Crystal Structure and Catalytic Behavior in Olefin Epoxidation of a One-Dimensional Tungsten Oxide/Bipyridine Hybrid

Article abstract of DOI:10.1021/acs.inorgchem.5b00928

The tungsten oxide/2,2′-bipyridine hybrid material [WO₃(2,2′-bpy)]·nH₂O (n = 1-2) (1) has been prepared in near quantitative yield by the reaction of H₂WO₄, 2,2′-bpy, and H₂O in the mole ratio of ca. 1:2:700 at 160°C for 98 h in a rotating Teflon-lined digestion bomb. The solid-state structure of 1 was solved and refined through Rietveld analysis of high-resolution synchrotron X-ray diffraction data collected for the microcrystalline powder. The material, crystallizing in the orthorhombic space group Iba2, is composed of a one-dimensional organic-inorganic hybrid polymer, _∞¹[WO₃(2,2′-bpy)], topologically identical to that found in the previously reported anhydrous phases [MO₃(2,2′-bpy)] (M = Mo, W). While in the latter the N,N′-chelated 2,2′-bpy ligands of adjacent corner-shared {MO₄N₂} octahedra are positioned on the same side of the 1D chain, in 1 the 2,2′-bpy ligands alternate above and below the chain. The catalytic behavior of compound 1 for the epoxidation of cis-cyclooctene was compared with that for several other tungsten- or molybdenum-based (pre)catalysts, including the hybrid polymer [MoO₃(2,2′-bpy)]. While the latter exhibits superior performance when tert-butyl hydroperoxide (TBHP) is used as the oxidant, compound 1 is superior when aqueous hydrogen peroxide is used, allowing near-quantitative conversion of the olefin to the epoxide. With H₂O₂, compounds 1 and [MoO₃(2,2′-bpy)] act as sources of soluble active species, namely, the oxodiperoxo complex [MO(O₂)₂(2,2′-bpy)], which is formed in situ. Compounds 1 and [WO(O₂)₂(2,2′-bpy)] (2) were further tested in the epoxidation of cyclododecene, trans-2-octene, 1-octene, (R)-limonene, and styrene. The structure of 2 was determined by single-crystal X-ray diffraction and found to be isotypical with the molybdenum analogue.

Full text of DOI:10.1021/acs.inorgchem.5b00928

Article
pubs.acs.org/IC
Crystal Structure and Catalytic Behavior in Olefin Epoxidation of a
One-Dimensional Tungsten Oxide/Bipyridine Hybrid
Tatiana R. Amarante, Margarida M. Antunes, Anabela A. Valente,* Filipe A. Almeida Paz,*
Martyn Pillinger, and Isabel S. Gonca̧ lves*
Department of Chemistry, CICECOAveiro Institute of Materials, University of Aveiro, Campus Universitar
́
io de Santiago,
3810-193 Aveiro, Portugal
S
* Supporting Information
ABSTRACT: The tungsten oxide/2,2′-bipyridine hybrid material
[WO3(2,2′-bpy)]·nH2O (n = 1−2) (1) has been prepared in near
quantitative yield by the reaction of H2WO4, 2,2′-bpy, and H2O in the
mole ratio of ca. 1:2:700 at 160 °C for 98 h in a rotating Teflon-lined
digestion bomb. The solid-state structure of 1 was solved and refined
through Rietveld analysis of high-resolution synchrotron X-ray
diffraction data collected for the microcrystalline powder. The material,
crystallizing in the orthorhombic space group Iba2, is composed of a
one-dimensional organic−inorganic hybrid polymer, 1[WO3(2,2′-
∞
bpy)], topologically identical to that found in the previously reported
anhydrous phases [MO3(2,2′-bpy)] (M = Mo, W). While in the latter
the N,N′-chelated 2,2′-bpy ligands of adjacent corner-shared {MO4N2}
octahedra are positioned on the same side of the 1D chain, in 1 the 2,2′-bpy ligands alternate above and below the chain. The
catalytic behavior of compound 1 for the epoxidation of cis-cyclooctene was compared with that for several other tungsten- or
molybdenum-based (pre)catalysts, including the hybrid polymer [MoO3(2,2′-bpy)]. While the latter exhibits superior
performance when tert-butyl hydroperoxide (TBHP) is used as the oxidant, compound 1 is superior when aqueous hydrogen
peroxide is used, allowing near-quantitative conversion of the olefin to the epoxide. With H2O2, compounds 1 and [MoO3(2,2′-
bpy)] act as sources of soluble active species, namely, the oxodiperoxo complex [MO(O2)2(2,2′-bpy)], which is formed in situ.
Compounds 1 and [WO(O2)2(2,2′-bpy)] (2) were further tested in the epoxidation of cyclododecene, trans-2-octene, 1-octene,
(R)-limonene, and styrene. The structure of 2 was determined by single-crystal X-ray diffraction and found to be isotypical with
the molybdenum analogue.
[MoO2Cl2(L)] is an alternative route to such com-
pounds.16,20,21 The catalytic potential of molybdenum(VI)
oxide−organonitrogen systems for the epoxidation of olefins
and the oxidation of secondary amines to nitrones has been
studied.13,16,20−22 Depending on the structure and composition
of the hybrid, as well as the catalytic reaction conditions, these
materials act either as sources of soluble active species or as
heterogeneous catalysts.
The structural inorganic chemistry of tungsten generally
parallels that of molybdenum. Hence, several isotypical MoVI/
WVI oxide−organonitrogen systems are known, such as
[MO3(py)],4 [MO3(pyz)0.5],6,7 [MO3(2,2′-bpy)],9,10 and
[MO3(4,4′-bpy)0.5].4,5,7 On the other hand, there are differ-
ences in the physicochemical properties of MoVI and WVI (such
as Pauling electronegativity, ionic radius, reduction potential,
Lewis acid strength of the cations, and MO bond strengths),
which can result in isotypical complexes or materials that
exhibit contrasting behaviors for certain applications. Indeed,
tungsten oxide-based materials, including inorganic−organic
INTRODUCTION
■
Metal oxide−organic hybrid materials have potential applica-
tions in many fields, such as catalysis, gas storage, electrical
conductivity, magnetism, photochromism, and optics.1−3
Group VIb (Mo, W) systems are of particular interest,
especially those in which N-donor molecules are directly
coordinated to the oxide substructure.1,2 In an early report,
layered molybdenum and tungsten trioxides were intercalated
by pyridine (py) and 4,4′-bipyridine (4,4′-bpy) under strict
anhydrous conditions to give the hybrid materials [MO3(py)]
(M = Mo, W) and [MoO3(4,4′-bpy)0.5], in which the
heterocyclic nitrogen is directly coordinated to the metal
atom of the oxide layer.4 The latter material was subsequently
prepared by hydrothermal reaction of MoO3, 4,4′-bpy, and
H2O at 150 °C.5 Indeed, hydrothermal reactions of MO3 (or
simpler precursors like Na2MO4) with organonitrogen ligands,
such as pyrazine (pyz),6−8 2,2′-bpy and its derivatives,8−13 4,4′-
bpy,7,14 1,10-phenanthroline,15 pyrazolylpyridines,16 4,4′-bipyr-
idylamine,17 and triazoles,5,18,19 have led to the isolation of a
series of Mo (W) hybrid materials with structures based on
one-dimensional (1D) chains, 2D networks, or 3D frameworks.
Hydrolysis of dioxomolybdenum(VI) complexes of the type
Received: April 27, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
FT-IR spectra were collected using KBr (Sigma-Aldrich, 99%, FT-
IR grade) pellets and a Mattson-7000 infrared spectrophotometer.
Attenuated total reflectance (ATR) FT-IR spectra were measured
using a Specac Golden Gate Mk II ATR accessory having a diamond
top plate and KRS-5 focusing lenses. Solid-state 13C cross-polarization
(CP) magic-angle-spinning (MAS) NMR spectra were recorded using
a Bruker Avance 400 spectrometer (9.4 T) at 100.62 MHz with 3.7 μs
1H 90° pulses, 1.5 ms contact time, spinning rates of 10−12 kHz, and
5 s recycle delays. Chemical shifts are quoted in parts per million
(ppm) from tetramethylsilane.
[WO3(2,2′-bpy)]·nH2O (1). A mixture of H2WO4 (0.49 g, 1.96
mmol), 2,2′-bpy (0.62 g, 3.97 mmol), and H2O (25 mL) was heated in
a rotating (15 rpm) Teflon-lined stainless steel digestion bomb at 160
°C for 98 h. After cooling down to ambient temperature, the resultant
microcrystalline white solid was separated from the pink aqueous
liquor (pH 6) by filtration, washed with an excess of water and diethyl
ether (4 × 10 mL), and finally dried at 80 °C. Yield: 0.80 g, 96%
(based on W). Anal. Calcd for C10H12N2O5W ([WO3(2,2′-bpy)]·
2H2O): C, 28.32; H, 2.85; N, 6.61. Found: C, 28.30; H, 2.63; N, 6.63.
