Skip to content

Asymmetric Induction and Asymmetric Synthesis

CSIR NET: Cram, Felkin–Anh and chelation models, 1,3-induction, chiral auxiliaries, chiral reagents and catalysts, double induction, kinetic resolution.

  • 18 min read
  • Updated
On this page · 16 sections
  1. Prerequisites
  2. The Hook: A Drug Made with a Chiral Phosphine
  3. What Asymmetric Induction Means
  4. 1,2-Induction: Nucleophilic Addition to α-Chiral Carbonyls
  5. 1,3-Induction and Directed Reactions
  6. Chiral Auxiliaries
  7. Chiral Reagents and Catalysts
  8. Double Asymmetric Induction
  9. Kinetic Resolution and Related Strategies
  10. Memory Tricks
  11. Don’t Confuse
  12. Common Mistakes
  13. Practice
  14. Quick Revision Table
  15. Key Takeaways
  16. What’s Next

Prerequisites

The Hook: A Drug Made with a Chiral Phosphine

What Asymmetric Induction Means

Asymmetric induction
The preferential formation of one stereoisomer at a new stereogenic unit, caused by a chiral feature already present in the substrate, a reagent, a catalyst or the environment. The induction is called 1,2-, 1,3- or 1,4- according to the distance between the existing stereocentre and the new one (counting the new centre as position 1 and moving along the chain to the controlling centre at position 2, 3 or 4, in the common usage).
  • Internal (substrate) induction gives diastereomers; the selectivity is a dr.
  • External induction (chiral reagent or catalyst acting on a prochiral substrate) gives enantiomers; the selectivity is an ee.
  • Relayed induction (chiral auxiliary): the auxiliary is attached, induces a diastereoselective reaction, and is removed. The overall result is an enantioselective synthesis, but the key step is diastereoselective, so the minor isomer can be removed by ordinary chromatography or crystallisation before cleavage.

Ways to obtain one enantiomer

graph TD
    G["Goal: one enantiomer"] --> CP["Chiral pool
start from amino acids,
sugars, terpenes, tartaric acid"] G --> RS["Resolution
diastereomeric salts,
chiral HPLC,
kinetic resolution"] G --> AS["Asymmetric synthesis"] AS --> SC["Substrate control
existing stereocentre
(Felkin-Anh, chelation)"] AS --> AU["Auxiliary control
Evans, Oppolzer,
Enders, Myers"] AS --> RC["Reagent control
stoichiometric chiral
reagent (Ipc2BH)"] AS --> CC["Catalyst control
Sharpless, Noyori,
CBS, organocatalysts,
enzymes"]

1,2-Induction: Nucleophilic Addition to α-Chiral Carbonyls

The classic test case: a nucleophile (hydride, Grignard reagent, enolate) adds to an aldehyde or ketone whose α-carbon carries three different groups, labelled L (large), M (medium) and S (small). The two faces of the carbonyl are diastereotopic, and the models below predict which is attacked.

Cram’s open-chain model

  • The reactive conformation places the largest group L anti to the carbonyl oxygen (L eclipsing the smaller carbonyl substituent R), with the oxygen flanked by M and S. In Cram’s original picture the metal-complexed oxygen is the bulkiest part of the carbonyl.
  • The nucleophile attacks the carbonyl carbon from the side of S.
  • Cram’s rule was empirical (from additions to 2-phenylpropanal-type substrates) and predicts the major isomer in most simple cases, but it uses an eclipsed reactant-like conformation and has no clear energetic basis.

The Felkin–Anh model

Felkin proposed a staggered, reactant-like transition state, and Anh added the trajectory of attack.

