Monolacunary K8SiW11O39-Catalyzed Terpenic Alcohols Oxidation with Hydrogen Peroxide

Article abstract of DOI:10.1007/s10562-018-2434-0

Abstract: Lacunar potassium undecasilicotungstate salt catalyst was highly efficient in oxidation reactions of terpenic alcohols with hydrogen peroxide. Carbonylic products were selectively obtained from alcohols such as borneol, nerol, geraniol and β-citronellol. The K₈SiW₁₁O₃₉ catalyst was synthesized in one pot reaction starting from precursor salts (KCl, Na₂WO₄, and Na₂SiO₃). The catalyst salt was characterized by FT-IR, TG/DSC, BET, XRD analyses and potentiometric titration. The main reaction parameters were assessed. Based on experimental data, a reaction pathway was proposed. In borneol oxidation, TON of 2720 was achieved, indicating the high catalytic activity. As far we know, it is the first time where the monolacunar catalyst is used without an additional introduction of metal transition into Keggin anion. A comparison of the catalytic performance of different lacunar silicotungstic acid salts exchanged with different cations was performed. The K₈SiW₁₁O₃₉ catalyst was used without loss activity.

Full text of DOI:10.1007/s10562-018-2434-0

Catalysis Letters
https://doi.org/10.1007/s10562-018-2434-0
Monolacunary K8SiW11O39-Catalyzed Terpenic Alcohols Oxidation
with Hydrogen Peroxide
Márcio José da Silva1 · Pedro Henrique da Silva Andrade1 · Sukarno Olavo Ferreira2 · Castelo Bandane Vilanculo1 ·
Cesar Macedo Oliveira1
Received: 19 February 2018 / Accepted: 20 May 2018
© Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
Lacunar potassium undecasilicotungstate salt catalyst was highly efficient in oxidation reactions of terpenic alcohols with
hydrogen peroxide. Carbonylic products were selectively obtained from alcohols such as borneol, nerol, geraniol and
β-citronellol. The K8SiW11O39 catalyst was synthesized in one pot reaction starting from precursor salts (KCl, Na2WO4,
and Na2SiO3). The catalyst salt was characterized by FT-IR, TG/DSC, BET, XRD analyses and potentiometric titration. The
main reaction parameters were assessed. Based on experimental data, a reaction pathway was proposed. In borneol oxidation,
TON of 2720 was achieved, indicating the high catalytic activity. As far we know, it is the first time where the monolacu-
nar catalyst is used without an additional introduction of metal transition into Keggin anion. A comparison of the catalytic
performance of different lacunar silicotungstic acid salts exchanged with different cations was performed. The K8SiW11O39
catalyst was used without loss activity.
Graphical Abstract
O
O
H
O
H
H
O
+
O
K
W
8Si 11O39
H
O
O
H
2O2
O
+
H
O
H
O
H
O
Keywords Lacunar silicotungstic acid potassium salt · Hydrogen peroxide · Catalytic oxidation · Terpenic alcohols
1 Introduction
* Márcio José da Silva
silvamj2003@ufv.br
Terpenic alcohols are an abundant, renewable and attrac-
tive raw material for synthesis of fine chemicals, drugs
and agrochemicals [1–3]. Especially, its carbonylic deriva-
tives always have organoleptic properties that make them
ingredient of flavours, fragrances and perfumes [4, 5].
1
Chemistry Department, Federal University of Viçosa,
Campus Universitário, Avenue P.H. Rolfs, s/n, Viçosa,
Minas Gerais Zip Code 36570-000, Brazil
2
Physics Department, Federal University of Viçosa, Viçosa,
Minas Gerais Zip Code 36570-000, Brazil
Vol.:(0123456789)
1 3

