Facile Construction of Carboxyl-Functionalized Ionic Polymer towards Synergistic Catalytic Cycloaddition of Carbon Dioxide into Cyclic Carbonates

The development of bifunctional ionic polymers as heterogeneous catalysts for effective, cocatalyst- and metal-free cycloaddition of carbon dioxide into cyclic carbonates has attracted increasing attention. However, facile fabrication of such polymers having high numbers of ionic active sites, suitable types of hydrogen bond donors (HBDs), and controlled spatial positions of dual active sites remains a challenging task. Herein, imidazolium-based ionic polymers with hydroxyl/carboxyl groups and high ionic density were facilely prepared by a one-pot quaternization reaction. Catalytic evaluation demonstrated that the presence of HBDs (hydroxyl or carboxyl) could enhance the catalytic activities of ionic polymers significantly toward the CO2 cycloaddition reaction. Among the prepared catalysts, carboxyl-functionalized ionic polymer (PIMBr-COOH) displayed the highest catalytic activity (94% yield) in the benchmark cycloaddition reaction of CO2 and epichlorohydrin, which was higher than hydroxyl-functionalized ionic polymer (PIMBr-OH, 76% yield), and far exceeded ionic polymer without HBDs groups (PIMBr, 54% yield). Furthermore, PIMBr-COOH demonstrated good recyclability and wide substrate tolerance. Under ambient CO2 pressure, a number of epoxides were smoothly cycloadded into cyclic carbonates. Additionally, density functional theory (DFT) calculation verified the formation of strong hydrogen bonds between epoxide and the HBDs of ionic polymers. Furthermore, a possible mechanism was proposed based on the synergistic effect between carboxyl and Br− functionalities. Thus, a facile, one-pot synthetic strategy for the construction of bifunctional ionic polymers was developed for CO2 fixation.


