A Metal-Organic Tetrahedron as a Redox Vehicle to Encapsulate Organic Dyes for Photocatalytic Proton Reduction

Article abstract of DOI:10.1021/jacs.5b00832

The design of artificial systems that mimic highly evolved and finely tuned natural photosynthetic systems is a subject of intensive research. We report herein a new approach to constructing supramolecular systems for the photocatalytic generation of hydr

Full text of DOI:10.1021/jacs.5b00832

Article
pubs.acs.org/JACS
A Metal−Organic Tetrahedron as a Redox Vehicle to Encapsulate
Organic Dyes for Photocatalytic Proton Reduction
Xu Jing, Cheng He, Yang Yang, and Chunying Duan*
State Key Laboratory of Fine Chemicals, Dalian University of Technology, Dalian 116024, China
S
* Supporting Information
ABSTRACT: The design of artificial systems that mimic highly evolved and finely
tuned natural photosynthetic systems is a subject of intensive research. We report
herein a new approach to constructing supramolecular systems for the photocatalytic
generation of hydrogen from water by encapsulating an organic dye molecule into the
pocket of a redox-active metal−organic polyhedron. The assembled neutral Co4L4
tetrahedron consists of four ligands and four cobalt ions that connect together in
alternating fashion. The cobalt ions are coordinated by three thiosemicarbazone NS
chelators and exhibit a redox potential suitable for electrochemical proton reduction.
The close proximity between the redox site and the photosensitizer encapsulated in
the pocket enables photoinduced electron transfer from the excited state of the
photosensitizer to the cobalt-based catalytic sites via a powerful pseudo-intramolecular
pathway. The modified supramolecular system exhibits TON values comparable to
the highest values reported for related cobalt/fluorescein systems. Control
experiments based on a smaller tetrahedral analogue of the vehicle with a filled
pocket and a mononuclear compound resembling the cobalt corner of the tetrahedron suggest an enzymatic dynamics behavior.
The new, well-elucidated reaction pathways and the increased molarity of the reaction within the confined space render these
supramolecular systems superior to other relevant systems.
processes occurring in aqueous solution and the presence of a
well-defined pocket to force the proton reduction center and
photosensitizer into close proximity.17,18 Most importantly, to
construct highly efficient and easily operated systems for
photocatalytic hydrogen evolution from water, the supra-
molecular vessels should be designed such that at least one of
the two essential components exhibit redox activity and/or
light-harvesting ability.19,20
The supramolecular assembly of predesigned inorganic and
organic building blocks is an excellent tool for constructing
well-defined, nanosized molecular vessels that catalyze special
chemical transformations.21−28 Inspired by the pocket feature
of natural enzymes, researchers have developed functional
coordination polyhedra with various structures and catalytic
activities to establish the magnificent catalysis of natural
enzymes.28−34 By incorporating three thiosemi-carbazone
(TSC) bidentate chelators with potential guest-accessible sites
into the tripod ligand backbone, we developed a new approach
to assemble metal−organic polyhedra that act as a supra-
molecular host and redox catalyst for the photocatalytic
generation of hydrogen from water (Scheme 1). The strong
coordinating ability of the NS chelators was expected to
enhance the stability of the molecular polyhedron and to afford
cobalt ions with redox potentials suitable for proton
reduction.35,36 The pocket of the polyhedron was well-defined
1. INTRODUCTION
The design of artificial catalysts that can compete with the
catalytic proficiency of enzymes is a subject of intensive
research.1,2 The conventional route to fabricating an effective
artificial enzyme is to reproduce the sometimes elusive
structure of the enzyme’s active site.3,4 Recently reported
models have also been designed according to the idea of
enhancing the reaction rate by increasing the concentration of a
substrate around its reactive center.5,6 Accordingly, self-
assembly reaction vessels based on reversible interactions
have been considered a new phase of matter, in which the
physicochemical properties of the molecules contained in the
“molecular flask” are considerably modified with respect to
those exhibited in the solid, liquid, or gas phase.7 New reaction
pathways have emerged for such substrate molecules inside
these containers, by enhancing the proximity between of the
substrate and the catalytic center and increasing the effective
molarity of the reaction.8,9
Inspired by the significant advances and exciting results,10,11
we considered whether a supramolecular strategy could provide
ways of mimicking the highly evolved and finely tuned natural
systems for photocatalytic water splitting,12,13 which is an
important process both in living systems and in solar energy
conversion.14,15 The essential components of artificial homoge-
neous systems for hydrogen evolution always involve a
photosensitizer for light absorption, a catalyst for proton
reduction, and an electron donor.16 The intrinsic difficulties of
mimicking natural systems thus include the feasibility of the
Received: January 25, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.5b00832
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
splitting represents the significantly different chemical shifts of
the protons inside and outside the polyhedron.
Scheme 1. Procedure for the Synthesis of the Molecular
Tetrahedron and Construction of the Artificial
Single-crystal analysis clearly showed the formation of a
neutral Co4L4 tetrahedron with an ideal C3 symmetry via the
connection of four deprotonated ligands and four cobalt ions in
an alternating fashion (Figure 2). Each cobalt ion is
Supramolecular System for Photocatalytic Proton Reduction
to encapsulate organic dye molecules for the construction of
supramolecular photocatalytic systems. The proximity between
the redox catalytic sites and the photosensitizer was expected to
be enforced, and the odds of unwanted electron-transfer
reactions were reduced.37,38 The modified supramolecular
system was significantly superior to systems based on a
mononuclear cobalt compound resembling the cobalt corner of
the polyhedron or on a smaller tetrahedral analogue with an
almost solid sphere for the light-driven generation of hydrogen
from water.
2. RESULTS AND DISCUSSION
Figure 2. Structure of the molecular tetrahedron Co-TFT, showing
the coordination geometries of the cobalt ions and the empty sphere
(red ball) for the encapsulation of an organic dye, e.g., a fluorescein
molecule, respectively. The cobalt, sulfur, nitrogen, and carbon atoms
are drawn in green, yellow, blue, and gray, respectively.
