Available online at www.sciencedirect.

com

ScienceDirect
Energy Procedia 63 (2014) 2082 – 2091

GHGT-12

In-situ XRD studies of dolomite based CO2 sorbents

Bjørnar Arstad1*, Anna Lind1, Kari Anne Andreassen1, Joanna Pierchala1, Knut
Thorshaug1, Richard Blom1.
1
SINTEF Materials and Chemistry, Forksningsveien 1, 0314 Oslo, Norway

Abstract

Calcined dolomite (a CaOMgO material derivative of the natural mineral dolomite) has the potential as a CO 2
sorbent working at high temperatures (500-650 °C) in processes like post combustion carbonate-looping, or in
sorbent enhanced reactions such as sorbent-enhanced water-gas shift (SEWGS) or sorbent-enhanced reforming
(SER). However, deactivation and performance loss is inevitable and much effort has been aimed towards
understanding and improving solid sorbents for various applications. This work presents a study of deactivation
trends of calcined dolomite and Zr-modified versions of calcined dolomite that appears to possess better cyclic
properties at rapid sorption-desorption cycling (which is a likely conditions in a real process using a CFB reactor).
Surface area measurements, thermogravimetric analyses (TGA), fixed bed reactor studies, in-situ XRD and in-situ
IR methods carried out at 600ºC in a gas containing 1-2vol% steam in 10% CO2 have been conducted to obtain
fundamental information about the de-activation mechanisms taking place.
© 2014
© 2013TheTheAuthors. Published
Authors. by Elsevier
Published Ltd. Ltd.
by Elsevier This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of GHGT.
Peer-review under responsibility of the Organizing Committee of GHGT-12
Keywords: CCS, CO2 capture, sorbent, dolomite, in-situ XRD, in-situ IR, TGA, fixed bed reactor, breakthrough measurements

1. Introduction

Reduction of anthropogenic CO2 emissions is important in order to mitigate global warming. One main source of
CO2 emissions is power production from fossil fuels, and huge efforts have been carried out to enable large scale
solvent based CO2 capture processes that can be retrofitted to existing power plants. These will likely be based on

* Corresponding author. Tel.: +47-98243498; fax: +47-22067350.

E-mail address: bjornar.arstad@sintef.no

1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Peer-review under responsibility of the Organizing Committee of GHGT-12
doi:10.1016/j.egypro.2014.11.224
 Bjørnar Arstad et al. / Energy Procedia 63 (2014) 2082 – 2091 2083

water solutions of amines [1]. Other technological solutions for removing CO2 are e.g. based on membranes or solid
sorbents but are at present less mature than solvent based processes. However, many of these show promising
economical potentials as exemplified by the high temperature solid looping sorbent process "Calcium Looping",
CaL, and "Sorption enhanced steam methane reforming", SE-SMR, which both are assessed by techno-economic
evaluations to have a reduced efficiency penalty compared to standard amine based processes [2,3,4,5].
The CaL process is a post combustion technology that is based on the idea that CaO(s) is reacting with CO 2(g) to
CaCO3(s) hence removing gas phase CO2 from a gas mixture, see eq. 1 below.

CaO(s) + CO2(g) ' CaCO3(s) ΔH298K = -179 kJ/mol eq. (1)

CaL has received large interests during the last 10-15 years, with several small (10-100 kW) and medium sized
units (1-2 MW) demonstrated [6,7], but there is still a concern connected to the cost, and particular, the lifetime of
the sorbent powder(lime stone).
Different from the CaL, but with essential similarities concerning sorbent use, is the SE-SMR. SE-SMR is a
process for improved hydrogen production with simultaneous CO2 capture since the reactions are driven towards H2
as CO2 is captured by the solid sorbent (typically CaO or calcined dolomite, CaOMgO). The hydrogen, which has a
purity of about 97-98 % if CaO is the active sorbent component, may then be further purified for the hydrogen
market or burned for electricity generation.
The CaL and SE-SMR processes are suited for circulating fluidized bed (CFB) reactor systems [6,7,8]. The most
essential features of a CFB CaL-process and the SE-SMR processe are shown schematically below:

