The Julia Olefination

What the Julia olefination does

The Julia olefination is a carbonyl olefination — it converts an aldehyde's C=O into a C=C double bond, welding a new carbon fragment on in the process. Its two coupling partners are a sulfone (R-CH₂-SO₂-Het, the carbanion source) and an aldehyde (R′-CHO, the electrophile). The net transformation is:

    R-CH₂-SO₂-Het  +  R′-CHO   ──base──→   R-CH=CH-R′   +  "SO₂ + HetO⁻"

What makes it distinctive among olefinations is where the two fragments come from. The Wittig needs a phosphonium salt; the Horner-Wadsworth-Emmons (HWE) needs a stabilizing ester or ketone on the phosphonate. The Julia sulfone is neither — it is a bench-stable, chromatographable, storable solid that can carry an arbitrarily complex carbon skeleton. That is why, when a synthesis needs to join two elaborate halves across a trans double bond, the Julia-Kocienski reaction is often the tool of choice.

There are two generations to keep straight. The classic Julia (Julia-Lythgoe) uses a phenyl sulfone and runs as a stepwise sequence — add, trap, then reduce. The modern Julia-Kocienski uses a heteroaryl sulfone (benzothiazolyl or, better, phenyltetrazolyl) and collapses the whole thing to the alkene in one flask, with no separate reduction. Both are covered below.

The classic Julia-Lythgoe mechanism

The original reaction is a four-operation sequence, and each operation earns its place:

1. Deprotonation. A strong base (n-BuLi, or an amide base) removes the proton α to the sulfone. The sulfone's two S=O groups stabilize the resulting carbanion by delocalizing the negative charge onto oxygen — the same reason a sulfone α-C-H has a pK_a around 29 (in DMSO), far more acidic than a plain alkane.
2. Addition to the aldehyde. The sulfone-stabilized carbanion attacks the aldehyde carbonyl carbon. The carbonyl π electrons collapse onto oxygen, generating a β-alkoxide next to the sulfone. This addition is reversible, so the diastereomer ratio here does not fix the final alkene geometry.
3. Trap the alkoxide. The β-alkoxide is acylated in situ — usually as a benzoate (PhCOCl / benzoic anhydride) or acetate, sometimes a mesylate. This turns a poor leaving group (alkoxide) into a good one and, crucially, freezes the intermediate so it can be isolated and purified.
4. Reductive elimination. A single-electron reductant — sodium amalgam (Na/Hg) in methanol at low temperature, or samarium(II) iodide (SmI₂) — cleaves the C-SO₂ bond to give a β-acyloxy radical/carbanion, which then eliminates the ester in an E1cb-like or radical-anion pathway. The two new sp² carbons form the double bond.

  1.  R-CH₂-SO₂Ph  +  base   →   R-CH⁻-SO₂Ph   (α-sulfonyl carbanion)
  2.  R-CH⁻-SO₂Ph  +  R′CHO  →   R-CH(SO₂Ph)-CH(O⁻)-R′   (β-alkoxide)
  3.  + PhCOCl              →   R-CH(SO₂Ph)-CH(OBz)-R′   (β-benzoyloxy sulfone)
  4.  + Na/Hg (or SmI₂)     →   R-CH=CH-R′  +  PhSO₂⁻  +  BzO⁻   (E-alkene)

The E-selectivity of the classic version is a feature of the reductive elimination step, not the addition: the radical-anion intermediate is configurationally mobile and relaxes to the lower-energy anti geometry before it eliminates, so the thermodynamically favored trans alkene predominates regardless of how the addition went. That thermodynamic control is why Julia-Lythgoe is reliably E-selective for 1,2-disubstituted alkenes.

The modern Julia-Kocienski mechanism

Kocienski's one-pot variant replaces the phenyl group with a heteroaryl that is electrophilic at the carbon bonded to sulfur. Now the intermediate doesn't need an external reductant — it self-destructs:

1. Deprotonate and add. As before: base removes the α-proton, the α-heteroaryl-sulfonyl carbanion adds to the aldehyde, giving a β-alkoxide.
2. Smiles rearrangement. The alkoxide oxygen attacks the ipso (C-S) carbon of the heteroaryl ring intramolecularly. The heteroaryl migrates from sulfur to oxygen — a Truce-Smiles-type rearrangement — cleaving the C(aryl)-S bond and leaving a sulfinate anion (R-CH(SO₂⁻)-CH(O-Het)-R′).
3. Elimination of SO₂. The sulfinate and the newly installed heteroaryloxide leave together in a syn-periplanar (E1cb-like) elimination, expelling sulfur dioxide and the heteroaryl alkoxide and forming the C=C bond.

  R-CH⁻-SO₂-PT  +  R′CHO
        │  addition
        ▼
  R-CH(SO₂-PT)-CH(O⁻)-R′          β-alkoxide
        │  Smiles: O attacks ipso C of PT ring, PT migrates S → O
        ▼
  R-CH(SO₂⁻)-CH(O-PT)-R′          β-(heteroaryloxy) sulfinate
        │  syn elimination: expel SO₂ + PT-O⁻
        ▼
  R-CH=CH-R′   +   SO₂   +   PT-O⁻      (E-alkene)

The whole cascade happens in one flask, often just on warming from −78 °C. Because the geometry is set at the elimination step — and under the standard non-chelating conditions the antiperiplanar arrangement of the bulky groups is favored — the PT-sulfone route is reliably E-selective. This is the version you will see in almost every modern total synthesis.

Reagents, bases, solvents, and conditions

• The sulfone. Made by S-alkylation of a heteroaryl thiol (2-mercaptobenzothiazole for BT; 1-phenyl-1H-tetrazole-5-thiol for PT) with an alkyl halide (or via Mitsunobu with an alcohol), then oxidation of the sulfide to the sulfone with m-CPBA, H₂O₂/ammonium molybdate, or Oxone. The PT-sulfone is the workhorse.
• The base. KHMDS, NaHMDS, or LiHMDS (the hexamethyldisilazides) are standard because they are strong, non-nucleophilic, and their bulky, poorly-coordinating character is exactly what favors E-alkene formation. n-BuLi and LDA are used for the classic phenyl-sulfone version.
• The counterion. This is a selectivity dial. Potassium (KHMDS) and non-chelating conditions give E; magnesium or lithium with a chelating solvent can push toward Z. Barbier-type conditions (premix sulfone + aldehyde, then add base) sometimes improve yield with sensitive aldehydes.
• The solvent. DME, THF, DMF, or toluene. Coordinating vs non-coordinating solvent shifts E/Z ratio, so the solvent is chosen alongside the base.
• Temperature. Typically −78 °C for the deprotonation/addition, then warm to room temperature to drive the Smiles/elimination cascade.
• Reductant (classic only). 6% sodium amalgam in buffered MeOH/THF at −20 °C, or SmI₂/HMPA. SmI₂ is milder and mercury-free, and it tolerates more functionality.

Scope, selectivity, and stereochemistry

The Julia-Kocienski reaction shines at making 1,2-disubstituted (E)-alkenes between two carbon fragments of any size. The stereochemical outcome is controlled by four coupled variables — sulfone type, base/counterion, solvent, and temperature:

• PT-sulfone + KHMDS + DME, −78 °C → E. The canonical E-selective recipe. Bulky potassium base and non-coordinating conditions favor the open, anti transition arrangement.
• BT-sulfone + chelating counterion → often Z. The older benzothiazolyl sulfones under lithium/magnesium chelation can favor Z, and were exploited for that when Z was the target.
• Aldehyde matters. Aromatic and α,β-unsaturated aldehydes tend to give higher E-selectivity than simple aliphatic ones; α-branched aldehydes can erode it.

Because the addition step is reversible and the elimination is stereodefining, the reaction is under Curtin-Hammett-type control — you are not locked into whatever diastereomer forms first. That is precisely why tuning the base and solvent, rather than the substrate, lets you steer geometry.

Julia vs Wittig vs HWE vs Peterson

	Julia-Kocienski	Wittig	Horner-Wadsworth-Emmons	Peterson
Carbanion source	Heteroaryl sulfone (PT/BT)	Phosphonium ylide	Phosphonate carbanion	α-silyl carbanion
Typical geometry	E (trans), tunable	Z with non-stabilized; E with stabilized	E (with ester)	E or Z (acid vs base workup)
Needs a reductant?	No (Kocienski) / yes (classic Lythgoe)	No	No	No
Byproduct	SO₂ + heteroaryl alkoxide	Ph₃P=O (hard to remove)	Dialkyl phosphate (water-soluble)	R₃Si-O⁻
Best for	Non-stabilized (E)-alkenes between complex fragments	Simple alkenes, semi-stabilized	α,β-unsaturated esters/ketones	Alkenes where Si is already present
Reagent stability	Sulfone is a stable, storable solid	Ylides often unstable	Phosphonates stable	Silanes stable
Handles ketones?	Poorly (aldehyde-optimized)	Yes	Modestly	Yes

Worked example: a polyene fragment coupling

Suppose you need the (E)-alkene joining two halves of a macrolide, both of which are precious. You install a PT-sulfone on one half and leave an aldehyde on the other:

    R-CH₂-SO₂-(1-phenyltetrazol-5-yl)  +  R′-CHO
        │  KHMDS (1.1 eq), DME, −78 °C, 30 min
        │  then warm to 25 °C, 1–2 h
        ▼
    R-CH=CH-R′   (E:Z typically 90:10 → >95:5)   +  SO₂  +  PT-OH (after workup)

• Order of addition. Cool the PT-sulfone in DME to −78 °C, add KHMDS to form the deep-colored anion, then add the aldehyde dropwise. (For fragile aldehydes, the Barbier variant — premix sulfone + aldehyde, then add base — can improve yield.)
• Workup. Quench with saturated NH₄Cl, extract, and chromatograph. The only organic byproduct is 1-phenyl-5-hydroxytetrazole (PT-OH), which separates cleanly from the alkene.
• Why not Wittig here? A non-stabilized Wittig ylide would give the Z-alkene, and the Ph₃P=O byproduct co-elutes with many products. HWE can't help without an ester at the reacting carbon.

This exact logic — PT-sulfone plus complex aldehyde, KHMDS, one pot — is how the trans double bonds in landmark polyene syntheses were forged, from the C1-C15 fragment couplings in the epothilone campaigns to numerous macrolide and polyketide targets.

Limitations and side reactions

• Self-condensation of BT-sulfones. The metalated benzothiazolyl sulfone can attack the electrophilic C2 of a second BT-sulfone, dimerizing before it ever meets the aldehyde. This was the central weakness that PT-sulfones were designed to fix — the tetrazole has no such electrophilic carbon.
• Ketones are poor partners. The reaction is aldehyde-optimized; ketone additions are slow, reversible, and low-yielding, with eroded selectivity.
• Enolizable and epimerizable aldehydes. Strong amide bases can deprotonate or epimerize α-stereocenters on the aldehyde. Controlled addition, low temperature, and sometimes the Barbier protocol mitigate this.
• E/Z drift. Selectivity is condition-dependent; a substrate that gives clean E under KHMDS/DME may erode with a different base, counterion, or solvent, so conditions usually need optimization per fragment.
• Mercury (classic route). Sodium amalgam is toxic and generates mercury waste. SmI₂ or the fully reductant-free Kocienski route are the modern, greener answers.

Historical discovery: who and when

The reaction is named for Marc Julia, the French chemist at the Institut Pasteur / ESPCI in Paris. In 1973, Julia and Jean-Marc Paris reported the original phenyl-sulfone sequence — deprotonate, add to the aldehyde, acetylate, and reductively eliminate with sodium amalgam. Basil Lythgoe at Leeds developed and popularized the reductive-elimination step, so the classic version is called the Julia-Lythgoe olefination.

In 1991, Sylvestre Julia (Marc's brother, also a chemist) reported that benzothiazol-2-yl (BT) sulfones could be used in a one-pot protocol, collapsing to the alkene without a separate reductant via the Smiles-rearrangement cascade. Then in 1998, Philip Kocienski and coworkers introduced the 1-phenyl-1H-tetrazol-5-yl (PT) sulfones, curing the self-condensation problem of the BT reagents and giving higher yields and better E-selectivity. The PT-sulfone, KHMDS-in-DME protocol that resulted is what most chemists mean today when they say "Julia-Kocienski."

Practical and industrial notes

• Convergence. Because both fragments can be arbitrarily complex and the sulfone is a stable solid, Julia-Kocienski is a favorite for convergent total synthesis: build the two halves separately, then join them late with a stereodefined trans double bond.
• Reagent shelf life. PT- and BT-sulfones are crystalline, chromatographable, and store for months — unlike many phosphonium ylides. This makes route planning and scale-up predictable.
• Green chemistry. The reductant-free Kocienski route eliminates the mercury waste of the classic Lythgoe version, and the byproducts (SO₂, a water-soluble heteroaryloxide) are easier to handle than triphenylphosphine oxide.
• Scale considerations. HMDS bases and cryogenic (−78 °C) conditions add cost at scale, but the clean byproduct profile and reliable E-selectivity often justify it for high-value pharmaceutical intermediates.