From Coconut Coir to Bio-Based Insulation Panels: Factorial Design and Genetic-Algorithm Exploration of Pulp Yield and Thermal Conductivity

Abstract

Coconut coir is an abundant lignocellulosic agro-industrial residue that can be processed into fibrous pulp for exploratory bio-based insulation panels. This study used a complete three-level, three-factor (3^3) factorial design to examine the effects of phosphoric-acid pretreatment concentration, sodium-hydroxide cooking concentration, and bleaching chlorometric degree on gravimetric pulp yield and apparent thermal conductivity. Second-order regression equations and coefficient-magnitude (Pareto) rankings were used as exploratory tools to describe treatment effects, and a genetic algorithm was used to identify a candidate compromise within the tested factor bounds. Observed pulp yields ranged from 20.1% to 74.5%, and apparent thermal conductivity ranged from 0.035 to 0.170 W×m^-1K^-1. The candidate conditions gave model-predicted values of 92.4% pulp yield and 0.041 W m^-1K^-1 apparent thermal conductivity; these values were not independently confirmed experimentally. Because compositional analysis of the recovered solid was not performed, the material is described as treated fibrous pulp rather than purified cellulose. The results provide an exploratory basis for further optimization, which should include composition, moisture, and independent confirmation measurements before practical insulation performance is claimed.