研究目的
To increase the uniformity of thin film semiconductor devices and quality of solar cell products while maintaining high production quality through the application of a run-to-run control algorithm to a multiscale, three-dimensional computational fluid dynamics model of plasma-enhanced chemical vapor deposition.
研究成果
The run-to-run control algorithm, using substrate temperature correction via an exponentially weighted moving average formulation, successfully drives the amorphous silicon thin film product to within 1% of the desired thickness set-point in both radial and azimuthal directions across the wafer surface within five to six batches of operation. This approach is effective for both optimized and poorly optimized PECVD geometries, offering a novel means of deposition control without the need for re-tooling.
研究不足
The computational demands of the 3D multiscale CFD model are significant, requiring days to weeks of continuous processing per batch. The model's accuracy and speed are also affected by the need to interpolate between known data points when updating boundary conditions and the inability to further decompose the microscopic kinetic Monte Carlo simulations.
1:Experimental Design and Method Selection:
The study employs a multiscale, three-dimensional computational fluid dynamics (CFD) model for plasma-enhanced chemical vapor deposition (PECVD) of amorphous silicon thin films. The model links macroscopic reactor scale behaviors to microscopic thin film growth using a dynamic boundary updated at each time step.
2:Sample Selection and Data Sources:
The model uses a cylindrical PECVD reactor geometry with showerhead holes arranged in a rectangular array, simulating the deposition of amorphous silicon thin films.
3:List of Experimental Equipment and Materials:
The simulation utilizes ANSYS software for creating the geometric mesh and solving the partial differential equations describing transport phenomena and chemical reactions in the gas phase. Three user-defined functions (UDFs) are developed to tailor the ANSYS solver to the PECVD of amorphous silicon.
4:Experimental Procedures and Operational Workflow:
The macroscopic PDEs are solved using finite volume methods, with the microscopic domain initialized based on the macroscopic results. The simulation advances until the end time of a batch deposition process.
5:Data Analysis Methods:
The effectiveness of the run-to-run control algorithm is evaluated by measuring the thin film thickness uniformity across the wafer surface after successive batch deposition cycles.
独家科研数据包,助您复现前沿成果,加速创新突破
获取完整内容