This Work consists of Scanning Probe Capacitance measurements made with an Atomic Force Microscope (AFM) tip as an electrode in contact with two different materials. In the first part we used a Gallium Arsenide wafer as sample composed with layers of different doping concentrations previously defined by simulating capacitance curves. These capacitance measurements were performed with Schottky and MOS junctions. The second part consists of measurements performed on a Silicon based solar cell.
We have also measured the capacitance of the same GaAs wafer in the same conditions on a large scale set up (device) which is well understood and easier to describe from the point of view of the Poisson's equation solution and these measurements were used for a comparison with the AFM measurements. Except by the surface impurities (such as oxides and water films) that can influence the capacitance curve behavior in the AFM, in the Schottky case, no significant differences between the AFM and the device measurements were found. On the other hand, in the MOS case, the capacitance curve behavior is completely different in both set ups (AFM and device). In the large scale set up, at 1kHz, no low frequency behavior is presented. On the other way around, the capacitance curve measured with the AFM tip presents a low frequency characteristic for measurements performed with frequencies up to 20kHz. The believed reason for this unexpected behavior is the spherical symmetry of the electrical field produced by the AFM tip, the high density of minorities right below the oxide/wafer interface due to the surface states of GaAs and the charges contained in the oxide that help attracting the minorities carriers to the surface. The cutoff frequency for this low frequency behavior of the capacitance has been estimated to be 3,85 kHz and the results of these measurements in the GaAs-MOS capacitor in the AFM will be submitted for publication in specialized scientific journals.
In the second part of this work, in chapter 5, we performed capacitance measurements along a Si p-n junction and, observing the change of behavior of the capacitance curves that go from high (while the tip was set in the p-type neutral bulk) to low frequency (as the tip approaches the space charge region of the junction), we were able to determine that the electrons diffused from the n-type approximately 288nm inside of the p-type Si (space charge region). From those measurements, we could also estimate the doping profiles of each Si type.