Despite the paramount importance of conductance measurements to bulk solid state physics, surface conductance is very poorly understood. A prominent example for the sorry state of the field is the Si(111)(7x7) reconstruction. Several claims to measure the surface conductance have been made but the resulting values span 5 orders of magnitude and in some cases the results even depend on the bulk doping - which is highly improbable. Many of these problems can be related to the indirect methods used. Only very recently, direct quantitative measurements were provided [1].
Experimentally, surface conductance can be measured in the same way as bulk conductance by using four point probes. Surface sensitivity can be achieved by choosing a sufficiently small distance between the probes. Different approaches have been suggested to achieve this, e.g. monolithic four point probes or multi-tip STM systems. The principle advantage of the latter is that variable probe spacing (and thus variable surface sensitivity) can be achieved. However, in addition to reliability concerns, the smallest probe spacing achievable with such instruments is somewhat limited. This limits the application of the instrument to materials with a poorly conducting bulk.
In this work, we present a multi-contact mono-cantilever probe, which differs from earlier monolithic four point probes by that fact that all contacts are mounted on a single cantilever [2]. In this approach, the minimum spacing can be reduced to the order of 100 nm, whilst keeping the cantilever size in the micrometer range. Furthermore, many contacts can be reliably formed on one cantilever, thus "variable spacing" can be achieved reliably by switching between the various probe configurations. The mono-cantilever also has significant mechanical advantages.
As an example, we will present measurements made on the low-index Bismuth surfaces. In contrast to the semimetallic bulk, these surfaces support metallic surface states, forming a quasi two dimensional metal [3]. Using measurements with different probe spacings, we are able to estimate a lower limit of the conductance through these surface states.
[1] J.W. Wells, et al. PRL. 97, 206803 (2006).
[2] Capres A/S (www.capres.com)
[3] Ph. Hofmann, Prog. Surf. Sci. 81, 191-245 (2006). |