Metal Embedded Porous Carbon for Efficient CO2 Cycloaddition under Mild Conditions

Nitrogen-doped porous carbon material was generated via thermal pyrolysis of zeolitic imidazole frameworks (ZIFs). The structure of the ZIF templates was tuned, so that the obtained product was an N-doped porous carbon-containing encapsulated metal nanoparticle. The hierarchical structural and unique properties of pyrolyzed materials are involved in further application, including catalysis. The as-synthesized porous carbon materials were applied as a catalyst for CO2 fixation on cyclic carbonates under near ambient pressure without solvent and co-catalyst. The zinc dispersion in highly porous carbon material, deriving from ZIF-8, exhibited a superior catalytic performance among the synthesized materials. The acid sites (Zn species) and the incorporated basic sites (N-species) present in the porous carbon material are essential for a high affinity for gas adsorption and CO2 conversion. Additionally, the catalyst was found to be very robust and stable during recycling studies as the catalytic performance remained high for seven cycles.


Introduction
Carbon materials have been wildly investigated and have broad applications in academic research and industrial applications. Among all existing carbon-based materials, porous carbon has been viewed as the most important material due to its specific characteristics, including (i) excellent chemical, thermal and mechanical stability, (ii) tunable porosity and surface functional chemistry, (iii) good electrical and thermal conductivity, (iv) high surface area, (v) structural heterogeneity and morphology, and (vi) low cost and uncomplicated handling in the manufacture [1,2]. Due to those captivating properties, porous carbon is used in various applications as an adsorbent, catalyst, electrode relevant to energy conversion, and environmental chemistry [3][4][5][6][7]. Different routes are used to synthesize porous carbon material, including activation (physical or chemical), carbonization of polymer aerogels, synthetic template procedure, etc. [8][9][10][11]. Simple thermal pyrolysis of various organic precursors under a controlled atmosphere is a general process to obtain porous carbon material [12].
Nevertheless, the disordered and non-interconnected structures of thermally synthesized carbon greatly limit their application. Recently, metal-organic frameworks (MOFs) and zeolite imidazolate frameworks (ZIFs) have been evidenced to be ideal sacrificial precursors for synthesizing various carbon-based materials. Metal-organic frameworks (MOFs) are among the most important and rapidly growing groups of porous materials. They are made of a combination of metal species and organic linkers. At the same time, zeolitic imidazole frameworks (ZIFs) represent one category of MOF materials comprised of imidazole linkers and metal ions (clusters), with a structure similar to zeolites. Due to various metal/organic linker species and the tunability of the frame, shape, and pore size, MOFs/ZIFs are very interesting and challenging [13][14][15]. Nevertheless, further development of suitable MOF materials for applications still requires improvement by several means, including chemical, physical, and thermal stability. Furthermore, the presence of the organic linkers in their framework possesses a high carbon content which can be seen as an excellent carbon source. Additionally, the metal ions/cluster and functional elements in the framework can transform the properties of carbon products during thermal decomposition (pyrolysis). A growing number of studies report the use of MOF/ZIF-template to synthesize porous carbon materials, especially in the energy and environmental areas, including batteries (e.g., lithium-ion, lithium-sulfur, and lithium-air batteries), supercapacitors, and gas adsorption and separation [16,17]. Although many studies on pyrolyzed MOF/ZIFtemplate have been reported in the past, few reports discuss their use in heterogeneous catalyses, such as oxidation, hydrogenation, dehydrogenation, bio-oil refining, and the Fischer-Tropsch process [18][19][20]. The metal/oxide nanoparticles doped porous carbon, derived from MOFs/ZIFs, could be advantageous for catalytic application in the heterogeneous phase. Using ZIF-67 as a template to derive cobalt in N-doped carbon (Co/CN) was reported as an efficient catalyst for organic transformations [21]. These hybrid materials exhibited excellent activity to convert nitroarenes under mild reaction conditions. However, the recyclability decreased in consecutive runs since the strongly adsorbed reaction species blocked the active sites. Concerning the CO 2 transformation, pyrolyzed ZnCo-ZIF with two metals embedded in a carbon matrix (Zn/Cn-ZIF-H 2 /Ar-1000) was recently applied as a robust heterogeneous catalyst for the CO 2 addition into epoxide [22]. The bimetallic system (Zn/Co) containing Lewis acid sites (metal ions) and basic sites from the N-doped carbon matrix, was demonstrated to be an active catalyst for the CO 2 addition reaction. The easy magnetic separation of the catalyst from the reaction mixture was an extra benefit of this robust heterogeneous catalyst.
Herein, we report on the development of N-doped porous carbonaceous materials derived from green synthesized ZIFs (ZIF-8, ZIF-67, and ZnCo-ZIF). The porous carbon materials act as robust heterogeneous catalysts with excellent activities and selectivities for the CO 2 fixation reaction under free-solvent and co-catalyst conditions at near atmospheric pressure. Interestingly, the N-containing carbonaceous material exhibits excellent catalytic performance. Metals dispersion and the N-sites doped in the hierarchical porous carbon, originating from the ZIF-8 material, are essential in the mechanism for the CO 2 fixation reaction. Moderate reaction conditions are required using the synthesized porous carbon as a catalyst for the CO 2 cycloaddition into epoxides (optimized at 80 • C, 1bar CO 2 pressure, solvent-free, and no co-catalyst). The designed catalyst has several advantages, e.g., highly selective, robust activity, solvent-free, recyclable, and can lead to a new approach in heterogeneous catalyst development for CO 2 transformation. Moderate reaction conditions were obtained by optimizing the catalyst loading, temperature, and reaction time under near ambient pressure.

Material Characterization
The ZIF materials (ZIF-8, ZIF-67, and Zn/Co-ZIF) were first synthesized and used as a template for N-doped porous carbon through the pyrolysis process. Using water as a solvent, a room temperature synthesis route was applied to synthesize the ZIFs template, since mild reaction conditions can be used in a green synthetic route. The synthesized ZIFs were characterized to confirm their characteristic properties, e.g., crystallinity, morphology, porosity, and thermal stability via X-ray diffraction (PXRD), SEM, BET, and TGA, respectively, as shown in Figures S2-S5. The characterization results revealed that the synthesized ZIFs had similar characteristic properties as reported in the literature [23,24]. After that, the ZIF used as template materials was pyrolyzed under controlled conditions. The pyrolysis process destroyed the crystalline structure of ZIFs (1000 • C for 1 h at a flow rate of 50 cm 3 ·min −1 of 5% H 2 in Ar) as confirmed by X-ray diffraction analysis (Figure 1a). The broad peak at about 23 • , observed in the pyrolyzed materials' diffraction pattern, was assigned to the carbon [002] diffraction. Next to the broad carbon peak, metallic crystal peaks were observed in the pyrolyzed material's diffraction pattern only for Co/CN and ZnCo/CN. In addition, the diffraction of metallic crystalline cobalt peaks (2θ of 44.3 • , 51.5 • , and 76.1 • ) was found to be more intense in Co/CN than ZnCo/CN, which can be related to the cobalt content of ZIF templates. Moreover, the metal sintering effect during the high pyrolysis temperature could also be a reason for the high diffraction intensity. On the contrary, a non-crystal diffraction pattern was observed for Zn/CN, which implied that the metal (Zn) is well-dispersed or that the sample is a metal-free porous carbon. As the pyrolysis temperature of Zn-ZIF is 1000 • C and zinc is known to evaporate at a temperature above 908 • C, there is a chance that the diffraction peaks of Zn may not be seen in the pyrolyzed sample [25,26]. The broad diffraction peak is assigned to amorphous carbon with a graphite-like carbon structure [27]. Furthermore, elemental analysis (ICP) was used to determine element content in the pyrolyzed material. The result revealed a tiny zinc amount in Zn/CN compared to the amount of zinc in the ZIF-8 template (Zn ≈ 28% wt.) ( Table 1). A well-dispersed tiny zinc content could result in the absence of zinc crystal diffraction peaks in Zn/CN. synthesized ZIFs had similar characteristic properties as reported in the literature [23,24]. After that, the ZIF used as template materials was pyrolyzed under controlled conditions. The pyrolysis process destroyed the crystalline structure of ZIFs (1000 °C for 1 h at a flow rate of 50 cm 3 ·min −1 of 5% H2 in Ar) as confirmed by X-ray diffraction analysis (Figure 1a).
The broad peak at about 23°, observed in the pyrolyzed materials' diffraction pattern, was assigned to the carbon [002] diffraction. Next to the broad carbon peak, metallic crystal peaks were observed in the pyrolyzed material's diffraction pattern only for Co/CN and ZnCo/CN. In addition, the diffraction of metallic crystalline cobalt peaks (2θ of 44.3°, 51.5°, and 76.1°) was found to be more intense in Co/CN than ZnCo/CN, which can be related to the cobalt content of ZIF templates. Moreover, the metal sintering effect during the high pyrolysis temperature could also be a reason for the high diffraction intensity.
