W{100} clean surface phase transition studied by core-level-shift spectroscopy: Order-order or order-disorder transition.

Abstract

W4f/sub 7/2/ core-level spectra are recorded from the high- and low-temperature phases on a ''flat'' W(100) crystal. In the high-temperature phase, the core-level binding energy associated with surface atoms is 370 +- 10 meV lower than that from the bulk atoms. This binding energy is found to be shifted by 15 +- 5 meV in the low-temperature phase. The magnitude of this shift and an excess broadening observed in the surface peak from the high-temperature phase give evidence for disorder in this phase, in which the surface atoms have similar local coordination to those of the low-temperature phase. The surface core-level shift of the low-temperature phase agrees well with the structural model proposed for this phase by Debe and King, in contrast with a previous conclusion. Assuming that defects can induce long-range order in the high-temperature phase which differs in nature from that in the low-temperature phase, one can explain the disagreement in the literature concerning the value of the binding energy associated with the W(100) surface atoms in the high-temperature phase, as well as the discrepancies over the observed shifts of this binding energy upon the low-temperature phase transition.