The CO₂ Absorption in Flue Gas Using Mixed Ionic Liquids

Abstract

Because of the appealing properties, ionic liquids (ILs) are believed to be promising alternatives for the CO₂ absorption in the flue gas. Several ILs, such as [NH₂emim][BF₄], [C₄mim][OAc], and [NH₂emim[OAc], have been used to capture CO₂ of the simulated flue gas in this work. The structural changes of the ILs before and after absorption were also investigated by quantum chemical methods, FTIR, and NMR technologies. However, the experimental results and theoretical calculation showed that the flue gas component SO₂ would significantly weaken the CO₂ absorption performance of the ILs. SO₂ was more likely to react with the active sites of the ILs than CO₂. To improve the absorption capacity, the ionic liquid (IL) mixture [C₄mim][OAc]/ [NH₂emim][BF₄] were employed for the CO₂ absorption of the flue gas. It is found that the CO₂ absorption capacity would be increased by about 25%, even in the presence of SO₂. The calculation results suggested that CO₂ could not compete with SO₂ for reacting with the IL during the absorption process. Nevertheless, SO₂ might be first captured by the [NH₂emim][BF₄] of the IL mixture, and then the [C₄mim][OAc] ionic liquid could absorb more CO₂ without the interference of SO₂.

1. Introduction

The CO₂ absorption of flue gas is an important process for reducing greenhouse gas [1]. To date, most flue absorptions are performed by using amine solvents. However, conventional absorption methods usually have some disadvantages: High equipment corrosion rate, high absorbent make-up rate due to the amine degradation by SO₂, NO₂, and O₂ in the flue gas, and high energy consumption during the regeneration process [2]. In the last decades, ionic liquids (ILs) have been used in many fields. Multifunctional ionic liquids are easily prepared, and the vapor pressure of ionic liquids can be neglected [3]. The other attractive properties of ILs include: High thermal stability, large electrochemical window, and high dissolve ability of compounds [4]. Blanchard et al. [5] have reported that certain ILs can considerably dissolve CO₂ gas. Since then, ILs for CO₂ capture have attracted much attention. For example, multi-N-containing ionic liquids can absorb much more SO₂/CO₂ in the flue gas than that of the limestone solvent [6]. Shiflett et al. [7,8] have found that the imidazolium-based ionic liquid [C₄mim][OAc] can markedly reduce the energy losses of CO₂ absorption comparing with those of the commercial monoethanolamine solvent. Guanidinium salt ILs (e.g., [TMG][L]) and functional ILs (e.g., [NH₂p-bmim][BF₄] and 2-(2-hydroxyethoxy) ammonium acetate) all show high efficiency for CO₂ and SO₂ capture [9,10].

The typical flue gas from coal-burning usually contains about 15 vol% CO₂, 10 vol% H₂O, and more than 2 vol% SO₂ [11]. Apart from CO₂, the effects of SO₂ on the flue gas absorption should be taken into account [12]. Most researchers consider that water has a little influence on the CO₂ capture, but the effects of SO₂ would not be neglected. For the impurities of the flue gas, it is found that the ILs are more likely to absorb SO₂ than CO₂. Specifically, the N element of the ionic liquid (IL) prefers to capture the SO₂ molecules, and then CO₂ molecules are repelled by the captured SO₂ [13]. Thus, the CO₂ absorption capacity of ionic liquids would be rapidly decreased over several cycles. Moreover, almost all of the ILs would exhibit much higher SO₂ absorption capacity than CO₂ due to the higher solubility of SO₂ in ILs. For instance, pure CO₂ solubility in the guanidinium-based ILs was only 0.4 mol/mol, while the SO₂ solubility in these ILs was as high as 2.5 mol/mol under the same conditions [10]. For an extreme case, the absorption capacities of pure SO₂ and CO₂ in the azole-based ILs (e.g., [P₆₆₆₁₄][Im]) were 3.5 mol/mol and 0.1 mol/mol, respectively [14]. Although SO₂ has stronger interactions with ILs than CO₂ does, the actual partial pressure of SO₂ in the flue gas is very low [15]. It is well accepted that the partial pressure of SO₂ is about two orders of magnitude lower than that of CO₂, and low SO₂ partial pressure usually leads to low absorption capacity for SO₂. That is, a lot of energy will be consumed to remove SO₂ and CO₂ from the flue gas, if we used a two-step absorption process (absorbing SO₂ at first, and then CO₂).

