Mechanochemical Encapsulation of Fullerenes in Peptidic Containers Prepared by Dynamic Chiral Self-Sorting and Self-Assembly

Article abstract of DOI:10.1002/chem.201504451

Molecular capsules composed of amino acid or peptide derivatives connected to resorcin[4]arene scaffolds through acylhydrazone linkers have been synthesized using dynamic covalent chemistry (DCC) and hydrogen-bond-based self-assembly. The dynamic character of the linkers and the preference of the peptides towards self-assembly into β-barrel-type motifs lead to the spontaneous amplification of formation of homochiral capsules from mixtures of different substrates. The capsules have cavities of around 800 ?³ and exhibit good kinetic stability. Although they retain their dynamic character, which allows processes such as chiral self-sorting and chiral self-assembly to operate with high fidelity, guest complexation is hindered in solution. However, the quantitative complexation of even very large guests, such as fullerene C₆₀ or C₇₀, is possible through the utilization of reversible covalent bonds or the application of mechanochemical methods. The NMR spectra show the influence of the chiral environment on the symmetry of the fullerene molecules, which results in the differentiation of diastereotopic carbon atoms for C₇₀, and the X-ray structures provide unique information on the modes of peptide-fullerene interactions.

Full text of DOI:10.1002/chem.201504451

DOI: 10.1002/chem.201504451
Full Paper
&
Supramolecular Chemistry
Mechanochemical Encapsulation of Fullerenes in Peptidic
Containers Prepared by Dynamic Chiral Self-Sorting and Self-
Assembly
+ [a]
+ [a]
[b, c]
Hanna Je˛drzejewska,[a]
´
Marek Szymanski , Michał Wierzbicki , Mirosław Gilski,
Marcin Sztylko,[a] Piotr Cmoch,[a] Aleksander Shkurenko,[d] Mariusz Jaskólski,*[b, c] and
Agnieszka Szumna*[a]
Abstract: Molecular capsules composed of amino acid or
peptide derivatives connected to resorcin[4]arene scaffolds
through acylhydrazone linkers have been synthesized using
dynamic covalent chemistry (DCC) and hydrogen-bond-
based self-assembly. The dynamic character of the linkers
and the preference of the peptides towards self-assembly
into b-barrel-type motifs lead to the spontaneous amplifica-
tion of formation of homochiral capsules from mixtures of
different substrates. The capsules have cavities of around
as chiral self-sorting and chiral self-assembly to operate with
high fidelity, guest complexation is hindered in solution.
However, the quantitative complexation of even very large
guests, such as fullerene C60 or C70, is possible through the
utilization of reversible covalent bonds or the application of
mechanochemical methods. The NMR spectra show the in-
fluence of the chiral environment on the symmetry of the
fullerene molecules, which results in the differentiation of
diastereotopic carbon atoms for C70, and the X-ray structures
provide unique information on the modes of peptide–fuller-
ene interactions.
3
800 Š and exhibit good kinetic stability. Although they
retain their dynamic character, which allows processes such
Introduction
groups are required to obtain predictable and functional 3D
structures. Consequently, the practical applications of peptidic
building blocks are often limited by their synthesis.[5] In this
regard, the utilization of relatively easily available, modified,
short peptides has proven to be more practical.[6] A promising
but largely unexplored approach involves the combination of
short peptides with dynamic covalent chemistry (DCC). This ap-
proach takes advantage of the natural tendency of peptides to
self-assemble as well as of the reversible character of chemical
reactions to amplify the formation of complex functional struc-
tures.[7] Dynamic structures based on peptides can also poten-
tially benefit from the unique features of the DCC approach,
such as the possibility of self-sorting, ligand-templating/com-
plexation, and self-healing. However, they also present signifi-
cant challenges, such as the need to strike a balance between
kinetic and thermodynamic parameters, or the need to control
the flow of energy to the system to cross various energy barri-
ers.
The chemical diversity and biocompatibility of peptides have
stimulated intensive research on their applications as supra-
molecular building blocks in the field of artificial nanomaterials.
Numerous examples of the application of peptide-based syn-
thetic materials have been reported, including targeted drug
delivery systems,[1] materials for regenerative medicine,[2] retro-
viral gene transfer, and the even capturing carbon dioxide
from fuel gas.[3] However, relatively long peptidic or peptidomi-
metic sequences[4] or the presence of additional interacting
+
[a] M. Szymanski, M. Wierzbicki,+ H. Je˛drzejewska, M. Sztylko, Dr. P. Cmoch,
´
Prof. A. Szumna
Institute of Organic Chemistry, Polish Academy of Sciences
Kasprzaka 44/52, 01-224 Warsaw (Poland)
E-mail: agnieszka.szumna@icho.edu.pl
[b] Dr. M. Gilski, Prof. M. Jaskólski
Faculty of Chemistry, A. Mickiewicz University
Umultowska 89b, 61-614 Poznan (Poland)
E-mail: mariuszj@amu.edu.pl
In this paper we present the formation of new, dynamic, and
yet highly stable peptidic capsules and show various ways to
cross the energy barriers of covalent and noncovalent process-
es (reversible chemical reactions, self-assembly, and complexa-
tion). Our de novo designed capsules possess acylhydrazone
linkers and peptides connected through their carbon termini,
which results in enhanced kinetic and thermodynamic stability.
On the other hand, high stability implies the existence of high
energy barriers that hamper the complexation of guests. We
show that the barriers can be surmounted by utilizing the re-
[c] Dr. M. Gilski, Prof. M. Jaskólski
Institute of Bioorganic Chemistry, Polish Academy of Sciences
Noskowskiego 12/14, 61-704 Poznan (Poland)
[d] Dr. A. Shkurenko
Institute of Physical Chemistry, Polish Academy of Sciences
Kasprzaka 44/52, 01-224 Warsaw (Poland)
[+] These individuals contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201504451.
