Ordered arrays of organometallic iridium complexes with long alkyl chains on graphite

Article abstract of DOI:10.1002/chem.200600972

Iridium(III) fac-tris(2-phenylpyridine) fac-[Ir(ppy)₃] complexes equipped with long alkyl chains were prepared to examine their apability to form organized arrays on the surface of highly oriented pyrolytic graphite (HOPG). The molecules form l

Full text of DOI:10.1002/chem.200600972

FULL PAPER
DOI: 10.1002/chem.200600972
Ordered Arrays of Organometallic Iridium Complexes with Long Alkyl
Chains on Graphite
Joe Otsuki,*[a] Tsubasa Tokimoto,[a] Yuki Noda,[a] Tomohiro Yano,[a]
Takeshi Hasegawa,[b] Xiaoming Chen,[c] and Yoshio Okamoto[c]
Abstract: Iridium
nylpyridine) fac-[Ir
N
slips). Infrared external reflection spec-
troscopy suggested that the adsorbed
alkyl chains adopt the trans-zigzag con-
formation in the nanoslip, although the
orientations of the zigzag plane of the
alkyl groups are mixed. Cyclic voltam-
metry indicates fast electron transfer
between the adsorbed molecules and
the substrate and significant intermo-
lecular electronic interactions. It was
found that annealing at high tempera-
tures is an effective method to prepare
ordered assemblies more than a few
micrometer scale (microslips). The ori-
entations of the nanoslips prepared
from the racemic mixture exhibited an
apparent 12-fold symmetry, while its
optically active enantiomer resulted in
more irregular domains with a six-fold
symmetry, implying an important role
of chirality on packing at the molecular
level and on the orientation of the do-
mains at larger scales. When drop-cast
from more concentrated solutions than
a few hundreds of micromolar, multi-
layers were obtained, in which the
alkyl chains in the molecules are more
or less perpendicular to the surface.
This structure can be transformed into
the nanoslips upon standing.
ACHTREUNG
equipped with long alkyl chains were
prepared to examine their capability to
form organized arrays on the surface of
highly oriented pyrolytic graphite
(HOPG). The molecules form lamellar
arrays at the 1-phenyloctane/HOPG in-
terface. From the analysis of the STM
images, it was concluded that the mole-
cules align with alkyl chains being in-
terdigitated. Similar lamellar arrays
were also obtained at the air/HOPG in-
terface upon drop-casting of toluene
solutions. The lamellar structure at the
molecular level leads to rectangular
two-dimensional crystalline domains a
few hundred nanometers long (nano-
Keywords:
nanostructures
iridium
· physisorption
·
·
self-assembly · surface chemistry
Introduction
of electro- and photo-active compounds deserve special at-
tention from the viewpoint of molecular electronic/photonic
applications. Alkylated compounds have a high affinity for
the surface of highly oriented pyrolytic graphite (HOPG),
and many of them form organized monolayers at the solid/
liquid interface, which can be imaged with scanning tunnel-
ing microscopy (STM) at molecular resolution under ambi-
ent conditions.[2] The adlayers of porphyrins and phthalocya-
nines with long alkyl chains have attracted intense attention
with this respect.[3,4] Another important class of photo/elec-
tro-active compounds are trischelating octahedral metal
complexes, wherein a d6 metal ion (e.g., Ru2+, Os2+, and
Ir3+) is chelated by three bidentate ligands (e.g. 2,2’-bipyri-
dine). In contrast to planar compounds represented by por-
phyrins and phthalocyanines, this type of complexes has
rarely been adsorbed on HOPG.[5] In fact, an ordered array
of an octahedral complex with alkyl chains on HOPG has
never been reported to date, to the best of our knowledge,
despite the fact that introducing alkyl chains is so powerful
that a variety of alkylated molecules have been made into
Organized molecular assemblies on surfaces would be a
basis for the bottom-up construction of molecule-based ma-
terials and devices.[1] Especially, molecular layers comprising
[a] Prof. J. Otsuki, T. Tokimoto, Y. Noda, T. Yano
College of Science and Technology
Nihon University
1-8-14 Kanda Surugadai, Chiyoda-ku, Tokyo 101-8308 (Japan)
Fax : (+81)3-3259-0817
E-mail: otsuki@chem.cst.nihon-u.ac.jp
[b] Prof. T. Hasegawa
Department of Chemistry
Graduate School of Science and Engineering
Tokyo Institute of Technology and PRESTO (Japan) Science
and Technology Corporation (JST)
Meguro-ku, Tokyo 152-8551 (Japan)
[c] Dr. X. Chen, Prof. Y. Okamoto
EcoTopia Science Institute
Nagoya University
Furo-cho, Chikusa-ku, Nagoya 464-8603 (Japan)
Chem. Eur. J. 2007, 13, 2311 – 2319
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2311

arrays.[2] It is anticipated that an octahedral complex may be
adsorbed on the HOPG surface if alkyl chains can be intro-
duced at three positions that are in the facial relationship
with each other. This is because the three facial points can
touch a planar surface simultaneously as these are on the
same side of the spherical complex. However, it is usually
difficult to introduce substituents selectively into facial posi-
tions, since the reaction of ligands and a metal ion would
usually end up in the formation of a mixture of meridional
and facial isomers.
Iridium
(III) fac-tris(2-phenylpyridinato-N,C2’), fac-[Ir-
A
Figure 1. Arrangement of Ir-Cn at the 1-phenyloctane/HOPG interface.
a) STM image of Ir-C18 (100”100 nm2; I=98 pA; V=À1.3 V). b) De-
pendence of the line separation (dn) on the carbon number (n) in the
alkyl chain.
(ppy)3], and its derivatives are potential candidates for
highly efficient organic light emitting diodes.[6] The reaction
of iridium precursors with 2-phenylpyridine yields predomi-
nantly fac-[Ir
(ppy)3] under high temperature conditions.[7]
G
Hence, the introduction of alkyl groups to either the pyri-
dine or phenyl ring at the para-position to the central metal
ion allows us to obtain complexes having alkyl chains at the
desired facial positions. We have prepared such complexes,
Ir-Cn (n=18, 22, 30), and investigated the molecular arrays
spontaneously formed on HOPG. Here, we report for the
first time a successful implementation of the long-alkyl-
chain strategy for an octahedral metal complex, which con-
sequently leads to organized two-dimensional assemblies of
organometallic iridium complexes.
fitted by a linear relationship with a correlation constant of
1.0: dn = 0.125”n + 1.25.