The effects of single IL on the CO₂ capture have been widely discussed in recent years [2,7,16,17,18,19]. However, there are few reports on the IL mixtures for CO₂ absorption, especially on the CO₂ capture from the flue gas. Therefore, if the IL mixture was employed, the CO₂ absorption capacity of the mixed ILs might be greater than that of the single ionic liquid. Specifically, if the SO₂ of flue gas was absorbed by one IL of the mixture, the negative influence of SO₂ on the whole IL mixture might be significantly decreased, and then more CO₂ could be captured.

In this work, we want to use IL mixtures to remove CO₂ and SO₂ from the flue gas. The amine-functional ionic liquid [NH₂emim][BF₄] and the imidazolium-based ionic liquid [C₄mim][OAc] were synthesized at first. Subsequently, the two ILs were mixed to investigate the CO₂ and SO₂ absorption. Here, [NH₂emim][BF₄] was mainly used to absorb SO₂ in the flue gas, while the [C₄mim][OAc] ionic liquid was employed to capture CO₂. In order to more clearly study the actual flue gas, the CO₂ absorption performance in the IL mixture was measured at the simulated flue gas with 15 vol% CO₂ and 2 vol% SO₂. The effect of SO₂ on the IL mixture and the interaction between CO₂ and SO₂ in the IL absorption were also studied. Furthermore, the absorption mechanism at the molecule level was investigated by the quantum chemical calculation and the instrumental analysis.

2. Results and Discussion

2.1. CO₂ and SO₂ Absorption Performance of the Single ILs

When the simulated flue gas (15% CO₂/85% N₂) only contains the CO₂ impurity, one mole [C₄mim][OAc] could absorb 0.298 mol CO₂ (

Table 1. Summary of CO₂ absorption capacity by single ionic liquids.

Ionic Liquids	T, K	Gas	Absorption Capacity, mol CO2/mol IL
[C4mim][OAc]	293	15% CO2/85% N2	0.298
[C4mim][OAc]	293	15% CO2/2% SO2/83% N2	0.204
[NH2emim][BF4]	293	15% CO2/85% N2	0.290
[NH2emim][BF4]	293	15% CO2/2% SO2/83% N2	0.180
[NH2emim][OAc]	293	15% CO2/85% N2	0.291
[NH2emim][OAc]	293	15% CO2/2% SO2/83% N2	0.171

). However, when SO₂ was mixed in the simulated flue gas (15% CO₂/2% SO₂/N₂), the CO₂ absorption capacity of [C₄mim][OAc] was reduced to 0.204 CO₂/mol IL. The result shows that SO₂ has a negative effect on the IL absorbing CO₂. The single IL [NH₂emim][OAc] and [NH₂emim][BF₄] were also used to capture the CO₂ of flue gas. Without the interference of SO₂, the CO₂ absorption capacities of [NH₂emim][OAc] and [NH₂emim][BF₄] were 0.291 mol CO₂/mol IL and 0.290 mol CO₂/mol IL, respectively. However, like the [C₄mim][OAc] ionic liquid, the CO₂ absorption capacities under exposure to 2% SO₂ would be markedly decreased to 0.171 mol CO₂/mol [NH₂emim][OAc] and 0.180 mol CO₂/mol [NH₂emim][BF₄], respectively. In short, the single ionic liquids all exhibit the CO₂ absorption capacities, but this capacity would be greatly weakened by the interference of SO₂.

