Prediction of the equilibrium moisture content based on the chemical composition and crystallinity of natural fibres.

Natural fibre-based materials offer various advantages compared to synthetic fibres, however their applications are limited mainly due to their hygroscopic properties, which are affected by their chemical composition, microstructure and the porosity of the plant cells of which the fibre is composed. Therefore, this work investigates the hygroscopic behavior of natural fibres to obtain a better understanding of the relation of the chemical composition of the fibres, their crystallinity, and their equilibrium moisture content. The crystallinity index was determined to include amorphous cellulose into the developed models. Nine biomass samples were selected (flax, hemp, jute, spruce, bamboo, corn stalks, palm leaves, rice husk and wheat straw) to construct models via linear regression to predict the moisture sorption behavior of natural fibres. Thorough statistical (ANOVA, RMSE) analysis showed that the developed models are relevant and descriptive. From all major plant cell wall constituents (lignin, crystalline cellulose, amorphous cellulose and hemicellulose), it is hemicellulose's hygroscopicity that is largely responsible for the moisture uptake of the fibres, with (amorphous) cellulose and lignin playing a (much) smaller role. This study has improved the understanding of the hygroscopic behavior of natural fibres, and is important for optimal application of these fibres in composite materials. • A linear models predicting the EMC absorption and desorption were constructed. • Hemicellulose is the main cell wall constituent affecting EMC desorption and absorption. • Non-crystalline cellulose affects the sorption significantly more than the overall cellulose content. [ABSTRACT FROM AUTHOR]