【作者】 张新星；

【导师】 赵丽娟；

【作者基本信息】 南开大学 ， 凝聚态物理， 2013， 博士

【摘要】 离子液体(IL)的溶剂化动力学过程具有典型的双相性。其快响应过程通常在1ps内结束,而剩余部分则在纳秒时域内弛豫。为观测IL的超快动力学过程,宽光谱荧光上转换系统(FLUPS)经调节,在多方面得到改进。通过采用波前倾斜后的红外门脉冲(1340nm)以及光度和时间零点校准,系统可以测量到80fs时间分辨率且高信噪比的宽带(425-750nm)荧光光谱。结合FLUPS和时间相关单光子计数(TCSPC)两种测量手段,可以完整地观测到荧光团在IL中的斯托克斯位移动力学过程。利用该手段,本文测量了香豆素153(C153)在21种咪唑、吡咯烷等类型离子液体中的完整溶剂化响应函数,其中快响应过程通常占整体动力学过程的10%-40%。快响应过程时间与离子对约化质量的相关性表明,这部分动力学过程是由离子的惯性运动引起的。溶剂化弛豫中的慢响应部分则在较宽时域上分布且与IL粘滞度相关联,证明了慢响应过程源于溶剂的扩散性结构重组。为进一步研究溶剂化动力学与溶剂本身介电弛豫过程的关系,简单介电连续场模型被引入到溶剂化响应函数的预测中。溶剂化响应函数的预测值通常快于实验值2-4倍。为了深入理解溶剂化效应与介电弛豫过程之间的相关性,本文还测量“基准”溶质C153在离子液体(1-丁基-3-甲基咪唑四氟硼酸盐)与两种极性溶剂(乙氰和水)混合液中的溶剂化响应函数,并且均未找到证明C153在这两种混合体系中优先溶剂化效应的有力证据。C153在上述混合体系中的溶剂化及转动时间与粘滞度呈现良好的相关性。此外,文中还利用混合体系的介电弛豫谱(频域范围为200MHz-89GHz)对溶剂化响应函数作了预测,并且通过与实验值比较,进一步讨论了介电连续场模型的正确性。在IL+乙氰混合液中,介电连续场模型的预测精度与纯离子液体中的情况相似：快响应部分的预测值与实验值相符,但慢响应过程的预测值过快。预测结果的准确度与混合体系中乙氰含量无关。与此相反,介电连续场模型无法正确预测IL+水混合体系的溶剂化响应函数。介电连续场模型对溶剂化响应函数的预测值仅依赖于溶剂的介电谱,与溶质探针无关。因而为进一步分析该理论,本文测量了另外一种溶致变色探针—4-氨基邻苯二甲酰亚胺(4AP)在离子液体中的溶剂化响应函数。总体来说,4AP的溶剂化动力学过程要系统性地慢于C153在同种IL中的观测值。这种差异性被认为是由溶质探针的自身运动造成的。文中所提出的“转动修正”利用转动相关函数,有效地缩小了两者所观测到的溶剂化响应函数间的差异。利用文献数据4-二甲基胺基-4’-氰芪(DCS)和C153,“转动修正”的正确性得到进一步验证。此外,仅通过调节介电谱数据中的电导率,介电连续场模型就能很好地预测转动修正后的溶剂化响应函数。除实验研究外,本文还利用了数值和解析计算扩展了介电连续场模型的应用。若将溶剂化响应函数用多指数函数的形式拟合,通过解析形式的模型算法,拟合参量可以直接转换为多德拜函数+电导率项形式的广义介电量。该方法预测出的介电响应函数与实验值非常一致,但电导率则普遍小于测量值。解析算法揭示了溶剂化时间(τsolv)与静态电导率σ0之间的简单关联性。文中通过公式推导验证了该关系的正确性,并利用C153在34种离子液体中相关数据得到了经验关系式：ln(<τsolv>/ps)=4.37-0.92ln(σ0/Sm-1)。更多还原

【Abstract】 Solvation dynamics in ionic liquids (ILs) is typically biphasic, consisting of a fast component completed in1ps and the rest extending up to the nanosecond domain. To observe the fast dynamics, an instrument for broadband fluorescence upconversion spectroscopy (FLUPS) is developed. With tilted1340nm gate pulses, photometric and time-zero calibration, highly time-resolved (80fs), broad spectral range (425-750nm) and completely background free spectra can be obtained.The full dynamic Stokes shift is measured by combining the techniques of FLUPS and time-correlated single photon counting (TCSPC). In this way, the complete solvation response function of coumarin153(C153) is determined in21imidazolium, pyrollidinium, and assorted other ILs. The fast component is found to account for10%-40%of the response. The time constant associated with this component is correlated to ion reduced mass, indicating that it is caused by the ions’inertial motions. A much slower component, which relaxes over a broad time range, completes the solvent relaxation. Its origins are connected to diffusive, structural reorganization, based on the fact that its time is well correlated to the IL viscosity. A simple dielectric continuum model is introduced to investigate the relationship between dielectric relaxation and solvation dynamics. The dielectric continuum model is found to over-estimate the speed of solvation by factors of2-4.To obtain additional perspective on the connection between solvation and dielectric relaxation, mixtures of an ionic liquid,1-butyl-3-methyl-imidazolium tetrafluoroborate ([Im41][BF4]), and two polar solvents, acetonitrile and water, are also studied. The physical properties of both mixtures vary systemically with the volume fraction of [Im41][BF4]. By using C153as a "standard" probe, the solvation response function is examined and no clear evidence is found to confirm the suspicion of preferential solvation. Both solvation and rotational times are nicely correlated to solution viscosity. Experimental dielectric data over the frequency range200MHz-89GHz are used to predict the solvation response and further test the dielectric continuum model. In the case of acetonitrile+IL mixtures, the accuracy of continuum model predictions was comparable to that in neat ILs:the fast component is well predicted while the speed of the slow part is overestimated. The quality of these predictions was equally good at high and low acetonitrile content. In contrast, the continuum model totally failed in the IL+water mixtures at high water content.The dielectric continuum model as applied above predicts the same dynamics for all dipolar solutes. To further test this prediction the solvation response of another solvatochromic probe,4-aminophthalamide (4AP), is measured in four ionic liquids. The4-AP response functions are systematically slower than those of C153in the same ionic liquids. The origin of this difference was thought to arise from the effect of solute motion on solvation. A correction for solute motion using measured rotational correlation functions significantly reduces the differences observed between C153and4AP. Comparisons between literature data on4-dimethylamino4’-cyanostilbene (DCS) and C153support the use of this rotational correction. The rotationally corrected solvation response functions of both C153and4AP can be reproduced using dielectric continuum predictions by allowing the conductivity used in the dielectric modeling to differ from experimental values.In addition to experimental studies the dielectric continuum model is investigated both numerically and analytically. An analytical method for inverting a multi-exponential representation of the solvation response to obtain a description of the permittivity expressed as a sum of a conductivity term+multiple Debye terms. The computed conductivity of C153in neat ILs is found to be systematically smaller than the bulk value, while the predicted permittivity agrees well with the data from bulk dielectric measurements. This analytical approach also reveals a simple relationship between the integral solvation time <τsolv> predicted by the continuum model and the static conductivity σ0. Furthermore, a more general derivation of the same correlation is provided and data on C153solvation in34neat ILs are presented to support this prediction and provide the empirical counterpart:ln(<τsolv>/ps)=4.37-0.92ln(σ0/Sm-1).更多还原