Dimeric boron complexes bearing isoquinolyl-pyrrole ligands

Abstract

In the present study, a novel series of dimeric boron complexes bearing 2-(isoquinol-1-yl)pyrrole ligands were synthesized and their optical properties were investigated. The linkage positions of several isoquinoline and pyrrole rings facilitated the control of their molecular orbital energy levels and structural rigidities, thereby determining their characteristic optical properties. Computational studies revealed the detailed relationship between structures and optical properties.

---

Boron dipyrromethene (BODIPY)¹ is one of the most valuable organic fluorochromes because of its high molar extinction coefficient and fluorescence emission efficiency. Consequently, BODIPYs have been explored in material and biological chemistry for various applications such as organic light-emitting devices,² organic lasers,³ solar cells,⁴ molecular probes,⁵ labelling reagents,⁶ imaging,⁷ and drug delivery.⁸ Moreover, they undergo diverse molecular modifications to control their optical properties for various applications. By contrast, BODIPY dimers have been investigated as a powerful approach for achieving BODIPYs exhibiting unique optical properties, such as bathochromically shifted absorption and fluorescence bands, enlarged Stokes shifts (Δ), and intersystem crossing.^9–11 In particular, the optical properties of directly connected BODIPY dimers were substantially controlled by their linkage positions, such as α-, β-, γ-, and meso-positions (Fig. 1A), exerting considerable impacts on their electronic properties and twisted structures. However, BODIPYs are not advantageous in both structural diversity and control of electronic properties due to the limited linkage position on the highly symmetric dipyrromethene. Therefore, novel boron complex cores, which can generate more diverse structures, are required to be explored.

Fig. 1 Chemical structures of (A) BODIPY, (B) an asymmetric boron complex bearing isoquinolylpyrrole ligands, and (C) dimer compounds iQnD and PD in the present work.

Asymmetric ligands comprising various heterocyclic compounds have emerged as attractive candidates for expanding the structural diversities of boron complexes. Furthermore, studies have revealed that controlling the electronic characteristics, rigidities, and packing structures of asymmetric boron complexes can impart the complexes with unique optical properties.^12,13 In addition, our previous systematic studies on asymmetric boron complexes bearing (iso)quinolinylpyrrole ligands revealed that their optical properties can be substantially controlled by slightly changing their π-conjugated structures.¹⁴ In particular, a 2-(isoquinol-1-yl)pyrrole-ligand-bearing monomeric boron complex 1 (Fig. 1B) exhibited remarkable optical properties, such as a large Δ, and highly efficient emission both in solution and in the solid state despite its simple structure. We also elucidated the dependence of the optical properties of 1 on the position of its aryl substituents.¹⁵ These results revealed that 1 could serve as a core for developing photofunctional materials. Accordingly, we synthesized dimeric boron complexes bearing 2-(isoquinol-1-yl)pyrrole ligands (Fig. 1C) and investigated their optical properties. Our findings revealed that their optical properties depended on the linkage position. In particular, the isoquinoline–isoquinoline-linked boron complex dimer (iQnD) and pyrrole–pyrrole-linked boron complex dimer (PD) exhibited substantially different properties. Furthermore, these interesting structure–optical property relationships were investigated via computational studies (time-dependent density functional theory, TD-DFT).