Introduction
Exploring efficient technologies to reduce carbon emission has become the focus of international concern due to the ever-increasing atmospheric CO 2 concentration [1,2]. Despite the difficulties, the utilization of CO 2 as a renewable building block for the production of value-added chemicals is deemed one of the most hopeful strategies [3,4]. Recently, numerous transformation approaches have been developed for CO 2 fixation [5,6]. As a result, catalytic conversion of CO 2 into valuable cyclic carbonates has attracted enormous attention owing to 100% atom efficiency and high product selectivity [7][8][9].
For chemical conversion of the inert CO 2 molecules, effective catalysts are essential to overcome the large thermodynamic energy barrier. In this context, various catalytic systems, such as metal salts [10][11][12], imidazolium salts [13][14][15], and ammonium salts [16][17][18], have been developed for CO 2 cycloaddition during the past few decades [19,20]. Generally, metal-based catalysis systems display high catalytic activity, but they also suffer from obvious shortcomings, such as the contamination of metal residue and difficulty of product separation [21][22][23]. One plausible approach to address these issues is to develop highly active metal-free catalytic systems [24][25][26]. Recent advances have demonstrated that organic To investigate the physicochemical properties of the prepared ionic polymers, a number of characterization analyses, including solid-state 13 C NMR spectra, FT-IR, XPS, TEM, SEM, SEM-mapping, N2 adsorption-desorption, and TGA, were carried out. As seen in Figure 1, solid-state 13 C NMR spectra of PIMBr, PIMBr-OH, and PIMBr-COOH showed similar signal patterns, suggesting their similar chemical structure. The characteristic peaks at about 128, 113, and 105 ppm were attributable to C2 and C5, C3-4, and To investigate the physicochemical properties of the prepared ionic polymers, a number of characterization analyses, including solid-state 13 C NMR spectra, FT-IR, XPS, TEM, SEM, SEM-mapping, N 2 adsorption-desorption, and TGA, were carried out. As seen in Figure 1, solid-state 13 C NMR spectra of PIMBr, PIMBr-OH, and PIMBr-COOH showed similar signal patterns, suggesting their similar chemical structure. The characteristic peaks at about 128, 113, and 105 ppm were attributable to C2 and C5, C3-4, and C1 atoms of the polymer molecules [34,35]. In the spectrum of PIMBr, the weak signals at around 42 and 47 ppm corresponded to C6-7 atoms of the saturated carbon [38]. Compared to the spectrum of PIMBr, the bands of C6-7 atoms in PIMBr-OH and PIMBr-COOH showed obvious blue shifts due to the presence of -OH and -COOH groups [38]. It was noteworthy that PIMBr-COOH showed a characteristic peak at around 165 ppm, which was related to the carbon atoms of the carboxyl group, thus suggesting the successful incorporation of the carboxyl group in PIMBr-COOH [47,48]. To investigate the physicochemical properties of the prepared ionic polymers, a number of characterization analyses, including solid-state 13 C NMR spectra, FT-IR, XPS,  TEM, SEM, SEM-mapping, N2 adsorption-desorption, and TGA, were carried out. As seen in Figure 1, solid-state 13 C NMR spectra of PIMBr, PIMBr-OH, and PIMBr-COOH showed similar signal patterns, suggesting their similar chemical structure. The characteristic peaks at about 128, 113, and 105 ppm were attributable to C2 and C5, C3-4, and C1 atoms of the polymer molecules [34,35]. In the spectrum of PIMBr, the weak signals at around 42 and 47 ppm corresponded to C6-7 atoms of the saturated carbon [38]. Compared to the spectrum of PIMBr, the bands of C6-7 atoms in PIMBr-OH and PIMBr-COOH showed obvious blue shifts due to the presence of -OH and -COOH groups [38]. It was noteworthy that PIMBr-COOH showed a characteristic peak at around 165 ppm, which was related to the carbon atoms of the carboxyl group, thus suggesting the successful incorporation of the carboxyl group in PIMBr-COOH [47,48]. The FT-IR spectra of the three samples in Figure 2 displayed characteristic peaks of imidazolium rings at 1553 and 1105 cm −1 , which were attributed to the stretching vibrations of C = C [49,50] and C-N + bonds [51], respectively. These observations suggested the successful incorporation of the imidazolium units in the prepared polymers. For the spectrum of PIMBr-COOH, the characteristic peak depicted at 1323 cm −1 was related to the C-O vibration of -COOH groups [52], indicating the successful incorporation of the carboxyl