Simultaneous Double-Pass Absorption and Fluorescence Excitation–Emission Matrix Spectroscopy for Measurements of Reaction Kinetics

Abstract

A novel Hadamard-transform excitation–emission matrix (EEM) spectrometer generates two-dimensional (2D) fluorescence matrices at a data acquisition rate of over 6 EEMs per minute and with a spectral resolution of 5.3 nm. Using Fresnel reflections from the sample cell, we could record optical transmission spectra synchronously with the 2D EEMs. The spectrometer was integrated into a custom-designed stopped-flow injection device to collect visible absorption and fluorescence EEM spectra of reacting solutions. Two different kinetic studies on two rapidly evolving chemical reactions with multiple overlapping spectral components were conducted by collecting over 8400 absorption spectra and EEMs. The third-order rate constant for the demetalation of chlorophyll-a to pheophytin-a was experimentally determined to be 450 ± 20 M–2·s–1 as derived from a parallel factor (PARAFAC) analysis where absorption and fluorescence data were combined. A PARAFAC analysis of data collected from the insertion of a copper atom into pheophytin-a resulted in several absorbing components and only a single fluorescing component. A reaction model with an association complex and a sitting-a-top (SAT) complex as intermediates explained the absorption data, resulting in a sequence of second-order reactions with rate constants of 4.0 ± 0.4, 2.7 ± 0.3, and 0.28 ± 0.02 M–1·s–1, respectively. The rate constant of the fluorescence decrease was determined to be 1.7 ± 0.2 M–1·s–1, which is consistent with the fluorescence component being attributed to both the pheophytin-a and the association complex.

HPLC-Detector Based on Hadamard-Transform Fluorescence Excitation-Emission-Matrix Spectroscopy

“…Adding absorption detection, either using a separate detector or the same programmable light source, may further resolve ambiguities and also help detect nonfluorescent analytes. We recently demonstrated a detector that uses a similar programmable light source to obtain HT-absorption information synchronously with HT-fluorescence data in kinetic analysis …”

“…For several years, parallel factor analysis (PARAFAC) has been the tool of choice for the analysis of EEM spectral time-series as its specific restrictions are well-matched to the photophysical restrictions of fluorescence spectroscopy within the assumptions of Kasha’s rule, i.e., that strong emission occurs in a single, comparably narrow wavelength range for each fluorophore, somewhat independent of the excitation wavelength . While a typical multivariate analysis problem consists of a small number of EEM matrices (<100) with a large number of fluorescent components (>10), EEM analysis of chemical kinetics or, as in the present case, of chemical separation involves collection of hundreds or thousands of EEM matrices having only a very limited number of fluorescent components at one time (<6) . The advantages of PARAFAC lie in its speed and precision with which components can be attributed to fluorophores under these circumstances.…”

“…The advantages of PARAFAC lie in its speed and precision with which components can be attributed to fluorophores under these circumstances. These advantages are not unique to PARAFAC, of course, and other algorithms have been shown to perform similarly well …”

“…Recently, we introduced a much faster EEM spectrometer achieving for the first time an acquisition rate of higher than 30 spectra per minute. − This EEM instrument uses a fully programmable 31-channel light source and a conventional 2048-channel fluorescence photodiode array spectrometer. The programmable light source produces 31 different combinations of 31 wavelength channels based on Hadamard-matrix multiplexing .…”

“…The programmable light source produces 31 different combinations of 31 wavelength channels based on Hadamard-matrix multiplexing . A fluorescence spectrum is recorded for each combination (“bar code”) of excitation wavelength channels, and the resulting data cube is demultiplexed using the inverse Hadamard transform yielding the EEM. , …”