The researchers believed that CO₂ and SO₂ of the flue gas would be absorbed simultaneously [13,20]. The [C₄mim][OAc] and [NH₂emim][BF₄] absorption results supported this conclusion. For example, an experiment was carried out of outlet SO₂ concentration vs. time to investigate the SO₂ absorption performance. In this study (Figure 1a), the simulated flue gas contained 15% CO₂, 2% SO₂, and 83% N₂. The concentrations of CO₂ and SO₂ of the outlet stream were simultaneously detected with time. It was found that [C₄mim][OAc] and [NH₂emim][BF₄] all can capture CO₂, but the outlet SO₂ concentration was hardly detected before 40 min (Figure 1a). It indicates that SO₂ could be completely absorbed by [C₄mim][OAc] and [NH₂emim][BF₄] during the absorption process. Additionally, an extreme case was investigated in which the simulated flue gas contained 80% CO₂, 2% SO₂, and 18% N₂. However, SO₂ was also not found at the gas stream of the outlet before 15 min. Compared with CO₂, SO₂ has higher dipole moments and molecular polarity, which often results in the strong affinity of SO₂ with ionic liquids [21,22].

The presence of SO₂ in flue gas usually leads to a competitive and negative influence on the separation of CO₂. Figure 2 shows the CO₂ absorption performance of [C₄mim][OAc] at the atmosphere of 15% CO₂/85% N₂ and 15% CO₂/2% SO₂/83% N₂, respectively. After 6 regeneration cycles, 1 mol [C₄mim][OAc] could absorb 0.255 mol CO_2, although the IL absorption capacity slightly decreased. In contrast, the absorption capacity of 1 mol [C₄mim][OAc] was only 0.10 mol CO₂ after the same number of cycles. In addition, the net SO₂ absorption experiment (2% SO₂/98% N₂) showed that [NH₂emim][BF₄] has high SO₂ absorption capacity (0.35 mol SO₂/mol IL), while 1 mol [C₄mim][OAc] only absorbs 0.18 mol SO₂ (Figure 1b). These results agreed well with previous studies [23,24,25].

FT-IR can investigate the interaction between IL and CO₂/SO₂ [26]. The spectra of [C₄mim][OAc] showed the changes after 15% CO₂ and 2% SO₂ absorption, respectively (Figure 3a−c). However, the [C₄mim][OAc] spectrum had minimal changes when it was used to remove pure CO₂. Although the appearance of the carbonyl band at 1720 cm⁻¹ shows that the acetate anion might be partly converted into the acetate acid [27], Shiflett considered that the amount of such a chemical reaction was minor and reversible [24]. Thus, the other reactions between CO₂ and the cation species in this work might not be detected within the wavenumber of 800−1600 cm⁻¹, as Figure 3a,c shown. Similarly, the FT-IR spectra of [NH₂emim][BF₄] almost have no change before and after CO₂ absorption (Figure 3d,f), indicating that the chemical reaction between the [NH₂emim] cation and CO₂ was not enough to be detected by FT-IR.

In contrast, the chemical reaction between SO₂ and [C₄mim][OAc] was much stronger than that of [C₄mim][OAc]−CO₂. Even if there was only a small amount of SO₂ (2%) in the flue gas, the FT-IR spectrum still shows marked changes (Figure 3b). Compared with the spectrum of fresh [C₄mim][OAc], the peak intensity at 1580 and 1371 cm⁻¹ decreased significantly as the [C₄mim][OAc] absorbing SO₂. Meanwhile, the peaks at 1720, 1321, 1254, 1144, and 950 cm⁻¹ newly appeared in the spectrum. The bands at 1321 and 1144 cm⁻¹ can be attributed to the stretching of SO₂ absorbed by the ionic liquid [26]. After SO₂ absorption, the new peak at 1720 cm⁻¹ shows the formation of a carbonyl group, which also indicates that most of the acetate ions were no longer associated with [C₄mim] cations. The intense band at 950 cm⁻¹ should be assigned to the vibrational mode of SO₃²⁻ or S₂O₅²⁻. It once again suggests that the interaction between the SO₂ and [C₄mim][OAc] ionic liquid was strong. Similarly, the peak intensity at 885 and 1543 cm⁻¹ changed markedly when the [NH₂emim][BF₄] absorbed SO₂. Particularly, two new peaks appeared at 968 and1367 cm⁻¹, which can be attributed to the interaction between the N elements of the IL and SO₂ [6].

