3d/4f Coordination Clusters as Cooperative Catalysts for Highly Diastereoselective Michael Addition Reactions

Article abstract of DOI:10.1021/acs.inorgchem.7b01011

Michael addition (MA) is one of the most well studied chemical transformation in synthetic chemistry. Here, we report the synthesis and crystal structures of a library of 3d/4f coordination clusters (CCs) formulated as [Zn^II₂Y^II

Full text of DOI:10.1021/acs.inorgchem.7b01011

Article
pubs.acs.org/IC
3d/4f Coordination Clusters as Cooperative Catalysts for Highly
Diastereoselective Michael Addition Reactions
Kieran Griffiths,† Athanassios C. Tsipis,‡ Prashant Kumar,† Oliver P. E. Townrow,† Alaa Abdul-Sada,†
Geoffrey R. Akien,§ Amgalanbaatar Baldansuren,⊥ Alan C. Spivey,∥ and George E. Kostakis*,†
†Department of Chemistry, School of Life Sciences, University of Sussex, Brighton BN1 9QJ, U.K.
‡Laboratory of Inorganic and General Chemistry, Department of Chemistry, University of Ioannina, 451 10 Ioannina, Greece
§Department of Chemistry, Lancaster University, Lancaster LA1 4YB, U.K.
⊥School of Chemistry, The University of Manchester, Manchester M13 9PL, U.K.
∥Department of Chemistry, Imperial College London, South Kensington Campus, London SW7 2AZ, U.K.
S
* Supporting Information
ABSTRACT: Michael addition (MA) is one of the most well
studied chemical transformation in synthetic chemistry. Here,
we report the synthesis and crystal structures of a library of
3d/4f coordination clusters (CCs) formulated as [ZnII YIII2L4-
2
(solv)X(Z)Y] and study their catalytic properties toward the
MA of nitrostyrenes with barbituric acid derivatives. Each CC
presents two borderline hard/soft Lewis acidic ZnII centers
and two hard Lewis acidic YIII centers in a defect dicubane
topology that brings the two different metals into a proximity
of ∼3.3 Å. Density functional theory computational studies
suggest that these tetrametallic CCs dissociate in solution to
give two catalytically active dimers, each containing one 3d and
one 4f metal that act cooperatively. The mechanism of catalysis
has been corroborated via NMR, electron paramagnetic resonance, and UV−vis. The present work demonstrates for the first
time the successful use of 3d/4f CCs as efficient and high diastereoselective catalysts in MA reactions.
(Figure 1).21−25 The crystallographic characterization of seven of
these CCs containing different metal combinations (i.e., 1ZnY,
INTRODUCTION
■
Nature provides many highly efficient but complex catalytic
systems that rely on synergistic, cooperative activation through
multiple metal centers. Examples include tyrosinase1 and
photosystem II.2 To emulate this behavior is hugely challenging,
and it is only in the past decade that polynuclear coordination
clusters (CCs),3−5 especially ones containing two metals,6−11
have emerged as viable platforms to achieve this. Bimetallic
cooperative catalysts have been described comprising different
types of metals encompassing alkali, transition, and lantha-
Figure 1. A cartoon representation of a 3d/4f CC. Red balls indicate
nides.12−17 Unsurprisingly, the precise topology of the cluster
O atoms.
and in particular the distance separating the two metals (typically
3.5−6 Å) have been suggested to be a key parameter for efficient
catalysis.8 Only a few 3d/4f bimetallic catalysts have been
1ZnSm, 1ZnEu, 1ZnGd, 1ZnDy, 1ZnTb, and 1ZnYb) has been
reported previously23 and reveals that in all cases the 3d and 4f
reported, but some of these have been shown to display high
diastereo- and enantioselectivity in nitro-Mannich14,17 and Henry12
metals are close together (3.3 0.1 Å).
reactions, and excellent efficiency in oxidation of alcohols18 and
In addition to developing a facile method for the synthesis of
these 3d/4f CCs, and structurally characterizing them, we have
water.19
previously demonstrated their stability in solution (by electro-
We have previously reported the synthesis of a small library
spray ionization-mass spectrometry (ESI-MS), electron para-
of heterometallic 3d/4f CCs possessing a rigid defect dicubane
topology, with the general formula [MII LnIII2(L1)4(solv)X(Z)Y]
magnetic resonance (EPR), and NMR and their catalytic
2
(1MLn), where H2L1, is (E)-2-(2-hydroxy-3-methoxybenzyli-
dene-amino)phenol, M is Co/Ni/Zn, Ln is Y20/Sm/Eu/Gd/
Dy/Tb/Yb, solv is EtOH/CH3CN/DMF, and Z is Cl/NO3/ClO4
Received: April 24, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
competence at room temperature and low catalyst loading
(1−2.5 mol %) in 2-furaldehyde/sec-amine domino reactions,21,22
in Friedel−Crafts type indole/aldehyde and indole/nitrostyrene
alkylation reactions,23,24 and in Petasis-Manich reactions.25
By maintaining the hard 4f center invariant and tuning the 3d center
(Co, Ni, Cu, or Zn), we have shown that valuable mechanistic
information can be extracted by correlating product distributions
and selectivity versus metal properties.26 Such correlations are not
possible using homometallic polynuclear 3d/3d or 4f/4f catalysts.
Michael additions (MAs) are some of the most well studied
chemical transformations in synthetic chemistry, and various
catalytic protocols displaying high levels of diastereo- and
enantioselectivity have been developed, including those
promoted by organocatalysts, simple salts, and complexes.27−30
Inspired by Shibasaki’s pioneering studies into nitro-Mannich
reactions catalyzed by a presumed dinuclear Cu/Sm CC formed
in situ,14,17 in which cooperative binding of the substrates to both
metal centers was proposed, we envisaged that our library of
catalysts might provide efficient catalysts for MA reactions. To this
end, we opted to investigate the MA reactions of nitrostyrenes and
1,3-dimethyl barbituric acid as a test system.31−33 To the best of our
knowledge, there is no previous report with the use of bimetallic
3d/4f as catalysts in MA reactions. Modification of the ligand and
the metal centers as well as in situ studies of the catalytic reaction
with UV Vis, NMR, and EPR spectroscopy have been carried out,
and the mechanistic insights provided by these methods were
augmented by density functional theory computation.
Crystal Structures. For nine of the CCs (2ZnY, 4ZnY,
7ZnY, 8ZnY, 10ZnY, 14ZnY, 15ZnY, 16ZnY, and 19ZnY),
crystals suitable for single crystal X-ray diffraction were obtained
by the vapor diffusion method using DMF and Et2O. The crystal
structures of all nine clusters are shown below (Figure 2). We use
the term “isoskeletal”34 to describe all these catalysts as they all
possess a defect dicubane topology. The 3d metal is in the center
and the 4f in the wings, as shown in the cartoon representation
(Figure 1). In accord with our previous findings, the 3d and 4f
metals are separated by 3.3 0.1 Å. Five out of six and six out of
eight coordination sites of the 3d and 4f centers respectively are
occupied by the ligand.
