Olefin epoxidation with hydrogen peroxide using octamolybdate-based self-separating catalysts

Article abstract of DOI:10.1039/c4gc01411a

Mo₈O₂₆^4- based organic polyoxomolybdate salts of general formula [Hmim]₄Mo₈O₂₆ (Hmim = 1-hexyl-3-methylimidazolium), [Dhmim]₄Mo₈O₂₆ (Dhmim = 1,2-dimethyl-3-hexylimidazolium) and [Hpy]₄Mo₈O₂₆·H₂O (Hpy = 1-hexylpyridinium) have been prepared and characterized. These compounds were applied as catalysts for olefin epoxidation using hydrogen peroxide (H₂O₂) as oxidant in CH₃CN. The polyoxomolybdate salts exhibit excellent catalytic performance and are also self-separating, a great advantage for catalyst recycling. The catalysts can be reused for at least 10 runs without significant loss of activity. This journal is

Full text of DOI:10.1039/c4gc01411a

Green Chemistry
View Article Online
View Journal
PAPER
Olefin epoxidation with hydrogen peroxide using
octamolybdate-based self-separating catalysts†
Cite this: DOI: 10.1039/c4gc01411a
Ming-Dong Zhou,*a,b Mei-Ju Liu,a Liang-Liang Huang,a Jian Zhang,a
Jing-Yun Wang,b Xue-Bing Li,c Fritz E. Kühn*b and Shu-Liang Zanga
4−
Mo8O26
based organic polyoxomolybdate salts of general formula [Hmim]4Mo8O26 (Hmim =
1-hexyl-3-methylimidazolium), [Dhmim]4Mo8O26 (Dhmim = 1,2-dimethyl-3-hexylimidazolium) and
[Hpy]4Mo8O26·H2O (Hpy = 1-hexylpyridinium) have been prepared and characterized. These compounds
were applied as catalysts for olefin epoxidation using hydrogen peroxide (H2O2) as oxidant in CH3CN. The
polyoxomolybdate salts exhibit excellent catalytic performance and are also self-separating, a great
advantage for catalyst recycling. The catalysts can be reused for at least 10 runs without significant loss
of activity.
Received 24th July 2014,
Accepted 18th October 2014
DOI: 10.1039/c4gc01411a
www.rsc.org/greenchem
in an industrial-scale process is so far not convincingly
established.8
Introduction
Epoxides are very important key raw materials for a wide
Polyoxometalates (POMs), owing to their good catalytic fea-
variety of chemicals such as glycols, glycol ethers, alkanol- tures such as high activity and selectivity, controllable redox
amines, and polymers.1 So far, various compounds including and acidic properties at atomic or molecular levels, have also
inorganic metal oxides and organometallic compounds have been studied as oxidation catalysts in the past few decades.9
been applied as catalysts for olefin epoxidations using oxygen, Among various POMs, the classic Ishii-Venturello system has
ozone, hydrogen peroxide or organic peroxides as oxidants.2 In attracted much interest in oxidation reactions and has been
particular, hydrogen peroxide-based catalytic epoxidation has extensively investigated.10 In the Ishii-Venturello system, the
received much attention since hydrogen peroxide is a cheap, Venturello anion ({PO4[WO(O2)2]4}3−) is combined with a qua-
mild and environmentally benign reagent that produces only ternary ammonium cation in the formation of an inorganic–
water as by-product.3 Today, H2O2 has already been success- organic hybrid salt, which can be used as a phase transfer
fully used in industry for propylene epoxidation using TS-1 catalyst using H2O2 as primary oxidant for epoxidation of
zeolite as heterogeneous catalyst.4 However, in the case of various olefins under rather mild conditions.11 It has addition-
homogeneous catalysis, although some molecular organome- ally been found that some Keggin- or Lindqvist-type hybrid
tallic compounds such as CH3ReO3 (MTO),5 MoO2X2L2,6 POMs such as [Cnmim]PW12O40 and [Cnmim]W6O19 can also
Cp*MoO2Cl7 etc. have proven to be excellent catalysts for epoxi- be used as catalysts for oxidations and esterifications.12
dations on a laboratory scale, they are not applied in larger
Despite these catalysts showing high activity towards epoxi-
scale industrial processes since these types of organometallic dations, it is not convenient to separate such catalysts in a
catalysts are rather expensive to produce and their reusability homogeneous system. Xi et al. have developed an interesting
“reaction-controlled phase-transfer catalysis” system using
[π-C5H5NC16H33]3[PO4(WO3)4] as a phase transfer catalyst for
aSchool of Chemistry and Material Science, Liaoning Shihua University,
propylene epoxidation, and the system was also proven to be
active for other olefins.13 In Xi’s system, the catalyst precursor
Dandong Road, No. 1, Fushun 113001, China. E-mail: mingdong.zhou@lnpu.edu.cn;
Fax: +86 24 56863837; Tel: +86 24 56860770
is insoluble, but the active species formed is soluble in the
reaction medium, and therefore the system manifests a solid–
liquid–solid phase transfer of the catalyst during the reaction.
Thus the catalyst is “self-separating” after the completion of
the reaction and can be recycled simply by filtration.13 Sub-
bMolecular Catalysis/Chair of Inorganic Chemistry, Catalysis Research Center,
Technische Universität München, Ernst-Otto-Fischer-Strasse 1, D-85747 Garching
bei München, Germany. E-mail: fritz.kuehn@ch.tum.de; Fax: +49 89 289 13473;
Tel: +49 89 289 13096
cKey Laboratory of Biofuels, Qingdao Institute of Bioenergy and Bioprocess
Technology, Chinese Academy of Sciences, Qingdao 266101, China.
