Optmisation of cadmium (II) removal onto biodegradable composite using artificial neural networks and response surface methodology: quantum chemical performance.

A multifunctional grafted cellulose nanocrystals derivative adsorbent (composites) with carboxyl, amide, and secondary amino groups was successfully developed for Cd2+ removal. The characteristics of CNCs, chitosan, and nanocomposites were determined using FTIR, TGA, SEM, and BET. The approaches of artificial intelligence and Response Surface Methodology modeling were employed, as well as how well they predicted response (adsorption capacity). The adsorption isotherm and kinetic models were applied to comprehend the process further. Statistical results demonstrated that The response surface model approach performed better than the artificial neural network model approach. The adsorption capacity was 440.01 mg/g with a starting pH of 5.65, a duration of contact of 315 minutes, a starting concentration of 333 mg/L, and an adsorbent dose of 16.93 mg. The FTIR examination revealed that the functional groups of the nanocomposites were equivalent to those of CNCs and chitosan; however, the nanocomposites were more thermally stable than CNCs and chitosan. The nanocomposites' SEM pictures revealed a porous structure, thin particle size, and needle-like shape. The Langmuir model explains the spontaneous nature of the adsorption process, and chemisorption served as the primary control. According to the Dubinin-Radushkevich Model, to adsorb Cd2+, the energy required is larger than 8 kJ mol -1, suggesting that the chemisorption mechanism was involved. The adsorption kinetics were established using the pseudo-second-order rate model. HOMO-LUMO energy binding differences were used to find the best locations for adsorption.