Applied Energy 140 (2015) 297303

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

CO2 capture at ambient temperature in a xed bed with CaO-based

sorbents
Firas N. Ridha a, Vasilije Manovic b, Arturo Macchi a, Edward J. Anthony a,c,
a
Department of Chemical and Biological Engineering, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada
b
Energy & Power Engineering Division, School of Engineering, Craneld University, Craneld, Bedfordshire MK43 0AL, United Kingdom
c
School of Applied Science, Craneld University, Craneld, Bedfordshire MK43 0AL, United Kingdom

h i g h l i g h t s

 CaO-based pellets can be used for CO2 capture at ambient temperature.

 Pre-hydration of pellets improves CO2 capture capacity signicantly.
 CO2 capture capacity of 0.5 g/g is obtained in a CaO-based xed bed.
 The capture capacity is sensitive to moisture and CO2 inlet concentration.

a r t i c l e i n f o a b s t r a c t

Article history: This work investigates post-combustion CO2 capture at ambient temperature in a xed bed by means of
Received 24 April 2014 CaO-based sorbents. Two sorbents were used: limestone and pellets prepared from powdered limestone
Received in revised form 10 November 2014 using calcium aluminate cement as a binder. The results showed that pre-hydration had a signicant
Accepted 14 November 2014
effect on CO2 capture performance of the two sorbents. For instance, after 8 h pre-hydration, the
Available online 18 December 2014
breakthrough time increased from 21 min to 660 min for lime, and from 19 min to 750 min for pellets.
The performance of pellets was more sensitive to hydration conditions than for the lime. At break-
Keywords:
through, full carbonation conversion over half of the reactor was achieved in a pre-hydrated bed of pellets
CO2 capture
Limestone
exposed to a feed with 0.5% CO2, resulting in an average specic capture of 0.51 g CO2/g bed material. This
Pelletization was considered a sufcient capture performance, with a distinct mass transfer zone (MTZ) located in the
CaO-based sorbents upper half of the reactor. However, increasing CO2 inlet concentration to 2% shortened the breakthrough
Fixed bed time and shifted the MTZ toward the entrance zone of the reactor. It was concluded that capturing CO2
from low-CO2 ue gases at ambient temperature using a xed bed of pre-hydrated CaO-based pellets is a
promising approach that has the potential to achieve reasonable capture performance at relatively low
cost. The proposed process can be used for CO2 capture from CO2-depleted ue gases (residual CO2) from
processes such as amine scrubbing and calcium looping. And it would allow for the possibility that cap-
ture could be increased to any given level required by new legislation for plant with carbon capture.
2014 Published by Elsevier Ltd.

1. Introduction feasible method for short-term mitigation of post-combustion

CO2 is the primary greenhouse gas emitted by human activities,
such as the combustion of fossil fuels for energy and transportation
[1]. Coal-red power plants are one of the major sources for CO2
emissions and, therefore, mitigation of CO2 emissions from these
plants is urgently required. Carbon capture and storage (CCS) has
been receiving increased attention because it is considered as a
Corresponding author at: School of Applied Science, Craneld University,
Craneld, Bedfordshire MK43 0AL, United Kingdom. Tel.: +44 01908 476165; fax:
+44 01234 751671.
E-mail address: b.j.anthony@craneld.ac.uk (E.J. Anthony).
CO2 from power plants and other stationary sources [24]. A typical
CO2 concentration in ue gas of such coal-red plants is 1215%,
depending on coal type and the technology used in the combustion
system [5]. Generally, the goal of most capture processes is to
achieve >90% CO2 capture from these ue gases. Achieving this
capture capacity will still leave a CO2-depleted ue gas containing
low amounts of CO2 (residual CO2), i.e. <2% CO2 that will be released
to the atmosphere. Although this residual CO2 still contributes to
the increasing anthropogenic CO2 emission, processes for its
capture have not been studied in order to further reduce the
amount of CO2 released to atmosphere.

