This document is the Accepted Manuscript version of a Published Work that appeared in final form in
ACS CATALYSIS copyright © American Chemical Society after peer review and technical editing by the
publisher. To access the final edited and published work see
https://pubs.acs.org/doi/abs/10.1021/acscatal.7b04198

Improving the stability of CeO2 catalyst by rare earth metal promotion and
molecular insights in the dimethyl carbonate synthesis from CO2 and
methanol with 2-cyanopyridine

Dragos Stoian†, ‡, §, Francisco Medina‡, Atsushi Urakawa*†
†

Institute of Chemical Research of Catalonia (ICIQ), The Barcelona Institute of Science and Technology, Av. Països
Catalans 16, 43007 Tarragona, Spain
‡

Department of Chemical Engineering, University Rovira i Virgili, Av. Països Catalans 26, 43007 Tarragona, Spain

*Email: aurakawa@iciq.es

1

Abstract
The unmatched efficiency of the direct dimethyl carbonate (DMC) synthesis from CO2 and
methanol over CeO2 catalysts in the presence of an organic dehydrating agent (2cyanopyridine, 2-CP) was recently reported with high DMC yield (>90%) in both batch
and continuous operations. However, the CeO2 catalyst gradually deactivates in the timescale of days due to suggested surface poisoning by 2-picolinamide (2-PA) produced by
hydration of 2-CP. This work seeks for active and stable CeO2-based catalysts and aims
to understand the material factors influencing the catalytic performance. Surface
modification of CeO2 by the addition of rare earth metal (REM) was found effective to
improve the catalyst stability. Surface basicity and reducibility of the Ce4+ species play
important roles in preventing catalyst deactivation by stabilizing the reactive methoxy
species in comparison to the poisoning species (2-PA or species alike). This has been
evidenced by in situ ATR-IR spectroscopy. CeO2 materials promoted with 1 wt% rare
earth metals (La, Gd, and Pr) greatly enhanced the catalyst stability while retaining the
high catalytic activity of CeO2. Among them, 1 wt% Pr promotion to CeO2 was the most
effective, affording 35% higher DMC yield in comparison to bare CeO2 after 150 h time
on stream under the optimized reaction condition of 30 bar and 120 °C.
Keywords:

carbon

dioxide,

methanol,

dimethyl

carbonate,

cyanopyridine, rare earth metal, surface properties, in situ ATR-IR.

2

cerium

oxide,

2-

1. Introduction
One of the major global challenges urged by the climate change and the desired paradigm
shift from the fossil-fuels dependent society to a more sustainable one is the carbon
management, i.e. how to efficiently utilize carbon dioxide (CO2) being emitted and
accumulated in the atmosphere. Catalytic transformation of CO2 into useful chemicals
such as organic and inorganic carbonates via non-reductive CO2 transformations has
been widely investigated for that aim.1-3 Among others, the synthesis of alkyl carbonates
through direct carboxylation reaction of alcohols like methanol (Reaction 1, synthesis of
dimethyl carbonate (DMC)) is very attractive in terms of atom economy.1, 4

DMC is a versatile chemical and represents an ecologically benign alternative to toxic
and corrosive methylating or carbonylating agents such as methyl halides (CH3X),
dimethyl sulfate (DMS), and phosgene (COCl2).1, 4-7 DMC is used in the synthesis of
polycarbonates and polyurethanes, as solvent replacing methyl ethyl ketone, butanone,
and tert-butyl acetate, and as electrolyte in lithium ion batteries.1-2, 4 Furthermore, DMC
has gained wide attention as an oxygenate fuel additive due to its high oxygen content,
thus increasing combustion performance.8
In the methanol carboxylation reaction with CO2, many heterogeneous catalysts have
been investigated to date. ZrO2, SiO2, Al2O3, TiO2, H-ZSM5, H-USY, H-MOR, ZnO, MoO3,
Bi2O3, MgO, Y2O3, HfO2, La2O3, Pr6O11, Ga2O3, GeO2, In2O3, Sb2O3, and hybrid ones
(e.g. H3PO4/V2O5 and Cu-Ni/V2O5-SiO2) have been tested, but without or with very low

3

activity (e.g. 0.3 mmol DMC gcat-1 h-1 for zirconia (ZrO2) materials).9-13 Occasionally,
dimethyl ether (DME) instead of DMC was produced via dehydration reaction of methanol,
which is commonly catalyzed by acidic sites, especially at elevated temperatures (>150
°C). According to previous studies, ZrO2 and ceria (CeO2) are the unique materials
effectively catalyzing the DMC formation.9-10,

14-21

However, the major challenge is

associated with the thermodynamics of Reaction 1, which is generally unfavorable due
to equilibrium limitation. Even under thermodynamically favorable high-pressure
conditions (ca. 400 bar), DMC yield is ca. 1% in accordance with the thermodynamic
expectation.22-24 Therefore, it is imperative to employ an effective mean to shift the
equilibrium to the product side for potential commercialization of DMC synthesis from CO2
and methanol. It is well known that water removal can shift the equilibrium improving DMC
yield. Chemical (e.g. 2, 2-dimethoxy propane) or physical water traps17, 25 and catalytic
membrane reactors26 have been employed for this purpose, but the improvement had
been relatively minor.
In 2013, Tomishige et al. described the DMC synthesis over CeO2 in a batch reactor using
a nitrile (2-cyanopyridine, 2-CP) as an organic dehydrating agent and reported an
exceptionally high DMC yield of 94%. In the reaction, CeO2 functioned as an effective
catalyst for both DMC synthesis and nitrile hydrolysis, forming the corresponding amide
(2-picolinamide, 2-PA, Reaction 2).27-28

4

They also verified the recyclability of 2-PA by dehydrating 2-PA back to 2-CP with
Na2O/SiO2 as catalyst. Together with the formation of DMC and 2-PA, very small amounts
of methyl picolinate (Me-PCN) and methyl carbamate (Me-CBM) were found as
byproducts. Me-PCN is co-produced with NH3 arising from the reaction between 2-PA
and methanol, while Me-CBM originates from the reaction between DMC and NH3
(Scheme S1, Supporting Information).
Inspired by the work of Tomishige et al., our group reported continuous DMC synthesis
over CeO2 in a fixed-bed reactor in the presence of 2-CP under low-to-high pressure
conditions (up to 300 bar).24 Although striking DMC yields >90% were achieved during
the continuous operation above 30 bar, a severe catalyst deactivation was noticed,
resulting in more than 50% drop in methanol conversion in 10 days with a small loss (89%) in DMC selectivity. Our following investigation allowed unravelling the cause of the
catalyst deactivation by visual and spectroscopic inspections using an optically
transparent fused quartz reactor. Most importantly, results showed molecularly adsorbed
2-PA over the deactivated CeO2 catalyst. However, the catalyst could be reactivated
when the adsorbed 2-PA was removed from the surface by a simple thermal treatment in
air at 300 °C which is close to the boiling point of 2-PA (284.1 °C).29
Motivated by the previous results and the superior DMC yield in the presence of 2-CP in
comparison to other dehydrating agents17, 30-31 using CeO2 as catalyst, the present study
seeks for stable and active CeO2-based catalyst in the reaction under continuous
operation as well as to understand material factors influencing the catalytic performance.
The uniquely high catalytic activity of CeO2 in this reaction may originate from the redox
capabilities of the cerium oxide in conjunction with its acid-base properties32-36. Therefore,
5

in this work common strategies to modify the chemical properties (acidity/basicity) of
CeO2 by the addition of metal/metal oxides to CeO2 to influence the oxygen storage
capacity, defect stability, and mobility of oxygen atoms were examined and their impacts
on catalyst stability were thoroughly investigated. Particularly, the effects of La,37-39 Gd,4041

and Pr42-45 promoters were investigated in depth to influence the oxygen

defects/mobility of CeO2 based materials while mildly altering the acid/base properties.
Besides the use of an ordinary stainless-steel tubular reactor, an optically transparent
fused-quartz tubular reactor was employed in order to visually assess the deactivating
state of the catalysts under the optimized reaction conditions (30 bar and 120 °C).29 The
catalysts before and after the reactions were characterized by several techniques to
identify critical material factors determining the catalytic activity and stability of the
materials. Furthermore, molecular insights into the surface adsorbed active species
responsible for the reaction and deactivation were gained by in situ attenuated total
reflection (ATR)-IR spectroscopy, elucidating the nature of surface poisoning process and
how REM promotion prevents it.
2. Experimental
2.1.

Materials

High surface area powder of CeO2 was kindly supplied by Daiichi Kigenso Kagaku Kogyo
Co. Ltd., Japan and used without any further treatment. The CeO2 material is nano-sized
(high surface area) but highly crystalline, fulfilling the important requisites as active
catalyst for the target reaction as reported by Tomishige et al. and for some CeO2
materials such materials states can be effectively obtained after calcination at 600 °C.19,
27

Rare earth metals (REM) promoted ceria materials (x wt% M on CeO2, where x = 0.25,
6

1, and 5 and M = La, Gd, and Pr) were synthesized by the incipient wetness impregnation
of the pristine CeO2 with an aqueous solution of the metal precursor (lanthanum (III)
nitrate hexahydrate, gadolinium (III) nitrate hydrate, and praseodymium (III) nitrate
hydrate, respectively). The metal nitrates (99.9% purity) were purchased from Alfa Aesar.
The impregnated materials were dried overnight at 80-90 °C in an oven and calcined at
400 °C for 4 h in a static air after a temperature ramp of 2 °C min-1. Methanol (≥99.9%,
HPLC grade) and 2-cyanopyridine (2-CP, 99%) were purchased from Sigma Aldrich. High
purity CO2 gas (>99.9993%) was purchased from Abelló Linde, Spain.
2.2.

Reaction system

The details of the reaction set-up and analytical method for the identification and
quantification of reaction products were described in our previous works.24, 29 In a typical
experiment, 300 mg of the catalyst (pelletized, crushed, and sieved to 200-300 µm particle
size) was loaded into the reactor tube (a stainless-steel tube with ID: 1.74 mm and OD:
3.17 mm). For the visual inspection experiments, a fused quartz reactor (ID: 2.0 mm and
OD: 3.0 mm) was used.29 The CO2 flow rate was kept at 6 NmL min-1. The mixture of
methanol and 2-CP at 2:1 molar ratio (methanol:CO2 = 1:2.5, molar ratio) was passed to
the reactor at 10 µL min-1 by means of an HPLC pump (Jasco, PU-2080 Plus).
All catalytic tests were performed under optimized reaction conditions where the DMC
yield is high while the reaction pressure is relatively mild (30 bar and 120 °C).29 We aimed
to perform the reaction for over 100 h, but occasionally the reaction had to be stopped
before due to the sealing issue of back pressure regulator (Jasco, BP-2080 Plus) caused
by the high-temperature kept at the regulator and the clogging or damage of the sealing

7

materials.29 Nevertheless, the reaction was performed more than 48 h to observe a clear
trend in catalytic activity, evaluate catalyst deactivation, and calculate its rate.

2.3.

Characterization

X-ray diffraction (XRD) measurements were made using a Siemens D5000 diffractometer
(Bragg-Brentano parafocusing geometry and vertical θ-θ goniometer) fitted with a curved
graphite diffracted-beam monochromator, incident and diffracted-beam Soller slits, a
0.06° receiving slit and scintillation counter as a detector. The angular 2θ diffraction range
was between 5 and 70°. The data were collected with an angular step of 0.05° at 9 s per
step (which resulted in a scan rate of 1°/3 min) and sample rotation with Cu-Kα radiation
at 40 kV and 30 mA.
Temperature programmed desorption (TPD) experiments were performed using a
Thermo Scientific™ TPDRO 1100 Advanced Catalyst Characterization System equipped
with a flow-through quartz reactor and a thermal conductivity detector with a tungsten
filament. In a typical experiment a catalyst material (100-150 mg) was firstly subjected to
the following pretreatments, (i) drying / thermal treatment under He at 400 / 600 °C for 1
h, (ii) CO2 (4% CO2 in He) / NH3 (5% NH3 in He) saturation at 80 °C for 30 min, and (iii)
flushing under He at 80 °C for 30 min. A TPD experiment consisted of ramping the
temperature from room temperature to 800 °C at a ramp of 10 °C min-1 under He flow at
20 mL min-1. The effluent gases were analyzed by a mass spectrometer (Pfeiffer Omnistar
GSD 301 C). An anhydrous magnesium perchlorate (Mg(ClO4)2) trap was used for CO2-

8

TPD experiments to absorb mainly H2O, whereas no trap was used for NH3-TPD
experiments.
Temperature programmed reduction experiments with hydrogen (H2-TPR), were
performed using the same apparatus as in the TPD experiments. 50-70 mg of a sample
was first pretreated at 100 °C for 1 h under N2 stream and subsequently heated from 50
to 800 °C at the ramp rate of 5 °C min-1 under the flow of 5% H2 in N2 at 20 ml min-1. A
soda lime (CaO + Na2O) trap was used to remove H2O and CO2.
The textural properties (BET surface area and BJH pore size distribution) of the materials
were determined by N2 physisorption experiments using Quantachrome Instruments,
AUTOSORB iQ. The material (100-150 mg) was firstly outgassed at 150 °C at 4 mbar to
clean the sample surface and pores prior to the measurements.
In situ ATR-IR measurements (ZnSe crystal, multiple reflections) were performed using a
PIKE HATR accessory on a Bruker Tensor 27 spectrometer equipped with a LN-MCT
detector at 4 cm-1 resolution. An aqueous slurry containing catalyst particles was
deposited via drop-casting method over the ZnSe crystal at room temperature and dried.
The reaction cell was then closed and the temperature was increased to 120 °C. The
catalyst was kept for 2-3 h under a flow of inert (N2). The measurements were performed
with the N2 flow saturated with methanol or methanol+2-CP mixture (15 mmol 2-CP in
methanol) at room temperature. For all the experiments the gas flow rate was 10 NmL
min-1.
Ex situ Raman measurements were carried out on a Thermo Scientific™ Nicolet™ iS™50
FT-IR spectrometer equipped with 1064 nm near-infrared excitation laser. The instrument

9

was preferably used to avoid the strong fluorescence shown by the samples after the
catalytic test with 532/785 nm laser.29 Ex situ Raman measurements (for fresh materials)
were carried out on a BWTEK dispersive i-Raman spectrometer equipped with 532 nm
excitation laser and a TE-cooled linear array detector.
FT-IR measurements of the samples before and after the reaction were performed on a
Bruker Alpha spectrometer equipped with a DTGS detector at 2 cm-1 resolution in an ATR
sampling configuration with a diamond internal reflection element.
Thermogravimetric analysis (TGA) of the samples before and after the reaction was
performed on a Mettler Toledo TGA/SDTA 851 equipment. In a typical experiment, 3-5
mg of a catalyst placed in an alumina crucible were subjected to heating from 25 to 800
°C at a ramp rate of 5 °C min-1 under a flow of synthetic air at 50 mL min-1.
3. Results and discussion
3.1.

Effects of rare earth metal promotion

We have thoroughly examined the effects of surface modifications by REM addition on
the stability of CeO2 catalyst during the DMC synthesis and on the physicochemical
properties of the catalyst materials. Different approaches to improve the stability, such as
Ce-Zr solid-solutions and precious metals (PM) promoted CeO2 (i.e. 1 wt% Rh, Ru, and
Pd) were also studied but these approaches were not effective in improving the stability;
rather they suppressed the catalytic activity of CeO2. Therefore, these results are
described in detail only in Supporting Information.
Figure 1a depicts the effects of La addition to CeO2 (0.25, 1, and 5 wt% La) on the longterm stability of the catalytic activity (methanol conversion, XMeOH and DMC selectivity,
10

SDMC) in the continuous DMC synthesis. Initially, slightly lower XMeOH values (ca. 88%)
were detected for all La-promoted CeO2 materials as compared to that of the bare CeO2
(92.5%), but all La-containing catalysts performed better than CeO2 after 60 h. About 74%
XMeOH was observed for 0.25 wt% La-CeO2 at 71 h and 5 wt% La-CeO2 showed ca. 75%
XMeOH at 86 h of the reaction. Remarkably, the most stable catalyst within the La-series
was 1 wt% La-CeO2 which maintained a high activity (>80% of XMeOH) even at 120 h, in
contrast to ca. 60% XMeOH with CeO2 at the same time on stream. The La addition also
impacted on the DMC selectivity. Initially, the higher the La loading, the poorer the SDMC
(>99% for bare CeO2 and ca. 85% for the 5 wt% sample). The SDMC values leveled off
with time to ca. 92-93% and the most stable sample, 1 wt% La-CeO2, displayed 94% SDMC
at 120 h of reaction.
The effects of Gd addition (0.25, 1, and 5 wt% Gd) to CeO2 on the long-term reactivity
are compared in Figure 1b. 0.25 and 1 wt% Gd catalysts exhibited a higher level of initial
XMeOH compared to the La-promoted catalysts and at the same level as that of CeO2.
However, further increase of Gd loading to 5 wt% lowered the initial XMeOH to 87%.
Importantly, all the Gd-containing catalysts displayed better long-term stability than CeO2.
81-82% XMeOH was observed for 0.25 and 5 wt% Gd-CeO2 at 68 and 50 h of the reaction,
respectively. Similarly to the case of the La-promoted CeO2, the most stable catalyst was
1 wt% Gd-CeO2 which maintained a very high activity (84% XMeOH) even at 100 h of the
reaction, which is significant considering 63% XMeOH using CeO2 at 100 h. Furthermore,
Gd addition resulted in high and stable SDMC (mostly >98%) at all loadings over the
reaction time tested. Based on the superior stability of the catalytic activity at high SDMC,
1 wt% was found to be the optimal Gd loading for the DMC synthesis.

11

Figure 1. Effects of REM promoters on the stability of methanol conversion (XMeOH, lower panels) and DMC selectivity
(SDMC, upper panels) in the DMC synthesis from CO2 and methanol in the presence of 2-CP at 120 °C, 30 bar for a. LaCeO2 (left panels), b. Gd-CeO2 (middle panels), and c. Pr-CeO2 (right panels). Different symbols are used to distinguish
the data points of catalyst materials as follows: black circles - bare CeO2, red squares – 0.25 wt% REM, green stars –
1 wt% REM, and blue triangles – 5 wt% REM. The dashed lines represent the trend lines to guide eyes for SDMC and
to calculate deactivation rates of XMeOH (Figure 2) with a linear fitting.

Furthermore, the effect of Pr addition (0.25, 1, and 5 wt% Pr) to CeO2 on the reactivity
was investigated (Figure 1c). The initial XMeOH slightly decreased more notably at
increased Pr loading, i.e. 92% (0.25 wt%), 90% (1 wt%), and 88% (5 wt%). However,
similar to Ga-promoted catalysts, XMeOH for all Pr-promoted catalysts became higher than
that of CeO2 after ca. 30 h due to the higher long-term stability. 83% XMeOH was observed
for 0.25 wt% Pr-CeO2 at 60 h, while 5 wt% Pr-CeO2 showed 80% XMeOH at 88 h. As in the
case of the other REM addition, the most stable catalyst was once again confirmed to be
at 1 wt% loading. The catalyst could maintain remarkable 82% XMeOH in comparison to
50% observed for CeO2 at 150 h. Moreover, Pr addition showed similarly stable SDMC as
12

observed for the Gd-promoted catalysts. 0.25 and 1 wt% Pr catalysts were highly
selective to DMC (>98% at ca. 150 h of the reaction for the 1 wt% sample), whereas 5
wt% Pr catalyst showed comparably lower SDMC, especially during the initial phase.
The above study of REM (La, Gd, Pr) promotion to CeO2 clearly shows its positive
impacts, leading to higher stability in both activity (XMeOH) and DMC selectivity, thus
significantly improving the overall performance. When 1 wt% Pr-CeO2 and CeO2 are
compared, an increase of around 35% in the DMC yield for the Pr-promoted catalyst was
obtained at ca. 150 h, which is significant and of high relevance in practice.
Figure 2 summarizes the results on the catalyst stability in terms of deactivation rate,
expressed by the drop in XMeOH (%) per hour. Since the deactivation was reasonably linear
for all REM-promoted samples (Figure 1), the slope was calculated with the least squares
method assuming a linear trend line of XMeOH. Obviously, the incorporation of REM was
positive in terms of long-term catalytic activity and the positive effects were more evident
for Gd- and Pr-CeO2. As noted above, the best materials were 1 wt% REM catalysts,
showing the lowest rate of deactivation (0.080, 0.075, and 0.055 % h-1 for La-, Gd-, and
Pr-CeO2 respectively) in comparison to 0.253 % h-1 of CeO2. In addition, it should be
reminded that 1 wt% Gd- and Pr-promoted catalysts were the best materials in terms of
long term stability in SDMC. When the REM loading is higher (at 5 wt%), the catalytic
activity is lowered due to more influential effects of pure REM oxides (e.g. La and Pr)
which show intrinsically lower catalytic activity compared to CeO2 for the hydration of 2CP to 2-PA.46

13

Figure 2. Rate of deactivation during the DMC synthesis (120 °C, 30 bar) expressed by the drop in XMeOH (%) per hour
for pristine CeO2 and REM-promoted CeO2 materials.

3.2.

Characterization of pristine and as-synthesized catalysts

The stability studies presented in section 3.1 showed the positive effects of REM as
promoter of CeO2 in the DMC synthesis in the presence of 2-CP, especially at 1 wt%
loading. In order to identify the origin of the enhanced stability of the REM-promoted
catalysts, the chemical and structural properties of the pristine and as-synthesized
catalysts as well as spent catalysts were investigated in detail.
Structural and textural properties
The XRD patterns of the pristine CeO2 and as-synthesized materials incorporating REM
are presented in Figure S2 (Supporting Information). CeO2 showed the peaks
corresponding to (111), (200), (220), and (311) planes of cubic fluorite phase. The
materials with REM showed negligible structural changes, indicating that metal promoters
were finely dispersed on the surface of CeO2.
14

The BET surface area and pore volume of all materials are presented in Table S1
(Supporting Information). Addition of REM to CeO2 consistently decreased the surface
area (the higher the loading, the lower the surface area), while the materials containing 5
wt% REM showed a clear reduction in the pore volume. Importantly, the changes in the
textural properties and bulk structure of the materials are not correlated with the catalytic
activity and stability.
Raman investigation
To elucidate the effects of REM addition on the CeO2 structure, the materials were further
characterized by Raman spectroscopy (Figure S3). The results suggest that the REMs
are atomically incorporated into CeO2 judging from the absence of REM sesquioxide
bands47 near 400-420 cm-1, corroborated with a minor shift and with an intensity change
of the F2g Raman active band of CeO2 at ca. 460 cm-1. Moreover, the characteristic band
due to oxygen vacancies of CeO2 at ca. 570 cm-1 emerges upon REM-doping.48 This
band is more pronounced at higher amount of doped REM. Among REM-CeO2 materials,
Pr-CeO2 shows the highest intensity ratio of the bands at 570 and 460 cm-1 (I570/I460),
commonly used as the measure of oxygen vacancies.49 Another indication in this direction
is the small band around 250 cm-1 which is more prominent for Pr-CeO2. These features
imply that in case of Pr-CeO2, oxygen vacancies are located closer to the surface where
catalysis takes place in comparison to La- and Gd-CeO2. Therefore, it can be inferred that
Pr is the best promoter to effectively modify the CeO2 surface structure by its atomic-level
dispersion, creating oxygen vacancies near the surface. Also 5 wt% Pr addition likely
leads to over-modification and the loss of active CeO2 surface required for the reaction
due to intrinsically lower activity of PrO2 compared to CeO2.46 Creation of oxygen
15

vacancies on CeO2 can lead to the creation of a unique surface state which enhances the
surface exposed Ce atoms and improves the stability of the catalyst as discussed later.

