A flexible nonporous heterogeneous catalyst for size-selective oxidation through a bottom-up approach

Article abstract of DOI:10.1002/anie.201005275

Size does matter: The nonporous tetra-n-butylammonium salt of silicodecatungstate, synthesized through a bottom-up approach, heterogeneously catalyzes the size-selective oxidation of various organic compounds, including olefins, sulfides, and organosilanes, with aqueous H₂O₂ in ethyl acetate. The catalyst can be easily separated by filtration and reused several times with retention of high catalytic activity. Copyright

Full text of DOI:10.1002/anie.201005275

Communications
DOI: 10.1002/anie.201005275
Heterogeneous Catalysis
A Flexible Nonporous Heterogeneous Catalyst for Size-Selective
Oxidation through a Bottom-Up Approach**
Noritaka Mizuno,* Sayaka Uchida, Keigo Kamata, Ryo Ishimoto, Susumu Nojima,
Koji Yonehara, and Yasutaka Sumida
The bottom-up approach has the potential to create novel
devices with a wide range of applications such as in
electronics, medicine, and energy, as the arrangement of
molecular building blocks into nanostructures can be con-
trolled.[1,2] It is still a great challenge to fabricate not only
devices but also heterogeneous catalysts with intended
structures and functions by a bottom-up approach,[3] while
biominerals such as shells and bones have been already
formed by the bottom-up approach through the self-assembly
of inorganic building blocks with organic molecules in
water.[4] The control of the self-organization of nanobuilding
blocks with well-defined sizes, shapes, and physical and
chemical properties would lead to progress in science and
technology. Various catalytically active sites, such as metal
nodes, framework nodes, and molecular species, can be
introduced into metal–organic frameworks (MOFs) through
self-assembly. Efficient size- and enantioselective catalysis by
crystalline and porous MOFs has been reported for reduction,
Polyoxometalates (POMs) are discrete early transition-
metal oxide cluster anions with applications in broad fields,
such as catalysis, materials, and medicine, because their
structures and chemical properties can be finely tuned by
choose of the constituent elements.[9–14] Various POMs such as
peroxometalates, lacunary POMs, and transition-metal-sub-
stituted POMs have been developed for H2O2- or O2-based
green oxidations. Therefore, POMs are suitable nanobuilding
blocks to construct heterogeneous oxidation catalysts.
Recently, the development of heterogeneous oxidation cata-
lysts based on POMs and the related compounds has been
attempted according to the following strategies: “solidifica-
tion” of POMs (formation of insoluble solid ionic materials
with appropriate countercations) and “immobilization” of
POMs through adsorption, covalent linkage, and ion
exchange. In most cases, however, the catalytic activities
and selectivities of the parent homogeneous POMs are
somewhat or much decreased by the heterogenization, and
there are only a few successful examples.[15–18] We are
interested in a bottom-up approach to the design and
synthesis of artificial heterogeneous catalysts with POMs
and herein report that the nonporous tetra-n-butylammonium
salt of [g-SiW10O34(H2O)2]4ꢀ ([(n-C4H9)4N]4[g-SiW10O34-
ꢀ
C C bond formation, and acid–base reactions, and hydrolytic
and oxidative stabilities are critical for the development of
MOF-based oxidation systems that are efficient, chemo- and
size-selective, and recyclable, and use the green oxidant
H2O2.[5–7] Therefore, the development of efficient, easily
recoverable, and recyclable heterogeneous oxidation catalysts
with H2O2 by a bottom-up approach has received particular
research interest.[3,8]
(H2O)2]·H2O, 1·H2O) synthesized through
a bottom-up
approach sorbs ethyl acetate (EtOAc), which is highly
mobile in the solid bulk of the compound, probably contri-
buting to the easy co-sorption of the olefins and H2O2. The
compound heterogeneously catalyzes size-selective oxidation
of various organic substances including olefins, sulfides, and
silanes with aqueous H2O2 in EtOAc. The compound can
easily be separated by filtration and reused several times with
retention of its high catalytic activity. The catalysis is truly
heterogeneous in nature because the filtrate after removal of
the solid catalyst is completely inactive. Notably, size-
selective oxidation catalysis is observed: small olefins are
much more preferentially epoxidized than large olefins. To
the best of our knowledge, this study provides the first
example for the heterogeneously catalyzed size-selective
liquid-phase oxidation with H2O2 by a POM-based catalyst.
