Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Impact of Oxygen Plasma Surface Treatment on Photoresist Adhesion in BaTiO3-Based Photonic Device Fabrication

Weiyou Kong, Weijia Kong, Lukas Chrostowski, Xin Xin

TL;DR

The paper addresses photoresist delamination on BaTiO$_3$ thin films caused by oxygen-plasma cleaning. It combines optical microscopy, EDS, and XPS to show that plasma induces a nanometer-scale BaCO$_3$-rich interphase at the resist/BaTiO$_3$ boundary, weakening adhesion without altering bulk stoichiometry. The adhesion loss is reversible by solvent cleaning but recurs with plasma exposure, highlighting a process fragility and guiding practitioners toward alternative cleaning methods or controlled plasma windows. This work provides a mechanistic link between plasma-induced surface chemistry and lithography failure, with practical guidance for reliability in ferroelectric oxide photonics.

Abstract

Oxygen-plasma pre-cleans are routine before fabrication, but on BaTiO3 thin films we observed catastrophic photoresist lift-off during mild rinsing and sonication. To explain the failure, we combined optical microscopy, EDS, and XPS. EDS showed no meaningful bulk stoichiometry change, whereas XPS revealed a nanometer-scale, plasma-induced shift in surface chemistry: hydroxylation and carbonate formation consistent with a BaCO3-rich interphase at the resist/BaTiO3 boundary. This chemically weak interphase, recreated upon each plasma step and removable by simple solvent cleaning, provides the mechanism for delamination. The key takeaway for practitioners is process guidance: avoid uncritical O2-plasma use on BTO; if cleaning is required, use alternative chemistries (e.g., UV-ozone) or carefully tuned plasma windows that preserve adhesion. More broadly, the study illustrates how lightweight analytics at the surface (correlative microscopy + surface spectroscopy) can pinpoint the root cause of yield-limiting defects in oxide photonics and translate directly into higher-reliability process recipes.

Impact of Oxygen Plasma Surface Treatment on Photoresist Adhesion in BaTiO3-Based Photonic Device Fabrication

TL;DR

The paper addresses photoresist delamination on BaTiO thin films caused by oxygen-plasma cleaning. It combines optical microscopy, EDS, and XPS to show that plasma induces a nanometer-scale BaCO-rich interphase at the resist/BaTiO boundary, weakening adhesion without altering bulk stoichiometry. The adhesion loss is reversible by solvent cleaning but recurs with plasma exposure, highlighting a process fragility and guiding practitioners toward alternative cleaning methods or controlled plasma windows. This work provides a mechanistic link between plasma-induced surface chemistry and lithography failure, with practical guidance for reliability in ferroelectric oxide photonics.

Abstract

Oxygen-plasma pre-cleans are routine before fabrication, but on BaTiO3 thin films we observed catastrophic photoresist lift-off during mild rinsing and sonication. To explain the failure, we combined optical microscopy, EDS, and XPS. EDS showed no meaningful bulk stoichiometry change, whereas XPS revealed a nanometer-scale, plasma-induced shift in surface chemistry: hydroxylation and carbonate formation consistent with a BaCO3-rich interphase at the resist/BaTiO3 boundary. This chemically weak interphase, recreated upon each plasma step and removable by simple solvent cleaning, provides the mechanism for delamination. The key takeaway for practitioners is process guidance: avoid uncritical O2-plasma use on BTO; if cleaning is required, use alternative chemistries (e.g., UV-ozone) or carefully tuned plasma windows that preserve adhesion. More broadly, the study illustrates how lightweight analytics at the surface (correlative microscopy + surface spectroscopy) can pinpoint the root cause of yield-limiting defects in oxide photonics and translate directly into higher-reliability process recipes.

Paper Structure

This paper contains 18 sections, 1 equation, 4 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Experimental details
  3. Specimen preparation
  4. Photoresist delamination test
  5. Energy-dispersive spectroscopy (EDS)
  6. X-ray photoelectron spectroscopy (XPS)
  7. Results
  8. Photoresist delamination result
  9. EDS result
  10. XPS results
  11. Discussion
  12. Photoresist delamination and its reversibility
  13. EDS evidence: absence of bulk compositional change
  14. XPS analysis: surface chemistry and carbonate formation
  15. Mechanism: plasma-induced BaCO3 interlayer causes resist delamination
  16. ...and 3 more sections

Figures (4)

  • Figure 1: Optical micrographs (2.5 mm × 3.5 mm field of view per panel) of a 5 mm × 5 mm BaTiO$_3$ wafer coated with ZEP520A photoresist, comparing conditions without and with oxygen-plasma pretreatment, before and after ultrasonication. Dashed lines outline the Uniform Photoresist Region, whose disappearance indicates complete film removal. (a,b) No plasma: before and after ultrasonication; (c,d) With plasma: before and after ultrasonication. The comparison highlights the loss of photoresist adhesion following oxygen-plasma exposure.
  • Figure 2: EDS mapping of a $5\,{mm} \times 5\,mm$ BaTiO$_3$ sample at an accelerating voltage of $5\,keV$. (a) Spectrum acquired from the as-prepared BaTiO$_3$ surface without oxygen plasma treatment. (b) Spectrum acquired from the BaTiO$_3$ surface after oxygen plasma treatment.
  • Figure 3: XPS spectra of a BaTiO$_3$ thin film that was sequentially cleaned with propanol and isopropanol and then immediately transferred into the XPS chamber for measurement. Panels: (a) Ba 3d, (b) O 1s, (c) Ti 2p, (d) C 1s.
  • Figure 4: XPS spectra of a BaTiO$_3$ thin film after cleaning with propanol and isopropanol, followed by an additional O$_2$ plasma surface treatment before immediate transfer into the XPS chamber. Panels: (a) Ba 3d, (b) O 1s, (c) Ti 2p, (d) C 1s.