Researchers developed a deep neural network framework that predicts hafnium oxide thin-film thickness, refractive index, and wet etch rate from atomic-layer-deposition process conditions and wafer location. The model generated process-property maps that could accelerate ALD optimization while supporting future digital twin-guided semiconductor manufacturing.
