Deconvolution of fluorescence spectra: contribution to the structural analysis of complex molecules.

Fluorescence spectroscopy is a sensitive analytical tool in the studies of both simple and complex molecular structures. In complex molecules, however, determining the number and position of components may give a specific insight into the structure, complementary to the other analytical techniques. We applied log-normal model to analyze fluorescence of simple monofluorophore molecule. In order to analyze spectra where both fluorophores and Raman emission bands were present, we developed a method obtained by combination of the symmetric, Gaussian, for Raman and asymmetric, log-normal model, for fluorescence, applicable to the molecules of different complexity. Technically, for each sample we varied excitation wavelength with 5 nm step and recorded the corresponding emission spectra. They were subsequently used for component analysis. Position of each component was plotted against the excitation wavelength. Applying this approach we could identify minimal number of components having stable positions, while their approximate probability density (APD) in a spectral series was correlated with the probable number of fluorophores in the molecule. The method was tested on molecules containing different number of fluorophores: monomers involved in the synthesis of plant polymer lignin-coniferyl alcohol (one fluorophore), ferulic acid (two fluorophores) and on lignin model compound produced from these monomers (many fluorophores). All investigated species belong to benzene-substituted class of compounds, and it is reasonable to assume that they have similar fluorescence band contour. We also report the results of environmental scanning electron microscopy (ESEM) studies showing multilayered dehydrogenative polymer (DHP) structure, in order to show complexity of the polymer. Our results present complementarity of these two approaches in the structural studies of the lignin model compound.