Single-Crystal-to-Single-Crystal (SCSC) Linear Polymerization of a Desymmetrized Anthraphane

Article abstract of DOI:10.1002/chem.201802513

In this work we present one of the rare cases of single-crystal-to-single-crystal (SCSC) linear polymerizations, resulting in a novel ladder-type polymer. The polymerization is based on the photoinduced [4+4]-cycloaddition reactions between trifunctional

Full text of DOI:10.1002/chem.201802513

A Journal of
Accepted Article
Title: Single-Crystal-to-Single-Crystal (SCSC) Linear Polymerization of
a Desymmetrized Anthraphane
Authors: Marco Servalli, Nils Trapp, and Dieter Schlüter
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201802513
Link to VoR: http://dx.doi.org/10.1002/chem.201802513
Supported by

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
Single-Crystal-to-Single-Crystal (SCSC) Linear Polymerization of
a Desymmetrized Anthraphane
Marco Servalli*[a], Nils Trapp[b], A. Dieter Schlüter[c]
Abstract: In this work we present one of the rare cases of single-
crystal-to-single-crystal (SCSC) linear polymerizations, resulting in a
novel ladder-type polymer. The polymerization is based on the
photoinduced [4+4]-cycloaddition reactions between trifunctional
anthracene-based monomers. The careful design of the monomer
anthraphane-tri(OMe), results in perfectly stacked anthracene pairs in
the crystal structure, with Schmidt’s distances d = 3.505-3.666 Å and
shift s = 1.109 Å, allowing a selective linear polymerization in
quantitative yields and in a matter of minutes, without compromising
the integrity of the single crystals. The obtained polyanthraphane-
tri(OMe), reveals moreover a very interesting and unprecedented
quantitative yields and in a single-crystal-to-single-crystal (SCSC)
fashion[15]
Topochemical polymerizations have long been
reported[16–22] together with SCSC reactions[23,24]
however
.
,
topochemical linear polymerization which proceed in a SCSC
fashion resulting in covalent polymers are still a rarity in the
literature[15,25–29]. Encouraged by the rich and highly selective
photoreactivity of compound 1, we are currently widening our
activities towards simple derivatives of the parent compound,
whereby structural modifications are being considered which
could have a direct influence on the geometry of anthracene pairs
in the crystal structure. This way it was hoped to find new packing
case of stereoisomerism, which is characteristic for polyanthraphanes. motifs that could help to widen the product diversity of
anthraphanes in their SCSC reactions.
By analyzing the packings of parent compound 1 it became
evident that the molecules had a tendency to interdigitate in the
crystal structure, often resulting in anthracenes pairs slightly or
Introduction
completely laterally off-set, pushed closer than necessary to the
central benzene rings of the respective other monomer. The
Anthraphane 1 is a novel D3h-symmetric hydrocarbon cyclophane
equipped with three photoreactive anthracene units designed for
observed offsets not only rendered some packings unreactive, but
topochemical reactions in single crystals[1–4] (Figure 1). The
they also affected considerably the rate of photodimerization in
reactions are brought about by photochemical [4+4]-
reactive packings. We thought that this offsets could be sterically
cycloadditions between face-to-face (ftf) stacked anthracene
pairs[5–8]. Compound 1 is particularly interesting because its three
counteracted by introducing small groups to at least one of the
central benzene rings of 1. For symmetry reasons and synthetic
equally reactive sites cause a unique potential versatility in terms
feasibility considerations, three of these groups were envisaged,
of solid state reactivity. It can not only afford dimers and linear
ladder polymers[9,10], but also two-dimensional polymers[11–13] can
be envisaged, depending on how the anthracene units are
arranged in the crystal structure and whether one, two or even
resulting in the design of anthraphane-tri(OMe), 2 (Figure 1). In
this work, we present the facile synthesis of anthraphane 2 and
disclose two new packing motifs (etf packing 3 and etf/ftf packing
5) this compound assumes in the single crystal, one of which
three of these units involve themselves in dimerization reactions.
being a particular case of polymorphism and also being reactive,
We therefore started a research program initially concentrating on
the study of the crystallization and packing behavior of parent
anthraphane 1. Seven main packing motifs were found, three of
furnishing in a SCSC fashion the novel linear ladder polymer
poly1Danthraphane-tri(OMe) PA2. Interestingly, not only do the
main structural skeletons of PA1 and PA2 differ considerably, but
which photoreactive[14]. They gave either the dimer dianthraphane
it also appears that the methoxy substituents play a key role in the
or a novel linear polymer, the poly1Danthraphane, PA1, in
polymerization. In contrast to PA1, which passed through an
intermediate dimer state, the crystal packing of monomer 2 allows
for a better overlap in the anthracene pairs and, thus, for a higher
[a]
[b]
[c]
Dr. M. Servalli
Department of Materials
ETH Zurich
Vladimir-Prelog-Weg 5, 8093 Zurich (Switzerland)
E-mail: marcosendymion@gmail.com
Dr. N. Trapp
Small Molecule Crystallography Center (SMoCC)
Department of Chemistry and Applied Biosciences
ETH Zurich
polymerization reactivity. Intermediate states are not observed
and photopolymerization proceeds smoothly in a matter of
minutes.
Vladimir-Prelog-Weg 1, 8093 Zurich (Switzerland)
Prof. Dr. A. D. Schlüter
Department of Materials
ETH Zurich
Vladimir-Prelog-Weg 5, 8093 Zurich (Switzerland)
Supporting information for this article is given via a link at the end of
the document.((Please delete this text if not appropriate))
Figure 1. Chemical structure of anthraphane
counterpart anthraphane-tri(OMe) 2 synthesized in this study.
