Advances in Homogeneous and Heterogeneous Catalytic ...

Advances in Homogeneous and Heterogeneous Catalytic AsymmetricEpoxidation

Q.-H. Xia,*,† H.-Q. Ge,† C.-P. Ye,† Z.-M. Liu,‡ and K.-X. Su§

Laboratory for Advanced Materials and New Catalysis, School of Chemistry and Materials Science, Hubei University, Wuhan 430062, China,Laboratory of Natural Gas Utilization and Applied Catalysis, Dalian Institute of Chemical Physics of Chinese Academy of Sciences, Dalian 116023,

China, and Jingmen Technological College, Jingmen 448000, China

Received June 30, 2004

Contents1. Introduction 16032. Homogeneous Systems 1605

2.1. Sharpless Systems 16052.1.1. Chiral Titanium Catalysts 16052.1.2. Chiral Vanadium Catalysts 1607

2.2. Porphyrin Systems 16092.2.1. Iron Porphyrins 16092.2.2. Ruthenium Porphyrins 16112.2.3. Manganese Porphyrins 16122.2.4. Molybdenum Porphyrins 1613

2.3. Salen Systems 16142.3.1. Manganese and Chromium Salens 16142.3.2. Cobalt Salens 16172.3.3. Palladium Salens 16172.3.4. Ruthenium Salens 1617

2.4. BINOL Systems 16182.4.1. Lanthanum BINOLs 16182.4.2. Ytterbium BINOLs 16192.4.3. Gadolinium and Samarium BINOLs 16202.4.4. Calcium BINOLs 1620

2.5. Chiral Carbonyl Compound Systems 16212.5.1. Simple Chiral Ketones 16212.5.2. Polyhydric Compounds Derived Ketones 16242.5.3. Chiral Aldehydes 1626

2.6. Chiral Iminium Salts 16272.7. Other Homogeneous Systems 1628

2.7.1. Mo-peroxo Complexes 16282.7.2. Lithium Complexes 16292.7.3. Magnesium Complexes 16292.7.4. Zinc Complexes 16302.7.5. Ruthenium Complexes 16312.7.6. Methyltrioxorhenium (MTO) 16322.7.7. Nickel Complexes 16322.7.8. Sulfonium Ylides 1632

3. Heterogeneous Systems 16353.1. Supported Sharpless Systems 1635

3.1.1. Insoluble Polymer-Supported TitaniumComplexes

1635

3.1.2. Organic−Inorganic Hybrid-supportedTitanium Complexes

1636

3.1.3. Silica-Supported Tantalum Complexes 16373.1.4. Silica-Supported Casein−Cobalt

Complexes1638

3.2. Supported Porphyrins Systems 16383.3. Supported Salen(Metal) Systems 1639

3.3.1. Polymer-Supported Salen Complexes 16393.3.2. Silica-Supported Salen Complexes 16423.3.3. MCM-41-Supported Salen Complexes 16423.3.4. Helical Polymer-Supported Salen

Complexes1645

3.3.5. “Ship-in-a-bottle” Trapped SalenComplexes

1645

3.3.6. Dendrimer-Supported Salen Complexes 16473.4. Phase Transfer Catalysis Systems 1647

3.4.1. Conventional Phase Transfer Catalysis(PTC)

1648

3.4.2. Fluorous Biphase Systems (FBS) 16523.5. Other Heterogeneous Systems 1653

3.5.1. Supported BINOL Systems 16533.5.2. Polyamino Acids Catalysis Systems 16533.5.3. Polymer-Supported Amino Acid Cu(II)

Complexes1654

3.5.4. Nanocrystalline MgO 16543.5.5. TBHP+KF/Alumina 1655

4. Conclusions 16555. Appendix: A Cross-Referenced Table 16566. References 1656

1. Introduction

Barry Sharpless was awarded the 2001 Nobel Prizeas a compliment to his great contributions in the fieldof asymmetric epoxidation, which has been recog-nized as one of the most important techniquesemerging in the last 30 years. These fundamentalfindings have largely expanded the scope of asym-metric synthesis and allowed a more targeted prepa-ration of pharmaceutical products, such as antibiot-ics, antiinflammatory drugs, and other medicines.

Chiral epoxides are very important building blocksfor the synthesis of enantiomerically pure complexmolecules, in particular, of biologically active com-pounds.1,2 Catalytic asymmetric epoxidation is anespecially useful technique for the synthesis of chiral

* Corresponding author. Phone and Fax: 0086-27-50865370; e-mail: [email protected].† Hubei University.‡ Dalian Institute of Chemical Physics of Chinese Academy ofSciences.§ Jingmen Technological College.

1603Chem. Rev. 2005, 105, 1603−1662

10.1021/cr0406458 CCC: $53.50 © 2005 American Chemical SocietyPublished on Web 04/19/2005

Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

compounds in both academia and industry becausea chiral catalyst molecule can act as an enzyme toinduce a million-level chiral product molecules.3 Thedevelopment of chiral catalysts capable of inducingasymmetric centers with high efficiency has alwaysbeen an important task for asymmetric synthesis.

The ability to produce desired organic compounds inenantiomerically pure forms from simple and readilyavailable precursors by using extremely smallamounts of chiral catalysts, generally with substrate/catalyst molar ratios of 100-1000 or higher, hastremendously practical implications.4,5 A variety ofcarbon-carbon double bonds, for example, those ofallylic alcohols, R- and â-unsaturated esters, andsimple alkenes, can be catalytically epoxidized withseveral metal catalysts.6-8

In 1965, Henbest et al. oxidized olefins with anenantiomeric excess (ee) of 10% using a chiral per-

Prof. Q.-H. Xia (Qinghua Xia) was born in 1965 in Hubei, China. Hegraduated from Department of Chemistry of Jingzhou Normal College in1984. From 1984 to 1987, he worked in Jingmen Petrochemical Complexof SINOPEC. In the autumn of 1987, he successfully passed the admissionexamination to enter Dalian Institute of Chemical Physics of ChineseAcademy of Sciences as a graduate of catalysis chemistry. There hepursued research in the fields of homogeneous and heterogeneouscatalysis and zeolite synthesis under the supervision of Prof. L.-B. Zheng,Prof. G.-Q. Chen, Prof. G.−W. Wang, and Prof. Q.-X. Wang and receivedhis M.Sc. degree in 1990 and Ph.D. degree in 1993. From 1993 to 1997,as a postdoctoral fellow he worked in Fudan University under thesupervision of Prof. Z. Gao and at the University of Tokyo under thesupervision of Prof. T. Tatsumi. In October 1995, he was promoted toAssociate Professor of Chemistry in Fudan University. In 1996, he workedfor NEDO of Japan as an Advanced Technology Research Scientist. From1997 to 2003, NUS, CPEC, and ICES of Singapore employed him as aResearch Fellow. In 1999, he and his family became permanent residentsof Republic of Singapore. From 2002, he started his new career as aScholar-level professor and an institute director in Hubei University tobuild up a Laboratory for Advanced Materials and New Catalysis. From2004, he was appointed as an adjunct professor by DICP and JingmenTechnological College. His research interests are mainly focused on thesynthesis and application of zeolites, porous molecular sieves, nanoma-terials and membranes, C1 chemistry, homogeneous and heterogeneouscatalysis, asymmetric hydrogenation and epoxidation, partial oxidation overTi-zeolites and others, environmental catalysis, etc.

Mr. H.-Q.Ge (Hanqing Ge) was born in 1977 in Hubei, China. He receivedhis Bachelor degree from School of Chemistry and Materials Science ofHubei University in 2002. Currently, he is in the third year of his Masterstudies on the asymmetric epoxidation of olefins on organic−inorganichybrid catalyst, under the supervision of Prof. Q.-H. Xia and Prof. C.-P.Ye.

Prof. C.-P. Ye (Chuping Ye) was born in 1952 in Hubei, China. Hegraduated from Department of Chemistry of Hubei University in 1975.Since then, he has worked in Hubei University to teach and performscientific research in organic chemistry. He was promoted to AssociateProfessor of Organic Chemistry in 1993 and to Professor of OrganicChemistry in December 2004. His research interests include appliedorganic chemistry, organic functional materials, adhesives, synthesis offine chemicals, etc.

Prof. Z.-M. Liu (Zhongmin Liu) was born in 1964 in Henan, China. Hegraduated from Department of Chemistry of Zhengzhou University in 1983.Then he continued his postgraduate and Ph.D. programs in the field ofcatalysis under the supervision of Prof. L.-B. Zheng, Prof. J. Liang, Prof.G.-Q. Chen, and Prof. G.-Y. Cai in Dalian Institute of Chemical Physicsof Chinese Academy of Sciences. He received his M.Sc. degree in 1986and Ph.D. in 1990 from DICP. In 1990, he started his scientific researchcareer in DICP, where he was promoted to Associate Professor in 1994and to Full Professor in 1996. In 1996, he worked in CNRS of France asa postdoctoral fellow. From 2000, he was appointed as an adjunctprofessor by Jilin University and Zhengzhou University. Currently, he is abig group leader, an assistant director of DICP, and an editorial boardmember for Chinese Journal of Catalysis and Journal of Natural GasChemistry. He was an international scientific committee member of 13thIZC and 14th IZC. His research interests are in the areas of heterogeneouscatalysis, C1 chemistry, porous zeolites and molecular sieves, natural gasutilization and applied catalysis, etc.

1604 Chemical Reviews, 2005, Vol. 105, No. 5 Xia et al.

Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

oxoacid as the oxidant, which was the first researchof asymmetric epoxidation.9 In the early 1980s,Katsuki and Sharpless made a milestone discovery,by which the enantiomers of an epoxide could beproduced efficiently and predictably from an allylicalcohol with a catalytic system consisting of diethyltartrate (DET), titanium tetraisopropoxide [Ti(O-Pri)4], and tert-butyl hydroperoxide (TBHP).10 There-after, much important progress has been madetoward the asymmetric epoxidation of other classesof alkenes.

In homogeneous catalytic systems, the reactionscan be efficiently carried out with high yields andee’s, but lengthy purification procedures are inevi-table and generally the expensive catalysts are quitedifficult to recover and recycle. However, in hetero-geneous systems, the reaction mixture can be easilyseparated from one another by simple filtration, andin some cases the catalysts can be reused severaltimes, thus bestowing on heterogeneous systemsmore convenient potential applications in industry.Since the solid-phase synthesis of oligopeptides wasfirst reported by Merrifield,11 many groups haveextensively developed various classes of heteroge-neously asymmetric epoxidation catalysts, such asinsoluble and soluble polymer-supported,12-22 organic-inorganic hybrid-supported,23-31 dendrimer -support-ed catalysts, etc.32-34 As compared with homogeneouscatalysts, heterogeneous ones also possess somedisadvantages, such as low reactivity and difficultpreparation, thus leading to the appearance of somenew catalysts combining homogeneous and hetero-geneous advantages, including “ship-in-a-bottle”trapped catalysts,35,36 fluorous biphase catalysts,37,38

room-temperature ionic liquids, etc.39

This review will systematically summarize theprogress of homogeneous and heterogeneous catalyticchiral epoxidation in the past decades and grant aconvenient and wide vision for the understanding ofpeers working in the area of asymmetric catalysis.

2. Homogeneous Systems

2.1. Sharpless SystemsAs shown in Scheme 1, four epoxides (racemic 2

and 3) from geraniol 1 are possibly made throughimproving either regio- or chemoselectivity, while theformation of each of the individual enantiomersrequires enantioselectivity.40 In 1957, Henbest et al.reported that the electronic deactivation at C1 coor-dinated by oxygen substituent caused the selectivecoordination of peroxoacids to the 6,7-position doublebond, leading to racemic 3.41 In 1973, Michaelson etal. cracked the regioselectivity problem representedby the epoxidation of geraniol. Early epoxidation ofgeraniol 1 catalyzed by transition metals with alkylhydroperoxides was highly selective for the 2,3-position compounds (racemic 2).42 In 1980, Katsukiet al. discovered titanium-catalyzed asymmetric ep-oxidation of olefins bearing allylic hydroxy groupsinto either 2 or ent-2, thereby solving one side of theproblem in enantioselectivity.10

The resolution of the problem in regioselectivityand enantioselectivity aroused the interest of manygroups in the research of asymmetric epoxidation. In1980s, Sharpless et al. first introduced chiral ligandsto transition metals Mo, Ti, and V, further developinga Sharpless asymmetric epoxidation system. Theysuccessfully converted allylic alcohols into asym-metric epoxides in high chemical yields with morethan 90% ee, under the catalysis of a transition metalcatalyst Ti(OPri)4 by using TBHP as the oxidant witha chiral additive DET. However, this system holds amain disadvantage, namely, a low turnover number,which stimulates interest in searching for moreefficient catalysts. For the epoxidation of allylicalcohols, many efficient systems have been reported,such as the catalytic systems consisting of vanadiumcatalysts and hydroxamic acids derivatives,43-47 butthe Sharpless protocol with titanium tetraisopro-poxide and tartrate esters ligands is still investigatedthe most.10,48

2.1.1. Chiral Titanium Catalysts

Over the past two decades, the use of titaniumalkoxide complexes as the catalysts for the asym-metric epoxidation of allylic alcohols has undergonea rapid expansion, especially since the discovery in1980 by Sharpless et al.10 The Sharpless epoxidationprocess seems to be well understood, in which theimpact of a chiral ligand on enantioselectivity is

A/Prof. K.-X. Su (Kexin Su) was born in 1949 in Hubei, China. Hegraduated from Department of Chemistry of Huazhong Normal Universityin 1977. From 1977 to 1984, he worked in Jingzhou Normal College asa Lecturer of Chemistry. Then he moved to Department of ChemicalEngineering of Jingmen Technological College where he presently works.He was promoted to Associate Professor of Chemistry in 1999. Hisresearch fields include organic synthesis, polymer materials, and chemicalprocess.

Scheme 1. Regio- and EnantioselectiveMonoepoxidations of Geraniola

a Adapted from ref 40 with permission. Copyright 2002 Wiley-VCH.

Catalytic Asymmetric Epoxidation Chemical Reviews, 2005, Vol. 105, No. 5 1605

Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

critical.49-51 Regardless of the metal-catalyzed meth-odology, the structure of hydroperoxides plays asignificant role in the level of enantioselectivityachievable for the asymmetric epoxidation, which hasresulted in the development of structurally diversealkyl hydroperoxides.52 In the early Sharpless epoxi-dation, the oxygen donor was achiral TBHP, andasymmetric induction resulted from the contributionof optically active tartrate with catalytic amounts aschiral auxiliary.10,50 We can imagine that if opticallyactive hydroperoxides are used, the hydroperoxideswill serve not only as the oxygen donors but also asthe sources of chirality without the need for tartrateas chiral auxiliary. When the optically active hydro-peroxide as the oxygen donor is used to provide thechiral environment, achiral diol ligands must bechosen as additives to limit the coordination orienta-tion between alkene molecules and the oxygen donorsduring the assembly of the loaded titanium complex.Generally, optically active hydroperoxides were ob-tained either by 1O2 photooxidation of thiazolidinederivatives53 and allylstannanes54 or by oxidation ofunsaturated glycosides.55 Under titanium catalysis,first attempts of the epoxidation of allylic alcoholswith chiral hydroperoxides only afforded ee’s of lessthan 20%, but sugar-derived hydroperoxides showedhigher enantioselectivities (up to 50% ee).56,57

In 1997, Adam et al. reported Ti-mediated asym-metric epoxidation of a variety of prochiral allylicalcohols with optically active hydroperoxides (4, 5,and 6b) as oxygen donors and multidentate ligandsas achiral additives, to achieve up to 50% ee in goodyields (63-97%).52 In the same year, Lattanzi et al.reported the application of several furylhydroperox-ides 6 in the Sharpless asymmetric epoxidation ofallylic alcohols.58 The results showed that the enan-tioselectivity was strongly dependent on the substitu-tion near the reactive sites, consistent with theobservations made by Corey59 and Sharpless et al.49,50

Tertiary or alkyl hydroperoxide has been proven toensure high ee’s.52,60,61 Anyway, an intrinsic limita-tion of this system is imposed by the essentiality ofcoordinated functional groups on the substrate toattain a high enantioselectivity. In 2002, Lattanzi etal. made a further finding that a high enantioselec-tivity could be obtained in the asymmetric epoxida-tion of allylic alcohols with tertiary furylhydroper-oxide 6a as shown in Table 1.62 Entries 1 and 4obtained high ee’s with the addition of stoichiometric

quantity of chiral catalyst, while the use of a subs-toichiometric amount of chiral catalyst in entries 2and 5 resulted in a slight drop in the enantioselec-tivity.

