Catalytic performance of a boron peroxotungstate complex under homogeneous and heterogeneous conditions

Article abstract of DOI:10.1016/j.cattod.2012.03.071

The preparation and characterization (FT-IR, FT-Raman, ¹¹B MAS NMR, diffuse reflectance, elemental analysis) of a novel boron peroxotungstate (BTBA)₄H[BW₄O₂₄] (BTBA = benzyltributylammonium) is reported, along with its use in the homogeneous oxidation of cis-cyclooctene, geraniol, linalool and (-)-carveol with H ₂O₂ as oxidant and acetonitrile as solvent. High catalytic activity was registered for all the substrates studied under homogeneous conditions, namely 99% of conversion of geraniol after 2 h, 93% for linalool after 5 h, 74% for cis-cyclooctene after 6 h, and 100% for (-)-carveol after 2 h of reaction. Some oxidation studies were carried out with the Venturello complex, [PW₄O₂₄]^3-, in the same conditions. Furthermore, the boron peroxotungstate (BW₄) was immobilized using two different strategies: (a) BW₄ anchored into a functionalized silica (aptesSiO₂) giving BW₄@aptesSiO₂ and (b) BW₄ encapsulated on a metal organic framework, commonly referred as MIL-101, giving BW₄@MIL-101. The catalytic activity of both heterogeneous materials was investigated for geraniol oxidation and the results were compared with those obtained with BW₄ under homogeneous conditions. The encapsulated boron peroxotungstate (BW₄@MIL-101) gave rise to the best results, reaching complete conversion of geraniol after 3 h of reaction and 78% selectivity for 2,3-epoxygeraniol. Additionally, this heterogeneous catalyst could be reused without appreciable loss of catalytic activity, affording similar 2,3-epoxygeraniol selectivity. The heterogeneous catalysts' stability was also investigated after the oxidation reactions by different characterization techniques.

Full text of DOI:10.1016/j.cattod.2012.03.071

Catalysis Today 203 (2013) 87–94
Contents lists available at SciVerse ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Catalytic performance of a boron peroxotungstate complex under homogeneous
and heterogeneous conditions
Isabel C.M.S. Santosa,∗, Salete S. Balulab, Mário M.Q. Simõesa, Luís Cunha-Silvab, M. Grac¸ a P.M.S. Nevesa,
Baltazar de Castrob, Ana M.V. Cavaleiroc, José A.S. Cavaleiroa
a QOPNA, Departamento de Química, Universidade de Aveiro, 3810-193 Aveiro, Portugal
b REQUIMTE, Departamento de Química e Bioquímica, Faculdade de Ciências, Universidade do Porto, 4169-007 Porto, Portugal
c CICECO, Departamento de Química, Universidade de Aveiro, 3810-193 Aveiro, Portugal
a r t i c l e i n f o
a b s t r a c t
The preparation and characterization (FT-IR, FT-Raman, 11B MAS NMR, diffuse reflectance, elemental anal-
ysis) of a novel boron peroxotungstate (BTBA)4H[BW4O24] (BTBA = benzyltributylammonium) is reported,
along with its use in the homogeneous oxidation of cis-cyclooctene, geraniol, linalool and (−)-carveol with
H2O2 as oxidant and acetonitrile as solvent. High catalytic activity was registered for all the substrates
studied under homogeneous conditions, namely 99% of conversion of geraniol after 2 h, 93% for linalool
after 5 h, 74% for cis-cyclooctene after 6 h, and 100% for (−)-carveol after 2 h of reaction. Some oxidation
Article history:
Received 15 November 2011
Received in revised form 2 March 2012
Accepted 22 March 2012
Available online 17 May 2012
Keywords:
Boron
studies were carried out with the Venturello complex, [PW4O24]
3−, in the same conditions.
Furthermore, the boron peroxotungstate (BW4) was immobilized using two different strategies: (a)
BW4 anchored into a functionalized silica (aptesSiO2) giving BW4@aptesSiO2 and (b) BW4 encapsu-
lated on a metal organic framework, commonly referred as MIL-101, giving BW4@MIL-101. The catalytic
activity of both heterogeneous materials was investigated for geraniol oxidation and the results were
compared with those obtained with BW4 under homogeneous conditions. The encapsulated boron per-
oxotungstate (BW4@MIL-101) gave rise to the best results, reaching complete conversion of geraniol after
3 h of reaction and 78% selectivity for 2,3-epoxygeraniol. Additionally, this heterogeneous catalyst could
be reused without appreciable loss of catalytic activity, affording similar 2,3-epoxygeraniol selectivity.
The heterogeneous catalysts’ stability was also investigated after the oxidation reactions by different
characterization techniques.
Peroxotungstate
Oxidation
Hydrogen peroxide
Heterogeneous
Catalysis
© 2012 Elsevier B.V. All rights reserved.
where X = P or As [9], and [M2O3(O2)4]2−, where M = Mo, W [10,11].
Several groups have studied the Venturello peroxocomplexes or
1. Introduction
The transformation of natural compounds into valuable inter-
mediates for organic synthesis is a significant line of study, both
in the laboratory and in industry. In particular, the search for effi-
cient and environmentally clean catalytic oxidative reactions has
been much studied [1,2]. Among the oxidants available, the use
of hydrogen peroxide (H2O2) in the catalytic oxidation of organic
compounds has been gaining special attention since it is a safe,
non-toxic, relatively low-cost reagent and easily manipulated after
the reactions. Additionally, the only reduction product expected is
water [3–6].
It is known that the reaction of hydrogen peroxide with certain
metals (e.g. MoVI, WVI, VV) can yield peroxocomplexes, which are
efficient oxidation catalysts in the presence of H2O2 [7,8], namely
{(XO4)[W(O)(O2)2]4}3−, the so-called Venturello peroxocomplex,
related systems [10–15]. Sergienko has reviewed the structures of
almost all of the W and Mo peroxocomplexes known [16]. However,
to our knowledge, no peroxotungsto- or peroxomolybdocomplex
containing boron have been described.