Compound 1·H2O was synthesized by a bottom-up
approach as described below. The silicodecatungstate [g-
SiW10O34(H2O)2]4ꢀ was synthesized in situ by the addition of
concentrated HNO3 to an aqueous solution of [g-SiW10O36]8ꢀ.
Then, tetra-n-butylammonium bromide [(n-C4H9)4N]·Br was
added to the solution, and white powder of 1·H2O was
formed.[19] The use of other cations, such as tetramethylam-
monium [(CH3)4N]+, formed single crystals.[19] The powder X-
ray diffraction (XRD) pattern, crystal structure, and space-
filling model of 1·H2O are shown in Figure 1a–c. The
[*] Prof. Dr. N. Mizuno, Dr. S. Uchida, Dr. K. Kamata, R. Ishimoto,
S. Nojima
Department of Applied Chemistry, School of Engineering, The
University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-8656 (Japan)
Prof. Dr. N. Mizuno, Dr. S. Uchida, Dr. K. Kamata
Core Research for Evolutional Science and Technology (CREST)
(Japan), Science and Technology Agency (JST), 4-1-8 Honcho,
Kawaguchi, Saitama 332-0012 (Japan)
E-mail: tmizuno@mail.ecc.u-tokyo.ac.jp
Dr. K. Yonehara, Y. Sumida
Advanced Materials Research Center, Nippon Shokubai Co., Ltd.,
5-8 Nishi Otabi-cho, Suita, Osaka 564-8512 (Japan)
[**] We are grateful to S. Kuzuya (The University of Tokyo) and M. Fujita
(The University of Tokyo) for their help in experiments. This work
was supported by the Core Research for Evolutional Science and
Technology (CREST) program of the Japan Science and Technology
Agency (JST), the Global COE Program (Chemistry Innovation
through Cooperation of Science and Engineering), the Development
in a New Interdisciplinary Field Based on Nanotechnology and
Materials Science Programs, and a Grant-in-Aid for Scientific
Research from the Ministry of Education, Culture, Science, Sports,
and Technology of Japan.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201005275.
9972
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 9972 –9976

Angewandte
Chemie
similar to that of 1·H2O, indicating that the structure is
maintained after desorption of the water of crystallization
(Figure S1 in the Supporting Information). The surface area
of 1 calculated with the Brunauer–Emmett–Teller (BET) plot
of the N2 adsorption isotherm (77 K, Figure S2) was 5.4 m2 gꢀ1
and low, which agreed with the nonporous crystal structure of
1. In addition, the particle size of 1 calculated with the BET
surface area (5.4 m2 gꢀ1) and density (2.17 gcmꢀ3) assuming
spherical particles was r= 0.26 mm, and agreed fairly well with
the SEM image (r= 0.25 mm) of 1.
