Preparation of core-shell coordination molecular assemblies via the enrichment of structure-directing codes of bridging ligands and metathesis of metal units

Article abstract of DOI:10.1021/ja508822r

A series of molybdenum- and copper-based MOPs were synthesized through coordination-driven process of a bridging ligand (3,3′-PDBAD, L¹) and dimetal paddlewheel clusters. Three conformers of the ligand exist with an ideal bridging angle between the two carboxylate groups of 0° (H₂α-L¹), 120°(H₂β-L¹), and of 90°(H₂γ-L¹), respectively. At ambient or lower temperature, H₂L¹ and Mo₂(OAc)₄ or Cu₂(OAc)₄ were crystallized into a molecular square with γ-L¹ and Mo₂/Cu₂ units. With proper temperature elevation, not only the molecular square with γ-L¹ but also a lantern-shaped cage with α-L¹ formed simultaneously. Similar to how Watson-Crick pairs stabilize the helical structure of duplex DNA, the core-shell molecular assembly possesses favorable H-bonding interaction sites. This is dictated by the ligand conformation in the shell, coding for the formation and providing stabilization of the central lantern shaped core, which was not observed without this complementary interaction. On the basis of the crystallographic implications, a heterobimetallic cage was obtained through a postsynthetic metal ion metathesis, showing different reactivity of coordination bonds in the core and shell. As an innovative synthetic strategy, the site-selective metathesis broadens the structural diversity and properties of coordination assemblies.

Full text of DOI:10.1021/ja508822r

Article
pubs.acs.org/JACS
Preparation of Core−Shell Coordination Molecular Assemblies via
the Enrichment of Structure-Directing “Codes” of Bridging Ligands
and Metathesis of Metal Units
Jinhee Park,‡,∥,⊥ Ying-Pin Chen,‡,§,⊥ Zachary Perry,‡ Jian-Rong Li,*,†,‡ and Hong-Cai Zhou*,†,‡,§
†Department of Chemistry and Chemical Engineering, College of Environmental and Energy Engineering, Beijing University of
Technology, Beijing, 100124, P. R. China
‡Department of Chemistry, Texas A&M University, P.O. Box 30012, College Station, Texas 77842, United States
§Department of Materials Science and Engineering, Texas A&M University, College Station, Texas 77843, United States
∥Nano Hybrid Technology Research Center, Creative and Fundamental Research Division, Korea Electrotechnology Research
Institute, Changwon 641-120, Republic of Korea
S
* Supporting Information
ABSTRACT: A series of molybdenum- and copper-based MOPs
were synthesized through coordination-driven process of a
bridging ligand (3,3′-PDBAD, L1) and dimetal paddlewheel
clusters. Three conformers of the ligand exist with an ideal
bridging angle between the two carboxylate groups of 0° (H2α-
L1), 120° (H2β-L1), and of 90° (H2γ-L1), respectively. At ambient
or lower temperature, H2L1 and Mo2(OAc)4 or Cu2(OAc)4 were
crystallized into a molecular square with γ-L1 and Mo2/Cu2 units.
With proper temperature elevation, not only the molecular square
with γ-L1 but also a lantern-shaped cage with α-L1 formed
simultaneously. Similar to how Watson−Crick pairs stabilize the helical structure of duplex DNA, the core−shell molecular
assembly possesses favorable H-bonding interaction sites. This is dictated by the ligand conformation in the shell, coding for the
formation and providing stabilization of the central lantern shaped core, which was not observed without this complementary
interaction. On the basis of the crystallographic implications, a heterobimetallic cage was obtained through a postsynthetic metal
ion metathesis, showing different reactivity of coordination bonds in the core and shell. As an innovative synthetic strategy, the
site-selective metathesis broadens the structural diversity and properties of coordination assemblies.
the structural complexity of MOPs is still limited; new
strategies need to be developed to enrich MOP chemistry.
Nature utilizes a variety of weak interactions such as H-
bonding, π−π stacking, and hydrophobic interactions to build
various biological systems with relatively simple building units
containing structure-directing “codes”. For example, DNA
double strands are constructed based on Watson−Crick pairs
which rely on complementary H-bonding interactions resulting
from specific base pairs. In coordination driven molecular self-
assemblies, the building blocks (organic linkers and metal-
containing nodes) can be designed to possess structure-
directing “codes” such as built-in multiple H-bonding
interaction sites, intrinsic coordination modes, and variable
coordination angles, leading to the formation of MOPs with
desired structural features and functionalities.1 As the “codes”
directing the structure become more complicated, the complex-
ity of the coordination assembly will increase.
INTRODUCTION
■
Over the past two decades, coordination driven molecular self-
assemblies have been studied extensively due to their beautiful
structures, unique physical/chemical properties, and potential
applications. Among such self-assembled structures, a variety of
metal−organic polygons/polyhedra (MOPs) have been de-
signed and synthesized, using inspirations drawn from not only
geometric entities such as Archimedean and Platonic solids but
also biological systems with high symmetry.1−8 Both interior
and exterior functionalizations9−11 of MOPs have been carried
out, allowing the isolation of the confined space to be explored
toward molecular recognition,10,12 sensing,13 catalysis,14−18
regioselective19,20 and self-sorting reactions,21 molecular
flasks,22,23 storage of reactive species,24 and other potential
applications.24,25 Structures of the resulting MOPs are often
controlled by the coordination geometry of the metal or metal
units, and the bridging angles of the organic ligands.1,5,25−29
Recently, a few new synthetic methods, such as bridging-ligand-
substitution strategy and the use of mixed ligand systems, have
been developed in our laboratory to expand the structural
diversity of MOPs.30,31 However, compared to natural systems,
Herein, we report a series of MOPs, including novel core−
shell coordination molecular assemblies, derived from dimetal
Received: August 26, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja508822r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
paddlewheel units (Cu2 and Mo2 units) and multiple
conformers of organic linkers, 3,3′-((pyridine-2,6-dicarbonyl)-
bis(azanediyl))dibenzoate (3,3′-PDBAD, L1, Figure 3a) and its
positional isomer 4,4′-((pyridine-2,6-dicarbonyl)bis-
(azanediyl))dibenzoate (4,4′- PDBAD, L2, Figure 3b). The
protonated ligand H2L1 was synthesized from the condensation
reaction of 2,6-pyridinedicarbonyl dichloride and 3-amino-
benzoic acid via the formation of amide bonds, which are well-
known for H-bonding interactions with H-bond donors and
acceptors. Three conformers of H2L1 exist, namely anti−anti
(designated H2α-L1 for simplicity) with an ideal bridging angle
between the two carboxylate groups of 0°, syn−syn (H2β-L1)
with a bridging angle of 120°, and anti−syn (H2γ-L1) of 90°
bridging angle (Figure 1). In principle, H2β-L1 and H2γ-L1 have
Figure 1. Three different conformers of 3,3′-H2PDBAD, L1.
