Development and applications of multidimensional fluorescence spectroscopy in time-dependent systems

Abstract

Fluorescence detection has been important in the analysis of time-dependent processes due to the high sensitivity and rapid response of the technique. However, the acquired data only provide information about the kinetic parameter being examined. The video fluorometer (VF) is a novel instrument capable of acquiring a complete Emission-Excitation Matrix (EEM) in 0.5 s. This instrument has proven effective for multicomponent analysis, yet little use has been made of its inherent speed. In this dissertation we will develop the video fluorometer as a probe of time-dependent phenomena. Since large amounts of data are acquired in a single kinetic run, a key consideration will be reduction of these large data sets. Alternative forms of data display will be discussed, particularly reduced forms of the data which allow quick preview of spectral properties. A concurrent need is the ability to spectrally deconvolute EEMs containing several components. This will be shown by exploiting the fact that a multicomponent EEM is a linear combination of single-component EEMs. The application of fluorescence quenching to data reduction will also be demonstrated. An in depth discussion of the problems inherent in the inner-filter effect will be included, along with a possible solution to the problem. These data reduction techniques will be applied to the study of the photochemical reaction of polychlorinated alkanes with anthracene. The reaction rate as a function of both chlorinated reactant and photoproduct will be examined along with the spectral properties of the product to arrive at a reasonable reaction mechanism. The VF will be applied to detection in High Performance Liquid Chromatography (HPLC). An interactive system will be designed to collect and preview the data to find chromatographic peaks. The spectra of these peaks can then be examined to identify each component. A 16-component mixture will demonstrate that one can identify components by spectral properties alone.