Compound 1 could act as an effective heterogeneous
catalyst for the oxidation of 1-hexene with H2O2 in insoluble
EtOAc to give 1,2-epoxyhexane in 92% yield. The epoxida-
tion of 1-hexene was completely stopped by the removal of 1
from the reaction solution (Figure S3). In addition, it was
confirmed by inductively coupled plasma atomic emission
spectroscopy that the tungsten content in the filtrate was less
than 0.2% of that in fresh 1. These results indicate that any
tungsten species that leached into the reaction solution is not
an active homogeneous catalyst and that the observed
catalysis is truly heterogeneous in nature. The epoxidation
did not proceed efficiently without EtOAc, suggesting that
EtOAc promotes the present heterogeneous oxidation catal-
ysis. Compound 1 sorbed (4.4 ꢁ 0.4) ꢁ 10ꢀ1 molmolꢀ1 of 1-
hexene from the mixture with EtOAc, and the amount was
two orders of magnitude larger than that of the surface
adsorption (ca. 4.0 ꢁ 10ꢀ3 molmolꢀ1) [Eq. (1)]. Compound 1
sorbed 2.3 ꢁ 0.1 molmolꢀ1 H2O2 from the mixture with
EtOAc [Eq. (2)]. These results show that both reactant and
oxidant are sorbed into the solid bulk of 1 by the presence of
EtOAc.
Figure 1. a) Experimental and calculated powder XRD patterns, b) crys-
tal structure, and c) space-filling model of 1·H2O. d) Experimental and
calculated powder XRD patterns, e) crystal structure, and f) space-
filling model of 1·EtOAc. Gray lines and black markers in (a) and (d)
show the differences between the experimental and calculated data
and the calculated peak positions, respectively. Light green and purple
polyhedra in (b) and (e) show the [WO6] and [SiO4] units, respectively.
Black sticks in (b) and (e) show the [(n-C4H9)4N]+ fragments. Light
green spheres in (c) and (f) show the tungsten and oxygen atoms of
[g-SiW10O34(H2O)2]4ꢀ. Purple, dark gray, and pink spheres in (c) and (f)
show the silicon, carbon, and hydrogen atoms, respectively.
ꢀ1
IðtÞ ¼ I0ð1ꢀTCH=T
Þ
fexpðꢀt=T11ðHÞÞꢀexpðꢀt=TCHÞg
ð3Þ
11ðHÞ
The powder XRD pattern of 1 (Figure 1a) was completely
changed by EtOAc sorption (Figure 1d). The structure
analyses of the powder XRD patterns showed that the crystal
system changed from tetragonal to cubic and the cell volume
(V/Z) increased from 2605 ꢀ3 to 2761 ꢀ3 with EtOAc
sorption. The increase in the cell volume (156 ꢀ3) agreed
well with the volume of the EtOAc molecule (162 ꢀ3)
calculated from the molecular weight and density of the
liquid, suggesting that 1 molmolꢀ1 of EtOAc was sorbed into
the crystal lattice of 1.
calculated peak positions and shapes agreed fairly well those
found experimentally (Figure 1a). The rather large difference
between the experimental and calculated data, especially at
higher angles, is probably because of the variations of electron
densities in the unit cell, which may be caused by the highly
disordered alkyl chains of the tetra-n-butylammonium cat-
ions. Therefore, the tetra-n-butylammonium cations were not
located with the powder XRD analysis but with the molecular
mechanics using a universal force field. In the crystal lattice of
1·H2O, the [g-SiW10O34(H2O)2]4ꢀ ions were arranged in a
tetragonal cell and surrounded by the [(n-C4H9)4N]+ cations.
