The Ireland-Claisen Rearrangement

What the Ireland-Claisen does

Start with an ordinary allylic ester — an acid esterified to an allylic alcohol, so the skeleton reads R–C(=O)–O–CH₂–CH=CH₂. The Ireland-Claisen rearrangement rewires that molecule into a γ,δ-unsaturated carboxylic acid: a new carbon-carbon bond forms between the ester's α-carbon and the far (γ) carbon of the allyl group, the allylic C–O bond breaks, and the double bond walks one carbon over. You get an acid whose alkene now sits two carbons out from the carboxyl.

The clever part is how the reaction manufactures a pericyclic array that the neutral ester simply does not have. A classical Claisen rearrangement needs an allyl vinyl ether — the six-atom C=C–O–C–C=C chain that folds and shifts. An ester's carbonyl is not a C=C, so it can't play. Ireland's insight was to enolize the ester and cap the enolate oxygen with a silyl group, converting the C=O into a genuine, electron-rich C=C. That silyl ketene acetal is the missing vinyl ether, and it rearranges at strikingly mild temperature.

   allyl ester            silyl ketene acetal              γ,δ-unsaturated acid
                LDA, −78 °C            [3,3] shift              H₃O⁺ / F⁻
  R-CH₂-C(=O)-O-CH₂-CH=CH₂  ─────────►  R-CH=C(OTMS)-O-CH₂-CH=CH₂  ──────────►  silyl ester  ─────►  R-CH(-CH₂-CH=CH₂)-COOH
                then TMSCl                                    (warm, ~25 °C)

The mechanism, arrow by arrow

Four operations, only one of which is the pericyclic event itself:

1. Deprotonate the ester α-carbon. LDA (lithium diisopropylamide) or LiHMDS, added at −78 °C, removes the most acidic proton — the one α to the carbonyl (pK_a ≈ 30). The electrons flow onto oxygen, giving a lithium ester enolate with a C=C between Cα and the carbonyl carbon and a negative charge on O.
2. Trap the enolate as a silyl ketene acetal. Chlorotrimethylsilane (TMSCl) or TBSCl is present in the pot; the enolate oxygen's lone pair attacks silicon, kicking out chloride. The oxygen is now an O–SiR₃ ether, and the C=C is locked in place. You have converted a carbonyl into a ketene acetal: C=C(OSiR₃)(OR′), where OR′ is the allyloxy group. This is the electron-rich vinyl ether that a Claisen needs.
3. Fold into a chair and fire the [3,3] shift. On warming, the six atoms of the array — the two ketene-acetal carbons, the acetal oxygen, and the three allyl carbons — curl into a six-membered chair transition state. Three electron pairs move around the ring in one concerted loop: the C(allyl)–O σ bond breaks, the ketene-acetal π bond becomes the new C–C σ bond, and the allyl C=C slides to the adjacent carbon. No ion is ever released.
4. Hydrolyze the silyl ester. The immediate product is the trimethylsilyl (or TBS) ester of the new acid. Dilute aqueous acid, or a fluoride source such as TBAF, cleaves the Si–O bond and delivers the free γ,δ-unsaturated carboxylic acid.

  The concerted [3,3] step (three curved arrows, one loop):

        O–SiR₃                        O–SiR₃
        |                             |
    Cα==C                         Cα--C           (ketene-acetal Cα=C consumed; O–Si stays single)
        \                             ‖
         O           ──chair──►        O           (this acetal O becomes the new C=O carbonyl)
         |                             ⋮
        C1                            C1           (C(allyl)–O σ bond breaks)
        |                             ‖
        C2==C3                        C2--C3        (allyl π walks; new Cα–C3 σ bond forms)
                                       |
                                      (new C–C bond ties Cα to C3)

Count the electrons: three arrows, six electrons, a Hückel-topology six-membered transition state — thermally allowed as a suprafacial-suprafacial process by the Woodward-Hoffmann rules, exactly like the parent Claisen and Cope rearrangements.

