Selectivity algorithm for the formation of two cryptand/paraquat catenanes

Article abstract of DOI:10.1021/ol9028463

(Figure Presented) Two [2]catenanes based on two different cryptand hosts, a dibenzo-24-crown-8-based cryptand and a bis(m-phenylene)-26-crown-8-based cryptand, have been synthesized and characterized. These two cryptand hosts only have a minor difference in chemical structure. However, this minor structural difference leads to big differences in the configurations and packing modes of the resulting catenanes. In addition, from the crystal structures of the two catenanes, we found that the cyclophane guest seems to have a selectivity algorithm and chooses the larger-size rings to go through and interlock with.

Full text of DOI:10.1021/ol9028463

ORGANIC
LETTERS
2010
Vol. 12, No. 4
760-763
Selectivity Algorithm for the Formation
of Two Cryptand/Paraquat Catenanes
Ming Liu, Shijun Li, Menglong Hu, Feng Wang, and Feihe Huang*
Department of Chemistry, Zhejiang UniVersity, Hangzhou 310027, P. R. China
fhuang@zju.edu.cn
Received December 9, 2009
ABSTRACT
Two [2]catenanes based on two different cryptand hosts, a dibenzo-24-crown-8-based cryptand and a bis(m-phenylene)-26-crown-8-based
cryptand, have been synthesized and characterized. These two cryptand hosts only have a minor difference in chemical structure. However,
this minor structural difference leads to big differences in the configurations and packing modes of the resulting catenanes. In addition, from
the crystal structures of the two catenanes, we found that the cyclophane guest seems to have a selectivity algorithm and chooses the
larger-size rings to go through and interlock with.
Chemists have been intrigued with the extent to which
chemical self-assembly can be explored.1 Self-assembly is
a process in which components spontaneously form ordered
aggregates.2 The studies of self-assembly are booming
nowadays in both biological and chemical systems.2 Actually,
the formation of single crystals from solution is a typical
and vivid example of chemical self-assembly. Single mol-
ecules in solution choose their own optimal way to assemble
and precipitate into solid materials driven by hydrogen
bonding, hydrophobic/hydrophilic, π-π stacking, and/or
other weak interactions.3 Therefore, by the study of the
information stored in the chemical compounds’ single
crystals, not only their structural features can be known but
also their self-assembly behavior can be understood. Great
progress has been achieved in exploiting different packing
modes influenced by the minor changes in chemical struc-
tures. On this basis, by appropriate design, some organic
crystals have exhibited applications in many fields.4
On the other hand, the formation of many self-assembly
systems relies on molecular recognition between hosts and
guests. A lot of supramolecular units such as pseudorotax-
anes, rotaxanes, and catenanes have been successfully
constructed by template-directed protocols.5 In recent years,
(4) (a) Garcia-Garibay, M. A. Acc. Chem. Res. 2003, 36, 491–498. (b)
Troisi, A.; Orlandi, G.; Anthony, J. E. Chem. Mater. 2005, 17, 5024–5031.
(c) Luo, X.; Ishihara, T. AdV. Funct. Mater. 2004, 14, 905–912.
(5) (a) Raymo, F. M.; Stoddart, J. F. Chem. ReV. 1999, 99, 1643–1663.
(b) Flood, A. H.; Nygaard, S.; Laursen, B. W.; Jeppesen, J. O.; Stoddart,
J. F. Org. Lett. 2006, 8, 2205–2208. (c) Vella, S. J.; Tiburcio, J.; Gauld,
J. W.; Loeb, S. J. Org. Lett. 2006, 8, 3421–3424. (d) Sobransingh, D.; Kaifer,
A. E. Org. Lett. 2006, 8, 3247–3250. (e) Liu, Y.; Bruneau, A.; He, J.; Abliz,
Z. Org. Lett. 2008, 10, 765–768. (f) Goldup, S. M.; Leigh, D. A.; Long,
T.; McGonigal, P. R.; Symes, M. D.; Wu, J. J. Am. Chem. Soc. 2009, 131,
15924–15929.
(1) Service, R. F. Science 2005, 309, 95.
(2) (a) Witesides, G. M.; Mathias, J. P.; Seto, C. T. Science 1991, 254,
1312–1319. (b) Zhang, S. Nat. Biotechnol. 2003, 21, 1171–1178. (c) Rieth,
S.; Wang, B.-Y.; Bao, X.; Badjic´, J. D. Org. Lett. 2009, 11, 2495–2498.
(d) Koshkakaryan, G.; Klivansky, L. M.; Cao, D.; Snauko, M.; Teat, S. J.;
Struppe, J. O.; Liu, Y. J. Am. Chem. Soc. 2009, 131, 2078–2079.
(3) (a) Desiraju, G. R. Crystal Engineering. The Design of Organic
Solids; Elsevier: Amsterdam, 1989. (b) Witesides, G. M.; Grzybowski, B.
Science 2002, 295, 2418–2421. (c) Northrop, B. H.; Khan, S. J.; Stoddart,
J. F. Org. Lett. 2006, 8, 2159–2162.
10.1021/ol9028463  2010 American Chemical Society
Published on Web 01/21/2010

using these supramolecular units, especially the ones based
on cyclodextrins6 and calixarenes7 and interlocked struc-
tures,8 as building blocks in crystal engineering has attracted
increasing interest not only because of their unique potential
applications but also due to their exemplary role of self-
assembly. Here we report two new [2]catenanes based on
two cryptand hosts, which have only a subtle structural
difference. It was found that this minor structural difference
influences the self-assembly process at two levels: (1) the
process of forming [2]catenanes and (2) the process of
assembly into 3D structure (i.e., forming single crystals).
Cryptands H1 and H29 are derivatives of dibenzo-24-
crown-8 (DB24C8) and bis(m-phenylene)-26-crown-8
(BMP26C8), respectively. They are different only by the
positions of the ether chains on the phenyl rings. Both of
them have been proven to be able to bind paraquat
derivatives (N,N′-dialkyl-4,4′-bipyridinium salts) strongly
and used in the efficient preparation of pseudorotaxanes
and rotaxanes.10,11
into their MeCN solutions. As shown in Figure 1, the host
and guest components of M1 have strong and compact
Figure 1. Ball-stick views of the X-ray crystal structures of
[2]catenanes M1 (a) and M2 (b). H1 and H2 are red, G is blue,
hydrogens are magenta, oxygens are green, and nitrogens are black.
PF6- counterions, other solvent molecules, and hydrogens except the
ones involved in hydrogen bonding were omitted for clarity. Hydrogen
bond parameters: H···O distance (Å), C-H···O(N) angle (deg),
C···O(N) distance (Å) A, 2.43, 145, 3.231; B, 2.54, 125, 3.167; C,
2.27, 140, 3.040; D, 2.59, 135, 3.316; E, 2.55, 139, 3.310; F, 2.55,
122, 3.135; G, 2.34, 166, 3.250; H, 2.47, 110, 2.929; I, 2.62, 146,
3.265; J, 2.47, 148, 3.543; K, 2.52, 135, 3.345; L, 2.38, 156, 3.255.
[2]Catenanes M1 and M2 were synthesized in 30.6% and
24.7% yields, respectively, based on the π-donor/π-acceptor
interaction of the cryptand H1 or H2 with paraquat deriva-
tives (Scheme 1).12
π-donor/π-acceptor interactions. The two aromatic rings of
the cryptand H1 are nearly parallel (torsion angle 11.5°) with
a centroid-centroid distance of 6.793 Å. The two electron-
deficient bipyridinium units in the cyclophane also show
good coplanarity: the torsion angle of the two pyridinium
rings inside the cryptand cavity is 12.3°, and the correspond-
ing value is 12.8° for the two outside pyridinium rings. The
interplanar separations between the phenyl rings on the
cryptand and their accompanying bipyridinium units are all
around 3.4-3.5 Å. These indicate good π-donor/π-acceptor
interactions between the cryptand H1 and the cyclophane
G. Six hydrogen bonds (A-F in Figure 1a) further stabilize
the interlocked structure. Specially, a ꢀ-pyridinium hydrogen
of the cyclophane G is hydrogen-bonded to the pyridine
nitrogen atom of H1. Actually, the design intention of
introducing a pyridine nitogen atom onto the third chains of
cryptands is to introduce a hydrogen bonding acceptor for
ꢀ-pyridinium hydrogens, which should increase host-guest
binding between cryptand hosts and paraquat derivatives.10,11
Furthermore, the inside phenyl ring of H1 is also in an
optimal position, which brings two [C-H···π] interactions
with the p-xylyl spacers of the tetracationic cyclophane G.
In the interlocked structure of M1, good integration of three
kinds of weak noncovalent interactions was achieved.
In the crystal packing structure of M1, M1 molecules form
chains, linear supramolecular polycatenanes (Figure 2a),
driven by hydrogen bonding (a pyridyl 4-hydrogen interacts
Scheme 1. Syntheses of Two Cryptand/Paraquat [2]Catenanes
Both of the [2]catenanes provided single crystals suitable
for X-ray analysis when grown by vapor diffusion of i-Pr2O
(6) Liu, Y.; You, C.-C.; Zhang, M.; Weng, L.-H.; Wada, T.; Inoue, Y.
Org. Lett. 2000, 2, 2761–2763.
(7) (a) Liu, Y.; Wang, H.; Zhang, H.-Y.; Wang, L.-H. Cryst. Growth
Des. 2005, 5, 231–235. (b) Guo, D. -S.; Su, X.; Liu, Y. Cryst. Growth
Des. 2008, 8, 3514–3517.
(8) (a) Loeb, S. J. Chem. Commun. 2005, 1511–1518. (b) Li, Q.; Zhang,
ˇ
W.; Miljanic´, O. S.; Knobler, C. B.; Stoddart, J. F.; Yaghi, O. M. Chem.
Commun. 2010, 46, 380–382.
(9) The synthetic route to H2 is shown in Scheme S1 (Supporting
Information). H1 can also be synthesized by a similar route, employing
methyl 3,4-dihydroxybenzoate as starting material.
(10) (a) Huang, F.; Switek, K. A.; Zakharov, L. N.; Fronczek, F. R.;
Slebodnick, C.; Lam, M.; Golen, J. A.; Bryant, W. S.; Mason, P. E.;
Rheingold, A. L.; Ashraf-Khorassani, M.; Gibson, H. W. J. Org. Chem.
2005, 70, 3231–3241. (b) Gibson, H. W.; Wang, H.; Slebodnick, C.; Merola,
(12) (a) Anelli, P. L.; Ashton, P. R.; Ballardini, R.; Balzani, V.; Delgado,
M.; Gandolfi, M. T.; Goodnow, T. T.; Kaifer, A. E.; Philp, D.; Pietraszk-
iewicz, M.; Prodi, L.; Reddington, M. V.; Slwain, A. M. Z.; Spencer, N.;
Stoddart, J. F.; Vicent, C.; Williams, D. J. J. Am. Chem. Soc. 1992, 114,
193–218. (b) Liu, M.; Li, S.; Zhang, M.; Zhou, Q.; Wang, F.; Hu, M.;
Fronczek, F. R.; Li, N.; Huang, F. Org. Biomol. Chem. 2009, 7, 1288–
1291. (c) Li, S.; Liu, M.; Zheng, B.; Zhu, K.; Wang, F.; Li, N.; Zhao, X.;
Huang, F. Org. Lett. 2009, 11, 3350–3353.
J.; Kassel, W. S.; Rheingold, A. L. J. Org. Chem. 2007, 72, 3381–3393
.
(11) (a) Zhang, J.; Huang, F.; Li, N.; Wang, H.; Gibson, H. W.; Gantzel,
P.; Rheingold, A. L. J. Org. Chem. 2007, 72, 8935–8938. (b) Wang, F.;
Zhou, Q.; Zhu, K.; Li, S.; Wang, C.; Liu, M.; Li, N.; Fronczek, F. R.;
Huang, F. Tetrahedron. 2009, 65, 1488–1494
.
Org. Lett., Vol. 12, No. 4, 2010
761

26-membered and 25-membered rings. The size effect and
steric hindrance effect were believed to be the main reasons
that introduced this interesting phenomenon. For example,
in M1, the kinetic effect may cause the bipyridinium salt a
slow threading of the smallest ring (24-membered ring).
However, this smallest ring can not accommodate the
bipyridinium unit, imposing strain on the molecule. Ulti-
mately, they find the most “comfortable” place with the help
of all kinds of weak interactions.
