The aim of this work has been to ascertain the effect of hemicellulose
removal and bagasse drying on selectivity, extension and rate of soda
delignification. To this end, three bagasse samples (standard, air dr
ied and partially hydrolysed) were pulped at different temperature and
active alkali doses according to a three-level factorial design. It w
as concluded that bagasse delignification occurs only in two phases ('
'bulk'' and ''residual''), lignin being removed from,the very beginnin
g of pulping. The initial polysaccharides removal stage, present in wo
od delignification, does not take place for bagasse. A two phases expo
nential equation has been found suitable for modeling bagasse delignif
ication kinetics. Results confirm that bagasse drying renders delignif
ication difficult because of diffusion restrictions, while hemicellulo
ses removal makes it easier.