Selective methoxylation of α-pinene to α-terpinyl methyl ether over Al3+ ion-exchanged clays

Article abstract of DOI:10.1016/j.apcata.2014.10.028

In this study, we report the use of clay-based catalysts in the methoxylation of α-pinene, for the selective synthesis of α-terpinyl methyl ether, TME. The main reaction products and intermediates were identified by GC-MS. The reaction conditions (stirring rate and catalyst load) that afford a kinetic regime were established. SAz-1 (Cheto, Arizona, USA) source clay and a montmorillonite (SD) from Porto Santo, Madeira Archipelago, Portugal, were modified by ion-exchange with Al³⁺ to produce catalysts with markedly different acidities and textural properties. The catalysts based on the high layer-charge SAz-1 montmorillonite proved to be the most active. Ion-exchange with Al³⁺, followed by thermal activation at 150°C, afforded the highest number of Br?nsted acid sites - a significant proportion of which were located in the clay gallery - and this coincided with the maximum catalytic activity. The influence of various reaction conditions, to maximize α-pinene conversion and selectivity, was studied over AlSAz-1. When the reaction was performed for 1 h at 60°C, the conversion reached 65% with 65% selectivity towards the mono-ether, TME. Similar conversions and selectivities required up to 50 h over zeolites and other solid acid catalysts. The kinetic dependencies of this reaction on temperature and reagent concentration, over the selected clays were also investigated. It was established that, in the temperature and reagent concentration regime studied, the reaction was first order with respect to α-pinene. The apparent activation energies over the two catalysts, calculated from Arrhenius plots, were almost identical at 72 kJ mol^-1.

Full text of DOI:10.1016/j.apcata.2014.10.028

Applied Catalysis A: General 489 (2015) 171–179
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Selective methoxylation of ␣-pinene to ␣-terpinyl methyl ether over
Al3+ ion-exchanged clays
C. Catrinescua,b,∗, C. Fernandesa,c, P. Castilhoa, C. Breend
a Centro de Química da Madeira (CQM), Centro de Ciências Exactas e da Engenharia da Universidade da Madeira, Campus Universitário da Penteada,
9000-390 Funchal, Portugal
b “Gheorghe Asachi” Technical University of Iasi, Faculty of Chemical Engineering and Environmental Protection, Department of Environmental Engineering
and Management, 73 D. Mangeron Blvd, 700050 Iasi, Romania
c Laboratório Regional de Engenharia Civil, Rua Agostinho Pereira de Oliveira, 9000-264 Funchal, Portugal
d Materials and Engineering Research Institute, Sheffield Hallam University, Howard Street, Sheffield S1 1WB, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 22 July 2014
Received in revised form 10 October 2014
Accepted 17 October 2014
Available online 26 October 2014
In this study, we report the use of clay-based catalysts in the methoxylation of ␣-pinene, for the selective
synthesis of ␣-terpinyl methyl ether, TME. The main reaction products and intermediates were identified
by GC–MS. The reaction conditions (stirring rate and catalyst load) that afford a kinetic regime were estab-
lished. SAz-1 (Cheto, Arizona, USA) source clay and a montmorillonite (SD) from Porto Santo, Madeira
Archipelago, Portugal, were modified by ion-exchange with Al3+ to produce catalysts with markedly
different acidities and textural properties. The catalysts based on the high layer-charge SAz-1 montmo-
rillonite proved to be the most active. Ion-exchange with Al3+, followed by thermal activation at 150 ◦C,
afforded the highest number of Brønsted acid sites – a significant proportion of which were located in
the clay gallery – and this coincided with the maximum catalytic activity. The influence of various reac-
tion conditions, to maximize ␣-pinene conversion and selectivity, was studied over AlSAz-1. When the
reaction was performed for 1 h at 60 ◦C, the conversion reached 65% with 65% selectivity towards the
mono-ether, TME. Similar conversions and selectivities required up to 50 h over zeolites and other solid
acid catalysts. The kinetic dependencies of this reaction on temperature and reagent concentration, over
the selected clays were also investigated. It was established that, in the temperature and reagent con-
centration regime studied, the reaction was first order with respect to ␣-pinene. The apparent activation
Keywords:
Ion-exchanged clays
Clay acidity
Catalysis
Pinene
Methoxylation
Terpinyl methyl ether
energies over the two catalysts, calculated from Arrhenius plots, were almost identical at 72 kJ mol−1
.
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
by steam distillation, from gum and sulphate turpentine which
presents two structural isomers (␣- and ␤-). ␣-Pinene is the most
widely encountered terpenoid in nature [2] and finds numerous
uses in the food, fragrance and pharmaceutical sectors as well
as in the synthesis of chemical intermediates. Thus, ␣-pinene is
considered a versatile building block for the synthesis of high-
value added chemicals, mainly through catalytic processes, such as
isomerization, epoxidation and pinene oxide isomerization, hydra-
tion and dehydroisomerization, esterification, etherification [2–4]
and also in a four-step synthesis of linalool from ␣-pinene [6,10].