H2O + N2 CO2 + H2O

MO(s) + CO2(g)
SE-SMR unit
MCO3(s)
Regeneration

Natural gas Steam reforming

Carbonation

Fuel burning in + water-gas shift H2

this reactor Steam + CO2 capture
provides heat for
regeneration of
sorbent
MO MCO3
Sorbent CO2
regeneration
Flue gas 500-700°C Gas/coal/biomass + O2

Figure 1. Left: Schematic drawing of the Ca-looping process, and right: Principle of the sorption enhanced steam methane reforming process

As in all processes that utilize powders sorbent deactivation in the CaL and SE-SMR processes are inevitable. As
a consequence of the somewhat limited lifetime of easily accessible natural minerals such as lime and dolomite,
much research has been aimed towards making new artificial (nano)materials. However, due to CaO's favorable
thermodynamic properties combined with relatively fast reaction rates with CO2 most recent research has been
focused on making materials that have CaO as the active phase, possibly fixed in a suitable matrix. Various
strategies for preparing improved CaO based materials are e.g.: Thermal pre-treatments of lime [9], hydration of
lime [10], doping CaO by KMnO4 [11], dispersion of CaO on various supports like e.g. Ca 12Al14O33 or CaTiO3 [12,
13], or co-precipitation of Ca with other elements (M) like Cr, Mn, Co, Cu, or Ce, for a nominal composition of
Ca0.9M0.1Ox [14], or CaO on MgO mixed with Ca 9Al6O18 [15], or CaCeZrOx [16]. Lately, we tried to deposit Ti-,
Zr-, and Al-nanoparticles on calcined dolomite in order to stabilize and improve cyclic performance without any real
success [17]. However, the idea was followed further and by focusing on other Zr-sources, different loaded amounts,
and other experimental conditions we hoped to improve the materials. In addition we coupled this work with studies
of deactivation trends of calcined dolomite and applied several quite different techniques during the experimental
work. After material preparations we used N2 BET and micropore volume measurements to assess surface and pore
volumes, powder XRD to investigate new phases present after Zr-modifications, thermogravimetric analyses (TGA)
to measure CO2 uptake capacities of the sorbents, dual fixed bed reactor for measuring CO 2 breakthrough curves and
to have an improved control of flow and reactivity compared to normal TGA analyses. The fixed bed rig has also
2084 Bjørnar Arstad et al. / Energy Procedia 63 (2014) 2082 – 2091

been used to study rapid adsorption-desorption cycling performance, which it is very suited for. In addtion we have
performed in-situ XRD and in-situ DRIFTS FTIR characterization of fresh and aged materials under realistic
conditions, including cycling between soprtion and desorption of CO 2, in order to unravel fundamental aspects of
the performance and decay trends. All sorption experiments for all methods have been at 600 °C using a 1-2% steam
in 10% CO2 balanced in N2 gas mixture. Regeneration was carried out using N2 gas at 850 or 900 °C.