M. J. da Silva et al.
Industrially, the production of carbonyl products has been
until today carried out with toxic stoichiometric oxidants,
which generate residues that should be disposal into
environment. Green oxidants such as hydrogen peroxide
are a benign environmentally alternative, nonetheless, it
requires the presence of metal catalyst to be activate [6, 7].
Polyoxometalates are a class highly versatile of cata-
lysts with acidic and redox properties that are used in
homogeneous and heterogeneous processes [8, 9]. Several
heteropolyacids are efficient catalysts in oxidation reac-
tions with oxygen or hydrogen peroxide [10]. Among the
polyoxometalates, the Keggin heteropolyacids (HPAs) are
most used in catalysis. The most common Keggin HPAs
are clusters of metal oxygen with anions of general for-
mulae Xn+M12O40n−m(Xn+ =P5+, Si4+; M=Mo, W; m=8).
Keggin HPAs are strong Bronsted acids but have also
redox properties, being thus catalysts used either in acidic
or oxidative processes [11]. However, they have low sur-
face area and are soluble solids in polar solvents, a draw-
back that hamper its use in certain reactions. To overcome
these late disadvantages, the HPA catalysts have been
supported on solid matrixes that have high surface area
[12]. This way, solid supported HPA catalysts have shown
active in acid-catalyzed reactions such esterification, trans-
esterification isomerization, as well as oxidation reactions
[13–17].
transition metal substituted lacunar HPA salt catalysts in
oxidation reactions of alcohols and olefins [22].
Lacunar Keggin type POMs can be easily synthesized
by removal of one W-O or M-O units from octahedral of
the Keggin heteropolyanion (i.e., XM12O40n−; n = 3 when
X = P5+; n = 4 when X = Si4+), giving mono lacunary het-
eropolyanions (XM11O39m−; m = 7 when X = P5+; m = 8
when X = Si4+) [23]. Depends on the pH of synthesis, di
or tri lacunary species can be formed. A common synthesis
route is to promote an increase of pH value of a solution
containing the Keggin heteropolyacid, resulting in the pro-
tons exchange by the metal cations present in the alkaline
medium. The extension of hydrolysis of the Keggin anion is
governed by the pH control [24]. Another process is to start
from precursors salts (i.e., tungstate or molybdate, silicate
or phosphate) and perform the synthesis in one-pot reaction
[25].
Lacunar heteropolyacid salts can contain highly charged
counterions such as small-sized metal cations or cations of
large radium. The firsts are compounds that are soluble in
water and polar organic solvents. Contrariwise, when larger
cations are used in the synthesis, the solid heteropoly salts
are insoluble. Indeed, charge, size, shape, and hydrophobic-
ity are some important features that offer strong interactions
between the ionic components [25, 26].
Song et al Keggin HPA salts synthesized
α-K5PW11O39(MOH2) (M = Mn2+, Co2+, Ni2+, and Zn2+)
and assessed its catalytic activity in liquid phase oxida-
tion of 2-propanol with hydrogen peroxide [27]. They
have found that the conversions followed the trends:
Organic reactions as esterification and oxidation generate
water as by-product; moreover, some reactants are used in
aqueous solutions (i.e., aqueous hydrogen peroxide). These
features may contribute for the solid-supported catalyst
leaching. An alternative approach to circumvent this prob-
lem is convert the Keggin heteropolyacids to solid salts. To
do it, the protons should be replaced by large radius cations
(ca. > 1.30 Å), a characteristic that can give them high sur-
face area and insolubility in polar solvents.
K5PWMn11O39 > K5PWCo11O39 > K5PWNi11O39
>
K5PWZn11O3, which was consequence of that reduction
potential and absorption edge energy of these catalysts.
Patel et al. described the that Keggin-type Cs+ salt of a
mono Mn+2-substituted phosphotungstate (Cs5PMnW11O39)
on styrene oxidation with tertbutyl hydroperoxide [28].
Conversions of 61 and 99% selectivity of benzaldehyde
were achieved. This same research group synthesized
Cs5PNiMo11O39 catalyst and assessed its activity on
Suzuki–Miyaura cross coupling reactions [29]. Although
it is not oxidative process, we highlight that those cesium
catalysts were soluble under reaction conditions (water: ethyl
alcohol mixture). However, those authors developed a pro-
cess to recovery the catalyst. They cooled reaction solution
to room temperature, and then extracted the organic products
with ethyl ether; consequently, the catalyst migrated to the
aqueous solutions. Afterwards, the dryness of the solution
provided the solid catalyst, which was repeatedly reused
without loss activity [29].
Although ammonium alkyl derivative organic cations
have being used, a procedure very common is the total or
partial replacement of the protons by the cesium [18]. The
partial exchange of protons by those cations make insolu-
ble the HPAs, increase its surface area and even retain its
acidity properties. Nonetheless, the level of substitution of
protons by cations such as Cs+ or K+ plays a pivotal rule
on catalytic activity of HPA salts [19]. In some cases, the
Cs+ HPA salts present difficulty to be removed from reac-
tion medium. Thus, they have been supported on silica or
mesoporous materials [19–21].
Another structural modification that allows improve
the activity of HPA catalysts in oxidation reactions is the
removal of a MO unity from the Keggin heteropolyanion
(i.e. M=Mo or W). In general, these lacunar catalysts have
a vacancy that can be filled by metal transition cation, which
potentially may enhance its catalytic activity. Recently,
several works in literature have explored the activity of
Lacunary heteropolyanions can also underwent dimeri-
zation or trimerization generating di or tri-lacunar species.
Mizuno et al. studied the epoxidation of 1-octene with H2O2
catalyzed by a series of tetrabutylammonium silicotungstate
1 3