The H3TFT ligand, 2,2′,2″-((nitrilotris(benzene-4,1-diyl))
tris(methanylylidene))tris(hydrazinecarbothioamide), was pre-
pared by the reaction of tris(4-formylphenyl)amine with
thiosemicarbazide in a 1:3 mol ratio. The reaction of H3TFT
with Co(CH3COO)2·4H2O in a DMF solution yielded 48% of
the molecular tetrahedron Co-TFT. Elemental and powder X-
ray analyses indicated that the bulk sample consisted of a pure
phase. Magnetic susceptibility measurements suggested a
diamagnetic behavior of the bulk sample, demonstrating the
presence of only CoIII ions in the Co-TFT polyhedron.
coordinated in a fac-configuration by three bidentate NS
chelators from three different ligands. The three anionic sulfur
donors are positioned on one side and hold great potential for
the attachment of a proton via hydrogen-bonding interactions,
thus shortening the distance between the redox sites and the
proton to be reduced. The average Co−S and Co−N bond
distances of approximately 2.30 and 1.96 Å, respectively, are in
good agreement with those of the related CoIII thiosemi-
carbazone complexes.39,40 The absence of any counterions
implies that the CS and C−N(H) bonds were all
transformed to C−S and CN bonds, respectively, via loss
of the imine protons during the coordination of the metal ions.
The measured C−S, C−N, and N−N bond distances
(Supporting Information Figures S1−S3) are all in the normal
range of single- and double-bond lengths, suggesting extensive
electron delocalization over the entire ligand skeleton.41,42 In
this case, the structure provides a conventional proton-transfer
path within the thiolate/thioamide tautomeric equilibrium in
the ligand backbone. The proton would be easily transferred
onto a suitable position during the proton reduction
process.43,44 These metal thiosemicarbazone (TSC) complexes
are therefore promising candidates for proton reduction
catalysts.45−48
The bonding of the ligands to the metal ions was also
confirmed by the splitting and shifting of the resonance signals
1
in the H NMR spectrum of Co-TFT (Figure 1). Specifically,
1
Figure 1. Partial H NMR spectra of the free ligand H3TFT (top
picture) and the molecular tetrahedron Co-TFT (bottom picture),
showing the disappearance of the imine proton C(S)−NH signals and
the splitting of the phenyl ring signals upon the coordination. The
peak at 7.92 ppm of the spectrum o f the complex Co-TFT is
assignable to the protons of the DMF molecules.
Importantly, the separation between the cobalt ions in the
tetrahedron was measured to be 15.58 Å, and the average
separation between two tertiary amine N atoms was measured
to be 11.36 Å. The inner volume of the pocket within the Co-
TFT tetrahedron was calculated as approximately 520 Å3, and
the rhombus opening exhibited two diagonal lengths of
approximately 15.6 and 10.1 Å. The structural feature of the
molecular tetrahedron most likely allows the encapsulation of
one fluorescein (Fl) molecule in its pocket, producing an
artificial supramolecular system for the light-driven generation
of hydrogen from water.
the disappearance of the imine proton C(S)−NH signals at
11.38 ppm in the free ligand H3TFT and the significant upfield
shifts of the NH2 protons from 8.15 and 7.91 ppm in the
thiosemi-carbazone moiety of the free ligand to one peak at
7.86 ppm in the spectrum of the complex Co-TFT suggest that
the bidentate moiety was coordinated to the metal ion. The
splitting of the phenyl ring signals at 7.76 and 7.06 ppm of the
ligand to 7.26 and 7.17 and 6.38 and 6.22 ppm, respectively, is
also an indicator of the formation of polyhedral species; this
B
DOI: 10.1021/jacs.5b00832
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
The ESI-MS spectrum of compound Co-TFT in the solution
exhibits an intense peak at m/z = 2417.19 that is assigned to the
single charged [Co4(TFT)4]− species, suggesting the formation
and stability of the M4L4 species in solution. When an
equimolar amount of fluorescein Fl was added into the solution
of Co-TFT, a new intense peak at m/z = 2749.23 was observed.
A simple comparison with the simulation results based on
natural isotopic abundances suggests that the peak is properly
assigned to a [Fl + Co4(TFT)4]− species (Figure 3). This result
implies the formation of a 1:1 stoichiometric complexation
species Fl ⊂ Co4(TFT)4 in the solution.
Figure 4. Partial NOESY spectrum of the Co-TFT (0.1 mM) and Fl
(0.3 mM) mixture in DMSO-d6, showing the potential interactions
between the protons of the phenyl ring of the fluorescein (HC) and the
tetrahedral cage (H4) (blue circles) and those between the protons of
the phenyl rings of the fluorescein (HD) and the molecular tetrahedron
(H3) (green circles).
Figure 3. ESI-MS spectra of Co-TFT in a methanol/acetonitrile
solution (top picture) and of Fl in the aforementioned solution
(bottom picture). The inserts show the measured and simulated
isotopic patterns at m/z = 2417.19 (top) and 2749.23 (bottom),
respectively.
Figure 5. (a) Cyclic voltammogram of 0.1 mM Co-TFT (black line)
and Co-TCT (red line), which is a smaller analogue of Co-TFT with
an almost solid sphere, in CH3CN containing 0.1 M TBAPF6. Scan
Rate: 100 mV/s. (b) Normalized fluorescence of Fl (10 μM, black
line) in an EtOH/H2O solution (1:1, pH = 11.6) and of the
aforementioned solution upon addition of Co-TFT(10 μM, red line)
or Co-TCT (10 μM, blue line).
The 1H NMR spectrum of Co-TFT (0.1 mM) in the
presence of an equimolar amount of Fl shows significant upfield
shifts of the signals associated with protons H3,6 (δ = 0.15 ppm)
and of other signals; these shifts provide an indicator39 for the
encapsulation of a molecule of Fl within the π-electron-rich
pocket of the Co-TFT tetrahedron to give the complexation
species Fl ⊂ Co4(TFT)4. The NOESY spectrum of the solution
of Co-TFT (0.1 mM) and fluorescein Fl (0.3 mM) shows
obvious H−H interactions between the phenyl rings of Fl (Hc)
and the tetrahedron (H4, blue cycles) and between the phenyl
rings of Fl (Hd) and the tetrahedron (H3, green cycles) (Figure
4). These results demonstrate the potential π−π interactions
between the phenyl ring of Fl and the phenyl ring of the
molecular tetrahedron, further confirming that the host−guest
supramolecular system was stabilized by specific noncovalent
interactions rather than by nonspecific static interactions.
