Ionothermal route to layered two-dimensional polymer-frameworks based on heptazine linkers

Article abstract of DOI:10.1021/ma101008c

The synthesis of three molecular building blocks based on 1,4-aryl, 4,40-biphenyl and 1,3-aryl each containing a triazine and a carbonitrile functionality and the thermally induced autocondensation of these into three respective heptazine-based, ordered,

Full text of DOI:10.1021/ma101008c

Macromolecules 2010, 43, 6639–6645 6639
DOI: 10.1021/ma101008c
Ionothermal Route to Layered Two-Dimensional Polymer-Frameworks
Based on Heptazine Linkers
Michael J. Bojdys,*,† Stephanie A. Wohlgemuth,† Arne Thomas,‡ and Markus Antonietti†
†Max Planck Institute of Colloids and Interfaces, Department of Colloid Chemistry, Research Campus Golm,
‡
€
14476 Potsdam, Germany, and Technische Universitat Berlin, Funktionsmaterialien, Englische Strasse 20,
10587 Berlin, Germany
Received May 7, 2010; Revised Manuscript Received July 14, 2010
Introduction
triazine group (an ammeline derivative) with dicyandiamide
(DCDA) in a base catalyzed reaction. The formation of the
Layered two-dimensional polymer-frameworks are fascinating
research targets. The development of reliable synthetic routes to
periodic, covalent molecular layers touches on structural, analy-
tical, technological and theoretical aspects of the natural sciences
and has a profound impact on fundamental research and its
realization, as seen for example in graphene research1 and the
recent field of 2D and 3D covalent frameworks.2-6 The linking of
organic molecules to long-range ordered moieties or materials is
classically limited to noncovalent interactions,7-9 while covalent
polymerization reactions usually only yield structurally rando-
mized networks.10,11
The most challenging hurdle in the synthesis of extended, yet
precisely defined 2D and 3D structures based on covalent che-
mistry is widely believed to be the requirement, that the reaction
linking individual organic constituents needs to be reversible,
allowing the scaffold to arrange to the thermodynamic, well
ordered product rather than the kinetic, amorphous structure.12,13
In addition, the design of the building blocks must allow for
growth into a nonperturbed, regular geometry (i.e., at least three
latent bond-forming sites enclosing an appropriate angle) and for
sufficient chemical stability.
One such promising candidate for a condensation and bonding
pattern of monomeric units was described first in 1834 by Justus
von Liebig for the condensation of cyanamides to several
amorphous C/N materials which he arbitrarily named melamine,
melam, melem and melon, with increasing temperature of forma-
tion.14 Yet it was very much later that the “possible formation
mechanisms of melem” were discussed.15,16 From literature, the
formation of the heptazine unit proceeds under formal deamina-
tion of the partaking species, and under ideal conditions no excess
of reactive molecules which would be available for side-reactions
is created. Note further that the condensation process requires no
more than a triazine and a carbonitrile group to proceed. Hence,
a self-propagating condensation mechanism employing a single
molecular precursor which contains both the triazine- and the
carbonitrile-functionality at its termini can be used (cf. Figure 1).
Since the reaction is performed in a closed system under constant
temperature all steps in the condensation process remain essen-
tially reversible.
stable, aromatic triazine unit drives the reaction to completion.
The aromatic carbonitriles with one triazine-functionality readily
precipitate from boiling 2-propanol after short reaction times (cf.
Experimental Section). Such a designed molecular precursor is
ready to undergo the thermal condensation outlined above, and
the particular 3-fold symmetry and planar arrangement of the
trisubstituted heptazine unit then allows the synthesis of planar,
2D frameworks which are expected to stack in a graphitic fashion
(cf. Figure 1).
