Ga-doped zinc oxide (ZnO:Ga) thin films were prepared on glass substrate by magnetron sputtering at room temperature (RT) and thermally annealed in hydrogen atmosphere for 1 h. The effects of film thickness and annealing temperature on sheet resistance, transmittance and figure of merit of ZnO:Ga thin films were analyzed and modeled using the artificial neural networks (NNets). The NNet models presented the good prediction on sheet resistance, transmittance and figure of merit of ZnO:Ga thin films and it was found that the electrical and optical properties of ZnO:Ga thin films were enhanced by thermal annealing. After NNet models were verified, genetic algorithm (GA) was used to search the optimized recipe for the desired figure of merit of ZnO:Ga thin films. The methodology allows us to estimate the optimal process condition with a small number of experiments.
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- Artificial Intelligence