E-mail: lixb@qibebt.ac.cn; Fax: +86 532 80662778; Tel: +86 532 80662757
†CCDC 997672 and 997673. For crystallographic data in CIF or other electronic
format see DOI: 10.1039/c4gc01411a
3−
3−
sequently, it was found that PW12O40 or PO4[(WO(O2)2]4
anion-based compounds also show self-separation phenomena
in catalysis.14 These kinds of catalysts combine the advantages
This journal is © The Royal Society of Chemistry 2014
Green Chem.

View Article Online
Paper
Green Chemistry
of both homogeneous and heterogeneous catalysts; thus they
are very promising catalysts for industrial applications.15
Recently, we have found that alkylimidazolium/alkylpyridin-
ium octamolybdates exhibit high activity towards olefin epoxi-
dations when using 30% H2O2 as oxidant. Moreover, these
compounds display self-separation properties in catalysis, and
the catalysts can be used at least 10 times without significant
Table 1 Physical property and characterization of compounds 1–3
Decomp.
temp. (°C)
Catalyst
IR (cm−1
)
[Hmim]4Mo8O26
[Dhmim]4Mo8O26
[Hpy]4Mo8O26·H2O
300
290
310
946
956
947
910
911
911
841
837
836
710
714
717
665
650
632
4−
loss of activity, indicating good catalyst stability. Mo8O26
temperature. TGA data indicate that all examined compounds
show negligible volatility and high thermal stability with a
decomposition onset temperature around 300 °C (see Table 1).
based salts are types of cheap and easily prepared POM
materials.16 However, not much work has been done with
respect to their catalytic performance so far,17 and almost no
studies have targeted their self-separating property. Although
FT-IR spectra were used to clarify the structure of the
4−
Deng et al.17a have mentioned that [(n-C4H9)4N]4[α-Mo8O26
]
Mo8O26
anions. All three compounds show similar IR
spectra with respect to Mo–O bond stretching. The IR spectra
exhibit several characteristic Mo–O stretching bands in the
region of 700–1000 cm−1 (see the data in Table 1). The values
are in good accordance with those published in the litera-
ture.16,17 The single-crystal X-ray structures were obtained for
compounds 1 and 3, which further confirm the results of the
spectroscopic analysis (Fig. 1 and 2).
could be self-precipitating in a sulfide oxidation catalytic
system, ethyl acetate had to be added to precipitate the cata-
lyst. In this work, a highly efficient, “green” method, allowing
for good catalyst recycling for olefin epoxidations using 30%
aqueous hydrogen peroxide as oxidant and alkylimidazolium
or alkylpyridinium octamolybdates as catalysts, is presented.
A summary of experimental data of the crystal determi-
nation can be found in the Experimental section. Selected
bond distances of the Mo8O26 anions of 1 and 3 are given in
4−
Results and discussion
Table 2. Both compounds show an array of eight edge-shared
MoO6 octahedra, which includes different types of Mo–O
bonds: terminal Mo–Ot, bridging Mo–Oμ2, Mo–Oμ3 and
Mo–Oμ5, indicating a typical β-octamolybdate.17b,18 According to
the X-ray structure of compound 1, the averaged bond distance
Synthesis and characterization
4−
Three Mo8O26
based organic polyoxomolybdate salts of
1-hexyl-3-methylimid-
formula [Hmim]4Mo8O26 (Hmim
=
azolium), [Dhmim]4Mo8O26 (Dhmim = 1,2-dimethyl-3-hexylimid-
azolium) and [Hpy]4Mo8O26·H2O (Hpy = 1-hexylpyridinium)
have been synthesized. The structures are shown in Scheme 1.
1
All three compounds were characterized by FT-IR, H NMR,
ESI MS (positive mode) and elemental analysis (see Experi-
mental section). The melting point and decomposition temp-
erature were determined with a micro-melting point apparatus
and by thermogravimetric analysis (TGA). All three compounds
are not conventional ionic liquids since they are solids below
100 °C. The melting point for 1 is about 158 °C, whereas com-
pounds 2 and 3 decompose directly at higher temperatures
but no melting process is observed below the decomposition
Fig. 1 ORTEP view of compound 1 showing vibrational ellipsoids at the
50% probability level. H atoms are omitted for clarity.
Fig. 2 ORTEP view of compound 3 showing vibrational ellipsoids at the
Scheme 1 Structures of compounds 1–3.
50% probability level. H atoms are omitted for clarity.
Green Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Table 2 Selected bond distances (Å) for compounds 1 and 3a
Mo–Ot
Mo–Oµ2
Mo–Oµ3
Mo–Oµ5
1
3
Mo(1)–O(6)
Mo(2)–O(5)
Mo(2)–O(7)
Mo(3)–O(12)
Mo(3)–O(8)
Mo(4)–O(11)
Mo(4)–O(13)
1.693(5)
1.690(5)
1.698(5)
1.701(5)
1.704(5)
1.696(6)
1.699(6)
Mo(1)–O(10)
Mo(2)–O(9)
Mo(3)–O(3)
Mo(4)–O(3)
Mo(4)–O(9)
Mo(4)–O(10)#1
1.748(5)
1.892(5)
1.892(5)
1.926(5)
1.934(5)
2.280(5)
Mo(1)–O(4)
Mo(1)–O(1)
Mo(2)–O(4)
Mo(2)–O(1)#1
Mo(3)–O(1)
Mo(3)–O(4)#1
1.941(4)
1.942(5)
2.002(4)
2.332(4)
2.008(5)
2.329(4)
Mo(1)–O(2)
Mo(1)–O(2)#1
Mo(2)–O(2)
Mo(3)–O(2)
Mo(4)–O(2)
2.136(4)
2.356(4)
2.334(4)
2.318(4)
2.455(4)
Mo(1)–O(5)
Mo(2)–O(6)
Mo(2)–O(7)
Mo(3)–O(9)
Mo(3)–O(10)
Mo(4)–O(13)
Mo(4)–O(12)
1.694(3)
1.698(3)
1.701(3)
1.701(3)
1.702(3)
1.694(3)
1.698(3)
Mo(1)–O(3)
Mo(2)–O(11)#1
Mo(3)–O(8)
Mo(4)–O(11)
Mo(4)–O(8)
Mo(4)–O(3)#1
1.748(3)
1.899(3)
1.882(3)
1.927(3)
1.934(3)
2.308(3)
Mo(1)–O(4)
Mo(1)–O(1)
Mo(2)–O(4)#1
Mo(2)–O(1)
Mo(3)–O(1)
Mo(3)–O(4)#1
1.962(3)
1.951(3)
2.002(3)
2.310(3)
2.001(3)
2.299(3)
Mo(1)–O(2)#1
Mo(1)–O(2)
Mo(2)–O(2)
Mo(3)–O(2)#1
Mo(4)–O(2)#1
2.135(2)
2.352(2)
2.376(2)
2.353(2)
2.388(2)
a Symmetry codes: #1 − x, −y, −z + 2 (1), #1 − x, −y + 2, −z + 2 (3).
