Fabrication of high-aspect-ratio, sub-micron gratings in AlGaInP/GaAs laser structures using a BCl3/Cl2/Ar inductively coupled plasma

Autor: Angela Sobiesierski, Peter Michael Smowton, Gareth Edwards, David I. Westwood
Rok vydání: 2007
Předmět:
Zdroj: Semiconductor Science and Technology. 22:1010-1015
ISSN: 1361-6641
0268-1242
DOI: 10.1088/0268-1242/22/9/006
Popis: A process for fabricating deep-etched nanostructures in AlGaInP and GaAs using a BCl3/Cl2/Ar inductively coupled plasma (ICP) is reported. The sidewall profile of 4 µm wide ridges, etched using a two-step process, is shown to be strongly dependent on the fraction of Ar present in the process gas and the mechanisms for etching vertically for the two materials are shown to be quite different. GaAs etches chemically in Cl-rich environments causing severe undercut whereas AlGaInP is strongly passivated. As the Ar fraction was increased and the etch became more physical, the unpassivated GaAs cap directly underneath the Ni mask etched more anisotropically, acquired a vertical sidewall between 40 and 75% Ar and remained vertical whilst the mask remained intact. AlGaInP behaved very differently, showing clear evidence of InCl3 passivation (a positive sidewall profile) at low Ar concentrations and slight undercutting (indicative of passivant removal and chemical, isotropic etching) at 90% Ar. The AlGaInP sidewall etched vertically with ~80–85% Ar. At 80% Ar, where the AlGaInP etch rate reached a peak value of (0.85 ± 0.05) µm min−1, a very large RMS surface roughness of (25.3 ± 1.0) nm was measured, but this is reduced (to ~1 nm) at 82–85% Ar while maintaining a high etch rate. Despite the difference in etch mechanisms, vertical sidewall features were achieved at the same Ar fraction not only on relatively wide ridges but also in a 1D array with 200 nm lattice constant and 20:1 aspect ratio in both materials. To the best of our knowledge, these are the deepest vertically etched structures on a scale suitable for fabrication of first-order photonic crystals in AlGaInP.
Databáze: OpenAIRE