Unraveling the spectral responses of reacting molecules that are distributed across the microdomains of microheterogeneous media requires specialized data analysis. Multichannel data acquisition facilitates data mining and mixture analysis because multivariate data sets are subject to matrix and tensor analysis methods that can separate signals into components without a priori knowledge of the number or nature of their spectral properties. 

Our goal in data mining/analysis software development has been to generate methods that preserve this generality while maximizing the incorporation of available information into the isolated component responses. Our past work focused on the characterization of complex media by resolution of responses of environment-sensitive spectral probes using data mining approaches. This strategy was used to separate the spectra and emission decays of differentially solvated polycyclic aromatic hydrocarbons (PAHs) in surfactant solutions, PAHs in mixtures, differentially solvated probe molecules molecules in melting lipid bilayers, hydrogen-bonded PRODAN excited states in protonated solvents, and tryptophan rotamers in proteins. Current work focuses on increasing the speed and efficiency of spectral resolution algorithms and development of software approaches to reducing data acquisition time. 

‍