TGA showed a weight loss of 6.7% up to 150 °C (calcd: for loss of
2H2O, 8.5%; for loss of 1.5H2O, 6.5%). Selected FT-IR (KBr, cm−1): ν
= 3397 (s, br), 3313 (s, br), 3112 (m), 3079 (m), 3035 (m), 1650
(m), 1600 (s), 1575 (w), 1567 (w), 1540 (vw), 1494 (m), 1475 (m),
1446 (s), 1313 (m), 1170 (m), 1160 (w), 1122 (w), 1029 (m), 979
(w), 939 (m), 883 (vs), 846 (s), 775 (vs), 655 (vs), 455 (m), 435 (w),
414 (m), 372 (m), 349 (s). 13C{1H} CP MAS NMR: δ = 123.8, 124.5,
125.7 (C3/C3′, C5/C5′), 138.8, 142.4 (C4/C4′), 148.4 (C6/C6′),
149.6, 150.6 (C2/C2′) ppm.
hybrids, have attracted increasing interest because of their
unique properties and corresponding applications in optics and
electronics.8,23−25 For example, Tallon and co-workers doped
the layered hybrid material [WO3(4,4′-bpy)0.5] with Na+, K+,
and Ca2+ ions so as to alter the electronic structure and increase
conductivity.23−25 Such materials may also exhibit unique
catalytic properties, taking into account the contrasting
behaviors often encountered for molecular MoVI/WVI cata-
lysts.26−30 In catalytic olefin epoxidation, for example, W-based
catalysts usually show a lower activity than Mo-based catalysts
in the presence of tert-butyl hydroperoxide (TBHP) as the
oxidant. However, the situation is often reversed when H2O2 is
the oxidant, with the W-based catalysts being superior.28−30
These differences may be due to a complex interplay of various
factors, such as different inhibiting effects of water (much
stronger for Mo than for W), greater oxophilicity, and Lewis
acidity of the WVI center.30
Inspired by our successful investigations into molybdenum
oxide−organonitrogen systems as catalysts or catalyst pre-
cursors for oxidation reactions, we have started to explore
tungsten oxide systems. An additional motivation for such
studies is the above-mentioned tendency for oxotungsten(VI)
catalysts to work well with aqueous H2O2, which is recognized
as a cheap, safe, and environmentally clean oxidant. Here, we
report the first account of the catalytic behavior of a tungsten
oxide−organonitrogen hybrid material, namely, [WO3(2,2′-
bpy)]·nH2O. The epoxidation of cis-cyclooctene has been
chosen as a model reaction, using TBHP or H2O2 as oxidant,
and emphasis has been placed on studying the stability of the
hybrid and identifying the nature of the catalytically active
species. In addition, the structure of the hybrid has been solved
and refined through Rietveld analysis of high-resolution
synchrotron X-ray diffraction data.
[WO(O2)2(2,2′-bpy)] (2). A mixture comprising [WO3(2,2′-bpy)]·
nH2O (0.25 g, 0.59 mmol with n = 2), 30% aq H2O2 (10.2 mL, 0.09
mol), and CH3CN (65 mL) was heated at 70 °C for 24 h. The
resultant white precipitate was separated by filtration, washed with
cold distilled water and ethanol, and finally vacuum-dried at ambient
temperature. Yield: 0.16 g, 65%. Anal. Calcd for C10H8N2O5W: C,
28.59; H, 1.92; N, 6.67. Found: C, 28.40; H, 2.01; N, 6.61. Selected
FT-IR (KBr, cm−1): ν = 3448 (vs, br), 3112 (m), 3085 (s), 1637 (w),
1608 (sh), 1600 (s), 1571 (m), 1560 (m), 1496 (m), 1473 (s), 1446
(s), 1423 (m), 1315 (m), 1243 (m), 1220 (w), 1176 (m), 1160 (m),
1047 (m), 1035 (m), 1022 (w), 944 (vs), 847 (sh), 835 (s), 775 (s),
728 (m), 665 (m), 649 (m), 595 (m), 541 (s), 416 (w). 13C{1H} CP
MAS NMR: δ = 122.1, 123.3, 125.4, 128.7 (C3/C3′, C5/C5′), 144.0,
146.7 (C4/C4′), 147.9, 149.5 (C6/C6′), 153.9, 155.4 (C2/C2′) ppm.
Crystals suitable for X-ray diffraction were obtained by slow
evaporation of the filtrate from the above reaction. The PXRD pattern
of the bulk product 2 was in excellent agreement with a simulated
pattern calculated from the single-crystal X-ray structural data (Figure
S1 in the Supporting Information).
X-ray Diffraction Studies. The structures of 1 and 2 were
determined by powder and single-crystal XRD, respectively. High-
resolution PXRD data suitable for crystal solution were collected at
100 K on the powder diffractometer at the ID22 beamline of the
European Synchrotron Radiation Facility (ESRF), Grenoble, France.
Tables 1 and 2 gather all the details pertaining to the X-ray data
collection, crystal data, and structure refinement for 1 and 2. The final
Rietveld plot for 1 is supplied in Figure 1. A complete description of
the procedures used in the powder and single-crystal XRD studies is
given in the Supporting Information. Structural drawings were created
using the software package Crystal Impact Diamond.33
EXPERIMENTAL SECTION
■
Materials and Methods. For synthesis, H2WO4 (puriss p.a.,
Fluka), 2,2′-bipyridine (98.0%, Fluka), acetonitrile (analytical reagent,
Lab-Scan), and diethyl ether (puriss p.a., Sigma-Aldrich) were acquired
from commercial sources and used as received. The material
[MoO3(2,2′-bpy)] was prepared as described previously.31 For the
catalytic experiments, the olefins cis-cyclooctene (95%), cyclododecene
(96%, mixture cis/trans), trans-2-octene (97%), 1-octene, (R)-
limonene (97%), and styrene (99%) were obtained from Sigma-
Aldrich and used as received. 1,2-Dichloroethane (99%, Aldrich),
ethanol (>99.9% analytical grade, Scharlau ACS), anhydrous α,α,α-
trifluorotoluene (≥99%, Sigma-Aldrich), acetonitrile (≥99%, Aldrich),
undecane (99%, Sigma-Aldrich), 5.5 M tert-butyl hydroperoxide in
decane (Sigma-Aldrich), 70% aq tert-butyl hydroperoxide (Aldrich),
and 30% aq H O (Riedel-de Haen) were acquired from commercial
̈
2
2
sources and used as received.
Elemental analysis for C, H, and N was performed at the University
of Aveiro with a Leco TruSpec 630-200-200 analyzer. Routine powder
X-ray diffraction (PXRD) data were collected at ambient temperature
on a Philips Analytical Empyrean (θ/2θ) diffractometer equipped with
a PIXcel1D detector, with automatic data acquisition (X’Pert Data
Collector software v4.2) using monochromatized Cu Kα radiation (λ =
1.5406 Å). Intensity data were collected by the step-counting method
(step 0.01°), in continuous mode, in the ca. 3.5° ≤ 2θ ≤ 50° range.
Scanning electron microscopy (SEM) images were collected using a
Hitachi SU-70 microscope operating at 20 kV. Samples were prepared
by deposition on aluminum sample holders followed by carbon coating
using an Emitech K 950 carbon evaporator. Thermogravimetric
analysis (TGA) was carried out using a Shimadzu TGA-50 instrument,
from ambient temperature to ca. 800 °C, under a continuous stream of
air at a flow rate of 20 mL min−1 and a heating rate of 5 °C min−1.
Catalytic Tests. The catalytic tests were carried out at 55 or 70 °C
under air in closed borosilicate batch reactors (5 mL capacity)
equipped with a magnetic stirrer (1000 rpm) and a valve for sampling.
Typically, the reaction vessel was loaded with 1, 2, or [MoO3(2,2′-
bpy)] (0.018 mmol); olefin (1.8 mmol); oxidant (3 mmol); and
cosolvent (2 mL, when applied). The cosolvent was either 1,2-
dichloroethane (DCE), ethanol (EtOH), α,α,α-trifluorotoluene
(TFT), or acetonitrile (CH3CN). The oxidant solution was 70% aq
tert-butyl hydroperoxide (TBHPaq), 5.5 M tert-butyl hydroperoxide in
decane (TBHPdec), or 30% aq H2O2. The reactor containing
[MO3(2,2′-bpy)] (M = Mo, W) or [WO(O2)2(2,2′-bpy)], olefin,
and cosolvent (when applied) was preheated at the desired reaction
B
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. X-ray Data Collection, Crystal Data, and Structure
Refinement Details for [WO3(2,2′-bpy)]·H2O (1)
Table 2. Crystal and Structure Refinement Data for
Compound [WO(O2)2(2,2′-bpy)] (2)
data collection
formula
C10H8N2O5W
420.03
formula weight
crystal system
space group
a (Å)
diffractometer
wavelength (Å)
temp (K)
ID22 beamline, ESRF, France
0.495958(7)
monoclinic
P21/n
100
6.2509(6)
12.5514(11)
13.6903(12)
91.851(3)
1073.55(17)
4
geometry
Debye−Scherrer
1.002−30.000
unit cell
b (Å)
2θ range (deg)
c (Å)
β (deg)
volume (Å3)
b
formula
crystal system
space group
a (Å)
C10H8N2O4W
orthorhombic
Iba2
Z
Dc (g cm−3
)
2.599
19.03471(18)
17.14447(16)
7.27897(4)
2375.42(3)
8
μ(Mo Kα) (mm−1
crystal size (mm)
crystal type
)
10.776
b (Å)
0.22 × 0.06 × 0.04
colorless needles
3.54−25.34
−7 ≤ h ≤ 7, −15 ≤ k ≤ 15, −16 ≤ l ≤ 16
14503
c (Å)
volume (Å3)
θ range
Z
index ranges
Dc (g cm−3
)
2.259(1)
reflections collected
independent reflections
completeness to θ = 25.24°
final R indices
Profile Parameters
1967 [Rint = 0.0694]
99.8%
profile function
zero shift (2θ°)
fundamental parameters approach
0.003(1)
R1 = 0.0316, wR2 = 0.0585
a,b
Refinement Details
[I > 2σ(I)]
a,b
no. of independent reflections
no. of global refined parameters
817
final R indices (all data)
R1 = 0.0460, wR2 = 0.0623
m = 0.0054, n = 10.1229
1.941 and −1.097 e Å−3
c
102
weighting scheme
a
Rp, Rwp, Rexp, GOF
9.44, 13.16, 0.91, 14.5
largest diff. peak and hole
a
b
2
2
R1 = ∑||Fo| − |Fc||/∑|Fo|. wR2 = (∑[w(Fo − Fc )2]/
Structure Reliability Factors
c
2
2
2
∑[w(Fo )2])1/2. w = 1/[σ2(Fo ) + (mP)2 + nP], where P = (Fo
+
a
RBragg
5.40
a
2Fc2)/3.
2
Rp = ∑i|yi,o − yi,c|/∑i|yi,o|, Rwp = [∑iwi(yi,o − yi,c)2/∑iwiyi,o
]
1/2, RBragg
2
= ∑n|In,o − In,c|/∑iIn,o, Rexp = [(n − p)/∑iwiyi,o ] ,
1/2, GOF = Rwp/Rexp
where yi,o and yi,c are the observed and calculated profile intensities and
In,o and In,c are the observed and calculated intensities, respectively.
The summations run over i data points or n independent reflections.
