Electron collision cross section data in plasma etching modeling

Abstract

Semiconductor chips are the cornerstone of the information age, which play a vital role in the rapid development of emerging technologies such as big data, machine learning, and artificial intelligence. Driven by the growing demand for computing power, the chip manufacturing industry has been committed to pursuing higher level of integration and smaller device volumes. As a critical step in the chip manufacturing processes, the etching process therefore faces great challenges. Dry etching (or plasma etching) process based on the low-temperature plasma science and technology is the preferred solution for etching the high-precision circuit pattern. In the low-temperature plasma, electrons obtain energy from the external electromagnetic field and transfer the energy to other particles through collision process. After a series of complex physical and chemical reactions, a large number of active particles such as electrons, ions, atoms and molecules in excited states, and radicals are finally generated, providing the material conditions for etching the substrate. Dry etching chamber is a nonlinear system with multiple space-time dimensions, multiple reaction levels and high complexity. Facing such a complex system, only by fully understanding the basic physical and chemical reaction of the etching process can we optimize the process parameters and improve the etching conditions, so as to achieve precision machining of the semiconductor and meet the growing demand of the chip industry for etching rate and yield. In the early days, the process conditions of dry etching were determined through the trial-and-error method, which is characterized by high cost and low yield. However, with the help of plasma simulation, nowadays people have been able to narrow the scope of experiment to a large extent, and find out efficiently the optimal process conditions in a large number of parameters. In this review, we first introduce the basic theory of the mostly used models for plasma simulation including kinetic, fluid dynamic, hybrid and global models, in which the electron collision cross sections are the key input parameters. Since the formation of the low-temperature plasma is driven by the electron-heavy particle collision processes, and the active species for plasma etching are generated in the reactions induced by electron impact, the accuracy and completeness of the cross-section data greatly affect the reliability of the simulation results. Then, the theoretical and experimental methods of obtaining the cross-section data of etching gases are summarized. Finally, the research status of the electron collision cross sections of etching atoms and molecules is summarized, and the future research prospect is discussed.