The slope in the equation, 0.125 nm, matches well with
the length per carbon atom along the axis of an alkyl chain
extended on HOPG (0.128 nm).[2a] The intercept, 1.25 nm, is
about the size of the Ir(ppy)3 core on the basis of the crystal
G
structure.[8] Hence, the STM image is consistent with a la-
mellar structure model, in which intervals are filled with ex-
tended alkyl chains oriented nearly perpendicular to the
bright lines. An alkyl chain may be snugly fit into the gap
made by a pair of alkyl chains of an Ir-Cn molecule, since
the width per close packed alkyl chain is 0.4–0.5 nm[9] and
the oxygen–oxygen distances on the periphery of the Ir-
A
C18 on HOPG is depicted in Figure 2 on the basis of these
Results and Discussion
Figure 2. Possible arrangement of Ir-C18 on HOPG consistent with the
STM images at the liquid/solid interface.
Assemblies at the liquid/solid interface: Samples for STM
observation at the liquid/solid interface were prepared by
depositing a drop of a 1-phenyloctane solution of Ir-Cn on
freshly cleaved HOPG via a syringe. Figure 1a shows an
STM image obtained for a 1 mm solution of Ir-C18 on
HOPG. The image shows several crystalline domains with
dimensions of tens of nanometers. In each domain, a paral-
lel line pattern is observed. The bright lines of different di-
rections meet at domain boundaries. It is most likely that
considerations. In this figure, the orientation of the alkyl-
chain plane (all represented by the flat orientation) and the
chirality of the Ir(ppy)3 core (represented by the L enantio-
A
mer) are not taken into account. These aspects will be dis-
cussed for assemblies at the air/solid interface in later sec-
tions.
these lines correspond to the row of aligned Ir(ppy)3 units,
A
while the gaps are filled by alkyl chains. The separation be-
tween the neighboring lines is d18 = 3.5 nm. It has been
found that the other Ir-Cn complexes also assemble them-
selves into similar parallel line patterns. The line separations
for Ir-C22 and Ir-C30 were found to be 4.0 and 5.0 nm, respec-
tively. These separations (dn in nm) are plotted against the
alkyl chain length (n) of Ir-Cn as in Figure 1b. The data are
Despite repeated attempts, images for individual Ir(ppy)3
A
units have not been obtained. This may probably be, at least
partly, due to an intrinsic difficulty to obtain a high resolu-
tion image for samples with large undulations. Other factors
which may be involved include a dense packing of the Ir-
AHCTREUNG
2312
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2311 – 2319

Organometallic Iridium Complexes
FULL PAPER
Assemblies at the air/solid interface—observation with dy-
namic force microscopy (DFM): We have also prepared or-
ganized assemblies of Ir-C18 at the air/solid interface by a
simple drop-cast method and investigated their morphology.
In a typical sample preparation, a 4 mL solution of Ir-C18 in
distilled toluene was dropped via a syringe on the surface of
HOPG (12”12 mm2), which was then air-dried. Thus pre-
pared substrate was subjected to DFM observations. The ob-
served images were relatively homogeneous over the whole
substrate, although the density of the molecules somewhat
depends on the macroscopic positions on the substrate.
Therefore, the following discussion is based on the most
widely observed morphologies for a given concentration of
the cast solution.
Figure 3 displays the surface morphologies as observed
with DFM. When the concentration of the cast solution is
less than ꢀ100 mm, needlelike protrusions are found scatter-
ing around on the surface, as shown by the 5”5 mm2 image
in Figure 3a. The needlelike objects are in fact nearly rec-
tangular crystalline domains (nanoslips) with well-defined
linear edges, with typical domain size being a few tens of
nanometer wide and ꢀ200 nm long, as shown in the magni-
fied image in Figure 3b. The cross section of the nanoslips
shows that the height is homogeneous at 0.63Æ0.11 nm.
This apparent height measured by DFM indicates that the
nanoslips have monolayer thickness. In fact, this height is
measurement conditions, is less sensitive to the 1 nm high Ir-
A
erted by an isolated blip is largely cancelled by the reduced
attractive force from the surrounding flat part of the mole-
cules, as the tip is retracted when tracing above the blip.[10]
The use of less concentrated cast solutions resulted in the
appearance of similar nanoslips, only that the surface densi-
ty was lower. In the surface assemblies prepared from more
concentrated solutions, on the other hand, differently orient-
ed nanoslips partly merge into a membranous layer. An ex-
ample is shown in Figure 3c, which was prepared from a
100 mm cast solution. This trend is more highlighted by the
image in Figure 3d, which was prepared from a more con-
centrated 250 mm solution. In this image most of the surface
is covered by the monolayer film of Ir-C18.
Observation with STM: The STM has revealed the inner
structure of the nanoslips. Figure 4a shows that each nano-
slip consists of a bundle of bright lines running along the
long axis of the slip. The gap between the rows is 3.6 nm,
which is identical with that observed for the array at the 1-
phenyloctane/HOPG interface. This suggests that the molec-
ular arrangement in the liquid/solid interface and in the air/
solid interface is identical as shown in Figure 2. Thus, the
bright lines most likely correspond to the rows of the
aligned Ir(ppy)3 units. The preferential growth directions in
U
between the height of the Ir
which is about 1 nm, and the thickness of an alkyl chain
(0.5 nm). This indicates that the DFM, under the present
A
the case of liquid/solid and air/solid interfaces appear differ-
ent, however; the domain grows more along the lines in the
air/solid interface, whereas the domains tend to be wider in
the direction perpendicular to the lines at the liquid/solid in-
terface. Despite repeated attempts, we were not able to
obtain a molecularly resolved STM image at the air/solid in-
terface either, as can be seen from the narrower area scan in
Figure 4b. More detailed analysis of the molecular arrange-
ment in the nanoslips will be described in a later section in
relation to the chirality of the molecule.
Figure 4. STM images of the two-dimensional nanoslips. a) Wide area
image (100”100 nm2; I=3.0 pA, V=À1.1 V). b) Close-up image (20”
20 nm2; I=3.0 pA, V=À0.9 V).
Infrared external-reflection spectroscopy (IR-ERS): The IR-
ERS was employed to obtain information on the conforma-
tion of alkyl chains in the assembly at the air/solid interface.
The measurements were performed with p-polarized IR ray
with an angle of incidence of 508 from the surface normal.