for terminal MovO is 1.697 Å. The bridging Mo–O–Mo (with Epoxidation of olefins
two coordinated oxygen) bonds have an averaged distance of
Compounds 1–3 were examined as catalysts for the epoxidation
of different olefins (including cyclooctene, cyclohexene,
styrene, 1-hexene and 1-dodecene) using 30% hydrogen per-
oxide as oxidant. Details concerning the catalytic reaction are
given in the Experimental section. Blank reactions show no
cyclooctene conversion with excess amounts of H2O2 in the
absence of catalyst at 60 °C.
The reaction was first carried out by using compound 1 as
catalyst and cyclooctene as standard substrate in different sol-
vents, at various temperatures and different amounts of
oxidant or catalyst. Data for cyclooctene conversion and
epoxide yield can be found in Tables 3 and 4. Data in entries
1–4 and 8 of Table 3 reflect solvent effects on catalytic per-
formance: EtOH- and CH3CN-mediated reactions resulted in
the best conversions and epoxide yields, whereas EtOAc,
benzene and toluene only gave a medium or low conversion
and yield. It was observed that compound 1 was insoluble
in all solvents examined at 60 °C. However, after adding
1.911 Å, the triple coordinated oxygen atoms exhibit a Mo–O
distance 1.942 Å, the five-coordinated oxygens exhibit an aver-
aged Mo–O bond distance of 2.336 Å. Beside these well-
established dominating bond distances, there are some “inter-
mediate” bond lengths of 2.002, 2.008, 2.332, 2.329 and
2.455 Å originating from non-regular three- and five-co-
ordinated oxygen atoms. Anomalous bond distances were
obtained for Mo(1)–O(10)–Mo(4) bridge with big variations
from 1.748 to 2.280 Å. Possibly the intermolecular interaction
with Hmim cations leads to more pronounced variations in
Mo–O bond distances than observed in “normal” [Mo8O26]4−
complexes. In aqueous solution, only three different Mo–O
bond distances are found, namely 1.70, 1.90 and 2.30 Å for
“isolated” [Mo8O26]4− anions.19 These values are in good
agreement with some of the averaged distances (1.697, 1.911
and 2.336 Å) obtained for compound 1. These values can be
characteristic of the most “regular” basic octahedral units. It is
interesting to note that the Mo–O bond stretching vibrations
are in an (almost) linear relationship with the Mo–O bond dis-
tances. It is known that there are several empirical relations
between stretching frequencies (or the derived force constants)
and the bond distances. A Badger type of relation exhibits a
linear correlation between (ν/1000)2 and 1/r3, where ν is the
stretching frequency (cm−1) and r is the bond distance (Å).
Accordingly, the series of IR bands for compound 1 at 946,
910, 840, 710 and 665 cm−1 (Table 1) can be correlated to
1.697, 1.911, 1.942, 2.049 and 2.336 Å Mo–O bond distances,
respectively. Very similar conclusions can be drawn for com-
pound 3, since both the IR spectra and the X-ray structures are
very similar to those of compound 1. Similarly, the Mo–O
stretching Raman frequencies of solid Na2Mo2O7 at 990 (IR),
939, 925, 881, 821 and 768 cm−1 20 can be correlated to the
Mo–O bond distances 1.700, 1.772, 1.832, 1.937, 2.058 and
2.436 Å, respectively.21 It is further noteworthy that the linear
correlation is valid for both the octahedral and tetrahedral
units of the anion structure.
Table 3 Epoxidation of cis-cyclooctene using compound 1 as catalyst
Conv.b
(%)
Yieldc
(%)
Temp.
(°C)
Entrya
Solvent
n(H2O2):n(olefin)
1
2
3
4
5
6
7
8
EtOH
EtOAc
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.0
1.5
2.0
94
45
11
56
51
60
60
90
90
81
98
99
92
44
10d
49
49
57
57
87
87
76
93
97
60
60
60
60
25
40
50
60
70
60
60
60
Benzene
Toluene
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
9
10
11
12
a Reaction conditions: cis-cyclooctene (2 mmol), [Hmim]4Mo8O26
(1.5 mol%), solvent (1 mL), t = 1 h. b Conversion to cyclooctene oxide
was calculated by GC analysis. c Isolated yield after column
chromatography. d Yield was calculated based on GC analysis.
This journal is © The Royal Society of Chemistry 2014
Green Chem.

View Article Online
Paper
Green Chemistry
Table 4 Epoxidation of cis-cyclooctene with different amounts of
compound 1 as catalyst
Catalyst amount
(mol%)
Conv.b
(%)
Yieldc
(%)
Entrya
1
2
3
4
5
6
7
0.25
0.5
1
1.5
2
2.5
3
46
72
88
31
70
83
93
98
97
98
98
100
100
100
a Reaction conditions: cis-cyclooctene (2 mmol), acetonitrile (1 mL),
n(H2O2):n(olefin) = 1.5, t = 1 h at 60 °C. b Conversion to cyclooctene oxide
was calculated by GC analysis. c Isolated yield after column
chromatography.
