The Synthesis of Dimethyl Carbonate by the Oxicarbonylation of Methanol Over Cu Supported on Carbon Norit

Article abstract of DOI:10.1007/s10562-014-1460-9

The catalytic activity of Norit supported Cu, Ni and Cu-Ni catalysts was investigated in the synthesis of dimethyl carbonate (DMC) by the oxidative carbonylation of methanol. Cu/Norit showed the best catalytic activity. The reaction was carried out in a continuous flow system at atmospheric pressure usually at 393 K. The main products were methyl formate, DMC and CO₂. The methanol conversion on Cu/Norit achieved in steady state was about 22 % and the DMC yield 13.2 %. Based on the XPS data we can establish that copper reduced to its metallic form during reduction but oxidized in the reaction mixture, and is mostly in the Cu⁺ state, with some Cu²⁺. It is possible that the DMC formation rate depends on the surface concentration of oxidized Cu and on the ratio of Cu⁺ and Cu²⁺. Based upon the IR measurements adsorbed DMC was found on the surface of the Cu/Norit catalyst during the catalytic reaction.

Full text of DOI:10.1007/s10562-014-1460-9

Catal Lett
DOI 10.1007/s10562-014-1460-9
The Synthesis of Dimethyl Carbonate by the Oxicarbonylation
of Methanol Over Cu Supported on Carbon Norit
•
G. Merza B. Laszlo A. Oszko G. Potari
•
•
•
´
´
E. Varga A. Erdohelyi
´
´ ´
•
}
Received: 26 September 2014 / Accepted: 5 December 2014
Ó Springer Science+Business Media New York 2014
Abstract The catalytic activity of Norit supported Cu, Ni
and Cu–Ni catalysts was investigated in the synthesis of
dimethyl carbonate (DMC) by the oxidative carbonylation
of methanol. Cu/Norit showed the best catalytic activity.
The reaction was carried out in a continuous flow system at
atmospheric pressure usually at 393 K. The main products
were methyl formate, DMC and CO2. The methanol con-
version on Cu/Norit achieved in steady state was about
22 % and the DMC yield 13.2 %. Based on the XPS data
we can establish that copper reduced to its metallic form
during reduction but oxidized in the reaction mixture, and
is mostly in the Cu? state, with some Cu2?. It is possible
that the DMC formation rate depends on the surface con-
centration of oxidized Cu and on the ratio of Cu? and
Cu2?. Based upon the IR measurements adsorbed DMC
was found on the surface of the Cu/Norit catalyst during
the catalytic reaction.
industrial chemical. Its usage is facilitated by its low tox-
icity, high biodegradability and low persistence [1]. DMC
can be used in industry mainly as a fuel additive [2], pre-
cursor for synthesis of carbonic acid derivatives, methy-
lating agent [3], methyl sulphate exchanger [4], or an
intermediate in the synthesis of polycarbonates and isocy-
anates [5], carbonylating agent [6], alkylating agent, polar
solvent [7], component in the synthesis of polyurethane [8].
Proving the possibilities of the green usage of DMC, a
new alternative route for the enzymatic coproduction of
biodiesel and glycerol carbonate was investigated by Seong
et al. [9] by the transesterification of soybean oil with
DMC.
The traditional DMC synthesis used toxic and hazardous
phosgene as reactant and suffered from several drawbacks
[10]. Numerous developments have been published for the
production of DMC including methanolysis of phosgene,
ester exchange process, liquid phase methanol carbonyl-
ation and methylnitrite carbonylation and gas-phase oxi-
dative carbonylation of methanol [11] although these
processes use corrosive, toxic and flammable gases.
Therefore the preparation and the usage of new catalysts
such as K2CO3 [12], ZrO2 [13], H3PW12O4–ZrO2 [14], Cu
modified (Ni, V, O) complex [15], ZrO2–CeO2 [16] were
forced. The disadvantages such as the need of high pressure
or the deactivation of the catalyst were also described and
the reported formation rate of DMC was low. Various
environmentally friendly chemical methods were devel-
oped like the transesterification of ethylene carbonate with
methanol [17], and the synthesis from urea and methanol
[18].
Keywords Oxidative carbonylation Á Dimethyl
carbonate Á Active carbon Á Supported Cu catalyst Á
Cu/Norit
1 Introduction
As an environmentally friendly compound dimethyl car-
bonate (DMC) has the potential to be used as a green
´
´
´
´ ´
G. Merza Á B. Laszlo Á A. Oszko Á G. Potari Á E. Varga Á
}
A. Erdohelyi (&)
Institute of Physical Chemistry and Materials Science,
´
´
University of Szeged, Aradi vertanuk tere 1, Szeged 6720,
Carbonaceous materials have been used widely in
catalysis because of their superior structural, mechanical,
chemical, thermal, and unique electrical transport
properties.
Hungary
e-mail: erdohelyi@chem.u-szeged.hu
123

