The Absorption Performance of Ionic Liquids–PEG200 Complex Absorbent for VOCs

: A novel complex absorbent composed of polyethylene glycol 200 (PEG200) and ionic liquids (ILs) was prepared for the absorption of volatile organic compounds (VOCs) such as dichloromethane (DCM) and benzene. We prepared complex absorbents composed of [EMIM][Cl], [BMIM][Cl], [HMIM][Cl], [BMIM][BF 4 ], [BMIM][PF 6 ], [BMIM][NTF 2 ], and PEG200, respectively, and studied the absorption properties of these six complex absorbents for DCM and benzene. The results show that under the optimized situation, the absorptivity of [HMIM][Cl]–PEG200 complex absorbent for DCM is 85.46% in the ﬁrst 5 min, and 87.15% for benzene. No obvious decay in the absorptivity of [HMIM][Cl]– PEG200 for DCM and benzene was observed in ﬁve cycles, indicating an impressive regeneration performance. Furthermore, the mechanism of ionic liquid absorption for VOC is explored by thermodynamic analysis and quantum chemical calculations. The theoretical calculation results show that the [HMIM][Cl]–DCM interaction is stronger than the [HMIM][Cl]–benzene interaction, which is consistent with the results of the absorption experiment. Moreover, the strong hydrogen bonds can be formed between both [HMIM][Cl]–DCM and [HMIM][Cl]–benzene. for DCM and benzene. The results show that when the concentration of DCM was 20,000 mg · m − 3 , the concentration of benzene was 5000 mg · m − 3 , the ﬂow rate of the inlet gas (Q) was 0.036 m 3 · h − 1 and the absorption temperature was 298.15 K, the absorptivities of 40


Introduction
Various volatile organic compounds (VOCs) are produced in the petrochemical industry, including benzene, dichloromethane, acetone, and phenol [1,2]. To date, more and more attention has been paid to the treatment of dichloromethane and benzene because they are highly toxic, leading to the risk of human cancer [3][4][5]. Therefore, the effective method of capturing and processing VOCs is the key to protect the ecological environment, improve human health, and achieve sustainable development [6]. At present, there are two ways to treatment the VOCs: destructive and recyclable techniques [7]. The technologies for destroying VOCs mainly include thermal or catalytic oxidation [8], biodegradation [9], and plasma methods [10]. These technologies convert VOCs into carbon dioxide and water, mainly through physical methods or chemical reactions. The recovery techniques include absorption [7,11], adsorption [12,13], condensation [14], and membrane separation [15,16]. These methods enrich or separate VOCs by physical or chemical methods. Two-stage condensation is often used in industrial processes to recover DCM and benzene. However, the concentration of DCM in the exhaust gas makes it difficult to meet emission standards [17]. Therefore, it is of great significance to find a new method that enables us to efficiently treat dichloromethane and benzene at the same time. Notably, the absorption method has been extensively used due to its simplicity, safety, recyclability, low energy consumption, and good price competitiveness. Besides, it is worth noting that the most important thing during the absorption is to choose the most appropriate absorbent for it. The common solutions used for removing VOC include water, alcohol, and ketone [18,19]. However, these absorbents have plenty of disadvantages such as low absorptivity, poor recyclability, high corrosiveness, and tendency to cause secondary pollution. Therefore, it has always been the focus of a great number of researchers to address the problems happening with applying the traditional absorbents for the absorption.
Ionic liquids (ILs) have been widely studied because of their extremely low vapor pressure, adjustable structure, high solubility for organic substances, and excellent thermal stability [20,21]. Obviously, ionic liquids with extremely low vapor pressure have lower energy consumption during the separation of solutes and solvents than traditional absorbents with high volatility. In recent years, many studies on the absorption of VOC by ILs have been reported [22,23]. For example, Wu et al. [24] synthesized six conventional ILs and measured the capacity of ILs to absorb DCM at different temperatures and partial pressures. The results showed that 1-butyl-3-methylimidazolium thiocyanate ( [25] determined the thermodynamic properties of the absorption process of benzene and acetone in eight ILs in detail. The results show that the anion of the IL plays the most important role in determining the solubility of benzene and acetone in ILs, and increases in the following order: Overall, these studies have confirmed that DCM and benzene have high solubility in ILs, but there were few reports on the dynamic absorption for DCM and benzene by ILs. Moreover, the high viscosity of ILs will lead to slow mass transfer, which is a challenge for large-scale application of ILs in the absorption process. Therefore, the key technology to achieve large-scale VOC treatment and recovery is to find a high-efficiency, cheap, and low-viscosity absorbent [26,27]. Until now, the use of ILs to absorb VOC has received extensive attention from researchers, but few studies on the absorption of DCM and benzene by ILs have been reported. In view of the above discussion, ILs are difficult for wide use as absorbents for VOC absorption because of their high viscosity. In this work, we reduced the viscosity of the IL by adding PEG200 to the IL. Subsequently, to explore the absorption performance of the complex absorbent for mixed gases composed of DCM and benzene, we conducted absorption experiments with DCM and benzene as the mixed gases. The experimental results showed that [HMIM][Cl]-PEG200 was an effective absorber for absorbing a gas mixture consisting of DCM and benzene. In addition, the effects of absorption temperature (T), inlet gas flow rate (Q), inlet gas concentration (C DCM, C Benzene ), and IL mass fraction on the absorptivity were measured. Finally, to explore the absorption mechanism in the absorption process, the interaction energy between VOC and IL was calculated using Gaussian 09 software.  6 ]) (≥99%) were purchased from Shanghai Chengjie Chemical Co., Ltd., Shanghai, China. Nitrogen (≥99.99%) was purchased from Wance Gas Co., Ltd., Tianjin, China. DCM, benzene, and polyethylene glycol 200 were purchased from Tianjin Kermel Technology Co., Ltd., Shanghai, China. All chemical reagent purity is analytical purity, no further processing and purification was performed before use.

Preparation Method of IL-PEG200 Complex Absorbent
The [HMIM][Cl]-PEG200 complex absorbent preparation process was as follows. A total of 12 g of [HMIM][Cl] and 8 g of PEG200 were mixed in a flask and stirred at ambient

Characterization
The IR spectra of the absorbents were obtained by Bruker Tensor 27 IR spectrometer with 16 scans in the range of 4000 to 400 cm −1 . Then, 1 H NMR spectra was acquired on a Bruker AV400 spectrometer in CDCl 3 . The viscosity of the absorbent was measured by using the rotating viscometer (NDJ-79, Shanghai Changji Geological Instrument Co., Ltd., Shanghai, China.) with different rotors.

Absorption Process
The experimental apparatus is illustrated in Scheme 1. The gaseous mixture of DCM and nitrogen and the gaseous mixture of benzene and nitrogen are obtained by two independent bubblers (3) and enter the gas mixer (5). The mass concentrations of DCM and benzene are adjusted by respective mass flow meters (2), respectively. The temperature of the gas is controlled by the oil bath (4). The gas flow is controlled by the mass flow meters. The mixed gas enters the gas washing bottle filled with absorbent, and the outlet gas is discharged into the atmosphere after passing through the tail gas treatment device. Samples are taken at the inlet and outlet of the gas washing bottle, and their concentrations are analyzed by gas chromatogram. The pipes before and after the absorption bottle are polytetrafluoroethylene pipes. The absorption bottle is preloaded with 40 g (43.24 mL) complex absorbent for each test. The diameter of the absorption bottle is 4 cm. In order to simulate the ratio of industrial waste gas, the mass concentration ratio of DCM and benzene is set to 4:1. After the absorption experiment, the regeneration of the absorbent is stripped by nitrogen at a temperature of 333.15 K.

Characterization
The IR spectra of the absorbents were obtained by Bruker Tensor 27 IR spectrometer with 16 scans in the range of 4000 to 400 cm −1 . Then, 1 H NMR spectra was acquired on a Bruker AV400 spectrometer in CDCl3. The viscosity of the absorbent was measured by using the rotating viscometer (NDJ-79, Shanghai Changji Geological Instrument Co., Ltd., Shanghai, China.) with different rotors.

Absorption Process
The experimental apparatus is illustrated in Scheme 1. The gaseous mixture of DCM and nitrogen and the gaseous mixture of benzene and nitrogen are obtained by two independent bubblers (3) and enter the gas mixer (5). The mass concentrations of DCM and benzene are adjusted by respective mass flow meters (2), respectively. The temperature of the gas is controlled by the oil bath (4). The gas flow is controlled by the mass flow meters. The mixed gas enters the gas washing bottle filled with absorbent, and the outlet gas is discharged into the atmosphere after passing through the tail gas treatment device. Samples are taken at the inlet and outlet of the gas washing bottle, and their concentrations are analyzed by gas chromatogram. The pipes before and after the absorption bottle are polytetrafluoroethylene pipes. The absorption bottle is preloaded with 40 g (43.24 mL) complex absorbent for each test. The diameter of the absorption bottle is 4 cm. In order to simulate the ratio of industrial waste gas, the mass concentration ratio of DCM and benzene is set to 4:1. After the absorption experiment, the regeneration of the absorbent is stripped by nitrogen at a temperature of 333.15 K. Scheme 1. Absorption experiment device (1-nitrogen bottle, 2-mass flow meters, 3-bubblers, 4-constant temperature water bath, 5-gas mixer, 6-sampling port, 7-absorption bottle).