Here, the iQnDs were synthesized via a Suzuki–Miyaura cross-coupling reaction (Scheme 1A). First, pyrrolylisoquinoline 2a bearing a chlorine atom at the C4 position of the isoquinoline ring was synthesized, following our previous report (for detailed structures, see ESI†). Next, it was subjected to Miyaura borylation with a Buchwald-type ligand in the presence of a Pd catalyst to yield the boronate ester 3a. Thereafter, the obtained 2a and 3a were subjected to the Suzuki–Miyaura cross-coupling reaction to produce a precursor of the dimeric ligand 4a. Finally, NaOMe-driven N-Boc deprotection, followed by boron complexation with BF₃·OEt₂ were performed in the presence of triethylamine to give iQ4D in 20% yield from 2a. Similarly, dimeric boron complexes iQ5D, iQ6D, and iQ7D were synthesized from pyrrolylisoquinoline 2b, 2c, and 2d, respectively. The dimer structure of iQ5D was confirmed by single crystal X-ray structure analysis. A single crystal of iQ5D from DCM/cyclohexane yielded the structure shown in Scheme 1A. Interestingly, iQ5D has two kinds of boron complex part, namely a planar unit (blue structure) and distorted unit (green structure), respectively (Fig. S1A–C, ESI†). Although their bond lengths and angles are slightly different (Tables S2 and S3, ESI†), they are almost equal to those of monomer 1.¹³ These structural differences between the boron complex units could be ascribed to a favored packing structure. Actually, the packing structure was formed via π–π interactions of planar units with each other and distorted units with each other (Fig. S1D, ESI†). The isoquinoline rings of iQ5D were linked almost orthogonally; the dihedral angle was ca. 82° (Fig. S1E, ESI†). Although we attempted the synthesis of PD based on the Suzuki–Miyaura cross-coupling reaction, the process was not efficient owing to the instability of the isoquinolylpyrrole-boronic acid ester toward water (for the detailed synthetic route, see ESI†). Therefore, 5 and 6 were subjected to the Negishi cross-coupling reaction (Scheme 1B) to obtain the desired precursor of the dimeric ligand 7 in a 92% yield. Finally, PD was obtained via N-Boc deprotection, followed by boron complexation in 6% yield.

Scheme 1 Synthesis of dimeric boron complexes (A) iQnD with an inset ORTEP drawing of iQ5D, and (B) PD.

To clarify the optical properties of iQnD and PD, their ultraviolet-visible (UV-Vis) absorption and fluorescence emission spectra were measured (Fig. 2). A solution of iQ4D exhibited an absorption band at a wavelength (λ^abs_max) of 455 nm (Fig. 2A, yellow line) and a fluorescence emission peak at λ^em = 513 nm postexcitation (Fig. 2B, yellow line). They are more bathochromically shifted in comparison with those of 1 (Fig. 2A and B, black broken line, respectively) due to the expanded π-skeleton; λ^abs_max = 455 (iQ4D) > 415 nm (1), λ^em = 513 (iQ4D) > 481 nm (1). Although iQ5D, iQ6D, and iQ7D in DCM also exhibited more bathochromically shifted absorption bands (Fig. 2A, blue, green, and purple lines, respectively) than 1, their redshift values were lower than that of iQ4D; λ^abs_max = 430 (iQ5D), 435(iQ6D), and 425 nm (iQ7D). Furthermore, although the λ^abs_max values of iQ5D, iQ6D, and iQ7D were similar, their λ^em values substantially differed, resulting in different emitting colors, from turquoise blue to light green (Fig. 2B, blue, green, and purple lines, respectively);¹⁶λ^em = 487 (iQ5D), 536 (iQ6D), and 503 nm (iQ7D). Notably, iQnD exhibited a high fluorescence quantum yield (Φ_F) despite their large Δ values (Table 1). Moreover, PD in DCM exhibited a drastically redshifted absorption band at 553 nm (Fig. 1A, red line). Although the fluorescence emission peak of PD in DCM (λ^em = 593 nm) was also redshifted compared with those of iQnD, its Δ value remarkably decreased (Δ = 1220 cm⁻¹). In addition, its Φ_F was 0.51. Although iQnD and PD also exhibited fluorescence even in the solid state (Fig. 2C), their Φ_F values were lower than those obtained in solution probably because of the expanded π-systems, which promoted quenching through π–π interactions.

Fig. 2 (A) UV-Vis spectra of 1, iQnD, and PD in DCM; (B) fluorescence emission spectra of 1, iQnD, and PD in DCM; and (C) fluorescence emission spectra of 1, iQnD, and PD in the solid state.

