Regulation of the frontier molecular orbitals and photophysical properties of boron tropolonate complexes by regioselective functionalization

Hikari Ogoshi a, Shunichiro Ito ab and Kazuo Tanaka *ab
aDepartment of Polymer Chemistry, Graduate School of Engineering, Kyoto University, Katsura, Nishikyo-ku, Kyoto 615-8510, Japan. E-mail: tanaka@poly.synchem.kyoto-u.ac.jp
bDepartment of Technology and Ecology, Graduate School of Global Environmental Studies, Kyoto University, Katsura, Nishikyo-ku, Kyoto 615-8510, Japan

Received 23rd July 2025 , Accepted 14th August 2025

First published on 22nd August 2025


Abstract

Here, we show that the C5 substitution in α-tropolone efficiently modulates the energy level of the highest occupied molecular orbital rather than the lowest unoccupied one. Furthermore, the C5-substitution imparts functional luminescent properties, such as aggregation-induced emission, room-temperature phosphorescence, and thermally activated delayed luminescence.


Precise control of the energy levels of frontier molecular orbitals (FMOs) in π-conjugated organic molecules is crucial for achieving optimal materials in practical optoelectronic applications.1–5 We have recently proposed that π-conjugated skeletons holding “isolated FMOs” are a promising electronic structure for meeting this demand, followed by modulation of optical and electric properties.6–11 An electronic structure in which either the highest occupied molecular orbital (HOMO) or the lowest unoccupied molecular orbital (LUMO) is distributed on a skeletal carbon in π-conjugated systems is called an “isolated HOMO (LUMO)” (Fig. 1a). In particular, we have demonstrated that by aza-substitution or regioselective substitution on an isolated FMO positions, the corresponding FMO energy level can be selectively perturbed.11 Previously, isolated FMOs have been found in several frameworks. Meanwhile, upon deeper structural investigation, we suggest that the isolated-FMO scaffolds have even π-electrons on odd atoms in their C2v-symmetry π-conjugated systems, called even-electron/odd-atom systems (ee/oa), as shown in Fig. 1b. Since either the HOMO or LUMO of ee/oa C2v systems has a nodal plane passing through skeletal carbons, we assumed that these molecules could have an isolated FMO.
image file: d5cc04171f-f1.tif
Fig. 1 (a) Concept of isolated FMOs. (b) Isolated FMO scaffolds in previous reports. (c) Chemical structures of Tp and TpB, and Kohn–Sham FMOs of TpB calculated at the B3LYP/6-31G(d,p) level.

This work focuses on α-tropolone (Tp, Fig. 1c),12 a cyclic α-hydroxyketone consisting of a 10e/9a system including seven carbon and two oxygen atoms, as a representative ee/oa system. Tp and its derivatives are characterized by their intramolecular polarization thanks to their 6π-aromatic resonance structure on their seven-membered ring. It was revealed that its difluoroboron complex (TpB, Fig. 1c) has enhanced aromaticity and dramatically increased photoluminescent quantum yield compared to Tp.13 We also demonstrated that TpB works as a powerful electron acceptor and that TpB-based donor–acceptor conjugated polymers show near-infrared emission in the film state.14 Importantly, as shown in Fig. 1c, TpB has the 10e/9a C2v system followed by isolated LUMO at C4 and C6 and isolated HOMO at C5 (see SI for computational detail). Although there are some reports on the C5-selective functionalization,15–17 the optical properties of these 5-substituted TpB derivatives are still unclear. These Tp-based luminescent materials could exhibit a lower photoluminescent quantum yield (PLQY) in aggregation states than in solution states.

Herein, we evaluated the effects of the C5 substitution of TpB on its optoelectronic properties to demonstrate the selective modulation of its isolated HOMO. By regioselective functionalization of isolated HOMO, it is possible to alter the HOMO level more effectively than the LUMO level, leading to a change in the energy gap. As a result, it was found that the HOMO/LUMO gap is tunable from 3.33 to 2.66 eV. Furthermore, various luminescent properties, including aggregation-induced emission (AIE),18 excimer-like emission,19 thermally activated delayed fluorescence (TADF),20 and room-temperature phosphorescence (RTP),21 were observed. Structure–property relationships on luminescent behaviors are also studied.