2.2. CO₂/SO₂ Absorption Properties in IL Mixtures

Due to the influence of SO₂, the CO₂ absorption capacity of single ionic liquids was greatly decreased. If [NH₂emim][BF₄] and [C₄mim][OAc] were simultaneously utilized, more CO₂ (with SO₂) in the flue gas might be captured. The CO₂ absorption capacities of IL mixtures at different mole fractions of [C₄mim][OAc] or [NH₂emim][BF₄] were displayed in Figure 4. Here, X is defined as the molar ratio of [NH₂emim][BF₄] to the IL mixture ([C₄mim][OAc]/[NH₂emim][BF₄]). Compared with the CO₂ absorption capacity of the single IL, the IL mixture could remove more CO₂. This result may be due to the presence of [NH₂emim][BF₄]. The small amount of SO₂ might be absorbed by [NH₂emim][BF₄] at first. Without the interference of SO₂, the CO₂ absorption capacity of the IL mixture was significantly enhanced. It was also found that the absorption of CO₂ did not increase significantly with the increase of the X value. When X was 0.3, the CO₂ absorption capacity of the [C₄mim][OAc]/[NH₂emim][BF₄] mixture reached up to the maximum. Similarly, when the IL mixture [C₄mim][OAc]/[NH₂emim][OAc] was used to remove CO₂/SO₂ of the flue gas, the poison effect of SO₂ on [C₄mim][OAc] was also greatly reduced. As Figure 4 shows, the CO₂ absorption of [C₄mim][OAc]/[NH₂emim][OAc] was about 0.4 mol CO₂/mol IL. Compared to the single [C₄mim][OAc], the absorption capacity of the IL mixture was improved.

The ¹H NMR spectrum (Figure 5) shows that SO₂ would interact with [NH₂emim][BF₄] and [NH₂emim][OAc]. The NMR data of fresh [NH₂emim][BF₄] and fresh [NH₂emim][OAc] are listed as follows:

Fresh [NH₂emim][BF₄], δ = 7.721 (s, 1H, unsaturated C−H in the imidazole ring, with N connected to the left and right), 7.641 (d, 1H, unsaturated C−H in the imidazole ring), 7.636 (d, 1H, unsaturated C−H in the imidazole ring), 4.295 (s, 3H, H₃C−N ring), 4.039 (t, 2H, H₂C−N ring), 3.112 (m, 2H, N−CH₂−C−N ring), and 1.878 (t, 2H, NH₂).

Fresh [NH₂emim][OAc], δ = 12.751 (s, 1H, unsaturated C−H in the imidazole ring, with N connected to the left and right), 7.865 (d, 1H, unsaturated C−H in the imidazole ring), 7.703 (d, 1H, unsaturated C−H in the imidazole ring), 3.577 (s, 3H, H₃C−N ring), 3.749 (t, 2H, H₂C−N ring), 2.693 (m, 2H, N−CH₂−C−N ring), 1.973 (t, 2H, NH₂) and 1.651 (s, 3H, CH₃ in OAc-).

In comparison to the ¹H NMR spectrum of the fresh [NH₂emim][BF₄] (Figure 5a), new resonance peaks at 8.10 ppm were found after SO₂ absorption, which indicates the formation of S^…N [28]. According to this result, it was considered that the interaction between SO₂ and [NH₂emim][BF₄] should mainly occur at the N element of the [NH₂emim] cation. For the case of the [NH₂emim][OAc] (Figure 5b), a typical peak of −COOH in the ¹H NMR spectrum moved from 12.75 to 11.83 ppm, and a new resonance peak was observed at 7.52 ppm after SO₂ absorption. These results suggest that the interaction between [NH₂emim][OAc] and SO₂ had occurred [25]. That is, the interaction between [OAc] and SO₂ leads to the moving from 12.75 ppm to 11.83, while the reaction of [NH₂emim] and SO₂ makes the new peak 7.52 ppm appearance.

