The Combes Quinoline Synthesis

What the Combes synthesis does

The Combes synthesis is a two-move way to punch a fresh pyridine ring onto an existing benzene ring, giving you a quinoline — benzene and pyridine fused at a shared edge. You start with a plain aniline (any primary aromatic amine) and a 1,3-diketone, most classically pentane-2,4-dione (acetylacetone). The two pieces snap together in a specific order:

1. Condense. The aniline nitrogen attacks one of the two carbonyls, and the pair loses a molecule of water. Instead of a simple imine, the 1,3-dicarbonyl gives a conjugated β-enaminone (a β-aminoenone) — the nitrogen ends up doubly stabilized by the still-intact second carbonyl. This step is easy and often needs no catalyst.
2. Cyclize. Strong acid protonates the remaining ketone, turning it into a hot electrophile. The aniline's own benzene ring reaches around and attacks that carbon at its ortho position — an intramolecular electrophilic aromatic substitution. A final loss of water rearomatizes everything, and out comes the quinoline.

The bookkeeping is worth memorizing because it tells you exactly which product you get. From a symmetric 1,3-diketone, the two carbonyl carbons become C-2 and C-4 of the new ring, the central CH₂ of the diketone becomes C-3, and the aniline nitrogen becomes N-1. Acetylacetone (two methyls flanking the CH₂) therefore delivers 2,4-dimethylquinoline.

The mechanism, arrow by arrow

Take the canonical case: aniline + pentane-2,4-dione. Watch the electrons.

Stage 1 — enaminone formation (no acid needed).

1. The lone pair on the aniline nitrogen attacks one carbonyl carbon of the diketone, forming a tetrahedral carbinolamine (hemiaminal).
2. Proton transfer and loss of water from that carbon would normally give an imine — but the neighbouring second carbonyl lets the system relax into a far more stable, fully conjugated β-enaminone. The product is 4-(phenylamino)pent-3-en-2-one: PhNH–C(CH₃)=CH–C(=O)CH₃. The N–H, the C=C, and the C=O form a push-pull, intramolecularly hydrogen-bonded array.

This vinylogous amide is stable enough to isolate. It is the reason the Combes gives clean regiochemistry: the nitrogen is now locked onto one specific carbon.

Stage 2 — acid-catalyzed ring closure (the demanding step).

1. Concentrated H₂SO₄ (or PPA) protonates the surviving ketone oxygen. The carbon of that C=O becomes strongly electrophilic — essentially a protonated-ketone (oxocarbenium) center.
2. The aniline benzene ring acts as the nucleophile. Its π electrons swing onto that electrophilic carbon at the position ortho to the nitrogen, forming a new C–C bond and a cyclohexadienyl cation (arenium/σ-complex) — exactly the intermediate you meet in a Friedel-Crafts acylation.
3. Loss of the ring proton (deprotonation of the sp³ carbon) restores the benzene ring's aromaticity, giving a bicyclic carbinol (a 4-hydroxy-1,2,3,4-tetrahydro-type species).
4. A final acid-promoted dehydration removes that hydroxyl as water, and tautomerization sets up the fully aromatic pyridine ring. The result is the aromatic quinoline.

   Ph-NH₂  +  CH₃-C(=O)-CH₂-C(=O)-CH₃            (aniline + acetylacetone)
        │  −H₂O (condensation, no catalyst)
        ▼
   Ph-NH-C(CH₃)=CH-C(=O)-CH₃                     (β-enaminone — isolable)
        │  conc. H₂SO₄, Δ
        │  1) protonate the ketone → strong electrophile
        │  2) ring attacks ortho (intramolecular SEAr)
        │  3) rearomatize, then −H₂O
        ▼
      2,4-dimethylquinoline                       (+ 2 H₂O total)

The net transformation loses two molecules of water: one in the condensation, one in the cyclodehydration. Both departures are why a strong dehydrating acid outperforms a merely strong Brønsted acid.

