Apr 17, 2012
Analysis compares growth mechanisms of Zn-polar ZnO nanowires on O-polar ZnO and on sapphire
Crystal polarity governs the shape of zinc oxide nanostructures: ZnO nanowires always grow in the  direction of the wurtzite structure, whereas pyramids grow in the [000-1] direction. These nanostructures are obtained on sapphire substrates, which do not select any polarity. More surprisingly, similar polarities are observed during the growth on (000-1) ZnO substrates. Nanowires sit on top of pyramids, with an inversion domain boundary that is a priori not energetically favourable.
Researchers from the microelectronics institute LETI in Grenoble (France) are exploring the possibility of making light-emitting diodes (LEDs) based on ZnO or GaN nanowires. The first step is to master nanowire growth, which is strongly influenced by crystal polarity. Next, core-shell quantum wells are grown on the nanowires. And finally, the nanowires are integrated into a device.
The team has compared the growth mechanisms of ZnO nanostructures grown by metal organic vapour phase epitaxy on sapphire, and on O-polar ZnO. Because the wurtzite structure of ZnO is not centro-symmetric, the  and [000-1] directions are not equivalent. Growth occurring along the  axis is said to be Zn-polar, while growth occurring along [000-1] is said to be O-polar.
The polarity is determined experimentally by comparing convergent beam electron diffraction patterns with simulations. On (0001) sapphire, O-polar pyramids and Zn-polar nanowires could be obtained: the shape of the nanostructures is thus determined by their growth direction. As sapphire is a non-polar material, it is not very surprising that ZnO of both polarities can be grown.
More surprising is the observation of Zn-polar nanowires on top of O-polar pyramids during the growth on O-polar ZnO or sapphire substrates. Nanowires and pyramids are separated by inversion domain boundaries, which move up during the growth of the nanowire and stay at the top of the pyramids. This movement of inversion domain boundaries, linked to the diffusion of zinc, oxygen and impurities, would be easier than the formation of vertical inversion domain boundaries.
The role of aluminium impurities coming from the substrate is proposed to explain the nucleation of these polarity inversions, and consequently the nanowires. The material could be more stable when impurities segregate at inversion domain boundaries than when they simply stay in a solid solution.
Next, the scientists aim to focus their work on the growth of organized nanowires, and on the difficult p-type doping of ZnO.
Full details on the work can be found in the journal Nanotechnology.
About the author
This work was carried out in the microelectronics institute LETI (Grenoble, France) of the French atomic and alternative energy commission (CEA). New technologies for solid-state lightning are explored, such as ZnO and GaN nanowire-based LEDs. Transmission electron microscopy studies were performed in the nanoscience institute, INAC, on the state-of-the-art nano-characterization platform at Minatec Campus. Guillaume Perillat-Merceroz and Robin Thierry did their PhD from 2008 to 2011 at CEA under the supervision of Guy Feuillet, Pierre-Henri Jouneau and Pierre Ferret, exploring the links between growth conditions, structural and optical properties of ZnO-based materials such as nanowires, core-shell quantum wells and ion implantation defects. Guillaume Perillat-Merceroz is now a postdoctoral researcher at EPFL (Lausanne, Switzerland), studying nitrides mainly with a transmission electron microscope equipped with a cathodoluminescence system.