AVS 66th International Symposium & Exhibition
    Plasma Science and Technology Division Tuesday Sessions
       Session PS+EM-TuM

Paper PS+EM-TuM6
CCP Dry Clean Process Development Using Quadrupole Mass Spectrometer and Optical Emission Spectroscopy

Tuesday, October 22, 2019, 9:40 am, Room B131

Session: Advanced FEOL
Presenter: Harutyun Melikyan, Micron Technology
Authors: H. Melikyan, Micron Technology
A.D. Martinez, Micron Technology
S.C. Pandey, Micron Technology
M. Koltonski, Micron Technology
G. Sandhu, Micron Technology
Correspondent: Click to Email

Dual frequency capacitive coupled plasmas provide flexible control of ion energy distributions, enabling high selectivity for etching of different materials, and flexibility to develop more efficient dry clean processes for higher productivity. It is common for each patterned wafer etch to have pre and post dry clean steps, with and without wafer, respectively. The primary focus of dry etch engineers is to design high etch rate processes, with minimized process time, and without changing the critical dimensions. Additionally, significant gains can be achieved by optimizing the dry clean recipes because they contribute significantly to the raw process time. In some cases, it can take a long time to achieve proper pre and post conditioning of the chamber.

In this work, we show that real-time monitoring techniques can be used to provide insight into the etch process byproducts, enabling intelligent development of the dry clean recipes. Integrating quadrupole mass spectrometry and optical emission spectroscopy with the capacitive coupled plasma reactor, and monitoring in parallel, we were able to identify 99% of the patterned wafer etch byproduct species with high confidence. Knowing the etch byproducts provides a clear path to an optimized dry clean process. We were then able to develop the dry clean recipe with appropriate precursor gases to increase volatile byproducts significantly (e.g. SiF4 at 33%) and minimize non-volatile byproducts (e.g. ammonium salts at <1%), resulting in a 70% reduction of dry clean process time.