RESULTS AND DISCUSSION
■
Synthesis and Physicochemical Characterization of
[WO3(2,2′-bpy)]·nH2O (n = 1−2) (1). Compound 1 was
obtained in excellent yield (96%) by the reaction of H2WO4,
2,2′-bpy, and H2O in the mole ratio of ca. 1:2:700 at 160 °C for
98 h in a Teflon-lined digestion bomb. Previously, Twu et al.
reported the isolation of [WO3(2,2′-bpy)] in 80% yield, using
exactly the same starting materials, reaction temperature, and
time.10 The higher yield obtained in the present work may be
because we performed the hydrothermal synthesis with
continuous rotation of the bomb, while Twu et al. apparently
carried out the reaction under static conditions. As will be
described below, performing the synthesis under dynamic
conditions leads not only to a higher yield but also to a unique
b
Statistical weights wi are usually taken as 1/yi,o. Hydrogen atoms from
the water molecule of crystallization have not been included.
temperature for 10 min, and afterward, the oxidant solution (also
preheated in the case of TBHP) was added to the reactor; the reaction
time was counted from this instant. The catalytic reactions were
monitored using a Varian 3800 GC equipped with a capillary column
(DB-5, 30 m × 0.25 mm) and a flame ionization detector. The
reaction products were identified by GC−MS (Trace GC 2000 Series
(Thermo Quest CE Instruments)-DSQ II (Thermo Scientific)
equipped with a capillary column (DB-5 MS, 30 m × 0.25 mm ×
0.25 μm), using He as carrier gas).
After each 24 h batch run (with EtOH as cosolvent), a solid was
separated by centrifugation at 3500 rpm, thoroughly washed with
ethanol, and dried overnight at ambient temperature and then under
vacuum for 1 h at 65 °C. The recovered solids were reused in
consecutive 24 h batch runs using similar initial mass ratios of
olefin:oxidant:catalyst:cosolvent to those used in the first run.
Contact tests (CTs) were carried out for 1 (with EtOH or CH3CN
as cosolvent) in order to assess whether the catalytic reaction was
homogeneous or heterogeneous. The CTs consisted of treating 1
under typical reaction conditions, but without the substrate, for 24 h at
70 °C. Subsequently, the (undissolved) solid was separated by
centrifugation, washed, and dried overnight at ambient temperature
and then under vacuum for 1 h at 65 °C, giving sol-CT-EtOH and sol-
CT-CH3CN. The solution obtained from the CT was passed through
a syringe equipped with a 0.2 μm VWR nylon membrane filter, giving
liq-CT-EtOH and liq-CT-CH3CN. The solids and solutions obtained
from the CTs were tested in the reaction of Cy under typical
conditions, at 70 °C.
structure for the 1D 1[WO3(2,2′-bpy)] polymer present in
∞
the crystal structure of 1.
Compound 1 was characterized by elemental analysis, TGA,
FT-IR spectroscopy, 13C{1H} MAS NMR, and PXRD,
including a complete structural determination using synchro-
tron X-ray data (described below). TGA of 1 revealed a weight
loss of 6.7% between ambient temperature and 150 °C, with a
plateau being reached at 100 °C (Figure S2 in the Supporting
Information), assigned to the removal of water molecules of
crystallization and residual physisorbed water (1−2 water
molecules per formula unit). In contrast, the tungsten oxide/
bipyridine hybrid material reported by Twu et al. was found to
be anhydrous.10 On the other hand, the two materials present
similar thermal behavior above 150 °C, with decomposition
taking place between 200 and 450 °C. For compound 1, three
overlapping steps contribute to a net weight loss of 37.0%. The
residual mass of 55.9% at 450 °C is in excellent agreement with
the theoretical value of 55.9%, calculated assuming loss of 2,2′-
C
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Final Rietveld plot of [WO3(2,2′-bpy)]·nH2O (1) (n = 1). Observed data points are indicated as a blue line, the best fit profile (upper
trace) and the difference pattern (lower trace) are drawn as solid red and gray lines, respectively. Blue and gray vertical bars indicate the angular
positions of the allowed Bragg reflections for 1 and the reported analogous compound with CCDC refcode LONSIQ, respectively.32 Relative
amounts (wt %) of the two samples present in the bulk material are 99.88(1):0.12(1) (1:LONSIQ). Refinement details are given in Table 1. The
inset depicts a perspective view along the [001] direction of the unit cell of 1.
bpy and formation of WO3, with n = 1.5 (water content) in the
starting material. SEM of 1 showed that the morphology
consists of long, thin needles (Figure 2).
CP MAS NMR spectrum of 1 displays eight separate signals in
the range of 120−150 ppm, suggesting that the two pyridyl
groups of the ligand are not equivalent in the crystal structure
of 1 (Figure 3). Indeed, the two signals at 123.8 and 148.4 ppm
are slightly broader than the remaining resonances; hence, each
of these two signals probably comprises two overlapping
Figure 2. Representative SEM images of (a) 1, (b) 2, and the solids
recovered after the (c) first (CH3CN as cosolvent) and (d) second
(EtOH as cosolvent) catalytic runs, with 1 as (pre)catalyst.
In the range of 500−950 cm−1, the FT-IR spectrum of 1
displays sharp bands at 775, 846, 883, and 939 cm−1 and a
broad, very intense absorption at 655 cm−1, similar to the data
published by Twu et al. for [WO3(2,2′-bpy)].10 The band at
775 cm−1 is assigned as a ligand (2,2′-bpy) mode, while the
other bands are assigned as W−O vibrations (νWO between
840 and 940 cm−1 and νW−O−W at 655 cm−1). The
molybdenum analogue, [MoO3(2,2′-bpy)], exhibits a similar
set of bands at 622, 760, 882, and 914 cm−1.20,31 The 13C{1H}
Figure 3. 13C{1H} CP MAS NMR spectra of (a) the hybrid 1 and (b)
the oxodiperoxo complex 2.
D
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 4. Schematic representation of the (a) 1D, neutral ∞1[WO3(2,2′-bpy)] polymer present in the crystal structure of 1 running parallel to the c-
axis, emphasizing the (b) zigzag nature of the polymer and the intermetallic distance. (c) Distorted octahedral coordination environment of the
crystallographically unique WVI center. Symmetry transformation used to generate equivalent atoms: (i) −x, 1 − y, −1 + z.
resonances (for C3/C3′ or C5/C5′, and C6/C6′, respectively),
giving the expected total of ten resonances.
and above the growth axis of the inorganic core of the polymer
(the torsion angle between consecutive moieties is ca. 175°),
being mutually placed in approximately parallel planes (Figure
5b). This behavior for organic linkers was previously reported
for the secondary [WO3(en)] phase described by Yang et al.,34
where en = ethylenediamine. Conversely, in the structure
reported by Yan et al.,32 the 2,2′-bpy ligands are on the same
side of the inorganic core of the chain, being slightly tilted
along that direction (the torsion angle between adjacent ligands
is ca. 50°). As shown in Figure 5a, consecutive 2,2′-bpy ligands
in LONSIQ are placed on average planes that subtend a mutual
dihedral angle of ca. 11.5°. Figure 5c overlays the two
topologically identical, but conformationally distinct,
High-Resolution Powder Synchrotron XRD Studies of
[WO3(2,2′-bpy)]·nH2O (1). The main crystalline product
obtained from the hydrothermal reaction of tungstic acid with
2,2′-bpy was formulated as [WO3(2,2′-bpy)]·nH2O on the
basis of high-resolution PXRD data (n = 1). As clearly observed
in the collected powder pattern, the material has trace amounts
of crystalline impurities (Figure 1). One of these impurities was
unequivocally identified as the related [WO3(2,2′-bpy)] phase
reported by Yan et al. (CCDC refcode LONSIQ), single-
crystals of which were obtained by the hydrothermal reaction of
Na2WO4, 2,2′-bpy, and H2O at 190 °C.32 The LONSIQ phase,
which is isotypical with the corresponding molybdenum
compound reported by Zubieta and co-workers,9 was obtained
in a microcrystalline form by Twu et al.10 as described above. A
second impurity (highlighted in Figure 1) has not yet been
identified, but it does not correspond to any of the starting
materials used in the synthesis or possible secondary products
the crystalline structure of which has been reported and
deposited in the Cambridge Structural Database. Rietveld
analysis assuming solely a mixture of 1 with LONSIQ (the
section of the powder pattern affected by the second impurity
was masked) clearly showed that the latter material is present in
trace amounts, accounting for only 0.12(1)% by weight of the
total sample (a value typically smaller than ca. 5% may be
considered as negligible).
1[WO3(2,2′-bpy)] polymers present in 1 and LONSIQ,
∞
emphasizing the geometrical differences described above.
The presence of distinct chemical moieties in the {WN2O4}
coordination polyhedron and the marked trans effect of the
terminal oxo groups contribute to the highly distorted
coordination geometry observed in 1: while the W−(N,O)
bond distances range from 1.755(16) to 2.334(7) Å, the cis and
trans (N,O)−W−(N,O) octahedral internal angles fall within
the ranges of 71.45(15)°−110.9(17)° and 153(2)°−164.4(5)°,
respectively (Table 3). These values are in very good agreement
with the geometrical data reported by Yan et al. for LONSIQ,
with the corresponding intervals being 1.714(9)−2.381(11) Å,
68.8(4)°−107.0(5)°, and 150.7(4)°−163.4(4)°.32 To further
compare these two structures and to evaluate any variation in
the degree of distortion of the WVI octahedra between 1 and
LONSIQ, an adaptation of the mathematical method proposed
by Baur35 was employed to calculate a distortion index (DI) for
bonds and angles:
The main structural feature of 1 concerns the presence of a
1D, neutral 1[WO3(2,2′-bpy)] hybrid polymer (Figure 4a).
∞
The basic building unit of this polymer is a six-coordinated WVI
metal center bonded to three oxo groups (two terminal and
one μ2-bridging) and one N,N′-chelated 2,2′-bpy organic
moiety. The coordination geometry of the resulting
{WN2O4} polyhedron resembles a highly distorted octahedron
(Figure 4c). Although the polymer in 1 is topologically identical
to that reported by Yan et al.,32 there is a marked distinction
concerning the orientation of the 2,2′-bpy organic ligands. In
the compound reported herein, the N,N′-chelated 2,2′-bpy
moieties of adjacent coordination polyhedra alternate below
n
i=1
∑
|d(W−(N,O)) − d(W−(N,O))average
|
i
DIBond
=
n
∑
|d(W−(N,O)) |
i
i=1
(1)
E
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 5. Comparison between the geometrical features of the 1D, neutral ∞1[WO3(2,2′-bpy)] polymers present in compound 1 and in the crystal
structure reported by Yan et al.32 (CCDC refcode LONSIQ).