On the contrary, a non-crystal diffraction pattern was observed for Zn/CN, which implied that the metal (Zn) is well-dispersed or that the sample is a metal-free porous carbon. As the pyrolysis temperature of Zn-ZIF is 1000 °C and zinc is known to evaporate at a temperature above 908 °C, there is a chance that the diffraction peaks of Zn may not be seen in the pyrolyzed sample [25,26]. The broad diffraction peak is assigned to amorphous carbon with a graphite-like carbon structure [27]. Furthermore, elemental analysis (ICP) was used to determine element content in the pyrolyzed material. The result revealed a tiny zinc amount in Zn/CN compared to the amount of zinc in the ZIF-8 template (Zn ≈ 28%wt.) ( Table 1). A well-dispersed tiny zinc content could result in the absence of zinc crystal diffraction peaks in Zn/CN.   The morphologies of the materials were investigated using a scanning electron microscope (SEM). The rhombic dodecahedron shape with a particle size varying from 50-200 nm, depending on the type of ZIF template, was observed (Figure 2a-c). After pyrolyzing the template material, shape deformation and shrinkage were obtained. At the same time, numerous smaller nanoparticles (bright spots) were detected, demonstrating the presence of Co nanoparticles for ZnCo/CN and Co/CN, respectively, as depicted in Figure 2e-f. Additionally, it is worth mentioning that carbon nanotubes (CNTs) appear on the Co/CN, inset Figure 2f. Interestingly, the Zn/CN observes a slight surface shrinkage that keeps the rhombic shape similar to the ZIF-8 template. More in-depth morphological analyses applying transmission electron microscopy (TEM) were performed, and shape deformation and shrinkage phenomena were observed in the pyrolyzed samples, namely, Zn/CN and Co/CN obtained from the ZIF-8 and ZIF-67 templates, respectively ( Figure 3). For the Zn/CN sample, no metallic zinc nanoparticles were observed, confirming the observation made by XRD (Figure 1b,c). The morphologies of the materials were investigated using a scanning electron microscope (SEM). The rhombic dodecahedron shape with a particle size varying from 50-200 nm, depending on the type of ZIF template, was observed (Figure 2a-c). After pyrolyzing the template material, shape deformation and shrinkage were obtained. At the same time, numerous smaller nanoparticles (bright spots) were detected, demonstrating the presence of Co nanoparticles for ZnCo/CN and Co/CN, respectively, as depicted in Figure 2e-f. Additionally, it is worth mentioning that carbon nanotubes (CNTs) appear on the Co/CN, inset Figure 2f. Interestingly, the Zn/CN observes a slight surface shrinkage that keeps the rhombic shape similar to the ZIF-8 template. More in-depth morphological analyses applying transmission electron microscopy (TEM) were performed, and shape deformation and shrinkage phenomena were observed in the pyrolyzed samples, namely, Zn/CN and Co/CN obtained from the ZIF-8 and ZIF-67 templates, respectively ( Figure 3). For the Zn/CN sample, no metallic zinc nanoparticles were observed, confirming the observation made by XRD (Figure 1b     The morphologies of the materials were investigated using a scanning electron microscope (SEM). The rhombic dodecahedron shape with a particle size varying from 50-200 nm, depending on the type of ZIF template, was observed (Figure 2a-c). After pyrolyzing the template material, shape deformation and shrinkage were obtained. At the same time, numerous smaller nanoparticles (bright spots) were detected, demonstrating the presence of Co nanoparticles for ZnCo/CN and Co/CN, respectively, as depicted in Figure 2e-f. Additionally, it is worth mentioning that carbon nanotubes (CNTs) appear on the Co/CN, inset Figure 2f. Interestingly, the Zn/CN observes a slight surface shrinkage that keeps the rhombic shape similar to the ZIF-8 template. More in-depth morphological analyses applying transmission electron microscopy (TEM) were performed, and shape deformation and shrinkage phenomena were observed in the pyrolyzed samples, namely, Zn/CN and Co/CN obtained from the ZIF-8 and ZIF-67 templates, respectively ( Figure 3). For the Zn/CN sample, no metallic zinc nanoparticles were observed, confirming the observation made by XRD (Figure 1b,c).   Interestingly, the TEM image of Zn/CN, Figure 3b, exhibits a uniform layered carbon matrix generated during the pyrolysis of ZIF-8. Moreover, the absence of zinc-based nanoparticles or clusters in the TEM image of Zn/CN suggests that Zn atoms could exist in an atomic dispersion or were eliminated during pyrolysis. While for Co/CN, metallic Co nanoparticles in various particle sizes (10-100 nm) were observed (Figure 3b). The TEM image demonstrates that the Co nanoparticles are embedded in layers of carbon (Figure 3b, inset picture), thus stabilizing the cobalt nanoparticles.