All the tungsten peroxocomplexes mentioned above have been
tested in the epoxidation of cyclic and linear alkenes with H2O2
under homogeneous conditions [8,11,17,18]. In the particular case
of cis-cyclooctene oxidation with H2O2, this has already been stud-
ied using peroxocomplexes {(XO4)[W(O)(O2)2]4}3−, where X = P
or As, and [W2O3(O2)4]2− as catalysts [10,11,19–21]. The studies
reported so far under homogeneous conditions were carried out
using chlorinated solvents or benzene as solvent, even though 100%
selectivity to 1,2-epoxycycloctane was always obtained, with no
allylic oxidation products as by-products.
However, the oxidation of geraniol, linalool, carveol or other
allylic alcohols with H2O2, catalysed by tungsten peroxocomplexes,
has received much less attention [2,4,11,22–26]. To the best of our
knowledge, the only reports for the catalytic oxidation of geraniol
∗
Corresponding author. Tel.: +351 234 370 200; fax: +351 234 370 084.
E-mail address: ivieira@ua.pt (I.C.M.S. Santos).
0920-5861/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2012.03.071

88
I.C.M.S. Santos et al. / Catalysis Today 203 (2013) 87–94
with H2O2 in a homogeneous medium is one where a cetylpyri-
dinium salt of the Venturello complex is used in CHCl3 [22] and our
own work using tretrahexylammonium salt of [W2O3(O2)4]2− in
CH3CN [11]. Mizuno and co-workers used the dodecyltrimethylam-
monium salt of [W2O3(O2)4(H2O)2]2− in aqueous solution [23–26]
in the same reaction. In all the previous cases, the main prod-
uct formed is usually 2,3-epoxygeraniol. The oxidation of linalool
was described using only the Venturello complex, giving rise to a
mixture of the corresponding hydroxy furan and hydroxy pyran
derivatives [22]. No reports on the oxidation of (−)-carveol with
H2O2 in the presence of peroxocomplexes were found.
at 337 nm and operating in a reflectron mode. MALDI-MS spec-
tra were acquired using ␣-cyano-hydroxycinnamic acid (Aldrich),
prepared in acetonitrile/(water:trifluoroacetic acid 0.1%) 50/50 as
matrix. The GC–MS analyses were performed on a Finnigan Trace
GC–MS (Thermo Quest CE instruments) using helium as the car-
rier gas (35 cm/s); GC-FID analyses were performed using a Varian
Star 3900 chromatograph and helium as the carrier gas (35 cm/s).
Fused silica capillary columns of the DB-5 type (30 m × 0.25 mm
i.d.; 25 ␮m film thickness) were used, in both cases.
2.2. Synthesis of the catalysts
The immobilization of homogeneous catalysts on solid supports
has attracted remarkable research interest since the resulting het-
erogeneous catalysts have the advantages of being easier to recover
and to recycle. Some reports concerning the immobilization of
peroxotungstates are known, employing inorganic materials such
as Ti-silicalites or layered double hydroxides (LDHs) as supports
[27,28]. Anion exchange of W compounds is another option, since
most of the peroxocomplex catalysts are anionic in their active form
[29]. Recently, some work has been done in the immobilization of
peroxocomplexes via ionic liquid-modified silica [25,30,31].
In the present work, we present a new boron peroxotungstate
(BTBA)4H[BW4O24] (BTBA = benzyltributylammonium) and its use
in the homogeneous oxidation of cis-cyclooctene, and some ter-
penic alcohols with two double bonds (geraniol, linalool and
carveol) with H2O2 as oxidant and acetonitrile as solvent. For com-
parison, linalyl acetate was also oxidized in the same conditions.
The boron peroxotungstate was immobilized using two different
strategies, namely anchored on a functionalized silica material
and encapsulated on a metal organic framework (MOF) commonly
referred as MIL-101 [32–35]. The heterogeneous catalytic activity
of these materials in the oxidation of geraniol was also determined
and compared with the homogeneous performance in similar
conditions.
(C19H34N)4HBW4O24 [BW4]. 12 mL of 30 wt.% H2O2 were added
to an aqueous solution of Na2WO4·2H2O (10 mmol, 20 mL). The pH
of the resulting solution was adjusted with HCl 6 M until the obtain-
ment of a colourless solution (pH = 2.8). To the resulting solution,
H3BO3 in water (0.26 M, 10 mL) was added dropwise, with stirring
during 30 min. Finally, an aqueous solution of benzyltributylam-
monium chloride (BTBACl, 20 mmol, 10 mL) was added. A white
solid was obtained and then filtered, washed and dried in a desic-
cator under vacuum. Yield, 88%. Anal. found (%): W, 33.4; B, 0.48;
C, 40.0; H, 6.62, N, 2.53. Calcd (%): W, 32.9; B, 0.48; C, 40.8; H, 6.13,
N, 2.50. Total weight loss found: 56.8%; calcd: 55.9%. FT-IR (cm−1):
954, 868, 840, 727, 703, 640, 569. FT-Raman (cm−1): 1006, 964, 912,
852, 622, 566, 335, 302, 258. MALDI-MS: 2236.4 Da, corresponding
to the expected formula (C19H34N)4HBW4O24
.
(C24H52N)3PW4O24 [PW4]. This compound was prepared by
an adaptation of a described procedure [10,36]. A solution of
Na2WO4·2H2O (3.76 g, 10.0 mmol) in water (15 mL) and 30% H2O2
(7 mL) was heated for 4 h at 60 ◦C, giving a colourless solution.
Then an aqueous solution of Na2HPO4 (0.36 g; 2.5 mmol in 5 mL of
H2O) was added. After cooling the solution to room temperature, a
solution of [N(C6H13)]Cl (1.56 g, 4.0 mmol) in toluene (40 mL) was
added dropwise. The results obtained for the characterization of
this compound are in agreement with those previously published
[10,36].
2. Experimental
MIL-101. This solid support was prepared by the adaptation
of the original method reported by Férey et al. [32]. A mixture
containing chromium(III) nitrate [Cr(NO3)3·9H2O, 2 mmol], 1,4-
benzenedicarboxylic acid (H2bdc, 2 mmol) and hydrofluoric acid
(100 ␮L) in H2O was stirred at room temperature. The resulting
suspension was transferred to an autoclave and heated at 493 K
for 9 h. After cooling, a significant amount of H2bdc (as crystals)
was observed with the desired material (green thin powder). The
purification of the MIL-101 material was performed by a double
DMF treatment, followed by a double treatment in ethanol. FT-IR
(cm−1): 3465, 1670, 1621, 1508, 1403, 1018, 748, 665, 588, 387.
