Chemistry Reference
In-Depth Information
Fig. 7.6 a Calculated potential-energy surface of OH as a function of the lateral displacements
of H atom from the top of the bridge site. The two minima correspond to the inclined geometries,
and the potential barrier between the minimum and the saddle point (for switch motion) is
0.14 eV. The potential minima are 3.94 eV in depth with respect to free OH. b Schematic
illustration of the energetically favorable pathway of the hydroxyl switching
structure (Fig.
7.4
), which is consistent with the previous calculations [
30
,
31
] and
experiment [
35
]. The geometrical parameters are almost same as those reported
previously.
There is a notable isotope effect between OH and OD. Although no remarkable
difference can be seen in the STM appearance between them (Fig.
7.5
a), the
bi-state fluctuation can be observed in the current trace measured over OD.
Figure
7.5
b shows a typical current trace measured over OD while the tip is fixed
over the one depression. This fluctuation was never observed for OH. Thus OH and
OD are distinguishable. It is plausible the bi-stable fluctuation of OD corresponds
to the switching between the two orientations as illustrated in Fig.
7.5
c. The STM
appearance of the paired depression aligned along [001] axis is interpreted as a
time averaged image of the switching. The STM simulation based on the Tersoff-
Hamann approach predicted that the side of the H atom appears brighter than the
opposite position though a depression feature could not be reproduced. I tenta-
tively assigned that the high-current (low-current) state is assigned to the orien-
tations of OD pointing toward (away from) the tip.
To elucidate the dynamical properties of hydroxyl, the potential energy surface
(PES) was calculated along the lateral coordinates of the H atom (Fig.
7.6
a). In the
calculations the Cu atoms in topmost two layers were optimized at each point. The
PES shows a double-well structure, where the two minima correspond to the two
(Footnote 1 continued)
constant of 3.64 Å, which is 0.8 % larger than the experimental value of 3.61 Å, was used to
construct the slabs. A (4 9 4) Monkhorst-Pack [H. J. Monkhorst and J. D. Pack, Phys. Rev. B 13,
5188 (1976).] k-point set was used to sample the surface Brillouin zone, and the Fermi level was
treated by the first order Methfessel-Paxton scheme [M. Methfessel and A. T. Paxton, Phys. Rev.
B 40, 3616 (1989).] with the 0.05 eV smearing width. During the structural optimization,
adsorbates and two topmost Cu layers were allowed to relax until the forces on them were less
than 0.05 eV/Å.