The alkoxylation of pinene is a less explored route to produce
1-methyl-4-(˛-alkoxyisopropyl)-1-cyclohexenes. These compounds
are used as flavours and fragrances for perfume and cosmetic
products, as additives for pharmaceuticals and agricultural chem-
icals, and also in the food industry. Among these functionalized
monoterpenes, the ␣-terpinyl methyl ether is of commercial inter-
est due to its pleasant grapefruit-like aroma. This compound is
conventionally produced via the alkoxylation of pinene or limonene
using mineral acids. The use of strong homogeneous liquid acids
The use of renewable feedstocks, such as monoterpenoids, to
produce fine chemicals via heterogeneous catalytic processes is
considered a sustainable approach which conforms with the princi-
ples of green chemistry [1]. Terpenes, such as limonene and pinene,
are an abundant renewable feedstock, used as starting materials for
the synthesis of fine chemicals, in food, cosmetic and pharmaceu-
tical industries. Terpenes are found not only in several essential
oils but also as the major constituents of pine resins and of some
byproducts of the pulp, paper and food (citrus) industries [2,3].
Pinene is a natural bicyclic monoterpene that can be separated,
∗
Corresponding author at: “Gheorghe Asachi” Technical University of Iasi, Faculty
of Chemical Engineering and Environmental Protection, Department of Environ-
mental Engineering and Management, 73 D. Mangeron Blvd, 700050 Iasi, Romania.
Tel.: +40 232 237594; fax: +40 232 237594.
E-mail addresses: ccatrine@ch.tuiasi.ro, ccatrine@uma.pt (C. Catrinescu).
http://dx.doi.org/10.1016/j.apcata.2014.10.028
0926-860X/© 2014 Elsevier B.V. All rights reserved.

172
C. Catrinescu et al. / Applied Catalysis A: General 489 (2015) 171–179
Table 1
is not recommended for industrial applications due to the associ-
ated corrosion and environmental challenges. Previous studies on
␣-pinene and limonene methoxylation have been principally per-
formed using zeolites [7] although clay-based catalysts can also
be used [8]. The alkoxylation of limonene over ␤ zeolite and ion-
exchanged clays afforded good yields (around 85%) whereas the
methoxylation of pinene, over the same ␤ zeolite, gave lower yields
(50%) of the same product. The methoxylation of pinene has been
also studied over acidic cation exchange resins [9] and sulphonic
acid-modified mesoporous silica [10]. However, more recent stud-
ies over poly(vinyl alcohol) containing sulfonic acid groups [11],
heteropolyacids immobilized on silica [12] and microporous and
mesoporous carbons [13] reported good selectivities, of ca. 60%, at
almost complete conversion.
Clay minerals, which are natural materials that cost significantly
less than the catalysts listed above, are versatile and environmen-
tally friendly catalysts that can be modified with relative ease to
promote a wide variety of organic reactions [14]. Beside the type of
clay, the nature, the locality and the extent of the isomorphic sub-
stitution strongly influences the layer charge of the clay and exerts
a major influence on the acidity and accessibility of the active sites.
These intrinsic characteristics of the clays, responsible for their
catalytic activity, can be readily improved using different meth-
ods such as ion-exchange, acid treatment and pillaring [15]. Thus,
Ni2+- and Al3+-exchanged montmorillonites, following appropri-
ate thermal activation procedures, are considered model Lewis and
Brønsted acids, respectively. The nature of the active sites has been
unequivocally determined using FT-IR spectra of adsorbed pyridine
[16] and verified by catalytic data.
Elemental composition for the purified host clays and for the Al3+ ion-exchanged
forms.
Clay
Chemical composition, % oxide
Na
Mg
Al
Si
K
Ca
Ti
Fe
NaSD
SAz-1
AlSD
2.6
0.1
0.1
0.1
3.4
6.7
2.9
6.3
20.5
20.0
22.5
23.2
57.4
59.7
54.4
68.1
0.9
0.2
0.8
0.1
0.5
3.2
0.1
0.0
3.3
0.3
5.0
0.3
10.3
1.8
12.8
1.7
AlSAz-1
The nitrogen adsorption–desorption isotherms at −196 ◦C were
determined on an Autosorb iQ from Quantachrome Instruments,
equipped with turbomolecular pumps for high vacuum attain-
ment, using helium (for dead space calibration) and nitrogen of
99.999% purity. Prior to the adsorption measurements, all the sam-
ples were outgassed for 5 h at 200 ◦C, achieved using a heating rate
of 1 ◦C min−1
.
