Variable temperature photoemission studies on clean Ge(100) have
revealed the presence of a 'metallic' surface state in the band gap
region [1]. A recent report clearly shows that the surface state is
actually located above the Fermi level by approx. 0.1 eV [2]. Several
explanations have been proposed to the extraordinary phenomenon,
e.g., electron doping due to thermally activated adatoms from step
edges, symmetrization of the dimers and thermal excitation of
electrons into unoccupied states.
We present new photoemission results obtained between 180 K and 830
K. Our measurements reveal a maximum of the intensity in the range
520 K - 570 K. At temperatures higher than 570 K we observe a gradual
decrease in the intensity. Angle resolved spectra show that the
surface state has a well defined k//-dependence that follows the
c(4x2) periodicity. Very similar results were obtained on both an
undoped and a 10 mΩcm n-type sample. The overall appearance of the
feature was found to be quite insensitive to sample preparation and
it was not particularly sensitive to contamination. LEED investigations in the same temperature range show how a nearly perfect c(4x2) pattern becomes streaky and finally turns in to a 2x1
pattern. The onset of the structure above the Fermi level takes place
just before all c(4x2) streaks have disappeared which corresponds to
a temperature slightly above room temperature. This is in contrast to
an earlier study where the onset was more strongly associated with
the transition to the 2x1 phase observed in LEED [1]. We find that
the emission above the Fermi level can be explained by thermal
occupation of a π* band derived from a c(4x2) ordering of the Ge
dimers. It is known from Si(100), that photoemission shows two
occupied surface state bands, significative of the c(4x2)
reconstruction, even though LEED only shows a 2x1 pattern. This
suggests that the c(4x2) band structure survives at higher
temperatures. At sufficiently high temperatures the c(4x2) bands are
expected to disappear and leave only the 2x1 band structure. This
would explain the decrease in the emission observed at higher
temperature.
[1] S.D. Kevan and N. G. Stoffel, PRL 53, 702 (1984)
[2] K. Nakatsuji et al., PRB 72, 241308(R) (2005) |