Figure 3. DFM images of the assemblies of Ir-C18 at the air/solid interface
on HOPG prepared by drop-casting toluene solutions. The full scale in
the height profiles is 1.5 nm. a) Prepared from a 25 mm solution; 5”
5 mm2. b) Prepared from a 25 mm solution; 1”1 mm2. c) Prepared from a
100 mm solution; 1”1 mm2. d) Prepared from a 250 mm solution; 1”1 mm2.
À
Figure 5 shows IR-ERS spectra in the C H stretching vibra-
Chem. Eur. J. 2007, 13, 2311 – 2319
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2313

J. Otsuki et al.
which makes the band location inaccurate. Therefore, the ns-
(CH2) band is more appropriate for the discussion, and we
can conclude that the adsorbed molecules have highly or-
dered alkyl chains. Furthermore, the in-skeleton asymmetric
methyl stretching vibration, na(CH3), band is found at
E
2960 cmÀ1. This vibration is in parallel with the trans-zigzag
plane of an alkyl chain, corroborating the all trans-zigzag
conformation.
On the other hand, the nanoslips in the higher density
sample prepared from a 100 mm solution are more in contact
with others forming a partly membranous planar layer. The
spectrum is broader, with the ns
2853 cmÀ1, higher than that for the lower density sample and
the na(CH2) band appears at almost at the same position
A
À
Figure 5. C H stretch region in the infrared spectra for two different
AHCTREUNG
samples of nanoslips of Ir-C18 on HOPG. The upper spectrum is for a
high density sample prepared from a 100 mm toluene solution, while the
lower spectrum is for a low density sample from a 50 mm solution. Typical
DFM morphologies (250”250 nm2) for respective samples are shown in
the insets.
(2924 cmÀ1). These spectral changes indicate that the alkyl
chains in the denser sample with a membranous morphology
contain gauche-rich conformations. Apparently, alkyl chains
must bend to fill the gaps made by differently oriented
nanoslips to form the planar membranous layer. This argu-
tion region for the nanoslips on the surface with different
densities. Typical surface morphologies of the IR samples
are shown as the insets in the figure.
IR-ERS measured on a nonmetallic surface is known to
have a unique surface selection rule that the surface parallel
ment is consistent with the na
This band location is assigned to the out-of-skeleton na
band, which evidences that the conformation has become
disordered in concert with the inferior order of the alkyl
chains.
ACHTREUNG
ACHTREUNG
transition moment yields
a negative absorbance band,
whereas the surface normal one yields positive absorbance
when the molecule has a specific orientation in the mono-
layer and the angle of incidence is less than Brewsterꢁs
angle. Since the angle of incidence in this study (508) is less
than the Brewsterꢁs angle (ca. 608) of the HOPG/organic
layer interface, the surface selection rule would be held. In
the spectra, however, all the bands appear positively, which
does not obey the rule. This strongly suggests that the alkyl-
chain planes have mixed rotation angles about the axis of
the chain. Two energy-minimized orientations have been
found for alkane molecules adsorbed on HOPG by molecu-
lar mechanics calculations.[9] In one orientation, the plane
including the zigzag carbons is parallel to the basal plane of
graphite, which is denoted as flat orientation. In the other
orientation, the zigzag carbons are on a plane perpendicular
to the surface, which is called vertical orientation. In the sit-
uation wherein these two orientations are mixed, the IR-
ERS spectra should resemble normal IR transmission spec-
tra.
Cyclic voltammetry: Cyclic voltammetry showed that Ir-C18
undergoes a reversible redox reaction at 0.20 V versus ferro-
cenium/ferrocene in a homogeneous acetonitrile/THF 1:2
solution, corresponding to the formal oxidation of the Ir3+
[12]
to Ir4+
.
Cyclic voltammetry was also performed for a
drop-cast and dried sample on HOPG by placing an electro-
lyte solution on the substrate covered with the nanoslips to
examine the electronic communication between the sub-
strate and the molecules in the nanoslips. Figure 6 shows
cyclic voltammograms measured in a 0.5m aqueous Na2SO4
solution for an Ir-C18 sample prepared from a 5 mm cast solu-
tion. The scan rate was varied from 0.1 to 1 VsÀ1. The peak
current values, which are obtained by subtracting the charg-
ing current from the data, are in proportion to the scan rate
Besides, peak maxima in the IR spectra provide informa-
tion on the conformation of the alkyl chains. The lower den-
sity sample prepared from a 50 mm solution is covered with
nanoslips that touch other nanoslips at the ends, as the
DFM image shows in the inset. The symmetric methylene
stretching vibration, ns
(CH2), band appears at 2850 cmÀ1,
E
which suggests that the alkyl chain has the all trans-zigzag
conformation.[11] On the contrary, the asymmetric methylene
stretching vibration, na
which suggests that the alkyl chain has gauche-rich confor-
mations.[11] The band location of the na
(CH2) band is, howev-
A
Figure 6. Cyclic voltammetry for Ir-C18 on HOPG. a) Scan rate dependent
cyclic voltammograms (0.1–1 VsÀ1). b) Relationship between the peak
current values and scan rates.
er, known to be influenced by the Fermi resonance band ori-
ginated from the methylene scissoring band at 1466 cmÀ1,
2314
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2311 – 2319

Organometallic Iridium Complexes
FULL PAPER
as shown in the inset, which confirms that the reaction is
due to the surface immobilized species. The anodic and
cathodic peak potentials are identical, irrespective of the
scan rate; this indicates that the electron transfer reaction
between the molecules and the substrate is fast compared
with the time scale of the experiment. Besides these ideal
behaviors, the full width at half maximum is ꢀ160 mV,
which is much larger than the value of 91 mV for noninter-
acting surface species. In principle, this can be due to sur-
face inhomogeneity and/or interaction among redox active
species. As judged from the images of DFM, and otherwise
highly symmetrical voltammograms, it is likely that the elec-
Orientation of the nanoslips and chirality: Upon drop-cast-
ing, the nanoslips are spread over a wide area on the sur-
face, as exemplified by the image in Figure 3a. The 2D-Fast
Fourier Transform (FFT) was employed to see if there are
any preferred orientations or any correlation with the sym-
metry of the HOPG surface underneath. The FFT on the
DFT image in Figure 3a is shown in Figure 8a, which indi-
trochemical environment for the Ir
A
homogeneous. Therefore, it can be deduced that there are
significant electronic interactions among the Ir
most likely along the molecular row.
A
Annealing effect: The nanoslips on HOPG at the air/solid
interface are stable; no apparent morphology changes are
noticed over weeks, as long as the sample is kept at room
temperature. However, we have found that annealing at
higher temperatures causes a large morphology change in
the surface assembly. Figure 7a shows a DFM image for a
Figure 8. Orientations of racemic monolayer slips on HOPG. a) 2D-FFT
of the DFM image in Figure 3a. Six orientations are discerned. b) STM
image (500”500 nm2; I=3 pA , V=À1.05 V). The lower inset shows the
FFT of this image, superimposed with the FFT of the underlying HOPG,
of which the original data are shown in the upper inset.
cates that the nanoslips are evenly distributed with a 12-fold
angular symmetry, or in six different orientations. The orien-
tations of the enlarged microslips on annealed samples also
show the 12-fold symmetry. In general, alkanes and simple
derivatives thereof adsorb on the HOPG surface with a six-
fold symmetry, reflecting the underlying symmetry of graph-
ite. The strong adsorption of alkane molecules on graphite
surface is explained by the fortuitous match of the distance
between the centers of the hexagons of the graphite lattice,
2.46 Š, and the distance between alternate methylene
groups of the hydrocarbon backbone, 2.51 Š.[2a] Thus, the
long axis of an alkyl chain is expected to orient parallel to
one of the three directions of the lines connecting the cen-
ters of the hexagons, giving rise to a six-fold symmetry, or
three different orientations. It may not be surprising,
though, that six different orientations arise if the growth di-
rections of the two-dimensional crystal are offset by an
angle somewhere between 0 and 308 from one of the sym-
metry axes of the underlying graphite. To give the apparent
12-fold symmetry as in the present case, the offset angle
must be ꢀ158.
Figure 7. Samples of Ir-C18 on HOPG, annealed at 1258C for 40 min.
a) DFM image (5”5 mm2). b) STM image (100”100 nm2; I=3 pA ; V=
À1.0 V).
drop-cast sample, which was then annealed in an oven for
40 min at 1258C, a temperature lower than the melting
point of Ir-C18 (1458C) by 208C. Annealing at temperatures
around the melting point resulted in a similar morphology
change. The nanoslips grow at the expense of other nano-
slips, giving rise to a network of long slips of two-dimension-
al linear crystalline domains (microslips). The growth of the
domain along the long axis stops when encountered with the
side of another domain, but some domains extend over a
few micrometers. The height of these microslips is identical
to that of the nanoslips before annealing. The STM image in
Figure 7b for one of these large area domains clearly shows
that the bright lines are along the long axis of the microslip
as is the case for nanoslips. The separation between the
bright lines is also identical to that of nanoslips before an-
nealing. Thus, the ordered domain can be obtained in a
large area by annealing.
Amodel of the molecular arrangement on the graphite
surface consistent with the above argument is presented in
Figure 9. In the model, the long axis of the alkyl chains in a
molecule is oriented in the direction commensurate with the
graphite surface, along the x axis indicated in the figure.
Then, the next molecule in the same row is placed slightly
shifted from the eclipsed position in such a way that the
growth direction of the crystal is offset by ꢀ158 from the y
axis. There can be another equivalent orientation with the
offset angle of about À158, giving rise to a pair of nanoslips
rotated by 308, which are mutually mirror images. There are
three equivalent directions rotated by 608 on the HOPG sur-
face for this pair of mirror images, which accounts for the
Chem. Eur. J. 2007, 13, 2311 – 2319
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2315

J. Otsuki et al.
total of six different orientations separated by 308, as experi-
mentally observed. Figure 8b shows an STM image and its
2D-FFT (lower inset) for an area of 500”500 nm2 contain-
ing nanoslips. The every 308 distribution is again clearly
demonstrated, although only three orientations are present
due to a limited number of nanoslips in the imaged area.
The underlying graphite surface was also scanned by chang-
ing the tunneling conditions (I=30 pA, V=0.03 V) and the
obtained image was treated with 2D-FFT, as shown in the
upper inset. Comparison of these FFTs reveals that the
nanoslip orientations and the symmetry axes of graphite are
offset in the way consistent with the nanoslip orientations
proposed in Figure 9, although the exact evaluation of the
relevant angles is difficult due to experimental uncertainties.
is significantly blurred reflecting the irregular shape of each
domain. ADFM image for another area containing more
extended domains is shown in Figure 10b. The shapes of the
domains are again somewhat irregular as compared with
racemic samples. The 2D-FFT shows more clearly three ori-
entations rotated by 608 each other. This is in stark contrast
to the racemic mixture, for which six different orientations
are obtained. The different morphologies between the two-
dimensional domains from racemic and optically active ma-
terials strongly suggest that their molecular arrangements in
the domains are different.
Figure 10c shows a DFM image for an annealed sample of
D-Ir-C18. Straight edges oriented 608 to each other are recog-
nized, in contrast to the 308 angle found for annealed sam-
ples of racemic mixture, see
Figure 7a. In addition, there are
many curved edges, which are
not found in racemic samples.
That the morphology of the op-
tically active material is differ-
ent from that of the racemic
mixture also for annealed sam-
ples indicates that the differ-
ence in structure is not merely
due to the kinetic effect, such
as the rate of surface migration,
but due to thermodynamics.
Our attempts to obtain an STM
image afforded only diffuse un-
clear images. We believe that
this is due to a more disor-
dered, and hence more mobile,
molecular arrangement as im-
Figure 9. Schematic drawing of the proposed arrangement of molecules in a pair of mirror image nano- and
microslips. There are other two sets of pairs rotated by 60 and 1208 from this pair, accounting for the total of
six different orientations. This drawing is not meant to represent the precise molecular arrangement but to il-
lustrate the orientation with respect to the underlying graphite surface and the racemic nature of the nano-
and microslips.
If the Ir(ppy)3 unit joining three alkyl chains were achiral,
G
the mirror image nanoslips would be isoenergetic with each
other. However, the sample used here is a racemic mixture
of the D and L enantiomers. Therefore, the chirality may
play a role in the assembly process and the resulting two-di-
mensional structure.[2b,13] An especially relevant question is
whether each nanoslip is a racemate, which contains both
enantiomers, or a conglomerate, which comprises of only
one of the enantiomers.
To shed light on this matter, we have isolated D- and L-
Ir-C18 enantiomers (enantiomeric excess > ꢀ90%) by
means of chiral column chromatography. Then the optically
active material was drop-cast in the same way as for the rac-
emic mixture. Here only the results obtained for the D
isomer are presented, since both enantiomers gave the same
results as expected. The monolayer domains scatter around
on the surface as in the racemic mixture as can be seen in
Figure 10a. However, the shapes of the domains are more ir-
regular than those prepared from the racemic mixture, com-
pare the image with that in Figure 3a. As for the orientation
distribution, three different orientations are apparent from
the 2D-FFT as shown in the inset, although the FFT image
Figure 10. DFM images (5”5 mm2) of monolayer assemblies of D-Ir-C18.
a) Image for an area with scattered molecular domains and its 2D-FFT.
b) Image for another area with larger domains and its 2D-FFT for the
upper part (above the crevice on the HOPG surface). c) Annealed (ca.
1258C for 20 min) sample.
2316
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2311 – 2319

Organometallic Iridium Complexes
FULL PAPER
plied from the irregular shapes of the domains.
On the basis of the above observations, we have conclud-
ed that the molecular arrangement in the two-dimensional
domains from the optically active Ir-C18 is different from
that of racemic Ir-C18. This in turn leads to the conclusion
that the nano- and microslips prepared from the racemic
mixture of Ir-C18 are racemates. This conclusion is incorpo-
rated in the schematic representation of a possible molecu-
lar arrangement in the nanoslips in Figure 9.
Figure 12. DFM images for a time evolution of Ir-C18 assemblies (5”
5 mm2). a) About 5 min after drop-casting. b) The same area an hour
later. 2D-FFT is shown in the inset.
Multilayer assemblies: When (racemic) assemblies were pre-
pared from solutions more concentrated than a few hun-
dreds of micromolar, a different morphology was observed.
Figure 11 shows a DFM image for a sample prepared by
room temperature, more and more needles appeared gradu-
ally, as the DFM image in Figure 12b shows, which was
taken an hour later for the same area. The 2D-FFT clearly
shows the 12-fold symmetry for the newly emerged nano-
slips. Thus, the layered structure transforms into the nano-
slips, which is apparently more stable on the HOPG sur-
face.[14]
Conclusions
In summary, we have shown that ordered arrays of iridium
cyclometallated complexes can be prepared on HOPG both
at the liquid/solid and air/solid interfaces. The key to the
success was to introduce alkyl chains to the metal complex
unit at the facial positions, which was made possible by the
selective reactivity of the iridium phenylpyridine cyclometal-
lation reaction.
Figure 11. DFM image and a height profile for multilayers of Ir-C18 at the
air/solid interface on HOPG (2.5”2.5 mm2).
drop-casting a 1 mm toluene solution. Molecules are assem-
bled to form a mosaic layered structure. The layers are ex-
tremely flat at the molecular level as the height profile
shows. The thicknesses of the layers are integer multiples of
a fundamental thickness of 2.6Æ0.3 nm. It is unlikely that
the molecules are lying flat as in the nano- or microslips,
since the repeat unit of 2.6 nm could not be explained. Also,
the structure of the layer seems not to be directed by the
HOPG surface but to be intrinsic to the molecule, since the
thickness of the first layer directly touching the substrate
surface is the same as that of the second and third layers,
which have no interaction with the surface. Thus, it is most
likely that the layers consist of close packed molecules in a
conformation with alkyl chains more or less perpendicular
to the surface, since the length of the alkyl chain of 18
carbon atoms, 2.3 nm, is close to the measured thickness of
the layer.
We have found a time evolution of the layered structure
to form nanoslips. Figure 12a shows a case in which a single
layer structure was observed on HOPG. The image was
taken about 5 min after drop-casting. The height of amor-
phous-shaped plateaus is 2.5 nm, which is identical to that
observed in the multilayer structure as discussed above.
Therefore, we assume their structures are common, in which
molecules are packed with the alkyl chains oriented upright.
There are also features that appear like needles in the
image, which are in fact the nanoslips. Upon standing at
The alkylated iridium complexes assemble into a lamellar
structure on HOPG both at the liquid/solid and air/sold in-
terfaces, in which the metal complex units align in rows. In
the air/solid interface molecular assemblies, which were pre-
pared by drop-casting toluene solutions of the racemic mix-
ture, give rise to monolayer rectangular domains with the
long axis parallel to the molecular rows (nanoslips). In the
nanoslips, it was indicated by IR-ERS that the adsorbed
alkyl chains adopt the trans-zigzag conformation but the ver-
tical and flat orientations of the hydrocarbon plane are
mixed. Cyclic voltammetry showed a fast electron transfer
between the iridium complexes in the nanoslips and the
HOPG substrate and a significant intercomplex electronic
interaction. It was found that annealing the nanoslip sample
at higher temperatures is a simple and powerful means to
prepare larger two-dimensional single crystalline domains
(microslips). The orientation of the nanoslips from the race-
mic mixture has a 12-fold angular symmetry, while that from
optically active complexes has a six-fold symmetry. Thus, the
chirality of molecules plays an important role in the packing
of the molecules and the angular distribution of the nano-
slips. When drop-cast from toluene solutions more concen-
trated than a few micromolar, molecular layers in which
alkyl chains are more or less perpendicular to the surface
were obtained, which may stack up into multilayer struc-
Chem. Eur. J. 2007, 13, 2311 – 2319
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2317

J. Otsuki et al.
(CH2)14), 0.88 ppm (t, J=6.8 Hz, 3H; CH3); CI-HRMS: m/z: calcd for
C24H42IO: 473.2274; found: 473.2377 [M+H]+.
tures. This molecular layer can be a precursor to the nano-
slips.
2-(3-Docosyloxyphenyl)pyridine: Aflask containing docosyloxy-3-iodo-
benzene (5.8 g, 11 mmol), 2-(tributylstannyl)pyridine (4.05 g, 11 mmol),
Electronic and optical properties of these ordered metal
complex arrays are of great interest, with potential applica-
tions in molecule-based devices. In particular, these ordered
arrays of highly luminescent species are suitable for investi-
gation of electroluminescence or tunneling-current induced
luminescence at the molecular level.[15–17]
and [Pd(PPh3)4] (635 mg, 0.55 mmol) was evacuated in vacuum and then
N
filled with N2. Dry toluene (77 mL) deaerated with N2 was introduced
into the flask, and the mixture was heated under reflux overnight. After
the mixture was filtrated, the solvent was evaporated, and the residue
was purified by chromatography (silica gel, CH2Cl2) to afford a white
solid (1.84 g, 35%). 1H NMR (CDCl3): d=8.69 (d, J=5.5 Hz, 1H; py6),
7.74 (m, 2H; py3, py4), 7.56 (m, 1H; Ph2), 7.53 (d, J=8.0 Hz, 1H; Ph6),
7.37 (t, J=8.0 Hz, 1H; Ph5), 7.23 (m, 1H; py5), 6.96 (dd, J=8.0, 2.4 Hz,
1H; Ph4), 4.05 (t, J=6.8 Hz, 2H; OCH2), 1.81 (quintet, J=6.8 Hz, 2H;
Experimental Section
OCCH2), 1.48 (quintet, J=6.8 Hz, 2H; OCCCH2), 1.4–1.2 (m, 36H;
+
(CH2)18), 0.88 ppm (t, J=6.8 Hz, 3H, CH3); LDI-MS: m/z: 480 [M+H]
;
Syntheses—general: 1H NMR spectra were recorded on a Jeol JNM-
GX400 spectrometer. Mass spectrometers used were a Shimadzu GCMS-
QP5050Aspectrometer for EI-MS, an Applied Biosystems Voyager
System 6123 for LDI-MS, and an Agilent G1969A spectrometer for CI-
MS. Gas chromatography and gel permeation chromatography were con-
ducted using a Shimadzu GC353B system with a TC-1 column and a Jai
LC-9201 recycle GPC apparatus, respectively. EI-HRMS measurements
were carried out by the Chemical Analysis Center of the College of Phar-
macy, Nihon University. Elemental analyses were performed by the
Chemical Analysis Center of the College of Science and Technology,
Nihon University.
elemental analysis calcd (%) for C33H53NO: C 82.16, H 11.13, N 2.92;
found: C 82.46, H 10.78, N 2.89.
2-(3-Octadecyloxyphenyl)pyridine: 1H NMR (CDCl3): d=8.69 (d, J=
5.5 Hz, 1H; py6), 7.74 (m, 2H; py3, py4), 7.56 (m, 1H; Ph2), 7.53 (d, J=
8.0 Hz, 1H; Ph6), 7.37 (t, J=8.0 Hz, 1H; Ph5), 7.23 (m, 1H; py5), 6.96
(dd, J=8.0, 2.4 Hz, 1H; Ph4), 4.05 (t, J=6.8 Hz, 2H; OCH2), 1.81 (quin-
tet, J=6.8 Hz, 2H; OCCH2), 1.48 (quintet, 6.8 Hz, 2H; OCCCH2), 1.4–
1.2 (m, 28H; (CH2)14), 0.88 ppm (t, J=6.8 Hz, 3H; CH3); EI-HRMS: m/
z: calcd for C29H45NO: 423.3501; found: 423.3501 [M]+; purity (GC)
=100%.
2-(3-Triacontyloxyphenyl)pyridine: Amixture of 3-(pyridin-2-yl)phenol
(55 mg, 0.32 mmol), 1-chlorotriacontane (100 mg, 0.22 mmol), and K2CO3
(400 mg, 2.89 mmol) in DMF was refluxed overnight. The reaction mix-
ture was partitioned between CH2Cl2 and H2O. The organic layer was
taken, MeOH was added, and CH2Cl2 was mildly evaporated to precipi-
tate out a white solid (174 mg). This crude material was used without fur-
ther purification in the preparation of Ir-C30. 1H NMR (CDCl3): d=8.69
(d, J=5.5 Hz, 1H; py6), 7.74 (m, 2H; py3, py4), 7.56 (m, 1H; Ph2), 7.53
(d, J=8.0 Hz, 1H; Ph6), 7.37 (t, J=8.0 Hz, 1H; Ph5), 7.23 (m, 1H; py5),
6.96 (dd, J=8.0, 2.4 Hz, 1H; Ph4), 4.05 (t, J=6.8 Hz, 2H; OCH2), 1.81
(quintet, J=6.8 Hz, 2H; OCCH2), 1.48
Iridium complexes, Ir-Cn, were prepared according to Scheme 1. The li-
gands with a long alkyl chain, 2-(3-alkoxyphenyl)pyridine, were obtained
by alkylation of 3-iodophenol followed by the Stille coupling with 2-(trib-
utylstannyl)pyridine.[18] Alternatively, the ligands could also be obtained
by alkylation of 3-(pyridin-2-yl)phenol,[19] obtained by the Stille coupling
of 3-iodophenol and 2-(tributylstannyl)pyridine.[18] The ligands thus pre-
pared were refluxed in ethylene glycol with [Ir(acac)3] to afford the de-
G
sired Ir-Cn. Optical resolution (> ꢀ90% ee) was achieved by chiral
HPLC on a Jasco chromatograph (PU-1580) system equipped with a
Chiralcel OD (Daicel) column by using hexane/2-propanol 96:4.
(quintet, J=6.8 Hz, 2H; OCCCH2),
1.4–1.2 (m, 72H; (CH2)36), 0.88 ppm (t,
J=6.8 Hz, 3H; CH3); LDI-MS: m/z:
591 [M+H]+; EI-HRMS: m/z: calcd
for C41H69NO: 591.5379; found:
591.5378 [M]+; purity (GPC) =96%.
Ir-C22: Amixture of 2-(3-decyloxyphe-
nyl)pyridine (1.23 g, 2.6 mmol) and [Ir-
(acac)3] (210 mg, 0.43 mmol) in ethyl-
ene glycol (4 mL) was heated under
reflux under Ar for 4 d. MeOH was
added to the reaction mixture to pre-
cipitate out the crude product, which
was purified by chromatography (silica
Scheme 1. Synthesis of Ir-Cn.
gel, CH2Cl2/hexane 1:1) to afford
a
1-Docosyloxy-3-iodobenzene: Asolution of 3-iodophenol (7.0 g,
32 mmol) and 1-bromodocosane (8.7 g, 22 mmol) in acetone (100 mL)
was heated under reflux in the presence of K2CO3 (11.6 g) overnight.
After the mixture was filtered, the solvent was evaporated, and the resi-
due was redissolved in CH2Cl2. MeOH was added to the solution, and
the solvent was mildly evaporated to remove CH2Cl2, resulting in the
precipitation of a white solid (9.91 g, 83%). 1H NMR (CDCl3): d=7.28–
7.23 (m, 2H; Ar2, Ar4), 6.98 (t, J=8.0 Hz, 1H; Ar5), 6.85 (ddd, J=8.0,
2.4, 0.8 Hz, 1H; Ar6), 3.91 (t, J=6.8 Hz, 2H; OCH2), 1.76 (quintet, J=
6.8 Hz, 2H; OCCH2), 1.43 (quintet, J=6.8 Hz, 2H; OCCCH2), 1.4–1.2
(m, 36H; (CH2)18), 0.88 ppm (t, J=6.8 Hz, 3H; CH3); EI-HRMS: m/z:
calcd for C28H49IO: 528.2828; found: 528.2828 [M]+; purity (GC)
>99.8%.
yellow solid (246 mg, 35%). 1H NMR
(CDCl3): d=7.80 (d, J=8.4 Hz, 3H; py3), 7.56 (t, J=8.4 Hz, 3H; py4),
7.52 (d, J=5.6 Hz, 3H; py6), 7.23 (d, J=2.4 Hz, 3H; Ph2), 6.83 (t, J=
5.6 Hz, 3H; py5), 6.62 (br, 3H; Ph5), 6.56 (dd, J=8.4, 2.4 Hz, 3H; Ph4),
3.93 (t, J=6.8 Hz, 6H; OCH2), 1.74 (quintet, J=6.8 Hz, 6H; OCCH2),
1.43 (quintet, J=6.8 Hz, 6H; OCCCH2), 1.4–1.2 (m, 108H, (CH2)18),
0.88 ppm (t, J=6.8 Hz, 9H; CH3); CI-HRMS: m/z: calcd for
C99H157IrN3O3: 1629.1860; found: 1629.1858 [M+H]+; purity (GPC)
>99%.
Ir-C18:
1H NMR (CDCl3): d=7.80 (d, J=8.4 Hz, 3H; py3), 7.56 (t, J=
8.4 Hz, 3H; py4), 7.52 (d, J=5.6 Hz, 3H; py6), 7.23 (d, J=2.4 Hz, 3H;
Ph2), 6.83 (t, J=5.8 Hz, 3H; py5), 6.62 (br, 3H; Ph5), 6.56 (dd, J=8.4,
2.4 Hz, 3H; Ph4), 3.93 (t, J=6.8 Hz, 6H; OCH2), 1.74 (quintet, J=
6.8 Hz, 6H; OCCH2), 1.43 (quintet, J=6.8 Hz, 6H; OCCCH2), 1.4–1.2
(m, 84H; (CH2)14), 0.88 ppm (t, J=6.8 Hz, 9H; CH3); LDI-MS: m/z:
1460 [M]+; elemental analysis calcd (%) for C87H132IrN3O3: C 71.56, H
9.11, N 2.88; found: C 71.40, H 8.85, N 2.72.
1-Iodo-3-octadecyloxybenzene: 1H NMR (CDCl3): d=7.23–7.28 (m, 2H;
H2, H6), 6.98 (t, J=8.0 Hz, 1H; H5), 6.85 (ddd, J=8.0, 2.4, 0.8 Hz, 1H;
H4), 3.91 (t, J=6.8 Hz, 2H; OCH2), 1.76 (quintet, J=6.8 Hz, 2H;
OCCH2), 1.43 (quintet, J=6.8 Hz, 2H; OCCCH2), 1.4–1.2 (m, 28H;
2318
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2311 – 2319

Organometallic Iridium Complexes
FULL PAPER
Ir-C30
:
1H NMR (CDCl3): d=7.80 (d, J=8.4 Hz, 3H; py3), 7.56 (t, J=
[3] a) X. Qiu, C. Wang, Q. Zeng, B. Xu, S. Yin, H. Wang, S. Xu, C. Bai,
J. Am. Chem. Soc. 2000, 122, 5550–5556; b) X. Qiu, C. Wang, S.
Yin, Q. Zeng, B. Xu, C. Bai, J. Phys. Chem. B 2000, 104, 3570–3574;
c) H. Wang, C. Wang, Q. Zeng, S. Xu, S. Yin, B. Xu, C. Bai, Surf. In-
terface Anal. 2001, 32, 266–270.
8.4 Hz, 3H; py4), 7.52 (d, J=5.6 Hz, 3H; py6), 7.23 (d, J=2.4 Hz, 3H;
Ph2), 6.83 (t, J=5.6 Hz, 3H; py5), 6.62 (br, 3H; Ph5), 6.56 (dd, J=8.4,
2.4 Hz, 3H; Ph4), 3.93 (t, J=6.8 Hz, 6H; OCH2), 1.74 (quintet, J=
6.8 Hz, 6H; OCCH2), 1.43 (quintet, J=6.8 Hz, 6H; OCCCH2), 1.4–1.2
(m, 216H; (CH2)36), 0.88 ppm (t, J=6.8 Hz, 9H; CH3); CI-HRMS: calcd
for C123H205IrN3O3: 1965.5620; found: 1965.5556 [M+H]+; purity (GPC)
>99%.
[4] a) T. Ikeda, M. Asakawa, M. Goto, K. Miyake, T. Ishida, T. Shimizu,
Langmuir 2004, 20, 5454–5459; b) J. Otsuki, E. Nagamine, T.
Kondo, K. Iwasaki, M. Asakawa, K. Miyake, J. Am. Chem. Soc.
2005, 127, 10400–10405; c) J. Otsuki, S. Kawaguchi, T. Yamakawa,
M. Asakawa, K. Miyake, Langmuir 2006, 22, 5708–5715.
[5] Afew examples of metal complexes on HOPG have been observed
by STM: a) L. Latterini, G. Pourtois, C. Moucheron, R. Lazzaroni,
J.-L. BrØdas, A. K.-D. Mesmaeker, F. C. De Schryver, Chem. Eur. J.
2000, 6, 1331–1336; b) D.-L. Shieh, K.-B. Shiu, J.-L. Lin, Surf. Sci.
2004, 548, L7–L12; c) J.-M. U. Ziener, J.-M. Lehn, A. Mourran, M.
Mçller, Chem. Eur. J. 2002, 8, 951–957; d) M. S. Alam, S. Strçmsdçr-
fer, V. Dremov, P. Müller, J. Kortus, M. Ruben, J.-M. Lehn, Angew.
Chem. 2005, 117, 8109–8113; Angew. Chem. Int. Ed. 2005, 44, 7896–
7900.
[6] a) M. A. Baldo, S. Lamansky, P. E. Burrows, M. E. Thompson, S. R.
Forrest, Appl. Phys. Lett. 1999, 75, 4–6; b) C. Adachi, M. A. Baldo,
S. R. Forrest, M. E. Thompson, Appl. Phys. Lett. 2000, 77, 904–906;
c) G. He, O. Schneider, D. Qin, X. Zhou, M. Pfeiffer, K. Leo, J.
Appl. Phys. 2004, 95, 5773–5777; d) K. Goushi, R. Kwong, J. J.
Brown, H. Sasabe, C. Adachi, J. Appl. Phys. 2004, 95, 7798–7802.
[7] A. B. Tamayo, B. D. Alleyne, P. I. Djurovich, S. Lamansky, I. Tsyba,
N. N. Ho, R. Bau, M. E. Thompson, J. Am. Chem. Soc. 2003, 125,
7377–7387.
[8] V. V. Grushin, N. Herron, D. D. LeCloux, W. J. Marshall, V. A.
Petrov, Y. Wang, Chem. Commun. 2001, 1494–1495.
[9] S. Yin, C. Wang, X. Qiu, B. Yu, C. Bai, Surf. Interface Anal. 2001,
32, 248–252.
[10] H. Onishi, A. Sasahara, H. Uetsuka, T. Ishibashi, Appl. Surf. Sci.
2002, 188, 257–264.
[11] R. G. Snyder, H. L. Strauss, C. A. Elliger, J. Phys. Chem. 1982, 86,
5145–5150.
[12] Y. Ohsawa, S. Sprouse, K. A. King, M. K. DeArmond, K. W. Hanck,
R. J. Watts, J. Phys. Chem. 1987, 91, 1047–1054.
[13] a) L. C. Giancarlo, G. W. Flynn, Acc. Chem. Res. 2000, 33, 491–501;
b) L. PØrez-Garcꢂa, D. B. Amabilino, Chem. Soc. Rev. 2002, 31, 342–
356.
Characterization of the assembly: STM-1 grade HOPG (12”12 mm) was
purchased from GE Advanced Ceramics. STM and DFM measurements
were carried out with an SII SPI3800N-SPA400 microscope under ambi-
ent conditions. Tips for STM measurements were made mechanically
with Pt/Ir 9:1 wire. The bias voltages in Figure captions refer to the sub-
strate voltage with respect to the tip. For the measurements at the liquid/
solid interface, a droplet of an Ir-Cn solution in 1-phenyloctane was ap-
plied on the surface just below the tip using a syringe, after a freshly
cleaved HOPG substrate was observed to confirm the atomic resolution
of the tip. Samples to be observed at the air/solid interface were prepared
by dropping a 4 mL toluene solution on the HOPG surface (12”12 mm),
which was then air-dried. Annealing was carried out in an oven kept at a
constant temperature (Æ5 K) for an indicated duration, after which the
temperature was lowered slowly at a rate of approximately 1 KminÀ1
.
DFM (dynamic force microscopy or a dynamic force mode of atomic
force microscopy) measurements were operated in a noncontact mode
with an aluminum-coated silicon cantilever (SII SI-DF3) with a resonant
frequency of 28–31 kHz, a tip radius of <10 nm, and a spring constant of
1.7 NmÀ1. Cyclic voltammetry was carried out with a Hokuto Denko
HZ-3000 voltammetric analyzer for dried samples on HOPG, which was
placed underneath an electrolyte solution (0.5m Na2SO4 in H2O). Pt and
Ag wires were used as the counter and pseudoreference electrodes, re-
spectively. IR-ERS spectra were measured on a Thermo Electron (Madi-
son, WI) Nicolet 6700 FT-IR spectrometer equipped with a liquid nitro-
gen cooled mercury cadmium telluride detector with a modulation fre-
quency of 60 kHz. The number of accumulations to improve the signal-
to-noise ratio was 3000, and the spectral resolution was 4 cmÀ1. The
sample was vertically placed on a Harrick (Ossining, NY) VRA-1DG re-
flection attachment, which was set in the spectrometer. The p-polarized
IR ray was generated by a Harrick PWG-U1R wire-grid polarizer. For
the collection of background spectra, a cleaved surface of the HOPG
block was used.
[14] We have observed a process of two-dimensional crystalline domain
formation for a related tripod-type molecule: J. Otsuki, S. Shimizu,
M. Fumino, Langmuir 2006, 22, 6056–6059.
[15] S. F. Alvarado, L. Rossi, P. Müller, P. F. Seidler, W. Riess, IBM J.
Res. Dev. 2001, 45, 89–100.
[16] a) M. Vacha, Y. Koide, M. Kotani, H. Sato, Chem. Phys. Lett. 2004,
388, 263–268; b) M. Vacha, Y. Koide, M. Kotani, H. Sato, J. Lumin.
2004, 107, 51–56.
[17] Z.-C. Dong, A. S. Trifonov, X.-L. Guo, K. Amemiya, S. Yokoyama,
T. Kamikado, T. Yamada, S. Mashiko, T. Okamato, Surf. Sci. 2003,
532–535, 237–243.
Acknowledgements
This work was partially supported by the Ministry of Education, Culture,
Sports, Science and Technology, Japan (Grant-in-Aid for Scientific Re-
search on Priority Areas, “Chemistry on Coordination Space” (18033050)
and the High-Tech Research Center Project for Private Universities) and
the Kurata Memorial Hitachi Science and Technology Foundation.
[18] M. van der Sluis, V. Beverwijk, A. Termaten, F. Bickelhaupt, H.
Kooijman, A. L. Spek, Organometallics 1999, 18, 1402–1407.
[19] M. Terashima, K. Seki, C. Yoshida, K. Ohkura, Y. Kanaoka, Chem.
Pharm. Bull. 1985, 33, 1009–1015.
[1] a) D. L. Allara, Nature 2005, 437, 638–639; b) J. V. Barth, G. Cos-
tantini, K. Kern, Nature 2005, 437, 671–679.
[2] For reviews, see: a) D. M. Cyr, B. Venkataraman, ; G. W. Flynn,
Chem. Mater. 1996, 8, 1600–1615; b) S. De Feyter, F. C. De Schryv-
er, Chem. Soc. Rev. 2003, 32, 139–150; c) S. De Feyter, J. Hofkens,
M. Van der Auweraer, R. J. M. Nolte, K. Müllen, F. C. Schryver,
Chem. Commun. 2001, 585–592; d) S. De Feyter, F. C. De Schryver,
J. Phys. Chem. B 2005, 109, 4290–4302.
Received: July 7, 2006
Published online: December 13, 2006
Chem. Eur. J. 2007, 13, 2311 – 2319
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2319