Cyclic voltammetry of the molecular tetrahedron Co-TFT
(1.0 mM) recorded in a DMF solution shows the coupled CoII/
CoI and CoIII/CoII reduction process (at approximately −1.02
V and as a shoulder at approximately −0.78 V (vs. Ag/AgCl),
respectively, Figure 5a).50 The CoII/CoI potential falls well
within the range of that of proton reduction in aqueous
media,51 indicating that the Co-TFT complex in its reduced
state is capable of directly reducing protons. The molecular
tetrahedron Co-TFT is also demonstrated to be an efficient
quencher of the excited state of Fl. The addition of 10 μM Co-
TFT to the solution of Fl (10 μM) in an EtOH/H2O solution
(1:1 in volume and pH = 11.6, ensuring the same pH condition
as that of the reaction mixture mentioned below for
photocatalytic hydrogen evolution) quenched approximately
60% of the emission intensity of Fl (Figure 5b). Because the
addition of Co-TFT (0.1 mM) to the solution of Fl (0.1 mM)
in EtOH/H2O solution (1:1 in volume, pH = 11.6) did not
cause any obvious changes in the absorption spectra, the
quenching process is likely attributable to a classical photo-
induced electron transfer (PET) from the excited state Fl* to
the redox catalyst Co-TFT.52 As a consequence, Co-TFT was
directly activated for proton reduction by the excited state Fl*
during irradiation and the encapsulation of fluorescein within
the pocket of the tetrahedron would be a powerful method of
establishing new homogeneous systems for photocatalytic
hydrogen evolution.
The luminescence at 525 nm of a Fl solution (10 μM)
containing Co-TFT (50 μM) decayed in a clearly exponential
fashion with the lifetime, similar to the luminescence of a
solution of free Fl (4.50 ns, Supporting Information Figure
S10). Two luminescent species apparently coexisted in the
titration mixture. One is the Fl species itself with its fluorescent
lifetime being maintained; the other is the host−guest
complexation species Fl ⊂ Co-TFT. The fact that the decay
behavior approximates well to a typical single-exponential
function suggests that the host−guest complexation species Fl
⊂ Co-TFT exhibits ignored luminescence. Upon addition of up
to 50 μM Co-TFT, the luminescent titration profile of Fl (10
μM) in the solution was consistent with the Hill plot,53 and the
best fit of the titration profile suggested a 1:1 host−guest
behavior with an association constant calculated as 3.38
C
DOI: 10.1021/jacs.5b00832
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
( 0.18) × 104 M−1 (Supporting Information Figure S12). The
encapsulation of Fl inside the pocket of Co-TFT likely
facilitated the PET via a fast pseudo-intramolecular pathway,54
where the unwanted energy- and electron-transfer processes
could potentially be avoided.
The irradiation of a solution containing Fl (2.0 mM), Co-
TFT (4.0 μM), and triethylamine (NEt3) (10% v:v) in a H2O/
EtOH (1:1 in volume) solution at 25 °C resulted in direct
hydrogen generation.55 A higher efficiency of hydrogen
production was achieved at pH = 11.0−12.0 (Supporting
Information Figure S15). Basic conditions thermodynamically
favor water oxidation and facilitate further hydrogen evolution;
with water replacing the sacrificial electron donors.56 Control
experiments revealed that the absence of any of these individual
components led to failure to produce hydrogen, thus
demonstrating that all three species are essential for hydrogen
generation. Of course, the artificial system could not function
well in the absence of light.
catalyst.47 Simultaneously, the close proximity between the
redox site and the photosensitizer within the confined space
further allows the PET to occur in a more powerful pseudo-
intramolecular pathway to avoid unwanted electron transfer.
To further determine whether the photocatalytic hydrogen
evolution occurred inside the pocket of Co-TFT through a
typical enzymatic fashion or outside of the pocket in a normal
homogeneous manner, an inhibition experiment was carried
out by adding a nonreactive species, adenosine triphosphate
(ATP), into the reaction mixture.58 As shown in Figure 7, the
When the concentrations of Fl (2.0 mM) and NEt3 (10% in
volume) were fixed, the volume of the hydrogen produced
exhibited a linear relationship with the concentration of the Co-
TFT catalyst in the range from 2.0 to 12.0 μM (Figure 6a). The
Figure 7. (a) Volume of hydrogen produced by systems containing Fl
(2.0 mM), NEt3 (10% v/v), and redox catalysts (red bars) in EtOH/
H2O (1:1 in v:v, pH = 11.6): Co-TFT (10.0 μM), Co-TCT (10.0
μM), and Co-DMT (40.0 μM); the cyan bars show the
aforementioned systems in the presence of ATP(2.0 mM). (b)
Fluorescence spectra of Fl(10 μM) in EtOH/H2O (1:1 in volume, pH
= 11.6) upon addition of 0.5 mM ATP (black line), of the solution
upon addition of Co-TFT (50 μM, blue line), and of the solution
upon addition of both Co-TFT (50 μM) and 0.5 mM ATP (red line);
the spectra show the recovery of emission in the presence of ATP
when excited at 470 nm.
addition of 2.0 mM ATP efficiently quenched the photo-
catalytic hydrogen production by the Fl (2.0 mM)/ Co-TFT
(10.0 μM)/NEt3 (10%) system. The TON value of the
hydrogen produced in the presence of ATP(2.0 mM) was only
18% of the original Fl/Co-TFT/NEt3 system under the same
experimental conditions. The photocatalytic reaction exhibited
a competitive inhibition behavior, which means that the
reaction occurred within the pocket of the molecular
tetrahedron Co-TFT.49 Notably, the addition of ATP (0.5
mM) to the solution mixture containing Fl (10 μM) and Co-
TFT (50 μM) resulted in an emission recovery of the same
band. The addition of ATP (0.5 mM) to the solution of Fl (10
μM) did not cause any quenching of the emission of Fl (Figure
7b, dark line). The recovery of the emission of Fl confirmed the
substitution of the encapsulated Fl molecules in the pocket of
the molecular tetrahedron Co-TFT by ATP molecules.
