EFTEM studies of Si nanowire networks in SiO2 for thin film PV cells

Si based nanostructures became in the last years a promising material for the PV cells. The quantum confinement effect of Si nanostructures allows for band gap engineering by size manipulation which can be used for optimization sun light absorption.
Here, we consider SiOx layers deposited by magnetron sputtering. By subsequent rapid thermal processing, SiOx decays by spinodal decomposition into a network of Si nanowires (NWs) in SiO2.
To get images of the morphology of Si NWs in SiO2 it is not sufficient to use mass contrast or lattice plane imaging. The Si and SiO2 phases can only be distinguished by energy filtering of the transmitted electrons (EFTEM). Here, the relative energy shifts of the plasmonic valence band resonances of Si and SiO2 are used. HR-EFTEM techniques are applied to study morphology and crystallinity of the Si NW networks fabricated from different metastable SiOx. To facilitate understanding of the TEM images, details of decomposition are studied using kinetic Monte-Carlo (KMC) simulations. For the EFTEM images, density-density correlations are calculated to determine the structure size of NW network, which are then compared with the 3D morphologies provided by KMC. Combining EFTEM with KMC allows us to predict and control the average size of the NWs.
Former studies and our electronic structure calculations provide a guideline for band gap optimization of Si NW networks, thus paving the way to band gap engineering via control of the mean NW diameter.