Chinese Chemical Letters  2014, Vol.25 Issue (01):1-8   PDF    
Asymmetric catalytic anhydride openings via carbon-based nucleophiles
Chen Zheng, Fen-Er Chen     
* Corresponding authors at:Department of Chemistry, Fudan University, Shanghai 200433, China
Abstract: The asymmetric desymmetrization of cyclic anhydrides via the addition of carbon-based nucleophiles has been the focus of considerable levels of interest because it leads to optically active products. Over the past 20 years, a variety of different catalytic asymmetric alkylation reactions have been developed for the desymmetrization of cyclic anhydrides using different metal reagents as nucleophiles and using chiral ligands. The purpose of this review is to provide an overview of significant developments in this field.
Key words: Cyclic anhydrides     Carbon-based nucleophiles     Chiral ligands     Asymmetric catalysis     Metal reagents    
1. Introduction

The desymmetrization of meso cyclic anhydrides via the addition of carbon-based nucleophiles represents a well-established and powerful synthetic tool in asymmetric synthesis because it allows for the construction of multiple stereogenic centers in one symmetry-breaking operation [1]. Over the past two decades, the asymmetric opening of anhydrides with carbon-based nucleophiles has been extensively investigated because of the great potential for the application of this technique in the asymmetric synthesis of important chiral intermediates and optically active fine chemicals (Scheme 1) [2].

Download:
Scheme 1. Asymmetric anhydride openings via carbon-based nucleophiles.
The asymmetric anhydride alkylation reaction using chiral Grignard reagents was pioneered by Real et al. [3] in 1993. Research in this area has been subsequently shifted to focus on the transition-metal catalyzed alkylation of anhydrides due to the mild reaction conditions and high stereoselectivity associated with these transformations. Nickel-, palladium- and rhodium-based systems have all been successfully applied as catalysts to the anhydride alkylation reaction. This review presents a comprehensive overview of both newly developed and well-established strategies in the field of asymmetric anhydride opening by the addition of carbon-based nucleophiles.

2. Asymmetric anhydride alkylation using organozinc reagents as nucleophiles

The ability of transition-metal complexes to mediate carbon- carbon bond-forming reactions makes them invaluable tools in organic synthesis [4]. Rovis et al. [5] provided a detailed description of the investigations involved in the asymmetric alkylation of anhydrides catalyzed by transition-metals. Nickel-, palladium- and rhodium-based catalysts have all been successfully applied to the desymmetrization of prochiral cyclic anhydrides.

2.1. Nickel-catalyzed anhydride alkylation

Non-stereoselective methods for the desymmetrization of cyclic anhydrides were initially studied by Rovis [5], providing a platform for the subsequent development of the asymmetric anhydride alkylation reaction. The 2,2'-bipyridyl (bipy) nickel complex was found to be highly effective in promoting the alkylation of succinic anhydrides, whereas the use of the (2- diphenylphosphino)ethylpyridine (pyphos)-derived nickel complex was successfully used to facilitate the alkylation of glutaric anhydrides to give the corresponding keto acids in high yields [5a,c]. Interestingly, the use of nickel allowed for the introduction of chiral ligands to provide enantioselectivity over the bond-forming event. As shown in Scheme 2, the use of the phosphino-oxazoline ligand (i-PrPHOX, 1) afforded an active catalyst for the alkylation of cyclohexane-1,2-dicarboxylic anhydride (2) leading to the formation of the keto acid product in 85% yield and 79% ee [5a].

Download:
Scheme 2. Ni-catalyzed asymmetric alkylation of 2.
The alkylation of cyclic anhydrides with diorganozinc reagents can be efficiently catalyzed by adding nickel in conjunction with an appropriate ligand. There are limitations to this protocol, including the fact that only one of the two zinc substituents is transferred during the reaction. To address this issue and conduct a thorough investigation of the Ni-catalyzed alkylation of anhydride, Rovis et al. employed mixed zinc reagents [6] as nucleophiles and evaluated their effects on the levels of enantioselectivity in the products. As shown in Table 1, particularly high levels of selectivity for Ph transfer were achieved when i-Pr2Zn or (TMSCH2)2Zn were used as the second zinc reagents (Table 1, entries 6-9), and even higher levels of enantioselectivity were observed when Et2Zn was used as the second zinc reagent (Table 1, entries 3-5). It is noteworthy that the asymmetric alkylation of anhydrides was proceeded with a lower enantioselectivity when Ph2Zn prepared in situ from the reaction of ArBr and n-BuLi with ZnCl2 as the zinc reagent [7].