FT-Raman (cm−1): 3074, 2935, 1611, 1458, 1146, 1040, 872, 812,
632.
2.1. Reagents and methods
Hydrogen peroxide (30 wt.% solution in water) was purchased
from Riedel-de-Haën, whereas cis-cyclooctene, geraniol, linalool,
and (−)-carveol were purchased from Sigma–Aldrich. All other
chemicals and solvents were used as received or distilled and
dried using standard procedures. Silica with surface area of
590–690 m2/g was used as received from Sigma–Aldrich.
Elemental analyses for W and B were performed by ICP spec-
trometry (University of Aveiro, Central Laboratory of Analysis)
and C, H, N elemental analyses were performed on a Leco CHNS-
932 apparatus. Weight loss was determined by thermogravimetric
analysis performed between 30 and 700 ◦C at 5 ◦C/min on a TGA-
50 Shimadzu thermobalance. Infrared absorption spectra were
recorded on a Mattson 7000 FTIR spectrometer, using KBr pellets.
Raman spectra were measured using a Brüker RFS100/s FT-Raman
spectrometer (Nd:YAG laser 1064 nm excitation). Powder X-ray
diffraction (XRD) was performed on powders deposited on sili-
con substrates, using a Philips X’Pert instrument operating with
Immobilization of BW4. Composite materials based on BW4 were
prepared by its immobilization in two different solid supports: sil-
ica (SiO2) and metal-organic framework (MIL-101):
BW4@aptesSiO2. To immobilize BW4 on SiO2, the surface of SiO2
was initially modified by the grafting methodology, adapted from
the literature [37]. The dried SiO2 (1 g) was refluxed for 24 h in
dry toluene (30 mL) with 3 mmol of 3-aminopropyltriethoxysilane
(aptes) under argon. The functionalized support aptesSiO2 was
then treated with nitric acid to protonate the amine groups on the
surface of the material. The protonated material is subsequently
washed, filtered and dried. The immobilization of the boron per-
oxotungstate was performed by stirring the mixture of treated
0.3 g of treated aptesSiO2 in 20 mL of acetonitrile with 30 mg BW4
for 24 h at room temperature. The resulting solid was filtered,
washed with acetonitrile and dried at room temperature. Anal.
found (%): W, 4.25; B, 0.06; Si, 30,6; loading of boron peroxo-
tungstate: 0.058 mmol per 1 g.
Cu-K␣ radiation (ꢀ = 1.54178 A) at 40 kV/50 mA. 11B MAS NMR
˚
spectra were recorded with a Bruker Avance 400 spectrometer
at 128.37 MHz, using spinning rates of 14 kHz. The 11B MAS NMR
spectra were recorded with 15◦ pulses and 5 s recycle delays. Chem-
ical shifts are quoted in ppm from BF3·O(C2H5)2. Matrix Assisted
Laser Desorption Ionization (MALDI) mass spectra (MS) were
acquired, in positive mode, with a MALDI-TOF/TOF Applied Biosys-
tems 4800 Proteomics Analyser (Applied Biosystems, Framingham,
MA, USA) instrument equipped with a nitrogen laser emitting

I.C.M.S. Santos et al. / Catalysis Today 203 (2013) 87–94
89
Scheme 1.
BW4@MIL-101. The encapsulation of BW4 in MIL-101 was
performed by adapting a published procedure [38]. Initially, an ace-
tonitrile solution of 10 mM of BW4 (50 mL) was prepared and then
0.5 g of MIL-101 was added and stirred for 24 h at room tempera-
ture. After that, the solid was filtrated and washed several times
with acetonitrile. Anal. found (%): W, 12.11; B, 0.17; loading of
boron peroxotungstate: 0.165 mmol per 1 g; FT-IR (cm−1): 3436,
1670, 1619, 1509, 1401, 1018, 954, 840, 750, 665, 582, 385. FT-
Raman (cm−1): 3072, 2931, 1613, 1455, 1144, 1006, 964, 872, 809,
635.
equipped with a magnetic stirrer, using H2O2 (30 wt.%) as oxi-
dant. The oxidation reactions of 1, 3, 4 and 5 were performed at
80 ◦C whereas the oxidation reactions of 2 were done at room
temperature and protected from light. The homogeneous reac-
tions were typically carried out as follows: the substrate and the
required amount of catalyst were dissolved in acetonitrile (1.5 mL)
and stirred; then the required amount of 30% (w/w) aqueous H2O2
was added to the reaction mixture. Aliquots were taken directly
from the reaction mixture and injected into the GC-FID or GC–MS
equipments for analysis of the starting materials and products. Sim-
ilar conditions were used in heterogeneous reactions. The studies
in heterogeneous conditions were performed only with geraniol
(2). The amount of catalyst used was 20 mg for BW4@MIL-101 and
40 mg for BW4@aptesSiO2 and the aliquots were centrifuged and
then injected into the GC-FID or GC–MS equipments. At the end
of the reactions the heterogeneous catalysts were separated from
2.3. Oxidation reaction procedure
The oxidation reactions of cis-cyclooctene (1), geraniol (2),
linalool (3), linalyl acetate (4) and (−)-carveol (5) (Scheme 1)
were carried out in acetonitrile, in a closed 5 mL reaction vessel

90
I.C.M.S. Santos et al. / Catalysis Today 203 (2013) 87–94
6.7
the reaction mixtures by centrifugation and washed with different
solvents (CH3CN and C2H5OH) to remove the remaining substrate
and oxidant, as well as the reaction products. The recovered cata-
lysts were dried under vacuum at room temperature and used in a
new reaction under identical experimental conditions, with read-
justment of all quantities, without changing the molar ratios and
reaction concentrations. The reaction products reported here were
identified as described elsewhere [11,39,40]. In the case of linalool
and (−)-carveol oxidation, the reaction products were identified
by comparison to the GC–MS library. Blank reactions were per-
formed for all substrates, confirming that no oxidation products
are obtained unless the catalyst and H2O2 are both present. Even in
the presence of the supports (aptesSiO2 or MIL-101) and H2O2 no
oxidation products were obtained. Acetamide was never observed
among the reaction products. Thus, the possibility of formation of
peroxyimidic acid, which would function as the oxidizing agent,
was excluded.
