Synthesis of two-dimensional phenylene-boroxine networks through in vacuo condensation and on-surface radical addition

Article abstract of DOI:10.1039/c1cc13896k

We report on covalent two-dimensional phenylene-boroxine hybrid-networks that were synthesized under ultra-high vacuum conditions from doubly functionalized monomers through thermally activated condensation prior to deposition and successive heterogeneously catalyzed radical addition on Ag(111). Structural properties were characterized in situ by high resolution Scanning-Tunneling-Microscopy (STM).

Full text of DOI:10.1039/c1cc13896k

View Article Online / Journal Homepage / Table of Contents for this issue

Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 12355–12357
www.rsc.org/chemcomm
COMMUNICATION
Synthesis of two-dimensional phenylene–boroxine networks through
in vacuo condensation and on-surface radical additionwz
ab
Stefan Schlogl, Thomas Sirtl,ab Johanna Eichhorn,ab Wolfgang M. Hecklabc and
¨
Markus Lackinger*abc
Received 30th June 2011, Accepted 10th October 2011
DOI: 10.1039/c1cc13896k
We report on covalent two-dimensional phenylene–boroxine
hybrid-networks that were synthesized under ultra-high vacuum
conditions from doubly functionalized monomers through
thermally activated condensation prior to deposition and successive
heterogeneously catalyzed radical addition on Ag(111). Structural
properties were characterized in situ by high resolution Scanning-
Tunneling-Microscopy (STM).
Ag(111). Subsequent thermal annealing initiates the on-surface
polymerization via radical addition (see Fig. 1 for a reaction
scheme). To this end, the aromatic DBPBA monomer is
doubly functionalized with two different functional groups:
one boronic acid group for condensation and two bromine
substituents for surface-mediated radical addition. On
Ag(111) radical addition of brominated precursor molecules
is kinetically suppressed at room temperature and thermal
annealing is required to initiate the polymerization through
homolysis of the bromine substituents.12 On the other hand,
the condensation of boronic acids can be thermally activated
in vacuo without the need for a catalyst.8 In the proposed
reaction scheme, the cyclo-condensation of DBPBA molecules
into TDBPB is the last synthesis step prior to deposition and
carried out in vacuo by a mere thermal treatment at 190 1C.
Boronic acids preferably react into boroxine rings through
cyclo-condensation of three entities, whereby three H2O molecules
are released. Because DBPBA is only monofunctionalized for
The synthesis of covalent two-dimensional organic materials is
a challenging but promising goal in material science. Single
layer graphene is the most abundant and best studied 2D
organic material.1 The inherent absence of an electronic band-gap
gives rise to intriguing physics, yet, other envisaged applications
require a band-gap that needs to be introduced by dopant side
groups.2 While graphene can meanwhile routinely be prepared
by exfoliation of graphite or graphite oxide,3 a bottom-up
synthetic approach through polymerisation of organic monomers
offers structural and chemical versatility.4–6 On-surface synthesis
of 2D covalent networks was accomplished through various
types of polymerization reactions,4 predominantly condensation7–9
and radical addition of halogenated precursors.10–12 Herein we
demonstrate the synthesis of covalent hybrid-networks comprised
of both phenyl and boroxine rings arranged on a hexagonal
lattice. Boroxine rings (B3O3) are hexagonal, planar, and
similar in size to phenyl rings.13 In this respect it is interesting
to study their interchangeability in covalent networks and the
impact on structural properties.14,15 Substitution of phenyl
with boroxine rings is also known to modify material properties,
e.g. renders polymeric materials flame retardant.16,17 Synthesis
of the proposed covalent hybrid-networks is based on
3,5-dibromophenylboronic acid (DBPBA). First, heating of
DBPBA thermally activates condensation which yields less
reactive TDBPB monomers that are then deposited onto
Fig. 1 Reaction scheme for synthesis of covalent phenylene–boroxine
hybrid-networks from 3,5-dibromophenylboronic acid (DBPBA)
monomers. First thermally activated condensation of (a) DBPBA in
the crucible of a Knudsen cell results in (b) 1,3,5-tris(30,50-dibromo-
phenyl)-boroxine (TDBPB). Deposition of TDBPB on Ag(111) and
further thermal activation yield (c) surface-stabilized triphenylene–
boroxine hexaradicals (TPBHR) that subsequently polymerize into
covalent phenylene–boroxine networks through radical addition.
Hereby the bromine substitution pattern of TDBPB allows for two
different stackings: (d) ABAB and (e) AA.
a Department of Physics & TUM School of Education, Tech. Univ.
Munich, Schellingstrasse 33, 80333 Munich, Germany.
E-mail: markus@lackinger.org; Tel: +49 89 2179 605
b Deutsches Museum, Museumsinsel 1, 80538 Munich, Germany
c Center for NanoScience (CeNS), Schellingstrasse 4, 80799 Munich,
Germany
w This article is part of the ChemComm ‘Molecule-based surface
chemistry’ web themed issue.
z Electronic supplementary information (ESI) available: Experimental
and calculational details; DFT results; LEED, NMR, additional STM,
and Raman data. See DOI: 10.1039/c1cc13896k
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 12355–12357 12355

condensation, the reaction does not yield interconnected networks
as for diboronic acids,13 but larger 1,3,5-tris(30,50-dibromophe-
nyl)-boroxine molecules (TDBPB, see Fig. 1(b)). This ‘‘in-crucible
synthesis’’ is feasible, because cyclo-condensation of the boronic
acid functionality is preferred over sublimation, whereby thermal
sublimation of the free boronic acid DBPBA is essentially not
possible. Instead a relatively high experimentally determined
sublimation temperature of 190 1C indicates thermal sublimation
of the larger threefold-symmetric TDBPB monomer that already
features both phenyl and boroxine rings. Since bromine function-
alities are not known to react at these relatively low temperatures
without the presence of a catalyst, further polymerization reac-
tions in the crucible are not expected, but thermal sublimation of
intact molecules. The thermal ‘‘in-crucible synthesis’’ of TDBPB
was verified by comparative Raman spectroscopy and NMR of
the initial and thermally treated DBPBA compound (see ESIz for
spectra and details). In additional experiments a DBPBA filled
crucible was used to deposit organic molecules onto graphite(001)
as well as onto Ag(111), with both substrates held at room
temperature. These experiments further substantiate the thermally
activated condensation of DBPBA into TDBPB and its subsequent
deposition. On inert graphite(001) substrates well-ordered self-
assembled trigonal monolayers with a lattice parameter of (1.50 ꢀ
0.10) nm were observed. In the STM topograph shown in Fig. 2(a)
threefold symmetric molecular entities can be identified, whose
size and symmetry are consistent with intact TDBPB molecules.
Moreover, six peripheral protruding spherical features can be
discerned and are assigned to the bromine substituents of intact
TDBPB. In the proposed structural model, cyclic halogen–
halogen–halogen bonds between three bromine substituents with
experimental bromine–bromine distances of (0.45 ꢀ 0.10) nm
stabilize the self-assembled monolayer (Fig. 2(b)). Halogen bonds
are attractive anisotropic electrostatic interactions between Cl, Br,
or I subsitutents that originate from a non-spherical charge
distribution at the halogen with a negative ring and a positive
end cap.18,19 Trigonal cyclic arrangements optimize electrostatic
interactions and were similarly found in both 3D and 2D
structures.18,20 Experimental Br–Br distances here are larger than
those found in bulk crystals (0.36 nm) and hint towards a surface
influence. STM measurements of TDBPB deposited onto Ag(111)
reveal a less well ordered arrangement of threefold symmetric
molecules (Fig. S3, ESIz). Again, size and symmetry of the
molecules and the discernible bromine substituents suggest
deposition of TDBPB rather than the precursor material DBPBA.
Albeit some monomers are already interconnected and some
contrast features point toward deposition of a low fraction of
unreacted DBPBA monomers, the great majority of TDBPB
molecules remains intact upon deposition onto Ag(111) held at
room temperature and the bromine substituents are not cleaved
off. Although the intermolecular arrangements exhibit only short-
range order, molecular aggregates are likewise stabilized by
halogen-bonds similar to the monolayers on graphite. LEED
measurements of similarly prepared samples show a hexagonal
pattern of rather broad reflections. The LEED pattern is consis-
tent with two orientational domains of (2O7ꢁ2O7) ꢀ R191 that
corresponds to a lattice parameter of a = 1.53 nm (Fig. S2, ESIz).
Yet, the diffraction pattern most likely does not arise from long-
range ordered monolayers, but from short range order, where
TDBPB dimers adopt a preferred intermolecular distance of 1.5 nm
and a preferred orientation with respect to the substrate. The
LEED lattice parameter is consistent with the nearest-neighbor
distance in STM measurements of less-well ordered samples.
Although in the STM experiments, the deposition rate was varied
from 0.02–0.1 monolayer per minute, no long-range ordered
monolayers were obtained on Ag(111), maybe because of DBPBA
impurities and first covalent interlinking reactions.
Because of its bromine substitution pattern and its planar
structure, TDBPB is a well suited monomer for the synthesis
of hexagonal phenylene–boroxine hybrid-networks through
surface-mediated polymerization. While annealing of TDBPB
monolayers up to 200 1C for 10 min on graphite(001) only
leads to thermally induced desorption, annealing on Ag(111)
(10 min at 130 1C) activates the dehalogenation reaction aided
by the catalytic properties of the surface,10,12,21 whereby
triphenylene–boroxine hexaradicals (TPBHR) are generated
(Fig. 1(c)). Subsequently, these radicals polymerize through
radical addition into covalent networks directly on the surface
(Fig. 3(a) and (b)). As illustrated in the reaction scheme the
polymerization of hexaradicals can proceed in two different
stereochemical arrangements (Fig. 1(d) and (e)). Two adjacent
monomers can interlink via two covalent intermolecular
bonds. Since the monomers possess a triangular footprint this
means they adjoin at their bases. Exclusive formation of these
two-fold covalent interconnects would result in a periodic
covalent network with ABAB stacking and P6mm symmetry
as depicted in Fig. 1(d). However, radical addition is also
possible in a way that one monomer forms two covalent bonds
with two different monomers on one side of its triangular
footprint. Continuation of this binding motif into extended
two dimensional networks yields AA stacking (Fig. 1(e)),
where the tips of the triangles are aligned with the center of
the baseline in the row above. The resulting periodic network
would still exhibit a hexagonal arrangement of boroxine and
phenyl rings, but would possess considerably lower symmetry
(P2mg). In the experiment random stacking is expected, since
the polymerization cannot be controlled, and consequently a
less regular distribution of boroxine rings within the hexagonal
network. Polymerization, i.e. the formation of covalent networks
upon thermal treatment is unambiguously verified by
Fig. 2 Self-assembled TDPBP monolayer on graphite(001). (a) STM
topograph (+0.60 V, 40 pA, correlation-averaged); as evident from
the scaled overlay, size and symmetry of the molecular entities are
consistent with threefold-symmetric TDBPB molecules and six
peripheral bromines of one molecule are highlighted by black arrows.
(b) Tentative model of the TDBPB monolayer, where molecules
interact through cyclic halogen–halogen–halogen bonds (indicated
by black arrows).
c
12356 Chem. Commun., 2011, 47, 12355–12357
This journal is The Royal Society of Chemistry 2011

(see Fig. S5, ESIz). Around these vacancies, the hexagonal
lattice is preserved and just one monomer is missing, leading to
AA stacking.
In summary, we show that a covalent hybrid-network
consisting of both phenyl and boroxine rings can be synthe-
sized entirely in UHV from a comparatively simple monomer.
The employed monomer was doubly functionalized with both
boronic acid groups and bromine substituents for condensation
and radical addition reactions, respectively. We find that
vacuum sublimation of even comparatively small boronic acids
is not possible, because increasing the crucible temperature
up to the sublimation point already thermally activates self-
condensation. Then again, the preferred and reliable self-
condensation of boronic acids into boroxine rings can also
serve as a design principle for carrying out the last synthesis
step in the crucible of the Knudsen cell and deposit more
complex monomers from abundantly available, structurally
simpler precursor molecules.
We thank Dr A. M. Gigler for Raman measurements and
S. Hug and Prof. B. V. Lotsch for NMR measurements.
Funding by the Nanosystems-Initiative Munich (NIM) cluster
of excellence, Elitenetzwerk Bayern (St. S.) and Verband der
Chemischen Industrie (T.S.) is gratefully acknowledged.
Fig. 3 (a) Overview and (b) close-up STM topographs (+0.7 V, 50 pA)
of annealed TDPBP monolayers on Ag(111). Thermally activated
dehalogenation and radical addition yield covalent networks. (c)
Line-profile along A–B in image (a); it indicates a lattice parameter,
i.e. interpore distance, of (0.78 ꢀ 0.10) nm. (d) Close-up of covalent
networks with molecular overlay. The aggregate is formed by exclusive
AB stacking.
Notes and references
1 A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191.
2 J. Wintterlin and M. L. Bocquet, Surf. Sci., 2009, 603, 1841–1852.
3 H. C. Schniepp, J. L. Li, M. J. McAllister, H. Sai, M. Herrera-
Alonso, D. H. Adamson, R. K. Prud’homme, R. Car, D. A. Saville
and I. A. Aksay, J. Phys. Chem. B, 2006, 110, 8535–8539.
4 A. Gourdon, Angew. Chem., Int. Ed., 2008, 47, 6950–6953.
5 D. F. Perepichka and F. Rosei, Science, 2009, 323, 216–217.
6 J. Sakamoto, J. van Heijst, O. Lukin and A. D. Schluter, Angew.
Chem., Int. Ed., 2009, 48, 1030–1069.
comparison of precise experimental STM derived structure
parameters with theoretical values. The line-profile in Fig. 3(c)
indicates a spacing between adjacent six-membered rings of
(0.78 ꢀ 0.10) nm. This value is in perfect agreement with the
0.78 nm as obtained from geometry optimized DFT calculations
of the P6mm structure (cf. ESIz).
7 C. H. Schmitz, J. Ikonomov and M. Sokolowski, J. Phys. Chem. C,
2009, 113, 11984.
The DFT calculations also yield an approximative value of
2.8 eV for the HOMO–LUMO energy gap. The electronic
states do not exhibit a significant dispersion in k-space. This
means that electrons are localized with low mobility and
almost infinite effective mass. Similar hexagonal networks that
only consist of phenylene rings show a less flat dispersion
relation. Comparison of those structurally similar networks
denotes effective electronic decoupling through boroxine
rings22 (details can be found in ESIz).
8 S. Weigelt, C. Busse, C. Bombis, M. M. Knudsen, K. V. Gothelf,
E. Lægsgaard, F. Besenbacher and T. R. Linderoth, Angew.
Chem., Int. Ed., 2008, 47, 4406.
9 N. A. A. Zwaneveld, R. Pawlak, M. Abel, D. Catalin, D. Gigmes,
D. Bertin and L. Porte, J. Am. Chem. Soc., 2008, 130, 6678–6679.
10 M. Bieri, M. Nguyen, O. Groning, J. Cai, M. Treier, K. Ait-
Mansour, P. Ruffieux, C. Pignedoli, D. Passerone, M. Kastler,
K. Mullen and R. Fasel, J. Am. Chem. Soc., 2010, 132,
16669–16676.
11 L. Grill, M. Dyer, L. Lafferentz, M. Persson, M. Peters and
S. Hecht, Nat. Nanotechnol., 2007, 2, 687–691.
12 R. Gutzler, H. Walch, G. Eder, S. Kloft, W. M. Heckl and
M. Lackinger, Chem. Commun., 2009, 4456–4458.
13 A. P. Cote, A. I. Benin, N. W. Ockwig, M. O’Keeffe, A. J. Matzger
and O. M. Yaghi, Science, 2005, 310, 1166–1170.
14 M. J. Farrall and J. M. J. Frechet, J. Org. Chem., 1976, 41,
3877–3882.
15 J. Pellon, L. H. Schwind, M. J. Guinard and W. M. Thomas,
J. Polym. Sci., 1961, 55, 161–167.
16 A. L. Korich and P. M. Iovine, Dalton Trans., 2010, 39, 1423–1431.
17 A. B. Morgan, J. L. Jurs and J. M. Tour, J. Appl. Polym. Sci.,
2000, 76, 1257–1268.
The covalent phenylene–boroxine networks do not exhibit
long-range order. Yet, in covalent networks that were synthesized
from structurally much simpler tritopic monomers as for
instance tribromophenylbenzene12,21 non-ideal strained covalent
rings composed of more or less than six monomers occur with
densities comparable to those of ideal rings. In polymerized
TDBPB networks strained rings are only observed occasionally
(cf. ESIz). Apparently, their formation becomes less likely due
to the complex bond topology of the six-fold bromine-
substituted monomer. Although the polymerized TDBPB
networks exhibit mostly regular rings, the resulting domain
size is comparatively small. Different azimuthal orientations to
the substrate impair coherent interconnection into long-range
ordered networks. In contrast to covalent networks from
simple tritopic monomers, now vacancies can be observed
18 E. Bosch and C. L. Barnes, Cryst. Growth Des., 2002, 2, 299–302.
19 S. L. Price, A. J. Stone, J. Lucas, R. S. Rowland and
A. E. Thornley, J. Am. Chem. Soc., 1994, 116, 4910–4918.
20 H. Walch, R. Gutzler, T. Sirtl, G. Eder and M. Lackinger, J. Phys.
Chem. C, 2010, 114, 12604–12609.
21 M. O. Blunt, J. C. Russell, N. R. Champness and P. H. Beton,
Chem. Commun., 2010, 46, 7157–7159.
22 D. Armstrong and P. Perkins, J. Chem. Soc. A, 1967, 790–792.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 12355–12357 12357