Thermal Studies. Thermogravimetric analysis for compounds
2ZnY−21ZnY (see Supporting Information, Figures S26−S45)
display similar behavior. As the temperature in increased, the
mass of the CCs decreases slowly up to 280−320 °C. This
decrease is between 6 and 13% depending on overall molecule
weight and corresponds to the loss of the two solvent DMF
molecules. Once this temperature is exceeded, the mass rapidly
decreases up to 400 °C, which is indicative of decomposition.
This decrease becomes less significant and plateaus after 600 °C,
and with all examples the remaining mass is constant by 800 °C.
The residual % mass is between 19 and 25% depending on initial
molecular mass and corresponds to the remaining metal oxides
(ZnO/Y2O3, 2:1).
Catalytic Experiments. Our initial benchmarking studies
leading up to the synthesis of the above library of ZnY CCs,
examined the reaction of 1,3-dimethyl barbituric acid with trans-
β-nitrostyrene in ethanol. This reaction proceeds only slowly
in the absence of a catalyst (i.e., 35% conversion in 1 h). Zn salts
alone do not promote the transformation, but some lanthanide
salts give moderate yields (see Supporting Information, Table S3).
Using a range of our previously reported 3d/4f CCs as catalysts
at a 2.5 mol % loading level, 1ZnSm, 1ZnEu, 1ZnTb, 1ZnGd,
and 1ZnDy were found to give moderate performance, but the
use of 1ZnY and 1ZnYb provided the MA product in quantitative
yield within 15 min at room temperature (see Supporting
Information, Table S4). Taking into account the relative prices of
Y and Yb, we selected to evaluate the performance of 1ZnY
further. After screening several solvents, we identified that the
ethanol/water (6:4) solvent system was optimal in terms of rate
and is also advantageous from an environmental impact35
perspective (Table S5). Then, we confirmed that reaction at room
temperature rather than at −5 °C or reflux gave the highest yields
(Table S6) and that loadings of 2.5 mol % rather than 5 or 10 mol %
gave highest yields (Table S7). Using these optimized conditions,
we then explored the scope of the reaction using
an array of substituted nitrostyrenes (Table 2). Compounds I−V
and VII have been reported previously.31 We were now in a position
to evaluate the performance of our library of CCs containing
the different ligands: 2ZnY−21ZnY (Table S8). The results
indicated the superiority of catalyst 17ZnY which was found to
give excellent yields even when using a loading level of just
0.5 mol % (Tables S9 and S10). The ligand in this CC contains
the allyl and bis-(4-tert-butyl-2-aminophenol) substructure
which we hypothesize may provide improved solubility relative
to the other ligands.
RESULTS AND DISCUSSION
■
Catalyst Synthesis. Using our already reported ZnY CC
1ZnY as a starting point, a library of 20 analogous ZnY CCs was
prepared using substituted (E)-2-(2-hydroxy-3-methoxybenzyli-
dene-amino)phenol ligands (H2L1-H2L21, Table 1). These
Table 1. Synthesis and Structures of Ligands H2L1−H2L21
Used to Prepare CCs 1ZnY−21ZnY
ligands wereeasily synthesized in one step by condensation of three
commercially available 5-substituted-3-methoxysalicylaldehydes
(5-H, 5-allyl, and 5-Br) with seven 2-aminophenol derivatives,
under ambient conditions and in high yields (Table 1).
The selected ligands were expected to offer very similar
coordination environments, such that the same core CC motif
would assemble and we would be able to identify the influence of
the second coordination sphere on their properties as catalysis.
The solid CCs were synthesized in one-step reactions using the
ligand H2L, Zn(NO3)2·6H2O, Y(NO3)3·x(H2O), and Et3N
under reflux for 2 h in excellent yields (>90%). In all cases,
In general, all these reactions give the products in very good to
quantitative yields. The structures of compounds VI, VIII, and
XII were confirmed by single crystal X-ray structure determi-
nation (see Supporting Information, Figure S1).
compounds formulated as [ZnII YIII2L4(NO3)2(DMF)2] solvent
2
were obtained, and they were characterized by thermogravimetric
analysis (TGA) and CHN analysis.
Mechanistic Studies. In order to gain information about the
possible mechanism of this reaction and to delineate the scope
B
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 2. Crystal structures of 2ZnY, 4ZnY, 7ZnY, 8ZnY, 10ZnY, 14ZnY, 15ZnY, 16ZnY, 19ZnY, and 1CuY. Color code: Zn, green; Y, mauve;
Cu, light blue; O, red; N, blue; C, gray; Cl, light green; Br, brown.
and limitations of the 17ZnY catalyst, we performed the following
set of reactions.
Influence of Bases. Two reactions were performed in the
presence of base (Et3N or NaOH); however, a significant
lowering of the catalytic conversion was observed. This outcome
can likely be attributed to the instability of the CC in the presence
of base which can effect conversion to other CCs (Table S11).
Tuning the 3d Metal. To further understand the role of the
metals in 17ZnY, we performed parallel reactions using
the isoskeletal compounds 1NiDy,36,37 1CoDy,36,37 and 1CuY
C
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
in the presence of 1ZnY was investigated by UV−vis spectroscopy in
a water/ethanol solution (see Supporting Information, Figure S2).
The trans-β-nitrostyrene exhibits a strong absorption at 320 nm.
The absorption was recorded over 70 min with 5 min intervals
between measurements. It was observed that the intensity of
the peak at 320 nm gradually decreased and the intensity of the
product (347 nm) gradually increased. The reaction proceeds
much more slowly in the absence of catalyst, demonstrating the
need for catalyst (see Supporting Information, Figure S3). trans-
β-Nitrostyrene and dimethylbarbituric acid were also left for
60mintoreact before2.5% mmolof 1ZnY wasadded; the reaction
proceeded faster in the presence of the catalyst (see Supporting
Information, Figure S4).
Table 2. Scope of Reactions of 1,3-Dimethyl Barbituric Acid
and Derivatives with trans-β-Nitrostyrenes Catalyzed by 1ZnY
EPR Studies. To gain further mechanistic insights, the following
Q-band EPR studies of 1ZnGd in solution at 15K were performed.
trans-β-Nitrostyrene or 1,3-dimethyl barbituric acid, dissolved in
EtOH, was added to ethanolic solutions of 1ZnGd (1:1, 20:1,
and 40:1 ratios). Additionally, a mixture of trans-β-nitrostyrene
and 1,3-dimethyl barbituric acid, dissolved in EtOH (1:1 ratio),
was added to ethanolic solutions of 1ZnGd (1:1, 20:1, and 40:1
ratios) (see Supporting Information, Figure S5). The experiments
in ratios of 20:1 and 40:1 were executed to mimic the catalytic
conversion (100:5 and 100:2.5) conditions and identify the
stability of the bimetallic system.
