Imaging fingerprinting of excitation emission matrices

The spectral fingerprinting of the excitation emission matrix (EEM) of fluorescent substances is demonstrated using polychromatic light sources and tri-chromatic image detectors. A model of the measured fingerprints explaining their features and classification performance, based on the polychromatic excitation of the indicators is proposed. Substantial amount of spectral information is retained in the fingerprints as corroborated by multivariate analysis and experimental conditions that favor such situation are identified. In average, for five different substances, the model shows a fitting goodness measured by the Pearson's r coefficient and the root mean square deviation of 0.8541 and 0.0247 respectively, while principal component classification patterns satisfactorily compare with the EEM spectroscopy classification and respectively explain 96% and 93% of the information in the fist two principal components. The measurements can be performed using regular computer screens as illumination and web cameras as detectors, which constitute ubiquitous and affordable platforms compatible with distributed evaluations, in contrast to regular instrumentation for EEM measurements.