less steric hindrance than H2α-L1. It is thus anticipated that at
low temperature, H2β-L1 and H2γ-L1 may dominate because
they are energetically favorable; in contrast, all conformers can
coexist at high temperature. Unfortunately, we did not observe
β-L1 throughout the experiments.
In this context (schemed in Figure 2), at low temperature the
reaction between H2L1 and Mo2(OAc)4 or Cu2(OAc)4 gave rise
to a molecular square with γ-L1 and Mo2 or Cu2 units.32 When
the temperature was elevated, both a lantern-shaped cage with
α-L1 and a molecular square with γ-L1 formed simultaneously.
The cage was encapsulated inside the square through
multipoint H-bonding interactions. The enrichment of
structure-directing “codes” of the bridging ligand has thus
resulted in the formation of a unique core−shell coordination
molecular assembly, where a metal−organic polyhedron is
encapsulated inside a metal−organic polygon. This strategy
may have general implications in the synthesis of more
complicated coordination assemblies, including molecular and
extended structures. The formation of a double shell structure
was reported by Fujita and co-workers,33 where linkers
consisting of dual bispyridines were adopted, leading to the
formation of a sphere-in-sphere complex by “orthogonal” self-
assembly. Clever and co-workers demonstrated an inter-
penetrated dimeric cage, showing allosteric binding ability of
halide anions.34 As for the present system, a molecular square
and cage are nested to form a core−shell structure via
multipoint H-bonding interactions, reminiscent of H-bonding
interactions in the Watson−Crick pairs of a DNA double helix.
Figure 2. Overview scheme of all self-assembled structures in this
work.
isostructural except the following minor differences. The former
is sitting on a 4/m site and has crystallographically imposed C4h
symmetry, while the compound 2 has C2h symmetry. The
packing between MOPs is most affected by the coordinated
solvent molecules which change distance between individual
MOPs in the structures. The top views and side views of the
two close-packed structures can be found in Supporting
Information Figures S7 and S8. Both are derived from four
untwisted M2 paddlewheel units coordinated by four pairs of γ-
L1 ligands, whose conformation is kinetically trapped at low
temperature (see Figure 3a,c). It is also worth mentioning that
the interior surface of the molecular square is functionalized
with 16 amide groups and four Mo2 (or Cu2) units, which may
enable their utility in cooperative catalysis.35
When the isomeric bridging ligand, L2 (4,4′-PDBAD, Figure
3b), instead of L1 (3,3′-PDBAD, Figure 3a) was assembled with
Cu2 units, an octahedral cage (Figure 3d), [Cu2(L2)2(DMA)2]6·
S (3) was afforded. MOP 3 has a structure similar to the
previously reported molecular octahedron, [Cu2(CDC)2]6,
where a carbazole dicarboxylate ligand, 9H-carbazole-3,6-
dicarboxylate (CDC) with a 90° bridging angle was used.36
In 3, the ligand L2 also adopts a conformation with an
approximately 90° bridging angle as shown in Figure 3b.
Twelve L2 ligands act as edges bridging six Cu2 vertices giving
rise to the octahedral cage. The internal cavity size (referring to
the diameter of the largest sphere that can be accommodated
within the cage) and the edge length of a triangular window in
the octahedral cage are 20.18 and 14.10 Å, respectively. A
comparison between 2 (the Cu2 analogue of 1) and 3 indicates
that even with the same bridging angle and the same metal
RESULTS AND DISCUSSION
■
A molybdenum-based molecular square, [Mo2(γ-L1)2(DEF)2]4·
S (1, S represents uncoordinated solvent molecules) was
isolated as yellow plate crystals from the reaction of H2L1 and
Mo2(AcO)4 in N,N-diethylformamide (DEF) at room temper-
ature for 2 days under a N2 atmosphere. Similarly, a reaction
between H2L1 and Cu2(AcO)4(H2O)2 in N,N-dimethylaceta-
mide (DMA) at 4 °C for 2 days afforded green octahedral
crystals of [Cu2(γ-L1)2(DMA)2]4·S (2), the Cu2+ analogue of 1.
The structures of 1 and 2 were determined by single-crystal X-
ray diffraction analysis. Compounds 1 and 2 are almost
B
dx.doi.org/10.1021/ja508822r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 3. Structures of molecular assemblies 1 (c) and 3 (d) as well as
their 90° bridging-angle ligands, L1 (a) and L2 (b), respectively (H
atoms of ligands and coordinated solvent molecules were omitted for
clarity). Color scheme: Mo, pink; Cu, cyan; O, red; N, blue; C, black.
The green polygon and polyhedron represent geometries of 1 and 2
when considering metal clusters as vertices and ligands as edges.