Fig. 3 Time-dependent yield of cyclooctene epoxide in the presence
■
▲
of compounds 1–3 as catalysts at 60 °C with 1.5 mol% ( , 1; ×, 3; , 2)
○
●
and 0.5 mol% ( , 1; *, 2; , 3) catalyst.
1.2 equiv. of H2O2 (30%) solution, the white catalyst powder detail for their catalytic behaviour for cyclooctene epoxidation,
changed its color to yellow (indicating the formation of a reac- with amounts of both 1.5 and 0.5 mol% of catalyst applied. All
tion product with hydrogen peroxide, most likely the actual compounds examined show quite high catalytic activity and
catalyst/active species) and dissolved very fast, forming a selectivity towards cyclooctene epoxide, no diol being observed
homogeneous solution when using EtOH and CH3CN as according to GC analysis. When adding 1.5 mol% of catalysts
solvents. In the case of EtOAc, benzene or toluene as solvents, 1–3 to the system, >70% cyclooctene epoxide was obtained
the catalyst could not be dissolved in the organic/water bipha- within 2 minutes and ca. 96% after 30 minutes. Both com-
sic system which is formed. This results in only a low conver- pounds 1 and 2 as catalysts lead to almost 100% epoxide yield
sion and epoxide yield. In CH3CN, a self-precipitation of the within 1.5 h. Compound 3 gives an epoxide yield of 93% after
catalyst after the completion of reaction took place, similar to 1.5 hours. The turnover frequencies (TOFs) for compounds 1–3
the phenomenon observed by Xi et al.13 with respect to are 622, 621 and 628 h−1, respectively (calculated after 5 min
[π-C5H5NC16H33]3[PO4(WO3)4] in catalyzed propylene epoxi- reaction time). Due to similar activity with 1.5 mol% of cata-
dation. Such catalysts combine the advantages of both homo- lyst, it is difficult to compare the catalyst performance in more
geneous and heterogeneous reactions. Since the catalyst detail. Therefore, the system was further examined with a
undergoes a solid–liquid–solid phase transfer during the reac- decreased catalyst amount (0.5 mol%). It can be seen from
tion, it can be easily recycled by filtration.
Fig. 3 that compound 1 shows higher activity than compounds
Based on the advantage of self-separation for the catalyst, 2 and 3 in the initial reaction phase, whereas 1 and 2 result in
CH3CN was applied as solvent for further studies. In a further similar epoxide yields after 1.5 hours. The TOFs with an
set of experiments, the temperature effect on catalytic perform- amount of 0.5 mol% of catalyst are in general higher than that
ance was studied. The reaction at 70 °C exhibits the highest with 1.5 mol% of catalyst, with values of 1488 (1), 984 (2) and
cyclooctene conversion and the best isolated epoxide yield 1080 (3) h−1 (calculated after 5 min of the reaction). Despite
(Table 3, entry 9). The corresponding data are slightly lower at compound 3 displaying a higher TOF than of compound 2, the
60 °C. Since the H2O2 molecule may decompose faster at final epoxide yield is lower. Moreover, the epoxide yields do
higher temperatures, it is nevertheless favorable to carry out not increase after 40 minutes for both amounts of catalyst
the reaction at 60 °C. Entries 8, 10–12 in Table 3 indicate the applied for compound 3. Similar to compound 1, compounds
effect of oxidant amount on catalytic performance. 1.0 to 2 and 3 also exhibited a self-separation phenomenon in
2.0 equiv. of H2O2 to substrate was applied in catalysis. It was CH3CN. It was observed that all the catalyst solid dissolved
observed that a molar equivalent amount of H2O2 was not immediately after adding H2O2 to the system at 60 °C; however
enough to complete the oxidation reaction, while on adding precipitation occurred already within 10 minutes for each reac-
≥1.5 equiv. of H2O2, more than 98% conversion can be tion with both 1.5 mol% and 0.5 mol% of catalyst applied,
achieved. Thus at least 1.5 equiv. of H2O2 is necessary to com- indicating a fast reaction rate and that most H2O2 molecules
plete the catalytic reaction.
had been consumed within 10 minutes of starting the reac-
Table 4 shows the effect of the applied catalyst amount on tion. As can also be observed from Fig. 3, all the reactions
the catalytic performance. In general, the higher the amount already give quite high epoxide yields after 2 minutes of reac-
of catalyst applied, the higher are conversion and yield. When tion, and the yields keep increasing rapidly before 10 minutes,
increasing the catalyst amount up to 2 mol%, the reaction and slow down thereafter. Due to the low concentration of
reaches almost 100% cyclooctene conversion, and no diol H2O2 after 10 minutes, the reaction activity decreases signifi-
byproduct was observed in the reactions.
cantly and catalyst precipitation occurs.
Based on the examination of the catalytic performance for
It is well established that {PO4[MO(O2)2]4}3− (M = Mo, W) is
compound 1, all three compounds were further examined in the active species when using PM12O40− based POM as catalyst
Green Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Fig. 5 Infrared spectra of pure compound 1 (a), after the reaction with
H2O2 (b), and recycled compound 1 (c).
Table 5 Epoxidation of different olefins
Fig. 4 ESI-MS (negative mode) measurement of the isolated active
species.