TG data were recorded on a Mettler TG50 thermobalance
equipped with a TC10A processor. Samples (∼10 mg) were trans-
ferred directly out of cyclohexylamine (CHA) vapour into the
thermobalance and the desorption traces were recorded at a
heating rate of 20 ◦C under a nitrogen flow of 25 cm3/min. Sam-
ples were conditioned for 15 min under flowing nitrogen to
reduce the amount of physisorbed CHA. Variable temperature dif-
fuse reflectance infrared Fourier transform spectra (VT-DRIFTS),
were recorded at room temperature, then at 25 ◦C increments
until 250 ◦C. Samples were held at a specific temperature for
15 min in a flow of dry nitrogen in a variable-temperature cell
(Graseby-Specac; maximum operating temperature 500 ◦C). The
spectrometer used was a Mattson Polaris operating at 4 cm−1 res-
olution and 256 scans.
In this work we report the synthesis of ␣-terpinyl methyl ether
via the methoxylation of ␣-pinene over clay catalysts. Based on
our recent results detailing the methoxylation of limonene [8], two
starting clays, with different compositions and properties, were
selected and these were activated by ion-exchange with Al3+ and
then thermally activated at 150 ◦C. This approach is known to afford
the maximum catalytic activity in Brønsted acid catalysed organic
reactions involving polar reagents.
2.3. Catalytic tests
␣-Pinene and n-decane (internal standard) received from Sigma
Aldrich were dried over anhydrous magnesium sulphate prior
to use. Anhydrous methanol was used as received from Sigma
Aldrich. The reactions were performed in a stirred 25 ml batch reac-
tor, equipped with a reflux condenser, under drying tube (CaCl2)
protection. Before the reaction, a known amount of catalyst was
thermally activated at 150 ◦C, in air, for 2 h in a vial. Before being
removed from the oven, the vials were stoppered and then placed in
a desiccator to cool and prevent rehydration. After being cooled at
room temperature (15 min) the catalyst powder was quickly trans-
ferred into the reaction vessel containing dry methanol, preheated
at the reaction temperature. The injection of pinene and n-decane
(internal standard) marked the start of the reaction. Samples were
taken periodically and the catalyst was removed by syringe filtra-
tion. The filter had no influence on the reaction products and no
further reaction took place during storage. The reaction products
were identified by GC–MS (Agilent 6890N/MSD GC–MS system) and
quantified by GC with FID, using a J&W Carbowax 20 M column and
n-decane as an internal standard.
In a separate test the effect of the larger amounts of water avail-
able in non-thermally activated clays was evaluated. These high
water content samples were stored in a desiccator over saturated
aqueous Ca(NO3)2 prior to their use as a catalyst.
Initial reaction rates were calculated utilizing the concentra-
tion versus time data collected during the first few minutes of
the reaction. During this initial time period, the concentration of
the reactants decreased almost linearly with time, and hence the
reaction rate i.e., the differential of the reactant concentration with
respect to time, was calculated from the slope of a linear fit to these
initial data.
2. Experimental
2.1. Catalyst preparation
SAz-1 (Cheto, Arizona, USA), received from The Clay Mineral
Society Source Clay repository (Purdue University), was suspended
in deionized water and the <2 ␮m size fraction was collected by
centrifugation. The raw bentonite (SD) was collected from the Serra
de Dentro deposit (Porto Santo - Madeira Archipelago, Portugal)
and purified [17,18] to give the Na-exchanged form (NaSD). The
major impurities were removed by low speed centrifugation (6 min,
at 600 rpm), to obtain the <2 ␮m size fraction, followed by the
removal of inorganic carbonates by incremental addition of a
0.5 M sodium acetate buffer until the clay suspension reached
pH 6.8. Then, the product was converted into the Na-exchanged
form using 1 M aqueous sodium chloride solution. Excess Cl− was
removed by dialysis and the solid clay was obtained after drying the
gel collected following centrifugation at 4500 rpm for 30 min.The
chemical composition for SAz-1 and NaSD is reported in Table 1.
SAz-1 and NaSD were treated three times with 0.3 M Al(NO3)3,
washed, dried and ground to give the AlSAz-1 and AlSD catalysts.
These samples were stored in a desiccator over saturated aqueous
Ca(NO3)2.
2.2. Characterization
The XRD patterns were recorded using a Shimadzu LabX
˚
XRD-6000 diffractometer with Cu K␣ radiation (ꢀ = 1.54184 A).

C. Catrinescu et al. / Applied Catalysis A: General 489 (2015) 171–179
173
Table 2
to a smaller extent, by the arrangement of the particles with respect
to each other (microstructure). Previous work revealed that the
isotherms for Al-SD and Al-SAZ-1 approached type IIb [8] and dis-
played hysteresis loops associated with slit shaped pores between
plate-shaped particles. The high surface areas were attributed to
the presence of narrow micropores [8].
