Advance in the study of limonene epoxidation with H2O 2 catalyzed by Cu(II) complex heteropolytungstates

Article abstract of DOI:10.1016/j.catcom.2012.04.030

The complex heteropolyoxotungstates of formula K₁₀[M ₄(H₂O)₂(PW₉O₃₄) ₂]·20H₂O (PWM) with M = Co(II), Zn(II), Cu(II) and Mn(II), and their precursor, Δ-Na₈HPW₉O ₃₄·19H₂O (Δ-PW₉), were evaluated as bulk in limonene epoxidation using H₂O₂ as oxidant, 1,2 dicloroethane as solvent and metiltricaprilamonium chloride (Aliquat 336) as a phase transfer agent at 2°C. In these biphasic conditions PWCu was the most active phase. Subsequently this phase was supported on γ-Al ₂O₃, the metal load was determined by AAS, ICP and X-ray fluorescence and the preservation of PWCu structure was confirmed by XPS and Raman microprobe. PWCu/γ-Al₂O₃ was evaluated at 70°C with acetonitrile as solvent. The pure PWCu showed a higher conversion to epoxide but lower selectivity due to the formation of secondary products.

Full text of DOI:10.1016/j.catcom.2012.04.030

Catalysis Communications 26 (2012) 117–121
Contents lists available at SciVerse ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Advance in the study of limonene epoxidation with H2O2 catalyzed by Cu(II)
complex heteropolytungstates
María G. Egusquiza a, , Carmen I. Cabello a,1, Irma L. Botto c, Horacio J. Thomas d, Sandra Casuscelli b,
⁎
Eduardo Herrero b, Delia Gazzoli e
a
“Centro de Investigación y Desarrollo en Ciencias Aplicadas Dr. J. J. Ronco”, CINDECA, CCT CONICET La Plata — Universidad Nacional de La Plata, Calle 47 N° 257, 1900-La Plata, Argentina
“Centro de Investigación y Tecnología Química”, CITeQ, Facultad Regional Córdoba, Universidad Tecnológica Nacional, Cruz Roja Argentina esq. Maestro Lopez, 5016-Córdoba, Argentina
“Centro de Química Inorgánica”, CEQUINOR, CCT CONICET La Plata — Universidad Nacional de La Plata, Calle 47 esq. 115, 1900-La Plata, Argentina
“Planta Piloto Multipropósito” PLAPIMU, CIC-PBA — Universidad Nacional de La Plata. Cam. Centenario y 508, 1897-Gonnet, Argentina
b
c
d
e
Dipartimento di Chímica, Università di Roma “La Sapienza”, Pzzale. Aldo Moro 5, Roma 00185, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
The complex heteropolyoxotungstates of formula K10[M4(H2O)2(PW9O34)2]·20H2O (PWM) with M=Co(II),
Zn(II), Cu(II) and Mn(II), and their precursor, Δ-Na8HPW9O34·19H2O (Δ-PW9), were evaluated as bulk in
limonene epoxidation using H2O2 as oxidant, 1,2 dicloroethane as solvent and metiltricaprilamonium chlo-
ride (Aliquat 336) as a phase transfer agent at 2 °C. In these biphasic conditions PWCu was the most active
phase. Subsequently this phase was supported on γ-Al2O3, the metal load was determined by AAS, ICP and
X-ray fluorescence and the preservation of PWCu structure was confirmed by XPS and Raman microprobe.
PWCu/γ-Al2O3 was evaluated at 70 °C with acetonitrile as solvent. The pure PWCu showed a higher conver-
sion to epoxide but lower selectivity due to the formation of secondary products.
Received 18 January 2012
Received in revised form 26 April 2012
Accepted 27 April 2012
Available online 16 May 2012
Keywords:
Complex heteropolytungstates
Limonene epoxidation
Hydrogen peroxide
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
other hand, an additional attribute of heteropolyoxometalates is their in-
herent stability against oxygen-donor species such as O2 and H2O2 [1–3].
The attainment of high added value products in conventional in-
dustrial processes requires strong oxidants, which are difficult to han-
dle and store; besides they produce a large number of by-products,
generally nitrogenous or sulfurate compounds or Cr3+ salts, which
are very difficult to eliminate and highly polluting. Hydrogen perox-
ide is considered as an ideal “oxidant” for these types of processes
since 47% of its weight is oxidant, it produces water as the only
byproduct and consequently the generations of inorganic salts are
avoided. Therefore, design and synthesis of adequate catalysts for ox-
idation processes allowing the use of this “ideal” oxidant are needed.