Monolacunary K8SiW11O39-Catalyzed Terpenic Alcohols Oxidation with Hydrogen…
salts in CH3CN at 305 K [30, 31]. They have found that
whereas the dilacunary salt, which was the most active cata-
Na8SiW11O39) adjusted the pH to 5.5 and this value was
maintained constant for 100 min.
lyst, other mono- and tri-vacant lacunary compounds had
Thus, obtain potassium silicotungstate salt, solid KCl (ca.
90 g) was added to the solution with gentle stirring. After
15 min, the precipitate was filtered through a sintered glass
filter. The catalyst was recrystallized in hot water. If neces-
sary, the insoluble material was quickly removed by filtra-
tion, and the salt was precipitated again by new addition of
solid KCI (ca. 80 g). Finally, the precipitate was filtered,
two times washed (ca. 50 mL portions) with 2 molL−1 KCl
aqueous solution, and air dried. The K8SiW11O39 catalyst
was heated to 473 K for 3 h in air prior the characterization
steps and catalytic tests.
4−
moderate activity. The saturated specie (i.e., SiW12O40
was almost inactive.
)
Ding et al. investigated the activity of dilacunary silico-
tungstate (i.e., K8(γ-SiW10O36)⋅13H2O) in oxidation reac-
tions of alcohols with 30% aqueous hydrogen peroxide [32].
The reaction was carried out in an aqueous/oil biphasic sys-
tem, which allowed easy recovery of a catalyst under reac-
tion mild conditions. High yields of ketones were obtained
at room temperature the dilacunary silicotungstate was five
times recycled without decrease of catalytic activity.
In this work, we carried out the reaction of terpenic alco-
hols in the presence of monolacunar silicotungstate catalyst.
The adequate choice of solvent (dimethylacetamide, DMA),
allowed that this catalyst achieve high conversions and selec-
tivity for carbonylic products. Borneol, a secondary bicyclic
terpenic alcohol important to synthesis of fine chemicals
was selected as model molecule. The main reaction vari-
ables (catalyst load, temperature, reactants stoichiometry)
were assessed. The reaction scope was extended to the other
terpenic alcohols. The catalyst was recovered and reused
without loss activity.
2.3 Characterization of Catalyst
The textural properties of lacunar potassium salt catalyst
were studied by H2 desorption/adsorption using NOVA
1200e High Speed, Automated Surface Area and Pore Size
Analyzer Quantachrome Instruments. Previously, the sam-
ple was degassed by 1 h. The specific surface area was cal-
culated by Brunauer-Emmett-Teller equation applied to the
desorption/ adsorption isotherms.
Powder X-rays diffraction pattern of catalyst was recorded
with a X-ray Diffraction System model D8-Discover Bruker
using Ni filtered Cu-kα radiation (λ=1.5418 Å), working at
40 kV and 40 mA. The measurements were done in intervals
of 0.05° with a counting time of 1.0 s in the 2θ range of
5–80 degrees.
2 Experimental Section
2.1 Chemicals
FT-IR/ ATR spectrum of the lacunar salt was recorded on
an FT-IR Varian 660 spectrometer with reflectance acces-
sory utilizing KBr plates under ambient conditions.
The potentiometric titration (Bel pH meter, model W3B)
measured the amount and the strength of the acid sites as
described by Pizzio et al. [33]. A suspension of catalyst in
CH3CN was stirred by 3 h and then titrated with n-butyl-
amine (ca. 0.05 N, toluene). The acidity of the catalyst meas-
ured by this technique allows us to evaluate the total number
of acid sites as well as their acidic strength.
All chemicals were purchased from commercial sources.
Sodium tungstate, sodium metasilicate were acquired from
Sigma–Aldrich (99 wt%). Potassium chloride and sodium
hydrogen carbonate were Vetec (99 wt%). Borneol (99 wt%,
racemic mixture Sigma Aldrich), β-citronellol (racemic mix-
ture 90–95 wt%, Sigma Aldrich), linalool (Sigma Aldrich,
99 wt%), nerol (Sigma Aldrich, 97 wt%) and geraniol (Sigma
Aldrich, 98 wt%) were used as received. Dimethylacetamide
(DMA) was acquired from Sigma–Aldrich. Hydrocloric acid
(36.5 wt%) was Quimica Moderna. The 34 wt% aqueous
H2O2 (Sigma) was the oxidant in all reactions.
2.4 Catalytic Runs
Typically, a 25 mL three-necked glass flask, equipped with
a sampling system and a reflux condenser was charged with
DMA (ca. 10 mL), terpenic alcohol (ca. 2.0 mmol), and an
adequate amount of K8SiW11O39 catalyst. The reaction was
carried out using magnetic stirring and heating to 353 K
temperature. After added hydrogen peroxide at adequate
molar ratio, the reaction was carried out for over 3 h.
2.2 Synthesis of Monolacunar Potassium
Silicotungstate Salt
The catalyst was synthesized as described in literature [32].
Na2SiO3 (1.1 g, 50 mmol) and Na2WO4 (182 g, 0.55 mol)
were separately dissolved in 100 mL of water. Under mag-
netic stir, the Na2SiO3 solution was poured into Na2WO4
To follow the reaction progress, aliquots were periodi-
cally collected and analyzed in a gas chromatograph (Shi-
madzu, GC 2010, capillary column, FID). The main prod-
ucts were identified through GC-MS analyses (Shimadzu
solution, and the slow addition of HCl (165 mL, 4 molL−1
)
resulted in the H4SiW12O40 acid. The dropwise addition of
NaHCO3 solution converted H4SiW12O40 to lacunar salt (i.e.,
1 3

M. J. da Silva et al.
MS-QP 2010 ultra, mass spectrometer, electronic impact
mode at 70 eV, coupled to a Shimadzu 2010 plus, GC).
Additionally, reaction products were co-injected in GC with
authentic samples.
shows that all these bands are present in FT-IR spectrum of
salt synthesized.
The absorption bands of lacunary salt are shifted to higher
wavenumber if compared to those of precursor heteropoly-
acid. The same phenomenon was observed by Romanelli
et all, when synthesized sodium lacunar salts [34]. The main
absorption bands of lacunar salt and silicotungstic acid are
placed at 991 e 1014 cm− 1 (νas W = O), 943 e 978 cm− 1
(νas Si-Oa), 852 e 910 cm−1 (νas W-Ob-W) e, 787 e 762 (νas
W-Oc-W), respectively.
The influence of main reaction variables (i.e. reactants
stoichiometry, temperature, catalyst load) was investigated.
Blank-reactions were performed for each molar ratio. Bor-
neol was selected as a model molecule. The reaction scope
was extended to the other terpenic alcohols (geraniol, nerol,
β-citronellol, α-terpineol, linalool).
Nonetheless, Pizzio et al concluded that alone FT-IR
it is not enough to confirm the transformation of the
2.5 Recovery and Reuse of the Catalyst
4−
8−
SiW12O40 to SiW11O39 lacunar phase during the salt
synthesis. They have found that the main bands of the latter
are placed at 1008, 959, 946, 882, 869, 794 and 720 cm−1
[33]. Those bands are close of them present in FT-IR of
lacunar potassium salt.
After the end of the reaction, the solid catalyst was three
times washed with dichloromethane to extract the remain-
ing products. After it has been Then it was heating in plate
to near dryness, it was dried at room conditions. The solid
catalyst was weighed and reused in another catalytic run.
3.1.2 Analyses of Powder X-Rays Diffraction Patterns
3 Results and Discussion
3.1 Catalysts Characterization
3.1.1 FT-IR Spectroscopy Analyses
It is important highlight that water molecules number per
unitary cell strongly affects the XRD patterns of HPAs or
its salts. On this sense, Derrick et al recorded data from
hydrates silicotungstic acid and showed the existence of sta-
ble phases of composition H4SiW12O40⋅nH2O where n=24,
14, 6 and 0 [35]. They have prepared H4SiW12O40⋅6H2O
heteropolyacid to 150 °C and through direct comparison of
XRD spectrum we conclude that their data fits well with
those recorded by us.
The literature attributes the main typical absorption bands of
monolacunary salt as follows: 994 cm−1 (Si-Oa), 945 cm−1
(W-Od), 875 cm−1 (W-Ob-W), 805 and 730 cm−1 (W-Oc-W)
[28]. On the other hand, Pizzio and Blanco have found that
FT-IR spectrum of H4SiW12O40 displayed the absorption
bands at 1020, 982, 926, 884 and 778 cm−1 [33]. The Fig. 1
The X-rays diffractograms of precursors (tungstate and
metasilicate), parent silicotungstic acid and the lacunar
potassium silicotungstic salt are presented in Fig. 2. The
main diffraction lines are placed in the region 5< 2θ < 35
degrees. Tungstate has monoclinic symmetry C2/m (Z=4).
90
75
60
45
H4SiW12O40
30
15
K8SiW11O39
1600 1400 1200 1000 800
wavenumber / cm-1
600
Fig. 1 FT-IR spectra of silicotungstic acid and lacunar potassium sili-
Fig. 2 Powder XRD patterns of K8SiW11O39 lacunar salt compared to
cotungstate salt
the precursors tungstate, silicate and the parent H4SiW12O40
1 3

