Multi-response optimization of ZnO thin films using Grey-Taguchi technique and development of a model using ANN

Spray pyrolysis deposited zinc oxide films have been prepared using zinc acetate precursor in iso-propyl alcohol (IPA). Based on deposition parameters, outputs were optimized for multiple performance characteristics such as thickness, roughness, and an optical transmittance of ZnO films using Grey-Taguchi technique. Optimum parameters from Grey relational analysis (GRA) were found to be 0.2 M, 1 bar, 365 °C, and 5 mL min−1. The average thickness, roughness, and an optical transmittance obtained at the optimum condition using GRA were 79 nm, 1.88 nm, and 85.96% respectively as compared to the results obtained at optimized condition by a Taguchi method were 95 nm, 2.45 nm, and 85.19%, respectively. This indicated that the design proposed by Grey relational technique is a feasible approach for determining the optimal parameters. Further, the best condition was predicted by artificial neural network (ANN) and it was observed using cascade forward back propagation algorithm, tangent sigmoid transfer function as a hidden layer, and 4-8-3 topology.