It is well known that heteropolyoxometalates present a number of
chemical and structural properties that make them attractive as oxidation
catalysts. They are structurally versatile in their ability to be substituted
with different metal ions and controlling the number of these. On the
A great number of catalysts have been recently reported for the
selective oxidation of organic substrates under benign environmental
conditions.
The epoxidation reaction of olefins has attracted particular inter-
est in recent years due to versatility of epoxides as intermediate
agents in organic synthesis. Particularly, production of oxygenated
terpenic derivatives and natural terpenoids is important for produc-
ing fragrances and food additives [4]. Among the latter, the limonene
deserves special attention due to its relatively low cost, feasibility of
becoming different oxygenated products of high added value, and
its easy attainment from a renewable natural source [5–7].
Several catalytic oxidation processes of limonene have been stud-
ied [8–10]; however, none of them has shown the necessary charac-
teristics for industrial scale application yet [11–14].
In previous studies the catalytic activity of a series of hetero-
polyoxotungstates of general formula, K10[M4(H2O)2(PW9O34)2]·20H2O
(PWM) with M=Co(II), Zn(II), Cu(II) and Mn(II), and their precursor,
Δ-Na8HPW9O34·19H2O (Δ-PW9), was analyzed in limonene oxidation
[15]. As the PWCu phase showed good activity and selectivity in the lim-
onene oxidation, kinetic studies were carried out changing the limo-
nene/H2O2/catalyst ratio. Moreover the PWCu phase supported on
alumina was evaluated as a heterogeneous catalyst.
⁎
Corresponding author. Fax: +54 221 421 1353/421 0711/422 0288/425 4277x125.
E-mail address: megus@quimica.unlp.edu.ar (M.G. Egusquiza).
Member of the research staff of CICPBA Argentina and Facultad de Ingeniería
1
U.N.L.P.
1566-7367/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2012.04.030

118
M.G. Egusquiza et al. / Catalysis Communications 26 (2012) 117–121
2. Experimental
2.1. Synthesis and characterization of catalysts
The synthesis of complex heteropolytungstates (PWM) was car-
ried out in aqueous solution, by condensation of lacunary derivatives
of the phosphotungstic acid Δ-PW9 thermally treated [16,17].
PWCu
10−
2½Δ−PW9O34
ꢀ
9− þ 4M2þ þ 2H2O→½PW9M2ðH2OÞO34ꢀ2
Precipitation of potassium salt of these complexes was due to the
addition of KCl excess. The products were purified by recrystalliza-
tion. The synthesized polyoxometalates were characterized using dif-
ferent physicochemical techniques previously reported [15,18].
The synthesis of the Cu(II) product, PWCu, is different compared
with the species containing other metallic ions such as Co(II),
Mn(II) or Zn(II) due to its thermal instability in solution. This phase
undergoes a thermolysis, which produces a mixture of the expected
phase and two structural isomers of the Keggin type of formula
PWCu / Al2O3
before reaction
K
5,5Na1,5[PW10Cu2(H2O)2O38]·13H2O, whose proportion depends on
the temperature and time conditions. Thus, the synthesis was con-
ducted at low temperature (60 °C) to obtain the best isomer–phase
ratio; as a consequence, it was not purified by crystallization.
The supported catalyst was prepared by an equilibrium impregna-
tion method, from an aqueous solution of PWCu, in contact with
300 mg of γ-Al2O3 by shaking for 24 h. The support presented the fol-
lowing textural parameters: specific area of 226 m2/g, 0.65 cm3/g vol-
ume of pore and particle size of 200 μm. After impregnation, the
solution and the solid were separated by centrifugation and the
solid was dried in oven for 24 h.
PWCu/ Al O
2
3
after reaction
200
400
600
800
[cm-1]
1000
1200
1400
Both the pure and the supported phases were characterized by
comparison through Raman microprobe by an Infinity (Jobin-Yvon)
with a photodiode detector and integrated to an optic microscope
excitation line Laser YAG Nd 532 nm and 2 cm−1 resolution and
XPS spectroscopy by a Spectrometer Leybold Heraeus LHS10, in FAT
mode with Al radiation (1486.6 eV). Also, the W and Cu charge was
determined by different methods of chemical and physical analysis
such as AAS, ICP and the X ray fluorescence. The atomic absorption
analysis was carried out using a Varian AA 240 equipment, and deter-
mination of heavy metals by X ray fluorescence was conducted using
an Innov-X Systems Alpha 4000 software LEAP (Light Element Analy-
sis Program).