Felkin–Anh model
  1. View along the Cα–C(=O) bond. In the reactive conformation the largest group L is perpendicular to the plane of the carbonyl, anti to the incoming nucleophile.
  2. The nucleophile approaches at the Bürgi–Dunitz angle (obtuse, roughly 105–110° to the C=O bond), not perpendicular to it.
  3. Of the two conformers with L perpendicular, the favoured one has the nucleophile passing next to S rather than M.
  4. Polar Felkin–Anh: an electronegative α-substituent (OR, Cl, NR₂) takes the role of L, irrespective of size, because the C–X bond perpendicular to the carbonyl lowers the π* energy through overlap with \( \sigma^*_{\text{C–X}} \).
  • The Felkin product is the diastereomer predicted by this model; the other is the anti-Felkin product.
  • For simple alkyl and aryl α-substituents, Cram and Felkin–Anh predict the same major product; Felkin–Anh is preferred because it rests on a sensible transition state and explains the trends (bulkier nucleophiles and larger L give higher selectivity).
  • Selectivities from substrate control alone are often modest (a few to one), and improve with bulkier nucleophiles, larger differences between M and S, and low temperature.

Chelation control (Cram’s cyclic model)

When the α-substituent is a Lewis-basic group (OR, NR₂) and the metal can chelate (Mg²⁺, Zn²⁺, Ti⁴⁺, Ce³⁺, Mn²⁺), the carbonyl oxygen and the α-heteroatom bind the metal in a five-membered chelate.

  • The chelate holds the C=O and the C–OR bond syn-coplanar.
  • The nucleophile attacks from the face opposite to the larger of the two remaining α-groups.
  • The result is usually the anti-Felkin diastereomer, i.e. the opposite of the polar Felkin–Anh prediction.
ConditionModel that appliesTypical outcome
α-OR with small, chelating protecting group (OMe, OBn, OMOM) and a chelating metal (MgBr, ZnBr₂, TiCl₄)chelationchelation (anti-Felkin) product
α-OR with a bulky silyl group (OTBDPS, OTIPS) or a non-chelating counter-ion (K⁺, BF₃)polar Felkin–AnhFelkin product
α-alkyl or α-aryl onlyFelkin–Anh (= Cram)Felkin product

The switch is a powerful synthetic tool: by changing only the protecting group or the Lewis acid, the same substrate can be steered to either diastereomer.

The Cornforth model

For α-halo carbonyl compounds, Cornforth proposed a reactive conformation in which the C–X bond is anti to C=O to minimise dipole repulsion; the nucleophile then attacks past the smaller remaining group. It and the polar Felkin–Anh model often predict the same major product. For CSIR NET, Felkin–Anh, polar Felkin–Anh and chelation are the three to know.

1,3-Induction and Directed Reactions

Reduction of β-hydroxy ketones

MethodMechanismProduct
Narasaka–Prasad: Et₂BOMe (or Bu₂BOMe), then NaBH₄boron chelates both oxygens in a six-membered ring; external hydride attacks from the less hindered facesyn-1,3-diol
Evans–Saksena: Me₄N⁺ BH(OAc)₃⁻the free OH exchanges onto boron; hydride is delivered intramolecularly from the same boronanti-1,3-diol

The two methods give complementary diastereomers from one substrate, which made 1,3-diol (polyketide) synthesis systematic.

Hydroxyl-directed reactions

  • Henbest epoxidation: a peracid epoxidises cyclohex-2-en-1-ol mainly cis (syn) to the OH, because the OH hydrogen-bonds to the peracid in the transition state. Acetylating the OH removes the effect, and the larger group then directs to the opposite face.
  • VO(acac)₂ / t-BuOOH and Ti(OiPr)₄ / t-BuOOH: the metal binds the allylic alcohol and the peroxide together; epoxidation is syn to the OH, usually with higher selectivity than the peracid.
  • Simmons–Smith cyclopropanation of allylic alcohols is also directed syn by zinc coordination.
  • In acyclic allylic alcohols, allylic 1,3-strain fixes the reactive conformation (the allylic C–H lies near the plane of the alkene when there is a cis substituent), which makes these directed reactions highly diastereoselective.