Numerical spectrum simulations of the EPR spectra of the
3d/4f CCs were performed. The conventional forms of spin
̂
Hamiltonians, i.e., Hspin = S · D · S + βeB0 · g · S, were considered
where the first term denotes the fine-structure splitting (zero-
field splitting - ZFS), and the second term denotes the electron
Zeeman interaction. For the S-state 4f7 (Gd3+) ions, they usually
exhibit an orbital singlet ground state, and the degeneracy 2J + 1
of each S multiplet can be partially or wholly lifted. Regarding the
complex nature of the observed experimental spectra, the spin−
orbit coupling (SOC) must have a significant orbital contribution
to the low-lying spin ground state. There is likely a competing
effect between the ZFS energy of S = 7/2 state and the crystal-
field energy of Gd3+ (4f7), including Zn2+ (3d10), even though the
latter is diamagnetic. The EasySpin computational simulation
package39 was used to allow a reasonable parameter set forD-values
and effective g′-factors to be obtained.
(see Supporting Information). The reactions of 1CuY and
1ZnDy yielded the desired product in very good yields (90%) while
with 1NiDy and 1CoDy had poor performance (Table 3). These
Table 3. Comparison of Efficacy for Compounds 1ZnY,
1ZnDy, 1NiDy, 1CoDy, and 1CuDy
entry
loading/%
catalyst
yield/%
time/min
Only the permissible transitions (ΔmS = 1) between the
electron spin mS = 1/2 and mS = 3/2 sublevels of Gd3+
were expected to be observed at X-band microwave frequency
whose excitation energy is obviously not enough to resolve the
transitions between high-spin multiplets. All the EPR spectra
were broadened significantly at 340 mT, while there was a sharp
transition at 1200 mT. Since, in orientation disordered powder-
type samples (including frozen solutions), paramagnetic species
are randomly oriented with respect to the applied magnetic field
B0, and their EPR spectra appear as the sum of the resonances
of all those orientations. In such conditions, the powder-type
spectra provide only limited information on the orientation of
the g-tensor principal axes in the molecular coordinate system
(see Supporting Information, Figure S6).
We were able to assign the electronic splitting to the partially
resolved transitions at 340 and 1200 mT. The corresponding
energy levels were obviously not well-separated, indicating that
mS could not become the main quantum numbers. Thus,
fictitious spin states are likely responsible for the observed
splitting separated at g′|| and g′⊥ regions. The numerically
simulated spectra produced rather good agreements with the
experimental ones (see Supporting Information, Figure S7), and
their corresponding parameters are given in Table 4. These are
1
2
3
4
5
2.5
2.5
2.5
2.5
2.5
1ZnY
quantitative
15
15
15
15
15
1ZnDy
1NiDy
1CoDy
1CuY
94
29
34
85
differences can be rationalized taking into account the Irving-
Williams38 stability series, in which for a given ligand the stability
of dipositive metal ion complexes increases in the order: CoII <
NiII < CuII < ZnII. It would therefore appear that binding
(i.e., activation) of at least one of the substrates via an O atom to a
3d metal center is required for efficient catalysis.
NMR Studies. Careful titration of DMF solutions of 1ZnY
with 0.25−10 equiv of trans-β-nitrostyrene or of 6-amino-1,3-
dimethyluracil led to no significant observable changes in the
1H NMR spectra. On the other hand, analogous titrations with
N,N′-diethyl-2-thiobarbituric acid led to significant differences:
both the N,N′-diethyl-2-thiobarbituric acid and 1ZnY peaks were
affected. Evidently binding of the individual substrates to the
1ZnY is not strong, but this presents no impediment to the
catalytic activity of the complex.
UV Vis Studies. The reaction of trans-β-nitrostyrene
(1 × 10−5 mmol) and dimethylbarbituric acid (1 × 10−5 mmol)
D
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 4. Spin Hamiltonian Parameters Obtained from
EasySpin Numerical Simulations
Table 5. Comparison of Efficacy for Compounds 1ZnY,
1ZnDy, 1NiDy, 1CoDy, and 1CuDy
effective g-factors
|D|
[MHz]
HWHM
[mT]
samples
g′1
g′2
g′3
1ZnGd with BA (1:1)
1ZnGd with NS (1:1)
1.961 0.563 0.563 158(30)
1.961 0.563 0.560 128(20)
36
32
46
46
33
46
46
40
46
1ZnGd with BA and NS (1:1) 1.951 0.570 0.560 108(10)
1ZnGd with BA (1:20)
1ZnGd with NS (1:20)
1.951 0.570 0.560 108(10)
1.961 0.572 0.552 96(10)
1ZnGd with BA and NS (1:20) 1.951 0.570 0.560 108(10)
1ZnGd with BA (1:40)
1ZnGd with NS (1:40)
1.951 0.570 0.560 108(10)
1.961 0.570 0.560 96(10)
1ZnGd with BA and NS (1:40) 1.951 0.570 0.560 108(10)
likely not the only parameter sets, and the further refinements
can probably be achieved. Importantly, the slight differences in
g-factors are taken as evidence in support of the Gd coordination
environment changes, particularly in the g′⊥ region. We believe
that the coordination number remains intact since the overall
molecular symmetry is unambiguously reflected on the experi-
mental g-tensor. On the other hand, the existence of the large
ZFS (hν ≪ D) can be tested using a series of high-order electron
spin operators in the spin Hamiltonians, including the extended
Stevens operators.40 We attempted such simulations to obtain
a plausible parameter set, but the simulated spectra did not
reproduce reasonable matches with the experiments (spectra not
shown). Only transitions within the ground Kramers doublet
of the J = 7/2 ground term of Gd3+ (8S7/2) are observed as
a fictitious (or effective) spin S′ = 1/2 system at X-band, with the
ground state g-factors of g′|| ≈ 1.961 and g′⊥ ≈ 0.561, regarding
the unknown rhombicity of ZFS and the ratio between SOC and
crystal-field energy.
a
dr = diastereoisomeric ratio.
1ZnEu, 1ZnGd, 1ZnDy, 1ZnTb, 1ZnYb, 1CoY, 1NiY, and
1CuY as catalysts gave drs and yields similar to those with trans-β-
methyl-nitrostyrene (entries 2−10).
Computational Studies. To gain insight into a plausible
mechanism of the MA of 1,3-dimethyl barbituric acid to trans-
β-nitrostyrenes catalyzed by 3d/4f CCs, we performed a series of
calculations on possible reaction pathways employing density
functional theory (DFT) methods and monitoring the natural
atomic charge distribution and the nature of the frontier molecular
orbitals (FMOs) on both the catalyst and the substrates.
Computational details with the relevant citations are given in the
Supporting Information. We selected to examine 1,3-dimethyl
barbituric acid and 6-amino-1,3-dimethyl barbituric acid with
(Z)-(2-nitrovinyl)benzene, (Z)-(2-nitroprop-1-enyl)benzene,
and (Z)-1-methoxy-4-(2-nitroprop-1-enyl)benzene. To avoid
any confusion when discussing the mechanistic details of the MA
reaction catalyzed by the ZnY bimetallic catalyst, the chemical
structures of the selected molecules with a labeling scheme
are shown in Figure 3. First, we calculated the thermodynamics
of the noncatalyzed MA reactions in aqueous solution employing
the PBE0/6-311++G(d,p) computational protocol to obtain
natural atomic charges calculated by natural bond orbital (NBO)
population analysis. The relevant FMOs are shown in Figure S7.