Monolacunary K8SiW11O39-Catalyzed Terpenic Alcohols Oxidation with Hydrogen…
Comparing diffraction patterns of synthesized salt
K8SiW11O39 with literature data (JCPD file no. 34-0205)
we have found that they perfectly correspond to the cubic
phase [22]. The increasing of width of peaks is evidence
of greater size of crystallites. All the three samples were
crystalline in nature.
consequently the unitary cell was also decreased. In addi-
tion, the introduction of potassium cations into heteropoly-
anion structure increased the amount of diffraction lines.
3.1.3 Measurement of Acidity Strength of HPA Catalysts
The powder X-ray diffraction patterns of H4SiW12O40
and K8SiW11O39 were compared (Fig. 2). It was found that
the synthesized salt has similar diffraction lines pattern
to the parent H4SiW12O40, which has main XRD peaks in
the region of 5°<2θ<30°. It is suggestive that the Keggin
anion is present in their framework of lacunar salt. These
lines were shifted to higher 2θ angles in X-rays diffraction
spectrum of lacunar salt. An increasing of values of 2θ
angles means that interplanar distance was decreased and
The plots of the electrode potential as a function of
mEq. N-butylamine per g of the catalyst is shown in
Fig. 3. The acidic strength of surface sites is assigned
as follow: very strong sites (Ei > 100 mV); strong sites
(0<Ei<100 mV); weak sites (100<Ei<0 mV) and very
weak sites (Ei<100 mV) [33].
It is observed that while the silicotungstic acid has very
strong acid sites, the lacunary potassium salt has very weak
acid sites (Fig. 3). Moreover, the profile of titration curve
of lacunar salt quickly drops, indicating the almost protons
were removed during the synthesis.
3.1.4 Analyses of Porosimetry of the Lacunar Potassium
Silicotungstate Salt
The porosity properties of lacunar potassium salt were inves-
tigated. Isotherms of adsorption/ desorption as well as the
distribution, volume and diameter of porous are shown in
Fig. 4.
In general, the surface area of heteropolyacid increases
when the protons are exchanged by potassium cations.
Indeed, while the surface area of H4SiW12O40 was < 3.
0 m2/g, the surface area of lacunar salt was 10.2 m2/g [27].
The isotherms reflect Type II characteristics, which are
typical of not porous solids, suggesting that there was physic
adsorption in multilayers. It was observed the total revers-
ibility, indicated by the absence of hysteresis cycle in the
Fig. 3 Potentiometric titration curves with n-butylamine of
H4SiW12O40 and K8SiW11O39 catalysts
(a)
(b)
Fig. 4 Isotherms of adsorption and desorption (a) and volume and diameters porous (b) of K8SiW11O39 catalyst
1 3

