Effect of the Presence of MEA on the CO2 Capture Ability of Superbase Ionic Liquids

: The miscibility of monoethanolamine (MEA) in five superbase ionic liquids (ILs), namely the trihexyl-tetradecylphosphonium benzotriazolide, trihexyl-tetradecylphosphonium benzimidazolide, trihexyl-tetradecylphosphonium 1,2,3-triazolide, trihexyl-tetradecylphosphonium 1,2,4-triazolide, and trihexyl-tetradecylphosphonium imidazolide was determined at 295.15 K using 1 H-NMR spectroscopy. The solubility of carbon dioxide (CO 2 ) in equimolar (IL + MEA) mixtures was then studied experimentally using a gravimetric technique at 295.15 K and 0.1 MPa. The effect of MEA on the CO 2 capture ability of these ILs was investigated together with the viscosity of these systems in the presence or absence of CO 2 to evaluate their practical application in CO 2 capture processes. The effect of the presence of MEA on the rate of CO 2 uptake was also studied. The study showed that the MEA can enhance CO 2 absorption over the ideal values in the case of [P 66614 ][123Triz] and [P 66614 ][Bentriz] whilst in the other systems the mixtures behave ideally. A comparison of the effect of MEA addition with the addition of water to these superbase ILs showed that similar trends were observed in each case for the individual ILs studied.


INTRODUCTION
Carbon dioxide is a greenhouse gas that is released into the atmosphere from flue gas streams. It is believed that CO2 currently contributes to 60 % of the greenhouse gases released into the atmosphere. 1 With global CO2 emissions set to increase there is a real and immediate need to develop a process that can effectively and efficiently capture CO2 before being released into the atmosphere. The sources of CO2 include mobile and stationary sources with the latter associated predominantly from power plants in the form of flue gas streams. As well as emissions of CO2, it is also present in significant concentrations in biogas and natural gas. In these systems, the CO2 limits the combustion of the methane and must be separated before the gas can be utilized as a fuel. In both emissions and methane containing streams, efficient separation is critical for both economic and environmental reasons. To enable these types of separations, industrial practice is currently dominated by aqueous alkanolamine solvents (e.g. 30 wt% aqueous monoethanolamine (MEA)) where a chemical reaction occurs between the CO2 and amine solution. This process is attractive due to the low cost of solvent, high reactivity and an absorption capacity of 1:2 mole of CO2 per mole of solvent ratio. 2,3 These advantages are offset by the volatile and corrosive nature of the aqueous solvent mixtures and the high energy consumption needed for regeneration with an enthalpy of regeneration of CO2, 4 ~ 66.7-76.9 kJ•mol -1 , leading to high operational costs and environmental impact. 5,6 Due to these disadvantages there has been a recent increase in interest in the search for a more suitable solvent for post combustion CO2 capture.
In this respect ionic liquids (ILs) have received an increasing amount of interest in recent years as an alternative solvent for CO2 capture due to their potential as environmentally friendly or "green" solvents. ILs have highly attractive properties such as high thermal stability, negligible vapor pressure and their ability to be task specific. 7 Blanchard et al. 8 were the first to report high CO2 solubility in imidazolium-based ILs reaching a CO2 mole fraction of 0. 6  Recently, a range of superbase ILs have been reported which showed efficient and reversible CO2 capture with an uptake greater than the 1:1 nCO2:nIL. [16][17][18] Unlike the amino acid-based ILs, these ILs have a limited number of hydrogen atoms on the anion and, therefore, only a small increase in viscosity upon CO2 absorption is observed. In addition, as flue gas/methane streams contain water it is also important to understand how these ILs behave under wet conditions. The presence of water has been shown to affect the physical properties of the IL and also has the potential to compete with the CO2 for absorption. 18 mixtures showed an optimum composition of 0.7 mol·dm -3 MEA and 0.3 mol·dm -3 of IL with a capacity of 0.534 nCO2:nabsorbent and suggested that the components mutually promote the CO2 solubility. Furthermore, this particular system showed good stability and CO2 uptake and release as a capacity of 0.480 nCO2:nabsorbent was still observed after the 4 th cycle. 28 These recent studies have shown that these mixtures can reduce the viscosity of the ILs and decrease absorption time with maximum absorption being reached in 60 min in some cases. 26,27,29 It was also shown that they can reduce the loss of material and energy consumption associated with current amine technology. [21][22][23] This solution presents itself as an attractive method to utilize the performance of amines whilst using the desirable properties of ILs.
In this study MEA was added to the previously studied superbase ILs to investigate the effect of a chemically absorbing co-solvent compared with our previous work with water as a nonreacting co-solvent. 17  and allowed to stir at room temperature overnight. The organic layer was then extracted and washed with ultrapure water (100 cm 3 ) repeatedly five times and dried under high vacuum.
All ILs were then dried under high vacuum (1 Pa at 50 °C) and then stored in an argon filled dry glovebox prior to use.
Name, source, initial and final purity of compounds used during this work are listed in Table S1 of the electronic supporting information (ESI).