In 2003, Lattanzi et al. first synthesized an enan-tiopure tertiary hydroperoxide 7 from camphor via astereospecific nucleophilic substitution of hydroxylgroup bound to the chiral carbon center of the alcoholmolecule by H2O2 (unlike TADOOH), in which thereactive peroxo (-OOH) group is directly bound tothe stereogenic carbon center.63 However, the so-prepared chiral tertiary hydroperoxide only offereda moderate ee up to 46% and a yield up to 59% forthe asymmetric epoxidation of allylic alcohols. It isnoteworthy that the chiral tertiary alcohol can beisolated at the end of epoxidation and then can berecycled for the synthesis of the hydroperoxide, thusproviding a valuable chiral resource-saving protocol.However, optically active hydroperoxides as oxygendonors in the presence of multidentate diols (asachiral ligands) always afford lower enantioselectivi-ties than the Sharpless modus operandi of employingTBHP as an achiral oxygen donor and C2-symmetrictartrate as chiral auxiliary.

In 2003, Perez et al. reported the synthesis of afamily of titanium(IV) alkoxide compounds, i.e., [Ti-(OPri)3(OR)]2 [OR ) AdamO (8), DAGPO (9), DAGFO(10), or MentO (11)], [Ti(OPri)2(OR)2]2 [OR ) AdamO(12), DAGPO (13), DAGFO (14), or MentO (15)], andTi(OR)4 [OR ) AdamO (16), DAGPO (17), DAGFO(18), or MentO(19)].64 These preparations could beperformed by using two different routes: (a) throughthe metathesis reaction of TiCl(OPri)3 and TiCl2-(OPri)2 with ROH in the presence of Et3N and (b)alternatively through the alcohol exchange of Ti-(OPri)4 with a high boiling-point alcohol ROH, whereRO ) adamantanoxi (AdamOH), 1,2:3,4-di-O-isopro-pylidene-R-D-galacto-pyranoxi (DAGPOH), 1,2:5,6-di-O-isopropylidene-R-D-glucofuranoxi (DAGFOH), or(-)-(1R,2S,5R)-menthoxi (MentOH). Generally, thesolution state of most Ti(OR )4 (OR ) OMe, OEt,OBun) appears to be an equilibrium among mono-,di-, and trinuclear species, and at room temperaturetrinuclear species are dominant; however, with theincrease of steric bulk of an alkoxide ligand such asOR ) OPri, OBut, monomeric species are predomi-nant. Sugars are natural chiral ligands and canreadily assemble transition metal complexes to forma receptor cavity with three-dimensional structuralcharacteristics. More importantly, the entity of thus-formed molecules possesses the intrinsic propertiesof both sugars and metal complexes. Some of thesechiral titanium(IV) Lewis acid compounds, e.g., 13

Table 1. Asymmetric Epoxidationa of Allylic AlcoholsCatalyzed by 6a

a For experimental details, see ref 62. Adapted from ref 62with permission. Copyright 2002 Elsevier. b Stoichiometricconditions. c Catalytic conditions.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

and 14 derived from diacetone galactose and diac-etone glucose, have been applied in the asymmetricepoxidation of cinnamyl alcohol to receive goodcatalytic activity and stereoselectivity (65% yield with22% ee for 13 and 60% yield with 17% ee for 14). Onthe basis of solid-state NMR and variable tempera-ture NMR experiments, they assumed the existenceof low-level substituted dimeric titanium compoundsTi(OPri)4-n(OR)n (n ) 1, 2) and highly substitutedmononuclear Ti(OR)4 in the solution and solid statesdue to the steric hindrance imposed by bulky alkoxideligands with a nuclearity greater than 8.65

2.1.2. Chiral Vanadium CatalystsA number of asymmetric epoxidations based on

chiral vanadium (V) complexes with TBHP have beenreported,4,43,45-47,66-69 in which the epoxidation ofallylic alcohols is the most well-known.42,66,67,69 Va-nadium(V) alkylperoxo complexes were widely ac-cepted as intermediates in the catalytic systems ofVO(acac)2/TBHP42,70-72 and VO(acac)2/ligand/TB-HP.46,73,74 Prior to the appearance of titanium-basedcatalysts, the Sharpless group had first developedasymmetric epoxidation (50% ee) catalyzed by thechiral vanadium hydroxamate complex 20.43 A fewyears later, they found again that when using pro-line-derived hydroxamic acid as the chiral ligand, upto 80% ee could be achieved.72 Additionally, in 1999Yamamoto et al. successfully applied a new chiralhydroxamic acid 21, derived from 2,2′-binaphthol,serving as monovalent ligand coordinated with avanadium complex in the asymmetric epoxidation ofallylic alcohols to obtain an ee up to 94%.45 To explorethe potentials of these new catalysts further, theyextensively tested the epoxidation of various substi-tuted allylic alcohols. The results showed that theepoxidation of 3,3′-disubstituted allylic alcohols cata-lyzed by VO(OPri)3 and 21c proceeded smoothly toyield the corresponding epoxides in moderate to goodyields (70�87%) with mediocre ee’s (41�78% ee).However, in the case of 2,3-disubstituted allylicalcohols, high ee around 90% ee was obtained,irrespective of bearing aromatic groups. It should benoted that unlike titanium tartrate system, molec-ular sieves to sequester water, which has a deleteri-ous effect on both the rate and selectivity, were notrequired in this case.49,75 It seems that severalcharacteristics of chiral vanadium complex played animportant role in increasing the rate and enantiose-lectivity, i.e., the starting oxidation state of vana-dium, the coordination ability of hydroxamic acids,and π-interaction or steric repulsion between the

metal-binding site and oxidant. To improve theefficiency of vanadium-based catalysts, Yamamoto etal. in 2000 developed a family of chiral hydroxamicacid ligands with a structure similar to the complex22,46 and found that the product selectivity increasedgradually with an increase of the steric hindrancein the side chain of amino acid, in which the bestresult was achieved when using tert-leucine-derivedhydroxamic acid as the ligand. When L-R-amino acidwas partly changed into imido group, the best result(87% ee) was achieved for the epoxidation of 2,3-diphenyl-prop-2-en-1-ol with 1,8-naphthalene-dicar-bonyl-protected hydroxamic acid. Whereas for thesame reaction, if aryl group near the metal-coordi-nated site was changed, the highest ee value couldreach 96% when using N-bis(1-naphthyl) methyl-substituted hydroxamic acid as the ligand.

The asymmetric epoxidation of homoallylic alcoholswas difficult to conduct with the other metal-basedcatalysts reported previously.76-78 However, vana-dium complexes could effectively catalyze the occur-rence of this reaction to yield the corresponding epoxyalcohols with good to high stereoselectivities.79 Yama-moto et al. prepared a chiral vanadium complex 23from vanadium triisopropoxide and an R-amino acid-based hydroxamic acid, which was a rather efficientcatalyst for the epoxidation of disubstituted allylicalcohols with up to 96% ee in high yields (almost>95%).45,46,78 In 2003, they reported again that thischiral catalyst could also be used for the asymmetricepoxidation of 3-monosubstituted homoallylic alco-hols with high enantioselectivities (84-91% ee) inmoderate yields (42-89%),80 as shown in Table 2.

Meanwhile, they observed that cumene hydroperox-ide (CMHP) as the oxidant benefited a high enanti-oselectivity, while TBHP led to an extremely lowenantioselectivity <8%.

In 2000, Bolm et al. first reported the use of avanadium-based catalyst bearing the paracyclophaneplanar-chiral ligand 24 in the asymmetric epoxida-tion of allylic alcohols.47 Even simple TBHP could beused as a terminal oxidant to yield epoxy alcohols

Table 2. Asymmetric Epoxidation of HomoallylicAlcohols Using 23 as Liganda

a Adapted from ref 80 with permission. Copyright 2003Wiley-VCH.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

with 71% ee for the epoxidation of (E)-2-methyl-3-phenyl-2-propen-1-ol with 24c. To identify the opti-mal ligand structure with respect to the substituentR at the nitrogen of 24, they tested the epoxidationof (E)-2-methyl-3-phenyl-2-propen-1-ol with VO-(OPri)3 and CMHP in toluene and found that all theligands 24 resulted in the corresponding epoxides inhigh yields >86% with moderate enantioselectivities32-52% ee. Interestingly, in contrast to Yamamoto’sresults with the ligand 21, bulky aliphatic substitu-ents rather than aromatic groups at the nitrogen of24 remarkably increased the enantioselectivity, andthe S-configured ligand led to an (S,S)-configuredepoxide as a predominant enantiomer.

In 2002, Wu et al. reported the synthesis andapplication of new vanadium catalysts bearing (+)-ketopinic acid-based chiral hydroxamic acid ligands25 and 26 for the asymmetric epoxidation of allylicalcohols.81 Owing to poorly enantiomeric selectivityof the chiral hydroxamic acid ligand 25 derived from(+)-ketopinic acid 27 with different substituents onthe nitrogen, its structure had to be modified throughincorporating bulkier substituents at C2 of the bor-nane skeleton so as to acquire a high enantioselec-tivity. This is a two-pronged strategy to study theeffect of an N-substituent in a modified steric envi-ronment of the bornane moiety on the selectivity andalso the impact of the size of the C2 substituent ofbornane on the selectivity. Chiral hydroxamic acid26 with required structural features was preparedstarting from (+)-ketopinic acid 27. Moderate to highee values (up to 89%) were obtained for the asym-metric epoxidation of allylic alcohols with vanadium-based catalysts coordinated by readily accessible newchiral hydroxamic acid ligands having diverselystructural features, accompanied by an importantcharacteristic that increasing the bulk of the N-substituent retarded the selectivity (Table 3).

In 2003, Bryliakov et al. carried out the oxidationof (-)-(R)-linalool with TBHP to form (2S,3R)- and(2R,3R)-configured monoepoxides with moderate di-astereomeric excess (de) in the presence of a vana-dium(V) complex derived from the interaction of[VO(acac)2] or [VO(OBun)3] with a new chiral pyra-zolylethanol ligand 28.9 Normally, this reaction pro-ceeded slowly at room temperature (only conversionof 55% within 49 h); however, the coordination ofchiral ligand 28 to the active center enhanced thereaction rate considerably with an improved diaste-reoselectivity, for which the maximal de value ofabout 56% was acquired at 20 °C in toluene underoptimal conditions. Adam et al. also reported thevanadium(V)-catalyzed asymmetric epoxidation ofprimary allylic alcohols with optically active TAD-DOL-derived hydroperoxides 29 and 30 (as theasymmetric controller) to obtain up to 72% ee.82 Afterthe completion of reaction, the optically active TAD-DOL could be completely recovered without any lossof enantiomeric purity, which could provide theopportunity of regenerating the chiral oxygen source.This demonstrated the effect of a hydrogen-bondedtemplate on the enantiofacial control in the vanadium-catalyzed epoxidation with a chiral hydroperoxide,in combination with an achiral hydroxamic acidligand. This novel finding has greatly stimulated thedevelopment of more effective hydroxy-functionalizedhydroperoxides for the vanadium-catalyzed asym-metric epoxidation. The initial examples of thetitanium-catalyzed asymmetric epoxidation of allylicalcohols with optically active sugar-derived52,57 orsimple secondary hydroperoxides56 achieved ee valuesof up to 50%. Such hydroperoxides have been recentlyutilized in the asymmetric Weitz-Scheffer epoxidationof R,â-enones to obtain high ee values of up to 90%.83

However, when sterically TADDOL-derived hydro-peroxide TADOOH was used, the highest enantio-selectivity of 98% ee was achieved in the Weitz-Scheffer epoxidation, the Baeyer-Villiger oxidation,and sulfoxidation.84 Also in 2003, TADOOH wassuccessfully used as a chiral oxygen source in the

Table 3. Asymmetric Epoxidation of Allylic AlcoholsUsing 26e as Liganda

a Adapted from ref 62 with permission. Copyright 2002Elsevier.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

catalytic asymmetric epoxidation of allylic alcoholsby an oxovanadium-substituted polyoxometalate.85

2.2. Porphyrin SystemsSo far, the transition metal-catalyzed enantio-

selective epoxidation has been of utmost importanceand widely studied over the pastdecades.45-47,52,62,81,86-92 However, in practice usefulcatalysts should (a) be easily prepared and stored,(b) display high reactivity (turnover frequency),selectivity, and durability, and (c) contain inexpen-sive and environmentally benign metals in the coor-dination sphere by means of sterically and electroni-cally tunable chiral ligands.93-95 The role of porphyrinsis important in a bewildering array of proteins;96 theirfunctions include O2 storage (myoglobin Mb), O2transport (hemoglobin Hb), oxidation of inactivatedcarbon-hydrogen bonds (cytochrome P450), and oxy-gen reduction (cytochrome c oxidase). The diversityof functions of these natural hemoproteins is dictatedby many factors, such as the number and nature ofaxial ligands, the spin and oxidation state of themetal center, the nature of the polypeptide chain, andthe geometry of the porphyrin ring.

Chiral metalloporphyrins have constituted an im-portant class of catalysts for the asymmetric epoxi-dation of alkenes, which occurs basically via highlyreactive oxometal (MdO) intermediates.97 A rigidmacrocyclic core and alterable periphery of porphy-rins make them attractive templates for buildingasymmetric catalysts. Chiral groups have been at-tached to porphyrins in various geometries, aimingat systems that might give high enantioselectivitiesand turnover numbers. Most studies on metallopor-phyrin catalysts are confined to the porphyrin com-plexes of iron,98-102 manganese,103 ruthenium,104-106

and molybdenum.103,107 Several different strategieshave been adopted to append optically active groupsonto the macrocyclic ring of metalloporphyrins.108,109

Supplementary consideration lies in the nature ofoxidants and metals (Fe, Ru, and Mn, etc.).110 Ofthree metal complexes with the same chiral porphy-rin ligand, much better results were obtained withiron and ruthenium than with manganese. It has tobe pointed out that when 2,6-dichloropyridine-N-oxide (Cl2PyNO) was used as the oxidant, the samechiral porphyrin ligand resulted in different ee valuesfor stoichiometric and catalytic reactions, possiblydue to the co-interaction of oxidant, axial ligand, andoxygen transfer.111 In contrast, for the same reactionssimilar ee was obtained using oxygen or iodosylben-zene as the oxidant. Berkessel105 and Che et al.106,112,113

also described high efficiency of the catalyst systemsinvolving different chiral porphyrin ligands for theasymmetric epoxidation of unfunctionalized olefins,reported previously by Halterman.114 When using 2,6-dichloropyridine-N-oxide as terminal oxidant, the

epoxide of 1,2-dihydronaphthalene was obtained with77% ee and 90% yield.105

2.2.1. Iron PorphyrinsIn the presence of molecular O2, synthetic iron(II)

porphyrins, which are not imposed with protectionupon both faces, tend to be oxidized irreversibly toform µ-oxo Fe(III) dimers.115 To avoid this unfavor-able oxidation, several model structures with pro-tected faces have been developed.116-118 In 1999,Collman et al. presented a highly efficient catalystbased on a novel chiral iron porphyrin 31.119 Thissystem gave very high enantioselectivities and turn-over numbers for the epoxidation of styrene deriva-tives (83% ee for styrene, 88% ee for pentafluorosty-rene, and 82% ee for m-chlorostyrene) and non-conjugated terminal alkenes. More significantly, thissystem manifested unusual chiral induction for non-conjugated terminal olefins, such as 3,3-dimethyl-butene and vinyltrimethylsilane, in which the eevalues for these two olefins exceeded the highestvalues obtained from any catalytic systems reportedpreviously, including salen-Mn derivatives.120-122

In 2000, a new series of porphyrins 32 thatincorporate four identical chiral binaphthyl deriva-tives in the meso-positions was synthesized by Regi-nato et al.100 Four iron(III) chloro complexes (Fe-32)have been tested as catalytic precursors in theasymmetric epoxidation of styrene with iodosylben-zene as a single oxygen atom donor. The chemicalyield of epoxide was about 47%; however, the ee valuewas strongly dependent on the structure of thecatalytic precursor, ranging from 21% with thecompound 32a (RRRR) to 57% with the compound 32c(RRââ), in agreement with the results reported byCollman et al.109 In all the runs, absolute configura-tion of the products was (S), suggesting that thestereochemical outcome of this reaction was primarilydecided by the absolute configuration of the binaph-thyl moiety rather than by the overall moleculararrangement.

