Fermi edge

The spectrum of the alloy in Figure 8.19 shows pronounced differences. The shape of the Fermi edge is different from that of Cu or Pd and proves to be sensitive to the constitution of the alloy. The peak due to formation of the 3/2 core state of Cu is shifted by 0.94 eV in the alloy and broadened slightly. The two Pd peaks are also shifted, but only slightly, and are narrowed to almost 50 per cent of their width in Pd itself... 

Figure 5.43. UP-spectra of Ag YSZ electrodes for (a) cathodic and (b) anodic polarization of the galvanic cell Ag YSZ Pd,PdO at 547°C. In (b), the shift of the Fermi edge of the small silver particles on YSZ under anodic polarization is shown enlarged (5x).24 Reprinted with permission from Wiley-VCH.

Although the dynamic final state effect can also explain the broadening of the core levels, the expected shift for the Fermi-edge and the core levels should still be quite similar. In fact, in numerous experiments [83,84,86,96,100,101] significantly different shifts were observed for the different spectral elements, which indicates that beyond the final state effects - which probably still give important contributions to shifts - initial state effects also have to be considered. 

On the other hand, the significantly different shift of the Ag core levels and the Fermi edge suggests that the observed changes cannot be explained by a simple final state effect. In addition, taking into account that the BE shift of the Ag 3d levels can be separated into a AF initial state contribution and a h.R final state contribution [96,164]... 

Fig. 21 a Normalized As K-edge XANES spectra for FeAs and some FeAsi j,Py members, measured in transmission mode, b Orbital projections of conduction states calculated from FeAs and FeAso.5oPo.50 (the Fermi edge is at OeV). Reprinted with permission from [61]. Copyright Elsevier... 

Figure 4.23. Photoemission process and UPS spectrum of a semiconducting material. F vacCs) and vac(d) represent the sample and detector vacuum levels, respectively. The Fermi edge of the metal is represented by the discontinuous line.

Fig. 10. A sample XPS valence-band spectrum of Aujj [48], sandwiched between a spectrum for bulk gold and for an Aui,L7X3 cluster. The latter two spectra were taken from Fig. 2 of Ref. [76]. The arrow denotes the position of the Fermi edge...

Multiplet splitting occurs in the core-level region as well as in the energy range which we called valence region in this case it can be due to photoemission from a localized state, although this state may be not very far from the Fermi edge. A typical example of... 

Figure 10 shows combined XPS-BIS results for a-U. All valence band spectra for U, (even if they suffer from poor resolution or if a shght oxygen contamination cannot be excluded) display a strong emission in a narrow band just below the Fermi edge Ep. 

This spectrum has been compared with the BIS measurement on Nd metaP i.e. of the homologous lanthanide. Trivalent Nd has a localized 4f initial state configuration. For U, a 5 f or a 5 f initial state configuration are usually assumed, with a tetravalent or a trivalent core respectively. While in Nd the 4f and 4f multiplet states, as evaluated in an atomic-like Russell-Saunders scheme, can be well recognized in the XPS/BIS combined results, and are well separated from a (weak) d-emission at the Fermi edge, in U the occupied states and the empty states spectra join in a continuous band at Ep. Therefore, only the symmetry of 5 f states, given by the position of the main peaks in the joint spectrum, can be recognized with certainty. 

Fig. 2. Photoemission spectra of KxC6o [8]. K3C60 (x = 0.1, 1.3, and 2.2) is a metal, which shows a clear Fermi edge. Also, there exists a small shoulder at 1.6 eY. On the other hand, K4C6o (x = 4.2) is an insulator, which does not show a Fermi edge. The two phases coexist for x = 2.8 and 3.7.

---