The main characteristics of the Al3+-exchanged clays.
Clay
d001, Å AsBET, m2 g−1 Acidity, mmol H+ g−1 clay CEC, meq 100 g−1
Al-SD
Al-SAz-1 14.8
14.8
140
96
1.2
1.64
81
120
The nature of the acid sites generated on the Al3+-exchanged
clays were explored by comparing the variable temperature dif-
fuse reflectance infrared Fourier transform spectra (VT-DRIFTS) of
pyridine-treated samples and the number of sites were estimated
using thermal desorption (TG) of cyclohexylamine (CHA). In accor-
dance with our previous studies [8,16,20], Al3+-exchanged clays
exhibited bands at 1635 and 1540 cm−1 attributed to pyridinium
ion, formed on Brønsted acid sites (BPYR), and some minor bands at
1613 and 1450 cm−1 that are diagnostic for pyridine co-ordinately
3. Results and discussion
3.1. Characterization
The chemical analysis of the original and ion-exchanged clays
(Table 1) confirmed that SAz-1 contained a large amount of struc-
tural magnesium – located in the octahedral sheet – and, in
comparison to SD, exhibited a higher cation exchange capacity
(CEC) which reflected a higher layer charge. The purified SD sam-
ple is an iron-rich clay that displayed a markedly higher specific
surface area than SAz-1. The most important results of the physical-
chemical characterization are summarized in Table 2.
bound to Lewis acid sites (LPYR). The peaks at 1596 and 1440 cm−1
,
in the spectra recorded at lower temperatures, report the presence
of H-bonded pyridine (HPYR). All these species contribute to the
intensity of the 1490 cm−1 band although the largest contribution
at higher temperatures (>120 ◦C) is from BPYR.
The XRD trace of the unpurified SD confirmed that the major
components were montmorillonite and magnesian calcite (C).
Smaller amounts of anorthite (A), feldspar (F) and quartz (Q)
were present as impurities (Fig. 1). However, most of the impu-
rities were removed during the purification step, as shown in
the trace obtained from NaSD. The powder XRD patterns of the
Al3+-exchanged clays revealed peaks corresponding to montmo-
The evolution of these diagnostic bands with temperature is pre-
sented in Fig. 2. The spectrum recorded at 50 ◦C for pyridine-treated
Al3+-exchanged samples revealed an already well-defined BPYR
peak at 1539 cm−1. This peak increased in intensity and reached
a maximum at 150 ◦C, then diminished progressively for activa-
tion at higher temperatures. The continued presence of this band
rillonites, with well defined d001 reflections (5.96◦ 2ꢁ, 14.8 A),
suggesting a well-ordered arrangement of the clay platelets, with
two layers of hydration water in the interlayer prior to activation
at 150 ◦C.
at 250 ◦C highlighted the strength of the Brønsted acid sites on Al3+
-
˚
exchanged clays. The spectrum of pyridine-treated Al-SD at 150 ◦C
exhibited strong bands, diagnostic of pyridine bound to Brønsted
acid sites, at 1490, 1540 and 1653 cm−1 [8,16].
The textural properties of Al3+- exchanged clays were inves-
tigated using nitrogen adsorption at −196 ◦C. The values of the
specific surface areas, obtained by applying the B.E.T. method [23]
to the corresponding isotherms, are presented in Table 2. These
represent the accessible surface in the unswollen state of the clay
and is determined by the microporosity resulting from the quasi-
crystalline overlap region and from the accessible areas of the
interlayer [8]. The total specific surface area might also be affected,
Thermal desorption of cyclohexylamine was used to quantify
the number of the acid sites on clay catalysts [16]. The technique
determines the weight loss between 280 and 440 ◦C and converts
it to the number of mmol of CHA desorbed. The relative ease of
obtaining this quantity has popularized its use, even though the
value obtained does not readily distinguish between cyclohexy-
lamine bound to Brønsted or Lewis acid sites [20]. The quantity of
CHA desorbed from Al-SD, in the appropriate temperature interval
AlSAz-1
AlSD
NaSD (purified)
Raw-SD
C
A
A
Q
F
5
10
15
20
25
30
35
40
2-theta, degrees
Fig. 1. The powder XRD patterns for raw, purified and Al3+-exchanged clays. Impurities: magnesian calcite (C), anorthite (A), feldspar (F) and quartz (Q).

174
C. Catrinescu et al. / Applied Catalysis A: General 489 (2015) 171–179
e
d
c
b
!
1650
1600
1550
1500
1450
1400
Wavenumbers, 1/cm
Fig. 2. VT-DRIFTS spectra of Al-SAz-1 following exposure to pyridine vapour and evacuation at (a) 50 ◦C, (b) 100 ◦C, (c) 150 ◦C, (d) 200 ◦C and (e) 250 ◦C.