Figure 4. (a) and (b) Structure of the core−shell molecular assembly
4 viewed from the c direction and a/b directions, respectively (H-bond
interactions are shown in thin green/red bonds); (c) structure of the
core molecular cage in 4; and (d) α-L1 in the core of 4 (non-H-
bonding H atoms of ligands and coordinated solvent molecules were
omitted for clarity). Panels a and c also show sites I, II, III, and IV of
Mo2 units. Color scheme: Mo, pink; O, red; N, blue; H, dark green; C,
black (in shell) or orange (in core).
node, the structures of the resulting molecular architectures
(herein the molecular square and octahedral cage) can be
totally different if the ligands contain intrinsically different
structure-directing codes.
Although 1−3 all contain carboxamide groups in their
organic linkers, MOP 3 is not soluble in common organic
solvents. MOPs 1 and 2, on the other hand, are soluble in
DMA and DEF, respectively, allowing the study of their
solution absorption spectra (Supporting Information Figure
S29). The absorption spectrum of 1 in DMA exhibits a
characteristic peak at 450 nm, corresponding to the δ → δ*
transition of the Mo−Mo quadruple bond.37 Because the Mo2
unit is sensitive to oxidation, the spectroscopic feature gradually
diminishes in air (Supporting Information Figure S30). The
absorption spectrum of 2 in DEF displays a feature at 713 nm
as shown in Supporting Information Figure S29. In addition,
preliminary CO2 sorption studies at 195 K were performed
with activated solid-state samples of 1 and 2 to confirm their
porosity (Supporting Information Figure S31). The BET and
Langmuir surface areas of 1 are 92.9 and 282.34 m2/g, and
those of 2 are 139 and 471 m2/g, respectively.
Sixteen H-bonds (shown in green/red thin line segments in
Figure 4a) between the core and shell have formed,
demonstrating the power of the structure directing “codes”,
which facilitate the formation of the core−shell structure and
stabilize the overall structure (Figure 4a,b). Attempted
syntheses of a stand-alone core structure or a Cu2-based
core−shell analogue were unsuccessful. In fact, at an elevated
temperature the mixture of Cu2+ ions and H2L1 formed a two-
dimensional (2-D) coordination polymer, 5, in which the L1
ligands adopt the γ conformation (Supporting Information
Figure S9).
Interestingly, we were able to utilize the core−shell structure
4 to prepare a heterobimetallic molecular assembly via
metathesis of the metal units. Metathesis has been widely
used to prepare relatively stable molecular assemblies by
replacing more labile metal ions with other metal ions by
forming kinetically inert coordination bonds.39,40 However, the
metathesis of MOPs in the form of single crystal to single
crystal transformation has been, to the best of our knowledge,
unreported likely due to their structural instability and low
porosity which could restrict ion flow in such single crystals. In
4, two types of distinct Mo2 units, one in the core and the other
in the shell are coordinated with two different conformers of L1.
Metal ion metathesis in a single crystal to single crystal
transformation can elucidate the reactivity difference of these
coordination bonds through the crystal structure analysis.
Crystals of freshly synthesized 4 were immersed in a 0.1 M
Cu(NO3)2 solution in N,N-dimethylformamide (DMF) under
an Ar atmosphere for 6 and 48 h. [With longer immersion time
(2 weeks), the crystals of 4 became amorphous, indicating
compound decomposition or transformation. This prevented
the structural determination at longer exchange times.] On the
In an attempt to induce conformational changes of L1 (thus
observe either α- or β-conformation), the reaction mixture of
H2L1 and Mo2(AcO)4 in DMA was heated to 85 °C under a N2
atmosphere. After 2 days, light yellow crystals of [Mo2(γ-
L1)2(DMA)2]4[Mo2(α-L1)2(DMA)2]2·S (4) were harvested
and its structure was solved by single-crystal X-ray diffraction
analysis. As expected, an additional conformer, α-L1, was found
in a coordination molecular assembly. Unexpectedly this
assembly was nested inside of the γ-L1 bridged square (Figure
4). As shown in Figure 4c, four α-L1 ligands bridges two Mo2
units, leading to a lantern-shaped molecular cage (the “core”).
The inner cavity size of the cage is 6.52 Å (distance between
the two inner Mo ions of the Mo2 units) × 16.43 Å (distance
between the two nitrogen atoms in the two pyridyl rings of the
two opposite α-L1 ligands).38 The γ-L1 ligands form a
molecular square (the “shell”) identical to 1. The lantern
shape molecular cage is encapsulated inside the square, yielding
a core−shell assembly (4). Assembly 4 exhibits C4h symmetry,
with a 4-fold axis passing through the Mo−Mo quadruple bond.
C
dx.doi.org/10.1021/ja508822r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
basis of energy-dispersive X-ray spectroscopic studies (EDS),
Cu/Mo molar ratios for Cu2+ exchanged 4 (denoted as 4′)
remained the same (1:1.1) after 6 or 48 h of exchange reaction
(Supporting Information Table S2). Cu and Mo mapping
revealed homogeneous metathesis throughout the crystals
(Supporting Information Figure S15). This partial metathesis
suggests that some of the Mo2 units in 4 undergo exchange
with the Cu2 units readily but the rest are too inert to undergo
metal metathesis under given conditions. Note that the same
metathesis conditions (even after 5 days only a very small
amount of uptake ∼5% was seen, which, without convincing X-
ray analysis, we are unable to accredit to this exchange
mechanism) are unsuccessful for other 3d transition metals,
such as Mn2+, Co2+, Ni2+, Zn2+ ions (Supporting Information
Table S3). We have attributed this process to the relatively slow
exchange rates41 as well as preferred coordination geometry and
the stability of the resulting dimetal paddlewheels. Therefore,
we will focus on the Cu2+ exchanged 4 in this work.
(2.099 Å before metathesis), respectively. On the basis of
CCDC database, 2.560 Å is the typical distance from Cu to Cu
in a Cu−Cu paddle wheel (2.6 Å); in contrast, 2.217 Å is
similar to Mo−Mo quadruple bond length (2.1 Å). The
torsional angles around the metal ions of the core and shell
structure are 15.50° (0.09° before metathesis) and 2.63° (0.01°
before metathesis), respectively (Figure 5b,c). These observa-
tions indicate that the Mo2 units in the core cage were almost
replaced by the Cu2 units, while those in the shell were partially
substituted.42
The core-selective metathesis observed in this work could be
attributed to the several reasons. First, a close inspection of the
packing of the core−shell structure from both the crystallo-
graphic a and b directions reveals that the site II of the Mo2
units in the core is directly accessible from a 1-D channel
(Figure 5e). Even with axial ligands on the two Mo atoms,
there is still enough space for solvated Cu2+ ions to easily reach
the Mo2 units, which is why the exchange reaction is fast. In
fact, once the Cu salt was added, an immediate color change of
the crystal was observed. The M2 units in the shell (sites III and
IV), on the other hand, are not directly accessible due to the
coordinated ligands. We postulate that during the metathesis
reaction, the quadruply bonded Mo2 units must be solvated and
then Cu2+ ions replace them. The excess Cu2+ ions could then
participate in the shell metathesis. The shell metathesis,
however, was not complete. Once the 1:1.1 molar ratio of
Cu/Mo has been reached, either elongating the reaction time
(2 weeks) or increasing the concentration of Cu2+ solution did
not lead to further metathesis. This phenomenon can be
described by a process of thermodynamic equilibrium. It has
been reported that metal centers connected by flexible ligand
environments can participate in competing equilibria due to
environmental conditions and cation identity.43,44 For the core,
the α-L1 ligand produces stress on the Mo2 units, which
energetically favors relatively flexible Cu2 units to release their
tension. We believe the quadruply bonded Mo−Mo species,
which is positionally locked due to orbital symmetry, prevents
the release of this tension. As there is no bond in the Cu2 unit,
the torsional strain can be relieved by a twist in the core
conformation. In contrast, the shell possesses less stress than
the core does thus a partial substitution between the Mo2 units
and Cu2+ ions is enough to release their tension.
To investigate where and how the metathesis occurred, the
structure of 4′ was characterized by single-crystal X-ray
diffraction analysis (Figure 5a, right), which revealed that the
heterobimetallic assembly, [Mo2(γ-L1)2]4[Cu2(α-L1)2]2·S (4′)
was the final product (Figure 5a). In the structure of 4′, the
interatomic distances of the M2 units in the core and shell
structures are 2.560 Å (2.100 Å before metathesis) and 2.217 Å
Anisotropic atomic displacement parameters (ADP, Uij,) of
the metals in 4 and 4′ were analyzed in order to confirm the
complete metathesis in the core and partial one in the shell.
ADP, namely, thermal ellipsoids, could be regarded as
spreading of electron diffusion around their equilibrium
positions on account of positional disorder, inverse-propor-
tionally reflecting confinement of atoms45−47 (see details in
Supporting Information section 2.4). It can be quantified and
compared by the ellipsoid volumes proportional to multiplicity
of eigenvalues of the ADP matrices (eq (2.5.9) in Supporting
Information ). In Figure 5d, the volume of thermal ellipsoids
before metathesis at site III and site IV is obviously larger than
that at site I and site II. Since the thermal ellipsoids reflect local
flexibility of a structure, which is potentially related to the
reactivity,46,47 Figure 5d implies that the reactivity of the four
sites is different (Supporting Information Figure S11). As a
result, the Mo2 units at sites I and II have constrained thermal
motion of the electrons, like a compressed spring, thus being
more reactive. Graphically, the shape of the thermal ellipsoids
at sites I and II transformed from flat ellipsoids to spherical
ellipsoids after metathesis, indicating a release of the torsional
Figure 5. (a) Metal ion metathesis in core−shell assembly 4 to give 4′.
Photographs show the color change of the crystals from red orange to
dark green during the metathesis; (b and c) views of the paddlewheel
metal-carboxylate centers in the core cages of 4 and 4′, respectively.
Color scheme: Mo, pink; Cu, cyan; O, red; N, blue; H, black green; C,
black (in shell) or orange (in core). (d) Thermal ellipsoids before
(colored) after metathesis (transparent) at sites I, II, III, and IV. (e)
The close packed structure of 4′ viewed along −b axis.
D
dx.doi.org/10.1021/ja508822r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
excess thionyl chloride was removed under vacuum and quenched with
ethanol and then water. Due to the reactivity of 2,6-pyridinedicarbonyl
chloride, the residual 2,6-pyridinedicarbonyl chloride was used without
further purification. 2,6-Pyridinedicarbonyl chloride (0.61 g, 3.0
mmol) was dissolved in 20 mL of anhydrous acetonitrile and added
dropwise to a solution of 3-aminobenzoic acid (0.96 g, 7.0 mmol) in
acetonitrile (20 mL). The resulting mixture was stirred at room
temperature for 24 h and then refluxed for 2 h. The precipitated white
solid was isolated by filtration and washed with 100 mL of acetonitrile
followed by 500 mL of water and then overnight drying under vacuum.
Yield for the two steps: 0.98 g, 81%. 1H NMR (300 MHz, DMSO-d6)
δ 11.19 (s, 2 H), 8.56 (s, 2 H), 8.45 (d, 2 H, J = 7.2 Hz), 8.34 (t, 1 H, J
= 7.6 Hz), 8.20 (d, 2H, J = 8.1 Hz). IR (ν max): 3379, 3329, 2816,
1678, 1604, 1589, 1519, 1454, 1408, 1315, 1288, 1242, 1184, 1118,
999, 933, 891, 856, 767, 740, 686 cm−1. Elemental Analysis (% calc/
found): C, 62.22/62.09; H, 3.73/3.76; N, 10.37/10.41. (Supporting
Information Figures S1−S3)
strain along c-axis. In addition, the H-bonds stabilized the core
structure during the metal substitution process, resulting in the
selective substitution of the core without the collapse of the
cage. It should be pointed out that the direct synthesis for such
a heterobimetallic molecular assembly from Cu2+ and Mo2+
mixtures yielded amorphous powders. Consequently, the core−
shell assembly can be used as a template for other
hereobimetallic molecular assemblies. The site-specific metal
substitution provides a good synthetic strategy to broaden the
structures of coordination driven molecular assemblies, which
are difficult to access through the direct assembly process.
Furthermore, the preparation of the heterobimetallic molecular
assemblies is a meaningful step in realization of enzyme-
mimetic supramolecules. In biological systems, various
heterometallic complexes are involved in cooperative enzymatic
reactions.48,49
Synthesis of 4,4′-H2PDBAD, L2. L2 was isolated from the same
procedure except using 4-aminobenzoic acid instead of 3-amino-
benzoic acid. Yield of 0.96 g, 79% for two steps. 1H NMR (300 MHz,
DMSO-d6) δ 11.28 (s, 2 H), 8.46 (d, 2 H, J = 7.5 Hz), 8.34 (t, 1 H, J =
7.5 Hz), 8.13 (d, 2 H, J = 7.8 Hz), 8.05 (d, 2H, J = 7.8 Hz). IR (ν
max): 3286, 3228, 2808, 1697, 1658, 1589, 1554, 1531, 1438, 1292,
1230, 1138, 1080, 999, 948, 898, 813, 756, 736, 659 cm−1. Elemental
Analysis (% calc/found): C, 62.22/61.97; H, 3.73/3.60; N, 10.37/
10.20. (Supporting Information Figure S4 ∼ S6)
CONCLUSIONS
■
We have reported a series of the MOPs, including the core−
shell molecular assemblies derived by the building blocks
(organic linkers and meal-containing nodes) possessing the
right structure-directing “codes” to facilitate the formation of
unfavorable conformations. The core−shell MOP was success-
fully employed as a template to prepare a heterobimetallic
assembly, in which the metal metathesis occurred favorably in
the core. The site-selective metal metathesis provides a novel
synthetic strategy to broaden the structures and properties of
coordination assemblies, including molecular and extended
structures. Synthetic strategies to take full advantage of the
built-in structure-directing codes to attain more complicated
molecular assemblies are currently being developed in our
laboratory.
Mo2(OAc)4 Synthesis. Under nitrogen, 2.99 g of Mo(CO)6 (11.3
mmol) was added to a stirring mixture of 100 mL glacial acetic acid
and 20 mL acetic anhydride. A vacuum was applied and the flask was
backfilled with nitrogen; this was repeated three times. The solution
was allowed to reflux for 32 h. At this point, stirring was stopped and
the flask was allowed to cool overnight, yielding crystals which
appeared to be green in color. After washing three times with ethanol,
the now bright yellow crystals were dried under reduced pressure.
Yield: 1.0176 g (2.2 mmol), 39%.
Synthesis of [Mo2(γ-L1)2(DEF)2]4·S 1. A mixture of H2L1 (20
mg), Mo(OAc)2 (20 mg), and N,N-diethylformamide (DEF) (1.5
mL) was placed in a 2 mL vial. The vial was purged with N2 and
allowed to sit in room temperature for 2 days. Yellow block crystals of
1 were obtained. IR (ν max): 3313, 3078, 1681, 1589, 1531, 1392,
1303, 1222, 1141, 1076, 979, 910, 840, 810, 756, 678 cm−1. Elemental
Analysis of activated 1 (% calc/found): C, 50.52/46.92; H, 2.62/3.66;
N, 8.42/9.01.
EXPERIMENTAL SECTION
■
Instruments. 1H nuclear magnetic resonance (NMR) data was
used to confirm the ligand syntheses, recorded on a Mercury 300
spectrometer at the Center for Chemical Characterization and Analysis
(CCCA). FT-IR analysis was conducted by use of an IRAffinity-1
instrument. The TGA data was obtained in a TGA-50 (SHIMADZU)
thermogravimetric analyzer with a heating rate of 5 °C/min under 25
mL/min N2 flow. Single crystal reflections were collected on a Bruker
single-crystal APEXII CCD Diffractometer with Mo Kα (λ = 0.71073
Å) at 110 K. The powder X-ray diffraction patterns (PXRD) were
recorded by a Bruker D8-Focus Bragg−Brentano X-ray Powder
diffractometer equipped with a Cu sealed tube (λ = 1.54178 Å) at a
scan rate of 0.5 s deg−1. Gas adsorption measurements were performed
by Micromeritics’ ASAP 2020 with extra-pure quality gases. Molar
ratio of copper and molybdenum of 4′ was determined by ICP-MS
(Agilent 7700x) analysis and SEM-EDS (FEI Quanta 600) equipped
with X-ray mapping and digital imaging. Vacuum dried powdery 4′ was
dissolved in 200 μL of 30% trace-metal grade HNO3 (Fisher
Scientific), and diluted up to 8 mL with deionized H2O for ICP
analysis.
Synthesis of [Cu2(γ-L1)2(DMA)2]4·S 2. H2L1 (8 mg), Cu(OAc)2
(8 mg), and N,N-dimethylacetamide (DMA) (1.5 mL) were added 2
mL vial. The reaction took place at 4 °C for 2 days. Green block
crystals were collected. IR (ν max): 3294, 3078, 1678, 1531, 1431,
1392, 1307, 1222, 1145, 1080, 999, 910, 837, 759, 678 cm−1. Elemental
Analysis of activated 2 (% calc/found): C, 54.02/51.72; H, 2.81/4.58;
N, 9.00/11.02.
Synthesis of [Cu2(L2)2(DMA)2]6·S 3. 2,6-Lutidine (0.1 mL) was
added to the mixture of H2L2 (40 mg), Cu(NO3)2 (25 mg), and DMA
(15 mL) in 20 mL vial. The reaction took place at room temperature
in 1 day. Blue plate crystals were collected. IR (ν max): 3260, 3101,
1662, 1600, 1527, 1384, 1315, 1176, 1110, 1080, 1014, 860, 779, 744,
682 cm−1. Elemental Analysis of activated 3 (% calc/found): C, 54.02/
50.65; H, 2.81/3.86; N, 9.00/8.97.
