【Introduction】

Ionic liquids attract much attention recently as a new class of liquids. They have many interesting properties that are not associated with usual molecular liquids. Recent intensive physicochemical studies have suggested that these properties originate from specific local structures formed in ionic liquids. However, there is no unified view on the nature of these local structures. In this study, I try to further elucidate the nature of local structures in ionic liquids with a focus placed on the lack of correlation between the macroscopic thermal property and the microscopic energy transfer that is observed with picosecond time-resolved Raman spectroscopy. The volume of the local structure formed in ionic liquids is estimated from the analysis based on the diffusion equation of heat.

Picosecond time-resolved Raman spectroscopy provides a direct means to monitor the microscopic energy transfer in solutions; the 1570 cm(-1) Raman band of S1 trans-stilbene can be used as a "picosecond Raman thermometer" because the peak position of this band changes linearly with temperature. With this "picosecond Raman thermometer," the cooling kinetics of trans-stilbene has been observed for a number of molecular liquids after it is photoexcited to the S1 state with excess vibrational energy. The observed cooling rates correlate very well with the thermal diffusivities, indicating that they are controlled by the thermal diffusion in the solvents around the S1 trans stilbene molecule.

In the present thesis (partially in the Master thesis work), I extend the picosecond time-resolved Raman study of vibrational cooling of S1 trans-stilbene to ionic liquids. The observed vibrational cooling rates and the thermal diffusivities are plotted for ionic liquids as well as for molecular liquids in Fig.1. There is no obvious correlation between the two quantities for the seven ionic liquids, while a good correlation is observed for the molecular liquids. The vibrational cooling rates in the ionic liquids are similar to those of normal alkanes, although their thermal diffusivities are considerably smaller. The presence of local structures in ionic liquids may well disturb a continuous heat transfer from the heat source S1 trans-stilbene to outer solvent molecules. If heat transfer is slow at the boundary between the neighboring local structures, macroscopic thermal diffusion is not determined by the microscopic energy transfer represented by the cooling rate observed by the Raman thermometer.

【Experimental】

I constructed a picosecond time-resolved Raman spectrometer with two independently tunable light sources for the pump and probe pulses (Fig. 2). The output from a femtosecond mode-locked Ti:sapphire oscillator was amplified by a Ti:sapphire regenerative-multipass amplifier. A half of the amplified pulse was separated and delivered to a BBO crystal. The frequency-doubled output from the BBO crystal was used to pump an optical parametric amplifier (OPA) to generate the Raman probe pulses at 592 nm. The residual of the 800 nm amplified pulse pumped another OPA, which generated the pump pulse of 297 nm or 325 nm. For determining the precise peak position of the Raman band in the excess energy dependence measurement, I constructed a 4-f band pass filter which improved the spectral resolution to 8 cm(-1). The time resolution estimated from pump/probe cross correlation was 4.0 ps. Wavelength of the pump pulse was set to 297 and 325 nm, corresponding to the excess energy of 2500 and 0 cm(-1), respectively. The pump and probe pulses, traveling the common light path after being combined with a dichroic mirror, were focused onto the sample solution, which was circulated through a dye laser jet nozzle. Rayleigh scattering and fluorescence in the ultraviolet region were eliminated before the spectrograph by an optical narrow band-rejection filter and an optical color filter. The scattered light was dispersed by a single imaging spectrograph and detected by a liquid-nitrogen cooled CCD detector. The sample concentration of trans-stilbene was 5.0×10-3 mol dm-3. The ionic liquid used in the experiment was C2mimTf2N (1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide).

【Results and Discussion】

Vibrational Cooling Process with and without Excess Energy

I have measured the vibrational cooling kinetics of S1 trans-stilbene with and without the vibrational excess energy, in order to examine whether or not the excess energy reaches the boundary of the local structure by diffusion. The observed time-resolved Raman spectra, with and without the excess energy, are shown in Fig.3. Two sets of cooling kinetics obtained from the two sets of time-resolved Raman spectra are shown in Fig. 4. The result for the 297 nm excitation, with an excess energy of 2500 cm(-1), shows a large temperature change reflecting the vibrational cooling process. No large change is observed for the excitation with the 325 nm light, which carries only very small excess energy. Peak positions after 50 ps with and without the excess energy agree with each other within the experimental uncertainties. This fact indicates that the temperature increased at the photoexcitation recovers to the room temperature in 50 ps. We thus conclude that the excess energy is fully dissipated within a local structure without reaching the boundary. The volume of the local structure in ionic liquids is large enough to keep the temperature increase at 50 ps below the detection limit.

Numerical Simulation of Thermal Diffusion in the Local Structure in C2mimTf2N

I have analyzed the observed vibrational cooling kinetics with a numerical model based on the diffusion equation of heat, in order to estimate the size of the local structure in C2mimTf2N. Thermal diffusion within a local structure is modeled with the heat dissipation from a heat source in a finite volume. The heat source is regarded as a homogeneous distribution of temperature rise in a rectangular solid. This heat source represents S1 trans-stilbene and the nearest solvent ions that share the excess energy immediately after the photoexcitation. The finite volume of the local structure in ionic liquids is modeled by a cube with a length L that corresponds to the diameter of the local structure. At the boundary of this cube, heat flow is reflected to the opposite direction from the incident. Under these conditions, I simulated the thermal diffusion in the local structure and compared the resultant temperature change with the observed cooling kinetics (Fig. 4). It is obvious that the simulated cooling kinetics with L =10 nm reproduces the observed cooling kinetics very well, showing a complete recovery to the room temperature at time delays larger than 50 ps. However, the simulated curve stops cooling before reaching the room temperature if the L value is 5.0 nm. Simulations with smaller L values indicate that the temperature does not cool down to the room temperature because the heat remains within the local structure. This comparison between the observed and the simulated cooling kinetics shows that the diameter of the local structure is expected to be 10 nm or larger. I have thus estimated the minimum size of the local structure in C2mimTf2N, 10 nm, with picosecond timeresolved Raman spectroscopy and a numerical simulation based on the diffusion equation of heat.

Fig 1. Relationship between the observed vibrational cooling rates and thermal diffusivity

Fig. 2 Layout of the constructed picosecond timeresolved Raman spectrometer

Fig. 3 Time-resolved Raman spectra of S1 transstilbene in C2mimTf2N after photoexcitation (a) excitation: 297 nm, (b) excitation: 325 nm

Fig. 4 Cooling kinetics with (filled circle) and without (open square) the excess energy and simulated with two diameters of the local structure,L, 5.0 (dotted curve) and 10 nm (dotted line), and the room temperature (solid line)

本論文は、ピコ秒時間分解ラマン分光法を用いた、イオン液体中の局所構造に関する研究について記述されており、全5章から構成される。

第1章では導入として、本研究の目的が、イオン液体の局所構造の理解にあり、それをミクロおよびマクロな熱的性質の比較から議論すること述べられている。溶液中でのマクロな熱の移動の速さを示す熱拡散率は、過渡回折格子法と呼ばれる手法で測定され、ミクロな熱拡散はS1 トランススチルベンを「ピコ秒ラマン温度計」として用いて動冷却過程を測定することで観測することが可能であることが述べられている。とくに、ピコ秒時間分解ラマン分光法を用いてイオン液体中のミクロな熱拡散過程を観測することが、局所構造に関する理解を深める上で重要であることが述べられている。

第2章では、 実験条件がまとめられており、主にピコ秒ラマン分光装置改良の詳細とその性能について述べられている。2台の光パラメトリック増幅器を用いることで、ラマンプローブ光とポンプ光の波長をそれぞれ自由に変えることが可能な装置である。また、自作した波長可変4-fバンドパスフィルターにより、プローブ光のスペクトルを狭帯域化し、より高い分解能でのピコ秒時間分解ラマンスペクトルの取得を可能にしたことが、例を挙げて示されている。

第3章では、ピコ秒時間分解ラマン分光法を用いて、イオン液体中のS1 トランススチルベンの振動冷却速度を測定し、過渡回折格子法で測定された熱拡散率と比較した結果が議論されている。イオン液体の場合、熱拡散率に比べて振動冷却速度が大きく、また、2つの量に相関がないことがわかった。この結果は、局所構造が存在するイオン液体中では、局所構造の境界を越える遅い熱の移動により、マクロなスケールで起こる熱拡散過程が遅くなるためであると解釈されている。

第4章では、イオン液体中で光励起余剰エネルギーの有無によるS1 トランススチルベンの振動冷却過程の違いを観測し、その結果を、熱拡散方程式の解と比較することにより、局所構造の大きさを見積もる試みが述べられている。S1 トランススチルベンから発生する熱が、境界に達する前にほとんど拡散してしまう程局所構造は十分に大きく、その大きさは10 nm以上であると結論されている。

第5章は以上の研究成果のまとめである。

本研究により、イオン液体中では、局所構造の存在により、ミクロな熱拡散とマクロな熱拡散が直接相関しないことが示された。イオン液体は、透明で局所構造が光学顕微鏡などでは検出不可能であることと、第4章の結果をあわせて、局所構造の大きさが10から100 nmのオーダの範囲であることを結論された。ピコ秒時間分解ラマン分光法を用いることで、溶液中のミクロな熱的性質と、マクロな熱的性質を区別して議論することが可能になったことは、溶液中の不均一な系を調べる新しい手法が開発されたことを意味する。このような新しい手法とその有用性を提示した本論文の内容は高く評価できる。

本論文第3章の一部は、Chemistry Letters誌に投稿済み(岩田耕一、〓田雄太、〓口宏夫との共著)である。第3-4章の主要部分は、Journal of Chemical Physics 誌に受理済み、公表予定(岩田耕一、西山嘉男、木村佳文、〓口宏夫との共著)である。これらの報文では、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、論文提出者吉田匡佑に博士(理学)の学位を授与することが適当であると認める。