Interlayer Friction and Superlubricity in Single-Crystalline Contact

Jul 30, 2018 - Interlayer friction between the atomic planes of 2D materials and heterostructures is a promising probe of the physics in their interla...
0 downloads 0 Views 8MB Size
Download PDF
www.acsnano.org

ACS Nano 2018.12:7638-7646. Downloaded from pubs.acs.org by UNIV OF SOUTH DAKOTA on 08/28/18. For personal use only.

Interlayer Friction and Superlubricity in Single-Crystalline Contact Enabled by TwoDimensional Flake-Wrapped Atomic Force Microscope Tips Yanmin Liu,† Aisheng Song,† Zhi Xu,‡,§ Ruilong Zong,⊥ Jie Zhang,† Wenyan Yang,† Rong Wang,† Yuanzhong Hu,† Jianbin Luo,† and TianBao Ma*,† †

State Key Laboratory of Tribology, Tsinghua University, Beijing 100084, People’s Republic of China Beijing National Laboratory for Condensed Matter Physics and Institute of Physics, Chinese Academy of Sciences, Beijing 100190, People’s Republic of China § School of Physical Sciences, University of Chinese Academy of Sciences, Beijing 100190, People’s Republic of China ⊥ National Center of Electron Spectroscopy in Beijing, Beijing 100084, People’s Republic of China ‡

S Supporting Information *

ABSTRACT: Interlayer friction between the atomic planes of 2D materials and heterostructures is a promising probe of the physics in their interlayer couplings and superlubricity. However, it is still challenging to measure the interlayer friction between well-defined 2D layers. We propose an approach of thermally assisted mechanical exfoliation and transfer to fabricate various 2D flake-wrapped atomic force microscopy (AFM) tips and to directly measure the interlayer friction between 2D flakes in single-crystalline contact. First, superlubricity between different 2D flakes and layered bulk materials is achieved with a friction coefficient as low as 10−4. The rotation angle dependence of superlubricity is observed for friction between graphite layers, whereas it is not observed between graphite and h-BN because of the incommensurate contact of the mismatched lattices. Second, the interlayer lateral force map between ReS2 layers is measured with atomic resolution, showing hexagonal patterns, as further verified by theoretical simulations. The tribological system constructed here offers an experimental platform to study interlayer couplings and friction between 2D flakes and layered bulk materials. KEYWORDS: superlubricity, thermally assisted mechanical exfoliation and transfer, 2D flakes, single-crystalline contact, interlayer coupling graphite flake onto the FFM tip, they found strong rotation angle dependence of friction force between the graphite layers. Furthermore, they observed the retrieval of a high friction state via the torque-induced reorientation and a transition from

uperlubricity1−4 is a state of vanishing friction that occurs when two crystalline surfaces slide over each other in incommensurate contact. Because of the weak interlayer interaction, graphene, carbon nanotubes, and other twodimensional (2D) materials offer potential opportunities to achieve superlubricity.5−10 Evidence for the superlubricity of graphite was provided by Dienwiebel et al.11 using a homemade frictional force microscope (FFM). By the attachment of a

S

© 2018 American Chemical Society

Received: December 23, 2017 Accepted: July 30, 2018 Published: July 30, 2018 7638

DOI: 10.1021/acsnano.7b09083 ACS Nano 2018, 12, 7638−7646

Article

Cite This: ACS Nano 2018, 12, 7638−7646

Article

ACS Nano

Figure 1. Preparation of 2D flake-wrapped AFM tips. (a) Conventional AFM tip scanning on the freshly exfoliated graphite substrate with terraces at room temperature. The black bulk represents the SiO2/Si substrate, the gray bulk represents the graphite crystal, and the blue atoms represent the graphite flake with terraces. (b) Tip fracture during the rapid heating process. (c) Wrapping of a graphite flake onto the fractured tip during the tip scanning process under high temperature. (d) Schematic drawings and SEM and TEM characterization of the fractured tip wrapped by graphite. (e) Schematic of the Raman spectroscopy measurement and the Raman spectra of both the graphite transferred on the tip and the graphite substrate. HRTEM image of an approximately 15 nm thick graphite transferred on the tip, where six graphene layers with a total thickness of 2 nm between the red dashed lines are highlighted.

surfaces.29 Despite these attempts and findings, the direct experimental measurement of the interlayer friction between 2D materials is still rarely reported. Recently, Liu et al. have obtained superlubricity between 2D materials (both graphene/graphene and graphene/h-BN) under applied normal loads by using a graphene-coated microsphere (GMS).30 The contact between the GMS and the crystalline substrate was found to be multiasperity contact of the polycrystalline graphene coating. Similarly, a series of methods have been proposed to coat graphene onto the atomic force microscopy (AFM) tip.31 However, it is very difficult to achieve a well-defined contact interface (e.g., single-crystalline contact and defect-free) between 2D materials to measure the intrinsic interlayer friction. Here, we endeavored to transfer 2D flakes onto AFM tips, motivated by the work of Dienwiebel and co-workers,11 as well as the AFM manipulation of 2D flakes.32 By using a thermally assisted mechanical exfoliation and transfer method (TAMET), a large-area graphite flake could be easily transferred to the AFM tip with fairly high reproducibility. In addition to the graphitewrapped tip, MoS2, TaS2, ReS2, and h-BN flake-wrapped tips were also prepared, all of which show superlow friction coefficients. While the interlayer friction between the graphite layers shows strong rotation angle dependence, the superlubricity between graphite and h-BN layers is achieved, regardless of the relative rotation angle. More interestingly, the interlayer lateral force map between ReS2 layers with atomic resolution shows a distinctive periodic hexagonal pattern, which is very different from the characteristic Re-chain structure when scanning by using a conventional AFM tip, coinciding with our theoretical modeling, further demonstrating the single-crystalline sliding contact geometry between the 2D layers. The tribological system used in this study offers an experimental platform to study interlayer couplings and friction between 2D flakes and layered bulk materials.