Synthesis of [Mo2(γ-L1)2(DMA)2]4[Mo2(α-L1)2(DMA)2]2·S 4. A
mixture of H2L1 (20 mg) and Mo(OAc)2 (20 mg) in 1.5 mL of DMA
was prepared in a 2 mL vial, purged with N2, and heated to 85 °C for 2
days. Light red block crystals were isolated. IR (ν max): 3290, 3082,
1658, 1589, 1531, 1388, 1300, 1222, 1138, 1076, 956, 914, 840, 813,
756, 675 cm−1. Elemental Analysis of 4 (% calc/found): C, 50.52/
48.14; H, 2.62/3.82; N, 8.42/8.73.
Metal Ion Metathesis for Cu−Core Mo−Shell [Mo2(γ-
L1)2]4[Cu2(α-L1)2]2·S 4′. Due to the oxygen sensitivity of Mo2+
complexes, the entire sample preparation steps were conducted in
an Ar atmosphere glovebox. The as-synthesized single crystals of 4
were briefly washed with fresh DMF. Approximately 20 mg of the
crystals 4 was soaked in 2 mL of 100 mM Cu(NO3)2 in DMF for 6 h,
Chemicals. Copper acetate monohydrate, copper nitrate, 2,6-
pyridinedicarboxylic acid, 3-aminobenzoic acid, 4-aminobenzoic acid,
molybdenumhexacarbonyl, thionyl chloride, N,N-diethylformamide,
and 2,6-lutidine were purchased from Alfa Aesar Chemicals.
Acetonitrile and glacial acetic acid were supplied by BDH. N,N-
Dimethylacetamide was obtained from Acros organic. N,N-Dimethyl-
formamide was obtained from Macron fine chemicals. All starting
materials were used as received without further purification.
Synthesis of 3,3′-H2PDBAD, L1. 2,6-Pyridinedicarboxylic acid
(2.0 g, 12.0 mmol) and thionyl chloride (40.0 mL, 0.55 mol) were
mixed in a dried 250 mL round-bottomed flask. N,N-Dimethylforma-
mide (1.0 mL, 12.9 mmol) was carefully added dropwise, and the
mixture was stirred and heated at 75 °C for 6 h under N2 flow. The
E
dx.doi.org/10.1021/ja508822r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. Crystal Data and Results of Structure Refinement
1
2
3
4
4′
5
Mo shell
Cu shell
octahedral cage
Mo core−shell
Mo core/Cu shell
Cu 2D sheet
953696
Cu2
CCDC
953691
Mo8
C168H104N24O56
4122.27
Orange Octahedron Green Octahedron Blue Square plates
953692
Cu8
C168H104N24O56
3863.07
953693
Cu12
C252H156N36O84
5794.61
953694
Mo12
C252H156N36O80
953695
Formula
Cu5.6Mo6.4
C252H156N36O78
C58H62N10O16
Formula weight
Color/Shape
Crystal system
Space group
a (Å)
6119.41
Yellow Rod
Tetragonal
I4/m
5905.97
Green Rod
Tetragonal
I4/m
1282.26
Green Chunk
Monoclinic
P21/c
Tetragonal
I4/m
Orthorhombic
Cmce
Triclinic
P1
̅
34.623(2)
34.623(2)
16.4266(10)
90.00
23.524(5)
31.791(7)
48.797(10)
90.00
26.273(5)
26.687(5)
29.106(6)
90.067(3)
108.930(2)
105.381(2)
18527(6)
1
31.220(10)
31.220(10)
24.656(8)
90.00
31.22(3)
31.22(3)
25.05(3)
90.00
8.9180(12)
15.890(2)
20.734(3)
90.00
b (Å)
c (Å)
α (deg)
β (deg)
90.00
90.00
90.00
90.00
97.404(2)
90.00
γ (deg)
V (Å3)
90.00
90.00
90.00
90.00
19691(2)
2
36493(13)
4
24032(14)
2
24416(44)
2
2913.6(7)
2
Z
dcalcd. (g/cm3)
0.695
0.703
0.519
0.846
0.803
1.462
μ (mm−1
F(000)
)
0.287
0.501
0.370
0.352
0.446
0.809
4128
7840
2940
6128
5950
1332
θmax [deg]
26.00
26.00
24.98
24.00
24.82
26.00
Completeness
Collected reflections
Unique reflections
Parameters
99.9%
99.5%
97.8%
99.7%
98.2%
100.0%
21574
103537
10019
261
94146
132545
63673
86231
25881
10619
176
18320
9669
5724
245
679
237
411
Restraints
15
39
116
21
23
9
Rint
0.0741
0.0626
0.1625
0.0881
0.1693
1.002
0.0802
0.1591
0.3077
0.2137
0.3287
1.009
0.1041
0.0815
0.1519
0.2168
0.1634
0.880
0.1934
0.1353
0.2555
0.2542
0.2847
0.994
0.1836
0.1434
0.2659
0.3304
0.3150
1.004
0.0706
0.0441
0.0764
0.0872
0.0832
1.005
R1 [I > 2σ(I)]
wR2 [I > 2σ(I)]
R1 (all data)
wR2 (all data)
GOF on F2
Δρmax/Δρmin [e·Å−3
]
2.753/−0.938
2.006/−1.030
0.751/−0.776
1.319/−0.780
1.136/−0.705
0.810/−0.808
1 day, 2 days, and 20 days. During the soaking process, the solution
was refreshed with fresh Cu(NO3)2 solution every day. Single crystal
X-ray diffraction analysis was performed on the crystals directly taken
out from the solution without further washing process. For EDS
analysis, the solution was exchanged with fresh DMF for 2 days to
remove any remaining residues. These same conditions were applied
to the nitrate salts of Mn, Co, Ni, and Zn, which showed no metathesis
after 2 days by EDS analysis.
and validation reply can be found in Supporting Information (section
2.1). The CIF files can be obtained free of charge from the Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/
cif. CCDC numbers and the crystal information are listed in Table 1.
