Synthesis of di-n-butyl carbonate from n-butanol: Comparison of the direct carboxylation with butanolysis of urea by using recyclable heterogeneous catalysts

Article abstract of DOI:10.1016/j.cattod.2016.02.005

The synthesis of di-n-butyl carbonate has been studied starting from n-butanol and either CO₂ or urea. A comparison of the two synthetic routes is reported. Several mixed oxides have been synthesized and tested with the aim of finding a catalyst active in mild conditions (T, t), recoverable and reusable. Different strategies to push the reaction toward the formation of the target product (di-n-butylcarbonate) have been applied and adapted to each case. The pervaporation membrane and chemical water traps are compared as techniques for water elimination and equilibrium shift in the direct carboxylation. Among the tested catalysts, 0.03Nb₂O₅/CeO₂ is the best in the case of the direct carboxylation of butanol, whereas 0.5MgO/ZnO results the best in terms of activity and robustness for the alcoholysis of urea.

Full text of DOI:10.1016/j.cattod.2016.02.005

G Model
CATTOD-10035; No. of Pages8
ARTICLE IN PRESS
Catalysis Today xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Synthesis of di-n-butyl carbonate from n-butanol: Comparison of the
direct carboxylation with butanolysis of urea by using recyclable
heterogeneous catalysts
Antonella Angelinia,b, Angela Dibenedettoa,b,∗, Stefania Fasciano , Michele Arestac
b
a
Department of Chemistry, University of Bari, Campus Universitario, Via Orabona 4, 70126 Bari, Italy
CIRCC, Via Celso Ulpiani 27, 70126 Bari, Italy
National University of Singapore, Department of Chemical & Biomolecular Engineering, 4 Engineering Drive, E5 #02-09 Singapore 117585
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis of di-n-butyl carbonate has been studied starting from n-butanol and either CO2 or urea.
A comparison of the two synthetic routes is reported. Several mixed oxides have been synthesized and
tested with the aim of finding a catalyst active in mild conditions (T, t), recoverable and reusable. Different
strategies to push the reaction toward the formation of the target product (di-n-butylcarbonate) have
been applied and adapted to each case. The pervaporation membrane and chemical water traps are
compared as techniques for water elimination and equilibrium shift in the direct carboxylation. Among
the tested catalysts, 0.03Nb2O5/CeO2 is the best in the case of the direct carboxylation of butanol, whereas
Received 28 October 2015
Received in revised form 4 January 2016
Accepted 1 February 2016
Available online xxx
Keywords:
Carboxylation of butanol
Alcoholysis of urea
Mixed oxides
0
.5MgO/ZnO results the best in terms of activity and robustness for the alcoholysis of urea.
©
2016 Elsevier B.V. All rights reserved.
Dibutylcarbonate
1
. Introduction
Dialkyl carbonates are chemicals largely used in the chemi-
dialkyl carbonates are required. However, new synthetic routes are
wished which may be safer and more sustainable.
Dialkyl carbonates can be prepared from CO2 through either
the direct carboxylation of alcohols (Eq. (1)) [1,19–22] or the
cal and polymer industry. Finding new synthetic methodologies,
which respond to the requisites of sustainability, is of paramount
importance for the chemical industry. In this paper, we focus on
di-n-butylcarbonate (D BC), a member of the dialkylcarbonates
family, that finds application as solvent and is an important chemi-
alcoholysis of urea, an active form of CO . (Eqs. (2a) and (2b))
2
◦
◦
(ꢀGf urea = −197.5 kJ/mol; ꢀGf
= −394.4 kJ/mol) Today urea
CO2
n
represents over 65% of products industrially produced from CO
[23].
2
n
cal intermediate for the synthesis of organic compounds [1]. D BC is
also used as phosgene substitute for polycarbonates synthesis [2],
as base material for lubricants, for metal degreasing and leather
processing [3].
D BC is generally prepared via synthetic routes such as:
phosgenation [1,4], oxidative carbonylation of butanol [5–7], trans-
esterification of other carbonates [8–18]. Although phosgene
CO2 + 2ROH → (RO)2CO + H2O
(1)
(2a)
(2b)
H2NCONH2 + ROH → H2NC(O)OR + NH3
H2NC(O)OR + ROH → (RO)2CO + NH3
n
Although both the direct and indirect route based on CO2 have
an environmental impact much lower than the commonly used
phosgenation and offer great advantage in terms of atom econ-
omy [24], they do not have industrial application because of some
bottlenecks.
technology is quite safe today, COCl is banned in several countries
2
and, most important, cannot be transported. The carbonylation
process has drawbacks due to the use of chlorides and NO. In
the transesterification process expensive precursors such as other
The direct carboxylation of alcohols (Eq. (1)) has thermody-
namic limitations that cause low conversion yield. At T > 400 K the
equilibrium concentration of the carbonate is as low as 0.1–1%, due
to the fact that, although the ꢀH is slightly negative, the adverse
∗ Corresponding author at: Department of Chemistry, University of Bari, Campus
Universitario, Via Orabona 4, 70126 Bari, Italy. Fax: +39 080 5443606.
E-mail address: angela.dibenedetto@uniba.it (A. Dibenedetto).
entropic factor due to the use of gaseous CO , which is found at the
2
end in a liquid product, makes the ꢀG positive [25].
http://dx.doi.org/10.1016/j.cattod.2016.02.005
0
920-5861/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: A. Angelini, et al., Synthesis of di-n-butyl carbonate from n-butanol: Comparison
of the direct carboxylation with butanolysis of urea by using recyclable heterogeneous catalysts, Catal. Today (2016),
http://dx.doi.org/10.1016/j.cattod.2016.02.005

G Model
CATTOD-10035; No. of Pages8
ARTICLE IN PRESS
A. Angelini et al. / Catalysis Today xxx (2016) xxx–xxx
2
The removal of water in reaction 1 helps to shift the chemi-
cal equilibrium toward the carbonate, leading to better carbonate
yields. Several organic and inorganic water traps have been used
2.1. Synthesis of the catalysts
ꢀ
ꢀꢀ
The notation [xM O/M O] is used for the mixed oxides, where x
ꢀ ꢀꢀ
[
26–28]. Attempts to remove water from the reaction mixture
indicates the molar ratio of M to M taken as one.
using pervaporation membanes in the case of ethanol have afforded
interesting results [20d].