Chem. Eur. J. 2016, 22, 3148 – 3155
3148
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
versible character of covalent bonds or by mechanochemical
methods. Although the mechanisms of chemical processes
under extreme mechanical forces are still poorly understood,
the high potential of mechanochemistry in controlling supra-
molecular and dynamic systems has recently been demonstrat-
ed.[8] We conclusively show the effective complexation of non-
intuitive guest molecules, fullerenes C60 and C70, inside our
peptidic capsules and analyze the fullerene–peptide interac-
tions.
Results and Discussion
Design and synthesis
We have previously demonstrated that short peptides can
serve as building blocks for the dynamic formation of capsules,
provided that they are properly pre-organized by, for example,
connection to a macrocyclic skeleton.[9] The peptides are con-
nected to the scaffolds by imine linkers through their nitrogen
termini. However, an intrinsically low energy barrier for the
imine formation/hydrolysis reaction precluded further applica-
tions of such capsules for complexation (under ambient condi-
tions, the imine linkers were prone to disruption even by
simple neutral molecules). A new generation of dynamic pepti-
dic capsules has been designed by using the resorcin[4]arene
scaffold 1[10] and peptidic fragments 2a–3b connected
through acylhydrazones as dynamic linkages (Figure 1). The
acylhydrazone linkers display an excellent balance between
facile reversibility and stability, as suggested[11] and exploited
by Sanders and co-workers.[12] Additionally, acylhydrazones are
compatible with water and biological environments,[13] and
their formation can be catalyzed under mild conditions in
media of various pH.[14] It should be noted that simple aromat-
ic acylhydrazones attached to the resorcinarene scaffold have
been previously used for the formation of molecular capsules,
however, not in a dynamic way.[15]
Figure 1. a) Synthesis of acylhydrazones 4a–5b. b) Full chemical structure of
a capsule hemisphere, with the atomic labelling scheme. The front arm is
shown in black, the remaining ones, shown in perspective, are in gray.
c) 1H NMR spectrum of (l-5a)2 (CDCl3, 300 K, 600 MHz).
The peptide hydrazides 2a–3b were synthesized by using
the EDCI/OXYMA (EDCI= 1-ethyl-3-(3-dimethylaminopropyl)-
carbodiimide, OXYMA= ethyl (hydroxyimino)cyanoacetate)
peptide coupling strategy (see the Supporting Information).
The reaction of 1 with peptide hydrazides 2a–3b gave tetra-
substituted acylhydrazones 4a–5b in high yields (85–97%, Fig-
ure 1a). Control experiments were performed to verify the re-
versibility of the reaction under the conditions employed.
When l-4a was mixed with l-4b and heated at 708C for
7 days, scrambling of the acylhydrazone fragments was ob-
served, which indicates the reversibility of the reaction under
such conditions (see Figures S24–S28 in the Supporting Infor-
mation).
previously reported for analogous functional groups.[16] In the
1
present case, the acylhydrazones 4a–5b exhibit simple H and
13C NMR spectra in nonpolar solvents, consistent with the C4
symmetry of the molecules, assuming single tautomeric forms
1
1
in all arms (Figure 1c). Analysis of the H–15N and H–13C HSQC
and HMBC spectra of 4a–5b allowed us to determine the posi-
tions of the hydrogen atoms and thus to assign the specific
tautomeric forms (see the Supporting Information). For all the
products 4a–5b, the resorcinol rings exist in the enol form
and the acylhydrazone groups assume the amide form (Fig-
ure 1b). The X-ray crystal structures of 4b and 5a also con-
firmed that all the CarÀOH bonds are of the same length,
which indicates the enol form of the resorcinol rings. Note that
the tautomeric form of the resorcin[4]arene skeleton in the
case of acylhydrazones 4a–5b is different from the keto–en-
amine form reported previously for peptidic capsules based on
imine linkages.[10]
Isomerism
Acylhydrazones 4a–5b have many possible isomers, including
regioisomers, tautomers, and atropisomers. The isomerism
originates from keto–enol and amide–iminol tautomerism and
from E/Z isomerization of the double and partial double bonds
at various positions. All the isomerization processes have been
Chem. Eur. J. 2016, 22, 3148 – 3155
www.chemeurj.org
3149
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Self-assembly
Ramachandran angles of natural b-sheet structures (f=
À118.48, y=107.28). The interstrand hydrogen-bonded cycles
are formed as 10-membered rings (orange motif). Notably, the
binding motif is not fully complementary; the strands are
grouped pairwise, with unmatched amide groups (Figure 2a),
which act to bridge the paired-strands through the recruit-
ment of hydrogen-bonded ethanol molecules.
The NMR spectra and X-ray structures confirm that all the syn-
thesized peptidic acylhydrazones 4a–5b form self-assembled
capsular dimers in solution and in the solid state. The ROESY
spectra of acylhydrazones 4a and 4b (amino acid derivatives)
each show the ROE effect between the protons f and h (see
Figures S5 and S10 in the Supporting Information). For acylhy-
drazones 5a and 5b (dipeptides), a ROE was observed be-
tween the protons f and h,h’’ (see Figures S15 and S20). The
proximity of these protons is consistent with the formation of
self-assembled structures. Additionally, the diffusion coeffi-
cients observed for the present acylhydrazones are similar to
the values reported for other dimeric capsules (see Table S1).
All these findings strongly suggest that in solution the acylhy-
drazones exist as dimers.