G. Merza et al.
Among the various catalysts CuCl2/NaOH/activated
carbon (AC) was reported to show good catalytic perfor-
mance in the synthesis of DMC. The effect of reaction
conditions and that of promoters on the reaction were
evaluated [19]. The optimal reaction temperature was set to
393–403 K. The catalytic activity increased with the
increase of OH/Cu molar ratio as compared to CuCl2/AC.
With the morphological analysis two different crystal
structures were observed on the surface of the catalyst:
orthorhombic and rhombohedral. The second one was more
favourable for DMC synthesis [20].
[32]. Nevertheless we did not find any data in the literature
about using carbon Norit as catalysts support operating in
DMC synthesis.
In this paper we focus on the e synthesis of DMC at
atmospheric pressure by vapour-phase oxidative carbon-
ylation of methanol on copper supported by AC (Norit), a
chloride free catalyst and on the catalytic properties during
the reaction.
2 Experimental
Typical results were obtained on CuCl/Carbon in the
oxidative carbonylation of methanol. It is clear that the
2-
2.1 Preparation of the Catalyst
CuCl3 surface species finally shifted to Cu0 and cuprous
chloride was identified as the active species in DMC pro-
duction [21]. Ren et al. [22] observed that the activation
temperature of the starch-based carbon supported Cu cat-
alyst significantly influenced the activity of the sample,
while the surface area, micropore volume and the total pore
volume changed and these properties resulted in the dif-
ferent efficiency. The AC supported PdCl2–CuCl2 was also
an efficient catalyst in this reaction but the activity of the
sample significantly decreased in time. The results showed
that the catalyst deactivation was due to the loss of chlorine
[23].
Norit supported Cu, Cu–Ni and Ni catalysts were prepared
via the conventional incipient wetness impregnation tech-
nique. The AC (Norit ROW 0.8 mm pellet from Alfa Ae-
sar) was impregnated with Cu(NO3)2Á3H2O or/and
Ni(NO3)2Á6H2O solution. In the bimetallic sample the Cu/
Ni ratio was 2. In a typical preparation process calculated
amount of the metal salt was dissolved in 25 % ammonia
solution and then the support was added to yield 10 wt%
metal content. The mixture was stirred for 24 h at room
temperature followed by ultrasonication for 2 h, and aging
for 24 h. The liquid was evaporated at 363 K. The dry
powder was calcined at 723 K for 3 h in N2 flow. The
samples were reduced in situ usually at 873 K in hydrogen
flow for 1 h. The BET surface area of the Cu/Norit was
1,070 m2/g, of the Ni/Norit 900 m2/g, of the Cu–Ni/Norit
951 and 1,212 m2/g of the pure support.
Cu–Ni supported on multi-walled carbon nanotubes
(MWCNT) was also used as catalyst for the direct syn-
thesis of DMC connected with the full characterization of
the sample. Relatively high methanol conversion and DMC
selectivity were reported, though the optimal experimental
conditions, like high pressure, have some inconveniences
[24]. In a recent paper we discussed the yield of DMC
formation which was higher on Cu/MWCNT than on Cu–
Ni/MWCNT catalyst [25].
The support was originally consisted of 0.8 mm o.d. and
3–4 mm long cylindrical pellets; this form was kept
throughout the experiments. The average size of the metal
particles was about 39 nm. Only 6 % of the Cu particles
are smaller than 10 nm and about 20 % of them are bigger
than 50 nm.
On a novel graphene nanosheet (GNS) Cu–Ni bimetallic
catalyst the catalytically active metal particles were dis-
persed on the GNS in the direct synthesis of DMC and this
effectively enhanced the productivity [26].
The metal clusters were roughly spherical on the surface
of amorphous carbon.
The mechanism of the oxidative carbonylation of
methanol was theoretically investigated by Ren et al. The
calculated results showed that the DMC formation pathway
should be as follows: monomethyl carbonate (MMC) spe-
cies produced by CO insertion to methoxide species, and
then MMC reacts with methanol to form DMC [27].
Earlier AC (Norit) was successfully used as a catalyst
support in different reactions. Solymosi et al. used Mo2C/
Norit in the decomposition and steam reforming of meth-
anol [28], for the decomposition of ethanol [29], and also
for the steam reforming and the decomposition of dimethyl
2.2 Reaction Procedure
The DMC synthesis by the oxidative carbonylation of
methanol was carried out in a continuous flow system in a
fixed bed quartz reactor at atmospheric pressure. The
reactor tube had an inner diameter of 7 mm with a length
of circa 20 cm.
Methanol was introduced into the system by bubbling
the mixture of CO and O2 through the methanol at 323 K.
Usually 1 g of the catalyst was used without dilution. After
the pre-treatment of the sample the reactor was cooled
down to the reaction temperature, usually 393 K. The ratio
of the reactants (CH3OH/CO/O2 ratio was 2/1/1) was the
same in all experiments. The flow rate was usually 16 ml/min.
´
ether (DME) [30]. Marban et al. investigated Norit-sup-
ported Rh-based catalyst for the methanol decomposition
reaction [31], Tolmacsov et al. tested Pt metals supported
on Norit for the decomposition and reforming of methanol
123

The Synthesis of Dimethyl Carbonate by the Oxicarbonylation
The reactants and the products were analysed on-line by an
Agilent 6890 gas chromatograph equipped with an Agilent
5975C VL MSD mass-spectrometer, thermal conductivity
and flame ionization detectors.
subjected to rolling ball background subtraction and con-
trast enhancement, and then the diameter of the metal
nanoparticles in the image was manually measured against
the calibrated TEM scale bar. Each diameter distribution
histogram was constructed from 100 individual nanoparti-
cle diameter measurements [33, 34].
2.3 Characterization of the Catalysts
For XPS studies the powder samples were pressed into
tablets with approximately 10 mm diameter and a few
tenth of mm thickness and placed into the load lock of the
spectrometer. Sample treatments were carried out in a high
pressure cell (catalysis chamber) directly attached to the
analysing chamber and isolated from that with a gate valve.
With the help of the sample manipulator and a second
insertion mechanism it was possible to transfer the samples
from the analysis chamber into the high pressure cell in
vacuum, without the reach of air. The samples were pre-
treated the same way as mentioned above. After each pre-
treatment step the samples were cooled down in nitrogen
flow to room temperature. Afterwards the flow was stop-
ped, the high pressure cell was evacuated and the sample
was taken back to the analysis chamber. XP spectra were
taken with a SPECS instrument equipped with a PHOIBOS
150 MCD 9 hemispherical energy analyser operated in the
FAT mode. The excitation source was the Ka radiation of a
magnesium anode (hm = 1,253.6 eV). The X-ray gun was
operated at 210 W power (14 kV, 15 mA). The pass
energy was set to 20 eV, the step size was 25 meV. Typ-
ically five scans were added to get a single spectrum. For
data acquisition and evaluation both manufacturers’
(Specslab2) and commercial (CasaXPS, Origin) software
packages were used.
3 Results and Discussion
3.1 Oxidative Carbonylation of Methanol
The catalytic reaction was intentionally performed at
atmospheric pressure at 393 K, on Cu/Norit to obtain a
‘greener’ way for DMC synthesis. For comparison the tests
were carried out also on Ni/Norit and on Cu–Ni/Norit
catalysts.
In the reactions CO2, DMC and methyl formate (MF)
were formed as main products in all cases. A small amount
of dimethoxy methane (DMM) and traces of DME were
also detected. Figure 1a shows the conversion of methanol
on Cu/Norit, which was increasing through the reaction, up
to 22 % after 7 h time on stream. Figure 1b demonstrates
that on Ni/Norit the methanol conversion was about 13 %
at the beginning of the reaction, in the first 2 h it decreased
continuously and after 150 min reached the quasi steady
state value, about 2.5 %. The conversion of methanol on
Cu–Ni/Norit increased up to 11 % in the first 130 min of
the reaction attained a maximum and remained nearly the
same (Fig. 1c). The CO conversion changed similarly to
the methanol conversion in all cases. On the Ni/Norit it
decreased in time but in the Cu containing samples the
conversion increased as a function of time on stream
(Fig. 1).
Infrared spectra were recorded with an Agilent Cary 670
type FTIR spectrometer equipped with diffuse reflectance
attachment (Harrick) with BaF2 windows with a wave
number accuracy of 4 cm-1. Typically 32 scans were
registered. The whole optical path was purged by CO2- and
H2O-free air generated by a Balston purge gas generator.
The catalysts were pre-treated as mentioned above then the
CH3OH ? CO ? O2 mixture was introduced into the cell
at the reaction temperature and the IR spectra were
recorded. Practically the same experimental conditions
were used as in the catalytic measurements.
On Cu/Norit the formation rate of CO2 in the first hour
of the reaction increased, then decreased, after 3 h it
remained nearly constant, while the formation rate of MF
slightly increased in the first hour then remained the same
(Fig. 2a). The formation of DMC started after the 100th
min of the reaction evolved with an enhanced rate until the
350th min afterwards started to decrease. On Ni/Norit the
rate of MF and CO2 formation decreased in time, the DMC
formation increased slightly at the same time although the
production was less than measured on Cu/Norit (Fig. 2b).
However, on Cu–Ni/Norit the rate of MF production sig-
nificantly increased and that of CO2 decreased slower in
time, still the DMC and the DMM formation increased only
a bit (Fig. 2c). It is clear from the above data, that under the
same circumstances the rate of DMC formation was more
than ten times higher on Cu/Norit as on Ni/Norit or on Cu–
Ni/Norit after 5 h of the reaction.
The BET surface areas of the catalysts were measured
using N2 adsorption at the temperature of the liquid
nitrogen by a BELCAT A instrument with single point
method.
The Cu particle size distribution was determined by
image analysis of the HRTEM (FEI Tecnai G2 20 X-Twin;
200 kV operation voltage, 180,0009 magnification,
125 pm/pixel resolution) pictures using the ImageJ soft-
ware. At least five representative images of equal magni-
fication, taken at different spots of the TEM grid were first
The selectivity of MF remarkably increased at the
beginning of the reaction on Cu/Norit. After reaching a
123

G. Merza et al.
Fig. 1 The conversion of CO
and methanol on Cu/Norit (a),
on Cu–Ni/Norit (b) and on Ni/
Norit (c) in the
CH3OH ? CO ? O2 (2:1:1)
reaction at 393 K, (The space
velocity was 780 h-1
)
Fig. 2 The rate of dimethyl
carbonate, methyl formate, CO2
and dimethoxy methane
formation on Cu/Norit (a), on
Ni/Norit (b) and on Cu–Ni/
Norit (c) in the
CH3OH ? CO ? O2 (2:1:1)
reaction at 393 K, (The space
velocity was 780 h-1
)
maximum 33 % MF selectivity after 150 min it started to
decrease and stabilized after 300 min at 17 %. In the first
100 min of the reaction only traces of DMC was formed
then its selectivity increased gradually, after 300 min
stayed nearly constant, about 60 %. On the other hand,
CO2 formation was the opposite, decreased drastically
from 100 % selectivity to 23 % after 300 min (Fig. 3a).
On Ni/Norit the MF (from 90 to 74 %) and the CO2
(from 22 to 19 %) selectivity decreased (Fig. 3b). In the
first 100–120 min of the reaction only a small amount of
DMC was detected, afterwards with increasing selectivity
it reached only 10 % in the 250 min of the reaction. Fig-
ure 3c shows the selectivity of the products on Cu–Ni/
Norit. DMC selectivity marginally and that of MF notably
increased in the reaction, while the selectivity of CO2
constantly decreased.
The yield of the DMC formation was the highest in the
steady state period at ambient pressure on Cu/Norit
(*13.2 %) and it decreased in the order of Cu/Nor-
it [ Cu–Ni/Norit [ Ni/Norit. The value obtained on Cu/
Norit is comparable with the yield obtained on Cu/Y zeo-
lite (13.1 %) [35] or on CuCl2 supported on AC (16.3 %)
[19] at 2.3 bar in an autoclave system, but it was definitely
higher than it was found on Cu–Ni/MWCNT at high
pressure [12] or on Cu/MWCNT at atmospheric pressure
[25].
When the catalyst was reduced at lower temperature (773
or 673 K) the induction time of DMC formation was
123

The Synthesis of Dimethyl Carbonate by the Oxicarbonylation
Fig. 3 The selectivity of
dimethyl carbonate, methyl
formate, CO2 and dimethoxy
methane on Cu/Norit (a), on Ni/
Norit (b) and on Cu–Ni/Norit
(c) in the CH3OH ? CO ? O2
(2:1:1) reaction at 393 K, (The
space velocity was 780 h-1
)
considerably longer (after the pre-treatment at 673 K it was
280 min), the DMC selectivity in the steady state was lower,
below 20 % instead of 60 %. Without any pre-treatment the
catalyst was totally inactive in the DMC synthesis.
When the methanol–oxygen reaction was followed
under the same experimental conditions at 393 K, the
methanol conversion was similar on either of the catalysts
though in these cases MF was the main product, only a
small amount of CO2 and traces of formaldehyde were
formed. The rate of products formation decreased in the
first 30 min of the reaction then remained stable. In the
methanol–CO reaction only CO2 was formed.
The pure support was also used as catalyst in the same
reaction using the same experimental conditions. In this case
only a small amount of CO2 was found as reaction product.
Fig. 4 The formation rate of CO2, dimethyl carbonate and methyl
formate as a function of temperature (a) (contact time = 3.75 g sec/
ml) and contact time (b) at 393 K in the CH3OH ? CO ? O2 (2:1:1)
reaction on Cu/Norit approximately at the steady state conditions
3.2 Catalytic Behaviour of Cu/Norit
The oxidative carbonylation of methanol was studied in
detail on Cu/Norit. The reaction temperature was varied
from 373 to 413 K. Each experiment was carried out on
fresh catalyst. The steady state values obtained at different
temperatures were compared. The rate of the products
formation increased in all cases when the reaction was
followed at higher temperature than 373 K. The amount of
CO2 increased continuously as a function of temperature
but the MF and DMC formation curves had maxima at 383
and 393 K, respectively (Fig. 4a). It should be noted that
the DMM formation rate significantly increased as a
function of temperature; at 413 K it was about the quarter
of the DMC production, while only traces were detected at
373 K. The selectivity of MF decreased by increasing the
temperature, while that of DMC changed by giving a
maximum curve.
The variation of the space velocity exerted divergent
influence on the products distribution of the oxidative
carbonylation of methanol. The conversion of methanol
and the amount of DMC and CO2 formed in the reaction
increased quasi linearly with rising the contact time. In
contrast the formation curve of MF had a maximum
(Fig. 4b).
3.3 XP Spectra
In order to follow the changes of the surface composition
and those of the oxidation states of the catalysts compo-
nents XPS measurements were carried out during the cat-
alytic reaction.
123

G. Merza et al.
High-resolution XP spectra of Cu/Norit were taken in
the Cu 2p, C 1s, O 1s and Cu LMM regions in all stages of
the treatment. The Cu 2p doublet in the as received sample
with Cu 2p3/2 at 933.6 eV showed the presence of
Cu(NO3)2 [36] and CuO which probably formed in the
decomposition of nitrate salt. Both the position of the Cu
2p3/2 peak and the presence of satellite features support this
assertion. The peak location shifted to 932.4 eV after
reduction at 873 K for 60 min and the satellites disap-
peared (Fig. 5). To distinguish between different copper
oxidation states we must also take into account the Auger
parameter, which in its simplified form of the sum of the
binding energy of the Cu 2p3/2 photo peak and the kinetic
energy of the L3MM Auger peak. This feature was located
at 918.8 eV after reduction, thus giving 1,851.2 eV for the
Auger parameter, so copper reduced to its metallic form
and its surface concentration significantly increased during
the reduction. In the course of the oxidative carbonylation
of methanol the Cu 2p3/2 binding energy and the peak
intensity did not change but the enhancement of a new peak
around 944 eV (as a satellite) could be seen. This peak was
observed even after 10 min of the reaction. After 60 min an
intensifying shoulder could also be observed on the high
binding energy side on both components of the Cu 2p
doublet. The Auger spectra were more similar to the one
taken in the as received state than to the spectrum taken
after reduction. So we may conclude that copper was
oxidized in the reaction mixture and was mostly in the Cu?
state, with some Cu2? also present on the surface. The
amount of this latter species increased with increasing
reaction time and was demonstrated by the intensifying
satellite peak and shoulder on the copper spectra.
Similar spectra were obtained for the Cu component in
the Cu–Ni/Norit catalyst. As regards the binding energy of
Ni 2p in the as received sample the presence of NiO
(855.6 eV) and a small amount of metallic Ni (852.7 eV)
[37] were detected. The Ni 2p spectra of the reduced cat-
alyst showed the presence of metallic Ni and Ni2? species
at 852.7 and 855.6 eV, respectively. This indicates the
existence of a fraction of unreduced Ni after the reduction
of the sample in the Cu–Ni/Norit. On the Ni/Norit sample
before the pre-treatment only the Ni2? was detected at
855.6 eV. On the reduced sample a new feature was also
observed at 854.5 eV in addition to the metallic Ni
(852.7 eV) which could be assigned also to Ni2? but in
different chemical environment [38].
The survey spectra showed a very intense C 1s peak.
High resolution carbon spectra taken on pure Norit and
Cu/Norit featured
a
narrow asymmetric peak
(FWHM = 1.3–1.4 eV) with a characteristic long tailing
and shoulder on the high binding energy side (Fig. 6). On
pure Norit an intense peak showed up at 284.5 eV in the as
received state with components at *286–287, *290
and *293 eV. The 284.5 eV peak corresponds to C–C and
Fig. 5 XP spectra of Cu 2p
(a) and Cu LMM Auger spectra
(b) recorded after the
CH3OH ? CO ? O2 reaction at
393 K on Cu/Norit; as received
(1), reduced sample (2), after
10th (3), 60th (4) and 150th (5)
minutes of the reaction
A
B
Cu LMM
Cu 2p
932.4
Cu(II)
A.p.[Cu] = 1851.3
Cu
Cu(II)
Cu(I)
A.p.[Cu(I)] = 1849.3
A.p.[Cu(II)] = 1851.1
5
4
5
3
4
3
2
2
1
1
905
910
915
920
925
968
960
952
944
936
928
Kinetic energy [eV]
Binding energy [eV]
123

The Synthesis of Dimethyl Carbonate by the Oxicarbonylation
C–H bonds, while the 286 and 290 eV peaks to oxygenated
carbon bonds like C–O (phenolic, alcoholic, etheric) and
O–C=O (carboxyl, ester), respectively. The component
with the highest binding energy may come from adsorbed
CO or CO2 or from the shake-up satellites from the p–p*
transitions in aromatic rings [39, 40].
After the oxidative carbonylation of methanol the line
shape changed considerably, an intense peak developed
around 530–531 eV even after 10 min time on stream but
there was no change in the intensity during the reaction.
The rendering of oxygen components is somewhat
ambiguous in the literature. Some [39–41] assign the peak
around 530–531 eV to oxygen in C=O bond, while Soria-
No changes were detected on pure Norit after a mild
reduction with H2 at 473 K for 30 min. The peak shape and
peak areas remained practically constant throughout the
experiment on the copper doped samples too; the change
was well within 10 %. However the intensity of the com-
ponent representing C–O groups markedly decreased.
These also mean that we could not detect the formation of
any new carbonaceous surface species during the reaction.
The copper doublet is easily identifiable on the wide scan
spectra but the O 1s region is hardly discernible, especially
on the spectra of the treated samples. However high-reso-
lution O 1s spectra showed that this region has a complex
shape. In the O 1s region of pure Norit a peak located at
532.5 eV was detected with a shoulder at 537.4 eV. After
reduction the intensity of the main peak reduced but its
position did not change. The minor component shifted to
536.3 eV and its intensity increased relative to the main
peak. The picture modified when spectra were recorded on
Cu/Norit after reduction at 693 K for 60 min. In this case
the O 1s spectrum could be fitted with four components, the
most intense one being located at 533.2 eV (Fig. 7.). At the
same time two minor shoulders could be identified at 536.6
and 530.6 eV. The peak fitting process required to suppose
that an additional component is present peaking at 534.4.
´
Sanchez et al. put this component to 531.7 eV [42]. Since
C=O bonds can occur in various groups at the edges of the
carbon structure, both assumptions can be true. Electrons
from copper–oxygen bonds can also contribute to this peak.
Following the work of Figueiredo et al. [39] the peak at
533.2 after reduction can be attributed to –O– bonds. This
peak vanished in the course of reaction, instead peaks
developed at 532.7–532.9 eV, with increasing intensity
during the reaction, characteristic of C=O bonds e.g. in
anhydride. OH groups may also contribute to this feature.
The peak at 534.0–534.4 eV may originate from carboxylic
groups, while that at 536.1–536.6 eV was from adsorbed
water. As already mentioned above, neither the intensity
nor the shape of the C 1s spectrum changed remarkably
during the reaction. Changes did occur in the structure of O
1s spectra however these are not reflected in the carbon line
shapes. The reason for this was that after reduction on the
sample the carbon to oxygen ratio at the surface was about
20:1. After 150 min of the reaction the oxygen content
doubled to about 10 %, but this was still too few to cause
major changes in the carbon spectra.
From the XP spectra the surface concentrations of
metals on the pre-treated samples were also determined. On
Fig. 6 XP spectra of C 1s
recorded after the
284.3
284.5
A
B
CH3OH ? CO ? O2 reaction at
393 K on Cu/Norit; as received
(1), reduced sample (2), after
10th (3), 60th (4) and 150th (5)
minutes of the reaction (a), the
pure Norit (b)
40000 cps
20000 cps
5.
4.
3.
286.7
290.0
2.
1.
293.4
295
294
292
290
288
286
284
282
280 300
290
285
280
Bindingenergy [eV]
Bindingenergy [eV]
123

G. Merza et al.
Fig. 7 XP spectra of O 1s taken
on Cu/Norit; reduced (a), after
10th (b), 50th (c) and 150th
(d) minutes of the
CH3OH ? CO ? O2 reaction at
393 K
O 1s
533.2
A
B
532.7
530.2
534.0
3000 cps
1500 cps
536.3
530.6
534.4
536.6
538
536
534
532
530
528
538
536
534
532
530
528
Binding energy [eV]
Binding energy [eV]
C
D
532.8
532.9
534.0
534.0
3000 cps
3000 cps
536.1
530.1
530.0
536.3
538
536
534
532
530
528
538
536
534
532
530
528
Binding energy [eV]
Binding energy [eV]
]
reduced Cu/Norit the Cu content was 0.8 %, on Cu–Ni/
Norit 2 % Cu and 4.8 % Ni was detected. On the surface of
Ni/Norit the metal concentration was 6.2 %.
subtracted, absorptions at 2,141, 1,767, 1,610, 1,495, 1,404,
1,293–1,291, 1,066–1,053, 1,036–1,034 and 1,026–1,021 cm-1
could be detected. The intensities of these bands changed in
different ways during the reaction. The absorption at
2,141 cm-1 diminuted, and the 1,767 and 1,292 cm-1 fea-
tures intensified in time. The intensity of the bands at 1,034
and 1,021 cm-1 first evolved, but after 30 min of the reac-
tion diminished (Fig. 8).
The bands at 2,141 cm-1 are in the CO region. In
contrast to earlier findings, the adsorption of CO on Cu2?
or on metallic Cu sites were predominantly weak and
3.4 DRIFT Spectra Registered During the Reaction
The infrared spectra recorded in the DRIFT cell during the
oxidative carbonylation of methanol on Cu/Norit at 393 K
showed mainly the gas phase spectra of CO and methanol. If
the spectra obtained on pure Norit under the same experi-
mental conditions in the 2,200–1,000 cm-1 range is
123

The Synthesis of Dimethyl Carbonate by the Oxicarbonylation
reversible at 300 K and above, still CO bonded most
strongly on Cu? [43, 44]. London and Bell [45] found a
sharp peak at 2,140 cm-1 during the adsorption of CO on
CuO, but Engeldinger et al. [39] observed the characteristic
Cu(I)–CO band at 2,060–2,146 cm-1 on Cu–Y depending
upon the Cu content. Our XPS results demonstrated that
Cu? and Cu2? also exist on the surface of the catalyst.
Based on these results we cannot unambiguously assign
this peak. The intensity of the absorption at 2,141 cm-1
minimised in time during the reaction and according to the
XPS results the amount of Cu2? grew on the surface, we
may conclude that the CO absorbs at 2,141 cm-1 and is
Wilmhurst [48] the bands between 1,740 and 1,770 cm-1
could be assigned as the C=O vibration of gas phase MF.
Although both DMC and MF have absorption in this
region, and an intense band was detected at
1,293–1,291 cm-1 but when MF was adsorbed only a weak
absorption was detected at 1,230 cm-1 [48]. From these
data we may suppose that the bands at 1,767 and
1,293–1,291 cm-1 showed the presence of DMC on the
surface of the catalyst.
The features at 1,459 and 1,404 cm-1 detected at the
beginning of the reaction are in the C–H deformation
region and can be attributed to the anti-symmetric vibration
of CH3 and CH2 groups of DMM, respectively [49].
The band at 1,610 cm-1 is due to the adsorbed water
formed in the reaction.
bonded rather to Cu? than to Cu2?
.
In order to assignate the bands formed during the reac-
tion below 2,000 cm-1 we registered the infrared spectra of
the main products on Cu/Norit catalyst. After the adsorp-
tion of DMC at room temperature bands were observed in
the 2,000–1,100 cm-1 region at 1,781, 1,768, 1,462, 1,454,
1,293 cm-1. Similar spectra were recorded after the
adsorption of DMC on multi-walled carbon nanotube
supported Cu sample [25]. The bands at 1,781 and
1,768 cm-1 could be attributed to m(C=O), those at 1,462
and 1,454 cm-1 to das(–CH3) in the cis–trans and cis–cis
structure of DMC. The intensive absorption at 1,293 cm-1
could be assigned to the mas(O–C–O) vibration. Kar et al.
[46] and Bohets and van der Veken [47] used a similar
assignation for identifying the absorption bands obtained
after DMC adsorption. In the 2,000 and 1,100 cm-1 region
in the spectra of MF adsorbed on Cu/Norit between peaks
were detected at 1,768, 1,754 and 1,743 cm-1 and a weak
absorption was observed at 1,208 cm-1. According to
The interaction of CH3OH and Cu is summarised briefly
for the assignation of the bands below 1,100 cm-1. Pre-
vious studies demonstrated that methanol interacted
weakly on clean Cu surface. It is necessary to activate the
surface by partial oxidation; 20 % surface coverage by
oxygen was required to achieve the maximum adsorption
of methanol [50]. Methanol was found to adsorb disso-
ciative on Cu? by the formation of methoxy species
without additional supply of oxygen [44]. These findings
demonstrate that the activation of methanol and CO occurs
on the same sites, on Cu?. The methanol adsorbed on Cu2O
produces bands at 1,064 and 1,030 cm-1 that was attrib-
uted to the C–O stretching and CH3 rocking mode of
methoxy, respectively [51]. According to these results the
bonds observed at 1,068 an 1,036 cm-1 could be assigned
to the C–O stretching and CH rocking mode of methoxy
Fig. 8 Infrared spectra
registered during the
0,01
0,01
CH3 ? CO ? O2 reaction at
393 K on Cu/Norit in the 10th
(1), 20th (2), 30th (3), 90th (4)
and 150th (5) minutes
5
5
4
4
3
2
3
2
1
1
2200
2100
2000
1800
1500
1200
900
Wavenumber (cm-1)
123

G. Merza et al.
bonded to the oxidized Cu. Methoxy on Cu (100) surface
reportedly absorb at 984 cm-1 at 210 K [52], the C–O lies
perpendicular to the surface. The band observed at
1,021–1,026 cm-1 could be attributed to the C–O stretch-
ing vibration of CH3–O bonded to another structure.
Cu0/AC surface [27]. The calculated results indicated that
DMC forms in the reaction of MMC with methanol.
Another theoretical analysis demonstrated that on
Cu2O(111) surface the CH3O inhibit the CO adsorption
together with the CO insertion to methoxide species for the
production of MMC. Eventually the main reaction pathway is
the reaction of these species with methoxy to form DMC [53].
The present paper supports a recent assumption in the
literature for the DMC synthesis, namely that which
emphasizes the strong connection with oxidized Cu as also
demonstrated by the results of XPS studies (Fig. 5).
From the above mentioned DMC formation pathways
our results verify a few statements, particularly where
MMC species react with methanol. The MMC was stated to
form by the CO insertion to the methoxy species however it
was not found on the surface during the reaction. It could
be explained by the high reactivity and low surface con-
centration. Still the dissociative adsorption of methanol on
Cu? by the formation of methoxy species [44] and the
strong CO adsorption to Cu? [43] were described in the
literature.
3.5 The Possible Route of DMC Formation
Cu/Norit showed remarkable catalytic properties in the
oxidative carbonylation of methanol. The methanol con-
version reached 22 % during the reaction, the CO con-
version changed as the methanol conversion (Fig. 1a). The
CO2 formation rate first increased, then decreased, finally
remained nearly constant. The MF formation rate after
100 min of the reaction remained quasi the same. The
DMC formation started only after a 100 min induction
period followed by a rapid formation until reaching the
steady state (Fig. 2a).
XPS results showed that the reduced catalyst was oxi-
dized in the reaction, thus the surface Cu was oxidized to
Cu? and Cu2? (Fig. 5) and the amount of the latter species
increased with increasing reaction time. The XPS studies of
the pure Norit reveals that there was no significant change
taking place on the surface of the support (Fig. 6) although
the recorded O 1s spectra showed the change of the oxygen
content of the catalyst (Fig. 7).
Cu oxidization during the reaction was demonstrated by
our XPS results (Fig. 5). First mainly Cu? species are
formed on the surface, but later the amount of Cu2?
increased in time. As it is illustrated by the DMC formation
curve, when only metallic Cu is on the surface, in the first
minutes of the reaction DMC was not produced. Later
when only Cu? was detected on the catalyst DMC was also
not detectable. By increasing the amount of Cu2? the DMC
formation rate also increased but after some hours it slowly
lessened (Fig. 3). The fully oxidized sample was totally
inactive in the DMC synthesis. From these results we may
conclude that both Cu? and Cu2? are necessary for the
DMC synthesis, and in addition their ratio is decisive in the
DMC formation. The changes of the oxidation state of Cu
during the reaction could result in the changes of the DMC
formation rate in time (Fig. 2) on Cu/Norit. Below is a
proposed reaction pathway for DMC synthesis in which the
following species
Infrared measurements revealed that the CO absorption
is more favourable on Cu? than on Cu2? and that the
activation of methanol is also occurring on Cu?.
Earlier it was shown that Cu? ions are the active species
in the DMC formation by oxidative carbonylation [6, 16].
A large number of experimental studies about the mecha-
nism and kinetics of DMC formation on CuCl [3, 6, 16] or
on Cu–zeolites [49] have been reported. DMC formation
was stated to occur via two possible pathways. One of them
is the CO addition to di-methoxide species, the other is the
reaction of MMC and methanol to produce DMC [49].
In a theoretical investigation of the DMC formation on
AC the reaction species were presented to be adsorbed on
123

The Synthesis of Dimethyl Carbonate by the Oxicarbonylation
References
have an important role in the first two steps of the reaction.
Insertion of CO into the latter species resulted in the DMC
formation. Our XPS results support this theory, Zhang and
Bell [49] suggested similar mechanism.
1. Li C, Zhang X, Zhang S (2006) Chem Eng Res Des 87:1
2. Wen LB, Xin CY, Yang SC (2010) Appl Energ 87:115
3. Shivarkar AB, Gupte SP, Chaudhari RV (2005) J Mol Catal A
226:49
4. Ono Y (1997) Appl Catal A 155:133
5. Tharun J, Dharman MM, Hwang Y, Roshan R, Park MS, Park
DW (2012) Appl Catal A 419:184
The results presented on Fig. 4b strongly suggest that
the formation of MF occurs via another possible route. The
O 1s signal of the catalysts indicates that the amount of
carboxyl groups increased on the surface during the reac-
tion so the direct interaction of the adsorbed methoxy
species with the carboxyl groups could produce the MF.
Methanol adsorption on oxidized Cu on CuO surface
results in the production of formate species [54] while
either reacts with methanol and so MF is formed or
decomposes to CO2. These possibilities lead to the MF rate
having a maximum as the function of temperature or
contact time. The formation of MF and the decomposition
or desorption of this product reduce the Cu2? centers which
could reoxidize with oxygen.
6. Li Z, Su K, Cheng B, Ming J, Zhang L, Xu Y (2011) Catal
Commun 12:932
7. Litaiem Y, Dhahbi M (2012) J Mol Liq 169:54
8. Kim I, Ahn JT, Ha CS, Yang CS, Park I (2003) Polymer 44:3417
9. Seong P, Jeon BW, Lee M, Cho DH, Kim D, Jung KS, Kim SW,
Han SO, Kim YH, Park C (2011) Enzyme Microb Technol
48:505
10. Giannoccaro P, Ravasio N, Aresta M (1993) J Organomet Chem
451:243
11. Keller N, Rebmann G, Keller V (2010) J Mol Catal A 317:1
12. Zhang B, Ding G, Zheng H, Zhu H (2014) Appl Catal B 152:226
13. Tomishige K, Sakaihori T, Ikeda Y, Fujimoto K (1999) Catal Lett
58:225
14. Jiang CJ, Guo YH, Wang CG, Hu CW, Wu Y, Wang EB (2003)
Appl Catal A 256:203
15. Wang XJ, Xiao M, Wang SJ, Lu YX, Meng YZ (2007) J Mol
Catal A 278:92
16. Tomishige K, Kunimori K (2002) Appl Catal A 237:103
4 Conclusion
´
´
17. Stoica G, Abello S, Ramırez JP (2009) Appl Catal A 365:252
18. Sun J, Yang B, Wang X, Wang D, Lin H (2005) J Mol Catal A
239:82
In this work Cu, Ni and Cu–Ni/Norit catalysts were pre-
pared, characterised, their catalytic activities were com-
pared in the oxicarbonylation of methanol. Based on the
experiments we can conclude the following:
19. Han MS, Lee BG, Suh I, Kim HS, Ahn BS, Hong SI (2001) J Mol
Catal A 170:225
20. Han MS, Lee BG, Ahn BS, Kim HS, Moon DJ, Hong SI (2003) J
Mol Catal A 203:137
21. Liu T, Chang CS (2007) J Chin Inst Chem Eng 38:29
22. Ren J, Guo C, Yang L, Li Z (2013) Chin J Catal 34:1734
23. Jiang R, Wang S, Zhao X, Wang Y, Zhang C (2003) Appl Catal
A 238:131
Cu/Norit was a significantly effective catalyst for DMC
synthesis. Under the optimal reaction conditions at atmo-
spheric pressure the highest selectivity of DMC was about
60 % and the highest conversion of methanol was 22 % with
the DMC yield of 13.2 % in the steady state of the reaction
which is comparable with the previously reported ones at
high pressure. The CO2 formation rate was increasing with
the temperature while MF and DMC formation rate changed
through a maximum curve verifying that 393 K is the opti-
mal temperature. The formation rate of DMC and CO2 lin-
early enhanced by changing the contact time the formation
rate of MF went through a maximum which is supposed to
be the by-product of the reaction.
24. Bian J, Xiao M, Wang S, Lu Y, Meng Y (2009) Appl Surf Sci
255:7188
´
(2014) J Mol Catal A 393:117
´
´
´ ´ ´
}
25. Merza G, Laszlo B, Oszko A, Potari G, Baan K, Erdohelyi A
26. Bian J, Wie XW, Wang L, Guan ZP (2011) Chin Chem Lett 22:57
27. Ren J, Wang W, Wang D, Zuo Z, Lin J, Li Z (2014) Appl Catal A
472:47
28. Barthos R, Solymosi F (2007) J Catal 249:289
´
´
´
29. Barthos R, Szechenyi A, Koos A, Solymosi F (2007) Appl Catal
A 327:95
´
30. Solymosi F, Barthos R, Kecskemeti A (2008) Appl Catal A
350:30
The XPS measurement results revealed that Cu/Norit
catalyst went through changes. After reduction the catalyst
was oxidized in the reaction, furthermore, the ratio of Cu?
and Cu2? was determining in the DMC formation rate. The
C 1s peak showed no significant change on the surface of
the catalyst but from the O 1s spectra the change of the
carbon and O2 content was visible.
´
´ ´
31. Marban G, Lopez A, Lopez I, Solıs TV (2010) Appl Catal B
´
99:257
32. Tolmacsov P, Gazsi A, Solymosi F (2009) Appl Catal A 362:58
´ ´
33. Potari G, Madarasz D, Nagy L, Laszlo B, Sapi A, Oszko A,
Kukovecz A, Erdohelyi A, Konya Z, Kiss J (2013) Langmuir
´
´
´
´
´
´
}
´
29:3061
´
34. Madarasz D, Potari G, Sapi A, Laszlo B, Csudai C, Oszko A,
Kukovecz A, Erdohelyi A, Konya Z, Kiss J (2013) J Phys Chem
´ ´
´
´
´
´
´
}
´
Chem Phys 15:15917
35. Nam JK, Choi MJ, Cho DH, Suh JK, Kim SB (2013) J Mol Catal
A 370:7
36. Wagner CD, Riggs WM, Davis LE, Moulder JF, Mullenberg GE
(1992) Handbook of X-ray Photoelectron Spectroscopy, Perkin-
Elmer, USA
Acknowledgments Financial support of this work by TAMOP
4.2.2.A-11/1/KONV-2012-0047 and HU RO/0901/090/2.2.2 project
is gratefully acknowledged. This research was supported by the
European Union and the State of Hungary, co-financed by the
European Social Fund in the framework of TAMOP-4.2.4.A/2-11/1-
2012-0001 ‘National Excellence Program’.
123

G. Merza et al.
37. Saw ET, Oemar U, Tan XR, Du Y, Borgna A, Hidajat K, Kawi S
(2014) J Catal 314:32
´ ´
38. Guczi L, Stefler G, Geszti O, Sajo I, Paszti Z, Tompos A, Schay Z
45. London JW, Bell AT (1973) J Catal 31:32
46. Kar PB, Ramanathan N, Sundararajan K, Viswanathan KS (2012)
J Mol Struct 1024:84
(2010) Appl Catal A 375:236
39. Figueiredo JL, Pereira MFR, Freitas MMA, Orfa˜o JJM (1999)
Carbon 37:1379
40. Polovina M, Babic B, Kaluderovic B, Dekanski A (1997) Carbon
35:1047
47. Bohets H, Veken BJ (1999) Phys Chem Chem Phys 1:1817
48. Wilmhurst JK (1957) J Mol Spectrosc 1:201
49. Zhang Y, Bell AT (2008) J Catal 255:153
50. Wachs IE, Madix RJ (1978) J Catal 53:208
51. Lin SD, Cheng H, Hsiao TC (2011) J Mol Catal A 342:35
52. Camplin JP, McCash EM (1996) Surf Sci 360:229
53. Zhang R, Song L, Wang B, Li Z (2012) J Comput Chem 33:1101
54. Poulston S, Rowbotham E, Stone P, Parlett P, Bowker M (1998)
Catal Lett 52:63
´
41. Puziy AM, Poddubnaya OI, Socha RP, Gurgul J, Wisniewski M
(2008) Carbon 46:2113
´
42. Sanchez MS, Valiente AM, Ruiz AG, Nevskaia DM (2010) J
Colloid Interface Sci 343:194
43. Dandekar A, Vannice MA (1998) J Catal 178:621
44. Engeldinger J, Domke C, Richter M, Bentrup U (2010) Appl
Catal A 382:303
123