In 2000, Rose et al. developed the application ofchiral iron-porphyrin 33,123 a “seat” porphyrin ob-tained by condensing a chiral binap diacid chloride


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

with aminophenyl porphyrin; for the asymmetricepoxidation of various terminal olefins with iodosyl-benzene, especially for styrene and pentafluorosty-rene, 59% ee and 85% ee were, respectively, ob-tained.124 In the case of bis-binaphthenylporphyrin34, the ee values for styrene and pentafluorostyrenewere 83 and 88%, respectively,119 while the best ee(92%) was obtained from the catalytic epoxidation oftert-butylethylene. Furthermore in 2004, they syn-thesized a new chiral binaphthyl-strapped iron-porphyrin 35, which exhibited unprecedented cata-lytic activity toward the enantioselective epoxidationof terminal olefins.125 For the epoxidation of styrene,typical ee values were measured to be 90-97%,whereas the turnover numbers (TON) averaged16 000. The C2-symmetric binap-strapped porphyrinshave proven to be efficient for the epoxidation ofterminal olefins.119,123 The presence of two rigid binapwalls efficiently directs the approach of olefin to themetal center, induces a good transfer of asymmetry,and inhibits the oxidative degradation of the catalystas well as the formation of an unreactive µ-oxodimer.126 In addition, C2-symmetric porphyrins areeasily prepared from readily available R2â2-tetrakis-(O-aminophenyl) porphyrin. The chemical yieldsbased on the consumed PhIO varied from about 85%in most cases to 96% in the case of styrene. Remark-ably, very good enantioselectivity was maintainedwhen the reactions were carried out at room tem-perature, e.g., the epoxidation of styrene at roomtemperature afforded the desired epoxide in 94% ee,and the catalyst could be reused in the second run.

Groves et al. have carried out an investigation onthe asymmetric epoxidation of styrene to (R)-(+)-styrene oxide (48% ee), catalyzed by optically activeiron porphyrins such as FeT(R,â,R,â-binap)PPCl withiodosylbenzene.127 Also, iron complexes of a binaph-

thyl capped porphyrin were used by Collman et al.128

for the epoxidation of 4-chloro-styrene (50% ee) and4-nitrostyrene (56% ee). Naruta et al. reported thecatalytic performance of an iron metalloporphyrinwith rigid backbones of binaphthyl groups using 1,5-diphenylimidazole as the cocatalyst and PhIO as theoxidant.129 Thereafter, they improved the selectivityof epoxides by designing “twin coronet” porphyrinligands,130-133 in which each face of the macrocycleis occupied by two binaphthyl units. Because of thedifference of the binding mode of chiral auxiliary,there were two topologically “eclipsed” and “stag-gered” isomers with D2 symmetry. Iron(III) complexof the eclipsed isomer was observed to be an enan-tioselective catalyst, robust enough to exhibit higherthan 500 TON with 74-96% ee for chiral epoxidation.For those olefins bearing electron-withdrawing sub-stituents, the ee value was measured to be 74% forpentafluorostyrene, 89% for 2-nitrostyrene, and 96%for 3,5-dinitrostyrene, respectively, which was thehighest stereoselectivity reported for styrene deriva-tives catalyzed by chiral metalloporphyrins. Theligands are usually prepared through two routes, oneof which is attaching the chiral superstructure toeach of the four atropisomers of the parent tet-raarylporphyrin,134 the other modifying the substit-uents on a single atropisomer.135,136

In 2003, Smith et al. observed that the ee valuesinduced by chiral iron porphyrins 36-39 dependedon the structure of catalysts and substrates, thetemperature, and the solvent and that replacing a2,6-di(1-phenylbutoxy)-phenyl group by a pentafluo-rophenyl group led to a marked improvement in theefficiency and stability of catalyst.137 Generally, thepresence of a pentafluorophenyl group on the por-phyrin ring can increase the catalytic activity in threeways: (1) The electron-withdrawing fluorines acti-vate the electrophilic high-valent oxoiron intermedi-ate. (2) By removing electron density from themacrocyclic ring, the halogens make the porphyrinless susceptible to be attacked by the active oxidant,leading to less self-destruction of catalyst. (3) Sincea pentafluorophenyl group is considerably less bulkythan 2,6-di(1-phenylbutoxy)-phenyl not only does thisallow easier access of the substrate to the oxoironcenter during the catalytic cycle, but it also removesoxidizable 1-phenylbutoxy groups, retarding intramo-lecular destruction of catalyst. They likely playimportant roles in oxidations; however, it is difficultto determine which of three effects is the mostimportant. Iron(III) tetra (pentafluorophenyl) por-phyrin was a significantly better catalyst than thenonfluorinated analogue iron (III) tetraphenylpor-phyrin. Lowering the temperature from 20 to 0 °Cresulted in a higher epoxidation yield and a smallbut significant increase in the ee values. Althoughthe overall yields and ee values showed an obviousdependence on the solvent, the major enantiomerformed with each substrate/catalyst combination wasunaffected by variations. Interestingly, no singlesolvent was optimal for all three systems.

In 2003, Boitrel et al. further developed the ap-plication of four atropisomers of an L-prolinoyl picketporphyrin iron complex 40 in the asymmetric epoxi-


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

dation of p-chlorostyrene and 1,2-dihydronaphtha-lene, in which low ee values ca. 34% were obtained.102

The results show that even in a flexible structure,too much hindrance implies a negative consequenceon the selectivity of epoxidation. It is worthwhile tobe mentioned that the enantioselectivity obtainedwith picket porphyrins was as high as that withstrapped ones, and in some cases, even higher.

2.2.2. Ruthenium Porphyrins

Ruthenium complexes coordinated with either D4-chiral porphyrins or homochiral porphyrins havebeen reported.105,106,138,139 In some cases, these chiralmetalloporphyrins could catalyze the oxidation ofstyrene derivatives with good ee values but lowyields.104,111,138 More importantly, for identical chiralporphyrin ligands the ruthenium complexes weremuch better at inducing the asymmetry than werethe more commonly used iron, manganese, or rhod-ium derivatives.104,110,140-146 In 1996, Gross et al.reported the first example of epoxidation catalyzedby a homochiral ruthenium porphyrin 41a104 andobserved a notable effect of solvent on the enanti-oselectivity of chiral epoxy styrene, for which theenantioselectivity induced by benzene (44% ee) wasmuch higher than by dichloromethane (4% ee).Recently, there was renewed interest in the oxidationreactions catalyzed by ruthenium(II) porphyrin com-plexes and simultaneously in the development of newchiral ruthenium porphyrins.147,148 The catalytic reac-tions were mainly focused on asymmetric epoxidationof olefins,105,139 and in some cases a gradual deactiva-tion of catalyst was observed, possibly ascribed to theformation of inactive carbonyl complexes from trans-dioxo(tetramesitylporphyrinato) ruthenium (VI).149 In1999, new members 41a-c of D2-symmetric ruthe-nium porphyrins were reported,111 which showed tobe the most selective catalysts for the asymmetricepoxidation of terminal and trans-disubstituted ole-fins. The enantioselectivity obtained with 41b wasmuch higher than with 41a and resulted in theexploration of ruthenium porphyrin complex 41c,which was also based on the X-ray crystalline struc-tures of both 41a and its ruthenium complex104,150

and on the molecular modeling investigation of 41b.Indeed, 41c was a very good catalyst, with higherchiral induction for terminal and trans-olefins butsignificantly poorer chiral induction for cis-olefins.

The epoxidation of styrene and its m- and p-chloro-substituted derivatives proceeded with 79-83% ee.151

A strategy to catalyze highly enantioselective ep-oxidation of unfunctionalized trans-disubstituted alk-enes would be to prepare reactive, isolable, and chiralmetal-oxo complexes, which can act upon trans-alkenes to give epoxides with better enantioselectivi-ties than cis-counterparts.104 Once such complexesare obtained, the relationship of structure-enanti-oselectivity established by studying stoichiometricepoxidations of alkenes could assist future design forbetter metal catalysts. In 1999, Zhang et al. appliedD2-symmetric chiral trans-dioxoruthenium(VI) por-phyrins 41a, d, and e, bifacially encumbered by fourchiral threitol units, to catalyze the enantioselectiveepoxidation of trans-â-methylstyrene (70% ee) andcinnamyl chloride (76% ee).112

Berkessel et al. in 1997 synthesized an enantio-merically pure ruthenium carbonyl porphyrin 42awith very high yield by refluxing the correspondingporphyrin ligand with Ru3(CO)12 in phenol.105 Thecatalytic asymmetric epoxidation of olefins over 42awith 2,6-dichloropyridine N-oxide (2,6-DCPNO), whichwas carried out at room temperature in benzeneunder argon, afforded epoxides in good yields up to88% with up to 77% ee. In 2003, they reported againthe use of three enantiomerically pure and electroni-cally tuned ruthenium carbonyl porphyrin catalysts42b-d in the asymmetric epoxidation of a variety ofolefinic substrates.152 Introduction of a CF3 substitu-ent in the remote position resulted in an improvedstability, and turnover numbers up to 14 200 withan ee value of 80% were achieved for the epoxidationin benzene using 2,6-DCPNO as the oxygen donor(1.1 equiv to olefin) at room temperature.

Dioxygen is the most appealing since it is economi-cal and environmentally friendly,153,154 but reports onthe use of dioxygen in the enantioselective epoxida-tion of alkenes are quite sparse. Chiral Mn (â-


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

ketoiminato) complexes could catalyze aerobic asym-metric epoxidation of alkenes, but aldehyde wasrequired as a sacrificial reductant.89 In 1998, Lai etal. reported the first example of aerobic enantiose-lective epoxidation of alkenes catalyzed by dioxoru-thenium(VI) complex 43 with a D4-symmetric chiralporphyrin ligand, which did not rely on the use of aco-reductant.106,155 The complex exhibited catalyticactivity toward aerobic enantioselective epoxidationof prochiral alkenes with an enantioselectivity up to73% ee under an oxygen pressure of 8 atm. However,the catalytic oxygenation of styrene using air in oneatm pressure in CH2Cl2 gave very low turnovernumbers; even less than 8 atm pressure of oxygenstyrene oxide was obtained with only 10 TON and70% ee. But when the amount of catalyst wasstoichiometric, the attainable ee value was 77%.Lowering the temperature decreased the epoxideyield. For the asymmetric epoxidation of alkenes, theee values of epoxides ranged from 40 to 77%, and thehighest ee value was due to the reaction of 1,2-dihydronaphthalene with the complex in dichloro-methane at -15 °C.

In 2003, Le Maux et al. synthesized two new C2-symmetric chiral porphyrins bearing cyclohexyl sub-stituents at the ortho-position of meso-phenyl groupsand used their ruthenium complexes 44-47 to cata-lyze the asymmetric epoxidation of styrene deriva-tives with 2,6-dichloropyridine-N-oxide, e.g., a mod-erate ee value of 35% was obtained for the 1,2-dihydronaphthalene substrate.156 The results showthat the presence of various cyclohexane rings aschiral entities on metalloporphyrins appears to be anattractive mechanism for controlling both the reac-tivity and the enantioselectivity. However, the poorreactivity of “chiral cyclohexyl short arms” has be-come an indication of the level of steric incumbrancethat can be tolerated in a reactive homochiral por-phyrin complex. Compressing the size of the ring,e.g., choosing cyclopentane, cyclobutane, and cyclo-propane rings, could be a good alternative to over-come this hurdle, as reported previously with chiralphosphines in the asymmetric hydrogenation.157-159

2.2.3. Manganese PorphyrinsTo achieve high turnover numbers, Mn complexes

of some chiral porphyrins have been synthesized andtested for asymmetric epoxidations.97,160-162 In 1993,Proess et al. reported the application of the chiralmanganese porphyrin complex 48 in the asymmetricepoxidation of olefins with sodium hypochlorite as anoxygen donor.134 As shown in Table 4, when the

reactions were carried out at the most stable pH )13 value of porphyrin, the reaction time was usuallylong. When the pH value of the reaction mixture wasadjusted to between 9.5 and 10.5 by the addition ofsolid NaHCO3, all the reactions were greatly acceler-ated, and the catalyst decomposed rapidly. Addition-ally, when the catalyst was added into the reactionmixture in several batches, the first portion ofcatalyst decomposed faster than the latter ones.

Collman et al. studied the synthesis and catalyticproperties of threitol-strapped manganese porphy-rins.163 When associated with bulky cocatalysts suchas 1,5-dicyclohexylimidazole, those catalysts effec-tively catalyzed the epoxidation of 4-chlorostyrene(70% ee) and cis-â-methylstyrene (77% ee). In 1996,Vilain-Deshayes et al. further investigated the asym-metric epoxidation of 4-chlorostyrene and 1,2-dihy-dronaphthalene catalyzed by a series of chiral Mn-porphyrins 49-54 bearing glycosyl substituents(glucose, maltose, lactose) at the ortho- or meta-position of the meso-phenyl groups with dilute H2O2or iodosylbenzene as the oxidant.162,164 The enanti-oselectivity was moderate ca. 29% ee and stronglydependent on the position of the glycosyl residue. Themeta-glycosylated catalysts were not able to inducea significant ee level; however, when the chiralmoieties were in ortho-positions, closer to the activecenter of the macrocycle, an environment able toinduce a high enantioselectivity could be generated.The use of natural sugar derivatives as chiral motives

Table 4. Asymmetric Epoxidation of Olefins Catalyzedby Porphyrin Complex 48 with NaClO as theOxidanta



Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

on metalloporphyrins would be possible to catalyzethe enantioselective epoxidation of poorly reactiveolefins with easily accessible oxidants such as H2O2.

Compounds derived from 2,2-cyclophane are thoughtto be very stable to the action of light, oxidation,acids, and bases and relatively high temperatures.165

In 1998, Rispens et al. reported homochiral, atropi-somerically pure manganese complexes 55 containinga [2,2]-para-cyclophane-4-carbaldehyde building blockgroup, which were used as the catalysts in theepoxidation of unfunctionalized olefins using aqueousNaClO, 30%-H2O2 and PhIO as oxygen donors.166 Upto 780 overall turnover numbers were obtained in thepresence of NaClO and PhIO, while with 30% H2O2only catalase activity was observed; unfortunately,no significant enantioselectivity was achieved. How-ever, when the Mn(III) complex of porphyrin 55b wasused in the epoxidation of unfunctionalized alkeneswith aqueous NaClO, the enantioselectivities in therange of 22-31% were obtained. These encouragingresults suggested that enantiopure [2,2]-para-cyclo-phane-4-carbaldehyde could be a good building blockfor the synthesis of new chiral porphyrins.

