The effect of alkoxide ionic liquids on the synthesis of dimethyl carbonate from CO2 and methanol over ZrO2-MgO

Article abstract of DOI:10.1007/s10562-011-0666-3

The use of carbon dioxide in the synthesis of chemicals, such as dimethyl carbonate (DMC), constitutes an environmentally attractive alternative to hazardous and toxic reagents. However, the direct synthesis of DMC from methanol and CO₂ is characterized by low yields due to the reaction equilibrium and the thermodynamic limitations (ΔG_298K ⁰ = +26.3 kj/mol). Alkoxide ionic liquids possessing alkylimidazolium and benzalkonium cations were prepared, characterised and tested together with ZrO₂-MgO catalyst for the synthesis of DMC from methanol and CO ₂. By using the novel ionic liquid as additives, ca. 12% conversion of methanol, and 90% selectivity to DMC was obtained at 120 °C and 7.5 MPa. The water abstracting potential of the ionic liquids influenced the conversion of methanol and the selectivity to DMC. The alkoxide ionic liquids were recovered and reused in DMC synthesis without loss in activity and selectivity.

Full text of DOI:10.1007/s10562-011-0666-3

Catal Lett (2011) 141:1254–1261
DOI 10.1007/s10562-011-0666-3
The Effect of Alkoxide Ionic Liquids on the Synthesis of Dimethyl
Carbonate from CO2 and Methanol over ZrO2–MgO
•
Valerie Eta Paivi Maki-Arvela Eero Salminen
•
•
¨
Tapio Salmi Dmitry Yu. Murzin Jyri-Pekka Mikkola
¨
•
•
Received: 2 June 2011 / Accepted: 7 July 2011 / Published online: 23 July 2011
Ó Springer Science+Business Media, LLC 2011
Abstract The use of carbon dioxide in the synthesis of
chemicals, such as dimethyl carbonate (DMC), constitutes
an environmentally attractive alternative to hazardous and
toxic reagents. However, the direct synthesis of DMC from
methanol and CO2 is characterized by low yields due to the
reaction equilibrium and the thermodynamic limitations
1 Introduction
Recently, dimethyl carbonate (DMC) has attracted atten-
tion as an important intermediate for polycarbonate resins
and a useful carbonylating and methylating agent due to its
versatile chemical reactivity and unique physical properties
[1, 2]. It is an environmentally benign building block and a
promising substitute for phosgene, dimethyl sulphate and
methyl iodide in organic syntheses due to its negligible
toxicity and rapid biodegradability. DMC is also consid-
ered as a potential candidate for meeting the oxygenate
specifications set for transportation fuels [3].
À
Á
DGo298K ¼ þ26:3 kJ=mol . Alkoxide ionic liquids pos-
sessing alkylimidazolium and benzalkonium cations were
prepared, characterised and tested together with ZrO2–
MgO catalyst for the synthesis of DMC from methanol and
CO2. By using the novel ionic liquid as additives, ca. 12%
conversion of methanol, and 90% selectivity to DMC was
obtained at 120 °C and 7.5 MPa. The water abstracting
potential of the ionic liquids influenced the conversion of
methanol and the selectivity to DMC. The alkoxide ionic
liquids were recovered and reused in DMC synthesis
without loss in activity and selectivity.
DMC is currently synthesised via the oxidative car-
bonylation or phosgenation of methanol [4–6]. Although
these processes are fully developed and highly efficient, the
use of toxic and corrosive reagents and the generation of
wastes are important setbacks requiring alternative tech-
niques that are environmentally friendly. CO2 can be
captured from various sources e.g. from fossil fuel-fired
plants [7] and used as a reactant and/or reaction medium
[8] for chemical syntheses. Consequently, the utilisation of
CO2 in the synthesis of DMC (Eq. 1) has attracted
academic interest for the following reasons: (i) replacement
of multistep processes [9] with more direct synthetic pro-
cedures and waste reduction, leading to a high atom
economy, (ii) implementation of alternative techniques to
processes currently based on toxic or expensive materials,
(iii) raw material diversification and, (iv) carbon recycling
[10]. The implementation of synthetic methodologies based
on CO2 could considerably contribute to the reduction of
CO2 emissions. Although the utilization of CO2 as a
feedstock in chemical syntheses will have only a minor
impact on the mitigation of greenhouse gases, at least in
short term, it forms an integral part of carbon management
by providing a strategic recycling pathway. Therefore,
Keywords Alkoxide ionic liquid ꢀ Dimethyl carbonate ꢀ
Carbon dioxide ꢀ Methanol ꢀ ZrO2–MgO
¨
V. Eta ꢀ P. Maki-Arvela ꢀ E. Salminen ꢀ T. Salmi ꢀ
D. Yu. Murzin (&) ꢀ J.-P. Mikkola
Laboratory of Industrial Chemistry and Reaction Engineering,
˚
Process Chemistry Centre, Abo Akademi University,
˚
Biskopsgatan 8, 20500 Turku/Abo, Finland
e-mail: dmurzin@abo.fi
J.-P. Mikkola
Technical Chemistry, Department of Chemistry,
˚
Chemical–Biological Center, Umea University,
˚
90187 Umea, Sweden
123

Effect of Alkoxide Ionic Liquids on the Synthesis of DMC
1255
research on the direct synthesis of DMC from methanol
and CO2 has been attempted, in order to replace toxic
carbon monoxide and phosgene representing the current
industrial state-of-the-art.
