Near infrared hyperspectral imaging and MCR-ALS applied for mapping chemical composition of the wood specie Swietenia Macrophylla King (Mahogany) at microscopic level

Abstract Near infrared (NIR) spectroscopy offers an efficient method for the characterization of solid wood. Nowadays, it is particularly relevant to the development of new methods that allow the determination of chemical composition in wood at the microscopic level, in order to enable the estimation of the distribution of the main chemical components at the mapped area. In this work, NIR hyperspectral imaging was applied for the determination of the distribution of holocellulose (cellulose + hemicellulose), lignin and extractives at the three grown directions (tangential, transversal and radial) of the specie Swietenia macrophylla King (Mahogany). The concentration maps were obtained by applying multivariate curve resolution-alternating least squares (MCR-ALS). NIR spectra recovered by MCR-ALS for cellulose, lignin and extractives showed band signals that agreed with the ones observed in reference spectra or in the literature. The mapped area presented different anatomical structures, which makes possible to observe the distributions of holocellulose, lignin and extractives in these different structures. The concentration maps for holocellulose showed that the fibers and the vascular line are the anatomical structures with the highest concentrations of this compound, which concentration varied from 30 to 91% (w/w). The concentration maps for extractives and lignin also showed high concentrations in vascular line, while rays showed the lowest concentrations. Considering the mapped area, lignin and extractives presented relative concentrations varying from 16 to 53% (w/w) and 1 to 16% (w/w), respectively. It was observed that the estimated concentrations in the maps agreed with the anatomic functions of the structures observed in the images.