incommensurate to commensurate contact that is energetically more favorable.12 The observation of the self-retracting of graphite mesas demonstrates a microscale superlubricity, except for that in the locked orientations with hexagonal symmetry.13,14 By scanning tunneling microscopy (STM), the tip can induce rotation of the graphene nanoflake into an incommensurate registry with the graphite substrate, leading to superlubric sliding of the nanoflake, which then spontaneously rotates back to commensurate ground states, again showing the instability of incommensurate contact between graphene layers.15 In addition to graphene and graphite, molybdenum disulfide (MoS2) has also shown excellent behaviors of superlubricity.16,17 An in situ transmission electron microscopy (TEM) characterization clearly captures the shear-induced interlayer sliding of MoS2 layers in the cross-section view, showing the low shear strength between the atomic layers with interplanar van der Waals interaction.18 Recently, friction tests on MoS2 have been performed by combining the in situ scanning electron microscope (SEM) technique with a Si nanowire force sensor, and a friction coefficient of 10−4 was measured during the sliding between incommensurate MoS2 monolayers.19 Thus, the realization of sustainable incommensurate sliding contact is the main focus to achieve superlubricity. It has been theoretically proposed by density functional theory (DFT) calculations, molecular dynamics (MD) simulations, and a registry index model that heterostructures composed of 2D layers with lattice mismatch and intrinsic incommensurate interfacial geometry can help achieve robust superlubricity.20−27 Preliminary evidence for these predictions was observed in a recent Raman spectrum measurement together with a modified linear chain model, which indicates a two-orders-of-magnitude decrease in the interlayer lateral force constant compared with their homogeneous bilayers.28 In another study, multilayer WS2 grains were manipulated by a probe under SEM observation to slide smoothly on graphite and hexagonal boron nitride (h-BN) 7639

DOI: 10.1021/acsnano.7b09083 ACS Nano 2018, 12, 7638−7646

Article

ACS Nano

Figure 2. Characterizations of the 2H-MoS2-wrapped tip after the friction experiment as described below. (a) Fractured tip wrapped by MoS2. (b) HRTEM image of the enlargement of region 1# near the tip apex denoted by the red rectangle in (a); the yellow line indicates the interface between 2H-MoS2 layers and the Si tip. (c) HRTEM image of the enlargement of region 2# near the tip apex denoted by the magenta rectangle in (a). (d) HAADF-STEM-EDS line-scanning characterization of the 2H-MoS2-wrapped tip.

Figure 3. Frictional characteristics of 2D flake-wrapped tips. (a) Friction between various 2D flake-wrapped tips and the graphite substrate. (b) Friction between the graphite wrapped tip (GWT) and various layered materials substrate. The experiment is conducted with a scan size of 600 nm and a scan rate of 5 Hz for (a) and (b). (c) Rotation angle dependence of GWT scanning on graphite (red line) or h-BN/Gr (black line) with a scan size of 5 nm under a normal load of 100 nN. The narrow peaks of high friction were observed at approximately 42 ± 2° and 100 ± 2°. Between these peaks, a wide angular range with ultralow friction close to the detection limit of the instrument was found. (d) Time evolution of the friction force: the black line represents the friction between the bare fractured tip and the graphite substrate (normal load is 800 nN), and the red line stands for the friction between the fractured tip wrapped by graphite and the graphite substrate (normal load is 1230 nN for the first 25459 s and then 1914 nN until 40448 s). The experiment is conducted with a scan size of 600 nm and a scan rate of 1 Hz. The inset is the pull-off force measurement at different stages during the TAMET process. Upper panel: pull-off force curve acquired by the tip before fracture at room temperature; middle panel: pull-off force curve acquired by the freshly fractured tip after heating at approximately 200 °C; lower panel: pull-off force curve acquired by the GWT at approximately 200 °C.

7640

DOI: 10.1021/acsnano.7b09083 ACS Nano 2018, 12, 7638−7646

Article

ACS Nano

RESULTS AND DISCUSSIONS Preparation of 2D Flake-Wrapped AFM Tips. The graphite-wrapped tips are prepared by a thermally assisted mechanical exfoliation and transfer method in an environmental chamber in an AFM. A schematic illustration of the preparation process is shown in Figure 1 and described as follows. (1) A thin piece of graphite is first exfoliated from the natural graphite and then transferred onto the SiO2/Si substrate. The following TAMET processes make use of the terrace structure on the piece of graphite; that is, there exist some large graphite flakes on the graphite substrate. (2) AFM scanning is conducted by using a commercial AC240 tip under the ambient atmosphere at room temperature. We focus on a scanning area to contain a terrace edge of a graphite flake (as shown schematically in Figure 1a) for the following exfoliation and transfer process in step 4. (3) The AFM tip is lifted away 50 μm from the sample, and the temperature of the sample is raised from 25 °C to 200 °C rapidly at the rate of 1 °C/s. During the heating process, the tip crashes against the sample, leading to the fracture of the tip near the apex,33 as shown in Figure 1b (more details are presented in the Methods section). After the fracture of the tip, a plateau emerges, as shown in Figure 1d and Supporting Information Figure S1. (4) The fractured tip is then used to scan the same region containing the terrace edge as in step 2 under a normal load of 500 nN. During the scanning process, the graphite flake is rolled up from the edge and exfoliated from the graphite substrate and then is transferred and wrapped onto the fractured tip, as schematically shown in Figure 1c. The heating can drive off the water adsorbed on the surface of the tip and the sample as well as promote the adhesion of the fractured tip onto the graphite substrate.34 We have measured the tip−sample adhesion via the force curves, as shown in the inset of Figure 3d. The fractured tip shows much higher adhesive force (75.2 nN) with the sample at 200 °C compared to that before fracture (13.4 nN) at room temperature. In addition to the role of heating, the exposure of a fresh surface of the fractured tip or the possible change in the contact area could also be contributory factors. After the wrapping process, the adhesive force between the graphitewrapped tip (GWT) and the sample is markedly reduced to 8 nN at 200 °C. The exfoliation of the graphite flake from the substrate is shown in Supporting Information Figure S2. (5) After the transferring process, the sample and tip are cooled to room temperature. A series of characterizations were conducted. Figure 1d shows the SEM and TEM images of the graphite-wrapped tip. The diameter of the tip apex is around 100−300 nm with fluctuations from each experiment. The tip apex is covered with a continuous graphite film. Raman spectroscopy on the tip further verifies the existence of graphite on the tip. The absence of a D peak in Figure 1e suggests a highly ordered layered structure with no or little defects, demonstrating the advantage of this TAMET method. Furthermore, highresolution TEM (HRTEM) of the graphite flake transferred to the tip shows the obvious layered structure of graphite with an interlayer spacing of approximately 3.33 Å. The TAMET method as described above could also be used to transfer other 2D flakes on the AFM tip. One example is the 2H-MoS2-wrapped AFM tip. As shown in Figure 2, the HRTEM image of the 2H-MoS2-wrapped tip shows the characteristic layered structure with an interlayer distance of about 6.17 Å35 (Figure 2b,c) near the tip apex denoted by the red and magenta

rectangles in Figure 2a. The yellow line in Figure 2b roughly shows the interface between the MoS2 and silicon tip, indicating that the MoS2 layers are well-adhered on the AFM tip apex. Furthermore, the high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and energy dispersive X-ray spectroscopy (EDS) line scanning have been conducted to show the elemental distribution near the tip apex, as shown in Figure 2d. From the line-scan profile, the abrupt transition from silicon to molybdenum and sulfur can be observed at the tip apex in a range of about 10 nm, further confirming the wrapped MoS2 layers at the tip apex. The small amount of oxygen and carbon is likely from the oxidized layer of the silicon tip and the amorphous carbon contamination layer from the electron beam in TEM, respectively. HRTEM images of AFM tips wrapped by other 2D flakes (1T′-ReS2, 1T-TaS2, and h-BN) are shown in Figures S3−S5 in the Supporting Information. Frictional Characteristics of the 2D Flake-Wrapped Tips. All the microscale friction measurements are conducted at room temperature by AFM in the lateral force mode. The average friction force is calculated by the half-width of the lateral force loop under different applied loads, as shown in Figure 3a. Five types of 2D flakes (2H-MoS2, 1T-TaS2, 1T′-ReS2, h-BN, and graphite) were wrapped on the tip by the same TAMET method and slid against the graphite substrate (Gr). The average friction force increases linearly with normal load for all cases according to the fitting lines in Figure 3a, the slopes of which are also denoted, representing the coefficient of friction.36 The offset friction force F0 when the applied load is 0 is related to the adhesive force between the tip and the substrate materials.36 As a result, all five groups of 2D flakes show ultralow friction coefficients of less than 0.002, among which the lowest is 0.0001 for the h-BN flake-wrapped tip in sliding contact with graphite. Although the superlow friction between the graphite and h-BN has been predicted theoretically,37 this work presents the direct measurement of the interlayer friction at the heterogeneous interface between graphite and h-BN layers as well as between graphite and other 2D flakes, such as transition metal dichalcogenides (TMDs); such measurements are otherwise quite challenging without the present method. Note that a very low friction coefficient of 0.0009 is obtained for the 1T′-ReS2 flake-wrapped tip sliding against the graphite substrate; this friction coefficient is much lower than those of 2H-MoS2/Gr and 1T-TaS2/Gr, exhibiting its best lubricating behavior among the TMDs. It has been reported recently that both the contact size and the interfacial interaction play an important role in determining superlubricity.7 In order to illustrate the role of interfacial interaction in the present experiment and eliminate the possible effects of the tip geometry and contact area on the interfacial friction trends shown in Figure 3a, a separate group of experiments were conducted by using the same graphitewrapped tip to slide against various layered materials (2H-MoS2, 1T-TaS2, 1T′-ReS2, graphite, and h-BN). As shown in Figure 3b, the friction coefficient between different pairs of 2D layered materials exhibits similar trends to that in Figure 3a. Specifically, the lowest friction coefficient of 0.0001 is obtained between graphite and h-BN, and the friction coefficient between graphite and 1T′-ReS2 (0.0008) is still the lowest when compared to 2HMoS2 and 1T-TaS2. Thus, the measured microscale friction behaviors are largely dependent on the interfacial interactions between the 2D flakes in well-defined contact geometry. 7641

DOI: 10.1021/acsnano.7b09083 ACS Nano 2018, 12, 7638−7646

Article

ACS Nano

Figure 4. Experimental and simulation results of the 1T′-ReS2/1T′-ReS2 interlayer frictional characteristics. The experiment is conducted with a scan size of 10 nm, a scan rate of 10 Hz, and a normal load of 230 nN. (a) Filtered lateral force image by using 1T′-ReS2-wrapped tip scanning on the 1T′-ReS2 crystal. (b) Enlargement of the region denoted by the yellow square in (a). (c) Simulated interlayer lateral force map between 1T′-ReS2 layers. (d) Corresponding atomic structure of the 1T′-ReS2/1T′-ReS2 interface used in the simulation to obtain (c), where sulfur atoms are not shown to emphasize the orientation of the Re-chain structure. The blue atoms denote the Re atoms in the 1T′-ReS2 layer wrapped on the tip, the blue arrow denotes the direction of Re-chains, the red atoms denote the Re atoms in the 1T′-ReS2 layer on the substrate, and the red dashed arrow denotes the direction of Re-chains. θ is the angle between the Re-chain and the tip fast scanning direction. θup = 140° is the angle between the Re-chain on the tip and the scanning direction; θlow = 150° is the angle between the Re-chain on the substrate and the scanning direction. (e) Friction loops along the red lines in (a) and (c). The upper panel is the experimental result, and the lower panel is the simulation result. (f, g, and h) Similar to (b), (c), and (d), except the 1T′-ReS2 layer on the substrate is rotated 19° counterclockwise, resulting in a different relative orientation between the layers θup/θlow = 140°/169° and different lateral force patterns.