methyl iodide and [C2-mim][Br] affording 11% yield of
DMC [30]. However, more research is required in order to
understand the reaction pathway and the role of methyl
iodide or other halogens in the reaction.
The removal of water from the reaction mixture is
necessary in order to obtain high yields of DMC via
relaxing thermodynamic restrictions and shifting of equi-
librium towards DMC formation. Herein, novel ionic liq-
uids containing the methoxide species as anions and
imidazolium or benzalkonium cations were prepared,
characterised and tested together with ZrO2–MgO as a
catalyst in the direct synthesis of dimethyl carbonate from
methanol and carbon dioxide. The effect of the ionic liq-
uids in water trapping and on the yield and selectivity to
DMC is presented. The performances of different alkoxide
ionic liquids in the synthesis of DMC are also compared.
Furthermore, the recovery and reuse of the alkoxide ionic
liquids in DMC synthesis was successfully demonstrated.
2CH3OH þ CO2 ꢀ ðCH3O)2CO þ H2O
ð1Þ
Typically, DMC is synthesised from methanol and CO2
in an autoclave at supercritical conditions (e.g. 180 °C,
20 MPa) catalysed by supported copper catalysts [11], Sn
(IV) and Ti(IV) alkoxides and organotin compounds
[12, 13], alkaline metal carbonates [14], Cu supported on
MgO–SiO2 [15], ZrO2 or modified ZrO2 [16–18]. Although
more than 90% selectivity to DMC was obtained under the
reaction conditions reported, the yields of DMC were
nevertheless lower than 0.5 mol% since the equilibrium of
the reaction is largely shifted towards the reactants
[16, 19]. In addition, the thermodynamic limitations
embedded in the direct reaction of CO2 and methanol,
the inertness of CO2, the hydrolysis of the DMC and the
deactivation of the catalyst are important factors
contributing to the low yields of DMC [19].
2 Experimental
Various additives e.g. 2,2-dimethylpropane [20] or
benzonitrile [21] have been used in order to improve the
yield of DMC beyond the equilibrium level resulting in a
13-fold increase in the amount of DMC synthesised.
Another technique involves a two-step reaction, performed
either as a one-pot reaction or as a set of separate reactions
using basic metal oxide catalysts. In the first step, ethylene
oxide or propylene oxide reacts with CO2 to form the
corresponding cyclic carbonate. The transesterification of
the cyclic carbonate with methanol then results in DMC
[20–24]. In our previous work, enhanced yields corre-
sponding to 7.2 mol % DMC were reported using methanol
and CO2 in the presence of butylene oxide, at 150 °C and
9.5 MPa [25]. However, only 53% selectivity to DMC was
obtained due to the interaction of butylene oxide, used as a
chemical water abstracting additive, with CO2 and
methanol.
2.1 Materials and Methods
1-methylimidazole, 1-chlorobutane, sodium methoxide,
benzylalkonium chloride (benzyl-dimethyl-tridecyl-aza-
nium chloride [BA][Cl]), methanol, toluene, zirconyl
nitrate, aqueous ammonium hydroxide (28%), butylene
oxide, dimethyl carbonate, magnesium nitrate, ammonium
hydroxide (60%), were purchased from Sigma Aldrich with
purities C99%. CO2 (99.99%) was purchased from AGA.
All chemicals were used as received. 1H NMR spectra were
recorded at 25 °C with a Bruker Avance 600 MHz spec-
trometer using DMSO-d6 as a solvent. 13C NMR was
measured using Bruker Avance 250 MHz. Thermal
decomposition temperatures were measured using TGA
(Cahn D-200) under N2 atmosphere. Samples between 80
and 100 mg were heated from 25 to 400 °C under constant
heating at 5 °C/min. Decomposition temperatures were
determined from the onset mass loss (5% decomposition)
of the sample.