1 and its desymmetrized
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
Results and Discussion
Sonogashira reactions (Scheme 1). In particular, small
substituents were also desirable to not sterically interfere with the
final copper-free Sonogashira cross-coupling between the hexa-
substituted benzene 6 and 8. Moreover, to suppress the offset in
the anthracene pairs in the crystals, simple model considerations
suggested the methoxy substituents in compound 2 to be suitable.
Synthesis of anthraphane-tri(OMe)
It was desirable to introduce subsituents on the triethynylbenzene
moiety that did not involve dramatic changes to the original
synthetic protocol developped for anthraphane 1[4]. As such,
methoxy groups were chosen, due to their compatibility with the
desylilation step and the basic conditions of the required
Scheme 1. Synthesis of the new anthraphane 2. a) Route for the synthesis of the methoxy-substituted triethynylbenzene 6, the key building block required in the
last step. b) Sequential Sonogashira cross-coupling reactions of ditriflate 7 with the triethynylbenzene core and of the resulting tritriflate 8 with the substituted
triethynylbenzene core 6 from route a).
The synthesis of the trimethoxy-substituted benzene core 6 was
adapted and modified from an already published procedure from
Hennrich[30,31] (Scheme 1a): in a first step, 1,3,5-trifluorobenzene
was quantitatively iodinated with periodic acid in sulfuric acid to
give compound 3[32]. Triple nucleophilic aromatic substitution of
the fluorides in anhydrous DMI with freshly prepared sodium
methoxide yielded the methoxy-substituted iodinated compound
4 in 85% yield. For the Sonogashira cross-coupling of 4 with
trimethylsilylacetylene, the reaction conditions were slightly
changed due to the particularly poor reactivity of 4 towards
oxidative addition[33]. Instead of the standard Pd(PPh3)4, the more
were obtained in 39% yield in form of an amorphous powder. The
structure of anthraphane-tri(OMe)
2
was unambiguously
confirmed by 1H-NMR and 13C-NMR spectroscopy, high-
resolution mass spectrometry and by SC-XRD.
Crystallization of 2 and its packings in the single crystal
Desymmetrization of anthraphane 1 to compound 2 involved
addition of three methoxy groups and thus a reduction of
symmetry from D3h to the C3v point group; as such, an increase in
solubility was expected, which in fact was observed and allowed
to measure a fully resolved 13C-NMR spectrum in solution (Figure
S5). The best crystallization method was slow cooling of nearly
saturated solutions. The solvents employed were THF, dioxane,
chloroform, 1-methylnaphthalene and nitrobenzene. The
amorphous powder obtained from the work-up was suspended in
the desired solvent, which was then heated close to the boiling
point until a clear solution was obtained. This solution was then
cooled to room temperature over 24 h (for details see page S14
in the SI). Single crystals were obtained in all cases and analyzed
by SC-XRD: three main packing motifs were obtained,
summarized in Figure 2.
aggressive and nucleophilic catalyst Pd(PtBu3)2 was used[34]
,
which was generated in situ from Pd(OAc)2 and by deprotonation
of HPtBu3BF4 with DIPA. For such an electron-rich substrate, the
reaction worked very well with yields up to 65%. In a final step,
desylilation of 5 with potassium carbonate in methanol afforded
quantitatively the target compound 6 in high purity.
With the substituted core in hand, 2 was then assembled by using
the same conditions employed for the synthesis of anthraphane
(Scheme 1b). Conveniently, the final step worked as nicely as in
the original synthesis and the work-up to isolate the monomer did
not have to be changed: from 640 mg of precursor 8, 203 mg of 2
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
Figure 2. The three packing motifs obtained for compound 2: a) etf packing 1, b) etf packing 3 and c) mixed etf/ftf packing 5. The same packing nomenclature from
the hydrocarbon anthraphane 1 was used[14]. From top to bottom: top view of a layer in the crystal structures, layer arrangement in the crystal structure, solvents
from which the packing can be obtained, optical micrograph of representative single crystals. The photoreactive anthracene units are colored in red, solvent
molecules are omitted for clarity. Please note that the displayed etf packing 1 motif is based only on a fragmentary structure solution due to excessive disorder in
the crystals (see text); for that reason the model is represented as “capped sticks” and the methyl groups were omitted. The etf packing 3 and etf/ftf packing 5 are
instead real crystal structures.
Before going into an analysis of the three different packing motifs
obtained, a comment concerning crystal quality is appropriate. In
particular all the crystals belonging to the etf packing 1 had a high
degree of disorder both in the main molecule and solvent
molecules incorporated in the structure, preventing fully resolved
X-ray structure analysis. The packing displayed in Figure 2a is
based on fragmentary structure solution only and should not be
regarded as a real and complete crystal structure; despite the
fragmentary solution however, the edge-to-face motif however
could be clearly discerned. The packings displayed in Figure 2b
and 2c correspond instead to real and solved crystal structures.
The etf packing 1 (the nomenclature of the packing motifs
was kept from reference [14]) was obtained from THF, dioxane
and chloroform in the form of yellow needles. Hexagonal plates
were instead obtained from nitrobenzene in
a case of
polymorphism (see below). In this packing all the anthracene
units are arranged in an edge-to-face fashion and cannot
therefore photoreact. This packing motif was also a very common
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
occurrence for the parent hydrocarbon anthraphane 1 and we
were therefore not surprised by the outcome. As indicated before,
the desymmetrized 2 in this packing always resulted in poor
quality crystals, in which the molecule intrinsically does not seem
to have a preferred orientation: while it always packs in layers, in
each layer some molecules have the substituted core oriented in
one direction and some in the opposite direction. While the etf
motifs were clearly discernible, the reliability factors always
resulted too high for an acceptable structural model. Crystals from
this packing were also very sensitive upon manipulation, easily
losing solvent when taken out of their mother liquor and
completely losing crystallinity when rinsed with other solvents.
This is likely due to the porous nature of this packing. Using slow
vapor diffusion as crystallization method did not yield better
quality crystals.
(Figure S7). It can be reasoned that the small impurities present
in the amorphous powder obtained after the synthetic workup are
responsible for obtaining the photoreactive etf/ftf packing 5; in fact
this result could be reproduced from three different batches of 2.
Removing the impurities by recrystallizing a second time results
instead in the non-reactive packing. However, by recrystallizing
the hexagonal platelets of the etf packing 1 and seeding the
obtained clear solution with a rhombohedral single crystal from
the first crystallization step, the photoreactive etf/ftf packing 5
could be obtained again selectively. It is noteworthy that these
single crystals can be rinsed with methanol and dried without
losing crystallinity; after one month stored in a vial they could be
again measured by SC-XRD. Having the photoreactive crystals in
hand, we then proceeded to investigate their photoreactivity.
Better results were obtained by using 1-methylnaphthalene
as crystallization solvent: yellow cubes were obtained and upon
SC-XRD analysis, a new packing motif was discovered, the etf
packing 3 (Figure 2b). As the name suggests, in this packing there
is again an exclusive edge-to-face relationship between the
anthracene units of 2. There are parallel anthracene units in the
structure but they are too offset to be photoreactive. The voids
between the molecules are filled by solvent molecules; two
methylnaphthalene molecules are π-π stacking with each other
(for more detail see Figure S11). The overall structure can be
regarded as arrays of cyclophanes arranged in layers and held
together by solvent molecules and a large parallel displaced π-π
stacking interaction between anthracene moieties. Molecules in
arrays are stacked on top of each other and have a preferred
orientation of their trimethoxybenzene moieties: between
neighboring arrays the orientation alternates itself (Figure S12).
This is the only structure in this study in which the trimethoxy-
benzene units of the molecules have a clear preferred orientation.
Crystallization from nitrobenzene afforded yellow
rhombohedral plates which upon SC-XRD analysis also resulted
in a new packing motif: etf/ftf packing 5 (Figure 2c). In this packing
motif, each molecule has two anthracene units face-to-face
stacking with the anthracene units of its neighbors, potentially
leading to linear polymer chains upon photoreaction. The
molecules again are packed in layers and the void between them
are filled with nitrobenzene molecules (Figure S15). Two
nitrobenzenes are π-π stacking with each other and with the third
anthracene blade not involved in ftf stacking. These solvent
molecules are disordered in their orientation, trying to avoid
contact with the methoxy groups: the two orientations have
roughly 80 : 20 occupation. This is a consequence of the main
molecule also being disordered: from this it can be concluded that
in average, in the crystal structure 80% of the cyclophanes have
the trimethoxy-benzene rings oriented on one side and the
remaining 20% oriented on the opposite side.
Structural analysis and SCSC Photopolymerisation of 2
Crucial for a topochemical SCSC reaction is that the reactive units
are in close proximity and properly aligned and oriented in the
crystal structure. In the case of anthracene pairs, they have to ftf-
stack properly with good overlapping between the p-orbitals at
their reactive 9, 10 positions. A geometrical analysis of the
anthracene pairs can help to assess their photoreactivity: useful
parameters are the distance between the respective C9 and C10
positions d9,10’ and d10,9’ (the so-called Schmidt’s distances[35]) and
the shift s, which describes the planar shift of the anthracene rings’
centroids from an ideally coplanar superimposed anthracene pair
and is therefore a measure of the overall offset of the pair. Since
the two anthracene units of 2 are ftf-stacking equivalently in the
crystal structure, only one pair is displayed in Figure 3a. The
anthracenes are not exactly coplanar having a plane angle of
approximately 3°; consequently two different distances d9,10’ and
d10,9’ are present, 3.666 Å and 3.505 Å. The shift value s accounts
for roughly 1.109 Å, corresponding to a minor offset in the pair.
These values suggest a stacking particularly suitable for a SCSC
photoreaction. Compared to the previous study on 1 and the
resulting polymer PA1, the geometrical parameters of 2 are much
better: for 1, d values were ranging from 3.695 – 3.948 Å and shift
values from 1.354 – 1.904 Å, above which reactivity was not
observed[15]. High values were in fact shown to dramatically
influence the SCSC reaction kinetics, by slowing down the
reactivity of poorly overlapped pairs and consequently causing
cracks in the crystals during the reaction, due to the pronounced
molecular movement necessary for the pairs to better overlap and
dimerize. In the current case we therefore expected a smooth and
fast SCSC photopolymerisation, without having to pay too much
attention to the irradiation conditions, such as temperature and
time.
An interesting case of polymorphism was discovered when
handling this nitrobenzene solvate: in an attempt to increase the
quality of the single crystals from the first crystallization batch, we
decided to recrystallize them again from nitrobenzene. After the
second crystallization however, hexagonal platelets were
obtained which surprisingly corresponded to the etf packing 1
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
Figure 3. a) Geometrical parameters of the stacked anthracene pairs used to assess the photoreactivity; b) one-step SCSC linear photopolymerization of the
nitrobenzene solvate in the etf/ftf packing 5. The reaction proceeds with minimal mismatches between the unit cell parameters of the monomer crystal and the
polymer crystal, resulting in a smooth and fast polymerization at room temperature. Optical micrographs of the single crystals before and after polymerization are
displayed in bright field mode and between crossed polarizers. Consumption of the anthracene units during polymerization results in crystal discoloration.
It is perhaps noteworthy to mention the cause of the nice
stacking of the pairs of 2: in our previous crystallization studies of
1 it was found in certain packings that the molecules had the
tendency to interpenetrate resulting in offset anthracene pairs.
Interpenetration can go as far as the hydrogen atoms of the
central benzene cores start to sterically repel with the hydrogen
atoms of the anthracene edge. In 2 instead, the substitution of the
hydrogen atoms of the central benzene core with methoxy groups
prevents interdigitation resulting in nicely stacked anthracene
pairs (Figure 4). This appears like an interesting element of crystal
engineering[36,37] which could be applied in the future for this kind
of molecules.
diffractometer’s goniometer head, ready to be measured after
irradiation. The irradiation wavelength was chosen as 465 nm
(same for anthraphane, tail-end irradiation[38]). The results are
summarized in Figure 3b. Photopolymerizations turned out to be
very fast. Already within 15 min of irradiation at room temperature
monomer single crystals smoothly convert into polymer single
crystals of PA2. The unit cell parameters are minimally affected
by the reaction: the a-axis lengthens by 0.4%, the b-axis shortens
by 1.6%, the c-axis lengthens by 0.5% and β is the only angle
affected, increasing by 5%. These minimal changes are a
consequence of the nicely overlapping anthracene pairs, and
prevent the crystals from cracking during the reaction. The quality
of the crystals after irradiation slightly increased, allowing to
model additional solvent molecules in the crystal structure, which
had to be masked in the monomer crystal (Figure S20). In the
Irradiations were performed at room temperature both in
bulk with dry single crystals placed in a sealed vial under argon
(Figure S8) or with a single crystal mounted on the pin of a
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
particular crystal measured, again the main molecule appeared
disordered in its orientation: approximately 77% of the molecules
had their trimethoxybenzene cores oriented in one direction and
23% in the opposite direction. Polymerization also resulted in
visible discoloration of the crystals due to the consumption of the
anthracene chromophores.
Figure 4. . Schematic offset of the anthracene pairs in crystal structures of anthraphanes 1 (left) and 2 (right). Substituents at the central benzene cores prevent
interpenetration with consequent slipping of the anthracene pairs.
Face indexing shed light into the orientations of the chains
in the crystal (Figure 5), showing that they run along the longest
dimension of the rhombohedra being typically 180 µm, whereas
the layers stack along the shorter dimension of 150 µm. The
length of 180 µm would ideally corresponds to linear chains with
approximately 150’000 repeat units and molar masses of 136
MDa. We however do not expect having accomplished such
outstanding molar masses because of crystal defects and the
phenomenon of mosaicity[39,40], which lead to grain boundaries
which may or may not interrupt the polymerization reactions.
However, we highlight the fact that the formation of this linear
polymer is unique to the single crystalline state. If attempted in
solution, not only the selectivity towards linear growth would not
be ensured due to the monomer’s trifunctionality, but also the
molecular weights would be much lower, since rigid monomer
such as 2 tend to precipitate out of the solution already at the
oligomeric state.
ppm and 51.6 ppm corresponding to the bridge carbons of the
anthracene dimers are clearly visible (Figure S23). Not
surprisingly the polymerized crystals are completely insoluble;
additional characterization of the polymer would require
exfoliation in a suitable solvent which is currently in progress.
Preliminary experiments show that exposure to NMP at room
temperature breaks down the mother crystals into needles (Figure
S28).
[4+4]-Cycloadditions are known to be reversible. Not
surprisingly, poly1Danthraphane-tri(OMe) PA2 upon thermal
treatment underwent depolymerization as confirmed by both 13C-
CP/MAS NMR ATR-FTIR spectroscopy (Figure S23 and S26):
heating the crystals under argon at 200°C for 24 h resulted in
complete depolymerization but also resulted in loss of crystallinity.
The crystals maintained their shapes without turning into powder
but showed poor diffraction upon SC-XRD. As expected,
irradiation of the depolymerized crystals did not result in
repolymerization. Differential scanning calorimetry (DSC), also
confirmed the thermal retro-polymerization by showing a typical
exotherm appearing above 200°C (Figure S27).
The polymer crystals were further characterized by
fluorescence spectroscopy (Figure S22), 13C-CP/MAS NMR-, and
ATR-FTIR spectroscopy (Figure S24). In the solid-state NMR
spectrum of the polymer crystals, the characteristic peaks at 49.2
Figure 5. Face indexing of a rhombohedral single crystal mounted on the pin of a goniometer head showing how the polymer chains are oriented within it.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
Finally, we turn the attention to the architecture of the polymer
PA2 itself: while it might seem analogous to the polyanthraphane
PA1, there is a fundamental difference between both polymers.
This difference goes beyond the presence of the methoxy
functional groups and concerns how the repeat units are
connected together in terms of stereochemistry (Figure 6).
Anthracenes have two faces, both of which are potentially
accessible for stacking and consequent photodimerisation: in the
case of polyanthraphane PA1, in the crystal structure the two
faces engaged in ftf stacking are the ones pointing away (exo)
from the unreacted anthracene unit, resulting in a poly-(exo,exo)-
anthraphane; in the present study however, the stacking faces
were the ones pointing towards the unreacted anthracene (endo),
resulting in a poly-(endo,endo)-anthraphane-tri(OMe), PA2. This
isomerism results in chains being approximately 26% thicker in
the exo,exo polymer (≈ 30.1 Å) with respect to the endo,endo
case (≈ 23.8 Å). Consequently the chains are also 20% shorter in
the exo,exo polymer compared to the endo,endo one.
Figure 6. Stereoisomerism in polyanthraphanes affects the molecular structure of the polymer chains. Depending on the stacking preference of anthracene pairs,
exo,exo polymers or endo,endo polymers were obtained, each with their characteristic molecular dimensions. The reported dimensions account for van der Waals
radii.
polyanthraphane-tri(OMe), PA2, the second polyanthraphane
reported so far. We found that substitution with methoxy groups
Conclusions
at the central benzene core of the cyclophane, can sterically force
the anthracene pairs into a near-perfect stacking, resulting in a
very smooth and fast polymerization process, with minimal
changes in the unit cell parameters of the crystals.
Depolymerization by retro-[4+4]-cycloaddition was found to be
initiated at temperatures above 200°C. This study also highlights
an unprecedented case of isomerism characteristic of
polyanthraphanes, which can result in endo,endo or exo,exo
configured polymers depending on which face of the anthracene
moieties the dimerisation reaction takes place. We would expect
In conclusion, we synthesized the desymmetrized anthraphane-
tri(OMe) 2 using the new building block 6 while keeping the
synthetic protocol developed for the original hydrocarbon
anthraphane 1 unchanged. Compound 2 was crystallized from
different solvents resulting in two new packing motifs, etf packing
3 and the photoreactive etf/ftf packing 5, which add to the
previously reported seven packings known for anthraphane.
Irradiation of the etf/ftf packing 5 single crystals resulted in a
SCSC linear photopolymerization via [4+4]-cycloaddition of ftf-
stacked
anthracenes,
affording
single
crystals
of
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
1,3,5-triiodo-2,4,6-trimethoxybenzene 4
such differences to impact for instance the solubility of the chains,
where in the exo,exo polymer the free anthracene units would be
more exposed to solvent interaction and have more degrees of
freedom. Endo,exo or exo,endo chains are still elusive, but would
also of course widen the architectural palette of polyanthraphanes.
Apart from that, such structural differences might be of interest for
construction of nanostructures with well-defined molecular
dimensions involving the still unreacted anthracene units provided
in fixed relative distances and orientations. Last but not least, we
note that crystal packing effects do not seem to be sufficient to
clearly steer the orientation of the trimethoxybenzenes units of 2
in a defined way. Being able to control the orientation would open
access to piezoelectric polymers having chains with specific
Sodium hydride (0.94 g, 23.5 mmol, 6 eq, 60% dispersion in mineral oil)
was suspended in 10 mL dry diethyl ether in a 100 mL Schlenk flask.
MeOH (1.00 mL, 23.5 mmol, 6 eq) diluted in 10 mL dry diethyl ether was
slowly added to the hydride under vigorous stirring. The reaction mixture
was then stirred at room temperature for 1 h until no more hydrogen
evolution was detectable. The volatiles were removed by a stream of
nitrogen and the residue was dried on HV for 10 min. 20 mL dry DMI were
then added to the residue and to the resulting suspension 1,3,5-trifluoro-
2,4,6-triiodobenzene 3 (2.00 g, 3.92 mmol, 1 eq) was added in small
portions over 30 min under vigorous stirring (caution: some foaming during
this exothermic reaction is possible). The obtained orange suspension was
stirred overnight at room temperature and then poured into 80 mL
saturated NaHCO3 solution. The white precipitate was collected by
filtration and washed with water until the pH of the filtrate resulted neutral.
Recrystallization from boiling methanol afforded compound 4 as white
needles (1.56 g, 2.86 mmol, 73%). Rf (30% EtOAc in hexane): 0.72, Mp:
176-178°C. 1H-NMR (300 MHz, CDCl3) δ/ppm: 3.86 (s, 9H). 13C-NMR
(75.5 MHz, CDCl3) δ/ppm: 161.5, 82.8, 60.9. HRMS (FT-MALDI): m/z
calcd for C9H9I3O3 [M-H]+: 545.7680; found: 545.7680.
permanent dipole moments[41,42]
.
Experimental Section
Materials and Methods
((2,4,6-trimethoxybenzene-1,3,5-triyl)tris(ethyne-2,1-
diyl))tris(trimethylsilane) 5
All reactions were carried out under nitrogen by using standard Schlenk
techniques and dry solvents unless otherwise noted. Dry diethyl ether, dry
methanol, dry DMI and dry toluene were purchased over molecular sieves
from Acros and used directly. Diisopropyl amine was dried by passing it
over a column of activated neutral aluminum oxide. Pd(PPh3)4 catalyst was
freshly prepared and stored in a glove-box in the dark under N2 at room
temperature. All reagents were purchased from Acros, Aldrich or TCI, and
used without further purification. Column chromatography for purification
of the products was performed by using Merck silica gel Si60 (particle size
40-63 μm). NMR was recorded on a Bruker AVANCE (1H: 300 MHz, 13C:
75 MHz) at room temperature. The signal from the solvents was used as
internal standard for chemical shift (1H: δ = 7.26 ppm, 13C: δ = 77.16 ppm
for chloroform, 1H: δ = 6.00 ppm, 13C: δ = 73.78 ppm for 1,1,2,2-
tetrachloroethane, 19F: δ = -164.9 ppm for hexafluorobenzene). High
resolution mass spectroscopy (HRMS) analyses were performed by the
MS-service of the Laboratory for Organic Chemistry at ETH Zurich with
spectrometers (ESI- and MALDI-ICR-FTMS: IonSpec Ultima Instrument).
1,3,5-triiodo-2,4,6-trimethoxybenzene 4 (180 mg, 0.33 mmol, 1.00 eq) was
placed in a dry 20 mL Schlenk tube along with catalyst Pd(OAc)2 (4.00 mg,
0.02 mmol, 0.05 eq), co-catalyst CuI (3.00 mg, 0.02 mmol, 0.05 eq) and
ligand tri-tert-butylphosphonium tetrafluoroborate (9.00 mg, 0.03 mmol,
0.10 eq). In a separate Schlenk tube, 10 mL dry diisopropylamine were
degassed by four cycles of freeze-pump-thaw and then added by syringe
to the reactants. Trimethylsilylacetylene (0.20 mL, 1.65 mmol, 5 eq) was
added and the reaction mixture was sealed and heated to 65°C for 48 h
under an argon atmosphere. Formation of a beige precipitate indicated the
start of the reaction. After cooling to room temperature, the reaction
mixture was filtered through a celite pad and concentrated to dryness.
Separation by column chromatography (3% EtOAc in hexane) afforded the
title compound 5 as a yellowish solid of purities high enough for the next
step (98 mg, 0.21 mmol, 65%). Rf: 0.37, Mp: 82-84°C. 1H-NMR (300 MHz,
CDCl3) δ/ppm: 4.02 (s, 9H), 0.24 (s, 27H). 13C-NMR (75.5 MHz, CDCl3)
δ/ppm: 163.5, 108.3, 103.5, 96.1, 59.3, 0.0. HRMS (FT-MALDI): m/z calcd
for C24H36O3Si3 [M]+: 456.1972; found: 456.1970.
Synthetic Procedures
1,3,5-triethynyl-2,4,6-trimethoxybenzene 6
Compound 5 (0.20 g, 0.44 mmol, 1 eq) was dissolved in a solvent mixture
of 5 mL THF and 2 mL MeOH. Potassium carbonate (7.00 mg, 0.05 mmol,
0.12 eq) and 0.5 mL H2O were added and the reaction mixture was stirred
at room temperature for 16 h. After removal of the solvents in vacuo, 5 mL
H2O were added to the residue, which was then extracted with DCM (3 x
15 mL). The combined organic phases were dried over MgSO4,
concentrated and subjected to flash column chromatography (10% EtOAc
in hexane) to afford the title compound 6 as a white solid (97 mg, 0.40
mmol, 92%). The product must be stored in the fridge protected from light
and under nitrogen (decomposition is characterised by a blue-coloration
and insolubility in organic solvents). Rf: 0.24 , Mp: 125-127°C. 1H-NMR
(300 MHz, CDCl3) δ/ppm: 4.06 (s, 9H), 3.46 (s, 3H). 13C-NMR (75.5 MHz,
CDCl3) δ/ppm: 165.7, 106.8, 85.8, 74.8, 61.7. HRMS (FT-MALDI): m/z
calcd for C15H13O3 [M-H]+: 214.0859; found: 214.0870.
1,3,5-trifluoro-2,4,6-triiodobenzene 3
Periodic acid (5.13 g, 22.5 mmol, 1.50 eq) was suspended in 35 mL
sulphuric acid at 0°C. Finely ground potassium iodide (11.2 g, 67.6 mmol,
4.50 eq) was added in portions over 5 min during which time iodine
vapours evolved. 1,3,5-Trifluorobenzene (1.56 mL, 15.1 mmol, 1.00 eq)
was then added by syringe at 0°C and the reaction mixture was then
heated to 70°C for 5 h, during which time additional sulphuric acid can be
added for better stirring. After cooling to room temperature, the reaction
mixture was poured into 350 g crushed ice. 200 mL diethyl ether were
added and the layers separated. The organic layer was washed once with
150 mL of a 15% solution of sodium thiosulfate, once with 150 mL water,
dried over MgSO4 and concentrated to dryness. The residue was purified
by sublimation: the temperature was kept at 50°C (p = 0.04 mbar) for 1 h,
during which time a brown layer coated the cold finger, then it was
increased to 110°C to allow sublimation of the product. Compound 3 was
obtained as a white crystalline solid (7.60 g, 14.9 mmol, 97%). Rf (hexane):
0.6, Mp: 156-158 °C. 13C-NMR (75.5 MHz, CDCl3) δ/ppm: 162.4 (dt, J =
243.5, 7.8 Hz), 63.9 (ddd, J = 34.9, 34.1, 3.9 Hz). 19F-NMR (282.5 MHz,
CDCl3) δ/ppm: 68.83. HRMS (FT-MALDI): m/z calcd for C14H13O2 [M-H]+:
213.0910; found: 213.0909.
Anthracene-1,8-ditriflate 7
Compound 7 was synthesised according to the literature procedure[4]. 1H-
NMR (300 MHz, CDCl3) δ/ppm: 8.90 (s, 1H), 8.61 (s, 1H), 8.07 (d, J = 7.8
Hz, 2H); 7.62-7.50 (m, 4H). 19F-NMR (282.5 MHz, CDCl3) δ/ppm: -76.36.
13C-NMR (75.5 MHz, CDCl3) δ/ppm: 145.7, 133.1, 128.9, 127.9, 125.6,
125.4, 118.9 (q, JCF = 320.3 Hz), 118.4, 114.1. HRMS (FT-MALDI): m/z
calcd for C16H8F6O6S2 [M]+: 473.9661; found: 473.9661. Mp: 111°C.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
(benzene-1,3,5-triyltris(ethyne-2,1-diyl))tris(anthracene-8,1-diyl)
tris(trifluoromethanesulfonate) 8
freshly crystallized batch and the crystals were rinsed three times with
methanol, then dried with a stream of nitrogen (the single crystals perfectly
withstand the washing procedure and can be stored dry over weeks
without losing their crystallinity). The single crystals were placed in a
sealed vial previously purged with argon and put in the LED reactor; every
once in a while the vial was shaken and turned in order to homogeneously
irradiate the mass of crystals.
Compound 8 was synthesised according to the literature procedure[4]. 1H-
NMR (300 MHz, C2D2Cl4) δ/ppm: 9.36 (s, 1H), 8.61 (s, 1H), 8.19 (s, 1H),
8.16-8.08 (m, 2H), 7.98 (d, J = 7.1 Hz, 1H), 7.63 (dd, J = 8.6 Hz, 7.0 Hz,
1H), 7.58-7.50 (m, 2H). 13C-NMR (75.5 MHz, C2D2Cl4) δ/ppm: 145.8, 134.5,
132.6, 131.8, 131.61, 131.57, 129.0, 128.6, 127.5, 126.0, 125.0, 124.3,
124.0, 121.3, 118.9, 118.7 (q, JCF = 321.0 Hz), 117.3, 94.0, 88.0. 19F-NMR
(282.5 MHz, C2D2Cl4) δ/ppm: -76.65. HRMS (FT-MALDI): m/z calcd for
C57H27F9O9S3 [M]+: 1122.0668; found: 1122.0673. Mp: decomposes above
270°C.
Supporting Information
NMR data and synthetic procedures to compound 2. Details on
crystallization procedure, polymorphism and experimental setup for SCSC
reactions. Crystallographic data for every co-crystal, with optical
micrographs and additional information on the structures. ATR-FTIR
spectra and UV/Vis emission spectra of the monomer and polymer crystals.
13CP-MAS NMR data. DSC measurements.
Anthraphane-tri(OMe) 2
Precursor 8 (640 mg, 0.57 mmol, 1.00 eq) was suspended in 380 mL dry
toluene (1.50 mM) with compound 6 (137 mg, 0.57 mmol, 1.00 eq) and dry
triethylamine (15.8 mL, 114 mmol, 200 eq). The reaction mixture was
degassed by cooling it to -80°C with an acetone-dry ice bath and then
performing five cycles of vacuum (10 min) and nitrogen backfilling.
Pd(PPh3)4 (197 mg, 0.17 mmol, 0.30 eq) was added with N2 counter-flow
and the suspension was degassed twice again and backfilled with argon
after the last cycle. After warming to room temperature, the reaction
mixture was put in a preheated bath at 80°C and stirred in the dark under
argon for 5 d. After cooling to room temperature, the reaction mixture was
filtered and the filtrate was concentrated to dryness. The brownish residue
was washed with MeOH to obtain a beige solid, which was collected by
filtration and rinsed with more MeOH until the filtrate resulted colourless.
The beige solid was then dissolved in tetrachloroethane and slowly
precipitated with MeOH to obtain the pure product 2 as a pale yellow solid
(203 mg, 0.22 mmol, 39%). The product can be recrystallized from
dioxane to obtain yellow needles. Rf (20%EtOAc in hexane): 0.27 (blue
Acknowledgements
We thank Dr. Kirill Feldman and Prof. Jan Vermant (Laboratory of
Polymer Technology, ETH Zürich) for access to the optical
microscopy equipment; many thanks go to Dr. Thomas Schweizer
(Institute of Polymer Chemistry, ETH Zürich) for providing the
PID-controller and crystallization apparatus and for building the
LED-photoreactor. Many thanks to Michael Solar (Laboratory of
Inorganic Chemistry, Small Molecule Crystallography Center,
ETH Zurich) for measuring the single crystals. The help of Dr.
René Verel (Laboratory of Inorganic Chemistry, ETH Zürich) for
the solid-state NMR measurements is greatly acknowledged.
1
fluorescence with λ = 366 nm), Mp: decomposes above 280°C. H-NMR
(300 MHz, C2D2Cl4) δ/ppm: 9.54 (s, 3H), 8.56 (s, 3H), 8.11 (d, J = 8.6 Hz,
6H), 7.90-7.80 (m, 6H), 7.77 (s, 3H), 7.60-7.50 (m, 6H), 4.14 (s, 9H). 13C-
NMR (75.5 MHz, C2D2Cl4) δ/ppm: 164.22, 134.10, 131.68, 131.44, 131.41,
131.29, 130.16, 130.11, 129.29, 128.97, 127.30, 125.30, 125.08, 123.93,
123.85, 121.31, 120.62, 107.41, 94.64, 92.09, 88.27, 85.42, 62.05. HRMS
(FT-MALDI): m/z calcd for C69H36O3 [M]+: 912.2659; found: 912.2657.
Conflict of Interest
The authors declare no conflict of interest.
Crystallization Procedure
Keywords: single-crystal-to-single-crystal polymerization•
anthracene • [4+4]-cycloaddition • cyclophane • crystal
engineering
For the screening crystallization procedure, 3 mg anthraphane were
suspended in 0.5 ml solvent inside a clean glass vial equipped with a screw
cap lined with Teflon. The suspension was briefly degassed by purging it
with argon and subsequently sealed and heated close to the boiling point
of the solvent in a sand bath. If a clear solution was not obtained, 0.1 mL
solvent were added and the degassing/heating process repeated. Once a
clear solution was obtained, the vial was controlled-cooled to room
temperature over 24 h, by using a PID controller coupled to the heating
plate (Figure S6). The crystallization process was carried out in a vibration-
free environment and exposure to light was avoided as much as possible.
For the preparation of the monomer crystals of etf/ftf packing 5, 15 mg of
2 were typically crystallized from 1.5 mL nitrobenzene.
[1]
[2]
B. M. D. Cohen, G. M. J. Schmidt, J. Chem. Soc. 1964, 1996–2000.
M. D. Cohen, G. M. J. Schmidt, F. I. Sonntag, J. Chem. Soc. 1964,
2000–2013.
[3]
[4]
[5]
[6]
[7]
G. M. J. Schmidt, Solid State Photochemistry, Verlag Chemie,
Weinheim, 1976.
M. Servalli, N. Trapp, M. Wörle, F. G. Klärner, J. Org. Chem. 2016,
81, 2572–2580.
H. Bouas-Laurent, J.-P. Desvergne, A. Castellan, R. Lapouyade,
Chem. Soc. Rev. 2000, 29, 43–55.
H. Bouas-Laurent, J.-P. Desvergne, A. Castellan, R. Lapouyade,
Chem. Soc. Rev. 2001, 30, 248–263.
I. Zouev, D. Cao, T. V. Sreevidya, M. Telzhensky, M. Botoshansky,
M. Kaftory, CrystEngComm 2011, 13, 4376–4381.
J. Ferguson, A. W.-H. Mau, Mol. Phys. 1974, 27, 377–387.
A. D. Schlüter, Adv. Mater. 1991, 3, 282–291.
Y. C. Teo, H. W. H. Lai, Y. Xia, Chem. Eur. J. 2017, 23, 14101–
14112.
Topochemical single-crystal-to-single-crystal (SCSC) reactions
[8]
[9]
[10]
For the SCSC polymerization, an in-house built cylindrical photoreactor
was used (Figure S10) equipped with 16 high power LEDs (465 nm) with
a nominal dissipation power of 16 x 2.8 W = 44.8 W and a luminous flux of
approximately 30 Lumen per LED at I = 400 mA. The single crystals were
either irradiated on the pin of a goniometer head after being mounted or in
bulk. The pin with the crystal was put in the middle of the reactor and to
regulate the temperature of irradiation (0°C or 20°C), a cooler from an X-
ray diffractometer was used, having the nitrogen stream just above the pin.
For bulk irradiation, the nitrobenzene mother liquor was removed from a
[11]
J. Sakamoto, J. Van Heijst, O. Lukin, A. D. Schlüter, Angew. Chem.
Int. Ed. 2009, 48, 1030–1069.
[12]
[13]
M. Servalli, Chimia 2017, 71, 359-368.
M. Servalli, H. C. Öttinger, A. D. Schlüter, Phys. Today 2018, 71,
40.
[14]
M. Servalli, N. Trapp, M. Solar, A. D. Schlüter, Cryst. Growth Des.
2017, 17, 3419-3432.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
[15]
M. Servalli, M. Solar, N. Trapp, M. Wörle, A. D. Schlüter, Cryst.
Growth Des. 2017, 17, 6510−6522.
[16]
M. Hasegawa, Y. Suzuki, J. Polym. Sci. Part B Polym. Lett. 1967, 5,
813–815.
[17]
[18]
G. Wegner, Zeitschrift für Naturforsch. 1969, 24 B, 824–832.
A. Matsumoto, T. Matsumura, S. Aoki, Macromolecules 1996, 29,
423–432.
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
T. Itoh, S. Nomura, T. Uno, Angew. Chem. Int. Ed. 2002, 41, 4306–
4309.
M. Li, A. D. Schlüter, J. Sakamoto, J. Am. Chem. Soc. 2012, 134,
11721–11725.
S. Nomura, T. Itoh, H. Nakasho, T. Uno, M. Kubo, K. Sada, K.
Inoue, M. Miyata, J. Am. Chem. Soc. 2004, 126, 2035–2041.
M. Zhu, L. Yin, Y. Zhou, H. Wu, L. Zhu, Macromolecules 2018, 51,
746–754.
S. L. Huang, T. S. A. Hor, G. X. Jin, Coord. Chem. Rev. 2017, 346,
112–122.
K. Novak, V. Enkelmann, G. Wegner, K. B. Wagener, Angew.
Chem. Int. Ed. 1993, 32, 1614–1616.
L. Dou, Y. Zheng, X. Shen, G. Wu, K. Fields, W.-C. Hsu, H. Zhou,
Y. Yang, F. Wudl, Science 2014, 343, 272–277.
W. L. Xu, M. D. Smith, J. A. Krause, A. B. Greytak, S. Ma, C. M.
Read, L. S. Shimizu, Cryst. Growth Des. 2014, 14, 993–1002.
Y.-Q. Lan, W.-W. He, S.-L. Li, Inorg. Chem. Front. 2018, 5, 279–
300.
J. W. Lauher, F. W. Fowler, N. S. Goroff, Acc. Chem. Res. 2008, 41,
1215–1229.
A. Matsumoto, T. Tanaka, T. Tsubouchi, K. Tashiro, S. Saragai, S.
Nakamoto, J. Am. Chem. Soc. 2002, 124, 8891–8902.
G. Hennrich, Tetrahedron 2004, 60, 9871–9876.
G. Hennrich, I. Asselberghs, K. Clays, A. Persoons, J. Org. Chem.
2004, 69, 5077–5081.
[30]
[31]
[32]
[33]
[34]
[35]
[36]
H. H. Wenk, W. Sander, Eur. J. Org. Chem 2002, 3927–3935.
M. Schilz, H. Plenio, J. Org. Chem. 2012, 77, 2798–2807.
G. C. Fu, Acc. Chem. Res. 2008, 41, 1555–1564.
G. M. J. Schmidt, Pure Appl. Chem. 1971, 27, 647–678.
G. R. Desiraju, J. J. Vittal, A. Ramanan, Crystal Engineering: A
Textbook, World Scientific Publishing Company, 2011.
G. R. Desiraju, J. Am. Chem. Soc. 2013, 135, 9952–9967.
V. Enkelmann, Mol. Cryst. Liq. Cryst. Sci. Technol. 1998, 313, 15–
23.
[37]
[38]
[39]
[40]
J. M. Thomas, CrystEngComm 2011, 13, 4304.
J. O. Williams, J. M. Thomas, Mol. Cryst. Liq. Cryst. 1972, 16, 371–
375.
[41]
[42]
J. E. Mark, P. J. Flory, J. Am. Chem. Soc. 1966, 88, 3702–3707.
W. Heywang, K. Lubitz, W. Wersing, Piezoelectricity, Springer,
Berlin, Heidelberg, 2008.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802513
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 2:
FULL PAPER
Marco Servalli*, Nils Trapp, A. Dieter
Schlüter
Page No. – Page No.
Single-Crystal-to-Single-Crystal
(SCSC) Linear Polymerization of a
Desymmetrized Anthraphane
Text for Table of Contents
This article is protected by copyright. All rights reserved.