Observation of Microscale Superlubricity in Graphite

Upon shearing a microscale lithographically defined graphite mesa, the sheared section retracts spontaneously to minimize interface energy. Here, we demonstrate a sixfold symmetry of the self-retraction and provide a first experimental estimate of the frictional force involved, as direct evidence that the self-retraction is due to superlubricity, where ultralow friction occurs between incommensurate surfaces. The effect is remarkable because it occurs reproducibly under ambient conditions and over a contact area of up to $10\times 10\mu {\mathrm{m}}^2$, more than 7 orders of magnitude larger than previous scanning-probe-based studies of superlubricity in graphite. By analyzing the sheared interface, we show how the grain structure of highly oriented pyrolitic graphite determines the probability of self-retraction. Our results demonstrate that such self-retraction provides a novel probe of superlubricity, and the robustness of the phenomenon opens the way for practical applications of superlubricity in micromechanical systems.

Figure 1

(a) After mechanical exfoliation, (1) a silicon dioxide (${\mathrm{SiO}}_2$) film is grown on the graphite surface by plasma-enhanced chemical vapor deposition, and the film is coated with photoresist. (2),(3) Microscopic ${\mathrm{SiO}}_2$ squares are defined by electron beam lithography and (4) used as a mask for reactive ion etching of the squares into the HOPG, to define graphite mesas. SEM images of the top and side views of the mesas are shown. (b) Illustration of a mesa being partially sheared with a micromanipulator, to form a self-retracting flake on a graphite platform. When the microtip is raised to release the flake, it automatically returns to its original position on the mesa. (c) Observation of this process in a vacuum in a SEM. (d) Observation of the same process under ambient conditions with an optical microscope.

Figure 2

In situ manipulation of a graphite flake in a SEM. (a)–(i) are the selected frames from Supplemental Movie 1 [24]; the orientation denoted by the arrow in each picture shows the “lock-in” orientation, where the arrow is always along the same side of the flake and the dashed squares denote the location of the graphite platform. (j) Plot of the lock-in orientations translated from (a)–(i), clearly indicating the $60°$ symmetry. In this plot, the $0°$ tick mark has been aligned with the direction of the arrow in Fig. 2a, so that subsequent high-symmetry tick marks at $60°$ spacing can be compared with visual estimates of other lock-in directions. Hatched areas indicate regions between lock-in orientations where self-retraction is observed to occur reproducibly (see the SM [24]).

Figure 3

(a),(b) The deformation of a microtip when moving a graphite flake that self-retracts. (c),(d) The attempt to move one that is in a lock-in state. The images are for loading (with the tip in contact with the flake during the attempt to shear) and unloading (with the tip removed from the flake surface). The images are selected from in situ Supplemental Movies 2 and 3 [24].

Figure 4

(a) STM scan of the typical exposed surface of a graphite mesa that exhibits self-retraction, which indicates nearly atomic-scale smoothness, and (b) the same for a mesa that does not exhibit self-retraction, showing the appearance of significant interfacial defects. (c) shows no color variation in optical microscope observations of partially sheared flakes that exhibit self-retraction. The image is obtained by deliberately locking in the flake. (d) Abrupt color variations for flakes that do not self-retract, indicating large steps at the interface. (e),(f) are schematic representations of the proposed graphite microstructure for mesas that exhibit self-retraction [in (e), where the grain boundary traverses the entire mesa] and those that do not [in (f), shearing occurs at the boundary between multiple grains, resulting in large interface steps and defects].