Experimental Determination of the Optical Density of States in Iron

Abstract

Energy distributions of photoemitted electrons and the spectral distribution of quantum yield from iron ($\phi =4.8\mathrm{}$ eV) have been measured to a maximum photon energy of 11.6 eV. These data are presented and interpreted in terms of the electronic structure of iron. No evidence is found in these data consistent with the assumption that conservation of k is an important selection rule. Rather, it is found that the data can be interpreted in a consistent manner if the optical transition probability is assumed to depend only on the initial and final densities of states. The results allow determination of the optical density of states in the regions $-6.0\mathrm{}\le (E-E_F)\le 0$ and $5.5\le (E-E_F)\le 11.6$ eV, where $E_F$ is the energy of the Fermi level. Maxima are found in the valence-band optical density of states at 0.35, 2.4, and 5.5 eV below $E_F$. This result is similar to that obtained in nickel, except the lowest-energy peak is not as strong and occurs at a lower energy in iron. The conduction-band optical density of states is approximately constant in the region observed. The iron samples were also coated with approximately one monolayer of cesium to reduce the work function ($\phi =1.55\mathrm{}$ eV) and thereby extend the range of measurements. Strong transitions are observed near $\hslash \omega =2\mathrm{}$ eV, for which the matrix elements vary markedly with $\hslash \omega$. The results, obtained at higher photon energies, are in reasonable agreement with the noncesiated data and suggest that the conduction-band optical density of states decreases monotonically by a factor of two between 2.5 and 5 eV above $E_F$.