Synthesis of oxindoles via reductive CO₂ fixation

Abstract

The synthesis of 3-aryl-3-hydroxy-2-oxindoles, which are a structural motif found in various natural products and pharmaceutically active compounds, was conducted via reductive coupling of (2-aminophenyl)(aryl)methanone derivatives and CO₂ as a key step. The conditions employing Mg with chlorotrimethylsilane in DMA are the best for the reductive coupling, where the aryl halide moiety is intact. This reaction proceeds well without the protection of the amino group. The reductive coupling and acid-catalyzed lactam formation can be performed in a one-pot reaction to give the oxindoles.

---

Introduction

Carbon dioxide (CO₂) is considered to be one of the causative agents for global warming. Therefore, the effective utilization of CO₂ collected from plant emissions is important in view of sustainable chemistry. Utilization of CO₂ as a C1 synthon for organic synthesis has been developed in this context.¹

Oxindoles are a structural motif found in various natural products and medicinally relevant molecules.² In this study, the synthesis of 3-aryl-3-hydroxy-2-oxindoles 1 is focused on, as the skeleton is included in several pharmaceutical compounds such as SM-130686 ³ and ECi8.⁴Scheme 1a shows the representative examples of the previously reported retrosynthesis for 1. Most commonly, oxindoles 1 are synthesized via 1,2-addition of an aryl anion equivalent to an isatin derivative.⁵ Intramolecular arylation to a ketone is also reported.⁶ Another approach involves aromatic nucleophilic substitution with an enolate of mandelic acid derivatives as a key step.⁷ In this context, our retrosynthesis was designed based on reductive CO₂ fixation with (2-aminophenyl)(aryl)methanone derivatives 2, followed by lactam formation (Scheme 1b). So far, an enormous number of synthetic reports for 1 have been published (Scheme 1b). However, to the best of our knowledge, such an approach via CO₂ fixation has not been reported.

Scheme 1 (a) Representative examples of the previously reported retrosynthesis for 3-aryl-3-hydroxy-2-oxindoles 1. (b) Our retrosynthesis for 1.

Reductive CO₂ fixation of diaryl ketones has been studied.^8–11 Such a reaction can be roughly classified into four reduction patterns, reduction by (1) electrodes,⁸ (2) alkali metals,⁹ (3) low-valent transition or rare-earth metals¹⁰ and (4) light.¹¹ Our group also has developed the reductive coupling of aldehydes, especially utilizing early-transition metals such as vanadium and titanium compounds as a catalyst in the presence of chlorotrimethylsilane.¹² Here, we report the synthesis of 3-aryl-3-hydroxy-2-oxindoles 1via reductive coupling of CO₂ and ketoamines 2 as a key step.

Results and discussion

Our investigation commences with the screening of the combination of an early-transition metal catalyst and Zn or Mg as a terminal reductant. First, Zn was used as a terminal reductant (Table 1). Ketoamine 2a reacted with CO₂ (balloon) in the presence of TiCl₄ (20 mol%), Zn (5 equivalents) and chlorotrimethylsilane (2 equivalents) in N,N-dimethylformamide (DMF) at room temperature, followed by treatment with an aqueous 3 M HCl solution. But, the desired oxindole 1a was not obtained, instead, indole 4 was formed (Table 1, entry 1). Indole 4 is considered to be formed via intramolecular McMurry coupling after formylation of the amine with DMF. The use of THF or DME gave the complex mixtures (Table 1, entries 2 and 3). When DMA was used as a solvent, the reactions with Cp₂TiCl₂, VCl₃, VBr₃ or Cp₂VCl₂ as a catalyst afforded the desired oxindole 1a in low yields (2–9%) with the alcohol 3 as a main product (Table 1, entries 4–7). The reaction without a catalyst also gave 1a in a low yield (Table 1, entry 8).

Table 1 Screening of an early-transition metal catalyst and a solvent for the synthesis of oxindole 1avia reductive coupling of 2a and CO₂ in the presence of Zn as a terminal reductant

Entry	Catalyst	Solvent	Yield^a/%
			Recovered 2a^b	Oxindole 1a	Alcohol 3^b
a Yield was calculated by using the integral ratio of the peaks for each compound and the internal standard (1,3,5-trimethoxybenzene) in the ^1H NMR spectrum of the crude mixture. b Starting substrate 2a and the byproducts 3 and 5 were present as a HCl salt in the aqueous layer with a HCl solution after the treatment with an aqueous 3 M HCl solution. To extract them from the aqueous layer, a saturated aqueous NaHCO3 solution was added, and they were extracted with ethyl acetate. c Reaction time was 20 h. d The formation of indole 4 and deoxygenated compound 5 was observed as by-products in this entry. e Not determined.
1^c^,^d	TiCl4	DMF	65	0	n.d.^e
2^c	TiCl4	THF	Complex mixture
3^c	TiCl4	DME	Complex mixture
4	Cp2TiCl2	DMA	20	3	13
5	VCl3	DMA	23	9	19
6	VBr3	DMA	22	4	44
7	Cp2VCl2	DMA	7	2	27
8	—	DMA	8	2	43

---

The screening was continued using Mg as a terminal reductant in DMA (Table 2). In this case, the desired oxindole 1a was obtained under the conditions with a catalyst shown in entries 1–6 in Table 2. But, the formation of alcohol 3 was observed in the presence of titanium or vanadium catalysts. Finally, the desired oxindole 1a was obtained in the absence of a catalyst in 95% yield (Table 2, entry 7).¹³ Decreasing the amount of chlorotrimethylsilane to 1 equivalent resulted in lowering of the yield of 1a to 70% (Table 2, entry 8). In the absence of both a catalyst and chlorotrimethylsilane, the product 1a was not detected, showing that chlorotrimethylsilane is essential (Table 2, entry 9). This is consistent with the previous results.¹² Instead of chlorotrimethylsilane, the use of collidine HCl salt provided 1a in a low yield (Table 2, entry 10). Keeping the equivalents of Mg at three was not a problem (97%, Table 2, entry 11). But, the further decreasing to one equivalent gave rise to a decrease of the yield of 1a (74%, Table 2, entry 12).

Table 2 Screening of a catalyst for the synthesis of oxindole 1avia reductive coupling of 2a and CO₂ in the presence of Mg as a terminal reductant

Entry	Catalyst	Yield^a/%
		Recovered 2a^b	Oxindole 1a	Alcohol 3^b
a Yield was calculated by using the integral ratio of the peaks for each compound and internal standard (1,3,5-trimethoxybenzene) in the ^1H NMR spectrum of the crude mixture. b Starting substrates 2a and 3 were present as a HCl salt in the aqueous layer with a HCl solution after the treatment with an aqueous 3 M HCl solution. To extract them from the aqueous layer, a saturated aqueous NaHCO3 solution was added, and they were extracted with ethyl acetate. c It was difficult to quantify. d Instead of chlorotrimethylsilane. e Isolated yield (containing a very small amount of ethyl acetate, see the ESI for the ^1H NMR spectrum).
1	TiCl4	—^c	44	11
2	VCl3	0	51	5
3	Cp2TiCl2	0	68	2
4	Cp2VCl2	0	80	16
5	Yb(OTf)3	7	52	16
6	VBr3	0	52	21
7	—	0	95	5
8	— (1 equiv. of Me3SiCl)	16	70	7
9	— (Without Me3SiCl)	27	0	48
10	— (With collidine HCl salt)^d	51	11	11
11	— (3 equiv. of Mg)	0	97 (99)^e	Trace
12	— (1 equiv. of Mg)	19	74	—^c

---

The scope and limitation of the substrates were investigated (Scheme 2). It is important to note that aryl halides such as fluoride, chloride and bromide can be used as a substrate to give oxindoles 1 in good yields in spite of the presence of Mg (96% for 1c, 93% for 1d, 88% for 1e, 76% for 1f and 96% for 1k). The presence of an electron donating group at the para position of the phenyl group was not a problem for this reaction (85% for 1g). The ester moiety was tolerated in this reaction to afford the corresponding product 1h in 83% yield. The product 1i with a furyl group was also synthesized in this reaction. N-Methylated substrates reacted well to give the corresponding products (76% for 1j and 96% for 1k). Instead of the aryl group, the methyl group at R³ did not provide a good result, and the yield of 1i was quite low (2%).

Scheme 2 Scope and limitation of the synthesis of oxindoles 1via reductive coupling of 2 and CO₂.

To gain an insight into the reaction path, aqueous work-up with a basic solution (aqueous saturated NaHCO₃ solution) instead of an aqueous 3 M HCl solution was carried out for the reaction of 2a. The aqueous solution was extracted with ethyl acetate. Both the organic and aqueous layers were separately treated with an aqueous 3 M HCl solution. From the aqueous layer (Scheme 3b), oxindole 1a was obtained in 64% yield, but 11% from the organic layer (Scheme 3a). These results indicate that the lactam formation in entry 7 in Table 2 and Scheme 2 mainly takes place in the process of the treatment with aqueous 3 M HCl solution.

Scheme 3 Aqueous work-up with a basic solution followed by acid treatment for the synthesis of oxindoles 1avia reductive coupling of 2a and CO₂.

Concerning the reductive coupling with CO₂, there are two plausible paths (Scheme 4),¹³ (1) one-electron reduction of the ketone, addition of the resulting radical anion species to CO₂, followed by one-electron reduction (path A) and (2) sequential two-electron reduction of the ketone and 1,2-addition of the carbanion (path B). The related papers presenting the mechanisms were also reported for both paths,^9a,14,15 which cannot be excluded for this reaction at present.

Scheme 4 Plausible paths for the synthesis of oxindoles 1avia reductive coupling of 2a and CO₂.

Conclusions

We have demonstrated the synthesis of 3-aryl-3-hydroxy-2-oxindoles 1via reductive coupling of ketoamines 2 and CO₂ as a key step. This synthetic approach is new although the synthesis of oxindoles 1 has been well studied. The conditions employing Mg with chlorotrimethylsilane in DMA are the best for the reductive coupling, where protection of the amino group is not required. In spite of using Mg, an aryl halide moiety such as aryl fluoride, chloride and bromide is intact in this reaction. Further studies including enantioselective synthesis and application to the synthesis of pharmaceutically active compounds are now underway.

Experimental

General

NMR spectra were recorded on a JEOL JNM-ECS 400 spectrometer. Chemical shifts were reported in ppm on the δ scale relative to a residual solvent (DMSO-d₆: δ = 2.50 for ¹H NMR and 39.52 ppm for ¹³C NMR) or tetramethylsilane (δ = 0 ppm for ¹H and ¹³C NMR) as an internal standard. Infrared spectra were recorded on a JASCO FT/IR-6200. Mass spectra were recorded on a JEOL JMS-700 spectrometer using the fast atom bombardment (FAB) or electron impact (EI) mode. Substrates 2a–f, k are available from commercial sources. Substrates 2g,¹⁶2i¹⁷ and 2j¹⁸ were prepared according to the literature. Substrate 2h was prepared via the Suzuki–Miyaura coupling reaction of 2d with 1.2 equivalents of 4-(ethoxycarbonyl)phenylboronic acid in the presence of 2 mol% Pd(PPh₃)₄ and aqueous Na₂CO₃ in toluene and ethanol at 80 °C overnight. Recrystallized 2h (CH₂Cl₂–hexane) after silica-gel column chromatography was used for the reaction.

General procedure

To a round-bottomed flask was added Mg (36.5 mg, 1.5 mmol). The atmosphere in the flask was exchanged with argon by repeating evacuation and purge. Then, DMA (4.1 mL, dried with MS4A before use) was added to the flask. The atmosphere in the flask was exchanged with CO₂ using a balloon by repeating evacuation and purge. Chlorotrimethylsilane (127 μL, 1.0 mmol, distilled over CaH₂ before use) and the 0.554 M DMA solution of 2 (0.9 mL, 0.5 mmol) were added to the mixture at room temperature. After the reaction mixture was stirred for 4 h, a 3 M aqueous HCl solution (5 mL) was added. The mixture was stirred for at least 30 min. The mixture was transferred to a separatory funnel. The product 1 was extracted with ethyl acetate twice. The organic layer was washed with brine, dried over Na₂SO₄ and filtered through filter paper. The filtrate was evaporated to give the crude product. To calculate the yield, 1,3,5-trimethoxybenzene (25 mg, 0.15 mmol) was added as an internal standard. The ¹H NMR spectra of the mixture were recorded. The yield was calculated by the integral ratio of the peaks for the product 1 and internal standard. Identification of the products 1 was conducted by comparison with the ¹H NMR data reported previously: 1a,⁷1b,^5c1c,¹⁹1g,²⁰1j^6a and 1k.^5c

Isolation of 1a. To remove DMA completely, the aqueous work-up was modified. After the treatment with a 3 M aqueous HCl solution, H₂O (200 mL) and ethyl acetate (50 mL) were added to the mixture in a separatory funnel. The aqueous layer was extracted with a mixed solution of ethyl acetate (20 mL) and hexane (20 mL) twice. The combined organic layer was washed with H₂O (100 mL × 2 and 200 mL). The organic layer was concentrated in vacuo to give pure 1a (111.4 mg, 0.495 mmol, 99% yield, containing a very small amount of ethyl acetate) without further purification. See the ESI‡ for the ¹H NMR spectrum.

Isolation of 1d, 1e and 1f. They were purified to remove the internal standard for the full characterization as below. To their DMSO solution was added H₂O to give a white precipitate. The precipitation was collected by filtration, and the residue was washed with CH₂Cl₂ to give the corresponding oxindole 1.

Isolation of 1h and 1i. Their crude mixtures were purified by preparative TLC (hexane/ethyl acetate = 1 : 2 for 1h and 1 : 3 for 1i).

Characterization data for the compounds whose ¹H NMR data are not reported previously

1d: ¹H NMR (400 MHz, DMSO-d₆) δ = 10.58 (s, 1 H), 7.44 (dd, J = 8.2, 1.8 Hz, 1 H), 7.38–7.257 (m, 5 H), 7.21 (d, J = 1.8 Hz, 1 H), 6.88 (d, 8.2 Hz, 1 H), 6.79 (s, 1 H) ppm; ¹³C NMR (100 MHz, DMSO-d₆) δ = 177.97, 141.24, 140.82, 136.13, 131.98, 128.28, 127.71, 127.40, 125.30, 113.65, 112.03, 77.33 ppm; IR(ATR) ν = 3369, 3199, 2358, 1704, 1473, 1182, 821, 737 cm⁻¹; HRMS(FAB) m/z: [M]⁺ calcd for C₁₄H₁₀BrNO₂: 302.9895; found: 302.9889.

1e: ¹H NMR (400 MHz, DMSO-d₆) δ = 10.67 (s, 1 H), 7.91 (ddd, 7.9, 7.9, 1.8 Hz, 1 H), 7.42–7.25 (m, 3 H), 7.06 (ddd, J = 11.5, 8.1, 1.1 Hz, 1 H), 7.00 (s, 1 H), 6.90 (d, J = 9.9 Hz, 1 H), 6.91 (s, 1 H) ppm; ¹³C NMR (100 MHz, DMSO-d₆) δ = 177.05, 158.39 (d, J = 245.4 Hz), 141.25, 134.29, 129.88 (d, J = 7.7 Hz), 129.27, 128.16 (d, J = 12.5 Hz), 128.01 (d, J = 3.8 Hz), 125.68, 124.30 (d, J = 2.9 Hz), 124.09, 115.16 (d, J = 21.1 Hz), 111.29, 74.25 ppm; IR(ATR) ν = 3410, 3237, 2367, 2328, 1713, 1485, 1049, 823, 755 cm⁻¹; HRMS(FAB) m/z: [M]⁺ calcd for C₁₄H₉ClFNO₂: 277.0306; found: 277.0308.

1f: ¹H NMR (400 MHz, DMSO-d₆) δ = 10.70 (s, 1 H), 8.06 (dd, J = 7.8, 1.8 Hz, 1 H), 7.49 (ddd, J = 7.8, 7.8, 1.8 Hz 1 H), 7.41–7.32 (m, 2 H), 7.29 (dd, J = 8.2, 2.3 Hz, 1 H), 7.03 (s, 1 H), 6.89 (d, J = 8.2 Hz, 1 H), 6.76 (d, J = 2.3 Hz, 1 H) ppm; ¹³C NMR (100 MHz, DMSO-d₆) δ = 176.45, 142.21, 137.89, 133.54, 130.28, 129.72 × 2, 129.29, 129.02, 126.89, 125.51, 123.82, 111.19, 76.07 ppm; IR(ATR) ν = 3420, 3237, 2361, 2342, 1711, 1483, 1470, 1437, 1036, 826, 746, 732 cm⁻¹; HRMS(FAB) m/z: [M]⁺ calcd for C₁₄H₉Cl₂NO₂: 293.0010; found: 293.0008.

1h: ¹H NMR (400 MHz, DMSO-d₆) δ = 10.58 (s, 1 H), 7.97 (d, J = 8.7 Hz, 2 H), 7.72 (d, J = 8.7 Hz 2 H), 7.68 (dd, J = 8.2, 2.3 Hz, 1 H), 7.46 (d, J = 1.8 Hz, 1 H), 7.36–7.24 (m, 5 H), 7.04 (d, J = 8.2 Hz, 1 H), 6.75 (s, 1 H), 4.31 (q, J = 7.3 Hz, 2H), 1.32 (t, J = 7.3 Hz, 3H) ppm; ¹³C NMR (100 MHz, DMSO-d₆) δ = 178.55, 165.63, 144.38, 142.47, 141.27, 134.76, 132.85, 129.87, 128.25, 128.13, 127.61, 126.25, 125.47, 123.18, 110.57, 77.41, 60.73, 14.23 ppm; IR(KBr) ν = 3308, 3061, 2980, 2933, 1728, 1713, 1606, 1484, 1278, 1175, 1103, 771 cm⁻¹; HRMS(EI) m/z: [M]⁺ calcd for C₂₃H₁₉NO₄: 373.1314; found: 373.1310.

1i: ¹H NMR (400 MHz, DMSO-d₆) δ = 10.44 (s, 1 H), 7.59 (brd, J = 1.8 Hz, 1 H), 7.30 (d, J = 7.3 Hz, 1 H), 7.24 (ddd, J = 7.8, 7.8, 0.9 Hz, 1 H), 6.98 (dd, J = 7.3, 7.3 Hz, 1 H), 6.85 (d, J = 7.3, 1 H), 6.73 (s, 1H), 6.40 (dd, J = 3.2, 1.8 Hz, 1 H), 6.30 (d, J = 3.2 Hz, 1 H) ppm; ¹³C NMR (100 MHz, DMSO-d₆) δ = 176.18, 153.01, 143.19, 141.69, 130.78, 129.56, 124.99, 121.87, 110.27, 109.88, 107.65, 73.48 ppm; IR(KBr) ν = 3310, 2822, 2361, 1709, 1684, 1624, 1474, 1311, 1112, 763 cm⁻¹; HRMS(EI) m/z: [M]⁺ calcd for C₁₂H₉NO₃: 215.0582; found: 215.0583.

2h: ¹H NMR (400 MHz, CDCl₃) δ = 8.02 (d, J = 8.7 Hz, 2 H), 7.75 (d, J = 2.3 Hz 1 H), 7.72–7.68 (m, 2 H), 7.62 (dd, J = 8.7, 2.3 Hz, 1 H), 7.56 (tt, J = 7.3, 1.4 Hz, 1 H), 7.51–7.46 (m, 4 H), 6.85 (d, J = 8.7 Hz, 1 H), 6.20 (brs, 2 H) ppm; ¹³C NMR (100 MHz, CDCl₃) δ = 198.89, 166.48, 150.78, 144.54, 139.77, 132.99, 132.88, 131.46, 130.13, 129.24, 128.42, 128.29, 127.25, 125.79, 118.34, 117.67, 60.88, 14.35 ppm; IR(KBr) ν = 3487, 3352, 3066, 2986, 2902, 1704, 1639, 1729, 1247, 1107, 775 cm⁻¹; HRMS(EI) m/z: [M]⁺ calcd for C₂₂H₁₉NO₃: 345.1365; found: 345.1360.

Acknowledgements

Financial support from JST (ACT-C) is acknowledged.

Notes and references

1. (a) T. Sakakura, J.-C. Choi and H. Yasuda, Chem. Rev., 2007, 107, 2365; (b) Carbon Dioxide as Chemical Feedstock, ed. M. Aresta, Wiley-VCH, Weinheim, 2010.
2. Reviews: (a) B. M. Trost and M. K. Brennan, Synthesis, 2009, 3003; (b) F. Zhou, Y.-L. Liu and J. Zhou, Adv. Synth. Catal., 2010, 352, 1381; (c) R. Dalpozzo, G. Bartoli and G. Bencivenni, Chem. Soc. Rev., 2012, 41, 7247.
3. (a) T. Tokunaga, W. E. Hume, T. Umezome, K. Okazaki, Y. Ueki, K. Kumagai, S. Hourai, J. Nagamine, H. Seki, M. Taiji, H. Noguchi and R. Nagata, J. Med. Chem., 2001, 44, 4641; (b) T. Tokunaga, W. E. Hume, J. Nagamine, T. Kawamura, M. Taiji and R. Nagata, Bioorg. Med. Chem. Lett., 2005, 15, 1789.
4. Y. Shen, J. Liu, G. Estiu, B. Isin, Y.-Y. Ahn, D.-S. Lee, A.-L. Barabási, V. Kapatral, O. Wiest and Z. N. Oltvai, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 1082.
5. Selected examples: (a) S. J. Angyal, E. Bullock, W. G. Hanger, W. C. Howell and A. W. Johnson, J. Chem. Soc., 1957, 1592; (b) Ref. 3a; ; (c) R. Shintani, M. Inoue and T. Hayashi, Angew. Chem., Int. Ed., 2006, 45, 3353.
6. Selected examples: (a) Y.-X. Jia, D. Katayev and E. P. Kündig, Chem. Commun., 2010, 46, 130; (b) L. Yin, M. Kanai and M. Shibasaki, Angew. Chem., Int. Ed., 2011, 50, 7620.
7. S. Barroso, G. Blay, L. Cardona, I. Fernández, B. García and J. R. Pedro, J. Org. Chem., 2004, 69, 6821.
8. (a) S. Wawzonek and A. Gundersen, J. Electrochem. Soc., 1960, 107, 537; (b) Y. Ikeda and E. Manda, Bull. Chem. Soc. Jpn., 1985, 58, 1723; (c) S.-F. Zhao, H. Wang, Y.-C. Lan, X. Liu, J.-X. Lu and J. Zhang, J. Electroanal. Chem., 2012, 664, 105; (d) R. Liu, G. Yuan, C. L. Joe, T. E. Lightburn, K. L. Tan and D. Wang, Angew. Chem., Int. Ed., 2012, 51, 6709; (e) S.-F. Zhao, M. Horne, A. M. Bond and J. Zhang, Phys. Chem. Chem. Phys., 2015, 17, 19247.
9. (a) Y. V. Kurbatov and A. S. Kurbatova, Russ. J. Org. Chem., 2001, 37, 1344; (b) A. Häuβermann, F. Rominger and B. F. Straub, Chem. – Eur. J., 2012, 18, 14174.
10. (a) Z. Hou, K. Takamine, O. Aoki, H. Shiraishi, Y. Fujiwara and H. Taniguchi, J. Chem. Soc., Chem. Commun., 1988, 668; (b) Z. Hou, K. Takamine, O. Aoki, H. Shiraishi, Y. Fujiwara and H. Taniguchi, J. Org. Chem., 1988, 53, 6077; (c) J. J. Eisch, P. O. Fregene and J. N. Gitua, J. Organomet. Chem., 2007, 692, 4647; (d) J. J. Eisch and P. O. Fregene, Eur. J. Org. Chem., 2008, 4482.
11. (a) M. Kanemoto, H. Ankyu, Y. Wada and S. Yanagida, Chem. Lett., 1992, 2113; (b) T. Ogata, K. Hiranaga, S. Matsuoka, Y. Wada and S. Yanagida, Chem. Lett., 1993, 983.
12. (a) T. Hirao, T. Hasegawa, Y. Muguruma and I. Ikeda, J. Org. Chem., 1996, 61, 36; (b) For a review: T. Hirao, Chem. Rev., 1997, 97, 2707; (c) T. Hirao, M. Asahara, Y. Muguruma and A. Ogawa, J. Org. Chem., 1998, 63, 2812; (d) T. Hirao, B. Hatano, M. Asahara, Y. Muguruma and A. Ogawa, Tetrahedron Lett., 1998, 39, 5247; (e) B. Hatano, A. Ogawa and T. Hirao, J. Org. Chem., 1998, 63, 9421; (f) T. Hirao, B. Hatano, Y. Imamoto and A. Ogawa, J. Org. Chem., 1999, 64, 7665; (g) T. Hirao, A. Ogawa, M. Asahara, Y. Muguruma and H. Sakurai, Org. Synth., 2005, 81, 26; (h) X. Xu and T. Hirao, J. Org. Chem., 2005, 70, 8594; (i) A. Miyasaka, T. Amaya and T. Hirao, Chem. – Eur. J., 2014, 20, 1615.
13. To check the possibility of intramolecular reductive coupling of trimethylsilyl arylcarbamate of 2a, an NMR experiment in DMSO-d₆ was carried out in the absence of Mg. As a result, the formation of trimethylsilyl arylcarbamate was not observed. Therefore, this path is excluded here.
14. Related conditions are reported for the reductive coupling of imines and CO₂: A. A. Sathe, D. R. Hartline and A. T. Radosevich, Chem. Commun., 2013, 49, 5040.
15. (a) I. Nishiguchi, M. Sakai, H. Maekawa, T. Ohno, Y. Yamamoto and Y. Ishino, Tetrahedron Lett., 2002, 43, 635; (b) T. Uchida, Y. Kita, H. Maekawa and I. Nishiguchi, Tetrahedron, 2006, 62, 3103.
16. (a) Synthesis: S. A. Chandler, P. Hanson, A. B. Taylor, P. H. Walton and A. W. Timms, J. Chem. Soc., Perkin Trans. 2, 2001, 214; (b) Characterization: K. Kobayashi, S. Fujita, S. Fukamachi and H. Konishi, Synlett, 2009, 3378.
17. V.-H. Nguyen, L. Vendier, M. Etienne, E. Despagnet-Ayoub, P.-A. R. Breuil, L. Magna, D. Proriol, H. Olivier-Bourbigou and C. Lorber, Eur. J. Inorg. Chem., 2012, 97.
18. (a) Synthesis: J. Huang, C. Wan, M.-F. Xu and Q. Zhu, Eur. J. Org. Chem., 2013, 1876; (b) Characterization: D. Susanti, L. L. R. Ng and P. W. H. Chan, Adv. Synth. Catal., 2014, 356, 353.
19. J. Gui, G. Chen, P. Cao and J. Liao, Tetrahedron: Asymmetry, 2012, 23, 554.
20. Z.-K. Xiao, H.-Y. Yin and L.-X. Shao, Org. Lett., 2013, 15, 1254.

---

Footnotes

† This paper is dedicated to Professor Barry M. Trost on the occasion of his 75th birthday.

‡ Electronic supplementary information (ESI) available: ¹H NMR spectrum for 1a, and ¹H and ¹³C NMR spectra for 1d, 1e, 1f, 1h, 1i and 2h. See DOI: 10.1039/c6qo00107f

---