The void formation and the poor adhesion of the original metal stack to the silicon dioxide substrates both enabled this unusual characterization method and identified the seriousness of the failure mode.

The inability to form good electrical contacts presents a frustrating challenge to those trying to take advantage of high crystalline perfection in GaN nanowires. On the one hand, the p-type doping can’t be measured and optimized because the nanowires can’t be contacted, and on the other hand, the contacts are poor in part because the doping is not optimized.

Adhesion layers

In the study, the team from NIST used different Ti-based adhesion layers to eliminate the delamination of the Ni/Au films after annealing, yet some degradation in the I-V response of the nanowires still occurred. XRD measurements of the Ni/Au layers before and after annealing on p-GaN films and SiO2 with and without an adhesion layer showed that the phase formation and evolution of the crystallinity of the metal films are dependent on the substrate material.

The key to preventing degradation during annealing was the use of an adhesion layer that included Al. The alloy formation of Al with Ni and Au seen in the XRD scans inhibits the mechanisms that are responsible for the void formation. The lack of void formation at the contact/NW interface after annealing when a Ti/Al/Ni adhesion layer is used was confirmed using TEM. The resulting increase in the contact area to the NW surface is believed to be responsible for the large decrease in resistance.

This work shows that alternative contacting schemes, rather than the industrial favourite, NiAu, are critical for contacting and optimizing p-type GaN nanowires.

Full details can be found the journal Nanotechnology.