Interconversion between discrete and a chain of nanocages: Self-assembly via a solvent-driven, dimension-augmentation strategy

Article abstract of DOI:10.1021/ja306150x

Using a ligand bearing a bulky hydrophobic group, a "shish kabob" of nanocages, has been assembled through either a one-fell-swoop or a step-by-step procedure by varying the dielectric constant of the assembly mixture. A hydrophobic solvent breaks down th

Full text of DOI:10.1021/ja306150x

Communication
pubs.acs.org/JACS
Interconversion between Discrete and a Chain of Nanocages: Self-
Assembly via a Solvent-Driven, Dimension-Augmentation Strategy
Tian-Fu Liu, Ying-Pin Chen, Andrey A. Yakovenko, and Hong-Cai Zhou*
Department of Chemistry, Texas A&M University, PO Box 30012, College Station, Texas 77842, United States
S
* Supporting Information
adjacent layers are mutually interdigitated to form adjustable
ABSTRACT: Using a ligand bearing a bulky hydrophobic
group, a “shish kabob” of nanocages, has been assembled
through either a one-fell-swoop or a step-by-step
procedure by varying the dielectric constant of the
assembly mixture. A hydrophobic solvent breaks down
the chain to discrete nanocages, while a hydrophilic
solvent reverses the procedure. Although the shish kabob
of nanocages has exactly the same chemical composition
and even the same Archimedean-solid structure as those of
its discrete analogue, its gas-adsorption capacity is
remarkably improved because assembly of a chain exposes
the internal surface of an individual cage. This dimension-
augmentation strategy may have general implications in
the preparation of porous materials.
chambers with dynamic sorption properties. The interdigitation
of the bulky hydrophobic groups plays a vital role for the novel
structures and the unique temperature-responsive feature. It is
interesting to contemplate whether the interdigitaion could
happen between adjacent nanocages. In this case, it would
represent an effective route to prevent the rearrangement of
discrete cages upon activation and consequently improve
structural stability and maintain channel continuity in solids.
With this motivation, we undertook the synthesis of MOPs
using ligands with bulky hydrophobic groups. A surfactant-like
amphiphilic ligand, 5-((triisopropylsilyl)ethynyl) isophthalate
(TEI), consisting of a bulky hydrophobic end and two
hydrophilic carboxylate functional groups (Figure 1a), was
etal−organic polyhedra (MOPs), also known as
Mcoordination nanocages, have attracted great interest
in the past two decades.1 Different from metal−organic
frameworks (MOFs), MOPs are discrete metal−organic
molecular assemblies with well-defined cavities. Their unique
structural and chemical properties make them ideal candidates
for many applications, such as catalysis, drug delivery, and
molecular capsulation.2 Significant progress has been made in
rational design of MOPs. Some design strategies including
edge-directed and face-directed self-assembly,3 symmetry
interaction model, and molecular library approach,4 as well as
bridging ligand-substitution strategy,5 have been systematically
investigated. However, compared with MOFs, one of the
daunting challenges for MOPs is their very limited gas uptake,
which severely impairs its application potential. It is mainly
because MOPs, as discrete molecular assemblies, form crystals
through packing and often lose crystallinity upon activation.
Even though most of the MOPs may have internal cavities, the
molecular cages usually rearrange after solvent removal.
Therefore, the array of MOPs loses channel continuity in
solids, which hampers the diffusion of gas molecules. Hence,
finding ways of fixing the relative position of nanocages to
construct continuous open channels would dramatically
improve the effectiveness of these nanocages in gas storage
and other applications.
Figure 1. (a) Protonated TEI. (b) Crystal structure of 1. (c) Crystal
structure of 2, some of the triisopropylsilyl-ethynyl groups being
omitted for clarity. (inset) The coordination mode of carboxylate
groups linking neighboring cages.
adopted for this purpose. Combining TEI and Cu(II), we
reported a micelle-like cuboctahedral nanocage (denoted as 1
hereafter, Figure 1b) with 24 triisopropylsilyl (TIPS) groups on
the exterior.7 Herein, with the same ligand and metal source, we
present a novel nanocage-based chain structure in which the
bulky groups of the adjacent cuboctahedral cages are mutually
In the last five years, our laboratory has utilized a number of
carboxylate ligands with bulky hydrophobic groups to make a
series of mesh-adjustable molecular sieves, whose mesh size can
be tuned by temperature changes.6 The hydrophilic ends
(carboxylate groups) bind metal clusters to afford two-
dimensional layers, and the bulky hydrophobic groups of
Received: June 28, 2012
Published: October 8, 2012
© 2012 American Chemical Society
17358
dx.doi.org/10.1021/ja306150x | J. Am. Chem. Soc. 2012, 134, 17358−17361

Journal of the American Chemical Society
Communication
interdigitated, and two of the metal vertices are further linked
by two oxygen atoms from neighboring cages to form a shish
kabob of nanocages (denoted as 2 hereafter, Figure 1c). In
general, in a hydrophobic solution, the hydrophobic ends of a
surfactant tend to have more contact with the solvent, whereas
the hydrophilic ends try to have less contact with the solvent;
this is the driving force for micelle (or inverse micelle)
formation.8 As expected, a micelle-like compound, nanocage 1
was obtained in a hydrophobic mixture of solvents
(benzene:methanol = 19:1). As shown in Figure 1b, 1 is a
coordination nanocage assembled from 12 Cu2(CO2)4
paddlewheel motifs and 24 isophthalate bridging ligands to
give a cuboctahedron. It has 8 triangular (ca. 8 Å) and 6 square
(ca. 12 Å) windows with an enclosed void with a diameter of
ca.15 Å. Remarkably, by employing the same starting materials
but in a hydrophilic mixture of solvents (DMF:water =1:2), the
exact same cuboctahedral nanocages were assembled except the
mutual interdigitation of the TIPS groups from adjacent cages,
giving rise to a shish kabob of cuboctahedra (Figure 1c). This
kind of interdigitation was also observed between the adjacent
chains as shown in Figure S3 in Supporting Information (SI). A
close examination of the structure revealed that the two closest
Cu(II) ions from two adjacent cages were linked through two
carboxylate group in a bidentate-bridging mode with Cu−O
bond lengths of 2.272(11) and 2.426(14) Å, respectively
(Figure 1, inset). Presumably the hydrophobic effect is the
driving force behind the formation of 2. It is well-known that
there exists unusually strong attractions between hydrophobic
groups in a hydrophilic environment.9 Evidently, by adhering to
each other, hydrophobic groups can minimize their contact
with a hydrophilic environment. Such behavior was also widely
observed in natural processes, such as protein folding and
membrane formation.10 In the current case, the hydrophobic
effect forces the coordination cages to aggregate and leads to
the formation of two Cu−O bonds between a pair of adjacent
cages, thereby extending the structure to a 1D chain.
shown together with the mother liquor (Figure 2, center, left).
When suspended in CH2Cl2, 2 dissolved quickly to give a teal
solution (Figure 2, center, middle). Matrix-assisted laser
desorption ionization time of flight (MALDI-TOF) mass
spectrometry reveals a high-intensity peak of m/z 9902.69,
which corresponds to a formula of Cu24(TEI)24·6H2O (1). Teal
crystals of 1 were obtained when the CH2Cl2 solution was
layered with DMF, and the mixture was allowed to stay (Figure
2, center, right). The subsequent powder X-ray diffraction
(PXRD) studies further confirmed the complete conversion of
2 to 1 (Figure S5).
The dissociation of the 1D chain into discrete nanocages
presumably happened due to a solvation effect of methylene
chloride. Furthermore, a sample of 1 was immersed in 1.5 mL
DMF/H2O (with a volume ratio of 1:2, the same as that for the
synthesis of 2) in a 2 mL vial, which was subsequently kept at
85 °C for three days, after which PXRD patterns were collected,
indicating a complete conversion of 1 to 2. These solvent-
driven structural transformations further confirm the hydro-
phobic effect as a driving force in the formation of the shish
kabob of nanocages. For the transformation from 2 to 1, the
solvation of the hydrophobic nanocages by methylene chloride
is an exothermic procedure, which compensates for the energy
needs of breaking the intercage Cu−O bonds. In contrast, when
the hydrophobic 1 was subject to heating in a hydrophilic
solvent, the hydrophobic effect made the cages aggregate into a
1D chain with the formation of intercage Cu−O bonds,
reducing the hydrophobic−hydrophilic contact.
This type of stimuli-responsive synthetic strategy has been
applied in the self-assembly processes of other structures. For
example, some amphiphilic block copolymers can self-assemble
into micelles in response to environmental changes, such as
solvent, temperature, and pH.11 Herein, the aggregation of
nanocages, mimicking the stimuli responsiveness in protein
folding and membrane formation, was observed in MOPs for
the first time. It also provides us a new synthetic strategy to
achieve higher dimensional self-assembled arrays.
As anticipated, the interdigitation and the Cu−O bonds
between neighboring cages effectively prevent the rearrange-
ment of cages in the solid state upon activation and preserve
the structural integrity of 2, which are demonstrated by the
much-improved porosity and gas sorption capacities compared
with 1 (Table 1). The total potential solvent-accessible volume
Keeping these considerations in mind, we set out to
investigate whether changing the dielectric constant of the
solvent mixture can drive structural transformation between 2
and 1. In the center of Figure 2, an as-prepared sample of 2 was
Table 1. Adsorption Capacities (cm3·g−1) of 1 and 2 for N2,
O2, H2 (at 77 K), CO2, and CH4 (at 195 K)
compound
N2
O2
H2
Ar
CO2
CH4
1
2
5
8
41
6
71
23
82
214
216
115
160
169
is 36% estimated by using the PLATON routine.12 N2
adsorption study on an activated sample of 2 revealed a BET
surface area of 739 m2·g−1 (Langmuir 843 m2·g−1), among the
highest reported for MOPs to date.13 It is worth noting that 2
has a pore volume of 0.292 cm3·g−1 estimated based on
sorption isotherms, comparable to that estimated by using the
PLATON routine (0.317 cm3·g−1).12 Moreover, 2 retains its
porosity and crystallinity after many adsorption−desorption
cycles. These findings demonstrate the high stability and
permanent porosity of 2.
Figure 2. Illustrations of the self-assembly processes and interconver-
sion of cuboctahedral cages, 1 (bottom-right), and the 1D chain, 2
(top). Bottom-left, a TEI ligand and a dicopper paddlewheel. Center,
the as-synthesized 2 in mother liquid (left); dissolving 2 in CH2Cl2
(middle) leads to the formation of 1; and recrystallization in CH2Cl2/
DMF (right) gives teal crystals of 1.
The gas sorption isotherms of H2, CO2, and CH4 for 2 all
exhibit type I behavior, and all the gas uptakes are significantly
17359
dx.doi.org/10.1021/ja306150x | J. Am. Chem. Soc. 2012, 134, 17358−17361