DMC has also been synthesised using room temperature
ionic liquids (RTILs). RTILs have demonstrated excellent
performance as catalysts or as reaction media in catalytic
reactions [26–28] for example, DMC has been synthesised
by the oxidative carbonylation of methanol at 120 °C and
2.4 MPa using Cu salts as catalysts and ionic liquids as
promoters, affording 17% conversion of methanol and
97.8% selectivity to DMC [26]. In addition, the transeste-
rification of propylene and ethylene carbonate at 140 °C,
occurred in the presence of [C2-mim][Cl] as a catalyst,
resulting in 15 and 75% yield, respectively [27, 29]. In the
transesterification, however, the low atom economy (59
wt%) and the use of toxic cyclic ethers make this process
unsatisfactory. As a result, the direct reaction of CO2 and
methanol has been performed at 7.4 MPa in the presence of
2.2 Preparation and Characterisation of Catalyst
The catalyst was synthesised by the co-precipitation of a
0.5 M solution of zirconyl nitrate and a 0.3 M solution of
magnesium nitrate hydrate in a 500 ml beaker by the drop-
wise addition of ammonium hydroxide under stirring until
the pH 9 was achieved. The precipitate obtained was aged
for 3 days while maintaining the pH at 9 by constant
addition of appropriate amounts of NH4OH. The precipi-
tate was washed several times with distilled water to
remove the unreacted NH4OH. The resulting cake was
123

1256
V. Eta et al.
dried at 80 °C and sieved in a 100 lm mesh. The powder
was calcined at 700 °C for 3 h in a muffle furnace to obtain
ZrO2–MgO. The specific surface area of the catalysts was
studied by measuring the nitrogen adsorption–desorption
isotherms at 77 K (Sorptomatic 1900) in liquid nitrogen.
Prior to measurements, the samples were outgassed for 3 h
at 150 °C to a residual pressure below 0.01 Pa. The total
surface area calculated according to the Brunauer–
Emmett–Teller (BET) [31] equation. The structural prop-
erties of the fresh and used catalysts were determined by
X-ray diffraction (XRD, Siemens D5000). SEM/EDX and
the quantitative spot analysis of the samples were measured
using the SEM/EDX (Oxford Instruments) at 20 keV for
7 s with a take-off angle of 36° and a dead time of 15.5 s.
18.8, 31.8, 36.3, 45.8, 47.5, 121.7, 122.9, 137.7; Analysis
C9H18N2O Calcd (%): C 63.5, H 10.6, N 16.4; found: C
63.2, H 10.5, N 16.6; Td5% = 191 °C.
2.3.3 [BA][MeO]
Benzylalkonium methoxide was synthesized by dissolving
0.03 mol of benzylalkonium chloride in 40 mL distilled
water and adding 0.03 mol of HNO3. The solution was
stirred for 1 h and chloroform was added. The organic
phase of the solution was washed with distilled water and
the chloroform removed in vacuo at 50 °C to obtain ben-
zylalkonium nitrate. A solution of sodium methoxide in
methanol was then reacted with benzylalkonium nitrate for
4 h to obtain benzylalkonium methoxide. 1H NMR
(600 MHz, DMSO-d6) d 7.68 (2H, m, J = 7.9 Hz), 7.52
(2H, m, J = 7.2 Hz), 7.35 (1H, m, J = 7.2 Hz), 4.58 (2H,
s), 3.29 (2H, t, J = 7.7 Hz), 3.07 (3H, s), 2.98 (6H, m),
1.88 (2H, t, J = 7.6 Hz), 1.34 (2H, t, J = 6.6 Hz), 1.25
(12H, m), 0.87 (3H, t, J = 7.0 Hz); 13C NMR (250 MHz,
DMSO-d6) d 13.6, 20.9, 21.7, 25.6, 5(28.5), 30.7, 49.5,
63.4, 64.5, 121.7, 129.1, 129.4, 130.5, 133.7; Analysis
C19H35NO Calcd (%): C 77.8, H 11.9, N 4.8; found: C
77.4, H 11.2, N 5.2; Td5% = 185 °C.
2.3 Preparation and Characterisation of ILs
Alkyl halide (0.15 mol) was added into a vigorously stir-
ring solution of 1-methylimidazole (0.15 mol) in toluene
(50 ml) at 0 °C and the solution was stirred at ca. 25 °C for
48 h. The mixture was placed in a freezer for 4 h to obtain
a solid or viscous liquid in toluene. The supernatant solvent
was decanted and the remaining solid or viscous oil was
re-crystallized using acetonitrile and dried in vacuo to
obtain 1-alkyl-3-methyl imidazolium halide in approxi-
mately 80% yield. Sodium methoxide (0.08 mol) in meth-
anol was mixed with dialkyl-imidazolium halide (0.1 mol)
in batches and allowed to react for 24 h to obtain a
yellowish solution. The precipitated sodium halide was
filtered from the solution. The filtrate was dried in vacuo at
70 °C for 24 h to obtain 1-alkyl-3-methylimidazolium
methoxide. The structure of the ionic liquids and comple-
2.3.4 Recovery of Ionic Liquid
The ionic liquid was recovered by filtration and washing of
the catalyst with dichloromethane. The solvent, methanol
and the reaction products formed were evaporated under
vacuum at 70 °C for 8 h. 1H NMR (600 MHz DMSO-d6) d
8.31(1H, s), 7.54 (1H, m), 7.48 (1H, m), 4.04 (2H, t,
1
tion of the reaction was confirmed by H NMR. The Na
J = 7.2 Hz), 3.77 (2H, m), 1.20 (2H, m), 0.82 (3H, m); 13
C
content of the alkoxide ILs determined by ICP was less
than 0.1 wt% while the amount of water determined by
Karl–Fisher titration was \250 ppm.
