Non-Linear Effects in Asymmetric Catalysis: A Useful … · Non-Linear Effects in Asymmetric Catalysis: A Useful Tool in Understanding Reaction Mechanisms ... phosphoramide ligands

Non-Linear Effects inAsymmetric Catalysis: A

Useful Tool in UnderstandingReaction Mechanisms

Group MeetingAaron Bailey12 May 2009

What is a Non-Linear Effect?In asymmetric catalysis, the ee (er) of theproduct can be predicted based on theenantiopurity of the chiral substrate (auxiliaryor ligand)

eemax= eeproduct/eeauxiliary

By using this equation a linear correlation isobtained (line a)

However, this equation cannot hold true for all reactions as aggregation of ligands occursfrequently to form homochiral and heterochiral (meso) complexes

When products are obtained in higher enantiopurity than predicted a (+)-NLE is observed(curve d & c)Likewise, if a lower enantiopurity is obtained that what is predicted a (–)-NLE is observed(curve e & b).

Kagan, H. B. et al. Angew. Chem. Int. Ed., 2009, 48, 456-494.Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357.

First Observed Reports

Asymmetric sulfideoxidation displayed(–)-NLE effectsdependent on purity andconcentration ligand

Curve A= stoichiometricamounts of reagents

Curve B= Sub-stoichiometric amounts(0.5 eq Ti)

Me

SMe

Me

SMe

O

Ti(Oi-Pr)4

DET, H2O, t-BuOOH

Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357

First Observed Reports

Asymmetric epoxidationdisplayed(+)-NLE effects

Ti(Oi-Pr)4

DET, H2O, t-BuOOH

O

OH

OH

Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357

Revision of Asymmetric ModelML*2 Model (Differences in Reactivity)

Metal complexes can (if enantioenriched) form homochiral systems(MLRR & MLSS) as well as heterochiral (meso) dimers (MLRS).

Revision of mathematical modeleemax≠ eeproduct/eeauxiliary must apply correction factor eeproduct=eemaxeeauxiliaryƒ

Ƒ=(1+β)/(1 + gβ) if fast equilibrium occurs between complexes

then β=

Where β is the relative amount of meso- to chiral-type complexes in solutionAnd g is the relative reactivity of the meso-type complex with respect to the chiral complex

Kagan, H. B. et al. J. Am. Chem. Soc. 1994, 116, 9430-9439.

Reservoir Model

eeeff= eeaux-αeeres/(1- α)

Note: Both models can operate independently but manyreactions display both such NLE’se.g. if a dimeric (meso) complex is completely inactive and anenergy sink ML2 equation simplifies toeeprod= ee0eeaux(1+β)

Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357

Higher Order Systems-ML3, ML4ML3 ML4

At least 4 possible structuresMLRRR MLRSR MLRSS MLSSSAll will be chiral

Homochiral complexgive greaterselectivity thatheterochiral complex

Heterochiralcomplex giveshigher selectivityand is morereactive

Five possible structure:MLRRRR, MLRRRS, MLRRSS, MLRSSS, MLSSSS

Assumes statistical distribution between ligands

Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357

Asymmetric Sulfide Oxidation

Me

SMe

Me

SMe

O

Ti(Oi-Pr)4

DET, H2O, t-BuOOH

From experiment Calculated data overlay with experimental

Observing NLE’s and using derived equations a mechanistic hypothesis can be proposed:Data consistent with ML4 system

Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357

Asymmetric EpoxidationTi(Oi-Pr)4

DET, H2O, t-BuOOH

O

OH

OH

Active catalyst in solution

Ligand “reservoir” or inactive dimer

Ti

O

Ti

O

O

CO2Et

CO2Et

O

OR

RO

OROR

CO2Et

CO2Et

meso

Ti

O

Ti

O

O

CO2Et

CO2Et

O

OR

RO

OROR

CO2Et

CO2Et

Homochiral dimer

Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357Sharpless, K. B., et al. J. Am. Chem. Soc. 1991, 113, 106-113.Sharpless, K. B., et al. J. Am. Chem. Soc. 1991, 113, 113-126.

Proposing Mechanisms using NLEs

A large (+)-NLE was observed using catalyticamount of DAIB ligand

A 1:1 ligand:Zn results in no alkylationOptimal reaction conditions used 1:2stoichiometry

Propose a mechanism that accounts for theobserved NLE.

Which model does it fit? MLn? Reservoir? Orboth?

O

H ZnEt2

NMe2

OH

Et

OH

toluene, 0 º, 6 h

97 %,

99:1 er

*

Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357.Noyori, R. et. al. J. Am. Chem. Soc. 1989, 111, 4028-4036.Noyori, R. et. al. J. Am. Chem. Soc. 1986, 108, 6071-6072.

Mechanistic Proposal I

Based on early kinetic, NMR, and stoichiometric studies the mechanism above was proposedThis does not necessarily account for the dramatic NLE observedNoyori, R. et. al. J. Am. Chem. Soc. 1989, 111, 4028-4036.

Reservoir Effect in ZnEt2 Additions

Homochiral and Heterochiral dimers forming in solutionExperimental observations sugges KDHomo=3x10–2

KDMeso=1x10–5

MO calculations predict meso complex energetically favored by 3.1 kcal/mol

Noyori, R. et. al. J. Am. Chem. Soc. 1995, 117, 6327-6335.

Absence of NLE in ZnEt2 AdditionsStudy wanted to examine stereochemical controlling elements offerrocene ligands.

