Synthesis of substituent-free dioxadiaza[8]circulene to investigate intermolecular interactions and photophysical properties

Aoi Nakagawa a, Wataru Ota b, Takumi Ehara c, Yusuke Matsuo a, Kiyoshi Miyata c, Ken Onda c, Tohru Sato *ad, Shu Seki *a and Takayuki Tanaka *a
aDepartment of Molecular Engineering, Graduate School of Engineering, Kyoto University, Kyotodaigakukatsura, Nishikyo-ku, Kyoto 615-8510, Japan. E-mail: tanaka@moleng.kyoto-u.ac.jp
bMOLFEX, Inc., 34-4 Takano Nishibiraki-cho, Sakyo-ku, Kyoto 606-8103, Japan
cDepartment of Chemistry, Faculty of Science, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan
dFukui Institute for Fundamental Chemistry, Kyoto University, 3, 4-4 Takano Nishibiraki-cho, Sakyo-ku, Kyoto 606-8103, Japan

Received 17th October 2024 , Accepted 6th November 2024

First published on 6th November 2024


Abstract

Peripherally unsubstituted dioxadiaza[8]circulene, as the first example of structurally identified pristine hetero[8]circulene, was synthesized by the substituent detachment reactions. The solid-state structures and photophysical properties were analysed to elucidate intermolecular interactions. Herzberg–Teller type emission was considered to explain the optical behavior.


Hetero[8]circulenes have recently been studied as unique heteroatom-incorporated polycyclic aromatic molecules due to their roughly planar structures, square-shaped geometry and functionality of the heteroatoms embedded at the periphery.1 Several functionalized hetero[8]circulenes especially those including furan, pyrrole and/or thiophene units were reported as novel electronic and optical materials.2–4 In most cases, however, hetero[8]circulenes have been decorated with peripheral substituents to increase solubility and/or to prevent crucial aggregation. The aggregation behaviours of tetraoxa[8]circulene and tetrathia[8]circulene have indeed been studied in detail,5,6 while tetraaza[8]circulenes bearing benzo-annulated structures were revealed to show hydrogen-bonding interaction between the outer-pointing NH group and solvent molecule as an inevitable factor to determine solid-state packing.7 This improves net solubility in hydrogen-bond accepting solvents such as THF and DMSO, hence no peripheral substituents are necessarily attached in the molecular design.

We became interested in the relationship between the number of NH sites and intermolecular interactions, that is, solvent effects on the solid-state structure and molecular packing. In our earlier reports, tetraaza[8]circulene (1) and oxatriaza[8]circulene (2) bearing four and two benzo moieties, respectively, were soluble in THF, DMSO and acetone, being allowed to be characterized by NMR, UV/Vis absorption and emission spectra and X-ray diffraction analysis (Fig. 1).7 Dithiadiaza[8]circulene (3) with two NH sites and four benzo moieties were only soluble in DMSO.8 Pittelkow and co-workers have recently reported the synthesis of hetero[8]circulene 4 including two nitrogen, one oxygen and one sulfur atoms embedded, while its solid-state structure and solvation aptitudes were not studied in detail.9 Noteworthy to say, a pristine tetraoxa[8]circulene (5) was obtained by detachment of t-butyl substituents, but the properties of 5 have been elusive mainly because of its insolubility in organic solvents.5 To complete the systematic assessments, we envisioned that pristine dioxadiaza[8]circulene (6) would be an important target to investigate intermolecular interactions and photophysical properties. In the literature, the skeleton of dioxadiaza[8]circulene is constructed by acid-catalyzed oxidative dimerization of 3,6-dihydroxycarbazole derivatives in good yields10 and this methodology, along with the substituent detachment protocol, would give rise to substituent-free dioxadiaza[8]circulene. In this paper, the synthesis, solid-state structure, aggregation behavior in solution, and photophysical properties were studied with a focus on the substituent effects at the periphery. To this end, Herzberg–Teller type emission is first proposed for hetero[8]circulenes.


image file: d4cc05539j-f1.tif
Fig. 1 Benzo-annulated and substituent-free hetero[8]circulenes. Hydrogen bonding mode between the NH site and solvent molecule is shown aside. The substituent detachment strategy to afford dioxadiaza[8]circulene 6 is indicated in the bottom right.

According to the report on the synthesis of dioxadiaza[8]circulene 6-Pr-tBu reported by Pittelkow,10 the synthesis of 6-PE-tBu, which is an N-phenylethyl substituted variant of 6-Pr-tBu, was conducted in 6 steps in 18% overall yields from carbazole (Scheme 1). Phenylethyl groups are employed as the substituent at the N-sites since base-promoted removal of this unit has been reported in the literature.11 Single crystals were obtained by slow-vapor diffusion of n-hexane into the solution of 6-PE-tBu in dichloromethane, and the structure has been revealed by X-ray diffraction (XRD) analysis (Fig. 2). The C–C bonds consisting of the eight-membered ring are in the range of 1.402–1.436 Å. Among them, the C–C bonds of phenyl moieties (C2–C3, C4–C5, C6–C7 and C8–C1) are shorter than those of tetrabenzotetraaza[8]circulene 1 (1.428–1.438 Å).7a The NICS(0) value at the center of the eight-membered ring is +8.47 ppm, being indicative of decent antiaromaticity (Fig. S8-3, ESI). The degree of the paratropicity is slightly larger than that of 1 (+6.82 ppm).7a


image file: d4cc05539j-s1.tif
Scheme 1 Synthesis of 6-PE-tBu.

image file: d4cc05539j-f2.tif
Fig. 2 (a) X-Ray crystal structure and selected bond lengths of 6-PE-tBu. The thermal ellipsoids were scaled to 50% probability level. (b) Packing diagram.

Detachment of the peripheral substituents of 6-PE-tBu was conducted by the addition of potassium hexamethyldisilazide to a solution of 6-PE-tBu in 1,4-dioxane at 120 °C to give 6-tBu in 84% yield (Scheme 2). The solubility of 6-tBu is good in many organic solvents such as CHCl3 and THF. Notably, detachment of the t-butyl groups from 6-PE-tBu gave an insoluble product presumably due to the lack of both NH sites and bulky substituents. Next, retro-Friedel–Crafts reaction was conducted for 6-tBu in 1,2-dichlorobenzene/toluene, which resulted in the complete removal of t-butyl groups in 20 min to afford target molecule 6 in 75% yield. Similarly to the case of 1 and 2, 6 is soluble in hydrogen-bond accepting solvents such as THF and DMSO. The 1H NMR spectrum of 6 in DMSO-d6 exhibited doublet signals at 7.81 and 7.91 ppm as well as a broad peak due to the NH proton at 11.93 ppm. Interestingly, in the mixture of 1,2-dichlorobenzene-d4 and THF-d8, a peak attributable to the NH proton is upfield-shifted depending on the temperature from 9.87 ppm at room temperature to 8.90 ppm at 373 K. Since other peaks do not change significantly, this peak shift is due to the dissociation of solvent molecules (Fig. 3).


image file: d4cc05539j-s2.tif
Scheme 2 Synthesis of 6.

image file: d4cc05539j-f3.tif
Fig. 3 Variable-temperature 1H NMR spectra of 6 in the mixture of 1,2-dichlorobenzene-d4 and THF-d8 (v/v = 9/1).

Single crystals were obtained by slow-vapor diffusion method and the solid-state structures of 6-tBu and 6 have been elucidated by XRD measurement (Fig. 4a and ESI). The structures of the [8]circulene cores are essentially the same, featuring the C–C bonds consisting of the eight-membered ring in the range of 1.404–1.438 Å. The packing diagram of 6-tBu is different from that of 6-PE-tBu (Fig. S6-2, ESI). While the crystal of 6-tBu contains solvent molecules (THF) in the asymmetric unit, 6-PE-tBu is packed without any solvent molecules. For 6, no solvent molecules are included in the crystal lattice, even though the single crystals were grown from THF. In fact, 6 is densely packed into a dimeric (sandwich) herringbone form with its π–π distance of 3.41 Å and the density of crystals is larger (1.59 g cm−3) than that of 6-PE-tBu (1.18 g cm−3) (Fig. 4b). This packing pattern is similar to that of porphine, an unsubstituted porphyrin core, implying the structural resemblance.12 It is noteworthy to remark that 6 is the first example of structurally well-identified hetero[8]circulenes without any peripheral substituents and benzo-fused structures.


image file: d4cc05539j-f4.tif
Fig. 4 (a) X-Ray crystal structure and bond lengths of 6. The thermal ellipsoids were scaled to 50% probability level. (b) Packing diagram.

The aggregation mode is analysed in their powder state prepared by drying the corresponding THF solutions. The PXRD patterns do not match well with simulated ones calculated from crystal structures for 6-tBu and 6 (Fig. S5-4, ESI). IR spectra were measured for the same powder samples (Fig. S5-1, ESI). Generally, peaks around 3500 cm−1 are ascribable to NH vibration, and those around 3000 cm−1 are C–H vibration of t-butyl groups. Indeed, 6-tBu exhibited a sharp peak at 3468 cm−1, indicating that the NH site was not hydrogen bonded to any hydrogen bond acceptor, being consistent with the PXRD analysis. Interestingly, 6 exhibited a combination of sharp and broad bands around 3400 cm−1, while the C–H vibration peaks around 3000 cm−1 were not observed. These results indicate that the powder sample of 6 contains at least two aggregation modes.13

Thermogravimetric analysis exhibited a 3 step weight loss around 57, 225 and 436 °C for 6-tBu due to solvent evaporation, while 6-PE-tBu and 6 did not show such a behavior (Fig. S6-6, ESI).

The UV/Vis absorption spectra of 6-PE-tBu, 6-tBu and 6 measured in THF displayed two major bands around 270 nm and 350–400 nm (Fig. 5). The absorption spectrum of 6-PE-tBu is almost identical to that of 6-Pr-tBu,10 while those of 6-tBu and 6 are slightly blue-shifted and sharpened. The emission bands were observed in the range of 400–600 nm with their quantum yields of ΦF = 0.22 for 6-PE-tBu, 0.17 for 6-tBu and 0.20 for 6. The shapes of the emission bands look complicated, especially for 6 (vide infra). The emission lifetimes (τF) obtained by time-correlated single photon counting (TCSPC) technique were estimated by first-order kinetic fitting to be 9.99 ns for 6-tBu and 10.18 ns for 6, while estimated values of τF = 3.79 and 8.17 ns for 6-PE-tBu was obtained by second-order kinetic fitting (Fig. S7-1, ESI). We further recorded time-resolved photoluminescence (TRPL) spectra for 6, which revealed that each emission band was not resolved within 2 ns regime (Fig. S7-2, ESI). This implies that the emission comes from single chromophore, excluding the possibility of excimer-type emission.14


image file: d4cc05539j-f5.tif
Fig. 5 UV/Vis absorption and emission spectra of (a) 6-PE-tBu, (b) 6-tBu and (c) 6 in THF. λex = 385 nm.

To gain insights about the photophysical properties, density functional theory (DFT) calculation was conducted at the level of B3LYP/6-311G(d,p). The optimized structure of 6 is almost identical to the one in the solid state (Fig. S8-2, ESI), and the selected molecular orbital (MO) diagrams and Kohn–Sham orbital representations are shown in Fig. 6a. Accordingly, the absorption bands in the range of 350–400 nm consist of the transitions from the HOMO, HOMO−1, HOMO−2 and HOMO−3 to the LUMO. Among them, those from HOMO−1 (au) and HOMO−2 (B1u) to the LUMO (B1u) are forbidden transitions located at 420 and 387 nm, respectively. The HOMO to LUMO transition appears at 390 nm with the oscillator strength of f = 0.24. This situation drove us to consider Herzberg–Teller emission in which the forbidden S1 emission can be partially allowed by intensity borrowing of the S2 state.15 Further, the complex emission bands in 6 can be fitted by considering the existence of molecules having hydrogen bonding with THF at the NH sites (Fig. S8-7, ESI).16 The simulated emission spectrum in Fig. 6b is in good accordance with the experimental one measured in THF when the calculated spectra with and without THF molecules are superimposed. To prove the solvent effect, we measured the absorption and emission spectra in a variable ratio of THF and CH2Cl2 (Fig. 6c and Fig. S7-5, ESI). As the content of CH2Cl2 increased, the emission spectra merged into a typical one with clear vibronic bands. Although essentially the same assignments can be made for 6-tBu, the spectral features are more complex due to the perturbation by peripheral substituents. Thus, the substituent-free, pristine dioxadiaza[8]circulene 6 offered a significant basis to unravel detailed photophysical properties of hetero[8]circulenes.


image file: d4cc05539j-f6.tif
Fig. 6 (a) Selected MO diagrams for 6. (b) Experimental (red) and calculated fluorescence spectrum for 6 with (dashed line) and without (solid line) THF molecules at T = 298 K. The broadening of the density of final vibronic states was expressed using the Gaussian function with a linewidth of 200 cm−1. (c) Emission spectra of 6 in a mixture of THF and CH2Cl2. λex = 385 nm.

In summary, substituent-free hetero[8]circulene 6 was successfully synthesized by the detachment of N-phenylethyl groups and t-butyl groups from 6-PE-tBu in good yields in 2 steps. The solid-state packing structures were analysed by XRD and IR spectra, showing that the included solvent molecules hydrogen bonded to the NH sites were easily evaporated in the powder state. However, this solvent coordination behaviour somehow helps to increase solubility and no aggregation effects were observed at least in the optical spectroscopy regime. Despite its decent antiaromaticity, 6 exhibit emission with a moderate quantum efficiency. The observed emission spectra in THF can be simulated by considering Herzberg–Teller type emission and superposition of THF coordinated and non-coordinated ones. Since similar solvent coordination behaviours were known in pentaaza[10]circulene and aza[n]helicenes,17 this unusual solvent effect will be further investigated in the future.

This work was supported by JSPS KAKENHI Grant Numbers JP22H00314, JP22K05253, JP23K23425, JP23K27296 and JP23K17942 for Kyoto University and JP20H05676, JP22H02159, JP23H03833, JP23H01977, JP23H04631 and JP23K20039 for Kyushu University Platform of Inter-/Transdisciplinary Energy Research (Q-PIT) Module-Research Program and by JST SPRING, Grant Number JPMJSP2110. Numerical calculations were partly performed at Information Initiative Center, Hokkaido University.

Data availability

The data supporting this article have been included in the ESI.

Conflicts of interest

There are no conflicts to declare.

Notes and references

  1. (a) T. Hensel, N. N. Andersen, M. Plesner and M. Pittelkow, Synlett, 2016, 498 Search PubMed; (b) Y. Miyake and H. Shinokubo, Chem. Commun., 2020, 56, 15605 RSC.
  2. C. B. Nielsen, T. Brock-Nannestad, T. K. Reenberg, P. Hammershøj, J. B. Christensen, J. W. Stouwdam and M. Pittelkow, Chem. – Eur. J., 2010, 16, 13030 CrossRef.
  3. S. K. Pedersen, V. B. R. Pedersen, F. S. Kamounah, L. M. Broløs, G. V. Baryshnikov, R. R. Valiev, K. Ivaniuk, P. Stakhira, B. Minaev, N. Karaush-Karmazin, H. Ågren and M. Pittelkow, Chem. – Eur. J., 2021, 27, 11609 CrossRef.
  4. V. B. R. Pedersen, S. K. Pedersen, Z. Jin, N. Kofod, B. W. Laursen, G. V. Baryshnikov, C. Nuckolls and M. Pittelkow, Angew. Chem., Int. Ed., 2022, 61, e202212293 CrossRef PubMed.
  5. T. Brock-Nannestad, C. B. Nielsen, M. Schau-Magnussen, P. Hammershøj, T. K. Reenberg, A. B. Petersen, D. Trpcevski and M. Pittelkow, Eur. J. Org. Chem., 2011, 6320 CrossRef.
  6. H. Murase, Y. Nagata, S. Akahori, H. Shinokubo and Y. Miyake, Chem. – Asian J., 2020, 15, 3873 CrossRef.
  7. (a) F. Chen, Y. S. Hong, S. Shimizu, D. Kim, T. Tanaka and A. Osuka, Angew. Chem., Int. Ed., 2015, 54, 10639 CrossRef PubMed; (b) Y. Morimoto, F. Chen, Y. Matsuo, K. Kise, T. Tanaka and A. Osuka, Chem. – Asian J., 2021, 16, 648 CrossRef; (c) Y. Matsuo, F. Chen, K. Kise, T. Tanaka and A. Osuka, Chem. Sci., 2019, 10, 11006 RSC.
  8. Y. Matsuo, T. Tanaka and A. Osuka, Chem. – Eur. J., 2020, 26, 8144 CrossRef.
  9. B. Lousen, S. K. Pedersen, P. Bols, K. H. Hansen, M. R. Pedersen, O. Hammerich, S. Bondarchuk, B. Minaev, G. V. Baryshnikov, H. Ågren and M. Pittelkow, Chem. – Eur. J., 2020, 26, 4935 CrossRef PubMed.
  10. T. Hensel, D. Trpcevski, C. Lind, R. Grosjean, P. Hammershøj, C. B. Nielsen, T. Brock-Nannestad, B. E. Nielsen, M. Schau-Magnussen, B. Minaev, G. V. Baryshnikov and M. Pittelkow, Chem. Eur. J., 2013, 19, 17097 CrossRef PubMed.
  11. (a) S. Akahori, A. Kaga, J. Kim, H. Yorimitsu, D. Kim, H. Shinokubo and Y. Miyake, Chem. – Asian J., 2022, 17, e202200244 CrossRef PubMed; (b) A. Kaga, K. Nogi and H. Yorimitsu, Chem. Eur. J., 2019, 25, 14780 CrossRef PubMed.
  12. L. E. Webb and E. B. Fleischer, J. Chem. Phys., 1965, 43, 3100 CrossRef.
  13. The powder sample of 6 was also prepared by dropping a THF solution into n-hexane, in which the ratio of the new peaks observed in PXRD and IR spectra increased. Drying this sample under vacuum for two days resulted in almost no change in these spectra. See Fig. S5 and S6 in the ESI, for details.
  14. We confirmed that the absorption and emission spectra were independent on the concentration and excitation wavelength. See Fig. S6-3 and S6-6 in the ESI..
  15. (a) G. Herzberg, E. Teller and Z. Physik, Chem Abt., 1933, 21, 410 Search PubMed; (b) J. N. Murrell and J. A. Pople, Proc. Phys. Soc. A., 1956, 69, 245 CrossRef.
  16. (a) W. Ota, M. Uejima and T. Sato, Bull. Chem. Soc. Jpn., 2023, 96, 582 CrossRef; (b) W. Ota, M. Uejima, N. Haruta and T. Sato, Bull. Chem. Soc. Jpn., 2024, 97, uoad020 CrossRef.
  17. (a) Y. Matsuo, K. Kise, Y. Morimoto, A. Osuka and T. Tanaka, Angew. Chem., Int. Ed., 2022, 61, e202116789 CrossRef PubMed; (b) Y. Matsuo, M. Gon, K. Tanaka, S. Seki and T. Tanaka, J. Am. Chem. Soc., 2024, 146, 17428 CrossRef.

Footnote

Electronic supplementary information (ESI) available: Experimental and computational details, as well as X-ray crystallographic data for 6-PE-tBu, 6-tBu and 6 are available. CCDC 2346717–2346719. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4cc05539j

This journal is © The Royal Society of Chemistry 2024
Click here to see how this site uses Cookies. View our privacy policy here.