The X-ray structure of (d,l-5a)2 (Figure 2b) also confirms the
formation of a dimeric self-assembled capsule. The binding
motif also involves the formation of four antiparallel b-sheet
“pairs of strands” (Figure 2d). However, an important difference
between the binding motifs of capsules (d-4b)2 and (d,l-5a)2
is that in the latter case the interstrand pairing leads to the for-
mation of 14-membered rings (Figure 2d,e, blue motif). The
second amino acid residue in each strand (Gly) plays a less de-
fined role as it interconnects the “pairs of strands” or mediates
interactions with other capsules and solvent molecules in the
crystal. It is important to note that although the sample was
crystallized from a racemate and the space group is centro-
symmetric, each individual capsule contains only homochiral
hemispheres. This strongly suggests a preference towards the
formation of homochiral dimers by the dipeptidic acylhydra-
zones.
The X-ray crystal structure of (d-4b)2 confirms the formation
of self-assembled dimers with two hemispheres interacting
through hydrogen bonds between their peptide backbones
(Figure 2a). The hydrogen-bonding motif involves the forma-
tion of four pairs of antiparallel strands, resembling the anti-
parallel b-sheet binding motif in a b-barrel (Figure 2c,e). The
conformation of the peptidic chains also corresponds to typical
Figure 2. X-ray crystal structures of a,c) (d-4b)2 and b,d) (d,l-5a)2 showing the characteristic binding motifs. e) Schematic representation of the capsules and
the b-sheet binding motifs, illustrating the differences in the relative arrangement of the peptide chains. NMR monitoring of the exchange of hemispheres:
1H NMR spectra of f) (l-4b)2, g) (l-4a)2, and h) (l-4a)2 +(l-4b)2 after 3 days and i) (l-4a)2 +(d-4b)2 after 3 days (CDCl3/MeOH, 95:5, 300 K, 600 MHz). Chiral self-
sorting during synthesis: 1H NMR spectra of j) (l-4a)2, k) (l-4b)2, l) (l-5a)2, m) the product of the reaction of 1+l-2a+l-2b, n) the product of the reaction of
1+l-2a+d-2b, o) the product of the reaction of 1+l-2b+l-3a, and p) the product of the reaction of 1+l-2b+d-3a (reaction conditions: 7 days, 708C,
CHCl3; NMR spectra recorded in CDCl3, 300 K, 600 MHz).
Chem. Eur. J. 2016, 22, 3148 – 3155
www.chemeurj.org
3150
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Stability
A plausible explanation for the amplification of homochiral
capsules over heterochiral or mixed capsules may involve the
concurrent involvement of various factors, for example, 1) the
preference of acylhydrazone linkages towards the planar E con-
formation about the NÀN bond (99.6% of the structures based
on a CSD search and structures presented in this paper, see
Figure S49 in the Supporting Information) and 2) the confor-
mational preferences of peptides within a b-sheet binding
motif according to the Ramachandran plot (especially for
Val).[18]
To probe the stability of the dimers and the preference for
chiral self-assembly, we analyzed two mixtures of capsules:
(l-4a)2 +(l-4b)2 (homochiral mixture) and (l-4a)2 +(d-4b)2
(heterochiral mixture) in various media. In CDCl3 at room tem-
perature the 1H NMR spectra remained unchanged in both
cases, even after 7 days. Heating of the mixtures resulted in co-
valent scrambling, without a stage in which hybrid capsules
made of two different hemispheres could be detected. To pro-
mote noncovalent exchange of the hemispheres, a small
amount of a polar solvent was added (5% v/v MeOH). After
3 days, the equilibration process had led to the formation of
a statistical mixture of hybrid capsules in the case of the ho-
mochiral mixture (Figure 2h), whereas no change was observed
in the composition of the heterochiral mixture (Figure 2i).
These results provide further evidence of the existence of di-
meric forms in solution, even after the addition of polar sol-
vents. They also indicate that homochiral dimers are highly
preferred. Additionally, it should be noted that the self-assem-
bled dimers have an unusually high kinetic stability in nonpo-
lar solvent (CDCl3) and relatively slow rates of hemisphere ex-
change (days) even after the addition of polar solvent.
Interestingly, no self-sorting was observed based on the
length of the peptides. The reaction of 1 with a mixture of l-
2b+l-3a (hydrazides of different length but the same chirali-
ty) led to a mixture of many products (Figure 2o). However,
the reaction of 1 with a mixture of l-2b+d-3a (hydrazides of
different length and different chirality) led to a well-defined
mixture of only two products (l-4b)2 +(d-5a)2 (Figure 2p).
Thus, in the present case, only the chirality but not the length
of the peptides plays a critical role in self-sorting. This is an in-
triguing observation because the complementarity of the bind-
ing motifs is the key factor in the self-sorting processes and re-
sults in the amplification of products composed of substrates
of the same lengths. Complementarity is especially important
for conformationally labile substrates, and, in fact, high-fidelity
self-sorting based on length has been previously observed for
imine-based peptidic capsules.[9] In the present case, the lack
of length-dependence can be explained by a comparison of
the binding motifs in capsules (l-4b)2 and (d,l-5a)2. In both
capsules, the conserved binding motifs involve only the first
amino acid residue in each strand. The presence of a second
amino acid residue, Gly, in (d,l-5a)2 does not lead to any sig-
nificant elongation of the (d,l-5a)2 capsule compared with
(l-4b)2, because it only forms additional interconnections be-
tween the “pairs of strands”. However, the binding motifs in (l-
4b)2 and (d,l-5a)2 are slightly different (Figure 2e). The differ-
ence can be minimized provided that a twist of the hemi-
spheres with respect to each other is possible, resulting in
a change in the binding motif from a 10- to a 14-membered
hydrogen-bonded ring. Assisted by such a rotation, strands
composed of amino acid derivatives and dipeptides could
easily fit into the same binding motif, thereby leading to inef-
fective length-based self-sorting.
Self-sorting
Various types of calixarene-based molecular capsules have
shown self-sorting ability at the stage of their noncovalent
self-assembly.[17] There are also several examples of the use of
peptides to induce chiral self-sorting.[7a,b] However, a combina-
tion of noncovalent stabilization and the dynamic nature of
chemical reactions that leads to effective self-sorting already at
the stage of covalent synthesis of such capsules has only been
reported by us for imine-based capsules.[9] In the present work,
we performed a series of experiments to test the possibility of
self-sorting for hydrazide substrates of different lengths and
different chirality. For the reaction of 1 with a mixture of hydra-
zides of the same length but different chirality, for example,
l-2a+d-2b, the exclusive formation of homochiral dimers was
observed (Figure 2n and Figures S30–32 in the Supporting In-
formation), in agreement with the previously observed prefer-
ence towards homochiral dimers (by both X-ray diffraction and
solution experiments). Effective chiral self-sorting also took
place when racemic mixtures of dipeptidic hydrazides were
used, for example, l-3a+d-3a, which led to the exclusive for-
mation of homochiral dimers (see Figure S30f). However, when
a mixture of hydrazides with the same length and the same
chirality was used, for example, l-2a+l-2b, no sorting was
observed (confirmed also by 13C NMR, ESI-MS, and HPLC, Fig-
ure 2m and Figures S30–32). These results indicate that the
chirality of the peptides governs the self-sorting process. The
lack of self-sorting in the case of two peptides of the same
chirality excludes the possibility that self-sorting is caused by
interactions that are specific to these two particular peptides
(e.g., CÀH···p interactions between the side-chains of Val and
Phe).
Complexation
The cavity volumes of the capsules were estimated by two dif-
ferent procedures. The calcVoid procedure[19] implemented in
3
3
OLEX2[20] gave 801 Š for (d-4b)2 and 758 Š for (d,l-5a)2 (de-
fault solvent radius 1.42 Š, shrink truncation radius 1.42 Š). The
3
Spaceball program[21] gave a value of 830 Š for (d-4b)2. Based
on shape and size complementarity, we predicted that fuller-
ene C60 (V=549 Š3)[22] could fit into these cavities. Fullerene is
also considered a good probe for monitoring the intrinsic
open/close processes of capsules due to its size and lack of
highly directional interactions with the external surfaces of
capsules.
Chem. Eur. J. 2016, 22, 3148 – 3155
www.chemeurj.org
3151
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Complexation experiments in CDCl3 did not show any traces
of C60 encapsulation within the (l-4a)2 or (l-5a)2 capsules,
even after several weeks (at RT and 708C). Complexes also
failed to form after the addition of various amounts of MeOH,
which was shown to promote capsule disassembly. However,
when the reaction between 1 and hydrazide l-2b or l-3a was
carried out in the presence of stoichiometric amounts of C60 or
C70, the quantitative formation of the respective complexes
C60ꢀ(l-4b)2, C60ꢀ(l-5a)2, and C70ꢀ(l-5a)2 was observed. The
formation of these complexes was unambiguously confirmed
are removed while the rest of the capsule remains intact.
Therefore, even though covalent scrambling occurs, the com-
plexation or release of a large fullerene guest is still hampered,
because it requires a considerable opening.
Although the complexation of fullerenes by the pre-assem-
bled capsules did not proceed in solution, successful complex-
ation could be achieved by using mechanochemical methods
in the solid state. When a mixture of (l-5a)2 and C60 was ball-
milled for 1 hour, the complexation of C60 was observed. This
result indicates that mechanochemical methods can be valua-
ble tools for loading cargo into molecular capsules of high ki-
netic stability and that they are able to overcome kinetic barri-
ers more effectively than in solution. Although mechanochemi-
cal encapsulation in micelles or liposomes is widely known, to
the best of our knowledge, this is the first demonstration of
mechanochemically induced encapsulation in molecular con-
tainers. The mechanism of action, as in many other mechano-
chemical processes, remains mostly unknown. However, we be-
lieve that explanations involving conformational changes and/
or chemical reversibility under mechanochemical conditions
due to local overheating are plausible.
1
by clear new sets of signals in the H and 13C NMR and ESI-MS
spectra (Figure 3a–g and Figures S35–S39 in the Supporting In-
formation). After the formation of the complexes, the ful-
lerenes were not released, even under conditions in which co-
valent scrambling is observed for free capsules (CDCl3, 7 days,
708C).
These results indicate that the capsules have a high thermo-
dynamic affinity towards the complexation of C60 and C70;
therefore the equilibrium is shifted towards the formation of
complexes. However, the complexation of these guests is limit-
ed by a high kinetic barrier that precludes complexation inside
capsules that have already been formed. But the complexation
also does not proceed under the specific conditions in which
covalent scrambling occurs (exchange of peptidic arms by the
rupture of covalent bonds). It may be rationalized by assuming
that covalent scrambling proceeds through a “one-at-a-time”
mechanism, that is, at a single moment only one or two arms
The C60ꢀ(l-5a)2 complex was characterized by NMR spec-
troscopy and X-ray crystallography (C60ꢀ(l-5a)2 and C60ꢀ(d-
5a)2 are enantiomeric but otherwise identical, see Figure S42
in the Supporting Information). In the 13C NMR spectrum, the
signal of encapsulated C60 appears as a sharp singlet shifted
upfield by 2.0 ppm compared with the signal of the free guest
1
Figure 3. Complexation of fullerenes as shown by the H (a–d) and 13C (e–g) NMR spectra of a) (l-5a)2, b) (l-5a)2 +C60 after ball-milling, c) the product of the
reaction of 1+l-3a+C60, d) the product of the reaction of 1+l-3a+C70, e) the product of the reaction of 1+l-3a+C60, f) the product of the reaction of
*
*
*
1+l-3a+C70, and g) C70
(
capsules, free fullerene, encapsulated fullerene; CDCl3, 600 MHz, 300 K). X-ray crystal structures of h,j) C60ꢀ(d-5a)2 and
i,k) C70ꢀ(d-5a)2. The capsules are shown in stick representation and the fullerenes as van der Waals surfaces. The surfaces in (j,k) are color-coded for host–
guest contacts after subtraction of the sum of the van der Waals radii: red: <0 Š, white: =0 Š, and blue: >0 Š. The scale in the top and bottom panels is dif-
ferent for better visibility. The lines and angle in (i) show the mutual orientation of the principal axes of C70 and (d-5a)2. l) Electrostatic potential mapped onto
the total electron density isosurface of monomeric acylhydrazone (ab initio calculations, DFT B3LYP).
