Biomass to chemicals: Rearrangement of β-pinene epoxide into myrtanal with well-defined single-site substituted molecular sieves as reusable solid Lewis-acid catalysts

Article abstract of DOI:10.1016/j.apcata.2010.03.056

The epoxide rearrangement of β-pinene oxide into myrtanal was studied in the presence of solid Lewisacid catalysts. Different metals such as Zr, Sn, Ti, Nb, Ta, Al, and Ga have been incorporated within network positions of zeolite Beta by isomorphous substitution. The materials have been characterized by powdered X-ray diffraction (XRD), IR spectroscopy and by in situ IR spectroscopy on adsorbed molecules (cyclohexanone and pyridine). Among the studied catalysts Zr-Beta gave the best selectivity (up to 94%) in acetonitrile as solvent at practically complete conversion. In this epoxide rearrangement the solvent has to be selected to balance the competitive adsorption of the product. Product desorption is enhanced with acetonitrile which results in a selectivity increase with only a small penalty on the rate of reaction. Zr-Beta has been reused in the batch mode and in a fixed-bed reactor for several times and analysis of the reused catalyst indicated that Zr-Beta is a robust catalyst for the epoxide rearrangement. Leaching of the metal or crystal degradation has not been detected.

Full text of DOI:10.1016/j.apcata.2010.03.056

Applied Catalysis A: General 380 (2010) 165–171
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Biomass to chemicals: Rearrangement of ␤-pinene epoxide into myrtanal with
well-defined single-site substituted molecular sieves as reusable solid Lewis-acid
catalysts
Olalla de la Torre, Michael Renz, Avelino Corma∗
Instituto de Tecnología Química (UPV-CSIC), Universidad Politécnica de Valencia–Consejo Superior de Investigaciones Científicas, Avenida de los Naranjos s/n, 46022 Valencia, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
The epoxide rearrangement of ␤-pinene oxide into myrtanal was studied in the presence of solid Lewis-
acid catalysts. Different metals such as Zr, Sn, Ti, Nb, Ta, Al, and Ga have been incorporated within network
positions of zeolite Beta by isomorphous substitution. The materials have been characterized by pow-
dered X-ray diffraction (XRD), IR spectroscopy and by in situ IR spectroscopy on adsorbed molecules
(cyclohexanone and pyridine). Among the studied catalysts Zr-Beta gave the best selectivity (up to 94%)
in acetonitrile as solvent at practically complete conversion. In this epoxide rearrangement the solvent
has to be selected to balance the competitive adsorption of the product. Product desorption is enhanced
with acetonitrile which results in a selectivity increase with only a small penalty on the rate of reaction.
Zr-Beta has been reused in the batch mode and in a fixed-bed reactor for several times and analysis of
the reused catalyst indicated that Zr-Beta is a robust catalyst for the epoxide rearrangement. Leaching of
the metal or crystal degradation has not been detected.
Received 21 January 2010
Received in revised form 8 March 2010
Accepted 25 March 2010
Available online 31 March 2010
Keywords:
Beta zeolite
Epoxide rearrangement
Heterogeneous catalysis
Solid Lewis acid
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
activity for others. For instance, materials involving titanium sites
incorporated into a siliceous zeolite as can be titanium silicalite-1
The development of reusable solid catalysts with isolated and
well-defined single sites is a matter of scientific and industrial inter-
est to achieve good activities and selectivities [1,2]. Zeolites are
very convenient materials to design single or multisite catalysts,
since well-defined Brönsted and Lewis acids (metals) can be intro-
duced within framework positions. The electronic characteristics
of these sites as well as the adsorption characteristics of the mate-
rial can be modulated during the synthesis or even by postsynthesis
treatments to adapt the solid to the specific reaction to be catalyzed.
The catalytic activity of Brönsted acids can be correlated with
their ability to form acid–base adducts through proton transfer. In
contrast, in the case of zeolites with Lewis-acid sites the molecular
interaction depends on several parameters [3] such as: the intrin-
sic acid strength (measured as the energy of the lowest unoccupied
molecular orbital—LUMO), a back-donation of electron density
from the Lewis acid to the empty orbitals of the base, and changes
in the electronic levels of the active site caused by its interaction
with the reactant molecules.