Table 3. Selected Bond Lengths (Å) and Angles (deg) for the
DIAngle
=
Crystallographically Independent WVI Center Present in
n
a
∑
|((N,O)−W−(N,O)) − ((N,O)−W−(N,O))ideal
|
[WO3(2,2′-bpy)]·nH2O (1) (n = 1)
i
i=1
n
∑
|((N,O)−W−(N,O)) |
W1−O1
W1−O1i
W1−O2
W1−O3
W1−N1
W1−N2
1.99(6)
O1−W1−O1i
O1−W1−O2
O1i−W1−O2
O1−W1−O3
O1i−W1−O3
O1i−W1−N1
O1i−W1−N2
O1−W1−N1
O1−W1−N2
O2−W1−O3
O2−W1−N1
O2−W1−N2
O3−W1−N1
O3−W1−N2
N1−W1−N2
153(2)
i
i=1
(2)
1.78(6)
110.9(17)
83.5(18)
99(2)
1.84(2)
where the subscripts i and average/ideal correspond to
individual and average/ideal bond lengths and angles (see
Table 3 for individual data). The DIBond values for the WVI
centers are 0.11 and 0.12 for 1 and LONSIQ, respectively,
while the DIAngle values for the cis/trans angles are 0.12/0.13
for 1 and 0.12/0.14 for LONSIQ. These results clearly
demonstrate that the rotation of the N,N′-chelated 2,2′-bpy
organic linkers along the W → μ2-O → W direction of the
inorganic backbone of the polymer does not significantly affect
the geometry of the {WN2O4} coordination polyhedron, which
is equally highly distorted in both compounds.
1.755(16)
2.323(3)
2.334(7)
100(2)
85.5(15)
78.2(19)
74.2(13)
77.8(17)
102.5(8)
162.4(11)
92.8(6)
93.0(5)
164.4(5)
71.45(15)
One main difference observed for the 1[WO3(2,2′-bpy)]
∞
a
polymers present in 1 and LONSIQ resides in the intermetallic
W···W distance along the W → μ2-O → W direction of the
polymer. While in LONSIQ this distance is 3.6766(8) Å, in 1
the distance is slightly longer, being 3.75(4) Å (Figure 4b). The
Symmetry transformation used to generate equivalent atoms: (i) −x,
1 − y, −1 + z.
F
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
“kink” (W−(μ2-O)−W) angle is also higher in the structure of
1 [171(3)° versus 147.7(5)° for LONSIQ]. These distinct
structural features seem to mostly arise from the orientation of
the pendant N,N′-chelated ligands, which influences the
supramolecular contacts present in the structure (see below).
Similarly, in the [WO3(en)] structure described by Yang et
al.,34 in which the ligands alternate above and below the growth
direction of the polymer, the intermetallic W···W distance is
greater [3.7247(8) Å] than in LONSIQ, with a W−(μ2-O)−W
angle of 152.3(11)°.
Compound 1 contains one water molecule of crystallization
per WVI center, while LONSIQ is completely anhydrous.
Because of the strong hydrophobic nature of the organic ligand,
these water molecules are engaged in mostly weak supra-
molecular contacts in the crystal structure. Even though the
alternate rotation of the 2,2′-bpy organic ligands along the
growth direction of the polymer gives rise to 1D channels in
which these water molecules reside, their inclusion may not be
the reason for the different conformations observed for 1 and
LONSIQ. Indeed, in the [WO3(en)] crystal structure, the en
ligands adopt the same alternate orientation and the structure
does not include any solvent molecules. Hence, the presence of
water molecules in the crystal structure of 1 probably results
from the need to essentially fill the available space, not being a
crucial and directing structural feature. Under the high-energy
X-ray beam of the synchrotron, these molecules may be easily
removed (even at low temperatures), thus promoting the
formation of partially dehydrated phases that could account for
the small amount of crystalline impurities.
Table 4. Influence of the Reaction Conditions on the
Epoxidation of cis-Cyclooctene in the Presence of 1
a
oxidant
cosolvent temp (°C) conversion (%) CyO selectivity (%)
TBHPdec
none
55
70
70
70
70
70
70
55
70
70
10
57
4
100
100
100
100
100
100
100
100
100
100
none
DCE
TFT
4
EtOH
CH3CN
CH3CN
CH3CN
CH3CN
EtOH
1
8
TBHPaq
H2O2
2
77
98
99
H2O2
H2O2
a
Reaction conditions: initial molar ratios of W:Cy:oxidant =
1:100:167, 2 mL of cosolvent (when applied), 24 h.
(pre)catalysts in the same reaction, e.g., cyclopentadienyl
compounds of the type [Cp′W(CO)3R] (Table 5).36
For 1, increasing the reaction temperature from 55 to 70 °C
enhanced the reaction rate (57% conversion at 24 h) without
affecting epoxide selectivity, which was always 100% (Table 4,
Figure 6). At this higher reaction temperature, the addition of a
cosolvent (DCE, TFT, EtOH, or CH3CN) had a detrimental
effect, with CyO yields at 24 h being in the range of 1−8%,
similar to the 2% conversion observed without catalyst. In
contrast, the molybdenum analogue [MoO3(2,2′-bpy)] exhibits
much higher activity when used as the catalyst for the same
reaction with TBHPdec as oxidant and DCE as cosolvent: 95%
conversion at 6 h/55 °C and 100% conversion at 1 h/75 °C.31
In the presence of 1, the reaction of Cy was sluggish for both
the organic (TBHPdec) and aqueous (TBHPaq) solutions of
TBHP. The reaction mixtures were always biphasic solid−
liquid. For all cosolvents, TBHP solutions, and reaction
temperatures tested, the PXRD (Figure 7) and FT-IR ATR
spectra (Figure S5, Supporting Information) of the solids
recovered by centrifugation of the reaction mixtures after a 24 h
batch run were similar to those of 1. Hence, compound 1 seems
to be fairly oxidatively and hydrolytically stable under the
TBHP-based conditions. The poor catalytic performance may
be partly due to the inability of 1 to activate TBHP for the
catalytic reaction.
The crystal packing of 1 (inset in Figure 1) is essentially
mediated by the need to effectively fill the available space
alongside with a number of cooperative, weak C−H···O
hydrogen-bonding interactions. The most notable ones concern
the interactions along the b-axis interconnecting adjacent
1[WO3(2,2′-bpy)] polymers, in which the N,N′-chelated
∞
2,2′-bpy molecule from one chain interacts with the two
terminal oxo groups of the neighboring one (see Figure S3a of
the Supporting Information). These interactions are, in general,
highly directional (at least two with interactions angles close to
ca. 180°; see Table S1, Supporting Information), ultimately
contributing to a strong cohesion of the polymers in the solid
state. These interpolymer connections are reinforced by the
C4−H14···O3 hydrogen-bonding connection along the [001]
direction of the unit cell (Figure S3b, Supporting Information).
The water molecules of crystallization lie, as mentioned above,
in the channels of the structure and interact weakly with
neighboring N,N′-chelated 2,2′-bpy molecules (Figure S4,
Supporting Information). Even though the ∠(DHA) inter-
action angles are typically greater than ca. 160°, the interatomic
distance between donor and acceptors are relatively long
(usually greater than ca. 3.4 Å; see Table S1, Supporting
Information).
When aqueous H2O2 was used as oxidant instead of TBHP,
the reaction of Cy in the presence of 1 was much faster, giving
CyO as the only product in 77% yield at 55 °C/24 h and 98%
yield at 70 °C/24 h, with CH3CN as cosolvent (Figure 6, Table
4). No cocatalyst is necessary to obtain fairly good catalytic
results for the system 1/H2O2. For example, it has been
reported for [WO(O2)(L)2] (L = N-cinnamyl-N-phenyl-
hydroxamic acid), tested as catalyst in the same reaction, that
no epoxidation reaction occurred without NaHCO3 as
cocatalyst (Table 5).37 Another example of a tungsten-based
(pre)catalyst that performs well for the epoxidation of Cy with
aq H2O2 in the absence of a cocatalyst is the cyclopentadienyl
oxotungsten(VI) complex [Cp*2W2O5], which gave an epoxide
yield of about 90% after ca. 100 min reaction at 55 °C (Table
5).38 For compounds of the type [Fe(L)(CH3CN)2](ClO4)2 (L
= aminopyridine ligand), the epoxidation of Cy with H2O2
(CH3CN as cosolvent) was much faster than that found for 1,
although in the former case a large excess of acetic acid was
used to promote the catalytic reaction (Table 6).42 Complexes
of the type [ReO3(CH3)L] [L = bis(fluorous-ponytailed)-2,2′-
bipyridines] led to faster reaction of Cy than 1 at 55 °C;
Catalytic Epoxidation of cis-Cyclooctene. Effect of
Oxidant, Cosolvent, and Temperature and Comparison
with Literature Data. Compound 1 was explored as a
(pre)catalyst for olefin epoxidation using cis-cyclooctene (Cy)
as a benchmark substrate. First, TBHPdec was used as oxidant,
no cosolvent was added, and the reaction was performed at 55
°C. Excellent selectivity (100%) toward cyclooctene oxide
(CyO) was observed for 1, but the catalytic activity was very
poor (10% conversion at 24 h reaction, Table 4). As mentioned
in the introduction, poor catalytic results are frequently
reported in the literature for tungsten compounds tested as
G
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 5. Comparison of the Catalytic Results for 1 with Literature Data for Selected Tungsten Complexes Tested as Catalysts in
the Reaction of cis-Cyclooctene with H2O2 or TBHP
a
reaction conditions
b
c
(pre)catalyst
solvent/cocatalyst
T (°C)
[Cy]0 (M)
[W]0 (M)
t (h)
conv (%)
sel (%)
ref
H2O2 as Oxidant
d
1
1
1
EtOH/−
CH3CN/−
CH3CN/−
CH3CN/−
70
55
70
25
25
55
0.7
0.007
0.007
24
24
24
99
100
100
100
−
tw
tw
tw
37
37
38
0.7
77
0.7
0.007
98
e
[WO(O2)(L)2]
[WO(O2)(L)2]
[Cp*2W2O5]
0.83
∼0.0004
∼0.0004
0.0054
∼0
72
e
CH3CN/NaHCO3
0.83
0.5
100
CH3CN + toluene/−
0.27
1.67
∼90
100
TBHP as Oxidant
1
−
−
55
70
at
2.4
0.024
0.024
0.011
0.011
0.011
0.011
0.020
0.020
4/24
4/24
2/10
13/57
0.02
−
−
−
100
100
100
−
−
−
−
−
1
2.4
f
[Cp*2W2O5]
[Cp*2W2O5]
[Cp*2W2O5]
[Cp*2W2O5]
[CpansaW(CO)3]
[CpW(CO)3CH3]
CDCl3/−
(CD3)2CO/−
CH3OD/−
CD3CN/−
−
1.15
1.15
1.15
1.15
2.03
2.03
14
30
30
30
30
36
36
f
at
7
f
at
6
f
at
2
55
55
4
4
<10
100
−
<10
100
a
Reaction conditions: T = reaction temperature (at = ambient temperature), [Cy]0 = initial molar concentration of Cy, [W]0 = initial molar
b
c
d
e
concentration of W, t = reaction time. Conversion of Cy. Selectivity toward CyO. tw = This work. L = N-cinnamyl-N-phenylhydroxamic acid.
f
Reaction time in days.
Figure 6. Dependence of Cy conversion and CyO yield on the time of
reaction in the presence of 1 with TBHPdec/(no cosolvent) at 55 °C
◇
□
(
) or 70 °C ( ), H2O2/CH3CN (*) at 55 °C, and H2O2/EtOH (Δ)
or H2O2/CH3CN (×) at 70 °C. For all catalytic tests, CyO selectivity
was always 100%.
however, twice the molar amount of metal was used in the case
of the rhenium catalysts, and halogenated cosolvents were
used.43
Iodometric titrations (ca. 10% error) for the systems 1/
H2O2/(EtOH or CH3CN) heated at 70 °C for 24 h (without
olefin) indicated no significant decomposition of the oxidant.
These results suggest that, for 1, H2O2 is productively
consumed in the epoxidation process (i.e., negligible H2O2
decomposition to water plus molecular oxygen).
Figure 7. PXRD patterns of 1 (a) and the corresponding solid
recovered from a 24 h batch run using TBHPdec/EtOH (b), TBHPaq/
CH3CN (c), H2O2/CH3CN (d), and H2O2/EtOH [(1/H2O2/EtOH)-
run1 (e), (1/H2O2/EtOH)-run 2 (f), (1/H2O2/EtOH)-run 3 (g)], at
70 °C. The remaining patterns are for the solids obtained from the
contact tests, namely, sol-CT-CH3CN (h) and sol-CT-EtOH (i), and
the solids recovered from a batch run with sol-CT-CH3CN and sol-
CT-EtOH [(sol-CT-CH3CN)-run1 (j) and (sol-CT-EtOH)-run1 (k),
respectively].
The catalytic results for 1/H2O2 are far superior to those
obtained for the molybdenum analogue [MoO3(2,2′-bpy)],
tested as catalyst under similar reaction conditions (Table 6):
83% and 25% conversion for 1 and [MoO3(2,2′-bpy)],
respectively, at 6 h reaction, with CyO as the only reaction
product. Other related molybdenum(VI) compounds contain-
ing the 2,2′-bipyridine ligand, such as the hybrid material
{[MoO3(2,2′-bpy)][MoO3(H2O)]}n,20 gave poorer results
when tested as catalysts in the same reaction (Table 6). A
H
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 6. Comparison of the Catalytic Results for 1 with Literature Data for Compounds Possessing Bipyridine-type Ligands,
a
Tested as Catalysts in the Reaction of cis-Cyclooctene with H2O2
reaction conditions
(pre)catalyst
solvent
EtOH
T (°C)
[Cy]0 (M)
[M]0 (M)
t (h)
conv. (%)
sel. (%)
ref
1
1
1
2
70
55
70
70
70
55
60
60
60
at
0.7
0.007
0.007
6/24
6/24
6/24
24
83/99
47/77
84/98
98
100
100
100
100
100
100
<1
tw
tw
tw
tw
tw
20
39
40
40
41
41
42
43
43
44
45
45
CH3CN
CH3CN
CH3CN
EtOH
0.7
0.7
0.007
0.7
0.007
[MoO3(2,2′-bpy)]
0.7
0.007
6/24
48
25/71
10
{[MoO3(2,2′-bpy)][MoO3(H2O)]}n
[Mo(O2)2(H2O)n] + 2eq 2,2′-bpy
[MoO(O2)2(2,2′-bpy)]
[MoO(O2)2(2,2′-bpyO2]
[MoO(O2)2(2,2′-bpy)]
[MoO(O2)2(4,4′-bpy)]
CH2Cl2
1.0
0.01
CHCl3
0.1
0.0024
0.011
18
12
b
[C8mim]PF6
[C8mim]PF6
CH2Cl2
0.45
0.45
1.22
1.22
0.09
0.1
18
61
38
c
0.011
18
65
100
100
100
100
100
100
100
100
100
0.0037
0.0037
∼0.0005
0.002
6
40
CH2Cl2
at
6
30
d
e
[Fe(L)(CH3CN)2](ClO4)2
CH3CN
at
0.08
5
100
>98
>98
98
[ReO3(CH3)(2,2′-bpy)]
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
at
f
[ReO3(CH3)(bpy-Fn)]
at
0.1
0.002
5
[ReO3(CH3)(4,4′-Me2-2,2′-bpy)]
[ReO3(CH3)(2,2′-bpy)]
[ReO3(CH3)(Me2-2,2′-bpy)]
25
at
0.88
0.95
0.95
0.0088
0.0095
0.0095
3
2/24
1
72/72
CH2Cl2
at
∼100
c
a
b
See footnotes a−d in Table 5 for abbreviations. [C8mim]PF6 = 1-n-octyl-3-methylimidazolium hexafluorophosphate. 2,2′-bpyO2 = 2,2′-bipyridine
d
e
f
N,N′-dioxide. L = bipyridine bipiperidine. Excess acetic acid was used. bpy-Fn = bis(fluorous-ponytailed) 2,2′-bipyridine with the substituent
group (CH2)3(CF2)5CF3.
similar trend has been reported for oxomolybdenum/tungsten
analogues possessing cyclopentadienyl ligands, where the W
compound was more effective than the Mo analogue for the
catalytic epoxidation of olefins and oxidation of thiophenes with
H2O2.30,38,46 A theoretical study (based on DFT calculations)
by Dinoi et al. indicated a lower activation barrier for
[Cp*2W2O5] than for its Mo analogue, which is consistent
with the higher catalytic activity of the former.38 It has been
proposed that the reaction mechanism for tungsten coordina-
tion compounds involves the primary activation of H2O2, where
an oxo ligand (WO) is protonated with concomitant
formation of a hydroxide−hydroperoxo intermediate possessing
the moiety {W(OH)(OαOβH)}, in which (OαOβH) is η2-
coordinated; this intermediate plays a role in the oxygen atom
transfer reaction to the olefin.38 On the other hand, for the
molybdenum-based catalytic system [Cp*MoO2Cl]/H2O2, the
active hydroperoxo intermediates formed may undergo isomer-
ization (via proton transfer) to give less active peroxo
derivatives.47
Identification of Active Species and Catalytic Stability
with H2O2. For the reaction systems 1/H2O2/(CH3CN or
EtOH), the mixtures were biphasic solid−liquid. The solids
were recovered after a 24 h batch run [giving (1/H2O2/
CH3CN)-run1 and (1/H2O2/EtOH)-run1] and characterized
by FT-IR ATR spectroscopy, PXRD, and SEM. These solids
recovered from a 24 h batch run with aqueous H2O2 as oxidant and
Figure 8. FT-IR ATR spectra of 1 (a) and the corresponding solids
exhibited comparable PXRD patterns (Figure 7) and FT-IR
ATR spectra (Figure 8), which are clearly different from those
of 1. Hence, 1 is converted into different species under these
reaction conditions. SEM revealed a change in morphology
from thin, long needles for 1 to larger, irregular blocks for the
recovered solids (Figure 2). The solid (1/H2O2/EtOH)-run1
was reused twice (with H2O2/EtOH, at 70 °C; details given in
the Experimental Section). The catalytic results were similar for
the three batch runs (Table 7). Moreover, the crystalline
structures (Figure 7), FT-IR ATR spectra (Figure 8), and
particle morphologies (Figure 2) of the solids recovered after
each run were similar. To assess more closely the rate of
CH3CN as cosolvent, at 55 °C (b) or 70 °C (c), and with EtOH as
cosolvent at 70 °C [(1/H2O2/EtOH)-run1 (d), (1/H2O2/EtOH)-run
2 (e), (1/H2O2/EtOH)-run 3 (f)]. The remaining spectra are for the
solids obtained from the contact tests, namely, sol-CT-CH3CN (g)
and sol-CT-EtOH (h), Prec-liq-CT-CH3CN (i), and the solids
recovered from a batch run with sol-CT-CH3CN and sol-CT-EtOH
[(sol-CT-CH3CN)-run1 (j) and (sol-CT-EtOH)-run1 (k), respec-
tively].
transformation of 1, additional reactions were performed for
each cosolvent at 70 °C, and the solids present in the reaction
mixtures were isolated after 15, 30, 60, and 120 min reaction
I
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
sol-CT-EtOH and sol-CT-CH3CN) and liquid phases (denoted
liq-CT-EtOH and liq-CT-CH3CN) were tested in the reaction
of Cy with H2O2. Catalytic results for sol-CT and liq-CT (96−
100% conversion at 24 h) were similar to those obtained for 1
(Tables 6 and 7). Hence, dissolved metal species play an
important role, and the reaction of Cy occurs via homogeneous
catalysis. Indeed, the solids sol-CT-EtOH and sol-CT-CH3CN
exhibited similar FT-IR spectral features (Figure 8) and
crystalline structures (Figure 7), matching the data obtained
for the solids recovered after 24 h batch catalytic runs. After
using sol-CT-EtOH and sol-CT-CH3CN in the catalytic tests,
the resultant recovered solids [denoted (sol-CT-EtOH)-run1
and (sol-CT-CH3CN)-run1] exhibited similar FT-IR spectral
features and PXRD patterns to those displayed by sol-CT-
EtOH and sol-CT-CH3CN (Figures 7 and 8).