M. J. da Silva et al.
adsorption–desorption isotherms of the lacunary catalyst
[27, 36].
one, attributed to decomposition of Si–O–W framework fol-
lowed by the noticeable peak in DSC curves around 610 °C.
The final products are an oxides mixture (Fig. 5).
The introduction of potassium cation as well as the
removal of an WO unit from Keggin anion resulted in a
decreasing of the decomposition temperature in relation to
the parent heteropolyacid, probably as consequence of loss
symmetry [27].
The pore volume (< 0.02 cm3 g− 1) as well as the pore
sizes showing values in the region between 0.90 and 1.15 nm
confirm that the lacunary potassium silicotungstate catalyst
is microporous. These porosity and isothermal adsorption/
desorption characteristics are supported by the literature data
[33, 36].
3.1.5 Thermal Analyses of Lacunary Potassium
Silicotungstate Salt
3.2 Catalytic Tests
3.2.1 Effect of Stirring Speed
Thermogravimetric analysis of lacunary potassium silico-
tungstate salt shows two regions of loss weight; the first one
before 200 °C assigned to loss of all water molecules (i.e.,
7 mols of water molecules per mol of catalyst). The second
The effect of stirring speed of the reaction solution was pre-
viously checked aiming to avoid the possible issue of mass
transfer limitations. We carried out the reactions with stir-
ring speed of 350, 700 and 1050 rpm. No significant change
was found when the catalyst load was 10 mol% (i.e., the
maximum load used). Therefore, all the runs were carried
with stirring rate of 350 rpm.
3.2.2 Effect of Catalyst Load
Recently, we have found that Na2WO4 catalyst was highly
efficient in oxidation reactions of alcohols with hydrogen
peroxide in DMA solutions [37]. Hida and Nogusa assigned
this efficiency the ability of solvent in stabilize a tetraperoxo-
tungstate intermediate, that is the specie catalytically active
in this reaction [38]. For this reason, the DMA was the sol-
vent on this present work. The effect of catalyst load was
assessed the kinetic curves are displayed in Fig. 6.
In general, the camphor was always the main product
formed in K8SiW11O39-catalyzed oxidation reactions of
Fig. 5 TG-DSC Curves of samples of lacunary catalysts. Black line is
the TG curve and blue line the DSC curve
Fig. 6 Effect of catalyst load on kinetic curves of K8SiW11O39-catalyzed oxidation reactions of borneol with H2O2. Reaction conditions: borneol
(1.34 mmol); H2O2 (4.02 mmol); temperature (333 K); reaction time (3 h); DMA (10 mL)
1 3

Monolacunary K8SiW11O39-Catalyzed Terpenic Alcohols Oxidation with Hydrogen…
borneol with H2O2 (ca. selectivity equal or higher than 90%)
(Scheme 1).
proposed that a specie tetraperoxotungstate is a key-interme-
diate on that reaction [38, 39].
We have found that the catalyst was highly active, conse-
quently, a great number of catalytic runs were performed to
assess the effect of concentration. All the runs performed are
listed in Table 1, including the TON achieved.
In the present work, after the synthesis of lacunar salt as
depicted in the Scheme 2, we suppose that only a tungsten
atom per mol of catalyst can be peroxidized. Therefore, we
can conclude that a peroxotungstate complex is the catalyti-
cally active specie on this oxidation reaction.
The lacunar catalyst was highly active; when a dras-
tic reduction of catalyst concentration of 10–0.125 mol%
was performed, only a little decreasing of conversion was
achieved. Consequently, the TON obtained was very signifi-
cant. When the catalyst concentration was 0.0125 mol%, a
conversion was 34%, however, the TON was 2720, indicat-
ing the high activity of catalyst.
Thus, we propose a reaction pathway were a peroxided
intermediate (1) has a highly electron-deficient tungsten
atom that may undergo nucleophilic attack of borneol
hydroxyl group (Scheme 2).
This attack then results in an opening of highly unsta-
ble three-center ring, where a hydrogen atom of hydroxyl
group of borneol will be transferred to the oxygen atom that
is bonded to the tungsten, giving OOH group and bornyl-
alkoxide (omitted by simplification). It is possible that in this
step both groups remain coordinated to the tungsten atom.
It is very common the formation of species like these (i.e.,
intermediates with peroxide and or alkoxide groups bonded
to tungsten atom) in tungsten-catalyzed oxidations [40–42].
Nonetheless, as we haven’t any experimental evidence cor-
roborating that this step has really occurred into tungsten
coordination sphere, it was not included it in the Scheme 2.
In the last step, the simultaneous elimination of water and
oxidized substrate (i.e., camphor) generate the lacunar heter-
opolyanion that can once more react with one mol of H2O2
and then participate of more one catalytic cycle (Scheme 2).
3.3 Insights on Reaction Pathway
Previously, when we investigated the mechanism of the
Na2WO4-catalyzed borneol oxidation by hydrogen perox-
ide, we have found that the addition of H2O2 to the sodium
tungstate/ DMA solution resulted in the disappearance of
strong band at 931.5 cm−1 wavenumber, which was present
in the FT-IR spectrum of Na2WO4 [37]. It was an evidence
that W=O double bonds that are typical of tungstate anion
were absent in the Na2WO4-catalyzed borneol oxidation by
hydrogen peroxide.
Another evidence of this peroxidation step was the
absorption bands at frequencies lower than 597 cm−1, which
are assigned to the vibrations of tungsten- oxygen single
bonds, typical of peroxotungstate complexes [39–41]. Based
on these findings and in the work of Hida and Nogusa, we
3.4 Assessment Other Variables
of K8SiW11O39‑Catalyzed Borneol Oxidation
Reaction With Hydrogen Peroxide
In oxidation reactions, the proportion substrate: oxidant is
always a key aspect; since the hydrogen peroxide is used in
aqueous solutions, an increase on its load may trigger con-
current reactions or even affect the catalyst activity.
To assess this effect, we carried out reactions with the
proportion of reactants varying of 1:1 to 1:5. The kinetic
curves are displayed in Fig. 7. As demonstrate in Fig. 7b,
with exception of reaction carried out with a stoichiometric
proportion (ca. 1:1), the reaction selectivity was not affected
by increasing on oxidant load. In all other proportions the
Scheme 1 Oxidation of borneol by H2O2 in the presence of
K8SiW11O39 catalyst in DMA solutions
Table 1 Effect of catalyst
Run
Conc. (mol%)
Conv. (%)
TON
Run
Conc. (mol%)
Conv. (%)
TON
load on conversion and TON
of K8SiW11O39-catalyzed
oxidation reactions of borneol
with H2O2
1
2
3
4
5
10.0
7.50
5.00
2.50
1.25
99
99
98
98
97
9.9
13.3
19.8
39.6
77.8
7
8
9
10
11
0.500
0.250
0.125
0.0625
0.0125
97
97
89
42
34
194
760
712
672
2720
Reaction conditions: borneol (1.34 mmol); H2O2 (4.02 mmol); temperature (353 K); reaction time (3 h);
DMA (10 mL)
1 3

