Synthesis and characterization of copper and aluminum salts of H3PMo12O40 for their use as catalysts in the eco-friendly synthesis of chromanes

Article abstract of DOI:10.1016/j.apcata.2009.12.029

Acid catalysts based on aluminum (or copper) salts of molybdophosphoric acid (H₃PMo₁₂O₄₀) were prepared. They were synthesized from a heteropolyacid solution to which Al₂(SO₄)₃, Al₂O(CH3COO)₄ or CuSO₄ was added. The obtained salts were texturally characterized and a low specific surface area, between 1 and 9 m²/g, was observed. The analysis of the prepared salts by ICP-AES spectrometry indicated that the molar content of Mo and P was that corresponding to the anion [PMo₁₂O₄₀]^3-. The presence of undegraded Keggin structure was confirmed by XRD, FT-IR, DRS and ³¹P MAS-NMR. The catalyst acidity was measured by means of potentiometric titration with a solution of n-butylamine in acetonitrile and by temperature-programmed desorption of pyridine. The AlPMo₁₂O₄₀ salt (from sulfate), which at 5 h of reaction showed the highest conversion of m-cresol (98%) and 75% and 25% selectivity to chromane 4 and 3, respectively, presented the highest total acidity and, simultaneously, acid sites with maximum acid strength very high. On the other hand, Cu_0.5H₂PMo₁₂O₄₀ catalysts at 5 h of reaction showed lower conversion of m-cresol than the former salt (90%) and also lower selectivity to chromane 4 and 3 (37% and 10%, respectively). In the latter case also 43% of open-chain compounds were observed. With the other salts, Al_0.5H_1.5PMo₁₂O₄₀, Cu₃(PMo₁₂O₄₀)₂ and AlPMo₁₂O₄₀ (from Al₂O(CH₃COO)₄), only the reaction intermediaries, open-chain compounds, were obtained because these salts presented low total acidity and acid sites with maximum acid strength very low.

Full text of DOI:10.1016/j.apcata.2009.12.029

Applied Catalysis A: General 375 (2010) 196–204
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Synthesis and characterization of copper and aluminum salts of H3PMo12O40 for
their use as catalysts in the eco-friendly synthesis of chromanes
b
b
b
a
a,
*
´
´
Daniela S. Mansilla , M. Rosario Torviso , Elba N. Alesso , Patricia G. Vazquez , Carmen V. Caceres
a Centro de Investigacio´n y Desarrollo en Ciencias Aplicadas ‘‘Dr. Jorge Ronco’’ (CINDECA), Universidad Nacional de La Plata, Facultad de Ciencias Exactas, Calle 47 N8 257,
1900 La Plata, Argentina
b
´
´
Departamento de Quı´mica Organica, Facultad de Farmacia y Bioquımica, Universidad de Buenos Aires, Argentina
A R T I C L E I N F O
A B S T R A C T
Article history:
Acid catalysts based on aluminum (or copper) salts of molybdophosphoric acid (H3PMo12O40) were
prepared. They were synthesized from a heteropolyacid solution to which Al2(SO4)3, Al2O(CH3COO)4 or
CuSO4 wasadded. Theobtained saltswere texturally characterized and a low specific surface area, between
1 and 9 m2/g, was observed. The analysis of the prepared salts by ICP-AES spectrometry indicated that the
Received 18 August 2009
Received in revised form 17 December 2009
Accepted 18 December 2009
Available online 11 January 2010
molar content of Mo and P was that corresponding to the anion [PMo12O40]
3À. The presence of undegraded
Keggin structure was confirmed by XRD, FT-IR, DRS and 31P MAS-NMR. The catalyst acidity was measured
by means of potentiometric titration with a solution of n-butylamine in acetonitrile and by
temperature-programmed desorption of pyridine. The AlPMo12O40 salt (from sulfate), which at 5 h of
reaction showed the highest conversion of m-cresol (98%) and 75% and 25% selectivity to chromane 4
and 3, respectively, presented the highest total acidity and, simultaneously, acid sites with
maximum acid strength very high. On the other hand, Cu0.5H2PMo12O40 catalysts at 5 h of reaction
showed lower conversion of m-cresol than the former salt (90%) and also lower selectivity to
chromane 4 and 3 (37% and 10%, respectively). In the latter case also 43% of open-chain compounds
Keywords:
Copper and aluminum salts
H3PMo12O40
Catalysts
Eco-friendly synthesis
Chromanes
were observed. With the other salts, Al0.5H1.5PMo12O40, Cu3(PMo12O40
) and AlPMo12O40 (from
2
Al2O(CH3COO)4), only the reaction intermediaries, open-chain compounds, were obtained because
these salts presented low total acidity and acid sites with maximum acid strength very low.
ß 2010 Elsevier B.V. All rights reserved.
1. Introduction
systems [5,6]. In these papers, their properties as catalysts for acid
or oxidation reactions have been highlighted. There are several
In nature there are many compounds with recognized
pharmacological activity, and most of them have been obtained
by extraction from vegetables. Reactions in liquid phase have been
carried out for their synthesis, and they have generally been
catalyzed with the aim of optimizing their preparation. Catalysis is
a key solution for the development of a new eco-friendly chemical
industry [1], in particular for the pharmaceutical industry, which is
one of the industries that generates greater proportion of waste
according to the contamination index. The heterogeneous catalytic
process as substitute for stoichiometric processes prevents the
generation of environmentally harmful effluents and by-products,
leading to clean technologies [2]. This calls for an innovative effort
in the design of new catalysts to obtain catalytic agents more
benign for the environment [3,4].
industrial processes that have used heteropolycompounds with a
Keggin-type structure, and the basic investigation continues [7,8].
These compounds are multielectronic oxidants and, simultaneous-
ly, they are acids with strength greater than that of the
conventional acids. The reactions in which they can be used, such
as cyclization, esterification, dehydration, oxidation of amines or
epoxidation of alkenes, are important for a large number of
industries related to fine chemicals, such as flavor, pharmaceutical
and food industries, among others.
For the preparation of catalysts from the above-mentioned
heteropolycompounds, their properties can be controlled by a
suitable selection of the heteropolyanion and its constituent
elements such as heteroatom, peripheral atom and countercation.
These catalysts can be used in homogeneous processes in liquid
phase, reactions in two-phase liquids (phase-transfer catalysis) or
in liquid–solid or gas–solid systems.
Several reviews have reported reactions catalyzed by hetero-
polycompounds, both in homogeneous and in heterogeneous
In the preparation of supported catalysts, heteropolyacid
stability in the impregnating solution is an important factor, since
it influences the species present in the solid and hence, the
properties of the final catalyst. Therefore in our research group, the
* Corresponding author. Tel.: +54 221 421 1353; fax: +54 221 425 4277.
E-mail address: ccaceres@quimica.unlp.edu.ar (C.V. Ca´ceres).
0926-860X/$ – see front matter ß 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2009.12.029

D.S. Mansilla et al. / Applied Catalysis A: General 375 (2010) 196–204
197
influence of different solvents on the species in solution and on the
support has been studied [9,10].
elimination of CO2. On cooling, the mixture was stirred for 2 h.
Then, the BaSO4 or Ba(CH3COO)2 solid was removed by filtration
and it was washed three times with distilled water. Al or Cu salts
were obtained by evaporation of remnant water after a week.
The use of different solids as support was investigated. This
study was very important due to the low specific surface area that
the bulk compounds present. It was also important due to the
advantage of supported catalysts of being easily separated in
liquid-phase reactions. Among other systems, the impregnation on
carbon, silica, titania and alumina [11,12], and on zirconia [13] has
been studied.
These salts were AlPMo12O40
, Al0.5H1.5PMo12O40, Cu0.5H2P-
Mo12O40, Cu3(PMo12O40 2 (all from Al2(SO4)3) and AlPMo12O40
)
(from Al2O(CH3COO)4). They were called AlPMo, Al0.5HPMo,
Cu0.5HPMo, CuPMo, and AlPMoAct, respectively.
In particular for their use in liquid-phase reactions, it is
convenient to obtain insoluble or solid-supported heteropolycom-
pounds that can be firmly immobilized. In this way additional
advantages over the homogeneous catalyst were obtained, due to
the facility of catalyst-product separation and their possibility of
re-use. The results of studies conducted and reported by different
authors and by our research group [14–20] have emphasized the
importance of developing processes employing this type of catalyst
to obtain products on a small scale, such as those produced for the
fine chemicals industries.
2.2. Salt characterization
2.2.1. Inductively coupled plasma atomic emission
spectroscopy (ICP-AES)
The analyses were performed using a Shimadzu Sequential
Plasma Spectrometer 1000, model III, and conventional pneumatic
nebulization (Meinhard), according to EPA 200.7 standard. The
operating conditions were plasma power, 1.2 kW; torch height,
15 mm; plasma gas (Ar) at 14 l/min; cooling Ar flow, 1.4 l/min; and
Ar carrier, 1 l/min.
On the other hand, substances with vitamin E activity are
homologues of 6-chromanol. The eight naturally occurring
compounds are divided into two groups, the tocopherols, which
have a saturated C16 isoprenoid side chain, and tocotrienols, which
have a triply unsaturated C16 side chain (Scheme 1). This vitamin
has a considerable biological role as antioxidant [21]. It is
especially important to prevent the formation of hydroperoxide
derivatives of the polyunsaturated fatty acid. The first synthesis of
tocopherols was performed by Karrer [22], using trimethylhy-
droquinone and allyl bromides with aluminum trichloride as
catalyst. The ZnCl2 was also used as catalyst for synthesis of
tocopherols [22–23]. The most important reaction of tocopherols is
2.2.2. Fourier transform infrared spectroscopy (FT-IR)
FT-IR spectra of solid samples were recorded using a Bruker IFS
66 FT-IR spectrometer, with pellet samples, and a measuring range
of 400–4000 cmÀ1
.
2.3. X-ray diffraction (XRD)
XRD patterns of solid samples were recorded. To obtain these
diagrams, a Philips PW-1390 device with built-in recorder was
used. Conditions were as follows: Cu Ka radiation; Ni filter; 20 mA
and 40 kV in the high voltage source; scanning angle (2u) from 58 to
the oxidation to
g
-hydroxyalcohylquinones, which have good
608; and scanning rate, 18/min.
activity to cure and prevent muscular dysphagia in animals.
The above-mentioned concepts show the importance of
studying the substitution of aluminum trichloride by hetero-
polycompounds with Lewis acid characteristics in the synthesis of
tocopherols. In the present work, the substitution of aluminum
trichloride by aluminum (or copper) salts of molybdophosphoric
acid was studied in the synthesis of chromanes.
2.4. Diffuse reflectance spectroscopy (DRS)
The solid samples were studied in the range 200–600 nm, using
UV–visible Varian Super Scan 3 equipment, fitted with a diffuse
reflectance chamber with BaSO4 inner surface.
2.4.1. Magic-angle spinning nuclear magnetic resonance (MAS-NMR)
Solid samples were analyzed by 31P MAS-NMR by means of
Varian Mercury Plus 300 equipment with a sample holder 7 mm in
diameter, a resonant frequency of 121.469 MHz, and a spinning
rate of 5 kHz. The measurements were carried out at room
temperature using 85% H3PO4 as external reference.
