(711c) Run-to-Run Control of PECVD Systems: Application to a Multiscale CFD Model of Amorphous Silicon Deposition

Uniform deposition of thin film layers remains a challenge in silicon processing industries due to the lack if in situ measurements and drift in the plasma composition caused by fouling [1],[2]. Specifically, in plasma-enhanced chemical vapor deposition (PECVD) of amorphous silicon (a-Si:H), growth rate non-uniformities greater than 20% across the surface of the wafer are common [3], which may lead to inefficient solar cells and microelectronics of poor quality. Process operators typically avoid this issue by preconditioning the reaction chamber and manually adjusting deposition conditions such that thickness non-uniformities may be reduced. Recently, Crose et al. developed a run-to-run based, offline control scheme which has been demonstrated to reduce the offset in product thickness from 5% to less than 1% within ten batches of operation [3]. While these results are promising, the multiscale model used relies on a first principles approach to solving the continuous mass, momentum and energy balances; however, recent advances in parallel computation strategies have made possible the application of multiscale computational fluid dynamics (CFD) to the modeling of PECVD reactors [4].

Consequently, we propose the application of the newly developed run-to-run control strategy to the simulated deposition of amorphous silicon (a-Si:H) thin films using a transient, multiscale CFD model. The two-dimensional axisymmetric model recently developed by Crose et al. [4] has been shown to accurately capture the complex behavior of the PECVD reactor and to reproduce experimentally observed non-uniformities with respect to the film thickness and porosity, and will be used throughout this work. A computationally efficient parallel processing scheme allows for the application of the run-to-run algorithm to 40 serial batch simulations which are shown to reduce the product offset from the 300 nm thickness set-point. Additional considerations are made with respect to the porosity and hydrogen content of the amorphous silicon product.

[1] Armaou A, Christofides PD. Plasma enhanced chemical vapor deposition: modeling and control. Chemical Engineering Science. 1999;54:3305-3314.

[2] Gabriel O, Kirner S, Klick M, Stannowski B, Schlatmann R. Plasma monitoring and PECVD process control in thin film silicon-based solar cell manufacturing. EPJ Photovoltaics. 2014;55202:1-9.

[3] Crose M, Kwon JSI, Tran A, Christofides PD. Multiscale modeling and run-to-run control of PECVD of thin film solar cells. Renewable Energy. 2017;100:129-140.

[4] Crose M, Tran A, Christofides PD. Multiscale computational fluid dynamics: methodology and application to PECVD of thin film solar cells. Coatings. 2017; 7 (2) , 22.