NMR (250 MHz, DMSO-d6) d 12.7, 31.7, 36.1, 45.9, 47.4,
120.1, 122.5, 136.9; Analysis C8H16N2O Calcd (%): C
61.5, H 10.2, N 17.9; found: C 62.1, H 10.8, N 17.4;
Td5% = 180 °C.
The recovered 1-alkyl-3-methylimidazolium hydroxide
was rejuvenated by addition of equimolar amounts of
sodium methoxide, to regenerate the 1-alkyl-3-methylimi-
dazolium methoxide.
2.3.1 [C2-mim][MeO]
1H NMR (600 MHz, DMSO-d6) d 8.32 (1H, s), 7.55 (1H,
m), 7.47 (1H, m), 4.11 (2H, m), 3.77 (3H, s), 1.37 (3H, t,
J = 7.4 Hz), 1.14 (3H, s); 13C NMR (250 MHz, DMSO-d6)
d 16.5, 36.1, 44.3, 46.1, 122.7, 123.8, 140.0; Analysis
C7H14N2O Calcd (%): C 59.2, H 9.8, N 19.7; found: C 58.9,
H 10.2, N 19.5; Td5% = 210 °C.
2.4 Catalytic Tests
The experiments were carried out in a laboratory scale
autoclave (Parr Inc.) with an inner volume of 300 mL
equipped with a stirrer and an electric heater. In a typical
sequence, 0.5 g of catalyst and 4 g of ionic liquid were
introduced into the reactor and methanol (247 mmol,
10 mL) was added. The content was purged with argon and
pressurised with CO2 to ca. 4.5 MPa. The reactor was
heated to the required temperature and stirred at 700 rpm
for 9 h. The reactor was cooled to 3 8C before venting out
2.3.2 [C4-mim][MeO]
1H NMR (600 MHz, DMSO-d6) d 8.32 (1H, s), 7.90 (1H,
d, J = 6.8 Hz), 7.48 (1H, d, J = 1.8), 4.16 (2H, m,
J = 7.3 Hz), 4.05 (3H, s), 1.34 (2H, m, J = 6.9 Hz), 1.12
(2H, m, J = 6.9 Hz), 0.85 (3H, s), 0.81 (3H, t,
J = 7.1 Hz); 13C NMR (250 MHz, DMSO-d6) d 12.9,
123

Effect of Alkoxide Ionic Liquids on the Synthesis of DMC
1257
Fig. 1 SEM images showing MgO crystallites on the surface of ZrO2. The molar ratio of Zr to Mg = 15. a Fresh ZrO2–MgO (5,0009) and
b catalyst after the reaction (5009)
excess CO2. During recycling of the ionic liquid, the mass
ratio of the methanol and the ionic liquid was kept constant
(e.g. 197 mmol methanol, 3.2 g IL, 0.8 g catalyst) and
fresh ZrO2–MgO was used. The product in the liquid phase
was analysed by gas chromatography (Agilent Technolo-
gies, 6890 N) using a 60 m 9 320 lm 9 0.5 lm capillary
column (J&W 123-506 E DB5) equipped with a flame
ionisation detector, using toluene as an internal standard.
The products were identified by GC–MS (Agilent Tech-
nologies, 6890N).
fumed in moist air with a characteristic smell. Similarly, ben-
zalkonium methoxide was synthesised from benzalkonium
nitrate and sodium methoxide in methanol.