Chiral Auxiliaries

Chiral auxiliary strategy
  1. Attach an enantiopure, recoverable chiral group to the substrate.
  2. Carry out the key bond-forming step; the auxiliary makes the competing transition states diastereomeric (high dr).
  3. Purify the major diastereomer (ordinary chromatography or crystallisation).
  4. Cleave the auxiliary (hydrolysis, reduction, transamidation) and recover it.
AuxiliaryDerived fromTypical useStereocontrol
Evans oxazolidinones (4-benzyl- or 4-isopropyl-oxazolidin-2-one)amino acids (via amino alcohols)N-acyl imide enolate alkylation; boron aldolthe C4 substituent shields one enolate face; Bu₂BOTf / i-Pr₂NEt gives the Z-boron enolate, and the Zimmerman–Traxler TS gives the “Evans syn” aldol
Oppolzer’s camphorsultamcamphoralkylations, aldols, Diels–Alder, conjugate additionsthe sultam’s rigid bicyclic frame blocks one face
Enders’ SAMP / RAMP hydrazones(S)- or (R)-prolineα-alkylation of ketones and aldehydesmetalated hydrazone chelated by the methoxymethyl group
Myers’ pseudoephedrine amidespseudoephedrineα-alkylation of amide enolateslithium alkoxide and enolate held in a rigid arrangement
8-Phenylmenthol esters(+)-pulegone (chiral pool; Corey)Diels–Alder, ene reactionsthe phenyl ring stacks over one face of the acrylate

Advantages: reliable, predictable, and the diastereomers can be separated to high purity. Drawbacks: two extra steps (attach and remove) and a stoichiometric amount of the chiral material.

Chiral Reagents and Catalysts

Reagent control (stoichiometric)

  • Diisopinocampheylborane (Ipc₂BH) and IpcBH₂, made from α-pinene: asymmetric hydroboration of alkenes (Brown).
  • Alpine-Borane (B-isopinocampheyl-9-BBN): asymmetric reduction of reactive ketones such as acetylenic ketones (Midland).
  • Chiral allyl- and crotylboranes / boronates (Brown’s Ipc₂B-allyl; Roush’s tartrate boronates): addition to aldehydes, where the alkene geometry of the crotyl reagent sets syn or anti (a Zimmerman–Traxler-type chair) and the chiral ligand sets the absolute configuration.

Catalyst control (substoichiometric)

ReactionCatalyst systemSubstratesNotes
Sharpless asymmetric epoxidation (SAE)Ti(OiPr)₄, diethyl tartrate (DET), t-BuOOHallylic alcoholspredictable face: with the allylic alcohol drawn with C=C vertical and CH₂OH at the lower right, L-(+)-DET delivers O from the bottom face, D-(−)-DET from the top
Sharpless asymmetric dihydroxylation (AD)catalytic OsO₄ (as K₂OsO₂(OH)₄), cinchona ligand, K₃Fe(CN)₆most alkenesAD-mix-α contains (DHQ)₂PHAL, AD-mix-β contains (DHQD)₂PHAL; the two ligands give opposite enantiomers
Jacobsen–Katsuki epoxidationchiral Mn(III)–salen, NaOClunfunctionalised alkenes, best for cis-disubstitutedcomplements SAE (no OH needed)
Knowles hydrogenationRh(I)–DIPAMPdehydroamino acid derivatives (enamides)L-DOPA process
Noyori hydrogenationRu(II)–BINAPβ-keto esters, functionalised alkenes and ketonesBINAP is axially chiral (no stereocentre)
CBS reductionchiral oxazaborolidine (from prolinol) + BH₃prochiral ketonesthe catalyst binds borane and the ketone, placing the larger ketone substituent away from the catalyst’s ring
Proline catalysis(S)-prolineintramolecular (Hajos–Parrish) and intermolecular aldol, Mannichenamine mechanism; the carboxylic acid directs the electrophile
MacMillan imidazolidinoneschiral secondary amine saltsDiels–Alder, conjugate additions of α,β-unsaturated aldehydesiminium activation lowers the LUMO
Enzymeslipases, esterases, dehydrogenasesresolutions, desymmetrisations, reductionshigh selectivity, often > 99% ee

Double Asymmetric Induction

When a chiral substrate meets a chiral reagent or catalyst, the two inductions combine.

  • Matched pair: both favour the same new configuration; selectivity is higher than either alone.
  • Mismatched pair: they oppose each other; selectivity is lower, and if the reagent is strong enough it still wins (reagent control).

