the free amine — a Hofmann variant used when you want the protected amine.

Modern hypervalent-iodine variants. The classic Br₂/NaOH conditions are harsh (strongly basic, oxidizing). The most-used modern replacement uses [bis(trifluoroacetoxy)iodo]benzene (PIFA) or (diacetoxyiodo)benzene (PIDA) in water/acetonitrile at room temperature. These deliver the electrophilic "N–X" character without free bromine, run near neutral pH, and are the go-to when the substrate carries base-sensitive or oxidizable groups. NBS (with base or DBU) and lead tetraacetate are older alternatives.

Scope, selectivity, and stereochemistry

Three facts govern how this reaction behaves:

• Only primary amides work. You need the –C(=O)–NH₂ group with two N–H hydrogens. Secondary and tertiary amides have no N–H to brominate and deprotonate a second time — they simply don't undergo the rearrangement.
• Retention of configuration. Because R migrates without ever fully separating from the framework (the migration is concerted with bromide loss), a stereocenter within R that is directly attached to the migrating carbon keeps its absolute configuration. Optically active amides give optically active amines. This is the cleanest experimental proof that no symmetric, freely-rotating nitrene intermediate exists.
• Migratory aptitude. The transition state builds partial positive charge on the migrating carbon, so groups that stabilize that positive character migrate faster: electron-rich aryls and more-substituted alkyls are good; strongly electron-poor aryls (e.g. p-nitrophenyl) migrate more slowly. Since there is only one group on a primary amide, aptitude sets the rate, not a product ratio.

A practical limit: with very large substrates and forcing basic conditions, side reactions (hydrolysis of the amide to the carboxylic acid, or over-oxidation) can eat into yield — which is exactly why the room-temperature PIFA variant took over for delicate molecules.

Hofmann vs the other 1,2-nitrogen rearrangements

Hofmann, Curtius, Schmidt, and Lossen all funnel through the same isocyanate and all give an amine one carbon shorter. They differ only in how they generate the electron-deficient nitrogen. The Beckmann is the odd one out — it makes an amide, not an amine.

Reaction	Starts from	Leaving group	Passes through isocyanate?	Product
Hofmann	Primary amide R-CO-NH₂	Br⁻ (from N-bromoamide)	Yes	Amine R-NH₂ (–1 C)
Curtius	Acyl azide R-CO-N₃	N₂	Yes	Amine R-NH₂ (–1 C)
Schmidt	Carboxylic acid + HN₃	N₂	Yes	Amine R-NH₂ (–1 C)
Lossen	Activated hydroxamic acid R-CO-NH-OX	Carboxylate ⁻OX	Yes	Amine R-NH₂ (–1 C)
Beckmann	Ketoxime R₂C=N-OH	H₂O (from N-OH)	No (nitrilium)	Amide (same C count)

If you already have the amide in hand, Hofmann is the cheapest of the four. Curtius is chosen when strong base is intolerable (azides are dangerous but base-free). Schmidt collapses acid → amine in one pot but uses hazardous hydrazoic acid.

Worked example: benzamide → aniline, and phthalimide → anthranilic acid

Textbook case. Benzamide gives aniline, one carbon shorter (the carbonyl carbon leaves as CO₂):

    C₆H₅-C(=O)-NH₂  +  Br₂  +  4 NaOH  ──→  C₆H₅-NH₂  +  2 NaBr  +  Na₂CO₃  +  2 H₂O
       benzamide (7 C)                          aniline (6 C)

Intermediate you can actually stop at: phenyl isocyanate, C₆H₅–N=C=O, the same intermediate used industrially to make phenyl-carbamate pesticides.

The classic large-scale case: anthranilic acid. Start from phthalimide, open it with base to phthalamic acid (a benzene ring carrying an amide on one carbon and a carboxylate on the adjacent carbon), then run the Hofmann on the amide. The amide nitrogen becomes an –NH₂, the carbonyl carbon leaves as CO₂, and you are left with 2-aminobenzoic acid — anthranilic acid:

    phthalimide  ──NaOH/H₂O──→  phthalamic acid  ──Br₂, NaOH (Hofmann)──→  anthranilic acid
    (C₆H₄(CO)₂NH)                (o-HOOC-C₆H₄-CONH₂)                        (o-H₂N-C₆H₄-COOH)

Anthranilic acid is a genuine commodity: it is the precursor to saccharin, to indigo and azo dyes, to the fragrance methyl anthranilate (grape flavour), and to drugs such as the fenamate anti-inflammatories. The Hofmann degradation of phthalimide was for decades the standard industrial route to it.

Limitations and side reactions

• Amide hydrolysis competes. Hot, strongly aqueous base can simply hydrolyze the primary amide back to the carboxylic acid and ammonia before it rearranges. Keeping bromination fast (good Br₂ dispersion, cold addition) and using the milder PIFA route both suppress this.
• Only primary amides. No N–H₂, no reaction — the second deprotonation and the leaving-group installation both need those hydrogens.
• Sensitive functional groups. Br₂/NaOH will brominate electron-rich arenes and phenols, oxidize thiols, and cleave some protecting groups. Choose the hypervalent-iodine variant for such substrates.
• Isocyanate side-trapping. If a nucleophile (an amine, an alcohol) is present in solution, it can intercept the isocyanate to give a urea or carbamate instead of the free amine — sometimes a nuisance, sometimes exactly what you want.
• Scale and safety. Elemental bromine is corrosive and toxic; the reaction evolves CO₂; and N-haloamides can be shock-sensitive if isolated. Industry keeps them in solution and never dry.

Discovery: August Wilhelm von Hofmann, 1881

The reaction is named for August Wilhelm von Hofmann (1818–1892), the German chemist who also gave his name to the Hofmann elimination (of quaternary ammonium salts — a different reaction) and to the Hofmann voltameter. In 1881 he reported that treating amides with bromine and alkali produced amines with one fewer carbon, and worked out that the reaction ran through an isocyanate. It is sometimes called the Hofmann degradation or Hofmann bromamide reaction.

The mechanistic detail that no free nitrene forms — that migration is concerted with loss of the leaving group — came later, from stereochemical studies showing retention of configuration at the migrating carbon. That same concerted-migration insight ties Hofmann to the Curtius (1890), Lossen, and Schmidt (1923) reactions, which all share the isocyanate intermediate.

Industrial and modern notes

• Anthranilic acid manufacture. As above, the Hofmann degradation of phthalimide was the workhorse industrial route for over a century, feeding the dye and saccharin industries. Some modern plants have shifted to catalytic ammoxidation routes, but Hofmann chemistry remains textbook and lab-standard.
• Carbamate agrochemicals. The isocyanate intermediate is valuable in its own right. Trapping R–N=C=O with an alcohol gives a carbamate; several carbamate insecticides and herbicides trace to isocyanate chemistry closely related to the Hofmann intermediate.
• Chain-shortening in total synthesis. When a target needs an amine exactly one carbon down from an available acid, and the stereocenter must survive, the Hofmann (or its base-free cousin, the Curtius) is the standard disconnection — you make the amide, then degrade.
• Green chemistry. The PIFA/PIDA hypervalent-iodine variants replaced elemental bromine and molar excesses of hot NaOH with room-temperature, near-neutral, water-compatible conditions — the main reason the reaction is still taught as practical rather than merely historical.