Interestingly, the addition of ATP (0.4 mM) to a solution of
the molecular tetrahedron Co-TFT(0.1 mM) caused obvious
upfield shifts of the aromatic protons relevant to the adenosine
ring and a significant splitting of the signals relevant to the
trisphenylamine protons in Co-TFT (Supporting Information
Figure S8). Apparently, π···π stacking interactions occurred
between the adenosine ring of ATP and the benzene rings of
the molecular tetrahedron Co-TFT, providing an indicator for
the encapsulation of ATP molecules in the pocket of the
tetrahedron. Meanwhile, Co-TFT (10 μM) in a DMF solution
exhibited a weak emission band centered at 450 nm when
excited at 370 nm. Upon addition of up to 0.2 mM ATP into
the aforementioned solution, an emission enhancement of this
band of up to 210% was observed. The Hill plot of the titration
profile suggests a 1:1 host−guest behavior, with an association
constant for the ATP ⊂ Co-TFT inclusion species calculated as
Figure 6. Light-driven hydrogen evolution of (a) the systems
containing Fl (2.0 mM), NEt3 (10% v/v), and Co-TFT in an
EtOH/H2O solution (1:1, pH = 11.6) with the concentration of Co-
TFT fixed at 2.0 μM (black), 4.0 μM (red line), and 6.0 μM (blue
line); and (b) of the systems containing Fl (2.0 mM), NEt3 (10% v/v),
and Co-TCT in an EtOH/H2O solution (1:1, pH = 11.6) with the
concentration of Co-TCT fixed at 5.0 μM (black), 10.0 μM (red line),
and 15.0 μM (blue line).
initial turnover frequency (TOF) was approximately 750 mol of
hydrogen per mole of catalyst per hour, and the calculated
turnover number (TON) was approximately 11,000 mol of
hydrogen per mole of catalyst. These values are comparable to
the highest values reported for related cobalt/fluorescein
systems,35−38 demonstrating the potential advantage of the
supramolecular system in photocatalytic hydrogen production
from water. When the concentration of Fl was varied, the TOF
and TON increased until a plateau value was reached at a
concentration of 2.0 mM. Beyond this plateau value, the further
addition of Fl caused little TON enhancement relevant to the
Co-TFT catalyst. The volume of hydrogen produced was
apparently not dependent on the concentration of Fl in the
solution but was dominated by the concentration of Fl
encapsulated inside the pocket of the molecular tetrahedron
Co-TFT.
From a mechanistic viewpoint, the encapsulation of one
molecule of the organic dye Fl inside the pocket of the
molecular tetrahedron Co-TFT first enforces the proximity
between the cobalt-based redox catalytic site and the
photosensitizer Fl. This supramolecular structure then allows
a direct PET process from the excited state Fl* to the redox
D
DOI: 10.1021/jacs.5b00832
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
1.66 0.12 × 105 M−1 (Supporting Information Figure S11).
This value is approximately 5 times larger than that of the Fl ⊂
Co-TFT inclusion species, suggesting the possibility of ATP
molecules substituting the encapsulated Fl molecules and
occupying the pocket of the molecular tetrahedron.
exhibited two diagonal lengths of 10.1 and 3.94 Å. Because the
van der Waals radius of a monatomic molecule is approximately
3.6 Å, the sphere of Co-TCT was clearly almost solid and the
size of the pocket was too small for an Fl molecule to be
encapsulated.60 Because Co-TCT and Co-TFT exhibited
almost the same coordination structures and redox potentials,
Co-TCT was considered an ideal reference compound for Co-
TFT in the investigation of photocatalytic hydrogen evolution
from water within a supramolecular system.
The photolysis of a solution containing 0.2 mM F1 and Co-
TCT (10.0 μM) in a solvent mixture containing NEt3 (10% in
volume) and EtOH/H2O (1:1 in volume) resulted in hydrogen
generation under the photocatalytic conditions optimized at pH
= 11.6. The TOF was approximately 450 mol of hydrogen per
mole of catalyst per hour, and the TON was approximately
4500 mol of hydrogen per mole of redox catalyst (Figure 6b).
The volume of the generated hydrogen was approximately 40%
that produced by Co-TFT under the same experimental
conditions and with an equal concentration of the redox
catalyst. Experiments also showed that the addition of 2.0 mM
ATP to the Co-TCT(10.0 μM)/Fl(2.0 mM)/NEt3 system did
not obviously affect hydrogen evolution (Figure 7a), demon-
strating the normal homogeneous dynamics behavior of the
system. Because the two tetrahedra exhibit almost exactly the
same structural features and redox potentials, the superiority of
the Co-TFT/Fl/NEt3 systems over Co-TCT/Fl/NEt3 is
attributed to the well-emerged new reaction pathways of the
Co-TFT/Fl/NEt3 systems and the increased molarity of the
reaction within the confined space.
To further investigate the potential factors that influence the
photocatalytic hydrogen evolution process, a tetrahedral
analogue with the same structural features but an almost solid
sphere was designed and prepared. The H3TCT ligand, 2,2′,2″-
(benzene-1,3,5-triyltris(methanylylidene))tris(hydra-zinecarbo-
thioamide), was prepared via the reaction of thiosemicarbazide
with benzene-1,3,5-tricarbaldehyde in a 3:1 mol ratio. The
reaction of H3TCT with Co(CH3COO)2·4H2O in DMF
produced Co-TCT in 42% yield. Elemental and powder X-
ray analyses indicated the bulk sample consisted of a single pure
phase. The ESI-MS spectrum of the Co-TCT complex in the
solution exhibited [Co4(TCT)4]+ peak at m/z = 1748.89,
suggesting the formation of the similar M4L4 species in the
solution. However, attempts to obtain signals relevant to the
host−guest complexation species failed whenever various
concentrations of Fl were added to the solution. Cyclic
voltammetry of Co-TCT exhibited a broad peak at approx-
imately −1.03 V (vs Ag/AgCl, Figure 5a) assignable to the
coupled CoII/CoI and CoIII/CoII processes. This potential is
consistent with the Co-TCT tetrahedron and permits the
exploration of redox-induced reactions near the H2/H+ couple.
Crystal structure analysis revealed that Co-TCT exhibits
ideal T symmetry, with its cobalt ions and ligands located on
the crystallographic C3 axis (Scheme 2). Each cobalt center has
the same coordination environment as those in Co-TFT
(Supporting Information Figure S4). The average distance
between cobalt ions was 10.18 Å. The separation between the
center of the tetrahedron and the central points of the meta-
substituted benzenes was approximately 3.37 Å, giving an inner
volume of approximately 160 Å3. The rhombus opening
The luminescence titration of Fl (10 μM) exhibited a
quenching of approximately 38% of the luminescence intensity
upon the addition of up to 50 μM Co-TCT (Supporting
Information Figure S13). The decrease in the lifetime (from 4.5
to 4.0 ns) of Fl upon addition of Co-TCT (50 μM) confirmed
a normal intermolecular quenching process (Figure 8a). Given
Scheme 2. Structures of the Molecular Tetrahedra of the
Cobalt Ions and the Different Radii of the Pockets (red
a
balls)
Figure 8. Luminescence decay of Fl (10.0 μM, red line) in EtOH/
H2O (1:1, pH = 11.6) and of the aforementioned solution upon
addition of (a) Co-TFT (50.0 μM) or (b) Co-TCT (50.0 μM). The
intensity was recorded at 525 nm; the excitation wavelength was 472.6
nm.
that the concentration of NEt3 (10%, 0.72 M) was much higher
than that of Co-TCT in the reaction mixture, a PET from the
NEt3 to the excited state of the organic dye (reduction
quenching) dominated the homogeneous photolysis instead of
direct quenching by Co-TCT. Meanwhile, the decrease in
fluorescence lifetime of Fl (0.1 mM) upon the addition of 0.72
M NEt3 (3.6 ns) confirmed that the PET process from Et3N to
the excited state Fl* dominated the photocatalytic processes.