Alternatively, the alcoholysis of urea is a promising synthetic
route [29–34] since urea is abundant (it is produced at over
2
.1.1. Synthesis of 2CaO/CeO2
CaCO3 (1.32 g) and Ce(NO ) ·6H O (2.85 g) were mixed at the
3 3
2
solid state using HEM. At the end of the milling (30 min at 700 rpm)
the mixture was calcined for 3 h at 823 K.
1
80 Mt/y from CO ), inexpensive, nontoxic, and environmentally
2
friendly. Upon interaction of urea with alcohol [Eq. (2a)] a urethane
is first formed, which is then transformed into the dialkyl carbonate.
2
.1.2. Synthesis of 0.03Nb O5/CeO
2 2
[
Eq. (2b)] Reaction 2a easily occurs even in absence of catalysts and
a) To 7 g of Ce(NO ) (NH ) dissolved in anhydrous ethanol
3
6
4 2
at lower temperatures than reaction 2b, which, instead, requires a
catalyst. Metal oxides (such as ZnO, CeO , MgO, CdO, La O [35])
(30 mL), 95 ␮L of [Nb(OEt)5]2 dissolved in 1 mL of EtOH were
2
2
3
added dropwise, keeping the mixture under stirring. A NH3(aq)
solution in water (1:10 v/v) was added dropwise until pH 9 was
reached. During the addition of the base the precipitation of
the hydroxides took place and the solution color changed from
red to yellow and finally to pinkish. The suspension was left
under stirring for 2 h and then it was centrifuged. The separated
hydroxides were washed with water (3 × 5 mL), dried at 503 K
for 12 h and, then, calcined at 823 K for 2 h in air.
are generally employed as catalyst under hard operative conditions
497–523 K, 7–20 h), which may affect the selectivity toward the
carbonate and make complex the separation process.
Herein, we report a comparative study on the synthesis of D BC
through the direct carboxylation of butanol and butanolysis of urea.
In the latter case, considered the higher yield of conversion, we
have also carried out a screening of different mixed oxides paying
attention to their recoverability and recyclability.
(
n
b) 2 g of CeO and 0.1 g of Nb O5 were ground for 30 min by HEM at
2
2
a rate of 700 rpm, reversing the rotation of the agata balls every
0 min. At the end of this process, the solid was calcined at 823 K
2
. Material and methods
1
for 3 h.
Chemicals were purchased from Aldrich (RP). All the commercial
metal oxides were calcined at 823 K for 3 h before use. Commercial
urea was reduced to a fine powder and dried under vacuum before
use in the reaction [36].
Pure D BC used for the calibration curve was obtained by trans-
esterification of DMC with 1-butanol (vide infra).
2
.1.3. Synthesis of xMgO/ZnO
xMgO/ZnO mixed oxides with x = 1; 0.5 were prepared by
HEM. Zn(NO ) ·6H O (2.02 g) was mixed with 1.73 or 0.88 g of
n
3 2
2
Mg(NO ) ·6H O in an agate reactor for 30 min. The obtained solids
3
2
2
were calcined at 823 K for 3 h.
Mixed oxides were prepared by High Energy Milling (HEM)
at 700 rpm by using a Planetary Micro Mill Fritsch Pulverisette
2
.1.4. Synthesis of ZrO /CeO
7
equipped with an agata basket and balls. Alternatively the co-
2
2
Z
r
O
/
C
e
O
w
a
s
p
r
e
p
a
r
e
d
b
y
H
E
M
.
1
g
o
f
Z
r
O
a
n
d
1
.
4
g
o
f
C
e
O
2
2
2
2
precipitation method was used, as specified.
were mixed in an agata reactor and ground for 30 min at a rate of
00 rpm, reversing the rotation of the agata balls every 10 min. At
GC–MS analyses were carried out with a gas chromatograph Shi-
madzu 17 A (capillary column: 30 m; MDN-5S; l 0.25 mm, 0.25 mm
film) coupled to a Shimadzu QP5050A mass spectrometer.
Kinetic studies on the reactions were carried out using a gas
chromatograph HP 6850 series equipped with a FID detector and a
capillary column ZB-WAX (30 m, 0.25 mm).
7
the end of this process, the solid was calcined at 823 K for 3 h.
2.2. Transesterification of DMC with n-butanol
Water content was measured by the Karl Fisher method, using
a Metrohm 785 DMP Titrino apparatus.
The reaction was carried out in a stainless steel autoclave at
463 K for 12 h using a DMC/ BuOH molar ratio of 1:4. The conver-
n
Infrared spectra were obtained with a spectrometer SHIMADZU
IR Prestige 21 using KBr discs. The sample was dispersed in Nujol.
Energy Dispersive X-ray diffraction (EDX) patterns were
recorded at room temperature on a Shimadzu EDX-720/800HS,
using a 5–50 kV Rh target X-ray generator and a Si (Li) detecting
system. For noise reduction, the detector was cooled using liquid
nitrogen.
sion was followed by GC. When DMC was completely converted,
the D BC was separated by double distillation. It had purity higher
than 99% (elemental analyses and GC–MS) and was suited to be
used for building a gas-chromatographic calibration curve.
n
2
2
.3. Direct carboxylation of n-butanol
CO2 and NH -chemisorption and TPD analyses were performed
3
.3.1. Reaction of n-butanol with CO2
−1
under He flow rate at 40 mL min using a Pulse ChemiSorb2750
The reaction was carried out in a 100 mL stainless steel auto-
Micromeritics instrument. Samples were pre-treated under N at
2
clave charged with 5 MPa of carbon dioxide. In a typical catalytic
test 0.1 g of catalyst were added to 4 mL of n-butanol. The reaction
was carried out at 433 K for 6 h or, alternatively, as specified. The
working pressure reached during the reaction was about 6 MPa.
