Switchable supramolecular polymers from the self-assembly of a small monomer with two orthogonal binding interactions

Article abstract of DOI:10.1021/ja200941a

The low molecular weight heteroditopic monomer 1 forms supramolecular polymers in polar solution as shown, for example, by infrared laser-based dynamic light scattering (DLS), small-angle neutron scattering (SANS), electron microscopy (TEM, cryo-TEM), and

Full text of DOI:10.1021/ja200941a

ARTICLE
pubs.acs.org/JACS
Switchable Supramolecular Polymers from the Self-Assembly of a
Small Monomer with Two Orthogonal Binding Interactions
Gerd Gr€oger,† Wolfgang Meyer-Zaika,‡ Christoph B€ottcher,§ Franziska Gr€ohn,*,|| Christian Ruthard,^ and
Carsten Schmuck*,†
†Institut f€ur Organische Chemie, and ‡Institut f€ur Anorganische Chemie, Universit€at Duisburg-Essen, Universit€atsstrasse 7,
45141 Essen, Germany
§Forschungszentrum f€ur Elektronenmikroskopie, Institut f€ur Chemie und Biochemie, Freie Universit€at Berlin, 14195 Berlin, Germany
Department of Chemistry and Pharmacy, Interdisciplinary Center for Molecular Materials (ICMM), Friedrich-Alexander-Universit€at
Erlangen-N€urnberg, Egerlandstrasse 3, D-91058 Erlangen, Germany
^Max Planck Institute for Polymer Research, Ackermannweg 10, D- 55128 Mainz, Germany
S Supporting Information
b
ABSTRACT: The low molecular weight heteroditopic mono-
mer 1 forms supramolecular polymers in polar solution as
shown, for example, by infrared laser-based dynamic light
scattering (DLS), small-angle neutron scattering (SANS), elec-
tron microscopy (TEM, cryo-TEM), and viscosity measure-
ments. Self-assembly of 1 is based on two orthogonal binding
interactions, the formation of a Fe(II)-terpyridine 1:2 metalꢀ
ligand complex and the dimerization of a self-complementary
guanidiniocarbonyl pyrrole carboxylate zwitterion. Both bind-
ing interactions have a sufficient stability in polar (DMSO) and
even aqueous solutions to ensure formation of linear polymers
of considerable length (up to 100 nm). The supramolecular polymerization follows a ringꢀchain mechanism causing a significant
increase in the viscosity of the solutions at millimolar concentrations and above. The linear polymers then further aggregate in
solution into larger globular aggregates with a densely packed core and a loose shell. Both binding interactions can be furthermore
switched on and off either by adding a competing ligand to remove the metal ion and subsequent readdition of Fe(II) or by reversible
protonation and deprotonation of the zwitterion upon addition of acid or base. The self-assembly of 1 can therefore be switched
back and forth between four different states, the monomer, a metal-complexed dimer or an ion paired dimer, and finally the polymer.
’ INTRODUCTION
(e.g., metal ions, protons).5 Supramolecular polymers not only
possess polymer-like properties (e.g., higher viscosity and
elasticity), but these properties can also adapt to changes in their
surroundings.6 A fascinating aspect is the possibility of supramo-
lecular polymers to self-heal structural defects because of the
reversible nature of the noncovalent interaction between the
individual monomers.7 One challenge in the field is to find
suitable small molecular weight monomers, which self-assemble
strongly enough in solution to obtain a sufficient degree of
polymerization. The degree of polymerization depends on the
stability of the noncovalent interaction responsible for polymer-
ization. The first examples of supramolecular polymers as
introduced by Lehn8 or Meijer9 were based on hydrogen-
bonding interactions. Especially the self-complementary ureido-
pyrimidone quadruple AADD H-bonding motif developed
by Meijer has attracted a lot of attention in this respect.10,11
However, themajordisadvantageofpurelyH-bondedsupramolecular
This work describes the reversible formation of linear main-
chain supramolecular polymers based on the self-assembly of a
small ditopic monomer 1 with two orthogonal binding interac-
tions, metalꢀligand binding and ion pair formation. Polymer
formation in this case can be reversibly switched on and off by
either the addition of acid or base or the addition or removal of
metal ions. Supramolecular polymers have become an increasing
field of research in the last 20 years as an interesting new class of
materials.1 Besides the functionalization of covalent polymers
with supramolecular binding sites, which can be used to further
aggregate the macromolecule into even larger structures,2 so-
called main-chain supramolecular polymers are especially
interesting.3 In contrast to traditional covalent polymers, in such
systems the individual monomers are held together by nonco-
valent interactions,4 and hence the supramolecular interaction is
part of the polymer backbone. Therefore, formation of the
polymer is reversible (at least under certain conditions) and
can be controlled, for example, by changes in concentration,
solvent, or temperature or by the addition of external stimuli
Received: February 9, 2011
Published: May 04, 2011
r
2011 American Chemical Society
8961
dx.doi.org/10.1021/ja200941a J. Am. Chem. Soc. 2011, 133, 8961–8971
|

Journal of the American Chemical Society
ARTICLE
Scheme 1. Synthesis of the Heteroditopic Monomer 1
polymers is the weakness of this interaction under competitive
polar conditions. Hydrogen bonds are only stable in nonpolar
organic solvents such as chloroform but dissociate in more polar
and especially protic solvents.12 Hence, self-assembled systems,
which are solely based on this interaction only, are not stable, for
example, in aqueous solutions.13 H-bonds as well as other
electrostatic interactions can, however, become more stable
withintheshielding surrounding of a lesspolar microenvironment
(e.g., hydrophobic pockets within an organic polymer or den-
drimer shell).14,15 More often, additional binding interactions
such as Coulomb,16 hydrophobic, or aromatic stacking inter-
actions17 are introduced to obtain more stable self-assembled
systems in water. Some recent examples are based on amphiphiles
with aromatic cores (such as perylenbisimides,18,19 π-conjugated
phenylenes,20 cyclotriveratrylenes,21 or hexacorobenzene,22 just
to name a few), functionalized dipolar dyes,23 cyclodextrins and
calixarenes,24 or macroionꢀcounterion interaction.25,16b A spe-
cial case is metalꢀligand interaction, which also has been used
extensively to construct supramolecular polymers.26 As the
strength of metalꢀligand binding can approach that of covalent
bonds, metallo-supramolecular polymers can be nearly as stableas
traditional covalent ones and therefore can also reach very large
degrees of polymerization.
Most often, supramolecular polymerization is based on one
main type of binding site or a set of binding interactions (e.g.,
H-bonding and π-stacking), which together are responsible for
aggregation. For example, a combination of ionic interaction and
π-stacking can lead to nonlinear supramolecular polymers with
various shapes.27 However, there are also examples in which two or
more binding sites independently, in an orthogonal way, lead to
stepwise polymerization.28 The first example for such a supramo-
lecular polymer with two orthogonal binding sites was introduced
by Schubert et al. who functionalized a polyethylene glycol polymer
with a self-complementary H-bonding AADD binding site on the
one end and a terpyridine ligand on the other end, respectively.29 In
this case, however, the monomer was already a polymer, and due to
the use of H-bonds for one binding site supramolecular polymer-
ization was limited to chloroform. Yashima combined metalꢀ
ligand binding with much stronger charge interactions instead of
H-bonds and presented linear metallo-supramolecular polymers,
which are either cationic or anionic due to amidinium or carbox-
ylate groups incorporated into the strands.30 Hence, two comple-
mentary strands form a helical supramolecular heteroduplex held
together by ion pairing. Despite these fascinating examples of
supramolecular polymers, it is still a challenge to develop new
molecules forming stable supramolecular polymers of sufficient
length in polar solutions and even more so to be able to deliberately
control their formation reversibly by external stimuli.31 In this
respect, we recently introduced a new type of a linear supramole-
cular polymer in which two orthogonal bindingsites are responsible
for polymerization within a single strand.32 We designed the small
heteroditopic molecule 1 in which a terpyridine ligand is attached
to a self-complementary guanidiniocarbonyl pyrrole carboxylate
zwitterion. This zwitterion was developed earlier by us and forms
very stable dimers (Kass > 108 Mꢀ1 in DMSO) held together by
H-bond assisted ion pairs.33 Addition of a metal ion such as Fe(II)
then further polymerized these ion paired dimers into a linear
supramolecular polymer. In a preliminary report,32 we could
demonstrate the formation of such linear polymer strands, for
example, on solid support by atomic force microscopy (AFM) and
in solution using DOSY NMR spectroscopy. We have now studied
the self-assembly of 1 in more detail using infrared laser-based
dynamic light scattering as well as small-angle neutron scattering
(SANS) and electron microscopy (TEM, cryoTEM). We provide
evidence for the mechanism of polymer formation and show how
polymer formation not only changes the macroscopic properties of
the solution (viscosity) but can be reversibly switched on and off by
the addition of acid or base and the addition or removal of the
metal ions.
’ RESULTS AND DISCUSSION
Synthesis of Monomer 1. The ditopic monomer 1 was
synthesized as shown in Scheme 1. The pyrrole tricarboxylic
acid 2 was coupled to N-Boc guanidine using HCTU as coupling
reagent. The ester group in the propyl side chain of 3 was then
selectively hydrolyzed with trimethyl tinhydroxide to obtain acid
4, which was reacted with the amine-functionalized terpyridine 5
to give the protected precursor 6. Ligand 5 is easily accessible
from reduction of p-nitrophenyl-terpyridine 7, which can be
8962
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971

Journal of the American Chemical Society
ARTICLE
Scheme 2. Self-Assembly of a Heteroditopic Monomer 1 Based on Two Orthogonal Binding Interactions, MetalꢀLigand Binding
and Ion Pair Formationa
a The system can be switched from monomer A to either a metal-complexed (B) or ion paired dimer (C), on to larger self-assembled aggregates D (such
as cyclic oligomers or linear polymers depending on the concentration of the solution).
synthesized following a literature protocol.34 After cleavage of the
protecting groups in 6 with TFA, the heteroditopic monomer 1
was obtained as a slightly yellow zwitterion after pH adjustment.
Self-Assembly of 1. As shown in Scheme 2, monomer 1 can
reversibly change from monomer A (e.g., in form of the proton-
ated chloride salt 1•HCl), to a 2:1 metal complex B or an ion
8963
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971

Journal of the American Chemical Society
ARTICLE
Figure 2. Concentration dependence of the relative viscosity of solu-
tions of polymer D, monomer A (protonated chloride salt of 1), and the
metal complex C in DMSO.
Figure 1. Change of the viscosity of a solution of ion paired dimer C
(2.5 mM in DMSO) upon the addition of aliquots of Fe(II), confirming
the formation of a more viscous polymer D due to the formation of a 1:2
metalꢀligand complex.
ion paired dimer C do not change further, when the metal ion is
added and D is formed. Hence, the two binding sites are
orthogonal; the binding properties of the one site are not affected
by the other.
paired dimer C and then further on to larger self-assembled
aggregates D depending on the absence and presence of two
chemical stimuli (pH value, Fe(II) ion). Either signal alone only
leads to the formation of dimers; only when both signals are
present simultaneously can larger aggregates be formed. Because
of the pKa values of the zwitterion (ca. 5 for the carboxylic acid
and 7 for the guanidinium cation, respectively), the zwitterion is
only present within a pH range from 5 to 7. This pH range allows
facile protonation and deprotonation of zwitterion 1 under
ambient conditions, thereby eliminating the self-complementar-
ity of this binding site. Neither the protonated cation nor the
deprotonated anion can dimerize in polar solution.35 The metal
ion can be removed by the addition of a competing ligand such as
HEEDTA (vide infra).
Viscosity Studies. As both interactions, metalꢀligand binding
and ion pair formation, are very stable, the formation of
supramolecular polymers with a significant degree of polymer-
ization even in polar solution was expected. This was first probed
using capillary viscosity measurements as the formation of a
supramolecular polymer should change the macroscopic proper-
ties of the solution.9,5 The viscosity measurements were per-
formed at 25 ꢀC in DMSO with a Schott visco pump II AVS 370
and a type I micro Ubbelohde capillary. Aliquots of iron(II)
chloride (80.0 mM stock solution) were added to a solution of
the ion paired dimer C (2.5 mM) in DMSO. After the mixture
was stirred for 20 min to ensure complete complexation, the
relative viscosity was measured (Figure 1). An increase of the
viscosity was observed until 0.5 equiv of Fe(II) was added, after
which further metal addition did not cause any more changes of
the viscosity. This is consistent with the self-assembly of the ion
paired dimers C present initially in solution into the polymer D
due to the formation of a 1:2 ironꢀterpyridine complex (as also
seen in the UV/vis titration experiments).
Self-assembly is a concentration-dependent process. There-
fore, in contrast to a conventional covalent polymer, also the
viscosity of solutions of 1 depends on the concentration as shown
again by dilution capillary viscosity experiments in DMSO (T =
30 ꢀC). For example, the relative viscosity for D significantly
increases to a value of ηrel = 1.22 at c = 8.5 mM (Figure 2), which
is neither the case for monomer A (c = 8.5 mM, ηrel = 1.02), nor
the 2:1 metal complex B (c = 8.5 mM, ηrel = 1.04), nor the
protected precursor 8 (data not shown).
A closer look at the concentration dependence of the viscosity
of polymer D allows for gaining further insight into the mechan-
ism of polymer formation.37 A double-logarithmic plot of the
specific viscosity (ηspecific) versus the concentration shows two
lines with slopes of 0.56 and 1.17, which intersect at a critical
polymerization concentration (CPC) of c ≈ 0.35 mM (Figure 3).
The appearance of a CPC is a strong indication for a ringꢀchain
polymerization mechanism.38 Monomer 1 is flexible enough to
form cyclic structures (see also Figure 5) as required for a
ringꢀchain supramolecular polymerization (Figure 4). Further-
more, the intermolecular association constants for both the
metalꢀligand complexation and the dimerization of the zwitter-
ion are large enough (K g 108 in DMSO) to give rise to the
Self-assembly of 1 is therefore a stepwise process, which can be
accomplished in two ways. At a pH of 5ꢀ7, compound 1 exists as
a zwitterion, which in solution self-assembles into the ion paired
dimer C.33 Because of the very high stability of this dimer (Kdim
>
108 Mꢀ1 in DMSO), no monomer can be detected even down to
micromolar concentrations (as tested by NMR and UV/vis
dilution studies, data not shown). Subsequent addition of 0.5
equiv of iron(II) chloride to a solution of this dimer C leads
immediately to the appearance of a violet-blue color, indicating
formation of the terpyridineꢀiron(II) 2:1 complex and thus of
self-assembled aggregates D (self-assembly route A f C f D).
Formation of D can also be achieved the other way around,
starting from the protonated cation of 1, the monomer A, which
cannot dimerize, then adding the metal to form the metal-
complexed dimer B and adjusting the pH afterward to get D
(self-assembly route A f B f D). The appearance of the violet-
blue color is typical for terpyridineꢀmetal complexes and is due
to the occurrence of a metal to ligand charge transfer band at λ =
580 nm.36 A UV/vis titration experiment in which the appear-
ance of this MLCT band was monitored quantitatively upon
addition of aliquots of Fe(II) confirmed the formation of the 2:1
complex and provided a stability of Kass > 108 Mꢀ2 under these
conditions.32 All binding steps can also be monitored by char-
acteristic proton shifts in the NMR spectra, which also showed
that neither the zwitterionic binding site nor the terpyridine
coordination site affect each other.32 For example, after metal
complexation to form B, the signals of the terpyridine ligand do
not show any further shift change when self-assembly occurs to
form D upon pH adjustment. Also, the characteristic shifts of the
8964
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971

Journal of the American Chemical Society
ARTICLE
Figure 5. Calculated energy minimized structures of a cyclic dimer, a
cyclic tetramer, and the linear polymer along with representative
molecular dimensions as obtained from force field calculations.
Figure 3. Double-logarithmic plot of the specific viscosity (ηspecific
)
versus the concentration of D, suggesting a ringꢀchain polymerization
mechanism. The insets show TEM images recorded below (0.1 mM
(A)) and above (1 mM (B)) the CPC, also confirming the predominant
formation of cyclic and linear structures under these conditions.
Figure 4. Schematic representation of the ringꢀchain polymerization
mechanism of monomer 1. The linker between the two binding sites is
flexible enough that small cyclic oligomers such as dimers or tetramers
can form, which are in equilibrium with linear ones. At low concentra-
tions, the cycles are preferred, whereas with increasing concentrations
the equilibrium shifts toward the linear polymers.
Figure 6. ESI-MS spectrum of D (0.2 mM in MeOH) showing a
predominant signal for a cyclic dimer and a smaller one for a cyclic
tetramer.
occurrence of a critical concentration as pointed out by
Ercolani.37b,38b At lower concentrations (below CPC), first the
formation of cyclic oligomers takes place, and only small changes
in the viscosity of the solution are observed. These cyclic
oligomers are in equilibrium with linear ones but are favored
entropically in dilute solutions. Above the CPC, chain growth at
the expense of the cycles becomes more favorable, and the
polymerization degree of the linear polymers increases signifi-
cantly as seen in the steeper slope of the specific viscosity
(Figure 3).
Electron Microscopy. Transmission electron microscopy
(TEM) confirmed the ringꢀchain polymerization model de-
duced from the viscosity data. A drop of a diluted solution of the
sample was brought on a holey carbon film coated copper grid.
After the excess of solvent was removed with a piece of filter
paper and the grid was dried, the grid was stained with an
ethanolic solution of uranyl acetate (1% by weight) for 2 min.
Excess of staining agent was removed by rinsing with ethanol.
TEM images taken at a concentration of 0.1 mM and hence
below the CPC show small cyclic structures with a diameter of ca.
2ꢀ4 nm (TEM image A in Figure 3). Above the CPC at 1 mM,
no more cyclic structures but adlayers of linear polymer strands
with a thickness of ca. 1 nm and a length of more than 20 nm are
observed (TEM image B in Figure 3). Similar adlayers were also
observed previously on mica surfaces in atomic force microscopy
images.32
To probe the feasibility of the formation of small cycles, we
performed molecular modeling studies using the software
package Macromodel 8.0 (OPLS force field, GBSA water
solvation model). Energy minimized structures (Figure 5)
were obtained from a Monte Carlo conformational search
with 50.000 steps for each structure. In each case, the resulting
minimum was found at least 10 times or more during the
simulation. These calculations confirm that the formation of
small cycles is geometrically possible in accordance with the
TEM images and the postulated ringꢀchain polymerization
mechanism. Furthermore, the calculated molecular dimen-
sions of both the cyclic oligomers as well as the polymer
strands are in excellent agreement with the size of the particles
observed in the TEM images.
The formation of small cycles in solution was also confirmed
by ESI MS (Figure 6). A strong signal for a cyclic dimer at m/z =
616.192 (charge þ2) as well as a less intense signal for a tetramer
at m/z = 833.578 (charge þ3) are observed. However, no signal
for a trimer is observed. This is not surprising as only an even
number of monomers 1 can form cyclic oligomers. Any odd
number of monomers could only form linear oligomers with
unsaturated binding sites at each end. The ESI spectra therefore
8965
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971

Journal of the American Chemical Society
ARTICLE
technique was used. Vitrified samples suitable for cryo-TEM
are limited with respect to their thickness to ca. 100 nm layers to
provide the necessary translucence for the electron beam. There-
fore, three-dimensional objects larger than this size range are
hardly accessible by this method. After negative staining with 1%
uranyl acetate, we could observe a further assembly of the linear
polymers into larger globular aggregates with diameters ranging
from 200 to 600 nm (Figure 7B). Within the core of these
aggregates, the strands are densely packed (these areas are
excluded from the stain and therefore appear as white spheres).
Toward the globules’ outer shell, the strands (still with an
apparent thickness of 1ꢀ2 nm and a length ranging from 35 to
100 nm) are more loosely arranged. In the areas between the
globular aggregates, individual polymer strands can be found as
seen in the cryo-TEM images without negative staining salt.
Taken all together, the cryo-TEM images therefore demonstrate
the formation of linear, semiflexible supramolecular polymers of
considerable length even in water and their tendency to further
aggregate into larger globular structures.
Light Scattering Studies. To confirm the findings from the
electron microscopy studies, we also studied the self-assembly of
1 using dynamic light scattering (DLS) at different concentra-
tions in DMSO (c = 0.5 and 6 mM). As the sample strongly
absorbs in the visible wavelength range (due to the colored
ironꢀterpyridine complex), a DLS analysis with typical visible
lasers normally used was not possible. We therefore used an IR
laser with 831.5 nm wavelength to perform the experiment, as it
was shown previously that this allows for the characterization of
colored samples by DLS.41 Measurements were performed at a
scattering angle of θ = 90ꢀ. Figure 8A displays results for the
0.5 mM solution. Autocorrelation function and decay time
distribution clearly show bimodal behavior. One peak corre-
sponds to a hydrodynamic radius in the size range of rH,app
=
Figure 7. (A) Cryo-TEM image of polymer D (c = 0.5 mM in water/
DMSO 95/5) showing linear polymer strands. (B) These polymers
further assemble into larger densely packed globular aggregates (white
sphere in the center) as proven by the negative staining preparation with
1% uranyl acetate. Globules in the diameter range of 200ꢀ600 nm can
be found.
200ꢀ300 nm, and the other corresponds to a hydrodynamic
radius above 10 μm. It was not possible to remove these very
large aggregates by various attempts to filter the solution. One
possibility might be that the large structures are completely
destroyed by shear forces during filtration and then re-form in
solution afterward, which is perhaps not unexpected for a highly
dynamic and reversible self-assembling system. Therefore, long
measuring times were chosen to include these very large particles
in the measurement (1800 s in the given example). Even though
the self-assembled polymer carries charges, the observed bimodal
distribution in DLS is not caused by polyelectrolyte effects as
evident from the visibility of very strongly scattering large
particles in the laser beam by eye. Possibly, the chloride
counterions associate with the polymers in DMSO so that the
remaining overall charge of the aggregate is not too high. In any
case, no polyelectrolyte effects complicate the size analysis by
DLS, and it is therefore possible to discuss (apparent) particle
sizes. Thereby, it was possible to obtain reproducible results: The
bimodal distribution is always detected, and an apparent hydro-
dynamic radius of rH = 235 ( 5 nm for the smaller species results.
However, as these data result from a measurement at a 90ꢀ
scattering angle only and no angular extrapolation was per-
formed, the real value likely is somewhat larger. It is reasonable
to assume that these particles correspond to similar globular
aggregates as seen in the TEM images in water (Figure 7B) even
though the DLS data were obtained from DMSO solution and
the TEM images from water.
also suggest the presence of cyclic rather than linear oligomers
under dilute conditions.
Cryo-TEM images,39,40 which provide direct structural data
from a vitrified aqueous solutions (c = 0.5 mM in 95:5 water/
DMSO), reveal linear semiflexible polymer strands just as well
(Figure 7A). The contrast of the data is extraordinarily high due to
the presence of iron ions in the aggregates. The estimated thickness
of the linear polymers is ca. 1ꢀ2 nm (in excellent agreement with
the TEM images and the modeling data), and their length on
average in water is between 30 and 100 nm. As one monomer has a
length of ca. 3.8 nm (according to modeling data), a polymer length
of 50 nm translates into an aggregation number of ca. 13 mono-
mers. Furthermore, as seen from a closer inspection of the images,
the polymer strands in solution have a tendency to undergo lateral
aggregation. For example, in the upper right corner of Figure 7A,
two strands can be seen, which aggregate into a double strand. This
further aggregation might be explained by hydrophobic interactions
between the strands or anion-induced aggregation of the overall
positively charged polymers. As the cryo-TEM images were taken
from an aqueous solution, it is not surprising that hydrophobic
aggregation into larger structures can take place.
To probe for even larger aggregates than just the linear
polymers, the complementary use of the negative staining
The very high scattering intensity of the large particles (r >
10 μM) does not allow the simultaneous detection of any smaller
8966
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971

Journal of the American Chemical Society
ARTICLE
Figure 8. (A) Dynamic light scattering of a solution of 1 in DMSO (c =
0.5 mM). Electric field autocorrelation function g1(τ) and distribution of
relaxation times A(τ) (wavelength λ = 831.5 nm, scattering angle θ =
90ꢀ). The duration of the measurement was 1800 s. (B) Same experi-
ment at a concentration of 6 mM. Multiple measurements with duration
of 5 s each were averaged after excluding runs with a higher intensity
caused by very large particles.
species such as individual polymers. Hence, the absence of faster
processes indicating small species in this sample does not allow
any conclusion whether these particles are not present or can
simply not be detected due to the extremely high scattering
intensity of the larger particles (scattering intensity scales with r6).
Therefore, to check for smaller species, a concentrated solution
(6 mM) was measured multiple times for 5 s, and runs that
detected very large particles were disregarded using the software
“dust filter”. This way, 22 runs were collected and averaged for
data analysis. The result is shown in Figure 8B. Different
experiments of this kind were performed, differing in accumula-
tion times (5, 10, and 30 s) and in selection criteria for the runs.
In all cases, consistently and reproducibly two further species
were detected at fast decay times besides the two large species
already seen before. The two small species have a hydrodynamic
radius of rH = 1ꢀ1.2 nm and 25ꢀ35 nm, respectively. For the
1 nm size range, no angular dependence is expected in DLS.
Hence, even though the measurements were only taken at one
scattering angle (as mentioned above), the data represent a
reliable measure for the real size of the particles. The radius of
25ꢀ35 nm for the second species again represents only an
apparent value, and the real radius might be somewhat larger
as no angular extrapolation could be performed.
Figure 9. SANS analysis of 1 (2 mM in d6-DMSO): (a) Scattering curve
I(q) (arbitrary units); (b) corresponding pair distance distribution
function P(r) indicating particles with a maximum dimension of
265 nm; and (c) corresponding radial density profile ΔF(x) showing
a coreꢀshell structure with an inner sphere with higher density and
about 60 nm radius and a total radius of about 130 nm.
The DLS measurements therefore further complement the
findings obtained from electron microscopy. The small species of
ca. 1 nm radius represents the cyclic oligomers (dimers, tetra-
mers, etc.), which were also seen in TEM (Figure 3B) and ESI
(Figure 6), confirming also the calculated size of such cycles
(Figure 5). The species with an apparent hydrodynamic radius of
25ꢀ35 nm are most likely the linear, semiflexible polymer
strands seen in TEM and cryo-TEM images. If one assumes stiff
linear strands with a diameter of ca. 1ꢀ2 nm, one can estimate a
polymer length of ca. 240ꢀ280 nm from an apparent hydro-
dynamic radius of 25 nm, corresponding to a polymerization
degree of ca. 70 monomers. Of course, the polymers are not
completely stiff but semiflexible (as seen in cryo-TEM), so this
8967
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971

Journal of the American Chemical Society
ARTICLE
Figure 10. UVꢀvis complexation studies ([1] = 0.06 mM in DMSO). The occurrence of a blue-violet color indicates formation of the metal complex
and thus the formation of self-assembled cyclic oligomers D from the ion paired dimer C. HEEDTA as a competing ligand removes the Fe(II) ions from
the terpyridineꢀmetal complex as evident from the decolorization of the sample. Re-addition of excess metal ions then re-forms the cyclic oligomers.
concentration.43 At millimolar concentrations, a degree of poly-
merization of ca. 102 requires a stability of the noncovalent
interaction, which holds the monomers together of Kass
g
107 Mꢀ1 37c
.
Indeed, both binding interactions, metalꢀligand
complexation and dimerization of the zwitterion, have an asso-
ciation constant of Kass g 108 in DMSO. The formation of
supramolecular polymers of this length from the self-assembly of
1 in DMSO is therefore reasonable.
Small Angle Neutron Scattering. To get further information
on the internal structure of the larger particles (the aggregated
polymer strands), small angle neutron scattering (SANS) was
performed. Samples for SANS were prepared in d6-DMSO with
a concentration of 2.5 g Lꢀ1 and transferred into optical quality
quartz cells with 2 mm path length. SANS studies were
performed at the FRM II, Munich, Germany, at the KWS 1
instrument of the J€ulich Center for Neutron Science. We used
three configurations withtheneutronwavelengthsλand sampleꢀ
detector distances d of λ = 0.45 nm/d = 1.6 m, λ = 0.45 nm/
d = 7.6 m, and λ = 1.2 nm/d = 7.6 m. A total scattering vector
range of 0.0143 nmꢀ1 < q < 2.8 nmꢀ1 was covered. Data were
corrected for empty quartz cell scattering, electronic background,
and detector uniformity and were converted to an absolute scale
using secondary standards. The data were further corrected by
subtracting the contributions from solvent scattering and inco-
herent background.
Figure 11. Relative viscosity of a solution of polymer D ([1] = 6 mM in
DMSO) upon stepwise addition of equal amounts of acid or base.
Supramolecular polymerization depends on the protonation state of the
zwitterion in 1 and can therefore be switched on and off by the addition
of acid and base.
number is only an estimate. Nevertheless, a degree of polymer-
ization of this order in DMSO is reasonable. Using the simplified
Carothers equation,42 the degree of polymerization DP for a
supramolecular polymer is roughly proportional to (Kass C)1/2
where Kass is the association constant for the intermolecular
binding of the monomers and C is the total monomer
,
3
8968
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971

Journal of the American Chemical Society
ARTICLE
Figure 12. Schematic representation of the self-assembly of monomer 1 in solution.
In the SANS measurement, again large particles with a
diameter of >200 nm were detected, confirming the results from
the DLS and cryo-TEM measurements. However, SANS now
allows one to further probe the internal structure of these
aggregates in contrast to DLS. For this purpose, the scattering
curve was Fourier transformed into the pair distance distribution
function P(r), which basically represents a distance histogram of
the particle and thus is dependent on and characteristic for the
particle shape and internal density profile.44 P(r) can be calcu-
lated without the need to assume any model in advance from the
scattering curve for finite size, randomly oriented particles (or
supramolecular assemblies, etc.) in solution. The resulting P(r) is
shown in Figure 9b. Most strikingly, the pair distance distribution
function P(r) obtained here cannot be described by radially
homogeneous particles, even assuming polydispersity. Further-
more, deconvolution into the radial density profile (Figure 9c)
shows that the particles consist of a core with a higher density
and a less dense shell. This result is consistent with the rG/r =
0.62 extracted from the SANS measurements prior to the
density profile calculation, which is typical of inhomogenous
“microgels”:45 A possible origin is a different density or branching
profile in the center of the particle as compared to its outer shell.
In other words, a denser network type structure is present in the
inner region and more “free chains” in the outer region of the
particle. This corresponds nicely to the TEM images, which also
showed a densely packed core surrounded by a loose shell of
polymer particles (Figure 7B). Interestingly, such density profiles
are also known for polyelectrolytes that associate with multi-
valent counterions into larger nanoparticles.46
Reversibility of Self-Assembly. As mentioned above, self-
assembly of 1 requires both the presence of the metal ion as well
as the correct pH range. Therefore, self-assembly can be rever-
sibly switched on and off by adding acid or base to the solution or
by removing or adding the metal ion. This reversibility was
probed by UVꢀvis spectroscopy in DMSO (Figure 10). Starting
from the blue-violet color of self-assembled cyclic oligomers D
(c = 0.06 mM in DMSO), addition of an excess of HEEDTA (10
equiv, N-(2-Hydroxyethyl)ethylenediaminetriacetic acid) led to
complete decomplexation and formation of the ion paired dimers
C as confirmed by the decolorization of the solution. HEEDTA is
a known strong competing ligand system for transition metals
like iron(II). It has a significantly higher affinity for Fe(II) than
does the terpyridine ligand (log K = 12.3 versus log K = 7,
respectively)47 and therefore quantitatively removes the metal
ion from the self-assembled oligomers D causing deaggregation
into the ion paired dimer C. If excess iron(II) is added again, the
blue-violet color and the MLCT-band reappear, indicating the
reformation of the self-assembled aggregates D from the ion
paired dimer C.
The effect of reversible protonation and deprotonation of the
zwitterion in 1 was probed using capillary viscosity studies. The
relative viscosity of η = 1.16 of the polymer D (DMSO solution,
c = 6.8 mM, 25 ꢀC) decreases significantly to η = 1.04 after the
addition of aqueous hydrochloric acid (100 μL, 1M) (Figure 11).
These values are the same as measured before for the polymer D
and the metal-complex B under these conditions (see also
Figure 2). Protonation of the zwitterionic binding site removes
its self-complementarity and thus leads to depolymerization. The
remaining metal-complexed dimer B, however, has a significantly
smaller viscosity than does polymer D. Upon subsequent addi-
tion of the same amount of sodium hydroxide (100 μL, 1 M), the
viscosity increases again. The carboxylic acid in B is deproto-
nated, and the self-complementary zwitterion is restored, which
immediately dimerizes and thus reforms the polymer D. Because
of the change of solvent composition (from pure DMSO to
aqueous DMSO), the relative viscosity of polymer D at this point
is smaller than the initial value (η = 1.09 instead of η = 1.16,
respectively), reflecting the dependence of the stability of the
zwitterionic dimer and hence the degree of polymerization on
solvent composition. This external switching between polymer D
and metal-complex B is fully reversible as further additions of acid
or base cause again the same changes of the viscosity.
’ CONCLUSION
We have shown here that a small monomer 1 with two
orthogonal binding interactions self-assembles into a main-chain
supramolecular polymer in polar solvents (DMSO, water). Both
binding interactions, the formation of an ironꢀterpyridine
complex and the dimerization of a self-complementary zwitter-
ion, are independent of each other, allowing the system to exist
in four different self-aggregated states: a monomer (A), a
8969
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971

Journal of the American Chemical Society
ARTICLE
metal-complexed dimer (B) or an ion paired dimer (C), and
finally either cyclic or linear self-assembled aggregates (D). Both
binding events can be switched on and off independently from
each other (e.g., by removing the metal with a competing ligand
and the addition of acid or base) so that reversible external
switching between these four states is possible. As both binding
interactions are very strong (Kass g 108 in DMSO), self-assembly
leads to supramolecular polymers D with a significant degree of
polymerization already at millimolar concentrations. Because of
the flexibility of the molecule, the formation of small cyclic
oligomers (e.g., dimers, tetramers, etc.) is also possible, so that
the supramolecular polymerization follows a classical ringꢀchain
polymerization model. The semiflexible polymer strands in
solution then further aggregate into larger spherical nanostruc-
tures of considerable size most likely due to hydrophobic or
aromatic stacking interactions between the strands. This stepwise
self-assembly process is summarized in Figure 12. All available
experimental data (UV/vis spectroscopy, viscosity, TEM, cryo-
TEM, DLS, SANS, and molecular modeling) are consistent and
fully confirm this model. Hence, the heteroditoic monomer 1 is
one of the few systems that forms supramolecular polymers of
considerable size even in polar solvents and that can be further-
more switched externally between aggregated and nonaggre-
gated state in two different ways. Such system might be of interest
for the development of stimuli responsive materials in the future.
(3) (a) Fox, J. D.; Rowan, S. J. Macromolecules 2009, 42, 6823–6835.
(b) Shimizu, L. S. Polym. Int. 2007, 56, 444–452. (c) Kyung Yang, A.;
Ambade, A. V.; Weck, M. Chem. Soc. Rev. 2011, 40, 129–137.
(4) (a) Schneider, H.-J. Angew. Chem., Int. Ed. 2009, 48, 3924–3977.
(b) Williams, D. H.; Westwell, M. S. Chem. Soc. Rev. 1998, 27, 57–64.
(c) Bissantz, C.; Kuhn, B.; Stahl, M. J. Med. Chem. 2010, 53, 5061–5084.
(5) Serpe, M. J.; Craig, S. L. Langmuir 2007, 23, 1626–1634.
(6) (a) de Greef, T. F. A.; Meijer, E. W. Nature 2008, 453, 171–173.
(b) Bosman, A. W.; Brunsveld, L.; Folmer, B. J. B.; Sijbesma, R. P.;
Meijer, E. W. Macromol. Symp. 2003, 201, 143–154.
(7) Cordier, P.; Tournilhac, F.; Souliꢀe-Ziakovic, C.; Leibler, L.
Nature 2008, 451, 977–980.
(8) (a) Lehn, J.-M. Polym. Int. 2002, 51, 825–839. (b) Berl, V.;
Schmutz, M.; Krische, M. J.; Khoury, R. G.; Lehn, J.-M. Chem.-Eur. J.
2002, 8, 1227–1244. (c) Kotera, M.; Lehn, J.-M.; Vigneron, J.-P. J. Chem.
Soc., Chem. Commun. 1994, 197–199. (d) Gulik-Krczywicki, T.;
Fouquey, C.; Lehn, J.-M. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 163–167.
(9) (a) Sijbesma, R. P.; Beijer, F. H.; Brunsveld, L.; Folmer, B. J. B.;
Hirschberg, J. H. K. K.; Lange, R. F.; Lowe, J. K. L.; Meijer, E. W. Science
1997, 278, 1601–1604. (b) Folmer, B. J. B.; Sijbesma, R. P.; Versteegen,
R. M.; van der Rijt, J. A. J.; Meijer, E. W. Adv. Mater. 2000, 12, 874–878.
(10) (a) Beijer, F. H.; Sijbesma, R. P.; Kooijman, H.; Spek, A. L.;
Meijer, E. W. J. Am. Chem. Soc. 1998, 120, 6761–6769. (b) S€ontjens,
S. M. H.; Sijbesma, R. P.; van Genderen, M. H. P.; Meijer, E. W. J. Am.
Chem. Soc. 2000, 122, 7487–7493.
(11) (a) Schmuck, C.; Wienand, W. Angew. Chem., Int. Ed. 2001,
40, 4363–4369. (b) Zimmerman, S. C.; Corbin, P. S. Struct. Bonding
2000, 96, 63–94.
(12) (a) Jeffrey, G. A. An Introduction to Hydrogen Bonding; Oxford
University Press: New York, 1997. (b) Hunter, C. A. Angew. Chem., Int.
Ed. 2004, 43, 5310–5324.
(13) (a) Rehm, T.; Schmuck, C. Chem. Commun. 2008, 801–813.
(b) Oshovsky, G. V.; Reinhoudt, D. N.; Verboom, W. V. Angew. Chem.,
Int. Ed. 2007, 46, 2366–2393. (c) Zayed, J.; Nouvel, N.; Rauwald, U.;
Scherman, O. A. Chem. Soc. Rev. 2010, 39, 2806–2816.
(14) (a) Obert, E.; Bellot, M.; Bouteiller, L.; Andrioletti, F.; Lehen-
Ferrenbach, C.; Bouꢀe, F. J. Am. Chem. Soc. 2007, 129, 15601–15605.
(b) Yoshikawa, I.; Sawayama, J.; Araki, K. Angew. Chem., Int. Ed. 2008,
47, 1038–1041. (c) Merschky, M.; Wyszogrodzka, M.; Haag, R.;
Schmuck, C. Chem.-Eur. J. 2010, 16, 14242–14246. (d) Mes, T.;
Smulders, M. M. J.; Palmans, A. R. A.; Meijer, E. W. Macromolecules
2010, 43, 1981–1991.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedure for
b
the synthesis of 1 as well as analytical and spectral characteriza-
tion data. This material is available free of charge via the Internet
at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
franziska.groehn@chemie.uni-erlangen.de; carsten.schmuck@
uni-due.de
(15) However, ethylene glycol only exhibits a stabilizing effect as
compared to aqueous solvents (as in ref 14c). In less polar solvents, it
destabilizes H-bonded systems due to competitive intramolecular
H-bonding: (a) de Greef, T. F. A.; Nieuwenhuizen, M. M. L.; Stals,
P. J. M.; Fitie, C. F. C.; Palmans, A. R. A.; Sijbesma, R. P.; Meijer, E. W.
Chem. Commun. 2008, 4306–4308. (b) de Greef, T. F. A.; Nieuwenhui-
zen, M. M.; Sijbesma, R. P.; Meijer, E. W. J. Org. Chem. 2010, 75,
598–610.
(16) (a) Rehm, T.; Schmuck, C. Chem. Soc. Rev. 2010, 39,
3597–3611. (b) Gr€ohn, F. Soft Matter 2010, 6, 4296–4302.
(17) Ryu, J.-H.; Hong, D.-J.; Lee, M. Chem. Commun. 2008,
1043–1054.
’ ACKNOWLEDGMENT
Financial support of our work from the DFG and the Fonds of
the chemical industry and the J€ulich Center for Neutron Science
(JCNS) is gratefully acknowledged. We also thank Prof. Christoph
Schalley and Henrik Winkler (FU Berlin, Germany) for measuring
the ESI-MS spectrum, Wilhelm Sicking (University of Duisburg-
Essen) for help with the molecular modeling calculations, and Dr.
Henrich Frielinghaus (JCNS, Germany) for help with the SANS
measurements.
(18) (a) Neelakandan, P. P.; Pan, Z.-Z.; Hariharan, M.; Zheng, Y.;
Weissman, H.; Rybtchinski, B.; Lewis, F. D. J. Am. Chem. Soc.
2010, 132, 15808–15813. (b) Krieg, E.; Shirman, E.; Weissman, H.;
Shimoni, E.; Wolf, S. G.; Pinkas, I.; Rybtchinski, B. J. Am. Chem. Soc.
2009, 131, 14365–14373. (c) Kaiser, T.; Wang, H.; Stepanenko, V.;
W€urthner, F. Angew. Chem., Int. Ed. 2007, 46, 5541–5544. (d) Zhang, X.;
Rehm, S.; Safont-Sempere, M. M.; W€urthner, F. Nat. Chem. 2009,
1, 623–629.
’ REFERENCES
(1) For selected reviews on supramolecular polymers, see:
(a) Brunsveld, B. J. B.; Folmer, E. W.; Meijer, E. W.; Sijbesma, R. P.
Chem. Rev. 2001, 101, 4071–4097. (b) Bouteiller, L. Adv. Polym. Sci. 2007,
207, 79–112. (c) Binder, W. H.; Zirbs, R. Adv. Polym. Sci. 2007, 207, 1–78.
(d) Rieth, S.; Baddeley, C.; Badjic, J. D. Soft Matter 2007, 3, 137–154.
(e) Armstrong, G.; Buggy, M. J. Mater. Sci. 2005, 40, 547–559. (f) Ciferri,
A. Supramolecular Polymers; Taylor & Francis: Boca Raton, Florida, 2005.
(g) Sherrington, D. C.; Taskinen, K. A. Chem. Soc. Rev. 2001, 30, 83–93.
(2) (a) Pollino, J. M.; Weck, M. Chem. Soc. Rev. 2005, 34, 193–207.
(b) Rotello, V.; Thayumanavan, S. Molecular Recognition and Polymers;
Wiley: New York, 2008.
(19) W€urthner, F. Chem. Commun. 2004, 1564–1579.
(20) (a) Garcia, F.; Sanchez, L. Chem.-Eur. J. 2010, 16, 3138–3146.
(b) Yagai, S.; Kubota, S.; Saito, H.; Unoike, K.; Karatsu, T.; Kitamura, A.;
Ajayaghosh, A.; Kanesato, M.; Kikkawa, Y. J. Am. Chem. Soc. 2009, 131,
5408–5410. (c) Garcia, F.; Aparicio, F.; Marenchino, M.; Campos-Olivas,
R.; Sanchez, L. Org. Lett. 2010, 12, 4264–4267.
8970
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971

Journal of the American Chemical Society
ARTICLE
(21) Percec, V.; Imam, M. R.; Peterca, M.; Wilson, D. A.; Heiney,
P. A. J. Am. Chem. Soc. 2009, 131, 1294–1304.
(22) Yin, M.; Shen, J.; Pisula, W.; Liang, M.; Zhi, L.; M€ullen, K. J. Am.
Chem. Soc. 2009, 131, 14618–14619.
(35) Schmuck, C. Eur. J. Org. Chem. 1999, 2397–2403.
(36) Hofmeier, H.; Schubert, U. S. Chem. Commun. 2004, 373–399.
(37) (a) Ciferri, A. Macromol. Rapid Commun. 2002, 23, 511–529.
(b) de Greef, T. F. A.; Smudlers, M. M. J.; Wolffs, M.; Schenning, A. P. H.
J.; Sijbesma, R. P.; Meijer, E. W. Chem. Rev. 2009, 109, 5687–5754.
(c) ten Cate, A. T.; Sijbesma, R. P. Macromol. Rapid Commun. 2002,
23, 1094–1112.
(38) (a) Hagy, M. C.; Chen, C.-C.; Dormidontova, E. E. Macro-
molecules 2007, 40, 3408–3421. (b) Ercolani, G.; Mandolini, L.;
Mencarelli, P.; Roelens, S. J. Am. Chem. Soc. 1993, 115, 3901–3908.
(c) Chen, C.-C.; Dormidontova, E. E. Macromolecules 2004, 37,
3905–3917. (d) Ercolani, G.; Di Stefano, S. J. Phys. Chem. B 2008,
112, 4662–4665.
(39) (a) Dubochet, J.; McDowall, A. W. J. Microsc. (Oxford, U.K.)
1981, 124, Rp3–Rp4. (b)Dubochet, J.; Chang, J. J.;Freeman, R.; Lepault,
J.; McDowall, A. W. Ultramicroscopy 1982, 10, 55–61. (c) Dubochet, J.;
Lepault, J.; Freeman, R.; Berriman, J. A.; Homo, J. C. J. Microsc. (Oxford,
U.K.) 1982, 128, 219–237.
(23) (a) Schmidt, R.; Uemura, S.; W€urthner, F. Chem.-Eur. J. 2010,
16, 13706–13715. (b) Yagai, S.; Kinoshita, T.; Higashi, M.; Kishikawa,
K.; Nakanishi, T.; Karatsu, T.; Kitamura, A. J. Am. Chem. Soc. 2007,
129, 13277–13287. (c) Yagai, S.; Higashi, M.; Karatsu, T.; Kitamura, A.
Chem. Mater. 2005, 17, 4392–4398. (d) Chen, Z.; Lohr, A.; Saha-M€oller,
C. R.; W€urthner, F. Chem. Soc. Rev. 2009, 38, 564–584.
(24) (a) Harada, A.; Takashima, Y.; Yamaguchi, H. Chem. Soc. Rev.
2009, 38, 875–882. (b) Becherer, M. S.; Schade, B.; B€ottcher, C.; Hirsch,
A. Chem.-Eur. J. 2009, 15, 1637–1648. (c) Koopmans, C.; Ritter, H.
J. Am. Chem. Soc. 2007, 129, 3502–3503. (d) Kretschmann, O.; Steffens,
C.; Ritter, H. Angew. Chem., Int. Ed. 2007, 46, 2708–2711.
(25) (a) Reinhold, F.; Kolb, U.; Lieberwirth, I.; Gr€ohn, F. Langmuir
2009, 25, 1345–1351. (b) Li, Y.; Yildiz, H. U.; M€ullen, K.; Gr€ohn, F.
Biomacromolecules 2009, 10, 530–540.
(40) For a recent review, see: Friedrich, H.; Frederik, P. M.; de With,
G.; Sommerdijk, N. A. J. M. Angew. Chem., Int. Ed. 2010, 49, 7850–7858.
(41) Ruthard, C.; Maskos, M.; Kolb, U.; Gr€ohn, F. Macromolecules
2009, 42, 830–840.
(26) Forsomerecentreviewsonmetallo-supramolecularpolymers, see:
(a) Dobrawa, R.; W€urthner, F. J. Polym. Sci., Part A: Polym. Chem. 2005,
43, 4981–4995. (b) Hofmeier, H.; Schubert, U. S. Chem. Soc. Rev. 2004,
33, 373–399. (c) Chiper, M.; Hoogenboom, R.; Schubert, U. S. Macromol.
Rapid Commun. 2009, 30, 565–578. (c) Shunmugam, R.; Gabriel, G. J.;
Khaled, A. A.; Tew, G. N. Macromol. Rapid Commun. 2010, 31, 784–793.
(d) Schubert, U. S.; Newkome, G. R.; Maners, I. . ACS Symposium Series
928, 2006. (e) Friese, V. A.; Kurth, D. G. Curr. Opin. Colloid Interface Sci.
2009, 14, 81–93.
(27) (a) Gr€ohn, F.; Klein, K.; Brand, S. Chem.-Eur. J. 2008, 14,
6866–6869. (b) Willerich, I.; Gr€ohn, F. Chem.-Eur. J. 2008, 14,
9112–9116. (c) Gr€ohn, F. Macromol. Chem. Phys. 2008, 209,
2295–2301. (d) Willerich, I.; Ritter, H.; Gr€ohn, F. J. Phys. Chem. B
2009, 113, 3339–3354. (e) Gr€ohn, F.; Klein, K.; Koynov, K. Macromol.
Rapid Commun. 2010, 31, 75–80.
(42) Martin, R. B. Chem. Rev. 1996, 96, 3043–3064.
(43) The Carothers equation is based on an isodesmic growth mode
(also called equal K model). However, for a high concentration of
monomers well above the CPC, the degrees of polymerization for an
isodesmic and a ringꢀchain polymerization are equal. Hence, as a first
approximation, the use of the Carothers equation is surely justified here.
(44) (a) Glatter, O. ActaPhys. Austriaca1977, 47, 83–102. (b) Glatter,
O. J. Appl. Crystallogr. 1977, 10, 415–421. (c) Glatter, O. J. Appl.
Crystallogr. 1980, 13, 577–584. (d) Gr€ohn, F.; Bauer, B. J.; Amis, E. J.
Macromolecules 2001, 34, 6701–6707.
(45) Kunz, D.; Thurn-Albrecht, T.; Burchard, W. Colloid Polym. Sci.
1983, 261, 635–646.
(46) Yildiz, U. H.; Koynov, K.; Gr€ohn, F. Macromol. Chem. Phys.
2009, 210, 1678–1690.
(28) (a) Hofmeier, H.; Schubert, U. S. Chem. Commun. 2005,
2423–2432. (b) Schmittel, M.; Mahata, K. Angew. Chem., Int. Ed.
2008, 47, 5284–5286.
(47) Martell, A. E.; Smith, R. M. Critical Stability Constants; Plenum
Press: New York, 1974.
(29) (a) Hofmeier, H.; El-ghayoury, A.; Schenning, A. P. H. J.;
Schubert, U. S. Chem. Commun. 2004, 318–319. (b) Hofmeier, H.;
Hoogenboom, R.; Wouters, M. E. L.; Schubert, U. S. J. Am. Chem. Soc.
2005, 127, 2913–2921.
(30) (a) Furusho, Y.; Yashima, E. Macromol. Rapid Commun. 2011,
32, 136–146. (b) Yashima, E.; Maeda, K.; Iida, H.; Furusho, Y.; Nagai, K.
Chem. Rev. 2009, 109, 6102–6211. (c) Yashima, E.; Maeda, K.; Furusho,
Y. Acc. Chem. Res. 2008, 41, 1166–1180.
(31) Some recent examples and reviews on switchable nanoassem-
blies: (a) Voskuhl, J.; Fenske, T.; Stuart, M. C. A.; Wibbeling, B.;
Schmuck, C.; Ravoo, B. J. Chem.-Eur. J. 2010, 16, 8300–8306.
(b) Rodler, F.; Linders, J.; Fenske, T.; Rehm, T.; Mayer, C.; Schmuck,
C. Angew. Chem., Int. Ed. 2010, 49, 8747–8750. (c) Du, J.; Tang, Y.;
Lewis, A. L.; Armes, S. P. J. Am. Chem. Soc. 2005, 127, 17982–17983.
(d) Versluis, F.; Tomatsu, I.; Kehr, S.; Fregonese, C.; Tepper, A. W. J.
W.; Stuart, M. C. A.; Ravoo, B. J.; Koning, R. I.; Kros, A. J. Am. Chem. Soc.
2009, 131, 13186–13187. (e) Ghosh, S.; Reches, M.; Gazit, E.; Verma, S.
Angew. Chem., Int. Ed. 2007, 46, 2002. (f) Menand, M.; Jabin, I. Chem.-Eur.
J. 2010, 16, 2159–2169. (g) Leung, K. C.-F.; Chak, C.-P.; Lo, C.-M.;
Wong, W.-Y.; Xuan, S.; Cheng, C. H. K. Chem. Asian J. 2009, 4, 364–381.
(h) Rowan, S. J. Polym. Prepr. 2009, 50, 339–340. (i) Beck, J. B.; Rowan,
S. J. J. Am. Chem. Soc. 2003, 125, 13922–13923. (j) Wang, C.; Guo, Y.;
Wang, Z.; Zhang, X. Langmuir 2010, 26, 14509–14511. (k) Wang, F.;
Zhang, J.; Ding, X.; Dong, S.; Liu, M.; Zheng, B.; Li, S.; Wu, L.; Yu,
Y.; Gibson, H. W. Angew. Chem., Int. Ed. 2010, 49, 1090–1094. (l) Yagai,
S.; Kitamura, A. Chem. Soc. Rev. 2008, 37, 1520–1529. (m) Willerich, I.;
Gr€ohn, F. Angew. Chem., Int. Ed. 2010, 49, 8104–8108.
(32) Gr€oger, G.; Stepanenko, V.; W€urthner, F.; Schmuck, C. Chem.
Commun. 2009, 698–700.
(33) Schmuck, C.; Wienand, W. J. Am. Chem. Soc. 2003, 125, 452–459.
(34) Mutai, T.; Cheon, J. D.; Arita, S.; Araki, K. J. Chem. Soc., Perkin
Trans. 2 2001, 7, 1045–1050.
8971
dx.doi.org/10.1021/ja200941a |J. Am. Chem. Soc. 2011, 133, 8961–8971