Table 7. Consecutive Batch Runs of Epoxidation of cis-
Cyclooctene with H2O2 using 1 as Precatalyst and Catalytic
Tests with Solid (sol) and Liquid (liq) Phases of the Contact
a
Tests for 1 with EtOH and CH3CN
sample
cosolvent conversion (%) CyO selectivity (%)
(1/H2O2/EtOH)-run1
(1/H2O2/EtOH)-run2
(1/H2O2/EtOH)-run3
sol-CT-EtOH
EtOH
EtOH
EtOH
EtOH
EtOH
CH3CN
CH3CN
99
97
100
100
100
100
100
100
100
99
96
liq-CT-EtOH
100
97
sol-CT-CH3CN
liq-CT-CH3CN
98
a
Reaction conditions: initial molar ratios of W:Cy:oxidant =
1:100:167, 2 mL of cosolvent, 24 h, 70 °C.
To check if the chemical nature of the dissolved metal
species was similar to that of the undissolved compound (sol-
CT-CH3CN), attempts were made to isolate metal species from
liq-CT-CH3CN. This was successfully accomplished by adding
n-hexane to the solution and leaving it overnight at 4 °C; the
precipitated solid was separated by centrifugation (3500 rpm),
washed with CH3CN, and dried, giving a solid denoted as Prec-
liq-CT-CH3CN. The FT-IR spectral features of the solids Prec-
liq-CT-CH3CN and sol-CT-CH3CN were very similar (Figure
8), supporting the above hypothesis that 1 is converted to a
single, stable compound and that the latter is responsible for
the observed homogeneous catalytic reaction. With the aim of
unambiguously identifying this compound, a large-scale (33-
fold increase) experiment was carried out for 1/H2O2/CH3CN
(without olefin; please see the Experimental Section for
details). After 24 h of reaction at 70 °C, the reaction mixture
was cooled to ambient temperature and filtered to isolate a
times. With CH3CN as cosolvent, the FT-IR spectrum of the
solid after 15 min was similar to that of the solid recovered
from a 24 h batch run [(1/H2O2/CH3CN)-run1] (Figure S6 in
the Supporting Information). On the other hand, with EtOH as
cosolvent, the FT-IR spectrum of the solid after 15 min
exhibited a mix of spectral features indicative of both 1 and (1/
H2O2/EtOH)-run1, suggesting that the transformation of 1 is
slower with this cosolvent than with CH3CN. Nevertheless,
after 1 h with EtOH as cosolvent, the transformation of 1 into
the new species was evidently complete. Hence, 1 seems to be
converted into a single, stable compound under the reaction
conditions used.
In order to check for the catalytic contribution of
(undissolved) metal species and their chemical nature, contact
tests (CTs) were carried out for 1/H2O2 as described in the
Experimental Section (without substrate, using EtOH or
CH3CN as cosolvent, at 70 °C). The resultant solids (denoted
a
Table 8. Olefin Epoxidation with H2O2 in the Presence of 1 or 2, Using CH3CN as Solvent
c
epoxide sel
b
d
conv (%)
(%)
other reaction products (%)
substrate
1
2
1
2
1
2
cis-cyclooctene
cyclododecene
trans-2-octene
98
83
92
98
79
93
100
100
64
100
100
79
−
−
−
−
2,3-octanediol
2-octene-4-one
3-octene-2-one (12)
3,4-epoxy-2-octanone (9)
3,5-octadiene-2-ol (6)
1,2-octanediol
heptanal
2,3-octanediol
2-octene-4-one
3-octene-2-one (6)
3,4-epoxy-2-octanone (5)
3,5-octadiene-2-ol (5)
1,2-octanediol
heptanal
1-octene
27
27
95
96
2-octenal
2-octenal
1-octene-3-one
heptanoic acid
1-octene-3-one
heptanoic acid
e
e
(R)-limonene
84
74
81
76
56
57
4-isopropenyl-1-methyl-1,2-cyclohexanediol (37)
5-isopropenyl-2-methyl-7-oxobicyclo[4.1.0]heptan-2-ol
styrene glycol (17)
4-isopropenyl-1-methyl-1,2-cyclohexanediol (36)
5-isopropenyl-2-methyl-7-oxobicyclo[4.1.0]heptan-2-ol
styrene glycol (10)
styrene
3
4
benzaldehyde (43)
benzaldehyde (44)
benzoic acid (12)
benzoic acid (14)
2-hydroxyacetophenone (23)
benzene acetaldehyde
2-hydroxyacetophenone (25)
benzene acetaldehyde
acetophenone
acetophenone
a
b
c
Reaction conditions: initial molar ratios of W:Cy:oxidant = 1:100:167, 2 mL of cosolvent, 24 h, 70 °C. Olefin conversion. Epoxide product
d
e
selectivity. Byproducts (results in parentheses are the selectivities for the main byproducts identified). The epoxide products were 1,2-epoxy-p-
menth-8-ene and l,2-8,9-diepoxy-p-menthane, and the molar ratio of 1,2-epoxy-p-menth-8-ene/1,2-8,9-diepoxy-p-menthane was 2.5 and 3.1 for 1 and
2, respectively.
J
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
colorless solid, which was identified as the oxodiperoxo
complex [WO(O2)2(2,2′-bpy)] (2) by elemental analysis,
single-crystal (see below) and powder XRD, FT-IR, and
NMR spectroscopy. A comparison of the FT-IR spectral
features of 2 with those for the recovered solids described
above confirms that complex 2 is the catalytically active species
formed from 1 in the presence of aqueous H2O2. The FT-IR
spectrum of 2 shows bands at 944 cm−1 [ν(WO)]; 847 and
835 cm−1 [ν(O−O)]; and 649, 595, and 541 cm−1 [ν(W−
O)peroxo], which are all characteristic of [MO(O2)2L] complexes
bearing one bidentate ligand (L).16,21 The 13C{1H} CP MAS
NMR spectrum of 2 displays 10 separate signals in the range of
120−155 ppm, indicating that the two pyridyl groups of the
ligand are not equivalent in the crystal structure of 2 (Figure 3).
In consonance with that observed for 1, the solid recovered
after a catalytic run (70 °C, 24 h) using [MoO3(2,2′-bpy)] as
precatalyst, 30% aq H2O2 as oxidant, and EtOH as cosolvent
was identified by FT-IR and PXRD as the complex [MoO-
(O2)2(2,2′-bpy)]. The recovered solid displayed bands in the
FT-IR spectrum at 534, 579, and 648 [ν(Mo−O)peroxo]; 856
[ν(O−O)]; and 935 cm−1 [ν(MoO)], in agreement with
published data for the oxodiperoxo complex (Figure S7 in the
Supporting Information).48 As will be described below, the
complexes [MO(O2)2(2,2′-bpy)] (M = Mo, W) are isotypical.
Accordingly, the PXRD pattern for the recovered solid closely
matched that for 2 (Figure S1, Supporting Information).
The solubility of 2 was apparently greater in CH3CN than in
EtOH since generally a lower amount of solid was recovered for
the former. These observations are consistent with the faster
conversion of 1 into 2 in CH3CN discussed above (monitored
by FT-IR). However, the kinetics of the epoxidation of Cy is
roughly comparable for the two cosolvents. Possibly, the type of
cosolvent influences the stability of transition states formed in
the catalytic cycle.
The reaction of Cy with H2O2, in the presence of 2 (CH3CN
as cosolvent), led to CyO as the only product formed with 98%
yield at 24 h, 70 °C (Table 6). These results are similar to those
obtained for 1. Hence, 1 is transformed in situ into 2, which is
responsible for the catalytic reaction. The catalytic results for 2
are far superior to those reported in the literature for the
molybdenum analogue [MoO(O2)2(2,2′-bpy)] tested in the
same reaction, under similar conditions (Table 6).40,41 This
trend somewhat parallels that observed for 1 and its
molybdenum analogue [MoO3(2,2′-bpy)] (discussed above).
The FT-IR spectrum of the solid recovered from the reaction of
Cy in the presence of 2 was similar to that for the original
complex (Figure S8, Supporting Information), suggesting that 2
is fairly stable.
styrene, the respective diol product (1,2-diol in the case of (R)-
limonene) was formed, which most likely occurred via
hydrolysis of the (relatively sensitive) epoxides. The poor
epoxide selectivity in the reaction of styrene was accompanied
by the formation of various byproducts, including 2-
hydroxyacetophenone, benzaldehyde, and benzoic acid, sug-
gesting that deeper oxidation occurred.
The results indicate that the cyclic olefin Cy is more reactive
than cyclododecene, which may be due to steric effects
associated with the epoxidation of the bulkier olefin. For the
C8 linear olefins 1-octene and trans-2-octene, the more
substituted olefin trans-2-octene is substantially more reactive
than the terminal one. A similar reactivity trend was observed
for the diene substrate (R)-limonene, for which selectivity to
1,2-epoxy-p-menth-8-ene was higher than that to l,2-8,9-
diepoxy-p-menthane. Hence, regioselectivity favored the
epoxidation of the endocyclic CC bond over the terminal
one. These results suggest that the catalytic reactions may be
governed by steric and/or electronic effects. The FT-IR ATR
spectra of the solids recovered after using 1 or 2 with the
different olefins were similar to that for 2 (Figure S8,
Supporting Information), which is consistent with the catalyst
stability of 2 discussed above.
Crystal Structure of [WO(O2)2(2,2′-bpy)] (2). The crystal
structure of 2 is isotypical with that of the MoVI analogue
published by Schlemper et al. in the early 1980s, as clearly
depicted by the unit cell parameters and space group symmetry
summarized in Table 2.49 A search of the literature reveals, to
the best of our knowledge, that this latter compound was the
first reported oxodiperoxo MoVI compound. Compound 2 is,
surprisingly, the first crystallographically determined structure
of a compound of this kind based on WVI centers. Since the
report by Schlemper et al., over a dozen related structures (with
MoVI) have been reported, most of them comprising N,N-
chelated organic linkers in the metal coordination sphere.
