English

• Open-cage fullerenes are fullerene derivatives formed through partial cleavage of the fullerene skeleton carbon-carbon bonds [1-3]. The size of the orifice in open-cage fullerenes are usually defined by the number of atoms on the rim of the orifice. Cleavage of one 5, 6-bond on C₆₀ forms a 9-membered orifice, whereas one 6, 6-bond forms a 10-membered orifice. Consecutive cleavage of three 6, 6-bonds around the same pentagon leads a 18-membered orifice. Insertion of a heteroatom such as nitrogen [4], oxygen [5] and sulfur [6] on the rim of orifice has been reported to expand the size of the orifice. Open-cage fullerenes serve as unique molecular containers because of their fully covalent bonded cage structure. Various neutral guest molecules have been inserted into the cavity of open-cage fullerenes with a suitable orifice size including He [7], O₂ [8], HF [9, 10], HCCH [11] and CH₃OH [12]. Recently halide anions [13] and LiF and [BeF]⁺ [14] have been inserted into open-cage fullerene derivatives.

  Preparation of open-cage fullerenes is a challenging task due to their unique spherical structure. Classical organic carbon-carbon bond breaking reactions usually results in complicated mixtures of products when applied to fullerenes. It is not uncommon that fullerene chemistry shows special reaction pathways [15-18]. The first open-cage fullerene derivative has a 11-membered orifice, which was reported back in 1995 by Wudl et al. [4]. Since then several methods have been reported for the preparation of open-cage fullerenes [1]. Singlet oxygen oxidation is one of the frequently used method, which results in two carbonyl groups on the rim of the orifice per singlet oxygen addition step. The fullerene-peroxide mediated cage-opening method has generated a number of open-cage fullerenes with a relatively large orifice [[19], [20]]. Anilines are very effective reagents for orifice expansion through complicated rearrangement processes [21]. Most of the reported methods were found through serendipity. Based on the reported methods, it is still not possible to rationally design the synthesis of open-cage fullerenes with a specific size of orifice and special functional groups around the rim of the orifice for functional studies. More information is still needed to selectively cleave the fullerene skeleton bonds and prepare open-cage fullerenes for practical application exploration. In the present work, we report the formation of 18-membered open-cage fullerene derivatives through a selective aniline mediated orifice expansion reaction.

  In our previous study we have shown that compound 1 react with anilines to form the 14-membered and the 18-membered open-cage fullerene derivatives 2 and 3 depending on the amount of anilines added (Scheme 1) [[22], [23]]. The bis-imino derivative 3 were formed readily in the presence of excess anilines. To obtain the mono-imino derivative 2, we had to add the aniline very carefully to avoid the formation of the bis-imino derivative 3. The imino groups in compound 3 are very stable towards hydrolysis. It was not possible to hydrolyze these imino groups into carbonyl groups. We tried to convert 2 into the mono imino analogue of compound 3 under various conditions, but only observed decarbonylation derivatives.

  Scheme 1

  

  Scheme 1.  Reactions of compound 1 with less hindered aniline derivatives.

  DownLoad: Full-Size Img PowerPoint

  All the anilines used previously has little steric hindrance because the substituent is located at either the para or meta-position on the benzene ring (Scheme 1). In an effort to make mono imino analogue of compound 3, we tested the reaction of 1 with 2, 6-diisopropylaniline which is quite sterically hindered. Instead of the orifice expanded derivative 2 for less hindered anilines, we isolated compound 4 as the major product (Scheme 2). As expected the more sterically hindered 2, 6-diisopropylaniline was much less reactive and there was no bis-imino derivative detected even in the presence of four equivalent of 2, 6-diisopropylaniline. Interestingly the purity of compound 1 was quite important for the formation of 4. We had to remove all the impurities by repeated preparative thin layer chromatography (PTLC) purification. Impurities present in compound 1 severely affected the yield of 4.

  Scheme 2

  

  Scheme 2.  Reaction of 1 with 2, 6-diisopropylaniline and subsequent reactions with NBS and Lewis acids.

  DownLoad: Full-Size Img PowerPoint

  Compound 4 has a 15-membered orifice including the oxygen atom on the rim of the orifice. To further expand the orifice, we tried various conditions to hydrolyze the acetal moiety. Both B(C₆F₅)₃ and FeCl₃ were effective at removing the acetyl group to form compound 5. The epoxy moiety was converted into a hydroxyl group along with the shift of the C = C double bond on the lifted pentagon ring. This double bond shift also took place in the epoxide opening process with N-bromosuccinimide (NBS) to form compound 6. The phenyl group was also brominated at the 4-postion during the reaction with NBS. The imino group remains unchanged under both the conditions for the formation of 5 and 6. Surprisingly there is a chloro atom on the lifted pentagon instead of the bromine atom. The original bromine atom was apparently replaced by a chlorine atom from chloride impurities such HCl in the solvents during the chromatography purification procedure. Silica gel could be another possible source for chloride. Flood et al. has reported that chloride on silica gel was extracted into their chloride recognition molecule [24].

  The structure of 4 is quite different from that of 2. The acetyl group is still present in 4. A possible mechanism is shown in Scheme 3 for the formation of 4. Unlike the less hindered anilines which act as nucleophilic reagents and attack the acetyl group in the first step, the present sterically hindered 2, 6-diisopropylaniline acts as an reductant in the first step. The initial electron transfer process forms intermediate A. Among the various resonance forms, structure A is the most favored because the negative chrge is next to the electron withdrawing carbonyl group, and the radical can form conjugation with the C=C double bond (an ally radical). The other alternative structure with the negative charge also next to the carbonyl group but on the oxygen bonded carbon is less favorable because the electron donating ether oxygen would make it less stable. The charged nature of the aniline radical cation and the radical anion A provides driving force for the formation of intermediate B even though steric hindrance is still quite high. The cleavage of the fullerene C—C bond may take place through the six-membered transition state shown in B. Cleavage of the tbutylperoxo O—O bond helps to reduce the steric hindrance around the bulky 2, 6-diisopropylanilino group. It is possible that the tbutylperoxo O—O bond was cleaved before the fullerene C—C bond.

  Scheme 3

  

  Scheme 3.  Proposed mechanism for the reaction between 2, 6-diisopropylaniline and compound 1.

  DownLoad: Full-Size Img PowerPoint

  The above electron transfer pathway is further supported by the reaction between 1 and 4, 4′-methylenedianiline, the major product of which is compound 7 analogous to compound 4. Presence of the two amino groups facilitates electron transfer process. In addition, the methyleneanilino group is also a bulky group and may prevent direct addition through interactions with the tbutylperoxo groups. Only after the electron transfer process to form the charged radical ions, the enhanced attractions between the charged radical species can overcome the steric hindrance. The remaining anilino group in 7 showed limited reactivity towards 1. There was no dimeric product detected. Steric hindrance of the fullerene cage and the addends on it apparently makes the aniline moiety in compound 7 unreactive towards another molecule of 1.

  All the new compounds are C₁ symmetric. Their NMR spectra agree with the structures shown in Scheme 2, Scheme 4. The phenyl group in the 2, 6-diisopropylaniline derivatives 4, 5 and 6 showed two sets of distinct signals for the two isopropyl groups, i.e., two multiplets for the two tertiary C—H protons and four doublets for the four CH₃ protons. So the rotation is restricted for both the N-Ar and the ⁱPr-Ar single bonds. The phenyl group in the 4, 4′-methylenedianiline derivative 7 showed free rotation, each of the phenyl rings showed just two doublets for the four protons. The carbonyl groups showed the expected number of signals above 177 ppm on the ¹³C NMR spectra. The two dimensional HMBC spectra of compounds 6 and 7 revealed the acetyl carbonyl carbon at 167.30 and 167.68 ppm respectively. The imino carbon is difficult to assign conclusively because its chemical shift is similar to the fullerene skeleton carbon signals. The expected molecular ion signals were observed on the mass spectra. The mass spectrum of 6 showed clearly the presence of one Cl and one Br atom.

  Scheme 4

  

  Scheme 4.  Reaction between compound 1 and 4, 4′-methylenedianiline.

  DownLoad: Full-Size Img PowerPoint

  The structure of compound 6 is further confirmed by single crystal X-ray diffraction analysis (Fig. 1). The unit cell contains two pairs of enantiomers. The hydroxyl group forms H-bond with the imino nitrogen atom. The N^…H-O distance is 2.861 Å. The phenyl ring is almost perpendicular to the lifted pentagon ring, leaving the two isopropyl groups away from the adjacent carbonyl oxygen atom on the pentagon. Space-filling model indicates that rotation of the phenyl group is hindered mainly by the carbonyl oxygen group. The C=O bond distance (1.191(4) Å) of the carbonyl group on the lifted pentagon is slightly shorter than those of the other two carbonyl groups (1.208(4), 1.203(4) Å), as a result of less distortion by the spherical cage structure.

  Figure 1

  

  Figure 1.  Single-crystal X-ray structure of 6. Ellipsoids were drawn at the 50% level. Color key: gray, carbon; red, oxygen; blue, nitrogen; green, chlorine; wine red, bromine. Dashed blue line indicates H-bond.

  DownLoad: Full-Size Img PowerPoint

  In conclusion, steric effect plays an important role in the reaction between aniline and open-cage fullerene derivative 1. Less hindered anilines act as nucleophiles and directly attack less crowded electron positive positions around the rim of the orifice. The bulky aniline follows an electron transfer pathway as the first step to form ionic radical intermediates, which than interact with each other to form an aniline derivative. As a result of their different reaction mechanism, the structure of the aniline added derivatives can be quite different. Unlike the less hindered anilines which easily form open-cage fullerenes with two imino groups, bulky anilines gave open-cage fullerenes with just one imino group, which is quite stable towards hydrolysis. Further work is underway to explore the chemical reactivity and the application of open-cage fullerene as ligands for metal complexes [25].

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Linlin Yu: Investigation. Xueli Liu: Investigation. Rui Gao: Investigation. Jialin Ming: Funding acquisition. Yi Qiu: Investigation. Jie Su: Investigation. Liangbing Gan: Writing – review & editing, Writing – original draft, Supervision, Project administration, Funding acquisition, Conceptualization.

  Acknowledgments

  This work is supported by Beijing National Laboratory for Molecular Sciences (No. BNLMS-CXXM-201904) and National Natural Science Foundation of China (NSFC, Nos. 22171009, and 22261039). The measurements of NMR and Mass Spectrometry were performed at the Analytical Instrumentation Center of Peking University. The help from PKUAIC (Ms. Hui Fu, Ms. Xiu Zhang and Ms. Xiaoran He) was acknowledged.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110382.

  
  
• 1. [1]

     G.C. Vougioukalakis, M.M. Roubelakis, M. Orfanopoulos, Chem. Soc. Rev. 39 (2010) 817–844. doi: 10.1039/B913766A

  2. [2]

     M. Murata, Y. Murata, K. Komatsu, Chem. Commun. 46 (2008) 6083–6094. doi: 10.1039/b811738a

  3. [3]

     L.B. Gan, D.Z. Yang, Q.Y. Zhang, et al., Adv. Mater. 22 (2010) 1498–1507. doi: 10.1002/adma.200903705

  4. [4]

     J.C. Hummelen, M. Prato, F. Wudl, J. Am. Chem. Soc. 117 (1995) 7003–7004. doi: 10.1021/ja00131a024

  5. [5]

     Z. Xiao, J.Y. Yao, D.Z. Yang, et al., J. Am. Chem. Soc. 129 (2007)16149–16162. doi: 10.1021/ja0763798

  6. [6]

     C.M. Stanisky, R.J. Cross, M. Saunders, et al., J. Am. Chem. Soc. 127 (2005) 299–302. doi: 10.1021/ja045328x

  7. [7]

     Y. Rubin, T. Jarrosson, G.W. Wang, et al., Angew. Chem. Int. Ed. 40 (2001) 1543–1546. doi: 10.1002/1521-3773(20010417)40:8<1543::AID-ANIE1543>3.0.CO;2-6

