Correlation between the basicity of solid bases and their catalytic activity towards the synthesis of unsymmetrical organic carbonates

Article abstract of DOI:10.1016/j.jcat.2004.10.024

The correlation between the basic properties of solid catalysts and the reaction rate of the selective synthesis of unsymmetrical organic carbonates via direct condensation of diethylcarbonate (DEC) and alcohols was investigated. A detailed kinetic study of the transesterification of 1-phenylethanol and DEC was made with catalysts with different basic strengths. The solids included fluorinated hydrotalcite, MgLa mixed oxides, CsF (pure or supported on α- and γ-aluminas), KF (pure or supported on α- and γ-aluminas), anatase, rutile, and zirconia. Basic properties were determined by the adsorption of CO₂, measured by gravimetry coupled with mass spectrometry and by calorimetry. Both techniques show a higher basicity for MgLa mixed oxides. The rate of the reaction was roughly proportional to the number of strongly basic sites present on the catalyst, except for CsF/α-Al ₂O₃, which showed higher activity for several different substrates in spite of a lower number and strength of basic sites. This is attributed to the high nucleophilicity of the alcoholate formed as an intermediate. The addition of water to the reaction medium induces two effects: inhibition of the reaction and the promotion of etherification of the substrate by ethanol as a side reaction. The catalysts can be reused several times with a small loss of activity.

Full text of DOI:10.1016/j.jcat.2004.10.024

Journal of Catalysis 229 (2005) 237–242
www.elsevier.com/locate/jcat
Correlation between the basicity of solid bases and their catalytic activity
towards the synthesis of unsymmetrical organic carbonates
∗
Bhaskar Veldurthy, Jean-Marc Clacens, François Figueras
Institut de Recherches sur la Catalyse du CNRS, 2 av. A. Einstein, 69626 Villeurbanne, France
Received 28 June 2004; revised 21 October 2004; accepted 25 October 2004
Abstract
The correlation between the basic properties of solid catalysts and the reaction rate of the selective synthesis of unsymmetrical organic
carbonates via direct condensation of diethylcarbonate (DEC) and alcohols was investigated. A detailed kinetic study of the transesterification
of 1-phenylethanol and DEC was made with catalysts with different basic strengths. The solids included fluorinated hydrotalcite, MgLa mixed
oxides, CsF (pure or supported on α- and γ -aluminas), KF (pure or supported on α- and γ -aluminas), anatase, rutile, and zirconia. Basic
properties were determined by the adsorption of CO , measured by gravimetry coupled with mass spectrometry and by calorimetry. Both
2
techniques show a higher basicity for MgLa mixed oxides. The rate of the reaction was roughly proportional to the number of strongly basic
sites present on the catalyst, except for CsF/α-Al O , which showed higher activity for several different substrates in spite of a lower number
2
3
and strength of basic sites. This is attributed to the high nucleophilicity of the alcoholate formed as an intermediate. The addition of water to
the reaction medium induces two effects: inhibition of the reaction and the promotion of etherification of the substrate by ethanol as a side
reaction. The catalysts can be reused several times with a small loss of activity.

2004 Elsevier Inc. All rights reserved.
Keywords: Catalysis by solid bases; Transesterification; Alcohols; Basicity measurements; Kinetics; Unsymmetrical organic carbonates
1
. Introduction
vents [8]. An environmentally benign preparation of alkyl
carbonates from alkyl halides and alcohols in the presence
of alkali metal carbonates or carbon dioxide has recently
been developed [9–11] and paves the way to other synthetic
methodologies. In homogeneous catalysis, unsymmetrical
dialkyl carbonates are produced by transesterification of car-
boxylic acid diesters with an allyl alcohol in the presence of
metal bases (inorganic carbonates, alcoholates, and hydrox-
ides), with crown ethers or polyethylene glycols added as
cocatalysts [12].
We recently reported MgLa mixed oxides [13] and CsF/
α-Al2O3 [14] as new solid bases for the synthesis of the title
compounds. These catalysts showed quite different behav-
iours: CsF/α-Al2O3 was more active and much less sensitive
than MgLa mixed oxides towards the structure of the sub-
strate. Here we attempt to establish a quantitative relation
between the basic properties measured by the gravimetric
A major aim for the chemical technology of the future
is the avoidance of noxious and environmentally unaccept-
able effluents, including salts and organic solvents [1–3]. Al-
though solutions to this problem have been found in the case
of acid catalysis, with more than 100 industrial processes
operated in the world, much less success has been encoun-
tered for basic catalysis, with only a few processes [4]. Het-
erogeneous catalysis is particularly attractive because it al-
lows the ready separation of the products from the catalyst.
A particular example of this is organic carbonates, which
are manufactured on a commercial scale and are used as
intermediates in pharmaceuticals [5], plasticisers, synthetic
lubricants [6], monomers for organic glasses [7], and sol-
*
and calorimetric adsorption of CO and the catalytic activ-
Corresponding author. Fax: (+33) 47244 5399.
2
E-mail address: figueras@catalyse.cnrs.fr (F. Figueras).
ity for the transesterification of 1-phenylethanol and DEC,
0021-9517/$ – see front matter  2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2004.10.024

238
B. Veldurthy et al. / Journal of Catalysis 229 (2005) 237–242
which was chosen as a model reaction in a solvent free sys-
tem.
from the variation in the concentration of alcohol. After
completion of the reaction the reaction mixture was filtered
and the catalyst was washed with DEC (2 × 2.5 ml) to wash
off the product adhering to the surface of the catalyst. The
combined filtrate fractions were concentrated under reduced
pressure to obtain a clean product. Excess DEC removed
from the reaction mixture can be recycled for subsequent
use. The product was analysed by H and C NMR, and
the purity and structure of the product were confirmed by
GC-MS analysis.
2
. Experimental
2
.1. Materials and methods
1
13
The solids included fluorinated hydrotalcite, MgLa mixed
oxides, CsF (pure or supported on α- and γ -aluminas), KF
pure or supported on α- and γ -aluminas), CaF2, anatase,
(
2
rutile, and zirconia. α-Alumina (SPH 512, 10.5 m /g) and
γ -alumina (SCP 350, 400 m /g) from Rhone Poulenc were
2.3. Recycling experiments
2
used as supports. The samples supported by aluminas or
titanias were prepared by aqueous impregnation of the com-
mercial support with 1 mmol/g of fluoride. HDT-F was ob-
tained by anionic exchange of hydrotalcite in the OH form
in an aqueous solution of KF [15,16] and MgLa mixed ox-
ide (Mg/La in the solid = 4.26) by coprecipitation from
a solution of corresponding nitrates [17,18]. CsF/α-Al2O3,
KF/α-Al2O3, HDT-F, and homogeneous analogues KF and
CsF were activated by simple drying at 393 K for 4 h, and
MgLa mixed oxide was calcined at 873 K for 5 h in air, and
the remaining solids were treated under nitrogen at 673 K
for 3–4 h prior to use.
The reuse of the catalyst was investigated with MgLa
mixed oxides and KF and CsF on α-Al2O3, with the use
of 1-phenylethanol and 4-chloro-sec-phenylethanol as sub-
strates. The reaction mixture was taken out with a syringe
equipped with a microfilter after completion of the reaction,
leaving the catalyst in the lowest possible amount of liquid,
and then a new charge of reactants was added.
3
. Results and discussion
3
.1. Characterisation and basic properties of the solids
The preparation of these solid bases has been reported
previously [19,20], and their basicity has been measured in
a qualitative form. A quantitative approach to basic proper-
ties is presented here, which measures the adsorption of CO2
by microcalorimetry at room temperature. The instrument
was a Calvet calorimeter coupled to a volumetric ramp [21].
The samples (0.1 g) were first evacuated at 673 K and then
brought into contact with small doses of gas; the differential
enthalpy of adsorption was then measured. This technique
makes it possible to measure the number of sites of a given
strength, evaluated by the enthalpy of adsorption. The results
were checked by differential thermal gravimetry (DTG) per-
formed with a Setaram DTA-DTG apparatus coupled with a
mass spectrometer, which allowed us to measure the adsorp-
tion of carbon dioxide on the solid.
The surface areas of different catalysts are reported in Ta-
ble 1; the homogeneous analogues (pure KF and pure CsF)
were not evaluated because of their low surface area. Large
differences in surface area are due to the choice of the sup-
port, since most often the acidity of the support changes
Table 1
Catalytic properties of various solid bases and of their homogeneous ana-
logues in direct condensation of sec-phenethyl alcohol with DEC: A model
a
reaction for the evaluation of the best catalyst
Catalyst
Surface
area
Number of
Reaction
time
(h)
Yield
(%)
b
basic sites
2
−1
−1
(m g
)
(µmol g
)
CsF/α-Al O
7
9.4
217
234
–
11.4
14
–
98
47.7
–
0.45
5.0
12
12
12
100
100
57
36
100
2
3
KF/α-Al O
2
3
b
CsF/γ -Al O
2
3
b
2
.2. Catalytic reaction of 1-phenylethanol and DEC
KF/γ -Al O
2
3
HDT-F
KF/TiO (Rutile)
c
6
12
88
2
An excess of DEC (33 mmol) and 1-phenylethanol
c
KF/TiO (Anatase)
44
–
–
–
–
–
–
12
12
12
12
8
2
(
2 mmol) were placed in a three-necked round-bottomed
flask equipped with a condenser. Freshly activated catalyst
0.1 g) was added, and the experiment was started with stir-
c
KF/ZrO
44
2
c
KF/CaF (20 wt%)
12
2
c
(
Pure KF
–
< 5
Pure CsF
Mg La mixed oxide
–
–
14.9
–
12
3.5
12
100
100
NR
ring in a nitrogen atmosphere at ∼ 403 K with the addition
of dodecane (0.05 ml) as an internal standard. Stirring was
continued until the completion of the reaction as monitored
by thin-layer chromatography. The composition of the re-
action medium was analysed by gas chromatography with
a Perkin–Elmer Autosystem equipped with a capillary col-
umn. The reaction selectively produces alkylcarbonate and
ethanol. Because of its low boiling point, the ethanol is dis-
tilled during the reaction. The conversion was determined
37.6
10.5
e
Pure α-Al O
2
3
d
Blank
–
–
12
NR
a
All the reactions performed on 2 mmol alcohol with 33 mmol of DEC
and 0.1 g of catalyst in nitrogen atmosphere at 403 K.
b
c
d
e
1
Yields are calculated with H NMR and GC.
Remaining is the starting material.
In absence of catalyst.
NR represents no reaction.

B. Veldurthy et al. / Journal of Catalysis 229 (2005) 237–242
239
Fig. 1. Differential heats of adsorption of CO on solid bases.
2
with its surface area, which also controls the basic proper-
ties. MgLa mixed oxides, such as the fluorides supported by
α-Al2O3 or titanias, show low surface areas and are non-
porous.
The basic properties of KF supported by α- or γ -alumina
have been reported recently [20], and the calorimetric re-
sults are compared here with those for HDT, HDT-F, CsF/
α-Al2O3, and MgLa mixed oxides (Fig. 1). CsF/α-Al2O3
showed a strong initial enthalpy of adsorption, with only
a few basic sites. The distribution of basic sites of KF/
α-Al2O3 is very close to that of CsF/α-Al2O3, with a
lower initial enthalpy. KF/γ -Al2O3 showed a weaker ba-
sic strength and a larger number of basic sites. This lower
basicity of KF/γ -Al2O3 has been attributed to the loss of
KF due to its reaction with the support with the formation of
fluoroaluminate, as detected by X-ray diffraction and solid-
state 19F-NMR spectroscopy [20]. A similar phenomenon
has been observed for CsF, which partly retains its integrity
on α-Al2O3 but disappears in reaction with the support on
γ -Al2O3 [14]. The high basicity of CsF and KF on α-Al2O3
is then attributed to a good dispersion of the fluoride at the
surface of an inert support, which makes it possible to pre-
serve part of the original fluoride. If hydrotalcite appears as
a weak base in this context, HDT-F is comparable to sup-
ported fluorides, showing a weaker initial heat of adsorption
and a larger number of basic sites.
Fig. 2. Differential thermal gravimetric studies of MgLa and CsF/α-Al O .
2
3
alyst therefore behaves like a pure La oxide, with a higher
thermal stability of the surface area.
The basicity was also estimated by measuring the degree
of carbonation of the solids by differential gravimetry. The
results are reported in Fig. 2 for MgLa and CsF/α-Al2O3,
where the intensities of the signal of the mass spectrome-
ter have been normalised to the mass of sample in order to
make a quantitative comparison. The two solids show sim-
ilar behaviours in DTG, with the occurrence of two steps
involving both dehydration and decarbonation. MgLa and
CsF both exhibit a peak for CO2 above 800 K, but the peak
relative to MgLa is intense and that relative to CsF/α-Al2O3
is very small. These results of DTG analysis of decarbona-
tion therefore confirm the results of calorimetry, pointing to
greater basicity of the MgLa mixed oxide.
The higher basic strength was observed for MgLa mixed
oxides with a lower number of basic sites. Pure La oxide is
known to be a strong base, on which CO2 is adsorbed with
−
1
an initial heat of adsorption of about 140 kJ mol up to a
−1
coverage of 26 µmolg . For higher coverages the heat of
−1
adsorption decreases fast and falls to 100 kJ mol at about
3.2. Reaction kinetics
−1
5
0 µmolg [22]. The MgLa mixed oxide shows an initial
heat of adsorption close to 150 kJ mol , which also de-
creases fast for a coverage above 20 µmolg . The XRD
pattern shows the presence of La oxide [18], and this cat-
−1
We investigated the influence of both the number and
strength of basic sites on the transesterification of DEC by
1-phenylethanol as a model reaction (Scheme 1).
−1

240
B. Veldurthy et al. / Journal of Catalysis 229 (2005) 237–242
Scheme 1. Model reaction for the synthesis of nonsymmetrical organic car-
bonates from an alcohol and DEC.
Table 2
Synthesis of carbonates from alcohols using MgLa mixed oxide, CsF/
α-Al O and KF/α-Al O as catalyst using the standard conditions
2
3
2 3
Alcohol
Reaction time
MgLa
CsF/α-Al O
KF/α-Al O
2
3
2
3
Phenylethanol
6 h
4 h
3.5 h
30 min
20 min
20 min
Sec-Phenethyl alcohol
Ethyl-1-(4-chloro)
phenethyl alcohol
5 h
7 h
(
+) Menthol
45 min
4 h
15 min
30 min
Ethyl tetrahydrofurfuryl
Fig. 3. First order plot of the conversion versus time on different catalysts,
and effect of water on CsF/α-Al O .
alcohol
2
3
A first determination of catalytic properties was made
by measurement of the time required to reach a reasonably
high conversion of 1-phenylethanol, as illustrated in Tables 1
and 2. On this basis it was observed that pure KF was prac-
tically inactive, as were α-Al2O3, KF on CaF2, ZrO2, and
anatase. KF/rutile showed high activity and selectivity to-
wards the desired product, as observed earlier in a Michael
addition reaction [20]. Pure CsF catalyses the reaction but re-
quires 12 h to reach full conversion, whereas CsF/α-Al2O3
achieves this in 45 min, showing a much higher basicity. If
we take into account the number of basic sites of these solid
bases (Table 1), CsF/α-Al2O converts 2 mmol of substrate
to the corresponding product with about 1 µmol of active ba-
sic sites, and then reaches turnovers of 2000; and the MgLa
mixed oxide achieves turnovers of 1300. Therefore it can be
concluded that the reaction is catalytic.
Fig. 4. Correlation of the reaction rate with the total number of active sites.
It appeared at this stage that MgLa mixed oxides with
a higher number of basic sites of higher strength showed
poorer catalytic properties than CsF/α-Al2O3 of lower basic-
ity. This behaviour is not particular to one reaction but was
confirmed on several substrates (Table 2) that reacted very
fast on this catalyst: a quantitative conversion was observed
typically in less than 30 min on CsF/α-Al2O3, whereas the
same result required hours with the MgLa mixed oxide. Sim-
ilarly, the small differences in basicity between CsF and KF
on α-alumina are not consistent with the large differences
in reaction time. Therefore, a kinetic study was done with
selected catalysts to investigate the correlation between the
reaction rate and the basic properties. The kinetic data for the
reaction in excess of DEC (ratio alcohol/DEC = 15) over
most catalysts could be fitted well to a pseudo-first-orderrate
law:
plot over a large range of alcohol conversions, as illustrated
in Fig. 3 for catalysts showing quite different activities. The
reaction rate constants k obtained from these results were
compared with the basic properties of the corresponding cat-
alysts (Fig. 4). Indeed, the catalytic activity is not correlated
with the total number of basic sites. This can be accounted
for by the fact that only those sites able to abstract a pro-
ton from the alcohol can play a catalytic role. The results
of Table 1 show a low activity for KF/γ -alumina, which has
−1
an initial energy of adsorption of about 120 kJ mol , sug-
gesting that the minimum basic strength required for the ab-
straction of a proton from 1-phenylethanol corresponds to an
−1
enthalpy of CO adsorption higher than 120 kJ mol . The
2
results were then replotted in Fig. 5, under the assumption
−1
that the active sites are those for which ꢀH > 120 kJ mol
.
A rough correlation does appear between the number of
these sites and the reaction rate, but CsF/α-Al2O3 falls well
away from of the line, which merits a special discussion.
Because of the higher acidity of alcohols compared with
DEC, the base-catalysed reaction most probably takes place
log(1/(1 − x)) = (ka/2.303)t,
where ka is the apparent first-order rate constant, x is the
fractional conversion of alcohol, and t is the reaction time.
A plot of log(1/(1 − x)) as a function of time gives a linear

B. Veldurthy et al. / Journal of Catalysis 229 (2005) 237–242
241
Table 3
Synthesis of carbonates from diols using (a) MgLa mixed oxides and (b)
CsF/α-Al O as catalysts
2
3
a
Entry
Alcohol
Reaction
time
Product
Yield
(%)
b
1
(a) 2 h
(a) 98
(b) 98
(b) 25 min
2
3
(a) 1 h
b) 20 min
(a) 97
(b) 97
(
(a) 45 min
b) 45 min
(a) 97
(b) 98
(
Fig. 5. Correlation between the reaction rate constant and the number of
−1
strong basic sites, adsorbing CO with an enthalpy > 120 kJ mol
.
2
4
(a) 45 min
b) 30 min
(a) 98
(b) 97
(
by proton abstraction by a negatively charged fluoride (or
oxide) moiety to generate an alkoxide anion stabilised at the
surface cation. The alkoxides conjugated with cesium are
believed to constitute “naked anions” exhibiting enhanced
nucleophilicity [5,23], and this mechanism accounts for the
high activity of cesium fluoride and the low rate sensitiv-
ity towards the structure of the alcohol. In contrast, MgLa
mixed oxides, even though they have comparable basicity,
were found to show a rather strong sensitivity to the structure
of the substrate. Weakly coordinated species should be more
negatively charged, then reactive, and less sensitive to steric
factors than an intermediate strongly bonded at the surface.
Cs shows a Pauling radius of 1.69 Å compared with 1.33 Å
for K, then a lower density of charge, accounting for the par-
ticular behaviour of this cation.
a
All the reactions performed on 2 mmol alcohol with 33 mmol of DEC
and 0.1 g of catalyst in nitrogen atmosphere at 403 K.
b
Isolated yields.
Table 4
Reuse of various catalysts with two different substrates
a
Alcohol
Experiment
number
Reaction
time (h)
Yield
(%)
b
CsF/α-Al O as catalyst
1-Phenylethanol
2
3
1
2
3
4
0.33
0.5
1.5
99
98
82
14
12
4
-Chloro-1-phenylethanol
1
3
4
0.41
1.0
12
98
97
21
This low sensitivity of CsF/α-Al2O3 is particularly in-
teresting for the fast and selective transesterification of
1
,2-diols into cyclic carbonates (Table 3, entries 1 and 2),
MgLa mixed oxides as catalyst
-Chloro-1-phenylethanol
which are usually prepared from phosgene and carbodiimi-
dazole [24] or bis(trichloromethyl) carbonate [25], whereas
4
1
2
5
3.5
4
7
98
97
98
2
,5-hexanediol and 1,4-cyclohexanediol yielded the corre-
sponding bis-carbonate as the sole reaction product (Table 3,
entries 3 and 4). Therefore, this methodology permits us
to offer a simple and efficient alternative to the existing
methodologies. Another interesting observation was that the
selectivity did not change with the conversion: according to
the literature [5], in basic media the cyclic carbonates can
undergo ring opening to give a mono-protected alcohol; this
was not observed in our reaction conditions. The high se-
lectivity towards the cyclic carbonate can be attributed to a
lower rate of the secondary reaction or to the use of carbon-
ate as solvent.
KF/α-Al O
2
3
1-Phenylethanol
1
2
3
4
5
7.5
12
12
99
98
92
16
a
All the reactions performed on 2 mmol alcohol with 33 mmol of DEC
in nitrogen atmosphere at 403 K.
b
Isolated yields.
general rule a lower activity is noticed in the second cy-
cle, and the time required to reach complete conversion in-
creases. From a practical point of view, however, a catalyst
reaching a 97% yield in a slightly longer time can still be
considered useful. With this criterion the catalyst can be re-
cycled two times in the case of KF/α-Al2O3 and three times
with CsF/α-Al2O3, with a longer reaction time required for
the third cycle (Table 4). A chemical analysis of the cata-
3
.3. Recycling experiments
The reuse of the catalyst was investigated with MgLa
mixed oxides and KF and CsF on α-Al2O3, with 1-phenyl-
ethanol and 4-chloro-sec-phenylethanol as substrates. As a

242
B. Veldurthy et al. / Journal of Catalysis 229 (2005) 237–242
lysts used (after four cycles) revealed no loss of Cs, K, or F,
and thus that the lower activity cannot be due to leaching.
With MgLa mixed oxides, the situation is similar, with little
change of activity after four recycles and a clear loss of ac-
tivity after the fifth recycling only. The longer reaction time
can be attributed to two main factors: some loss of catalyst
remaining in suspension in the reaction mixture and inhibi-
tion by traces of water present in each new feed of reactants.
4. Conclusions
In conclusion, a quantitative relation does exist between
the catalytic activity and the basic properties of the catalysts
employed for the transesterification reaction. MgLa mixed
oxides and CsF/α-Al2O3 were the best catalysts and could
be recycled a few times. The high activity observed with
CsF/α-Al2O3 may be due to the higher nucleophilicity of
the alcoholate formed as an intermediate, leading to a faster
rate.
3
.4. Inhibition effect of water on reaction kinetics
This hypothesis was checked by measurement of the ki-
netics in the presence of water (added initially at the level
−1
of 1 mmolmmol of alcohol). It was observed that water
inhibits but does not completely suppress the reaction: the
rate is very low for the reaction catalysed by CsF/α-Al2O3
Acknowledgments
(
Fig. 3), showing kinetic inhibition, but the yield after 1 h
Thanks are due to the technical services of IRC, espe-
cially for chemical and GC-MS analysis.
reached 57% with high selectivity. Up to this time, the ki-
netics was still first order and corresponded to a much lower
rate constant. After this time, the ether corresponding to the
alcohol, not detected in the absence of water, appeared as a
by-product in large quantities (about 30%). With the MgLa
mixed oxides, the same phenomenon was observed, with a
yield reaching 52% after 2 h, followed by a slow rate with
the appearance of ether as a by-product (20%). The forma-
tion of ether by reaction of diethylcarbonate with a diol,
References
[1] L. Krumenacker, S. Ratton, Actualite Chimique (1986) 29.
[2] W.F. Holderich, Stud. Surf. Sci. Catal. 75 (1993) 127.
[3] J.M. Thomas, R. Raja, G. Sankar, B.F. Johnson, D.W. Lewis, Chem.
Eur. J. 7 (2001) 2973.
[4] K. Tanabe, W.F. Holderich, Appl. Catal. A 181 (1999) 399.
[5] J.P. Parrish, R.N. Salvatore, K.W. Jung, Tetrahedron 56 (2000) 8207.
[6] S. Gryglewicz, F.A. Oko, G. Gryglewicz, Ind. Eng. Chem. Res. 42
(2003) 5007.
1
,4:3,6-dianhydrosorbitol, has been reported to be catalysed
at 523 K by sodium methoxide or organic bases such as
diazobicyclo[2,2,2]octane (DABCO) [26]. It was also ob-
served earlier that in the absence of water in the reaction
medium, esters were readily hydrolysed on hydrated hydro-
talcites at about 323 K [27], a phenomenon not observed on
dehydrated solids and therefore attributed to the presence of
hydroxyls at the surface. Indeed, CsF/α-Al2O3 is dehydrated
in vacuum at about 400 K (Fig. 2) and readily rehydrated in
air, and faster in water. A simple mechanism for the forma-
tion of ether by a base-catalysed reaction could be based on
the following scheme, involving the hydrolysis of DEC, fol-
lowed by decarboxylation of the anion to an alkoxide, which
can react on the alcoholic substrate to form an ether:
[7] J.L.R. Williams, K.R. Dunham, US 2843567, 1958.
[
[
8] F. Mizia, F. Rivetti, US 20020056468, 2002.
9] Y. Ono, Appl. Catal. A: General 155 (1997) 133.
[10] D. Delledonne, F. Rivetti, U. Romano, Appl. Catal. A: General 221
(2001) 241.
[
[
11] P. Tundo, M. Selva, Acc. Chem. Res. 35 (2002) 706.
12] Y. Proux, M. Pellegrina, Fr 2608812, 1988.
[13] B. Veldurthy, F. Figueras, Chem. Commun. 10 (2004) 734.
[14] J.-M. Clacens, D. Genuit, B. Veldurthy, G. Bergeret, L. Delmotte,
A. Garcia-Ruiz, F. Figueras, Applied Catalysis, B: Environmental 53
(2004) 95.
[15] B.M. Choudary, M. Lakshmi Kantam, V. Neeraja, K. Koteswara Rao,
F. Figueras, L. Delmotte, Green Chem. 3 (2001) 257.
[
16] F. Figueras, B. M. Choudary, M. Lakshmi Kantam, V. Neeraja, K. Ko-
teswara Rao, WO 0166246, 2001.
−
−
EtO–CO–OEt + OH → EtOH + EtOCOO ,
−
−
EtOCOO → CO2 + EtO ,
EtO + ROH → EtOR + OH .
[17] J. Palomeque, J.-M. Clacens, F. Figueras, J. Catal. 211 (2002) 103.
[
[
[
18] F. Figueras, H. Kochkar, L.K. Mannepalli, Fr 2834228, 2003.
19] J. Palomeque, J. Lopez, F. Figueras, J. Catal. 211 (2002) 150.
20] J.-M. Clacens, D. Genuit, L. Delmotte, A. Garcia-Ruiz, G. Bergeret,
R. Montiel, J. Lopez, F. Figueras, J. Catal. 221 (2004) 483.
21] P.C. Gravelle, Adv. Catal. 22 (1972) 191.
−
−
Scheme 2. Possible mechanism for ether formation.
[
[
[
[
22] A. Auroux, A. Gervasini, J. Phys. Chem. 94 (1990) 6371.
23] F. Chu, E.E. Dueno, K.W. Jung, Tetrahedron Lett. 40 (1999) 1847.
24] S.K. Kang, J.H. Jeon, K.S. Nam, C.H. Park, H.W. Lee, Synth. Com-
mun. 24 (1994) 305.
The continuous decrease in activity for transesterification
upon recycling can then be accounted for by, on one hand,
the kinetic effect of the water contained in the reactants, and,
on the other hand, by the development of the side reaction of
hydrolysis of DEC and etherification of the alcoholic sub-
strate.
[25] R.M. Burk, M.B. Roof, Tetrahedron Lett. 34 (1993) 395.
[26] J.N. Greenshields, Application: US 4770871, 1988.
[27] J. Lopez, F. Figueras, unpublished results.