The electrostatic force microscope (EFM) maps the spatial variation of the potential energy difference between a tip and a sample which results from anon-uniform charge distribution, variations in surface workfunction, etc. Unfortunately, the resolution of the EFM has never compared with that of other scanning probe methods such as atomic force microscopy. We thought perhaps part of the problem was that it is difficult to find sample surfaces which are highly corregated in terms of the potential energy. Since silver nanocrystallites on graphite should carry a positive charge (the work function of silver is 500 meV negative of graphite), these surfaces are ideal for probing the resolution which is attainable using the EFM mode.

Silver nanoparticles on graphite basal plane surfaces were concurrently imaged using electrostatic force microscopy (EFM) and non-contact atomic force microscopy. EFM images - one shown below center - were obtained having a lateral resolution 4-5 nm, and a resolution perpendicular to the surface of 1 nm. The dependence of the contrast in the EFM data for the silver nanoparticles as a function of the applied tip bias was consistent with a positive charge for silver nanocrystals on the graphite surface - qualitatively as expected by theory.

At left is shown an NC-AFM image showing the topography of electrochemically deposited silver nanoparticles having an average height of about 20Å on graphite. The image window is 0.5 x 0.5 µm² and was recorded at 0.3 Hz. As commonly observed, an increased density of particles can be found along the step edge, whereas on the basal plane a smaller areal density of particles is seen. The center image shows the spatial variation of the EFM signal acquired concurrently with the topography image at left. The image has been processed by means of a line-by-line subtraction in order to remove large steps which were introduced by the DC bias steps which were applied during the image acquisition. The contrast of the large silver particle inverts relative to background as the sign of the applied DC bias changes. In this instance, the bias was switched from +10 V to -8 V. At right is shown the applied DC bias (left) and the unprocessed EFM signal (right) plotted as a function of position in the slow scan direction. The expected correlation between the sign and the amplitude of the applied DC bias and the resulting EFM signal is seen.