One building block, two different supramolecular surface-confined patterns: Concentration in control at the solid-liquid interface

Article abstract of DOI:10.1002/anie.200705322

(Figure Presented) New tool for surface modification: The 2D pattern formed by physisorbed dehydrobenzoannulene molecules on a graphite surface depends on their concentration in solution. The concentration dependence is directly related to the difference in stability between the linear andthe honeycomb polymorphs andtheir respective molecular densities (see picture).

Full text of DOI:10.1002/anie.200705322

Communications
DOI: 10.1002/anie.200705322
Self-Assembly on Surfaces
One Building Block, Two Different Supramolecular Surface-Confined
Patterns: Concentration in Control at the Solid–Liquid Interface**
Shengbin Lei, Kazukuni Tahara, Frans C. De Schryver, Mark Van der Auweraer, Yoshito Tobe,*
and Steven De Feyter*
Self-assembly of nanometer-sized building blocks at surfaces
and interfaces is of increasing interest for nanotechnology
research.[1,2] Recently surface-confined two-dimensional (2D)
molecular networks, especially those with void spaces, so-
called “2Dporous networks”, have attracted much atten-
tion.[3,4] The porous networks are typically sustained through
hydrogen bonds,[5] metal–ligand coordination,[6] or even van
der Waals interactions.[7] Cavity sizes ranging from 1 nm up to
5 nm have been reported for single- or multicomponent
molecular systems. These 2Dporous networks are used as
hosts to immobilize functional units as guest molecules in a
repetitive and spatially ordered arrangement.[8]
The interplay of adsorbate–adsorbate and adsorbate–
substrate interactions is crucial for the outcome of surface-
confined self-assembling processes.[10] Other factors such as
temperature and the nature of the solvent (the latter aspect
obviously only at the solid–liquid interface) play important
roles as well.[5e,11,12] Surprisingly, the effect of solute concen-
tration on the self-assembly of physisorbed systems has never
Scheme 1. a) Chemical structure of alkoxylated dehydrobenzo[12]annu-
lenes (DBA-OCn). b,c) Space-filling models of the conformations of
been probed systematically to the best of our knowledge,[13] in
contrast to chemisorbed systems.[14]
DBA-OC16 in a porous honeycomb pattern (b) andin a close-packed
linear pattern (c). In the latter case, two alkoxy chains are desorbed.[9]
We report a systematic study on the concentration-
dependent formation of surface-confined 2Dnetworks at
the interface of highly oriented pyrolitic graphite (HOPG)
and 1,2,4-trichlorobenzene (TCB). The building blocks are
alkoxylated dehydrobenzo[12]annulenes (DBAs)
(Scheme 1). They have been reported to form 2Dporous
networks when the alkoxy chain is less than twelve carbon
atoms long.[15] DBAs with longer alkoxy chains preferentially
form close-packed linear structures. Here we show that by
adjusting the DBA concentration in solution, the ratio of the
two polymorphs can be controlled such that either a regular
2Dporous honeycomb network (at low concentrations) or a
dense-packed linear network (at high concentrations) is
formed. The concentration dependency of the two poly-
morphs has been probed systematically and is modeled. A
practical outcome of this study is that, thanks to the
“concentration-in-control” concept, appropriate alkylated
compounds can be used to form surface-confined 2Dnano-
porous monolayers, the pore size of which is determined by
the length of the alkoxy chains. The formation of honeycomb
networks with a giant pore size of 5.4 nm is demonstrated, and
this is by no means an absolute limit.
[*] Dr. S. Lei, Prof. Dr. F. C. De Schryver, Prof. Dr. M. Van der Auweraer,
Prof. Dr. S. De Feyter
Division of Molecular andNanomaterials
Department of Chemistry
andINPAC—Institute of Nanoscale Physics andChemistry
Katholieke Universiteit Leuven (K.U. Leuven)
Celestijinenlaan 200F, 3001 Leuven (Belgium)
Fax: (+32)16-327-990
E-mail: steven.defeyter@chem.kuleuven.be
As an example, Figure 1 shows typical scanning tunneling
microscopy (STM) images, recorded at the solid–liquid
interface, of monolayers formed at two different concentra-
tions of DBA-OC16 in TCB. The bright features are the DBA
cores. The darker areas contain the alkoxy chains or, when a
honeycomb network is formed, adsorbed mobile solvent
molecules. The insets highlight the main structural motifs.
