Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

Carbon capture by absorption – Path covered and ahead MARK

a,⁎ a b c
I. Sreedhar , Tanisha Nahar , A. Venugopal , B. Srinivas
a
Department of Chemical Engineering, BITS Pilani Hyderabad Campus, Hyderabad 500078, India
b
I & PC Division, Indian Institute of Chemical Technology, Hyderabad 500007, India
c
Department of Chemical Engineering, GVP College of Engineering, Visakhapatnam, India

A R T I C L E I N F O A BS T RAC T

Keywords: Global warming and associated climate change has resulted in serious efforts towards reducing greenhouse gas
Absorption emissions, primarily carbon dioxide through carbon capture. There are various technical options in pre- and
Carbon capture post-combustion modes available viz., adsorption, absorption, membrane separation, chemical looping
Solvent combustion with and without oxygen uncoupling and cryogenic separations. Among all these, absorption
Contactors
technology which could be deployed as a post-combustion option to be integrated with power plant, has been
Modeling
commercialized with amines as solvents long back. But there is a long way to go to improve this process in terms
Kinetics
Thermodynamics of economic viability due to large regeneration costs involved and to make it more environmental friendly with
minimum toxic and corrosion problems due to solvents. In this review, absorption based carbon capture has
been reviewed comprehensively and critically in terms of various aspects like solvents and their synthesis
protocols, performance analysis of different solvent systems, contactors, kinetics and thermodynamics,
modeling and simulation studies etc. Every section has been discussed in terms of the trends and developments
observed including the contemporary status besides citing future challenges and prospects to improve the
technology.

1. Introduction preferred at relatively higher pressures and concentrations of CO2. Post

The impact of global warming due to greenhouse gas emissions,
primarily carbon dioxide is a serious concern all over the world. Efforts
are being made continuously to limit the degree of future climate
changes due to the increase in greenhouse gas emissions. The rise in
temperature would eventually lead to the rise in sea level, if not
controlled, increase the probability of occurrence of storms and floods.
The International Panel of Climate Change (IPCC) estimates that by the
year 2100, the CO2 content in the atmosphere shall reach 570 ppmv,
the sea level will increase by 3.8 m and the global mean temperature
will rise by 2 ºC with major consequences on the environment [1–4].
With the continuous rise in world population, demand for energy was
expected to shoot up. Burning of fossil fuels (coal, petroleum and
natural gas) continues to cater to the energy demand which accounts
for massive CO2 emissions into the atmosphere, thus leading to
significant increase in global warming. In order to reduce this emis-
sions significantly, carbon capture (CC) has been identified as the key
solution and hence under the focus of scientists, technologists and
environmentalists worldwide [5].
The three major modes employed in CC were pre-combustion,
oxyfuel combustion (OFC) and post combustion. Pre-combustion was
combustion options were expensive but could be used for low
concentrations of CO2 and could be integrated with the existing power
plants for effective CC. A comparative study between pre- and post-
combustion technologies was done in integration with a coal combus-
tion plant [6]. To achieve CC efficiency of 90%, L/G ratio was found as
2.55 mol solvent/mol syngas and 1.13 mol solvent/mol syngas,
solvent regeneration heat was found to be 2.18 GJ/tCO2 and 2.8 GJ/
tCO2 for pre- and post-combustion respectively. Oxy fuel combustion
involves blowing pure oxygen over the fuel for complete combustion
and producing relatively pure steam and CO2. Out of the above three
modes, post combustion carbon capture was widely used. The different
technologies of post-combustion CC were absorption (chemical and
physical), adsorption, cryogenic separation, membrane separation and
chemical looping combustion.
In adsorption, liquid or gas adheres onto solid adsorbents which
could later be regenerated by the application of temperature, vacuum
or pressure swing operations. Activated carbon, metal oxides, alumina,
phosphates and zeolites [7–11], metal organic frameworks [12–15],
micro-porous carbons and polymers [16–20] and modified amines
[21–23] were some of the different adsorbents reported to be used in
this process. Membrane separation of carbon dioxide from flue gases

⁎
Corresponding author.
E-mail address: isreedhar2001@yahoo.co.in (I. Sreedhar).

http://dx.doi.org/10.1016/j.rser.2017.03.109
Received 10 November 2016; Received in revised form 13 March 2017; Accepted 23 March 2017
1364-0321/ © 2017 Elsevier Ltd. All rights reserved.
I. Sreedhar et al.
depends on the difference in selectivity and permeability. Some of the
membranes like organics consisting of polymers and inorganics con-
sisting of zeolite, ceramic, metals, carbon etc. were used for the
separation. It also acts as a contacting unit for the solvent and the
gases which move in counter- current fashion and CO2 gets absorbed
selectively on the matrix of the membrane [8,24,25]. Oxyfuel combus-
tion (OFC) is the latest technology which has not yet been commercia-
lized on large scale. Pure oxygen ensures minimal production of
undesirable gases like NOx, resulting in the increased selectivity for
CO2. Cryogenic separation involves compression and cooling of the gas
mixture with CO2 at various levels to advance phase change of CO2 along
with other constituents
which could later be separated by distillation [26].
This process wouldn’t require chemical absorbents; could be carried
out at atmospheric pressure and was compatible with both the pre- and
post-combustion modes of CC. The major drawbacks of this technology
include the drop of heat transfer rate due to the built-up of thick CO2
layers, plugging caused due to the solidification of CO2 and ice
formation [25,27,28]. In CLC, the oxygen required for combustion
was obtained from metal oxides to avoid direct contact between fuel
and air. No additional separation step was required in this process and
formation of NOx was also minimum, leading to enhanced CC
efficiency. But it is yet to be commercialized to the industry level
[26]. Combining oxy-combustion with post-combustion CC could also
potentially reduce the overall energy consumption by 25% when
compared to an oxy-combustion process [29].
Absorption CC, the main focus of this paper is a technical option
that could be used in both pre- and post-combustion modes of CC and
this technology has been commercialized too. Absorption could be both
physical and chemical. Physical absorption depends on solubility of
CO2 in the solvent and was preferred at higher pressures [30]. Solvents
like rectisol, selexol etc. have been reported to be used as physical
absorbents [27]. However the concentration of CO2 in flue gases from
coal fired power plants was reported to be upto 15% [31–34] at low
pressure conditions, hence, chemical absorption was preferred owing
to higher absorption capacity at low CO2 partial pressure [35,36]. A
typical chemical absorption system consists of three components i.e.
solvent, absorber and stripper. Flue gases from different CO2 sources
like coal power plants come under contact with lean solution in the
counter-current manner in the absorber. CO2 was then absorbed by the
solvents, leading to a lower CO2 concentration in the flue gas. The
solvent rich CO2 was then regenerated in the stripper. The regenerated
lean solution was returned back to the absorber and compressed CO2
was collected at the top of the stripper and transported. Chemical
absorption has been the most successful technology that has been
commercialized from many years but it is yet to be scaled up in CO2
recovery in power plants. It could be easily fitted with the existing
power plants. However, less CO2 loading capacity, elevated equipment
size and corrosion rate, energy penalties during absorbent regeneration
were few major drawbacks of chemical absorption which could be
resolved by improvement in absorbents and operations involved in the
process [37]. Use of aqueous alkanolamines was the most commercially
successful and widely practiced technique in chemical absorption based
CC [38]. Monoethanolamine (MEA) was one of the earliest alkanola-
mines used for CC. It has a high reaction rate, good absorption capacity
and was cheap. However some of the major drawbacks were high
energy penalties for regeneration of solvent, degradation in oxidizing
environment, and corrosive effects [7–11,39–43]. Other alkanolamines
such as diethanolamine (DEA) have also been used for absorption but
they suffered from similar limitations. Researchers have also proposed
the use of hybrid solvents like methanol with DEA or MEA [44,45] and
N- methyl-2-pyrrolidone (NMP) with diethylene glycol (DEG) and
MEA [46]. These could perform better than the single solvents due to
their enhanced diffusivity and solubility. Combination of MEA and
DEA has been reported to give some success in improving carbon
loading capacity, withstanding corrosion and reduction in energy
penalties during regeneration [11,43,47–51]. The use of sterically-
1081
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
hindered amines like 2-amino-2-methyl-1propanol (AMP) were also
reported in literature because of good absorption capacity, degradation
and corrosion resistance, lower energy penalties and higher selectivity
[43,52–55]. Combinations of AMP with DEA and DEA with MDEA
were also reported [56,57].
Piperazine (PZ)/cyclic diamines were used because of good absorp-
tion capacity, higher resistance towards corrosion and degradation and
lower regeneration costs [26]. Dash et al. [58] reported that 90% CC
could be achieved using 22 wt% AMP and 8 wt% PZ [58–62]. Use of
aqueous potassium carbonate (K2CO3) promoted with inorganic salts
like arsenates, borates, silicates and vandates; organics like amines and
their derivatives, alkaline amino acids like arginine and biological
enzymes like carbonic anhydrase were also described as effective
absorbents due to reduced enthalpy, degradation and cost require-
ments [63–72]. K2CO3 with MgO as support was reported to attain a
99.4% efficiency of CO2 capture [73]. Ionic liquids integrated with
functional groups like amino acids exhibited good CO2 absorption
ability, lower energy penalties for regeneration, higher selectivity and
higher thermal and chemical stabilities [74–76]. The use of ammonia
as solvent for CC has been reported to have higher rate of reaction with
lower regeneration energy than traditional MEA solvent [77–79].
To enhance CC with minimal energy penalty, extensive research has
been done and reported for the development of efficient gas–liquid
contactor configurations, solvent systems and stripper configurations
[26]. The central idea was to maximize the surface area and mass
transfer for the processes of absorption and desorption. Different
absorber configurations like packed bed (PB), bubble column, spray
column, rotating packed bed (RPB) and tray towers were employed.
Some of the retrofit options like use of multiple columns, vapor
recompression and heat integration in stripping section, use of split
flows and matrix stripping were reported to improve the efficiency [80–
87].
In this review article, the importance of carbon capture and the
various technical options available in achieving the same so as to
mitigate a very grave concern of global warming have been initially
discussed in the preceding the introduction section. In the sections that
follow, the developmental trends on following aspects related to
absorption based carbon capture were discussed thoroughly besides
citing future challenges and prospects in each of them. Every section
has been summarized by a table that captures the overall findings.
• Solvents-classification, synthesis protocols and the comparison of
their performance in CC.
• Promoters employed along with solvents to enhance their perfor-
mance.
• Contactors employed in absorption based CC-type, scale, critical
design parameters, merits and demerits of each of them.
• Modeling studies done to analyze, validate and predict the technol-
ogy under more rigorous conditions. Various models were discussed
in terms of their governing equations, assumptions, performance,
advantages and limitations.
• Thermodynamic and kinetic aspects of the process employing
various solvent systems were presented and discussed as these were
very vital in contactor design and scale-up.
2. Absorbents and their performance
2.1. Absorbents and their classification
An absorbent employed in the CC process could be classified as two
types: (a) chemical where chemical reactions occur to separate carbon
dioxide and (b) physical where separation occurs by physical methods
with mass transfer occurring at the gas-liquid interface. It depends on
the solubility of gases, surface area of the interface and operating
conditions like temperature, pressure. Fig. 1 highlights the classes of
absorbents used for CC. The chemical or reactive absorbents involve
I. Sreedhar et al. Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

ammonia solution and treated under hydrothermal conditions [93].

Fig. 1. Types of absorbents employed in CC.
amines, blends, ionic liquids, aqueous solvents, ammonia etc. The
blends were so formulated to combine the positive characteristics of
different absorbents and simultaneously suppressing their negative
characteristics. The physical absorbents include solvents like Rectisol,
Selexol etc. To be employed as an absorbent, a solvent should have the
desired properties viz., high reactivity and absorptivity with CO2, high
stability under extensive thermal and continuous chemical exposure,
low vapor pressure, easy regenerability, low environmental impact and
economical to use.
2.2. Synthesis protocols
The blends of different absorbents could be made using various
synthesis protocols. The Oil/Water emulsion was prepared by mixing
lipophilic amine with water and an emulsifier [88]. Ionic liquid with
non-ionic surfactants were prepared by putting them together into a
magnetic stirrer to ensure that they were adequately dissolved [89].
Immobilization technique of nano-sized calcium oxide was done by
preparing nano-particulate calcium carbonate (PCC) in ethanol with
ultra-sonication. Resulting suspension was dropped onto the dried
ceramic fabric (yttria or aluminium) using a transfer pipette. The
impregnated fabric was then dried and sintered [90–92]. For poly-
ethylene polyamine (PEI) anchored onto the surface of Fe3O°4
nanoparticles, the following steps were undertaken. Fe3O4 particles
were prepared by co-precipitatingco-precipitating Fe2+ and Fe3+ ions in
PEI was then dissolved in dry DMSO by stirring, with the addition of
propyl isocyanate. The silane modified PEI solution and Fe3O4
nanoparticles were mixed in toluene dried by NaH. The Fe3O4 grafted
with PEI was isolated by centrifugation, washed with ethanol to replace
DMSO and toluene. Finally, to remove unreacted PEI, mixture was
washed with de-ionized water. The powder obtained was dried before
use [94]. PZ and phosphates were added as additives into tertiary or
hindered amines like aqueous N-methylmonoethanolamine (MMEA) to
form a complex absorbent [95]. Reagent grade amines were obtained
from commercial sources. Anhydrous PZ was melted with distilled and
de-ionized water and the required amine [96].
CaO was used as an absorbent for CC in Natural Gas Combined
Cycle (NGCC) power plants [97]. Micron sized CaO powder was
calcined at 1173 K in an oven for 3 h in static air. To enable extrusion
at mild pressures, the remaining amount of water was fixed. The paste
was extruded by a plunger extruder. The products were then cut into
equal pellets which were then calcined again in an oven for 3 h at
1173 K.
Spherical nanoparticles of SiO2 and Al2O3 were mixed in methanol
and employed as absorbents [97]. Powdered nanoparticles were added
to methanol and dispersed for 2 h using an ultrasonicator. During the
absorption-regeneration cycles, to maintain the stability of these nano-
absorbents, they were constantly dispersed by processes like hydro-
dynamics, inflow and release of bubbles of gas and convention.
Zn-cyclen was a synthetically created enzyme which was used as an
absorbent. The aza-macrocyclic scaffold was treated with a perchlorate
salt of zinc and the mixture was heated to 60 oC. 5 ml cyclen in ethanol
was added to 10 ml of Zn(ClO4)2 in a 1:1 molar ratio and kept for a
period of over 1.5 h. The product was recovered by vacuum filtration
and then washed with ethanol. MS and H1 NMR 400 were used to
verify the completion [98]. Table 1 gives an overview of various
synthesis protocols of absorbents employed in CC.
2.3. Comparison of their performance
Amines like monoethanolamine (MEA) and diethanolamine (DEA)
were the earliest and the most regularly employed absorbents because
of their low cost, high reactivity and high rate of absorption. But they

Table 1
Overview of synthesis protocols of different absorbent systems in CC.

Solvent Raw materials Synthesis procedure Operating Remarks Refs.

conditions

O/W emulsion Lipophilic amine, water, -Mixing NA -ionic liquid and non-ionic surfactants [88,89]
emulsifier -Separation separated by magnetic stirrer
o
Immobilization of CaO Calcium carbonate, ethanol, -Suspension by mechanical Drying at 100 C, -ultrasonication to improve dispersion, [90–93]
ceramic fabric stirring Sintering at 800 oC reduce agglomerates
(yttria/aluminium) -Ultrasonication
-Drop into fabric
-Drying and sintering
PEI on Fe3O4 Fe+2, Fe+3, DMSO, propyl -Precipitation of ions NA -NH3 helps precipitation of ions [93,94]
isocyanate, toluene, ethanol -Dissolution of PEI in DMSO -Ethanol replaces DMSO and toluene
-Mixing and drying -Deionized water removes unreacted PEI
-Centrifugation -powdered form
-Washing by ethanol
-Drying
PZ+MMEA PZ, distilled and deionized water, -Anhydrous PZ melted NA NA [95,96]
MMEA -CO2 sparges gravimetrically
CaO pellets CaO powder -calcination 1173 K -used in NGCC power plant [97]
-extusion -extruder has 3 mm die
-cutting -
-calcination
SiO2/Al2O3 in methanol SiO2 (Sigma-Aldrich), Al2O3 (Alfa- -Ultrasonication 0 oC -SiO2– 15 nm [97]
Aesar), methanol -Al2O3– 45 nm
-circulation ultrasonicator used
Zn cyclen enzyme Cyclen, ethanol, Zn(ClO4)2 -Mixing and heating 60 oC -raw materials soluble in ethanol, final [98]
-Vacuum filtration product is not

1082
I. Sreedhar et al.
suffer from various drawbacks like degradation when in oxidative
atmosphere, intensive regeneration energy requirement, limited car-
bon dioxide loading capacity and corrosive nature with foaming and
fouling characteristics [7–11,39–43]. A study on optimizing CC using
MEA as a solvent was based on three parameters-solution circulation
rates, CO2 residual mole fraction of regenerated solution and lean
solution temperature [99]. It was reported that the costs could be
lowered with lower solution circulation rates, higher CO2 residual mole
fractions and higher lean solution temperature. Thus, a general view of
the utility and energy requirements warrants the lowest possible costs.
Methyldiethanolamine (MDEA), a hybrid of MEA and DEA has been
used with some success because of its improved carbon dioxide loading
capacity, higher corrosion and degradation resistance and reduced
regeneration costs but at lower absorption rates [11,43,47–51].
Diisopropanolamine (DIPA) was used because of its high corrosion
resistance and lower regeneration costs but it had similar drawbacks
[9,11,40]. 2-diethylaminoethanol (DEAE) [100], which could be pre-
pared from renewable sources [101,102], has good chemical stability,
high cyclic capacity, high loading capacity yet low heat of reaction
[103]. It was reported to be a good substitute to the traditional tertiary
alkanolamines. The kinetics [104] and factors affecting CO2 capacity
like basicity [105], temperature [106], density and viscosity [107] were
studied. Weiland equation [108] was employed to derive correlations
for the viscosities of CO2 loaded and unloaded solutions. Xu et al.
[109–111] employed 1,4-butanediamine (BDA) as a promoter to DEAE
and found that it accelerated the rate of reaction and also reduced the
sensible heat requirements. Azhgan et al. [112] used 1,5-diamino-2-
methylpentane (DAMP) as absorbent in CC. The equilibrium CO2
loading of DAMP, MDEA and MEA were 0.86, 0.43 and 0.54
respectively at 303 K. DAMP was less volatile than MEA and hence,
resulted in lower solvent loss in the CO2 removal process. Several
amine-based absorbents like 2-amino-2-ethyl-1,3-propanediol
(AEPD), triethylenetriamine (TETA), tetraethylenepentamine (TEPA),
2-amino-2-hydroxymethyl-1, 3-propanediol (THAM) were reported to
decrease regeneration energy consumption significantly. Results
showed that more hydroxyl groups, shorter length of carbon chain
bonded at the active nitrogen atom and the existence of methyl groups
improved the CO2 stripping rate [113]. Various new tertiary amines
like 4-(dimethylamino)-2-butanol (DMAB), 4-(dipropylamino)-2-buta-
nol (DPAB), 4-(dibutylamino)-2-butanol (DBAB), 4-((2-hydroxyethyl)
(methyl)amino)-2-butanol (HEMAB) and 4-((2-hydroxyethyl)(ethyl)
amino)-2-butanol (HEEAB) were tried and reported as potential
absorbents for CC [114]. Out of these, DMAB, HEEBAB, HEMAB were
reported to have high absorption capacities (0.88, 0.68, 0.44 mol CO2/
mol amine), high absorption rates (0.082, 0.142, 0.111 mol CO2/min).
While keeping the heats of absorption low, these also had fast
regeneration rates, proving to be promising solvents for blends in
CC. 2-ethylaminoethanol (2EAE) was reported to have very fast
reaction kinetics, good CO2 absorption and low heat of absorption
when compared to several tertiary amines namely MDEA, 1-dimethy-
lamino-2-propanol (1DMA2P), 2-dimethylaminoethanol (2DMAE), 3-
dimethylamino-1-propanol (3DMA1P), N,N,N,N-tetramethyl-1,3-pro-
panediamine (TMPDA) [115]. Xiao et al. [116] described certain
general trends for CC performance of tertiary amines. For tertiary
amines, the ethyl group led to high equilibrium CO2 solubility ;
presence of a hydroxyl group or a heterocyclic structure reduced the
equilibrium CO2 solubility; side carbon chain enhanced amine activity;
presence of a hydroxyl group at the γ position improved CO2 absorp-
tion and hydroxyl group increased the heat of absorption.
Sterically hindered
amines have a hundred-fold higher rate of absorp-
tion than tertiary amines [117]. Sherman et al. [118] studied the use of
a stearically hindered amine - 2-piperidineethanol (2PE) and analyzed
its performance. 2PE was thermally and oxidatively robust [119,120],
exhibit fast kinetics, low eco-toxicity, easy bio-degradability but was
highly viscous [121] and carcinogenic. 2-amino-2-methyl-1-propanol (AMP)
was tried for CC. Though it had higher absorption capacity, higher
1083
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
selectivity, higher corrosion and degradation resistance, easy regener-
ability; it exhibited low absorption rate [43,52–55]. Aqueous mixtures
of AMP with DEA and DEA with MDEA were also reported [56,57].
Several new absorbents have been identified by blending amines
like primary-tertiary amines (MEA-MDEA), secondary-tertiary amines
(DEA-MDEA). These blends combine the high reaction rate of primary
or secondary amines with the high equilibrium capacity, lower enthalpy
of tertiary amines. Gao et al. [122] compared 30 wt% MEA-methanol
blend with 30 wt% MEA in a complete process of absorption desorption
in a pilot plant. Results reported that the former had a faster rate of
CO2 absorption and lower regeneration energy requirement than the
latter (MEA-Methanol – 3.22 MJ/kg CO2, MEA – 4.04 MJ/kg CO2).
Shamiri et al. [123] examined the solubility of CO2 in 30 wt% MEA and
glycerol (0–20 wt%) over temperatures ranging from 313 to 333 K and
partial pressures of CO2 varying from 500 to 1500 kPa. It was observed
that at low pressures and low glycerol concentrations, the CO2
absorption capacity of glycerol+MEA was greater than MEA over all
temperature ranges. The same was not true for higher pressures ( >
1000 kPa). Lower concentrations of glycerol ( < 10 wt%) at lower
pressures ( < 1000 kPa) increased solubility whereas higher concentra-
tions of glycerol (15 and 20 wt%) led to a decrease in solubility of CO2.
The Gibbs-Helmholtz expression [124] was used to evaluate the
isosteric heats of absorption in MEA-Glycerol blends (5 wt% glycerol,
30 wt% MEA). CO2 loadings were found to be dependent on enthalpy
for this particular solvent, thus, indicating a different absorption
mechanism. Ramazani et al. [125] developed a new blend potassium
lysinate (PL) with MEA and determined the equilibrium solubility
using VLE for temperatures from 303 to 323 K and partial pressure of
CO2 from 0 to 50 kPa. The new blend gave a higher absorption capacity
than MEA.
Cyclic amine Piperazine (PZ) was used as a promoter due to its
rapid formation of carbamates with carbon dioxide. PZ and phosphates
were added as additives in tertiary or hindered amines like aqueous N-
methylmonoethanolamine (MMEA) to form a complex absorbent. It
was selected because of its low cost, easy availability, less corrosive
nature due to weaker alkalinity as compared to MEA. It exhibited high
absorptivity upto 0.49 mol/L and low solution circulation rates. Lower
circulation rate resulted in reduction in pumping energy costs and
therefore, a reduction in regeneration energy. The loading capacities
were in the order of (MMEA+PZ) > (MMEA+K3PO4) > (MMEA
+K2HPO4) > MMEA > MEA > (MMEA+KH2PO4) [95]. A study by Du
et al. [96] evaluated the pKa values of these blends. At a higher pKa
(10.5 for tropine), amines were entirely protonated and behave like
strong alkalis like NaOH. At a low pKa, amines don’t protonate and the
effect was similar to PZ. For the amine to buffer the solvent at both the
rich and lean conditions, an optimum pKa value was desired. It was
found to be 9.1 which was achieved by blending mono-tertiary amine
with PZ. The absorption rate of CO2 in these PZ/tertiary amine blends
was observed to be lower than PZ alone. CO2 absorption rate was
studied in a blended solution of PZ and DEAE [126]. At T=303 K,
WDEAE=0.23, on changing WPZ from 0 to 0.026, the absorption rate
increased from 1.62 to 4.67 (kmol m-2 s-1). For DEAE-PZ aqueous
solution, the viscosity of the solution increased with increase in CO2
loading and showed an exponential decrease with increase in tempera-
ture [100]. Shuangchen et al. [127] investigated the effect of blended
PZ+Ni (II) on CC in a bubbling reactor. CO2 efficiency was found to be
72% when 2 wt% NH3 was mixed with 0.025 mol/L PZ and 0.05 mol/L
Ni(II), which was greater than what was achieved by 3 wt% NH3. The
reaction between PZ and CO2 occurred according to the zwitterion
mechanism. Using XRD and UV visible spectrophotometer, a reaction
mechanism was proposed. In 4 wt% NH3, solution with additives, PZ
coexisted with [Ni(NH3)x]2+ and NiCl2 existed in complex forms. The
complex traps free NH3 and hence reduces escape of NH3. XRD results
indicated that the precipitate formed was Ni(OH)2. CC performance by
a PZ derivative 1-(2-aminoethyl)piperazine (AEP) which contained all
primary, secondary and tertiary amine groups, was explored by Choi
I. Sreedhar et al.
et al. [128]. At 333 K, AEP was found to display 2.2 times greater CO2
loading than MEA. Its apparent rate of absorption was same as DEA.
Inspite of all types of functional amine groups present reacting in the
order: primary > secondary > tertiary, AEP did not form a bicarbamate
but AEP-1-carbamate and AEP-7-carbamates were formed. Not much
information is available about the physical properties, nitrosamine
formation, mass transfer, oxidative degradation of AEP.
Blend of AMP and PZ has been promising as it led to a 20% decline
in regeneration energy and 45% reduction in the circulation rate of
absorbent, apart from significant increase in thermal and oxidative
degradation resistances. Dash et al. [58], reported in their simulation
studies that 90% CC was achieved by using a 22 wt% AMP and 8 wt%
PZ solution. Khan et al. [129] gradually increased the concentration of
Piperazine in the AMP-PZ blend to study its performance in a packed
column. With temperatures ranging from 298 to 308 K, CO2 partial
pressure ranging from 10 to 15 kPa and PZ composition ranging from
2 wt% to 10 wt%), 99.63% of CO2 absorption was achieved at 10 wt%
PZ. Regeneration process was carried out at 363–383 K and at vacuum
conditions of 50–55 kPa. It was noted that as the concentration of PZ
in the blend increased, the regeneration efficiency reduced. Maximum
regeneration efficiency of 98.93% was achieved by 2 wt% PZ. AMP+PZ
blends (with 2, 5, 8, 10 wt% PZ) showed better CO2 loading capacities
than 30 wt% MEA solutions. The reboiler duty for the least PZ
concentration blend was found to be the lowest. Shariff et al. [130]
conducted CC from CO2-rich Natural Gas using a 7 wt% PZ +23 wt%
AMP blend in a packed absorption column at high pressures of 0.1–
5.05 MPa. It was reported that the performance was enhanced at
higher pressures because of higher driving forces for mass transfer. A
liquid flow rate of 4.33 m3/m2 h exhibited the best performance.
Increase in L/G increases the driving force, but an optimum L/G ratio
should be obtained for cost optimization in circulation and reboiler
heat duties. There is large scope for optimization of amine concentra-
tion, as higher the viscosity, the higher would be the solvent cost and
corrosion of the equipment. Investigation showed that blend of DEA
and PZ increase the CC efficiency by more than 20% compared to DEA.
Based on the CC study [131] of DEA, MEA, EDA, PZ as absorbents, it
was reported that velocity and CO2 content of gas showed a negative
impact on decarburization, whereas concentration and velocity of
absorbent had an opposite effect. The CC efficiency order was found
to be DEA < MEA < EDA < PZ.
A blend of DETA-sulfolane-H2O was tried as an absorbent [132] in
CC. Diethylenetriamine (DETA) has two primary and one secondary
amino groups. In comparison to MEA, DETA exhibited higher reactiv-
ity [133,134], lower heat of reaction even with better CO2 absorbing
capacity [135], higher rate of mass transfer [136,137] and lower heat
duty for regeneration [138]. Sulfolane, an aprotic polar solvent, acts as
a physical additive into the chemical absorption process, because of its
high stability and resistance to corrosion. It had been previously used
in blends as MDEA-sulfolane-H2O and AMP-sulfolane-H2O [132] in
which phase-split was observed by Luo et al. [132] at high CO2
loadings. CO2 and DETA appeared in the upper phase, sulfolane
remained in the lower phase. This was due to limited solubility of
ionic products of CO2 with DETA than in sulfolane. A blend of 20%
DETA, 40% sulfolane, 40% H2O was tested in a biphasic cyclic process.
The CO2 loading was found to be 35% higher in the blend than in 30%
MEA solutions. However, after absorption, only 60% of the blend with
the composition of 30% DETA, 10% sulfolane, 60% H2O could be
regenerated.
Nwaoha et al. [139] studied tri-solvent blends of AMP-PZ-MEA as
potential absorbents in CC. The overall concentration of AMP+PZ was
constant at 3 kmol/m3 to inhibit precipitation, while the total solvent
concentration was 6 kmol/m3. The tri-solvent blend reported high
cyclic capacities, low heat duties when compared to 5 kmol/m3 of MEA
but higher initial desorption rate 93–125% higher than MEA). Saiwan
et al. [140] reported a tri-solvent blend of AMP-MDEA-DETA. whose
overall concentration was kept constant at 4.5 kmol/m3. All of them
1084
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
were reported to be superior to 5 kmol/m3 of MEA in terms of
equilibrium CO2 loadings and absorption capacities, initial desorption
rates, cyclic loadings and capacities and regeneration energies. For the
first 60 min, only blends with AMP/MDEA ratio of 1 and 2 had a higher
initial absorption rate than MEA. It also had lower corrosion potential
and lower costs considering the lower circulation rates than MEA.
Quad-solvent blends have also been studied as an alternative, though
very scarce data was available. An aqueous quad-solvent blend of MEA-
TETA-AMPD-PZEA was studied [141]. When compared to a 5 kmol/
m3 MEA solution, it had a 25% lower regeneration energy and 27%
lower liquid/gas flow rate. An experimental analysis has been carried
out on a blend of AMP-PZ-MEA-DETA [142], with the total concen-
tration kept constant at 5 kmol/m3. Compared to similar concentration
MEA solutions, they had a higher CO2 absorption capacity (37–43%)
and higher equilibrium loading (39.8–45.7%). There was no precipita-
tion on cooling the blend. The performance of quad-solvents could
further be enhanced by blending two bicarbonate forming (sterically
hindered or tertiary monoamines) and two promoters (primary or
secondary amines) [143].
Aqueous ammonia, owing to its high carbon capture efficiency, easy
availability, low regeneration costs, high absorption capacity, ability to
remove SO2, NOx along with CO2 simultaneously and the value added
salts that could be used as fertilizers [8,62,144–148]; has been
employed as an absorbent in a pilot scale project [67,68]. However,
ammonia slip was the main drawback associated with this technology.
In order to overcome the problem, Alstom et al. [149] conducted the
process under chilled conditions on a laboratory scale, which was later
patented by Eli gal et al. [150]. Giuffrida et al. [151] studied an
advanced air-blown IGCC plant integrated with a post-combustion CC
based on chilled and cooled aqueous ammonia scrubbing. Ammonia
slip was controlled by acid wash from the H2S recovered from the
desulphurization unit of IGCC. The by-product thus obtained was
ammonium sulphate which could be used as a fertilizer. In comparison
to the case with no CC, IGCC efficiency increased from 6.3 to 7.2 and
from 7.6 to 9.1 percentage points in cooled (15–20 oC) and chilled
(7 oC) mode respectively. In the former case, specific primary energy
consumption for CO2 was as low as 2.27 MJ/kgCO2, independent of the
amount of CC in the range of 80–90%. Hence, the integrated plant had
a better effect when operated to cooled mode. However, the results
were reported to be specific to only this case where high Sulphur coal
was fired in IGCC plants. To inhibit the evaporative ammonia losses in
the CC process, the use of metal ion additives (M(II) – Cu (II), Zn (II),
Ni (II)) has been reported [152]. The complex formation of ammonia
with the metal ions reduced the free ammonia concentration leading to
less evaporative losses but with a proportional reduction in the CO2
absorption rate. It also leads to a reduction in the cyclic capacity and
increase in the regeneration energy. It was concluded that if the
precipitated solid or slurry formed during the CC process could be
dealt with, metal ions would suppress ammonia loss and lead to
production of insoluble side products (CuCO3, ZnCO3, NiCO3). If it
couldn’t be dealt with, then the reduction of ammonia would be
recommended for addressing the loss of ammonia issue.
Sharifzadeh et al. [153] studied the efficiency of an amine promoted
buffer salt (APBS) – CDRMax developed by Carbon Clean Solutions
Limited (CCSL) for CC. In comparison to MEA, CDRMax required 25%
less steam, 58% less column packing, 61% less cooling water and 31%
less pumping energy per unit CO2 captured. CDRMax was reported to
have a dual benefit of faster kinetics of amines, thus a lower equipment
size, and a lower regeneration energy of buffer, thus lower operating
costs. It also had a higher CO2 loading capacity which reduced the
circulation rate significantly.
The effect of addition of enzyme carbonic anhydrase (CA) had been
studied on MEA, MDEA, AMP, K2CO3 [154]. The results with and
without the enzyme were benchmarked against a standard 30 wt%
MEA solution across various temperatures. It was reported that there
was a drastic increase in the liquid side mass transfer coefficients of
I. Sreedhar et al. Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

MDEA and K2CO3. The effect of AMP was moderate whereas it was
barely visible on MEA. It was thus concluded that the addition of CA
only benefited systems which could form bicarbonates. An increase in
temperature led to a reduction in liquid side mass transfer coefficients
of DEA and K2CO3, but an increase for AMP.
Aqueous potassium carbonate (K2CO3) was also attempted as an
absorbent because of its low cost, low enthalpy requirements, low
toxicity, low solvent losses and high degradation resistance. To
counterbalance the low rate of mass transfer, promoters like inorganic
salts of arsenites, vanadates, borates and silicates; biological enzymes
like carbonic anhydrase; organics like amines and their derivatives and
alkaline amino acids like arginine have been reported to be employed
[63–72]. An efficiency of 99.4% CO2 has been exhibited by potassium
carbonate supported on MgO [73]. Extensive research has been done
on using perchloric acid (HClO4) [155,156], bromine (Br2) and
hypobromous acid (HBrO) [157], telluric acid (Te(OH)6), MEA, DEA,
PZ, amino acids, arsenious acids, boric acid, vanadates, carbonic
anhydrase [157] as promoters. Shrier and Danckwerts [158] reported
ethylaminoethanol (EAE) as the most efficient promoter. EAE could be
easily synthesized, had a high thermal and oxidative resistance. Saeed
et al. [98] presented the use of Zn-cyclen (Zinc complex containing the
cyclic ligand – 1,4,7,10-tetracyclodode) as a promoter to K2CO3. The
promoter resembles the natural enzyme CA but was its improved
version. It was more stable, has a lower molecular weight and much
longer lifetimes. The kinetics increased by ten folds on using this mimic
enzyme. A thermally stable CA promoter (NZCA) was tested through a
stopped flow method in a wetted wall column [159]. The catalysis
activation energy was found to be 51 kJ/mol at 298–328 K. The
corrected first order rate coefficient increased ten folds on addition of
5.2 µM NZCA to 30 wt% K2CO3 solution. Even after continuously
running the solution for 8 h at a pH of 10.6–10.8 at 323 K, NZCA
was found to be stable retaining 70% of its catalysis efficiency.
Though these promoters have been used extensively in industrial
plants, they had various drawbacks like degradation, corrosion, eva-
poration. Table 2 documents the performance of various promoters
[157]. Table 3 describes the merits and demerits of using various
promoters.
Ionic liquids have been employed because of their low vapor
pressure, moderate viscosity and high thermal stability. They were
effective in absorbing CO2 with a CO2/H2 selectivity of over 100 [160].
Their regeneration was simple and absorption was high owing to
flexible structures in the desired reversible reaction with CO2 [161].
Imidazolium based ionic liquids were employed due to their unique
properties viz., non-explosive nature and non-inflammability, higher
reaction velocity and selectivity, negligible vapor pressure, higher
chemical and thermal stability, less expensive, high efficiency and
compatibility with membrane technology [26,162–164]. Imidazole
ionic liquids contain anions such as BF4-, PF6-, Tf2N-. The absorptivity
of various anions with the same cation and under the same operating
conditions was in the order: [NO3] < [DCA] < [BF4] < [PF6] < [TfO] <
[Tf2N] < [methide]. Results also showed that increase in carbon chain
was conducive to absorption [31]. It was also later confirmed that the
order of the CC capacity was [MeSO4]- < [BF4]- < [PF6]- < [Tf2N]- [165].
The absorptivity was less than 0.1 mol CO2/mol IL (ionic liquid). They
were often blended with amines for better absorption performance. But
due to higher viscosity and costs, these blends were suitable only at
laboratory level. Their absorptivity was almost 0.5 mol CO2/mol IL.
With the objective of reducing costs and viscosity, non-ionic surfac-
tants were added to aqueous solutions of ionic liquids. Five ILs viz.,
[bmim][PF6], [emim][OTf], [bmim][OAc], [hmim][NTf2] and [bmim]
[BF4] and three non-ionic surfactants viz., Tween 80, X-100 and PEG-
400 were under study. The absorption capacity of different surfactants
was in the order: X-100 > Tween 80 > PEG-400. Absorbent with the
combination of Tween 80-[emim][OTf]-water had a high absorption
capacity of 0.798 mol/L at room temperature and low pressure. It had
the advantages of good reproducibility, low viscosity, low cost and low
consumption when compared to ionic liquids. It could be used again
after decomposition and regeneration and had the potential to become
a novel absorbent for carbon dioxide [9]. For these Imidazolium based
ionic liquids, Henry’s constant was used to evaluate the solubility of
CO2, where a smaller Henry’s constant means higher solubility [166]. It
was reported that [Bmim][TCM] had lower viscosity yet a higher
absorption capacity when compared to other ILs like [Bmim][BF4],
[Bmim][PF6], [Bmim][Tf2N], [Bmim][DCA] [160]. [bmim][TCM] was
even thermally stable upto operating temperatures as high as 150 oC.
Dai et al. [160] used [Bmim][TCM] as an absorbent and studied the
effects of operating parameters like temperature, flow rates, pressures
on absorption capacity in a tubular glass membrane contractor. At an

Table 2
Summary of various promoters.

Promoter (Concentration, wt%) Concentration (wt%) K2CO3 concentration (wt%) Temperature (K) Accelerationa

MEA 0.5 1.8 291 0.2

5 30 336 15
10 30 336 45
5 25 294 6
DEA 5 25 294 2.6
– – 363 4–5
2 20 353 1.6
2 25 323–363 ~3
5 25 323–363 ~6
PZ 5 20 333 10
MDEA 5 25 294 1
Arginine 0.077 M 35 322 0.44
0.387 M 35 322 1.35
Histidine 0.104 M 35 322 1.54
Glycine (1 M) 1M 30 333 22
Sarcosine (1 M) 1M 30 333 45
Proline (1 M) 1M 30 333 14
Carbonic Anhydrase 300 mg/L 20 298 8.8–11.3
300 mg/L 20 313 5.2–6.4
300 mg/L 20 323 3.4–4.0
300 mg/L 20 313–333 2–6
55 mg/L 30 313 0.3
300 mg/L 20–30 298 6–20
300 mg/L 20–30 323 2–8

a
Acceleration=(Absorption rate in promoted K2CO3 solution)/(Absorption rate in unpromoted K2CO3 solution at same operating conditions).

1085
I. Sreedhar et al.
Table 3
Merits and demerits of promoters used in K2CO3 solution.
Promoter Performance Toxicity
Arsenic acid Very good Very poor
Boric acid Good Very good
Vanadate Good Good
MEA- Very good Good
DEA Very good Good
PZ Very good Good
Glycine Very good Very good
Sarcosine Very good Very good
CA Very good Very good
CA mimicking compounds Good Good
operating temperature of 80 oC and pressure of 20 bar, the overall CO2
flux achieved was 2.5×10-4 mol m-2 s-1. Zareiekordshouli et al. [167]
compared the thermo-physical properties and CO2 absorption capa-
cities of the amine-functionalized cation IL [Amim][Tf2N] and its non-
functionalized equivalent [bmim][Tf2N]. The former had a higher
density, viscosity, molecular weight at certain temperature ranges at
1 atm. The latter showed typical phase behavior like a physical
absorbent. Because of the presence of an amine group, [Amim]
[Tf2N], it was proved to be a better absorbent than its counterpart.
At 308.15 K, it improved the absorption of [bmim][Tf2N] by a factor of
7.4 at 1.1 bar and 3 at 8 bar. While both of them displayed good
regenerability, at the operating conditions (308.15 K and upto 8 bar
pressure), both of them could not exceed the absorption capacity
achieved by 30% MEA solution. At 308.15 K, with a pressure around
6 bar, the loading capacity of [Amim][Tf2N] was comparable to that
obtained from 20% DEA at 323 K. However, its viscosity increased
significantly after absorption. Also, it was found that synthesis of ILs at
a larger scale for scale-up to commercial level was cost intensive. Hence
there is great scope to develop novel and cost effective synthesis
protocols of ILs which are highly promising. Fukumoto et al. [168]
developed a series of amino acid based ionic liquids from twnety
naturally occurring amino acids because of their easy availability, bio-
degradability and bio-compatibility. Amino acids ionic liquids (AAILs)
enhance the absorption capacity of aqueous amines. Various AAILs
have been developed and reported in the recent years like terabutyl-
phosphonium amino acid ([P(C4)4][AA]), trihexyl (tetradecyl) ammo-
nium lysinate ([N66614][Lys]), 1-hexyl-3-methylimidazolium glycinate
([Hmim][Gly]), tetramethylammonium glycinate ([N1111][Gly]), tetra-
methylammonium Lysinate ([N1111][Lys]), tetraethylammonium glyci-
nate ([N2222][Gly]), tetraethylammonium Lysinate ([N2222][Lys]), 1-
butyl-3-methylimidazolium glycinate ([Bmim][Gly]) etc. [168–174].
Out of the above mentioned AAILs, [N1111][Gly] was considered to be a
good choice as a solvent and was reported to have increased the
absorption rate by a significant amount [175–179]. Its glycinate ion
isolates and provides protons to the carbonate/bicarbonates in the
reaction, hence, helps stabilize [180]. Though, the solution viscosity
increased slightly when compared to the lean solution, the reaction
kinetics followed fast pseudo-first order reaction [181]. Owing to
flexibility in structure and good recyclability, ionic liquids have also
been recently reported to be employed as promoters for CO2 absorption
with K2CO3. Fu, in his work [180] reported the CO2 loadings of one
such solution of K2CO3-[N1111][Gly] at temperatures ranging from
303 K to 323 K using the precipitation of BaCO3 method. At 313 K, for
a mass fraction of K2CO3 of 0.15, the mass of CO2 absorbed increased
from 4.03 g CO2/100 g solution to 4.75 g CO2/100 g solution, when
mass fraction of [N1111][Gly] changed from 0 to 0.075. [N1111][Gly]
here not only acted as a promoter but also absorbed some CO2 itself
[181]. The results were validated using previous works [182–184] and
Weiland equation [185] was used to fit the viscosities. After studying
the effects of viscosity, solution composition, vapor pressure of CO2 and
temperature on the absorption capacity, certain conclusions were made
1086
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
Stability Volatility Corrosivity Pilot plant
Very good Very good Good Good
Good Very good Good Poor
Very poor Very good Very good Unclear
Poor Poor Poor Good
Poor Poor Poor Good
Good Good Good Good
Poor Very good Poor Good
Good Very good Unclear Unclear
Very poor Very good Unclear Unclear
Unclear Very good Unclear Unclear
[186–189]. Concentration of absorbent and viscosity exhibited a
contradictory effects on the absorption capacity. An increase in
concentration or a decrease in viscosity led to a higher absorption rate
but with an increase in concentration, the viscosity also increases which
may inhibit the absorption process. Xu et al. [190] synthesized six dual
functional ionic liquids (DFILs) with cation as metal chelate and anion
as pyrazolide - [Na(MDEA)2][Pyra], [Na(AEE)2][Pyra], [Na(DEA)2]
[Pyra], [Na(MDEA)][Pyra], [Na(AEE)][Pyra] and [Na(DEA)][Pyra].
They were synthesized by treating alkanolamines with pyrazole sodium
salt and their CO2 absorption was investigated at 353.2 K.
Characterization techniques like 13C NMR spectra, FT-IR show that
CO2 reacted with a quasi-aza-crown ether cation and [Pyra]- anion
simultaneously. It was reported that DFILs with alkanolamine/pyr-
azole sodium salt ratio of two had a larger CO2 absorption capacity than
those with the molar ratio of one. They were reused over 5 cycles but
showed no signs of loss in absorption capacity. [Na(MDEA)2][Pyra],
[Na(DEA)2][Pyra] were reported to be potential candidates for CC at
high temperatures. Though attempts to synthesize ILs with lower
viscosity were underway, there is still a long way to go in this direction.
Solid adsorbents and membranes incorporated with ILs could also be
an alternative [191] like [P66614][2-Op] on a MCM-41 sieve that
showed higher absorption rate than pure IL [192].
Binding organic liquids (BOLs) exhibited tunable physico-chemical
properties and required low regeneration energy [193–198]. Alcohol-
amine-water blends and polyethylenimine (PEI) immobilized on silica
(SiO2) support have been employed as novel absorbents for carbon
capture [199,200]. PEI anchored onto the surface of Fe3O4 nanopar-
ticles has also been reported as a novel absorbent [94]. It had the
advantage of lower energy consumption as the super paramagnetic
nanoparticles act as supporters and could be easily separated using a
magnet. At 40 °C and 100 kPa, their absorption capacity was
0.373 mol/L. Besides a simple synthesis procedure and excellent CC
capacity, the absorbent could be easily and fully regenerated at
relatively low temperatures (110 °C) and exhibit good stability over
repetitive absorption-desorption cycles. Eftaiha et al. [201] suggested
the use of chitin-acetate (CA) in dimethyl sulfoxide (DMSO) as a
potential candidate to absorb CO2 owing to its high nitrogen content. It
was found that the hydroxyl group at the C-6 of the repeating unit
reacts with CO2 to form organic carbonate. This was stabilized through
ionic interaction and hydrogen bonding via a supramolecular interac-
tion with the neighboring hydroxyl and ammonium groups.
Calculations with the Density Functional Theory (DFT) supported the
CA/CO2 adduct formation. It was reported to have a maximum
sorption capacity of 3.63 mmol CO2/g sorbent at 10% (w/v) measured
within an in situ ATR-FITR. Aoyagi et al. [202] studied the CC and
release behavior of silatranyl amidines - 1-[3-(Silatranyl)propyl]-
1,4,5,6-Tetrahydropyrimidine, 1-[3-(Silatranyl)propyl]-2-imidazoline
and N,N-dimethyl-N′-[3-(Silatranyl)propyl]formamidine, in DMSO
solution in dry conditions. A six membered cyclic amidine captured
CO2 significantly owing to its excellent amidine moiety, at atmospheric
pressure and temperature resulting in the formation of the correspond-
I. Sreedhar et al.
ing amidinium bicarbonate. The trapped CO2 was released under Ar
atmosphere at 60 oC. A five member cyclic amidine group captured and
released CO2 with low efficiency. On the other hand, the acyclic
amidine with silatranyl did not capture CO2 owing to its poor acyclic
amidine moiety. Since, the silatranyl group did not have an impact on
the capture-release process, it proves to be an excellent candidate for
anchoring to silicon surfaces. Future work is directed to synthesize
conductive ionic materials based on these amidines with silatranyl
group.
An O/W emulsion absorbent blended with lipophilic amine has
been reported recently [88]. Absorbents N-methylcyclohexylamine
(MCHA) and 2,6-dimethylpiperidine (DMPD) were used with Tween
80 emulsifier. Absorbent could be renewed at lower temperatures
(333–353 K) using lipophilic amines, therebydemostrating low thermal
degradation and less corrosion. Owing to higher surface area and base
strength of lipophilic amine, the rates of absorption of MCHA and
DMPD emulsion absorbents were about 80–132% and 58–83% higher
than aqueous MEA solution at 313 K respectively. Calcium oxide (CaO)
was also tried as an absorbent. Though the process efficiency improved
and hydrogen yield increased five times over coal pyrolysis [203], it had
some drawbacks pertaining to its lifetime and durability. The reaction
involved substantial change in specific volume leading to severe
mechanical stresses in porous absorbent. Also, calcination carried out
at elevated temperature caused sintering of particles [204,205]. To
improve the structural integrity of porous structures, various new
forms like dolomite (CaCO3.MgCO3) were used; CaO dispersed in
porous inert calcium titanate (CaTiO3) matrix, CaO impregnated in
porous alumina granules and titanium ethoxide, or mixed with
mayenite (Ca12Al14O33) or nano-sized alumina (Al2O3) particles were
developed and tested. Lee et al. [92] immobilized nano-sized CaO on
fibrous ceramic based fabrics (yttria and alumina mat). With simplified
procedure and improved performance, cyclic reactivity of CaO was
preserved. The sample had a degree of carbonation of 55% using yttria
fabric as substrate and did not degrade even upto twelve cycles, while
the value dropped by about 8% from the maximum value of 59% after
that. Yttria fabric as a substrate for calcium oxide absorbent was
superior to the alumina fabric at severe calcination conditions due to
the formation of Ca12Al14O33 by reaction between CaO and alumina at
temperatures above 800 oC [90]. CaO has also been tested as an
absorbent in a radio frequency (RF) heated reactor. For a mixture
containing 30% CO2 (by volume), the process of absorption and
desorption was repeated for over twenty times. CaO demonstrated
stability and an absorption efficiency of 99.9%. During the reverse flow
mode, CO2 desorption decreased the partial pressure of CO2, thus
reducing the cycle time by 20% and preventing CO2 re-absorption [97].
Zhu et al. [206] studied CC in a vertical bio-reactor using CA
immobilized in alginate polymers as successful absorbents. It followed
a biomimetic route, enhanced enzyme activity, increased chemical
resistance, pH and thermal stability. It could be reused for six cycles
retaining upto 61% of its original activity. Its surface was characterized
for any possible structural changes through SEM. Pure aqueous
solvents like aqueous ammonia, aqueous potassium carbonate have
been used too.
Nanoparticles (SiO2 and Al2O3) were added to methanol to enhance
the absorption process and the performance has been evaluated in a
CO2 absorption and regeneration. It was found that there was an
optimum amount of nanoparticles that could be added for mass
transfer enhancement [207–209]. This was because of two reasons,
firstly, the nanoparticles interfere with the gas-liquid contact area,
leading to reduction in movement of particles; secondly, addition of
nanoparticles in general leads to an increase in viscosity thereby
leading to a decrease in diffusion. The absorption rates decreased with
the progress of the absorption process. After about 8000 s, the
absorption rates obtained were constant – 0.046 g/s for methanol,
0.042 gm/s for Al2O3/methanol, 0.050 g/s for SiO2/methanol. Al2O3/
methanol displayed a reduction in absorption rate by 9% after four
1087
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
cycles when compared to pure methanol, whereas SiO2/methanol
displayed an increase in absorption rate by 8% after five cycles. For
the combined process, the performance of Al2O3/methanol was low
because of carbonates or bicarbonates formed on CO2 bonding with
Al2O3. Cluster particle size played an integral role, as the surface effect
of the nanoparticles dominated the thermal effect. Also, it was
concluded that the absorption and regeneration processes should take
place simultaneously to use nanoabsorbents in reality [210].
Biphasic solvents systems could be classified as two types: amines
in alcohol (non-aqueous systems) and less soluble amines (alkanola-
mines/lipophilic amines) [211]. Systems like DMX [212] and thermo-
morphic biphasic solvents (TBS) [213] were a new class of blends. In
the DMX process, two immiscible phases were formed – CO2 rich and
CO2 lean phases. To cut down on the reboiler duty, i.e., to increase the
stripper efficiency, the CO2 rich phase was directly transported to the
stripper column. Reboiler duty in the DMX process was 2.3 GJt-1 CO2
while a conventional MEA process required 3.7 GJt-1 CO2 [214]. In
TBS systems, MAPA+DEEA was stripped at temperatures below its
boiling point, reducing regeneration energy requirements and enabling
the use of less expensive utilities for regeneration [215]. In general,
when compared to MEA, biphasic solvent systems have been reported
to have lower regeneration temperatures and higher regeneration
depth. Wang et al. [216] developed phase-change absorbents with
N,N-dimethylbutylamine (DMBA) and N,N-diethylaminoethanol
(DEEA). A wetted wall column was used to examine the impact of
temperature, gas flow rate, CO2 loading, DMBA/DEEA composition
ratio on absorption rate and kinetics. It was also revealed that 4 M
DMBA+2 M DEEA exhibited higher rate of absorption and stability
than 2 M DMBA+4 M DEEA.
Bai et al. [217] studied the impact of types of bonding, -NH content,
flowability and particle size on CC using series of liquid-Ds (polyhedral
oligomeric silsesquioxanes derivatives). Polyetheramine M2070 and
polyetheramine M1000 (similar in structure but different in molecular
weights) were used as canopies whereas 3-(trihydroxylsilyl)-1-pro-
pane-sulfoni with sulfonic acid group (SIT8378.3) and γ-(2,3-epox-
ypropoxy)-propyltrimethoxysilane with epoxy group (KH560) were
used as corona structures. The latter could form an ionic or covalent
bond with the former, thus the effect of bonding could be studied.
Polyetheramine D2000 and M2070 had similar weights but different –
NH content, thus the effect of –NH content could be studied. It was
reported that C-liquid-Ds (involving combinations of KH560 with
either M2070 or M1000) had better recyclability. It was also concluded
that the –NH content increased CC by liquid-Ds.
Physical absorbents like Selexol, Rectisol and Purisol were also
reported to be employed in CC. Rectisol had the highest carbon dioxide
capture capacity and absorption rate. However, CC by Rectisol was
implemented only at low temperatures which required refrigeration of
solvent, thereby increasing the capital and operating costs of the plant.
It was reported that for the operating temperature of 303–333 K, the
Henry’s law constant of Selexol was far higher than water, indicating
more feasibility of Selexol than water [26,218].
Tables 4 and 5 capture the summary of the single and blended
absorbents respectively with their performance, efficiency, merits and
demerits,
Solvent selection and development of an appropriate synthesis
protocol play a critical role in the successful deployment of absorption
based CC. Right from MEAs introduced long back at the inception,
there have been significant research efforts in identifying novel
absorbents and their blends including the blends of primary, second-
ary, tertiary and sterically hindered amines, potassium carbonate,
amine blends with carbonates, ionic liquids with and without anion
functionalization, ammonia based solvents, piperazine, BOLs, biphasic
solvents etc. Though absorption based CC has been commercialized
with few solvents, the major hurdle in the deployment of this
technology was energy intensive regeneration and environmental
pollution or corrosion issues associated with the process and degrada-
 I. Sreedhar et al.
Table 4
Overview of single solvents used in CC.

Absorbent CC (%), AC (kg/ Merits Demerits Remarks

kg)

MEA 100,0.45 low cost, high reactivity energy intensive, pilot, spray column provides superior absorption than packed
mol/mol degrade and corrode in oxidizing
atmosphere, limited loading capacity
DEA low cost, high absorption rate energy intensive, corrode in oxidizing
atmosphere, limited loading capacity
DIPA resistant to corrosion, reduced regeneration costs low absorption rate
DEAE NA Good chemical stability, low heat of reaction prepared from renewable sources, substitute to tertiary alkanolamines,
BDA used as promoter
DAMP 0.86 Higher absorption capacity, less volatile
AEPD, TETA, TEPA, THAM decrease regeneration energy significantly More hydroxyl group, shorter carbon chain, existence of methyl group
elevates CO2 stripping rate
DMAB 0.88 High absorption rates, low heat of absorption, fast
HEEBAB 0.68 regeneration rates
HEMAB 0.44
2PE Thermally and oxidatively robust, fast kinetics, low Highly viscous, carcinogenic
ecotoxicity, easily biodegradable
1088

AMP 90, NA high absorption capacity, high selectivity, easy regenerability, low absorption rate linear reduction in temperature with height of column
corrosion and degradation resistant
PZ ~100, 0.32 resistant to corrosion and cyclic amine, successful as a promoter, employed with MEA, DETA and
thermal degradation, rapid formation to form carbamates, AMP
reduced regeneration costs
AEP 2.2 times MEA AC Higher absorption capacity Contains all 3 – primary, secondary and tertiary amines, no bicarbamate
formed
NH3 95–99, 1.2 high absorption capacity, Ammonia slip pilot scale project
easy availability, low regeneration costs, removes SO2,NOx,

Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

used as fertilizers
K2CO3 99.4, NA low enthalpy requirements, low mass transfer rates inorganic salts and organics added to compensate for low mass transfer
less cost, low toxicity, high degradation resistance
o
Ionic liquids 99.11 at 60 C, non-inflammability, non- explosivity, high thermal and imidazolium based; increase in carbon chain conducive to absorption,
0.1 mol/mol chemical stability, high reaction selectivity, low VP, density and viscosity increase with concentration & decrease with
economical temperature
Binding organic liquids highly energy efficient during regeneration tunable physico- chemical properties
CaO high process efficiency sintering of particles, high mechanical hydrogen yield five times the conventional pyrolysis
stress developed
CaO in RF heated reactor Resistance to degradation and high absorption efficiency of 9 times lesser environmental heat losses
99.9%
CA immobilized in alginate Increased chemical, pH resistance and enhanced enzyme
polymers activity
 I. Sreedhar et al.
Table 5
Overview of solvent blends used in CC.

Adsorbent CC (%), AC (kg/kg) Merits Demerits Remarks

MDEA 99, NA improved CO2 loading, resistant to corrosion and degradation, lower absorption rates does not polymerize unlike primary and
reduced regeneration costs secondary amines
MEA+methanol NA Lower regeneration energy requirement Pilot plant
MEA+glycerol NA Low pressure and low concentrations – absorption capacity higher High pressures and high concentration of Gibbs Helmholtz evaluates isosteric heats of
glycerol – decrease in CO2 solubility absorption
MEA+PL Higher absorption capacity than MEA
MMEA+PZ NA, 0.49 mol/L easy availability, low cost, low corrosiveness due to weaker alkalinity
low circulation rate
DEAE+PZ Increase in WPZ by 0.026, absorption rate increases 3 fold Viscosity increases with increase in CO2
loading, decrease in temperature
PZ+Ni(II) 72% Complex formation, Reduction in NH3 escape Bubbling reactor
AMP+PZ 90, NA Pilot plant, packed bed, 22 wt% AMP+8 wt% PZ
1089

DETA (20%)+sulfolane Higher reactivity, less heat of reaction, high mass transfer, less heat Phase splitting happens
(40%)+H2O(40%) of regeneration, stability and corrosion resistant
AMP+PZ+MEA (50%) High cyclic capacities, low heat duties AMP+PZ kept at 50% of solvent to avoid
precipitation
IL+amines > 99, 0.5 mol/mol better absorption performance high viscosity, high costs suitable for lab research
IL+non-ionic surfactant > 99, 0.798 mol/L good reproducibility, low potential absorbent for commercializing
viscosity, low cost, used again after decomposition
PEI on SiO2 support NA, NA High absorption efficiency, high desorption
PEI+Fe3O4 NA, 0.373 mol/L lower energy consumption, simple synthesis, fully regenerated at low

Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

temperatures, stable over repetitive absorption- desorption
MCHA+Tween80 NA, 0.87 mol/mol CO2 loading decreases with increasing
(313 K) temperature at same partial pressure
DMPD+Tween80 NA, 0.72 mol/mol lower activation energy than alkanolamine
(313 K)
CaO onto alumina fabric structural integrity improved, 55% of carbonation achieved degrades after 12 cycles due to formation of
Ca12Al14O33
CaO onto yttria fabric structural integrity improved, 59% of carbonation achieved
I. Sreedhar et al.
tion of the solvents. The major target for the economic viability of this
technology was its successful integration with power generation unit in
the post combustion mode. This requires absorbents with ideal
attributes viz., high thermal and degradation resistance, easy regenera-
tion, high capture capacity, oxidative stability and absorption rate, low
heat of absorption, low viscosity, reactivity, toxicity and volatility.
There is also a need of developing novel solvent regeneration options
besides making temperature and pressure swing operation, employing
novel PH swing (electrochemical) too. Also there is a strong need to
focus on the identification of solvents that are suitable for multi-
component removal instead of only CO2 separation from flue gas.
Solvent reclamation using various methods like solvent purging,
mechanics based or activated carbon based filtration, thermal reclama-
tion, online neutrallization, employing ion exchange or electro-dialysis
reclaimers.
3. Contactors used in absorption based CC
At the operational level, the flue gases at atmospheric pressure
enter the absorber column and simultaneously solvent from the
stripper section was pumped into the absorber in the counter-current
fashion. CO2 was absorbed into the solvent which was further pumped
to the top of stripper which operates at 100–150 °C. In the stripper, the
reverse process takes place by the virtue of the thermal swing which
then results in removal of CO2 from the solvent. The regenerated
solvent was again pumped into the absorber column and then CO2 was
compressed and stored. This absorption/stripping technology was an
energy demanding process and the overall cost involved in this process
was nearly 52–77 US$/tonne CO2 . The required energy for regenera-
tion for CC from coal-fired power plant varies from 3.24 to 4.2 GJ/
tonne CO2 [219]. Around 60% of the total energy consumption was
involved in solvent regeneration [37]. In order to reduce the energy
penalty and increase the process efficiency, extensive literature about
gas-liquid contactor configurations, stripping configurations and sol-
vent systems have been reported.
3.1. Absorber configurations
For PCC, various kinds of conventional absorption configurations
like tray column, packed bed column, spray column, and bubble
column have been employed. However the use of the conventional
technology has led to an increase in the electricity demand because of
which new absorber configurations like rotating packed bed, mini
channel reactor, membrane and vortex contactors have been proposed
and developed in the recent years.
The conventional absorption process involved the direct contact of
acid gas with absorber (generally a high column) which has the lean
solvent in it through which the mass transfer happens at the liquid -gas
interface [220]. The industrial absorbers (gas–liquid contactors) were
of two types [221]:
1. Differential contactors: It includes packed columns, bubble col-
umns, spray towers, where mass transfer occurs throughout the
length of the contactor and there is no gas– liquid equilibrium at any
point.
2. Stage-wise contactors: It includes plate columns and rotating disc
contactors, where gas–liquid equilibrium was attained at each stage
and the mass transfer happens only in a part of the volume.
Some of the different gas-liquid contactors which have been
employed for the post combustion carbon capture shall be discussed.
Tray columns were one of the widely employed configurations for
carbon capture. It also formed an important basis for the equilibrium
based models. The CO2 uptake depends on the tray efficiency profile.
Typical efficiencies reported were 337% for the bottom tray, 14% for
top tray and a maximum efficiency of 45% on the 15th tray from the
1090
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
bottom in the high pressure absorption [220]. Increase in operational
pressures up to 17 bar resulted in improved efficiency of column but at
the same time it reduced the process safety and was accompanied with
high operational cost and high pressure drop. Therefore packed
columns were suggested as a reliable alternative for the process of
carbon capture [222].
Packed columns were widely used in numerous gas separation
technologies and chemical industries. Packed bed columns have lower
residence times and lower bottom temperatures which facilitated better
separation of heat sensitive mixtures than that of the tray columns
[223]. The important parameters involved in the design of packed
columns were column diameter, type and nature of packing, and
packing height. Packing section was the heart of the packed bed
absorption process which provided the surface area for the gas and
liquid phases to contact. The sole purpose of the packing was to
increase the efficiency for a given capacity, at a reasonable cost. An
ideal packing material has large and uniform surface area which
improves the Gas-Liquid contact, maximum void space per unit
column volume that lowers the resistance to the gas, minimum friction
and cost [221]. There were two types of material available for packing:
Random packing (Pall ring, IMTP, Raschig rings) and Structured
Packing (Flexipac, Mellapak, Gempak, BX). The overall mass transfer
coefficient was higher in structured packing than that of the random
packing [224] due to larger contact area provided by structured
packing for gas and liquid phases225]. Various packings such as
Mellapak have been reported to reduce pressure drop in heat and
mass transfer operations and also proven to provide excellent perfor-
mance for columns with diameters upto 15 m [222,226]. The corruga-
tion angle was kept as 30° for X series of Mellapak and as 45° for the Y-
types [227]. These geometrical changes enabled major increase in
capacity but have also resulted in considerable reduction in purity of
the product [225,228]. BX, Flexipac 1Y and Mellapak 350Y were
recommended for flue gas treatment in coal and gas fired power plants
[223]. Apart from the type of packing, other major factors affecting the
packed bed column were physical properties of solvent, partial pressure
of CO2 and total pressure of the system, absorbent concentration and
CO2 loading on the solvent [229]. The physical properties of solvent
mainly like the surface tension had influence on the mass transfer and
also resulted in the increases of effective packing area [230]. Increase
in partial pressure of CO2 also improved the mass transfer efficiency
[35,231,232]. The flow rate of liquid had been reported to give a
positive impact on the carbon capture efficiency where as in case of gas
flow rate, clarity is still needed [35,230–234]. Concentration of
absorbent was found to have influence on CC efficiency, temperature,
heat of reaction and the CO2 vapor pressure [36,230,235]. A high
temperature of absorption resulted in an increase in viscosity which
leads to reduction in solubility of CO2 [233]. Wang et al. [236] studied
CC in packed column with α-Al2O3 ceramic foam packings using 30 wt
% diglycolamine (DGA) as solvent. They permitted higher flow rates of
gas and liquid, with an upper bound to liquid velocity so that flooding
was averted. Foams with lower PPI had less tendency to flood.
Reducing gas velocity led to increase in CO2 absorption and reduction
in operating efficiency. The bottom of the column experienced a hot
spot (temperature < 5 oC) where L/G ratio was found to be very low.
With increase in CO2 loading from 0 to 0.35 mol CO2/mol DGA, free
amine reduced and CO2 equilibrium pressure increased in liquid,
hence, the absorption efficiency decreased. At CO2 loading as high as
0.5 mol CO2/mol DGA, CO2 equilibrium pressure crossed inlet CO2 gas
pressure and desorption occured.
Bubble columns have also been reported as one of the powerful
process for absorption of CO2 owing to its simple construction, better
control of the liquid residence time, higher heat and mass transfer
coefficient and better solid handling capacity [237,238]. In bubble
column reactor, the liquid height to diameter ratio plays a crucial role
for improving the capture performance. It has been reported that large
liquid height-to-diameter ratio (H/D) increased CO2 capture efficiency
I. Sreedhar et al.
due to increase in gas bubble holdup time [239]. However these
columns have very low gas flow rate, gas holdup and the gas superficial
velocity, thus involving a large reactor size [238].
Use of spray tower for CO2 capture has been reported to provide
various advantages like reduction in investment and operating costs
because of the absence of expensive packing, very little pressure drop,
which provided the opportunity to use precipitating solvents and a
relatively higher flexibility of the process [240]. The key parameter for
the spray column was the mass transfer coefficient which decreased
with a decrease in partial pressure and the loading of CO2, increased
with an incremental increase in flow rate of liquid and gas. The
different types of nozzles also provided different mass transfer coeffi-
cient ranges because of the variation in hydrodynamic capacity of the
spray nozzle. The spray columns were also reported to have an edge
over packed column due to their larger interfacial area provided.
However spray columns offer short contacting time for carbon capture
[241]. Zhao et al. [242] developed a multistage spray reactor to
improve the absorption of CO2 in ammonia. A CFD simulation showed
that the vortex flow based model included complex multidimensional
velocity profiles (V-shaped tangential velocities, non-uniform axial
velocity profile) which enhanced the contact area, mixing and hence,
increased mass transfer. The CC efficiency increased from 73% to
80.3% with increase in liquid flow rate while it decreased from 86.7% to
76.1% with increase in gas flow rate. The vortex flow simulation led to
an increase in 7–15% of CC efficiency and 18–33% in gas phase mass
transfer coefficients.
All these conventional methods were mature and efficient but still
collectively pose major drawbacks in terms of size, mass transfer
resistance at the gas-liquid interface, operating and electrical costs,
corrosion of equipment, foaming, entrainment and loss of solvent, thus
paving the way for the non-conventional methods.
Rotating packed bed was one of the novel methods which has been
reported to combat the above drawbacks significantly. In a RPB, liquid
comes in contact with gas on the surface of packing under the influence
of the centrifugal field. The liquid was split into tiny droplets and film
under high gravity and subsequently the mass transfer and liquid gas
contact area increased. Use of the rotating packed bed with the mixture
of DETA and PZ for 10% CO2 has been reported to have advantage over
the packed bed because of increased liquid gas contact area, mass
transfer rate and small size for attaining similar capture efficiency, of
greater than 90% [37,243–246]. The performance of high gravity based
RPBs depends on various factors like liquid and gas flow rates, type and
concentration of liquid, temperature of reaction and CO2 concentration
[247]. Improvised packings in RPBs like disc plate packings [248]
multistaged spraying packings [249], helical packings [250], zigzag
packing [251] and split-packings [252,253] have been tried. Wong
et al. [254] compared PB and RPB absorbers based on a pilot plant
study using 30 wt% MEA as a solvent. Both exhibited similar perfor-
mances with similar energy consumption. However, the volume of RPB
was one-third of PB absorber. The results obtained using a 20 wt%
MEA solvent were also in agreement with simulation studies using
ASPEN rate-based model for PB and ASPEN custom modeler for RPB.
Similar results were reported by Yu et al. [255] who studied the effect
of contactor configurations on absorption capacities. He used ’11m′
MEA as an absorbent and presented the results on various configura-
tions like single RPB, single PB, and RPB+PB, PB+RPB, RPB+RPB in
series. To achieve the same efficiency using the same reactor volume,
RPB was better than PB. RPB+RPB in series demonstrated further
improved performance. It had an increase of 20.6% in the amount of
treated gas over a single RPB of equivalent volume. The driving force
for mass transfer increased. Similarly, for two RPBs in series, the
regeneration energy reduced significantly by 9.5%. As usual, the
efficiency reduced with increase in gas flow rate but quantitatively,
the amount of CO2 captured increased. On replacing MEA with ‘4m′
DETA +‘4m′ PZ, the amount of gas treated increased to 80–140%. It
was also concluded that CO2 capture amount, rather than CO2 capture
1091
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
efficiency must be used as a decisive parameter – in DETA+PZ, even
though CC efficiency reduced from 90% to 80%, the amount of CO2
captured still increased by 75%. A glycerol+NaOH blend [256] was also
tested in a RPB. Under high rotating speed (1500 rpm), the absorption
percentage and enhancement factor were doubled. At the same NaOH
concentration, the coefficients for mass transfer increased five times
after the addition of glycerol, owing to the intense centrifugal field in
viscous media.
Recently, immense research efforts were directed on novel con-
tactors like mini-channel reactors, vortex tube contactors and mem-
brane contactors. Mini channel reactor was reported to offer high
surface area to volume ratio with almost 100% capture efficiency using
DEA [257]. In vortex tube contactors, the separation was facilitated by
forcing expanded gas and the liquid solvent into the rotational flow
field which increased the turbulence and resulted in effective CO2
transfer into the absorbent. It significantly reduced the regeneration
requirements and fluxes of absorbents, thus reducing the capital and
operating expenses. The cost of gas compression involved in this
process could be reduced with modification in aperture nozzle design
[258].
Membrane contactor technology was also one of the promising
alternatives which has been successfully reported to overcome the
challenges posed by the conventional ones [259–262]. This process
provided the advantages of high chemical selectivity and compact
structure of membrane technology. In this, the liquid absorbent
recovers the CO2 that was separated from flue gas by the virtue of its
diffusion through the membrane. In gas-liquid contactors, porous
hydrophobic membranes were used as non-selective barriers that help
define the physical gas-liquid interface. This could thus be operated
with negligible pressurization with a large surface area/volume at a
very low energy consumption [263–268]. The usage of conventional
absorbents having low surface tension like MDEA, MEA, DEA leads to
membrane pore wetting, subsequently damaging the porous structure.
To avoid pore wetting, which also would lead to an undesirable increase
in mass transfer resistances, hydrophobic membranes were employed
[269]. Meanwhile, the compatibility of ionic liquids with polymeric
membranes is still under study. Using solvents like carbonate solutions
or water tends to avoid these problems but results in low kinetics of
CO2 hydration. To improve the hydration efficiency, bio-catalyst
carbonic anhydrase (CA) was used. It was immobilized to enhance
stability and enable reusability [270–272]. To test the feasibility of the
process, few bio-reactors like bio-catalytic membrane reactors
[273,274], Integrated Vacuum Carbonate Absorption Process
(IVCAP) reactor [270,275], Janus bio-catalytic membrane reactor
[276] bubbling reactor [277], hydrogel reactor [278,279] etc. have
been constructed in the past. Zhang et al. [270,275] developed an
IVCAP reactor series where nanoparticles were suspended in the
carbonate solution put in a sealed glass container with pure CO2.
Significant hydration was observed and hence continuous operation of
the reactor was considered very difficult. In hydrogel reactor, CA was
encapsulated in the hydrogel, poured into the interstitial voids between
two successive Hydrophobic Polyvinylidene Fluoride (PVDF) mem-
branes. This posed two problems: first, the hydrogel imposed resistance
to mass transfer; secondly, process of regeneration of denatured
enzymes wasn’t easy. In an attempt to overcome these drawbacks of
the above two models, a combination of the bio-catalytic process in a
gas-liquid membrane contactor was attempted. Arazawa et al. [274]
developed the CA coated hollow fiber membrane which was used as an
artificial lung. Yong et al. [280–282] developed bio-catalytic membrane
reactor series wherein immobilized CA was coated onto the surface of
polymeric membranes (using the layer to layer approach or intermedi-
ate supports like SiO2 nanoparticles). Huo et al. [276] developed a
Janus-like hydrophilic/super-hydrophobic membrane contactor (hy-
drophilic layer to enhance immobilization of enzymes and super-
hydrophobic layer to address the pore wetting. CA was strategically
immobilized at the interface of gas and liquid to minimize the
I. Sreedhar et al.
resistance to mass transfer. Though they exhibited improved hydration
efficiency, easy reusability and regenerability of the enzymes , a
complicated process of fabrication and lower enzyme loading limits
its application. Huo et al. [283] covalently immobilized CA on TiO2
nanoparticles. Bio-catalytic membrane contactors (both virgin and
superhydrophobic polypropylene membranes) were developed and
investigated for the effects of superhydrophbic modifications, enzyme
concentration, liquid and gas flow rates and temperatures and pH on
contractor performance. Results were reported on both the cases. On
the superhydrophobic modification, the water contact angle changed
from 105 to 160 and the penetration pressure increased from 3.05 bar
to 3.75 bar. The SEM results indicated a slight yet tolerable reduction
in the pore size of the membrane, since at the interface, evaporation of
water led to condensation in capillary. The virgin membrane lost all
efficiency after 60 min while the latter, even after 10 cycles of 30 min
each, retained CO2 hydration activity. When compared to the conven-
tional absorption equipment, it had the attributes of large contact area,
flexibility in operations, independent control over liquid and gas, a
compact size, ease of scale-up, modularity and high surface-area-to-
volume ratio [284–286].
Studies were reported on Non-dispersive absorption using Gas
Liquid Hollow Fiber Membrane Contractors (GLHFMC) [287–296].
Hollow fiber contractors (HFCs) had many fibers packed tightly in a
membrane module, with two inlets for solvent and gas feed and 2
outlets [297]. They provide a very large surface area per contact
volume, thus enabling better separation of gas-liquid mixtures. The
concentration gradient between the two phases becomes the driving
force. This poses one disadvantage – the absorbent liquid wets the
pores of hollow fiber membranes, increasing the mas transfer resis-
tance and affecting the life of GLHFMC. To eliminate this problem,
hydrophobic membranes had been proposed. Poly Tetrafluoro
Ethlylene (PTFE) membrane was developed by Esato and Eiseman
[298]. However, it was expensive [299]. Qi and Cussler [300–302]
employed polypropylene HFCs for the first time and reported the
results using NaOH as absorbent. Recently PVDF polymer has been
used to develop GLHFMC because of their solubility, good chemical
and thermal resistances [303]. Adding to PVDF and PTFE, PEI
[304,305], Poly Ether Sulfolane (PES) [306], PolySulfolane (PSf)
[307,308], and PolyPropylene (PP) [300,301] have also been used to
make GLHFMC. PVDF, PSf and PES were very expensive and though
PP was cheaper, it had a very low chemical resistance [300]. Fashandi
et al. [299] studied Poly Vinyl Chloride (PVC) as another polymer to
make GLHFMC because of its good processability. NMP was used as a
solvent in this. PVC/NMP solution with polymer concentration of 20 wt
% was prepared by extrusion through the annulus of spinneret into
coagulation bath of tap water. An optimum take up speed was found
out which pronounced the effect of increase in absorbent flow rate,
thereby enhancing CO2 capture. Using 10ethyl-3-methylimidazolium
acetate [emim][Ac] [309], it was reported that the use of polysulfone
(Ps) hollow fiber contractor was better than a polypropylene (PP)
hollow fiber membrane. Ps was also reported to possess higher thermal
stability than PP. Field tests using activated MDEA were done in a poly
ether ether ketone (PEEK) hollow fiber membrane [523] having a
diameter of 4 in.. It was reported to have a CO2 removal greater than
90%. This was further scaled to 8 in. modules. They were found to have
an intrinsic permeance of 2150–2670 GPU and achieved a CO2
removal rate of 91.3% . An increase in solvent temperature, flow
velocity and feed pressure led to an increase in CO2 flux and rate of
capture. PEEK HFMCs were effective in CC from low concentration
feeds, thereby exhibiting good potential for natural gas or coal CC.
Membrane fouling in membrane based contactors was caused not
only by pore wetting but was also induced by other components in the
flue gas (SOx, NOx and water vapor). Zhang et al. [310] studied the
synergistic effects of water vapor coupling with NOx and SOx on
PolyPropylene HFM performance. It was found that as long as the
concentration of NO2 was lower than CO2, the effect of NO2 on
1092
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
membrane fouling could be neglected. H2O had a capillary condensa-
tion effect on CO2 absorption and reduced CO2 absorbing capacity by
18.2%. SO2 had a competitive effect on CO2 absorption and reduced it
by 29.3%. EDS and FTIR analyses indicated that complete absorption
of SO3 by MEA was not possible, hence, a new sulfonic group might be
generated which could acidify and change the fundamental surface
wettability of the PP membrane. The SO2 exposure could be avoided by
sweeping nitrogen. Fly ash blocked the pored of membrane, thus
affecting its performance.
Saeed et al. [98] studied the absorption process of CO2 using a Zn-
cyclen (a mimic enzyme to CA) promoted K2CO3 solvent in a hydro-
phobic porous glass membrane. Hydrophobicity was reported to be
retained for a long period , avoiding pore wetting and thus providing
high mass transfer rates. The mass transfer characteristics of the
membrane were studied using NaOH as it resulted in a faster reaction.
The membrane was found to be the limiting factor in the process and
contributed to around 85% of the entire mass transfer resistance.
Certain combustion reactions produce a mixture of Volatile Organic
Compounds (VOCs, n-hexane) and CO2 thus polluting the environ-
ment. Diego et al. [311] studied the introduction of a second liquid
phase (MDEA) in a two-phase partitioning bioreactor (TPPR)
[312,313]. This liquid-liquid phase formed stable emulsions and
improved the mass transfer rate between the gas-liquid and organic-
liquid phases [314]. Dependencies of CO2 capture on gas flow rate,
presence of inerts and amine concentrations were reported. CO2
absorption occured through gas-aqueous interface while n-hexane
absorption occured through gas-organic interface.
In conventional heating modes, methane was burnt to heat up the
surrounding air and this hot air was employed as a heat transfer
medium to heat the reactor. Instead of the conventional heating mode
used in reactors, a radio frequency heated reactor was tried and tested
using CaO as an absorbent [97]. RF heating used electromagnetic (EM)
energy to selectively heat materials, giving higher penetration depths
and better safety aspects when compared to microwave (MW) heating.
RF irradiation was rapid and heated the entire reactor instantaneously.
So, conventional heating methods required 6.5 times more energy than
RF heating modes. An efficiency of 99.9% CO2 absorption was reported
in both the methods, but the former resulted in nine times higher
environmental heat losses. The transition time between absorption and
desorption was also reduced by a factor of three in the latter. However,
there is still a great scope to improve the CO2 desorption rate in the RF
heated reactors. The resistance offered by the membrane due to
excessive wetting and fouling were some of the challenges posed by
this technology but the ongoing studies about the use of ceramic
membrane, development of hollow fiber membrane contactor using
strong membrane substance (e.g., fluorinated polyamides PEEK, PTEF
and inorganic), use of new solvents with membrane like ionic liquids
have been reported to minimize the above drawbacks [259,315].
Future work should be aimed at finding the optimal module design
and geometry for membranes. Efforts should also be directed towards
its application in regeneration process in order to establish itself as the
commercialized product that could potentially reduce the cost of
carbon capture [285].
3.2. Stripper configurations
Over the past few years, different stripper configurations for
optimum energy consumption in carbon capture processes have been
reported in literature. Simple stripper was used as the baseline
configuration in industries which was operated at 160 kPa and later
the CO2 was further compressed to 1000 kPa in five stages. The
stripping at higher pressure also led to rise in temperature of stripper
which made the process easier [316]. The Vacuum stripper configura-
tion was similar to the simple stripper. It was operated at the pressure
30 kPa and the reboiler temperature was around 60–80 °C. The low
pressure steam ensured lower electricity demand and lower tempera-
 I. Sreedhar et al.
Table 6
Summary of contactors employed for CC.

Contactors Critical Design Parameter Efficiency Merits Demerits Scale Remarks Refs.

Tray Column -types of tray used - gas-liquid flow rates Around 45% Conventional and simple way for CC -large amount of pressure drop Various pilot plants have been set Not suitable for [220,328]
–pressure –column diameter and height -elevated column size to process up and is also being employed at the industrial
-tray spacing -number of trays huge amount of volume the industrial level. scale
- Difficult to design and support a
large number of trays.
-High Operational Cost
-Plugging
Packed Column -Types of Packing used Up to 95% Conventional and most widely used -elevated device size to process huge Used at industrial level [223,238,328–330]
-Packing Height and diameter technology for CC amount of volume.
-HETP Gas-Liquid -lower residence time and bottom -high pressure drops
flow rate temperature ensures a better -plugging
-column pressure separation -difficult to clean packing and repair
-concentration temperature of inlets the damage without emptying the
column
-packing also offers increased
resistance to mass transfer
-high operational cost
Bubble Column – liquid height to diameter ratio Up to 95% -simple construction - – low gas flow rates, superficial Industrial level and at lab scale [237,238,330–332]
– Gas-liquid flow rate higher heat and velocity and gas hold up capacity
– inlet concentration, pressure and mass transfer coefficient pose a requirement of large device
temperature -control of the liquid residence time size
– total reactor volume -improved solid handling capacity – Difficult to eliminate exothermic
1093

– Gas-liquid flow rate heat.

– inlet concentration, pressure and – poor solvent-gas flow distribution
temperature – channeling
– total reactor volume
Rotating -rotary design 75–99% -reduced device size with the similar Lab scale, yet to be [37,243–246,349–351]
Packed Bed -speed capture efficiency to that of the commercialized at large scale
-packing density conventional ones
-mass transfer rate -improved mass transfer and
increment in the contacting area

Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

Vortex tube -Vortex inlet Up to 95% -Improved mass transfer by high Lab scale, has the potential to be [258]
-Tube design turbulence, small thickness of film commercialized at large scale.
-Tube length and small droplets due to high speed
acceleration
-Simple design
-Lessen the requirement for solvent
regeneration
-Requires low maintenance
-Reduction in cost
Membrane -type of membrane used and packing Upto 90% -High Chemical selectivity -wetting Pilot plants have been developed [261,262,284,285,315]
Contactor fraction -flexible operations -fouling but still has to be commercialized
-interfacial area -independent liquid and gas
-pressure and flow of gas and liquid -compact size
-overall mass transfer coefficient -ease of scale- up
-solvent loading capacity -modularity
-high surface
-area-to-volume ratio
I. Sreedhar et al.
ture kept a check on corrosion and amine degradation but with slower
kinetics [316].
Modified split flow scheme developed by Aroonwilas and Veawab
resulted in the reduction of steam consumption [317]. Studies were
performed on multi-pressure and vapor recompression stripping with
vapor recompression by Jassim and Rochelle and reduction in total
equivalent work by 3–11% in comparison to simple stripper column
was reported [318]. Matrix stripper configuration was also proposed
which resulted in 8% less total equivalent work in comparison with a
base line configuration [319]. Later on, several modifications were
made on matrix stripper which further resulted in the reduction of total
power required by 4.5% [320]. Additionally, several other improve-
ments such as absorption flash stripping, intercooler absorption,
stripper condenser with liquid bypass, multi- pressure with split,
internal exchange stripper,and vacuum stripper were reported in
literature and resulted in reduction of the energy penalty when
compared to the baseline MEA capture process but these configura-
tions didn’t use the latent heat from overhead section efficiently [319–
325]. Thus several new improvements to the intercooling compression
system, introduction of split flow to the vapor recompression (SFVR)
and improved split flow to vapor recompression (ISFVR) were reported
in the literature which resulted in reduction of number of compressor
stages, cooling water and energy consumption as the rich stream was
partially heated by the overhead vapor before entering into the stripper
[325]. Continuous Research in the direction of advanced stripping/
absorption configurations were going on which might have the
significant positive effects in the reduction of energy requirements for
carbon capture. Lin and Rochelle [326] developed an Advanced Flash
Stripper (AFS). In the AFS, the stripping stream was eliminated already
by rich solvent bypass stream, so AFS was reported to have a lower
optimum heat of absorption over a simple one [327].
Several novel methods of regeneration process have been reported
[328]. One of them was the pH swing method [329]. Feng et al. [330]
studied the addition of various weak acids (suberic acid, phthalic acid,
oxalic acid) in the regeneration unit and its effect on the release rate of
CO2 and regeneration energy. It was observed that with the addition of
acids, the CO2 release rate increased. Similar results were reported by
Du et al. [331] who conducted the same study adding suberic acid,
phthalic acid, adipic acid and sebacic acid to different absorbents –
MEA, DEA, MDEA. Not many studies have been reported on organic
acids.
A major flipside of using ammonia as an absorbent was the
difficulty of regenerating NH3 from NH4HCO3. In view of the problem,
Huang and Chang et al. [332] added a basic ion-exchange resin to the
product solution to adsorb carbonic acid. The ammonia regeneration
efficiency (ARE) of different resins (A-7, A-21, A-830, IRA-67, IRA-96,
WA-30) were found out. For the recovery of the resin from the product,
the mixture was heated in water and N2 was bubbled to strip CO2. The
sustainability of resin was also examined by running several adsorp-
tion-desorption cycles.
One more method reported was the addition of a highly volatile
immiscible solvent as a stripping carrier. A major contributor to the
heat duty of reboiler was the latent heat of vaporization. Addition of
such a stripping carrier led to a reduction in water vapor generation
thus lowering the operating temperature. The steam in the reboiler also
dilutes CO2 in the regeneration unit, wherein the vaporized solvent
provides the driving force and enhances the release of CO2 from the
absorbent [333]. Frimpong et al. [334] used pentene as a stripping
carrier with 30 wt% MEA as an absorbent. It was found that the effect
was more pronounced at lower temperatures (90 oC). Tobiesen and
Svendsen et al. [335] studied the CC process with the addition of an
inert immiscible component in the desorber. A computational model
developed showed the opposite results, wherein the hydrocarbon added
in the desorber increased the reboiler duty, instead of reducing it.
Owing to limitations in reduction potential and corrosive nature of
absorbents at high temperatures, membranes have been proposed to be
1094
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
used in the regeneration process. Various methods like membrane flash
process [336–338], membrane vacuum generation [339,340], mem-
brane electrolysis [341] have been tested. However, they couldnt be
commercialized owing to huge costs of membranes. Inspite of several
alternatives being tested in the market, none of these are economic
enough to be commercialized for full-scale applications.
A recent case-study [342] suggested a solar assisted post-combus-
tion CC process (SPCC), thereby directly eliminating the expensive
desorber system and reducing the capital expenditure by 15–30%.
However, considerable amount of solar energy would be wasted during
the transfer of heat from radiations to rich solvent. It was therefore,
proposed to set up a solar thermal system, use a solar collector pipe
(parabolic trough pipe) which would directly heat the rich solvent,
disassociating the CO2-solvent bonds. A cost analysis showed that there
could be an economic benefit of $12.2 M when compared to a SPCC.
However, a lot of CC modeling has to be done in integration with the
electric market and RECs fluctuation in prices in this area. Table 6
below summarizes and compares various contactors employed in
absorption based CC with critical design parameters, efficiency, scale,
merits and demerits.
In general, the gas-liquid contactors can be optimized by: contin-
uous and counter-current liquid flow to increase the driving force, i.e.,
concentration difference; membrane so chosen should have a small
mass transfer resistance Membrane contractors, though are compactly
structured and more efficient, but their fabrication is complicated. One
potential way of developing a Janus bio-catalytic membrane is a
superficial single-sided polydopamine deposition method [277].
Fabrication process is currently under investigation.
Contactors play a vital role in the successful deployment of any
technology. In the absorption based CC, there have been various
reactors of different scales employed like conventional packed columns,
bubble and tray columns, rotating packed bed reactor, micro-channel
reactor, membrane contactors etc. The main features that any ideal
contactor system needs to have are high efficiency, scalability, low cost,
high safety, easy operation, provide better contact, low energy require-
ment etc. A lot of current research has been directed from conventional
packed columns towards membrane contactors as we could achieve CC
through dual modes i.e., membrane separation and absorption. This
requires the development of effective membrane materials compatible
with different solvent systems. Also there is a strong need to develop
membranes that are not prone to wetting and fouling that would
degrade their performance. Different membrane modules are being
configured to facilitate this process. Also there are many research
efforts that have gone into the enhancement of process intensification
factor, plug in energy integration and possible flow sheet modifications
to bolster the efficiency, scalability, environmentally benign nature and
more importantly commercial viability [343-348].
4. Thermodynamic and kinetic models
Effective absorption of CO2 in various solvents involves under-
standing of both thermodynamic and chemical nature of the process.
This was known as reactive absorption where reactions occur simulta-
neously with absorptive mass transfer.
The two different approaches to modeling were: (a) equilibrium
models where vapor-liquid equilibrium was assumed at each stage; and
(b) rate-based (non-equilibrium) models which were more accurate
since actual rates of reaction and mass transfer resistances were
considered, because physical equilibrium was difficult to attain. The
former involved a quick mass transfer, hence was applicable to solvents
with high rates of reaction like primary amines. It employed MESH
equations (Material balance, Equilibrium equations, mole fractions
Sum conditions and Heat balance). The latter was more widely
employed where an enhancement factor was incorporated to amend
the equilibrium computations. It was usually the ratio of coefficients of
chemical and physical mass transfer, obtained by either fitting experi-
I. Sreedhar et al.
mental data or theoretically by simplifying models using assumptions.
Maxwell-Stefan equations from kinetic theory of gases were modified to
describe mass transfer resistances [26].
The process fluctuations and other disturbances were the causes for
the development of dynamic modeling. However, this was not always
feasible because of the scarcity of dynamic experimental data, thereby
inaccuracy in validation of models. Hence, steady state models were
often convenient.
4.1. Mechanisms of absorption
To determine the kinetics of chemical absorption of carbon dioxide,
it was important to know the reaction mechanism between carbon
dioxide and the absorbent. With amine absorbents, three main
mechanisms [352,353] were possible:
1. Zwitterion mechanism: First proposed by Caplow and enhanced by
Danckwerts [354], this mechanism demonstrates that the CO2 and
solvent form a zwitterion first which was deprotonated further by
base to form carbamate [355]. This mechanism was generally
applied to primary, secondary and sterically hindered amines.
CO2+R1R2NH↔R1R2NH+COO-
R1R2NH+COO-+B (R1R2NH2OH-+H2O)↔R1R2NCOO-+BH+
2. Ter-molecular mechanism: It proposes that formation of zwitterions
and the process of de-protonation occured simultaneously.
CO2+R1R2NH+B↔R1R2NCOO-+BH+
3. Base-catalysed hydration mechanism: This occured when the absor-
bents were tertiary alkanolamines, which couldn’t directly react with
carbon dioxide. These in turn have a base- catalytic effect on the
process of hydration of carbon dioxide and may lead to amine
dissociation.
CO2+R1R2R3N+H2O↔R1R2R3NH++HCO3-
If the reaction was irreversible and the deprotonation of zwitterion
was the rate determining step (RDS), both (1) and (2) becomes
identical.
Equilibrium constants and rate constants (as a function of tem-
perature) were the fundamental parameters for an absorption model,
as they help determine the liquid phase composition. They could be
derived from either experimental data or correlations for different
solvents, linking them to temperature.
A lot of research has been done on using promoters to enhance the
performance of novel absorbents. These promoters differed greatly in
their physicochemical properties but had some similarities in their
structures. They all had O- or OH groups or they have the capability to
act as Lewis bases with CO2 [156,356–358].
a. Promoter deprotonates to form activated species:
Promoter→Activated promoter+H+
b. CO2 adds to activated specie to form intermediate:
Activated promoter+CO2→Intermediate
c. Promoter regenerates:
Intermediate+H2O→Promoter+HCO3+
For reactions occurring through zwitterion mechanism (instead of
termolecular reaction), step (a) follows step (b).
4.2. Thermodynamic models
An efficient thermodynamic model was required to describe VLE,
which was the basis of any absorption process. Precisely for CO2
absorption, a thermodynamic model was essential for determining the
1095
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
partial pressure of CO2 and quantifying the regeneration energy
required. Over the years, various thermodynamic models have been
developed to determine the partial pressure of CO2 and to enumerate
the regeneration energy. Kent-Eisenberg equilibrium model was the
first model developed in 1976 describing absorption of carbon dioxide
by MEA solvents [359]. Two phase equilibrium was defined using
Henry constants. As the model could not predict the liquid phase
composition using the model, it was enhanced by Krirpiphat and
Tontiwachwuthikul in 1995, determining the solubility of CO2 in AMP
solutions [352].
An applied Gibbs energy model – eNRTL method (Electrolyte Non
Random-Two Liquid) developed by Chen et al. [360,361] correlates
mean ionic activity coefficients of mixed solvent electrolyte systems. It
was based on two main assumptions of considering high repulsive
forces between ions and local electro-neutrality. An ASPEN Plus
package includes this. In a study [362], it was used for modeling the
VLE data for absorption of CO2 in Piperazine (PZ) (2–8 mass%)
activated aqueous solutions of MDEA (30–50 mass%). A physical
solvent called sulfolane was added to the above hybrid solution and
studied over the temperature range of 308–328 K and CO2 partial
pressure in the range of 1–1400 kPa. The experimental results and
results obtained from the model were in good agreement. The liquid
phase speciation of CO2 loaded aqueous MDEA, (MDEA+PZ), (MDEA
+PZ+sulfolane) were different [362]. Bicarbonate being the major
product showed that the carbamate reversion of PZ-carbamate was
favored. Results also further showed that addition of sulfolane had a
positive outcome on the solubility of CO2 in MDEA based solvents but
it differed with variation in composition and temperature. Therefore,
the operating conditions had to be tuned to get the best solubility.
Several versions of Gibbs energy based model, like, the EoS/GE
model were used determine the Gibbs energy through equations of
state. Chunxi and Furst developed the first model which considered
electrostatic interactions in liquid phase and hence included a specially
adapted equation of state [363]. Through yet another model -
UNIQUAC, activity coefficients were calculated using the excess
Gibbs free energy. Based on the same model, UNIFAC assumed
molecules as a set of groups, hence getting data for segments would
essentially be used to predict the activity of any molecule which
UNIQUAC does not [364].
Arshad et al. [365] used the Extended UNIQUAC framework and
developed a model describing the thermodynamics of H2O-DEEA-CO2
(DEEA – 2-(diethylamino)ethanol), H2O-MAPA-CO2 (MAPA – 3-
(methylamino)propylamine) and H2O-DEEA-MAPA-CO2 (liquid-liquid
phase split) systems. Ninety four model parameters and six thermo-
dynamic properties were evaluated from around 1500 sets of experi-
mental data available including vapor-liquid equilibrium, liquid-liquid
equilibrium, solid-liquid equilibrium, heat of absorption, vapor pres-
sure, excess enthalpy. The parameters calculated were valid across
temperatures from −25 to 200 oC, CO2 partial pressure from 0 to
945 kPa and concentrations of MAPA, DEEA, CO2 upto
23,131,33 mol kg-1 H2O-1 respectively.
Instead of using Gibbs energy, SAFT (Statistical Association Fluid
Theory) model determined Helmholtz excess energy. These equations
were clustered in the mean field theories. Hence, when the local
properties of the fluid have big fluctuations, this led to lower accuracy
[366].
Table 7 [367] captures the thermodynamic models researched so
far.
4.3. Modeling and simulation studies
Following is a review of the models developed using the reaction
kinetics and thermodynamic models involved in the process of CC.
Software like gPROMS, Aspen Plus, Aspen HYSYS, ProMAX, Tsweet,
MATLAB with ODE solvers and Simulink have been engaged to
understand modeling. To understand the hydrodynamic phenomenon
I. Sreedhar et al. Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

Table 7
Summary of the thermodynamic models.

VLE Models Remarks

Non rigorous models Kent Eisenberg [352] -non-idealities processed as equilibrium constants K
-fugacity and activity coeff=1
-valid for loading between 0.2 and 0.7
-not valid for acid gases

Rigorous models – Excess Gibbs Pitzer [367] -activity coeff=f(ion concentration)

Energy -single solvent or dilute electrolyte systems
-complicated calculations of activity coeffs
Deshmukh-Mather [368] -based on Guggenheim eq. and Debye-Huckel theory
-fugacity and activity coeffs easily predicted
-defines relation between CO2 loading and carbamate formations
-fails to converge sometimes
e-NRTL [360] -solvents: alkanolamines and water
-non-ideality of systems defined from pure to saturated solutions
-ionic species infinitely diluted in water
-correlates mean ionic activity coeffs of mixed electrolyte systems
Extended UNIQUAC model [365] -combines Debye-Huckel theory with original UNIQUAC model
-simpler than eNRTL
-calculates VLE and thermal properties of non-electrolytes or electrolytes

Equation of State (EoS) based model
Electrolyte EoS [363]
Extended SAFT [366]
-Helmholtz energy: electrolyte part for ions and non-electrolyte for molecules
-one-parameter model
-suitable for high pressures
-accounts for contributions of hard sphere repulsive force, chain formation, dispersion forces,
association to Helmholtz function
-no need of explicit description of chemical equilibrium and its dependence on temperature and
density

in absorbers, CFD based models were employed. Luo et al. and Aboudheir et al. [377,378] in their works, this work used
Viscosity of the fluid has an impact on the rate of mass transfer. the termolecular reaction mechanism and takes into account the
Zhang and Yu [369] proposed calculating viscosity from equations reversibility of reaction, which was very much plausible at increased
based on the scaling principle and free volume theory. It deemed valid concentrations of MEA or high CO2 loadings. The reaction was written
for pure hydrocarbons over various temperatures and pressures. as:
Extending it over solutions by combining the chain fluid model of
CO2+MEA +Base←→MEACOO-+BaseH+
diffusion [370,371] with Stokes-Einstein equation [372] could be a
possibility. But considering the vast variety of ions in such solutions, it Both the models were validated. The concentration based model
was fairly difficult to develop such equations. Hence, Weiland equation gave larger deviations for higher concentrations (7–9 M MEA) whereas
has been used in most models including absorbents like DEAE, [N1111] the activity based model held well for both high and low concentra-
[Gly] as it successfully took into account of the interdependence tions. The models were valid over temperature ranges of 293–343 K
between temperature, pressure, mass fraction of amine and CO2 and CO2 loading range of 0–0.5 mol CO2/mol amine. The models were
loading capacities. able to predict the absorption rates, desorption rates with accuracy.
w1 w1 Also, the termolecular reaction mechanism very aptly describes the
ηmix = η1 + η2 CO2-MEA reaction.
w1+w2 (w1+w2)
Kim et al. [379] used the carbamate formation reaction, i.e., an
ηi [(aiw + bi ) T + (ciW + di )] w additional amine model to study the effect of blends on CO2 absorption.
=exp ( *f ( α , w )
ηwater T2 Methods like Density Functional Theory (DFT), Own n-layered
Integrated molecular orbital and molecular mechanics (ONIOM) were
f (α , w )= α (eiw+fiT +gi )+1
used to model MEA based solvents [380,381]. The optimal geometry of
where η water is the viscosity of pure water , reactants, the products and the transition state were obtained from
Becke’s 3 parameterized Lee-Yang-Parr exchange (B3LYP) [382], M06-
w = w1 + w2, ai , bi , ci , di , ei , fi , gi are adjustable parameters.
2X functionals [383,384] and MP2 [385]. Water or amines lowered the
Zareikordshouli et al. [167] in his study of using ILs like [bmim] activation energy of the solution by improving the intermolecular
[Tf2N] and [Amim][Tf2N] as absorbents, used the Vogel-Fulcher- interaction and removing proton from amine. AMP and DEA were
Tammann (VFT) equation [373,374] to study the temperature depen- found to reduce the performance capacity of MEA. The results were
dence of viscosity. validated using experimental data.
Chu et al. [386] modeled CC in a packed column using MEA as an
⎛ B ⎞
η = A exp ⎜ ⎟ absorbent. MEA was taken as pseudo-continuous in the pseudo-single-
⎝T − T0 ⎠ liquid model [387]. Heat transfer, mass and momentum balance
where A, B, T 0 are adjustable parameters equations, fluid dynamics related to gas and liquid phases were
simulated based on the Representative Elementary Volume (REV)
The characteristic exponent φ in the equation proposed by Ghatee model. The simulated model was verified against experiments per-
et al. [375] correlating the surface tension and viscosity was found out formed by Tontiwachwuthikul et al. [388]. Good agreement was
to be 0.3 in this case. reported showing the reliability of the models and assumptions used.
Putta et al. [376] developed and validated two new kinetic models The porosity and surface area per unit volume of the packing material
based on concentration and activity, using the penetration theory and should be as high as possible (like NO. 8). The optimal operating
global optimization. Unlike the assumptions of irreversibility used by

1096
I. Sreedhar et al.
pressure was found out to be 1 atm and the optimal height of column
should be 8 m.
Rode [389] had developed a one-dimensional, multicomponent
adiabatic model for CO2 absorption on MEA using hollow fiber
membrane contractors. It was validated using both experimental data
at lab-scale and pilot-scale and the results were compared to an
isothermal model to study the influence of heat release on performance
of contractor. 60% deviations were obtained indicating the necessity of
adiabatic modeling in industry. However, as capillary condensation was
imminent because of rapid heat transfer, so wetting still remains a
problem in microporous and composite membranes. Mass transfer
coefficients in gas phase [390], membrane phase [391,392], and liquid
phase [393] have been calculated and correlations have been devel-
oped.
Mores et al. [394] assumed a two film model, ideality in mixtures,
neglected solvent vaporization and had developed an equilibrium stage
model simulating carbon dioxide absorption by MEA solvents and used
it to optimize the operating conditions comparing the CO2 removal
efficiencies. Activity coefficients used to calculate chemical equilibrium
constants are adopted from Aboudheir et al. [395]. Stage efficiency
correlations as a function of an enhancement factor and physical
properties of liquid were used to express the thermodynamic non-
idealities. Equations developed by Onda et al. [396], Wilson [397],
Bravo and Fair [398] were used to calculate the effective mass transfer
area. Finally constraints were applied to calculate Henry’s constants,
enthalpies, viscosities, vapor pressures, reaction heats, surface tension
and fugacity coefficients among many other variables. A desorber
model considering the same constraints was also developed.
Simulation results by Austgen and Rochelle [399] and Aboudheir
et al. [400] validated this equilibrium model. The absorber column
was validated using the experimental measurements from pilot plant
and simulation using HYSYS process simulator. According to sensitiv-
ity studies, of the three approaches discussed above (Onda et al.,
Wilson, Bravo and Fair), the correlations for gas and liquid-film mass
transfer coefficients were better predicted using the Onda et al. [396]
For temperature and vapor flow rate, correlations developed by Bravo
were in good agreement with the experimental results. The stripper
configuration was also validated using experimental data from
Tobienssen et al. [400], Freguia and Rochelle [401] and Dugas [402].
It was observed that the optimal value while minimizing CO2 recovery/
amine flow rate decreases as the height of the column increases.
Khan et al. [328] developed a typical rate-based steady state model
using the two film theory based on MEA as a solvent defining the
theoretical column stages. The overall absorption rate was obtained
from Levenspiel correlation corrected by an enhancement factor [403].
Again, coefficients for gas and liquid film mass transfer and interfacial
area for packings are obtained from Onda [396]. Several assumptions
like adiabatic operation, plug flow, constant flow rates, negligible axial
diffusion, constant pressure and negligible evaporation of solvent etc.
were made in developing the model. The results were validated using
Tontiwachwuthikul pilot plant which was in good agreement with the
predicted values
[53]. The concentration of CO2 for the bottom part of the
column was under predicted which was important because major
absorption was believed to take place in the bottom part. Because of
the neglected solvent evaporation, the solvent temperature also de-
viated from the experiment. These results were also validated against
the Aroonwilas pilot plant [404] with structured packing. They were in
good agreement. Gas liquid-film mass transfer coefficients provided
better predictions than the correlations given by Bravo et al. and
Higbie’s penetration theory [405].
A rigorous rate-based model considering both kinetically controlled
and instantaneous reactions was introduced by Kucka [406] using MEA
as an absorbent. Where most models used an enhancement factor, this
model calculated interfacial flux explicitly and defined heat and mass
transfer coefficients based on film theory. Mass action law based on
activities of species is applied to consider the instantaneous reactions.
1097
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
Reaction constants and equilibrium constants were derived via activity
based correlations. Electrolyte influence was also included by relating
film equations with diffusion and reaction kinetics. Assuming local
electro-neutrality and neglecting interactions through diffusion, mass
transfer across the film is guided by Maxwell-Stefan equations. Thus,
film component and heat fluxes were quantified. For the liquid phase,
eNRTL method described the thermodynamic non-idealities whereas
for the gas phase, SRK equation of state was used. Two data sets from
Tontiwachwuthikal pilot plant and one from an industrial plant were
used to validate the model. Temperature profile for the liquid phase
and composition in gas phase were plotted along column, which
satisfactorily matched with the experimental data. Compared to other
models where enhancement factor was used, these results were in
better agreement with the industrial data.
Another rate based steady state model was developed by Al-Baghli
et al. [407] studying the solubility of CO2 and H2S in MEA and DEA
solvents. Several assumptions like two film theory, stages approach,
ideal phases, and reaction occurring entirely in liquid film, constant
temperature and composition of gas and liquid leaving the tray and on
the tray were made. Interface equilibrium was defined by Henry’s law.
An enhancement factor used to describe reaction kinetics was calcu-
lated from a system of five non-linear differential equations defining
reactions and diffusion in liquid film. Kent-Eisenberg model was used
to get concentration of species in bulk of liquid. Danckwerts approach
[408] was used to calculate reaction rates. Economopoulos algorithm
was used to determine the weir height, diameter of tower, tray
hydraulics, tray spacing and pressure drop across them. MEA required
less number of stages than that of DEA. For CO2, the enhancement
factor for both solvents increased marginally at the top of tower. For
H2S, it increased sharply until the concentration reached zero and
decreased henceforth towards the top stages.
The first dynamic model, developed by Kvamsdal et al. [409] in
2008 simulated the process of absorption of CO2 in MEA on gPROMS
tool and was validated against a steady state data from University of
Texas (UT) at Austin. Subsequently, it was validated against both
steady state and transient state (pilot plant set up in NTNU/SINTEF).
Several assumptions for wetted column including plug flow, linear
pressure drop, two film theory, ideal gas phase, zero accumulation,
axial differential energy and mass balances, thermal equilibrium
between packing and liquid, condensation of water only at interface
and walls and effective contact area equal to packing material specific
area etc. were made in the development of the model. The developed
tool tests two scenarios in dynamic analysis: change in flue gas and
liquid flow rate during start up and load reduction cases. Mass balances
with transfer coefficients given by Onda et al. [396] and heat balances
were solved for bulk phases. An enhancement factor was used.
Coefficients of reaction rates were derived based on Freguia and
Rochelle [401], corrected by activity coefficients. To determine the free
carbon dioxide mole fraction at equilibrium and concentration of free
MEA, an equilibrium speciation model [412] was used. Equilibrium
pressure was obtained through activity coefficients of MEA and CO2
which were derived from Wilson equation.
Based on the work of Kvamsdal et al. [409], Svendsen et al. [410]
Harun [411] proposed another dynamic model for absorption of
carbon dioxide on MEA. Two film theory was applied and the column
was discretized and energy and mass balance equations were applied
on each component in each stage. Different equations were described
for the enhancement factor of absorber and stripper. A simple
equilibrium model was used to model the reboiler, where compositions
in liquid and vapor phases were calculated at various temperatures and
pressures using the P-T flash calculations. Heat exchangers describing
exchange between hot lean amine from the stripper and cold rich
amine from the absorber were simulated on a mathematical model. It
was validated against steady state data. The absorber model was
validated using experimental results obtained from Lawal et al.
[412]. Results showed that the model under predicted the tempera-
I. Sreedhar et al.
tures along the column. The stripper model was validated against
simulation results obtained from Aspen Plus RADFRAC model. Results
showed that the model over predicted the temperatures along the
column. Satisfactory results were obtained with Aspen Plus for the heat
exchanger and reboiler. A partial load reduction scenario was opted for
dynamic simulations. It was concluded that with a linear decrease in
flue gas flow rate, the liquid to gas ratio (L/G) increased. The carbon
dioxide loading reduced with time.
The conventional rate-based screening model was used to deter-
mine the mass transfer performances of various absorbents, thereby
helping decide the best absorbent. The equilibrium based screening
methods have been applied to single solvents and blends like AEEA
[413], MDEA-PZ [414]. Based on these models, Aronu et al. [415–417]
developed a quicker method for comparison between TEPA, MAPA,
TMBPA and their blends. The absorption and desorption processes
were carried out at 40 oC and 80 oC respectively with CO2 loading
varying from 0% to 90% (desorption initiated again by raising
temperature). To decide on the optimal absorbent amongst the three,
Aronu et al. used cyclic capacity of absorbents and loading difference
between lean and rich solutions. Chowdhury et al. [418,419] developed
a similar model based on cyclic capacity and heat of reaction. The
processes were carried out at 40 oC and 70 oC. CO2 loadings from zero
to equilibrium value would require an infinite amount of energy and
height of desorption tower. Even though the absorption rate was high
when the loading was close to zero, the reboiler duty was also extremely
high. At equilibrium, the reboiler duty becomes low but the absorption
rate became very low. Therefore, considering the absorption rate and
reboiler duty, the process was done in the central area of the range of
loading. These models could be used for selecting solvent or evaluating
the mass transfer performance of wetted tower or packed tower.
However, it failed to evaluate the overall CC performance [420–423].
Luo et al. [424] developed a multiple screening method and evaluated
the repeatability and stability of solvent. It has a combined CO2
absorption and desorption process being carried out at 40 oC and
80 oC respectively. This screening process was executed on 5 solvents
and the CO2 capture performance was AMP > DEA > DEEA > MEA >
MDEA. This result was also validated using an ASPEN Plus model in a
pilot plant scale. 30 wt% AMP presented a 41.5% higher capture than
30 wt% MEA. The screening method was also implemented on MEA/
DEEA blends of various ratios. The optimal molar ratio was found to be
2.5:2.5 which gave a performance of 31.8% higher than 30 wt% MEA
solutions. However, the results obtained from this model were semi-
quantitative.
A predictive tool for MDEA gas scrubbing process was developed by
modifying RATEFRAC (Aspen Plus) by Bolhar-Nordenkampf et al.
[425]. Results were also validated with data from a commercial MDEA
plant based in Germany. Calculations were done on each tray and two
film theory was applied. Henry’s law was used to get the mole fraction
at the interface. Overall energy and mass transfer rates were derived
using Khrishnamurthy equations. Maxwell-Stefan equations were used
to determine the two phase mass transfer coefficients. It predicted the
H2S values better than the equilibrium simulation. The same prediction
model based on equilibrium stage model was also developed for
desorption equilibrium because reactions occured at high temperatures
which had equilibrium like reaction rates. An enhancement factor by
Danckwerts [354] was used to describe the reaction kinetics and hence
relate the physical mass transfer coefficients to actual transfer coeffi-
cients.
Gilassi et al. [426] developed a mass transfer model in COMSOL for
CC in HFMC using MEA, DEA, MDEA as absorbents. It took into
account laminar flow conditions, axial and radial diffusions across
membranes and convention in shell and tube sides. The nonlinear
PDEs were solved using finite element method in CFD. MEA showed
the best performance amongst the three.
Zhang et al. [427] developed a 2D mathematical model for
membrane gas absorption (MGA) from N2/CO2 mixture using EDA,
1098
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
DEA, MEA and PZ as absorbents. Certain assumptions were made for
the calculations: ideal gas behaviors, non-wetting of membrane,
isothermal conditions, steady-state condition, Henry’s law at gas-liquid
interface. The velocity profile within the hollow fibers were given by
Seader et al. [428], while for the shell side were given by Happel’s free
surface theory [429]. Applying boundary conditions, the model was
solved using COMSOL Multiphysics. A finite element method (FEM)
incorporating error controls and adaptive meshing techniques.
PARDISO solver was used which was high-performing, memory
efficient, robust and easy to use for solving large sparse symmetric
linear equations on shared and distributed memory processors. This
could even be used for other acid gases removal from biogas, natural
gas and syngas. A similar study was done by Saidi et al. [430] using 4-
diethylamino-2-butanol (DEAB) in a gas-liquid HFMC. This was
validated and in good agreement with already available experimental
data in literature for MEA, DEA and MDEA [431,432]. Results
obtained were compared with MEA, DEA, TEA, MDEA and the effect
of parameters like concentration, liquid and gas flow rates, CO2 partial
pressure, temperature etc. was studied. Sensitivity analysis revealed
that significant mass transfer occurs in the gas phase. The CO2
absorption was at par with MEA and higher than DEA, TEA and
MDEA. Increasing temperature, liquid flow rate, concentration of
absorbent and decreasing gas flow rate enhances CO2 removal.
Norouzbahari et al. [433] developed a non-equilibrium rate based
stage model using PZ as an absorbent. Most rate based models
assumed an ideal plug flow regime (no axial/longitudinal mixing) in
the column. But the non-ideality affects the separation and the effects
were more pronounced in industrial packed columns. This led to
stagnation of liquid within voids and hence a decrease in the driving
force for mass transfer. This study based on the two film theory,
includes the effect of axial mixing in the component mass balance PDE
in liquid phase and used the most accurate kinetics, mass transfer and
hydraulic correlations (Billet and Schultes [434]). The thermodynamic
modeling of the above solution involves the γ-φ approach to calculate
various ionic and molecular concentrations of species at the interface
and in the bulk. The modified Pitzer’s equation [435] was used to
calculate the activity coefficients in the liquid electrolyte solution. The
VLE condition was applied to H2O and CO2 using extended Raoult’s
law and Henry’s law. PZ has a very low vapor pressure, hence, was
ignored. Artificial Neural Network (AAN) model [436] was applied to
accurately obtain the missing initial values of PZ concentration, inter-
face CO2 loading at a given temperature and CO2 partial pressure.
Finally, Newton Raphson method was used to solve for the non-linear
VLE equation. The numerical method-of-lines (MOL) was used to
transform the kinetics partial differential and non-linear algebraic
equations to Ordinary differential and non-linear algebraic equations
with only t as an independent variable. 4th order Runge-Kutta method
was then applied [437]. The results were validated against experi-
mental data with an average absolute relative deviation percentage
(AARD) of 6.59%. The consideration of non-idealities reduced the
AARD from 6.59% to 2%.
Gabrielsen et al. [438] developed a rate based model predicting the
solubility of CO2 in AMP solutions and heat of absorption as a function
of temperature and loading. Several assumptions including two film
theory, constant temperature between liquid bulk and interface, ideal,
liquid phase ideality and adiabatic conditions neglecting solvent
evaporation and axial dispersion were made. Mass transfer coefficients
and gas-liquid interfacial area were obtained from Billet and Schultes
[439] correlations. Enhancement factor proposed by Krevelen and
Hoftijzer [440] was used. Partial pressure of carbon dioxide as a
function of carbon dioxide loading was calculated using equilibrium
constants [441]. These values were validated against experiments by
Tontiwachwuthikul et al. Even though no experimental data was
available to validate, but still heat of absorption for CO2 in AMP was
calculated using the Gibbs-Helmholtz equation [441]. AMP, being a
sterically hindered amine, followed the zwitterion mechanism in
I. Sreedhar et al.
reaction with CO2. Results were in agreement with experimental data
except for a slight over prediction of temperature at the bottom of
column.
Yang et al. [442] simulated thirteen bubble column reactors with
different geometries, varying the orifice size, orifice number, operating
conditions etc., using ammonia as a solvent. The chemical equilibrium
was modeled based on studies of Darde et al. [443] while the mass
transfer and hydrodynamics models were modeled using the
Representative Elementary Volume method (REV) [444–446].
Ammonia slip was also modeled which was verified against experi-
ments conducted by Ma et al. [447]. It was reported that an increase in
the orifice size led to an increase in CO2 transfer and reduction in the
mass transfer coefficient for ammonia escape, reducing the overall CO2
capture efficiency. An increase in the orifice number leads to an
increase in CO2 mass transfer, ammonia escape and an overall
enhancement of CO2 capture efficiency. The optimal pressure was
found to be the atmosphere pressure. The optimal height-to-diameter
ratio was reported to be 5.76.
Khan et al. [129] performed experiments on AMP-PZ blends and
the effectiveness of using PZ as a promoter (2–10 wt% of PZ). The
experimental data thus obtained was plotted in MATLAB using the
Response Surface Methodology (RSM) model for non-linear fitting. All
5 variables (amine concentration, liquid flow rate, gas flow rate, partial
pressure of CO2 and temperature) were normalized by their highest
values and a quadratic equation was used for fitting. Excellent
agreement was observed. Guo and Zhang [26,448] and Dash et al.
[58], used the rate fraction in ASPEN Plus for studying ammonia and
AMP-PZ blends. Sarkar et al. [449], explored the design of a fluidized
bed regenerator using multiphase CFD employing solid absorbents
under well mixed flow conditions and plug flow. Chikukwa et al. [450],
studied the dynamic behavior developing models based on carbon
dioxide absorption and stripping using MEA as solvent by varying flue
gas composition and flow rate, capture level of CO2, reboiler duty, L/G
ratio etc. Though the Pandya model [451] of the continuous differential
reactor (CDC) formulated in 1983 had been proven to be very useful in
modeling and simulation of packed bed reactors, it had its short-
comings. It was based on some inconsistent assumptions and simpli-
fications. Mario Llano-Restrepo [452] developed new equations of
material and energy balances for liquid phase and came up with a
recast version of Pandya’s model. Combination of this with the most
suitable correlations for kinetic constant, liquid-phase speciation,
enhancement factor, heat capacity, density, viscosity, effective inter-
facial area, heat of reactive absorption, surface tension of CO2-loaded
aqueous MEA solutions provides an accurate prediction of temperature
profiles [451]. However, correlations for heat of reactive absorption are
based on experimental data and were available for very few pressures
and temperatures, extrapolating them might be dangerous.
For blends of PZ/tertiary amines (IMI, HEMor, DMAEE, HMPD,
HEPD, DEA-PDL, Tropine etc.), a very generic Aspen Plus model was
developed by Du et al. [277] based on a previous Aspen Plus model of
PZ/MDEA using eNRTL thermodynamic framework. It predicted the
VLE data of CO2 based on the pKa of the tertiary amine. The effects of
polarity and pKa on the solubility on CO2 were studied. For same pKa
and polarity, hindered amines showed similar VLE results.
Puxty et al. [453] developed a steady state model in MATLAB
illustrating CO2 mass transfer in single systems (MEA, PZ and NH3) or
mixtures (PZ-AMP, PZ-NH3, PZ being used as promoter). Mixture
results were predicted using results obtained from single systems. This
model was validated against experimental results from a wetted wall
column. Several assumptions including two film theory, neglecting
solvent evaporation and axial dispersion, constant physical properties
were made. Kinetics varied from fast to kinetically controlled reactions.
Equilibrium and rate constants were corrected by activity coefficients.
System of PDEs was defined by 2nd Fick’s law and diffusion flux for
different column heights was derived using 1st Fick’s law. For single
systems, predicted and experimental data were in satisfactory agree-
1099
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
ment. For ammonia particularly, increase in ammonia led to high
volatility and uncertainty in concentration of ammonia concentration.
For mixtures, an excellent agreement was found between predicted and
experimental data proving that the behavior of mixtures could be
predicted by individual systems.
Sherman et al. [118] modeled CO2 absorption in 2PE to validate the
importance of hindered amine carbamate reaction in the reaction. The
model was developed in Aspen Plus V8.4 based on the asymmetric
eNRTL model for liquid phase and ESRK model for vapor phase. AMP
was chosen as it has similar carbamate stability (2PE has a higher
carbamate stability). It was concluded that the rapid mass transfer in
hindered amines occured because of carbamate formation. The rate of
carbamate formation was unaffected by stearic hindrance and is
calculated using Bronsted correlation (hindered amines react in the
same manner as unhindered amines).
Wang et al. [454] developed a multiscale reaction model of CC
taking place in a circulating fluidized bed with K2CO3 as an absorbent.
The cluster effect was taken into account while developing the reaction
kinetics. It was assumed that the reaction rates vary across phases –
dense, interface and dilute. A cluster structured drag model was
employed to provide local parameters for the reaction. This model
performed better than the conventional ones. It was reported that the
non-uniform distribution of solid sorbents, their resident time and
temperature fields caused due to clusters affect the absorption rate.
Savage et al. [455] reported the mechanisms of reactions involved
during absorption and desorption in 25 wt% K2CO3. Several promoters
have been used to accelerate the process of CC using K2CO3 as an
absorbent. Reaction kinetics for promoters like MEA, MDEA, DEA
[456,457], boric acid [458], PZ [459,460], arginine [461], pyrovana-
date [462], sodium [463] have been studied extensively over the years.
Limited studies were reported on EAE which was considered as the
best promoter (APCE solvent) until the kinetics of the reactions using a
fall in pressure method in a glass stirred cell reactor were examined.
Zwitterion mechanism was assumed. The overall reaction was deemed
to be of second order (first order with respect to EAE and CO2). The
Henry’s constant decreased by 5% and 31% when the EAE concentra-
tions in the solvent went from 0.6 to 2kmol/m3 and temperature went
from 303 to 318 K respectively. The above mentioned led to an increase
in rate of CO2 absorption by 35.1% and 47.6% respectively. However
the experiments above were conducted without any CO2 loading.
Du et al. [464] simulated the performance of 2 ILs - [bmim][BF4]
and [bmim][PF6]. Thermodynamic data like density, heat capacity and
vapor pressure were calculated. Simulation results obtained from
Aspen Plus using the e-NRTL method. VLE was modeled using RK
and NRTL model. The VLE data calculated was verified with the
experimental data and shown to be in good agreement. ILs had
negligible solvent loss – 0.299 g/tCO2 for [bmim][BF4] and 0.391 g/
tCO2 for [bmim][PF6], whereas a value of 178 g/tCO2 for MEA. The
energy consumptions were found to be 26.7% and 24.8% lower than
MEA-based process respectively.
Bazargani et al. [165] developed several models overviewing CO2
absorption in various ILs like [bmim][BF4], [bmim][PF6], [bmim]
[MeSO4],[bmim][Tf2N]. EoS based thermodynamic model, i.e.,
Perturbed Hard Sphere Chain Equation of State (PHSC) was applied.
From the regression of PVT data of CO2 and ILs, 3 scaling constants –
hard-sphere diameter, number of segments per molecule and non-
bonded interaction energy, were calculated. On fitting the CO2 solubi-
lity data, the one binary interaction parameter was computed. It was
concluded that PHSC EoS was effective in calculating CO2 solubility in
the ILs except at high pressures. Also, it increased with increase in
pressure and decrease in temperature.
Wu et al. [465] developed a model to evaluate the performance of
CC with 1-butyl-3-methylimidazolium glycinate ([C4mim][Gly]) as an
absorbent in a double stirred cell reactor. The density was measured
using pycnometer and viscosity by Ubbelohde meter. The effect of
temperature on diffusion coefficient [466,467] and physical solubility
I. Sreedhar et al.
[468] was taken into account. Based on values in literature [469], the
gas film and liquid film mass transfer coefficients [470–474] were
calculated at varying temperatures from 293 to 323 K. Its absorption
rate was found to be higher than DEA but lower than MEA. High
temperature enabled better absorption and desorption. The
Enhancement factor E was found to be linearly dependent on square
root of absorbent concentration and equations for reaction rate
constant were devised.
Pelayo et al. [475] conducted an economic analysis of bio-methane
production costs using 3 ionic liquids – 1-ethyl-3-methylimidazolium
bis[(trifluoroethyl)sulfonyl]imide ([C2MIm][Tf2N]), 1-hexyl-3-methyli-
midazolium bis[(trifluoromethyl)sulfonyl]imide ([C6MIm][Tf2N]) and
trihexyl(tetradecyl)phosphonium bis[(trifluoromethyl)sulfonyl]imide
([P66614][Tf2N] on Aspen Plus and Aspen Process Economic Analyzer
(APEA). Since, ILs were not present in the Aspen Plus data base, the
COSMO-SAC (Conductor-like Screening Model with Segment Activity
Coefficient) property method [476] was used which unlike UNIFAC or
UNIQUAC, assumed phase equilibrium on individual atoms instead of
big functional groups. The equipment described includes a multistage
compressor for compression of biogas, a packed absorption column for
CO2 absorption, a flash evaporator for regeneration of solvent, a
centrifugal pump for recirculation of solvent, a pre-absorber cooler
and a turbine for recovery of electricity. Inspite of having the lowest
CO2 absorption capacity, [C2MIm][Tf2N] gave the highest plant
efficiency and lower costs of production. It needed the lowest invest-
ment and operating cost and the highest production rate. Efficiency
varied from 71% to 86% and costs varied from $9.2–11.3 GJ-1 (LHV).
The optimum pressure of plant operation was found to be 20 bar.
Increasing it led to increase in absorption capacity as well as operating
costs. These costs estimated were a 40–51% higher than a same scale
MEA based CC.
In a study by Alper et al. [477], BOLs were developed. Kinetic
parameters were estimated in terms of pseudo-first order homogenous
reaction between CO2 and BOLs in 1-hexanol by a stopped-flow
method with detection of conductivity. An amidine 1,5-diazabicy-
clo[4.3.0]non-5-ene (DBN) and guanidines 1,5,7-triazabicyclo[4.4.0]
dec-5-ene (TBD) and 2-tert-butyl-1,1,3,3-tetramethylguanidine
(BTMG) were developed and their CC properties were estimated in a
temperature range of 288–308 K. A modified termolecular reaction
mechanism between BOLs and a linear alcohol was used to justify the
experimental kinetics of the model. Additionally, quantum chemical
calculations were performed using B3LYP, CCSD, MP2 methods [478–
482] to get the energetic and structural framework of the single step
termolecular reaction. Computational part was carried out using
Gaussian 09 quantum chemistry software package and molecular
visualizations were carried out using GaussView 5.0.9 [483,484].
Limited research was reported on developing mathematical models
for optimization of CO2 absorption process in hollow fiber contractors.
Faiz and Al-Marzouqi [476] used MEA as an absorbent and developed
a 2-dimensional mathematical model based on continuity equation for
removal of CO2 and H2S from gas streams in HFCs. Kieffer et al. [485]
simulated mass transfer in a liquid-liquid membrane contractor but for
laminar conditions. They developed a 2-dimensional mass transfer
model which helped predict concentrations in the reactor. Rezakazemi
et al. [486] studied the absorption process of CO2 in HFCs using
[N1111][Gly] as absorbent. They divided a single fiber of HFC into three
distinct zones: tube, shell and membrane and equations were derived
for every zone. These mass transfer and momentum equations were
then solved on COMSOL Multiphysics package (version 3.5). Finite
Element Method was employed for discretization and UMFPACK was
set as the solver [487–489]. The 2D model (axial and radial directions)
was validated against experimental data obtained from past studies
[490]. It was reported that increasing liquid velocity and reducing gas
flow rate enhanced the absorption process.
Another rate based model developed by Ghaemi et al. [491]
simulated CO2 absorption in partially carbonated ammonia aqueous
1100
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
solutions. Several assumptions like two film theory, ideality in mixtures
and 100 stages were assumed. Axial temperature profiles, energy
balances, heat transfer rates were determined. Considering unsteady
state heat and mass transfer, species concentration was calculated
using initial boundary value. Interfacial diffusion was defined by
Maxwell-Stefan equations. Interfacial compositions were calculated
from equations based on semi-empirical Pitzer equation. Rate con-
stants (temperature dependent) were obtained from Sharma and
Danckwerts [408]. Relations by Onda [396] defined the mass transfer
coefficients. This model was validated with the experimental data
obtained from steady state measurements at a pilot plant. Dynamic
models were solved by both numerical and analytical methods. It was
observed that the use of enhancement factor led to higher deviations
for many parameters measured.
A general model for all amine based absorbents was developed by
Jayarathna et al. [492]. Four different sub-models were developed for
absorber, reboiler, stripper and condenser. Mass and energy balances
were considered same for all columns, changing only the boundary
conditions. Several assumptions like constant flow rates, neglected
interfacial mass transfer of oxygen and nitrogen, linear pressure drop,
ideal phases, reactions in the liquid film only, adiabatic system, each
control volume of each phase behaved like a continuous stirred tank
etc. were made. Condenser was modeled by adding a reflux drum to it.
Reboiler was modeled by a fixed fraction of vapor to feed and the heat
duty was calculated. Mass transfer was considered to occur as a result
of concentration differences. Henry’s law coefficients from Kent-
Eisenberg model were used to determine the concentrations at
equilibrium. The arte coefficients for forward reaction were derived
from correlations given by Jamal et al. [493], and Freguia and Rochelle
[401] thus estimating the overall rate. This model was validated against
experimental data [402,409,494]. Various parameters like CO2 load-
ings and removal efficiency, heat duty of reboiler, flow rates and
temperature profiles were calculated and found to be in good agree-
ment with experimental measurements. For dynamic cases, numerous
cases of disturbances were validated against steady state data.
Based on a steady state study by Kucka et al. [406], Kenig et al.
[495] developed a dynamic model pertaining to absorption of CO2 on
aqueous NH3 solutions. For bulk phases, liquid hold up was considered
in mass balances. For film region, a steady state mass transfer could be
assumed. Assuming local electro-neutrality, the electric driving force
and diffusion using Nerst Planck equations was used to describe the
effect of electrolytes. Interfacial fluxes were defined by both Maxwell-
Stefan equations (no electrolyte effect considered) and the Nerst Planck
equations. Fluxes obtained from diffusion differ insignificantly from
those obtained by Maxwell-Stefan equations. This model was validated
against experimental results from ammonia hydrogen sulphide circula-
tion scrubbing (ASCS) process in Germany. The rates of a significant
reaction between CO2 and hydroxyl ions were defined by Astarita,
Savage and Bisio [496]. The optimal column height was investigated
and axial concentration profiles were plotted and validated. Dynamic
simulations have a growth in ammonia concentration in the feed case
suddenly. Concentration profiles were plotted as a function of time.
Results showed that the selectivity in chemisorption reduced on
increasing concentration of solvent.
Considering certain reasonable assumptions, Gao et al. [497]
developed a mathematical model estimating the coefficients of mass
transfer. The model was CO2/NaOH absorption process occurring in a
RPB with wire mesh packing. The simulation was run on MATLAB and
the simulation results were satisfactorily validated using self-experi-
mentations and previously published work. Results were as follows:
Temperature positively affected reaction rate constant, liquid mass
transfer coefficient and diffusion coefficient; Parameters of tempera-
ture, Higee factor, pressure and liquid flow rate had positive effects on
absorption efficiency while CO2 concentration at inlet and gas flow rate
had negative effects on the same.
Physical absorption process using Rectisol as solvent has been
I. Sreedhar et al. Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

(continued on next page)

reviewed by Gatti et al. [498], by simulating staged regeneration and

[415–417]
[418,419]
Refs. auto-refrigeration via mechanical vapor compression on Aspen Plus.
[376]
[379]
[386]
[389]

[394]
[328]

[406]

[407]
[409]

[411]
[412]

[424]

[425]
[426]
[433]
Larachi et al. [499] used emulsions of DEA with ionic liquids in a
stirred cell contractor to understand the influence of ionic concentra-

Dynamic model using set of PDEs, packed bed using IMTP 40 and Mellapak 250Y, validated using
Steady state, Validated using two pilot plants (Tontiwachwuthikul et al. [51] and Aroonwilas) and
tion, temperature and agitation speed which proved to be the most
vividly impactful. A thermodynamic approach for absorption of CO2 in

Steady state, Validated using pilot plant (Tontiwachwuthikul et al. [51]) and industrial plant
Concentration model gave larger deviations at higher concentrations, activity based versatile

one industrial column (Pintola et al. [450]), different mass transfer coefficients correlations
aqueous solutions of 2-methylpiperidine had been studied. The en-

pilot plant from Separation Research Program (SRP) at University of Texas, Austin [360]
REV and pseudo single liquid model, optimum pressure and height of column calculated

thalpies and gas solubility was determined as a function of tempera-

Steady state, Validated using a pilot plant study at University of Texas, Austin [360]

ture, pressure and amine composition. A rigorous model based on γ-Φ

approach was developed making adjustments according to the available

Non equilibrium stage model, most accurate correlations used, AAN model
VLE and enthalpy data. The consistency of this representation was
expected to be achieved with an upcoming study of speciation at

Multiple screening method, combined CO2 absorption and desorption

equilibrium using NMR techniques [500].

Dynamic model, validated using SRP at University of Texas, Austin

Steady state, Validated using commercially operated MDEA plant

Dynamic model, validated by simulation data from Aspen Plus Lawal et al. [412] simulated a dynamic model containing both

Laminar flow, COMSOL, axial and radial diffusion considered

absorber and desorber linked together by recycle. It studied the post
combustion CC in coal fired power plants. Several assumptions like two
HFMC used, validated against lab-scale and pilot scale

film theory, plug flow, negligible vapor hold up, interfacial phase
equilibrium, linear pressure drop, adiabatic system, reactions reaching
equilibrium etc. were made. The regenerator model consists of two
flash drums – condenser and reboiler. Mass transfer coefficients given
by Onda et al. [396]. The equilibrium compositions were calculated
based on liquid and vapor fugacity coefficients calculated using the
NRTL model in Aspen Plus. The steady state validation of this model
DFT, ONIOM, B3LYP, MP2

was done against the data from University of Texas (UT) at Austin.
Dynamic analyses proved that the molar L/G ratio affected the
(Pintola et al. [450])

absorber performance more than the actual solvent and flue gas flow
rates whereas reboiler duty affected the performance of the regenera-
tor.
Steady state
Remarks

In membrane contactors, Rezaei et al. [501] employed Quantitative

Quicker

Structure property Relationship (QSPR) technique which used struc-

tural variables for modeling sixteen amines. Genetic Algorithm –
Multiple Linear Regression method was used to select proper variables
Effect of disturbances on temperature profile and absorption

temperature profiles, water mass fraction, CO2 capture level,

total heat duty/CO2 recovery; CO2 recovery/amine flow rate

and develop the model. The model was validated with a standard error
Mass transfer coefficients in gas, membrane and liquid

estimate equal to 0.0572 and 0.0496 and the coefficient of determina-

column stages to achieve a specific level of separation

tion equal to 0.986 and 0.989 for the absorption and desorption
process respectively. In order to explain the relationship between
CO2 and MEA concentrations along column

descriptors and CO2 loadings, mechanistic approaches were used.

CO2 concentration and temperature profile

CO2 and H2S concentrations along column

Stearic hindrance affected desorption more than absorption.

L/G ratio, CO2 absorption and loading

Repeatability and stability of solvent

To model the pressure drop and mass transfer in the gravity based
Cyclic capacity, loading difference

rotational packed bed reactor, Zhao et al. [502] proposed various

Cyclic capacity, heat of reaction
Carbamate formation reaction

models like statistical regression, mechanistic theory, numerical simu-

reboiler duty through time
Concentration and activity

lation or intelligence based approach.. The numerical simulations could

Evaluation parameter

be either chemistry based or CFD based, depending on the degree of

coupling [503,504]. The former failed to describe the flow pattern and
the mixing of multiple phase flow. Though costly, the latter was
extensively used as it presented a complete model of fluid flow, mixing
and mass transfer. It involved the simulation of coefficients of mass
phases

rates

transfer, trajectories of droplets etc. using the Eulerian-Lagrangian

method [505]. Intelligence based models correlated the influencing
AMP, DEA, DEEA, MEA,

factors to the performance parameters and were based on Artificial

TEPA, MAPA, TMBPA

Neural networks (AAN), Support Vector Regression (SVR) paired with

MEA, DEA, MDEA

heuristic search methods like Genetic Algorithm (GA), Particle Swarm

Optimization (PSO), Simulated Annealing (SA). These helped provide
MEA and DEA
Summary of modeling and simulation studies.

an accurate prediction based on the test data using the training based
on sample data. Reports suggested that AAN gave closer estimations of
Solvent

MDEA
MDEA

mass transfer coefficients with accuracies of 5% for ammonia and

MEA

MEA
MEA
MEA
MEA
MEA

MEA
MEA

MEA

MEA

PZ

ethanol stripping, 11% for VOC absorption [506]. However, it had

drawbacks like over-fitting, unrepresentative input selection, local
Bolhar-Nordenkampf et al.

minimization. The SVR model, when compared to the ANN model,

displayed better accuracy in prediction [507]. Modeling and parameter
Norouzbahari et al.

estimation has been done for the absorption of CO2 in aqueous MEA
Chowdhury et al.
Kvamsdal et al.
Model name

solution in a pilot scale packed column [509] However, this was a

Gilassi et al.
Harun et al.
Baghli et al.
Mores et al.

Aronu et al.
Kucka et al.

Lawal et al.
Putta et al.

Khan et al.
Rode et al.

complete data driven model and did not provide a complete overview of
Kim et al.
Chu et al.

Luo et al.
Table 8

the processes of mixing, reaction and mass transfer.

Table 8 [352,452] contains the overview of the steady and dynamic
models studied. Various models have their own correlations developed

1101
I. Sreedhar et al. Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

(continued on next page)

for an enhancement factor, overall mass and energy balances and

[508][509]

[53,510]
transfer rates, molar flux for CO2, pressure drop, reboiler equations etc.
Refs.

[438]
[129]

[453]

[118]
[454]
[165]
[465]
[475]
[491]

[492]

[495]

[497]
[451]

[491]
Modeling studies were very much needed to understand the
absorption based carbon capture using different solvents under differ-
ent process conditions. These studies enable us to understand the

packed bed using Pall rings, set of ODEs+liquid film used, liquid phase CO2 loading mole ratio
Steady state, Validated using a wetted wall apparatus, sensitive to surface tension of solution,
interactions at macro and micro levels between solvent and gas

Dynamic model suing PDEs, packed bed using IMTP 40, validated using pilot plant study at

packed bed using Berl saddles, set of ODEs (CDC model), gas phase temperature missing
molecules. Most of the modeling studies reported were either equili-
brium or rate based focusing at macromolecular level at steady states.
More rigorous and advanced dynamic models are needed to under-
stand the phenomenon at molecular level with all complexities like

packed bed using Pall rings, set of ODEs, validated using Vanichseni [448]
difficult kinetic, thermodynamic and hydrodynamic scenarios (turbu-
Steady state, Validated using pilot plant (Tontiwachwuthikul et al. [52])

lence, non-ideal flows etc), unsteady state conditions, multi-component

Dynamic model, validated using pilot plant of coke plant in Germany

removal systems etc. The modeling studies if done rigorously, enables
us to screen best performing solvents and identifying future directions
Dynamic model, validated using absorption column pilot plant

of research.
Asymm eNRTL for liquid phase, ESRK model for gas phase

Packed bed using raschig rings, set of ODEs (CDC model)

5. Conclusions and challenges

The absorption based carbon capture has been reviewed with

henry's law constant and diffusivity of CO2

reference to various aspects like solvents and their synthesis protocols;

their performance analysis in CC; influence of different promoters to
further enhance the solvents’ performance; different contactor config-
University of Texas, Austin [360]
RSM model for non-linear fitting

urations employed in absorption based CC with the details of critical

design parameters, operating conditions, scale and performance ana-
EoS thermodynamic model

lysis; modeling and simulation studies done to understand and validate

the technology under various process conditions and thermo-kinetic
studies to understand the main factors that influence the same. This
Cluster effect

COMSO-SAC

review is a comprehensive one and gives an update for prospective

Remarks

researchers to understand the developmental trends, gaps existing in

MATLAB

missing

various aspects along with future challenges that may be summarized

as follows
axial profile of mass transfer and reaction rates, dynamic and

•
Amine concentration, liquid and gas flow rate, CO2 partial

CO2 concentration across the film and temperature profile

liquid and gas phase temperature, mole fraction of CO2 in

liquid and gas phase temperature, mole fraction of CO2 in

Though there has been a great deal of research efforts in identifying

dynamic temperature profiles, CO2 loading and removal

liquid phase temperature, CO2 loading mole ratio, mole

liquid phase diffusivity of CO2, partial pressure of CO2

novel solvents like blended amines, biphasic solvents, sterically

CO2 concentration, liquid temperature along column

hindered amines, amines with structural modifications. ionic liquids

with suitable functionalization, there is still a great scope for scale
steady state profiles in discretized liquid film
calculated vs. measured CO2 absorption flux

Temperature, diffusion coefficient, solubility

up of successful solvents at the lab scale

Simple synthesis protocols need to be developed to synthesize
the solvents to achieve the desired attributes like solubility, low
regeneration energy, favorable mass transfer, kinetics and thermo-
bulk gas phase, pressure of CO2
efficiency, re-boiler heat duty

dynamics,, biodegradability and low toxicity, low viscosity and hence

Coefficients of mass transfer

better transport properties and also improved chemical and thermal

Evaluation parameter

pressure, temperature

stabilities.
Carbamate formation

•
CO2 solubility data

Economic analysis

Dynamic modeling and simulations in place of steady state models

Absorption rate

bulk gas phase

are required for optimizing the CC process with respect to various

along column

critical parameters like gas and liquid flow rates, concentration,

temperature, CO2 partial pressure thereby permitting successful
integration of the CC technology with power generation or any other
CO2 source with all its complexities like load fluctuations and start-
up issues.
Amines/ammonia mixture

• Better monitoring and control of solvent degradation products is

Partially carbonated

Aqueous ammonia

required by flue gas conditioning, operating absorber and regen-

erator at lowest possible temperature and by following an integrated
approach in designing the absorber and scrubbing sections.
ammonia

•
AMP-PZ

Novel cost effective analytical tools need to be developed for

Solvent

K2CO3

NaOH

assessing solvents’ attributes mentioned above that affect their

MEA

MEA

MEA

MEA
AMP

MEA
2PE

ILs
ILs
ILs

performance under the most rigorous conditions so that further

research may be planned as per the outcome.
•
Tontiwachwuthikul et al.

To extract the both selectivity through absorption and compactness

by membranes, membrane contactors are best suited for post
Table 8 (continued)

combustion CC by absorption including the regeneration process.

Jayarathna et al.
Gabrielsen et al.

Bazargani et al.
Sherman et al.
Model name

NAVanichseni

Hence efforts need to be directed to identify novel, robust and non-

Ghaemi et al.

Ghaemi et al.
Pandya et al.
Pelayo et al.
Wang et al.

Kenig et al.
Puxty et al.
Khan et al.

porous membrane materials configured in the appropriate modular

Gao et al.
Wu et al.

design with high resistance to wetting and fouling. Long time studies
need to be done for commercialization [522].
• Process intensification needs to be extensively applied in absorber

1102
I. Sreedhar et al. Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

and stripper sections with suitable solvents employing reactor [40] Isaacs E, Otto FD, Mather AE. J Chem Eng Data 1977;22:71–3.
[41] Lepaumier Picq D, Carrette PL. Energy Procedia 2009;1:893–900.
systems like rotating paced bed instead of conventional packed [42] Shen KP, Li MH. J Chem Eng Data 1992;37:96–100.
bed reactors so that the capital cost could be significantly reduced. [43] Sartori , Savage DW. Ind Eng Chem Fundam 1983;22:239–49.
Modeling and simulation studies are also required in this area for [44] Archane, Gicquel L Provost E, Fürst E, Chem. Eng. Res. Des. 86. 592–9.
[45] Usubharatana P, Tontiwachwuthikul P. Energy Procedia 2009;1:95–102.
effective process control and scale-up. [46] Vaidya PD, Mahajani VV. Ind Eng Chem Res 2005;44:1868–73.
• There is a strong need to use energy integration, process intensifica- [47]
[48]
Bishnoi S, Rochelle GT. Chem Eng Sci 2000;55:5531–43.
Austgen DM, Rochelle GT, Chen CC. Ind Eng Chem Res 1991;30:543–55.
tion, flow sheet modifications/retrofitting and CO2 utilization in this
[49] Barzagli F, Mani F, Peruzzini M. Energy Environ Sci 2010;3:772–9.
absorption based CC technology to make it commercially viable. [50] Bishnoi S, Rochelle GT. Fluid Phase Equilib 2000;168:241–58.
• As more focus has been directed towards absorption, desorption of [51] Xu GW, Zhang CF, Qin SJ, Gao WH, BH. Ind Eng Chem Res 1998;37:1473–7.
CO2 from rich solvents needs greater attention by researchers by [52] Kim YE, Lim JA, Jeong SK, Yoon YI, Bae ST, Nam SC. Bulk Or Chem Soc
2013;34:783–7.
employing novel techniques like acid addition, dual alkali method,
[53] Tontiwachwuthikul P, Meisen A, Lim CJ. Chem Eng Sci 1992;47:381–90.
membrane technology, stripper carriers so as to minimize the [54] Aboudhei , Tontiwachwuthikul P, Idem R. Ind Eng Chem Res 2006;45:2553–7.
regeneration energy requirement that had become a major hurdle [55] Alper E. Ind Eng Chem Res 1990;29:1725–8.
for this CC technology. [56] Rinker EB. [Ph.D. dissertation]. University of California, Santa Barbara; 1997.
[57] Mandal BO, Bandyopadhyay SS. Chem Eng Sci 2005;60:6438–51.
• Effective promoters viz., organic, inorganic or enzymatic ones [58] Dash SK, Samanta AN, Bandyopadhyay SS. Int J Greenh Gas Control
compatible to various solvents need to be identified so as have the 2014;21:130–9.
desired kinetics and thereby lower equipment cost in the [59] Chen X, Closmann F, Rochelle GT. Energy Procedia 2011;4:101–8.
[60] Rochelle G, Chen E, Freeman S, Wagener DV, Xu Q, Voice A. Chem Eng J
process.Future research should also focus on high temperature 2011;171:725–33.
carbon capture for a successful integration with power sector like [61] Adeosun Abbas Z, Abu-Zahra MRM. Energy Procedia 2013;37:300–5.
employing high temperature resistant inerts like coal fly ash in [62] Gouedard , Picq D, Launay F, Carrette PL. Int J Greenh Gas Control
2012;10:244–70.
adsorption based CC [524]. [63] Kumar N, Rao DP. Gas Sep Purif 1989;3:152–5.
[64] Le Q, Xu J, Shi Y. Acta Phys Chim Sin 1992;8:753–9.
References [65] Ahmadi M, Gomes VG, Ngian K. Sep Purif Technol 2008;63:107–15.
[66] Endo K, Nguyen QS, Kentish SE, Stevens GW. Fluid Phase Equilib
2011;309:109–13.
[1] National Research Council. National Academies Press, Washington; 2010. [67] Guo D, Thee H, Silva G, Chen J, Fei W, Kentish SE. Environ Sci Technol
[2] Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB. . Cambridge 2011;45:4802–7.
University Press, New York; 2007. [68] Shen S, Feng X, Zhao R, Ghosh UK, Chen A. Chem Eng J 2013;222:478–87.
[3] Edmonds J. Proceedings of the 9th international conference on greenhouse gas [69] Cullinane T, Rochelle GT. Ind Eng Chem Res 2006;45:2531–45.
control technologies. Washington; 2008. [70] Khodayari. [Ph.D. thesis]. University of Illinoisat Urbana-Champaign; 2010.
[4] Metz B, Davidson OR, Bosch PR, Dave R, Meyers LA. . Cambridge University [71] Lu Y, Ye X, Zhang Z, Khodayari A, Djukadi T. Energy Procedia 2011;4:1286–93.
Press, New York; 2007. [72] Russo ME, Olivieri G, Marzocchella A, Salatino P, Caramuscio P, Cavaleiro C. Sep
[5] Freund P. Proc Inst Mech Eng A: J Power Energy 2003;217:1–7. Purif Technol 2013;107:331–9.
[6] Dinca C, Slavu N, Badea A. J Energy Inst 2017:1–11. [73] Lee SC, Choi BY, Lee TJ, Ryu CK, Ahn YS, Kim JC. Catal Today 2006;111:385–90.
[7] Van Der Zwaan B, Smekens K. Environ Model Assess 2009;14:135–48. [74] Mores P, Scenna N, Mussati S. Int J Greenh Gas Control 2012;6:21–36.
[8] Abass , Olajire A. Energy 2010;35:2610–28. [75] Wappel D, Gronald G, Kalb R, Draxler J. Int J Greenh Gas Control
[9] Rao B, Rubin ES. Environ Technol 2002;36:4467–75. 2010;4:486–94.
[10] Davidson RM. IEA Clean Coal Centre CCC/125. 2007. [76] Guo BS, Jing GH, Zhou ZM. Int J Greenh Gas Control 2015;34:31–8.
[11] Aroonwilas A, Veawab A. Ind Eng Chem Res 2004;43:2228–37. [77] Bai H, Yeh AC. Ind Eng Chem Res 1997;36(6):2490–3.
[12] Britt D, Furukawa H, Wang B, Glover TG, Yaghi OM. Proc Natl Acad Sci USA [78] Shuangchen Ma. Chin J Soc Power Eng 2012;32(1):52–8.
2009;106:20637–40. [79] Shuangchen Ma. Chem Ind Eng Prog 2012;5:1143–9.
[13] Luebke R, Eubank JF, Cairns AJ, Belmabkhout Y, Wojtas L, Eddaoudi M. Chem [80] Reddy S, Scherffius J, Francuz V. Patent No. WO2007075485A2. 2007.
Commun 2012;48:1455–7. [81] Oyenekan BA, Rochelle GT. Proceedings of the 8th international conference on
[14] Gassensmith JJ, Furukawa H, Smaldone RA, Forgan RS, Botros YY, Yaghi OM. J greenhouse gas control. Trondheim, Norway; 2006.
Am Chem Soc 2011;133:15312–5. [82] Oyenekan BA, Rochelle GT. Ind Eng Chem Res 2006;45:2457–64.
[15] Yao Q, Su J, Cheung O, Liu Q, Hedin N, Zou X. J Mater Chem 2012;22:10345–51. [83] Rochelle G. Patent No.WO2004080573 A1.2004.
[16] Silvestre JA, Wahby A, Sepulveda AE, Martinez ME, Kaneko K, Rodriguez FR. [84] Rochelle GT, Oyenekan BA. Patent No. US2008/0127831. 2008.
Chem Commun 2011;47:6840–2. [85] Reddy S, Gilmartin J. Francuz. PatentNo.WO2007075466A2. 2007.
[17] Laybourn Dawson R, Clowes R, Iggo JA, Cooper AI, Khimyak YZ. Polym Chem [86] Geleff SA. PatentNo.WO2004073838A1. 2004.
2012;3:533–7. [87] Mimura T, Yagi Y, Iijima. Patent No. CA2557454A1. 2006.
[18] Rabbani MG, El-Kaderi HM. Chem Mater 2012;24:1511–7. [88] Jeon SB, Cho SW, Lee SS, Jang SY, Oh KJ. J Taiwan Inst Chem Eng
[19] Kiskan , Antonietti M, Weber J. Macro Mol 2012;45:1356–61. 2014;45:2673–80.
[20] Furukawa H, Yaghi OM. J Am Chem Soc 2009;131:8875–83. [89] Zhang W, Ye L, Jian J. J Environ Chem Eng 2015;3:227–32.
[21] Belmabkhout Y, Sayari A. Adsorpt 2009;15:318–28. [90] Lee MS, Goswami Y, Kothurkar N, Stefanakos EK. Powder Tech 2015;274:313–8.
[22] Bacsik Z, Ahlsten N, Ziadi A, Zhao G, Garcia-Bennett AE, Matute B. Langmuir [91] Lee MS, Goswami Y, Kothurkar N, Stefanakos EK. Proc Energy Sustain
2011;27:11118–28. 2008:10–4.
[23] Si X, Jiao C, Li F, Zhang J, Wang S, Liu S. Energy Environ Sci 2011;4:4522–7. [92] Lee MS, Goswami Y, Stefanakos EK. Int J Hydrog Energy 2009;34:745–52.
[24] Brunetti A, Scura F, Barbieri G, Drioli E. J Membr Sci 2010;359:115–25. [93] Mak SY, Chen Y. Rapid Commun 2005;26:1567–71.
[25] Gupta M, Coyle I, Thambimuthu K. . CANMET Energy Technology Centre, Natural [94] Liu Z, Du Z, Zou W, Li H, Mi J, Zhang C. Chem Eng J 2013;223:915–20.
Resources, Canada; 2003. [95] Lu G, Hua AC, Bao LL, Liu SX, Zhang H, Xu ZW. Sep Purif Tech 2011;82:87–92.
[26] Sreenivasulu B, Gayatri DV, Sreedhar I, Raghavan KV. Renew Sust Energy Rev [96] Du Y, Yuan Y, Rochelle GT. Chem Eng Sci 2016:397–404.
2005;41:1324–50. [97] Fernandez J, Sotenko M, Derevschikov V, Lysikov A, Rebrov EV. Chem Eng
[27] GHG IEA. . IEAGHG, IEAGHG/SR2, Cheltenham. UK; 1993. Process 2016;108:17–26.
[28] Axel M, Xiaoshan S. Energy Convers Manag 1997;38:37–42. [98] Saeed M, Deng LJ. Membr Sci 2016.
[29] Vega F, Sanna A, Maroto-Valer MM, Navarrete B, Abad-Correa D. Int J Greenh [99] Kazemi A, Mehrabani-Zeinabad A. Energy 2016;112:353–63.
Gas Control 2016;54:160–7. [100] Fu D, Wang LM, Mi CL, Zhang P. J Chem Therm 2016;101:123–9.
[30] Chakma A. Energy Sources 1999;21:51–62. [101] Vaidya PD, Kenig EY. Chem Eng Technol 2009;32:556–63.
[31] Thomas C Kerr HR. Elsevier, Oxford; 2005. [102] Vaidya PD, Kenig EY. Chem Eng Sci 2007;62:7344–50.
[32] EA. OECD/IEA, France; 2008. [103] Chowdhury FA, Yamada H, Higashii T, Goto K, Onoda M. Ind Eng Chem Res
[33] Dave N, Do T, Puxty G, Rowland R, Feron PHM, Attalla MI. Energy Procedia 2013;52:8323–31.
2009;;1:949–54. [104] benitez-garcia J, ruiz-ibanez G, Al-Ghawas HA, Sandall OC. Chem Eng Sci
[34] Yan S, Fang M, Zhang W, Zhong W, Luo Z, Cen K. Energy Convers Manag 1991;46:2927–31.
2008;49:3188–97. [105] Kim C, Savage D. Chem Eng Sci 1987;42:1481–7.
[35] Aroonwilas , Tontiwachwuthikul P. Sep Purif Technol 1997;12:67–79. [106] Little R, Van Swaaij W, Versteeg G. AIChe J 1990;36:1633–40.
[36] Schwendig. Leimkuhler H-J. (Ed.). 2010. 391–406. [107] Zhang JF, Fennell PS, Trusker JM. J Chem Eng Data 2015;60:2392–9.
[37] Cheng YH, Chih HH, Chung TS. Aerosol Air Qual Res 2012;12:745–69. [108] Weiland RH, Dingman JC, Cronin DB, Browning GJ. J Chem Eng Data
[38] Sumedh S, Warudkara , Kenneth R, Michael S, Wonga B, Hirasakia GJ. Int J 1998;43:378–82.
Greenh Gas Control 2013;16:342–50. [109] Xu ZC, Wang SJ, Chen CH. J Chem Eng Data 2013;58:1633–40.
[39] Dawodu OF, Meisen A. J Chem Eng Data 1994;39:548–52. [110] Xu ZC, Wang SJ, Chen CH. Int J Greenh Gas Control 2013;16:107–15.

1103
I. Sreedhar et al. Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

[111] Xu ZC, Wang SJ, Chen CH. Ind Eng Chem Res 2013;52:9790–802. 2014;7:897–902.
[112] Azhgan M, Farsi M, Eslanloueyan R. Int J Greenh Gas Control 2016;51:409–14. [173] Wei CC, Puxty G, Feron P. Chem Eng Sci 2014;107:218–26.
[113] Wanga Z, Fanga M, Pana Y, Yanb S, Luoa Z. Energy Procedia 2013;37:1085–92. [174] Yu H, Wu YT, Jiang YY, Zhou Z, Zhang ZB. New J Chem 2009;33:2385–90.
[114] Singto S, SUpap T, Idem R, Tontiwachwuthikul P, Tantayanon S, Al-Marri J , et al. [175] Zhang F, Fang G, W YT, Wu YT, Li AM, Zhang ZB. Chem Eng J 2010;160:691–7.
Sep Purif Technol. [176] Zhang F, Ma JW, Zhou Z, Wu YT, Zhang ZB. Chem Eng J 2012;181–182:222–8.
[115] Hadri N, Quang DV, Goetheer ELV, Abu Zahra MRM. Appl Energy [177] Zhang F, Gao Y, Wu XK, Ma JW, Wu YT, Zhang ZB. Chem Eng J 2013;223:371–8.
2017;185:1433–49. [178] Gao Y, Zhang F, Huang K, Ma JW, Wu YT, Zhang ZB. Int J Greenh Gas Control
[116] Xiao M, Liu H, Idem R, Tontieachwuthikul P, Liang Z. Appl Energy 2013;19:379–86.
2016;184:219–29. [179] Zhou ZM, Jing GH, Zhou LJ. Chem Eng J 2012;204–206:235–43.
[117] Sartori G, Savange DW. Ind Eng Chem Fundam 1983;22:239–49. [180] Fu D, Xie JJ. Chem Thermodyn 2016.
[118] Sherman BJ, Ciftja AF, Rochelle GT. Chem Eng Sci 2016;153:295–307. [181] Jing GH, Zhou LJ, Zhou ZM. Chem Eng J 2012;181–182:85–92.
[119] Freeman SA. University of texas at Austin; 2011. [182] Park SB, Shim CS, Lee H, Lee KH. Fluid Phase Equilib 1997;134:141–9.
[120] Voice AK. University of texas at Austin; 2013. [183] Gonçalves FA, Kestin J. Int J Thermophys. 1981. 2. 315–22.
[121] Li I. University of Texas at Austin; 2015. [184] Correia RJ, Kestin J, Khalifa HE. J Chem Eng Data 1980;25:201–6.
[122] Gao J, yin J, Zhu F, Chen X, Tong M, Kang W, et al. Sep Purif Technol [185] Weiland RH, Dingman JC, Cronin DB. J Chem Eng Data 1998;43:378–82.
2016;167:17–23. [186] Xue ZM, Zhang ZF, Han J, Chen Y, Mu TC. Int J Greenh Gas Control
[123] Shamiri A, Shafeeyan MS, Tee HC, Leo CY, Aroua MK, Aghamohammadi N. J Nat 2011;5:628–33.
Gas Sci Eng. 2016. 35. 605–13. [187] Gao Y, Zhang F, Huang K, Ma JW, Wu YT, Zhang ZB. Int J Greenh Gas Control
[124] Svendsen HF, Hessen ET, Mejdell T. Chem Eng J 2011;171:718–24. 2013;19:379–86.
[125] Ramazani R, Samsani A, Jahanmiri A, van der Bruggen B, Mazinani S. Int J [188] Fu D, Zhang P, Mi C. Energy 2016;101:288–95.
Greenh Gas Control 2016;51:29–35. [189] Fu D, Zhang P. Energy. 2015. 87. 165–72.
[126] Vaidya PD, Kenig EY. Ind Eng Chem Res 2008;47:34–8. [190] Qian W, Xu Y, Xie B, Ge Y, Shu H. Int J Greenh Gas Control 2017;56:194–201.
[127] Shuangchen M, Gongda C, Sijie Z, Jiaqi W, Ran G, Lan M, et al. Appl Energy [191] Luo X, Wang C. Curr Opin Green Sustain Chem 2017;3:33–8.
2016;169:597–606. [192] Cheng J, Li YN, Hu LQ, Zhou JH, Cen KF. Energy Fuels 2016;30:3251–6.
[128] Choi JH, Kim YE, Nam SC, Yun SH, Yoon YI, Lee JH. Korean J Chem Eng 2016. [193] Heldebrant DJ, Koech PK, Ang MTC, Liang C, Rainbolt JE, Yonkera CR. Green
[129] Khan AA, Halder GN, Saha AK. Int J Greenh Gas Control 2016;44:217–26. Chem 2010;12:713–21.
[130] Hairul NAH, Shariff AM, Bustam MA. Fuel Proc Tech 2017;157:20–8. [194] Heldebrant DJ, Koech PK, Rainbolt JE, Zheng F, Smurthwaite T, Freeman CJ.
[131] Zhang ZJ. Nat Gas Sci Eng 2016;31:589–95. Chem Eng J 2011;171:794–800.
[132] Luo W, Guo D, Zheng J, Gao S, Chen J. Int J Greenh Gas Control 2016;53:141–8. [195] Heldebrant DJ, Koech PK, Yonker CR. Energy Environ Sci 2010;3:111–3.
[133] Hartono A, da Silva EF, Svendsen HF. Chem Eng Sci 2009;64:3205–13. [196] Heldebrant DJ, Yonker CR, Jessop PG, Phan L. Energy Environ Sci
[134] Salvi AP, Vaidya PD, Kenig FY. Can J Chem Eng 2014;92:2021–8. 2008;1:487–93.
[135] Kim YE, Moon SJ, Yoon YI, Jeong SK, Park KT, Bae ST, et al. Sep Purif Technol [197] Heldebrant DJ, Yonker CR, Jessop PG, Phan L. Chem A Eur J 2009;15:7619–27.
2014;122:112–8. [198] Helderbrant DH, Koech PK, Rainbolt JE, Zheng FC. Energy Procedia
[136] Fu K, Chen G, Sema T, Zhang X, Liang Z, Idem R, et al. Chem Eng Sci 2011;4:216–23.
2013;100:195–202. [199] Lin PH, Wong DSH. Int J Greenh Gas Control 2014;26:69–75.
[137] Fu K, Sema T, Liang Z, Liu H, Idem R, Tontiwachwuthikul P. Ind Eng Chem Res [200] Hoffman S, Hammache S, Gray ML, Fauth DJ, Pennline HW. Fuel Process
2012;51:12058–64. Technol 2014;126:173–87.
[138] Zhang X, Fu K, Liang Z, Rongwong W, Yang Z, Idem R, et al. Fuel [201] Eftaiha AF, Alsoubani F, Assaf KI, Troll C, Rieger , Khaled AH, et al. Carbo Poly
2014;136:261–7. 2016;152:163–9.
[139] Nwaoha C, Saiwan C, Tontiwachwuthikul P, Supap T, Rongwong W, Idem R, et al. [202] Aoyagi N, Endo T. Tetrahedron 2017. http://dx.doi.org/10.1016/
Nat Gas Sci Eng 2016. j.tet.2017.01.012.
[140] Nwaoha C, Saiwan C, Tontiwachwuthikul P, Supap T, Rongwong W, Idem R, et al. [203] Lin S, Harada M, Suzuki Y, Hatano H. Fuel 2002;81:2079–85.
Int J Greenh Gas Control 2016;53:292–304. [204] Barker R. J Appl Chem Biotechnol 1973;23:733–42.
[141] Esmaeili H, Roozbehani B. Int J Greenh Gas Control 2014;30:212–5. [205] Stanmore BR, Gilot P. Fuel Process Technol 2005;86:1707–43.
[142] Nwaoha C. [MSc Thesis]. The Petroleum and Petrochemical College, [206] Zhu Y, Li W, Sun G, Tang Q, Bian H. Int J Greenh Gas Control 2016;49:290–6.
Chulalongkorn. University, Bangkok, Thailand; 2015. [207] Park SW, Choi BS, Kim SS, Lee BD, Lee JW. J Ind Eng Chem 2008;14:166–74.
[143] Nwaoha C, Supap T, Idem R, Saiwan C, Tontiwachwuthikul P, Al-Marri J, et al. [208] Turanov AN, Tolmachev YV. Heat Mass Transf 2009;45:1583–8.
Petroleum 2016:1–27. [209] Feng XM, Johnson DW. Int J Heat Mass Transf 2012;55:3447–53.
[144] Choi WJ, Min BM, Seo JB, Park SW. Ind Eng Chem Res 2009;48:4022–9. [210] Lee JW, Pineda IT, Lee JH, Kang YT. Appl Energy 2016;178:164–76.
[145] Pellegrini G, Strube R, Manfrida G. Energy 2010;35:851–7. [211] Zhuang Q, Clements B, Dai J, Carrigan L. Int J Greenh Gas Control
[146] Puxty G, Rowland R, Attalla M. Chem Eng Sci 2010;65:915–22. 2016;52:449–60.
[147] Yeh AC, Bai H. Sci Total Environ 1999;228:121–33. [212] Carrette PL, Cadours R, Boucot P, Mougin P, Prigent M, Gibert A , et al. 10th
[148] Zhao B, Su Y, Tao W, Li L, Peng Y. Int J Greenh Gas Control 2012;9:355–71. Meeting of the IEA international post-combustion CO2 capture network. 24–25th
[149] Black S, Paelinck P, Kozak F, Petig A, Morris E. Proceedings of Pennwell oil sands May 2007. Lyon, France.
conference. Calgary, Alverta, Canada; 2008. [213] Zhang J, Nwani O, Tan Y, Agar DW. Chem Eng Res Des 2011;89(8):1190–6.
[150] Inc EIG. WO Patent No WO/2006/02285. 2006. [214] Raynal L, Alix P, Bouillon PA, Gomez A, de Nailly MlF, Jacquin M, et al. Trapy J
[151] Bonalumi D, Giuffrida A. Energy 2016;117:439–49. Energy Procedia 2011;4:779–86.
[152] Li L, Conway W, Burns R, Maeder M, Puxty G, Clifford S, et al. Int J Greenh Gas [215] Agar D, Tan Y, Zhang X. Patent WO/2008/015217. 2008.
Control 2017;56:165–72. [216] Wang L, An S, Li Q, Yu S, Wu S. Chem Eng J 2017;314:681–7.
[153] Sharifzadeh M, Bumb P, Shah N. Appl Energy 2016;163:423–35. [217] Bai H, Zheng Y, Yang R. Mater Des 2016;99:145–54.
[154] Gladis A, Gundersen MT, Fosbol PL, Woodley JM, Solms N. Chem Eng J [218] Chen WH, Chen SH, Hung CI. Appl Energy 2015;111:731–41.
2017;309:772–86. [219] International Energy Agency. Report No. PH4/33. 2004.
[155] Cents AHG, Brilman DWF, Versteeg GF. Chem Eng Sci 2001;56:1075–83. [220] Kohl AL, Nielsen RB. Gas purification [Fifth]. Houston: Gulf Professional
[156] Dennard AE, Williams RJP. J Chem Soc A 1966:812–6. Publishing; 1997.
[157] Hu G, Nicholas NJ, Smith KH, Mumford KA, Ketish SE, Stevens GW. Int J Greenh [221] Perry R, DW . Green Perry's chemical engineers'. New York: Handbook McGraw-
Gas Control 2016;53:28–40. Hill; 2007.
[158] Shrier A, Danckwerts P. Ind Eng Chem Fundam 1969;3:415–23. [222] Salem , Reyhani FA. Int J Greenh Gas Control 2015;42:237–45.
[159] Hu G, Smith KH, Nicholas NJ, Yong J, Kentish SE, Stevens GW. Chem Eng J [223] Udara SPR, Arachchige, Morten CM. ISSN 1450-216X. 87. 2012. 117–26.
2017;307:49–55. [224] Montigny D, Tontiwachwuthikul P, Chakma A. Can J Chem Eng 2001:79.
[160] Dai Z, Deng L. Int J Greenh Gas Control 2016;54:59–69. [225] Bender P, Moll A, Linde AG, Hollriegelskreuth I. Chem 2002.
[161] Ayala AE, Simoni LD, Lin YD, Brennecke JF, Stadtherr MA. Comput Aided Chem [226] Strigle RF Jr.. Gulf Publishing Co, Texas. 2nd edition; 1994.
Eng 2006;21:463–8. [227] Road PN, Shirur T. Pune; 2000.
[162] Wappel D, Gronald G, Kalb R, Draxler J. Int J Greenh Gas Control [228] Brunazzi E, Pagliant A, Spiegel L, Tolaini F. Ind Chem Mater Sci 2001.
2010;4:486–94. [229] Tan LS, Shariff AM, Lau KK, Bustam MA. J Ind Eng Chem 2012;18:1874–83.
[163] Rogers RD. Nature 2007;447:917–8. [230] Aroonwilas A, Veawab A, Tontiwachwuthikul P. Ind Eng Chem Res
[164] Davis JH. Chem Lett 2004;33:1072–7. 1999;38:2044–50.
[165] Bazargani Z, Sabzi F. J Mol Liq 2016;223:235–42. [231] Godini HR, Mowla D. Chem Eng Res Des 2008;86:401–9.
[166] Anthony JL, Maginn EJ, Brennecke JF. J Phys Chem B 2002;106:7315–20. [232] Zhang X, Wang J, Zhang CF, Yang YH, Xu JJ. Ind Eng Chem Res
[167] Zareiekordshouli F, Lashani-zadehgan A, Darvishi P. Int J Greenh Gas Control 2003;42:118–22.
2016;53:328–37. [233] Aboudheir, Demontigny D, Tontiwachwuthikul P, Chakma A. SPE gas technology
[168] Kenta F, Masahiro Y, Hiroyuki O. J Am Chem Soc 2005;127:2398–9. symposium. Calgary, Alberta, Canada; 1998.
[169] Zhang JM, Zhang SJ, Dong K, Zhang YQ, Shen Q, Lv XM. Chemistry [234] Yan S, Fang M, Zhang W, Zhong W, Luo Z, Cen K. Energy Convers Manag
2006;12:4021–6. 2008;49:3188–97.
[170] Sistla YS, Khanna A. Chem Eng J 2015;273:268–76. [235] Setameteekul , Aroonwilas A, Veawab A. Sep Purif Technol 2008;64:16–25.
[171] Guo HX, Zhou ZM, Jing GH. Int J Greenh Gas Control 2013;16:197–205. [236] Wang Z, Gupta M, Warudkar S, Cox KR, Hirasaki GJ, Wong MS. Ind Eng Chem
[172] Saravanamurugan S, Kunov-Kruse AJ, Fehrmann R, Riisager A. Chemsuschem Res 2016.

1104
I. Sreedhar et al. Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

[237] Fleischer Becker S, Eigenberger G. Chem Eng Sci 1996;5(10):1715–24. [298] Esato K, Eiseman B. J Thorac Cardiovasc Surg 1975;65:690–7.
[238] Patent WO. 201212 A1. 1886. [299] Fashandi H, Ghodsi A, Saghafi R, Zarrebini R. Int J Greenh Gas Control
[239] Zhaoa Sub Y, Penga Y. Int J Greenh Gas Control 2013;17, [481–48]. 2016;52:13–23.
[240] Seyboth , Zimmermann S, Heidel B, Scheffknecht G. Energy Procedia Vol [300] Tontiwachwuthikul P, Chakma A. J Membr Sci 2006;266:99–107.
2014;63:1667–77. [301] Khaisri S, Tontiwachwuthikul P, Jiraratananon R. Sep Purif Technol
[241] Kuntz J, Aroonwilas A. Energy Procedia 2008. 2009;65:290–7.
[242] Zhao B, Su Y, Cui G. Energy 2016;102:106–17. [302] Qi Z, Cussler EL. I. Mass transfer in the liquid. J Membr Sci 1985;23:321–32.
[243] Munjal S, Dudukovic MP, Ramachandran PA. Chem Eng Sci 1989;44:2245–56. [303] Liu F, Hashim NA, Liu Y, Abed Moghareh MR, Li K. J Membr Sci 2011;375:1–27.
[244] Fowler R, Gendes KF, Nyguard HF. 21st Annual offshore technology conference. [304] Naim R, Ismail AF, Cheer NB, Abdullah MS. Chem Eng Res Des 2014;92:1391–8.
Houston, TX; 1989. [305] Rahbari-Sisakht M, Ismail AF, Rana D, Matsuura T. Sep Purif Technol
[245] Lin CC, Liu WT, Tan CS. Ind Eng Chem Res 2003;42:2381–6. 2012;98:472–80.
[246] Yu H, Tan CS. Energy Procedia 2013;37:455–60. [306] Ghasem N, Al-Marzouqi M, Zhu L. Sep Purif Technol 2012;92:1–10.
[247] Zhao B, Tao W, Zhong M, Su Y, Cui G. Renew Sustain Energy Rev 2016;65:44–56. [307] Rahbari-Sisakht M, Ismail AF, Matsuura T. Sep Purif Technol 2012;88:99–106.
[248] Meeuwse M, van der Schaaf J, Kuster BSM, Schouten JC . Chem Eng Sci. 65(1). [308] Rahbari-Sisakht M, Ismail AF, Matsuura T. Sep Purif Technol 2012;86:215–20.
466–71. [309] Gomez-Coma L, Garea A, Irabien A. Int J Greenh Gas Control 2016;52:401–9.
[249] Pan ZQ, Zhang YJ, Deng XH, Huang KJ. South China Univ Technol [310] Zhang L, Li J, Zhou L, Liu R, Wang X, Yang L. Ind Eng Chem Res 2015.
2006;34:67–71. [311] Gomez-Diaz D, Lopez-Rivas D, Navaza JM, Couvert A. Chem Eng
[250] Chen ZQ, Xiong SX, Wu JG. J Chem Ind Eng 1995;46:388–92. 2016;39(4):751–7.
[251] Wang GQ, Xu ZC, Yu YI, Ji JB. Chem Eng Proc: Proc Intensif 2008;47:2131–9. [312] Daugulis AJ, Boudreau NG. Biotechnol Lett 2003;25:1421–4.
[252] Chnadra A, Goswami PS, Rao DP. Ind Eng Chem Res 2005;44:4051–60. [313] Mun oz R, Daugulis AJ, Herna´ndez M, Quijano G. Biotechnol Adv
[253] Reddy KJ, Gupta A, Rao DP. Ind Eng Chem Res 2006;45:4270–7. 2012;30:1707–20.
[254] Chamchan N, Chang JY, Hsu HC, Kang JL, Wong DSH, Jang SS, et al. J Taiwan [314] Galindo E, Pacek AW, Nienow AW. Biotechnol Bioeng 2000;69:213–21.
Inst Chem Eng. 2016. 000. 1–7. [315] Xinh Oyenekan BA. University of Texas; 2007.
[255] Yu CH, Chen MT, Chen H, Tan CS. Appl Energy 2016;175:269–76. [316] Aroonwilas , Veawab A. Int J Greenh Gas Control 2007;1(2):143–50.
[256] Chiang CY, Lee DW, Liu HS. J Taiwan Inst Chem Eng 2017;000:1–8. [317] Aroonwilas , Veawab A. Int J Greenh Gas Control 2007;1:43–50.
[257] Ganapathy H, Shooshtari A, Dessiatoun S, Alshehhi M, Ohadi M. Appl Energy [318] Jassim MS, Rochelle GT. Ind Eng Chem Res 2006;45(8):2465–72.
2014;119:43–56. [319] Oyenekan BA, Rochelle GT. AIChE J 2007;53(12):3144–54.
[258] Raterman KT, McKellar M, Podgorney A, Stacey D, Turner T, Stokes B , et al. [320] Schach MO, Schneider RD, Schramm H, Repke JU. Ind Eng Chem Res
Idaho National Engineering and Environmental Laboratory. 2010;49(5):2363–70.
[259] Gabelman , Hwang ST. J Membr Sci 1999;159:61–106. [321] Chang H, Shi CM. Sep Sci Technol 2005;40(4):877–909.
[260] Mavroudi M, Kaldis SP, Sakellaropoulos GP. Fuel 2003;82:2153–9. [322] TobiesenFA Svendsen HF, Mejdell T. Ind Eng Chem Res 2007;46(23):7811–9.
[261] Dindore VY, Brilman DWF, Versteeg GF. Chem Eng Sci 2005;60:467–79. [323] Plaza JM, Chen E, Rochelle GT. AIChE J 2010;56(4):905–14.
[262] Klaassen R, Feron PHM, Jansen AE. Chem Eng Res Des 2005;83:234–46. [324] Alie C, Backham L, Croiset E, Douglas PL. Energy Convers Manag
[263] Zhang Y, Wang R. Curr Opin Chem Eng 2013;2:255–62. 2005;46(3):475–87.
[264] Al-Marzouqi MH, El-Naas MH, Marzouk SAM, Al-Zarooni MA, Abdullatif N, Faiz [325] ai Y, Lin A, Yang J, Tu ST, Yan J. J Membr Sci 2015;496:1–12.
R. Sep Purif Technol 2008;59:286–93. [326] Lin YJ, Madan T, Rochelle GT. Faraday Discuss 2017, [in press].
[265] Tasselli F, Drioli E. J Membr Sci 2007;301:11–8. [327] Lin YJ, Rochelle GT. Int J Greenh Gas Control 2016;53:169–77.
[266] Lang WZ, Xu ZL, Yang H, Tong W. J Membr Sci 2007;288:123–31. [328] Li T, Keener TC. Int J Greenh Gas Control 2016;51:290–304.
[267] Lee HK, Jo HD, Choi WK, Park HH, Lim CW, Lee YT. Desalination [329] Eimer D, Sjøvoll M, Eldrup N, Heyn R, Juliussen O, McLarney M, et al. In:
2006;200:604–5. Proceedings of the 3rd Nordic symposium on CO2. Trondheim; 2003.
[268] Li JL, Chen BH. Sep Purif Technol 2005;41:109–22. [330] Feng B, Du M, Dennis TJ, Anthony K, Perumal MJ. Energy Fuels 2010;24:213–9.
[269] Rongwong W, Fan C, Liang Z, Rui Z, Idem RO, Tontiwachwuthikul P. J Membr Sci [331] Du M, Feng B, An H, Liu W, Zhang L. Korean J Chem Eng 2011;29:362–8.
2015;490:236–46. [332] Huang H, Chang SG. Energy Fuels 2002;16:904–10.
[270] Zhang S, Lu Y, Ye X, Int J. Greenh Gas Control 2013;13:17–25. [333] Erga O, Juliussen O, Lidal H. Energy Convers Manag 1995;36:387–92.
[271] Zhang YT, Zhang L, Chen HL, Zhang HM. Chem Eng Sci 2010;65:3199–207. [334] Frimpong RA, Remias JE, Neathery JK, Liu K. Int J Greenh Gas Control
[272] Luo J, Meyer AS, Mateiu RV, Pinelo M. New Biotechnol 2015;32:319–27. 2012;9:124–9.
[273] Bednár A, Nemestóthy N, Bakonyi P, Fülöp L, Zhen G, Lu X, et al. Chem Eng J [335] Tobiesen FA, Svendsen HF. Ind Eng Chem Res 2006;45:2489–96.
2016;303:621–6. [336] Simioni M, Kentish SE, Stevens GW. J Membr Sci 2011;378:18–27.
[274] Arazawa DT, Kimmel JD, Finn MC, Federspiel WJ. Acta Biomater [337] Okabe K, Kodama S, Mano H, Fujioka Y. Energy Procedia 2009;1:1281–8.
2015;25:143–9. [338] Teramoto M, Kitada S, Ohnishi N, Matsuyama H, Matsumiya N. J Membr Sci
[275] Zhang S, Zhang Z, Lu Y, Rostam-Abadi M, Jones A. Bioresour Technol 2004;234:83–94.
2011;102:10194–201. [339] Wang Z, Fang M, Pan Y, Yan S, Luo Z. Energy Procedia 2013;37:1085–92.
[276] Hou J, Dong G, Xiao B, Malassigne C, Chen VJ. Mater Chem A 2015;3:3332–42. [340] Fang M, Wang Z, Yan S, Cen Q, Luo Z. Int J Greenh Gas Control 2012;9:507–21.
[277] Yang HC, Hou J, Wan LS, Chen V, Xu ZK. Adv Mater Interfaces 2016;3:500774. [341] Zhao X, Liu N, Wang Y, Fei W, Stevens GW. Ind Eng Chem Res 2011;50:8620–31.
[278] Zhang YT, Zhi TT, Zhang L, Huang H, Chen HL. Polymer 2009;50:5693–700. [342] Khalilpour R, Milani D, Qadir A, Chiesa M, Abbas A. Renew Energy
[279] Zhang YT, Dai XG, Xu GH, Zhang L, Zhang HQ, Liu JD, et al. AIChE J 2017;1004:60–75.
2012;58:2069–77. [343] Khan M, Krishnamoorthi V, Mahmud T. Chem Eng Res Des 2011;89:1600–8.
[280] Yong JKJ, Cui J, Cho KL, Stevens GW, Caruso F, Kentish SE. Langmuir [344] Weiland RH, Hatcher NA, Nava JL Optimized Gas Treating, Inc.,
2015;31:6211–9. [345] Wang M, Lawal A, Stephenson P, Sidders J, Ramshaw C, Yeung H. Chem Eng Res
[281] Yong JKJ, Stevens GW, Caruso F, Kentish SE. J Chem Technol Biotechnol Des 2011;89:1609–24.
2015;90:3–10. [346] Bai HL, Yeh AC. Ind Eng Chem Res 1997;36:2490–3.
[282] Hou J, Ji C, Dong G, Xiao B, Ye Y, Chen VJ. Mater Chem A 2015;3:17032–41. [347] Yeh AC, Bai HL. Sci Total Environ 1999;228:121–33.
[283] Hou J, Zulkifli MY, Mohammad M, Zhang Y, Razmjou A, Chen VJ. Membr Sci [348] Kollera M, Wappelb D, Trofaierb N, Gronald G. Energy Procedia 2011;4:1777–82.
2016. [349] Lin CC, Lin YH, Tan CS. J Hazard Mater 2010;175:344–51.
[284] Bounaceur R, Castel C, Rode S, Roizard D, Favre E. Chem Eng Res Des [350] Wanga M, Atuman SJ, Ramshaw C, Eimer D, Nuhu M. Appl Energy
2012;90:2325–37. 2015;158:275–91.
[285] Zhao S, Paul HM, Deng L, Favre E, Chabanon E, Yan S, et al. J Membr Sci. 2016. [351] Zhao B, Su Y, Tao W. Appl Energy 2014;136:132–42.
511. 180–206. [352] Koronakin IP, Prentza L, Papaefthimiou V. Renew Sust Energ Rev
[286] Mansourizadeh A, Jazebizadeh MH, Vaseghi MR, Aghili A. J Polym Res 2015;15:547–66.
2015;23:1–10. [353] Zhiwu L, Fu K, Idem R, Tontiwachwuthikul P. CJCHE: 318; 2015.
[287] Luis P, Gerven TV, van der Bruggen V. Prog Energy Combust Sci 2012;38:419–48. [354] Danckwerts PV. Gas–liquid reactions. New York: McGraw-Hill; 1970.
[288] Mansourizadeh A, Ismail AF, Matsuura T. J Membr Sci 2010;353:192–200. [355] Zhao B, Su Y, Tao W, Li L, Peng Y. Int J Greenh Gas Control 2012;9:355–71.
[289] Li K, Kong JF, Wang D, Teo WK. AIChe J 1999;45:1211–9. [356] Phan DT, Maeder M, Burns RC, Puxty G. Environ Sci Technol 2014;48:4623–9.
[290] Atchariyawut S, Feng C, Wang R, Jiraratananon R, Liang DT. J Membr Sci [357] Cullinane JT, Rochelle GT. Fluid Phase Equilib 2005;277:197–213.
2006;285:272–81. [358] Guo D, Thee H, Tan CY, Chen J, Fei W, Kentish S, et al. Environ Sci Technol
[291] Rezaei M, Ismail AF, Hashemifard SA, Matsuura T. Chem Eng Res Des 2011;45:4802–7.
2014;92:2449–60. [359] Vrachnos A. National Technical University of Athens, Athens. [Ph.D. thesis].
[292] Rahbari-Sisakht M, Ismail AF, Rana D, Matsuura T, Emadzadeh D. Sep Purif 2007.
Technol 2013;116:67–72. [360] Chen C, Britt HI, Boston JF, Evans LB. AIChE J 1982;28:588.
[293] Wongchitphimon S, Wang R, Jiraratananon R. J Membr Sci 2011;381:183–91. [361] Chen C, Evans LB. AIChE J 1986;32:444.
[294] Xu A, Yang A, Young S, Tontiwachwuthikul P. J Membr Sci 2008;311:153–8. [362] Sukanta D, Syamalendu B, Intl J. Greenh Gas Control 2016;44:227–37.
[295] Ren J, Wang R, Zhang HY, Liang DT, Tay JH. J Membr Sci 2006;281:34–344. [363] Chunxi L, Furst W. Chem Eng Sci 2000;55:2975.
[296] Ghodsi A, Fashandi H, Zarrenbini M, Abolhasani MM, Gorji M. RSC Adv [364] Falck S. Report in DIK. 2002.
2015;5:92234–53. [365] Arshad MW, von Solms N, Thomsen K. Int J Greenh Gas Control
[297] Lin SH, Chiang PC, Hsieh CF, Li MH, Tung KL. J Chin Inst Chem Eng 2016;53:401–24.
2008;39:13–21. [366] Diamantonis . Energy Fuels 2011;25(7):3334–43.

1105
I. Sreedhar et al. Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107

[367] Pitzer KS. Thermodynamics of electrolytes. J Phys Chem 1973;77(2):268–77. [424] Luo X, Liu S, Gao H, Liao H, Tontiwachwuthikul P, Liang Z. Sep Purif Technol
[368] Deshmukh R, Mather A . Chem Eng Sci. 1998, 1, 36(2). 355–62. 2016;169:279–88.
[369] Zhang XG, Yu YX. Fluid Phase Equilib 2010;295:237–43. [425] Bolhàr-Nordenkampf M, Friedl A, Koss U, Tork T. Chem Eng Process
[370] Yu YX, Gao GH. Fluid Phase Equilib 1999;166:111–24. 2004;43:701–15.
[371] Yu YX, Gao GH. Fluid Phase Equilib 2011;1736:165–79. [426] Gilassi S, Rahmanian N. Int J Chem React Eng 2015.
[372] Dullien FAI. Trans Faraday Soc 1963;59:856–68. [427] Zhang ZJ. Nat Gas Sci Eng 2016;31:589–95.
[373] Vranes M, Papovic S, Pavlovic K, Gadzuric S. J Chem Thermodyn [428] Seader JD, Henley EJ, Roper DK. Separation process principles. Canada: John
2012;57(4):1072–7. Wiley & Sons; 2011.
[374] Weingartner H. Angew Chem Int Ed 2008;47:654–70. [429] Happel J. AIChE J 1959;5(2):174e177.
[375] Ghatee MH, Zare M, Zolghadr AR, Moosavi F. Fluid Phase Equilib [430] Saidi M. J Memb Sci 2017;524:186–96.
2010;291:188–94. [431] Nii S, Takeuchi H. Gas Sep Purif 1994;8:107–14.
[376] Putta KR, Pinto DDD, Svendsen HF, Knuutila HK. Int J Greenh Gas Control [432] Yan SP, Fang MX, Zhang WF, Wang SY, Xu ZK, Luo ZY, et al. J Memb Sci
2016;53:338–53. 2006;285:272–81.
[377] Aboudheir A, Tontiwachwuthikul P, Chakma A, Idem R. Chem Eng Sci [433] Norouzbahari S, Shahhosseini S, Ghaemi A. RSC Adv 2016.
2003;58:5195–210. [434] Billet R, Schultes M. Chem Eng Res Des 1999;77:498–504.
[378] Luo X, Hartono A, Svendsen HF. Chem Eng Sci 2015;123:57–69. [435] Pitzer KS. J Phys Chem 1973;77:268–77.
[379] Kim S, Shi H, Lee JY. Int J Greenh Gas Control 2016;45:181–8. [436] Norouzbahari S, Shahhosseini S, Ghaemi A. J Nat Gas Sci Eng 2015;24:18–25.
[380] Dapprich S, Komaromi I, Byun KS, Morokuma K, Frisch MJ. J Mol Struct: [437] Brenan KE, Campbell SL, Petzold LR. Numerical solution of initial-value problems
THEOCHEM 1999;461–462:1–21. in differential-algebraic equations. Soc Ind Appl Math 1996.
[381] Maseras F, Morokuma K. J Comput Chem 1995;16:1170–9. [438] Gabrielsen J, Michelsen M, Stenby E, Kontogeorgis G. AIChE J 2006;52:3443–51.
[382] Lee C, Yang W, Parr RG. Phys Rev B 1988;37:785–9. [439] Billet , Schultes M. Chem Eng Technol 1993;16:1–9.
[383] Zhao Y, Truhlar D. Theor Chem Acc 2008;120:215–41. [440] Krevelen DW, Hoftijzer PJ. Rec Trav Chim 1948;67:563–86.
[384] Zhao, Truhlar. Acc Chem Res. 41. 2008. 157–61. [441] Gabrielsen J, Michelsen ML, Stenby EH, Kontogeorgis GM. Ind Eng Chem Res
[385] Moller C, Plesset MS. Phys Rev 1934;46:618–22. 2005;44:3348–54.
[386] Chu F, Yang L, Du X, Yang Y. Energy 2016;109:495–505. [442] Chu F, Yang L, Du X, Yang Y. Appl Energy 2017;190:1068–80.
[387] Liu GB, Yu KT, Yuan XG, Liu CJ, Guo QC. Chem Eng Sci 2006;61:6511–29. [443] Darde V, Thomsen K, Well WJMV, Bonalumi D, Valenti G, Macchi E. Int J Greenh
[388] Tontiwachwuthikul P, Meisen A, Liu CJ. Chem Eng Sci 1992;47(2):381–90. Gas Control 2012;8:61–72.
[389] David Z, Bouchra B, Sabine R, Thibaut N, Camel M, Christophe C, et al. Jpn [444] Chu F, Yang L, Du X, Yang Y. Energy 2016;109:495–505.
Memb Sci 2015;493:106–19. [445] Liu GB, Yu KT, Yuan XG, Liu CJ, Guo QC. Chem Eng Sci 2006;61:6511–29.
[390] Li JI, Chen BH. Sep Purif Technol 2005;41:109–22. [446] Pham DA, Lim Y, Jee H, Ahn E, Jung Y. Chem Eng Sci 2015;132:259–70.
[391] Kreulen H, Smolders CA, Versteeg GF, Vanswaaij WPM. Chem Eng Sci [447] Ma S, Chen G, Zhu S, Han T, Yu W. Appl Energy 2016;162:354–62.
1993;48:2093–102. [448] Zhang M, Guo Y. Appl Energy 2013;111:142–52.
[392] Boributh S, Assabumrungrat S, Laosiripojana N, Jiraratananon R. J Membr Sci [449] SarkarA Pan W, Suh DM, Huckaby ED, Sun X. Powder Technol 2014;265:35–46.
2011;380:21–33. [450] Chikukwa A, Enaasen N, Kvamsdal HM, Hillestad M. Energy Procedia
[393] Yang MC, Cussler EL. AIChe J 1986;32:1910–6. 2012;23:82–91.
[394] Mores P, Scenna N, Mussati S. Post-combustion CO2 capture process: equilibrium [451] Pandya JD . Chem Eng Commun. 19. 34–61.
stage mathematical model of the chemical absorption of CO2 into monoethano- [452] Mario L, Eduard A. Int J Greenh Gas Control 2015;42:258–87.
lamine (MEA) aqueous solution. Chem Eng Res Des 2011;89:1587–99. [453] Puxty , Rowland R, Attalla M. Energy Proc 2011;4:1369–76.
[395] Aboudheir A, Tontiwachwuthikul P, Chakma A, Idem R. Kinetic of reactive [454] Wang S, Wang Q, Chen J, Liu G, Lu H. L Sun Fuel 2015.
absorption of carbon dioxide in high CO2-loaded, concentrated aqueous mono- [455] Savage D, Sartori G, Astarita G. Chem Soc 1984;77:17–31.
ethanolamine solutions. Chem Eng Sci 2003;58:5195–210. [456] Behr P, Maun A, Deutgen K, Tunnat A, Oeljeklaus G, Görner K. Energy Procedia
[396] Onda K, Takeuchi H, Okumoto Y. Mass transfer coefficients between gas and 2011:85–92.
liquid phases in packed column. J Chem Eng Jpn 1968;1(1):56–61. [457] Tseng P, Ho W, Savage D. AIChE J 1988;6:922–31.
[397] Wilson I. Gas liquid contact area of random and structured packing [M.S. Thesis]. [458] Ghosh UK, Kentish SE, Stevens GW. Energy Procedia 2009;1:1075–81.
Texas: University of Texas at Austin; 2004. [459] Cullinane JT, Rochelle GT. Chem Eng Sci 2004;17:3619–30.
[398] Bravo JL, Fair JR. Generalized correlation for mass transfer in packed distillation [460] Cullinane JT, Rochelle GT. Ind Eng Chem Res 2006;8:2531–45.
columns. Ind Eng Chem Process Des Dev 1982;21:162. [461] Shen S, Feng X, Zhao R, Ghosh UK, Chen A. Chem Eng J 2013:478–698.
[399] Austgen DM, Rochelle GT. Model of vapor–liquid equilibria for aqueous acid gas– [462] Nicholas NJ, da Silva G, Kentish S, Stevens GW. Ind Eng Chem Res
alkanolamine systems using the electrolyte-NRTL equation. Ind Eng Chem Res 2014;8:3029–39.
1989;28:1060–73. [463] Knuutila H, Juliussen O, Svendsen HF. Chem Eng Sci 2010;23:6077–88.
[400] Tobienssen A, Juliussen O, Svendsend H. Chem Eng Sci 2008;63:2641–56. [464] Ma T, Wang J, Du Z, Abdeltawab AA, M Al-Enizi A, Chen X, et al. Int J Greenh
[401] Freguia S, Rochelle GT. AIChE J 2003;49:1676–86. Gas Control 2017;58:2223–31.
[402] Dugas ER. [M.S. Thesis]. University of Texas at Austin; 2006. [465] Guo H, Zhou Z, Jing G. Int J Greenh Gas Control 2013;16:197–205.
[403] Levenspiel. Chemical reaction engineering. 3rd ed. 1999. [466] Zhou ZM, Jin GH, Zhou LJ. Chem Eng J 2012;204–206:235–43.
[404] Aroonwilas , Chakma A, Tontiwachwuthikul P, Veawab A. Chem Eng Sci [467] Camper D, Becker C, Koval C, Noble R. Ind Eng Chem Res 2005;44:1928–33.
2003;58:4039–53. [468] Ahmady A, Hashim MA, Arona MK. J Chem Eng Data 2010;55:5733–8.
[405] Higbie . Trans Am Inst Chem Eng 1935;31:365. [469] Gao X, Zhou I, Zhou Z. Chem Eng J 2012;181:85–92.
[406] Kucka L, Muller I, Kenig EY, Gorak A. Chem Eng Sci 2003;58:3571–8. [470] Wu ZK, Zhang Y, Lei WK, Yu P, Luo YB. Chem Eng J 2015;264:744–52.
[407] Al-Baghli NA, Pruess SA, Yesavage VF, Selim MS. Fluid Phase Equilib [471] Chen C, Liu B, Liu G, Chai C. Tianjin Sc. Tech. Press, Tianjin; 2011.
2001;185:31–43. [472] Gao X, Ding XL, Du Z. Appl Energy 2010;87:2647–51.
[408] Danckwerts PV, Sharma MM. Chem Eng 1987:244–80. [473] Chu H, Chien TW, Li SY. Sci Total Environ 2001;275:298–300.
[409] Kvamsdal M, Zakeri A, Einbu A, Chikukwa A, Hillestad M. Energy Proc [474] Garcia-Gutierrez P, Jacquemin J, MrCrellis C, Dimitriou I, Taylor SFR, Hardacre
2011;4:1526–33. C. RWK Allen Energy Fuels 2016;30(6):5052–64.
[410] Hoff KA, Juliussen O, Falk-Pedersen O, Svendsen HF. Ind Eng Chem Res [475] Aspen technology. Aspen physical property system – physical property models.
2004;43:4908–21. Brlington, USA; 2013.
[411] Harun N, Douglas PL, Ricardez-Sandoval L, Croiset E. Energy Proc [476] Faiz R, Al-Marzouqi M. Sep Purif Technol 2011;76:351–61.
2011;1(4):1478–85. [477] Orhan OY, Tankal H, Kayi H, Alper E. Int J Greenh Gas Control 2016;49:379–86.
[412] Lawal , Wang M, Stephenson P, Koumpouras G, Yeung H. Fuel [478] Headgordon M, Pople JA, Frisch MJ. Chem Phys Lett 1988;153:503–6.
2010;89:2791–801. [479] Frisch MJ, Headgordon M, Pople JA. Chem Phys Lett 1990;166:275–80.
[413] Ma’mun S, Svendsen HF, Hoff KA, Juliussen O. Energy Convers Manag [480] Møller CP, Plesset MS. Phys Rev 1934;46:0618–22.
2007;48:251–8. [481] Paldus J, Cizek J. Chem Phys Lett 1969;3:1–3.
[414] Chen X, Rochelle GT. Chem Eng Res Des 2011;89:1693–710. [482] Scuseria GE, Janssen CL, Schaefer HF. J Chem Phys 1988;89:7382–7.
[415] Aronu UE, Svendsen HF, Hoff KA, Juliussen O. Energy Proced 2009;1:1051–7. [483] Frisch MJ, Schlegel GW, Scuseria HB, Robb GE, Cheeseman MA, Scalmani JR,
[416] Aronu UE, Svendsen HF, Hoff KA. Int J Greenh Gas Control 2010;4:771–5. et al. Revision A1. Wallingford, CT: Gaussian; 2009.
[417] Aronu UE, Hoff KA, Svendsen HF. Chem Eng Res Des 2011;89:1197–203. [484] Dennington R, Keith T, Milliam J. GaussView, Ver. 5.0.9. Shawnee Mission, KS:
[418] Chowdhury FA, Okabe H, Shimizu S, Onoda M, Fujioka Y. Energy Proced Semichem Inc; 2009.
2009;1:1241–8. [485] Kieffer R, Charcosset C, Puel F, Mangin D. Comput Chem Eng 2008;32:1325–33.
[419] Chowdhury FA, Yamada H, Higashii T, Goto K, Onoda M. Ind Eng Chem Res [486] Rezakazemi M, Razavi SMR, Albadarin AB, Shirazian S. Chem Eng Proc 2016.
2013;52:8323–31. [487] Sohrabi MR, Marjani A, Moradi S, Davallo M, Shirazian S. Appl Math Model
[420] Idem R, Wilson M, Tontiwachwuthikul P, Chakma A, Veawab A, Aroonwilas A, 2011;35:174–88.
et al. Ind Eng Chem Res 2006;45:2414–20. [488] Shirazian S, Ashrafizadeh SN. Ind Eng Chem Res 2011;50:2245–53.
[421] Gabrielsen J, Svendsen HF, Michelsen ML, Stenby EH, Kontogeorgis GM. Chem [489] Shirazian S, Moghadassi A, Moradi S. Simul Model Pract Theory 2009;17:708–18.
Eng Sci 2007;62:2397–413. [490] Lin SH, Chiang PC, Hsieh CF, Li MH, Tung KL. J Chin Inst Chem Eng
[422] Naami A, Edali M, Sema T, Idem R, Tontiwachwuthikul P. Ind Eng Chem Res 2008;39:13–21.
2012;51:6470–9. [491] Ghaemi A, Shahhosseini S, Maragheh MG. Chem Eng J 2009;149:110–7.
[423] Bishnoi S, Rochelle GT. Chem Eng Sci 2000;55:5531–43. [492] Jayarathna A, Lie B, Melaaen MC. Int J Greenh Gas Control 2013;4:282–90.

1106
I. Sreedhar et al.
[493] Jamal , Meisen A, Lim CJ. Chem Eng Sci 2006;61:6590–603.
[494] Zhang Y, Chen H, Chen C, Plaza JM, Dugas R, Rochelle GT. Ind Eng Chem Res
2009;48:9233–46.
[495] Kenig EY, Schneider R, Gorak A. Chem Eng Sci 1999;54:5195–203.
[496] Astarita G, Savage DW, Bisio A. Gas treating with chemical solvents. New York:
Wiley; 1983.
[497] Gao XY, Liu L, Hu ML, Xiang Y, Chu GW, Zou HK, et al. Chem Eng Proc 2016.
[498] Gatti M, Martelli E, Marechal F, Consonni S. Appl Therm Eng 2014;70:1123–40.
[499] Hasib-Ur-Rahman M, Siaj M, Larachi F. Int J Greenh Gas Control
2012;6:246–52.
[500] Coulier Y, Lowe A, Tremaine PR, Coxam JY, Ballerat-Busserolles K. Absorption of
CO2 in aqueous solutions of 2-methylpiperidine: heats of solution and modelling.
Int J Greenh Gas Control 2016;47:322–9.
[501] Rezaei B, Riahi SJ. Nat Gas Sci Eng 2016;33:388–96.
[502] Zhao B, Tao W, Zhong M, Su Y, Cui G. Renew Sustain Energy Rev 2016;65:44–56.
[503] Yi F. Beijing University of Chemical Technology. [In Chinese]. 2008.
[504] Qian Z, Xu L, Cao H, Guo K. Ind Eng Chem Res 2009;48:9261–7.
[505] Xu M, Zhang JW, Shen ZG, Chen JF, Zhao J. J Beijing Univ Chem Technol
2004;31:30–5, [In Chinese].
[506] Saha D. Heat Mass Transf 2009;45:451–7.
[507] Zhao B, Su Y, Tao W. Appl Energy 2014;136:132–42.
[508] Escobillana GP, Saez JA, Pérez-Correa JR, Neuburg HI, Kershenbaum LS. Can J
Chem Eng 1991;69:969–77.
1107
Renewable and Sustainable Energy Reviews 76 (2017) 1080–1107
[509] Vanichseni S. [Ph.D. thesis]. Imperial College, London; 1977.
[510] Tontiwachwuthikul P. [Ph.D. thesis]. University of British Columbia, Vancouver,
Canada; 1990.
[511] Pintola , Tontiwachwuthikul P, Meisen A. Gas Sep Purif 1993;7:47–52.
[512] Alatiqi , Sabri MF, Bouhamra W, Alper E. Gas Sep Purif 1994;8:3–11.
[513] Kucka , Richter J, Kenig EY, Górak A. Sep Purif Technol 2003;31:163–75.
[514] Kucka , Müller I, Kenig EY, Górak A. Chem Eng Sci 2003;58:3571–8.
[515] Aroonwilas , Tontiwachwuthikul P, Chakma A. Sep Purif Technol
2001;24:403–11.
[516] DeMontigny D, Aboudheir A, Tontiwachwuthikul P, Chakma A. Ind Eng Chem Res
2006;45:2594–600.
[517] Liu GB, Yu KT, Yuan XG, Liu CJ, Guo QC. Chem Eng Sci 2006;61:6511–29.
[518] Dugas RE. [M.S.E. Thesis]. The University of Texas at Austin; 2006.
[519] Dugas R, Alix P, Lemaire E, Broutin P, Rochelle G. Energy Proc 2009;1:103–7.
[520] Afkhamipour , Mofarahi M. Int J Greenh Gas Control 2013;15:186–99.
[521] Afkhamipour , Mofarahi M. Int J Greenh Gas Control 2014;25:9–22.
[522] Sreedhar I, Vaidhiswaran R, Bansi MK, Venugopal A. Renew Sustain Energy Rev
2017;68(1):659–84.
[523] Li S, Pyrzynski TJ, Klinghoffer NB, Tamale T, Zhong Y, Aderhold JL, et al. J Memb
Sci 2017;527:92–101.
[524] Sreenivasulu B, Sreedhar I, Reddy BM, Raghavan KV. Energy and Fuels
2017;31:785–94.