Reagents, silanes, and real conditions

The recipe is compact but every reagent earns its place:

• Base. LDA is the default (1.0-1.1 equiv), generated in situ from n-BuLi + diisopropylamine at 0 °C, then cooled to −78 °C before the ester is added. LiHMDS is a common substitute when the ester bears base-sensitive groups.
• Silylating agent. TMSCl is fast and cheap; TBSCl (with a touch of HMPA to boost reactivity) gives a hardier, isolable ketene acetal. TMS ketene acetals often rearrange in the same flask on warming; TBS versions can be purified first.
• Solvent. THF is standard. Crucially, adding roughly 23% HMPA (hexamethylphosphoramide) — or the less toxic DMPU — flips the enolate geometry (see below). Temperature: deprotonate and silylate at −78 °C, then warm to 25-65 °C to drive the rearrangement.
• Workup. Quench with dilute HCl or aqueous NH₄Cl, or treat with TBAF/fluoride, to cleave the silyl ester and free the carboxylic acid. Extract and purify as the acid or, after diazomethane/TMS-diazomethane, the methyl ester.

Because the ketene acetal is so electron-rich, the shift is comfortable near room temperature — a dramatic improvement over the ~200 °C needed for a plain allyl vinyl ether Claisen. That mildness is what makes the reaction compatible with the sensitive functionality of a real total synthesis.

Stereochemistry: solvent chooses the diastereomer

This is the reaction's marquee feature. Two stereochemical relationships stack up:

1. Enolate geometry is set by solvent. Deprotonating the ester with LDA in pure THF gives predominantly the (Z)-ester enolate, which is trapped as the (E)-silyl ketene acetal. Adding ~23% HMPA (or DMPU) reverses the selectivity to the (E)-enolate, hence the (Z)-silyl ketene acetal. The additive changes the aggregation and the geometry of the lithium/base transition state during deprotonation.
2. The chair relays geometry into the product. Because the six atoms rearrange through a rigid chair, each ketene-acetal geometry (E or Z) maps cleanly onto one relative configuration of the two new stereocenters. Pick the solvent, pick the enolate geometry, and you have effectively picked syn or anti product with high diastereoselectivity.
3. Chirality is transferred, not created from nothing. A stereocenter or defined alkene in the starting allylic alcohol is relayed 1,3 across the forming bond. Feed a single enantiomer of the allylic alcohol and the suprafacial, ion-free chair delivers a single, predictable enantiomer of the acid — "1,3-chirality transfer."

   Solvent            Enolate geometry    Silyl ketene acetal    Product relative config
   ─────────────────  ──────────────────  ─────────────────────  ───────────────────────
   THF only           (Z)-enolate         (E)-ketene acetal      one diastereomer (e.g. anti)
   THF + 23% HMPA     (E)-enolate         (Z)-ketene acetal      the other (e.g. syn)

The upshot: from a single ester you can access either diastereomer of a γ,δ-unsaturated acid just by choosing whether HMPA is in the flask. Very few reactions give that kind of switchable stereocontrol on a simple functional-group handle.

How it compares to related [3,3] shifts

	Claisen (classical)	Ireland-Claisen	Cope	Overman
Starting array	Allyl vinyl ether	Silyl ketene acetal (from allyl ester)	1,5-diene	Allylic trichloroacetimidate
What moves	O across allyl	C-C forms; O→acid	C-C bond migrates	N onto carbon
Product	γ,δ-unsaturated carbonyl	γ,δ-unsaturated acid	Isomeric 1,5-diene	Allylic amide/amine
Typical temperature	~180-200 °C	−78 °C → ~25-65 °C	150-250 °C	60-140 °C (or Pd, RT)
Stereocontrol handle	Substrate geometry	Enolate geometry (THF vs HMPA)	Substrate geometry	Chair face / chiral Pd
New stereocenters	Up to 1	Up to 2 contiguous	Up to 2	1
Concerted?	Yes	Yes	Yes	Yes (thermal)

The Ireland variant's edge is threefold: the mildest temperature of the thermal Claisen family, the ability to install a carboxylic acid directly, and switchable diastereocontrol from a one-line change in solvent.

Worked example: an allyl propanoate to a 2-methyl-4-pentenoic acid

Take allyl propanoate, CH₃CH₂–C(=O)–O–CH₂–CH=CH₂. The α-carbon bears the methyl-bearing carbon that will become the new stereocenter.