In order to further investigate the effects of SO₂ on the CO₂ absorption capacity of ionic liquids, the CO₂ absorption performance of fresh IL and after SO₂-saturated IL are illustrated in Figure 6. Specifically, fresh [C₄mim][OAc] and fresh [NH₂emim][BF₄] ionic liquids were used to absorb SO₂ at first. When the IL was saturated by SO₂, the CO₂ absorption performance of the IL was investigated. It was found that the SO₂-saturated [C₄mim][OAc] did not have the ability to absorb CO₂. The concentration of CO₂ at the outlet was almost equal to that at the inlet. While fresh [C₄mim][OAc] can absorb CO₂ even after 60 min. The similar results were also observed using fresh [C₂mim][OAc] and SO₂ saturated [C₂mim][OAc] to absorb CO₂ [24]. Shiflet et al. considered that the interaction between the [OAc] anion and CO₂ plays an important role in the CO₂ removal of [OAc]−based ionic liquids [8]. However, the presence of SO₂ makes a great impact on the CO₂ absorption of [C₄mim][OAc]. In contrast, when [NH₂emim][BF₄] was saturated by SO₂, the [NH₂emim][BF₄] still had the CO₂ absorption capacity. As Figure 6b shows, SO₂-saturated [NH₂emim][BF₄] could capture about 2−7% CO₂ during the absorption process.

2.3. Quantum Chemical Calculation on the Interaction of IL Mixture with CO₂/SO₂

The absorption capacity of CO₂ in the IL mixtures was higher than that of the single ionic liquid. This may be related to the interactions between ILs and CO₂/SO₂ molecules. Thus, the interaction of the [C₄mim][OAc] anion and [NH₂emim][BF₄] with CO₂/SO₂ was deeply investigated through quantum chemical calculation, which might be helpful to understand well the roles of CO₂ and SO₂ in the IL absorption. In this work, the structure of [NH₂emim][BF₄] and [C₄mim][OAc] was optimized on the basis of DFT-D3 calculation at first. The configuration of the IL with the lowest energy was considered as the optimized structure. Additionally, the structures of [C₄mim][OAc] and [NH₂emim][BF₄] with CO₂ and SO₂ were also investigated (Figure 7). The structural parameters for the IL−CO₂/SO₂ complexes are listed in

Table 2. Structural parameters for the complexes.

Structural Parameters	CO2−[NH2emim][BF4]	SO2−[NH2emim][BF4]	[C4mim][OAc]−CO2	[C4mim][OAc]−SO2
C−O, Å	1.19		1.21
∠O−C−O, °	166		146
S−O, Å		1.46		1.49
∠O−S−O, °		113.5		112.6

.

In general, the CO₂/SO₂ gas molecules around the anions and cations were related to the absorption reaction. As Figure 7a shows, there was a strong interaction between the N atom and the S atom of SO₂. Due to the complexation of SO₂^…N, the average angle of SO₂ was 116.0°. Compared to 119.5° of the pure SO₂ molecule, the bending degree of O=S=O was increased. Similarly, [NH₂emim][BF₄] also leads to an impact on the CO₂ structure. The angle of CO₂ was bent from 180° to 166°, and the bond length of C−O was extended from 1.16 Å to 1.19 Å. For the cases of [C₄mim][OAc], the interaction between the O and the C atom of carbon dioxide was also strong due to the negatively charged oxygen (O atom) in the [OAc] anion. The curvature of CO₂ could be increased by the interaction of the [OAc] anion. The average angle of CO₂ was bent to 146°, and the bond length of C−O was elongated to 1.21 Å. The configurations in Figure 7 also suggest that [NH₂emim] cation and [OAc] anion are the active sites for CO₂/SO₂ absorption.

To some extent, the interaction energy and absorption enthalpy might reflect the absorption capacity of the ILs. It was found that the NH₂emim] cation and [OAc] anion were the main active sites for the absorption of CO₂ and SO₂. In order to save the calculation cost and reduce the interference of other ions, herein, only the thermodynamic data of [OAc]−CO₂, [OAc]−SO₂, CO₂−[NH₂emim], and SO₂−[NH₂emim] complexes were compared (

Table 3. Thermochemical parameters and charge transfer of the ion−CO₂/SO₂ complexes.