(TS-1) or Ti-Beta, are excellent catalysts for the activation of hydro-
gen peroxide on the metal to perform olefin epoxidation reactions
[4], while the metal active site of Sn-Beta does not activate hydro-
gen peroxide, but interacts with a carbonyl group of the reactant
molecule by accepting electron density. However, this electron
density is not transferred into empty orbitals of the tin center but
into anti-bonding molecular orbitals centered at the adjacent oxy-
gen atoms [5]. With the help of this additional electron density
hydrogen peroxide can be bound to these oxygen atoms via hydro-
gen bonding. Both reactants adsorbed onto the zeolite constitute
a perfect starting point for the Baeyer–Villiger (BV) oxidation of
the carbonyl compound (see Fig. 1). Thus, although both metals, Ti
and Sn, are tetra-valent Lewis acids and have isomorphically sub-
stituted silicon atoms in the same zeolite framework, the catalytic
activity of the titanium material is excellent for olefin epoxidation
and very poor for the BV oxidation. On the other hand, the activity
of Sn-Beta for epoxidations with hydrogen peroxide is negligible
while it performs very well the BV oxidation of cyclic ketones.
If we now select a third type of reaction which is also cat-
alyzed by Lewis acids, i.e. the Meerwein–Ponndorf–Verley (MPV)
reduction [6], it has been observed that the catalytic performance
of different metal substituted zeolites can change with the type
of substrate. When cyclohexanone was contacted with 2-butanol,
Sn-Beta gave a four-fold activity with respect to Zr-Beta, whereas
the activity was almost inverse for the mixture benzaldehyde/2-
The relative importance of the different parameters involved in
the catalytic phenomena with Lewis-acid sites dictates why some
Lewis-acid catalysts are effective for one reaction and show no
∗
Corresponding author. Tel.: +34 96 387 78 00; fax: +34 96 387 78 09.
E-mail address: acorma@itq.upv.es (A. Corma).
0926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.03.056

166
O. de la Torre et al. / Applied Catalysis A: General 380 (2010) 165–171
2. Experimental
2.1. Catalyst preparation and characterization
The syntheses of metal-containing zeolite Beta materials were
performed in fluoride medium from gels of the following compo-
sition [14]:
SiO2:xM : 0.54TEAOH : 0.54HF :yH2O
where x was 0.01 or 0.02, and y was varied between 4 and 7.
The reactants were zirconyl chloride, Sn(IV) chloride pentahydrate,
Ti(IV), Nb(V) and Ta(V) ethoxides, Al(III) isopropoxide, and Ga(III)
nitrate as different metal sources and for all zeolites in common
tetraethylorthosilicate as the silica source, tetraethylammonium
hydroxide (TEAOH, 35 wt% in water) and HF (50 wt% in water). Dea-
luminated zeolite Beta seeds were also employed for the synthesis.
The crystallization was carried out in Teflon-lined stainless-steel
autoclaves at 140 ◦C under rotation for different periods of time.
After the desired time, the autoclaves were cooled down to room
temperature and the zeolites were collected by filtration, washed
with water and dried at 100 ◦C overnight. The as-synthesized
metal-containing zeolite Beta materials were calcined in static air
at 580 ◦C for 3 h in an oven in order to remove the occluded organic.
Crystallinity and phase identification of the materials was done
by powder X-ray diffraction (XRD) in a PANalytical CUBIX diffrac-
tometer equipped with a PW3050 goniometer (Cu K␣ radiation)
provided with a variable divergence slit. The metal content of the
as-synthesized and calcined samples was determined by chemical
analysis (Varian 715-ES ICP-Optical Emission Spectrometer) after
dissolution of the solids in a HNO3/HF solution. The following sili-
con/metal ratios were determined: Si/Zr = 130 and 115, Si/Sn = 130,
Si/Ti = 132, Si/Nb = 109, Si/Ta = 78, Si/Ga = 45, and Si/Al = 103. The
XRD patterns of the calcined samples (580 ◦C/3 h) confirmed high
crystallinity and peaks of the corresponding metal oxides were
not observed (patterns not shown). Nitrogen adsorption exper-
iments on the calcined samples (580 ◦C/3 h) gave an isotherm
very similar to that of pure silica-Beta with micropore volumes of
Fig. 1. Energy of the LUMO of the metal complexes with cyclohexanone coordinated
to the metal and hydrogen peroxide to the adjacent oxygen atom of the hydroxy
group calculated for the Ti-Beta and the Sn-Beta catalyst. For Sn-Beta the LUMO is
the ␲ anti-bonding orbital of the carbonyl group that can be attacked by the HOMO of
the hydrogen peroxide. For Ti-Beta the LUMO is located at the metal and therefore
the hydrogen peroxide tends to coordinate to the metal instead of attacking the
carbonyl bond.
butanol [3]. From the examples given above, plus much results
reported in the literature on Lewis-acid reactivity [2,7,8], it appears
difficult to predict a priori or by simple analogy the best active site
for a given reaction.
Recently we have shown that Lewis acids can be used for the
transformation of biomass into fine chemical products [9,10]. Myr-
tanal has been isolated from natural products, such as the roots
of Greek Paeonia taxa among others [11]. These plants have been
used through antiquity, as antiseptic and for cleansing wounds. Up
to 25% of myrtanal was found in essential oils from Geum urbanum
roots [12]. Synthetically it has been obtained from ␤-pinene oxide
with 45% yield using solid catalysts such as aluminium oxides [13].
Owing to the industrial interest of the reaction, we have stud-
ied here the possibilities of metal zeolites with well-defined single
isolated sites, to catalyze the synthesis of myrtanal from ␤-pinene
oxide by epoxide rearrangement (Eq. (1)).
0.20–0.21 cm3 g−1 and BET surface areas of 450–475 m2 g−1
.
2.2. Metal characterization
In order to ascertain the metal coordination and consequently
to support their location in framework positions the following
techniques were used: UV–visible DR for Ti (not shown) and EXAFS-
XANES for Sn, Nb, and Ta. The Lewis acidity was evidenced by
adsorption–desorption of pyridine and/or cyclohexanone followed
by IR spectroscopy.
Infrared (IR) measurements were performed with a Nicolet 710
FTIR spectrometer using vacuum cells. The spectra in the frame-
work vibration region were done using the KBr pellet technique,
whereas experiments with adsorption of pyridine and cyclohex-
anone as probe molecules were performed in self supported wafers
of 10 mg cm−2 that were degassed overnight under vacuum (10−2
to 10−3 Pa) at 400 ◦C. The spectra were recorded and then pyri-
dine or cyclohexanone was admitted and, after equilibration,
the samples were outgassed for 1 h at increasing temperatures
(150/250/350 ◦C for pyridine and 25/50/100/200 ◦C for cyclohex-
anone). After each desorption step, the spectrum was recorded at
room temperature and the background subtracted in the pyridine
case.
(1)
It will be shown that by properly selecting the catalyst and reac-
tion conditions it is possible to achieve 94% yield of the desired
myrtanal product which is much higher than any other value
reported up to now.
2.3. Catalytic reaction
␤-Pinene oxide was prepared by a standard epoxidation method
and was employed in 96% purity (1H NMR). 600 mg of the epoxide

O. de la Torre et al. / Applied Catalysis A: General 380 (2010) 165–171
167
was dissolved in 3 g of solvent. A 50-mg sample of the catalyst was
added and the reaction mixture heated to 80 ◦C while stirring. The
reaction was monitored by gas chromatography with a Varian 3900
GC equipped with a Carbowax column (15 m length, 0.32 mm inner
diameter and 0.25 ␮m film).
Myrtanal was identified by mass spectroscopy and NMR spec-
troscopy on a sample distilled from the reaction mixture: MS m/z
(%): 152(2) [M]+, 29(47), 39(58), 41(100), 67(91), 69(80), 81(70),
82(63), 83(37), 123(54); 13C NMR (75 MHz, CDCl3) ı = 204.0, 49.0,
40.5, 40.2, 39.4, 26.2, 23.9, 23.5, 20.4, 12.9.