2. Experimental
2.1. Salts of molybdophosphoric acid synthesis
The salts of H3PMo12O40 were prepared by the technique
described by Silvani and Burns [24]. In a typical preparation
5.48 mmol of H3PMo12O40 (MPA) was dissolved in 12 ml of
distilled water. The necessary mmol of Al2(SO4)3Á18H2O,
Al2O(CH3COO)4 or CuSO4ÁH2O, to get the desired salt, were
added to the previous solution. The solution was continuously
stirred and heated to 50–60 8C. Then, 8.22 mmol of BaCO3 was
slowly added; the reaction was continued to the complete
2.5. Acidity measurements
2.5.1. Potentiometric titration
The catalyst acidity was measured by means of potentiometric
titration. A known mass of solid (0.05 g) suspended in 45 ml of
acetonitrile was stirred for 8 min. Later the suspension was titrated
with a solution of n-butylamine in acetonitrile (0.025 N) at a flow
rate of 0.025 cm3/min. The electrode potential (mV) variation was
measured using a titrating device with a Metrohm 806 exchange
unit.
2.5.2. Temperature-programmed desorption of pyridine and
trimethylpyridine
Also the catalyst acidity was assessed by temperature-
programmed desorption of probe molecules (pyridine and
trimethylpyridine). The pyridine is a probe molecule that allows
determining the total acidity of a compound, and the trimethyl-
pyridine makes it possible to determine the sites with Brwnsted
acidity. Two hundred milligrams of the catalyst was immersed in a
closed vial containing pure pyridine (or trimethylpyridine) for 4 h.
Scheme 1. Synthesis of chromanes.

198
D.S. Mansilla et al. / Applied Catalysis A: General 375 (2010) 196–204
Then the catalyst was taken out from the vial and excess pyridine
(or trimethylpyridine) was removed by evaporation at room
temperature under a fume hood. The sample was then charged to a
quartz microreactor, and a constant nitrogen flow (40 ml/min) was
set up. Weakly adsorbed pyridine (or trimethylpyridine) was first
desorbed in a first stage of stabilization by heating the sample at
110 8C for 2 h. The temperature of the oven was then raised to
600 8C at a heating rate of 10 8C/min. The reactor outlet was
directly connected to a flame ionization detector to measure the
desorption rate of pyridine or trimethylpyridine.
123.8, 123, 12, 121.5, 121.3, 116.5, 113.6, 39.5, 32, 29.7, 29.5, 26.5,
25.5, 21, 20, 17.6, 16 ppm.
(7) 1H NMR (13C NMR) (CDCl3):
d = 7.00 (d, J = 8.0 Hz, 1H)
(129.8), 6.65 (d, J = 2.5 Hz, 1H) (117), 6.60 (dd, J = 2.5 Hz, J = 8.0 Hz,
1H) (112.5), 5.20 (bt, 1H) (123), 5.10 (bt, 1H) (124), 3.30 (d,
J = 6.9 Hz, 2H) (31), 2.25 (s, 3H) (19.5), 2.10 (m, 2H) (39.5), 2.07 (m,
2H) (26.7), 1.72 (s, 3H) (16), 1.70 (s, 3H) (25.7), 1.62 (s, 3H) (17.6)
ppm. Chemical shifts for 13C were determined by means of HMQC
(CDCl3).
The product yield was calculated as the mass ratio between the
obtained purified amount of product and the theoretical mass
expected for the total conversion of m-cresol.
2.6. Textural properties
When the best reaction conditions were determined, the
traditional catalysts ZnCl2 and AlCl3 were tested.
Specific surface area (SBET), pore volume, and mean pore
diameter of solid samples were determined by nitrogen adsorp-
tion/desorption techniques using Micromeritics Accusorb 2100E
equipment.
Also under these optimal conditions, all the synthesized salts
were tested following the reaction by high performance liquid
chromatography (HPLC) for 5 h, with a Konik 500-A device, fitted
with a UV detector (204 nm), C-18 column and acetonitrile as
solvent at 0.5 cm3/min flow rate.
2.7. Synthesis of chromanes
The conversion by HPLC was calculated as the ratio between
mmol of transformed m-cresol and 100 mmol of initial reagent,
and the product selectivity as the ratio between mmol obtained of
each product and 100 mmol of final reaction solution.
AlPMo catalyst leaching was evaluated. The reaction was
carried out under the optimal conditions during 10 min, the
catalyst was filtered, and then the solution was stirred at reflux
temperature, and it was evaluated if the reaction was completed.
On the other hand, experiences about re-use were realised.
Reactions were carried out in a glass batch reactor. In all of
them, the ratios 1:1:0.033 mmol of m-cresol:geraniol:catalyst
were used. The m-cresol and geraniol were solubilized in the
chosen solvent (chloroform, heptane, hexane and heptane/ethyl
carbonate), then the catalyst was added. The mixture was stirred at
reflux temperature, and the reaction was followed by thin-layer
chromatography (TLC) for 5 h. The TLC test was carried out using
layers of aluminum with silica gel 60 F254, the hexane:ethyl
acetate mixture (ratio 2:1) as eluent, and the reagents as reference.
On completion, the catalyst was separated by filtration. The crude
reaction was washed with 10% NaOH, then with water and dried
(Na2SO4 anh.). The solvent was removed in vacuum (except in the
case of heptane/ethyl carbonate). In the case of the reaction with
heptane/ethyl carbonate, following filtration the phases were
separated in a decanting vial. The carbonate phase was extracted
with heptane. Then the heptane phase was treated with 10% NaOH,
and the same procedure as the above-mentioned for the rest of the
solvents was applied.
3. Results and discussion
The analysis of the prepared salts by ICP-AES spectroscopy
indicated that the molar content of Mo and P is that corresponding
3À
to the anion [PMo12O40
]
.