−
1
2
0 mL min flow and 773 K. The ratio strong basic to strong acid
sites was determined on the basis of the amount of carbon diox-
ide and ammonia desorbed during TPD experiments between 473
and 773 K. (Fig. 7) The ratio of strong basic to acid sites was calcu-
lated considering the area relevant to desorbed CO and NH in the
n
The yield in D BC was determined gas-chromatographically using
2
3
n
n-tridecane as internal standard and expressed either as % D BC or
temperature interval 473–773 K (Eqs. (3) and (4)). The gas evolved
below 473 is considered to give weak–medium interactions with
the catalyst.
as % conversion of butanol (Fig. 1).
2
.3.2. Water elimination using CH CN
3
Area CO2 (473−773K)
Area CO2(298−773K)
To the reaction mixture as in Section 2.3.1, CH CN was added
Strong basic sites = [
Strong acid sites = [
] × Total CO2 Adsorbed (mL/g)
(3)
(4)
3
as specified in Table 2. The conversion of n-butanol was almost
doubled using an excess of CH3CN, but the reaction mixture was
much more complex as by-products are formed as discussed below.
Area NH3 (473−773K)
Area NH3 (298−773K)
] × Total NH3 Adsorbed (mL/g)
Please cite this article in press as: A. Angelini, et al., Synthesis of di-n-butyl carbonate from n-butanol: Comparison
of the direct carboxylation with butanolysis of urea by using recyclable heterogeneous catalysts, Catal. Today (2016),
http://dx.doi.org/10.1016/j.cattod.2016.02.005

G Model
CATTOD-10035; No. of Pages8
ARTICLE IN PRESS
A. Angelini et al. / Catalysis Today xxx (2016) xxx–xxx
3
Fig. 1. nBuOH conversion into D BC: dependence from time (a) and temperature (b).
n
2
.3.3. Direct carboxylation of n-butanol coupled to pervaporation
A Na-type pervaporation ceramic membrane (Sulzer) was
ethanol using the same operative conditions and the same catalyst
[20d]. Several factors could affect such result such as the solubility
of CO2 in the alcohol, the viscosity of the medium and, essentially,
the acidity of the alcohol.
located in a stainless steel housing covered with a heating ribbon
set at 403 K. A Speck pump was used to circulate the feed through
the membrane at about 300 mL/min. The permeate was collected
in a cold trap. The vacuum pump applied to the permeate produced
n
The dependence of the D BC yield on time and temperature
n
is reported in Fig. 1. The best yield of D BC is observed after 6 h.
−4
a final pressure of 10 MPa.
Increasing the temperature above 433 K, new products are formed.
At 453 and 483 K butyraldehyde and butyraldehyde dibutyl acetal
were formed together with dibutyl carbonate and the selectivity
toward the latter decreases from 100% at 433 K to 55% at 483 K.
This makes more complex the separation process and requires a
higher energy expenditure in post reaction treatment.
The catalytic reaction was carried out as reported in Section 2.3.
Then it was circulated inside the pervaporation equipment in order
to reduce the water content until 100–150 ppm. The dehydrated
retentate was used in a new carboxylation run.
2
.4. Butanolyisis of urea
The new species identified in the reaction mixture are products
of oxidation of butanol. The latter reaction occurs at expenses of the
mixed oxides (Scheme 1), which are reduced to a lower oxidation
state or even to metals, which do show a reduced catalytic activ-
ity. However, a second negative aspect linked to a higher reaction
temperature is the loss of the catalysts.
The best operative conditions in terms of yield and selectivity
are, thus, 433 K and 6 h. Other catalysts were tested in such condi-
tions, as reported in Table 1.
0.03Nb2O5/CeO2 was synthesized in two different ways (by co-
precipitation, Entry 1; or by HEM, Entry 2) in order to verify if the
synthetic methodology could affect the catalytic performance.
The two catalysts are almost equivalent, with a very slight higher
carbonate yield (0.32 vs 0.29%) obtained using 0.03Nb2O5/CeO2
synthesized by co-precipitation. The difference can be due to the
2.4.1. Butanolyisis of urea in a closed batch reactor
The reaction was carried out in a 100 mL stainless steel autoclave
connected to a pressure regulation valve through a ∼35 cm long
stainless steel tube used as refrigerant. In a typical run, 0.5 g of
urea were added to 7.6 mL of alcohol (molar ratio alcohol/urea = 10)
and 70 mg of catalyst (weight ratio urea/catalyst = 7). The reaction
was carried out for 7 h, or as specified. When the working pressure
reached 1 MPa, the valve was open and NH3 was allowed to escape
and was collected. As the reaction progressed, a lowering of the
pressure was observed due to the lower amount of NH3 released.
After the autoclave was cooled to room temperature, the reac-
n
tion mixture was analysed by GC and GC–MS. The yield in D BC is
expressed with respect to urea (the limiting reagent).
2
.4.2. Butanolyisis of urea with intermediate total elimination of
NH3
The reaction was performed as reported in Section 2.4.1 up to
h. The reactor was, then, cooled down to room temperature and
Table 1
DnBC synthesis by carboxylation of nBuOH using different mixed oxides. (Operative
conditions: BuOH = 4 mL, P = 5 MPa of CO2, T = 433 K, t = 6 h, 100 mg of catalyst).
7
Conv. of nBuOH into DnBC [%]
Entry
Catalyst
opened in order to allow a complete removal of NH . The reactor
was closed and heated again to 453 K and the reaction continued
for the specified time (see Fig. 6).
3
1
2
4
3
7
5
6
0.03Nb2O5/CeO2 coprecip
0.32
0.29
0.3
0
0.03
0
0.03Nb2O5/CeO2HEM
ZrO2/CeO2
0.5MgO/ZnO
ZrO2
3
. Results and discussion
La2O3
TiO2
0
.1. Synthesis of D BC via direct carboxylation of nbutanol
n
3
Our previous studies [20d] on the synthesis of diethyl carbon-
Table 2
Direct carboxylation of n-butanol catalysed by 0.03Nb2O5/CeO2 in presence of
CH3CN as dehydrating agent.
ate via the direct carboxylation of ethanol demonstrated the higher
catalytic activity of 0.03Nb O5/CeO with respect to both the pure
2
2
Conv. of nBuOH into D BC [%]
n
components and other mixed oxides reported in the literature.
.03Nb O5/CeO was used as catalyst in the carboxyation of n-
Entry
Reaction conditions
0
2
2
1
2
3
4
T = 433 K, 6 h
0.32
0.5
0.56
0.59
butanol starting with the same operative conditions used in the
case of ethanol (0.1 g of catalyst, 4 mL of ROH, 408 K, 3 h). It is worth
to note that the carboxylation of butanol is slower than that of
T = 433 K, 6 h, CH3CN:BuOH = 1:1
T = 433 K, 6 h, CH3CN:BuOH = 2:1
T = 433 K, 20 h, CH3CN:BuOH = 2:1
Please cite this article in press as: A. Angelini, et al., Synthesis of di-n-butyl carbonate from n-butanol: Comparison
of the direct carboxylation with butanolysis of urea by using recyclable heterogeneous catalysts, Catal. Today (2016),
http://dx.doi.org/10.1016/j.cattod.2016.02.005

G Model
CATTOD-10035; No. of Pages8
ARTICLE IN PRESS
4
A. Angelini et al. / Catalysis Today xxx (2016) xxx–xxx
O
H
C
2
H
C
2
oxidation
2BuOH
BuO
OBu
H
2
C
CH
CH2
CH3
C
H C
3
H C
2
OH
H C
H C
2
3
H
H C
2
Scheme 1. Formation of butyraldehyde and butyraldehyde dibutyl acetal from BuOH.
Table 3
higher surface area of particles formed using the coprecipitation
method with respect to HEM [37].
In order to try to increase the D BC yield, acetonitrile was added
as an organic water-trap to overcome the thermodynamic limita-
n
D BC synthesis by butanolysis of urea using different pure and mixed oxides. (Oper-
ative conditions: BuOH = 8 mL, T = 453 K, t = 7 h, 70 mg of catalyst).
n
n
Entry Catalyst
Strong basic/strong acid sites (nB/nA) D BC yield [%]
tions of the reaction. (Table 2) [21]. It is worth to note that CH CN,
1
2
3
4
5
6
7
8
ZnO
MgO
CaO
CeO2
2CaO/CeO2
0.5MgO/ZnO
MgO/ZnO
0.9
1.2
2.7
0.75 [20d]
2.47 [36]
1 [36]
0.8
9.1
7.5
7.4
3.6
10.0
8.8
3
while improving the conversion of butanol, causes the formation
of several products such as acetamide-AA (Eq. (5)), butyl acetate-
BA (Eq. (6)) and trace of butyl carbamate-BC (Eq. (7)) as revealed
by GC–MS analyses, making more difficult and energetically more
demanding the recovery of the target product.
3.5
5.7
0.03Nb2O5/CeO2 0.45 [20d]
Using an excess of acetonitrile for long time of reaction (Table 2,
n
Entry 4) the D BC yield was almost doubled with respect to the
case in which it was absent (Entry 1). But, concurrently, also the
formation of BA and BC was largely observed.
synthetic methodology. In Table 3, the activity of the tested cata-
lystsis reported and expressed as conversion of urea into D BC. The
n
corresponding pure oxides have been also tested for comparison.
Under the reaction conditions, similar to those used in the
ethanolysis of urea in our previous works [36,39], the most active
catalyst results to be 2CaO/CeO2 (Table 3, Entry 5), despite of the
(5)
(
(
6)
7)
n
Despite the positive effect on the formation of D BC, the fact that
the selectivity is lowered, does not make such methodology a win-
ning option [38] also because the hydrated form of the water trap
must be dehydrated and the water trap re-circulated, in order to
reduce waste formation. Therefore, we have used a pervaporation
membrane for the elimination of water. Under such circumstances
the conversion of butanol was raised up to 0.55% after only one
moderate activity of the relative pure oxides CaO and CeO2 (Entries
3 and 4).
Its activity is higher than that of ZnO. Table 4 compares the activ-
ity of most active catalysts in three consecutive runs. The slight
deactivation observed for 2CaO/CeO2 (Table 4, Entry 1) is due to a
only partial conversion of CaO into CaCO3.
“
carboxylation–pervaporation–carboxylation” cycle without pro-
This is confirmed by FTIR spectra (Fig. 2) of the solids recovered
−
1
duction of co-products. The membrane is enough selective toward
at the end of the reaction (a signal at 1754 cm due to carbonate
moiety is quite evident). By calcining the solid before re-use CaCO3
is converted back to CaO and the catalyst retains its performance
as already observed in the ethanolysis of urea [36].
n
water: 9.8% butanol and 3.9% D BC were lost with the permeate.
Moreover, the membrane does not need any activation after use.
Such technique seems to be more suited for boosting the produc-
n
tion of D BC from butanol and CO , than the addition of chemical
2
traps, although the total conversion remains modest.
Table 4
Study of recyclability of the catalyst. (Operative conditions: BuOH = 8 mL, T = 453 K,
t = 7 h, 70 mg of catalyst).
.2. Synthesis of D BC via nButanolysis of urea
n
3
n
n
n
Entry
Catalyst
D BC yield
D BC yield
D BC yield
The butanolysis of urea was performed by testing several dif-
I cycle [%]
II cycle [%]
III cycle [%]
ferent mixed oxides, which may be characterized by having either
intermediate acid-base properties with respect to single oxides or
even show quite different features. We have combined metals with
strong acid (Nb), basic (Ca, Mg) or intermediate acid-base (Zn, Ce)
properties in order to identify which of them were more suitable
for the butanolysis of urea. The recoverability and recyclability of
the most interesting catalysts were also studied as such proper-
ties have key importance for up-scaling and exploitation of the
a
1
2
3
4
5
2CaO/CeO2
10
8.9
18.4
7.4
9.2
13.8
7.5
5.6
8.4
9.4
12.9
a
ZnO
9.1
9.1
8.8
8.8
b
ZnOcalc
a
0.5MgO/ZnO
0.5MgO/ZnOcalc
b
a
The catalyst was separated from the liquid phase by centrifugation, dried under
vacuum and used in the following cycle.
b
The catalyst was calcined before its re-use in catalysis.
Please cite this article in press as: A. Angelini, et al., Synthesis of di-n-butyl carbonate from n-butanol: Comparison
of the direct carboxylation with butanolysis of urea by using recyclable heterogeneous catalysts, Catal. Today (2016),
http://dx.doi.org/10.1016/j.cattod.2016.02.005

G Model
CATTOD-10035; No. of Pages8
ARTICLE IN PRESS
A. Angelini et al. / Catalysis Today xxx (2016) xxx–xxx
5
Fig. 3. FTIR spectrum of ZnO after use in the first catalytic cycle: the band at
−
1
2
231 cm is attributed to bound NCO.
Fig. 2. FTIR spectra of 2CaO/CeO2 mixed oxide after three consecutive reaction
cycles.
Table 5
Leaching of the metal centre in three consecutive cycles of reaction.
Catalyst
CaO/CeO2
Leached metal
% Ca
I cycle
II cycle
III cycle
2
1.2
0.4
1.54
0.36
0.27
1.6
0.15
1.95
0.8
0.25
0.9
%
Ce
ZnO
.5MgO/ZnO
% Zn
%Zn
0
3.65
Under the operative conditions, the catalysts do not undergo
significant leaching, (Table 5) and reveal to behave as real hetero-
geneous catalysts.
0
.03Nb O5/CeO , the most acid catalyst among those reported
Fig. 4. Effect of reaction temperature on DnBC yield. (nBuOH/urea molar ratio = 10:
2
2
in Table 3 [20d], does not show a satisfying catalytic performance
Table 3, Entry 8), confirming the literature data that say that cat-
alysts with strong acidity, such as Al O3 and ZrO , are worst than
urea/catalyst ratio (w/w) = 7, time = 7 h, catalyst: 0.5MgOZnO).
(
2
2
basic catalysts such as CaO, MgO and La O [35] in the alcoholysis
2
3
of urea.
Using ZnO (Table 3, Entry 1) and 0.5MgO/ZnO mixed oxide
n
(
Table 3, Entry 6) almost the same D BC yield was obtained but the
recyclability of the two catalyst shows a different behavior. (Table 4,
Entries 2–5)
ZnO shows an irregular trend over three reaction cycles: its cat-
alytic activity increases after the first reaction cycle, and decreases
in the third one. (Table 4, Entry 2) If the solid is calcined at the end of
each cycle such irregular trend is no more observed (Table 4, Entry
3
).
Fig. 5. Effect of nBuOH/urea molar ratio on D BC yield (urea/catalyst ratio
(
n
As reported in the literature [39,40], ZnO is prone to form
w/w) = 7:1, T = 453 K, time = 7 h, catalyst: 0.5MgO/ZnO).
homogenous coordination complexes such as Zn(NCO) (NH3)2
2
with ammonia and isocyanate (Eq. (9)) formed upon decomposition
of urea at high temperature (above 435 K, Eq. (8)). The homoge-
neous catalyst has been shown to be the active species and can be
lost with the liquid phase in the case of ethanol [40c] in which it is
soluble.
be converted in other species not yet identified but not active in
catalysis.
In order to fully elucidate the behavior of ZnO, we have carried
out the butanolysis of urea in presence of fresh ZnO for different
reaction times, longer than 7 h: a higher yield of carbonate was
(
H N) CO → NH + HNCO
(8)
(9)
2
2
3
n
found (D BC yield about 23.2%) after 14 h.
0
.5MgO/ZnO, differently from all the other mixed oxides, retains
ZnO + 2HNCO + 2NH → Zn(NCO) (NH ) + H O
3
2
3 2
2
about the same catalytic activity in three consecutive cycles even
if the catalyst is re-used without calcination. (Table 4, Entries 4–5)
This suggests that the nature of the catalyst does not change dur-
ing catalysis. It was, then, used in the investigation of the effect of
further reaction parameters.
Conversely, in the butanolysis of urea Zn(NCO) (NH ) is
2
3 2
formed, as demonstrated by FTIR spectrum (Fig. 3), but it shows
a low solubility in the reaction medium as confirmed by the EDX
data (Table 5). Therefore, because the active species is not lost in the
liquid phase and remains in the solid separated from the reaction
mixture, it is prompt to act as catalyst in the second run with-
out any induction time, producing a higher carbonate yield. The
reason of the deactivation of the catalyst in the third cycle is still
under investigation. FTIR spectra suggest that Zn(NCO) (NH ) can
The results of a study on the influence of the temperature on
the final yield is shown in Fig. 4. Increasing the temperature up to
n
497 K, the D BC yield increases.
At higher reaction temperature, the carbonate yield reaches
really interesting values (32.1 and 36.2% at 473 and 497 K respec-
2
3 2
Please cite this article in press as: A. Angelini, et al., Synthesis of di-n-butyl carbonate from n-butanol: Comparison
of the direct carboxylation with butanolysis of urea by using recyclable heterogeneous catalysts, Catal. Today (2016),
http://dx.doi.org/10.1016/j.cattod.2016.02.005

G Model
CATTOD-10035; No. of Pages8
ARTICLE IN PRESS
A. Angelini et al. / Catalysis Today xxx (2016) xxx–xxx
6
n
Fig. 6. Trend of D BC yield (%) vs reaction time (h) in presence of 0.5MgO/ZnO mixed oxide as catalyst using different reaction conditions. (Urea/catalyst w/w ratio = 7:1,
T = 453 k).
and 34.5% after 14 and 24 h respectively, since removing ammonia
from the reaction medium allows to shift the equilibrium to right
[
36].
In conclusion, the best conversion yield of urea (>35%) is
observed at 497 K, working with a BuOH/urea molar ratio equal
to 10, and carrying out the reaction for 7 h, upon continuous NH3
removal by bubbling N2 with strict butanol recovery [41].
3.3. Comparison between alcoholysis of urea and direct
carboxylation of n-butanol
By comparing the results obtained with the butanolysis of urea
and the direct carboxylation of butanol for the synthesis of n-butyl
carbonate two important aspects can be highlighted:
n
1
) The two routes afford quite different yields of D BC, the butanol-
ysis of urea being more efficient. Even using a dehydrating agent,
the yield in carbonate obtained in the direct carboxylation pro-
cess is close to two orders of magnitude lower than that obtained
through the alcoholysis of urea. This can be explained by the
higher stability and lower reactivity of the CO2 molecule with
respect to urea, and the relevant thermodynamic constraints.
) The same catalysts show completely different activity in the two
processes. In fact, the most active catalyst for the butanolysis of
urea, namely 0.5MgO/ZnO, is not active in the direct carboxyla-
Fig. 7. NH3-TPD profiles of 0.5MgO/ZnO and 0.03Nb2O5/CeO2 catalysts [20d,36].
tively) but increasing amounts of by-products are formed. GC–MS
analysis on the liquid phase reveals the presence of butyraldehyde
dibutyl acetal and N-butyl butylcarbamate (N-BBC). The first could
be formed from the oxidation of butanol as also observed in the
direct carboxylation. N-Butyl butylcarbamate (N-BBC) comes from
the akylation of butylcarbamate by butylcarbonate. Therefore, the
increased conversion of butanol is accompanied by a loss of selec-
tivity, not a good feature, and decomposition of catalysts.
2
tion of butanol, in which catalysts like 0.03Nb O5/CeO perform
2
2
much better. It is worth to underline that the latter shows a
poor activity in the alcoholysis of urea. The explanation of such
behavior can be found in the different reaction mechanisms of
the two processes. (Scheme 2).
n
The effect of butanol/urea initial molar ratio on D BC yield is
n
shown in Fig. 5. Increasing the BuOH/urea molar ratio from 5:1 to
n
1
0:1 there is an increase in D BC yield from 4 to 8.8%. However a
further increase of the alcohol does not produce a further significant
Looking at the reaction mechanisms reported in the literature
increase of the carbonate yield which is 8.8 and 10.9% at 10:1 and
[20c,d,42], it is possible to note that the role of the basic sites is the
same in the two processes: they interact with the alcohol molecule
to form the alcoholate moiety. The main difference is in the role of
the acid sites since in pathway A they interact with alcohol to form
the alkyl cation, whereas in path B they coordinate carbamate to
favor the nucleophylic attack of the alcoholate.
n
1
5:1 butanol/urea molar ratio, respectively. So the molar ratio 10:1
seems to be the best to use.
A kinetic study of the reaction catalysed by 0.5MgO/ZnO (Fig. 6)
carried out in the best operative conditions in a closed vessel shows
that the rate of carbonate production increases quite linearly up
n
to 7 h and then reaches a pseudo-equilibrium (D BC yield = 10%
If NH -TPD profiles of 0.03Nb O5/CeO and 0.5MgO/ZnO are
3
2
2
conversion). A different trend is observed when the reaction is car-
ried out with a more efficient NH3 elimination: in such conditions
higher D BC yields can be reached at longer reaction times: 32.0%
compared [20c,36], it is possible to observed that the former has
a larger number of acid sites (signal at 450 and 555 K) with respect
to the latter that shows only a small amount of strong acid sites
n
Please cite this article in press as: A. Angelini, et al., Synthesis of di-n-butyl carbonate from n-butanol: Comparison
of the direct carboxylation with butanolysis of urea by using recyclable heterogeneous catalysts, Catal. Today (2016),
http://dx.doi.org/10.1016/j.cattod.2016.02.005

G Model
CATTOD-10035; No. of Pages8
ARTICLE IN PRESS
A. Angelini et al. / Catalysis Today xxx (2016) xxx–xxx
7
Scheme 2. Putative reaction pathway for (A) direct carboxylation of alcohols, (B) alcoholysis of urea.
(
at 485 K). Such different acid properties agree with the differ-
[8] X. Feng, Q. Wang, G. Wang, F. Qiu, Chin. J. Catal. 27 (3) (2006) 195–196.
9] S. Han, M. Luo, Z. Wu, CN 102351709 (2012).
10] Yong-Xin, CN 1569811 A (2005).
[
ent behavior of the two catalysts in the direct carboxylation and
butanolysis of urea.
[
[
[
[
[
[
11] Z. Li, K. Su, Y. Diao, B. Cheng, CN 104030885 (2014).
12] D. Tao, Z. Yang, F. Ouyang, S. Xu, Z. Li, CN 103709032 (2014).
13] D. Srinivas, R. Srivastava, P. Ratnasamy, US 20070083062 A1 (2007).
14] H. Tian, Y. Zhu, Y. Hao, H. Wang, J. Liu, H. Huang, CN1699328 (2005).
15] G. Wang, X. Feng, Q. Wang, Y. Wang, Y. Zeng, CN 1775346 (2006).
[16] Y. Li, C. Xu, X. Chen, CN 1569811 (2005).
17] X. Gui, J. Shen, Spec. Petrochem. 28 (3) (2011) 13–17.
4
. Conclusions
Di-n-butyl carbonate (D BC) was synthesized reacting butanol
n
n
[
[
with either CO2 or urea. Catalysts with different acid/basic proper-
ties are required in the two synthetic methodologies.
18] F. Ouyang, Z.-Z. Wang, Y. Zhou, Z. Cheng, Z.-H. Lu, Z. Yang, D.-J. Tao, App. Catal.
A: Gen. 492 (2015) 177–183.
The direct carboxylation of n-butanol is favored by more acid
[19] (a) A. Dibenedetto, A. Angelini, Synthesis of organic carbonates, in: R. van
Eldik, M. Aresta (Eds.), Adv. in Inorg. Chem., vol. 66, Academic Press, 2013, pp.
catalysts such as 0.03Nb O5/CeO2 and by water elimination. How-
2
25–81;
ever, even if it is the most interesting route for the synthesis of
carbonates thanks to the simplicity of the process, the obtainable
D BC yields are always lower than that obtained by the butanolysis
(
b) A. Dibenedetto, A. Angelini, P. Stufano, J. Chem. Technol. Biotechnol. 89 (3)
(2014) 334–353.
n
[20] (a) M. Aresta, A. Dibenedetto, C. Pastore, I. Papai, G. Schubert, Top. Catal. 40
(
2006) 71–81;
of urea, an indirect route to use CO2.
(b) M. Aresta, A. Dibenedetto, C. Pastore, C.B. Corrado, B.A. Aresta, S. Cometa,
E. De Giglio, Catal. Today 137 (2008) 125–131;
On the other hand, basic catalysts are required for the butanol-
(
c) M. Aresta, A. Dibenedetto, C. Pastore, A. Angelini, B.M. Aresta, I. Papai, J.
Catal. 269 (2010) 44–52;
d) A. Dibenedetto, M. Aresta, A. Angelini, J. Ethiraj, B.M. Aresta, Chem. Eur. J.
ysis of urea. 2CaO/CeO2 results to be the most active catalyst in the
first cycle of reaction, whereas the 0.5MgO/ZnO is the most robust
since its catalytic activity is approximately the same during three
consecutive catalytic tests both when the catalyst is re-used such
as in a next run, or it is calcined before re-use: such mixed-oxide
can be easily recycled.
(
18 (2012) 10324–10334.
[
21] (a) M. Honda, M. Tamura, Y. Nakagawa, K. Nakao, K. Suzuki, K. Tomishige, J.
Catal. 318 (2014) 95–107;
(
b) M. b), A. Honda, B. Suzuki, K. Noorjahan, K. Fujimoto, K. Suzuki, Chem.
Commun. (2009) 4596–4598;
Several reaction parameters were investigated using
.5MgO/ZnO. It was found that a smaller catalyst amount can
(c) M. Honda, S. Kuno, N. Begum, K. Fujimoto, K. Suzuki, Y. Nakagawa, K.
Tomishige, App. Catal. A: Gen. 384 (2010) 165–170.
22] D.T. Nowlan, C.V. McNeff, US Patent US 0288309 A1 (2011).
23] M. Aresta, A. Dibenedetto, A. Angelini, J. CO2 Util. 3–4 (2013) 65–73.
0
[
[
ensure a better catalytic performance most probably because the
active metal center are employed in the carbonate formation and
side reactions are prevented.
[24] M. Aresta, A. Dibenedetto, I. Tommasi, Energy Fuels 15 (2001) 269–273.
[
[
[
25] M. Aresta, A. Dibenedetto, A. Angelini, Comprehensive Inorganic Chemistry II,
in: Jan Reedijk, Kenneth Poeppelmeier (Eds.), From Carbon Dioxide to
Valuable Products under Homogeneous Catalysis, vol. 6, 2015, pp. 563–586
(Chapter 6.19).
26] (a) M. Honda, M. Tamura, Y. Nakagawa, S. Sonehara, K. Suzuki, K. Fujimoto, K.
Tomishige, ChemSusChem 6 (8) (2013) 1341–1344;
n
A higher reaction temperature leads to an increase in D BC yield
although it consistently affects the selectivity toward the carbon-
ate.
Using a low BuOH/urea initial molar ratio a lower carbonate
concentration is obtained, while increasing the molar ratio higher
yields are obtained. The most suited ratio is 10.
(
b) M. Honda, S. Kuno, S. Sonehara, K.I. Fulimoto, K. Suzuki, Y. Nakagawa, K.
Tomishige, ChemCatChem 3 (2011) 365–370.
27] (a) T. Sakakura, J.C. Choi, H. Yasuda, Chem. Rev. 107 (2007) 2365–2387;
(
6
b) T. Sakakura, Y. Saito, J.-C. Choi, T. Masuda, T. Sako, T. Oriyama, J. Org. Chem.
4 (1999) 4506–4508;
(c) T. Sakakura, Y. Saito, M. Okano, J.C. Choi, T. Sako, J. Org. Chem. 63 (1998)
095–7096.
n
The profile of the reaction kinetics points out that the D BC yield
tends to an equilibrium value at longer reaction time (24 h).
7
[
28] (a) M. Aresta, A. Dibenedetto, E. Fracchiolla, P. Giannoccaro, C. Pastore, I.
Papai, G. Schubert, J. Org. Chem. 70 (16) (2005) 6177–6186;
Acknowledgements
M. Aresta, A. Dibenedetto, P. Stufano, Patent MI2009A001221 (2009);
(
2
b) M. Aresta, A. Dibenedetto, A. Angelini, I. Papai, Top. Catal. 58 (1) (2015)
–14.
29] Y. Deng, X. Ma, Y. Tian, L. He, L. Lu, X. Ma, CN 102557948 (2010).
One of us (SF) thanks CIRC srl for a grant. AA thanks Total for
partial support. VALBIOR-Apulia Region is thanked for the use of
some equipment.
[
[30] Y. Deng, Y. Ma, F. Shi, J. Qi, T. He, X. Ma, L. Lu, CN 102093221 (2011).
31] Y. Deng, L. Wang, X. Tian, Y. Wang, S. Liu, L. Lu, Y. He, X. Ma, CN 102464588
2012).
[32] M. Mizukami, Y. Arai, H. Harada, T. Ohshida, H. Ohgi, US 6031122 (2000).
[
(
References
[
[
[
33] R. Saley, R. Michaelson, E.N. Suciu, B. Kuhlmann, US 5565603 (1996).
34] F.J. Sowa, N.J. Cranford, US 2834799 (1958).
35] (a) M. Wang, H. Wang, N. Zhao, W. Wei, Y. Sun, Cat. Commun. 7 (2006) 6–10;
[
1] (a) G.S. Abbas, S. Swaminathan, Chem. Rev. 96 (1996) 951–976;
b) W.C. b), N. Peng, F.K. Zhao, W. Xiao, Y.H. Sun, Pure Appl. Chem. 84 (2011)
03–620;
c) S. Huang, B. Yan, S. Wang, X. Ma, Chem. Soc. Rev. 44 (2015) 3079–3116.
(
6
(
(
(
(
4
(
b) D. Wang, B. Yan, X. Zhai, L. Zhou, Fuel Proc. Technol. 88 (2007) 807–812;
c) Q. Li, W. Zhang, N. Zhao, W. Wei, Y. Sun, Catal. Today 115 (2006) 111–116;
d) S. Xin, L. Wang, H. n Li, K. Huang, F. Li, Fuel Proc. Technol. 126 (2014)
[2] R.L. Conn, J.C. Williams, J.E.C. Willham, EP 1472088 EP (2006).
[3] P. Wang, S. Liu, L. Lu, X. Ma, Y. He, Y. Deng, RSC Adv. 5 (2015) 62110–62115.
[4] (b) M.A. Pacheco, C.L. Marshal, Energy Fuels 11 (1997) 2–29;
H.C. Stevens, US 2787631 A (1957).
53–459;
e) W. Joe, H.J. Lee, U.G. Hong, Y.S. Ahn, C.J. Song, B.J. Kwon, I.K. Song, J. Ind.
Eng. Chem. 18 (2012) 1730–1735.
[
36] A. Angelini, A. Dibenedetto, D. Curulla-Ferre, M. Aresta, RSC Adv. 5 (2015)
[5] C. Min-Sheng, T. Chia-Jung, C. Chih-Wei, US 2008/0300419Al (2008).
[6] L. Dinghua, L. Xiaoqin, C. Zhensong, CN102302950 (2012).
[7] K. Takeshi, T. Masao, Y. Nobuhiro, M. Kenji, EP528498 (1993).
8
8401–88408.
Please cite this article in press as: A. Angelini, et al., Synthesis of di-n-butyl carbonate from n-butanol: Comparison
of the direct carboxylation with butanolysis of urea by using recyclable heterogeneous catalysts, Catal. Today (2016),
http://dx.doi.org/10.1016/j.cattod.2016.02.005

G Model
CATTOD-10035; No. of Pages8
ARTICLE IN PRESS
A. Angelini et al. / Catalysis Today xxx (2016) xxx–xxx
8
[37] (a) J.-S. Lee, S.-C. Choi, J. Europ. Ceram. Soc. 25 (14) (2005) 3307–3314;
(2010) 381–389;
(
b) O.A. Kirichenko, G.W. Graham, W. Chun, R.W. McCabe, Stud. Surf. Sci.
(c) W. Zhao, W. Peng, D. Wang, N. Zhao, J. Li, F. Xiao, W. Wei, Y. Sun, Catal.
Commun. (2009) 655–658.
Catal. 118 (1998) 411–420.
[
[
38] M. Aresta, ICCDU 2015, Singapore, July 5–9, 2015, PL 6.
39] A. Dibenedetto, A. Angelini, M. Aresta, S. Fasciano, M.E. Cucciolito, F. Ruffo,
B.M. Aresta, D. Curulla Ferré, E. De Giglio, App. Catal. A: Gen. 493 (2015) 1–7.
40] (a) J. Zhang, F. Wang, W. Wei, F. Xiao, Y. Sun, Korean J. Chem. Eng. 27 (6)
[41] M. Aresta, A. Dibenedetto, F. Nocito, C. Ferragina, J. Catal. 268 (1) (2009)
106–114.
[42] A. Dibenedetto, A. Angelini, S. Fasciano, I. Papai, D. Curulla-Ferre ı´ , M. Aresta, J.
CO2 Util. 8 (2014) 27–33.
[
(
2010) 1744–1749;
(
b) H. Wang, M. Wang, W. Zhao, W. Wei, Y. Sun, React. Kinet. Mech. Cat. 99
Please cite this article in press as: A. Angelini, et al., Synthesis of di-n-butyl carbonate from n-butanol: Comparison
of the direct carboxylation with butanolysis of urea by using recyclable heterogeneous catalysts, Catal. Today (2016),
http://dx.doi.org/10.1016/j.cattod.2016.02.005