The constituent ions were densely packed in the unit cell, and
1·H2O was nonporous with a normalized unit cell volume (V/
Z) of 2605 ꢀ3. The water of crystallization in 1·H2O was
desorbed by the evacuation or treatment with dry He gas flow
at 298–333 K to form the guest free phase [(n-C4H9)4N]4[g-
SiW10O34(H2O)2] (1). The powder XRD pattern of 1 was
During the spin locking step in 13C cross-polarization
magic-angle spinning (CPMAS) NMR spectroscopy, there is
magnetization transfer from 1H to 13C through heteronuclear
dipolar interactions. The cross-polarization process is de-
scribed by Equation (3) with an assumption TCH ! T11(H)
,
where I(t), I0, TCH, and T11(H) represent the intensity of the
signal at contact time of t, theoretical maximum intensity of
Angew. Chem. Int. Ed. 2010, 49, 9972 –9976
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 9973

Communications
the signal, cross-polarization time constant, and proton spin-
lattice relaxation time, respectively.[20,21]
As a result, the signal intensity of the less mobile carbon
atom reaches a maximum at relatively short contact times
(shorter TCH). Therefore, the mobilities of the [(n-C4H9)4N]+
cations of 1 as a host and EtOAc molecules as a guest were
investigated with variable contact time CPMAS NMR
spectroscopy. The changes in the 13C CPMAS NMR spectra
of 1·EtOAc at 333 K with an increase in contact times are
shown in Figure S4. In the range of short contact times
(ꢂ 5 ms), only the signals for the C1–C4 carbon atoms of [(n-
C4H9)4N]+ were observed. At a contact time of 10 ms, the
signals for [(n-C4H9)4N]+ almost disappeared and those for
EtOAc appeared. In the range 10–100 ms, the intensities of
the signals for EtOAc monotonically increased. The cross-
polarization time constants for the C1, C2, C3, and C4 carbon
atoms of [(n-C4H9)4N]+ were 0.11, 0.12, 0.16, and 0.22 ms,
respectively (Figure 2). These values are comparable to those
Figure 3. Peak intensities of the 13C CPMAS NMR spectra of 1·EtOAc
at 333 K as functions of contact time. a) Methylene group
(d=60.0 ppm), b) methyl group of acetate moiety (d=20.0 ppm), and
c) methyl group of ethoxy moiety (d=14 ppm).
Figure 4. Recycling of 1 for the oxidation of a) propene, b) ethylpropyl-
sulfide, and c) ethyldimethylsilane. Reaction conditions: a) 1 (2
mol%), propene (6 atm), 60% H2O2 (1 mmol), EtOAc (6 mL), 333 K,
3 h. b) 1 (0.7 mol%), ethylpropylsulfide (3 mmol), 60% H2O2
(1 mmol), EtOAc (3 mL), 313 K, 2 h. c) 1 (2 mol%), ethyldimethylsi-
lane (3 mmol), EtOAc (6 mL), 333 K, 3 h.
scale epoxidation of 1-butene with H2O2 in the butyl acetate
solvent, and the retention of high catalytic performance in a
large-scale epoxidation raises the prospect of the present
system for the industrial heterogeneous epoxidation [Eq. (4)].
Since the reaction rates for the heterogeneous system were
smaller than those for the homogeneous one, longer reaction
times were needed. In addition, the heterogeneous system
required higher reaction temperature than the homogeneous
system (especially for the larger olefins). Therefore, non-
productive decomposition of H2O2 proceeds in some cases,
resulting in a slight decrease in yields for the present
heterogeneous system.
Figure 2. Peak intensities of the 13C CPMAS NMR spectra of 1·EtOAc
at 333 K as a function of contact time. a) C1 (d=57.9 ppm), b) C2
(d=24.6 ppm), c) C3 (d=20.3 ppm), and d) C4 (d=14.9 ppm) of (n-
C4H9)4N+.
of the alkylammonium surfactants in mesoporous silica.[22]
Notably, the cross-polarization time constants for the carbon
atoms of EtOAc were approximately 60 ms and much longer,
suggesting high mobility of the EtOAc molecules (Figure 3).
The epoxidation of nonactivated propene in EtOAc with
H2O2 catalyzed by 1 also proceeded efficiently to form 1,2-
epoxypropane in 89% yield (Figure 4). After the epoxidation,
1 could easily be separated from the reaction mixture by
simple filtration. The recovered 1 after the epoxidation could
be reused at least four times without significant loss of
catalytic activity (95% (1st reuse), 86% (2nd reuse), and 91%
(3rd reuse)). In addition, the present system could be applied
to the heterogeneous and recyclable oxidation of ethyl-
propylsulfide and ethyldimethylsilane to the corresponding
sulfoxide and silanol (Figure 4). 1,2-Epoxybutane was
obtained in 90% yield with 90% selectivity in a 240 mmol
Notably, the size-selective heterogeneous epoxidation
proceeded with 1 in EtOAc, while no size-selectivity was
observed for the 1-catalyzed homogeneous competitive
=
epoxidation in acetonitrile in accord with the p(C C)
9974 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 9972 –9976

Angewandte
Chemie
HOMO energies (Table S1). In the case of the heterogeneous
competitive epoxidation of 1-hexene and various C3–C12
terminal olefins, the yield ratios of the C3–C12 epoxides to
1,2-epoxyhexane monotonically decreased to 0.04 with an
increase in the molecular cross-sectional areas of the C3–C12
olefins (Figure 5). The amounts of 1-hexene and 1-dodecene
in Table S2. The reactivity order was different from each
other. In the heterogeneous system, the reaction rates for the
terminal olefins were increased with respect to those for the
branched olefins. Such an order was also observed for
epoxidation catalyzed by TS-1 (titanium silicalite), for
which restricted transition-state shape selectivity and diffu-
sion effects were observed.[23] In addition, the competitive
epoxidation of propene and cyclododecene gave 1,2-epoxy-
propane and 1,2-epoxycyclododecane in 66% and 14%
yields, respectively, whereas 1,2-epoxycyclododecane was
obtained preferentially in the homogeneous system
[Eq. (6)]. Not only olefins but also sulfides (ethylpropylsul-
fide vs. dibutylsulfide) and silanes (ethyldimethylsilane vs.
tripropylsilane) were oxidized size-selectively in EtOAc
solvent [Eqs. (7) and (8)] and the small substrates were
preferentially oxidized to the corresponding sulfoxide and
silanol. Since the EtOAc molecules are highly mobile in the
solid bulk of 1, it is probable that the sorption of olefins and
H2O2 and the size-selectivity is largely enhanced by its
presence. Such a unique sorption property of 1 resulted in
heterogeneous, recyclable, and size-selective epoxidation
catalysis with H2O2. To the best of our knowledge, size-
selective oxidation catalyzed by POM-based compounds has
never been reported to date.
Figure 5. Plots of the ratios of the C3–C12 epoxides to 1,2-epoxyhexane
for competitive epoxidation of various C3–C12 terminal olefins and 1-
hexene against the molecular cross-sectional areas of the olefins. The
details are shown in Table S1.
sorption from a mixture in EtOAc were (4.4 ꢁ 0.8) ꢁ 10ꢀ1 and
(4.0 ꢁ 3.0) ꢁ 10ꢀ2 molmolꢀ1, respectively [Eq. (5)], which is in
accord with the much faster epoxidation of 1-hexene than 1-
dodecene in Figure 5. The yield ratios of the C3–C12 epoxides
to 1,2-epoxyhexane against the molecular cross-sectional
areas of the C3–C12 olefins for the heterogeneous compet-
itive epoxidation of 1-hexene and various C3–C12 terminal
olefins in the presence of [(n-C5H11)4N]4[g-SiW10O34(H2O)2]
monotonically decreased to 0.05 with an increase in the
molecular cross-sectional areas in a similar way to that of [(n-
C4H9)4N]4[g-SiW10O34(H2O)2] (Figure S5). Therefore, the
size-selectivity was not much changed by the change of the
counter cation from [(n-C4H9)4N]+ to [(n-C5H11)4N]+.
Received: August 24, 2010
Published online: November 17, 2010
Keywords: heterogeneous catalysis · oxidation ·
.
polyoxometalates · size-selectivity · structure elucidation
The ratio of C8 epoxide to C6 epoxide for the competitive
epoxidation of 1-hexene and 1-octene remained almost
unchanged as a function of time (Figure S6), suggesting
little variation in the size-selectivity during the epoxidation.
In addition, the ratios of C8 epoxide to C6 epoxide varied
little with changes in the size and dipole moment of solvents
(Figure S7).
[1] W. Lu, C. M. Lieber, Nat. Mater. 2007, 6, 841 – 850.
