Elaboration des composites a base de cendres de balles de riz et d’une argile smectique : application dans l’adsorption du β-carotene et des acides gras libres de l’huile de palme

Abstract

The present work aimed to synthesize silica-smectite composites via three different synthetic pathways. Thus, the 20 µm enriched fraction of a smectite clay from Bana was respectively mixed with (1) silica sludge from rice husk ash after acid treatment (composite CSS1), (2) an aqueous sodium silicate solution from the alkaline dissolution of silica sludge (composite CSS2), and (3) a nanosilica powder obtained after the hydroxylation/polymerization of a sodium silicate solution (composite CSS3). Products were investigated by XRD; FTIR spectroscopy, SEM/EDS, and SSA/BET measurements. All composites showed a heterogeneous morphology, where the distribution of silica particles in the clay matrix changed with each synthetic pathway: For CSS1, a predominantly three dimensional-like structure with a phyllosilicate matrix skeleton was obtained. For CSS2, an amorphous compound with limited intergranular cohesion containing silicate agglomerates intercalated between clay sheets was found and CSS3 has a packed morphology with a regular distribution of silica phases in the clay matrix . In all the synthesized composites, the amorphous silica phase was identified, with a potential higher reactivity and SSA of 228, 257, and 300 m²/g for CSS1, CSS2 and CSS3, respectively. Correspondingly, the microstructure evidenced both an increased porosity and an increase in chemically active sites due to the presence of additionnal silanol and siloxane groups from silica. Consequently, the obtained products are potential multifunctional materials. As application, they were used in the adsorption of palm oil β-carotene and free fatty acids (FFA). For β-carotene adsorption, the optimal conditions for all the adsorbents were 2 % (w/w), 40 min and 95 °C for adsorbent dosage, contact time, and temperature respectively. These conditions yielded the bleaching capacities of 64 % for CSS1, 79 % for CSS2 and 92 % for CSS3. The process was found to follow pseudo-2nd order, intraparticle diffusion and Freudlich models and be spontaneous and endothermic. For FFA adsorption, Response Surface Methodology has been investigated considering (20-60 min), (1-2 %) and (60 - 90 °C) as factors‘ domains in Box-Berhken matrix. All those factors were significant according to Pareto‘s charts. Parameters from ANOVA were R2 = 99.31 % and p-values < 0.05 indicating that more than 99% of acid value variation could be explained by the model based on 95 % confidence. The theoritical optimal conditions were found to be t = 60 min, T = 90 °C and d = 2 %. A good correlation between the predicted and experimental values was also observed.