Development of Borate-Based Basic Ionic Liquid for Room Temperature CO₂ Capture

Abstract

As part of CO₂ capture strategies, ionic liquid-based CO₂ absorbents have gained attention for their tunable properties to lower the energy costs for CO₂ capture. In this study, a series of borate-based nonamine functionalized ionic liquids (ILs), incorporated with magnesium acetylacetonate, were developed and investigated for its CO₂ capture capability at moderate temperature under ambient pressure. Nuclear magnetic resonance and Fourier transform infrared spectroscopy confirmed the successful incorporation of acetylacetonate ligands into the fluorinated-lithium borate ionic liquids. Comprehensive analyses of the physical and thermochemical properties revealed that the synthesized ILs remain stable below 200 °C, with the borate structure and acetylacetonate ligands intact. The ILs functionalized with fluorinated alcohol and magnesium acetylacetonate enhance the CO₂ uptake capacity by 55% in comparison with the original lithium borate ILs, suggesting the enhanced cooperative interactions responsible for improved CO₂ capture performance. The carbon capture mechanism was identified to proceed via physical absorption, as evidenced by minimal changes in the characterization results and viscosity after CO₂ absorption. The enthalpy of CO₂ absorption (ΔH _a) for the synthesized ILs were determined experimentally by using differential scanning calorimetry to be in the range from −12.4 kJ mol^–1 to −18.9 kJ mol^–1, which are much lower than that of conventional amine solutions (e.g., MEA: −82 kJ mol^–1) and amine-based ILs ((e.g., [Bmim]­[Ac]: −45.8 kJ mol^–1). These findings suggest that lithium borate-acetylacetonate ILs offer a promising approach for a CO₂ capture system under ambient conditions.

1. Introduction

The emissions of increasing amounts of greenhouse gases have contributed to ongoing global warming issue, with carbon dioxide (CO₂) being the most critical substance. According to the annual report of the National Oceanic and Atmospheric Administration (NOAA), the global average concentration of CO₂ reached a record high of 422.8 ppm in 2024. If no substantial actions are implemented to reduce emission of CO₂, it is projected that the global temperature could increase significantly beyond 2 °C by the end of the century. − Carbon dioxide capture, utilization, and storage (CCUS) technologies have been developed to have the potential to reduce the CO₂ emissions by 20%. , Amine-based solvents are the most commonly used in CCUS postcombustion CO₂ capture method due to their high reactivity and thermal stability. , However, the high energy requirement by the solvent regeneration, equipment corrosion issues, solvent degradation, environmental hazards, and high capital costs limit the applications. − Therefore, this has prompted the development of alternative absorbents with lower energy requirements to address these limitations.

CO₂ capture through the usage of ionic liquids (ILs) has gained increasing attentions, with numerous publications investigating CO₂ absorption using various types of ILs. − ILs are organic salts that composed entirely of cations and anions that exist in the liquid state at or near room temperature, with melting point below 100 °C. − Their low melting points arise from the use of bulky, asymmetric cations and weakly coordinating anions, which disrupt crystal packing and allow the salt to remain in liquid. , Compared to conventional aqueous amine-based solvents such as monoethanolamine (MEA), which rely on chemisorption through strong acid–base reactions, ILs offer several distinct advantages as absorbents of CO₂, including low saturated vapor pressure, a wide temperature range, excellent thermal and chemical stability, and high tunability with the ease of changing cation and anion combinations, catering to the specific requirements by the applications. − Unlike MEA or the solid adsorbents, such as amine-functionalized mesoporous silica and CaO, , which requires high regeneration energy due to strong CO₂ binding, many ILs promote weaker, reversible interactions with CO₂. The absorption enthalpy of CO₂ with ILs can vary widely, ranging from weak to strong interactions depending on their chemical structure and absorption mechanisms, which may involve physical, chemical, or a combination of both. These advantages make ILs promising alternatives for efficient CO₂ capture.

ILs can be categorized into two different types: (1) amino-functionalized and (2) nonamine-based ILs. The ILs consist of different combinations of cations and anions, with imidazolium-based being the most extensively studied as cations, followed by ammonium, phosphonium, piperidinium, and pyrrolidinium. − These cations are paired with a wide variety of anions, such as fluorinated-based (i.e., [BF₄]⁻, [PF₆]⁻ and [TfO]⁻), halides (i.e., [Cl]⁻, [Br]⁻ and [I]⁻), and amino acid-based anions (i.e., [Gly], [Lys] and [Ala]). Orhan et al. investigated the effects of various cation–anion combinations and found that the imidazolium-based IL [Bmim]­[Ac] exhibited the highest CO₂ equilibrium loading of 0.89 mol CO₂ per mol amine, followed by [Bmim]­[Tf₂N] with 0.77 mol CO₂ per mol amine at 2 bar.

