On the contraction of the W(001)-(1 × 1) surface using LEED intensity analysis

Three recent independent attempts at deducing the W(001)-(1 × 1) surface structure by LEED beam intensity analysis have yielded contractions of the topmost layer spacing of 6 ± 6%, 11 ± 2%, 4.4 ± 3% normal to the surface plane. We investigate possible reasons for the discrepancies by comparing published experimental and theoretical profiles of these workers as well as our own. Our main conclusions are that the direct comparison of experimental data of different investigators shows deviations which are comparable to the changes in the calculated profiles for various surface contractions. Also the deviations between calculated intensity profiles using different (but still realistic) assumed scatteting potentials are comparable to the changes in the calculated profiles for various surface contractions. The main uncertainty in the scattering potential is the choice between the Slater free electron exchange-correlation term (coefficient α = 1) or the Kohn-Sham version ($\text{α =}\text{2}\text{3}$) or a value in between. For tungsten the corrections due to relativistic atomic scattering must also be considered. These uncertainties in the calculated and experimental profiles lead to the conclusion that the surface layer contraction of W(001)-(1 × 1) is not known at the present time. To assess the potential of LEED in deducing surface structures of this type further comprehensive analyses are required where the uncertainties in the theoretical scattering potential are also examined.