Notably, the direct quenching of the excited state Fl* by the
molecular tetrahedron produced Fl+ species and avoid the
formation of unstable PS− radical anions, thus prolonging the
a
Cobalt, sulfur, nitrogen, and carbon atoms are drawn in green, yellow,
blue, and gray, respectively.
E
DOI: 10.1021/jacs.5b00832
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
lifetime of the system.61 The well-emerged new PET pathway
contributed to the superiority of the supramolecular system
over normal homogeneous systems.
3. CONCLUSION
In summary, a new strategy for the construction of artificial
photocatalytic systems for the generation of hydrogen from
water by encapsulating an organic photosensitizer in the pocket
of a redox metal−organic polyhedron was reported. Within the
confined space, these essential components were forced closer
together, and the pseudo-intramolecular PET from the excited
state of the photosensitizer to the redox cobalt sites was
modified to avoid unwanted electron-transfer processes.
Control experiments based on the mononuclear compound
and a smaller tetrahedral analogue with a solid sphere suggested
that the superiority of the supramolecular systems over others
was attributable to the well-elucidated new reaction pathways
and to the increased molarity of the reaction within the
confined space. New supramolecular artificial photocatalytic
systems with improved efficiency were created by further
modification of the structures and redox potentials of the
metal−organic polyhedrons to match the electronic config-
uration and energy levels of the organic guests.
Whereas the cobalt ions exhibited uncommon coordination
geometry, the preparation of a mononuclear cobalt complex
containing a tristhiosemicarbazone moiety resembling the
corner of the molecular tetrahedron as another reference
material was rigorously pursued. The ligand HDMT, 2-(4-
(dimethylamino) benzylidene) hydrazinecarbothioamide, was
synthesized according to method described in the literature.62
The cobalt complex Co-DMT was synthesized in approx-
imately 66% yield by reacting Co(CH3COO)2·4H2O with
ligand HDMT with in DMF. Elemental analysis and powder X-
ray analysis indicated that the bulk sample consisted of a single
pure phase. Single-crystal structure analysis of revealed that the
cobalt center was affixed to three sulfur atoms and three
nitrogen atoms. These donors belonged to three bidentate
chelators, two of which were five-membered rings and the other
of which was a four-membered ring (Figure 9).
4. EXPERIMENTAL SECTION
4.1. Materials and Methods. All chemicals were of reagent-grade
quality, were obtained from commercial sources, and were used
1
without further purification. H NMR spectra were measured on a
Varian INOVA 400 and 100 M spectrometer. ESI-MS was carried out
on an HPLC-Q-Tof mass spectrometer. The elemental analyses of C,
H, and N were performed on a Vario EL III elemental analyzer. UV−
vis spectra were measured on an HP 8453 spectrometer. The solution
fluorescence spectra were measured on a JASCO FP-6500. The
solution of fluorescein was prepared in EtOH/H2O 1:1 (v:v), whereas
the solution of Co-TFT was prepared in DMF. The high-
concentration stock solution of ATP (10.0 mM) was prepared directly
in 9:1 DMF/H2O (v:v). Electrochemical measurements were
performed on a ZAHNER ENNIUM electrochemical workstation
with a conventional three-electrode system consisted of a homemade
Ag/AgCl electrode as the reference electrode, a piece of platinum silk
0.5 mM diameter as the counter electrode, and a glassy carbon
electrode as the working electrode. The measurements were
performed at room temperature after the solutions were degassed
with nitrogen.
Figure 9. Molecular structure with the atomic-numbering scheme of
Co-DMT.
Photoinduced hydrogen evolution was carried out in a 40 mL flask.
Various amounts of the catalyst, Fl, and NEt3 in 1:1 EtOH/H2O were
added to obtain a total volume of 5.0 mL. The flask was sealed with a
septum and degassed by bubbling nitrogen for 15 min under
atmospheric pressure at room temperature. The pH of this solution
was adjusted to a specific pH by adding H2SO4 or NaOH and was
measured with a pH meter. The samples were subsequently irradiated
by a 500 W xenon lamp. The reaction temperature was maintained at
293 K using a water filter to absorb heat. The generated photoproduct
of H2 was characterized using a GC 7890T gas chromatograph
equipped with a 5 Å molecular sieve column (0.6 m × 3 mm) and a
thermal conductivity detector; nitrogen was used as the carrier gas.
The amount of hydrogen generated was determined by the external
standard method. Hydrogen was not measured in the resulting
solution, and the slight effect of the generated hydrogen gas on the
pressure of the flask was neglected for the calculation of the volume of
hydrogen gas.
4.2. Syntheses and Characterizations. Preparation of H3TFT.
2,2′,2″-((Nitrilotris(benzene-4,1-diyl))tris(methanylylidene))tris-
(hydrazinecarbothioamide): Five drops of acetic acid was added to a
mixture of tris(4-formylphenyl) amine (0.33 g, 1.0 mmol) and
thiosemicarbazine (0.27 g, 3.0 mmol) in a methanol solution. After the
mixture was refluxed for 6 h, the yellow precipitate formed was
collected by filtration, washed with methanol, and dried in vacuum.