Chem. Eur. J. 2016, 22, 3148 – 3155
www.chemeurj.org
3152
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
molecule (Figure 3e). The symmetry rules for C60 predict that
all carbon atoms are homotopic, therefore, their chemical
shifts are not differentiated upon encapsulation in a chiral cap-
sule. The presence of a single signal for encapsulated C60 indi-
cates that spinning of the fullerene within the capsule is fast
on the NMR timescale. Despite this high rotational freedom in
solution, in the solid state the guest molecule is perfectly or-
dered. The C60 molecule in the crystal structure was modeled
and refined by using only two soft geometrical restraints (SADI
for equivalence of the two unique CÀC bonding distances),
and the experimental electron density showed a well-resolved
guest structure at full occupancy (Figure 3h and Figure S46). A
van der Waals representation of the host–guest complex indi-
cates that the geometric match is very good. However, the in-
tramolecular distances between the host and guest molecules
are systematically higher than the sum of the corresponding
van der Waals radii (Figure 3j). Interestingly, the shortest distan-
ces are observed between C60 and the acylhydrazone groups
(3.37 Š for Cfullerene···C=N).
In contrast to C60 complexation, the larger C70 guest complex
could only just be detected, even after prolonged mechano-
chemical procedures. However, the fullerene molecule could
be stoichiometrically complexed when the chemical reaction
of 1 and l-3a was carried out in the presence of C70. A com-
parison of the 13C NMR spectra of free C70 and C70ꢀ(l-5a)2
shows significant complexation-induced shifts and symmetry
changes. The 13C NMR spectrum of C70ꢀ(l-5a)2 contains seven
signals of equal intensity that have been attributed to com-
plexed C70 (Figure 3f). The spectrum of free C70 contains five
signals, two of which have an intensity twice that of the
others, in accordance with the D5h symmetry (in all solvents re-
ported, including the spectrum in CDCl3 recorded in this work,
Figure 3g). The change in symmetry upon complexation can
be attributed to the influence of the chiral environment. In
chiral environments, enantiotopic carbon atoms (related by
mirror planes) become diastereotopic and have different chem-
ical shifts. Thus, the resulting symmetry of C70 in a chiral envi-
ronment is D5. In addition, the hindered rotation of C70 within
the cavity would result in a further lowering of the symmetry
of both the C70 guest and the capsule. Because such a symme-
try reduction is not observed here, it must be concluded that
the tumbling of C70 in the cavity is fast. The X-ray crystal struc-
ture of C70ꢀ(d-5a)2 shows a well resolved C70 molecule inside
the cavity. Interestingly, the long axis of the C70 prolate sphe-
roid is not aligned with the long axis of the capsule but has an
approximate angle of tilt of 298 (angle in Figure 3i). A compari-
son of the three X-ray structures of (d-5a)2, C60ꢀ(d-5a)2, and
C70ꢀ(d-5a)2 indicates that the capsule skeletons have very sim-
ilar geometries with a symmetry close to C4 (deviations <1%,
see Figure S48 and movie in the Supporting Information). The
association motif is conserved and the internal cavity volumes
are similar in all structures. This indicates that the tilting of C70
inside the cavity is not caused by a distortion of the capsule
shape, induced, for example, by packing forces or, indeed, by
the guest molecule itself. The tilt must result, therefore, from
favorable host–guest interactions. Indeed, analysis of the van
der Waals contacts indicates that the host–guest distances are
shorter for C70ꢀ(l-5a)2 than for C60ꢀ(l-5a)2 (Figure 3k). The
shortest distances are observed between C70 and the acylhy-
drazone groups and glycine methylene hydrogen atoms of
(l-5a)2 (3.20 Š for Cfullerene···C=N and 2.38 Š for Cfullerene···HGly).
The stoichiometric and quantitative complexation of ful-
lerenes by peptide–resorcinarene capsules is non-intuitive due
to the apparently different chemical nature of the host and
guest molecules. Previously reported fullerene receptors were
based mainly on extended complementary aromatic surfaces,
for example, cyclotriveratrylenes and unmodified calixarenes,
cycloparaphenylenes,[23] corannulenes,[24] phthalocyanines,[25]
porphyrins,[26] anthracenes,[27] and even anthanthrene (six con-
jugated aromatic rings) macrocycles.[28] Even though large pro-
teins have also been used to stabilize fullerenes in water, the
stabilizing forces are largely unknown.[29] It has been postulat-
ed that the internal hydrophobic cavities of those proteins are
mainly responsible for the interactions. In the present capsules,
the shortest molecular interactions can be precisely deter-
mined, therefore they shed some light on the peptide–fuller-
ene interactions. Surprisingly, the shortest intramolecular dis-
tances are observed not between the fullerenes and (as ex-
pected) the aromatic rings, but involve the acylhydrazone
groups or the glycine methylene protons (part of the peptide
backbone). To elucidate the participation of the acylhydrazone
groups in the fullerene interactions, we performed ab initio
DFT calculations (at the B3LYP level of theory) on a monomeric
acylhydrazone unit (Figure 3l).[30] Calculations of the electrostat-
ic potential indicated that a slightly negative potential extends
from the aromatic ring to the acylhydrazone double bond.
Therefore, interactions of this electron-rich fragment with ful-
lerenes (which can be considered p acceptors) seem to be fa-
vorable. However, other factors may also contribute to the
overall stabilization of these complexes (e.g., solvophobic in-
teractions, as was previously suggested for the complexation
of C60 in naphthalenediimide nanotubes).[31]
Conclusion
A series of molecular capsules based on amino acid and pep-
tide derivatives attached to macrocyclic scaffolds have been
synthesized. The dynamic character of the acylhydrazone
moiety, used as the linker, and effective self-assembly enabled
highly selective syntheses of the capsules from racemic mix-
tures of hydrazides by high-fidelity chiral self-sorting. The cap-
sules have greatly enhanced thermodynamic and kinetic stabil-
ity compared with the previously reported capsules based on
imine linkers. The stability enhancement stems from the ro-
bustness of the noncovalent binding motif and from the
higher hydrolytic stability of the covalent linker. Although the
capsules retain their dynamic character, thereby allowing pro-
cesses such as chiral self-sorting and chiral self-assembly to
occur with high precision, other processes, like complexation
by pre-assembled capsules are markedly hampered. Therefore,
the encapsulation of fullerene buckyballs, which match the
size and shape of the cavity of the pre-assembled capsules,
cannot be accomplished in solution. However, as conclusively
demonstrated in this work, the complexation energy barrier
Chem. Eur. J. 2016, 22, 3148 – 3155
www.chemeurj.org
3153
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
anisotropic displacement parameters for all non-hydrogen atoms.
All hydrogen atoms were included at geometrically predicted coor-
dinates. For (d,l-5a)2, C60ꢀ(l-5a)2, and C70ꢀ(l-5a)2, stereochemical
restraints were applied to the bond lengths in the peptidic chains
and the fullerene molecules. Soft restraints were also applied to
thermal parameters as necessary (in disordered regions). Highly
disordered solvent molecules were removed from the electron
density by means of the Solvent Masking procedure in OLEX2.[20]
can be successfully overcome by utilizing the reversible charac-
ter of the covalent bonds, or by mechanochemical methods.
Experimental Section
The use of CDCl3 or CHCl3 as solvent was dictated by the experi-
mental conditions (e.g., in situ experiments in NMR tubes). The use
of deuterated solvent had no influence on the properties of the in-
vestigated capsules.
Synthesis of capsules: Tetraformylresorcin[4]arene 1[10] (0.05 mmol,
41 mg) and the respective hydrazide 2a–3b (0.2 mmol) were dis-
solved in CHCl3 (2 mL) in a sealed pressure vial. The mixture was
heated at 708C for 20 h. Then the reaction mixture was cooled
down and filtered. The solution was evaporated under reduced
pressure and vacuum-dried. Acylhydrazones 4a–5b (as noncova-
lent dimers (l-4a)2–(d-5b)2) were obtained as yellow solids in
yields of 92–98%.
CCDC 1060617 ((d-4b)2), 1060618 ((d,l-5a)2), 1060619 (C60ꢀ(l-5a)2),
and 1060620 (C70ꢀ(l-5a)2) contain the supplementary crystallo-
graphic data for this paper. These data are provided free of charge
by The Cambridge Crystallographic Data Centre.
Acknowledgements
This work was supported by the Foundation for Polish Science
(A.S. and H.J., grant POMOST/2011-4/10), the National Science
Center (M.J. and M.G., grant 2013/10M/NZ1/00251) and the
Wroclaw Centre for Networking and Supercomputing (grant
no. 299). The research leading to these results was funded by
the European Community’s Seventh Framework Programme
(FP7/2007-2013) under BioStruct-X (grant agreement no.
283570). We thank MAX IV Laboratory in Lund for the alloca-
tion of synchrotron radiation beamtime. We would like to ac-
Scrambling experiment: (l-4a)2 (0.006 mmol, 9.8 mg) and (l-4b)2
(0.006 mmol, 8.6 mg) were dissolved in CDCl3 (1 mL). The mixture
was stirred at 708C for 7 days in a sealed pressure vial. After cool-
1
ing, the sample was analyzed by H NMR spectroscopy and ESI-MS.
The presence of mixed products (scrambled) was detected.
Self-sorting
experiments:
Tetraformylresorcin[4]arene
1 (0.05 mmol, 41 mg) and two different hydrazides (0.1 mmol each)
were dissolved in CHCl3 (2 mL) in a sealed pressure vial. The reac-
tion mixture was stirred at 708C for 7 days. Then, the mixture was
cooled down and filtered. The solution was evaporated under re-
´
knowledge Prof. Kinga Suwinska for her help with in-house X-
1
duced pressure. The solid was analyzed by H and 13C NMR spec-
ray data collection, and Dr. Michał Michalak and Paweł Czer-
´
winski for their assistance with HPLC and ball-milling experi-
troscopy.
ments, respectively.
Complexation of fullerenes using mechanochemical methods:
(l-5a)2 (0.005 mmol, 16.4 mg) and C60 (0.005 mmol, 3.6 mg) were
placed in a grinding vial. The sample was ball-milled at a milling
rate of 500 minÀ1 for 1 h. Then the sample was dissolved in CDCl3
(1 mL) and filtered. The products were analyzed by NMR spectros-
copy. The complexation of C70 was studied by an analogous proce-
dure. The partial complexation of C60 (ca. 30%) and C70 (ca. 5%)
was detected.
Keywords: chirality · fullerenes · mechanochemistry · peptide
mimics · supramolecular chemistry
[1] a) D. G. Fatouros, D. A. Lamprou, A. J. Urquhart, S. N. Yannopoulos, I. S.
Vizirianakis, S. Zhang, S. Koutsopoulos, ACS Appl. Mater. Interfaces 2014,
6, 8184–8189; b) J. J. Panda, V. S. Chauhan, Polym. Chem. 2014, 5,
4418–4431.