It is well established that the α-carbon of barbituric acid has a
reactive hydrogen atom and is quite acidic (pKa = 4.01) because
of the aromatic stabilization of the carbanion. DFT calculations at
the PBE0/6-311++G(d,p) level of theory revealed that the keto
form of 1,3-dimethyl barbituric acid in aqueous solution can
coexist in equilibrium with the diketo−enol form, which is found
at 8.7 kcal/mol higher in energy than the triketo form (Figure 3).
Notice that ab initio and DFT calculations on the tautomers of
barbituric acid showed that the triketo form is found to be the
most stable form in the gas phase and in solution, which is in
agreement with the experimental result.41−45 On the other hand, the
aminodiketo form of 6-amino-1,3-dimethyl barbituric acid is the most
stable form relative to the aminoketo−enol and diketo-ammonium
forms, whicharefoundtobe47.0and63.6kcal/molhigherinenergy.
According to NBO population analysis the keto and enolic
oxygen atoms of the diketo−enolic tautomer acquire higher
negative natural atomic charges by 0.03 up to 0.06 |e| relative to
the corresponding oxygen atoms of the triketo tautomer. On the
other hand, the α-carbon atom in the enolic tautomer acquires
The effects of the hyperfine fields created by 155Gd and 157Gd
(both I = 3/2) nuclei on the alignment of the electron spins are
negligible, and the electron problem can be treated independently.
67Zn (I = 5/2) nucleus has only a contribution to inhomogeneous
line broadenings. In addition, hyperfine interaction of ligands is
overwhelmed and readily masked by such broadening effects.
Test of Diastereoselectivity. When we employed trans-β-methyl-
nitrostyrene instead of trans-β-nitrostyrene in the reaction with
1,3-dimethyl barbituric catalyzed by 1ZnY (2.5 mol %), excellent
yields of product VI were obtained within 15 min (Table 2). The
NMR data showed the presence of one diastereoisomer (>20:1 dr)
(see Supporting Information). A single crystal X-ray structure
determination on this product obtained after filtration of the
reaction in the presence of the catalyst revealed this to have the
(R*,R*) relative configuration. Next, we explored the influence of
the lanthanide on the diastereoselectivity of this reaction (Table 5).
We ran parallel reactions using 1ZnSm, 1ZnEu, 1ZnGd, 1ZnDy,
1ZnTb, and 1ZnYb as catalysts under otherwise identical
conditions (entries 1−7). Only slight differences in diastereose-
lectivity were observed (20:1 for 1ZnY to >20:1 for 1ZnSm) which
might be attributed to the different ionic radii of the Lanthanides.
Subsequently, we investigated 1CoY, 1NiY, and 1CuY as catalysts
and observed that all three CCs gave ∼20:1 drs, but that 1CoY
and 1NiY gave significantly lower yields (entries 8−10). We also
performed the reaction of cis-β-methyl-4-methoxy-nitrostyrene with
1,3-dimethyl barbituric acid with 1ZnY as catalyst to yield product
XXII (Table 4). The NMR data again showed the presence of one
diastereoisomer, which X-ray characterization revealed also to have
the (R*,R*) relative stereochemistry (Figure S2). Use of 1ZnSm,
E
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(44.6% having 2p character) and the adjacent keto-oxygen atom
(13.5% with 2p character). Both the nature of the HOMOs and
the negative charge distribution suggest that the triketo form of the
1,3-dimethyl barbiturato ligand could be coordinated to a metal
center through the keto-oxygen donor atoms, while the diketo−
enolic form through the α-carbon atom. Similarly, the nature of
HOMO of the 6-amino-1,3-dimethyl barbiturate substrate
supports coordination through the α-carbon atom.
The noncatalyzed MA reactions of the triketo tautomer of
1,3-dimethyl barbituric acid to (Z)-(2-nitrovinyl)benzene,
(Z)-(2-nitroprop-1-enyl)benzene, and (Z)-1-methoxy-4-(2-nitro-
prop-1-enyl)benzene are predicted to be exothermic by 11.9, 9.9,
and 10.9 kcal/mol respectively, while the exothermicity of the
addition of the diketo−enol tautomer of 1,3-dimethyl barbituric
acid to (Z)-(2-nitrovinyl)benzene, (Z)-(2-nitroprop-1-enyl)-
benzene, and (Z)-1-methoxy-4-(2-nitroprop-1-enyl)benzene are
20.7, 18.6, and 19.6 kcal/mol respectively at the PBE0/6-311+
+G(d,p) level of theory. The noncatalyzed MA reactions of the
6-amino-1,3-dimethyl barbiturate substrate to (Z)-(2-nitrovinyl)-
benzene are predicted to be also exothermic by 13.2 kcal/mol.
Considering the nature of the lowest unoccupied molecular orbital
(LUMO) of the (Z)-(2-nitrovinyl)benzene, (Z)-(2-nitroprop-1-
enyl)benzene, and (Z)-1-methoxy-4-(2-nitroprop-1-enyl)benzene
being a π* MO mainly localized on the NO2 moiety (51.2%, 57.5%,
and 56.9%) and on the C(7) unsaturated carbon atom (22.8%,
21.4%, and 21.8%), the MA reactions under study are FMO-
controlled supported by HOMO−LUMO interactions.
Figure 3. Natural atomic charges and the relevant FMOs of the triketo
and diketo−enol forms of 1,3-dimethyl barbituric acid, the aminodiketo,
aminoketo−enol, and diketo-ammonium forms of 6-amino 1,3-dimethyl
barbituric acid and the (Z)-(2-nitrovinyl)benzene, (Z)-(2-nitroprop-1-
enyl)benzene, and (Z)-1-methoxy-4-(2-nitroprop-1-enyl)benzene cal-
culated at the PBE0/6-311++G(d,p) level of theory in aqueous
solutions. The C(7) labels on nitro-styrenes are in bold.
An alternative plausible reaction pathway involves the
nucleophilic attack of the C(4) electrophilic center of the
protonated trans-β-nitrostyrenes, namely, (Z)-(2-nitrovinyl)-
benzene, (Z)-(2-nitroprop-1-enyl)benzene, and (Z)-1-methoxy-
4-(2-nitroprop-1-enyl)benzene and by the nucleophilile 1,3-di-
methyl barbiturate carbanion (Figure 4).