Methods
The purity of the synthesized ILs and IL:MEA mixtures were analyzed using 1 H-NMR using a Bruker Avance 400 MHz Ultra shield Plus NMR spectrometer as reported in Figure S1 of the

CO2 Absorption and Desorption Experiments
All samples were prepared, stored and weighed out into a small vial with a septum lid (1.9 cm 3 ) in an argon filled glovebox. The absorption measurements of the pure and water saturated systems were described in previous work. 18 The measurement of the CO2 uptake in the IL:MEA mixtures followed the same gravimetric procedure. The sample (~ 0. where ki is equilibrium constant of the reaction i (with i = 1, 2, 3, or 4) To solve this problem, several assumptions have been made, such as: (i) the formation of the IL:MEA complex is driven by an 1:1 molar ratio (eq. 1); (ii) the solubility of the CO2 in free IL was calculated on the basis of an 1:0.23 mole of IL per mole of CO2 ratio based on experimental data for pure IL system (eq. 2); (iii) the solubility of the CO2 in free MEA was determined on the basis of an 1:0.42 mole of IL per mole of MEA ratio based on experimental data for the pure MEA system; (iv) the solubility of the CO2 in the IL:MEA complex was assumed to follow an 1:1 mole of IL:MEA complex per mole of CO2 ratio.
Equation 5 was then used to calculate the equilibrium constant k1 (eq. 1).
where k1 is the equilibrium constant of the eq. 1 and x defines the number of moles of IL:MEA complex formed on the basis of an 1:1 molar ratio.    Figure 2 shows the comparison of the pure IL and equimolar IL:MEA systems with the industrial benchmark for CO2 capture process. It shows that each of these systems gives a better CO2 capacity than the aqueous amine system currently implemented.

Viscosity
Previous studies [27][28][29] Table  S3 of the ESI, that a small increase in viscosity has been generally observed in the case of the selected superbase ILs in the absence and presence of water and this should be examined for other potential IL systems. Figure 3 compares the absorption kinetics for the pure IL, (water + IL) at the water saturation level and equimolar (IL + MEA) systems for the case of [124Triz] as represented as % of the saturation number of moles of CO2 absorbed with respect to time (Table S4 of the ESI). No significant differences were observed for the three systems studied suggesting that the rate of absorption of CO2 is not just dependent on the viscosity. This is further demonstrated in Figure 4 as no correlation was found by fitting the viscosity of the equimolar (IL + MEA) mixtures before the CO2 uptake as the function of their initial rate of CO2 uptake. This indicates that mass transfer of CO2 is not rate determining for the uptake and, therefore, it is likely due to the structure of the system and varying strengths of interaction between the ion pairs is also playing a significant role.

Synergy between the IL and MEA Components
The five equimolar (IL + MEA) mixtures were probed to investigate their synergistic properties.
That is, are the components of the absorbent working together to increase the CO2 capture ability or are the two components working independently to absorb CO2. In Table 2 and MEA behave in a synergistic manner with enhanced CO2 solubility compared with the ideal case. Figure 5 also shows that the highest synergy is achieved at the 0.5:0. is not widely discussed in literature with many groups reporting the reduction in viscosity when an amine is present as a main advantage of these systems for CO2 capture. [24][25][26] A comparison of these systems was also performed with a physically absorbing IL (e.g.
[ The enhanced CO2 capture observed suggests that the CO2 competes effectively with the -OH group on the MEA to reprotonate the anion and both the IL and MEA components chemically absorbing the CO2. This is similar to what was reported in the previous work where the H2O is also seen to compete effectively with the CO2. 18 It is likely that this is an entropy driven reaction and the presence of a co-solvent causes a weakening of the interaction between the anion and cation giving the CO2 greater opportunity for interaction with the anion of the IL, as noted for the system in the presence of water. 18 It was reported previously that the H2O disrupts the anioncation interaction where an enhancement was observed allowing increased interaction of the anion with CO2. In general, this occurs for the anions which have low CO2 solubility in the pure IL system. It is, therefore, possible that the MEA acts in a similar manner although in this case a stable complex is formed which is unlikely to be the case in the H2O containing system.

Desorption and Regeneration
Regeneration tests were carried out using an equimolar [P66614] [Bentriz]:MEA mixture at desorption temperatures of 353.15 K and 393.15 K. As shown in Table 5 it was found that at 353.15 K, the CO2 capacity decreased by 47 % after three cycles and no synergistic effect was observable. This is thought to be due to the fact that, at this temperature, not all of the CO2 has been removed which reduces the number of active sites for CO2 capture. When the temperature of desorption was increased to 393.15 K, it can be seen that the CO2 absorption capacity decreased slightly but even after three cycles a significant synergistic effect was observable.
After desorbing CO2 at 393.15 K it is necessary to add an average of 3.84 wt% MEA to restore an equimolar system. However, the amount of fresh MEA needed in the system could be reduced by condensing the MEA removed in the gas phase back into the system.

CONCLUSIONS
In this work, five trihexyl-tetradecylphoshonium-based ionic liquids paired with superbasic Importantly, whilst the addition of the MEA co-solvent decreased the viscosity compared with the pure IL, a large increase was seen on absorption of CO2 which was much higher than that found using water as the co-solvent. After careful investigation of these systems it is apparent that water has more potential as a co-solvent compared with MEA for CO2 absorption. This is due to the large increase in viscosity upon CO2 absorption in the presence of MEA making it unsuitable for industrial use.

SUPPORTING INFORMATION AVAILABLE
1 H-NMR spectra for synthesized ILs are available in Figure S1. Name, source, initial and final purity of compounds used during this work are listed in Table S1. Data of the relative volatility of the MEA in each of the investigated mixtures, the viscosity of (water + IL) mixtures and the CO2 solubility in each of the investigated mixture as the function of time are tabulated in Tables S2-S4, respectively.