Martin Pfeffermann1, Renhao Dong2, Robert Graf1, Wojciech Zajaczkowski1, Tatiana Gorelik3, Wojciech Pisula1, Akimitsu Narita1, Klaus Müllen1, Xinliang Feng2. 1. Max Planck Institute for Polymer Research , Ackermannweg 10, D-55128 Mainz, Germany. 2. Center for Advancing Electronics Dresden and Department of Chemistry and Food Chemistry, Technische Universität Dresden , Mommsenstraße 4, D-01062 Dresden, Germany. 3. Institute for Physical Chemistry, Johannes Gutenberg Universität Mainz , Welderweg 11, D-55099 Mainz, Germany.
Abstract
Utilizing dynamic self-assembly and self-sorting to obtain large-area, molecularly precise monolayered structures represents a promising approach toward two-dimensional supramolecular organic frameworks (2D SOF) or 2D supramolecular polymers. So far, related approaches suffer from small domain sizes, fragility and weak long-range internal order. Here we report on the self-assembly of a host-guest enhanced donor-acceptor interaction, consisting of a tris(methoxynaphthyl)-substituted truxene spacer, and a naphthalene diimide substituted with N-methyl viologenyl moieties as donor and acceptor monomers, respectively, in combination with cucurbit[8]uril as host monomer toward monolayers of an unprecedented 2D SOF. Featuring orthogonal solubility, the participating molecules self-assemble at a liquid-liquid interface, yielding exceptionally large-area, insoluble films, which were analyzed by transmission electron microscopy, atomic force microscopy and optical microscopy to be monolayers with a thickness of 1.8 nm, homogeneously covering areas up to 0.25 cm(2), and featuring the ability to be free-standing over holes of 10 μm(2). Characterization with ultraviolet-visible absorption spectroscopy, solid-state nuclear magnetic resonance spectroscopy, infrared spectroscopy, and grazing incidence wide-angle X-ray scattering allowed for confirmation of a successful complexation of all three monomers toward an internal long-range order and gave indications to an expected hexagonal superstructure. Our results extend the existing variety of two-dimensional soft nanomaterials by a versatile supramolecular approach, whereas the possibility of varying the functional monomers is supposed to open adaptability to different applications like membranes, sensors, molecular sieves, and optoelectronics.
Utilizing dynamic self-assembly and self-sorting to obtain large-area, molecularly precise monolayered structures represents a promising approach toward two-dimensional supramolecular organic frameworks (2D SOF) or 2D supramolecular polymers. So far, related approaches suffer from small domain sizes, fragility and weak long-range internal order. Here we report on the self-assembly of a host-guest enhanced donor-acceptor interaction, consisting of a tris(methoxynaphthyl)-substituted truxene spacer, and a naphthalene diimide substituted with N-methyl viologenyl moieties as donor and acceptor monomers, respectively, in combination with cucurbit[8]uril as host monomer toward monolayers of an unprecedented 2D SOF. Featuring orthogonal solubility, the participating molecules self-assemble at a liquid-liquid interface, yielding exceptionally large-area, insoluble films, which were analyzed by transmission electron microscopy, atomic force microscopy and optical microscopy to be monolayers with a thickness of 1.8 nm, homogeneously covering areas up to 0.25 cm(2), and featuring the ability to be free-standing over holes of 10 μm(2). Characterization with ultraviolet-visible absorption spectroscopy, solid-state nuclear magnetic resonance spectroscopy, infrared spectroscopy, and grazing incidence wide-angle X-ray scattering allowed for confirmation of a successful complexation of all three monomers toward an internal long-range order and gave indications to an expected hexagonal superstructure. Our results extend the existing variety of two-dimensional soft nanomaterials by a versatile supramolecular approach, whereas the possibility of varying the functional monomers is supposed to open adaptability to different applications like membranes, sensors, molecular sieves, and optoelectronics.
