International Journal of Greenhouse Gas Control 81 (2019) 83–92

Contents lists available at ScienceDirect

International Journal of Greenhouse Gas Control

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

Experimental investigation of CO2 solubility and its absorption rate into T

promoted aqueous potassium carbonate solutions at elevated temperatures
Dimitrios Grimekisa, , Sotirios Giannoulidisa, Konstantina Manoua, Kyriakos D. Panopoulosb,
⁎

Sotirios Karellasa
a
Laboratory of Steam Boilers and Thermal Plants, National Technical University of Athens (NTUA), 9 Heroon Polytechniou Str., 15780, Zografou, Athens, Greece
b
Chemical Process and Energy Resources Institute, Centre for Research and Technology Hellas (CERTH), 57001, Thermi, Thessaloniki, Greece

ARTICLE INFO ABSTRACT

Keywords: Aqueous potassium carbonate solutions are suitable for capturing CO2 from flue gases and process gases with a
Carbon dioxide solubility low heat duty for regeneration. However, their absorption rate and capacity can be further improved by pro-
Potassium carbonate moters. In this study, an experimental investigation of CO2 solubility and absorption rate into pure and promoted
MEA potassium carbonate solutions in a batch stirred reactor at elevated temperatures is presented. Measurements for
Piperazine
pure K2CO3 solvents were carried out at 80 °C, 100 °C and 120 °C for 1.81 M and 2.53 M potassium carbonate
MDEA
concentrations. Solubility results were successfully validated against electrolyte-NRTL model predictions. The
Glycine
experiments were also used to provide a reference for evaluating the effect of promoter addition on CO2 solu-
bility and absorption rate. Subsequently, four additives, namely Methyl diethanolamine (MDEA),
Monoethanolamine (MEA), Piperazine (PZ) and Glycine were tested at various concentrations in a 1.81 M po-
tassium carbonate solution at a temperature range of 100–120 °C. Results indicated an increase in solvent ca-
pacity and absorption rate for PZ- and MEA-promoted K2CO3 solutions. Glycine addition benefits the absorption
rate mostly at low loadings, although the solvent capacity is not affected positively. MDEA has a detrimental
effect on solvent capacity at higher concentrations, while no positive effect on absorption rate was found.

1. Introduction solution as a result of providing heat to the reboiler and/or due to

pressure reduction. For the chemical absorption of CO2, types of amines
Carbon dioxide (CO2) separation processes based on absorption and potassium carbonate are widely investigated. Solvent screening and
have gained interest over the last decades due to climate change future evaluation is essential in order to determine the suitability of different
projections (Sreedhar et al., 2017). Apart from post-combustion appli- liquid solvents with respect to a number of criteria, such as solvent
cations, CO2 can be captured from gas streams prior to the production capacity, absorption rate, required reboiler duty for regeneration,
of synthetic fuels (Atsonios et al., 2013; Koytsoumpa et al., 2015). toxicity, corrosion potential, cost and thermochemical stability
During substitute natural gas (SNG) production, intermediate process (Dawodu and Meisen, 1996). Specific types of amines are considered for
steps, such as methanation, require CO2 removal from syngas after their enhanced solvent capacity and increased absorption rates (Wang
gasification to reduce the hydrogen requirement for syngas conversion et al., 2011). However, the utilization of amine solutions is associated
to methane, while inorganic and organic sulfur removal is needed for with certain drawbacks such as high reboiler duty, material corrosion
catalyst protection (Edinger et al., 2017; Neubert et al., 2017). Fur- and solvent degradation issues. Aqueous solutions of potassium carbo-
thermore, the SNG content in acidic gases (such as CO2 and H2S) must nate can overcome several of these problems. Operation at elevated
be low before the gas is fed to the grid. The state-of-the-art option for temperatures, i.e. close to the atmospheric boiling point, is beneficial
acid gas removal is by liquid solvents (Koytsoumpa et al., 2015), which for reduced energy requirement for regeneration (Ye and Lu, 2014),
is carried out by gas-liquid contacting equipment. Two separation col- while potassium carbonate is thermochemically stable and non-volatile
umns are necessary, one for acid gas absorption and one for acid gas (Kothadaraman, 2010). In addition, K2CO3 solutions are characterized
recovery. In the absorber, CO2 is transferred from the gas to the liquid by a low corrosion potential (Hu et al., 2016). On the other hand, one of
phase, while in the desorption column, CO2 is stripped out of the the major drawbacks of aqueous potassium carbonate solutions is the

⁎
Corresponding author at: Laboratory of Steam Boilers and Thermal Plants, Building “O”-School of Mechanical Engineering, National Technical University of
Athens (NTUA), 9 Heroon Polytechniou Str., 15780, Zografou, Athens, Greece.
E-mail address: [email protected] (D. Grimekis).

https://doi.org/10.1016/j.ijggc.2018.12.008
Received 10 July 2018; Received in revised form 16 November 2018; Accepted 11 December 2018
1750-5836/ © 2018 Elsevier Ltd. All rights reserved.
D. Grimekis et al. International Journal of Greenhouse Gas Control 81 (2019) 83–92

low CO2 absorption rate compared to primary amines (Borhani et al., however, PZ addition to the potassium carbonate solution induces a
2015b). depression of the CO2 equilibrium partial pressure. Kim et al. (Kim
Measurements of CO2 equilibrium solubility in concentrated aqu- et al., 2012) carried out a study on CO2 solubility in PZ- and 2-me-
eous potassium carbonate solutions at temperatures ranging from 110 thylpiperazine(2-MPZ)-promoted K2CO3 aqueous solutions over a
to 140 °C and 70–110 °C have been published for a conversion of K2CO3 temperature range of 40–80 °C in a batch stirred reactor. Increasing the
to KHCO3 from 0 to more than 80% (Benson, 1956; Tosh et al., 1959). promoter concentration had a greater impact on CO2 solubility com-
Water vapor pressure was found to be nearly constant, except when the pared to increasing potassium ions in the mixture. Over the investigated
temperature exceeded 120 °C, which caused an initial increase and a temperature range, the absorption rate of CO2 was similar for aqueous
sharp decrease for conversions higher than 40%. A sharper increase in K2CO3-2-MPZ and K2CO3-PZ.
CO2 partial pressure for more concentrated mixtures was noted, while Glycine (C2H5NO2) and other amino acids have not been studied
equilibrium constants were found to be nearly constant for more dilute extensively as promoters of potassium-containing solvents. For glycine
solutions for each pair of concentration and temperature. Recently, as a promoter, laboratory scale results for the kinetics of absorption at
experimental test rigs such as stirred reactors (Mazinani et al., 2015) 40 °C, 60 °C and 80 °C can be found in Thee et al. (Thee et al., 2014) and
and wetted-wall columns (Thee et al., 2012b) have been used in studies results from a small-scale pilot plant have been published by Smith
dealing with CO2 absorption into liquid solvents. et al. (Smith et al., 2015). Moreover, studies on different types of amino
Studies of CO2 absorption into aqueous potassium carbonate acids in potassium-containing solutions are those published by Aronu
(Bohloul et al., 2014; Park et al., 1997) and aqueous amine solutions (Li et al. (Aronu et al., 2010) and Shen et al. (Shen et al., 2013). For the
et al., 2015; Singh et al., 2011) can be found, while recent investiga- case of MDEA, measurements at inlet CO2 partial pressures in the range
tions greatly focus on mixtures of amines or amine/potassium carbo- 0–20 kPa and at ambient temperature were published by Nii and Ta-
nate blends in order to seek favorable characteristics such as low heat keuchi (Nii and Takeuchi, 1994).
duty for regeneration, fast absorption rates and high solvent capacity From the aforementioned literature, it can be concluded that further
(Cullinane, 2005). The maximum concentration of potassium carbonate research is needed for the application of promoted aqueous K2CO3 so-
is typically limited to 50 wt.% K2CO3 or about 5.5 M K2CO3 based on lutions at “warm” gas cleaning conditions (i.e. at higher temperatures).
published density measurements (Knuutila et al., 2010). A higher Since K2CO3 solutions are particularly suitable for high-temperature
concentration must be avoided due to the low bicarbonate (KHCO3) absorption (Kohl and Nielsen, 1997), the suitability of amines for long-
solubility, which can cause precipitation and fouling problems. In term operation needs particular attention. Oxidative degradation may
practice, moderate concentrations of 20–40 wt.% or maximum molar be observed in the absorber column when flue gas is processed, while
concentrations up to 4 M K2CO3 are considered more appropriate. thermal degradation usually occurs in the stripper section. According to
Promoter concentrations for amines are usually in the 1–10 wt.% range Rochelle (Rochelle, 2012), thermal degradation is rapid in the
(Borhani et al., 2015b; Hu et al., 2016), while for amino acid promoters, 120–150 °C temperature range for most amines. Thus, the upper tem-
experimental investigations have also targeted a wide concentration perature limit for the current investigation was set to 120 °C, which is
range (Thee et al., 2014). known to be an acceptable value for units operating with MEA, taking
According to Thee et al. (Thee et al., 2012b), the addition of 1.1 M into account the amine make-up requirement. A short overview of
and 2.2 M MEA in an aqueous solution of 30 wt.% K2CO3 at 63 °C can studies related to amine thermal degradation for various temperature
increase the absorption rate by a factor of 16 and 45 respectively. It is ranges is provided in Table 1. Corrosion, toxicity and other major issues
also stated that the effect of primary amines is more significant com- which need to be addressed for the safe and reliable operation of ab-
pared to secondary or tertiary amines. In a recent study, Ramazani et al. sorption units containing amines are discussed by Hu et al. (Hu et al.,
(2016) measured the solubility of CO2 in MEA-promoted K2CO3 solu- 2016).
tions with three molar ratios and a total concentration of 2.5 M at 40 °C In this study, the effect of four additives on CO2 solubility and ab-
and concluded that a high molar ratio of MEA to potassium carbonate is sorption rate into aqueous K2CO3 solutions was investigated. A first
not beneficial for CO2 solubility (expressed as the ratio of CO2 mole- experimental campaign was performed with 1.81 M and 2.53 M aqu-
cules in the liquid phase divided by the total solvent molecules in the eous potassium carbonate solutions (corresponding to 25 g and 35 g
solution). However, this molar ratio increase was found to positively K2CO3 per 100 mL of solution, respectively) at 80–120 °C in a stirred
influence the CO2 absorption rate. On the other hand, it was concluded cell reactor and the results were compared and validated with VLE
that the addition of 2-(1-piperazinyl)-ethylamine (PZEA), piperazine results obtained through thermodynamic modeling of the CO2-H2O-
(PZ) and tetraethylenepentamine (TEPA) in aqueous potassium carbo- K2CO3-KHCO3 system with the electrolyte-NRTL model in Aspen Plus®
nate solutions results to an increase in CO2 specific absorption capacity. software. After assessing these preliminary results, which were also
TEPA was proposed as the most efficient promoter for increasing both used as a reference for comparison, the effect of a primary amine
solubility and CO2 absorption rate at ambient temperature. Concerning (Monoethanolamine, MEA), a tertiary amine (Methyl diethanolamine,
PZ as a promoter, Cullinane and Rochelle (Cullinane and Rochelle, MDEA), a diamine (Piperazine, PZ) and an amino acid (Glycine, Gly) on
2005) presented a vapor-liquid equilibrium (VLE) study to demonstrate CO2 solubility was evaluated at elevated temperatures (100 °C and
the effect of potassium to piperazine ratio on CO2 solubility. For the 120 °C). Because most existing studies have investigated such mixtures
promoted solvent, they found that the trend of the solubility curves is at lower temperatures (Cullinane and Rochelle, 2004; Hu et al., 2016;
similar to the one obtained for aqueous PZ and aqueous K2CO3, Kim et al., 2012; Lee et al., 2016; Nii and Takeuchi, 1994; Ramazani

Table 1
Thermal degradation studies of investigated amines.
Compound Temperature (oC) Thermal degradation Source
potential

MEA 111-125 First order rate constant (Davis, 2009)

k1 = 2.9·10−8 s−1
Piperazine 162 First order rate constant k1 = 2.9·10−8 s−1 (Freeman, 2011)
MDEA 119-129 First order rate constant k1 = 2.9·10−8 s−1 (Rochelle, 2012)
Glycine (Na salts) 125-145 More than an order of magnitude higher than MEA (Huang et al., 2013)

84
D. Grimekis et al. International Journal of Greenhouse Gas Control 81 (2019) 83–92

et al., 2016; Thee et al., 2014, 2012b), the present investigation focused CO2 + RNH2 RNH+2 COO (8)
on higher temperatures for which solubility results are scarce for those
promoters (Cullinane and Rochelle, 2005), while this could be an ad-
RNHCOO + H2 O RNH2 + HCO3 (9)
vantage for the integration of a K2CO3-based absorption process in a hot RNH+3 + H2 O RNH2 + H3 O+ (10)
gas cleaning configuration. Furthermore, the current investigation
provides experimental results for CO2 solubility in Gly-promoted K2CO3 The promotion mechanism for piperazine (PZ, C4H10N2) can simi-
solutions for which very few studies are available (Lee et al., 2015). larly be described through carbamate formation. However, PZ is a
Absorption rates are also compared among promoted solutions to pro- diamine and thus both forms, PZCOO− and PZ(COO−)2, are present.
pose efficient mixtures for high-temperature CO2 absorption. The reactions for CO2 absorption into PZ-promoted K2CO3 solutions are
given in Eqs. ((11)–(14)) with the addition of Eqs. (5–7). In this case,
2. Reaction mechanisms Eqs. (6), (11) and (13) are rate-limiting steps (Sreedhar et al., 2017).
Further details can be found in the study of (Cullinane and Rochelle,
2.1. Reaction mechanism for CO2 absorption into aqueous pure K2CO3 2004).
solutions PZ+ H2 O+ CO2 PZCOO + H3 O+ (11)

The overall chemical reaction of the aqueous pure K2CO3 solution PZH+ + H2 O PZ+ H3 O+ (12)
with CO2 takes place according to Eq. (1). PZCOO + H2 O+ CO2 PZ(COO ) 2 + H3 O+ (13)
K2 CO3(l) + H2 O(l) + CO2(aq) 2KHCO3(aq) (1)
H+PZCOO + H2 O PZCOO + H3O+ (14)
The mechanism is described through the physical solubility of CO2 When Glycine (Gly, NH2CH2COOH) is added in aqueous solutions, it
and hydrolysis, potassium carbonate dissolution in water, carbonate can be present in a protonated (Gly+), zwitterionic ( Gly+) and anionic
hydrolysis, carbonate and bicarbonate formation, as well as water io- form ( Gly ) (Thee et al., 2014). At sufficiently high pH values, the
nization (Borhani et al., 2015b). Since potassium carbonate and po- concentration of the protonated form is very low. The reaction me-
tassium bicarbonate are both strong electrolytes, the metal is present chanism can be described through the protonation of the anionic form,
only in the form of potassium ions. The carbonate ion reacts with water the formation of a carbamate and the carbamate hydrolysis to bi-
to generate hydroxyl ions, which then react with CO2. A list of these carbonate, as shown in Eqs. ((15)–(17)), respectively. The reaction
reactions is presented in Table 2. For a pH value higher than 9, the mechanism when glycine is used with aqueous potassium carbonate is
reaction shown in Eq. (6) is the rate-limiting step, while the chemical further analyzed in the study of (Lee et al., 2016).
reaction of Eq. (5) is instantaneous.
Gly + H3 O+ Gly + + H2 O (15)
2.2. Reaction mechanisms for CO2 absorption into promoted K2CO3 Gly + CO2 + H2 O GlyCOO + H3 O+ (16)
solutions
GlyCOO + H2 O Gly+ HCO3 (17)
Despite the diverse physical and chemical properties of additives,
The CO2 reaction mechanism for MDEA (CH3N(CH2CH2OH)2) dif-
there are also similarities such as having OH or O− groups or acting as
fers in that no carbamate formation takes place as in the previous cases.
Lewis bases with CO2 being a Lewis acid (Dennard and Williams, 1966;
Instead, it is based on the generation of hydroxyl anions through MDEA
Phan et al., 2014). Previous studies have suggested a variety of pro-
hydrolysis and the kinetic reaction of MDEA with CO2 and water, as
motion mechanisms and one of the principal conclusions drawn is that
shown in Eqs. ((18)–(19)), in addition to Eqs. ((5)–(7)). The rate-lim-
promoters can influence the reaction between CO2 and H2O (Hu et al.,
iting reactions are shown in Eq. (6) and Eq. (19). A more thorough
2016). Astarita et al.(Astarita et al., 1981) presented a generalized
description of the CO2 reaction mechanisms in aqueous K2CO3 solutions
model to explain the effect of specific amines or inorganic promoters
with MDEA, as well as aqueous solutions with MDEA, is given by
through the formation of an intermediate species due to the reaction of
(Borhani et al., 2015a) and (Pacheco and Rochelle, 1998), respectively.
the promoter with CO2.
Concerning the additives used in this study, reaction mechanisms MDEA+ H2 O MDEAH+ + OH (18)
have been studied and proposed in the literature. Monoethanolamine
(MEA, NH2CH2CH2OH) can react with CO2 to form a zwitterion inter- MDEA+ CO2 + H2 O MDEAH+ + HCO3 (19)
mediate which subsequently loses a proton to form a carbamate. The
reaction mechanism consists of reactions in Eqs. ((8)–(10)) with the 3. Experimental section
addition of Eqs. (5–7) and that of the deprotonation of the zwitterion
intermediate with free MEA or other compounds which act as bases in 3.1. Materials and solvent preparation
the solution. Rate-limiting reaction steps (kinetic reactions) are Eq. (6)
and Eq. (8). Further details can be found in the study of (Thee et al., Nitrogen (N2) with a purity higher than 99.8% and CO2 with a
2012b). purity higher than 99.7% were supplied by Air Liquide Hellas SAGI.
Potassium carbonate (K2CO3, anhydrous, ≥ 99.0% purity), Piperazine
Table 2 (PZ, anhydrous, ≥ 99.0% purity), Monoethanolamine (MEA, ≥ 99.0%
Chemical reactions for CO2 absorption into H2O-K2CO3 solutions.
purity), Glycine (Gly, anhydrous, ≥ 99.0% purity) and Methyl die-
Reactions thanolamine (MDEA, ≥ 99.0% purity) were supplied by Merck Hellas
SA. All gases and chemicals were used “as-received”, i.e. without fur-
CO2(g) CO2(aq) (2)
ther purification. Deionized water was used for the preparation of
CO2(aq) + 2H2 O(l) HCO3(aq) + H3O+ (3)
(aq) mixtures and volumetric measurements were performed with a grad-
2K + 2 (4)
K2CO3(s) (aq) + CO3(aq) uated cylinder ( ± 0.5 mL accuracy). Moreover, a precision balance
CO23(aq) + H2 O(l) HCO3(aq) + OH (aq) (5) (Kern-ALJ 220-5DNM) with an accuracy of ± 0.0001 g was used. The
OH (aq) + CO2(aq) HCO3(aq) (6) total solution volume was constant (400 mL) for all experiments.
2H2 O(l) H3O+
(aq) + OH(aq)
(7) Mixtures of PZ/K2CO3 and MEA/K2CO3 were tested at 100 °C and
120 °C, while MDEA- and Gly-containing mixtures were evaluated at

85
D. Grimekis et al. International Journal of Greenhouse Gas Control 81 (2019) 83–92

Table 3 distribution in the liquid phase and simultaneously reduce the time
Temperature and concentration test matrix of investigated pure and promoted interval to equilibrium. A back pressure regulator was installed at the
K2CO3 solutions. pipeline exit for pressure adjustment and safety purposes. Pressure and
K2CO3 Promoter type/ Promoter/K2CO3 Temperature (oC) temperature values were continuously monitored and recorded by a
concentration (M) concentration (M) molar ratio (-) data acquisition system.

1.81 No/- – 80, 100, 120

2.53 No/- – 80, 100, 120
1.81 PZ/0.58 0.32 100, 120 3.3. Experimental procedure
1.81 PZ/1.16 0.64 100, 120
1.81 MEA/0.82 0.45 100, 120 Initially, the solvent was transferred to the equilibrium cell and the
1.81 MDEA/0.25 0.14 100
vacuum pump was used to evacuate it at about 400 mbar (absolute
1.81 MDEA/0.84 0.46 100
1.81 Gly/0.67 0.37 100
pressure) and remove air contaminants and oxygen. While still at am-
bient conditions, nitrogen was added to slightly pressurize the equili-
brium cell in order to avoid boiling in the vessel at higher temperatures.
Pressurization at 1.2 bar and 2.0 bar was applied for 100 °C and 120 °C
experiments, respectively, while no additional nitrogen was added for
100 °C. The test matrix for pure K2CO3 and promoted solutions is shown experiments at 80 °C. The stirring rotational speed was constant at
in Table 3. 180 rpm for all tests.
Afterward, the heating phase began and lasted about 3 h in order to
3.2. Experimental test rig and measurement equipment ensure that pressure and temperature had stabilized and an initial VLE
state had been achieved. During the heating of both cells, the pre-
The experimental campaigns were conducted in the laboratory-scale heating vessel was purged with CO2 in order to remove atmospheric air.
test rig depicted in Fig. 1. The apparatus mainly consists of two pressure After purging for 2 min, CO2 was stored in the preheating cell under
vessels, the preheating and the equilibrium cell, which are made of pressure (8 bar maximum storage pressure) and it was heated at the
316L stainless steel. Both cells are cylindrical and have a diameter of desired temperature level for at least 30 min. Subsequently, the valve to
100 mm and a height of 195 mm. Electrically heated jackets were de- the equilibrium cell opened and CO2 was transferred from the pre-
signed to fit the vessels and preserve a constant temperature with an heating to the equilibrium cell within about 1 s, causing a peak in cell
accuracy of ± 0.1 K with PID temperature control. Electric heat tracing pressure at the beginning of each injection. The CO2 solubility was
was additionally installed for flange and tube heating. The pressure was measured by the pressure decay method. After the pressure inside the
measured with two WIKA S-10 pressure transmitters (0–6 bar absolute equilibrium cell reached a maximum value, a pressure decrease fol-
pressure, accuracy ± 0.03 bar), one for each cell. The temperature was lowed, characterized by an initial sharper drop which gradually became
measured with K-type thermocouples, which had been previously ca- smoother until the pressure stabilized. The sequence of injections
librated with a Pt100 temperature sensor in boiling water. In the continued until the desired partial pressure range was scanned.
equilibrium cell, where gaseous and liquid phases co-exist, the tem- Pressure values were recorded every 2 s. An equilibrium state was as-
perature was measured with two thermocouples, one for each phase. A sumed to occur when the pressure measurement deviation from the
diaphragm vacuum pump (KNF Laboport N86 KT.18), which is capable average value recorded was not greater than 0.01 bar for a time interval
of establishing 100 mbar pressure (absolute), was connected to the of 40 min. All the major steps of the experimental procedure are
piping system for residual gas removal. Furthermore, a magnetic hot- highlighted in Fig. 2, which presents a typical pressure profile recorded
plate stirrer was used in order to ensure a uniform temperature inside the equilibrium cell.

Fig. 1. Schematic diagram of the apparatus for solvent screening (C1: preheating cell, C2: equilibrium cell, PR: pressure regulator, BPR: back pressure regulator, VP:
diaphragm vacuum pump, MS: magnetic stirrer, P: pressure transmitter, T: thermocouple).

86
D. Grimekis et al. International Journal of Greenhouse Gas Control 81 (2019) 83–92

Fig. 2. Pressure profile and experimental procedure for CO2 solubility measurement.

3.4. CO2 solubility calculation and error analysis *

PCO V
2 g, eq
nCO2, ge =
RTeq (23)
The CO2 solubility in a liquid solvent can be derived by con-
tinuously or semi-continuously analyzing the CO2 content of the gas- where Vg , eq is the gas-phase volume of the equilibrium cell and Teq is the
eous or liquid phase (Freeman et al., 2010; Shen et al., 2013). If an temperature recorded at equilibrium conditions.
analyzer is not to be used, the solubility can be derived by comparing For the interpretation of results, CO2 solubility denoted as CCO2 (mol
the initial to the equilibrium pressure in a batch feeding mode (Li et al., CO2/L of initial solvent) is used since it provides a comparative measure
2014; Ramazani et al., 2016). After sufficient time after the injection of of solvent capacity per unit volume. It is defined as the ratio of the
CO2 into the absorption cell (which contains the liquid under study) has amount of CO2 absorbed, nCO2, j, sol , to the initial solvent volume Vs , as
elapsed, an equilibrium state is achieved. The CO2 equilibrium pressure shown in Eq. (24).
(PCO
*
2
) can be derived as:
nCO2, j, sol
* CCO2 =
PCO 2
= Peq Ps (20) Vs (24)

where Peq is the total equilibrium cell pressure after a CO2 injection j For the solution loading definition, different formulas can be found
and Ps is the total equilibrium cell pressure before the first CO2 injec- in the literature (Bohloul et al., 2014; Kim et al., 2012). Herein, solution
tion. The value of this residual pressure Ps was assumed to remain loading (α) is defined as the ratio of the CO2 quantity absorbed (in
constant throughout the course of each experiment that adds con- moles) to the sum of the equivalent K2CO3 quantity (in moles), n K2 CO3 ,
secutive amounts of CO2. and the amount of promoter (in moles), npromoter , if any (thus excluding
The quantity of CO2 contained in the solution after each injection j water moles from this definition).
is described by Eq. (21).
nCO2, j, sol
nCO2, j, sol = nCO2, j + nCO nCO2, ge a=
1, sol 2, gs (21) n K2 CO3 + npromoter (25)

It is, therefore, the sum of the existing amount in solution, nCO2, j 1, sol In order to estimate the uncertainty of solubility values which are
after the (j-1) injection and equilibrium state, and the current difference reported in this work, an error propagation analysis was carried out
between the CO2 quantity initially present in the gas phase, nCO2, gs and taking into account available uncertainties of volume, pressure and
the final equilibrated gas-phase CO2 quantity, denoted as nCO2, ge . temperature measurements for the preheating and equilibrium cell, as
The expression for the transferred CO2 moles from the preheating to well as volume measurement accuracy for the liquid solvent. The un-
the equilibrium cell, nCO2, t , can be formulated as shown in Eq. (22): certainty estimation expression given by Moffat (Moffat, 1988) was
used, which is shown in Eq. (26):
Pb Vpre Pa Vpre
nCO2, t =
RTb RTa (22) N
1
2 2
CCO2
CCO2 = xi
where Vpre is the total volume of the preheater cell, Pb, Tb are the i=1
xi (26)
pressure and temperature of the preheater before the valve opens and
Pa, Ta are the pressure and temperature of the preheater after the valve where N is the number of all x i variables subject to measurements and
has been closed and indications have stabilized. For the nCO2, gs calcu- x i is the measurement uncertainty for each variable. The solubility
lation on intermediate points, the remaining CO2 moles from the pre- measurement uncertainty was calculated to be ± 4% arising from the
vious injection are also considered. accuracy of the equipment used. Pressure measurement accuracy sig-
When equilibrium is achieved, the quantity of CO2 in the gas phase nificantly contributed to the overall uncertainty result, as also noted in
of the equilibrium cell (nCO2, ge) is given by Eq. (23): other studies (Bohloul et al., 2014).

87
D. Grimekis et al. International Journal of Greenhouse Gas Control 81 (2019) 83–92

Fig. 3. Experimental results and electrolyte-NRTL model predictions with default pair interaction parameters for the 1.81 M K2CO3 aqueous solution at (a) 80 °C and
(b) 120 °C.

4. Results and discussion parameters of the model were determined through data regression in
Aspen Plus® software according to the methodology and experimental
4.1. Test rig reliability evaluation: comparison with default and regressed data sets described by Hilliard (Hilliard, 2008). Apart from CO2 solu-
electrolyte-NRTL model results bility data from Tosh et al. (Tosh et al., 1959), the regressed model
includes experimental data for the mean ionic activity coefficient, vapor
The reliability of the experimental test rig was evaluated by mea- pressure depression and heat capacity of the solution.
suring CO2 solubility in pure K2CO3 aqueous solutions. The results were As shown in Fig. 4, the experimental results of CO2 solubility in the
compared against solubility predictions from the electrolyte-NRTL 2.53 M K2CO3 aqueous solution agree well with the regressed electro-
model coupled with the Soave-Redlich-Kwong (SRK) equation of state. lyte-NRTL model results. The relative deviations were smaller com-
This model was developed for liquid solutions that contain electrolytes pared to those obtained with the default Aspen Plus® parameters, which
and describes the excess Gibbs energy by taking into account long-range was more evident when a ranged between 0.2 mol CO2/mol K2CO3 and
ion-ion interactions, mixed solvent composition (Born correction) and 0.4 mol CO2/mol K2CO3 in Fig. 4a, where the regressed model curve fits
local interactions according to the NRTL theory (Chen and Evans, 1986; better the experimental results. The aforementioned comparison vali-
Chen and Song, 2004). The model adjustable parameters can either be dated that the experimental apparatus and measurement procedure can
set to default values available in the Aspen Plus® database or be mod- be considered reliable for the solvent screening process of pure and
ified through data regression analysis as described by Borhani et al. promoted K2CO3 mixtures.
(2015a). Fig. 3 indicates that the experimental results in this study are
in good agreement with the model results for the pure K2CO3 aqueous
4.2. CO2 solubility in pure K2CO3 aqueous solutions
solution when the default Aspen Plus® databank parameters are used.
The relative deviation between the model and experimental results was
The experimental results for the aqueous K2CO3 solutions are pre-
particularly small for the lower solution concentration (1.81 M K2CO3).
sented in Table 4 as pairs of equilibrium CO2 partial pressure and CO2
In the case of 2.53 M K2CO3 aqueous solution, the relative deviation
solubility values, represented as the quantity of CO2 absorbed (moles)
between predicted electrolyte-NRTL and experimental results was more
per liquid solution volume (L). The solvent capacity increased when
pronounced and thus, a regression procedure was performed to adjust
potassium carbonate concentration increased from 1.81 M to 2.53 M for
the molecule-electrolyte binary parameters of the activity coefficient
the same equilibrium CO2 partial pressure and temperature. On the
model based on experimental data from the literature. Apart from its
other hand, when solution loading was used (Fig. 5), defined as the
importance for predicting VLE curves with a higher accuracy, this
ratio of CO2 amount in the liquid phase (in moles) to the amount of
methodology is widely used when modeling absorption columns with
potassium carbonate (in moles), the loading declined for a higher so-
detailed reaction kinetics, mass transfer correlations and comprehen-
lution concentration. A similar trend was described by Bohloul et al.
sive thermodynamic models for the liquid and gaseous phases (Borhani
(Bohloul et al., 2014). In the aforementioned study, the effect of CO2
et al., 2015a; Koronaki et al., 2015). Therefore, the adjustable
physical solubility in water was compared to the solubility in the K2CO3

Fig. 4. Experimental results and electrolyte-NRTL model predictions with regressed binary interaction parameters for 2.53 M K2CO3 aqueous solution at (a) 80 °C and
(b) 120 °C.

88
D. Grimekis et al. International Journal of Greenhouse Gas Control 81 (2019) 83–92

Table 4
Experimental results of CO2 solubility in pure K2CO3 aqueous solutions.
1.81 M K2CO3/80 °C 1.81 M K2CO3/100 °C 1.81 M K2CO3/120 °C

CCO2 (mol CCO2 (mol CCO2 (mol

2 (kPa)
* * *
PCO PCO 2 PCO 2
CO2/L) (kPa) CO2/L) (kPa) CO2/L)

2.1 0.50 4.2 0.35 3.3 0.33

8.8 0.72 7.0 0.49 6.5 0.40
17.5 0.91 13.1 0.64 10.5 0.46
33.6 1.09 20.4 0.76 12.3 0.52
55.4 1.27 27.7 0.85 14.4 0.57
81.3 1.32 32.4 0.91 18.2 0.63
23.3 0.69
28.2 0.75
32.7 0.80
38.9 0.86
48.0 0.90

2.53 M K2CO3/80 °C 2.53 M K2CO3/100 °C 2.53 M K2CO3/120 °C

CCO2 (mol CCO2 (mol CCO2 (mol

2 (kPa)
* * *
PCO PCO 2 PCO 2
CO2/L) (kPa) CO2/L) (kPa) CO2/L)

5.1 0.71 4.1 0.48 5.4 0.57

9.6 0.91 7.3 0.63 8.1 0.63
16.0 1.10 11.8 0.75 9.4 0.68
24.7 1.26 15.0 0.85 14.3 0.74
32.2 1.39 17.8 0.92 16.5 0.78
50.7 1.56 27.3 1.08 18.9 0.83
75.6 1.71 39.2 1.22 22.4 0.88
Fig. 5. Effect of temperature on CO2 solubility in (a) 1.81 M and (b) 2.53 M
55.3 1.36 25.1 0.92 pure K2CO3 aqueous solutions.
77.3 1.49 29.3 0.97
33.6 1.02
38.4 1.07
42.2 1.11 The additives that do not contribute favorably to CO2 solubility
46.7 1.15 compared to pure K2CO3 solvents at 100 °C can be identified from
52.3 1.19
Fig. 6. Glycine addition to the solution caused a shift of the solubility
55.9 1.23
61.7 1.24 curve to the left, indicating a decrease in solubility compared to the
66.3 1.30 1.81 M K2CO3 case. In addition, the slope of the solubility curve was
found to be steeper compared to the reference K2CO3 solvent and the
mixture only achieved a solubility of 0.88 mol CO2/L solvent at a
maximum CO2 equilibrium partial pressure of 89 kPa. However, CO2
aqueous solution and it was concluded that the chemical reaction solubility was comparable for pure 1.81 M K2CO3 and Gly-promoted
dominates CO2 absorption into the liquid phase. solutions at partial pressures below 10 kPa. A similar amino acid (ar-
For the same K2CO3 concentration, an increase in temperature ne- ginine) was studied by Shen et al. who found that, while its effect on
gatively influenced CO2 solubility in all cases. In Fig. 5a, increasing capacity was positive at partial pressures below 1 kPa and low solution
temperature from 80 °C to 100 °C for 1.81 M K2CO3 and CO2 partial loadings, it greatly diminished when loading increased to a > 0.3 (Shen
pressure equal to 10 kPa caused a decrease in CO2 loading by about et al., 2013). The behavior was attributed to bicarbonate formation
24%, while the reduction between 80 °C and 120 °C was 33 %, re- from carbonate and carbamate formation from active promoters in the
spectively. For the higher K2CO3 concentration, Fig. 5b shows a similar solution. For potassium glycinate used as a promoter, Lee et al. mea-
trend with a slightly weaker effect of temperature. An analogous be- sured a CO2 partial pressure decrease when glycine concentration in-
havior was also reported by Tosh et al. (Tosh et al., 1959) for potassium creased and proposed that this can be attributed to the fact that the
carbonate concentrations of 1.7 M and 2.8 M. In their study, a tem- additive acts as an additional mode of CO2 capture in the solution. It
perature increase from 70 °C to 110 °C resulted in a loading reduction of must be noted however that a different range of solution loadings was
about 34% and 29%, respectively. This effect can be attributed both to evaluated and that glycine was deprotonated by mixing with potassium
the exothermic overall reaction in Eq. (1), as well as to the effect of hydroxide (Lee et al., 2015). MDEA exhibited a different behavior as an
higher temperatures on CO2 physical solubility. Fig. 5 also shows that at additive based on its concentration: a) for the case of 1.81 M K2CO3/
80 °C and for a constant partial pressure, the resulting loading reduction 0.25 M MDEA, the solubility curve was found to be placed between
between 1.81 M and 2.53 M K2CO3 solutions was about 14%. This de- those of 1.81 M K2CO3 and 2.53 M K2CO3, thus slightly improving ca-
crease was even lower at higher temperatures. pacity compared to the 1.81 M K2CO3 reference case, while b) the so-
lubility of CO2 decreased when the MDEA concentration increased to
4.3. CO2 solubility in promoted K2CO3 aqueous solutions 0.84 M. This can be attributed to the different chemical reaction me-
chanism in the system (Borhani et al., 2015a), in comparison with the
The results for the promoted solutions are similarly presented in a one governing the interaction of primary amines with CO2 which is
logarithmic scale for CO2 partial pressure in the vertical axis. described through carbamate formation.
Exponential curves are fitted to all VLE data series to better visualize On the other hand, MEA and PZ promoters were both found to
their slopes. From Figs. 6 and 7, it can be seen that CCO2 increases as significantly contribute to the CO2 solubility increase at 100 °C com-
CO2 partial pressure increases. It is important to note that additives did pared to 1.81 M K2CO3. The solubility curves obtained for 0.82 M MEA,
not influence CO2 solubility in the same way, but differences were 0.58 M PZ and 1.16 M PZ in 1.81 M K2CO3 were found to be shifted on
found based on additive type, concentration and temperature. the right of the curve of 1.81 M K2CO3 (without promoter addition).

89
D. Grimekis et al. International Journal of Greenhouse Gas Control 81 (2019) 83–92

Fig. 6. Effect of additives on CO2 solubility in 1.81 M K2CO3 solutions at 100 °C.

Moreover, CO2 solubility in these mixtures was found to be also higher are to be compared between pure K2CO3 solutions and promoted mix-
than that of the 2.53 M K2CO3 solution. Among these mixtures, 1.81 M tures, the addition of MEA seems to have a detrimental effect on so-
K2CO3/1.16 M PZ yielded the highest solubility results, followed by lution loading compared to the reference pure 1.81 M K2CO3 solution,
1.81 M K2CO3/0.58 M PZ and finally 1.81 M K2CO3/0.82 M MEA. When already at CO2 partial pressures above 10 kPa at 100 °C. The equili-
comparing the solutions 1.81 M K2CO3 and 1.81 M K2CO3/1.16 M PZ, a brium curve at 120 °C is positioned above that of the pure K2CO3
CCO2 value equal to 0.85 mol CO2/L was obtained for CO2 partial aqueous solutions at the partial pressure range investigated. A similar
pressures of 28 kPa and 5 kPa respectively, corresponding to more than effect was described by Ramazani et al. (Ramazani et al., 2016), for
an 80% relative decrease in equilibrium partial pressure. Concerning molar ratios of MEA promoter to K2CO3 between 0.25 and 1.5 at 40 °C.
the effect of PZ as an additive, Cullinane et al. (Cullinane and Rochelle,
2005) reported that increasing PZ concentration in a solution of con- 4.4. Effect of additives on CO2 absorption rate
stant K+ concentration can benefit CO2 solubility and this is reflected
through a less steep solubility curve. The increased solubility can be In order to compare the absorption rates for the promoted solutions,
attributed to the conversion of active PZ to PZCOO− at low/medium results are reported for a similar initial CO2 loading, as well as similar
loadings and PZ(COO−)2 at higher loadings (see also Eqs.((11)–(14))). partial pressure at the beginning of the pressure decay curve. Results for
For the case of MEA, the improved CO2 absorption efficiency can be fresh solvents and for intermediate points of a = 0.3 are discussed, since
attributed to the increased OH− ions, which increase the solution pH as a typical absorption/desorption process operates with only partially
measured by (Thee et al., 2012b), as well as to the zwitterion me- stripping the absorbed acid gas so as to lower the reboiler heat duty
chanism for carbamate formation between MEA and CO2 shown in Eqs. (Moioli and Pellegrini, 2015; Thee et al., 2012b). While for the current
((8)–(10)). experimental method this entails that not all starting loading and par-
At 120 °C, the 1.81 M K2CO3/1.16 M PZ solution was again found to tial pressure values are identical for the intermediate loading points, the
contribute to a higher CO2 solubility compared to the rest of the mix- comparison can provide a good insight on the suitability of different
tures. Glycine and MDEA were excluded from investigations at 120 °C. promoters for enhancing the overall reaction rate. For a more accurate
From Fig. 7, it can be concluded that 1.81 M K2CO3 solutions with quantification of reaction kinetics, experiments with a device which
0.58 M PZ and 0.82 M MEA exhibit similar results to those measured for ensures a measurable surface area for mass transfer are required to
the 2.53 M K2CO3 aqueous solution. An outcome of this comparison at determine the CO2 flux (Thee et al., 2012a).
120 °C is the fact that MEA performs slightly better in terms of capacity For a starting a value of about 0.3 at 100 °C, a time duration be-
compared to the 1.81 M K2CO3/0.58 M PZ mixture at higher partial tween 18–20 min was needed to reach equilibrium for the 1.81 M
pressures, which was not the case at 100 °C. When solution loadings (a ) K2CO3 solution, while a shorter time interval was required to attain
equilibrium state for the first CO2 injection (∼14 min), which is an-
ticipated due to faster reaction kinetics for the fresh solvent compared
to the loaded one. This is caused by a higher availability of hydroxyl
ions to react with CO2, as shown in the kinetic reaction of Eq. (6). A
similar observation on the effect of increasing solution loading on ab-
sorption rate was made for the promoted solutions independently of
their type and concentration. The high absorption rate for 1.81 M
K2CO3 at both temperatures is attributed to a higher CO2 partial pres-
sure at the beginning of the pressure decay process. Addition of PZ and
MEA was found to greatly improve the absorption rate compared to the
other promoters used. As shown in Figs. 8 and 9, the solution promoted
with 1.16 M PZ performed better and it was followed by the mixtures
1.81 M K2CO3/0.58 M PZ and 1.81 M K2CO3/0.82 M MEA. At 100 °C,
the average absorption rate over the entire time interval towards the
equilibrium point for a starting loading equal to 0.3 mol CO2/mol blend
was 0.0047 mol/min for the mixture 1.81 M K2CO3/1.16 M PZ. This
Fig. 7. Effect of additives on CO2 solubility in 1.81 M K2CO3 solutions at 120 °C. value was about 11% higher compared to the mixture promoted with

90
D. Grimekis et al. International Journal of Greenhouse Gas Control 81 (2019) 83–92

promoted with MDEA. Pure glycine seems to have no remarkable im-

pact on CO2 absorption at high solution loadings and a slight im-
provement for the fresh solvents can be pointed out based on Fig. 8.
According to Thee et al., the pretreatment of glycine towards a de-
protonated amino acid can establish a reaction mechanism similar to
the one described for alkanolamines with the formation of a carbamate
intermediate (Thee et al., 2014). The K2CO3 solution with the depro-
tonated amino acids was found to exhibit improved reaction kinetics
compared to primary amines and it is therefore suggested for future
investigation. A strong base must be used in this case, such as KOH, and
the reaction for deprotonation of the zwitterions is represented by Eq.
(27).
+
NH2 R1R2COO + OH NHR1R2COO + H2 O (27)
The absorption rate curves for PZ and MEA at 120 °C were compared
Fig. 8. Effect of additives on CO2 absorption rate into fresh solvents to the one obtained for K2CO3 without additives. As shown in Fig. 9 for
(PCO2, gs = 180 kPa for promoted solutions, PCO2, gs = 200 kPa for the pure K2CO3 the fresh solvents, the time to equilibrium after the first injection of CO2
aqueous solution, T = 100 °C).
was in a range of 6–7 min for the promoted solutions and 11–12 min for
the pure K2CO3 aqueous solution. Piperazine promoted solutions were
found to enhance the absorption rate to a greater extent at this higher
temperature, similarly to the observations made at 100 °C. However, at
this higher temperature level, similar absorption rates for MEA and PZ
were observed and thus both promoters may be used for high-tem-
perature CO2 absorption. According to (Thee et al., 2012b), the ap-
parent reaction rate constant of CO2 in a MEA-promoted K2CO3 solution
can be increased by increasing both MEA concentration and solution
temperature.

5. Conclusions

In this study, an experimental investigation of CO2 solubility and

absorption rate into pure and promoted K2CO3 solvents was conducted.
For K2CO3 aqueous solutions without promoters, tests were carried out
at 80 °C, 100 °C and 120 °C with two different K2CO3 concentrations and
experimental results were successfully validated against simulation
Fig. 9. Effect of additives on CO2 absorption rate into fresh solvents results obtained with the electrolyte-NRTL model fitted with default
(PCO2, gs = 80 kPa for promoted solutions, PCO2, gs = 90 kPa for the pure K2CO3 electrolyte pair parameters in Aspen Plus®, as well as regressed para-
aqueous solution, T = 120 °C). meters based on experimental data from previous studies in order to
improve the accuracy of the activity coefficient model. Glycine, MDEA,
MEA and PZ were tested at 100 °C as additives, while the latter two
promoters were also tested at 120 °C. As far as K2CO3 solvents are
0.82 M MEA. For the fresh solvents at 100 °C, those containing PZ (both concerned, solubility increases with a decrease in temperature or with a
concentrations), MEA, Gly and MDEA (at its higher concentration) denser solvent. Concerning the addition of promoters, the use of PZ
equilibrated after 7, 10, 15 and 20 min respectively, while at a = 0.3, greatly improves CO2 solubility and absorption rate at the same time.
the promotion effect of a higher PZ concentration was more prominent The use of MEA is also beneficial for both solubility and absorption rate,
compared to the lower concentration, mainly due to a higher amount of but to a lesser extent compared to PZ. On the other hand, MDEA can
free amine molecules which can react with CO2 molecules. At this potentially improve solubility results when added in lower quantities,
loading level, MEA followed PZ promoted solutions with equilibration however, when used at higher concentrations, it lowers CO2 solubility.
times in the order of 14 min for the former and 10 min for the latter. Taking into account that also no remarkable effect on CO2 absorption
The structure of PZ molecules consisting of two amino groups and a rate was measured, MDEA is not recommended as a promoter for
high pKa value has been proposed to be responsible for its performance K2CO3-H2O solutions. Finally, pure glycine addition, even at a low
in terms of absorption rate (Cullinane and Rochelle, 2005; Ramazani molarity, deteriorated CO2 solubility.
et al., 2016). For the mixture containing MDEA as an additive, no
particular enhancement effect on the absorption rate was noted. Taking Acknowledgments
into account the solubility results for the MDEA/K2CO3 mixture of this
study, MDEA is not recommended as a promoter for CO2 absorption at The authors greatly acknowledge the financial support provided by
this temperature range, which is in agreement with the conclusions of the European Commission’s Research Fund for Coal and Steel (RFCS)
Borhani et al. (Borhani et al., 2015a) in the framework of a simulation under Grant Agreement No. RFCR-CT-2013-00008/“CO2freeSNG2.0”.
analysis for MDEA promoted potassium carbonate. Moreover, ac-
cording to Nii and Takeuchi (Nii and Takeuchi, 1994), MDEA can only References
slightly improve the CO2 flux compared to other amines in aqueous
solutions of potassium carbonate and the enhancement factor was Aronu, U.E., Svendsen, H.F., Hoff, K.A., 2010. Investigation of amine amino acid salts for
found to be much smaller compared to other amines at ambient tem- carbon dioxide absorption. Int. J. Greenhouse Gas Control 4, 771–775. https://doi.
org/10.1016/j.ijggc.2010.04.003.
perature conditions. Astarita, G., Savage, D.W., Longo, J.M., 1981. Promotion of CO2 mass transfer in car-
The absorption rate for Gly-promoted solutions was found to be bonate solutions. Chem. Eng. Sci. 36, 581–588. https://doi.org/10.1016/0009-
comparable to the reference case solution and similar to the solution 2509(81)80146-7.

91
D. Grimekis et al. International Journal of Greenhouse Gas Control 81 (2019) 83–92

Atsonios, K., Christodoulou, C., Koytsoumpa, E.-I., Panopoulos, K., Kakaras, E., 2013. Lee, A., Mumford, K.A., Wu, Y., Nicholas, N., Stevens, G.W., 2016. Understanding the
Plant design aspects of catalytic biosyngas conversion to higher alcohols. Biomass vapour–liquid equilibrium of CO2 in mixed solutions of potassium carbonate and
Bioenergy 53, 54–64. https://doi.org/10.1016/j.biombioe.2013.01.035. potassium glycinate. Int. J. Greenhouse Gas Control 47, 303–309. https://doi.org/10.
Benson, H., 1956. Improved process for CO2 absorption uses Hot carbonate solutions. 1016/j.ijggc.2016.02.005.
Chem. Eng. Progr. 52, 433–438. Li, J., Lin, X., Ning, P.-G., Cao, H.-B., Zhang, Y., 2014. Measurement and modelling of the
Bohloul, M.R., Peyghambarzadeh, S.M., Lee, A., Vatani, A., 2014. Experimental and solubility of carbon dioxide in aqueous 1, 8-p-menthane-diamine solution. J.
analytical study of solubility of carbon dioxide in aqueous solutions of potassium Chem.Thermodyn. 71, 64–70. https://doi.org/10.1016/j.jct.2013.11.018.
carbonate. Int. J. Greenhouse Gas Control 29, 169–175. https://doi.org/10.1016/j. Li, H., Le Moullec, Y., Lu, J., Chen, J., Marcos, J.C.V., Chen, G., Chopin, F., 2015. CO2
ijggc.2014.08.009. solubility measurement and thermodynamic modeling for 1-methylpiperazine/
Borhani, T.N.G., Akbari, V., Hamid, M.K.A., Manan, Z.A., 2015a. Rate-based simulation water/CO2. Fluid Phase Equilib. 394, 118–128. https://doi.org/10.1016/j.fluid.
and comparison of various promoters for CO2 capture in industrial DEA-promoted 2015.03.021.
potassium carbonate absorption unit. J. Ind. Eng. Chem. 22, 306–316. https://doi. Mazinani, S., Ramazani, R., Samsami, A., Jahanmiri, A., Van der Bruggen, B.,
org/10.1016/j.jiec.2014.07.024. Darvishmanesh, S., 2015. Equilibrium solubility, density, viscosity and corrosion rate
Borhani, T.N.G., Azarpour, A., Akbari, V., Alwi, S.R.W., Manan, Z.A., 2015b. CO2 capture of carbon dioxide in potassium lysinate solution. Fluid Phase Equilib. 396, 28–34.
with potassium carbonate solutions: A state-of-the-art review. Int. J. Greenhouse Gas https://doi.org/10.1016/j.fluid.2015.03.031.
Control 41, 142–162. https://doi.org/10.1016/j.ijggc.2015.06.026. Moffat, R.J., 1988. Describing the uncertainties in experimental results. Exp. Therm Fluid
Chen, C.C., Evans, L.B., 1986. A local composition model for the excess Gibbs energy of Sci. 1, 3–17. https://doi.org/10.1016/0894-1777(88)90043-X.
aqueous electrolyte systems. AlChE J. 32, 444–454. https://doi.org/10.1002/aic. Moioli, S., Pellegrini, L.A., 2015. Improved rate-based modeling of the process of CO2
690320311. capture with PZ solution. Chem. Eng. Res. Des. 93, 611–620.
Chen, C.C., Song, Y., 2004. Generalized electrolyte‐NRTL model for mixed‐solvent elec- Neubert, M., Treiber, P., Krier, C., Hackel, M., Hellriegel, T., Dillig, M., Karl, J., 2017.
trolyte systems. AlChE J. 50, 1928–1941. https://doi.org/10.1002/aic.10151. Influence of hydrocarbons and thiophene on catalytic fixed bed methanation. Fuel
Cullinane, J.T., 2005. Thermodynamics and Kinetics of Aqueous Piperazine with 207, 253–261. https://doi.org/10.1016/j.fuel.2017.06.067.
Potassium Carbonate for Carbon Dioxide Absorption. PhD Thesis. The University of Nii, S., Takeuchi, H., 1994. Removal of CO2 and/or SO2 from gas streams by a membrane
Texas at Austin, Texas, USA. absorption method. Gas Sep. Purif. 8, 107–114. https://doi.org/10.1016/0950-
Cullinane, J.T., Rochelle, G.T., 2004. Carbon dioxide absorption with aqueous potassium 4214(94)80017-0.
carbonate promoted by piperazine. Chem. Eng. Sci. 59, 3619–3630. https://doi.org/ Pacheco, M.A., Rochelle, G.T., 1998. Rate-based modeling of reactive absorption of CO2
10.1016/j.ces.2004.03.029. and H2S into aqueous methyldiethanolamine. Ind. Eng. Chem. Res. 37, 4107–4117.
Cullinane, J.T., Rochelle, G.T., 2005. Thermodynamics of aqueous potassium carbonate, https://doi.org/10.1021/ie980123g.
piperazine, and carbon dioxide. Fluid Phase Equilib. 227, 197–213. https://doi.org/ Park, S.-B., Shim, C.-S., Lee, H., Lee, K.-H., 1997. Solubilities of carbon dioxide in the
10.1016/j.fluid.2004.11.011. aqueous potassium carbonate and potassium carbonate/poly (ethylene glycol) solu-
Davis, J., 2009. Thermal Degradation of Aqueous Amines Used for Carbon Dioxide tions. Fluid Phase Equilib. 134, 141–149. https://doi.org/10.1016/S0378-3812(97)
Capture. The University of Texas at Austin, Texas, USA PhD Thesis. 00052-6.
Dawodu, O.F., Meisen, A., 1996. Degradation of alkanolamine blends by carbon dioxide. Phan, D.T., Maeder, M., Burns, R.C., Puxty, G., 2014. Catalysis of CO2 absorption in
Can. J. Chem. Eng. 74, 960–966. https://doi.org/10.1002/cjce.5450740620. aqueous solution by inorganic oxoanions and their application to post combustion
Dennard, A., Williams, R., 1966. The catalysis of the reaction between carbon dioxide and capture. Environ. Sci.Technol. 48, 4623–4629. https://doi.org/10.1021/es500667s.
water. J. Chem. Soc. A 812–816. https://doi.org/10.1039/J19660000812. Ramazani, R., Mazinani, S., Jahanmiri, A., Van der Bruggen, B., 2016. Experimental in-
Edinger, P., Grimekis, D., Panopoulos, K., Karellas, S., Ludwig, C., 2017. Adsorption of vestigation of the effect of addition of different activators to aqueous solution of
thiophene by activated carbon: A global sensitivity analysis. J. Environ. Chem. Eng. potassium carbonate: absorption rate and solubility. Int. J. Greenhouse Gas Control
5, 4173–4184. https://doi.org/10.1016/j.jece.2017.07.041. 45, 27–33. https://doi.org/10.1016/j.ijggc.2015.12.003.
Freeman, S., 2011. Thermal Degradation and Oxidation of Aqueous Piperazine for Carbon Rochelle, G.T., 2012. Thermal degradation of amines for CO2 capture. Current Opinion
Dioxide Capture. The University of Texas at Austin, Texas, USA PhD Thesis. Chem. Eng. 1, 183–190. https://doi.org/10.1016/j.coche.2012.02.004.
Freeman, S.A., Dugas, R., Van Wagener, D.H., Nguyen, T., Rochelle, G.T., 2010. Carbon Shen, S., Feng, X., Ren, S., 2013. Effect of arginine on carbon dioxide capture by po-
dioxide capture with concentrated, aqueous piperazine. Int. J. Greenhouse Gas tassium carbonate solution. Energy Fuels 27, 6010–6016. https://doi.org/10.1021/
Control 4, 119–124. https://doi.org/10.1016/j.ijggc.2009.10.008. ef4014289.
Hilliard, M., 2008. A Predictive Thermodynamic Model for an Aqueous Blend of Singh, P., Brilman, D.W., Groeneveld, M.J., 2011. Evaluation of CO2 solubility in po-
Potassium Carbonate, Piperazine, and Monoethanolamine for Carbon Dioxide tential aqueous amine-based solvents at low CO2 partial pressure. Int. J. Greenhouse
Capture from Flue Gas. Chemical Engineering department. The University of Texas at Gas Control 5, 61–68. https://doi.org/10.1016/j.ijggc.2010.06.009.
Austin, Texas, USA PhD Thesis. Smith, K., Lee, A., Mumford, K., Li, S., Thanumurthy, N., Temple, N., Anderson, C.,
Hu, G., Nicholas, N.J., Smith, K.H., Mumford, K.A., Kentish, S.E., Stevens, G.W., 2016. Hooper, B., Kentish, S., Stevens, G., 2015. Pilot plant results for a precipitating po-
Carbon dioxide absorption into promoted potassium carbonate solutions: A review. tassium carbonate solvent absorption process promoted with glycine for enhanced
Int. J. Greenhouse Gas Control 53, 28–40. https://doi.org/10.1016/j.ijggc.2016.07. CO2 capture. Fuel Process. Technol. 135, 60–65. https://doi.org/10.1016/j.fuproc.
020. 2014.10.013.
Huang, Q., Bhatnagar, S., Remias, J.E., Selegue, J.P., Liu, K., 2013. Thermal degradation Sreedhar, I., Nahar, T., Venugopal, A., Srinivas, B., 2017. Carbon capture by absorp-
of amino acid salts in CO2 capture. Int. J. Greenhouse Gas Control 19, 243–250. tion–path covered and ahead. Renewable Sustain. Energy Rev. 76, 1080–1107.
https://doi.org/10.1016/j.ijggc.2013.09.003. https://doi.org/10.1016/j.rser.2017.03.109.
Kim, Y.E., Choi, J.H., Nam, S.C., Yoon, Y.I., 2012. CO2 absorption capacity using aqueous Thee, H., Smith, K.H., da Silva, G., Kentish, S.E., Stevens, G.W., 2012a. Carbon dioxide
potassium carbonate with 2-methylpiperazine and piperazine. J. Ind. Eng. Chem. 18, absorption into unpromoted and borate-catalyzed potassium carbonate solutions.
105–110. https://doi.org/10.1016/j.jiec.2011.11.078. Chem. Eng. J. 181, 694–701. https://doi.org/10.1016/j.cej.2011.12.059.
Knuutila, H., Juliussen, O., Svendsen, H.F., 2010. Density and N2O solubility of sodium Thee, H., Suryaputradinata, Y.A., Mumford, K.A., Smith, K.H., da Silva, G., Kentish, S.E.,
and potassium carbonate solutions in the temperature range 25 to 80 °C. Chem. Eng. Stevens, G.W., 2012b. A kinetic and process modeling study of CO2 capture with
Sci. 65, 2177–2182. https://doi.org/10.1016/j.ces.2009.12.015. MEA-promoted potassium carbonate solutions. Chem. Eng. J. 210, 271–279. https://
Kohl, A.L., Nielsen, R., 1997. Gas Purification. Gulf Professional Publishing. doi.org/10.1016/j.cej.2012.08.092.
Koronaki, I., Prentza, L., Papaefthimiou, V., 2015. Modeling of CO2 capture via chemical Thee, H., Nicholas, N.J., Smith, K.H., da Silva, G., Kentish, S.E., Stevens, G.W., 2014. A
absorption processes-An extensive literature review. Renewable Sustainable Energy kinetic study of CO2 capture with potassium carbonate solutions promoted with
Rev. 50, 547–566. https://doi.org/10.1016/j.rser.2015.04.124. various amino acids: glycine, sarcosine and proline. Int. J. Greenhouse Gas Control
Kothadaraman, A., 2010. Carbon Dioxide Capture by Chemical Absorption: A Solvent 20, 212–222. https://doi.org/10.1016/j.ijggc.2013.10.027.
Comparison Study. PhD Thesis. Massachusetts Institute of Technology. Tosh, J.S., Field, J., Benson, H., Haynes, W., 1959. Equilibrium Study of the System
Koytsoumpa, E.-I., Atsonios, K., Panopoulos, K., Karellas, S., Kakaras, E., Karl, J., 2015. Potassium Carbonate, Potassium Bicarbonate, Carbon Dioxide, and Water. Bureau of
Modelling and assessment of acid gas removal processes in coal-derived SNG pro- Mines, Pittsburgh, Pa.(USA).
duction. Appl. Therm. Eng. 74, 128–135. https://doi.org/10.1016/j.applthermaleng. Wang, M., Lawal, A., Stephenson, P., Sidders, J., Ramshaw, C., 2011. Post-combustion
2014.02.026. CO2 capture with chemical absorption: A state-of-the-art review. Chem. Eng. Res.
Lee, A., Wolf, M., Kromer, N., Mumford, K.A., Nicholas, N.J., Kentish, S.E., Stevens, G.W., Des. 89, 1609–1624. https://doi.org/10.1016/j.cherd.2010.11.005.
2015. A study of the vapour–liquid equilibrium of CO2 in mixed solutions of po- Ye, X., Lu, Y., 2014. Kinetics of CO2 absorption into uncatalyzed potassium carbonate–-
tassium carbonate and potassium glycinate. Int. J. Greenhouse Gas Control 36, bicarbonate solutions: effects of CO2 loading and ionic strength in the solutions.
27–33. https://doi.org/10.1016/j.ijggc.2015.02.007. Chem. Eng. Sci. 116, 657–667. https://doi.org/10.1016/j.ces.2014.05.050.

92