Basicity and acidity of the catalysts
As mentioned in Introduction, several acid-base heterogeneous catalysts have been
tested in the direct carboxylation of methanol with CO2. While the basic sites are generally
required to activate CO2 molecule, both acidic and basic sites are often reported to be
mandatory for methanol activation (CH3O- and CH3+ formation). It has been widely
reported that ZrO2, CeO2, and CeO2–ZrO2 solid solution show both acidic and basic
characters.50 CeO2 possesses Lewis acidic (Ce4+) and basic (O2-) sites and the material
may have exactly the right acidity-basicity balance to catalyze the reaction, more precisely
the two reactions of DMC synthesis and hydration reaction of 2-CP.27-28 Besides the acidbase property, the unique feature of the CeO2 material compared to many other oxides is
the reducibility of the catalyst and lattice/surface oxygen mobility. It is not yet clear but the
latter features may play pivotal roles in DMC synthesis and related to the catalyst stability.
Similarly, it has been previously suggested that addition of Cu in low amount (0.1 to 0.5
wt%) may work cooperatively with CeO2 catalysts to accelerate the partial reduction of
Ce4+ sites to Ce3+, enhancing the DMC yield by 20-60%.51 For these reasons, acidity and
basicity characterization by NH3- and CO2-TPD and material reducibility by H2-TPR were
performed to evaluate possible correlations between material properties and catalytic
performance.

16

According to Hutter et al. 52 who investigated the adsorption mechanism of CO2 on CeO2
(111) by DFT calculations, three stable configurations of CO2 on the CeO2 surface have
been identified: (i) monodentate carbonate, (ii) bidentate carbonate, and (iii) linearly
adsorbed species. Among these three configurations, the linear species are represented
as a physically adsorbed CO2 (no hybridization of orbitals of the linear CO2) and the
monodentate species are the most stably adsorbed ones. On the other hand, both monoand bi-dentate carbonate species are represented by bent CO2 configurations having the
C atom chemically bonded to O atom on the surface. The stronger the interaction of CO2
with the CeO2 surface, the higher its desorption temperature, i.e. the low (20-200 °C),
medium (200-450 °C), and high temperature (>450 °C) desorption peaks are assigned to
linear, bidentate and monodentate carbonates, respectively.53 In the same manner, the
low (20-200 °C), medium (200-450 °C), and high temperature (>450 °C) desorption peaks
in NH3-TPD measurements will be attributed to the weak, medium, and strong acid sites,
respectively.
CO2- and NH3-TPD profiles of all catalysts are presented in Figure S4 in Supporting
Information and the amounts of adsorbed-desorbed CO2 and NH3 are summarized in
Table S2 (Supporting Information), respectively. Generally, the addition of a foreign atom
on/in the CeO2 structure reduced the amount of adsorbed CO2/NH3 and thus the total
number of acidic and basic sites. In other words, the bare CeO2 material is more acidic
and more basic compared to the doped/promoted CeO2 materials, judging from the
number of the sites. There were clear and consistent trends in the changes of acidity and
basicity by the addition of REM to CeO2 (Figure S4 B). In all cases, the total number of
acidic and basic sites became smaller, and the degree of the decrease was less at lower

17

REM loading without affecting the characteristics of the strength of acidity and basicity,
as evident from the similar desorption profiles of CO2- and NH3-TPD for the 0.25 wt%
REM-CeO2 to those of CeO2. Upon increasing the REM loading to 1 and 5 wt%, there
were notable changes in the nature of the strong acidic and basic sites with the formation
of two types of strong basic sites (desorption at ca. 500 and 700 °C). For acidic sites,
mildly strong sites (desorption at ca. 500 °C) were reduced but strong sites (at ca. 750
°C) were enhanced. Thus, it can be concluded that higher loading of REM reduces the
number of weak-moderate acidic/basic sites but increases that of strong acidic/basic
sites.
With the aim to identify the role of acidity, basicity and their synergy in the reaction, we
evaluated the relations between acidity, basicity, or a unifying parameter (i.e.
basicity/acidity ratio calculated from the total amount of the adsorbed probe molecules)
of the examined materials and their catalytic performance and stability. When the
basicity/acidity ratio is larger than 1 (i.e. more basic catalysts; CeO2 and REM-promoted
CeO2), all the materials display good catalytic activity at the initial phase of the reaction.
In contrast, when the basicity/acidity ratio is <1 (i.e. more acidic) the materials such as
CeO2-ZrO2 solid solutions, ZrO2 and PM-CeO2 series exhibited lower activity (Figure 3).
It is widely reported that DMC synthesis from CO2 and methanol requires the presence of
both acidic and basic sites.10 This study comes as a confirmation to that statement as
well as an indication of the importance of a proper balance between acidity and basicity
of catalyst. Weak acidity is important in the selective DMC synthesis since the formation
of common by-product, dimethyl ether (DME), is more easily catalyzed over strong acid
sites, especially at higher temperatures.54 Moreover, 2-PA (weak base) adsorption over

18

the acidic sites of CeO2 leads to a severe catalyst deactivation.29 Importantly, basicity
seems playing more important roles by activating the CO2 molecule and formation of
CH3O- species and monodentate methyl carbonate (MMC, CH3OCOO-metal atom) as
possible reaction intermediates.9, 19, 55 The relatively high basicity of CeO2 against its
acidity is mildly influenced by the addition of REM unlike the materials with comparably
lower basicity when Zr and PM are added (Table S2, Supporting Information). This
basicity/acidity ratio is thus a good indicator to evaluate the activation capability of catalyst
for the reaction and it should be higher than the threshold (in our case >1).

Figure 3. Initial DMC yield against basicity/acidity ratio for all examined materials of this study at 120 °C, 30 bar.
Different symbols are used to distinguish the data points of catalyst materials as follows: squares for pure CeO2 (green),
pure ZrO2 (red) and CeO2-ZrO2 solid solutions (orange), blue stars for PM-CeO2, and the black circles for REM-CeO2.
The data points can be extracted from Figure S4/Table S2.

19

Figure 4. Rate of deactivation in the DMC synthesis plotted against (a) acidity and (b) basicity of CeO2 and REM-CeO2
materials. Filled circles are for pristine CeO2, while the half-empty circles correspond to the REM promoted CeO2. The
orange rectangles mark the values for the 1 wt% REM promoted materials.

Furthermore, possible correlations of the rate of deactivation with the parameters related
to acidity or basicity were sought for the promising REM-promoted materials. Figure 4
presents the rate of deactivation for CeO2 and REM-CeO2 as a function of the basicity
and acidity expressed by the total number of sites determined by the CO2- and NH3-TPD

20

experiments. Albeit no clear correlation was found for the deactivation rate with acidity,
the influence of the basicity on the deactivation was consistent; the lower the basicity, the
lower the rate of deactivation. This implies that basicity likely induces the surface
poisoning and thus deactivation, which is rather non-intuitive due to the expected stronger
interaction of 2-PA with acidic sites. This point will be discussed in detail later with the
identification and quantitative analysis of surface adsorbed species during the reaction.
Reducibility of the catalysts
The reducibility of the CeO2-based catalysts reflected by the oxygen mobility56-58 were
studied by H2-TPR. The H2-TPR profiles of pristine CeO2 and REM-promoted CeO2 are
presented in Figure S5 and Figure S6 (Supporting Information), respectively. Figure 5
presents a comparison of the profiles for the best performing materials (1 wt% REMCeO2) and pristine CeO2. Table S3 (Supporting Information) summarizes the H2 uptake
of all the samples investigated.
The reduction of CeO2 is known to proceed in two steps; the first one at ca. 500 °C,
corresponding to the reduction of surface Ce (IV) atoms more relevant for the catalytic
processes, and the second one at ca. 800 °C, corresponding to the bulk reduction of the
material (elimination of O2- from the lattice and formation of Ce2O3).34, 58-59 The reducibility
of ceria is known to be strongly dependent on the CeO2 crystallite size.33, 60 The CeO2
used in this study has a high surface area with a crystallite size of ca. 10 nm (as calculated
from the Scherrer equation using the (111) reflection peak) and it presents clearly the two
major peaks assignable to surface reduction (peak at 487 °C) and bulk reduction (peak
at 816 °C) processes (Figure S5).

21

Figure 5. H2-TPR profiles (corresponding to the surface reduction region) for CeO2 and 1 wt% REM-CeO2 materials.

Clear and consistent changes in the material reducibility were observed for the REM
promoted CeO2 materials (Figure S6). At 0.25 wt% REM loading, the changes in
reducibility were not pronounced. At 1 and 5 wt% REM loading, the appearance of highly
reducible features of the materials was evident with a new shoulder-peak at ca. 310-380
°C for all REM types (5 wt% La displayed two shoulders). Interestingly, the presence of
such a shoulder-peak near the low temperature CeO2 reduction has been attributed to
the existence of Ce4+ located in different chemical environment, such as the subsurface
region.61-63 Even more importantly, in the case of the catalysts with high stability in the
reaction (i.e. 1 wt% REM loading) the surface reduction peak was shifted to slightly lower
temperatures as compared to the case of CeO2 (Figure 5), reported previously due to the
enhanced redox property of CeO2, for example by the function of Pr.64 Precisely, the low
temperature peak which started to rise at 235 °C for CeO2 shifted to lower reduction-onset
temperature of 210, 170, and 220 °C for 1 wt% La-, Gd-, and Pr-CeO2, respectively.
22

Among 5 wt% REM-CeO2 materials, only the La-CeO2 showed the same behavior
reducing the onset temperature of surface reduction at ca. 160 °C, whereas for Gd- and
Pr-CeO2 showed higher onset temperature compared to the corresponding 1 wt% REMCeO2. Also, it is worth noting that 0.25 and 1 wt% REM loading did not affect considerably
the shape of the surface reduction peak, while at 5 wt% REM loading, the materials
displayed a small decrease of the peak. This is in a good agreement with the textural
properties where increasing the REM loading to 5 wt% resulted in a decrease of ca. 25%
in the SBET, indicating less CeO2 available on the surface.
In summary, at 1 wt% REM loading the reducibility of CeO2 is enhanced while retaining
the good reducibility of surface CeO2 which may be responsible for the catalytic activity.
On the other hand, at 0.25 wt% REM loading does not sufficiently influence material
properties determining the catalytic activity (i.e. acid-base properties and reducibility),
while 5 wt% REM loading leads to an over-enhanced basicity and diminished active
surface area (notable loss in surface area and pore volume) besides intrinsic lower activity
of REM oxides compared to CeO2.46
3.3.

Characterization of pristine and 1 wt% REM-CeO2 before and after the
reaction

The long-term stability of CeO2 catalyst in the direct DMC synthesis in the presence of 2CP was markedly improved using REM promoters (i.e. La, Gd, and Pr), but there was still
a gradual and substantial deactivation of the catalyst materials over more than 100 h of
reaction. To better understand and explain the effects of the REM promoters and their
functions during the reaction, bare CeO2 and 1 wt% REM-CeO2 after the reaction were
characterized by visual, spectroscopic and thermal analyses. The reactions (120 °C, 30
23

bar, for 30 h) were performed in a fused quartz reactor to enable visual inspection and
Raman spectroscopy of the materials within the reactor. Prior to the analyses, the
catalysts were thoroughly washed with methanol at 120 °C at 3 mL min-1 for 15 min in the
reactor to remove the crystallites of 2-PA deposited over the catalyst surface, accordingly
to our previous study.29 This washing procedure is necessary to identify strongly adsorbed
surface molecular species causing catalyst deactivation by Raman and ATR-IR
spectroscopic studies.29

Visual inspection and spectroscopic investigation
Figure 6 shows the photographs of the catalysts before and after the reaction with
subsequent hot methanol washing. Initially, all materials displayed a light-yellow color
characteristic of pristine CeO2 except Pr-CeO2 in a reddish tint. After the reaction, all the
materials turned darker and there seems a correlation between the degree of catalyst
deactivation and the degree of the color change. This tendency is the most noticeable for
CeO2 turning into dark maroon and less color change was observed for 1 wt% La-CeO2.
Even less change in color was observed for 1 wt% Gd-CeO2. 1 wt% Pr-CeO2 changed its
color to brick-red due to the reddish color of the as-synthesized material.

24

Figure 6. Photographs of CeO2 and REM-CeO2 materials before (top) and after (below) the reaction at 120 °C, 30 bar
for 30 h and a subsequent hot methanol washing at 120 °C for 15 min.

Interestingly, the degree of the color changes was in good agreement with the changes
in the surface area (SBET) and pore volume of the materials after the reaction (Table 1).
32.5% SBET loss was observed for CeO2, while the La, Gd, and Pr promoted samples
showed ca. 29% (La and Gd) and 24% (Pr) SBET loss. Moreover, the decrease in the pore
volume of the materials was perfectly in accordance with the rate of deactivation (Figure
2); CeO2 losing 31.5% of its pore volume and La, Gd, and Pr samples losing 18.3%,
12.1%, and 9.2%, respectively. No structural change in the crystallinity of the materials
was detected by XRD after 30 h of the reaction (Figure S9, Supporting Information).

25

Table 1. BET surface area and pore volume of CeO2 and REM-CeO2 before and after the reaction (120 °C, 30bar, for
30 h) and subsequent hot methanol washing at 120 °C.

Material
Pristine
CeO2
1 wt% LaCeO2
1 wt% GdCeO2
1 wt % PrCeO2

Surface area (m2 g-1)

Decrease Pore volume (cm3 g-1) Decrease
(%)
Before
After
(%)

Before

After

160

108

32.5

0.197

0.135

31.5

135

97

28.1

0.186

0.152

18.3

146

104

28.8

0.199

0.175

12.1

139

106

23.7

0.184

0.167

9.2

Ex situ Raman and ATR-IR measurements were performed to clarify the cause of the
catalyst deactivation by identifying the surface adsorbed species (Figure S10). Both
Raman and IR studies indicate that 2-PA-like species (more precisely, 2-PA monoanion,
Py-CONH-) are adsorbed on the surfaces in accordance with our previous study of bare
CeO2 where band assignments are described,29 implying that the REM promoters do not
influence the adsorption mode and origin of deactivation. Based on the intensity of the
Raman bands, the amount of 2-PA adsorbed on the surface of the pristine CeO2 was
suggested to be higher than that on the surface of REM-CeO2 materials. This is in good
agreement with the catalytic activity; catalyst deactivation is more pronounced when
surface is more poisoned by a larger amount of adsorbed 2-PA. It is important to mention
that the four materials re-gained their initial color and catalytic activity upon mild
calcination treatment at 300 °C in air as demonstrated for CeO2.29

26

TGA
Furthermore, thermogravimetric analysis (TGA) was performed for the catalyst materials
before and after the reaction and the results are summarized in Table 2 (TGA profiles are
shown in Figure S11, Supporting Information). All the materials before the reaction
showed only one desorption peak, assigned to the removal of surface impurities (i.e.
physically adsorbed water). On the other hand, an additional feature of a peak slightly
below 300 °C was observed for the materials after the reaction. This peak corresponds
to the 2-PA elimination from the CeO2 surface as the boiling point of 2-PA is 284.1 °C. It
is interesting to note that the peak slightly shifts towards lower temperatures (ca. 270 °C)
for 1 wt% REM-CeO2 promoted materials as compared that of pristine CeO2 (ca. 280 °C),
as determined by peak calculation taking the first derivative of the weight loss curve. This
is a good indication of weaker interaction between the 2-PA and the catalyst surface when
CeO2 is promoted with 1 wt% REM. The lower amount of 2-PA adsorption indicated by
the Raman study was also confirmed by TGA (Table 2) where the REM promotion clearly
reduced the degree of the weight loss for the catalysts after the reaction. These results
agree with the changes of surface area and pore volume; the less 2-PA adsorbed, the
less changes in the surface area and pore volume. REM promotion at the effective
amount (1 wt%) leads to the reduction in basicity (Figure 4) and decreases the interaction
between 2-PA and the catalyst surface. This results in reducing the rate of poisoning by
2-PA, thus enhancing the catalyst stability when 2-CP is used as dehydrating agent.
When the basicity is sufficiently high compared to the acidity, the high activity of CeO2
itself is retained also enhanced and become ideal for this reaction.

27

Table 2. Summary of TGA of the materials (CeO2 and 1 wt% REM) before and after the reaction.

Material
CeO2
1 wt% La-CeO2
1 wt% Gd-CeO2
1 wt% Pr-CeO2

Peak
no.
1
2
1
2
1
2
1
2

Before reaction
Weight
loss / %
7.23
5.34
5.77
5.49
-

Peak
max / °C
50.8
41
38.9
41.7
-

After reaction
Weight
loss / %
2.83
5.21
2.01
4.64
2.83
3.59
3.02
3.39

Peak
max / °C
31.5
277.5
30.5
272.9
30.5
268.5
39.3
268.5

Another important factor influencing the catalyst stability is the reducibility of the catalyst
surface, which was promoted by REM, especially at 1 wt%, as discussed above. Redox
properties change the charge state of atoms and this can consequently modulate the
acid-base properties. When the reducibility is enhanced while retaining the active CeO2
sites by REM (presumably those characterized by the H2-TPR peak <500 °C), it is
possible that the catalyst surface poisoned by 2-PA may change its redox state (Ce4+ 
Ce3+) flexibly under the reaction condition and create an environment facilitating 2-PA
desorption from the surface. This hypothesis is supported by the lower desorption
temperature of 2-PA in TGA for 1 wt% REM samples as discussed above. This would be
naturally positive for the catalyst stability and it is speculated that 1 wt% Gd or Pr
promotion is particularly effective in inducing such a function while retaining the active
catalytic sites of CeO2.

28

3.4.

Understanding the deactivation mechanism: In situ ATR-IR studies

To gain deeper molecular insights into the deactivation mechanism and consequently the
nature of active sites, we performed a series of in situ ATR-IR studies to identify the types
of methoxy species formed over the catalyst surfaces and their relation to surface
poisoning process by 2-PA or species alike. For comparison, we have studied chemical
interactions of methanol and 2-CP with the surfaces of CeO2 and 1 wt% Pr-CeO2. Figure
7 displays ATR-IR spectra obtained by alternatingly passing methanol vapor and N2 over
CeO2 or Pr-CeO2 at 120 °C. For both materials, as soon as the methanol vapor passes
over the catalysts, the methoxy bands at ca. 1110 and 1060 cm-1 appeared and they
gradually disappeared under N2 flow accompanying redshifts which are indicative of
stronger interaction of methoxy species with the surface at lower coverage. Marked
differences were observed for the two catalysts in the absorbance, the relative ratio of the
two bands, stability of the methoxy species under N2 and band positions observed.

29

Figure 7. (right) In situ ATR-IR spectra of surface species formed by alternatingly passing methanol vapor in N2 and
N2 over CeO2 (top) and 1 wt% Pr-CeO2 (bottom) at 120 °C. The background was taken in N2 before admission of
methanol. The scale is in absorbance. (left) The temporal evolution of the area of the two bands assigned to terminal
methoxy (t-OCH3, ca. 1110 cm-1) and bridged methoxy (b-OCH3, ca. 1060 cm-1).

It is widely acknowledged that methanol can be easily dissociated on the acid−base sites
of CeO2 to form surface methoxy species (on the Lewis acid sites, Cen+) and proton (on
the Lewis base sites, O2-) even at room temperature.65 The observed two bands can be
unambiguously assigned to the C–O stretching modes of terminal (t-OCH3, at ca. 1110
cm-1) and bridged (b-OCH3, ca. 1060 cm-1) methoxy species.19, 65 For the synthesis of
DMC from CO2 and methanol, t-OCH3 was reported to play key roles since monomethyl
carbonate (MMC, CH3O-COO-), the suggested intermediate, can be formed by the

30

reaction between CO2 and t-OCH3.19 On the other hand, the existence of b-OCH3 species
is related to the presence of oxygen defect sites on CeO2.66-68
Figure 7 evidences higher amounts of adsorbed methoxy species of both types over Prpromoted CeO2. This may be closely related to the less number of basic sites as observed
for 1wt% REM-CeO2 (Figure 4), which can be interpreted as lower number of surface
oxygen atoms of CeO2 and thus consequently a higher number of surface exposed Ce
atoms where the methoxy species can be formed. Interestingly, the temporal evolutions
of the band areas of the two methoxy species (Figure 7) show that the ratios of t-OCH3
to b-OCH3 over the surface are different for the two catalysts. The Pr promotion enhanced
the ratio of t-OCH3 to b-OCH3 from around 1 (CeO2) to around 1.5 (Pr-CeO2). Since defect
sites are expected to enhance b-OCH3, the enhanced relative population of t-OCH3
observed for Pr-CeO2 is counter-intuitive. A closer look into the spectral features (Figure
7) shows that the b-OCH3 band on Pr-CeO2 (1055 cm-1) is redshifted and broadened in
comparison to that on CeO2 (ca. 1060 cm-1). The broadness is indicative of the presence
of various similar geometrical configurations of adsorbed methoxy species, thus of the
existence of more defective adsorption sites over Pr-CeO2. Pr can be directly involved in
the enhanced interaction of methoxy with the surface as indicated by the redshift, but
possibly methoxy species may be preferably adsorbed on the hollow Ce sites on Pr-CeO2.
The latter interpretation is consistent with the increased number of t-OCH3 compared to
b-OCH3 observed for Pr-CeO2, since involving more Ce atoms by hollow site adsorption
(3 coordinating Ce atoms) than bridged site adsorption (2 coordinating Ce atoms) can
lead to less b-OCH3 species and thus increase t-OCH3 population comparably.

31

Another important observation during this methanol sorption study is the stability of the
methoxy species. Upon switching from methanol vapor to N2 at the reaction temperature
(120 °C), the number of surface methoxy species decreased, but to a different extent for
the two materials. Within the duration of our experiments (N2 flow for ca. 90 min), t-OCH3
disappeared almost completely, whereas about 40% of b-OCH3 (assuming the band area
is proportional to the concentration of the methoxy species) remained on the bare CeO2
after ca. 90 min of N2 flow, clearly showing higher stability of b-OCH3.69 Strikingly, the
stability of both methoxy species are greatly enhanced by Pr-promotion. The higher
stability of b-OCH3 compared to t-OCH3 was also confirmed for Pr-CeO2, showing even
>70% of the b-OCH3 present on the surface after ca. 90 min of N2 flow at 120 °C.
To further clarify the importance of the enhanced stability of the methoxy species in terms
of catalyst stability, we performed another in situ ATR-IR study by alternatingly passing a
vapor containing 2-CP and methanol and that of methanol over the two materials at 120
°C to monitor surface chemical processes causing catalyst poisoning (Figure 8).

32

Figure 8. In situ ATR-IR spectra of surface species formed by alternatingly passing the vapor of methanol + 2-CP and
that of methanol, both saturated in N2, over CeO2 (top) and 1 wt% Pr-CeO2 (bottom) at 120 °C. The background was
the material under N2. The scale is in absorbance.

For CeO2, as soon as the mixture of 2-CP and methanol passed over the catalysts the
appearance of characteristic t-OCH3 and b-OCH3 bands at ca. 1110 and 1060 cm-1 was
confirmed, followed by the gradual but significant decrease of the bands with time. This
decrease of the methoxy bands accompanied the emergence of the bands characteristic
of 2-PA anion (Py-CONH-), assigned as 1560 cm-1 (δNH), 1390 cm-1 (νC-N) and 750 cm-1

33

(δring), as the surface intermediate in the hydrolysis of 2-CP towards 2-PA. Upon switching
to methanol vapor as an attempt to “clean” the CeO2 surface, the spectrum remained
unaltered; the number of 2-PA-like species remained constant and the number of the
methoxy species could not be recovered. Another admission of the vapor of 2-CP and
methanol led to further decrease of the methoxy species and increase of the 2-PA-like
species in exchange. Between the two methoxy species, the decrease of the band of tOCH3, which is suggested to be responsible for DMC formation, was more pronounced.
It is interesting to note that this t-OCH3 band could be recovered slightly by passing the
vapor containing only methanol again, while b-OCH3 band could not as much. These
results prove that 2-PA-like species and methoxy species competitively adsorb on the
same surface sites of CeO2, thus causing surface poisoning by reducing the number of
catalytically active methoxy species. The surface sites responsible for the formation of bOCH3 are more resistant to the poisoning likely due to the stronger adsorption strength of
b-OCH3 than t-OCH3, while the recovery of the surface sites is difficult once they are
poisoned.
The identical study with Pr-CeO2 shows highly contrasting results with clear molecular
insights into the REM promotion effects (Figure 8). As in the case of CeO2, the 2-PA-like
species were formed with time when the vapor of 2-CP and methanol was passed over
Pr-CeO2 but to a much lesser extent. Obviously, the stability of both methoxy species
were greatly enhanced and thus made the surface highly resistant to the poisoning.
Passing methanol vapor could not recover the surface methoxy species, indicating that
when the sites are poisoned, the recovery is very difficult, in agreement with the slower
but continuous decrease in the catalytic activity of the REM-promoted catalysts.

34

3.5.

Active sites and deactivation mechanism

Based on the above studies, Figure 9 summarizes mechanistic views on the active sites
and deactivation of CeO2 during the DMC synthesis from CO2 and methanol in the
presence of 2-CP. Methanol molecules adsorb over CeO2 as t-OCH3 and b-OCH3 species
(Figure 9a). When 2-CP reaches the surface, it is adsorbed as 2-PA like species (Figure
9b). A possible structure of adsorbed 2-CP over the defect site of CeO2 was recently
reported by Tamura et al.65 and this state closely resembles the molecular structure of 2PA and most likely this or the chemical state where N atom is hydrogenated/protonated
is the state we have noted as 2-PA like species in this work. It is important to point out
that this state of surface species can block both t-OCH3 as well as b-OCH3 sites. It was
evidently shown that the surface sites of CeO2 is competitively accessed by methanol and
2-CP. Over the surface of pure CeO2, the 2-PA-like species are formed rapidly by
occupying the sites capable of methoxy formation as shown spectroscopically (Figure 8).
Since this 2-PA-like species hinders also the formation of t-OCH3 species, thus the
catalyst deactivates over time.
This study clearly demonstrated that REM-promotion could slow down the blocking of the
active sites by 2-PA-like species by strengthening the adsorption of methoxy species,
especially that of b-OCH3 species. REM promoters enhanced the number of surface
defects and sites allowing multiple coordination of Ce to stabilize the methoxy species.
Also, the enhanced redox properties of REM-CeO2 especially at 1 wt% REM loading
seem optimum to achieve the most efficient stabilization of methoxy species while
retaining sufficient number of t-OCH3 sites for DMC synthesis. Besides, the enhanced
oxygen mobility by REM may also slow down the formation of 2-PA-like species since the
35

oxygen defects in subsurface may facilitate conversion of the 2-PA-like species back to
2-CP.

Figure 9. Simplified reaction scheme for DMC synthesis over CeO2 in the presence of 2-CP. While DMC is formed by
the reaction between CO2 and t-OCH3 (a), the adsorption of 2-CP and/or the subsequent hydrolysis step produces an
intermediate (dark red) that leads to the poisoning of the methanol adsorption sites (b). Surface hydrogen
atoms/protons are not shown for the sake of brevity.

4. Conclusions
Effects of different promoters to CeO2 on the long-term behavior of catalytic performance
in continuous DMC synthesis from CO2 and methanol in the presence of a dehydrating
agent (2-CP) were evaluated. Among them, 1 wt% rare earth metal (REM: La, Gd, or Pr)promoted catalysts, especially Gd-CeO2 and Pr-CeO2, exhibited the highest stability
without deteriorating the catalytic performance of pristine CeO2. The visual, spectroscopic

36

(IR and Raman) and TGA investigations clarified that the surface adsorption of 2-PA
produced by the hydration reaction of 2-CP is the cause of the catalyst deactivation and
there was a clear relation between the amount of 2-PA adsorbed and the degree of
catalyst deactivation. The origins of the promoter effects were studied in detail by
characterizing the textural properties, acidity-basicity, and reducibility of the catalyst
materials. The right balance in basicity and acidity, especially sufficiently high
basicity/acidity ratio, is found necessary to achieve enhanced catalytic activity. The pure
CeO2 has the right acidity-basicity balance and this characteristic is retained for REMpromoted CeO2 materials. Regarding the catalyst stability, when the number of basic sites
is smaller and the catalyst reducibility is higher (i.e. easier to create surface defects and
increased mobility of sub-surface O atoms to/from the CeO2 surface), there were clear
improvements. Remarkably, compared to CeO2, 1 wt% Pr-CeO2 and 1 wt% Gd-CeO2
exhibited 35% higher DMC yield after 150 h and 25% higher DMC yield after 100 h of
time on stream, respectively. This greatly enhanced catalyst stability by the REM
promoters in the light of the previously found reactivation strategy (calcination at 300 °C
in air) render the process more attractive in practice to produce DMC from CO2 and
methanol. Furthermore, molecular insights into the catalyst deactivation and thus the
nature of active sites were uncovered by in situ ATR-IR spectroscopic studies. They
clearly show that REM promoters enhance adsorption strength of methoxy species and
thus prevents the 2-PA-like species from adsorbing and thus blocking the sites for DMC
synthesis.

37

Acknowledgements
We thank the Generalitat de Catalunya for financial support through the CERCA
Programme and recognition (2014 SGR 893) and MINECO (CTQ2016-75499-R
(AEI/FEDER-UE)) for financial support and support through Severo Ochoa Excellence
Accreditation 2014–2018 (SEV-2013-0319).
Supporting Information
The Supporting Information is available free of charge on the ACS Publications website
at …. Reaction schemes for the formation of observed bi-products, catalytic tests with Zr
and precious metal (Rh, Rh, Pd) promoted CeO2, XRD patterns (before and after the
reaction) and Raman spectra of CeO2 and REM-CeO2 materials, CO2-/NH3-TPD and H2TPR results, Raman and IR spectra of the catalysts after the reaction, TGA results of the
catalysts before and after the reaction, BET surface area and pore volume of all studied
materials.
AUTHOR INFORMATION
Corresponding author
aurakawa@iciq.es
Present address
§

The Swiss-Norwegian Beamlines (SNBL) at ESRF, CS40220, 38043 Grenoble

CEDEX 9, France

38

References
1.
Shaikh, A.-A. G.; Sivaram, S., Organic Carbonates. Chem. Rev. 1996, 96, 951–976.
2.
Sakakura, T.; Choi, J.-C.; Yasuda, H., Transformation of Carbon Dioxide. Chem. Rev.
2007, 107, 2365–2387.
3.
Sakakura, T.; Kohno, K., The Synthesis of Organic Carbonates from Carbon Dioxide.
Chem. Commun. 2009, 1312-1330.
4.
Ono, Y., Catalysis in the Production and Reactions of Dimethyl Carbonate, an
Environmentally Benign Building Block. Appl. Catal. A-Gen. 1997, 155, 133–166.
5.
Aresta, M.; Quaranta, E., Carbon Dioxide: A Substitute for Phosgene. ChemTech 1997,
27, 32-40.
6.
Keller, N.; Rebmann, G.; Keller, V., Catalysts, Mechanisms and Industrial Processes for
the Dimethyl Carbonate Synthesis. J. Mol. Catal. A-Chem. 2010, 317, 1-18.
7.
Delledonne, D.; Rivetti, F.; Romano, U., Developments in the Production and Application
of Dimethyl Carbonate. Appl. Catal. A-Gen. 2001, 221, 241–251.
8.
Pacheco, M. A.; Marshall, C. L., Review of Dimethyl Carbonate (DMC) Manufacture and
Its Characteristics as a Fuel Additive. Energ. Fuel. 1997, 11, 2–29.
9.
Tomishige, K.; Ikeda, Y.; Sakaihori, T.; Fujimoto, K., Catalytic Properties and Structure of
Zirconia Catalysts for Direct Synthesis of Dimethyl Carbonate from Methanol and Carbon Dioxide.
J. Catal. 2000, 192, 355–362.
10.
Tomishige, K.; Sakaihori, T.; Ikeda, Y.; Fujimoto, K., A Novel Method of Direct Synthesis
of Dimethyl Carbonate from Methanol and Carbon Dioxide Catalyzed by Zirconia. Catal. Lett.
1999, 58, 225-229.
11.
Tomishige, K., Direct Synthesis of Dimethyl Carbonate from Methanol and Carbon Dioxide
over Solid Oxide Catalysts. Curr. Top. Catal. 2002, 3, 81 - 101.
12.
Wu, X. L.; Xiao, M.; Meng, Y. Z.; Lu, Y. X., Direct Synthesis of Dimethyl Carbonate on
H3PO4 Modified V2O5. J. Mol. Catal. A-Chem. 2005, 238, 158–162.
13.
Wu, X. L.; Meng, Y. Z.; Xiao, M.; Lu, Y. X., Direct Synthesis of Dimethyl Carbonate (DMC)
Using Cu-Ni/VSO as Catalyst. J. Mol. Catal. A-Chem. 2006, 249, 93–97.
14.
Ikeda, Y.; Sakaihori, T.; Tomishige, K.; Fujimoto, K., Promoting Effect of Phosphoric Acid
on Zirconia Catalysts in Selective Synthesis of Dimethyl Carbonate from Methanol and Carbon
Dioxide. Catal. Lett. 2000, 66, 59-62.
15.
Ikeda, Y.; Asadullah, M.; Fujimoto, K.; Tomishige, K., Structure of the Active Sites on
H3PO4/ZrO2 Catalysts for Dimethyl Carbonate Synthesis from Methanol and Carbon Dioxide. J.
Phys. Chem. B 2001, 105, 10653–10658.
16.
Tomishige, K.; Furusawa, Y.; Ikeda, Y.; Asadullah, M.; Fujimoto, K., CeO2–ZrO2 Solid
Solution Catalyst for Selective Synthesis of Dimethyl Carbonate from Methanol and Carbon
Dioxide. Catal. Lett. 2001, 76, 71-74.
17.
Tomishige, K.; Kunimori, K., Catalytic and Direct Synthesis of Dimethyl Carbonate Starting
from Carbon Dioxide Using CeO2-ZrO2 Solid Solution Heterogeneous Catalyst: Effect of H2O
Removal from the Reaction System. Appl. Catal. A-Gen. 2002, 237, 103–109.
18.
Hofmann, H. J.; Brandner, A.; Claus, P., Direct Synthesis of Dimethyl Carbonate by
Carboxylation of Methanol on Ceria-Based Mixed Oxides. Chem. Eng. Technol. 2012, 35, 2140–
2146.
19.
Yoshida, Y.; Arai, Y.; Kado, S.; Kunimori, K.; Tomishige, K., Direct Synthesis of Organic
Carbonates from the Reaction of CO2 with Methanol and Ethanol over CeO2 Catalysts. Catal.
Today 2006, 115, 95–101.
20.
Aresta, M.; Dibenedetto, A.; Pastore, C.; Cuocci, C.; Aresta, B.; Cometa, S.; Giglio, E. D.,
Cerium (IV) Oxide Modification by Inclusion of a Hetero-Atom: A Strategy for Producing Efficient
and Robust Nano-Catalysts for Methanol Carboxylation. Catal. Today 2008, 137, 125–131.
39

21.
Aresta, M.; Dibenedetto, A.; Pastore, C.; Angelini, A.; Aresta, B.; Pápai, I., Influence of
Al2O3 on the Performance of CeO2 Used as Catalyst in the Direct Carboxylation of Methanol to
Dimethyl Carbonate and the Elucidation of the Reaction Mechanism. J. Catal. 2010, 269, 44–52.
22.
Santos, B. A. V.; Silva, V. M. T. M.; Loureiro, J. M.; Barbosa, D.; Rodrigues, A. E.,
Modeling of Physical and Chemical Equilibrium for the Direct Synthesis of Dimethyl Carbonate at
High Pressure Conditions. Fluid Phase Equilibr. 2012, 336, 41–51.
23.
Santos, B. A. V.; Pereira, C. S. M.; Silva, V. M. T. M.; Loureiro, J. M.; Rodrigues, A. E.,
Kinetic Study for the Direct Synthesis of Dimethyl Carbonate from Methanol and CO2 over CeO2
at High Pressure Conditions. Appl. Catal. A-Gen. 2013, 455, 219–226.
24.
Bansode, A.; Urakawa, A., Continuous DMC Synthesis from CO2 and Methanol over a
CeO2 Catalyst in a Fixed Bed Reactor in the Presence of a Dehydrating Agent. ACS Catal. 2014,
4, 3877–3880.
25.
Santos, B. A. V.; Silva, V. M. T. M.; Loureiro, J. M.; Rodrigues, A. E., Adsorption of H2O
and Dimethyl Carbonate at High Pressure over Zeolite 3A in Fixed Bed Column. Ind. Eng. Chem.
Res. 2014, 53, 2473–2483.
26.
Li, C.-F.; Zhong, S.-H., Study on Application of Membrane Reactor in Direct Synthesis
DMC from CO2 and CH3OH over Cu–KF/MgSiO Catalyst. Catal. Today 2003, 82, 83–90.
27.
Honda, M.; Tamura, M.; Nakagawa, Y.; Sonehara, S.; Suzuki, K.; Fujimoto, K.-i.;
Tomishige, K., Ceria-Catalyzed Conversion of Carbon Dioxide into Dimethyl Carbonate with 2Cyanopyridine. ChemSusChem 2013, 6, 1341–1344.
28.
Honda, M.; Tamura, M.; Nakagawa, Y.; Nakao, K.; Suzuki, K.; Tomishige, K., Organic
Carbonate Synthesis from CO2 and Alcohol over CeO2 with 2-Cyanopyridine: Scope and
Mechanistic Studies. J. Catal. 2014, 318, 95–107.
29.
Stoian, D.; Bansode, A.; Medina, F.; Urakawa, A., Catalysis under Microscope: Unraveling
the Mechanism of Catalyst De- and Re-Activation in the Continuous Dimethyl Carbonate
Synthesis from CO2 and Methanol in the Presence of a Dehydrating Agent. Catal. Today 2017,
283, 2-10.
30.
Eta, V.; Maki-Arvela, P.; Leino, A.-R.; Kordas, K.; Salmi, T.; Murzin, D. Y.; Mikkola, J.-P.,
Synthesis of Dimethyl Carbonate from Methanol and Carbon Dioxide: Circumventing
Thermodynamic Limitations. Ind. Eng. Chem. Res. 2010, 49, 9609–9617.
31.
Zhang, Z.-F.; Liu, Z.-W.; Lu, J.; Liu, Z.-T., Synthesis of Dimethyl Carbonate from Carbon
Dioxide and Methanol over CexZr1-xO2 and [EMIM]Br/Ce0.5Zr0.5O2. Ind. Eng. Chem. Res. 2011, 50,
1981–1988.
32.
Trovarelli, A., Catalysis by Ceria and Related Materials. Imperial College Press: 2002.
33.
Trovarelli, A., Catalytic Properties of Ceria and CeO2-Containing Materials. Catal. Rev.
1996, 38, 439-520.
34.
Trovarelli, A., Structural and Oxygen Storage/Release Properties of CeO2-Based Solid
Solutions. Comment. Inorg. Chem. 1999, 20, 263-284.
35.
Gorte, R. J., Ceria in Catalysis: From Automotive Applications to the Water–Gas Shift
Reaction. AIChE J. 2010, 56, 1126–1135.
36.
Steele, B. C. H.; Heinzel, A., Materials for Fuel-Cell Technologies. Nature 2001, 414, 345352.
37.
Reddy, B. M.; Katta, L.; Thrimurthulu, G., Novel Nanocrystalline Ce1−xLaxO2−δ (x = 0.2)
Solid Solutions: Structural Characteristics and Catalytic Performance. Chem. Mater. 2010, 22,
467–475.
38.
Moura, J. S.; Fonseca, J. d. S. L.; Bion, N.; Epron, F.; Silva, T. d. F.; Maciel, C. G.; Assaf,
J. M.; Rangel, M. d. C., Effect of Lanthanum on the Properties of Copper, Cerium and Zirconium
Catalysts for Preferential Oxidation of Carbon Monoxide. Catal. Today 2014, 228, 40–50.
39.
Miki, T.; Ogawa, T.; Haneda, M.; Kakuta, N.; Ueno, A.; Tateishi, S.; Matsuura, S.; Sato,
M., Enhanced Oxygen Storage Capacity of Cerium Oxides in Cerium Dioxide/Lanthanum
Sesquioxide/Alumina Containing Precious Metals. J. Phys. Chem. 1990, 94, 6464–6467.
40

40.
Cho, B. K., Chemical Modification of Catalyst Support for Enhancement of Transient
Catalytic Activity: Nitric Oxide Reduction by Carbon Monoxide over Rhodium. J. Catal. 1991, 131,
74-87.
41.
Lee, J. G.; Park, J. H.; Shul, Y. G., Tailoring Gadolinium-Doped Ceria-Based Solid Oxide
Fuel Cells to Achieve 2 W cm−2 at 550 °C. Nat. Commun. 2014, 5:4045.
42.
Martínez-Arias, A.; Hungría, A. B.; Fernández-García, M.; Iglesias-Juez, A.; Soria, J.;
Conesa, J. C.; Anderson, J. A.; Munuera, G., Operando DRIFTS Study of the Redox and Catalytic
Properties of CuO/Ce1−xTbxO2−δ (x = 0–0.5) Catalysts: Evidence of an Induction Step During CO
Oxidation. Phys. Chem. Chem. Phys. 2012, 14, 2144-2151.
43.
Reddy, B. M.; Thrimurthulu, G.; Katta, L., Design of Efficient CexM1−xO2−δ (M = Zr, Hf, Tb
and Pr) Nanosized Model Solid Solutions for CO Oxidation. Catal. Lett. 2011, 141, 572-581.
44.
Tang, Y.; Zhang, H.; Cui, L.; Ouyang, C.; Shi, S.; Tang, W.; Li, H.; Lee, J.-S.; Chen, L.,
First-Principles Investigation on Redox Properties of M-Doped CeO2 (M = Mn, Pr, Sn, Zr). Phys.
Rev. B 2010, 82, 125104
45.
Ahn, K.; Yoo, D. S.; Prasad, D. H.; Lee, H.-W.; Chung, Y.-C.; Lee, J.-H., Role of Multivalent
Pr in the Formation and Migration of Oxygen Vacancy in Pr-Doped Ceria: Experimental and FirstPrinciples Investigations. Chem. Mater. 2012, 24, 4261–4267.
46.
Tamura, M.; Wakasugi, H.; Shimizu, K.-i.; Satsuma, A., Efficient and Substrate-Specific
Hydration of Nitriles to Amides in Water by Using a CeO2 Catalyst. Chem. Eur. J. 2011, 17,
11428–11431.
47.
Zarembowitch, J.; Gouteron, J.; Lejus, A. M., Raman Spectra of Lanthanide Sesquioxide
Single Crystals with A-Type Structure. Phys. stat. sol. (b) 1979, 94, 249-256.
48.
Westermann, A.; Geantet, C.; Vernoux, P.; Loridant, S., Defects Band Enhanced by
Resonance Raman Effect in Praseodymium Doped CeO2. J. Raman Spectrosc. 2016, 47, 1276–
1279.
49.
Guo, M.; Lu, J.; Wu, Y.; Wang, Y.; Luo, M., UV and Visible Raman Studies of Oxygen
Vacancies in Rare-Earth-Doped Ceria. Langmuir 2011, 27, 3872-3877.
50.
Cutrufello, M. G.; Ferino, I.; Solinas, V.; Primavera, A.; Trovarelli, A.; Auroux, A.; Picciau,
C., Acid–Base Properties and Catalytic Activity of Nanophase Ceria–Zirconia Catalysts for 4Methylpentan-2-ol Dehydration. Phys. Chem. Chem. Phys. 1999, 1, 3369-3375.
51.
Wada, S.; Oka, K.; Watanabe, K.; Izumi, Y., Catalytic Conversion of Carbon Dioxide into
Dimethyl Carbonate Using Reduced Copper-Cerium Oxide Catalysts as Low as 353 K and 1.3
MPa and the Reaction Mechanism. Front. Chem. 2013, 1, 1-8.
52.
Hahn, K. R.; Iannuzzi, M.; Seitsonen, A. P.; Hutter, J., Coverage Effect of the CO2
Adsorption Mechanisms on CeO2(111) by First Principles Analysis. J. Phys. Chem. C 2013, 117,
1701–1711.
53.
Choudhary, V. R.; Rane, V. H., Acidity/Basicity of Rare-Earth Oxides and their Catalytic
Activity in Oxidative Coupling of Methane to C2-Hydrocarbons. J. Catal. 1991, 130, 411-422.
54.
Ikeda, Y.; Furusawa, Y.; Tomishige, K.; Fujimoto, K., Selective Conversion of Carbon
Dioxide and Methanol to Dimethyl Carbonate Using Phosphoric Acid-Modified Zirconia Catalysts.
ACS Symp. Ser. 2002, 809, 71–84.
55.
Jung, K. T.; Bell, A. T., An in Situ Infrared Study of Dimethyl Carbonate Synthesis from
Carbon Dioxide and Methanol over Zirconia. J. Catal. 2001, 204, 339–347.
56.
Migani, A.; Vayssilov, G. N.; Bromley, S. T.; Illas, F.; Neyman, K. M., Dramatic Reduction
of the Oxygen Vacancy Formation Energy in Ceria Particles: A Possible Key to their Remarkable
Reactivity at the Nanoscale. J. Mater. Chem. 2010, 20, 10535-10546.
57.
Sutradhar, N.; Sinhamahapatra, A.; Pahari, S.; Jayachandran, M.; Subramanian, B.; Bajaj,
H. C.; Panda, A. B., Facile Low-Temperature Synthesis of Ceria and Samarium-Doped Ceria
Nanoparticles and Catalytic Allylic Oxidation of Cyclohexene. J. Phys. Chem. C 2011, 115, 7628–
7637.

41

58.
Meher, S. K.; Rao, G. R., Polymer-Assisted Hydrothermal Synthesis of Highly Reducible
Shuttle-Shaped CeO2: Microstructural Effect on Promoting Pt/C for Methanol Electrooxidation.
ACS Catal. 2012, 2, 2795–2809.
59.
Yao, H. C.; Yao, Y. F. Y., Ceria in Automotive Exhaust Catalysts: I. Oxygen Storage. J.
Catal. 1984, 86, 254-265.
60.
Giordano, F.; Trovarelli, A.; Leitenburg, C. d.; Giona, M., A Model for the TemperatureProgrammed Reduction of Low and High Surface Area Ceria. J. Catal. 2000, 193, 273–282.
61.
Prymak, I.; Kalevaru, V. N.; Wohlrab, S.; Martin, A., Continuous Synthesis of Diethyl
Carbonate from Ethanol and CO2 over Ce–Zr–O Catalysts Catal. Sci. Technol. 2015, 5, 23222331.
62.
Pojanavaraphan, C.; Luengnaruemitchai, A.; Gulari, E., Effect of Catalyst Preparation on
Au/Ce1−xZrxO2 and Au–Cu/Ce1−xZrxO2 for Steam Reforming of Methanol. Int. J. Hydrogen Energ.
2013, 38, 1348–1362.
63.
Dobrosz-Gómez; Gómez-García, M. Á.; Rynkowski, J. M., CO Oxidation over Au/CeO2ZrO2 Catalists: The Effect of the Support Composition of the Au-Support Interaction. Kinet. Catal.
2010, 51, 823–827.
64.
Jin, Q.; Shen, Y.; Zhu, S., Praseodymium Oxide Modified CeO2/Al2O3 Catalyst for
Selective Catalytic Reduction of NO by NH3. Chin. J. Chem. 2016, 34, 1283—1290.
65.
Tamura, M.; Kishi, R.; Nakayama, A.; Nakagawa, Y.; Hasegawa, J.-y.; Tomishige, K.,
Formation of a New, Strongly Basic Nitrogen Anion by Metal Oxide Modification. J. Am. Chem.
Soc. 2017, 139, 11857−11867.
66.
Badri, A.; Binet, C.; Lavalley, J.-C., Use of Methanol as an IR Molecular Probe to Study
the Surface of Polycrystalline Ceria. J. Chem. Soc., Faraday Trans. 1997, 93, 1159-1168.
67.
Tamura, M.; Satsuma, A.; Shimizu, K.-i., CeO2-Catalyzed Nitrile Hydration to Amide:
Reaction Mechanism and Active Sites. Catal. Sci. Technol. 2013, 3, 1386-1393.
68.
Tamura, M.; Tomishige, K., Redox Properties of CeO2 at Low Temperature: The Direct
Synthesis of Imines from Alcohol and Amine. Angew. Chem. Int. Ed. 2015, 54, 864 –867.
69.
D'Angelo, A. M.; Wu, Z.; Overbury, S. H.; Chaffee, A. L., Cu-Enhanced Surface Defects
and Lattice Mobility of Pr-CeO2 Mixed Oxides. J. Phys. Chem. C 2016, 120, 27996–28008.

42

Supporting Information

Improving the stability of CeO2 catalyst by rare earth metal promotion and
molecular insights in the dimethyl carbonate synthesis from CO2 and methanol
with 2-cyanopyridine

Dragos Stoian†, ‡, §, Francisco Medina‡, Atsushi Urakawa*†
†

Institute of Chemical Research of Catalonia (ICIQ), The Barcelona Institute of Science and Technology, Av. Països
Catalans 16, 43007 Tarragona, Spain
‡

Department of Chemical Engineering, University Rovira i Virgili, Av. Països Catalans 26, 43007 Tarragona, Spain
*Email: aurakawa@iciq.es

Present address
§

The Swiss-Norwegian Beamlines (SNBL) at ESRF, CS40220, 38043 Grenoble CEDEX 9, France

43

Scheme S1. By-products formation (methyl picolinate, Me-PCN and methyl carbamate, Me-CBM) in the continuous
DMC synthesis in the presence of a dehydrating agent (2-cyanopyridine, 2-CP). 2-CP hydrolysis leads to the formation
of the corresponding amide, 2-picolinamide, 2-PA.

Experimental: materials

High surface area powders of CeO2 and CeO2-ZrO2 solid solutions (78-22, 50-50, and
25-75 wt%) were kindly supplied by Daiichi Kigenso Kagaku Kogyo Co. Ltd., Japan and
used without any further treatment. High surface area ZrO2 (1/8” pellets) was purchased
from Alfa Aesar and used as received. Precious metals (1 wt% Ru, Rh, Pd) were
synthesized by the incipient wetness impregnation of the pristine CeO2 with an aqueous
solution of the metal precursor (chlorides). The metal chlorides (99.9% purity) were
purchased from Alfa Aesar. The impregnated materials were dried overnight at 80-90 °C
in an oven and calcined at 400 °C for 4 h in a static air after a temperature ramp of 2 °C
min-1.

44

Direct methanol carboxylation: effect of Zr and precious metal (PM)

Figure S1 (a) shows the catalytic performance in terms of methanol conversion (XMeOH)
and DMC selectivity (SDMC) during the initial phase (0-2 h) and at 24-26 h of the DMC
synthesis using CeO2, ZrO2 and CeO2-ZrO2 solid solutions. The reaction time, more
precisely time on stream, is shown as 2 h interval due to the residence time of the liquid
solution in the reaction system. The liquid sample was analyzed every 2 h when enough
liquid product (ca. 1 mL) was collected (when one number is shown for the reaction time,
the initial value of the interval is taken). Although the addition of Zr to CeO2 has been
reported to positively affect the catalytic activity1-5 for the reaction, in this work negative
effects of Zr incorporation were obvious. Regarding the initial activity (0-2 h), the decrease
in XMeOH became consistently more prominent at higher Zr content, decreasing from
92.4%, 88.4%, 72.9% to 37.3% for 0, 22, 50, and 75 wt% ZrO2 amount in the solid
solution, respectively. Moreover, the use of high surface area ZrO2 gave only a trace
amount of DMC. After 24 h, catalyst deactivation was observed for all catalysts and it was
more pronounced for Ce-Zr [50/50] (ca. 21% activity drop compared to 6-8% drops of the
other samples). On the other hand, the drop in catalytic activity did not affect SDMC
significantly, showing >95% with the formation of small amounts of Me-PCN (from 1.0 to
4.3%) and Me-CBM (0.4 to 1.6%). Interestingly, at higher ZrO2 content in the solid
solutions, the decrease in SDMC was smaller over 24 h (ca. 5% SDMC decrease for Ce-Zr
[100/0], while only ca. 1% decrease for Ce-Zr [25/75]), indicating positive influences of Zr
promotion on the catalyst stability in terms of DMC selectivity.
Figure S1 (b) presents the catalytic performance using CeO2-supported Ru, Rh and Pd
catalysts. Clearly, these results indicate strongly negative influences of the PMs on the
45

performance. For 1 wt% Pd-CeO2 and 1 wt% Rh-CeO2, ca. 63% XMeOH was initially
observed, while the addition of 1 wt% Ru resulted in much lower activity of 34% XMeOH.
Regarding the catalyst stability, it is very striking to note that both 1 wt% Rh-CeO2 and 1
wt% Ru-CeO2 catalysts showed a tremendous loss in activity after 24 h (XMeOH of ca. 7
and 12%, respectively). In contrast, a slight increase in XMeOH (from 63 to 70%) was
detected for 1 wt% Pd-CeO2. The incorporation of the PMs consistently impacted
negatively on DMC selectivity, showing ca. 75-80% SDMC for the three catalysts.
Interestingly, unlike CeO2 and CeO2-ZrO2 solid solutions, all ceria-supported PM catalysts
displayed an increase in SDMC by ca. 10% after 24 h of the reaction. Still, SDMC values
were considerably lower compared to that observed for CeO2 and the by-products were
1.6-20.4% Me-PCN and 1.1-5.4% Me-CBM. The high selectivity to Me-PCN indicates that
2-PA is formed via the hydration reaction of 2-CP and that the reaction between 2-PA and
methanol (Scheme S1) is promoted by the PMs or by the sites created by the specific
interactions between CeO2 with the PMs. The reaction is similar to urea methanolysis
where -NH2 group is replaced by -OCH3 group and both acidic and basic catalysts are
known to catalyze the reaction.6-7 CeO2-based materials are known to be active in the
reaction8-9, and it is implied that promotion of CeO2 by the PMs can be highly effective for
methanolysis reactions.
Generally, both Zr and PMs incorporation showed negative influences on the catalytic
performance, especially on the reactivity (XMeOH). The lower reactivity of the catalysts at
higher Zr contents implies that the reactivity originates from CeO2 and its surface. In case
of PMs addition, the impact on the reactivity was drastic despite the relatively low amount
of the PMs (1 wt%) with respect to CeO2. Considering the much lower initial reactivity and

46

the very low catalytic stability of Rh-CeO2 and Ru-CeO2 compared to CeO2, it can be
concluded that the PMs alter the chemical surface properties (e.g. acid-base, redox)
completely, as confirmed later. Nonetheless, the effects of incorporating both Zr and PMs
in CeO2 were found negative for the activity and stability in the reaction.

Figure S1. Methanol conversion and DMC selectivity during the continuous DMC synthesis at 120 °C, 30 bar at time
on stream of 0-2 and 24-26 h using a. Ce-Zr mixed oxides and b. precious metal-CeO2 as catalyst. Black and grey
columns represent methanol conversion and DMC selectivity, respectively.

47

Characterization: structural and textural properties (XRD, Raman, N2 physisorption)

The XRD patterns of the pristine CeO2, CeO2-ZrO2 solid solutions, ZrO2 and assynthesized materials incorporating PM and REM are presented in Figure S2. CeO2
showed the peaks corresponding to (111), (200), (220), and (311) planes of cubic fluorite
phase, while ZrO2 exhibited the peaks for a monoclinic phase with a small amount of a
tetragonal phase. CeO2-ZrO2 solid solutions preserved a single cubic fluorite phase
characteristic of CeO2 without a detectable monoclinic/tetragonal phase characteristic of
ZrO2. It should be noted that the characteristic XRD peaks of CeO2-ZrO2 solid solutions
slightly shifted to higher angles compared to the peaks for cubic fluorite phase of CeO2.
This can be explained by the insertion of zirconium atoms in the CeO2 cubic matrix with
the shrinkage of its unit cell (Ce4+ ionic radius is 0.97 Å, whereas Zr4+ ionic radius is 0.82
Å) and by the deformation of the cubic phase to be transformed into the tetragonal phase
for higher zirconium loading.10 The materials with PM and REM showed negligible
structural changes, indicating that metal promoters were finely dispersed on the surface
of the support.
Figure S3 shows the Raman spectra of 1wt% (left) and 5 wt% (right) REM-doped CeO2
materials in comparison to the spectrum of pristine CeO2. The major changes were
observed for Pr-doped CeO2 materials. The origin of the spectral changes and influence
on catalytic activity are discussed in the main text.
The BET surface area and pore volume of all materials are presented in Table S1. Higher
amount of incorporated Zr or REM to CeO2 consistently decreased the surface area,
although there was no clear trend in the pore volume with the type and loading of REM.
In case of the PM-loaded materials, the surface area was also reduced and it was most
48

pronounced for Pd-CeO2. The pore volume of the three PM-CeO2 samples were similarly
reduced by ca. 10%. Importantly, the changes in the textural properties and bulk structure
of the materials are not correlated with the catalytic activity and stability.

Figure S2. X-ray diffraction (XRD) patterns of the pristine CeO2, CeO2-ZrO2 solid solutions, ZrO2 and as-synthesized
materials incorporating precious metal and REM promoters.

49

Figure S3. Raman spectra (excitation laser: 532 nm) of pristine CeO2 (red) and REM (La (orange), Gd (blue) and Pr
(green))-CeO2. (left) 1 wt% REM-CeO2 materials and (right) 5 wt% REM-CeO2.

Characterization: acidity and basicity

Generally, the addition of a foreign atom on/in the CeO2 structure reduced the amount of
adsorbed CO2/NH3, thus the total number of acidic and basic sites. The exceptions were
CeO2-ZrO2 solid solutions and ZrO2 for which the weak acidity was greatly enhanced
when the ZrO2 content was >50 wt% due to more pronounced influence of highly acidic
ZrO211-14 in the solid solutions (Figure S4 A). Zr-addition to CeO2 resulted in weakening
of the strong acidity/basicity as indicated by the small amount of adsorbed NH3/CO2,
respectively, at temperature above ca. 500 °C.
On the other hand, PM addition to CeO2 induced remarkable changes in the
acidity/basicity of the material (Figure S4 A). Upon Rh and Ru addition to CeO2, weak
and moderate acidic sites were decreased, but mildly strong acidic sites (desorption at
ca. 520 °C) became less and, in contrast, highly strong acidity (desorption at ca. 700 °C)
was enhanced. For Rh- and Ru-CeO2, the number of strong basic sites were reduced

50

significantly, although the formation of stronger basic sites characterized by the
desorption peaks at 680 and 760 °C was observed for Ru-CeO2. Pd addition resulted in
very different changes in both acidity and basicity from those of Rh and Ru addition.
Notably, the weak-moderate basicity of CeO2, characterized by the broad CO2 desorption
peaks in the range of 100-450 °C, was largely suppressed and, instead, strong basic sites
(desorption at ca. 640 °C) were formed. The similarity of acid-base properties of Rh- and
Ru-CeO2 in comparison to Pd-CeO2 was consistently reflected in the catalytic activity and
stability trend (Figure S1). This implies that acid-base properties are playing roles in the
DMC synthesis from CO2 and methanol in the presence of 2-CP.

A

Figure S4 (A). Temperature programmed desorption (TPD) profiles of the CeO2-ZrO2 solid solutions and precious
metals (PM) promoted CeO2 materials. Upper panels correspond to those of CO2-TPD, whereas the lower ones to
those of NH3-TPD.

51

B

Figure S4 (B). Temperature programmed desorption (TPD) profiles of the rare earth metals (REM) promoted CeO2
materials. Upper panels correspond to those of CO2-TPD, whereas the lower ones to those of NH3-TPD.

Figure S5. H2-TPR profiles of the high surface area CeO2 used throughout this study. Both surface (low temperature)
and bulk (high temperature) reduction peaks can be clearly identified.

52

Figure S6. H2-TPR profiles of the REM-CeO2 materials used throughout this study. Surface reduction peak is
represented.

As well known, Zr addition to CeO2 induced drastic changes in the reducibility of CeO21517

(Figure S7). The TPR profiles of the CeO2-ZrO2 solid solutions displayed only one main

broad reduction peak in the region between 500–580 °C at different peak positions (507,
543, and 576 °C for CeO2-ZrO2 [78-22], [50-50], and [25-75], respectively). In contrast,
the high surface area ZrO2 used in the present study showed no reduction peaks under
the condition investigated (Figure S7). Although the overall reducibility of CeO2-ZrO2 is
enhanced by facilitating the bulk reduction compared to CeO2, it should be noted that the
reduction of surface Ce atoms is more facile for CeO2 than CeO2-ZrO2 at lower
temperatures (<400 °C, Figure S5). The catalytic tests clearly showed that CeO2 and not
ZrO2 is responsible for activation of CO2 and methanol (Figure S1), and this indicates
that the reducibility of surface Ce may play important role in the reaction mechanisms
besides the balance between basic/acidic sites. The importance of redox sites in the
catalytic cycle have been investigated by operando spectroscopic means and will be
communicated separately.
PM addition to CeO2 resulted in totally different TPR profiles (Figure S8). The three
catalysts are characterized by the appearance of a new large peak in the TPR profiles at
53

very low temperatures around 100-180 °C, characteristic of Ru, Rh, and Pd oxides
reduction. The peak corresponding to the CeO2 surface reduction was reduced drastically
for the three PM-CeO2 materials. As in the cases of the catalytic activity (Figure S1) and
TPD results (Figure S4), the reducibility of Ru- and Rh-CeO2 presented some similarities,
showing the reduction peak at ca. 100 °C, while a higher reduction peak ca. 180 °C was
observed for Pd-CeO2. This indicates possible correlations among catalyst reducibility,
acidity-basicity, and catalytic performance. Based on the H2-TPR profiles of PM-CeO2,
the H2 consumption is high at low temperatures below 200 °C considering that it is the
reduction of 1 wt% PM. This observation together with lowered reduction peak at ca. 500
°C imply the promoted reduction of surface Ce by the catalytic function of PM, facilitating
the reactivity of surface oxygen of CeO2. Nevertheless, this high reducibility of surface
CeO2 did not positively affect the catalytic performance of PM-CeO2 (Figure S1) and
further investigation is required to elucidate the PM effects on the reactivity and product
selectivity.

Figure S7. H2-TPR profiles of the CeO2-ZrO2 solid solutions (black for pure CeO2 and orange for pure ZrO2).

54

Figure S8. H2-TPR profiles of the precious metals (PM) promoted CeO2 materials (surface reduction region).

Figure S9. XRD patterns of the used materials (CeO2 and 1 wt% REM promoted CeO2) after the 30 h catalytic
comparison test.

55

Figure S10. Raman (left) and ATR-IR (right) studies of the catalysts after the reaction and methanol washing. The
spectral contribution due to CeO2 has been subtracted and the Raman spectra have been normalized with respect
to CeO2 F2g mode at ca. 460 cm-1.

Figure S11. TGA of CeO2 and 1 wt% REM promoted CeO2 before and after the reaction.

56

Table S1. The BET surface area (m2 g-1) and pore volume (cm3 g-1) of the pristine CeO2, CeO2-ZrO2 solid solutions,
ZrO2 and as-synthesized materials incorporating precious metal and rare earth metal promoters.

Material type

CeO2-ZrO2 solid
solutions

Precious metals
(PM) – CeO2

Rare earth
metals (REM) –
CeO2

BET surface area (m2 g-1)

Pore volume (cm3 g-1)

160
75
59
58
97
141
147
123
149
135
122
156
146
123
152
139
126

0.197
0.247
0.199
0.230
0.320
0.173
0.176
0.177
0.175
0.186
0.172
0.198
0.199
0.177
0.180
0.184
0.173

CeO2
CeO2-ZrO2 [78-22]
CeO2-ZrO2 [50-50]
CeO2-ZrO2 [25-75]
ZrO2
1 wt% Rh-CeO2
1 wt% Ru-CeO2
1 wt% Pd-CeO2
0.25 wt% La-CeO2
1 wt% La-CeO2
5 wt% La-CeO2
0.25 wt% Gd-CeO2
1 wt% Gd-CeO2
5 wt% Gd-CeO2
0.25 wt% Pr-CeO2
1 wt% Pr-CeO2
5 wt% Pr-CeO2

Table S2. The amounts of adsorbed-desorbed CO2 and NH3 expressed in μmol gcat.-1 (a measure of the basicity and
acidity of the materials) for the catalysts tested in the long term studies of DMC synthesis.

Material type

CeO2-ZrO2 solid
solutions

Precious metal
(PM) – CeO2

Rare earth metal
(REM) – CeO2

μmol CO2 gcat.-1

μmol NH3 gcat.-1

520.8
140.1
137.7
149.0
313.6
259.6
176.8
118.1
411.7
380.3
399.0
410.8
377.1
386.2
414.5
374.4
400.6

396.2
511.2
710.0
316.3
1410.8
255.8
299.5
360.1
259.3
317.9
308.7
364.3
243.0
286.1
213.8
317.6
293.4

CeO2
CeO2-ZrO2 [78-22]
CeO2-ZrO2 [50-50]
CeO2-ZrO2 [25-75]
ZrO2
1 wt% Rh-CeO2
1 wt% Ru-CeO2
1 wt% Pd-CeO2
0.25 wt% La-CeO2
1 wt% La-CeO2
5 wt% La-CeO2
0.25 wt% Gd-CeO2
1 wt% Gd-CeO2
5 wt% Gd-CeO2
0.25 wt% Pr-CeO2
1 wt% Pr-CeO2
5 wt% Pr-CeO2
57

Table S3. H2 uptake (μmol gcat.-1) of the CeO2-ZrO2 solid solutions, precious metals (PM) promoted CeO2, and rare
earth metals (REM) promoted CeO2 catalysts used in long-term DMC synthesis. For the PM and REM promoted
materials only the contribution for the surface reduction has been included.

H2 consumption
in μmol gcat.-1

Material type

CeO2-ZrO2 solid
solutions

Precious metals
(PM) – CeO2

Rare earth
metals (REM) –
CeO2

CeO2
CeO2-ZrO2 [78-22]
CeO2-ZrO2 [50-50]
CeO2-ZrO2 [25-75]
ZrO2
1 wt% Rh-CeO2
1 wt% Ru-CeO2
1 wt% Pd-CeO2
0.25 wt% La-CeO2
1 wt% La-CeO2
5 wt% La-CeO2
0.25 wt% Gd-CeO2
1 wt% Gd-CeO2
5 wt% Gd-CeO2
0.25 wt% Pr-CeO2
1 wt% Pr-CeO2
5 wt% Pr-CeO2

58

376 (surface + bulk = 733)
401
532
408
no reduction
355
274
533
372
434
472
406
479
369
404
420
371

References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.

Tomishige, K.; Furusawa, Y.; Ikeda, Y.; Asadullah, M.; Fujimoto, K., Catal. Lett. 2001, 76,
71-74.
Tomishige, K.; Kunimori, K., Appl. Catal. A-Gen. 2002, 237, 103–109.
Hofmann, H. J.; Brandner, A.; Claus, P., Chem. Eng. Technol. 2012, 35, 2140–2146.
Lee, H. J.; Park, S.; Song, I. K.; Jung, J. C., Catal. Lett. 2011, 141, 531-537.
Lee, H. J.; Joe, W.; Song, I. K., Korean J. Chem. Eng. 2012, 29, 317-322.
Wang, M.; Wang, H.; Zhao, N.; Wei, W.; Sun, Y., Catal. Commun. 2006, 7, 6–10.
Wang, M.; Zhao, N.; Wei, W.; Sun, Y., Ind. Eng. Chem. Res. 2005, 44, 7596–7599.
Joe, W.; Lee, H. J.; Hong, U. G.; Ahn, Y. S.; Song, C. J.; Kwon, B. J.; Song, I. K., J. Ind.
Eng. Chem. 2012, 18, 1018–1022.
Joe, W.; Lee, H. J.; Hong, U. G.; Ahn, Y. S.; Song, C. J.; Kwon, B. J.; Song, I. K., J. Ind.
Eng. Chem. 2012, 18, 1730–1735.
Sergent, N.; Lamonier, J.-F.; Aboukaïs, A., Chem. Mater. 2000, 12, 3830–3835.
Tanabe, K., Mater. Chem. Phys. 1985, 13, 347-364.
Yamaguchi, T., Sekiyu Gakkaishi 1993, 36, 250-267.
Tanabe, K.; Yamaguchi, T., Catal. Today 1994, 20, 185-197.
Khaodee, W.; Jongsomjit, B.; Assabumrungrat, S.; Praserthdama, P.; Goto, S., Catal.
Commun. 2007, 8, 548–556.
Prymak, I.; Kalevaru, V. N.; Wohlrab, S.; Martin, A., Catal. Sci. Technol. 2015, 5, 23222331.
Fornasiero, P.; Dimonte, R.; Rao, G. R.; Kaspar, J.; Meriani, S.; Trovarelli, A.; Graziani,
M., J. Catal. 1995, 151, 168-177.
Murota, T.; Hasegawa, T.; Aozasa, S.; Matsui, H.; Motoyama, M., J Alloy. Compd. 1993,
193, 298-299.

59