Reagents, catalyst, and real conditions

• The amine. Any primary aromatic amine with at least one free ortho position: aniline itself, toluidines, anisidines, halo-anilines. Electron-rich anilines close fastest (the ring is the nucleophile).
• The 1,3-dicarbonyl. Symmetric diketones (acetylacetone, 3,5-heptanedione) give a single 2,4-disubstituted quinoline. Unsymmetrical 1,3-diketones and β-ketoesters/β-ketoaldehydes can give the Conrad–Limpach or Knorr quinolones instead, depending on which carbonyl condenses and the temperature — a distinction Combes chemistry shares with those cousins.
• The cyclization acid. Concentrated sulfuric acid is the historical reagent, but it can char sensitive substrates. Modern practice prefers polyphosphoric acid (PPA), P₂O₅, or anhydrous HF — all strong, hot, water-scavenging media that push the dehydrative ring closure and give cleaner, higher-yielding products.
• Temperature. Condensation is fine at 20–60 °C. Cyclization typically runs at 100 °C or above (PPA is often used at 120–140 °C).
• Stoichiometry. The acid is used in large excess — it is a reagent and a solvent, not a catalyst, because it consumes the water produced.

Scope, selectivity, and regiochemistry

Two selectivity questions decide whether the Combes is a good choice:

• Which carbonyl closes the ring? For a symmetric diketone the answer is trivial — both are equivalent. For unsymmetrical 1,3-dicarbonyls, the nitrogen tends to condense with the more electrophilic/less hindered carbonyl, and the other carbonyl does the cyclization. Getting a single regioisomer requires either symmetry or a large steric/electronic bias.
• Which ortho carbon of the aniline gets attacked? With aniline or a para-substituted aniline both ortho positions are equivalent (or only one is free), so ring closure is clean. With a meta-substituted aniline the two ortho positions are different, and you get a mixture of 5- and 7-substituted quinolines — the substituent's directing effect breaks the tie only partially.

Stereochemistry is essentially a non-issue: the product quinoline is flat and fully aromatic, so no stereocenters survive. The β-enaminone intermediate does have E/Z geometry about its C=C, but it equilibrates and does not affect the outcome.

The reaction dies on strongly electron-poor anilines (nitroanilines, polyhalogenated anilines) — the deactivated ring can no longer act as the nucleophile in the S_EAr closure, exactly as Friedel-Crafts fails on nitrobenzene.

Combes vs the other classic quinoline syntheses

	Combes	Skraup	Friedländer	Doebner–Miller
Amine partner	Aniline (primary aromatic amine)	Aniline	2-Aminoaryl ketone (ortho-acyl aniline)	Aniline
Carbonyl partner	1,3-Diketone	Glycerol (→ acrolein in situ)	Ketone/aldehyde with α-CH₂	α,β-Unsaturated aldehyde/ketone
Ring-closing step	Intramolecular SEAr on the aniline ring	SEAr after Michael addition	Aldol/Knoevenagel-type + condensation	Conjugate addition then SEAr
Typical product	2,4-Disubstituted quinoline	Unsubstituted quinoline	2,3-Disubstituted quinoline	2-Methylquinoline (quinaldine)
Key intermediate	β-Enaminone (isolable)	β-Anilinopropionaldehyde	Schiff base / aldol	Aniline–enal adduct
Cyclization conditions	Conc. H₂SO₄ / PPA / P₂O₅	H₂SO₄ + oxidant (nitrobenzene, As₂O₅)	Base or acid, milder	HCl, oxidant
Oxidant needed?	No (product already aromatic)	Yes (to aromatize)	No	Usually yes
Exotherm/hazard	Moderate (hot strong acid)	Severe — famously violent	Mild	Moderate
First reported	1888	1880	1882	1881

Worked example: aniline + acetylacetone → 2,4-dimethylquinoline

This is the textbook Combes run and the one Combes himself did in 1888.

  Step 1 (enaminone):
     PhNH₂ (1.0 eq) + CH₃COCH₂COCH₃ (1.0 eq)
        neat or in EtOH, RT–60 °C, −H₂O
     → PhNH-C(CH₃)=CH-COCH₃   (4-anilinopent-3-en-2-one)

  Step 2 (cyclodehydration):
     β-enaminone + conc. H₂SO₄ (large excess, solvent)
        100 °C, 1–2 h,  −H₂O
     → 2,4-dimethylquinoline

• Reagents. Aniline 1.0 equiv, pentane-2,4-dione 1.0 equiv for the condensation; then the isolated enaminone is added portionwise to a large excess of concentrated sulfuric acid (or PPA) for the ring closure.
• Why isolate the enaminone? Running the whole thing in one pot with sulfuric acid can protonate the aniline (turning off its nucleophilicity) and give tars. Making the enaminone first, under neutral conditions, then cyclizing it cleanly, is the reliable protocol.
• Workup. Pour the hot acid mixture cautiously onto crushed ice, basify with NaOH/ammonia, extract the free quinoline base into ether or DCM, and distil or recrystallize.
• Yield. With PPA rather than sulfuric acid, 2,4-dimethylquinoline is obtained in good yield (frequently 60–85% for the cyclization); sulfuric acid alone gives lower, more variable yields because of charring.