It has been reported that the rate of organicoxidations catalyzed by metalloporphyrins are stronglyenhanced by the addition of an axial ligand, such aspyridine-type bases.167 For the asymmetric epoxida-tions, the addition of axial ligand to catalytic systemsis well-known to be a convenient and efficient strat-egy to enhance enantioselectivity. For instance, highenantioselectivity and good chemical yield were real-ized in the epoxidation of 2,2-dimethylchromenederivatives using achiral salen-Mn as the catalyst inthe presence of chiral bipyridine N,N′-dioxide.168 Also,axial ligation is important for effective asymmetrytransfer during the alkene epoxidation catalyzed bysingle-faced chiral porphyrins because they can actas blocking agents for the unhindered face to forcethe incoming substrate to interact with the hinderedchiral pocket.169,170 In 2002, Lai et al. investigated the

effect of multifarious organic bases (Scheme 2) inchiral Mn-porphyrin-catalyzed alkene epoxidation.171

They observed that for substituted pyridines, theenantioselectivity of cis-â-methylstyrene epoxidationfollowed a linear free energy relationship (chiralcatalyst 56), which is for pyridines bearing electron-donating groups as well. Compared to the case of noamine additive, when 4-N,N-(dimethylamino)pyri-dine (DMAP) and KHSO5 (Oxone) were employed asthe axial ligand and the oxidant, a significant im-provement in the enantioselectivity from 43 to 81%ee for the epoxidation of cis-â-methylstyrene with astereospecificity of 86.5% was achieved. This was alsothe first successful use of Oxone in the metallopor-phyrin-catalyzed asymmetric epoxidation. Very re-cently, Smith et al. in 2003 prepared the chiral Mn-porphyrin catalyst 57 to afford a very low ee valueof 7.8% for styrene, 5.8% for cis-hept-2-ene, and 0.8%for 2-methylbut-2-ene, respectively.137

2.2.4. Molybdenum Porphyrins

In 1970, Srivastava et al. first published thepreparation of molybdenum porphyrins.107 Numerousmolybdenum complexes with porphyrin ligands havebeen synthesized since then; however, to our knowl-edge all the reported complexes bear nonchiral por-phyrin macrocycles, which prevents them from func-tioning as catalysts in asymmetric epoxidations. In2001, Liu et al. initiated investigations on chiralmolybdenum porphyrins.103 It had been well docu-mented that molybdenum-catalyzed epoxidationswith alkyl hydroperoxides 58 usually involved [Mo-

Scheme 2. Various Organic Bases Used in ChiralMn-Porphyrin-Catalyzed Alkene Epoxidationa

a Adapted from ref 171 with permission. Copyright 2002 GeorgThieme Verlag.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

(η2-O2)] or [Mo-OOR] active intermediates;172 there-fore, the synthesis and structural characterization of[MoVO(P*)(OMe)] bearing a chiral D4-symmetric P*ligand (H2P* ) 5,10,15,20-tetrakis [(1S,4R,5R,8S)-1,2,3,4,5,6,7,8-octahydro-1,4:5,8-dimethanoanthracene-9-yl] porphyrin) was performed, along with thecatalytic behavior of 58 for the epoxidation of aro-matic alkenes with TBHP, which represented thefirst molybdenum porphyrin-catalyzed asymmetricepoxidation of alkenes with an ee value up to 29%(Table 5). These findings have highlighted the po-

tential use of chiral peroxo or alkylperoxo complexesof Mo-porphyrins for the asymmetric epoxidation ofalkenes.

2.3. Salen SystemsAsymmetric epoxidation of unfunctionalized alk-

enes holds a considerable interest in organic synthe-sis because the resulting enantiomerically pure ep-oxides are versatile intermediates in the preparationof several classes of compounds.

Difficulty in the construction of an effective por-phyrin catalyst is partially attributable to its π-con-jugated planar structure, which does not allow thepresence of stereogenic carbons in the porphyrin ring.In contrast to the porphyrin complex, the salencomplex can contain stereogenic carbons at C1′′, C2′′,C8, and C8′ carbons (Scheme 3).159 Since these ste-reogenic centers locate close to the metal center,therefore, higher asymmetric induction can be an-ticipated by using optically active salen complexesas catalysts. Facile synthesis of salen ligands from anumber of readily available chiral diamines has ledto a successful endeavor, which in a short time spanmade it possible to screen a large number of ligandsto optimize the catalyst structure. Very lengthy stepsinvolved in chiral porphyrin synthesis have limitedextensive applications of chiral porphyrins as cata-lysts. For industrial purposes, salen-Mn catalysts arepreferred since manganese itself is a relatively

nontoxic metal, and manganese complexes are supe-rior to iron complexes for the selective epoxidationof olefins, chiefly because of fewer side reactions overmanganese complexes.

2.3.1. Manganese and Chromium SalensThe pioneering studies performed by Katsuki and

Jacobsen et al. have led to a variety of chiral salen-Mn(III) catalysts, which have made an efficientbreakthrough in epoxidizing nonfunctionalized alk-enes with different oxidants at high enantio-selectivity.86-88,91,120,121 A variety of alkenes have,thus, been epoxidized enantioselectively with simpleoxidants such as PhIO, NaClO, H2O2, and Oxone, andan optimal enantioselectivity for a given substratecould be achieved by choosing the proper catalyst andreaction conditions.121,173 Rather straightforward ac-cess to chiral salen ligands, based on the condensa-tion of diamines with salicylaldehyde derivatives,provides a unique opportunity to finely tune the stericand electronic properties of the catalyst.174 Moreimportantly, chiral salen systems offer an additionaladvantage over the known chiral porphyrins becausethe asymmetric centers in salen complexes locatecloser to reactive metal sites than those in porphyrincomplexes.86,87,121,175,176

As the model to mimic the function of cytochromeP450, a transition metal Schiff-base has several ad-vantages: (a) easy availability, (b) low cost, and (c)high epoxidation activity toward a wide range ofunfunctionalized olefins. This should make it to be avery attractive candidate for laboratory and indus-trial uses.177 The discovery of chiral salen-Mn(III)complexes has, for the first time, allowed a conve-nient and inexpensive route for the efficient asym-metric epoxidation of unfunctionalized olefins.86-88,120

The tert-butyl substituted catalyst 59 consisting ofboth commercially available enantiomers is not themost enantioselective but is much easier to pre-pare.176 Note that the Jacobsen catalyst is air-stableand can be stored for a long period without ap-preciable decomposition.178 Indeed, with OSi(Pri)3 inC5 and C5′, the ee value is higher for cis- or cyclicolefins, of which chromenes are ideal substrates, andfor those substrates Katsuki catalyst 60 is alsoeffective.120,176

During the past decade, a large number of chiralC2-symmetric salen-type Schiff-base complexes havebeen synthesized as the catalysts for various asym-metric reactions including enantioselective epoxida-tion of unfunctionalized alkenes.8,179-181 These sym-

Table 5. Asymmetric Epoxidation of AromaticAlkenes with TBHP Catalyzed by Complex 58a


Scheme 3. Asymmetric Carbons in Porphyrin andSalena

a Adapted from ref 173 with permission. Copyright 1993 Wiley-VCH.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

metrical catalysts are generally highly efficient andstraightforward, while the unsymmetrical Schiff-baseligands are much less accessible, which has, accord-ingly, led to the scarcity of reports of chiral unsym-metrical salen ligands.182-185 Up to now, all themethods published for chiral unsymmetrical salenligands have involved a stepwise synthesis of non-C2-symmetric salen ligands containing two differentdonor units via mono-imines. Unfortunately, thosesyntheses were basically unreliable, and even afterpurification mono-imines prepared from salicylalde-hyde and 1,2-diamine were always contaminatedwith various amounts of C2-symmetric bis-imine.186

Very recently, Daly et al. succeeded in avoiding thisinherent hurdle by trapping mono-aldimine as atartrate salt;185 however, its further reaction withdifferent salicylaldehydes still produced an unsym-metrical bis-imine in low yield. So far, the mostreliable route for the preparation of non-C2-sym-metric salen ligands has been involved in the use ofa statistical method.187 Other chiral unsymmetricalsalen-like catalysts for the asymmetric epoxidationof alkenes reported recently include a biomimetic Mn-dihydrosalen complex188,189 and Mn-picolinamide-salicylidene complexes.190 A series of unsymmetricalchiral salen-Mn complexes 61 and 62 were alsoreported by Kureshy191,192 and Kim.193

Although the obtained ee value over the unsym-metrical Schiff-base complex is lower than that onthe corresponding symmetrical one, this synthesismethod could open up the possibility of designingother chiral unsymmetric Schiff-base complexes withdifferent steric and electronic properties on differentsubunits. Generally, structural and electronic proper-ties of salen ligands play important roles in control-ling the activity of catalysts.120,194 In almost all thesalen complexes, two identical salicylaldehyde de-rivative moieties are connected to both sides of onediamine in the ligands to grant a modified structuraland electronic property.87,120,121,182,194,195 In 2000, Kimet al. synthesized fully symmetrical macrocyclic

chiral salen Schiff-bases 63 through condensing twodiamine molecules with two 2-diformylphenols.196

These new macrocyclic chiral salen-Mn catalystsdisplayed moderate activity and enantioselectivity.

In 2001, Ahn et al. carried out the asymmetricepoxidation of olefins over the new sterically hinderedsalen-Mn (III) complex 64.197 Experimentally, thepreparation of salen-Mn (III) complex 64 might beeasier than Katsuki catalyst 60. On the other hand,the presence of polar ester groups at R2 in 64 wouldenhance the rate of phase transfer of oxidant mol-ecules in an aqueous/alkene biphasic reaction system;consequently, the increase of turnover frequency inthe epoxidation reactions should be anticipated.197

Furthermore, they also found that the chirality indiamine moiety was a requirement for a high enan-tioselectivity and that the absolute configuration ofstyrene oxide was controlled by the chirality in thediamine bridge rather than by the absolute configu-ration of C3 and C3′ in BINOL unit and that theenantioselectivity was as high as 96-99% ee for cis-methylstyrene and 2,2-dimethylchromene.197 It isnoteworthy that the epoxidation of cis-â-methylsty-rene over the catalyst (R, S)-64 achieved 96% ee,relatively higher than those obtained in any otherNaClO-promoted reactions at 0 °C, which furtherproves the crucial factor governing the absoluteconfiguration and ee values of epoxides to be thechirality of diamine groups at C3 and C3′ in salenligand. Very recently, Murahashi et al. in 2004reported the novel salen-Mn complex 65 bearing achiral binaphthyl strapping unit as an effectivecatalyst for the epoxidation of alkenes with iodosyl-benzene (up to 93% ee).198

Under these considerations, dimeric homochiralsalen-Mn(III) complexes 66 were synthesized andused as catalysts by Janssen,182 Kureshy,199-201andLiu et al.202 in the enantioselective epoxidation ofunfunctionalized olefins with various oxidants, inwhich excellent conversion was obtained with thehighest ee value up to 99%. Monomeric Jacobsencomplex with a catalyst loading of 2 mol % gave aconversion of 96% with 97% ee within 9 h for theoxidation of cyano chromene in the presence of4-phenyl pyridine N-oxide; however, the dimericcomplex 66a required only 2.5 h for 100% conversionwith >99% ee under identical reaction conditions,even using simple pyridine N-oxide as the additive.A low catalyst loading down to 0.6 mol % was stillsufficient for achieving satisfactory conversion andselectivity within 6 h. Further decreasing the catalystloading to 0.2 mol % caused a reduction in thereaction rate and ee value (in 24 h), while the catalystloading of below 0.2 mol % deteriorated the reactionrate and selectivity rapidly. The enhanced activity


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

of dimeric catalyst 66a shows that two units do notaffect the isolation but have a synergic interaction.Axial bases have a pronounced effect on the reactivityand selectivity of enantioselective epoxidation and ina biphasic reaction system possess a dual role instabilizing the oxo intermediate and transportingoxidant into the organic phase.203 It has been re-ported that a monomeric catalyst with appropriateN-oxide derivative additives could activate and sta-bilize the catalyst.203 Comparatively, the epoxidationof styrene over the dimeric catalyst 66a proceededfaster in the presence of either 4-phenyl pyridineN-oxide (4-PPNO) or 4-(3-phenyl propyl) pyridineN-oxide (4-PPPNO) than adding pyridine N-oxide(PyNO), without any visible improvement in thechiral induction. In addition, the recovered catalystby precipitation using hexane without undergoingfurther purification showed a visible decrease in thereactivity, without any loss of enantioselectivity.

Complexes 66b and 66c were also used in theepoxidation of unfunctionalized alkenes.200 All the

reactions were completed in less than 15 min; how-ever, the reactions using Jacobsen’s monomeric cata-lyst were complete after 25 min. As expected, dimericcomplexes showed an enhanced activity. The enan-tioselectivity of reactions was always high, irrespec-tive of the loading level of catalyst, even for catalystloading of only 0.5 mol %. A dimeric catalyst alsoshowed better repeatability and stability than did amonomeric one. Complex 66a could be efficientlyused as many as five cycles in a biphasic reactionsystem, and the activity of the recovered catalystshowed a gradual decrease with recycling numbers,possibly due to either minor degradation or solubleloss of the catalyst during the reaction.

It is well-known that the chiral center on salen-Mn complexes induces the enantioselectivity in theasymmetric epoxidation of olefins. However, Hashi-hayata et al. in 1997 first developed a new asym-metric epoxidation system consisting of an achiralsalen-Mn complex and a chiral amine, in whichcommerically available (-)-sparteine displayed thehighest asymmetric induction.204 It should be men-tioned that the addition of water into the reactionmedium improved the enantioselectivity (up to 73%ee) for the epoxidation of chromene derivative. In1999, the same group reported again high enanti-oselectivity and good chemical yield in the epoxida-tion of 2,2-dimethylchromene derivatives using achiralsalen-Mn(III) complex 67 as the catalyst in thepresence of axially chiral bipyridine N,N′-dioxide.168

Their research has suggested a new concept of chiralactualizer, which could have the potential to open anew avenue for asymmetrically catalytic reactions.

The chiral catalysts with salen ligands derivedfrom appropriately modified common carbohydrateshave been found to be as active as those based on1,2-cyclohexanediamine. There was a report made byBorriello et al. in 2004 regarding the synthesis of newchiral salen-type ligands 68 and 69 derived fromD-glucose and D-mannose.205 These two compoundswere obtained by introducing appropriate functionsat the C2 and C3 positions of the sugar ring. Theefficiency of the corresponding Mn complexes in theepoxidation of styrenes was shown, e.g., for function-alized cis-â-methylstyrene an ee value of 86% wasachieved. These promising results deserve furtherinvestigation aiming at the exploration of otherfunctional groups present naturally on the skeletonof carbohydrates. For instance, -OH groups notinvolved in metal coordination may be used for otherpurposes, such as anchoring onto a polymeric matrixor improving the solubility in alternative solvents.

Chromium-salen complexes have been extensivelyapplied in stereoselective alkene epoxidations,181,206-208

kinetic resolution of epoxides,209-211 alcohol oxida-tions,212 asymmetric addition of organometallic re-agents to aldehydes,213-215 and asymmetric hetero-


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

Diels-Alder reactions.216 Distinctly different fromsalen-Mn-catalyzed alkene epoxidations (for recentfindings on mechanisms see refs 217-221), chromium-salens give good enantioselectivities for (E)-alkenes,and the established oxygen-transfer CrVdO speciesis relatively stable and EPR active. For more detailedunderstanding of the mechanism of either Mn- or Cr-salens mediated chiral epoxidation, please refer to arecent review by Gilheany et al.181

Recently, Gilheany et al. investigated the substitu-ent effect on the salen ring, which focused mainly oncombinations with respect to all-important 8 posi-tions and on a comprehensive study of effect of onesubstituent at all the positions.222 The results showedthat the enantioselectivity was increased at 3,3- and6,6-positions and decreased at 4,4- and 5,5-positionsby halo-substitution on the salen ring. Addition oftriphenylphosphine oxide significantly increased theselectivity of the complexes coming from 3,3- or 5,5-substitution rather than 4,4- or 6,6-substitution, andthe use of a nitrate counterion was beneficial in mostcases. Additionally, the enantioselectivity could beelevated by the addition of many different types ofphosphoryl compounds, among which tri-arylphos-phine oxides were the most effective, e.g., tris(3,5-dimethylphenyl) phosphine oxide led to the highestee value of 93%.222b A ceiling effect of additive hasbeen observed, in which bulky additives were inef-fective, accompanied by a saturation effect withrespect to the increase of concentration; however,there was no electronic effect of substituents insidethe additive. It was found that the presence ofcompounds containing extended π-electron systemsstrongly affected the ee (20% ee).208b In certain cases,such additives appeared to stabilize active oxidantand to slow the reaction. Unsubstituted and methyl-substituted imidazoles were found to be beneficialadditives, while imidazoles with aromatic substitu-ents were very detrimental. These results haveprovided actual support for Katsuki’s views on theimportance of π-interactions in the epoxidationscatalyzed by analogous salen-Mn complexes. Com-pounds containing SdO and CdO bonds also showedan impact on the enantioselectivity to a lesser extent;

however, whether phosphine sulfides on borane couldsurvive contact with the oxidant would depend on thesalen substitution pattern. A phosphorus ylide wasfound to stabilize a Cr(V) oxo species, approximatelyto double its lifetime.