-2.6
BW4 - experimental
BW4 - simulated
26
13
0
-13
-26
δ (ppm)
Fig. 1. 11B MAS NMR spectrum of solid compound BW4 (experimental – top and
simulated – bottom). The signal of weak intensity at −7.6 ppm may be due to the
3. Results and discussion
existence of an impurity, most likely from BW12O40
.
3.1. Synthesis and characterization of (C19H34N)4HBW4O24
[BW4]
Dimeric oxo-bridged peroxotungstates containing WO(O2)2
units are formed in aqueous solutions of sodium tungstate and
H2O2 with the decreasing of pH [36,41]. In the presence of some
oxoacids (e.g. H3PO4, H3AsO4), the products can incorporate the
central element of that acid, giving rise to species containing one
or more than one WO(O2)2 units linked to a XOn group, like the PO4
3−
3−
)
in the Venturello complex, {(PO4)[WO(O2)2]4} (or [PW4O24
]
[36]. Besides the choice of the oxoacid and the pH in solution,
another important factor influencing the isolation of the peroxo-
complexes is the countercation chosen. So far, no peroxocomplexes
that incorporate boron in its structure had been reported. In a pre-
vious study it was found that the dimeric species [W2(O)3(O2)4]2−
was precipitated from solutions obtained by mixing aqueous solu-
tions of Na2WO4·2H2O and 30% H2O2 with pH adjustment to 2.8,
followed by the addition of boric acid, and of tetrahexylammonium
chloride [11]. In the present work we have used the same method-
ology of Ref. [11], but using benzyltributylammonium chloride. The
white solid obtained was analysed by infrared, Raman and 11B
solid-state NMR spectroscopy, elemental and thermogravimetric
analyses and all the data are in accordance with a new peroxocom-
Fig. 2. FT-IR spectra of BW4, MIL-101, BW4@MIL-101 and BW4@MIL-101-ac, in the
plex with the following general formulae (C19H34N)4HBW4O24
.
wavenumber region between 400 and 1800 cm−1 (ac = after catalysis).
The mass analysis (MALDI-MS) for BW4 gave the expected ion
[M]+ = 2236.4 Da, which is also consistent with the formula given
above.
evident at 964 and 852 cm−1, respectively. The v(B–O) vibration
appears at 912 cm−1, which is in agreement with the literature data
for this bands [45].
It was not possible to obtain single crystals of BW4, in order to
identify the structure of the new compound synthesized. The corre-
sponding powder X-ray diffractogram shows the high crystallinity
of the sample, since most of the observed peaks are well defined
and have high intensity, as shown in Fig. 4. However, the identifi-
cation was not possible by comparison of the XRD pattern with the
indexed ones in the database [46].
The presence of boron was confirmed by elemental analysis
and 11B solid MAS NMR spectroscopy (Fig. 1). It is possible to
observe a second-order quadrupole powder pattern, which was
simulated yielding an isotropic chemical shift, ıiso = 10.4 ppm, and
a quadrupole coupling constant CQ = 1.201 MHz. These results show
that boron is in a relatively distorted BO4 environment [42,43].
The peroxocomplexes have characteristic FT-IR spectra, pre-
senting bands in the range of 500–650 cm−1, attributable to
asymmetric and symmetric vibrations of the peroxo bonds [W(O2)]
[12,13]. BW4 shows the peroxo W(O2) bands at 590 and 640 cm−1
(Fig. 2). It is also possible to observe the W O and O–O vibrations
at 954 and 868, 840 cm−1, respectively. The characteristic band of
the v(B–O), in polyoxometalate anions, is usually overlapped with
the v(W–Ob–W) vibration, and this band typically appears near
920 cm−1 [44]. BW4 shows a weak band at 921 cm−1, that is likely
to be attributed to the v(B–O) vibration. The FT-Raman spectrum
of BW4 (Fig. 3) confirms the presence of the peroxo bond [W(O2)]
at 622 and 566 cm−1, and the W O and O–O vibrations are also
3.2. Synthesis and characterization of the supported catalysts
The peroxotungstate BW4 was immobilized following two
different strategies: (i) immobilization on a silica support func-
tionalized with 3-aminopropyltriethoxysilane (BW4@aptesSiO2);
(ii) encapsulation of BW4 into the nanocages of a metal-organic-
framework MIL-101 (BW4@MIL-101). From the elemental analyses
it is possible to determine average loadings of 0.058 and

I.C.M.S. Santos et al. / Catalysis Today 203 (2013) 87–94
91
Fig. 3 confirms the presence of BW4 in this material. In the
BW4@MIL-101 FT-Raman spectrum it is possible to observe the
presence of the support and the two small bands near 1006 and
964 cm−1 can be attributed to the BW4 peroxocomplex. These
results, together with the FT-IR data and the elemental analysis,
suggest that the encapsulation of the BW4 into the MIL-101 support
was effective.
The powder XRD pattern of the material BW4@MIL-101 (Fig. 4,
top) indicates that the crystalline structure of the solid support was
retained after the encapsulation of the peroxo complex, since the
main diffraction peaks are maintained unchanged. As expected,
the crystalline structure of BW4 is completely transformed dur-
ing the insertion into the MIL-101, since no diffraction peaks of
the complex are visible in the diffractogram of BW4@MIL-101. This
observation confirms that all the peroxo complex present in the
material are in non-crystalline form, most probably due to encap-
sulation in the MIL-101 cavities.
3.3. Homogeneous catalysis
The oxidation of cis-cyclooctene (1), geraniol (2), linalool (3),
linalyl acetate (4) and (−)-carveol (5) (Scheme 1 and Table 1)
with H2O2 were performed in the presence of catalytic amounts of
the peroxocomplexes BW4 and PW4, using acetonitrile as solvent.