Table 1
Ni-catalyzed asymmetric opening of 4.
In 2005, two possible mechanistic routes to this asymmetric reaction were proposed by Rovis group [5c]. As shown in Scheme 3, Pathway A involves the initial formation of an alkyl nickel intermediate Ⅰ from direct alkyl group transfer from the zinc reagent to the catalyst, whereas pathway B involves oxidative addition of the low-valent nickel complex to the electron-deficient C-O bond of the cyclic anhydride to provide carboxylate Ⅱ. Intermediate Ⅰ could then react with the anhydride directly yielding intermediate Ⅲ which would provide the desired product upon collapse. Transmetalation of Ⅱ yields Ⅳ which can then provide the final product along with regeneration of the active catalyst by reductive elimination.

Download:
Scheme 3. The proposed mechanistic pathways.
2.2. Palladium-catalyzed anhydride alkylation

An improved protocol for the enantioselective desymmetrization of cyclic anhydrides using palladium-based catalysts was also reported (Scheme 4) [8]. The enantioselective ring-opening of cyclic anhydrides was initially investigated using (S)-2,2'-bis(diphenylphosphino)- 1,1'-binaphthyl [(S)-BINAP] as the ligand and Pd(OAc2 as the palladium source, with the corresponding keto acid product obtained in 67% yield and 77% ee. The enantioselectivity of the transformation was improved by biphenyl ligands 6 and 7. However, other changes of the electronic nature of the aromatic phosphine portion of the ligand failed to increase the enantioselectivity.

Download:
Scheme 4. Ligand screen for the asymmetric arylation of 4.
In addition, the asymmetric arylation of substrate 4 proceeded smoothly at lower temperatures to give the desired product with an improved yield and comparable level of enantioselectivity (Scheme 5).

Download:
Scheme 5. Pd-catalyzed asymmetric arylation of 4.
Furthermore, the palladium-catalyzed addition of diphenylzinc to structurally diverse succinic anhydrides was carried out smoothly at ambient temperature to give the corresponding keto acids in high yields (61-89%) with excellent enantioselectivity (89-97%) (Table 2).

Table 2
Pd-catalyzed asymmetric arylation of succinic anhydrides.
Under similar conditions, dimethylzinc was used as the nucleophile to provide the desired product in 78% yield and 64% ee (Table 3, entry 1). The addition of a catalytic amount of 4- fluorostyrene afforded the desired keto acid in 80% yield and 91% ee (Table 3, entry 2). In addition, the efficiency of the reaction was highly dependent on the molar ratio of the palladium to the ligand ratio (Table 3, entries 3-5).

Table 3
Pd-catalyzed asymmetric alkylation of 4.
2.3. Rhodium-catalyzed anhydride alkylation

Although the diorganozinc reagents have been used to provide excellent results in both nickel- and palladium-catalyzed methodologies, the commercial availability of the diorganozinc reagents is quite limited in comparison. For this reason, research efforts in this area have been shifted toward the use of zinc nucleophiles generated in situ. Pleasingly, a procedure of enantioselective desymmetrization of anhydrides using organozinc nucleophiles prepared in situ has been successfully developed using rhodium catalysis. Unfortunately, the use of zinc nucleophiles formed in situ was incompatible with the palladiumcatalyzed methodology [9].

2.3.1. Rhodium/phosphoramidite catalyst

In 2007, Rovis et al. [10] reported the successful asymmetric alkylation of cis-2,3-dimethylsuccinic anhydride (16) with 3,4,5- trimethoxyphenylzinc triflate in the presence of Taddol-PNMe2 18 to give the desired product in 85% yield and 87% ee. The 3,4,5- trimethoxyphenylzinc triflate reagent involved in this reaction was generated from the corresponding aryl bromide by lithiumhalogen exchange and subsequent reaction with zinc triflate. The reaction was found to be tolerant of a wide range of nucleophiles (Table 4), providing the corresponding keto acids in good yields (74-88%) with high enantioselectivity (80-88% ee). Furthermore, the newly developed methodology was also successfully applied to the total synthesis of three eupomatilone lignans [11].

