Decarboxylation of dialkyl carbonates to dialkyl ethers over alkali metal-exchanged faujasites

Article abstract of DOI:10.1039/c0gc00536c

Non-toxic DAlCs, especially lighter dimethyl- and diethyl-carbonate, are regarded as very green alkylating reagents, particularly when coupled with metal-exchanged Y- and X-faujasites as catalysts. These reactions are selective, free from wastes or byproducts, and often require no additional solvent other than the carbonate. Nonetheless, this paper demonstrates that the operating temperature and the nature of the faujasite must be carefully chosen in order to avoid DAlC decomposition. In fact, at temperatures ranging from 150 to 240°C, faujasites can promote decarboxylation of light DAlCs to the corresponding ethers CH₃OCH₃ and CH₃CH ₂OCH₂CH₃ plus CO₂. Heavier DAlCs (dipropyl- and dioctyl-carbonate) undergo a similar decomposition pathway, followed by further reactions to the corresponding alcohols (n-propanol and n-octanol) and alkenes [propylene and octene(s)]. These transformations not only consume DAlCs, but also give rise to dangerously flammable ethers, as well as undesirable alcohols, alkenes and CO₂. The present work reports an original investigation of the decarboxylation of DAlCs on faujasites with the aim of providing operative boundaries to the experimental conditions to minimise unwanted decomposition. The reaction is strongly affected by the nature of the catalyst: the more basic zeolites, NaX and CsY, are by far more active systems than NaY and LiY. However, solid K₂CO₃ proves to be rather inefficient. The temperature also plays a crucial role: for example, the onset of the decarboxylation of DMC requires a temperature of ～30°C lower than that for DEC and DPrC. Overall, awareness that certain zeolites cause decomposition of DAlCs under conditions similar to the ones used for DAlC-promoted alkylations allows determination of the correct experimental boundaries for a safer and more productive use of DAlCs as alkylating agents. The Royal Society of Chemistry.

Full text of DOI:10.1039/c0gc00536c

Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
Volume 13 | Number 4 | April 2011 | Pages 725–1040
ISSN 1463-9262
COVER ARTICLE
Selva et al.
Decarboxylation of dialkyl
carbonates to dialkyl ethers over
alkali metal-exchanged faujasites
1463-9262(2011)13:4;1-V

Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 863
www.rsc.org/greenchem
PAPER
Decarboxylation of dialkyl carbonates to dialkyl ethers over alkali
metal-exchanged faujasites†
Maurizio Selva,* Massimo Fabris and Alvise Perosa
Received 10th September 2010, Accepted 8th December 2010
DOI: 10.1039/c0gc00536c
Non-toxic DAlCs, especially lighter dimethyl- and diethyl-carbonate, are regarded as very green
alkylating reagents, particularly when coupled with metal-exchanged Y- and X-faujasites as
catalysts. These reactions are selective, free from wastes or byproducts, and often require no
additional solvent other than the carbonate. Nonetheless, this paper demonstrates that the
operating temperature and the nature of the faujasite must be carefully chosen in order to avoid
◦
DAlC decomposition. In fact, at temperatures ranging from 150 to 240 C, faujasites can promote
decarboxylation of light DAlCs to the corresponding ethers CH
3
OCH
3
and CH
3
CH
2
OCH
2
CH
3
plus CO . Heavier DAlCs (dipropyl- and dioctyl-carbonate) undergo a similar decomposition
2
pathway, followed by further reactions to the corresponding alcohols (n-propanol and n-octanol)
and alkenes [propylene and octene(s)]. These transformations not only consume DAlCs, but also
give rise to dangerously flammable ethers, as well as undesirable alcohols, alkenes and CO . The
2
present work reports an original investigation of the decarboxylation of DAlCs on faujasites with
the aim of providing operative boundaries to the experimental conditions to minimise unwanted
decomposition. The reaction is strongly affected by the nature of the catalyst: the more basic
zeolites, NaX and CsY, are by far more active systems than NaY and LiY. However, solid K
2
CO
3
proves to be rather inefficient. The temperature also plays a crucial role: for example, the onset of
◦
the decarboxylation of DMC requires a temperature of ~30 C lower than that for DEC and
DPrC. Overall, awareness that certain zeolites cause decomposition of DAlCs under conditions
similar to the ones used for DAlC-promoted alkylations allows determination of the correct
experimental boundaries for a safer and more productive use of DAlCs as alkylating agents.
Introduction
the combination of DAlCs with a class of zeolites, namely
alkali metal exchanged faujasites as catalysts, has proven to
4
Light dialkyl carbonates (ROCO
cially dimethyl carbonate (DMC), are excellent green substitutes
of toxic reagents and solvents. Thanks to their dual electrophilic
reactivity at both the carbonyl and the alkyl carbons, they
find major applications in organic synthesis as safe building
blocks for the replacement of phosgene and alkyl halides in
carboxyalkylation and alkylation processes. In addition, their
solvency properties – recently reviewed by B o¨ rner et al. – clearly
2
R, DAlCs, R = C
1
–C ), espe-
3
be successful to improve the outcome of several O- and N-
5
alkylation reactions of different nucleophiles. As an example,
1
Scheme 1 details some DMC-mediated methylations carried
out in the presence of NaY or NaX faujasites. Two significant
aspects emerge: (i) The use of zeolites steers the reaction towards
an exclusive methylation process, for example eqn (a) and
1,2
(b) show that both anilines and benzyl-type alcohols, even
3
when deactivated by steric or electronic effects, yield only the
corresponding N- and O-methyl derivatives.
An added value for amines is the very high selectivity
indicate that DAlCs should always be taken into consideration
as alternative media for new transformations. In this context,
(
up to 98% at complete conversion) towards mono-N-methyl
anilines, which cannot be otherwise achieved using conventional
alkylation methods. (ii) Faujasite catalysts allow an excellent
chemoselectivity in the methylation of ambident nucleophiles.
Scheme 1 (left) exemplifies the case of aminophenols, aminoben-
zoic acids, hydroxybenzyl alcohols, and hydroxythiophenols
Dipartimento di Scienze Molecolari e Nanosistemi dell’Universit a` Ca’
Foscari Venezia, Calle Larga S. Marta, 2137–30123, Venezia, Italy.
E-mail: selva@unive.it
†
Electronic supplementary information (ESI) available: (i) A synopsis
of MS signals for major gaseous and liquid products, including
DME, CO , propylene, diethyl ether, diproyl ether, isomeric octenes,
and glycidol; (ii) an IR spectrum for a reaction mixture from the
2
[
eqn (c)–(f), respectively], whose reactions with DMC proceed
via a complete discrimination between the two nucleophilic
2
decarboxylation of DMC, including dimethyl ether, CO , and DMC. See
DOI: 10.1039/c0gc00536c
terms of the substrates: methylation reactions take place at
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 863–872 | 863

Table 1 Faujasite catalysts used in the reaction of DMC
a
b
Starting zeolite (Na %)
Product
Ionic exchange (%)
NaY (8)
NaY (8)
LiY
CsY
67
56
a
b
The Na content was evaluated through atomic adsorption (AA). The
percentage of ionic exchange (from NaY) was evaluated by atomic
adsorption (Na) and emission (Li and Cs).
It should be noted that the decarboxylation of organic
carbonates is documented in the literature: for example, Pd-
complexes catalyze the transformation of allyl alkyl carbonates
Scheme 1 DMC-mediated methylations in the presence of faujasite
catalysts.
6
to unsaturated esters and ketones, thermal treatment and/or
basic/radical initiators induce the loss of CO
2
from carbonate
polymers and alkyl aryl carbonates to produce mixtures of
1
a,7
the amine, alcohol, or thiol groups, while the aromatic OH
substituents as well as the acid functions are fully preserved from
possible alkylation or transesterification processes. Although
these transformations are relatively energy intensive (they usu-
ally occur at 150–200 C), DMC acts not only as a safe alkylating
agent, but also as a reaction solvent, and no by products except
olefins, alcohols and polyether diols, hydrotalcite or alumina
catalysts allow the synthesis of dialkyl ethers from heavy (C10
21) dialkyl carbonates. However, to the best of our knowledge,
–
8
C
an investigation of the behavior of light dialkyl carbonates
with respect to the decarboxylation reaction has never been
reported particularly in the presence of zeolite systems. This
paper clearly demonstrates that such a reaction can proceed
extensively when dimethyl-, diethyl-, and dipropyl-carbonates
(DMC, DEC, and DPrC, respectively) are the substrates and
alkali metal-exchanged Y and X faujasites (NaY and NaX) are
the catalysts. The corresponding ethers have been observed in
all cases though the onset of the decarboxylation is affected
by the structure of carbonate and the nature of the zeolites.
For a more complete comparison, also different catalysts such
◦
for MeOH and CO are obtained.
2
The investigation of these reactions indicates that the for-
mation of the desired alkylation products may take place
through two concurrent mechanisms. This is especially true
for alcohols whose etherification with DAlCs over faujasites,
proceeds through a sequence of tetrahedral and S 2-type
N
5d
processes followed by a decarboxylation reaction. Scheme 2
describes the reaction of benzyl alcohol with DMC.
as hydrotalcite and K
2
CO , and reagents such as dioctyl-
3
and glycerol-carbonates have been examined. A mechanistic
hypotheses is formulated on the basis of the adsorption of
DMC over the NaY surface. Overall, this systematic study of the
experimental conditions for the decarboxylation of light DAlCs
provides accurate boundaries to circumvent the reaction, and as
a consequence, to improve substantially the outcome of DAlCs-
mediated alkylations.
Scheme 2
Results
Initially, benzyl methyl carbonate (BMC) and benzyl methyl
ether (BME) originate from the attack of benzyl alcohol at
both the carbonyl and methyl carbons of DMC [paths a and
b, respectively]. Then, if the mixture is allowed to react further,
BMC undergoes a quantitative decarboxylation and BME is
obtained as a sole product in isolated yields up to 96%.
Catalysts
Four different faujasites, namely MY (M = Li, Na, Cs) and
NaX were used as catalysts. Two of them, NaY and NaX, were
commercially available samples. LiY and CsY were prepared
through conventional ionic exchange reactions, starting from
NaY and LiCl or CsCl as precursors. The metal content of
the catalysts was determined by atomic adsorption (Na) and
An intriguing question arises from this example: if reaction
9
conditions allow BMC to lose CO , then would the decarboxy-
2
lation of DMC be possible as well? If so, this raises a rather
detrimental issue: an undesired consumption of the alkylating
agent/solvent (DMC) would take place along with the co-
10
emission (Li, Cs), performed according to known procedures
(Table 1). Hydrotalcite (HT: KW2000, Mg0.7Al0.3
O1.15) and
K
2
CO were also from commercial suppliers.
3
generation of CO in addition to that formed by the alkylation
2
reaction (Scheme 2, path b). Moreover, this behavior would also
open a more general window on the scope and limitations of the
combined use of dialkyl carbonates and faujasites; in fact, higher
homologues such as DEC and DPrC, would be expected to react
similarly to DMC. One concludes that for a more productive use
of light DAlCs as alkylating agents, their decarboxylation – if
occurring – should be avoided or minimized as much as possible.
Reaction of dimethyl carbonate over faujasites, hydrotalcite, and
CO
K
2
3
DMC was used as a model substrate to begin the investigation.
Before each reaction, the catalysts (faujasites, hydrotalcite, and
◦
K
2
CO
3
) were dehydrated under vacuum (8 mbar) at 70
C
overnight. Initial experiments were carried out in a 120 mL
8
64 | Green Chem., 2011, 13, 863–872
This journal is © The Royal Society of Chemistry 2011

autoclave charged with a mixture of DMC (10 mL, 119 mmol)
and the faujasite (0.6 g of NaY or NaX; the weight ratio,
Q, catalyst : DMC was 0.056), and heated under stirring for
Fig. 1 indicated that NaY faujasite allowed sufficiently slow
reactions with well distributed conversions (6–75%) over a broad
range of temperature (150–220 C). Therefore, the results with
◦
◦
11
6
h at different temperatures of 150–220 C. Under these
NaY seemed a convenient choice to compare the activity of
different catalysts. The investigation was extended to Li- and
12
conditions, hydrotalcite (HT: 0.6 g) was also tested as a catalyst.
14
Once cooled to rt, a residual pressure up to 20 bar was
measured inside the reactor. In order to recover the gaseous
products of the reaction, a dedicated experimental setup was
arranged: the autoclave was purged by a low temperature trap-
to-trap apparatus (see Experimental for further details). The
gaseous mixture was collected in an air-tight Drechsel bottle
Cs-exchanged Y faujasites and K
2
CO . Under the conditions
3
above described (DMC: 10 mL; weight ratio Q = cat/DMC =
◦
0.056), reactions were performed at 200 C for 6 h. An additional
experiment was carried out in the absence of any catalyst. The
results are reported in Fig. 2, where for a more complete view
the cases of NaY, NaX and HT are also included.
(
200 mL) and weighed. Both the gas phase and residual liquid
in the autoclave were analysed by GC-MS. In the first one
gas), only two products, CO and dimethyl ether (DME),
(
2
13
were observed. This unambiguously proved that the DMC
underwent a decarboxylation reaction (Scheme 3).
Scheme 3
The progress of such a reaction was evaluated from the total
(
weighed) amounts of CO and DME: this allowed to calculate
2
the conversion of DMC in each experiment. The results are
reported in Fig. 1.
◦
Fig. 2 Decarboxylation of DMC over different catalysts at 200 C.
DMC by itself was rather stable to the decarboxylation. The
reaction was observed only in the presence of a catalyst, and in
all experiments CO and dimethyl ether were the sole detected
2
products (Scheme 3). However, the outcome was greatly affected
by the nature of the added solid: the conversion of DMC was
very modest (4%) on K
2
CO , while it progressively raised from
3
2
6% to 94% with the increase of the basic character of Y-
faujasites, i.e. from LiY to NaY and to CsY, respectively. This
latter (CsY) offered an activity comparable to that of NaX and
HT.
Importantly, results of both Fig. 1 and 2 also provided an in-
dicative threshold temperature, below that the decarboxylation
of DMC was negligible over different catalysts. This marked
operative limits where DMC can be safely and efficiently used
as an alkylating agent without significant decomposition and
Fig. 1 Decarboxylation of DMC over sodium faujasites (NaY and
without formation of CO .
2
NaX) and hydrotalcite (HT) at different temperatures and after 6 h.
Effect of the faujasite amount and of the reaction time on the
decarboxylation of DMC
Although the three catalysts were all effective for the decar-
boxylation of DMC, their performance was quite different. For
example, at 180 C, the reaction conversion was of 80%, 54%,
◦
Commercial NaY and NaX catalysts were chosen to continue
the investigation. Two sets of experiments [(a) and (b)] were
and 7%, over NaX, HT, and NaY, respectively. In the same
◦
◦
order, at 200 C, the unreacted DMC was 4%, 8%, and 31%.
carried out: (a) at 200 C and for 6 h, DMC (10 mL, 119 mmol)
These results indicated a preliminary trend of catalytic activity:
NaX > HT > NaY.
was set to react with increasing quantities of both NaY and
NaX, by varying the weight ratio Q = cat : DMC from 0.015
◦
The GC-MS analysis of the residual DMC in the autoclave
showed traces of methanol whose amount did not exceed 3%
even at high conversions (up to 90%). The presence of the alcohol
was ascribed to hydrolysis of the starting carbonate, due to water
to 0.084; (b) at 200 C, a mixture of DMC (10 mL, 119 mmol)
and the catalyst (NaY or NaX; Q ratio = 0.056), was set to react
at different times, from 1 to 18 h, respectively. The results are
shown in Fig. 3(a) and (b), respectively. Both sets of reactions
confirmed that NaX was by far, a more active catalyst than NaY.
8
adsorbed on the catalysts.
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 863–872 | 865

Competitive methylation and decarboxylation reactions over
NaX
Fig. 1–3 and Table 1 proved that extensive decarboxylation of
DMC took place under the same conditions used to carry out
several methylation reactions mediated by DMC and catalyzed
by faujasites (Scheme 2). This led to the plausible hypothesis
that dimethyl carbonate could exhibit its methylating activity
while, simultaneously, it underwent a decarboxylation reaction.
An experiment was performed to verify such an aspect. A
mixture of benzyl alcohol (5.0 mL, 48.4 mmol), DMC (15.0 mL,
1
78.2 mmol), and NaX (0.52 g; Q = cat : DMC = 0.032)
◦
was set to react at 200 C, in a 120 mL autoclave, for 12 h.
Once cooled to rt, the reactor was purged according to the
procedure above mentioned (Fig. 1): a mixture of gaseous
products (13.3 g) composed of CO and DME, was recovered.
2
The GC-MS analysis of the residual slurry content of the
reactor revealed that the starting benzyl alcohol was totally
15
converted as shown in Scheme 2. These results confirmed the
competitive occurrence of the decarboxylation of DMC and the
O-methylation of the alcohol: the two processes consumed ~70%
and 25%, respectively, of the initial DMC amount.
Reaction of diethyl- and dipropyl-carbonates over Na-faujasites
The reactions of higher homologues of DMC such as diethyl-
and dipropyl-carbonates (DEC and DPrC, respectively), were
investigated in the presence of NaX and NaY faujasites. Before
each run, both catalysts were carefully dehydrated under vacuum
◦
(
8 mbar) at 70 C, overnight. Experiments were carried out in
the same 120 mL autoclave used for DMC.
In the case of DEC, a mixture of the carbonate (10 mL,
8
3 mmol) and the faujasite (0.55 g of NaY or NaX; the weight
◦
Fig. 3 The decarboxylation of DMC carried out at 200 C, over NaY
ratio, Q, catalyst : DEC was 0.056), was set to react for 6 h at
different temperatures of 180–240 C. Once cooled to rt, the
reactor was dipped in a cooling bath at -55 C, and vented. A
and NaX. (a) effect of the catalyst amount; (b) effect of the reaction time
◦
(
Q = 0.056).
◦
GC-MS analysis of the released gas showed the sole presence of
Fig. 3a [top; set (a)] proved that the decarboxylation of DMC
CO . The autoclave was allowed to return gradually to ambient
2
was a truly catalytic process affected by the zeolite loading: the
reaction conversion smoothly increased from 7 to ~50% as the
NaY amount was augmented by a factor of 6 (Q = 0.084). By
contrast, even with a modest Q ratio of 0.028, the use of NaX
allowed a substantially quantitative conversion.
temperature, and the reaction mixture was analyzed by GC-
MS. Diethyl ether (DEE) was observed as the main product
(
90–93%), confirming that also the decarboxylation of DEC
took place over Na-faujasites (Scheme 4).
Fig. 3b [bottom; set (b)] showed the effect of the reaction time:
at a Q ratio of 0.056, the decarboxylation process was complete
in ~3 h in the presence of NaX (gray profile), while even after
1
8 h, 20% of the initial DMC was still unreacted when NaY was
the catalyst (black profile).
The conversion of DMC was always determined by
Scheme 4
the weighed amounts of the reaction products (CO
DME).
2
and
Minor amounts (3–5%) of ethanol were detected as a by-
product. Fig. 4 reports the conversion of DEC and the selectivity
towards the formation of DEE (both parameters evaluated by
GC-MS) vs. the reaction temperature.
In order to test whether the thermal activation of faujasites
affected their performance, two aliquots of NaY were heated in
-
1
◦
a stream of dry air (10 mL min ), at 200 and 400 C, respectively,
for 4 h. Then, these samples were used as catalysts to carry out
Two aspects emerged: (i) the performance of the two faujasites
resembled that shown for DMC. At comparable conditions, the
DEC conversion was always higher over NaX with respect to
NaY (black profiles); (ii) the onset of the decarboxylation of
◦
two additional reactions with DMC at 200 C, for 6 h. The
weight ratio Q = NaY:DMC was set to 0.056. In both cases, the
reaction conversion (~30%) did not appreciably differ from that
◦
◦
evaluated when the catalyst was simply dried at 70 C (compare
DEC required a temperature ~30 C higher than that for DMC
Fig. 1, 2, and 3a).
(compare Fig. 1). In other words, diethyl carbonate appeared
8
66 | Green Chem., 2011, 13, 863–872
This journal is © The Royal Society of Chemistry 2011

the formation of both propanol and propylene was instead, the
18
cracking of DPE according to Scheme 6:
Scheme 6
This hypothesis was substantiated by the following consid-
erations. The full decarboxylation of DPrC corresponded to
the release of 56 mmol (2.45 g) of CO which, from the total
2
weight of gaseous products, allowed to calculate the formation
of 30 mmol (1.25 g) of propylene. The stoichiometry of Scheme 6
indicated that the same molar amount (30 mmol) of n-propanol
had to be present in the liquid products, thereby implying a
residual (liquid) DPE quantity of 26 mmol. Accordingly, a
propanol:DPE ratio of 65 : 35 was obtained in good agreement
with the GC ratio (62 : 38) observed for these two compounds.
Fig. 4 Decarboxylation of DEC over NaY and NaX faujasites.
less prone than DMC to the loss of CO . The selectivity to DEE
was, on average, very high (>90%, gray profiles); though, as for
DMC, a minor hydrolysis of DEC due to water adsorbed on the
faujasites yielded ethanol as a co-product.
2
Other dialkyl carbonates
Dipropyl carbonate (DPrC) was prepared by adjusting a gen-
eral procedure reported for the synthesis of dialkyl carbonates.
To verify whether the decarboxylation over Na-faujasites held
true also for high molecular weight and/or viscous organic
16
The reaction of DPrC was carried out using the procedure
above described for DEC, only in the presence of the more
active NaX as the catalyst: the carbonate (8.8 mL, 56 mmol)
and the faujasite (0.3 g of NaX; the weight ratio, Q, catalyst :
DPrC was 0.036), was set to react at temperatures of 220
carbonates, two model compounds such as dioctyl carbonate
19
(
DOC) and glycerine carbonate (GlyC) were chosen. The
reactions of DOC (2.0 g, 10.8 mmol) and GlyC (2.0 g, 17.0 mmol)
◦
were explored at 200 C for 6 h, in the presence of NaX faujasite
(the weight ratio, Q, catalyst : carbonate, was 0.06). In both
◦
and 240 C for 6 and 12 h, respectively. The GC-MS analysis
cases, the viscosity of the reagents required the use of a solvent:
cyclohexane (9 mL) and dimethoxyethane (9 mL) were used
for DOC and GlyC, respectively. Experiments were carried out
according to the procedure detailed for DEC.
of the reaction products showed that the outcome of these
experiments was remarkably different from that of DMC and
DEC. DPrC underwent the expected decarboxylation reaction,
though the process was no longer selective on formation of the
corresponding ether. Regardless of the conversion of DPrC,
In the case of DOC, the GC-MS analysis of the reaction
mixture showed the presence of unreacted carbonate (42%),
dioctyl ether (7%) and comparable amounts of n-octanol and
the gas vented from the autoclave was a mixture of CO and
2
propylene, while liquid products were n-propanol and di-n-
propyl ether (DPE) (Scheme 5).
20
isomeric octenes (28 and 23%, respectively, ) (Scheme 7).
Scheme 7
Scheme 5
The good analogy with the behavior of DPrC (Scheme 5)
indicated that also the decarboxylation of DOC took place along
with the plausible decomposition of the corresponding dioctyl
ether to octanol and octenes. The reaction of GlyC instead, did
not offer any clear interpretation: a complex mixture of several
products was observed, most of them remained unidentified
compounds. However, glycerol and trace amounts (5%) of
glycidol were detected (Scheme 8).
◦
At 220 C, after 6 h, the liquid phase contained unreacted
DPrC (61%) along with comparable amounts of n-PrOH and
DPE (22 and 17%, respectively). The gas phase (2.0 g) was
17
composed by CO and propylene. Both propylene and n-PrOH
2
formed even at a moderate conversion of the starting carbonate.
At 240 C, the conversion of DPrC was complete after
◦
1
2 h: a gaseous mixture (3.70 g) of propylene and CO , and
2
a liquid mixture of n-propanol (62%) and DPE (38%) were
recovered. A first observation was that the presence of such
a remarkable amount n-propanol hardly derived from the
hydrolysis of DPrC. If so, the required quantity of water (0.31 g,
1
7 mmol) adsorbed on the catalyst would have been comparable
to the weight of the catalyst itself. A plausible explanation for
Scheme 8
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 863–872 | 867

Discussion
More specifically, according to the acid–base scale proposed
21,23
by Barthomeuf et al.,
Fig. 2 shows that the increase of the
The decarboxylation of DMC and the nature of the catalyst
reaction conversion is in good agreement with the trend of
basic strength of faujasites: the more basic the zeolite (NaX
CsY > NaY ≥ LiY), the higher its decarboxylating activity.
The formation of gaseous products during the decarboxylation
of DAlCs, particularly of DMC, requires a non-conventional
experimental setup. In order to provide a complete reaction
analysis, including conversions and yields, the whole apparatus
must be designed not only to characterize final products, but also
to recover them quantitatively. An accurate tuning/calibration
of the system is required, especially for the balance between
the scale of the reaction, the size of trapping bottles for gases,
and vacuum cycles to be applied (details are given in the
Experimental section). This technical barrier is perhaps one of
the limitations why a detailed investigation for such a reaction
was never previously reported for these substrates.
As far as the role of the catalyst, the adsorption phenomena of
DMC over zeolites should be examined. Na-faujasites are often
described as amphoteric catalysts due to the presence of basic
oxygen atoms of aluminosilicate framework and weak Lewis
acid metal cations. The latter in particular, are considered to
be responsible for the interaction of dimethyl carbonate with
the zeolite surface: according to IR and Raman investigations
~
An hypothesis for this behaviour is based on the following
reasoning. Alkali metal-exchanged faujasites, particularly the
cheap and safe NaX and NaY, are claimed as some of the
24
most effective adsorbents for CO
an excellent potential reaching an adsorption efficiency up to
.5 mmol of CO per g of solid. Moreover, the affinity for carbon
dioxide (under moderate pressures) has been demonstrated to
2
capture/storage: they have
5
2
24,25
increase with the basicity of the faujasites.
In our case,
this feature is expected to produce two opposite effects: (i) it
plausibly facilitates the initial breakdown of DMC (DAlC) on
NaX with respect to NaY (Scheme 3); (ii) it may determine a
preferred saturation of the surface of more basic faujasites by
CO
lead to a drop of the catalyst activity. It should be noted,
however, that CO adsorption on zeolites is highly reversible
>90%) and the adsorption capacity decreases with a rise of
2
released from the decarboxylation reaction, which would
21
2
(
24
temperature. Under the investigated conditions (at around
◦
22
200 C), the dynamic of the adsorption process is likely to inhibit
reported by us and by others, the adsorption of DMC over
Na-faujasites takes place as shown in Scheme 9.
the saturation of the catalytic surface by CO , thus leaving a
2
sufficient solid basicity to discriminate the activity of different
faujasites.
It should be noted however, that the basicity per se is not
a sufficient requisite for a rapid decarboxylation reaction. Solid
K
2
CO , in fact, is a rather poor catalyst (Fig. 2) probably because
3
of its inefficiency in the adsorption of DMC. By contrast, the
high conversion with hydrotalcite (HT, Fig. 2) is ascribable to
26
Scheme 9
the moderate amphoteric properties exhibited by this material.
DMC acts as a base to form acid–base complexes (I and II)
with the Lewis acidic sites (Na+) of the catalyst. In both I and
The decarboxylation of higher dialkyl carbonates
II species, the O–CH
3
and C–OCH
3
bonds are weakened. This
In the case of diethyl- and dipropyl-carbonate, their adsorption
over faujasites, their decarboxylation mechanism, as well as the
relative catalytic activity of NaX and NaY (Fig. 4 for DEC),
can be explained according to the discussion proposed for
fact implies an electrophilic activation of DMC after adsorption
over faujasites, and in the last instance it can explain the catalytic
activity of such solids in methylation/carboxymethylation reac-
tions (Scheme 2).
The occurrence of adducts I and II also offers an interpre-
tation for the decarboxylation reaction examined in this paper.
Scheme 10 shows a plausible mechanistic hypothesis via a polar
intra- and inter-molecular reactions (left and right, respectively).
For simplicity only compound I is considered. The elongation of
27
DMC. The same holds for the selectivity in the reactions of
DMC and DEC: the presence of alcohols (MeOH and EtOH) is
compatible with minor hydrolysis of the two carbonates due to
28
residual water adsorbed on the catalysts. Major differences
concern two aspects. (i) The loss of CO2 from both DEC
and DPrC is more energy intensive than for DMC. The
corresponding decarboxylation reactions take place at 240 and
both O–CH
3
and C–OCH bonds in the adsorbed DMC favours
3
◦
their cleavage with the extrusion of CO
2
; the recombination
200 C, respectively (compare Fig. 1a and 4). This result is
-
d
+d
(
through a nucleophilic attack) of OCH
3
and CH
3
groups
consistent with the general trend observed for a number of
alkylation and carboxyalkylation processes mediated by dialkyl
carbonates, where a dramatic drop of reactivity always occurs
if DMC is replaced with heavier DAlCs (except for dibenzyl
produces dimethyl ether. Whatever the mechanism, results of
Fig. 1–3 leave no doubts about the superior catalytic efficiency
of NaX with respect to NaY, for the decarboxylation of DMC.
1,2
carbonate). An increase of the steric hindrance around both
alkyl and carbonyl carbons of the reacting carbonates accounts
for this behavior. (ii) The decarboxylation of DPrC is not a
selective process. The expected product, dipropyl ether (DPE),
forms along with two side compounds such as n-propanol and
propylene (Scheme 5). Although some hydrolysis of DPrC is
possible, this reaction hardly accounts for the sizable amount
of the alcohol (up to 62%) detected in the liquid mixture.
Scheme 10
8
68 | Green Chem., 2011, 13, 863–872
This journal is © The Royal Society of Chemistry 2011

Instead, the cracking of DPE catalyzed by NaX offers the most
reasonable pathway to explain the presence of both n-propanol
and propylene (Scheme 5). It should be noted that the conversion
of dialkyl ethers to the corresponding olefins has been claimed in
the literature, by using alkali and alkali earth metal-exchanged
corresponding dimethyl and diethyl ethers in a very high yield.
Heavier carbonates instead, give different products: dipropyl
carbonate provides a mixture of dipropyl ether, n-propanol
and propylene; likewise, dioctyl carbonate provides dioctyl
ether, n-octanol, and octene(s), respectively. The mechanism
of adsorption of DMC over Na-faujasites, offers a rationale
to discuss the onset of the decarboxylation reaction, while the
acid–base properties of zeolites allow an interpretation for the
catalytic activity of the tested catalysts: the more basic the
17
faujasites.
In line with previous investigations,
1,2,5
these results clearly
exemplify how the reactivity of light DAlCs, particularly DMC,
can be remarkably different from that of their heavier homo-
logues. A difference that may be determined also by the nature of
the catalysts used. These considerations are further corroborated
by the behavior of dioctyl-carbonate (DOC). Although the
presence of a reaction solvent (CyH) does not allow a direct
comparison with DPrC, the reaction of DOC catalyzed by NaX
shows a product distribution quite similar to DPrC, yielding
faujasite, the better its capability to favor the release of CO
2
during the reaction, the better its performance. In the case of
dipropyl- and dioctyl-carbonate, once decarboxylation takes
place, cracking of the corresponding dialkyl ethers catalysed
by Na-faujasites is a plausible pathway to explain the presence
of alcohols and alkenes as co-products.
dioctyl ether, n-octanol and isomeric octenes (Scheme 7). At
Overall, the results reported in this work, highlight two
significant aspects: they elucidate the possible limitations of
the use of dialkyl carbonates in the presence of faujasite
catalysts and they offer boundaries to the reaction conditions,
i.e. temperature and nature of the catalysts, in order to avoid side
reactions by which an extensive consumption of DAlCs may take
place along with the formation of highly flammable ethers such
◦
1
80–250 C, the decarboxylation of DOC has been described
8
also with a hydrotalcite catalyst. In this case however, reaction
products were only dioctyl ether and n-octanol: the authors did
not observe octene(s) and the formation of the alcohol (up to
2
0% at complete conversion) was attributed to the hydrolysis of
the reactant carbonate.
Dimethyl ether (DME) – due to its structure – generated
in the reaction of DMC, cannot decompose according to the
cracking reaction of Scheme 6. Also the hydrolysis of DME
to MeOH is not a plausible process: this requires a very high
as DME and DEE, and the co-generation of undesired CO . In
2
order to control the efficiency and safety of DAlCs-promoted
alkylations, one must be aware that a threshold temperature for
decomposition of organic carbonates should be established case
by case for any given catalytic system.
Finally, water adsorbed on faujasites does not seem to have a
role on the decarboxylation of DAlCs.
◦
temperature (usually ≥300 C), strongly acid catalysts and
29
a water(steam)/DME ratio > 2.5, clearly not compatible
with conditions here investigated. Therefore, the presence of
water adsorbed on the catalyst is unlikely to alter the product
distribution in the reaction of DMC. In this respect, it should
also be noted that the conversion of DMC is not affected by
Experimental
◦
different thermal treatments (at 70, 200, and 400 C, respectively)
General
used for NaY catalyst: from the mechanistic standpoint, this
suggests that (different) amounts of water adsorbed by the
faujasite do not enter the decarboxylation step of DMC.
In the reaction of DEC, not even trace amounts of ethylene
are observed, meaning that diethyl ether (DEE) does not
decompose over the catalyst. No clear reasons account for this
behavior, though the trend of alkene stability (RCH CHR >
All reagents used [K
carbonate (DEC), glycerine carbonate (GlyC), dimethoxyethane
DME), cyclohexane, n-propanol, propyl chloroformate, and n-
2
CO
3
, dimethylcarbonate (DMC), diethyl-
(
octanol] were ACS grade and were employed without further
purification. GlyC was a generous gift from Huntsman Cor-
poration. Faujasites NaY and NaX were commercial samples
from Aldrich (art. # 334448 and 28,359-2); LiY and CsY were
synthesized from NaY by an ion exchange using LiCl and CsCl
30
CH
2
CHR > CH
2
CH ) can play a role. Alike to DME,
2
also the hydrolysis of DEE to EtOH should not occur under the
examined conditions.
9a
precursors. The percentage of ion exchange was determined
9b
according to a method previously reported by us: based on
the initial Na content, the % exchange was 67% and 56%, for
Li and Cs, respectively. Hydrotalcite KW2000 was from Kyowa
Conclusions
Chemical Industry Co. Tokyo, (Mg0.7Al0.3O1.15) with a specific
The combined use of dialkyl carbonates (DAlCs) and alkali
metal exchanged faujasites is a powerful synthetic tool to pro-
mote the alkylation of a number of N-, O-, and S-nucleophiles
with excellent selectivities, above 95%, at quantitative conver-
sions. However, this paper demonstrates that the experimental
conditions, especially the high temperature often required for
such processes, may also trigger important side reactions: in the
2
-1
surface area 202 m g . The catalysts (faujasites, hydrotalcite,
◦
and K
2
CO
3
) were stored under vacuum (8 mbar) at 70 C. In two
additional experiments, two samples of NaY were activated in a
-
1
◦
stream of dry air (10 mL min ), at 200 and 400 C, respectively.
GC-MS (EI, 70 eV) analysis were run using a HP5/MS capillary
column (30 m). The infrared spectra have been recorded on the
Bruker Vertex 70 FTIR spectrometer equipped with a Ge/KBr
beamsplitter, a Globar source, and a DTGS detector.
◦
range of 150–240 C, a decarboxylation of the reactant dialkyl
carbonates may take place concurrently to the desired alkylation
reactions. Furthermore, the decarboxylation of DAlCs may
be followed by other processes whose occurrence is strictly
dependent on the nature of DAlCs. In particular, dimethyl-
Dipropyl carbonate (DPrC)
DPrC was prepared by adjusting a general method for the
15
and diethyl-carbonate undergo the loss of CO
2
to produce the
synthesis of dialkyl carbonates: a 250 mL 2-necked flask
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 863–872 | 869

equipped with a condenser and a dropping funnel, was charged
purged by a low temperature trap-to-trap apparatus. The system
is outlined in Fig. 5.
with a mixture of n-propanol (0.33 mmol, 20 g) and pyridine
◦
(
0.34 mmol, 27 g). The flask was cooled at 0 C and propyl
chloroformate (0.31 mmol, 38 g) was added dropwise under
vigorous stirring. The mixture was allowed to reach rt. Then, aq.
HCl (5%, 30 mL) was added. The organic phase was extracted
with diethyl ether (2 ¥ 120 mL), and dried over Na
2
SO . The
4
◦
product DPrC was distilled (bp = 83 C/86 mbar; Y = 35 g,
1
7
7%) and characterized by GC-MS and H NMR.
1
H NMR (CDCl3, 400 MHz) d (ppm): 4.07 (t, 4H, J = 6.7
Hz), 1.68 (m, 4H), 0.95 (t, 6H, J = 7.4 Hz). GC-MS (EI, 70 eV),
+
m/z: 146 ([M] , <1%), 63 (42), 59 (27), 43 (100), 42 (16), 41
(
28). Spectroscopic properties were in agreement with those of
an authentic commercial sample of DPrC.
2
Fig. 5 The apparatus used for the recovery of gaseous CO and
dimethyl ether. A: autoclave; CB: cooling bath; CF: cold finger; B:
air-tight Drechsel bottle; V: valve; VG: vacuum gauge; T: temperature
control.
Dioctylcarbonate (DOC)
DOC was prepared by the transesterification reaction of n-
octyl alcohol with DMC, accordingly to a procedure already
reported. Starting from 26 g (0.2 mol) of n-octanol, DOC was
8
A vacuum line was equipped with a cold finger (250 mL, CF)
isolated in a 68% yield (20 g) and characterized by GC-MS and
and a Drechsel bottle (250 mL, B). By valves V
1
–V , the whole
2
1
H NMR.
system was set to a pressure £1 Torr. This was continuously
monitored by a vacuum gauge (VG). Then, the autoclave (A)
containing the reaction mixture, the finger CF, and the bottle B
1
H NMR (CDCl , 400 MHz) d (ppm): 4.11 (t, 4H, J = 6.7 Hz),
3
1
(
(
.65 (m, 4H), 1.33 (m, 20H), 0.87 (t, 6H, J = 6.8 Hz). GC-MS
+
EI, 70 eV), m/z: 286 ([M] , <1%), 113 (32), 84 (22), 83 (26), 71
were simultaneously dipped in three cooling baths (CB
1
, CB
) set at -60 C, -196 C, and -196 C, respectively. The
valves V –V were closed and the autoclave was connected to
CF by opening V : once the pressure of the system reached a
constant value (~150 Torr), V was opened to vent N charged
2
,
◦
◦
◦
100), 70 (33), 69 (37), 57 (74), 56 (35), 55 (46), 43 (49), 42 (24),
and CB
3
4
1 (64). Spectroscopic properties were in agreement with those
1
2
8
reported in the literature.
3
1
2
The decarboxylation of DMC: general procedure
in the reactor at the beginning of the experiment. Then, V
1
◦
was closed and the bath CB
temperature, V
1
was brought to -20 C. At this
A 120 mL stainless steel autoclave fitted with two valves, a
pressure gauge and a thermocouple for temperature control,
was charged with DMC and the catalyst. The autoclave was
3
allowed to purge the reaction products, CO and
2
DME, from the autoclave. Both compounds were condensed in
CF along with a very small fraction (£0.5 mL) of the residual
closed and, after a careful air removal by 3 vacuum/N
2
DMC. V
CB
selectively CO
remained in CF. Eventually, V
3
was closed and acetone was added to the bath
cycles, it was electrically heated at the desired temperature. The
reaction mixture was kept under magnetic stirring throughout
the experiments. The following conditions and amounts of
DMC/catalyst were used.
Figure 1: 10 mL (10.7 g, 119 mmol) of DMC and 600 mg of
the catalyst (weight ratio, Q, catalyst : DMC = 0.056). Catalysts
were NaX and NaY faujasites, and hydrotalcite (HT: KW2000).
The reaction temperature was in the range of 150–220 C, and
the reaction time was 6 h.
Figure 2: 10 mL (10.7 g, 119 mmol) of DMC and 600 mg of the
catalyst (weight ratio, Q, catalyst : DMC = 0.056). Six catalysts
were used: LiY, NaY, CsY and NaX faujasites, hydrotalcite (HT:
◦
2
until it reached -50 C. Then, V
2
was opened to convey
2
and DME from CF to the bottle B, while DMC
was closed. The bottle B was
3
disconnected from the vacuum line, allowed to return gradually
to ambient temperature, and weighed. The gaseous mixture in
the bottle was then analysed by GC-MS.
The conversion of DMC reported in Fig. 1–3 was calculated
◦
by the total mass (and molar) amounts of CO
2
and DME,
according to the stoichiometry given in Scheme 3. MS spectra of
both CO
2
and DME gave a match quality higher than 97% when
31
compared to standard products of the Wiley-MS database. The
structure of DME was further confirmed by the comparison of
the IR spectrum of the reaction mixture with that of an authentic
sample of DME (details are given in the ESI†).
◦
KW2000), and K
2
CO
3
. The reaction temperature was 200 C
and the reaction time was 6 h.
Figure 3a: 10 mL (10.7 g, 119 mmol) of DMC were set to react
with increasing quantities of both NaY and NaY. The weight
ratio, Q, catalyst : DMC was varied from 0.015 to 0.084. The
The decarboxylation of DEC and DPrC
◦
reaction temperature was 200 C and the reaction time was 6 h.
The decarboxylation of both DEC and DPrC were carried out
according to the procedure described for DMC. The following
conditions and amounts of dialkyl carbonate/catalyst were
used.
Figure 3b: 10 mL (10.7 g, 119 mmol) of DMC and 600 mg of
the catalyst (weight ratio, Q, catalyst : DMC = 0.056). Catalysts
were NaY or NaX faujasites. The reaction temperature was
◦
2
00 C, and the reaction time was in the range of 1–18 h.
DEC, Figure 4: 10 mL (83 mmol) of DEC and 550 mg of the
catalyst (weight ratio, Q, catalyst : DEC = 0.056). Catalysts were
NaX and NaY faujasites. The reaction temperature was in the
At the end of each experiment, the autoclave was cooled at
room temperature. In order to recover the gaseous products of
the reaction, CO and dimethyl ether (DME), the autoclave was
◦
2
range of 180–240 C, and the reaction time was 6 h. The GC-MS
8
70 | Green Chem., 2011, 13, 863–872
This journal is © The Royal Society of Chemistry 2011

analysis of the gas phase showed the presence of CO
product: its (weighted) mass amount allowed calculation of the
conversion of DEC.
2
as a sole
this work. Mr. Alessandro Baldan is also acknowledged for his
support in IR spectra.
The GC-MS analysis of the residual liquid mixture in the
autoclave showed two products: diethyl ether (major, >90%
at complete conversion) and ethanol (trace amounts, 3–5%).
The analysis of the liquid phase also offered the possibility to
determine the conversion of DEC whose values matched (within
Notes and references
1
(a) A.-A. Shaik and S. Sivaram, Chem. Rev., 1996, 96, 951–976; (b) P.
Tundo and M. Selva, Acc. Chem. Res., 2002, 35, 706–716; (c) V. K.
Ahluwalia, and M. Kidwai, In New Trends in Green Chemistry,
Kluwer Academic and Anamaya Pub., New Delhi, India, 2004;
(d) Carbon Dioxide as Chemical Feedstock, M. Aresta, Ed.; Wiley
VCH, Weinheim, 2010, Ch. 7.
±
3%) those calculated from the weight of CO . The structure of
2
diethyl ether and ethanol were confirmed by their mass spectra
and by comparison with authentic commercial samples.
DPrC, Scheme 5: 8.8 mL (56 mmol) of DPrC and 300 mg of
the faujasite (the weight ratio, Q, catalyst : DPrC = 0.036). The
2
(a) M. Selva, P. Tundo, A. Perosa and S. Memoli, J. Org. Chem., 2002,
6
7, 1071–1077; (b) A. Loris, A. Perosa, M. Selva and P. Tundo, J. Org.
Chem., 2004, 69, 3953–3956; (c) M. Selva, P. Tundo, A. Perosa and
F. Dall¢Acqua, J. Org. Chem., 2005, 70, 2771–2777; (d) M. Selva, A.
Perosa, P. Tundo and D. Brunelli, J. Org. Chem., 2006, 71, 5770–5773;
◦
temperature was set at 220 and 240 C for 6 and 12 h, respectively.
(
e) M. Selva and A. Perosa, Green Chem., 2008, 10, 457–464.
CO
spectra of both compounds gave a match quality higher than
5% when compared to standard products of the Wiley-MS
2
and propylene were detected as gaseous products. MS
3
4
B. Schaffner, F. Schaffner, S. P. Verevkin and A. Borner, Chem. Rev.,
2010, 110, 4554–4581.
Alkali metal exchanged Y- and X-faujasites are a class of zeolites in
which the negative charge of the aluminosilicate framework is coun-
terbalanced by an alkali metal cation. For example, NaY and NaX
faujasites possess the same 3D-structure, but they differ for the Si/Al
9
31
database. However, the conversion of DPrC could not be
17
estimated from the weighted amounts of CO and propene.
2
The GC-MS analysis of the residual liquid mixture in the
autoclave showed two products: dipropyl ether (17–40%) and
n-propanol (23–60%). The structure of dipropyl ether and
n-propanol were confirmed by their mass spectra and by
comparison with authentic commercial samples.
content. Their general formulas are Na56[(AlO
2
)
56(SiO
2
)
136]·250H
2
O
and Na86[(AlO
2
)
86(SiO
2
)
106]·264H
2
O, respectively. For morphological
details and other properties, see: (a) F. Schwochow and L. Puppe,
Angew. Chem., Int. Ed. Engl., 1975, 14, 620–628; (b) G. C. Bond,
in Heterogeneous Catalysis Principles and Applications, Oxford
University Press, NewYork, USA, 2nd edn, 1987, pp. 104–110.
(a) M. Selva, A. Bomben and P. Tundo, J. Chem. Soc., Perkin Trans.
5
1
, 1997, 1041–1045; (b) M. Selva, P. Tundo and A. Perosa, J. Org.
Chem., 2003, 68, 7374–7378; (c) M. Selva, P. Tundo and T. Foccardi,
J. Org. Chem., 2005, 70, 2476–2485; (d) M. Selva and P. Tundo, J. Org.
Chem., 2006, 71, 1464–1470; (e) M. Selva, E. Militello and M. Fabris,
Green Chem., 2008, 10, 73–79; (f) M. Selva, A. Perosa and M. Fabris,
Green Chem., 2008, 10, 1068–1077; M. Selva and M. Fabris, Green
Chem., 2009, 11, 1161–1172.
The reaction of dioctyl carbonate (DOC) and glycerol carbonate
(GlyC)
The reaction of both DOC and GlyC were carried out with the
same procedure used for DMC. However, due to the viscosity of
the starting carbonates, a solvent was necessary. The following
conditions and amounts of dialkyl carbonate/catalyst were
used.
6 (a) J. Tsuji, K. Sato and H. Okumoto, J. Org. Chem., 1984, 49, 1341–
1
344; (b) J. Tsuji and I. Minami, Acc. Chem. Res., 1987, 20, 140–145.
7
(a) D. A. Loy, J. V. Beach, B. M. Baugher, R. A. Assink, K. J. Shea,
J. Tran and J. H. Small, Chem. Mater., 1999, 11, 3333–3341; (b) M.
Wu, J. Guo and H. Jing, Catal. Commun., 2008, 9, 120–125; (c) Y. S.
Park and W. K. Lee, Bull. Korean Chem. Soc., 1997, 18(4), 360–361.
P. Tundo, F. Aric o` , A. E. Rosamilia and S. Memoli, Green Chem.,
DOC, Scheme 7: 2.0 g (10.8 mmol) of DOC, 120 mg of NaX
faujasite (the weight ratio, Q, catalyst : carbonate, was 0.06),
and cyclohexane (9 mL) as solvent. The reaction temperature
8
9
2
008, 10, 1182–1189.
◦
was set at 240 C, and the reaction time was 6 h.
The ion exchange reaction was carried out as described by: (a) M.
Onaka, K. Ishikawa and Y. Izumi, Chem. Lett., 1982, 1783; (b) M.
Selva, P. Tundo and A. Perosa, J. Org. Chem., 2002, 67, 9238–9247.
The GC-MS analysis of the gas phase showed CO as
2
a sole product. The GC-MS analysis of the residual liquid
mixture in the autoclave showed the presence of unreacted DOC
1
1
0 Analytical Methods Manual, Varian Inc., 1989.
◦
1 The temperature range (150–200 C) was chosen according to
(
(
42%), dioctyl ether (7%), n-octanol (28%), and isomeric octenes
mixture of 1-octene, 2-octene and other octenes: total of 23%).
the operating conditions usually employed for DMC-mediated
methylation reactions.
2 Hydrotalcite is an aluminium-magnesium hydroxycarbonate.
1
The structure of dioctyl ether, n-octanol, 1-octene and 2-octene
was confirmed by their mass spectra and by comparison with
authentic commercial samples.
Natural
hydrotalcite
possesses
the
general
formula
and Al
2+
3+
Mg Al (OH)16CO
6
2
3
·4H O. However, both Mg
2
can be isomorphously substituted by other metal ions,
forming new hydrotalcite-like materials of general formula
GlyC, Scheme 8: 2.0 g (17 mmol) of GlyC, 120 mg of NaX
faujasite (the weight ratio, Q, catalyst : carbonate, was 0.06), and
dimethoxyethane (9 mL) as solvent. The reaction temperature
2+
3+
x+
n-
[
M
1-x
M
x
(OH)
2
]
A
x/n ·mH
2
O, where M denotes metal ions,
A denotes exchangeable anions with valence n, and x is within
0
.17–0.33. See, for example: A. Bhattacharyya, V. W. Chang and
◦
was set at 240 C, and the reaction time was 6 h.
D. J. Schumacher, Appl. Clay Sci., 1998, 13, 317–328.
1
3 It should be noted for clarity that at rt, a refluxing liquid was visible
at the bottom of the Drechsel bottle. Once the bottle was opened, its
content was completely released to the atmosphere. The liquid was
The GC-MS analysis of the residual liquid mixture in the
autoclave showed the formation of several products. Among
them, glycerol (10–12%) and glycidol (3–5%) were detected.
The structure of both compounds were confirmed by their mass
spectra and by comparison with authentic commercial samples.
◦
compressed dimethyl ether (bp = 24.8 C), as confirmed by the GC-
MS analysis. In addition, this also explained why the GC response
of the gaseous mixture showed a slight excess of CO
CO /DME signals was, on average, 1.2.
4 K CO was reported as a catalyst for a number of DMC-mediated
alkylations (see refs. 1–2). Therefore, the knowledge of the behavior
of K CO as a catalyst for the decarboxylation of DMC was rather
important.
5 A 9 : 1 mixture of benzyl methyl ether and dibenzyl ether [(PhCH
2
: the ratio of
2
1
1
2
3
Acknowledgements
2
3
MIUR (Italian Ministry for Education, University and Re-
search) is gratefully acknowledged for the financial support of
2 2
) O;
DBE] was obtained. The small amount (10%) of dibenzyl ether
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 863–872 | 871

plausibly derived from two reactions, both catalyzed by the NaX
faujasite: the dehydration of benzyl alcohol (see ref. 5d), and/or
the transesterification of benzyl alcohol to dibenzylcarbonate
22 (a) T. Beutel, J. Chem. Soc., Faraday Trans., 1998, 94, 985–993; (b) F.
Bonino, A. Damin, S. Bordiga, M. Selva, P. Tundo and A. Zecchina,
Angew. Chem., Int. Ed., 2005, 44, 4774–4777.
(
PhCH
2
OCO
2
CH
2
Ph) followed by a decarboxylation process. In our
23 B. Su and D. Barthomeuf, Stud. Surf. Sci. Catal., 1995, 94, 598–603.
24 (a) G. Maurin, P. L. Llewellyn and R. G. Bell, J. Phys. Chem. B, 2005,
109, 16084–16091; (b) D. Bonenfant, M. Kharoune, P. Niquette, M.
Mimeault and R. Hausler, Sci. Technol. Adv. Mater., 2008, 9, 013007.
25 S. V. Bordawekar and R. J. Davis, J. Catal., 2000, 189, 79–90.
26 (a) J. Shen, J. M. Kobe, Y. Chen and J. A. Dumesic, Langmuir, 1994,
10, 3902–3908; (b) J. Shen, M. Tu and C. Hu, J. Solid State Chem.,
1998, 137, 295–301.
27 To the best of our knowledge, the interactions of faujasites with
dialkyl carbonates other than DMC have not been investigated so
far. However, the adsorption of DEC, DPrC, and DOC on such
catalysts is expected to proceed as for DMC, via the formation of
acid–base complexes similar to those shown in Scheme 9.
28 Alkali metal faujasites of both X- and Y-type are hygroscopic solids,
see: Organic Solid State Reactions, ed. F. Toda, Kluwer Academic
Publishers, Dordrecht, The Netherlands, 2002, pp. 160–161.
29 (a) K. Faungnawakij, Y. Tanaka, N. Shimoda, T. Fukunaga, S.
Kawashima, R. Kikuchi and K. Eguchi, Appl. Catal., A, 2006,
304, 40–48; (b) K. Faungnawakij, R. Kikuchi, T. Fukunaga and K.
Eguchi, J. Phys. Chem. C, 2009, 113(43), 18455–18458.
case, the presence of DBE as a side-product left a minor uncertainty
on the final conversion of DMC.
1
1
6 M. Matzner, R. P. Kurkjy and R. J. Cotter, Chem. Rev., 1964, 64,
6
45–687.
7 Unfortunately, the GC signals of CO
perfectly resolved and an estimation of the CO
2
and propylene could not be
amount was not
2
possible. As a consequence, the conversion of DPrC could not be
calculated from the weight of gaseous products.
8 (a) D. D. Rosenfeld, D. D. Pete, L. L. Iaccino, and B. Hammond,
Exxon Chemical Company, WO/1993/025504 (December 23, 1993);
1
1
(
b) G. I. Cox, D. D. Pete, and D. D. Rosenfeld, Exxon Chemical
Company, WO/1995/011209 (April 27, 1995).
9 (a) J. A. Kenar, G. Knothe, R. O. Dunn, T. W. Ryan III
and A. Matheaus, J. Am. Oil Chem. Soc., 2005, 82(3), 201–
2
05; (b) Huntsman, Technical Bulletin (http://193.221.91.170/
performance _ products / Media / JEFFSOL%C2%AE _ Glycerine _
Carbonate.pdf).
2
2
0 Faujasites have been reported to catalyse the isomerization of olefins,
see for example: Y. Ono and T. Baba, Catal. Today, 1997, 38, 321–
3
37.
30 J. McMurry, In Organic Chemistry, 6th Ed., Brooks Cole, Pub, 2003.
31 Wiley Library of Mass Spectral Data, 2002: software of the GC-MS
instrument.
1 (a) D. Barthomeuf, J. Phys. Chem., 1984, 88, 42–45; (b) D.
Barthomeuf, Catal. Rev., 1996, 38(4), 521–612.
8
72 | Green Chem., 2011, 13, 863–872
This journal is © The Royal Society of Chemistry 2011