Modeling pseudo-second-order kinetics of orange peel-paracetamol adsorption process using artificial neural network

In this work, an artificial neural network (ANN) was developed to model the Pseudo- Second Order (PSO) kinetics of orange peel-paracetamol adsorption process. The orange peel used for the adsorption process was prepared, activated, and characterized using Scanning Electron Microscopy (SEM) and Fourier Transform Infrared Spectroscopy (FTIR) techniques respectively. Batch adsorption experiment was carried out to obtain the concentrations of paracetamol (PCM) adsorbed on orange peel activated carbon (OPAC) at different operating conditions which include contact time (0–330 ​minutes), initial PCM concentration (10–50 ​mg/L), and temperature (30–50 ​°C). Then, the experimental data was used to compute PSO kinetics of the orange peel – paracetamol adsorption process. To predict the PSO kinetics, different ANN structures were investigated. The optimal ANN structure which uses 18 hidden neurons, hyperbolic tangent sigmoid transfer function (tansig) at the input layer, linear transfer function (purelin) at the output layer, and Levenberg Marquardt as its backpropagation algorithm demonstrated the optimal prediction ability. Specifically, the optimal ANN model gave Mean Average Error (MAE), Mean Square Error (MSE), Root Mean Square Error (RMSE), and R2 values of 0.0515, 0.0064, 0.0798 and 1.0000 respectively when compared with the experimental data. The results obtained showed that ANN can be used to effectively model PSO kinetics of orange peel-paracetamol adsorption process.