If the two contributions to \( \Delta\Delta G^\ddagger \) are additive, the ratios multiply:

$$\text{dr}_{\text{matched}} \approx \text{dr}_{\text{substrate}} \times \text{dr}_{\text{reagent}}, \qquad \text{dr}_{\text{mismatched}} \approx \frac{\text{dr}_{\text{reagent}}}{\text{dr}_{\text{substrate}}}$$

This is the basis of “reagent control” strategies (Masamune): use a reagent so selective that it overrides any substrate preference, so that either diastereomer can be made at will.

Kinetic resolution

A chiral catalyst or reagent converts one enantiomer of a racemate faster than the other. The selectivity factor is

$$s = \frac{k_{\text{fast}}}{k_{\text{slow}}} = \frac{\ln[(1-c)(1-\text{ee}_{\text{SM}})]}{\ln[(1-c)(1+\text{ee}_{\text{SM}})]}$$

where c is the fractional conversion and \( \text{ee}_{\text{SM}} \) the ee of the recovered starting material. Mass balance links the two ees: \( c\cdot\text{ee}_{P} = (1-c)\cdot\text{ee}_{\text{SM}} \).

  • Maximum yield of either enantiomer is 50%.
  • The ee of the recovered starting material rises steadily with conversion and can approach 100% for any s > 1 if you are willing to sacrifice yield. The ee of the product is highest at low conversion and falls as the slow enantiomer starts to react.
  • Worked numbers for s = 20: at c = 50%, ee(SM) = ee(P) ≈ 79%; at c = 60%, ee(SM) ≈ 98% and ee(P) ≈ 65%.
  • Sharpless epoxidation is an effective kinetic resolution of racemic secondary allylic alcohols; lipase-catalysed acylation and hydrolysis are the commonest laboratory resolutions.

Beyond the 50% limit:

  • Dynamic kinetic resolution (DKR): the substrate enantiomers interconvert rapidly (for example through an enolisable α-stereocentre) while one reacts selectively. All the material can be funnelled through the fast-reacting enantiomer, so yields up to 100% are possible. Noyori’s Ru–BINAP hydrogenation of α-substituted β-keto esters is the textbook case (two new stereocentres, one fixed by DKR).
  • Desymmetrisation: a meso or prochiral substrate has enantiotopic groups; converting one of them selectively (enzymatic hydrolysis of one ester of a meso diester, for example) gives a single enantiomer in up to 100% yield.

Memory Tricks

Don’t Confuse

ThisIs not this
Felkin product (polar FA: OR perpendicular)Chelation product (OR syn-coplanar with C=O in a chelate): usually the opposite diastereomer
Chiral auxiliary: stoichiometric, covalently attached, removed laterChiral catalyst: substoichiometric, not incorporated
Kinetic resolution: racemate, max 50% eachDynamic kinetic resolution: racemising substrate, up to 100%
Desymmetrisation: meso/prochiral substrateResolution: racemic substrate
Narasaka–Prasad: syn-1,3-diolEvans–Saksena: anti-1,3-diol
SAE: allylic alcohols only (Ti/DET)Jacobsen–Katsuki: unfunctionalised alkenes (Mn-salen)
Matched pair: selectivities reinforceMismatched pair: selectivities oppose
DIPAMP: P-stereogenicBINAP: axially chiral

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not taken from any CSIR NET paper. Use R = 8.314 J K⁻¹ mol⁻¹ and T = 298 K.

Level 1: Foundation

  1. In the Felkin–Anh model, where is the largest α-substituent placed relative to the carbonyl group?
  2. Which ligand class is present in AD-mix-β?
  3. What is the maximum theoretical yield of one enantiomer in a classical kinetic resolution?
Answers
  1. Perpendicular to the plane of the carbonyl, anti to the incoming nucleophile.
  2. (DHQD)₂PHAL, a dihydroquinidine-derived phthalazine ligand. AD-mix-α contains (DHQ)₂PHAL.
  3. 50% (without in-situ racemisation).