To enhance the CO₂ absorption capacity at mild conditions, amine functional groups have been incorporated into IL structures. ,, The amine groups can be integrated into the cation (cation-functionalized IL), the anion (anion-functionalized IL), or both (dual-functionalized IL) in the ILs. This strategic modification introduces active sites that enhance CO₂ interaction, selectivity, and absorption kinetics while maintaining recyclability. − The anion of an IL has been identified as a critical factor in determining CO₂ absorption performance, as CO₂ solubility is strongly influenced by the nature and reactivity of the anion. − While cation-functionalized ILs can chemically bind CO₂, they often exhibit a reaction stoichiometry of 1:2 for CO₂ absorption which may limit process efficiency and increase solvent consumption. In contrast, anion functionalized with amine groups has demonstrated improved stoichiometry to 1:1 CO₂: IL molar ratio. Gurkan et al. showed that ILs based on trihexyl­(tetradecyl)­phosphonium [P₆₆₆₁₄] as the cation paired with amino acid-derived anions ([Gly]⁻, [Lys]⁻ and [Met]⁻) have achieved this stoichiometry ratio, with a CO₂ uptake of 1.26, 1.60, 0.90 mol CO₂ per mol IL respectively. ,, These results presented the dominant role of the anion in enhancing the CO₂ capture capacity. However, the chemisorption of CO₂ by amine-containing ILs typically leads to a significant increase in viscosity. For instance, the viscosity of [P₆₆₆₁₄]­[Gly] increased from 360 to 17000 cP upon CO₂ uptake. The drastic rise in viscosity is attributed to the complexed formation of a hydrogen-bonded network between the IL and CO₂. Despite the remarkable CO₂ absorption performance of amine-functionalized ILs, the desorption process requires substantial energy due to strong chemical bonding, byproduct formation, and high viscosity.

To overcome these issues, recent studies have explored whether nonamine functionalized ILs can achieve efficient CO₂ capture while avoiding the formation of hydrogen-bonded networks. Imidazolium-based ILs such as [Bmim]­[PF₆], and [C_nmim]­[Tf₂N] primarily undergo physical absorption. Blanchard et al. reported [Bmim]­[PF₆] exhibited 0.19 mol CO₂ per mol IL at 1 bar, with physisorption interaction between the IL and CO₂. The exploration of high-pressure conditions to improve CO₂ solubility based on Henry’s Law were further investigated by studying the phase behavior of CO₂ in [Bmim]­[PF₆] under varying pressure conditions. ,,− Their findings indicated that the IL exhibited a CO₂ solubility of 0.231 mol of CO₂ per mol of IL at 15.17 bar, which significantly increased to 0.729 mol of CO₂ per mol of IL at 95.67 bar. Similarly, Shin et al. found that [C₂mim]­[Tf₂N] absorbed 0.24 mol CO₂ per mol IL at 1 bar, compared to 0.30 mol CO₂ per mol IL for [C₈mim]­[Tf₂N], indicating that longer alkyl chains enhance uptake. Both ILs exhibited sharp increases at elevated pressure, reaching 0.761 mol CO₂ per mol IL and 0.845 mol CO₂ per mol IL, respectively.

In addition to physical absorption, some nonamine ILs undergo chemical absorption. Dai et al. reported that superbase-derived protic IL, [MTBD]­[Im], exhibited rapid and reversible CO₂ uptake with equimolar absorption capacity. In a subsequent study of anion reactivity, the phosphonium hydroxide-derived ILs were shown to allow fine-tuning of CO₂ absorption efficiency through the selection of weak proton donors with different pK _a values. For instance, the absorption capacities of [P₆₆₆₁₄]­[Pyrr], [P₆₆₆₁₄]­[Im], [P₆₆₆₁₄]­[Ind], [P₆₆₆₁₄]­[Triz], [P₆₆₆₁₄]­[BenTriz], and [P₆₆₆₁₄]­[Tetz] were 1.02, 1.00, 0.98, 0.95, 0.17, and 0.08 mol of CO₂ per mole of IL, respectively. Importantly, these ILs maintained low viscosity during absorption, supporting rapid mass transfer and offering the potential for scalable CO₂ capture applications.

Thermodynamic comparisons further illustrate the key differences. Amine-functinalized IL, such as [P₆₆₆₁₄]­[Met], has an enthalpy of −64 kJ mol^–1, whereas nonamine ILs undergoing chemical absorption ([P₆₆₆₁₄]­[Pyrr]: −91 kJ mol^–1 ;[P₆₆₆₁₄]­[Im]: −89.9 kJ mol^–1) display even stronger interactions. In contrast, [C₂mim]­[Tf₂N], a representative of nonamine ILs that undergoes physical absorption, is around −7.81 kJ mol^–1. Physical absorption is governed mainly by van der Waals forces and allows for easier desorption due to the relatively low energy needed to break these interactions. Consequently, these findings suggest that nonamine functionalized ILs has emerged as an alternative energy to address the issues of viscosity and energy requirements.