For the recycling tests in the batch mode 6.00 g of the epoxide in
30 g of acetonitrile solvent and 500 mg of the Zr-Beta catalyst were
heated to 80 ◦C for 2 h. The reaction mixture was filtered and the
recovered solid catalyst re-activated by calcination at 580 ◦C for 3 h.
For the subsequent cycle the amount of solvent and epoxide were
scaled to the available quantity of catalyst. After 8 recycles 50% of
the catalyst had been lost due to the catalyst handling (6.25% on an
average per cycle). The crude reaction mixtures were analyzed as
described above.
For the fixed-bed continuous-flow reactor study Zr-Beta-2 was
pelletized and sieved (0.4–0.6 mm) and 800 mg of the catalyst
placed in a stainless-steel tube (14 cm, 4 mm inner diameter). The
reactor tube was heated to 80 ◦C and the feed (17 wt% of ␤-pinene
oxide in acetonitrile) was passed at a rate of 0.5 mL/min that cor-
responded to a weight hourly space velocity (WHSV) of 30 h−1 and
a contact time of 2 min. The product was collected at the end of
the reactor at room temperature. When the conversion started to
decrease the reaction was stopped and the catalyst re-activated by
passing an air flow (30 mL/min) at 500 ◦C for 20 min. After cool-
ing down to 80 ◦C the reaction was continued. The crude reaction
mixtures were analyzed off-line as described above.
3. Results and discussion
3.1. Catalyst characterization
All Beta catalysts were synthesized with hydrogen fluoride as
mineralizer and were activated by calcination in air (3 h, 580 ◦C).
The zeolite structure was confirmed by powdered X-ray diffrac-
tion, and the incorporation of the metal into crystallographic
network positions has been demonstrated before by extended X-
ray absorption fine structure (EXAFS) and X-ray absorption near
edge structure (XANES), working with Ti [15], Sn [16], Nb [17], and
Ta [17] substituted Beta samples. In the original papers it has been
shown that when the metals were incorporated into the zeolite
framework, the resultant zeolites present Lewis acidity as demon-
strated by adsorption of a probe molecule (cyclohexanone; cf. Fig. 2)
for the metal substituted samples used in this work. Additionally,
as a fast and simple rejection technique we have used IR spec-
troscopy since the spectra of metal substituted zeolites show a band
at 960 cm−1 (not shown). In other words, the presence of a band at
960 cm−1 is a necessary fact if a metal is incorporated, but it is not
sufficient to ascertain the incorporation [15,18,19].
For comparison purposes we have also prepared the Beta alumi-
nosilicate (Si/Al = 103) and a Ga silicate (Si/Ga = 45). These samples
show both Lewis and Brönsted as evidenced by pyridine adsorption
and IR spectroscopy (see Fig. 3).
3.2. Activity tests
In a preliminary screening we have seen here that Zr-Beta, Sn-
Beta, and Ti-Beta give high catalytic activity for this reaction and
therefore these three solids have been chosen to study the influence
of the nature of solvents. Hence, ␤-pinene oxide was dissolved in
different solvents and contacted with the three Lewis-acidic Beta-
Fig. 2. In situ infrared spectra of cyclohexanone adsorbed onto different Beta zeo-
lites. In the figure the corresponding spectra are displayed after zeolite activation
(straight line), after desorption at 100 ◦C (dash line) and after desorption at 200 ◦
(dash dot line; 150 ◦C for Zr-Beta).
C

168
O. de la Torre et al. / Applied Catalysis A: General 380 (2010) 165–171
Fig. 3. In situ infrared spectra (a) of the hydroxide region of Al-Beta and Ga-Beta before adsorption of pyridine (thick line) and after adsorption of pyridine and desorption
at 150 ◦C (thin line) and (b) of the difference for pyridine adsorbed onto Al-Beta and Ga-Beta after desorption at 150 ◦C and the corresponding zeolite background spectra.
zeolites, i.e. Zr-Beta, Sn-Beta, and Ti-Beta. The best selectivity was
observed with the Zr-Beta catalyst in acetonitrile solvent (Table 1,
entry 1). At a reaction temperature of 80 ◦C and after a reaction
time of 2 h myrtanal (2) was achieved with 94% selectivity at almost
complete conversion. With the Sn-Beta catalyst similar results were
obtained with almost 90% selectivity at the same conversion level
(Table 1, entry 2). A significantly lower conversion at the same
reaction time was observed for Ti-Beta (Table 1, entry 3).
A comparison experiment using acetonitrile (2 h), nitromethane
(30 min) or 1,4-dioxane (30 min) as solvent indicated that a com-
petitive adsorption between acetonitrile and the reactant on the
Lewis-acid center occurs. This is well in line with the Lewis-basic
properties of acetonitrile, that make this molecule suitable to be
used as probe molecule to study Lewis acidity [20]. On the other
hand nitromethane and 1,4-dioxane are more weakly adsorbed on
the Lewis-acid sites than acetonitrile.
With Sn-Beta in most solvents a lower product selectivity was
observed with respect to the other two catalysts. The affinity of the
tin center towards the carbonyl group of the product myrtanal (2)
might be a possible explanation for this lower selectivity. In our pre-
vious theoretical calculations [3] for the interaction of the carbonyl
bond of adsorbed cyclohexanone with the three different metals,
i.e. Sn, Zr, and Ti, interaction is strongest for Sn-Beta. Hence, due to
a longer adsorption time of the product on the Lewis-acid center
in the case of Sn-Beta, the probability for further rearrangement
reactions into structural isomers will be increased. When using
acetonitrile as solvent, this can compete successfully with the car-
bonyl group of the myrtanal for adsorption, and facilitates product
desorption, increasing thereof the selectivity from 70 to 89%.
The influence of solvent properties on product selectivity with
the other two catalysts, i.e. Zr- and Ti-Beta, is not very pronounced.
For Zr-Beta, the selectivities towards myrtanal (2) vary only within
Table 1
␤-Pinene epoxide rearrangement in different solvents catalyzed by Zr-, Sn-, and Ti-Beta zeolite at 80 ◦C.
Entry
Catalyst
Solvent
Time [h]
Conv. [%]
Product selectivity
Yield 2 [%]
2 [%]
3 [%]
4 [%]
Other [%]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Zr-Beta
Sn-Beta
Ti-Beta
Zr-Beta
Sn-Beta
Ti-Beta
Zr-Beta
Sn-Beta
Ti-Beta
Zr-Beta
Sn-Beta
Ti-Beta
Zr-Beta
Sn-Beta
Ti-Beta
Acetonitrile
Acetonitrile
Acetonitrile
Acetonea
2.0
2.0
2.0
2.0
2.0
2.0
1.0
1.0
1.0
0.5
0.5
0.5
0.5
0.5
0.5
98
98
74
98
90
99
97
94
89
86
77
70
84
89
80
89
87
76
86
90
82
87
2
2
2
2
4
2
2
4
2
2
4
1
2
2
1
1
2
2
3
7
10
10
13
6
7
11
7
92
87
64
75
63
83
86
80
77
87
73
86
90
80
85
11
Acetonea
13
8
2
5
2
3
6
3
3
6
3
Acetonea
Toluene
Toluene
Toluene
Nitromethane
Nitromethane
Nitromethane
1,4-Dioxane
1,4-Dioxane
1,4-Dioxane
>98
87
>98
96
>98
>98
97
8
14
10
5
10
9
98
a
Reaction temperature 56 ◦C.

O. de la Torre et al. / Applied Catalysis A: General 380 (2010) 165–171
169
Table 2
Comparison of different metal-containing Beta zeolites as catalysts for the ␤-pinene epoxide rearrangement in acetonitrile as solvent after 2 h reaction time at 80 ◦C.
Entry
Catalyst
Si/M ratio [mol/mol]
Conv. [%]a
Product selectivity
Yield 2 [%]
2 [%]
3 [%]
4 [%]
Other [%]
1
2
3
4
5
6
7
8
Zr-Beta
Sn-Beta
Ti-Beta
Nb-Beta
Ta-Beta
Al-Beta
Ga-Beta
Si-Beta
130
98
98
74 (84)
38 (41)
36 (40)
46 (53)
37 (42)
<5%
94
89
86
70
72
74
63
0
2
2
7
7
8
9
1
2
2
7
5
5
6
5
7
92
87
64
27
26
34
23
130
132
109
78
103
45
10
16
16
13
22
–
a
Values in parenthesis are the conversions after 4 h reaction time. Selectivities after 4 h are very close to those after 2 h.
Table 3
Recycling experiments for Zr-Beta zeolite as catalyst for the ␤-pinene epoxide rearrangement in acetonitrile as solvent. The catalyst was filtered out after each run of 2 h at
80 ◦C and re-activated by calcination (580 ◦C/3 h).
Entry
Cycle
Conv. [%]a
Product selectivity
Yield 2 [%]
2 [%]
3 [%]
4 [%]
Other [%]
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
97
>98
>98
98
>98
>98
>98
>98
>98
92
94
93
91
93
94
94
93
93
2
2
1
2
1
1
2
2
2
1
1
1
1
1
1
1
1
1
5
3
5
6
6
4
3
5
4
92
94
93
91
93
94
94
93
93
a range of 7% (except for acetone) reaching 94% with acetonitrile
(cf. Table 1). The results for Ti-Beta were similar, with a selectivity
range from 84% (acetone) to 89% (toluene).
of secondary rearrangements with the corresponding penalty on
selectivity. With this last catalyst (Sn-Beta) best results are obtained
when using acetonitrile as solvent, possibly due to the competitive
adsorption of acetonitrile on the metal Lewis-acid site and a more
efficient desorption of the desired product.
Other metal-containing Beta zeolites were also tested using
acetonitrile as solvent. Penta-valent metals such as niobium and
tantalum, incorporated into the Beta framework by isomorphous
substitution, show catalytic activity for the epoxide rearrangement
reaction, although with moderate conversion and lower selectivity
than before (cf. Table 2, entries 4 and 5). When Nb and Ta are incor-
porated into network positions of Beta zeolite, EXAFS and XANES
spectroscopy indicate that the metal bears a M O bond and is coor-
dinated to three silyloxy framework groups [17].
Brönsted acidic zeolites such as Al-Beta and Ga-Beta gave con-
versions and selectivities considerably lower than Zr-Beta (cf.
Table 2, entries 1, 6, and 7). Therefore, it can be concluded that
Brönsted-acidic zeolites are not suitable catalysts for the present
epoxide transformation. The incorporation of a metal is the origin
of the catalytic activity since all silica Beta zeolite does not give
considerable conversion under the studied reaction conditions (cf.
Table 2, entry 8).
3.3. Stability and recyclability of Zr-Beta
The reaction was carried out at a larger scale with Zr-Beta as
catalyst and acetonitrile as solvent. The catalyst was filtered out
after the reaction and re-activated by calcination in air at 580 ◦C for
3 h. From the results in Table 3 it can be seen that even after nine
reuses, conversion was practically complete after 2 h reaction time
and selectivity remained high (approximately 95%) and constant,
within the experimental error. Hence, in nine experiments 37.2 g of
epoxide 1 were transformed with 0.500 g of Zr-Beta catalyst which
corresponds to a turnover number (TON; defined as mol of reactant
converted per mol of metal) of 3878.
After the nine cycles the recovered catalyst was characterized
again and compared with the fresh catalyst. The X-ray diffraction
pattern of the re-used sample was identical to the one of the fresh
sample (cf. Table 4). SEM images of the fresh, calcined Zr-Beta cat-
alyst and the same material after the 9 catalytic cycles were very
similar and degradation of the zeolite crystals was not detected (cf.
Fig. 4a and b, respectively). ICP analysis indicated the same Si/Zr
ratio before and after the reactions.
In summary, the epoxide rearrangement of ␤-pinene oxide into
myrtanal occurred with highest selectivity (94%) at full conversion
with Zr-Beta in acetonitrile as solvent. The Lewis-acid properties of
Zr-Beta are the most adequate, providing high activity and effective
desorption of the product. In the case of Sn centers, the stronger
binding of the myrtanal product to the metal favours the occurrence
Table 4
Analysis of fresh and used catalysts.
Catalyst
TON [mol/mol]a
Crystallinity [%]b
Si/Zr ratio [mol/mol]c
Fresh calcined Zr-Beta
Zr-Beta catalyst used in 9 cycles in a batch reactor
Fresh calcined Zr-Beta-2
–
100
100
98
131
129
116
113
3878
–
2130
Zr-Beta-2 catalyst used in 6 cycles in a continuous-flow reactor
100
a
Mole of substrate converted per mol of metal sites (on an average).
Compared with a commercial sample of Al-Beta; experimental error 5%.
Determined by ICP-OES.
b
c

170
O. de la Torre et al. / Applied Catalysis A: General 380 (2010) 165–171
Fig. 4. SEM images of Zr-Beta catalyst. (a) Fresh calcined Zr-Beta, (b) Zr-Beta catalyst after re-uses in batch mode, (c) fresh Zr-Beta-2 catalyst, and (d) Zr-Beta-2 after use in
the fixed-bed continuous reactor.
Fig. 5. Epoxide rearrangement in the presence of Zr-Beta-2 in a fixed-bed continuous-flow reactor.
Taking all the above into account, it should be possible to use
Zr-Beta catalyst in a fixed-bed continuous reactor. Hence, 800 mg
of Zr-Beta-2 was placed in a tubular stainless-steel reactor and ␤-
pinene oxide dissolved in acetonitrile was passed at 80 ◦C. Again,
excellent results were obtained for the conversion and selectivity of
the ␤-pinene oxide rearrangement (cf. Fig. 5). The conversion was
complete and the selectivity towards myrtanal (2) was 92%. After
80 min time on stream the conversion started to decrease and a
short reactivation procedure was performed. To do that the reactor
was heated to 500 ◦C and air (30 mL/min) was passed through for
20 min. Afterwards, the reaction was continued. By applying the
reactivation procedure periodically, the catalyst could be used in
several cycles at full conversion with a selectivity level of 92–96%.
Thus 32.9 g of substrate were converted that corresponds to a TON
of 2130 that certainly could be larger if the reaction was contin-
ued (cf. Table 4). Analysis of the used catalyst confirmed again the
high stability of the catalyst. The used catalyst gave the same X-
ray diffraction pattern as the fresh one and SEM images show very
similar crystal size and morphology for the fresh, calcined material
and the reused catalyst (cf. Fig. 4c and d).
In previous work, we have seen that for Lewis-acid catalyzed
reactions which present no diffusion constraints for reactants and
products, Sn- or Zr-Beta are more active than the corresponding
metal MCM-41 material. For the rearrangement of ␤-pinene oxide
the results obtained here, under exactly the same reaction con-
ditions (Table 1, entries 10–12; nitromethane solvent), as well as
from the results obtained previously [10] we can say that M-Beta
is also more active and selective than M-MCM-41 for production
of myrtanal. This result confirms that there is no reaction control
by diffusion for the rearrangement of ␤-pinene oxide to myrtanal
with Beta zeolite, and this type of catalyst can be of interest for
preparative chemistry as well as for industrial applications.
4. Conclusions
Solid Lewis-acid sites in metal-containing Beta zeolite have been
identified as active and selective for the epoxide rearrangement
of ␤-pinene oxide to myrtanal. Among Zr-Beta, Sn-Beta, Ti-Beta,
Ta-Beta, Nb-Beta, Al-Beta, and Ga-Beta, the former gives the best
selectivity towards myrtanal (2) with yields up to 96% which is

O. de la Torre et al. / Applied Catalysis A: General 380 (2010) 165–171
171
much higher than any of the values previously reported in the
patent and scientific literature. Materials involving Brönsted acid
sites such as Al-Beta or Ga-Beta are much less suitable materials
since they lack both activity and selectivity. For making better use
of these solid Lewis-acid catalysts the solvent has to be selected
to balance the competitive adsorption of the product. In this way,
favouring product desorption is enhanced which results in a selec-
tivity increase with only a small penalty on the rate of reaction.
Zr-Beta can be re-activated and used several times in a batch
reactor or in a continuous-flow reactor achieving several thousands
of TON (cycles per metal center; on an average). In conclusion, Zr-
Beta is an active, selective and robust catalyst for the synthesis of
myrtanal from ␤-pinene oxide, giving the highest yields reported
up to now.
[3] M. Boronat, A. Corma, M. Renz, P.M. Viruela, Chem. Eur. J. 12 (2006) 7067–
7077.
[4] P. Ingallina, M.G. Clerici, L. Rossi, G. Bellussi, Stud. Surf. Sci. Catal. 60 (1991)
31–39.
[5] M. Boronat, A. Corma, M. Renz, G. Sastre, P.M. Viruela, Chem. Eur. J. 11 (2005)
6905–6915.
[6] C.F. de Graaw, J.A. Peters, H. van Bekkum, J. Huskens, Synthesis (1994)
1007–1017.
[7] H. Yamamoto (Ed.), Lewis Acid Reagents: A Practical Approach, Oxford Univer-
sity Press, New York, 1999.
[8] M. Santelli, J.-M. Pons (Eds.), Lewis Acids and Selectivity in Organic Synthesis,
CRC Press, 1995.
[9] A. Corma, M. Renz, Arkivoc viii (2007) 40–48, http://www.arkat-usa.org/get-
file/19130/.
[10] A. Corma, M. Renz, M. Susarte, Topics Catal. 52 (2009) 1182–1189.
[11] V. Papandreou, P. Magiatis, I. Chinou, E. Kalpoutzakis, A.-L. Skaltsounis, A. Tsar-
bopoulos, J. Ethnopharmocol. 81 (2002) 101–104.
[12] C. Vollmann, W. Schultze, Dtsch. Apoth. Ztg. 135 (1995) 32–44.
[13] A. Kergomard, J. Philibert, Bull. Soc. Chim. France (1959) 1381–1385.
[14] S. Valencia, A. Corma, US Patent 5968473 (assigned to UOP), 1999.
[15] T. Blasco, M.-A. Camblor, A. Corma, P. Esteve, J.M. Guil, A. Martínez, J.A.
Perdigón-Melón, S. Valencia, J. Phys. Chem. B 102 (1998) 75–88.
[16] S.R. Bare, S.D. Kelly, W. Sinkler, J.J. Low, F.S. Modica, S. Valencia, A. Corma, L.T.
Nemeth, J. Am. Chem. Soc. 127 (2005) 12924–12932.
[17] A. Corma, F.X. Llabrés i Xamena, C. Prestipino, M. Renz, S. Valencia, J. Phys.
Chem. C 113 (2009) 11306–11315.
Acknowledgements
This work was supported by the CICYT (MAT-2006-3798164),
Prometeo 2008, Project Consolider-Ingenio 2010 and Fundación
Areces. O.T. is in particular grateful to the Generalitat Valenciana
(programa FPI) and the Spanish Ministry for Science and Innovation
(programa FPU) for doctoral fellowships.
[18] G. Bellussi, M.S. Rigutto, Stud. Surf. Sci. Catal. 85 (1994) 177–213.
[19] G. Ricchiardi, A. Damin, S. Bordiga, C. Lamberti, G. Spano, F. Rivetti, A. Zecchina,
J. Am. Chem. Soc. 123 (2001) 11409–11419.
[20] M. Boronat, P. Concepción, A. Corma, M. Renz, S. Valencia, J. Catal. 234 (2005)
111–118.
References
[1] A. Corma, Catal. Rev. Sci. Eng. 46 (2004) 369–417.
[2] A. Corma, H. Garcia, Chem. Rev. 103 (2003) 4307–4365.