ZIFs are intrinsically porous materials, and their pyrolysis inevitably causes the escape of volatile compounds and metals, producing different levels of porosities. The porosity properties of pyrolyzed materials were analyzed by N 2 physisorption analysis. The adsorption below the relative pressure (P/P 0 ) < 0.05 significantly decreases for the Co/CN and ZnCo/CN compared to their templates (Figures 4 and S6). These results indicate a diminished micropore content of pyrolyzed materials. Consequently, the surface area of those pyrolyzed products is significantly reduced compared to their pristine templates ( Table 1). On the contrary, the N 2 isotherm at a low relative pressure (P/P 0 < 0.05) increases significantly for Zn/CN (>two times) compared to the ZIF-8 template. The template structure decomposes, and the organic linker (2-Methyl-Imidazole; 2-MIM) in the template framework gets transformed into carbon material during the pyrolysis process. Simultaneously, zinc elimination from the framework occurs, maintaining the pore structure, and causing high adsorption of N 2 on Zn/CN. The specific surface area values increased in the order Co/CN < ZnCo/CN < Zn/CN (Table 1, and Figure S6). Therefore, the N 2 isotherms are of type IV for the pyrolyzed materials, while the small hysteresis at the medium relative pressure (0.45 < P/P 0 < 1.0) suggests the presence of mesopores. These isotherms confirm the pyrolyzed materials' hierarchical porosity (mixed micro-mesoporosity). The material characteristics regarding the pores, spanning from the micro to the larger pores (meso or macropore), are consistent with reports on pyrolyzed MOFs or ZIFs [28]. Consequently, as given in Table 1, pyrolyzed ZIF-8 (Zn/CN) exhibited the highest surface area (1525 m 2 ·g −1 ) followed by ZnCo/CN (205 m 2 ·g −1 ), and Co/CN (151 m 2 ·g −1 ), respectively. Additionally, a difference in mesopore size distribution was observed in the order of Zn/CN > ZnCo/CN > Co/CN, respectively ( Figure S6), which could be attributed to the metal (Zn) elimination during the pyrolysis process. Moreover, the metal elimination during the pyrolysis process could enlarge the porosity compared to the metal embedded pyrolyzed product. The enhanced pore size (micro to mesopore) might probably be due to the metal ions/cluster mobility (e.g., agglomeration, sintering) and generating additional pores during the framework decomposition at the high temperature of the pyrolysis process. However, the high thermal stability of cobalt resulted in the metal agglomeration into crystalline nanoparticles and the embedment in carbon materials and blocking of micropores, reducing the pyrolyzed product's total pore and surface area. Interestingly, the TEM image of Zn/CN, Figure 3b, exhibits a uniform layered carbon matrix generated during the pyrolysis of ZIF-8. Moreover, the absence of zinc-based nanoparticles or clusters in the TEM image of Zn/CN suggests that Zn atoms could exist in an atomic dispersion or were eliminated during pyrolysis. While for Co/CN, metallic Co nanoparticles in various particle sizes (10-100 nm) were observed ( Figure 3b). The TEM image demonstrates that the Co nanoparticles are embedded in layers of carbon ( Figure  3b, inset picture), thus stabilizing the cobalt nanoparticles.