http://dx.doi.org/10.1016/j.apenergy.2014.11.030
0306-2619/ 2014 Published by Elsevier Ltd.
298 F.N. Ridha et al. / Applied Energy 140 (2015) 297303
Currently, the commercially applied method for post-combus-
tion CO2 capture is based on amine-scrubbing processes, which
operate at low temperatures, e.g. <50 C on wide ranges of gas
sources with CO2 concentration of up to 15% [6,7]. Although this
technology has been in use for many years, there are a number
of issues that limit its application, such as the degradation of the
expensive amine solutions due to impurities in ue gas, corrosion
in the system, and high energy demand for the amine regeneration
[8,9]. Another route for capturing CO2 is swing adsorption pro-
cesses (by pressure (PSA), temperature (TSA), or vacuum (VSA))
[1013]. Typically, such systems use relatively expensive molecu-
lar sieve zeolites, activated carbons, and amine-based sorbents.
The application of these technologies for capturing CO2 from
low-CO2 sources raises questions about the economic feasibility
of such approaches [14]. Capturing CO2 from air (or the so-called
direct air capture (DAC)) is perhaps the extreme example of CO2
capture from low-CO2 sources [15]. Wet procedures for DAC have
been proposed, where CO2 is captured at ambient temperature
using sodium-based and potassium-based solutions [16].
Limestone, as a CaO-based sorbent, is an abundant material and
can be obtained in large quantities at low cost. Theoretically, 1 g of
CaO can capture 786 mg of CO2, which represents a promising
large capacity (e.g. activated carbon typically captures 80 mg of
CO2 per g of adsorbent [17]) via the following reaction:
CaOs CO2g $ CaCO3s 1
After hydration (Eq. (2), the Ca(OH)2 product upon calcination has a
higher pore volume and pore surface area than the parent material
and can be used for CO2 capture [18].
CaOs H2 Og $ CaOH2s 2 tion in the bed efuent was measured by a CO2 analyzer. Two ana-
Although carbonation and the effect of moisture in concrete
constructions have been investigated in detail [1922], to date we
have found no recent literature on using CaO-based sorbents in
xed beds as a means of capturing residual CO2 from ue gases at
ambient or near ambient temperature. The implicit assumption
was that this reaction would be too slow at such temperatures for
practical systems. In this work, a xed-bed system with CaO-based
sorbents was used for capturing CO2 from low-CO2 gas streams (<2%
CO2). CO2 inlet concentration was varied to investigate its effect on
the CO2 capture performance of the xed bed. The concentration
time relationship or breakthrough curve of CO2 in the efuent
stream was obtained, and selected systems were modeled to obtain
kinetic parameters. The mass transfer zone (MTZ) was also analyzed
to obtain a better understanding of the capture system.
2. Experimental
2.1. Sorbents
Two sorbents were used in this work: Cadomin limestone and
pellets prepared from calcined Cadomin powder using 10% calcium
aluminate cement as a binder. The pelletization was carried out in
a mechanical pelletizer (Glatt GmbH) in batch mode, where lime
and cement were mixed vigorously for a few seconds followed
by addition of water under continuous mixing. The mixing was
facilitated by two blades: an agitator at the bottom of the vessel
and a chopper on the side. The rotation speed of the blades was
controllable, thus the mixing speed was adjusted to obtain the
desired product consistency. Both natural and pelletized sorbents
were sieved to a particle size of 300850 lm. This particle size
was chosen because in preliminary experiments with smaller-
sized particles there were problems related to: (1) expansion of
sorbent particles due to the formation of Ca(OH)2 and CaCO3 which
have larger molar volumes than CaO (respectively 33.5 and 36.9
versus 16.7 cm3/mol), and (2) agglomeration inside the column
causing the operating pressure for the xed bed to increase.
2.2. Fixed-bed set-up
The experimental set-up used in this work is shown in Fig. 1. All
units were connected using 1=4 00 Teon tubes.
A stainless steel xed-bed reactor of column size (Dc) 10.8 mm
i.d. and length (L) 140 mm was used. Thus, the design parameter
of Dc/dp and L/Dc were 19 and 13, respectively, where dp is the
mean particle diameter. These design values were used to avoid wall
effects [23] and to minimize axial dispersion [24]. To prevent
elutriation of small sorbent particles, a mesh disc layered with
lter paper (with a pore size of 1.2 lm) was positioned at each
end of the column occupying a combined depth of 5 mm (thus bed
effective length was 135 mm). The mesh at the bottom of the col-
umn also supported the packing material and acted as a distributor.
2.3. Methods
The sides of the loaded column were lightly tapped to settle the
packing (loose bulk density is 10% lower than compacted bulk
density). The column was then attached to the capture system.
The capture process at ambient temperature (22 C) was started
by opening the feed tank valve and adjusting the rotameters to a
ow rate of 650 mL/min, which was maintained constant through-
out the experiment. After passing through the bubbler, the humid-
ied gas was then introduced to the bed. The capture pressure was
100 kPa gauge pressure, which was maintained constant by
adjusting the controlling valves of the rotameters. CO2 concentra-
lyzers with different detection ranges were used to ensure high
accuracy of measurements: 05000 ppm, and 03%. The measure-
ment of breakthrough was stopped when CO2 concentration in the
bed efuent was considered to be too low to reach a complete
breakthrough curve in a reasonable period of time. After recording
the breakthrough data, the system was shut down and the material
from the column was removed for analysis. The MTZ was deter-
mined by analyzing the hydration and carbonation conversions
across the length of the bed, and obtained at the end of the break-
through measurement (50% of breakthrough). CO2 inlet concentra-
tion was varied from 0.5% to 2% (N2 balance). The pre-hydration
experiment was performed by owing N2 through the bubbler to
the column at ambient temperature and a ow rate of 650 mL/min.
Samples were taken from the system as required and analyzed
for conversion in a Mettler Toledo TGA/SDTA851 thermogravimet-
ric analyzer (TGA). In this analysis, a sample of 10 mg was heated
in the TGA from ambient temperature to 900 C at a heating rate of
50 C/min under N2 ow of 40 mL/min. Carbonation and hydration
conversions were then calculated from the calcination prole.
Vent 7 2
9 6
8
4 3
5 1
Fig. 1. Schematic of the experimental set-up; (1) gas cylinder, (2) ow control
panel, (3) high-pressure rotameter, (4) pressure gauge, (5) bubbling humidier, (6)
xed bed, (7) rotameter, (8) dehumidier, and (9) CO2 analyzer.
 F.N. Ridha et al. / Applied Energy 140 (2015) 297303 299

2.4. Characterization and pellets at 25 C. It is important to note that both sorbents in

The physical properties of calcined sorbents were determined
by N2 isotherm measurement at 196 C (Micromeritics TriStar
II-3020). The calcination was carried out at 850 C for 2 h in air,
followed by heating at 250 C in N2 overnight. Pore surface area
was obtained employing the multiple-point BrunauerEmmett
Teller (BET) model, while pore volumes were calculated using the
BarrettJoynerHalenda (BJH) model [25]. Surface morphology of
selected sorbents was examined by scanning electron microscopy
(SEM, Hitachi S-3400 N) to investigate the effect of the hydration
and carbonation reactions on the surface condition of the sorbent.
3. Results and discussion
3.1. Characterization
Important physical properties of sorbents are shown in Table 1.
The results indicate that pellets have superior porosity compared
to lime. Pellets exhibited 30% higher pore surface area and 11%
higher pore volume than lime. This is consistent with previous
results [26], and is attributed mainly to the presence of Al2O3 in
the cement (70%), and the subsequent formation of mayenite
(Ca12Al14O33) after calcination at 850 C (prior to loading).
Mayenite tends to form a mesoporous-based structure at high tem-
peratures, and thereby enhances the porosity of the pellets. Also,
the pellets had a slightly lower bulk density than lime.
For CO2 capture with CaO-based sorbents, pores with a
diameter narrower than 150 nm are the dominant contributor to
the carbonation reaction; these pores have a high surface area
and reasonable pore volume that can accommodate the bulky
product, CaCO3. Pores larger than 150 nm are expected to play a
lesser role based on work on carbonation at high temperatures
for calcium looping [17,27].
The effect of sorbent properties on CO2 capture capacity can be
these tests were calcined, and therefore Ca(OH)2 was assumed to
be absent. In general, CO2 uptake increased rapidly in the rst ve
minutes, which indicates a fast reaction stage. After the relatively
fast initial reaction, a slower reaction stage controlled by diffusion
in the pores or in the product layer takes place, resulting in a
slower reaction rate [28]. This is presumably due to the fact that
the carbonation reaction in micropores is hindered by the limited
space that restricts the growth of CaCO3.
In-situ hydration of calcined sorbent is more effective and easier
to perform than ex-situ hydration followed by loading. Fig. 3
illustrates a comparison of hydration conversions between both sor-
bents as a function of hydration time. As expected, a longer
pre-hydration period results in higher conversion. At the top of the
reactor (Fig. 3a), pellets were more hydrated than lime particles
whereas the opposite was found at the bottom of the reactor
(Fig. 3b) for pre-hydration time longer than 8 h. This is attributed
to the addition of cement, which reduced CaO content available for
hydration in pellets. As the hydration conversion in Fig. 3 was calcu-
lated on the basis of Ca(OH)2 decomposition, hydration conversion
in lime was expected to exceed that of the pellets. This occurs only
when the pre-hydration duration is long enough to result in a full
conversion of CaO to Ca(OH)2 at the bottom of the reactor.
3.2. Breakthrough curves
CO2 outlet concentrationtime proles are depicted in Fig. 4.
Breakthrough curves for pre-hydrated beds were mostly
(a) 100
Lime
80 Pellets
Hydration (%)

seen in Fig. 2, which shows TGA results of CO2 uptake with lime 60

Table 1 40
Critical physical properties of sorbents.

Sorbent Surface area Pore volume Bulk densitya

BET (m2/g) BJH (cm3/g) (g/cm3) 20

Lime 12.1 0.057 0.87

Pellets 15.6 0.063 0.80
0
a
Density of calcined material. 0 1 2 8
Pre-hydration duration (h)

1.8 100
(b)
Lime
1.5
80 Pellets
CO2 capture (%)

1.2
Hydration (%)

60
0.9

0.6 40
Lime
0.3 Pellets 20

0
0 5 10 15 20 0
0 1 2 8
Time (min)
Pre-hydration duration (h)
Fig. 2. Ambient TGA results for dry-basis CO2 capture (dry sorbent and gas) with
0.5% CO2 in feed. Fig. 3. Hydration levels at both ends of the reactor; (a) top, and (b) bottom.
300 F.N. Ridha et al. / Applied Energy 140 (2015) 297303

symmetrical and relatively steep. These curves show a clear effect of 1.2
pre-hydration on the breakthrough curve: the longer the 2% CO2
pre-hydration, the longer the breakthrough time for both sorbents. 1

CO2 in bed effluent (%)

For pre-hydration of 0, 1, 2, and 8 h, CO2 breakthrough occurred after
21, 36, 53, and 660 min for lime beds and after 19, 34, 230, and 0.8
750 min for beds of pellets, respectively. These results were
obtained with a CO2 inlet concentration of 0.5%. The bed clearly
0.6
becomes more efcient after sufciently long pre-hydration
resulting in a greater hydration conversion, and enhanced capture.
With 0 and 1 h of pre-hydration, the breakthrough proles for 0.4
0.5% CO2
lime particles are steeper than those of pellets. Although it appears
that lime particles have faster overall reaction kinetics than pellets, 0.2 1% CO2
the capture process is still deemed relatively slow. For longer
pre-hydration times, i.e. between 2 h and 8 h, the pellets and lime 0
particles exhibited relatively similar breakthrough slopes. In 10 100 1000
general, the overall difference in breakthrough curve slopes Time (min)
implies that mass transfer rates are different, despite both sorbents
Fig. 5. Effect of CO2 inlet concentration on breakthrough curves of pellets pre-
possessing similar mean particle size range. With the exception of
hydrated for 8 h. Solid line represents the prediction by Eq. (4).
0 h of pre-hydration, breakthrough times with pellets were longer
than those with lime. This indicates the superiority of hydrated/
partially-hydrated pellets over hydrated/partially-hydrated lime
for CO2 capture in a xed bed. It is worth noting that all capture shown in Fig. 5. The breakthrough time decreased from 757 to 296
tests in a xed bed in this work were performed using a humid to 30 min as CO2 inlet concentration was increased from 0.5 to 1 to
gas stream regardless of the pre-hydration condition of the bed. 2%. In addition, the slopes of the breakthrough curves were similar
The presence of moisture in the gas stream is necessary to promote and less sharp for the 0.5% and 1% CO2 inlet concentrations than for
CO2 capture by CaO-based sorbents at ambient temperature the 2% CO2 inlet concentration.
[29,30]. According to Pontiga et al. [31], the mechanism of such a The short breakthrough time and rapid increase of CO2
capture process begins with solution of gaseous CO2 in the water concentration in the bed efuent for the case of 2% CO2 inlet con-
nano-droplets on the surface of the sorbent forming carbonic acid. centration is interesting. The moisture in the gas stream, necessary
This step promotes sorbent carbonation through the formation of to dissolve the large quantity of gaseous CO2 on the surface of the
movable spherules. sorbent, is limited. Thus, CO2 ows through the bed of primarily
It should be noted that the lime sorbent contains more CaO than dry-surface Ca(OH)2 and leaves the system. It can be concluded
the pellets and this explains the slightly longer breakthrough times that CO2 capture performance in a xed bed of CaO-based pellets
for lime compared to that of pellets (a delay of 2 min). The effect can be signicantly compromised by lack of sufcient moisture
on the breakthrough curves of CO2 concentration in the feed gas is in the feed stream.

3000 2500
0h 1h
2500 2000

2000
1500
1500
1000
1000
CO2 in bed effluent (ppm)

500
500

0 0
0 10 20 30 40 0 10 20 30 40 50

400 3000
2h 8h
2500
300 Samples were
2000 taken for SEM

200 1500

1000
100
500

0 0
0 50 100 150 200 250 300 0 200 400 600 800 1000
Time (min)

Fig. 4. Effect of pre-hydration on breakthrough curves of beds subjected to 0.5% CO2 in feed. (s) pellets, and (4) lime. Solid line represents the prediction by Eq. (4).
 F.N. Ridha et al. / Applied Energy 140 (2015) 297303 301

Table 2
Kinetic parameters for key systems. (a) 100
Sorbents CO2 inlet concentration (%) Rate constant, k (min1) s (min) Carbonaon
80
Lime 0.5 0.076 745

Conversion (%)
Pellets 0.5 0.045 897
Pellets 2.0 0.684 35.42
60

40

(a) 35
20 Hydraon
30

25
0
Conversion (%)

0 20 40 60 80 100 120 140

20 Hydraon
Distance from bottom, z (mm)
15

10
(b) 100
Carbonaon 80
5

Conversion (%)
0
0 20 40 60 80 100 120 140 60
Distance from bottom, z (mm)

40
(b) 100

80 20
Conversion (%)

60 0
0 20 40 60 80 100 120 140
Distance from bottom, z (mm)
40
Fig. 7. Effect of CO2 inlet concentration on bed axial conversion of pellets pre-
hydrated for 8 h; (a) 1% CO2, and (b) 2% CO2.
20

in Table 2. The lime bed exhibited a k value of 0.076 min1, which

0
is higher than that for pellet bed, i.e. 0.045 min1. This is consistent
0 20 40 60 80 100 120 140
with the fact that the slope of the breakthrough curve for the lime
Distance from bottom, z (mm)
bed is steeper than that for the pellet bed, as has been discussed
Fig. 6. Effect of pre-hydration on bed axial conversion of pellets with 0.5% CO2 inlet
previously. In addition, increasing CO2 inlet concentration to 2%
concentration; (a) 0 h, and (b) 8 h. resulted in a signicant increase in k value, i.e. 0.684 min1. These
results indicate that this rate constant, as dened in Eq. (3), is
dependent on both sorbent type and CO2 inlet concentration. How-
A simplied model [32] was applied to compare global reaction ever, previous results [32] showed that the proportionality constant
kinetics: (sk) is nearly constant and independent of CO2 concentration.
Unfortunately the limited number of data points available in this
dQ
 kQ1  Q 3 work does not allow for more detailed analysis.
dt
where dQ=dt is the change in the breakthrough concentration, Q is 3.3. Axial capture prole
the probability for sorption (dened as 1  C b =C i ), and k is the rate
constant. Integration of Eq. (1) followed by rearrangement results Determining the MTZ is helpful in understanding the progress
in: of the capture process along the axial direction of the bed (ow
1 Cb direction). Here, it is important to note, that the current system
t s ln 4 possesses a non-conventional MTZ. MTZ in the present work is
k Ci  Cb
dened as the effective bed zone for CO2 capture, that is the
where s is the time required to obtain 50% breakthrough (i.e. Cb/ portion of the bed with Ca(OH)2 pellets with a wet surface.
Ci = 0.5), Ci and Cb are CO2 concentrations in the inuent and efu- Whereas in the portion of the bed with dry CaO-based or Ca(OH)2
ent streams, respectively. k is determined from linear regression, -based pellets the carbonation rate is too slow for measurable cap-
while s is determined experimentally. Eq. (4) was applied on three ture. Fig. 6 shows the effect of hydration on the progress of the MTZ
key systems, and the model tting is shown in Fig. 4 (8 h) and Fig. 5 through the reactor for 0.5% CO2 inlet concentration. In Fig. 6a, it
(2% CO2). The model ts the experimental breakthrough data (aver- can be seen that the hydration conversion decreased gradually
age absolute relative error = 1.86%) very well. Kinetic parameters from 30.2% at the bottom of the reactor to 4.4% at the top, while
obtained from the model for beds pre-hydrated for 8 h are shown the MTZ was located in the bottom section of the reactor (from
302 F.N. Ridha et al. / Applied Energy 140 (2015) 297303

z = 0 to z = 40 mm), with carbonation conversion of 8.2% at the bot- Table 3

tom of the reactor. These conversions are fairly low, which explains CO2 capture in a xed bed of pellets.

the fast breakthrough observed (Fig. 4 and 0 h). The poor perfor- CO2pre- Figure Bed specic capture Bed total Bed average
mance of the system with 0 h of pre-hydration can be explained hydration (g CO2/g bed material) capture conversion
by the lack of surface moisture on the Ca(OH)2 pellets as CaCO3 duration (g CO2) (%)

and Ca(OH)2 formation proceeded simultaneously in the reactor. 0.5%0 h 6a 0.02 0.22 3.5
This can be clearly seen in Fig. 6b. After 8 h of pre-hydration, 0.5%8 h 6b 0.51 5.06 81.3
1%8 h 7a 0.29 2.91 46.7
pellets in the lower half of the reactor (from z = 0 to z = 60 mm) dis- 2%8 h 7b 0.07 0.67 10.8
played a maximum conversion to CaCO3 of nearly 90% with no
detectable amount of Ca(OH)2. This explains the long breakthrough
time obtained here (Fig. 4 and 8 h). The remaining portion of the
The overall CO2 capture performance can be determined by
bed (from z = 60 to z = 135 mm) contained the MTZ. The carbon-
calculating CO2 captured for the entire bed. Table 3 shows the total
ation conversion decreased gradually then plunged after
capture after calculating the area under the curves in Figs. 6 and 7
z = 100 mm.
up to the point of breakthrough. CO2 capture after the
Increasing CO2 inlet concentration resulted in a strong effect on
breakthrough point is expected to be small in these cases espe-
the MTZ, as shown in Fig. 7. In Fig. 7a, increasing CO2 concentration
cially for the lime bed, which implies that capture capacities in
to 1% widened the MTZ over the entire length of the bed. This is an
Table 3 correctly represent the performance of the bed up to the
undesirable feature for the operation of a xed bed system as an
breakthrough point. The system of 0.58 h (0.5% CO2 with 8 h
important design consideration is to minimize the length of the
pre-hydration) exhibited the best performance with a capture
MTZ for better utilization of the sorbent. Further increase in CO2
capacity 74% higher than that of 1%8 h, and 22 times and 6 times
inlet concentration to 2% shifted the MTZ to the bottom of the
higher than those of 0.5%0 h and 2%8 h, respectively. The bed
bed (from z = 0 to z = 10 mm), as shown in Fig. 7b. This is consistent
average specic capture of 0.51 g CO2/g bed obtained in the
with the short breakthrough time exhibited in Fig. 5. This also
0.5%8 h period appears superior to that of 0.1 g/g captured by
highlights the fact that the Ca(OH)2 formed after 8 h pre-hydration
amine-based adsorbents at 25 C [33]. In addition, the current sys-
is not the only parameter required for sufcient CO2 capture. Mois-
tem captured more CO2 than seen for a high-temperature TGA
ture in the gas stream is also essential to promote the mass transfer
environment where Cadomin limestone particles (1 mm) exhib-
process. It is worth noting that the carbonation conversion of 44%
ited a rst-cycle capture capacity of 0.43 g/g when exposed to
at the bottom of the reactor (Fig. 7b) is most likely attributable to
15% CO2 (N2 balance) at 650 C for 20 min [34].
the presence of a partially moisturized lter paper, which in turn
The surface morphology of pellets withdrawn from the top
wets the layer of pellets in contact with the lter paper.
(z = 135 mm) and bottom (z = 0 mm) of the reactor were found
to have a signicantly different texture, as shown in Fig. 8. In
general, these images are consistent with the results of Fig. 7b.
For instance, Fig. 8a shows that sample taken from the top
(a) appeared to be mainly composed of Ca(OH)2. On the other hand,
Fig. 8b shows the sample taken from the bottom was primarily
composed of Ca(OH)2 and CaCO3.

4. Conclusions

This work examines the use of CaO-based sorbents for ambient

temperature CO2 capture from a low-CO2 stream, e.g. <2%, using a
Ca(OH)2
xed bed reactor. CO2 capture was performed with dry sorbents
(no pre-hydration) in the TGA, and the capture trends indicated a
distinct advantage of pellets over lime. However, when the capture
process was carried out in the xed bed with 0.5% CO2 in the gases
10 m supplied to the bed, lime appeared to perform better than pellets,
represented by a steeper breakthrough curve. Breakthrough time
was also consistent, 21 min for limestone compared to 19 min for
(b) pellets. When pre-hydration was employed, the capture
performance of the bed was signicantly improved for these two
sorbents. CO2 capture efciency increased with increasing
pre-hydration time. For instance, for 8 h pre-hydration, CO2 break-
through occurred after 660 and 750 min in xed beds of lime and
pellets, respectively. In all pre-hydrated systems, pellets performed
better than lime. Increasing the CO2 concentration in the feed from
0.5% to 2% shortened the breakthrough time from 757 min to only
30 min and shifted the MTZ to the bottom of the reactor (entrance
zone), thus contributing to the rapid breakthrough of CO2. This rep-
resented a signicant reduction in capture performance, and was
attributed to insufcient moisture in the gas stream. It appears that
capturing CO2 at ambient temperature using CaO-based sorbents is
10 m a feasible process that can be deployed using relatively cheap
reagents. Although the carbonation reaction is slow at low
Fig. 8. SEM micrographs of pellets from the bed tested with 2% CO2 inlet temperatures, careful selection of operating conditions can result
concentration; (a) top, and (b) bottom. in a promising capture performance. For future tests, the effect of
 F.N. Ridha et al. / Applied Energy 140 (2015) 297303
sorbent regeneration on the performance of the system over multi-
ple cycles should be investigated.
References
[1] Bauer N, Bosetti V, Hamdi-Cherif M, Kitous A, McCollum D, Mjean A, et al. CO2
emission mitigation and fossil fuel markets: dynamic and international aspects
of climate policies. Technol Forecast Soc Change 2015;90:24356.
[2] Boot-Handford ME, Abanades JC, Anthony EJ, Blunt MJ, Brandani S, Dowell NM,
et al. Carbon capture and storage update. Energy Environ Sci 2014;7:13089.
[3] Miller BG. Clean coal engineering technology. Oxford: Elsevier; 2011 [Chapter
10].
[4] Siefert NS, Chang BY, Litster S. Exergy and economic analysis of a CaO-looping
gasier for IGFCCCS and IGCCCCS. Appl Energy 2014;128:6074.
[5] Yongping Y, Rongrong Z, Liqiang D, Kavosh M, Patchigolla K, Oakey J.
Integration and evaluation of a power plant with a CaO-based CO2 capture
system. Int J Greenh Gas Control 2010;4:60312.
[6] Giuffrida A, Bonalumi D, Lozza G. Amine-based post-combustion CO2 capture
in air-blown IGCC systems with cold and hot gas clean-up. Appl Energy
2013;110:4454.
[7] Snderby TL, Carlsen KB, Fosbl PL, Kirboe LG, von Solms N. A new pilot
absorber for CO2 capture from ue gases: measuring and modelling capture
with MEA solution. Int J Greenh Gas Control 2013;12:18192.
[8] Gouedard C, Picq D, Launay F, Carrette PL. Amine degradation in CO2 capture. I.
A review. Int J Greenh Gas Control 2012;10:24470.
[9] Pearson P, Hollenkamp AF, Meuleman E. Electrochemical investigation of
corrosion in CO2 capture plants-inuence of amines. Electrochim Acta
2013;110:5116.
[10] Dantas TLP, Luna FMT, Silva Jr IJ, Torres AEB, de Azevedo DCS, Rodrigues AE,
et al. Carbon dioxidenitrogen separation through pressure swing adsorption.
Chem Eng J 2011;172:698704.
[11] Kulkarni AR, Sholl DS. Analysis of equilibrium-based TSA processes for direct
capture of CO2 from air. Ind Eng Chem Res 2012;51:863145.
[12] Su F, Lu C, Chung AJ, Liao CH. CO2 capture with amine-loaded carbon
nanotubes via a dual-column temperature/vacuum swing adsorption. Appl
Energy 2013;113:70612.
[13] Hedin N, Andersson L, Bergstrom L, Yan J. Adsorbents for the post-combustion
capture of CO2 using rapid temperature swing or vacuum swing adsorption.
Appl Energy 2013;104:41833.
[14] Aaron D, Tsouris C. Separation of CO2 from ue gas: a review. Sep Sci Technol
2005;40:32148.
[15] Ranjan M, Herzog HJ. Feasibility of air capture. Energy Proc 2011;4:286976.
[16] Mahmoudkhani M, Keith DW. Low-energy sodium hydroxide recovery for CO2
capture from atmospheric air-thermodynamic analysis. Int J Greenh Gas
Control 2009;3:37684.
[17] Rashidi A, Yusupa S, Loong LH. Kinetic studies on carbon dioxide capture
activity using activated carbon. Chem Eng Trans 2013;35:3616.
303
[18] Blamey J, Paterson NPM, Dugwell DR, Stevenson P, Fennell PS. Reactivation of a
CaO-based sorbent for CO2 capture from stationary sources. Proc Combust Inst
2011;33:267381.
[19] Talukdar S, Banthia N. Carbonation in concrete infrastructure in the context of
global climate change: development of a service lifespan model. Constr Build
Mater 2013;40:77582.
[20] Yoon I-S, Copuroglu O, Park K-B. Effect of global climatic change on
carbonation progress of concrete. Atmos Environ 2007;41:727485.
[21] Alexander et al., editors. Concrete repair, rehabilitation and retrotting
II. London: Taylor & Francis Group; 2009. ISBN 978-0-415-46850-3.
[22] Colombo GV. Study of CO2 sorbents for extravehicular activity. NASA
CR114632. Report under contract no NAS2-6050. Moffett Field, California;
1973.
[23] ASTM D6646-03. Standard test method for determination of the accelerated
hydrogen sulde breakthrough capacity of granular and pelletized activated
carbon; 2008.
[24] Delgado JMPQ. A critical review of dispersion in packed beds. Heat Mass
Transfer 2006;42:279310.
[25] Lowell S, Thomas MA, Thommes M. Characterization of porous solids and
powders: surface area, pore size, and density. The Netherlands: Kluwer
Academic Publishers; 2004.
[26] Manovic V, Anthony EJ. Long-term behavior of CaO-based pellets supported by
calcium aluminate cements in a long series of CO2 capture cycles. Ind Eng
Chem Res 2009;48:890612.
[27] Fennell PS, Pacciani R, Dennis JS, Davidson JF, Hayhurst AN. The effects of
repeated cycles of calcination and carbonation on a variety of different
limestones, as measured in a hot uidized bed of sand. Energy Fuels
2007;21:207281.
[28] Alvarez D, Abanades JC. Determination of the critical product layer thickness in
the reaction of CaO with CO2. Ind Eng Chem Res 2005;44:560815.
[29] Shih SM, Ho CH, Song YS, Lin JP. Kinetics of the reaction of Ca(OH)2 with CO2 at
low temperature. Ind Eng Chem Res 1999;38:131622.
[30] Beruto DT, Botter R. Liquid-like H2O adsorption layers to catalyze the Ca(OH)2/
CO2 solidgas reaction and to form a non-protective solid product layer at
20 C. J Eur Ceram Soc 2000;20:497503.
[31] Pontiga F, Valverde JM, Moreno H, Duran-Olivencia FJ. Dry gassolid
carbonation in uidized beds of Ca(OH)2 and nanosilica/Ca(OH)2 at ambient
temperature and low CO2 pressure. Chem Eng J 2013;222:54652.
[32] Tsai WT, Chang CY, Ho CY, Chen LY. Simplied description of adsorption
breakthrough curves of 1,1-dchloro-1-uoroethane (HFCF-14b) on activated
carbon with temperature effect. J Colloid Interface Sci 1999;214:4558.
[33] Gray ML, Champagne KJ, Fauth D, Baltrus JP, Pennline H. Performance of
immobilized tertiary amine solid sorbents for the capture of carbon dioxide.
Int J Greenh Gas Control 2008;2:38.
[34] Ridha FR, Manovic V, Wu Y, Macchi A, Anthony EJ. Post-combustion CO2
capture by formic acid-modied CaO-based sorbents. Int J Greenh Gas Control
2013;16:218.