Proposed TS

Due to linear correlation between ee ofligand and product, authors could ruleout possibility of inactive aggregates(MLn) or reservoir effect

Bolm, C. et al. J. Org. Chem. 1998, 63, 7860-7867.

Other NLE in Organometallic Rxns

H

OOMe

TMSO

OH

N

HO

Me

CO2H

MeO

1. L1/A1/Ti(OiPr)4

2. TFAO

O

Ph

L1

A1

O

BnOH

OTMS

StBu

N

ON

N

O

Cu

2+

2 SbF6–

CH2Cl2

BnO

OH O

StBu

Ph Ph

Both examples show significant (+)-NLE’s

Elucidation of the mechanisms was achieved by observing NLE

Both are thought to proceed via the reservoir model- meso-Dimer favored by 2.9 kcal/mol

Ding, K. et al. Chem. Eur. J. 2002, 8, 5033-5042.

Evans, D. A. et al. J. Am. Chem. Soc. 1999, 121, 669-685.

NLE in Organocatalytic Reactions

H

O

Me

O

N

H

Me

Me

Me

Me

O

H

O

Three possible roles of catalystsJørgensen, K. A. et al. J. Org. Chem. 2003, 68, 4151-4157.

(–)-NLE Provides Mech. InsightsPresence of (–)-NLE indicates more thanone molecule of chiral amine present instereodefining step

Eq 5 is proposed mechanism with iminiumintermediate and enamine formation occurring

Matched-Mismatched case proposed to beresponsible for lower selectivity

N

Me

Jørgensen, K. A. et al. J. Org. Chem. 2003, 68, 4151-4157.

Juliá-Colonna EpoxidationO

H2O2, NaOH

poly-L-leucine

OO

Originally proposed catalytic pocket shownKelly, D. R. et al. Chem. Comm. 2004, 2021-2022.Kelly, D. R. et al. Chem. Comm. 2004, 2018-2020.

(+)-NLE Observed for PLLBased on ML2 system described previously, assuming statistical mixture of monomer and dimeramino acid residues as well as the heterodimer chains are not-reactive the following equationcan be deduced

Kelly, D. R. et al. Chem. Comm. 2004, 2021-2022.Kelly, D. R. et al. Chem. Comm. 2004, 2018-2020.

Revision of Mechanism

5 Leucine residues matched theamount of asymmetric amplificationobserved as well as the rate (amount ofactive catalyst)

Stereocontrolling element

H-bonding assists epoxide formation

Kelly, D. R. et al. Chem. Comm. 2004, 2021-2022.Kelly, D. R. et al. Chem. Comm. 2004, 2018-2020.

Asymmetric Allylations

Kinetic data coupled with NLE led to conclusion that twophosphoramide ligands present in stereodetermining step

Mechanistic experiments utilizing bisphosphoramideligands further supported hypothesis of two ligandspresent in stereodefining step

Denmark, S. E. et al. J. Am. Chem. Soc. 2000, 122, 12021-12022.Denmark, S. E. et al. J. Org. Chem. 2006, 71, 1513-1522.

Asymmetric Openings of meso-Epoxides

Observation of (–)-NLE along with kinetic datasuggests two phosphoramides present inmechanismWas able to rule out reservoir effect and productinhibition

Bisphosphoramide catalysts did not improveselectivities

Rate=–dprod/dt=K[HMPA]2[epoxide]

Denmark, S. E. et al. Adv. Syn. Cat. 2007, 349, 567-582.

Proposed Mechanism

Denmark, S. E. et al. Adv. Syn. Cat. 2007, 349, 567-582.

False Positives in NLE Systems

Initial studies by Kaganled to the conclusionthat a (–)-NLE waspresent

Recent studies have concluded the absenceof a NLE

Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357List, B. et al. J. Am. Chem. Soc. 2003, 125, 16-17.List. B. et al. J. Am. Chem. Soc. 2003, 125, 2475-2479.

Solubility Differences

Homochiral dimers less stabilized by H-bonding compared to HeterochiralBut how does this explain observed (–)-NLE?

Hayashi, Y. et al. Angew. Chem. Int. Ed. 2006, 45, 4593-4597.

Method of Amino Acid AdditionIn experiments, L-Proline and D-Proline were added separately causing the solids to have aconglomerate equilibrium in the solvent

Blackmond, D. G. et al. Angew. Chem. Int. Ed. 2006, 45, 7989-7992Kagan, H. B. et al. J. Am. Chem. Soc. 1986, 108, 2353-2357

Autoinduction/AutocatalysisA + B

A + B

Cat. P

Cat.-P P

OEt

OEtH

O O

Et

L (5 %)

ZnEt2 (10 %)

MS, toluene

O

EtOEt

EtO

OH

Trost, B. M. et al. J. Am. Chem. Soc. 2004, 126, 2660-2661.

L=

NLE and Autocatalysis- A BriefIntroduction

N

N

H

O

Zn

N

N

OH

N

N

OH

(0.0005 % ee)

99 % ee

Based on kinetic data observed adimeric complex is believed to bethe mechanism; however, thishypothesis is speculative anddebate is ongoing

Soai, K. et al. Tet. Asymmetry, 2001, 12, 1965-1969Soai, K. et al. Nature, 1995, 378, 767-768Soai, K. et al. Chirality, 2002, 14, 548-554.

Conclusions• The observation of NLE’s can have significant mechanistic

implications

• Careful analysis of the effect coupled with kinetic data,theoretical calculations, and isolation of intermediates can helpelucidate mechanisms

• The fundamental theory behind NLE’s has contributed to muchof the debate centered around asymmetricamplification/autocatalysis and the evolution of chirality theories