蚀刻(微加工)
材料科学
原子层沉积
微电子
等离子体刻蚀
聚合物
表面粗糙度
沉积(地质)
等离子体
氟碳化合物
纳米技术
图层(电子)
分析化学(期刊)
化学
复合材料
有机化学
物理
古生物学
沉积物
生物
量子力学
作者
Xifeng Wang,Mingmei Wang,Peter Biolsi,Mark J. Kushner
出处
期刊:Journal of vacuum science & technology
[American Institute of Physics]
日期:2021-03-16
卷期号:39 (3)
被引量:11
摘要
Fabricating sub-10 nm microelectronics places plasma processing precision at atomic dimensions. Atomic layer etching (ALE) is a cyclic plasma process used in semiconductor fabrication that has the potential to remove a single layer of atoms during each cycle. In self-limiting ideal ALE, a single monolayer of a material is consistently removed in each cycle, typically expressed as EPC (etch per cycle). In plasma ALE of dielectrics, such as SiO2 and Si3N4, using fluorocarbon gas mixtures, etching proceeds through deposition of a thin polymer layer and the process is not strictly self-terminating. As a result, EPC is highly process dependent and particularly sensitive to the thickness of the polymer layer. In this paper, results are discussed from a computational investigation of the ALE of SiO2 on flat surfaces and in short trenches using capacitively coupled plasmas consisting of a deposition step (fluorocarbon plasma) and an etch step (argon plasma). We found that ALE performance is a delicate balance between deposition of polymer during the first half cycle and etching (with polymer removal) during the second half cycle. In the absence of complete removal of the overlying polymer in each cycle, ALE may be transient as the polymer thickness grows with each cycle with a reduction in EPC until the thickness is too large to enable further etching. Small and statistical amounts of polymer left from a previous cycle can produce statistical variation in polymer thickness on the next cycle, which in turn can lead to a spatially dependent EPC and ALE roughness. Based on synergy between Ti (sputtering time) and Tp (passivation time), dielectric ALE can be described as having three modes: deposition, roughening surface (transitioning to etch-stop), and smooth surface with steady-state EPC.
科研通智能强力驱动
Strongly Powered by AbleSci AI