Fig. 1. Raman spectra for the phase PWCu, PWCu/γ-Al2O3 and PWCu/γ-Al2O3 post-
reaction.
Conversion for both reactions was calculated according to the fol-
lowing expression:
Conversion ð%Þ
¼ ½∑products ðmol%Þꢀ 100=½∑products ðmol%Þ þ limoneneðLÞ ðmol%Þꢀ:
⋅
Selectivity was calculated according to the following expression:
Selectivity ð%Þ ¼ ½producti ðmol%Þ 100ꢀ=½∑products ðmol%Þꢀ
⋅
2.2. Catalytic evaluation
where i represents a certain product.
3. Results and discussion
The reaction catalyzed by bulk PWM, with metiltricaprilamonium
chloride (Aliquat 336) as phase transfer agent, was carried out at
atmospheric pressure, at 2 °C and using dichloroethane as solvent
[15]. Aliquots of organic phase were taken at different times of the re-
action and analyzed by gaseous chromatography and GC mass. The
relative uncertainties of the measurements were tested with repeated
determinations. The percent relative uncertainty (CV (%)) of the re-
sult was calculated dividing the corresponding absolute uncertainty
by the average of the measurements.
The reaction catalyzed by supported catalyst in the absence of PTA
(phase transfer agent) was carried out like a typical reaction, limo-
nene (L, 3.3 mmol) was dissolved in acetonitrile (2.04 g) and H2O2
35 wt.% in water (0.84 mmol) was added. The reaction started by
adding the supported catalyst (76.3 mg) to the reactor thermostated
at 70 °C. The non-converted hydrogen peroxide was determined by
iodometric titration.
3.1. Structural and spectroscopic behavior of the PWCu phase
The heteropolytungstates of formula [PW9M2(H2O)2O34]
10− are de-
scribed as a “sandwich-type” structure where two polymeric fragments
Table 1
Atomic rates by XPS among the elements present in the PWCu.
PWCu
Value
Theoretical
Experimental
K/Cu
K/P
K/W
K/Na
Na/Cu
Cu/W
O2−/W
1.75
3.50
0.39
2.33
0.75
0.22
3.78
5.05
7.10
1.30
7.47
0.77
0.22
3.92
Aliquots were taken at different times of the reaction, filtered and
then analyzed by gaseous chromatography and GC mass analysis as in
the biphasic system.

M.G. Egusquiza et al. / Catalysis Communications 26 (2012) 117–121
119
Table 2
Table 3
Atomic rates by XPS among the elements present in the PWCu/γ-Al2O3.
Conversion and selectivity to epoxide limonene with a L/H2O2/catalyst rate=1000/
2000/1 for Δ-PW9 and PWM phases.
PWCu/
Al2O3
Value
Phase
Conv. %a
Select, epox. %b
Theoretical
Experimental
Δ-PW9
PWCo
PWZn
PWMn
PWCu
68.2
0.3
0.6
16.9
96.1
89.3
–
Na/W
Cu/W
K/Cu
Na/Cu
W/Al
K/Al
0.15
0.22
2.75
0.75
–
0.56
0.54
1.48
1.04
0.03
0.02
0.01
0.01
100.0
100.0
89.0
a
–
CVb7%.
b
Na/Al
Cu/Al
–
CVb5%.
–
compared with the respective theoretical bulk ratios in Table 1 for
pure PWCu and in Table 2 for the PWCu phase supported on alumina.
The XPS Cu/W ratio of the original PCuW phase was close to the
stoichiometric one (Table 1), whereas a much higher Cu/W ratio
was obtained for the PWCu supported phase (Table 2).
of the precursor PW9 are condensed by means of the intercalation of a
tetranuclear cluster of divalent metal coordinated octahedrically to
the oxygen [18].