Yield: 85%. Anal. calcd for C24H24N10S3·H2O: H, 4.41; C, 52.53; N,
25.53. Found: H, 4.60; C, 52.10; N, 25.05. 1H NMR (DMSO-d6, ppm)
δ 11.38 (s, 3H, NH), 8.15 (w, 3H, −NH2), 8.01(s, 3H, CHN), 7.91
The irradiation of a solution containing 0.2 mM Fl and Co-
DMT (40.0 μM) in a solvent mixture containing Et3N (10% in
volume) and EtOH/H2O (1:1 in volume) resulted in hydrogen
generation at pH = 11.6. Approximately 2.75 mL of hydrogen
was produced after irradiating for 24 h, and the addition of ATP
(2.0 mM) did not change the volume of hydrogen produced
(Figure 7). In comparison to the 12.3 mL of hydrogen
produced in the Co-TFT/Fl/NEt3 system under the same
reaction conditions (10 μM of Co-TFT, ensuring the same
concentration of cobalt ions), Co-DMT exhibited a consid-
erably lower catalytic activity. The superiority of the Co-TFT/
Fl/NEt3 system over the Co-DMT/Fl/NEt3 system is
attributed to the new reaction pathways of the Co-TFT/Fl/
NEt3 system and to the increased molarity of the reaction
within the confined space. The luminescence titrations of Fl
(10 μM) quenched approximately 35% of the emission upon
the addition of up to 0.2 mM Co-DMT. The decrease in
fluorescence lifetime (from 4.5 to 4.0 ns) of Fl (0.1 mM) upon
the addition of Co-DMT (0.2 mM) confirmed that the PET
occurred in a normal bimolecular manner. Because the decrease
in fluorescence lifetime of Fl upon the addition of 0.72 M NEt3
(3.6 ns) is quite larger than that of Co-DMT, a reduction
quenching by NEt3 clearly dominates the homogeneous
photolysis rather than oxidative quenching by Co-DMT.
F
DOI: 10.1021/jacs.5b00832
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(w, 3H, −NH2), 7.76(d, 6H, phenyl), 7.06 (d, 6H, phenyl). 13C NMR
(DMSO-d6, ppm) δ 178.5, 148.6, 146.5, 127.9, 124.8.
distances and allowed to ride on the parent non-hydrogen atoms.
Several bond distances in two of the DMF molecules were constrained
at ideal values, and the thermal parameters of adjacent atoms in these
molecules were constrained to be similar.
Preparation of Co-TFT. H3TFT (54.8 mg, 0.10 mmol) and
Co(CH3COO)2·4H2O (24.9 mg, 0.10 mmol) were dissolved in DMF
(5 mL). After the solution was refluxed for 4 h, the reaction mixture
was cooled to room temperature. Black crystals of Co-TFT were
obtained by diffusing methanol into the aforementioned DMF
solution. Yield: approximately 48% (based on the crystals dried
under vacuum). Anal. calcd for Co4C96H84N40S12·4C3H7NO: H, 4.16;
For Crystal Data of Co-TCT. The non-hydrogen atoms were refined
anisotropically. The hydrogen atoms within the ligand backbones and
the DMF molecules were fixed geometrically at calculated distances
and allowed to ride on the parent non-hydrogen atoms. Two nitrogen
atoms, the carbon atom, and the sulfur atom of the thiourea moiety
were disordered into two parts with each S. O. F. fixed as 0.25; the
corresponding bond distance in the two disordered parts were
constrained at the same value. Several bond distances in the DMF
molecules were constrained at ideal values, and the thermal parameters
of adjacent atoms in these molecules were constrained to be similar.
For Crystal Data of Co-DMT. The non-hydrogen atoms were
refined anisotropically. The hydrogen atoms within the ligand
backbones and the DMF molecules were fixed geometrically at
calculated distances and allowed to ride on the parent non-hydrogen
atoms. Three carbon atoms on one of the DMF molecules were
disordered into two parts, and their S. O. F. was refined with free
variables; the thermal parameters on adjacent atoms in the disordered
DMF molecules were constrained to be similar.
1
C, 47.85; N, 22.73; Found: H, 4.36; C 47.38; N 22.47. H NMR
(DMSO-d6, ppm) 8.11 (s, 3H, CHN), 7.78 (br, 6H, −NH2), 7.26
(d, 3H, phenyl), 7.17 (d, 3H, phenyl), 6.38 (d, 3H, phenyl), 6.22(d,
3H, phenyl). 13C NMR(DMSO-d6, ppm); δ 163.5, 149.6, 148.5, 128.9,
125.8.
Preparation of H3TCT. 2,2′,2″-(Benzene-1,3,5-triyltris
(methanylylidene))tris(hydrazinecarbothio-amide): Five drops of
acetic acid were added to a mixture of benzene-1,3,5-tricarbaldehyde
(0.16 g, 1.0 mmol) and thiosemicarbazine (0.27 g, 3 mmol) in
methanol solution, and the mixture was refluxed for 4 h. The yellow
precipitate formed was collected by filtration, washed with methanol,
and dried under vacuum. Yield: 90%. Anal. calcd for C12H15N9S3·
2H2O: H, 4.59; C, 34.52.; N, 30.19. Found: H, 4.50; C, 34.64; N,
1
30.30. H NMR (DMSO-d6, ppm) δ 11.63 (s, 3H, NH), 8.30 (s, 3H,
CHN), 8.22 (s, 3H, phenyl), 8.08 (br, 6H, −NH2). 13C NMR
ASSOCIATED CONTENT
* Supporting Information
Crystal data (CIF file), experimental details, tables, and figures.
This material is available free of charge via the Internet at
http://pubs.acs.org.
■
(DMSO-d6, ppm) δ: 178.6, 148.8, 134.1, 131.4.
S
Preparation of Co-TCT. H3TCT (38.1 mg, 0.10 mmol) and
Co(CH3COO)2·4H2O (24.9 mg, 0.10 mmol) were dissolved in DMF
(10 mL). After the solution was refluxed for 4 h, the reaction mixture
was cooled to room temperature. Black crystals of Co-TCT were
obtained by diffusing methanol into the aforementioned DMF
solution. Yield: approximately 42%. Anal. calcd for Co4C48H48N36S12:
AUTHOR INFORMATION
Corresponding Author
*E-mail: cyduan@dlut.edu.cn
■
1
H, 2.77; C, 32.95; N, 28.82; Found: H, 2.84; C 33.16; N 28.62. H
NMR (DMSO-d6, ppm) 8.03(s, 3H, CHN), 7.87 (br, 6H, −NH2),
7.22(br, 3H, phenyl). ESI-Ms, M+: 1748.89. 13C NMR (DMSO-d6,
ppm); δ 163.6, 147.6, 133.2, 130.6.
Notes
The authors declare no competing financial interest.
Preparation of Co-DMF. 2-(4(Dimethylamino)benzylidene)-
hydrazinecarbothio-amide (HDMT, 22.2 mg, 0.1 mmol) and Co-
(CH3COO)2·4H2O (9.0 mg, 0.35 mmol) were dissolved in DMF (10
mL) and stirred overnight at room temperature. Black crystals of Co-
DMT were obtained by diffusing ether into the DMF solution. Yield:
approximately 66%. Anal. calcd for CoC30H39N12S3: H, 5.44; C, 49.85;
N, 23.25; Found: H, 5.90; C 49.62; N 22.88. ESI-MS, M+:723.19.
