Enabling continuous capture and catalytic conversion of flue gas CO2
to syngas in one process
Luis F. Bobadilla1†, José M. Riesco-García2, Germán Penelás-Pérez2 and Atsushi
Urakawa1*
1

Institute of Chemical Research of Catalonia (ICIQ), The Barcelona Institute of Science
and Technology, Av. Països Catalans 16, 43007 Tarragona, Spain
2

Repsol, Centro de Tecnología, Carretera Nacional N-V, 28935 Móstoles, Spain
* Corresponding author: aurakawa@iciq.es (Atsushi Urakawa)

†

Current address: Instituto de Tecnología Química, Universidad Politécnica de Valencia
– CSIC, Avenida de los Naranjos s/n, 46022 Valencia, Spain

ABSTRACT
Albeit a variety of available strategies for CO2 conversion to useful chemicals and fuels,
most technologies require relatively pure CO2, especially without oxygen and water.
This requires additional steps of CO2 capture and purification before its efficient
conversion. This necessity increases energy requirement, leading to poorer carbon
footprints and higher capital expenditures lowering the viability of overall CO2
conversion processes. We have developed an effective technology which combines CO2
capture and conversion processes using isothermal unsteady-state operation and a
catalyst consisting of earth-abundant chemical elements (FeCrCu/K/MgO-Al2O3).
Diluted CO2 streams common in process flue gases, even containing oxygen and water,
can be fed to the process and relatively pure product stream such as syngas, i.e. carbon
oxides (CO and CO2) and hydrogen mixture, can be produced. A possible scheme of

reactor integration for continuous CO2 abatement and conversion based on a tworeactors system is presented.
1. Introduction
Success in the quest for sustainable solutions to environmental threats induced by the
atmospheric carbon dioxide (CO2) accumulation largely relies on advances in chemical
sciences and technological innovations [1]. One of the most promising strategies for the
mitigation of anthropogenic CO2 emission is its utilization by converting CO2 into
chemical fuels and value-added chemicals such as methanol and polymers [2]. Prior to
the chemical transformation, however, CO2 must be first captured from emission
sources such as power, refinery, chemical, steel, and ceramic plants. This stage
represents an increase from 40 to 80% of the total capital cost for a conventional postcombustion carbon capture and storage (CCS) process [3]. These steps drastically
increase capital costs and investments for CO2 transformation. Only the capture step is
estimated to increase the energy requirements of a power plant by 25–40% [4].
Most chemical processes are designed to be operated under steady-state conditions.
In case of heterogeneous catalytic processes, process performance is conventionally
optimized under a given temperature, pressure, concentration, and flow conditions.
Having this approach as the main stream, over the past 50 years intensive research and
development work in academia and industry was devoted to improve the overall
reaction performance using so-called unsteady-state operation. For a number of
catalytic reactions, the advantages of unsteady-state operation and improvement in
catalytic process performances were shown theoretically and also demonstrated in
practice by forcing external parameter(s) such as concentration, temperature, pressure,
and flow direction to change. Using such operations, the profiles of the catalyst states,

coverage of active surface chemical species, concentrations, and temperatures in
reactors can be influenced and controlled to a great extent, thus providing more
favourable conditions for better process performance [5-7]. One of the catalytic
technologies based on unsteady-state operation close to our daily life is NOx storagereduction (NSR) catalysis pioneered by TOYOTA and widely used in automotive
industry [8, 9]. The major advantage of NSR technology is to chemically reduce
nitrogen oxides in fuel-lean and oxygen-rich conditions. Without transiently switching
between NOx storage phase and reduction phase (as a short pulse of increased fuel for
the latter), efficient NOx reduction is not possible.

Among possible strategies proposed in the literature, chemical looping process
represents an attractive technology that offers means to effectively purify CO2 for
subsequent conversion by changing the environment of catalyst within the reactor
system [4]. Within the closed-system, the catalyst experiences unsteady-state reaction
conditions. The fundamental concept of the process is based on the conversion of metal
oxide sorbent to metal carbonate by in situ removal of flue gas CO2 where calciumbased materials are typically used as the sorbent. The major general challenges of such
processes are the required complex engineering, material resistance and drastic
temperature changes for catalyst regeneration [4]. Generally this technology requires an
additional reactor for CO2 transformation unless the two conversion steps of CO2 to CO
and H2 oxidation to H2O are combined in one process (i.e. net reverse water-gas shift
(RWGS) reaction operated under unsteady-state condition) as demonstrated by Kuhn
using perovskite-type oxides [10, 11]. The approach is promising although further
material and process improvements are required to convert CO2 to CO efficiently. Very
recently, Farrauto has reported dual-functional materials based on ruthenium and
calcium oxide which capture CO2 from a model flue gas stream and subsequently