Within a certain concentration range, the surface pattern
changes from a dense-packed linear motif to a porous 2D
honeycomb system when the concentration is decreased. The
molecular models (Figure 1c) reflect the intermolecular
Dr. K. Tahara, Prof. Dr. Y. Tobe
Division of Frontier Materials Science
Graduate School of Engineering Science, Osaka University
Toyonaka, Osaka 560-8531 (Japan)
E-mail: tobe@chem.es.osaka-u.ac.jp
[**] This research has been supportedby the Belgian Federal Science
policy through IAP 6/27, the Fundfor Scientific Research—Flanders
(FWO), anda Grant-in-Aidfor Scientific Research from the Ministry
of Education, Culture, Sports, Science, and Technology (Japan). The
authors thank Shuhei Furukawa for fruitful discussions.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
2964
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 2964 –2968

Angewandte
Chemie
Figure 1. A series of STM images obtainedat the 1,2,4-trichloroben-
zene/graphite interface of monolayers formedat two DBA-OC 16
concentrations: a) 1.1”10ꢀ4 molLꢀ1 (Vbias =1.01 V, Iset =220 pA) and
b) 5.7”10ꢀ6 molLꢀ1 (Vbias =0.80 V, Iset =170 pA). These images show
the typical transition from a close-packedlinear pattern to a nano-
porous honeycomb pattern when the DBA concentration is decreased.
The inserts in (a) and(b) show high-resolution images of linear
(Vbias =ꢀ0.67 V, Iset =490 pA) andhoneycomb patterns ( Vbias =ꢀ0.58 V,
Figure 2. Honeycomb structure formedby DBA-OC 20 at the TCB/
graphite interface at 2.4”10ꢀ6 molLꢀ1 (Vbias =0.75 V, Iset =36 pA). The
size of the cavities reaches 5.4 nm. A line profile along the white line
is shown below the STM image.
Iset =670 pA), respectively. c) Scheme illustrating the structure andthe
transition of linear-type to honeycomb-type patterns.
chains, an apparently exponential decrease of the surface
coverage of the honeycomb pattern is observed when the
DBA concentration is increased. The longer the alkoxy
chains, the more pronounced the concentration-dependent
decrease of the surface coverage of the honeycomb pattern.
The honeycomb pattern is observed only in a very small range
of concentrations for DBA-OC18. For DBA-OC20, it is even
interactions. This concentration-dependent behavior proved
to be general for the alkoxy-substituted DBA derivatives,
though the concentration range for which this transition is
observed depends on the length of the alkoxy chains: DBA-
OC12 forms almost exclusively surface-
confined honeycomb structures up to a
concentration of 7.4 ” 10ꢀ4 molLꢀ1, while
concentrations below 2.4 ” 10ꢀ6 molLꢀ1
are absolutely necessary to create honey-
comb patterns with DBA-OC20. The
repeating period for the latter honeycomb
network is 6.3 nm and the pore diameter
reaches 5.4 nm, which is to the best of our
knowledge one of the largest values ever
reported (see Figure 2). These extremely
large nanocavities are supposed to be able
to trap large guest molecules or supra-
molecular assembles by means of host–
guest interactions.[16]
The dependence of the surface cover-
age of 2Dporous networks on the DBA
concentration and alkoxy chain length is
shown in Figure 3a and b. The honeycomb
surface coverage of DBAs with short
alkoxy chains, that is, DBA-OC12 and
DBA-OC14, apparently decreases linearly
with increasing DBA concentration for
the concentration range tested (from 7.1 ”
10ꢀ4 molLꢀ1 to 3.6 ” 10ꢀ5 molLꢀ1 and 6.4 ”
10ꢀ4 molLꢀ1 to 6.4 ” 10ꢀ6 molLꢀ1 for
DBA-OC12 and -OC14, respectively.).
Figure 3. a) Dependence of the surface coverage (q) of the honeycomb pattern on the DBA
concentration. b) Plot of the dependence of the honeycomb surface coverage of DBA-OC18.
c) Plot of ln{Yh/[DBA]} versus ln{(1ꢀYh)/[DBA]}. The thick solidlines show the linear fits.
&
d) The dependency of slope m ( experimental; & expectedfrom the area per unit cell (see
*
Table S1 in the Supporting Information)) andintercept ln K ( ) on the alkoxy chain length n.
However, for DBAs with longer alkoxy The solidlines are linear fits of m andln K.