The deprotonation of the triketo form of 1,3-dimethyl
barbituric acid to form the 1,3-dimethyl barbiturate carbanion
is predicted to be endothermic by 45.1 kcal/mol. Of note is a
remarkable increase in the negative atomic charge of the keto
oxygen atoms in the carbanion relative to the corresponding keto
oxygen atoms of the triketo tautomer. On the other hand the
negative natural atomic charge on the deprotonated α-carbon
atom of the carbanion decreases by 0.04|e|. Interestingly, the
HOMO of the nucleophile is primarily localized on the
deprotonated α-carbon atom (52.3%) and the two adjacent
keto oxygen atoms (28.9%) being π*-type MOs, while the
less negative natural atomic charge by 0.114 |e| relative to the
α-carbon atom of the triketo tautomer. In the 6-amino 1,3-dimethyl
barbiturate substrate the keto oxygen atoms acquire higher
negative natural atomic charges by 0.055 up to 0.080 |e| relative
to the corresponding oxygen atoms of the triketo tautomer of the
1,3-dimethyl barbiturate substrate, while the α-carbon atom
acquires less negative natural atomic charge by 0.12 |e|.
The highest occupied molecular orbital (HOMO) in the triketo
tautomer is mainly localized on the keto-oxygen atoms adjacent to
the α-carbon atom (64.4% having 2p character), while in the
diketo−enolic tautomer is mainly localized on the α-carbon atom
Figure 4. Natural atomic charges and relevant FMOs of the 1,3-dimethyl barbiturate carbanion and the protonated (Z)-(2-nitrovinyl)benzene,
(Z)-1-methoxy-4-(2-nitrovinyl)benzene, and (Z)-(2-nitroprop-1-enyl)benzene calculated at the PBE0/6-311++G(d,p) level of theory in aqueous
solutions. The C(4) labels on the protonated nitro-styrenes are in bold.
F
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
LUMOs of the protonated (Z)-(2-nitrovinyl)benzene, (Z)-(2-
nitroprop-1-enyl)benzene, and (Z)-1-methoxy-4-(2-nitrovinyl)-
benzene are mainly localized on the electrophilic C(2) carbon
atom (42.0%, 36.1%, and 42.3% respectively) (Figure 5). It is
The dual descriptor is another useful function used to reveal
reactive sites. Formally, the definition of the dual descriptor Δf
has a close relationship with the Fukui function:
Δf(r) = f +(r) − f −(r)
Δf(r) predicts both types of reactive sites simultaneously. It is
argued that when Δf > 0, the site is favorable to nucleophilic
attack, whereas when Δf < 0, the site is favorable for electrophilic
attack.
Excellent linear relationships are obtained for the yield (%) vs.
Δf C(7) and yield (%) versus f−C(7) correlations. Good linear
relationships are also obtained for the yield (%) vs. PA and yield
(%) vs. μ correlations (μ = (εHOMO + εLUMO)/2) (Figure 5).
These correlations show that the nucleophilic character of the
C(7) carbon atom of the trans-β-nitrostyrenes affects the efficacy
of the MA reaction. Upon protonation the C(2) carbon atom in
the protonated trans-β-nitrostyrenes becomes an electrophilic
center. The calculated electronic descriptors related with the
electrophilic character of the C(2) carbon atom in the protonated
trans-β-nitrostyrene molecules are given in the Supporting
Information (Table S13). Excellent linear relationships are
obtained for the yield (%) vs. f−C(2) and yield(%) vs. QC(2)
correlations (f+C(2) is the condensed Fukui function for
electrophilic attack) (Figure 6). However, poor linear relationship
Figure 5. Linear relationships between yield (%) vs. ΔfC(7), yield (%)
vs. f−C(7), yield (%) vs. PA and yield (%) vs. μ correlations calculated at
the PBE0/6-311++G(d,p) level of theory in aqueous solutions.
apparent that the nucleophilic attack at the electrophilic C(2) carbon
atom by the carbanion is a FMO-controlled process supported by the
interaction of the HOMO(nucleophile) with the LUMO-
(electrophile). The nucleophilic attack at the electrophilic C(2)
carbon atom of the protonated (Z)-(2-nitrovinyl)benzene, (Z)-(2-
nitroprop-1-enyl)benzene, and (Z)-1-methoxy-4-(2-nitrovinyl)-
benzene by the 1,3-dimethyl barbiturate carbanion is predicted to
be strongly exothermic by 61.4, 58.7, and 49.1 kcal/mol, respectively.
To probe the electrophilic/nucleophilic character of the trans-
β-nitrostyrene molecules, selected electronic descriptors (con-
densed Fukui function, f−C(7),46,47 condensed dual descriptor
Figure 6. Linear relationships between yield (%) vs. f−C(2) and yield
(%) vs QC(2) correlations calculated at the PBE0/6-311++G(d,p) level
of theory in aqueous solutions.
Δf C(7),48 proton affinity, PA, and chemical potential, μ49,50
)
were calculated by the PBE0/6-311++G(d,p)/PCM computa-
tional protocol in order to find out how the electrophilic/
nucleophilic character of the electrophilic/nucleophilic centers
of the trans-β-nitrostyrenes affects the efficacy of the MA reaction
(see Supporting Information, Table S12).
The Fukui function, which is an important concept in
conceptual DFT, has been widely used in prediction of reactive
site of molecules. The Fukui function is defined as
is obtained for the yield (%) versus f−C(2) correlation, since the
C(2) carbon atom in the protonated trans-β-nitrostyrene
molecules is not susceptible to electrophilic attack.
The above correlations show that the electrophilic character of
the C(2) carbon atom would affect the efficacy of the MA
reaction, since it determines the strength of the nucleophilic
attack by the 1,3-dimethyl barbiturate carbanion nucleophile.
In summary, the MA reaction should involve first a proton
transfer from the 1,3-dimethyl barbituric acid to the C(NO2)Me
carbon atom, which is accompanied by charge density
redistribution rendering the C(2) carbon atom electrophilic
center that is attacked by the nucleophile 1,3-dimethyl
barbiturate carbonation formed.
To gain insights into the activation of the substrates of the MA
reaction catalyzed by the bimetallic ZnY catalysts we explored all
possible coordination modes of the substrates to the metal
centers of the catalysts by means of DFT calculations at the
PBE0/Def2-TZVP(Zn,Y)∪6-31G(d,p)(E) (E = main group
element) level of theory in aqueous solution using a
representative catalyst, 7ZnY. The estimated binding energy of
nitrate to the Y metal center is 30.8 kcal/mol. Therefore,
dissociation of the nitrate seems to be difficult. However,
examination of the defect dicubane structure of the catalyst
suggests that its dissociation into dinuclear species in solution
may well be possible. It can be seen that the two dimeric species
⎡
⎤
∂ρ(r)
∂N
f(r) =
⎢
⎣
⎥
⎦
ν
where N is number of electrons in the present system. The
constant term ν in the partial derivative is the external potential.
In the finite difference approximation, the Fukui function can be
calculated unambiguously for three situations:
Nucleophilic attack: f +(r)
= ρN+1(r) − ρN(r)
≈ ρLUMO(r)
Electrophilic attack: f −(r)
= ρN(r) − ρN−1(r)
≈ ρHOMO(r)
G
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
in the defect dicubane structure are loosely associated, for the
Zn−O distances of 2.331(2) are longer than the Zn−O distances
of 2.083(2) of the loosely associated DMF with Zn (Figure 7).
of the 7ZnY-barbiturate substrate adducts, with the barbiturate
substrate approaching the Y metal center, failed. In contrast, a
local minimum was located on the potential energy surface
(PES) of a 7ZnY-barbiturate adduct involving coordination of
barbiturate to Zn metal atom of the catalyst upon dissociation of
the coordinated DMF ligand. Note that the estimated binding
energy of DMF to the Zn metal center is 10.7 kcal/mol at the
PBE0/Def2-TZVP(Zn,Y)∪6-31G(d,p)(E) (E = main group
element) level of theory in aqueous solutions. The binding of the
barbiturate substrate to Zn metal center is very weak, and the
estimated binding energy was found to be 2.8 kcal/mol, which is
indicative of weak noncovalent interactions. The calculated
distance between the Zn and the keto O atom of the triketo form
of the 1,3-dimethyl barbituric acid was found to be 2.140 Å.
On the other hand, trans-β-nitrostyrene substrates showed a
preference to be coordinated to Y metal center of the 7ZnY
catalyst, via a unidentate coordination mode,51 and the estimated
binding energy was predicted to be 4.3 kcal/mol at the PBE0/
Def2-TZVP(Zn,Y)∪6-31G(d,p)(E) (E = main group element)
level of theory in aqueous solutions. The calculated distance
between the Y and the O atom of the unidentate trans-β-
nitrostyrene substrate was found to be 2.445 Å. On approaching
the trans-β-nitrostyrene substrate to Zn central atom resulted in a
local minimum corresponding to a very weak association of the
barbiturate substrate with the 7ZnY catalyst, but the binding
energy was marginal (0.2 kcal/mol). Then we optimized the
geometry of the adduct formed upon interaction of the trans-β-
nitrostyrene substrate with the Y metal center of the 7ZnY-
barbiturate adduct. In this adduct the barbiturate and trans-β-
nitrostyrene substrates are both coordinated to the Zn and Y
metal centers of the 7ZnY-catalyst, respectively. The estimated
Zn−O and Y−O bond distances are predicted to be 2.135
and 2.454 Å, respectively. For this adduct the estimated
binding energy of the trans-β-nitrostyrene−yttrium interaction
is 4.1 kcal/mol. The equilibrium geometries of the 7ZnY-
barbiturate and 7ZnY-trans-β-nitrostyrene adducts optimized at
the PBE0/Def2-TZVP(Zn,Y)∪6-31G(d,p)(E) (E = main group
element) level of theory in aqueous solutions are given in the
Supporting Information (Figure S9).
Figure 7. Defect dicubane structure of the catalyst 4ZnY with the
estimated intermolecular Y···O and Zn···O distances between the
dinuclear species indicative of a loose association of the dimers in the
defect dicubane structure. Color code: Zn, green; Y, mauve; O, red; N,
blue; C, gray.
Compare also with the Zn−O distances of 2.044(2) between Zn
and the bridging O atom. Similarly, the Y−O distances are very
long and indicative of weak interactions.
Then, we calculated the natural atomic charges on the metal
centers of the ZnY bimetallic catalysts which have been
characterized crystallographically using their crystal structures
in aqueous solution and analyzed the MOs for only the
representative catalyst 7ZnY. The natural atomic charges along
with the efficacy for these catalysts are compiled in Table 6, while
Table 6. Natural Atomic Charges on the Metal Centers of
3d/4f ZnY Bimetallic Catalyst Calculated at the PBE0/Def2-
TZVP(Zn,Y)∪6-31G(d,p)(E) (E = Main Group Element)
Level of Theory in Aqueous Solutions
It can be concluded that the barbiturate and trans-β-
nitrostyrene substrates for the MA reaction catalyzed by the
ZnY bimetallic catalyst are cooperatively activated through their
interaction with the Zn and Y metal centers of the catalyst
following the proposed mechanism shown in Scheme 1.
catalyst
yield (%)
QZn
QY
2ZnY
100
100
98
1.308
1.311
1.326
1.320
1.325
1.322
1.330
1.331
1.307
1.935
1.929
1.942
1.945
1.932
1.937
1.927
1.926
1.935
16ZnY
19ZnY
10ZnY
14ZnY
8ZnY
In the proposed mechanism, initially the barbiturate substrate
is coordinated to the Zn metal center of the 7ZnY catalyst after
dissociation of the coordinated DMF solvent molecule and then
coordination of the trans-β-nitrostyrene substrate to the Y metal
center of the catalyst. This could be a stepwise process, or both
substrates could be simultaneously coordinated to the metal
centers of the catalyst yielding the intermediate Im. The
coordination of the substrates to the Zn and Y metal centers
yielding Im is a slightly exothermic process, and the estimated
exothermicity is predicted to be 6.9 kcal/mol at the PBE0/Def2-
TZVP(Zn,Y)∪6-31G(d,p)(E) (E = main group element) level of
theory. The two substrates upon coordination to the metal
centers are brought into proximity such that a proton can be
transferred from the α-carbon atom of the barbiturate to the
CH(NO2) carbon atom of the coordinated trans-β-nitrostyrene
followed by the nucleophilic attack of the resulting barbiturate
carbanion to the electrophilic C(7) carbon atom of the trans-β-
nitrostyrene to yield the intermediate 1Im. These processes are
shown by the arrows drawn on the relevant Im structures of the
86
86
73
7ZnY
55
4ZnY
43
15ZnY
42
the three-dimensional (3D) plots of FMOs of the representative
catalyst7ZnY are shown in the Supporting Information (Figure S8).
Inspection of the FMOs of 7ZnY catalyst reveals that the
virtual acceptor FMOs (LUMO, LUMO+1, LUMO+2, LUMO
+3, etc.) are of π*-type located on the ligands. As no acceptor
orbital is located on the metal centers of the ZnY bimetallic
catalyst coordination of the MA substrates to the metal centers
supported by the interaction of the acceptor orbitals with the
donor orbitals (HOMO, HOMO−1, HOMO−2, etc.) of the
substrates is not feasible. Therefore, the substrates interact with
the bimetallic catalyst through noncovalent interactions (electro-
static anddispersion forces). Allattemptstooptimize thegeometry
H
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
by the arrows drawn on the relevant Im structures of the catalytic
cycle (Scheme 1). The MA product would then be weakly
associated with the Y metal center through the uncoordinated
O donor atom of the N(O)OH moiety (Figure 8). The formation
of this isomeric intermediate is also a slightly exothermic process,
the estimated exothermicity being 4.8 kcal/mol. The dissociation
energy for the release of the -N(O)OH product is 13.9 kcal/mol,
and this released product readily isomerizes to the more stable
(by 16.6 kcal/mol) -NO2 product. Comparison of the two
possible reaction pathways reveals that the first pathway
involving the formation of 1Im is more feasible than that
involving the formation of 2Im.
