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

[26a-E104-3] Formation energy prediction of large-scale finite-sized extended interstitial defects in Si crystal using an ANN interatomic potential

〇Yusuke Noda¹, Masato Ohbitsu², Tatsuya Yokoi³, Eiji Kamiyama¹, Takuto Ushiro², 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-principles calculation, interatomic potential

In this study, we systematically evaluated energetic stability of finite-sized extended interstitial defects (EIDs) in silicon crystals using an artificial neural-network (ANN) interatomic potential. We compared formation energies of compact clusters, (311)rod-like defects, and Frank-loop defects, respectively, and then we theoretically revealed that the Frank-loop defects become most stable for the large number of interstitial Si atoms.