A copper-catalyzed domino reaction to construct functionalized indolizinones

Abstract

An efficient copper-catalyzed domino reaction to 3,8a-disubstituted indolizinones has been firstly developed, and the protocol uses pyridine ketones and terminal alkynes as the starting materials, overriding the isolation of propargylic alcohols.

---

The development of efficient and versatile strategies for the synthesis of heterocycles has always been an active research topic in organic chemistry.¹ In this regard, transformations that employ readily available substrates and go through a “one-pot” reaction to provide access to multiply functionalized heterocycles are highly desirable. Indolizinones are N-heterocycles that have only recently appeared in the chemical literature,^2–5 and are useful precursors to indolizidine natural products, pharmaceutical agents and new materials.^2b Methods for the synthesis of indolizinones have been reported by other groups with salts or complexes of platinum,² copper,³ silver,⁶ gold⁶ and iodine.⁴ This transformation is proposed to proceed by a cyclization/1,2-migration process of propargylic pyridines. Indolizinones are also accessible from propargylic alcohols using a greener protocol. Kim group^5c and Sarpong group^2b discovered that heating tertiary alcohols in ethanol afforded indolizinone products in excellent yields. The preparation of propargylic alcohols, utilizing basic lithium reagents under special conditions, was necessary in all of the previous reports. Other methods for synthesizing propargylic alcohols would employ heavy metal catalysts such as zinc or indium in the presence of bulky ligands,⁷ or tetrabutylammonium fluoride (TBAF) catalyst⁸ in dry solvents. Here, we firstly reported copper-catalyzed domino reaction to construct 3,8a-disubstituted indolizinones starting from pyridine ketones and terminal alkynes, overriding the preparation and isolation of propargylic alcohols (Scheme 1).

Scheme 1 Methods for the synthesis of indolizinone derivatives.

Our initial studies focused on identifying the optimal conditions (Table 1). Di-2-pyridyl ketone 1a and phenylacetylene 2a could be smoothly converted to indolizinone 3aa in 57% yield with 10 mol% of CuBr₂ as the catalyst, 2 equiv. of NEt₃ as the base, and dioxane as the solvent at 110 °C during 24 hours (entry 1). Without the participation of NEt₃, the yield decreased to 38% (entry 2), so was catalyzed by CuI (comparing to entry 5), indicating that the base could accelerate the reaction rate. While without the addition of CuBr₂, the reaction couldn't happen (entry 3), showing that copper-salt as the catalyst was necessary. We screened other catalysts, and found Cu(OAc)₂ was the best, which led to 95% yield after 4 hours (comparing entries 1 and 5–8). To our delight, the reaction conditions were easily handled. For example, the reaction operated in freshly dried dioxane under 1 atm nitrogen atmosphere also provided 94% yields after 4 hours (entry 9). Besides of dioxane, acetonitrile or toluene as the solvent also provided excellent yields (entries 10 and 11), albeit the inexpensive “green” water led to a low yield of 25% (entry 12).

Table 1 Optimization of conditions for copper-catalyzed domino reaction of di-2-pyridyl ketone 1a with phenylacetylene 2a to form indolizinone 3aa^a

Entry	Cat.	Base	Solvent	Time [h]	Temp. [°C]	Yield^b [%]
a Reaction conditions: 1a (0.3 mmol), 2a (0.6 mmol), catalyst (0.03 mmol), base (0.6 mmol), solvent (2 ml) in sealed tube.b Isolated yield.c Under highly pure nitrogen atmosphere and in freshly dried dioxane.
1	CuBr2	NEt3	Dioxane	24	110	57
2	CuBr2	—	Dioxane	24	110	38
3	—	NEt3	Dioxane	24	110	nr
4	CuI	—	Dioxane	24	110	42
5	CuI	NEt3	Dioxane	24	110	69
6	CuBr	NEt3	Dioxane	24	110	80
7	Cu(NO3)2	NEt3	Dioxane	24	110	65
8	Cu(OAc)2	NEt3	Dioxane	4	110	95
9^c	Cu(OAc)2	NEt3	Dioxane	4	110	94
10	Cu(OAc)2	NEt3	CH3CN	4	110	89
11	Cu(OAc)2	NEt3	Toluene	4	110	86
12	Cu(OAc)2	NEt3	H2O	4	110	25

---

With optimized conditions in hand, we first examined the scope of terminal alkynes with 1a (Table 2). Excellent yields of the desired products were obtained with substituted phenylacetylene bearing electron-donating or electron-withdrawing groups. Strangely, para-substituted groups (except of methyl group shown as in entry 2) at the benzene ring made the reaction proceed slowly (entry 3–5, and 8). For example, the para-chloride substrate provided the corresponding indolizinone 3ae in 93% yields after 24 hours (entry 5), but meta-chloride (entry 6) and ortho-chloride (entry 7) substrates gave indolizinones 3af and 3ag in 93% and 96% yields respectively, just after 4 hours. The structure of indolizinone 3af was confirmed by the X-ray single crystal method (see Fig. S2†). Pyridylacetylene with N atom at different positions (entries 9–11), and 2-naphthylacetylene (entry 12), and 4-phenyl phenylacetylene (entry 13) led to good yields from 61% to 100%, but in need of longer times and/or increasing temperature to 130 °C. Aliphatic alkynes such as cyclohexylacetylene (entry 14), 4-phenyl-1-butyne (entry 15), and 1-octyne (entry 16) all afforded products with excellent yields in 1 hour.

Table 2 Substrate scope for terminal alkynes catalyzed by Cu(OAc)₂^a

Entry	2	3	Temp.	Time	Yield^b [%]
a Reaction conditions: 1a (0.3 mmol), 2 (0.6 mmol), Cu(OAc)2 (0.03 mmol), NEt3 (0.6 mmol), dioxane (2 ml) in a sealed tube.b Isolated yield.
1		3aa	110	4	95
2		3ab	110	4	100
3		3ac	110	24	85
4		3ad	110	24	78
5		3ae	110	24	93
6		3af	110	4	93
7		3ag	110	4	96
8		3ah	110	24	91
9		3ai	130	24	64
10		3aj	110	72	79
11		3ak	130	18	61
12		3al	130	5	100
13		3am	130	1	92
14		3an	110	1	90
15		3ao	110	1	94
16		3ap	110	1	80

---

Then we investigated the scope of pyridine ketones, another six different substrates were examined (Table 3), bis(5-methyl-2-pyridyl)ketone 1b,⁹ trifluoro-(2-pyridyl) ketone 1c and 2-pyridyl-4-pyridyl ketone 1d all led to indolizinones with yields from 33% to 82%, and bis(5-methyl-2-pyridyl) ketone 1b showed the similar reactivity as 1a. Terminal alkynes in varied types could react with 1b or 1c, providing the corresponding indolizinones 3ba–3cn with yields from 53% to 82%. However, the reaction starting from the substrates phenyl-(2-pyridyl) ketone 1e, methyl-(2-pyridyl) ketone 1f, and pyridine-2-carbaldehyde 1g couldn't happen under the established conditions. These results indicated that the structure of pyridine ketones played a very important role in deciding whether the reaction could happen. We deduced that two 2-pyridyl N atoms in the substrates 1a or 1b chelating to the copper ion, especially comparing to the substrate 1d, thus facilitates nucleophilic attack of the terminal alkyne, while CF₃⁻ in the substrate 1c as a strong electron-withdrawing group played the similar role.

Table 3 Substrate scope for pyridine ketones catalyzed by Cu(OAc)₂^a

Entry	1	2	3	Temp.	Time	Yield^b [%]
a Reaction conditions: 1 (0.3 mmol), 2 (0.6 mmol), Cu(OAc)2 (0.03 mmol), NEt3 (0.6 mmol), dioxane (2 ml) in a sealed tube.b Isolated yield.
1		2a	3ba	110	1	77
2	1b	2c	3bc	110	24	68
3	1b	2l	3bl	110	24	82
4	1b	2n	3bn	110	1	80
5		2a	3ca	110	20	70
6	1c	2l	3cl	110	36	57
7	1c	2n	3cn	110	96	53
8		2a	3da	110	63	33
9		2a	—	110	24	—
10		2a	—	110	24	—
11		2a	—	110	24	—

---

In order to confirm the chelation between 1a and Cu(OAc)₂, we obtained the gem–diol complex in 36% yield by the slow volatilization at room temperature (note: all the reagents were used without any purification after purchase and the exclusion of water was not required), from the solution of 1a and Cu(OAc)₂ in the ratio of 1 : 1 in dioxane. In the solid state, the gem–diols was formed through Cu²⁺-promoted in-situ hydration, and each OH group O atom chelated to Cu²⁺ cooperating with two pyridine N atoms, and finally provided an mononuclear inclined octahedron with the ligand and metal salts in the ratio of 2 : 1 (Fig. 1). The gem–diol form (Py)₂C(OH)₂ from pyridine ketone 1a was reported to be stabilized at the presence of BF₄⁻,¹⁰ ClO₄⁻,¹¹ and here CH₃COO⁻, indicating that the hydrolysis was independent on the relative basicity of a counter-ion.¹² We isolated the crystal sample, and it was further converted to indolizinone 3aa qualitatively, when treated with phenylacetylene 2a. Besides, previous reports have shown that di-2-pyridyl ketone could coordinate with metal ions including Cu²⁺, providing a dinuclear complex in dry solvents.¹³ The specific coordination chemistry of di-2-pyridyl ketone to Cu²⁺ lent itself to form the important intermediate propargylic alcohols bearing pyridine rings, which might be converted further into indolizinone.

Fig. 1 ORTEP representation of the cation of gem–diol complex between 1a and Cu(OAc)₂.

On the basis of these preliminary results, we speculated that propargylic pyridine was a key intermediate. We isolated the propargylic pyridine 4aa in 31% yield after 0.5 hour during the reaction process (Scheme 2A), under the established conditions, and it would be converted into 3aa further quantitatively (Scheme 2C). The formation of propargylic pyridine decided whether the final conversion to indolizinone could succeed, and it was dependent on the structure of pyridine ketone. To explore the importance of the propargylic alcohols during the reaction process, we then prepared propargylic pyridine 4ea starting from the phenyl-(2-pyridyl) ketone 1e by the participation of n-BuLi (Scheme 2B). Under the established conditions the new propargylic pyridine 4ea could be converted into 3ea efficiently, with yields of 88%.

Scheme 2 (A) Isolation of propargylic pyridine 4aa. (B) Preparation of propargylic pyridine 4ea. (C) Copper-catalyzed transformations of 4aa and 4ea under the established conditions.

As shown in Scheme 3, we proposed a reasonable mechanism. Firstly, propargylic alcohol bearing pyridine rings I is formed, through nucleophilic attack of a alkynylcopper(I) species,^15,16 with the assistance of copper ion and base. Coordination of the carbonyl oxygen to a metal ion (direct polarization) and coordination of the pyridyl group (induced polarization) can promote such polarization of the carbonyl C=O bond.^10,14 Then, subsequent cyclization/1,2-migration processes of propargylic alcohol in the presence of Cu(I)³ or Cu(II)¹⁷ happen to afford the indolizinone 3 finally.

Scheme 3 Possible mechanism for Cu(OAc)₂-catalyzed domino synthesis of 3,8a-disustituted indolizinones.

Conclusion

To the best of our knowledge, the work reported herein represents the first example of copper-catalyzed domino reaction to construct indolizinones starting from pyridine ketones and terminal alkynes in one-pot. Pyridine ketones containing two 2-pyridyl groups chelating to the copper ion or a strong electon-withdrawing CF₃⁻ group promoted polarization of the carbonyl C=O bond thus facilitated the nucleophilic attack of the terminal alkyne, and provided the corresponding indolizinones after cyclization and 1,2-migration. The specific coordination chemistry of di-2-pyridyl ketone to metal ions endows its carbonyl group with increased electrophilicity, which will lead to a growing number of interests in the future.

Acknowledgements

The authors wish to thank the National Natural Science Foudation of China (Grant No. 21302188, 91027033) and Chinese Academy of Sciences (Grant No. KJCX2-EW-N06-01 and XDA09030307) for financial support.

References

1. For a recent general review, see: J. A. Joule and K. Mills, Heterocyclic Chemistry, Blackwell Science Ltd., Cambridge, MA, 4th edn, 2000.
2. (a) C. R. Smith, E. M. Bunnelle, A. J. Rhodes and R. Sarpong, Org. Lett., 2007, 9, 1169; (b) A. R. H. Narayan and R. Sarpong, Org. Biomol. Chem., 2012, 10, 70; (c) S. T. Heller, T. Kiho, A. R. H. Narayan and R. Sarpong, Angew. Chem., Int. Ed., 2013, 52, 11129; (d) E. M. Bunnelle, C. R. Smith, S. K. Lee, S. W. Singaram, A. J. Rhodes and R. Sarpong, Tetrahedron, 2008, 64, 7008.
3. B. Yan, Y. B. Zhou, H. Zhang, J. J. Chen and Y. H. Liu, J. Org. Chem., 2007, 72, 7783–7786.
4. J. Choi, G. H. Lee and I. Kim, Synlett, 2008, 1243–1249.
5. (a) I. Kim and K. Kim, Org. Lett., 2010, 12, 2500; (b) H. Cho and I. Kim, Tetrahedron, 2012, 68, 5464; (c) I. Kim, J. Choi, S. Lee and G. H. Lee, Synlett, 2008, 15, 2334.
6. I. V. Seregin, A. W. Schammel and V. Gevorgyan, Org. Lett., 2007, 9, 3433.
7. Selected references: (a) P. G. Cozzi, Angew. Chem., Int. Ed., 2003, 42, 2895; (b) N. K. Anad and E. M. Carreira, J. Am. Chem. Soc., 2001, 123, 9687; (c) X. Li, G. Lu, W. H. Kwok and A. S. Chan, J. Am. Chem. Soc., 2002, 124, 12636; (d) C. Wolf and S. Liu, J. Am. Chem. Soc., 2006, 128, 10996; (e) D. E. Frantz, R. Fassler, C. S. Tomooka and E. M. Carriera, Acc. Chem. Res., 2000, 33, 373; (f) Z. Xu, R. Wang, J. Xu, C. Da, W. Yan and C. Chen, Angew. Chem., Int. Ed., 2003, 42, 5747.
8. V. R. Chintareddy, K. Wadhwa and J. G. Verkade, J. Org. Chem., 2011, 76, 4482.
9. J. P. Collman, M. Zhong, Z. Wang and M. Rapta, Org. Lett., 1999, 13, 2122.
10. C. Q. Wan, H. J. Yan and T. C. W. Mak, Inorg. Chim. Acta, 2015, 432, 13.
11. V. Tangoulis, C. P. Raptopoulou, A. Terzis, S. Paschalidou, P. Perlepes and E. G. Bakalbassis, Inorg. Chem., 1997, 36, 3996.
12. (a) C. Q. Wan and T. C. W. Mak, Inorg. Chim. Acta, 2008, 361, 1496; (b) C. Q. Wan, G. S. Li, X. D. Chen and T. C. W. Mak, Cryst. Growth Des., 2008, 8, 3897.
13. (a) R. A. Mariezcurrena, A. W. Mombrú, L. Suescun, E. Kremer and R. González, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 1999, 55, 1989; (b) A. Basu, A. R. Saple and N. Y. Sapre, J. Chem. Soc., Dalton Trans., 1987, 1797; (c) G. S. Papaefstathiou and S. P. Perlepes, Comments Inorg. Chem., 2002, 23, 249.
14. E. C. Constable, Metals and Ligand Reactivity, VCH, Weinheim, Germany, 1996, p. 46.
15. Copper(II) salts such as Cu(OAc)₂ readily undergo reduction to copper(I) in the presence of a terminal alkyne, and the chelation of di-2-pyridyl ketone to Cu(II) may facilitate the generation of Cu(I). The research about the conversion from Cu(II) to Cu(I) in the presence of terminal alkynes with pyridyl substrates, please see: (a) W. S. Brotherton, H. A. Michaels, J. T. Simmons, R. J. Clark, N. S. Dalal and L. Zhu, Org. Lett., 2009, 11, 4954; (b) G. C. Kuang, P. M. Guha, W. S. Brotherton, J. T. Simmons, L. A. Stankee, B. T. Nguyen, R. J. Clark and L. Zhu, J. Am. Chem. Soc., 2011, 133, 13984; (c) Z. Yuan, G.-C. Kuang, R. J. Clark and L. Zhu, Org. Lett., 2012, 14, 2590.
16. We detected the 1,4-diphenylbutadiyne as one by-product by GC-MS to testify indirectly the generation of the Cu(I) species, after the completion of the reaction between 1a and 2a under the optimized conditions (see Fig. S1†). The possible generation process of the Cu(I) species was shown as in below, according to ref. 16a:
    .
17. As shown in Scheme 2c, Cu(II) salts could catalyze the cyclization/1,2-migration of propargylic alcohol to indolizinones.

---

Footnote

† Electronic supplementary information (ESI) available: Synthetic procedures of indolizinones, and copies of NMR spectra of all new compounds. CCDC 1428333 and 1428334. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5ra20119e

---