Table 4
Rh-catalyzed asymmetric alkylation of 16.
A variety of different anhydrides were found to be treated using this chemistry, including substrates containing backbone olefins and strained rings, which produced the desired products with moderate yields and reasonable enantioselectivity (76-83% ee) (Table 5).

Table 5
Rh-catalyzed asymmetric arylation of succinic anhydrides.
2.3.2. Rhodium/PHOX catalyst

Although nickel- and palladium-based catalysts are largely ineffective for the addition of alkyl nucleophiles to glutaric anhydrides, Rovis et al. [12] developed a rhodium-catalyzed enantioselective alkylation reaction for the desymmetrization of meso glutaric anhydrides. An initial period of reaction optimization revealed that the alkylation of anhydride 25 with tertbutylPHOX 29 provided the best results in the presence of [Rh(nbd)Cl]2 at 25 ℃ (Table 6).

Table 6
Catalyst and ligand optimization of alkylation of 25.
The reaction was also tolerant of a broad range of functionalities, including alkyl substituents, esters, and alkyl chlorides (Scheme 6).

Download:
Scheme 6. Rh-catalyzed asymmetric alkylation.
In addition, a variety of other 3,5-disubstituted glutaric anhydrides were also well tolerated under the reaction conditions. For example, the bicyclic anhydride 30 produced the corresponding keto acid in 87% yield and 85% ee, whereas the bis-acetate anhydride 32 gave the desired product in 65% yield and 84% ee (Scheme 7).

Download:
Scheme 7. Rh-catalyzed asymmetric alkylation of glutaric anhydrides.
Although this methodology proceeded efficiently and selectively with alkyl zinc nucleophiles, it was not compatible when aryl zinc nucleophiles were used. In 2009, Rovis et al. [13] reported an interesting observation pertaining to the potential use of aryl zinc nucleophiles in this particular reaction. Thus, the desymmetrization of cyclohexenecarboxylic anhydride (4) with an aryl zinc nucleophile produced the corresponding keto acid in 78% yield and 76% ee (Scheme 8).

Download:
Scheme 8. Rh-catalyzed asymmetric arylation of 4.
Another remarkable result was achieved when the desymmetrization of glutaric anhydride 25 was performed using bis(homoenolate) zinc as the nucleophile. In this particular case, when the alkylation was carried out with a substrate concentration of 0.15 mol/L, the desired product 37 was obtained in 49% yield and 95% ee, with the rearrangement byproduct 38 isolated from the reaction mixture in 30% yield and 0% ee (Scheme 9). Interestingly, however, when a substrate concentration of 0.3 mol/L was used, the corresponding keto acid was produced in 75% yield and 95% ee, with none of the byproduct being observed. When the reaction was performed with i-Pr-PHOX 1 as the ligand, none of the desired product was obtained, with the rearrangement product being isolated as the sole product of the reaction.

Download:
Scheme 9. Rh-catalyzed asymmetric opening of 25.
The rhodium/PHOX catalyst system efficiently catalyzes the enantioselective desymmetrization of cyclic anhydrides, particularly 3,5-dimethylglutaric anhydride, using alkyl zinc halide nucleophiles. In contrast, the rhodium/phosphoramidite catalyst system has been used for the efficient desymmetrization of succinic anhydrides using in situ prepared aryl zinc triflate reagents. Unfortunately, alkyl zinc reagents are incompatible with this particular catalyst system.