4.3. Crystallography. The intensities were collected on a Bruker
SMART APEX CCD diffractometer equipped with a graphite-
monochromated Mo−Kα (λ = 0.71073 Å) radiation source; the
data were acquired using the SMART and SAINT programs.63,64 The
structures were solved by direct methods and refined on F2 by full-
matrix least-squares methods using the SHELXTL version 5.1
software.65
ACKNOWLEDGMENTS
■
This work was supported by the National Natural Science
Foundation of China (21421005 and 21231003) and the Basic
Research Program of China (2013CB733700) and the Program
for Changjiang Scholars and Innovative Research Team in
University (IRT1213).
REFERENCES
■
(1) Kirby, A. J.; Hollfelder, F. From Enzyme Models to Model Enzymes;
RSC Publishing: Cambridge, 2002.
(2) Lehn, J. M. Science 2002, 295, 2400−2403.
(3) Pauling, L. Nature 1948, 161, 707−709.
(4) Tabushi, I. Acc. Chem. Res. 1982, 15, 66−72.
(5) Yoshizawa, M.; Tamura, M.; Fujita, M. Science 2006, 312, 251−
254.
Crystal Data for Co-TFT. Co4C134H179N52O15.50S12 Mr = 3386.73,
rhombohedral, space group R-3c, dark-red block, a = 24.741(1) Å, c =
109.349(6) Å, V = 57967(5) Å3, Z = 12, Dc = 1.164 g cm−3, μ(Mo−
Kα) = 0.530 mm−1, T = 200(2)K. 11,347 unique reflections [Rint
=
(6) Oluth, M. D.; Bergman, R. G.; Raymond, K. N. Science 2007, 316,
0.0846]. Final R1 [with I >2σ(I)] = 0.0735, wR2 (all data) = 0.2339.
CCDC no. 973241.
85−88.
(7) Yoshizawa, M.; Klosterman, J. K.; Fijuta, M. Angew. Chem., Int.
Ed. 2009, 48, 3408−3438.
Crystal Data for Co-TCT. Co4C40H50N40O8S12 Mr = 2079.84, cubic,
space group I-43m, dark-red block, a = 20.884(1) Å, V = 9108.9 (6)
Å3, Z = 2, μ(Mo−Kα) = 0.531 mm−1, T = 200(2) K. 1439 unique
reflections [Rint = 0.1088]. Final R1 [with I > 2σ(I)] = 0.0670, wR2 (all
data) = 0.1615. CCDC no. 1000791.
Crystal Data for Co-DMT. CoC42H67N16O4S3, Mr = 1015.23,
monoclinic, space group P21/c, dark-red block, a = 9.712(1),b =
24.697(2), c = 21.697(2) Å, V = 5195.9(7) Å3, Z = 4, Dc = 1.299 g
cm−3, μ(Mo−Kα) = 0.506 mm−1, T = 296(2) K. 8094 unique
reflections [Rint = 0.0939]. Final R1 [with I > 2σ(I)] = 0.0787, wR2 (all
data) = 0.2005. CCDC no. 1000792.
(8) Loner, A. L.; Marquez, C.; Dickman, M. H.; Nau, W. M. Agnew.
Chem. Int. Ed. 2011, 50, 545−548.
(9) Lee, J. W.; Samal, S.; Selvapalam, N.; Kim, H. J.; Kim, K. Acc.
Chem. Res. 2003, 36, 621−630.
(10) Raynal, M.; Ballester, P.; Vidal-Ferran, A.; Leeuwen, P. W. N. M.
Chem. Soc. Rev. 2014, 43, 1660−1733.
(11) Raynal, M.; Ballester, P.; Vidal-Ferran, A.; Leeuwen, P. W. N. M.
Chem. Soc. Rev. 2014, 43, 1734−1787.
(12) Mahata, K.; Frischmann, P. D.; Wurthner, F. J. Am. Chem. Soc.
̈
2013, 35, 15656−15661.
For Crystal Data of Co-TFT. Except for methanol (solvent) and
water molecules, the other non-hydrogen atoms were refined
anisotropically. The hydrogen atoms within the ligand backbones
and the DMF molecules were fixed geometrically at calculated
(13) Furutani, Y.; Kandori, H.; Kawano, M.; Nakabayashi, K.;
Yoshizawa, M.; Fujita, M. J. Am. Chem. Soc. 2009, 131, 4764−4768.
(14) Gray, H. B.; Maverick, A. W. Science 1981, 214, 1201−1205.
G
DOI: 10.1021/jacs.5b00832
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(15) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U.S.A. 2006,
103, 15729−15735.
(51) Artero, V.; Chavarot-Kerlidou, M.; Fontecave, M. Angew. Chem.,
Int. Ed. 2011, 50, 7238−7266.
(52) Jiang, P. J.; Guo, Z. J. Coord. Chem. Rev. 2004, 248, 205−229.
(53) Connors, K. A. Binding Constants; John Wiley: New York, 1987.
(54) Nelissen, H. F. M.; Kercher, M.; De Cola, L.; Feiters, M. C.;
Nolte, R. J. M. Chem.Eur. J. 2002, 8, 5407−5414.
(55) Zhang, P.; Wang, P.; Na, Y.; Li, X. Q.; Jiang, Y.; Sun, L. C.
Dalton Trans. 2010, 39, 1204−1206.
(16) Esswein, A. J.; Nocera, D. G. Chem. Rev. 2007, 107, 4022−4047.
(17) Wang, J.; He, C.; Wu, P. Y.; Wang, J.; Duan, C. Y. J. Am. Chem.
Soc. 2011, 133, 12402−12405.
(18) Porel, M.; Chuang, C. H.; Burda, C.; Ramamurthy, V. J. Am.
Chem. Soc. 2012, 134, 14718−14721.
(19) Kitamoto, K.; Sakai, K. Angew. Chem., Int. Ed. 2014, 53, 4618−
4622.
(56) Yamade, Y.; Miyahigashi, T.; Kotani, H.; Ohkubo, K.; Fukuzumi,
S. J. Am. Chem. Soc. 2011, 133, 16136−16145.
(20) Frischmann, P. D.; Mahata, K.; Wurthner, F. Chem. Soc. Rev.
