Electronic Spectra of Trigonal and Disordered Phases of Tellurium and Selenium. II. Experiment

The electronic structures of trigonal tellurium and trigonal selenium have been investigated by photoemission, using conventional energy-distribution curves (EDC's) and their second derivatives which reveal significant fine structure, embedded, but hardly observable, in EDC's. The two $p$ valence bands ($p_1 \mathrm{and} p_2$) and the second conduction band (SCB) are explored for both materials except for $p_1$ in the case of Se. In particular, the bottom of the SCB is positioned 4.6 and 6.1 eV above the upper edge of $p_2$ in Te and Se, respectively. The spectral structure is interpreted in terms of partial contributions to the total density of states (DOS) of restricted regions of the Brillouin zone, namely, about the $\Delta$ axis and point $H$, as obtained theoretically in Paper I of this work. A similar study is carried out on disordered films of Te and Se, deposited in ultrahigh vacuum onto a flame-polished silica substrate. A comparison between the photoemission spectra obtained from these films and those obtained from trigonal Te and Se reveal a sharp attenuation of the conduction-band-DOS structure in agreement with the theoretical results, Paper I. The valence-band total DOS in these films exhibit more subtle disorder effects, characterized by (i) the loss of the contribution of point $H$ and (ii) the preservation of the $\Delta$ contribution. In addition, it is shown that, depending on the thermal treatment of the films, two sets of disordered configurations may be obtained in Te: one [$a-\mathrm{Te}\mathrm{}\left(1\right)\mathrm{}$] which can effectively be approximated by disordering the trigonal Te lattice, and the other [$a-\mathrm{Te}\mathrm{}\left(2\right)\mathrm{}$] which represents a further departure from this lattice and could, therefore, be approached via deformation of a different (possibly simple cubic) lattice. Because of the relatively good fit achieved in the experimental electronic structure of the disordered Se films using the trigonal approach, these films do not seem to contain a majority of ${\mathrm{Se}}_8$ rings but might rather consist of a mixture of small weekly distorted Se chains, possibly amalgamated with a small proportion of rings.