Table 1 Optical properties of 1, iQnD, and PD

	State	λ ^absmax/nm (ε/10^4 M^−1 cm^−1)	λ ^em/nm	Δ/cm^−1	Φ F
a Data from our previous report (ref. 14a).
1	in DCM	415 (1.72)	481	3310	0.90^a
	Solid	—	521	—	0.34^a
iQ4D	in DCM	455 (4.17)	513	2480	0.70
	Solid	—	630	—	0.02
iQ5D	in DCM	430 (4.08)	487	2720	0.99
	Solid	—	600	—	0.12
iQ6D	in DCM	435 (2.89)	536	4330	0.74
	Solid	—	634	—	0.01
iQ7D	in DCM	425 (2.65)	503	3650	0.67
	Solid	—	575	—	<0.01
PD	in DCM	553 (4.23)	593	1220	0.51
	Solid	—	662	—	nd

---

To further investigate the optical properties of the synthesized dimeric boron complexes, TD-DFT (TD-SCF/B3LYP/6-31+G(d,p)) calculations were performed to estimate their π–π* transitions. As shown in Fig. 3, the energy levels of the highest-occupied molecular orbital (HOMO) and lowest-unoccupied molecular orbital (LUMO) of monomer 1 were −5.99 and −2.65 eV, respectively. Furthermore, the lowest energy band of 1 resulted from the HOMO–LUMO transition (Table 2): E(1) = 415 (in DCM) and 409 nm (TD-DFT calc.). Conversely, the energy diagram of the iQnDs revealed that their HOMO−1 and HOMO as well as LUMO and LUMO+1 exhibited similar energy levels (Fig. 3, blue area). These similarities were ascribed to the degeneracy of the HOMO and LUMO, respectively, derived from the independent boron complex unit. Therefore, the HOMO−1 and LUMO+1 of iQnD were involved in their lowest energy bands (Table 2). On the whole, the energy levels of the HOMO−1, HOMO, LUMO, and LUMO+1 of iQnD decreased compared with those of 1. Notably, the decreasing degrees of its LUMO and LUMO+1 energy levels were higher than that of its HOMO and HOMO−1 levels, resulting in the redshift of the absorption band.¹⁷ Their oscillator strengths (f) generally correlated with its molar extinction coefficient. iQ6D exhibited the most favored transition from HOMO to LUMO+2 as shown in Table 2. This calculation result is consistent with the measured UV-vis spectrum, which showed a large peak at 339 nm (Fig. 2A, green line). Furthermore, the TD-DFT calculations revealed that the lowest energy band of PD resulted from its HOMO–LUMO transition and that its HOMO level increased remarkably compared with that of 1 (Fig. 3, red area): HOMO = −5.22 (PD) > −5.99 eV (1). The HOMO and HOMO−1 orbitals of PD indicated the strong interaction between the localized spatial distributions on the pyrrole rings (Fig. S5, ESI†). This interaction led to the resolving of the degeneracy, which resulted in an enhanced energy level of the HOMO. This increased HOMO level facilitated the significant redshift of the absorption band: E(PD) = 553 (in DCM) and 558 nm (TD-DFT calc.). Thus, these calculations revealed that the isoquinoline–isoquinoline and pyrrole–pyrrole linkages could effectively modulate the LUMO and HOMO levels, respectively, and these properties could be ascribed to the electronic properties of monomer 1 whose LUMO and HOMO spatial distributions were eccentrically located on the isoquinoline and pyrrole rings, respectively.

Fig. 3 Energy diagrams of 1, iQnD, and PD by TD-DFT calculations.

Table 2 Experimental and theoretical absorption wavelengths of 1, iQnD, and PD, and transition assigned by TD-DFT calculations

	λ ^in DCM/nm (ε/10^4 M^−1 cm^−1)	λ ^TD-DFT	Assignment
1	415 (1.72)	409 (f = 0.292)	HOMO to LUMO
iQ4D	455 (4.17)	432 (f = 0.173)	HOMO to LUMO+1
		423 (f = 0.608)	HOMO−1 to LUMO
iQ5D	430 (4.08)	425 (f = 0.691)	HOMO to LUMO+1
iQ6D	435 (2.89)	467 (f = 0.280)	HOMO to LUMO
	339 (3.88)	326 (f = 0.517)	HOMO to LUMO+2
iQ7D	425 (2.65)	445 (f = 0.387)	HOMO to LUMO
PD	553 (4.23)	558 (f = 1.05)	HOMO to LUMO

---

As aforementioned, PD exhibited a smaller Δ (1220 cm⁻¹) than 1 and iQnD (ca. 3000–4000 cm⁻¹). In addition, the full width at half maximum of its emission band (2010 cm⁻¹) was smaller than those of 1 and iQnD (ca. 2500–3000 cm⁻¹). As these results indicated the structural rigidity of PD, we performed computational studies to confirm the structural properties of the corresponding boron complexes. Our previous reports revealed that the rigidities of cyclic structures comprising coordinate bonds (N and B) remarkably affects their fluorescence emission properties.¹³ Therefore, the bond lengths of the five-membered structure containing the N and B atoms of 1 and PD in the ground state (S₀ state, DFT/B3LYP/6-31+G(d,p)) and excited state (S₁ state, TD-SCF/B3LYP/6-31+G-(d,p)) were calculated. The changes in the bond lengths of PD between the S₀ and S₁ states were remarkably smaller than those of 1 (Table S5, ESI†). These calculations revealed that the structural relaxation in PD was suppressed to ensure a smaller Δ value. Furthermore, regarding the optimized structure of PD in the S₀ state, the distance between the F atom of BF₂ and the H atom on the β-position of pyrrole was 2.34 Å, which allowed the F⋯H interactions that suppressed the free rotation around the single C–C bond linking the boron complexes (Fig. S6, ESI†). This result also supported the PD rigidity to suppress its structural relaxation.

In conclusion, we synthesized a novel series of dimeric boron complexes bearing 2-(isoquinol-1-yl)pyrrole ligands. The optical properties of iQnD and PD were mainly controlled by decreasing and increasing their LUMO and HOMO energy levels, respectively. Furthermore, our TD-DFT-based computational studies revealed that PD exhibited a remarkably enhanced structural rigidity, which could be crucial for preparing highly efficient fluorescent materials with high color purity in the long-wavelength range. These distinct changes in the optical properties owing to the linkage position were induced by the asymmetric structures comprising heterogeneous heterocycles. This approach could be efficient for developing photofunctional materials. Further syntheses, detailed optical properties, and application studies will be reported in the near future.

This work was supported by JSPS KAKENHI Grant Number JP21K14613 and the 2023 DIC Award in Synthetic Organic Chemistry, Japan.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

Data for this article, including synthetic details, NMR and fluorescence spectra, X-ray crystallographic analysis, electrochemical properties, and calculation details, are available in the ESI.†

Notes and references

1. (a) A. Loudet and K. Burgess, Chem. Rev., 2007, 107, 4891–4932; (b) N. Boens, B. Verbelen and W. Dehaen, Eur. J. Org. Chem., 2015, 6577–6595.
2. M. Chapran, E. Angioni, N. J. Findlay, B. Breig, V. Cherpak, P. Stakhira, T. Tuttle, D. Volyniuk, J. V. Grazulevicius, Y. A. Nastishin, O. D. Lavrentovich and P. J. Skabara, ACS Appl. Mater. Interfaces, 2017, 9, 4750–4757.
3. (a) I. García-Moreno, F. Amat-Guerri, M. Liras, A. Costela, L. Infantes, R. Sastre, F. L. Arbeloa, J. B. Prieto and Í. L. Arbeloa, Adv. Funct. Mater., 2007, 17, 3088–3098; (b) E. Palao, G. Duran-Sampedro, S. D. L. Moya, M. Madrid, C. García-López, A. R. Agarrabeitia, B. Verbelen, W. Dehaen, N. Boens and M. J. Ortiz, J. Org. Chem., 2016, 81, 3700–3710.
4. (a) S. P. Singh and T. Gayathri, Eur. J. Org. Chem., 2014, 4689–4707; (b) M. A. Filatov, Org. Biomol. Chem., 2020, 18, 10–27.
5. J. Zhang, N. Wang, X. Ji, Y. Tao, J. Wang and W. Zhao, Chem. – Eur. J., 2020, 26, 4172–4192.
6. (a) J.-S. Lee, N.-y Kang, Y. K. Kim, A. Samanta, S. Feng, H. K. Kim, M. Vendrell, J. H. Park and Y.-T. Chang, J. Am. Chem. Soc., 2009, 131, 10077–10082; (b) X. Zhang, Y. Xiao, J. Qi, J. Qu, B. Kim, X. Yue and K. D. Belfield, J. Org. Chem., 2013, 78, 9153–9160.
7. T. Kowada, H. Maeda and K. Kikuchi, Chem. Soc. Rev., 2015, 44, 4953–4972.
8. (a) A. Sharma, A. Khatchadourian, K. Khanna, R. Sharma, A. Kakkar and D. Maysinger, Biomaterials, 2011, 32, 1419–1429; (b) F. Rancan, A. Todorova, S. Hadam, D. Papakostas, E. Luciani, C. Graf, U. Gernert, E. Rühl, B. Verrier, W. Sterry, U. Blume-Peytavi and A. Vogt, Eur. J. Pharm. Biopharm., 2012, 80, 76–84.
9. Direct linked BODIPY dimers; (a) A. B. Nepomnyashchii, M. Bröring, J. Ahrens and A. J. Bard, J. Am. Chem. Soc., 2011, 133, 8633–8645; (b) A. B. Nepomnyashchii, M. Bröring, J. Ahrens and A. J. Bard, J. Am. Chem. Soc., 2011, 133, 19498–19504; (c) Z. Kang, F. Lv, Q. Wu, H. Li, Z. Li, F.-X. Wu, Z. Wang, L. Jiao and E. Hao, Org. Lett., 2021, 23, 7986–7991; (d) Z. Kang, Q. Wu, X. Guo, L. Wang, Y. Ye, C. Yu, H. Wang, E. Hao and L. Jiao, Chem. Commun., 2021, 57, 9886–9889; (e) L. Wang, Q. Wu, Z. Kang, X. Guo, W. Miao, Z. Li, H. Zuo, H. Wang, H. Si, L. Jiao and E. Hao, Org. Lett., 2023, 25, 5055–5060.
10. π-System linker inserted BODIPY dimers; (a) Y. Cakmak and E. U. Akkaya, Org. Lett., 2009, 11, 85–88; (b) S. Saino, M. Saikawa, T. Nakamura, M. Yamamura and T. Nabeshima, Tetrahedron Lett., 2016, 57, 1629–1634.
11. π-Conjugated cyclic structure fused BODIPY dimers; (a) C. Yu, L. Jiao, T. Li, Q. Wu, W. Miao, J. Wang, Y. Wei, X. Mu and E. Hao, Chem. Commun., 2015, 51, 16852–16855; (b) Q. Gong, Q. Wu, X. Guo, H. Li, W. Li, C. Yu, E. Hao and L. Jiao, Org. Lett., 2021, 23, 7661–7665.
12. (a) Y. Ren, X. Liu, W. Gao, H. Xia, L. Ye and Y. Mu, Eur. J. Inorg. Chem., 2007, 1808–1814; (b) J. F. Araneda, W. E. Piers, B. Heyne, M. Parvez and R. McDonald, Angew. Chem., Int. Ed., 2011, 50, 12214–12217; (c) K. Yamamoto, W. Imai, S. Kanamori, K. Yamamoto and Y. Nakamura, J. Org. Chem., 2023, 88, 4003–4007.
13. (a) C. Yu, Y. Sun, L. Jiao and E. Hao, Synlett, 2024, 37–54; (b) C. Jin, X. Yang, W. Zhao, Y. Zhao, Z. Wang and J. Tan, Coord. Chem. Rev., 2024, 513, 215892.
14. (a) T. Nakano, A. Sumida and K. Naka, J. Org. Chem., 2021, 86, 5690–5701; (b) T. Nakano, A. Sumida and K. Naka, Eur. J. Inorg. Chem., 2021, 3148–3157.
15. (a) T. Nakano and S. Fujikawa, J. Org. Chem., 2022, 87, 11708–11721; (b) T. Nakano and S. Fujikawa, ChemPhotoChem, 2024, 8, e202300189.
16. Photographs showing the emission colors of the boron complexes are exhibited in Fig. S2 and S3, ESI†.
17. The similarity of the HOMO levels among 1 and iQnD was also exhibited by electrochemical measurement in Fig. S4 and Table S4, ESI†.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 2456352. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d5cc03519h

---