Tailoring Large Pores of Porphyrin Networks on Ag(111) by Metal–Organic Coordination

Article abstract of DOI:10.1002/chem.201602154

The engineering of nanoarchitectures to achieve tailored properties relevant for macroscopic devices is a key motivation of organometallic surface science. To this end, understanding the role of molecular functionalities in structure formation and adatom

Full text of DOI:10.1002/chem.201602154

DOI: 10.1002/chem.201602154
Full Paper
&
Porphyrins |Hot Paper|
Tailoring Large Pores of Porphyrin Networks on Ag(111) by Metal–
Organic Coordination
+
[a]
+ [a, b]
[a]
[c]
[c, d]
Felix Bischoff , Yuanqin He ,
Knud Seufert, Daphnꢀ Stassen, Davide Bonifazi,*
[a]
[a]
Johannes V. Barth, and Willi Auwꢁrter*
Abstract: The engineering of nanoarchitectures to achieve
tailored properties relevant for macroscopic devices is a key
motivation of organometallic surface science. To this end,
understanding the role of molecular functionalities in struc-
ture formation and adatom coordination is of great impor-
tance. In this study, the differences in formation of Cu-medi-
ated metal–organic coordination networks based on two
pyridyl- and cyano-bearing free-base porphyrins on Ag(111)
are elucidated by use of low-temperature scanning tunnel-
ing microscopy (STM). Distinct coordination networks evolve
via different pathways upon codeposition of Cu adatoms.
The cyano-terminated module directly forms 2D porous net-
works featuring fourfold-coordinated Cu nodes. By contrast,
the pyridyl species engage in twofold coordination with Cu
and a fully reticulated 2D network featuring a pore size ex-
2
ceeding 3 nm only evolves via an intermediate structure
based on 1D coordination chains. The STM data and comple-
mentary Monte Carlo simulations reveal that these distinct
network architectures originate from spatial constraints at
the coordination centers. Cu adatoms are also shown to
form two- and fourfold monoatomic coordination nodes
with monotopic nitrogen-terminated linkers on the very
same metal substrate—a versatility that is not achieved by
other 3d transition metal centers but consistent with 3D co-
ordination chemistry. This study discloses how specific mo-
lecular functionalities can be applied to tailor coordination
architectures and highlights the potential of Cu as coordina-
tion center in such low-dimensional structures on surfaces.
[6]
Introduction
by surfaces in a controlled ultrahigh-vacuum (UHV) setting.
Also in this 2D scenario, the functionality of metal–organic co-
ordination networks is represented by the combination of
metal nodes, offering, for example, active sites for energy con-
Supramolecular coordination chemistry is a vivid field of re-
search as the combination of organic or metal–organic ligands
and metal centers yields structures and properties that are not
achievable by the individual building blocks. Recent examples
highlighting the potential of metal–organic compounds in-
[7]
version chemistry, and the ligands, dictating the pore sizes,
[8]
the confinement of adsorbates, and the (magnetic) coupling
[9]
between the metals. Specifically, the coordination number
and symmetry at the nodes are decisive for the topology of
[
1]
clude reports alluding to information storage and process-
[
2]
[3]
[4]
[5]
[10]
ing, energy storage, catalysis, and molecular electronics.
the resulting metal-directed architectures. Despite the many
[6-10]
Considerable efforts have been dedicated to adapting design
principles from such three-dimensional (3D) supramolecular
structures to a two-dimensional (2D) environment represented
reports on surface-anchored metal–organic architectures,
strategies to engineer and deliberately tailor assemblies still
need to be further developed and refined. For example, ex-
tended 2D networks featuring a grid-like structure with square
[11]
+
+
pores and mononuclear nodes are rarely reported. In this re-
spect, molecules offering fourfold symmetry might yield ad-
vantages compared to the frequently applied ditopic linear
linkers. Hereby, tetrapyrroles as porphyrins are ideal candi-
dates, which proved to be stable and versatile building blocks
[
a] F. Bischoff, Y. He, Dr. K. Seufert, Prof. J. V. Barth, Prof. W. Auwꢀrter
Physik Department E20, Technische Universitꢀt Mꢁnchen
8
5748 Garching (Germany)
E-mail: wau@tum.de
+
[b] Y. He
Institute for Advanced Study, Technische Universitꢀt Mꢁnchen
5748 Garching (Germany)
[12]
for self-assembled molecular structures on surfaces. The cen-
tral macrocycles, which host two hydrogens or a metal center,
add functionality to the system, as they can be used as molec-
8
[c] D. Stassen, Prof. D. Bonifazi
Department of Chemistry and Namur Research College
University of Namur (Belgium)
[13]
ular switches, can undergo metalation reactions directly on
[
14]
[15]
[d] Prof. D. Bonifazi
a surface, and have potential for heterogeneous catalysis
School of Chemistry, Cardiff University
Main Building, Park Place, Cardiff CF10 3AT (UK)
E-mail: bonifazid@cardiff.ac.uk
[16]
and spintronics. Most importantly, the tetrapyrrole macrocy-
cle can be substituted by a wide variety of terminal moieties,
offering vast possibilities to steer intermolecular and metal–or-
+
[
] These authors contributed equally.
[11a,12a–c,17]
ganic interactions.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201602154.
Chem. Eur. J. 2016, 22, 1 – 10
1
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
[
11c,14b]
Herein we present a low-temperature scanning tunneling
microscopy (STM) study that compares the Cu-directed assem-
blies of two de novo-synthesized porphyrins, functionalized
with biphenylene–cyano (2H–TPCN) and phenylene–pyridyl
substituents (2H–TPyPP), respectively, on Ag(111). Although
both molecules feature nitrogen-terminated ligands and as-
semble into similar organic arrays, they respond markedly dif-
ferently to the copper atoms. TPCN directly forms 2D metal–or-
ganic networks with small pores and fourfold coordination
nodes, whereas TPyPP follows a hierarchic pathway from 1D
metal coordination chains to an open porous 2D metal–organ-
ic network with linear twofold-coordinated metal centers.
Monte Carlo simulation and comparison of Co- and Cu-direct-
ed networks, indicate that the origin of the two- and fourfold
coordination motif results from an interplay between the steric
hindrance at the coordination center and the inherent coordi-
nation properties of Cu.
of 2H–TPCN on Ag(111) and BN/Cu(111),
whereas 2H–
TPyPP is addressed in this study for the very first time. Com-
pared to commercially available tetraphenyl- (TPP) or tetrapyri-
dylporphyrins (TPyP) featuring only one phenyl or pyridyl unit
in each meso substituent, these novel modules introduce an
additional degree of conformational freedom as the terminal
2
ring R can rotate around the CÀC single bond connecting it
1
to R (cf. Figure 1a and e). Upon surface confinement, this en-
hanced molecular flexibility strongly influences the molecular
self-assembly and coordination characteristics (see below). The
adsorption of TPP and TPyP species on Ag(111) induces
a saddle-shaped macrocycle deformation where the terminal
[13,18]
rings are rotated typically 50–608 out of the surface plane.
