Journal of Environmental Sciences 20(2008) 14–27

Progress in carbon dioxide separation and capture: A review

Hongqun Yang1,∗, Zhenghe Xu1 , Maohong Fan2,* , Rajender Gupta1 ,
Rachid B Slimane3 , Alan E Bland4 , Ian Wright5

1. Department of Chemical and Materials Engineering, University of Alberta, Edmonton,

AB, T6G 2G6, Canada. E-mail: [email protected]
2. School of Materials Science and Technology, Georgia Institute of Technology, Atlanta, GA, 30332, USA
3. Gas Technology Institute, Des Plaines, IL 60018, USA
4. Western Research Institute, Laramie, Wyoming, 82072, USA
5. Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831, USA

Abstract
This article reviews the progress made in CO2 separation and capture research and engineering. Various technologies, such as
absorption, adsorption, and membrane separation, are thoroughly discussed. New concepts such as chemical-looping combustion and
hydrate-based separation are also introduced briefly. Future directions are suggested. Sequestration methods, such as forestation, ocean
fertilization and mineral carbonation techniques are also covered. Underground injection and direct ocean dump are not covered.

Key words: coal; flue gas; carbon dioxide; sorbent; absorption; adsorption; membrane; separation; sequestration

Introduction house effect as per mass of gases. There are increasing

concerns for global warming caused by the effects of GHG,
The 20th century has seen the rapid increase of popula- particularly CO2 . International Panel on Climate Change
tion and explosive growth in energy consumption. As more (IPCC) predicts that, by the year 2100, the atmosphere
countries becoming industrialized, it is expected that more may contain up to 570 ppmv CO2 , causing a rise of mean
energy will be consumed in 21st century. EIA predicts global temperature of around 1.9°C and an increase in
57 percent increase of energy demand from 2004 to 2030 mean sea level of 38 m (Stewart and Hessami, 2005). Also
(EIA, 2007). accompanied is species extinction.
Table 1 shows the comparisons of energy use, popula- The total amount of carbon on earth is constant and its
tion and per capita consumption in 1900 and 2001 (Song, distribution among lithosphere, atmosphere and biosphere
2006). In current stage over 85 percent of world energy was relatively balanced until the advent of era of industrial-
demand is supplied by fossil fuels. Fossil-fueled power ized civilization. The CO2 concentration in the atmosphere
plants are responsible for roughly 40 percent of total CO2 is increasing. Fig. 1 shows the change of atmospheric CO2
emissions, coal-fired plants being the main contributor level over the years between 1000 and 1997 and actual CO2
(Carapellucci and Milazzo, 2003). Environmental issues level during 1958–2004 (Song, 2006). CO2 level increased
due to emissions of pollutants from combustion of fossil from 280 ppmv in 1000 to 295 ppmv in 1900 based on
fuels have become global problems, including air toxics Antarctica ice core data. It increased to 315 ppmv in 1958
and greenhouse gases (GHG). The CO2 emission from and further to 377 ppmv in 2004 based on actual data
human activity was on the order of 7 Gt/a in the late logged in Hawaii.
1990’s (Yamasaki, 2003). This includes the combustion of There are three options to reduce total CO2 emission into
fossil fuels in all major industries and other factors such as the atmosphere, i.e., to reduce energy intensity, to reduce
deforestation and desertification. carbon intensity, and to enhance the sequestration of CO2 .
Through the studies of the past five decades, particularly The first option requires efficient use of energy. The second
the past 15 years, increased GHG levels in atmosphere option requires switching to using non-fossil fuels such as
is believed to cause global warming. Among these GHG, hydrogen and renewable energy. The third option involves
CO2 is the largest contributor in regard of its amount the development of technologies to capture and sequester
present in the atmosphere contributing to 60 percent more CO2 .
of global warming effects (Yamasaki, 2003), although To enhance the sequestration of CO2 , options are sought
methane and chlorofluorocarbons have much higher green- to accelerate the fixation of carbon in lithosphere and bio-
sphere, either CO2 fixation by the enhancement of natural
*Corresponding author. E-mail: [email protected]. sinking process, such as forestation, ocean fertilization and
E-mail: [email protected].
No. 1 Progress in carbon dioxide separation and capture: A review 15

Table 1 World energy use, population and per capita in 1900 and 2001 intensive step of CCS, many research have targeted at
(Song, 2006). improving the current technologies or developing new
Energy 1900 Use 2001 Use
approaches of CO2 separation and capture. This paper
source MTOEa % or Unit MTOEa % or Unit
reviews recent progress of CO2 separation and capture.
Various technologies are introduced although geological
Coal 501 55 2,395 24
Petroleum 18 2 3,913 39
carbon injection and ocean dump are not covered.
Natural gas 9 1 2,328 23
Nuclear 0 0 662 6 1 CO2 capture schemes for fossil-fueled power
Renewablesb 383 42 750 8
Total 911 100% 1,004.8 100% plants
Population 1,762 Million 6,153 Million
Per capita E use 0.517 TOEa 1.633 TOE In a global context, among all the industries emit-
Global CO2 emission 534 MMTCEa 6,607 MMTCE
Per capita CO2 emission 0.30 MTCE 1.07 MTCE
ting CO2 , fossil-fueled power plants generate the largest
Atmospheric CO2 295 ppmva 371 ppmv amount of CO2 emission and that accounts for about
Life expectancy 47.3 Years 77.2 Years 33–40 percent of the total (Carapellucci and Milazzo,
aTOE: Ton oil equivalent; MTOE: million ton of oil equivalent; MMTCE: 2003; Stewart and Hessami, 2005). CO2 needs to be
million metric ton of carbon equivalent; MTCE: metric ton of carbon separated and captured from the flue gases of such point
equivalent; ppmv: part per million by volume. b including hydroelectric sources before direct sequestration. For these power plants,
power, biomass, geothermal, solar and wind energy.
CO2 separation and capture processes can be divided
into several scenarios: post-combustion processes for a
tradition coal-fired power plant, pre-combustion processes
for gasification or reforming and oxy-fuel processes. Oxy-
fuel combustion is sometimes referred to as oxy-firing
or oxy-combustion. Newly emerged technologies, such as
chemical-looping combustion (CLC), significantly reduce
the complexity of separating CO2 from a gas stream.
1.1 Pulverized coal combustion CO2 separation and
capture process
In this post-combustion process, CO2 is separated from
a flue gas environment containing NO x and SO2 . Fig.2
shows this process flow (Feron and Hendriks, 2005).
One way to do post-combustion capture is to use
chemical absorption, such as monoethanolamine (MEA)
absorption. This technique has been widely used in natural
gas industry for over 60 years. Another advantage of this
technique is it produces a relatively pure CO2 stream.
Although the size and cost of the required absorber would
Fig. 1 Atmospheric CO2 concentrations during 1000–2004 based on the
analysis of ice cores and logged atmospheric CO2 concentrations during
be comparable to those of an SO2 scrubber, the absorber
1958–2004 (Song, 2006). would consume one-quarter to one-third of the total steam
produced by the plant, reducing its generating capacity by
mineral carbonation, or direct artificial CO2 sequestration, the same amount. Also the footprint of the host plant is
such as injection into geological formations and ocean. increased by 60% (Elwell and Grant, 2006). The National
Carbon dioxide sequestration provides a mid-term so- Energy Technology Laboratory (NETL) estimated in 2000
lution to mitigate environment impacts and allows human that this scheme would increase the cost of electricity
continue to use fossil energy until renewable energy tech- production by 70% (Elwell and Grant, 2006). An alter-
nologies mature. Since CO2 capture and sequestration is a native post-combustion CO2 capture method is the use of
relatively expensive mitigation option, it can be regarded membrane gas separation technologies.
as an insurance policy (Herzog and Drake, 1996). Riahi
et al. had a modeling study of CO2 capture and seques-
tration incorporating factors of economic, demographic,
energy demand and alternative policy (Riahi et al., 2004).
They concluded that carbon capture and sequestration
(CCS) is one of the obvious priority candidates for long-
term technology policies and enhanced R&D efforts to
protect human against the risks associated with high en-
vironmental impacts of climate changes even widespread
deployment of CCS requires decades to come. Fig. 2 Pulverized coal combustion CO2 capture (Feron and Hendriks,
Since the CO2 separation is the first and most energy 2005).
16 Hongqun Yang et al. Vol. 20

1.2 Gasification CO2 separation and capture process

In this pre-combustion process, fuels are first converted
into a mixture of CO2 and H2 through a reforming (natural
gas) or gasification (coal) process and the subsequent
shift-reaction. CO2 can be separated from the conversion
product stream and H2 can then be burned in gas turbine or
be used by fuel cell. Fig.3 shows this process flow (Feron
Fig. 4 Oxy-firing CO2 capture process (Feron and Hendriks, 2005).
and Hendriks, 2005). Gasification partially oxidizes coal
to produce a gaseous fuel, which is essentially a hydrogen
The main advantage of this technology is the elimination
and carbon monoxide mixture. When syngas is used to
of NO x control equipment and the CO2 separation step.
fuel a plant similar to a traditional combined-cycle power
The boiler size can be reduced because only oxygen is
plant, the process is referred to as integrated gasification
supplied for combustion. The size of subsequent equip-
combined-cycle (IGCC).
ments, such as SO2 scrubber, can also be reduced. The
Several methods can be used to capture CO2 . The
disadvantage is the corrosion of equipment by increased
leading option for CO2 capture is an absorption process,
SO2 concentration in the exhaust gas stream.
in which the solvent can be a chemical one, such as
This technology is not mature and its operating and
MEA absorption process, or physical one, such as pressure
maintenance and capital costs would be comparable to
swing adsorption (PSA). Physical absorption is a mature
that of post-combustion technology. The capital cost is
technology and has been used in the Great Plains Synfuels
estimated to be USD $2,040/kW (Elwell and Grant, 2006).
plant in North Dakota, U.S. for 20 years (Elwell and
Specifically, the oxygen separation plant would consume
Grant, 2006). This synfuel plant consume lignite coal and
about 23% to 37% of the total plant output and costs about
produces synthetic natural gas. Membrane technology has
the same as a chemical absorber.
been used in the FutureGen power plant (Elwell and Grant,
2006). 1.4 Chemical-looping combustion process
Pre-combustion capture is potentially less expensive
Chemical-looping combustion (CLC) is a novel process
than post-combustion capture. IGCC power plants ap-
with inherent CO2 capture. It has also been called unmixed
plying pre-combustion capture are more efficient than
combustion since direct contact between fuel and com-
pulverized coal-fired plants and would be the choice for
bustion air is avoided. Instead, an oxygen carrier brings
new plants. U.S. Department of Energy (DOE) estimated
oxygen from air to fuel. Suitable oxygen carriers are small
in 2000 that the use of pre-combustion capture would
particles of metal oxide such as Fe2 O3 , NiO, CuO or
increase the cost of electricity production by 25%. The
Mn2 O3 . A basic CLC system is shown in Fig.5 (Rydén
long-term goal is to reduce the production penalty to 10%
and Lyngfelt, 2006). The CLC has two reactors, one for air
(Elwell and Grant, 2006).
and one for fuel. The oxygen carrier circulates between the
1.3 Oxy-combustion CO2 separation and capture pro-
cess
In the oxy-firing process, pure O2 is separated from
air and sent to energy conversion unit and combines with
partially recycled flue gas of concentrated CO2 to keep
the furnace temperature below the allowable point. The
combustion takes place in an environment of O2 /CO2
mixture. The resulting flue gas is high-purity CO2 stream.
Fig.4 shows this process flow (Feron and Hendriks, 2005).
The exhaust gas stream is free of nitrogen components.
Particulates and sulfur compounds are first removed from
the exhaust stream using widely adopted techniques. After
SO2 removal, the exhaust gas stream is approximately 90%
CO2 by volume on a dry basis. Further separation of CO2
is not necessary. CO2 can then be compressed for storage Fig. 5 Chemical-looping combustion (Rydén and Lyngfelt, 2006). Me:
metal; MeO: metal oxide as oxygen carrier.
or transportation.

Fig. 3 Gasification CO2 capture process (Feron and Hendriks, 2005).

No. 1 Progress in carbon dioxide separation and capture: A review 17

reactors. In the air reactor, the carrier is oxidized by oxygen in the form of particulate organic matter either by the
according the following Reaction (1). In the fuel reactor, death of phytoplankton or after grazing. This sequestration
the metal oxide is reduced by fuel, which is oxidized to process could be enhanced by ocean fertilization. Ocean
CO2 and H2 O, according to the following Reaction (2). fertilization refers to the practice of increasing limiting
nutrients to stimulate the production of phytoplankton.
O2 + 2Me −→ 2MeO (1) Ocean fertilization has been challenged to interfere with
1 the marine ecosystem which could lead to fatal impact.
Cn Hm + (2n + m)MeO −→ nCO2
2 (2) In addition, the decomposition of sinking organic matters
1 1 could produce other stronger greenhouse gases, such as
+ mH2 O + (2n + m)Me
2 2 methane or nitrogen monoxide. This method needs to be
The amount of energy released or required in the further studied before it is put into practice.
reactors depends on these two reactions, as well as the 2.3 Photosynthesis process
temperature of reactions. Normally, Reaction (1) is strong-
ly exothermic. Reaction (2) is usually endothermic but for Most of the current CCS options offer short- to mid-term
some combinations of fuel and oxygen carrier it may be solutions with associated drawbacks. For instance, geolog-
slightly exothermic. In principle, all major fuels can be ic and oceanic injection only delays the release of CO2
utilized in a CLC. into atmosphere. Authors believe that only by switching
CLC has several advantages compared with convention- to renewable fuels can solve the CO2 emission problem.
al combustion. The exhaust gas stream from air reactor Biological carbon sequestration using technologies such as
is harmless, consisting mainly of N2 . In a well-designed controlled photosynthetic reactions may help to alleviate
system, there should be no thermal formation of NO x since GHG problems in a sustainable way.
the regeneration of oxygen carrier takes place without Stewart et al. studied a system that combines solar
flame and at moderate temperatures. The exhaust gas from energy collection and delivery through a fiber optic system
the fuel reactor consists of CO2 and H2 O. Separation to stimulate the growth of biological organisms in a
of CO2 can be done by a condenser. This is the major bio-generator to produce useful by-products from carbon
advantage with CLC which avoids the huge energy penalty dioxide (Stewart and Hessami, 2005).
necessary in traditional amine scrubbing process to capture The photo-bioreactor system makes use of the natural
CO2 . photosynthesis process to convert light, heat and car-
bon dioxide to useful products, such as carbohydrates,
hydrogen and oxygen. The type of product depends on
2 Carbon fixation
the biological strains used in the photo-bioreactor. The
following equation illustrates a photosynthetic reaction:
There are both natural and artificial ways to capture
or fix the carbon to avoid or delay emission into the 6CO2 (aq) + 6H2 O(l) + light + heat −→
atmosphere, such as forestation, ocean fertilization, photo- (3)
synthesis, mineral carbonation, and hydrate. Geological C6 H12 O6 (aq) + 6O2 (g)
injection and direct ocean dump are not covered. Cyanobacteria or micro-algae have been suggested to
2.1 Forestation perform the role of photosynthesis. In order to promote
uniform growth of the organisms, the distribution of light
It is estimated that 1.4±0.7 Gt-carbon is captured by flux in the wavelength range of 400–700 nm needs to be
terrestrial systems from atmosphere via photosynthesis delivered to the bioreactor. The efficient distribution of
(Yamasaki, 2003). Increasing this carbon flux is achieved light throughout the photo-bioreactor will affect the carbon
by forestation, reforestation of arid lands and greening of dioxide uptake rates. Despite 50 years of development
deserts. At the beginning of this sequestration, the carbon of closed-system photo-bioreactor systems, commercial
flow to terrestrial vegetation is positive and the “stock” viability has not yet been achieved. At present, open pond
of carbon is increased. When fully developed, the net systems produce around 100 t of biomass annually at a cost
capture becomes zero because the balance of capturing and of around USD $10,000/metric ton (Stewart and Hessami,
release. The potential capacity for carbon sequestration 2005).
in terrestrial systems is estimated to be 5-10 Gt-carbon
annually (Yamasaki, 2003). 2.4 Mineral carbonation, natural or biomimetic
2.2 Ocean fertilization Another major CO2 fixation process in nature is the
chemical weathering of rocks, such as silicates, containing
Ocean stores more CO2 than terrestrial vegetation. The calcium or magnesium. The silicate rocks could be turned
ocean contains about 38,000 Gt-carbon, and about 1.7±0.5 into carbonates by reacting with CO2 following this mech-
Gt is taken up annually from the atmosphere (Yamasaki, anism (Maroto-Valera et al., 2005):
2003). The production of phytoplankton at 50–100 Gt-
carbon annually is much higher than that of terrestrial (Mg,Ca) x Siy O x+2y + xCO2 → x(Mg,Ca)CO3 + ySiO2 (4)
vegetation (Yamasaki, 2003). Part of the carbon would be
released back into the atmosphere by respiration process, Mineral carbonation results in the storage of CO2 in
and the remaining would descend into the deeper ocean solid form as a stable, environmentally benign mineral
18 Hongqun Yang et al. Vol. 20

carbonate. The energy state of mineral carbonate is 60 to promising enzyme to increase the rate of precipitation of
180 kJ/mol lower than CO2 , which is 400 kJ/mol lower carbonate mineral, with industrial wastewater as a cation
than carbon (Maroto-Valera et al., 2005). Consequently, source.
sequestration in the form of a carbonate ensures long-term According to a study performed by Haywood et al.,
fixation of CO2 rather that temporary storage. Also avoided without radical alteration to the various process routes,
is the risk of any accidental release of stored CO2 leaking it seems unlikely that CO2 sequestration within a stable
from underground. mineral carbonate is a realistic proposition (Haywood et
The weathering of rocks happens in nature over geologic al., 2001). Widespread application of this technology will
time scales. Artificial ways were first proposed by Seifritz require a huge supply of metal oxides–sufficient to cause
in 1990 to emulate this natural process (Liu et al., 2005). major impacts upon iron, manganese, chromium and nickel
Two methods are possible to accelerate the reaction rates, mining industries.
i.e., to enhance the dissolution of mineral ions using
2.5 In-situ CO2 capture
stronger acids, or to react under higher CO2 pressure.
This artificial process is called CO2 mineralization, or Hughes et al. explored the in-situ CO2 capture using
mineral carbonation. Many researchers studied the min- carbonation mechanism at high temperature (–700°C ) at
eral carbonation process (Druckenmiller and Marotovaler, atmosphere pressure in a dual fluidized bed combustion
2005; Haywood et al., 2001; Liu et al., 2005; Maroto- system (Hughes et al., 2005). The following reactions take
Valera et al., 2005; Stolaroff et al., 2005). The capacity place inside the reactors:
of this fixation is huge, even though the use of acid may
cause problems in both the reaction facilities and in the CaO + CO2 FGGGB
GGG CaCO3 (5)
environment, and higher CO2 pressure consumes intensive
energy. Similar to CLC, there are two reactors in this system.
The carbonation of silicate minerals by mechanochem- Primary fuel combustion take place in the carbonator and
ical process was investigated by University of Utah and sorbent CaO reacts with CO2 in the same reactor achieving
University of Idaho (Nelson, 2004; Nelson and Prisbrey, in-situ CO2 capture. CaCO3 is heated to regenerate CaO in
2004). Natural and synthetic silicate compounds were calciner and the heat is provided by burning a secondary
ground for a range of times in the presence of gaseous CO2 . fuel such as petroleum coke. The overall thermal efficiency
Studied materials include naturally-occurring materials, has been shown to be comparable to that of a current
such as forsterite, lizardite, and wollastonite. Synthetic combustion system without CO2 capture. CANMET has
material magnesium silicate was also studied. It was shown extensively studied similar approaches (Sun et al., 2007;
that a significant change takes place in the lizardite variety Manovic and Anthony, 2007; MacKenzie et al., 2007).
of serpentine after 15 to 20 min of intense grinding
in the presence of gaseous CO2 . Analyses showed that 2.6 Hydrate-based separations
small amounts of carbon are fixed by grinding lizardite, Hydrate-based separation as a promising option is worth
forsterite, and wollastonite, and magnesium silicate, in the mentioning. This process begins by forming the hydrate
presence of gaseous CO2 . Considering the energy input by exposing the exhaust gas streaming containing CO2 to
to grind the materials to a size that can fix CO2 , this water under high pressure. As the hydrate forms, the CO2
technology should have limited use in practice. is capture. Thy hydrate is then separated and dissociated,
Maroto-Valer et al. studied mineral carbonation through thereby releasing CO2 . This technology has an energy
surface activation to promote and accelerate carbonation penalty as small as 7% (Elwell and Grant, 2006).
reaction rates and efficiencies at moderate reaction condi-
tions (Maroto-Valera et al., 2005). Their results indicate
3 Carbon separation technical options
that the surface area of the raw serpentine can be signifi-
cantly increased through physical and chemical activation
Capture of CO2 contributes 75 percent to the overall
methods. The chemical activations were more effective
CCS cost and CCS increases the electricity production cost
than physical activation at increasing the surface area.
by 50 percent (Feron and Hendriks, 2005). Although these
Sulfuric acid was found to be the most effective acid
numbers may vary with different CCS schemes, cutting
used during the chemical activation. The steam activation
the capture cost is the most important issue for the CCS
promotes the CO2 capture by serpentine. The most promis-
process to be acceptable to the energy industry. Hence,
ing results came from the carbonation of the extracted
this article mainly focuses on the progress in technologies
Mg(OH)2 solution, where the carbonation efficiency was
of CO2 separation and capture from a gas mixture from
estimated to be at least 53 percent.
the chemical conversion point of view. There are many
Liu et al. developed a biomimetic process, in which
options for CO2 separation and capture, and these include
a biological catalyst, the enzyme carbonic anhydrase,
adsorption, absorption, membrane and biotechnology. The
was used to accelerate an aqueous processing route to
optimum CO2 capture scheme could be determined by
carbonate formation (Liu et al., 2005). The enzyme was
analyzing costs in the context of power generation.
immobilized on chitosan-alginate beads. Simulated brine
The absorption/stripping process, using amine solutions
containing calcium was used to provide cations for car-
such as MEA, is a commercialized technology used in
bonate formation. Results show carbonic anhydrase is a
natural gas industry for 60 years and is regarded as the
No. 1 Progress in carbon dioxide separation and capture: A review 19

most mature process. The CO2 recovery rate is 98% for duty for solvent regeneration can constitute up to 70% of
MEA (Yamasaki, 2003). The stability of the amines and the total operating costs in a CO2 capture plant (Idem et
the energy consumption of the stripping process have large al., 2006). The MEA process disadvantages include: (1)
room of further improvement. low carbon dioxide loading capacity; (2) high equipment
The adsorption process is based on the same principle corrosion rate; (3) amine degradation by SO2 , NO2 , HCl,
but using porous solid adsorbents such as zeolites and HF, and oxygen in flue gas which induce a high absorbent
activated carbon, and chemical reactions between the makeup rate; (4) high energy consumption during high
adsorbent and CO2 may or may not occur during the temperature absorbent regeneration (Fauth et al., 2005;
separation process. Both pressure swing adsorption and Resnik et al., 2004; Yeh et al., 2005).
temperature swing processes are widely used. Besides MEA, diethanolamine (DEA) and methyl-
CO2 adsorption/absorption capacity data from various diethanolamine (MDEA) are often used as absorbents.
references are summarized in Table 2. The proposed mechanism of reactions between CO2 and
The large-scale membrane separation process has an amines are shown in Fig.6. According to this mechanism,
advantage over other processes. Membrane materials with the majority of the CO2 captured will result in the forma-
high selectivity for CO2 over oxygen or nitrogen have tion of bicarbonate in the liquid amine capture system. In
been developed, including polymeric materials and inor- aqueous media, there is a requirement of 2 mol-amine/mol-
ganic materials. Porous membranes with supporting amine CO2 for the formation of stable bicarbonate compounds
solutions could be effective for the separation of CO2 resulting in the capture of CO2 .
(Yamasaki, 2003). At present stage, a sufficiently high
selectivity with a large flux has not been realistic for
membrane separation process.
3.1 Amine absorption process
MEA absorption process is the most extensively stud-
ied CO2 absorption system. Various aspects of MEA
absorption were recently studied, such as modifying the
absorbents, improve the process design, or optimizing the
process operation (Abanades et al., 2004; Aboudheir et Fig. 6 Proposed reaction sequence for the capture of carbon dioxide by
al., 2006; Al-Juaied and Rochelle, 2006; Bello and Idem, liquid amine-based systems (Gray et al., 2005).
2006; Goff and Rochelle, 2006; Idem et al., 2006; Jassim
and Rochelle, 2006; Lawal and Idem, 2006; Ma’mun et Mixed amines have been reported to maximize the
al., 2006; Oyenekan and Rochelle, 2006; Ramachandran desirable qualities of the individual amines. The use of
et al., 2006; Rao and Rubin, 2006; Silva and Svendsen, mixed amines is to have a solution consisting of tertiary
2006; Supap et al., 2006; Tanthapanichakoon et al., 2006; and primary amines or tertiary plus secondary amines
Tobiesen and Svendsen, 2006). This paper summarizes that retains much of the reactivity of primary amines
only representative studies. or secondary amines at similar or reduced circulation
Natural gas industry uses MEA to absorb CO2 from rates but offers low regeneration costs similar to those of
natural gas. There are commercial MEA absorption pro- tertiary amines. There are reports of substantial reduction
cesses with which CO2 is removed from combustion flue in energy requirements and modest reduction in circulation
gas stream. Such processes allow the MEA solution to rates for amine blends relative to the corresponding single
be contacted with flue gas in an absorber where CO2 is amine system of similar total amine concentration (Idem
absorbed by the solution. MEA reacts with CO2 in the gas et al., 2006). Idem et al. compared the performance of
stream to form MEA carbamate. The CO2 -rich MEA solu- aqueous 5 kmol/m3 MEA with that of an aqueous 4/1
tion is then sent to a stripper where it is reheated to release molar ratio MEA/MDEA blend of 5 kmol/m3 total amine
almost pure CO2 . The MEA solution is then recycled to concentration as a function of the operating time using
the absorber (Stewart and Hessami, 2005). This process is two pilot CO2 capture plants. Their results indicate that a
generally uneconomic as it requires large equipment size huge heat duty reduction can be achieved by using a mixed
and intensive energy input. It is widely known that the heat MEA/MDEA solution instead of a single MEA solution in

Table 2 CO2 sorbent performance reported by references

Sorbent Capacity Flue gas composition References

Aqueous ammonium 1.20 g CO2 /g NH3 15 vol% CO2 , 85 vol% N2 Resnik et al., 2004;
Yeh et al., 2005
Aminated mesoporous silica 0.45–0.6 mol CO2 /mol amine 100% CO2 Knowles et al., 2005
Aminated SBA-15 1528–4188 µmol CO2 /g sorbent 10%CO2 , 90% He, with about 2% H2 O Gray et al., 2005
PEI-impregnated MCM-41 45 ml (STP) CO2 /g adsorbent 15% CO2 , 4% O2 , 81% N2 Xu et al., 2005
PEI-impregnated MCM-41 246 mg CO2 /gPEI or 82 mg CO2 /g sorbent N/A Song, 2006; Xu et al., 2002
Anthracite activated carbon 65.7 mg CO2 /g adsorbent N/A Maroto-Valer et al., 2005
Lithium silicate 360 mg CO2 /g sorbent 100% CO2 Kato et al., 2005
20 Hongqun Yang et al. Vol. 20

an industrial environment of a CO2 capture plant, although The CO2 absorption capacity of MAE (0.75 mol-
this benefit depends on whether the chemical stability of CO2 /mol-MAE) is greater than that of MEA (0.5 mol-
the solvent can be maintained. CO2 /mol-MEA). The maximum CO2 absorption capacity
of an amine is 0.5 mol-CO2 /mol-amine if the reaction
3.2 Ammonium absorption process
product is carbamate and 1.0 if the reaction product is
Yeh et al. reported the Aqua Ammonia Process which bicarbonate. This group of researchers had not identified
avoids the drawbacks of the MEA processes (Resnik et al., the secondary alkali to regenerate the primary alkali MAE.
2004; Yeh et al., 2005). This process uses aqueous ammo-
3.4 Molecular sieve adsorbent
nia as CO2 sorbent with the capability of multi-component
control. Flue gas needs to be pretreated by oxidizing SO2 Molecular sieves are a range of specially designed
and NO to SO3 and NO2 , respectively. The flue gas then sieves that separate molecules based on their molecular
reacts with aqueous ammonia in a wet scrubber. The weight or molecular size. This technology is believed to
regeneration of ammonium requires heat input to ther- be cost-effective and can be adapted to a variety of carbon
mally decompose ammonium bicarbonate and ammonium sequestration schemes (Stewart and Hessami, 2005). There
carbonate. They estimated that this process saves energy were many research activities aiming to improve the CO2
up to 60 percent compared to MEA process. The major adsorption by chemically treat the molecular sieve surface.
by-products from this process include ammonium sulfate, Adsorbents based on high surface area inorganic
ammonium nitrate, ammonium bicarbonate. Ammonium supports that incorporate basic organic groups, usually
sulfate and ammonium nitrate are well-known fertilizers. amines, are of particular interest. The interaction between
The market for ammonium bicarbonate is uncertain but the basic surface and acidic CO2 molecules is thought to
can be thermally decomposed to recycle ammonium. This result in the formation of surface ammonium carbamate
process potentially promotes the burning of cheap and under anhydrous conditions and in the form of ammonium
abundant high-sulfur coals. bicarbonate and carbonate species in the presence of water.
Refer to Fig.7 for a schematic of reactions between CO2
3.3 Dual-alkali absorption approach
and amines within an adsorbent environment. Similar to
A modified Solvay dual-alkali approach was proposed amine absorption process, the CO2 adsorption capacity is
(Huang et al., 2001). The Solvay process employs a dual- 0.5 mol CO2 /mol surface-bound amine group without the
alkali approach with ammonia (primary alkali) as a catalyst presence of water, 1.0 mol CO2 /mol surface-bound amine
to aid the reaction of CO2 with sodium chloride for the
production of sodium carbonate:

CO2 + NaCl + NH3 + H2 O −→ NaHCO–3 ↓ + NH4 Cl

(6)
Product sodium carbonate is obtained by heating sodi-
um bicarbonate. The ammonia is recovered by reacting
ammonium chloride with lime (secondary alkali), where
limestone serves as the source of lime.
2NH4 Cl + Ca(OH)2 −→ 2NH3 + CaCl2 + 2H2 O (7)
When applying this process to capture carbon dioxide
from flue gas, several drawbacks are obvious. The use
of limestone for the regeneration of ammonia renders the
process ineffective mainly because of the consumption
of limestone, production of CO2 , and intensive energy
requirement during calcination. Also for every two moles
of CO2 captured, one mole of CO2 is released from
calcination of limestone. This can be seen from the overall
reactions:
CaCO3 −→ CaO + CO2 (8)

2NaCl + 2CO2 + CaO + H2 O −→ 2NaHCO3 + CaCl2

(9)
In this study, methylaminoethanol (MAE) was chosen as
an effective primary alkali with the following reaction:
CO2 + NaCl + HOCH2 CH2 (CH3 )NH + H2 O ←→ Fig. 7 Surface reactions of amine groups with CO2 (Knowles et al.,
(10)
NaHCO−3 ↓ + HOCH2 CH2 (CH3 )NH·HCl 2005).
No. 1 Progress in carbon dioxide separation and capture: A review 21

with the presence of water. Mesoporous substrates, such

as silica (Knowles et al., 2005; Knowles et al., 2006),
SBA-1 (Yoshitake et al., 2003), SBA-15 (Gray et al.,
2005), MCM-41 (Song, 2006; Xu et al., 2005; Yoshitake
et al., 2003) and MCM-48 (Huang et al., 2003), are
attractive because they possess pores that are large enough
to be accessed by molecules with amino groups. Both
the porosity and surface functional groups facilitate the
capture of CO2 . A typical study is summarized below.
3.5 “Molecular basket” adsorbent for CO2 separation
Song (2006) and Xu et al. (2002) studied a novel CO2
“molecular basket” adsorbent. This adsorbent is based
on mesoporous molecular sieve of MCM-41 impregnated Fig. 9 “Molecular basket” concept for highly-selective high-capacity
with polyethylenimine (PEI). Fig.8 shows the structures CO2 adsorbent (Song, 2006).
of MCM41 and PEI. Fig.9 shows the concept of this
molecular basket. With the increase in PEI loading, the capacity of the activated anthracites at higher temperature,
surface area, pore size and pore volume of the loaded due to the introduction of alkaline nitrogen groups on the
MCM-41 decrease. When the loading is higher than 30 surface.
wt.%, the mesoporous pores began to be filled with PEI and 3.7 CO2 adsorbents based on lithium compounds
the adsorbent shows a synergetic effect on the adsorption
of CO2 by PEI. At PEI loading of 50 wt.%, the highest CO2 Lithium zirconate (Li2 ZrO3 ), with a favorable CO2
adsorption capacity of 246 mg/g PEI is obtained, which is sorption characteristics, has been investigated as a high
30 times higher than that of MCM-41 and is about 23 times temperature CO2 absorbent (Fauth et al., 2005). This
that of the pure PEI. Chemical impregnation proved to be a technology, based on the chemical reaction using Li2 ZrO3
better way to prepare the molecular basket. This study also to capture CO2 , is illustrated in the following reaction:
shows that this molecular basket can selectively capture Li2 ZrO3 (s) + CO2 (g) ←→ Li2 CO3 (s) + ZrO2 (s) (11)
CO2 for the separation of CO2 from simulated flue gas,
the separation of CO2 from natural gas-fired and coal-fired The reaction is reversible in the temperature range of
boiler flue gas (Song, 2006). 450–590°C. The direction can be easily reversed by a
simple temperature swing approach. The formation of
3.6 Adsorption by activated carbon
eutectic carbonate composed of Li2 CO3 and K2 CO3 can
Anthracites are known to produce high surface area ac- accelerate the CO2 absorption reaction. A number of
tivated carbon. Maroto-Valer et al. studied the CO2 capture binary and ternary eutectic salt-modified lithium zirconate
behavior of steam-activated anthracite (Maroto-Valer et sorbents were identified and evaluated for high temperature
al., 2005). They found that CO2 capture does not show a CO2 capture. The combination of binary alkali carbon-
linear relationship with the surface area. The highest CO2 ate, binary alkali/alkaline earth carbonate, ternary alkali
adsorption capacity was 65.7 mg CO2 /g adsorbent for the carbonate and ternary alkali carbonate/halide eutectic to
anthracite activated at 800°C for 2 h with a surface area of Li2 ZrO3 noticeably improved the CO2 uptake rate and CO2
540 m2 /g. The anthracite with the highest surface area of sorption capacity. Formation of a eutectic molten carbonate
1,071 m2 /g only had a CO2 adsorption capacity of 40 mg layer on the outer surface of reactant Li2 ZrO3 particles
CO2 /g adsorbent. This could be explained by certain size facilitates the transfer of gaseous CO2 during the sorption
pores being effective for CO2 adsorption. Also the NH3 process.
treatment and PEI impregnation increased the CO2 capture Lithium silicate (Li4 SiO4 ) was studied by Nakagawa
et al. for their CO2 adsorption behaviors (Essaki et al.,
2004; Kato et al., 2005). They found that the capacity
of lithium silicate is much larger than that of lithium
zirconate. Lithium silicate adsorbs CO2 below 720°C and
releases CO2 above 720°C by the following mechanism:

Li4 SiO4 + CO2 ←→ Li2 SiO3 + Li2 CO3 (12)

Desired features, such as large capacity, rapid absorp-

tion, wide range of temperature and concentrations of CO2 ,
and stability, make this compound a strong candidate for
developing commercially competitive CO2 adsorbent.
3.8 CO2 capture based on membrane separation
Membranes have been widely used in various industrial
Fig. 8 Structures of MCM-41 and PEI (Song, 2006). separations for the last two decades. Industrial applications
22 Hongqun Yang et al. Vol. 20

are currently dominated by polymeric membranes. Recent forms channels of a certain size. Large channels will allow
research directed at the development and application of faster diffusion of gases through a membrane at the cost of
inorganic membranes is advancing faster because of the less selectivity.
demand in new application fields, such as fuel cells, mem- Membranes are a low cost means of separating gases
brane reactors, and other high-temperature separations. when high purity is not vital. There are a number of issues
Inorganic or polymeric membrane separation processes associated with the capture of carbon dioxide from flue
are expected to be more efficient that conventional CO2 gas which limit the use of membranes. The concentration
separation processes (Damle and Dorchak, 2001; Xu et al., of carbon dioxide in flue gases is low, which means
2001). that large quantities of gases will need to be processed.
Both polymeric and metallic membranes can be used The high temperature of flue gases will rapidly destroy
to produce clean fuel from a syngas obtained either from a membrane, so the gases need to be cooled to below
coal gasification or reforming natural gas or methanol. 100°C prior to membrane separation. The membranes need
Syngas containing H2 is delivered at pressures and tem- to be chemically resistant to the harsh chemicals contained
peratures dependent on the type of fuel processor. Cooling within flue gases, or these chemicals need to be removed
is required for polymeric membranes. If the fuel gas is prior to the membrane separation. Additionally, creating
composed mainly of CO and H2 , various polymers can be a pressure difference across the membrane will require
used for hydrogen separation, but actual selectivities are significant amounts of power.
lower due to polymer plasticization (Carapellucci and Mi- Polymers studied in various studies include: poly-
lazzo, 2003). For inorganic membranes, palladium/silver acetylenes (Stern, 1994), polyaniline (Illing et al., 2001),
alloys are the most efficient. Their 100 percent selectivity poly (arylene ether)s (Xu et al., 2002), polyarylates (Pixton
for hydrogen guarantees a pure hydrogen permeant. The and Paul, 1995), polycarbonates (Aguilar-Vega and Paul,
separation based on metallic membranes is more efficient 1993), polyetherimides (Li and Freeman, 1997), poly
as the pressure drop across membrane is smaller. (ethylene oxide) (Lin and Freeman, 2004), polyimides
The diffusional characteristics of a particular gas (Stern et al., 1989), poly(phenylene ether) (Aguilar-Vega
through a given membrane depend on membrane physi- and Paul, 1993), poly(pyrrolone)s (Zimmerman and Koros,
cal and chemical properties, the nature of the permeant 1999) and polysulfones (Aitken et al., 1992). Fig.10 shows
species. The third factor, interaction between membrane molecular structures of some commonly used polymers.
and permeant, refers to the sorptivity or solubility of the The performance of some polymeric membranes are
gas in the membrane (Shekhawat et al., 2003). summarized in Table 3 mainly separating post-combustion
It is important to construct membranes into a practical flue gas with CO2 /N2 being the main components (Powell
configuration. Most polymeric membrane gas separation and Qiao, 2006). Single-stage membrane systems are not
studies are performed on dense homogeneous membranes. capable of high capture efficiency and CO2 can be further
Such a configuration provides the permeability with a concentrated by a second membrane stage.
closer relationship to the intrinsic permeability of the
Table 3 Performance of polymeric membranes separating CO2 /N2
polymer. However, the construction of a gas separation
(Powell and Qiao, 2006)
module in practical applications requires a more complex
membrane structure. Dense membranes tend to be sig- Material Permeance Selectivity
nificantly thicker than the selective layer of asymmetric (m3 /(m2 ·Pa·s) αCO2 /N2
membranes. This leads to dense membranes possessing Polyimide 735 43
considerably lower gas fluxes than alternative membrane Polydimethylphenylene oxide 2750 19
structures. Often used is asymmetric composite hollow Polysulfone 450 31
Polyethersulfone 665 24.7
fiber or flat sheet configuration. Poly(4-vinylpyridine)/polyetherimide 52.5 20
Polyacrylonitrile with poly 91 27.9
3.9 Polymeric membranes
(ethylene glycol)
Generally, gas molecules transport through a polymer- Poly(amide-6-b-ethylene oxide) 608 61
ic membrane by a solution-diffusion mechanism. Other
mechanisms include a molecular sieve effect and Knud- Materials for effective separation of gases can follow
sen diffusion (Powell and Qiao, 2006). These transport one of two overall strategies: increasing the rate of dif-
mechanisms are briefly introduced in inorganic membrane fusion of carbon dioxide through the polymeric structure
section. The terms permeability and selectivity are used to and increasing the solubility of carbon dioxide in the
describe the performance of a gas separation membrane. membrane. The introduction of mixed-matrix membranes
There appears to be a trade-off between selectivity and may allow superior performance which combines the ad-
permeability. Gas molecules tend to move through free vantages of polymeric and inorganic membranes materials.
volumes–the gaps between polymeric structures. Because
3.10 Inorganic membranes
of the movement of the polymer chains, a channel between
gaps can be formed allowing gas molecules to move from Inorganic membranes can be classified into two cate-
one gap to another and thus gas molecules can effectively gories based on structure: porous and dense. In porous
diffuse through the membrane structure. Selective trans- inorganic membranes, a porous thin top layer is casted on a
port of gases can be achieved by use of a polymer which porous metal or ceramic support, which provides mechan-
No. 1 Progress in carbon dioxide separation and capture: A review 23

surface diffusion, capillary condensation, and molecular

sieving (Shekhawat et al., 2003; Luebke et al., 2006).
Please refer to the references for detailed explanation.

Fig. 11 Transport mechanism through microporous membranes

(Shekhawat et al., 2003).

3.11 Carbon membranes

Carbon membranes for gas separations are typically
produced by the pyrolysis of thermosetting polymers. The
pyrolysis temperature, typically in the range of 500 to
1,000°C, depends upon the type of precursor material and
dictates the separation performance of the carbon mem-
branes (Shekhawat et al., 2003). Pyrolysis of polymeric
compounds leads to carbon material with a narrow pore
size distribution below molecular dimensions (< 1 nm),
Fig. 10 Examples of polymer molecular structures used for CO2 which makes it possible to separate gases with very similar
separation (Powell and Qiao, 2006). molecular sizes. The predominant transport mechanism
of most carbon membranes is molecular sieving. The
ical strength but offers minimum mass-transfer resistance. selection of precursor polymer, the membrane prepara-
Alumina, carbon, glass, silicon carbide, titania, zeolite, and tion method, and the carbonization process determine the
zirconia membranes are mainly used as porous inorganic performance of carbon molecular sieve membranes. The
membranes supported on different substrates, such as α- mechanical stability can be improved by supporting a thin
alumina, γ-alumina, zirconia, zeolite, or porous stainless carbon membrane on a porous support material, such as α-
steel. Surface modification by covalently bonding a layer alumina. The high thermal and chemical stability of these
of selected compounds with appropriate functional groups membranes provide hope in gas separation applications,
is one of the more convenient ways to alter membrane such as separation of CO2 in flue gas emissions from power
performance. Such modification can increase the perfor- plants.
mances by changing the mean pore size and promoting
3.12 Alumina membranes
an eventual specific interaction between the surface of
the membrane and the permeating molecules to enhance The generally mesoporous structure of alumina dictates
permeation. The dense inorganic membranes consist of that transport within such membranes by a Knudsen diffu-
a thin layer of metal, such as palladium and its alloys, sion mechanism (Shekhawat et al., 2003). With mixtures
or solid electrolytes, such as zirconia. These membranes such as CO2 /N2 , where the gases have similar mass, and
are highly selective for hydrogen or oxygen separation. CO2 /H2 , where selectivity toward the heavier component
Gas transport occurs via solution-diffusion mechanism or is required, alumina is undesirable as a material for
charged particles in dense membranes. The low permeabil- membrane separation layer. Alumina finds its use in the
ity across the dense inorganic membranes limits its wide separation of gases mainly as support, where its sound
applications as compared to porous inorganic membranes. structural properties, and chemical and hydrothermal sta-
There are four main transport mechanisms by which bilities beyond 1,000°C make it very desirable. A few
gas separation using porous inorganic membranes can be attempts have been made to modify alumina membranes
described, described by Fig.11, i.e. Knudsen diffusion, to facilitate CO2 surface diffusion with limited success. In
24 Hongqun Yang et al. Vol. 20

order to achieve high separation factors in systems like al., 2003), polyimide-silica (Kusakabe et al., 1996),
CO2 /N2 , an interaction between one of the gases in the Nafion-zirconium oxide (Apichatachutapan et al., 1996),
mixture and the membrane surface can be introduced by HSSZ-13-polyetherimide (Husain and Koros, 2007), acry-
chemical modification of separation layers. lonitrile butadiene styrene-activated carbon (Anson et al.,
2004). It is known that the permeability of a gas through a
3.13 Silica membranes
zeolite-filled polymeric membrane depends on the intrinsic
Silica is a viable starting material for the fabrication properties of the zeolite and the polymer (Tantekin-
of CO2 selective membranes, primarily because of its Ersolmaz et al., 2000). However, their performance suffers
innate stability and easy modification of its structures. from defects caused by poor contact at the molecular
Unlike alumina, which tends to undergo phase transition sieves/polymer interface. This can allow the gases to flow
at relatively low temperature, or carbon, which can exhibit non-selectively around the solid particles. Sonication and
substantial changes in pore size in oxidizing environments, decantation techniques have been suggested to overcome
silica shows exceptional thermal, chemical, and structural this issue (Vu et al., 2003). The successful implementation
stability in both oxidizing and reducing environment. of this membrane development depends on both, the se-
Sol-gel method and chemical vapor deposition (CVD) lection of polymeric matrix and inorganic zeolite, and the
technique are used to prepare silica membranes. elimination of interfacial defects (Anson et al., 2004).
A similar concept is hybrid membrane where a
3.14 Zeolite membranes
porous inorganic support material is surface-modified with
Zeolites are crystalline aluminosilicates with a uniform chemicals which have good affinity with CO2 . Strictly
pore structure and a minimum channel diameter range of speaking, the hybrid membrane might better be named
0.3 to 1.0 nm. The presence of molecular-sized cavities surface-modified inorganic membrane so that not to be
and pores make the zeolites effective as shape-selective confused with mixed-matrix membrane. This helps CO2
materials for a wide range of separation applications. This separation in two ways: porous inorganic materials al-
ability to selectively adsorb molecules by size and polarity lows large flux while the chemical provides selectivity
is the key to the unusual efficiency of synthetic zeolites as (Luebke et al., 2006; Shekhawat et al., 2003). The
the basis for gas separation. Separation occurs in zeolite studied chemical-inorganic pairs include polyether-silica
membranes by both molecular sieving and surface diffu- (Kim et al., 2005), trichlorosilane-γ-alumina (Luebke
sion mechanisms. et al., 2006), organosilane-Vycor glass (Singh et al.,
Zeolite membranes are usually prepared by in-situ hy- 2004), tetrapropylammonium-silica (Yang et al., 2002),
drothermal synthesis on porous stainless steel, α-alumina, titania-trimethoxysilaen (Abidi et al., 2006), γ-alumina-
or γ-alumina support tubes or disks for the gas permeation trimethoxysilane (Abidi et al., 2006), hexagonal meso-
studies. Membranes of various zeolites, such as ZSM-5 porous silica-aminopropylhydroxysilyl (Knowles et al.,
(Kusakabe et al., 1996), Y type (Kusakabe et al., 1997), 2005; Chaffee, 2005).
silicalite (Lin et al., 2000), A type (Aoki et al., 2000),
3.16 Facilitated transport membranes
P type (Dong and Lin, 1998), modernite (Bernal et al.,
2000), and silicoaluminophosphate (Poshusta et al., 2000) Facilitated transport membranes (FTM) have received
have been synthesized on porous supports. Typically the a lot of attention in gas separations because they offer
heat of adsorption of gases on most zeolites increase in higher selectivity and larger flux (Shekhawat et al., 2003).
this order: H2 < CH4 < N2 < CO2 (Poshusta et al., 2000). Higher selectivity in FTM is achieved by incorporating
This order is consistent with the electrostatic properties of a carrier agent into membrane, which reacts reversibly
each molecule. Carbon dioxide preferentially permeates in with the penetrating species. In addition to the solution-
CO2 /N2 , CO2 /CH4 , and CO2 /H2 mixtures at low tempera- diffusion mechanism of the polymeric membranes, FTMs
tures, because CO2 adsorbs more strongly on zeolites than also involve a reversible complexing reaction. The per-
the other gases. For the CO2 /N2 and CO2 /CH4 mixtures, meating species dissolves in the up stream portion of
CO2 is smaller in size and thus permeates faster at elevated the membrane and reacts with the carrier agent inside
temperature. However, the selectivities are the same for the the membrane to form a complex. This complex diffuses
CO2 /N2 or CO2 /CH4 systems at elevated temperatures due across the membrane, and then releases the permeants on
to the absence of competitive adsorption. the downstream side of the membrane, while the carrier
agent is simultaneously recovered and diffuses back to the
3.15 Mixed-matrix and hybrid membranes feed side. FTMs can be in a form of fixed carrier mem-
These molecular sieves possess superior gas trans- branes, solvent-swollen polymer membrane, and mobile
port properties, but have significant problems with their carrier membranes (Shekhawat et al., 2003).
processibility. Incorporation of molecular sieves with-
in a polymer membrane possibly provides both the 4 Conclusions
processibility of polymers and selectivity of molecular
sieves (Nomura et al., 1997). Many research direct at Fossil fuels continue to be the primary energy source,
mixed-matrix membranes; the polymer-zeolite pairs in- at least for this century. There are many technical options
clude polydimethylsiloxane-silicalite (Tantekin-Ersolmaz for separation and/or capture of CO2 from combustion flue
et al., 2000), polyimide-carbon molecular sieve (Vu et gas and other industrial effluents. Other than geological
No. 1 Progress in carbon dioxide separation and capture: A review 25

injecting and direct ocean dump, forestation and ocean flue gases. Industrial and Engineering Chemistry Research,
fertilization can be utilized to avoid or delay the carbon 45(8): 2569-2579.
emission into atmosphere. Mineral carbonation should Bernal M P, Piera E, Coronas J, Menendez M, Santamaria J, 2000.
have limited applications. Amine absorption process is Mordenite and ZSM-5 hydrophilic tubular membranes for
mature but has large room of improvement. Adsorption the separation of gas phase mixtures. Catalysis Today, 56:
221–227.
provides another option although it has many problems
Carapellucci R, Milazzo A, 2003. Membrane systems for CO2
need to be addressed. Membrane separation processes
capture and their integration with gas turbine plants. Pro-
provide several advantages over other conventional sepa- ceedings of the Institution of Mechanical Engineers Part A:
ration techniques; a membrane combining high flux, high Journal of Power and Energy, 217: 505–517.
selectivity and high stability is not realistic at this stage but Chaffee A L, 2005. Molecular modeling of HMS hybrid materials
mixed-matrix membranes provide hopes. for CO2 adsorption. Fuel Processing Technology, 86(14-
Although emerging technologies, such as zero-emission 15): 1473–1486.
power plant, significantly reduce the complexity of sep- Damle A S, Dorchak T P, 2001. Recovery of carbon dioxide
arating CO2 , many other industrial processes still emit in advanced fossil energy conversion processes using a
substantial amounts of carbon dioxide which needs to membrane reactor. Journal of Energy and Environment
be separated; CO2 separation and capture technologies Research, 1(1): 77–89.
Dong J, Lin Y S, 1998. In situ synthesis of P-type zeolite-
will continue to evolve. Membrane process as energy-
membranes on porous α-alumina supports. Industrial and
saving, space-saving, easy to scale-up, could be the future
Engineering Chemistry Research, 37: 2404–2409.
technology for CO2 separation. Druckenmiller M L, Maroto-Valer M M, 2005. Carbon sequestra-
tion using brine of adjusted pH to form mineral carbonates.
References Fuel Processing Technology, 86: 1599–1614.
EIA, 2007. International Energy Outlook 2007. US DOE.
Abanades J C, Rubin E S, Anthony E J, 2004. Sorbent cost http://www.eia.doe.gov/oiaf/ieo/index.html, accessed No-
and performance in CO2 capture systems. Industrial and vember 2007.
Engineering Chemistry Research, 43(13): 3462–3466. Elwell L C, Grant W S, 2006. Technology options for capturing
Abidi N, Sivadea A, Bourret D, Larbot A, Boutevin B, Guida- CO2 -Special Reports. Power 150(8).
Pietrasanta F, Ratsimihety A, 2006. Surface modification of Essaki K, Nakagawa K, Kato M, Uemoto H, 2004. CO2 ab-
mesoporous membranes by fluoro-silane coupling reagent sorption by lithium silicate at room temperature. Journal
for CO2 separation. Journal of Membrane Science, 270: of Chemical Engineering of Japan, 37(6): 772–777.
101–107. Fauth D J, Frommell E A, Hoffman J S, Reasbeck R P, Pennline
Aboudheir A, Tontiwachwuthikul P, Idem R, 2006. Rigorous H W, 2005. Eutectic salt promoted lithium zirconate: Novel
model for predicting the behavior of CO2 absorption into high temperature sorbent for CO2 capture. Fuel Processing
AMP in packed-bed absorption columns. Industrial and Technology, 86(14-15): 1503–1521.
Engineering Chemistry Research, 45(8): 2553-2557. Feron P H M, Hendriks C A, 2005. CO2 capture process
Aguilar-Vega M, Paul D R, 1993. Gas transport properties of principles and costs. Oil and Gas Science and Technology-
polycarbonates and polysulfones with aromatic substitu- Rev, IFP 60(3): 451–459.
tions on the bisphenol connector group. Journal of Polymer Goff G S, Rochelle G T, 2006. Oxidation inhibitors for copper and
Science Part B: Polymer Physics, 31(11): 1599–1610. iron catalyzed degradation of monoethanolamine in CO2
Aguilar-Vega M, Paul D R, 1993. Gas transport properties of capture processes. Industrial and Engineering Chemistry
polyphenylene ethers. Journal of Polymer Science Part B: Research, 45(8): 2513–2521.
Polymer Physics. 31(11): 1577–1589. Gray M L, Soong Y, Champagne K J, Pennline H, Baltrus J P,
Aitken C L, Koros W J, Paul D R, 1992. Effect of structural Stevens R W, Khatri R Jr, Chuang S S C, Filburn T, 2005.
symmetry on gas transport properties of polysulfones. Improved immobilized carbon dioxide capture sorbents.
Macromolecules, 25(13): 3424–3434. Fuel Processing Technology, 86(14-15): 1449–1455.
Al-Juaied M, Rochelle G T, 2006. Absorption of CO2 in aque- Haywood H M, Eyre J M, Scholes H, 2001. Carbon fioxide
ous diglycolamine. Industrial and Engineering Chemistry dequestrtion as dtable varbonate minerals–Environmental
Research, 45(8): 2473–2482. barriers. Environmental Geology, 41: 11–16.
Anson M, Marchese J, Garis E, Ochoa N, Pagliero C, 2004. ABS Herzog H J, Drake E M, 1996. Carbon dioxide recovery and
copolymer-activated carbon mixed matrix membranes for disposal from large energy systems. Annual Review of
CO2 /CH4 separation. Journal of Membrane Science, 243: Energy and the Environment, 21: 145–166.
19–28. Huang H P, Shi Y, Li W, Chang S G, 2001. Dual alkali approaches
Aoki K, Kusakabe K, Morooka S, 2000. Separation of gases with for the capture and separation of CO2 . Energy and Fuels,
an A-type zeolite membrane. Industrial and Engineering 15(2): 263–268.
Chemistry Research, 39: 2245–2251. Huang H Y, Yang R T, Chinn D, Munson C L, 2003. Amine-
Apichatachutapan W, Moore R B, Mauritz K A, 1996. Asym- grafted MCM-48 and silica xerogel as superior sorbents
metric nafion/(zirconium oxide) hybrid membranes via in for acidic gas removal from natural gas. Industrial and
ditu dol-gel vhemistry. Journal of Applied Polymer Science, Engineering Chemistry Research, 42(12): 2427–2433.
62(2): 417–426. Hughes R W, Lu D Y, Anthony E J, Macchi A, 2005. Design,
Bello A, Idem R O, 2006. Comprehensive study of the ki- process simulation and construction of an atmospheric dual
netics of the oxidative degradation of CO2 loaded and fluidized bed combustion system for in situ CO2 capture
concentrated aqueous monoethanolamine (MEA) with and using high-temperature sorbents. Fuel Processing Technol-
without sodium metavanadate during CO2 absorption from ogy, 86(14-15), 1523–1531.
26 Hongqun Yang et al. Vol. 20

Husain S, Koros W J, 2007. Mixed matrix hollow giber Liu N, Bond G M, Abel A, McPherson B J, Stringer J, 2005.
membranes made with modified HSSZ-13 zeolite in Biomimetic sequestration of CO2 in carbonate form: Role
polyetherimide polymer matrix for gas separation. Journal of produced waters and other brines. Fuel Processing
of Membrane Science, 288: 195–207. Technology, 86: 1615–1625.
Idem R, Wilson M, Tontiwachwuthikul P, Chakma A, Veawab Luebke D, Myers C, Pennline H, 2006. Hybrid membranes for
A, Aroonwilas A, Gelowitz D, 2006. Pilot plant studies selective carbon dioxide separation from fuel gas. Energy
of the CO2 capture performance of aqueous MEA and and Fuels, 20(5): 1906–1913.
mixed MEA/MDEA solvents at the university of regina MacKenzie A, Granatstein D L, Anthony E J, Abanades J C,
CO2 capture technology development plant and the bound- 2007. Economics of CO2 capture using the calcium cycle
ary dam CO2 capture demonstration plant. Industrial and with a pressurized fluidized bed combustor. Energy and
Engineering Chemistry Research, 45(8): 2414–2420. Fuel, 21: 920–926.
Illing G, Hellgardt K, Wakeman R J, Jungbauer A, 2001. Prepa- Ma’mun S, Jakobsen J P, Svendsen H F, Juliussen O,
ration and characterisation of polyaniline based membranes 2006. Experimental and modeling study of the solubil-
for gas separation. Journal of Membrane Science, 184: 69– ity of carbon dioxide in aqueous 30 mass % 2-((2-
78. aminoethyl)amino)ethanol solution. Industrial and Engi-
Jassim M S, Rochelle G T, 2006. Innovative absorber/stripper neering Chemistry Research, 45(8): 2505–2512.
configurations for CO2 capture by aqueous Manovic V, Anthony E J, 2007. Steam reactivation of spent
monoethanolamine. Industrial and Engineering Chemistry CaO-based sorbent for multiple CO2 capture cycles. Envi-
Research, 45(8): 2465–2472. ronmental Science and Technology, 41: 1420–1425.
Kato M, Nakagawa K, Essaki K, Maezawa Y, Takeda S, Kogo Maroto-Valer M M, Fauth D J, Kuchta M E, Zhang Y, Andresen
R, Hagiwara Y, 2005. Novel CO2 absorbents using lithium- J M, 2005. Activation of magnesium rich minerals as
containing oxide. International Journal of Applied Ceramic carbonation feedstock materials for CO2 sequestration. Fuel
Technology, 2(6): 467–475. Processing Technology, 86: 1627–1645.
Kim H, Lim C, Hong S, 2005. Gas permeation Maroto-Valer M M, Tang Z, Zhang Y, 2005. CO2 capture by
properties of organic-inorganic hybrid membranes activated and impregnated anthracites. Fuel Processing
prepared from hydroxyl-terminated polyether and 3- Technology, 86(14-15): 1487–1502.
isocyanatopropyltriethoxysilane. Journal of Sol-Gel Nelson M G, 2004. Carbon dioxide sequestration by
Science and Technology, 36(2): 213–221. mechanochemical carbonation of mineral silicates-Final
Knowles G P, Delaney S W, Chaffee A L, 2006. Report. USDOE.
Diethylenetriamine[propyl(silyl)]-functionalized (DT) Nelson M G, Prisbrey K A, 2004. Mechanochemical methods for
mesoporous silicas as CO2 adsorbents. Industrial and CO2 sequestration. SME annual meeting, Feb 23-25 2004,
Engineering Chemistry Research, 45(8): 2626–2633. Denver, CO, United States. Preprints, 923–925.
Knowles G P, Graham J V, Delaney S W, Chaffee A L, 2005. Nomura M, Yamaguchi T, Nakao S I, 1997. Silicalite membranes
Aminopropyl-functionalized mesoporous silicas as CO2 modified by counterdiffusion CVD technique. Industrial
adsorbents. Fuel Processing Technology, 86: 1435–1448. and Engineering Chemistry Research, 36: 4217–4223.
Kusakabe K, Ichiki K, Hayashi J, Maeda H, Morooka S, Oyenekan B A, Rochelle G T, 2006. Energy performance of
1996. Preparation and characterization of silica-polyimide stripper configurations for CO2 capture by aqueous amines.
composite membranes coated on porous tubes for CO2 Industrial and Engineering Chemistry Research, 45(8):
separation. Journal of Membrane Science, 115: 65–75. 2457–2464.
Kusakabe K, Kuroda T, Murata A, Morooka S, 1997. Formation Pixton M R, Paul D R, 1995. Gas transport properties of
of a Y-type zeolite membrane on a porous a-alumina tube polyarylates. Part I: Connector and pendant group effects.
for gas separation. Industrial and Engineering Chemistry Journal of Polymer Science Part B: Polymer Physics, 33(7):
Research, 36: 649–655. 1135–1149.
Kusakabe K, Yoneshige S, Murata A, Morooka S, 1996. Pixton M R, Paul D R, 1995. Gas transport properties of polyary-
Morphology and gas permeance of ZSM-5 type xeolite lates. Part II: Tetrabromination of the bisphenol. Journal
membrane gormed on a porous a-alumina dupport tube. of Polymer Science Part B: Polymer Physics, 33(9): 1353–
Journal of Membrane Science, 116: 39–46. 1364.
Lawal A O, Idem R O, 2006. Kinetics of the oxidative degrada- Poshusta J C, Tuan V A, Pape E A, Noble R D, Falconer J
tion of CO2 loaded and concentrated aqueous MEA-MDEA L, 2000. Separation of light gas mixtures using SAPO-34
blends during CO2 absorption from flue gas streams. membranes. AIChE Journal, 46(4): 779–789.
Industrial and Engineering Chemistry Research, 45(8): Powell C E, Qiao G G, 2006. Polymeric CO2 /N2 gas separation
2601–2607. membranes for the capture of carbon dioxide from power
Li Y, Ding M, Xu J, 1997. Gas separation properties plant flue gases. Journal of Membrane Science, 279: 1-49.
of aromatic polyetherimides from 1,4-bis(3,4- Ramachandran N, Aboudheir A, Idem R, Tontiwachwuthikul
dicarboxyphenoxy)benzene dianhydride and P, 2006. Kinetics of the absorption of CO2 into mixed
3,5-diaminobenzic acid or its esters. Journal of Applied aqueous loaded solutions of monoethanolamine and
Polymer Science, 63(1): 1–7. methyldiethanolamine. Industrial and Engineering Chem-
Lin H, Freeman B D, 2004. Gas solubility, diffusivity and istry Research, 45(8): 2608–2616.
permeability in poly(ethylene oxide). Journal of Membrane Rao A B, Rubin E S, 2006. Identifying cost-effective CO2 control
Science, 239: 105–117. levels for amine-based CO2 capture systems. Industrial and
Lin X, Kita H, Okamoto K I, 2000. Silicalite membrane Engineering Chemistry Research, 45(8): 2421–2429.
preparation, characterization, and separation performance. Resnik K P, Yeh J T, Pennline H W, 2004. Aqua ammonia
Industrial and Engineering Chemistry Research, 40: 4069– process for simultaneous removal of CO2 , SO2 and NO x .
4078. International Journal of Environmental Technology and
No. 1 Progress in carbon dioxide separation and capture: A review 27

Management, 4(1-2): 89–104. matrix membranes. Journal of Membrane Science, 175:

Riahi K, Rubin E S, Schrattenholzer L, 2004. Prospects for 285–288.
carbon capture and sequestration technologies assuming Tanthapanichakoon W, Veawab A, McGarvey B, 2006. Electro-
their technological learning. Energy, 29: 1309–1318. chemical investigation on the effect of heat-stable salts on
Rydén M, Lyngfelt A, 2006. Using steam reforming to produce corrosion in CO2 capture plants using aqueous solution
hydrogen with carbon dioxide capture by chemical-looping of MEA. Industrial and Engineering Chemistry Research,
combustion. International Journal of Hydrogen Energy, 31: 45(8): 2586-2593.
1271–1283. Tobiesen F A, Svendsen H F, 2006. Study of a modified
Shekhawat D, Luebke D R, Pennline H W, 2003. A review of car- amine-based regeneration unit. Industrial and Engineering
bon dioxide selective membranes–A topical teport. National Chemistry Research, 45(8): 2489–2496.
Energy Technology Laboratory, United States Department Vu D Q, Koros W J, Miller S J, 2003. Effect of condensable
of Energy. impurity in CO2 /CH4 gas feeds on performance of mixed
Silva E F, Svendsen H F, 2006. Study of the carbamate stability matrix membranes using carbon molecular sieves. Journal
of amines using ab initio methods and free-energy pertur- of Membrane Science, 221: 233–239.
bations. Industrial and Engineering Chemistry Research, Vu D Q, Koros W J, Miller S J, 2003. Mixed matrix membranes
45(8): 2497–2504. using carbon molecular sieves: II. Modeling permeation
Singh R P, Way J D, McCarley K C, 2004. Development of a behavior. Journal of Membrane Science, 211: 335–348.
model surface flow membrane by modification of porous Xu X, Song C S, Andresen J M, Miller B G, Scaroni A W, 2002.
vycor glass with a fluorosilane. Industrial and Engineering Novel polyethyleneimine-modified mesoporous molecular
Chemistry Research, 43(12): 3033–3040. sieve Of MCM-41 type as adsorbent for CO2 capture.
Song C, 2006. Global challenges and strategies for control, Energy and Fuels, 16: 1463–1469.
conversion and utilization of CO2 for sustainable develop- Xu X, Song C, Miller B G, Scaroni A W, 2005. Adsorption
ment involving energy, catalysis, adsorption and chemical separation of carbon dioxide from flue gas of natural gas-
processing. Catalysis Today, 115: 2–32. fired boiler by a novel nanoporous “molecular basket”
Stern S A, 1994. Polymers for gas separations: The next decade. adsorbent. Fuel Processing Technology, 86(14-15); 1457–
Journal of Membrane Science, 94: 1–65. 1472.
Stern S A, Mi Y, Yamamoto H, Clair A K, 1989. Struc- Xu Z K, Dannenberg C, Springer J, Banerjee S, Maier G, 2002.
ture/permeability relationships of polyimide membranes. Novel poly(arylene ether) as membranes for gas separation.
Applications to the separation of gas mixtures. Journal of Journal of Membrane Science, 205: 23–31.
Polymer Science Part B: Polymer Physics, 27(9): 1887– Xu Z, Wang J, Chen W, Xu Y, 2001. Separation and fixation of
1909. carbon dioxide using polymeric membrane. In Proceedings
Stewart C, Hessami M, 2005. A study of methods of carbon of 1st National Conference on Carbon Sequestration, Wash-
dioxide capture and sequestration–the Sustainability of ington, DC, 2001.
a photosynthetic bioreactor approach. Energy Conversion Yamasaki A, 2003. An overview of CO2 mitigation options for
and Management, 46: 403–420. global warming – Emphasizing CO2 sequestration options.
Stolaroff J K, Lowry G V, Keith D W, 2005. Using CaO- and Journal of Chemical Engineering of Japan, 36(4): 361–375.
MgO-rich industrial waste streams for carbon sequestration. Yang S M, Lee Y E, Hyun S H, Lee C H, 2002. Organic-
Energy Conversion and Management, 46: 687–699. templating approach to synthesis of nanoporous dilica
Sun P, Grace J R, Lim C J, Anthony E J, 2007. Removal of CO2 vomposite membranes (I): TPA-templating and CO2 depa-
by Ca-based sorbents in the presence of SO2 . Energy and ration. Journal of Materials Science, 37(12): 2519–2525.
Fuels, 21: 163–170. Yeh J T, Resnik K P, Rygle K, Pennline H W, 2005. Semi-
Supap T, Idem R, Tontiwachwuthikul P, Saiwan C, 2006. Anal- batch absorption and regeneration studies for CO2 capture
ysis of monoethanolamine and its oxidative degradation by aqueous ammonia. Fuel Processing Technology, 86(14-
products during CO2 absorption from flue gases: A compar- 15): 1533–1546.
ative study of GC-MS, HPLC-RID, and CE-DAD analytical Yoshitake H, Yokoi T, Tatsumi T, 2003. Adsorption of chromate
techniques and possible optimum combinations. Industrial and arsenate by amino-functionalized MCM-41 and SBA-1.
and Engineering Chemistry Research, 45(8): 2437–2451. Chemistry of Materials, 15: 4536–4538.
Tantekin-Ersolmaz S B, Atalay-Oral Ç, Tatlier M, Erdem- Zimmerman C M, Koros W J, 1999. Entropic selectivity analysis
Senatalar A, Schoemanb B, Sterte J, 2000. Effect of zeolite of a series of polypyrrolones for gas separation membranes.
particle size on the performance of polymer-zeolite mixed Macromolecules, 32(10): 3341–3346.