Back to Highlights

(2011) Transport in Ultra-Thin Silicon Membranes


graph of conductance and energy versus voltage

Fig. A Top: Square symbols are the conductance of a 120 nm thick silicon membrane as a function of a back-gate voltage. The blue line is the simulated conductance without a direct contribution from mobile electrons in the surface pi* band. The blue curve includes such a contribution, and it is a much better fit to the experimental data. The lower panel shows the calculated Fermi level at both the surface and the back interface.

Nanostructures are extremely sensitive to their surface properties. In the past year we have made extensive measurements of thin silicon films in ultra-high vacuum. We now have strong evidence that ultra-thin silicon with a 2x1 surface reconstruction is heavily influenced by the surface pi* band. Fig. A shows experimental measurements (black squares) of the conductance as a function of a back gate voltage for a membrane of thickness 120nm. There is a clear minimum in conductance for gate voltages near zero. We have shown that this minimum is significantly less deep than it otherwise would be, because of the role of the surface pi* band. In order to understand this effect, we have developed a model of the interaction between the surface and the thin “bulk” of the membrane, and calculations for the model have been performed by students in Professor Knezevic’s group. 

By comparing the data and the theoretical simulations, it is clear that the surface pi* band affects the membrane conductance in two different ways. First, thermally excited electrons populating the pi* band leave behind holes in the bulk of the membrane, and these holes are highly mobile, contributing to the conductance of the membrane (blue curve in Fig. A, top). Second, the electrons occupying the pi* band themselves are also mobile, and that mobility contributes to the conductance of the membrane (red curve in Fig. A, top). The Fermi level as a function of gate voltage is shown in Fig. A, bottom. The calculations shown in Fig. A are still preliminary, as details of the model for the surface states are still being developed, and in this part of the work we are benefiting from interactions with Professor Franz Himpsel of IRG3.