M. J. da Silva et al.
Scheme 2 Synthesis of
lacunar salt (a) and prob-
ably reaction pathway (b) of
K8SiW11O39-catalyzed oxida-
tion of borneol by H2O2
(a)
(b)
(a)
(b)
Fig. 7 Effect of oxidant load on kinetic curves (a) and conversion and
camphor selectivity (b) of K8SiW11O39-catalyzed oxidation reactions
of borneol with H2O2. Reaction conditions: borneol (1.34 mmol);
reaction time (3 h); K8SiW11O39 (1.25 mol%); temperature (353 K);
DMA (10 mL)
conversions were equal or higher than 93%. The effects of
temperature on kinetic curves of K8SiW11O39-catalyzed oxi-
dation reactions of borneol with H2O2 are shown in Fig. 8.
An increase of temperature resulted in a higher initial rate
of reaction as well as a higher final conversion.
selectivity was strongly affected by changes on temperature.
Indeed, conversion and camphor selectivity have an almost
equal behavior (Fig. 8a).
We have found that in reactions at temperatures lower
than 343 K, species alkylperoxides were also formed, com-
promising the camphor selectivity (Fig. 8b). These com-
pounds were not detectable by GC analysis, but had its
On the other hand, whereas to observed when we var-
ied the stoichiometry of reactants, the reaction camphor
1 3

Monolacunary K8SiW11O39-Catalyzed Terpenic Alcohols Oxidation with Hydrogen…
(a)
(b)
Fig. 8 Effect of temperature on kinetic curves (a) and conversion and camphor selectivity (b) of K8SiW11O39-catalyzed oxidation reactions of
borneol with H2O2. Reaction conditions: borneol (2.00 mmol); H2O2 (6.00 mmol); reaction time (3 h); K8SiW11O39 (1.25 mol%); DMA (10 mL)
formation confirmed through triphenylphosphine tests, as
proposed by Shulpin et al [43]. These results agree with
literature [35].
We can conclude that optimized conditions for the
K8SiW11O39-catalyzed borneol oxidation with hydrogen
peroxide to achieve ca. 90% of conversion and selectivity
were as follows; borneol: hydrogen peroxide molar ratio
(2:1); temperature (343 K), catalyst load and reaction time
(5 mol% and 90 min respectively), or if a longer time of reac-
tion will be acceptable (0.5 mol%, 180 min, respectively).
3.5 Catalyst Reuse
The K8SiW11O39 is a solid catalyst and was insoluble in the
reaction medium (i.e. DMA). To recover it, we employed
the same procedure used to purify the reaction products (see
Fig. 9 Recovery rates of K8SiW11O39 catalyst and conversion of
experimental, Sect. 2.5). We achieved a recovery rate vary-
the oxidation reactions of borneol with H2O2. Reaction condi-
ing from 88 to 92% along three cycles of reuse. This high
recovery suggests that procedure used was efficient. There-
fore, it was possible to reuse this solid catalyst in another
reaction. The results of reuse and recovery rate of the cata-
lyst are shown in Fig. 9.
tions: borneol (2.00 mmol); H2O2 (6.00 mmol); reaction time (3 h);
K8SiW11O39 (5.0 mol%); temperature (353 K); DMA (10 mL)
The primary alcohols (β-citronellol) and allylic (geraniol,
nerol), tertiary (linalool and α-terpineol) were the substrates
selected. All these alcohols have double bonds, which can
be potentially functionalized (i.e., through epoxidation or
nucleophilic addition reactions).
The stability of catalyst was verified via FT-IR analyses
(Fig. 10). The main absorption bands of K8SiW11O39 cata-
lyst remained intact after three cycles of reuse.
3.6 Effect of Substrate on K8SiW11O39‑Catalyzed
Terpenic Alcohols Oxidation Reaction
with Hydrogen Peroxide
The kinetic curves and selectivity and presented in
Fig. 12a and b, respectively. Expectedly, the reactions with
the tertiary alcohols (linalool and α-terpineol) achieved low
conversions (ca. 50 and 47%, respectively). Although the
hydroxyl group of these alcohols cannot be oxidized, they
have double bonds that underwent nucleophilic addition
The reaction scope was extended to the other terpenic alco-
hols (Fig. 11).
1 3

M. J. da Silva et al.
reactions of water present in the oxidant. In addition, a mix-
ture of epoxides was also detected (Fig. 12b).
The allylic alcohols are more reactive than those inacti-
vated, consequently, the selectivity for carbonylic products
of β-citronellol (i.e. β-citronellal) was lower than those of
geraniol and nerol (i.e., geranial and neral, respectively)
(Fig. 12a).
This effect also affected the conversion; while geraniol
and nerol (i.e. geometric isomers) achieved almost complete
conversion within three reaction hours, the conversion of
β-citronellol was only 72% (Fig. 12a).
Linalool and α-terpineol were converted to a complex
mixture of minority products, formed through nucleophilic
addition reactions of water to the double bonds, for these
reasons we excluded of Fig. 12.
Conversely, geraniol and nerol were selectively converted
to two products; epoxides and aldehydes (Fig. 12b). Equally
to the β-citronellol, they have a tri-substituted double bond
less reactive, probably due to a hindrance steric. This trisub-
stituted double bond, whereas to two trisubstituted double
Fig. 10 FT-IR spectra of fresh and reused K8SiW11O39 catalyst
Fig. 11 Structural formulae of
the terpenic alcohols
(a)
(b)
Fig. 12 Kinetic curves (a) and reaction selectivity (b) of K8SiW11O39-catalyzed oxidation reactions of terpenic alcohols with H2O2. Reaction
conditions: terpenic alcohol (2.0 mmol); H2O2 (6.0 mmol); K8SiW11O39 catalyst (1.25 mol%); temperature (363 K); DMA (10 mL)
1 3

Monolacunary K8SiW11O39-Catalyzed Terpenic Alcohols Oxidation with Hydrogen…
Table 2 Effect of heteropolyanion in the catalytic activity of lacunar
potassium salts in the borneol oxidation reactions with hydrogen per-
oxide
on Mo or W(VI) atoms of the heteropolyanions give the
active species. Certainly, a higher electronic density on these
addenda atoms should to improve this step.
Run
Catalyst
Conversion (%) Camphor
The activity of silicotungstic acid salts generated by
the total exchange of the protons by different cations was
assessed and the kinetic curves are displayed in Fig. 13.
Among the catalysts assessed, only those salts contain-
ing TBA (i.e., tert-butyl ammonium) and Na+ cations were
almost soluble. The cesium and ammonium salts were insol-
uble. Therefore, it is reasonable suppose that the solubility
was a positive characteristic that favored the activity of het-
eropoly salt catalysts. The literature describes that a higher
acid strength is also important in these reactions. Indeed, the
TBA was also a highly active catalyst [49, 50].
selectivity
(%)
1
2
3
K8SiW11O39
K7PW11O39
K7PMo11O39
98
95
34
98
93
93
Reaction conditions: borneol to H2O2 (1:3); lacunar salt catalyst
(5 mol%); temperature (333 K); DMA (10 mL)
bonds of geraniol and nerol that were epoxidized, has two
methyl groups and remained intact after the oxidation
reactions.
4 Conclusion
3.7 Effect of Catalyst Nature: Assessments
on the Impact of Heteropolyanion
and Countercation on Activity of Catalyst
in Borneol Oxidation Reaction with Hydrogen
Peroxide
One of the most significant reactions in organic chemistry
is oxidation reaction of alcohols, which provide a variety
of products such as carbonyl compounds and epoxides.
Especially, when the substrate are renewable raw material
and the oxidant is environmentally benign, these reactions
assume greater importance. In this work, we investigate the
oxidation of terpenic alcohols with hydrogen peroxide over
lacunar catalyst (i.e., K8SiW11O39). This catalyst was effi-
ciently synthesized in one pot reaction using the adequate
precursors (Na2WO4, Na2SiO3 and KCl salts) at pH 5.5.
The catalyst was characterized by FT-IR, BET, XRD and
potentiometric titration. In oxidation reactions of borneol,
the catalyst achieved high TON (ca. 2720) and high cam-
phor selectivity (ca. 90%). Based on experimental data, a
reaction pathway was proposed, involving a tungstoperoxide
intermediate. Geraniol and nerol were almost completely
Due to high efficiency of the lacunar silicotungstic potas-
sium salt, we have decided to investigate the effect of oth-
ers heteropolyanions. We synthesized the potassium salts of
phosphotungstic and phosphomolybdic acids and assessed
their catalytic activity in borneol oxidation with hydrogen
peroxide. The main results are displayed in Table 2.
The literature has described the success of lacunar Keggin
heteropolyacid salts (i.e., containing tungsten and molyb-
denum as addenda atoms) in oxidations with hydrogen
peroxide [44]. However, the type of substrate and the reac-
tion conditions (i.e., homogeneous or biphasic, pH, addi-
tives) sometimes difficult to understand why the tungsten or
molybdenum salt was the most active catalyst [45]. Herein,
in addition to this effect, the heteroatom present in the Keg-
gin anion seems to have also played a vital role for the effi-
ciency of catalyst.
The tungsten HPAs were more active than molybde-
num (Table 2). This result agrees with the literature, which
described that the W-based polyoxometalate catalysts are
more active than the Mo catalysts [46]. Nonetheless, this
point is still controversial; the literature has described works
where the activity of molybdenum or tungsten heteropoly-
acids has been assessed even for heteropolyacids simultane-
ously containing the two metals [47]. Therefore, there are
examples where molybdenum catalysts are also effective
[48].
In according with the Scheme 2, the addenda atom of the
catalyst (i.e., Mo or W) should be peroxidized and then pro-
mote the transferring of oxygen of peroxide to the substrate
[45]. Therefore, the nucleophilic attack of hydrogen peroxide
Fig. 13 Effect of countercation in kinetic curves of lacunar undeca-
silicotungstate salt-catalyzed oxidation reactions of borneol with
H2O2. Reaction conditions: borneol to H2O2 (1:3); lacunar salt cata-
lyst (5 mol%); temperature (333 K); DMA (10 mL)
1 3