Conv.b
(%)
Yieldc
(%)
Entrya
Olefin
1
2
3
4
Cyclohexene
Styrene
1-Hexene
96
100
99
44d
91e
96
and H2O2 as oxidant for epoxidation.10 However, not much work
has been done on the examination and isolation of catalytic
active species for [Mo8O26]4− anion-based catalysts. Galindo
et al.22 successfully isolated a tetraperoxo-octamolybdate derivative
with the formula [Hdmpz]4[Mo8O22(O2)4(dmpz)2]·2H2O as an
intermediate during the preparation of the oxodiperoxo
species [MoO(O2)2(dmpz)2]. Another reported peroxo-octa-
molybdate is [NH4]4[Mo8O24(O2)2(H2O)2]·4H2O.23 To the best of
our knowledge, these are the only two peroxo-octamolybdates
reported in the literature. Since the isolation of these inter-
mediates is governed by serendipity, an efficient synthetic
route to this compound has still not been established and
complex mixtures were obtained in all cases. The reaction
between compound 1 and H2O2 was studied in order to
examine the catalytic species in our work. FT-IR and negative
ESI-MS measurements were performed to detect the possible
active species towards epoxidation. When adding 30% H2O2 to
a suspension of compound 1 in acetonitrile at 60 °C, the solid
dissolves completely and finally results in a yellow solution,
indicating the formation of an active species. Nevertheless, the
ionic peak representing the peroxo-octamolybdate could not
be detected in the ESI-MS spectra. Instead, a peak indicating
the dissociated oxodiperoxomolybdenum species MoO(O2)2
(m/z = 177.08) was observed, which may be the possible active
species in catalysis (Fig. 4).
1-Dodecene
91
87
a Reaction conditions: cis-cyclooctene (2 mmol), [Hmim]4Mo8O26
(1.5 mol%), acetonitrile (1 mL), n(H2O2):n(olefin) = 1.5, t = 4 h at 60 °C.
b Conversion to cyclooctene oxide was calculated by GC analysis.
c Isolated yield after column chromatography. d 1,2-Cyclohexanediol
was detected as major byproduct. e The main product was benzyl
aldehyde. The selectivity to styrene oxide was less than 2%.
4−
estingly, the cluster structure of Mo8O26 is recovered after
H2O2 is completely consumed. As can be seen from the spec-
trum of recycled compound 1 (Fig. 5c), no vibrational changes
are observed for the recycled compound, indicating good stabi-
lity of the catalyst.
Besides cyclooctene, the catalytic system was also applied to
other olefins (Table 5). Excellent activity for different olefins
can also be achieved, even for terminal olefins such as
1-hexene and 1-dodecene. However, a significant amount of
diol can be detected in the case of the epoxidation of cyclohexene,
while the oxidation of styrene resulted in the formation of
benzyl aldehyde as the main product.
Recycling and reuse of compounds 1–3 as epoxidation
catalysts
Fig. 5 shows the IR spectra of compound 1 (5a), after the
reaction with excess H2O2 (5b), and the recycled compound 1 Owing to the self-precipitation feature of the catalysts,
(5c). It can be observed that the spectrum of the catalyst pre- compounds 1–3 were recycled directly by filtration after
cursor has MovO vibrations at 946 cm−1, and bridging the reaction. Comparing to the recycling process for
Mo–O–Mo vibrations in the range of 665–910 cm−1. However, [(n-C4H9)4N]4[α-Mo8O26]17a reported by Deng et al., no additional
after the reaction with excess H2O2, the bridging Mo–O–Mo organic solvent is necessary to precipitate the solid in our
vibration peaks disappear. Instead, a peroxo-bond absorption system. The recovered solid was just washed with water, and
peak ν(O–O) at 854 cm−1 (stretching vibration) can be observed then used directly as catalyst for the next run. The recycling
(Fig. 5b).13a The disappearance of bridging Mo–O–Mo yields for all the catalysts were higher than 95%. The losses are
vibrations further confirms the dissociation of the cluster largely ascribed to the workup and may be quantitative after
structure, which is in accordance with the ESI-MS data. Inter- some optimization. Fresh cis-cyclooctene and H2O2 were then
This journal is © The Royal Society of Chemistry 2014
Green Chem.

View Article Online
Paper
Green Chemistry
Characteristic data
Compound 1: [C10H19N2]4Mo8O26 (1852.00). Yield, 81% (white
crystal). Elemental analysis calcd: C, 25.93, H, 4.13, N, 6.05;
found: C, 25.82, H, 3.98, N, 6.01. IR (KBr, cm−1): see Table 1.
1H NMR (500 MHz, DMSO-d6, r.t., ppm): δ = 9.13 (s, 1H, mz-
H2), 7.77 (s, 1H, mz-H4), 7.70 (s, 1H, mz-H5), 4.18 (t, 2H, CH2),
3.88 (s, 3H, NCH3), 1.80 (m, 2H, CH2), 1.28 (m, 6H, 3 × CH2),
0.87 (m, 3H, CH3). ESI-MS (methanol, m/z, %): cation, 167.1
+
Fig. 6 Reuse of compounds 1–3 for epoxidation of cis-cyclooctene
at 60 °C.
(C10H19N2 , 100%).
Compound 2: [C11H21N2]4Mo8O26 (1908.88). Yield, 76%
(white crystal). Elemental analysis calcd: C, 27.69, H, 4.44, N,
5.87; found: C, 27.52, H, 4.38, N, 5.88. IR (KBr, cm−1): see
Table 1. 1H NMR (500 MHz, DMSO-d6, r.t., ppm): δ = 7.64
(s, 1H, mz-H4), 7.61 (s, 1H, mz-H5), 4.10 (t, 2H, CH2), 3.75 (s, 3H,
NCH3), 1.70 (m, 2H, CH2), 1.27 (s, 6H, 3 × CH2), 0.87 (t, 3H,
added for a new reaction cycle. Fig. 6 shows the reuse of com-
pounds 1–3 as catalysts for cis-cyclooctene epoxidation at 60 °C
with 1.5 mol% of catalysts for 1 hour. Compounds 1 and 2 can
be reused for at least ten cycles without significant decrease of
yield, proving that this system has persistent activity during
the recycling experiments.
+
CH3). ESI-MS (methanol, m/z, %): cation, 181.1 (C11H21N2 ,
100%).
Compound 3: [C11H18N]4Mo8O26·H2O (1858.59). Yield, 82%
(white crystal). Elemental analysis calcd: C, 28.43, H, 4.01, N,
3.01; found: C, 28.45, H, 3.98, N, 3.04. IR (KBr, cm−1): see
Table 1. 1H NMR (500 MHz, DMSO-d6, r.t., ppm): δ = 9.13
(d, 2H. py-H2,6), 8.62 (t, 1H, py-H4), 8.17 (t, 2H, py-H3,5), 4.61
(t, 2H, CH2), 1.91 (m, 2H, CH2),1.28 (m, 6H, 3 × CH2), 0.85 (t, 3H,
CH3). ESI-MS (methanol, m/z, %): cation, 164.1 (C11H18N+,
100%).
Experimental
General methods
All chemicals were purchased from Sinopharm Chemical
Reagent Co. Ltd of China or Acros and used without further
purification. 1H NMR spectra were recorded with a 500 MHz
Bruker Avance DPX-500 spectrometer. IR spectra were recorded
with a PerkinElmer Frontier FT-IR spectrometer. Catalytic runs
were monitored by GC methods with an Agilent 6890 instru-
ment using a capillary column (30 m × 0.25 mm × 0.25 μm).
Microanalyses of the obtained products were performed with a
Flash EA 1112 series elemental analyzer. Melting points were
determined by an X-6 micro melting point apparatus (Beijing
Tech Instrument Co. Ltd). Thermogravimetric analysis (TGA)
and differential scanning (DSC) analysis were conducted utiliz-
ing a Netzsch-STA 409 PC system, and typically about 10 mg of
General procedure for the epoxidation of olefins
In a typical reaction, compound 1 (55.6 mg, 30 μmol) was
applied as catalyst and added to 1 mL of CH3CN in a reaction
vessel in air. Olefin (2 mmol) was added, followed by the
addition of H2O2 at 60 °C to start the reaction. The course of
the reactions was monitored by quantitative GC analysis.
Samples were taken at regular time intervals, diluted with
acetonitrile, and treated with a catalytic amount of MgSO4 and
MnO2 to remove water and to destroy the unreacted peroxide.
The resulting slurry was filtered and the filtrate injected onto a
GC column. The conversion of olefins and the formation of
epoxides were calculated from calibration curves (r2 > 0.999)
recorded prior to the reaction. For the recycling experiment,
the white solid was filtered after cooling down the reaction
residue to room temperature. The solid was washed with water
3 times and then dried under vacuum to obtain pure recycled
catalyst. Fresh substrate and H2O2 were then added for a new
reaction cycle.
each sample was heated from 25 °C to 1000 °C at 10 K min−1
.
ESI-MS spectra were recorded by an LCQ Fleet ESI mass
spectrometer. The imidazolium or pyridinium bromine salts
were prepared according to the literature.24
Synthesis and characterization
Compounds 1–3 were prepared according to the published pro-
cedures,17b as follows. A solution of Na2MoO4·2H2O (4.8 g,
20 mmol) was dissolved in 30 mL of water, and then the solu-
tion was acidified with HCl (37%, ca. 10 mL) to a pH value of
Single-crystal X-ray structure determinations
4.5. The resulting mixture was refluxed for 1 h. After cooling Details of the X-ray experiment, crystal parameters, data collec-
down the solution to room temperature, imidazolium or pyri- tions and refinements are summarized in Table 6. Single
dinium bromine salts (ca. 10 mmol) were added and a white crystals were mounted on
a Bruker XRDR3M/ESYSTCM
solid was precipitated immediately from the solution. After diffractometer operating at 50 kV and 30 mA equipped with a
30 min, the water solution was evaporated to about 5 mL at MoKα radiation source (λ = 0.71073 Å). Data collections were
100 °C. Then the precipitate was filtered and thoroughly performed at 273(2) K and 188(2) K with a ω/φ diffraction
washed three times successively with water and ethanol. measurement method and reduction was performed using the
Recrystallization of the solid from acetonitrile afforded white SMART and SAINT software with frames of 0.3° oscillation in
crystals.
the θ range 1.5 < θ < 26.2°. The structures were solved by direct
Green Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Table 6 Crystal data and structure refinement for compounds 1 and 3
1
3
Empirical formula
Formula weight
Temperature (K)
C
40H76N8Mo8O26
C44H74Mo8N4O27
1858.59
188(2)
1852.00
273(2)
Wavelength (Å)
Crystal system, space group
Unit cell dimensions
0.71073
0.71073
Monoclinic, P2(1)/n
a = 16.3997(11) Å
b = 10.8045(7) Å
c = 17.5797(12) Å
α = 90°
Monoclinic, P2(1)/n
a = 15.6345 (9) Å
b = 12.1211 (7) Å
c = 16.5893 (10)Å
α = 90°
β = 96.3800(10)°
γ = 90°
3095.7(4)
2, 1.905
1.645
β = 94.6030(10)°
γ = 90°
3133.7(3)
2, 1.970
1.628
Volume (Å3)
Z, calculated density (mg m−3
)
Absorption coefficient (mm−1
F(000)
)
1680
1836
Crystal size (mm)
θ range for data collection (°)
Limiting indices
0.2 × 0.18 × 0.16
1.80 to 26.04
−20 ≤ h ≤ 20
−12 ≤ k ≤ 13
−20 ≤ l ≤ 21
19112/6079 [R(int) = 0.0411]
99.5
0.23 × 0.17 × 0.10
1.72 to 26.01
−19 ≤ h ≤ 17
−14 ≤ k ≤ 14
−18 ≤ l ≤ 20
19651/6162 [R(int) = 0.0390]
99.8
Reflections collected/unique
Completeness to θ = 26.01° (%)
Refinement method
Full-matrix least-squares on F2
6079/30/370
1.041
Full-matrix least-squares on F2
6162/0/381
Data/restraints/parameters
Goodness-of-fit on F2
1.025
Final R indices [I > 2σ(I)]
R1 = 0.0539
wR2 = 0.1392
R1 = 0.0753
R1 = 0.0345
wR2 = 0.0720
R1 = 0.0460
R indices (all data)
wR2 = 0.1552
1.653 and −1.253
wR2 = 0.0773
0.658 and −0.458
Largest diff. peak and hole (e Å−3
)
methods and all non-hydrogen atoms were subjected to aniso- dissociated oxodiperoxo compounds might be the catalytic
tropic refinement by full-matrix least-squares on F2 using the active species for the epoxidation reactions.
SHELXTL package.25
Crystallographic data (excluding structure factors) for the
Acknowledgements
structures reported in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary
publication nos. CCDC-997672 (1) and CCDC-997673 (3).
M.D.Z. and S.L.Z. thank the National Science Foundation of
China (21101085), the National Science & Technology Pillar
Program (2012BAF03B02) and the Program for Liaoning
Excellent Talents in University (LJQ2012031) for financial
Conclusions
Three [Mo8O264−]-based polyoxomolybdate salts of general
support. X.B.L. thanks the “100 Talents” program of Chinese
Academy of Sciences (KJCX2-EW-H05) for financial support.
formula [Hmim]4Mo8O26 (Hmim
= 1-hexyl-3-methylimid-
azolium), [Dhmim]4Mo8O26 (Dhmim = 1,2-dimethyl-3-hexylimid-
azolium) and [Hpy]4Mo8O26·H2O (Hpy = 1-hexylpyridinium)
have been prepared and used to catalyze the epoxidation of
different olefins including cyclooctene, cyclohexene, styrene
and 1-octene. They can be used as efficient catalysts when
using 30% of H2O2 as oxidant and acetonitrile as solvent. In
general, octamolybdate-based catalysts exhibit high catalytic
performance towards the epoxidation of cis-cyclooctene. More
importantly, since these insoluble compounds form soluble
active species after adding oxidant and are able to self-precipi-
tate after the completion of the reaction, the compounds can
be used as self-separating catalysts. Recycling experiments
indicate that the catalysts can be recycled and reused at least
10 times without significant loss of activity, indicating good
stability of the catalysts. Negative ESI-MS data indicate that the
Notes and references
1 K. Weissermel and H. J. Arpe, Industrial Organic Chemistry,
VCH, Weinheim, 3rd edn, 1997.
2 (a) S. Huber, M. Cokoja and F. E. Kühn, J. Organomet.
Chem., 2014, 751, 25–32; (b) S. A. Hauser, M. Cokoja and
F. E. Kühn, Catal. Sci. Technol., 2013, 3, 552–561.
3 (a) N. Mizuno, K. Yamaguchi and K. Kamata, Coord.
Chem. Rev., 2005, 249, 1944–1956; (b) G. Grigoropoulou,
J. H. Clark and J. A. Elings, Green Chem., 2003, 5, 1–7.
4 (a) G. Schindler, C. Walsdorff, R. Koerner and H. Goebbel,
WO, 2007000396, 2007; (b) W. Cheng, X. Wang, G. Li,
X. Guo and S. Zhang, J. Catal., 2008, 255(2), 343–346;
(c) C. Shi, B. Zhu, M. Lin and J. Long, Catal. Lett., 2009, 133
(1–2), 70–75; (d) J. Zhuang, G. Yang, D. Ma, X. Lan, X. Liu,
This journal is © The Royal Society of Chemistry 2014
Green Chem.

View Article Online
Paper
Green Chemistry
X. Han, X. Bao and U. Mueller, Angew. Chem., Int. Ed., 2004, 15 I. I. E. Markovits, W. A. Eger, S. Yue, M. Cokoja,
116(46), 6537–6541.
5 (a) C. C. Romão, F. E. Kühn and W. A. Herrmann, Chem.
Rev., 1997, 97, 3197–3246; (b) F. E. Kühn, A. Scherbaum
C. J. Münchmeyer, B. Zhang, M.-D. Zhou, A. Genest,
J. Mink, S.-L. Zang, N. Rösch and F. E. Kühn, Chem. – Eur.
J., 2013, 19, 5972–5979.
and W. A. Herrmann, J. Organomet. Chem., 2004, 16 (a) Y. Shi, X. Ren, S. Ren, F. Fu, J. Wang and G. Xue,
689, 4149–4164; (c) W. A. Herrmann, A. M. J. Rost,
J. K. M. Mitterpleininger, N. Szesni, W. Sturm,
R. W. Fischer and F. E. Kühn, Angew. Chem., Int. Ed., 2007,
46, 7901–7903; (d) E. Tosh, J. K. M. Mitterpleininger,
A. M. J. Rost, D. Veljanovski, W. A. Herrmann and
F. E. Kühn, Green Chem., 2007, 12, 1296–1298.
6 F. E. Kühn, A. M. Santos and M. Abrantes, Chem. Rev.,
2006, 106, 2455–2475.
7 (a) K. R. Jain, W. A. Herrmann and F. E. Kühn, Coord.
J. Chem. Crystallogr., 2010, 40, 985–988; (b) T. Dong,
F.-W. Chen, M.-H. Cao and C.-W. Hu, Chem. Res. Chin.
Univ., 2011, 27(1), 11–14; (c) S. Ikegami and A. Yagasaki,
Materials, 2009, 2, 869–875; (d) Q. Li, P. Wu, P. Yin,
J. Zhang, L. Shi and Y. Wei, J. Cluster Sci., 2010, 21,
181–186; (e) H. Gao, J. i. Yu, J. Du, H. Niu, J. Wang,
X. Song, W. Zhang and M. Jia, J. Cluster Sci., 2014, 25(5),
1263–1272; (f) D. Xiao, H. An, E. Wang and L. Xu, J. Mol.
Struct., 2005, 738, 217–225.
Chem. Rev., 2008, 252, 556–568; (b) A. Raith, P. Altmann, 17 (a) C. Yang, Q. Jin, H. Zhang, J. Liao, J. Zhu, B. Yu and
M. Cokoja, W. A. Herrmann and F. E. Kühn, Coord. Chem.
Rev., 2010, 254, 608–634.
8 (a) Q. Zhang, S. Zhang and Y. Deng, Green Chem., 2011, 13,
2619–2637; (b) A. Behr, A. J. Vorholt, K. A. Ostrowski and
T. Seidensticker, Green Chem., 2014, 16, 982–1006.
9 (a) N. Mizuno and K. Yamaguchi, Chem. Rev., 2006, 6,
12–22; (b) I. V. Kozhevnikov, Catalysis by Polyoxometalate,
John Wiley & Sons, Chichester, UK, 2002; (c) K. Kamata,
K. Yonehara, Y. Sumida, K. Yamaguchi, S. Hikichi and
N. Mizuno, Science, 2003, 300, 964–966; (d) B. Zhang, S. Li,
A. Pöthig, M. Cokoja, S. L. Zang, W. A. Herrmann and
J. Deng, Green Chem., 2009, 11, 1401–1405;
(b) M.-L. Guo and H.-Z. Li, Green Chem., 2007, 9, 421–423;
(c) S. L. Linguito, X. Zhang, M. Padmanabhan, A. V. Biradar,
T. Xu, T. J. Emge, T. Asefaab and J. Li, New J. Chem., 2013,
37, 2894–2901; (d) C. A. Gamelas, P. Neves, A. C. Gomes,
A. A. Valente, C. C. Romão, I. S. Goncalves and M. Pillinger,
Catal. Lett., 2012, 142, 1218–1224; (e) J. Du, J. Yu, J. Tang,
J. Wang, W. Zhang, W. R. Thiel and M. Jia, Eur. J. Inorg.
Chem., 2011, 2361–2365; (f) C. J. Carrasco, F. Montilla,
E. Álvarez, M. Herbert and A. Galindo, Polyhedron, 2013,
54, 123–130.
F. E. Kühn, Z. Naturforsch. B: Chem. Sci., 2013, 68b, 18 (a) W. J. Kroenke, J. P. Fackler Junior and A. M. Mazany,
587–579; (e) L. R. Graser, S. Jürgens, M. E. Wilhelm, M. Cokoja,
W. A. Herrmann and F. E. Kühn, Z. Naturforsch. B: Chem.
Sci., 2013, 68b, 1138–1142.
Inorg. Chem., 1983, 22, 2412; (b) E. M. III McCarron,
J. F. Whitney and D. B. Chase, Inorg. Chem., 1984, 23, 3275–
3280; (c) M. Michalec, K. Stadnicka and S. A. Hodorowicz,
Cryst. Res. Technol., 2007, 42, 91–97.
10 (a) Y. Qiao and Z. Hou, Curr. Org. Chem., 2009, 13,
1347–1365; (b) Y. Ishii, K. Yamawaki, T. Yoshida, T. Ura 19 D. G. Lyxel, L. Pettersson and I. Persson, Inorg. Chem.,
and M. Ogawa, J. Org. Chem., 1987, 52, 1868–1870. 2001, 40, 584–592.
11 (a) C. Venturello, E. Alneri and M. Ricci, J. Org. Chem., 20 G. D. Saraiva, W. Paraguassu, M. Maczka, P. T. C. Freire,
1983, 48, 3831–3833; (b) C. Venturello and R. D. Aloiso,
J. Org. Chem., 1988, 53, 1553–1557; (c) I. V. Kozhevnikov,
F. F. de Sousa and J. M. Filho, J. Raman Spectrosc., 2011, 42,
1114–1119.
G. P. Mulder, M. C. Steverink-de Zoete and M. G. Oostwal, 21 D. J. Jovanović, I. L. Validžić, M. Mitrić and J. M. Nedeljković,
J. Mol. Catal., 1998, 134, 223–228. J. Am. Ceram. Soc., 2009, 92, 2467–2470.
12 (a) Y. Leng, J. Wang, D. Zhu, X. Ren, H. Ge and L. Shen, 22 M. Herbert, F. Montilla, E. Álvarez and A. Galindo, Dalton
Angew. Chem., Int. Ed., 2009, 48, 168–171; (b) L. Liu, Trans., 2012, 41, 6942–6956.
C. Chen, X. Hu, T. Mohamood, W. Ma, J. Lin and J. Zhao, 23 (a) A. J. Ward, G. J. Arrow, T. Maschmeyer, A. F. Masters,
New J. Chem., 2008, 32, 283–289; (c) B. S. Chhikara,
R. Chandra and V. Tandon, J. Catal., 2005, 230, 436–439.
13 (a) Z. Xi, N. Zhou, Y. Sun and K. Li, Science, 2001, 292,
P. Turner and J. K. Clegg, Acta Crystallogr., Sect. E: Struct.
Rep. Online, 2009, 65, i53; (b) S. Olson and R. Stomberg,
Z. Kristallogr., 1997, 212, 699.
1139–1141; (b) X. Yang, S. Gao and Z. Xi, Org. Process Res. 24 J. G. Huddleston, A. E. Visser, W. M. Reichert,
Dev., 2005, 9, 294–296.
14 (a) H. Li, Z. S. Hou, Y. X. Qiao, B. Feng, Y. Hu, X. R. Wang
H. D. Willauer, G. A. Broker and R. D. Rogers, Green Chem.,
2001, 3, 156.
and X. G. Zhao, Catal. Commun., 2010, 11, 470–475; 25 (a) G. M. Sheldrick, SHELXTL97, University of Göttingen,
(b) A. C. Cole, J. L. Jensen, I. Ntai, K. L. T. Tran, K. J. Weaver,
D. C. Forbes and J. H. Davis, J. Am. Chem. Soc., 2002, 124,
5962–5963.
Germany, 1997; (b) SMART V5.618 Software for the CCD
Detector System, Bruker Analytical X-ray Systems, Inc.,
Madison, WI, 1998.
Green Chem.
This journal is © The Royal Society of Chemistry 2014