Application of the kinetic and neuro-fuzzy logic models for methylene blue adsorption behaviour onto FlyAsh or GGBS geopolymer adsorbents in fixed-bed systems

: This study reports on the adsorption properties of alkali-a ctivated alumino-silicates derived from low calcium coal fly ash (G 1 ) or high calcium GGBS (G 2 ), prepared solely with an alkaline solution of NaOH/Na 2 SiO 3 (silica modulus, Ms 1.1). In order to confirm the adsorbents structural characteristics, FTIR spectral analysis, powder XRD patterns, and SEM-EDS for morphology were used. The equilibrium adsorption capacity (q e ) for Methylene Blue (MB) was evaluated for both samples using a flow reactor, under optimized conditions: 10 cm bed height, 5 mL/min flow rate, and an initial MB concentration of 10 ppm at pH 10. Under these conditions, adsorption capacities of 48 mg/g for G 1 and 38 mg/g for G 2 were achieved. Further, DOE-RSM data obtained with a quadratic model with a high regression coefficient (R 2 = 0.953) and an insignificant lack of fit (0.2426). Dye sorption equilibrium isotherms were modeled to explain the adsorption mechanism. Chemisorption of MB showed an excellent fit to the Thomas model, R 2 value of 0.991. In addition to experimental studies, Adaptive Neuro-Fuzzy Inference System integrated with a Genetic Algorithm (ANFIS-GA) was employed to model the nonlinear adsorption process and optimize the membership function parameters, thereby improving predictive performance. Upon training and simulation, the ANFIS-GA model achieved a high correlation coefficient (R 2 ) in the range of 0.96 – 0.97, and a low mean squared error (MSE). Thus, this waste based, low-cost adsorbents support SDGs related to water, including SDG#6 (sustainable wastewater treatment), SDG#12 (resource valorization and circular economy practices), SDG#15 (helps protect terrestrial ecosystems from contaminated effluent discharge) and Industry 4.0.