Introduction

Dipyrrin is a monovalent bidentate ligand with two pyrrole groups, a substructure of porphyrin. Dipyrrin is very famous for its BF2 complex (BODIPY), which shows strong light absorption at visible region and high fluorescence quantum yield, and it is intensely studied as light-harvesting antennas, bioprobes, laser dyes, and so on. Dipyrrin can also coordinate to various metal ions to form bis- or tris(dipyrrinato)metal complexes, and some of them such as Zn(II), Cu(II), Al(III), Ga(III) complexes are also known to show fluorescence. Such metal complexes are useful not only for pigment, but also supramolecular linker units, so that metal complexes are advantageous over BODIPYs in this regard. To focus on the fluorescence property, Zn(II) complexes are known to the highest ΦF among metal complexes. In my Ph. D. work, I studied syntheses of some BODIPYs and bis(dipyrrinato)Zn(II) complexes and their physical properties. Also, through the study of various complexes, I revealed the detailed photochemical processes, which is essential to utilize them as fluorescent materials.

Meso-alkynyl BODIPYs

Many BODIPYs with π-conjugated substituents have been reported, for example, alkynyl groups at 2,6- positions, and/or vinyl groups at 3,5- or 1,7- positions. However, there are only a few examples for substitution of a π-conjugated functional group such as alkynyl and vinyl groups at the meso- (8-) position. In order to reveal the detailed properties of meso- alkynyl substituted BODIPYs, I synthesized BODIPY monomers 1 and dimers 2 (Figure 1).

UV-visible absorption spectra of 1a showed a peak at 554 nm in hexane, which is red-shifted about 40 nm compared to the reference BODIPY 3a (Figure 2). Fluorescence of 1a was observed at 570 nm with high fluorescence quantum yield (ΦF = 0.86) in toluene. Generally, ΦF of BODIPY diminishes when a rotatable aryl group was introduced to the meso- position because of thermal rotation deactivation; however, high ΦF of 1a showed no such influence of the phenylethynyl group.

DFT calculation of 1a showed that HOMO was completely localized on the dipyrrinato ligand, while LUMO was distributed around both the dipyrrinato ligand and the phenylethynyl group. As a result, the energy level of HOMO did not change from that of 3a, while LUMO was lower reflecting the extension of π-conjugation. This behavior was also revealed by cyclic voltammetry: the oxidation potentials of 1a and 3a were almost identical (~0.9 V), while the reduction potential of 1a (–1.21 V) was positively shifted than that of 3a (–1.57 V).

Chapter 3. Heteroleptic Bis(dipyrrinato)Zn(II) Complexes

As mentioned before, Zn(II) complex is the most fluorescent among bis or tris(dipyrrinato)metal complexes, but the highest reported value was 0.36, far inferior to that of BODIPY which sometimes reaches almost 1.0. Meanwhile, mono(dipyrrinato)Zn(II) shows high ΦF, indicating that the effect of zinc atom, such as heavy atom effect, is not the major reason of low ΦF. This disadvantage has degraded bis(dipyrrinato)Zn(II) complexes as a fluorescence material.

In this work, I developed the charge-transfer hypothesis for fluorescence quenching in bis(dipyrrinato)Zn(II) complexes. The concise expression is as follows: photoexcitation of bis(dipyrrinato)Zn(II) complexes gave 1π-π* excited state, in which one electron moves from one ligand to the other, to form charge-separated states (CS). 1π-π* excited state and CS are in equilibrium and decay from 1π-π* excited state accompanies fluorescence, while decay from CS goes thermally. As a result, the degree of contribution of 1π-π* excited state determines the ΦF.

Based on this hypothesis, I developed heteroleptic bis(dipyrrinato)Zn(II) complexes for the purpose of the improvement of ΦF of bis(dipyrrinato)Zn(II) complexes. All heteroleptic complexes 4-7 were synthesized from the corresponding dipyrrinato ligands and zinc acetate (Figure 3). The structures were determined by X-ray crystallographic analysis. Although Zn(II) is known to be highly substitution active, heteroleptic complexes are stable in solution: heating a CDCl3 solution of 4a resulted in neither decomposition nor disproportionation.

UV-visible absorption spectra for heteroleptic complexes showed two absorption peaks corresponding to the π-π* transition of each ligands, and no shift from the absorption maxima of the corresponding homoleptic complexes. On the other hand, fluorescence spectra showed only one peak derived from the ligand with lower excitation energy (Figure 4). Excitation spectra for the most of heteroleptic complexes were revealed to be identical with their absorption spectra, indicating that when the ligand with shorter absorption wavelength was excited, the excitation energy was quantitatively transferred from this ligand to the other. The ΦF of heteroleptic complexes showed higher values than those of the corresponding homoleptic complexes except 4b. Especially,ΦF of 4a, 5b, and 6 reached to the comparative value of those of the corresponding BODIPYs. On the other hand, 4b showed very weak fluorescence.

As a representative of heteroleptic complexes 4-7, the increase and decrease of ΦF in 4a and 4b are explained by charge-transfer hypothesis as described below (Figure 5). DFT calculation revealed that homoleptic complexes have two degenerated HOMOs and LUMOs assignable to the π and π* orbitals of dipyrrin ring. In heteroleptic complex 4a, the HOMO and the LUMO localized on the π-conjugated dipyrrin ligand, and HOMO–1 and LUMO+1 localized on the companion ligand. On the other hand, HOMO and HOMO–1 of 4b were reversed compared to 4a because of the introduction of electron donating alkyl groups into the companion ligand. In heteroleptic complex 4a, CS became unstable than homoleptic complex 10, because of the energy difference between HOMO and HOMO–1, or LUMO and LUMO+1, which resulted in the increment of contribution of π-π* excited state and higher ΦF. In contrast, reversed ordering of HOMO and HOMO–1 in 4b resulted in stabilization of the CS and in weak fluorescence.

The existence of CS was suggested by other experimental results. All bis(dipyrrinato)Zn(II) complexes largely degrade their ΦF in more polar CH2Cl2 in which charge-separated states are generally more stabilized than in toluene: such tendency is not so evident in BODIPYs. It is worth noting that degradation in polar solvent is more evident in homoleptic complexes rather than heteroleptic ones, reflecting the efficient frontier orbital ordering. Moreover, direct observation of CS was attempted by time-resolved UV-visible absorption spectroscopy. Transient absorption spectrum for homoleptic complex 8a showed the positive change at around 545 nm, which is assignable to the absorption peak of one electron oxidized and/or reduced product of 8a.

I also studied solid-phase synthesis of heteroleptic bis(dipyrrinato)Zn(II) complexes, for reducing the generation of homoleptic complexes as the side product during the solution phase synthesis. Immobilization of a terminal ligand 11 to a polystyrene resin, subsequent coordination of a Zn(II) ion and another terminal ligand, and cleavage from the resin by saponification successfully gave heteroleptic complex 15, although subgeneration of a homoleptic complex 16 could not be completely eliminated (Scheme 1).

Heteroleptic Dipyrrin-Azadipyrrin Zn(II) Hybrid Complex

Azadipyrrin is another family of the dipyrrin, in which in which a methine carbon atom of dipyrrin at meso- position is substituted by a nitrogen atom. Azadipyrrin is also known to form a bis(azadipyrrinato)Zn(II) complex, however, all of them are reported to be non-fluorescent. I assumed that the fluorescence of bis(azadipyrrinato)Zn(II) complex quenches due to charge-separation pathway as shown in bis(dipyrrinato)Zn(II) complexes, and also I adopted the strategy used for bis(dipyrrinato)Zn(II) complexes into the azadipyrrin-Zn(II) complex. Heteroleptic azadipyrrin-Zn(II) complex 17 (Figure 6) bearing a π-conjugated azadipyrrinato ligand and a simple dipyrrinato ligand was synthesized by the same way as heteroleptic bis(dipyrrinato)Zn(II) complexes. The structure of 17 was revealed by X-ray crystallographic analysis and zinc center was distorted from tetrahedral because of π-π interaction between the phenyl group and the pyrrole ring. 17 showed sharp absorption band at 485 nm and 656 nm, and exhibited fluorescence exclusively at 672 nm. Although the fluorescence quantum yield was very low (ΦF = 0.00043), considering that the corresponding bis(azadipyrrinato)Zn(II) complex 18 was totally non-emissive, the result indicates the success of strategy for the improvement of the fluorescence quantum yield.

Dipyrrin Zn(II) Complexes bearing Supporting Ligands

Chapter 3 showed the improvement of the fluorescence quantum yield of bis(dipyrrinato)Zn(II) complexes, however, this can adopt to the heteroleptic complexes bearing π-conjugated dipyrrin as a fluorescent ligand. Therefore, for the purpose to improve the fluorescence of the Zn(II) complexes with simple dipyrrin as fluorescent ligand, the use of a supporting ligand with shorter absorption wavelength than a dipyrrin was investigated. At first, a dipyrrinato Zn(II) complex 19 with an iminopyrrolyl supporting ligand was developed (Figure 7). 19 seemed to be highly fluorescent, however, this was unstable in solution and could not be isolated. Next, complex 20 bearing a bisoxazoline ligand was synthesized. 20 was highly fluorescent in solution (ΦF = 0.70 in toluene) and stable enough to be isolated. Considering that previously reported mono(dipyrrinato)Zn(II) complexes are usually highly unstable like 19, development of 20 expands the potential use of dipyrrinatoZn(II) complexes as photochemical materials.

Conclusion

In conclusion, new BODIPYs and dipyrrinatoZn(II) complexes were developed. Their photophysical properties were revealed along with electrochemical and theoretical studies. The charge-separation mechanism in the photochemical process of bis(dipyrrinato)Zn(II) complexes was proposed. The strategies for the improvement of the fluorescence quantum yields based on the hypothesis were successful, suggesting the reasonability of the existence of the charge-separated states. In addition, the direct observation of the charge-separated state proved the hypothesis.

Figure 1. Novel meso-alkynyl BODIPYs 1 and 2

Figure 2. Absorption and emission spectra for 1a, 2a and 3

Figure 3. Heteroleptic and homoleptic bis(dipyrrinato)Zn(II) complexes

Figure 4. Absorption and emission spectra for 4a, 8a and 10

Figure 5. Illustration of charge-separation hypothesis

Scheme 1. Solid-phase synthesis of a heteroleptic bis(dipyrrinato)Zn(II) complex

Figure 6. Heteroleptic and homoleptic azadipyrrinatoZn(II) complexes

Figure 7. Fluorescent mono(dipyrrinato)Zn(II) complexes

本論文は六章と付録からなり、第一章では研究の背景と目的、第二章はメソアルキニルBODIPYの合成と物性、第三章はヘテロビスジピリン亜鉛錯体の合成と物性、第四章はヘテロアザジピリン亜鉛錯体の合成と物性、第五章はモノジピリン亜鉛錯体の合成と物性、第六章に研究成果のまとめと展望を述べられている。以下に各章の概要を示す。

第一章では研究の背景と目的について述べている。金属錯体は様々な光機能材料として用いられており、それぞれの目的に応じた異なる光化学過程が利用されている。従って、金属錯体の未知の光化学過程を明らかにすることは、これらを応用する上で必要不可欠であると言える。ジピリンは強い光吸収を有する一価の二座配位子であり、強い蛍光を有するホウ素錯体(BODIPY)で有名である。一方、ジピリン金属錯体は、金属-配位子結合を活用した超分子構造体を構築できる点でBODIPYより優れているが、蛍光特性という点においては極めて劣っている。従って、ジピリン金属錯体の蛍光特性の改善およびそのメカニズムの解明は、これらを光機能性材料に応用する上で重要な課題であると述べている。本論文では、新規な蛍光性ジピリン錯体、特にビスジピリン亜鉛錯体を開発すること、また、それらの蛍光特性の研究を通じてジピリン錯体の光化学過程を明らかにすることを目的としている。

第二章では、メソアルキニルBODIPYについて述べている。メソ位にフェニルエチニル基を導入したBODIPYを合成し、吸収・発光の長波長シフトからπ共役系が拡張されていることを示した。電気化学測定および理論計算より、メソ位へのアルキニル基の導入に伴い、HOMOはほとんど変化せず、LUMOは低下することを明らかにした。

第三章では、ヘテロビスジピリン亜鉛錯体について述べている。既報のホモビスジピリン亜鉛錯体の蛍光量子収率が低いことを、配位子間の電荷分離によるクエンチであるという仮説を立て、これを基に、蛍光特性の改善を目的としたヘテロビスジピリン亜鉛錯体を設計している。ヘテロ錯体の性質として、対応するホモ錯体と変わらない吸収・発光波長を有すること、高い蛍光量子収率を示したこと、配位子間で定量的なエネルギー移動が起こること、電気化学により見積もられた電荷分離状態のエネルギーと蛍光量子収率とが妥当な相関を示すことについて述べている。また、時間分解分光による電荷分離状態の直接観測を行うことで、電荷分離仮説を実証したと述べている。

第四章では、第三章を踏まえ、メソ位のメチン炭素が窒素に置き換わったジピリン類縁体であるアザジピリンについて、第三章で用いた戦略が適応できるかどうかを調べることを目的とした、ヘテロアザジピリン-ジピリン亜鉛錯体の合成について述べている。ヘテロ錯体は、量子収率こそ低いものの、ホモ錯体では観測されなかった蛍光が復活し、第三章での戦略がアザジピリンにも適用できたと述べている。

また、第五章でも第三章を踏まえ、最小構成となるジピリン配位子を発光部位とした、安定なジピリン亜鉛錯体を実現しうる補助配位子の探索を行ったと述べている。補助配位子としてイミノピロールおよびビスオキサゾリン配位子を用い、いずれも強い蛍光を有すること、ビスオキサゾリン配位子を用いた場合には高い熱安定性を同時に実現できたと述べている。

第六章では以上の結果を総括し、今後の展望について述べている。

以上、本論文では新規な強蛍光性のπ共役BODIPYおよびジピリン亜鉛錯体を開発することに成功し、またこれらの蛍光特性と電子構造との関連性を調べることにより、重要な光化学過程を明らかにしたと記述している。本博士論文においては、これらの光化学過程を制御するためのフロンティア軌道操作についても述べており、エネルギー移動を利用したキャリア輸送、電子移動を利用した光合成模倣システム、配位子-金属結合による光機能性配位性高分子など様々な応用が出来ると期待される。なお、本論文は、第二章は坂本良太、北河康隆、奥村光隆、西原寛との共同研究、第三章は坂本良太、玉井尚登、北河康隆、奥村光隆、西原寛との共同研究、第四章は坂本良太、岸田正彬、高良祐亮、林幹大、土屋瑞穂、柿沼純子、武田拓真、平田圭祐、荻野誠也、河原佳祐、八木俊樹、池平秀、中村智也、磯村真由子、外山未琴、市川早紀、北河康隆、奥村光隆、西原寛との共同研究、第五章は坂本良太、西原寛との共同研究であり、一部は学術誌にて出版されたものであるが、論文提出者が主体となって実験、解析を行ったもので、論文提出者の寄与が十分であると判断する。

したがって、博士(理学)の学位を授与できるものと認める。