Previous studies have demonstrated that the Cu derivative PWCu
is mixed with two isomers with Keggin structure. The Jahn Teller ef-
fect leads to a clear availability of P\O bonds of lacunary precursor
facilitating the instability of the Cu phase in aqueous medium unlike
the behavior observed in the Zn and Co phases [15–18]. Cracking
and further reordering of the structure result in the formation of
two structures.
Fig. 1 shows the Raman microprobe spectra corresponding to the
pure and the supported PWCu phases (the latter being plotted after
and before the reaction). In these spectra, a shift of the main lines
corresponding to the vibrational modes of W_O (960 cm−1) bonds
was observed, indicating a weakening of the modes due to PWCu-
supported surface interaction, which involved a complex process of
adsorption that led to certain distortion of the structure.
The line assigned to symmetric stretching for the P\O bond of PO4
group (around 1100 cm−1) showed a weak intensity in the pure
phase; however it showed that this group preserved its typical tetra-
hedral symmetry in the supported species.
These findings indicate that the interaction between the PWCu
phase and the support caused destruction and/or a preferential orienta-
tion of the units and that Cu was concentrated on the support surface.
However, the bulk composition of the supported PCuW material,
determined by X-ray fluorescence elemental analysis, quite agreed
(0.68) with surface composition (0.54 by XPS), revealing that some
structural changes had occurred during the adsorption process.
The analysis by AAS for the supported sample accounted for 14% of
adsorbed W in γ-Al2O3. This value was obtained by simple balance of
masses from the values of metal adsorbed before and after impregna-
tion. Due to interferences among the elements it was impossible
to determine Cu by this method. Also, analysis of PWCu supported on
alumina by X ray fluorescence indicated an adsorption of (g of metal/
100 g support) W=16.5% and Cu=3.88% showing a Cu/W ratio=
0.23. Values of 14 g of W/100 g of solid were obtained by the ICP
method.
Tables 1 and 2 show the atomic rates among the elements present
in the pure phase and in PWCu phase supported on alumina deter-
mined by XPS. By this method, regions Na1s, K2p, Cu2p, O1s, C1s,
P2p, Al2s, Al2p and W4f were recorded. Na1s and K2p bands were
clearly observed, while P2p was difficult to detect. The copper was
detected as Cu(II) and W as W(VI), in both phases.
3.2. Epoxidation of limonene
The catalytic activity was evaluated for the phases, including Δ-
PW9, in a molar ratio L/H2O2/catalyst ratio=1000/2000/1 according
to the experience of previous works [20] (Fig. 2). Table 3 shows the
conversion and selectivity observed for each phase for 40 h of reac-
tion time. Fig. 3 shows the activity for all phases used as catalysts.
PWCu and Δ-PW9 were the most active phases, showing high se-
lectivity to epoxide. PWCo, PWZn and PWMn phases showed low
catalytic activity. The PWMn system showed good selectivity with
no other reaction product observed. The results showed a high con-
version of limonene for the PWCu complex, much higher than that
of the remaining phases, with a high selectivity to epoxide in short pe-
riods. However, as reaction advanced, selectivity to epoxide decreased
slightly by increasing the production of secondary compounds.
The surface XPS atomic ratios among the various elements, calcu-
lated from the peak areas and atomic sensitivity factors [19], are
120
PWCu
100
80
PW9
60
40
PWMn
20
0
PWCo, PWZn
35 40
0
5
10
15
20
25
30
t (hs.)
Fig. 2. Diagram of the reaction of the epoxidation of limonene, where (L) limonene;
Fig. 3. Limonene conversion as a function of the reaction time for Δ-PW9 and PWM
(E) 1,2-epoxilimonene; (D) diol; (H) hydroperoxide; (Car) carveol; (C) carvone.
with M=Co(II), Zn(II), Cu(II) and Mn(II).

120
M.G. Egusquiza et al. / Catalysis Communications 26 (2012) 117–121
Table 4
Table 5
Conversion and selectivity to epoxide according to L/H2O2/PWCu ratio.
Conversion and selectivity of the PWCu and PWCu/γ-Al2O3 phase at 7 h of reaction
time.
L/H2O2/PWCu
Conv.a
(% in 40 h)
Conv. (% ) (dilution)
*
Selec. Eb
(% in 40 h)
Catalysts
Conversion of
limonene (%)
Conversion
of H2O2
Selectivity to
epoxide (%)
Efficiency
of H2O2
1000/2000/1
2500/5000/1
5000/10,000/1
96.1
36.2
21.6
96.08 (1)c =96.1
*
95.6
96.6
94.1
36.16 (2.5)c =90.4
*
PWCu with PTAa
PWCu–γ-Al2O3
without PTAb
12.24
17.63
–
97.71
59.84
60.8
–
21.59 (5)c =107.9
*
91.6
68.4
a
CVb7%.