Besides the recent work of Herbert et al. with 3-methylpyrazole
and pyrazole,40,50 all remaining known structures are based on
N,N′-chelated organic linkers.51
The asymmetric unit of compound 2 is composed of a whole
molecular unit as depicted in Figure 9a. The crystallographically
independent WVI center is coordinated to two peroxo groups,
one terminal oxo moiety, and, as in compound 1, an N,N′-
chelated 2,2′-bpy organic ligand. The W−(N,O) bond
distances were found in the 1.726(5)−2.297(6) Å range
(Table 9), with the shorter being attributed to the terminal oxo
group and the longer ones to the connections to the organic
ligand. The overall {WN2O5} coordination sphere can be
simplified by solely considering the centers of gravity of the two
coordinated peroxo groups, leading to a distorted trigonal
bipyramidal coordination environment, as depicted in Figure
9b. Table 9 summarizes the most relevant geometrical
information concerning these two ways of considering the
coordination environment of the WVI centers in 2.
Even though the [WO(O2)2(2,2′-bpy)] complex is rich in
atoms capable of acting as strong acceptors in hydrogen-
bonding interactions, the molecule only has these atoms bound
to carbon, and therefore, only weak supramolecular interactions
mediate the crystal packing of 2. A series of weak C−H···O
interactions (Table S2 in the Supporting Information) connect
neighboring [WO(O2)2(2,2′-bpy)] complexes in the crystal
structure, with the most important interactions being
represented in Figure S9 (Supporting Information). The
cooperative nature of these interactions, together with the
Reactivity of Other Olefins in the Presence of 1 or 2.
The catalytic performances of the two compounds were further
investigated for the epoxidation of different olefins, namely,
cyclododecene, trans-2-octene, 1-octene, styrene, and (R)-
limonene, using H2O2 as oxidant and CH3CN as cosolvent, at
70 °C, 24 h (Table 8). In general, with each substrate, 1 and 2
led to comparable conversions, similar to that observed with Cy
as substrate. These results are consistent with the above
hypothesis that 2 was responsible for the catalytic reaction.
The epoxide selectivity was excellent (100%) for the
substrates Cy and cyclododecene, very high for 1-octene
(87−95%), fair for trans-2-octene (64−79%) and (R)-limonene
(56−57%), and very poor for styrene (3−4%). Some of the
reaction byproducts identified by GC−MS are listed in Table 8.
For the substrates 1-octene, trans-2-octene, (R)-limonene, and
K
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
fact that most are relatively directional [interaction ∠(DHA)
angles typically above ca. 150°] and with short D···A distances
(see Table S2, Supporting Information), boosts their
importance in the crystal packing of 2. On the other hand,
the coordinated 2,2′-bpy ligands are also engaged in a series of
π···π stacking interactions, as depicted in Figure S10 of the
Supporting Information (for geometrical data, see Table S2,
Supporting Information). A view of the crystal packing is given
in Figure S11 (Supporting Information).
CONCLUSIONS
■
A new structural variant of a tungsten oxide/2,2′-bipyridine
hybrid material with the composition [WO3(2,2′-bpy)]·nH2O
has been prepared and characterized. The use of a dynamic
hydrothermal synthesis appears to be responsible for the
formation of the unique crystal structure in 1, which is
composed of 1[WO3(2,2′-bpy)] polymeric chains in which
∞
the 2,2′-bpy ligands are located alternately above and below the
growth axis of the chain rather than just on one side, as is found
in the previously reported anhydrous phases [MO3(2,2′-bpy)]
(M = Mo, W), which were isolated after static hydrothermal
syntheses. The catalytic behavior of 1 has been studied in detail
for the epoxidation of the model substrate cis-cyclooctene.
When aqueous H2O2 is used as the oxidant, 1 is quickly
converted to the partially soluble oxodiperoxo complex
[WO(O2)2(2,2′-bpy)] (2), which is the active species
responsible for the selective epoxidation reaction in the
homogeneous phase. Hence, for this particular reaction,
compound 1 is useful as a stable and easily and safely storable
catalyst precursor. Compounds 1 and 2 promoted the reaction
of different olefins (cyclododecene, trans-2-octene, 1-octene,
styrene, and (R)-limonene) using H2O2 as oxidant; with each
olefin, the two compounds led to similar catalytic results. When
compared with molybdenum, tungsten oxide/organonitrogen
hybrid materials have been less well studied, with unsubstituted
pyridine and bipyridine molecules being the most common
ligands employed. Work is ongoing in our laboratory to expand
Figure 9. (a) Schematic representation of the discrete complex present
in the asymmetric unit of compound [WO(O2)2(2,2′-bpy)] (2). Non-
hydrogen atoms are represented as thermal ellipsoids drawn at the
50% probability level and hydrogen atoms as small spheres with
arbitrary radii. (b) Simplified polyhedral representation of the WVI
coordination environment, evidencing the distorted trigonal bipyr-
amidal coordination environment when the centers of gravity of the
peroxo groups are taken as the central coordination site. Selected bond
lengths and angles are given in Table 8
Table 9. Selected Bond Lengths (in Å) and Angles (in deg) for the Crystallographically Independent WVI Metal Center Present
a
in Compound [WO(O2)2(2,2′-bpy)] (2)
W1−O1
W1−O2
W1−O3
W1−O4
1.964(5)
1.916(5)
1.909(5)
1.953(5)
1.726(5)
86.1(2)
W1−N1
W1−N2
2.180(6)
2.297(6)
W1−Cg(O2,1)
W1− Cg(O2,2)
O4−W1−O1
O4−W1−N1
O4−W1−N2
O5−W1−O1
O5−W1−O2
O5−W1−O3
O5−W1−O4
O5−W1−N1
O5−W1−N2
N1−W1−N2
1.792(3)
1.783(3)
157.2(2)
86.4(2)
W1−O5
O1−W1−N1
O1−W1−N2
O2−W1−O1
O2−W1−O4
O2−W1−N1
O2−W1−N2
O3−W1−O1
O3−W1−O2
O3−W1−O4
O3−W1−N1
O3−W1−N2
Cg(O2,1)−W1− Cg(O2,2)
Cg(O2,1)−W1−O5
Cg(O2,1)−W1−N1
Cg(O2,1)−W1−N2
78.6(2)
45.3(2)
78.6(2)
133.0(2)
130.2(2)
86.4(2)
100.0(2)
104.7(2)
106.5(2)
101.8(2)
91.8(2)
132.8(2)
89.9(2)
45.1(2)
163.1(2)
71.3(2)
130.2(2)
85.9(2)
132.51(2)
103.38(2)
107.93(2)
81.83(2)
Cg(O2,2)−W1−O5
Cg(O2,2)−W1−N1
Cg(O2,2)−W1−N2
105.29(2)
108.13(2)
81.56(2)
a
Cg(O2,1): Center of gravity of the O1−O2 coordinated peroxo group. Cg(O2,2): Center of gravity of the O3−O4 coordinated peroxo group.
L
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(8) Zhang, X.; Hejazi, M.; Thiagarajan, S. J.; Woerner, W. R.;
Banerjee, D.; Emge, T. J.; Xu, W.; Teat, S. J.; Gong, Q.; Safari, A.;
Yang, R.; Parise, J. B.; Li, J. J. Am. Chem. Soc. 2013, 135, 17401−17407.
(9) Zapf, P. J.; Haushalter, R. C.; Zubieta, J. Chem. Mater. 1997, 9,
2019−2024.
the structural diversity and catalytic potential of tungsten
oxide/organic hybrid materials by using other bipyridine
derivatives or related organonitrogen compounds as ligands
for the metal oxide substructure.
(10) Twu, J.; Fang, T.-H.; Hsu, C.-F.; Yu, Y.-Y.; Wang, G.-J.; Tang,
C.-W.; Chen, K.-H.; Lii, K.-H. J. Mater. Chem. 1998, 8, 2181−2184.
(11) Lu, Y.; Wang, E.; Yuan, M.; Li, Y.; Hu, C. J. Mol. Struct. 2003,
649, 191−195.
ASSOCIATED CONTENT
* Supporting Information
■
S
Detailed description of X-ray diffraction studies, Figures S1−
S11, Tables S1 and S2, and crystallographic data in CIF format
for compounds 1 and 2. The Supporting Information is
available free of charge on the ACS Publications website at
DOI: 10.1021/acs.inorgchem.5b00928. Crystallographic data
for compounds 1 and 2 reported in this Article have been
deposited at the Cambridge Crystallographic Data Centre as
CCDC 1054220 and 1054221, respectively. These data can be
obtained free of charge from the Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/conts/retrieving.html.
(12) Kim, J.; Lim, W. T.; Koo, B. K. Inorg. Chim. Acta 2007, 360,
2187−2191.
(13) Amarante, T. R.; Neves, P.; Valente, A. A.; Paz, F. A. A.; Fitch,
A. N.; Pillinger, M.; Gonca
4628.
̧ lves, I. S. Inorg. Chem. 2013, 52, 4618−
(14) Cui, C.-P.; Dai, J.-C.; Du, W.-X.; Fu, Z.-Y.; Hu, S.-M.; Wu, L.-
M.; Wu, X.-T. Polyhedron 2002, 21, 175−179.
(15) Zhou, Y.; Zhang, L.; Fun, H.-K.; You, X. Inorg. Chem. Commun.
2000, 3, 114−116.
(16) Amarante, T. R.; Neves, P.; Gomes, A. C.; Nolasco, M. M.;
Ribeiro-Claro, P.; Coelho, A. C.; Valente, A. A.; Paz, F. A. A.; Smeets,
S.; McCusker, L. B.; Pillinger, M.; Gonca
53, 2652−2665.
̧ lves, I. S. Inorg. Chem. 2014,
AUTHOR INFORMATION
Corresponding Authors
*A.A.V. e-mail: atav@ua.pt.
*F.A.A.P. e-mail: filipe.paz@ua.pt.
*I.S.G. e-mail: igoncalves@ua.pt.
■
(17) Zapf, P. J.; LaDuca, R. L., Jr.; Rarig, R. S., Jr.; Johnson, K. M., III;
Zubieta, J. Inorg. Chem. 1998, 37, 3411−3414.