2.3.2. Cobalt SalensMuch attention has been paid to the development

of direct and selective epoxidation of olefins by theuse of molecular oxygen and a suitable reductant,such as a primary alcohol,223 aldehyde,224 and cyclicketone225 that can accept one oxygen atom frommolecular oxygen to enable the reaction. In 1997,Kureshy et al. first reported the asymmetric epoxi-dation of nonfunctionalized prochiral olefins by com-bined use of an atmospheric pressure of molecularoxygen and a reductant of isobutylaldehyde catalyzedby chiral salen-Co (II) complexes 70 with or withoutPyNO co-oxidant.226 An up to 55% ee and a yield of90% were achieved for the catalytic epoxidation oftrans-3-nonene. Anyway, it should be pointed outthat the presence of a catalytic amount of pyridineN-oxide notably improved the enantioselectivity with-out any change in the configuration.

2.3.3. Palladium SalensC2-symmetric 1,1-binaphthyl unit and tetradentate

Schiff bases (salen) constitute two of the most im-portant types of chiral auxiliaries in metal-mediatedasymmetric catalysis. There are a variety of Pd (II)Schiff base complexes that have been reported in theliterature, of which only those bearing bidentateSchiff base ligands are well documented.227 Althoughthe salen-Pd (II) complex was first isolated in 1963,tetradentate Schiff base complexes of this metal ionhave still been sparse, and their reactivity towardthe alkene epoxidation remains unexplored.227,228 In2000, Zhou et al. provided the first example regardingthe asymmetric epoxidation of alkenes catalyzed bya palladium complex PdII-71.229 With a molar ratioof catalyst/styrene/TBHP ) 1:5:20, the reaction overthe complex PdII-71 gave rise to a 16% conversion ofstyrene after 6 h, affording styrene epoxide in theselectivity of 37% with 17% ee and benzaldehyde inthe selectivity of 50% (based on the consumed sty-rene). Under identical conditions, 27% of p-fluorosty-rene was converted to p-fluorostyrene epoxide in theselectivity of 51% with 71% ee and to p-fluorobenz-aldehyde in the selectivity of 32%.

2.3.4. Ruthenium SalensIn 1999, Takeda et al. published the synthesis and

catalytic application of (ON+)(salen) ruthenium (II)complex [(ON)Ru-salen complex] 72 in the asym-


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

metric epoxidation of 6-acetamido-2,2-dimethyl-7-nitrochromene in the presence of various oxidants,for which 2,6-dichloropyridine N-oxide as terminaloxidant was preferred.230 All the tested conjugatedolefins showed high enantioselectivities greater than80% ee, irrespective of their substitution patterns;however, the enantioselectivity decreased with elon-gated reaction time. This suggested the gradualdecomposition of the (ON)Ru-salen complex 72 dur-ing the reaction to generate fragmental Ru-specieswith a lower asymmetric induction. Furthermore, thereaction was accelerated by exposure to sunlight. Thesense of asymmetric induction by the (ON)Ru-salencomplex 72 was similar to that of the Mn-salencomplex bearing the same salen ligand.

In summary, metallosalens are considerably effec-tive catalysts for the asymmetric epoxidation ofconjugated cis-di-, tri-, and some tetra-substitutedolefins.231 Metallosalens can be easily synthesizedfrom corresponding salicylaldehyde derivatives, di-amines, and metal ions. The main advantages include(1) the commercial availability of a wide variety ofchiral and/or achiral salicylaldehydes, chiral di-amines, and metal ions, and (2) the easy preparationand use of metallosalen complex catalysts. Thisresults in the convenient application of such catalystsin the industry. And, when these salen-complexes areemployed to catalyze the epoxidation of functional-ized substituted olefins, the satisfied results can stillbe achieved.

2.4. BINOL Systems

2.4.1. Lanthanum BINOLsThe catalytic asymmetric epoxidation of R,â-

unsaturated carbonyl compounds is an importanttransformation in organic synthesis because it con-structs two adjacent chiral carbon centers simulta-neously, and the corresponding enantiomericallyenriched epoxy compounds can be easily convertedto many types of useful chiral compounds.232-239

Although many research groups have developedefficient catalytic asymmetric epoxidations of R,â-

unsaturated ketones,232,233,236,238-242 there are onlylimited examples with respect to the epoxidation ofR,â-unsaturated esters as substrates catalyzed byeither a salen-Mn complex243 or a chiral ketone.244-248

In both cases, the substrate was only â-aryl-substi-tuted R,â-unsaturated ester. Shi et al. reportedexcellent results for a few â-alkyl-substituted sub-strates but without trans-â-alkyl-substituted R,â-unsaturated esters due to the poor chemoselectivi-ty.248 Recently, the use of lanthanoid-BINOL com-plexes to catalyze the enantioselective epoxidation ofenones with hydroperoxides was studied in greatdetail.240-242,249-251 This catalytic system was ap-plicable for both alkyl and aryl substituted enones,giving the product in high yields with good enantio-selectivity.

Shibasaki catalyst, which resulted from an equimo-lar mixture of (R)-(+)- 1,1′-bi-2-naphthol (BINOL)and La(OPri)3, could catalyze the asymmetric epoxi-dation of a wide range of (E)-enones in the presenceof 4 Å molecular sieves.240 When chalcone was ep-oxidized by cumene hydroperoxide oxidant, 93% ofchemical yield and 83% ee value were obtained.Ytterbium catalysts showed a good chiral inductionfor the epoxidation of alkyl R,â-enones, in which 96%ee was obtained with external addition of triph-enylphosphine oxide (15 mol %),242 and the positiveeffect of addition of water on the enantioselectivitywas observed.249 In 1998, Daikai et al. found that thecoordination of an external ligand to chiral ytterbiumcomplex [Yb-(R)-(BNP)3] not only increased the solu-bility of catalyst but also largely enhanced theenantioselectivity of hetero-Diels-Alder reaction,242,252

indicative of the importance of saturated coordinationto lanthanoid with an appropriate number of ligandsfor the deoligomerization of lanthanoid complexes. Inaddition, they further investigated the effect of ad-ditives on lanthanoid complex-catalyzed asymmetricepoxidation of enone that was developed by Shibasakiet al.232,233,253-255 Among the additives tested, triph-enylphosphine oxide showed the best result (96% ee)and a notable ligand acceleration of the reaction rate.One plausible explanation could be responsible forsuch a remarkable ligand effect, i.e., the interactionof ligand with Ph3PdO would produce a nonpoly-merically uni-structured catalyst, further leading tothe occurrence of epoxidation on the coordinationsphere of lanthanum, where the reaction sites mightbe quite close to the chiral binaphthyl ring due tothe steric hindrance of bulky phosphine oxide ligand.It was also found that when CMHP was used in placeof TBHP, the enantioselectivity was further raisedto over 99% ee, and the amount of catalyst could bereduced to 0.5 mol % without any serious loss in theenantioselectivity. More importantly, all of the re-agents required for this asymmetric epoxidation arecommercially available, and low reaction tempera-tures usually required to attain a high enantioselec-tivity are not necessary, thus making it a highlypractical protocol. To understand the mechanism ofhighly enantioselective epoxidation, the relation ofenantiomeric purity of chiral ligand to that of theproduct was investigated. The results have shownthat the ligand with only 40% ee could induce the


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

product with more than 99% ee. This phenomenonstrongly suggests that the active catalyst may not bemonomeric, instead it has a particular structure thathardly changes during the reaction because of itsthermodynamic stability.

In 2003, a highly practical way was again reportedfor obtaining R,â-epoxy ketones with a high opticalpurity (Table 6).251 The chiral lanthanum complex

self-assembled in-situ from lanthanum triisopro-poxide, (R)-BINOL, triarylphosphine oxide, and alkylhydroperoxide (1:1:1:1) efficiently catalyzed the ep-oxidation of R,â-unsaturated ketones with TBHP orCMHP at room temperature into the correspondingepoxy ketones with an enantioselectivity of >99% ee.Actually, this catalyst system has been successfullyscaled up to the level of 30-80 kg, in which theproduct yield could reach 90% with a high ee valueof >98%.256 Also, this homochiral dimeric binuclearµ-complex, La/(R)-BINOL/Ph3PO/ROOH (1:1:1:1), wasproposed as an active catalyst for the reaction inScheme 4, accompanied with a possible catalytic cycleas shown in Scheme 5.251 One of two lanthanum ionswould function as a Lewis acid to activate thesubstrate, and the peroxide attached to the othermight be delivered as the active oxidant, thus, tocontrol the stereochemistry of epoxidation. For theepoxidation of chalcone with CMHP, the catalyst La-(OPri)3/(R)-BINOL/Ph3PO/CMHP (1:1:1:1) affordedunusual yield (98%) and ee value (>99%). The addi-tion of triphenylphosphine oxide seems to stabilize

those intermediates, while excessive CMHP couldjeopardize their structures because of the formationof an oligomeric structure.

In 2001, Nemoto et al. reported the application ofthe asymmetric catalyst, La-BINOL-Ph3AsdO (1-5mol %), for the epoxidation of a variety of enones,including dienone and cis-enone.241 The reaction wascompleted in a reasonable time to attain the corre-sponding epoxy ketones in 99% yield and >99% ee.And recently, Ohshima et al. in 2003 claimed the firstexample of a general asymmetric epoxidation of R,â-unsaturated carboxylic acid derivatives, catalyzed byLa-BINOL-Ph3AsdO complex with TBHP.257 Thisnew catalytic system has been extended to theepoxidations of R,â-unsaturated carboxylic acid 4-phe-nylimidazolides. All the epoxidations of â-aryl-typesubstrates could end in a reasonable reaction time(3.5-6 h); however, only â-aryl-type substrates witheither an electron-withdrawing group or an electron-donating group on the aromatic ring were smoothlyepoxidized to obtain a good ee (89-93% ee). Thiscatalyst system was also effective for highly reactiveâ-alkyl substituted substrates. Both primary andsecondary alkyl substituted substrates gave rise tothe products in high yield (72-93%) with slightly lowee (79-88% ee). In particular, this reaction wasapplicable to substrates that are functionalized witha C-C double bond or a ketone without over-oxidation.

2.4.2. Ytterbium BINOLsLa-catalysts and Yb-catalysts can be used in a

complementary manner for the asymmetric epoxida-tion of aromatic and aliphatic substituted enones(94% ee) at room temperature. In 1998, Watanabeet al. discovered that the addition of 4 Å molecularsieves to the catalyst was considerably effective foracquiring the product in high yield and that even thepresence of a small amount of water obviouslyimproved the ee value.249 For example, 83% yield and85% ee were achieved for the asymmetric epoxidationof 4-methyl-1-phenyl-pent-1-en-3-one with the addi-tion of H2O, while the absence of H2O led to only 81%yield and 70% ee. However, the absence or presenceof H2O did not show any visible impact on La-catalyst. The role of water can be rationalized withthe coordination of water molecules to Yb atom,thereby controlling the orientation of hydroperoxidesto form an appropriate asymmetric environment.Recently, Chen et al. successfully developed a seriesof 6,6′-disubstituted BINOL ligands (S)-73 and thecorresponding ytterbium complex catalysts.258 Espe-cially, the ytterbium complex resulting from Yb-

Table 6. Asymmetric Epoxidation of Various Enoneswith La-BINOL Complexes and TBHPa,b

a Adapted from ref 251 with permission. Copyright 2003Wiley USA. b CMHP was used as the oxidant in entry 11-22.

Scheme 4. Proposed Structure of the ActiveCatalysta

a Adapted from ref 251 with permission. Copyright 2003 WileyUSA.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

(OPri)3 and (S)-6,6′-diphenyl-BINOL in THF wasfound to be an efficient catalyst for the asymmetricepoxidation of R,â-unsaturated enones with CMHP,in which the best results were 97% ee and 91% yieldfor the product (E)-1,3-diphenylprop- 2,3-epoxy-1-one(Table 7).

2.4.3. Gadolinium and Samarium BINOLsThe introduction of substituents into 6,6′-positions

of BINOL might exert a profound impact on the

activity and enantioselectivity of the catalyst due tothe modified electronic and steric properties.259-262 In2001, Chen et al. achieved epoxy chalcone with 95%ee over 5 mol % of Gd(OPri)3-(S)-6,6′-diphenyl-BINOLcatalyst at room temperature.263 In this system,CMHP was obviously advantageous over TBHP, e.g.,the ee value decreased from 95% with CMHP to 86%with TBHP for the production of epoxy chalcone. Thechiral ligand could be quantitatively recovered andreused with a negligible loss of chemical yield andee. In 2003, Kinoshita et al. developed the asym-metric epoxidation of R,â-unsaturated N-acyl pyr-roles, which are highly reactive and versatile sub-strates as ester surrogates.264 Although the functionalgroup tolerance of the Wittig reaction is good undermild conditions, this olefination is generally consid-ered to be somewhat inefficient as a result of theproduction of Ph3PO as a byproduct. They hypoth-esized that the reuse of waste Ph3PO in the epoxi-dation reaction as a modulator for Sm-H8-BINOLcomplex would partially compensate the disadvan-tage of the Wittig reaction. Thus, a sequential reac-tion from Wittig olefination to catalytic asymmetricepoxidation took place, in which Ph3PO yielded in thefirst (Wittig) reaction was reused as an additive inthe second reaction epoxidation. This process pro-vided an efficient access to prepare optically activepyrrolyl epoxides from a variety of aldehydes in highyields with excellent enantioselectivities. Under op-timal conditions (THF/toluene and Ph3PO), the reac-tion proceeded smoothly with a low catalyst loadingof 1 mol % to afford the epoxide 74 in 94% yield and99% ee in 18 min. The reaction reached the comple-tion over 5 mol % of the catalyst in the presence ofCMHP (less explosive and reactive than TBHP)within 12 min with 91% yield and >99.5% ee.

2.4.4. Calcium BINOLsAs described above, the asymmetric epoxidation of

R,â-unsaturated enones is also a powerful strategy

Scheme 5. Speculated Catalytic Cyclea

a Adapted from ref 251 with permission. Copyright 2003 Wiley USA.

Table 7. Asymmetric Epoxidation of EnonesCatalyzed by Yb-BINOL Complexes 73 with CMHP atRoom Temperaturea



Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

for the synthesis of enantiomerically enriched organiccompounds.265 In 2003, Kumaraswamy et al. firstdeveloped a novel and efficient chirally modifiedcalcium complex 75 derived from commerically avail-able low-cost CaCl2 and potassium salt of (S)-6,6′-diphenyl BINOL, for the asymmetric epoxidation ofR,â-unsaturated enones (Table 8).266 Introduction ofsubstituents to the 3-position and 6,6′-positions ofBINOL exhibited profound effects not only on theactivity but also on the enantioselectivity of epoxychalcone.258,263 In an effort to increase the enantiose-lectivity of chalcone, a wide range of substituted-BINOL and C2-symmetric ligands were prepared.The results showed that complex 75 offered a higheree value than did the 2′-substituted BINOL calciumcomplex 76 and the C2-symmetric complex 77. Theincrease in the enantioselectivity was not only dueto the electronic effect of 6,6′-aryl substitution aspredicted by Vries et al.258,263 but also due to theincrease in bond angle between two naphthyl rings,which may provide a favorable coordination environ-ment at the metal-ligand site.

