Comment on: '(2 x 1)-(1 x 1) phase transition on Ge(001): quasi-chemical approximation and Monte Carlo simulations' [Surface Science 563 (2004) 99-109]

摘要

It is well known that the Ge(001) surface reconstructs to form dimers, allowing a reduction of the dangling bonds by 50% compared to the bulk terminated case. At moderate temperatures the di-mers appear symmetric giving rise to a (2x1) structure. At lower temperature the dimers are buckled in an alternating fashion, leading to a further reduction of the symmetry forming a c(4 x 2)-reconstructed surface. Around room temperature both phases coexist in an extended temperature range. In all published (and all our unpublished) STM images the dimers have been observed to align in dimer-rows without exception. Broad agreement has been obtained on the details of these low and intermediate temperature surface structures. In contrast, the nature of the reversible phase transition (2 x 1)-(1 x 1) occurring close to the melting temperature is still heavily disputed. It is consistently reported to take place around 900-1100 K. However, persistent disagreement prevails on the extent of the dimer stability throughout the phase transition. Some authors believe that the dimer concentration is essentially maintained and make domain wall proliferation responsible for the phase transition, while others claim that the phase transition is the direct result of massive dimer break-up. The results of Cvetko et al. [8] are consistent with this latter view. The first conclusion is indirect and essentially based on, and follows, the early photo-emission observation of a finite density of states near the Fermi-level throughout the transition, which has been attributed to the presence of symmetric dimers.