Morphological characteristics and composition of lipophilic extractives and lignin in Brazilian woods from different eucalypt hybrids

Abstract: The morphological and chemical characteristics of the woods from several eucalypt hybrids from the Brazilian Genolyptus program were studied. The hybrids selected for this study were Eucalyptus grandis × E. urophylla (IP), E. urophylla × E. urophylla (U1×U2), E. grandis ×[E. urophylla × E. globulus] (G1×UGL), and [E. dunnii × E. grandis]× E. urophylla (DG×U2). The analyses of the lipophilic extractives indicated a similar composition in all eucalypt hybrids, which were dominated by sitosterol, sitosterol esters and sitosteryl 3β-d-glucopyranoside. These compounds are responsible for pitch deposition during kraft pulping of eucalypt wood. Some quantitative differences were found in the abundances of different lipid classes, the wood from U1×U2 having the lowest amounts of these pitch-forming compounds. The chemical composition and structure of lignins were characterized by Py-GC/MS and 2D-NMR that confirmed the predominance of syringyl over guaiacyl units and only showed traces of p-hydroxyphenyl units in all the woods, with the highest S/G ratio for G1×UGL. The 2D-NMR spectra gave additional information about the inter-unit linkages in the lignin polymer. All the lignins showed a predominance of β-O-4′ ether linkages (75–79% of total side-chains), followed by β–β′ resinol-type linkages (9–11%) and lower amounts of β-5′ phenylcoumaran-type, β-1′ spirodienone-type linkages or β-1′ open substructures. The lignin from the hybrid G1×UGL presented also the highest proportion of β-O-4′ linkages, and therefore, it is foreseen that the wood from this hybrid will be more easily delignifiable than the other selected Brazilian eucalypt hybrids. In complement to these chemical analyses, the morphological characterization of fibers, vessels and fines revealed that hybrid eucalypt clone DG×U2 presented the most interesting properties for the manufacture of paper pulps and biofuels. [Copyright &y& Elsevier]