Abstract
The improvement in ALD growth rate has always been challenging due to its slow atomic-scale depositions. Although Al(2)O(3) ALD is one of the most widely used ALD processes, the effects of its process parameters on growth rate have not been systematically analyzed using statistical approaches. These statistical methods offer better efficiency and effectiveness compared to traditional techniques for studying complex processes like ALD. This paper presents a systematic investigation and optimization of four process parameters on growth rate of Al(2)O(3) ALD thin films using a full factorial design of experiments (DOE) approach. Statistical analysis revealed that deposition temperature is the only statistically significant factor in Al(2)O(3) ALD process, while argon gas flow rate, pulsing time and purging time are tested nonsignificant. Significant interactions were found between deposition temperature and purging time, and between pulsing time and purging time, with all other interactions being nonsignificant. Optimal process settings for higher deposition rate were identified: the temperature and gas flow rate are set at lower levels, while pulsing time and purging time are set at higher levels.