Noyori Asymmetric Hydrogenation

What Noyori hydrogenation does

Ordinary catalytic hydrogenation — H₂ over Pd/C, PtO₂ (Adams' catalyst), or Raney nickel — happily reduces a double bond, but it is blind to handedness. When the substrate is prochiral (the C=C or C=O has two different faces, and reducing it creates a new stereocenter), a flat achiral metal surface delivers hydrogen to either face with equal probability. The result is a 50:50 racemate: equal amounts of the (R) and (S) product. For a chiral drug, half of that is at best dead weight and at worst harmful.

Noyori's insight was to replace the achiral surface with a single metal atom wrapped in a chiral ligand. A ruthenium or rhodium center holds a molecule of BINAP, a diphosphine that is locked into a left- or right-handed twist. That twist carves out a chiral pocket around the metal. When the prochiral substrate docks and H₂ is split and delivered, one face is heavily favored — so one enantiomer pours out at 95–99% enantiomeric excess (ee). Pick (R)-BINAP or (S)-BINAP and you choose which mirror image you get, from the very same starting material.

   Prochiral substrate     +   H₂    ──[Ru/Rh–(R)-BINAP]──▶   single enantiomer (up to 99% ee)

   (achiral Pd/C would give a 50:50 racemate instead)

The BINAP ligand: chirality with no stereocenter

BINAP is 2,2′-bis(diphenylphosphino)-1,1′-binaphthyl. Two naphthalene rings are joined by a single C–C bond, and each bears a diphenylphosphino (–PPh₂) group at the 2-position. The two phosphorus atoms chelate the metal as a bidentate diphosphine, forming a seven-membered ring with the metal at the apex.

What makes it remarkable is that BINAP has no sp³ stereocenter at all. Its chirality is axial — a form of atropisomerism. The two bulky naphthyl halves are too big to swing past each other around the connecting bond, so rotation is frozen and the molecule is permanently locked in one of two mirror-image twists, designated (R)-BINAP and (S)-BINAP. The dihedral twist between the naphthyl planes (roughly 70–90°) projects the four phenyl groups on phosphorus into a rigid, C₂-symmetric arrangement. That C₂ symmetry is the design trick: it halves the number of distinct substrate-binding orientations, so the two competing diastereomeric transition states are cleanly separated in energy — which is exactly what high ee requires.

        PPh₂                     the P–M–P bite plus the frozen
       /                         binaphthyl twist creates a chiral
   [naphthyl]                    "propeller" of four phenyl groups
       \\                          around the metal — a handed pocket
   [naphthyl]                    that discriminates the two faces of
       \\                          the prochiral substrate.
        PPh₂

Mechanism I — hydrogenating a functionalized alkene (inner sphere)

The original 1986 Ru-BINAP chemistry excels on alkenes that carry a coordinating group next to the double bond — enamides, itaconic acids/esters, β-keto esters, and allylic/homoallylic alcohols. That extra group is not a spectator; it anchors the substrate to the metal and sets up which face is exposed.

1. Chelation. The substrate binds the Ru(II) center through both its C=C π-bond and its coordinating heteroatom (the amide oxygen of an enamide, or the carboxylate of itaconic acid). This two-point binding forces one specific prochiral face toward the metal.
2. Heterolytic H₂ activation. Ru–BINAP splits H₂ heterolytically: one hydrogen becomes a metal hydride (Ru–H, a source of H⁻) and the other leaves as a proton (H⁺) picked up by a coordinated carboxylate or solvent. The two hydrogens therefore have different characters — hydride and proton — not two identical H atoms.
3. Migratory insertion. The Ru–H hydride migrates onto one carbon of the coordinated C=C, forming a new C–H bond and a Ru–alkyl σ-bond. Because the substrate face was fixed in step 1, this hydride is delivered to a single face — this is the enantio-determining step.
4. Reductive/protonolysis release. The remaining proton is delivered to the second carbon (via reductive elimination or protonolysis of the Ru–C bond), completing the CH–CH unit and releasing the saturated, enantiopure product. Ru(II) is regenerated and grabs the next substrate.

The electron bookkeeping is a classic organometallic cycle: heterolytic H₂ cleavage, then migratory insertion of a hydride into a coordinated alkene, then C–H bond-forming release. Handedness is imprinted at the insertion step, where the rigid BINAP pocket makes delivery to one face several kcal/mol lower in energy than to the other.

Mechanism II — hydrogenating a ketone (outer-sphere, bifunctional)

Simple ketones — an aryl methyl ketone like acetophenone — have no handy coordinating group. Noyori's 1995 breakthrough for these was the trans-RuCl₂(BINAP)(1,2-diamine) catalyst, activated by a base (KOH or KOtBu) and run under H₂. Here the mechanism is completely different: it is outer-sphere and bifunctional — the ketone never bonds to the metal.

