In a semiconducting compound with a rather small electronic band gap, one would expect a great interplay between atomic disordering and electronic properties. For example, the increase in the disordering is reflected in a higher scattering probability of the mobile electrons and consequently in a decrease in the conductivity. This process competes with an increasing overlap of the non-bonding orbitals, and hence the formation of conduction paths, through which the carriers can move freely [1]. Evidence of this competitive behaviour can be observed in all of the elemental semiconductors, which upon melting undergo an electronic transition towards a (semi-)metallic state.

Zintl compounds are semiconductors which possess a peculiar local atomic arrangement, and therefore it is particularly interesting to study the interplay between the electronic and structural degrees of freedom. One can produce Zintl phases by alloying alkali metals (AM) with post-transition metals (PTM). Due to the high difference in the electronegativities of the components (*), the equiatomic mixture forms compounds in which the PTM atoms cluster in polyanionic units, surrounded by the AM cations. The occurrence of the polyanions in such alloys is necessarily accompanied by the formation of covalent bonds, from which the origin of the semiconducting behaviour can be traced. Several neutron diffraction (ND) studies have been performed in the last decades on the melting of this class of semiconductors [2]. These show a persistence of the negatively charged units in the liquid state (at least up to the dissociation temperature), together with the absence of any obvious increase in the conductivity.

To investigate this ability of the alkali metals to prevent extensive electronic orbital overlap, we performed an X-ray absorption spectroscopy (XAS) measurement at beamline BM29. This experiment simultaneously probed the local atomic arrangement and the electronic state of a particular Zintl compound (KTe) under high temperature and high pressure (300-1000 K, 0.1-5 GPa), to understand if any electronic transition can be driven by applying an external force.

These extreme conditions were applied through a two-anvil large-volume press of the Paris-Edinburgh type. The experimental setup included the parallel detection of the diffraction pattern to monitor the sample condition and to check for the continued absence of chemical contaminants during the P-T cycles. To overcome the difficulties arising from the high corrosivity of the alkali metals and the extreme conditions, we have developed a technique by which it is possible to protect the sample from oxygen and moisture, but still allows for the performance of proper XAS experiments.

With this experimental setup we have collected for the first time XAS spectra at several temperatures and pressures of solid and liquid KTe (see Figure 87). We have confirmed the ND results of Fortner [3] who observed a high degree of structural ordering in the liquid at ambient pressure. Our data clearly indicate the existence of Te dimers in the molten state up to 3 GPa. In addition, our measurements show a clear shift of the Te K-edge of approx 1.5 eV, upon melting at the constant pressure of 3 GPa (see Figure 88). An analogous behaviour, even at ambient pressure, is well known to occur in elemental semiconductors, such as Ge or Te [4], where the structural disordering due to the liquid state results in an increased overlap of the orbitals. We postulate that this observation is consistent with the filling of the covalent band gap at the Fermi level, and supports our initial hypothesis as to the role of pressure in this and related systems.

References
[1] J.E. Enderby, A.C. Barnes, Rep. Prog. Phys., 53, 85 (1990).
[2] See the review in: W. van der Lugt, J. Phys.: Condens. Matter, 8, 6115 (1996).
[3] J. Fortner et al., Phys. Rev. Lett., 69, 1415 (1992).
[4] S. De Panfilis, A. Filipponi, Europhys. Lett., 37, 397 (1997).

Authors
S. De Panfilis (a), M. Borowski (a), C. Meneghini (b), M. Minicucci (c), L. Comez (c), T. Neisius (a), A. Polian (d), J.-P. Itié (d).

(a) ESRF
(b) LNF-INFN, Frascati (Italy)
(c) Dip. di Matematica e Fisica, Universitá di Camerino (Italy)
(d) Lab. Physique des Milieux Condensés, Université P. et M. Curie, Paris (France)