Bulk and surface charge states of ${\mathrm{K}}_3{\mathrm{C}}_{60}$

We detect a significant angle-dependence in the core level and valence line shapes of photoelectron spectra of single crystal ${\mathrm{K}}_3{\mathrm{C}}_{60}$. This allows the identification of bulk and surface components in the data, and allows us to explain the anomalous line shapes observed for this system. The states near the Fermi level are associated with the bulk of the sample. There is strong evidence of an insulating surface layer, which we ascribe to intermolecular electron correlations. These results simplify the interpretation of previous, apparently conflicting observations.

Figure 1

C $1s$ photoemission spectra of ${\mathrm{K}}_3{\mathrm{C}}_{60}$, measured at the indicated angles.

Figure 2

(a) C $1s$ photoemission spectra of ${\mathrm{K}}_3{\mathrm{C}}_{60}$ at the indicated emission angles, scaled to give similar intensities for the shoulder at $284.2\mathrm{eV}$. (b) The dashed line is the difference spectrum between normal and grazing emission as scaled in (a). The other curves are discussed in the text. (c) Normal emission spectrum and surface line shape, scaled to extract the bulk component. (d) Approximate bulk line shape obtained as the difference derived from the spectra in (c). The dashed line illustrates a plausible main line for a metallic line shape. Also shown are suggested corrections (arrows) due to the influence of scattered electrons. See the text for more details.

Figure 3

C $1s$ photoemission spectra at the indicated emission geometries, with the relevant surface and bulk contributions obtained from the line-shape analysis shown in Fig. 2. Both components are recombined to fit the normal and grazing emission data, with the result shown as a solid line. See the text for a discussion.

Figure 4

(a) Normal emission valence PES spectrum of ${\mathrm{K}}_3{\mathrm{C}}_{60}$. The solid line is a model given by the sum of the illustrated images of the normal emission C $1s$ spectrum (dashed lines), placed according to the peaks of pure ${\mathrm{C}}_{60}$ as suggested by the vertical lines. (b) Comparison with a pure ${\mathrm{C}}_{60}$ spectrum (shifted in energy) to help illustrate the uniform shifts. See the text for more details.

Figure 5

Valence and core level data of ${\mathrm{K}}_3{\mathrm{C}}_{60}$ taken at the indicated photon energies. The $1487\mathrm{eV}$ data (Ref. 14) were taken at 45° emission with unpolarized light, the $110\mathrm{eV}$ data at normal emission and with the light polarization along the emission direction. The $110\mathrm{eV}$ data are broadened (solid line) to mimic the effects of the poorer instrumental resolution at $1487\mathrm{eV}$. (a) Normalized core level spectra. Taking differences in emission angle and IMFP into account both spectra are expected to have similar surface and bulk contribution. (b) The valence PES are scaled to match at the shoulder as indicated by the long arrows. The small arrows indicate the surface peak intensities. See the text for more details.

Figure 6

Schematic of the lattice sites of the ${\mathrm{K}}_3{\mathrm{C}}_{60}$ surface and immediate subsurface regions. Fulleride sites are indicated by molecular images, K ions by letters and/or numbers. The charge states of the molecules are indicated by the absence or presence (and type) of circles around the molecular images, as specified in the legend, along with details of the local charge configurations for particular sites. (a) View from the top of the ${\mathrm{K}}_3{\mathrm{C}}_{60}$ (111) surface, in which only the outermost layer of fulleride ions is visible, seen to consist of ${\mathrm{C}}_{60}$ molecules in charge states $-1$ and $-2$. The potassium ions of the three layers below are shown as well. To help illustrate the stacking sequence, three cross sections are taken, indicated by the horizontal lines. (b) Side view of the three cross sections suggested in (a). Height in this figure corresponds to vertical placement, and horizontal placement depicted by the three cross sections is indicated by the dashing of the lines through a given set of sites. This figure allows one to assess the qualitative variation in Madelung potential. (c) 3D illustration of the arrangement of the slices indicated by the horizontal dark lines in (a) and (b).

Figure 7

LUMO-derived band of ${\mathrm{K}}_3{\mathrm{C}}_{60}$ for the indicated emission angles. The light polarization is always parallel to electron emission direction. The model shown is described in the text.

Figure 8

Comparison of the present LUMO-derived photoemission spectrum to that of a ${\mathrm{K}}_3{\mathrm{C}}_{60}$ monolayer (Ref. 22). The difference spectrum shows intensity at higher binding energy in the region of the expected surface contribution, as explained in the text.

Figure 9

(a), (b) K $2p$ photoemission of ${\mathrm{K}}_3{\mathrm{C}}_{60}$ at the indicated emission angles, taken at $h\nu =365\mathrm{eV}$. The contributions of inelastically scattered photoelectrons were subtracted after estimation of a Shirley background. The separation of bulk tetrahedral and octahedral sites as previously determined in Ref. 59 is indicated. (c) The difference between normal and grazing emission shown as a dashed line was determined using the illustrated rescaled spectra of (a) and (b). The difference should be a good approximation to the surface line shape. The analysis is not as straight forward as in the C $1s$ region, which is rationalized in the present model as due to the existence of several K surface sites.

Figure 10

K $3p$ PES at the indicated excitation energies.

Figure 11

Angle-dependent PES spectra of the ${\mathrm{K}}_3{\mathrm{C}}_{60}$ LUMO-derived band at the indicated angles and polarization settings.

Figure 12

Valence PES of ${\mathrm{K}}_3{\mathrm{C}}_{60}$ taken at $110\mathrm{eV}$ at the indicated emission angles. The light polarization direction was parallel to the electron emission direction. The height of peaks and shoulders is indicated by braces. The ratio of peak height to shoulder intensity is 1.6 in normal emission and 3.4 in grazing emission.

Figure 13

The IMFP of photoelectrons for the indicated samples. See the text for more details.

Figure 14

Calculated IMFPs from the TPP-2M formulas for the set of materials as shown in Table as a function of kinetic energy.