	OAc−CO2	OAc−SO2	OAc−CO2−SO2
ΔE, kJ/mol	−40.7	−113.4	−140.0
ΔH, kJ/mol	−46.5	−125.1	−151.4
ΔG, kJ/mol	−1.6	−70.8	−72.1
net charge transfer, e	−0.510	−0.382	−0.035(CO2)/−0.316(SO2)
	CO2−[NH2emim]	SO2−[NH2emim]	SO2−CO2−[NH2emim]
ΔE, kJ/mol	−33.8	−123.9	−156.8
ΔH, kJ/mol	−36.3	−126.7	−160.1
ΔG, kJ/mol	−7.2	−55.1	−61.9
net charge transfer, e	−0.312	−0.399	−0.308(CO2)/−0.334(SO2)

). It was found that the interaction energy and absorption enthalpy of [OAc]−CO₂−SO₂ complex were less than the sum of the energy and the enthalpy for [OAc]−CO₂ and [OAc]−SO₂, suggesting that CO₂ and SO₂ would competitively react with [OAc] anion.

In contrast, the interaction energy and absorption enthalpy of [NH₂emim]−CO₂−SO₂ complex were approximately equal to the sum of those for the [NH₂emim]−CO₂ and [NH₂emim]−SO₂ complexes, indicating that the competitive reaction between [NH₂emim]−CO₂ and [NH₂emim]−SO₂ was not obvious. The absorption reaction might also lead to a change in charge distribution. It was found that the amount of net charge transfer from CO₂ to [NH₂emim] in [NH₂emim]−CO₂−SO₂ complex was almost equal to that of [NH₂emim]−CO₂, suggesting that [NH₂emim] had strong interactions with either SO₂ or CO₂. However, due to the impact of SO₂, the net charge transfer from [OAc] to CO₂ was significantly reduced from −0.510 in the [OAc]−CO₂ complex to −0.035 in [OAc]−CO₂−SO₂ complex, respectively, which might account for the decrease of CO₂ absorption capacity for [C₄mim][OAc] in Table 1.

Table 4. Thermochemical parameters for the IL−CO₂/SO₂ complexes ^a.

	[C4mim][OAc]−CO2	[C4mim][OAc]−SO2	CO2−[NH2emim][BF4]	SO2−[NH2emim][BF4]
ΔE, kJ/mol	−26.4 (−21.9)	−80.1 (−70.5)	−19.2 (−16.8)	−85.4 (−76.3)
ΔH, kJ/mol	−30.5	−93.2	−29.1	−91.5
ΔG, kJ/mol	−2.7	−40.6	8.1	−37.2

^a Values of brackets were calculated by the continuum universal solvation model (SMD).

collects the thermochemical data of [NH₂emim][BF₄]−CO₂ and [NH₂emim][BF₄]−SO₂ complexes. In order to consider the solvent effect of ionic liquids, the continuum universal solvation model (SMD) was used in the calculation. Based on the SMD model, the interaction energies of [NH₂emim][BF₄]−CO₂ and [NH₂emim][BF₄]−SO₂ system were −16.8 and −76.3 kJ/mol, respectively. They were lower than those in the gas phase (−19.2 and −85.4 kJ/mol). Notably, the difference between the interaction energy ([NH₂emim][BF₄]−CO₂ and [NH₂emim][BF₄]−SO₂) in the gas phase (59.5 kJ/mol) was very consistent with the energy difference (66.2 kJ/mol) using the SMD model. In the liquid phase, the interaction energy between [NH₂emim][BF₄] and SO₂ was slightly greater than that of [NH₂emim][BF₄]−CO₂, suggesting that [NH₂emim][BF₄] tends to react with SO₂ rather than with CO₂ during the absorption process. Similarly, this phenomenon could also be observed in the thermodynamic data of the [C₄mim][OAc]−CO₂ and [C₄mim][OAc]−SO₂ complexes. In short, SO₂ was more active than CO₂ in the reaction with ionic liquids, and the [NH₂emim][BF₄] may be more likely to absorb SO₂.

