Ultrahigh strength and shear-assisted separation of sliding nanocontacts studied in situ

Abstract

The behavior of materials in sliding contact is challenging to determine since the interface is normally hidden from view. Using a custom microfabricated device, we conduct in situ, ultrahigh vacuum transmission electron microscope measurements of crystalline silver nanocontacts under combined tension and shear, permitting simultaneous observation of contact forces and contact width. While silver classically exhibits substantial sliding-induced plastic junction growth, the nanocontacts exhibit only limited plastic deformation despite high applied stresses. This difference arises from the nanocontacts' high strength, as we find the von Mises stresses at yield points approach the ideal strength of silver. We attribute this to the nanocontacts' nearly defect-free nature and small size. The contacts also separate unstably, with pull-off forces well below classical predictions for rupture under pure tension. This strongly indicates that shearing reduces nanoscale pull-off forces, predicted theoretically at the continuum level, but not directly observed before.

Fig. 1. A custom-designed in situ apparatus enables observation of nanoscale single asperity friction^,.

a Schematic of the stainless steel frame that holds the NEMS device, which is mounted at the sample location of a TEM holder. Four gold wires, used to drive the two electrostatic NEMS actuators, are shown along with a hole for passage of the TEM beam. b Schematic of the silicon-based NEMS device, showing two orthogonally-oriented cantilevers to measure friction and normal forces, and electrostatic actuators to move asperities in lateral and vertical directions. The inset shows the TEM view of the contact. The upper asperity is the one connected to the cantilever for measuring the friction force. c Example of a single asperity sliding experiment observed by TEM. i, The upper asperity is actuated in the lateral direction. Initially, the asperities are not in contact. ii, The lower asperity has been pulled into tensile contact with the upper asperity due to attractive forces. iii, iv, v, The upper asperity slides laterally across lower asperity. vi, The junction separates. Videos of this experiment are available (See Supplementary Movie).

Fig. 2. TEM images demonstrate that only limited nanoscale plastic deformation occurred due to contact, sliding, and separation.

For the experiment depicted in Fig. 1, the shape before the contact a was compared with the shape after the separation b, c depicts the difference between a, b. The lines shown are traced manually while magnifying the image. Despite the high stresses, the plastic displacements are less than 1.0 nm in size.

Fig. 3. Forces (from the NEMS device), contact width (from TEM images), and resulting calculated stresses as a function of sliding distance.

a Friction and load forces during an asperity friction measurement. The friction force acts parallel to the direction of the actuation, and the load force acts perpendicular to the friction force, as shown in Fig. 1.b In contrast, the shear force acts parallel to the plane of contact, whose orientation changes during sliding. Similarly, the normal force acts perpendicular to the plane of contact, i.e., perpendicular to the shear force (Fig. 1c, iv). c the contact width, measured as the shortest width of the junction. d The tensile normal stress, shear stress, and von Mises stress, derived from the values of the normal force, the shear force, and the contact width as shown in Eqs. (1)–(4). The indices i-vi corresponds to the panels in Fig. 1. A video of this experiment is available (See Supplementary Movie). The error bars representing the uncertainty of each experimental value arose from the resolution of the TEM and NEMS actuator, and the calculations are performed as described in Supplementary Discussion 2.

Fig. 4. The experimental values of the von Mises stress at slip instabilities far exceed those of bulk silver.

They reached 24–59% of silver’s theoretical strength according to DFT calculations, rivaling values observed for single crystal or pentatwinned nanowires in tension. Triangles: values of the von Mises stress measured at observed yield points. Dashed lines and circles show literature values discussed in the text. The error bars representing the uncertainty of each experimental value arose from the resolution of the TEM and NEMS actuator, and the calculations are performed as described in Supplementary Discussion 2.

Fig. 5. The experimental pull-off forces during sliding are well below predictions from adhesive contact mechanics models for separation under pure tension, indicating that shear stresses strongly assist the separation process.

Blue squares: experiments. Red circles: predicted pull-off force from the Maugis-Dugdale model using values of Tabor’s parameter µT determined for each experiment. Solid line: predicted pull-off force from the DMT model. Dashed line: predicted pull-off force from the JKR model (see Supplemental Figs Table 2). A value of W = 2.0 J/m² is used for all calculations, as discussed in the text. The experimental pull-off forces are generally well below all of these predicted values, indicating that the applied shear force is playing a role in promoting the separation process. The error bars representing the uncertainty of each experimental value arose from the resolution of the TEM and NEMS actuator, and the calculations are performed as described in Supplementary Discussion 2.