Origin of the ~15° etch-angle deviation in 4H-SiC RIBE triangular photonics

Determine the physical and process mechanisms responsible for the approximately 15° deviation between the realized etch angle of triangular cross-section structures and the substrate holder tilt angle during reactive ion beam etching of bulk 4H-SiC using SF6/O2 chemistry, and ascertain whether nickel hard-mask edge shrinkage and sidewall/apex over-etching during undercut formation are the dominant contributors.

Background

The paper demonstrates wafer-scale fabrication of suspended triangular cross-section photonic devices in bulk 4H-SiC using reactive ion beam etching (RIBE) with a tilted and rotating substrate holder. Ideally, the etch angle of the triangular cross-section should match the mechanical tilt angle of the substrate holder, enabling predictable device geometry and straightforward mask compensation.

However, measurements consistently show that the realized etch angle deviates from the expected value by about 15°. This unresolved deviation limits precise control of photonic device geometry and motivates identifying the underlying mechanisms, for which the authors suggest mask edge shrinkage and sidewall/apex over-etch during extended undercut etching as potential causes.

References

The origin of the 15${\circ}$ deviation in the etch angle from the ideal scenario, where the etch angle is equal to the tilt angle, is yet to be fully understood. Some possible explanations for the origins of this issue could be shrinking of the metal mask at the edges during the etch or over-etching of the sidewalls and the apex of the triangular cross-section during the additional etch time required to create an undercut, which would require further investigations.

— Wafer-Scale Integration of Freestanding Photonic Devices with Color Centers in Silicon Carbide  (2405.07498 - Majety et al., 2024) in Discussion