${\mathrm{C}}_{60}$ Molecular Bearings and the Phenomenon of Nanomapping

Inspired by suggestions of ${\mathrm{C}}_{60}$ “nanobearings,” we have measured sliding friction on fixed and rotating ${\mathrm{C}}_{60}$ layers to explore whether a lubricating effect is present. We refer to this general phenomenon as “nanomapping,” whereby macroscopic attributes are mapped in a one on one fashion to nanoscale entities. Our measurements are the first to directly link friction to a documented molecular rotation state. Friction is, however, observed to be higher for rotating layers, in defiance of the ball-bearing analogy. Thus, no direct mapping of macro- to nanoscale attributes can be established.

Figure 1

Schematic of the QCM experiment. On Ag(111) or Cu(111), ${\mathrm{C}}_{60}$ molecules in the 2nd layer will rotate rapidly in random, independent directions, with rotational diffusion constants of $1.8\times {10}^{10}{\mathrm{s}}^{-1}$ [8, 11]. On Ag(111), ${\mathrm{C}}_{60}$ molecules in the 1st layer will ratchet slowly between preferred orientations. On Cu(111), ${\mathrm{C}}_{60}$ molecules in the 1st layer will not rotate [9, 12].

Figure 2

Average methanol frequency shift and coverage data for four ${\mathrm{C}}_{60}$ monolayer (solid circles) and four ${\mathrm{C}}_{60}$ bilayer (open triangles) samples. The data sets for monolayer vs bilayer ${\mathrm{C}}_{60}$ were box-averaged to give a mean frequency shift, and the error bars represent 1 standard deviation from the mean.

Figure 3

The average slip times for methanol on monolayer ${\mathrm{C}}_{60}$ (solid circles) and bilayer ${\mathrm{C}}_{60}$ (open triangles) samples. The four data sets were box-averaged to give a mean slip time, and the error bars represent 1 standard deviation from the mean.

Figure 4

The average slip times for krypton sliding on ${\mathrm{C}}_{60}$ monolayer (solid circles) and ${\mathrm{C}}_{60}$ bilayer (open triangles) films. The 4 data sets were box-averaged to give a mean slip time, and the error bars represent 1 standard deviation from the mean. Because of the box-averaging of the data, the liquid-solid transition usually present in krypton isotherms cannot be seen here, although it was present in the raw data sets.