Pattern recognition as a tool to discriminate softwood and hardwood bark fractions with different particle size

The aim of this study was to characterize the chemical composition of different granulometric fractions obtained by milling of the bark of two softwood [Norway spruce, Picea abies (L.) Karst., and Scots pine, Pinus sylvestris L.] and two hardwood species (birch, Betula pendula Roth, and eucalypt, Eucalyptus globulus Labill.), and to discriminate between them on the basis of their chemical composition content via pattern recognition techniques [principal component analysis (PCA), cluster analysis (CA), and discriminant analysis (DA)]. Bark chemical composition differed between species, and chemical variables could be used to differentiate between them. Size reduction yields granulometric fractions that are not chemically homogeneous and that can also be discriminated. Therefore, potential applications of bark in valorization programs have to carefully consider the species-specific composition and their size reduction patterns. PCA and CA were adequate tools to characterize the different bark fractions within each species. DA allowed identifying the bark samples according to species and independently from particle size. Pattern recognition statistical methods were shown to be useful tools to analyse bark fractions and chemically discriminate species and fractions.