Abstract
A three‐dimensional π‐conjugated chiral cage with six [5]helicene units (a triple helicene cage) was synthesized for the first time. Taking advantage of the Yamamoto coupling reaction, the triflate‐substituted triple [5]helicene, a strained and preorganized precursor, was dimerized to afford the target compound. Single‐crystal X‐ray diffraction analysis revealed the unique structural features of the triple helicene cage: a cage‐shaped rigid structure with outer helical grooves and an inner chiral cavity. All‐P and all‐M enantiomers were separated successfully by HPLC over a chiral column and their chiroptical properties were characterized by circular dichroism spectra.
Keywords: aromaticity, chirality, multiple helicene, polycyclic aromatic hydrocarbons, π-conjugated cages
Closed and three‐dimensional (3D) π‐conjugated hydrocarbons such as fullerenes and carbon nanotubes have received extensive attention in numerous research fields because of their aesthetic structures, unique properties, and various applications.1 Development of the bottom‐up strategy for the synthesis of 3D nanocarbons is a hot topic in organic chemistry and material sciences. In 2017, the carbon nanobelt, a milestone compound in this field, was reported.2 However, construction of closed and 3D polycyclic aromatic hydrocarbons (PAHs) with fused ring systems remains a significant challenge. Cage‐shaped conjugated hydrocarbons have been investigated over the last few decades as model compounds in nanoarchitecture studies3a, 3b or as precursors of fullerenes.3c–3g Recently, cage‐shaped phenylene multirings have been reported independently by the Itami4 and Yamago5 groups (Figures 1 A and 1 B). These molecules consist exclusively of discrete benzene rings, in which para‐phenylenes are linked by 1,3,5‐tri‐substituted benzene intersections. In these highly distorted compounds, the strain across the whole molecule was usually generated in the last step of their synthesis using relatively less strained precursors. Although these elegant methods have offered new strategies for the synthesis of para‐phenylene‐based multirings, they are not applicable to the construction of 3D highly strained fused‐ring systems. Recently, we have reported threefold‐symmetrical multiple helicenes, such as the triple [5]helicene T5H,6 a structural isomer of hexabenzotriphenylene HBTP 7 and hexapole [5]helicene H5H 8 (Figure 1). These multiple helicenes have highly distorted structures, because of the repulsion of multiple [5]helicene units. In particular, the thermodynamically most stable isomer of T5H, synthesized through oxidative photocyclization,9 can provide a unique platform with three functional groups at the tip of its peripheral moieties positioned on the same side of the molecule.
Figure 1.

Examples of cage‐shaped π‐conjugated hydrocarbons and threefold‐symmetrical multiple helicenes.
We considered that the structural features of T5H must represent a promising building unit to construct fused‐ring 3D cage structures, because T5H is preorganized for the synthesis of cage molecules. HBTP and H5H are not always suitable building units for this particular purpose, because they suffer from possible formation of undesired isomers or conformations during their synthesis through the pyrolysis or cyclotrimerization of arynes. Herein, we report the synthesis and properties of a triple helicene cage (THC 1) that consists of 20 benzene rings with six [5]helicene units, and represents the first example of fused‐ring π‐conjugated 3D cage‐shaped molecules.
Our strategy for the synthesis of THC includes oxidative photocyclization10 for the construction of the T5H backbone and a subsequent aryl–aryl coupling reaction (Scheme 1). The yields of each step from compounds 2 to 7 were moderate to excellent. Ni0‐mediated Yamamoto coupling of triflate 7 gave the target compound THC in 4 % yield.11 The low yield in the last step is probably because a racemic precursor was used. Generally, the reaction of two compounds with the same chirality dimerize, whereas those with different chirality form oligomers. Although the solubility of the isolated THC is low in common organic solvents such as benzene, toluene, dichloromethane, and chloroform, it is moderately soluble in carbon disulfide.
Scheme 1.

Synthesis of triple helicene cage THC 1. Reagents and conditions: i) 1 m HCl, 96 %; ii) stearoyl chloride, TEA, 99 %; iii) hν, I2, K2CO3, 51 %; iv) LiAlH4, 96 %; v) Tf2O, pyridine, quant.; vi) Ni(COD)2, bpy, 1,5‐COD then 7, 4 %.
The 1H NMR spectrum of THC consists of only seven peaks in carbon disulfide/[D2]tetrachloroethane (Figure 2 a, black), indicating its highly symmetrical structure. The 1H NMR spectrum of THC showed almost the same characteristics as those of [5]helicene12 and T5H,6 except for Hf and Hg. THC is chiral and each enantiomer was successfully separated by HPLC over a chiral column. The ee value of the first and second fractions was 97 % (Figure S23). 1H NMR spectra of the racemic form and purified second eluted enantiomer of THC measured at 295 and 233 K were compared (Figures 2, S21, and S22). Signals assigned to Hf and Hg showed larger upfield shifts than other signals when the temperature decreased. These two signals are susceptible to environmental change because Hf and Hg reside at ortho positions of the biphenyl moiety, which is the most flexible part of this rigid molecule. Figure 2 a shows that the two 1H NMR spectra measured at 295 K were almost identical. In contrast, the spectral region at approximately 7.7 ppm showed small differences for these samples measured at 233 K. The spectral differences observed between two stereochemically different samples may reflect differences in the aggregation status between homochiral and heterochiral molecules. This postulate is based on the packing structure of THC (Figure 4).
Figure 2.

1H NMR spectra of THC of racemic (black) and the second eluted enantiomer (red) in carbon disulfide/[D2]tetrachloroethane: a) 295 K; b) 233 K.
Single crystals for X‐ray diffraction analysis were obtained by slow evaporation of carbon disulfide/benzene/toluene solutions of THC at room temperature.13 The crystal structure of THC revealed that its unit cell contained two crystallographically independent pairs of enantiomers [pairs α (C1−) and β (C85−); space group: P‐1]. THC exhibited a pseudo‐threefold rotational axis through the centroid of the central benzene ring of the T5H moieties. The rigid PAH skeleton ensures that THC retains a highly symmetrical structure of the inner cavity, even in the crystal; otherwise, the hollow cavity is usually in a collapsed state. The most unique structural features of THC were the cage‐shaped structure with outer helical grooves and the inner chiral cavity (Figures 3 a, 3 b, and S24). Six [5]helicene moieties in THC have the same helicity, all‐P [(P)‐THC] and all‐M [(M)‐THC] forms. THC in the α and β pair has 43 and 52 Å3 of inner void space, respectively. The distances between the two benzene rings located centrally at the top and bottom of the T5H moieties are 6.32 Å (for α) and 6.33 Å (for β). Structural analysis revealed that THC has a cavity like a three‐leaf clover, including a 3.36 Å diameter inscribed sphere. The electron density inside the cage is too low to identify any guests. Even N,N‐dimethylformamide, the solvent for the synthesis of THC, is larger than the cavity size. The windows of THC are very narrow and small molecules such as carbon disulfide used in the recrystallization process cannot gain access. The dihedral angles, defined as the angle formed by the two benzene rings located at the terminal edges of each [5]helicene subunit, of the six [5]helicene moieties (average: 57.7° for α and 58.3° for β, Table S1) are larger than the three [5]helicene moieties of T5H (average: 52.8°). Three single bonds between two T5H units hoist opposite [5]helicene arms toward each other to make the dihedral angles wider in THC (Figure S25).
Figure 3.

X‐ray structures of (P)‐THC‐β: a) side and b) top view. ORTEP drawings are shown at 50 % probability. Hydrogen atoms are omitted for clarity. Approximate distances are provided within the double‐headed arrows. The solvent accessible void (1.2 Å probe) is depicted in orange. c) Individual and d) averaged HOMA values (black numbers in rings) and dihedral angles (red numbers) are shown.
Distances between two carbon atoms at the tip of the fjord region in the [5]helicene moieties of THC were 3.000 Å (average for α) and 3.007 Å (average for β) (Table S2), which are longer than those in T5H (average: 2.927 Å). The C−C bond between two benzene rings of biphenyl moieties in THC (Table S2) were normal values for biphenyl compounds. To evaluate the local aromaticity of the individual rings, the harmonic oscillator model of aromaticity14 (HOMA) values were calculated. Central green‐colored benzene rings A and K (Figures 3 c and 3 d; for ring naming see Table S1) have the lowest HOMA value (average: 0.402 for α and 0.320 for β) in THC, which is similar to that in T5H (0.302). In contrast, HOMA values of the central benzene ring in HBTP (0.711) and H5H (0.687) are far higher than that found in THC. These differences are derived from symmetric properties where HBTP and H5H have D 3 symmetry, whereas T5H has lower C 3 symmetry that causes more distortion at ring AAV (AAV denotes the averaged properties of rings A and K; Table S1) in THC. Bond alternation of ring AAV in THC is larger than those in HBTP and H5H (Table S2). Although the bond lengths at fused positions of ring AAV in THC are comparable with those in HBTP and H5H, those of non‐fused positions in THC are longer than those in HBTP and H5H. The HOMA values for rings BAV and CAV are moderate [0.618 and 0.500 (for α), 0.631 and 0.535 (for β)] and the HOMA value for ring DAV is high with values of 0.846 (for α) and 0.825 (for β). These values are almost the same as those in T5H, indicating that strains caused by annulation are effectively dispersed across the whole molecule. The crystal structure of THC matched well with that of the optimized structure obtained by DFT calculations [B3LYP/6–31‐G(d)], indicating that there is little influence residing in the crystal (Table S3). Strain energies of THC and T5H were calculated to be 79.6 and 38.2 kcal mol−1, respectively (DFT at the same level of theory). An increase in the strain energy by 3.2 kcal mol−1 corresponds to the distortion caused by cage formation (Figure S26). Structural similarity between T5H and THC implies that our synthetic strategy using a distorted and preorganized precursor to construct a cage compound is effective. Three helical grooves in THC of both enantiomers intercalate with each other (Figure 4 a). One (P)‐THC is surrounded by three (M)‐THCs, and vice versa, to form a lamellar structure. A void space surrounded by six THC molecules, which is filled with solvent molecules (see the Supporting Information) is connected to each other to form a 1D column, thereby forming a crystal honeycomb structure (Figures 4 b and 4 c). Strong π–π interactions make this compound barely soluble in various solvents.
Figure 4.

Crystal packing of THC. (P)‐THC and (M)‐THC are shown in blue and pink, respectively. a) The shortest C−C distances between neighboring enantiomers are provided. b) View along the a axis. c) View along the c axis. Solvent molecules are omitted for clarity.
The photophysical properties of THC were measured in chloroform (Figure 5). Determining the precise extinction coefficient by UV/Vis spectral analysis was difficult because of low solubility. The absorption maximum of THC was observed at 346 nm, which is 6 nm shorter than that of T5H. This small blueshift is presumably caused by a decrease in the effective conjugation. THC exhibits weak fluorescence at 470 nm (λex=346 nm) with a low fluorescence quantum yield of 0.017, which is comparable to those of T5H (φ=0.026), H5H (φ=0.039), and other helicenes.15 In the circular dichroism (CD) spectrum of THC, the first peak of the two fractions eluted by chiral HPLC exhibits a positive Cotton effect over the ranges of 366–430 and 285–342 nm, whereas a negative Cotton effect was observed at 342–366 and below 285 nm (Figure 5). The second eluting peak showed a mirror‐imaged CD spectrum against the first one. Simulated CD spectra of (P)‐ and (M)‐THC prepared by using TDDFT calculations match those observed for the first and second eluted peaks, respectively (Figure S27). The anisotropy factors (g values: g=Δϵ/ϵ) of the first eluting enantiomer of THC are +13.3×10−3 (400 nm) and +13.9×10−3 (309 nm). These g values are higher than those of T5H [(P)‐isomer, g=+7.8×10−3 at 281 nm), H5H [(P,M,P,M,P,M)‐isomer, g=−4.8×10−3 at 324 nm], (P)‐(+)‐[5]helicene (g=+4.2×10−3 at 310 nm), and (P)‐(+)‐[6]helicene (g=+9.2×10−3 at 324 nm).16 The high g values observed for THC arise from its characteristic structure with six [5]helicene units that have the same helicity.
Figure 5.

Optical properties of THC and T5H in chloroform. Upper panel: UV/Vis absorption (solid line) and fluorescence spectra (dotted line) of THC (blue) and T5H (green). Lower panel: CD spectra of THC (solid line) and T5H (dotted line). The gray bars show the TDDFT‐derived rotatory strength values for (P)‐THC at the B3LYP/6‐31G(d) level.
The calculated HOMO and LUMO energy levels for THC were −5.12 and −1.64 eV, respectively (Figure S28). The former is higher than that for T5H (−5.29 eV), and the latter is lower than that obtained for T5H (−1.60 eV) at the same level of theory, which means that the HOMO–LUMO energy gap of THC (3.48 eV) is reduced when compared with that of T5H (3.69 eV).
In summary, the synthesis of THC, a 3D cage‐shaped molecule composed of six [5]helicene units, is reported. The key to the successful synthesis of THC is the use of the rigid and pre‐organized T5H precursor in the Yamamoto coupling reaction. The structure of THC was unambiguously determined by single‐crystal X‐ray diffraction analysis. THC is chiral with unique structural features; outer helical grooves and an inner chiral cavity caused by the six [5]helicene moieties. Our synthetic strategy to construct this cage‐shaped molecule provides a new method to synthesize novel 3D π‐conjugated molecules. The inherently chiral THC should facilitate the study of 3D chiral nanocarbons.
Conflict of interest
The authors declare no conflict of interest.
Supporting information
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Supplementary
Acknowledgements
This work was supported by MEXT‐Supported Program for the Strategic Research Foundation at Private Universities, 2012–2016 (Grant S1201034). This work was also supported in part by a Faculty of Science Grant‐in‐Aid for Scientific Research (2017) from Toho University. The computations were performed using the Research Center for Computational Science, Okazaki, Japan.
T. Matsushima, S. Kikkawa, I. Azumaya, S. Watanabe, ChemistryOpen 2018, 7, 278.
References
- 1.
- 1a. Segawa Y., Ito H., Itami K., Nat. Rev. Mater. 2016, 1, 15002; [Google Scholar]
- 1b. Segawa Y., Yagi A., Matsui K., Itami K., Angew. Chem. Int. Ed. 2016, 55, 5136–5158; [DOI] [PubMed] [Google Scholar]; Angew. Chem. 2016, 128, 5222–5245; [Google Scholar]
- 1c. Yamago S., Kayahara E., Iwamoto T., Chem. Rec. 2014, 14, 84–100; [DOI] [PubMed] [Google Scholar]
- 1d. Kawase T., Kurata H., Chem. Rev. 2006, 106, 5250–5273. [DOI] [PubMed] [Google Scholar]
- 2. Povie G., Segawa Y., Nishihara T., Miyauchi Y., Itami K., Science 2017, 356, 172–175. [DOI] [PubMed] [Google Scholar]
- 3.
- 3a. Wu Z., Lee S., Moore J. S., J. Am. Chem. Soc. 1992, 114, 8730–8732; [Google Scholar]
- 3b. Högberg H.-E., Thulin B., Wennerström O., Tetrahedron Lett. 1977, 18, 931–934; [Google Scholar]
- 3c. Rubin Y., Parker T. C., Kahn S. I., Holliman C. L., McElvany S. W., J. Am. Chem. Soc. 1996, 118, 5308–5309; [Google Scholar]
- 3d. Rubin Y., Parker T. C., Pastor S. J., Jalisatgi S., Boulle C., Wilkins C. L., Angew. Chem. Int. Ed. 1998, 37, 1226–1229; [DOI] [PubMed] [Google Scholar]; Angew. Chem. 1998, 110, 1353–1356; [Google Scholar]
- 3e. Tobe Y., Nakagawa N., Naemura K., Wakabayashi T., Shida T., Achiba Y., J. Am. Chem. Soc. 1998, 120, 4544–4545; [Google Scholar]
- 3f. Tobe Y., Nakagawa N., Kishi J., Sonoda M., Naemura K., Wakabayashi T., Shida T., Achiba Y., Tetrahedron 2001, 57, 3629–3636; [Google Scholar]
- 3g. Tobe Y., Umeda R., Sonoda M., Wakabayashi T., Chem. Eur. J. 2005, 11, 1603–1609. [DOI] [PubMed] [Google Scholar]
- 4.
- 4a. Matsui K., Segawa Y., Namikawa T., Kamada K., Itami K., Chem. Sci. 2013, 4, 84–88; [Google Scholar]
- 4b. Matsui K., Segawa Y., Itami K., J. Am. Chem. Soc. 2014, 136, 16452–16458. [DOI] [PubMed] [Google Scholar]
- 5. Kayahara E., Iwamoto T., Takaya H., Suzuki T., Fujitsuka M., Majima T., Yasuda N., Matsuyama N., Seki S., Yamago S., Nat. Commun. 2013, 4, 2694. [DOI] [PubMed] [Google Scholar]
- 6. Saito H., Uchida A., Watanabe S., J. Org. Chem. 2017, 82, 5663–5668. [DOI] [PubMed] [Google Scholar]
- 7.
- 7a. Barnett L., Ho D. M., Baldridge K. K., Pascal R. A., J. Am. Chem. Soc. 1999, 121, 727–733; [Google Scholar]
- 7b. Peña D., Pérez D., Guitián E., Castedo L., Org. Lett. 1999, 1, 1555–1557. [Google Scholar]
- 8. Hosokawa T., Takahashi Y., Matsushima T., Watanabe S., Kikkawa S., Azumaya I., Tsurusaki A., Kamikawa K., J. Am. Chem. Soc. 2017, 139, 18512–18521. [DOI] [PubMed] [Google Scholar]
- 9.
- 9a. Mallory F. B., Mallory C. W., Org. React. 1984, 30, 1–456; [Google Scholar]
- 9b. Mallory F. B., Wood C. S., Gordon J. T., J. Am. Chem. Soc. 1964, 86, 3094–3102; [Google Scholar]
- 9c. Wood C. S., Mallory F. B., J. Org. Chem. 1964, 29, 3373–3377. [Google Scholar]
- 10. Matsushima T., Kobayashi S., Watanabe S., J. Org. Chem. 2016, 81, 7799–7806. [DOI] [PubMed] [Google Scholar]
- 11.Recent reports describing the use of Yamamoto coupling to construct 3D molecules:
- 11a. Mys̈liwiec D., Kondratowicz M., Lis T., Chmielewski P. J., Stépień M., J. Am. Chem. Soc. 2015, 137, 1643–1649; [DOI] [PubMed] [Google Scholar]
- 11b. Gu X., Gopalakrishna T. Y., Phan H., Ni Y., Herng T. S., Ding J., Wu J., Angew. Chem. Int. Ed. Angew. Chemie Int. Ed. 2017, 56, 15383–15387. [DOI] [PubMed] [Google Scholar]
- 12. Ayalon A., Rabinovitz M., Tetrahedron Lett. 1992, 33, 2395–2398. [Google Scholar]
- 13.CCDC https://www.ccdc.cam.ac.uk/services/structures?id=doi:10.1002/open.201800006 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from http://www.ccdc.cam.ac.uk/.
- 14.
- 14a. Kruszewski J., Krygowski T. M., Tetrahedron Lett. 1972, 13, 3839–3842; [Google Scholar]
- 14b. Krygowski T. M., Cyrański M. K., Chem. Rev. 2001, 101, 1385–1419. [DOI] [PubMed] [Google Scholar]
- 15. Sapir M., Donckt E. V., Chem. Phys. Lett. 1975, 36, 108–110. [Google Scholar]
- 16. Nakai Y., Mori T., Inoue Y., J. Phys. Chem. A 2012, 116, 7372–7385. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Supplementary
