The Johnson-Claisen Rearrangement

What the Johnson-Claisen does

The Johnson-Claisen (or "orthoester Claisen") rearrangement is a two-carbon chain extension dressed up as a functional-group swap. You start with an allylic alcohol — an -OH sitting one carbon away from a C=C double bond — and you finish with a γ,δ-unsaturated ester: an ester whose carbonyl is separated from a new C=C double bond by exactly two carbons.

Two things happen simultaneously, which is what makes the reaction so useful:

• A new carbon-carbon bond is forged at the far end of the old allyl system — the carbon that used to be the terminal alkene carbon.
• The double bond walks over by one position (an allylic transposition). The carbon that used to carry the -OH becomes an alkene carbon in the product.

Formally it is the aliphatic Claisen rearrangement of an allyl vinyl ether — a [3,3]-sigmatropic shift. The clever part of the Johnson variant is how it builds that allyl vinyl ether in one pot from a cheap allylic alcohol and a cheap orthoester, without a strong base or a metal.

    R-CH=CH-CH₂-OH   +   CH₃-C(OEt)₃   ──H⁺ cat., 140 °C──→   R-CH(-CH₂CO₂Et)-CH=CH₂
       allylic alcohol      triethyl               γ,δ-unsaturated ester
                            orthoacetate           (new C-C bond + transposed C=C)

The step-by-step mechanism

The overall transformation is a cascade of acid-catalyzed equilibria that funnels into one irreversible sigmatropic step. Track the electron flow through four stages:

1. Transesterification onto the orthoester. The weak acid protonates one -OEt of the orthoester; that oxygen leaves as ethanol, generating a stabilized oxocarbenium ion R'C(OEt)₂⁺. The allylic alcohol's oxygen lone pair attacks this cation, and after loss of a proton you have a mixed orthoester: the allyl group is now one of the three alkoxy substituents on the orthoester carbon.
2. Elimination to the ketene acetal. Acid protonates a second -OEt; it leaves as ethanol, generating an oxocarbenium again. This time an α-C-H is lost instead of adding a nucleophile, giving a C=C. The result is a mixed ketene acetal — CH₂=C(OEt)(O-allyl) — which is precisely an allyl vinyl ether. This is the species primed to rearrange.
3. The [3,3]-sigmatropic shift. Now no catalyst is needed. Six electrons reorganize around a six-membered chair-like ring in one concerted step: the O-C(allyl) σ-bond breaks, a new C-C σ-bond forms between the vinyl-ether terminal CH₂ and the allyl terminus, and both π-bonds slide over. Aromatic-like bond reorganization, but through a saturated pericyclic transition state.
4. Collapse to the ester. Because the rearranging carbon already carries the second oxygen (the -OEt), the shift delivers the ester directly: the bridging oxygen re-forms its double bond to become the carbonyl C=O, and the -OEt stays put as the ester alkoxy group. The product simply is the γ,δ-unsaturated ethyl ester; no further workup chemistry is required beyond distilling away ethanol and excess reagent.

  Step 1  allyl-OH + CH₃C(OEt)₃  ──H⁺──→  CH₃C(OEt)₂(O-allyl)   + EtOH   (mixed orthoester)
  Step 2  CH₃C(OEt)₂(O-allyl)    ──H⁺──→  CH₂=C(OEt)(O-allyl)   + EtOH   (ketene acetal)
  Step 3  CH₂=C(OEt)(O-allyl)    ──[3,3]──→  chair TS  ──→  new C-C bond, C=C transposed
  Step 4  → γ,δ-unsaturated ethyl ester  (R-CH(CH₂CO₂Et)-CH=CH₂)

The whole point of continuously distilling off ethanol is that steps 1 and 2 are equilibria; removing ethanol by Le Chatelier drives them forward and keeps feeding the irreversible step 3.

The chair transition state and stereochemistry

Every synthetically important feature of the Johnson-Claisen — its (E)-selectivity, its faithful chirality transfer, its predictable regiochemistry — comes straight out of the six-membered chair-like transition state. Draw the allyl vinyl ether folded so the two termini that will bond are stacked over one another, and you get a ring reminiscent of cyclohexane.

• (E)-alkene preference. A substituent on the forming double bond prefers the pseudo-equatorial position of the chair to escape 1,3-diaxial strain. That single conformational preference dictates that the new trisubstituted (or disubstituted) alkene emerges predominantly (E). Ratios of 90:10 (E:Z) or higher are routine for simple cases; that reliability is why Johnson built terpenoid double bonds this way.
• 1,3-chirality transfer. If the starting allylic alcohol carries a stereocenter at the carbinol carbon, the rigid chair relays that information to a new stereocenter formed two carbons away, with high fidelity. The reaction is suprafacial-suprafacial, as required by the Woodward-Hoffmann rules for a thermal [3,3] shift (4n+2 = 6 electrons, thermally allowed).
• No competing boat. The chair is strongly preferred over the boat transition state (typically by several kcal/mol), which is what makes the stereochemical outcome clean rather than a scramble.

Reagents, catalyst, and conditions

The classic recipe is almost aggressively simple, which is much of its appeal:

• Orthoester (the acyl source). Triethyl orthoacetate, CH₃C(OEt)₃, is the default and delivers a -CH₂CO₂Et unit. Trimethyl orthoacetate works too. Higher orthoesters (triethyl orthopropionate, triethyl orthobutyrate) install a substituted α-carbon and set an additional stereocenter. Typically used in 3-10 fold excess.
• Acid catalyst. A small amount (0.05-0.3 equiv) of a weak carboxylic acid — propionic acid is the textbook choice; others use pivalic acid, 2,4-dinitrophenol, or hydroquinone-buffered acids. The acid must be weak enough not to simply dehydrate the allylic alcohol or polymerize the ketene acetal.
• Temperature. 130-160 °C, often neat or in a high-boiling inert solvent. The [3,3] shift itself has an activation barrier around 30-35 kcal/mol, which is why real heat is required.
• Ethanol removal. A short-path distillation head lets the liberated ethanol boil off as it forms, dragging the equilibria toward the ketene acetal.
• No inert atmosphere heroics. Unlike the Ireland variant, there is no enolate to protect, no LDA, no low temperature. This is a heat-and-stir reaction.

Johnson-Claisen vs the other Claisen ester variants

	Johnson-Claisen	Ireland-Claisen	Eschenmoser-Claisen
Key reagent	Orthoester, e.g. CH₃C(OEt)₃	Allyl ester + LDA + TMSCl	MeC(OMe)₂NMe₂ (DMA-DMA)
Reactive intermediate	Ketene acetal (in situ)	Silyl ketene acetal	Ketene N,O-acetal
Catalyst / base	Weak acid (propionic acid)	Strong base (LDA)	None / mild heat
Temperature	130-160 °C	-78 °C then warm	110-160 °C
Product	γ,δ-unsaturated ester	γ,δ-unsaturated acid/ester	γ,δ-unsaturated amide
Alkene stereocontrol	(E)-selective (chair)	(E) or (Z) selectable (enolate geometry)	(E)-selective (chair)
Operational cost	Very low — heat & stir	High — anhydrous, cryogenic	Low-moderate
Best for	Robust, scalable ester synthesis	Fine stereocontrol, aldol-like sites	Amide targets, alkaloids

All three, together with the amide-forming Overman and the ketone-forming Carroll rearrangements, belong to the same family: convert an allylic alcohol into a chain-extended, double-bond-transposed carbonyl compound via a [3,3] shift. Which one you pick is dictated by the oxidation state you want in the product and how much stereocontrol you need.

Worked example: a Johnson terpene building block

Take (E)-2-methyl-2-buten-1-ol (also called tiglyl alcohol: HOCH₂-C(CH₃)=CH-CH₃) and heat it with triethyl orthoacetate and a drop of propionic acid at 140 °C, distilling off ethanol:

    CH₃-CH=C(CH₃)-CH₂-OH  +  CH₃C(OEt)₃  ──propionic acid (0.1 eq), 140 °C──→
                                        EtO₂C-CH₂-CH(CH₃)-C(CH₃)=CH₂
                                        ethyl 3,4-dimethyl-4-pentenoate (a γ,δ-ester)
                                        + 2 EtOH (distilled off)

• Reagents. Allylic alcohol 1.0 equiv, triethyl orthoacetate 5 equiv, propionic acid ~0.1 equiv.
• Conditions. Reflux/heat to 140 °C in a flask fitted with a distillation head; collect the ethanol distillate over 2-6 h.
• Outcome. The old C2=C3 double bond transposes to the former carbinol carbon (giving the 1,1-disubstituted isopropenyl terminus, the classic isoprene-unit cap), while a new -CH₂CO₂Et arm is installed at the old alkene terminus. Because this particular product alkene is 1,1-disubstituted, it has no E/Z to set — the (E)-selectivity of the reaction shows up whenever the new double bond is 1,2-di- or trisubstituted instead.
• Workup. Cool, wash out the acid and excess orthoester, and distill or chromatograph the ester.

