Skip to main content
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2014 Jul 14;16(15):4036–4039. doi: 10.1021/ol501869s

Enantioselective Synthesis of Polycyclic Nitrogen Heterocycles by Rh-Catalyzed Alkene Hydroacylation: Constructing Six-Membered Rings in the Absence of Chelation Assistance

Xiang-Wei Du 1, Avipsa Ghosh 1, Levi M Stanley 1,*
PMCID: PMC4144754  PMID: 25020184

Abstract

graphic file with name ol-2014-01869s_0009.jpg

Catalytic, enantioselective hydroacylations of N-allylindole-2-carboxaldehydes and N-allylpyrrole-2-carboxaldehydes are reported. In contrast to many alkene hydroacylations that form six-membered rings, these annulative processes occur in the absence of ancillary functionality to stabilize the acylrhodium(III) hydride intermediate. The intramolecular hydroacylation reactions generate 7,8-dihydropyrido[1,2-a]indol-9(6H)ones and 6,7-dihydroindolizin-8(5H)-ones in moderate to high yields with excellent enantioselectivities.


The hydroacylation of alkenes in the presence of a transition metal catalyst has been extensively investigated as a direct route to ketones from simple starting materials.1 Despite the importance of ketones as synthetic building blocks and their potential as entry points to an array of chemical architectures, the hydroacylation of alkenes remains underdeveloped and underutilized relative to other metal-catalyzed hydrofunctionalizations of alkenes.2

Intramolecular hydroacylations to generate five-membered carbocycles are the most established class of alkene hydroacylations,3 and many enantioselective hydroacylations of substituted 4-pentenals and 2-vinylbenzaldehydes form chiral, nonracemic cyclopentanones and dihydroindenones.4 Recent strategies also enable the synthesis of six-, seven-, and eight-membered carbocycles and heterocycles through intramolecular alkene hydroacylation reactions.5 Despite these achievements, intramolecular alkene hydroacylations to generate nitrogen heterocycles remain rare,5a5c,6 and hydroacylation reactions to form rings of greater than five atoms are often driven by strain release5c,5e,5h or rely on heteroatom functionality contained at specific sites within the substrate molecules to stabilize acylrhodium(III) hydride intermediates and prevent catalyst decomposition.5b,5d,5g,6b

The potential to develop alkene hydroacylation as a platform for synthesis of medicinally important nitrogen heterocycles led us to study hydroacylations of indole- and pyrrole-2-carboxaldehydes containing N-vinyl and N-allyl substitution. We recently reported Rh-catalyzed hydroacylation of N-vinylindole-2-carboxaldehydes to form dihydropyrroloindolones in high yields with excellent enantioselectivities.7 Hydroacylations of N-allylindole-2-carboxaldehydes have proven more challenging because these processes involve the formation of a six-membered ring instead of a five-membered ring. During our studies, Douglas reported the first example of alkene hydroacylation involving N-allylindole-2-carboxaldehydes (eq 1).5a These hydroacylation reactions are enabled by transient generation of a 2-aminopicoline-based aldimine that stabilizes the acylrhodium hydride intermediate. However, this approach to chelation assistance requires a complex mixture of catalyst precursors and additives, and highly enantioselective hydroacylations involving 2-aminopicoline-based aldimines have not been reported.

graphic file with name ol-2014-01869s_0002.jpg 1

We now report catalytic, enantioselective hydroacylations of N-allylindole-2-carboxaldehyes and N-allylpyrrole-2-carboxaldehydes (eq 2). These hydroacylations occur to form dihydropyridoindolones and dihydroindolizinones in moderate to high yields and represent the first examples of highly enantioselective, transition metal-catalyzed hydroacylation to form six-membered rings in the absence of chelation assistance.

To test whether intramolecular hydroacylations would occur to generate six-membered rings in the absence of chelation assistance, we studied the reaction of 1-(2-methylallyl)-1H-indole-2-carboxaldehyde 1a catalyzed by complexes prepared in situ from [Rh(COD)Cl]2, (R)-BINAP L1, and a variety of silver salts (Table 1).8 We found the hydroacylation of 1a did not occur in the presence of rhodium catalysts with chloride or mesylate counterions (entries 1 and 2) and formed dihydropyridoindolone 2a in low yield when the catalyst contained a triflate counterion (entry 3).

Table 1. Identification of Catalysts for Hydroacylation of 1-(2-Methylallyl)-1H-indole-2-carboxaldehyde 1a.

graphic file with name ol-2014-01869s_0005.jpg

entry ligand AgX conv (%)a yield 2a (%)b ee (%)c
1 L1 5 0
2 L1 AgOMs 2 0
3 L1 AgOTf 18 17 (12) 94
4 L1 AgPF6 69 69 (63) 96
5 L1 AgBF4 99 86 (83) 95
6 L1 AgSbF6 82 82 (79) 96
7 L2 AgBF4 99 99 (94) 97
8 L3 AgBF4 99 99 (97) 87
9d L2 98 93 (90) 96
a

Conversion of 1a determined by 1H NMR spectroscopy.

b

Yield of 2a determined by 1H NMR spectroscopy. Isolated yield of 2a is shown in parentheses.

c

Determined by chiral HPLC analysis.

d

Reaction performed with 5 mol % [Rh(COD)2]BF4 as a catalyst precursor.

The intramolecular hydroacylation of 1a occurred in higher yields and formed 2a with higher enantioselectivities when the rhodium catalyst contained a weakly coordinating counterion.9 The reaction of 1a generated 2a in 63–83% yield with 95–96% enantiomeric excess in the presence of rhodium complexes with hexafluorophosphate, tetrafluoroborate, and hexafluoroantimonate counterions (entries 4–6). The identity of the counterion had minimal effect on the enantioselectivity of the hydroacylations. However, catalysts containing tetrafluoroborate and hexafluorantimonate counterions led to significantly higher yields of 2a.

To improve the yield and selectivity of our model reaction, we studied the impact of catalysts prepared from additional BINAP derivatives on the reaction of 1a. The rhodium(I) complexes of (R)-Tol-BINAP L2 and (R)-Xyl-BINAP L3 catalyze the hydroacylation of 1a to form 2a in higher yields (94% and 97% yield) than the Rh complex of the parent ligand L1 (compare entries 7 and 8 with entry 5). However, the hydroacylation of 1a occurred with the highest enantioselectivity when the reaction was conducted with the Rh complex of (R)-Tol-BINAP. The hydroacylation of 1a occurs with similar enantioselectivity and forms 2a in 90% yield when the reaction is performed with a catalyst generated from [Rh(COD)2]BF4 and (R)-Tol-BINAP (entry 9), suggesting the role of the Ag(I) salt is limited to anion exchange to generate the active catalyst. In all cases, the formation of six-membered ketone 2a was favored; the formation of a five-membered ketone product was not observed.10

The absolute configuration of 2a was determined after bromination of 2a with N-bromosuccinimide to generate 3 in nearly quantitative yield (Scheme 1). The absolute configuration of 3 was determined to be (S) by X-ray crystallographic analysis.

Scheme 1. Absolute Stereochemistry and Structure of 3.

Scheme 1

Table 2 summarizes the results of hydroacylations with 1-(2-methylallyl)-indole-2-carboxaldehydes containing substitution at the 3-, 4-, 5-, 6-, and 7-positions on the indole core. In general, hydroacylations of 1-(2-methylallyl)-indole-2-carboxaldehydes containing electron-donating substituents, electron-withdrawing substituents, and halogens at the 4-, 5-, 6-, and 7-positions occur with excellent enantioselectivity (entries 1–8). Hydroacylations of 4-MeO-, 5-MeO-, and 6-MeO-substituted 1bd formed 2bd in high yields (84–96%) with excellent enantioselectivities (97–99% ee, entries 1–3). The hydroacylation of 4,7-dimethoxy-substituted 1-(2-methylallyl)-indole-2-carboxaldehyde 1e occurred with high enantioselectivity, but the corresponding dihydropyridoindole 2e was isolated in 53% yield (entry 4).11 Hydroacylations of 1fi containing halogens or electron-withdrawing groups at the 5- and 6-positions occurred with excellent enantioselectivities (95–98% ee), and 2fi were isolated in 65–92% yields (entries 5–8). A 3-substituted 1-(2-methylallyl)-indole-2-carboxaldehyde 1j was also an excellent substrate for hydroacylation. The reaction of 1j formed 2j in 93% yield with 93% ee (entry 9).

Table 2. Rh-Catalyzed Enantioselective Hydroacylation of 1-(2-Methylallyl)-indole-2-carboxaldehydes 1bj.

graphic file with name ol-2014-01869s_0006.jpg

entry 1 R1 R2 2 yield 2 (%)a ee (%)b
1 1b H 4-MeO 2b 96 97
2 1c H 5-MeO 2c 95 99
3 1d H 6-MeO 2d 84 97
4 1e H 4,7-(MeO)2 2e 53 95
5 1f H 5-Cl 2f 92 98
6 1g H 6-Cl 2g 83 96
7 1h H 5-NO2 2h 89 96
8 1i H 6-CF3 2i 65 95
9 1j Et H 2j 93 93
a

Isolated yield of 2.

b

Determined by chiral HPLC analysis.

The results of intramolecular hydroacylations of N-allylindole-2-carboxaldehydes containing a range of 2-substituted allyl units, are shown in Table 3. Hydroacylations of 1kq containing alkyl, benzyl, aryl, and ester substituents at the central carbon of the allyl unit occurred with excellent enantioselectivities (95–97% ee), but these reactions formed dihydropyridoindoles 2kq in modest yields (55–75% NMR yields, 23–56% isolated yields). The relatively low isolated yields of 2kq result from a combination of competitive decarbonylation of 1kq and challenging product purifications from reaction mixtures containing unreacted 1kq and decarbonylation products. Yields of decarbonylation products ranged from 5% to 15%.

Table 3. Enantioselective Hydroacylation of N-Allylindole-2-carboxaldehydes 1kq.

graphic file with name ol-2014-01869s_0007.jpg

entry R (1) 2 conv (%)a yield 2 (%)b ee (%)c
1d Et (1k) 2k 80 53 (75) 97
2d n-hexyl (1l) 2l 73 42 (61) 96
3 CH2Ph (1m) 2m 82 45 (55) 97
4 Ph (1n) 2n 72 37 (64) 97
5 4-Me-C6H4 (1o) 2o 71 31 (63) 96
6 4-Cl-C6H4 (1p) 2p 83 56 (66) 95
7 CO2Et (1q) 2q 65 23 (56) 96
a

Conversion of 1 determined by 1H NMR spectroscopy.

b

Isolated yield of 2. NMR yield of 2 is listed in parentheses.

c

Determined by chiral HPLC analysis.

d

AgBF4 was used in place of AgSbF6.

The ability to form dihydropyridoindoles by enantioselective hydroacylations of N-allylindole-2-carboxaldehydes led us to investigate analogous hydroacylations of N-allylpyrrole-2-carboxaldehydes containing a range of substituted allyl units (Table 4). The hydroacylations of N-allylpyrrole-2-carboxaldehydes 4ac containing alkyl substitution at the central carbon of the allyl unit formed dihydroindolizinones 5ac in modest-to-good yields (51–79%) with 94–97% enantiomeric excess (entries 1–3). The hydroacylation of 4d (R = CH2Ph) did not occur to high conversion in the presence of 5 mol % catalyst. However, the reaction of 4d formed 5d in 52% yield with 92% enantiomeric excess when the reaction was run in the presence of 10 mol % rhodium catalyst (entry 4).

Table 4. Enantioselective Hydroacylation of N-Allylpyrrole-2-carboxaldehydes 4ah.

graphic file with name ol-2014-01869s_0008.jpg

entry R (4) 5 yield 5 (%)a ee (%)b
1c Me (4a) 5a 79 97
2 Et (4b) 5b 70 96
3 n-hexyl (4c) 5c 51 94
4d CH2Ph (4d) 5d 52 92
5 Ph (4e) 5e 96 97
6 4-Me-C6H4 (4f) 5f 86 97
7 4-Cl-C6H4 (4g) 5g 85 95
8d CO2Et (4h) 5h 98 98
a

Isolated yield of 5.

b

Determined by chiral HPLC analysis.

c

Reaction run at 100 °C using AgBF4 in place of AgSbF6.

d

Reaction run in the presence of 10 mol % catalyst.

In general, the N-allylpyrrole-2-carboxaldehydes are less prone to decarbonylation than the related N-allylindole-2-carboxaldehydes. These results are particularly evident for pyrroles 4eg with aryl substitution at the central carbon of the allyl unit. The hydroacylations of 4eg (R = Ph, 4-Me-C6H4, and 4-Cl-C6H4) occurred with excellent enantioselectivities (95–97% ee), and the heterocyclic ketone products 5eg were isolated in 85–96% yields (entries 5–7). These results contrast hydroacylations of N-allylindole-2-carboxaldehydes with aryl substitution at the central carbon of the allyl unit (compare entries 4–6 in Table 3 with entries 5–7 in Table 4). Reactions of indoles 1np require 10 mol % catalyst to reach high coversion due to competing decarbonylation, while reactions of the analogous pyrroles 4eg require only 5 mol % catalyst to reach full conversion and decarbonlyation side products are not observed.12 The hydroacylation of pyrrole 4h containing an electron-withdrawing group at the central carbon of the allyl unit (R = CO2Et) generated 5h in 98% yield with 98% ee (entry 8).

The synthetic utility of our enantioselective hydroacylation reactions has been demonstrated through a rapid asymmetric synthesis of the nonsteroidal aromatase inhibitor MR 20492 (Scheme 2).13 Enantioselective hydroacylation of N-allylindole-2-carboxaldehyde 4g formed dihydroindolizinone 5g in 85% yield with 95% ee (Table 4, entry 7). Aldol condensation of 5g with pyridine-4-carboxaldehyde generated (S,Z)-MR 20492 in 57% yield.

Scheme 2. Enantioselective Synthesis of (S,Z)-MR 20492.

Scheme 2

In summary, we have developed catalytic, enantioselective hydroacylations of N-allylindole- and N-allylpyrrole-2-carboxaldehydes. These hydroacylation reactions are catalyzed by a readily available Rh complex, occur in the absence of chelation assistance, and form six-membered heterocyclic ketones in moderate-to-excellent yields from a variety of indole and pyrrole substrates. The utility of our method is demonstrated in a straightforward asymmetric synthesis of the nonsteroidal aromatase inhibitor MR 20492. Studies to expand the scope of transition-metal-catalyzed hydroacylation reactions that occur in the absence of chelation assistance and to extend these methods to additional carbocyclic and heterocyclic scaffolds are ongoing in our laboratory.

Acknowledgments

We thank ISU, the ISU Institute for Physical Research, and the NSF (CAREER 1353819) for financial support of this work. L.S. thanks the NIH for a Pathways to Independence Award (GM95697). We thank Dr. Arkady Ellern (ISU) for X-ray diffraction data collection and structure determination.

Supporting Information Available

Experimental procedures, characterization for all new compounds, and crystallographic data for compound 3. This material is available free of charge via the Internet at http://pubs.acs.org.

The authors declare no competing financial interest.

Funding Statement

National Institutes of Health, United States

Supplementary Material

ol501869s_si_001.pdf (37.9MB, pdf)
ol501869s_si_002.cif (16.7KB, cif)

References

  1. a Willis M. C. Chem. Rev. 2010, 110, 725. [DOI] [PubMed] [Google Scholar]; b Leung J. C.; Krische M. J. Chem. Sci. 2012, 3, 2202. [Google Scholar]; c Park Y. J.; Park J.-W.; Jun C.-H. Acc. Chem. Res. 2008, 41, 222. [DOI] [PubMed] [Google Scholar]; d Fu G. C. In Modern Rhodium-Catalyzed Organic Reactions; Evans P. A., Ed.; Wiley-VCH: New York, 2005; p 79. [Google Scholar]
  2. Hartwig J. F.Organotransition Metal Chemistry; University Science Books: Sausalito, CA, 2010. [Google Scholar]
  3. For selected examples, see:; a Okamoto R.; Tanaka K. Org. Lett. 2013, 15, 2122. [DOI] [PubMed] [Google Scholar]; b Vautravers N. R.; Regent D. D.; Breit B. Chem. Commun. 2011, 47, 6635. [DOI] [PubMed] [Google Scholar]; c Larock R. C.; Oertle K.; Potter G. F. J. Am. Chem. Soc. 1980, 102, 190. [Google Scholar]; d Lochow C. F.; Miller R. G. J. Am. Chem. Soc. 1976, 98, 1281. [Google Scholar]
  4. For selected and recent examples, see:; a Hoffman T. J.; Carreira E. M. Angew. Chem., Int. Ed. 2011, 50, 10670. [DOI] [PubMed] [Google Scholar]; b Kunda K.; McCullagh J. V.; Morehead A. T. Jr. J. Am. Chem. Soc. 2005, 127, 16042. [DOI] [PubMed] [Google Scholar]; c Marce P.; Diaz Y.; Matheu M. I.; Castillon S. Org. Lett. 2008, 10, 4735. [DOI] [PubMed] [Google Scholar]; d Barnhart R. W.; Wang X.; Noheda P.; Bergens S. H.; Whelan J.; Bosnich B. J. Am. Chem. Soc. 1994, 116, 1821. [Google Scholar]; e Wu X.-M.; Funakoshi K.; Sakai K. Tetrahedron Lett. 1992, 33, 6331. [Google Scholar]
  5. a Beletskiy E. V.; Sudheer C.; Douglas C. J. J. Org. Chem. 2012, 77, 5884. [DOI] [PubMed] [Google Scholar]; b Bendorf H. D.; Ruhl K. E.; Shurer A. J.; Shaffer J. B.; Duffin T. O.; LaBarte T. L.; Maddock M. L.; Wheeler O. W. Tetrahedron Lett. 2012, 53, 1275. [Google Scholar]; c Crépin D.; Dawick J.; Aïssa C. Angew. Chem., Int. Ed. 2010, 49, 620. [DOI] [PubMed] [Google Scholar]; d Coulter M. M.; Dornan P. K.; Dong V. M. J. Am. Chem. Soc. 2009, 131, 6932. [DOI] [PubMed] [Google Scholar]; e Aïssa C.; Fürstner A. J. Am. Chem. Soc. 2007, 129, 14836. [DOI] [PubMed] [Google Scholar]; f Sato Y.; Oonishi Y.; Mori M. Angew. Chem., Int. Ed. 2002, 41, 1218. [DOI] [PubMed] [Google Scholar]; g Bendorf H. D.; Colella C. M.; Dixon E. C.; Marchetti M.; Matukonis A. N.; Musselman J. D.; Tiley T. A. Tetrahedron Lett. 2002, 43, 7031. [Google Scholar]; h Aloise A. D.; Layton M. E.; Shair M. D. J. Am. Chem. Soc. 2000, 122, 12610. [Google Scholar]; i Gable K.; Benz G. A. Tetrahedron Lett. 1991, 32, 3473. [Google Scholar]; j Hoshimoto Y.; Hayashi Y.; Suzuki H.; Ohashi M.; Ogoshi S. Angew. Chem., Int. Ed. 2012, 51, 10812. [DOI] [PubMed] [Google Scholar]
  6. a Castaing M.; Wason S. L.; Estepa B.; Hooper J. F.; Willis M. C. Angew. Chem., Int. Ed. 2013, 52, 13280. [DOI] [PMC free article] [PubMed] [Google Scholar]; b Arnold J. S.; Mwenda E. T.; Nguyen H. M. Angew. Chem., Int. Ed. 2014, 53, 3688. [DOI] [PubMed] [Google Scholar]
  7. Ghosh A.; Stanley L. Chem. Commun. 2014, 50, 2765. [DOI] [PubMed] [Google Scholar]
  8. Although substrate 1 lacks a heteroatom capable of coordinating the rhodium center, it is possible that a second substrate molecule stabilizes the acylrhodium(III) hydride intermediate as proposed by Fairlie and Bosnich:Fairlie D. P.; Bosnich B. Organometallics 1988, 7, 946. [Google Scholar]
  9. Phan D. H.; Kim B.; Dong V. M. J. Am. Chem. Soc. 2009, 131, 15608. [DOI] [PubMed] [Google Scholar]
  10. The hydroacylation of N-allylindole-2-carboxaldehyde, which lacks substitution at the 2-position of the allyl unit, occurs exclusively with anti-Markovnikov selectivity to form the achiral dihydropyridoindolone product.
  11. 1,4-Dimethoxy-7-methylpyrido[1,2-a]indole was isolated as a byproduct in 30% yield. For related cyclizations of o-allylbenzaldehydes to form naphthalenes, see:agdale A. R.; Park J. H.; Youn S. W. J. Org. Chem. 2011, 76, 7204. [DOI] [PubMed] [Google Scholar]
  12. Reactions of substrates 1n and 4e occur in the presence of 5 mol % catalyst to approximately 50% conversion in 1.5 h suggesting the reactivity of indole and pyrrole substrates are similar in the absence of catalyst deactivation.
  13. a Dallemagne P.; Sonnet P.; Enguehard C.; Rault S. J. Heterocycl. Chem. 1996, 33, 1689. [Google Scholar]; b Sonnet P.; Enguehard J. G.; Dallemagne P.; Rault R. B. Bioorg. Med. Chem. Lett. 1998, 8, 1041. [DOI] [PubMed] [Google Scholar]; c Auvray P.; Sourdaine P.; Moslemi S.; Séralini G. E.; Sonnet P.; Enguehard C.; Guillon J.; Dallemagne P.; Bureau R.; Rault S. J. Steroid. Biochem. Mol. Biol. 1999, 70, 59. [DOI] [PubMed] [Google Scholar]; d Sonnet P.; Dallemagne P.; Guillon J.; Enguehard C.; Stiebing S.; Tanguy J.; Bureau R.; Rault S.; Auvray P.; Moslemi S.; Sourdaine P.; Séralini G. Bioorg. Med. Chem. 2000, 8, 945. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

ol501869s_si_001.pdf (37.9MB, pdf)
ol501869s_si_002.cif (16.7KB, cif)

Articles from Organic Letters are provided here courtesy of American Chemical Society

RESOURCES