ZIFs are intrinsically porous materials, and their pyrolysis inevitably causes the escape of volatile compounds and metals, producing different levels of porosities. The porosity properties of pyrolyzed materials were analyzed by N2 physisorption analysis. The adsorption below the relative pressure (P/P0) < 0.05 significantly decreases for the Co/CN and ZnCo/CN compared to their templates ( Figure 4 and Figure S6). These results indicate a diminished micropore content of pyrolyzed materials. Consequently, the surface area of those pyrolyzed products is significantly reduced compared to their pristine templates ( Table 1). On the contrary, the N2 isotherm at a low relative pressure (P/P0 < 0.05) increases significantly for Zn/CN (>two times) compared to the ZIF-8 template. The template structure decomposes, and the organic linker (2-Methyl-Imidazole; 2-MIM) in the template framework gets transformed into carbon material during the pyrolysis process. Simultaneously, zinc elimination from the framework occurs, maintaining the pore structure, and causing high adsorption of N2 on Zn/CN. The specific surface area values increased in the order Co/CN < ZnCo/CN < Zn/CN (Table 1, and Figure S6). Therefore, the N2 isotherms are of type IV for the pyrolyzed materials, while the small hysteresis at the medium relative pressure (0.45 < P/P0 <1.0) suggests the presence of mesopores. These isotherms confirm the pyrolyzed materials' hierarchical porosity (mixed micro-mesoporosity). The material characteristics regarding the pores, spanning from the micro to the larger pores (meso or macropore), are consistent with reports on pyrolyzed MOFs or ZIFs [28]. Consequently, as given in Table 1, pyrolyzed ZIF-8 (Zn/CN) exhibited the highest surface area (1525 m 2 ·g −1 ) followed by ZnCo/CN (205 m 2 ·g −1 ), and Co/CN (151 m 2 ·g −1 ), respectively. Additionally, a difference in mesopore size distribution was observed in the order of Zn/CN > ZnCo/CN > Co/CN, respectively ( Figure S6), which could be attributed to the metal (Zn) elimination during the pyrolysis process. Moreover, the metal elimination during the pyrolysis process could enlarge the porosity compared to the metal embedded pyrolyzed product. The enhanced pore size (micro to mesopore) might probably be due to the metal ions/cluster mobility (e.g., agglomeration, sintering) and generating additional pores during the framework decomposition at the high temperature of the pyrolysis process. However, the high thermal stability of cobalt resulted in the metal agglomeration into crystalline nanoparticles and the embedment in carbon materials and blocking of micropores, reducing the pyrolyzed product's total pore and surface area.   The elemental analysis revealed that Zn/CN has the highest C percentage over ZnCo/CN and Co/CN (Tables 1 and S2). With the increasing pyrolysis temperature, zinc was eliminated from ZIF-8 and, to a lesser extent, from ZnCo-ZIF. Cobalt remained in ZnCo/CN and Co/CN after pyrolysis due to the higher evaporation temperature of cobalt. These effects could originate from the presence of different elements in the template materials. Nevertheless, pyrolysis reduced the N percentage in the samples compared to the percentage in their template. Co/CN has the highest N percentage, followed by ZnCo/CN and Zn/CN, respectively. The frameworks' thermal stability from high to low stability is given by: ZIF-67 > ZIF-8 > ZnCo-ZIF (TGA, Figure S4). These results imply that the strong M-N interaction enhances the N-content in pyrolyzed materials. The surface functional groups on pyrolyzed materials can differ according to the ZIF template's chemical environment, resulting in different chemical properties [29]. FT-IR analysis was performed on the pyrolyzed materials to gain information regarding their chemical bonding and surface functional groups. The FT-IR spectra reveal that the original M-N stretch modes (M: Zn or Co) at 421 cm −1 disappears in the pyrolyzed materials ( Figure S7). This result confirms that the framework in ZIF templates was destroyed during the pyrolysis process. However, peaks at 1608 and 1250 cm −1 , derivative stretching and bending mode of N-H, and C-N, respectively, confirm chemical bonding between carbon and nitrogen in the pyrolyzed materials [30]. Interestingly, all pyrolyzed materials typically exhibit a band at 3423 cm −1 , assigned to the N-H and/or O-H symmetric stretching vibration [31,32]. Furthermore, the band characteristic of the presence of an aromatic ring peak at 1608 cm −1 for the carbon material (C=C stretching vibration) is observed in the pyrolyzed materials.