In the crystal structure of [2]catenane M2, the two aromatic
rings of cryptand H2 are nearly parallel (torsion angle 7.9°);
the centroid-centroid distance is 6.793 Å. As the π-electron-
accepting group, the bipyridinium units inside the cryptand
have good coplanarity with a torsion angle of 12.9°. The
interplanar separations between this π-acceptor unit and the
accompanying π-donor units (phenyl rings on cryptand H2)
are 3.217 Å (with the phenyl ring inside the cyclophane
cavity) and 3.378 Å (with the outside one). These parameters
indicate strong donor/acceptor interactions. Six hydrogen
bonds further stabilize the interlocked structure (G-L in
Figure 1b). Just like the case of M1 (Figure 1a), a
ꢀ-pyridinium hydrogen of the inside bipyridinium units forms
two hydrogen bonds with the pyridine nitrogen atom and an
ester oxygen atom on the third chain of the cryptand host in
the crystal structure of M2 (Figure 1b). Furthermore, two
strong [C-H···π] interactions can also be found between the
two hydrogens on the inside phenyl ring of the cryptand host
and the two p-xylyl spacers of the tetracationic cyclophane
G in M2 (Figure 1b).
Figure 2. Packing representation of M1’s self-assembly to form
the crystalline solid. (a) M1 molecules form a supramolecular main-
chain polycatenane-like structure driven by hydrogen bonding. (b)
The polycatenane chains grow to a “planar” structure driven by
π-donor/π-acceptor interactions. (c) The “planar” structures as-
semble to a 3D structure driven by the hydrogen bonding interac-
-
tions with the help of numerous PF6 anions.
The torsion angle of the bipyridinium rings outside the
cryptand (H2) cavity is 35.0°; they show poor coplanarity.
Therefore, in the packing mode, M2 molecules cannot grow
by continuous π-donor/π-acceptor stacking. Alternately, they
pack by the π-π stacking between the aromatic rings at four
directions in one plane (a phenyl ring of the cyclophane
stacks with a phenyl ring on the cryptand of the neighboring
[2]catenane, and then one pyridinium ring of the uncovered
bipyridinium unit stacks with a phenyl ring on the cyclophane
of the third [2]catenane molecule). In this way, they grow
to the plane-like structure. It is worth noting that some PF6
anions and solvent molecules help to form the “plane” by
H-bonding interactions. Finally, the planes assemble to the
3D structure by H-bonding interactions with the help of
with an ether oxygen atom on the neighboring cryptand).
Then, these steplike chains stacked to form “planes” driven
by continuous π-donor/π-acceptor interactions between the
outside pyridinium and phenyl rings (Figure 2b). This is a
common characteristic of reported donor/acceptor [2]cat-
enanes.12 At last, numerous PF6 anions occupy voids in the
planes through hydrogen bonds (C-H···F) to form an orderly
3D structure (Figure 2c). Therefore, there are at least two
kinds of weak interactions involved in the self-assembly
process of single-crystal formation from M1.
To our surprise, in the crystal structure of M2, the
cyclophane G threads through the two ethylene glycol chains
on cryptand H2 instead of through the other opening.
Cyclophane G is interlocked with the bis(m-phenylene)-26-
crown-8, and the 2,6-pyridyl linkage reinforces the interlock-
ing. In fact, this kind of cryptand/paraquat [2]catenane, in
which the guest ring is interlocked with the cryptand
molecular cage, can be looked at as the guest ring interlocked
by two of the three rings of the cryptand cage at the same
time. If the three rings of the cryptand cage have different
sizes, there will be some isomers when the interlocked pattern
is different. Therefore, there are three potential isomers for
[2]catenane M1 because of the different sizes of the three
rings in H1 but only two potential isomers for M2 because
of its higher symmetry. However, it seems that the guest
ring has a selectivity algorithm and chooses the two larger-
size rings to interlock with. In M1, the cyclophane ring chose
the larger rings, the 27-membered and 25-membered rings.
In M2, the cyclophane ring also chose the larger rings, the
-
numerous PF6 and solvent molecules (Figure 3).
1
M1 and M2 were further characterized by H NMR and
13C NMR spectroscopy, low- and high-resolution electrospray
ionization mass spectrometry, and UV-vis spectroscopy (see
Supporting Information). After the formation of the [2]ca-
tenanes, the signals of protons on both the cryptands and
cyclophanes split severely, which is a signature of the
reduced symmetry of the interlocked structures. However,
unlike the broad peaks of previously reported catenanes
prepared from bis(m-phenylene)-32-crown-10-based crypt-
ands,12b,c the signals of protons of M1 and M2 were quite
sharp. It implies that in the solution of M1 or M2 the
dynamic processes of the interlocked components are slower.