group in PIMBr-COOH. The bands at around 3386 cm −1 presented in the three samples were attributed to the peak of the absorbed H2O. Indeed, the adsorption of atmospheric H2O was a common phenomenon in PILs due to their hygroscopic properties [38,53]. Thus, FT-IR analyses further proved the successful fabrication of target ionic polymers. The FT-IR spectra of the three samples in Figure 2 displayed characteristic peaks of imidazolium rings at 1553 and 1105 cm −1 , which were attributed to the stretching vibrations of C = C [49,50] and C-N + bonds [51], respectively. These observations suggested the successful incorporation of the imidazolium units in the prepared polymers. For the spectrum of PIMBr-COOH, the characteristic peak depicted at 1323 cm −1 was related to the C-O vibration of -COOH groups [52], indicating the successful incorporation of the carboxyl group in PIMBr-COOH. The bands at around 3386 cm −1 presented in the three samples were attributed to the peak of the absorbed H 2 O. Indeed, the adsorption of atmospheric H 2 O was a common phenomenon in PILs due to their hygroscopic properties [38,53]. Thus, FT-IR analyses further proved the successful fabrication of target ionic polymers. XPS was conducted to further analyze the chemical structure of the prepared polymers. As depicted in Figure 3a, C, N, O, and Br elements were identified in the XPS full spectra of PIMBr, PIMBr-OH, and PIMBr-COOH. The presence of the O element in PIMBr could be attributed to the absorbed H2O in the polymer network, which was consistent with the result of the FT-IR analysis. As seen in Figure 3b, the N 1s spectra exhibited two different peaks at 401 eV and 397 eV, corresponding to the contribution of ionic XPS was conducted to further analyze the chemical structure of the prepared polymers. As depicted in Figure 3a, C, N, O, and Br elements were identified in the XPS full spectra of PIMBr, PIMBr-OH, and PIMBr-COOH. The presence of the O element in PIMBr could be attributed to the absorbed H 2 O in the polymer network, which was consistent with the result of the FT-IR analysis. As seen in Figure 3b, the N 1s spectra exhibited two different peaks at 401 eV and 397 eV, corresponding to the contribution of ionic and non-ionic N atoms of the imidazolium rings, respectively [54][55][56]. The C 1s spectra of PIMBr-OH and PIMBr-COOH in Figure 3c can be deconvoluted into three peaks at around 286, 285, and 284 eV, which were attributable to C-O, C-N, and C-C bonds, respectively [38,54,57]. The Br 3d spectra in Figure 3d were fitted into two peaks related to the 3d 3/2 and 3d 5/2 binding energy of Br − [58]. Therefore, XPS analyses further demonstrated the successful fabrication of target polymers. Element distribution of PIMBr-COOH was also investigated by EDS mapping. As illustrated in Figure 4, N, O, and Br elements were homogeneously distributed in the ionic polymer framework, suggesting that the active sites of the prepared catalyst were well-distributed. For solid catalysts bearing synergistic catalytic sites, the homogeneous distribution of the active sites was important for achieving synergistic catalytic performance. XPS was conducted to further analyze the chemical structure of the prepared polymers. As depicted in Figure 3a, C, N, O, and Br elements were identified in the XPS full spectra of PIMBr, PIMBr-OH, and PIMBr-COOH. The presence of the O element in PIMBr could be attributed to the absorbed H2O in the polymer network, which was consistent with the result of the FT-IR analysis. As seen in Figure 3b, the N 1s spectra exhibited two different peaks at 401 eV and 397 eV, corresponding to the contribution of ionic and non-ionic N atoms of the imidazolium rings, respectively [54][55][56]. The C 1s spectra of PIMBr-OH and PIMBr-COOH in Figure 3c can be deconvoluted into three peaks at around 286, 285, and 284 eV, which were attributable to C-O, C-N, and C-C bonds, respectively [38,54,57]. The Br 3d spectra in Figure 3d were fitted into two peaks related to the 3d3/2 and 3d5/2 binding energy of Br − [58]. Therefore, XPS analyses further demonstrated the successful fabrication of target polymers. Element distribution of PIMBr-COOH was also investigated by EDS mapping. As illustrated in Figure 4, N, O, and Br elements were homogeneously distributed in the ionic polymer framework, suggesting that the active sites of the prepared catalyst were well-distributed. For solid catalysts bearing synergistic catalytic sites, the homogeneous distribution of the active sites was important for achieving synergistic catalytic performance.   The porous properties of the prepared polymers were measured by nitrogen physisorption at 77 K. As summarized in Table 1, PIMBr and PIMBr-OH showed low specific surface areas of 8.47 and 11.20 m 2 g −1 , respectively, which were similar to those of ionic polymers synthesized by the quaternization reaction [59,60]. PIMBr-COOH re- The porous properties of the prepared polymers were measured by nitrogen physisorption at 77 K. As summarized in Table 1, PIMBr and PIMBr-OH showed low specific surface areas of 8.47 and 11.20 m 2 g −1 , respectively, which were similar to those of ionic polymers synthesized by the quaternization reaction [59,60]. PIMBr-COOH revealed a moderate surface area of 36.51 m 2 g −1 that was higher than that of PIMBr-OH. The relatively low surface areas of the three polymers could be ascribed to their high ionic density and flexible propylene linkers. High ionic density led to strong electrostatic interactions between the polymer chains [59,60], while the flexibility of the propylene linkers resulted in framework interpenetration during the framework formation. The morphologies of the prepared polymers were analyzed SEM ( Figure 5). It is worth noting that PIMBr-COOH showed a sponge-like morphology that comprised irregular reunion particles. This rough surface structure endowed the polymer with abundant absorption and active sites for CO 2 fixation. PIMBr and PIMBr-OH displayed irregular agglomeration structures of large particles, and the particle size was much greater than that of PIMBr-COOH. A representative TEM image of PIMBr-COOH ( Figure 5d) revealed that the diameter of the aggregated particles was in a range of tens to hundreds of nanometers. As polymer-based solid catalysts, the thermal stability of the polymers is a critical aspect of their practical application. Therefore, TGA was conducted to analyze their thermal stability over the temperature range of 25-600 °C. As depicted in Figure 6, three samples showed similar thermogravimetric behavior. The initial weight loss (2-5 wt%) that occurred below 150 °C could be assigned to the removal of the absorbed water, a common phenomenon for ionic polymers [49,61]. For PIMBr and PIMBr-OH, the main weight loss took place above 250 °C. Compared to PIMBr and PIMBr-OH, PIMBr-COOH exhibited a lower decomposition temperature and a higher weight loss rate. These could be owing to the high oxygen content and easy decomposition of -COOH groups. TGA results indicated that the three samples could be stable at 220 °C. For CO2 cycloaddition, the reaction was generally performed below 150 °C; therefore, the prepared polymers were stable enough to promote such CO2 conversion. As polymer-based solid catalysts, the thermal stability of the polymers is a critical aspect of their practical application. Therefore, TGA was conducted to analyze their thermal stability over the temperature range of 25-600 • C. As depicted in Figure 6, three samples showed similar thermogravimetric behavior. The initial weight loss (2-5 wt%) that occurred below 150 • C could be assigned to the removal of the absorbed water, a common phenomenon for ionic polymers [49,61]. For PIMBr and PIMBr-OH, the main weight loss took place above 250 • C. Compared to PIMBr and PIMBr-OH, PIMBr-COOH exhibited a lower decomposition temperature and a higher weight loss rate. These could be owing to the high oxygen content and easy decomposition of -COOH groups. TGA results indicated that the three samples could be stable at 220 • C. For CO 2 cycloaddition, the reaction was generally performed below 150 • C; therefore, the prepared polymers were stable enough to promote such CO 2 conversion. thermal stability over the temperature range of 25-600 °C. As depicted in Figure 6, three samples showed similar thermogravimetric behavior. The initial weight loss (2-5 wt%) that occurred below 150 °C could be assigned to the removal of the absorbed water, a common phenomenon for ionic polymers [49,61]. For PIMBr and PIMBr-OH, the main weight loss took place above 250 °C. Compared to PIMBr and PIMBr-OH, PIMBr-COOH exhibited a lower decomposition temperature and a higher weight loss rate. These could be owing to the high oxygen content and easy decomposition of -COOH groups. TGA results indicated that the three samples could be stable at 220 °C. For CO2 cycloaddition, the reaction was generally performed below 150 °C; therefore, the prepared polymers were stable enough to promote such CO2 conversion.

Catalytic Performance
With the target ionic polymers in hand, their catalytic performance was evaluated in the CO2 cycloaddition reaction to certify the rational design of HBDs-functionalized PILs. At the beginning of the study, catalytic activities of PIMBr, PIMBr-OH, and PIMBr-COOH were compared in the cycloaddition of CO2 into less-volatile epichlorohydrin (model reaction). As summarized in Table 2, 1 mol% of HBDs-free PIMBr af-

Catalytic Performance
With the target ionic polymers in hand, their catalytic performance was evaluated in the CO 2 cycloaddition reaction to certify the rational design of HBDs-functionalized PILs. At the beginning of the study, catalytic activities of PIMBr, PIMBr-OH, and PIMBr-COOH were compared in the cycloaddition of CO 2 into less-volatile epichlorohydrin (model reaction). As summarized in Table 2, 1 mol% of HBDs-free PIMBr afforded a 54% yield of target cyclic carbonate at 100 • C for 4 h ( Table 2, entry 1). Notably, under the same reaction conditions, PIMBr-OH and PIMBr-COOH provided much higher yields, reaching up to 76% and 94%, respectively (Table 2, entries 2 and 3). Given that PIMBr, PIMBr-OH, and PIMBr-COOH had similar chemical structures and ionic densities, the higher catalytic activities of PIMBr-OH and PIMBr-COOH can be due to the synergistic catalytic effect between hydroxyl/carboxyl HBDs and Br − [62][63][64], which was able to promote the rate-determining step (i.e., epoxide ring opening) of the reaction, thereby accelerating the reaction rate. Compared to PIMBr-OH, the higher catalytic activity of PIMBr-COOH should be ascribed to its stronger hydrogen bonding interaction with epichlorohydrin molecular [47]. To confirm these assumptions, DFT calculations were carried out. As depicted in Figure 7, the results of DFT calculation were consistent with those of catalytic evaluation. Hydrogen bonds with the bond lengths of 1.930 and 1.740 Å were formed between epichlorohydrin, PIMBr-OH, and PIMBr-COOH. The shorter H-bond length between epichlorohydrin and PIMBr-COOH suggested a stronger hydrogen bond interaction. Accordingly, the bond length of the C-O bond in the epichlorohydrin ring was lengthened from 1.440 Å to 1.441 and 1.442 Å after interaction with PIMBr-OH and PIMBr-COOH, respectively. Therefore, the results of the catalytic evaluation and DFT calculations verified the importance of suitable HBDs in boosting the catalytic efficiency of ionic polymers. 1.930 and 1.740 Å were formed between epichlorohydrin, PIMBr-OH, and PIMBr-COOH. The shorter H-bond length between epichlorohydrin and PIMBr-COOH suggested a stronger hydrogen bond interaction. Accordingly, the bond length of the C-O bond in the epichlorohydrin ring was lengthened from 1.440 Å to 1.441 and 1.442 Å after interaction with PIMBr-OH and PIMBr-COOH, respectively. Therefore, the results of the catalytic evaluation and DFT calculations verified the importance of suitable HBDs in boosting the catalytic efficiency of ionic polymers.   suggested a stronger hydrogen bond interaction. Accordingly, the bond length of the C-O bond in the epichlorohydrin ring was lengthened from 1.440 Å to 1.441 and 1.442 Å after interaction with PIMBr-OH and PIMBr-COOH, respectively. Therefore, the results of the catalytic evaluation and DFT calculations verified the importance of suitable HBDs in boosting the catalytic efficiency of ionic polymers.   Having the optimal catalyst in hand, the influences of reaction parameters, such as temperatures, CO 2 pressure, and reaction time, were investigated to optimize the reaction conditions. As illustrated in Table 3, the results indicate that reaction temperature had a great impact on the catalytic activity. The yield of target cyclic carbonate exhibited an obvious increment when the reaction temperature increased from 40 to 100 • C ( Table 3, entries 1-3). At 100 • C, the effect of CO 2 pressure was also evaluated. As summarized in Table 3, the product yield decreased gradually as the CO 2 pressure reduced from 1 MPa to 0.1 MPa. Notably, 77% yield and high selectivity of target carbonate were still observed at ambient CO 2 pressure. The good catalytic performance under low CO 2 pressure should be assigned to the high ionic density and porous structure of the PIMBr-COOH catalyst. Under ambient CO 2 pressure, the effect of reaction time was also checked. As displayed in Figure 8, the product yield increased from 31% to 96% when the reaction time extended from 1 to 6 h. Therefore, we obtained the optimal reaction conditions: PIMBr-COOH (1 mol%), 100 • C, CO 2 pressure (0.1 MPa), and 6 h. an obvious increment when the reaction temperature increased from 40 to 100 °C ( Table  3, entries 1-3). At 100 °C, the effect of CO2 pressure was also evaluated. As summarized in Table 3, the product yield decreased gradually as the CO2 pressure reduced from 1 MPa to 0.1 MPa. Notably, 77% yield and high selectivity of target carbonate were still observed at ambient CO2 pressure. The good catalytic performance under low CO2 pressure should be assigned to the high ionic density and porous structure of the PIMBr-COOH catalyst. Under ambient CO2 pressure, the effect of reaction time was also checked. As displayed in Figure 8, the product yield increased from 31% to 96% when the reaction time extended from 1 to 6 h. Therefore, we obtained the optimal reaction conditions: PIMBr-COOH (1 mol%), 100 °C, CO2 pressure (0.1 MPa), and 6 h.  Apart from catalytic activity, the recyclability of the solid catalysts is also a crucial indicator for evaluating their potential in industrial applications. Therefore, the recycling performance of PIMBr-COOH was evaluated under optimal reaction conditions. At the end of each catalytic run, the solid catalyst was centrifugally separated, washed with ethanol, dried under a vacuum, and then directly engaged in the next catalytic cycle. As shown in Figure 9, PIMBr-COOH showed nice recyclability and could be used at least six times. The initial activity decrease after the first run might be attributed to the partial pore blocking during the first catalytic run. For comparison, the catalytic performance of some previously reported metal and cocatalyst-free heterogeneous catalysts is listed in Table 4. Notably, PIMBr-COOH showed better catalytic performance than most of the Apart from catalytic activity, the recyclability of the solid catalysts is also a crucial indicator for evaluating their potential in industrial applications. Therefore, the recycling performance of PIMBr-COOH was evaluated under optimal reaction conditions. At the end of each catalytic run, the solid catalyst was centrifugally separated, washed with ethanol, dried under a vacuum, and then directly engaged in the next catalytic cycle. As shown in Figure 9, PIMBr-COOH showed nice recyclability and could be used at least six times. The initial activity decrease after the first run might be attributed to the partial pore blocking during the first catalytic run. For comparison, the catalytic performance of some previously reported metal and cocatalyst-free heterogeneous catalysts is listed in Table 4. Notably, PIMBr-COOH showed better catalytic performance than most of the previously reported solid and metal-free catalysts. In terms of the salient features, such as facile preparation, high numbers of ionic active sites, controlled spatial positions of dual active sites, high catalytic activity, and effective catalyst recyclability, the designed bifunctional ionic polymer was shown to be a promising heterogeneous and environmentally friendly catalyst for CO 2 cycloaddition. previously reported solid and metal-free catalysts. In terms of the salient features, such as facile preparation, high numbers of ionic active sites, controlled spatial positions of dual active sites, high catalytic activity, and effective catalyst recyclability, the designed bifunctional ionic polymer was shown to be a promising heterogeneous and environmentally friendly catalyst for CO2 cycloaddition.  This  To investigate the general applicability of the PIMBr-COOH catalyst in the CO 2 cycloaddition reaction, epoxides bearing different functional groups were tested. As shown in Table 5, epoxides with electron-withdrawing groups, such as bromide and alkoxy, worked well and reacted with CO 2 smoothly to afford excellent yields of corresponding products in 6 h ( Table 5, entries 1-4). Epoxides bearing electron-donating alkyl groups could also provide excellent yields of target cyclic carbonates, but prolonged reaction times are needed ( Table 5, entries 5 and 6). As a representative example of sterically hindered epoxide, styrene oxide provided a high yield of corresponding cyclic carbonate in 6 h ( Table 5, entry 7). These results further revealed that the prepared PIMBr-COOH is a promising heterogeneous catalyst for CO 2 cycloaddition reaction.

Plausible Reaction Mechanism
Based on the experimental results and literature data [44,48,49,70], a plausible mechanism for CO2 cycloaddition reaction using PIMBr-COOH as a catalyst was proposed (Scheme 2). Firstly, the carboxyl group in PIMBr-COOH coordinated with the oxygen atom in epichlorohydrin, leading to the formation of a hydrogen bond and thereby activating the C-O bond of the epichlorohydrin ring. Subsequently, the epichlorohydrin ring was opened via the nucleophilic attack of Br − on the β-carbon atom with little steric hindrance in epichlorohydrin, leading to the formation of a bromoalkoxide intermediate. Then, an alkylcarbonate anion was produced through the reaction between the bromoalkoxide intermediate and CO2. Lastly, target cyclic carbonate was generated via the intermolecular ring-closing of the alkylcarbonate anion. Meanwhile, PIMBr-COOH was regenerated and participated in the next catalytic cycle.

Plausible Reaction Mechanism
Based on the experimental results and literature data [44,48,49,70], a plausible mechanism for CO 2 cycloaddition reaction using PIMBr-COOH as a catalyst was proposed (Scheme 2). Firstly, the carboxyl group in PIMBr-COOH coordinated with the oxygen atom in epichlorohydrin, leading to the formation of a hydrogen bond and thereby activating the C-O bond of the epichlorohydrin ring. Subsequently, the epichlorohydrin ring was opened via the nucleophilic attack of Br − on the β-carbon atom with little steric hindrance in epichlorohydrin, leading to the formation of a bromoalkoxide intermediate. Then, an alkylcarbonate anion was produced through the reaction between the bromoalkoxide intermediate and CO 2 . Lastly, target cyclic carbonate was generated via the intermolecular ring-closing of the alkylcarbonate anion. Meanwhile, PIMBr-COOH was regenerated and participated in the next catalytic cycle.
thereby activating the C-O bond of the epichlorohydrin ring. Subsequently, the epichlorohydrin ring was opened via the nucleophilic attack of Br − on the β-carbon atom with little steric hindrance in epichlorohydrin, leading to the formation of a bromoalkoxide intermediate. Then, an alkylcarbonate anion was produced through the reaction between the bromoalkoxide intermediate and CO2. Lastly, target cyclic carbonate was generated via the intermolecular ring-closing of the alkylcarbonate anion. Meanwhile, PIMBr-COOH was regenerated and participated in the next catalytic cycle.

Scheme 2.
Possible catalytic mechanism for the cycloaddition of CO2 with epichlorohydrin catalyzed by PIMBr-COOH(* means repeating units of the polymer). Scheme 2. Possible catalytic mechanism for the cycloaddition of CO 2 with epichlorohydrin catalyzed by PIMBr-COOH(* means repeating units of the polymer).

Synthesis of PIMBr, PIMBr-OH and PIMBr-COOH
HBDs-functionalized PILs were fabricated via a one-pot quaternization strategy. In detail, TIB (1.73 mmol, 0.48 g) and 3-bromo-2-(bromomethyl)propionic acid (2.32 mmol, 0.57 g) were added to N,N-dimethylformamide (30 mL) in a 100 mL pressure-resistant tube. After replacing the interior air with N 2 , the tube was transferred to an oil bath, stirred, and heated under 100 • C for 24 h. Subsequently, the tube was removed from the oil bath and cooled naturally to room temperature. Next, the resultant solid was washed thoroughly with ethanol, filtered, and dried under vacuum (80 • C) for 24 h. Finally, 1.03 g of white solid (98% yield) were obtained and named PIMBr-COOH (1.03 g). Following the same synthesis procedures of PIMBr-COOH, PIMBr and PIMBr-OH were also successfully fabricated in nearly quantitative yields by taking the place of 3-bromo-2-(bromomethyl)propionic acid with other 1,3-dibromo compounds, i.e., 1,3-dibromopropane and 1,3-dibromo-2propanol, respectively.

Catalytic Performance
In a typical reaction of catalytic CO 2 cycloaddition, catalysts (PIMBr-COOH, PIMBr-OH, or PIMBr) and epoxide were successively added to a 25 mL high-pressure reactor equipped with a magnetic stirrer. After purging with CO 2 several times, the reactor was charged with CO 2 to 1 MPa and reacted at a pre-set temperature for the required time. After reaction completion, the system was cooled to room temperature. The resultant mixture was separated by centrifugation, and the obtained liquid was analyzed by GC and GC-MS using dimethylacetamide as the internal standard.

Characterization
Solid-state 13 C NMR measurements were analyzed on a Bruker Avance 400 NMR spectrometer (Zurich, Switzerland) with a 4 mm rotor spinning module. FT-IR measure-ments were recorded on a Nicolet Nexus FT-IR spectrometer (Madison, WI, USA) using the KBr pellet technique over a range of 4000-400 cm −1 at 4 cm −1 resolution and 32 scans. X-ray photoelectron spectroscopy (XPS) analyses of the prepared PILs were performed on a Thermo Fisher Scientific K-Alpha spectrometer (Waltham, MA, USA) equipped with Al K radiation (1486.68 eV). Porous properties of the prepared PILs were measured at 77 K on a Micromeritics ASAP2460 after degassing at 100 • C for 8 h. A scanning electron microscope (FE-SEM, Hitachi S-4800, accelerateel voltage: 5 kV (Tokyo, Japan) with energy-dispersive X-ray spectroscopy (EDS) and a transmission electron microscope (TEM, FEI Talos F200X), operating at a high voltage of 200 kV, were used to investigate the morphology of the materials. Thermogravimetric analysis (TGA, Netzsch, Bavaria, Germany) was carried out using a NETZSCH STA 449F3 over a temperature range from 25 to 600 • C at a heating rate of 10 • C/min under a dynamic N 2 atmosphere. GC and GC-MS analyses were recorded on SCION 456-GC and Agilent 6890/5973 GC-MS, respectively.

DFT Calculation
DFT calculations were performed to investigate the hydrogen-bond interactions between epichlorohydrin and designed catalysts using the Gaussian 09 package. The geometry of all molecules was optimized by the B3LYP-GD3 level using the 6-31G* [25] basis set [71]. The same level of frequency analysis was used to ensure the minimum potential energy surface of the optimized geometry.

Conclusions
In summary, novel ionic polymers with different types of HBDs, high numbers of imidazolium ionic sites, and controlled spatial positions of dual active sites were facilely constructed by a one-pot quaternization strategy. Impressively, the presence of hydroxyl and carboxyl groups boosted the catalytic activities of ionic polymers significantly toward the cycloaddition reaction of CO 2 into epoxides. Carboxyl-functionalized ionic polymers displayed higher catalytic activity than those of HBDs-free and hydroxyl-functionalized ionic polymers. DFT calculation demonstrated that the formation of hydrogen bonds between epoxide with HBDs of ionic polymers facilitated the ring-opening step, and the higher catalytic activity of PIMBr-COOH was mainly attributed to its stronger hydrogen bond interaction with the epoxide substrate. Additionally, PIMBr-COOH exhibited good recyclability and was able to be reused at least five times. Under ambient CO 2 presence, a series of substituted epoxides were smoothly transformed into cyclic carbonates in excellent yields under solvent-, cocatalyst-, and metal-free reaction conditions. This research thus represents a reliable strategy for direct fabrication of carboxyl-functionalized ionic polymers towards efficient CO 2 conversion.

Conflicts of Interest:
There are no conflicts to declare.