Where the Combes shows up

• 2,4-Disubstituted quinoline scaffolds. The Combes is the go-to when you specifically want substituents at both C-2 and C-4 — a pattern that Skraup (which gives unsubstituted quinoline) and Friedländer (2,3-pattern) cannot deliver. That 2,4-disubstitution appears in many antimalarial and antibacterial quinoline pharmacophores.
• Fluorinated and functionalized quinolines. Because the diketone is chosen freely, aryl-1,3-diketones and fluorinated diketones feed directly into Combes chemistry to make ligands and materials-science building blocks (quinolines are common metal-chelating and OLED emitter cores).
• Mechanistic teaching. The Combes is a favorite lecture example because it cleanly separates two textbook operations — enamine/enaminone condensation and Friedel-Crafts-type intramolecular acylation — into two isolable, teachable stages.
• 4-Quinolones by a close cousin. Swap the diketone for a β-ketoester and you slide into Conrad–Limpach / Knorr territory, which builds the 4-quinolone core of fluoroquinolone antibiotics (ciprofloxacin, levofloxacin). The Combes and these relatives together define classic acid-cyclization quinoline chemistry.

Limitations and side reactions

• Regiochemical scrambling on meta-anilines. As noted, a meta substituent leaves two inequivalent ortho positions, so 5- and 7-substituted quinolines both form. Plan the aniline substitution pattern to avoid this, or accept a separation.
• Deactivated anilines don't close. Nitro, cyano, and multiple halogens strip the ring of the electron density it needs for the S_EAr step; the enaminone forms but never cyclizes.
• Charring in concentrated sulfuric acid. Sensitive substrates decompose to intractable tars in hot conc. H₂SO₄. PPA, P₂O₅, or HF are the standard fixes.
• Aniline protonation. If you try a true one-pot reaction, the strong acid can protonate the aniline nitrogen before it condenses, shutting down enaminone formation. Do the condensation first, under neutral conditions.
• Unsymmetrical diketones. These can give mixtures of regioisomeric enaminones and hence mixtures of quinolines. Symmetric 1,3-diketones sidestep this entirely.

History: Alphonse Combes, 1888

The French chemist Alphonse Combes reported the reaction in 1888, condensing aniline with acetylacetone and cyclizing the resulting β-aminoenone with sulfuric acid to 2,4-dimethylquinoline. It arrived during the golden decade of quinoline chemistry: Zdenko Hans Skraup had published his glycerol route in 1880, Oscar Doebner & Wilhelm von Miller their unsaturated-carbonyl variant in 1881, and Paul Friedländer his ortho-aminoaryl-ketone route in 1882. Quinoline itself had been isolated from coal tar by Runge in 1834 and from quinine degradation by Gerhardt in 1842, and the rush to make substituted quinolines was driven by the dye industry and by the search for synthetic antimalarials to replace natural quinine. Combes's contribution was the clean two-stage logic — an isolable enaminone, then an intramolecular Friedel-Crafts-style closure — that still makes it the textbook route to 2,4-disubstituted quinolines.

Safety and practical notes

• Concentrated acid handling. The cyclization uses large volumes of hot concentrated sulfuric acid or PPA. Add substrate to acid, never acid to substrate; quench onto ice, and expect strong exotherms.
• Aniline toxicity. Aniline and many substituted anilines are toxic and absorbed through skin (methemoglobinemia risk). Work in a fume hood with gloves.
• HF variant. Anhydrous HF gives excellent cyclizations but is acutely dangerous (deep tissue burns, systemic fluoride toxicity); it belongs only in specialized facilities. PPA is the safer default for most benches.
• Water is the enemy. Since the ring closure is a dehydration, keep the cyclization medium anhydrous; adventitious water reverses the equilibrium and drops the yield.