Twinning superlattices in indium phosphide nanowires

Superconductivity: unlikely pairs In most superconductors, the pairing-up of electrons responsible for resistance-free conductivity is driven by vibrations of the solid's crystal lattice. But there are other superconducting materials in which the 'glue' responsible for binding electrons is thought to have a very different origin: quantum fluctuations of spin or charge. An unusually 'violent' generalization of such a pairing mechanisms, in which spin and charge instabilities combine forces, has been identified in the unconventional superconductor CeRhIn 5 . These intimately coupled fluctuations significantly disrupt the flow of electrons in their normal unpaired state, yet also provide the quantum-mechanical glue necessary for generating superconducting pairs. In this paper, the crystal structure and stacking fault density of semiconducting nanowires composed of the same material are controlled by doping, leading to twinning superlattices. Periodic arrays of rotational dislocations lead to crystal heterostructures in indium phosphide and gallium phosphide nanowires. Semiconducting nanowires offer the possibility of nearly unlimited complex bottom-up design 1 , 2 , which allows for new device concepts 3 , 4 . However, essential parameters that determine the electronic quality of the wires, and which have not been controlled yet for the III–V compound semiconductors, are the wire crystal structure and the stacking fault density 5 . In addition, a significant feature would be to have a constant spacing between rotational twins in the wires such that a twinning superlattice is formed, as this is predicted to induce a direct bandgap in normally indirect bandgap semiconductors 6 , 7 , such as silicon and gallium phosphide. Optically active versions of these technologically relevant semiconductors could have a significant impact on the electronics 8 and optics 9 industry. Here we show first that we can control the crystal structure of indium phosphide (InP) nanowires by using impurity dopants. We have found that zinc decreases the activation barrier for two-dimensional nucleation growth of zinc-blende InP and therefore promotes crystallization of the InP nanowires in the zinc-blende, instead of the commonly found wurtzite, crystal structure 10 . More importantly, we then demonstrate that we can, once we have enforced the zinc-blende crystal structure, induce twinning superlattices with long-range order in InP nanowires. We can tune the spacing of the superlattices