Cooling and Particulate Challenges for Next Generation Atomic Layer Etch Technologies
High Aspect Ratio (HAR) etch is fast emerging as the dominant method for fabricating next generation silicon micro-structures such as 3D-NAND, MRAM and DRAM. As 3D-NAND and other technologies scale further, achieving anisotropic HAR etch at commercially feasible yields and process times is a paramount challenge. The advancing process requirements of HAR etch are driving the need for colder wafer temperatures as well as finer particle control.
While introduced in the 1980s , HAR cryogenic etching has gained recent momentum because it allows simultaneous etching and sidewall passivation, resulting in significantly fewer process steps than other methods. Currently understood process requirements, coupled with practical heat transfer limitations and the exothermic nature of certain reactions in the cryogenic etching process, point to a wafer temperature < -85°C for HAR cryogenic etching. Accounting for temperature gradients, a desirable process window between -110°C and -120°C is indicated . Use of liquid nitrogen is impractical in a clean room environment, motivating a search for other cooling solutions.
In addition to the cooling challenge, particle contamination is one of the biggest factors affecting wafer yield in semiconductor manufacturing processes such as etch and deposition. As these processes move into the atomic layer domain, particle control becomes increasingly important in the management of defects and yield.
It has been shown that particle formation during multi-layer plasma etching is activated by the presence of H2O; moreover, the number of particles generated is related to the partial pressure of H2O in the chamber . This presentation proposes the combination of cryo and turbomolecular pumping (TMP) as a means of particle mitigation, by minimizing the H2O partial pressure.
In this presentation, we will also show how Mixed Refrigerant Joule-Thompson (MR-JT) cryochillers can be an ideal solution for HAR cryogenic etching, with several advantages in energy consumption, physical footprint, and reliability. Second, particle transport through the TMP has been modeled to show that the number of chamber particles is directly affected by the pumping mechanism. The results of this model and experimental trials are reported, and it is shown that chamber particles can be more effectively removed by novel TMP technology.
 Tachi S, Tsujimoto K and Okudaira S “Low-temperature reactive ion etching and microwave plasma etching of silicon,” Applied Physics Letters, vol. 52, issue 8, 616, 1988.
 N. Ito, T. Moriya, F. Uesugi, M. Matsumoto, S. Liu, and Y. Kitayama, “Reduction of particle contamination in plasma-etching equipment by dehydration of chamber wall,” Japanese Journal of Applied Physics, vol. 47, no. 5, pp. 3630-3634, 2008.