The interaction between IL mixture ([C₄mim][OAc]/[NH₂emim][BF₄]) and CO₂/SO₂ has also been investigated by the quantum chemistry calculation. The optimized structure is displayed in Figure 8. In the mixed ionic liquids, it is found that SO₂ was close to [NH₂emim][BF₄], while the CO₂ molecule was near the [C₄mim][OAc] ionic liquid. Specifically, SO₂ would have interacted with the N atom on the [NH₂emim] cation, and CO₂ was more likely to react with the [OAc] anion. This result might explain why the IL mixture can more effectively absorb the CO₂ of flue gas. Because of the existence of [NH₂emim][BF₄], SO₂ may be first captured by [NH₂emim]. Without the interference of SO₂, the [C₄mim][OAc] ionic liquid then could absorb more CO₂.

3. Materials and Methods

3.1. Materials

The simulated flue gas was obtained by pure gas CO₂, SO₂, and N₂ (purity of >99.99 wt%). They were all purchased from Beifen (China) Gas Technology Company. 1-butyl-3-methylimidazolium tetrafluoroborate ([C₄mim][BF₄], 99 wt%) was obtained from Sigma-Aldrich Chemical Co., but the ionic liquid 1-butyl-3-methylimidazolium acetate ([C₄mim][OAc], 99 wt%) were purchased from Lanzhou Greenchem ILs, LICP, CAS, China. Additionally, [NH₂emim][BF₄] and [NH₂emim][OAc] have been synthesized by ourselves in this work. The used materials were as follows: 1-methylimidazole (C₄H₆N₂), 2-bromoethylamine hydrobromide (C₂H₇Br₂N), sodium acetate anhydrous (CH₃COONa), 1-methylimidazole (C₄H₆N₂), acetic acid (CH₃COOH), and sodium borate (NaBF₄). They were all provided by Sinopharm Chemical Reagent Co., Ltd., China, with purity over 98 wt%.

3.2. Ionic Liquid Preparation

In this work, the ionic liquids [NH₂emim][BF₄] and [NH₂emim][OAc] were prepared by ourselves according to the method used in the literature [17,18,29]. First, the [NH₂emim] cation was prepared by the reaction of 1-methylimidazole and 2-bromoethylamine hydrobromide under reflux for 12 h. Second, the [NH₂emim]-based IL was simply synthesized by ion exchange with NaBF₄ or NaOAc/CH₃COOH in ethanol, and then the ethanol was removed in vacuum. The structures of the ILs were confirmed by proton nuclear magnetic resonance (¹H NMR, Bruker WB400 AMX spectrometer, Billerica, MA, USA). Here, deuterated chloroform (CDCl₃) was used as a solvent, and tetramethylsilane (TMS) was employed as an internal standard for ¹H NMR measurement.

3.3. CO₂ and SO₂ Absorption

As Figure 9 shows, CO₂ and SO₂ absorption experiments were performed in a 30 mL reactor immersed with a water-bath temperature controller. The temperatures were controlled at 293 K for absorption and 353 K for desorption, respectively. The simulated flue gas was a mixture of N₂, CO₂, and SO₂ in accordance with a certain proportion. As a typical absorption process, 10 mL IL or IL mixtures were added to the reactor at first. Subsequently, 15 vol% CO₂, 2 vol% SO₂ and 83 vol% N₂ were mixed in storage. The intake speed of the mixed gas was controlled at 60 mL/min, and the absorption pressure was controlled at 101.3 kPa. The concentrations of CO₂ and SO₂ were analyzed by a gas analyzer (MRU NOVA2000) at the outlet. To investigate the IL regeneration, CO₂ or SO₂ saturated IL was also loaded in the reactor. Desorption was performed at 353 K under a pure N₂ gas atmosphere for 30 min.

The amount of absorbed CO₂ or SO₂ was calculated by the following equation.

$$A_{\mathrm{gas}}=\frac{M_{\mathrm{IL}}{\rho }_{\mathrm{gas}}Q{\int }_{t_1}^{t_2}(C_0-C_{\mathrm{gas}}(t))\mathrm{d}t}{m_{\mathrm{IL}}{M}_{\mathrm{gas}}}$$

where A_gas is the molar amount of CO₂ or SO₂ in the ionic liquid; Q is the flow rate of the gas stream; C₀ and C_gas are the CO₂ or SO₂ concentrations at the inlet and the outlet streams, respectively; t₁ refers to the beginning time of the absorption process; when the CO₂ and SO₂ concentration at the outlet stream returns to the initial concentration, the time is t₂; M_IL and M_gas are the molecular weight of IL and CO₂ (or SO₂), respectively; m_IL is the weight of the ILs, and ρ_gas is the density of CO₂ or SO₂.

After the IL was saturated by CO₂ and SO₂, the complex structure was investigated by the FTIR and NMR technologies. The FTIR spectra of the samples were analyzed on an FTIR spectrometer (PerkinElmer, Frontier 2500). In addition, the structure changes of the [NH₂emim]-based IL after absorption were also detected by the NMR spectrometer (Bruker WB400 AMX, 300 MHz) using chloroform-d (CDCl₃) as a solvent and tetramethylsilane (TMS) as an internal standard.

3.4. Theoretical Calculation

A quantum chemical calculation was used in this work to study the interaction between the ionic liquid and CO₂ with SO₂. All calculations were carried out by the Gaussian 16 program [30]. For IL calculations, Li et al. [31] suggested that the density function of the Minnesota family [32] (e.g., M06-2X) with a diffusion function basis set (e.g., 6-311++G(d,p)) might give reasonable results. If dispersion-corrected density functionals (e.g., gd3bj, DFT-D3) were used, more reliable results could be obtained [33]. Therefore, the geometry optimization and frequency analysis of all ILs and IL mixtures were performed at the M06-2X/6-311++G(d,p) level and correction with Grimme’s method. In order to calculate the interaction energy of the IL complexes, the basis set superposition error (BSSE) method was employed to correct the energy results [34]. The effect of the solvent should be taken into consideration in the theoretical calculation of the ionic liquids. It was found that the SMD solvation model proposed by Truhlar et al. can be used for the IL calculation very well [35,36]. Thus, the density functional theory (M06-2X and dispersion-corrected method) with the SMD model was also used to calculate the interaction energy of IL−CO₂/SO₂.

4. Conclusions

The CO₂ and SO₂ absorption of the flue gas in ionic liquids were investigated by the experimental method and theoretical calculation. The single ionic liquids, such as [NH₂emim][BF₄] and [C₄mim][OAc], all showed good CO₂ absorption performance for the simulated flue gas without SO₂ interference. However, SO₂ was more likely to react with the active sites of the ILs. When SO₂ was in the flue gas, the CO₂ absorption capacity of the single ionic liquid would be significantly inhibited. It was found that the interference of SO₂ on the CO₂ absorption performance might be markedly reduced by using the ionic liquid mixtures. The CO₂ absorption capacity of the IL mixture [C₄mim][OAc]/[NH₂emim][BF₄] was about 0.4 mol CO₂/mol IL even at an atmosphere of 15% CO₂/2%SO₂/83% N₂, which was greater than that of single [C₄mim][OAc] (0.204 mol CO₂/mol IL). There was a competitive relationship between CO₂ and SO₂ during the absorption process. The single ILs prefer to capture SO₂ rather than remove CO₂, due to the stronger interaction energy of SO₂ and the ILs. The experimental and calculated results suggested that the [OAc] anion and [NH₂emim] cation are the main active sites for CO₂ and SO₂ absorption. A lower absorption enthalpy of the IL−SO₂ or IL−CO₂ system usually means low absorption capacity. Thus, for the IL mixture [C₄mim][OAc]/[NH₂emim][BF₄], the quantum calculation results indicated that [NH₂emim][BF₄] might be more likely to absorb SO₂ of the flue gas and CO₂ was easily removed by the [C₄mim][OAc].