3. Asymmetric anhydride alkylation using Grignard reagents as nucleophiles

The first desymmetrization of cyclic anhydrides using carbon nucleophiles was reported by Real et al. [3] with chiral Grignard reagents being employed as the carbon-based nucleophiles. The addition of (-)-ephedrine-derived Grignard reagents 39a to anhydride 41 produced a solution of the salts 42a and 42b, which were further converted to 43 and 44 (66% ee) in 56% overall yield by the addition of NaBH4 followed by acid hydrolysis. Under the same conditions, the use of the (+)-pseudoephedrine-derived Grignard reagent 39b provided aldehyde 44 (>99.4% de) in 65% overall yield and 99% ee. The reaction with the compound 39c under the same conditions provided product 44 in 64% overall yield and 99.2% ee (Scheme 10).

Download:
Scheme 10. Asymmetric arylation of 41 via Grignard reagents.
The pioneering work of Real has been followed by a series of particularly interesting investigations from a number of different groups, including those of Fu, Rovis, and Harada. Fu et al. [14] found that the inclusion of commercially available (-)-sparteine allowed for the enantioselective ring-opening reaction of 3-phenylglutaric anhydride (45) with phenyl magnesium chloride to give the corresponding δ-keto acid in 63% yield and 88% ee (Scheme 11) [14].

Download:
Scheme 11. Asymmetric arylation of 45 via Grignard reagents.
Under the same conditions, a variety of different 3-substituted glutaric anhydrides reacted with PhMgCl/(-)-sparteine to give the corresponding keto acids in high levels of enantioselectivity (87- 92% ee) and good yields (51-91%) (Table 7).

Table 7
Asymmetric openings of anhydrides via Grignard reagents
Following Fu’s protocol, the optimal N-Me (+)-sparteine-like diamine 52 was evaluated with success in the asymmetric opening of one meso anhydride using phenylmagnesium chloride by O’Brien group [15]. Reaction of phenylmagnesium chloride/52 with meso anhydride 53 in toluene at -78 ℃ for 20 h generated a 78% yield of keto acid with 78% ee (Scheme 12).

Download:
Scheme 12. Asymmetric opening of 53 via Grignard reagents.
4. Asymmetric anhydride alkylation using methallylstannane as nucleophiles

In 2007, Harada [16] reported the enantioselective ringopening reaction of cyclic anhydrides with methallylstannane in the presence of an oxazaborolidine, which acts as a Lewis acid catalyst [17]. As shown in Scheme 14, the treatment of anhydride 55 with 2 equivalent of methallylstannane (56) in the presence of OXB 57a (30 mol%) at room temperature in CH2Cl2 provided the desired alkylation product 58. Subsequent treatment with base led to the sequential fragmentation and epimerization of the product, followed by acidic work-up and esterification to give the keto ester 60 in 91% yield and 16% ee (Scheme 13).

Download:
Scheme 13. Asymmetric opening of 55 via methallylstannane.
As shown in Table 8, when the bicyclic (55, 61) and tricyclic (62, 63) anhydrides were subjected to the ring-opening reaction at ambient temperature, they gave the trans-acetyl esters in good yields (80-96%). Unfortunately, however, no appreciable enantioselectivity was observed in these reactions (16-29% ee) (Table 8, entries 1, 3, 5, and 7). The enantioselectivity of the transformation could be improved slightly at lower temperatures, although the use of lower temperatures led to a reduction in the yield (Table 8, entries 2, 4, and 6).

Table 8
Asymmetric opening of anhydrides via methallylstannane.
An increase in the enantioselectivity was observed when structurally modified OXB catalysts were used. This increase in enantioselectivity was attributed to the enhanced Lewis acidity of the modified OXB catalysts, which provided enhanced selectivity at lower temperatures. For example, the treatment of the tricyclic anhydride 62 with 56 in the presence of OXB-BCl 64 at -78 ℃ gave the corresponding trans-keto ester in 80% ee and 13% yield (Scheme 14).

Download:
Scheme 14. Asymmetric opening of 62 via methallylstannane.
5. Conclusion

A number of different approaches and protocols have been developed over the past two decades for the highly enantioselective ring-opening of cyclic anhydrides for asymmetric C-C bondforming reactions. Herein, we have reviewed developments in the enantioselective desymmetrization of cyclic anhydrides with carbon-based nucleophiles. Although good yields and high enantioselectivity have been achieved using Grignard reagents and organozinc-based nucleophiles, the successful alkylation of anhydrides with high enantioselectivity using carbon nucleophiles remains challenging. The key challenges of this approach are the accessibility of the reagents and catalysts, as well as the ease of product purification. Furthermore, the discovery of new catalysts and chiral ligands that could allow for significant expansion in the current scope of the substrates and nucleophiles amenable to this approach represents an even greater challenge. To meet these challenges, considerable efforts are required to develop a broader mechanistic understanding of the reasons behind the success of the currently efficient catalytic systems, and to allow for the continuous exploration of a large range of catalysts.

