Title: Multiscale modeling of PECVD generated Silicon films with Kinetic Monte Carlo and LAMMPS molecular dynamics
Presenter: Andreas Bick
Affiliation: Scienomics SARL, 17 Square Edouard VII, 75009 Paris, France
Co-author: L.D. Peristeras
Affiliation: Scienomics SARL, 17 Square Edouard VII, 75009 Paris, France
Co-authors: D.G. Tsalikis, V.G. Mavrantzas, E. Amanatides and D. Mataras
Affiliation: Department of Chemical Engineering, University of Patras, Patras, GR26500, Greece
Plasma enhanced chemical vapor deposition (PECVD) is a widely used technique for growing films for application in a broad range of technologies such as electronics 1, optoelectronics 2, photovoltaics 3,4 and is of significant industrial interest since it allows fast deposition on thin films even at low temperatures 5,6. Although PECVD has been investigated extensively, there are still important aspects that have not been completely elucidated 6 such us growth phenomena upon different radical impinging the surface and plasma surface interactions, i.e. the phenomena occurring on the surface upon impingement of chemically reactive radicals and energetic species from the plasma 7.
Our main objective is to elucidate the microscopic mechanisms as well as the interplay between atomic level and macroscopic design parameters associated with the development of nano- or micro-scale crystalline regions in the grown film. The ultimate goal is to use multi-scale modeling as a design tool for tackling the issue of local crystallization and its dependence on operating variables. At the heart of our simulation approach is a very efficient, large-scale kinetic Monte Carlo (kMC) algorithm which allows generating samples of representative Si films based on a validated chemistry model. In a second step, the generated film will be subjected to an atomistic Molecular Dynamics (MD) simulation study which restores the molecular details lost or ignored in the kMC model. The atomistic simulations are computationally very demanding; they are, however, an important ingredient of our work: we will use it to back-map the coarse grained model employed in the kMC simulations to an all-atom model which is further relaxed through detailed NPT MD. This tunes local structure thus also important morphological details associated with the presence of crystalline and amorphous regions (and the intervening interfacial domains) in the grown film.
The developed kinetic Monte Carlo (KMC) model is able to simulate and study the kinetics of film growth at the conditions of interest (pressure=4 mbar, power = 310 mW/cm2, total flow rate =1 slm (standard litres per minute) and silane mole fraction in gas phase =1 - 6 %) in a medium scale plasma reactor. The model is based on a carefully chosen set of reacting or active radicals (species) in the gas phase impinging the film and a detailed set of surface reactions, without excluding any important information (e.g. surface diffusion, defects) and can access film thicknesses in the order of several tens of nanometers using moderate computational resources generating representative initial structures depending on the operational conditions of interest.
The initial structures generated with the kMC model will be used as input for the atomistic simulations. Atomic-scale simulation work of plasma-surface interactions in the PECVD of Si thin films has been based on an empirical description of interatomic interactions in the Si:H systemaccording to Tersoff’s 8-10 potential for Si, as extended by Ohira 11-13 to incorporate Si–H, H–H, and the corresponding three-body interactions. The extension of the potential to include the presence of hydrogen adopted the Tersoff parametrization to fit results of ab initio calculations for the structure and energetics of SiHx, x = 4, species in the gas phase (Ohira 11-13). The quantitative accuracy of the interatomic potential of Ohira 11-13 for the study of radical–surface interactions was tested exhaustively through extensive comparisons 14-18 of its predictions with experimental measurements and results of ab initio calculations. These comparisons also indicate that the interatomic potential of Ohira 11-13 is capable of quantitative descriptions of radical surface diffusion processes on hydrogen-covered crystalline and amorphous Si surfaces. In our work the atomistic simulations are computationally very demanding; the structures generated via kMC are of the order of tens of nanometers, consisting of several hundreds of thousands silicon and hydrogen atoms. LAMMPS was selected as the MD simulator package that is able to run massively parallel simulations on modern HPC systems. The tersoff style potential was extended to model the specific functional form used by Ohira 19.
The difference with functional form supported in LAMMPS is the existence of aij parameter in the repulsive part which in most of the cases can be considered equals to one.
This work is as part of ongoing research work funded by the European Commission under the 7th Framework Programme (FP7-ENERGY-2009-TREN-2) titled: Demonstration of high performance processes and equipments for thin film silicon photovoltaic modules produced with lower environmental impact and reduced cost and material use. The presented simulation work was supported by the LinkSCEEM-2 project, funded by the European Commission under the 7th Framework Programme through Capacities Research Infrastructure, INFRA-2010-1.2.3 Virtual Research Communities, Combination of Collaborative Project and Coordination and Support Actions (CP-CSA) under grant agreement no RI-261600.
1 R. Reif, Plasma enhanced chemical vapor deposition of thin films for microelectronics processing in Handbook of Plasma Processing Technology: Fundamentals, Etching, Deposition, and Surface Interactions, (S. M. Rossnagel, J. J. Cuomo, and W. D. Westwood, Noyes Eds., Park Ridge, NJ, 1990).
2 L. Crowley, Sol. St. Technol. 35, 94 (1992).
3 A. Shah, P. Torres, R. Tscharner, N. Wyrsch and H. Keppner. Science 285, 692 (1999).
4 W. C. Sinke, MRS Bull. 18, 18 (1993).
5 A. Matsuda. Jpn. J. Appl. Phys. 43, 7909 (2004).
6 D. Maroudas, Adv. Chem. Eng. 28, 251 (2001).
7 D. B. Graves, M. J. Kushner, J. W. Gallagher, A. Garscadden, G. S. Oehrlein and V. A. Phelps, Panel on Database Needs in Plasma Processing ( National Academy Press, Washington, DC, 1996).
8 J. Tersoff, Phys. Rev. Lett. 56, 632 (1986).
9 J. Tersoff, Phys. Rev. B 37, 6991 (1988).
10 J. Tersoff, Phys. Rev. B 39, 5566 (1989).
11 T. Ohira, T. Inamura, and T. Adachi, Mater. Res. Soc. Symp. Proc. 336, 177 (1994).
12 T. Ohira, O. Ukai, T. Adachi, Y. Takeuchi and M. Murata, Phys. Rev. B 52, 8283 (1995).
13 T. Ohira, O.Ukai, M. Noda, Y. Takeuchi and M. Murata, Mater. Res. Soc. Symp. Proc. 408, 445 (1996).
14 S. Ramalingam, D. Maroudas and E. S. Aydil, Appl. Phys. Lett. 72, 578 (1998).
15 S. Ramalingam, D. Maroudas and E. S. Aydil, J. Appl. Phys. 84, 3895 (1998).
16 S. Ramalingam, D. Maroudas, E. S. Aydil, and S. P. Walch, Surf. Sci. 418, L8 (1998).
17 S. Ramalingam, D. Maroudas and E. S. Aydil, J. Appl. Phys. 86, 2872 (1999b).
18 S. Ramalingam, P. Mahalingam, E. S. Aydil and D. Maroudas, J. Appl. Phys. 86, 5497 (1999).
19 T. Ohira, O.Ukai, M. Noda, Surface Science 48, 216, (2000).