We synthesized the seven 5-substituted TpB derivatives via regioselective amination and Sandmeyer-type reactions (Fig. 2a; see the SI for their synthesis) and initially evaluated their absorption and photoluminescence (PL) properties using diluted solutions (Fig. 2b, Table 1 and Fig. S1, S6). Compared to TpB, two remarkable spectral changes were observed in the absorption spectra. Firstly, it was shown that all substituents can induce the redshift of the longest onset wavelength (λonsetabs). The bands were observed in the order of PhCF3B < BrB < PhB < OMeB < IB < PhOMeB < NMe2B in the range from 394 to 466 nm. From this trend, it is roughly suggested that the stronger electron-donating group induced the larger redshift. The delicate balance between resonance (electron-donating) and inductive (electron-withdrawing) effects of the halogens might perturb the order. Secondly, the intensity ratio between the two absorption bands in TpB (305 and 364 nm) was varied by the substituent effect as represented in IB, BrB, and OMeB, and the single broad bands were detected from PhB, PhCF3B, PhOMeB, and NMe2B. It is indicated that the aromatic and the strong electron-donating substituents significantly affect the electronic structures of the TpB scaffold.


image file: d5cc04171f-f2.tif
Fig. 2 (a) Chemical structures TpB-based complexes examined. (b) UV-vis absorption and emission spectra of synthesized complexes in 1.0 × 10−5 M chloroform solutions. (c) UV-vis diffuse reflectance and emission spectra of synthesized complexes in crystalline states.
Table 1 Photophysical properties of the synthesized compounds in dilute chloroform solutions and crystals
Solution Crystal
λ onsetabs [thin space (1/6-em)] /nm λ PL/nm PLQYc λ onsetdiff [thin space (1/6-em)] /nm λ PL/nm PLQYc
a Onset wavelength of absorption spectra. b Onset wavelength of diffuse reflectance spectra. c Absolute photoluminescence quantum yield. For BrB and OMeB, the absolute PLQY value was decomposed into two fractions based on PL integration ratio.
TpB 372 378 0.05 367 386 0.04
IB 403 401 <0.01 429 599 <0.01
BrB 395 404 <0.01 424 423 <0.01
584 <0.01
PhCF3B 394 395 0.02 391 410 0.03
PhB 396 412 <0.01 397 430 0.05
OMeB 402 401 0.07 430 423 <0.01
566 0.02
PhOMeB 420 469 <0.01 431 474 0.27
NMe2B 466 497 0.05 529 600 <0.01


All substituents also induced the redshift of the onset wavelength (λonsetPL) in the PL spectra. The PL bands were observed in the order of PhCF3B < IB < BrB < PhB < OMeB < PhOMeB < NMe2B, and the order was the same as that of λonsetabs, except for BrB. The decreases in the PLQY by halogen and aryl substituents are probably because of the heavy atom effect, which facilitates the intersystem crossing from singlet to triplet excited states, and additional intramolecular motions, respectively. Meanwhile, the OMe and NMe2 groups did not significantly affect the PLQY.

We conducted cyclic voltammetry to estimate the LUMO level (Fig. S22 and Table S26). In addition, the HOMO level of each complex was estimated from the LUMO level and the optical gap which was determined from the absorption spectra (Fig. S23). As expected, the HOMO level increases as the electron-donating ability of the substituent becomes stronger. On the other hand, the LUMO level is changed less significantly by the C5-substituent than the HOMO. This result clearly indicates that the C5-substituent affects the HOMO level more efficiently, owing to the isolated HOMO of TpB.

The solvent dependency of their optical properties was also tested with toluene, chloroform, and acetonitrile solutions (Fig. S2–S9 and Tables S7–S14). All 5-substituted complexes, except PhOMeB, showed less significant changes in their λonsetabs and λPL, similarly to TpB. The PL band of PhOMeB exhibited the bathochromic shift with increasing solvent polarity (Fig. S8 and Table S13), suggesting that its excited state should have the intramolecular charge transfer (ICT) likely between the electron-deficient TpB moiety14 and the electron-donating 4-methoxyphenyl group.