These substrates (except 4) were chosen to evaluate the catalytic
activity of the new peroxocomplex BW4 towards the epoxidation
of an alkene and allylic alcohols with hydrogen peroxide, given
that peroxocomplexes are known as good epoxidation catalysts.
Under the conditions used, cis-cyclooctene gives only the corre-
sponding epoxide (1a), geraniol gives mainly the 2,3-epoxide (2a),
linalool affords predominantly the furan hydroxyl ether (3a), the
6,7-epoxide (4a) is the main product for linalyl acetate oxidation
and (−)-carveol affords carvone (5a) and 2,3-epoxide (5b) as the
major products (Scheme 1). For comparison, the Venturello com-
plex, (C24H52N)3PW4O24, was used in similar conditions.
The first stage on the catalytic studies was to evaluate the
oxidation reaction of cis-cyclooctene (1, Scheme 1), as a model sub-
strate. The conversion of cis-cyclooctene in the presence of BW4
reached 76% after 6 h of reaction at 80 ◦C, achieving only the 1,2-
epoxycyclooctane, corresponding to a turnover number (TON) of
456 (Table 1). This result is less interesting than that obtained with
[W2O3(O2)4]2− in similar conditions [11].
Fig. 3. FT-Raman spectra of BW4, MIL-101, BW4@MIL-101 and BW4@MIL-101-ac,
in the wavenumber region between 500 and 1700 cm−1 (ac = after catalysis).
BW4@MIL-101
MIL-101
BW4
Geraniol (2, Scheme 1), linalool (3, Scheme 1) and carveol (5,
Scheme 1) are allylic alcohols that offer several possible places of
oxidative attack, namely at the two double bonds. These three dif-
ferent alcohols allow us to evaluate how the position of the hydroxy
group can influence the outcome of the oxidation reactions.
The oxidation of geraniol took place at room temperature. The
GC analyses of the reaction mixtures (Table 1) show that 2,3-
epoxygeraniol (2a) was the main product, with selectivity between
76% and 90% depending on the amount of H2O2 used and reaction
time, and conversion always above 95%. Two other unidentified
products were obtained, possibly cyclic derivatives [47], together
with a small amount of diepoxide. For the higher conversion, 330
TON and 1022 TOF (h−1, determined after 10 min) were obtained,
after 2 h of reaction.
The regioisomeric product distribution seems to be dominated
by the relative electron density of the two double bonds, that
favours epoxidation of geraniol at the C6–C7 position [22,40],
because electron density at C2–C3 double bond is significantly
depleted by the inductively electron-accepting hydroxyl group
[22]. When preferential epoxidation at the C2–C3 position has been
found, a complex involving the metal centre, the oxidant and the
substrate, that may coordinate through the OH group might be
implicated [22,48,49]. Since 2,3-epoxygeraniol was obtained as the
3
7
11
15
19
23
2θ (º)
Fig. 4. Powder X-ray diffraction patterns for BW4, MIL-101 and BW4@MIL-101.
0.165 mmol per 1 g of material, respectively. In the case of the
immobilization onto the silica support the peroxocomplex is prob-
ably supported by electrostatic interactions with the ammonium
groups. In the case of the MOF support the peroxocomplex appears
to be encapsulated.
In order to verify that the active peroxo species BW4 remain
intact after the immobilization into the supports, vibrational spec-
tra were recorded (FT-IR and FT-Raman spectra). In the case of
BW4@aptesSiO2 both of these techniques were not able to pro-
vide good information as the vibrations of the supporting material
interfered with the typical peroxo vibrations. Fig. 2 shows the FT-IR
spectra of the peroxocomplex, of the MIL-101 and of the supported
material BW4@MIL-101, in the wavenumber region between 400
and 1800 cm−1. When comparing the three spectra it is possible
to find, in the BW4@MIL-101 material, all the bands of the support
and also two extra small bands near 954 and 840 cm−1, which were
attributed to the W O and O–O vibrations, respectively, in the BW4
peroxocomplex.

92
I.C.M.S. Santos et al. / Catalysis Today 203 (2013) 87–94
Table 1
Homogeneous epoxidation of the several substrates with H2O2 in the presence of the peroxocomplexes.
Substrate
Catalyst
Sub:H2O2
Time (h)
Conversion (%)a
TONb
Products selectivity (%)a
cis-Cyclooctenec
1a
BW4
PW4
3:3
3:3
6
6
76
74
456
444
100
100
Geraniold
2a
90
89
76
70
2b
0
0
0
16
2c
1
1
2
3
BW4
BW4
BW4
PW4
1:1
1:2
1:4.5
1:4.5
5
3
2
2
95
96
99
95
316
320
330
316
Linaloole
3a
59
63
3b
17
18
3c
23
0
3d
1
19
BW4
PW4
1:2
1:2
5
5
93
82
310
273
Linalyl Acetatee
4a
90
78
4b
9
19
4c
1
3
BW4
PW4
1:2
1:2
6
6
84
86
280
286
Carveole
5a
56
52
5b
37
24
5c
1
20
BW4
PW4
1:2
1:2
2
2
100
100
333
333
a
Determined by gas chromatography.
b
c
mol of total products/mol of catalyst used.
Reaction conditions: catalyst 5 ␮mol; CH3CN (1.5 mL); cis-cyclooctene 3 mmol; H2O2 30% 3 mmol; 80 ◦C.
d
e
Reaction conditions: catalyst 3 ␮mol; CH3CN (1.5 mL); geraniol 1 mmol; H2O2 30% 4.5 mmol; room temperature and protected from light.
Reaction conditions: catalyst 3 ␮mol; CH3CN (1.5 mL); substrate 1 mmol; H2O2 30% 2 mmol; 80 ◦C.
main product, in the presence of the new peroxocomplex BW4, the
presence of such intermediate may be suggested.
100
80
60
40
20
0
The oxidative reactions of linalool (3, Scheme 1) were carried at
80 ◦C. In this case, the favoured double bond was the C6–C7 position,
since it is through the 6,7-epoxide that the hydroxy furan 3a and
hydroxy pyran 3b are formed [22], indicating that in this case the
OH group does not possibly influence the selectivity. So the main
product obtained in the oxidation of linalool is 3a (Table 1), with
59% of selectivity, after 5 h of reaction, and 93% of conversion (310
TON). In order to elucidate the mechanism of formation of 3a via
6,7-epoxylinalool, the oxidation of linalyl acetate (4) was studied
(Scheme 1). Also in this case the double bond at the C6–C7 position
was favoured, affording mainly the 6,7-epoxide 4a. The cyclization
of 4a cannot occur, due to the acetate group.
The oxidative reactions of (−)-carveol (5, Scheme 1) were car-
ried at 80 ◦C. In this case, the oxidation products occurred with 100%
conversion after 2 h and originated predominantly carvone (5a) and
2,3-epoxide (5b). As in the case of geraniol, the epoxidation at the
double bond nearer to the hydroxyl group is preferred.
BW4@MIL-101
BW4@aptesSiO2
BW4
0
1
2
3
Time (h)
Fig. 5. Kinetic profiles for the oxidation of geraniol with H2O2 in the presence of
homogeneous and heterogeneous catalysts.
Many papers have described the use of peroxotungstate based
catalysts for epoxidations of allylic and homoallylic alcohols with
H2O2. The obtained products are, generally, those indicated in
Scheme 1 but conversion and selectivities are much dependent
on the reaction conditions, namely solvent, concentrations, and
H2O2/substrate molar ratio. Mizuno et al. have presented and
reviewed the results obtained in several papers [8,17,18,50]. In par-
ticular, hydroxy furan 3a and hydroxy pyran 3b were previously
obtained in the presence of a peroxotungstophosphate plus hydro-
gen peroxide [22]. The results obtained for geraniol in Table 1 show
that the increase in the H2O2/substrate molar ratio leads to higher
conversion in a shorter time, but with lower selectivity. In identical
conditions, the conversion and selectivity for 2a are slightly higher
in reactions with BW4 than with PW4.
The studies carried out with the Venturello complex, PW4, in
order to compare the catalytic performance of PW4 and BW4, were
carried out in the conditions for which BW4 had originated the
higher conversion values. The results are presented in Table 1. It is
possible to see that the obtained conversions for all the substrates
were not very different, being equal or marginally higher for BW4
than for PW4 for the allylic alcohols. BW4 or PW4 give rise only to
some differences in the selectivities, even though the main prod-
ucts are always the same. In particular, it may be noted that the
yield of the 6,7-epoxy 2b and 5c seems to be higher in the presence
of PW4 than of BW4, and that the yield of the diepoxy 3c is signifi-
cant for BW4 but null for PW4. As the outcome of the reaction may
depend on several variables, no explanations for these facts were
attempted. In spite of these differences, the behaviour of BW4 and
PW4 are, overall, similar.
3.4. Heterogeneous catalysis
Heterogeneous catalysts based on peroxotungstate BW4 were
tested for the reaction where homogeneous BW4 presented higher
catalytic performance, namely geraniol epoxidation. The reac-
tion conditions were similar to that used for the homogeneous
system. The catalytic performance of the two heterogeneous cata-
lysts was investigated. Fig. 5 compares the kinetic profiles for the
homogeneous BW4 system and for the two heterogeneous cata-
lysts: BW4@aptesSiO2 and BW4@MIL-101. The main difference of

I.C.M.S. Santos et al. / Catalysis Today 203 (2013) 87–94
93
(instead of 89%). After 3 h of reaction, when the oxidation of geraniol
was complete in the presence of BW4@MIL-101, only 80% of con-
version was achieved when the solid was removed. Similarly, for
BW4@aptesSiO2 system, the conversion was 56% after 2 h instead
of 67% in the presence of heterogeneous catalyst. These results may
suggest that a partial leaching of the active centres (BW4) occurs
from both supports and are lost for the solution. Both method-
ologies of immobilization are based on electrostatic interaction or
physical absorption. This leaching phenomena was confirmed by
elemental analysis, since the percentages found, after three cycles,
in the case of BW4@MIL-101 were W, 6.46; B, 0.09, and in the case of
BW4@aptesSiO2 were W, 1.31; B, 0.02; Si, 29.1, which correspond to
a leaching of the BW4 peroxocomplex of 47% and 69%, respectively.
4. Conclusion
Fig. 6. Conversion and selectivity for 2,3-epoxygeraniol for the homogeneous sys-
tem and different recycling cycles for the oxidation of geraniol after 2 h of reaction,
for BW4@MIL-101.
A new boron peroxotungstate (BTBA)4H[BW4O24] was pre-
pared and characterized (FT-IR, FT-Raman, 11B MAS NMR, diffuse
reflectance, MALDI-MS, elemental analysis). This boron peroxo-
tungstate (BW4) was tested in the homogeneous oxidation of
cis-cyclooctene, geraniol, linalool and (−)-carveol with H2O2 as
oxidant and acetonitrile as solvent. High catalytic activity was regis-
tered for all the substrates studied under homogeneous conditions.
Complete conversion of geraniol and carveol was achieved after 2 h
of reaction at room temperature and 80 ◦C, respectively, whereas
conversion of linalool reached 93% after 5 h. The selectivity for the
products obtained is related with the position of the hydroxyl group
in the substrate. For both geraniol and carveol, the epoxidation at
the double bond nearer to the hydroxyl group is preferred. This is
attributed to the possible formation of an intermediate involving
the OH group, which is not favourable for linalool.
these two heterogeneous catalysts consists on the support where
BW4 was immobilized and also on the methodology of immo-
bilization, which led to different loadings. In the homogeneous
and heterogeneous systems the selectivity was similar and 2,3-
epoxygeraniol (2a) was the main product (78 and 75% of selectivity
with BW4@MIL-101 and BW4@aptesSiO2, respectively). A slightly
increase in yield for the main product was also observed with reac-
tion time. A complete conversion of geraniol was observed for
the homogeneous system after 2 h of reaction and the same was
achieved in the presence of BW4@MIL-101 after 3 h of reaction. At
this time, the BW4@aptesSiO2 system gives 75% of conversion, in
the conditions used. Turnover frequencies (TOF/h−1) were calcu-
lated after 10 min of reaction and the higher result was obtained
for BW4 (1022 TOF), followed by the BW4@MIL-101 (501 TOF) and
the lower activity was found to BW4@aptesSiO2 (394 TOF).
The solid catalysts can be easily separated by filtration to be
used in a new reaction under identical experimental conditions. The
catalysts were recycled in order to test their stability and activity in
consecutive reaction cycles. Fig. 6 illustrates the reusability of the
heterogeneous system BW4@MIL-101 for geraniol oxidation with
H2O2 after 2 h of reaction. It can be seen that this catalyst is active for
at least three consecutive cycles and almost complete conversion
was found also for the second reaction cycle. Similar behaviour was
observed for BW4@aptesSiO2 for the first two reaction cycles. The
conversion obtained after 2 h of reaction is 65% for the first cycle,
56% for the second cycle and 25% for the third cycle. The selectivity
for different cycles is analogous, showing a slightly increase during
reaction time for the 2,3-epoxygeraniol.
The same substrates were oxidized in the presence of the
3−
Venturello complex, {(PO4)[W(O)(O2)2]4}
(PW4), in the same
conditions used for BW4. The results obtained with these two per-
oxotungatates were not very different. Similar conversion values
were obtained with both catalysts, but selectivities for the different
products varied to some extent. The novel boron peroxotungstate
(BW4) was immobilizedusingtwo differentstrategies:(a)anchored
into functionalized silica (aptesSiO2) giving BW4@aptesSiO2 and (b)
encapsulated on a metal organic framework, commonly referred
as MIL-101, giving BW4@MIL-101. The catalytic activity of both
heterogeneous materials was investigated for geraniol oxida-
tion and the results were compared with those obtained with
BW4 under similar, but homogeneous conditions. The encapsu-
lated boron peroxotungstate (BW4@MIL-101) gives rise to the
best results, reaching complete conversion of geraniol after 3 h of
reaction, along with 78% selectivity for 2,3-epoxygeraniol. Further-
more, BW4@MIL-101 could be reused without appreciable loss of
catalytic activity, affording similar 2,3-epoxygeraniol selectivity.
BW4@aptesSiO2 and BW4@MIL-101 were characterized after three
catalytic cycles and its stability was also investigated by different
characterization techniques, such as FT-IR and FT-Raman, which
indicate absence of major catalyst decomposition.
The
heterogeneous
catalysts
BW4@MIL-101
and
BW4@aptesSiO2 recovered after the three cycles were char-
acterized by different techniques, in order to check their stability
after reaction. The stability was confirmed by FT-IR and FT-Raman
spectroscopy, where the characteristic bands can be observed
[Fig. 2, BW4@MIL-101-ac; Fig. 3, BW4@MIL-101-ac], the pattern
being the same before and after catalysis.
In order to further investigate the catalytic activity of the het-
erogeneous catalysts versus the putative homogeneous reaction
after leaching of the active species from the support, the solid cat-
alyst was separated from the reaction mixture by careful filtration
after 10 min of reaction for BW4@MIL-101 system and after 1 h for
BW4@aptesSiO2. These times were chosen taking into account that
conversion should be higher than 20%. The geraniol conversion
at the end of the reaction (2 h for both catalysts: BW4@MIL-101
and BW4@aptesSiO2) decreases after solid filtration, for both sys-
tems. For BW4@MIL-101, the conversion after 2 h was only 64%
Acknowledgements
Special thanks are due to Professor João Rocha for his help with
11B MAS NMR and to Dr. Rosário Domingues for the MALDI-MS
spectra. Thanks are due to Fundac¸ ão para a Ciência e a Tecnologia
(FCT, Portugal) and FEDER for funding QOPNA through the Project
PEst-C/QUI/UI0062/2011, for REQUIMTE financial support through
the project PEst-C/EQB/LA0006/2011, and for CICECO financial sup-
port, and the R&D project PTDC/CTM/100357/2008.

94
I.C.M.S. Santos et al. / Catalysis Today 203 (2013) 87–94
References
[26] N. Mizuno, S. Hikichi, K. Yamaguchi, S. Uchida, Y. Nakagawa, K. Uehara, K.
Kamata, Catalysis Today 117 (2006) 32–36.
[27] E. Duprey, J. Maquet, P.P. Man, J.-M. Manoli, M. Delamar, J.M. Brégeault, Applied
Catalysis A 128 (1995) 89–96.
[28] B.F. Sels, D.E. De Vos, P.A. Jacobs, Tetrahedron Letters 37 (1996) 8557–8560.
[29] B.F. Sels, A.L. Villa, D. Hoegaerts, D.E. De Vos, P.A. Jacobs, Topics in Catalysis 13
(2000) 223–229.
[30] S.X. Ying, W.J. Fa, Journal of Molecular Catalysis A: Chemical 280 (2008)
142–147.
[31] X. Shi, X. Han, W. Ma, J. Wei, J. Li, Q. Zhang, Z. Chen, Journal of Molecular Catalysis
A: Chemical 341 (2011) 57–62.
[32] G. Férey, C. Mellot-Draznieks, C. Serre, F. Millange, J. Dutour, S. Surble, I. Mar-
giolaki, Science 309 (2005) 2040–2042.
[33] N.V. Maksimchuk, M.N. Timofeeva, M.S. Melgunov, A.N. Shmakov, Y.A. Chesa-
lov, D.N. Dybtsev, V.P. Fedin, O.A. Kholdeeva, Journal of Catalysis 257 (2008)
315–323.
[34] Y. Zhang, V. Degirmenci, C. Li, E.J.M. Hensen, ChemSusChem 4 (2011) 59–64.
[35] R. Canioni, C. Roch-Marchal, F. Sécheresse, P. Horcajada, C. Serre, M. Hardi-Dan,
G. Férey, J.-M. Grenèche, F. Lefebvre, J.-S. Chang, Y.-K. Hwang, O. Lebedev, S.
Turner, G. Van Tendeloo, Journal of Materials Chemistry 21 (2011) 1226–1233.
[36] C. Venturello, R. D’Aloisio, Journal of Organic Chemistry 53 (1988) 1553–1557.
[37] A. Bordoloi, F. Lefebvre, S.B. Halligudi, Journal of Catalysis 247 (2007) 166–175.
[38] Z. Hou, N. Theyssen, A. Brinkmann, K.V. Klementiev, W. Grünert, M. Bühl, W.
