PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
March 3, 2026Scientific Reports0 citationsOpen Access

Annealing and passivation study of germanium on silicon (GOS) mid-infrared waveguide for sensing applications

RARachel C. F. AngJGJia Sheng GohLTLandobasa Y. M. Tobing

Key Points

  • The study demonstrates a reduction in propagation loss by up to 17 times at roughly 5.85 microns, enhancing sensor efficiency.
  • Annealing treatments under forming gas significantly help in mitigating optical losses in germanium on silicon waveguides.
  • Using atomic layer deposition for passivation with aluminum nitride minimizes oxidation, though it introduces additional waveguide losses.
  • These findings highlight potential improvements in infrared sensing technologies through optimized GOS waveguide devices.

Abstract

Germanium on silicon (GOS) is an excellent platform for non-dispersive infrared sensing (NDIR) due to its broad mid-infrared (mid-IR) transparency. However, its optical waveguide propagation loss as a photonic sensing medium and its susceptibility to oxidation are concerns to be addressed. Herein, we study the effect of annealing GOS waveguide devices under forming gas and passivation using atomic layer deposition (ALD) of aluminum oxide (Al2O3) and aluminum nitride (AlN) on waveguide loss. Our findings showed that annealing helped reduce propagation loss as high as 17x at wavelength of ~ 5.85 μm and passivation with AlN was effective in minimising oxidation of germanium (Ge) in ambient, albeit at the expense of higher waveguide loss originating from the MIR absorption in the AlN film itself. Nevertheless, these results provide insights towards improving the performance and robustness of a GOS waveguide-based sensor.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Ang et al. (2026) studied this question.

synapsesocial.com/papers/69a7656dbadf0bb9e87d911dhttps://doi.org/10.1038/s41598-026-35766-1
Ask AI
Helpful
Bookmark
Share
View Full Paper