Kwan-Sun Yoon

Kinetic monte carlo modeling of boron diffusion in strained silicon

We discuss the boron diffusion in a biaxial tensile-strained {001} Si and SiGe layer using the kinetic Monte Carlo (KMC) method. We created strain in silicon by adding a germanium mole fraction to the silicon in order to perform a theoretical analysis. The generation of strain in silicon influences the diffusivity as well as the penetration profile during the implantation. The strain energy of the charged defects was calculated from ab initio calculation whereas the diffusivity of boron was extracted from the Arrhenius formula. Hereby, the influence of the germanium content on the dopant diffusivity was estimated. Our KMC study revealed that the diffusion of the B atoms was retarded with increasing germanium mole fraction in the strained silicon layer. Furthermore, we derived the functional dependence of the in-plane strain as well as the out-of-plane strain on the germanium mole fraction, which is based on the distribution of equivalent stress along the Si/SiGe interface.

Atomistic modelling for boron diffusion in strained silicon substrate

We discuss the issue of boron diffusion in biaxial tensile strained {001} Si and SiGe layer with kinetic Monte Carlo (KMC) method. We created strain in silicon by artificially adding a germanium mole fraction to the silicon in order to perform a theoretical analysis. The strain energy of the charged defects was calculated from ab initio calculation whereas the diffusivity of boron was extracted from the Arrhenius formula. Hereby, the influence of the germanium content on the diffusivity of the impurity atom was estimated. Our KMC study revealed that the diffusion of the boron atoms was retarded with increasing germanium mole fraction in the strained silicon layer. Furthermore, we derived the functional dependence of the in-plane strain as well as the out-of-plane strain as a function of the germanium mole fraction, which is based on the distribution of equivalent stress along the Si/SiGe interface.

Ab-initio study on indium diffusion in silicon substrate under hydrostatic stress

In this paper, we report our ab-initio study on indium diffusion in strained Si. We investigated the minimum energy path as well as the migration energy for indium atoms in the hydrostatic strained silicon wherein the strain was incorporated by constructing a silicon layer on SiGe substrate with 20% germanium mole fraction. The ab-initio calculation allowed us to figure out the lowest-energy structure (InS+Sii Td) and the next lowest energy structure (Ini Td) during indium diffusion. The energy of migration barrier was also predicted from the nudged elastic band and climbing nudged elastic band method. Consequently, we found that the energy barrier was a little higher than the case of unstrained Si.

Determination of Evolution Path for B/sub m/I/sub n/ Clusters in Atomistic Model

In this paper, we report our study on the determination on the evolution path and dominant BmIn clusters during boron diffusion using a simple atomistic model for describing the kinetic Monte Carlo (KMC). It has been known that clusters generated after ion implantation play a decisive role in the enhanced boron diffusion at the tail region in contrast to the immobile property at the peak region. Our model, which is based on the simple continuum model [2], takes the intermediate clusters into account as well as dominant clusters for understanding the evolutionary behaviour of boron interstitial clusters during boron diffusion. We found out that the intermediate clusters such as B3I3 and B2I3 play a significant role despite that the lifetime of the corresponding intermediate clusters are relatively short due to low binding energies. Additionally, our investigation revealed that the density of BI2 clusters increased at the beginning of the annealing process while the density of B3I increased at a later stage. Also, we found the evolution path of dominant clusters from B2I to B3I during annealing. Finally, KMC simulation results were compared with experimental data, which demonstrated an excellent agreement