• PDF / 73,219 Bytes
• 6 Pages / 612 x 792 pts (letter) Page_size
• 60 Downloads / 149 Views

DOWNLOAD

REPORT

---

E9.17.1

Barrier to migration of the intrinsic defects in silicon in different charged system using First-principles calculations Jinyu Zhang, Fujitsu R&D Center Co. Ltd, Room B1003, Eagle Run Plaza No.26 Xiaoyun Road, Chaoyang District Beijing, China Yoshio Ashizawa and Hideki Oka Fujitsu Laboratories Ltd. 50 Fuchigami, Akiruno, Tokyo, 197-0833, Japan ABSTRACT Using density functional theory (DFT) calculations within the generalized gradient approximation (GGA), we have investigated the structure, energies and diffusion behavior of Si defects including interstitial, vacancy, FFCD and divacancy in various charged states. INTRODUCTION Intrinsic point defects in silicon have attracted a great deal of interest because of the technological importance of the material. In particular they strongly influence the diffusion of dopant atoms at elevated temperatures during device manufacture. Both experimental and theoretical techniques have been brought to bear on these defects but fundamental gaps in our understandings still exist. Unfortunately it has not been possible to detect silicon self-interstitials directly and we must rely on theoretical techniques to elucidate their microscopic nature. METHOD Our calculations have been performed using the CASTEP[1] density functional electronic structure package. We have used the PBE[2] generalized gradient approximation (GGA) for the exchange-correlation functional with a plane-wave basis set. We have used a Vanderbilt ultra-soft pseudopotential[3] for silicon with a supercell of 64 atoms and a plane-wave-basis set with a kinetic-energy cutoff of 330 eV for all calculations. A 23 Monkhorst[4] set was applied to k-point sampling. To minimize the electronic energy, a density mixing scheme is used,[5] whereas for minimization of ionic energy, Hellmann-Feynman theorem is utilized to calculate forces and a conjugate gradient scheme used.[6] The relaxation of all configurations used in these calculations proceeds until the Hellmann-Feynman force does not exceed 0.03 eV/Å. To investigate the saddle point of the diffusion path, Linear Synchronous Transit (LST) and Quadratic Synchronous Transit (QST) [7] are used. We checked the convergence with respect to the size of the basis with different basis-set energy cut-offs from 160eV to a maximum 400eV. The results indicate that the energy difference between the defect structures are converged to 0.1eVwhen cutoff energy is larger than 280eV. A supercell of 216 atoms and a kinetic-energy cutoff of 250 eV are used to check convergence of some configurations. Because of this large cell size, we could use only the gamma point for the electronic structure part. To calculate the formation energy of charged defects, we followed the process suggested by Jeong et al.[8]

E9.17.2

RESULTS AND DISCUSSION 1. Self-interstitial Many theoretical studies of self-interstitials in silicon have been performed. There are some differences between the results of the various LDA calculations, but the consensus view is that the split-, hexagonal (H), and tetrahedral (