1. Build the metal-ligand cooperative site. Base and H₂ convert the precatalyst into a species carrying a Ru–H hydride and an N–H proton on the diamine ligand. The metal supplies a hydride (δ⁻ H) and the amine N–H supplies an acidic proton (δ⁺ H). This is the "bifunctional" pair.
2. Concerted six-membered transition state. The C=O approaches the Ru–H / N–H pair edge-on. In one concerted step, the Ru–H hydride adds to the carbonyl carbon while the N–H proton adds to the carbonyl oxygen, through a cyclic six-membered transition state: Ru···H···C=O···H···N. No Ru–O or Ru–C bond ever forms.
3. Enantiocontrol from the chiral scaffold. The BINAP twist and the chiral diamine together only permit the ketone to slot in with one prochiral face presented to the hydride. That geometric gating is what selects (R)- or (S)-alcohol at 95–99% ee.
4. Regenerate. The alcohol product diffuses away, and H₂ heterolytically reloads the Ru–H / N–H pair for another turn.

Because the ketone is reduced by an ionic, hydride-plus-proton delivery rather than by binding and inserting, this chemistry tolerates other reducible groups beautifully: an isolated C=C double bond elsewhere in the molecule is untouched, because it has no lone pair to accept the proton. This chemoselectivity — reduce C=O, leave C=C alone — is the mirror image of what a Pd/C surface would do, and it is a direct consequence of the outer-sphere mechanism.

Catalyst, reagents, and conditions

• Metal precursor. For alkenes: Ru(II) sources such as [RuCl₂(benzene)]₂ or Ru(OAc)₂ combined with BINAP; for classic Knowles-style work and some enamides, Rh(I) with BINAP or DIPAMP. For ketones: trans-RuCl₂[(S)- or (R)-BINAP][(S,S)- or (R,R)-diamine], where the diamine is typically DPEN (1,2-diphenylethylenediamine) or DAIPEN.
• The chiral ligand. (R)-BINAP or (S)-BINAP — one enantiomer of the ligand for each enantiomer of product. Other axially chiral (atropisomeric) diphosphines such as SEGPHOS and MeO-BIPHEP work on the same principle; DIPAMP (P-stereogenic) and DuPhos (C-stereogenic phospholanes) achieve the same end through a different kind of chirality.
• Reductant. H₂ gas, from a few atmospheres up to ~100 atm depending on substrate. Ketone hydrogenations under the diamine system often run at only 4–10 atm.
• Base (ketone system only). KOH or KOtBu, catalytic, to build the active Ru–H / N–H species.
• Solvent. Methanol, ethanol, or isopropanol; sometimes CH₂Cl₂. Alcohols shuttle protons and dissolve H₂ well.
• Temperature. Frequently room temperature to ~50 °C — mild by industrial standards.
• Loading. As low as 0.001–0.1 mol% catalyst, thanks to turnover numbers of 10⁴–10⁵.

Scope, selectivity, and stereochemistry

The key requirement is a prochiral unsaturation — a C=C or C=O whose reduction creates a stereocenter. Beyond that, scope splits by mechanism:

• Functionalized alkenes (inner sphere): β-keto esters (→ β-hydroxy esters), α-(acylamino)acrylates (→ α-amino acids), itaconic acid derivatives, geraniol/nerol allylic alcohols, and enamides. The coordinating group is essential — it anchors the face.
• Ketones (outer sphere): aryl ketones, α,β-unsaturated ketones (reduced at C=O only, leaving C=C intact), and many dialkyl ketones. No directing group needed.
• Dynamic kinetic resolution: for α-substituted β-keto esters that racemize fast at the α-carbon under the reaction conditions, Ru-BINAP can convert a racemic starting material into a single diastereomer and enantiomer of the product — controlling two stereocenters at once.

Stereochemistry is dictated purely by ligand handedness. With a fixed substrate, switching from (R)-BINAP to (S)-BINAP flips the product from (R) to (S). The face selectivity comes from the several-kcal/mol energy gap between the two diastereomeric transition states enforced by the C₂-symmetric pocket; a gap of only ~2.7 kcal/mol at room temperature already corresponds to ~99:1 selectivity (≈98% ee).

Noyori hydrogenation vs. related methods

	Noyori (Ru/Rh-BINAP)	Achiral H₂ (Pd/C, PtO₂)	Sharpless epoxidation / dihydroxylation
Bond made chiral	C–H at a new stereocenter	C–H, but racemic	C–O (epoxide or diol)
Substrate	Prochiral C=C (functionalized) or C=O	Any C=C / C=O	Allylic alcohols / alkenes
Chirality source	Axial BINAP + chiral metal complex	None (achiral surface)	Chiral tartrate (Ti) or cinchona (Os)
Typical ee	95–99%	0% (racemic)	90–99%
Catalyst amount	Catalytic, TON 10⁴–10⁵	Catalytic	Catalytic (Ti/Os) + stoich. oxidant
Reductant / oxidant	H₂ gas	H₂ gas	TBHP or K₃Fe(CN)₆ (oxidations)
Chooses either enantiomer?	Yes — swap (R)/(S)-BINAP	No	Yes — swap tartrate / ligand
Nobel recognition	Noyori & Knowles, 2001	—	Sharpless, 2001

