Palladium-catalyzed heteroannulation of [60]fullerene with N-(2-arylethyl) sulfonamides via C–H bond activation

Abstract

The palladium-catalyzed heteroannulation of [60]fullerene with various N-(2-arylethyl) sulfonamides afforded a variety of [60]fullerene-fused tetrahydrobenzazepines. These reactions were initiated by C–H bond activation and followed by cyclization. In addition, further transformation and electrochemistry of the obtained [60]fullerene-fused tetrahydrobenzazepines were investigated.

---

Due to their immense potential applications in materials science and biological science, fullerenes and their derivatives have attracted significant attention.¹ A vast number of chemical reactions have been discovered to functionalize fullerenes over the past two decades.² Among them, transition-metal-mediated or -catalyzed reactions of [60]fullerene (C₆₀) have attracted increasing attention.³ Recently, our group has investigated reactions of C₆₀ mediated by transition metal salts such as Mn(OAc)₃,^3a,c Fe(ClO₄)₃,⁴ Cu(OAc)₂,⁵ Pb(OAc)₄ ⁶ and Ag₂CO₃ ⁷ to obtain a variety of novel fullerene products. Functionalization of C₆₀ through a Pd-catalyzed C–H bond activation strategy has also been disclosed. However, this protocol is relatively underdeveloped and limited to a few Pd-catalyzed heteroannulations of C₆₀ with anilides,^8a benzamides,^8b arylsulfonic acids,^8cN-benzyl sulfonamides,^8d phenylethyl/benzyl alcohols^8e and N-sulfonyl-2-aminobiaryls.⁹ Therefore, there is still a demand to prepare more fullerene compounds with different appended moieties through Pd-catalyzed C–H bond activation.

On the other hand, Pd-catalyzed ligand-directed C–H bond activation has emerged as one of the most powerful tools to construct C–C and C–X (X = heteroatom) bonds in organic synthesis.¹⁰ 2-Arylethylamines have been applied to construct heterocycles via C–H activation reactions.¹¹ Orito et al. reported the Pd-catalyzed direct aromatic carbonylation of 2-arylethylamines.^11a The Yu group developed Pd-catalyzed intramolecular C–H aminations to prepare indolines from 2-arylethyl triflamides and 2-arylethyl 2-pyridylsulfonamides.^11b–d However, using 2-arylethylmine derivatives to construct seven-membered tetrahydrobenzazepine via a C–H activation strategy has been unknown until now.

As the formation of C₆₀-fused seven-membered-ring compounds through Pd-catalyzed C–H bond activation presumably proceeds through a relatively scarce eight-membered palladacycle intermediate, the synthesis of C₆₀-fused seven-membered-ring compounds remains a great challenge.^8e,9 In continuation of our Pd-catalyzed reactions of C₆₀,⁸ herein we report a novel heteroannulation of C₆₀ with various N-(2-arylethyl) sulfonamides to give the rare C₆₀-fused tetrahydrobenzazepines through a Pd-catalyzed C–H activation protocol. Furthermore, these C₆₀-fused tetrahydrobenzazepines were found to undergo TfOH-promoted rearrangements.

