Ultrasonic assisted removal of methylene blue on ultrasonically synthesized zinc hydroxide nanoparticles on activated carbon prepared from wood of cherry tree: Experimental design methodology and artificial neural network

Abstract

Abstract The zinc hydroxide nanoparticles was synthesized and loaded on activated carbon prepared from wood of cherry tree (Zn (OH)2-NPs-AC). Prepared NP based adsorbent was used to remove methylene blue (MB) from aqueous medium. The dependency of MB concentration, pH, adsorbent doses and sonication time on the extent of adsorption was investigated and optimized using response surface methodology (RSM) based on central composite design. Analysis of variance (ANOVA) was made to calculate coefficient of determination (R2). The best operation of conditions were determined for MB concentration (12.5 mg L− 1), pH (6), adsorbent mass (0.025 g) and sonication time (6.5 min). In addition, all the experimental data was used to train the artificial neural network (ANN) model. Performance evaluation of the ANN model by means of squared error (MSE), average absolute percent deviation (AAD%) and correlation coefficient (R2) depicted the experimental value of MSE = 0.0529, AAD = 0.1894% and R2 = 0.98. These values were better than that of obtained from RSM model (MSE = 2.7107, AAD = 1.470%, R2 = 0.9142). It was noted that the equilibrium isotherm data followed Langmuir model with high adsorption capacity. The adsorption kinetics was efficiently represented by combination of pseudo second order and intraparticle diffusion models.