[2] H. Cꢂlfen, M. Antonietti, Angew. Chem. 2005, 117, 5714 – 5730;
Angew. Chem. Int. Ed. 2005, 44, 5576 – 5591.
[3] W. Schmidt, ChemCatChem 2009, 1, 53 – 67.
[4] J. D. Hartgerink, E. Beniash, S. I. Stupp, Science 2001, 294,
1684 – 1688.
[5] A. Corma, H. Garcꢃa, F. X. Llabrꢄs i Xamena, Chem. Rev. 2010,
110, 4606 – 4655.
The reactivity for the heterogeneous and homogeneous
epoxidation of a series of C6 olefins catalyzed by 1 are shown
[6] J. Y. Lee, O. K. Farha, J. Roberts, K. A. Scheidt, S. B. T. Nguyen,
J. T. Hupp, Chem. Soc. Rev. 2009, 38, 1450 – 1459.
Angew. Chem. Int. Ed. 2010, 49, 9972 –9976
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9975

Communications
[7] L. Alaerts, J. Wahlen, P. A. Jacobs, D. E. De Vos, Chem.
[15] J. T. Rhule, W. A. Neiwert, K. I. Hardcastle, B. T. Do, C. L. Hill,
J. Am. Chem. Soc. 2001, 123, 12101 – 12102.
Commun. 2008, 1727 – 1737.
[8] R. A. Sheldon, M. C. A. van Vliet, Fine Chemicals through
Heterogeneous Catalysis, Wiley-VCH, Weinheim, 2001.
[9] R. Neumann, Prog. Inorg. Chem. 1998, 47, 317 – 370.
[10] Special Thematic Issue on Polyoxometalates (Ed.: C. L. Hill),
Chem. Rev. 1998, 98, 1 – 390.
[11] I. V. Kozhevnikov, Catalysts for Fine Chemical Synthesis:
Catalysis by Polyoxometalates, Wiley, Chichester, 2002.
[12] M. T. Pope in Comprehensive Coordination Chemistry II, Vol. 4
(Eds.: J. A. McCleverty, T. J. Meyer), Elsevier Pergamon,
Amsterdam, 2004, pp. 635 – 678.
[16] K. Yamaguchi, C. Yoshida, S. Uchida, N. Mizuno, J. Am. Chem.
Soc. 2005, 127, 530 – 531.
[17] R. Neumann, H. Miller, J. Chem. Soc. Chem. Commun. 1995,
2277 – 2278.
[18] B. Sels, D. De Vos, M. Buntinx, F. Pierard, A. Kirsch-De
Mesmaeker, P. Jacobs, Nature 1999, 400, 855 – 857.
[19] K. Kamata, K. Yonehara, Y. Sumida, K. Yamaguchi, S. Hikichi,
N. Mizuno, Science 2003, 300, 964 – 966.
[20] C. A. Fyfe, A. C. Diaz, H. Grondey, A. R. Lewis, H. Forster,
C. A. Jyfe, J. Am. Chem. Soc. 2005, 127, 7543 – 7558.
[21] S. H. Liang, I. D. Gay, Langmuir 1985, 1, 593 – 599.
[22] L. Q. Wang, J. Liu, G. J. Exarhos, B. C. Bunker, Langmuir 1996,
12, 2663 – 2669.
[13] C. L. Hill in Comprehensive Coordination Chemistry II, Vol. 4
(Eds.: J. A. McCleverty, T. J. Meyer), Elsevier Pergamon,
Amsterdam, 2004, pp. 679 – 758.
[14] N. Mizuno, K. Kamata, S. Uchida, K. Yamaguchi in Modern
Heterogeneous Oxidation Catalysis (Ed.: N. Mizuno), Wiley-
VCH, Weinheim, 2009, pp. 185 – 216.
[23] M. G. Clerici, P. Ingallina, J. Catal. 1993, 140, 71 – 83.
9976 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 9972 –9976