A Low-Dimensional Metallic State on Au-Adsorbed Ge(001) Surface
Komori Group
Surface electronic states on bulk semiconductors are suitable for the study of low-dimensional electronic properties because the metallic state in the bulk band gap can provide an ideal low-dimensional metal. Moreover, we can expect to study one-dimensional (1D) surface electronic properties such as Tomonaga-Lüttinger liquid (TLL) and Peierls instability when the surface atomic structure is highly anisotropic. As one of the surfaces with a 1D atomic structure, the Au-adsorbed Ge(001) surface has recently attracted much attention and its electronic and atomic structures have been studied in detail using scanning tunneling microscopy/spectroscopy (STM/STS), and angle-resolved photoemission spectroscopy (ARPES). An interesting STS result of this system was that the surface local density of states (LDOS) depends on the energy E as (E - EF)α with α = 0.5 [1]. This was attributed to TLL while the metallic surface state has an anisotropic two-dimensional (2D) dispersion in the ARPES study [2]. We have studied this surface state by using high-resolution STS to understand the origins of these interesting features [3].
Figure 1 shows the results of STM/STS for the Au-adsorbed Ge(001) surface consisting of the well-known atomic chain structure. The atomic images largely depend on the sample bias voltage. The symmetry of the surface structure has been considered to be c(8×2) while a short range order of a 4×8 superstructure can be seen on the surface. The tunneling spectra always show a dip at EF as in Fig. 1d. However, by fitting the data to the (E-EF)α dependence, α is always larger than 0.7 at the ordered 4×8 area while α is closed to 0.5 at the areas including point defects and domain boundaries of the 4×8 superstructure. Thus, the square-root behavior is ascribed to the surface disorder.
We have measured differential conductivity (dI/dV) maps in the bias-voltage (Vb) region corresponding to the metallic surface state to study the spatial distribution of the surface LDOS. The results shown in Fig. 2 exhibit no 1D channel of high LDOS, which is expected for a 1D surface state. The observed LDOS is consistent with the ARPES observation of the 2D metallic state.
References
- 1. C. Blumenstein et al., Nat. Phys. 7, 776 (2011).
- 2. K. Nakatsuji et al., Phys. Rev. B 80, 081406 (2009).
- 3. J. Park et al., Phys. Rev. B 90, 165410 (2014).