The control of thermal oxide formation in Si is attracting great interest and importance due to the recent requirement of miniaturization in Si devices. This also demands detailed understanding of the atomic-scale processes during Si thermal oxidation. It is widely consented that Si thermal oxidation consists of a diffusion process of oxidant in the oxide already formed and its reaction process at the SiO2/Si interface. Regarding the reaction process, we have revealed that the barrier-height of the O2 insertion into Si-Si bonds of silicon substrate at the flat and abrupt interface (0.2 eV) [1] is at variance with the experimentally reported activation energy (2.0 eV). [2] To resolve this discrepancy, other microscopic mechanisms such as the effects of accumulation of interfacial strain by the oxidation and its release mechanisms, [3] and polymorphism of amorphous SiO2 [4] are predicted. Besides these predictions, the verification based on atom-scale calculations has remained far less examined. Here, we perform first-principles total-energy calculations that clarify atomic processes of O2 incorporation into the silicon substrate at SiO2/Si interfaces with oxidation-induced strain. The calculated barrier-heights of an O2 (~0.9 eV) demonstrate that the oxidation-induced strain effect is thought to be the key to clarify the origin. While the calculated barrier-heights for the strained interfaces are larger than that for the strain-released one (0.2 eV), the values are still lower than that estimated by the experiments (2.0 eV). These results reveal that the oxidation-induced strain suppresses the reaction of oxygen but the oxygen molecule incorporation processes themselves cannot explain the activation energy, implying that the release of the strain around the interface as well as the location of O2 molecules in the oxide is substantial for atom-scale understanding of the reaction mechanisms in the thermal oxidation of silicon.
References: [1]T. Akiyama and H. Kageshima, Surf. Sci. 576, L65 (2005).[2]B.E. Deal and A.S. Grove, J. Appl. Phys. 36, 3770 (1965).[3] H. Kageshima and K. Shiraishi, Phys. Rev. Lett. 81, 5936 (1998).[4]H. Kageshima et al., Jpn. J. Appl. Phys. 45, 7672 (2006).
|