However, the formation of Liebig’s melem in this thermally
induced condensation reaction was reported to be at around 370
and 390 °C, which was verified by a differential scanning calori-
metry study on the condensation of cyanamide (cf. Supporting
Information).15-17 This discards all standard solvents. Kanatzidis
and co-workers have previously pointed out the limited choice of
appropriate media for synthetic applications at intermediate
temperatures (i.e., 150-350 °C) and successfully applied alkali-
metal polysulfide melts in the synthesis of low-dimensional
ternary chalcogenides.18 The eutectic mixture of lithium and
potassium halides (e.g., LiCl/KCl 45:55 wt %, Tm=352 °C, or
LiBr/KBr 52:48 wt %, Tm=348 °C)19 has been known for some
time as a medium for electrochemical processes, in particular in
high-temperature galvanic cells20 and more recently as a solvent
for salts of lanthanides and actinides.21 Sundermeyer et al.
showed already in the 1960s that known organic chemistry can
be performed in molten salts.22,23 Among the successfully synthe-
sized compounds were carbonyl and fluorocarbonyl pseudohalo-
genides24 as well as cyanides, cyanates and thiocyanates of both
silicon and carbon.25 In this context, Sundermeyer explicitly
points out the good solvating properties of the eutectic mixture
of LiCl and KCl with respect to nitrides, carbides, cyanides,
cyanates, and thiocyanates, and we have previously used the
LiCl/KCl eutectic as a solvent in the synthesis of a crystalline,
graphitic carbon nitride species.4 For our means the LiBr/
KBr eutectic was identified as a good solvent due to its high-
temperature stability and a melting point below the polyconden-
sation point of s-heptazine. From laboratory experience, it was
given preference over the chloride analogue due to its superior
solvation properties for the small molecular precursors employed
in the synthesis and subsequent aggregates of higher molecular
weight.
The nature of the group R chosen as a linker between the
triazine- and the carbonitrile-functionality (as seen in Figure 1) is
accessible to the synthetic chemist. The compounds used are
bisubstituted aromatics, preferably with two cyano-groups in
para positions. One of the nitriles can then be converted into a
In the work at hand, we report the synthesis of three molecular
building blocks based on 1,4-aryl, 4,40-biphenyl and 1,3-aryl each
containing a triazine and a carbonitrile functionality and the
thermally induced autocondensation of these into three respec-
tive heptazine-based, ordered, 2D, layered polymer frameworks
(HBFs) in an eutectic salt melt of lithium and potassium bromide
*Corresponding author. Telephone: þ49 331 567 9569. E-mail:
m.j.bojdys.02@cantab.net.
r 2010 American Chemical Society
Published on Web 07/28/2010
pubs.acs.org/Macromolecules

6640 Macromolecules, Vol. 43, No. 16, 2010
Bojdys et al.
Figure 1. (Upper part, boxed) Proposed reaction mechanism for the condensation of melem.15,16 (Upper part) Formal condensation mechanism of a
monomer containing a triazine- and carbonitrile functionality to a heptazine based framework. (Lower part) Condensation motifs of 2D heptazine-
based frameworks (HBFs). The heptazine linker is shown as a planar triangle (blue) and the aromatic spacer as a coplanar rectangle (red), with 1, 2, and
3 being constituents of the precursors ArCNTz, BiPhCNTz and 1,3-ArCNTz and of the frameworks HBF-1, HBF-2, and HBF-3, respectively.
as solvent of choice. We thereby show that by making use of
heptazine-chemistry, the toolbox of the synthetic chemist can be
expanded to link organic building blocks entirely by strong
covalent bonds and to give not only chemically and thermally
stable, but also well-ordered, lightweight materials.
and washed with cold water until neutral. After removal of the
solvent and drying in vacuo, 4-(4,6-diamino-1,3,5-triazin-2-yl)-
benzonitrile was obtained as a white powder. Yield: 7.50 g, 35.3
mmol, 92.4% for C10H8N6. Anal. Calcd for C10H8N6: C, 56.60;
H, 3.80; N, 39.60. Found: C, 56.39; H, 3.79; N, 36.82. In addi-
tion, the compound was verified as pure by 1H and 13C NMR
(cf. Supporting Information).
Synthesis of HBF-1. A quartz ampule of the dimensions l =
120 mm, o.d. = 30 mm, i.d. = 27 mm was charged with 4-(4,6-
diamino-1,3,5-triazin-2-yl)benzonitrile (0.50 g, 2.4 mmol) and a
mixture of lithium bromide and potassium bromide (3 g, 52:48
wt %) under inert atmosphere. The ampule was sealed under
vacuum, heated and kept at the terminal temperature of 430 °C
for 48 h. After cooling, the homogenously brown-black block of
salt and product was stirred for overnight in distilled water to
Experimental Section
Synthesis of 4-(4,6-Diamino-1,3,5-triazin-2-yl)benzonitrile
(ArCNTz). 1,4-Dicyanobenzene (5.0 g, 38.2 mmol, 98%) and
dicyandiamide (3.25 g, 38.2 mmol, 99%) were heated to 105 °C
in 2-propanol (140 cm3). KOH (0.36 g, 6.4 mmol) was dissolved
in 2-propanol (10 cm3) and added dropwise to the reaction
mixture which was then refluxed at 105 °C for 1 h. The solution
was poured into cold water and the white precipitate collected

Article
Macromolecules, Vol. 43, No. 16, 2010 6641
dissolve the salt, and was subsequently washed with water, THF
and acetone to remove residual salt and poorly condensed
residues. HBF-1 was obtained as a brown-black powder and
dried in vacuo at 150 °C. Yield: 0.27 g, 0.47 mmol, 58.8% for
C30H12N14. Anal. Calcd for C30H12N14: C, 63.38; H, 2.52; N,
34.49. Found: C, 57.84; H, 2.52; N, 33.81.
The high nitrogen contents observed for the three networks
discount a multitude of possible binding motifs which are not
based on heptazine units e.g., a binding motif based on triazine
;
(C3N3) linkers, as seen for the covalent triazine frameworks
CTF-1 and CTF-2,5,26 would imply lower nitrogen contents for
each framework (i.e., nitrogen content for HBF-1 would be 21.86
wt % as to observed 33.81 wt %, for HBF-2 would be 12.27 wt %
as to observed 23.00 wt %, and for HBF-3 would be 21.86 wt %
as to observed 31.72 wt %). With thermogravimetric analysis
yielding residual masses of approximately 1-1.5% at 900 °C
under synthetic air for the three respective systems, it is highly
unlikely that the inorganic components of the eutectic salt melt
are regularly incorporated in the framework (cf. Supporting
Information).
Extensively progressed condensation and the covalent linking
of the frameworks by heptazineunits were additionally probed by
attenuated total reflection Fourier transform infrared (ATR FT-
IR) and magic angle spinning nuclear magnetic resonance (MAS
NMR) spectroscopy. Vibrational spectra of HBF-1, HBF-2
and HBF-3 showed that the broad ammonia bands at around
3200 cm-1, indicative of secondary and primary amines (and
their intermolecular hydrogen bonding) which are still present for
the mono- and oligomeric building blocks at the start of the
reaction and at lower temperatures, next to disappear for the high
temperature condensation products. Since aromatic C-H bonds
are an integral part of the aromatic linkers, some remaining
bands in this region are expected. The carbonitrile bands around
2250 cm-1 also gradually disappear as the condensation is driven
to completion. The aromatic C-C band at 1600 cm-1 remains
visible throughout the process, which corroborates the results
from elemental microanalysis that no aromatic groups are
eliminated. The characteristic breathing mode of the triazine
and heptazine units at 800 cm-1 is also preserved throughout the
process. Combined with the results of the elemental microana-
lysis, it is safe to assume that the material we are looking at consist
of condensed domains at least on the nanometer scale with few
terminal groups or defects (cf. Figure 2).
Synthesis of 40-(4,6-Diamino-1,3,5-triazin-2-yl)biphenyl-4-
carbonitrile (BiPhCNTz). 4,40-Biphenyldicarbonitrile (3.0 g,
14.25 mmol, 97%) and dicyandiamide (1.21 g, 14.24 mmol,
99%) were heated to 110 °C in 2-propanol (100 cm3). KOH
(0.13 g, 2.32 mmol) was dissolved in 2-propanol (10 cm3) and
added dropwise to the reaction mixture which was then refluxed
at 110 °C for 18 h. The solution was poured into cold water and
the yellow precipitate collected and washed with cold water until
neutral. After removal of the solvent and drying in vacuo, 40-(4,6-
diamino-1,3,5-triazin-2-yl)biphenyl-4-carbonitrile was obtained
as a yellow powder. Yield: 3.86 g, 13.39 mmol, 94.0% for C16-
H12N6. Anal. Calcd for C16H12N6: C, 66.66; H, 4.20; N, 29.15.
Found: C, 65.61; H, 4.27; N, 26.30. In addition, the compound
was verified as pure by 1H and 13C NMR (cf. Supporting
Information).
Synthesis of HBF-2. A quartz ampule of the dimensions l =
120 mm, o.d. = 30 mm, i.d. = 27 mm was charged with 40-(4,6-
diamino-1,3,5-triazin-2-yl)biphenyl-4-carbonitrile (0.34 g, 1.18
mmol) and a mixture of lithium bromide and potassium bro-
mide (5 g, 52:48 wt %) under inert atmosphere. The ampule was
sealed under vacuum, heated and kept at the terminal tempera-
ture of 400 °C for 24 h. After cooling, the homogenously black
block of salt and product was stirred for overnight in distilled
water to dissolve the salt, and was subsequently washed with
water, THF and acetone to remove residual salt and poorly
condensed residues. HBF-2 was obtained as a black powder and
dried in vacuo at 150 °C. Yield: 0.23 g, 0.29 mmol, 74.6% for
C48H24N14. Anal. Calcd for C48H24N14: C, 72.35; H, 3.04; N,
24.61. Found: C, 67.37; H, 3.56; N, 23.00.
Synthesis of 3-(4,6-Diamino-1,3,5-triazin-2-yl)benzonitrile
(1,3-ArCNTz). 1,3-Dicyanobenzene (5.0 g, 38.2 mmol, 98%)
and dicyandiamide (3.25 g, 38.2 mmol, 99%) were heated to
110 °C in 2-propanol (130 cm3). KOH (0.36 g, 6.4 mmol) was
dissolved in 2-propanol (10 cm3) and added dropwise to the
reaction mixture which was then refluxed at 110 °C for 12 h. The
solution was poured into cold water and the white precipitate
collected and washed with cold water until neutral. After
removal of the solvent and drying in vacuo, 3-(4,6-diamino-
1,3,5-triazin-2-yl)benzonitrile was obtained as a white powder.
Yield: 7.82 g, 36.85 mmol, 96.5% for C10H8N6. Anal. Calcd for
C10H8N6: C, 56.60; H, 3.80; N, 39.60. Found: C, 56.26; H, 3.86;
N, 36.68. In addition, the compound was verified as pure by 1H
and 13C NMR (cf. Supporting Information).
Synthesis of HBF-3. A quartz ampule of the dimensions l =
120 mm, o.d. = 30 mm, i.d. = 27 mm was charged with 3-(4,6-
diamino-1,3,5-triazin-2-yl)benzonitrile (0.50 g, 2.4 mmol) and a
mixture of lithium bromide and potassium bromide (3 g, 52:48
wt %) under inert atmosphere. The ampule was sealed under
vacuum, heated and kept at the terminal temperature of 400 °C
for 48 h. After cooling, the homogenously brown-black block of
salt and product was stirred for overnight in distilled water to
dissolve the salt, and was subsequently washed with water, THF
and acetone to remove residual salt and poorly condensed
residues. HBF-3 was obtained as a brown-black powder and
dried in vacuo at 150 °C. Yield: 0.28 g, 0.49 mmol, 61.3% for
C30H12N14. Anal. Calcd for C30H12N14: C, 63.38; H, 2.13; N,
34.49. Found: C, 57.62; H, 2.15; N, 31.72.
The solid-state 13C MAS NMR spectra for HBF-1, HBF-2,
and HBF-3 are presented in Figure 3. The resonances observed at
approximately 169 and 163 ppm for each framework are indica-
tive of the two different carbon environments expected for a
trisubstituted heptazine, with the former connected directly to the
aryl linker and the latter forming the core of the heptazine unit.
The resonances of the bridging aromatic linkers are situated
between 140 and 120 ppm. The observed number of chemical
environments corresponds to the expected number for heptazine-
based frameworks and are in agreement with previous reports in
literature for heptazine-based materials.27-30
To assess the crystal structure of HBF-1, HBF-2, and HBF-3
the powder X-ray patterns of the best-ordered condensation
products were chosen, namely the most thoroughly condensed
products obtained at 430, 410, and 400 °C, respectively. The
observed peak positions for HBF-1 generally give a good corre-
lation with the calculated in-plane Bragg peak positions of
hexagonal unit sheet with the parameters (cf. Supporting In-
formation). The hexagonal unit cell is populated in the a-b plane
with heptazine units bridged by phenylene linkers (cf. Figure 4)
giving chemically sound bond lengths and angles. In this model
unit cell the first low-angle peak is assigned to an in-plane
˚
reflection, yielding unit cell parameters of a = b = 18.590 A.
This is much larger than the unit cell parameters calculated and
observed for the corresponding triazine (C3N3) bridged network
Results and Discussion
5
˚
(i.e., CTF-1, a = b = 14.547 A) or for the isoelectronic borozine
31
˚
Elemental analysis of HBF-1, HBF-2, and HBF-3 gives the
(B3O3) bridged network (i.e., COF-1, a = b = 15.420 A)
chemical formulas C30.0N15.04H17.5, C48.0N14.05H30.2, and C30.0
N14.16H13.3 which is in good correspondence to the theoretical
values of C30N14H12, C48N14H24, and C30N14H12, respectively.
-
further supporting the formation of the larger heptazine unit in
the frameworks. The packing motif of these sheets, however, is
not trivial to deduce, although the principal packing peak at

6642 Macromolecules, Vol. 43, No. 16, 2010
Bojdys et al.
Figure 2. FTIR spectra of different condensation stages of HBF-1, HBF-2 and HBF-3 and their respective monomers (A, B, and C, respectively) in
LiBr/KBr melt at varying temperatures.
Figure 3. 13C MAS NMR spectra of (A) HBF-1, (B) HBF-2, and (C) HBF-3 with indicated spinning sidebands (/ in black).

Article
Macromolecules, Vol. 43, No. 16, 2010 6643
Figure 4. Observed (black) and calculated (blue) PXRD profiles with Bragg peak positions (green) of (A) HBF-1, (B) HBF-2, and (C) HBF-3.
Calculated reflections with a mixed-in l-component are marked (/ in blue). Underlying atomic connectivity pattern with carbon, nitrogen, and
hydrogen is represented as gray, blue, and white spheres, respectively.
26-27° in terms of 2θ for HBF-1, corresponds to approximately
from one sheet to the next, varying not only by translation but
potentially also by torsion. It is also a distinct possibility that
several domains of varying degrees of interplanar order are
present at any one-time for each framework, rendering them
“phase impure” with respect to the c-direction of each unit-cell.
Nonetheless, the near-equivalence in (hk0) reflections for each
system and a chemically reasonable occupation of any singular
sheet supports the assertion that the synthesized HBF materials
are sufficiently well described by the modeled structures (cf.
Figure 4 and Supporting Information). Assuming the calculated
unit-cell parameters in a- and b-direction given above and in the
Supporting Information a layer-to-layer packing distance of
˚
3.3 A an interplanar stacking distance of the order of magni-
;
˚
tude of graphite (3.35 A).
Also HBF-2 and HBF-3 show crystallinity in the XRD
pattern, while especially the broad peaks in HBF-2 point to a
rather ill-defined crystal structure. Unfortunately, there is no
chemically intuitive, obvious pattern and no clear hint from unit-
cell indexing (see Supporting Information) as to how the indivi-
dual sheet should be related. This is especially true for HBF-2 for
which the linking biphenyl group will experience steric repulsion,
and as a result most likely will twist out of plane. Thus, crystal-
lization of the HBF-2frameworkto long-range ordered crystals is
inherently difficult. It can be speculated whether a singular
monodomain of long-range order in the c-direction can be
achieved at all for systems with a very weak interplanar correla-
tion and a near-infinite number of possibilities of how to relate
˚
approximately 3.3 A, the theoretical densities of HBF-1, HBF-
2, and HBF-3 are 0.803, 0.592, and 1.055 g cm-3, respectively.
Volumetric density measurements for the dried frameworks gave
values of 0.984, 0.777, and 1.033, which further corroborates the

6644 Macromolecules, Vol. 43, No. 16, 2010
Bojdys et al.
Figure 5. Representative SEM images of HBF-1, HBF-2, and HBF-3 (A1, A2, and A3, respectively) and TEM images, showing (B1) TEM image
ꢀ
showing a Moire pattern and (B1, inset) the Fourier transform as observed for HBF-1, (B2) TEM image and (B2, inset) the Fourier transform showing
a packing repeat of layers as observed for HBF-2, and (B3) TEM image and (B3, inset) the Fourier transform showing a packing repeat of layers as
observed for HBF-3.
structural model and the fact that the described pores are
essentially empty.
(cf. Supporting Information). In each case the nanometer-sized
superstructures remained visible, indicating, that they are inter-
ꢀ
Scanning electron microscopy (SEM and transmission elec-
tron microscopy (TEM) reveal further indicators of the atomic,
molecular and supramolecular order of the frameworks. The
selection of images of HBF-1 in Figure 5 shows lamellar features
(A1 and B1), indicative of the interplanar correlation. However,
the sheer order of magnitude of the distances between ordered
repeats is in the range of several nanometers, i.e., an n-fold of the
modeled unit-cell dimensions in each direction. In order to
discard any artifacts produced by camera or substrate, pictures
were taken at different camera lengths and magnifications and
with samples lying directly on carbon support as well as off it
ference (Moire) patterns of real, atomic (or molecular) highly
ordered structures. Torsional distortion of layers which can freely
rotate with respect to one-another as mentioned previously in
the structural discussion based on PXRD data is a likely cause
of Moire interference and was reported previously for different
;
;
ꢀ
systems.32-34 Parts B2 and B3 of Figure 5 show the representative
TEM images of HBF-2 and HBF-3, respectively. Although the
outer shape of the aggregates in the SEM overviews (cf. Figure 5,
parts A2 and A3) does not seem to reveal any sharp edges known
for crystalline systems, higher resolved TEM images indicate
an ordering on molecular length scale with parallel “fringes”

Article
Macromolecules, Vol. 43, No. 16, 2010 6645
separated by 1.2 nm (approximately four layers) for HBF-2
(Figure 5, part B2) and 0.64 nm (approximately two layers) for
HBF-3 (Figure 5, part B3).
Acknowledgment. This work was funded by the Project
House “ENERCHEM” of the Max Planck Society. We would
€
like to thank Soren Selve from the TU Berlin for his aid and
patience with TEM imaging and Regina Rothe from the Max
Planck Institute of Colloids and Interface for the water adsorp-
tion measurements.
The architectural stability and porosity of HBF-1, HBF-2, and
HBF-3 was initially studied by measuring nitrogen adsorption,
however the results in the order of magnitude of approximately
30 m2 g-1 external surface area indicated that the pores are
essentially not accessible from the outside. This is a standard
problem of organic frameworks where smaller amounts of
organic side products or monomers can easily block a majority
of the channel system in a dynamic fashion. The fact that the
pores are in principle open to the outside can be deduced from the
fact that we were able to remove all but tiny traces of salt by
simple washing with water.
Supporting Information Available: Figures showing NMR
spectra, FTIR spectra, TGA plots, TEM images, DSC plots,
and PXRD analysis, and a table of elemental microanalysis
data.This material is available free of charge via the Internet at
http://pubs.acs.org.
References and Notes
Subsequently, a water adsorption study was performed on the
guest-free materials, and commercially available graphite and
clay (kaolinite) were chosen as reference systems. Samples of
“as-synthesized” HBF-1, HBF-2, HBF-3, graphite and clay were
evacuated at 10-5 Torr vacuum pressure and heated to 180 °C for
12 h to remove solvents. The samples were then used for
measurement of the isotherm from 0 to 1 bar water vapor
pressure, which shows gradual uptake at p/p0 from 0.05 to 0.6,
and some features of accessible micropores for HBF-3 (cf.
Supporting Information). The slow rise in the isotherm occurring
at higher pressures is due to the existence of a small population of
external mesopores between the crystallites; this feature is not
uncommon for particles with platelet morphology.35 Unfortu-
nately, the Brunauer-Emmett-Teller (BET) model cannot be
reliably applied to water adsorption, so no apparent surface area
can be given. A powder X-ray diffraction experiment was
performed to monitor the effect of water on the dried and
evacuated frameworks. Although it should be noted, that the
general intensity of the all reflections is attenuated due to low
contrast between the constituent light atoms (C, N, H) and water,
the general finding is that the uptake of water is accompanied
by a broadening of the peak at 26 to 27° (in terms of 2θ), which
was identified as an indicator of layer-to-layer ordering, while the
principal (hk0) reflection persist (cf. Supporting Information).
This suggests that the water uptake is accompanied by
swelling and disordering of individual layers of the framework.
Overall, this study shows that the materials HBF-1, HBF-2,
and HBF-3 behave analogous to previously known layered
materials (especially clays) with respect to water uptake, and
that these laterally continuous polymers can principally be
separated into mono- or at least oligo-layers by physical
exfoliation.
(1) Geim, A. K.; Novoselov, K. S. Nat. Mater. 2007, 6 (3), 183–191.
(2) Yaghi, O. M.; O’Keeffe, M.; Ockwig, N. W.; Chae, H. K.;
Eddaoudi, M.; Kim, J. Nature 2003, 423 (6941), 705–714.
(3) El-Kaderi, H. M.; Hunt, J. R.; Mendoza-Cortes, J. L.; Cote, A. P.;
Taylor, R. E.; O’Keeffe, M.; Yaghi, O. M. Science 2007, 316 (5822),
268–272.
(4) Bojdys, M. J.; Muller, J. O.; Antonietti, M.; Thomas, A. Chem.
Eur. J. 2008, 14, 8177–8182.
;
(5) Kuhn, P.; Antonietti, M.; Thomas, A. Angew. Chem., Int. Ed. 2008,
47, 3450–3453.
(6) Kuhn, P.; Forget, A.; Su, D. S.; Thomas, A.; Antonietti, M. J. Am.
Chem. Soc. 2008, 130, 13331–13337.
(7) Blake, A. J.; Champness, N. R.;Hubberstey, P.; Li, W. S.;Withersby,
M. A.; Schroder, M. Coord. Chem. Rev. 1999, 183, 117–138.
(8) Colvin, V. L.; Goldstein, A. N.; Alivisatos, A. P. J. Am. Chem. Soc.
1992, 114, 5221–5230.
(9) Decher, G. Science 1997, 277 (5330), 1232–1237.
(10) Kim, Y. H. J. Polym. Sci., Polym. Chem. 1998, 36, 1685–1698.
(11) Voit, B. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 2505–2525.
(12) Sakamoto, J.; van Heijst, J.; Lukin, O.; Schluter, A. D. Angew.
Chem., Int. Ed. 2009, 48, 1030–1069.
(13) Yaghi, O. M.; Li, H. L.; Davis, C.; Richardson, D.; Groy, T. L. Acc.
Chem. Res. 1998, 31, 474–484.
(14) Liebig, J. v. Ann. Pharm. 1834, 10.
(15) Hosmane, R. S.; Rossman, M. A.; Leonard, N. J. J. Am. Chem.
Soc. 1982, 104, 5497–5499.
(16) May, H. J. Appl. Chem. 1959, 9, 340–344.
(17) Lotsch, B. V.; Schnick, W. Chem. Eur. J. 2007, 13, 4956–4968.
;
(18) Kanatzidis, M. G.; Park, Y. Chem. Mater. 1990, 2, 99–101.
(19) Solomons, C.; Goodkin, J.; Gardner, H. J.; Janz, G. J. J. Phys.
Chem. 1958, 62, 248–250.
(20) Laitinen, H. A.; Tischer, R. P.; Roe, D. K. J. Electrochem. Soc.
1960, 107, 546–555.
(21) Cassayre, L.; Serp, J.; Soucek, P.; Malmbeck, R.; Rebizant, J.;
Glatz, J. P. Electrochim. Acta 2007, 52, 7432–7437.
(22) Sundermeyer, W. Angew. Chem. 1965, 77, 241–258.
(23) Sundermeyer, W. Chem. uns. Zeit 1967, 1, 150–157.
(24) Verbeek, W.; Sundermeyer, W. Angew. Chem. 1967, 79, 860–861.
(25) Sundermeyer, W. Z. Anorg. Allgem. Chem. 1961, 310, 50–52.
(26) Bojdys, M. J.; Jeromenok, J.; Thomas, A.; Antonietti, M. Adv.
Mater. 2010, 22, 2202–2205.
Conclusion
In this manuscript, a simple and convenient pathway toward a
family of regular, two-dimensional polymer structures based on
heptazine units, HBF-1, HBF-2, and HBF-3, was presented.
Furthermore, and foremost, a new way of linking organic
building blocks into well-ordered aggregates purely by strong
covalent bonds in a manner which overcomes the thermodynamic
and kinetic limitations known from crystallization and conden-
sation processes is employed. The procedure to prepare the
triazine- and carbonitrile-functionalized precursors for thermally
induced autocondensation is likely to be easily adaptable to a
multitude of molecules of 2D or 3D symmetry. All three
synthesized frameworks and other possible heptazine-based
structures are intellectually appealing, and they hold exiting
prospects as lightweight materials optimized for low k-dielectrics,
thermal insulation and phonon damping. Further work has also
to address the question of pore opening by etching or purification
processes.
(27) Jurgens, B.; Irran, E.; Senker, J.; Kroll, P.; Muller, H.; Schnick, W.
J. Am. Chem. Soc. 2003, 125, 10288–10300.
(28) Miller, D. R.; Swenson, D. C.; Gillan, E. G. J. Am. Chem. Soc.
2004, 126, 5372–5373.
(29) El-Gamel, N. E. A.; Seyfarth, L.; Wagler, J.; Ehrenberg, H.;
Schwarz, M.; Senker, J.; Kroke, E. Chem. Eur. J. 2007, 13,
;
1158–1173.
(30) Holst, J. R.; Gillan, E. G. J. Am. Chem. Soc. 2008, 130, 7373–7379.
(31) Cote, A. P.; Benin, A. I.; Ockwig, N. W.; O’Keeffe, M.; Matzger,
A. J.; Yaghi, O. M. Science 2005, 310 (5751), 1166–1170.
(32) Bassett, G. A.; Menter, J. W.; Pashley, D. W. Proc. R. Soc. London,
Ser. a: Math. Phys. Sci. 1958, 246 (1246), 345.
(33) Dawson, I. M.; Follett, E. A. C. Proc. R. Soc. London, Ser. a:
Math. Phys. Sci. 1959, 253 (1274), 390.
(34) Gevers, R.; Delavignette, P.; Blank, H.; Amelinckx, S. Phys. Status
Solidi 1964, 4 (2), 383–410.
(35) Rouquerol, F.; Rouquerol, J.; Sing, K. Adsorption by Powders and
Porous Solids; Academic Press: London, 2002.