We chose the reaction of C₆₀ with N-phenethyl-p-toluenesulfonamide (1a) as the model reaction to screen the optimal conditions. At the onset, C₆₀ (36 mg, 0.05 mmol) was allowed to react with 3 equiv. of 1a in the presence of 3 equiv. of K₂S₂O₈, 1 equiv. of mesitylenesulfonic acid dihydrate (MesSA) and 20 mol% of Pd(OAc)₂ in 6 mL of o-dichlorobenzene (ODCB) at 80 °C for 10 h. To our satisfaction, the desired fullerotetrahydrobenzazepine 2a was obtained in 6% yield (Table 1, entry 1). Replacing MesSA with p-toluenesulfonic acid (PTSA) led to 2a in 14% yield (Table 1, entry 2). When the reaction was performed in a mixture of PTSA (1 equiv.) and trifluoroacetic acid (TFA, 0.3 mL),⁹ the product yield could be further improved to 18% yield (Table 1, entry 3). However, TFA itself also provided a comparative yield with much more recovered C₆₀ (Table 1, entry 4 vs. entry 3). The use of acetic acid (AcOH) or pivalic acid (PivOH) to replace TFA was unsuccessful (Table 1, entries 5 and 6). All these results suggested that TFA was more suitable for this reaction. The critical role played by TFA was believed to be to participate in the in situ formation of a more reactive Pd(TFA)₂ species from the Pd(OAc)₂ precatalyst,¹² thus promoting the current heteroannulation of C₆₀. After the acid was selected, several oxidants were screened. Compared to K₂S₂O₈, oxone, AgOAc and p-benzoquinone (BQ), Cu(OAc)₂ afforded the best result with a yield of 30% (Table 1, entry 10 vs. entries 4 and 7–9). Notably, the fruitless attempt to replace TFA with PTSA further confirmed that TFA was the best acid for this reaction (Table 1, entry 11). In addition, the desired reaction did not take place in the presence of DMSO, CH₃CN or DMF (Table 1, entries 12–14). Reducing the Pd(OAc)₂ loading decreased the product yield dramatically (Table 1, entry 15 vs. entry 10). Gratifyingly, decreasing the amount of TFA from 0.3 mL to 0.2 mL could give a slightly better yield (32%) (Table 1, entry 16 vs. entry 10). Therefore, 0.05 mmol of C₆₀, 20 mol% of Pd(OAc)₂, 3 equiv. of 1a, 3 equiv. of Cu(OAc)₂ with ODCB–TFA (6.2 mL, v/v = 30 : 1) as the solvent at 80 °C were selected as the optimal reaction conditions.

Table 1 Screening conditions for the reaction of C₆₀ with 1a^a

Entry	Acid	Oxidant	Solvent (mL)	Yield of 2a^b (%)
a Unless otherwise specified, all reactions were performed with 0.05 mmol of C60, 0.15 mmol of 1a, 20 mol% of Pd(OAc)2, 0.15 mmol of oxidant and 0.3 mL of acid in 6 mL of ODCB at 80 °C for 10 h. b Isolated yield, values in parentheses were based on consumed C60. c 1 equiv. of the solid acid was used. d 0.3 mL of AcOH was used. e 193 mg of PivOH was used. f 10 mol% of Pd(OAc)2 was used, and the reaction time was 18 h. g 0.2 mL of TFA was used. Ts = p-toluenesulfonyl, DMSO = dimethyl sulfoxide, DMF = N,N-dimethylformamide.
1^c	MesSA	K2S2O8	ODCB (6)	6 (30)
2^c	PTSA	K2S2O8	ODCB (6)	14 (50)
3^c	PTSA/TFA	K2S2O8	ODCB (6)	18 (36)
4	TFA	K2S2O8	ODCB (6)	18 (60)
5^d	AcOH	K2S2O8	ODCB (6)	Trace
6^e	PivOH	K2S2O8	ODCB (6)	Trace
7	TFA	Oxone	ODCB (6)	8 (36)
8	TFA	AgOAc	ODCB (6)	< 3%
9	TFA	BQ	ODCB (6)	7 (58)
10	TFA	Cu(OAc)2	ODCB (6)	30 (55)
11^c	PTSA	Cu(OAc)2	ODCB (6)	Trace
12	TFA	Cu(OAc)2	ODCB–DMSO (5 : 1)	0
13	TFA	Cu(OAc)2	ODCB–CH3CN (5 : 1)	0
14	TFA	Cu(OAc)2	ODCB–DMF (5 : 1)	0
15^f	TFA	Cu(OAc)2	ODCB (6)	17 (49)
16^g	TFA	Cu(OAc)2	ODCB (6)	32 (59)

---

With the optimal conditions in hand, the scope of this annulation was explored using a variety of substrates as illustrated in Table 2. We were pleased to find that all of the examined N-(2-arylethyl) sulfonamides 1a–1i could be applied to furnish the desired C₆₀-fused tetrahydrobenzazepines 2a–2i in synthetically valuable yields. Substrates bearing either an electron-withdrawing group or an electron-donating group on the 2-arylethylamine ring worked well and gave the desired products 2b–2g in 12–47% yields (Table 2, entries 2–7). The substrate with an electron-withdrawing chloro or bromo group at the para position of the phenyl ring (1b, 1c) gave a lower yield based on consumed C₆₀ because it tended to generate some fullerene byproducts, which were most probably formed from the reaction of C₆₀ with bulk ODCB (Table 2, entries 2 and 3). The efforts to increase the product yields and inhibit the formation of byproducts by adjusting the reaction temperature and reaction time proved fruitless. The substrate 1d with a methyl group at the para position of the phenyl ring provided 2d in 40% yield (Table 2, entry 4), while the substrate 1f with a strong electron-donating methoxy group at the para position of the phenyl ring resulted in a decreased yield (Table 2, entry 6). When substrates were substituted at the meta position (1e, 1g), products resulting from the reactions at the less sterically hindered positions were regioselectively obtained in 37% and 47% yields, respectively (Table 2, entries 5 and 7). In addition, 2-phenethylamine with the methanesulfonyl (Ms) group attached to the nitrogen atom (1h) gave a comparable product yield to that of 1a (Table 2, entry 8 vs. entry 1). It is of interest to note that the tosylamide of L-phenylalanine (1i) was also reactive under our optimal conditions, and a novel C₆₀-fused amino acid derivative was isolated in 24% yield (Table 2, entry 9).

Table 2 Pd-catalyzed heteroannulation of C₆₀ with N-(2-arylethyl) sulfonamides 1a–1i^a

Entry	Substrate 1	Product 2	Yield^b (%)	Recovered C60 (%)
a Unless otherwise specified, all reactions were performed with 0.05 mmol of C60, 0.15 mmol of 1a–1i, 0.01 mmol of Pd(OAc)2, 0.15 mmol of Cu(OAc)2 and 0.2 mL of TFA in 6 mL of ODCB at 80 °C for 10 h. b Isolated yield, values in parentheses were based on consumed C60. c The reaction was performed for 6 h.
1			32 (59)	46
2			24 (39)	38
3			12 (34)	65
4			40 (65)	38
5			37 (61)	39
6			20 (77)	74
7^c			47 (63)	25
8			37 (56)	34
9			24 (69)	65

---

All products 2a–2i were unambiguously characterized using HRMS, ¹H NMR, ¹³C NMR, FT-IR and UV-vis spectra. The ESI mass spectra of 2a–2i gave the correct molecular ion peaks. In their ¹³C NMR spectra, the observation of at least 51 lines in the range of 133–156 ppm for the sp²-carbons of the C₆₀ skeleton, and two peaks at 78–80 ppm and 70–72 ppm for the two sp³-carbons of the fullerene cage, is consistent with the C₁ symmetry of their molecular structures. The IR spectra of 2a–2i showed two strong absorptions at 1330–1350 cm⁻¹ and 1140–1160 cm⁻¹ due to the sulfonamide group. Furthermore, their UV-vis spectra displayed a characteristic peak at 433–437 nm, which is the diagnostic absorption for the 1,2-adduct of C₆₀.^7,9

Based on our experimental results and previously suggested mechanisms in the literature,^8,9 a plausible mechanism for the formation of the C₆₀-fused tetrahydrobenzazepines is shown in Scheme 1. Initially, N-(2-arylethyl) sulfonamide 1 is coordinated to the Pd(II) species, followed by ortho C–H activation to produce the intermediate A. Insertion of C₆₀ into the arylpalladium bond in A yields the intermediate B. Subsequent reductive elimination of the intermediate B generates C₆₀-fused tetrahydrobenzazepine 2 and Pd(0). The latter is reoxidized to a Pd(II) species by Cu(OAc)₂ to complete the catalytic cycle.

Scheme 1 Proposed reaction mechanism for the formation of C₆₀-fused tetrahydrobenzazepines 2.

The alternative possible five-membered-ring [60]fulleroindanes resulting from annulation at the benzylic position and the ortho position of N-(2-arylethyl) sulfonamides were not observed, indicating the preference of the ligand-directed C–H activation pathway. It should be noted that the formation of [60]fulleroazepines via Pd-catalyzed C–H bond activation has been reported.⁹ However, only rigid N-sulfonyl-2-aminobiaryls were used as the substrates, and a hybrid acid system (PTSA/TFA) was crucial for the efficient formation of the seven-membered-ring products.⁹ In our present work, flexible N-(2-arylethyl) sulfonamides incorporating an alkyl chain, even the chiral amino acid moiety, were employed as the substrates, and only TFA was required as the acid additive (Table 1, entry 3 vs. entry 4). In addition, Pd(OAc)₂/Cu(OAc)₂ in ODCB–TFA (7 mL, v/v = 6 : 1), which was similar to our system, was totally inert in Chuang's work.⁹ A related study on C₆₀-fused tetrahydroazepinones and -azepinonimines formed from the Cu(OAc)₂-promoted heteroannulations of C₆₀ with N-sulfonylated o-amino-aromatic methyl ketones or O-alkyl oximes was recently disclosed.¹³ Besides the sulfonamide group, another functional group such as ketone or oxime group was required to facilitate the radical pathway.

Further transformations of representative C₆₀-fused tetrahydrobenzazepines 2a and 2i were also investigated (Scheme 2). Intriguingly, treatment of C₆₀-fused tetrahydrobenzazepine 2a with 10 equiv. of trifluoromethanesulfonic acid (TfOH) at ambient temperature for 45 min afforded a mixture of fullerotetrahydronaphthalene 3a and fulleroindane 4a (4 : 1) in a total yield of 82%. The attempt to obtain 3a and 4a directly from the Pd-catalyzed reaction of C₆₀ with 1a (Table 1, entry 16) by adding 10 equiv. of TfOH led to a yield of only 5% along with 79% of recovered C₆₀, showing extremely low efficiency of the direct transformation. Similarly, treatment of C₆₀-fused tetrahydrobenzazepine 2i with 10 equiv. of TfOH for 2 h afforded an 8.1 : 1 mixture of 3i and 4i in 55% yield. Both 3i and 4i bear a phenylalanine moiety which is a biologically active motif, hinting potential application of 3i and 4i in biomedical sciences. In addition, the preparation of C₆₀-fused tetrahydronaphthalene¹⁴ and indane^8d,14,15 derivatives is still limited; this method provides a new route to synthesize these two types of compounds.

Scheme 2 Transformation of C₆₀-fused tetrahydrobenzazepines 2a and 2i.

The half-wave reduction potentials of 2a–2i along with 5 (Fig. 1) and C₆₀ were measured by cyclic voltammetry and are summarized in Table 3. We noted that the reduction potentials of C₆₀-fused tetrahydrobenzazepines were dependent on the substituents on the 2-arylethylamine moiety. Products bearing an electron-withdrawing group are reduced more easily, whereas products with electron-donating groups give more negative first redox reduction potentials (2b–2cvs.2d–2g). Compound 2h bearing the mesyl group has a more positive first redox reduction potential than that substituted by the tosyl group (2hvs.2a). Compared with the seven-membered-ring compound 2a, the first reduction potential of its six-membered-ring analogue 5 shows a positive shift (–1.120 V for 5vs. −1.133 V for 2a), exhibiting an effect of the C₆₀-fused ring size. In addition, the first reduction potentials of 2a–2i and 5 are 40–73 mV negatively shifted compared to that of C₆₀. In consideration of their high solubility in common solvents such as CS₂, CHCl₃, chlorobenzene and o-dichlorobenzene, C₆₀-fused tetrahydrobenzazepines may have potential application in organic photovoltaic devices when combined with suitable polymer donors.^16,17 A close examination of the cyclic voltammograms indicated that the first two redoxes of 2a–2i were reversible, while their third redoxes were irreversible. Therefore, this unique electrochemical property may be exploited to generate ring-opened trianions of 2a–2i by controlled potential electrolysis (CPE), and then undergo nucleophilic reactions with various electrophiles to give other diversified fullerene derivatives.¹⁸

Fig. 1 Structure of compound 5.

Table 3 Half-wave reduction potentials (V) of 2a–2i, 5 and C₆₀^a

Compound	E 1	E 2	E 3
a Potential values versus Fc/Fc^+ reference electrode. Conditions: ca. 1 mM of the title compound and 0.1 mM of n-Bu4NClO4 in anhydrous ODCB; reference electrode: SCE; working electrode: Pt; auxiliary electrode: Pt wire; scanning rate: 20 mV s^−1. b Saturated solution of 2i.
2a	−1.133	−1.533	−2.004
2b	−1.128	−1.532	−2.041
2c	−1.124	−1.528	−2.013
2d	−1.135	−1.522	−2.008
2e	−1.132	−1.517	−1.997
2g	−1.140	−1.535	−2.043
2g	−1.149	−1.545	−2.048
2h	−1.118	−1.509	−1.992
2i	−1.122	−1.515	−1.964
5	−1.120	−1.514	−2.000
C60	−1.076	−1.460	−1.925

---

In summary, we have successfully synthesized the C₆₀-fused tetrahydrobenzazepines by the Pd-catalyzed heteroannulation of [60]fullerene with various N-(2-arylethyl) sulfonamides via a C–H bond activation strategy. The rare seven-membered products are supposed to be generated via a hard-to-form eight-membered palladacycle intermediate. In the presence of TfOH, the C₆₀-fused tetrahydrobenzazepines could be converted to the C₆₀-fused tetrahydronaphthalene and indane derivatives.

Acknowledgements

We are grateful for financial support from the National Natural Science Foundation of China (no. 21132007) and the Specialized Research Fund for the Doctoral Program of Higher Education (no. 20123402130011).

Notes and references

1. For reviews, see: (a) M. Prato, Top. Curr. Chem., 1999, 199, 173; (b) F. Diederich and M. Gomez-Lopez, Chem. Soc. Rev., 1999, 28, 263; (c) E. Nakamura and H. Isobe, Acc. Chem. Res., 2003, 36, 807; (d) D. M. Guldi, F. Zerbetto, V. Georgakilas and M. Prato, Acc. Chem. Res., 2005, 38, 38.
2. For selected reviews, see: (a) A. Hirsch, Synthesis, 1995, 895; (b) C. Thilgen and F. Diederich, Chem. Rev., 2006, 106, 5049; (c) F. Giacalone and N. Martín, Chem. Rev., 2006, 106, 5136; (d) M. Murata, Y. Murata and K. Komatsu, Chem. Commun., 2008, 6083.
3. For reviews, see: (a) G.-W. Wang and F.-B. Li, J. Nanosci. Nanotechnol., 2007, 7, 1162; (b) Y. Matsuo and E. Nakamura, Chem. Rev., 2008, 108, 3016; (c) G.-W. Wang and F.-B. Li, Curr. Org. Chem., 2012, 16, 1109; (d) M. D. Tzirakis and M. Orfanopoulos, Chem. Rev., 2013, 113, 5262.
4. For a review, see: F.-B. Li and G.-W. Wang, Sci. China Chem., 2012, 55, 2009.
5. G.-W. Wang and F.-B. Li, Org. Biomol. Chem., 2005, 3, 794.
6. F.-B. Li, T.-X. Liu, Y.-S. Huang and G.-W. Wang, J. Org. Chem., 2009, 74, 7743.
7. C.-L. He, R. Liu, D.-D. Li, S.-E. Zhu and G.-W. Wang, Org. Lett., 2013, 15, 1532.
8. (a) B. Zhu and G.-W. Wang, Org. Lett., 2009, 11, 4334; (b) S.-C. Chuang, V. Rajeshkumar, C.-A. Cheng, J.-C. Deng and G.-W. Wang, J. Org. Chem., 2011, 76, 1599; (c) F. Li, T.-X. Liu and G.-W. Wang, Org. Lett., 2012, 14, 2176; (d) Y.-T. Su, Y.-L. Wang and G.-W. Wang, Chem. Commun., 2012, 48, 8132; (e) W.-Q. Zhai, R.-F. Peng, B. Jin and G.-W. Wang, Org. Lett., 2014, 16, 1638.
9. V. Rajeshkumar, F.-W. Chan and S.-C. Chuang, Adv. Synth. Catal., 2012, 354, 2473.
10. For selected recent reviews on C–H activations, see: (a) T. W. Lyons and M. S. Sanford, Chem. Rev., 2010, 110, 1147; (b) C. S. Yeung and V. M. Dong, Chem. Rev., 2011, 111, 1215; (c) K. M. Engle, T.-S. Mei, M. Wasa and J.-Q. Yu, Acc. Chem. Res., 2012, 45, 788; (d) B.-J. Li and Z.-J. Shi, Chem. Soc. Rev., 2012, 41, 5588.
11. (a) K. Orito, A. Horibata, T. Nakamura, H. Ushito, H. Nagasaki, M. Yuguchi, S. Yamashita and M. Tokuda, J. Am. Chem. Soc., 2004, 126, 14342; (b) J.-J. Li, T.-S. Mei and J.-Q. Yu, Angew. Chem., Int. Ed., 2008, 47, 6452; (c) T.-S. Mei, X. Wang and J.-Q. Yu, J. Am. Chem. Soc., 2009, 131, 10806; (d) T.-S. Mei, D. Leow, H. Xiao, B. N. Laforteza and J.-Q. Yu, Org. Lett., 2013, 15, 3058.
12. (a) S. White, B. L. Bennett and D. M. Roddick, Organometallics, 1999, 18, 2356; (b) C. S. Yeung, N. Borduas, X. Zhao and V. M. Dong, Chem. Sci., 2010, 1, 331.
13. T.-X. Liu, Z. Zhang, Q. Liu, P. Zhang, P. Jia, Z. Zhang and G. Zhang, Org. Lett., 2014, 16, 1020.
14. T.-X. Liu, F.-B. Li and G.-W. Wang, Org. Lett., 2011, 13, 6130.
15. (a) M. Saunders, H. A. Jiménez-Vázquez, R. J. Cross, E. Billups, C. Gesenberg and D. J. McCord, Tetrahedron Lett., 1994, 35, 3869; (b) K. Kokubo, S. Tochika, M. Kato, Y. Sol and T. Oshima, Org. Lett., 2008, 10, 3335; (c) Y.-T. Su and G.-W. Wang, Org. Lett., 2013, 15, 3408.
16. For a recent review, see: C.-Z. Li, H.-L. Yip and A. K.-Y. Jen, J. Mater. Chem., 2012, 22, 4161.
17. M. Hashiguchi, N. Obata, M. Maruyama, K. S. Yeo, T. Ueno, T. Ikebe, I. Takahashi and Y. Matsuo, Org. Lett., 2012, 14, 3276.
18. (a) R. Liu, F. Li, Y. Xiao, D.-D. Li, C.-L. He, W.-W. Yang, X. Gao and G.-W. Wang, J. Org. Chem., 2013, 78, 7093; (b) Y. Xiao, S.-E. Zhu, D.-J. Liu, M. Suzuki, X. Lu and G.-W. Wang, Angew. Chem., Int. Ed., 2014, 53, 3006.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental procedures and characterization data. See DOI: 10.1039/c4qo00106k

---