ChemComm
Communication
twice as low as that of porphyrin moiety II, which indicates that
the former belongs to a porphyrin dimer situated in the bottom
layer. It has an apparent height of 0.8–1.0 Å, which is typically
observed for a monolayer of Cu–Mn–porphyrin dimers physi-
sorbed on HOPG (ESI,† Fig. S2).
Porphyrin moiety II belongs to a dimer in the top layer and
interacts in a J-type geometry with porphyrin moiety I. An
H-type assembly of two porphyrins on top of each other is
observed at the location labelled III. Its apparent height is twice
that of porphyrin moiety I and only slightly higher than the
offset-stacked porphyrins at location II. This difference might
be the result of a somewhat larger steric hindrance between two
porphyrins in an H-type assembly, or of a higher conductivity
through the coplanar p-stacked aromatic systems. In the
consecutive scan, the geometry between the porphyrins in the
top and bottom layers has changed completely to an H-type
structure (Fig. 4B). This geometry change was accompanied and
confirmed by changes in the apparent height profiles, and is
apparently the result a lateral movement of porphyrin dimers in
the top layer, along the lamella (Fig. 4A - B). Although a fully
H-type geometry is not expected to be the energetically most
favourable one, it seems that when a transformation to such
type assembly occurs somewhere in the top layer, it is subse-
quently transferred throughout the entire domain (ESI,† Fig. S3
and S4). We have no evidence that the translation of the top
layer is a direct result of scanning by the tip, since STM images
show the two different aggregation types at different locations
in a single scan line (ESI,† Fig. S3A, B and E). However, the tip
might have induced the movement of the first porphyrin
dimers into their new geometry, after which the packing
gradually changed over the entire domain. After its formation,
the full H-type packing of the porphyrins was observed during
only four consecutive scans (15 min), which indicates that it is
metastable. The porphyrin dimers in the top layer subsequently
moved again along the lamella into a new double layer geometry,
(Fig. 4B - C), resembling the first one. This packing remained
stable for hours. In this final packing the offset-stacked
porphyrin moieties were positioned at the other side of the
porphyrins in the bottom layer when compared to the packing
in Fig. 4A. This difference was also confirmed by the apparent
height profiles. The observed movement of apparently the whole
top layer domain is another indication of the relatively weak
interactions between the two molecular layers. The coherence of
the top layer, which behaves like a two-dimensional sheet, is
attributed to sufficiently strong intermolecular van der Waals
interactions between the alkyl chains.
Fig. 3 Three consecutive STM images (contrast-enhanced, Vb = À450 mV,
Is = 5 pA, Dt = 5 min, scale bar 3 nm) of a bilayer of Cu–Mn porphyrin dimer
2 at the same location at the graphite/1-phenyloctane interface, revealing
the dynamics between the two layers. The arrows point from the Cu to the
MnCl centre within a dimer. Dimers that reoriented with respect to the
previous scan are indicated by a red arrow.
the interactions between the top and bottom layer are relatively
weak, allowing porphyrin dimers from the top layer to desorb and
dissolve into the supernatant solution, while the same or another
molecule from that solution can subsequently fill the vacancy.
This incoming dimer can either orient itself in the same direction
as the previously present one, or adopt an opposite orientation.
We never observed vacancies in the top layer, which indicates
that it is energetically unfavourable to leave the vacancies that are
created by desorption of a porphyrin dimer unoccupied, and that
the dynamic exchange process of desorption and adsorption is
fast on the STM scanning scale. It is also possible that a vacancy
is filled by lateral movement of neighbouring porphyrin dimers,
but in that case we would have expected to observe vacancies at
other locations in the top layer.
Intriguingly, the top layer was also able to reorient itself as a
whole, over rather large domains of 50 Â 50 nm2 (Fig. 4 and
ESI,† Fig. S3). The STM images in Fig. 4 show magnifications of
such a moving domain, revealing a dynamic process which we
attribute to a rearrangement of the top layer. Consider the
porphyrin moieties labelled I, II and III along the cross-section
in Fig. 4A. The apparent height of porphyrin moiety I is almost
Future work will be directed towards investigating the
reactivity of Cu–Mn porphyrin dimer 2 in oxidation reactions,
and in particular whether this reactivity is different in a
monolayer or a bilayer arrangement of the molecules at the
Fig. 4 Top: three consecutive STM images (Vb = À450 mV, Is = 5 pA,
Dt = 5 min, scale bar 3.4 nm) of a bilayer of Cu–Mn porphyrin dimer 2 at the solid/liquid interface.
graphite/1-phenyloctane interface, showing a repositioning of the top layer with
This research was supported by the Council for the Chemical
respect to the bottom layer, first to an H-type (A - B) and then to a J-type
assembly (B - C). Red squares represent porphyrin dimers in the top layer, blue
squares those in the bottom layer. The red arrows indicate the direction in which
porphyrin dimers in the top layer (red squares) move along a lamellar array.
Sciences of the Netherlands Organisation for Scientific Research
(CW-NWO) (VIDI grant 700.58.423 and ECHO grant 700.57.023),
by the European Research Council (ERC Starting grant NANO-
Bottom: cross sections corresponding to the yellow dashed traces in the images. CAT – 259064), and by the Ministry of Education, Culture and
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 7291--7294 | 7293