| This work included two studies: (1) employing multi-way fluorescence spectra, i.e. excitation-emission matrices (EEMs) to investigate four Humic substances (HS) samples, two obtained from Aldrich and Fluka, and two i.e. 1S102H and 1S103H obtained from International Humic Substances Society (IHSS); (2) employing multi-way fluorescence spectra to investigate fluorescence resonance energy transfer in micelles.;In HS study, the pH-dependence of the four HS samples were investigated within a broad pH range, from pH=2 through pH=11, by means of chemometrics approach, Parallel factor analysis (PARAFAC), that modeled the steady-state EEMs into the superposition of individual component's contributions. For samples obtained from Aldrich and Fluka, PARAFAC modeling revealed that each of them contained two primary fluorescent components. These components showed broad and featureless emission spectra and structured excitation spectra, which is a typical characteristic of HS fluorescence at room temperature in aqueous solutions. Besides two fluorescent components, each of 1S102H and 1S103H contained a third component that showed tryptophan-like spectra. The emission of the tryptophan-like component was structured, blue-shifted compared to tryptophan normal emission in water. The tryptophan-like component emission in 1S102H and 1S103H did not show obvious pH-dependence. It was concluded that tryptophan-like fluorophore in 1S102H and 1S103H existed in a relatively non-polar environment. Strong "shielding" of the tryptophan-like component by HS was suspected, which protected the tryptophan-like component from influence of pH.;Besides the pH-dependence study, the Fluka HS sample, referred to as Fluka humic acid (FHA) binding with Cu(II) was investigated. Cu(II) binding induced fluorescence quenching of FHA. After employing PARAFAC to recover the individual component's fluorescence quenching, the Weber-Ryan HS-metal binding model was utilized to recover the Cu(II) binding parameters for each component in FHA. Such study was respectively carried out at pH=3, 4, 5 and 6. Recovered binding stability constants of the two components increased with pH, showing almost linear plot for pH vs. logarithm of stability constants. The slope of the plot was less than unity. A conclusion was drawn that each component in FHA was not "pure", containing protonated and deprotonated fluorescent binding sites. The recovered binding capacity towards Cu(II), however decreased from pH=2 to pH=4, an interesting phenomena that was compared to the literature reports and discussed.;Time-resolved multi-way fluorescence spectra, i.e. time-resolved EEMs (TREEMs) were collected and modeled by PARAFAC for FHA at pH=5, in absence of Cu (II) and in presence of 20 muM Cu(II) respectively. This study, as a supplementary work to the steady-state EEMs mentioned above, revealed that there were two primary fluorescent components in FHA, in accordance with steady-state results. Further decay time analysis revealed that in each primary component there were two sub-components showing distinct decay times. As Cu(II) was added to FHA solution, the decay times of the sub-components did not show obvious change, while fluorescence intensities accordingly decreased, indicative of static quenching induced by Cu(II). The existence of two sub-components in each primary component seemed to be in accordance with steady-state study results.;In study of fluorescence energy transfer, steady-state kinetics related to multi-way fluorescence was for the first time derived. Theory analysis first time predicted that PARAFAC could properly model the EEMs in presence of energy transfer. The EEMs should be a superposition of contributions from three "components": (1) donors, (2) acceptors directly excited by incident photons, and (3) acceptors excited by energy transfer from donors. In contrast, conventional PARAFAC theory in EEMs stated that only the EEMs composed of "independent" components could be modeled by PARAFAC.;In order to verify the theory, energy transfer between fluorophores in micelles was used as a model system to investigate. Biphenyl was selected as the energy donor, and 2,5-diphenyloxazole was selected as the energy acceptor. Surfactant sodium dodecyl sulfate (SDS) was used to form micelles. EEMs were collected for the fluorophores in SDS micellar solutions. PARAFAC modeling clearly showed that only three components were appropriate to fit the EEMs, consistent with steady-state kinetics prediction. One component corresponded to the acceptors excited by energy transfer in micelles. The other two components corresponded to donors and acceptors directly excited by incident photons, respectively.;In PARAFAC modeling for energy transfer EEMs, good fit gave negative core consistency (CC), a diagnostic parameter currently accepted and used for PARAFAC modeling in the literature. In the literature, CC diagnostic theory stated that proper PARAFAC modeling should have relatively high and positive CC. The observation in this work implied that currently used CC diagnostic might not be applicable to EEMs of systems having energy transfer. (Abstract shortened by UMI.)... |
