We present a general model for the vapor-liquid-solid nanowire (NW) growth rates which accounts for adatom diffusion from the substrate and sidewalls into the Au catalyst drop as well as the Gibbs-Thomson effect of elevated chemical potential in the drop with a curved surface. The growth model is compared with the experimental length-diameter dependences for InP and Si NWs grown via metal organic chemical vapor deposition (MOCVD) and GaAs nanowires grown via molecular beam epitaxy (MBE). We show that MBE growth is affected mainly by adatom diffusion from the substrate, whereas MOCVD growth is affected mainly by direct Au drop impingement and sidewall diffusion. The Gibbs-Thomson effect is shown to limit growth for smaller diameter nanowires. Fits for diffusion lengths and the Gibbs-Thomson radii are determined which explain the experimental length-diameter dependence observed.

The incorporation of metal atoms into silicon nanowires during metal-particle-assisted growth is a critical issue for various nanowire-based applications. Here we have been able to access directly the incorporation and redistribution of metal atoms into silicon nanowires produced by two different processes at growth rates ranging from 3 to 40 nm s⁻¹, by using laser-assisted atom probe tomography and scanning transmission electron microscopy. We find that the concentration of metal impurities in crystalline silicon nanowires increases with the growth rate and can reach a level of two orders of magnitude higher than that in their equilibrium solubility. Moreover, we demonstrate that the impurities are first incorporated into nanowire volume and then segregate at defects such as the twin planes. A dimer-atom-insertion kinetic model is proposed to account for the impurity incorporation into nanowires.

Incorporation of nanostructures is a recent trend in the photovoltaic community, aimed at improving light absorption and consequently cell efficiency. In this regard, semiconductor nanowires provide an attractive research platform for a new generation of cost-effective and efficient solar cells. Thanks to their unique geometry, silicon nanowires enhance light trapping and anti-reflection effects by means of multiple scattering between individual nanowires, and by coupling the light into confined eigenmodes over a broad range of the solar spectrum. Moreover, radial junction solar cells built around nanowires decouple the light absorption and carrier collection directions, which allows for a higher internal field and better carrier collection. Thus, arrays of radial junction solar cells bring advantages of high efficiency with reduced material amount. This is particularly attractive for devices based on hydrogenated amorphous and microcrystalline silicon thin films. In this paper, after reviewing different approaches to fabricate silicon nanowires, we focus on nanowires grown using the plasma-assisted vapour–liquid–solid method because of the simplicity and compatibility with current silicon thin-film technology. Their application to a-Si : H based radial junction solar cells has already resulted in ∼8% of stable devices with an absorber layer thickness of only 100 nm. Moreover, current challenges and perspectives such as the use of a microcrystalline silicon absorber are also reviewed.

Up-scaling silicon nanowire (SiNW)-based functionalities requires a reliable strategy to precisely position and integrate individual nanowires. We here propose an all-in-situ approach to fabricate self-positioned/aligned SiNW, via an in-plane solid-liquid-solid growth mode. Prototype field effect transistors, fabricated out of in-plane SiNWs using a simple bottom-gate configuration, demonstrate a hole mobility of 228 cm²/V s and on/off ratio >10³. Further insight into the intrinsic doping and structural properties of these structures was obtained by laser-assisted 3 dimensional atom probe tomography and high resolution transmission electron microscopy characterizations. The results could provide a solid basis to deploy the SiNW functionalities in a cost-effective way.