Stochastic Optimization of TiO2-Graphene Nanocomposite by Using Neuro-Regression Approach for Maximum Photocatalytic Degradation Rate

TiO2 is one of the most common materials for photocatalytic applications due to its stability, affordability, and photoactive efficiency. However, it has some drawbacks, such as limited solar radiation response and quick recombination of excitons. Using graphene could be one of the methods to enhance the photocatalytic properties of TiO2. This study intends to optimize the photocatalytic performance of TiO2/Graphene (TiO2/G) nanocomposite by using neuro-regression analysis. In the analysis, the effect of some hydrothermal synthesis parameters, namely, amount of graphene oxide, ethanol/water ratio, and hydrothermal reaction time on the photocatalytic activity of TiO2/G nanocomposite, have been investigated. The parameters were determined from a literature study focused on overcoming the drawbacks of TiO2 by combining it with graphene oxide. Nelder-Mead, Simulated Annealing, Differential Evolution, and Random Search algorithms are used to obtain the optimum synthesis parameters for maximum photocatalytic activity in the optimization process. The results are indicated that all algorithms give the realizable value for design variables and photodegradation rate.