band gap for the higher-mass fullerenes. In all cases the onset energies are much lower than the HOMO-LUMO gap derived from PES/IPES experiments (see §17.1.1), consistent with the excitonic nature of the absorption edge in EELS. From the difference between the EELS and PES/IPES gap values, the exciton binding energy is estimated to be between ~0.5 and 0.8 eV for C60, C70, and C84 [17.26], Also given in Table 17.1 are the values for the static dielectric constant e0 and the Is carbon core level energy for various fullerenes in comparison to graphite.
High-resolution electron energy loss spectroscopy (HREELS) has been used to study the vibrational spectra of fullerenes (see §11.5.9), because this technique is also sensitive to the so-called silent modes, which are forbidden in the first-order Raman and infrared spectra (see §11.5.3 and §11.5.4). Since the energy of the vibrational modes is much smaller than the energies of electron beams used in electron energy loss spectroscopy (EELS), the high-resolution version of EELS is used to study the vibrational spectra.
The high-resolution EELS technique has been used to study the vibrational spectra of C60 and doped C60 compounds, as well as for monolayers of C60 on surfaces. Measurements of the red shifts or blue shifts of these modes for monolayer C60 coverage on substrates indicate whether charge is transferred to the C60 or from the C60 to the surface. This topic is further discussed in §11.5.9.
17.2.6. EELS Spectra for Alkali Metal-Doped. Fullerenes
Electron energy loss spectra have been reported up to 40 eV for alkali metal-doped Rb3C60 and Rb6C60 as well as for Rb3C70 and Rb6C70, where the nominal stoichiometries for doped films were inferred from x-ray characterization [17.21]. The energy loss spectra for Rb^Q,, (see Fig. 17.11) show a marked dependence on alkali metal concentration, although the spectra for RbxC60 and RbxC70 for similar x values show many similarities. As shown in Fig. 17.11, analysis of MXC60 data to yield e^w, q) and e2((o, q) indicates a large negative €,(0) for Rb3C60, as expected for a metallic system, while the energy loss spectra for C60 and Rb6C60 indicate semiconducting behavior. In contrast, the EELS spectra for the nominal composition 70 do not indicate conducting behavior, and this is attributed to the much lower maximum in the electrical conductivity (by about two orders of magnitude) of alkali metal-doped C70 relative to that of Rb3C60. Also of interest is the very similar behavior of the energy loss spectra for different alkali metals M^C60 for similar x values, indicative of the small hybridization between the fullerene and alkali metal wave functions, as is also observed in many other experimental studies. Of interest are the relatively high values for the low-frequency dielectric function for Rb6C60 [e,(0) = 7.1] and for Rb6C70 [e,(0) = 8.0] [17.21]. For all Rb,C60 and RbxC70 thin films that were studied (x = 0,3,6), tt -> n* structure was observed below 8 eV and a a* or (tt + cr) -*■ (tt + a)* structure above 8 eV [17.21]. Using the data in Fig. 17.11, the frequency dependence of the optical conductivity can be calculated, and the results for Rb3C60 show a peak at 0.5 eV and a second smaller peak at 1.1 eV. The lower-energy peak is identified with
Was this article helpful?