Level 2: Part B Style

Practice Q1. A β-hydroxy ketone is reduced (i) with Et₂BOMe followed by NaBH₄ and (ii) with Me₄N⁺ BH(OAc)₃⁻. The major products are, respectively,

(a) syn-1,3-diol and anti-1,3-diol (b) anti-1,3-diol and syn-1,3-diol (c) syn-1,3-diol in both (d) anti-1,3-diol in both

Answer: (a)

(i) Narasaka–Prasad: a boron chelate fixes a half-chair and external hydride gives the syn diol. (ii) Evans–Saksena: hydride is delivered intramolecularly from an acyloxyborohydride bound to the β-OH, giving the anti diol.

Practice Q2. A chiral aldehyde reacts with an achiral reagent with dr 4:1. An enantiopure chiral reagent reacts with an achiral analogue of the aldehyde with 25:1 selectivity. Assuming additive transition-state energy contributions, the matched and mismatched combinations give approximately

(a) 29:1 and 21:1 (b) 100:1 and 6:1 (c) 100:1 and 1:6 (d) 25:1 and 4:1

Answer: (b)

Matched: 4 × 25 = 100:1 (\( \Delta\Delta G^\ddagger \) 3.4 + 8.0 = 11.4 kJ mol⁻¹). Mismatched: 25/4 = 6.25:1 (8.0 − 3.4 = 4.5 kJ mol⁻¹), still in the reagent’s favour. Option (c) would require the substrate bias to exceed the reagent bias.

Practice Q3. Treatment of cyclohex-2-en-1-ol with mCPBA gives mainly

(a) the epoxide trans to the OH, because of steric hindrance (b) the epoxide cis to the OH, because of hydrogen bonding to the peracid (c) a 1:1 mixture of cis and trans epoxides (d) cyclohexane-1,2,3-triol directly

Answer: (b)

The Henbest effect: the allylic OH hydrogen-bonds to the peracid and delivers oxygen to its own face. Acetylating the OH removes the directing effect.

Practice Q4. Which statement about the Evans oxazolidinone boron aldol (Bu₂BOTf, i-Pr₂NEt, then an aldehyde) is correct?

(a) It forms an E-enolate and gives the anti aldol. (b) It forms a Z-enolate and gives a syn aldol whose absolute configuration is set by the auxiliary. (c) It is enantioselective because the boron reagent is chiral. (d) It gives a racemic syn aldol.

Answer: (b)

Dibutylboron triflate with a hindered amine gives the Z-boron enolate of the N-acyl oxazolidinone. A Zimmerman–Traxler chair gives syn relative configuration (simple diastereoselectivity), and the C4 substituent of the auxiliary selects one of the two syn enantiomers (induced diastereoselectivity, “Evans syn”). The boron reagent is achiral; the auxiliary supplies the chirality.

Level 3: Part C Style

Practice Q5. Two related α-alkoxy ketones are treated with MeMgBr in THF at −78 °C:

Substrateα-Protecting groupRatio X : Y
AOCH₂Ph (benzyl)95 : 5
BOSiPh₂tBu (TBDPS)20 : 80

X and Y are the same pair of diastereomeric tertiary alcohols in both experiments.

Which statements are correct?

  1. X is the chelation-controlled product and Y the Felkin (polar Felkin–Anh) product.
  2. For substrate A, \( \Delta\Delta G^\ddagger \approx 7.3 \) kJ mol⁻¹.
  3. The bulky silyl ether prevents the five-membered chelate.
  4. Substrate B shows a higher diastereoselectivity than substrate A.

(a) 1, 2 and 3 only (b) 1 and 3 only (c) 2 and 4 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 3 only
  1. Correct. The small, Lewis-basic benzyloxy group chelates Mg and gives X strongly; the bulky, weakly basic TBDPS ether cannot chelate, and the polar Felkin–Anh product Y takes over. Chelation and Felkin give opposite diastereomers.
  2. Correct. \( RT\ln(95/5) = 8.314\times298\times\ln 19 = 7.3 \) kJ mol⁻¹.
  3. Correct. Steric bulk and the reduced Lewis basicity of a silyl ether oxygen both disfavour chelation.
  4. Wrong. 80:20 is 4:1 (3.4 kJ mol⁻¹), lower than 19:1.

