Reducing Conductivity Variability in Si Nanowires via Surface Passivation for Nanoelectronics

Ensembles of semiconductor nanowires grown via the bottom-up vapor–liquid–solid (VLS) mechanism, especially those that are lightly doped or nominally undoped, can exhibit large nanowire-to-nanowire variations in electrical conductivity. This broad conductivity distribution, attributed to uncontrolled surfaces and the large surface area of nanowires, limits the fabrication of homogeneous ensembles of nanoelectronic devices, including transistors, photovoltaics, and biosensors. While methods to control surfaces are well understood for planar surfaces, the diversity of surface structures in a nanowire ensemble introduces new processing and characterization challenges. Here, we employ electro-orientation spectroscopy, a high-throughput, solution-based method, to measure the conductivity distributions and quantify the variability of as-synthesized and postprocessed Si nanowire ensembles. Our measurements reveal a conductivity distribution with an unusual, highly skewed non-Gaussian shape, whose variability is best quantified with a log-normal coefficient of variation (COV). We demonstrate a reduction in the COV up to 2.6× as a function of increasing conformal Al2O3 thickness. The decreased COV and accompanying increase in mean conductivity are consistent with a narrower distribution of surface-state densities upon passivation. Our findings highlight the surface-dependent variations inherent to bottom-up nanowire processing, and the need for advanced processes and analytical tools to control these variations for nanoelectronic applications.