In this work, the focus is the development of a new type of nonamine functionalized ILs and testing for their potential capability for low energy CO₂ capture. Harada et al. first proposed that lithium sodium borate (Li_0.5Na_0.5)₃BO₃ molten salt can achieve a high CO₂ capacity of 7.3 mmol g^–1 in 1 min of reaction with CO₂ because of the fast reaction kinetics through enhanced ion mobility and mass transport, which allows CO₂ molecules to diffuse throughout the material. Meanwhile, Guzman Gonzalez et al. synthesized a series of lithium-based ILs for the applications as single-component electrolytes in lithium batteries. This study focuses on the synthesis and characterization of lithium borate-based ILs, including two novel nonamines functionalized ILs, alongside comparative evaluation with the reference ILs. The CO₂ absorption capacity was evaluated at ambient temperature to investigate the feasibility of the material. In addition, we examined key material properties, including viscosity and thermal stability, as they are critical factors influencing the CO₂ absorption performance. With the developed ILs, the energy consumption could be significantly reduced, with operating absorption temperature at room temperature, unlike molten salts that require high temperatures for absorption.

2. Experimental Section

2.1. Materials

Hexane (≥96%), anhydrous magnesium sulfate (MgSO₄, ≥ 98%), 2,2,2-trifluoroethanol (≥98%), and tetrahydrofuran (THF, ≥ 99.5%) were purchased from FUJIFILM Wako Pure Chemical Co., Ltd., Japan. Triethylene glycol monomethyl ether (≥98%), n-butyl lithium (ca. 15% in hexane, ca. 1.6 mol/L), and diethyl ether (≥99.5%) were purchased from Tokyo Chemical Industry (TCI) Co., Ltd., Japan. Borane tetrahydrofuran complex solution (BH₃-THF, 1.0 M in THF) and magnesium acetylacetonate (Mg­(acac)₂, ≥ 98%) were purchased from Sigma-Aldrich. All of the above chemicals except hexane were used without further purification. Hexane was dried with anhydrous MgSO₄ before synthesis. The experiments utilized deionized water. N₂ (≥99.995%) and CO₂ (≥99.5%) were supplied by Suzuki Shokan Japan.

2.2. Synthesis Procedure

Four different types of lithium borate-based ILs were prepared with variation with the inclusion of fluorinated alcohol (FA) or Mg­(acac)₂. The chemical structures of the ILs synthesized in this study are illustrated as shown in Figure S1, with the reference lithium borate IL labeled as LBIL, lithium borate IL with FA labeled as LBIL-FA, lithium borate IL with only Mg­(acac)₂ labeled as LBIL-Ac, and lithium borate IL with both FA and Mg­(acac)₂ labeled as LBIL-FA-Ac. The reference LBIL and LBIL-FA were previously reported by Guzmán-González et al.

2.2.1. Synthesis of LBILs

A previously reported methodology was used for the synthesis of the ILs. Five ml of dried hexane and triethylene glycol monomethyl ether (10 mmol, 3.6 mL) were added into a 100 mL three-neck flask. The system was purged with a N₂ flow during the reaction. The solution was stirred and subsequently cooled in a liquid N₂ bath, preventing solidification of the system. BH₃-THF (5 mmol, 5 mL) was carefully added dropwise as the reaction is highly exothermic. The reaction mixture was then slowly warmed to room temperature and stirred for 30 min. Afterward, the reaction mixture was cooled again in a liquid N₂ bath. Subsequently, 2,2,2-trifluoroethanol (5 mmol, 3.6 mL) as FA (R^F–OH), was added dropwise, and the system was warmed to room temperature again for 1 h. After the reaction, it was cooled again in a liquid N₂ bath and carefully added 5 mmol of n-butyl lithium (R^C-Li). The reaction mixture was stirred at room temperature for another 2 h. Finally, the mixture was washed with cold diethyl ether and placed in a rotary evaporator at 40 °C for 24 h under vacuum atmosphere. It was further dried in an oil bath at 50 °C for 24 h under vacuum. The obtained final product will be present as a transparent liquid. The reference ionic liquid LBIL was synthesized using the same synthesis method without the addition step of FA. The synthetic route is shown in Figure .

1.

Synthetic route for reference LBIL and LBIL-FA.

2.2.2. Synthesis of LBILs with Mg­(acac)₂

The same synthesis method will be used for synthesizing LBIL-Ac and the LBIL-FA-Ac compound. The addition of Mg­(acac)₂ was incorporated during the IL synthesis, after the reaction with the organolithium compound. 0.15 g of Mg­(acac)₂ powder was stirred and dissolved in 5 mL of THF. Afterward, 1 mL of Mg­(acac)₂ in THF solution was slowly added into the reaction mixture and mixed for 2 h before proceeding for washing and drying steps. Yellowish liquid samples were obtained as the final products. The synthetic route for addition of Mg­(acac)₂ is shown in Figure .

2.

Synthetic route for references LBIL-Ac and LBIL-FA-Ac.

2.3. Material Characterization and Physicochemical Properties of Ionic Liquid

2.3.1. Characterization

The series of lithium-borate ILs were characterized using ¹H and ¹³C Nuclear Magnetic Resonance (NMR) (Avance III 400 MHz, Bruker) and Fourier transform infrared spectroscopy (FT-IR) (Nicolet iS20, Thermo Fischer Scientific). Chloroform-d (≥99.6%, TCI) was used as the NMR solvent for all the samples tested. Infrared spectra were collected using the KBr pellet method with data collected over 64 scans at wavelengths ranging from 400 to 4000 cm^–1 and a spectral resolution of 4 cm^–1.

