Enhanced biosorption of fluoride by extracted nanocellulose/polyvinyl alcohol composite in batch and fixed-bed system: ANN analysis and numerical modeling

The present investigation attempted to examine the defluoridation feasibility onto the extracted nanocellulose/PVA polymer composites. Nanocellulose were derived from sugarcane bagasse and blended with PVA (polyvinyl alcohol) polymer matrix. The defluoridation potential of nanocellulose/PVA was observed to be significantly dependent on the various operational factors including pH, time interval, etc. the Temkin isotherm (R² = 0.989) as well as the Langmuir isotherm equation (R² = 0.982) could well fit with the investigational data. Following the Langmuir isotherm, the maximum monolayer adsorption capacity for fluoride elimination at 25°C was obtained as 11.363 mg g⁻¹. The nature of rate-limiting steps involved in defluoridation process might be effectively predicted by pseudo-second-order kinetics. Values of thermodynamic state properties achieved as of the thermodynamic analysis showed that the defluoridation process was spontaneous, exothermic, and feasible. The diffusion and mass transfer study were estimated by following the Boyd’s model. Average effective diffusion coefficient (Dₑ) at various initial fluoride concentrations (4–10 mg L⁻¹) was obtained as 15.3343×10⁻⁷ m²s⁻¹ and the estimated magnitude of the mass-transfer coefficient (Kf) was 0.0346×10⁻⁹ m s⁻¹ (temperature = 298 K, C₀= 6 mgL⁻¹). An ANN (artificial neural network) model applied to optimize and simulate the defluoridation procedure. Furthermore, continuous flow column reactor was conducted to investigate the practical applicability of composites in the defluoridation process. The Yoon-Nelson and the Thomas model exhibited excellent conformity with the breakthrough curves. Nanocellulose/PVA satisfactorily eliminated fluoride from its aqueous solution and can be considered as a suitable bio-sorbent for defluoridation.