Ultrahigh Critical Current Density across Sliding Electrical Contacts in Structural Superlubric State

In conventional sliding electrical contacts (SECs), large critical current density (CCD) requires a high ratio between actual and apparent contact area, while low friction and wear require the opposites. Structural superlubricity (SSL) has the characteristics of zero wear, near zero friction, and all-atoms in real contact between the contacting surfaces. Here, we show a measured current density up to $17.5\mathrm{GA}/{\mathrm{m}}^2$ between microscale graphite contact surfaces while sliding under ambient conditions. This value is nearly 146 times higher than the maximum CCD of other SECs reported in literatures ($0.12\mathrm{GA}/{\mathrm{m}}^2$). Meanwhile, the coefficient of friction for the graphite contact is less than 0.01 and the sliding interface is wear-free according to the Raman characterization, indicating the presence of the SSL state. Furthermore, we estimate the intrinsic CCD of single crystalline graphite to be $6.69\mathrm{GA}/{\mathrm{m}}^2$ by measuring the scaling relation of CCD. Theoretical analysis reveals that the CCD is limited by thermal effect due to the Joule heat. Our results show the great potential of the SSL contacts to be used as SECs, such as micro- or nanocontact switches, conductive slip rings, or pantographs.

Figure 1

Experimental setup. (a)–(c) The process of transferring graphite flake. We shear the Pt-cap graphite mesa with tungsten probe precisely controlled by a micromanipulator. The $4\mathrm{\mu }\mathrm{m}\times 4\mathrm{\mu }\mathrm{m}$ flake is stuck to the tip of probe and placed on the surface of a freshly cleaved $14\mathrm{\mu }\mathrm{m}\times 14\mathrm{\mu }\mathrm{m}$ substrate. Scale bar, $10\mathrm{\mu }\mathrm{m}$.(d) The illustration of our experimental setup.

Figure 2

The measurements of CCD between graphite interfaces. (a) Raman spectroscope mappings of the graphite substrate’s sliding surface after sliding with different current densities from small to large (10.5, 15.6, 16.7, and $17.8\mathrm{GA}/{\mathrm{m}}^2$). We scanned the Raman spectroscope mapping after each test. Scale bar, $2\mathrm{\mu }\mathrm{m}$. (b) Raman spectroscope curves of different points marked in (a) with the Raman shift from 1000 to $2400{\mathrm{cm}}^{-1}$. (lines 1 to 17 from bottom to top, marked with a red circle at peak D of line 17, The lines not shown are similar to lines 1, 5, 9, and 13). (c) The relationship between friction and normal load before ($0\mathrm{GA}/{\mathrm{m}}^2$) and after (16.7 and $17.8\mathrm{GA}/{\mathrm{m}}^2$) different current densities.

Figure 3

The scaling relation of CCD and comparison with existing reports. (a) The relationship between CCD and sizes of graphite flakes. For each size, multiple samples are tested to obtain the intrinsic value of CCD. (b) The relationship between the square of CCD ${(J_{\mathrm{c}}^L)}^2$ and the size of the graphite flake. (c) The relationship between CCD and the size of the contact. The red and purple points represent our and previous work [9, 10, 11, 15, 29, 44, 45], respectively. The blue line is the fitting according to Eq. (3). (d) The relationship between the CCD and the friction coefficient. The red and purple points represent our and previous work [10, 11, 15, 44, 45], respectively.

Figure 4

The durability of SSL SEC carrying large current. (a) The transmitting current during more than 2000 cycles under the 0.5 V bias voltage. The amplitude and frequency of the cycle are $1\mathrm{\mu }\mathrm{m}$ and 3 Hz. (b) Raman mapping of the sliding area. The brightness represents the signal intensity around the defecting peak ($\sim 1350{\mathrm{cm}}^{-1}$). (c) Raman signals of different points in (b) with the Raman shift from 1000 to $2400{\mathrm{cm}}^{-1}$.