Machine learning simulation of Cr (VI) separation from aqueous solutions via a hierarchical nanostructure material.

• Separation of Cr(VI) via a hierarchical ordered composite material. • Artificial neural network simulation of Cr(VI) removal from water using adsorption. • Validation and training the ANN using measured data of Cr removal. We developed a novel methodology for simulation and computation of species adsorption on the surface of ordered nanoporous materials. The main aim was to develop an advanced and high-performance predictive model for prediction of adsorption behaviour of novel nanostructured materials which can be used for a variety of species. The simulations were conducted using artificial intelligence technique by considering the pH and adsorbent dosage as the inlet variables to the model for prediction of the output. The computations were carried out in order to predict the equilibrium amount of species in the solution. The study was performed on the adsorption of Cr (VI) ions using nanocomposite of layered double hydroxide-metal organic framework. Due to the ordered and high surface area of this composite material, it was considered for this computational study and removal of Cr (VI) ions. The collected data was used to train the artificial intelligence network, and the statistical analysis was performed for training and validation of the network in order to assess the accuracy of the trained model. The R2 value more than 0.999 was obtained for the fitting the Cr (VI) adsorption data by the model. A great agreement was obtained between the model's findings and the measured data for removal of Cr (VI) from water. Furthermore, the effect of adsorption parameters on the equilibrium concentration of Cr was studied using the validated model. It was revealed that the adsorption removal was increased by increasing the adsorbent dosage, while it decayed by enhancing the pH of solution. [ABSTRACT FROM AUTHOR]