Physical and mechanical characterization of a new lignocellulosic material developed from <i>Borassus aethiopum</i> fruit shells with an epoxy matrix

This study investigates the reinforcement potential of Borassus aethiopum (BA) fruit shells in an epoxy matrix, with particular emphasis on the influence of particle size and filler loading on the physico-mechanical properties of the resulting bio-composites. Raw BA shells were characterized using scanning electron microscopy (SEM), X-ray micro-computed tomography, Vickers micro-indentation (HV0.1), Fourier-transform infrared spectroscopy (FTIR), and thermogravimetric analysis (TGA). Microstructural observations revealed a functionally graded and interlaced fibrous architecture typical of highly lignified lignocellulosic tissues. The three-dimensional micro-tomographic analysis highlighted a highly heterogeneous surface morphology with altitude variations ranging from −111 µm to +134 µm and peak-to-valley amplitudes exceeding 200 µm. These topographical features indicate a hierarchical organization of thick fiber bundles resulting in a naturally rough surface that can promote mechanical interlocking with polymer matrices. The shells also exhibited a heterogeneous micro-hardness distribution and an actual density ranging from 1.21 to 1.36 g·cm -3 . Bio-composites were fabricated using three particle size distributions: T1 (0.5–1.25 mm), T2 (1.25–2.5 mm), and T3 (2.5–4 mm), at filler loadings of 60, 70, and 80 %. Mechanical characterization through three-point bending tests and Shore D hardness, together with water absorption measurements, demonstrated that particle size strongly governs composite performance. The highest mechanical efficiency was obtained for 70 % T3 particles, yielding a modulus of elasticity of 8.7 ± 2.14 GPa, whereas 80 % T1 composites exhibited the lowest stiffness (2.87 ± 0.94 GPa) and the highest water uptake. These findings demonstrate that larger particles improve stiffness and dimensional stability, while smaller particles increase hygroscopic sensitivity. .