Fabrication and field emission properties of vertical, tapered GaN nanowires etched via phosphoric acid

Abstract

The controlled fabrication of vertical, tapered, and high-aspect ratio GaN nanowires via a two-step top-down process consisting of an inductively coupled plasma reactive ion etch followed by a hot, 85% H3PO4 crystallographic wet etch is explored. The vertical nanowires are oriented in the direction and are bound by sidewalls comprising of semipolar planes which are at a 12° angle from the [0001] axis. High temperature H3PO4 etching between 60 °C and 95 °C result in smooth semipolar faceting with no visible micro-faceting, whereas a 50 °C etch reveals a micro-faceted etch evolution. High-angle annular dark-field scanning transmission electron microscopy imaging confirms nanowire tip dimensions down to 8–12 nanometers. The activation energy associated with the etch process is 0.90 ± 0.09 eV, which is consistent with a reaction-rate limited dissolution process. The exposure of the type planes is consistent with etching barrier index calculations. The field emission properties of the nanowires were investigated via a nanoprobe in a scanning electron microscope as well as by a vacuum field emission electron microscope. The measurements show a gap size dependent turn-on voltage, with a maximum current of 33 nA and turn-on field of 1.92 V nm−1 for a 50 nm gap, and uniform emission across the array.

Wet-based digital etching on GaN and AlGaN

“…3(a)). No clear change on the morphology of the GaN vertical tips sidewalls was observed after piranha etch, and the estimated sidewall slope remains about 72 -75 degree, which is very close to the value reported by the H 3 PO 4 work 4 . It should be noted that this slope is based on 60-degree-tilted SEM images rather than from direct cross-sectional images, so a minor error could exist in this estimation.…”

“…It is more difficult to estimate the tip width for n + AlGaN tips because of SEM charging effects once the size is below 20 nm. The sharpest tip radius 4 . The shrinking rates on AlGaN (C1 and C2) are higher than GaN (A), and the sidewall surface of the AlGaN tips becomes rougher after piranha clean and after DE (Fig.…”

“…Therefore, the Al 2 O 3 could be formed on the semi-polar high-Al AlGaN sidewalls and be etched by the mixture of H 2 SO 4 and H 2 O 2 21 , while the strong polarization of the c-plane (0001) AlGaN surface limits the oxidization and thus the etching rate. The increasing roughness on the sidewalls of AlGaN vertical tips could indicate microfaceting due to orientation-dependent etching rates 4,22,23 or the probable segregation in AlGaN alloys 24,25 . DE experiments on semi-polar or non-polar AlGaN materials in the future would allow to further investigate the hypothesis of lattice-plane-dependent etching rates.…”

Etching of the SiGaxNy Passivation Layer for Full Emissive Lateral Facet Coverage in InGaN/GaN Core–Shell Nanowires by MOVPE

“…One hypothesis is that the 450 nm emission is due to a greater incorporation of indium on the edges, which has been previously detected and ascribed to a combination of strain relaxation occurring at the edges 58 along with the formation of nanometric a-plane facets. 59,60 Another explanation could be that H 3 PO 4 etches preferentially the a-planes over m-planes in GaN, which is supported by a recent study on etching barrier index calculations, 49 as introduced in ref 46 to approximately predict relative resistance to the etching of each crystallographic plane. A similar conclusion could be made for the top of the KOH SQW sample but not for most of its sidewalls.…”

“…Diluted buffered oxide etch (BOE), potassium hydroxide (KOH), and phosphoric acid (H 3 PO 4 ) have been chosen as they were all reported to etch SiGa x N y . Moreover, KOH is able to etch GaN, especially N-polar c -planes and semi-polar planes. − H 3 PO 4 was also reported to etch semi-polar planes in GaN NWs. ,− More precisely, Deeb et al showed by surface photovoltage measurements that H 3 PO 4 treatment removes the surface states and reveals bulk-like properties of the underlying GaN. The exact protocols were as follows: the BOE was diluted 80 times, and the sample was treated for 10 s at room temperature.…”