Two-dimensional (2D)
soft nanomaterials, which are defined as atomically
or molecularly thin layers of atoms or molecules periodically arranged
in two orthogonal dimensions,[1−3] are attracting increasing interest
since the isolation of graphene as their most prominent example.[4−8] With their unique electrical, mechanical, thermal, and optical properties
in combination with large surface areas, they have high potential
for a variety of applications such as sensing, membranes, biotechnology,
catalysis, and (opto-)electronic devices.[1,9−11] For example, free-standing monolayers of covalent
organic frameworks (COFs),[6,7,12,13] as well as metal–organic
frameworks (MOFs),[14−18] have recently been reported as 2D polymers and supramolecular polymers
with applicability in templating,[12] photoelectric
conversion,[14] and electrocatalytic hydrogen
evolution.[15] Although promising results
have been achieved with the networks comprising covalent bonds, 2D
supramolecular polymers represent a promising approach toward molecularly
precise structures over large areas.[9] Apart
from the more sophisticated 2D MOFs with coordination bonds, 2D supramolecular
organic frameworks (SOFs) built with other noncovalent bonds represent
an incipient type of 2D supramolecular polymer. In this field, while
remarkable results were already published on the assembly of 2D SOFs
on liquid–solid interfaces, promoting hydrogen bonding as well
as van der Waals interactions,[19−21] only a few examples comprising
host–guest enhanced interactions were found to be feasible
for obtaining 2D SOFs.[22−24] More precisely, host–guest enhanced interactions
of dipoles[22,23] and donors and acceptors[24] by means of cucurbit[8]uril (CB[8]), respectively,
have been utilized to build strong complexes. This interaction motif
is able to sustain the planarity of supramolecularly interacting groups
by stacking the guests face on inside the cavity of CB[8] and can
therefore maintain a two-dimensional backbone of planar molecules,
preventing out-of-plane polymerization.In the case of the host–guest enhanced donor–acceptor
interactions in CB[8], association constants of Ka ≥ 1011 M–2 were
reported in literature,[26] depending on
the electron affinity of participating molecules.[27] Considering the reversibility of the supramolecular assembly
and its self-sorting behavior toward a thermodynamic minimum, SOFs
offer a promising approach to fabricate large-area and molecularly
precise 2D nanomaterials. Nevertheless, it is astounding to find thus
far only a few reports on the formation of 2D SOFs, which all failed
to demonstrate a long-range internal order.[22−24] To the best
of our knowledge, all previous 2D SOFs were prepared in aqueous media,
where they suffered from small domain sizes and low processability.
Moreover, the preparation of free-standing 2D SOFs without substrate
supports remains elusive, although it is imperative for their applications
as membranes, for example, in ultrasensitive pressure sensors and
size excluding films.[1,4,28] This
is presumably due to the intrinsic defects or a limited strength of
supramolecular interactions, which cannot solely sustain the SOF structures,
leading to fragility.[22−24]Herein, we report the interface synthesis and
characterization
of unprecedented large-area free-standing monolayers of 2D SOF, based
on a host–guest enhanced donor–acceptor interaction
inside CB[8]. The C3-symmetric donor monomer
Np-Trx was provided by the electron rich, planar truxene (Trx) spacer
with three 6-methoxynaphthalene (Np) donor units attached to its periphery
(Figure ). The acceptor
molecule MV-NDI on the other hand was obtained by connecting two N-methyl-4,4′-bipyridin-1-ium (MV, from N-methyl viologenyl) acceptor moieties to an electron deficient naphthalene
diimide (NDI). Flexible ethylene spacers were introduced between the
NDI and MV units to compensate potentially occurring tensions in the
expected superstructure. The electron-donating and -withdrawing spacer
molecules were provided in the donor and acceptor monomer, respectively,
in order to maximize the interactions and to make our 2D SOF more
robust and capable of free-standing. By design, the self-assembly
of donor Np-Trx and acceptor MV-NDI inside the cavity of CB[8] is
expected to form a supramolecular honeycomb-network 2D SOF (Figure ). Furthermore, to
minimize tensions to the assembling 2D SOF and to prevent potentially
occurring defects to result in out-of-plane polymerization, the orthogonal
solubility of donor and acceptor monomers was utilized for applying
the stabilizing and 2D-templating effect of a liquid–liquid
interface, considering the possibility of face-on arrangements of
polyaromatic hydrocarbons to an interface.[16,29] To our best knowledge, this is the first example of a 2D SOF assembled
at an interface, which enabled the formation of insoluble 1.8 nm thick
monolayered films of exceptionally large areas of more than 0.25 cm2. We show an indication for successful complexation in the
monolayer and provide indirect proof of successful complexation in
solid state of bulk 2D SOF powder. Multilayered 2D SOF was found to
comprise a long-range internal order, which is in agreement with the
formation of a hexagonal arrangement of its superstructure.
Figure 1
Schematic illustration
of the assembly of the C3-symmetric, planar
donor molecule Np-Trx, the linear
acceptor molecule MV-NDI, and the host-molecule CB[8] toward the hexagonal
superstructure 2D SOF. The molecular structures of Np-Trx (left),
MV-NDI (middle), and CB[8] (right) are shown at the top. Below each
molecule, a schematic representation is visible with blue and red
parts representing the donor and acceptor moieties, respectively,
for visualizing the proposed superstructure. The inset represents
a proposed molecular model of one hexagon of the superstructure (geometry
optimized by UFF method[25]).
Schematic illustration
of the assembly of the C3-symmetric, planar
donor molecule Np-Trx, the linear
acceptor molecule MV-NDI, and the host-molecule CB[8] toward the hexagonal
superstructure 2D SOF. The molecular structures of Np-Trx (left),
MV-NDI (middle), and CB[8] (right) are shown at the top. Below each
molecule, a schematic representation is visible with blue and red
parts representing the donor and acceptor moieties, respectively,
for visualizing the proposed superstructure. The inset represents
a proposed molecular model of one hexagon of the superstructure (geometry
optimized by UFF method[25]).
Results and Discussion
Synthesis of 2D SOF
The donor precursor Np-Trx was
synthesized, starting with an acid catalyzed cyclotrimerization of
commercially available 5-bromo-1-indanone to obtain 2,7,12-tribromotruxene[30] in 84% yield. A Suzuki coupling reaction with
6-methoxy-2-naphthaleneboronic acid gave the target compound Np-Trx
after workup in yields of 70% (Scheme , top). Synthesis of the acceptor monomer MV-NDI was
performed commencing with commercially available 1,4,5,8-naphthalenetetracarboxylic
anhydride. Imidization with 2-aminoethanol and subsequent reaction
of the crude intermediate with bromine and phosphorus trichloride
gave N,N-bis(2-bromoethyl)naphthalene
diimide in 69%. Introduction of N-methyl-4,4′-bipyridin-1-ium
substituents in the periphery of N,N-bis(2-bromoethyl)naphthalene diimide was possible by nucleophilic
substitution of bromide with 4,4′-bipyridine, followed by methylation
with methyl iodide to obtain MV-NDI in 39% yield (Scheme , bottom). The molecular structure
of MV-NDI, CB[8], and Np-Trx was designed to feature orthogonal solubilites,
so the fabrication of 2D SOF was possible to perform at the interface
of water and toluene. For initial experiments, a common liquid–liquid
Langmuir–Blodgett trough was utilized for the synthesis of
2D SOF. However, due to a long time required for the assembly of the
components, that is, 4 h for a monolayer, and the low vapor pressure
of toluene in combination with the high surface area of the Langmuir–Blodgett
trough, it was not possible to reproducibly fabricate the 2D SOF.
Therefore, we turned to the use of a sealable, custom-made glass trough
for the assembly and deposition process (Figure S1, SI). The trough was placed on a vibration isolation platform
to avoid ruptures in the layer by peripheral vibrations. For the formation
of 2D SOF, an aqueous solution of preassembled MV-NDI with CB[8] was
provided in the glass trough, followed by overlaying a solution of
Np-Trx in toluene (Figure ). After the resulting interface was allowed to rest for assembly,
the two-phase system was drained through the porous glass bottom,
which was equipped with an appropriate substrate to allow for the
deposition of the resulting insoluble 2D SOF on its surface. Following
the assembly process online, for example, with a Wilhelmy-plate, was
not possible, for which depositions after incremental time spans were
performed and the time dependency of the resulting assembly was followed
by atomic force microscopy (Figure S2, SI). The ratio of concentrations of the utilized solutions was thereby
found to be crucial for obtaining a continuously homogeneous layer.
As the combination of MV-NDI with CB[8] in water is only able to build
a dumbbell shaped complex, but not to build a macroscopic assembly
without the presence of Np-Trx in toluene, the aqueous solution was
applied in excess to provide a high concentration of polymerizable
acceptor molecules at the interface, whereas the concentration of
Np-Trx was varied to control the regularity and size of the resulting
assembly. Specifically, when the concentration of Np-Trx is >4
×
10–5 mol/L in toluene, small fragments were obtained
from the assembly process, which tended toward multilayer formation
before interconnection of the fragments. In contrast, concentrations
<8 × 10–6 mol/L resulted in connected but
randomly perforated small layers. In the concentration range between
8 × 10–6 and 4 × 10–5 mol/L, homogeneous monolayers were achieved, where the concentration
of 2 × 10–5 mol/L was found to take advantage
of both a reproducible monolayer formation and a small assembly time
of 4 h.
Scheme 1
Reaction Paths To Obtain the Donor Monomer, Np-Trx (top), and
the
Acceptor Monomer, MV-NDI (bottom)
Figure 2
Schematic illustration of the self-assembly process and the deposition
procedure. The pictures depict the glass trough in the different states
of the process. (a) An aqueous solution of preassembled MV-NDI + CB[8]
is provided. (b) The two-phase system after addition of a solution
of Np-Trx in toluene is shown. (c) The self-assembled structure at
the interface after an assembly time of 4 h is visible. (d) The isolated,
assembled structure after draining the solutions through a valve at
the bottom is illustrated.
Schematic illustration of the self-assembly process and the deposition
procedure. The pictures depict the glass trough in the different states
of the process. (a) An aqueous solution of preassembled MV-NDI + CB[8]
is provided. (b) The two-phase system after addition of a solution
of Np-Trx in toluene is shown. (c) The self-assembled structure at
the interface after an assembly time of 4 h is visible. (d) The isolated,
assembled structure after draining the solutions through a valve at
the bottom is illustrated.
Atomic Force Microscopy (AFM), Optical Microscopy (OM), and
Transmission Electron Microscopy (TEM) Studies
The assembly
process over time was investigated by draining the assembling solutions
onto Si/SiO2-wafer substrates after incremental time spans
and subsequent AFM analysis, which reveals separated aggregates after
1 h, small porous domains after 2 h, a continuous but defective film
after 3 h, and finally a homogeneous film after 4 h (Figure S2, SI). Remarkably enough, AFM analysis on a scratched
film of the 2D SOF on a SiO2/Si-wafer, deposited after
4 h of assembly time, reveals a layer thickness of 1.8 nm (Figure a), which is in good
agreement with the outer diameter of one CB[8] molecule[31] and complies with the height expected for CB[8]-containing
polymers with CB[8] standing vertically on the substrate.[22,23] The presence of a monolayer is corroborated by persistent height
differences of 1.6 nm, which were measured during monitoring the assembly
process (Figure S12a–c, SI). These
results thus validate that the 2D SOF has been successfully obtained
as monolayer sheets. Large areas of homogeneously covered substrate
were visible by OM. Figure b shows a picture of monolayered 2D SOF, homogeneously covering
0.25 cm2 of Si/SiO2 wafer substrate. In order
to demonstrate the contrast with the bare substrate, the covering
layer was provided with a scratch. The homogeneity of the sheet is
corroborated by a small wrinkle of the layer, which is denoted with
an arrow. TEM analysis of 2D SOF was carried out on lacey-carbon supported
copper grids. The TEM image in Figure c reveals a thin free-standing film of 2D SOF homogeneously
covering an area of over 200 μm2 carbon support.
The large holes in the upper region are uncovered and serve as benchmark.
Remarkably, the deposited monolayer of 2D SOF was found to be free-standing
on the lacey-carbon support, covering holes of sizes up to ∼10
μm2 without rupturing, suggesting a significant mechanical
stability, which can be attributed to the strong supramolecular interaction
in combination with a long-range order in this system. This mechanical
stability is further corroborated by the canvas-like layer in Figure d. Although it is
connected to the carbon support at only one side and one point on
the opposite site, it still shows the free-standing feature. The wrinkled
edge probably occurred due to strain. The small dots visible on this
free-standing layer are probably dried residues from the precursor
monomers. The absence of split edges in both TEM images (Figure c,d) further corroborates
the presence of monolayered material. In accordance with other reported
monolayered material,[4,32−34] our 2D SOF
suffers from decomposition under high electron beam intensities, which
hinders high-resolution TEM measurements to probe the molecular structure.
Figure 3
(a) Tapping-mode
AFM height picture of the monolayered 2D SOF with
a scratch. The scale bar is 1 μm. The inset shows the cross-section
perpendicular to the scratch with a depth of 1.8 nm. For comparison,
the molecular structure of CB[8] with an outer diameter of 1.75 nm[31] is illustrated on the right, proving the monomolecular
thickness of the layer. (b) OM picture of 2D SOF, homogeneously covering
a Si/SiO2 wafer. The scratch was applied to reveal the
contrast with the bare surface, and the arrow indicates a wrinkle
of the layer for contrast. The scale bar is 150 μm. (c,d) TEM
images of 2D SOF on lacey-carbon supported copper grids. Panel c shows
a monolayer of 2D SOF, homogeneously covering large areas of the TEM
grid. Arrows point out uncovered areas. The scale bar is 5 μm.
(d) TEM-picture of slightly wrinkled, monolayered 2D SOF, free-standing
between the lacey-carbon support. The scale bar is 2 μm.
(a) Tapping-mode
AFM height picture of the monolayered 2D SOF with
a scratch. The scale bar is 1 μm. The inset shows the cross-section
perpendicular to the scratch with a depth of 1.8 nm. For comparison,
the molecular structure of CB[8] with an outer diameter of 1.75 nm[31] is illustrated on the right, proving the monomolecular
thickness of the layer. (b) OM picture of 2D SOF, homogeneously covering
a Si/SiO2 wafer. The scratch was applied to reveal the
contrast with the bare surface, and the arrow indicates a wrinkle
of the layer for contrast. The scale bar is 150 μm. (c,d) TEM
images of 2D SOF on lacey-carbon supported copper grids. Panel c shows
a monolayer of 2D SOF, homogeneously covering large areas of the TEM
grid. Arrows point out uncovered areas. The scale bar is 5 μm.
(d) TEM-picture of slightly wrinkled, monolayered 2D SOF, free-standing
between the lacey-carbon support. The scale bar is 2 μm.
Ultraviolet–Visible
(UV–vis) and Infrared (IR)
Absorption Spectroscopy
UV–vis absorption measurements
were carried out on a monolayer film of 2D SOF deposited onto a quartz
wafer surface (Figure a, black). For comparison, drop-cast films of its precursor Np-Trx
and preassembled precursors MV-NDI + CB[8] were also measured on quartz
wafers (Figure a,
blue and red, respectively). Figure a reveals a broad absorption of the 2D SOF from 600
to 1000 nm with apparently increased absorbance with respect to its
precursors, which can be assigned to a charge-transfer band of interacting
donor and acceptor molecules inside CB[8].[26,35,36] This result therefore indicates the efficient
complexation of monomers within 2D SOF. Notably, the absorption peaks
of the NDI units at 383 and 365 nm remain unshifted in 2D SOF, suggesting
that the NDI cores were not involved in the complexation. IR spectroscopy
of a bulk powder (see Supporting Information for preparation) was performed on 2D SOF in comparison with its
precursors for indirect proof of the stoichiometry. With IR spectroscopy
not being sufficiently sensitive in its rotational–vibrational
structure upon changes within this noncovalent complexation motif,
one expects the spectrum of 2D SOF to be a stoichiometrical superposition
of the individual spectra of participating components. Because the
spectrum of bulk 2D SOF (Figure b, black) is in agreement with the stoichiometrical
superposition of CB[8]/MV-NDI/Np-Trx = 6:3:2 (Figure b, blue), a quantitative presence of the
individual components in the 2D SOF can be assumed (see Figure S3, SI for the full spectrum).
Figure 4
Solid-state
characterizations of the insoluble 2D SOF. (a) Solid-state
UV–vis absorption spectra of Np-Trx (blue), MV-NDI + CB[8]
(red), and 2D SOF (black). To be comparable, the spectra of MV-NDI
+ CB[8] and 2D SOF were normalized to the 0 → 1 absorption
of the naphthalene diimide spacer, and the spectrum of Np-Trx was
normalized to its local maximum at 333 nm. (b) IR spectrum of 2D SOF
(black) in comparison with a spectrum formed by stoichiometric superposition
of those of Np-Trx, MV-NDI, and CB[8] (blue).
Solid-state
characterizations of the insoluble 2D SOF. (a) Solid-state
UV–vis absorption spectra of Np-Trx (blue), MV-NDI + CB[8]
(red), and 2D SOF (black). To be comparable, the spectra of MV-NDI
+ CB[8] and 2D SOF were normalized to the 0 → 1 absorption
of the naphthalene diimide spacer, and the spectrum of Np-Trx was
normalized to its local maximum at 333 nm. (b) IR spectrum of 2D SOF
(black) in comparison with a spectrum formed by stoichiometric superposition
of those of Np-Trx, MV-NDI, and CB[8] (blue).
Solid-State Nuclear Magnetic Resonance (SS-NMR) Analysis
The presence of a donor–acceptor interaction of monomers was
further supported by solid-state 13C{1H} correlation
NMR spectroscopy studies. Because the amount of 2D SOF material necessary
for solid state 2D 13C{1H} Lee–Goldburg (LG) heteronuclear
correlation (HETCOR) NMR spectroscopy was too high to be obtained
by deposition of monolayers with the general procedure, we used an
alternative method by stirring an aqueous solution of preassembled
CB[8] and MV-NDI vigorously with a solution of Np-Trx in toluene.
Thus, the liquid–liquid interface was increased and constantly
renewed to obtain a larger amount of bulk 2D SOF material by filtration.
Comparing the 2D 13C{1H} LG-HETCOR NMR spectrum
of the 2D SOF (Figure a, black) with those of its respective precursors (Figure b–d; color code, see Figure ), the correlation
signals in the spectrum of the complex are shifted compared with the
corresponding peaks in the correlation spectra of the precursors.
The monitored chemical shift changed upon the complexation, resulting
from modulations of the local electron density at the observed nuclear
spin site. Therefore, decreasing chemical shift values indicate a
stronger electronic shielding, which is induced by an increased electron
density. Therefore, the chemical shifts of the electron donor Np-Trx
can be expected to experience a low field shift upon complexation,
while those of the acceptor molecule MV-NDI should clearly exhibit
high field shifts at the sites involved in the complexation. The recorded 13C{1H} LG-HETCOR NMR spectra show correlation signals
of the Np-Trx precursor between 6 and 8 ppm in the 1H dimension
and between 120 and 130 ppm in the 13C dimension (Figure c, blue). These signals
observe a pronounced downfield shift in both dimensions, 1H as well as 13C, to the resulting chemical shift values
of 8–10 ppm in the 1H dimension and 125–135
ppm in the 13C dimension in the correlation spectrum of
2D SOF (Figure c,
black). The strong correlation signals of MV-NDI are observed in the
pure material at 9.2 ppm 1H chemical shift and a 13C chemical shift of 144 and 129 ppm, respectively.
Figure 5
(a) 13C{1H} Lee–Goldburg (LG) heteronuclear
correlation (HETCOR) spectrum of 2D SOF (color code, see Figure ). In order to illustrate
changes due to the complexation, superpositions of panel a with the
corresponding 13C{1H} LG-HETCOR spectra of the
pure precursors (b) MV-NDI, (c) Np-Trx, and (d) CB[8] are given.
(a) 13C{1H} Lee–Goldburg (LG) heteronuclear
correlation (HETCOR) spectrum of 2D SOF (color code, see Figure ). In order to illustrate
changes due to the complexation, superpositions of panel a with the
corresponding 13C{1H} LG-HETCOR spectra of the
pure precursors (b) MV-NDI, (c) Np-Trx, and (d) CB[8] are given.In contrast to the signals of
Np-Trx, the correlation signals of
MV-NDI are shifted upfield to 6.7 ppm 1H- and 130 ppm 13C-chemical shift and 6.3 ppm 1H- and 116 ppm 13C chemical shift, respectively (Figure b). These changes in chemical shifts correspond
to an increased electronic shielding of MV-NDI and a deshielding of
Np-Trx, which are expected to originate from the electron donating
effect of the methoxynaphthyl moiety of Np-Trx in combination with
the electron withdrawing behavior of the N-methyl-4,4′-bipyridinyl
group of MV-NDI. The changes in the signals of CB[8] upon complexation
to 2D SOF (see Figure d) are much weaker compared with the changes in MV-NDI and Np-Trx.
The correlation signal of the CH2 groups observed at ∼55
ppm in the 13C dimension as well as the carbonyl signals
observed at 157 ppm in the 13C dimension are almost unchanged
in 2D SOF compared with pure CB[8]. Additional solid state NMR experiments
probing the molecular mobility (recoupled polarization transfer heteronuclear
dipolar order (REPT-HDOR) sideband patterns see Figure S14, SI) were able to reveal rigidity of the CB[8]
molecules in 2D SOF, which we attribute to CB[8] hosting both acceptor
and donor moieties. The correlation signals of the CB[8]-CH sites
observed at 72 ppm in the 13C dimension, however, reveal
a heterogeneous behavior. The major contribution of the signal remains
unchanged, similar to the other signals of CB[8], which is in accordance
with the literature, whereas CB[8] generally exhibits only small peak
shifts upon comparable complexations.[22] Nevertheless, a small fraction of CB[8] experiences a clear deshielding
trend by being shifted 2 ppm in the 1H and 3 ppm in the 13C dimension to the low field. REPT-HDOR sideband patterns
were able to contribute this signal to a CB[8] species, hosting only
one guest inside its cavity (Figure S14, SI). With respect to its chemical shift, the guest is supposed to be
the acceptor moiety of MV-NDI. Therefore, this fraction corresponds
to defects or layer edges, which are expected in a higher amount due
to the deviating preparation method of the bulk 2D SOF powder. Finally,
the LG-HETCOR NMR spectrum of 2D SOF revealed very intense signals
of CB[8], medium signals for MV-NDI, and only weak signals for Np-Trx,
which was in qualitative agreement with the expected stoichiometry
of CB[8]/MV-NDI/Np-Trx = 6:3:2 in the proposed supramolecular assembly.
Even though the LG-HETCOR method is not quantitative, this stoichiometry
agrees with IR spectroscopy (Figure b). The visible fine structure of NMR spectra further
indicates the presence of ordered structure of 2D SOF (Figure b–d).
So far, our
attempts, including high-resolution TEM and scanning
tunneling microscope (STM) experiments, were not successful to elucidate
the internal structure of 2D SOF. High resolution TEM experiments
suffered from a visible decomposition of the layer under the necessarily
high irradiation doses, making it unfeasible to obtain the molecular
structure. We attribute the unsuccessful STM experiments on the other
side to a poor interaction between the 2D SOF and the substrate. With
regard to CB[8] having a small contact area with the substrate’s
surface while standing perpendicular on the substrate, in combination
with large distances between the CB[8] monomers, the 2D SOF probably
suffers from a weak interaction to the substrate. The 2D SOF is therefore
expected to feature high dynamics at the interface, precluding the
possibility to analyze the internal structure. Thus, to gain deeper
insight into the supramolecular organization within the 2D SOF layers,
grazing incidence wide-angle X-ray scattering (GIWAXS) was performed.
GIWAXS can provide details about the crystalline packing of weakly
scattering organic materials in thin films, whereas well-ordered films
typically exhibit diffraction patterns with a high number of intensive
elliptical reflections.[37] In contrast to
ordinary XRD measurements, GIWAXS probes the molecular packing in
both in-plane (q) and
out-of-plane (q) directions
with respect to the film surface. Conventionally, reflections along
the out-of-plane and in-plane direction appearing along the q axis and q axis, respectively, are indexed as
(00l), and (h00) and (0k0), respectively. The shape, intensity, and position of diffraction
peaks are related to the molecular packing within the film. Remarkably
enough, the GIWAXS pattern for a 122 nm thick 2D SOF multilayer film
(the preparation method was described in Supporting Information) reveals a highly ordered structure confirmed by
a large number of distinct reflections (Figure a). In order to exclude possible reflections
from the precursors, their GIWAX patterns were also analyzed for comparison
(Figure S4, bottom, SI). The meridional
reflections of 2D SOF localized at q = 0.395 and 0.422 Å–1 (Figure b) correspond to d-spacings of 1.59 and 1.48 nm and were assigned to the
(001) and (001′) interlayer planes, respectively. Following
reflections up to the fourth order indicates the long-range order
of the molecules in the out-of-plane direction of the film. We attribute
the d-spacing of 1.48 nm with respect to the diameter
of CB[8] of 1.75 nm[31] to a lamellar, Bernal-stacked
arrangement. The splitting of the meridional reflections might be
related to a marginally elliptical distortion of the CB[8] molecules,
which then show two different axes along and orthogonal to the planes
of the donor and acceptor moieties. The elliptical deformation of
the CB[8] is in agreement with 13C mobility measurements,
which show a restricted rotation of CB[8] as host around its guests
(Figure S14, SI). Reflections on the equatorial
plane (q) are typically
a combination of two lateral packing directions and are more difficult
to assign. Here we observed a pronounced periodicity, which suggests
for the first visible reflection in the small-angle range at q = 0.381 Å–1 the index of (300) (Figure b). Corresponding to a d-spacing of this
reflection of 1.64 nm, the first order reflection along this axis
then would have a d-spacing of 4.92 nm. A relatively
simple periodic sequence of reflections point to a high symmetry of
the structure within the film, which is either cubic or hexagonal
according to the theoretical peak distribution derived from Bragg’s
law. The cubic structure, however, would contradict the symmetry of
constituting molecules, which would rather adopt a hexagonal packing.
Indexing the 4.92 nm reflection as (100) within a hexagonal lattice
results in a unit cell parameter along this axis of 5.68 nm. The presence
of peaks up to 10th order within this indexing corroborates a well-organized
in-plane structure with a long-range order of molecules. The proposed
hexagonal unit cell of 5.68, 5.68, and 1.5 nm is in good agreement
with the 2D hexagonal structure schematically illustrated in Figure .
Figure 6
(a) GIWAXS pattern of
a layer-by-layer deposited, 122 nm thick
multilayer of 2D SOF. (b) One-dimensional integrations of the meridional
(top) and the equatorial (bottom) diffractions in panel a. Arrows
indicate the positions of the particular (hkl) scattering
intensities.
(a) GIWAXS pattern of
a layer-by-layer deposited, 122 nm thick
multilayer of 2D SOF. (b) One-dimensional integrations of the meridional
(top) and the equatorial (bottom) diffractions in panel a. Arrows
indicate the positions of the particular (hkl) scattering
intensities.
Conclusion
In
summary, we have successfully fabricated an unprecedented 2D
SOF by self-assembly of a host–guest enhanced donor–acceptor
interaction in the cavity of CB[8] at a liquid–liquid interface.
Monolayer films of 2D SOF with dimensions up to 0.25 cm2 were achievable by horizontal deposition and were found to be free-standing
over gaps of 10 μm2. We attribute the homogeneity
and the mechanical strength to the fundamental design of the molecules,
comprising electron withdrawing and donating effects, respectively,
to enhance donor–acceptor interactions in combination with
an intentionally introduced flexibility to one monomer, allowing a
strong complexation of the host–guest enhanced donor–acceptor
interaction toward a highly regular superstructure. Although it was
not possible to elucidate the internal structure unambiguously, analysis
of the X-ray scattering pattern indicates the formation of a hexagonal,
layered arrangement. Nevertheless, further efforts are necessary to
confirm the packing model unambiguously. The regularity of the 2D
superstructure in combination with the availability of free-standing
monolayers makes 2D SOF most promising for applications like an ultrasensitive
sensor or a size-excluding membrane. The facile, solution-based approach
is readily applicable to larger scales. Regarding the feasible exchangeability
of the truxene and the naphthalene diimide as implemented functional
spacers, respectively, with spacer molecules of different sizes, the
resulting 2D SOF derivatives are expected to be custom-tailorable
to a large variety of pore sizes. Furthermore, the visible charge-transfer
band in UV/vis spectroscopy indicates the rich optoelectronic properties
of 2D SOF, thus paving the way for using 2D SOFs in the field of optoelectronic
devices.[9,38,39]
Authors: Artur Ciesielski; Andrea Cadeddu; Carlos-Andres Palma; Adam Gorczyński; Violetta Patroniak; Marco Cecchini; Paolo Samorì Journal: Nanoscale Date: 2011-07-27 Impact factor: 7.790
Authors: Urs Rauwald; Frank Biedermann; Stéphanie Deroo; Carol V Robinson; Oren A Scherman Journal: J Phys Chem B Date: 2010-07-08 Impact factor: 2.991
Authors: Elizabeth L Magnotti; Spencer A Hughes; Rebecca S Dillard; Shengyuan Wang; Lillian Hough; Arshad Karumbamkandathil; Tianquan Lian; Joseph S Wall; Xiaobing Zuo; Elizabeth R Wright; Vincent P Conticello Journal: J Am Chem Soc Date: 2016-12-12 Impact factor: 15.419