In Fig. 1 the FT-IR spectrum of molybdophosphoric acid is
shown. Its characteristic bands at 1064 cmÀ1 due to stretching POa,
962 cmÀ1 to stretching Mo = Od, 870 cmÀ1 to stretching Mo–Ob–
Mo, and 775 cmÀ1 to stretching Mo–Oc–Mo are observed. The
subindexes indicate oxygen bridging Mo and the heteroatom (a), at
corners (b) and edges (c) belonging to MoO6, and terminal oxygen
(d) [25].
The reaction products were purified by preparative thin-layer
chromatography (p-TLC), using the same mixture as that used in
TLC as elution solvent. The p-TLC was performed on
a
20 cm  20 cm glass plate coated with silica gel 60 F254
(0.50 mm).
The obtained compounds were identified by proton and carbon
nuclear magnetic resonance spectroscopy (1H NMR and 13C NMR),
using TMS as reference, and CDCl3 as solvent. The spectra were
recorded in Bruker AM 500 equipment.
The FT-IR spectra of prepared Al and Cu salts show the
characteristic bands of molybdophosphoric anion, which demon-
strates that the Keggin structure is intact. In Fig. 1 the spectra (in
the range 600–1300 cmÀ1) of salts AlPMo, Al0.5HPMo, Cu0.5HPMo,
CuPMo, and AlPMoAct are shown.
Spectral data for reaction products are as follows:
(3) 1H NMR (CDCl3):
d = 7.03 (t, J = 7.7 Hz, 1H), 6.74 (d,
J = 7.7 Hz, 1H), 6.66 (d, J = 7.7 Hz, 1H), 5.27 (bs, 1H), 2.63 (dd,
J = 5.0, J = 16.4, 1H), 2.36 (dd, J = 16.4, J = 13.2 Hz, 1H), 2.26 (s, 3H),
2.00 (m, 2H), 1.73 (dd, J = 5, J = 13.2 Hz, 2H), 1.36 (m, 2H), 1.22 (s,
3H), 1.05 (s, 3H), 0.95 (s, 3H) ppm.
(4) 1H NMR (13C NMR) (CDCl3):
d = 7.04 (d, J = 8.2 Hz, 1H)
(130.3), 6.65 (d, J = 2.3 Hz, 1H) (116.5), 6.62 (dd, J = 2.5 Hz,
J = 8.2 Hz, 1H) (112.2), 5.30 (bs, 1H) (120.3), 2.90 (dd, J = 5.0,
J = 13.8, 1H), 2.40 (dd, J = 9.4 Hz, J = 13.8 Hz, 1H) (34.2), 2.32 (s, 3H)
(19), 1.90 (m, 2H) (49.50), 1.40 (m, 2H, 2H) (28), 1.00 (s, 3H) (28.4),
0.93 (s, 3H) (27) ppm. Chemical shifts for 13C were determined by
means of HMQC (CDCl3).
(5 and 6) 1H NMR (CDCl3):
d = 7.00 (t, J = 8.0 Hz, 1H), 6.98 (d,
J = 8.0 Hz, 1H), 6.76 (d, J = 8.0 Hz, 1H), 6.69 (d, J = 8.0 Hz, 1H), 6.64
(s, 1H), 5.32 (bt, 1H), 5.17 (bt, 1H), 5.05 (m, 2H), 3.38 (d, J = 7.0 Hz,
2H), 3.33 (d, J = 7.0, 2H), 2.30 (s, 3H), 2.28 (s, 3H), 2.08 (m, 8H), 1.79
(s, 3H), 1.76 (s, 3H), 1.69 (s, 3H), 1.67 (s, 3H), 1.58 (s, 3H), 1.56 (s,
3H). (13C NMR) (CDCl3):
d = 154.5, 138.5, 137.7, 132, 130, 127, 124,
Fig. 1. FT-IR spectra of MPA and its Al or Cu salts.

D.S. Mansilla et al. / Applied Catalysis A: General 375 (2010) 196–204
199
Table 1
31P MAS-NMR results.
Solid
Chemical shifts (ppm)
MPA
À3.77
À3.19
À3.31
À3.60
À3.38
À3.51
AlPMo
Al0.5HPMo
AlPMoAct
CuPMo
Cu0.5HPMo
observed (Fig. 2), which indicates that the salts present a
crystalline structure. According to bibliographic data, the crystal-
line structure corresponds to H3PMo12O40Á13H2O and it presents
triclinic symmetry [26]. Then it is possible to assure that the
Keggin-type heteropolyanion is undegraded.
By diffuse reflectance spectroscopy it is possible to verify the
presence of the [PMo12O40]
3À species for the Al and Cu salts of the
MPA. The characteristic absorption bands, according to biblio-
graphic data [27], at 220 and 260 nm of the tetrahedral Mo, and
bands attributed to oxygen-metal transfers at 210–230 nm and
another one that extends from 240 up to 545 nm indicating Mo in
octahedral position, can be observed in Fig. 3 for AlPMo and
Cu0.5HPMo salts of the MPA. Analogous spectra were obtained for
the other salts prepared.
The results obtained by 31P MAS-NMR are shown in Table 1.
All the chemical shifts appear at higher magnetic fields than that
of the external reference used (H3PO4). The spectrum of the
commercial sample MPA has only one sharp line at À3.77 ppm
(Fig. 4d). For the Al and Cu salts of the MPA, a small chemical
shift displacement to lower fields was observed (Fig. 4). This
may be due to the interaction of Cu2+ and Al3+, instead of H+,
with the hydration water molecules that the HPAs have in their
Fig. 2. X-ray diffraction patterns of MPA and its Al or Cu salts.
The FT-IR spectrum corresponding to the copper salt presents
the following characteristics:
-
-
Splitting of the band corresponding to stretching POa (1075 and
1065 cmÀ1).
Shift of the band corresponding to stretching Mo-Ob-Mo
(870 cmÀ1).
These differences with the aluminum salt (although both
cations are in the secondary structure) are because the cation Cu2+
has a greater ionic radius, and it is generally found as Cu(OH)6. This
hydration degree can be conserved when it is incorporated to the
secondary structure.
