Skip to main content
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2015 Jul 20;137(29):9438–9442. doi: 10.1021/jacs.5b05528

Phosphine-Catalyzed Doubly Stereoconvergent γ-Additions of Racemic Heterocycles to Racemic Allenoates: The Catalytic Enantioselective Synthesis of Protected α,α-Disubstituted α-Amino Acid Derivatives

Marcin Kalek 1, Gregory C Fu 1,*
PMCID: PMC4577964  PMID: 26192217

Abstract

graphic file with name ja-2015-05528p_0015.jpg

Methods have recently been developed for the phosphine-catalyzed asymmetric γ-addition of nucleophiles to readily available allenoates and alkynoates to generate useful α,β-unsaturated carbonyl compounds that bear a stereogenic center in either the γ or the δ position (but not both) with high stereoselectivity. The utility of this approach would be enhanced considerably if the stereochemistry at both termini of the new bond could be controlled effectively. In this report, we describe the achievement of this objective, specifically, that a chiral phosphepine can catalyze the stereoconvergent γ-addition of a racemic nucleophile to a racemic electrophile; through the choice of an appropriate heterocycle as the nucleophilic partner, this new method enables the synthesis of protected α,α-disubstituted α-amino acid derivatives in good yield, diastereoselectivity, and enantioselectivity.

Introduction

In recent years, nucleophilic catalysis by chiral tertiary phosphines has emerged as a powerful strategy for the enantioselective synthesis of a wide array of useful compounds.1 Exemplary of this approach is the phosphine-catalyzed addition of nucleophiles to the γ-position of electron-deficient alkynes and allenes,2 which furnishes α,β-unsaturated carbonyl compounds that are well-suited for further stereoselective functionalizations (eqs 13 and 24).5

graphic file with name ja-2015-05528p_0002.jpg 1

To date, systematic studies of phosphine-catalyzed enantioselective γ-additions have been limited to processes that control a single stereocenter, either in the γ or in the δ position (eqs 1 and 2, respectively).3,4 We recently decided to attempt to substantially enhance the utility of this strategy by developing a method that would control the stereochemistry at both termini of the newly created bond (eq 3). To the best of our knowledge, there has been only a single isolated example of a phosphine-catalyzed γ-addition between such partners, a coupling that proceeded with modest diastereoselectivity (2.5:1) and good enantiomeric excess (91–92%).3c

graphic file with name ja-2015-05528p_0003.jpg 3

Although we were concerned that this effort might be stymied by the propensity of the electrophilic allene/alkyne to isomerize to a 1,3-diene6 rather than form a carbon–carbon bond with a hindered tertiary nucleophile, we were particularly attracted by the potential to generate products that are more stereochemically rich than in previous studies and to achieve a doubly stereoconvergent reaction of two racemic coupling partners, one with a stereogenic center and the other with a stereogenic axis. In this report, we describe the achievement of our objective, specifically, a phosphine-catalyzed γ-addition that furnishes α,α-disubstituted α-amino acid derivatives,7,8 an important family of target molecules, with very good diastereoselectivity and enantioselectivity (eq 4).

graphic file with name ja-2015-05528p_0004.jpg 4

Results and Discussion

Upon exploring a variety of reaction parameters for a model coupling between a 1,3-oxazol-5(4H)-one and an allenoate, we have been able to develop conditions that furnish the desired γ-addition product in good yield, diastereoselectivity, and enantioselectivity in the presence of phosphepine 1(9) (Table 1, entry 1). No γ-addition is observed in the absence of the phosphine (entry 2), and related phosphepines (2 and 3) as well as spirophosphine 4(10) do not provide satisfactory results (entries 3–5). A modestly inferior yield, diastereoselectivity, and/or enantioselectivity are obtained at room temperature or in the absence of the phenol (entries 6 and 7).11

Table 1. Phosphine-Catalyzed Doubly Stereoconvergent γ-Additions: Effect of Reaction Parametersa.

graphic file with name ja-2015-05528p_0007.jpg

graphic file with name ja-2015-05528p_0017.jpg

a

All data are the average of two experiments.

b

Combined yield of the two diastereomers.

c

Determined through analysis by 1H NMR spectroscopy.

d

ee of the major diastereomer. A negative ee value signifies the predominant formation of the enantiomer of the illustrated product.d

Having identified effective conditions, we examined the scope of this new method for the catalytic asymmetric synthesis of α,α-disubstituted α-amino acid derivatives. With regard to the allenoate coupling partner (Table 2),12 the choice of the R2 group of the ester has only a modest impact on the course of the reaction (entries 1–4). A variety of substituents in the γ-position of the allenoate (R1) are tolerated, and functional groups such as a silyl ether, a cis alkene, an alkyne, a thiophene, and a primary alkyl chloride are compatible with the reaction conditions (entries 5–13). On a gram scale (1.17 g of product), the catalytic enantioselective γ-addition illustrated in entry 10 proceeds in 90% yield, 15:1 dr, and 94% ee.

Table 2. Phosphine-Catalyzed Doubly Stereoconvergent γ-Additions: Scope with Respect to the Allenoatea.

graphic file with name ja-2015-05528p_0009.jpg

graphic file with name ja-2015-05528p_0010.jpg

a

All data are the average of two experiments.

b

Yield of purified product, isolated as a mixture of diastereomers.

c

Determined through analysis by 1H NMR spectroscopy of the unpurified reaction mixture.

d

ee of the major diastereomer.

e

Catalyst loading: 10%.

We have also examined the scope with respect to the nucleophilic partner (the 1,3-oxazol-5(4H)-one) in the doubly stereoconvergent γ-addition (Table 3).13 As the substituent R in the 4 position increases in size from Me to i-Bu, a small loss in ee is observed, although the yield and the diastereoselectivity remain high (entries 1–3). Functional groups such as an olefin, an ether, an imide, a protected indole, and a sulfide are tolerated. The enantiomeric excess of the product is independent of whether racemic or enantioenriched 1,3-oxazol-5(4H)-one is employed as the nucleophile.

Table 3. Phosphine-Catalyzed Doubly Stereoconvergent γ-Additions: Scope with Respect to the 1,3-Oxazol-5(4H)-onea.

graphic file with name ja-2015-05528p_0011.jpg

graphic file with name ja-2015-05528p_0012.jpg

a

All data are the average of two experiments.

b

Yield of purified product, isolated as a mixture of diastereomers.

c

Determined through analysis by 1H NMR spectroscopy of the unpurified reaction mixture.

d

ee of the major diastereomer.

e

The diastereomeric products were isolated separately.

An attempt to achieve a doubly stereoconvergent γ-addition of a 1,3-oxazol-5(4H)-one bearing an aromatic substituent in the 4 position furnished essentially none of the desired product (<1% yield). Building on our hypothesis that the greater stability of the conjugate base of this coupling partner might be attenuating its nucleophilicity, we decided to explore the reactivity of a more electron-rich family of heterocycles, specifically, 2-amino-3,5-dihydro-4H-imidazol-4-ones. We were pleased to determine that, under related conditions, an array of γ-additions of this family of aryl-substituted nucleophiles can be achieved in good yield, diastereoselectivity, and enantioselectivity (Table 4). In addition to serving as protected α,α-disubstituted α-amino acid derivatives, 2-amino-3,5-dihydro-4H-imidazol-4-ones are interesting targets in their own right.14

Table 4. Phosphine-Catalyzed Doubly Stereoconvergent γ-Additions: 2-Amino-3,5-dihydro-4H-imidazol-4-ones as Nucleophilesa.

graphic file with name ja-2015-05528p_0013.jpg

graphic file with name ja-2015-05528p_0014.jpg

a

All data are the average of two experiments.

b

Yield of purified product, isolated as a mixture of diastereomers.

c

Determined through analysis by 1H NMR spectroscopy of the unpurified reaction mixture.

d

ee of the major diastereomer.

The products of our phosphine-catalyzed stereoconvergent γ-additions can be transformed into other interesting compounds. For instance, selective hydrolytic ring opening of the 1,3-oxazol-5(4H)-one reveals an N-protected α,α-disubstituted α-amino acid (eq 5).7 Furthermore, ruthenium-catalyzed dihydroxylation, followed by in situ lactonization, proceeds with high diastereoselectivity to afford a product that bears four consecutive stereocenters wherein the newly introduced oxygens are differentiated (eq 6).

graphic file with name ja-2015-05528p_0005.jpg 5

A possible mechanism for these phosphine-catalyzed γ-additions of heterocycles to allenoates is outlined in Figure 1. In the first step of the catalytic cycle, the nucleophilic phosphine catalyst adds to the β position of the allenoate to furnish zwitterion A, which upon protonation by the 1,3-oxazol-5(4H)-one affords ion pair B.15 Next, the deprotonated heterocycle adds γ to the carbonyl group of the phosphonium salt, generating ylide C; if the phosphonium salt (B) is instead deprotonated in the δ position, then the undesired diene can be formed.6a Tautomerization (CD) followed by fragmentation yields the desired γ-addition product and regenerates the phosphine catalyst.

Figure 1.

Figure 1

Outline of a possible mechanism for the phosphine-catalyzed γ-addition of 1,3-oxazol-5(4H)-ones to allenoates (for the sake of simplicity, all steps are drawn as irreversible, and alkenes are illustrated as single isomers).

By following the progress of a phosphine-catalyzed enantioselective γ-addition over time, we have determined that there is a modest kinetic resolution of the racemic allenoate (selectivity factor ∼4).16 We have separated the enantiomers of the allenoate, and we have confirmed that they do indeed react at different rates and that they generate the γ-addition product with the same enantiomeric excess, consistent with the mechanism outlined in Figure 1 wherein the original stereochemistry of the allenoate is lost in the formation of intermediate A. We have determined that the unreacted 1,3-oxazol-5(4H)-one (pKa ∼19 in DMSO)17 is essentially racemic throughout the course of the reaction. The ee of the product remains constant during the γ-addition process.

The rate law for the phosphine-catalyzed γ-addition of a 1,3-oxazol-5(4H)-one to an allenoate is first-order in the catalyst and in the allenoate, and it is zeroth-order in the nucleophile and in the phenol; moreover, we have established through a 31P NMR spectroscopic study that the resting state of the catalyst is the free phosphine. Taken together, these observations are consistent with a mechanism wherein the first step of the catalytic cycle, the addition of the phosphine catalyst to the allenoate to form zwitterion A, is the turnover-limiting step.18

Conclusions

We have developed the first phosphine-catalyzed γ-addition reactions in which two adjacent stereogenic centers are controlled with very good diastereoselectivity and enantioselectivity; this doubly stereoconvergent process employs two racemic coupling partners, one with a stereogenic center and the other with a stereogenic axis. The method provides ready access to an array of protected, unsaturated α,α-disubstituted α-amino acid derivatives, which can undergo selective transformations that further enlarge the diversity of products that can be generated through this method. A mechanistic investigation establishes that, during the course of the γ-addition, there is modest kinetic resolution of the racemic allenoate, whereas the nucleophile remains racemic. The available data are consistent with the addition of the phosphine catalyst to the allenoate being the turnover-limiting step of the catalytic cycle. Further studies of phosphine-catalyzed stereoselective reactions are underway.

Experimental Section

General Procedure

An oven-dried 20 mL vial was charged with catalyst 1 (6.8 mg, 0.018 mmol, 5%), 2-chloro-6-methylphenol (4.2 μL, 5.0 mg, 0.035 mmol, 10%), and the 1,3-oxazol-5(4H)-one (0.35 mmol). The vial was capped with a PTFE-lined septum cap and evacuated/backfilled with nitrogen (3 cycles). Diisopropyl ether (anhydrous; 3.5 mL) was added via syringe, and then the vial was cooled to 0 °C. Next, the allenoate (0.42 mmol, 1.2 equiv) was added via syringe, and then the reaction mixture was stirred at 0 °C for 24 h. To deactivate the catalyst, a solution of tert-butyl hydroperoxide (5.0–6.0 M in decane; 50 μL) was added. The resulting mixture was stirred at 0 °C for 10 min, and then it was allowed to warm to rt. The mixture was concentrated under reduced pressure, and the residue was purified by column chromatography.

Acknowledgments

Support has been provided by the National Institutes of Health (National Institute of General Medical Sciences: R01-GM57034) and the Swedish Research Council (fellowship for M.K.: Dnr 350-2012-6645). We thank Dr. Nathan D. Schley, Dr. Michael K. Takase (X-ray Crystallography Facility; a Bruker KAPPA APEX II X-ray diffractometer was purchased via NSF CRIF:MU Award CHE-0639094), Dr. David G. VanderVelde (NMR Facility), Dr. Scott C. Virgil (Center for Catalysis and Chemical Synthesis, supported by the Gordon and Betty Moore Foundation), and Dr. Daniel T. Ziegler for assistance.

Supporting Information Available

Experimental procedures and compound characterization data. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.5b05528.

The authors declare no competing financial interest.

Supplementary Material

ja5b05528_si_001.cif (100.2KB, cif)
ja5b05528_si_002.pdf (4.8MB, pdf)

References

  1. For recent reviews and leading references, see:; a Xiao Y.; Sun Z.; Guo H.; Kwon O. Beilstein J. Org. Chem. 2014, 10, 2089–2121 10.3762/bjoc.10.218. [DOI] [PMC free article] [PubMed] [Google Scholar]; b Fan Y. C.; Kwon O. In Science of Synthesis: Asymmetric Organocatalysis; List B., Maruoka K., Eds.; Georg Thieme Verlag: Stuttgart, 2012; Vol. 1, pp 723–782. [Google Scholar]
  2. For initial studies of γ-addition processes that generated achiral or racemic products, see:; a Trost B. M.; Li C.-J. J. Am. Chem. Soc. 1994, 116, 3167–3168 10.1021/ja00086a074. [DOI] [Google Scholar]; b Trost B. M.; Li C.-J. J. Am. Chem. Soc. 1994, 116, 10819–10820 10.1021/ja00102a071. [DOI] [Google Scholar]
  3. For examples of processes that generate a γ stereocenter with good enantiomeric excess, see:; a Chung Y. K.; Fu G. C. Angew. Chem., Int. Ed. 2009, 48, 2225–2227 10.1002/anie.200805377. [DOI] [PMC free article] [PubMed] [Google Scholar]; b Smith S. W.; Fu G. C. J. Am. Chem. Soc. 2009, 131, 14231–14233 10.1021/ja9061823. [DOI] [PMC free article] [PubMed] [Google Scholar]; c Sinisi R.; Sun J.; Fu G. C. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 20652–20654 10.1073/pnas.1003597107. [DOI] [PMC free article] [PubMed] [Google Scholar]; d Sun J.; Fu G. C. J. Am. Chem. Soc. 2010, 132, 4568–4569 10.1021/ja101251d. [DOI] [PMC free article] [PubMed] [Google Scholar]; e Fujiwara Y.; Sun J.; Fu G. C. Chem. Sci. 2011, 2, 2196–2198 10.1039/c1sc00414j. [DOI] [PMC free article] [PubMed] [Google Scholar]; f Lundgren R. J.; Wilsily A.; Marion N.; Ma C.; Chung Y. K.; Fu G. C. Angew. Chem., Int. Ed. 2013, 52, 2525–2528 10.1002/anie.201208957. [DOI] [PMC free article] [PubMed] [Google Scholar]; g Fang Y.-Q.; Tadross P. M.; Jacobsen E. N. J. Am. Chem. Soc. 2014, 136, 17966–17968 10.1021/ja5117638. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. For examples of processes that generate a δ stereocenter with good enantiomeric excess, see:; a Chen Z.; Zhu G.; Jiang Q.; Xiao D.; Cao P.; Zhang X. J. Org. Chem. 1998, 63, 5631–5635 10.1021/jo9721756. [DOI] [Google Scholar]; b Wang T.; Yao W.; Zhong F.; Pang G. H.; Lu Y. Angew. Chem., Int. Ed. 2014, 53, 2964–2968 10.1002/anie.201307757. [DOI] [PubMed] [Google Scholar]; c Chen J.; Cai Y.; Zhao G. Adv. Synth. Catal. 2014, 356, 359–363 10.1002/adsc.201300695. [DOI] [Google Scholar]
  5. For leading references, see: Catalytic Asymmetric Conjugate Reactions; Cordova A., Ed.; Wiley–VCH: Weinheim, 2010. [Google Scholar]
  6. a Trost B. M.; Kazmaier U. J. Am. Chem. Soc. 1992, 114, 7933–7935 10.1021/ja00046a062. [DOI] [Google Scholar]; b For a review, see: Kwong C. K.-W.; Fu M. Y.; Lam C. S.-L.; Toy P. H. Synthesis 2008, 2008, 2307–2317 10.1055/s-2008-1067173. [DOI] [Google Scholar]
  7. For a brief review with leading references on the synthesis and the significance of α,α-disubstituted α-amino acids, see:Metz A. E.; Kozlowski M. C. J. Org. Chem. 2015, 80, 1–7 10.1021/jo502408z. [DOI] [PubMed] [Google Scholar]
  8. For an example of a bioactive 1,3-oxazol-5(4H)-one, see:Pinto I. L.; West A.; Debouck C. M.; DiLella A. G.; Gorniak J. G.; O'Donnell K. C.; O'Shannessy D. J.; Patel A.; Jarvest R. L. Bioorg. Med. Chem. Lett. 1996, 6, 2467–2472 10.1016/0960-894X(96)00456-8. [DOI] [Google Scholar]
  9. a This family of phosphepines was originally developed to serve as chiral ligands for metal-catalyzed enantioselective processes. For an overview, see: Gladiali S.; Alberico E.; Junge K.; Beller M. Chem. Soc. Rev. 2011, 40, 3744–3763 10.1039/c0cs00164c. [DOI] [PubMed] [Google Scholar]; b For the first application of such a phosphepine as a chiral nucleophilic catalyst, see: Wurz R. P.; Fu G. C. J. Am. Chem. Soc. 2005, 127, 12234–12235 10.1021/ja053277d. [DOI] [PubMed] [Google Scholar]
  10. a This spirophosphine was originally developed to serve as a chiral ligand for metal-catalyzed enantioselective processes. See: Zhu S.-F.; Yang Y.; Wang L.-X.; Liu B.; Zhou Q.-L. Org. Lett. 2005, 7, 2333–2335 10.1021/ol050556x. [DOI] [PubMed] [Google Scholar]; b For the first application of this spirophosphine as a chiral nucleophilic catalyst, see: Reference (3a)
  11. In a preliminary study, the use of TBME as the solvent led to a somewhat lower yield (∼82%) and enantioselectivity (∼88% ee).
  12. In preliminary studies, when the γ-substituent of the allenoate (R1) was Ph, Bn, or cyclopentylmethyl, unsatisfactory results were obtained.
  13. Under our standard conditions, an attempt to employ a 1,3-oxazol-5(4H)-one with R = i-Pr led to a low yield of the γ-addition product.
  14. For examples of natural products/bioactive compounds that include a 2-amino-3,5-dihydro-4H-imidazol-4-one subunit, see:; a Gadwood R. C.; Kamdar B. V.; Dubray L. A. C.; Wolfe M. L.; Smith M. P.; Watt W.; Mizsak S. A.; Groppi V. E. J. Med. Chem. 1993, 36, 1480–1487 10.1021/jm00062a022. [DOI] [PubMed] [Google Scholar]; b Tsukamoto S.; Kato H.; Hirota H.; Fusetani N. J. Nat. Prod. 1996, 59, 501–503 10.1021/np960113p. [DOI] [PubMed] [Google Scholar]; c Edrada R. A.; Stessman C. C.; Crews P. J. Nat. Prod. 2003, 66, 939–942 10.1021/np020503d. [DOI] [PubMed] [Google Scholar]
  15. We hypothesize that the phenol may serve as a proton shuttle in this process.
  16. For other phosphine-catalyzed enantioselective γ-additions in which kinetic resolution of the allene has been examined, see:; a No kinetic resolution: References (3b, 3d); b Kinetic resolution: References (3c, 3e)
  17. Estimated through computations: B3LYP/6-311+G(2d,2p); DFT-D3; CPCM; gas-phase G correction.
  18. Under the same conditions but in the absence of the 1,3-oxazol-5(4H)-one, the allenoate isomerizes to the 1,3-diene (Reference (6)) at a rate and with a selectivity factor (kinetic resolution of the allenoate) that are essentially identical to those for the γ-addition reaction, consistent with the two processes sharing the same turnover-limiting step (formation of A).

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ja5b05528_si_001.cif (100.2KB, cif)
ja5b05528_si_002.pdf (4.8MB, pdf)

Articles from Journal of the American Chemical Society are provided here courtesy of American Chemical Society

RESOURCES