Synthesis, characterization, and catalytic behavior of dioxomolybdenum complexes bearing AcAc-type ligands

Article abstract of DOI:10.1002/ejic.201201350

A series of [MoO₂(acac′)₂] [acac′ = acetylacetonato-type ligand: dibenzoylmethane (3), 1-benzoylacetone (4), bis(p-methoxybenzoyl)methane (5), 2-acetylcyclopentanone (6), 2-acetylcyclohexanone (7), and 2-acetyl-1-tetralone (8)] complexes have been synthesized in yields of 44-83 % by a simple synthetic method by using sodium molybdate and the desired acac-type ligand as starting materials. All the complexes were characterized by IR, UV/Vis, NMR, and high-resolution ESI-MS, and for compounds 3, 4, and 8, solid-state structures were obtained by X-ray diffraction. All the complexes contain a cis-dioxomolybdenum moiety, as proven by the characteristic Mo=O vibrations in the IR spectra and the occurrence of four sets of signals in the NMR spectra of the complexes bearing asymmetrical ligands (4 and 6-8), and confirmed by the solid-state structures. The complexes were found to be active as catalysts in the dehydration of 1-phenylethanol to styrene using technical-grade toluene as the solvent in air at 100 °C. The highest catalytic activity was found for [MoO₂{(tBuCO) ₂CH}₂] (2), followed by [MoO₂{(C ₆H₅CO)₂CH}₂] (3). Both complexes were also found to be active in the dehydration of other alcohols, including allylic, aliphatic, and homoallylic alcohols, as well as secondary and tertiary alcohols, with 2 generally showing better activity and selectivity than 3. These catalytic results were compared with those previously obtained with the metal-based catalyst Re₂O₇ and the benchmark acid catalyst H₂SO₄. The results were dependent on the substrate: By using 2, good selectivities but lower activities were generally obtained with tertiary alcohols, whereas good activities but lower selectivities were obtained with secondary alcohols. The industrially important dehydration of 2-octanol to octenes was very efficiently catalyzed by 2. Overall, the [MoO ₂(acac′)₂] complexes reported herein could offer a cheaper and more abundant metal-based catalyst alternative to the previously reported rhenium-based catalytic system for the dehydration reaction. Copyright

Full text of DOI:10.1002/ejic.201201350

FULL PAPER
DOI:10.1002/ejic.201201350
Synthesis, Characterization, and Catalytic Behavior of
Dioxomolybdenum Complexes Bearing AcAc-Type
Ligands
[
a]
[a]
[b]
Ties J. Korstanje, Emma Folkertsma, Martin Lutz,
[a]
[a]
Johann T. B. H. Jastrzebski, and Robertus J. M. Klein Gebbink*
Keywords: Sustainable chemistry / Dehydration / Alcohols / Molybdenum / Alkenes / O ligands
A series of [MoO
2
(acacЈ)
2
] [acacЈ = acetylacetonato-type li-
for [MoO {(tBuCO)
2
2 2 2 6 5
CH} ] (2), followed by [MoO {(C H -
gand: dibenzoylmethane (3), 1-benzoylacetone (4), bis(p-
methoxybenzoyl)methane (5), 2-acetylcyclopentanone (6), 2-
acetylcyclohexanone (7), and 2-acetyl-1-tetralone (8)] com-
plexes have been synthesized in yields of 44–83% by a sim-
ple synthetic method by using sodium molybdate and the de-
CO) CH} ] (3). Both complexes were also found to be active
2
2
in the dehydration of other alcohols, including allylic, ali-
phatic, and homoallylic alcohols, as well as secondary and
tertiary alcohols, with 2 generally showing better activity and
selectivity than 3. These catalytic results were compared with
sired acac-type ligand as starting materials. All the com- those previously obtained with the metal-based catalyst
plexes were characterized by IR, UV/Vis, NMR, and high-
resolution ESI-MS, and for compounds 3, 4, and 8, solid-state
structures were obtained by X-ray diffraction. All the com-
plexes contain a cis-dioxomolybdenum moiety, as proven by
the characteristic Mo=O vibrations in the IR spectra and the
occurrence of four sets of signals in the NMR spectra of the
complexes bearing asymmetrical ligands (4 and 6–8), and
confirmed by the solid-state structures. The complexes were
found to be active as catalysts in the dehydration of 1-phen-
ylethanol to styrene using technical-grade toluene as the sol-
vent in air at 100 °C. The highest catalytic activity was found
2 7 2 4
Re O and the benchmark acid catalyst H SO . The results
were dependent on the substrate: By using 2, good selectivi-
ties but lower activities were generally obtained with tertiary
alcohols, whereas good activities but lower selectivities were
obtained with secondary alcohols. The industrially important
dehydration of 2-octanol to octenes was very efficiently cata-
2 2
lyzed by 2. Overall, the [MoO (acacЈ) ] complexes reported
herein could offer a cheaper and more abundant metal-based
catalyst alternative to the previously reported rhenium-based
catalytic system for the dehydration reaction.
Introduction
lulosic biomass to be a convenient resource for the chemical
industry, part of the functionality needs to be removed or
converted and its oxygen content needs to be lowered. One
of the methods for accomplishing this is by dehydration of
the hydroxy groups to yield olefinic moieties (Scheme 1).
As a result of the increasing scarcity of fossil resources,
currently the main provider of building blocks for the chem-
ical industry, considerable research effort is being focused
on the use of biomass as an alternative source of chemical
building blocks. Lignocellulosic biomass, in particular,
could provide sufficient sustainable carbon-based material
[
1,2]
on an annual basis.
The major challenge in the substitu-
tion of fossil resources by lignocellulosic biomass lies in the
structural differences between the two: Fossil resources con-
tain mostly underfunctionalized, oxygen-poor scaffolds,
whereas lignocellulosic biomass is highly functionalized and strong (solid) acid catalysts, such as mineral acids, zeo-
oxygen-rich in the form of hydroxy groups. For lignocel- lites, or metal oxides.
Scheme 1. Dehydration of alcohols to olefins.
Commonly, the dehydration reaction is performed with
[3]
[
4]
[5,6]
The major drawback of these
methods is their low selectivity and low functional group
tolerance as well as their acidic nature, which causes reactor
corrosion and safety concerns. More selective catalysts for
the dehydration of alcohols under mild conditions are
therefore desirable.
[
a] Organic Chemistry & Catalysis, Debye Institute for
Nanomaterials Science, Utrecht University,
Universiteitsweg 99, 3584 CG Utrecht, The Netherlands
Fax: +31-30-2523615
E-mail: r.j.m.kleingebbink@uu.nl
Homepage: http://www.uu.nl/science/occ
[
b] Crystal and Structural Chemistry, Bijvoet Center for
Biomolecular Research, Utrecht University,
Padualaan 8, 3584 CH Utrecht, The Netherlands
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201201350.
Few transition-metal-catalyzed dehydration reactions
have been reported; and those that have are based on the
[
7]
[8]
[9–13]
use of zinc, ruthenium, or rhenium.
We have pre-
viously reported on the use of rhenium-based catalysts in
Eur. J. Inorg. Chem. 2013, 2195–2204
2195
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
the dehydration reaction; we found high-valence rhenium oxide also showed little activity (entry 3), although rheni-
catalysts to be highly active as well as selective in the dehy- um(VII) oxide had performed very well in this reaction.
dration of various alcohols.[
14,15]
Comparison with tradi- Molybdenum(VI) dichloride dioxide, in contrast, showed
tional (solid) acid catalysts such as sulfuric acid, acid resins, good activity, giving almost complete conversion after 24 h
and zeolites showed that rhenium(VII) oxide, the most and a high initial rate of 5.43 mmolh–1 but with a poor
active rhenium-based catalyst, is more active and more se- selectivity for styrene (entry 4), and accordingly a poor
lective than any of the acidic catalysts tested. However, the mass balance. This is likely caused by oligo- or polymeriza-
limited natural abundance of rhenium could restrict the use tion of the styrene formed under the given reaction condi-
of rhenium-based catalysts. Molybdenum, like rhenium, is tions. [MoO (acac) ] (1) and [MoO {(tBuCO) CH} ] (2),
2
2
2
2
2
located in the middle of the transition-metal block of the both [MoO (acacЈ) ] complexes, showed similar activity, but
2
2
Periodic Table and as such is active in oxygen atom transfer somewhat improved selectivity for styrene (entries 5 and 6).
OAT) reactions such as oxidation and deoxygenation. In In comparison with sulfuric acid (entry 7), the two molybd-
addition, [MoO (acac) ] has been reported as a catalyst for enum complexes bearing acac-type ligands show higher ac-
(
2
2
[16]
the dehydration of tertiary alcohols.
In nature, molybd- tivity with a 2.5-fold lower catalyst loading, although the
enum-containing enzymes, usually bound by the pterin co- styrene selectivity was somewhat higher with sulfuric acid.
factor, are known to be active in OAT reactions, for exam- The trend in activity of the tested compounds appears to
ple, in dmso reductase and sulfite oxidase.[ In this paper follow the previously observed trend in rhenium-catalyzed
we report on our investigations into the use of molybd- dehydration, which is linked to the Lewis acidity of the cat-
17]
[
15]
enum-based catalysts for the dehydration of alcohols to ole- alyst. Molybdenum metal is a poor Lewis acid, as is the
fins.
anionic molybdate salt, and both show low activity. The
high-valence molybdenum complexes are stronger Lewis ac-
ids and show good activity. The only complex that falls out
of this trend is molybdenum(VI) oxide, which is a strong
Lewis acid, but shows low activity.
Results and Discussion
Catalysis Ϫ Commercially Available Complexes
Initially, our investigation of molybdenum-based cata-
lysts was focused on the dehydration of 1-phenylethanol to
styrene using commercially available molybdenum com-
plexes (Table 1). Screening for solvent, temperature, and the
Synthesis and Characterization of the [MoO (acacЈ) ]
Complexes
2
2
Because the two [MoO (acacЈ) ] complexes 1 and 2 were
2
2
presence or absence of air with [MoO (acac) ] (1) as catalyst the best performing catalysts tested in our initial dehy-
2
2
revealed the following optimal reaction conditions: Techni- dration tests, we decided to adapt the acac-type ligand to
cal-grade toluene as the solvent, 100 °C as the reaction tem- optimize the catalytic behavior of these complexes.
perature, and an ambient atmosphere (1 atm. air), which is
A simple synthetic method adapted from previously re-
using sodium molybdate as the metal
precursor and the acetylacetonate preligand in stoichiomet-
[
18,19]
similar to the optimal conditions found earlier by us for the ported methods
rhenium-based catalysts.[
14]
By using these conditions, we tested various commer- ric amounts in an acidic water/ethanol mixture yielded two
[
19]
cially available molybdenum materials in the dehydration previously reported (3 and 4)
and four novel (5–8)
reaction. Molybdenum metal (Table 1, entry 1) showed little [MoO (acacЈ) ] complexes in moderate-to-good yields (Fig-
2
2
activity in the dehydration of 1-phenylethanol, as did so- ure 1). An alternative method involving halide abstraction
dium molybdate (entry 2). Surprisingly, molybdenum(VI) from MoO Cl with the help of Ag(OTf) and subsequent
2
2
2
Table 1. Dehydration of 1-phenylethanol to styrene catalyzed by commercially available molybdenum complexes.[a]
[b,c]
1
Entry
Catalyst
Mo0
Catalyst loading Time
Conversion[b] Styrene selectivity[b] Turnover number[b]
Initial rate
[mmolh ]
–
[mol-%]
[h]
[%]
[%]
(TON)
1
2
3
4
5
6
7
5
5
5
1
1
22
22
22
24
24
22
24
23
12
35
8
5
2
Ͻ0.1
Ͻ0.1
Ͻ0.1
5.43
5.16
6.77
3.65
2 4
Na MoO
MoO
MoO
3
32
96
97
Ͼ99
Ͼ99
10
21
30
28
39
6
2
Cl
2
96
97
99
25
[MoO
[MoO
2
(acac)
2
] (1)
CH}
2
{(tBuCO)
2
2
] (2)
1
2.5
H SO
2 4
[
1
a] Reaction conditions: 2 mmol 1-phenylethanol, 0.02–0.1 mmol catalyst, 250 μL pentadecane (as internal standard), 10 mL toluene,
00 °C. [b] Based on GC. [c] Rate of consumption of starting material during the first 5 min of reaction.
Eur. J. Inorg. Chem. 2013, 2195–2204
2196
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
2 2
Figure 1. [MoO (acacЈ) ] complexes 1–8.
–1
addition of the desired ligand did not yield the desired com- symmetrical vibration at around 927 cm . The very small
pounds. difference between the Mo=O vibrational energies of 3 and
All the isolated complexes show a very strong doublet in 5 is surprising as the p-methoxy groups on the phenyl rings
–
1
the IR spectrum in the 895–940 cm region (Table 2), of 5 have a strong electron-donating effect, which is ex-
which can be attributed to the symmetrical and anti-sym- pected to influence the electron density on the metal center
metrical vibrations of the MoO moiety. The observed fre- and therefore the Mo=O bond energy. The similarity be-
2
quency range is indicative of the cis conformation of the tween these two complexes indicates that the π systems of
MoO moiety; the trans conformation would be lower in the phenyl rings do not influence the binding to the metal,
2
–
1
energy (790–820 cm ) due to less overlap of the lone pairs likely because it is not delocalized onto the enolate π sys-
of the oxo ligand with the d orbitals of the metal, which tem.
reduces the π donation of the oxo ligands and thus lowers
Considerable differences in the maximum absorption en-
[20]
the Mo=O bond order.
The cis conformation is most ergies are observed in the UV/Vis spectra of the complexes,
0
[19]
common for dioxo d metal complexes of this type. Con- varying from 289 nm for compound 7 to 380 nm for com-
siderable differences in the Mo=O vibrational energies of pound 5. These transitions can be attributed to π–π* transi-
the complexes are observed, ranging from 893 to 907 cm–1 tions in the enolate ligands. Furthermore, the colors of the
for the anti-symmetrical Mo=O vibration and from 927 to isolated complexes are different, ranging from yellow (2) to
–
1
940 cm for the symmetrical Mo=O vibration. The anti- dark blue (6). The two complexes bearing cycloalkyl deriva-
symmetrical and symmetrical Mo=O vibrations in these tives (6 and 7) both show an additional absorption at 777
complexes are correlated, with the exception of complex 7, and 778 nm, respectively. This absorption can be attributed
which has the lowest anti-symmetrical but the highest sym- to the presence of small amounts of polyoxomolybdates,
metrical Mo=O vibration energy. The IR spectra of the four also known as molybdenum blue, which are known to have
complexes bearing an aromatic moiety in the ligand back- strong absorptions in this region due to an intervalence
[
21]
bone (3–5 and 8) show similar characteristics with an anti- charge-transfer (IVCT) transition.
Therefore it is likely
–
1
symmetrical Mo=O vibration at around 898 cm and a that the inherent colors of 6 and 7 are not blue and green,
respectively, but yellow, similar to the color of the other
complexes. We also observed the rapid decomposition of
Table 2. IR and UV/Vis data of the [MoO
sorted by the symmetrical Mo=O vibration.
2 2
(acacЈ) ] complexes,
complexes 6 and 7 by NMR spectroscopy (see below),
which explains the occurrence of these species and thus
their apparent color.
Infrared
antisym Mo=O sym Mo=O
UV/Vis
max [nm] (ε [Lcm–1 mol–1])
λ
–
1
–1
When comparing the trends in the symmetrical Mo=O
vibrations and the UV/Vis absorption maxima, it can be
observed that the Mo=O vibration energy increases with
increasing UV/Vis absorption maximum. This trend can be
rationalized as follows: A low-energy absorption corre-
sponds to a small energy difference between the π and π*
orbitals of the ligand as a result of an elevated orbital en-
ergy of the HOMO in the case of an electron-rich system.
[
cm ]
[cm ]
3
7
2
6
1
3
4
5
8
893
907
901
902
898
899
899
897
940
936
935
932
929
927
927
927
289 (1.65ϫ10 ), 778 (329)
322 (6.71ϫ10 )
332 (4.38ϫ10 ), 777 (264)
322 (6.84ϫ10 )
376 (2.97ϫ10 ), 346 (2.01ϫ10 )
346 (4.73ϫ10 )
380 (4.08ϫ10 ), 363 (4.73ϫ10 )
376 (1.47ϫ10 )
3
5
3
4
4
4
4
4
4
Eur. J. Inorg. Chem. 2013, 2195–2204
2197
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Likewise, an electron-rich ligand system would donate more a result of the moderate solubility of the complexes in or-
electron density to the empty d orbitals of the metal center, ganic solvents, combined with the good solubility of the
thereby lowering the π donation of the lone pairs of the oxo free ligand, 13C NMR spectroscopic data were only success-
ligands to the d orbitals of the metal, thereby resulting in a fully obtained for complexes 3 and 4. For complexes 5–7,
lower bond order and thus a lower vibrational energy.
In the case of the complexes bearing unsymmetrical li-
gands, that is, 4 and 6–8, three different geometrical isomers
incorporating a cis-MoO moiety can be formed, each of
2
which has an enantiomer (Figure 2). This is illustrated by
the NMR spectra of these compounds, which contain four
sets of signals (one set for OC6–22 and OC6–33 each, and
another two sets for OC6–23[ due to the nonequivalence
of the ligands in this conformation). The complexes bearing
symmetrical ligands (3 and 5) have only one geometrical
isomer, as illustrated by a single signal for the central CH
proton. For the aryl protons of 3 and 5, a double set of
signals is observed due to the nonequivalence of the two
aryl rings in the dibenzoylmethane ligands. During the
NMR measurements, some decomposition of the com-
22]
plexes was observed, hampering the acquisition of 13
C
NMR spectra. Various solvents were tested to minimize the
decomposition during the measurement, with dry, oxygen-
Figure 2. Three different diastereoisomers of 4: OC6–22, OC6–23,
free [D ]thf proving to be the optimal solvent. However, as and OC6–33 (only the Δ enantiomers are shown).[
22]
8
Figure 3. X-ray crystal structures of 3 (left), 4 (middle), and 8 (right). Ellipsoids are drawn at the 50% probability level. Hydrogen atoms
have been omitted for clarity.
Table 3. Details of the crystal structure data collection for 3, 4, and 8.
3
4
8
Formula
C
574.42
yellow
0.33ϫ0.33ϫ0.24
110(2)
30
H
22MoO
6
C
450.28
yellow
0.30ϫ0.09ϫ0.06
150(2)
20
H
18MoO
6
C
502.36
yellow
0.16ϫ0.13ϫ0.03
150(2)
24 6
H22MoO
M
r
Crystal color
Crystal size [mm ]
T [K]
3
Crystal system
Space group
a [Å]
b [Å]
c [Å]
monoclinic
monoclinic
P2 /c (no. 14)
1
monoclinic
Cc (no. 9)
10.6197(8)
27.201(2)
7.6064(6)
113.841(2)
2009.8(3)
4
P2
1
/c (no. 14)
9.31711(8)
13.26512(16)
20.2515(8)
96.358(1)
2487.54(10)
4
9.8489(3)
12.4838(3)
16.3720(4)
116.188(1)
1806.34(9)
4
β [°] 3
V [Å ]
Z
D
–3
calcd. [gcm ]
1.534
0.65
1.656
0.65
1.660
0.65
–
1
(sinθ/λmax) [Å ]
Reflections measured/unique
Parameters/restraints
71874/5708
334/0
0.0209/0.0541
0.0271/0.0582
–
37210/4154
246/0
0.0271/0.0590
0.0419/0.0654
–
18504/4149
283/2
0.0194/0.0472
0.0207/0.0477
0.15(2)
R
R
1
/wR
/wR
2
[IϾ2σ(I)]
1
2
[2a7l]l reflections]
[
Flack x
S
1.066
–0.30/0.57
1.023
–0.49/0.59
1.081
–0.24/0.37
–3
ρ(min/max) [eÅ ]
Eur. J. Inorg. Chem. 2013, 2195–2204
2198
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
significant amounts of protonated ligand were also ob- Table 5. Dehydration of 1-phenylethanol to styrene catalyzed by
[a]
1
2 2
[MoO (acacЈ) ].
served in their H NMR spectra. Together with the occur-
rence of the protonated ligand, we also observed a more
Complex Time Conversion Styrene selec-
Initial rate
[mmolh ]
[h]
[%]/TON[
b]
tivity [%]
[b]
–1 [b,c]
symmetrical molybdenum species, likely due to
a
[
MoO (acacЈ)] species. As a result of the observed sensitiv-
2
3
4
5
6
7
8
24
24
24
20
20
24
98
99
Ͼ99
98
Ͼ99
98
30
29
34
35
30
30
5.43
3.89
0.73
2.32
3.77
1.21
ity towards decomposition for some of the complexes, ana-
lytically pure compounds were difficult to obtain, although
we succeeded in obtaining X-ray crystal structures for three
of the complexes.
We determined the solid-state structures of compounds
, 4, and 8 by X-ray crystal structure analysis (Figure 3). A
[
a] Reaction conditions: 2 mmol 1-phenylethanol, 0.02 mmol cata-
3
lyst, 250 μL pentadecane (used as internal standard), 10 mL tolu-
summary of the crystallographic data are given in Table 3
ene, 100 °C. [b] Based on GC. [c] Rate of consumption of starting
and selected bond lengths and angles are given in Table 4. material during the first 5 min of reaction.
In all cases a distorted octahedral coordination around the
molybdenum is observed, which is in accordance with the
previously determined solid-state structures of similar com-
pounds, including 1,[
23,24]
2, and 3. Although the crys-
[25]
[26]
All six complexes showed significant catalytic activity in
tal structure of 3 was reported earlier, the structure pre- the dehydration of 1-phenylethanol to styrene, reaching
sented here is of much higher quality. In the measured crys- nearly complete conversion within 20–24 h, similar to the
tals, only one isomer was found, OC6–33 for both 4 and 8. results described above for 1 and 2. The observed selectivity
The solid-state structures confirmed that all the complexes for styrene was also very similar, that is, between 29 and
contain a cis-MoO2 moiety, as was shown by IR spec- 35%, with complexes 5 and 6 being the most selective. In
troscopy. The Mo1–O5 and Mo1–O6 bond lengths terms of activity, a larger difference is observed: Complex
–1
[
1.6984(16)–1.7057(12) Å] show the presence of a double 3 showed an initial rate of 5.43 mmolh , which is similar
bond and thus confirm the presence of the oxo groups. The to 1, whereas 5 showed an almost eight-fold lower activity.
two oxygen atoms of the acac-type ligand are not equiva- Complexes 4 and 7 showed somewhat lower activity than
lent, as shown by the significant differences in the Mo–O 3, but still higher than the benchmark catalyst sulfuric acid.
bond lengths, with the longer Mo–O bond trans to the oxo On the basis of these results, it was decided to use the two
ligand. This is most likely caused by the strong trans influ- most active catalysts (2 and 3) to explore the substrate
ence of the oxo ligands, which weakens and thus lengthens scope in the dehydration reaction.
the C–O bond trans to the oxo ligands.
A wide variety of alcohols, ranging from allylic, ali-
phatic, and homoallylic alcohols as well as secondary and
tertiary alcohols, were tested in the dehydration reaction
with 2 or 3 as the catalyst (Table 6). The tertiary allylic and
aliphatic alcohols and both the secondary and tertiary all-
ylic alcohol reacted readily at 100 °C, with both catalysts
showing quite similar catalytic performance. In the case of
Table 4. Selected bond lengths and angles for 3, 4, and 8.
Bond lengths [Å]
3
4
8
Mo1–O1
Mo1–O2
Mo1–O3
Mo1–O4
Mo1–O5
Mo1–O6
2.1661(11)
1.9847(11)
1.9981(10)
2.1496(11)
1.7057(12)
1.6991(11)
2.2049(16)
1.9847(15)
2.0052(14)
2.1653(16)
1.6984(16)
1.7012(16)
2.1905(16)
1.9873(15)
2.0094(15)
2.1463(16) the tertiary allylic alcohol 1-vinylcyclohexanol, good results
1.7025(16) were obtained with full conversion and a good selectivity
1.7036(16)
for the desired olefin after 24 h with 2 as catalyst and a
Bond angles [°]
moderate selectivity with 3 as catalyst (entry 1). The tertiary
aliphatic alcohol 3-ethylpentan-3-ol, however, showed a
much lower conversion and initial rate with both catalysts,
but again a good selectivity for 3-ethylpent-2-ene with 2 as
catalyst and a moderate selectivity with 3 as catalyst (en-
try 2). In the case of the secondary allylic alcohol 1-octen-
3-ol, no dehydration products were observed. Instead a 1,3-
transposition to 2-octen-1-ol was observed (entry 3). After
O1–Mo1–O2
O2–Mo1–O6
O6–Mo1–O3
O3–Mo1–O1
O1–Mo1–O5
O5–Mo1–O6
O6–Mo1–O4
O4–Mo1–O1
O1–Mo1–O6
O4–Mo1–O5
81.06(4)
93.38(5)
99.17(5)
83.13(4)
89.14(5)
104.69(6)
91.17(5)
75.25(4)
165.72(5)
163.85(5)
81.35(6)
94.08(7)
98.44(7)
83.46(6)
86.83(7)
104.06(8)
92.17(7)
77.22(6)
168.82(7)
163.04(7)
78.82(6)
93.31(7)
99.91(7)
85.03(6)
87.20(7)
103.90(8)
93.97(7)
75.67(6)
167.54(7)
161.24(7)
1
2
5–60 min, selectivities of 90 and 79% were obtained with
2 and 25% conversion with 2 and 3, respectively, as cata-
lyst. When the reaction time was increased, the conversion
was higher, but the selectivity lower. The cyclic secondary
allylic alcohol 2-cyclohexen-1-ol, however, did undergo de-
hydration to 1,3-cyclohexadiene with high conversion and
activity with both catalysts but with low selectivity (en-
Catalysis with [MoO (acacЈ) ]
2
2
After successfully synthesizing complexes 3–8, they were try 4). In all cases, no other products were observed by GC,
tested as catalysts in the dehydration reaction of 1-phenyl- which indicates the formation of higher-boiling products re-
ethanol to styrene (Table 5).
sulting from oligomerization of the olefins.
Eur. J. Inorg. Chem. 2013, 2195–2204
2199
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Table 6. Dehydration of allylic, aliphatic, and homoallylic alcohols catalyzed by 2 or 3.[a]
[
2
a] Reaction conditions: 2 mmol substrate, 0.02 mmol 2 or 3, 250 μL pentadecane (used as internal standard), 10 mL toluene, 100–150 °C,
4 h, all values averaged over two runs. [b] Based on GC. [c] Rate of consumption of starting material during the first 5 min of reaction.
[
d] Determined by NMR spectroscopy using 1,3,5-trimethoxybenzene as internal standard. [e] Not determined. [f] Mixture of isomers,
quantified after hydrogenation using 10 wt.% Pd on activated carbon. [g] For catalyst 3: mixture of 3- and 4-octenes were obtained.
An interesting substrate was 3-methyl-5-hexen-3-ol (en- process could be industrially interesting for the production
try 5). This tertiary homoallylic alcohol did not show any of bio-based 1-octene, which is used as a co-polymer for the
reaction at 100 °C with 2, but with 3, some conversion was production of linear low-density polyethylene (LLDPE).[
29]
obtained. With 2, the reaction did proceed at 150 °C with
good conversion, but in both cases the reaction products dration of these alcohols shows that 2 is the superior cata-
could not be identified by GC. lyst, especially at the higher temperatures required for the
A comparison between 2 and 3 as catalysts in the dehy-
The secondary aliphatic alcohols did not show any reac- secondary alcohols to react. An explanation for this differ-
tion at 100 °C with either catalyst, but upon increasing the ence can be found in the stability of the complexes, with 3
temperature to 150 °C and by carrying out the reaction in decomposing more rapidly at 150 °C than 2, thus giving
an autoclave, these alcohols did convert. The secondary lower conversions.
homoallylic alcohols 1-octen-4-ol and 1,6-heptadien-4-ol
gave olefinic products, although the conversions were low (1 mol-%) as catalysts with those previously obtained with
By comparing the results obtained here with 2 and 3
[
15]
with both catalysts after 24 h (entries 6 and 7). In the first both Re O (0.5 mol-%) and H SO (2.5 mol-%),
some
2
7
2
4
case, a low selectivity for octadienes (10–14%) was obtained differences in reactivity can be observed. For tertiary
with both catalysts, but in the latter case good selectivities alcohol substrates (entries 1 and 2), a lower activity (using
of 94–99% for heptatrienes were observed. In the case of the initial rate as the measure) is observed in the molybd-
the secondary aliphatic alcohol 2-octanol, poor conversion enum-catalyzed dehydration reactions, whereas with sec-
was obtained with 3 as catalyst, but very good results were ondary alcohols (entries 3 and 4), higher rates than with
obtained with 2 as catalyst, achieving 88% conversion after H SO , yet lower than with Re O are observed. Notwith-
2
4
2
7
2
4 h and in total 92% selectivity for octenes (entry 8). As 2- standing the higher activities observed for the molybdenum
[28]
octanol can be obtained from castor oil, this dehydration catalyst compared with H SO , the conversions are much
2
4
Eur. J. Inorg. Chem. 2013, 2195–2204
2200
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
lower. Remarkably, the tertiary homoallylic alcohol 3- pelled as acetylacetone. This step could provide an explana-
methyl-5-hexen-3-ol (entry 5) does not react with the tion for the high activities of 2 and 3, as these complexes
molybdenum catalyst until 150 °C, whereas with both carry the most bulky ligands, which facilitates the ligand-
Re O and H SO , the reaction is observed at 100 °C. For exchange step. Next, the olefin product is expelled from
2
7
2
4
secondary homoallylic alcohol substrates (entries 6 and 7), the molybdenum center with the formation of a hydroxy-
conversions are again lower with the molybdenum catalyst, molybdenum species. From this species, two possible path-
but in the case of 2-octanol (entry 8), the conversion after ways can occur. One is the direct coordination of another
2
4 h is comparable to that of Re O and H SO .
alcohol molecule to reproduce the alkoxymolybdenum in-
2
7
2
4
In terms of product selectivity, significant differences are termediate following removal of water. The other possibility
also observed. For the tertiary allylic alcohol 1-vinylcy- is coordination of acetylacetone to reproduce the original
clohexanol (entry 1) and the secondary allylic alcohol 1- [MoO (acacЈ) ] complex, which also involves the expulsion
2
2
octen-3-ol, the selectivities with the molybdenum catalyst of water. In any of these ligand-exchange steps, the catalyst
are in between those obtained with Re O7 and H SO , could be prone to decomposition due to oxidation, dimer-
2
2
4
whereas with the aliphatic tertiary alcohol 3-ethyl-3-pent- ization, or oligomerization, as described above, thereby re-
anol (entry 2) and the secondary allylic alcohol 2-cy- ducing the overall catalytic activity.
clohexen-1-ol (entry 4), the selectivities are lower. On the
other hand, for secondary homoallylic alcohols (entries 6
and 7), mixed results were obtained: For 1-octen-4-ol, the
selectivity is much lower, whereas the dehydration of 1,6-
heptadien-4-ol results in excellent selectivity, and much
higher than with Re O or H SO . In the industrially most
2
7
2
4
interesting case, 2-octanol, the molybdenum catalyst shows
good results in terms of selectivity, matching that of Re O7
2
and surpassing H SO .
2
4
Following these observations, an explanation for the
poor activity observed with tertiary alcohols might be
found in steric factors, as tertiary alcohols are sterically
quite hindered and the molybdenum catalysts are also quite
demanding in a steric sense due to the presence of four tert-
butyl (2) or phenyl (3) groups. This could hamper substrate
approach and binding and, accordingly, the overall catalytic
activity. Furthermore, the good initial rates in combination
with the moderate conversions with some substrates suggest
the decomposition of the catalyst under the reaction condi-
[
30]
tions. It is known that 1 can thermally decompose.
One
of the proposed structures of this decomposition product is
V
a [Mo O (acac) ] species, based on EPR measurements. In
2
3
4
2 2
Scheme 2. Proposed catalytic cycle for the [MoO (acacЈ) ]-cata-
these measurements, a free radical, possibly an acetyl- lyzed alcohol-to-olefin dehydration reaction.
acetone radical, is also observed.[ Another proposed de-
31]
composition pathway could involve a photochemical reac-
[
32]
V
tion, yielding a [Mo O (acacЈ)] species, also by a radical
2
pathway. Finally, any type of polyoxomolybdate could be
formed by oxidation under the reaction conditions as these
are known to give an intense blue color.[ During the dehy-
Conclusions
21]
We have reported herein the synthesis of a series of
dration reactions we observed that all tested [MoO2- [MoO (acacЈ) ] complexes and their characterization by
2
2
(
acacЈ) ] complexes change color, in the case of 1 from yel- using various spectroscopic techniques (IR, UV/Vis,
2
low to green, and in the case of 2 from green to dark blue, NMR), high-resolution ESI-MS, and, for complexes 3, 4,
which indicates the transformation or decomposition of the and 8, X-ray diffraction. These complexes were used in the
complexes. This might be the cause for the moderate con- dehydration of 1-phenylethanol to styrene alongside a
versions in some cases, although good conversions were ob- number of commercially available molybdenum materials
tained in other cases.
and were found to be active in this reaction. The most
Based on the observations described herein we have pro- active catalysts tested in this study were [MoO {(tBuCO) -
2
2
posed a catalytic cycle for the dehydration reaction CH} ] (2) and [MoO {(C H CO) CH} ] (3), which accord-
2
2
6
5
2
2
(Scheme 2). As the molybdenum starting complex ingly were tested in the dehydration of a series of alcohol
[
MoO (acacЈ) ] is coordinatively saturated, one of the acac- substrates, varying from allylic to aliphatic and homoallylic
2
2
type ligands has to be expelled to allow for coordination of as well as both secondary and tertiary alcohols. The cata-
the alcohol. In this ligand exchange, the alcohol hydroxy lytic performances of both catalysts were compared on the
proton is transferred to the acac-type ligand, which is ex- basis of activity and selectivity with the results obtained
Eur. J. Inorg. Chem. 2013, 2195–2204
2201
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
[
15]
4
4
–1
–1
previously with Re O or H SO as catalyst
and were (2.97ϫ10 ), 346 nm (2.01ϫ10 Lmol cm ). MS (ESI): calcd. for
2
7
2
4
+
[
M – O]+ 560.0529; found 560.0577; calcd. for [M + H] 577.0556;
found to give moderate results with some substrates,
whereas with other substrates, such as the dehydration of
-octanol to octenes using 2, comparable results to Re O7
+
found 577.0548; calcd. for [M + Na] 599.0375; found 599.0496;
calcd. for [M + MeCN + Na]+ 640.0641; found 640.0761.
