A novel phenolic ionic liquid for 1.5 molar CO₂ capture: combined experimental and DFT studies

Abstract

A phenolic-based ionic liquid (IL), 1-(2-hydroxyethyl)-2,3-dimethylimidazolium phenoxide has been introduced for 1.5 molar carbon dioxide (CO₂) absorption at ambient conditions without using special methodologies or precautions. The structure of the IL was characterized by various methods such as FT-IR, ¹H NMR, ¹³C NMR, elemental analysis (EA) and thermogravimetric analysis (TGA). The ability of this IL for CO₂ uptake reaches to its maximum value after 2 h. The IL has been reused fr several times with constant efficiency. Density functional theory (DFT) calculations at the B3LYP/6-311++G** level of calculation have been carried out to gain more structural knowledge about the interaction of CO₂ with IL. Based on the results of experimental and theoretical investigations, a reasonable mechanism is proposed for CO₂ absorption.

---

Introduction

Carbon dioxide (CO₂) is a harmful greenhouse gas with 50–200 years life-time in the atmosphere.¹ Accumulation of CO₂ in the atmosphere resulted in global warming and hence strongly threatens human life. Another challenge in the field of carbon dioxide is to reduce its amount to an acceptable level during sweetening operations of sour gas.² Several methods have been developed for CO₂ capture including applying porous materials and membranes.^2,3 Despite few advantages such as low corrosion and low regeneration energy, the capacity and efficiency of these solid adsorbents (especially at low concentration of CO₂) are not desirable.⁴ Moreover, the application of the membranes is generally limited for purification of natural gas. In sweetening operations, the selectivity of this strategy is not desirable.⁴

The current industrial method for CO₂ capture comprises using aqueous amine solution such as ethanolamine.⁵ Although this compound has acceptable sorption capacity and selectivity, however it generally suffers from some serious drawbacks including losing the absorbent, equipment corrosion and high energy consumption during regeneration process.⁶

Ionic liquids (ILs) are known as versatile chemicals with widespread applications such as solvent, catalysts and gas separation.⁷ Maybe the most important application of ILs in gas separation processes refer to CO₂ capture chemistry. The first report of application of IL for CO₂ capture was introduced in 2002 which an imidazolium-based ionic liquid with N-substituted ethylamine moiety.⁸ In this report, the amine functionality of the mentioned IL interacted with CO₂ to form an IL-based carbamate moiety. By heating the CO₂-treated-IL, the captured carbon dioxide was released and thus the initial IL can be simply regenerated. After this interesting discovery, various strategies have been developed to improve and extend this method.⁹

Today, the major problem which stands against the progress of using ILs for CO₂ capture is the absorption capacity of these compounds. In the case of most amine-based ILs, each mol of IL only absorbs 0.5 mol of CO₂. To address the importance of this issue, Wang and coworkers reviewed the possibilities of enhancement of CO₂ capture by ILs via functionalization of the ILs.⁴ In this regard, by manipulating the structure of the cations and anions in the ILs, the amounts of CO₂ capture capacity could increase to 1 mol mol⁻¹ IL⁴ or at the highest report (to the best of our knowledge) this value was reported near to 1.25 mol mol⁻¹ IL.¹⁰

On the other hand, in the most cases the nature of interaction between the IL and CO₂ is not developed in molecular level. Investigations for the interactions of CO₂ with IL and absorption mechanism would give fruitful information about the CO₂ capture behavior and thus designing more conscious IL systems with improved activity. However, to the best of our knowledge only few examples were published which considered theoretical sides of the CO₂ absorption phenomenon using ionic liquids.¹¹ To address this issue, in an interesting example; Zhang et al. synthesized a novel dual amino-functionalized cation-tethered IL and studied the CO₂ absorption behaviors at various conditions as well as the absorption mechanism by means of density functional theory (DFT) calculations.¹²

Phenolic ILs are new class of compounds with improved absorption capacity up to 1 mol CO₂ per mol of IL. For example In 2012, some phenolic ILs were introduced for equimolar CO₂ capture.¹³

The current study aims to improve the absorption capacity of the phenolic ILs to 50% of their current values via the synthesis of a novel 1-(2-hydroxyethyl)-2,3-dimethylimidazolium phenoxide (IL 3). The IL is simply prepared with cheap and commercially available starting materials. DFT calculation with high level of calculations has been performed to elucidate the interactions of the carbon dioxide with IL 3 and give useful insight about the carbon dioxide absorption mechanism.

Results and discussion

Schematic illustration for the synthesis of the ILs is shown in Scheme 1. IL 1 was synthesis by one step reaction between 1,2-dimethylimidazole and 2-chloroethanol at 110 °C in toluene as solvent according to our previous reported procedure¹⁴ The anion exchange was performed in CH₂Cl₂ by using equimolar amounts of KOH.¹⁵ The resulted IL 2 was reacted with equimolar amounts of phenol to form IL 3 and water as a by-product of this neutralization reaction. The reason in which “2-methylated” imidazole was selected as a starting material is that it is significantly more stable than “2-H” imidazole under the applied basic condition. In the case of “2-H” imidazole, the IL is susceptible to form carbene under the basic media.¹⁶

Scheme 1 Schematic illustration for preparation of IL 3.

¹H NMR, ¹³C NMR and elemental analysis (EA) showed high purity of the synthesized IL (see Experimental section). Moreover, thermogravimetric analysis (TGA) showed that IL 3 was stable near to 250 °C and its decomposition was completed nearly at 350 °C (Fig. 1).

Fig. 1 TGA of IL 3.

After the preparation of the IL, the CO₂ uptake experiment was tested for IL 3 at ambient temperature (25 °C) and pressure with the gas flow rate of 100 ± 5 mL min⁻¹. The results of the absorption and desorption experiments are shown in Fig. 2.

Fig. 2 The results of the CO₂ absorption/desorption experiments.

As can be deduced from Fig. 2, the absorption reaches to its maximum value after 2 h. Before recording the weight of the IL 3 in each section, the sample was placed at the laboratory standard condition and under mild stirring to remove physisorbed carbon dioxide. For desorption, the mixture was heated at 45 °C by applying vacuum (100 mbar) for 15 min. The results of desorption experiment showed that the IL can readily release the captured carbon dioxide after 15 min (Fig. 2). For further studying desorption of carbon dioxide, the experiment was also examined under vacuum condition without applying heating and at room temperature (25 °C). In this condition, desorption took place only at relatively longer time (1 h). Moreover, at the solely thermal condition, the IL could release the CO₂ after 30 min. Consequently, to achieve effective desorption, heating the mixture should be accompanied by applying vacuum.

During the reaction, the viscosity of the IL 3 was gradually increased and the color of the liquid was slightly changed from reddish-brown to light brown. Fig. 3 shows the photograph of the CO₂-treated IL 3 during the reaction. Changing the color of the IL gives an interesting opportunity for “self-indicatoring” property of the IL 3 for CO₂ absorption reaction.

Fig. 3 Photograph of IL 3 during the CO₂ absorption experiment.

Regeneration of IL for CO₂ experiment is an important issue for gas separation. To investigate this issue, after the first run of absorption–desorption experiment, the recovered IL was subjected to another CO₂ uptake experiment. It was found that the IL 3 can be reused for at least eight cycles without noticeable loss of efficiency (Fig. 4). ¹H NMR, ¹³C NMR and FT-IR of the IL 3 after eight cycle of the CO₂ absorption are relatively identical and did not show significant difference compared to the initial IL. Worthy to note that according to our observations, the capacity for CO₂ uptake of IL 3 is 1.5 mol mol⁻¹ IL. A small deviation from this value may be attributed to the weight of very small amounts of the dissolved gas into the IL.

Fig. 4 Recycling results of IL 3 for CO₂ capture experiment.

FT-IR spectroscopy is a useful instrument for comparison about the changes of the structure of IL 3 before and after CO₂ capture experiment. In this regard two comparative FT-IR spectra for IL 3 were recorded before and after CO₂ absorption experiment (Fig. 5). For the spectrum which was recorded after CO₂ uptake, a characteristics band was observed at 1722 cm⁻¹ which is attributed to the C=O stretching frequencies of the carbonate moiety which was formed from the reaction of phenoxide ion of IL 3 and CO₂ during the absorption reaction.

Fig. 5 Comparative FT-IR spectra of IL 3 before and after CO₂ treatment.

¹³C NMR analysis is another useful method for screening the absorption of CO₂ by IL 3. The carbon atom frequency of carbonate in phenyl carbonate specie routinely observed between 155–170 ppm depending to the structure of the phenoxide based ILs. The spectrum of the IL 3 after CO₂ treatment, shows a new peak at δ = 168 ppm which belongs to the carbonate and verifies the successful CO₂ absorption (Fig. 6). This newly observed band is comparable to the previously values of the carbonate species of the phenolic ILs reported by Wang et al. (161 ppm)¹³ and Wu et al. (157 ppm).¹⁷

Fig. 6 ¹³C NMR of IL 3 before (a) and after (b) CO₂ uptake experiment (* = 168 ppm).

Phenolic ILs are among the most important compounds which have been utilized for CO₂ absorption.^13,17,18 Unlike the method which reported by Davis and co-workers,⁸ the absorption capacity of these compounds have been improved to uptake equimolar amount of CO₂. Our preliminary investigations for absorption property of IL 3 showed that each mol of IL can absorb 1.5 mol CO₂/mol IL which showed 50% improvement on the absorption capacity. Having this data in hand, we then tried to investigate the details of our observation for the absorption behavior for IL 3.

One of the most important parts is to find the active interacting sites of the IL. Hence, the structure of the IL 3 was optimized in the presence of two molecules of CO₂ which were placed in various positions around the IL and the structures were optimized at B3LYP-6-311++G** level of calculation. According to the optimized structure (Fig. 7), the carbon dioxide molecules have two distinct orientations around the IL 3.

Fig. 7 The structure of IL 3 with two carbon dioxide molecules optimized at B3LYP/6-311++G**. The distances are stated in Å.

As expected, the CO₂ (1) is moved toward the phenoxide ion of IL. The major tendency for this orientation is attributed to the negative charge of phenoxide and partially positive charge of the carbon dioxide. Moreover, CO₂ (1) is stabilized through a hydrogen bond with the –OH functionality of the IL. The ability of phenoxide ion for CO₂ capture was also proved by FT-IR and ¹³C NMR spectroscopy of the carbonate moiety which was formed during the CO₂ treatment experiment. The distance between phenoxide and CO₂ (1) is 2.547 Å. The mentioned approaching led to deviation of the linear structure of carbon dioxide with O–C–O bond angle 180° to a bent molecule with O–C–O bond angle of 172°. On the other hand, CO₂ (2) is oriented toward the –OH functionality of the IL 3. Again, the linear structure of CO₂ (2) is deviated to the O–C–O bent band with the angle of 177°. Consequently, the IL 3 is capable to absorb CO₂ molecules with two major possible sites which were shown in Fig. 7. Based on both experimental and computational investigations, we then tried to propose a reasonable mechanism for carbon dioxide absorption by IL 3 (Scheme 2).

Scheme 2 The proposed mechanism for CO₂ absorption of IL 3.

Our finding proved that when the phenolic IL 3 was exposure to CO₂ gas, phenyl carbonate species was formed. Two major possibilities can consider for the second site of IL (i.e. –OH) for capturing the CO₂ molecule. First, it could directly react and form “alkyl carbonic acid” specie via a subsequence proton transfer.¹⁹ However, no band related to the formation of alkyl carbonic acid was observed in ¹³C NMR analysis. This indicates that the –OH functionality of IL did not directly react with CO₂ to form IL-attached alkyl carbonic acid group.¹⁹ Accordingly, a mechanism for carbon dioxide absorption was proposed in Scheme 2. In this mechanism the phenoxide ion reacted with CO₂ molecules followed by “entrapping” or “complexation” of a CO₂ molecule between two hydroxyl moieties of two molecules of IL 3. The latter interaction is responsible for an additional 0.5 molar absorption of CO₂ by IL 3. To provide evidence about the ability of –OH groups for CO₂ absorption, IL 2 was subjected to carbon dioxide gas at identical conditions. The result indicated that it could uptake 0.54 mol CO₂ per mol of IL 2 which has very good agreement to the value based on the proposed mechanism.

Table 1 provides comparison of the capacity of some phenolic ILs with this work. As shown in Table 1, the maximum absorption of CO₂ per mol of IL is near to 1 mol per mol of IL. Although, the ILs based on trihexyl(-tetradecyl) phosphonium [P₆₆₆₁₄] cation could absorb CO₂ in relatively shorter reaction time. The electron donating groups on phenoxide ion (entries 3, 4 and 9) facilitate the absorption compared to electron withdrawing (entry 5) substituent. By increasing the reaction temperature, the absorption was significantly decreased which may be attributed to simultaneously desorption of CO₂.

Table 1 CO₂ absorption capacity of different phenolic ILs

Entry	Ionic liquid	Time [min]	Temp. [°C]	Capacity^a	Reference
a Mol CO2 per mol of IL.b MTBDH^+: protonated form of 7-methyl-1,5,7-triazabicyclo[4.4.0]dec-5-ene.c [P66614]: trihexyl(-tetradecyl) phosphonium IL with phenoxide (PhO) anion which have substitution in various positions.
1	IL 3	120	25	1.58	This work
2	[MTBDH^+][PhO^−]^b	300	23	0.49	18a
3	[P66614][4-Me-PhO]^c	30	30	0.91	13
4	[P66614][4-H-PhO]^c	30	30	0.85	13
5	[P66614][4-NO2-PhO]	20	30	0.30	13
6	[P66614][3-Cl-PhO]	30	30	0.72	13
7	[P66614][4-Cl-PhO]	30	50	0.65	13
8	[P66614][4-Cl-PhO]	30	70	0.50	13
9	[P66614][3-NMe2-PhO]	30	30	0.94	13

---

Conclusions

In summary, IL 3 was introduced as a simple, cheap and practical compound for CO₂ absorption. It was capable to uptake 1.5 equivalents CO₂ which to the best of our knowledge is the first report of improving the amounts of CO₂ uptake of the phenolic ILs to 50% of their current value. The possibility of the recycling was explored and found that it can recycle and reuse for eight times without losing the absorption efficiency. FT-IR spectrum of the IL 3@CO₂ showed a characteristic peak at 1722 cm⁻¹ which was belong to C=O stretching frequency of the phenyl carbonate moiety which forms during the CO₂ capture experiment. The presence of this specie was also proved by ¹³C NMR of IL 3 by observation a peak at δ = 168 ppm. DFT calculations showed that both phenoxide and –OH functionality are the active sites of the IL 3 for interaction with carbon dioxide. Based on the experimental finding and theoretical modeling, a mechanism for CO₂ absorption was proposed which indicates the participation of both covalent and non-covalent interactions for CO₂ absorption. Further investigations about the reaction mechanism are ongoing in our research group.

Experimental

Materials and equipments

All chemicals were supplied from Acros, Sigma-Aldrich, and Merck and used without further purification. ¹H NMR and ¹³C NMR spectra were recorded on the Bruker DRX 500 MHz. FT-IR spectra were recorded with ABB Bomem MB100 Fourier Transform Infrared Analyzer. Thermogravimetric analysis (TGA) was acquired under a nitrogen atmosphere with a TGA Q 50 thermogravimetric analyzer. Elemental analysis (EA) was performed utilizing Perkin Elmer, USA (2400, Series II).

Synthesis of the ionic liquids

To a round bottom flask equipped with a condenser, 1,2-dimethylimidazol (1 equiv.) and 2-chloroethanol (1.1 equiv.) were added and the mixture was refluxed in the super dry toluene for 24. Then, the mixture was decanted and the upper phase was removed. The residue was washed with dry toluene three times and dried in a vacuum oven overnight to produce IL 1. At the next stage, IL 1 was dissolved in CH₂Cl₂ and equimolar amount of powdered KOH was slowly added to this mixture at room temperature. After stirring for 24 h, the mixture was filtrated and the solution (IL 2 + CH₂Cl₂) was concentrated by rotary to afford IL 2. Then, IL 2 was dissolved in ethanol followed by gradually adding equimolar amount of phenol. The mixture was vigorously stirred for 24 h at room temperature. Finally, the water byproduct was removed by rotary and the resulted low-viscous oil was dried in vacuum oven in the presence of P₂O₅ as a dehydrating agent to afford IL 3 as reddish-brown oil. Spectra data: ¹H NMR (500 MHz DMSO-d₆): 2.55 (s, 3H), 3.64 (t, 2H), 3.70 (s, 3H), 4.16 (t, 2H), 5.44 (broad, 1H), 6.68 (t, 1H), 6.80 (d, 2H), 7.08 (t, 2H) and 7.64 (d. 2H); ¹³C NMR (125 MHz DMSO-d₆): 10.55, 35.55, 51.14, 60.48, 116.16, 119.51, 122.05, 122.95, 130.08, 145.59 and 158.37; elemental analysis (EA): calculated: C (66.64%), H (7.74%), N (11.96); found: C (65.60%), H (7.62), N (11.77).

CO₂ absorption experiment of IL 3

The absorption experiment was performed according to the previously reported method.²⁰ In a typical experiment, 7 g of IL 3 was added to a dry sample tube. Nitrogen was used as an inert gas before CO₂ treatment experiment. The purity of the utilized CO₂ and N₂ gases was determined to be 99.95%. Absorption was carried out at 25 °C for 2 h and at atmosphere pressure with a gas flow rate of 100 ± 5 mL min⁻¹. Desorption experiment was performed in a rotary evaporator at 45 °C by applying vacuum (≈100 mbar) for 15 min. During the CO₂ uptake experiment, the mixture of IL 3/CO₂ was stirred at 600 rpm to provide a uniform gas/liquid connection. The changes in the weight of the IL 3 was carefully recorded each 20 min with an analytical scale with accuracy of ±0.001 g.

Computational investigations

All of the structures were built using Spartan software package.²¹ The structures were optimized at B3LYP/6-311++G**. The frequencies of the optimized structures were checked to ensure that each structure was a real minimum. The lengths of hydrogen bonds are reported in angstrom (Å).

Notes and references

1. Overview of Greenhouse Gases, Carbon Dioxide Emissions, United States Environmental Protection Agency, http://www.epa.gov/climatechange/ghgemissions/gases/co2.html.
2. M. T. Ho, G. Leamon, G. W. Allinson and D. E. Wiley, Ind. Eng. Chem. Res., 2006, 45, 2546.
3. (a) R. Bounaceur, N. Lape, D. Roizard, C. Vallieres and E. Favre, Energy, 2006, 31, 2556; (b) M. B. Hagg and A. Lindbrathen, Ind. Eng. Chem. Res., 2005, 44, 7668.
4. C. Wang, X. Luo, X. Zhu, G. Cui, D. Jiang, D. Deng, H. Lia and S. Dai, RSC Adv., 2013, 3, 15518.
5. G. T. Rochelle, Science, 2009, 325, 1652.
6. B. R. Strazisar, R. R. Anderson and C. M. White, Energy Fuels, 2003, 17, 1034.
7. P. Wasserscheid and T. Welton, Ionic Liquid in Synthesis, Wiley-VCH, Weinheim, 2007.
8. E. D. Bates, R. D. Mayton, I. Ntai and J. H. Davis Jr, J. Am. Chem. Soc., 2002, 124, 926.
9. (a) Z. Yang, Y. Zhao and L. He, RSC Adv., 2011, 1, 545; (b) P. Sharma, S. D. Park, K. T. Park, S. C. Nam, S. K. Jeong, Y. I. Yoon and I. H. Baek, Chem. Eng. J., 2012, 193–194, 267.
10. B. F. Goodrich, J. C. de la Fuente, B. E. Gurkan, D. J. Zadigian, E. A. Price, Y. Huang and J. F. Brennecke, Ind. Eng. Chem. Res., 2011, 50, 111.
11. (a) J. Chen, W. Li, X. Li and H. Yu, Phys. Chem. Chem. Phys., 2012, 14, 4589; (b) R. Lo and B. Ganguly, New J. Chem., 2012, 36, 2549; (c) A. Liu, R. Ma, C. Song, Z. Yang, A. Yu, Y. Cai, L. He, Y. Zhao, B. Yu and Q. Song, Angew. Chem., Int. Ed., 2012, 51, 11306; (d) H. Tang and C. Wu, ChemSusChem., 2013, 6, 1050.
12. J. Zhang, C. Jia, H. Dong, J. Wang, X. Zhang and S. Zhang, Ind. Eng. Chem. Res., 2013, 52, 5835.
13. C. Wang, H. Luo, H. Li, X. Zhu, B. Yu and S. Dai, Chem.–Eur. J., 2012, 18, 2153.
14. M. Vafaeezadeh and M. Mahmoodi Hashemi, Chem. Eng. J., 2013, 221, 254.
15. B. C. Ranu and S. Banerjee, Org. Lett., 2005, 7, 3049.
16. A. K. L. Yuen, A. F. Masters and T. Maschmeyer, Catal. Today, 2013, 200, 9.
17. X. Zhang, K. Huang, S. Xia, Y. Chen, Y. Wu and X. Hu, Chem. Eng. J., 2015, 274, 30.
18. (a) C. Wang, H. Luo, D. Jiang, H. Li and S. Dai, Angew. Chem., Int. Ed., 2010, 49, 5978; (b) C. Wang, X. Luo, H. Luo, D. Jiang, H. Li and S. Dai, Angew. Chem., Int. Ed., 2011, 50, 4918.
19. T. S. Chamblee, R. R. Weikel, S. A. Nolen, C. L. Liotta and C. A. Eckert, Green Chem., 2004, 6, 382.
20. J. Aboudi and M. Vafaeezadeh, J. Adv. Res., 2014 DOI:10.1016/j.jare.2014.02.001.
21. Spartan 10V102, Wavefunction Inc., Irvine, CA, 2010.

---