References
[1] (a) S.R. Magnuson, Two-directional synthesis and its use in natural product synthesis, Tetrahedron 51 (1995) 2167-2213; (b) P. Magnus, I.K. Sebhat, Synthesis of the antitumor alkaloid (+)-pancratistatin using the β-azidonation reaction via a prochiral 4-arylcyclohexanone derivative, J. Am. Chem. Soc. 120 (1998) 5341-5342; (c) J.D. White, J. Kim, N.E. Drapela, Enantiospecific synthesis of (+)-byssochlamic acid, a nonadride from the ascomycete byssochlamys fulva, J. Am. Chem. Soc. 122 (2000) 8665-8671; (d) E. Garcia-Urdiales, I. Alfonso, V. Gotor, Enantioselective enzymatic desymmetrizations in organic synthesis, Chem. Rev. 105 (2005) 313-354.
[2] (a) B.M. Trost, D.L. van Vranken, Asymmetric transition metal-catalyzed allylic alkylations, Chem. Rev. 96 (1996) 395-422; (b) A.C. Spivey, B.I. Andrews, Catalysis of the asymmetric desymmetrization of cyclic anhydrides by nucleophilic ring-opening with alcohols, Angew. Chem. Int. Ed. 40 (2001) 3131-3134; (c) Y. Chen, P. McDaid, L. Deng, Asymmetric alcoholysis of cyclic anhydrides, Chem. Rev. 103 (2003) 2965-2983; (d) C. Bolm, I. Atodiresei, I. Schiffers, M. Kanai, M. Shibasaki, Asymmetric alcoholysis of meso-anhydrides mediated by alkaloids, Org. Synth. 82 (2005) 120-125; (e) I. Atodiresei, I. Schiffers, C. Bolm, Stereoselective anhydride openings, Chem. Rev. 107 (2007) 5683-5712.
[3] S.D. Real, D.R. Kronenthal, H.Y. Wu, A novel and highly efficient desymmetrization of a meso-anhydride by a chiral grignard reagent, Tetrahedron Lett. 34 (1993) 8063-8066.
[4] F. Diederich, P.T. Stang, Metal-Catalyzed Cross-coupling Reactions, Wiley-VCH, Weinheim, 1998.
[5] (a) E.A. Bercot, T. Rovis, A mild and efficient catalytic alkylative monofunctionalization of cyclic anhydrides, J. Am. Chem. Soc. 124 (2002) 174-175; (b) E.M. O'Brien, E.A. Bercot, T. Rovis, Decarbonylative cross-coupling of cyclic anhydrides: introducing stereochemistry at an sp3 carbon in the cross-coupling event, J. Am. Chem. Soc. 125 (2003) 10498-10499; (c) E.A. Bercot, T. Rovis, Highly efficient nickel-catalyzed cross-coupling of succinic and glutaric anhydrides with organozinc reagents, J. Am. Chem. Soc. 127 (2005) 247-254; (d) R.L. Rogers, J.L. Moore, T. Rovis, Alkene-directed regioselective nickel-catalyzed cross-coupling of cyclic anhydrides with diorganozinc reagents, Angew. Chem. Int. Ed. 46 (2007) 9301-9304.
[6] (a) P. Jones, C.K. Reddy, P. Knocel, Conjugate Michael additions with mixed diorganozincs, Tetrahedron 54 (1998) 1471-1490; (b) C. Bolm, N. Hermanns, J.P. Hildebrand, K. Muñiz, Asymmetric catalytic phenyl transfer to aldehydes: enantioselective synthesis of diarylmethanols, Angew. Chem. Int. Ed. 39 (2000) 3465-3467; (c) A.E. Jensen, P. Knochel, Nickel-catalyzed cross-coupling between functionalized primary or secondary alkylzinc halides and primary alkyl halides, J. Org. Chem. 67 (2002) 79-85; (d) D. Soorukram, P. Knochel, Enantioselective synthesis ofα-Ionone derivatives using an anti SN20 substitution of functionalized zinc organometallics, Org. Lett. 6 (2004) 2409-2411.