[2] a) J. B. Matson, S. I. Stupp, Chem. Commun. 2012, 48, 26–33; b) H. Hos-
seinkhani, P. D. Hong, D. S. Yu, Chem. Rev. 2013, 113, 4837–4861; c) R.
Ravichandran, M. Griffith, J. Phopase, J. Mater. Chem. B 2014, 2, 8466–
8478.
[3] D. Li, E. M. Jones, M. R. Sawaya, H. Furukawa, F. Luo, M. Ivanova, S. A. Si-
evers, W. Wang, O. M. Yaghi, C. Liu, D. S. Eisenberg, J. Am. Chem. Soc.
2014, 136, 18044–18051.
[4] G. W. Collie, K. P. Ziach, C. M. Lombardo, J. Fremaux, F. Rosu, M. Decos-
sas, L. Mauran, O. Lambert, V. Gabelica, C. D. Mackereth, G. Guichard,
Nat. Chem. 2015, 7, 871–878.
Complexation of fullerenes during chemical reaction: Tetrafor-
mylresorcin[4]arene
1 (0.02 mmol, 16.5 mg), l-3a (0.08 mmol,
22.2 mg), and C60 (0.01 mmol, 7.2 mg) were dissolved in CHCl3
(2 mL) in a sealed pressure vial. The reaction mixture was stirred at
708C for 4 days. The mixture was cooled down and filtered. Then
the solution was evaporated under reduced pressure. The products
were analyzed by NMR spectroscopy. The complexation of C70 was
studied by an analogous procedure. Hydrazides d-3a and l-2b
were tested in a similar way. In all cases, the quantitative formation
of stoichiometric complexes was detected.
X-ray structure determination: Single crystals of (d-4b)2 were
grown by the slow diffusion of ethanol vapor into a chloroform so-
lution. Single crystals of (d,l-5a)2 were grown by the slow diffusion
of isopropanol vapor into a chloroform solution. Single crystals of
C60ꢀ(l-5a)2 were grown by the slow diffusion of acetonitrile vapor
into a chloroform solution. Single crystals of C70ꢀ(l-5a)2 were
grown by the slow diffusion of methanol vapor into a chloroform
solution. X-ray measurements were carried out by using an Agilent
SuperNova diffractometer (CuKa radiation) for (d-4b)2 and C70ꢀ(l-
5a)2 and at the MaxLab II synchrotron beamline 1911-3 (Lund,
Sweden) at l=0.8000 Š for (d,l-5a)2 and C60ꢀ(l-5a)2. The struc-
tures were solved by using SHELXS[32] (the capsules) or by molecu-
lar replacement using the PHASER[33] module of the CCP4 suite[34]
and capsule skeletons as molecular probes (in the case of fullerene
complexes).[35] The structures were refined by using SHELXL with
[5] B. E. I. Ramakers, J. C. M. Hest, D. W. P. M. van Lçwik, Chem. Soc. Rev.
2014, 43, 2743–2756.
[6] a) L. Adler-Abramovich, E. Gazit, Chem. Soc. Rev. 2014, 43, 6881–6893;
b) H. Cui, M. J. Webber, S. I. Stupp, Biopolymers 2010, 94, 1–18; c) S.
Fleming, R. V. Ulijn, Chem. Soc. Rev. 2014, 43, 8150–8177.
[7] a) N. Ponnuswamy, F. B. L. Cougnon, J. M. Clough, G. Dan Pantos, J. K. M.
Sanders, Science 2012, 338, 783–785; b) N. Ponnuswamy, F. B. L. Coug-
non, G. Dan Pantos¸, J. K. M. Sanders, J. Am. Chem. Soc. 2014, 136, 8243–
8251; c) J. M. A. Carnall, C. A. Waudby, A. M. Belenguer, M. C. A. Stuart,
J. J. P. Peyralans, S. Otto, Science 2010, 327, 1502–1506; d) J. W. Sado-
wnik, R. V. Ulijn, Curr. Opin. Biotechnol. 2010, 21, 401–411; e) R. L.
Schoch, L. E. Kapinos, R. Y. H. Lim, Proc. Natl. Acad. Sci. USA 2012, 109,
16911–16916; f) A. Pal, M. Malakoutikhah, G. Leonetti, M. Tezcan, M.
Colomb-Delsuc, V. D. Nguyen, J. van der Gucht, S. Otto, Angew. Chem.
Int. Ed. 2015, 54, 7852–7856; Angew. Chem. 2015, 127, 7963–7967;
g) A. Moure, S. V. Luis, I. Alfonso, Chem. Eur. J. 2012, 18, 5496–5500;
Chem. Eur. J. 2016, 22, 3148 – 3155
www.chemeurj.org
3154
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
h) Y. Krishnan-Ghosh, S. Balasubramanian, Angew. Chem. Int. Ed. 2003,
42, 2171–2173; Angew. Chem. 2003, 115, 2221–2223.
[22] G. B. Adams, M. O’Keeffe, R. S. Ruoff, J. Phys. Chem. 1994, 98, 9465–
9469.
ˇˇ ´
[23] T. Iwamoto, Y. Watanabe, H. Takaya, T. Haino, N. Yasuda, S. Yamago,
Chem. Eur. J. 2013, 19, 14061–14068.
[24] A. S. Filatov, M. V. Ferguson, S. N. Spisak, B. Li, C. F. Campana, M. A. Pet-
rukhina, Cryst. Growth Des. 2014, 14, 756–762.
[25] I. Sµnchez-Molina, C. G. Claessens, B. Grimm, D. M. Guldiand, T. Torres,
Chem. Sci. 2013, 4, 1338–1344.
[8] a) T. Friscic, Chem. Soc. Rev. 2012, 41, 3493–3510; b) D. W. R. Balke-
nende, S. Coulibaly, S. Balog, Y. C. Simon, G. L. Fiore, C. Weder, J. Am.
Chem. Soc. 2014, 136, 10493–10498.
[9] H. Je˛drzejewska, M. Wierzbicki, P. Cmoch, K. Rissanen, A. Szumna,
Angew. Chem. Int. Ed. 2014, 53, 13760–13764; Angew. Chem. 2014, 126,
13980–13984.
[26] X. Fang, Y. Z. Zhu, J. Y. Zheng, J. Org. Chem. 2014, 79, 1184–1191.
[27] N. Kishi, M. Akita, M. Yoshizawa, Angew. Chem. Int. Ed. 2014, 53, 3604–
3607; Angew. Chem. 2014, 126, 3678–3681.
[10] M. Grajda, M. Wierzbicki, P. Cmoch, A. Szumna, J. Org. Chem. 2013, 78,
11597–11601.
[11] G. R. L. Cousins, S. A. Poulsen, J. K. M. Sanders, Chem. Commun. 1999,
1575–1576.
[28] T. Matsuno, S. Sato, R. Iizuka, H. Isobe, Chem. Sci. 2015, 6, 909–916.
[29] a) H. Wu, L. Lin, P. Wang, S. Jiang, Z. Dai, X. Zou, Chem. Commun. 2011,
47, 10659–10661; b) M. Calvaresi, F. Arnesano, S. Bonacchi, A. Bottoni,
V. Calò, S. Conte, G. Falini, S. Fermani, M. Losacco, M. Montalti, G. Natile,
L. Prodi, F. Sparla, F. Zerbetto, ACS Nano 2014, 8, 1871–1877.
[30] Gaussian 09, Revision B.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, and D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[31] N. Ponnuswamy, G. D. Pantos¸, M. M. J. Smulders, J. K. M. Sanders, J. Am.
Chem. Soc. 2012, 134, 566–573.
[12] a) R. T. S. Lam, A. Belenguer, S. L. Roberts, C. Naumann, T. Jarrosson, S.
Otto, J. K. M. Sanders, Science 2005, 308, 667–669; b) F. B. L. Cougnon,
J. K. M. Sanders, Acc. Chem. Res. 2012, 45, 2211–2221.
[13] a) J. A. Prescher, C. R. Bertozzi, Nat. Chem. Biol. 2005, 1, 13–21; b) V. T.
Bhat, A. M. Caniard, T. Luksch, R. Brenk, D. J. Campopiano, M. F. Greaney,
Nat. Chem. 2010, 2, 490–497.
[14] a) A. Dirksen, S. Dirksen, T. M. Hackeng, P. E. Dawson, J. Am. Chem. Soc.
2006, 128, 15602–15603; b) P. Crisalli, E. T. Kool, J. Org. Chem. 2013, 78,
1184–1189.
[15] a) Y. S. Park, J. Park, K. Paek, Chem. Commun. 2015, 51, 6006–6009;
b) Y. S. Park, J. Park, K. Paek, Chem. Commun. 2013, 49, 6316–6318.
´
danac, D. Kontrec, S. Miljanic, Spectrochim. Acta Part A
¯
´
[16] a) N. Galic, I. Bro
2013, 107, 263–270; b) A. Manimekalai, N. Saradhadevi, A. Thiruvalluvar,
Spectrochim. Acta Part A 2010, 77, 687–695; c) V. I. Minkin, A. V. Tsuka-
nov, A. D. Dubonosov, V. A. Bren, J. Mol. Struct. 2011, 998, 179–191.
[17] a) H. Gan, B. C. Gibb, Chem. Commun. 2012, 48, 1656–1658; b) M. Chas,
G. Gil-Ramirez, P. Ballester, Org. Lett. 2011, 13, 3402–3405; c) M. Chas, G.
Gil-Ramirez, E. C. Escudero-Adan, J. Benet-Buchholz, P. Ballester, Org.
Lett. 2010, 12, 1740–1743; d) D. Ajami, J. L. Hou, T. J. Dale, E. Barrett, J.
Rebek, Proc. Natl. Acad. Sci. USA 2009, 106, 10430–10434; e) E. S. Bar-
rett, T. J. Dale, J. Rebek, J. Am. Chem. Soc. 2008, 130, 2344–2350; f) D.
Braekers, C. Peters, A. Bogdan, Y. Rudzevich, V. Bohmer, J. F. Desreux, J.
Org. Chem. 2008, 73, 701–706; g) A. Wu, L. Isaacs, J. Am. Chem. Soc.
2003, 125, 4831–4835.
[32] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112–122.
[33] A. J. McCoy, R. W. Grosse-Kunstleve, P. D. Adams, M. D. Winn, L. C. Storo-
ni, R. J. Read, J. Appl. Crystallogr. 2007, 40, 658–674.
[34] M. D. Winn, C. C. Ballard, K. D. Cowtan, E. J. Dodson, P. Emsley, P. R.
Evans, R. M. Keegan, E. B. Krissinel, A. G. W. Leslie, A. McCoy, Acta Crys-
tallogr. Sect. D 2011, 67, 235–242.
[18] S. Hovmçller, T. Zhou, T. Ohlson, Acta Crystallogr. Sect. D 2002, 58, 768–
776.
[19] B. Rees, L. Jenner, M. Yusupov, Acta Crystallogr. Sect. D 2005, 61, 1299–
1301.
[35] M. Wierzbicki, M. Gilski, K. Rissanen, M. Jaskolski, A. Szumna, CrystEng-
Comm 2014, 16, 3773–3780.
[20] O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard, H. Pusch-
mann, J. Appl. Crystallogr. 2009, 42, 339–341.
Received: November 5, 2015
[21] M. Chwastyk, M. Jaskolski, M. Cieplak, FEBS J. 2014, 281, 416–429.
Published online on January 25, 2016
Chem. Eur. J. 2016, 22, 3148 – 3155
www.chemeurj.org
3155
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim