The modeling of CO2 absorption in ionic liquids using Artificial Neural Network

In this work, the potential application of Artificial Neural Network (ANN) was studied to predict the absorption of Carbon Dioxide (CO2) in Ionic Liquid (IL) solutions over wide-ranging operating conditions. A few physical properties had been chosen as input data which were temperature, partial pressure of CO2, molecular weight, acentric value, critical temperature and critical pressure of IL. A sample of 184 experimental data points of the solubility of CO2 is collected from literatures (for training, validation and testing stages) to acquire the network. In order to obtain the best developed ANN model, the trained network comprising of transfer function, training function, number of neurons and hidden layers had been manipulated. The best network (MSE=2.9336×10−5, MRE=0.007297, R=0.99977 and R2=0.9994) was trained by Levenberg-Marquardt backpropagation algorithm with Tan-sigmoid transfer function having two hidden layers, 6 and 12 neurons for first and second layer respectively. Besides, every single predicted data was compared to its respective experimental data and recorded the highest percentage deviation only less than 6%. Moreover, the extension capability of the model was investigated by additional data 104 data sets from three different types of IL ([bmim][PF6], [emim][Tf2N] and [C4mim] [DCA]). The results indicate that the acquired ANN model has power to forecast precisely the CO2 absorption in different types of IL.