Figure 1

(a) Schematic of the junction: ${\mathrm{C}}_{60}$ monolayer is supported on a thin ${\mathrm{Al}}_2{\mathrm{O}}_3$ film grown on the NiAl(110) crystal. Ag atoms are deposited on top of ${\mathrm{C}}_{60}$ (it is possible, however, that Ag atoms could adsorb below the ${\mathrm{C}}_{60}$ layer; this could not be determined from the experiment). The carbon skeletons and electron clouds of ${\mathrm{C}}_{60}$ are shown schematically. (b) STM image ($V_{\mathrm{b}}=2.5\mathrm{V}$, $I=0.1\mathrm{nA}$) of a Ag-doped ${\mathrm{C}}_{60}$ monolayer on the oxide film. The sample bias voltage $V_{\mathrm{b}}$ and tunneling current $I$ were chosen to be low enough for the ${\mathrm{C}}_{60}$ monolayer and $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$ complexes to be stable during the experiment. Arrows show individual $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$ complexes. ${\mathrm{C}}_{60}$ molecules in the monolayer have different orientations and appear as protrusions of various shapes (ordered triangular lattice in the image). (c) Curve $A$: $dI/dV$ spectrum of a ${\mathrm{C}}_{60}$ monolayer on the oxide film. The tunneling gap was set with $V_{\mathrm{b}}=2.5\mathrm{V}$ and tunneling current $I=0.1\mathrm{nA}$. Curve $B$: $dI/dV$ spectrum measured on a $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$ complex, gap set with $V_{\mathrm{b}}=1.5\mathrm{V}$, $I=0.1\mathrm{nA}$. Curve $C$: the fine features in curve $B$. Curve $D$: $dI/dV$ spectrum measured on the bare oxide film (gap set with $V_{\mathrm{b}}=2.5\mathrm{V}$, $I=0.1\mathrm{nA}$). The curves were scaled (the corresponding factors shown in the figure) and offset for clarity. (d) Tunneling giving rise to $P_1{}^{\prime }$. The ionized state of $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$, as well as the NiAl and tip Fermi levels $E_{\mathrm{F}}{}^{\mathrm{NiAl}}$ and $E_{\mathrm{F}}{}^{\mathrm{tip}}$ are shown for the unbiased ($E_{\mathrm{F}}{}^{\mathrm{NiAl}}=E_{\mathrm{F}}{}^{\mathrm{tip}}$) and biased ($E_{\mathrm{F}}{}^{\mathrm{NiAl}}\ne E_{\mathrm{F}}{}^{\mathrm{tip}}$) junction, with $E_{\mathrm{F}}{}^{\mathrm{NiAl}}$ level taken as a reference. The voltage drop in the junction is shown by a sloped solid line. The voltage drop in the ${\mathrm{C}}_{60}$ monolayer is not shown for simplicity.Reuse & Permissions

Figure 2

Spatial distributions of $dI/dV$ intensity measured around $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$ for a set of bias voltages $V_{\mathrm{b}}$, as indicated ($I=0.1\mathrm{nA}$ for all images). The image in the upper-left corner is the corresponding STM topography ($V_{\mathrm{b}}=2.2\mathrm{V}$, $I=0.1\mathrm{nA}$), with $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$ seen as a bright protrusion. Each $dI/dV$ image was recorded using the same $V_{\mathrm{b}}$ for topography and measurement of the $dI/dV$ signal. The edge of a ring from another $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$ complex is seen in the lower-right corner (for $V_{\mathrm{b}}=2.1–2.4\mathrm{V}$). A third $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$ is located in the upper-left corner (for $V_{\mathrm{b}}=2.2,2.3\mathrm{V}$).Reuse & Permissions

Figure 3

(a) $dI/dV$ map from Fig. 2 ($V_{\mathrm{b}}=2.2\mathrm{V}$) showing 41 points where the $dI/dV$ spectra in (b) were recorded (gap set by $V_{\mathrm{b}}=2.2\mathrm{V}$, $I=0.1\mathrm{nA}$ for all spectra). The curves in (b) were offset for clarity. The broad peak seen in spectrum “1” is due to another $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$, as described in the caption of Fig. 2. (c) $P_1{}^{\prime }$ evolution extracted from (b). The arrows in (a) and (c) show the locations where $P_1{}^{\prime }$ is “in resonance” with $V_{\mathrm{b}}$. Spatial regions I in (a) and (c) correspond to uncharged $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$, whereas in region II the complex is ionized. The two regions are separated by the “ring.”Reuse & Permissions

Figure 4

(a) STM topography of two $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$, marked as $A$ and $B$ ($V_{\mathrm{b}}=2.45\mathrm{V}$, $I=0.1\mathrm{nA}$), showing 50 points where the $dI/dV$ spectra in (d) were recorded. (b) $dI/dV$ image recorded simultaneously with (a). The different sizes of the two rings are caused by slightly different ionization energies of the two complexes. (c) Diagram showing three spatial regions (I, II, and III) corresponding to different charge occupancies of the two complexes (see text). (d) $dI/dV$ spectra measured across the two $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$ (gap set by $V_{\mathrm{b}}=2.4\mathrm{V}$, $I=0.1\mathrm{nA}$) The curves were offset for clarity. (e) Positions of peaks from (d) are plotted versus distance from point 1 (black circles). Gray squares show the data from a different set of $dI//dV$ spectra [see (f)]. Arrows show the locations where $V_{\mathrm{b}}$ is in resonance with one of the peaks (this defines the border between regions II and III). The dashed lines show extrapolated positions of the peaks for a hypothetical system of two noninteracting $\mathrm{Ag}({\mathrm{C}}_{60}{)}_3$. (f) The detailed structure of the “avoided crossing” point, showing switching of complex $A$. Letters $A$, $A^{\prime }$, $B$, and $B^{\prime }$ mark different branches of the data.Reuse & Permissions