2
2
were obtained. Amongst other dehydrative transformations,
2 6 5 3 2
[MoO (C H COCHCOCH ) ] (4): The general synthetic procedure
this opens up the possibility of using a cheaper and more was followed using 1-benzoylacetone (2.58 g, 5.9 mmol) as the li-
abundant metal, such as molybdenum, for the industrially gand. The desired product was obtained as a light-green solid
important dehydration of 2-octanol to octenes without sac- (81%, 2.90 g, 6.42 mmol) as three different conformational isomers.
Single crystals suitable for X-ray crystallography were obtained by
rificing selectivity.
slow evaporation of a dichloromethane solution at room tempera-
1
ture. H NMR ([D
8
]thf): δ = 2.17, 2.24, 2.26, 2.29 (s, total 6 H,
3
3
CH ), 6.57, 6.63, 6.66, 6.67 (s, total 2 H, CH), 7.44 (t, J = 7.7 Hz,
H, m-ArH), 7.46–7.53 (m, 4 H, m-ArH, p-ArH), 7.89 (t, J =
.6 Hz, 2 H, o-ArH), 7.98 (t, J = 6.6 Hz, 2 H, o-ArH) ppm. The
assignments were confirmed by COSY NMR spectroscopy.
3
2
7
Experimental Section
3
1
3
General: Bis(p-methoxybenzoyl)methane was prepared by a litera-
C
[
33]
ture procedure. All other starting materials were obtained from
commercial sources and used without further purification. Dry di-
ethyl ether and thf were obtained from an MBraun MB SPS-800
8 3
NMR ([D ]thf): δ = 25.7, 26.0, 27.1, 28.2 (CH ), 99.7, 100.7, 101.2,
101.3 (CH), 127.6–129.3 (o-ArC, m-ArC overlapping), 132.9, 133.1,
133.7, 133.8 (p-ArC), 136.4, 136.6, 137.1, 137.2 (ipso-ArC), 178.0,
187.0, 187.1, 187.2 187.9, 188.0, 188.1, 198.0 (Ar-CO,
solvent purification system, [D
sodium/benzophenone and stored over 4 Å molecular sieves.
and C NMR spectra were recorded at 298 K with a Varian AS
00 MHz NMR spectrometer at 400 and 100 MHz, respectively. (m), 767 (s), 700 (vs), 683 (vs) cm . UV/Vis (thf): λ (ε) = 346.0
Chemical shifts are reported in ppm and referenced against the
residual solvent signal. IR spectra were recorded with a Perkin–
Elmer Spectrum One FT-IR spectrometer operated in ATR mode.
UV/Vis spectra were recorded with a Varian Cary50 Scan UV/Vis
spectrometer. GC analysis was performed with a Perkin–Elmer Au-
tosystem XL Gas Chromatograph equipped with an Elite-17 col-
umn (30 mϫ 0.25 mmϫ 0.250 μm) and a flame ionization detector.
GC–MS analysis was performed on a Perkin–Elmer Autosystem
XL Gas Chromatograph equipped with an AT-50 column (30mϫ
8
]thf was dried by distillation from
1
H
COCH ) ppm. IR: ν˜ = 1592 (m), 1574 (m), 1497 (vs), 1480 (vs),
3
1
3
1450 (s), 1357 (vs), 1283 (vs), 1103 (m), 927 (vs), 899 (vs), 778
–
1
4
4
–1
–1
+
(4.73ϫ10 Lmol cm ). MS (ESI): calcd. for [M – O] 436.0213;
found 436.02332.
[
MoO
cedure was followed using bis(p-methoxybenzoyl)methane (0.48 g,
.69 mmol) as the ligand. The desired product was obtained as an
2 3 6 5 2 2
{(CH OC H CO) CH} ] (5): The general synthetic pro-
1
1
orange solid (52%, 0.31 g, 0.44 mmol). H NMR ([D
3
(
8
]thf): δ =
3
.77, 3.88 (s, 12 H, OCH ), 6.85 (d, J = 9.2 Hz, 4 H, m-ArH), 7.03
3
3
3
d, J = 8.8 Hz, 4 H, m-ArЈH), 7.19 (s, 2 H, CH), 7.96 (d, J =
3
8
.8 Hz, 4 H, o-ArH), 8.12 (d, J = 8.8 Hz, 4 H, o-ArЈH) ppm. The
assignments were confirmed by COSY NMR spectroscopy. IR: ν˜
1600 (m), 1544 (m), 1480 (vs), 1456 (vs), 1439 (s), 1300 (m), 1258
vs), 1227 (vs), 1169 (vs), 1126 (m), 1021 (m), 926 (m), 899 (m),
0
.25 mmϫ 0.25 μm) and a Perkin–Elmer TurboMass Upgrade. ESI
mass spectra were recorded in acetonitrile with a Waters LCT
Premier XE KE317 Micromass Technologies spectrometer. Owing
to the poor stability of these complexes in solution, as described in
the Results and Discussion section, no suitable elemental analysis
data was obtained.
=
(
–
1
4
8
(
42 (m), 787 (vs) cm . UV/Vis (thf): λ (ε) = 380 (4.08ϫ10 ), 363
4
–1
–1
+
4.73ϫ10 Lmol cm ). MS (ESI): calcd. for [M + H] 697.0980;
+
found 697.989; calcd. for [M – O] 680.0953; found 680.0941.
General Synthetic Procedure: A modified literature method was
used for the synthesis of complexes 3–8.[ A solution of dibenzo-
ylmethane (3.38 g, 15.1 mmol) in absolute ethanol (60 mL) was
added during 5 min to a colorless solution of sodium molybdate
dihydrate (1.09 g, 7.65 mmol) in 0.5 m HCl (45 mL, 22.6 mmol). A
yellow precipitate was formed and the suspension was cooled to
19]
[MoO {(C O)COCH ] (6): The general synthetic procedure
was followed using 2-acetylcyclopentanone (0.41 g, 3.25 mmol) as
2
5
H
6
3 2
}
the ligand. The desired product was obtained as a dark-green solid
1
(53%, 0.28 g, 0.74 mmol). H NMR ([D
8
]thf): δ = 1.85–1.97 (m, 4
), 2.29–2.56
) ppm. The assign-
H, CH
2
CH
2
CH
2
), 2.05, 2.06, 2.09, 2.09 (s, 6 H, CH
), 2.63–2.72 (m, 4 H, COCH
3
–30 °C overnight. The yellow solution was decanted and the bright-
(m, 4 H, CCH
2
2
yellow solid was dried in a vacuum desiccator over phosphorus
pentoxide overnight. The dried solid was washed three times with
diethyl ether until the filtrate remained colorless (3ϫ 15 mL) and
the solid obtained was dried in vacuo and stored under nitrogen.
ments were confirmed by COSY NMR spectroscopy. IR: ν˜ = 3299
(w), 1595 (m), 1489 (vs), 1272 (m), 1241 (s), 935 (s), 901 (vs), 831
–
1
5
(m), 791 (m), 727 (m) cm . UV/Vis (thf): λ (ε) = 332 (4.38ϫ10 ),
–
1
–1
+
777 (264 Lmol cm ). MS (ESI): calcd. for [M + CH
44.0324; found 444.0285.
3
CN + Na]
4
2 6 5 2 2
[MoO {(C H CO) CH} ] (3): The desired product was obtained as
a bright-yellow solid (63%, 2.72 g, 4.74 mmol). Single crystals suit-
able for X-ray crystallography were obtained by slow vapor dif-
fusion of diethyl ether into a dichloromethane solution at room
2 6 8 3 2
[MoO {(C H O)COCH } ] (7): The general synthetic procedure
was followed using 2-acetylcyclohexanone (1.02 g, 7.28 mmol) as
the ligand, with the exception that the ethanol was removed in
1
temperature. H NMR ([D
8
]thf): δ = 7.34–7.38 (m, 6 H, CH, m-
vacuo before cooling overnight. The desired product was obtained
3
1
ArЈH), 7.46–7.60 (m, 8 H, m-ArH, p-ArH, p-ArЈH), 8.02 (d, J = as a dark-blue solid (44%, 0.66 g, 1.61 mmol). H NMR ([D
.6 Hz, 4 H, o-ArЈH), 8.16 (d, J = 7.7 Hz, 4 H, o-ArH) ppm. The
assignments were confirmed by COSY NMR spectroscopy.
NMR ([D ]thf): δ = 97.7 [C(O)CH], 128.7 (o-ArC), 129.0 (o-ArЈC),
29.2 (m-ArЈC), 129.3 (m-ArC), 133.1 (p-ArЈC), 133.8 (p-ArC),
37.2 (ipso-ArЈC), 137.7 (ipso-ArC), 180.6 (CЈO), 188.4 (CO) ppm.
8
]thf):
), 2.06,
), 2.23–2.46 (m, 8 H, COCH
) ppm. The assignments were confirmed by COSY NMR
spectroscopy. IR: ν˜ = 3351 (br w), 2959 (w), 2878 (w), 1563 (s),
3
7
δ = 1.62–1.68 (m, 8 H, CH
2.06, 2.13, 2.15 (s, 6 H, CH
CCH
2 2 2 2 2 2 2 2
CH CH CH , CH CH CH CH
1
3
C
3
2
,
8
2
1
1
–
1
1474 (s), 1290 (vs), 940 (s), 893 (vs), 800 (m), 772 (s) cm . UV/Vis
3
–1
–1
IR: ν˜ = 3061 (w), 1594 (m), 1544 (s), 1506 (vs), 1474 (vs), 1437 (s),
(thf): λ (ε) = 289 (1.65ϫ10 ), 778 (329 Lmol cm ). MS (ESI):
354 (m), 1313 (s), 1290 (vs), 1228 (s), 1063 (m), 929 (m), 898 (vs), calcd. for [M + CH
CN + Na]+ 472.0638; found 472.0552; calcd.
for [M + Na]+ 431.0372; found 431.0361.
1
7
3
–1
55 (m), 708 (m), 676 (m) cm . UV/Vis (thf): λ (ε) = 376
Eur. J. Inorg. Chem. 2013, 2195–2204
2202
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
[
MoO
was followed using 2-acetyl-1-tetralone (2.48 g, 13.2 mmol) as the fairs, Agriculture and Innovation and the Netherlands Ministry of
ligand. The desired product was obtained as a yellow solid (80%, Education, Culture and Science.
.65 g, 5.28 mmol). Single crystals suitable for X-ray crystallogra-
phy were obtained by slow evaporation of a dichloromethane solu-
2
{(C12
H
8
O)COCH
3
}
2
] (8): The general synthetic procedure
Smart Mix Program of the Netherlands Ministry of Economic Af-
2
[
1] R. D. Perlack, L. L. Wright, A. F. Turhollow, R. L. Graham,
B. J. Stokes, D. C. Erbach, Biomass as Feedstock for a Bio-
energy and Bioproducts Industry: The Technical Feasibility of a
Billion-Ton Annual Supply, U.S. Department of Energy and
U.S. Department of Agriculture, Oak Ridge, TN, 2005.
2] T. Werpy, G. Peterson, Top Value Added Chemicals from Bio-
mass, vol. 1, Results of Screening for Potential Candidates from
Sugars and Synthesis Gas, U.S. Department of Energy, Oak
Ridge, TN, 2004.
1
tion at 2 °C. H NMR ([D
H, Ar-CH ), 2.89 (m, 4 H, CCH
ArH), 7.31 (t, J = 8.0 Hz, 2 H, 3-ArH), 7.38 (t, J = 7.2 Hz, 2 H,
8
]thf): δ = 2.28 (s, 6 H, CH
3
), 2.77 (m, 4
3
2
2
), 7.23 (d, J = 7.6 Hz, 2 H, 5-
3
3
3
4
-ArH), 7.97 (d, J = 7.2 Hz, 2 H, 2-ArH) ppm. The assignments
were confirmed by COSY NMR spectroscopy. IR: ν˜ = 2941 (w),
890 (w), 2838 (w), 1587 (m), 1562 (m), 1464 (vs), 1448 (vs), 1357
[
[
2
(
–
1
s), 1295 (s), 927 (s), 897 (vs), 737 (s), 720 (m) cm . UV/Vis (thf):
4
–1
–1
λ (ε) = 376 (1.47ϫ10 Lmol cm ). MS (ESI): calcd. for
3] M.-H. Lee, S.-W. Lee, Y.-M. Jeon, D.-Y. Park, J.-Y. Ryu, Novel
method for preparing styrenic olefins, LG Chemical Ltd., World
Patent 2005/035468, 2005.
+
[
[
3
M + CH CN + Na] 568.0640; found 568.0566; calcd. for
+
+
M + Na] 527.0374; found 527.0419; calcd. for [M – O] 488.0527;
found 488.0450.
[4] W. R. Moser, R. W. Thompson, C.-C. Chiang, H. Tong, J. Ca-
tal. 1989, 117, 19–32.
X-ray Crystallography: Single crystals of suitable dimension were
used for data collection. Reflections were measured with a Nonius–
KappaCCD diffractometer (compounds 3 and 4) with rotating an-
ode and graphite monochromator (λ = 0.71073Å) or with a Bruker
Kappa ApexII diffractometer (compound 8) with sealed tube and
Triumph monochromator (λ = 0.71073Å) up to a resolution of
[
5] H. Dirkzwager, M. van Zwienen, Process for the preparation of
styrenes, Shell International Research, World Patent 99/58480,
1999.
[6] G. Larsen, E. Lotero, L. M. Petkovic, D. S. Shobe, J. Catal.
1997, 169, 67–75.
[7] M. H. Lee, E. T. Choi, D. Kim, Y. M. Lee, Y. S. Park, Eur. J.
Org. Chem. 2008, 5630–5637.
sinθ/λmax) = 0.65 Å– at a temperature of 110(2) (3) or 150(2) K
1
(
(
[34]
[8] M. Schlaf, P. Ghosh, P. J. Fagan, E. Hauptman, R. M. Bullock,
Angew. Chem. 2001, 113, 4005; Angew. Chem. Int. Ed. 2001,
4 and 8). The reflections were integrated with EVAL14
(3),
[
35]
[36]
EVAL15 (4), or SAINT (8) software. Absorption corrections
based on multiple measured reflections were performed with SAD-
ABS.
40, 3887–3890.
[
9] Z. Zhu, J. H. Espenson, J. Org. Chem. 1996, 61, 324–328.
[
37]
The structures were solved by direct methods using
[10] J. Ziegler, M. Zdilla, A. Evans, M. M. Abu-Omar, Inorg. Chem.
[
38]
2
SHELXS-97 and refined with SHELXL-97 against F for all re-
flections. Non-hydrogen atoms were refined freely with anisotropic
displacement parameters. All hydrogen atoms were located in dif-
ference Fourier maps (3 and 4) or introduced at calculated posi-
tions (8) and refined with a riding model. The structure of 8 was
2009, 48, 9998–10000.
[11] E. Arceo, J. Ellman, R. Bergman, J. Am. Chem. Soc. 2010, 132,
11408–11409.
[12] S. Vkuturi, G. Chapman, I. Ahmad, K. M. Nicholas, Inorg.
Chem. 2010, 49, 4744–4746.
refined as an inversion twin[
0
was performed with the PLATON program.
27]
[13] I. Ahmad, G. Chapman, K. M. Nicholas, Organometallics
resulting in a twin fraction of
2011, 30, 2810–2818.
.15(2). Geometry calculations and checking for higher symmetry
[
[
[
[
[
14] T. J. Korstanje, J. T. B. H. Jastrzebski, R. J. M. Klein Gebbink,
[39]
ChemSusChem 2010, 3, 695–697.
15] T. J. Korstanje, E. F. de Waard, J. T. B. H. Jastrzebski, R. J. M.
Klein Gebbink, ACS Catal. 2012, 2, 2173–2181.
CCDC-910320 (for 3), -910321 (for 4), and -910322 (for 8) contain
the supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
16] M. L. Kantam, A. D. Prasad, P. L. Santhi, Synth. Commun.
1
993, 23, 45–48.
17] C. Kisker, H. Schindelin, D. Rees, Annu. Rev. Biochem. 1997,
6, 233–267.
6
Typical Procedure for the Dehydration of Alcohols to Olefins: A
solution of substrate (2 mmol) and pentadecane as internal stan-
dard (250 μL, 192.3 mg) in toluene (10 mL) was added to the cata-
lyst. The reaction flask was sealed with a septum, placed in a pre-
heated aluminum block-heater at 100 °C, and stirred magnetically.
Samples for GC analysis were removed by syringe, filtered through
Florisil, and eluted with ethyl acetate. For the reactions analyzed
18] H. Vrubel, V. Verzenhassi, S. Nakagaki, F. Nunes, Inorg. Chem.
Commun. 2008, 11, 1040–1043.
[19] F. Moore, R. Rice, Inorg. Chem. 1968, 7, 2510–2514.
[20] W. Griffith, T. Wickins, J. Chem. Soc. A 1968, 400–404.
[21] A. A. Bugayev, S. E. Nikitin, Opt. Commun. 2000, 180, 69–71.
[22] M. Brown, B. Cook, T. Sloan, Inorg. Chem. 1975, 14, 1273–
1278.
by NMR, [D
8
]toluene was used as the solvent and the yield was
[23] B. Kamenar, M. Penavic, C. K. Prout, Cryst. Struct. Commun.
1973, 2, 41–44.
determined against 1,3,5-trimethoxybenzene (0.67 mmol, 112.1 mg)
as internal standard. When using a liquid catalyst, the same pro-
cedure was used, but the catalyst was added to the solution by
Finnpipette. For the reactions at 150 °C, the same procedure was
used but in a 50 mL Parr autoclave using a glass insert and me-
chanical stirring.
[
[
[
[
24] O. Krasochka, Y. Sokolova, L. Atovmyan, J. Struct. Chem.
1
975, 16, 648–650.
25] M. Nass, M. Schurmann, H. Preut, N. Krause, Z. Kristallogr.
New Cryst. Struct. 2001, 216, 461–464.
26] B. Kojic-Prodic, Z. Rudic-Toros, D. Grdenic, L. Golic, Acta
Crystallogr., Sect. B 1974, 30, 300–305.
27] H. Flack, Acta Crystallogr., Sect. A 1983, 39, 876–881.
Supporting Information (see footnote on the first page of this arti-
cle): H, C, and COSY NMR, IR, UV/Vis, and ESI-MS spectra
for complexes 1–8.
1
13
[28] A. Vasishtha, R. Trivedi, G. Das, J. Am. Oil Chem. Soc. 1990,
6
7, 333–337.
29] P. van Leeuwen, N. Clement, M. Tschan, Coord. Chem. Rev.
011, 255, 1499–1517.
[
2
[
[
[
30] H. Gehrke Jr., J. Veal, Inorg. Chim. Acta 1969, 3, 623–627.
31] I. Bernal, Chem. Ind. 1966, 1343–1344.
32] M. Peterson, R. Vanatta, R. Richman, Inorg. Chem. 1983, 22,
1982–1985.
Acknowledgments
This research was performed within the framework of the CatchBio
program. The authors gratefully acknowledge the support of the
[33] W. Franek, Monatsh. Chem. 1996, 127, 895–907.
Eur. J. Inorg. Chem. 2013, 2195–2204
2203
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
[
34] A. J. M. Duisenberg, L. M. J. Kroon-Batenburg, A. M. M.
[37] G. M. Sheldrick, SADABS: Area-Detector Absorption Correc-
tion, v. 2.10, University of Göttingen, Germany, 1999.
Schreurs, J. Appl. Crystallogr. 2003, 36, 220–229.
[
38] G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–122.
39] A. L. Spek, Acta Crystallogr., Sect. D 2009, 65, 148–155.
Received: November 8, 2012
[35] A. M. M. Schreurs, X. Y. Xian, L. M. J. Kroon-Batenburg, J.
Appl. Crystallogr. 2010, 43, 70–82.
[
[
36] SAINT-Plus, Bruxer AXS Inc., Madison, WI, 2001.
Published Online: February 8, 2013
Eur. J. Inorg. Chem. 2013, 2195–2204
2204
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim