In Figure 3d, the scanned area at the center of the image is observed as a shallow 10058-F4 manufacturer hollow, the cross-sectional profile of which revealed its depth to be approximately 1.0 nm. In contrast, the multiple scans (ten scans) using a Pt-coated cantilever in SOW created a clear square hollow, as shown in Figure 3e,f. The etched depth of the 1.0 × 1.0 μm2 central area in Figure 3f was about 4.0 nm from a cross-sectional profile. The mechanism of inducing the

difference between image (d) and image (f) in Figure 3 is as follows. As mentioned previously, we scanned a cantilever in the contact mode. Taking into account the catalytic activity of metals (e.g., Pt) enhancing the reactions in Equations (1) and (2), we suppose that, at each moment during the scan, a Ge surface in contact with a Pt probe is preferentially oxidized in water in the presence of dissolved oxygen. This is schematically drawn in Figure 4a. Owing to the soluble nature of GeO2, the scanned area exhibits a square hollow, as shown in Figure 3f. In Figure 3b,d,f taken after the ten scans, no PF-01367338 pyramidal pits such as those shown in Figure 1 are observed. This is because we did not fix the cantilever at only one surface site, but rather scanned it over a micrometer area, which is much larger than the etched depth, as schematically depicted in Figure 4b. Figure 5a,b shows a summary of etched depth as a function of pressing force on the n-type and p-type Ge(100) surfaces, respectively.

Because the plots in Figure 5 slightly fluctuate, it is hard to fit them using a simple straight line or a curve. This is probably due to the difference

in probe apex among the sets of experiments performed. However, Figure 5 clearly indicates that (1) the catalytic activity of metals (e.g., Pt) has a much greater effect on Ge etching than that of the mechanical machining caused by a pressurized cantilever, and (2) dissolved oxygen in water is the key molecule in metal-assisted etching. Namely, it is easy to imagine that the Ge surface was machined mechanically to some extent by the pressed cantilever on Ge. In Figure 5, the etched IKBKE depth increases slightly at a larger pressing force even with a Si cantilever in SOW (light gray filled circles) or a Pt-coated cantilever in LOW (gray filled circles). This indicates that the mechanical etching of Ge occurs, but its effect is very small. On the other hand, a drastic increase in etched depth is observed with a Pt-coated cantilever in SOW (blue filled circles) at each pressing force, which is probably induced by the catalytic effect of Pt mediated by dissolved oxygen in water. One may think that the difference in etched depth between the blue and gray (or light gray) filled circles increases with increasing pressing force in Figure 5. This is as if the catalytic effect is enhanced at greater pressing forces. As for the reason for this enhancement, we imagine that the probe apex became blunter at larger forces.