Analysis of VOC
Absorptivity (η) was used to evaluate the absorptivity of IL complex absorbent for DCM and benzene. The absorptivity calculation formula is as follows: where η refers to the absorptivity of DCM or benzene in the IL, Cg,i refers to the inlet concentration of DCM or benzene (mg·m −3 ), and Cg,o refers to the outlet concentration of DCM or benzene (mg·m −3 ).

Analysis of VOC
Absorptivity (η) was used to evaluate the absorptivity of IL complex absorbent for DCM and benzene. The absorptivity calculation formula is as follows: where η refers to the absorptivity of DCM or benzene in the IL, C g,i refers to the inlet concentration of DCM or benzene (mg·m −3 ), and C g,o refers to the outlet concentration of DCM or benzene (mg·m −3 ). The concentration of DCM and benzene was measured by GC (SP-3420A, CHINA), hydrogen flame detector, KB-5 capillary column, column temperature 358.15 K, sampler

The Saturated Absorption Capacity of Absorption
During the dynamic absorption experiment, the absorbent continued to absorb DCM and benzene, and the total mass of the absorbent continued to increase. When the inlet gas concentration and the outlet gas concentration were almost equal and there was no significant change within half an hour, the absorbent at this time was considered to reach the dynamic absorption equilibrium. At this time, the ratio of the mass of DCM and benzene absorbed by the absorbent to the mass of the absorbent before absorption is recorded as the saturated absorption capacity. The saturated absorption capacity of absorption was experimentally measured when the DCM concentration, the benzene concentration, absorption temperature, inlet gas flow rate, and mass fraction of IL were fixed at 20,000 mg·m −3 , 5000 mg·m −3 , 298.15 K, 0.036 m 3 ·h −1 , and 60%, respectively.

COSMO-SAC Model
The COSMO for segment activity coefficient (COSMO-SAC) model, derived from the conductor-like screening model (COSMO) [28,29], was first proposed by Lin and Sandler, and is often used to calculate the interaction between molecules and predict the infinite dilution activity coefficient of the solute in the solvent [29][30][31]. First, the σ-spectrum is obtained by quantum chemistry calculations, but this process requires a lot of time. In addition, the time required for calculation increases exponentially as the number of atoms in the molecule increases. Therefore, the time required for the calculation is reduced by calculating the σ-spectra of the anions and cations of the IL separately. Finally, the sigma spectra of IL are obtained by adding the sigma spectra of anions and cations [32,33]. The anions and cations of the IL are drawn by Materials Studio [34]. The optimized anions and cations of the IL are obtained from using the density functional theory with nonlocal VWN-BP functional at the DNP v4.0 basis set [35,36]. After the geometry optimization, the screening charge density (σ*) was calculated by the COSMO model. The surface charge density of standard segment σ m was calculated by Equation (2). In the formula, r eff is the adjustable parameter, r 2 eff = a eff π (a eff = 7.50 Å 2 ), r n is the average radius of the segment, and d mn is the distance of the segment.
The probability distribution of surface charge density obtained by calculation is also called the sigma profile. The definition of the sigma profile for a single molecule is as follows: where A i (σ) is the surface area with a charge density of value σ, and A i is the total surface area of species i. For a mixture, the σ-profile is obtained by a weighted average of the σ-profile of individual components (Equation (4)).
Finally, the sigma profiles of the IL and the solute obtained in the thermodynamic model are used to calculate the activity coefficient of the solute i in the solution. The detailed equation is as follows: where ∆W refers to the interaction energy between fragments and is presented in detail in original papers (COSMO-SAC(2007) [30], COSMO-SAC(2010) [32], and COSMO-SAC(ion)) [37].
where R is the universal gas constant, T is the temperature, x i is the mole fraction of component i, n i is the total number of segments in component i (n i = A i /a eff ), a eff is the standard surface segment area (empirical parameter) (a eff = 0.075 nm 2 ), r i is the standard volume parameter of component i, q i is the standard area parameter of component i, and z is the coordination number (z = 10).