2.5. Chiral Carbonyl Compound SystemsThe intermediary of dioxiranes formed in oxidation

reactions was first suggested by Baeyer and Villigerin 1899 in the monoperoxysulfate oxidation of men-thone into the corresponding lactone.267 Experimentalconfirmation of this unusual structure was not avail-

able until 1977 when the parent dioxirane wasdetected during gas-phase ozonolysis of ethylene, andits existence was thus rigorously established.268-270

In the realm of solution-phase oxidation, however,the seminal breakthrough came in 1974, when Mont-gomery observed that ketones could effectively cata-lyze the decomposition of Oxone271 and the oxidationof halides and dyes.272 Montgomery’s speculation didnot establish that dioxiranes were the active inter-mediates, which was experimentally evidenced by 18Olabeling studies of Curci.273 These investigators dem-onstrated conclusively that dioxiranes were indeedproduced in situ from the interaction between Oxoneand ketones and that these intermediates wereresponsible for the accelerated decomposition of Ox-one.273 Since the isolation of dioxiranes is oftendifficult, therefore, the in-situ generation of dioxi-ranes is generally required, for which excess Oxone,a ketone and an appropriate buffer are employed.Recently, much attention has been focused on theepoxidation of alkenes with Oxone by either chiralketones244,274-280 (via a dioxirane273,281-283) or chiral-iminium salts284,285 (via an oxaziridinium species),which is a possible alternative to solve the problem.

2.5.1. Simple Chiral KetonesDioxiranes, either isolated or generated in situ

from KHSO5 (Oxone) and ketones (Scheme 6), have

shown to be efficient oxidants for the asymmetricepoxidation of unfunctionalized olefins. Owing tothe rapidness, mildness, and safety of the reaction,this system has been extensively investi-gated.108,273,281-283,286-295 The first chiral ketone-catalyzed asymmetric epoxidation was reported in1984 by Curci,274 who carried out the epoxidation(12.5% ee) of (E)-â-methylstyrene or 1-methylcyclo-hexene over the catalyst (+)-isopinocamphone 78 or(S)-(+)-3-phenyl-butan-2-one 79 in a biphasic mixtureof CH2Cl2/H2O buffered to a pH of 7-8, using Bu4-NHSO4 as a phase transfer catalyst. Since electron-deficient ketones were generally more reactive for theepoxidation, a trifluoromethyl group was accordinglyincorporated into the structure of ketones to preparehighly active ketones 80 and 81 with high yields.276

Yang et al. synthesized C2-symmetric cyclic chiralketones 82-84, derived from 1,1′-binaphthyl-2,2′-dicarboxylic acid.244,277,296 The design was based onthe following considerations:239 (a) C2-symmetry wasintroduced to limit competing reaction modes of the

Table 8. Asymmetric Epoxidation of r,â-UnsaturatedEnones with TBHP over the Catalyst 75 in theSolvent of Cyclohexane/Toluene (9:1) in the Presenceof 4 Å Molecular Sievesa


Scheme 6. Catalytic Cycle Involving Dioxiranesa

a Adapted from ref 291 with permission. Copyright 1998 TheAmerican Chemical Society.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

attack on dioxirane, which has two faces for theoxygen transfer. (b) Che chiral element was placedaway from the catalytic center (the carbonyl group)while avoiding the occurrence of any substitution atR-carbon since R-carbons are prone to the racemiza-tion, and the steric hindrance at R-carbon decreasesthe catalyst activity. (c) Two electron-withdrawingester groups were introduced to activate the carbonylgroup. The resulting ketone was stable and recover-able in high yield with up to 95% ee.

Ketone (R)-83 was easily prepared in one step from(R)-6,6-dinitro-2,2-diphenic acid and 1,3-dihydroxy-acetone using the Mukaiyama reagent. (R)-82a and(R)-83 showed similar enantioselectivities for theepoxidation. Ketone (R)-84 (10 mol %) was appliedin the asymmetric epoxidation of trans-stilbene underthe same reaction conditions with (R)-82a. It wasfound that the epoxidation reaction proceeded with10% of conversion (determined by 1H NMR) after 6h, and the ee value of trans-stilbene epoxide wasabout 12%. The poor activity of the ketone could beassociated with the fact that an ether group is aweaker electron-withdrawing group than an estergroup. The ester groups giving rise to a rigid and C2-symmetric structure of cyclic ketones seemed to beessential for effective asymmetric epoxidation. Ascompared to ketone 82, ketones 85 and 86 made bySong et al. in 1997 showed a low enantioselectivityof e59% ee, possibly because a replacement of theester group with the ether one that is a weakerelectron-withdrawing group could bring the carbonylgroup closer to the chiral center.280 But Adam’s

ketones 87 and 88 exhibited a catalytic activitysimilar to 85 and 86.297 Also, Yang et al. havedemonstrated the potential of C2-symmetric chiralketones for the catalytic asymmetric epoxidation oftrans-olefins and trisubstituted olefins.296 Epoxida-tion reactions could be performed with only 10 mol% of ketone catalysts, which could be recovered andreused without any loss of activity and chiral induc-tion. Convincing evidence for a spiro transition stateof dioxirane epoxidation has been proposed throughthe 18O-labeling experiment, in which chiral diox-iranes were found to be the intermediates in epoxi-dation reactions catalyzed by chiral ketones.

In 1999, Denmark et al. reported a highly activeand enantioselective seven-membered carbocyclicchiral ketone 89 with a chiral center closer to thecarbonyl group, in which the fluorine substitution atthe R-carbon caused a dramatic impact on the epoxi-dation rate, leading to an ee up to 94%.298 In 2002,Stearman et al. synthesized a series of chiral binaph-thyl ketone catalysts 90 with variable distributionsof fluorine atoms close to the carbonyl group.299 Theasymmetric epoxidation of trans-â-methylstyreneover catalysts 90 were all run under the sameconditions with 10 mol % catalyst at pH ) 9.3 and-15 °C. Parent ketone 90a catalyzed the epoxidationof trans-â-methylstyrene to gain 35% conversion with46% ee, while Song’s ketone 85 merely offered 29%ee. This was encouraging since the nonfluorinatedversion of Denmark’s catalyst 89a only achievedabout 5% conversion under similar conditions. Monof-luorinated ketone 90b led to 57% conversion andincreased the ee to 80%, while the R,R′-difluorinatedcatalyst 90c catalyzed the complete conversion ofsubstrate to epoxide with 86% ee. However, thetrifluoroketone catalyst 90d did not show any im-provement in the ee with a complete conversion.Denmark previously described a dramatic stereoelec-tronic effect of fluorinated cyclohexanone derivativesthat are constrained conformationally, with axialfluorination resulting in little activation of the ketonemoiety.298 This is consistent with the action of thethird fluorine atom in the catalyst 90d, whichpresumably occupies a pseudoaxial position, thusoffering no advantage as compared to the difluori-nated catalyst 90c. Tetrafluorinated ketone 90e,which exists almost exclusively as the hydrate, gave


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

rise to the results similar to the parent ketone 90a.The strong preference for the formation of hydratehas made the catalyst 90e less available for thedioxirane formation, hence decreasing the conversionand ee. Comparatively, the best catalyst is the R,R′-difluoroketone 90c, which has catalyzed the epoxi-dation of trans-â-methylstyrene with 100% of con-version and 86% ee at a catalyst loading of 10 mol%.

In the particular case of electron-deficient olefins,significant results have been obtained over deriva-tives of (-)-quinic acid and tropinone as well. Bor-tolini et al. in 2001 designed the ketone 91 (bile acid)with a chiral and steroid skeleton that can conferconformational rigidity, preventing any distortionand forcing three keto functional groups into differentdirections.300 In addition, the carboxylic functionalgroup on a lateral chain makes the ketone 91 solublein a slightly basic aqueous solution and anchorableonto a suitable support. As shown in Table 9, the

asymmetric epoxidation of different cinnamic acidderivatives in water-NaHCO3 has been performedin the system consisting of dehydrocholic acid (anoptically active ketone) and Oxone to attain a productwith an ee value of 75%. The temperature was foundto be an important factor for the oxidations withdioxiranes generated in situ.301,302

In 2000, Solladie-Cavallo et al. published theirresults with respect to the epoxidation of methylp-methoxycinnamate, tert-butyl p-methoxycinnama-

te, and trans-stilbene over R-halogenated ketones92.246 When the epoxidation was catalyzed by usingR-chloro ketones 92c and 92d having a methyl groupin the R-position instead of an isopropyl group, thechlorine in the axial or equatorial position signifi-cantly enhanced the efficiency of ketone. For ex-ample, when the chlorine occupies the axial position,73% of conversion could be realized, notably higherthan 20% of conversion with R-chlorine in the equa-torial position. The same behavior was also observedover R-fluoro ketones 92e and 92f, where R-fluorinein the axial position of 92f largely promoted almostthe complete conversion of substrate, while that inthe equatorial position of 92e led to only 43% ofconversion. It, thus, appears that the ketones withan axial halogen (Cl or F) are more efficient thanthose with an equatorial one. Also, it was shown thatthe promotion effect of axial fluorine was strongerthan axial chlorine.

In 2002, chiral ketones 93a and 93b bearing a1-aza-7-oxabicyclo[3,5,0]decane skeleton and their C2-symmetric analogue 94 were prepared as chiraldioxirane precursors for the asymmetric epoxidationof olefins with Oxone.303 Ketone 93b, bearing adiphenyl steric wall, was not effective for the chiralinduction with a quite poor selectivity. For example,the epoxidation of trans-stilbene with Oxone cata-lyzed by a stoichiometric amount of 93a in acetoni-trile-dimethoxymethane (DMM)-water (2:1:2) af-forded (S,S)-stilbene oxide with 60% ee, while thereaction over 93b produced (R,R)-stilbene oxide withonly 6% ee. However, for the same reaction ketone94 showed a good enantioselectivity up to 83% ee.These results have suggested that the Coulombrepulsion between carbonyl and ether oxygen atomsis operative as an electronic wall rather than as asteric wall. It, thus, seems that the maximal ef-ficiency of chiral ketone catalysts could be reachedby both choosing efficient substituents and adjustingthe reaction conditions, based on the substrates andthe ketone catalysts. The activating effects of fluorinesubstitution adjacent to the ketone carbonyl moietyare apparent for methyl (trifluoromethyl) dioxirane304

and dioxiranes derived from 2-fluoro-cyclohexa-nones,298 2-fluoro-1-tetralones, and 2-fluoroindan-1-ones.275 These results were influential in the designand subsequent use of oxidizing agents resulting fromketones 95,239,286-295 92e,245,246 and more recently90c.299

Table 9. Asymmetric Epoxidation of ProchiralCinnamic Acid Derivatives with Ketone 91a



Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

Armstrong et al. have made important contribu-tions to the progress of this field, particularly, theyrightly extolled the virtues of R-fluoro-N-ethoxycar-bonyltropinone 96 as an efficient and readily recycledcatalyst for the epoxidation of alkenes with Oxone(83% ee).245 To improve the enantioselectivity furtherand to clarify the factors responsible for the asym-metric induction, many attempts to prepare bicyclo-[3,2,1]octanone derivatives 95 and 97 with alterna-tive R-substitution have been carried out.239,286-295

Moreover, R-fluoroketone 98 displayed a much betterability to catalyze the asymmetric epoxidation oftrans-stilbene with Oxone, for which enantiomericallyenriched ketone 98 with 80% ee gave rise to stilbeneepoxide with 74% ee.305,306

2.5.2. Polyhydric Compounds Derived KetonesIn 1996, Shi et al. first reported the asymmetric

epoxidation of trans- and trisubstituted olefins withOxone over a fructose-derived ketone 99, which washighly efficient and mild, accompanied by high eevalues.278 Ketone 99 is an efficient epoxidation cata-lyst with high ee values for a variety of trans- andtrisubstituted olefins; however, it is not effectivetoward R,â-unsaturated esters due to its decomposi-tion under reaction conditions, presumably via aBaeyer-Villiger oxidation. The ketone was designedbased on the following ideas: (a) stereogenic centersclose to the active sites can result in an efficientstereochemical communication between the substrateand the catalyst; (b) the presence of a fused ring anda quaternary R-carbon attached to the carbonyl groupminimizes the epimerization of stereogenic centers;(c) one face of the catalyst is sterically blocked to limitpossible competing approaches. Carbohydrate-de-rived chiral ketones as the epoxidation catalysts canbe rationalized by the following reasoning: (a) car-bohydrates are chiral and readily available; (b)carbohydrates are highly substituted by oxygengroups with an inductive effect to activate the ketonecatalysts, which would possess a good reactivity; (c)carbohydrate-derived ketones could have rigid con-formations because of an anomeric effect, whichwould be desirable for a high enantioselectivity.

Subsequently, during the period from 1999 to 2001,Shi et al. extensively revealed high enantioselectivi-ties (89-99% ee) of D-fructose-derived chiral ketone99 for the asymmetric epoxidation of olefins withhydrogen peroxide or Oxone as primary oxidantunder mild conditions.307-316 The results showed that

hydrogen peroxide other than Oxone could be usedto generate dioxiranes and that the volume of solventrequired was significantly reduced as compared tothe case of potassium monoperoxysulfate (Oxone) asprimary oxidant. When the reactions were carried outin the solvents, such as DMF, THF, CH2Cl2, EtOH,or dioxane instead of CH3CN, GC detected only traceamounts of epoxides (<1%), suggesting that thecoexistence of hydrogen peroxide and CH3CN pro-moted the formation of catalytic dioxirane. Mostlikely, in the presence of CH3CN the actual oxidantresponsible for the formation of dioxirane was aperoxyimidic acid intermediate as depicted in Scheme7. A control experiment showed that in the absence

of any ketone, the addition of 1.0 M K2CO3 gainedonly 1% of conversion the reaction was stirred for 5h at 0 °C, while under identical conditions theintroduction of simple acetone led to a conversion of61%, implying that dioxiranes could also be gener-ated from other ketones rather than merely limitedin ketone 99.

However, for the epoxidation of cis-olefins ketone99 showed a rather poor enantioselectivity; for ex-ample, only 39% ee was obtained for cis-â-methyl-styrene with a major (1R,2S)-enantiomer. Scheme 8

shows that there are likely two major competing spirotransition states A and B formed in this catalyticsystem. A low ee value for cis-olefins suggests thatthe ketone catalyst cannot provide an essentiallystructural environment in these two transition statesto differentiate sufficiently between phenyl and meth-yl groups on the olefin. To solve this problem, theyfurther prepared a new chiral ketone 100,311 a

Scheme 7. Formation of Peroxyimidic AcidIntermediate (dioxirane)a


Scheme 8. Structure of Competing SpiroTransition States A and Ba

a Adapted from ref 315 with permission. Copyright 2001 TheAmerican Chemical Society.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

nitrogen analogue of 99, which achieved generallyhigh ee values for a class of acyclic and cyclic cis-olefins.

Owing to the extreme usefulness of epoxides de-rived from terminal olefins, asymmetric epoxidationof unfunctionalized terminal olefins has receivedintensive interest.317-319 Metal-complex catalystssuch as chiral porphyrin and salen complexes havebeen extensively studied for the epoxidation of ter-minal olefins, and in a number of cases the enantio-selectivity reaches about 80%, even a high up to>90% ee in certain cases. In 2001, Shi et al. discov-ered that chiral ketones 100a-c were effectivecatalysts for the epoxidation of terminal olefins.313

The substitution on the nitrogen of ketone 100aobviously enhanced the enantioselectivity, with upto 85% ee obtained for substituted styrene substrates(Table 10). As illustrated in Scheme 8, spiro A and

spiro B are the most plausible transition states, andseemingly spiro A is advantageous over B for thosesubstrates containing a π-bond system, where groupswith a π-bond system prefer proximal to the spiro

oxazolidinone. To probe the interaction further, ke-tones 100d-h with substituents on the phenyl groupwere prepared and tested as catalysts for the epoxi-dation of cis-â-methylstyrene, styrene, and 1-phenyl-cyclohexene.320 In the case of cis-â-methylstyrene, theee value increased from 83% for the electron-donatingMeO group to 90% for the electron-withdrawingsulfone and nitro groups. The interaction between thephenyl group of olefin and the oxazolidinone moietyof catalyst in the transition state spiro A could beinfluenced by the electronic nature of substituentson N-phenyl groups and favored by electron-with-drawing groups. A lower ee value obtained with 100his probably because the N-phenyl group in 100h isno longer coplanar with oxazolidinone due to thepresence of an ortho-nitro group in the phenyl group.The substituent effect on the enantioselectivity waseven more explicitly displayed in the epoxidation of1-phenylcyclohexene, for which the (R,R)-isomer wasobtained with p-MeO and p-Me groups, and the (S,S)-isomer with p-MeSO2 and p-NO2 groups. Clearly,electron-withdrawing groups on the N-phenyl groupin the catalyst further enhance the attractive inter-action between phenyl groups of olefins and oxazo-lidinones in the transition state of catalyst.

In 2002, Shi et al. further developed a chiral ketone101, a readily available acetate analogue of 99, whichwas active and highly enantioselective for the epoxi-dation of R,â-unsaturated esters in the presence ofOxone.248 They replaced the fused ketal of 99 withmore electron-withdrawing groups so as to reduce theBaeyer-Villiger decomposition and to enhance thestability and reactivity of ketone. The results in Table11 showed that an ee ranging 90-97% could beobtained for those substrates containing an electron-withdrawing group, particularly for trisubstitutedR,â-unsaturated esters and conjugated enynes. Theinteraction of Oxone with ketones 102 and 103,derived from 2-quinic acid, could also generate chiraldioxiranes, which then epoxidized enones (Table12).321,322

In 2002, Shing et al. investigated the asymmetricepoxidation of various alkenes catalyzed by threeD-glucose-derived ulose catalysts 104a-c with n-Bu4-NHSO4 as the oxidant,323 where ketone 104a dis-played a better enantioselectivity than did 104b and104c, e.g., the best ee values obtained for trans-stilbene oxide were 71% ee for the ketone 104a, 11%ee for 104b, and 26% ee for 104c, respectively. In2003, they reported again the catalytic applicationsof L-erythro-2-uloses 105, L-threo-3-uloses 106, and107 prepared from L-arabinose, for the asymmetricepoxidation.324 The anomeric aglycone steric sensorswith suitable size (e.g., uloses 105a and 105d)exhibited good stereochemical communication towardthe oxidation of trans-stilbene (up to 90% ee). Chemi-cal yields of epoxides with uloses 105 were poor (only

Table 10. Asymmetric Epoxidation of TerminalOlefins with Ketone 100a

a Adapted from refs 313 and 320 with permission. Copyright2001 and 2003 The American Chemical Society.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

13% yield) because of the decomposition of ulosecatalysts during the reaction. However, L-threo-3-uloses 106 and 107 could overcome the decomposi-tion, ascribable to the electron-withdrawing effect ofthe ester group(s) at the R-position. In the same year,readily available arabinose-derived 4-uloses 108,containing a tunable butane-2,3-diacetal as the stericsensor, was further synthesized, which showed anincreased enantioselectivity up to 93% ee along withincreasing the size of the acetal alkyl group in thecatalytic asymmetric epoxidation of trans-disubsti-tuted and trisubstituted alkenes.325 Ulose 108c wasthe best catalyst and chiral inducer for the epoxida-tion of a simple trans-disubstituted alkene (93% ee).

2.5.3. Chiral AldehydesAlthough ketone dioxiranes have been well-known

as excellent oxidants for decades,281-283,288 only in

recent years were chiral ketone dioxiranes detectedto be prominent oxidants for asymmetric epoxida-tions.239,247,270,299,303,320,323,326-329 In contrast, aldehydedioxiranes (RHCO2) still remain elusive, even thoughthey have been involved in biological processes andsubstantiated in theoretical studies.330,331 For ex-ample, C4a-flavin hydroperoxide has been used in thebioluminescent process of bacterial luciferase to reactwith an aldehyde generating the corresponding di-oxirane as a high-energy intermediate, which has,however, been neither isolated nor characterized.331

So far, the only known aldehyde dioxirane is dihy-drodioxirane (H2CO2), but it is obtained from theozonolysis of ethane at low temperature,270 ratherthan from formaldehyde directly. In 2003, Bez et al.first reported the chiral epoxidation with Oxone overoptically active aldehydes 109 synthesized fromfructose, which is the first direct evidence for theinvolvement of aldehyde dioxiranes.332 Although ke-tone dioxirane may be readily prepared from theoxidation of ketone by Oxone, the preparation ofaldehyde dioxirane by this way is tabooed. This isbecause aldehyde is reportedly prone to become thecorresponding acid undergoing an Oxone oxida-tion.333,334 It is clearly observed from Table 13 thataldehyde 109b yields consistently better enantiose-lectivity than aldehyde 109a for those substrates,except for terminal and cyclic alkenes.

In summary, chiral carbonyl compounds can beeasily synthesized from some cheap natural com-pounds and have shown especially high enantiose-lectivities for the asymmetric epoxidation of trans-olefins, which is relatively valuable for industrialapplications. Additionally, these catalysts display

Table 11. Asymmetric Epoxidation of r,â-UnsaturatedEsters with Ketone 101a

a Adapted from refs 248 with permission. Copyright 2002The American Chemical Society.

Table 12. Asymmetric Epoxidation of r,â-UnsaturatedOlefins with Oxoneover Ketones 102 and 103a

a Adapted from refs 321 and 322 with permission. Copyright1997 and 1999 The American Chemical Society.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

extensive applicability for almost all the substrates,such as functionalized and unfunctionalized substi-tuted olefins (please refer to Table 45 in the ap-pendix).

2.6. Chiral Iminium SaltsOxaziridinium salts as an electrophilic oxygen

source was first reported in 1976 by Milliet,335,336whichshowed extreme reactivity for the oxygen transfer tonucleophilic substrates, such as sulfides and alk-enes.337,338 Oxaziridinium salts, either used in iso-lated form or generated in situ from iminium saltsand Oxone, are powerful electrophilic oxidants.338,339

These organic salts were prepared either by thequaternization of the corresponding oxaziridines orby the peracid oxidation of iminium salts. Chiraliminium salts have been employed to catalyze theasymmetric epoxidation of olefins with Oxone as theprimary oxidant, in which a good chemical yield buta moderate enantioselectivity was obtained, possiblydue to a poor chiral recognition by the flat aryl ringattached to the central carbon of iminiumsalts.284,285,319,340-344 Aryl-substituted iminium saltsare more stable and easily isolated than alkyl-substituted ones; however, the intrinsic requirementfor an aryl ring severely limits the design andcatalytic use of chiral iminium salts. Alternatively,iminium salts can be generated in situ from thecondensation of either a ketone or an aldehyde witha secondary amine, which proceeds under slightlyacidic conditions that benefit the dehydration ofaminol. In the reaction, iminium salts first react withOxone to form oxaziridinium salts, which then trans-fer oxygen to epoxidize olefins, accompanied with aregeneration of iminium salts; subsequently, therestored iminium salts react again with an additional1 equiv of Oxone to generate oxaziridinium salts andto complete a catalytic cycle.

Oxaziridines, nitrogen analogues of dioxiranes,have constituted an important class of organic oxi-

dants under nonaqueous conditions. More recently,Aggarwal et al. have described the catalytic asym-metric epoxidation of simple alkenes mediated by abinaphthalene-derived iminium salt,284 which cata-lyzed the synthesis of both 1-phenylcyclohex-1-eneoxide with 71% ee and trans-stilbene oxide with 31%ee. Also, Armstrong et al. showed that even acycliciminium salts could be efficient catalysts for theasymmetric epoxidations by Oxone.285,342,345 The firstenantiopure oxaziridinium salt 110 was synthesizedby the quaternization of an oxaziridine derived froma chiral imine that was prepared from norephedrinein four steps.319,340 This salt and the correspondingiminium salt could catalyze the asymmetric epoxi-dation of alkenes using Oxone as the stoichiometricoxidant to obtain an ee value of ca. 40%, but no cis-epoxide products were detected, suggesting a single-step oxygen transfer process. And the enantioselec-tivity was higher for disubstituted olefins than formono- or trisubstituted ones. The presence of twoaromatic rings on the opposite ends of disubstituteddouble-bond seems to favor the “transfer of chirality”from oxaziridinium salt to the epoxide via the oxygentransfer.

Minakata et al. in 2000 applied a chiral ketimin-ium salt 111 in the asymmetric epoxidation of3-phenyl-prop-2-en-1-ol, but the enantioselectivitywas not high (39% ee).344 Page et al. in 2000 inves-tigated one group of dihydroisoquinolinium saltcatalysts, active at a loading as low as 0.5 mol % butlow enantioselective with only ca. 40% ee in theepoxidation of alkenes.343 Chiral primary amines(Scheme 9) could rapidly react with 2-(2-bromoethyl)

benzaldehyde to produce oil compounds of dihydro-isoquinolinium bromide salts, which were purifiedwith difficulty by conventional methods. This problemwas eventually solved by the counteranion exchange,which took place simply by adding an appropriateinorganic salt into the reaction mixture before theworkup. As inorganic additives, fluoroborate, hexa-fluorophosphate, perchlorate, and periodate salts ortheir derivatives were not ideal, but tetraphenyl-borate salts 112 derived from sodium tetraphenyl-borate were preferred.

Table 13. Enantioselective Epoxidation with ChiralAldehydes 109a

a Adapted from refs 332 with permission. Copyright 2003Elsevier.

Scheme 9. Structure of Chiral Primary Aminesa

a Adapted from ref 343 with permission. Copyright 2000 TheRoyal Society of Chemistry.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

Page et al. in 2001 further examined the catalyticapplication of dihydroisoquinolinium salts containingalcohol, ether, and acetal functional groups in thenitrogen substituent in the asymmetric epoxidationof olefins (ca. 60% ee).346 In addition, (1S,2R)-norephedrine derivative 113 catalyzed the epoxida-tion of 1-phenylcyclohexene with 30% ee, the corre-sponding catalyst derived from (1S,2R)-2-amino-1,2-diphenylethanol 114 led to 1-phenylcyclohexene oxidein 24% ee, while (1R,2S)-aminoindanol derivative 115showed a lower asymmetric induction (<14% ee). Theresults suggested that the size of an ether substituentin such catalysts is not particularly important for theasymmetric induction during the oxygen transfer toalkene. The selectivity may be partially assigned tointernal stabilization of the positive charge by adja-cent oxygen atom, which increases the conforma-tional and rotational rigidity of the molecule. Imin-ium salt 116 induced chiral R-methyl stilbene oxidewith 52% ee at 0 °C, notably higher than 15% ee ofisopinocamphenylamine-derived catalyst 117.

In 2002, Page et al. synthesized two new chiraliminium salts 118 and 119, of which dibenzaze-pinium salt 119 induced an ee value of ca. 60%.347

The results showed that these two new seven-membered ring catalysts were more reactive than thesix-membered ring catalyst 116 in the epoxidationreaction. In all cases, the catalyst 118 gave a pooreree value than did the catalyst 116, while the catalyst119 induced a somewhat different ee value from thecatalyst 116. For example, the catalyst 119 achievedan improved ee value of 60% for the substrate ofphenylcyclohexene but a similar ee value of 59% tothe catalyst 116 for triphenylethylene and even alower ee value of 15% for trans-stilbene, while thecatalyst 116 showed an extremely low enantioselec-tivty. In 2001, Wong et al. developed a new approachto catalyze the asymmetric epoxidation of olefins,which utilized chiral iminium salts 120 and 121,generated in situ from chiral amines and aldehydes,

as the catalysts.348 Epoxidation reactions were con-ducted with 20 mol % of amines and aldehydes toobtain epoxides with 65% ee. In 2002, Lacour et al.observed that the combined use of TRISPHAT coun-terions 122 with catalytic amounts of 18-Crown-6allowed an improvement of enantioselectivty in thecase of iminium precatalyst 123 under biphasic CH2-Cl2/water conditions, in which the ratio of enanti-omers increased from 2.4:1 to 7.2:1 for the epoxida-tion of 1-phenyl-dihydronaphthalene.349

2.7. Other Homogeneous SystemsAlkene epoxidations catalyzed by transition-metal

peroxo species have been studied for quite a longtime. There have been several example reactions,such as the epoxidation of olefins over polyoxo tung-state (PW12O24

3-) in the presence of quarternaryammonium salts,307-316 over active molybdenum per-oxo complexes (to yield the corresponding epoxidesin 50% yield and 40% ee),350 and over methyltriox-orhenium (MTO) with H2O2.93

2.7.1. Mo-peroxo ComplexesIn 2000, Park et al. published their studies on the

enantioselective epoxidation of styrene derivativesover transition metal (Mo)-peroxo complexes 124.351

As shown in Table 14, Mo-peroxo complexes achievedyields of 40-50% with 30-80% ee, and a significantsolvent effect on the epoxidations was observed; forinstance, the addition of solvents CH2Cl2 and isooc-tane inhibited alkenes from being epoxidized byTBHP. In 2003, Zhao et al. synthesized severalmolybdenum(VI)-cis-dioxo complexes bearing sugar-derived chiral Schiff-base ligands in a general for-mula of MoO2(L)(Solv) 125,352 where L ) N-sali-cylidene-D-glucosamine, N-salicylidene- 1,3,4,6-tetra-acetyl-R-D-glucosamine, N-5-chlorosalicylaldehyde-1,3,4,6-tetraacetyl-R-D-glucosamine, N-salicylalde-hyde-1,3,4,6-tetraacetyl-â-D-glucosamine,N-5-chloro-


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

salicylaldehyde-1,3,4,6-tetraacetyl-â-D-glucos-amine, N-salicylidene-4,6- O-ethylidene-â-D-gluco-pyranosylamine, and Solv ) methanol or ethanol.These complexes have been employed in the asym-metric epoxidation of olefins, to obtain a moderateenantiomeric induction of about 30% ee for cis-â-methyl styrene. As expected, for the epoxidation ofcis- and trans-â-methylstyrene, the common observa-tion is that the catalytic activity and asymmetricinduction for cis-substrates is much better than thatfor trans-analogues.

2.7.2. Lithium ComplexesIn 2003, Tanaka et al. developed the asymmetric

epoxidation of R,â-unsaturated carbonyl compoundswith alkylperoxides in good ee, catalyzed by externalchiral tridentate aminodiether-lithium peroxides 126-138 (Table 15).353 This work has revealed two novelfindings: (a) a slow addition of alkylperoxide benefitsthe catalytic asymmetric reaction; (b) the lone-pairelectron differentiating the coordination of carbonyloxygen to lithium is a critical factor for a highenantioselectivity. A reasonable explanation could beresponsible for the latter, i.e., a bulky substituent onthe carbonyl group of ketone prevented the lone-pairelectron from coordination at one side of the CdCdouble bond, hence directing a predominant coordi-nation to lithium at the other side opposite to thebulky substituent. On the basis of these experimentalfacts, the researchers have come to the conclusionsthat (a) a chiral ligand not only activates the lithiumperoxide but also controls the absolute stereochem-istry of nucleophilic epoxidation reaction, (b) a tri-dentate aminodiether ligand with a hemilabile meth-

oxyphenyl side chain benefits a high enantioselectivity,(c) the lone-pair electron differentiating the coordina-tion of carbonyl oxygen to lithium is actually opera-tive, and (d) a slow addition procedure of alkylper-oxide is preferred. This has, to date, become a certainbasis of developing more sophisticated conjugateaddition-type, ligand-controlled catalytic asymmetricepoxidation of R,â-unsaturated carbonyl compounds.

2.7.3. Magnesium ComplexesJackson et al. reported earlier that chalcone could

be converted into the corresponding epoxide withgood to excellent ee values by an inexpensive catalyst,prepared from DET and dibutylmagnesium.233 How-ever, this system was not especially effective for theepoxidation of aliphatic enones, only resulting in poorconversions. In 2001, they modified the originalprocedure so that it was effective for the asymmetricepoxidation of simple aliphatic enones with a highee value.354 They found that the addition of a 4 Å

Table 14. Asymmetric Epoxidation of AlkenesCatalyzed by Mo-peroxo Complexes 124a,b withTBHPa

a Adapted from ref 351 with permission. Copyright 2000 TheKorean Chemical Society.

Table 15. Asymmetric Epoxidation of r,â-UnsaturatedOlefins Catalyzed by an External Chiral TridentateAminodiether-Lithium Peroxidea

a Adapted from refs 353 with permission. Copyright 2003Elsevier.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

molecular sieve allowed a reasonable conversion ofthe substrate into the corresponding epoxide and thatit was possible to prepare the presumed catalystprecursor 139 simply by the addition of L-(+)-DETto dibutylmagnesium soluble in heptane and followedby removing the solvent and drying. As describedabove, the complex 139 was rather effective for theasymmetric epoxidation of aliphatic enones withTBHP in the presence of a 4 Å molecular sieve;however, there was one drawback, i.e., the conversionto the epoxide was initially fast and subsequentlysluggish, which could be improved by a portionwiseaddition of the solid catalyst 139 (Table 16).

2.7.4. Zinc ComplexesIn 1996, Enders et al. discovered that (E)-R,â-

unsaturated ketones could be asymmetrically epoxi-dized to yield trans-epoxides using a stoichiometricquantity of diethylzinc and a chiral alcohol, underoxygen atmosphere.232 Recently, Yu et al. achieved

up to 99% ee for these R,â-unsaturated ketones inthe presence of chiral ligand (1R,2R)-N-methylpseu-doephedrine 140 and chiral polybinaphthyl alcohols141 or 142, with almost a quantitative yield ofrecovery of the catalyst from the reaction mixture(Table 17).355

On the basis of the Weitz-Scheffer mechanism,Enders and Pu proposed a similar reaction pathwayfor their asymmetric epoxidations, namely, diethylz-inc reacts with chiral ligand (A) to give a zincalkoxide (B) with the release of ethane, and then thereaction of B with oxygen gives rise to a chiral metalperoxide (C), as depicted in Scheme 10.238 Advantagesof this route are that the cheapest oxidant such asoxygen or even air can be used and that the chiralligand can be recycled by the extraction into acid(HClaq) subsequently followed by both the basifica-tion with NaOHaq and the re-extraction into ether.Enders’ reagent has been extensively applied in theasymmetric epoxidation of chalcone and derivatives,alkyl-substituted enones, â-alkylidene-R-tetralones,and some (E)-nitroalkenes.356 This has provided anaccess to 3-substituted trans-2-nitro oxiranes with

Table 16. Asymmetric Epoxidation of AliphaticEnones Catalyzed by Complex 139 with TBHPa,b

a Di-tert-butyl tartrate as the ligand; n.d.; not determined.b Adapted from ref 354 with permission. Copyright 2001 TheRoyal Society of Chemistry.

Table 17. Asymmetric Epoxidation of Various Olefinswith Diethylzinc and Oxygen in the Presence ofChiral Alcohols 140-142a

a Adapted from refs 355 with permission. Copyright 1999The American Chemical Society.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

excellent de values of g98% and good ee values of36-82%.

2.7.5. Ruthenium ComplexesC2-symmetric chiral oxalamides have shown to be

attractive ligands for the enantiocontrol of metal-catalyzed reactions, in particular for the oxidations.Because of the π-acceptor property of R-dicarbonylunit and oxazoline rings, these ligands are supposedto be quite resistant to the oxidation and hencesuitable for the preparation of stable high-valentmetal complexes. The first example was given byYang et al., who employed a chiral N-donor bidentateligand and RuCl3 catalyst to obtain only a 21% eevalue.357 In 1998, End et al. developed the catalyticasymmetric epoxidation of olefins over Ru complexeswith chiral oxalamides 143 using NaIO4 oxidant.358,359

Among the oxalamides tested, isopropyloxazolinederivative 143b was found to be the most activeligand. The results showed that in the asymmetricepoxidation of (E)-stilbene, byproduct benzaldehydewas rapidly formed in the beginning of the reactionand that the ee value of the epoxide was very low inthe beginning but increased significantly along withprolonging the reaction time. However, if the catalystfirst underwent a pretreatment by the oxidant for amoment before contacting the substrate, the catalytic

performance, as expected, was significantly improvedwith 74% yield and 69% ee.

Ruthenium complexes with chiral N,O- and N,N-donor ligands have been tested as the asymmetricepoxidation catalysts.105,110,360 In 1998, Stoop et al.prepared five-coordinated ruthenium complexes con-taining tetradentate chiral ligands 144-146 with aP2N2 donor, effective for the asymmetric epoxidationof alkenes with hydrogen peroxide.361,362 An ee valueof 42% was obtained for the asymmetric epoxidationof styrene catalyzed by Ru-145, which was the firstexample of the asymmetric epoxidation over ruthe-nium complexes using hydrogen peroxide as primaryoxidant.

Tse et al. investigated the asymmetric epoxidationof trans-stilbene over Nishiyama’s Ru-(pyridine bisox-azoline) (pyridinedicarboxylic acids) 147a and 147b,where pyridinebisoxazoline (pybox) ligands were eas-ily prepared from natural amino acids.363,364 Unfor-tunately, the previous work of this system revealedseveral drawbacks, such as a low reactivity (96 hneeded for a complete conversion) and the limitedcatalyst. However, successful epoxidations have beenmade using trans-stilbene as the substrate, for whichincreasing the concentration of oxidizing agent (PhI-(OAc)2) resulted in an elevated yield but a reducedenantioselectivity, and water played a crucial role inthe reaction, possibly due to the enabling of theoxidation of 18-electron Ru(S,S-iPr2-pybox)(pydic)147a via a ligand dissociation. The addition ofstoichiometric amount of water increased the ee valueto 71% in the case of a low concentration of PhI(OAc)2as the stoichiometric oxidant, which, however, wasremained as a major disadvantage. Hence, in 2004,Bhor et al. developed a brand new system usingTBHP as the oxidant instead of PhI(OAc)2, for whichRu(pybox)(pydic) complexes 147-149 were used asthe catalysts for the epoxidation of olefins with TBHPto achieve an excellent yield up to 97% with a feasibleee value of 65% at room temperature.365

Scheme 10. Reaction Pathway Involving ZincComplexesa

a Adapted from ref 238 with permission. Copyright 2000 TheRoyal Society of Chemistry.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

Ruthenium complexes with nitrogen ligands havebeen recognized by a number of researchers aspowerful oxidizing catalysts, extensively applied inthe oxidative transformations.97 Since the pioneeringwork of Balavoine et al. in the epoxidation of olefinscatalyzed by a RuCl3-2,2′-bipyridine-NaIO4 sys-tem,366,367 many approaches have been proposed torender the efficiency of this system.368-370 In 2002,Pezet et al. investigated the synthesis and catalyticapplication of chiral ruthenium sulfoxide 150 for theasymmetric epoxidation of olefins.371 As presented inTable 18, most of the reactions have achieved high

yields and moderate ee values at room temperature;however, in the case of trans-â-methyl styrene anespecially high ee value of 94% is obtained, which isthe best result, so far, in the ruthenium-catalyzedepoxidation of alkenes. As the stoichiometric andcatalytic oxidants, oxo complexes of ruthenium withpolypyridine ligands (cis-[Ru-(bpy)2(py)O]2+) havebeen employed in a variety of organic and inorganicreactions.372-375 Optically pure cis-∆-[Ru-(bpy)2(py)O]has also been used in the stoichiometric oxidation ofmethyl p-tolyl sulfide.376 On the basis of these experi-ments, it is assumed that in the reaction, the oxygenatom transfer occurs via the in-situ formation of theoxo-ruthenium complex (RudO) intermediate, fur-ther confirmed by the displacement of chlorine ligandwith two chiral centers (sulfoxide and metal center)to form a catalyst.

2.7.6. Methyltrioxorhenium (MTO)Recently, an impressive amount of work has been

carried out on the catalytic property of methyltrioxo-rhenium (VII) (MTO) and more specifically on thepossibility as a selective epoxidation catalyst using

H2O2 as the oxidant.377-381 The transformation ofepoxide into 1,2-diols can be suppressed by replacinghydrogen peroxide with urea/hydrogen peroxide ad-duct (UHP) as the primary oxidant, which enablesthe epoxidations in nonaqueous media, whereby theepoxide ring-opening is largely avoided.381,382 A sec-ond alternative is based on the tendency of organo-rhenium(VII) oxides to increase the coordinationnumbers to form Lewis acid-base adducts.383,384 Inthis way, the complexes formed with certain nitrogen-containing bases have shown to be excellent catalystsfor the epoxidations in the presence of excessamine.93,385,386 In an attempt to achieve chirality,several chiral rhenium(VII) adducts were synthesizedby adding (S)-(+)-2-aminomethylpyrrolidine, (R)-(+)-phenyl ethylamine or L-prolinamide into differentsolutions of MTO. Outstandingly, the addition ofthese chiral auxiliaries induced the enantioselectiveand diastereoselective oxidations of several prochiralolefins with low-to-moderate level of stereoinductiv-ity.

2.7.7. Nickel ComplexesThe first example of the epoxidations of olefins over

nickel(II) complexes with iodosylbenzene was madeby Kochi.387 Later, Burrows et al. reported thesynthesis, crystalline structure, and catalytic ap-plication of nickel(II) complexes for the epoxidationof a wide variety of olefins in the presence of iodo-sylbenzene or hypochlorite.388-390 Dinuclear nickel-(II) complexes were also synthesized and used as theepoxidation catalysts by Gelling.391 Kureshy et al.successfully applied the symmetrical and nonsym-metrical Ni(II) square planar complexes 151 and 152in the enantioselective epoxidation of unfunctional-ized olefins, such as 1-hexene, 1-octene, trans-4-octene, styrene, 4-chlorostyrene, 4-nitrostyrene, and1,2-dihydronaphthalene using NaClO or O2 as theoxidant, in which 58% ee was achievable.392,393

2.7.8. Sulfonium YlidesThe recorded ways for preparing chiral epoxides

are chiefly as follows: (a) enantiofacially differentialoxidation of a prochiral CdC bond98,120,394,395 and (b)enantioselective alkylidenation of a CdO bond, viaa ylide route or a Darzens reaction. Although theformer has afforded much success, the structuralrequirement of substrates is a limited prerequisite.

Table 18. Asymmetric Epoxidation of OlefinsCatalyzed by Ru-Complex 150 with THBP at 25 °Ca

a Adapted with permission from ref 371. Copyright 2002 TheRoyal Society of Chemistry.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

The latter provides an alternative approach in solvingthis problem. Enantioselective epoxidation via a ylideroute is an approach, easily performed but relativelyundeveloped to date.396,397 Chiral sulfides, via thecorresponding sulfonium ylides, are increasinglybecoming important sources of chirality for asym-metric transformation, enantioselective epoxidation,and cyclopropanation.398,399 As an efficient method tosynthesize chiral epoxides, the base-promoted asym-metric epoxidation of aldehydes over chiral sulfideshas been developed.400-405

In 1996, Li et al. reported that D-(+)-camphor-derived sulfides 153-156 were rather efficient cata-lysts for the preparation of trans-2,3-diaryloxirane(up to 98% yield and 77% ee).406 When benzylatedsulfides 153a-156a were employed in the ylidereaction, a stoichiometric enantioselective epoxida-tion was realized. The opposite asymmetric induction,which led to the synthesis of both (+)- and (-)-trans-diaryloxiranes, was achieved only when exo-alkylthio-substituted sulfide 153 or 154a and endo-alkylthio-substituted sulfide 155 or 156 were used, respectively.This represents an efficient preparation of enantio-merically enriched (+)- and (-)-trans-diaryloxiranesfrom ylides with both high yields and reasonable eevalues, although both (+)- and (-)-enantiomers havealso been obtained in low yields from other routes.407,408

The effect of asymmetric induction of 153a (exo-benzylthio) was generally better than that of 155aor 156a (endo-benzylthio). It is noteworthy that theconversion of a free OH group at the C2-position to amethoxy group dramatically decreased both the yieldand the ee value of products. The investigationshowed that increasing the amount of sulfides didnot influence the ee values but shortened the reactiontime and improved the yields. As expected, theopposite asymmetric induction was again observedfrom 153b, which contains an exo-methylthio group,and from 155b or 156b, which contains an endo-methylthio group. A nonbonded interaction betweenthe free OH in the ylides resulting from sulfides (153,155, and 156) and the carbonyl group of aldehydescontrols the access of substrates to ylidic carbonpreferentially at one specified face, thus leading to amore efficient asymmetric induction than in the caseof ylide coming from methyl-protected hydroxylatedsulfides 154, which cannot cause such an interaction.

In 1999, Hayakawa et al. succeeded in highlyenantioselective synthesis of epoxides from aldehydescatalyzed by the sulfur ylide derived from chiralâ-hydroxy esters.402 However, the sulfur ylide derived

from chiral sulfide 157 resulted in (R,R)-stilbeneoxide in 64% yield with only 7% ee, possibly due tothe large distance from the chiral center to thereaction site. Accordingly, tetrahydrothiophene de-rivative 158 was prepared as the catalyst instead ofthe tetrahydrothiopyrane derivative, which notablyincreased the ee value of the product to 78%; inparticular, the highest enantioselectivity reached92% ee in the reaction of tolaldehyde.

Chiral pyridyl alcohols, prepared by reducing thecorresponding pyridyl ketones enantioselectively,have found numerous applications, e.g., as ligandsin the metal-mediated asymmetric catalysis,409,410 asresolving agents,411 and as starting materials forpreparing more advanced chiral ligands.412,413 Simi-larly, chiral furyl hydroperoxides have also been usedin the asymmetric epoxidation of allylic alcohols52 andsulfides.58,414 In 2000, Solladie-Cavallo et al. carriedout the asymmetric synthesis of epoxides from pure(R,R,R,SS)-(-)-sulfonium salt 159, commercially avail-able aldehydes and a phosphazene base [EtP2dEt-NdP(NMe2)2(NdP(NMe2)3)].403 They found that EtP2provided an exceptionally high enantioselectivitygenerally over 97% ee (Table 19) and that EtP2H+/

TfO- could be recovered by simply varying thesolvent polarity to regenerate the expensive baseEtP2 as shown in structure 159.

In 2001, Aggarwal et al. reported a unique processfor the direct coupling of two different aldehydes toform epoxides with the control of the relative andabsolute stereochemistry.415 High enantioselectivities(up to 94% ee) were achieved for the asymmetricepoxidation of a wide range of aromatic, heteroaro-matic, unsaturated, and aliphatic aldehydes cata-lyzed by chiral sulfide 160 with tosylhydeazone salts(tosyl ) toluene-4-sulfonyl) in the presence of Rh2-(OAc)4. It was found that the sulfide loading couldbe reduced from 20 to only 5 mol % without affecting

Table 19. Asymmetric Synthesis of Epoxides fromSulfonium Salt Using EtP2 Basea

a Adapted from ref 403 with permission. Copyright 2000Wiley-VCH.


Dow

nloa

ded

by U

NIV

OF

MO

NT

AN

A o

n Se

ptem

ber

18, 2

009

| http

://pu

bs.a

cs.o

rg

Pub

licat

ion

Dat

e (W

eb):

Apr

il 19

, 200

5 | d

oi: 1

0.10

21/c

r040

6458

Advances in Homogeneous and Heterogeneous Catalytic ...

Documents

Transcript of Advances in Homogeneous and Heterogeneous Catalytic ...