Ion-exchange process for the removal of Ni (II) and Co (II) from wastewater using modified clinoptilolite: Modeling by response surface methodology and artificial neural network

Abstract

The present work attempts to study ion-exchange process for the removal of Ni (II) and Co (II) from wastewater using modified clinoptilolite with succinic acid and Ethylenediamine tetraacetic acid (EDTA). The clinoptilolite was characterized using (FTIR) infrared spectroscopy, (SEM) scanning electron microscope and (XRD) X-ray diffraction analysis. The process variables were modelled and optimized using surface response methodology-central composite design (RSM-CCD) and Artificial Neural Network (ANN). The Statistical metric such as absolute average deviation (AAD), average relative errors (ARE), coefficient of determination (R2), hybrid fractional error function (HYBRID), marquart’s percentage standard deviation (MPSD), mean squared error (MSE), pearson’s Chi-square measure (χ2), root means square errors (RMSE), sum of the absolute errors (SAE) and sum of the squares of errors (SSE) were evaluated and the models appeared acceptable. However, the statistical results indicated that ANN model is superior to the RSM-CCD model approach. The optimization results of the process variables by RSM-CCD model were obtained at pH of 6, an initial concentration of 425 ​mg/L, clinoptilolite mass of 6 ​g, particle size of 1.25 ​mm and a temperature of 40 ​°C. The maximum removal percentage was 92.80 and 33.67 for Ni (II) and Co (II), respectively.