EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Plasma information–based characterization of wedge-shaped bowed etching profile driven by ion dynamics with the tailored voltage waveform bias

Taejun Park, Jihoon Park, Ji-Won Kwon, Ingyu Lee, Gon-Ho Kim, Seolhye Park

Physics of Plasmas · 01.11.2026 · 33 (11)

DOI: 10.1063/5.0349969 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Physics of Plasmas, online erschienen am 04.11.2026. Verfasst von Taejun Park und 5 weiteren Autor:innen. Die Arbeit nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Research Council of Science and Technology, Ministry of Trade, Industry and Energy, Korea Semiconductor Research Consortium.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Wedge-shaped bowing during high-aspect-ratio Si trench etching under tailored voltage waveform (TVW)-driven dual-frequency capacitively coupled plasma conditions is analyzed from a plasma physics perspective. TVW operation modifies RF sheath dynamics, producing a bi-modal ion energy distribution with an enhanced high-energy component. When reflected from tapered SiO2 mask sidewalls, these ions induce localized energy-flux focusing near the upper trench region, resulting in sharp wedge-shaped profile deformation. Combining in situ diagnostics with feature-scale transport modeling, a plasma-information index, PIBow, is introduced by normalizing the ratio of reflected high-energy ion energy flux to transported O radical flux to a characteristic profile-transition reference scale. This dimensionless index quantifies the competition between ion-driven etching and radical-mediated passivation, and wedge-forming and non-wedge-forming conditions are systematically organized around the normalized reference state PIBow≈1 for the analyzed system. Furthermore, a plasma information–based virtual metrology framework then transforms optical emission spectroscopy and voltage-current sensor measurements into physics-informed plasma variables for wedge geometry prediction. The model reproduces the measured wedge length with a high coefficient of determination (adj. R2 = 0.981) and yields an interpretable process-control window governed by the balance between energetic-ion delivery and O-radical-mediated passivation. An angle-dependent etch rate model is also proposed to describe wedge angle formation, reproducing the measured dependence on the operating conditions. Wedge-shaped bowing thus represents a distinct regime of profile evolution governed by coupled ion dynamics and transport-limited radical supply under TVW conditions, and the methodology presented here offers a scalable framework for understanding, predicting, and engineering plasma–surface interactions in plasma-assisted manufacturing.

Förderung

  • National Research Council of Science and Technology
  • Ministry of Trade, Industry and Energy
  • Korea Semiconductor Research Consortium
  • Government funds, Republic of Korea
  • Ministry of Science and ICT, South Korea
  • Brain Korea 21 FOUR Program
  • Korean government
  • New Faculty Startup Fund from Seoul National University

Zitieren

APA

Park, T., Park, J., Kwon, J., Lee, I., Kim, G., & Park, S. (2026). Plasma information–based characterization of wedge-shaped bowed etching profile driven by ion dynamics with the tailored voltage waveform bias. Physics of Plasmas, 33(11). https://doi.org/10.1063/5.0349969

Vancouver

Park T, Park J, Kwon J, Lee I, Kim G, Park S. Plasma information–based characterization of wedge-shaped bowed etching profile driven by ion dynamics with the tailored voltage waveform bias. Physics of Plasmas. 2026;33(11). doi:10.1063/5.0349969

BibTeX
@article{Park2026,
  author = {Park, Taejun and Park, Jihoon and Kwon, Ji-Won and Lee, Ingyu and Kim, Gon-Ho and Park, Seolhye},
  title = {Plasma information–based characterization of wedge-shaped bowed etching profile driven by ion dynamics with the tailored voltage waveform bias},
  journal = {Physics of Plasmas},
  year = {2026},
  volume = {33},
  number = {11},
  doi = {10.1063/5.0349969}
}
Kostenlos starten →