CO2 Solubility in Hybrid Solvents Containing 1-Butyl-3

Article pubs.acs.org/jced

CO2 Solubility in Hybrid Solvents Containing 1‑Butyl-3methylimidazolium Tetrafluoroborate and Mixtures of Alkanolamines Khalid Osman,*,† Deresh Ramjugernath,† and Christophe Coquelet‡ †

Thermodynamics Research Unit, School of Engineering, University of KwaZulu-Natal, Durban, South Africa MINES ParisTech, PSL Research University, CTP-Centre Thermodynamic of Processes 35, Rue Saint Honoré, 77305 Fontainebleau, France

‡

S Supporting Information *

ABSTRACT: To reduce the rate of climate change, feasible and energy-efficient solutions need to be found to capture CO2 at low pressure from flue gas emitted by various industries and energy sectors worldwide. The use of solvents to selectively absorb CO2 is a promising option for CO2 capture. This research investigated the solubility of CO2 in hybrid solvents containing the 1-butyl-3-methyl imidazolium tetrafluoroborate [bmim][BF4] ionic liquid with mixtures of up to three alkanolamine solvents, namely monoethanolamine (MEA), diethanolamine (DEA), and methyl-diethanolamine (MDEA). Gravimetric analysis was used to measure equilibrium CO2 solubility in the hybrid solvents containing various compositions of the above components at CO2 partial pressures of 0.05 MPa to 1.5 MPa and temperatures of 303.15 K to 323.15 K. CO2 solubility in these solvents was benchmarked against pure ionic liquids, as well as conventional alkanolamine solvents, and modeled using the Posey−Tapperson−Rochelle model for the alkanolamines present and the SRK equation of state for the ionic liquid present in the hybrid solvents. It was found that the hybrid solvents achieved significantly higher CO2 solubility at low pressure than pure ionic liquids and conventional alkanolamine solvents. Modeling, however, was found to be less accurate for hybrid systems than data modeled for pure ionic liquid systems.

■

INTRODUCTION Various industries, environmentalists, and the public at large, are concerned about the amount of carbon dioxide (CO2) being emitted by industrial processes due to the global warming effects of CO2 and its potential to accelerate climate change. In 2008, nearly 30000 Mega tons (Mt) of CO2 was emitted into the atmosphere,1 an increase of nearly 100 % compared to CO2 emissions in 1971. IPCC (2013)2 found that, from 1750 to 2011, 365 Gt of CO2 were emitted into the atmosphere by fossil fuel and cement industries. The CO2 concentration in the atmosphere at the beginning of 2013 was 391 ppm.2 An abundance of coal in many countries has resulted in these countries basing their infrastructure on electricity supplied by coal power plants. Moreover, the currently high cost of renewable energy due to the need of abundant energy storage implies that coal power plants will still dominate, or account for a large percentage, of a country’s electricity supply.1,3 Thus, a solution needs to be found to ensure that, while coal and other fossil fuel industries continue to dominate the energy sector, they may reduce their CO2 emissions. A promising midterm solution is carbon capture and storage (CCS), which involves capturing CO2 from flue gas emitted by various industries, transporting CO2, and the storage of CO2 through various methods including sequestration. 1 Well © XXXX American Chemical Society

developed and feasible techniques of CO2 compression, transportation, and sequestration exists currently.4 However, capturing CO 2 from flue gas still requires significant optimization. The process of CO2 absorption using solvents was found to have many advantages over other CO2 capture techniques. The main distinguishing advantages of the technique included the high level of development associated with industrial gas absorption and the high potential for feasible CO2 capture due to the number of possible solvents that may prove efficient for continuous industrial CO2 absorption and desorption.5 The most popular solvents studied for CO2 capture are alkanolamines and carbonate-based solvents. These solvents are chemical solvents which result in high CO2 absorption rate and capacity. However, they are also highly corrosive and can only be used if diluted with water.6 Water increases the heat capacity of the solvent, resulting in high amounts of energy being required for desorption. It was determined by Lecomte et al.7 that to conduct CO2 capture with a conventional 100 w = 30 monoethanolamine solvent with 50 % solvent regeneration Received: March 24, 2015 Accepted: June 18, 2015

A

DOI: 10.1021/acs.jced.5b00273 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

alkanolamines with ionic liquids on CO2 solubility instead of benchmarking pure ionic liquids. Equilibrium CO2 solubility measurements were undertaken using gravimetric analysis. Various solvent compositions were considered and these are provided in Table 1.

would incur an energy penalty of 3.75 GJ/ton CO2. Alternative solvents need to be found that can either replace alkanolamine solvents, or be blended with alkanolamines to reduce desorption energy costs. Solvents that offer higher CO2 solubility would result in less solvent being necessary for CO2 capture and regeneration. Other prospective solvents for CO2 capture include ionic liquids which are neutral liquids composed entirely of cations and anions. They are physical solvents which absorb CO2 through a rearrangement of solvent molecules to accommodate CO2 molecules. There has been growing interest in the use of ionic liquids, either in their pure state or blended with other solvents such as alkanolamines, to capture CO2.5,8 Previous studies have indicated that ionic liquids containing imidazolium cations are more selective to CO2 absorption over most flue gas components.9 This increased selectivity is beneficial for the recovery and ultimate storage of CO2. Also, studies have indicated that ionic liquids with fluorinated anions achieve significantly higher CO2 absorption than nonfluorinated ionic liquids.5 Nevertheless, compared to chemical solvents, ionic liquids achieve lower CO2 absorption capacity at low pressure.10 CO2 is only absorbed physically, requiring high flue gas pressure to achieve efficient CO2 absorption. This study focused on combining an ionic liquid with conventional alkanolamines to create hybrid solvents in an attempt to improve CO2 solubility at low pressure. Research in mixing ionic liquids with alkanolamines has received limited attention thus far. Zhang et al.11 modified the cation synthesis of ionic liquids to include alkanolamines in the cation itself. The study did not achieve CO2 absorption superior to that of pure imidazolium-based fluorinated ionic liquids. Limited studies were conducted on MDEA + ionic liquid + H2O mixtures by Ma et al.12 Although loading results were promising, overall absorption was low due to high H2O dilution. A corrosion study of BF4 + MDEA hybrid solvents13 achieved very encouraging results showing low solvent corrosiveness toward various steels.5 Camper et al.14 measured CO2 absorption in hybrid solvents containing MEA and DEA with undisclosed RTILs. The results indicated superior absorption over amine functionalized ionic liquids. Although the above results were promising overall, the study of hybrid solvents containing alkanolamines and ionic liquids was found to be very limited in comparison to pure chemical and physical solvents, as well as other hybrid solvents such as alkanolamine−methanol mixtures, which proved to be too volatile for solvent regeneration and recyclable use.15 In this research, we aimed to investigate the use of hybrid solvents further. The alkanolamines used in this study were monoethanolamine (MEA), diethanolamine (DEA), and methyl diethanolamine (MDEA), and the ionic liquid tested was [Bmim][BF4]. Instead of diluting the alkanolamines with H2O as is conventionally done, this combination intended to use the ionic liquid to replace H2O in conventional alkanolamine solvents. The ionic liquid served not only as a diluent for the corrosive alkanolamines but also as an absorbent of CO2. The ionic liquid [Bmim][BF4] was chosen due to its availability and relatively low cost in comparison to other ionic liquids. It was also found by Osman5 to be of low corrosiveness. Greater fluorination of the anion or greater cation chain lengths have been reported to achieve superior CO2 absorption to that of [Bmim][BF4], but the focus of this study was to investigate the effects of blending

Table 1. Mass Composition (100 w) of Hybrid Samples Measured for CO2 Solubilitya mass composition (100 w) sample

MEA

1 2 3 4 5 6

29.3 33.0 31.8 31.6 30.3 29.8

DEA

MDEA

[Bmim][BF4]

10.4 21.7 12.8

70.7 50.8 56.1 58.0 48.0 45.7

16.2 12.1

11.7

Standard uncertainty: u(m) = ± 0.1 mg; combined uncertainty: u(x) = 100 w = ± 0.6. a

The alkanolamines were selected in order to determine the effect of combining primary, secondary, and tertiary amines with the ionic liquid. MEA is a primary amine, DEA is a secondary amine, and MDEA is a tertiary amine. Moreover, the above alkanolamines were well studied for CO2 absorption.10,16,17 The [Bmim][BF4] ionic liquid was chosen for its CO2-selective imidazolium cation and fluorinated anion, which increased CO2 absorption.5,18 The compositions in Table 1 were chosen to investigate the effects of different alkanolamine−ionic liquid compositions on CO2 absorption. Conventional solvents utilize MEA at 100 w = 30, with 100 w = 70 water.10,16 The advantage of MEA is that it is a primary amine which reacts with CO2 at a comparatively high absorption rate, with a reasonably high CO2 absorption capacity.6 It was thus decided that all hybrid solvents studied must include 100 w = 30 MEA. Because of the high corrosiveness of MEA,10 the composition of MEA in the solvent was limited to 100 w = 30. Varying compositions of the secondary and tertiary amine were investigated. Although secondary and tertiary amines have been reported to achieve lower CO2 absorption rates, they were also reported to achieve higher CO2 absorption capacities than primary amines.16 These amines were also reported to be less corrosive than primary amines and could hence be used to increase the total amine composition in the solvent to up to 100 w = 50.16,17 Their addition to the solvent at high compositions, however, was also reported to increase the viscosity of the solvent and inhibit CO2 diffusion.19,20 Their inclusion in the solvent was thus limited. Addition of DEA alone to the ionic liquid−MEA mixture was first investigated up to 100 w = 16.2. Thereafter addition of MDEA alone was investigated, followed by a combination of DEA and MDEA. The equilibrium CO2 solubility results were modeled using the Redlich−Kwong equation of state for the ionic liquid and Posey−Tapperson−Rochelle model for the alkanolamine components.

■

MATERIALS AND EXPERIMENTAL PROCEDURE Materials. Carbon dioxide (CO2) and nitrogen (N2) gas were purchased from Afrox Ltd. (South Africa) with a stated minimum purity of 99.9 % by volume. CO2 absorption was measured, while all solvents were also tested for absorption with N2, in order to provide the correction for buoyancy effects. B



Article

Table 2. Densities, ρ, and Refractive Index, n, of Chemicals Together with Literature Values at Experimental Temperature and Atmospheric Pressurea ρ (g·cm−3)

n at 293.15 K T chemical

exp

lit. d

[Bmim][BF4]

1.42423

1.42475

MEA

1.454128

1.4539e

DEA

1.474742

1.4776e

MDEA

1.46956

1.46935f

K

exp

lit.

303.15 313.15 323.15 303.15 313.15 323.15 303.15 313.15 323.15 303.15 313.15 323.15

1.2038 1.1964 1.1893 1.0077 0.9997 0.9917 1.0905 1.0842 1.0777 1.0306 1.0248 1.0176

1.2039b 1.1963b 1.1887b 1.0085 1.0009 0.9931 1.0847c 1.0774c 1.0250c 1.0174c

Standard uncertainties: err(T) = 0.03 K, err(p) = 0.04 MPa; combined expanded uncertainty: Uc(ρ) = 0.0003 g·cm−3, Uc(n) = 0.0005. bFrom ref 19. cFrom ref 20. dFrom ref 21. eFrom ref 22. fFrom ref 23.

a

Figure 1. Diagram of intelligent gravimetric analyzer (IGA-01) for gas solubility measurements.

1-Butyl-3-methyl imidazolium tetrafluoroborate [Bmim][BF4], monoethanolamine, diethanolamine, and methyl diethanolamine were purchased from DLD Scientific Ltd. The purity of [Bmim][BF4] was stated by the supplier to be ≥ 98 % while the purity of all alkanolamines was stated to be ≥ 99 %. The supplier’s method of testing purity was stated to be H

NMR. The purity of all chemicals was tested and confirmed using an Atago RX-7000 CX refractometer. The density of [Bmim][BF4] and all alkanolamines were tested using an Anton Paar DMA 5000 M vibrating U-tube densitometer that was calibrated using Ultrapure water obtained using an Elga Purelab Option-Q Millipore Device and dried air. C



Article

probe with a resolution of ± 0.01 K was located inside the reactor next to where the sample holder is suspended to read temperature. Temperature control accuracy was to 0.05 % of desired conditions. For this research, the apparatus conducted measurements at a system pressure in the range of 0.05 MPa to 1.5 MPa. Isotherms of 303.15 K, 313.15 K, and 323.15 K were considered because it has been noted by various literature5,18 that high temperature absorption is not feasible due to absorption being an exothermic process. Moreover, the temperature for the gravimetric measurements was limited to 323.15 K to avoid losses of the alkanolamines by vaporization during low pressure measurements. Because of the small sample size, each equilibrium point took up to 5 h to reach equilibrium. The apparatus was connected to a personal computer, and all measurements including temperature, pressure, and sample weight were recorded using specialized software suited for the apparatus. Because of significant changes in solvent density with pressure, a buoyancy correction was applied, as with all gravimetric measurements on systems containing liquid.24−27 Nitrogen gas was used for the buoyancy correction and the method of buoyancy correction is provided as Supporting Information. Further information is provided in the work of Macedonia et al.26 and Shiflett et al.27 Analysis was made by focusing on absorption measurements. However, desorption measurements were also conducted to determine if any hysteresis existed and to provide an insight regarding the possibility of pressure swing absorption and desorption for the systems studied in this work. Correlation of Solubility Data. The modeling of CO2 solubility in pure ionic liquids and in conventional alkanolamine solvents was found to be abundant in literature. However, no modeling was attempted for CO2 in water-free hybrid solvents containing ionic liquids and alkanolamines. A combination of two simple models was attempted in this work in order to model CO2 absorption in the hybrid solvents. The solubility of CO2 in pure [Bmim][BF4] alone, measured in a previous work,18 was used to estimate the amount of CO2 absorbed specifically by the ionic liquid component of the hybrid solvent. CO2 solubility in the remaining alkanolamine components of the solvent were then modeled separately. In this manner, rather than treating the systems studied as ternary or quaternary systems, the systems were treated as an ideal combination of a binary system of CO2− ionic liquid and a pseudobinary system of CO2−alkanolamines and modeled. The modeling was thereafter combined. Gas absorption in ionic liquids has been modeled by numerous sources in literature utilizing various models. Osman5 provides a detailed analysis of models used by various literature sources appropriate to various systems. The absorption data of CO2 in the ionic liquid component of systems containing hybrid solvents were modeled using a generic nonelectrolyte Redlich−Kwong equation of state (RK-EOS). The model was successfully applied in the work of Osman5 and Bahadur et al.18 for CO2 and O2 absorption in pure [Bmim][BF4] and other ionic liquids. The model was easy to program, requiring very low computation time, and yet provided good correlation to within 0.3 % of measured partial pressures. Shiflett et al.27 also applied the RK-EOS for CO2 absorption in [Bmim][BF4] and [Bmim][PF6] ionic liquids due to the presence of nonvolatile components which contained no known critical point conditions. Systems containing electrolytes

Measured densities and refractive indices of all chemicals are provided in Table 2, along with previous measurements in literature.19−23 Water content of all chemicals was also tested using a Metrohm 702 SM Titrino meter and was found to be 600 ppm for [Bmim][BF4], 400 ppm for monoethanolamine, and 500 ppm for diethanolamine and methyl-diethanolamine. Sample Preparation. First, 100 mg of solvent was needed for solubility measurement by gravimetric analysis. However, it was impossible to accurately create the hybrid solvents at precise composition at such low quantity. Thus, 1 g of hybrid solvent solutions were prepared by individually measuring appropriate masses for each component using a Mettler balance with a resolution of ±0.001 g and combining each component in a sample vial. The 1 g preparations ensured accurate measurement of each component and effective mixing and reduced the error caused by inherent impurities of the chemicals. High solvent masses in the vial also ensured negligible effects of contact with the air. The accuracy in mass measurement was ± 1 mg for each component. This resulted in a combined uncertainty of up to ± 4 mg in sample mass depending on the number of components in the hybrid sample. Because of impurities stated by the suppliers of all chemicals, the combined uncertainty in mass composition was 100 w = 0.6. In all cases, the impurity was stated to be water and removed by the gravimetric analyzer upon degassing. Because of the volatility of amines in the solvent, particularly MEA, extra MEA was initially blended into the solvent and degassing was conducted at 293.15 K. Losses of sample were observed to be minimal for most samples. For measurements where significant losses were observed (higher than known water content), the losses were measured in real-time and assumed to be that of MEA, the most volatile amine. The composition of MEA in the sample was thus adjusted. Table 1 contains the final compositions of each solvent after degassing. Solubility Measurements. All gas solubility measurements for this research were conducted by gravimetric analysis using an Intelligent Gravimetric Analyzer (IGA-01) designed and constructed by Hiden Analytical Ltd. A description of the apparatus is also available in previous works including Osman (2014)5 and Bahadur et al. (2014).18 Figure 1 shows the setup of the gravimetric analyzer. The IGA consists of a small sample reactor cell into which hybrid solvents were placed. The cell is very small, allowing for small volumes of material to be studied at a time. Samples approximately 100 mg in mass were loaded. The sample holder was suspended by tungsten and gold wires inside a stainless steel reactor. The wires were attached to a microbalance which has a resolution of ± 0.1 μg. The weight of the sample holder was countered using a counterweight so that only the weight of the sample may be tracked. The reactor was sealed using copper gaskets. The samples were first degassed at 293.15 K using a Vacuubrand GMH-MD1 vacuum pump and an Edwards WRGS-NW35 turbomolecular pump to achieve vacuum below 100 Pa. Thereafter, CO2 gas was injected into the reactor and the reactor pressure was controlled using two precalibrated pressure controllers. One pressure controller controlled flow out using the vacuum pumps, while the other controlled gas flow and pressure into the reactor. The uncertainty in pressure reading was ± 1.25· 10−4 MPa. The control accuracy of the two pressure controllers was stated to be 0.025 % of desired conditions. Reactor temperature was controlled using a Polyscience SD07R-20 refrigerated recirculating water bath which was adequate for the temperature range in this work. A Pt100 D



Article

Table 3. Properties of [Bmim][BF4] Obtained by Lydersen−Joback−Reid Group Contribution Method31,32

[Bmim][BF4]

molar mass

Tb

TC

PC

VC

g·mol−1

K

K

bar

cm3·mol−1

226.02

391.0

523.25

18.88

645.61

a(T ) RT − V−b V (V + b)

(1)

where a(T) is a function of species mole fraction and temperature. The van der Waals−Berthelot mixing formula was used in this work, as successfully applied for similar conditions.30 N

∑

a(T ) =

i,j=1

τij ⎞ ⎛ a ia j ⎜1 + ⎟(1 − kij)x ixj ⎝ T⎠

(2)

where kij =

lijl ji(xi + xj) l jixi + lijxj

; ai = 0.427480

R2Tci2 αi(T ) Pci

and where k ⎛ TC T⎞ β − ⎜ ⎟ ∑ k TC ⎠ ⎝T k=0 ≤3

αi(T ) =

and b=

1 2

N

∑ (bi + bj)(1 − mij)(1 − kij)xixj i,j=1

where bi = 0.08664

RTci Pci

AmineH+ + HCO−3 ↔ Amine + CO2 (aq) + H 2O

Carbonate (CO3 ) and hydroxide (OH ) ion concentration were assumed to be negligible. The assumed presence of H2O contributes to inaccuracy in correlation because CO2 diffusion occurs in the ionic liquid instead of water. This represents a limitation in the model, and this model may serve as merely an empirical model rather than a theoretical model for the systems studied. Further analysis is thus imperative to ascertain the chemistry between the CO2−ionic liquid−alkanolamine systems in order to achieve more accurate modeling. The equilibrium constant for the above reaction is given by Posey et al.34 and Dicko et al.:35

Tb AM + BM ∑ nΔTM − (∑ nΔTM)2

(3)

where Tb = 198.2 + ∑ nΔTbM PC =

M [CM + ∑ nΔPM]2

(5)

−

2−

mij = mji and mii = 0 and lij, lji, τij, and mij are binary interaction parameters. Coefficients βk are adjustable fitting parameters. lij, lji, τij, mij, and βk are obtained by regression of measured data P-T-x data. Because of [Bmim][BF4] decomposition before reaching its critical point, critical temperature and pressure for the ionic liquid could not be found.26 Critical temperature and pressure were thus computed using a modified Lydersen−Joback−Reid group contribution method:31,32 TC =

0.6234

properties obtained using the above contribution method are provided in Table 3 for [Bmim][BF4]. CO2 absorption in the alkanolamine components of each hybrid solvent was modeled using the Posey−Tapperson− Rochelle model, as was successfully applied in the work of Osman et al.33 This model was developed by Posey et al.34 It is a simple empirical model, which assumes the entire reaction mechanism between CO2 and all alkanolamines to be a single absorption reaction. Posey et al.34 tested the model accuracy on systems involving a gas mixture of CO2 and H2S in single alkanolamine solvents of MDEA:H2O and DEA:H2O at various concentrations. Dicko et al.35 confirmed the model to be relatively accurate for systems involving MDEA at concentrations of up to 100 w = 50 despite its simplicity. Osman36 utilized the model for CO2 absorption in blends of H2O, DEA, and MDEA. Inaccuracies of as low as 0.01 % of the measured data were achieved with very low computation time. More complex models, known to provide a more accurate correlation for alkanolamine systems, were invalid for the system in this research due to the absence of H2O. Complex models assume reaction mechanisms between H2O, CO2, and the alkanolamines and therefore could not be used for this research. The extent of CO2 absorption in the hybrid solvent was measured and the extent of CO2 absorption in the ionic liquid component of the hybrid solvent could be known by measurements conducted in a previous work.5 Thus, the extent of absorption in the alkanolamines collectively in the hybrid solvent could be estimated. However, CO2 absorption in each alkanolamine component in the hybrid solvent could not be determined by gravimetric analysis alone. The only option with such circumstances was to assume a single reversible reaction and model the CO2 solubility data based on this simplification. Model computation was easy to apply to any P-T-x data. The following reaction was assumed for CO2 with all alkanolamines in the hybrid solvent.

such as the above systems have also been reported in literature to be successfully modeled using ordinary equations of state.28−30 In addition to good estimation and model simplicity, the RKEOS was chosen because it was also proven to be successful at temperature and pressure conditions similar to that measured in this research.27 The RK-EOS is given by the following:27 P=

ω

Ln(K CO2) = a +

b + cLT C O Amine + d(LT C O Amine)0.5 T (6)

(4)

with COAMINE = Amine concentration neglecting the presence of acid gases.

where AM = 0.5703, BM = 1.0121, CM = 0.2573, and EM = 6.75.31 n is the number of occurrences of any particular functional group in the molecule. Group contributions for ΔTbM, ΔTM, and ΔPM are provided in the work of Alvarez and Valderrama.32 Critical

= E

Amine Amine + Ionic Liquid

(7) DOI: 10.1021/acs.jced.5b00273 J. Chem. Eng. Data XXXX, XXX, XXX−XXX


Article

Thereafter, PCO2 was calculated using the following formula: PCO2 = XCO2K CO2

LT (1 − LT )

(8)

LT is CO2 loading in mol [CO2/mol alkanolamine]. Parameter “a” in eq 6 represents an overall correction factor, “b” represents temperature correction, and “c” and “d” account for amine concentration in the solvent. Parameters a to d were found by regression of measured data. CO2 partial pressure was calculated and compared to measured results. All modeling was conducted using Matlab R2011b.

■

RESULTS AND DISCUSSION OF MEASUREMENTS The results of the solubility measurements are presented in Figure 2a to c. CO2 solubility in pure [Bmim][BF4] and pure and [Bmim][Tf2N] ionic liquids, measured in a previous work,18 as well as CO2 solubility in a conventional chemical solvent containing MEA:H2O at 100 w = 30:7037 was also plotted in Figure 2a to c in order to provide a comparison between pure and hybrid solvents. The actual P-T-x data are available as Supporting Information. An observation found in all figures, and consistent with literature, was that CO2 absorption increased with increasing CO2 partial pressure. This is evident in Figure 2a to c above for absorption in all solvents and is consistent with physical and chemical equilibrium.31 For chemical absorption in alkanolamines, increasing pressure favored the primary, secondary, and tertiary reaction mechanism between CO2 and alkanolamines, thereby increasing CO2 absorption. In the case of the ionic liquid in the solvent, increasing pressure favored the diffusion and absorption of CO2 into the liquid phase. A further observation was that while CO2 absorption increased quite significantly with increasing pressure in the case of pure ionic liquids as measured in Bahadur et al.,18 there was a more gradual increase in the case of CO2 absorption in alkanolamine− ionic liquid hybrid solvents studied in this work. Figure 2a to c show that for absorption in pure [Bmim][BF4] and [Bmim][Tf2N], a high increase in CO2 partial pressure resulted in a high increase in CO2 absorption. In Figure 2a, the equilibrium CO2 mole fraction increased from 0.023 to 0.283 in the case of [Bmim][Tf2N] and from 0.027 to 0.183 in the case of [Bmim][BF4] from 0.05 MPa to 1.5 MPa at 303.15 K. By contrast, for hybrid solvents containing approximately 60 % ionic liquid by mass, equilibrium CO2 mole fraction increased only by 0.064 across the same pressure range. The limited CO2 absorption capacity of alkanolamine components of the hybrid solvents is the likely reason for this observation. With alkanolamine solvents, CO2 reacts with the alkanolamine, and once all the alkanolamine is converted to ensure equilibrium, there is not going to be any further absorption regardless of increased pressure. This was also observed in the works of Osman36 and Osman et al.33 Ionic liquids on the other hand are physical solvents. Absorption merely occurs through a rearrangement of solvent molecules to accommodate the solute. Pressure thus has a significant effect on the absorption of gases into the ionic liquid solvent. Higher amounts of alkanolamines in the hybrid composition resulted in increased pressure having a lower effect on CO2 absorption. This is illustrated in Figure 2b. CO2 mole fraction in the sample containing MEA:[Bmim][BF4] at 100 w = 29.3:70.7 at 303.15 K increased by 0.1 when comparing results at 0.05 MPa

Figure 2. Absorption of CO2 in [Bmim][BF4]:alkanolamine hybrid solvents at (a) 303.15 K; (b) 313.15 K; (c) 323.15 K. ●, MEA: [Bmim][BF4] at 100 w = 29.3:70.7; □, MEA:MDEA:[Bmim][BF4] at 100 w = 31.6:10.4:58; ■, MEA:MDEA:[Bmim][BF4] at 100 w = 30.3:21.8:48; △, MEA:DEA:[Bmim][BF4] at 100 w = 31.8:12.1:56.1; ▲, MEA:DEA:[Bmim][BF4] at 100 w = 33:16.2:50.8; ◆, MEA:DEA:MDEA:[Bmim][BF4] at 100 w = 29.8:11.7:12.8:45.7; +, pure [Bmim][BF4];18 ×, Pure [Bmim][Tf2N];18 ○, MEA:H2O at 100 w = 30:70.37

and 1.5 MPa, a difference significantly higher than for hybrid solvents containing alkanolamines at 100 w = 40 composition. All hybrid solvents achieved higher CO2 solubility at low pressure than the pure ionic liquids, as well as the conventional alkanolamine solvent of 100 w = 30 MEA in H2O.10,16 This was a very encouraging observation and is shown in Figure 2a to c at all isotherms, with Figure 2b specifically showing superior CO2 absorption at 313.15 K compared to the conventional alkanolamine solvent. This result has significant industrial implications. F



Article

Table 4. Enthalpy and Entropy of Absorption for all CO2 in All Hybrid Solvent Samples Studied in This Work Δh

standard deviation

ΔS

standard deviation

solvent

kJ·mol−1

kJ·mol−1

J·mol−1·K−1

kJ·mol−1

[Bmim][BF4] + MEA at 100 w = 70.7:29.3 MEA + DEA + [Bmim][BF4] at 100 w = 33:16.2:50.8 MEA + DEA + [Bmim][BF4] at 100 w = 31.8:12.1:56.1 MEA + MDEA + [Bmim][BF4] at 100 w = 31.6:10.4:58 MEA + MDEA + [Bmim][BF4] at 100 w = 30.3:21.8:48 MEA + DEA + MDEA + [Bmim][BF4] at 100 w = 29.8:11.7:12.8:45.7 [Bmim][BF4]36 [Bmim][PF6]36

−14.9 −3.9 −5.4 −9.9 −15.9 −5.1 −15.9 −16.1

0.4 0.4 1.1 0.6 4.5 1.4

−48.2 −12.4 −17.1 −31.6 −50.9 −16.4 −52.4 −53.2

1.8 1.4 3.6 1.8 14.4 4.5

In the energy sector, pulverized coal power plant flue gas is available at 0.1 MPa to 0.17 MPa.38 CO2 is available at pressure below atmospheric pressure. It is thus beneficial if the solvent in the absorption process would be able to significantly absorb CO2 at low pressure, and this is what the hybrid solvents achieved in this research. The hybrid solvents also achieved higher CO2 absorption at low pressure than other pure liquids studied in literature, including the [MOA][Tf2N] ionic liquid and [Bmim][MeSO4] ionic liquid studied in the work of Bahadur et al.18 At all isotherms, all hybrid solvents achieved higher CO2 absorption than [MOA][Tf2N] for pressures up to 0.4 MPa. The hybrid solvent containing MEA:DEA:[Bmim][BF4] at 100 w = 31.8:12.1:56.1 achieved higher CO2 absorption for pressures up to 1 MPa. For absorption at pressure up to 1 MPa, hybrid solvents containing [Bmim][BF4] were found to be beneficial over conventional alkanolamine solvents and pure ionic liquids. Diluting the alkanolamines with [Bmim][BF4] instead of water benefited the CO2 solubility achieved. Diluting alkanolamines such as MEA with water reduced the overall corrosiveness of the solvent and facilitated diffusion of CO2 into the solvent. However, as this study revealed, diluting alkanolamines with noncorrosive ionic liquids instead of water not only reduced the overall corrosiveness of the solvent and facilitated diffusion but also increased CO2 absorption because the ionic liquid also absorbed CO2. It was found that the hybrid solvent containing MEA:DEA: [Bmim][BF4] at 100 w = 31.8:12.1:56.1 achieved the highest equilibrium CO2 solubility of all hybrid solvents across all isotherms. However, the sample containing MEA:DEA:[Bmim][BF4] at 100 w = 33:16.2:50.8 achieved the lowest CO2 absorption from all hybrid solvents studied in this work. This shows that while the addition of a secondary amine such as DEA in compositions of up to 100 w = 10 increased CO2 absorption, further increases in DEA compositions achieved lower CO2 absorption. This is likely due to the high viscosity of DEA which impedes diffusion. Density and viscosity measurements were conducted by DiGuillo et al.39 for many alkanolamines, including those which are studied in this work. At 303.15 K, DEA viscosity was measured to be 356 cP, while that of MEA and MDEA was 14.86 cP and 57 cP, respectively. The effect of alkanolamine composition explained above for solvents containing DEA was also observed for hybrid solvents containing MDEA. At 303.15 K, Figure 2a shows that the second highest CO2 solubility was achieved in the sample containing MEA:MDEA:[Bmim][BF4] at 100 w = 31.6:10.4:58. However, an increase in MDEA composition resulted in lower absorption for the sample containing MEA:MDEA:[Bmim][BF4] at 100 w = 30.3:21.8:48. The effect was more pronounced at higher

temperatures of 313.15 K and 323.15 K, as shown in parts b and c of Figure 2, respectively. In the case of samples containing MDEA, higher temperatures significantly reduce the CO2 absorption capacity of samples containing MDEA. The reaction mechanism between MDEA and CO2 is different to primary and secondary amines. Moreover, in the case of systems containing only alkanolamines and ionic liquids as studied in this work, with no presence of water, the tertiary reaction mechanism does not occur because MDEA needs to react directly with water. Details of reaction mechanisms between CO2 and alkanolamines are presented in Osman.5 The lower CO2 absorption achieved may have occurred due to MDEA absorbing CO2 only physically rather than chemically. A comparison between parts a to c of Figure 2 confirms that absorption decreased with increasing temperature. This is consistent in literature for both chemical and physical solvents because CO2 absorption is an exothermic reaction. It was also noted that temperature had varying effects on different samples depending on their composition of ionic liquid and alkanolamines. Samples containing higher amounts of MDEA achieved much lower CO2 absorption when increasing the temperature from 303.15 K to 323.15 K. CO2 mole fraction in the sample containing MEA:MDEA:[Bmim][BF4] at 100 w = 30.3:21.8:48 achieved the second highest CO2 absorption at 303.15 K, yet by 323.15 K, it achieved the third lowest CO2 absorption of all hybrid solvents studied. A significant reduction in CO 2 absorption was noted even in the sample containing a low composition of MDEA such as MEA:MDEA:[Bmim][BF4] at 100 w = 31.6:10.4:58. At 303.15 K, this sample achieved CO2 absorption almost as high as the sample containing MEA:DEA: [Bmim][BF4] at 100 w = 31.8:12.1:56.1. Yet at 313.15 K and 323.15 K, a significant difference is noted in absorption achieved between these two hybrid solvents. To further analyze the temperature dependence of each system, enthalpy and entropy of absorption was also calculated for each system. Enthalpy of absorption is given by the following equation:40 ⎛ ∂ ln x 2 ⎞ L ⎟ Δh2̅ [kJ·mol−1] = h2̅ − h2G = −R ⎜ ⎝ ∂ 1/T ⎠ P −1

(9)

−1

where R is the gas constant in [J mol ·K ], T is the system temperature in [K], h̅L2 is the enthalpy of the gas absorbed in the liquid, hG2 is the enthalpy of the pure gas in the gas phase, and x2 is the liquid mole fraction of the absorbed gas. The entropy of absorption is given by the following equation:40 ⎛ ∂ ln x 2 ⎞ ⎟ Δ s2̅ [J·mol−1·K−1] = s 2̅ L − s2G = R ⎜ ⎝ ∂ ln T ⎠ P G

(10)



Article

where sL2̅ is the entropy of the gas absorbed in the liquid, sG2 is the entropy of the pure gas in the gas phase, and x2 is the liquid mole fraction of the absorbed gas. Enthalpy and entropy of absorption for all systems in this work and are presented in Table 4. Solvents containing approximately 100 w = 20 DEA had the lowest enthalpy and entropy of absorption and also the lowest deviation of these values. While this may indicate relatively lower temperature dependence than other systems, actual CO2 absorption was low anyway, as shown in Figure 2a to c. Low standard deviations suggest little effect of the chemical solvent in the hybrid solvent mixture. This is possibly due to low diffusion of CO2 in DEA due to high DEA viscosity. Another observation was that enthalpies and entropies of absorption measured for systems containing MDEA had the highest standard deviation as shown in Table 4. This indicated a varying effect on absorption due to the alternative, tertiary reaction mechanism that occurs between MDEA and CO2.41,42 Enthalpy and entropy of absorption was intermediate for systems containing only the primary alkanolamine MEA combined with [Bmim][BF4]. This may suggest intermediate temperature dependence and high ordering, possibly due to the high reactivity of MEA in comparison to MDEA and DEA. The enthalpy and entropy of absorption of all hybrid solvents was lower than that of pure [Bmim][BF4] and pure [Bmim][PF6] measured by Cadena et al.41 as well as conventional MEA solvent at 100 w = 30 composition measured by Kim and Svendsen.43 Moreover, the deviations exhibited for the hybrid solvents were much lower than for CO2 absorption in the conventional 100 w = 30 MEA solvent, which was shown to be highly dependent on the CO2 loading in the solvent.43 Absolute values for enthalpy of absorption were low in the case of hybrid systems in comparison to pure ionic liquids as shown in Table 4 and significantly lower in comparison with alkanolamine−H2O systems.16 This is an interesting result and is suggestive of a reaction mechanism that differs greatly from that of conventional CO2−alkanolamine−H2O systems. It is also suggestive of a lower CO2 absorption rate compared to CO2− alkanolamine−H2O systems. A kinetic analysis would be required to further investigate this. The most encouraging conclusion drawn from this study of CO2 absorption in hybrid solvents is that hybrid solvents containing an ionic liquid and alkanolamines achieved significantly superior CO2 absorption at low pressure than pure physical solvents such as pure ionic liquids and conventional alkanolamine solvents diluted with H2O such as MEA:H2O at 100 w = 30:70. The industrial implication of this result is that flue gas from an industrial source would not need to be highly compressed to increase the pressure and facilitate efficient CO2 capture. CO2 capture can occur at lower pressures from atmospheric to 0.5 MPa, rather than much higher pressures of 1.5 MPa. Ultimately compression costs will be lower. High equilibrium CO2 mole fractions achieved at low pressure also implies that if compression costs are not high, then the flue gas may be compressed while less solvent may be required and the CO2 capture process may be scaled down and require less solvent regeneration and recirculation energy. Further investigation is required to quantify this. CO2 absorption measurements were the key focus of study in this research. However, CO2 desorption was also investigated for each hybrid solvent using gravimetric analysis assess the repeatability and hence validity of equilibrium solubility data. Desorption data are available as Supporting Information in

Tables S1 to S6. The desorption data generally showed fairly good repeatability. However, for solvents containing MDEA, CO2 equilibrium desorption data was consistently reported to be 1 % to 3 % higher than absorption data, indicating that solvents containing MDEA may not be completely recyclable. Further investigation is however required for this.

■

RESULTS AND DISCUSSION OF SOLUBILITY DATA CORRELATION The accurate modeling of CO2 absorption in alkanolamine− [Bmim][BF4] hybrid solvents proved to be challenging. Because of the lack of information on the chemistry of these hybrid systems and their reaction mechanism, simpler models were utilized to model the data. The RK-EOS modeled CO2 partial pressure in the [Bmim][BF4] component of the hybrid solvent, while the Posey−Tapperson−Rochelle modeled CO2 partial pressure in the alkanolamine components of the hybrid solvent. Figures 3 to 8 provide the experimentally measured absorption data, together with model estimates for all systems studied in this work at all isotherms (303.15 K to 323.15 K).

Figure 3. Experimental results together with Posey−Tapperson− Rochelle and RK-EOS estimates for the system of CO2 in [Bmim][BF4]:MEA at 100 w = 70.7:29.3. ■, 303.15 K; ▲, 313.15 K; ◆, 323.15 K. *Dotted lines indicate model estimates.

Regressed RK-EOS parameters and Posey−Tapperson− Rochelle parameters for the systems above are presented in Table S-8 in the Supporting Information, and estimated numerical data are available alongside measured data in Tables S-2 to S-7 of the Supporting Information provided. Because the systems were treated as an ideal combination of a binary system

Figure 4. Experimental results together with Posey−Tapperson− Rochelle and RK-EOS estimates for the system of CO2 in MEA:DEA: [Bmim][BF4] at 100 w = 33:16.2:50.8. ■, 303.15 K; ▲, 313.15 K; ◆, 323.15 K. *Dotted lines indicate model estimates. H



Article

Figure 8. Experimental results together with Posey−Tapperson− Rochelle and RK-EOS estimates for the system of CO 2 in MEA:DEA:MDEA:[Bmim][BF4] at 100 w = 29.8:11.7:12.8:45.7. ■, 303.15 K; ▲, 313.15 K; ◆, 323.15 K. *Dotted lines indicate model estimates.

Figure 5. Experimental results together with Posey−Tapperson− Rochelle and RK-EOS estimates for the system of CO2 in MEA:DEA: [Bmim][BF4] at 100 w = 31.8:12.1:56.1. ■, 303.15 K; ▲, 313.15 K; ◆, 323.15 K. *Dotted lines indicate model estimates.

All PCO2 values were recorded in MPa. To analyze the accuracy in estimation of absorption in each ionic liquid, a root-meansquare error was calculated by taking an average of the errors of each data point. n

root‐mean‐square‐error =

Figure 7. Experimental results together with Posey−Tapperson− Rochelle and RK-EOS estimates for the system of CO2 in MEA:MDEA: [Bmim][BF4] at 100 w = 30.3:21.8:48. ■, 303.15 K; ▲, 313.15 K; ◆, 323.15 K. *Dotted lines indicate model estimates.

of CO2−ionic liquid and a pseudobinary CO2−alkanolamine system, these binary interaction parameters for the RK-EOS was limited to CO2 and the [Bmim][BF4]. Interactions between CO2 and the alkanolamines constitute a reactive system and were treated using the Posey−Tapperson−Rochelle model. The inaccuracy in model estimation was calculated by the following equation for each data point: PCO2(experimental) − PCO2(calculated) PCO2(experimental)

n

(12)

where n = number of data points for each alkanolamine−ionic liquid hybrid system. n = 21 because each isotherm contained seven points. It is noteworthy that model estimates for equilibrium points at low pressure of 0.05 MPa and 0.1 MPa were highly inaccurate using the proposed models. This can be noted visually upon inspection of Figures 3 to 8 or by analyzing calculated data in Tables S-2 to S-7 of the Supporting Information provided. Inaccuracies in estimation of 50 % to 128 % for data measured at pressure of 0.05 MPs and 0.1 MPa, indicated the failure of the combined RK−Posey−Tapperson−Rochelle model to predict CO2 partial pressure in the systems studied in this research. It can be concluded that the model is inapplicable for hybrid systems of this nature at low pressure. Root-mean-square error was thus calculated neglecting data at 0.05 MPa and 0.1 MPa. The combining of [Bmim][BF4] with alkanolamines in the absence of water achieved a profound increase in CO2 absorption at low pressure, which the RK−Posey−Tapperson−Rochelle model failed to account for. This is likely due to the lack of information on the chemistry of the hybrid systems. Equilibrium CO2 solubility in conventional alkanolamines diluted with water is usually very low at low pressure due to the high dilution of the alkanolamine with water.6 While water assists in diffusion of CO2, the actual absorption of CO2 in water is negligible.33 Also, solubility of CO2 in pure ionic liquids is low at low pressure. This is noted in Figure 2a to c containing CO2 solubility in pure [Bmim][BF4] and [Bmim][Tf2N] measured in a previous work.18 The model generally predicts CO2 partial pressures that are greater than the measured results as shown in Figures 3 to 8. Regressed parameters shown in Table S8 of the Supporting Information are of relatively high magnitude, particularly in the case of parameters C and D for the Posey−Tapperson−Rochelle model. These parameters account for the concentration of the alkanolamines in the solvent, which has been accurately described because the composition of the alkanolamines is collectively up to 100 w = 50. By contrast, interaction parameters for the RK-EOS describing CO2 solubility in the ionic liquid are of relatively lower magnitude, possibly suggesting that most of

Figure 6. Experimental results together with Posey−Tapperson− Rochelle and RK-EOS estimates for the system of CO2 in MEA:MDEA: [Bmim][BF4] at 100 w = 31.6:10.4:58. ■, 303.15 K; ▲, 313.15 K; ◆, 323.15 K. *Dotted lines indicate model estimates.

error (%) =

∑i = 1 error i2

·100 (11) I



Article

The Posey−Tapperson−Rochelle model (eq 6), when differentiated, may be used to estimate the enthalpy of absorption of CO2 in the solvent:

the CO2 molecules are absorbed by the alkanolamine component of the solvent. The high CO2 solubility achieved in this work at low pressure suggests that absorption is not occurring simply through absorption of CO2 into [Bmim][BF4] and the alkanolamines separately. It indicates that significant interaction and alternative reactions occur between the alkanolamines, ionic liquid, and CO2. Although [Bmim][BF4] is known to be stable,8 the high solubility achieved at low pressure indicates that [Bmim][BF4] is at least facilitating diffusion of CO2 for the reaction with alkanolamines, with superior diffusion reaction kinetics to that of CO2 with water. The approach of treating these systems as an ideal combination of binary and pseudobinary systems was thus found to be inaccurate. A more complex model that can incorporate all interactions and account for reactive absorption is thus required to provide greater accuracy and representation. This is the subject of further study. Moreover, the assumed reaction for the Posey−Tapperson− Rochelle model is for alkanolamine−water systems. The attempt to apply the system concerned was to investigate whether alkanolamine−ionic liquid systems would behave in the same manner and could be modeled along similar reaction mechanisms. The results indicate substantial inaccuracy and limitations in this model’s applicability. The mechanism of CO2 absorption in alkanolamine−ionic liquid systems has thus been proven to behave significantly different to that for alkanolamine− water systems. The error presented in Table S-8 of the Supporting Information indicates that even neglecting the data at low pressure, error in partial pressure estimation as high as 9.645 % of measured values were still noted. Apart from the lack of information regarding the chemistry of the systems, the limitations of the Posey−Tapperson−Rochelle model consistent with literature28 could still be observed. CO2 partial pressure for systems containing [Bmim][BF4] and MEA only, were modeled with lower accuracy than systems containing MEA and DEA. This was especially evident for the system containing 100 w = 33 MEA with 100 w = 16.2 DEA. This is consistent with observations by Posey et al.34 and Dicko et al.,35 which showed more accurate correlation for CO2 in blends of MEA and DEA, than for systems containing only MEA. The limitations of the model were highlighted for systems containing MDEA, which were modeled with the lowest accuracy. 7.4 % to 7.7 % error was noted for systems containing MDEA, compared to 5.9 % to 6.4 % for systems containing MEA and DEA. It is important to note that as explained in Osman,5 Weiland et al. (1993),44 and Benamor and Aroua,42 the reaction mechanism between CO2 and primary or secondary amines is the same, but with tertiary amines, the reaction mechanism is different. However, the Posey−Tapperson−Rochelle model does not account for the differing reaction mechanism between CO2 and the tertiary amine MDEA. A single reaction is assumed for all amines as shown above. Any combined model that includes the Posey−Tapperson−Rochelle model would not provide accurate prediction for CO2 absorption in tertiary amines. This was also observed in Osman36 for systems containing differing concentrations of MDEA and DEA. It is also noted that the higher amount of MDEA contained in a system, the less accurately that system was modeled. The system containing MEA:DEA:MDEA:[Bmim][BF4] at 100 w = 29.8:11.7:12.8:45.7 as shown in Figure 8 was modeled with greater accuracy than the system containing MEA:MDEA: [Bmim][BF4] at 100 w = 30.3:21.8:48 shown in Figure 7.

ΔHabs = bR

(13)

When estimated for each of the systems studied, using the parameters found in Table S8 of the Supporting Information, differences in enthalpies of absorption range from 14 % in the case of MEA−[Bmim][BF4] systems to 606 % on the case of MEA−DEA−[Bmim][BF4] systems, further indicating a significant deviation in the mechanism of absorption from conventional alkanolamine−H2O systems.

■

CONCLUSIONS [Bmim][BF4] was combined with MEA, DEA, and MDEA at different compositions to create six hybrid solvents of varying ionic liquid and alkanolamine composition and investigated for CO2 solubility at pressurse of 0.05 MPa to 1.5 MPa and temperatures of 303.15 K, 313.15 K, and 323 K. Hybrid solvents achieved superior CO2 absorption over conventional alkanolamine solvents as well as pure ionic liquids studied in previous works and in other sources in the literature. Of all the hybrid solvents studied, the solvent containing MEA:DEA:[Bmim][BF4] at 100 w = 31.8:12.1:56.1 achieved the highest CO2 absorption. While the addition of DEA and MDEA in small quantities benefitted CO2 solubility, increasing the DEA and MDEA composition above 100 w = 10 achieved lower CO2 absorption. Calculations of enthalpy and entropy of absorption showed that CO2 absorption in solvents containing MDEA was decreased more significantly with increasing temperature than other in the case of other hybrid solvent systems. The modeling of CO2 absorption in hybrid solvents proved difficult. The RK−Posey−Tapperson−Rochelle Model did not successfully model CO2 solubility at pressure for systems of 0.1 MPa and lower, achieving high inaccuracies in CO2 partial pressure of 50 % to 128 % of measured data. When neglecting data at low pressure, inaccuracy amounted to 5.88 % to 9.65 % of measured CO2 partial pressure, for pressures of 0.4 MPa to 1.5 MPa. Systems containing MEA and DEA were the most accurately modeled of all systems studied with a root-meansquare error of 5.878 % for systems containing [Bmim][BF4]. Systems containing only MEA as the alkanolamine component were less accurately modeled than systems containing MEA and DEA, while all systems containing MDEA were modeled with the lowest accuracy. A greater understanding of the chemistry of hybrid solvent systems in imperative in order to apply more complex models, and may form a good basis for future work.

■

ASSOCIATED CONTENT

S Supporting Information *

An explanation of the buoyancy correction procedure is provided, together with measured masses of each component in the gravimetric analyzer in Table S-1 and a graphical illustration using Figure S-1. Tabulated values of the absorption data presented in Figures 2−8 of the manuscript are available. This includes measured absorption data together with desorption data, as well as calculated data using the models discussed in this work. This data is labeled as Tables S-2 to S-7. Model parameters together with error values are in Table S-8. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jced.5b00273. J



■

Article

(16) Ma’mun, S.; Nilsen, R.; Svendsen, H. F. Solubility of carbon dioxide in 30 mass % monoethanolamine and 50 mass % methyldiethanol amine solutions. J. Chem. Eng. Data 2005, 50, 630− 634. (17) Austgen, D. M.; Rochelle, G. T.; Chen, C. C. Model of vapour− liquid equilibria for aqueous acid gas-alkanolamine systems. 2. Representation of H2S and CO2 solubility in aqueous MDEA and CO2 solubility in aqueous mixtures of MDEA with MEA or DEA. Ind. Eng. Chem. Res. 1991, 30, 543−555. (18) Bahadur, I.; Osman, K.; Coquelet, C.; Naidoo, P.; Ramjugernath, D. Solubilities of Carbon Dioxide and Oxygen in the Ionic Liquids Methyl Trioctyl Ammonium Bis(trifluoromethylsulfonyl)imide, 1Butyl-3-Methyl Imidazolium Bis(trifluoromethylsulfonyl)imide, and 1Butyl-3-methyl Imidazolium Methyl Sulphate. J. Phys. Chem. B 2015, 119, 1503−1514. (19) Seddon, K. R.; Torres, M. J.; Moens, L.; Abrahams, M. A. In Clean Solvents; ACS Symposium Series; American Chemical Society, Washington, DC, 2001; Volume 819. (20) Rebolledo-Libreros, M. E.; Trejo, A. Density and Viscosity of Aqueous Blends of Three Alkanolamines: N-Methyldiethanolamine, Diethanolamine, and 2-Amino-2-methyl-1-propanol in the Range of (303 to 343) K. J. Chem. Eng. Data 2006, 51, 702−707. (21) Lee, H. S.; Seo, M. D.; Kang, J. W.; Yang, D. R. Measurement and Correlation of the Solubility of Carbon Dioxide in the Mixtures of Aqueous Monoethanolamine Solution and Benzoic Acid. J. Chem. Eng. Data 2012, 57, 3744−3750. (22) Ethanolamines; The Dow Chemical Company: Midland, MI, 2003; p 6, http://msdssearch.dow.com/ PublishedLiteratureDOWCOM/dh_017d/0901b8038017d302. pdf?filepath=amines/pdfs/noreg/111-01375.pdf&fromPage=GetDoc (accessed on October 31, 2014). (23) Vahidi, M.; Moshtari, B. Dielectric data, densities, refractive indices, and their deviations of the binary mixtures of Nmethyldiethanolamine with sulfolane at temperatures 293.15−328.15 K and atmospheric pressure. Thermochim. Acta 2013, 551, 1−6. (24) Roper, M., Quotation Reference: IQ111015; The IGA Series Intelligent Gravimetric Analysers; Hiden Isochema Ltd., Warrington, UK, 2011. (25) Roper, M., The Determination of the Pore Size Distribution of an Activated Carbon using Dubinin−Astakhov Analysis of CO2 Adsorption at 273 K. Hiden Isochema Ltd.: Warrington, UK, 2005; http://www. hidenisochema.com/library/application_notes/#.VXdH5fmqqko (accessed January 11, 2011). (26) Macedonia, M. D.; Moore, D. D.; Maginn, E. J. Adsorption Studies of Methane, Ethane, and Argon in the Zeolite Mordenite: Molecular Simulations and Experiments. Langmuir 2000, 16, 3823− 3834. (27) Shiflett, M. B.; Yokozeki, A. Solubilities and diffusivities of carbon dioxide in ionic liquids: [bmim][PF6] and [bmim][BF4]. Ind. Eng. Chem. Res. 2005, 44, 4453−4464. (28) Tillner-Roth, R.; Friend, D. G. A Helmholtz free-energy formulation of the thermodynamic properties of the mixture (water + ammonia). J. Phys. Chem. Ref. Data 1998, 27, 63−96. (29) Yokozeki, A. Theoretical performance of various refrigerantabsorbent pairs in a vapor-absorption refrigeration cycle by the use of equation of state. Appl. Energy 2005, 80, 383−399. (30) Yokozeki, A. Solubility of refrigerants in various lubricants. Int. J. Thermophys. 2001, 22, 1057−1071. (31) Valderrama, J. O.; Robles, P. A. Critical properties, normal boiling temperatures, and acentric factors of fifty ionic liquids. Ind. Eng. Chem. Res. 2007, 46, 1338−1344. (32) Alvarez, V. H.; Valderrama, J. O. A modified Lydersen−Joback− Reid method to estimate the critical properties of biomolecules. Alimentaria 2004, 254, 55−66. (33) Osman, K.; Coquelet, C.; Ramjugernath, D. Absorption Data and Modeling of Carbon Dioxide in Aqueous Blends of Bis(2hydroxyethyl)methylamine (MDEA) and 2,2-Iminodiethanol (DEA): 25 % MDEA + 25 % DEA and 30 % MDEA + 20 % DEA. J. Chem. Eng. Data 2012, 57, 1607−1620.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge University of KwaZulu-Natal for a doctoral scholarship for Dr K. Osman. This work is based upon research supported by the South African Research Chairs Initiative of the Department of Science and Technology and the National Research Foundation.

■

REFERENCES

(1) Key World Energy Statistics2010; International Energy Agency: Paris, 2010. (2) Intergovernmental Panel on Climate Change (IPCC), 2013: Summary for Policymakers. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; Stocker, T. F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V., Midgley, P. M., Eds.; Cambridge University Press: Cambridge, UK, and New York, NY, 2013. (3) Eskom Power Generation, EskomCoal Power; Eskom Holdings Ltd.: Johannesburg, South Africa, 2011; http://www.eskom.co.za/ AboutElectricity/ElectricityTechnologies/Pages/Coal_Power.aspx (accessed Feb 7, 2014). ̂ (4) L’Agence de l’Environnement et de la Maitrise de l’Energie (ADEME) CO2 Capture and Storage in the Subsurface: A Technological Pathway for Combating Climate Change, Charbonnages de France, the Paris School of Mines and BRGM: Paris, 2007; ISBN: 978-2-7159-2438. (5) Osman K., Carbon Dioxide Removal from Coal Power PlantsA Review Of Current Capture Techniques and an Investigation of Carbon Dioxide Absorption Using Hybrid Solvents, Doctoral Thesis, University of KwaZulu Natal: Durban, South Africa, 2014. (6) Figueroa, J. D.; Fout, T.; Plasynski, S.; McIlvried, H.; Srivastava, R. D. Advances in CO2 capture technologyThe U.S. Department of Energy’s carbon sequestration program. Int. J. Greenhouse Gas Control 2008, 2, 9−20. (7) Lecomte F., Broutin P., Lebas E. Le Captage du CO2: Des Technologies Pour Réduire Les Émissions De Gaz À Effet De Serre (The Capture of CO2: Technologies for Reducing the Emissions of Greenhouse Gases); Technip: Paris, 2010; ISBN 978-2-7108-0938-8. (8) Arshad M. W. CO2 capture using Ionic Liquids. Master Thesis. Department of Chemical and Biochemical Engineering, Technical University of Denmark, 2009; http://etd.dtu.dk/thesis/240068/ CO2Captureusingionicliquid.pdf (accessed March 28, 2014). (9) Lei, Z.; Dai, C.; Chen, B. Gas Solubility in Ionic Liquids. Chem. Rev. 2014, 114, 1289−1326. (10) Jou, F.; Otto, F. D.; Mather, A. E. Vapor−Liquid Equilibrium of Carbon Dioxide in Aqueous Mixtures of Monoethanolamine and Methyldiethanolamine. Ind. Eng. Chem. Res. 1994, 33, 2002−2005. (11) Zhang, Y.; Zhang, S.; Lu, X.; Zhou, Q.; Fan, W.; Zhang, X. Dual amino-functionalized phosphonium ionic liquids for CO2 capture. Chem.Eur. J. 2009, 15, 3003−3011. (12) Ma, J.; Zhou, Z.; Zhang, F.; Fang, C.; Wu, Y.; Zhang, Z.; Li, A. Ditetraalkylammonium Amino Acid Ionic Liquids as CO2 Absorbents of High Capacity. Environ. Sci. Technol. 2011, 45, 10627−10633. (13) Zhang, S.-J.; Zhang, X.-P.; Zhao, Y.-S.; Zhao, G.-Y.; Yao, X.-Q.; Yao, H.-W. A novel ionic liquids-based scrubbing process for efficient CO2 capture. Sci. China Chem. 2010, 53, 1549−1553. (14) Camper, D.; Bara, J. E.; Gin, D. L.; Noble, R. D. RoomTemperature Ionic Liquid−Amine Solutions: Tunable Solvents for Efficient and Reversible Capture of CO2. Ind. Eng. Chem. Res. 2008, 47, 8496−8498. (15) Nerula, S. C.; Ashraf, M., Carbon dioxide separation, Process Economics Program; SRI International: Menlo Park, CA, 1987. K



Article

(34) Posey, M. L.; Tapperson, K. G.; Rochelle, G. T. A simple model for prediction of acid gas solubilities in alkanolamines. Gas Sep. Purif. 1996, 10, 181−186. (35) Dicko, M.; Coquelet, C.; Jarne, C.; Northrop, S.; Richon, D. Acid gases partial pressures above a 50 wt % aqueous methyldiethanolamine solution experimental work and modelling. Fluid Phase Equilib. 2010, 289, 87−97. (36) Osman, K., Carbon Dioxide Capture Methods for Industrial Sources: A Literature Review, Energy Efficiency and Feasibility Study; University of KwaZulu Natal: Durban, South Africa, 2011. (37) Park, S. H.; Lee, K. B.; Hyun, J. C.; Kim, S. H. Correlation and Prediction of the Solubility of Carbon Dioxide in Aqueous Alkanolamine and Mixed Alkanolamine Solutions. Ind. Eng. Chem. Res. 2002, 41, 1658−1665. (38) , DOE/NETL Advanced Carbon Dioxide Capture R&D Program: Technology Update; National Energy Technology Laboratory (NETL): Pittsburgh, PA, 2013; http://www.netl.doe.gov/File%20Library/ Unassigned/CO2-Capture-Tech-Update-2013.pdf (accessed February 7, 2014). (39) DiGuillo, R. M.; Lee, R.; Schaeffer, S. T.; Brasher, L. L.; Tela, A. S. Densities and Viscosities of the Ethanolamines. J. Chem. Eng. Data 1992, 37, 239−242. (40) Prausnitz, J. M.; Lichtenthaler, R. N.; Azevedo, E. G., Molecular Thermodynamics of Fluid Phase Equilibria, 3rd ed.; Prentice Hall: Upper Saddle River, NJ, 1999. (41) Cadena, C.; Anthony, J. L.; Shah, J. K.; Morrow, T. I.; Brennecke, J. F.; Maginn, E. J. Why Is CO2 So Soluble in Imidazolium-Based Ionic Liquids? J. Am. Chem. Soc. 2004, 126, 5300−5308. (42) Benamor, A.; Aroua, M. K. Modelling of CO2 solubility and carbamate concentration in DEA, MDEA and their mixtures using the Deshmukh−Mather model. Fluid Phase Equilib. 2005, 231, 150−162. (43) Kim, I.; Svendsen, H. F. Heat of Absorption of Carbon Dioxide (CO 2 ) in Monoethanolamine (MEA) and 2-(Aminoethyl)ethanolamine (AEEA) Solutions. Ind. Eng. Chem. Res. 2007, 46, 5803−5809. (44) Weiland, R. H.; Chakravarty, T.; Mather, A. E. Solubility of carbon dioxide and hydrogen sulphide in aqueous alkanolamines. Ind. Eng. Chem. Res. 1993, 32, 1419−1430.

L


CO2 Solubility in Hybrid Solvents Containing 1-Butyl-3

Recommend Documents