Journal of the American Chemical Society
Communication
higher than those of 1 (Figure 3 and Table 1). Meanwhile, the
N2, O2, and Ar sorption isotherms of 2 exhibit hysteretic
a nanocage chain structure. The hydrophobic bulky group plays
a vital role in the formation of the novel structure. By changing
the hydrophobicity/hydrophilicity of the solvent mixture for
assembly, reversible transformation between the shish kabob
(2) and discrete nanocages (1) can be achieved. The solvent-
responsive process mimics the stimuli responsiveness of natural
processes, such as micelle formation, protein folding, and
membrane construction. More importantly, the stability and
gas-adsorption capacity of 2 were greatly improved compared
with those of 1. This provides a strategy for the assembly of
coordination nanocages with higher stability and permanent
porosity, which may have general implications in the control of
assembly processes through dimension augmentation.
ASSOCIATED CONTENT
■
S
* Supporting Information
Full details for sample preparation and characterization results,
and crystallographic data (CIF). This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
zhou@mail.chem.tamu.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the U.S. Department of Energy
(DE-SC0001015 and DE-AR0000073), the National Science
Foundation (CBET-0930079), and the Welch Foundation (A-
1725). We are thankful to Dr. Joseph H. Reibenspies and Dr.
George. M. Sheldrick to providing access to SHELXL 2012
software.
Figure 3. The sorption isotherms of N2, O2, H2, Ar at 77 K, and CO2,
CH4 at 195 K, for 1 and 2 (equilibrium time is 5 s for both 1 and 2).
behavior on the desorption-branch with an equilibrium time of
5 s during the measurement. To explore the cause of the
hysteresis, we measured Ar isotherms in different conditions.
The results indicated that the hysteresis loop was effectively
decreased when raising the adsorption temperature from 77 to
87 K. A similar phenomenon was observed when adsorbate was
introduced slowly into the pore by extending the equilibrium
time. However, the hysteresis cannot be completely eliminated
even when the equilibrium time was extended to 20 s (Figure
S9). This type of sorption behavior most likely arises from a
porous structure with small openings connecting large cavities,
leading to the trapping of relatively large gas molecules at low
temperatures.7 At higher temperatures, such as in the cases of
195 and 87 K or when the gas molecules are smaller, as in the
case of hydrogen albeit at 77 K, the hysteresis vanishes,
reminiscent of the adsorption behavior of the mesh-adjustable
molecule sieves.6,7 This phenomenon corroborates well with
the crystal structure, in which the nanocage possesses two types
of openings that are much smaller than the size of the cavity.
More importantly, the enhancement of gas adsorption capacity
of various gases of up to 43 folds (for N2) implies the power of
dimension augmentation. In going from 0D to 1D, the internal
surfaces of the molecular cages became more exposed. It is
tempting to extrapolate this dimension augmentation strategy
further to 2D and 3D; this work is currently under way in our
laboratory.
REFERENCES
■
(1) (a) Eddaoudi, M.; Kim, J.; Wachter, J. B.; Chae, H. K.; O’Keeffe,
M.; Yaghi, O. M. J. Am. Chem. Soc. 2001, 123, 4368. (b) Ghosh, K.;
Hu, J.; White, H. S.; Stang., P. J. J. Am. Chem. Soc. 2009, 131, 6695.
(c) Chen, C.; Zhang, J.; Su, C. Eur. J. Inorg. Chem. 2007, 2997.
(d) Lim, S. H.; Su, Y.; Cohen, S. M. Angew. Chem., Int. Ed. 2012, 51,
5106. (e) Liu, M.; Liao, W.; Hu, C.; Du, S.; Zhang, H. Angew. Chem.,
Int. Ed. 2012, 51, 1585. (f) Sun, Q.-F.; Sato, S.; Fujita, M. Nature
Chem. 2012, 4, 330.
(2) (a) Fer
Dutour, J.; Surble,
́
ey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F.;
S.; Margiolaki, I. Science 2005, 309, 2040. (b) Sun,
́
Q.-F.; Iwasa, J.; Ogawa, D.; Ishido, Y.; Sato, S.; Ozeki, T.; Sei, Y.;
Yamaguchi, K.; Fujita, M. Science 2010, 328, 1144. (c) Abrahams, B. F.;
FitzGerald, N. J.; Robson, R. Angew. Chem., Int. Ed. 2010, 49, 2896.
(d) Wang, Z. J.; Brown, C. J.; Bergman, R. G.; Raymond, K. N.; Toste,
F. D. J. Am. Chem. Soc. 2011, 133, 7358. (e) Mal, P.; Breiner, B.;
Rissanen, K.; Nitschke, J. R. Science 2009, 324, 1697. (f) McManus, G.
J.; Wang, Z.; Zaworotko., M. J. Cryst. Growth Des. 2004, 4, 11.
(3) Seidel, S. R.; Stang., P. J. Acc. Chem. Res. 2002, 35, 972.
(4) (a) Leininger, S.; Olenyuk, B.; Stang, P. J. Chem. Rev. 2000, 100,
853. (b) Tranchemontagne, D. J.; Ni, Z.; O’Keeffe, M.; Yaghi, O. M.
Angew. Chem., Int. Ed. 2008, 47, 5136.
(5) Li, J.-R.; Zhou, H.-C. Nat. Chem. 2010, 2, 893.
(6) (a) Ma, S.; Sun, D.; Wang, X.-S.; Zhou, H.-C. Angew. Chem., Int.
Ed. 2007, 46, 2458. (b) Ma, S.; Sun, D.; Yuan, D.; Wang, X.-S.; Zhou,
H.-C. J. Am. Chem. Soc. 2009, 131, 6445.
(7) Zhao, D.; Yuan, D.; Krishna, R.; van Baten, J. M.; Zhou, H.-C.
Chem. Commun. 2010, 46, 7352.
In conclusion, we have presented a shish kabob of nanocages
in which the discrete nanocages are mutually interdigitated by
bulky hydrophobic groups and linked by Cu−O bonds to form
(8) (a) Nagarajan, R.; Ganesh, K. J. Chem. Phys. 1989, 90, 5843.
(b) Alawi, S. M.; Akhter, M. S. J. Mol. Liquids 2011, 160, 63.
17360
dx.doi.org/10.1021/ja306150x | J. Am. Chem. Soc. 2012, 134, 17358−17361