2. Experimental

The starting material in this work is the natural mineral dolomite (CaMg(CO 3)2) from Seljelid, Norway. In order
to prepare dolomite as an acceptor for CO2 the oxide version was formed by heat treatment (calcination) of the
mineral for 12 hours at 900°C in 10% H 2. This procedure transformed the dolomite into an oxide form that can be
termed CaOMgO, or calcined dolomite (there is close to a 1:1 ratio of Mg and Ca). Two other Zr modified samples,
with Zr nominal loadings of 0.5wt.% and 1wt.%, were made by impregnating the calcined dolomite with Zr-nitrate
in line with incipient wetness methods. After impregnation the samples were calcined in air for 15 hours at 900 °C.
Higher loadings were also made but these were inferior but are mentioned later (vide infra).
BET surface and micorpore volume were measured by nitrogen adsorption at 77K for all samples using a
BELSORP Mini instrument.
In order to asses the cyclic CO2 uptake/desorption capacity and kinetics, we studied the materials by Thermo
Gravimetric analysis (TGA) and a fixed bed rig for CO2 breakthrough measurements. Not all materials were tested
in both apparatuses. The gas composition used in these instruments for testing the materials were in all cases a mix
of 10vol.% CO2 and 1vol.% H2O balanced with N2. This was obtained by bubbling the 10% CO2 in N2 mix through
water cooled to about 11°C with a total pressure of 1 atm. The TGA experiments were carried out in a Netzsch
STA449 F1 instrument using the following conditions: Fist, the temperature was raised to 900 °C at 10ºC/min in N 2.
The temperature was kept constant for 3h before it was lowered to 600 °C at a rate of 10 °C/min. After 5 minutes at
600 °C in N2 the sample was exposed to the reactive gas for 20 minutes. After the sorption step the temperature was
raised to 900 °C for 30 minutes in pure nitrogen in order to regenerate the sorbent. Then the temperature was
lowered to 600 °C. The steps after the initial 3h at 900 were repeated 60 times. The outlet gas was analyzed by an
online mass spectrometer (MS) and an in-house developed Labview program controlled the whole gas feeding
sequence automatically.
In order to asses further the CO2 uptake/desorption capacity and kinetics, we tested the materials in a fixed bed
reactor and measured breakthrough curves. Figure 2 shows schematically the principle of the dual fixed bed test rig
and a typical raw data output.

30

25

20
ml/min CO2

15

10

5
Sorption De-sorption
0
0 20 40 60 80 100 120 140 160 180

Time / minutes

Figure 2. Left: Principle of the breakthrough rig and example of breakthrough curve. Right: The black line is the MS CO 2 signal (actually the
m/z=44 signal)
 Bjørnar Arstad et al. / Energy Procedia 63 (2014) 2082 – 2091 2085

The inner diameter of the reactor tubes where 1 cm and we applied a 4 cm high bed. The amount sorbent was
typically between 2.3 and 2.4 grams. The reactor contained the sorbent while the dummy reactor contained quartz.
Temperatures in the bed were measured at four points along the bed height. The gas flows were 50ml/min through
both reactors, one stream was pure N2 and the other was 10% CO2 in N2 saturated with H2O at 10-11°C, i.e. app. 1%
water similar to the TGA experiments. The outlet gas was analyzed by an online MS and an in-house developed
Labview program controlled the whole sequence automatically. In the initial phase of a test the reactor tubes were
heated to 600 °C by an electrically furnace. Pure N2 was entering the reactor with the sorbent while the CO2
containing gas entered the dummy reactor. At this stage the continuous CO2 MS signal (m/z = 44) shows the
intensity of 5 ml/min CO2 (the first minutes of the MS trace above). After ten minutes at 600 °C a valve switched so
that the CO2 containing gas enters the reactor tube with sorbent and the pure N2 gas enters the dummy reactor. The
valve switch is seen by a spike in the MS signal. During sorption CO2 is reacting with the sorbent and is removed
from the gas phase and the MS signal from CO2 (m/z = 44) declines. After some time with adsorption of CO2 the
MS signal starts to increase termed a CO2 breakthrough, i.e. the sorbent does not absorb all the CO 2 entering the
powder bed. Further on in time, the MS signal from CO2 rises almost to the initial level observed before the sorption
took place. Other reaction conditions, like e.g. higher partial pressure of CO 2, more added steam, or other
temperatures will changes this picture but the general features of the curve would still be the same. After a
predetermined time (10 or 120 minutes as in our cases reported here) with CO2 gas over the sorbent the gases are
switched again and the temperature was increased to 850 °C. During the heating period in N 2 (regeneration) the
sorbent released the previously absorbed CO2 and this is indicated as the de-sorption region in the figure. After
desorption was complete a cycle had passed. Experiments were done with varying bed heights, reaction times and
flows in order to ensure reliable and proper data. In one experiment reported below the adsorption time was 120
minutes and 13 cycles were carried out, while for the other experiments 90 cycles were carried out with 10 minutes
adsorption times. In these 10 minutes adsorption cycle experiments, there was in practice no CO2 breakthrough
when switching over to regeneration conditions and it took about a week to complete all 90 cycles. The absorption
capacity was evaluated using an in-house developed Excel macro software. The conversion of sorbent is also
evaluated and is defined as the moles CO2 adsorbed pr. mol CaO unit in the sample and is based on the following
estimate of total capacity. Since the molecular weight of CaOMgO(calcined dolomite) is 96 g/mol CaO constitutes
100%* 56/96=71% of the weight. 1 gram sorbent contains around 710 mg CaO which corresponds to 12.7 mmol
and is also the amount CO2 possible to adsorb if given enough time.
The X-ray diffraction (XRD) measurements were performed on a PANalytical Empyrean diffractometer. The
system is equipped with a PIXcel3D solid state detector. The measurements were carried out in reflection geometry
using CuKα1 radiation (α= 1.54060 Å) and a step size of 0.013 degrees. An Anton Paar XRK 900 in-situ high-
temperature (25-900°C) and high pressure (1-10 bar) reactor cell coupled to an automated gas switching system was
used for the in-situ XRD experiments. The pressure was 1 atm. and the feed gas employed was a 10% CO2 in N2
mix saturated with H2O at 10-11°C, i.e. app. 1% water, for a total flow 100 ml/min. Ca. 150-160 mg sample powder
was used for each experiment. For the plotting and analyses of the in-situ XRD the reflection at 2θ=29.3° for CaCO3
and at 2θ=37.2°for CaO. The experimental reaction condition protocol is schematically shown below.

30 min 30 min 30 min 30 min

850 °C

600 °C
10 60 min 1060 min
10 10 60 min

Heating/cooling rate: 10°C/min

--- Air
--- N
2
--- CO + H O
2 2

20 min 20 min
25 °C
Figure 3. The reaction condition protocol for the in-situ XRD experiments.
2086 Bjørnar Arstad et al. / Energy Procedia 63 (2014) 2082 – 2091

FTIR spectra were recorded on a Bruker Vertex 70 equipped with a MCT-detector, DRIFTS-cell (HVC-DRP-2),
temperature controller, and Praying-Mantis from Harrick. The spectra were recorded at a resolution of 4 cm-1, and
each spectrum is the average of four scans. All spectra were collected in reflectance mode, and we used a spectrum
of the solid samples under pure N2 collected just prior to their exposure to the gas mixture as background spectra.
The solid samples were loaded into the sample cup and heated under flowing N2 according to the same temperature
program as used for PXRD. When the temperature was stable at 600°C, we carefully switched from a pure N 2-flow
to a flow of 10% CO2 diluted in N2 that was bubbled through distilled H2O before it entered the DRIFTS-cell. Due
to the dead volume in the gas-feeding system and DRIFTS-dome, we define t = 0 when the first visible bands
appeared in the spectra.

3. Results and discussion

3.1. Deactivation studies of calcined dolomite

Table 1. Material summary for deactivation studies of calcined dolomite

Type BET area, m2/g Micropore volume, cm3/g

Calcined dolomite 19.2 0.26
Wet-90, 90 cycles wet CO2 9.6 0.18
Wet-13, 13 cycles wet CO2 9.2 0.15

The fixed bed experiments were carried out such that the total CO2 capacity that had been adsorbed by 90 short
cycles was about the same as in 13 long adsorption cycles. The adsorbed amounts pr. cycles for these two
experiments are shown in figure 4. The total CO2 uptake was 77 mmol for the 90 cycle experiment and 83 mmol for
the 13 cycle experiment. The total CO2 uptake, BET surface and micropore volume of these used materials were
therefore quite similar.

8,00 1,0
7,00 0,9
mmol CO2 / g sample

mmol CO2 / g sample

0,8
6,00
0,7
5,00 0,6
4,00 0,5
3,00 0,4
0,3
2,00
0,2
1,00
0,1
0,00 0,0
1 2 3 4 5 6 7 8 9 10 11 12 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86
Cycle number Cycle number

Figure 4. Left: 13 cycles wet CO2 with conversion going from about 59% to ~40%. Right: 90 cycles wet CO2 with conversion about 7 %

The fixed bed experiments clearly shows that by employing most of the available capacity in each cycle a clear
deactivation trend is observed with a steady reduction in the cyclic capacity for each cycle, which is in contrast to
the data from the short adsorption time experiment as the performance is apparently quite stable for the 90 cycles.
Representative SEM images of the used samples are shown below but did not give any hints to the different
trends. Hence the deactivation profile during long cycle experiments is likely not as a result of big morphological
changes.
 Bjørnar Arstad et al. / Energy Procedia 63 (2014) 2082 – 2091 2087

Figure 5. SEM images after 13 long ad-des cycles (left) and 90 short ad-des cycles. The SEM images show that there are no apparent big
differences in morphology for the samples.

In an attempt to further understand the differences in performance depending upon the adsorption time in-situ XRD
experiments were carried out on the two used samples and a calcined unused dolomite. Intensities of selected
reflections (as described in part 2) are plotted as histograms. Scan number is a complete XRD data set and is a sum
of several scans again.

Figure 6. in-situ XRD plot showing the CaCO3 and CaO evolution through 3 ad-des cycles. Left: Calcined dolomite. Middle: After 13 cycles wet
CO2 . Right: After 90 cycles wet CO2.

As one can se the unused calcined dolomite shows much faster and complete reaction with CO2 to CaCO3. The
90 cycle material has been deactivated compared to the unused calcined dolomite but not as much as the long cycle
material. This is complementary circumstantial evidence indicating that applying a smaller adsorption capacity leads
to a longer lifetime, even at about the same captured CO2 amounts. This might also well be similar in other sorbents
and is not really surprising considering that the less CO 2 adsorbed the less strain on the materials during the
reactions.

3.2. Improving performance of calcined dolomite by Zr impregnation

Understanding deactivation of sorbents for an eventual improvement has been an important topic for long time.
Based upon the idea that an added element might act as a promoter we impregnated the calcined dolomite with
various loading of Zr. The Zr modified materials were made with the aim of improving cyclic stability, however, by
impregnating a compound like calcined dolomite, which has a relative small surface area, the available surface area
for reactions might become lower. Table 2 below shows this effect but it is not until 5wt.% Zr loading the BET and
micropore volume decrease significantly. At 10 wt.% Zr loading the effect is similar to the 5wt.% material. Neither
of these will be further described here as they had low performance compared to the other samples.
2088 Bjørnar Arstad et al. / Energy Procedia 63 (2014) 2082 – 2091

Table 2. Zr modified material summary

Type BET area, m2/g Micropore volume, cm3/g

0.5 wt.% Zr 16.1 0.24
1 wt.% Zr 11.7 0.10
5 wt.% Zr 5.7 0.06

The impregnated samples were analyzed with powder XRD and for all materials with added Zr there had been
formed an extra perovskite phase that could be e.g. CaZrO3. This phase is only a minor component and has no
significant reactivity towards CO2 but it may be detrimental to the sorption process anyway. In figure 7 below a
typical XRD is shown.

x x x

10 20 30 40 50 60 70 80 90 100
2 Theta
Figure 7. XRD data for the 1wt.% Zr sample. There are small reflections that are identified as a perovskite phase and could maybe be e.g. CaZrO3
(some reflections are marked with x). This phase is seen in all samples impregnated with Zr. Other peaks are from CaO (the active phase) and
MgO.

In order to assess basic cyclic sorption performance we tested the materials in a thermogravimetric analyses
instrument (TGA) and a fixed bed reactor as described in section 2. Below are the TGA data for the 0.5wt.% and
1wt.% samples listed in Table 2. The 5wt.% Zr sample is not shown due to significant less CO 2 uptake capacity
(around stable 11-12 wt.%) and at 10wt.% the capacity is further down to about 6-7 wt.% CO2 uptake.

40 40

30 30
Weight % increase

Weight % increase

20 20

10 10

0 0
0 2000 4000 6000 0 2000 4000 6000
Time / min Time / min

Figure 8. TG data for the 0.5wt.% Zr (left), and the 1% Zr (right) modified materials. Adsorption at 600°C using 10% CO2, 1% H2O balanced in
N2. Desorption at 850°C in N2
 Bjørnar Arstad et al. / Energy Procedia 63 (2014) 2082 – 2091 2089

For further cyclic performance studies the 0.5wt.%, and the 1% Zr sample were tested in the fixed bed reactor at so
short CO2 uptake times that the conversion was below 10%, which is somewhat different from the relative long
TGA adsorption times. These experiments should be closer to industrial conditions as e.g the initial CaL article from
Shimizu et al. evaluated the CaL process using only 10% sorption capacity [3].

1,0 1,0 1,0

0,9 0,9 0,9
0,8 0,8 0,8
0,7 0,7 0,7
mmol CO2/g sample

mmolCO2/g sample

mmol CO2/g sample

0,6 0,6 0,6
0,5 0,5 0,5
0,4 0,4 0,4
0,3 0,3 0,3
0,2 0,2 0,2
0,1 0,1 0,1
0,0 0,0 0,0
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86
Cycle number Cycle number Cycle number

Figure 9. Left: Fixed bed data for calcined dolomite (the same data as shown in figure 4 right but included here for better comparisons with the Zr
materials) , 0.5wt.% Zr (middle), and the 1% Zr modified calcined dolomite (right). Adsorption at 600°C using 10% CO2, 1% H2O balanced in
N2. 10 minutes adsorption time for each of the 90 cycles. Desorption at 850°C using N2. The conversion of calcined dolomite is quite stable at
around 6.7% while the 1% modified sample was around 7.7%. The 0.5% Zr sample appears to be between these two materials if the somewhat
unstable capacity form cycle to cycle is averaged.

As can be seen from the fixed bed reactor data there is a transition from the calcined dolomite, through the
0.5wt.% Zr sample to the 1wt.% sample with a further loss in performance when testing the 5 and 10 wt.% Zr
materials (not shown).
Unused 1wt.%Zr and the same material after usage in 90 short adsorption cycle were studied by in-situ XRD.
The figure below shows these results.

Figure 10. in-situ XRD plot showing the CaCO3 and CaO evolution through 3 ad-des cycles. Left: Unused 1%Zr material. Right: 1%Zr after 90
cycles wet CO2 in the fixed bed reactor. There are no big differences hence the material hasn't deactivated much during 90 cycles.

Comparisons of the in-situ XRD data for the 1wt.% Zr sample before and after 90 short adsorption cycles shows
few, if any large changes. This is a strong indication that during use this material deactivates less compared to
calcined dolomite even if it shows a higher working cyclic capacity.
t-resolved FTIR-spectra of calcined dolomite and the 1wt.% Zr sample at 600 °C exposed to a flow of 10% CO 2
in N2 saturated with H2O are shown in Fig. 11 below. Calcined dolomite is at the top and the Zr sample is at the
bottom. As seen from the figure, the spectra were clearly different, and this difference suggests that the
impregnation of 1 wt. % Zr alters the surface chemistry of the calcined dolomite.
2090 Bjørnar Arstad et al. / Energy Procedia 63 (2014) 2082 – 2091

0,1 Reflectance 0,1 Reflectance

4000 3500 3000 2500 2000 1500 1000 1800 1600 1400 1200 1000

0,1 Reflectance 0,1 Reflectance

4000 3500 3000 2500 2000 1500 1000 1800 1600 1400 1200 1000
Wavenumber [cm-1] Wavenumber [cm-1]

Figure 11. t-resolved FTIR-spectra collected during exposure to 10% CO2 in N2 bubbled through H2O to calcined dolomite (upper) and 1 wt% Zr
impregnated on calcined dolomite (lower). The spectra on the right hand side show the details for the region 1900 – 1000 cm-1.

The region 1900 – 1000 cm-1 is particularly revealing (enlarged at the right in figure 11). Calcined dolomite
showed several bands; particularly at 1784, 1617, 1546, 1440, 1352, 1293, and 1066 cm-1. Furthermore, their
development with time was different. As seen, the bands that initially were the most dominant became less dominant
with time. Finally, they appeared as mere shoulders on the dominant band centered at 1440cm -1. The same
observation was not made for the sample containing 1wt% Zr impregnated on calcined dolomite. As seen, there
were three distinguishable bands centered at 1784, 1477, and 1413 cm-1 at all times.
Although a full in-depth assignment of the bands is beyond the scope of this work, some tentative and general
comments are warranted. First of all, we note that the number of bands observed for the sample impregnated with Zr
shows the fewest number of bands, and we interpret this as a lower number of carbonate adsorption modes. It may
be speculated that this is due to the formation of a Zr-containing surface layer that is less CO2-reactive than the
surface in calcined dolomite, and this interpretation would be in line with a perovskite layer as detected by PXRD.
The band centered at 1784 cm-1 is the only one in common for the two solids, and it may be assigned to bridging
bidentate carbonate [18]. Both spectra also show bands that may be tentatively assigned to monodentate carbonate
[19]. We also note a band that grows slightly upwards at 3691, and 3674 cm -1 for the calcined dolomite and the 1
wt% Zr sample, respectively. This indicates that CO2 may interact with some residual surface OH-groups after
calcination, and that these OH-groups are different for the samples.

4. Conclusion

In this article we have described cyclic performance and deactivation trends of calcined dolomite and Zr
modifications of the material in CO2 sorption reactions. We found that above a certain level of Zr (more than 5wt.%)
the surface and pores are likely blocked and the performance found inadequate. However, by using either 0.5wt% or
1wt.% Zr the BET surface are mostly still intact and the pore volume is only reduced by a factor of two. It is
noteworthy that the 1% Zr sample showed a stable performance but with a lower total capacity compared to calcined
dolomite in the TG apparatus but in the fixed bed reactor using short adsorption times the conversion was actually
 Bjørnar Arstad et al. / Energy Procedia 63 (2014) 2082 – 2091 2091

better. The relative improvement was 15% and indicates that impregnating calcined dolomite, or maybe CaO, with
small amounts of Zr may lead to very interesting cyclic performance properties at short adsorption times (which is
more relevant from an industrial application point of view). However, Zr impregnated sorbents might not be used if
e.g. CaL is used in conjunction with cement production since the spent sorbent is then used directly into the actual
cement manufacturing so the Zr-modified sorbents might therefore be more suited in e.g. SE-SMR or CaL in
conjunction with power production. Both the in-situ techniques (XRD and FTIR) applied in this work showed great
potential for further fundamental studies.

Acknowledgments

This publication has been produced with support from the BIGCCS Centre, performed under the Norwegian
research program Centres for Environment-friendly Energy Research (FME). The authors acknowledge the
following partners for their contributions: Aker Solutions, ConocoPhilips, Det Norske Veritas, Gassco, Hydro,
Shell, Statkraft, Statoil, TOTAL, GDF SUEZ and the Research Council of Norway (193816/S60).

References

[1] Rochelle GT. Amine Scrubbing for CO2 Capture. Science 2009;325:1652-4.
[2] Shimizu T, Hirama T, Hosoda H, Kitano K, Inagaki M, Tejima K. A twin fluid-bed reactor for removal of CO2 from
combustion processes. Chem. Eng. Res.Des. 1999;77:62-68.
[3] Zhao M, Minett AI, Harris AT. Energy Environ. Sci. 2013;6:25-40.
[4] Harrison DP. Sorption-enhanced hydrogen production: A review. Ind. Eng Chem. Res. 2008;47:6486-6501.
[5] Ochoa-Fernández E, Haugen G, Zhao T, Rønning M, Aartun I, Børresen B, Rytter E, Rønnekleiv M, Chen D. Green Chem.
2007;9:654-662.
[6] CaOling project, http://www.caoling.eu/, 7Framework Programme EU
[7] Orth M, Ströhle J, Epple B. Int. Conf. on Chemical Looping, 26th to 28th of September 2012, Darmstadt, Germany.
[8] Arstad B, Prostak J, Blom R. Continous hydrogen production by sorption enhanced steam methane reforming
(SE-SMR) in a circulating fluidized bed reactor: Sorbent to catalyst ratio dependencies. Chem. Eng. J.
2012;189-190:413-421.
[9] Manovic V, Anthony EJ. Thermal activation of CaO-based sorbent and self-reactivation during CO2 capture
looping cycles. Environ. Sci. Technol. 2008;42:4170-4174.
[10] Fennell PS, Davidson JF, Dennis JS, Hayhurst AN. Regeneration of sintered limestone sorbents for the
sequestration of CO2 from combustion and other systems. J. Energy Inst. 2007;80:116-119.
[11] Hughes RW, Lu DY, Anthony EJ, Wu YH. Improved Long-Term Conversion of Limestone-Derived Sorbents
for In Situ Capture of CO2 in a Fluidized Bed Combustor. Ind. Eng. Chem. Res. 2004;43:5529-5539.
[12] Li ZS, Cai NS, Huang YY, Han HJ. Synthesis, Experimental Studies, and Analysis of a New Calcium-Based
Carbon Dioxide Absorbent. Energy Fuel 2005;19:1447-1452.
[13] Aihara M, Nagai T, Matsushita J, Negishi Y, Ohya H. Development of porous solid reactant for thermal-
Energy storage and temperature upgrade using carbonation/decarbonation reaction. Appl. Energy 2001;69:225-
238.
[14] Lu H, Smirniotis PG, Calcium Oxide Doped Sorbents for CO2 Uptake in the Presence of SO2 at High
Temperatures. Ind. Eng. Chem. Res. 2009;48:5454 -5459.
[15] Chen H, Zhao C. Development of a CaO-based sorbent with improved cyclic stability for CO2 capture in
pressurized carbonation. Chem. Eng. J. 2011;171:197-205.
[16] Sultana KS, Chen D, Enhanced hydrogen production by in situ CO2 removal on CaCeZrOx nanocrystals. Catal.
Today 2011;171:43-51.
[17] Arstad B, Spjelkavik A, Andreassen KA, Lind A, Prostak J, Blom R. Studies of Ca-based high temperature
sorbents for CO2 capture. Ener. Proc. 2013;37:9-15.
[18] Philipp R, Fujimoto K. FTIR spectroscopic study of carbon dioxide adsorption/desorption on magnesia/calcium oxide catalysts. J. Phys.
Chem. 1992; 96:9035-9038.
[19] Stevens RW, Siriwardane RV, Logan J. In situ Fourier transform infrared (FTIR) investigation of CO2 adsorption onto zeolite materials.
Energy & Fuels 2008;22:3070-3079.