From their crystal structures, this phenomenon could be
understandable as various noncovalent interactions between
762
Org. Lett., Vol. 12, No. 4, 2010

nm, respectively, corresponding to their red and orange
colors. This difference may arise from their different modes
of interlocking. In M2, the cyclophane threads through the
crown ether part of the cryptand host, while in M1 it taco
complexes the crown ether part of the cryptand host. What’s
more, there are stronger π-donor/π-acceptor interactions
both between the two components and in the packing mode
of M1.
In summary, we have reported the syntheses and charac-
terizations of two new cryptand/paraquat [2]catenanes, which
only have a subtle difference in the cryptand structures.
However, this minor difference led to big differences in the
configurations and the packing modes of their corresponding
catenanes, M1 and M2. The X-ray crystal structures of the
catenanes showed the guest and host components interlocked
in very different ways. Furthermore, M1 and M2 adopted
totally different ways in assembling to the 3D solid state
structures. The subtle structural change in the hosts exert a
“butterfly effect”,15 bringing about major changes in crystal
engineering of the corresponding supramolecular systems.
In addition, from the crystal structures of M1 and M2, we
found that the guest molecule acts very intelligently; it seems
to have a selectivity algorithm and chooses the larger-size
rings to interlock with. These studies provide us with more
information about complicated self-assembly processes,
especially the important role of noncovalent interactions. On
the other hand, the donor/acceptor catenanes have exhibited
huge potential as building blocks in supramolecular archi-
tectures, which are enhanced by this investigation of their
packing process.
Figure 3. Packing representation of M2’s self-assembly to form
the crystalline solid. (a) M2 molecules grow by alternating π-π
stacking interactions to form a “line” structure. (b) The “lines” then
grow to the “planar” structure by π-π stacking interactions and
H-bonding interactions. (c) The “planar” structures assemble to the
3D structure by H-bonding interactions with the help of numerous
-
PF6 and solvent molecules.
Acknowledgment. This work was supported by the National
Natural Science Foundation of China (20774086 and 20834004),
National Basic Research Program (2009CB930104), Chinese
Universities Scientific Fund (2009QNA3008), and the Project-
sponsored by SRF for ROCS, SEM (J20080410). We thank
Dr. Xiaohe Miao for her kind assistance with X-ray crystal-
lography.
the host and guest components exist in M1 and M2. These
weak interactions hinder the relative movements of inter-
locked components, so the protons cannot exchange quickly
and coalesce to broad signals. Variable-temperature NMR
spectra studies13 of M2 showed that the activation barrier,
for the pirouetting process of the ethylene glycol chains of
the cryptand around the cyclophane G, was 19.3 kcal·mol-1;
this is higher than those of reported π-donor/π-acceptor
catenanes.14
Supporting Information Available: Synthetic procedures
and characterizations of compounds, crystal data for cat-
enanes M1 and M2, and other materials. This material is
available free of charge via the Internet at http://pubs.acs.org.
The UV-vis spectra of catenanes M1 and M2 showed
charge-transfer absorption bands centered at 410 and 352
OL9028463
(13) The method is called the “coalescence method”. See: Sutherland,
I. O. Annu. Rep. NMR Spectrosc. 1971, 4, 71–235.
(14) In typical crown ether-based donor/acceptor [2]catenanes, the
activation barrier is about 12-16 kcal·mol-1. See: ref 12 and: Halterman,
R. L.; Martyn, D. E.; Pan, X.; Ha, D. B.; Frow, M.; Haessig, K. Org. Lett.
2006, 8, 2119–2121.
(15) “Butterfly effect” is a phenomenon wherein small variations of the
initial condition result in significantly different outcomes. See: Gao, J.;
Delos, J. B. Phys. ReV. A 1992, 46, 1455–1467. Karkuszewski, Z. P.;
Jarzynski, C.; Zurek, W. H. Phys. ReV. Lett. 2002, 89, 170405.
Org. Lett., Vol. 12, No. 4, 2010
763