Optical properties in crystalline states were investigated (Fig. 2c and Table 1). As previously reported,13 the crystals of TpB exhibited fluorescence at 386 nm, which was similar to its solution. In contrast to the solution-state behaviors, remarkably different PL properties were observed in the solid state, including AIE from PhB, PhCF3B, and PhOMeB, significantly redshifted PL from OMeB and NMe2B, RTP from IB and BrB, and TADF from PhB.

Firstly, we found that the PLQY of the aryl-substituted complexes in the crystalline state was larger than that in solution, indicating an AIE character. It is likely due to the suppression of molecular motion. This hypothesis was supported by the PLQY enhancement in PMMA dispersion films compared to the solutions (Fig. S15–S17 and Tables S20–S22), where molecular motions are expected to be restricted. It should be noted that PhOMeB exhibited the most efficient emission in the crystalline state with a PLQY of 0.27 among the Tp-based complexes. To gain further insights into these AIE properties, photophysical measurements were performed using TpB, PhB, and PhOMeB in THF/water mixed solutions with varying water fractions (fw = 0–99%) (Fig. S21 and Table S25). Accordingly, as the water content increased, the PL intensity of TpB monotonically decreased, indicating that the typical ACQ behavior was observed. Meanwhile, for PhB and PhOMeB, the PL intensity decreased as the water fraction ranged from 0 to 80%, and λPL shifted to the longer wavelength region. Under such lower water contents (0–80%), the polarity of the solvent should increase upon addition of water, resulting in lower PL intensity and a bathochromic shift. Upon addition of more water (fw = 90–99%), the PL intensity increased, and λPL became close to that of the pure THF solution and the crystals in the case of PhOMeB. These higher water contents should lead to the formation of aggregates, resulting in enhanced emission.

Secondly, the crystals of OMeB and NMe2B exhibited new emission bands at longer wavelengths, with peaks at 566 nm and 600 nm, respectively. However, these emission bands were not observed in solutions or PMMA-dispersed films (1 wt% of the complex). The longer photoluminescent lifetime components were observed from the crystals of OMeB (15.4 ns) and NMe2B (8.1 ns) (Tables S18 and S19). These observations suggest that the additional PL bands should be attributed to excimer-like emission originating from intermolecular interactions. Diffuse reflectance spectra of their crystals showed an additional new absorption band in the longer wavelength regions (400–500 nm for OMeB, 460–520 nm for NMe2B), which were absent in their solution-state absorption spectra. This result clearly shows the formation of a static excimer-like structure in the ground state. Therefore, the intermolecular interactions in their crystals should be responsible for generating the additional low-energy states.

Thirdly, the crystals of IB and BrB showed the PL bands in the longer wavelength region (500–700 nm). The microsecond-order lifetimes (18.3 μs for IB and 169 μs for BrB; Tables S16 and S17) revealed that these PL bands should be attributed to RTP, supported by the heavy atom effect of iodine and bromine. In the solution states, these phosphorescent bands were not observed. This is likely because of the rapid non-radiative quenching of triplet excited states caused by molecular motions. On the other hand, their PMMA-dispersed films also exhibited a fluorescence band accompanied by a similar RTP, suggesting the phosphorescence decay rates might also be accelerated in the crystalline states by the external heavy atom effect.

To evaluate the electronic structures of the synthesized complexes, we performed density functional theory (DFT), time-dependent DFT (TD-DFT), and extended multistate complete active state second-order perturbation theory (CASPT2) calculations (see SI for computational details). Although the TD-DFT results overestimated the vertical transition energies, the trend in the S1 energy matched one in the experimentally determined optical gap, except for the position of IB (Fig. 3a). The deviation of IB might originate from the use of the pseudopotential for iodine. Furthermore, to evaluate the higher excited states more precisely within the relatively high density of states of TpB,14 the vertical transition energies for Sn states were also calculated for TpB and NMe2B by CASPT2, which considers multiple-electron excitations. The CASPT2 energies matched the positions of their absorption bands quite well compared to those by TD-DFT (Fig. 3b and c). Meanwhile, both CASPT2 and TD-DFT calculations showed that the S1 and S2 states are mainly assigned to HOMO → LUMO and HOMO−1 → LUMO+1, and HOMO → LUMO+1 and HOMO−1 → LUMO transitions, respectively (Tables S46 and S47). Therefore, in this study, TD-DFT calculations were employed to evaluate the photophysical properties, thereby saving computational resources. On the other hand, it is worth noting that the CASPT2 calculations indicated that the double (and triple) excitations also contribute to the excited states of the TpB-based complexes.


image file: d5cc04171f-f3.tif
Fig. 3 (a) Relationship between experimentally determined optical gap and calculated S1 state energy. Vertical transition energies of (b) TpB and (c) NMe2B calculated with CASPT2 (magenta) and TD-DFT (blue). Gray solid lines represent experimental absorption spectra. (d) Calculated Kohn–Sham molecular orbital energy diagram and Kohn–Sham molecular orbital distributions (isovalue = 0.03). H and L denote HOMO and LUMO, respectively.