Figure 5. Experimental and simulation results of commercial silicon tip/1T′-ReS2 frictional characteristics. The experiment is conducted with a scan size of 10 nm, a scan rate of 10 Hz, and a normal load of 200 nN. (a) Filtered lateral force image by tip scanning on the 1T′-ReS2 crystal. (b) Enlargement of the region denoted by the yellow square in (a). (c) Simulated lateral force map. (d) Corresponding atomic structure of the 1T′ReS2 crystal used in the simulation to obtain (c), where sulfur atoms are not shown to emphasize the orientation of the Re-chain structure. The red atoms denote the Re atoms in the 1T′-ReS2 layer on the substrate, and the red solid arrow denotes the direction of Re-chains. θ = 120° is the angle between the Re-chains and the tip fast scanning direction. (e) Friction loops along the red lines in (a) and (c). The upper panel is the experimental result, and the lower panel is the simulation result. (f, g, and h) Similar to (b), (c), and (d), except the 1T′-ReS2 layer on the substrate is rotated 72° clockwise, resulting in a different relative orientation between the tip and 1T′-ReS2 crystal (θ = 48°) and different lateral force patterns.

graphite-wrapped tip and the substrate and record the average friction with different rotation angles between graphite and hBN, as shown in Figure 3c (the rotation angle is a relative value

To study the effect of the relative rotation angle on the superlubricity of 2D heterostructures, we rotate the h-BN substrate to change the relative orientation between the 7642

DOI: 10.1021/acsnano.7b09083 ACS Nano 2018, 12, 7638−7646

Article

ACS Nano

with the sliding potential energy surface calculated by empirical potentials (simulation details are presented in the Methods section) to unravel the origin of the lateral force pattern and its dependence on the Re-chain orientation in both the upper and lower 1T′-ReS2 layers. For ease of discussion, we define the angle between the tip fast scanning direction and the Re-chain direction in the 1T′-ReS2 crystal as θ (Figure 4d and h), where θup denotes the angle between the scanning direction and the Re-chain direction in the 1T′-ReS2-wrapped tip (upper), and θlow denotes the angle between the scanning direction and the Re-chain direction in the 1T′-ReS2 crystal on the substrate (lower). In the simulations, we change θup and θlow and find that both angles can affect the lateral force pattern. The experimental lateral force image in Figure 4a can be well reproduced by simulation when θup = 140° and θlow = 150°. The similarity between the two can be viewed more clearly by a locally enlarged experimental lateral force image (Figure 4b) and the corresponding simulation result (Figure 4c). The black circles in Figure 4b and c highlight the maximum lateral force in the map, while the red triangles, green blocks, and blue diamonds indicate the second, third, and fourth largest lateral force, respectively. The pink inverted triangles show the valley of the map. Figure 4e shows the friction loops along the red line in Figure 4a and c, where the upper panel is the experimental result and the lower panel is the simulation result. The results agree well in detail (e.g., the number of stick−slip events in a period). The atomistic structure of the tribo-pair is shown in Figure 4d, where the sulfur atoms are not displayed for clarity. The blue atoms represent the Re atoms in the 1T′-ReS2 layer wrapped on the tip, the blue arrow is the direction of Re-chains, the red atoms represent the Re atoms in the 1T′-ReS2 layer on the substrate, and the red dashed arrow indicates the direction of its Re-chains. Both black arrows represent the fast scanning direction of the AFM tip. When the 1T′-ReS2 substrate is rotated by 19° counterclockwise experimentally, the interlayer lateral force pattern changes, as shown in Figure 4f. Correspondingly, the lower 1T′ReS2 layer is also rotated in the simulation by the same angle; when θlow becomes 169°, the lateral force pattern also changes, as shown in Figure 4g, in agreement with the experimental result. The corresponding interlayer atomistic structure is shown in Figure 4h. Other examples can be found in Supporting Information Figure S3. These results further confirm the assumption of the experimentally achieved interlayer friction between 1T′-ReS2 layers. In addition, the load dependence of the lateral force pattern between ReS2 layers has been studied and is shown in Figure S8 in the Supporting Information. For comparison, the lateral force images by using commercial silicon AFM tip sliding against the 1T′-ReS2 crystal show chainlike patterns in Figure 5 and Supporting Information Figure S7, in contrast to those in Figure 4. We assume that the chain-like pattern reflects the atomic structure of the 1T′-ReS2 layer on the substrate. This assumption is further verified by comparing both the experimental and simulation results, as shown in Figure 5b and c. As shown in Figure 5e, both friction loops obtained experimentally and theoretically along the red lines in Figure 5a and c show less detailed variation in a period than in Figure 4e. The stick−slip behavior can be simply attributed to the atomic structure and morphological corrugation of the 1T′-ReS2 layer on the substrate, as shown in Figure 5d. Similar results can be obtained when the 1T′-ReS2 layer on the substrate is rotated by 72° clockwise, and the angle between the Re-chain and the

where the absolute value does not have a physical meaning). The average friction is nearly independent of the rotation angle for the graphite-wrapped tip sliding on h-BN, as shown by the black line. For comparison, we also studied the graphite-wrapped tip sliding on a graphite substrate, clearly observing two narrow peaks of high friction at approximately 42 ± 2° and 100 ± 2°, with the peak values of 568 ± 55 pN and 572 ± 38 pN, respectively. This 60 degree rotational symmetry of the frictional properties is typical for graphite, representing commensurate contacts between the layers at these specific rotation angles. Meanwhile, for other rotation angles, superlow friction can be obtained, representing the incommensurate contacts. This observation is consistent with a previous experiment by Dienwiebel et al.11 and recent theoretical predictions.22 The intrinsic incommensurability between graphite and h-BN explains the relative rotation angle independence of superlubricity.37 Furthermore, the GWT shows a much longer lifetime compared with the bare fractured tip. For the bare fractured tip, friction already starts to increase after 190 s, and the graphite substrate wears out after 250 s. On the other hand, for the GWT, ultralow friction is preserved for a relatively long period: first with a normal load of 1230 nN for 25459 s and then with an increased load of 1914 nN without breaking the experiment for another 14989 s. The friction force still remains basically unchanged (the change at 25459 s is due to the increase of normal load), showing no sign of wear or breakdown of superlubricity. In addition, in the present experiment, the torque-induced reorientation is not observed, demonstrating the strong adhesion between the tip and the graphite flake wrapped on it, preventing the graphite flake from rotating toward the commensurate contact with the graphite substrate during the sliding process. This strong adhesion originates from the tip fracture and sample heating during the TAMET process, as described previously in step 4; thus, the graphite flake can be firmly attached to the AFM tip with much longer duration. ReS2/ReS2 Interlayer Frictional Characteristics. In another set of experiments, we measure the atomic-scale friction by using the 1T′-ReS2-wrapped tip to slide against another piece of 1T′-ReS2 crystal exfoliated and transferred on the silicon wafer. Thus, both the upper and lower sliding materials are 1T′ReS2 layers. Surprisingly, the atomic-resolution lateral force maps, as shown in Figure 4 and Supporting Information Figure S6, show different patterns from those by using a commercial silicon tip (AC240), as shown in Figure 5 and Supporting Information Figure S7. While the lateral force maps show obvious chain-like patterns for conventional tip scanning on 1T′-ReS2 in Figure 5a (the distance between the chains is b = 0.56 nm), the lateral force maps show more hexagonal-like patterns for 1T′-ReS2/1T′-ReS2 (with a side length of a = 0.65 nm, and the vertices are denoted by the black circles, corresponding to the largest lateral force during sliding) in Figure 4a. The chain-like lateral force pattern reflects the characteristic Re-chain structure in 1T′-ReS2 on the substrate, where the distance across the Re-chain is 0.56 nm. However, by using the 1T′-ReS2-wrapped tip, the lateral force pattern reflects the sliding behavior between the 1T′-ReS2 layers, which results in the more complex pattern than the simply chain-like pattern. The relation between the side length of the hexagon a and interchain distance b is simply b = a × sin(π/3). We assume that the lateral force images in Figure 4 originate from the interlayer sliding between the 1T′-ReS2 layers. To verify this point, we conducted simulations based on the Prandtl−Tomlinson model 7643

DOI: 10.1021/acsnano.7b09083 ACS Nano 2018, 12, 7638−7646

Article

ACS Nano scanning direction changes to θ = 48°, as shown in Figure 5h. In order for a comparison with the 1T′-ReS2-wrapped tip, the bare fractured tip instead of the sharp tip was also used to measure the lateral force image (Figure S9 in the Supporting Information), which shows similar patterns to that in Figure 5. The obvious contrast between the interlayer friction and tip− sample friction here is partially attributed to the characteristic Re-chain atomic structure on the ReS2 surface. Generally, for interlayer sliding between 2D flakes, the lateral force pattern should be a modulation of the two crystalline structures. However, more experimental and theoretical works are required to further study this modulation effect. The measurement of the atomic-resolution interlayer lateral force pattern here provides direct proof that this TAMET method can be utilized to accomplish a single-crystalline sliding contact at the nanoscale, thereby enabling one to measure the interlayer couplings and friction of 2D flakes or 2D heterostructures to unravel interficial physical phenomena.

the half-width of the lateral force loop averaged from three measurements and each for eight scan lines. Characterization. The Raman spectra of the GWT were collected on a Horiba HR800 instrument with a laser wavelength of 514 nm and 50× objective lens under ambient conditions; the laser spot size was 1 μm. The morphology of the GWT was observed by SEM (Hitachi SU8220). HRTEM was conducted (JEOL 2010F) with an accelerating voltage of 200 kV to study the lamellar structure of the attached graphite flake on the AFM tip. HRTEM and HAADF-STEM-EDS were conducted (JEOL 2100F) with an accelerating voltage of 200 kV to study the lamellar structure and elemental analysis of the MoS2 flake wrapped on the AFM tip. Simulation Details. The interlayer friction model was built by an upper 1T′-ReS2 nanoflake (6.4 Å × 5.6 Å) sliding against the lower 1T′ReS2 substrate (114.8 Å × 101.7 Å) in a direction consistent with the tip fast scanning direction in the experiments. The crystalline orientations of both the 1T′-ReS2 nanoflake and the substrate were varied to study the effects of the relative angles (θup and θlow) between the Re-chains and the scanning direction on the interlayer friction, as described in the main text. The x−y scanning size of the nanoflake was 38.3 Å × 33.9 Å to ensure that the edge effect of the substrate can be neglected. However, the choice of the size of the upper nanoflake is a balance between computational cost and accuracy. In contrast, the bare tip scanning against the 1T′-ReS2 crystal was modeled by a single Ar atom sliding against the 1T′-ReS2 crystalline substrate in a direction consistent with the tip fast scanning direction in the experiments. An Ar atom, rather than an actual tip with finite size, was utilized to qualitatively represent the potential energy variation felt by the tip during scanning, which should be physically meaningful, as demonstrated in previous studies.40−43 Moreover, the same scanning size was used as the interlayer sliding model. Note that in the present simulation the potential energy surface (PES) during sliding was calculated by Lennard-Jones potential (the potential parameters are listed in Table 1) instead of DFT calculations because of the extremely

CONCLUSIONS In this paper, a thermally assisted mechanically exfoliation and transfer method was proposed to transfer various 2D flakes onto the AFM tip, through which the interlayer friction can be achieved between single-crystalline 2D flakes at the contact area. Superlubricity between graphite layers with angular dependence was observed, and superlubricity between graphite flake and the bulk h-BN substrate was obtained without any angular dependence. Moreover, the wrapped graphite flake shows good adhesion with the tip; thus, the superlubricity is preserved over a long test period, without the observation of torqueinduced reorientation. More importantly, the interlayer lateral force map between 1T′-ReS2 layers was measured with atomic resolution. The periodic hexagonal patterns of the lateral force map are very different from the chain-like patterns obtained by using a conventional AFM tip, coinciding with our theoretical modeling, further demonstrating the single-crystalline sliding contact geometry between the 2D flakes. The tribological system used in this study offers an experimental platform to study the interlayer couplings and friction between 2D flakes and layered bulk materials.

Table 1. L-J Potential Parameters σ/Å ε/eV

S−S44

Re−S45

Re−Re46

S−Ar45

Re−Ar45

3.535 0.011

3.138 0.012

2.740 0.013

3.470 0.011

3.073 0.012

expensive computational cost of the latter. The scanning Ar atom and both the 1T′-ReS2 nanoflake and the 1T′-ReS2 substrate were considered as rigid bodies to simplify the calculations. The potential energy surface (Uinter) for the interlayer sliding at the constant load of 2 nN between the 1T′-ReS2 nanoflake and the 1T′ReS2 substrate when scanning in both x and y directions can be expressed by eq 1: ÅÄÅ ÑÉ ÅÅij σ yz6 ij σ yz12 ÑÑÑ Åjj S − S zz Ñ − S S j z Å Uinter = 4∑ ∑ εS − SÅÅÅjj zz − jjj zzz ÑÑÑÑ j z j z Å Ñ r r Å nj nj n j ÅÅÇk { k { ÑÑÑÖ ÄÅ É 6 12 Ñ ÅÅÅji σ y i σRe − S zy ÑÑÑÑ z j Å − Re S zz − jjj zz Ñ + 4 ∑ ∑ εRe − SÅÅÅjjj j rin zz ÑÑÑÑ ÅÅj rin zz n i { k { ÑÑÖ ÅÅÇk É ÅÄÅ Ñ 6 12 ÅÅij σ yz ij σ yz ÑÑÑÑ Å + 4 ∑ ∑ εRe − SÅÅÅÅjjjj Re − S zzzz − jjjj Re − S zzzz ÑÑÑÑ j rmj z ÑÑ ÅÅÅjk rmj z{ m j k { ÑÑÖ ÅÇ ÄÅ É 6 12 Ñ ÅÅÅji σ y i σRe − Re zy ÑÑÑÑ z j Å − Re Re zz − jjj zz Ñ + 4 ∑ ∑ εRe − ReÅÅÅjjj j rmi zz ÑÑÑÑ ÅÅj rmi zz m i { k { ÑÑÖ ÅÅÇk (1)

METHODS Preparation of 2D Flake-Wrapped Tip. The TAMET process was conducted in a Cypher ES environmental AFM (Asylum Research). During the rapid heating of the sample from 25 to 200 °C, the uniformly painted silver paint (5−6 mg, Leltsllber 200) on the bottom side of the sample for fixture could expand with the generation of many gas bubbles. As a result, the sample was raised by 50−60 μm, so that the AFM tip crashed against the sample, leading to the fracture of the tip near the apex, as shown in Figure 1b. Various 2D flakes (bulk hBN, natural graphite, and 2H-MoS2 from XFNANO; other TMDs from HQ graphene) were wrapped onto the tip by TAMET. Friction Force Microscopy Tests. The friction tests were conducted by Cypher (Asylum Research) AFM in the lateral force mode under ambient atmosphere, at a temperature of 23 ± 2 °C. The relative humidity is 56 ± 4% for the friction coefficient measurement and wear test in Figure 3d. The normal spring constant of the tip at room temperature was 2.56 ± 0.82 N m−1, as determined by the thermal noise method.38 The normal spring constants of the freshly fractured tip and graphite-wrapped fractured tip at 200 °C are 1.67 and 1.52 N m−1, respectively, for the measurement of pull-off force curves in the inset of Figure 3d. The lateral sensitivity of the tip was 1.85 ± 1.0 μN V−1 by the grating method.39 The friction force under each load was calculated by

i and j are the number of Re atoms and S atoms in the substrate material, respectively. m and n are the number of Re atoms and S atoms in the layer wrapped on the tip, respectively. When the Ar atom slid against the 1T′-ReS2 crystal, the PES Utip was acquired by the L-J potential formula at the constant load of 2 nN. Utip can be expressed by eq 2: 7644

DOI: 10.1021/acsnano.7b09083 ACS Nano 2018, 12, 7638−7646

ÄÅ É 6 12 Ñ ÅÅi ij σ yz ÑÑÑÑ ÅÅjj σS − Ar yzz − S Ar j z zz − jj z Ñ Utip = 4 ∑ εS − ArÅÅÅÅjjj jj r zzz ÑÑÑÑ ÅÅj rj zz j ÅÅÇk { k j { ÑÑÑÖ ÄÅ É 6 12 Ñ ÅÅi i σRe − Ar zy ÑÑÑÑ Åjj σRe − Ar zyz j Å zz Ñ zz − jjj + 4 ∑ εRe − Ar ÅÅÅjj j ri zz ÑÑÑÑ ÅÅÅjk ri z{ i k { ÑÑÖ ÅÇ

Article

ACS Nano

wrapped AFM tip. Computations were conducted on the “Explorer 100” cluster system of the Tsinghua National Laboratory for Information Science and Technology.

REFERENCES (2)

(1) Hirano, M.; Shinjo, K. Atomistic Locking and Friction. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 41, 11837−11851. (2) Kim, D. E.; Suh, N. P. On Microscopic Mechanisms of Friction and Wear. Wear 1991, 149, 199−208. (3) Meyer, E.; Gnecco, E. Superlubricity on The Nanometer Scale. Friction 2014, 2, 106−113. (4) Guo, W.; Yin, J.; Qiu, H.; Guo, Y.; Wu, H.; Xue, M. Friction of Low-Dimensional Nanomaterial Systems. Friction 2014, 2, 209−225. (5) Zhang, R.; Ning, Z.; Zhang, Y.; Zheng, Q.; Chen, Q.; Xie, H.; Zhang, Q.; Qian, W.; Wei, F. Superlubricity in Centimetres-Long Double-Walled Carbon Nanotubes under Ambient Conditions. Nat. Nanotechnol. 2013, 8, 912−916. (6) Berman, D.; Deshmukh, S. A.; Sankaranarayanan, S. K. R. S.; Erdemir, A.; Sumant, A. V. Macroscale Superlubricity Enabled by Graphene Nanoscroll Formation. Science 2015, 348, 1118−1122. (7) Dietzel, D.; Brndiar, J.; Š tich, I.; Schirmeisen, A. Limitations of Structural Superlubricity: Chemical Bonds versus Contact Size. ACS Nano 2017, 11, 7642−7647. (8) Berman, D.; Erdemir, A.; Sumant, A. V. Approaches for Achieving Superlubricity in Two-Dimensional Materials. ACS Nano 2018, 12, 2122−2137. (9) Klemenz, A.; Gola, A.; Moseler, M.; Pastewka, L. Contact Mechanics of Graphene-Covered Metal Surfaces. Appl. Phys. Lett. 2018, 112, 061601. (10) Xu, Q.; Li, X.; Zhang, J.; Hu, Y. Z.; Wang, H.; Ma, T. B. Suppressing Nanoscale Wear by Graphene/Graphene Interfacial Contact Architecture: A Molecular Dynamics Study. ACS Appl. Mater. Interfaces 2017, 9, 40959−40968. (11) Dienwiebel, M.; Verhoeven, G. S.; Pradeep, N.; Frenken, J. W. M.; Heimberg, J. A.; Zandbergen, H. W. Superlubricity of Graphite. Phys. Rev. Lett. 2004, 92, 126101. (12) Filippov, A. E.; Dienwiebel, M.; Frenken, J. W. M.; Klafter, J.; Urbakh, M. Torque and Twist against Superlubricity. Phys. Rev. Lett. 2008, 100, 046102. (13) Vu, C. C.; Zhang, S.; Urbakh, M.; Li, Q.; He, Q. C.; Zheng, Q. Observation of Normal-Force-Independent Superlubricity in Mesoscopic Graphite Contacts. Phys. Rev. B: Condens. Matter Mater. Phys. 2016, 94, 081405. (14) Zheng, Q.; Liu, Z. Experimental Advances in Superlubricity. Friction 2014, 2, 182−192. (15) Feng, X.; Kwon, S.; Park, J. Y.; Salmeron, M. Superlubric Sliding of Graphene Nanoflakes on Graphene. ACS Nano 2013, 7, 1718−1724. (16) Martin, J. M.; Donnet, C.; Le Mogne, T.; Epicier, T. Superlubricity of Molybdenum Disulphide. Phys. Rev. B: Condens. Matter Mater. Phys. 1993, 48, 10583−10586. (17) Miura, K.; Kamiya, S. Observation of The Amontons-Coulomb Law on The Nanoscale: Frictional Forces between MoS2 Flakes and MoS2 Surfaces. Europhys. Lett. 2002, 58, 610−615. (18) Oviedo, J. P.; KC, S.; Lu, N.; Wang, J.; Cho, K.; Wallace, R. M.; Kim, M. J. In Situ TEM Characterization of Shear-Stress-Induced Interlayer Sliding in The Cross Section View of Molybdenum Disulfide. ACS Nano 2015, 9, 1543−1551. (19) Li, H.; Wang, J.; Gao, S.; Chen, Q.; Peng, L.; Liu, K.; Wei, X. 2D Materials: Superlubricity between MoS2 Monolayers. Adv. Mater. 2017, 29, 1701474. (20) Hod, O. The Registry Index: A Quantitative Measure of Materials’ Interfacial Commensurability. ChemPhysChem 2013, 14, 2376−2391. (21) Wang, L. F.; Ma, T. B.; Hu, Y. Z.; Zheng, Q.; Wang, H.; Luo, J. Superlubricity of Two-Dimensional Fluorographene/MoS2 Heterostructure: A First-Principles Study. Nanotechnology 2014, 25, 385701. (22) Ansari, N.; Nazari, F.; Illas, F. Role of Structural Symmetry Breaking in The Structurally Induced Robust Superlubricity of

i and j are the number of the Re atoms and S atoms in the substrate material, respectively. According to the PT model, the motions of the AFM tip along the x and y directions are expressed in eqs 3a and 3b, respectively: m

m

∂ 2x ∂x ∂U + mμ + = Kx(x0 − x) ∂t ∂x ∂t 2 ∂ 2y ∂t 2

+ mμ

∂y ∂U + = K y(y0 − y) ∂t ∂y

(3a)

(3b)

where U was calculated by eqs 1 and 2 for the 1T′-ReS2 flake and the Ar atom sliding, respectively; m is the mass of the tip wrapped by the 1T′ReS2 crystal or the bare tip (m = 1 × 10−12 kg); (x, y) is the position of the tip; and (x0, y0) is the position of the driver in the x, y direction. Kx and Ky are the stiffness of the system in the x and y directions, respectively (Kx = Ky = 1.25 N/m). μ is the damping coefficient (μ = 2 k /m ). The fourth-order Runge−Kutta method was adopted to solve eqs 3a and 3b, and the lateral force was obtained by eq 4: f = Kx(x0 − x)

(4)

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.7b09083. Figures S1−S9 as described in the text (PDF)

AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. ORCID

Jianbin Luo: 0000-0002-5132-0712 TianBao Ma: 0000-0001-8016-9241 Author Contributions

T.-B.M. and J.L. proposed and supervised the project; Y.L. and T.-B.M. designed the experiments; Y.L. prepared the 2Dmaterial-wrapped AFM tip and conducted the frictional measurement; A.S. and Y.-Z.H. performed the theoretical simulation; Z.X. and R.Z. prepared the TEM sample holder for the AFM tip and performed the TEM and HRTEM characterization; W.Y. and R.W. performed the SEM characterization, and Y.L., A.S., J.Z., and T.-B.M. analyzed the data, interpreted the results, and wrote the text. All the authors participated in discussions of the research. Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS The authors would like to acknowledge the support of the National Natural Science Foundation of China (Grant Nos. 51527901, 51422504, 51375010) and the Austrian-Chinese Cooperative R&D Projects, FFG and CAS (No. 112111KYSB20150002). The authors also thank Chao Ma in the National Center of Electron Spectroscopy in Beijing (NCESBJ) for conducting characterizations of the 2D flake7645

DOI: 10.1021/acsnano.7b09083 ACS Nano 2018, 12, 7638−7646

Article

ACS Nano Graphene and H-BN Homo- and Hetero- Junctions. Carbon 2016, 96, 911−918. (23) Mandelli, D.; Leven, I.; Hod, O.; Urbakh, M. Sliding Friction of Graphene/Hexagonal−Boron Nitride Heterojunctions: A Route to Robust Superlubricity. Sci. Rep. 2017, 7, 10851. (24) Guerra, R.; van Wijk, M.; Vanossi, A.; Fasolino, A.; Tosatti, E. Graphene on H-BN: to Align or Not to Align? Nanoscale 2017, 9, 8799. (25) Guo, Y.; Qiu, J.; Guo, W. Reduction of Interfacial Friction in Commensurate Graphene/H-BN Heterostructures by Surface Functionalization. Nanoscale 2016, 8, 575−580. (26) Ouyang, W.; Ma, M.; Zheng, Q.; Urbakh, M. Frictional Properties of Nanojunctions Including Atomically Thin Sheets. Nano Lett. 2016, 16, 1878−1883. (27) Lebedev, A. V.; Lebedeva, I. V.; Popov, A. M.; Knizhnik, A. A. Stacking in Incommensurate Graphene/Hexagonal-Boron-Nitride Heterostructures Based on Ab Initio Study of Interlayer Interaction. Phys. Rev. B: Condens. Matter Mater. Phys. 2017, 96, 085432. (28) Wang, L.; Zhou, X.; Ma, T.; Liu, D.; Gao, L.; Li, X.; Zhang, J.; Hu, Y.; Wang, H.; Dai, Y.; Luo, J. Superlubricity of A Graphene/MoS2 Heterostructure: A Combined Experimental and DFT Study. Nanoscale 2017, 9, 10846−10853. (29) Kobayashi, Y.; Taniguchi, T.; Watanabe, K.; Maniwa, Y.; Miyata, Y. Slidable Atomic Layers in Van Der Waals Heterostructures. Appl. Phys. Express 2017, 10, 045201. (30) Liu, S. W.; Wang, H. P.; Xu, Q.; Ma, T. B.; Yu, G.; Zhang, C.; Geng, D.; Yu, Z.; Zhang, S.; Wang, W.; Hu, Y. Z.; Wang, H.; Luo, J. Robust Microscale Superlubricity under High Contact Pressure Enabled by Graphene-Coated Microsphere. Nat. Commun. 2017, 8, 14029. (31) Hui, F.; Chen, S.; Liang, X.; Yuan, B.; Jing, X.; Shi, Y.; Lanza, M. Graphene Coated Nanoprobes: A Review. Crystals 2017, 7, 269. (32) Island, J. O.; Steele, G. A.; can der Zant, H. S. J.; CastellanosGomez, A. Mechanical Manipulation and Exfoliation of Boron Nitride Flakes by Micro-Plowing with An AFM Tip. arXiv preprint arXiv 2015, 1501.06437. (33) Bhaskaran, H.; Gotsmann, B.; Sebastian, A.; Drechsler, U.; Lantz, M. A.; Despont, M.; Jaroenapibal, P.; Carpick, R. W.; Chen, Y.; Sridharan, K. Ultralow Nanoscale Wear through Atom-by-Atom Attrition in Silicon-Containing Diamond-Like Carbon. Nat. Nanotechnol. 2010, 5, 181−185. (34) Dean, C. R.; Young, A. F.; Meric, I.; Lee, C.; Wang, L.; Sorgenfrei, S.; Watanabe, K.; Taniguchi, T.; Kim, P.; Shepard, K. L.; Hone, J. Boron Nitride Substrates for High-Quality Graphene Electronics. Nat. Nanotechnol. 2010, 5, 722−726. (35) Hwang, H.; Kim, H.; Cho, J. MoS2 Nanoplates Consisting of Disordered Graphene-Like Layers for High Rate Lithium Battery Anode Materials. Nano Lett. 2011, 11, 4826−4830. (36) Ogletree, D. F.; Carpick, R. W.; Salmeron, M. Calibration of Frictional Forces in Atomic Force Microscopy. Rev. Sci. Instrum. 1996, 67, 3298−3306. (37) Leven, I.; Krepel, D.; Shemesh, O.; Hod, O. Robust Superlubricity in Graphene/H-BN Heterojunctions. J. Phys. Chem. Lett. 2013, 4, 115−120. (38) Butt, H.-J.; Jaschke, M. Calculation of Thermal Noise in Atomic Force Microscopy. Nanotechnology 1995, 6, 1−7. (39) Varenberg, M.; Etsion, I.; Halperin, G. An Improved Wedge Calibration Method for Lateral Force in Atomic Force Microscopy. Rev. Sci. Instrum. 2003, 74, 3362−3367. (40) Steiner, P.; Gnecco, E.; Krok, F.; Budzioch, J.; Walczak, L.; Konior, J.; Szymonski, M.; Meyer, E. Atomic-Scale Friction on Stepped Surfaces of Ionic Crystals. Phys. Rev. Lett. 2011, 106, 186104. (41) Holscher, H.; Ebeling, D.; Schwarz, U. D. Friction at AtomicScale Surface Steps: Experiment and Theory. Phys. Rev. Lett. 2008, 101, 246105. (42) Zhong, W.; Tomanek, D. First-Principles Theory of AtomicScale Friction. Phys. Rev. Lett. 1990, 64, 3054−3057. (43) Shi, R.; Gao, L.; Lu, H.; Li, Q.; Ma, T. B.; Guo, H.; Du, S.; Feng, X. Q.; Zhang, S.; Liu, Y.; Cheng, P.; Hu, Y. Z.; Gao, H. J.; Luo, J. Moiré Superlattice-Level Stick-Slip Instability Originated from Geometrically

Corrugated Graphene on A Strongly Interacting Substrate. 2D Mater. 2017, 4, 025079. (44) Meloni, S.; Pieretti, A.; Bencivenni, L.; Albertini, V. R.; Sadun, C.; Caminiti, R. SO2Cl2, SOCl2: Energy Dispersive X-ray Diffraction, Ab Initio and Molecular Dynamics Calculation. Comput. Mater. Sci. 2001, 20, 407−415. (45) Neek-Amal, M.; Asgari, R.; Tabar, M. R. R. The Formation of Atomic Nanoclusters on Graphene Sheets. Nanotechnology 2009, 20, 135602. (46) Mokhalingam, A.; Kumar, D.; Srivastava, A. Mechanical Behaviour of Graphene Reinforced Aluminum Nano composites. Mater. Today: Proceedings 2017, 4, 3952−3958.

7646

DOI: 10.1021/acsnano.7b09083 ACS Nano 2018, 12, 7638−7646