The Ohira-Bestmann Reagent

What the reagent does

You have an aldehyde and you want a terminal alkyne — a carbon-carbon triple bond ending in an acidic C-H. The Ohira-Bestmann reagent does this in one flask, replacing the aldehyde's C=O with a C≡CH and adding one carbon to the chain:

    R-CHO   ──[Ohira-Bestmann reagent], K₂CO₃, MeOH, RT──→   R-C≡C-H

What makes it beloved in a synthesis lab is what it does not require. The obvious way to build an alkyne from an aldehyde is the classic Seyferth-Gilbert homologation, which needs a strong, cryogenic base (potassium tert-butoxide or n-butyllithium at −78 °C). Those conditions cheerfully destroy esters, free alcohols, base-sensitive stereocenters, and half the functional groups on a real advanced intermediate. The Ohira-Bestmann protocol reaches the same reactive species with nothing but potassium carbonate in methanol at room temperature. Suddenly you can homologate an aldehyde sitting next to an ester, a Boc group, or an epoxide.

The reagent itself is a yellow oil: dimethyl (1-diazo-2-oxopropyl)phosphonate, CH₃-C(=O)-C(=N₂)-P(=O)(OMe)₂. Read it as three functional groups on one carbon — a diazo group, an acetyl group, and a dimethyl phosphonate. The acetyl group is the trick that lets a weak base do the job.

The mechanism, arrow by arrow

The transformation runs through five distinct steps. Only the first is unique to the Ohira-Bestmann variant; steps 2–5 are the Seyferth-Gilbert pathway, now reached under mild conditions.

1. Unmask the reagent (acetyl cleavage). K₂CO₃ generates a small amount of methoxide in methanol. Methoxide attacks the acetyl carbonyl of the reagent; the tetrahedral intermediate collapses to expel methyl acetate and leave the dimethyl (diazomethyl)phosphonate carbanion, (MeO)₂P(=O)-C⁻(=N₂). This anion is the very species the Seyferth-Gilbert reagent forms only after deprotonation by a strong base — here it is unveiled by a mild transesterification instead.
2. Add to the aldehyde. The diazo-stabilized carbanion is a soft nucleophile. It adds to the electrophilic carbonyl carbon of RCHO. The carbonyl π electrons flow onto oxygen, giving a β-alkoxide diazophosphonate — a new C-C bond, with the negative charge now on the former aldehyde oxygen.
3. Form and fragment the oxaphosphetane. Exactly as in a Wittig-type olefination, the alkoxide oxygen closes onto the phosphorus to make a strained four-membered oxaphosphetane ring (P-O-C-C). It fragments by retro-[2+2]: the P=O bond forms fully and the ring cleaves, expelling dimethyl phosphate (the phosphorus by-product) and leaving a vinyl diazo species, R-CH=C=N₂ / R-C(=N₂)=CH... — essentially a diazo group on a carbon now doubly bonded to the R-bearing carbon.
4. Lose nitrogen → alkylidene carbene. The diazo group is a superb leaving group as N₂ gas. Its departure leaves an electron-deficient carbon bearing a double bond to the neighboring carbon: an alkylidene carbene (a vinylidene carbene), R-CH=C:. This is the pivotal reactive intermediate.
5. 1,2-shift forges the triple bond. The carbene is desperate for two more electrons. The C-H bond on the adjacent (R-bearing) carbon migrates — a 1,2-hydrogen shift — sliding its bonding pair into the empty carbene orbital. That single migration converts the C=C=: into H-C≡C-, delivering the terminal alkyne R-C≡C-H and completing the homologation.

  reagent:   CH₃-C(=O)-C(=N₂)-P(=O)(OMe)₂
                    │  MeO⁻ (from K₂CO₃/MeOH)
                    ▼
  step 1:    CH₃CO₂Me  +  ⁻C(=N₂)-P(=O)(OMe)₂      (unmasked anion)
  step 2:    RCHO + ⁻C(=N₂)P(O)(OMe)₂ → R-CH(O⁻)-C(=N₂)-P(=O)(OMe)₂
  step 3:    oxaphosphetane → (MeO)₂P(=O)OH  +  R-CH=C=N₂  (vinyl diazo)
  step 4:    R-CH=C=N₂  →  R-CH=C:  +  N₂↑           (alkylidene carbene)
  step 5:    R-CH=C:  ──1,2-H shift──→  R-C≡C-H       (terminal alkyne)

Note where the atoms end up: the aldehyde's carbon becomes the internal alkyne carbon (still bearing R), and the aldehyde's hydrogen is the very hydrogen that does the 1,2-shift, ending up on the terminal ≡CH. The extra carbon of the alkyne comes from the reagent's diazo carbon. It is a genuine one-carbon homologation.