Schmidt, B. Spliethoff, B. Tesche, C. Weidenthaler, W. Leitner, Journal of Catal-
ysis 257 (2008) 315–323.
[39] I.C.M.S. Santos, M.M.Q. Simo˜es, M.M.M.S. Pereira, R.R.L. Martins, M.G.P.M.S.
Neves, J.A.S. Cavaleiro, A.M.V. Cavaleiro, Journal of Molecular Catalysis A: Chem-
ical 195 (2003) 253–262.
[40] R.R.L. Martins, M. Neves, A.J.D. Silvestre, M.M.Q. Simoes, A.M.S. Silva, A.C. Tome,
J.A.S. Cavaleiro, P. Tagliatesta, C. Crestini, Journal of Molecular Catalysis A:
Chemical 172 (2001) 33–42.
[41] M.H. Dickman, M.T. Pope, Chemical Reviews 94 (1994) 569–584.
[42] A.R. Couto, C.N. Trovão, J. Rocha, A.M.V. Cavaleiro, J.D.P. de Jesus, Journal of the
Chemical Society, Dalton Transactions (1994) 2585–2586.
[43] K.F.M.G.J. Scholle, W.S. Veeman, Zeolites 5 (1985) 118–122.
[44] C. Rocchiccioli-Deltcheff, M. Fournier, R. Franck, R. Thouvenot, Inorganic Chem-
istry 22 (1983) 207–216.
[45] R. Thouvenot, C. Rocchiccioli-Deltcheff, P. Souchay, C.R. Acad, Comptes Rendus
de L Academie Des Sciences 278 (1974) 455.
[46] International Centre for Diffraction Data PDF-04/Full File 2004.
[47] T.M. Khomenko, D.V. Korchagina, V.A. Barkhash, Russian Journal of Organic
Chemistry 37 (2001) 793–801.
[48] T. Katsuki, in: M. Beller, C. Bolm (Eds.), Transition Metals for Organic Synthesis,
Wiley-VCH, Weinheim, 1998, p. 260.
[49] W.R. Thiel, in: M. Beller, C. Bolm (Eds.), Transition Metals for Organic Synthesis,
Wiley-VCH, Weinheim, 1998, p. 290.
[1] R.A. Sheldon, J.K. Kochi, Metal: Catalyzed Oxidation of Organic Compounds,
Academic Press, New York, 1981.
[2] W. Adam, T. Wirth, Accounts of Chemical Research 32 (1999) 703–710.
[3] G. Strukul, Catalytic Oxidations with Hydrogen Peroxide as Oxidant, Kluwer,
Dordrecht, 1992.
[4] C.W. Jones, Applications of Hydrogen Peroxide and Derivatives, The Royal Soci-
ety of Chemistry, 1999.
[5] W.R. Sanderson, Pure and Applied Chemistry 72 (2000) 1289–1304.
[6] R. Noyori, M. Aoki, K. Sato, Chemical Communications (2003) 1977–1986.
[7] M. Beller, C. Bolm, Transition Metals for Organic Synthesis, vol. 2, Wiley-VCH,
Weinheim, 1998, p. 260.
[8] N. Mizuno, K. Yamaguchi, K. Kamata, Coordination Chemistry Reviews 249
(2005) 1944–1956.
[9] C. Venturello, R. D’Aloisio, J.C.J. Bart, M. Ricci, Journal of Molecular Catalysis 32
(1985) 107–110.
[10] A.C. Dengel, W.P. Griffith, B.C. Parkin, Journal of the Chemical Society, Dalton
Transactions (1993) 2683–2688.
[11] I.C.M.S. Santos, F.A. Almeida Paz, M.M.Q. Simões, M.G.P.M.S. Neves, J.A.S. Cav-
aleiro, J. Klinowski, A.M.V. Cavaleiro, Applied Catalysis A: General 351 (2008)
166–173.
[12] N.J. Campbell, A.C. Dengel, C.J. Edwards, W.J. Griffith, Journal of the Chemical
Society, Dalton Transactions (1989) 1203–1208.
[13] C. Aubry, G. Chottard, N. Platzer, J.M. Brégeault, R. Thouvenot, F. Chauveau, C.
Huet, H. Ledon, Inorganic Chemistry 30 (1991) 4409–4415.
[14] D.C. Duncan, R.C. Chambers, E. Hecht, C.L. Hill, Journal of the American Chemical
Society 117 (1995) 681–691.
[15] L. Salles, J.Y. Piquemal, R. Thouvenot, C. Minot, J.M. Bregeault, Journal of Molec-
ular Catalysis A: Chemical 117 (1997) 375–387.
[16] V.S. Sergienko, Crystallography Reports 53 (2008) 18–46.
[17] K. Kamata, T. Hirano, S. Kuzuya, N. Mizuno, Journal of the American Chemical
Society 131 (2009) 6997–7004.
[18] N. Mizuno, K. Kamata, K. Yamaguchi, Topics in Catalysis 53 (2010) 876–893.
[19] J. Prandi, H.B. Kagan, H. Mimoun, Tetrahedron Letters 27 (1986) 2617–2620.
[20] A.J. Bailey, W.P. Griffith, B.C. Parkin, Journal of the Chemical Society, Dalton
Transactions (1995) 1833–1837.
[21] N.M. Gresley, W.P. Griffith, B.C. Parkin, A.J.P. White, D.J. Williams, Journal of the
Chemical Society, Dalton Transactions (1996) 2039–2045.
[22] S. Sakaguchi, Y. Nishiyama, Y. Ishii, Journal of Organic Chemistry 61 (1996)
5307–5311.
[23] K. Kamata, K. Yamaguchi, S. Hikichi, N. Mizuno, Advanced Synthesis and Catal-
ysis 345 (2003) 1193–1196.
[24] K. Kamata, K. Yamaguchi, N. Mizuno, Chemistry: A European Journal 10 (2004)
4728–4734.
[25] K. Yamaguchi, C. Yoshida, S. Uchida, N. Mizuno, Journal of the American Chem-
ical Society 127 (2005) 530–531.
[50] N. Mizuno, K. Yamaguchi, Chemical Record 6 (2006) 12–22.