Adsorption of crystal violet dye onto a chemically modified NaOH–KOH orange peel–based adsorbent: Experimental study of kinetics and isotherms supported by ANN-based prediction

The discharge of water containing dyes has been a global problem for many years, and the treatment of this water has therefore become an essential requirement, necessitating well-researched techniques. This study presents the performance and adsorption capacity for crystal violet (CV) dye in an aqueous solution using an adsorbent produced from orange peels (OPCMA) chemically modified with sodium hydroxide (NaOH) and potassium hydroxide (KOH). A parametric study was experimentally conducted to optimize the studied parameters, such as contact time, pH, adsorbent dosage, initial dye concentration, and temperature. The adsorption study was followed by modeling of the adsorption isotherms and kinetics. In addition, an artificial neural network (ANN) implemented using JMP Pro 17 software was used to validate the experimentally determined dye removal rate and to estimate the effects of the studied parameters on the dye removal percentage. The experimental study showed a removal rate of 88.18% after 60 minutes, at a solution pH of 5.79, an adsorbent mass of 0.175 g, an initial dye concentration of 100 mg/L, and ambient temperature. Modeling of the adsorption isotherms showed that the Langmuir model was the most appropriate (R² = 0.996), giving a maximum adsorption capacity of 58.51 mg/g, while the pseudo-second-order model also showed a very good correlation coefficient (R² = 0.996), reflecting the excellent adequacy of the model. According to the thermodynamic study, the crystal violet removal process was spontaneous and exothermic. Modeling of the experimental data on the adsorption of crystal violet dye by OPCMA using a three-layer artificial neural network model showed that this model was valid and representative, with a sufficiently high correlation coefficient between the experimental data and the values predicted by the model (R² = 0.997).