Reagents, base, and conditions

• The reagent. Dimethyl (1-diazo-2-oxopropyl)phosphonate, CAS 90965-06-3. A pale-yellow oil, commercially available or prepared in one step from dimethyl (2-oxopropyl)phosphonate by diazo transfer (usually with p-acetamidobenzenesulfonyl azide, p-ABSA, or tosyl azide, plus a base). Typical loading 1.1–2.0 equiv.
• The base. Anhydrous K₂CO₃, 2–3 equiv, is standard. Cs₂CO₃ is used for sluggish substrates; it is more soluble and gives faster acetyl cleavage. The base does not deprotonate anything acidic on the reagent — it exists only to generate a catalytic pool of methoxide.
• The solvent. Anhydrous methanol is required, and it is mechanistically active, not inert: MeOH is the nucleophile that cleaves the acetyl trigger. Mixed MeOH/THF or MeOH/Et₂O is common when the substrate is poorly soluble. Ethanol can be substituted but is slower.
• Temperature and time. Add the reagent to the aldehyde/K₂CO₃ slurry at 0 °C, then stir at room temperature 2–16 h. Sensitive or easily-epimerized aldehydes are often run entirely at 0 °C.
• Work-up. Quench with water or saturated NH₄Cl, extract, and chromatograph. The by-products — methyl acetate, dimethyl phosphate salts, and N₂ gas — are all water-soluble or volatile and wash away cleanly.

Scope, selectivity, and stereochemistry

The reaction is broad across aldehyde classes and famously mild:

• Aromatic aldehydes (benzaldehyde → phenylacetylene) work in 80–95% yield. Electron-poor aromatics react fastest because the carbonyl is more electrophilic.
• Aliphatic aldehydes homologate well; branching at the α-carbon slows the addition step but rarely stops it.
• α,β-Unsaturated (conjugated) aldehydes give conjugated enynes — cinnamaldehyde becomes (E)-but-1-en-3-yne-derived products cleanly, with the alkene geometry preserved.
• Chirality is preserved. Because the mildest conditions run at 0 °C with a weak base, α-stereocenters and even epimerizable centers survive. This is the single biggest reason the reagent displaced Seyferth-Gilbert in total synthesis: you can homologate a chiral aldehyde next to a stereocenter without racemization.

There is no new stereocenter created — the product is a linear alkyne — so "stereochemistry" here means retention of existing centers, which the mild conditions deliver. Chemoselectivity is excellent: esters, lactones, carbamates (Boc, Cbz), silyl ethers, free hydroxyls, epoxides, and even ketones generally survive, because none of them react with the diazophosphonate anion faster than the aldehyde does.

Ohira-Bestmann vs the alternatives

	Ohira-Bestmann	Seyferth-Gilbert	Corey-Fuchs
Base needed	K₂CO₃ (mild)	KOtBu / n-BuLi (strong)	PPh₃ then n-BuLi (strong)
Temperature	0 °C → RT	−78 °C	0 °C then −78 °C
Number of pots	One	One	Two (dibromoolefination, then elimination)
Substrate	Aldehydes (not ketones)	Aldehydes & some ketones	Aldehydes
Carbon change	+1 (terminal alkyne)	+1 (terminal alkyne)	+1 (terminal alkyne)
Functional-group tolerance	Excellent (esters, OH, epoxides survive)	Poor (strong base destroys many)	Moderate (CBr₄/PPh₃ + strong base)
Epimerization risk	Very low	High	Moderate
By-products	MeOAc, (MeO)₂PO₂H, N₂	(MeO)₂PO₂⁻, N₂	CHBr₃, Ph₃PO, LiBr
Reagent handling	Bench-stable oil, mild	Same reagent, needs cryo base	CBr₄ / PPh₃, then organolithium

Worked example: an aldehyde to a terminal alkyne

Homologate 4-nitrobenzaldehyde to 1-ethynyl-4-nitrobenzene, a common Sonogashira coupling partner:

    4-O₂N-C₆H₄-CHO  +  (MeO)₂P(O)C(=N₂)C(O)CH₃  (1.3 eq)
        ──K₂CO₃ (2.0 eq), MeOH, 0 °C → RT, 4 h──→  4-O₂N-C₆H₄-C≡C-H   (~90%)

• Charge the aldehyde (1.0 equiv) and anhydrous K₂CO₃ (2.0 equiv) into dry methanol under nitrogen, cool to 0 °C.
• Add the Ohira-Bestmann reagent (1.3 equiv) dropwise. Gas evolution (N₂) is visible as the reaction proceeds.
• Stir 30 min at 0 °C, then warm to room temperature for 3–4 h; monitor by TLC until the aldehyde spot is consumed.
• Quench with water, extract into ether, wash, dry, and chromatograph to give the arylacetylene in roughly 90% yield.

The nitro group, which would be reduced or attacked under many strong-base conditions, is completely untouched. That electron-poor ring actually accelerates the addition step, so this substrate is among the fastest-reacting.

Real-world applications

• Total synthesis of natural products. The reagent is a workhorse for installing terminal alkynes on advanced, sensitive intermediates — the alkyne is later used for a Sonogashira coupling, a click reaction, an alkyne metathesis, or a ring-closing enyne metathesis. It appears in syntheses of epothilones, amphidinolides, and many polyketides precisely because it doesn't touch the fragile ester/stereocenter framework.
• Feedstock for click chemistry. A terminal alkyne is the partner in the copper-catalyzed azide-alkyne cycloaddition (CuAAC). Chemists routinely take an aldehyde-bearing scaffold, run Ohira-Bestmann to expose a clean ≡CH handle, then "click" on an azide-tagged label, drug, or affinity probe.
• Bioconjugation and chemical biology. Because the mild conditions preserve base-sensitive functionality, the reagent is used to append alkyne handles to sugars and peptidomimetic building blocks that are later attached to biomolecules.
• Materials and pharma building blocks. Arylacetylenes made this way feed into conjugated polymers, OLED emitters, and drug scaffolds that rely on a Sonogashira step to stitch the alkyne to an aryl halide.

Limitations and side reactions

• Ketones basically don't work. The extra steric bulk and lower electrophilicity of a ketone carbonyl let reagent decomposition outrun addition. Use classic Seyferth-Gilbert or dibromoolefination for internal alkynes from ketones.
• Enolizable aldehydes. Very acidic α-protons can let a trace of aldol or enolization compete; keeping the run cold (0 °C) and short minimizes it.
• Moisture. Excess water hydrolyzes the reagent and quenches the methoxide pool, killing yields. Anhydrous methanol and dry K₂CO₃ matter.
• Diazo hazards. The reagent is a diazo compound. It is far more stable than diazomethane, but it is still a potentially explosive, potentially toxic functionality — handle behind a shield, avoid friction and strong acids, and do not concentrate large amounts. The safer, less shock-sensitive character (relative to diazomethane) is one reason it caught on.
• Carbene side paths. If the substrate carries a C-H that can do a 1,5-insertion or a nearby double bond, the alkylidene carbene can occasionally cyclize or do a C-H insertion instead of the desired 1,2-shift. For simple aldehydes the 1,2-H shift dominates overwhelmingly.

Historical discovery

The lineage starts with Dieter Seyferth, who in 1971 showed that dimethyl (diazomethyl)phosphonate, deprotonated with base, reacts with carbonyls to give alkynes. John C. Gilbert developed the synthetic method into a reliable aldehyde-to-alkyne homologation in the late 1970s and early 1980s — hence "Seyferth-Gilbert homologation" — but it demanded potassium tert-butoxide at −78 °C.

Seiji Ohira published the crucial fix in 1989 (Synthetic Communications): put an acetyl group on the reactive carbon so that a weak base (K₂CO₃) in methanol could unmask the reagent in situ, no strong base required. Hans-Jürgen Bestmann and coworkers at Erlangen refined and championed the reagent and conditions in a widely cited 1996 Synlett paper, standardizing the K₂CO₃/MeOH protocol. Because both chemists shaped the reagent that bears the modification, it is universally called the Ohira-Bestmann reagent (and the reaction, the Ohira-Bestmann or Seyferth-Gilbert-Ohira-Bestmann homologation).