The bands corresponding to the protonic sites
were observed in the zone 4000–1500 cmÀ1. The spectrum shows a
broad band near 3400 cmÀ1 corresponding to
(OH), and that
corresponding to
(H2O) near 1600 cmÀ1. In the copper salts,
y(OH) and d(H2O)
y
d
Cu0.5HPMo and CuPMo, this band allows corroborating the above-
mentioned with respect to the presence of OH.
In the X-ray diffraction patterns of aluminum and copper salts,
well-defined diffraction lines of the heteropolyacid bulk were
Fig. 4. 31P NMR-MAS spectra (a) AlPMo, (b) Al0.5HPMo, (c) Cu0.5HPMo, (d) MPA, (e)
Fig. 3. DRS spectra of AlPMo and Cu0.5HPMo salts.
CuPMo and (f) AlPMoAct.

200
D.S. Mansilla et al. / Applied Catalysis A: General 375 (2010) 196–204
Fig. 6. Total acidity and Lewis acidity by temperature-programmed desorption of
pyridine and trimethylpyridine respectively.
The fact that the acidic properties of heteropolycompounds in
the solid state are sensitive to countercations explains the results
obtained (Table 2, Figs. 5 and 6). The total acidity of aluminum
salts were higher than that corresponding to copper salts due to
the acidity increases with an increase in the calculated charge on
the peripheral oxygen atom of the polyanion. This also explains the
lower total acidity of Al0.5HPMo salt with respect to AlPMo salt
because part of Al is substituted for H. However, this effect is not
very important when Cu is replaced by H. On the other hand, the
total acidity of AlPMoAct salt is lower than that of AlPMo salt due to
the use of a different precursor for the salt preparation, from
Al2O(CH3COO)4) and Al2(SO4)3), respectively. This preparation
condition is also important in governing the acidic properties.
Also programmed thermal desorption tests of trimethylpyr-
Fig. 5. Potentiometric titration curves for Al or Cu salts of MPA.
secondary structure. That is to say that by this technique it is
3À
also possible to verify the presence of the [PMo12O40
in the Al and Cu salts.
]
species
The acidity of the salts determined by potentiometric titration,
with a solution of n-butylamine in acetonitrile, is shown in Fig. 5.
As a criterion for interpreting the results obtained, it is suggested
that the initial electrode potential (E) indicates the maximum
acid strength of the acid sites and the value (meq amine/g solid)
where the plateau is reached indicates the total number of acid
sites. The acid strength of these sites may be classified according
to the following scale: E > 100 mV (very strong sites),
0 < E < 100 mV (strong sites), À100 < E < 0 (weak sites), and
E < À100 mV (very weak sites) [28]. The area under the curves of
Fig. 5 was calculated to know the total acidity.
idine for the prepared salts were performed to determine Brwnsted
acidity sites. The Lewis acid sites are thus obtained through the
difference between the areas under the thermal desorption curves
of pyridine (total acidity) and the corresponding areas under the
thermal desorption curves of trimethylpyridine (Brwnsted acidity).
In Table 2, the maximum acid strength of the acid sites and total
acidity of the different catalysts are shown. It is possible to observe
that the Cu0.5HPMo salt presented a maximum acid strength
slightly higher than that of the AlPMo salt. However, the AlPMo salt
had the highest total acidity. On the other hand, Al0.5HPMo,
AlPMoAct and CuPMo salts showed values of maximum acid
strength lower than that corresponding to the first two salts. These
catalysts presented total acidity lower than that of the AlPMo salt.
The values of maximum acid strength of the acid sites observed for
all salts were in the range 402–879 mV, therefore all prepared
catalysts have very strong sites.
Programmed thermal desorption of the pyridine probe mole-
cule allows determining the total acidity of the samples. The areas
under the thermal desorption curves of pyridine, for the different
catalysts, are shown in Fig. 6, where the area values with respect to
the highest one are plotted. The AlPMo salt has the highest total
acidity, whereas the other salts have a much lower total acidity.
The total acidity arrangement of the different salts obtained by this
technique is similar to that obtained by the potentiometric
titration technique (Table 2).
These area values relative to the highest one are plotted for the
different catalysts in Fig. 6. In this figure it is possible to observe
that the AlPMo salt has a larger number of sites with Lewis acidity
than the other catalysts.
With respect to the textural properties of the synthesized
aluminum and copper salts, the specific surface area of the AlPMo
salt (8.5 m2/g) is generally eight times higher than that of the other
catalysts, even for the similar salt obtained from aluminum
acetate. On the other hand, the rest of the salts present a lower
surface area than that of MPA (close to 3.0 m2/g). Since these
surface areas are small, 1.1–8.5 m2/g, it would be advisable to
support the catalysts on a porous solid.
With the objective of proving the catalytic activity of these MPA
salts in the synthesis of chromanes, the reaction between m-cresol
(1) and geraniol (2) was selected, Scheme 2. In such reaction, the
pyranosic ring is formed in only one stage by alkylation of the
aromatic ring and a subsequent addition of phenolic OH to the
double bond. In this particular case, as the aromatic ring was
substituted by a methyl group, the closed ring can be in ortho or
para position with respect to such group. Then it is possible that
two chromane isomers be formed. These are 2-(4-methylpent-3-
enil)-2,5-dimethylchromane (3) and 2-(4-methylpent-3-enil)-2,7-
dimethylchromane (4) (Scheme 2).
Table 2
Acidity measurements by potentiometric titration with of n-butylamine.
Open-chain reaction intermediaries can also be formed
(Scheme 3, 5–7). Their structure was confirmed by nuclear
magnetic resonance spectroscopy (1H NMR, HMQC and HMBC).
A reaction time of 5 h and the m-cresol:geraniol:catalyst ratio of
1:1:0.033 mmol were used to find the optimum operating
conditions and to compare the activity of the different catalysts.
The solvent and the reaction temperature were changed.
Solid
Maximum acid
strength (mV)
Total acidity (mV meq
n-butylamine/g solid)
AlPMo
779
661
473
402
879
1100
562
490
438
478
Al0.5HPMo
AlPMoAct
CuPMo
Cu0.5HPMo

D.S. Mansilla et al. / Applied Catalysis A: General 375 (2010) 196–204
201
Scheme 2. Reaction of m-cresol (1) and geraniol (2).
Fig. 7. Conversion of m-cresol using AlPMo as catalyst. Reaction conditions: m-
cresol:geraniol:catalyst 1:1:0.033 mmol, hexane, T = 69 8C.
of geraniol. Using AlPMo as catalyst, the increase of geraniol ratio
did not influence the yield.
Once the best reaction conditions had been set, the catalytic
activity of the other salts was evaluated. To follow the evolution of
reagents and products, reaction samples were taken at different
times and analyzed by HPLC. Retention times of the different
compounds are 5.45 min (m-cresol); 6.40 min (geraniol); 7.60 min
(open-chain compounds); 8.31 min (chromane 4); 9.40 min
(chromane 3). The three open-chain compounds had the same
retention time when they were analyzed by HPLC, so they are
grouped together to determine their selectivity.
The results for each catalyst are shown below.
1. AlPMo
In this case the two isomers 3 and 4, as well as the open-chain
compounds 5–7, were obtained.
The conversion and product distribution as a function of
reaction time obtained by HPLC are shown in Figs. 7 and 8,
respectively.
In the first 20 min of reaction, a 96% conversion of m-cresol is
achieved, reaching 98% at 1 h 20 min (Fig. 7).
As regards the products obtained (Fig. 8), selectivity at
20 min is 58% of chromane 4, 25% of chromane 3, and 13% of
open-chain compounds (5–7). Up to 2 h 40 min the percentage
of chromane 4 increases up to 75%, thereafter it keeps constant.
Scheme 3. Open-chain compounds.
Table 3 summarizes the experimental results when the AlPMo
salt was used as catalyst. The yields obtained via purification by p-
TLC are the sum of the two chromanes. The best yield (45%)
resulted when hexane (boiling point: 69 8C) was used as solvent.
This is why it is the solvent of choice to test the rest of the salts. In
the reaction carried out in hexane when the Cu0.5HPMo salt was
used as catalyst, the yield determined by preparative layer of
isolated products was 29% (sum of chromanes). It can be seen that
it is lower than that obtained when the AlPMo salt was used as
catalyst. The following variable to be determined was the amount
Table 3
Yields of isolate chromane 3 and 4 determined by preparative thin layer using
AlPMo as catalyst.
Solvent
Temperature (8C)
Yield (%)
Chloroform
61
98
26
37
18
45
Heptane
Heptane/diethyl carbonate
Hexane
126
69
Fig. 8. Selectivity of products using AlPMo as catalyst. Reaction conditions: m-
cresol:geraniol:catalyst 1:1:0.033 mmol, hexane, T = 69 8C.

202
D.S. Mansilla et al. / Applied Catalysis A: General 375 (2010) 196–204
Fig. 9. Conversion of m-cresol using Cu0.5HPMo as catalyst. Reaction conditions: m-
cresol:geraniol:catalyst 1:1:0.033 mmol, hexane, T = 69 8C.
Fig. 11. Conversion of m-cresol using Al0.5HPMo as catalyst. Reaction conditions:
m-cresol:geraniol:catalyst 1:1:0.033 mmol, hexane, T = 69 8C.
The percentage of chromane 3 reaches 25% at 20 min, and then it
keeps constant in time. The maximum percentage of open-chain
compounds would be formed in the first 20 min of reaction.
Then such compounds would transform into the corresponding
chromanes since at 2 h they can no longer be detected.
3. Al0.5HPMo
In this case, only open-chain products are obtained (Fig. 11).
There is no considerable formation of products up to 2 h of
reaction. Then there is an exponential increase of open-chain
products up to 4 h. At 5 h the conversion is 80%, which is the
yield of open-chain compounds. This value keeps constant in
time.
With this catalyst it was tested whether geraniol reacted
with itself under the reaction conditions mentioned above
(catalyst, temperature, solvent and time reaction). It could be
seen that geraniol reacted completely giving autocondensation
products that were detected by HPLC and 1H NMR. This is why
the absence of these products in the reaction under study
indicates that the rate of reaction of geraniol with m-cresol is
higher than the rate of reaction of geraniol autocondensation
The difference between the aluminum salt with partial
substitution of protons and the one with total substitution of
protons is that in the former case the reaction starts more
slowly, and at 5 h m-cresol conversion is lower (80% versus
98%). Besides, it can only produce the alkylation reaction but
not the subsequent ring closure for chromane formation, as
occurs in the AlPMo catalyst, which basically gives chromane 3
and 4.
2. Cu0.5HPMo
The conversion of m-cresol is 85% in 1 h. It increases up to
90% at 3 h, and then it keeps constant (Fig. 9).
4. CuPMo
Fig. 10 shows that during the first 3 h of reaction, open-chain
compounds rather than chromanes are formed. At 4 h there is
27% of chromane 4, 5% of chromane 3, and 58% of open-chain
compounds. At 5 h the following percentages are reached: 37%
chromane 4, 10% chromane 3, and 43% of the rest of the
compounds.
At 1 h of reaction, there is 81% conversion of m-cresol into
open-chain products, the maximum being reached at 4 h (84%)
(Fig. 12).
5. AlPMoAct
Fig. 10. Selectivity of products using Cu0.5HPMo as catalyst. Reaction conditions:
m-cresol:geraniol:catalyst 1:1:0.033 mmol, hexane, T = 69 8C.
Fig. 12. Conversion of m-cresol using CuPMo as catalyst. Reaction condtions: m-
cresol:geraniol:catalyst 1:1:0.033 mmol, hexane, T = 69 8C.

D.S. Mansilla et al. / Applied Catalysis A: General 375 (2010) 196–204
203
lower than corresponding to AlPMo12O40 (prepared from sulfate)
and Cu0.5H2PMo12O40 catalysts. The products obtained with these
catalysts were only the reaction intermediaries (open-chain
compounds). The different catalytic behavior is because they
showed values of maximum acid strength lower than that
corresponding to the AlPMo12O40 (prepared from sulfate) and
Cu0.5H2PMo12O40 catalysts, and they presented total acidity lower
than that of the AlPMo salt.
On the other hand, the values of Lewis acidity determined
by programmed thermal desorption tests (Fig. 6) showed
that the AlPMo salt has a greater number of sites with Lewis
acidity than the other catalysts. Also it is important to bear in
mind the fact that copper has lower Lewis acidity than
aluminum [29,30].
Under the same test conditions used with the above salts, ZnCl2
and AlCl3 catalysts, which are traditionally used in the synthesis
reactions described, were tested. There was no reaction in either of
the two cases. This can be because these salts are used in
stoichiometric amounts in these studies, and in this case a 3.3%
with respect to the reactants was used.
Fig. 13. Conversion of m-cresol using AlPMoAct as catalyst. Reaction conditions: m-
cresol:geraniol:catalyst 1:1:0.033 mmol, hexane, T = 69 8C.
The results of the catalyst leaching experiences showed that the
values of reactive conversion and products yields remained
constant when the catalyst was filtered. Then, it is possible to
assure that the reaction in these conditions is heterogeneous.
On the other hand, the values of reactive conversion and
products yields remained nearly constant after the re-use of the
catalyst, 97–99% and 80–90%, respectively.
In this case there is 87% conversion of m-cresol to open-chain
compounds at 5 h of reaction, which keeps constant in time
(Fig. 13).
It is worth pointing out how different these products are from
those obtained with the same salt prepared from sulfate. Here
there is nothing but formation of open-chain compounds. The
different behavior between the two salts is due to their very
different acidity, as was explained previously.
4. Conclusions
Summing up we can say that under the conditions set, the only
catalysts that gave the above-mentioned chromane 3 and 4 are
Acid catalysts based on the aluminum (or copper) salts of
molybdophosphoric acid were synthesized. Using different char-
acterization techniques it was possible to determine that these
salts present a crystalline structure and the Keggin-type hetero-
polyanion is undegraded. The salts prepared presented different
acid characteristics. The AlPMo12O40 salt (from sulfate), presented
the highest total acidity and, simultaneously, acid sites with
maximum acid strength very high. The Cu0.5H2PMo12O40 catalyst
showed acid sites with the highest maximum acid strength but its
total acidity was low. On the other hand, the rest of the salts
presented low total acidity and acid sites with maximum acid
strength very low.
AlPMo12O40, prepared from sulfate, and Cu0.5H2PMo12O40
.
With the rest of the salts only the reaction intermediaries 5–7
were obtained, where alkylation occurred but the ring was not
closed (Scheme 3). The three open-chain compounds have the
same retention time when they are analyzed by HPLC; that is why
only total selectivity can be determined. By preparative layer, using
the mixture hexane:ethyl acetate (10:1) one of them (compound 7)
and a mixture of the other two (compounds 5 and 6) could be
isolated. Taking advantage of the higher efficiency of AlPMo to
close the ring, both fractions were treated with this catalyst. No
reaction was observed with compound 7, whereas the mixture of
compounds 5 and 6 yielded chromane 3 and 4. These results,
together with those from spectroscopic methods, allowed us to
determine the structures unequivocally (Scheme 3). Moreover,
these results confirm the different behavior of the catalysts.
The results from these experiences can be explained from the
standpoint of the catalyst acidity. The catalysts showed different
acid behavior, as can be seen in potentiometric titration and
programmed thermal desorption tests (Table 2, Figs. 5 and 6).
The AlPMo12O40 salt (prepared from sulfate), which at 5 h
showed the highest conversion of m-cresol (98%), and 75% and 25%
selectivity to chromane 4 and 3, respectively, presented the highest
total acidity, as well as acid sites with maximum acid strength very
high.
The prepared catalysts were used in the reaction between m-
cresol and geraniol for the synthesis of chromanes. The conversion
of m-cresol and product selectivity is a function of the acid
characteristics of each salt.
The AlPMo12O40 salt (from sulfate), at 5 h of reaction showed
the highest conversion of m-cresol (98%) and 75% and 25%
selectivity to chromane 4 and to chromane 3, respectively, because
it presented the highest total acidity and, simultaneously, acid sites
with maximum acid strength very high. On the other hand, the
Cu0.5H2PMo12O40 catalyst at 5 h of reaction showed lower
conversion of m-cresol than the former salt (90%) and also lower
selectivity to chromane 4 and to chromane 3 (37% and 10%,
respectively). In the latter case also 43% of open-chain compounds
were observed. With the other salts, only the reaction intermediar-
ies, open-chain compounds, were obtained.
On the other hand, a better behavior of the Keggin hetero-
polycompounds with respect to ZnCl2 was obtained. In the latter
case, under the same operating conditions, no reaction was
observed.
In short, eco-friendly catalysts, insoluble in the reaction
solution, were obtained for the synthesis of chromanes. With
the aim of increasing the yield, due to their low specific surface
area, it is planned to support these catalysts on mesoporous
materials.
As for the Cu0.5H2PMo12O40 catalyst, at 5 h it showed a
conversion of m-cresol lower than the former (90%) and lower
selectivity to chromane 4 and 3 (37% and 10%, respectively).
Moreover, in this case there were 43% of open-chain compounds.
This catalytic behavior is on account of the fact that although this
catalyst has acid sites with a maximum acid strength slightly
higher than that of the catalyst AlPMo12O40 (prepared from
sulfate), it has lower total acidity.
As for the Al0.5HPMo, AlPMoAct and CuPMo salts, at 5 h they
showed conversions of m-cresol 80%, 87% and 84%, respectively,

204
D.S. Mansilla et al. / Applied Catalysis A: General 375 (2010) 196–204
[15] E. Alesso, R. Torviso, L. Finkielsztein, B. Lantan˜o, G. Moltrasio, J. Aguirre, L.
Acknowledgments
´
´
Pizzio, P. Vazquez, C. Caceres, M. Blanco, H. Thomas, J. Chem. Res. M (2001)
1232–1245.
The authors thank Mrs. L. Osiglio and Mrs. G. Valle for their
experimental contribution, V. Benitez and C. Pieck for temperature
programmed desorption tests, and project PICT 2003 No. 14-14482
of ANPCyT for financial support.
[16] E. Alesso, R. Torviso, M. Erlich, L. Finkielsztein, B. Lantan˜o, G. Moltrasio, J. Aguirre,
P. Va´zquez, L. Pizzio, C. Ca´ceres, M. Blanco, H. Thomas, Synth. Commun. 32 (24)
(2002) 3803–3812.
[17] E. Alesso, B. Lantan˜o, R. Torviso, L. Finkielsztein, G. Moltrasio, J. Aguirre, L. Pizzio, P.
Va´zquez, C. Ca´ceres, M. Blanco, Abstract Book 10th Brazilian Meeting on Organic
Synthesis, 2003, p. 90.
[18] L. Pizzio, P. Va´zquez, C. Ca´ceres, M. Blanco, E. Alesso, M. Erlich, R. Torviso, L.
Finkielsztein, B. Lantan˜o, G. Moltrasio, J. Aguirre, Catal. Lett. 93 (1–2) (2004)
67–73.
References
[1] G. Centi, P. Ciambelli, S. Perathoner, P. Russo, Catal. Today 75 (2002) 3–15.
[2] R.A. Sheldon, R.S. Dowing, Appl. Catal. A: Gen. 189 (1999) 163–183.
[3] G. Centi, S. Perathoner, Catal. Today 77 (2003) 287–297.
[4] P.T. Anastas, M. Kirchhoff, T.C. Williamson, Appl. Catal. A: Gen. 221 (2001) 3–13.
[5] T. Okuhara, N. Mizuno, M. Misono, Appl. Catal. A: Gen. 222 (2001) 63–77.
[6] F. Cavani, Catal. Today 41 (1998) 73–86.
[19] L.R. Pizzio, P.V. Va´zquez, C.V. Ca´ceres, M.N. Blanco, Appl. Catal. A: Gen. 256 (1–2)
(2003) 125–139.
[20] R. Torviso, E. Alesso, G. Moltrasio, L. Pizzio, P. Va´zquez, C. Ca´ceres, M. Blanco, Actas
del XIV Simposio Nacional de Quı´mica Orga´nica (2003), p. SO-141.
[21] L. Parker, S. Weber, G. Rimbach, J. Nutr. 131 (2001) 369S–373S.
[22] P. Karrer, Hel. Chim. Acta 21 (1938) 520–525;
[7] I.V. Kozhevnikov, Chem. Rev. 98 (1998) 171–198.
P. Karrer, Hel. Chim. Acta 21 (1938) 820–825.
[8] N. Mizuno, M. Misono, Chem. Rev. 98 (1998) 199–218.
[23] N. Hirose et al., US 6020505, 2000.
´
´
´
[9] P.G. Vazquez, M.G. Gonzalez, M.N. Blanco, C.V. Caceres, Stud. Surf. Sci. Catal. 91
[24] E. Silvani, R.C. Burns, J. Mol. Catal. A: Chem. 219 (2004) 327–342.
[25] P. Va´zquez, M. Blanco, C. Ca´ceres, Catal. Lett. 60 (1999) 205–215.
[26] M. Fournier, C. Feumi-Jantou, C. Rabia, G. Herve´, S. Launay, J. Mater. Chem. 2 (9)
(1992) 971–983.
(1995) 1121–1130.
[10] M.A. Castillo, P.G. Va´zquez, M.N. Blanco, C.V. Ca´ceres, J. Chem. Soc. Faraday Trans.
92 (17) (1996) 3239–3246.
[11] L.R. Pizzio, C.V. Ca´ceres, M.N. Blanco, Appl. Catal. A: Gen. 167 (2) (1998) 283–294.
[27] M. Fournier, C. Louis, M. Che, P. Chaquin, D. Masure, J. Catal. 119 (1989) 400–
414.
[28] G. Pecchi, Tesis, Universidad de Concepcio´n, Chile, 1980.
[29] G. Busca, in: J.L. Fierri (Ed.), Metal Oxides: Chemistry and Applications, CRC Press,
New York, 2005, pp. 247–318.
´
´
[12] P.G. Vazquez, M.N. Blanco, C.V. Caceres, Catal. Lett. 60 (1999) 205–215.
[13] L.R. Pizzio, P.V. Va´zquez, C.V. Ca´ceres, M.N. Blanco, Stud. Surf. Sci. Catal. 130
(2000) 953–970.
[14] P. Va´zquez, L. Pizzio, C. Ca´ceres, M. Blanco, H. Thomas, E. Alesso, B. Lantan˜o, L.
Finkielsztein, G. Moltrasio, J. Aguirre, J. Mol. Catal. 161 (2000) 223–232.
[30] G. Busca, Phys. Chem. Chem. Phys. 1 (1999) 723–736.