Quantum Chemistry Calculation
We carried out density functional theory (DFT) calculations on the binding energy of each system by Gaussian 09 [38]. The molecules (IL, DCM, and benzene) were optimized by B3LYP/6-311G* basis set and the energy of each system (I-DCM and IL-benzene) was calculated [38][39][40][41]. Then, the binding energy of each IL to DCM and benzene was calculated by the following equation: where E A is the energy (eV) of a single molecular model, E B is the energy (eV) of adsorbed molecule B, and E A−B is the energy (eV) of A-B adsorption system.

Absorbent Screening
In order to find an absorbent with high absorptivity for DCM and benzene, the absorptivity of complex absorbents containing different ILs were tested (Figure 1). Among them, the mass fraction of IL in the IL-PEG200 complex absorbent is 60%. As shown in Figure 1a

Influence of Different Mass Fractions of IL in Absorbent
The saturated absorption capacity of pure IL for the mixture of DCM and benzene reached 114.97 mg·g −1 (Table 1), and it was far greater than that of PEG200 for the mixed gas. In addition, it could be seen from the experimental results in Figure 2 (Figure 2). When the mass fraction of the IL in the complex absorbent dropped from 100% to 70%, the absorptivity (at 5 min) of the complex absorbent for DCM and benzene increased. In addition, the absorptivity (at 5 min) of the complex absorbent for DCM and benzene even reached 72.31% and 75.37%, respectively, when the mass fraction of the IL in the absorbent was 60%. Obviously, diluting the IL with PEG200 reduced the viscosity of the absorbent (Table  2), thus improving the bubble dispersion and allowing the solute to better contact the absorbent. However, when the mass fraction of IL in the complex absorbent was further reduced from 60% to 50%, the absorptivity (at 5 min) of the complex absorbent for DCM and benzene decreased to 36.12% and 46.87%, respectively. This was because, although PEG200 was mixed with IL to effectively improve the mass transfer efficiency, with the further increase of the content of PEG200, the decrease of IL concentration would lead to a decrease in the dissolution capacity of complex absorbers. In short, the increase in mass transfer efficiency failed to offset the decrease in absorption performance caused by the adverse effect of the decrease in the concentration of IL. Therefore, in the process of absorbing DCM and benzene by the complex absorbent, the optimal mass fraction of the IL in the absorbent was 60%. Thus, in subsequent experiments, the amount of IL was fixed at 60% of the total mass (40 g) of the absorbent.

Influence of Variable Parameters on VOC Absorptivity in the Complex Absorption
The effect of the absorption performance (C DCM , C Benzene , T, and Q) of [HMIM][Cl]-PEG200 for DCM and benzene at different parameters was investigated.
The absorptivity of [HMIM][Cl]-PEG200 for DCM and benzene at different inlet gas concentrations was explored. As shown in Figure 3, with the increase of inlet gas concentration, there was no significant change in the absorptivity for benzene, but the DCM absorptivity (at 5 min) decreased from 72.31% to 62.53% when the DCM inlet gas concentration increased from 20,000 mg·m −3 to 50,000 mg·m −3 . It is worth noting that, with the prolongation of the experimental time, the absorptivity decreases more slowly at the inlet concentration of 20,000 mg·m −3 and 5000 mg·m −3 for DCM and benzene. Usually, based on the two-layer film theory, the greater difference in concentration between the gas phase and the gas-liquid phase, the greater the driving force for mass transfer. However, the mass transfer efficiency in practical application is limited by the physical properties and contact area of absorbent. Therefore, the increase of inlet gas concentration will lead to the decrease of absorptivity [42]. Moreover, as the experiment time increased, the concentration of DCM and benzene in the absorbent increased, while the concentration difference between the gas phase and the gas-liquid interface decreased. As is known, this concentration difference was the driving force for the diffusion of DCM molecules and benzene molecules. To sum up, the higher the inlet gas concentration, the faster the absorptivity decreases in the absorption process. The absorptivity at different absorption temperatures was experimentally measured when the DCM concentration, the benzene concentration, the mass fraction of IL in complex absorbent, and the inlet gas flow rate were fixed at 20,000 mg·m −3 , 5000 mg·m −3 , 60%, and 0.036 m 3 ·h −1 , respectively. As shown in Figure 4, when the temperature increases from 298.15 K to 303.15 K, the absorptivity (at 5 min) for DCM and benzene decreases from 72.31% and 75.37% to 56.35% and 59.54%, respectively. As the absorption temperature increases, the Henry's law constants of DCM and benzene in the complex absorbent become larger, indicating that the solubility of DCM and benzene in [HMIM][Cl]-PEG200 decreases. Therefore, the solubility has a significant effect on the absorptivity, causing the absorptivity to decrease with the increase of absorption temperature. As shown in Figure 5, the absorptivity (at 5 min) for DCM and benzene increases to varying degrees as the gas flow rate decreases from 0.036 m 3 ·h −1 to 0.012 m 3 ·h −1 . Among them, the DCM absorptivity (at 5 min) increases most obviously, from 72.31% to 85.46%, while the benzene absorptivity (at 5 min) increases from 75.37% to 87.15%. The reason is that the residence time of the mixed waste gas in the complex absorbent increases when the gas flow rate decreases, which facilitates the contact of the gas and liquid phases. Therefore

Recovery Performance
The recyclability is an important indicator to evaluate absorbent, and the recycling performance of the [HMIM][Cl]-PEG200 complex absorbent was investigated. As can be seen from Figure 6, after [HMIM][Cl]-PEG200 was repeatedly used five times, the absorptivity for DCM and benzene at 5 min were 85.25% and 85.95%, respectively, which was only 0.21% and 1.2% lower than that of fresh [HMIM][Cl]-PEG200. These results confirm that the complex absorbent had excellent reusable performance. It can be seen from Figure 7

Thermodynamic Calculation and Quantitative Calculation
In order to explore the absorption mechanism of [HMIM][Cl], DFT theoretical calculation was carried out for IL, DCM, and benzene systems. Firstly, the infinite dilution activity coefficient and Henry's law constant were important thermodynamic parameters of the absorption process [43]. The infinite dilution activity coefficients of DCM or benzene in IL were calculated by the COSMO-SAC model. Moreover, the solubility of DCM or benzene in ILs is a result of the phase equilibrium [44][45][46][47]. Therefore, the Henry's constant, Hi(T), is defined as: In the formula, H i (kPa) refers to Henry's coefficient, γ ∞ i refers to the activity coefficients at infinite dilution of solute i in the IL, and p s i (kPa) refers to the saturated vapor pressure of DCM or benzene at temperature T (K). Table 3 shows that the infinite dilution activity coefficients of DCM in [HMIM][Cl] at different temperatures were all less than 1, which indicates that negative deviations from Raoult's law involve solutes which can be better absorbed by ILs. In addition, as shown in Tables 3 and 4, the Henry's law constants and infinite dilution activity coefficients of DCM and benzene increase with the increase of temperature, which indicates that the solubility of DCM and benzene decreases with increasing temperature. Apparently, the absorption of DCM and benzene is physical absorption. In summary, the increase in temperature is not conducive to the absorption of DCM and benzene, which is consistent with the results of the dynamic absorption experiment. The thermodynamic properties of absorbing DCM and benzene in an IL-PEG200 complex absorbent related to the Henry's law constant can be calculated according to the well-known thermodynamic relations [48]: where ∆G, ∆H, and ∆S are the Gibbs free energy, enthalpy, and entropy, respectively. Table 5 shows that all the enthalpy values were negative, demonstrating that the dissolution of DCM and benzene in [HMIM][Cl] is exothermic. In addition, the |∆H| of DCM and benzene is smaller than the |∆S| of DCM and benzene. Therefore, it can be judged that the dissolution of DCM and benzene in [HMIM][Cl] is controlled by entropy. The larger the ∆S value, the more favorable the absorption process. The partial molar excess enthalpy of the system was calculated by infinite dilution activity coefficient [49][50][51]. According to the Gibbs-Helmholtz equation, partial molar excess enthalpy (H i E,∞ ) is calculated as follows: where R is the gas constant. The smaller the value of H E,∞ i , the greater the interaction force between the solute and the solvent. It can be seen from Table 5

Interaction Energy Analysis between IL and Each Absorbed Substance
The results of quantum chemistry calculations ( Figure 9) showed that the strong-toweak order of binding energy of IL-DCM and IL-benzene corresponded to the results of the dynamic absorption experiment. It is worth noting that the binding energy of DCM-

Conclusions
In summary, we designed a variety of IL-PEG200 absorbents and screened out [HMIM][Cl]-PEG200 with high absorptivity for DCM and benzene. The results show that when the concentration of DCM was 20,000 mg·m −3 , the concentration of benzene was 5000 mg·m −3 , the flow rate of the inlet gas (Q) was 0.036 m 3 ·h −1 , and the absorption temperature was 298. 15