A real application: Takasago's L-menthol process

The most celebrated industrial use is not even a hydrogenation of H₂ across a double bond — it is a Rh-BINAP-catalyzed asymmetric isomerization, but it runs on the identical BINAP-chirality principle and is the flagship demonstration of the technology at scale.

   diethylgeranylamine  ──[Rh–(S)-BINAP]──▶  (R)-citronellal enamine  ──H₃O⁺──▶  (R)-citronellal
                          (TON > 100,000)                                          |
                                                                                   | ZnBr₂ cyclize, then H₂
                                                                                   ▼
                                                                             (−)-menthol  (L-menthol)

• The chiral step. A prochiral allylic amine (diethylgeranylamine) is isomerized with a cationic Rh–(S)-BINAP catalyst that shifts the double bond and, in doing so, creates the single new stereocenter of (R)-citronellal at high ee. The Rh–(R)-BINAP enantiomer would give the mirror image.
• Efficiency. The Rh-BINAP isomerization runs at turnover numbers above 100,000 — a whisper of catalyst converts an ocean of substrate.
• Scale. Takasago has produced on the order of thousands of tons of L-menthol per year by this route since the 1980s — a large fraction of the world's synthetic menthol for toothpaste, cough drops, and flavorings.

The archetypal hydrogenation example is equally clean: methyl acetoacetate (a β-keto ester) reduced by Ru–(R)-BINAP under a few atmospheres of H₂ in methanol gives methyl (R)-3-hydroxybutanoate at ~99% ee — a chiral building block that classical chemistry could only reach by fermentation or resolution.

Where it shows up in medicine and industry

• (S)-Naproxen. Ru-BINAP hydrogenation of a prochiral acrylic-acid precursor gives the (S)-anti-inflammatory at very high ee; the (R)-enantiomer is essentially inactive as an NSAID and is associated with liver toxicity, so single-enantiomer synthesis is not a luxury but a safety requirement.
• β-Amino acids and β-lactam intermediates. Asymmetric hydrogenation of β-(acylamino)acrylates and β-keto esters supplies enantiopure β-amino-acid and β-hydroxy building blocks used in the paclitaxel (Taxol) C-13 side chain and various protease inhibitors.
• Carbapenem antibiotics. Ru-BINAP dynamic kinetic resolution of β-keto esters builds the syn-amino-alcohol stereochemistry of key penem/carbapenem intermediates in a single step.
• Levofloxacin and other fluoroquinolones. Chiral alcohol intermediates set by Noyori ketone hydrogenation feed antibacterial manufacture.
• Fragrances and vitamins. Beyond L-menthol, BINAP hydrogenation sets stereocenters in intermediates toward vitamin E and various single-enantiomer fragrance molecules.

Limitations and practical cautions

• Directing group needed for alkenes. The inner-sphere alkene chemistry needs a coordinating group (amide, ester, carboxylate, hydroxyl) near the double bond. An unfunctionalized, isolated trisubstituted alkene is a poor Ru-BINAP substrate; those are the domain of Crabtree-type Ir catalysts instead.
• Oxygen and moisture sensitivity. Ru(0)/Ru(II)-BINAP species and the base-activated diamine complexes are air-sensitive; reactions run under inert atmosphere with degassed solvents. Trace O₂ oxidizes the phosphine and kills activity.
• Substrate poisons. Strong metal binders — free thiols, some amines, halide-rich impurities, CO — can coordinate the ruthenium and shut the catalyst down. Purity of substrate and H₂ matters.
• High-pressure H₂ hazard. Many alkene runs use tens of atmospheres of hydrogen; flammable-gas and high-pressure vessel precautions apply.
• Cost and recovery of the metal. Ruthenium/rhodium and enantiopure BINAP are expensive; the economics only work because turnover numbers are enormous, and industrial processes often include metal-recovery steps to recycle the precious metal.
• Matched vs mismatched pairs. In dynamic kinetic resolutions the ligand handedness must match the substrate's racemization; a mismatched (R)/(S) combination erodes both yield and ee.

Historical discovery: who and when

Ryoji Noyori and Hidemasa Takaya synthesized BINAP in 1980 at Nagoya University. In 1986 Noyori's group reported that Ru(II)-BINAP complexes hydrogenated functionalized alkenes with unprecedented enantioselectivity, and through the late 1980s and 1990s extended the platform to β-keto esters, allylic alcohols, and — with the 1995 trans-RuCl₂(BINAP)(diamine) systems — to simple ketones by the bifunctional outer-sphere mechanism.

The idea built on William S. Knowles, who at Monsanto in the late 1960s and 1970s created the first practical chiral hydrogenation catalyst (Rh with the chiral phosphine DIPAMP) and used it to manufacture L-DOPA, the Parkinson's-disease drug — the first industrial asymmetric catalytic process. Noyori generalized that proof-of-concept into a broad, high-ee, high-turnover technology.

In 2001 the Nobel Prize in Chemistry was awarded to William S. Knowles and Ryoji Noyori "for their work on chirally catalysed hydrogenation reactions," together with K. Barry Sharpless "for his work on chirally catalysed oxidation reactions." The unifying lesson of that prize: a single small chiral molecule, used catalytically, can imprint handedness on millions of product molecules.