[26a-E104-1~11] 15.7 Crystal characterization, impurities and crystal defects

[26a-E104-2] Prediction of preferential growth modes of large-scale vacancy clusters in Si crystal using an ANN interatomic potential

〇Yusuke Noda¹, Takuto Ushiro², Tatsuya Yokoi³, Eiji Kamiyama¹, Masato Ohbitsu², Hiroki Nagakura², Koji Sueoka¹ (1.Faculty of Computer Science and Systems Engineering, Okayama Pref. Univ., 2.Graduate School of Computer Science and Systems Engineering, Okayama Pref. Univ., 3.Graduate School of Engineering, Nagoya Univ.)

Keywords:silicon, first-principle calculation, interatomic potential

In this study, we evaluated energetic stability of large-scale vacancy (V) clusters in silicon crystals using an artificial neural-network (ANN) interatomic potential. Formation energies of V clusters including up to 1,000 vacancies were examined, and the ANN interatomic potential revealed preferential growth modes of V clusters. And, we found a relatively clear correlation between formation energies and the number of dangling bonds for various V clusters.