  8. [8]

     Z.S. Zhou, H.F. Han, Z.J. Chen, et al., Angew. Chem. Int. Ed. 58 (2019) 17690–17694. doi: 10.1002/anie.201911631

  9. [9]

     A. Krachmalnicoff, R. Bounds, S. Mamone, et al., Nat. Chem. 8 (2016) 953–957. doi: 10.1038/nchem.2563

  10. [10]

      L. Xu, H.J. Ren, S.S. Liang, et al., Chem. Eur. J. 21 (2015) 13539–13543. doi: 10.1002/chem.201502306

  11. [11]

      C.S. Chen, W.Y. Yeh, Chem. Eur. J. 22 (2016) 16425–16428. doi: 10.1002/chem.201604114

  12. [12]

      T. Futagoishi, M. Murata, A. Wakamiya, et al., Angew. Chem. Int. Ed. 56 (2017) 2758–2762. doi: 10.1002/anie.201611903

  13. [13]

      S.J. Sun, Z. Liu, F. Colombo, et al., Angew. Chem. Int. Ed. 61 (2022) e2022120.

  14. [14]

      R. Gao, Z. Liu, Z.Y. Liu, et al., Angew. Chem. Int. Ed. 62 (2023) e2023001.

  15. [15]

      L.B. Gan, D.J. Zhou, C.P. Luo, et al., Chin. Chem. Lett. 6 (1995), 199–200.

  16. [16]

      C.M. Du, J.H. Xu, Y.L. Li, et al., Chin. Chem. Lett. 12 (2001), 683–686.

  17. [17]

      X.X. Yan, C. Niu, Z.C. Yin, et al., Sci. Bull. 67 (2022), 2406–2410. doi: 10.1016/j.scib.2022.11.023

  18. [18]

      Z.C. Yin, M.J. Li, C. Niu, et al., Angew. Chem., Int. Ed. 62 (2023), e202304321. doi: 10.1002/anie.202304321

  19. [19]

      L.B. Gan, Chin. J. Chem. 36 (2018) 991–994. doi: 10.1002/cjoc.201800298

  20. [20]

      For a recent example, Y.Q. Wang, Y.R. Wu, Y.M. Yu, et al., Org. Chem. Front. 11 (2024) 3026–3032. doi: 10.1039/d4qo00449c

  21. [21]

      S-i. Iwamatsu, S. Murata, Synlett. 14 (2005) 2117–2129. doi: 10.1055/s-2005-872238

  22. [22]

      Q.Y. Zhang, Z.S. Jia, S.M. Liu, et al., Org. Lett. 11(2009) 2772–2776;. doi: 10.1021/ol9009305

  23. [23]

      Y.M. Yu, X. Xie, T. Zhang, et al. J. Org. Chem. 76 (2011) 10148–10153. doi: 10.1021/jo202014q

  24. [24]

      Y. Liu, W. Zhao, C.H. Chen, et al., Science 365 (2019), 159–161. doi: 10.1126/science.aaw5145

  25. [25]

      For a recent example, R. Gao, Z. Liu, Z.Y. Liu, et al., J. Am. Chem. Soc. 145 (2023) 18022–18028.

• 

  Scheme 1  Reactions of compound 1 with less hindered aniline derivatives.

  下载: 全尺寸图片 幻灯片
  

  Scheme 2  Reaction of 1 with 2, 6-diisopropylaniline and subsequent reactions with NBS and Lewis acids.

  下载: 全尺寸图片 幻灯片
  

  Scheme 3  Proposed mechanism for the reaction between 2, 6-diisopropylaniline and compound 1.

  下载: 全尺寸图片 幻灯片
  

  Scheme 4  Reaction between compound 1 and 4, 4′-methylenedianiline.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  Single-crystal X-ray structure of 6. Ellipsoids were drawn at the 50% level. Color key: gray, carbon; red, oxygen; blue, nitrogen; green, chlorine; wine red, bromine. Dashed blue line indicates H-bond.

  下载: 全尺寸图片 幻灯片
•