(Table 2) was lower than that desorbed from Al-SAz-1, which cor-
related well with the difference in CEC between SAz-1 (120 meq
100 g−1) and SD (81 meq 100 g−1).
3.2.2. Reaction products and intermediates
Scheme 1 illustrates the main reaction products and intermedi-
ates identified using GC–MS (Supporting information). Pinene (1)
reacted with methanol over the acid sites available on the clay
surface to form terpinyl methylether, TME, (8) as the main reac-
tion product. Other compounds were also identified in the complex
reaction mixture and the most important intermediates were iden-
tified as: (i) bi-cyclic terpenes: camphene (9) and unreacted pinene
(1); (ii) monocyclic terpenes: limonene (5), terpinolene (6) and ter-
pinene (7), and (iii) bicyclic ethers (␣-fenchyl methyl ether, and
bornyl methylethers).
A significant difference between SAz-1 and SD is the extent of
replacement of aluminium by magnesium in the octahedral layer,
resulting in different layer charges which are identified by the dif-
ferent cation exchange capacities (CEC) and acidities, as shown in
the last two columns in Table 2. The location and the density of
the layer charge are the main characteristics that are expected to
influence the catalytic activity, provided that the interlayer space
is accessible to the reagents.
In order to explain the formation of these intermediates, it is
reasonable to assume that protons generated by the polarization
of water [5,19,20], or methanol [8], by the small, highly-charged
Al3+-cations will protonate pinene (1), leading to the pinyl ion
(2). Another possibility is the generation of methoxonium ions
from methanol, which will act as the protonating agent of pinene,
leading to the same pinyl ion (2). Then, the acid-catalysed pro-
cess proceeds via two parallel pathways: (i) ring expansion, via
the bornyl ion (4), giving rise to bi- and tricyclic isomerization
products (camphene, 9) and, after the addition of methanol, to
bicyclic ethers (methyl fenchyl ether, 10, methyl bornyl ether,
11), and (ii) via the terpinyl ion (3), yielding monocyclic ter-
penes as isomerization products (limonene, 5, terpinolene, 6,
terpinene, 7) and terpinyl methylethers as methoxylation products
(terpinyl methylether, 8). This range of intermediates and prod-
ucts conforms with the reaction scheme proposed by Pito et al.
[11,12].
A typical set of concentration versus time data, obtained for
the alkoxylation of ␣-pinene over Al-SAz-1 and Al-SD catalysts, is
presented in Fig. 4. This figure also presents the yield of pinene
and terpinyl methylether, as a function of time elapsed, along with
the content of the other products identified above. The solid lines
linking the decrease in ␣-pinene concentration at early times iden-
tify the ranges used for the calculation of the initial reaction rates,
estimated as described in the experimental section.
3.2. Catalytic tests
3.2.1. Potential mass transfer limitations
A preliminary investigation was conducted to identify the
experimental conditions that ensure a true kinetic regime for the
tests, i.e. the absence of external mass transfer limitations. Initial
rate measurements were carried out at different stirring rates, cata-
lyst loadings and temperatures. The extent of the external diffusion
has been verified by performing experiments at increasing stirring
rates (100, 250, 500 and 750 and 1000 rpm at 60 ◦C). The reaction
rate increased up to 500 rpm then approached an asymptotic value
beyond which the external mass transfer effects were negligible.
Having adopted a reaction temperature of 60 ◦C and a stirring
rate of 750 rpm, the kinetic regime was confirmed by perform-
ing experiments at increasing catalysts loads. The absence of
mass transfer limitations at this temperature also guaranteed their
absence at lower temperatures, where the reaction was slower. The
experimental results were plotted using 1/r = f(1/m) charts (Fig. 3),
where r is the reaction rate and m is the mass of catalyst used
[21,22]. The plots were linear, for 0.50–4.0% (w/v) range, and passed
through the origin, confirming that the mass transfer resistance
was negligible in this range. The value of the reaction rate per
unit catalyst mass (3 × 10−6 mol s−1 g−1) was constant over this
interval and decreased to values below 2 × 10−6 mol s−1 at catalyst
loadings in excess of 10% w/v. Significantly lower reaction rates
(<10−7 mol s−1 g−1) were reported by Pito et al. [11,12] and Matos
[13] in catalytic tests performed over extended reaction time, of
30–60 h.