Saddle-shape deformations—induced by steric hindrance be-
1
tween the macrocycle and rotated moieties R —are also ex-
1
pected for both 2H–TPCN and 2H–TPyPP. However, as R acts
2
as a spacer, geometrically decoupling the terminal rings R
2
from the macrocycle, a rather parallel alignment of R with the
Ag(111) surface is anticipated as both, individual benzene and
[
19]
Results
pyridyl rings adsorb on Ag(111) in a planar fashion.
Figure 1b, c, f and g show high-resolution STM images of
H–TPCN and 2H–TPyPP/Ag(111) representing occupied states.
2
H–TPCN and 2H–TPyPP modules on Ag(111)
2
The porphyrins investigated in this study are tetrapyrrolic mac-
rocycles substituted at all four meso positions either with bi-
phenylene–cyano or with phenylene–pyridyl moieties. Structur-
al models of these two porphyrin derivatives, namely tetra[(4-
cyanophenyl)phen-4-yl]porphyrin (2H–TPCN) and tetra[(4-pyri-
dylphenyl)phen-4-yl]porphyrin (2H–TPyPP) are depicted in Fig-
ure 1a and e (see also the Supporting Information). We recent-
ly reported on the successful deposition and characterization
Both species present four peripheral lobes associated to the
meso substituents and a donut shape that is assigned to the
macrocycle. The latter shows twofold symmetry and appears
with a depression in the center, as observed previously for
[13]
free-base TPP on Ag substrates. The elongated meso sub-
stituents of 2H–TPCN are reflected in the larger apparent size
of the molecule and an increased intermolecular distance com-
pared to 2H–TPyPP (see below). Nevertheless, the overall ap-
Figure 1. Structural models of the porphyrin derivatives (a,e). The higher parts of rotated molecular moieties are highlighted in orange for better comparison
to the magnification on single molecules (b,f) within self-assembled, dense-packed islands on Ag(111) (c,g). The green lines highlight the molecular axis
through the upward rotated pyrroles. A single molecule is outlined in red in (c) and (g) as a guide to the eye and the blue squares indicate the unit cells that
include one molecule for both compounds. The substrate dense-packed directions are represented by yellow lines. Scan parameters: b) U=À1 V, I=30 pA;
c) U=À1 V, I=70 pA; f,g) U=À1 V, I=0.1 nA. d,h) Sketch of the assembly structure. The models were created with HyperChem and the molecular dimen-
sions were extracted after geometry optimization of the free molecule within the semi-empirical AM1 framework.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
2
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
pearance of both species is rather similar, as the cyano group
Formation of metal–organic coordination networks
[20]
does not contribute considerably to the STM contrast. Based
on a comparison of submolecular features, presented in Fig-
ure 1b and f with structural models, a tentative conformation
of the meso substituents’ orientation can be inferred. Whereas
To investigate the response of the porphyrin species to metal
adatoms, Cu was deposited with submonolayer, monomolecu-
lar coverages at room temperature. Figure 2 shows the fully re-
ticulated metal–organic coordination networks and the corre-
sponding structural models. For TPCN, a highly regular porous
network with a rectangular unit cell of size c=(21.9 Æ 0.5) ꢃ
and d=(24.0 Æ 0.5) ꢃ evolves, featuring domains that extend
over hundreds of square nanometers with a low defect density
(Figure 2a). The long-range order and the symmetry of the
network are reflected in the autocorrelation plot and a sharp
fast Fourier transform (FFT) pattern (Figure 2b). High-resolution
images (Figure 2c) reveal details of the network structure, with
2
1
R looks disk-like, indicating a planar adsorption, R appears as
2
a narrow bridge connecting the macrocycle and R . The asym-
1
metric appearance of R with respect to the axis through the
meso position of the macrocycle (white lines in Figure 1b and
1
f) provides an indication for the rotation of R . The upper part
of the phenyl group dominates the image contrast, in full
agreement with previously reported high-resolution STM data
(
see the Supporting Information of ref. [13]). Steric hindrance
1
between the rotated R and the pyrroles of the macrocycle
2
2
leads to its saddle-shape deformation. R appears as a broad
a pore size of approximately 86 ꢃ exposing bare Ag. Clearly,
protrusion symmetric with respect to the axis connecting op-
posing legs, in line with the contrast reported for terminal pyr-
every node—assigned to a Cu adatom—links four TPCN units
through their carbonitrile termini, resulting in a 1:1 stoichio-
metric ratio between TPCN and Cu. The projected N–Cu dis-
tance is (1.6 Æ 0.5) ꢃ, in agreement with previously reported
[21]
idyl groups adsorbed parallel to the Ag(111) surface.
We
2
thus conclude that the R rings are aligned approximately par-
allel to the surface. Note that the larger apparent height of the
legs compared to the macrocycle prevails only at small bias
voltages and thus is assigned to an electronic effect (cf. Fig-
ure S1 in the Supporting Information). As discussed below, the
adsorption geometry of the terminal pyridyl rings in TPyPP is
decisive for the distinct coordination behavior from TPyP, in
which the pyridyl group is rotated considerably out of the sur-
face plane.
[
22]
interfacial cyano–Cu coordination. Within the network, the
molecules appear slightly compressed compared to the organ-
ic phase; their aspect ratio deviates from unity, resulting in an
“X”-like shape.
This is also reflected in the rectangular unit cell, which dif-
fers from the square unit cell reported for Gd-coordinated
[14b]
TPCN networks.
The reduced symmetry might be induced
by the flexibility of the meso substituents combined with the
[23]
favorable hollow site absorption of Cu adatoms on Ag(111).
Indeed, a simple model overlay of the coordination network
onto a lattice representing the Ag(111) substrate demonstrates
that a highly regular, commensurate (8ꢄ80) structure can
evolve with Cu adatoms exclusively at hollow sites (see the
Supporting Information, Figure S5 for details). Apparently, the
energy gained by formation of the commensurate network—
enabled by the specific dimensions of 2H–TPCN—exceeds the
energy costs for deforming the molecule. As usual for 3d tran-
sition metals, the coordination center is not visualized in
Self-assembly of 2H–TPCN and 2H–TPyPP on Ag(111)
After room-temperature deposition on Ag(111), both modules
self-assemble into highly ordered, extended two-dimensional
islands (Figure 1c and g; corresponding structural models are
shown in Figure 1d and h). Both assemblies feature a square
unit cell with internal angles of (90 Æ 1)8 (marked in blue in
Figure 1c and g) with side lengths a =(20.4 Æ 0.5) ꢃ for 2H–
TPCN and b=(18.2 Æ 0.5) ꢃ for 2H–TPyPP. In addition, a dis-
tinct metastable structure characterized by a rhombic unit cell
can be achieved when depositing 2H–TPCN at high flux (see
the Supporting Information, Figures S2 and S3). For both por-
phyrin modules, the molecular axis through the two upward-
[24]
STM. However, an indirect fingerprint of the metal coordina-
tion is observed, as the coordinated terminal groups of TPCN
appear higher than those that are noncoordinated (see the
Supporting Information, Figure S4). For TPCN, fully reticulated
metal–organic coordination networks were obtained, coexist-
ing with dense-packed organic islands and large Cu clusters on
the Ag(111) support under all employed preparation condi-
tions. Thus, the yield for the metal–organic network formation
is not optimal at the given preparation temperature and
copper flux. However, additional architectures based on a si-
multaneous expression of metal–organic and organic bonding
bent pyrroles of the macrocycle (green lines in Figure 1a
and e) is either aligned with the < 1¯ 1¯ 2> or the < 1¯ 10> high-
symmetry directions of the Ag(111) lattice (green lines in Fig-
ure 1b and f). Whereas 2H–TPCN mostly aligns along < 1¯ 1¯ 2>,
[
18b]
as reported for Co–TPP/Ag(111),
no preference is discernible
[25]
for 2H–TPyPP. Despite these distinct azimuthal orientations in-
duced by the Ag(111) surface, site-specific interactions do not
prevail over lateral intermolecular interactions and the 2H–
TPCN and 2H–TPyPP arrays are not commensurate with the
underlying Ag(111) lattice, as revealed by bias-dependent
imaging and dI/dV spectroscopy (see the Supporting Informa-
tion, Figure S1). Both assemblies are stabilized by lateral non-
covalent interactions between neighboring nitrogen–phenyl-
ene groups.
motifs were never observed for TPCN and Cu. This was con-
firmed by the Monte Carlo simulations (see below).
For TPyPP, an extended metal–organic coordination network
also evolves upon exposure to Cu, characterized by a mole-
cule/Cu adatom ratio of 1:2 (Figure 2e and g). All four pyridyl
termini of 2H–TPyPP are connected with the adjacent mole-
cules by pyridyl–Cu–pyridyl coordination motifs, forming
a square unit cell with a side length of e=(28.2 Æ 0.5) ꢃ. The
projected N–Cu distance amounts to (1.9 Æ 0.5) ꢃ, in agree-
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
3
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 2. Formation of metal–organic networks upon deposition of Cu atoms. The blue squares indicate the unit cells and one molecular unit is outlined in
red as a guide to the eye in (c) and (g). For TPCN (a–d) every coordination node is surrounded by four molecules and the unit cell consists of one molecule
and one Cu atom. In contrast, the unit cell of TPyPP/Cu (e–h) consists of one molecule and two Cu atoms. The regularity of the metal–organic networks is re-
flected in autocorrelation plots (upper images in b and f) and sharp spots in FFT images (lower images in b and f). d,h) Model sketches of the networks. Dif-
ferences in the molecular appearance are assigned to the interaction of the macrocycle with Cu adatoms. The yellow stars represent the substrate’s dense-
packed directions. Scan parameters: a) U=0.7 V, I=50 pA; c) U=0.2 V, I=0.2 nA; e) U=0.9 V, I=80 pA; g) U=À0.2 V, I=80 pA.
[
21a,24a]
ment with the previous reports.
Similar head-on, twofold
archic pathway upon increasing the (local) density of Cu ada-
toms. After depositing small amounts of Cu adatoms onto
a submonolayer of 2H–TPyPP/Ag(111), a porous array appears
that is characterized by chain-like sub-structures (Figure 3a). A
close inspection reveals that it expresses simultaneously
metal–organic and organic bonding motifs like those described
in ref. [25]. As visualized in the corresponding structural model
(Figure 3c), TPyPP tectons are dense-packed in one direction
(organic bonding, marked with f) and form a head-on configu-
ration along the other direction (marked with e), which is as-
signed to a pyridyl–Cu–pyridyl coordination bond, in analogy
to the fully reticulated network (cf. Figure 2e). The network is
thus formed by 1D metal–organic chains that mutually interact
through lateral noncovalent interactions between neighboring
nitrogen–phenylene groups and follow the dense-packed sub-
strate directions. The structure features a rhomboid unit cell of
size e=(28.0 Æ 0.5) ꢃ and f=(18.2 Æ 0.5) ꢃ including an angle
of (60 Æ 1)8 and a TPyPP/Cu adatom ratio of 1:1.
Cu-mediated coupling motifs of pyridyl moieties are well-
[
11a,21a,b,24a,26]
known in surface-confined coordination chemistry.
Compared to the dense-packed organic arrays, TPyPP modules
within the metal–organic network are rotated by 458. Appa-
rently, the energy arising from metal coordination exceeds the
energy penalty from deviation from the original adsorption ori-
entation, underlining the weak site-specific molecule–substrate
interaction. The network domains extend over hundreds of
square nanometers and exhibit long-range regularity (see auto-
correlation plot and FFT pattern in Figure 2 f). However, high-
resolution STM data reveal that the pores vary in size and
shape (Figure 2 g). This local disorder is attributed to the
flexibility of both the pyridyl–Cu–pyridyl motif, which
features bond angles deviating from 1808, and the meso sub-
[
21a,b]
stituents.
Thus, a variety of pore shapes that deviate from
a perfect square can coexist, which classifies this architecture
[
27]
as a 2D short-range disordered crystalline network. The aver-
2
age pore size amounts to 340 ꢃ . To our knowledge, such
When the Cu dosage is increased and (locally) exceeds a 1:1
ratio of Cu adatoms to molecules, the organic bonds are trans-
formed into pyridyl–Cu–pyridyl coordination bonds and the
structure evolves into a fully reticulated 2D metal–organic net-
work. Figure 3b and the corresponding structural model in Fig-
ure 3d show the transition from 1D coordination chains to
a 2D coordination network. Both architectures coexist locally
and TPyPP can form coordination bonds from Cu to two,
three, or four adjacent molecules. Clearly, the meso substitu-
ents engaged in coordination bonds appear brighter than their
non-coordinated counterparts (cf. Figure 3a and Figure S4 in
a large area is unprecedented for homomolecular surface-sup-
ported porphyrin-based architectures. Consequently, the Cu-di-
rected TPyPP network might serve as a template to trap and
[
28]
order large adsorbates or even molecular aggregates. To this
end, the intrinsic flexibility opens perspectives for hosting and
sorting specific molecular guest species, enabling an adaptive
behavior of the pores, thus representing a two-dimensional an-
[
29]
alogue of a “soft porous crystal”. In contrast to the fully re-
ticulated TPCN coordination architecture that evolves directly
from the organic islands, the TPyPP coordination follows a hier-
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
4
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
TPCN the addition of small amounts of adatoms already leads
to the formation of a 2D coordination network (Figure 4 f). Im-
portantly, the simulations reflect all assemblies observed exper-
imentally and correctly reproduce a key difference between
TPyPP and TPCN, namely the fact that 1D metal–organic
chains emerge exclusively from TPyPP, even though they
would be allowed for TPCN.
Figure 3. For TPyPP, depending on the (local) Cu density, metal coordination
evolves in one or two directions. a) 1D coordination along one specific direc-
tion, indicated here by the black arrow labeled with “e”, coexists with organ-
ic interactions along “f”. Coordination will firstly be completed along one di-
rection before starting in another direction (as shown in b). c,d) Models. As
a guide to the eye, some molecules are outline in red. The yellow stars rep-
resent the substrate dense packed directions. Scan parameters:
a,b) U=À0.7 V, I=0.1 nA.
the Supporting Information), in line with the results discussed
for TPCN.
Monte Carlo modeling
The experimental results reveal striking differences in the for-
mation of metal–organic coordination networks and the corre-
sponding coordination motifs, despite the similarities of the
systems, that is, mononuclear Cu nodes coordinated to nitro-
gen of quasi-fourfold symmetric porphyrins on Ag(111). To ra-
tionalize the experimental findings, Monte Carlo modeling was
performed, as structure formation is known to be correctly re-
produced in such simple simulations for a variety of functional
Figure 4. a–d) Snapshots of the structure formation of porphyrin linkers with
metal atoms modeled in Monte Carlo simulations. a) 300 molecules and no
metal atoms. Without metal atoms both species yield the same assembly.
b) 300 TPyPP and 150 metal atoms. c) 300 TPyPP and 600 metal atoms.
d) 300 TPCN and 300 metal atoms. The lower-left inset in (a) highlights the
two interactions included in the simulations. The bottom-right insets in (a–
d) provide a magnification of the dominating pattern. e,f) Phase transition
diagrams for TPyPP and TPCN respectively, as a function of the relative
number of units (no. of metal atoms)/(no. of molecules) and the ratio of in-
[
17,30]
[31]
O C
teraction energies E /E . Due to the finite island sizes, the transition to the
tectons, including porphyrins
and phthalocyanines. For
coordinated phase (green) proceeds for TPCN already at values of (no. of
metal atoms)/(no. of molecules)<1 (cf. Figure S7 in the Supporting Informa-
tion).
these simulations, both TPyPP and TPCN are represented by
a fourfold symmetric cross and metal adatoms are depicted as
circles (e.g., inset in Figure 4a), following a representation pre-
[
32]
viously introduced for similar systems. Within the description
of this simplified model, the two porphyrin species are identi-
cal. To distinguish the two modules, TPyPP is restricted to form
one- and twofold coordination nodes only, as observed in the
experiment, and TPCN is allowed to engage in one-, two-,
three-, or fourfold coordination. Snapshots of the Monte Carlo
simulations are depicted in Figure 4a–d (see Methods section
for details). Naturally, the metal-free dense-packed islands are
identical for the two species (Figure 4a). After including
a small amount of adatoms, TPyPP forms metal–organic chains
To further characterize the system, additional Monte Carlo
simulations were carried out for different interaction energy
ratios E /E (van der Waals interaction vs. metal-coordination).
O
C
The results are summarized in phase transition diagrams for
TPyPP and TPCN, shown in Figure 4e and 4 f, respectively. To
generate these plots, the resulting structures are color-coded;
blue for the dense-packed organic network, red for the 1D co-
ordination chains, and green for the 2D metal–organic net-
work. Note that the 2D fully reticulated coordination networks
are different for the two species (TPyPP in a twofold fashion
and TPCN in a fourfold fashion), despite being both represent-
(Figure 4b) that evolve into an open-porous 2D coordination
network (Figure 4c) when more metal is added. In contrast, for
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
5
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
ed in green (see insets in Figure 3c and d, respectively). The
most prominent difference between the two diagrams is the
lack of 1D coordination for TPCN—although it is allowed—re-
gardless of the ratio of E /E evidenced by the lack of red in
providing rotational flexibility to the termini is crucial for the
formation of linear pyridyl–Cu–pyridyl binding motifs and for
the construction of large-pore Cu-mediated coordination net-
works. Spatial constraints at the coordination center, which
can be tuned by the geometric footprint of the terminal moiet-
ies, can be deliberately exploited to control the coordination
number and thus the topology of the network architectures.
Additionally, the Monte Carlo simulations show that the spatial
constraints of the pyridyl groups not only influence the final
architectures, but also induce an energetic preference for the
formation of 1D coordination chains and therefore are the
origin of the hierarchic assembly protocol in the case of TPyPP.
Regarding the role of the metal center, our experiments
reveal that single Cu adatoms - somewhat neglected in on-sur-
face coordination chemistry to date - can link four ligands in
a quasi square-planar arrangement. Fourfold mononuclear 3d-
metal nodes on metal substrates reminiscent of the square-
O
C
Figure 4 f. Apparently, a separation of organic islands and fully
reticulated fourfold coordinated arrays is energetically favored
over mixed organic/twofold coordination assemblies. Further-
more, the simulations yield no 1D metal–organic chains for
TPCN at any probed temperature (see Methods section and
Figures S7 and S8 in the Supporting Information), thus ruling
out any influences of the experimental preparation conditions.
The TPyPP structure formation proceeds via 1D chains (red,
Figure 4d) for all calculated E /E ratios, despite the possibility
O
C
to directly form 2D coordination networks. Therefore, in the
case of terminal groups expressing only in a twofold coordina-
tion motif, the observed chaining seems energetically clearly
favored even for large relative values of E . In the modeling,
C
[36]
the only difference between the two species is the permitted
coordination geometry at the node; no effects of the electron-
ic structure are considered. As the simulations qualitatively re-
produce the experimentally observed networks, it suggests
that the different formation pathways and the resulting net-
work topologies are mainly determined by distinct spatial con-
straints at the coordination center.
planar coordination motif have been observed for Mn,
[33a,c]
[20c,33b,c]
[9]
Fe,
Co
and Ni. Recently, on-surface fourfold coordi-
nation was achieved in porphyrin-based metal–organic net-
[
14b]
works by either applying lanthanide centers
that support
[37]
high coordination numbers or by introducing a boron nitride
[11c]
spacer layer.
To our knowledge, a coordination number of
[38]
four on surfaces was only reported for Cu dimers in metal–
organic networks and for distorted Cu–carboxylate com-
[39]
plexes. This study introduces the first coordination network
based on a square-planar motif based on monoatomic Cu cen-
ters. Accordingly, on Ag(111), Cu adatoms can form coordina-
Discussion
Based on the above findings, the following conclusions can be
drawn. Without steric hindrance, Cu adatoms favor a fourfold
coordination to the nitrogen termini of porphyrins on Ag(111).
In the experiments, this situation is realized for the slender
cyano moieties of TPCN. Restricting the nodal symmetry to
twofold by introducing spatial constraints due to planar or
near-planar pyridyl groups in TPyPP, assemblies featuring 1D
coordination chains can be achieved for molecule to adatom
ratios below 1:1. Increasing the adatom concentration induces
formation of additional pyridyl links, thus yielding a fully reticu-
lated porous coordination network. Regarding ligands, the ob-
served linear pyridyl–Cu–pyridyl motif was tentatively assigned
[21a,b,24a,40]
[41]
tion bonds to two,
three, or four nitrogen atoms.
