366
W. Piskorz and F. Zasada
for different STM settings, qualitatively different (with different motif, dimeric or
monomeric) images are obtained.
In [215], the possibility of formation of surface attached metal-organic frameworks (SURMOF) was noted and the HR-TEM image was simulated. The Zn-MOF5 bulk model was built based on the X-ray diffraction study of Li et al. [216]: the
unit cell was cubic (Fm ¯
3m space group) with Zn 32 O 104 C 192 H 96 stoichiometry, and
the initial lattice constant of 25.86 Å. The almost commensurate matching of rutile
and Zn-MOF-5 was noted; hence, the mutual mismatch of the supercell as a function of the number of MOF unit cells was discussed. The MOF cell was found
much softer than rutile cell (calculated bulk moduli of 18.41 GPa and 216.00 GPa,
respectively, both values are close to experimental data). The adhesion of Zn-MOF5 layer to rutile substrate (−0.32 eV nm
−2 ) was found only possible when dispersion forces (addition of −0.39 eV nm
−2 ) are included in the computations to overcome the stress (+0.69 eV nm
−2 ) imposed by interfacial strain (ε[001] = 0.31% and
ε[1 ¯
10] = 2.86%). The HR-TEM simulations were performed with use of the JEMS
software [217, 218].
The other surface-assisted formation of self-organised structures adsorbed on the
rutile surface, both the dispersion forces driven [70] and covalent bonded structures [213, 219], were noticed, the latter stemming from the polycondensation catalysed by protons adsorbed on the rutile surface in the form of hydroxyl groups.
The former case, i.e. the monolayer formed densely due to the intermolecular and
molecule-surface attraction concerned the 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA) on TiO 2 -(110) surface [70]. The STM images for the low-coverage
case were modelled within the Tersoff–Hamann approximation, and it can be seen
that the PTCDA molecules adsorb in the intact state. For the monolayer, several
adsorption modes were tested and, given the experimentally (LEED and STM) found
packing—c(6 × 2), one structure with significant bending of PTCDA molecules
along the protruding O b ridges (see [70] Fig.6) was found. The modelling of the
STM in agreement with experiment was possible only when the Bardeen theory
simulation was performed (see [70], Fig. 6d, e).
The self-organisation of morphological effects, i.e. the formation of nanoripples
by Ar
+ beams in UHV conditions and the reversible reorientation of them by 90
◦
during temperature change, was observed with STM and modelled at GGA level by
Kolmer et al. [188] Several channels of diffusion were postulated, and the calculation
of the energy barriers leads to the conclusion that diffusion of Ti adatoms alone,
characterised by the barrier of 1.14 eV along the [001] direction (“in-channel”), and
ever higher in the other directions (2.79 eV for “cross-channel”, 3.25 eV across the
[001] step), cannot be responsible for the mass movement required for the formation
of ripples. The assistance of the mobile oxygen surface species (diffusion barrier
of 0.42 eV) dramatically lowers the energy barriers (e.g. 0.71 eV for concerted
“in-channel” hop of Ti(O)). For the lowest T = 150 K, the ripples formation is
ruled mainly by the erosion of the ascending step edges as the diffusion processes
are virtually hindered ([188], Fig. 3). For slightly higher temperature, 300 K, the
assisted diffusion of Ti(O) becomes intensive for “in-channel” and the “cross [1–
11]” modes while keeps hindered for “cross-channel” and the “cross [001]”. The
W. Piskorz and F. Zasada
for different STM settings, qualitatively different (with different motif, dimeric or
monomeric) images are obtained.
In [215], the possibility of formation of surface attached metal-organic frameworks (SURMOF) was noted and the HR-TEM image was simulated. The Zn-MOF5 bulk model was built based on the X-ray diffraction study of Li et al. [216]: the
unit cell was cubic (Fm ¯
3m space group) with Zn 32 O 104 C 192 H 96 stoichiometry, and
the initial lattice constant of 25.86 Å. The almost commensurate matching of rutile
and Zn-MOF-5 was noted; hence, the mutual mismatch of the supercell as a function of the number of MOF unit cells was discussed. The MOF cell was found
much softer than rutile cell (calculated bulk moduli of 18.41 GPa and 216.00 GPa,
respectively, both values are close to experimental data). The adhesion of Zn-MOF5 layer to rutile substrate (−0.32 eV nm
−2 ) was found only possible when dispersion forces (addition of −0.39 eV nm
−2 ) are included in the computations to overcome the stress (+0.69 eV nm
−2 ) imposed by interfacial strain (ε[001] = 0.31% and
ε[1 ¯
10] = 2.86%). The HR-TEM simulations were performed with use of the JEMS
software [217, 218].
The other surface-assisted formation of self-organised structures adsorbed on the
rutile surface, both the dispersion forces driven [70] and covalent bonded structures [213, 219], were noticed, the latter stemming from the polycondensation catalysed by protons adsorbed on the rutile surface in the form of hydroxyl groups.
The former case, i.e. the monolayer formed densely due to the intermolecular and
molecule-surface attraction concerned the 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA) on TiO 2 -(110) surface [70]. The STM images for the low-coverage
case were modelled within the Tersoff–Hamann approximation, and it can be seen
that the PTCDA molecules adsorb in the intact state. For the monolayer, several
adsorption modes were tested and, given the experimentally (LEED and STM) found
packing—c(6 × 2), one structure with significant bending of PTCDA molecules
along the protruding O b ridges (see [70] Fig.6) was found. The modelling of the
STM in agreement with experiment was possible only when the Bardeen theory
simulation was performed (see [70], Fig. 6d, e).
The self-organisation of morphological effects, i.e. the formation of nanoripples
by Ar
+ beams in UHV conditions and the reversible reorientation of them by 90
◦
during temperature change, was observed with STM and modelled at GGA level by
Kolmer et al. [188] Several channels of diffusion were postulated, and the calculation
of the energy barriers leads to the conclusion that diffusion of Ti adatoms alone,
characterised by the barrier of 1.14 eV along the [001] direction (“in-channel”), and
ever higher in the other directions (2.79 eV for “cross-channel”, 3.25 eV across the
[001] step), cannot be responsible for the mass movement required for the formation
of ripples. The assistance of the mobile oxygen surface species (diffusion barrier
of 0.42 eV) dramatically lowers the energy barriers (e.g. 0.71 eV for concerted
“in-channel” hop of Ti(O)). For the lowest T = 150 K, the ripples formation is
ruled mainly by the erosion of the ascending step edges as the diffusion processes
are virtually hindered ([188], Fig. 3). For slightly higher temperature, 300 K, the
assisted diffusion of Ti(O) becomes intensive for “in-channel” and the “cross [1–
11]” modes while keeps hindered for “cross-channel” and the “cross [001]”. The