Practice Q6. A racemic secondary alcohol undergoes lipase-catalysed acylation with selectivity factor s = 20. The reaction is stopped at 60% conversion.

Which statements are correct?

  1. The recovered alcohol has an ee of about 98%.
  2. The ester product has an ee of about 65%.
  3. At most 40% of the starting material can be recovered as alcohol at this conversion.
  4. Running to 60% rather than 50% conversion improves the ee of both the recovered alcohol and the ester.

(a) 1, 2 and 3 only (b) 1 and 4 only (c) 2 and 3 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 3 only
  1. Correct. Solving \( 20 = \ln[(0.40)(1-\text{ee})]/\ln[(0.40)(1+\text{ee})] \) gives ee(SM) ≈ 0.977.
  2. Correct. Mass balance: ee(P) = (1 − c) ee(SM)/c = 0.40 × 0.977/0.60 ≈ 0.65.
  3. Correct. 40% of the material remains unconverted.
  4. Wrong. At 50% conversion both ees are about 79%. Going to 60% raises ee(SM) to about 98% but lowers ee(P) to about 65%, because more of the slow enantiomer is acylated.

Practice Q7. Consider the following statements about asymmetric catalysis.

  1. In Noyori’s hydrogenation of a racemic α-substituted β-keto ester, a single stereoisomer can be obtained in more than 50% yield because the α-stereocentre epimerises through the enol under the reaction conditions.
  2. Sharpless asymmetric epoxidation of (E)-2-hexen-1-ol with L-(+)-DET and D-(−)-DET gives the two enantiomeric epoxy alcohols.
  3. BINAP owes its chirality to a stereogenic carbon atom.
  4. A chiral catalyst of 50% ee always gives product of exactly half the ee obtained with the enantiopure catalyst.

Which are correct?

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2 and 4 only (d) 1, 2, 3 and 4

Answer: (a) 1 and 2 only
  1. Correct. This is dynamic kinetic resolution: fast racemisation of the substrate plus selective reduction of one enantiomer.
  2. Correct. The two tartrate enantiomers deliver oxygen to opposite faces of the alkene.
  3. Wrong. BINAP is axially chiral (hindered rotation about the binaphthyl bond); it has no stereogenic carbon.
  4. Wrong. Non-linear effects (positive or negative) are well documented; a linear relation is not guaranteed.

Quick Revision Table

IdeaFact
Asymmetric inductionexisting chiral element biases a new stereocentre; 1,2-, 1,3-, 1,4- by distance
Cram (open-chain)C=O anti to L; attack from the side of S
Felkin–AnhL perpendicular, anti to Nu; Bürgi–Dunitz approach past S
Polar Felkin–Anhelectronegative group acts as L
Chelation (Cram cyclic)small OR/NR₂ + Mg, Zn, Ti, Ce; anti-Felkin product
Bulky silyl etherblocks chelation; Felkin product
1,3-diolsNarasaka–Prasad syn; Evans–Saksena anti
Directed epoxidationHenbest (peracid, H-bond); VO(acac)₂; syn to OH
AuxiliariesEvans (Z-boron enolate → Evans syn), Oppolzer sultam, SAMP/RAMP, pseudoephedrine
SAETi(OiPr)₄, DET, TBHP; allylic alcohols; L-(+)-DET bottom face
ADAD-mix-α (DHQ)₂PHAL; AD-mix-β (DHQD)₂PHAL
HydrogenationKnowles Rh–DIPAMP (L-DOPA); Noyori Ru–BINAP
OtherCBS (ketones); Jacobsen–Katsuki (unfunctionalised alkenes); proline (aldol); MacMillan (iminium)
Double inductionmatched ≈ product of ratios; mismatched ≈ quotient
Kinetic resolutions = k_fast/k_slow; max 50%; ee(SM) rises with conversion
DKR, desymmetrisationup to 100% yield

Key Takeaways

What’s Next

Advertisement
Advertisement