This diversity in on-surface coordination numbers discriminates
Cu from other 3d metals such as Co, for which threefold coor-
dination reminiscent of the trigonal planar motif known from
3D coordination chemistry prevails. Even for cross-like TPCN
molecules on Ag(111), Co-coordination results in a random
metal–organic network in which three- and fourfold nodes co-
[14b]
exist,
thus ruling out a dominating role of the molecular
symmetry on the resulting metal–organic architecture. Indeed,
a quantitative analysis of coordination geometries of d-block
metals in 3D supramolecular complexes and solid-state struc-
tures shows a frequent occurrence of square planar and
square pyramidal motifs for Cu, which only play a negligible
[
11a,21a,b,24a]
to steric hindrance by several studies.
Only when re-
laxing these constraints by rotating terminal pyridyl rings out
of the surface plane, for example, by using the TPyP modules,
a fourfold coordination to mononuclear centers can be ach-
[42]
role for Co. Of course, one should be well aware that the
metallic surface can drastically influence the coordination be-
havior, allowing for non-integer oxidation states, coordination
spheres unachievable in solution, and coordinatively unsaturat-
ed centers exposing apical sites to vacuum. This is exemplified
by the coordination of cyano to Co with dicarbonitrile–poly-
[
33]
ieved
square-planar motifs are well known for pyridyl complexes in
(see the Supporting Information, Figure S6). Such
[
34]
3
D coordination chemistry, but uncommon in a 2D environ-
[
35]
ment. Their rare occurrence on metallic supports might be
attributed to several aspects: A large adatom–nitrogen dis-
tance with respect to the surface induced by the rotation of
the pyridyl ring out of the surface plane, weakening the pyrid-
yl–adatom interaction, the nature of the coordinating metal
center (see below) or simply the limited number of studies ad-
dressing molecular modules featuring rotated terminal pyridyl
moieties. To achieve fourfold coordination with coplanar ad-
sorbing moieties, a terminal group inducing minor steric con-
straints (e.g., cyano) is required. In this sense, the pyridyl–
phenylene substitution of the de novo-synthesized 2H–TPyPP
[
8a]
phenyl linkers
or to Cu with 9,10-anthracenedicarbonitrile
[43]
molecules. Furthermore, the metallic surface might mimic an
[
44]
additional ligand and thus reduce the coordination number
[8a]
in the 2D adsorbate systems. From this point of view, the
twofold pyridyl–Cu–pyridyl link translates to a T-shaped coordi-
nation sphere of Cu with three ligands (two pyridyl groups
and Ag). Indeed such T-shaped motifs are observed for Cu cen-
ters in 3D complexes, in contrast to Co preferring tetrahedral
[
42,45]
or pyramidal geometries.
Additionally, TPyP molecules
[33a]
form a fourfold coordination motif with Fe
but a twofold
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
6
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[
11a]
coordination with Cu
on Au(111), which fits observations
Experimental Section
from 3D chemistry where Fe, similar to Co, strongly prefers co-
ordination to four or more partners. Thus, in addition to the
important steric constraints induced by the pyridyl rings ad-
sorbed approximately parallel to the surface, as discussed
above, the naturally preferred coordination geometries of Cu
itself might contribute to the stabilization of the linear pyridyl–
Cu–pyridyl motif prevalent on coinage metal surfaces and at
the same time support fourfold coordination in the case of
TPCN.
Experimental procedures
All experiments were performed in a custom-designed ultrahigh
vacuum chamber housing a commercial STM from Createc operat-
ed at 5 K. The base pressure during experiments was below 3ꢄ
À10
+
1
0
mbar. Repeated cycles of Ar sputtering and annealing to
7
25 K were used to prepare the Ag(111) single crystal. 2H–TPyPP
and 2H–TPCN molecules were dosed from a thoroughly degassed
quartz crucible held at 760 K. During deposition, the sample was
kept at room temperature. Cu was evaporated from a home-built,
water-cooled cell by resistively heating a W filament supporting
a Cu wire of high purity (99.9999%). All STM images were recorded
in constant current mode by using an electrochemically etched
tungsten tip prepared by sputtering and controlled dipping into
the Ag(111) substrate. In the figure captions, voltage U refers to
the bias voltage applied to the sample. The WsXM program
Conclusion
(www.nanotec.es) was used to process the STM raw data.
In summary, by exploiting the preferred coordination geome-
tries of copper in combination with a deliberate porphyrin
functionalization, we were able to design extended, 2D, grid-
like metal–organic coordination networks on Ag(111). Both
TPCN and TPyPP thus offer a basis for the fabrication of bimet-
[
14b]
[11a]
allic
and mixed-valence
open porous networks through
[
14b]
Monte Carlo simulation
orthogonal insertion of metals.
Additionally, for TPyPP, the
large pore size and the flexibility of the pyridyl–Cu–pyridyl
links result in a 2D structure reminiscent of a soft porous crys-
tal, providing opportunities to act as a template for the selec-
A square lattice with 100ꢄ100 points is used as substrate due to
the shape of the molecules. Both TPyPP and TPCN are represented
by a cross that occupies five lattice sites, whereas a metal atom
fills one lattice site (e.g., inset in Figure 4a). Only two intermolecu-
lar interactions are considered and are limited to be short-ranged
and directional, that is, they reach one lattice site along the direc-
tion of the molecular substituents. The first is the van der Waals in-
[
12g]
tive adsorption of molecular guests or nanostructures.
To rationalize the formation of distinct Cu-mediated struc-
tures from TPCN and TPyPP, we performed Monte Carlo simula-
tions and related the experimental findings to reports on Co-
mediated coordination networks and metal–organic complexes
in solution chemistry. This comparison reveals that the choice
of coordinating metal is decisive for the emerging coordination
motif; for example, replacing Co by Cu in TPCN coordination
assemblies results in a highly regular network rather than
a random structure. Here, coordination geometries in 3D
metal–organic complexes can provide some clues for an ap-
propriate selection of suitable metal for a targeted motif. Co
preferentially binds in a tetrahedral fashion and therefore is
not a promising candidate for the formation of linear, twofold
coordination motifs on surfaces. In contrast, Cu was identified
by our study as a versatile center supporting different coordi-
nation numbers and geometries. By using TPCN, we achieved
the first surface-based coordination network based on a four-
fold motif and a mononuclear Cu center. Additionally, our
study indicates that the ligand properties (e.g., rotated pyridyl
vs. planar pyridyl vs. cyano moieties) must fit the targeted
nodal geometry and thus can be used to tailor the resulting
network structure and their formation pathway through spatial
constraints. Furthermore, we demonstrated the benefits of
basic Monte Carlo simulations in selecting suitable molecular
modules for metal–organic architectures prior to the actual ex-
periment. Consequently, our study introduces prospects for
the programmed design and selection of molecular and mon-
oatomic building blocks for surface-confined supramolecular
networks and thus contributes to controlled engineering of
metal–organic/organic architectures.
teraction with energy E , which occurs when two molecules align
O
in a dense-packed fashion (see top right inset in Figure 4a) and
the second describes metal coordination with interaction energy EC
between molecules and metal adatoms. It can only be formed
when the metal atom is placed on a lattice site right at the end of
a molecular substituent. For TPyPP, the metal atom is restricted to
form one- or twofold coordination to mimic steric hindrance. Thus,
a metal atom can express coordination bonds with a maximum of
two TPyPP units, which have to be on opposite sides of the metal
atom. In contrast, a metal atom can coordinate up to four TPCN
units. The molecule–substrate interaction is neglected, which is
a reasonable approximation, as the experimental observations
reveal no relevant site-specific molecule–substrate bonding. In the
simulations, E is kept constant at a value of 40 (expressed in kT
C
units of energy), while EO is varied from 5 to 40. These values are
selected by considering both theoretical values of the bonding en-
ergies and experimental results. According to previous reports, the
bonding energy of a CuÀN coordination bond varies from 0.5 to
[46]
2
eV for 3D systems and the energies of noncovalent bonds,
[47]
[48]
here T-type or PARI interactions, range from 40 to 80 meV. As
described above, molecules are deposited at room temperature,
which means the formation temperature of the self-assembly
should be less than or equal to this value. However, stable dense-
[49]
packed islands of TPP have been observed at room temperature,
indicating that at this temperature the bonding energies already
exceed the thermal energy. Therefore it is reasonable to choose
a temperature range of 200–250 K in the simulation. By using
these values to express the bonding energies in kT units, we get
E =40 and E =5 (see above). The number of molecules is kept
C
O
constant at 300. The simulation procedure follows the protocol de-
scribed in ref. [50].
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
7
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
[
[
13] W. Auwꢁrter, K. Seufert, F. Bischoff, D. Ecija, S. Vijayaraghavan, S. Joshi, F.
Acknowledgements
Klappenberger, N. Samudrala, J. V. Barth, Nat. Nanotechnol. 2011, 7, 41–
46.
We thank C. A. Palma and F. Klappenberger for fruitful discus-
sions. This work was supported by the European Research
Council (ERC) Advanced Grant MolArt (no. 247299) and the
Technische Universitꢁt Mꢅnchen Institute for Advanced Study
funded by the German Research Foundation (DFG) via the
German Excellence Initiative. D.B. gratefully acknowledges the
EU through the ERC Starting Grant „COLORLANDS“ project, the
FRS-FNRS (FRFC contracts no. 2.4.550.09), D.S. thanks the FNRS
for her doctoral fellowship. W.A. acknowledges funding by the
DFG via a Heisenberg professorship and by the ERC Consolida-
tor Grant NanoSurfs (no. 615233).
14] a) H. Marbach, Acc. Chem. Res. 2015, 48, 2649–2658; b) J. I. Urgel, D.
Ecija, W. Auwꢁrter, D. Stassen, D. Bonifazi, J. V. Barth, Angew. Chem. Int.
Ed. 2015, 54, 6163–6167; Angew. Chem. 2015, 127, 6261–6265; c) K.
Diller, A. C. Papageorgiou, F. Klappenberger, F. Allegretti, J. V. Barth, W.
Auwꢁrter, Chem. Soc. Rev. 2016, 45, 1629–1656.
15] B. Hulsken, R. Van Hameren, J. W. Gerritsen, T. Khoury, P. Thordarson,
M. J. Crossley, A. E. Rowan, R. J. M. Nolte, J. A. A. W. Elemans, S. Speller,
Nat. Nanotechnol. 2007, 2, 285–289.
16] C. Wꢁckerlin, K. Tarafder, J. Girovsky, J. Nowakowski, T. Hꢁhlen, A.
Shchyrba, D. Siewert, A. Kleibert, F. Nolting, P. M. Oppeneer, T. A. Jung,
N. Ballav, Angew. Chem. Int. Ed. 2013, 52, 4568–4571; Angew. Chem.
2013, 125, 4666–4669.
[
[
[
[
17] T. Lin, X. S. Shang, J. Adisoejoso, P. N. Liu, N. Lin, J. Am. Chem. Soc. 2013,
1
35, 3576–3582.
18] a) W. Auwꢁrter, A. Weber-Bargioni, A. Riemann, A. Schiffrin, O. Grçning,
R. Fasel, J. V. Barth, J. Chem. Phys. 2006, 124, 194708; b) W. Auwꢁrter, K.
Seufert, F. Klappenberger, J. Reichert, A. Weber-Bargioni, A. Verdini, D.
Cvetko, M. Dell’Angela, L. Floreano, A. Cossaro, G. Bavdek, A. Morgante,
A. P. Seitsonen, J. V. Barth, Phys. Rev. B 2010, 81, 245403.
Keywords: Monte Carlo simulations · porous materials ·
porphyrins · scanning tunneling microscopy · supramolecular
chemistry
[
[
19] P. Avouris, J. E. Demuth, J. Chem. Phys. 1981, 75, 4783–4794.
20] a) S. Gottardi, K. Mꢅller, J. C. Moreno-Lꢇpez, H. Yildirim, U. Meinhardt, M.
Kivala, A. Kara, M. Stçhr, Adv. Mater. Interfaces 2014, 1, 1300025; b) M.
Marschall, J. Reichert, K. Seufert, W. Auwꢁrter, F. Klappenberger, A.
Weber-Bargioni, S. Klyatskaya, G. Zoppellaro, A. Nefedov, T. Strunskus, C.
Wçll, M. Ruben, J. V. Barth, ChemPhysChem 2010, 11, 1446–1451; c) M.
Marschall, J. Reichert, A. Weber-Bargioni, K. Seufert, W. Auwꢁrter, S.
Klyatskaya, G. Zoppellaro, M. Ruben, J. V. Barth, Nat. Chem. 2010, 2,
131–137.
[
[
1] G. A. Craig, M. Murrie, Chem. Soc. Rev. 2015, 44, 2135–2147.
2] C. Coulon, V. Pianet, M. Urdampilleta, R. Clꢀrac in Molecular Nanomag-
nets and Related Phenomena (Ed.: S. Gao), Springer, Berlin, Heidelberg,
2015, p. 164.
[3] a) L. Lux, K. Williams, S. Ma, CrystEngComm 2015, 17, 10–22; b) F.-S. Ke,
Y.-S. Wu, H. Deng, J. Solid State Chem. 2015, 223, 109–121.
[
[
4] K. Leus, Y.-Y. Liu, P. V. D. Voort, Catal. Rev. Sci. Eng. 2014, 56, 1–56.
5] M. Ruben, J. Rojo, F. J. Romero-Salguero, L. H. Uppadine, J.-M. Lehn,
Angew. Chem. Int. Ed. 2004, 43, 3644–3662; Angew. Chem. 2004, 116,
[21] a) D. Heim, D. Ecija, K. Seufert, W. Auwꢁrter, C. Aurisicchio, C. Fabbro, D.
Bonifazi, J. V. Barth, J. Am. Chem. Soc. 2010, 132, 6783–6790; b) D.
Heim, K. Seufert, W. Auwꢁrter, C. Aurisicchio, C. Fabbro, D. Bonifazi, J. V.
Barth, Nano Lett. 2010, 10, 122–128; c) D. Ecija, M. Marschall, J. Reichert,
A. Kasperski, D. Nieckarz, P. Szabelski, W. Auwꢁrter, J. V. Barth, Surf. Sci.
2016, 643, 91–97; d) T. Kaposi, S. Joshi, T. Hoh, A. Wiengarten, K. Seu-
fert, M. Paszkiewicz, F. Klappenberger, D. Ecija, L. Dordevic, T. Marango-
ni, D. Bonifazi, J. V. Barth, W. Auwꢁrter, ACS Nano 2016, 10, 7665–7674.
[22] M. Pivetta, G. E. Pacchioni, E. Fernandes, H. Brune, J. Chem. Phys. 2015,
142, 101928.
3
728–3747.
[
6] a) J. V. Barth, G. Costantini, K. Kern, Nature 2005, 437, 671–679; b) L.
Bartels, Nat. Chem. 2010, 2, 87–95; c) N. Lin, S. Stepanow, M. Ruben,
J. V. Barth in Templates in Chemistry III (Eds.: P. Broekmann, K.-H. Dçtz,
C. A. Schalley), Springer, Berlin, Heidelbergr, 2008, pp. 1–44.
7] a) R. Gutzler, S. Stepanow, D. Grumelli, M. Lingenfelder, K. Kern, Acc.
Chem. Res. 2015, 48, 2132–2139; b) D. Grumelli, B. Wurster, S. Stepa-
now, K. Kern, Nat. Commun. 2013, 4, 2904.
[
[
8] a) U. Schlickum, R. Decker, F. Klappenberger, G. Zoppellaro, S. Klyat-
skaya, M. Ruben, I. Silanes, A. Arnau, K. Kern, H. Brune, J. V. Barth, Nano
Lett. 2007, 7, 3813–3817; b) R. Decker, U. Schlickum, F. Klappenberger,
G. Zoppellaro, S. Klyatskaya, M. Ruben, J. V. Barth, H. Brune, Appl. Phys.
Lett. 2008, 93, 243102; c) D. Kꢅhne, F. Klappenberger, W. Krenner, S.
Klyatskaya, M. Ruben, J. V. Barth, Proc. Natl. Acad. Sci. USA 2010, 107,
[23] a) M. Mi n´ kowski, M. A. Załuska-Kotur, Surf. Sci. 2015, 642, 22–32; b) S.
Sikandar Hayat, M. Alcꢈntara Ortigoza, M. A. Choudhry, T. S. Rahman,
Phys. Rev. B 2010, 82, 085405.
[24] a) S. L. Tait, A. Langner, N. Lin, S. Stepanow, C. Rajadurai, M. Ruben, K.
Kern, J. Phys. Chem. C 2007, 111, 10982–10987; b) N. Henningsen, R.
Rurali, C. Limbach, R. Drost, J. I. Pascual, K. J. Franke, J. Phys. Chem. Lett.
2011, 2, 55–61.
21332–21336; d) S. Nowakowska, A. Wꢁckerlin, S. Kawai, T. Ivas, J. Now-
akowski, S. Fatayer, C. Wꢁckerlin, T. Nijs, E. Meyer, J. Bjçrk, M. Stçhr, L. H.
Gade, T. A. Jung, Nat. Commun. 2015, 6, 6071; e) R. Zhang, G. Lyu, C.
Chen, T. Lin, J. Liu, P. N. Liu, N. Lin, ACS Nano 2015, 9, 8547–8553.
9] N. Abdurakhmanova, T.-C. Tseng, A. Langner, C. S. Kley, V. Sessi, S. Stepa-
now, K. Kern, Phys. Rev. Lett. 2013, 110, 027202.
[25] S. Vijayaraghavan, D. Ecija, W. Auwꢁrter, S. Joshi, K. Seufert, M. Drach, D.
Nieckarz, P. Szabelski, C. Aurisicchio, D. Bonifazi, J. V. Barth, Chem. Eur. J.
2013, 19, 14143–14150.
[26] T. R. Umbach, M. Bernien, C. F. Hermanns, L. L. Sun, H. Mohrmann, K. E.
Hermann, A. Krꢅger, N. Krane, Z. Yang, F. Nickel, Y.-M. Chang, K. J.
Franke, J. I. Pascual, W. Kuch, Phys. Rev. B 2014, 89, 235409.
[27] a) D. Ecija, S. Vijayaraghavan, W. Auwꢁrter, S. Joshi, K. Seufert, C. Aurisic-
chio, D. Bonifazi, J. V. Barth, ACS Nano 2012, 6, 4258–4265; b) H. Spill-
mann, A. Kiebele, M. Stçhr, T. A. Jung, D. Bonifazi, F. Cheng, F. Diederich,
Adv. Mater. 2006, 18, 275–279.
[28] C.-A. Palma, J. Bjçrk, F. Klappenberger, E. Arras, D. Kꢅhne, S. Stafstrçm,
J. V. Barth, Nat. Commun. 2015, 6, 6210.
[29] a) S. Horike, S. Shimomura, S. Kitagawa, Nat. Chem. 2009, 1, 695–704;
b) T. Sirtl, S. Schlçgl, A. Rastgoo-Lahrood, J. Jelic, S. Neogi, M. Schmittel,
W. M. Heckl, K. Reuter, M. Lackinger, J. Am. Chem. Soc. 2013, 135, 691–
695.
[
[
10] G. E. Pacchioni, M. Pivetta, H. Brune, J. Phys. Chem. C 2015, 119, 25442–
5448.
2
[
11] a) Y. Li, J. Xiao, T. E. Shubina, M. Chen, Z. Shi, M. Schmid, H.-P. Steinrꢅck,
J. M. Gottfried, N. Lin, J. Am. Chem. Soc. 2012, 134, 6401–6408; b) S. Ste-
panow, M. Lingenfelder, A. Dmitriev, H. Spillmann, E. Delvigne, N. Lin, X.
Deng, C. Cai, J. V. Barth, K. Kern, Nat. Mater. 2004, 3, 229–233; c) J. I.
Urgel, M. Schwarz, M. Garnica, D. Stassen, D. Bonifazi, D. Ecija, J. V.
Barth, W. Auwꢁrter, J. Am. Chem. Soc. 2015, 137, 2420–2423.
[
12] a) W. Auwꢁrter, D. ꢆcija, F. Klappenberger, J. V. Barth, Nat. Chem. 2015, 7,
105–120; b) J. M. Gottfried, Surf. Sci. Rep. 2015, 70, 259–379;c) S. Moh-
nani, D. Bonifazi, Coord. Chem. Rev. 2010, 254, 2342–2362; d) L.-A.
Fendt, M. Stçhr, N. Wintjes, M. Enache, T. A. Jung, F. Diederich, Chem.
Eur. J. 2009, 15, 11139–11150; e) N. Wintjes, J. Lobo-Checa, J. Hornung,
T. Samuely, F. Diederich, T. A. Jung, J. Am. Chem. Soc. 2010, 132, 7306–
[30] a) T. Lin, X. S. Shang, P. N. Liu, N. Lin, J. Phys. Chem. C 2013, 117, 23027–
23033; b) T. Lin, Q. Wu, J. Liu, Z. Shi, P. N. Liu, N. Lin, J. Chem. Phys.
2015, 142, 101909.
7
311; f) C. Iacovita, P. Fesser, S. Vijayaraghavan, M. Enache, M. Stçhr, F.
[31] A. Kasperski, P. Szabelski, Adsorption 2013, 19, 283–289.
[32] A. Kasperski, D. Nieckarz, P. Szabelski, Surf. Sci. 2015, 641, 269–277.
[33] a) T. Lin, G. Kuang, W. Wang, N. Lin, ACS Nano 2014, 8, 8310–8316;
b) M. E. Garah, A. Ciesielski, N. Marets, V. Bulach, M. W. Hosseini, P.
Diederich, T. A. Jung, Chem. Eur. J. 2012, 18, 14610–14613; g) F. Sedona,
M. Di Marino, M. Sambi, T. Carofiglio, E. Lubian, M. Casarin, E. Tondello,
ACS Nano 2010, 4, 5147–5154.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
8
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Samorꢉ, Chem. Commun. 2014, 50, 12250–12253;c) B. Wurster, D. Gru-
melli, D. Hçtger, R. Gutzler, K. Kern, J. Am. Chem. Soc. 2016, 138, 3623–
[42] D. Venkataraman, Y. Du, S. R. Wilson, K. A. Hirsch, P. Zhang, J. S. Moore,
J. Chem. Educ. 1997, 74, 915.
3
626.
[43] G. Pawin, K. L. Wong, D. Kim, D. Sun, L. Bartels, S. Hong, T. S. Rahman, R.
Carp, M. Marsella, Angew. Chem. 2008, 120, 8570–8573.
[
34] a) G. J. Long, P. J. Clarke, Inorg. Chem. 1978, 17, 1394–1401; b) Y. Agnus,
M. Labarelle, R. Louis, B. Metz, Acta Crystallogr. Sect. A 1994, 50, 536–
[44] K. Flechtner, A. Kretschmann, H.-P. Steinrꢅck, J. M. Gottfried, J. Am.
Chem. Soc. 2007, 129, 12110–12111.
[45] W. Hieringer, K. Flechtner, A. Kretschmann, K. Seufert, W. Auwꢁrter, J. V.
Barth, A. Gçrling, H.-P. Steinrꢅck, J. M. Gottfried, J. Am. Chem. Soc. 2011,
5
6
38; c) D. J. Hamm, J. Bordner, A. F. Schreiner, Inorg. Chim. Acta 1973, 7,
37–641.
[
[
35] A. Dmitriev, H. Spillmann, N. Lin, J. V. Barth, K. Kern, Angew. Chem. 2003,
15, 2774–2777.
1
33, 6206–6222.
1
[
46] a) C. F. Faul, M. Antonietti, Adv. Mater. 2003, 15, 673–683; b) J. V. Barth,
Annu. Rev. Phys. Chem. 2007, 58, 375–407; c) D.-Y. Wu, B. Ren, Y.-X.
Jiang, X. Xu, Z.-Q. Tian, J. Phys. Chem. A 2002, 106, 9042–9052.
36] a) T.-C. Tseng, C. Lin, X. Shi, S. L. Tait, X. Liu, U. Starke, N. Lin, R. Zhang,
C. Minot, M. A. Van Hove, J. I. Cerdꢈ, K. Kern, Phys. Rev. B 2009, 80,
155458; b) N. Abdurakhmanova, A. Floris, T.-C. Tseng, A. Comisso, S. Ste-
[
47] a) O. Bludsk, M. Rube sˇ , P. Soldꢈn, P. Nachtigall, J. Chem. Phys. 2008, 128,
panow, A. De Vita, K. Kern, Nat. Commun. 2012, 3, 940.
37] G. Lyu, Q. Zhang, J. I. Urgel, G. Kuang, W. Auwꢁrter, D. Ecija, J. V. Barth,
N. Lin, Chem. Commun. 2016, 52, 1618–1621.
114102; b) S. Tsuzuki, K. Honda, T. Uchimaru, M. Mikami, K. Tanabe, J.
[
[
[
[
Am. Chem. Soc. 2002, 124, 104–112.
[
[
48] E. Arras, A. P. Seitsonen, F. Klappenberger, J. V. Barth, Phys. Chem. Chem.
Phys. 2012, 14, 15995.
49] a) L. Scudiero, D. E. Barlow, K. W. Hipps, J. Phys. Chem. B 2000, 104,
38] A. Breitruck, H. E. Hoster, C. Meier, U. Ziener, R. J. Behm, Surf. Sci. 2007,
601, 4200–4205.
39] N. Lin, A. Dmitriev, J. Weckesser, J. V. Barth, K. Kern, Angew. Chem. Int.
Ed. 2002, 41, 4779–4783; Angew. Chem. 2002, 114, 4973–4977.
11899–11905; b) K. Comanici, F. Buchner, K. Flechtner, T. Lukasczyk, J. M.
Gottfried, H.-P. Steinrꢅck, H. Marbach, Langmuir 2008, 24, 1897–1901.
50] D. Nieckarz, P. Szabelski, J. Phys. Chem. C 2013, 117, 11229–11241.
40] A. Langner, S. L. Tait, N. Lin, R. Chandrasekar, M. Ruben, K. Kern, Angew.
Chem. Int. Ed. 2008, 47, 8835–8838; Angew. Chem. 2008, 120, 8967–
[
8970.
[41] S. L. Tait, A. Langner, N. Lin, R. Chandrasekar, O. Fuhr, M. Ruben, K. Kern,
Received: May 6, 2016
ChemPhysChem 2008, 9, 2495–2499.
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
9
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Porphyrins
F. Bischoff, Y. He, K. Seufert, D. Stassen,
D. Bonifazi,* J. V. Barth, W. Auwꢀrter*
&& – &&
Tailoring Large Pores of Porphyrin
Networks on Ag(111) by Metal–
Organic Coordination
The grids are alright: A scanning tun-
neling microsopy study combined with
Monte Carlo modeling of two similar,
but distinctly functionalized, porphyrin
tectons reveals how steric hindrance at
Cu coordination nodes guides the as-
sembly of porous grid-like networks
with unprecedented morphology and
pore size.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
10
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!