The photochemical reaction of aromatic carbonyl compounds shows significant varieties in their paths and kinetics depending on their substituents and solvents. The reaction paths and kinetics are determined by the nature of the excited states involved. Therefore, for understanding the mechanism and dynamics of photochemical reactions, it is important to know the nature of the excited electronic sates, including their electronic configurations and energy level ordering. Particularly, the low-lying excited triplet states are of great importance because most the photochemical reaction of aromatic carbonyl compounds proceeds via the low-lying excited triplet states.

In order to observe the low-lying excited triplet states, time-resolved infrared spectroscopy, in which T-T transitions are observed, has been found to be useful. Because T-T transitions are spin-allowed, they are comparable in intensity with S-S transitions. Furthermore, time-resolved infrared spectroscopy is applicable to a room-temperature solution sample. Thus, time-resolved infrared spectroscopy is a powerful tool for observing the low-lying excited triplet states.

In spite of high sensitivity and versatility, the existing time-resolved infrared spectrometer has significant deficiency to be improved: a limited spectral range below 4000 cm-1, a long measurement time and a low spectral resolution. In addition, the data processing using a separated digitizer is considerably inefficient.Utilizing a high-throughput monochromator covering wide spectral range and a-state-of-the-art data processor,the performance of the apparatus would be significantly improved.

In the present study, the author intends to:

1. Improve the performance of the time-resolved infrared spectrometer.

2. Establish the methodology for observing close-lying excited states directly by spectroscopy.

3. Elucidate the mechanism of the substituent and solvent dependence of photoreduction of aromatic carbonyl compounds.

This thesis is composed of the following seven chapters: Chapter 1 gives the general introduction.

In chapter 2, the time-resolved infrared spectrometer developed in this study is described. Figure 1 shows the block diagram of the present apparatus. Utilizing the high-throughput monochromator, whose f-number is 4.3 and whose focal length is 500 mm, the throughput of the present spectrometer becomes four times as high as that of the previous one. As a result, the minimum limit of detection less than 5x10-6 with the spectral resolution of 8cm-1 is achieved (figure 2). In addition, the high-speed digitizer mounted in the PC raises the upper limit of the repetition rate five times higher than and makes the number of the data elements that can be dealt with at a time twice more than those of the previous one.

Some tips for the noise reduction techniques and the sample cell system are also described here.

In section 3, the photoreduction reaction of benzophenone (scheme 1) is studied by the nanosecond time-resolved infrared spectroscopy. Making good use of it, the author successfully observed the whole reaction of the photoreduction of benzophenone, from the laser excitation to the formation of the products, except for the formation and the depletion of benzophenone in the S1 state in real time. This kind of pursuit of reaction cannot be achieved by any other spectroscopy and was achieved for the first time.

The time region from ~ -1 to 4 μs of the time-resolved spectra were analyzed by singular value decomposition technique. As a result, it was found that the time-resolved spectra could be reproduced by two components whose time constants were infinity and 1.9 x 106 s-1. Consequently, the rate of the hydrogen atom abstraction kA is revealed as 1.9 x 107 s-1 mol-1 dm3 for benzophenone.

The individual spectra of the transient species and the product species, namely benzophenone in the T1 state, benzophenone ketyl radical, and benzopinacol, are extracted from the time-resolved spectra (figure 3).Especially, the infrared spectrum of benzophenone ketyl radical is the first vibrational spectrum reported,because the Raman spectrum is hindered by strong emission from the ketyl radical itself.

In chapter 4, the infrared spectra of benzophenone in the lowest excited triplet state are reviewed in detail.The decay rate of the transient species heavily dependent on the concentration of O2 gas confirmed the assignment for the transient species to the lowest excited triplet state.

Within the vibrational spectrum, the CO stretch band was not observed, which is rather strange in terms of the polarity in the carbonyl group in the ground state.

The T-T electronic transition between the low-lying excited triplet states were discovered in the middle infrared region. It was suggested that, from this T-T transition observed in the infrared region, the energy of the low-lying excited triplet states, such as the T2 and T3 states, can be determined.

In chapter 5, time-resolved infrared spectra of photoexcited acetophenone and its derivatives, namely 4'-CF3-, 4'-CH3-, and 4'-CH3O-acetophenone and 2'-acetonaphthone (figure 4), were measured (figure 5). It was found that the higher-wavenumber and the lower-wavenumber bands did not shift with the substituents while the intensity ratio between them varied significantly. From this substituent dependence of the spectra, it was suggested that the thermal equilibrium between the lowest nπ* and the lowest ππ* states is established for the compounds examined here. Also, the higher-wavenumber and the lower-wavenumber bands were assigned to the nπ* -nπ* and ππ*-ππ* transitions, respectively (figure 6).

From this point of view, the mechanism of the substituent dependence of the photoreduction reaction rate was also explained. It was suggested that the partial reactivity of the ππ* states, which have been previously ascribed to the state mixing with the nπ* state, was ascribable to a partial population of the nπ* state that was involved in the thermal equilibrium.

The solvent dependence and the temperature dependence of the transient infrared spectra of photoexcited 4'-CH3O-acetophenone were also examined. The relative intensity of the ππ*-ππ* band increased as the polarity of the solvent increased, which was also explained by the thermal equilibrium. Finally, the possibility of the equilibrium between the lowest nπ* and the lowest ππ* states was confirmed by the temperature dependence of the relative intensity of the band.

In chapter 6, the vibrational bands in the transient infrared spectra of photoexcited acetophenones were studied. It was found that the vibrational bands disappear for the compounds the lowest nπ* and the lowest ππ* states of which is supposed to be isoenergetic or significantly close to each other, namely acetophenone and 4'-CH3-acetophenone. This fact supports the idea that the lowest nπ* and the lowest ππ* states is close to each other enough to come to thermal equilibrium

Finally, the conclusion of the present study is given in chapter 7.

Figure 1. Block diagram of time-resolved infrared spectrometer constructed in this study.

Figure 2. An example of the time-resolved infrared spectrum with best S/N.

Scheme 1. Photoreduction of benzophenone.

Figure 3. Individual infrared spectra of species extracted from the time-resolved infrared spectra.

Figure 4. Compounds examined in this study. R: CF3-, H-, CH3-, and CH3O-.

Figure 5. Transient infrared spectra of 4'-substituted acetophenones and 2'-acetonaphthone.

Figure 6. Model diagram of the low-lying excited triplet states. The establishment of the thermal equilibrium between the lowest nπ* and the lowest ππ* states were supposed (center).)

本論文は、高性能時間分解近・中赤外分光装置の開発と、その応用、とくに芳香族カルボニル化合物の低エネルギー励起三重項状態の観測と、光化学反応機構解明への応用について記述されており、全7章から構成される。

第1章では、芳香族カルボニル化合物の光還元反応機構解明の鍵である低エネルギー励起三重項状態の観測が、既存の分光法では困難であり、この目的のために時間分解近・中赤外分光法が有用であることが述べられている。また、既存の装置の問題点と、その解決策が挙げられている。

第2章は実験についてであり、開発した時間分解近・中赤外分光装置の詳細及びその性能評価について述べられている。波数分解能8 cm-1において吸光度差5×10-6以下の検出下限、1200 Hzの繰り返し速度、11000までをカバーする広範なスペクトル測定領域を達成した。

第3章では、時間分解赤外分光法の有用性の一例として、ベンゾフェノンの光還元反応を、光励起から最終生成物の生成までを追跡した結果を述べている。T1状態ベンゾフェノン及びケチルラジカルの赤外スペクトルが初めて得られた。

第4章ではT1状態ベンゾフェノンの赤外スペクトルについてさらに詳細に扱っている。中赤外から近赤外領域へと続く非常に幅広なT-T吸収バンドが発見された。これにより、時間分解赤外分光法が、これまでの研究で用いられてきた分光法では観測が困難であった低エネルギー励起三重項状態の観測方法として有用であることが示された。

第5章では、アセトフェノン類を主とする芳香族カルボニル化合物の励起三重項状態の観測と、光化学反応活性の置喚基効果が議論されている。観測されたアセトフェノン類の時間分解赤外スペクトルの置換基依存性と、光還元反応速度の置換基依存性を関連させて考察することで、nπ*、ππ*状態間に熱平衡が達成されているというモデルを提唱した。本モデルにより時間分解赤外スペクトルの溶媒依存性や温度依存性がうまく説明できることが示されている。

第6章では、T1状態アセトフェノン類の赤外振動スペクトルについて述べている。アセトフェノン及びその4'-メチル置換体では振動バンドが消失することがわかり、これはnπ*とππ*状態の接近によると推論し、熱平衡の存在を支持するものと論じている。第7章は、以上の研究成果のまとめである。

本研究により、時間分解近・中赤外分光装置の性能が大幅に向上され、さらに、低エネルギー励起三重項状態の新しい観測方法として、その有用性が示された。今回取り扱われた芳香族カルボニル化合物は光化学のプロトタイプとして重要であり、その光化学反応機構に関して従来の定説を覆し得る新たな実験的証拠を呈示した本論文の業績は高く評価できる。

本論文第4章はChemical Physics Letters誌に公表済み(濱口宏夫、佐藤伸との共著)であるが、論文提出者が主体となって実験および解析を行なっており、その寄与が十分であるので、学位論文の一部とすることに何ら問題はないと判断する。

以上の理由から、論文提出者藪本宗士に博士(理学)の学位を授与することが適当であると認める。