Résumé
In a biorefinery plant diagram, a grinding step is usually the prerequisite for biological conversion, allowing material densification and homogenization for transport and storage, or an increase of the accessible surface. Powders produced from plant biomass are heterogeneous in relation to native plant heterogeneity, and during grinding, dissociation often occurred at the tissue scale. The distribution of tissue within a single particle is required to understand, control and optimize the plant deconstruction through dry fractionation process. Therefore the tissular origin of powders must be identified at the particle level. If recognition of tissues is easy in sections of plant organs, it is not straightforward in powders. Taking advantage of the autofluorescence properties of cell wall components, we propose to use multispectral image analysis to identify the tissular origin of lignocellulosic particles. Maize stem was chosen as model of grass lignocellulose. The fluorescence variability of maize stem sections was first investigated. Images of the stem sections were acquired and chemometric approaches was implemented to reveal fluorescence variability without any a priori. Fluorescent profiles were extracted for the main tissues. Images of ground tissues, isolated from the same stem, were acquired using the same procedure. The fluorescence intensity profiles were analysed using principal component analysis. Similar variability was found in fluorescence profiles extracted from powders and tissues confirming the potential of fluorescence multispectral imaging to predict the tissue origin of particles.