(435a) Detailed Kinetic Modeling of Thermal Decomposition of Silane Via Automatic Mechanism Generation

The gas phase thermal decomposition of precursors such as silane is an important step in chemical vapor deposition for semiconductor material fabrication. Understanding the reaction pathways involved in this process can help design processes to better control the quality of the materials produced by CVD. We have extended the open-source software Reaction Mechanism Generator (RMG) [1] so that it can automatically generate detailed kinetic models for silicon hydrides and, eventually, a wide range of silicon-containing species. This work draws from and builds upon previous efforts in automatic mechanism generation for silicon hydrides, specifically in the field of silicon nanoparticle formation [2].

RMG's database was updated to include data for thermodynamics and kinetics important in silane thermal decomposition. Benson's group additivity values for thermochemistry of Si-containing molecules were added [3], and hydrogen bond increment (HBI) values for radical species were derived using quantum chemistry. Four reaction families, which RMG uses to propose reactions and estimate their kinetics, were enabled for silicon hydride chemistry:

Because existing experimental and theoretical kinetic data for some of these families is sparse, additional transition state theory (TST) rates were calculated, using an automated kinetics calculator. The automated TST calculator uses a group additive scheme and distance geometry to predict transition state geometries [4], DFT to optimize and characterize them, and the software package Cantherm to apply conventional transition state theory.

Using the updated database, a detailed model was built for thermal decomposition of SiH₄. To validate the RMG-generated model, reactor simulations were compared to flow tube experimental data from Onischuk et al. [5]. Simulation conditions were matched to the experimental conditions, with base conditions of y₀(SiH₄) = 1.6e-4 in argon, T = 913 K, and P = 39 kPa. The SiH₄ concentration profile variation with residence time matches reasonably well with the experimental data within a 20 K uncertainty in temperature. SiH₄ concentration dependence on temperature also matches well with experiment.

The experimental SiH₄ and Si₂H₆ concentration profiles, normalized to initial SiH₄ concentration, appear to vary with initial SiH₄ concentration suggesting a second order dependence on SiH₄ concentration. However, the simulation results approximate a first order dependence on SiH₄ concentration. Sensitivity analysis and flux analysis were employed to understand and resolve this discrepancy between the initial simulations and experiment.

The newly updated RMG and validated model allows simulation of silicon hydride thermal decomposition at different process conditions, reducing the need for future experiments. Importantly, this work provides a framework for generating detailed kinetic models of other, less well-studied CVD precursors.

References

[1] C.W. Gao, J.W. Allen, W.H. Green, R.H. West. Reaction Mechanism Generator: Automatic construction of chemical kinetic mechanisms. Computer Physics Communications, 203, 212-225, (2016) http://dx.doi.org/10.1016/j.cpc.2016.02.013.

[2] H.W. Wong, X. Li, M.T. Swihart, L.J. Broadbelt. Detailed kinetic modeling of silicon nanoparticle formation chemistry via automated mechanism generation. Journal of Physical Chemistry A, 108(46), 10122â??10132, (2004) http://dx.doi.org/10.1021/jp049591w

[3] M.T. Swihart, M. T and S.L. Girshick. Thermochemistry and kinetics of silicon hydride cluster formation during thermal decomposition of silane. Journal of Physical Chemistry B, 103(1), 64â??76, (1999) http://dx.doi.org/10.1021/jp983358e

[4] P.L. Bhoorasingh and R.H. West. Transition state geometry prediction using molecular group contributions. Phys. Chem. Chem. Phys., 17, 32173-32182, (2015) http://dx.doi.org/10.1039/C5CP04706D.

[5] A.A. Onischuk, V.P. Strunin, M.A. Ushakova, V.N. Panfilov. Studying of silane thermal decomposition mechanism. Int. J. Chem. Kinetics, 30(2), 99-110, (1998) http://dx.doi.org/10.1002/(SICI)1097-4601(1998)30:2<99::AID-KIN1>3.0.CO;2-O