M. J. da Silva et al.
21. Rana S, Mallick S, Rath D, Parida KM (2012) J Chem Sci
124(5):1117
converted and selectively gave epoxides and aldehydes at
equimolar proportions. A selectivity of ca. 60% β-citronellal
was obtained. The trisubstituted bond double present in the
geraniol, nerol and β-citronellol that contain two methyl
groups remained intact. In addition to be highly active and
selective, this solid catalyst was efficient recovered and
reused without loss activity. The effects of the heteropoly-
anion and of the cation present in the lacunar salt were stud-
ied. Tungsten lacunar potassium salts were more active than
one molybdenum. Soluble lacunar silicotungstic acid salts
(i.e., Na+ and TBA cations) were more active than insoluble
22. Dong X, Yu C, Wang D, Zhang Y, Wu P, Hu H, Xue G (2017)
Matter Res Bull 85:152
23. Liu S, Chen L, Wang G, Liu J, Gao Y, Li C, Shan H (2015) J Ener
Chem 25:85
24. da Silva MJ, de Oliveira CM (2018) Curr Catal 7:26
25. Coronel NC, da Silva MJ (2018) J Clust Sci 29:195
26. Hill CL (2007) J Mol Catal A 262:2
27. Choi JH, Kim JK, Park DR, Kang TH, Song JH, Song IK (2013)
J Mol Catal A 371:111
28. Pathan S, Patel A (2013) Dalton Trans 42(32):11600
29. Patel K, Shringarpure P, Patel A (2011) Transition Met Chem
36(2):171
+
salts (i.e., Cs and NH4 ).
30. Kamata K, Yonehara K, Sumida Y, Yamaguchi K, Hikichi S,
Mizuno N (2003) Science 300(5621):964
31. Kamata K, Kotani M, Yamaguchi K, Hikichi S, Mizuno N (2007)
Chem Eur J 13(2):639
Acknowledgements The authors are grateful for the financial support
from CAPES, CNPq and FAPEMIG (Brazil).
32. Ma Y, Zhang Y, Ding W, Zhao Y (2010) Catal Commun 11(9):853
33. Pizzio LR, Blanco MN (2007) Microporous Mesoporous Mat
103(1–3):40
References
34. Frenzel R, Sathicq AG, Blanco MN, Romanelli GP, Pizzio LR
(2015) J Mol Catal A 403:27
35. Berry FJ, Derrick GR, Mortimer M (2014) Polyhedron 68:17
36. Moffat JB (2001) Metal-oxygen clusters: the surface and catalytic
properties of heteropoly oxometalates. Plenum, New York
37. Viana LAS, da Silva GRN, da Silva MJ (2017) Catal Lett 148:374
38. Hida T, Nogusa H (2009) Tetrahedron 65:270
39. Das SP, Ankireddy SR, Boruah JJ, Islam NS (2012) RCS Adv
2(18):7248
1. Gallezot P (2007) Catal Today 121:76
2. da Silva MJ, Ayala DAM (2006) Catal Sci Techniol 6(9):3197
3. Costa VV, Rocha KAS, Oliveira LCA, Kozhevnikova EF, Koz-
hevnikov IV, Gusevskaya EV (2016) RCS Adv 6(49):43217
4. Monteiro JLF, Veloso CO (2004) Top Catal 27:169
5. da Silva MJ, Carari DM, da Silva AM (2015) RCS Adv
5(14):10529
40. Ishimoto R, Kamata K, Mizuno N (2012) Angew Chem Int Ed
51(19):4662
6. Guo Z, Liu B, Zhang Q, Deng W, Yang Y (2014) Chem Soc Rev
43(10):3480
41. Kazuhiko S, Masao A, Junko T, Klaus Z, Ryoji N (1999) Bull
Chem Soc Jpn 72(10):2287
7. Wang SS, Yang GY (2015) Chem Rev 115(11):4893
8. Anastas PT, Kirchhoff MM (2002) Acc Chem Res 35(9):686
9. Kozhevnikov IV (2007) J Mol Catal A 262(1–2):86
10. Cavani F (1998) Catal Today 41(1–3):73
42. Bortolini O, Conte V, Di Furia F, Modena G (1986) J Org Chem
51(14):2661
43. Shulpin GB, Kozlov YN, Shulpina LS, Petrovskiy PV (2010) Appl
Organomet Chem 24(6):464
11. Zhou Y, Chen G, Long Z, Wang J (2014) RCS Adv 4(79):42092
12. Gupta P, Paul S (2014) Catal Today 236:153
13. Narkhede N, Singh S, Patel A (2015) Green Chem 17(1):89
14. de Meireles LP, Rocha KAS, Kozhevnikov IV, Gusevskaya EV
(2011) Appl Catal A 409:82
44. Mizuno N, Yamaguchi K, Kamata K (2005) Coordin Chem Rev
249:1944
45. Patel K, Shringarpure P, Patel A (2011) Trans Metal Chem 36:171
46. Trakarnpruk W, Rujiraworawut K (2009) Fuel Process Technol
90:411
15. Alsalme A, Kozhevnikova EF, Kozhevnikov IV (2010) Appl Catal
A 390(1–2):219
47. Mohammad H, Alizadeh K, Tayebee R (2005) J Braz Chem Soc
16:108
16. Bang Y, Park DR, Lee YJ, Jung JC, Song IK (2011) Korean J
Chem Eng 28(1):79
48. Akimoto M, Ikeda H, Okabe A, Echigoya E (1084:) J Catal 89:196
49. Venturello C, D’Aloiso R, Bart JC, Ricci M (1985) J Mol Catal
32:107
17. Kharat N, Pendleton P, Badalyan A, Abedini M, Amini MM
(2001) J Mol Catal A 175(1–2):277
18. Alsalme M, Wiper PV, Khimyak YZ, Kozhevnikova EH, Koz-
hevnikov IV (2010) J Catal 276(1):181
50. Surjyakanta Rana S, Mallick L, Mohapatra G, Bishwa Bidita Var-
adwaj KM, Parida (2012) Catal Today 198:52
19. Ibrahim SM (2013) Modern Res Catal 2:110
20. Yang W, Billy J, Taârit YB, Védrine JC, Essayem N (2002) Catal
Today 73(1–2):153
1 3