Chitosan–Resole–Pectin Aerogel in Methylene Blue Removal: Modeling and Optimization Using an Artificial Neuron Network

• Published in: ChemEngineering 2023-09, Vol.7 (5), p.82
• Main Authors: Flores-Gómez, Jean, Villegas-Ruvalcaba, Mario, Blancas-Flores, José, Morales-Rivera, Juan
• Format: Article
• Language: English
• Subjects: Acids; Adsorbents; Adsorption; aerogel; Aerogels; Algorithms; Analysis; ANN; Artificial intelligence; Artificial neural networks; Cation exchanging; Chemical composition; Chemical engineering; Chemical oxygen demand; Chitosan; chitosan–resole–pectin; Correlation coefficients; Dyes; Fourier transforms; Functional groups; Graphene; Heat treatment; Heuristic methods; Hydrogen; Mathematical models; Methylene blue; Modelling; Neural networks; Neurons; Optimization; Parameters; Pectin; Phenols; Pollutants; Polysaccharides; Porous materials; removal; Sol-gel processes; Spectrum analysis
• ISSN: 2305-7084; 2305-7084
• DOI: 10.3390/chemengineering7050082

In this study, a novel chitosan–resole–pectin aerogel (CS–R–P) was created from a sol–gel reaction with a solution of Cs and P with resole by a freeze-drying technique, and this adsorbent was proposed for the removal of methylene blue (MB). In addition, with the use of an artificial intelligence technique known as an artificial neural network (ANN), this material was modeled and optimized. Its physical morphology and chemical composition were also characterized with FTIR and XPS, and its adsorption properties were analyzed. For modeling the adsorption process, three main parameters were used: the chitosan–resole–pectin concentration (45–75%), thermal treatment (6–36 h), and known concentrations of methylene blue (25–50 and 100 mg/L), established on the Box–Behnken design. The ANN was coupled with the improved gray wolf optimization (IWGO) metaheuristic algorithm, achieving a correlation coefficient of R2 = 0.99. The characterization indicates that the surface of the aerogels was micro- and mesoporous, the resole gave physical stability, and the polysaccharide base delivered the functional groups necessary for dye adsorption; the aerogels were successful dye adsorbents with a qe of 12.44 mg/g. Finally, the physical and chemical sorption was ascertainable with an adsorption that followed pseudo-second-order kinetics. The MB adsorption was clearly occurring though cation exchange and hydrogen binding as observed in the chemical composition. The ANN with the gray wolf optimizer was used for the prediction of the best operating parameters for MB removal, applying the following conditions—the CS–R–P aerogel concentration (52/30/18), the thermal treatment (9.12 h), and the initial concentration of methylene blue (37 mg/L)—achieving a 94.6% removal. These conclusions suggest that using artificial intelligence such as an ANN can provide an efficient and practical model for maximizing the removal action of new aerogels based on chitosan.