Journal of the American Chemical Society
Communication
(9) (a) Southall, N. T.; Dill, K. A.; Haymet, A. D. J. J. Phys. Chem. B
2002, 106, 521. (b) Meyer, E. E.; Rosenberg, K. J.; Israelachvili, J. Proc.
Natl. Acad. Sci. U.S.A. 2006, 103, 15739.
(10) (a) Dill, K. A. Biochemistry 1990, 29, 7133. (b) Tanford, C. The
Hydrophobic Effect: Formation of Micelles and Biological Membranes;
Wiley Interscience: New York, 1980.
(11) Li, C.; Gu, C.; Zhang, Y.; Lang, M. Polym. Bull. 2012, 68, 69.
(12) Spek, A. L. Acta Crystallogr., Sect. A: Found. Crystallogr. 1990, 46,
194.
(13) (a) Lu, Z.; Knobler, C. B.; Furukawa, H.; Wang, B.; Liu, G.;
Yaghi, O. M. J. Am. Chem. Soc. 2009, 131, 12532. (b) Furukawa, H.;
Kim, J.; Ockwig, N. W.; O’Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc.
2008, 130, 11650.
(14) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.;
Pierotti, R. A.; Rouquer
1985, 57, 603.
́
ol, J.; Siemieniewska, T. Pure Appl. Chem.
17361
dx.doi.org/10.1021/ja306150x | J. Am. Chem. Soc. 2012, 134, 17358−17361