Angew. Chem. Int. Ed. 2008, 47, 2964 –2968
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 2965

Communications
ð3Þ
ð4Þ
mꢀh
¼
mꢀ0h þ k T lnYh
impossible to obtain a full concentration-dependence curve
since the assembly changes from a close-packed linear pattern
to a submonolayer honeycomb pattern in a very small
concentration range (4.8 ” 10ꢀ6 molLꢀ1 to 2.4 ” 10ꢀ6 molLꢀ1).
This is in line with an increased tendency for close-packing for
DBAs with longer alkoxy chains, as observed previously.[15a,17]
Self-assembly at the solid–liquid interface is controlled by
the interplay of adsorbate–adsorbate, adsorbate–substrate,
adsorbate–solvent, and solvent–substrate interactions. The
assembly process is dynamic and depends on the adsorption–
desorption equilibrium. The building blocks, alkoxylated
DBAs, are quite unique. The six alkoxy chains attached to
the triangular core are extended radially, and the distance
between two alkoxy chains on the same edge is 0.9 nm, which
is nearly perfect for interdigitation of the alkoxy chains.[15,18]
In the honeycomb pattern, the six alkoxy chains on the rim of
the DBA cores spread out radially and are interdigitated with
those of neighboring molecules; this maximizes both the
adsorbate–adsorbate and adsorbate–substrate interactions. In
the close-packed linear structure, normally only four alkoxy
chains on two edges of the triangular core are adsorbed on the
surface (Scheme 1c and Figure 1c), and the remaining two
chains are extended into the solution phase.[15] This decreases
both the adsorbate–adsorbate and adsorbate–substrate inter-
actions at the level of single molecule. A rough estimation of
these energy considerations is shown in Table S1 in the
Supporting Information. The porous honeycomb motif is
favored at the level of a single molecule, taking into account
the balance of the van der Waals interactions with the
substrate and between interdigitating alkoxy chains and the
desolvation of the alkoxy chains. However, considering the
total free energy gain of the system taking into account the
surface density of adsorbed molecules, a close packing is more
favorable.
ꢀ
ꢀ0
ml ¼ ml þ k T lnYl
drawn in analogy, where Yh and Yl are the fractions of the
monolayer area occupied by the honeycomb and linear
patterns, respectively.
Combining Equations (1)–(4) yields Equations (5) and (6)
hꢀl
Yh
Yl
h
¼ K½DBAꢁ
ð5Þ
ð6Þ
l
ð =h
Þ
ꢀ
ꢁ
lꢀh
l
ꢀmꢀ0h þ ð =hÞ mꢀ0l ꢀ mꢀ0sol
h
K ¼ exp
:
k T
At close to complete coverage Equation (7) holds, which
implies Equation (8).
Yh þ Yl ¼ 1
ð7Þ
ð8Þ
ꢂ
ꢃ
ꢂ
ꢃ
Yh
1ꢀYh
ln
¼ m ln
þ ln K , with m ¼ l=h
½DBAꢁ
½DBAꢁ
ꢄ
ꢅ
ꢄ
ꢅ
Yh
½DBAꢁ
1ꢀYh
Plotting ln
versus ln
(Figure 3c) allows us to
½DBAꢁ
determine experimentally the slope m, which is the ratio of
the molecular densities of the linear and the honeycomb
pattern.
The intercept, given in Equation (9) reflects the adsor-
ꢀ
ꢁ
lꢀh
l
ꢀmꢀ0h þ ð =hÞ mꢀ0l ꢀ mꢀ0sol
h
ð9Þ
ln K ¼
k T
bate–adsorbate, adsorbate–substrate, and solute–solvent
interactions. The dependency of m and lnK on the length of
the alkoxy chains of the DBA derivatives is plotted in
Figure 3d.
To understand the effect of DBA concentration on the
polymorphs, we evaluated the relevant adsorption–desorp-
tion processes under thermodynamic control. The adsorp-
tion–desorption equilibrium determines the surface coverage
ratio of the honeycomb and linear patterns. Suppose one
fraction of the surface is covered by the honeycomb pattern
and another fraction by the linear pattern. Both physisorbed
patterns are in equilibrium with a solution with concentration
[DBA], and we suppose there are no specific interactions
The experimental results match the theoretical model
ꢄ
ꢅ
Yh
½DBAꢁ
very well. A nearly linear dependency of ln
on
ꢄ
ꢅ
1ꢀYh
ln
is detected for all the DBAs. The value of m
½DBAꢁ
determined by a linear fit of the experimental results is
around 2 and increases linearly with increasing alkoxy chain
length, consistent with the expected increase of the pore
fraction for DBA derivatives with longer alkoxy chains. The
experimentally determined intercept lnK shows a linear
dependency on the length of the alkoxy chain length as well,
consistent with the expected monotonic change in the
adsorbate–adsorbate, adsorbate–substrate, and solute–sol-
vent interactions upon increasing the alkoxy chain length.
between the molecules. Hence, m = m and m = m , where m ,
¯
¯
¯
¯
¯
h
sol
l
sol
h
m , and msol are the chemical potentials of a DBA molecule in
¯
¯
l
the honeycomb pattern, in the linear pattern, and in solution,
respectively. The conversion of a unit area of linear patterns
into a unit area of honeycomb patterns at equilibrium is
described by Equation (1), where l and h are the number of
To understand the physical meaning of the intercept we
l mꢀl ¼ h mꢀh þ ðlꢀhÞmꢀsol
ð1Þ
can change the expression to Equation (10). Here l(m0ꢀm0 )
¯
¯
l
sol
h k T lnK ¼ lðmꢀ0l ꢀmꢀs0olÞꢀhðmꢀ0hꢀmꢀ0sol
Þ
molecules per unit area in the linear and honeycomb patterns,
respectively. For dilute solutions Equation (2) holds, where
ð10Þ
and h(m0ꢀm0 ) are the change of free energy caused by the
¯
¯
sol
h
mꢀsol ¼ mꢀ0sol þ k T ln½DBAꢁ
ð2Þ
adsorption of a unit area of linear and honeycomb patterns,
respectively, at the interface under standard conditions. Thus
hkTlnK reflects the difference in stability between both
polymorphs. The values of hkTlnK calculated from the
m0 corresponds to m at standard conditions (ideal molar
¯
¯
sol
sol
solution). The relationships in Equations (3) and (4) can be
2966 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 2964 –2968

Angewandte
Chemie
effect on the quantitative data treatment as the concentration
experimental result are included in Table S1 in the Supporting
Information (in parentheses in the bottom line). For DBA-
OC12, the two polymorphs are predicted to be nearly equally
stable under standard conditions, while for those DBA
derivatives with longer alkoxy chains, the linear pattern is
that favored.
intervals were much larger. For each concentration, typically ten to
fifteen large scale images (120 ” 120 nm2) were recorded at different
locations. Subsequently, the average honeycomb surface coverage
and standard deviation was determined.
Received: November 20, 2007
The concentration dependency of the self-assembly of
DBAs can therefore be understood as arising from the
different stabilities and molecular densities of the two
polymorphs. The energy difference between both polymorphs
determines the systemꢀs sensitivity towards concentration:
the larger the energy difference, the more dramatic the
concentration dependence.
Published online: March 12, 2008
Keywords: interfaces · nanostructures ·
.
scanning probe microscopy · supramolecular chemistry
[1] a) J. V. Barth, G. Costantini, K. Kern, Nature 2005, 437, 671;
b) H. Wu, Y. Song, S. Du, H. Liu, H. Gao, L. Jiang, D. Zhu, Adv.
Mater. 2003, 15, 1925; c) V. Balzani, A. Credi, M. Venturi,
Nanotoxicology Today 2007, 2(2), 18; d) E. R. Kay, D. A. Leigh,
F. Zerbetto, Angew. Chem. 2007, 119, 72; Angew. Chem. Int. Ed.
2007, 46, 72; e) N. Katsonis, T. Kudernac, M. Walko, S. J.
van der Molen, B. J. van Wees, B. L. Feringa, Adv. Mater. 2006,
18, 1397; f) L. Piot, D. Bonifazi, P. Samori, Adv. Funct. Mater.
2007, 17, 3689.
[2] a) S. De Feyter, A. Gesqui›re, M. M. Abdel-Mottaleb, P. C. M.
Grim, F. C. De Schryver, C. Meiners, M. Sieffert, S. Valiyaveettil,
K. Mullen, Acc. Chem. Res. 2000, 33, 520; b) M. Hietschold, M.
Lackinger, S. Griessl, W. M. Heckl, T. G. Gopakumar, G. W.
Flynn, Microelectron. Eng. 2005, 82, 207; c) F. Chiaravalloti, L.
Gross, K. H. Rieder, S. M. Stojkovic, A. Gourdon, C. Joachim, F.
Moresco, Nature Mat. 2007, 6, 30; d) J. K. Gimzewski, C.
Joachim, R. R. Schlittler, V. Langlais, H. Tang, I. Johannsen,
Science 1998, 281, 531.
[3] a) S. Stepanow, N. Lin, J. V. Barth, K. Kern, J. Phys. Chem. B
2006, 110, 23472; b) Y. Ye, W. Sun, Y. Wang, X. Shao, X. Xu, F.
Cheng, J. Li, K. Wu, J. Phys. Chem. C 2007, 111, 10138; c) D.
Payer, A. Comisso, A. Dmitriev, T. Strunskus, N. Lin, C. Wöll, A.
DeVita, J. V. Barth, K. Kern, Chem. Eur. J. 2007, 13, 3900; d) R.
Otero, M. Schöck, L. M. Molina, E. Lægsgaard, I. Stensgaard, B.
Hammer, F. Besenbacher, Angew. Chem. 2005, 117, 2310;
Angew. Chem. Int. Ed. 2005, 44, 2270.
[4] a) N. Lin, S. Stepanow, F. Vidal, J. V. Barth, K. Kern, Chem.
Commun. 2005, 1681; b) S. Stepanow, N. Lin, F. Vidal, A. Landa,
M. Ruben, J. V. Barth, K. Kern, Nano Lett. 2005, 5, 901.
[5] a) M. Stöhr, M. Wahl, C. H. Galka, T. Riehm, T. A. Jung, L. H.
Gade, Angew. Chem. 2005, 117, 7560; Angew. Chem. Int. Ed.
2005, 44, 7394; b) S. J. H. Griessl, M. Lackinger, F. Jamitzky, T.
Markert, M. Hietschold, W. M. Heckl, J. Phys. Chem. B 2004,
108, 11556; c) M. Ruben, D. Payer, A. Landa, A. Comisso, C.
Gattinoni, N. Lin, J. P. Collin, J. P. Sauvage, A. De Vita, K. Kern,
J. Am. Chem. Soc. 2006, 128, 15644; d) J. A. Theobald, N. S.
Oxtoby, M. A. Phillips, N. R. Champness, P. H. Beton, Nature
2003, 424, 1029; e) G. Pawin, K. L. Wong, K. Y. Kwon, L. Bartels,
Science 2006, 313, 961.
In summary, the experiments and modeling reveal a clear
dependence of the nature of the physisorbed 2Dmolecular
surface pattern on the concentration of the building blocks in
solution. This can be associated with the well-known dilution
principle which has already been put into context with the
concentration dependence of surface self-assemblies pro-
duced “in vacuo”.[5a] The self-assembly occurs under thermo-
dynamic control. The concentration dependence is directly
related to the difference in stability between the linear and
the honeycomb polymorphs and their respective molecular
densities. The longer the alkoxy chains of the DBA deriva-
tives, the more sensitive the surface coverage of the porous
honeycomb structures is to the concentration of the building
blocks. These results not only provide insight in the thermo-
dynamics of the formation of surface-confined 2D(porous)
nanopatterns at the solid–liquid interface, but also some
general guidelines for the successful control of polymorphism
at such surfaces. Creating different surface patterns in a
controlled way merely by adjusting the concentration of the
building blocks in the liquid reservoir has the potential to be a
powerful and simple surface-modification tool. In this specific
study, concentration control of the appropriate alkoxy-chain-
containing compounds supports the formation of 2Dnano-
porous structures. The diameter of the pore reaches a value of
5.4 nm, which is the largest value reported so far at a liquid–
solid interface but by no means a fundamental limit. We
would like to stress that this “concentration-in-control”
concept is general and not restricted to this particular set of
molecules. This is also a cautionary lesson: concentration
effects should always be probed systematically at the solid–
liquid interface especially in the case of polymorphism.
Experimental Section
[6] a) S. Stepanow, M. Lingenfelder, A. Dmitriev, H. Spillmann, E.
Delvigne, N. Lin, X. Deng, C. Cai, J. V. Barth, K. Kern, Nature
Mat. 2004, 3, 229; b) M. A. Lingenfelder, H. Spillmann, A.
Dmitriev, S. Stepanow, N. Lin, J. V. Barth, K. Kern, Chem. Eur. J.
2004, 10, 1913.
[7] a) G. Schull, L. Douillard, C. Fiorini-Debuisschert, F. Charra, F.
Mathevet, D. Kreher, A.-J. Attias, Nano Lett. 2006, 6, 1360; b) D.
BlØger, D. Kreher, F. Mathevet, A.-J. Attias, G. Schull, A. Huard,
L. Douillard, C. Fiorini-Debuischert, F. Charra, Angew. Chem.
2007, 119, 7548; Angew. Chem. Int. Ed. 2007, 46, 7404; c) H.
Spillmann, A. Kiebele, M. Stöhr, T. A. Jung, D. Bonifazi, F.
Cheng, F. Diederich, Adv. Mater. 2006, 18, 275; d) D. Bonifazi,
A. Kiebele, M. Stöhr, F. Cheng, T. Jung, F. Diederich, H.
Spillmann, Adv. Funct. Mater. 2007, 17, 1051.
The DBA derivatives were synthesized according to previously
[12a]
reported methods.
For STM measurements, the DBAs were
dissolved in 1,2,4-trichlorobenzene (TCB) at a concentration of
1 mgmLꢀ1. Solutions of lower concentration were obtained by
sequential dilution of this original solution. An aliquot of 8 to 9 mL
of solution was applied to the surface of a piece of freshly cleaved
highly oriented pyrolytic graphite (HOPG, grade ZYB, Advanced
Ceramics Inc., Cleveland, USA). STM measurements were per-
formed with a PicoSPM (Agilent) instrument operating in constant-
current mode using a mechanically cut Pt/Ir (80/20) tip. All experi-
ments were carried out at 218C. Solvent evaporation led to a
maximum increase in DBA concentration of about 15% during a
measuring period, which lasted 30 to 40 min. This had only a slight
Angew. Chem. Int. Ed. 2008, 47, 2964 –2968
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 2967

Communications
[8] a) S. Stepanow, N. Lin, J. V. Barth, K. Kern, Chem. Commun.
[12] Z. Li, B. Han, L. J. Wan, T. Wandlowski, Langmuir 2005, 21,
6915.
2006, 2153; b) J. A. Theobald, N. S. Oxtoby, N. R. Champness,
P. H. Beton, T. J. S. Dennis, Langmuir 2005, 21, 2038; c) S. J. H.
Griessl, M. Lackinger, F. Jamitzky, T. Markert, M. Hietschold,
W. M. Heckl, Langmuir 2004, 20, 9403; d) J. Lu, S. B. Lei, Q. D.
Zeng, S. Z. Kang, C. Wang, L. J. Wan, C. L. Bai, J. Phys. Chem. B
2004, 108, 5161.
[13] K. Kim, K. E. Plass, A. J. Matzger, Langmuir 2005, 21, 647.
[14] F. Schreiber, Prog. Surf. Sci. 2000, 65, 151.
[15] a) K. Tahara, S. Furukawa, H. Uji-I, T. Uchino, T. Ichikawa, J.
Zhang, W. Mamdouh, M. Sonoda, F. C. De Schryver, S.
De Feyter, Y. Tobe, J. Am. Chem. Soc. 2006, 128, 16613; b) S.
Furukawa, H. Uji-I, K. Tahara, T. Ichikawa, M. Sonoda, F. C.
De Schryver, Y. Tobe, S. De Feyter, J. Am. Chem. Soc. 2006, 128,
3502.
[9] a) Y. Kaneda, M. E. Stawasz, D. L. Sampson, B. A. Parkinson,
Langmuir 2001, 17, 6185; b) K. Perronet, F. Charra, Surf. Sci.
2004, 551, 213.
[10] a) S. De Feyter, F. De Schryver, Top. Curr. Chem. 2005, 258, 205;
b) S. B. Lei, C. Wang, S. X. Yin, H. N. Wang, F. Xi, H. W. Liu, B.
Xu, L. J. Wan, C. L. Bai, J. Phys. Chem. B 2001, 105, 10838.
[11] a) L. Kampschulte, M. Lackinger, A. K. Maier, R. S. K. Kishore,
S. Griessl, M. Schmittel, W. M. Heckl, J. Phys. Chem. B 2006,
110, 10829; b) M. Lackinger, S. Griessl, W. M. Heckl, M.
Hietschold, G. W. Flynn, Langmuir 2005, 21, 4984; c) X. H.
Kong, K. Deng, Y. L. Yang, Q. D. Zeng, C. Wang, J. Phys. Chem.
C 2007, 111, 9235.
[16] These voids can host molecular clusters of guest molecules with
well-defined number and geometry; these results will be
reported elsewhere.
[17] P. Wu, Q. Zeng, S. Xu, C. Wang, S. Yin, C. L. Bai, ChemPhys-
Chem 2001, 2, 750.
[18] K. Tahara, C. A. Johnson II, T. Fujita, M. Sonoda, F. C.
De Schryver, S. De Feyter, M. M. Haley, Y. Tobe, Langmuir
2007, 23, 10190.
2968 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 2964 –2968