reduce the captured CO2 to methane at the same temperature in the same reactor [12].
Although the production of methane under the unsteady-state operation was
demonstrated and even the catalyst functions in the presence of water vapour, the low
CO2 capture efficiency and high loading of previous ruthenium are the important points
for further improvement.
Herein, we report a highly efficient dual-function catalyst material and process
concept of catalytic CO2 conversion strategy through CO2 capture and subsequent
reduction combined in one process, utilizing unsteady-state operation denoted here as
CO2 capture-reduction (CCR). In CCR, two distinct regimes, (i) capture of CO2 on a
storage component of a catalyst and (ii) release of CO2 or direct catalytic reduction of
stored CO2 by reducing gas (here hydrogen) and catalyst regeneration, are operated
alternately and isothermally. The important differences from the current state-of-the-art
are the very high capture efficiency enabled by potassium component of the catalyst,
earth-abundant metal components (Fe, Cr, Cu) and high activity and selectivity for CO2
reduction to CO, and the same duration for the CO2 capture and reduction phases. The
last point is crucial for an advanced process integration concept, enabling continuous
and complete CO2 capture and reduction of captured CO2 in one process using multiplereactors.
2. Experimental
2.1.

Materials and chemicals

PURAL®MG 20 hydrotalcite (Mg/Al molar ratio is 0.29) provided by Sasol
Germany GmbH, Inorganic Specialty Chemicals was employed as the pristine material
of the support. Potassium carbonate from Panreac was used as precursor of potassium
component. All the metal precursors salts used in this study were the nitrate form
purchased from Sigma-Aldrich. Deionized water was used for catalyst synthesis. The

gases were purchased from Linde at the quality of >99.9993% for CO2 and >99.999%
for the other gases.
2.2.

Catalyst synthesis

The hydrotalcite supported FeCrCu-K catalyst was prepared by sequential
impregnation using the incipient wetness method. Firstly, a thermal treatment of
hydrotalcite support at 600 °C for 3 h was carried out in order to obtain the
homogeneous mixed oxides of MgO and Al2O3. Afterwards, three metals (Fe, Cr and
Cu) were impregnated using a solution prepared by adding the necessary volume of
deionized water to the metal nitrates precursor salts. The resulting solid was dried
overnight at 80 °C followed by calcination at 500 °C for 5 h. Finally, a solution of
potassium carbonate was impregnated over the solid obtained and it was dried and
calcined under identical conditions. The elemental composition of the catalyst was
determined by ICP analysis (Table S1 – Supplementary Material).
2.3.

Reaction procedure

The experimental system mainly consists of four sections; (i) gas feed system, (ii)
switching valve, (iii) reactor, (iv) gas detection system (IR/MS). In (i) there are two gas
lines which allows preparing a gas mixture, one containing CO2, synthetic air, and
nitrogen (capture phase gas), and another one containing H2 (reduction phase gas) at
desired concentrations and flow rates. In order to investigate the effect of water, the
nitrogen flow can be passed through the water saturator maintained at 30 °C. These two
lines enter the (ii) switching valve and only one of the two gas lines enters the (iii)
catalytic reactor and the other line goes to a gas exhaust. There is a by-pass line at the
reactor to measure ‘blank’ response of the gas flow under the same conditions for the
flow controllers and detectors so that the results can be precisely calibrated and

quantified. The effluent gas stream was passed into the transmission gas cell of an IR
spectrometer (Bruker ALPHA) for quantitative product analysis. All valve switching
and spectrum acquisition were synchronized by means of dedicated LabVIEW program
in order to facilitate the data analysis and also precisely average over a number of CCR
cycles to increase the S/N of the data.
In each run 1 g of the catalyst was used. Prior to the testing it was pressed, crushed,
and sieved to the range of 200-300 µm. The catalyst was heated up to 450 °C under 45
mL min-1 N2 flow and then activated at the temperature under unsteady-state conditions
(30 cycles) with alternating flows of 45 mL min-1 of N2 saturated with water and pure
H2 for 90 s for each phase. This activation procedure was chosen to aim at avoiding
over-reduction of the catalyst and partially reducing the hematite (Fe2O3) to magnetite
(Fe3O4) as well as any CrO3 present in the catalyst to Cr2O3 [13]. Moreover, we
performed a blank test in each experiment using a fixed bed of an inert material (SiC)
with the same particle size (200-300 µm) and passing the same feed gas compositions
used in the catalytic tests.
3. Results and discussion
Figure 1 illustrates representative concentration profiles of CO2 and the product (CO
in this case) in the CCR process based on switching between capture phase containing
5.8% CO2 in N2 (the first half period) at 27 mL min-1 and pure H2 gas (the second half
period) at 65 mL min-1 passing through a reactor and the duration of each phase was
107.5 s. The reactor contained a CCR catalyst (1 g) consisting of Fe, Cr, Cu (reducing
components) and K (CO2 capture component) supported on calcined Mg and Al-based
hydrotalcites and it was maintained at 550 °C. The concentration profiles under the
unsteady-state operation after a stabilization period are compared with that of CO2 in a
blank test using a reactor packed with inert SiC pellets. It is evident that in the presence

of the CCR catalyst CO2 is efficiently trapped over the surface of alkali adsorption sites
expectedly in the form of surface carbonates during the CO2 capture phase.
Subsequently under the reducing condition by switching to H2 atmosphere, most of the
captured CO2 reacts with H2 over the catalytically active sites, releasing the reduction
product (CO) accompanied by some release of unconverted CO2.

Figure 1. A pictorial representation of the function and a possible mechanism of CCR
on the catalyst material and representative concentration profiles during the unsteadystate operation. In comparison, a CO2 concentration profile measured under the identical
condition using an inert SiC material filled in the reactor is also shown.
The advantages of CCR process with the use of the specific catalyst are obvious.
During the capture phase, no or very low concentration CO2 is detected in the effluent
stream of the reactor. On the other hand, the effluent stream in the reduction phase
contains mainly carbon oxides, mostly CO with some amount of CO2, and unreacted H2.
The simplicity of the CCR process stems from the condition for CO2 capture and
reduction; both are performed at the same temperature. This implies that identification
of an excellent catalyst is of major importance for isothermal CCR. In addition, in the

presented case, both periods have the same duration and the process was operated at
atmospheric pressure. These features are important in practice; the former for process
integration (vide infra) and the latter to cope with most effluent stream released from
various processes in industry.
One possible scheme to implement CCR process at a diluted-CO2 emission source is
illustrated in Figure 2. It consists of two CCR reactors where one is capturing CO 2 and
the other is reducing captured CO2 and regenerating the catalyst for subsequent CO2
capture (Mode A). Upon saturation of catalyst with CO2 and/or at a defined time period,
the functions of the two reactors are reversed; the first one for CO2 reduction and the
second one for CO2 capture (Mode B). By synchronizing the switching of gas flow
direction entering and exiting the two reactors with a possible time delay, it is possible
to continuously clean CO2 from the effluent stream and produce valuable products (here
syngas) also continuously.

Figure 2. Integrated two-reactor CCR process for continuous CO2 capture and
reduction.

A successful CCR catalyst must satisfy the following requirements:
•

High capacity to adsorb/absorb CO2 at operation temperature

•

Fast and complete reduction/release of stored CO2 and thus fast regeneration for

subsequent CO2 capture at the same temperature
•

High conversion of CO2 stored on the catalyst, e.g. in the form of carbonates

•

Selective hydrogenation to desired product(s)

•

High levels of reversibility and durability of CO2 capture and release/reduction

processes
•

Composed of abundant and economic chemical elements and materials

Very important properties of a successful catalyst are fast reducibility of captured
CO2 and catalyst regenerability within the time-scale of CO2 capture, bearing the
possible integrated implementation configuration (Figure 2) in mind. Achieving
required high reactivity, fast regenerability, and high CO2 absorption capacity at the
same temperature are challenging. The degree of catalyst regeneration must be high and
reproducible so that the function of catalyst as CO2 absorber is retained over a long-term
operation. Besides, product selectivity is of central importance to avoid further product
separation steps. It is also worth mentioning that ideally hydrogen used in CCR process
should be produced by a process powered by low carbon footprint energy sources, e.g.
water electrolysis and photolysis powered by renewable energy or nuclear power.
Regarding the catalyst composition, promising catalyst materials for CCR process
would consist of transition metal, alkali/alkali-earth metal carbonates supported over
metal oxides according to our previous experience on comparable NOx storagereduction catalytic processes [14-17]. The alkali/alkali-earth carbonates such as K2CO3

and BaCO3 can be reduced to respective oxides and hydroxides and they function as
CO2 capture and storage material in a wide temperature range. Moreover,
nanocrystalline carbonates such as nanosized BaCO3 and K2CO3 can be an excellent
reversible CO2 storage components because their decomposition temperatures drop
from 1350 and 890 °C in the form of large bulk crystallites, respectively, to ca. 200-300
°C, likely due to enhanced surface carbonates composition (Supplementary Materials).
The temperatures are expected to be even lower under reducing atmosphere of H2. The
type of metal oxides used as support will also alter the CO2 capture property drastically
owing to their unique acidity, basicity, porosity, hydrophilicity, and reactivity, thus
influencing the interaction strength and modes with CO2. Support materials themselves
can also participate in CO2 capture, forming surface carbonates and bicarbonates. On
the other hand, depending on the active metal, there are several possible products
synthesized from the reactions of the surface species like carbonates and formates
formed by CO2 capture or released CO2 with a chemical reductant like hydrogen as used
in this work. For example, Ni-based catalysts are selective in the production of methane
while that Cu and FeCr-based catalysts are effective for CO production via RWGS
reaction.
After screening and optimizing various active catalyst components for CO2 capture
and hydrogenation of stored CO2 (Supplementary Materials), we developed an active
CCR catalyst consisting of Fe, Cr, Cu, and K supported over mixed Mg and Al oxides
obtained by calcining a Mg- and Al-containing hydrotalcite material (SASOL, PURAL
MG20, Mg/Al = 0.32) [18]. The non-precious, transition metal components and Kpromoted Mg-Al oxides support are crucial for their RWGS activity and to achieve high
capacity for CO2 capture, respectively. Potassium promotion has been reported to
significantly improve CO2 reversible sorption capacity at the temperature we aimed at

(300–500 °C) [19, 20]. The details of the catalyst composition and synthesis are
described in Supplementary Materials and key functions of Cu and K as efficient CO2
reduction and capture, respectively, have been investigated by space- and time-resolved
spectroscopy and reported elsewhere [21].

Figure 3. CO2 capture efficiency, CO2 conversion, and CO selectivity as a function of
(A) CO2 concentration (5.8-9.5%) at 450 ºC and (B) reaction temperature with 5.8%
CO2. Gas composition - Capture phase: CO2 diluted in nitrogen (ideal condition), CO2
diluted in nitrogen saturated with 4% of water vapour (effect of water), CO2 diluted in
nitrogen with 4 % of oxygen (effect of oxygen) and CO2 diluted in nitrogen with 5% of
oxygen and 4% of water (realistic condition). Gas hourly space velocity (GHSV) - 1620
mL gcat-1 h-1. Reduction phase: Pure hydrogen with GHSV of 3900 mL gcat-1 h-1. The
CCR period length was 215 s, i.e. 107.5 s for CO2 capture and reduction phase.

In a typical post-combustion capture process, which includes conventional process
heaters and industrial utility boilers and represents most of today’s fossil-fuel-based
electricity generation, CO2 is captured from flue gas that contains 4-8 vol% CO2 for a
natural-gas-fired power plant and 12-15 vol% for coal-fired power plants, at
atmospheric pressures [3]. Other common major chemical components of flue gas are
nitrogen, oxygen, and water. Therefore, we tested the FeCrCu/K/PMG-20 catalyst in the
CCR process under model conditions with only CO2 (in N2) in the capture phase and
under more realistic conditions which additionally include O2 and H2O vapor,
mimicking post-combustion exhaust gas streams. The CCR performance was studied as
a function of CO2 concentration (Fig. 3A) and temperature (Fig. 3B) under the ideal and
realistic conditions. In all cases, only carbon monoxide and methane were detected, with
a dominant production of the former. One of the most important parameters assessing
the CCR performance is CO2 capture efficiency. This value was calculated based on the
amount of CO2 absorbed/adsorbed during the capture phase, indicating how much CO2
is abated from the CO2 containing gas stream using an integrated system as shown in
Figure 2. There are some portion of unconverted CO2 released mostly in the reduction
phase (Figure 1). The other important parameters are product selectivity and CO2
conversion. The latter value was calculated from the amount of CO2 entered over one
CCR cycle and converted to the products.
Generally, the inclusion of oxygen and water negatively impacted on CO 2 capture
efficiency and CO2 conversion. Nevertheless, the capture efficiency remained almost
100% when 5.8% CO2 was employed under both ideal and realistic conditions between
450-550 °C (Figure 3B) and remained high (>90%) at all inlet CO2 concentrations
investigated (Figure 3A). The lower capture efficiency at higher CO2 concentration is
due to the earlier saturation of the catalyst with CO2 using the CCR period investigated

(thus release of CO2 towards the end of the capture phase) because of the higher amount
of CO2 passed through the CCR reactor per unit time and also a lower amount of CO2
capture capacity under the realistic conditions likely due to competitive adsorption of
oxygen and water over the catalyst. In other words, the CO2 capture efficiency at higher
CO2 concentration can be improved by increasing the amount of catalyst or shortening
the duration of the capture phase. This process flexibility is important to cope with
effluent steams from different processes with varying concentrations of CO2 and other
gases. The remarkable effect of the realistic conditions was the lowered CO2 conversion
by 20-35% compared to the ideal conditions, resulting in larger release of unconverted
CO2 in the reduction phase. Possibly, the presence of water on the surface hindered the
RWGS activity because presence of water is more favorable for the reverse reaction, i.e.
WGS (CO+H2O ⇌ CO2+H2), but this is unlikely because of the consistently higher
selectivity to CO under the realistic condition. It is speculated that the presence of water
and O2 present in the capture phase influences the state of active sites (e.g. by surface
oxidation or adsorption) in a way that hinder RWGS reaction. Besides, higher CO2
concentration as well as higher temperature was found to be more favorable for CO
selectivity. This is explained by the fact that the methanation reaction is an exothermic
reaction which is favored thermodynamically at low temperatures, whereas RWGS is an
endothermic reaction thus favored at high temperatures.

Figure 4. (A) Evolution of CO2 conversion, capture efficiency and CO selectivity
during 750 CCR cycles (45 h) under the ideal and realistic conditions at 550 ºC with
5.8% CO2 in capture phase. (B) Representation of the concentration profiles of the last
CCR cycle.
Other important process factors in practice are stability and regenerability of the
catalyst under unsteady-state operation in a long term. Catalytic tests over 750 CCR
cycles (ca. 45 h) were performed at 550 °C (the temperature where the highest CO 2
conversion and CO selectivity were obtained) under the ideal and realistic conditions
using the FeCrCu/K/PMG-20 catalyst. Figure 4 shows the CCR performance indicators
as a function of CCR cycle number. The catalyst was found to be stable over the period
of time under both conditions. This high stability as well as high carbon balance
(100±5%) suggest that the captured CO2 is totally reduced or released, and CO2 is not
accumulated over a long time cumulatively as carbonates, coke, formate or another
surface carbonaceous species which can lead to catalyst deactivation. The only major
difference between the CCR performances under the ideal and realistic conditions lies in
the product selectivity. As discussed previously, under the realistic condition CO2

conversion is lowered significantly and CO2 was released in the reduction phase. This
lower CO2 conversion is generally unfavored, but this may not be a problem depending
on a targeted product to be synthesized from the syngas containing carbon oxides. The
H2/COx ratio of the gas stream in the reduction phase is ca. 40 and this value is higher
than the stoichiometric ratio of methanol synthesis or Fischer-Tropsch (FT) reaction.
Lower H2/COx ratios will be targeted in our future work. Actually, such a high value
can be considered beneficial as demonstrated for high-performance methanol synthesis
due to thermodynamic and kinetic advantages [22] and unconverted H2 can be recycled
(Figure 2). Moreover, the direct employment of bio-syngas containing H2, CO, CO2 and
CH4 has been developed for the production of Fischer-Tropsch (FT) hydrocarbons via
one-pass through the reactor using unconverted syngas to generate electricity in a gas
turbine combined cycle [23, 24].
4. Conclusions
We have demonstrated a strategy to combine CO2 capture/purification (>99%) step
and further conversion step in one process using unsteady-state operation under
isothermal conditions to produce syngas. Depending on composition of catalysts and
operation condition, the product selectivity and variety can be tuned and altered and
thus the scope of the strategy is extremely broad. CCR offers a mean to simplify the
CO2 utilization strategy by potentially facilitating production of useful chemical directly
at small and large CO2-emisssion sites.
Acknowledgements
We thank the financial support from Repsol. MINECO is acknowledged for support
through CTQ2012-34153 and Severo Ochoa Excellence Accreditation 2014-2018
(SEV-2013-039).

References
[1]

R. Schlögl, Angew. Chem. Int. Ed. 50 (2011) 6424-6426.

[2]

G. Centi, E.A. Quadrelli, S. Perathoner, Energy Environ. Sci. 6 (2013) 1711-

1731.
[3]

K. Bennaceur, in: T.M. Letcher (Ed.), Future Energy (Second Edition), Elsevier,
Boston, 2014, pp. 583-611.

[4]

D.M. D'Alessandro, B. Smit, J.R. Long, Angew. Chem. Int. Ed. 49 (2010) 60586082.

[5]

G.K. Boreskov, Y.U.S. Matros, Catal. Rev. 25 (1983) 551-590.

[6]

Y.S. Matros, G.A. Bunimovich, V.O. Strots, in: G.F. Froment, K.C. Waugh
(Eds.), Stud. Surf. Sci. Catal., Elsevier, 1997, pp. 141-155.

[7]

P. Silveston, R.R. Hudgins, A. Renken, Catal. Today 25 (1995) 91-112.

[8]

W.S. Epling, L.E. Campbell, A. Yezerets, N.W. Currier, J.E. Parks, Catal. Rev.
46 (2004) 163-245.

[9]

S. Roy, A. Baiker, Chem. Rev. 109 (2009) 4054-4091.

[10]

Y.A. Daza, R.A. Kent, M.M. Yung, J.N. Kuhn, Ind. Eng. Chem. Res. 53 (2014)
5828-5837.

[11]

Y.A. Daza, D. Maiti, R.A. Kent, V.R. Bhethanabotla, J.N. Kuhn, Catal.Today
258, Part 2 (2015) 691-698.

[12]

M.S. Duyar, M.A.A. Treviño, R.J. Farrauto, Appl. Catal. B Environ. 168–169
(2015) 370-376.

[13]

Byron Smith R J, Muruganandam Loganathan, M.S. Shantha, Int. J. Chem.
React. Eng. 8 (2010) 1542-6580.

[14]

A. Urakawa, N. Maeda, A. Baiker, Angew. Chem. Int. Ed. 47 (2008) 9256-9259.

[15]

N. Maeda, A. Urakawa, A. Baiker, J. Phys. Chem. C 113 (2009) 16724-16735.

[16]

N. Maeda, A. Urakawa, R. Sharma, A. Baiker, Appl. Catal. B Environ. 103
(2011).

[17]

L.F. Bobadilla, O. Marie, P. Bazin, M. Daturi, Catal.Today 205 (2013) 24-33.

[18]

K.D. Klaus Noweck, Jan Schiefler, Andrea Brasch, Process for producing
hydrotalcites and the metal oxides thereof, Sasol Germany Gmbh, United States,
2001.

[19]

S. Walspurger, L. Boels, P.D. Cobden, G.D. Elzinga, W.G. Haije, R.W. van den
Brink, ChemSusChem 1 (2008) 643-650.

[20]

J.M. Lee, Y.J. Min, K.B. Lee, S.G. Jeon, J.G. Na, H.J. Ryu, Langmuir 26 (2010)
18788-18797.

[21]

T. Hyakutake, W.v. Beek, A. Urakawa,

J. Mater. Chem. A, (2016) DOI:

10.1039/C5TA09461E.
[22]

A. Bansode, A. Urakawa, J. Catal. 309 (2014) 66-70.

[23]

K.-W. Jun, H.-S. Roh, K.-S. Kim, J.-S. Ryu, K.-W. Lee, Appl. Catal. A Gen.
259 (2004) 221-226.

[24]

S. Srinivas, R.K. Malik, S.M. Mahajani, Energy Sustain. Dev. 11 (2007) 66-71.

Supplementary Material
Enabling continuous capture and catalytic conversion
of flue gas CO2 to syngas in one process

Catalyst optimization
Chemicals
PURAL®MG hydrotalcites (Sasol Germany GmbH, Inorganic Specialty Chemicals)
were employed as support. The following hydrotalcites were considered (their Mg/Al
molar ratio is given in parentheses): PURAL®MG 5 (0.08), MG 20 (0.29), MG 30
(0.44), MG 63 (1.89) and MG 70 (2.59). Additionally, a sample consisting only of Al,
PURAL®SB, was also included in this study. K2CO3 from Panreac was used as
precursor of potassium. All the metal precursors salts used in this study were in the
form of nitrates purchased from Sigma-Aldrich.

Synthesis of hydrotalcites-based catalysts
Cu was chosen as the active metal to find the best hydrotalcite-derived mixed oxide
support. The calcined-hydrotalcite-supported Cu-K catalysts were prepared by
sequential impregnation using the incipient wetness method. The catalysts were
prepared in the following steps.
1) Thermal treatment of hydrotalcite support at 600 °C for 3 h in order to obtain the
homogeneous mixed oxides.
2) Impregnation of first precursor: Cu solution was prepared by adding the
necessary volume of deionized water to the metal precursor salts. After the
impregnation the catalyst was dried overnight at 80 °C followed by calcination at 500
°C for 5 h.
3) Impregnation of second precursor: K2CO3 solution was prepared by adding the
necessary volume of deionized water. After the impregnation the catalyst was thermally
treated in the same way as in step 2.
Furthermore, we have prepared two multimetallic catalysts containing Fe-Cr-Cu (Fe/Cr
molar ratio = 10 and Fe/Cu molar ratio = 38) and Cu-Zn (Cu/Zn molar ratio = 2.6), respectively.
The preparation of calcined-hydrotalcite-supported K-multimetallic catalysts was similar to that
of the Cu catalysts described above. In step 2, the nitrate salts of both metals are dissolved in the
same solution and they are co-impregnated.

Catalytic tests
In each run 600 mg of the catalyst was used. Prior to the testing it was pressed,
crushed, and sieved to the range of 200-300 µm. The catalysts are previously reduced at
450 °C for 1 h under 10% H2/He v/v at 40 mL min-1, except for the Fe-Cr based

catalysts. The pretreatment of the latter catalysts was carried out by partially reducing
the hematite (Fe2O3) to magnetite (Fe3O4) using the addition of the process gas mixtures
to activate the catalyst. This also converts any CrO3 present in the catalyst to Cr2O3 [1].
Several CO2 capture/reduction cycles (typically 50) are examined to obtain reproducible
catalytic performances (quasi steady-state). The selected conditions for the evaluation of
the catalytic tests were:
•

Capture phase: 50 mL min-1 of 10 % v/v of CO2 in helium with Gas hourly
space velocity (GHSV) - 500 mL gcat-1 h-1

•

Reduction phase: Pure hydrogen with GHSV of 3900 mL gcat-1 h-1.

•

Period time: 77 s (38.5 s in each phase)

•

Temperature: 450 °C

•

Pressure: 1 bar

Effect of Mg/Al molar ratio
In order to evaluate the effect of the Mg/Al molar ratio on the catalyst performance,
all hydrotalcites provided by SASOL and previously calcined were impregnated with 10
wt% of Cu and 10 wt% of K and compared with the series without potassium. Figure S1
shows the results of CO2 conversion of both series tested. It should be noted that the
presence of potassium increased notably the conversion in all the range of Mg/Al molar
ratio. Furthermore with an increase in Mg/Al molar ratio the conversion of CO2
gradually increased up to 10-20 wt% of MgO and above this loading the conversion
decreased. The presence of excess of MgO likely resulted in the formation of highly
stable MgCO3 species difficult to be reduced.

Figure S1. Effects of MgO percent in MgO-Al2O3 mixed oxides on CO2 conversion in
CCR with and without K.

Effect of potassium loading
We have investigated the effect of potassium loading over the 10 wt% Cu/Pural MG20
catalyst in the range 0-20 wt% of K. Figure S2 shows the catalytic performance as a
function of K loading under the CCR condition. The optimal K loading was found to be
10 wt%.

Figure S2. . Effect of K loading on the CO2 conversion

Multimetallic effect
Long test runs (500 cycles – 12h) using Cu/K/Pural MG20, Cu-Zn/K/Pural MG20
and Fe-Cr-Cu/K/Pural MG20 materials were performed in order to investigate the longterm stability and choose the optimal catalyst for the CCR process. Figure S3 compares
the CO2 conversion and the selectivity into CO and CH4 as a function of the time for the
three catalysts. We can observe that the CO2 conversion of the three materials were
similar and stable, while the CH4 selectivity of Cu monometallic and Cu-Zn bimetallic
catalysts increased with time possibly due to sintering of Cu. For this reason we have
decided to focus on FeCrCu trimetallic catalyst in order to investigate the efficiency of
the CCR process.

Figure S3. CO2 conversion and selectivity to CO and CH4 for Cu/K/PMG-20, CuZn/K/PMG-20 and Fe-Cr-Cu/K/PMG-20 catalysts

Catalyst characterization
Characterization methods
The elemental composition was determined by ICP analysis.
Nitrogen isotherms at 77 K were measured on Quantochrome Autosorb 1-MP
analyzer. Prior to analysis, the sample was degassed in vacuum at 573 K for 12 h.
The XRD pattern was recorded on Bruker AXS D8 Advance diffractometer equipped
with a Cu tube, a Ge (1 1 1) incident beam monochromator (λ = 0.1541 nm), and a
Vantec-1 PSD operated in transmission mode. Data were recorded in the range of 5-70°
2θ with a step size of 0.02° and a counting time of 4 s per step. For CO2-TPD, the
sample (100 mg) were firstly reduced in 5% H2 in N2 at 20 mL min-1 for 1 h at 450 °C,
followed by exposure of 4% CO2 in inert at 80 ºC for 1 h at 20 mL min-1.

TPD was performed under 5% H2 in N2 flow at 20 mL min-1 to investigate the
stability of CO2 adsorbed under a reducing environment as in CO2 hydrogenation in
CCR process. The temperature was increased from 50 up to 900 °C. The effluent gases
were analyzed on a mass spectrometer (MS), Pfeifer Omnistar GSD 301 C, without any
trap.
In order to estimate the CO2 capture capacity we have calculated the amount of CO2
captured in the equilibrium at 400 °C. The equilibrium CO2 sorption uptake was
measured by CO2 chemisorption using Quantachrome Autosorb iQ system.
Chemisorption is a strong specific interaction that affects only the surface parts
occupied by atoms that can chemically interact with the gas. It allows determining the
amount of surface active sites and possibly the amount of CO2 chemisorbed and
physisorbed. Before CO2 sorption, moisture and CO2 on samples was removed under
hydrogen flow for 1 h at 450 °C. After the temperature was decreased to 400 °C under
nitrogen flow, CO2 was adsorbed at 400 °C until the system reaches the equilibrium
pressure (800 Torr).
Characterization results
Table S1 shows the chemical compositions and textural properties (BET surface
area, pore size and pore volume) of the FeCrCu/K/PMG20 catalyst.
Table S1. Textural properties and elemental analysis
BET surface area
m2 g-1
111

Pore volume
cm3 g-1
2.11

Al2O3
66.53

Elemental analysis / wt.%
MgO
Fe
Cr
Cu
15.80
6.91
0.58
0.20

K
9.98

Figure S4 shows XRD patterns of FeCrCu/K/PMG20 catalyst. The diffraction peaks
obtained can be assigned to the following phases: hematite Fe2O3 (JCPDS 73-0603),

Cr2O3 (JCPDS 84-0315), monoclinic K2CO3 (JCPDS 16-820), cubic K2O (JCPDS 23493), spinel MgAl2O4 (JCPDS 21-1152) and γ-Al2O3 (JCPDS 10-425). Diffraction
peaks associated to Cu phases were not observed likely due to its high dispersion and
low concentration.

Figure S4. XRD pattern of FeCrCu/K/PMG20 catalyst

In order to investigate the stability of CO2 adsorbed and stored as carbonates, CO2TPD was performed after pre-reduction of the catalyst to render storage sites in their
active oxide form prior to CO2 adsorption (their decomposition was confirmed by MS)
and subsequent CO2 adsorption. CO2 desorption was studied under a hydrogen
atmosphere in order to study the stability of such adsorbed and carbonate species under
a reductive environment as in CCR. Figure S5 shows the CO2-TPD profiles in the
presence of hydrogen. CO2 desorption shows three peaks (I), (II) and (III) at 200, 400
and 650 °C, respectively. These three peaks can be ascribed to three different surface
carbonate species with different stability. Bansode et al. [2] assumed that peak (I)
originates from the decomposition of surface formates and peaks (II) and (III) from that
of surface carbonates. Importantly, the major desorption is characterized by peaks (I)
and this suggests the high reactivity of CO2 captured as carbonate/formate species and

that the desorption site corresponding to these peaks play crucial roles in determining
the catalytic performance in the CCR process.

Figure S5. CO2-TPD profile of FeCrCu/K/PMG20 catalyst under hydrogen

The equilibrium CO2 uptake was measured by CO2 chemisorption. The amounts of
chemisorbed and physisorbed CO2 over the FeCrCu/K/PMG20 catalyst are shown in
Table S2.
Table S2. CO2 sorption capacity estimated by CO2 equilibrium chemisorption at 400 °C

CO2 chemisorbed
mmol g-1
0.140

CO2 physisorbed
mmol g-1
0.157

CO2 total
mmol g-1
0.297

References
[1] Byron Smith R J, Muruganandam Loganathan, M.S. Shantha, Int. J. Chem. React.
Eng., 8 (2010) 1542-6580.
[2] A. Bansode, B. Tidona, P.R. von Rohr, A. Urakawa, Catal. Sci. Technol., 3 (2013)
767-778.