Scheme 1. A Proposed Mechanism Involving the Cooperative
Activation of Both the Barbiturate and trans-β-Nitrostyrene
Substrates via Their Coordination to the Zn and Y Metal
Centers of the ZnY Catalyst
The diastereoselectivity can be explained by considering the
very loose association of the two substrates with the catalyst,
where their orientation is mainly determined by maximization of
the noncovalent interactions between the substrates, leading
always to the formation of the R*/R* diastereoisomer
irrespective of which isomer of the nitrostyrene is used. These
noncovalent interactions for the catalyst−substrate complex 2Im
are shown in the Supporting Information (Figure S10). The
loose association of the two substrates with the catalyst would
also account for our findings using UV−vis spectroscopy that the
coordination environments of the catalyst and the catalyst +
substrate are indistinguishable.
To support the proposed mechanism (Scheme 1), ESI-MS
data on a sample taken during the reaction of β-methyl
nitrostyrene and barbituric acid catalyzed by 1ZnGd gave
peaks with m/z values that support the binding of the two
substrates (see Supporting Information, Figures S12−15).
CONCLUSIONS
■
In summary, the present study demonstrates for the first time the
successful use of 3d/4f CCs as efficient catalysts for MA
reactions. Although modification of the ligands in these CCs
does not appear to significantly influence their structure or their
catalytic properties, tuning the metal centers has provided
evidence that two metal centers act cooperatively to orchestrate
these rapid and highly diastereoselective reactions. DFT
calculations have been used to illuminate a possible mechanism
in which the CCs dissociate in solution to yield a pair of
catalytically active dimers each containing a 3d and a 4f metal
center. A range of spectroscopic techniques have been used to
corroborate this notion. In particular, EPR studies indicate that
the coordination number of both metals is unchanged in
solution, confirming that the structural integrity of the dual metal
scaffold remains intact and therefore able to effect cooperative
catalysis. Previous studies by Shibasaki,14,17 using 3d/4f
complexes as catalysts, have employed complexes prepared
in situ, for which no structural data either in the solid or solution
states were obtained. By contrast, this current work employs
structurally characterized crystalline tetrameric precatalysts and
provides experimental evidence that these retain structural
integrity in solution, albeit as 3d/4f metal containing dimers.
Given the proximity of the two Lewis acidic metal centers in
these structures, this supports the notion of cooperative catalysis
contributing to the strong catalytic behavior they display. Future
work will focus on collecting additional kinetic and spectroscopic
evidence to illuminate the detailed mode of catalysis for these
MA reactions and identifying new reactions for which the
proximity of a tunable pair of 3d and 4f metal centers can provide
a unique rate and/or selectivity characteristics.
catalytic cycle (Scheme 1). This intermediate comprises the MA
product weakly bound to the Y metal center through the O donor
atom of the NO2 moiety (Figure 8). The formation of this
Figure 8. Equilibrium geometries of possible intermediates formed in
the MA reaction of the barbiturate to trans-nitrostyrene catalyzed by a
ZnY bimetallic catalyst calculated at the PBE0/6-311++G(d,p) level of
theory in aqueous solution.
intermediate is a slightly exothermic process, the estimated
exothermicity being 6.4 kcal/mol. The coordinated product in
1Im immediately dissociates yielding the catalyst with release of
1.0 kcal/mol. Alternatively the proton could be transferred from
the α-carbon atom of the barbiturate substrate to the
uncoordinated O atom of the NO2 moiety of the trans-β-
nitrostyrene, resulting in transformation of the NO2 group into
an N(O)OH group, followed by the nucleophilic attack of the
barbiturate carbanion on the electrophilic C(7) carbon atom of
the trans-β-nitrostyrene yielding the intermediate 2Im. The
proton transfer and the nucleophilic attack processes are shown
I
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(12) Li, Y.; Deng, P.; Zeng, Y.; Xiong, Y.; Zhou, H. Org. Lett. 2016, 18,
1578−1581.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on theACS
Publications website at DOI: 10.1021/acs.inorgchem.7b01011.
■
S
(13) Handa, S.; Nagawa, K.; Sohtome, Y.; Matsunaga, S.; Shibasaki, M.
Angew. Chem., Int. Ed. 2008, 47, 3230−3233.
(14) Handa, S.; Gnanadesikan, V.; Matsunaga, S.; Shibasaki, M. J. Am.
Chem. Soc. 2010, 132, 4925−4934.
(15) Mechler, M.; Frey, W.; Peters, R. Organometallics 2014, 33, 5492−
5508.
Experimental procedures, ES-MS, X-ray, computational
details and NMR data (PDF)
Accession Codes
(16) Sohtome, Y.; Kato, Y.; Handa, S.; Aoyama, N.; Nagawa, K.;
Matsunaga, S.; Shibasaki, M. Org. Lett. 2008, 10, 2231−2234.
(17) Handa, S.; Gnanadesikan, V.; Matsunaga, S.; Shibasaki, M. J. Am.
Chem. Soc. 2007, 129, 4900−4901.
CCDC 1540023−1540036 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing
data_request@ccdc.cam.ac.uk, or by contacting The Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge
CB2 1EZ, UK; fax: +44 1223 336033.
(18) Maayan, G.; Christou, G. Inorg. Chem. 2011, 50, 7015−7021.
(19) Evangelisti, F.; More, R.; Hodel, F.; Luber, S.; Patzke, G. R. J. Am.
Chem. Soc. 2015, 137, 11076−11084.
(20) YIII is not a lanthanide but has ionic radii similar to HoIII and
catalytic behaviour to DyIII.
AUTHOR INFORMATION
Corresponding Author
*E-mail: G.Kostakis@sussex.ac.uk.
ORCID
Oliver P. E. Townrow: 0000-0001-9556-6450
Alan C. Spivey: 0000-0001-5114-490X
George E. Kostakis: 0000-0002-4316-4369
■
(21) Griffiths, K.; Gallop, C. W. D.; Abdul-Sada, A.; Vargas, A.;
Navarro, O.; Kostakis, G. E. Chem. - Eur. J. 2015, 21, 6358−6361.
(22) Griffiths, K.; Kumar, P.; Mattock, J. D.; Abdul-Sada, A.; Pitak, M.
B.; Coles, S. J.; Navarro, O.; Vargas, A.; Kostakis, G. E. Inorg. Chem.
2016, 55, 6988−6994.
(23) Griffiths, K.; Kumar, P.; Akien, G.; Chilton, N. F.; Abdul-Sada, A.;
Tizzard, G. J.; Coles, S.; Kostakis, G. E. Chem. Commun. 2016, 52,
7866−7869.
(24) Kumar, P.; Lymperopoulou, S.; Griffiths, K.; Sampani, S. I.;
Kostakis, G. E. Catalysts 2016, 6, 140.
Author Contributions
The manuscript was written through contributions of G.E.K.,
A.C.T., A.B., and A.C.S. All authors have given approval to the
final version of the manuscript.
(25) Kumar, P.; Griffiths, K.; Lymperopoulou, S.; Kostakis, G. E. RSC
Adv. 2016, 6, 79180−79184.
(26) Hegedus, L. S. Coord. Chem. Rev. 2000, 204, 199−307.
(27) Heravi, M. M.; Hajiabbasi, P.; Hamidi, H. Curr. Org. Chem. 2014,
18, 489−511.
Funding
EPSRC (UK) for funding (Grant Number EP/M023834/1).
Notes
(28) Hamashima, Y.; Sodeoka, M. Handbook of C-H Transformations:
Applications in Organic Synthesis; Wiley-VCH, 2005.
(29) Thirumalaikumar, M. Org. Prep. Proced. Int. 2011, 43, 67−129.
The authors declare no competing financial interest.
(30) Krause, N.; Hoffmann-Roder, A. Synthesis 2001, 2001, 171−196.
̈
ACKNOWLEDGMENTS
■
(31) Howlader, P.; Das, P.; Zangrando, E.; Mukherjee, P. S. J. Am.
Chem. Soc. 2016, 138, 1668−1676.
We thank the EPSRC (UK) for funding (Grant Number EP/
M023834/1), the University of Sussex, for offering a Ph.D.
position to K.G., the EPSRC UK National Crystallography
Service at the University of Southampton52 for the collection of
the crystallographic data for compounds 4ZnY, 7ZnY, 8ZnY,
10ZnY, 14ZnY, 15ZnY, 16ZnY, VIII, and XII and the EPSRC
UK National EPR Service at The University of Manchester.
(32) Al-Najjar, H. J.; Barakat, A.; Al-Majid, A. M.; Mabkhot, Y. N.;
Weber, M.; Ghabbour, H. A.; Fun, H.-K. Molecules 2014, 19, 1150−
1162.
(33) Seeliger, F.; Berger, S. T. A.; Remennikov, G. Y.; Polborn, K.;
Mayr, H. J. Org. Chem. 2007, 72, 9170−9180.
(34) Griffiths, K.; Dokorou, V. N.; Spencer, J.; Abdul-Sada, A.; Vargas,
A.; Kostakis, G. E. CrystEngComm 2016, 18, 704−713.
(35) Prat, D.; Hayler, J.; Wells, A. Green Chem. 2014, 16, 4546−4551.
(36) Mondal, K. C.; Sundt, A.; Lan, Y.; Kostakis, G. E.; Waldmann, O.;
Ungur, L.; Chibotaru, L. F.; Anson, C. E.; Powell, A. K. Angew. Chem.,
Int. Ed. 2012, 51, 7550−7554.
REFERENCES
■
(1) Mirica, L. M.; Vance, M.; Rudd, D. J.; Hedman, B.; Hodgson, K. O.;
Solomon, E. I.; Stack, T. D. P. Science 2005, 308, 1890−1892.
(2) Kanady, J. S.; Tsui, E. Y.; Day, M. W.; Agapie, T. Science 2011, 333,
733−736.
(37) Mondal, K. C.; Kostakis, G. E.; Lan, Y.; Wernsdorfer, W.; Anson,
C. E.; Powell, A. K. Inorg. Chem. 2011, 50, 11604−11611.
(38) Irving, H.; Williams, R. J. P. J. Chem. Soc. 1953, 3192.
(39) Stoll, S.; Schweiger, A. J. Magn. Reson. 2006, 178, 42−55.
(40) Misra, S. K. Multifrequency Electron Paramagnetic Resonance.
Theory and Applications - Chapter 19 2011, 1022.
̌
(3) Nesterov, D. S.; Chygorin, E. N.; Kokozay, V. N.; Bon, V. V.; Boca,
R.; Kozlov, Y. N.; Shul’pina, L. S.; Jezierska, J.; Ozarowski, A.; Pombeiro,
A. J. L.; Shul’pin, G. B. Inorg. Chem. 2012, 51, 9110−9122.
(4) Okamura, M.; Kondo, M.; Kuga, R.; Kurashige, Y.; Yanai, T.;
Hayami, S.; Praneeth, V. K. K.; Yoshida, M.; Yoneda, K.; Kawata, S.;
Masaoka, S. Nature 2016, 530, 465−468.
(41) Wang, X.-L.; Hou, L.-L.; Zhang, J.-W.; Zhang, J.-X.; Liu, G.-C.;
Yang, S. CrystEngComm 2012, 14, 3936−3944.
(5) Bilyachenko, A. N.; Dronova, M. S.; Yalymov, A. I.; Lamaty, F.;
Bantreil, X.; Martinez, J.; Bizet, C.; Shul’pina, L. S.; Korlyukov, A. A.;
Arkhipov, D. E.; Levitsky, M. M.; Shubina, E. S.; Kirillov, A. M.; Shul’pin,
G. B. Chem. - Eur. J. 2015, 21, 8758−8770.
(6) Ma, J. A.; Cahard, D. Angew. Chem., Int. Ed. 2004, 43, 4566−4583.
(7) Matsunaga, S.; Shibasaki, M. Chem. Commun. 2014, 50, 1044−
1057.
(42) Buckingham, D. A.; Clark, C. R.; McKeown, R. H.; Wong, O. J.
Am. Chem. Soc. 1987, 109, 466−476.
(43) Delchev, V. B. J. Struct. Chem. 2004, 45, 570−578.
(44) Senthilkumar, K.; Kolandaivel, P. J. Comput.-Aided Mol. Des. 2002,
16, 263−272.
(45) Schmidt, M. U.; Bruning, J.; Glinnemann, J.; Hutzler, M. W.;
̈
̈
Morschel, P.; Ivashevskaya, S. N.; van de Streek, J.; Braga, D.; Maini, L.;
̈
(8) Park, J.; Hong, S. Chem. Soc. Rev. 2012, 41, 6931−6943.
(9) Buchwalter, P.; Rose, J.; Braunstein, P. Chem. Rev. 2015, 115, 28−
́
Chierotti, M. R.; Gobetto, R. Angew. Chem., Int. Ed. 2011, 50, 7924−
126.
7926.
(10) Mankad, N. P. Chem. - Eur. J. 2016, 22, 5822−5829.
(11) Pye, D. R.; Mankad, N. P. Chem. Sci. 2017, 8, 1705−1718.
(46) Parr, R. G.; Yang, W. J. Am. Chem. Soc. 1984, 106, 4049−4050.
(47) Ayers, P. W.; Levy, M. Theor. Chem. Acc. 2000, 103, 353−360.
J
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
́
(48) Morell, C.; Grand, A.; Toro-Labbe, A. J. Phys. Chem. A 2005, 109,
205−212.
(49) Parr, R. G.; Donnelly, R. A.; Levy, M.; Palke, W. E. J. Chem. Phys.
1978, 68, 3801−3807.
(50) Kohn, W.; Becke, A. D.; Parr, R. G. J. Phys. Chem. 1996, 100,
12974−12980.
(51) Men
́
dez, I.; Rodríguez, R.; Polo, V.; Passarelli, V.; Lahoz, F. J.;
̃
García-Orduna, P.; Carmona, D. Chem. - Eur. J. 2016, 22, 11064−11083.
(52) Coles, S. J.; Gale, P. A. Chem. Sci. 2012, 3, 683−689.
K
DOI: 10.1021/acs.inorgchem.7b01011
Inorg. Chem. XXXX, XXX, XXX−XXX