CVb5%.
b
a
Molar ratio: L/H2O2/catalyst (1000/2000/1) at 2 °C.
Molar ratio: L/H2O2 (4/1), 76 mg of catalyst at 70 °C.
c
b
Dilution factor.
environmental point of view since a non toxic solvent, replacing the
chlorinated solvents, was used in this system.
In response to increased activity and selectivity towards
epoxilimonene obtained with PWCu, we selected this catalyst to
study the influence of the molar ratio of L/H2O2/catalyst on the
course of oxidation reaction while the other parameters were
unchanged. These results are presented in Table 4.
4. Conclusions
Fig. 4 shows the conversion as a function of time for each of the
L/H2O2/catalyst under study.
Among the heteropolytungstates, PWM with M(II)=Mn, Co, Cu
and Zn, only the phase with Cu(II) and its precursor phase Δ-PW9
showed catalytic activity in the epoxidation to limonene using H2O2
under biphasic conditions. The PWCu phase has proved to be the
most active catalyst for the oxidation of limonene. Thus, these results
give evidence about its chemical properties and different structural
characteristics related to the presence of Cu(II). Among these features
the following can be mentioned: a) the Jahn–Teller effect of Cu(II)
which allows a higher exposition of the P\O group; and b) the high
redox capacity of Cu(II) in an oxidative process.
The redox process induces the electronic transference which con-
tributes to the oxygen addition on the C_C, leading to the formation
of the limonene epoxide.
The different physico-chemical analyses conducted both on the pure
phase and on the alumina supported phase indicate that the Cu/W ratio
remains similar to the theoretical ratio according to the chemical com-
position of the pure phase. Also, the Raman and XPS spectroscopic tech-
niques show the preservation of the structure on the surface.
The best catalytic activity using PWCu phase as catalyst was ob-
served by the molar ratio L/H2O2/cat=1000/2000/1.
The catalytic evaluation of the ã-Al2O3 supported PWCu system in-
dicated that the reaction conducted in the absence of PTA showed a
higher conversion than the reaction under biphasic conditions. How-
ever, the selectivity to epoxilimonene obtained with PWCu/Al2O3
was lower than PWCu under biphasic system due to the hydrolysis
reaction of the oxirane ring. Although selectivity to epoxilimonene
obtained with the supported catalyst has been low, the results are en-
couraging since the heterogeneous system avoids the use of chlori-
nated solvents and the previous steps to the synthesis are easier.
Therefore, considering the high conversion of limonene achieved, fur-
ther research should be conducted to reduce the by-products.
As it can be seen, the limonene conversion decreased by increasing
the molar ratio L/H2O2/catalyst achieving a value close to 20 mol%.
However, when the conversion values were calculated considering
the dilution factor of the catalyst corresponding to the limonene con-
centration, equivalent conversions were obtained. These results indi-
cated the existence of a linear relationship between activity and
catalyst concentration. This fact demonstrated a “molecular reaction”
which is typical of a homogeneous conventional catalyst in which
the association of its molecular or ionic units in organic phase could
not take place. Regarding the product selectivity, the epoxilimonene
was obtained as major product in all the cases. Therefore, a molar
ratio L/H2O2/cat=1000/2000/1 appeared as the optimum value.
On the other hand, the catalytic activity of the phase PWCu
supported on alumina was evaluated taking into account the impor-
tance of using a heterogeneous catalyst for the environment. In addi-
tion, another important point is the desirable replacement of the
chlorate solvents with an environmentally friendly solvent such as
acetonitrile. Table 5 and Fig. 5 show comparative conversions of the
catalytic evaluation in the limonene epoxidation for the two systems
under study after a reaction time of 7 h. The results indicated that the
selectivity to epoxilimonene for the bulk catalyst using PTA was much
higher due to the fact that the biphasic system prevented the hydro-
lysis of the oxirane ring. However, in the heterogeneous system
(supported catalyst) a higher conversion of limonene was obtained.
