MOCVD Growth and Structural Properties of ZnS Nanowires: A Case Study of Polytypism.

Controlling the morphology, orientation, and crystal phase of semiconductor nanowires is crucial for their future applications in nanodevices. In this work, zinc sulfide (ZnS) nanowires have been grown by metalorganic chemical vapor deposition (MOCVD), using gold or gold-gallium alloys as catalyst. At first, basic studies on MOCVD growth regimes (mass-transport, zinc- or sulfur- rich conditions) have been carried out for ZnS thin films. Subsequently, the growth of ZnS nanowires was investigated, as a function of key parameters such as substrate temperature, S/Zn ratio, physical state and composition of the catalyst droplet, and supersaturation. A detailed analysis of the structural properties by transmission electron microscopy (TEM) is given. Depending on the growth conditions, a variety of polytypes is observed: zinc-blende (3C), wurtzite (2H) as well as an uncommon 15R crystal phase. It is demonstrated that twinning superlattices, i.e., 3C structures with periodic twin defects, can be achieved by increasing the Ga concentration of the catalyst. These experimental results are discussed in the light of growth mechanisms reported for semiconductor nanowires. Hence, in this work, the control of ZnS nanowire structural properties appears as a case study for the better understanding of polytypism in semiconductor 1D nanostructures.

Why is it difficult to grow spontaneous ZnO nanowires using molecular beam epitaxy?

Surface diffusion is known to be of prime importance in the growth of semiconductor nanowires. In this work, we used ZnMgO layers as markers to analyze the growth mechanisms and kinetics during the deposition of ZnMgO/ZnO multilayered shells by molecular beam epitaxy on previously grown ZnO nanowire cores (so called core-shell heterostructures). Specifically, the influence of the O2 flow sent into the plasma cell on the adatom surface mobility was investigated. By carefully measuring the growth rate on the lateral facets as well as on the top of the nanowires, it is concluded that the surface diffusion length of adatoms, within the used MBE growth conditions, is very low. Such poor surface mobility explains why so few works can be found related to the spontaneous growth (without catalyst) of ZnO nanowires by MBE, contrary to other deposition techniques.