ASSOCIATED CONTENT
* Supporting Information
■
S
Crystallography data and figures of the MOPs and the
coordination polymer. Calculations of atomic displacement
parameters of the MOPs. SEM-EDS, FT-IR, PXRD, TGA,
UV−vis, and CO2 adsorption data of MOPs. This material is
available free of charge via the Internet at http://pubs.acs.org.
Single Crystal X-ray Crystallography. All MOP crystals were
mounted onto a loop from mother liquid for single crystal X-ray data
collection without further treatment. Diffractions were measured on a
Bruker Smart Apex diffractometer equipped with a Mo−Kα sealed-
tube X-ray source (λ = 0.71073 Å) and liquid nitrogen stream (110 K).
The data frames were recorded using the program APEX2 and
processed using the program SAINT (v7.68A) routine within APEX2
(v2012.2.0). Absorption and beam corrections based on the multiscan
technique were applied to the integrated data using SADABS (v2008/
1). All the structures were solved by direct method using SHELXS and
refined by full-matrix least-squares on F2 using SHELXL software.
Carbon atoms with enormous thermal ellipsoids were refined with
the help of EADP restrains on each ligand to clear the warnings about
nonpositive definite matrices. All non-hydrogen atoms needed to be
refined with anisotropic displacement parameters during the final
cycles. Organic hydrogen atoms were located in calculated positions
with isotropic displacement parameters set to 1.2 × Ueq of the attached
atoms. The solvent molecules were highly disordered, and attempts to
locate and refine the solvent peaks were not always successful.
Contributions to scattering due to these solvent molecules were
removed using the SQUEEZE routine of PLATON;50 structures were
then refined again using the data generated. . The refinement details
AUTHOR INFORMATION
Corresponding Authors
zhou@tamu.edu
■
jrli@bjut.edu.cn
Author Contributions
⊥These authors contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
J. Park was supported through Energy Frontier Research
Center funded by U.S. Department of Energy (DE-
SC0001015). Y.-P. Chen was partially paid on U.S. Department
of Energy (DE-FC36-07GO17033) and partially on National
F
dx.doi.org/10.1021/ja508822r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(25) Ni, Z.; Yassar, A.; Antoun, T.; Yaghi, O. M. J. Am. Chem. Soc.
2005, 127, 12752−12753.
Science Foundation (NSF CBET-0930079). Z. Perry thanks
the support of the U.S. Department of Energy ARPA-e (DE-
AR0000249) and the Office of Naval Research (N00014-13-1-
0753) as well as Texas A&M University. J.-R. Li -thanks the
support from the Natural Science Foundation of China (No.
21271015, 21322601), the Program for New Century Excellent
Talents in University (No. NCET-13-0647), and the Beijing
Municipal Natural Science Foundation (No. 2132013). H.-C.
Zhou gratefully acknowledges support from U.S. Department of
Energy (DE-SC0001015, DE-AR000073, and DE-FC36-
07GO17033). The synthesis and characterization of the
molecular squares reported here were funded by U.S.
Department of Energy (DE- SC0001015, DE-AR000073, and
DE-FC36-07GO17033) and Welch Foundation grant (A-
1725). The FE-SEM acquisition was supported by the NSF
grant DBI-0116835, the VP for research office, and the TX Eng.
Exp. Station. The authors thank Dr. Paul Lindahl for the
performance and valuable discussions on ICP-MS analysis.
(26) Northrop, B. H.; Zheng, Y.-R.; Chi, K.-W.; Stang, P. J. Acc.
Chem. Res. 2009, 42, 1554−1563.
(27) Cotton, F. A.; Lin, C.; Murillo, C. A. Acc. Chem. Res. 2001, 34,
759−771.
(28) Eddaoudi, M.; Kim, J.; Wachter, J. B.; Chae, H. K.; O’Keeffe, M.;
Yaghi, O. M. J. Am. Chem. Soc. 2001, 123, 4368−4369.
(29) Smulders, M. M. J.; Riddell, I. A.; Browne, C.; Nitschke, J. R.
Chem. Soc. Rev. 2013, 42, 1728−1754.
(30) Li, J.-R.; Zhou, H.-C. Nat. Chem. 2010, 2, 893−898.
(31) Li, J.-R.; Zhou, H.-C. Angew. Chem., Int. Ed. 2009, 48, 8465−
8468.
(32) Cotton, F. A.; Daniels, L. M.; Lin, C.; Murillo, C. A. J. Am.
Chem. Soc. 1999, 121, 4538−4539.
(33) Sun, Q.-F.; Murase, T.; Sato, S.; Fujita, M. Angew. Chem., Int. Ed.
2011, 50, 10318−10321.
́
(34) Freye, S.; Hey, J.; Torras-Galan, A.; Stalke, D.; Herbst-Irmer, R.;
John, M.; Clever, G. H. Angew. Chem., Int. Ed. 2012, 51, 2191−2194.
(35) Park, J.; Li, J.-R.; Chen, Y.-P.; Yu, J.; Yakovenko, A. A.; Wang, Z.
U.; Sun, L.-B.; Balbuena, P. B.; Zhou, H.-C. Chem. Commun. 2012, 48,
9995−9997.
REFERENCES
(36) Li, J.-R.; Timmons, D. J.; Zhou, H.-C. J. Am. Chem. Soc. 2009,
131, 6368−6369.
■
(1) Chakrabarty, R.; Mukherjee, P. S.; Stang, P. J. Chem. Rev. 2011,
111, 6810−6918.
(2) Cook, T. R.; Zheng, Y.-R.; Stang, P. J. Chem. Rev. 2012, 113,
734−777.
(37) Radius, U.; Breher, F. Angew. Chem., Int. Ed. 2006, 45, 3006−
3010.
(38) Li, J.-R.; Yakovenko, A. A.; Lu, W.; Timmons, D. J.; Zhuang, W.;
Yuan, D.; Zhou, H.-C. J. Am. Chem. Soc. 2010, 132, 17599−17610.
(39) Kim, Y.; Das, S.; Bhattacharya, S.; Hong, S.; Kim, M. G.; Yoon,
M.; Natarajan, S.; Kim, K. Chem.Eur. J. 2012, 18, 16642−16648.
(40) Das, S.; Kim, H.; Kim, K. J. Am. Chem. Soc. 2009, 131, 3814−
3815.
(41) Helm, L.; Merbach, A. E. Chem. Rev. 2005, 105, 1923−1960.
(42) Chisholm, M. H. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 2563−
2570.
(3) Sun, Q.-F.; Sato, S.; Fujita, M. Nat. Chem. 2012, 4, 330−333.
(4) Han, Y.; Li, J.-R.; Xie, Y.; Guo, G. Chem. Soc. Rev. 2014, 43,
5952−5981.
(5) Tranchemontagne, D. J.; Ni, Z.; O’Keeffe, M.; Yaghi, O. M.
Angew. Chem., Int. Ed. 2008, 47, 5136−5147.
(6) Prakash, M. J.; Lah, M. S. Chem. Commun. 2009, 3326−3341.
(7) Perry, J. J., IV; Perman, J. A.; Zaworotko, M. J. Chem. Soc. Rev.
2009, 38, 1400−1417.
(8) Fujita, M.; Yazaki, J.; Ogura, K. J. Am. Chem. Soc. 1990, 112,
5645−5647.
̌ ̌ ́ ̌
(9) Hamilton, T. D.; Papaefstathiou, G. S.; Friscic, T.; Bucar, D.-K. i.;
(43) Cui, X.; Khlobystov, A. N.; Chen, X.; Marsh, D. H.; Blake, A. J.;
Lewis, W.; Champness, N. R.; Roberts, C. J.; Schroder, M. Chem.
̈
Eur. J. 2009, 15, 8861−8873.
(44) Aime, S.; Botta, M.; Fasano, M.; Marques, M. P. M.; Geraldes,
C. F. G. C.; Pubanz, D.; Merbach, A. E. Inorg. Chem. 1997, 36, 2059−
2068.
MacGillivray, L. R. J. Am. Chem. Soc. 2008, 130, 14366−14367.
(10) Fang, Y.; Murase, T.; Sato, S.; Fujita, M. J. Am. Chem. Soc. 2012,
135, 613−615.
(45) Friedman, R.; Caflisch, A. Proteins: Struct., Funct., Bioinf. 2008,
(11) Schmidtendorf, M.; Pape, T.; Hahn, F. E. Angew. Chem., Int. Ed.
2012, 51, 2195−2198.
73, 814−827.
(46) Yuan, Z.; Zhao, J.; Wang, Z.-X. Protein Eng. 2003, 16, 109−114.
(47) Gana, R.; Rao, S.; Huang, H.; Wu, C.; Vasudevan, S. BMC
Struct. Biol. 2013, 13, 6−20.
(12) Nakamura, T.; Ube, H.; Shiro, M.; Shionoya, M. Angew. Chem.,
Int. Ed. 2013, 52, 720−723.
(13) Clegg, J. K.; Cremers, J.; Hogben, A. J.; Breiner, B.; Smulders,
M. M. J.; Thoburn, J. D.; Nitschke, J. R. Chem. Sci. 2013, 4, 68−76.
(14) Pluth, M. D.; Bergman, R. G.; Raymond, K. N. Science 2007,
316, 85−88.
(48) Sasai, H.; Arai, T.; Satow, Y.; Houk, K. N.; Shibasaki, M. J. Am.
Chem. Soc. 1995, 117, 6194−6198.
(49) Anaya de Parrodi, C.; Walsh, P. J. Angew. Chem., Int. Ed. 2009,
48, 4679−4682.
(15) Pluth, M. D.; Bergman, R. G.; Raymond, K. N. Angew. Chem.,
Int. Ed. 2007, 46, 8587−8589.
(16) Hastings, C. J.; Pluth, M. D.; Bergman, R. G.; Raymond, K. N. J.
Am. Chem. Soc. 2010, 132, 6938−6940.
(50) Sheldrick, G. M. SHELXTL, Version 6.14, Structure
Determination Software Suite, Bruker AXS: Madison, WI, 2003.
(17) Brown, C. J.; Miller, G. M.; Johnson, M. W.; Bergman, R. G.;
Raymond, K. N. J. Am. Chem. Soc. 2011, 133, 11964−11966.
(18) Lu, W.; Yuan, D.; Yakovenko, A.; Zhou, H.-C. Chem. Commun.
2011, 47, 4968−4970.
(19) Kumazawa, K.; Biradha, K.; Kusukawa, T.; Okano, T.; Fujita, M.
Angew. Chem. 2003, 115, 4039−4043.
(20) Yoshizawa, M.; Tamura, M.; Fujita, M. Science 2006, 312, 251−
254.
(21) Smulders, M. M. J.; Jimen
Int. Ed. 2012, 51, 6681−6685.
́
ez, A.; Nitschke, J. R. Angew. Chem.,
(22) Yoshizawa, M.; Klosterman, J. K.; Fujita, M. Angew. Chem., Int.
Ed. 2009, 48, 3418−3438.
(23) Osuga, T.; Murase, T.; Fujita, M. Angew. Chem., Int. Ed. 2012,
51, 12199−12201.
(24) Mal, P.; Breiner, B.; Rissanen, K.; Nitschke, J. R. Science 2009,
324, 1697−1699.
G
dx.doi.org/10.1021/ja508822r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX