Review on the mass transfer performance of CO₂ absorption by amine-based solvents in low- and high-pressure absorption packed columns

Abstract

The gas-phase volumetric overall mass transfer coefficient (K_Ga_V) plays a key role in the assessment of an absorption packed column's performance since it determines the height of an absorber column. The effective and useable data provided by K_Ga_V is necessary for designing and scaling up absorption packed columns. This study provides the first comprehensive review of mass transfer performance in terms of K_Ga_V for CO₂ (K_GCO2a_V) absorption into amine solutions for absorber columns with random and structured packing. To date, researchers associated with the K_GCO2a_V parameter have focused on two main fields: experimental works and developing empirical correlations. For experimental works, K_GCO2a_V has been evaluated in the literature for a large number of conventional and improved amines over a range of operating parameters in laboratory-scale packed columns. In addition, researchers have developed empirical correlations for K_GCO2a_V based on operating parameters affecting K_GCO2a_V and physical properties. The details of research determining the K_GCO2a_V have been reviewed for low- and high-pressure absorption packed columns. Finally, directions for future research of the mass transfer performance for absorber packed columns in the CO₂ capture process have been discussed.

---

Morteza Afkhamipour

Morteza Afkhamipour received his MSc Degree (2011) in Chemical Engineering from PGU. He is presently a PhD student in PGU and his research interests are CO2 capture, gas sweetening, gas absorption, experimental design, artificial neural networks, amine solution thermodynamics, mass transfer and modeling and simulation. He has published eight papers in international journals.

Masoud Mofarahi

Masoud Mofarahi is a Full Professor in the area of Adsorption technology and CO2 emission Capture. He graduated in Chemical Engineering at Shiraz University, obtained his PhD in Chemical Engineering at Tarbiat Modares University in 2003. Currently, he is a member of the scientific board of Persian Gulf University. His main fields of research are Gas Adsorption, Pressure Swing Adsorption, CO2 Capture, Nano-fluids and Thermodynamics of Amine Solutions. He has participated in more than 10 national and international projects in the areas of Environment and Energy.

---

1. Introduction

Today, fossil fuels play a major role in the production and supply of energy in the world. With global economic development and population growth, the utilization of these fuels has increased extensively. As a result, the consumption of fossil fuels is causing a sharp increase of CO₂ emissions into the atmosphere.^1,2 CO₂ is the main greenhouse gas responsible for global warming and climate change. Although CO₂ is corrosive to exposed equipment and has a low heating value, it can be used in various industries after treatment.³ Research has become essential in achieving an effective process of CO₂ removal from industrial exhaust gas streams.⁴ CO₂ is usually produced in different concentrations by three main processes including pre-combustion, post-combustion, and oxyfuel combustion.^5,6 Currently, the only CO₂ separation process implemented at a fully commercial scale is the post-combustion CO₂ capture process. The foremost instances are post-combustion CO₂ capture in the TMC Mongstad in Norway, which can capture 300 000 tons of CO₂ per year and BD3 Sask Power in Canada, which can capture 1 million tons of CO₂ per year.⁶ Depending on the conditions of the post-combustion capture process, different CO₂ separation technologies can be used such as absorption, adsorption, and membrane and cryogenic separation.^7,8 One of the most widespread and cost-effective processes in the capture of impurities from various gas streams is the absorption method using chemical solvents. Among chemical absorption processes, the amine processes are among the most important. They have been applied in numerous refineries worldwide for eliminating CO₂ and hydrogen sulfide from ammonia-synthesis gas, natural gas, crude hydrogen, fossil-fuel fired power plants, petroleum industry, and town gas streams.⁹ Over the past decades, different separator devices for CO₂ capture processes, including columns with tray, spray, packing types, and membrane contactors have been extensively studied and have received significant attention.^10–13 Among these separators, packed columns (random or structured) are well known to have good mass transfer performance features. To accurately estimate the exact size of packed column height necessary for the mass transfer of CO₂ from the gas phase to the liquid phase in case of absorption with a chemical reaction, the accurate prediction of the mass transfer coefficients is very important.¹¹ To raise the designer's confidence and provide the best data for scaling up and designing packed columns in CO₂ capture plants, a good understanding of the basics of modeling and designing is fundamental.¹¹ Currently, CO₂ absorption into amine solutions in packed columns is considered to be one of the most complicated systems because of relationships between the mass/heat transfer, hydromechanics and thermodynamics.¹⁴ To calculate the mass transfer rate in an absorber column, whether random or structured, empirical or semi-empirical correlations have been developed by researchers in terms of mass transfer coefficients in gas and liquid phases and interfacial areas. However, these developed correlations can be inaccurate in some case and sometimes they cannot be applied to specific systems such as CO₂–amine systems. A review on several mass transfer correlations in packed columns has been performed in the literature.^15,16 Razi, et al.¹⁷ have assessed these correlations in a rate-based model for CO₂ capture with a monoethanolamine (MEA) solution to show the uncertainty associated with using correlations in large-scale CO₂ capture plants. They concluded that the interactions among the transport parameters, mass transfer coefficients, effective interfacial areas and kinetics are too complex for the mass transfer correlations to be applied with a sufficient level of confidence.¹⁷ So far, none of these reviews has provided a review of mass transfer performance in terms of the K_GCO2a_V parameter for CO₂ absorption into amine solutions in packed columns. In existing studies related to the K_GCO2a_V for CO₂ absorption at a laboratory scale, packed columns have been experimentally investigated for a large number of amine-based solvents across a range of operating parameters. The base of researchers' studies was the prediction of K_GCO2a_V, the investigation of effects of operating parameters on it, and the correlation between K_GCO2a_V and operating process parameters. This study presents a review of studies of K_GCO2a_V parameter at low- and high-pressure packed columns for CO₂ absorption into amine solutions in packed columns. The advantage of directly using K_GCO2a_V when calculating the mass transfer rate is that it makes it unnecessary to calculate the individual mass transfer coefficients in gas and liquid phases and other parameters such as the enhancement factor and Henry's law solubility constant. In addition, the K_GCO2a_V can be applied to specific systems; for example, Ziaii, et al.¹⁸ have used such developed empirical correlation in a CO₂ capture simulation with MEA solution for an absorber packed column.

2. Overview of amine solvents used in the analysis of K_GCO2a_V

2.1. Conventional amines

Notable development in CO₂ capture processes using reactive solvents has been seen in the past decade due to their ability to offer near-full absorption and desorption of CO₂. Among different reactive solvents such as amines, potassium carbonate, and ammonia, that have been studied for CO₂ capture processes, amines are considerably well-developed.¹⁹

MEA is the most well-known amine, with a high reaction rate with CO₂ and a low cost.²⁰ These advantages of MEA can decrease the height of an absorber column and facilitate large-scale operations.^14,21 However, MEA has some disadvantages such as high-energy consumption for its regeneration, low absorption capacity, and degradation and corrosion problems.^7,22 To overcome these disadvantages, a number of important amine solvents have been commercially utilized such as diethanolamine (DEA), N-methyldiethanolamine (MDEA), 2-amino-2-methyl-1-propanol (AMP), and piperazine (PZ).²³

As can be seen from Table 1, the aforementioned amines each have their advantages and disadvantages. Clearly, there is no particular amine solvent with all ideal characteristics for CO₂ capture processes. Studies have so far focused on the improvement of amine solvents in order to reach a high CO₂ capture performance and a low cost of operation.²⁴ In recent years, mixing of conventional amines has shown a considerable improvement of absorption and desorption in CO₂ capture processes.²⁵ The higher absorption capacity, faster kinetics and lower energy consumption for stripping of CO₂ are good characteristics of mixed amines. Research of mixed amines, such as MEA–MDEA and AMP–PZ, has demonstrated a great enhancement of the kinetics, thermodynamics, mass transfer, as well as energy consumption for regeneration.^26–28

Table 1 Properties of conventional and newly developed amines⁴⁵

Chemical name	MEA	DEA	MDEA	AMP	PZ, anhydrous	PZ, 65%	1DMA2P	DETA	DEEA
Molecular formula	C2H7NO	C4H11NO2	C5H13NO2	C4H11NO	C4H10N2	C4H10N2	C5H13NO	C4H13N3	C6H15NO
Formula weight	61.08	105.14	119.16	89.14	86.13	86.13	103.16	103.17	117.19
Freezing point °C	10.3	28	−21.0	26.0	110	41	−85.0	−39.0	−70
Boling point °C	170	268	247.1	165	146	116	96.0	207	161
Density at 25 °C	1.01	1.092	1.03	0.934	0.877	1.03	0.913	0.952	0.884
Vapor pressure (at 20 °C) mm Hg	0.48	<0.01	0.01	<1	0.1	6.28	8	0.08	1
Water solubility	Miscible	Miscible	Miscible	Miscible	14 wt%	Miscible	Miscible	Miscible	Miscible
Absorption rate	High	Medium	Low	Medium	High	High	Low	Medium	Low
Absorption capacity	Medium	Medium	High	High	High	High	High	High	Medium
Heat of absorption	High	High	Medium	High	High	High	Medium	High	Medium

---

2.2. Newly developed amines

Recently, the focus of studies has turned to new and promising amines, such as N,N-diethylethanolamine (DEEA), 4-diethylamino-2-butanol (DEAB), diethylenetriamine (DETA), and 1-dimethylamino-2-propanol (1DMA2P).^23,29–33 DEEA is a tertiary-type amine, which is a potential candidate for CO₂ bulk removal.³⁴ This amine can be made from low-priced resources such as agricultural products and residues.³³ Kim and Savage,³⁵ Benitez-Garcia, et al.³⁶ and Li, et al.³⁷ have reported data of CO₂ absorption rates by DEEA at different temperatures and concentrations. Li, et al.³⁷ have concluded that the rate of absorption of CO₂ in a DEEA solution is higher than in an MDEA solution. Also, Vaidya and Kenig³⁸ have shown that the absorption capacity of a DEEA solution, in terms of CO₂ loading, approached a value of 1 mol CO₂/mol DEEA. DEAB is an amino alcohol solvent which was synthesized based on a normal molecular design approach.³⁹ Sema, et al.⁴⁰ have studied and compared the CO₂ absorption capacity of DEAB with conventional amines such as MEA, DEA, AMP, and MDEA, and their results showed that the absorption of CO₂ in a DEAB solution requires a lower DEAB concentration for the same CO₂ removal efficiency as conventional amines. They also concluded that the rate of absorption of CO₂ in a DEAB solution is higher than in an MDEA solution and lower than in an MEA solution. DETA, a polyamine comprising two primary and one secondary amine groups, has shown to have a faster reaction rate and a higher absorption capacity compared with conventional amines.^41,42 Recently, Chowdhury, et al.⁴³ and Liang, et al.⁴⁴ have shown that a new tertiary amine, 1-dimethylamino-2-propanol (1DMA2P), has good potential for CO₂ absorption because of its superior performance. Kadiwala, et al.³⁰ have shown that 1DMA2P has a faster reaction rate than MDEA but slower than MEA. Chowdhury, et al.⁴³ have reported that the CO₂ loading of 1DMA2P (at low CO₂ partial pressures) is about twice as high as that of an MDEA solution. Table 1 shows a summary of conventional and newly developed amines' properties⁴⁵ used in analyses of K_GCO2a_V in packed columns.

3. Mass transfer theory

At a particular point of an absorber column, mass transfer occurs because of a chemical potential gradient between gas and liquid phases. The mass transfer ends when equilibrium is reached. In other words, when the net mass transfer becomes zero.⁴⁶ Nevertheless, the question is at what rate can the mass be transferred? This problem can be associated with the mass transfer coefficient.⁴⁷ The mass transfer coefficient is an important parameter in designing absorber columns.^11,14 Knowledge of this parameter can help a designer accurately calculate the height of an absorber column. In an absorber packed column in a post-combustion CO₂ capture plant, the removal efficiency of CO₂ absorption by amine solutions can be determined by the gas–liquid contact degree, physicochemical properties and hydrodynamics of the absorber column, amine reactivity degree and operating parameters related to the gas and amine solution.¹⁷ Chemical absorption of CO₂ into an amine solution can be described by the two-film theory.⁴⁸ This theory proposes that there are two thin films near the gas and liquid phase interfaces, which separate them from the liquid and gas bulk phases. This theory assumes that bulk phases are in equilibrium and all resistances of mass and heat transfer exist in the two films.⁴⁸ In most cases, when CO₂ moves from the gas to the liquid phase, a chemical reaction between CO₂ and the amine solution can take place in the liquid film or liquid bulk.⁴⁹ According to Fig. 1 and based on the two-film theory, the reaction between CO₂ and the amine solution can be characterized as infinitely fast rate or very slow rate.⁴⁸ Depending on the relative values of the reaction rate constants, mass transfer coefficients of gas and liquid phases, concentration ratio of reactants and CO₂ equilibrium solubility, reactions occur in a narrow zone within the film or through the film and bulk of the liquid. The two-film theory is prevalently used in rate-based models.^49–55 A significant number of them consider that the reaction takes place within the liquid film, when the reaction is assumed to be infinitely fast (i.e., CO₂ absorption into the MEA solution).

Fig. 1 Location of the chemical reaction between CO₂ and an amine solution based on the two-film theory.

3.1. Determination of K_GCO2a_V in an absorption packed column

At steady-state conditions, the absorbed mass flux of CO₂ (N_CO2) across the gas–liquid interface can be represented in terms of K_G and the difference between the CO₂ partial pressure in the gas bulk (Py_CO2) and the CO₂ partial pressure at the gas–liquid interface , as shown in eqn (1).^56,57

It is obvious from eqn (1) that N_CO2 is greatest when approaches zero and Py_CO2 is at a maximum value. In the same way, N_CO2 is zero when Py_CO2 is equal to .The significance of K_G can be seen from eqn (1)—for a given driving force, a greater K_G can give greater N_CO2 into the amine solution.⁵⁸ Since the driving force of mass transfer occurs at a small distance from the film, the concentration in the interface and, subsequently, K_G are difficult to measure in an absorption packed column because of the variations in the interfacial area with varying gas and liquid flow rates.⁵⁹ Therefore, it is more convenient and useful to express N_CO2 based on the unit volume of the absorption packed column rather than the interfacial area unit, as follows:⁵⁹

In eqn (2), N_CO2a_V can be obtained from K_GCO2a_V and the difference between Py_CO2 and the CO₂ partial pressure in the gas phase in equilibrium with the CO₂ concentration in the liquid bulk . To calculate N_CO2a_V, the mass balance according to the rate-based model, considering a small differential height of packing (dZ) of the absorption packed column (Fig. 2), can be written as follow:⁵⁹

Fig. 2 Differential elements of packing for an absorption column.

By substituting eqn (3) into eqn (2), K_GCO2a_V can be determined using eqn (4):

Most researchers have used eqn (4) to determine K_GCO2a_V from experiments on absorption packed columns.^{32,33,58–61} In eqn (4), the gas flow rate (G) and cross-section area of the column (A) as well as the packed column pressure (P) are known, and only two terms—the driving force and the derivative of the CO₂ molar ratio—have to be determined. The term in eqn (4) can be obtained from the equilibrium solubility data of CO₂ into the amine solution. Often, is assumed to be zero due to a fast reaction between CO₂ and the amine solution.^{14,32,49,58,60–65} The derivative of the CO₂ molar ratio can be determined by measuring the CO₂ concentration (molar fraction) profile in the gas phase along the height of the absorber packed column. By converting molar fraction values to molar ratio values of CO₂, the term is calculated by plotting Y_CO2 against the packing height of the absorber column (Z), as shown in Fig. 3.⁵⁹

Fig. 3 Determination of the molar ratio slopes by measuring the CO₂ concentration profile in the gas phase along the height of an absorber column.

When the CO₂ concentration is measured at the inlet and outlet of an absorber packed column, the average values of K_GCO2a_V can be obtained from eqn (5) suggested by Dey and Aroonwilas.⁶⁶

The advantage of using directly K_GCO2a_V when simulating the CO₂ absorption process is avoiding the need to calculate the enhancement factor and individual mass transfer coefficients in liquid and gas phases. The correlations between individual mass transfer coefficients are dependent on dimensionless numbers such as the Remolds and Schmidt numbers, as well as some hydrodynamic properties of the absorber column; mostly however, they were not developed to use in specific systems (i.e., amine system and packing type). In our previous work,⁵¹ we have applied mass transfer correlations from literature, which were not developed for amine systems, in the simulation of CO₂ absorption in amine solutions. We have performed sensitivity analyses of individual mass transfer coefficients and kinetics constants of the CO₂ reaction in amine solutions (the kinetic constant is used for calculating the enhancement factor). We concluded that two mass transfer correlations had the best prediction for an absorber column profiles compared with other applied mass transfer correlations from the literature. Therefore, when K_GCO2a_V is used directly in modeling CO₂ absorption into amine solutions, there is no need to evaluate the sensitivity of the absorber model.

4. Experimental studies determining K_GCO2a_V in low-pressure absorber packed columns

Several experimental works for obtaining K_GCO2a_V for conventional and newly developed amines have been conducted by researchers for absorption columns packed with random and structured packing. The reason for determining the mass transfer coefficient based on the gas phase is easy and hassle-free measurement of the CO₂ concentration along the height of the packed column. In most of these experimental studies, first K_GCO2a_V was determined and then the effects of operating parameters on it. In addition, the empirical correlations for K_GCO2a_V have been developed based on the effects of operating parameters in absorber columns. Table 2 provides a list of published works related to the determining of K_GCO2a_V in absorber packed columns operated in low-pressure conditions. For each list, detailed information about the solvent type, packing type, the height and diameter of packed columns, and operating parameters is given. In the following, the review of the mass transfer performance in terms of K_GCO2a_V by amine-based solvents operating in low-pressure absorber packed columns has been carried out.

Table 2 Summary of operating parameters for absorber packed columns in low pressure for determining K_GCO2a_V

Solvent	Packing type	Packing height (m)	Column diameter (D)	Inlet CO2 concentration (kPa)	Gas flow rate	Amine flow rate	Inlet liquid temperature (°C)	Inlet gas temperature (°C)	Amine concentration (mol L^−1)	Inlet CO2 loading (mol CO2/mol amine)	Ref.
MEA and AMP	Ceramic Berl saddles	6.55	0.1	11.5–19.5	11.1–14.8 (mol m^−2 s^−1)	9.5–13.5 (m^3 m^−2 h^−1)	N/A	14–20	1.2–3.8	0.0–0.37	(Tontiwachwuthikul et al., 1992)
AMP	EX	1.1	0.019	4.7–15.15	12.82 (mol m^−2 s^−1)	9.73 (m^3 m^−2 h^−1)	N/A	N/A	1.14	0.027	(Aroonwilas and Tontiwachwuthikul, 1997)
AMP	EX	1.1	0.019	4–10	46.2–96.8 (kmol m^−2 h^−1)	6.1–14.8 (m^3 m^−2 h^−1)	N/A	N/A	1.1–3	0.027–0.439	(Aroonwilas and Tontiwachwuthikul, 1998)
MEA	16 mm Pall rings	2.4	0.1	14.78–15.23	119.5–199.1 (L min^−1)	0.971–2.093 (L min^−1)	24.1–26	24.8–26.3	6–7	0.195–0.225	(Demontigny et al., 2001)
MEA	IMTP-15	2.4	0.1	5–19.9	94.7–249.2 (L min^−1)	0.867–3.134 (L min^−1)	23.5–28	23.4–31.2	2.98–9.02	0.077–0.354	(Demontigny et al., 2001)
MEA	Gempak 4A	2.4	0.1	14.90–20.30	119.7–191.9 (L min^−1)	0.948–2.968 (L min^−1)	22.5–28.7	24.4–28	3–7	0.12–0.227	(Demontigny et al., 2001)
MEA and AMP	EX	1.1	0.019	Up to 15	30–97.5 (kmol m^−2 h^−1)	4.9–29.3 (m^3 m^−2 h^−1)	20–37	N/A	1.1–5.2	N/A	(Aroonwilas et al., 1999)
MEA and AMP	Gempak 4A	0.98–2.21	0.1	Up to 15	30–97.5 (kmol m^−2 h^−1)	4.9–29.3 (m^3 m^−2 h^−1)	20–37	N/A	1.1–5.2	N/A	(Aroonwilas et al., 1999)
MEA and AMP	BX	1.02	0.25	Up to 15	30–97.5 (kmol m^−2 h^−1)	4.9–29.3 (m^3 m^−2 h^−1)	20–37	N/A	1.1–5.2	N/A	(Aroonwilas et al., 1999)
MEA, AMP, DEA, MDEA and DIPA	DX	2	0.02	10	48.2 (m^3 m^−2 h^−1)	4.8–10 (m^3 m^−2 h^−1)	N/A	N/A	3	0–0.4	(Aroonwilas and Veawab, 2004)
MEA and MEA–MDEA	DX	0.165–0.825	0.02	5–15	7.26–10.13 (L min^−1)	4.8–15.3 (m^3 m^−2 h^−1)	30–50	N/A	3–7	0–0.29	(Setameteekul et al., 2006)
MEA–MeOH	DX	0.4	0.034	15	5 (L min^−1)	0.02–0.1 (L min^−1)	N/A	N/A	5	0–0.5	(Usubharatana et al., 2006b)
MEA–AMP	DX	0.5	0.02	5–15	100 (kmol m^−2 h^−1)	2.5–5 (m^3 m^−2 h^−1)	N/A	N/A	3–5	0–0.59	(Dey and Aroonwilas, 2009)
MEA, MEA–MDEA, DEAB and DEAB–MEA	DX	2.15	0.0275	14.4–14.9	4.5–5.7 (L min^−1)	0.025–0.118 (L min^−1)	20–41	19.6–30.3	1.1–7.17	0–0.3	(Naami et al., 2012)
DETA	Dixon rings	1.14	0.028	14.8–15.8	28.78–46.62 (kmol m^−2 h^−1)	2.65–7.56 (m^3 m^−2 h^−1)	30–50	N/A	1–4	0.05–0.819	(Fu et al., 2012)
DETA	DX	1.2	0.028	8.8–14.1	22.2–40.4 (kmol m^−2 h^−1)	1.95–4.87 (m^3 m^−2 h^−1)	30–50	N/A	1–4	0.184–0.826	(Fu et al., 2013)
MEA–MeOH	DX	1.25	0.028	6.7–13.8	24.37–63.54 (kmol m^−2 h^−1)	2.92–16.09 (m^3 m^−2 h^−1)	10	11–17	2.5–5	0–0.373	(Fu et al., 2015)
1DMA2P	DX	1.4	0.028	13.2–14.5	28.02 (kmol m^−2 h^−1)	4.07 (m^3 m^−2 h^−1)	40	N/A	2	0.2–0.42	(Liang et al., 2015a)
1DMA2P	Dixon rings	1.4	0.028	13.3–14.5	28.75–46.62 (kmol m^−2 h^−1)	2.65–7.56 (m^3 m^−2 h^−1)	30–60	N/A	1–3	0–0.373	(Wen et al., 2015)
NH3	N/A	2	0.015	2.8–8	1400–2300 (m^3 m^−2 h^−1)	20–30 (m^3 min^−1)	27	N/A	0.27–0.72	N/A	(Li et al., 2014)
MEA–MeOH	Sulzer BX 500, Mellapale y 500 and Pall rings	3	0.15	15	1.7–6.5 (m^3 h^−1)	20–50 (L h^−1)	20–50	N/A	5	0.2–0.4	(Gao et al., 2016)
DEEA	DX and Dixon rings	1.7	0.028	3–15	30.5–43.52 (kmol m^−2 h^−1)	3.9–11.7 (m^3 m^−2 h^−1)	27–60	N/A	1–4	0.05–0.02	(Xu et al., 2016)

---

4.1. K_GCO2a_V of conventional amines

Tontiwachwuthikul, et al.⁶⁷ have reported the mass transfer performance of CO₂ absorption into MEA and AMP by measuring the temperature and concentration along the height of the absorption packed column. They performed several experiments in a laboratory-scale absorber packed column with 12.7 mm ceramic Berl saddles packing type. The column was made of six packed bed sections with a total packing height of 6.55 m and a 0.1 m diameter. The profiles along the height of the packed column were obtained for different liquid to gas ratios, inlet CO₂ concentration in feed of flue gas, and amine concentrations. They did not determine the K_GCO2a_V values directly; rather, they modeled the packed column for CO₂ absorption with an MEA solution by applying a rate-based model. Aroonwilas and Tontiwachwuthikul⁶⁸ have studied experimentally K_GCO2a_V for CO₂ absorption into AMP solution. They performed their experiments on a laboratory-scale absorption column of 1.1 m packing height and a 0.019 m diameter, packed with EX-type structured packing with a specific surface area of approximately 1700 m² m⁻³. Their experimental results showed that the values of K_GCO2a_V at fixed operating parameters were unaffected by the CO₂ partial pressure over a range of 3–15 kPa. They also compared the values of K_GCO2a_V for two cases—columns packed with ceramic Berl saddles packing and EX-type structured packing at the same operating conditions. They found that the value of K_GCO2a_V for EX-type structured packing was six times greater than for the ceramic Berl saddles packing. Afterwards, Aroonwilas and Tontiwachwuthikul⁶³ have studied experimentally the K_GCO2a_V for CO₂ absorption into AMP solution under operating conditions different from their previous work.⁶⁸ They found that the effect of the CO₂ partial pressure on the values of K_GCO2a_V changed slightly at pressures above 6 kPa and the values of K_GCO2a_V reduced from 1 to 6 kPa. For flow rates in the absorber column, the K_GCO2a_V was unaffected by the gas flow rate but the liquid flow rate had a pronounced effect on K_GCO2a_V, which increased the values of K_GCO2a_V by increasing the liquid flow rate in the range of 6.1–14.8 m³ m⁻² h⁻¹. In addition, by increasing the CO₂ loading of the AMP solution, the K_GCO2a_V values decreased. Demontigny, et al.⁵⁸ have reported experimental data of CO₂ absorption into ultra-highly concentrated MEA solutions (up to 9 kmol m⁻³) and investigated the effects of process parameters on K_GCO2a_V in three pilot-scale absorption columns packed with random (16 mm Pall Ring and IMTP-15) and structured packing (Gempak 4A). The diameter and height of the absorption packed columns were 0.1 and 2.4 m, respectively. Their results showed that the values of K_GCO2a_V increased with increasing liquid flow rates and were unaffected by the gas flow rate. By increasing the CO₂ partial pressure and CO₂ loading, the values of K_GCO2a_V decreased. In relation to the MEA concentration, which was one of their important works, by increasing the MEA concentration up to 4 kmol m⁻³, the values of K_GCO2a_V decreased with a mild slope but increased in the range of 4–9 kmol m⁻³. They also studied the effect of the packing type on the K_GCO2a_V values, and found that structured packing (Gempak 4A) had a better performance compared with random packing (16 mm Pall Ring and IMTP-15). When comparing 16 mm Pall Ring packing with IMTP-15 packing, the IMTP-15 had greater K_GCO2a_V values. Aroonwilas, et al.⁶⁹ have performed experiments on the performance of three types of structured packing (laboratory-scale (EX), pilot-scale (Gempak 4A), and industrial-scale (SulzerBX)) in terms of the K_GCO2a_V coefficient. The experimental data was reported for CO₂ absorption into sodium hydroxide (NaOH), MEA and AMP solutions. The laboratory-scale absorption packed column was packed with 20 packing elements of EX and had a total packing height of 1.1 m and a 0.019 m diameter. The pilot-scale absorption packed column was packed with Gempak 4A stainless steel and the packing height varied between 0.98 and 2.21 m, and the absorber had a 0.1 m diameter. The third case was an industrial-scale absorption–desorption unit in which an absorber column was packed with six elements of Sulzer BX gauze structured packing, and the column had a total packing height of 1.02 m and a 0.25 m. Their results indicated that the values of K_GCO2a_V increased with an increasing liquid flow rate and liquid concentration and were unaffected by the gas flow rate. The values of K_GCO2a_V decreased with increasing CO₂ concentrations up to 15%. The values of K_GCO2a_V increased with solvent temperature from 20 °C to 37 °C and decreased with temperatures from 40 °C to 65 °C. When comparing structured packing (Gempak 4A) and IMTP-25 packing, the Gempak 4A provided two times greater K_GCO2a_V values. Aroonwilas and Veawab⁶⁵ have comprehensively investigated the performance of conventional amines such as MEA, DEA, DIPA (diisopropanolamine), MDEA, and AMP; in addition, they have investigated blends including MEA–MDEA, DEA–MDEA, MEA–AMP, and DEA–AMP. They performed the experiments in a laboratory-scale absorption column with a 2 m packing height and a 0.02 m diameter with 36 DX-type elements of structured packing. Their results were presented based on the CO₂ removal efficiency, absorber height requirement, effective mass-transfer area, and K_GCO2a_V, under identical conditions for the liquid flow rate and CO₂ loading. Their result showed that the CO₂ removal efficiency in a CO₂ loading of zero was in the order MEA > DEA > AMP > DIPA > MDEA. The value of 100% of CO₂ removal efficiency was obtained for MEA, DEA, and AMP, requiring a 0.75, 1.75, and 2.0 m height of the packed column, respectively. Therefore, MEA showed a better performance in comparison with other studied amines. For blended amines, the value of 100% of CO₂ removal efficiency was obtained for MEA–AMP, DEA–AMP, MEA–MDEA, and DEA–MDEA, requiring a 1.2, 2.3, 3.3, and 5.4 m height of the packed column, respectively. The authors also assessed the performance in terms of the effective mass-transfer area under identical processing parameters and found that MEA provided the highest mass-transfer area among the tested amines, including DEA, DIPA, and MDEA. They also showed that the values of K_GCO2a_V at different CO₂ loadings for MEA were higher compared with other tested amines such as DEA, AMP, DIPA, and MDEA. Setameteekul, et al.⁷⁰ studied the mass transfer performance for CO₂ absorption in a MEA and MDEA blended amine. The experiments were performed based on the factorial experimental design method (a statistical method), and conducted more than 106 tests with three replications in an absorption column packed with a DX-type packing. The packing height varied between 0.165 and 0.825 m, and the absorber had a 0.02 m diameter. The results of the work by Setameteekul, et al.⁷⁰ indicated that the solvent temperature and solvent concentration have the largest effects on the K_GCO2a_V values and the other process parameters have smaller effects. Dey and Aroonwilas⁶⁶ used the blended MEA–AMP amine to determine K_GCO2a_V by using only two data sampling points of CO₂ concentrations at the bottom and top of an absorber column. In fact, they obtained the average values of K_GCO2a_V for an absorber column packed with a DX-type structured packing. Their results showed that the K_GCO2a_V values increased with an increasing liquid flow rate, temperature, and total amine concentrations, and decreased with increasing CO₂ partial pressure of the feed gas and CO₂ loading of the amines. The addition of higher concentrations of MEA in the mixed MEA–AMP amine led to an increase in the K_GCO2a_V values, except at high CO₂ loadings. This was because of a lower reaction rate of CO₂ with AMP compared with MEA.

Jeon, et al.⁷¹ have studied the mass transfer performance and effect of adding ammonia (NH₃) to AMP and MDEA. They determined the K_GCO2a_V values in an absorption packed column with a 1.5 m packing height and a 0.05 m diameter by testing two packing types including 6 mm ceramic Raschig rings, and a wire gauze laboratory-structured packing. They showed that the K_GCO2a_V values at a CO₂ partial pressure of 15 kPa increased for both mentioned systems by using structured packing, and increased even more by adding NH₃ from 1 wt% to 3 wt%. They also showed that the K_GCO2a_V values increased at lower CO₂ partial pressure and higher liquid-to-gas ratios. The overall conclusion of their work was that adding NH₃ to AMP and MDEA and using structured packing produced higher K_GCO2a_V values. Li, et al.⁷² have performed experiments for CO₂ absorption using an NH₃ solution to determine the K_GCO2a_V values in an absorber column packed with a novel structured packing with diversion windows type. The height of the absorber column and column diameter were 2.4 m (packing height: 2 m) and 0.15 m, respectively. Their results showed that K_GCO2a_V was enhanced by increasing the liquid flow rate and its concentration. However, the K_GCO2a_V values decreased when the CO₂ partial pressure increased to 8 kPa, and were unaffected by the gas flow rate. Kang, et al.⁷³ have tested various packing types including ceramic Raschig rings, Berl saddles, a structured gauze packing and a hybrid of Raschig rings and a structured packing in different ratios to investigate the mass transfer performance of a CO₂–MEA–AMP system. Their results showed that CO₂ removal efficiencies of Raschig rings, Berl saddles, and the structured packing materials provided higher values for the MEA than the AMP solution, and that the structured packing had a greater efficiency than the random packing. They improved the performance of single random and structured packing materials by mixing them in ratios of 1 : 1, 2 : 1, and 1 : 2. The optimal performance was obtained for the 2 : 1 ratio (structured packing/Raschig rings). The K_GCO2a_V parameter decreased in the order 2 : 1 hybrid packing > structured packing > Raschig rings > Berl saddles.

4.2. K_GCO2a_V of hybrid amines

Usubharatana, et al.⁷⁴ have studied K_GCO2a_V for CO₂ absorption by using a hybrid solution containing MEA and methanol. Their experiments were carried out on a laboratory-scale absorption column with a 0.4 m packing height and a 0.034 m diameter, packed with DX-type structured packing with a specific surface area of around 900 m² m⁻³. They showed that by increasing the methanol concentration in the MEA solution, K_GCO2a_V increased. By increasing the liquid flow rate, K_GCO2a_V increased and led to a decrease of the methanol carryover because of its vaporization to the top of the absorber column. Fu, et al.⁶⁰ have obtained experimentally K_GCO2a_V by using a hybrid solution containing MEA and methanol. They performed the experiment on a structured absorber column (packing height 1.25 m; column diameter 0.028 m) packed with a DX packing. They reported the CO₂ concentration and temperature data along the height of the packed column (seven points along column) for 33 tested runs. The effects of key process parameters including the MEA concentration, CO₂ loading, amine flow rate and gas flow rate were investigated, and it was found that K_GCO2a_V decreased with increasing CO₂ loading, gas flow rate (in higher concentrations of the MEA solution), and CO₂ concentration in the inlet feed gas. The authors suggested using a high liquid-to-gas ratio and a low temperature of the feed gas and amine in the absorber column in order to prevent the vaporization of methanol at the top of the absorber column. Gao, et al.⁷⁵ have studied the effects of different process parameters on the K_GCO2a_V values using a hybrid solvent: MEA–methanol. The experiments performed on an absorber column (packing height 3 m; column diameter 0.15 m) packed with three different packing types including Sulzer BX500, Mellapale Y500, and Pall rings 16 × 16. Their results indicated that the Sulzer BX500 had higher K_GCO2a_V values than the Mellapale Y500. The reason for this was the good uniform distribution of gas and liquid on the packing surface. In addition, their results showed that (1) the K_GCO2a_V values increased as the CO₂ lean loading decreased and the hybrid solution temperature, hybrid solution flow rate and gas flow rate increased, (2) the optimal temperature for reducing methanol evaporation was 20 °C, and (3) operating at a high liquid-to-gas ratio led to a reduction of methanol evaporation as suggested by Fu, et al.⁶⁰ study.

4.3. K_GCO2a_V of new amines

Naami, et al.⁵⁹ have studied experimentally K_GCO2a_V for CO₂ absorption using a DEAB solution. They also investigated absorption using MEA, MDEA, DEAB, and blended solutions included MDEA–MEA and DEAB–MEA. The authors performed their experiments on an absorber column packed with DX-type structured packing (height 2.15 m; diameter 0.0275 m). It was concluded that the presence of MEA in the DEAB solution increased the K_GCO2a_V values. The authors also investigated the effect of the liquid flow rate on the K_GCO2a_V values, and found that by increasing it in a narrow range, K_GCO2a_V values increased rapidly. Nevertheless, the K_GCO2a_V values were unaffected by the gas flow rate for the above-mentioned amines. The absorption capacity and cyclic capacity between DEAB and MDEA were also compared in the study, and it was found that the new amine, DEAB, has a higher CO₂ absorption and cyclic capacity than MDEA, which causes a reduction in the amine circulation rate and energy consumption for amine regeneration. Fu, et al.⁶¹ have determined experimentally K_GCO2a_V for CO₂ absorption into a DETA solution. They performed the experiment on a random absorber column (packing height 1.14 m; column diameter 0.024 m) packed with Dixon rings which have a specific surface area of around 2400 m² m⁻³. Their results showed that the K_GCO2a_V values for DETA were higher compared with MEA, and the K_GCO2a_V values increased as the DETA flow rate, DETA concentration, and inlet temperature increased. However, the K_GCO2a_V values decreased when the CO₂ loading of DETA increased. Fu, et al.⁷⁶ have investigated the K_GCO2a_V parameter using a DETA solution. They performed the experiment on a structured absorber column (packing height 1.7 m; column diameter 0.028 m) packed with a DX packing which has a specific surface area of around 900 m² m⁻³. They compared the K_GCO2a_V values between DETA–CO₂ and MEA–CO₂ systems, and found that the K_GCO2a_V values of DETA were higher compared with MEA. In addition, they showed that by increasing the DETA flow rate, concentration, and inlet temperature, the K_GCO2a_V values increased. Nevertheless, the K_GCO2a_V values decreased as the CO₂ loading of DETA increased. Liang, et al.⁴⁴ have investigated the mass transfer performance of CO₂ absorption by a 1DMA2P solution. They obtained K_GCO2a_V for a structured absorber column (packing height 1.4 m; column diameter 0.028 m) packed with DX packing. The authors compared the K_GCO2a_V values for 1DMA2P–CO₂ with those for MEA–CO₂ and MDEA–CO₂ systems, and found that the K_GCO2a_V values of the MEA solution were higher compared with 1DMA2P and MDEA solutions. They ranked them as MEA > 1DMA2P > MDEA. In addition, they showed that by increasing the CO₂ loading for the three mentioned systems, the K_GCO2a_V values decreased. Following the study of Liang, et al.,⁴⁴ Wen, et al.⁷⁷ obtained the K_GCO2a_V values in an absorber column packed with Dixon rings for 1DMA2P–CO₂ system. They compared the K_GCO2a_V values for 1DMA2P–CO₂ with those for the MDEA–CO₂ system, and found that the K_GCO2a_V values of the new amine solution were higher compared with the MDEA solution. In addition, the effects of the gas flow rate, amine concentration, amine flow rate, CO₂ loading, and amine temperature on the K_GCO2a_V values were investigated and the results indicated that the K_GCO2a_V values increased with increasing amine concentrations and amine flow rates, but decreased with increasing CO₂ loading. The inert gas flow rate had little effect on the K_GCO2a_V values. For the amine temperature, the K_GCO2a_V values increased up to 323 K in temperature range of 303–333 K, and decreased as amine temperature went above 323 K. Xu, et al.³³ have conducted experiments for CO₂ absorption in a DEEA solution. The experiments were carried out for both random and structured packing columns, namely DX and Dixon rings. The authors obtained the K_GCO2a_V values in an absorber column with a 1.7 m packing height and a 0.028 m diameter. The effects of important process parameters on the K_GCO2a_V values were investigated. The results showed that the DEEA concentration, CO₂ loading, and liquid flow rate had the most pronounced effect on K_GCO2a_V, whilst the gas flow rate had a negligible effect. The authors compared the K_GCO2a_V values between MDEA–CO₂ and DEEA–CO₂ systems, and showed that the K_GCO2a_V values for DEEA were higher compared with the MDEA solution. They also showed that the DX-type structured packing enhanced the K_GCO2a_V values and cyclic capacity compared with the Dixon rings packing. Chen, et al.⁷⁸ have used the experimental design method including the Taguchi method to select blended amines (MEA–DEAE, MEA–MDEA, MEA–DIPA, and MEA + AMP) as promising solvents for CO₂ capture in a packed column (packing height 1.2 m; column diameter 0.05 m). The optimum points and effects of three key process parameters including the liquid flow rate, gas flow rate and amine concentration on the absorption rate and K_GCO2a_V were studied. Sixteen runs were carried out by the Taguchi method at four levels for four factors to obtain the response values (absorption efficiency, absorption rate, scrubbing factor, and K_GCO2a_V). In case of K_GCO2a_V, the values were obtained using a two-film model. The results of the study indicated that the gas flow rate and amine flow rate are significant parameters, while the type of amine and amine concentration showed little effect on the K_GCO2a_V values.

5. Experimental studies determining K_GCO2a_V in high-pressure absorber packed columns

In the previous section, we reviewed the experimental studies for determining K_GCO2a_V in low-pressure absorber packed columns. The evaluation of the K_GCO2a_V coefficient under high pressure is essential for removing CO₂ from natural gas streams. Few studies are available in the literature for determining K_GCO2a_V in high-pressure absorber packed columns.

Abdul Halim, et al.⁶² performed experiments in order to remove CO₂ from a mixture of CO₂ and methane using an MEA solution under a pressure of 50 bar. The experiments were performed in a packed column (packing height 2.04 m; column diameter 0.046 m) with a packing type of Sulzer gauze, which has a surface area of around 500 m² m⁻³. The aim of their work was to determine K_GCO2a_V and the CO₂ removal efficiency under high pressure. The K_GCO2a_V values were obtained at a fixed CO₂ concentration of 20% mol, an amine concentration range of 1–4 kmol m⁻³, an amine flow rate range of 4.51–8.1 m³ m⁻² h⁻¹, a gas flow rate range of 18.89–35.08 kmol m⁻² h⁻¹, and feed temperature range of 27 °C to 45 °C. They validated their experiments' reliability by the work of Maneeintr, et al.⁷⁹ under low-pressure conditions. Their results indicated that high pressure, amine flow rate, and amine concentration had a significant effect on K_GCO2a_V, and that by increasing these parameters, the K_GCO2a_V values would increase. They also showed that the K_GCO2a_V values were unaffected by the gas flow rate and that the optimal point for the temperature of the inlet amine to the absorption column was found to be 40 °C. Halim, et al.⁸⁰ performed their experiments for removing CO₂ from a mixture of CO₂ and methane using an AMP–PZ solution in a packed column (packing height 2.04 m; column diameter 0.046 m) with a packing type of Sulzer gauze. The aim of their work was to determine the K_GCO2a_V values under high pressures (10–40 bar). The K_GCO2a_V values were obtained at a fixed CO₂ concentration of 40% mole, a total amine concentration of 30% wt (PZ with 7 wt% and AMP with 23 wt%), an amine flow rate range of 2.89–3.97 m³ m⁻² h⁻¹, a gas flow rate range of 33–51 kmol m⁻² h⁻¹, and a feed temperature range of 30 °C to 35 °C. Their results indicated that high pressure, amine flow rate, and amine concentration had a large effect on K_GCO2a_V, and; the K_GCO2a_V values were unaffected by the gas flow rate. Following the work of Halim, et al.,⁸⁰ Hairul, et al.⁸¹ performed experiments under different conditions of operating parameters. Their setup for an absorber column was similar the one in the work of Halim, et al.⁸⁰ They determined K_GCO2a_V under different operating conditions over a pressure range of 10–50 bar, a CO₂ concentration range of 30% to 50% mol, an amine concentration range of 3–9 wt% for PZ and 23–30 wt% for AMP, an amine flow rate range of 2.89–4.33 m³ m⁻² h⁻¹, a gas flow rate range of 33–40 kmol m⁻² h⁻¹, and feed temperature range of 30 °C to 35 °C. Their results showed that by increasing pressure above 20 bar, K_GCO2a_V increased, and by increasing the CO₂ concentration in the feed gas, K_GCO2a_V decreased. In addition, the performance of AMP in removing CO₂ from natural gas was compared with the AMP–PZ solution, and results showed the AMP–PZ system to be superior in terms of CO₂ removal efficiency.

6. General investigation of operating parameters affecting K_GCO2a_V

In the above-mentioned reviewed works on K_GCO2a_V, initially, the K_GCO2a_V values were obtained experimentally using the concentration profile of CO₂ in the gas phase and the mass balance equation, and then the effects of process parameters on it were investigated. In previous sections, we have shown how are the response values affected by increasing and decreasing one operating parameter. An understanding of how K_GCO2a_V of different amine-based solvents changes with different operating parameters and configurations of pilot-plant is significant in evaluating and optimizing CO₂ removal processes.^14,82 The variables affecting K_GCO2a_V in the above-mentioned reviewed works have been described in detail as follows.

6.1. CO₂ partial pressure

Increasing the CO₂ partial pressure in the gas feed to the absorber column has two effects on K_GCO2a_V. First, increasing the CO₂ partial pressure can lead to an increase of the partial pressure gradient⁸³ because of the consummation of more actively free-amine molecules and, as a result, a decrease of K_GCO2a_V can occur (according to eqn (2)). Second, increasing the CO₂ partial pressure can intensify the gas flow turbulence in the absorber column and, as a result, increase K_GCO2a_V.⁸⁴ In the above-mentioned reviewed works, researchers showed that by increasing the CO₂ partial pressure in the gas feed, K_GCO2a_V decreased. This effect shows that amine solutions, especially those which have higher reaction kinetic constant values (such as MEA or PZ), have good performance under lower partial pressure than under higher partial pressure.^85,86 This effect indicates that amine solutions have a high CO₂ removal efficiency in a lower partial pressure of CO₂ in the gas feed. This occurrence shows that the liquid phase resistance dominates the mass transfer performance of absorption into amine solutions.⁸⁷

6.2. Gas flow rate

The turbulence increases in the gas phase because an increase in the gas flow rate can lead to an increase of K_GCO2a_V.^60,77,78 However, experiments by many researchers, which were pointed out in previous sections, differed from the above-mentioned prediction by Fu, et al.⁶⁰ and Chen, et al.⁷⁸ works. They showed that the liquid film could control the process of CO₂ absorption in an amine solution and, as a result, K_GCO2a_V in such a system is unaffected by the gas flow rate.

6.3. Liquid flow rate

One of the key parameters, which can affect the mass transfer performance, is the liquid flow rate. Many researchers showed that when the liquid flow rate increased, K_GCO2a_V increased as well. The reasons for this effect are (1) an increase in the amine flow rate can cause an increase in free active molecules of amines for high CO₂ absorption, (2) an increase in the amine flow rate has the greatest effect on the surface of the packing, increasing the wet surface area between the amine and gas phases, and (3) an increase in the mass transfer coefficient in the liquid phase decreases the mass transfer resistance in the liquid phase and, as a result, increases the mass transfer coefficient in the gas phase.^{50,63,65,66,68} However, increasing this factor above the optimum point can lead to a loss of amines and to high-energy consumption for amine regeneration.

6.4. Liquid concentration

In general, increasing the amine concentration causes an increase of the K_GCO2a_V values. This is because of the availability of an extra amount of amine molecules for CO₂ absorption at the interface of the gas and liquid, and this increases the possibility for CO₂ to react with amines over a larger active surface area.^33,60,76 As mentioned before, the absorption of CO₂ in amine solutions is a process, which is controlled by the liquid phase. As a result, this phenomenon decreases the resistance in the liquid phase and increases the mass transfer coefficient in the gas phase. However, increasing the concentration leads to an increase of the viscosity, which can hinder the diffusion of CO₂ into amines. Under these conditions, the balance between increasing K_GCO2a_V and the cost involved should be considered by increasing the amine concentration.

6.5. Liquid temperature

Another other key parameter is liquid temperature, which can have an effect on K_GCO2a_V, the reaction kinetic, and equilibrium solubility. According to the Arrhenius equation,⁸⁸ the reaction kinetic constant for the reaction between CO₂ and an amine solution is temperature dependent, and by increasing the temperature a higher reaction rate constant is achieved.⁸⁹ Consequently, the enhancement factor can be increased and consequently, K_GCO2a_V can be increased as well.⁹⁰ Under absorption conditions, this effect can be reversed at higher temperatures when the reaction between CO₂ and amines becomes reversible (it approaches desorption conditions), and it can decrease K_GCO2a_V. In addition, the higher temperature can lead to an increase in the vapor pressure of CO₂ above the amine solution's and this can cause an increase in the Henry's law solubility constant; as result, a decrease of K_GCO2a_V and the solubility of CO₂ in amine solutions can occur.⁹¹ The balance between the above-mentioned parameters should be considered for increasing K_GCO2a_V.

6.6. CO₂ loading

An increase of CO₂ loading in amine solutions leads to a decrease in the existing active amine concentration, which consequently decreases K_GCO2a_V.^{33,71,74,76,77} This is obvious when the amount of CO₂ loading in the lean amine solution is high, the mass transfer driving force from the gas phase to the liquid phase will decrease and, in these cases, increasing the liquid flow rate to compensate the low absorption rate is not an effective method. The optimum way for decreasing the CO₂ loading in amine solutions is heating the amine solution to increase K_GCO2a_V.

6.7. Absorption pressure

Increase of pressure in an absorber column can lead to a decrease of CO₂ concentration in equilibrium with the amine solution and, as a result, the driving force for mass transfer can increase. Halim, et al.⁸⁰ showed that the absorption of CO₂ (from natural gas) in an AMP–PZ solution under a pressure range of 20–40 bar increased K_GCO2a_V. However, the increasing absorber pressure in CO₂ capture post-combustion processes will increase the cost of operation.⁹² Increasing K_GCO2a_V in CO₂ capture post-combustion processes by increasing pressure is not cost-effective.

6.8. Packing type

Packing is helpful in making more time for gas–liquid contact throughout the CO₂ absorption process, so it can increase the surface area and K_GCO2a_V in the packed column. Recently, researchers showed that using a structured packing in the absorption column could create higher surface are as compared with random packing.⁹³ This is because of greatly higher geometric wet surface areas per structured packing volume unit. However, the packing surface area should not be the only selection criteria for creating a higher K_GCO2a_V in the packed column.⁹⁴ The other parameters affecting packing, such as packing arrangement pattern, angle of corrugation, void fraction, and height of crimp, should be also considered when designing the absorption packed column, to minimize the pressure drop and capacity of the liquid entrainment.

7. Empirical correlations for K_GCO2a_V in packed columns

Several mass-transfer coefficient correlations are available for absorber columns packed with random and structured packing.^12,95–98 These developed correlations differ in their accuracy and system-specific applicability. By having these correlations and calculating the enhancement factor, K_GCO2a_V can be obtained. However, these correlations can increase the error in calculations, for example in rate-based models, and one needs to perform a sensitive analysis for the mentioned correlations, which may not be applicable to CO₂–amine systems.

The review on these correlations was done by Wang, et al.¹⁶ for random and structured packing columns. Herein, we have reviewed the empirical correlations of K_GCO2a_V in absorber packed columns, which were obtained experimentally, mainly from analyzing the effects of operating parameters on K_GCO2a_V in CO₂–amine systems. The reviewed correlations in this study are presented in Table 3 and the corresponding values of operating parameters for developed correlations have been listed in Table 2.

Table 3 Review of the developed empirical correlations of K_GCO2a_V in low-pressure absorber packed columns

Solvent	Packing type	Ref.	Correlations
MEA	Ceramic rings, glass rings, steel rings and ceramic saddles	(Kohl and Riesenfeld, 1985)
MEA	16 mm Pall rings, IMTP-15 and 4A Gempak	(Demontigny et al., 2001)
AMP	EX	(Aroonwilas and Tontiwachwuthikul, 1998)
MEA	DX	(Setameteekul et al., 2006)	KGae = 4.106 − 0.370A − 0.077B + 0.044C − 0.001758D − 4.74E + 0.00215A^2 + 0.004B^2 − 0.00162C^2 + 6.105E^2 − 0.02AC +…
MEA–AMP	DX	(Dey and Aroonwilas, 2009)
DEAB	DX	(Sema et al., 2013)
MEA	Dixon rings	(Fu et al., 2012)
DETA	Dixon rings	(Fu et al., 2012)
MEA–MDEA	DX	(Naami et al., 2013)
1DMA2P	Dixon rings	(Wen et al., 2015)
NH3	Packing with diversion windows	(Li et al., 2014)
DEEA	Dixon ring	(Xu et al., 2016)
DEEA	DX

---

Kohl and Riesenfeld⁹⁹ have developed an empirical correlation of K_GCO2a_V for CO₂ absorption in an MEA solution. The correlation applies to an absorption column packed with random packing. In their study, the K_GCO2a_V correlation is as function of the amine flow rate (L), amine loading (α), equilibrium loading (α_e), amine concentration (M), viscosity (μ), CO₂ partial pressure, and amine temperature. Their empirical correlation is

In the above equation, the F value is the packing factor, and Kohl and Riesenfeld⁹⁹ have reported F values for some random packing types. For example, the F values have been reported as 0.0021 and 0.003 for 1 inch ceramic saddles and 3/8 inch ceramic saddles, respectively. Demontigny, et al.⁵⁸ have developed an empirical correlation of K_GCO2a_V for a CO₂–MEA system based on the work of Kohl and Riesenfeld.⁹⁹ The correlation is as function of the same parameters developed by Kohl and Riesenfeld⁹⁹ except for the viscosity term as this term was not considered in the work of Demontigny, et al.⁵⁸ The developed equation in their work is

Demontigny, et al.⁵⁸ have reported that the eqn (7) had errors when predicting results using experimental data and is valid only up to a concentration of 3 kmol m⁻³. They reported that the errors arose from the CO₂ loading data. Despite of variations of CO₂ loading along the height of the packed column, the CO₂ loading value was assumed in its saturation point (0.5 mol CO₂/mol amine). Aroonwilas and Tontiwachwuthikul⁶³ have developed an empirical correlation of K_GCO2a_V for a CO₂–AMP system based on the work of Demontigny, et al.⁵⁸ The correlation obtained for an absorber column packed with a laboratory DX-type packing. The correlation is as function of the same parameters developed in the Demontigny, et al.⁵⁸ study. The developed correlation of K_GCO2a_V for a CO₂–AMP system byAroonwilas and Tontiwachwuthikul⁶³ is

They showed that the error between experimental data of K_GCO2a_V and results predicted by the above equation was 16.5%. Setameteekul, et al.⁷⁰ have developed a K_GCO2a_V correlation based on the experimental design factorial method. The K_GCO2a_V correlation was obtained for two systems including CO₂–MEA and CO₂–MEA–MDEA as a function of the amine flow rate, amine loading, amine concentration, amine temperature, and CO₂ partial pressure. The correlation is based on the results of interaction of operating parameters, and for the CO₂–MEA system, it is obtained from the equation below.

K_Ga_e = 4.106 − 0.370A − 0.077B + 0.044C − 0.001758D − 4.74E + 0.00215A² + 0.004B² − 0.00162C² + 6.105E² − 0.02AC +… (9)

It should be noted that the above equation was obtained after truncation of insignificant parameters. Insignificant parameters were those, which had less probable values in the analysis of variance results. The results obtained by the factorial method had an error of 18.39% for the CO₂–MEA system and a high error reported for the CO₂–MEA–MDEA system. Dey and Aroonwilas⁶⁶ have developed the K_GCO2a_V correlation for CO₂ absorption into a MEA–AMP solution in an absorber column packed with a laboratory DX-type packing as

In eqn (10), AMP/MEA is the molar ratio of the amine mixture, α is the CO₂ loading, x_CO2 is the molar fraction of CO₂ in liquid, C_s is the amine concentration, L1 is the amine flow rate and T is the amine temperature. The regressed coefficients (K, A, B, C, D, E, and F) were obtained for different AMP/MEA molar ratios of the amine mixture, and the errors were reported as 6.75%, 10.05%, 11.72% and 12.2% for MEA, MEA : AMP = 1 : 2, MEA : AMP = 1 : 1, and MEA : AMP = 2 : 1, respectively. Sema, et al.³² have developed the K_GCO2a_V correlation for CO₂ absorption into a DEAB solution. They have reported that the correlation is valid for an absorber column packed with a DX-type packing. Their results showed that the reported error for the developed correlation (eqn (11)) against experimental data was 14.6%.

Fu, et al.⁶¹ have developed a predictive correlation for K_GCO2a_V in a CO₂–DETA system. This correlation is valid for an absorber column packed with Dixon rings. The developed correlation was obtained based on the work of Demontigny, et al.⁵⁸ The correlation was validated the K_GCO2a_V; first, by developing the correlation according to the work of Kohl and Riesenfeld⁹⁹ for a CO₂–MEA system, and then, the correlation was validated for a CO₂–DETA system. These correlations are, for the CO₂–MEA (eqn (12)) and CO₂–DETA (eqn (13)) systems:

The predicted results by the correlations above showed errors of 16% and 14%, for CO₂–DETA and CO₂–MEA systems, respectively, against experimental results. Fu, et al.⁷⁶ have applied an artificial neural network (ANN) in order to estimate of K_GCO2a_V values for absorption of CO₂ into a DETA solution in an absorber column packed with laboratory Ex-type packing. In the ANN model, 8 parameters such as gas flow rate, CO₂ partial pressure, liquid flow rate, amine concentration, amine density, amine viscosity, diffusion of CO₂ in an amine solution, and the cycling loading of amines are defined as input parameters and K_GCO2a_V as the output parameter. To develop the ANN model, 75% and 25% of the dataset were used for training and testing, respectively. Their results showed that the ANN model could predict experimental data very well, with an error of 7.6% for the CO₂–DETA system in a packed column. Naami, et al.¹⁰⁰ have developed a correlation for the K_GCO2a_V in a CO₂–MEA–MDEA system in an absorber packed with a DX-type packing over MDEA–MEA concentrations of 27/3, 25/5, and 23/7 wt%. The developed correlation was based on the work of Dey and Aroonwilas⁶⁶ (eqn (10)), and is

In eqn (14), MDEA/MEA is the molar ratio of the amine mixture, α is the CO₂ loading of the amine, x_CO2 is the molar fraction of CO₂ in the liquid, C_s is the amine concentration, L1 is the amine flow rate and T is the amine temperature. The regression coefficients (K, A, B, C, D, E, and F) were obtained for different MDEA/MEA molar ratios of the amine mixture, and the errors were reported as 20.9%, 21.7%, and 22.8% for MEA/MDEA = 3/27 wt%, MEA/MDEA = 5/25 wt%, and MEA/MDEA = 7/23 wt%, respectively. Wen, et al.⁷⁷ have correlated the K_GCO2a_V data in an absorber column packed with Dixon rings for a 1DMA2P–CO₂ system. Their developed correlation was based on the Demontigny, et al.⁵⁸ study:

Their results showed that the developed correlation had an error of 9.8% when predicting experimental data. Li, et al.⁷² have developed the K_GCO2a_V correlation in an absorber column packing with diversion windows type in a NH₃–CO₂ system. The correlation is based on important parameters such as the liquid flow rate (L), NH₃ concentration and CO₂ partial pressure:

In addition, the modeling and simulation of absorption of CO₂ in a NH₃ solution was carried out using the computational mass transfer model with a developed correlation of K_GCO2a_V (eqn (16)). The authors did not report the error values between the developed correlation and experimental data, but trends of the developed correlation showed that it had a small error in comparison with the experimental data. Xu, et al.³³ have developed two correlations of K_GCO2a_V for absorber columns packed with Dixon rings and DX-type packing materials. They developed the correlation for a CO₂–DEEA system based on the work of Demontigny, et al.⁵⁸ as follows:

Their results showed that the errors obtained by eqn (17) (Dixon rings packing) and eqn (18) (DX-type packing) were 3% and 8%, respectively.

8. Conclusions

In this study, a review has been provided on K_GCO2a_V associated with amine-based solvents in absorption packed columns. As a first step, we have reviewed the experimental determination of K_GCO2a_V, previously done by researchers. With measuring CO₂ concentrations in the gas phase along the height of absorber columns and using the two-film theory, K_GCO2a_V was obtained by researchers for columns packed with random and structured packing materials. Details of pilot-plant data related to the determination of K_GCO2a_V have been reported, and the determination of K_GCO2a_V has been reviewed for various amine-based solvents (conventional amines, hybrid amines, newly developed amines) over a range of operating parameters of low- and high-pressure absorber columns. Second, we have reviewed and described the effects of operating parameters on the K_GCO2a_V data in absorber packed columns. In most studies, authors showed that the K_GCO2a_V values are unaffected by the gas flow rate, and by increasing the liquid flow rate, amine concentration, and column pressure, the K_GCO2a_V values increased. Increasing the CO₂ loading of amines and the partial pressure of CO₂ in the gas feed lead to a K_GCO2a_V decrease. Third, we have reviewed the developed empirical correlations of K_GCO2a_V for absorber low-pressure columns. The most developed correlations are functions of operating parameters of the absorber column and somewhat depend on the physical properties such as viscosity. It should be noted that developed correlations, which were reviewed in this study, are only based on specific systems (such as packing type and amine solvent). The advantages of these correlations are that one does not need to calculate the enhancement factor and perform a sensitivity analysis of mass transfer coefficients in both liquid and gas phases.

9. Prospects

In recent years, CO₂ removal using amine solutions has attracted extensive consideration by many researchers. One of the most important subjects is to evaluate the mass transfer performance in absorption packed columns in terms of mass transfer coefficients. As we have discussed in this study, the determination of the K_GCO2a_V term can help the designer gain a deeper understanding of such a system because of the simplicity of using CO₂ concentration measurements in the gas phase for absorber columns. However, some attempts are needed to gain a better understanding of the mass transfer performance in terms of K_GCO2a_V. Following are the possible future directions concerning the analysis and evolution of K_GCO2a_V in absorption packed columns:

• Determination of K_GCO2a_V for a pilot-plant and plant data.

• Determination of K_GCO2a_V for a stripper column under high temperature.

• More analyses are needed of K_GCO2a_V parameter in high-pressure conditions.

• Sophisticated equipment is required for high quality measuring of CO₂ along the packed column which is used for determination of K_GCO2a_V.

• Considering other gases in the flue gas feed to the absorber when determining K_GCO2a_V.

• Simultaneous effects of operating parameters on K_GCO2a_V need to be investigated, for example using statistical methods.

• Determination of K_GCO2a_V using a optimization technique such as the work of Ji, et al.¹⁰¹ who did this for the mass transfer coefficient in the liquid phase.

• More research on mixed amines is needed; for example, the effects of activators on the amine solutions for the determination of K_GCO2a_V.

• More research on the packing type for determination of K_GCO2a_V is needed, especially for plant data.

• The empirically developed correlations should not depend only on the operating parameters but should also depend on physical properties, while for the CO₂–amines systems reaction kinetics are also important.

• Applying the empirically developed correlation of K_GCO2a_V directly in the rate-based model and considering the errors.

• Modeling and optimization K_GCO2a_V are needed to find the optimum operating parameters effecting K_GCO2a_V.

Nomenclature

aV	Effective interfacial area
AC	Cross-sectional area of the column
AMP	2-Amino-2-methyl-1-propanol
C	Amine concentration
	CO2 concentration at interface
DEA	Diethanolamine
DEEA	N,N-Diethylethanolamine
DEAB	4-Diethylamino-2-butanol
DETA	Diethylenetriamine
DMA2P	1-Dimethylamino-2-propanol
F	Packing factor
g	Gas
G	Gas flow rate
IMTP	Intalox metal tower packing
KGaV	Gas-phase volumetric overall mass transfer coefficient
L	Liquid flow rate
MEA	Monoethanolamine
MDEA	Methyldiethanolamine
PZ	piperazine
	Partial pressure of CO2 at interface

References

1. A. J. McMichael, R. E. Woodruff and S. Hales, Lancet, 2006, 367, 859–869.
2. N. Panwar, S. Kaushik and S. Kothari, Renewable Sustainable Energy Rev., 2011, 15, 1513–1524.
3. S. A. Rice, Environments, 2014, 3, 7–15.
4. E. S. Rubin, H. Mantripragada, A. Marks, P. Versteeg and J. Kitchin, Progr. Energy Combust. Sci., 2012, 38, 630–671.
5. T. N. G. Borhani, A. Azarpour, V. Akbari, S. R. W. Alwi and Z. A. Manan, Int. J. Greenhouse Gas Control, 2015, 41, 142–162.
6. Z. H. Liang, W. Rongwong, H. Liu, K. Fu, H. Gao, F. Cao, R. Zhang, T. Sema, A. Henni and K. Sumon, Int. J. Greenhouse Gas Control, 2015, 40, 26–54.
7. M. Mofarahi, Y. Khojasteh, H. Khaledi and A. Farahnak, Energy, 2008, 33, 1311–1319.
8. P. Usubharatana, D. McMartin, A. Veawab and P. Tontiwachwuthikul, Ind. Eng. Chem. Res., 2006, 45, 2558–2568.
9. A. L. Kohl and R. Nielsen, Gas purification, Gulf Professional Publishing, 1997.
10. J. Kuntz and A. Aroonwilas, Ind. Eng. Chem. Res., 2008, 47, 145–153.
11. Z. H. Liang, T. Sanpasertparnich, P. P. Tontiwachwuthikul, D. Gelowitz and R. Idem, Carbon Manage., 2011, 2, 265–288.
12. R. F. Strigle, Random Packings and Packed Towers, 2nd edn, Gulf Publishing Company, Houston, TX, 1987.
13. P. Tontiwachwuthikul and A. Chakma, J. Membr. Sci., 2006, 277, 99–107.
14. K. Fu, G. Chen, Z. Liang, T. Sema, R. Idem and P. Tontiwachwuthikul, Ind. Eng. Chem. Res., 2014, 53, 4413–4423.
15. N. Razi, O. Bolland and H. Svendsen, Int. J. Greenhouse Gas Control, 2012, 9, 193–219.
16. G. Wang, X. Yuan and K. Yu, Ind. Eng. Chem. Res., 2005, 44, 8715–8729.
17. N. Razi, H. F. Svendsen and O. Bolland, Int. J. Greenhouse Gas Control, 2014, 26, 93–108.
18. S. Ziaii, S. Cohen, G. T. Rochelle, T. F. Edgar and M. E. Webber, Energy Procedia, 2009, 1, 4047–4053.
19. J. Salazar, U. Diwekar, K. Joback, A. H. Berger and A. S. Bhown, Energy Procedia, 2013, 37, 257–264.
20. A. Bandyopadhyay, Clean Technol. Environ. Policy, 2011, 13, 269–294.
21. M. Iijima, T. Nagayasu, T. Kamijyo and S. Nakatani, Mitsubishi Heavy Ind. Tech. Rev., 2011, 48, 26.
22. F. Y. Jou, A. E. Mather and F. D. Otto, Can. J. Chem. Eng., 1995, 73, 140–147.
23. K. Maneeintr, R. O. Idem, P. Tontiwachwuthikul and A. G. H. Wee, Energy Procedia, 2009, 1, 1327–1334.
24. G. Puxty, R. Rowland, A. Allport, Q. Yang, M. Bown, R. Burns, M. Maeder and M. Attalla, Environ. Sci. Technol., 2009, 43, 6427–6433.
25. G. Puxty and R. Rowland, Environ. Sci. Technol., 2011, 45, 2398–2405.
26. S. Bishnoi and G. T. Rochelle, Chem. Eng. Sci., 2000, 55, 5531–5543.
27. S. Ma'mun, H. F. Svendsen, K. A. Hoff and O. Juliussen, Energy Convers. Manage., 2007, 48, 251–258.
28. B. Mandal, M. Guha, A. Biswas and S. Bandyopadhyay, Chem. Eng. Sci., 2001, 56, 6217–6224.
29. Y.-C. Chang, R. B. Leron and M.-H. Li, J. Chem. Thermodyn., 2013, 64, 106–113.
30. S. Kadiwala, A. V. Rayer and A. Henni, Chem. Eng. J., 2012, 179, 262–271.
31. X. Luo, N. Chen, S. Liu, W. Rongwong, R. O. Idem, P. Tontiwachwuthikul and Z. Liang, Int. J. Greenhouse Gas Control, 2016, 53, 160–168.
32. T. Sema, A. Naami, K. Fu, G. Chen, Z. Liang, R. Idem and P. Tontiwachwuthikul, Chem. Eng. Sci., 2013, 100, 183–194.
33. B. Xu, H. Gao, X. Luo, H. Liao and Z. Liang, Int. J. Greenhouse Gas Control, 2016, 51, 11–17.
34. D. D. D. Pinto, J. G. M. S. Monteiro, B. Johnsen, H. F. Svendsen and H. Knuutila, Int. J. Greenhouse Gas Control, 2014, 25, 173–185.
35. C. Kim and D. Savage, Chem. Eng. Sci., 1987, 42, 1481–1487.
36. J. Benitez-Garcia, G. Ruiz-Ibanez, H. A. Al-Ghawas and O. C. Sandall, Chem. Eng. Sci., 1991, 46, 2927–2931.
37. J. Li, A. Henni and P. Tontiwachwuthikul, Ind. Eng. Chem. Res., 2007, 46, 4426–4434.
38. P. D. Vaidya and E. Y. Kenig, Chem. Eng. Technol., 2009, 32, 556–563.
39. P. Tontiwachwuthikul, A. G. Wee, R. Idem, K. Maneeintr, G.-j. Fan, A. Veawab, A. Henni, A. Aroonwilas and A. Chakma, US Pat., US7910078 B2, Grant Application number US 11/843,958, 2011.
40. T. Sema, A. Naami, R. Idem and P. Tontiwachwuthikul, Ind. Eng. Chem. Res., 2011, 50, 14008–14015.
41. A. Hartono, K. A. Hoff, T. Mejdell and H. F. Svendsen, Energy Procedia, 2011, 4, 179–186.
42. A. Hartono and H. F. Svendsen, Energy Procedia, 2009, 1, 853–859.
43. F. A. Chowdhury, H. Yamada, Y. Matsuzaki, K. Goto, T. Higashii and M. Onoda, Energy Procedia, 2014, 63, 572–579.
44. Y. Liang, H. Liu, W. Rongwong, Z. Liang, R. Idem and P. Tontiwachwuthikul, Fuel, 2015, 144, 121–129.
45. All physical properties listed in Table 1 were taken from MSDS (Material Safety Data Sheet), http://ScienceLab.com.
46. E. L. Cussler, Diffusion: mass transfer in fluid systems, Cambridge university press, 2009.
47. R. Taylor and R. Krishna, Multicomponent mass transfer, John Wiley & Sons, 1993.
48. O. Levenspiel, Ind. Eng. Chem. Res., 1999, 38, 4140–4143.
49. F. Khan, V. Krishnamoorthi and T. Mahmud, Chem. Eng. Res. Des., 2011, 89, 1600–1608.
50. M. Afkhamipour and M. Mofarahi, Int. J. Greenhouse Gas Control, 2013, 15, 186–199.
51. M. Afkhamipour and M. Mofarahi, Int. J. Greenhouse Gas Control, 2014, 25, 9–22.
52. T. N. G. Borhani, M. Afkhamipour, A. Azarpour, V. Akbari, S. H. Emadi and Z. A. Manan, J. Ind. Eng. Chem., 2016, 34, 344–355.
53. T. N. G. Borhani, V. Akbari, M. K. A. Hamid and Z. A. Manan, J. Ind. Eng. Chem., 2015, 22, 306–316.
54. J. Gabrielsen, H. F. Svendsen, M. L. Michelsen, E. H. Stenby and G. M. Kontogeorgis, Chem. Eng. Sci., 2007, 62, 2397–2413.
55. P. Mores, N. Scenna and S. Mussati, Int. J. Greenhouse Gas Control, 2012, 6, 21–36.
56. R. B. Bird, Appl. Mech. Rev., 2002, 55, R1–R4.
57. W. M. Deen, Analysis of transport phenomena (topics in chemical engineering), Oxford University Press, New York, 1998.
58. D. Demontigny, P. Tontiwachwuthikul and A. Chakma, Can. J. Chem. Eng., 2001, 79, 137–142.
59. A. Naami, M. Edali, T. Sema, R. Idem and P. Tontiwachwuthikul, Ind. Eng. Chem. Res., 2012, 51, 6470–6479.
60. K. Fu, W. Rongwong, Z. Liang, Y. Na, R. Idem and P. Tontiwachwuthikul, Chem. Eng. J., 2015, 260, 11–19.
61. K. Fu, T. Sema, Z. Liang, H. Liu, Y. Na, H. Shi, R. Idem and P. Tontiwachwuthikul, Ind. Eng. Chem. Res., 2012, 51, 12058–12064.
62. H. N. Abdul Halim, A. M. Shariff, L. S. Tan and M. A. Bustam, Ind. Eng. Chem. Res., 2015, 54, 1675–1680.
63. A. Aroonwilas and P. Tontiwachwuthikul, Ind. Eng. Chem. Res., 1998, 37, 569–575.
64. A. Aroonwilas, P. Tontiwachwuthikul and A. Chakma, Sep. Purif. Technol., 2001, 24, 403–411.
65. A. Aroonwilas and A. Veawab, Ind. Eng. Chem. Res., 2004, 43, 2228–2237.
66. A. Dey and A. Aroonwilas, Energy Procedia, 2009, 1, 211–215.
67. P. Tontiwachwuthikul, A. Meisen and C. J. Lim, Chem. Eng. Sci., 1992, 47, 381–390.
68. A. Aroonwilas and P. Tontiwachwuthikul, Sep. Purif. Technol., 1997, 12, 67–79.
69. A. Aroonwilas, A. Veawab and P. Tontiwachwuthikul, Ind. Eng. Chem. Res., 1999, 38, 2044–2050.
70. A. Setameteekul, A. Veawab and A. Aroonwilas, 2006 IEEE EIC Climate Change Conference, 2006.
71. S.-B. Jeon, H.-D. Lee, M.-K. Kang, J.-H. Cho, J.-B. Seo and K.-J. Oh, J. Taiwan Inst. Chem. Eng., 2013, 44, 1003–1009.
72. W. Li, X. Zhao, B. Liu and Z. Tang, Ind. Eng. Chem. Res., 2014, 53, 6185–6196.
73. M.-K. Kang, I.-D. Kim, B.-J. Kim, J.-S. Kang and K.-J. Oh, Ind. Eng. Chem. Res., 2015, 54, 5853–5861.
74. P. Usubharatana, A. Veawab, A. Aroonwilas and P. Tontiwachwuthikul, 2006 IEEE EIC Climate Change Conference, 2006.
75. J. Gao, J. Yin, F. Zhu, X. Chen, M. Tong, W. Kang, Y. Zhou and J. Lu, Sep. Purif. Technol., 2016, 163, 23–29.
76. K. Fu, G. Chen, T. Sema, X. Zhang, Z. Liang, R. Idem and P. Tontiwachwuthikul, Chem. Eng. Sci., 2013, 100, 195–202.
77. L. Wen, H. Liu, W. Rongwong, Z. Liang, K. Fu, R. Idem and P. Tontiwachwuthikul, Chem. Eng. Technol., 2015, 38, 1435–1443.
78. P.-C. Chen, M.-W. Yang, C.-H. Wei and S. Z. Lin, Int. J. Greenhouse Gas Control, 2016, 45, 245–252.
79. K. Maneeintr, R. O. Idem, P. Tontiwachwuthikul and A. G. Wee, Ind. Eng. Chem. Res., 2010, 49, 2857–2863.
80. H. Halim, A. Shariff and M. Bustam, Sep. Purif. Technol., 2015, 152, 87–93.
81. N. Hairul, A. Shariff and M. Bustam, Int. J. Greenhouse Gas Control, 2016, 49, 121–127.
82. M. Afkhamipour and M. Mofarahi, Int. J. Greenhouse Gas Control, 2016, 49, 24–33.
83. D. Bailey and P. Feron, Oil Gas Sci. Technol., 2005, 60, 461–474.
84. B. Xu, H. Gao, X. Luo, H. Liao and Z. Liang, Int. J. Greenhouse Gas Control, 2016, 51, 11–17.
85. Y. K. Salkuyeh and M. Mofarahi, Int. J. Energy Res., 2012, 36, 259–268.
86. Y. K. Salkuyeh and M. Mofarahi, Int. J. Energy Res., 2013, 37, 973–981.
87. A. Aroonwilas, A. Chakma, P. Tontiwachwuthikul and A. Veawab, Chem. Eng. Sci., 2003, 58, 4037–4053.
88. H. S. Fogler, Elements of chemical reaction engineering, 3rd edn, Prentice-Hall international, New Jersey, 1999.
89. J. E. Crooks and J. P. Donnellan, J. Chem. Soc., Perkin Trans. 2, 1989, 331–333.
90. C. Kale, A. Górak and H. Schoenmakers, Int. J. Greenhouse Gas Control, 2013, 17, 294–308.
91. J. A. Rocha, J. L. Bravo and J. R. Fair, Ind. Eng. Chem. Res., 1996, 35(5), 1660–1667.
92. P. Mores, N. Rodríguez, N. Scenna and S. Mussati, Int. J. Greenhouse Gas Control, 2012, 10, 148–163.
93. J. A. Kean, H. Turner and B. Price, Oil Gas J., 1991, 89, 41–47.
94. Y. Taitel, D. Bornea and A. Dukler, AIChE J., 1980, 26, 345–354.
95. R. Billet and M. Schultes, Chem. Eng. Res. Des., 1999, 77, 498–504.
96. J. Gualito, F. Cerino, J. Cardenas and J. Rocha, Ind. Eng. Chem. Res., 1997, 36, 1747–1757.
97. B. Hanley and C. C. Chen, AIChE J., 2012, 58, 132–152.
98. K. Onda, H. Takeuchi and Y. Okumoto, J. Chem. Eng. Jpn., 1968, 1, 56–62.
99. A. L. Kohl and F. Riesenfeld, Gas Purification, Gulf Publishing Company, Houston Texas, USA, 1985.
100. A. Naami, T. Sema, M. Edali, Z. Liang, R. Idem and P. Tontiwachwuthikul, Int. J. Greenhouse Gas Control, 2013, 19, 3–12.
101. X. Ji, W. Kritpiphat, A. Aboudheir and P. Tontiwachwuthikul, Can. J. Chem. Eng., 1999, 77, 69–73.

---