Calculated Kohn–Sham FMO levels and their distributions are shown in Fig. 3d. The trend of the calculated HOMO and LUMO energy levels was basically consistent with the experimentally estimated values (Fig. S23 and Table S26). Significantly, only the HOMO of each complex has electron distribution on the substituent while the LUMO does not, because the TpB scaffold has an isolated HOMO on its 5-position. Therefore, their HOMO level was affected more significantly than their LUMO, as expected. As a result, the compound with the higher HOMO had the lower S1 state energy. Hence, in summary, the regulation of the FMO energy levels by the regioselective substitution of TpB was achieved.

It was shown that the balance between the resonance and inductive effects of the substituents determined the trend in their HOMO level. IB and BrB had similar HOMO levels to TpB because of the electron-donating resonance and electron-withdrawing inductive effects of the halogens. PhCF3B also possessed HOMO at a similar level because of the balance between the extension of the π-conjugation and the electron-withdrawing effect of CF3. In contrast, PhOMeB had a higher HOMO than PhB and PhCF3B originating from the destabilization effect of the electron-donating methoxy group. OMeB and NMe2B had significantly higher HOMO levels than TpB because the strong electron-donating groups are directly introduced to the 7-membered ring. The HOMO of NMe2B was located at a higher level than that of OMeB owing to the stronger electron-donating ability of its dimethylamine group.

We also calculated the transition energies of dimers by using the results of single-crystal X-ray diffraction analysis of TpB, PhB, OMeB, PhOMeB, and NMe2B (Table S44). For all the complexes, dimers exhibited smaller transition energies than single molecules, which is consistent with the fact that each complex showed fluorescence at longer wavelengths in the crystalline state compared to the solution state, indicating that intermolecular interactions cause changes in optical properties in the crystalline state.

In conclusion, we report a novel isolated FMO scaffold, tropolone, having the ee/oa system. We introduced various functional groups on the 5-position via regioselective amination and following Sandmeyer-type reactions. It was revealed that the optical properties of boron tropolonate complexes are tunable by regioselective substitution of the isolated HOMO. Absorption and emission wavelengths were redshifted by the effect of each substituent. By directly introducing a strong electron-donating group, NMe2B exhibited an emission band in the green region in solution and in the orange region in the crystalline state. Only the emission character of PhOMeB was attributed to the CT character, caused by the donor–acceptor interactions between the methoxyphenyl donor and TpB acceptor moieties. In crystalline states, significant luminescent properties were observed. Halogenated derivatives exhibited RTP, PhB showed TADF, and OMeB showed excimer-like emission. Phenylated derivatives showed AIE properties due to the rotation of phenyl groups in solutions. DFT calculations and CV measurements revealed that selective control of HOMO levels by regioselective substitution of isolated HOMO was achieved with OMeB and NMe2B where electron-donating groups were directly introduced. These results suggest that TpB could be a promising candidate for the development of functional luminescent materials with tunable electronic properties according to a preprogrammed design.

A part of the computation time was provided by the SuperComputer System, Institute for Chemical Research, Kyoto University. This work was partially supported by the Foundation for the Promotion of Ion Engineering (for S. I.), Grant-in-Aids for Early-Career Scientists (for S. I., JSPS KAKENHI Grant Numbers 23K13793) and for Scientific Research (B) (for K. T., JSPS KAKENHI Grant Number, 24K01570), and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. RS-2024-00406152).

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the SI. Supplementary information: Experimental details, single-crystal analysis, photophysical properties, electrochemical measurements, computational details, and NMR spectra. See DOI: https://doi.org/10.1039/d5cc04171f

CCDC 2426035, 2426044, 2426051 and 2426365 contain the supplementary crystallographic data for this paper.22–25

Notes and references

  1. Y. Huo, J. Lv, M. Wang, Z. Duan, H. Qi, S. Wang, Y. Liu, L. Peng, S. Ying and S. Yan, J. Mater. Chem. C, 2023, 11, 6347–6353 RSC.
  2. Y. Xu, Q. Wang, X. Cai, C. Li, S. Jiang and Y. Wang, Angew. Chem., Int. Ed., 2023, 62, e202312451 CrossRef CAS PubMed.
  3. A. G. Lvov, M. Herder, L. Grubert, S. Hecht and V. Z. Shirinian, J. Phys. Chem. A, 2021, 125, 3681–3688 CrossRef CAS PubMed.
  4. Y. Wu, Y. Zhao and Y. Liu, Acc. Mater. Res., 2021, 2, 1047–1058 CrossRef CAS.
  5. H. Bürckstümmer, E. V. Tulyakova, M. Deppisch, M. R. Lenze, N. M. Kronenberg, M. Gsänger, M. Stolte, K. Meerholz and F. Würthner, Angew. Chem., Int. Ed., 2011, 50, 11628–11632 CrossRef PubMed.
  6. M. Gon, K. Tanaka and Y. Chujo, Angew. Chem., Int. Ed., 2018, 57, 6546–6551 CrossRef CAS PubMed.
  7. H. Watanabe, J. Ochi, K. Tanaka and Y. Chujo, Eur. J. Org. Chem., 2020, 777–783 CrossRef CAS.
  8. Y. Kawano, Y. Ito, S. Ito, K. Tanaka and Y. Chujo, Macromolecules, 2021, 54, 1934–1942 CrossRef CAS.
  9. S. Ito, Y. Ito, T. Kazuo and Y. Chujo, Polymer, 2022, 239, 124463 CrossRef CAS.
  10. H. Takahashi, H. Watanabe, S. Ito, K. Tanaka and Y. Chujo, Chem. – Asian J., 2023, 18, e202300489 CrossRef CAS PubMed.
  11. M. Gon, S. Ito, K. Tanaka and Y. Chujo, Bull. Chem. Soc. Jpn., 2021, 94, 2290–2301 CrossRef CAS.
  12. W. von, E. Doering and L. H. Knox, J. Am. Chem. Soc., 1951, 73, 828–838 CrossRef.
  13. H. Ogoshi, S. Ito and K. Tanaka, Bull. Chem. Soc. Jpn., 2023, 96, 452–460 CrossRef CAS.
  14. H. Ogoshi, H. Takahashi, S. Ito and K. Tanaka, Macromolecules, 2024, 57, 3462–3469 CrossRef CAS.
  15. A. Ito, H. Muratake and K. Shudo, J. Org. Chem., 2009, 74, 1275–1281 CrossRef CAS PubMed.
  16. T. Nozoe, S. Seto, S. Ebine and S. Ito, J. Am. Chem. Soc., 1951, 73, 1895 Search PubMed.
  17. J. Potenziano, R. Spitale and M. E. Janik, Synth. Commun., 2005, 35, 2005–2016 CrossRef CAS.
  18. J. Mei, N. L. C. Leung, R. T. K. Kwok, J. W. Y. Lam and B. Z. Tang, Chem. Rev., 2015, 115, 11718–11940 CrossRef CAS PubMed.
  19. K. Tanaka, Dalton Trans., 2024, 53, 9240–9247 RSC.
  20. H. Nakanotani, Y. Tsuchiya and C. Adachi, Chem. Lett., 2021, 50, 938–948 CrossRef CAS.
  21. S. Hirata, Adv. Opt. Mater., 2017, 5, 1700116 CrossRef.
  22. H. Ogoshi, S. Ito and K. Tanaka, CCDC 2426035: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2mfh6s.
  23. H. Ogoshi, S. Ito and K. Tanaka, CCDC 2426044: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2mfhh2.
  24. H. Ogoshi, S. Ito and K. Tanaka, CCDC 2426051: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2mfhq9.
  25. H. Ogoshi, S. Ito and K. Tanaka, CCDC 2426365: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2mftvr.

This journal is © The Royal Society of Chemistry 2025
Click here to see how this site uses Cookies. View our privacy policy here.