High-precision atomic force microscopy with atomically-characterized tips

Abstract

Traditionally, atomic force microscopy (AFM) experiments are conducted at tip–sample distances where the tip strongly interacts with the surface. This increases the signal-to-noise ratio, but poses the problem of relaxations in both tip and sample that hamper the theoretical description of experimental data. Here, we employ AFM at relatively large tip–sample distances where forces are only on the piconewton and subpiconewton scale to prevent tip and sample distortions. Acquiring data relatively far from the surface requires low noise measurements. We probed the CaF2(111) surface with an atomically-characterized metal tip and show that the experimental data can be reproduced with an electrostatic model. By experimentally characterizing the second layer of tip atoms, we were able to reproduce the data with 99.5% accuracy. Our work links the capabilities of non-invasive imaging at large tip–sample distances and controlling the tip apex at the atomic scale.

Quantifying the evolution of atomic interaction of a complex surface with a functionalized atomic force microscopy tip

“…When comparing the images obtained with a single-atom metal tip in Ref. 39 to the images obtained with the CO tip presented here, we find in agreement to previous studies an inversion of the AFM contrast in the electrostatic imaging regime 12,27 . This contrast inversion is a result of the opposite effective tip apex polarities when imaging an ionic lattice.…”

“…Previous works incorporating electrostatic tip-sample interactions used a single negative point charge to represent the CO tip apex 24 , 27 . To quantify the success of such a simple point charge model, we compare our data to a calculation where the sample atoms are represented by point charges and the CO tip apex by a single negative point charge , where e denotes the elementary charge (see “ Methods ”) 39 .

Figure 6 ( a ) Experimental constant-height image of the (111) surface recorded with a CO-functionalized tip, processed with a 78 pm 78 pm Gaussian low-pass filter 44 .

…”

“…As shown in Ref. 39 , we calculate the agreement between experiment and calculation as , where RQD is the relative quadratic deviation over one surface period in the line profiles. For the data shown in Fig.…”

“…Graph: Plot of the calculated force contrast and responsible tip-sample interactions as a function of tip-sample distance z above (111). ( e , f ) Side and top view of the (111) surface, consisting of neutral F –Ca –F triple layers 32 , 39 . …”

“…(b) A CO-terminated tip has been modeled with a negative point charge q. Graph: Plot of the calculated force contrast F and responsible tip-sample interactions as a function of tip-sample distance z above CaF 2 (111). (e,f) Side and top view of the CaF 2 (111) surface, consisting of neutral F − -Ca 2+ -F − triple layers 32,39 . Figure 1a-c illustrates the measurement cycle: after the tip is successfully terminated with a CO molecule on the Cu(111) surface, the tip is characterized by scanning over a second CO molecule.…”