Swap in a substrate whose transposed double bond ends up trisubstituted — say a (2E)-3-methyl allylic alcohol such as a geraniol fragment — and the same chair transition state now delivers that trisubstituted alkene predominantly (E), which is exactly the geometry Johnson needed for isoprenoid chains.

Johnson's own landmark use was iterative: each Johnson-Claisen extends a chain by a two-carbon, stereodefined trisubstituted-alkene unit — exactly the repeating motif of an isoprenoid chain. Stitch several together and you have assembled the polyene precursor for a biomimetic polyene cyclization, the strategy Johnson used to build steroid skeletons such as those en route to progesterone in a single acid-triggered cascade. The reaction's ability to set alkene geometry reliably is what made those cyclizations stereospecific.

Scope, selectivity, and limitations

• Substrate scope. Primary and secondary allylic alcohols both work; the more substituted the allyl terminus, the more valuable the (E)-selectivity becomes. Propargylic and homoallylic alcohols do not engage — the transposing double bond has to be exactly allylic to the -OH.
• Acid sensitivity. Because a Brønsted acid and high heat are present, acid-labile groups (acetals, some silyl ethers, tertiary allylic alcohols prone to ionization) can misbehave. Tertiary allylic alcohols often just dehydrate or give SN1 side products rather than the clean rearrangement.
• Competing ionization. Allylic cations formed by protonation/loss of water can give allylic rearrangement products or elimination alkenes. Keeping the acid weak and catalytic suppresses this.
• Thermal load. The requirement for 130-160 °C rules out very thermally fragile substrates; the Ireland-Claisen (which runs cryogenically) is the answer there.
• Regiochemistry is fixed by the framework. Unlike an aldol, you cannot choose which end reacts — the [3,3] topology dictates the new bond's position. This is a feature (predictable) as much as a limitation (inflexible).

Historical discovery

The parent reaction is Ludwig Claisen's 1912 discovery that allyl vinyl ethers rearrange thermally to γ,δ-unsaturated carbonyls — the aliphatic Claisen rearrangement. For decades the bottleneck was making the required allyl vinyl ether, which was often awkward.

William S. Johnson at Stanford solved that bottleneck in 1970 (Johnson, W. S.; Werthemann, L.; Bartlett, W. R.; Brocksom, T. J.; Li, T.; Faulkner, D. J.; Petersen, M. R. J. Am. Chem. Soc. 1970, 92, 741). His insight was to generate the ketene acetal — the vinyl-ether component — in situ from a cheap orthoester under simple acid catalysis, so the whole allyl-vinyl-ether-plus-rearrangement sequence collapses into one hot pot. Johnson pursued it explicitly as a tool for building isoprenoid double bonds with defined geometry, and it became a cornerstone of his celebrated biomimetic polyene cyclization approach to steroids. The Ireland (1972-76) and Eschenmoser (1964-69) variants round out the family, each trading operational simplicity for a different product oxidation state or a finer grip on stereochemistry.

Practical and safety notes

• Orthoesters are moisture-sensitive. Triethyl orthoacetate hydrolyzes to ethyl acetate and ethanol on contact with water/acid, so keep glassware dry and use a genuinely weak acid; a strong acid trace will destroy the reagent before it can do useful chemistry.
• Drive the equilibria. If yields stall, the usual culprit is incomplete ethanol removal — a proper short-path head and a slow gentle distillation matter more than the exact acid.
• Choose the acid deliberately. Propionic acid is mild and volatile (easy to remove); phenolic acids like 2,4-dinitrophenol or hydroquinone additions also double as radical inhibitors that suppress polymerization of the ketene acetal.
• Scale and greenness. The reaction is atom-economical in spirit — the only stoichiometric byproduct is ethanol — and needs no exotic metals, which is why it survives as a scalable, inexpensive route to γ,δ-unsaturated esters in process chemistry.
• Downstream flexibility. The ethyl ester product is a versatile handle: hydrolyze to the acid, reduce to the alcohol/aldehyde, or enolize for a subsequent aldol — and the terminal alkene is ripe for ozonolysis, hydroboration, or metathesis.