[7] J.B. Johnson, R.T. Yu, P. Fink, E.A. Bercot, T. Rovis, Selective substituent transfer from mixed zinc reagents in Ni-catalyzed anhydride alkylation, Org. Lett. 8 (2006) 4307-4310.
[8] E.A. Bercot, T. Rovis, A palladium-catalyzed enantioselective alkylative desymmetrization of meso-succinic anhydrides, J. Am. Chem. Soc. 126 (2004) 10248-10249.
[9] J.G. Kim, P.J. Walsh, From aryl bromides to enantioenriched benzylic alcohols in a single flask: catalytic asymmetric arylation of aldehydes, Angew. Chem. Int. Ed. 45 (2006) 4175-4178.
[10] J.B. Johnson, E.A. Bercot, C.M. Williams, T. Rovis, A concise synthesis of eupomatilones 4, 6, and 7 by rhodium-catalyzed enantioselective desymmetrization of cyclic meso anhydrides with organozinc reagents generated in situ, Angew. Chem. Int. Ed. 46 (2007) 4514-4518.
[11] (a) A.R. Carroll, W.C. Taylor, Constituents of Eupomatia species. XII. Isolation of constituents of the tubers and aerial parts of Eupomatia bennettii and determination of the structures of new alkaloids from the aerial parts of E. bennettii and minor alkaloids of E. laurina, Aust. J. Chem. 44 (1991) 1615-1626; (b) A.R. Carroll, W.C. Taylor, Constituents of Eupomatia species. XIV. The structures of eupomatilone-1, -2, -3, -4, -5, -6 and -7 isolated from Eupomatia bennettii, Aust. J. Chem. 44 (1991) 1705-1714.
[12] M.J. Cook, T. Rovis, Rhodium-catalyzed enantioselective desymmetrization of meso-3,5-dimethyl glutaric anhydride: a general strategy to syn-deoxypolypropionate synthons, J. Am. Chem. Soc. 129 (2007) 9302-9303.
[13] J.B. Johnson, M.J. Cook, T. Rovis, Ligand differentiated complementary Rh-catalyst systems for the enantioselective desymmetrization of meso-cyclic anhydrides, Tetrahedron 65 (2009) 3202-3210.
[14] R. Shintani, G.C. Fu, Highly enantioselective desymmetrization of anhydrides by carbon nucleophiles: reactions of Grignard reagents in the presence of (-)-sparteine, Angew. Chem. Int. Ed. 41 (2002) 1057-1059.
[15] M.J. Dearden, M.J. McGrath, P. O'Brien, Evaluation of (+)-sparteine-like diamines for asymmetric synthesis, J. Org. Chem. 69 (2004) 5789-5792.
[16] J. Suzuki, T. Harada, Oxazaborolidinone-catalyzed alkylative ring-opening reaction of cyclic anhydrides with methallylstannane, Synthesis 15 (2006) 2483-2488.
[17] (a) D. Seebach, R.E. Marti, T. Hintermann, Polymer-and dendrimer-bound Ti-TADDOLates in catalytic (and stoichiometric) enantioselective reactions. Are pentacoordinate cationic Ti complexes the catalytically active species? Helv. Chim. Acta 79 (1996) 1710-1740; (b) D. Seebach, G. Jaeschke, K. Gottewald, et al., Resolution of racemic carboxylic acid derivatives by Ti-TADDOLate mediated esterification reactions -a general method for the preparation of enantiopure compounds, Tetrahedron 53 (1997) 7539-7556; (c) G. Jaeschke, D. Seebach, Highly enantioselective ring opening of cyclic meso anhydrides to isopropyl hemiesters with Ti-TADDOLates: an alternative to hydrolytic enzymes? J. Org. Chem. 63 (1998) 1190-1197.