The most important feature of this catalytic system was the
high selectivity towards the main product, ␣-terpinyl methyl ether,
especially at high pinene conversions. This behaviour contrasts
with that of limonene alkoxylation, when high selectivity was
observed during the early stages of reaction and decreased at high

C. Catrinescu et al. / Applied Catalysis A: General 489 (2015) 171–179
175
Fig. 3. The influence of the catalyst load on the initial reaction rate over AlSAz-1 and AlSD. Inset: Early time behaviour. Reaction conditions: 10 ml methanol, 0.5 ml pinene,
60 ◦C, 750 rpm.
limonene conversions. The system studied here is of considerable
practical importance because it offers good selectivities towards
the mono-ether (65%), even at high pinene conversion (higher than
60%). Moreover, the Al3+-clay catalysts were able to produce these
high yields in only 1 h whereas other solid acids [11–13] required
up to 50 h to produce similar results.
The selectivity towards ␣-terpinyl methyl ether (Fig. 4) was sim-
ilar, irrespective of the nature of the host clay, suggesting that the
same mechanism was operating in both cases. This is consistent
with the proposed model, with protons being produced from water
molecules that have been strongly polarized by the exchangeable
Al3+-cations present in the interlayer space of the clay [8,16,19,20].
The data in Fig. 4 clearly demonstrate that the catalytic activ-
ity was profoundly influenced by the nature of the starting clay
(SAz-1 or SD). Catalytic activities over clays, whether they arise
from Brønsted or Lewis acidity, generally correlate with the cation-
exchange capacity (CEC) of the base clay, provided the reactant can
enter the gallery. Thus, SAz-1, a well-known montmorillonite of
relatively high CEC (120 meq (100 g clay)−1), would provide more
acid sites than a similarly exchanged SD sample (CEC 81 meq
(100 g clay)−1). The catalytic activities recorded in this reaction
clearly reflected the relative abundance of acid sites on the catalyst
surface. Thus, for these Al3+-exchanged clays the order of activities
was the same as the order of acid site concentrations, as determined
by CHA desorption experiments (Table 2), which were also related
to the CEC of the selected clay.
3.2.3. Influence of the thermal activation temperature
Fig. 5, which presents the pinene conversion over AlSAz-1
catalyst pretreated at different temperatures, supported the gen-
erally accepted model that the acid character of an Al3+-exchanged
montmorillonite is strongly dependent on the thermal activation
procedure. The deliberately hydrated samples, displayed a very
low catalytic activity, suggesting that the acid sites were blocked
by water making them inactive towards the reactants. Increasing
the pretreatment temperature to 80 and 100 ◦C caused a marked
increase in activity. The maximum Brønsted acidity was generated
upon thermal activation at 150 ◦C, as shown in the VT-DRIFTS spec-
tra of the pyridine treated Al3+-clays, but only a slight increase in
activity was observed at this temperature. Increasing the thermal
Scheme 1. The main reaction products and intermediates identified by GC–MS in the reaction mixture.

176
C. Catrinescu et al. / Applied Catalysis A: General 489 (2015) 171–179
Fig. 4. The change in composition of the reaction mixture during pinene methoxylation over (a and b) AlSAz-1 and (c and d) AlSD. Reaction conditions: 200 mg catalyst,
10 ml methanol, 0.5 ml pinene, 60 ◦C.
activation temperature above 150 ◦C, is known to cause a reduction
in the Brønsted acidity and an increase in the Lewis acidity in accor-
dance with the model in which the exchange ions become electron
pair accepting, or Lewis acid, sites as the directly coordinated water
is driven off [16]. Therefore, a pretreatment temperature of 150 ◦C
was selected for further studies, in order to avoid the uncertainty
regarding the Brønsted/Lewis acid balance and to take advantage
of the maximum number of the Brønsted acid sites.
The variation of selectivity towards terpinyl methylether, TME,
with the extent of pinene conversion was also scrutinized. The data
in Fig. 5B clearly illustrate that the activation temperature exerted
little influence on selectivity, especially at low pinene conversion.
Therefore, while the activation temperature exerted a consider-
able effect on the reaction rate, the selectivity toward TME was
not influenced in any significant way.
3.2.4. Influence of the reaction temperature
The influence of the reaction temperature on pinene con-
version and selectivity towards the mono-ether is presented
in Fig. 6A and B. The catalysed reactions were carried out at
different temperatures (35, 45, 55 and 65 ◦C) over 100 mg of
Al-SAz-1 catalyst while the pinene: methanol molar ratio and
the catalyst loading were kept constant. As anticipated, pinene
conversion increased with the temperature, under otherwise iden-
tical conditions. The same trend was observed for limonene
conversion in a related system [8]. Increasing the tempera-
ture up to 65 ◦C, did not lead to a decrease in selectivity, as
observed in the previous study on limonene methoxylation [8].
The selectivity to the mono-ether (Fig. 6), at constant conversion,
seems to be largely unaffected by raising the reaction tempera-
ture.
70
60
50
40
70
60
50
40
250 C
30
150 C
150 C
30
250 C
20
100 C
100 C
80 C
20
80 C
10
0
10
0
10
20
30
40
50
60
0
10
20
30
40
50
60
70
Time, min
Conversion, wt. %
Fig. 5. The effect of the Al-clay activation temperature on the methoxylation of ␣-pinene: (A) ␣-pinene conversion vs. time and (B) the correlation selectivity vs. conversion.
Reaction conditions: 200 mg cat, 10 ml methanol, 0.5 ml pinene, 60 ◦C.

C. Catrinescu et al. / Applied Catalysis A: General 489 (2015) 171–179
177
Fig. 6. The effect of the reaction temperature on the methoxylation of ␣-pinene: (A) ␣-pinene conversion vs. time and (B) the correlation selectivity vs. conversion. Reaction
conditions: 100 mg Al-SAz-1, 0.5 ml Pinene, 10 ml MeOH.
70
60
50
40
30
20
10
0
70
65
60
55
50
45
40
35
30
25
20
0.3 mol/l
0.6 mol/l
1.05 mol/l
0.3 mol/l
0.6 mol/l
1.05 mol/l
0
10
20
30
40
50
60
0
10
20
30
40
50
60
70
Time, min.
Conversion, wt. %
Fig. 7. The effect of the initial pinene concentration on the methoxylation of ␣-pinene: (A) ␣-pinene conversion vs. time and (B) the correlation selectivity vs. conversion.
Reaction conditions: 200 mg catalyst, 10 ml methanol, 60 ◦C.
3.2.5. Influence of initial ˛-pinene concentration
0.6 mol/L, only caused a slight decrease in pinene conversion at a
fixed reaction time; this trend was further accentuated at a higher
initial concentration (1.05 mol/L). Note that the initial reaction rate
was higher when the initial concentration of pinene was higher,
The effect of the initial ␣-pinene concentration on the conver-
sion and on the selectivity towards the mono-ether is presented
in Fig. 7. Doubling the initial pinene concentration, from 0.3 to
Fig. 8. ␣-Pinene concentration versus time during the methoxylation of ␣-pinene over AlSAz-1 and AlSD. Reaction conditions: 200 mg catalyst, 10 ml methanol, 0.5 ml pinene,
60 ◦C.

178
C. Catrinescu et al. / Applied Catalysis A: General 489 (2015) 171–179
-6
-6.5
-7
AlSAz-1
AlSD
-7.5
-8
-8.5
-9
-9.5
-10
0.00295
0.003
0.00305
0.0031
0.00315
0.0032
0.00325
1/T, K-1
Fig. 9. Arrhenius plot for methoxylation of ␣-pinene over AlSAz-1 and AlSD. Reaction conditions: 100 mg catalyst, 10 ml methanol, 0.5 ml pinene.
under otherwise identical conditions. Again, the selectivity to the
mono-ether, at similar conversion, was not affected by the initial
pinene concentration.
activation temperature had a strong influence on the acidity of
the clay surface and was, consequently, a significant factor in the
control of the reaction rate. In addition, other reaction parameters,
such as the reaction temperature and the initial concentration of
␣-pinene could be used to optimize the catalytic activity. The selec-
tivity towards the mono-ether at constant conversion was much
less dependent on these parameters.
The most important feature of this catalytic system was the
high selectivity towards the main product, ␣-terpinyl methyl ether,
especially at high pinene conversions. In this sense, the selectivity
towards TME (65%), obtained over Al3+-exchanged clays, is simi-
lar to that obtained by Pito et al. [11,12] and Mato et al. [13] (60%)
but at a much higher reaction rate. On the other hand, Hoelderich
[7] suggested that zeolite ␤ is more active than the clay-based
catalysts, at the expense of the selectivity, which is significantly
poorer (54%), especially at high conversions. A different trend was
observed in the related alkoxylation of limonene, over the same
catalysts, where the selectivity decreased as the limonene conver-
sion progressed. The reaction followed pseudo first-order kinetics,
with similar activation energies for both catalysts.
3.3. Kinetic study
The linear form of the logarithmic plot of ␣-pinene concen-
tration versus time, demonstrated that the reaction over the clay
catalysts followed pseudo first-order kinetics; i.e. first order with
respect to ␣-pinene and zero-order with respect to methanol,
because the latter was present in large excess. The values of the
reaction rate constants, determined from the slopes of the two lines
of negative slope in Fig. 8, confirmed that the rate of reaction at 60 ◦C
was faster over Al-SD than Al-SAz.
An Arrhenius plot of these reaction rate constants is shown
Fig. 9. The slopes of the ln r versus 1/T lines were similar for both
Al-exchanged clays, and an activation energy of 71.7 kJ/mol was
obtained for the alkoxylation reaction.
The similarity in activation energies suggested that the differ-
ence in rate constants must be attributed to a difference in the
pre-exponential factor. An acceptable explanation for the faster
rate over AlSD is that the significantly higher external surface area
makes it easier for the reactant and product molecules to arrive at
the Brønsted acid sites in the gallery. This suggests that the rate
determining step involves transport from the liquid phase to the
interlayer space perhaps via a ‘stagnant’ layer of reaction medium
close to the clay surface. Alternatively, the larger distance between
exchange sites in AlSD, which has a lower density of exchange sites
per unit surface, may control or contribute to easier access to the
reaction sites in the clay interlayer. The high charge density of the
AlSAz sample may result in interlayer congestion which results in
a slower rate of turnover at the acid sites.
Acknowledgements
This work was supported by Portuguese funds (FCT
–
Fundac¸ ão para a Ciência e a Tecnologia, projects PTDC/CTM-
CER/121295/2010 and PEst OE/QUI/UI0619/2011).
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.apcata.
2014.10.028.
References
4. Conclusion
[1] P.T. Anastas, L.B. Bartlett, M.M. Kirchhoff, T.C. Williamson, Catal. Today 55
(2000) 11–22.
[2] Y. Noma, Y. Asakawa, in: K.H.C. Baser, G. Buchbauer (Eds.), Handbook of Essen-
tial Oils: Science, Technology, and Applications, CRC Press, Boca Raton, 2010,
pp. 585–736.
[3] E.V. Gusevskaya, ChemCatChem 6 (2014) 1506–1515.
[4] M. Besson, P. Gallezot, C. Pinel, Chem. Rev. 114 (2014) 1827–1870.
[5] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411–2502.
[6] P. Mäki-Arvela, B. Holmbom, T. Salmi, D.Y. Murzin, Cat. Rev. 49 (2007) 197–340.
[7] K. Hensen, C. Mahaim, W.F. Hölderich, Appl. Catal. A: Gen. 149 (1997) 311–329.
[8] C. Catrinescu, C. Fernandes, P. Castilho, C. Breen, M.M.L. Carrott, I.P.P. Cansado,
Appl. Catal. A: Gen. 467 (2013) 38–46.
Modified clays have been proposed as active and selective cata-
lysts for the synthesis of ␣-terpinyl methyl ether by alkoxylation of
␣-pinene with methanol. The GC–MS analysis of the reaction mix-
tures showed that ␣-terpinyl methyl ether was obtained with good
selectivity over the Al3+-exchanged clay catalysts. AlSAz-1 was sig-
nificantly more active than the similarly exchanged SD, and this
behaviour was rationalized taking into account the higher number
of Brønsted acid sites present in this high charge clay, as shown
by the acidity measurements (VT-DRIFTS and TG of CHA). The

C. Catrinescu et al. / Applied Catalysis A: General 489 (2015) 171–179
179
[9] M. Yoshiharu, M. Masahiro, Jpn. Kokai 75 (1976) 948.
[10] J.E. Castanheiro, L. Guerreiro, I.M. Fonseca, A.M. Ramos, J. Vital, Stud. Surf. Sci.
Catal. 174 (2008) 1319–1322.
[15] F. Bergaya, B.K.G. Theng, G. Lagaly, Handbook of Clay Science, vol. 1, Elsevier,
2006, pp. 261 (Chapter 7).
[16] C. Breen, Clay Miner. 26 (1991) 473–486.
[11] D.S. Pito, I.M. Fonseca, A.M. Ramos, J. Vital, J.E. Castanheiro, Chem. Eng. J. 147
(2009) 302–306.
[12] D.S. Pito, I. Matos, I.M. Fonseca, A.M. Ramos, J. Vital, J.E. Castanheiro, Appl. Catal.
A: Gen. 373 (2010) 140–146.
[17] L.J. Arroyo, H. Li, B.J. Teppen, S.A. Boyd, Clays Clay Miner. 53 (2005) 512–520.
[18] S. Kaufhold, R. Dohrmann, M. Klinkenberg, S. Siegesmund, K. Ufer, J. Colloid
Interface Sci. 349 (2010) 275–282.
[19] C. Breen, Clay Miner. 26 (1991) 487–496.
[13] I. Matos, M.F. Silva, R. Ruiz-Rosas, J. Vital, J. Rodríguez-Mirasol, T. Cordero,
J.E. Castanheiro, I.M. Fonseca, Micropor. Mesopor. Mater. 199 (2014)
66–73.
[14] J.M. Adams, R.W. McCabe, Handbook of Clay Science, vol. 1, Elsevier, 2006, pp.
541–581 (Chapter 10.2).
[20] P. Komadel, M. Janek, J. Madejova, A. Weekes, C. Breen, J. Chem. Soc. Faraday
Trans. 93 (1997) 4207–4210.
[21] T.K. Sherwood, Pure Appl. Chem. 10 (1965) 595–610.
[22] H. Hichri, A. Accary, J. Andrieu, Chem. Eng. Process. 30 (1991) 133–140.
[23] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–319.