2.3.2. Viscosity

The viscosity of the samples was analyzed by using a modular compact rheometer (MCR302e, Anton Paar). The viscosity measurements were conducted at 20 °C, and the shear rates were varied from 1 to 1000 s^–1 to capture the viscosity behavior.

2.3.3. Thermal Properties

The thermal onset temperature (T _onset) and decomposition temperature (T _decomp) were measured using a thermogravimetric analyzer (TGA-550, TA Instruments). Average weights of 5–10 mg for every sample were placed on a platinum sample pan for testing with a gas flow rate of 100 mL min^–1. The sample was first pretreated to remove any residual solvents and moisture present at 70 °C for 2 h under 100% N₂ gas before starting the actual analysis. For determination of T _onset and T _decomp, the system was ramped from 30 to 350 °C at a ramping rate of 5 °C/min under N₂ atmosphere. The experimental temperatures of thermal decomposition are presented in terms of weight loss (%) and temperature (°C). T _onset was obtained from the intersection of the baseline weight and the tangent of the weight dependence on the temperature curve as decomposition occurs.

Absorption enthalpy (ΔH _a) were determined through differential scanning calorimetry (DSC) using a simultaneous DSC/thermogravimetric analyzer (SDT 650, TA Instruments). One to two droplets of IL sample were placed on a platinum sample pan for the testing with a gas flow rate of 100 mL min^–1, with the same pretreatment procedure. CO₂ was performed at 30 °C under 100% CO₂ for 60 min.

2.3.4. CO₂ Capture

CO₂ capture performance was evaluated through thermogravimetric measurements, which were conducted on a platinum sample pan using a thermogravimetric analyzer (TGA-550, TA Instruments) with a gas flow rate of 100 mL min^–1. Approximately 10 mg of sample underwent the same pretreatment procedure as the analysis of thermal properties. For CO₂ uptake measurements, the sample was exposed to 100% CO₂, with the absorption temperature performed at 30 °C for 60 min. Desorption of the residual was followed by an additional 60 min at 55 °C under 100 N₂ flow.

3. Results and Discussion

3.1. Role of Magnesium Acetylacetonate in CO₂ Absorption

Magnesium acetylacetonate was chosen to be incorporated into the lithium borate-based ILs due to its possible ability to enhance CO₂ absorption through increased active sites to capture CO₂ uptake and improved chemical stability. Acetylacetonate is a bidentate ligand derived from acetyl acetone and forms stable octahedral complexes through both oxygen atoms that create a six-membered chelate ring that supports the stability and reactivity of the metal complex. One of the key characteristics of acetylacetonate is keto–enol tautomerization. Tautomerization involves the transfer of a hydrogen atom from one molecular site to another, leading to changes in the molecular skeleton, electron density distribution, and chemical properties. − The enolate form of acetylacetonate has negatively charged oxygen and nucleophilic carbon, both of which can interact with the electrophilic carbon in CO₂, forming a new C–C bond. , The ligand exists in equilibrium between its keto and enol forms, with the enol form being stabilized by intramolecular hydrogen bonding. The nucleophilic hydroxyl group (−OH) and the adjacent electron-rich CC bond make the enol form particularly reactive toward CO₂, enabling hydrogen bonding interactions between the −OH group and CO₂. , With the cooperative interaction between the magnesium center and acetylacetonate ligand, Mg²⁺ acts as a Lewis acid that accepts the electrons and coordinates with CO₂, where CO₂ donates electrons from the oxygen atoms or the π-electrons in the system. Meanwhile, the acetylacetonate ligands function as Lewis base, contributing to a CO₂ “push-pull” mechanism of electron density, facilitating electron redistribution and promoting efficient physisorption.

The hypothesized reaction mechanism for the incorporation of acetylacetonate ligands involves substitution of the n-butyl group on the boron atom with the acetylacetonate ligand, forming a new boron–oxygen bond. Depending on steric and electronic factors, the ligand may coordinate to boron in a monodentate form via one oxygen atom. The main benefit of incorporation of magnesium acetylacetonate is that not only increases the number of active sites for CO₂ capture but also enhances the material’s stability and tunability, as the ligand structure can be modified by varying the cation and anion. ,

3.2. Characterization

Based on the NMR results, significant changes in the chemical shifts were observed after incorporation with magnesium acetylacetonate (Figures S2–S9, Table ). A new chemical shift appeared at 2.02–2.05 ppm, and 2.66–2.73 ppm, which are attributable to the acetylacetonate ligand. The 2.02–2.05 ppm region is consistent with the methyl group attached to OCH₃ of the acetylacetonate ligand. The peak signals at 2.66–2.73 ppm correspond to the protons  to the carbonyl group (CO) of the conjugated acetylacetonate ligand. This downfield shift arises from the deshielding effect of the adjacent CO group, which withdraws electron density from the alpha-protons, causing their signal to appear at a higher chemical shift. Additionally, the chemical shifts of the BOCH₂CF₃ were observed to have an upfield shift from 3.81 to 3.73 ppm upon magnesium acetylacetonate incorporation. This change can be explained by a coordination induced redistribution of electron density, which was reported by Butera et al. on how the metal ion coordination affects the H NMR chemical shift values. The electron distribution at borate-ether oxygen atoms alters the polarization environment experienced by the CH₂ protons, which resulted in slight changes in chemical shift. In the ¹³C NMR spectrum, the disappearance of the signals between 10 and 20 ppm after incorporation indicates that the alkyl chain of the ionic liquid is no longer present in the product.

1. Chemical Shift Assignment of the ¹H NMR Spectra of LBILs.

LBIL/LBIL-FA	δ (ppm)	LBIL-Ac/LBIL-FA-Ac	δ (ppm)
Terminal-CH3	0.918–0.955	Terminal-CH3	0.882–0.956
CH2	1.256–1.556	CH2	1.256–1.559
OCH3 (triethylene glycolmethyl ether)	3.388	OCH3 (acetylacetonate)	2.020–2.050
(OCH2CH2)3O	3.550–3.725	α-CH2	2.660–2.730
BOCH2CF3	3.809–3.833	OCH3 (triethylene glycolmethyl ether	3.388
		Overlapping (OCH2CH2)3O and BOCH2CF3	3.560–3.720

The FTIR spectrum of the synthesized ionic liquids confirms their chemical structure, with key vibrational bands at 2880, 1456, 1350, 1283, 1104 cm^–1 assigned to the triethylene glycol monomethyl ether chain and borate core − (Figure , Table ). A broad feature near 1653 to 1658 cm^–1 was present in all the samples and is consistent with the residual water rather than the structural vibrations of the ILs reported in other literature. , Both LBIL and LBIL-FA samples exhibited strong peak signals at 1283 cm^–1, which is attributed primarily to C–O stretching vibrations of the borate-ether framework. After the incorporation of Mg­(acac)₂, the intensity of this band significantly reduced in LBIL-Ac and LBIL-FA-Ac, consistent with coordination of Mg²⁺ to the oxygen atoms of the ether chains, which reduces the vibration activity of C–O groups. Furthermore, an additional peak appeared at 1616 cm^–1 in LBIL-Ac and LBIL-FA-Ac samples, likely due to the acetylacetonate ligand. The stability properties of the metal complex arises from the conjugated π system of acetylacetonate ligand. The absence of distinct CO stretching frequency, typically found around 1700 cm^–1 in the FTIR spectra is a key indicator of delocalized structure. , The presence of a peak at 1616 cm^–1 is characteristic of a chelated acetylacetonate ligand and does not correspond to a discrete CO or CC stretching vibration. − Instead, its position and intensity are consistent with a coupled vibrational mode of the chelated acetylacetonate ligand involving the conjugated O–C–C–C–O system. The CC bond within the acetylacetonate ligand represents the delocalization of π electrons across the O–C–C–C–O chain.

3.

FT-IR spectra of LBILs under N₂ at 25 °C.

2. FT-IR Peak Assignment of LBILs.

FT-IR peaks (cm –1 )	Assignments	FT-IR peaks (cm –1 )	Assignments
3396	v(O–H)	1283	v(C–O)
2880	v(C–H)	1248	v(C–F), v(C–O)
1653–1658	δ(H–O–H), v as(CO3 2–)	1200	v(B–O), v(C–O)
1616	v(CC), v(CO)	1103–1105	v(B–O), v(C–O–C)
1455–1457	δ(CH2), δ(CH3)	1028	v(C–O)
1350	δ(CH3), v(B–O)

3.3. Viscosity

Viscosity is one of the important parameters for ILs as it significantly influences the CO₂ mass transfer efficiency and the energy required for transport and regeneration. Lower viscosity facilitates faster mass transfer, allowing the CO₂ molecules to diffuse more rapidly within the liquid, thereby enhancing absorption rates. In contrast, high viscosity ILs hinder CO₂ diffusion, creating slower absorption kinetics and a lower overall efficiency. The viscosities of the LBILs samples were measured across various shear rates, as shown in Figure . The average viscosity of LBIL and LBIL-FA was 47.37 mPa.s and 33.62 mPa.s, respectively, indicating that the inclusion of FA has reduced the viscosity. Yang et al. has reported 2,2,2-trifluoroethanol can reduce the viscosity of ILs. Furthermore, further reduction of the viscosity was observed in LBIL-FA-Ac (12.39 mPa.s) with the inclusion of both FA and Mg­(acac)₂. The acetylacetonate ligand from Mg­(acac)₂ has replaced the hydrocarbon groups present in the IL, which disrupted the hydrophobic interactions and made the structure less bullky. The effect is conceptually similar to how Pd­(acac)₂ reduces heavy oil viscosity by breaking down the large hydrocarbon molecules. Nonetheless, these ILs samples exhibited a low viscosity comparable to conventional solvents and other ionic liquids reported in the literature (Figure S10).

4.

Viscosity of LBILs samples across the shear rate from 1 to 1000 s^–1.

3.4. Thermal Properties

Determining the decomposition temperature of ILs is essential to establishing suitable operating conditions for CO₂ absorption. Typically, the thermal stability of ILs is characterized by the onset temperature, T _onset, which is determined from the intersection of the baseline weight and the tangent to the weight-loss curve during decomposition. , However, ILs were found to degrade at a temperature significantly lower than the T _onset, making it crucial to assess whether the material can withstand the temperature required for CO₂ desorption and regeneration. , If decomposition occurs at or below the operating temperature, then the material would be unsuitable for prolonged use. The TGA curve observed two distinct stages of weight loss. The first stage shows gradual weight reduction between 105 and 220 °C, attributed to evaporation of residual solvent and moisture, together with the onset of ether chain and acetylacetonate ligand degradation. This is consistent with reports showing that the ether chains and the chelate rings of acetylacetonate degrading around 200 °C. , Farooq et al. also have synthesized a series of acetylacetonate-based magnetic ionic liquids (MILs) and tested that the thermal stabilities of these MILs were at the lower range of 130 to 215 °C. The second stage corresponds to the major decomposition step, where there is complete breakdown of the ionic liquid structure. T _onset was determined to be around 240 °C, while the decomposition temperature occurred at 275 °C (Figure a).

5.

Decomposition results for LBIL-FA-Ac measured from a starting temperature of 25 to 350 °C at a heating rate of 5 °C/min under 100% N₂, with determination of (a) T _onset and T _decomp; (b) FT-IR spectra of LBIL-FA-Ac for a thermal analysis test at a temperature from 25 to 350 °C under 100% N₂ flow.

To investigate the structural changes during heating, FT-IR was performed under N₂ at temperature varied from 25 to 300 °C. The IR spectra of LBIL-FA-Ac revealed a gradual decrease in peak intensity as temperature increased (Figure b). The intensity of the characteristic CC bond associated with Mg­(acac)₂ at 1610 cm^–1 was gradually decreased as temperature increased and completely disappeared above 200 °C. The peak at 1100 cm^–1, associated with ether chains and acetylacetonate, also decreases with increasing temperature, likely possible ether chain breakdown and acetylacetonate ligand decomposition. These FT-IR results align well with the staged decomposition observed in TGA, suggesting that the IL maintains structural integrity only at temperatures below 200 °C, where the borate complex and acetylacetonate remain stable.

3.5. Absorption Enthalpy

The enthalpy of the CO₂ absorption provides insights into the nature of interactions between the IL and CO₂. DSC revealed absorption enthalpy values in the range −12.4 kJ mol^–1 to −18.9 kJ mol^–1 for the LBILs synthesized in this study, with the highest value of −18.9 kJ mol^–1 for LBIL-FA-Ac (Figure S11). The values are comparable to those reported for other ILs, such as [Bmim]­[BF₄] (−15.9 kJ mol^–1), as shown in Figure S12. ,

3.6. CO₂ Capture Performance

3.6.1. Effects of Mg­(acac)₂ and Fluorinated Alcohols

The CO₂ absorption–desorption profiles of the LBILs samples reveal clear differences in the working capacity (Figure ). The results demonstrated that LBIL-FA-Ac containing both FA and Mg­(acac)₂, exhibited the highest absorption capacity of 0.9375 mmol g^–1 in 60 min of absorption time. This demonstrates that the combination of FA and Mg­(acac)₂ enhances the level of CO₂ uptake. Comparing LBIL and LBIL-FA-Ac, the combination resulted in a 55% increase in the CO₂ absorption capacity. The introduction of Mg­(acac)₂ significantly enhanced the CO₂ uptake capacity. According to the findings discovered by Sampson et al, Mg²⁺ plays a critical role in stabilizing electron-rich [CO₂]⁻ intermediates and facilitating C–O bond cleavage. This behavior is attributed to the metal–ligand cooperativity, where intramolecular interactions between a metal center and a nearby ligand site work synergistically for CO₂ activation. , Such systems are described as offering a “push-pull” mechanism, with Lewis acidic metal centers stabilizing the negative charge on CO₂. , In addition, it was reported that substituents that decrease ligand basicity favors CO₂ capture. Therefore, the Mg²⁺ center may act as a Lewis acid, providing additional coordination sites to interact with the electron-rich oxygen of CO₂ within the IL. This coordination likely promotes physisorption within the IL matrix by stabilizing the presence of CO₂ through donor–acceptor interactions. Similarly, the effect of FA on the absorption of CO₂ can also be assessed by comparing LBIL with LBIL-FA and LBIL-Ac with LBIL-FA-Ac. The presence of FA moderately increased the rate of CO₂ uptake. Hazra et al. reported that materials with interconnected discrete pores embedded with −CF₃ functionality exhibit strong CO₂-MOF interaction. The surface functionalized with −CF₃ groups created hydrogen bonding and Lewis acid–base interactions with the fluorine atom of −CF₃ group. Based on these reports, the moderate enhancement of uptakes is likely due to the presence of the trifluoromethyl (−CF₃) group, which can help with the facilitation of CO₂ dissolution and interactions. During the desorption phase, all samples exhibited a decline in CO₂ uptake, with LBIL-FA-Ac also demonstrating the fastest release. The working capacity of LBIL-FA-Ac remained the highest among the samples, with minimal residual, while LBIL still retained some absorbed CO₂ after 60 min of desorption. Additionally, the mass loss during the desorption for LBIL-FA-Ac was approximately only 1.5%, suggesting good thermal stability of the IL.

6.

CO₂ absorption–desorption results for the LBIL samples. Absorption test was performed at 30 °C under 100% CO₂ and desorption was performed at 55 °C under 100% N₂.

The CO₂ uptake capacity achieved for LBIL-FA-Ac (0.94 mmol g^–1) is higher than that of several nonamine functionalized ILs reported in other literature (Figure S13), such as [C₂min]­[Tf₂N] (0.61 mmol g^–1) and [Bmim]­[BF₄] (0.42 mmol g^–1). , Notably, this value is comparable to the CO₂ uptake of these ILs measured under higher pressures, with [Bmin]­[Tf₂N] (0.90 mmol g^–1) and [Bmim]­[BF₄] (0.80 mmol g^–1), demonstrating that LBIL-FA-Ac achieves superior CO₂ capture performance under mild ambient conditions.

3.6.2. Effects of Concentration of Mg­(acac)₂

To examine the effect of concentration of Mg­(acac)₂ in LBIL-FA-Ac samples on CO₂ uptake capacity, the concentration was varied from 0.03 to 0.15 g of Mg­(acac)₂ (Figure ). LBILs samples of 0.03 and 0.05 g Mg­(acac)₂ were observed to reach a maximum working capacity of approximately 0.90 mmol g^–1 at 60 min absorption time. However, further increasing the Mg­(acac)₂ concentration to 0.15 g resulted in a decline in uptake capacity. Goodwin et al. found that high concentrations of metal salts in ionic liquids can undergo ion aggregation and network formation beyond a critical gelation threshold. In such regimes, the formation of percolating ion networks and extended reduces the availability of free, mobile metal centers capable of coordination. , Since Mg²⁺ plays a central role in providing the coordination sites for CO₂, oversaturation of Mg­(acac)₂ may disrupt the IL structure with fewer Lewis acid sites, making them inaccessible for CO₂ interactions environments, which could explain the observed decline in CO₂ uptake beyond a concentration of 0.05 g Mg­(acac)₂.

7.

CO₂ absorption results for LBIL-FA-Ac samples with various concentrations of Mg­(acac)₂ per mL of THF from 0.03 to 0.15 g. Absorption test was performed at 30 °C under 100% CO₂.

3.6.3. Pressure Dependence of CO₂ Capture Performance

The CO₂ capture performance of LBIL-FA-Ac was evaluated at pressures ranging from 0.05 to 1 bar to assess the uptake capacity at reduced pressures (Figure ). The CO₂ uptake capacity exhibits a clear pressure-dependent trend. The absorption capacity increases as the partial pressure of CO₂ rises, from 0.26 mmol g^–1 at 0.05 bar to 0.9375 mmol g^–1 at 1 bar. Notably, at 0.15 bar, which is typically the CO₂ partial pressure in postcombustion flue gas, was able to achieve CO₂ uptake capacity of 0.60 mmol g^–1. Comparing with other imidazolium-based ILs reported in the literature, such as [Bmim]­[Tf₂N] (0.22 mmol g^–1 at 4.2 bar), [Bmim]­[Pf₆] (0.39 mmol g^–1 at 5.29 bar), [Emim]­[Tf₂N] (0.13 mmol g^–1 at 2.13 bar), the LBIL-FA-Ac IL demonstrates significantly higher CO₂ uptake capacity at muchs lower pressures. , This superior low-pressure performance highlights its potential for industrial applications in which CO₂ is present at partial pressures well below ambient conditions.

8.

CO₂ absorption test at various pressures ranging from 0.05 to 1 bar with absorption temperature at 30 °C for 60 min.

3.6.4. Performance of CO₂ Cycle Stability

The CO₂ uptake capacity of the material was evaluated over five consecutive cycles to evaluate its reusability (Figure ). The results show a steady decline with each subsequent cycle, reaching 0.389 mmol g^–1 by the fifth cycle. This represents almost 60% loss in capacity from the initial value. The significant decrease in the CO₂ uptake capacity indicates that the material has poor reusability and long-term stability under the tested conditions. This suggests that the material might undergo physical degradation of the material or blockage of active sites during the repeated absorption–desorption cycles. The reduced reusability of LBIL-FA-Ac is likely attributed to minor evaporation losses during the CO₂ cycle test. Although pure ionic liquids are generally nonvolatile, functionalized or blended IL systems can exhibit gradual mass loss due to evaporation of volatile cocomponents or loosely bound ligands. During the desorption process at 55 °C, slight evaporation or partial loss of Mg­(acac)₂ associated species may occur. Since Mg­(acac)₂ plays a crucial role in CO₂ uptake capacity, this could be the observed decline in CO₂ uptake upon regeneration. Similar behavior was observed and reported by Taylor et al., who discussed the effect of evaporation of the volatile components in the IL-based systems, which has affected its CO₂ uptake capacity. Their study demonstrated that the IL-glycol mixtures suffer from significant volatility and encapsulation of the mixtures have mitigated evaporation losses by around 40%, and able to maintain CO₂ capture capacity. Encapsulating the LBIL-FA-Ac system could be a promising strategy to improve its reusability to mitigate the evaporation of crucial parts of the ionic liquids systems. While the long-term stability and reusability of the LBIL-FA-Ac sample were not ideal, the present study provides a valuable foundation for understanding its CO₂ capture performance. These results highlight opportunities for further improvement, such as optimizing regeneration conditions, tuning the ligand environment by increasing chelate strength, or using supported/hybrid systems to enhance material immobilization and stability.

9.

CO₂ cycle with absorption temperature at 30 °C under 100% CO₂ for 60 min and desorption was performed at 55 °C under 100% N₂ for 60 min.

3.7. CO₂ Absorption Mechanism

Most ILs reported are found to be predominantly absorb CO₂ physically, though chemical absorption can also occur depending on the cation and anion present. To investigate the underlying mechanism of CO₂ absorption in LBILs samples, the interaction between IL and CO₂ was examined using FT-IR (Figure ) and ¹³C NMR (Figure ) spectra at ambient temperature by bubbling the LBIL samples with CO₂. The FT-IR spectra of the CO₂ saturated sample showed no significant new peaks compared to the lean sample, but upon CO₂ absorption, a notable change was observed in the spectrum. The band at 1616 cm^–1, corresponding to the conjugated CC and CO bonds of the chelated acetylacetonate ligand, disappeared, while the band intensity around 1650 cm^–1 has increased. This suggests that the CO₂ molecules may perturb the electron density within the acetylacetonate chelate and disrupt the vibrational coupling, causing a redistribution of the band intensity. The observed blue-shifting of the acetylacetonate-related vibration is consistent with weak CO₂-ligand interactions that alter local polarization and vibrational coupling. Similar blue-shifts of carbonyl type bands upon CO₂ exposure have been reported, where these are often attributed to dielectric effects that are significantly greater than specific interactions between CO₂ and carbonyl groups. − The broadening and decreased intensity in aldehydic proton and methyl acetate vibrational were attributed to intermolecular C–H···O as well as dielectric and donor–acceptor interactions between CO₂ and electron-rich oxygen sites. ,

10.

FT-IR spectra for LBILs samples of the CO₂ lean and CO₂ saturated samples.

11.

¹³C NMR results for LBILs samples of CO₂ lean and CO₂ saturated.

The band at 1620 cm^–1 is typically associated with CO stretching vibrations and is often used as an indicator of bicarbonate (HCO^3–) formation, which was not observed in the CO₂ saturated samples. Similarly, the band at 1452 cm^–1 may be attributed to asymmetric C–O stretching of carbonate species. In order to confirm any formation of carbonate species, further investigation using ¹³C NMR results revealed no new peaks in the 160 ppm region, which would be a significant characteristic of carbonate species. The absence of these signals strongly indicates that no chemical reaction with CO₂ occurred. Furthermore, the viscosity after the CO₂ absorption was also investigated. The viscosities of both the CO₂ lean and CO₂ saturated samples either decreased or remained similar for all ILs (Figure S14). In the case of LBIL and LBIL-FA, the viscosity was shown to decrease, as CO₂ molecules dissolve physically into the ionic liquid. Experimental studies have shown that physical absorption of CO₂ weakens ion–ion interactions, resulting in a decrease in viscosity. On the other hand, LBIL-FA-Ac samples exhibited an average viscosity of 13.48 mPa.s in the CO₂ lean sample and 12.53 mPa.s after CO₂ saturation. This is because of the effect of the acetylacetonate ligand, which has already dispersed the molecular interactions within the ILs, resulting in minimal changes after CO₂ saturation. This suggests that no significant formation of carbonate species occurred, which could otherwise have led to viscosity changes.

The absorption enthalpy values of the material can also be good indicators of the determination of the absorption mechanism. For all LBILs, the enthalpy values obtained (Sec ) were relatively low (<40 kJ mol^–1), indicating that CO₂ interacts with the ILs predominantly through physical absorption rather than chemisorption. Together, these results suggest that the CO₂ absorption in LBIL samples proceeds mainly via physical absorption.

4. Conclusions

In this study, a lithium borate-based ionic liquid incorporating Mg­(acac)₂ was developed and evaluated for its CO₂ capture performance. The introduction of Mg­(acac)₂ provided additional cooperative sites, enhancing the interaction between CO₂ and the ionic liquid through a cooperative push–pull mechanism. While the overall CO₂ uptake of the LBILs was lower than that of conventional, the designed ionic liquid demonstrated comparable or superior performance to other reported ionic liquids under milder conditions, highlighting its potential for low energy CO₂ capture. Moreover, the relatively weak interaction between CO₂ and the ionic liquid facilitates the easy desorption of CO₂ without requiring high thermal input. The balance between moderate uptake and energy efficiency could offer promise for practical CO₂ capture applications. Future efforts may focus on optimizing the ionic liquid formulation or integrating with porous supports to further enhance the stability and performance.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c09731.

• Chemical structures of LBILs; NMR spectra of synthesized LBILs; DSC data for LBILs; comparison of viscosity and CO₂ uptake capacity with other materials; viscosity measurements for LBILs upon CO₂ saturation (PDF)

The authors declare no competing financial interest.