Then, the selectivity to epoxilimonene decreased with the occurrence
of by‐products, such as carvone (C), carveol (Car), diepoxide (not in
the scheme) and glycol (D).
In all cases the mass balance was close to values which fell within
the error of the method.
Finally, although work must continue to improve selectivity in the
heterogeneous system, the results obtained are encouraging from the
100
80
100
1000/2000/1
80
60
60
40
2500/5000/1
5000/10000/1
PWCu/γ -Al2O3
40
20
20
PWCu
0
0
0
10
20
30
40
0
1
2
3
4
5
6
7
t (hs.)
t (hs)
Fig. 4. Oxidation of limonene as a function of time for the different molar ratio L/H2O2/PWCu.
Fig. 5. Limonene conversion as a function of time catalyzed by PWCu and PWCu/γ-Al2O3.

M.G. Egusquiza et al. / Catalysis Communications 26 (2012) 117–121
121
[9] A. Corma, M.E. Dómine, F. Rey, M.T. Navarro, S. Valencia, S. Casuscelli, E. Herrero,
Proceedings of XVII Simposio Ibero-americano de Catálise, Porto, Portugal, 2000,
p. 581.
[10] M.E. Crivello, C.F. Pérez, J.D. Fernández, S.G. Casuscelli, E.R. Herrero, Proceedings
of XVII Simposio Ibero-americano de Catálise, Porto, Portugal, 2000, p. 501.
[11] A.D. Silva, M.L. Patitucci, H.R. Bizzo, E. D'Elia, O.A.C. Antunes, Catalysis Communi-
cations 3 (2002) 435.
Acknowledgment
We greatly appreciate Prof. Edmond Payen's cooperation in all-
owing the use of Raman equipment at the University of Lille (France)
and for his valuable comments.
[12] P.A.L. Villa, A.F. Taborda, C. Monte de Correa, Journal of Molecular Catalysis A:
Chemical 185 (2002) 269.
[13] C. Venturello, E. Alneri, M. Ricci, Journal of Organic Chemistry 48 (1983) 3831.
[14] C. Venturello, R.D. Aloisio, Journal of Organic Chemistry 53 (1988) 1553.
[15] S. Casuscelli, E. Herrero, M. Crivello, C. Pérez, M.G. Egusquiza, C.I. Cabello, I.L.
Botto, Catalysis Today 107–108 (2005) 230.
[16] R. Massart, R. Contatn, J.M. Fruchart, J.P. Ciabrini, M. Fournier, Inorganic Chemis-
try 16 (1977) 2916.
[17] R.G. Finke, M.W. Droege, P.J. Domaille, Inorganic Chemistry 26 (1987) 3886.
[18] C.I. Cabello, M.G. Egusquiza, I.L. Botto, G. Minelli, Materials Chemistry and Physics
87 (2004) 264.
References
[1] R. Neumann, M. Gara, Journal of the American Chemical Society 116 (1994) 5509.
[2] R. Neumann, M. Gara, Journal of the American Chemical Society 117 (1995) 5066.
[3] M.T. Pope, A. Müller, Polyoxometalate Chemistry from Topology via Self-
assembly to Applications, Kluwer Academic Publishers, The Netherlands, 2001.
[4] R.A. Sheldon, New Developments in Selective Oxidation, Elsevier, Amsterdam,
1990, p. 1.
[5] K. Bauer, D. Garbe, H. Surburg, Common Fragrance and Flavor Materials: Prepara-
tion, Properties and Uses, 2nd ed. VCH Publishers, NY, 1990.
[6] C. Blanco, E.E. Stashenko, M. Combariza, J.R. Martínez, Journal of Chromatography.
A 697 (1995) 5011.
[19] C.D. Wagner, L.E. Davis, M.V. Zeller, J.A. Taylor, R.M. Raymond, L.H. Gale, Surface
and Interface Analysis 3 (1981) 211.
[20] M. Bösing, A. Nöh, I. Loose, B. Kerbs, Journal of the American Chemical Society 120
(1998) 7252.
[7] A.M. Santa, J.C. Vergara, L.A. Palacio, A. Echavarría, Catalysis Today 80 (2008) 133.
[8] M.A. Armendia, V. Borau, C. Jiménez, J.M. Luque, J.M. Marinas, J.R. Ruiz, F.J. Urbano,
Applied Catalysis A: General 216 (2001) 256.