Improve the Accuracy of Scanning Kelvin Probe Microscopy by Eliminating the Cantilever Effect
- PDF / 465,032 Bytes
- 6 Pages / 595 x 842 pts (A4) Page_size
- 70 Downloads / 194 Views
O11.7.1
Improve the Accuracy of Scanning Kelvin Probe Microscopy by Eliminating the Cantilever Effect Zhitao Yang and Michael G. Spencer School of Electrical and Computer Engineering, Cornell University Ithaca, NY 14850, U.S.A. ABSTRACT Scanning Kelvin probe microscopy (SKPM) is widely used to measure surface work functions and electrostatic potentials on nanoscale circuits, devices and materials. However, the accuracy of scanning Kelvin probe microscopy is reduced by a cantilever effect, which is due to a large capacitance gradient associated with the cantilever. We introduce an aperture structure to quantitatively moderate the strength of the cantilever effect. In this approach, the cantilever effect is eliminated and the true surface potential can be extracted by solving a set of linear equations. Experimental results show that this approach yields very accurate surface potentials when there is only a single potential source within the aperture. In the case of multiple potential sources, this method significantly improves accuracy as well. A mobile aperture structure mounted on a micromanipulator can make this approach more versatile. INTRODUCTION Scanning Kelvin probe microscopy (SKPM) is one of the most important characterization tools in nanotechnology. In conjunction with atomic-force microscopy (AFM), it can be used as a voltmeter to quantitatively measure the local nanoscale electrostatic potential distribution on the surface of integrated circuits, devices and semiconductor materials [1]. As device dimensions are rapidly shrinking, SKPM is playing a more and more important role. However, the accuracy of SKPM is not sufficient in some cases, especially when the local potential variation is large on a nanometer scale. One of the principal reasons for this is the cantilever effect. The cantilever effect, which is due to the stray capacitances contributed by the cantilever (and the tip cone), results in the surface potential image being “smeared-out” and the potential magnitude appearing smaller than the true value. In other words, both the contrast and the resolution of the potential image are degraded, since the observed potential is a locally weighted average over all surface potentials. The weighting factor is the derivative of the capacitance [2]. The cantilever effect has been studied by simulations and experiments [2-5]. Long and slender but slightly blunt tips on narrow cantilevers have been used to obtain good potential images [2]. The accuracy can also be improved by decreasing the tip-sample distance [3]. It has been proposed that the force gradient (rather than the electrostatic force) can be used as the signal source to increase the resolution [6,7]. In our work, a linear model is introduced to characterize the cantilever effect. An aperture structure on a shielding layer is proposed to quantitatively moderate the strength of the cantilever effect. By applying different DC biases to the shielding layer and solving a series of linear equations, the normalized derivative of the capacitance and the
Data Loading...
