Modelling and Optimization of Oxalic Acid Removal by Layered Double Hydroxide

The main purpose of this study is the investigation of the optimization of the conditions of oxalic acid (OxA) adsorption using layered double hydroxide (LDH), modeling the adsorption with both the response surface methodology (RSM) and an artificial neural network (ANN). Mg-Al LDH was synthesized via the co-precipitation method and characterized by Fourier transform infrared spectroscopy (FTIR), inductively coupled plasma mass spectrometry (ICP-MS) and X-ray diffraction (XRD) techniques. The equilibrium time and kinetic model data required to realize the adsorption process design were examined. The process time, initial acid concentration, temperature, and adsorbent dosage as the independent variables were chosen while measuring the percentage of OxA removal. Modeling these results with both RSM and ANN techniques resulted in an ANN model showing a slightly better coefficient of determination than the RSM model. The models yielded consistent results for the optimal conditions of the process.