  CH₃CH₂-C(=O)-O-CH₂-CH=CH₂

  1) LDA (1.05 eq), THF, −78 °C            → (Z)-ester enolate
  2) TBSCl (1.2 eq), −78 °C → warm         → (E)-silyl ketene acetal:  CH₃CH=C(OTBS)-O-CH₂-CH=CH₂
  3) warm to 25-40 °C                       → [3,3] chair shift
  4) H₃O⁺ (or TBAF)                         → 2-methyl-4-pentenoic acid  CH₃-CH(COOH)-CH₂-CH=CH₂

• Bond bookkeeping. The α-carbon of the ester (formerly CH₃CH₂–) becomes Cα of the acid, and a new C–C bond ties it to the terminal (γ) allyl carbon. The allylic C–O bond is gone; the double bond has walked to become a terminal vinyl on the acid chain.
• Yield and selectivity. Ireland-type rearrangements of simple allyl esters routinely run in 70-90% yield with 90:10 or better diastereoselectivity when the solvent is chosen deliberately; running the same substrate in THF/HMPA inverts the major diastereomer.
• Why it's useful here. In one pot you built a C–C bond, set an α-methyl stereocenter, installed a terminal alkene for later cross-metathesis or hydroboration, and delivered a free acid handle — four synthetic operations folded into one rearrangement.

Named applications in synthesis

• Ireland's own ionophore and macrolide work. Robert Ireland deployed the ester-enolate Claisen extensively in polyether antibiotic and macrolide campaigns (e.g., toward monensin-type and lasalocid-type fragments), exploiting the reliable relay of enolate geometry into contiguous stereocenters on complex chains.
• Terpene and steroid side chains. The reaction is a standard way to append a stereodefined, alkene-bearing acid arm — the "isoprenoid" fragment — onto an allylic alcohol drawn from a terpene skeleton, because the 1,3-chirality transfer preserves existing configuration.
• Amino-acid and unnatural-acid synthesis. Glycine- and alanine-derived allyl esters rearrange to γ,δ-unsaturated α-amino acids, a route to conformationally constrained and non-proteinogenic residues for medicinal chemistry.
• Ring-contraction and macrocyclization strategies. Because the shift transfers stereochemistry through a rigid chair, it is used to set remote stereocenters on rings and to build medium-ring acids where classical enolate alkylation would be sluggish or unselective.
• Continuous-flow scale-up. The low-temperature deprotonation/silylation and mild rearrangement have been adapted to continuous-flow reactors, improving heat and mixing control for the cryogenic enolization step on process scale.

Limitations and side reactions

• Cryogenic enolization is finicky. Clean enolate geometry demands accurate −78 °C, dry solvents, and careful stoichiometry; warming during deprotonation erodes the E/Z ratio and therefore the diastereoselectivity.
• Silyl ester hydrolysis on workup. TMS esters can hydrolyze prematurely or during silica chromatography; if the ketene acetal must survive isolation, switch to the sturdier TBS group.
• Competing [1,3] shifts and ionization. Substrates whose allylic carbon can ionize to a stabilized cation (benzylic, or highly substituted allylic systems, especially when overheated) can leak into a stepwise pathway that scrambles regiochemistry and erodes the enantiomeric excess. Keep the temperature as low as the rate allows, or move to a metal-catalyzed variant.
• Boat vs chair ambiguity in rings. Cyclic silyl ketene acetals (for example cyclohexenyl systems) can rearrange through either a chair or a boat transition state, giving only moderate stereoselectivity where an acyclic case would be clean.
• HMPA toxicity. HMPA is a suspected carcinogen; DMPU is the usual less-hazardous substitute when the (E)-enolate manifold is needed, though the E/Z ratios can differ slightly.

Who, when, and why it mattered

Robert E. Ireland introduced the ester-enolate Claisen rearrangement in 1972 (R. E. Ireland and R. H. Mueller, J. Am. Chem. Soc. 1972, 94, 5897), and he and his coworkers spent the following decade mapping out the stereochemistry — most influentially the discovery that the silyl ketene acetal's geometry, and therefore the product configuration, could be switched simply by adding HMPA to the deprotonation. That work (notably the detailed 1976 and 1980s JACS and JOC studies) turned a laboratory curiosity into a predictable stereochemical tool.

The rearrangement's lineage runs back to Ludwig Claisen, who described the thermal allyl-vinyl-ether rearrangement in 1912, and it is a first cousin of the all-carbon Cope rearrangement. Ireland's contribution was recognizing that an ester could be coaxed into the Claisen manifold by turning its carbonyl into a silyl-masked enol — a maneuver that combined C–C bond formation, acid installation, and stereocontrol in a single, exceptionally mild step. It remains one of the most-used sigmatropic reactions in complex-molecule synthesis.