Noncontact atomic force microscopy imaging mechanism on Ag(110): Experiment and first-principles theory

In this study, we analyze the noncontact atomic force microscopy (NC-AFM) imaging mechanism on the Ag(110) surface by experiment and ab initio theory. The experimental NC-AFM images exhibit atomic-scale resolution in the topography and dissipation signal. Interestingly, the maximum of the damping signal is between the maxima of the topography image. Comparing the geometry of the Ag(110) surface with the topography of a simulated NC-AFM image, we found that the first surface layer silver atoms are imaged as maxima in the topographic NC-AFM images. The overall structure and the corrugation height of the theoretical NC-AFM image are in good agreement with the experimental ones. Furthermore, the analysis of the short-range tip-sample interactions calculated at specific lattice sites revealed strong hysteresis effects. Our simulations also indicate that clean silicon tips might become contaminated with silver atoms during a NC-AFM experiment. Indeed, the NC-AFM experiments showed that the Ag(110) surface is difficult to image probably due to contamination of the silicon tip during the imaging process.

Figure 1

(Color online) Topographic scan of Ag(110) surface with NC-AFM at a frequency shift of $\Delta f=-95.7\mathrm{Hz}$ with a scan size of $200\times 200{\mathrm{nm}}^2$. This image shows a terrace with a typical diameter of $100\mathrm{nm}$. As previously reported in STM experiments (Ref. 36), fringes can be observed at the step edges.

Figure 2

(Color online) (a) Raw data of topography measured on a $2\times 2{\mathrm{nm}}^2$ scan on Ag(110) at a frequency shift of $\Delta f=-265.8\mathrm{Hz}$ and an amplitude of $A=9\mathrm{nm}$. The bright maxima are identified as the top layer atoms of the Ag(110) surface. (b) Simultaneously measured energy-dissipation signal showing a contrast opposite to the topography. Lateral drift was corrected according to the expected values for the lattice constants using the software implemented in the microscope software.

Figure 3

(Color online) (a) Line profile along the [100] direction as indicated by the dotted line in Fig. 2 showing the simultaneously measured topography (blue) and dissipation signals (red). The topography maxima coincide exactly with the dissipation minima. [(b) and (c)] Line profiles along the $\left[1\overline{1}0\right]$ (top) and $\left[1\overline{1}0\right]$ (bottom) lines, respectively. The underlying dashed lines are fits to the measured line profiles used to determine the atomic corrugation along the different crystallographic directions.

Figure 4

(Color online) Top view of the Ag(110) surface. For clarity, the first-layer silver atoms are plotted in gray, while the second-layer atoms are displayed in yellow. The letter A indicates the position of a first-layer Ag atom ($A$ site). The $B$ site marks the position of a second-layer Ag atom and the $C$ site corresponds to the hollow site. The saddle point between two top layer atoms is denoted as $D$ site.

Figure 5

(Color online) The calculated tip-sample interaction forces at a first-layer Ag atom ($A$ site). Bond-formation and bond-breaking processes on the approach and retraction paths lead to an energy dissipation of about $0.5\mathrm{eV}$. The calculated atomic structure at three distinguished tip-sample positions marked as (i), (ii), and (iii) in the force curve are shown at the bottom of the figure. The relaxed geometry depicted in (i) corresponds to a bond-formation process at a tip-sample distance of $3.6\mathrm{\AA }$ when the AFM tip approaches the Ag(110) surface. A bond-breaking process takes place on the retraction path as illustrated in (ii) and (iii) for two consecutive tip-sample distances of 4.2 and $4.4\mathrm{\AA }$, respectively.

Figure 6

(Color online) The calculated tip-sample interaction forces at a second-layer Ag atom ($B$ site). There is almost no hysteresis between the approach and retraction paths. The ball-and-stick model on the right shows the atomic structure of the tip-sample geometry the position marked by an arrow in the force curve. This relaxed configuration obtained in the repulsive regime shows the formation of multibonds between the tip and surface. In particular, this presence of multiple bonds is a fingerprint of AFM operating in the repulsive part of the tip-sample interaction forces.

Figure 7

(Color online) The calculated tip-sample interaction forces at a hollow site ($C$ site). In this case, the AFM tip extracts a Ag atom from the surface, and thus becomes contaminated with silver. Three atomic structures at the tip-sample positions marked as (i), (ii), and (iii) in the force curve are shown at the bottom of the figure. On the approach path, at a tip-sample separation distance of $3.2\mathrm{\AA }$, a chemical bond between the silicon tip and a surface Ag atom is formed, as illustrated by the tip-surface geometry (i). When the tip moves away from the Ag(110) surface, the silicon tip interacts quite strongly with the surface on the whole retraction path. The geometry (ii) shows such a situation for a tip-sample distance of $3.4\mathrm{\AA }$. As displayed by the tip-surface relaxed geometry (iii), at a final distance of $6\mathrm{\AA }$, the silver atom is completely detached from the surface, as demonstrated by the charge density analysis displayed in Fig. 8.

Figure 8

(Color online) Charge density map in a plane cut along the [001] direction of the ${\mathrm{Si}}_4{\mathrm{H}}_3\text{−}\mathrm{Ag}\left(110\right)$ system. The initially clean silicon tip picked up a silver atom from the surface on the retraction path above the hollow site [see Fig. 7(i)]. The absence of the charge density between the Ag atom attached to the silicon tip and the surface clearly marks a bond-breaking process. The charge density plots have been done using the XCRYSDEN program (Ref. 55).

Figure 9

(Color online) (a) The normalized frequency shift at different lattice sites. For the calculations of these curves, we considered also the long-range van der Waals forces. The black arrow marks the corrugation for a normalized frequency shift of $\gamma =-5.5\mathrm{fN}{\mathrm{m}}^{1∕2}$. (b) A simulated NC-AFM image for this value of the normalized frequency shift $(1\times 1{\mathrm{nm}}^2)$. A comparison with the surface structure of Ag(110) allows the identification of the maxima as the first-layer Ag atoms. (c) The scan lines calculated for the positions marked in (b). A comparison with the experimental scan lines shown in Fig. 3 reveals reasonable agreement between theory and experiment.