̈
(57) McCormick, T. M.; Calitree, B. D.; Orchard, A.; Kraut, N. D.;
Bright, F. V.; Detty, M. R.; Eisenberg, R. J. Am. Chem. Soc. 2010, 132,
15480−15483.
2013, 42, 1847−1870.
(21) Chakrabarty, R.; Mukherjee, P. S.; Peter J. Stang, P. J. Chem. Rev.
2011, 111, 6810−6918.
(58) He, C.; Wang, J.; Zhao, L.; Liu, T.; Zhang, J.; Duan, C. Y. Chem.
Commun. 2013, 49, 627−629.
(22) Leininger, S.; Olenyuk, B.; Stang, P. J. Chem. Rev. 2000, 100,
853−907.
(59) McKee, T.; McKee, J. R. Biochemistry: The Molecular Basis of
Life, 3rd ed.; McGraw-Hill: New York, 2003.
(23) Seidel, S. R.; Stang, P. J. Acc. Chem. Res. 2002, 35, 972−983.
(24) Stang, P. J.; Olenyuk, B. Acc. Chem. Res. 1997, 30, 502−518.
(25) Debata, N. B.; Tripathy, D.; Chand, D. K. Coord. Chem. Rev.
2012, 256, 1831−1945.
(60) Liu, Y.; Lin, Z. H.; He, C.; Zhao, L.; Duan, C. Y. Dalton Trans.
2010, 39, 11122−11125.
(61) Han, Z.; McNamara, W. R.; Eum, M. S.; Holland, P. L.;
Eisenberg, R. Angew. Chem., Int. Ed. 2012, 51, 1667−1670.
(62) Sampath, N.; Mathews, R.; Ponnuswamy, M. N.; Kang, L. Mol.
Cryst. Liq. Cryst. 2010, 518, 151−159.
(63) SMART Data collection software, version 5.629; Bruker AXS Inc.:
Madison, WI, 2003.
(64) SAINT, Data reduction software, version 6.45 Bruker AXS Inc.:
Madison, WI, 2003.
(65) Sheldrick, G. M. SHELX-97: Program for crystal structure
(26) Smulders, M. M. J.; Riddell, I. A.; Browne, C.; Nitschke, J. R.
Chem. Soc. Rev. 2013, 42, 1728−1754.
(27) Wiester, M. J.; Ulmann, P. A.; Mirkin, C. A. Angew. Chem., Int.
Ed. 2011, 50, 114−137.
(28) Meeuwissen, J.; Reek, J. N. H. Nat. Chem. 2010, 2, 615−621.
(29) Wang, Z. J.; Clary, K. N.; Bergman, R. G.; Raymond, K. N.;
Toste, F. D. Nat. Chem. 2013, 5, 100−103.
(30) He, Q. T.; Li, X. P.; Lu, Y.; Yu, Z. Q.; Wang, W.; Su, C. Y.
Angew. Chem., Int. Ed. 2009, 48, 6156−6159.
analysis; University of Gottingen: Gottingen, Germany, 1997.
̈
̈
̀
(31) Vriezema, D. M.; Aragones, M. C.; Elemans, J. A. A. W.;
Cornelissen, J. J. L. M.; Rowan, A. E.; Nolte, R. J. M. Chem. Rev. 2005,
105, 1445−1489.
(32) Wang, Z. J.; Clary, K. N.; Bergman, R. G.; Raymond, K. N.;
Toste, F. D. Nat. Chem. 2013, 5, 100−103.
(33) Inokuma, Y.; Kawano, M.; Fujita, M. Nat. Chem. 2011, 3, 349−
358.
(34) Bolliger, J. L.; Belenguer, A. M.; Nitschke, J. R. Angew. Chem.,
Int. Ed. 2013, 52, 7958−7962.
(35) Zarkadoulas, A.; Koutsouri, E.; Mitsopoulou, C. A. Coord. Chem.
Rev. 2012, 256, 2424−2434.
(36) McNamara, R. W.; Han, Z. J.; Yin, C. J.; Brennessel, W. W.;
Holland, P. L.; Eisenberg, R. Proc. Natl. Acad. Sci. U.S.A. 2012, 109,
15595−15599.
(37) Konigstein, C. A. J. Photochem. Photobiol. 1995, 90, 141−152.
(38) Na, Y.; Pan, J. X.; Wang, M.; Sun, L. C. Inorg. Chem. 2007, 46,
3813−3815.
(39) Paterson, B. M.; White, J. M.; Donnelly, P. S. Dalton Trans.
2010, 39, 2831−2837.
(40) Vinuelas-Zahínos, E.; Luna-Giles, F.; Torres-García, P.;
̃
́ ́
Fernandez- Calderon, M. C. Eur. J. Med. Chem. 2011, 46, 150−159.
(41) Zhao, Y. G.; Guo, D.; Liu, Y.; He, C.; Duan, C. Y. Chem.
Commun. 2008, 5725−5727.
(42) Li, M. X.; Cai, P.; Duan, C. Y.; Lu, F.; Xie, J.; Meng, Q. J. Inorg.
Chem. 2004, 43, 5174−5176.
(43) Zhang, T.; Liu, G. F.; Liu, L.; Jia, D. Z.; Zhang, L. Chem. Phys.
Lett. 2006, 427, 443−448.
(44) Chang, T. M.; Tomat, E. Dalton Trans. 2013, 42, 7846−7849.
(45) Lazarides, T.; McCormick, T. M.; Du, P. W.; Luo, G. G.;
Lindley, B.; Eisenberg, R. J. Am. Chem. Soc. 2009, 131, 9192−9194.
(46) Zhang, P.; Wang, M.; Dong, J. F.; Li, X. W.; Wang, F.; Wu, L. Z.;
Sun, L. C. J. Phys. Chem. C 2010, 114, 15868−15874.
(47) Dong, J. F.; Wang, M.; Zhang, P.; Yang, S. Q.; Liu, J. Y.; Li, X.;
Sun, L. C. J. Phys. Chem. C 2011, 115, 15089−15096.
(48) McCormick, T. M.; Han, Z. J.; Weinberg, D. J.; Brennessel, W.
W.; Holland, P. L.; Eisenberg, R. Inorg. Chem. 2011, 50, 10660−10666.
(49) Biros, S. M.; Yeh, R. M.; Raymond, K. N. Angew. Chem., Int. Ed.
2008, 47, 6062−6064.
(50) Baffert, C.; Artero, V.; Fontecave, M. Inorg. Chem. 2007, 46,
1817−1824.
H
DOI: 10.1021/jacs.5b00832
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX