Hydroacylation and Related Topics Dong Group Seminar

Brandon Reinus

Wed, Sept. 5th 2012

Why?

¡ Looking at the reaction, it is a highly atom- economical approach to synthesizing ketones

¡ Umpolung (ex: deprotonating dithioacetals)

¡ Using acrylate derivatives generates a 1,4 diketone relationship, a hard relationship to establish using classical organic synthesis. Presentation Overview

1. Hydroformylation (extremely brief) 2. Rh-Catalyzed Hydroacylation ¡ Intramolecular ¡ Intermolecular ¡ Other 3. NHC Catalyzed Hydroacylation ¡ Benzoin reaction ¡ Stetter reaction ¡ other Part 1 : Background

Reppe Roelen Science of Synthesis, Stereoselective Synthesis 1, 2011, pg.409 Hydroformylation Metal-Organic Cooperative Catalysis Park et al.

Introduction SCHEME 2 Among the many elegant examples of transition metal cat- alyzed activation of C-HandC-Cbonds,1 chelation-as- sisted protocols have recently attracted increasing attention in organometallic chemistry. Directed metalation processes have been demonstrated by valuable applications in organic synthesis, showing remarkable efficiency and chemoselectivity.2 In general, a chelation-assisted proto- col facilitates the formation of either the kinetically or ther- modynamically favored five- or six-membered metallacycle; aprepositionedcoordinatinggroupinducesspatialproxim- ing decarbonylation, their structures are too specific to apply ity between the C-HorC-Cbondsandthetransitionmetal for common aldehydes. center.1,2 Despite the magnificent usefulness in activating otherwise stable C-HandC-Cbonds,amajordrawbackof the chelation-assisted protocol is the need for a coordina- (1) tion group in a substrate. As a consequence, this introduces the necessity of additional steps to remove the coordinat- ing groups, and the method often suffers from limitations in employing noncoordinating or weakly coordinating sub- (2) strates. In particular, the C-Hbondactivationprocessof aldehydes, a key step of hydroacylation of olefins, can be seriously hampered by the lack of a strong coordinating The turning point of general chelation-assisted hydroa- group in a substrate leading to more critical problems such cylation was realized from the example of hydroacylation as decarbonylation (Scheme 1).3 using aldimine made up of 2-amino-3-picoline (1)andben- Because the principal challenge associated with the C-H zaldehyde.7 There is a very important consideration in our bond cleavage of an aldehyde involves suppressing the focus on a metal-organic cooperative catalysis (MOCC). For decarbonylation process of the acyl-metal hydride spe- instance, if incorporation of a new catalytic cycle of aldi- cies, some catalyst systems under a high pressure of car- mine formation into the conventional hydroacylation of bon monoxide or ethylene gases have been devised to aldimine is possible, we anticipated that a general and stabilize acyl-metal hydride species.4 However, these sys- decarbonylation-free hydroacylation of olefins with a vari- tems are not general and efficient due to the harshness of ety of aldehydes could be devised (Scheme 2). the reaction conditions and limitation of the olefins that can Fascinated by the concept of MOCC, we have investigated be used. acatalyticsystemconsistingofatransitionmetalandan An alternative approach to stabilize the acyl-metal hydride organic catalyst as a new chelation-assisted hydroacylation species was inspired by the isolation of cyclometalated com- protocol. In the quest for an optimized system composed of a 5 plexes derived from 8-quinolinecarboxaldehyde (eq 1) and Wilkinson’s catalyst, (Ph3P)3RhCl (2)and1,wefoundthatcom- 2-(diphenylphosphanyl)benzaldehyde6 (eq 2), which were a mon aldehydes can be successfully applied in this transfor- milestone of chelation-assistedPart hydroacylation. 2 : Rh-Catalyzed Although it is mation. 8 In addition, surprising efficiency enhancements could apparent that these aldehydesHydroacylation were very effective in avoid- be achieved when benzoic acid and aniline were introduced SCHEME 1

More on Decarbonylation: Organometallics 1999, 18, 5311 Adv. Synth. Catal. 2006, 348, 2148 JACS 2008, 130, 5206 Chem. Commun. 2008, 6215

Decarbonylation can be surpressedVol. by 41, using No. 2 highFebruary pressures 2008 222-234 of ethyleneACCOUNTS OF CHEMICAL RESEARCH 223 or CO or by generating a metallacycle

Accounts of Chemical Research p222, 2008

726 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 1 Scheme 4

Scheme 5

Scheme 2

Scheme 6

Historical Reactions Scheme 3 726 Chemical Reviews, 2010, Vol. 110, No. 2 ¡ Sakai (1972) Willis which the acyl unit is positioned trans to the chloride ligand (Scheme 4).8 The stability of complex 5 is due to the slow Scheme 1 Scheme 4 dissociation of the trimethylphosphine ligands.9 Heating 726726 ChemicalChemical Reviews, Reviews, 2010, 2010, Vol. Vol. 110, 110, No. No. 2 2 WillisWillis complex 5 to 50 °Cresultedinintramolecularhydroacylation to provide cyclopentanone in 72% yield. An experiment SchemeScheme 1 1 ¡ Milstein (1982)SchemeScheme 4 4 using a catalytic amount of complex 5 achieved three synthetically useful methods based on this reaction are turnovers; although only a low number of turnovers were known.4 In general, the most significant advances in hydroa- achieved,*Trimethyl this represented the first example of catalysis that cylation chemistry have involved developing strategies, or didphosphine not rely onis slow the presence of ethylene (for example, see Schememethods, 5 to limit this undesired decarbonylation pathway. Schemeto dissociate 5). ¡ Miller and Coworkers (1976) SchemeSchemeThe review 5 5 is divided according to the transformation Scheme 2 being considered, that is, intramolecular alkene hydroacy- 2.2.*Also Catalytic isolated Systems ethylene insertion lation, intermolecular alkene hydroacylation, etc., and when 2.2.1. Cyclopentanone Synthesis SchemeScheme 2 2 needed also by catalyst type. During these sections, detailed products mechanistic arguments relevant to the key hydroacylation Miller and co-workers demonstrated that intramolecular ¡ Larock (1980) Schemestep 6 will not be presented; rather, these are collected together alkene hydroacylation could be performed using substo- SchemeSchemein section 6 6 6. The scope of this review is comprehensive from ichiometric amounts of rhodium complexes if ethylene- the first report in 1972 through June 2009. A number of saturated solvent was employed. 4-Pentenal could be cyclized reviews that deal with C-H activation cover some aspects to provide cyclopentanone in 72% yield using only 10 mol of hydroacylation,1 as do several metal-specific reviews;5 %Wilkinson’scomplexinethylenesaturatedchloroform however, a comprehensive treatment that covers all substrate (Scheme 5).10 Several products originating from intermo- types and all catalyst systems has yet to appear.6 Reactions lecular reaction of ethylene and pentenal were isolated in that generate acyl metal intermediates similar to 1 indirectly Chemicallow yieldReviews, (see 2010, Scheme p. 725 32).11,12 via simple carbonylation or carbonylation-type reactions are Larock observed a similar improvement in the efficiency not considered.7 of a range of intramolecular hydroacylation reactions upon the use of ethylene saturated solvent. In a departure from 2. Intramolecular Alkene Hydroacylation the Miller system, he found that catalysts generated in situ Scheme 3 from [Rh(COD)Cl]2 and P(4-MeC6H4)3,P(4-MeOC6H4)3,or Scheme 3 13 Scheme 3 which2.1. Stoichiometric the acyl unit is positioned Systemstrans to the chloride ligand P(4-Me2NC6H4)3 were most efficient. With the relatively which the acyl8 unit is positioned trans to the chloride ligand high catalyst loadings of 50 mol % in methylene chloride at (Scheme 4). 8The stability of complex 5 is due to the slow whichdissociation(Scheme theSakai acyl disclosed 4). unit ofThe is the positioned stability the trimethylphosphine first of exampletrans complexto of the5 ligands. ais metal-mediated chloride due9 toHeating the ligand slow room temperature, the method could be used for the synthesis hydroacylation8 reaction in a study on prostanoid synthesis.9 3 (Schemecomplexdissociation 4). 5Theto 50 of stability°Cresultedinintramolecularhydroacylation the trimethylphosphine of complex 5 is ligands. due to theHeating slow of a range of substituted, spirocyclic, and fused cyclopen- dissociationtocomplexThe provide reactions of5 to cyclopentanone the 50 involved° trimethylphosphineCresultedinintramolecularhydroacylation treatment in 72% of a yield. range ligands. An of 4-enals experiment9 Heating with tanone products (Scheme 6). No other olefin or acetylene stoichiometric amounts of Wilkinson’s complex at room complexusingto5 provideto a 50 catalytic cyclopentanoneCresultedinintramolecularhydroacylation amount of in complex 72% yield.5 achieved An experiment three additive proved as effective. It was also reported that adding usingtemperature a catalytic° and delivered amount cyclopentanones of complex 53achievedin up to 34% three water had no effect on the reaction yield although the catalyst synthetically useful methods based on this reaction areto provideturnovers; cyclopentanone although only a in low 72% number yield. of turnovers An experiment were syntheticallyknown.4 In general, useful the methods most significant based on advances this reaction in hydroa- are achieved,turnovers;yield together this although representedwith similar only theamounts a low first number example of cyclopropane of of turnovers catalysis products that were was destroyed by the presence of oxygen. For the most known.cylation4 In chemistry general, the have most involved significant developing advances strategies, in hydroa- orusingdidachieved,4 aoriginating catalyticnot rely this on from represented amount the presence decarbonylation of the of complex first ethylene example pathways5 (forachieved of example, (Scheme catalysis see three 3). that reactive, generally unsubstituted, substrates, catalyst loadings synthetically usefulcylationmethods, methods chemistry to limit basedthis have undesired involved on this decarbonylation developing reaction strategies, are pathway. orturnovers;SchemedidVariation not although 5).rely of on the theonly metal presence a complex low of number ethylene allowed of Milstein (for turnovers example, to use were see a of 10 mol % could be used. For 4-enals, substituents in either 4 known. In general,methods,The the review to most limit is significant this divided undesired according advances decarbonylation to in the hydroa- transformation pathway.achieved,Schemesimilar this reaction 5). represented to isolate the andfirst characterize example of an catalysis acyl rhod- that the 2- or 5-position significantly lowered the reaction cylation chemistrybeingThe have review considered, involved is divided that developing is, according intramolecular strategies, to the alkene transformation or hydroacy-did not2.2.ium(III) rely Catalytic on hydride the Systemspresence intermediate; of ethylene treatment of (for 4-pentenal example, with see efficiency. Efforts to extend the methodology to the synthesis RhCl(PMe3)3 at room temperature delivered complex 5 in of alternative ring sizes were unsuccessful. methods, to limitbeinglation, this considered, intermolecular undesired that decarbonylation alkene is, intramolecular hydroacylation, pathway. alkene etc., hydroacy- and whenScheme2.2.1.2.2. 5). Catalytic Cyclopentanone Systems Synthesis lation,needed intermolecular also by catalyst alkene type. hydroacylation, During these sections, etc., and detailed when 2.2.1. Cyclopentanone Synthesis The reviewneeded ismechanistic divided also by arguments according catalyst type. relevant to During the to transformation the these key sections, hydroacylation detailed Miller and co-workers demonstrated that intramolecular being considered,mechanisticstep that will not is, arguments be intramolecular presented; relevant rather, alkene to these the are hydroacy-key collected hydroacylation together2.2. CatalyticalkeneMiller hydroacylation and Systems co-workers could demonstrated be performed that using intramolecular substo- lation, intermolecularstepin section will alkenenot 6. be The presented; hydroacylation, scope of rather, this review these etc., is are comprehensive and collected when together from2.2.1.ichiometricalkene Cyclopentanone hydroacylation amounts Synthesis of could rhodium be complexesperformed ifusing ethylene- substo- needed also byin catalystthe section first 6.report type. The Duringscopein 1972 of thesethrough this review sections, June is comprehensive 2009. detailed A number from of saturatedichiometric solvent amounts was employed. of rhodium 4-Pentenal complexes could be if cyclized ethylene- thereviews first report that deal in with1972 C through-H activation June 2009. cover A some number aspects of tosaturated provide solvent cyclopentanone was employed. in 72% 4-Pentenal yield using could only be 10 cyclized mol mechanistic arguments relevant1 to the key hydroacylation 5 Miller and co-workers demonstrated that intramolecular reviewsof hydroacylation, that deal withas C doH several activation metal-specific cover some reviews; aspects %Wilkinson’scomplexinethylenesaturatedchloroformto provide cyclopentanone in 72% yield using only 10 mol step will not be presented; rather, these- are collected together alkene hydroacylation10 could be performed using substo- however, a comprehensive1 treatment that covers all substrate5 (Scheme%Wilkinson’scomplexinethylenesaturatedchloroform 5). Several products originating from intermo- of hydroacylation, as do several metal-specific6 reviews; in section 6. Thetypes scope and of all this catalyst review systems is comprehensive has yet to appear. fromReactionsichiometriclecular reactionamounts10 of of ethylene rhodium and pentenal complexes were if isolated ethylene- in however, a comprehensive treatment that covers all substrate (Scheme 5). Several products11,12 originating from intermo- the first reporttypesthat in 1972 generate and all through catalyst acyl metal systems June intermediates 2009. has yet A to similar number appear. to 61 ofReactionsindirectlysaturatedlowlecular solvent yield reaction (see was Scheme ofemployed. ethylene 32). 4-Pentenal and pentenal could were be isolated cyclized in reviews that dealvia with simple C carbonylation-H activation or carbonylation-type cover some aspects reactions areto provideLarock cyclopentanone observed a similar in 72%11,12 improvement yield using in the only efficiency 10 mol that generate acyl7 metal intermediates similar to 1 indirectly low yield (see Scheme 32). of hydroacylation,vianot simple1 considered.as do carbonylation several metal-specific or carbonylation-type reviews; reactions5 are%Wilkinson’scomplexinethylenesaturatedchloroformof aLarock range observed of intramolecular a similar hydroacylation improvement reactions in the efficiency upon the use of ethylene saturated solvent. In a departure from however, a comprehensivenot considered. treatment7 that covers all substrate (Schemeof a 5). range10 Several of intramolecular products hydroacylation originating from reactions intermo- upon the Miller system, he found that catalysts generated in situ types and all catalyst2. Intramolecular systems has Alkene yet to appear. Hydroacylation6 Reactions lecularthe reaction use of ethylene of ethylene saturated and solvent. pentenal In were a departure isolated from in from [Rh(COD)Cl]2 and P(4-MeC6H4)3,P(4-MeOC6H4)3,or 2. Intramolecular Alkene Hydroacylation the Miller system, he found11,12 that catalysts13 generated in situ that generate acyl2.1. metal Stoichiometric intermediates Systems similar to 1 indirectly low yieldP(4-Me (see2NC Scheme6H4)3 were 32). most efficient. With the relatively from [Rh(COD)Cl]2 and P(4-MeC6H4)3,P(4-MeOC6H4)3,or via simple carbonylation or carbonylation-type reactions are Larockhigh observed catalyst loadings a similar of 50 improvement mol % in methylene13 in the chloride efficiency at 2.1.7 Sakai Stoichiometric disclosed the Systems first example of a metal-mediated roomP(4-Me temperature,2NC6H4)3 thewere method most could efficient. be usedWith for the the synthesis relatively not considered. of3 a range of intramolecular hydroacylation reactions upon hydroacylation reaction in a study on prostanoid synthesis. ofhigh a range catalyst of loadings substituted, of 50 spirocyclic, mol % in and methylene fused cyclopen- chloride at Sakai disclosed the first example of a metal-mediatedthe useroom of ethylene temperature, saturated the method solvent. could be In used a departure for the synthesis from The reactions involved treatment of a range of 4-enals with3 tanone products (Scheme 6). No other olefin or acetylene 2. Intramolecularhydroacylationstoichiometric Alkene reaction amounts Hydroacylation in of a study Wilkinson’s on prostanoid complex synthesis. at roomthe Millerof a rangesystem, of substituted, he found that spirocyclic, catalysts and generated fused cyclopen- in situ The reactions involved treatment of a range of 4-enals with additive proved as effective. It was also reported that adding temperature and delivered cyclopentanones 3 in up to 34%from [Rh(COD)Cl]watertanone had products no effect2 and (Scheme on P(4-MeC the reaction 6). No6H yield4 other)3,P(4-MeOC although olefin the or acetylenecatalyst6H4)3,or stoichiometric amounts of Wilkinson’s complex at room yield together with similar amounts of cyclopropane productsP(4-MewasadditiveNC destroyedH proved) were by as the effective. most presence efficient. It was of also oxygen.13 With reported For the thatthe relatively mostadding 2.1. Stoichiometrictemperature Systems and delivered cyclopentanones 3 in up to 34% 2 6 4 3 4 originating from decarbonylation pathways (Scheme 3).high catalystreactive,water had loadings generally no effect unsubstituted, of on 50 the mol reaction % substrates, in yield methylene although catalyst chloride the loadings catalyst at Sakai disclosedyieldVariation together the first of with the example similar metal complex amounts of a metal-mediated allowed of cyclopropane Milstein products to use a ofwas 10 destroyedmol % could by be the used. presence For 4-enals, of oxygen. substituents For in the either most 4 originating from decarbonylation pathways (Scheme 3).room temperature, the method could be used for the synthesis hydroacylationsimilar reaction reaction in a study to isolate on prostanoid and characterize synthesis. an acyl3 rhod- thereactive, 2- or generally 5-position unsubstituted, significantly substrates, lowered catalyst the reaction loadings ium(III)Variation hydride of the intermediate; metal complex treatment allowed of Milstein 4-pentenal to use withof a a rangeefficiency.of 10 ofmol substituted, Efforts% could to be extend used. spirocyclic, the For methodology 4-enals, and substituents fused to the synthesis cyclopen- in either The reactions involved treatment of a range of 4-enals with tanone products (Scheme 6). No other olefin or acetylene similarRhCl(PMe reaction3)3 at to room isolate temperature and characterize delivered an complex acyl rhod-5 in ofthe alternative 2- or 5-position ring sizes weresignificantly unsuccessful. lowered the reaction stoichiometricium(III) amounts hydride of Wilkinson’s intermediate; treatmentcomplex of at 4-pentenal room withadditiveefficiency. proved Efforts as effective. to extend It wasthe methodology also reported to the that synthesis adding temperature andRhCl(PMe delivered3)3 at cyclopentanones room temperature3 deliveredin up to complex 34% 5 inwater hadof alternative no effect ring on the sizes reaction were unsuccessful. yield although the catalyst yield together with similar amounts of cyclopropane products was destroyed by the presence of oxygen. For the most 4 originating from decarbonylation pathways (Scheme 3). reactive, generally unsubstituted, substrates, catalyst loadings Variation of the metal complex allowed Milstein to use a of 10 mol % could be used. For 4-enals, substituents in either similar reaction to isolate and characterize an acyl rhod- the 2- or 5-position significantly lowered the reaction ium(III) hydride intermediate; treatment of 4-pentenal with efficiency. Efforts to extend the methodology to the synthesis RhCl(PMe3)3 at room temperature delivered complex 5 in of alternative ring sizes were unsuccessful. Transition Metal Catalyzed Hydroacylation Chemical Reviews, 2010, Vol. 110, No. 2 727

Table 1. Scope of Bosnich’s Cationic Rh(I) Cyclizationsa

Figure 1.

The cyclization of 4-pentenal has been achieved using only 3 mol % of the complex Co(dppe)(PPh3)2 (dppe ) 1,2- bis(diphenylphosphino)ethane) as a catalyst.14-16 The authors propose the intermediacy of a cobalt(II) acyl hydride in analogy to the related rhodium(III) intermediates. This appears to be a potentially useful system although its application to other substrates has been limited. The development of cationic rhodium complexes as intermolecular hydroacylation catalysts by Bosnich repre- sented a significant breakthrough in alkene hydroacylation chemistry because it allowed low catalyst loadings to be routinely employed.17,18 Bosnich evaluated a number of catalyst systems, based on generic structure 6,thatcombined bidentate phosphine ligands with a cationic rhodium center (Figure 1). The optimal system was found to be [Rh(dppe- )]ClO4, which could be isolated as the arene-bridged dimer 7 or generated in situ as a disolvated species corresponding to 6.19 These complexes were applied to the cyclization of 4-pentenals; the catalyst [Rh(dppe)]ClO4 was effective for the hydroacylation of a variety of differently substituted 4-pentenals (Table 1).17 As can be seen, substituents in the 2-, 3-, 4-, and 5-positions were all possible although cyclizations with the 5-substituted substrates were signifi- cantly slower and generally required a higher catalyst loading. Transition Metal Catalyzed HydroacylationForthe Chemical majority Reviews, 2010,of substrates Vol. 110, No., 2 a727 1 mol % catalyst loading Intramolecularprovided rapid reactions at room temperature. Reactions The use of Table 1. Scopeionic of Bosnich’s liquids Cationic as Rh(I) the Cyclizationsreactiona media for similar cyclizations a b 1 20 [Rh(dppe)]2(ClO4)2, CD3NO2, 20 °C. Determined by GC and H has been reported. c The majorityTransition of Metal recent Catalyzed reports featuring Hydroacylation intramolecular NMR methods. 65 °C. Chemical Reviews, 2010, Vol. 110, No. 2 727 hydroacylation have employed the cationic Rh complexes popularized by Bosnich and originally developed by Osborne Scheme 7 Table 1. Scope of Bosnich’s Cationic Rh(I) Cyclizationsa and Schrock21 as the catalysts of choice (see section 2.4 for applications). Peters has shown that zwitterionic Rh com- Figure 1. plexes can also be used as effective complexes in hydroa- 22 The cyclization of 4-pentenal has been achieved using only cylation reactions leading to cyclopentanones. For example, 3 mol % of the complex Co(dppe)(PPh3)2 (dppe ) 1,2- neutral rhodium complex 8, featuring an anionic borate bis(diphenylphosphino)ethane) as a catalyst.14-16 The authors ligand, was shown to be an effective catalyst in the propose the intermediacy of a cobalt(II) acyl hydride in analogy to the related rhodium(III) intermediates. This conversion of enal 9 into cyclopentanone 10 (Scheme 7). appears to be a potentially useful system although its Direct comparison with [Rh(dppe)(NCMe)2][PF6](11)showed application to other substrates has been limited. that the zwitterionic complex 8 delivered 40 times the The development of cationic rhodium complexes as turnover frequency. The zwitterionic complexes also dis- intermolecular hydroacylation catalysts by Bosnich repre- delivered the target compounds with fewer side products but played greaterFigure tolerance 1. to coordinating solvents, with with slower rates and poorer yields. sented a significant breakthrough in alkene hydroacylation 728 Chemical Reviews, 2010, Vol. 110, No. 2 Willis chemistry because it allowed low catalyst loadings to be effective reactions still being achieved in benzene, THF, or The majority of substituted 4-pentenals used in the Bosnich routinely employed.17,18 Bosnich evaluated a number of acetonitrile.The The cyclization cationic complex of 4-pentenal11 was poorly has activebeen achieved in study using were only prepared via Claisen rearrangements of the catalyst systems, based on generic structure 6,thatcombined these solvents. Enal 9 was the only substrateScheme evaluated. 8 Scheme 11 bidentate phosphine ligands with a cationic rhodium center 3 mol % of the complex Co(dppe)(PPhTandem )Reaction:(dppe ) 1,2- Morgan and Kundu have reported the preparation and3 2 (Figure 1). The optimal system was found to be [Rh(dppe- bis(diphenylphosphino)ethane) as a catalyst.14-16 The authors )]ClO4, which could be isolated as the arene-bridged dimer application of a neutral bimetallic rhodium/titanium complex 7 or generated in situ as a disolvated species corresponding for intramolecularpropose hydroacylation. the intermediacy23 Complex of a12 cobalt(II)(Figure 2) acyl hydride in to 6.19 These complexes were applied to the cyclization of was readilyanalogy prepared to in situ the from related the corresponding rhodium(III) hydroxy- intermediates. This 4-pentenals; the catalyst [Rh(dppe)]ClO4 was effective for phosphine,appears Ti(OiPr) to4, and be [Rh(COD)Cl] a potentially2 and was useful shown system to although its the hydroacylation of a variety of differently substituted promote the cyclization of two enals to the corresponding 17 application to other substrates has been limited. 4-pentenals (Table 1). As can be seen, substituents in the cyclopentanones. Control experiments, relative to Bosnich- 2-, 3-, 4-, and 5-positions were all possible although The development of cationic rhodium complexes as cyclizations with the 5-substituted substrates were signifi- type cationic Rh systems, showed that the new complex Figure 2. cantly slower and generally required a higher catalyst loading. intermolecular hydroacylation catalysts by Bosnich repre- For the majority of substrates, a 1 mol % catalyst loading sented a significant breakthrough in alkene hydroacylation provided rapid reactions at room temperature. The use of Organometallics 1988, 7, 936-945 ionic liquids as the reaction media for similar cyclizations chemistry because it allowed low catalyst loadings toAngew be Chem, Int. Ed. 2003, 42, 2385 a b 1 20 [Rh(dppe)]2(ClO4)2, CD3NO2, 20 °C. Determined by GC and H has been reported. c 17,18 Table 2. Effect of Diene Geometry on Cycloheptanone Synthesis The majority of recent reports featuring intramolecular NMR methods. 65 °C. routinely employed. Bosnich evaluated a number of hydroacylation have employed the cationic Rh complexes catalyst systems, based on generic structure 6,thatcombined popularized by Bosnich and originally developed by Osborne Scheme 7 and Schrock21 as the catalysts of choice (see section 2.4 for bidentate phosphine ligandsScheme with 9 a cationic rhodium center applications). Peters has shown that zwitterionic Rh com- (Figure 1). The optimal system was found to be [Rh(dppe- plexes can also be used as effective complexes in hydroa- cylation reactions leading to cyclopentanones.22 For example, )]ClO4, which could be isolated as the arene-bridged dimer neutral rhodium complex 8, featuring an anionic borate 7 or generated in situ as a disolvated species corresponding ligand, was shown to be an effective catalyst in the 19 conversion of enal 9 into cyclopentanone 10 (Scheme 7). to 6. Direct comparison with [Rh(dppe)(NCMe)2][PF6](11)showed These complexes were applied to the cyclization of that the zwitterionic complex 8 delivered 40 times the substrate 22 (%) 23 (%) 24 (%) starting material (%) turnover frequency. The zwitterionic complexes also dis- delivered the target compounds4-pentenals; with fewer side the products catalyst but Scheme [Rh(dppe)]ClO 10 4 was effective for played greater tolerance to coordinating solvents, with with slower rates and poorer yields. (4E,6E) 62 13a 6 the hydroacylation of a variety of differently substituted a effective reactions still being achieved in benzene, THF, or The majority of substituted 4-pentenals used in the Bosnich17 (4Z,6E) 65 11 5 acetonitrile. The cationic complex 11 was poorly active in 4-pentenals (Table 1). As can be seen, substituents in the study were prepared via Claisen rearrangements of the (4E,6Z) trace 49b trace 38 these solvents. Enal 9 was the only substrate evaluated. 2-, 3-, 4-, and 5-positions were all possible although Morgan and Kundu have reported the preparation and a b application of a neutral bimetallic rhodium/titanium complex cyclizations with the 5-substituted substrates were signifi- E-alkene. Z-alkene. for intramolecular hydroacylation.23 Complex 12 (Figure 2) was readily prepared in situ from the corresponding hydroxy- cantly slower and generally required a higher catalyst loading. phosphine, Ti(OiPr)4, and [Rh(COD)Cl]2 and was shown to For the majority of substrates, a 1 mol % catalyst loading Scheme 12 promote the cyclization of two enals to the corresponding cyclopentanones. Control experiments, relative to Bosnich- provided rapid reactions at room temperature. The use of type cationic Rh systems, showed that the new complex Figure 2. ionic liquids as the reaction media for similar cyclizations a b 1 20 [Rh(dppe)]2(ClO4)2, CD3NO2, 20 °C. Determined by GC and H has been reported. c The majority of recent reports featuring intramolecular NMR methods. 65 °C. hydroacylation have employed the cationic Rh complexes Scheme 7 popularized by Bosnich andappropriate originally allyl developed vinyl ethers. by Osborne Eilbracht has established that 21 and Schrock as the catalystswith ofthe choice use of (see a suitable section catalyst, 2.4 for it is possible to couple applications). Peters hashydroacylation shown that zwitterionic with a Claisen Rh rearr com-angement to achieve a one- plexes can also be used aspot effective conversion complexes of allyl vinyl in hydroa- ethers to cyclopentanones, thus cylation reactions leadingavoiding to cyclopentanones. the isolation22 For of example, potentially sensitive aldehydes.24 neutral rhodium complexScheme8, featuring 8 gives twoan anionic examples; borate the use of RhCl(COD)(dppe) It is possible to access larger ring systems using hydroa- ligand, was shown toas be a ancatalyst effective and a CO catalyst atmosphere in the was found to be optimal. cylation if suitable additional functionality is incorporated into the substrate. Mori has developed the use of dienal conversion of enal 9 intoThe cyclopentanone use of alternative10 catalysts(Scheme such 7). as RuCl2(PPh3)3 required Direct comparison with [Rh(dppe)(NCMe)2][PF6](11)showed substrates to allow access to cycloheptanone products the use of higher pressures of CO (50 bar) to achieve similar 27,28 that the zwitterionic complexconversions.8 delivered A polystyrene-supported 40 times the η5-cyclopentadienyl (Scheme 11). Treatment of dienal 17 with a rhodium(I) turnover frequency. Therhodium zwitterionic catalyst complexes has also also been dis- used in relateddelivered transforma- the target compoundscatalyst with formed fewer metallocyclohexane side products but 18 as expected; from played greater tolerancetions. to25 coordinating solvents, with with slower rates and poorerhere the yields. reaction could either proceed to the usual cyclo- effective reactions still being achieved in benzene, THF, or The majority of substitutedpentanone 4-pentenals product used or isomerize in the Bosnich to rhodacycle 19 by way of acetonitrile. The cationic2.2.2. complex Larger11 was Ring poorly Systems active in study were prepared viaπ-allylrhodium Claisen rearrangements species 20. Reductive of the elimination from 19 these solvents. Enal 9 was the only substrate evaluated. would then provide a cycloheptanone product. Morgan and Kundu haveThe reported formation the of preparation ring systems and larger than cyclopentanones The authors observed that diene geometry in substrate 21 application of a neutral bimetallicusing intramolecular rhodium/titanium hydroacylation complex is not a trivial exercise. had a significant effect on the product distribution (Table for intramolecular hydroacylation.Ring closure23 Complex of these larger12 (Figure rings 2)is generally slower than for 2); if the diene contained an E alkene at C-6 then the required was readily prepared in situfive-ring from the formation corresponding and decarbonylation hydroxy- can become problem- cycloheptanones were obtained; however, a Z alkene at C-6 atic. In addition, if a five-membered closure is possible, then produced the cyclopentanone product. Cyclopentanone for- phosphine, Ti(OiPr)4, and [Rh(COD)Cl]2 and was shown to promote the cyclization ofthis two pathway enals will to the usually corresponding be followed. For example, reaction mation was rationalized on the basis that formation of a cyclopentanones. Controlof experiments, 5-hexenal does relative not provide to Bosnich- the cyclohexanone but delivers π-allyl intermediate similar to 20 would be unfavorable for the cyclopentanone as the sole product (Scheme 9).12 type cationic Rh systems, showed that the new complex Figure 2. the Z-alkene due to steric repulsion. Provided dienes with Six-membered ring formation has been achieved using a E-geometry at C-6 were employed, the preparation of rigid carbohydrate-derived scaffold (Scheme 10).26 The cycloheptanones was efficient for a number of substrates. authors had anticipated that treatment of enal 13 with the The cationic rhodium systems favored by Bosnich were used catalyst [RhCl(PPh3)2]2 would provide a cyclopentanone; throughout. For an application of this methodology to natural however, cyclohexanone 14 was obtained as the exclusive product synthesis, see Scheme 31. product. The authors reasoned that cyclopentanone formation Shair and co-workers have developed an elegant method is disfavored due to the ring strain that would be present in for the synthesis of cyclooctenones using intramolecular the resulting 5,5,5-tricyclic product. This was reinforced by hydroacylation.29 Eight-membered ring formation was achieved the much slower formation of cyclopentanone 15 from enal by the incorporation of a cyclopropane ring in the substrate. 16. This was not a general process and was only applicable The proposed mechanism is shown in Scheme 12. The key to a limited range of substrates. step is the fragmentation and isomerization of rhodacycle 728 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 8 Scheme 11

728 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 8 Scheme 11 728 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 8 Scheme 11

Table 2. Effect of Diene Geometry on Cycloheptanone Synthesis

Scheme 9

Table 2. Effect of Diene Geometry on Cycloheptanone Synthesis Scheme 10 Tablesubstrate 2. Effect22 (%) of Diene23 (%) Geometry24 (%) on Cycloheptanone starting material Synthesis (%) Larger Rings a Scheme 9 (4E,6E) 62 13 6 (4Z,6E) 65 11a 5 Scheme 9 (4E,6Z) trace 49b trace 38 a E-alkene. b Z-alkene.

Scheme 12 substrate 22 (%) 23 (%) 24 (%) starting material (%) SchemeScheme 10 10 substrate 22 (%) 23 (%) 24 (%) starting material (%) a (4E,6(4E)E,6E) 62 62 13a 136 6 (4Z,6(4E)Z,6E) 65 65 11a 115a 5 (4E,6(4Z)E,6Z) trace trace 49b 49traceb trace 38 38 appropriate allyl vinyl ethers. Eilbracht has established that a E-alkene.a E-alkene.b Z-alkene.b Z-alkene. with the use of a suitable catalyst, it is possible to couple hydroacylation with a Claisen rearrangement to achieve a one- Scheme 12 Transition Metal Catalyzedpot conversion Hydroacylation of allyl vinyl ethers to cyclopentanones, thus Scheme 12 Chemical Reviews, 2010, Vol. 110, No. 2 729 avoiding the isolation of potentially sensitive aldehydes.24 Scheme 13 Scheme 8 gives two examples; the use of RhCl(COD)(dppe)Scheme 15It is possible to access larger ring systems using hydroa- as a catalyst and a CO atmosphere was found to be optimal. cylation if suitable additional functionality is incorporated into the substrate. Mori has developed the use of dienal Chemical Reviews, 2010, Vol. 110, No. 2 The use of alternative catalysts such as RuCl2(PPh3)3 requiredWillis 728 substrates to allow access to cycloheptanone products the use of higher pressures of CO (50 bar) to achieve similar 27,28 Scheme 8 Schemeappropriateconversions. 11 allyl A vinyl polystyrene-supported ethers. Eilbracht hasη5-cyclopentadienyl established that (Scheme 11). Treatment of dienal 17 with a rhodium(I) withrhodium the use catalyst of a suitable has also catalyst, been used it inis relatedpossible transforma- to couple catalyst formed metallocyclohexane 18 as expected; from appropriate25 allyl vinyl ethers. Eilbracht has established that here the reaction could either proceed to the usual cyclo- withhydroacylation thetions. use of a with suitable a Claisen catalyst, rearrangement it is possible to achieve to a one-couple pot conversion of allyl vinyl ethers to cyclopentanones, thus pentanone product or isomerize to rhodacycle 19 by way of hydroacylationavoiding2.2.2. Larger the with isolation Ring a Claisen Systems of potentially rearrangement sensitive to achieve aldehydes. a24 one- π-allylrhodium species 20. Reductive elimination from 19 would then provide a cycloheptanone product. potScheme conversion 8 gives of two allyl examples; vinyl ethers the use to of cyclopentanones, RhCl(COD)(dppe) thus It is possible to access larger ring systems using hydroa- The formation of ring systems larger than cyclopentanones 24 The authors observed that diene geometry in substrate 21 avoidingas a catalyst the isolation and a CO ofatmosphere potentially was sensitive found to be aldehydes. optimal. cylation if suitable additional functionality is incorporated using intramolecular hydroacylation is not a trivial exercise. hadinto a the significantIt issubstrate. possible effect Mori to on access the has product developed larger distribution ring the systems use (Table of dienal using hydroa- SchemeTheRing use 8 givesclosure of alternative two of these examples; catalysts larger rings such the is use as generally RuCl of RhCl(COD)(dppe)2(PPh slower3)3 required than for 2); if the diene contained an E alkene at C-6 then the required the use of higher pressures of CO (50 bar) to achieve similar substratescylation to if allow suitable access additional to cycloheptanone functionality products is incorporated as a catalystfive-ring and formation a CO and atmosphere decarbonylation was can found become to be problem- optimal. cycloheptanones27,28 were obtained; however, a Z alkene at C-6 conversions. A polystyrene-supported η5-cyclopentadienylScheme(Scheme 16 into 11).the substrate.Treatment Mori of dienal has developed17 with a rhodium(I) the use of dienal The useatic. of In alternative addition, if catalysts a five-membered such as closure RuCl is2(PPh possible,3)3 required then producedcatalyst theformed cyclopentanone metallocyclohexane product. Cyclopentanone18 as expected; for- from rhodiumthis pathway catalyst will has usually also beenbe followed. used in For related example, transforma- reaction mationsubstrates was rationalized to allow on the access basis that to formation cycloheptanone of a products theTable use 2. of25 Effect higher of Diene pressures Geometry of CO on Cycloheptanone (50 bar) to achieve Synthesis similar here the reaction could27,28 eitherChem proceed reviews to the 2010, usual p.725 cyclo- tions.of 5-hexenal does not provide the cyclohexanone5 but delivers (Scheme 11). Treatment of dienal 17 with a rhodium(I) conversions. A polystyrene-supported η -cyclopentadienyl πpentanone-allyl intermediate product similar or isomerize to 20 would to rhodacycle be unfavorable19 by forway of the cyclopentanone as the sole product (Scheme 9).12 the Zcatalyst-alkene due formed to steric metallocyclohexane repulsion. Provided dienes18 as with expected; from Scheme 9 rhodium catalyst has also been used in related transforma- π-allylrhodium species 20. Reductive elimination from 19 2.2.2.25 Six-membered Larger Ring Systemsring formation has been achieved using a E-geometryhere the at reaction C-6 were could employed, either the proceed preparation to the of usual cyclo- tions. 26 would then provide a cycloheptanone product. rigid carbohydrate-derived scaffold (Scheme 10). The cycloheptanonespentanone was product efficient or isomerize for a number to ofrhodacycle substrates. 19 by way of Theauthors formation had anticipated of ring systems that treatment larger than of enal cyclopentanones13 with the TheThe cationic authors rhodium observed systems that favored diene geometryby Bosnich in were substrate used 21 using intramolecular hydroacylation is not a trivial exercise. hadπ a-allylrhodium significant effect species on the20 product. Reductive distribution elimination (Table from 19 2.2.2.catalyst Larger [RhCl(PPh Ring Systems3)2]2 would provide a cyclopentanone; throughout. For an application of this methodology to natural Ringhowever, closure cyclohexanone of these larger14 ringswas is obtained generally as slower the exclusive than for product2); ifwould the synthesis, diene then contained see provide Scheme an aE 31.cycloheptanonealkene at C-6 then product. the required Thefive-ringproduct. formation formation The authors of ring and reasoned decarbonylation systems that larger cyclopentanone can than become cyclopentanones formation problem- cycloheptanonesShairThe and co-workers authors were observed have obtained; developed that however, diene an elegant a geometryZ alkene method at in C-6 substrate 21 Scheme 10 substrateatic. In addition,22 (%) if a23 five-membered(%) 24 (%) closure starting is material possible, (%) then produced the cyclopentanone product. Cyclopentanone for- usingis intramolecular disfavored due hydroacylationto thea ring strain that is not would a trivial be present exercise. in for thehad synthesis a significant of cyclooctenones effect onthe using product intramolecular distribution (Table Scheme 14 (4thisE,6E pathway) 62 will usually 13 be followed.6 For example, reaction mation was29 rationalized on the basis that formation of a Ring(4Z,6 closuretheE) resulting 65 of these 5,5,5-tricyclic larger 11a rings5 product. is generally This was slowerreinforced than by for hydroacylation.2); if the dieneEight-membered contained ring an formationE alkene was at C-6achieved then the required of 5-hexenalthe much slower does not formation provideb of the cyclopentanone cyclohexanone15 butfrom delivers enal byπ-allyl the incorporation intermediate of similar a cyclopropane to 20 would ring in be the unfavorable substrate. for five-ring(4E,6Z) formation trace and 49 decarbonylationtrace can become 38 12 problem- cycloheptanones were obtained; however, a Z alkene at C-6 the16 cyclopentanone. This was not aas general the sole process product and (Scheme was only 9). applicable The hydroacylationsThethe proposedZ-alkene mechanism due were to steric all is performed shown repulsion. in Scheme usingProvided 12. Wilkinson’s The dienes key with atic.a InESix-memberedto-alkene. addition, a limitedb Z-alkene. range if aring five-membered of formation substrates. has closure been achieved is possible, usingcomplex then a stepE-geometry as isproduced thecatalyst fragmentation at the atC-6 cyclopentanone room were and temperature employed, isomerization product. the ofin preparation rhodacycle methylene Cyclopentanone of for- thisrigid pathway carbohydrate-derived will usually be followed. scaffold (Scheme For example, 10).26 reactionThechloride.cycloheptanones Themation method was allowed was rationalized efficient a variety for on aof the number seven- basis of and that substrates. eight- formation of a ofScheme 5-hexenalauthors 12 had does anticipated not provide that treatmentthe cyclohexanone of enal 13 butwith delivers thememberedThe ringsπ cationic-allyl to intermediate berhodium prepared systems (Scheme similar favored to 16). by20 BosnichAlkynewould werebeas well unfavorable used for 12 thecatalyst cyclopentanone [RhCl(PPh as3) the2]2 would sole product provide (Scheme a cyclopentanone; 9). as alkenethroughout. containingthe Z-alkene For substrates an application due to could steric of this be repulsion. employed methodology Provided although to natural dienes with Six-memberedhowever, cyclohexanone ring formation14 was has obtained been as achieved the exclusive usingthe use a ofproduct allylE-geometry synthesis,sulfides was at see C-6unsuccessful Scheme were 31. employed, and was attributed the preparation of rigidproduct. carbohydrate-derived The authors reasoned scaffold that cyclopentanone (Scheme formation 10).26toThe competitiveShaircycloheptanones and cleavage co-workers of was the have allyl efficient developed C-S for an bond. a elegant number Control method of substrates. authorsis disfavored had anticipated due to the that ring straintreatment that would of enal be13 presentwithexperiments in the forThe the conducted synthesis cationic using rhodium of cyclooctenones substrates systems in favored using which intramolecular by the Bosnich sulfur were used the resulting 5,5,5-tricyclic product. This was reinforced by hydroacylation.29 Eight-membered ring formation was achieved catalystthe much [RhCl(PPh slower formation3)2]2 would of cyclopentanone provide a cyclopentanone;15 from enalatom wasby replaced thethroughout. incorporation by an For oxygen of an a application cyclopropane or carbon of ring atom this in methodology confirmedthe substrate. to natural appropriate allyl vinyl ethers. Eilbracht has established that however, cyclohexanone 14 was obtained as the exclusive product synthesis, see Scheme 31. with the use of a suitable catalyst, it is possible to couple 16. This was not a general process and was only applicablethat S-chelationThe proposed wasmechanism a requirement is shown for cyclization. in Scheme 12. A recent The key hydroacylation with a Claisen rearrangement to achieve a one- product.to a limited The authors range of reasoned substrates. that cyclopentanone formationreport fromstep isShair Dong, the fragmentation and concerned co-workers and with isomerization have the developed enantioselective of rhodacycle an elegant method pot conversion of allyl vinyl ethers to cyclopentanones, thus is disfavored due to the ring strain that would be presentsynthesis in offor medium-ring the synthesis ethers of cyclooctenones and sulfides, has using demon- intramolecular avoiding the isolation of potentially sensitive aldehydes.24 the resulting 5,5,5-tricyclic product. This was reinforcedstrated by thathydroacylation. it is possible29 toEight-membered employ ether-linked ring formation substrates was achieved Scheme 8 gives two examples; the use of RhCl(COD)(dppe) the muchIt is possible slower to formation access larger of ring cyclopentanone systems using15 hydroa-fromsimilar enal toby those the shown incorporation in Scheme of a cyclopropane 16 in hydroacylation ring in the substrate. as a catalyst and a CO atmosphere was found to be optimal. cylation if suitable additional functionality is incorporated 31 16into. This the was substrate. not a Mori general has process developed and the was use only of dienal applicablechemistry (seeThe Scheme proposed 29, mechanism section 2.4). is shown in Scheme 12. The key The use of alternative catalysts such as RuCl2(PPh3)3 required to a limited range of substrates. The Brookhartstep is group the fragmentation have reported and a single isomerization example of of rhodacycle the use of higher pressures of CO (50 bar) to achieve similar substrates to allow access to cycloheptanone products 27,28 Co(I)-catalyzed intramolecular hydroacylation of a vinylsi- conversions. A polystyrene-supported η5-cyclopentadienyl (Scheme 11). Treatment of dienal 17 with a rhodium(I) rhodium catalyst has also been used in related transforma- catalyst formed metallocyclohexane 18 as expected; from lane leading to the synthesis of a seven-membered ketone 32 tions.25 25 into ring-expandedhere the reaction rhodacycle could26 either.Reductiveelimination proceed to the usual cyclo- (see Scheme 57 for intermolecular examples). from 26 wouldpentanone provide product the required or isomerize cyclooctanone. to rhodacycle 19 by way of 2.2.2. Larger Ring Systems Treatmentπ of-allylrhodium the relevant species cyclopropane-containing20. Reductive elimination sub- from 19 2.3. Stereoselective Reactions strates withwould cationic then rhodium provide a catalysts cycloheptanone under product. an ethylene The formation of ring systems larger than cyclopentanones The authors observed that diene geometry in substrate 21 2.3.1. Diastereoselective Systems using intramolecular hydroacylation is not a trivialatmosphere exercise. deliveredhad a significant cyclooctenones effect on the in product good to distribution moderate (Table Ring closure of these larger rings is generally sloweryields than for (Scheme2); if 13). the diene Wilkinson’s contained an complexE alkene was at C-6 ineffective then the required The abundance of defined stereochemical motifs in natural five-ring formation and decarbonylation can become problem-as a catalyst.cycloheptanones were obtained; however, a Z alkene at C-6 products and designed bioactive molecules has driven the atic. In addition, if a five-membered closure is possible,To then exploreproduced the mechanism the cyclopentanone in operation, product. the authors Cyclopentanone studied for- development of diastereoselective and enantioselective reac- this pathway will usually be followed. For example,the reaction reactionsmation of both was the rationalizedE and Z isomers on the basis of a that deuterium- formation of a tions. Sakai and co-workers were the first to explore of 5-hexenal does not provide the cyclohexanone but delivers -allyl intermediate similar to 20 would be unfavorable for labeled substrate.π The experiments led to two D-containing 33 the cyclopentanone as the sole product (Scheme 9).12 the Z-alkene due to steric repulsion. Provided dienes with diastereoselective intramolecular hydroacylation reactions. Six-membered ring formation has been achievedproducts using a in aE ratio-geometry that exactly at C-6 matched were employed, the E/Z theratio preparation of the of They extended their early study (section 2.1) to include the rigid carbohydrate-derived scaffold (Scheme 10).substrates.26 The Thecycloheptanones authors proposed was efficient that the forE isomer a number generated of substrates. cyclization of 3,4-disubstituted 4-pentenals (Scheme 17). The authors had anticipated that treatment of enal 13 withthe expected the The cyclooctenone cationic rhodium directly, systems via favored the mechanism by Bosnich were in used reactions proceeded with good levels of diastereoselectivity catalyst [RhCl(PPh3)2]2 would provide a cyclopentanone;Scheme 12 (throughout.E-27 f 28 For, Scheme an application 14). However, of this methodology to account to natural to deliver cis-3,4-disubstituted cyclopentanones. The origin however, cyclohexanone 14 was obtained as the exclusivefor the unexpectedproduct scrambling synthesis, see of Scheme the D-label, 31. they proposed of the selectivity is attributed to allylic strain, with transition product. The authors reasoned that cyclopentanone formationisomerization ofShair the andZ co-workersalkene in haveZ-27 developedvia five-membered an elegant method state 34 being favored over 35. is disfavored due to the ring strain that would be presentrhodacycles in 29forand the30 synthesisleading of to cyclooctenones intermediate using31, featuring intramolecular the resulting 5,5,5-tricyclic product. This was reinforced by hydroacylation.29 Eight-membered ring formation was achieved The use of a relatively high loading of Wilkinson’s the much slower formation of cyclopentanone 15 froman E enal-configuredby the alkene. incorporation of a cyclopropane ring in the substrate. complex allowed a variety of synthetically useful substituents 16. This was not a general process and was only applicableBendorf andThe co-workers proposed mechanism have prepared is shown a series in Scheme of medium- 12. The key to be tolerated, including keto-, chloro-, and hydroxyl groups to a limited range of substrates. ring sulfur heterocyclesstep is the fragmentation using intramolecular and isomerization hydroacyla- of rhodacycle (Table 3). Aldehydes masked as lactol units could also be tion.30 Their strategy was based on the formation of chelation- used successfully (entry 4). In all cases, the cis diastereomers stabilized intermediates such as 32 and 33 (Scheme 15). were obtained as the sole products. Reductive elimination from rhodacycles such as 33 would It was subsequently demonstrated that the use of chiral deliver the required medium-ring systems. Rh(I) catalysts in cyclizations analogous to those shown in 730 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 17 Scheme 18

Scheme 19

Table 3. Cis-Selective Cyclopentanone Synthesisa

catalyst and substrate pairings were reversed when the neutral complexes were employed.

2.3.2. Enantioselective Systems While investigating aldehyde decarbonylation using chiral rhodium catalysts, James and Young observed the first example of enantioselective intramolecular hydroacylation via kinetic resolution.37,38 Reaction of racemic pentenal 36 with a catalyst incorporating enantiomerically pure (S,S)- chiraphos yielded the corresponding cyclopentanone with up to 69% ee (Scheme 18). This level of selectivity was only Diastereo and Enantioselectiveachieved at low conversions (17%); at conversions of a RhCl(PPh3)3 (30 mol %), CH2Cl2, rt. 50-60%, the enantioselectivity was reduced to 40%. The reactionsTransition were Metal conducted Catalyzed Hydroacylation using only 1 mol %∼ of catalyst, Chemical Reviews, 2010, Vol. 110, No. 2 731 Table 4. Stereoselective Synthesis of 3,4-Disubstituted although a temperature of 150 C was needed to achieve Table 5. Enantioselective Reactions° Using Cationic Catalysts Cyclopentanones reasonable conversions. One reason for this low reactivity catalyst loadings were needed. These selectivities follow from was the presence of the quaternary center at C-2. the desymmetrization studies described above. Bosnich has also explored kinetic resolution processes Sakai and co-workers were the first to report the cyclization using cationic complexes and incorporated D-labeling studies 34,35 of achiral 4-pentenals using enantiomerically pure catalysts. to investigate the origin of the enantioselectivity in the Several ligand systems were investigated with trans-1,2-bis- 46 entry R ligand ee (%) cyclizations of 4-pentenals. Reaction of racemic 3-phenyl- [(diphenylphosphino)methyl]cyclohexane (DIPMC) delivering pentenal 39 with Rh[(S)-BINAP]ClO delivered the expected 1a,b Me (S,S)-MeDuphos 94 4 the highest2a,b selectivities.iPr For example, (S,S)-MeDuphos cyclization of phenyl- 96 cyclopentanone 40 in 51% yield along with 42% of 4-phe- substituted3a,b pentenalcyclopentyl37 furnished ( theS,S)-MeDuphos cyclopentanone product 96 nylpentenal 41 and 7% of the isomerized alkenes 42 (Scheme in 71%4c yield andtBu 76% ee (Scheme (S)-BINAP 19).35 Relatively> high99 21). The formation of 4-phenylpentenal is consistent with entry substrate catalyst time (h) yield (%) cis/trans c catalyst5 loadingsSiMe of3 25 mol %(S were)-BINAP employed, and>99 the the mechanism proposed by Bosnich for achiral hydroacy- 13R Rh[(S)-BINAP]ClO4 1.5 85 <1:>99 6c Ph (S,S)-chiraphos 78 reactionsc were conducted at room temperature. lation (see Scheme 77). The conclusion reached by Bosnich 23R Rh[(R)-BINAP]ClO4 4 74 97:3 7 4-MeO-Ph (S,S)-chiraphos 75 c 33S Rh[(R)-BINAP]ClO4 2 86 <1:>99 Both8 Sakai andC(O)Me Bosnich found (S)-BINAP that the use of cationic 87 is that no single step is responsible for the stereoinduction, c 43S Rh[(S)-BINAP]ClO4 5 82 96:4 rhodium9 complexesC(O)Ph incorporating (S)-BINAP chiral ligands allowed 94 rather a number of reversible steps all contribute. Given the b 10 CO2Et (S)-BINAP >99 complexity of the mechanism, it was not possible to propose highly enantioselectiveb cyclizations of achiral 4-pentenals 11 CO2iPr (S)-BINAP >99 asimplemodelthataccountsfortheobservedenantioselectivity.47,48 Table 3 allowed double diastereoselection to be explored, using low catalyst loadings. In the first report, Sakai found a PF - salt (5 mol %) used. b Acetone. c CH Cl . The desymmetrization systems described by Sakai (Table employingfirstneutralandthencationicrhodiumcomplexes.34-36 that the6 use of the catalyst [Rh(BINAP)]ClO2 2 4 allowed the cyclization of certain substrates to deliver enantioselectivities 6) are less complex in that isomerization products corre- Sakai found that the use of the cationic complex Rh(BI- up to 99%.36 Later, Bosnich and co-workers evaluated a range sponding to aldehydes 41 and 42 were not seen. Conse- NAP)ClO allowed the selective synthesis of all four possible 4 of bidentate chiral ligands and found that BINAP, chiraphos, quently the authors proposed a series of models based on stereoisomers of 3,4-disubstituted cyclopentanones (Table the steric interactions between the BINAP ligand and the 36 and DuPhos all generated selective catalysts for certain 49 4). For example, reaction of the (3R)-substrate with 5 mol substrates (Table 5).39-41 An impressive range of substituents substrates and the corresponding rates of reaction. % of the catalyst employing (S)-BINAP delivered the trans could be tolerated ranging from sterically demanding electron- configured product with >99:1 selectivity. Conversely, cy- 2.4. Applications of Intramolecular Alkene donating groups such as SiMe3 to muchChem smaller reviews electron- 2010, p.725 clization of the same substrate with the (R)-catalyst provided withdrawing groups such as C(O)Me. In all cases, conver- Hydroacylation the cis product with 97:3 selectivity. The situation was sions of >95% were achieved with enantioselectivities for The majority of reports discussed in the preceding section reversed for the (3S)-configured substrate. A similar pattern thefor majority substrates of substrates with tertiary falling alkyl in theor ester 94-99% and ketone range. substit- Aryl- have focused on either reaction or catalyst development. of selectivity was obtained with the neutral complexes substituteduents, while products DuPhos delivered was superior the lowest for selectivities substrates (70 containing-75% Given the success of a number of these systems, it is not Rh(BINAP)Cl; however, significantly higher catalyst load- ee).primary To obtain or secondary the optimal alkyl levels groups. of yield and enantioselec- surprising that these methods are starting to be employed as ings (50 mol %) were needed, and the levels of selectivity tivity, it was crucial to match ligand selection to individual key reactions in synthetic schemes. For example, the ability were inferior. Importantly, the matched and mismatched classThe of substrate. Sakai group For further example, investigated BINAP was the cyclization most effective of the 4-aryl-substituted pentenals using BINAP-derived catalysts to construct cyclopentanone systems efficiently using simple and discovered an unusual effect of ortho-halo substituents.42 acyclic precursors and low catalyst loadings has resulted in In general, cyclization of 4-aryl-susbtrates with (R,R)- a number of applications of the reactions to the synthesis of configured catalyst delivered (R)-configured products. How- target systems. Sakai’s discovery of the hydroacylation ever, when substrates containing an ortho-halo substituent reaction was achieved during an investigation into the were employed the (S)-configured products were obtained. synthesis of prostanoid derivatives, and in a series of The highest selectivities obtained were 83% ee. publications, he has continued these endeavors. In particular, he has exploited the availability of enantiomerically pure The Sakai and Suemune groups significantly extended the enals from terpenes such as limonene and limonen-10-ol in scope of these enantioselective cyclizations by adopting an anumberofsyntheses.Forexample,(-)-limonen-10-ol was enantioselective desymmetrization approach and employing converted through a series of steps to provide enantiomeri- substrates that led to the formation of two stereocenters cally pure enal 43 (Scheme 22).50 Treatment of enal 43 with 43,44 (Table 6). Cationic rhodium perchlorate BINAP com- Wilkinson’s complex provided ketone 44 in 88% yield as a plexes delivered trans-configured products in good yields single diastereomer. Ketone 44 was then advanced to a key with high enantioselectivity, while the corresponding neutral intermediate in the synthesis of 11-deoxyprostaglandin. complexes provided the cis-cyclopentanones, again with good Similar sequences have been employed to prepare intermedi- yields and selectivities. The neutral complexes required 50 ates in routes toward iridomyrmecin, isoiridomyrmecin,51 mol % catalyst loadings compared with the 5 mol % carbacyclin,52 8-isoprostanoic acid53 and the nepetalactones.54 employed with the cationic systems; however, by correct Eilbracht has utilized his tandem Claisen rearrangement/ catalyst selection, it was possible to produce any of the four hydroacylation sequences in the synthesis of a number of possible product stereoisomers. The method was further key intermediates for natural product syntheses. For example, extended to deliver products containing a quaternary stereo- apotentialintermediate(45)forthesynthesisofsolavetivone center; however, with these substrates the neutral complexes was prepared in a tandem sequence starting from vinyl ether were no longer active catalysts. The cationic complexes 46 (Scheme 23). Treatment of ether 46 with RhCl(COD- delivered the trans-configured products with excellent enan- )(dppe) in benzonitrile and 150 °Cprovidedcyclopentanone tioselectivities. 45 in 35% yield by way of aldehyde 47.55 The relatively The same group also explored the kinetic resolution of low yield obtained for the tandem conversion was attributed nonsymmetric racemic diene-aldehydes (Scheme 20).45 Re- to acid-promoted decomposition. The same group has used action of aldehyde 38 with Rh[(S)-BINAP]ClO4 delivered asimilarstrategyintheformalsynthesisoferythrodieneand trans-substituted cyclopentanone in 46% yield and an excel- spirojatamol.56 lent >95% ee. The use of the corresponding neutral com- Undheim and co-workers have utilized cyclopentanone- plexes allowed access to the cis-configured products with forming hydroacylation reactions in the synthesis of spirane- good enantioselectivity although the regioselectivity between bridged bisarenes57 and in the preparation of spirane-bridged the Ph- and Me-substituted alkenes was poor, and high indenes needed for the preparation of semititanocenes.58 For 732 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

732 Chemical Reviews, 2010, Vol. 110, No. 2 Table 6 Willis

Table 6

entry R catalyst time (h) yield (%) cis/trans ee (%)a config 1b Me [Rh((R)-BINAP)]Cl 72 25 97:3 >95 3S,4R 2b Me [Rh((S)-BINAP)]Cl 72 31 97:3 >95 3R,4S 3 Me [Rh((R)-BINAP)]ClO4 1 81 3:97 >95 3S,4S entry R catalyst time (h) yield (%) cis/trans ee (%)a config 4 Me [Rh((S)-BINAP)]ClO4 1 84 4:96 >95 3R,4R 1b Me [Rh((R)-BINAP)]Cl 72 255 Ph 97:3 [Rh((S)-BINAP)]ClO>954 3S,42R 70 18:82 >95 3S,4S 6 Ph [Rh((R)-BINAP)]ClO4 2 76 17:83 >95 3R,4R 2b Me [Rh((S)-BINAP)]Cl 72 31 97:3 >95 3R,4S 7 iBu [Rh((R)-BINAP)]ClO4 3 74 2:98 >95 3S,4S 3 Me [Rh((R)-BINAP)]ClO4 1 818 iBu 3:97 [Rh((S)-BINAP)]ClO>954 3S,44S 77 2:98 >95 3R,4R 4 Me [Rh((S)-BINAP)]ClO4 1 84 4:96 >95 3R,4R a Of major product. b Catalyst used at 50 mol %. 5 Ph [Rh((S)-BINAP)]ClO4 2 70 18:82 >95 3S,4S 6 Ph [Rh((R)-BINAP)]ClO4 2 76 17:83 >95 3R,4R 7 iBu [Rh((R)-BINAP)]ClO4 3Scheme 74 20 2:98 >95 3S,4S Scheme 23 732 Chemical Reviews, 2010, Vol. 110, No. 2 Willis 8 iBu [Rh((S)-BINAP)]ClO4 4 77 2:98 >95 3R,4R a Of major product. b Catalyst used at 50 mol %. Table 6

Scheme 20 Scheme 23

Scheme 21 entry R catalyst time (h) yield (%) cis/trans ee (%)a config 1b Me [Rh((R)-BINAP)]Cl 72 25 97:3 >95 3S,4R 2b Me [Rh((S)-BINAP)]Cl 72 31 97:3 >95 3R,4S 3 Me [Rh((R)-BINAP)]ClO4 1 81 3:97 >95 3S,4S Application4 Me [Rh((S)-BINAP)]ClO 4 1 84 4:96 >95 3R,4R 5 Ph [Rh((S)-BINAP)]ClO4 2 70 18:82 >95 3S,4S Scheme 21 6 Ph [Rh((R)-BINAP)]ClO4 2 76 17:83 >95 3R,4R 7 iBu [Rh((R)-BINAP)]ClO 3 74 2:98 95 3S,4S Scheme 22 4 > 8 iBu [Rh((S)-BINAP)]ClO4 4 77 2:98 >95 3R,4R a Of major product. b Catalyst used at 50 mol %. Scheme 24

Scheme 20 Scheme 23

Scheme 22 example, enal 48 was converted to ketone 49 using a dppe- derived catalyst (Scheme 24). Importantly, to achieve good Schemeyields 24 the reactions needed to be performed under solvent- free conditions for eight days. The use of benzonitrile or Scheme 21 nitromethane as solvent resulted in significant decarbonyla- tion. The second example also employed solvent-free condi- tions and shows the preparation of ketone 50 from enal 51, in 90% yield. In this second example, the use of benzonitrile as solvent lead to the formation of a significant byproduct. These two examples both employ complex enal substrates, example, enal 48 was converted to ketone 49 using a dppe- featuring quaternary carbon centers, a ketal, and an indene ring system and, despite the relatively high temperatures and derived catalyst (Scheme 24). Importantly, to achieve good Scheme 22 alcohol 55, an intermediate in the synthesis of brefeldin A extended reaction times needed, illustrate the good functional yields the reactions needed to be performed under solvent- previously reported by Gais.60 group tolerance of the hydroacylation process. Scheme 24 free conditions for eight days. The use of benzonitrile or Akineticresolutionusingacationicrhodiumcomplexwas Related enal systems were recently employed by Castillo´n nitromethane as solvent resulted in significant decarbonyla- employed by Rousseau and Mioskowski in a formal synthesis in the enantioselective synthesis of carbocyclic nucleosides tion. The second example also employed solvent-free condi- of brefeldin A (Scheme 25).59 Reaction of racemic aldehyde (Scheme 26).61 For example, treatment of enal 56 with a tions and shows the preparation of ketone 50 from enal 51, 52 with Rh[(S)-BINAP]ClO4 (0.9 mol %) triggered an cationic Me-DuPhos-containing catalyst, as described by in 90% yield. In this second example, the use of benzonitrile intramolecular hydroacylation reaction to deliver a 1:1 Bosnich, delivered cyclopentanone 57 with >95% ee. Ad- as solvent lead to the formation of a significant byproduct. mixture of cyclopentanones 53 and 54; both were obtained vancement of ketone 57 throughChem reduction reviews and2010, a p.725 lipase- withexample, 96% enal enantiomeric48 was converted enrichment. to ketone The 49transusing-configured a dppe- mediated dynamic kinetic resolution ultimately allowed These two examples both employ complex enal substrates, productderived catalyst was advanced (Scheme through 24). Importantly, a short sequence to achieve to deliver good access to carbocyclic nucleoside 58. featuring quaternary carbon centers, a ketal, and an indene yields the reactions needed to be performed under solvent- ring system and, despite the relatively high temperatures and free conditions for eight days. The use of benzonitrile or alcohol 55, an intermediate in the synthesis of brefeldin A extended reaction times needed, illustrate the good functional nitromethane as solvent resulted in significant decarbonyla- previously reported by Gais.60 group tolerance of the hydroacylation process. tion. The second example also employed solvent-free condi- Relatedtions and enal shows systems the preparation were recently of ketone employed50 from by enal Castillo51, ´n Akineticresolutionusingacationicrhodiumcomplexwas in 90% yield. In this second example, the use of benzonitrile employed by Rousseau and Mioskowski in a formal synthesis in the enantioselective synthesis of carbocyclic nucleosides as solvent61 lead to the formation of a significant byproduct. of brefeldin A (Scheme 25).59 Reaction of racemic aldehyde (SchemeThese 26). two examplesFor example, both employ treatment complex of enal enal substrates,56 with a 52 with Rh[(S)-BINAP]ClO4 (0.9 mol %) triggered an cationicfeaturing Me-DuPhos-containing quaternary carbon centers, catalyst, a ketal, andas described an indene by ring system and, despite the relatively high temperatures and intramolecular hydroacylation reaction to deliver a 1:1 Bosnich, delivered cyclopentanone 57 with >95% ee. Ad-alcohol 55, an intermediate in the synthesis of brefeldin A extended reaction times needed, illustrate the good functional vancement of ketone 57 through reduction and a lipase-previously reported by Gais.60 mixture of cyclopentanones 53 and 54; both were obtained group tolerance of the hydroacylation process. with 96% enantiomeric enrichment. The trans-configured mediatedAkineticresolutionusingacationicrhodiumcomplexwas dynamic kinetic resolution ultimately allowed Related enal systems were recently employed by Castillo´n product was advanced through a short sequence to deliver accessemployed to carbocyclic by Rousseau nucleoside and Mioskowski58. in a formal synthesis in the enantioselective synthesis of carbocyclic nucleosides of brefeldin A (Scheme 25).59 Reaction of racemic aldehyde (Scheme 26).61 For example, treatment of enal 56 with a 52 with Rh[(S)-BINAP]ClO4 (0.9 mol %) triggered an cationic Me-DuPhos-containing catalyst, as described by intramolecular hydroacylation reaction to deliver a 1:1 Bosnich, delivered cyclopentanone 57 with >95% ee. Ad- mixture of cyclopentanones 53 and 54; both were obtained vancement of ketone 57 through reduction and a lipase- with 96% enantiomeric enrichment. The trans-configured mediated dynamic kinetic resolution ultimately allowed product was advanced through a short sequence to deliver access to carbocyclic nucleoside 58. 734 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 30 Scheme 32 734 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 30 Scheme 32

Intermolecular Hydroacylation Scheme 33 Scheme 31 Scheme 33 Scheme 31

Scheme 34 Scheme 34

cess to thecess required to the required fused-ring fused-ring system. system. In order In order to achieve to achieve the second of these two stages, it was necessary to include the second of these two stages, it was necessary to include aldehydesaldehydes to to contain contain an an alkene alkene unit unit suggested suggested that that internal internal a quaternarya quaternary center in center the tether. in the tether. For example, For example, treatment treatment of of alkene coordination was a prerequisite for reactivity. alkene coordination was a prerequisite for reactivity. Chem reviews 2010, p.725 tetraene 68 with [Rh(dppe)]ClO4 at 65 °C resulted in initial Marder and Millstein extended the utility of intermolecular tetraene 68 with [Rh(dppe)]ClO4 at 65 °C resulted in initial hydroacylation to generate ketone 69, followed by cycloi- Marderhydroacylation and Millstein reactions extended with the the discovery utility that of intermolecular the indenyl- hydroacylationsomerization to generate to deliver ketone the isomeric69, followed ketones by70 cycloi-and 71 in 5 hydroacylationrhodium complex reactions [Rh(η with-C9H the7)(C discovery2H4)2]catalyzedtheaddi- that the indenyl- somerization to deliver the isomeric ketones 70 and 71 in 5 19% and 7% yield, respectively. When a substrate lacking rhodiumtion of complex a range of [Rh( aromaticη -C9H aldehydes7)(C2H4) to2]catalyzedtheaddi- ethylene; a single 19% andthe 7% quaternary yield, respectively. center was employed, When a substrate the initial hydroacyla-lacking tionexample of a range is shown of aromatic in Scheme aldehydes 33.70 Provided to ethylene; relatively a high single the quaternarytion proceeded center was as expected; employed, however the initial the cycloisomerization hydroacyla- examplepressures is shown of ethylene in Scheme (1000 psi 33. at70 25Provided°C) were relatively employed, high 1 tion proceededfailed. asPerforming expected; the however cascade process the cycloisomerization at reflux temperature pressuresturnover of numbers ethylene of (1000 ca. 4 psi h- atwere 25 ° observedC) were employed,with no failed. Performingallowed ketone the cascade70 to be process obtained at as reflux the exclusive temperature product turnoverdecomposition numbers of the of catalyst ca. 4 h to-1 metallicwere rhodium. observed A with range no allowed ketonein 44%70 yield.to Ketone be obtained70 was as advanced the exclusive through product an 11-step of simple aromatic aldehydes were employed, and methyl sequence to deliver epiglobulol. decomposition of the catalyst to metallic rhodium. A range in 44% yield. Ketone 70 was advanced through an 11-step formate was also shown to produce hydroacylation adducts, ofalthough simple aromatic at a much aldehydes slower rate were (2-3 employed, turnovers in and 24 h). methyl sequence to deliver epiglobulol. 3. Intermolecular Alkene Hydroacylation formateThe was first alsointermolecular shown to hydroacylation produce hydroacylation systems, described adducts, althoughby Miller, at68 aemployed much slower enals rate as the (2 aldehyde-3 turnovers component in 24 (see h). 3. Intermolecular3.1. Rhodium Alkene Catalysis Hydroacylation TheScheme first 32), intermolecular and their success hydroacylation was attributed systems, to intramo- described bylecular Miller, coordination68 employed of enals the alkene as the to aldehyde aid in catalyst component stabiliza- (see Extending hydroacylation chemistry to intermolecular Schemetion. The 32), use and of their heteroatom success chelation was attributed has emerged to intramo- as a 3.1. Rhodiumsystems has Catalysis proven to be challenging, with the main difficulty lecularsuccessful coordination strategy of for the intermolecular alkene to aid hydroacylation in catalyst stabiliza- and a Extendingremaining hydroacylation the prevention chemistry of decarbonylation to intermolecular pathways. number of systems have been reported. Suggs first isolated During Miller’s intramolecular study using pentenal sub- tion. The use of heteroatom chelation has emerged as a systems has proven to be challenging, with the main difficulty successfulacyl rhodium(III) strategy complex for intermolecular71 from the reaction hydroacylation of quinoline- and a strates (see Scheme 5), he observed that exchanging the 8-carboxaldehyde and Wilkinson’s complex, and found that remainingcatalyst the prevention from Wilkinson’s of decarbonylation complex to Rh(acac)(C pathways.H ) number of systems have been reported. Suggs first isolated During Miller’s intramolecular study using pentenal sub-2 4 2 after treatment with AgBF4 and octene, the hydroacylation allowed products from the intermolecular reaction with acyladduct rhodium(III) could be isolated complex (Scheme71 from 34). the71,72 reactionLater, modification of quinoline- strates (see Scheme 5), he observed that exchanging10,68 the ethylene to be isolated (Scheme 32). For example, 8-carboxaldehydeof the method to and employ Wilkinson’s [RhCl(COE) complex,2]2 as and the found catalyst that catalystreaction from Wilkinson’s of pentenal complex with Rh(acac)(C to Rh(acac)(C2H4)2 in2 ethylene-H4)2 afterfollowed treatment by treatment with AgBF with4 trimethylphosphiteand octene, the hydroacylation allowed this allowedsaturated products chloroform from the provided intermolecular a mixture of reaction heptenones, with with adductstoichiometric could be isolated process to (Scheme be applied 34). to71,72 a rangeLater, of modification electroni- the E-hept-5-en-2-one as the major10,68 component in 39% yield. ethylene to be isolated (Scheme 32). For example, ofcally the method varied alkenes, to employ delivering [RhCl(COE) hydroacylation2]2 as adducts the catalyst in Cyclopentanone was detected in only 1% yield. Reactions 73 reaction of pentenal with Rh(acac)(C2H4)2 in ethylene- followedyields byup treatment to 75%. withAttempts trimethylphosphite to use substoichiometric allowed this saturatedwith chloroform saturated provided aldehydes a provided mixture no of hydroacylation heptenones, with adducts; stoichiometricamounts of Wilkinson’s process to be complex applied resulted to a range in significantly of electroni- however, 4-hexenals were successfully employed. Subse- reduced yields. the E-hept-5-en-2-one as the major component in 39% yield. cally varied alkenes, delivering hydroacylation adducts in quently, Vora showed that Z-4-heptenal could also be Arelatedprotocol,catalyticinrhodium,employedthe73 Cyclopentanonecombined was with detected ethylene, in using only identical 1% yield. conditions, Reactions to deliver yieldschelating up abilityto 75%. of a phosphineAttempts group to to use allow substoichiometric intermolecular with saturated6-nonen-3-one aldehydes in provided a 26% yield. no hydroacylation69 The requirement adducts; for the amountshydroacylation; of Wilkinson’so-diphenylphosphino complex resulted benzaldehyde in significantly could be however, 4-hexenals were successfully employed. Subse- reduced yields. quently, Vora showed that Z-4-heptenal could also be Arelatedprotocol,catalyticinrhodium,employedthe combined with ethylene, using identical conditions, to deliver chelating ability of a phosphine group to allow intermolecular 6-nonen-3-one in a 26% yield.69 The requirement for the hydroacylation; o-diphenylphosphino benzaldehyde could be 736 Chemical Reviews, 2010, Vol. 110, No. 2 Willis Transition Metal Catalyzed Hydroacylation Chemical Reviews, 2010, Vol. 110, No. 2 737

Scheme 39 Scheme 42 SchemeScheme 40 45

Transition MetalChelation Catalyzed Hydroacylation Assisted Chemical Reviews, 2010, Vol. 110, No. 2 735 Scheme 35 Table 7.

Scheme 46

Scheme 43

Scheme 36 entrya R1 R2 yield (%) SchemeScheme 41 47 loadings to those in the basic process (Table 7) were 1 Ph Bu 98 employed, with 2-amino-4-picoline taking the place of 2 Ph Pr 83 Transition Metal Catalyzed Hydroacylation Chemical Reviews, 2010, Vol. 110, No. 2 737 2-amino-3-picoline. A related process employing allylic 3 Ph Hex 99 pyridyl amines as substrates has also been reported.89 4 Ph tBu 84 Scheme 42 Scheme 45 5 Ph SiMe 95 Afurthersimplificationoftheprocesswastoallowsimple 3 6 Ph C6F5 98 alcohols to be employed as the aldehyde precursors (Scheme 7 Ph PhOCH2 95 90 39). The rhodiumScheme complex 44 first catalyzes the oxidation of 8 4-MeO-C6H4 Bu 79 the alcohol to the aldehyde, using an equivalent of alkene 9 4-CF3-OC6H4 Bu 71 as the oxidant, before promoting the hydroacylation reaction In section 2.2.2, an example10 of S-chelation 4-Me2N- toC6 facilitateH4 Bu 60 the preparation of a variety11 of cyclic ketones 4-Ph-C was6H4 described Bu 95 as described previously. The most efficient catalyst was found b (Scheme 16). The Willis12 group has exploited PhCH2CH S-chelation2 to Bu 71 to be RhCl3 · H2O (3.3 mol %), used in combination with develop intermolecular reactivity. !-Methylsulfide-substituted triphenylphosphine and a stoichiometric amount of 2-amino- propanal could be combineda Five with equivalents a range of of alkenefunctionalized used. b Ten percent of an aldol product also isolated. 4-picoline. Slight modification of the system allowed primary electron-poor alkenes to generate hydroacylationChem reviews 2010, adducts p.725 in 112 amines to be employed as aldehyde equivalents.91 Methanol good yields (Scheme 45). The authors proposed the combinedScheme with 46 a range of neutral alkenes using 5formation mol % of of a five-membered S-chelate-stabilized acyl 74 Scheme 37 could also be employedWilkinson’s as a substrate, complex generating as catalyst formal- in benzene as 90rhodium°C. An species to account for the observed reactivity. The 92 dehyde in situ, and thenexample undergoing reaction double is hydroacylation. shown in Scheme 35.cyclizationuse of to [Rh(dppe)]ClO generate4 theallowed natural reactions product. to be performed Despite at the The Jun group has exploitedWhile theproviding excellent insight metal-directing into the requirementsrelatively for50 stabiliza-° lowC. The yield use obtained of simple electron-neutral for the hydroacylation alkenes, such step as of effect of picolyl imines to develop a range of cascade process the synthesis,1-octene, resulted no epimerization in reduced yields. was reported, and the Scheme 43 tion of the key rhodium-acyl units, the use of quinoline-The same and group has extended the concept of S-chelate involving C-H and C-C bond activation; hydroacylation/ methodology allowed the use of commercial aldehydes. The 93 phosphine-substituted aldehydes94 were limiting instabilization that the and introduced !-thioacetal-substituted alde- ortho-alkylation, ring-openingproducts necessarilyof cycloalkenones, incorporatedand the these largeJun methodology coordinatinghydes as effective has hydroacylation also been applied substrates to (Scheme the synthesis 46).113,114 of conversion of dienesgood to yields cycloalkanones to be realized;95 howeverare some mixtures notable of linear and18 101 groups.Scheme The 47 Jun group significantly advancedF-labeledThe chelation- complex ketones [Rh(dppe)]ClOand to4 was the again functionalization used as catalyst to of examples. branched isomers were obtained. The use of 1,6-dienespolymers.promote102,103 reaction with electron-poor alkenes. It was dem- producedstabilized extremely hydroacylation low yields of adducts.methods A with variety the of realization that Efforts to aid catalyst recovery have also been reported. Severalonstrated account that the and thioacetals review present articles in the documenting hydroacylation the 2-hydroxypicolyl aldehydes imines, could formed be employed in situ, successfully; could serve how- asadducts chelating- could be derivatized by hydrolysis to reveal the For example, polystyrene-supported diphenylphosphine75 was ever,aldehyde the absence equivalents. of a free hydroxyl Suggs grouphad in this previously positiondevelopment showncorresponding and that application carbonyl or of reduced picolyl with imine Raney intermediates nickel to 6,104-107 used in combinationresulted with in RhCl no reaction.3 · H2O as a tandem alcohol in hydroacylationdeliver the methylene and related derivatives. reactions Both are available. the sulfide- and imines such96 as 72 formed stable rhodium-iminoacyl com- oxidation/hydroacylationModificationplexes catalyst. and of the exploited catalystThe catalyst allowed these in the could intermolecular selective be produc-The hydroacylation Miurathioacetal-substituted group has undertaken aldehydes have detailed also been studies combined on the recovered and usedtion four of either times regioisomer; before deactivation. the use of [RhCl(COE) More 2]2, P(useo- ofsuccessfully salicylaldehyde with 1,2- as and a 1,3-disubstituted chelating substrate allenes forto deliver alkyne reactions (Scheme 36). Treatment of the imines with Wilkin- 115 recently, the Jun groupMePh) has3, AgClO described4, and theNaOAc use delivered of a H-bonding the linear isomer, !,γ-enones as products. 108,109 Scheme 44 son’s complex (5 mol %) and an alkene inhydroacylation THF at elevated (see SchemeScheme 71). 38 However, reactions solvent system to allowwhile catalyst the use recycling of [RhCl(COE) in primary2]2 in combination alcohol withof the sameWeller aldehyde and Willis with have the recently majority reported of an alkenes improved were NH(itemperature,Pr)2 as base delivered delivered, the branched after isomer hydrolysis, (Scheme the 43). hydroacylationcatalyst system that allows poorly reactive neutral alkenes hydroacylation systems. A 4,4In section!-bipyridyl/phenol 2.2.2, an example solvent of S-chelationpoor.109 toThe facilitate exception was reaction with vinyltriethylsilylane, The authorsadducts. also Isolated undertook labeling picolyl studies imines using have 1-deuter- been utilizedto be combined in with !-S-aldehydes in good yields (Scheme the preparation of a variety of cyclic ketones was described116 combination generatedated aldehydes a homogeneous and observed system considerable at elevated scrambling, withwhich,47). whenFor76 combined-78 example, with reaction salicylaldehyde between !-MeS-propanal and a catalyst and hydroacylation-basedsynthesesofacylferrocenederivatives(Scheme 16). The Willis group has exploited S-chelation to temperatures; however,deuterium upon incorporation cooling, two seen immiscible separately phases in the R-methylgeneratedhexene, from employing [RhCl(COD)] a catalyst2 and incorporating the ligand the 1,1 ligand!-bis(diphe- DPEp- groupand, and when!develop-methylene enoates intermolecular position were of employed reactivity. the products as!-Methylsulfide-substituted theshown alkene in components,hos, delivered the hydroacylation adduct in 70% yield. The were generated, allowing easy catalyst and product separa- nylphosphino)ferrocene79 (dppf), provided the hydroacylation tion. This approachScheme allowedin the43.propanal A synthesis up proposed to eight could mechanism of cycles be1,4-dicarbonyl combined isof shown catalysis with in systems. Scheme a range 44;adduct of functionalizedauthors in good propose yield (Scheme that the flexible, 41). Simple hemilabile allenes nature were of the also doubleThe chelationelectron-poor Jun97 group involving was alkenes two-point able to to generate demonstrate interactions hydroacylation of that the it was adductsDPEphos possible in ligand allows it to adopt a number of coordination with minimal loss of activity. A similar concept has112 been shown to be competent substrates, as illustrated by the rhodium withgood both yields the aldehyde (Scheme and98,99 the 45). dieneThe and aauthors series proposedmodes and the geometries and ultimately provides a longer-lived used to develop aof H-bonding reversibleto generate steps catalyst suitable leading system; to picolyl an irreversiblethe imines high in reductive situformation fromcatalyst, the of corre-!, resultingγ-enone in74 increased.Theuseofrefluxingtoluenewas activity. In addition to the spondingformation aldehydes of a five-memberedand amines.80 S-chelate-stabilizedThe optimized catalyst acyl temperatures neededelimination forrhodium reaction are proposed species disrupt to account to account a H-bonding for the for high the observedreactivityneeded reactivity.greater in order The activity to achieve achieved full with conversion the DPEphos for catalyst both system, types of and deuteriumsystem involved incorporation treating seen with a mixture this system. of the aldehydethe and ability the to generate an active catalyst from the simple network between 2,6-diaminouse of pyridine [Rh(dppe)]ClO and a4 barbiturate-allowed reactions toreaction. be performed The authors at proposed the intermediacy of a chelate bound rhodium, allowingalkene efficient50 with°C. Thehydroacylation Wilkinson’s use of simple complex reactions. electron-neutral (2 mol %),such alkenes, picolyl as 75 such amineto as account for the minimal decarbonylation Cooling the reactions(20 to mol1-octene, room %), temperature benzoic resulted inacid reduced allows (6 mol yields. the %), andobserved. aniline (60 mol H-bonds to reform and%) generates in tolueneThe same two at 130 group immiscible°C(Table7). has extended phases.81 By the this concept method,More of S-chelate recently a varied the Suemunebyproduct group was has investigated also isolated the together use with the expected Efficient catalyst recyclingrangestabilization was of alkenes again possible. and aldehydesintroduced ! could-thioacetal-substituted be combinedof salicylaldehydes in alde- good in combinationproduct. This with diene byproduct substrates. formation110,111 is typical when alkyl 113,114 The Kim group hasto utilized excellenthydes the as yields. Jun effective methodology Heteroaromatic hydroacylation in the substrates aldehydesAs (Scheme82 canand be 46). enoates seen from83 thealdehydes examples are presented used as in substrates Scheme 42, with the conditions shown good yields to be realized; however mixtures of linear and The complex [Rh(dppe)]ClO1004 was again used as catalyst to branched isomers were obtained.syntheses The use of of 1,6-dienes indolizidinescould (- also)-167B be employed and (-)-209D. as substrates.The it was possible to usein very Table mild 7. To conditions address this to issue, achieve the Jun group has recently Jun conditions were translatedpromote directly reaction and with applied electron-poor to the alkenes.excellent It was yields dem- for the addition of salicylaldehydes to a produced extremely low yields of adducts. A variety of Experimentationonstrated that the had thioacetals established present that in the the formation hydroacylation of the reported a modified procedure, employing cyclohexylamine 2-hydroxy aldehydes could be employedunion successfully; of enantiomerically how-picolyl pureadducts imine alkene could was73 the beand slowderivatized heptanal step of by to the hydrolysis originalvariety to process. reveal of 1,4- the The and 1,5-dienes.and p-trifluoromethylbenzoic The use of Wilkinson’s acid in place of aniline and 87 ever, the absence of a free hydroxyldeliver group the in this corresponding positiondescribed ketonecorresponding multicomponent in 35% yield carbonyl (Scheme catalystor reduced 40). mixture withcomplex Raney was needed nickel (20 mol to to %) allowedbenzoic the acid, reactions respectively. to be performedThese modified conditions resulted in no reaction. Reductive conditions (Hpromote2, Pd/C)deliver the allowed the facile methylene CBZ transimination cleavage derivatives. and mechanism, Bothat the ambient sulfide-described temperature. and in Employingallowed the these efficient conditions use of allowed alkyl aldehydes by limiting the Modification of the catalyst allowed the selective produc-Schemethioacetal-substituted 37, which provided aldehydes more rapid have access also been to the combined required formation of aldol side products. tion of either regioisomer; the use of [RhCl(COE)2]2, P(o-picolylsuccessfully imines. The with 1,2- overall and 1,3-disubstituted system is noteworthy allenes to deliver for the In efforts to expand the utility of the process the Jun group 115 MePh)3, AgClO4, and NaOAc delivered the linear isomer,relatively!,γ-enones low amounts as products. of rhodium complex employed and has extended their methodology to allow the direct use of while the use of [RhCl(COE)2]2 in combination withthe excellentWeller yields and Willis obtained. have Recently, recently reported the use an of improved montmo- allyl alcohols as substrates.88 In this new system, the same NH(iPr) as base delivered the branched isomer (Scheme 43). catalyst system that allows poorly reactive neutral alkenes 2 rillonite K10 clay as a cocatalyst to aid in imine formation rhodium catalyst was employed and mediates an initial allyl The authors also undertook labeling studies using 1-deuter- to be combined84 with !-S-aldehydes in good yields (Scheme ated aldehydes and observed considerable scrambling, withhas been47).116 reported,For example,as reaction has the between use of! microwave-MeS-propanal heating and alcohol to aldehyde isomerization, before hydroacylation 85,86 deuterium incorporation seen separately in the R-methylunderhexene, solvent-free employing conditions. a catalyst incorporating the ligand DPEp- takes place. As can be seen from Scheme 38, a good selection group and !-methylene position of the products shown in Thehos, final delivered example the in hydroacylation Table 7 employed adduct anin 70% alkyl yield. aldehyde The of substituted allyl alcohols could be combined with a range Scheme 43. A proposed mechanism is shown in Scheme 44;as theauthors substrate propose and that is the notable flexible, because hemilabile an aldol-derivednature of the of alkyl aldehydes. Similar reaction conditions and catalyst double chelation involving two-point interactions of the DPEphos ligand allows it to adopt a number of coordination rhodium with both the aldehyde and the diene and a series modes and geometries and ultimately provides a longer-lived of reversible steps leading to an irreversible reductive catalyst, resulting in increased activity. In addition to the elimination are proposed to account for the high reactivity greater activity achieved with the DPEphos catalyst system, and deuterium incorporation seen with this system. the ability to generate an active catalyst from the simple Transition Metal Catalyzed Hydroacylation Chemical Reviews, 2010, Vol. 110, No. 2 735

Scheme 35 Table 7.

Transition Metal Catalyzed Hydroacylation Chemical Reviews, 2010, Vol. 110, No. 2 735

Scheme 35 Table 7.

Scheme 36 entrya R1 R2 yield (%) 1 Ph Bu 98 2 Ph Pr 83 3 Ph Hex 99 4 Ph tBu 84 Scheme 36 entrya R1 R2 yield (%) 5 Ph SiMe3 95 6 Ph C6F5 198 Ph Bu 98 7 Ph PhOCH2 295 Ph Pr 83 8 4-MeO-C6H4 Bu3 79 Ph Hex 99 4 Ph tBu 84 9 4-CF3-OC6H4 Bu 71 5 Ph SiMe3 95 10 4-Me2N-C6H4 Bu 60 6 Ph C6F5 98 11 4-Ph-C6H4 Bu7 95 Ph PhOCH 95 12 PhCH CH Bu 71b 2 2 2 8 4-MeO-C6H4 Bu 79 9 4-CF -OC H Bu 71 a Five equivalents of alkene used. b Ten percent of an aldol product3 6 4 10 4-Me2N-C6H4 Bu 60 also isolated. 11 4-Ph-C6H4 Bu 95 12 PhCH CH Bu 71b combined with a range of neutral alkenes using 5 molAldimines % of 2 2 Wilkinson’s complex as catalyst in benzene as 90 C.74 An Scheme 37 a Five equivalents of alkene used. b Ten percent of an aldol product ° also isolated. example reaction is shown in Scheme 35. combined with a range of neutral alkenes using 5 mol % of While providing insight into the requirements for stabiliza- 74 Scheme 37 tion of the key rhodium-acyl units, the use of quinoline-Wilkinson’s and complex as catalyst in benzene as 90 °C. An example reaction is shown in Scheme 35. phosphine-substituted aldehydes were limiting in that the While providing insight into the requirements for stabiliza- products necessarily incorporated these large coordinatingtion of the key rhodium-acyl units, the use of quinoline- and groups. The Jun group significantly advancedphosphine-substituted chelation- aldehydes were limiting in that the stabilized hydroacylation methods with the realizationproducts that necessarily incorporated these large coordinating picolyl imines, formed in situ, could serve asgroups. chelating- The Jun group significantly advanced chelation- aldehyde equivalents. Suggs75 had previouslystabilized shown that hydroacylation methods with the realization that imines such as 72 formed stable rhodium-iminoacylpicolyl com- imines, formed in situ, could serve as chelating- 75 plexes and exploited these in intermolecular hydroacylationaldehyde736 Chemical equivalents. Reviews, Suggs2010, Vol.had 110, previously No. 2 shown that Willis reactions (Scheme 36). Treatment of the imines withimines Wilkin- such as 72 formed stable rhodium-iminoacyl com- plexesScheme and 39 exploitedScheme these in 38 intermolecular hydroacylation Scheme 40 son’s complex (5 mol %) and an alkene in THFreactions at elevated (Scheme 36). Treatment of the imines with Wilkin- temperature, delivered, after hydrolysis, the hydroacylationson’s complex (5 mol %) and an alkene in THF at elevated Scheme 38 adducts. Isolated picolyl imines have beentemperature, utilized in delivered, after hydrolysis, the hydroacylation hydroacylation-basedsynthesesofacylferrocenederivativesadducts.76- Isolated78 picolyl imines have been utilized in and, when enoates were employed as the alkene components,hydroacylation-basedsynthesesofacylferrocenederivatives76-78 in the synthesis of 1,4-dicarbonyl systems.79 and, when enoates were employed as the alkene components, 79 The Jun group was able to demonstrate that it wasin the possible synthesis of 1,4-dicarbonyl systems. to generate suitable picolyl imines in situ from theThe corre- Jun group was able to demonstrate that it was possible 80 to generate suitable picolyl imines in situ from the corre- sponding aldehydes and amines. The optimizedsponding catalyst aldehydes and amines.80 The optimized catalyst system involved treating a mixture of the aldehydesystem and involved the treating a mixture of the aldehyde and the alkene with Wilkinson’s complex (2 mol %), picolylalkene amine with Wilkinson’s complex (2 mol %), picolyl amine (20 mol %), benzoic acid (6 mol %), and aniline(20 (60 mol mol %), benzoic acid (6 mol %), and aniline (60 mol %) in toluene at 130 °C(Table7).81 By this method,%) in a variedtoluene at 130byproduct°C(Table7). was81 By also this method, isolated a varied together withbyproduct the expected was also isolated together with the expected range of alkenes and aldehydes could be combinedrange in of good alkenes andproduct. aldehydes This could byproduct be combined formation in good is typicalproduct. when This byproductalkyl formationChem is typical reviews when2010, p.725 alkyl 82 to excellent83 yields. Heteroaromatic aldehydes82 and enoates83 aldehydesScheme 41are used as substrates with the conditions shown to excellent yields. Heteroaromatic aldehydes andloadings enoates to thosealdehydes in the basic are used process as substrates (Table 7) with were the conditions shown could also be employed as substrates. could also be employedin Table as substrates. 7. To address this issue, the Junin group Table has 7. To recently address this issue, the Jun group has recently employed,Experimentation with had2-amino-4-picoline established that the taking formation the place of the of reported a modified procedure, employing cyclohexylamine Experimentation had established that the formation2-amino-3-picoline. of the reported A related a modified process procedure, employing employing allylic cyclohexylamine picolyl imine was the slow step of the original process.89 The and p-trifluoromethylbenzoic acid in place of aniline and picolyl imine was the slow step of the original process.describedpyridyl The amines multicomponent asand substratesp-trifluoromethylbenzoic catalyst has also mixture been was reported. needed acid to in placebenzoic of aniline acid, respectively. and 87 These modified conditions 87 described multicomponent catalyst mixture waspromote neededAfurthersimplificationoftheprocesswastoallowsimple the to facile transiminationbenzoic acid, mechanism, respectively. describedThese in modifiedallowed the conditions efficient use of alkyl aldehydes by limiting the promote the facile transimination mechanism, describedSchemealcohols 37, to in which be employed providedallowed as more the aldehyde rapid efficient access precursors use to ofthe alkyl required (Scheme aldehydesformation by limiting of aldol the side products. Scheme 37, which provided more rapid access to thepicolyl39). required90 The imines. rhodium Theformation overall complex system first of aldol catalyzes is noteworthy side the products. oxidation for the of In efforts to expand the utility of the process the Jun group picolyl imines. The overall system is noteworthyrelativelythe for alcohol the low to amounts the aldehyde,In of efforts rhodium using to expand complex an equivalent the employed utility of alkeneof and the processhas extended the Jun theirgroup methodology to allow the direct use of relatively low amounts of rhodium complex employedtheas excellent the oxidant, and yields beforehas obtained. promoting extended Recently, the their hydroacylation the methodology use of montmo- reaction to allowallyl the alcohols direct use as substrates. of 88 In this new system, the same rilloniteas described K10 clay previously. as a cocatalyst The most to efficient aid in imine catalyst formation88 was found rhodium catalyst was employed and mediates an initial allyl the excellent yields obtained. Recently, the use of montmo- 84allyl alcohols as substrates. In this new system, the same hasto beenbe RhCl reported,3 · H2O (3.3as has mol the %), use used of microwave in combination heating with alcohol to aldehyde isomerization, before hydroacylation rillonite K10 clay as a cocatalyst to aid in imine formation rhodium catalyst85,86 was employed and mediates an initial allyl 84 undertriphenylphosphine solvent-free conditions. and a stoichiometric amount of 2-amino- takes place. As can be seen from Scheme 38, a good selection has been reported, as has the use of microwave4-picoline. heating Slight modificationalcohol to of aldehyde the system isomerization, allowed primary before hydroacylation 85,86 The final example in Table 7 employed an alkyl aldehyde of substituted allyl alcohols could be combined with a range under solvent-free conditions. asamines the substrate to be employed andtakes is notable as place. aldehyde As because can equivalents. be an seen aldol-derived from91 Methanol Scheme 38,of alkyl a good aldehydes. selection Similar reaction conditions and catalyst The final example in Table 7 employed an alkylcould aldehyde also be employedof substituted as a substrate, allyl alcohols generating could formal- be combined with a range 92 as the substrate and is notable because an aldol-deriveddehyde in situ, andof then alkyl undergoing aldehydes. double Similar hydroacylation. reaction conditionscyclization and catalyst to generate the natural product. Despite the The Jun group has exploited the excellent metal-directing relatively low yield obtained for the hydroacylation step of effect of picolyl imines to develop a range of cascade process the synthesis, no epimerization was reported, and the involving C-H and C-C bond activation; hydroacylation/ methodology allowed the use of commercial aldehydes. The ortho-alkylation,93 ring-opening of cycloalkenones,94 and the Jun methodology has also been applied to the synthesis of conversion of dienes to cycloalkanones95 are some notable 18F-labeled ketones101 and to the functionalization of examples. polymers.102,103 Efforts to aid catalyst recovery have also been reported. Several account and review articles documenting the For example, polystyrene-supported diphenylphosphine was development and application of picolyl imine intermediates 6,104-107 used in combination with RhCl3 · H2O as a tandem alcohol in hydroacylation and related reactions are available. oxidation/hydroacylation catalyst.96 The catalyst could be The Miura group has undertaken detailed studies on the recovered and used four times before deactivation. More use of salicylaldehyde as a chelating substrate for alkyne recently, the Jun group has described the use of a H-bonding hydroacylation (see Scheme 71).108,109 However, reactions solvent system to allow catalyst recycling in primary alcohol of the same aldehyde with the majority of alkenes were hydroacylation systems. A 4,4!-bipyridyl/phenol solvent poor.109 The exception was reaction with vinyltriethylsilylane, combination generated a homogeneous system at elevated which, when combined with salicylaldehyde and a catalyst temperatures; however, upon cooling, two immiscible phases generated from [RhCl(COD)]2 and the ligand 1,1!-bis(diphe- were generated, allowing easy catalyst and product separa- nylphosphino)ferrocene (dppf), provided the hydroacylation tion. This approach allowed up to eight cycles of catalysis adduct in good yield (Scheme 41). Simple allenes were also with minimal loss of activity.97 A similar concept has been shown to be competent substrates, as illustrated by the used to develop a H-bonding catalyst system;98,99 the high formation of !,γ-enone 74.Theuseofrefluxingtoluenewas temperatures needed for reaction disrupt a H-bonding needed in order to achieve full conversion for both types of network between 2,6-diamino pyridine and a barbiturate- reaction. The authors proposed the intermediacy of a chelate bound rhodium, allowing efficient hydroacylation reactions. such as 75 to account for the minimal decarbonylation Cooling the reactions to room temperature allows the observed. H-bonds to reform and generates two immiscible phases. More recently the Suemune group has investigated the use Efficient catalyst recycling was again possible. of salicylaldehydes in combination with diene substrates.110,111 The Kim group has utilized the Jun methodology in the As can be seen from the examples presented in Scheme 42, syntheses of indolizidines (-)-167B and (-)-209D.100 The it was possible to use very mild conditions to achieve Jun conditions were translated directly and applied to the excellent yields for the addition of salicylaldehydes to a union of enantiomerically pure alkene 73 and heptanal to variety of 1,4- and 1,5-dienes. The use of Wilkinson’s deliver the corresponding ketone in 35% yield (Scheme 40). complex (20 mol %) allowed the reactions to be performed Reductive conditions (H2, Pd/C) allowed CBZ cleavage and at ambient temperature. Employing these conditions allowed 580 C. GONZÁLEZ-RODRÍGUEZ AND M. C. WILLIS

78 % ee) were obtained for reactions employing (S)-BINAP (6) as the chiral ligand; however, the reac- tions were not completely regioselective, and linear, achiral products were also isolated. The Tanaka group has reported a highly enantioselective Rh-catalyzed intermolecular hydro- acylation process using unfunctionalized aldehydes (7) and 1,1-substituted acrylamides (8) as the reac- tion components (Scheme 6) [14]. In this system, instead of relying on chelating aldehydes to stabilize Stereoselectivethe key reaction intermediates, stabilization was achieved by bidentate coordination of the acrylamide substrates to the metal center [15]. In addition, this coordination was also invoked to account for the Intermolecularhigh enantioselectivity observed (up to 98reactions % ee). The reactions employ the P-chiral bisphosphine lig- and QuinoxP* (9).

Scheme 6 Tanaka’s acrylamide-based enantioselective hydroacylation.

ENANTIOSELECTIVE INTERMOLECULARVery limited in HYDROACYLATIONscope, still REACTIONS EMPLOYING ␤-S-CHELATINGneeds ALDEHYDES a lot of work In 2008 our group reported an enantioselective intermolecular hydroacylation reaction employing β-S-substituted aldehydes in combination with 1,3-disubstituted allenes [16]. Excellent enantio- selectivities and yields were obtained using a Me-Duphos-derived cationic rhodiumPure Appl. Chem.catalyst Vol. [17]. 83, p. 577,By 2011 employing 1,3-disubstituted allenes 11 as more reactive equivalents of disubstituted alkenes, it was pos- sible to obtain chiral β,γ-enones 12 via a dynamic kinetic asymmetric transformation involving in situ racemization of the allenes (Scheme 7). The process was more selective when the rigid aromatic alde- hyde 10 was employed, allowing ees as high as 96 % to be obtained. Both electron-withdrawing or -donating groups could be tolerated on the allene substrates. Extension of the chemistry to tri-sub- stituted allene substrates resulted in products with high ee, but unfortunately in only low yields.

Scheme 7 The Willis group’s enantioselective allene hydroacylation chemistry.

© 2011, IUPAC Pure Appl. Chem., Vol. 83, No. 3, pp. 577–585, 2011 742 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 66 Scheme 68 Alkynes Intramolecular Transition Metal Catalyzed Hydroacylation Chemical Reviews, 2010, Vol. 110, No. 2 741

Scheme 61 Scheme 63

Scheme 67

Transition Metal Catalyzed Hydroacylation Chemical Reviews, 2010, Vol.Scheme 110, No. 64 2 741 Scheme 62 Scheme 61 Scheme 63 During their study of S-chelation-assisted intramolecular Scheme 69 alkene hydroacylation reactions toward medium-ring systems (see Scheme 16), the Bendorf group also reported several examples of intramolecular alkyne hydroacylations.30 Seven- and eight-membered R-alkylidenes arising from cis addition to the alkynes were obtained in moderate to good yields. Scheme 70 4.2. Enantioselective Processes

Scheme 64 TheScheme Fu group 65 has investigated the use of enantiomerically Scheme 62 pure Rh catalysts in the cyclization of ynals.143 Because no 3 sp centers are generated in theseChem reactions, reviews 2010, the p.725 group has also shown, confirming the high selectivity of the process. investigated kinetic resolution and desymmetrization proce- The group found that it was necessary to employ aromatic dures. Racemic 4-alkynals bearing substituents in the 3-posi- aldehyde 88 to obtain high enantioselectivity. The authors tion underwent smooth kinetic resolution, providing enan- also established that the process was not a simple kinetic tiomerically enriched cyclopentenones with good selectivities; resolution of the starting racemic allene but rather proceeded by a dynamic kinetic asymmetric transformation, involving an example transformation is shown in Scheme 66. A racemization of the allene during the course of the reaction. methoxy substituent was employed to allow a two-point Scheme 65 substrate-catalyst interaction. Catalysts incorporating the 4. Intramolecular Alkyne Hydroacylation ligand iPr-DuPhos were optimal for substrates incorporating tertiaryrhodium-catalyzed centers, while thosetrans withaddition a quaternary intramolecular center hydroacy- gave 5. Intermolecular Alkyne Hydroacylation also shown, confirming the high selectivity of the process. highestlation selectivity of 4-alkynals using BINAP-derived has been the catalysts. subject of Selectivity a theoretical 4.1. Catalytic Achiral Systems 141 The group found that it was necessary to employ aromatic factorsstudy. (s) of 20-40 could be achieved, allowing either the 5.1. Catalytic Achiral Processes aldehyde 88 to obtain high enantioselectivity.Perhaps The due authors to the desire to generate organic molecules The Tanaka group has studied the cyclization of 5- and also established that the process was not a simple kinetic unreacted starting material or cyclized product to be obtained Although not employed in alkene hydroacylation reactions, containing stereogenic centers, the study of alkynes in with6-alkynals high levels ofand stereoinduction. discovered complementary Achiral substrates reactivity bearing to the Ni catalysts have been shown to be effective for intermo- resolution of the starting racemic allenehydroacylation but rather proceeded reactions has received significantly less Fu chemistry described above;142 they observed cis addition by a dynamic kinetic asymmetric transformation, involving a quaternary center were employed in desymmetrization lecular hydroacylation of alkynes. Tsuda demonstrated that attention than the alkene-based process. An early report from to alkynes to generate exo-methylene cycloakanones. For racemization of the allene during the course of the reaction. reactions. The use of Tol-BINAP-derived catalysts allowed Ni(0) complexes could be used to combine simple aryl and Larock noted that attempts to promote intramolecular hy- example, treatment of a range of 5-alkynals with [Rh(BI- enantiomerically enriched cyclopentenones to be obtained alkyl aldehydes with symmetrical internal alkynes (Scheme 4. Intramolecular Alkyne Hydroacylationdroacylation of a range of ynals using Wilkinson’s complex NAP)]BF4 generated R-alkylidenecyclopentanones in good 69).145 The use of n-alkyl phosphines as ligands allowed rhodium-catalyzed 13trans addition intramolecularwith high levelshydroacy- of selectivity in good yield; an illustrative type catalysts were uniformly unsuccessful. However, the yields (Scheme 63). The corresponding dppe-derived catalyst formation of the enone products with high E-selectivity. Aryl Fu group has more recentlylation reported of 4-alkynals the cyclization has been the of subjectexamplewas of a significantlyis theoretical shown in Schemeless efficient. 67. The reactions were conducted 4.1. Catalytic Achiral Systems 141 or bulky alkyl phosphines generated mixtures of enone and 4-alkynals using cationic rhodiumstudy. complexes.137 Treatment at room temperature, and substitution at the 3-, 4-, and During their investigation into the kinetic resolution dienone products. Unsymmetrical alkynes led to regioiso- Perhaps due to the desire to generateof a organic range of molecules 4-alkynals withThe [Rh(dppe)] Tanaka group(BF has) in studied acetone the cyclization6-positions of 5- was and possible although the use of propargylic ether containing stereogenic centers, the study of alkynes in 6-alkynals and discovered2 4 2 complementarychemistry reactivity described to the above, the Fu group discovered a rare meric product mixtures. Although no detailed mechanistic hydroacylation reactions has receivedat room significantly temperature less providedFu chemistry good yields described of above; the desired142 they observedsubstituentscis addition (4-position) resulted in reduced144 yields. Reaction studies were undertaken, the authors propose a mechanism cyclopentenones (Scheme 61). An impressive range of exampleof corresponding of a catalytic parallel 6-alkynals kinetic under resolution. identical conditionsTreatment gener- attention than the alkene-based process. An early report from to alkynes to generate exo-methyleneof cycloakanones. racemic 4-alkynals For such as 90 with a catalyst incorporating involving an acyl nickel hydride intermediate. Larock noted that attempts to promotesubstituents intramolecular could hy- be tolerated,example, including treatment pendent of a alkene range of and 5-alkynalsated withR-alkylidenecyclohexanone [Rh(BI- products (Scheme 64). This Tol-BINAP led to the formation of cyclopentenone 91 Jun has applied his rhodium-based methodology to the droacylation of a range of ynals usingalkyne Wilkinson’s functionalities. complex TheNAP)]BF choice of4 catalystgenerated andR-alkylidenecyclopentanones solvent was represented in good a significant advance because until this report intermolecular hydroacylation of alkynes. The catalytic 13 together with cyclobutenone 92 (Scheme 68). Both cycloal- type catalysts were uniformly unsuccessful.found toHowever, be crucial the to theyields success (Scheme of these 63). The reactions. corresponding For dppe-derivedthe routine catalyst preparation of six-membered ketones using system generated from Wilkinson’s complex, 2-amino-3- kanones were obtained with high ee. The two products were Fu group has more recently reportedexample, the cyclization the use of of the equivalentwas significantly BINAP-derived less efficient. catalyst The reactionshydroacylation were conducted chemistry had relied upon conformational methylpicoline, and benzoic acid was effective for the 4-alkynals using cationic rhodium complexes.137 Treatment at room temperature, and substitution at the 3-, 4-, and in combination with methylene chloride as solvent lead to proposedrestrictions to arise or chelation from differing assistance modes (see of section addition 2.2.2). of At- combination of aryl and alkyl aldehydes with terminal of a range of 4-alkynals with [Rh(dppe)]alternative2(BF4)2 dienein acetone products,6-positions originating was from possible an isomerization although the userhodium of propargylictempts hydride to etherextend across the the methodology alkyne unit of to intermediate the cyclization93; of a 146 at room temperature provided good yields of138,139 the desired substituents (4-position) resulted in reduced yields. Reaction alkynes (Scheme 70). Interestingly, the major products in process. cis addition7-alkynal, leads leading ultimately to the to the formation cyclobutenone, of a seven-membered while trans the majority of examples studied were the branched R,!- cyclopentenones (Scheme 61). An impressiveIn order torange achieve of theof corresponding 4-alkynal to 6-alkynals cyclopentenone under identical conditionsring, were gener- unsuccessful and resulted in the return of starting substituents could be tolerated, including pendent alkene and ated R-alkylidenecyclohexanone productsaddition (Scheme leads 64). to This the cyclopentenone product. The process enones 94.Selectiveformationofthelinearisomerswasonly alkyne functionalities. The choice of catalystconversion, and solvent the hydroacylation was represented mechanism a significant must involve advance an becausecouldmaterial. until tolerate thisa report variety of aryl, hetero aryl, and alkyl groups possible with the use of alkynes bearing sterically demanding found to be crucial to the success ofunusual these reactions.trans addition For ofthe the routine rhodium preparation hydride across of six-membered the at the ketonesDuring 5-position, these using while studies, it the was Tanaka necessary group to observed maintain that a when substituents in combination with alkyl aldehydes. Achieving example, the use of the equivalent BINAP-derivedalkyne (Scheme catalyst 62). Labelinghydroacylation and cross-over chemistry experiments had relied3-methoxy uponthe conformational 5- group and 6-alkynal to achieve cyclizations efficient parallel were kinetic conducted resolution. at elevated selectivity for formation of the branched enones across a in combination with methylene chloridesupport as solvent this hypothesis. lead to Arestrictions single example or chelation of a similar assistancetrans (see sectiontemperatures, 2.2.2). At- a double-bond migration occurred to deliver alternative diene products, originatingaddition from an isomerization process has alsotempts been observed to extend bythe the methodology Nicolaou to thethe cyclization corresponding of a cyclopentenones and cyclohexenones. The process.138,139 group; during their studies7-alkynal, on the synthesis leading to of the dynemicin, formation of a seven-memberedoptimized procedure involved performing the whole process In order to achieve the 4-alkynalan attempted to cyclopentenone aldehyde decarbonylation,ring, were unsuccessful employing and stoichio- resulted in the returnat 80 of°C starting (Scheme 65). Experimentation established the conversion, the hydroacylation mechanismmetric must amounts involve of an Wilkinson’smaterial. complex, resulted in the double-bond migration was rhodium-mediated and not a unusual trans addition of the rhodiumformation hydride of across an the alkyne hydroacylationDuring these studies, product. the Tanaka140 The group observedsimple that thermal when isomerization. alkyne (Scheme 62). Labeling and cross-over experiments the 5- and 6-alkynal cyclizations were conducted at elevated support this hypothesis. A single example of a similar trans temperatures, a double-bond migration occurred to deliver addition process has also been observed by the Nicolaou the corresponding cyclopentenones and cyclohexenones. The group; during their studies on the synthesis of dynemicin, optimized procedure involved performing the whole process an attempted aldehyde decarbonylation, employing stoichio- at 80 °C (Scheme 65). Experimentation established the metric amounts of Wilkinson’s complex, resulted in the double-bond migration was rhodium-mediated and not a formation of an alkyne hydroacylation product.140 The simple thermal isomerization. 742 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 66 Scheme 68

DOI: 10.1002/anie.201105795 Heterocycle Synthesis An Alkyne Hydroacylation Route to Highly Substituted Furans** Scheme 67 Philip Lenden, David A. Entwistle, and Michael C. Willis*

Heterocyclic compounds, such as the bioactive natural catalyst systems, and generate no by-products because of their products pukalide and nakadomarin A, and the hugely 100% atom-economy.[3–5] In addition, a broad range of successful drug molecules ranitidine (Zantac) and atorvasta- aldehydes can now be employed, and particularly in the tin (Lipitor) are of great importance in pharmaceutical, case of alkyne hydroacylation, significant substitution of the agrochemical, and other fine-chemical applications unsaturatedScheme component 69 can be tolerated, thus allowing for [1] During their study of S-chelation-assisted intramolecular (Scheme 1). alkeneWhile hydroacylation there exists reactions a range toward of medium-ring methods for systemsthe regioselective production of highly substituted complex (see Scheme 16), the Bendorf group also reported severalmolecules in one catalytic intermolecular carbon–carbon examples of intramolecular alkyne hydroacylations.30 Seven-bond forming step. Herein, we demonstrate the utility of and eight-membered R-alkylidenes arisingAlkynes from cis additionintermolecular Intermolecular alkyne hydroacylation in the efficient syn- to the alkynes were obtained in moderate to good yields. thesis of di- and trisubstituted furans and related heterocycles. Scheme 70 g-Hydroxy-a,b-enones are known to undergo acid-cata- 4.2. Enantioselective Processes lyzed dehydrative cyclization to form furans, and this trans- The Fu group has investigated the use of enantiomericallyformation has been exploited by several research groups,[6] 143 pure Rh catalysts in the cyclization of ynals. Because nomost notably in the recent work from Donohoe et al.[7] The sp3 centers are generated in these reactions, the group has investigated kinetic resolution and desymmetrization proce-intermolecular hydroacylation of an aldehyde with readily dures. Racemic 4-alkynals bearing substituents in the 3-posi-available propargylic alcohols would permit the synthesis of tion underwent smooth kinetic resolution, providing enan-g-hydroxy-a,b-enones with 100% atom efficiency; coupling tiomerically enriched cyclopentenones with good selectivities;this carbon–carbon bond formation with an acid-catalyzed an example transformation is shown in Scheme 66. Adehydrative cyclization would allow for the regioselective methoxy substituent was employed to allow a two-pointsynthesis of di- or trisubstituted furans. The associated Scheme 1. Examplessubstrate of significant-catalyst heterocycle-containing interaction. Catalysts natural incorporating prod- the ucts and pharmaceuticals.ligand iPr-DuPhos were optimal for substrates incorporatingdisconnection is novel for this type of heterocycle (Scheme 2). tertiary centers, while those with a quaternary center gave Our5. initialIntermolecular investigations Alkyne to Hydroacylation realize the above route to highest selectivity using BINAP-derived catalysts. Selectivityfurans focused on the combination of propargyl alcohol 1a 5.1. Catalytic Achiral Processes transforming relativelyfactors (s) of complex 20-40 could starting be achieved, materials allowing intoApplication: sub- either the stituted heterocycles,unreacted and starting for thematerial functionalization or cyclized product of existing to be obtained Although not employed in alkene hydroacylation reactions, heterocycles, therewith high is less levels precedent of stereoinduction. for methodology Achiral substrates which bearing Ni catalysts have been shown to be effective for intermo- a quaternary center were employed in desymmetrization lecular hydroacylation of alkynes. Tsuda demonstrated that involves the direct,reactions. regiodefined The use of Tol-BINAP-derived synthesis of highly catalysts substi- allowed Ni(0) complexes could be used to combine simple aryl and [2] tuted heterocyclesenantiomerically from simple enriched starting cyclopentenones materials. As to be such, obtained alkyl aldehydes with symmetrical internal alkynes (Scheme 145 new methods forwith the high synthesis levels of of selectivity substituted in good heterocycles yield; an illustrative (or 69). The use of n-alkyl phosphines as ligands allowed their precursors)example are potentiallyis shown in Scheme of significant 67. value. This is formation of the enone products with high E-selectivity. Aryl Schemeor 2. bulkyAn alkyne alkyl hydroacylation phosphines generated route to mixtures furans. of enone and particularly trueDuring if such their methods investigation do not suffer into the from kinetic the same resolution dienone products. Unsymmetrical alkynes led to regioiso- drawbacks aschemistry the traditional described above,syntheses the Fu of group 1,4-dicarbonyl discovered a rare meric product mixtures. Although no detailed mechanistic Chem reviews 2010, p.725 compounds, theexample classic of a catalytic substrates parallel for kinetic the resolution. preparation144 Treatment of and thestudiesS-chelating were undertaken, alkyl aldehyde the authors2a propose(Table a 1). mechanism The hydro-Angew. Chem. Int. Ed. 2011, 50, 10657 furans, thiophenes,of racemic and 4-alkynals pyrroles; such these as 90 syntheseswith a catalyst are incorporating often acylativeinvolving union an of acyl these nickel two hydride substrates intermediate. using a dppe-derived Jun has applied his rhodium-based methodology to the either step- orTol-BINAP atom-inefficient. led to The the recent formation advances of cyclopentenone achieved 91Rh catalyst proceeded without incident. However, attempts together with cyclobutenone 92 (Scheme 68). Both cycloal- intermolecular hydroacylation of alkynes. The catalytic in intermolecular alkene and alkyne hydroacylation chemis- to achievesystem a dehydrative generated from cyclization Wilkinson’s using complex, TFA provided 2-amino-3- only kanones were obtained with high ee. The two products were methylpicoline, and benzoic acid was effective for the try means thatproposed these transformations to arise from differing are now modes ideal ofmethods addition ofa small amount of the desired furan 3a (entry 1). Reducing to exploit for the synthesis of heterocyclic molecules, because the timecombination for the of cyclization aryl and alkylevent aldehydes to 1 hour, with and terminal then to rhodium hydride across the alkyne unit of intermediate 93; alkynes (Scheme 70).146 Interestingly, the major products in they employ simplecis addition substrates leads ultimately and commercially to the cyclobutenone, available while trans10 minutes,the majority increased of examples the yield studied of furan were up the to branched 50% (entriesR,!- 2 addition leads to the cyclopentenone product. The processand 3).enones After94 exploring.Selectiveformationofthelinearisomerswasonly the use of several alternative acids, it could tolerate a variety of aryl, hetero aryl, and alkyl groupswas foundpossible that with the the use use of ofp alkynes-TSA bearing increased sterically the yield demanding to 66% [*] P. Lenden, Dr.at theM. C. 5-position, Willis while it was necessary to maintain a(entriessubstituents 4–7). Given in combination the known with alkylacid aldehydes. sensitivity Achieving of simple 3-methoxy group to achieve efficient parallel kinetic resolution. selectivity for formation of the branched enones across a Department of Chemistry, University of Oxford alkyl-substituted furans we speculated that the purification of Chemistry Research Laboratory the furan product by chromatography on silica gel might be Mansfield Road, Oxford, OX1 3TA (UK) responsible for the moderate yields of the isolated products.[8] E-mail: [email protected] Homepage: http://mcwillis.chem.ox.ac.uk/MCW/Home.html Accordingly, although the use of neutral alumina offered no Dr. D. A. Entwistle advantage, purifications employing Florisil (magnesium sili- Research API, Pfizer Global Research and Development cate) or triethylamine-doped silica allowed the furan to be Sandwich, Kent, CT13 9NJ (UK) isolated in significantly increased yields (entries 8–10). [**] This work was supported by the EPSRC and Pfizer. With the optimized conditions established, a short series Supporting information for this article is available on the WWW of 2,5-disubstituted furans was prepared (Scheme 3). Prop- under http://dx.doi.org/10.1002/anie.201105795. argylic aryl and heteroaryl substituents could be introduced

Angew. Chem. Int. Ed. 2011, 50, 10657 –10660  2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 10657 Highlights

DOI: 10.1002/anie.201000159 Phthalide Synthesis Catalytic Intramolecular Ketone Hydroacylation: Enantioselective Synthesis of Phthalides Michael C. Willis* asymmetric catalysis · hydroacylation · ketones · 736 Chemical Reviews, 2010, Vol. 110, No. 2 Willis lactones · rhodium Scheme 39 Scheme 40

The development of a new transition-metal-catalyzed trans- formation can often allow a traditional reaction, perhaps one that employs stoichiometric amounts of reagents and harsh reaction conditions, to be replaced by a selective, atom economic process which proceeds under mild reaction con- ditions. This is certainly the case with the recent report from Dong and co-workers of a catalytic enantioselective synthesis of phthalide compounds.[1] The disproportionation of alde- hydes has traditionally been achieved under basic conditions and is known as the Cannizzaro reaction.[2] A Lewis acid promoted variant—the Tishchenko reaction—is a related Highlights transformation that converts two carbonyl units into the [3] corresponding ester. BothScheme processes 41 suffer from the for- Scheme 1. Synthesis of seven-membered lactone derivatives based on loadings to those in the basic processmation (Table of side 7) products were and have not been readily adapted to enantioselective ketone hydroacylation. employed, with 2-amino-4-picolinecatalytic taking the asymmetric place of reactions.[4] DOI:As an 10.1002/anie.201000159 alternative to these Phthalide Synthesis 2-amino-3-picoline. A related processprocesses employing the Dong allylic research group has developed catalytic 89 pyridyl amines as substrates has also beenenantioselective reported.Other intramolecular substrates ketone hydroacylation reac- lactones (Scheme 1).[5] To achieve efficient reactions the Afurthersimplificationoftheprocesswastoallowsimpletions.[1,5] substrates employed in these transformations were required Catalytic Intramolecularalcohols to be employed as the Ketone aldehyde precursorsTransition-metal-catalyzed Hydroacylation: (Scheme hydroacylation reactions usu- to feature a chelating ether group. Developing a ketone 90 39). The rhodium complex first catalyzesally proceed the oxidation through of initial formation of an acyl metal hydroacylation route to phthalide compounds presented a the alcohol to the aldehyde, using an equivalent of alkene Enantioselective Synthesis ofhydride Phthalides species, and subsequent addition of this species across greater challenge, as the substrates needed to target the as the oxidant, before promoting the hydroacylation reaction Michael C. Willis*as described previously. The most efficienta multiple catalyst was bond. found Hydroacylation reactions of alkenes and phthalide architecture would no longer allow the presence of to be RhCl · H O (3.3 mol %), usedalkynes in combination are now with well-studied processes, with a number of a coordinating ether group. 3 2 [6] triphenylphosphine and a stoichiometricenantioselective amount of 2-amino- variants already reported. However, the The reaction shown in Scheme 2 illustrates the basic asymmetric catalysis · hydroacylation ·4-picoline. ketones · Slight modification of the systemcorresponding allowed primary processes requiring addition of the acyl metal process used to access enantiomerically enriched phthalide lactones · rhodium amines to be employed as aldehyde equivalents.species across91 Methanol a ketone or an aldehyde—carbonyl hydro- compounds. Ketobenzaldehydes such as 1 were the key could also be employed as a substrate,acylation—are generating formal- much less studied.[7] The key intermediates for substrates for the process, and delivered five-membered 92 dehyde in situ, and then undergoing doubleboth hydroacylation. types of reaction, thecyclization acyl metal to hydride generate species, the natural are lactones product.2 Despite, phthalides, the through the proposed ketone hydro- The Jun group has exploited the excellentformed metal-directing by the oxidative additionrelatively of the low metal yield catalyst obtained to for the the hydroacylation step of he development of a neweffect transition-metal-catalyzed of picolyl imines to develop trans- a rangealdehyde of cascade C H process bond; a processthe synthesis, that can be no considered epimerization as a was reported, and the T À formation can often allow ainvolving traditional C reaction,-H and C perhaps-C bond one activation;type hydroacylation/ of C H activation.methodology In common allowed with a the number use of of commercial aldehydes. The 93 À94 that employs stoichiometricortho amounts-alkylation, of reagentsring-opening and harsh of cycloalkenones,reactions thatand are the based onJun C methodologyH functionalization, has also been hydro- applied to the synthesis of 95 18 À 101 reaction conditions, to beconversion replaced by of dienes a selective, to cycloalkanones atom acylationare some reactions notable offer advantagesF-labeled in terms ketones of step and atom to the functionalization of examples. 102,103 economic process which proceeds under mild reaction con- economy. polymers. Efforts to aid catalyst recovery have also been reported. Dong and co-workers originallySeveral account developed and a review catalytic articles documenting the ditions. This is certainly theFor case example, with the polystyrene-supported recent report from diphenylphosphine was development and application of picolyl imine intermediates enantioselective intramolecular ketone hydroacylation reac- 6,104-107 Dong and co-workers of a catalyticused in enantioselective combination with synthesis RhCl3 · H2O as a tandem alcohol in hydroacylation and related reactions are available. of phthalide compounds.[1] oxidation/hydroacylationThe disproportionation ofcatalyst. alde-96 Thetion catalyst as a route could to enantiomerically be The Miura enriched group seven-membered has undertaken detailed studies on the hydes has traditionally beenrecovered achieved underand used basic four conditions times before deactivation. More use of salicylaldehydeFew examples, as a chelating C-O bond substrate formation for alkyne and is known as the Cannizzarorecently, reaction. the Jun[2] groupA Lewis has described acid the use of a H-bonding hydroacylation (see Scheme 71).108,109 However, reactions [*] Dr. M. C. Willis promoted variant—the Tishchenkosolvent system reaction—is to allow catalyst a related recycling in primary alcohol of the same aldehyde with the majority of alkenes were hydroacylation systems. A 4,4!-bipyridyl/phenolDepartment solventof Chemistry, Universitypoor.109 ofThe Oxford exception was reaction with vinyltriethylsilylane, transformation that convertscombination two carbonyl generated units a into homogeneous the systemChemistry at Research elevated Laboratory Chem reviews 2010, p.725 [3] which, when combined with salicylaldehyde and a catalystAngew. Chem. Int. Ed. 2010,49,6026-6027 corresponding ester. Bothtemperatures; processes suffer however, from upon the cooling, for- twoSchemeMansfield immiscible 1. Synthesis Road, phases Oxford, of seven-membered OX1 3TA (UK) lactone derivatives based on Fax: (+ 44)1865-285-002 generated from [RhCl(COD)]2 and the ligand 1,1!-bis(diphe- mation of side products andwere have generated,not been readily allowing adapted easy catalyst to enantioselective and product separa- ketone hydroacylation.nylphosphino)ferrocene (dppf), provided the hydroacylation [4] E-mail: [email protected] Scheme 2. Enantioselective formation of phthalides; cod =cycloocta- catalytic asymmetric reactions.tion. ThisAs an approach alternative allowed to theseup to eight cyclesHomepage: of catalysis http://mcwillis.chem.ox.ac.uk/MCW/Home.htmladduct in good yield (Scheme 41). Simplel,5-diene. allenes were also processes the Dong researchwith group minimal has developed loss of activity. catalytic97 A similar concept has been shown to be competent substrates, as illustrated by the 98,99 enantioselective intramolecularused ketone to develop hydroacylation a H-bonding reac- catalystlactones system; (Schemethe high 1).[5] Toformation achieve of efficient!,γ-enone reactions74.Theuseofrefluxingtoluenewas the temperatures needed for reaction6026 disrupt a H-bonding  2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2010, 49, 6026 – 6027 tions.[1,5] substrates employed in theseneeded transformations in order to achieve were full required conversion for both types of network between 2,6-diamino pyridine and a barbiturate- reaction. The authors proposed the intermediacy of a chelate Transition-metal-catalyzedbound hydroacylation rhodium, allowing reactions efficient usu- hydroacylationto feature reactions. a chelating ethersuch group.as 75 to Developing account for a the ketone minimal decarbonylation ally proceed through initialCooling formation the reactions of an acyl to room metal temperaturehydroacylation allows route the toobserved. phthalide compounds presented a hydride species, and subsequentH-bonds addition to reform of this speciesand generates across twogreater immiscible challenge, phases. as the substratesMore recently needed the Suemune to target group the has investigated the use a multiple bond. HydroacylationEfficient reactions catalyst recycling of alkenes was and againphthalide possible. architecture wouldof salicylaldehydes no longer allow in the combination presence with of diene substrates.110,111 alkynes are now well-studiedThe processes, Kim group with has a numberutilized the of Juna methodology coordinating in ether the group.As can be seen from the examples presented in Scheme 42, 100 enantioselective variants alreadysyntheses reported. of indolizidines[6] However, (-)-167B the and (The-)-209D. reactionThe shownit in was Scheme possible 2 illustrates to use very the mild basic conditions to achieve Jun conditions were translated directly and applied to the excellent yields for the addition of salicylaldehydes to a corresponding processes requiring addition of the acyl metal process used to access enantiomerically enriched phthalide union of enantiomerically pure alkene 73 and heptanal to variety of 1,4- and 1,5-dienes. The use of Wilkinson’s species across a ketone ordeliver an aldehyde—carbonyl the corresponding ketone hydro- in 35%compounds. yield (Scheme Ketobenzaldehydes 40). complex (20such mol as %)1 allowedwere the the key reactions to be performed [7] acylation—are much less studied.ReductiveThe conditions key intermediates (H2, Pd/C) for allowedsubstrates CBZ cleavage for the and process,at ambient and delivered temperature. five-membered Employing these conditions allowed both types of reaction, the acyl metal hydride species, are lactones 2, phthalides, through the proposed ketone hydro- formed by the oxidative addition of the metal catalyst to the aldehyde C H bond; a process that can be considered as a À type of C H activation. In common with a number of À reactions that are based on C H functionalization, hydro- À acylation reactions offer advantages in terms of step and atom economy. Dong and co-workers originally developed a catalytic enantioselective intramolecular ketone hydroacylation reac- tion as a route to enantiomerically enriched seven-membered

[*] Dr. M. C. Willis Department of Chemistry, University of Oxford Chemistry Research Laboratory Mansfield Road, Oxford, OX1 3TA (UK) Fax: (+ 44)1865-285-002 E-mail: [email protected] Scheme 2. Enantioselective formation of phthalides; cod =cycloocta- Homepage: http://mcwillis.chem.ox.ac.uk/MCW/Home.html l,5-diene.

6026  2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2010, 49, 6026 – 6027 My take:

¡ Intramolecular- start with cationic or other coordinatively unsaturated Rh(I)

¡ Intermolecular ¡ Chelation- toss-up both are used in the literature, see if anyone has used similar substrates, if not lean towards starting with cationic Rh(I) ¡ Aldimines – start with Wilkinson’s catalyst or other neutral Rh (I) Extending NHC-Catalysis to Uncommon Electrophiles Biju et al. applications, NHCs are mainly used for the umpolung of SCHEME 1. Proposed Mechanism of Benzoin Condensation aldehydes. In these reactions, addition of the NHC to the aldehyde finally results in the generation of an acyl anion equivalent, the Breslow intermediate. The benzoin conden- sation and Stetter reaction are the two most well-known transformations, which employ the Breslow intermediate as key intermediate (eq 1). Generally, imidazolium, thiazolium, or triazolium salt-derived NHCs have been used successfully for umpolung reactions, and during recent years there has been an increased interest in NHC-catalyzed transformations and many new reactions have been developed. The purpose of the present Account is to provide an update about the recent developmentsPart 3 ; inNHC NHC organocatalysis. Catalyzed Mainly, the RXNS Account is focused on NHC-catalyzed reactions of unconventional elec- trophiles developed in our laboratory, but adequate descrip- tion of related work carried out by others is also given.

Part 1 reacts as a nucleophile with another molecule of aldehyde to furnish the final product, the R-hydroxy ketone 5,andthe Part 2 original NHC catalyst 1 is regenerated (Scheme 1). The phenomenal success of NHCs in organocatalysis can be attributed primarily to their electronic properties leading Unconventional Part 3 to different modes of action in catalysis (Figure 1). The From a historical perspective, a reportreactions by Ukai et al. in 1943 demonstrating that thiazolium salts could be used as pronounced nucleophilicity of NHCs allows the addition to catalysts in the benzoin reaction constitutes an early exam- electrophiles such as aldehydes leading to the formation of ple for the involvement of azolium salts in organocatalysis.3 the tetrahedral intermediate (A). The azolium moiety is Breslow proposed a mechanistic explanation for the thiazo- strongly electron-withdrawing, acidifying the R-position (E). lium salt-catalyzed benzoin condensation in 1958.4 In this Alternatively, the same tetrahedral intermediate can under- mechanism, the catalytically active species was represented go hydride transfer in the presence of easily reducible as a thiazolium zwitterion, the resonance structure of an NHC, substrates such as activated carbonyl compounds resulting and the reaction was postulated to proceed via the enaminol in an acyl azolium species and a reduced reaction partner intermediate 3,the“Breslow intermediate” (Scheme 1). How- (H). The enaminol moiety (Breslow intermediate) prepared in ever, the existence of carbenes as catalytically active species mode (E) is highly nucleophilic and acts as an acylating agent in these processes was only realized almost three decades in a variety of NHC-catalyzed transformations (B, then F). later when the synthesis of stable phosphinocarbene was Furthermore, the addition of an NHC to R,β-unsaturated reported by Bertrand and co-workers in 19885 and the isola- aldehydes can lead to a dienaminol intermediate, rendering 7 tion and characterization of stable NHC was unequivocally the β-carbon atom nucleophilic (C). Additionally, the enam- established by Arduengo and co-workers in 1991.6 Although inol can trigger an elimination, if there is a leaving group at NHCs were used long before these findings, these seminal the R-position of the aldehyde (D). In this process, a nucleo- discoveries marked a true breakthrough and initiated exten- philic enol(ate) can form, in which the attached azolium sive research in the application of NHCs in catalysis. species acts as a bystander (I).8 Moreover, the catalyst can Following the proposal of Breslow,4 a key step in benzoin increase the electrophilicity in acyl azolium species and act condensation is the nucleophilic attack of the in situ generated as a good leaving group, re-entering the catalytic cycle (G). carbene 1 to the aldehyde, leading to the tetrahedral inter- A variety of NHC-catalyzed redox processes lead to the mediate 2, which undergoes proton transfer to the nucleo- formation of acyl azolium species using an oxidant.9 Finally, philic enaminol intermediate 3. This acyl anion equivalent 3 the Breslow intermediate can also act in a dual push pull À

Vol. 44, No. 11 ’ 2011 ’ 1182–1195 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1183 Benzoin CondensationExtending NHC-Catalysis to Uncommon Electrophiles Biju et al. applications, NHCs are mainly used for the umpolung of The Old SCHEME 1. Proposed MechanismThe New of Benzoin Condensation aldehydes. In these reactions, addition of the NHCCyanohydrin to the aldehyde finally results in the generation of an acyl anion equivalent, the Breslow intermediate. The benzoin conden- sation and Stetter reaction are the two most well-known transformations, which employ the Breslow intermediate as key intermediate (eq 1). Generally, imidazolium, thiazolium, or triazolium salt-derived NHCs have been used successfully for umpolung reactions, and during recent years there has been an increased interest in NHC-catalyzed transformations and many new reactions have been developed. The purpose of the present Account is to provide an update about the recent developments in NHC organocatalysis. Mainly, the Account is focused on NHC-catalyzed reactions of unconventional elec- trophiles developed in our laboratory, but adequate descrip- tion of relatedJustus work carried Von out Liebig by others is also given. Breslow Nikolay Zinin Int.

reacts as a nucleophile with another molecule of aldehyde to furnish the final product, the R-hydroxy ketone 5,andthe original NHC catalyst 1 is regenerated (Scheme 1). The phenomenal success of NHCs in organocatalysis can be attributed primarily to their electronic properties leading From a historical perspective, a report by Ukai et al. in to different modes of action in catalysis (Figure 1). The 1943 demonstrating that thiazolium salts could be used as pronounced nucleophilicity of NHCs allows the addition to catalysts in the benzoin reaction constitutes an early exam- electrophiles such as aldehydes leading to the formation of ple for the involvement of azolium salts in organocatalysis.3 the tetrahedral intermediate (A). The azolium moiety is Breslow proposed a mechanistic explanation for the thiazo- strongly electron-withdrawing, acidifying the R-position (E). lium salt-catalyzed benzoin condensation in 1958.4 In this Alternatively, the same tetrahedral intermediate can under- mechanism, the catalytically active species was represented go hydride transfer in the presence of easily reducible as a thiazolium zwitterion, the resonance structure of an NHC, substrates such as activated carbonyl compounds resulting and the reaction was postulated to proceed via the enaminol in an acyl azolium species and a reduced reaction partner intermediate 3,the“Breslow intermediate” (Scheme 1). How- (H). The enaminol moiety (Breslow intermediate) prepared in ever, the existence of carbenes as catalytically active species mode (E) is highly nucleophilic and acts as an acylating agent in these processes was only realized almost three decades in a variety of NHC-catalyzed transformations (B, then F). later when the synthesis of stable phosphinocarbene was Furthermore, the addition of an NHC to R,β-unsaturated reported by Bertrand and co-workers in 19885 and the isola- aldehydes can lead to a dienaminol intermediate, rendering 7 tion and characterization of stable NHC was unequivocally the β-carbon atom nucleophilic (C). Additionally, the enam- established by Arduengo and co-workers in 1991.6 Although inol can trigger an elimination, if there is a leaving group at NHCs were used long before these findings, these seminal the R-position of the aldehyde (D). In this process, a nucleo- discoveries marked a true breakthrough and initiated exten- philic enol(ate) can form, in which the attached azolium sive research in the application of NHCs in catalysis. species acts as a bystander (I).8 Moreover, the catalyst can Following the proposal of Breslow,4 a key step in benzoin increase the electrophilicity in acyl azolium species and act condensation is the nucleophilic attack of the in situ generated as a good leaving group, re-entering the catalytic cycle (G). carbene 1 to the aldehyde, leading to the tetrahedral inter- A variety of NHC-catalyzed redox processes lead to the mediate 2, which undergoes proton transfer to the nucleo- formation of acyl azolium species using an oxidant.9 Finally, philic enaminol intermediate 3. This acyl anion equivalent 3 the Breslow intermediate can also act in a dual push pull À

Vol. 44, No. 11 ’ 2011 ’ 1182–1195 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1183 Extending NHC-Catalysis to Uncommon Electrophiles Biju et al.

Extending NHC-Catalysis to Uncommon Electrophiles Biju et al.

this approach by utilizing acylsilanes as coupling partners for in benzoin and acyloin reactions allowing the selective aldehydes in a highly regiospecific cyanide-catalyzed cross formation of R-hydroxyketones although in low yield and silyl benzoin reaction.12 The noteworthy features of this with a narrow substrate scope. The common methods to transformation include the in situ protection of the second- synthesize this important class of compounds are often non- ary hydroxyl group as a silyl ether and the inhibition of the C C bond forming hydrolyses or oxidations. Therefore, their À homobenzoin formation. Subsequently, the Johnson group synthesis via an NHC-catalyzed formation of a new C C À reported the enantioselective cross silyl benzoin reaction bond would represent an attractive alternative. Along our using (R,R)-TADDOL-derived metallophosphite (eq 3).13 research in the field of dual NHC-metal-catalysis, we have FIGURE 1. Most Prominent Modes of Action in NHC-Organocatalysis. Recently, Scheidt and co-workers reported the use of O-silyl recently developed a new thiazolium salt 8 with a seven- thiazolium carbinol as acyl anion precursor by its reaction with membered ring in the backbone and a sterically demanding fashion (see chapter 4): the positively polarized proton of the cross-benzoinaldehyde in the reaction presence have of CsF received (eq 4).14 As a basic lot of conditions attention re-mesityl substituent on nitrogen.18 The synthesis of 8 was enaminol can interact with the reaction partner and with- cently,can be and avoided a recent by liberating review onthe NHC-organocatalysis acyl anion equivalent with covers a theaccomplished in two steps following a modified procedure draw electron-density from it, activating it for the attack of literaturefluoride source, on this this subject. method2f,10 is superiorThus, infor the synthesizing following, even we willof Bach and co-workers (eq 6).19 the enamine part of the Breslow intermediate (J). The focusdifficult on cross-acyloin recent developments products (coupling inthefieldofintermolecular of one or two aliphatic electronic and steric properties of NHCs can be tuned over cross-benzoinaldehydes). Another reactions. approach In thi ofsprocess,amixturewithupto masking one component to avoid its reaction with the NHC has already been demon- a wide range by choosing different nitrogen heterocycles as eight different products (four pairs of enantiomers) can be strated in aza-benzoin reactions.15 Additionally, Enders and well as by the proper choice of substituents on nitrogen and anticipated (eq 2), which renders the intermolecular cross- Henseler reported the use of trifluoromethyl ketone as an the backbone. Slight structural differences can have a dra- benzoinelectrophile reaction in a cross-benzoin especially challenging reaction using in terms the carbene of chemos- matic effect on catalytic activity and selectivity of carbenes. electivity.precursor 7 Indeed,.16a The excellent a general selectivity solution for for the aR-hydroxy- chemoselectiveR- 11 NHC-catalyzedtrifluoromethylCross-Benzoin ketones cross-benzo is a resultin reaction of the is reversibility still elusive. of the 2. Recent Advances in Benzoin Reaction homobenzoin reaction. Therefore, products of a broad range Extending NHC-Catalysis to Uncommon Electrophiles Biju et al. The benzoin reaction, the coupling of two aldehydes, is one of aromatic and heterocyclic aldehydes could be converted During our investigation, we found that thiazolium salt 8 of the first transformations found to be catalyzed by NHCs.3 into the corresponding products in excellent yield (eq 5). By gave the best results for the hydroxymethylation of a broad introducing a new chiral triazoliumcatalyst,thiscross-benzoin 20 Since the benzoin reaction is a valuable strategy to form new TABLE 1. NHC-Catalyzed Hydroxymethylationrange of of aromatic Aldehydes and aliphatic aldehydes with formalde- reaction using heteroaromatic aldehydes succeeded in an hyde (Table 1).20 Our mechanistic experiments revealed C C bonds leading to the formation of R-functionalized À asymmetric fashion in good to excellent yields and moderate that, besides other effects, the nucleophilic attack of the carbonyl compounds, this unique process and its mechan- to good enantioselectivities, which could be improved by Breslow intermediate onto formaldehyde is fast, leading to 10a 16b ism have been intensively studied. Moreover, the effort to crystallization. the selective formation of the observed product 10. 5614 Chemical Reviews, 2007, Vol. 107, No. 12 Enders et al. develop a highly asymmetric homobenzoin reaction led to One strategy for selective cross-benzoin reactions is the Inspired by an early work of Stetter and Dambkes,€ 21 Zeitler, the development of numerous chiral NHC precursors. How- modificationTable 3. Enantioselective of one aldehyde Synthesis of componentr-Amidoketones to by control Miller theConnonsuccessfully and co-workers used by Suzuki22 recently et al. for studied a diastereoselective the cross-acyloin and Co-workers ever, the asymmetric benzoin reaction and the intramolecular chemoselectivity. Johnson and co-workers demonstratedreactionintramolecular between crossed aromatic aldehyde and aliphatic-ketone benzoinaldehydes reaction (eq 7). in The yield ee the course of an elegant synthesis of preanthraquinones.72 R1 R2 R3 time (%) (%) formationBy employing of only the one highly cross- functionalizedacyloin product isoxazole was achieved55 as p-Cl-Ph Ph Ph 2 h 100 78 usingsubstrate, bulky the triazolium-derived tetracyclic R-hydroxyketone catalyst 11 (catalyst56 was obtained control) in ’ ’ – ’ ’ 1184 ACCOUNTS OF CHEMICAL RESEARCH 1182 1195 2011 Vol. 44, No. 11 p-Cl-Ph p-MeO-Ph Ph 2 h 90 87 combinationby base-promoted with an cyclocondensation aromatic aldehyde in good possessing yield. The a directing high o-NO2-Ph p-MeO-Ph Ph 15 min 77 82 diastereoselectivity was induced by the pre-existing stereo- o-NO2-Ph p-MeO-Ph i-Pr 15 min 63 79 substituentcenters in the in the substratesortho-position (Scheme (substrate 19). control). For the p-Cl-Ph Ph i-Pr 2 h 97 75 p-Cl-Ph 2,4-dimethoxyphenyl Ph 1 h 80 81 latter, bromine proved to be beneficial as it features the proper Ph Ph Ph 2 h 15 83 stericScheme demand 19. Intramolecular as well as the Crossed right Benzoinelectronic Reaction characteristics. by Suzuki and Co-workers Interestingly, the same chemoselectivity using triazolium cat- employed chiral peptidic thiazolium salts such as the 67 alyst has very recently beenin observed moderateto by Yang excellent and yields co- (49 98%) and moderate to compound (S,S,S,S)-51 (see Scheme3. Recent 16). They Advances observed in the Stetter11 Reaction À that the structure and stoichiometryEver of since the base the significantly seminal work of Stetter in 1973, the NHC- good enantioselectivities (56 78% ee).27a Impressively, workers for the acyloin reaction of simple aromatic aldehydes À influenced the activity of the catalyst. The hindered base 23 catalyzed addition of aldehydeswith to acetaldehyde. Michael acceptors, the the enantiomeric excess of the products could be en- pentamethyl piperidine (PEMP) gave the highest ee values Stetter reaction, is widely used as a catalytic pathway for the (Scheme 16, Table 3). hanced up to 99% ee by a single recrystallization (eq 8). You and co-workers have recentlysynthesis shown of 1,4-bifunctional that aromatic compounds such as 1,4-dike- The presence of the N-benzyl substituent in 12 was crucial aldehydes 4 can also be coupledtones, to unactivated 4-ketonitriles, imines and52 4-ketoesters.24 This is valuable, for activity and high levels of selectivity in this intermole- 2 3 (R ,R ) aryl) to form R-aminoketones 45 under thermo-“ ” 68 since it leads to an unnatural functional groupSuzuki distance, and Coworkerscular Stetter reaction, demonstrating the importance of the dynamic control. In contrast, in the investigations17 of Murry Chem.In Rev. an earlier2007, 107, report, 5606-5655 Inoue andwhich co-workers is difficultdemon- to realize usingIn order traditional to develop methods. a general method, our group started 27c Accountset al.of Chem. and Miller Res. and1182-1195, co-workers, 2011 the product was formed investigations on simple aldehydeN-substituent ketones 57 as of substrates triazolium-based catalysts. Indepen- stratedunder the kinetic use of control, paraformaldehyde as the imineStetter as must a and reaction react co-workers faster partner with succeededfor the to carbene-catalyzed selectively cross- crosseddently, intramolecular Rovis and benzoin co-workers reported the asymmetric the Breslow intermediate than with another aldehyde mol- couple a variety of aromatic andcondensation. aliphatic aldehydes Independently, with bothintermolecular our group and Suzuki Stetter and reaction of a glyoxamide and ecule. By utilizing the thiazolium chloride 10a,awiderange co-workers reported that various five- and six-membered28 avarietyofMichaelacceptorsinanintermolecularfash-Vol. 44, No. 11 ’ 2011 ’ 1182–1195 ’ ACCOUNTSalkylidenemalonates. OF CHEMICAL RESEARCH ’ 1185However, excellent enantioselec- of substrates was successfully used in this reaction with cyclic acyloins 58 can indeed be obtained as their racemic ion. The catalytic intramolecular version of this reaction varying yields (Scheme 17). Studies to understand the mixtures by employing commerciallytivities available were obtained thiazolium only for a morpholine-derived reaction mechanism revealed at least a partial thermodynamic 74 was developed by Ciganek in 1995salts asleading precatalysts to the (Scheme forma- 20).glyoxamide.The tetralone A variety58a was of β-substituted alkylidenemalo- control, as the desired product wastion also of formedbenzo-annulated with benzoin furanonesobtained and pyranones.in a significantly25 The higher yield than tetralone 58b as substrate. nates underwent this reaction in good yield with high Stetter reaction can be utilized forwith the interchanged controlled formation functionalities. This may be due to the activating effect of the aromaticasymmetric aldehyde in the induction precursor in the presence of a phenylala- Scheme 17. Intermolecular Aldehydeof new-Imine stereocenters Cross-coupling in the product. The intramolecular by You and Co-workers of 58a. Substrates lacking the obviouslynine-derived favorable NHC catalyst benz- 14 (eq 9). A limitation of this Stetter reaction and its variantsannulation have receivedin the resulting a lot acyloin (58d) showed only method was the use of alkylidene malonate with two ester of attention recently, and a numbermoderate ofyields reviews because have the competing intermolecular reac- groups, which diminishes the opportunity for further deri- addressed various aspects of thistion reaction.was also observed.2c,f,26 In The view quinone 58e was isolated starting from the corresponding dialdehyde.vatization. The R-hydroxyketone In view of this, the Rovis group recently devel- of these excellent reviews, a detailedis only formed discussion as an intermediate of the and undergoes air oxidation oped a highly enantioselective and diastereoselective intramolecular Stetter reactionto is the not quinone attempted58e during in this workup. intermolecular Stetter reaction of a glyoxamide and alky- Account. The next goal, of course, has been to develop an enantio- lidene ketoamides leading to 1,4-dicarbonyl compounds, Although NHC catalysts and reactionselective variant. protocols Unfortunately, are well the chiral bicyclic triazolium salt (S)-33, that had been found to be an excellent catalyst established for the enantioselective intramolecular ver- which are amenable for further derivatization to syntheti- 3.6. Intramolecular Crossed Benzoin for the enantioselective intermolecular benzoin condensation, 29 Condensations sion, the asymmetric intermolecularprovedStetter to be reaction ineffective still in thecally intramolecular useful intermediates reaction. in organic synthesis (eq 10). Searching for alternative catalysts,Subsequently, our group synthesized they used the nitroalkenes as viable Michael In contrast to the previouslyremains described a formidable intermolecular challenge. This is partly because novel triazolium salts 59 and acceptors60 starting in intermolecularfrom easily Stetter reaction using a new condensation reaction, intramolecularthe Michael crossed acceptors benzoin reac- containingaccessible a β-substituent enantiopure (except polycyclic γ-lactams. The precatalyst tions have been developed for achalcones) much shorter usually time. In show 1976, diminished59 led reactivity to excellent in the results Stet- in thechiral intramolecular NHC precatalyst crossed15.Byutilizingthestereoelectronic Cookson and Lane reported a cyclization of glutaric alde- benzoin condensation of aldehydeas ketones well as57 steric, as shown effects in induced by the fluoro and iso- hydes 53 with thiazolium saltster as reaction. precatalysts Recently, (55) to Enders the and co-workers reported the asymmetric intermolecular69 StetterScheme reaction 21. The of acyloins aromatic58 bearingpropyl a quaternary substituents stereocenter in 15,ahighlyenantioselectiveStetter corresponding hydroxylcyclopentanones 54! and 54!!. Un- were synthesized with very good yields and excellent - fortunately, asymmetric dialdehydesaldehydes (53b c and) only chalcones yielded catalyzedenantiomeric by the excess NHC (93- deri-98% ee).reaction75 The precatalyst of nitroalkenes60,easy and heteroaryl aldehydes was R - - 30 the isomeric -ketoles 54!b cvedand from54!!b triazoliumc (Scheme salt 18).12 leadingavailable to from 1,4-diketones the very cheap13 chiralitydeveloped source L-pyroglutamic (eq 11). However, the use of heteroaromatic Synthetic thiazolium salts (10c), developed by Stetter and 70 acid, proved to be even more active and the yields were his co-workers, similar to thiamine itself, have been consistently excellent, albeit accompanied by lower enan- 1186 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1182–1195 ’ 2011 ’ Vol. 44, No. 11 Scheme 18. Intramolecular Acyloin Condensation by tiomeric excesses (63-84% ee). The substrates of the Cookson and Lane enantioselective intramolecular crossed benzoin conden- sation were varied to widen the scope of the reaction. Promising results were achieved with substrates where the aldehyde and the ketone function are interchanged and with substrates where a five-membered ring is formed. The stereochemical outcome was explained by the transition state shown in Scheme 21. The Si-face of the Breslow intermedi- ate, which is formed as its (E)-isomer, is sterically shielded by the tetrahydronaphthalene residue of the tetracyclic Extending NHC-Catalysis to Uncommon Electrophiles Biju et al. Stetter Reaction

TABLE 2. NHC-Catalyzed Intermolecular Stetter Reaction32

aldehyde was crucial for high levels of reactivity and efficiently merged. A variety of aldehydes reacted with the selectivity. dehydroamino ester 16 in the presenceMichael of NHCacceptors generated with beta substituents usually are bad2c,31 reaction partners from L-phenyl alaninol derived triazolium salt 17 yield- ing R-amino acid derivatives 18 in excellent yield and Ciganek, Synthesis, 1995, 1311 stereoinduction (Table 2).32 Accounts of Chemical Research, 2011, 1182 The mechanism and mode of asymmetric induction are still unclear, but can be rationalized as follows. First, the reaction between the free carbene derived from 17 and the aldehyde leads to the formation of a nucleophilic Breslow intermediate (Scheme 2). The Michael acceptor 16 ap- proaches from the bottom face in an anti fashion, most likely supported by a hydrogen bond between the enol

hydrogen and the carbonyl oxygen of the Michael acceptor Downloaded by: University of Texas at Austin. Copyrighted material. (19, Scheme 2). In this process, ester enolate 21 bearing a new but transient stereocenter is formed highly stereoselec- tively. The stereochemistry is relayed to the R-position by a stereoselective protonation of the transiently formed eno- late. Finally, the NHC is released, destroying the initially formed stereocenter and forming the final product 18. Alternatively, a concerted transition state 20 could directly generate intermediate 22. This reaction is remarkable be- cause of the rather low electronic activation of substrate 16 and because it represents the first asymmetric intermolecu- lar Stetter reaction that generates only an R-stereocenter.

4. NHC-Catalyzed Reaction of Aldehydes with Unconventional Electrophiles We envisioned the synthesis of enantioenriched R-amino Although the NHC-catalyzed umpolung of aldehydes and acid derivatives by an intermolecular enantioselective Stet- the subsequent interception of the nucleophilic acyl anion ter reaction using N-acylamido acrylate 16 as the Michael intermediates with various electrophiles, such as aldehydes, acceptor. In this process, the two important steps, the C C ketones, imines, and activated, CdC double bonds, is well- À bond formation between the Breslow intermediate and the known, the analogous transformations with unconven- Michael acceptor as well as an asymmetric protonation are tional electrophiles such as unactivated C C multiple bonds À

Vol. 44, No. 11 ’ 2011 ’ 1182–1195 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1187 Extending NHC-Catalysis to Uncommon Electrophiles Biju et al.

and -withdrawing substituents on the aromatic ring as well as SCHEME 2. Proposed Mechanistic Pathways for the Enantioselective Stetter Reaction32 the substituent on the allyl moiety (Table 3). A likely mechanism starts with the addition of NHC to the carbonyl group of 2-allyloxy benzaldehyde to form Breslow intermediate 28 (Scheme 3). In a concerted transition state 29, this could add as a nucleophile to the olefin of the allyl moiety and the resulting zwitterionic tetrahedral intermedi- ate 30 liberates the desired product and the catalyst. The similarity to Conia-ene type transition states (same polarity of reacting olefins) and reverse-Cope elimination type ones as suggested by Rovis (five-membered transition state) is striking.36,37 It should be noted in this context that the possibi- lity of addition of Breslow intermediates4 to Michael acceptors in a concerted manner analogous to the reverse Cope elimina- (eq 12) and activated alkyl halides was largely unexplored. tion was previously mentioned by Rovis and Read de Alaniz.37 In contrast, the transition metal catalyzed hydroacylation of Furthermore, density functional theory calculations were per- unactivated C C multiple bonds, the insertion into the 38 À 33 formed by Grimme et al. to get more insight into the crucial Cformyl-H bond of aldehydes is established. Extending NHC-Catalysis to Uncommon Electrophilesstep ofBiju the et al. reaction, that is, whether the proton migration or the C Cbondformationoccursfirst.Thesestudiessupporta À 38 TABLE 4. NHC-Catalyzed Enantioselectiveconcerted Hydroacylation of butUnactivated highly Olefins asynchronous38 transition state. Subsequently, we applied this methodology to the highly asymmetric NHC-catalyzed intramolecular hydroacylation 4.1. Hydroacylation of Electron-Neutral Double Bonds. of unactivated olefins leading to the formation of an all- The intramolecular nucleophilic addition reaction of acyl carbon quarternary stereocenter in the biologically and anion equivalents to enol ethers catalyzed by the NHC pharmaceutically important chromanone structure.38 The generated from readily available thiazolium salt 24 was carbene generated from 17 showed exceptionally high developed by She and co-workers,34 leading to the forma- reactivity and selectivity affording differently substituted tion of benzofuranones 25 in excellent yield (eq 13). The H-Acylation of Unactivatedchromanones, most of them with 99% ee (Table 4). This exact mechanism of this transformation (concerted or step- extraordinarily high level of enantioinduction also speaks in wise involving an oxonium species) is not clear. Compounds favor of a concerted mechanism. Another indication comes First intermolecular NHC-Catalyzed Hydroacylation of Cyclopropenes39 TABLE 5. from the first intermolecular hydroacylation of cyclopropenes that we have reported very recently (Table 5).39 Experiments with deuterated substrates demonstrated that the hydroacy- lation is a syn addition process. With this latter transformation, Extending NHC-Catalysis to Uncommon Electrophiles Biju et al. Extending NHC-Catalysis to Uncommon Electrophilesa generalBiju NHC-catalyzed et al. intermolecular hydroacylation of truly unactivated olefins seems to be within reach. TABLE 4. NHC-Catalyzed Enantioselective Hydroacylation4.2. of Unactivated Hydroacylation Olefins38 of Electron-Neutral Triple Bonds. TABLE 3. NHC-Catalyzed Hydroacylation of Unactivated Double Bonds35 Recent investigations in our laboratory revealed novel In view of these interesting results, we then focused our 40 reactivity profiles of the carbene generatedTABLE from 6. NHC-Catalyzed the thiazo- Hydroacylation ofattention Unactivated Internal on Alkynes the hydroacylation of unactivated triple bonds. lium salt 8. We commenced with the NHC-organocatalyzed Treatment of unactivated internal alkynes 31 with the

cyclization of 2-allyloxy benzaldehydes 26 to the corre- carbene generated from 8 by deprotonation with K2CO3 sponding chromanones 27, the intramolecular hydroacyla- resulted in the smooth formation of benzylidene chroma- tion of unactivated CdC-double bonds.35 Gratifyingly, none 32, bearing a synthetically valuable exocyclic olefin, as among the wide range of NHCs screened, the carbene a single isomer.40 Variations on both aromatic rings are well generated from 8 by deprotonation with DBU showed the tolerated, providing good yields for electron-donating and best reactivity, providing the desired functionalized chroma- electron-withdrawing substituents (Table 6). Interestingly, nones in moderate to excellent yield. This new methodology competition experiments carried out using electronically was applied to a range of substrates withafforded electron-donating the chromanones in gooddifferent to excellent alkynes yield 70% revealed with a trans/cis (1)ratio reversible of 3:1. In addition, formation a variety of the (Table 7). Gratifyingly, a substituent on the progargylic coupling aldehydes 36 including heterocyclic and aliphatic Accounts of Chemical Research, 2011, 1182 moiety was also well tolerated, giving the product in ones have been examined, and in all cases the reaction

1188 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1182–1195 ’ 2011 ’ Vol. 44, No. 11 1190 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1182–1195 ’ 2011 ’ Vol. 44, No. 11 39 SCHEME 3. Proposed Mechanism for the Hydroacylation of Unactivated DoubleTABLE Bonds 5. 35,38First intermolecular NHC-Catalyzed Hydroacylation of Cyclopropenes

TABLE 6. NHC-Catalyzed Hydroacylation of Unactivated Internal Alkynes40

Breslow intermediate and (2) a key role of the alkyne in the enone 35. It was envisioned that this will constitute a dually rate determining step, since electron-poor alkynes reacted NHC-catalyzed hydroacylation cascade comprising an initial faster than electron-rich ones. hydroacylation of an unactivated triple bond and a subse- We continued with the organocatalyzed hydroacylation quent intermolecular Stetter reaction (eq 15). of unactivated terminal alkynes. Surprisingly, treatment of

33a with 8 and K2CO3 did not result in the formation of the expected enone, but it led to the formation of chromanone 34a, presumably by the addition of a second molecule of 33a to the intermediate enone (eq 14).

affordedReaction the of chromanonesthe 2-propargyloxy in good aldehydes to excellent33 with yield a 70% with a trans/cis ratio of 3:1. In addition, a variety of second(Table coupling 7). Gratifyingly, aldehyde a substituent catalyzed on by the NHC progargylic generated coupling aldehydes 36 including heterocyclic and aliphatic frommoiety8 resulted was also in thewell formation tolerated, ofgiving chromanone the product with in a ones have been examined, and in all cases the reaction 1,4-diketone motif in excellent yield. It is noteworthy that

these1190 ’ reactionsACCOUNTS OF require CHEMICAL only RESEARCH rather’ 1182– low1195 ’ catalyst2011 ’ Vol. 44, loading No. 11 In view of this unprecedented reactivity, we decided (5 mol %) and proceed with high level of selectivity. A to employ a second aldehyde as the coupling partner for variety of substrates with different substitution patterns

Vol. 44, No. 11 ’ 2011 ’ 1182–1195 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1189 Extending NHC-Catalysis to Uncommon Electrophiles Biju et al.

TABLE 4. NHC-Catalyzed Enantioselective Hydroacylation of Unactivated Olefins38

Extending NHC-Catalysis to Uncommon Electrophiles Biju et al. TABLE 5. First intermolecular NHC-Catalyzed Hydroacylation of Cyclopropenes39

TABLE 7. NHC-Catalyzed Hydroacylation Stetter Cascade40 À

TABLE 6. NHC-Catalyzed HydroacylationAlkynes of Unactivated Internal Alkynes and40 Arynes

TABLE 8. NHC-Catalyzed Hydroacylation of Arynes41

Extending NHC-Catalysis to Uncommon Electrophiles Biju et al.

TABLE 7. NHC-Catalyzed Hydroacylation Stetter Cascade40 À afforded the chromanones in good to excellent yield 70% with a trans/cis ratio of 3:1. In addition, a variety of (Table 7). Gratifyingly, a substituent on the progargylic coupling aldehydes 36 including heterocyclic and aliphatic moiety was also well tolerated, giving the product in ones have been examined, and in all cases the reaction

1190 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1182–1195 ’ 2011 ’ Vol. 44, No. 11

resulted in the smooth formation of the expected chroma- triflate 38 using 2.0 equiv each of KF and 18-crown-6 in the none derivative. presence of carbene generated from 8 by deprotonation Accounts of Chemical Research, 2011, 1182 4.3. Hydroacylation of Arynes. Encouraged by these using KOt-Bu resulted in the formation of the aryl ketones results, we then turned our attention to a class of highly 39 in moderate to excellent yield (Table 8).41 This process TABLE 8. NHC-Catalyzed Hydroacylation of Arynes41 reactive intermediates in organic synthesis, namely, arynes. provides an attractive transition metal-free synthetic strat- Although arynes have been extensively utilized in transi- egy to a wide range of aryl ketones and is particularly tion-metal-catalyzed reactions, their application in organo- appealing by virtue of the high levels of chemoselectivity catalytic processes is scarce presumably due to the inherent observed. reactivity of arynes toward nucleophiles. Interestingly, Mechanistic studies indicate that the electronic nature of the insertion of arynes into the C Hbondofalde- the aldehydes plays a prominent role in the rate-determin- formylÀ hydes, the intermolecular hydroacylation of arynes, was ing step,42 with electron-poor aldehydes reacting faster than unknown. Again, the proper choice of the carbene was key the electron-rich ones. Also, competition experiments car- to success. The reaction of a wide variety of aldehydes ried out using electronically dissimilar arynes revealed with the aryne generated in situ from 2-trimethylsilylaryl no preference in product formation, indicating that the

Vol. 44, No. 11 ’ 2011 ’ 1182–1195 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1191

resulted in the smooth formation of the expected chroma- triflate 38 using 2.0 equiv each of KF and 18-crown-6 in the none derivative. presence of carbene generated from 8 by deprotonation 4.3. Hydroacylation of Arynes. Encouraged by these using KOt-Bu resulted in the formation of the aryl ketones results, we then turned our attention to a class of highly 39 in moderate to excellent yield (Table 8).41 This process reactive intermediates in organic synthesis, namely, arynes. provides an attractive transition metal-free synthetic strat- Although arynes have been extensively utilized in transi- egy to a wide range of aryl ketones and is particularly tion-metal-catalyzed reactions, their application in organo- appealing by virtue of the high levels of chemoselectivity catalytic processes is scarce presumably due to the inherent observed. reactivity of arynes toward nucleophiles. Interestingly, Mechanistic studies indicate that the electronic nature of the insertion of arynes into the C Hbondofalde- the aldehydes plays a prominent role in the rate-determin- formylÀ hydes, the intermolecular hydroacylation of arynes, was ing step,42 with electron-poor aldehydes reacting faster than unknown. Again, the proper choice of the carbene was key the electron-rich ones. Also, competition experiments car- to success. The reaction of a wide variety of aldehydes ried out using electronically dissimilar arynes revealed with the aryne generated in situ from 2-trimethylsilylaryl no preference in product formation, indicating that the

Vol. 44, No. 11 ’ 2011 ’ 1182–1195 ’ ACCOUNTS OF CHEMICAL RESEARCH ’ 1191 Metal Vs Organic Metals:

¡ Have to worry about decarbonylation ¡ Rh is expensive, other metals are not as well documented ¡ Enantioselective reactions are pretty straightforward

¡ Screening catalysts is pretty straightforward

NHC:

¡ Can react with a wider range of substrates (benzoin reaction, stetter, and other substrates)

¡ Made from very accessible materials.

¡ Catalyze many transformations, so you have to be aware of possible side reactivity, or use that to your advantage..

¡ Have to find an NHC that works for you Thank You

Questions for Me? Questions for You..

O [Rh] H or R2 R

X

N O S Base H Dioxane, 2hr, heat X H

O Wilkinson's Amino pyridine H + Toluene, heat, 24hr Assume Aq. Workup Rh(I) P(Cy)3 + XS N NH

Ph Questions for You.. Extending NHC-Catalysis to Uncommon Electrophiles Biju et al.

TABLE 3. NHC-Catalyzed Hydroacylation of Unactivated Double Bonds35

730 Chemical Reviews, 2010, Vol. 110, No. 2 Willis

Scheme 17 Scheme 18 Answers for You..

SCHEME 3. Proposed Mechanism for the Hydroacylation of Unactivated Double Bonds35,38 O [Rh] H or R2 R Scheme 19 Table 3. Cis-Selective Cyclopentanone Synthesisa

X

N O S Base H catalyst and substrate pairings were reversed when the neutral Dioxane, 2hr, heat complexes were employed. X H 2.3.2. Enantioselective Systems

O O Wilkinson's While investigating aldehyde decarbonylation using chiral BreslowAmino intermediate pyridine and (2) a key role of the alkyne in the enone 35. It was envisionedrhodium that this catalysts, will constitute James a dually and Young observed the first R H + rate determiningToluene, heat, step, 24hr since electron-poor alkynes reacted NHC-catalyzed hydroacylationexample cascade of comprisingenantioselective an initial intramolecular hydroacylation 37,38 faster than electron-rich ones. hydroacylation of an unactivatedvia kinetic triple resolution. bond and a subse-Reaction of racemic pentenal 36 with a catalyst incorporating enantiomerically pure (S,S)- We continued with the organocatalyzed hydroacylationO quent intermolecular Stetter reaction (eq 15). Rh(I) chiraphos yielded the corresponding cyclopentanone with up of unactivatedP(Cy)3 terminal alkynes. Surprisingly, treatment of + XS to 69% ee (Scheme 18). This level of selectivity was only N NH 33a with 8 and K2CO3 did not result in the formation of the achieved at low conversions (17%); at conversions of a RhCl(PPh ) (30 mol %), CH Cl , rt. Ph expected enone, but it led to the3 formation3 of chromanone2 2 50-60%, the enantioselectivity was reduced to 40%. The 34a, presumably by the addition of a second molecule of reactions were conducted using only 1 mol %∼ of catalyst, 33a to the intermediateTable enone 4. Stereoselective (eq 14). Synthesis of 3,4-Disubstituted although a temperature of 150 °C was needed to achieve Cyclopentanones reasonable conversions. One reason for this low reactivity Reaction of the 2-propargyloxywas the presence aldehydes of the33 with quaternary a center at C-2. second coupling aldehydeSakai catalyzed and co-workers by NHC generated were the first to report the cyclization from 8 resulted in the formationof achiral of 4-pentenals chromanone using with enantiomerically a pure catalysts.34,35 1,4-diketone motif in excellentSeveral yield. ligand It is systems noteworthy were that investigated with trans-1,2-bis- these reactions require only[(diphenylphosphino)methyl]cyclohexane rather low catalyst loading (DIPMC) delivering In view of this unprecedented reactivity, we decided (5 mol %) and proceed withthe highest high level selectivities. of selectivity. For example, A cyclization of phenyl- substituted pentenal 37 furnished the cyclopentanone product to employ a second aldehyde as the coupling partner for variety of substrates with different substitution patterns 35 entry substrate catalyst time (h) yield (%) cis/trans in 71% yield and 76% ee (Scheme 19). Relatively high catalyst loadings of 25 mol % were employed, and the 13R Rh[(S)-BINAP]ClO4 1.5 85 <1:>99 Vol. 44, No. 11 ’ 2011 ’ 1182–1195 ’reactionsACCOUNTS OF were CHEMICAL conducted RESEARCH ’ at1189 room temperature. 23R Rh[(R)-BINAP]ClO4 4 74 97:3 33S Rh[(R)-BINAP]ClO4 2 86 <1:>99 Both Sakai and Bosnich found that the use of cationic 43S Rh[(S)-BINAP]ClO4 5 82 96:4 rhodium complexes incorporating chiral ligands allowed highly enantioselective cyclizations of achiral 4-pentenals Table 3 allowed double diastereoselection to be explored, using low catalyst loadings. In the first report, Sakai found employingfirstneutralandthencationicrhodiumcomplexes.34-36 that the use of the catalyst [Rh(BINAP)]ClO4 allowed the cyclization of certain substrates to deliver enantioselectivities Sakai found that the use of the cationic complex Rh(BI- up to 99%.36 Later, Bosnich and co-workers evaluated a range NAP)ClO allowed the selective synthesis of all four possible 4 of bidentate chiral ligands and found that BINAP, chiraphos, stereoisomers of 3,4-disubstituted cyclopentanones (Table 36 and DuPhos all generated selective catalysts for certain 4). For example, reaction of the (3R)-substrate with 5 mol substrates (Table 5).39-41 An impressive range of substituents % of the catalyst employing (S)-BINAP delivered the trans could be tolerated ranging from sterically demanding electron- configured product with >99:1 selectivity. Conversely, cy- donating groups such as SiMe3 to much smaller electron- clization of the same substrate with the (R)-catalyst provided withdrawing groups such as C(O)Me. In all cases, conver- the cis product with 97:3 selectivity. The situation was sions of >95% were achieved with enantioselectivities for reversed for the (3S)-configured substrate. A similar pattern the majority of substrates falling in the 94-99% range. Aryl- of selectivity was obtained with the neutral complexes substituted products delivered the lowest selectivities (70-75% Rh(BINAP)Cl; however, significantly higher catalyst load- ee). To obtain the optimal levels of yield and enantioselec- ings (50 mol %) were needed, and the levels of selectivity tivity, it was crucial to match ligand selection to individual were inferior. Importantly, the matched and mismatched class of substrate. For example, BINAP was most effective Application of C–H and C–C bond activation in organic synthesis 581

(5)

of 23 can restrict the conformation of the benzyl group, and helps the Ru catalyst to approach to the benzylic C–H bond after the precoordination of the catalyst onto the pyridyl nitrogen atom of 23.

C–C BOND ACTIVATION OF UNSTRAINED KETONES A variety of chemical bonds can be activated by transition-metal complexes. However, the activations of carbon–carbon sigma bonds are much more difficult than those of carbon–hydrogen bonds. It is well known that the inertness of C–C sigma bond originates not only from its thermodynamic stability, but also from its kinetic inertness. Highly ring-strained cyclopropanes have frequently been used as a sub- strate for C–C bond activation since the use of them is beneficial kinetically as well as thermodynami- cally [3a]. Strained carbonyl compounds such as cyclobutanones are also used since the strained C–C single bond α to a carbonyl group is weaker than other C–C bonds [22]. Recently, it was demonstrated that the carbon–carbon bond of unstrained ketones could be cleaved by a rhodium catalyst using the chelation-assisted strategy. For example, when benzylacetone (26) was heated with 3,3-dimethylbut-1-ene in the presence of 1 and 2, 5,5-dimethylhexan-2-one (27) was isolated as a major product along with a trace amount of styrene (eq. 6) [23]. The alkyl-exchanged ketone 27 must be derived from the replacement of a phenethyl group of 26 with the 3,3-dimethylbutyl C-Cgroup through meets the cleavage of C-H an C–C bond α to the carbonyl group. As illustrated in Fig. 3, the reac- tion commences with the formation of ketimine 28 from the condensation of 26 with 2. Precoordination of a rhodium complex to the pyridyl group in 28 can facilitate the cleavage of the C–C bond α to the formaldehydeimino group. Thus, the α C–C bond in 28andcould be cleaved by the Rh(I) complex to afford metallacycle 29. β-Hydride elimination in 29 followed by the hydrometallation of 31 onto 3,3-dimethylbut-1-ene acetaldehydefurnishes (iminoacyl)rhodium(III) alkyl 32. Reductive elimination in 32 produces ketimine 33, which is hydrolyzed into the alkyl-exchanged ketone 27.

Application of C–H and C–C bond activation in organic synthesis 583 (6) activation would give another ketimine 45. Finally, hydrolysis of both 44 and 45 produces the unsym- metrical ketone 42 and the symmetrical ketone 41, respectively. As shown in eq. 7, C–C bond activations of unstrained cycloalkanone derivatives were also in- vestigated. A mixture of ring-contracted cycloalkanones, 2-methylcyclohexanone (37) and 2-ethyl- cyclopentanone (38), was obtained after hydrolysis when cycloheptanoketimine 34 was treated with (8) [Rh(C8H14)2Cl]2 (35) and tricyclohexylphosphine (36) in the absence of any external olefins [24]. Interestingly, any ring-opened product was not detected. The ring-contracted products 37 and 38 seem to be formed via (imino)acylrhodium(III) hydride 39, which could be generated through the C–C bond cleavage584 in 34 by a Rh(I) catalyst and subsequentC.-H. β-hydride JUN AND elimination J. H. LEE in the resulting intermediate.

(9) ©2004 IUPAC, Pure and Applied Chemistry 76, 577–587

CLEAVAGE OF C–C TRIPLE BOND A carbon–carbon triple bond of alkynes is one of the strongest chemical bonds. Therefore, the cleavage of alkyne triple bonds is restricted to a few examples including alkyne-ligand scission on transition- metal complexes [28a], oxidative cleavage [28b], alkyne metathesis [28c], and others [28d]. During the course of our investigation of the chelation-assisted activation of C–H and C–C bond, we found that the C–CFig. 4 tripleA plausible bond explanation of alkynes for the couldformation be of cleavedboth 41 and through 42. the hydroiminoacylation of alkynes using allyl- amine derivatives. SYNTHESISThe reaction OF CYCLOALKANONES of allylamine 40 withFROM but-2-yne DIENES AND(46) ALLYLAMINESin the presence of 1, cyclohexylamine (47), and 11 afforded a mixture of ketimine 48 and aldimine 49 determined by a GC-analysis (eq. 10) [29]. Acidic Intramolecular hydroacylation provides the most promising way to prepare cyclopentanones from pent- hydrolysis4-enal [26]. ofHowever, 48 yielded its application 1-phenylpentan-3-one has been limited to (the49 )preparation in a quantitative of five-membered yield. It rings is obvious be- that both ethyl- idenecause themoiety competing of 49 decarbonylationand ethyl group of the acylof 48 metalare hydride derived intermediate from the prevails cleavage during of the the forma- C–C triple bond of 46. Thetion ofproposed other larger mechanism rings. for this unusual transformation is depicted in Fig. 5. The transformation be- gins withAttempts the to isomerization utilize formaldehyde of 40 as ato substrate 43, which in the couldchelation-assisted undergo double hydroiminoacylation hydroacylation with 46 to give failed. Since decarbonylation cannot occur in the reaction with allylamine 40, we envisaged that its ap- αplication,,β-unsaturated as a synthetic ketimine equivalent 51. of The formaldehyde, conjugate into addition the cyclization of 47 into with dienes51 followed could furnish by retro-Mannich-type fragmentationcycloalkanones with in thevarious resulting sizes through β-amino the consecutive ketimine C–H 52 andfurnishes C–C bond enamine activation 53 (eq.along 9). When with aldimine 49. The enamine40 was allowed 53 is to isomerized react with nonsubstituted into ketimine dienes 54 such, which as penta-1,4-diene is then transiminated and hexa-1,5-diene, by cyclohexylamine not only 47 to give anotherthe corresponding ketimine cycloalkanones, 48 with the butregeneration ring-contracted of products1. Finally, were acidicalso obtained. hydrolysis For example, of both the 48 and 49 furnishes reaction of 40 with penta-1,4-diene was performed in the presence of [Rh(C8H14)2Cl]2 (35) and PCy3 the(36) finalto afford product cyclohexanone 50 and acetaldehyde,and 2-methylcyclopentanone respectively. in a 87 and 13 % yield, respectively [27]. Cycloheptanone, 2-methylcyclohexanone, and 2-ethylcyclopentanone were obtained in a ratio of 38:40:22 from the reaction of 40 with hexa-1,5-diene. However, the reaction with 1,4- or 1,5-dienes bearing substituents at C2 or C3 position gave substituted cyclohexanones or cycloheptanones, respec- tively, as a sole product without forming ring-contracted products.

(10) ©2004 IUPAC, Pure and Applied Chemistry 76, 577–587

A variety of alkynes can be successfully applied into this unique transformation to afford various ketones in good to excellent yields. However, an allylamine as a substrate cause a limitation to its ver- satile use. We were intrigued by the possibility of development of a more common protocol for the C–C triple bond cleavage using aldehyde instead of allylamine 40. It is speculated that the key inter- mediate of the alkyne triple bond cleavage reaction, an α,β-unsaturated ketimine such as 51 (Fig. 5), could also be derived from an aldehyde and an alkyne using a cocatalyst of 1 and 2. To examine this hypothesis, acetaldehyde was allowed to react with dodec-6-yne in the presence of 1, 2, 47, and alu- minum chloride. As expected, octan-2-one was obtained in a good yield after hydrolysis (eq. 11) [30].

©2004 IUPAC, Pure and Applied Chemistry 76, 577–587