ð2Þ
The NMR studies of the methoxide ionic liquids
revealed that the calculated elemental composition of the
ionic liquids correspond to the proposed structures, with
high purity. The low water content of the ionic liquids,
determined by the Karl–Fischer titration, improved the
dehydrating potentials of the additives. The thermal
stability of the synthesised methoxide ionic liquids
(Table 1) showed lower decomposition temperatures in
comparison to the conventional ionic liquids due to an
increase in the anion hydrophilicity and decrease in the
ionic liquid viscosity e.g. the decomposition temperatures
of [C2-mim][Cl] decreases from 285 °C [32] to 210 °C
for the less viscous [C2-mim][MeO]. Since the thermal
stability of the synthesised ionic liquids is dependent on
the amount of water, a low operating temperature
ensured the stability of the ionic liquid owing to the
water produced during the reaction. The synthesis of
DMC from methanol and CO2 catalysed by ZrO2–MgO
was, thus, performed at 120 °C in the presence of the
methoxide ionic liquids. DMC was obtained as the major
product, with minor amounts of dimethyl ether detected
in the exhaust gas as the by-product. Both the catalyst
and the methoxide ionic liquids affected the conversion
of methanol and selectivity to DMC as demonstrated in
Sect. 3.2.1
3 Results and Discussion
3.1 Catalyst Characterisation
The X-ray diffraction (XRD) patterns of ZrO2–MgO
exhibited crystal patterns with peaks of high intensity at
30° and 50° assigned to a typical tetragonal phase of ZrO2,
and the peaks 43° corresponding to MgO. The particles of
MgO were dispersed in the bulk of the catalyst. These
results show that the mixed oxides resulted in t-ZrO2 sta-
bilised by MgO, which is the important phase in the syn-
thesis of DMC. The average crystal size of the particles
was 13.6 nm. The BET surface area determined was
55 m2/g. SEM/EDX results also confirm the presence of
mixed oxides (Fig. 1).
3.2 Alkoxide Ionic Liquids as Additives for DMC
Synthesis
The alkoxide ionic liquids were synthesised according to Eq. 2,
and used in the synthesis of DMC from CO2 and methanol. By
using sodium methoxide in the metathesis, 1-ethyl-3-methyl-
imidazolium methoxide 2a and 1-butyl-3-methylimidazolium
methoxide 2b were obtained. The hygroscopic viscous liquids
123

1258
V. Eta et al.
Table 1 DMC synthesis from methanol and CO2 at 120 °C over ZrO2–MgO
Entry
Additive
Tonset/ °C
Methanol
conversion (%)
DMC
selectivity (%)
TON
1
2
3
4
5
6
7
8
–
–
–
–
–
a[C4-mim][MeO]
[C4-mim][MeO]
[C2-mim][MeO]
[BA][MeO]
191
191
210
185
190
–
1.4
12.1
12.6
10.5
11.4
5.5
6
97
90
89
69
81
71
68
0.07
0.44
0.47
0.62
0.52
0.04
–
bRecycled [C4-mim][MeO]
CH3ONa
[C4-mim][Cl]/CH3ONa
240
Reaction conditions: 247 mmol methanol, 8 MPa CO2, 1 g ZrO2–MgO, 4 g IL, 120 °C, 9 h
Tonset Temperature at which the IL decreased by 5 wt%
a
No catalyst used
b
3.2 g of IL (1-ethyl-3-methylimidazolium methoxide [C2-mim][MeO], 1-butyl-3-methylimidazolium methoxide [C4-mim][MeO], ben-
zylalkonium methoxide [BA][MeO])
Since the maximum theoretical amount of DMC at
120 °C is only ca. 0.03 mol/L using ZrO2–MgO, the
methoxide ionic liquid was used as an additive to cir-
cumvent the thermodynamic restrictions via the removal of
water and consequently shifting the equilibrium towards
enhanced yields of DMC. The effects of the ionic liquids,
catalyst, reaction conditions and the recycling of the ionic
liquid on the conversion of methanol and selectivity to
DMC are discussed below. The turnover number (TON)
reported herein, is defined as the number of moles of DMC
formed per mol of ionic liquid used.
methanol. The activated species subsequently reacted with
CO2, activated by the ionic liquid, resulting in the forma-
tion of DMC. The catalyst was recovered via filtration and
drying after the experiment.
In order to investigate the influence of the methoxide
ionic liquids, DMC was synthesised in the presence of
dialkylimidazolium or benzylalkonium methoxide as
additives and the results reported on the basis of TON
(Table 1). The TON, a measure of the amount of meth-
oxide species used, was calculated so as to account for the
contribution of the ionic liquid used in the reaction. The
TON increased from 0.44 to 0.47 as the methoxide species
increased from 23.5 to 28.1 mmol for [C4-mim][MeO] and
[C2-mim][MeO], respectively. Conversely, the TON for
[BA][MeO] was 0.62 corresponding to 13.6 mmol of
methoxide species. In addition, the TON of the recycled
[C4-mim][MeO] was 0.52 corresponding to 18.8 mmol of
methoxide species. This shows that at the same methanol
conversion level, an increase of the amount of methoxide
species increases the TON since the methoxide species
promoted the formation of DMC via the removal of water.
The reaction of the methoxide species of the ionic liquids
with water produced during the reaction resulted in the
shift of the equilibrium towards DMC formation. For dif-
ferent levels of methanol conversions, the TON signifi-
cantly changed since it depends on the amount of DMC
formed and the ionic liquids used in the reaction.
3.2.1 Effect of the Catalyst and the Ionic Liquid
on DMC Synthesis
The results obtained from the synthesis of DMC in dif-
ferent ionic liquid–catalyst systems at 120 °C are summa-
rised in Table 1. The amount of DMC formed was
dependent on the amount/composition of the ionic liquid
and the presence or absence of the catalyst. In the absence
of both the catalyst and the ionic liquid, DMC formation
was nil. However, 1.4% conversion of methanol and 97%
selectivity to DMC was obtained in the presence of the
[C4-mim][MeO] without any catalyst. The conversion of
methanol increased to 12.1%, with 90% selectivity to DMC
when the catalyst and the ionic liquid were both present in
the reaction (Table 1, entries 1–3). The results show that
the additives together with the catalyst are necessary for the
reaction and showed a high performance at 120 °C and
8 MPa. The stabilisation of the tetragonal phase of ZrO2 by
MgO and the resulting increase in the basicity [25] of the
catalyst facilitated the activation and reaction of methanol.
The transfer of a proton to the metal oxide of the catalyst
and adsorption of methanol on the catalyst active sites
as methoxy-species resulted in the activated form of
A comparative study of the methoxide ionic liquids was
performed using the only sodium methoxide, or the con-
ventional [C4-mim][Cl] together with sodium methoxide at
120 °C catalysed by ZrO2–MgO. The amount of DMC
formed using these additives was lower that that obtained
with synthesised methoxide ionic liquids (Table 1, entry 7,
8). This indicates specific roles of the cation, anion and the
catalyst in the activation of methanol and CO2 and the
123

Effect of Alkoxide Ionic Liquids on the Synthesis of DMC
1259
trapping of water by the ionic liquid. Although the solu-
bility of CO2 in ionic liquids and the activation of methanol
have been reported as important initial steps for the reac-
tion of methanol and CO2 [26], these results show that only
ionic liquids with specific properties show high activity in
system resulted in enhanced yield and selectively to DMC
([70%). The new method shows promising results
obtained at a lower reaction temperature and pressure, thus,
saving energy and in addition, enabling the recovery and
recycling of the ionic liquid.
O
2CH3OH
+
+
CO2
+
+
N
N
R1
R
H3C
CH3
R
N
O
N
R1
R
N
N
R1
O
O
HO-
MeO-
O-
5
ð3Þ
4
2
3
a
R = CH3, R1 = CH3CH2
R = CH3, R1 = CH3CH2CH2CH2
b
3.2.2 Effect of Temperature on DMC Synthesis
the synthesis of DMC. Therefore, the performance of the
methoxide ionic liquids may be attributed to (i) the
abstraction of the C2 acidic proton of imidazolium cation
by the strongly basic methoxide to yield resonance stabi-
lised N-heterocyclic carbene [33–36], (ii) the formation of
dialkylimidazolium carboxylate due to the reaction of the
carbene with CO2 and affording the corresponding zwit-
terionic dialkylimidazolium carboxylates [33, 34], (iii) the
generation of dialkylimidazolium hydroxide, via the reac-
tion dialkylimidazolium methoxide with the water, and (iv)
the reaction of methoxide ion with the water formed during
the reaction, thereby shifting the reaction equilibrium
towards DMC formation. This is confirmed by a small
increment in the amount of methanol during 5 h using
sodium methoxide as additive due to the reaction of the
water produced with the methoxide species.
In order to investigate the stability of the ionic liquid and the
effect of increasing amounts of methoxide species in the
solution, the reaction was carried out in the presence of
1-butyl-3-methylimidazolium methoxide at different tem-
peratures. The results show that the conversion of methanol
increased as the temperature increased from 120 to 150 °C.
However, the selectivity to DMC decreased by a factor of
two as the temperature increased (Table 2, entries 1–2).
This indicates that at high temperatures, the conversion of
methanol increases due to the dehydration of methanol
resulting in dimethyl ether [37]. In addition, the ionic liquid
may decompose or promote polymerisation reactions,
leading to a low selectivity to DMC. A mild temperature
(120 °C) was chosen to ensure that the alkoxide ionic liq-
uids remained stable during 9 h, and also as a means of
minimising the energy used.
The insertion of CO2 to the dialkylimidazolium meth-
oxide may occur as a result of the abstraction of the C2
acidic proton, resulting in dialkylimidazolium carboxylate.
DMC and dialkylimidazolium hydroxide are formed by the
transesterification of the dialkylimidazolium carboxylate
with methanol (Eq. 3). When comparing the performance
of this novel alkoxide ionic liquid system and the previous
reported results [25] using butylene oxide as a chemical
trap for water for the synthesis of DMC (7.2% yield of
DMC, 53% selectivity), it can be observed that the novel
3.2.3 Effect of Ionic Liquid Loading on DMC Synthesis
DMC was synthesised from methanol and CO2 using dif-
ferent amounts of ionic liquids. Increasing the ionic liquid
loading resulted to an increase in the conversion of meth-
anol and the yield of DMC albeit a decrease in the selec-
tivity (Table 2, entries 1, 3). An increase in the amount of
Table 2 Effect of the reaction
temperature and ionic liquid
loading on DMC synthesis
Entry
Ionic liquid
Reaction
temp/°C
Methanol
conversion
(%)
DMC
selectivity
(%)
TON
Reaction conditions: 247 mmol
methanol, 8 MPa CO2, 1 g
ZrO2–MgO, 4 g IL, 9 h
1
2
3
[C4-mim][MeO]
[C4-mim][MeO]
[C4-mim][MeO]a
120
150
120
12.1
13.3
14.8
90
48
86
0.44
0.31
0.33
a
8 g ionic liquid
123

1260
V. Eta et al.
ionic liquid resulted in a non-linear increase in the con-
version of methanol. The TON also decreased from 0.44 to
0.33 due to the increase in the amount of methoxide species
without a corresponding proportional increase in the
amount of DMC formed. This indicates that increasing the
amount of ionic liquid was not beneficial due to the
increase in the liquid viscosity and changes in the flow
patters in the reactor albeit an efficient stirring (700 rpm).
Calculations of the phase equilibria of the binary mixture
of methanol and CO2 [38] show that about 9–12 mL of
CO2 is present in the liquid phase as liquid CO2. Therefore,
the maximum energy from the impeller was transferred to
the fluid in order to eliminate diffusion resistances. How-
ever, due to the re-establishment of the equilibrium of the
reaction and the mass transfer resistances due to the
increase in viscosity of the solution, an optimum amount of
ionic liquid (4 g) was chosen in order to obtain a high
methanol conversion and TON.
CH ONa
3
CH3OH
R1
+
HO-
4
R1
+
N
N
R
R
N
N
MeO-
4
2
R = CH3, R1 = CH3CH2
R = CH3, R1 nBu
a
=
b
The performances of the fresh and regenerated ionic
liquids were similar in the conversion of methanol and the
yield of DMC. These results show that the regenerated
ionic liquid could be reused several times with only a
minor loss in activity. It should, nevertheless be stated that
a quantitative assessment of the environmental effects of
imidazolium salts e.g. [C4-mim][Cl] shows environmental
and aquatic toxicity [39], therefore proper recycling or
disposal of these ILs are required in order to make their
application sustainable.
3.3 Recycling of Ionic Liquid
The hydroxide ionic liquid confirmed by NMR studies was
recovered by filtration, washing and distillation of the
liquid phase. The yield of the recovered IL was ca. 85 wt%.
The recovered ionic liquid reacted with a solution of
sodium methoxide in methanol resulted in the regeneration
of methoxide ionic liquid (Eq. 4). The results obtained with
the rejuvenated ILs are shown in Fig. 2. About 10 wt% of
the IL was lost as a result of differences in mass between
methoxide and hydroxide ILs and other losses due to
handling.
4 Conclusions
Novel systems containing methoxide anions together with
alkylimidazolium and benzalkonium cations were pre-
pared, characterised and tested in the direct synthesis of
DMC from methanol and CO2. Combination of the meth-
oxide ionic liquids and an active catalyst (ZrO2–MgO)
afforded high methanol conversion (12%) and high selec-
tivity to DMC (90%) based on the methanol consumed at
relatively low temperature (120 °C). Furthermore, the
methoxide ionic liquids acted as a water trapping medium,
thus, shifting the reaction equilibrium towards DMC for-
mation, and at the same time limiting the formation of by-
products. The ionic liquids were recovered and reused
without loss in the performance.
1.5
1.2
0.9
˚
Acknowledgments This work is part of the activities of Abo
Akademi Process Chemistry Centre within the Finnish Centre of
Excellence (2006–2011), and the KETJU research programme
(2006–2010) by the Academy of Finland. The authors are grateful to
the Academy of Finland for the financial support under the grants
120853, 124357 and 128626. This work is also associated with the
˚
activities of Umea University Chemical-Biological Centre whereupon
1.34
1.32
financial support from Bio4Energy programme, Kempe Foundations
and Knut and Alice Wallenberg Foundation are acknowledged.
1.31
0.6
0.3
References
0
Fresh ionic
liquid
Regenerated 1
Regenerated 2
1. Tundo P, Selva M (2002) Acc Chem Res 35:706–716
2. Tundo P (2001) Pure Appl Chem 73:1117–1124
3. Pacheco MA, Marshall CL (1997) Energy Fuels 11:2–29
4. King ST (1997) Catal Today 33:173–182
Fig. 2 Comparison of the fresh and regenerated [C4-mim][MeO] for
DMC synthesis. Reaction conditions after regeneration 1: 197 mmol
methanol, 8 MPa CO2, 0.8 g ZrO2–MgO, 3.2 g IL, 120 °C, 9 h
5. Aresta M, Quaranta E (1997) Chem Tech 27:32–40
123

Effect of Alkoxide Ionic Liquids on the Synthesis of DMC
6. Babad H, Zeiler AG (1997) Chem Rev 73:75–91
1261
´
¨
25. Eta V, Maki-Arvela P, Leino A-R, Kordas K, Salmi T, Murzin D,
Mikkola J-P (2010) Ind Eng Chem Res 49:9609–9617
26. Dong WS, Zhou X, Xin C, Liu C, Liu Z (2008) Appl Catal A Gen
334:100–105
27. Ju HY, Manju MD, Kim KH, Park SW, Park DW (2007) Korean
J Chem Eng 4:917–919
¨
7. Notz R, Tonnies I, McCann N, Scheffknecht G, Hasse H (2011)
Chem Eng Technol 34:163–172
8. Krush A, Vogel H (2008) Chem Eng Technol 31:23–32
9. Bruggink A, Schoevaart R, Kieboom T (2003) Org Process Res
Dev 7:622–640
10. Aresta M (2002) Fuel Chem Div Preprints 47:255
11. Lee MY, Park DC (1991) Stud Surf Sci Catal 66:631–640
12. Kizlink J, Pastucha I (1995) Collect Czech Chem Commun
60:687–692
28. Virtanen P, Karhu H, Toth G, Kordas K, Mikkola J-P (2009)
J Catal 263:209–219
29. Ju HY, Manju MD, Park DW, Choe Y, Park SW (2007) React
Kinet Catal Lett 90:3–9
13. Kizlink J, Pastucha I (1994) Collect Czech. Chem Commun
59:2116–2118
30. Wang H, Lu B, Cai QH, Wu F, Shan YK (2005) Chin Chem lett
16:1267–1270
14. Fang S, Fujimoto K (1996) Appl Catal A 142:L1–L3
15. Li C-F, S-He Zhong (2003) Catal Today 82:83–90
16. Tomishige K, Sakaihiro T, Ikeda Y, Fujimoto K (1999) Catal Lett
58:225–229
17. Jung KT, Bell AT (2001) J Catal 204:339–347
18. Tomishige K, Ikeda Y, Sakaihori T, Fujimoto K (2000) J Catal
192:355–362
31. Brunauer S, Emmett PH, Teller E (1938) J Am Chem Soc
60:309–319
32. Huddleton JG, Visser AE, Reichert WM, Willauer HD, Broker
GA, Rogers RD (2001) Green Chem 3:156–164
33. Tommasi I, Sorrentino F (2005) Tetrahedron Lett 46:2141–2145
34. Tommasi I, Sorrentino F (2006) Tetrahedron Lett 47:6453–6456
35. Holbrey JD, Reichert WM, Tkatchenko I, Bouajila E, Walter O,
Tommasi I, Rogers RD (2003) Chem Commun 2003, 28–29. doi:
10.1039/B211519K
36. Duong HA, Tekavec NT, Arif AM, Louie J (2004) Chem Com-
mun 1:112–113
37. Hong ST, Park HS, Lim JS, Lee Y-W, Masakazu Anpo M, Kim
J-D (2006) Res Chem Intermed 32:737–747
¨
19. Eta V, Maki-Arvela P, Murzin DYu, Salmi T, Mikkola J-P (2008)
Progress in heterogeneous catalysis. In: Marmaduke DL (ed)
Nova Science Publishers. Hauppauge, NY, pp 135–155
20. Tomishige K, Kunimori
237:103–109
K (2002) Appl Catal A Gen.
21. Honda M, Kuno S, Sonehara S, Fujimoto K, Suzuki K, Nakagawa
Y, Tomishige K (2011) Chem Cat Chem 3:365–370
22. Bhanage BM, Fujita S, Ikushima Y, Arai M (2001) Appl Catal A
Gen 219:259–266
38. Ballivet-Tkatchenko D, Chambrey S, Keiski R, Ligabue R,
Plasseraud L, Richard P, Turunen H (2006) Catal Today
115:80–87
¨
39. Ilgen F, Ott D, kralisch D, Palmberger A, Konig B (2009) Green
Chem 11:1948–1954
23. De C, Lu B, Lv H, Yu Y, Bai Y, Cai Q (2009) Catal Lett
128:459–464
24. Cui H, Wang T, Wang F, Gu C, Wang P, Dai Y (2004) Ind Eng
Chem Res 43:7732–7739
123