(18) Chuang, J.; Ouellette, W.; Zubieta, J. Inorg. Chim. Acta 2008,
361, 2357−2364.
(19) Lysenko, A. B.; Senchyk, G. A.; Lincke, J.; Lassig, D.; Fokin, A.
̈
Notes
A.; Butova, E. D.; Schreiner, P. R.; Krautscheid, H.; Domasevitch, K. V.
Dalton Trans. 2010, 39, 4223−4231.
(20) Abrantes, M.; Amarante, T. R.; Antunes, M. M.; Gago, S.; Paz,
F. A. A.; Margiolaki, I.; Rodrigues, A. E.; Pillinger, M.; Valente, A. A.;
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge funding by FEDER (Fundo Europeu de
Desenvolvimento Regional) through COMPETE (Programa
Operacional Factores de Competitividade). National funding
Gonca
(21) Amarante, T. R.; Neves, P.; Paz, F. A. A.; Valente, A. A.;
Pillinger, M.; Goncalves, I. S. Dalton Trans. 2014, 43, 6059−6069.
(22) Abrantes, M.; Goncalves, I. S.; Pillinger, M.; Vurchio, C.;
̧ lves, I. S. Inorg. Chem. 2010, 49, 6865−6873.
̧
̧
through the FCT (Fundaca̧ o para a Ciencia e a Tecnologia)
̂
̃
Cordero, F. M.; Brandi, A. Tetrahedron Lett. 2011, 52, 7079.
(23) Udin, I.; Fox, M. R.; Martin, H.; Gainsford, G. J.; Kennedy, J.;
Markwitz, A.; Telfer, S. G.; Jameson, G. B.; Tallon, J. L. Chem.
Commun. 2010, 46, 4261−4263.
within the projects FCOMP-01-0124-FEDER-029779 (FCT ref
PTDC/QEQ-SUP/1906/2012) and FCOMP-01-0124-
FEDER-041282 (FCT ref EXPL/CTM-NAN/0013/2013) is
thanked. This work was developed in the scope of the project
CICECOAveiro Institute of Materials (FCT ref UID/CTM/
50011/2013), financed by national funds through the FCT/
MEC and cofinanced by FEDER under the PT2020 Partner-
ship Agreement. The FCT and the European Union are
acknowledged for postdoctoral grants to T.R.A. (SFRH/BPD/
97660/2013) and M.M.A. (SFRH/BPD/89068/2012) co-
funded by MCTES and the European Social Fund through
the program POPH of QREN. The FCT and CICECO are
acknowledged for financial support toward the purchase of the
single-crystal diffractometer. The authors are grateful to the
ESRF (Grenoble, France) for approving the experiment CH-
4254.
(24) Chong, S. V.; Tallon, J. L. J. Phys. Chem. Solids 2010, 71, 303−
308.
(25) Udin, I.; Chong, S. V.; Telfer, S. G.; Kennedy, J.; Jameson, G. B.;
Waterland, M. R.; Tallon, J. L. J. Phys. Chem. C 2012, 116, 3787−3792.
(26) Amini, M.; Haghdoost, M. M.; Bagherzadeh, M. Coord. Chem.
Rev. 2014, 268, 83−100.
(27) Jørgensen, K. A. Chem. Rev. 1989, 89, 431−458.
(28) Jimtaisong, A.; Luck, R. L. Inorg. Chem. 2006, 45, 10391−10402.
(29) Tang, J.; Wang, L.; Liu, G.; Liu, Y.; Hou, Y.; Zhang, W.; Jia, M.;
Thiel, W. R. J. Mol. Catal. A: Chem. 2009, 313, 31−37.
(30) Sozen-Aktas,̧ P.; Manoury, E.; Demirhan, F.; Poli, R. Eur. J.
̈
Inorg. Chem. 2013, 2728−2735.
(31) Amarante, T. R.; Neves, P.; Coelho, A. C.; Gago, S.; Valente, A.
A.; Paz, F. A. A.; Pillinger, M.; Gonca
29, 883−892.
̧ lves, I. S. Organometallics 2010,
REFERENCES
(32) Yan, B.; Xu, Y.; Goh, N. K.; Chia, L. S. Inorg. Chem. Commun.
2000, 3, 379−382.
■
(1) Hagrman, P. J.; Hagrman, D.; Zubieta, J. Angew. Chem., Int. Ed.
1999, 38, 2638−2684.
(33) Brandenburg, K. DIAMOND, Version 3.2f; Crystal Impact GbR:
Bonn, Germany, 1997−2010.
(2) Hagrman, D.; Hagrman, P. J.; Zubieta, J. Comments Inorg. Chem.
1999, 21, 225−261.
(34) Yang, S. H.; Li, G. B.; Tian, S. J.; Liao, F. H.; Lin, J. H. Cryst.
Growth Des. 2007, 7, 1246−1250.
(3) Hagrman, P. J.; Finn, R. C.; Zubieta, J. Solid State Sci. 2001, 3,
745−774.
(35) Baur, W. H. Acta Crystallogr., Sect. B 1974, 30, 1195−1215.
(36) Zhao, J.; Santos, A. M.; Herdtweck, E.; Kuhn, F. E. J. Mol. Catal.
̈
(4) Johnson, J. W.; Jacobson, A. J.; Rich, S. M.; Brody, J. F. J. Am.
Chem. Soc. 1981, 103, 5246−5247.
A: Chem. 2004, 222, 265−271.
(37) Maiti, S. K.; Dinda, S.; Banerjee, S.; Mukherjee, A. K.;
Bhattacharyya, R. Eur. J. Inorg. Chem. 2008, 2038−2051.
(38) Dinoi, C.; Ciclosi, M.; Manoury, E.; Maron, L.; Perrin, L.; Poli,
R. Chem.Eur. J. 2010, 16, 9572−9584.
(5) Hagrman, P. J.; LaDuca, R. L., Jr.; Koo, H.-J.; Rarig, R., Jr.;
Haushalter, R. C.; Whangbo, M.-H.; Zubieta, J. Inorg. Chem. 2000, 39,
4311−4317.
(6) Xu, Y.; Lu, J.; Goh, N. K. J. Mater. Chem. 1999, 9, 1599−1602.
(7) Yan, B.; Xu, Y.; Goh, N. K.; Chia, L. S. Chem. Commun. 2000,
2169−2170.
(39) Herbert, M.; Montilla, F.; Galindo, A. J. Mol. Catal. A: Chem.
2011, 338, 111−120.
M
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(40) Herbert, M.; Montilla, F.; Galindo, A.; Moyano, R.; Pastor, A.;
Alvarez, E. Dalton Trans. 2011, 40, 5210−5219.
(41) Afsharpour, M.; Mahjoub, A. R.; Amini, M. M. Appl. Catal. A:
Chem. 2007, 327, 205−210.
(42) Mikhalyova, E. A.; Makhlynets, O. V.; Palluccio, T. D.; Filatov,
A. S.; Rybak-Akimova, E. V. Chem. Commun. 2011, 48, 687−689.
(43) Saladino, R.; Ginnasi, M. C.; Collalto, D.; Bernini, R.; Crestini,
C. Adv. Synth. Catal. 2010, 352, 1284−1290.
(44) Altmann, P.; Kuhn, F. E. J. Organomet. Chem. 2009, 694, 4032−
̈
4035.
(45) Zhou, M.-D.; Jain, K. R.; Gunyar, A.; Baxter, P. N. W.;
̈
Herdtweck, E.; Kuhn, F. E. Eur. J. Inorg. Chem. 2009, 2907−2914.
̈
(46) Ciclosi, M.; Dinoi, C.; Gonsalvi, L.; Peruzzini, M.; Manoury, E.;
Poli, R. Organometallics 2008, 27, 2281−2286.
(47) Comas-Vives, A.; Lledos
́
, A.; Poli, R. Chem.Eur. J. 2010, 16,
2147−2158.
(48) Gamelas, C. A.; Gomes, A. C.; Bruno, S. M.; Paz, F. A. A.;
Valente, A. A.; Pillinger, M.; Romao, C. C.; Goncalves, I. S. Dalton
̧
̃
Trans. 2012, 41, 3474−3484.
(49) Schlemper, E. O.; Schrauzer, G. N.; Hughes, L. A. Polyhedron
1984, 3, 377−380.
(50) Herbert, M.; Alvarez, E.; Cole-Hamilton, D. J.; Montilla, F.;
Galindo, A. Chem. Commun. 2010, 46, 5933−5935.
(51) (a) Figueiredo, S.; Gomes, A. C.; Fernandes, J. A.; Paz, F. A. A.;
Lopes, A. D.; Lourenco
Organomet. Chem. 2013, 723, 56−64. (b) Amarante, T. R.; Paz, F. A.
A.; Gago, S.; Goncalves, I. S.; Pillinger, M.; Rodrigues, A. E.; Abrantes,
̧ , J. P.; Pillinger, M.; Gonca̧ lves, I. S. J.
̧
M. Molecules 2009, 14, 3610−3620. (c) Carreiro, E. D.; Yong-En, G.;
Burke, A. J. Inorg. Chim. Acta 2006, 359, 1519−1523. (d) Brito, J. A.;
́
Gomez, M.; Muller, G.; Teruel, H.; Clinet, J. C.; Dunach, E.; Maestro,
̃
M. A. Eur. J. Inorg. Chem. 2004, 4278−4285. (e) Hinner, M. J.;
Grosche, M.; Herdtweck, E.; Thiel, W. R. Z. Anorg. Allg. Chem. 2003,
629, 2251−2257. (f) Glas, H.; Spiegler, M.; Thiel, W. R. Eur. J. Inorg.
Chem. 1998, 275−281. (g) Thiel, W. R.; Priermeier, T. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 1737−1738. (h) Thiel, W. R.; Angstl, M.;
Priermeier, T. Chem. Ber. 1994, 127, 2373−2379. (i) Herrmann, W.
A.; Thiel, W. R.; Kuchler, J. G.; Behm, J.; Herdtweck, E. Chem. Ber.
1990, 123, 1963−1970.
N
DOI: 10.1021/acs.inorgchem.5b00928
Inorg. Chem. XXXX, XXX, XXX−XXX