Control of the electronic and optical properties of aminoxyl radicals via boron complexation

Abstract

Stable radicals have attracted increasing attention in recent years because of their unique electronic and optical characteristics. Aminoxyl radicals are one of the most widely studied stable radicals to date, but their applications in opto-functional materials have yet to be explored in detail. Our group previously reported the boron complexes of aminoxyl radicals exhibit near-infrared (NIR) absorption. In this work, an aminoxyl radical without boron-complexation was synthesized to elucidate the effects of boron coordination on the properties of the aminoxyl radicals. The results of electron spin resonance spectroscopy, ultraviolet-visible-NIR absorption measurements, and density functional theory calculations indicated that boron complexation facilitated spin delocalization over the radical π-frameworks. Furthermore, a π-extended aminoxyl radical-boron complex exhibited a significant wavelength-shift to longer wavelengths in the NIR-II absorption region, thereby reflecting its larger π-conjugated radical skeleton.

---

Introduction

In recent years, air-stable organic radicals have attracted increasing attention because of their importance in fundamental aspects of organic chemistry and functional materials.¹ Although organic radicals are generally highly reactive and labile species, they can be stabilized by delocalization and steric protection of the reactive spin centres. Due to their open shell electronic structures, stable radicals exhibit unique properties that are not found in ordinary closed-shell compounds, which allows them to be widely studied as key components in functional organic materials and in opto-electronic applications.² Organic radicals generally possess small SOMO–LUMO (singly occupied molecular orbital, SOMO; lowest unoccupied molecular orbital, LUMO) and HOMO–SOMO (highest occupied molecular orbital, HOMO) energy gaps. This allows them to absorb lower-energy light irradiation, such as near infrared (NIR) light, leading to the application of stable radicals in NIR dyes.³ Furthermore, stable radicals exhibiting electro- and magneto-luminescence properties have recently appeared as an emerging class of opto-functional organic materials.^4,5 These optical properties are expected to accelerate the development of stable radical compounds with new structures, reactivities, and functions.

Aminoxyl radicals (R₂N-O˙), which are also known as nitroxyl radicals or nitroxides, are some of the most widely studied stable radicals.⁶ More specifically, the physical, chemical, and electrochemical stabilities of aminoxyl radicals have led to their application in redox catalysts,⁷ batteries,⁸ and organic magnetic materials.⁹ In addition, aminoxyl radicals can form Lewis acid–base-type complexes with various metal ions. Such complexes have shown to exhibit spin–spin interactions between the aminoxyl ligands and the magnetic transition metal ions.¹⁰ Consequently, paramagnetic aminoxyl radical complexes have been investigated as candidate materials for molecular magnets.

However, limited examples are known of aminoxyl radical-main group element complexes. As one example, Hayton et al. synthesized an isolable complex of 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO) and AlCl₃, which demonstrated an enhanced oxidation ability compared to that of TEMPO alone.¹¹ In another study, Graves et al. reported the electrochemical properties of aluminium and gallium complexes of 2-pyridyl-substituted aminoxide ligands,^12,13 although the corresponding aminoxyl radical complexes were not synthesized. Since aminoxyl radicals generally exhibit weak optical absorption characteristics, their applications in optical materials, such as NIR dyes and luminophores, have been limited.

Recently, our group reported aminoxide-boron complex 1 (Fig. 1),¹⁴ which take part in reversible, two-step one-electron oxidation processes that involve oxidation to the corresponding cationic aminoxyl radical and dicationic complexes. The chemical oxidation of aminoxide complex 1 using tris(4-bromophenyl)ammoniumyl hexachloroantimonate [(4-BrC₆H₄)₃N]SbCl₆ (“Magic Blue”) yielded the cationic aminoxyl radical complex [1]SbCl₆ as dark violet air- and moisture-stable crystals. Due to the donor–acceptor interactions between the Ar₂N groups and the aminoxyl radical-boron complex moieties, [1]SbCl₆ complex exhibits intense absorption in the NIR-II region (λ_max 1052 nm). However, the effects of boron-complexation on the properties of the aminoxyl radicals were not elucidated. In addition, the effects of the donor–acceptor-type structure on the properties of the aminoxyl radical were not clarified.

Fig. 1 Aminoxyl radical compounds ([1]SbCl₆, [2]SbCl₆, 3, and 4) and their related aminoxyl complexes (1 and 2) investigated in this work. Compounds 1 and [1]SbCl₆ have been previously reported in the literature.¹⁴

Thus, in the current study, a novel aminoxyl radical-boron complex [2]SbCl₆ bearing an extended conjugated skeleton structure is synthesized. In addition, aminoxyl radicals 3 and 4, which contain no boron coordination, are also synthesized. Furthermore, the electrochemical properties, electron spin resonance (ESR) spectra, molecular structures, and optical absorption characteristics of these aminoxyl radical compounds are compared to elucidate the effects of boron complexation and the donor–acceptor-type skeleton on the properties of such aminoxyl radical-boron complexes.

Results and discussion

Synthesis

Synthesis of the new aminoxyl radical compounds was performed as outlined in Scheme 1 (see the ESI† for details). Oxidation of the neutral and closed-shell aminoxide complex 2 with Magic Blue in CH₂Cl₂ (DCM) afforded cationic aminoxyl complex [2]SbCl₆ as dark brown crystals. Boron-free aminoxyl radicals 3 and 4 were synthesized starting from compounds 5 and 6via hydroxylamine intermediates 7 and 8, respectively.

Scheme 1 Synthesis of the aminoxyl radical compounds.

Electrochemical analysis of the aminoxide complexes

The cyclic voltammograms recorded for aminoxide complexes 1 ¹⁴ and 2 are shown in Fig. 2, wherein it can be seen that both the complexes exhibited reversible and two-step single electron oxidations. The first oxidation potential of 2 (+0.55 V) was almost identical to that of 1 (+0.57 V), while the second oxidation potential of 2 (+0.95 V) was shifted to the kathodic side compared to that of 1 (+1.12 V). This difference was attributed to the reduced Coulomb repulsion between the cationic charges in the more π-extended skeleton of 2.

Fig. 2 Cyclic voltammograms recorded for aminoxide complexes 1 ¹⁴ and 2 measured at 20 °C and at a scan rate of 0.1 V s⁻¹ in CH₂Cl₂ (1 mM 1 or 2) containing Bu₄NPF₆ (0.1 M). Working electrode: glassy carbon; counter electrode: Pt wire; reference electrode: Ag/AgNO₃ in MeCN.

Single crystal X-ray diffraction (scXRD) analysis of the radical compounds

The molecular structures of radical compounds [2]SbCl₆, 3, and 4 were determined using single crystal X-ray diffraction (scXRD), as was that of neutral complex 2. The scXRD results are presented in Fig. 3a, while some selected structural parameters are summarized in Fig. 3b. For comparison, the molecular structures of the previously reported compounds 1 and [1]SbCl₆ are also included.¹⁴ For [1]SbCl₆, the N1–O (1.376(3) Å), N1–pyridine (1.362(2) Å), and N3–pyridine (1.402(2) Å) bonds were shortened compared to those of the corresponding neutral complex 1, suggesting delocalization of the radical spin density through the aminoxyl π-skeleton. In contrast, boron-free radical 3 exhibited a remarkably short N1–O bond (1.280(3) Å), while its N1–pyridine and N3–pyridine bonds were longer than those in 1 and 1⁺ (the cationic component of [1]SbCl₆), indicating the localization of its radical spin population on the N1–O moiety. In addition, the N1–O, N1–pyridine, and N3–pyridine bond lengths in compounds 2 and 4 were similar to those observed for 1 and 3, respectively. Furthermore, it was found that the N1–O (1.425(3) Å) and N1–pyridine (1.363(3) 1.425(3) Å) bonds of [2]SbCl₆ were similar to those of 2, while the N3–pyridne (1.395(3)) and pyridine-benzene (1.463(3) Å) bonds were shortened compared to those of 2. These structural features of [2]SbCl₆ suggest that the radical spin was not substantially distributed on the N–O moiety.

Fig. 3 (a) Molecular structures of the aminoxyl radical compounds ([1]SbCl₆,¹⁴[2]SbCl₆, 3, and 4) and the aminoxide complexes (1 ¹⁴ and 2). (b) Selected bond lengths (Å) for the radicals ([1]SbCl₆, [2]SbCl₆, 3, and 4) and the neutral complexes (1 and 2). The corresponding bond lengths for the DFT-optimized structures (CAM-B3LYP/6-311G(d)) are shown in parentheses.

The molecular structures of these compounds were further investigated by density functional theory (DFT) calculations using CAM-B3LYP/6-311G(d) level of theory.¹⁵ The scXRD-determined coordinates of the radical compounds were employed as the initial coordinates for the structural optimization calculations. The obtained optimized coordinates corresponded to the equilibrium structures, as confirmed by the frequency calculations. Moreover, the experimentally determined structural parameters for compounds 1, 2, [1]SbCl₆, 3, and 4 were well reproduced with the aid of DFT optimizations. Additionally, the DFT-optimized structure for 2⁺ (the cationic component of [2]SbCl₆) indicated that the N1–O (1.374 Å) and N1–pyridine (1.321 Å) bond lengths were shorter than those of 2 (N1–O: 1.407 Å, N1–pyridine: 1.358 Å), suggesting that radical spin delocalization occurred through the aminoxyl π-skeleton of 2⁺ as in the case of 1⁺. The discrepancies between the experimental and calculated structures may be attributed to the effects of packing in the single crystals of [2]SbCl₆.

Electron spin resonance (ESR) spectra of the radical compounds

The X-Band ESR spectra of the aminoxyl complexes ([1]SbCl₆ ¹⁴ and [2]SbCl₆) and the boron-free radicals (3 and 4) were recorded at 20 °C in DCM (Fig. 4). Both boron-free radicals exhibited triplet ESR signals due to the nitrogen hyperfine coupling that is typical in aminoxyl radicals. In contrast, boron complexes [1]SbCl₆ and [2]SbCl₆ exhibited broad ESR signals, with the latter demonstrating no clear hyperfine coupling pattern. This result indicates that spin delocalization occurred over the π-conjugated aminoxyl radical skeleton.

Fig. 4 X-band ESR spectra of the aminoxyl complexes ([1]SbCl₆ ¹⁴ and [2]SbCl₆) and the boron-free radicals (3 and 4) measured at 20 °C in DCM (*: Mn marker signals).

To investigate spin delocalization in these radicals in greater detail, the Mulliken spin populations of the radicals were calculated by the DFT calculations (Fig. 5). These DFT calculations were performed on the DFT-optimized geometries of 1⁺, 2⁺, 3, and 4 as well as on the XRD geometry of 2⁺. For comparison, the spin population of TEMPO was also calculated. The spin populations on the O atoms of the boron-free radicals (3: 0.521, 4: 0.530) were comparable to that of TEMPO (0.526), while the spin populations on the N1 atoms in 3 (0.366) and 4 (0.359) were reduced compared to that of TEMPO (0.438). These differences were attributed to delocalization of the spin densities from the N1 atoms to the pyridine rings in 3 and 4. Consequently, in these radicals, the N2 atoms carry negligible spin densities, i.e., 0.011 and 0.002 for compounds 3 and 4, respectively. The spin populations on the O atoms of the boron complexes (1⁺: 0.097, 2⁺: 0.023 (Opt), 0.006 (XRD)) were found to be significantly lower than those of the boron-free radicals, while the spin populations of 1⁺ on the N1 (0.307) and N2 (0.280) atoms were comparable. Furthermore, the spin population on the N1 atom of 2⁺ (0.100 (Opt), 0.042 (XRD)) was determined to be significantly lower than that on the N2 atom (0.339 (Opt), 0.388 (XRD)). The spin density surfaces also showed that the unpaired electrons of 1⁺ and 2⁺ were substantially delocalized over the π-conjugated systems of the aminoxyl radicals.

Fig. 5 Mulliken spin populations for the radical compounds at the DFT-optimized structures (CAM-B3LYP/6-311G(d) level of theory). For 2⁺, the Mulliken spin populations calculated on the XRD-determined coordinate are shown in parentheses. Spin density surfaces of the radical compounds are also shown (isovalue: 0.000400 Å⁻³).

These findings suggest that complex 1⁺ could be described as the resonance hybrid of two cationic N-centred radicals, namely 1⁺-A and 1⁺-B (Fig. 6). In the case of 2⁺, the contribution of radical cation form 2⁺-B may be more important than that of 2⁺-A, considering the spin populations at the N1 and N2 atoms and the spin density surface. In both cases, the contribution of the O-centred radical cation forms, i.e., 1⁺-C and 2⁺-C are considered to be marginal.

Fig. 6 Resonance structures for the radical complexes 1⁺ and 2⁺.

Ultraviolet–visible-near infrared (UV-vis-NIR) spectra of the radical compounds

The ultraviolet-visible-near infrared (UV-vis-NIR) spectra of the radical compounds were subsequently measured in DCM at 20 °C (Fig. 7). The optical data are summerized in Table 1, which also includes data for the closed-shell complexes 1 ¹⁴ and 2. The UV-vis spectrum of 2 can be found in the ESI.† As presented in Fig. 7, radical complex [2]SbCl₆ exhibited a significant wavelength shift to longer wavelength along with a broad absorption (λ_max 1184 nm, ε 1.0 × 10⁴ M⁻¹ cm⁻¹) compared to [1]SbCl₆, reflecting an extension of the π-conjugated skeleton. The time-dependent (TD) DFT calculations on the optimized coordinate of 2⁺ (Opt) reproduced the observed absorption characteristics (λ_calcd 1128 nm, f 0.5694), although a weaker absorption and a shift to shorter wavelengths was predicted for the XRD coordinate of 2⁺ (λ_calcd 905 nm, f 0.3424), suggesting that 2⁺ (Opt) corresponds to the energy minimum structure of [2]SbCl₆ in solutions at 20 °C. The longest absorption wavelength of boron-free radical 3 was comparable to that of 4 and was significantly shorter than those of [1]SbCl₆ and [2]SbCl₆, indicating that elongation of the π-conjugated skeleton did not affect the SOMO–LUMO energy gaps in the boron-free radicals. In addition, boron-complexation was found to substantially increase the oscillator strengths (f) of the radicals. In summary, the boron-complexation of the π-conjugated aminoxyl radicals effectively stabilizes the lowest-energy excited states and increase the oscillator strengths, resulting in significantly modified absorption properties.

Fig. 7 UV-vis-NIR spectra of the radical compounds in DCM at 20 °C (1.0 × 10⁻⁵ M). X: signals originating from the measurement device.

Table 1 UV-vis-NIR absorption data for the radical compounds ([1]SbCl₆, [2]SbCl₆, 3, and 4) and the closed shell compounds (1 and 2) in DCM

Compound	λ max [nm]	ε [M^−1 cm^−1]	λ calcd ^b [nm]	f
a Ref. 14. b Lowest energy excitations obtained by the TD-DFT calculations (CAM-B3LYP/6-311+g(d)) on the DFT-optimized geometries. c Calculated oscillator strengths for the lowest energy excitations. d Calculated value for the DFT-optimized coordinate. e Calculated value for the scXRD coordinate.
[1]SbCl6 ^a	1052	1.7 × 10^4	925	0.3157
[2]SbCl6	1184	1.0 × 10^4	1128 (Opt)^d	0.5694 (Opt)^d
			905 (XRD)^e	0.4324 (XRD)^e
3	593	2.7 × 10^3	473	0.0008
4	573	2.3 × 10^3	482	0.0002
1	424	2.5 × 10^3	378	0.0500
2	350	3.3 × 10^4	339	0.1253

---

Conclusions

In this contribution, the effects of boron complexation on the optical properties of aminoxyl radicals were investigated by comparing the molecular structures, spin delocalization properties, and ultraviolet-visible-near infrared (UV-vis-NIR) absorption characteristics of the boron complexes of π-conjugated aminoxyl radicals and their corresponding boron-free radicals. Evaluation of the structural parameters and electron spin resonance properties combined with density functional theory calculations suggested that the unpaired electrons of the boron-aminoxyl radical complexes were mainly distributed on the N atoms of the aminoxyl and Ar₂N moieties, while the contribution of the oxygen-centred radicals was negligible. Therefore, the spin distribution of the boron complexes was found to be remarkably different from that in the boron-free aminoxyl radicals. The longest absorption bands of the boron complexes were observed in the NIR-II region, which were shifted to substantially longer-wavelengths than those of the boron-free radicals, thereby reflecting the effective spin delocalization occurring in the boron complexes. The application of these boron-aminoxyl radical complexes in NIR dyes and photo-thermal conversion applications is currently being investigated and will be described due course. The development of multi-radicals through the combination with the boron-aminoxyl radical complex is also under investigation.

Author contributions

This research project was designed by TA and TK. Synthesis, characterisation, data analysis were performed by TK. PY, MN, RH, HF contributed to the synthesis of the compounds. ESR measurements were conducted under the supervision of TO and YN. KM, TS, and YM carried out the characterisation including elemental analysis and crystallography. KM and RI carried out the electrochemical analysis and calculations, respectively. All the authors contributed to the manuscript preparation.

Data availability

The data supporting this article have been included as part of the ESI.† Crystallographic data for compounds CCDC 2353274 for (2), 23532776 for ([2]SbCl₆), 2353277 for (3), and 2353278 for (4) have been deposited at the CCDC and can be obtained from https://doi.org/10.1039/D4DT02973A.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by JSPS KAKENHI (Grant No. 21K05033 (T. A.), 21K05163 (H. F.), 23H01943 (T. S.), 24K23070 (K. M.), and 24K08382 (T. A.)). T. A. thanks the Kurata Grants from the Hitachi Global Foundation, the Shin-Sozai Joho Zaidan Grant, the Murata Science Foundation, the Iwatani Naoji Foundation, the Iketani Science and Technology Foundation, the Amano Institute of Technology, the Tokyo Kasei Chemical Promotion Foundation, and the International Collaborative Research Program of Institute for Chemical Research, Kyoto University (2024-134). The computations were performed at the Research Center for Computational Science, Okazaki, Japan (Project: 21-IMS-C164, 22-IMS-C164, 23-IMS-C170, and 24-IMS-C172, T. A.). The authors would also like to thank Prof. Shin-ichiro Kato (University of Shiga Prefecture) for guidance during the NMR measurements of the air-sensitive materials and for preparation of the Magic Blue reagents. Mr Masao Sasaki (University of Tsukuba) is acknowledged for his help in the elemental analysis. The authors would like to thank Dr Yuiga Nakamura at JASRI and BL02B1 of SPring-8, where XRD experiments were conducted under proposal number 2024A1851. Dedicated to Prof. Takayuki Kawashima on the occasion of his 77th birthday (Kiju).

References

1. I. Ratera and J. Veciana, Chem. Soc. Rev., 2012, 41, 303; Z. X. Chen, Y. Li and F. Huang, Chem, 2021, 7, 288.
2. A. Mizuno, R. Matsuoka, T. Mibu and T. Kusamoto, Chem. Rev., 2024, 124, 1034; S. Gao, Z. Cui and F. Li, Chem. Soc. Rev., 2023, 52, 2875; S. Kasemthaveechok, L. Abella, J. Crassous, J. Autschbach and L. Favereau, Chem. Sci., 2022, 13, 9833; P. Murto and H. Bronstein, J. Mater. Chem. C, 2022, 10, 7368.
3. L. Zheng, W. Zhu, Z. Zhou, K. Liu, M. Gao and B. Z. Tang, Mater. Horiz., 2021, 8, 3082; M. Yano, Y. Inada, Y. Hayashi, M. Nakai, K. Mitsudo and Y. Kashiwagi, Dyes Pigm., 2022, 197, 109929; M. Yano, Y. Inada, Y. Hayashi, T. Yajima, K. Mitsudo and Y. Kashiwagi, Chem. Lett., 2020, 49, 685; B. Lü, Y. Chen, P. Li, B. Wang, K. Müllen and M. Yin, Nat. Commun., 2019, 10, 767.
4. X. Ai, E. W. Evans, S. Dong, A. J. Gillett, H. Guo, Y. Chen, T. J. H. Hele, R. H. Friend and F. Li, Nature, 2018, 563, 536; M. Ito, S. Shirai, Y. Xie, T. Kushida, N. Ando, H. Soutome, K. J. Fujimoto, T. Yanai, K. Tabata, Y. Miyata, H. Kita and S. Yamaguchi, Angew. Chem., Int. Ed., 2022, 61, e202201965.
5. S. Kimura, T. Kusamoto, S. Kimura, K. Kato, Y. Teki and H. Nishihara, Angew. Chem., Int. Ed., 2018, 57, 12711; K. Kato, S. Kimura, T. Kusamoto, H. Nishihara and Y. Teki, Angew. Chem., Int. Ed., 2019, 58, 2606; S. Kimura, S. Kimura, H. Nishihara and T. Kusamoto, Chem. Commun., 2020, 56, 11195.
6. O. Ouari and D. Gigmes, in Nitroxides; Synthesis, Properties and Applications, ed. O. Ouari and D. Gigmes, the Royal Society of Chemistry, London, 2021, pp. 1–6; G. I. Likhtenshtein, in Ntroxides: Brief History, Fundamentals, and Recent Developments, Springer Series in Material Science, Springer Nature Switzerland, Cham, 2020, vol. 292.
7. W. Wertz and A. Studer, Green Chem., 2013, 15, 3116.
8. H. Nishide and K. Oyaizu, Science, 2008, 319, 737; E. P. Tomlinson, M. E. Hay and B. W. Boudouris, Macromolecules, 2014, 47, 6145; Y. Xie, K. Zhang, M. J. Monteiro and Z. Jia, ACS Appl. Mater. Interfaces, 2019, 11, 7096.
9. R. Chiarelli, A. Rassat and P. Rey, J. Chem. Soc., Chem. Commun., 1992, 1081; R. Chiarelli, M. A. Novak, A. Rassat and J. L. Tholence, Nature, 1993, 363, 147.
10. T. Ishida, S. Ito, Y. Homma and Y. Kyoden, Inorganics, 2021, 9, 10; A. Caneschi, D. Gatteschi, R. Sessoli and R. Rey, Acc. Chem. Res., 1989, 22, 392.
11. J. J. Schpaniak, A. M. Wright, R. A. Lewis, G. Wu and T. W. Hayton, J. Am. Chem. Soc., 2012, 134, 19350; T.-A. D. Nguyen, A. M. Wright, J. S. Page, G. Wu and T. W. Hayton, Inorg. Chem., 2014, 53, 11377.
12. K. Osanai, A. Okazawa, T. Nogami and T. Ishida, J. Am. Chem. Soc., 2006, 128, 14008; A. Okazawa, Y. Nagaichi, T. Nogami and T. Ishida, Inorg. Chem., 2008, 47, 8859; A. Okazawa, T. Nogami and T. Ishida, Chem. Mater., 2007, 19, 2733; A. Okazawa, D. Hashizume and T. Ishida, J. Am. Chem. Soc., 2010, 132, 11516.
13. A. M. Poitras, J. A. Bogart, B. E. Cole, P. J. Carroll, E. J. Schelter and C. R. Graves, Inorg. Chem., 2015, 54, 10901; J. M. Kirsh, A. J. Woodside, B. C. Manor, P. J. Carroll, P. R. Rablen and C. R. Graves, Inorganics, 2018, 6, 50; T. M. Herb, A. M. Poitras, K. G. Richardson, B. E. Cole, J. A. Bogart, P. J. Carroll, E. J. Schelter and C. R. Graves, Polyhedron, 2016, 114, 194.
14. M. Nakamura, R. Hyakutake, S. Morisako, T. Sasamori, Y. Mizuhata, N. Tokitoh, K. Nakashima, H. Fukumoto and T. Agou, Dalton Trans., 2022, 51, 13675.
15. M. J. Frisch, et al., Gaussian 16, Revision C.02, Gaussian, Inc., Wallingford, CT (USA), 2019. The complete citation for this program is included in the ESI.†.

---

Footnote

† Electronic supplementary information (ESI) available: General experimental procedure details and copies of the NMR and HRMS spectra. See DOI: https://doi.org/10.1039/d4dt02973a

---