XPS analysis of the pyrolyzed samples was performed to investigate the elemental composition and chemical states of various elements. The survey analyses revealed Zn, Co, C, O, and N in the samples, as shown in Figure 5a. The high-resolution N 1s spectra were deconvoluted into three peaks with centers at 397.5, 400.5, 401.6 eV, describing the contribution of coordinated quaternary pyridinic-N, pyrrolic-N, and ammonium-N, respectively ( Figure 5b). Different ZIF templates exhibited a variation in the ratio of N-species under the same pyrolysis conditions. The deconvolution of N-spectra reveals an increasing proportion of pyrrolic N species according to Co/CN < ZnCo/CN < Zn/CN. The opposite trend is observed for the ammonium-N species according to Co/CN > ZnCo/CN > Zn/CN. Notable, pyrrolic N species were reported earlier in the literature, having an excellent effect on the catalytic activity for the CO 2 cycloaddition with epoxides [22,33]. These pyrrolic N species, obtained from pyrolysis of ZIFs, serve as base sites, the active sites in the catalytic reaction (more discussion was provided under the section reaction mechanism). The Zn 2p high-resolution spectra display two peaks at 1021.1 eV and 1044.2 eV, assigned to Zn 2p 3/2 and Zn 2p 1/2 in Zn/CN and ZnCo/CN ( Figure S8). Zn/CN revealed an excellent catalytic activity among the N-doped porous carbon samples compared with ZnCo/CN and Co/CN. During pyrolysis, the ZIFs materials were heated at high temperatures under a controlled gas atmosphere. 2-methylimidazole, used to construct the ZIF material, served as the primary source of carbon formation with nitrogen content (CN). The characteristic materials indicate that pyrolysis conditions generated a different type of metal in N-doped porous carbon, which could involve the formation of active sites. Moreover, N-species are doped in the carbon, resulting in functionalized/doped carbon matrices creating unique properties of the generated carbon materials. The influence of different pyrolysis conditions on ZIF material (Zn/Co-ZIF synthesized via the spray drying method) was investigated in a previous report [22]. Metal oxide was generated in the presence of oxygen (air) during the thermal process. In comparison, during the pyrolysis process under an inert atmosphere, the metal ions/clusters became mobile and formed nanoparticles embedded in the carbon matrix derived from the organic ligand decomposition. Moreover, under a reduced atmosphere such as H 2 /Ar mixture atmosphere, metallic species were obtained, which exhibited the best catalytic performance for the CO 2 cycloaddition with epoxides into cyclic carbonates [28]. During pyrolysis, the ZIFs materials were heated at high temperatures under a controlled gas atmosphere. 2-methylimidazole, used to construct the ZIF material, served as the primary source of carbon formation with nitrogen content (CN). The characteristic materials indicate that pyrolysis conditions generated a different type of metal in N-doped porous carbon, which could involve the formation of active sites. Moreover, N-species are doped in the carbon, resulting in functionalized/doped carbon matrices creating unique properties of the generated carbon materials. The influence of different pyrolysis conditions on ZIF material (Zn/Co-ZIF synthesized via the spray drying method) was investigated in a previous report [22]. Metal oxide was generated in the presence of oxygen (air) during the thermal process. In comparison, during the pyrolysis process under an inert atmosphere, the metal ions/clusters became mobile and formed nanoparticles embedded in the carbon matrix derived from the organic ligand decomposition. Moreover, under a reduced atmosphere such as H2/Ar mixture atmosphere, metallic species were obtained, which exhibited the best catalytic performance for the CO2 cycloaddition with epoxides into cyclic carbonates [28].

Catalytic Activity
Due to the regular arrangement of elemental species in ZIF precursors, the carbon composites contain nanostructured species prone to be formed during the in situ pyrolysis process. Furthermore, the N present in the imidazole linker could effectively functionalize the carbon structure. These facile synthetic approaches provide a new way to prepare porous carbon. Carbon materials derived from ZIFs have been widely explored for electrochemical energy storage applications with specific surface areas and capacitance performances [34][35][36]. The pyrolyzed materials demonstrate excellent chemical and thermal resistance, which are clear advantages over the template materials (ZIFs). Thus, promising applications not possible with common MOFs (ZIFs) materials are now within reach. Inspired by those reasons, the functional carbon materials were applied as heterogeneous catalysts for CO2 fixation. Furthermore, milder reaction conditions (near ambient pressure) were used to pursue a "green reaction" without solvents and a co-catalyst. A "green reaction" would reduce byproducts, pollution, and risks to human health. Consequently, a highly selective heterogeneous catalyst was made without the need for downstream processes. The catalyst also exhibited excellent recyclability behavior in catalytic performance and stability, confirming that a truly heterogeneous catalytic system was created.

Catalytic Activity
Due to the regular arrangement of elemental species in ZIF precursors, the carbon composites contain nanostructured species prone to be formed during the in situ pyrolysis process. Furthermore, the N present in the imidazole linker could effectively functionalize the carbon structure. These facile synthetic approaches provide a new way to prepare porous carbon. Carbon materials derived from ZIFs have been widely explored for electrochemical energy storage applications with specific surface areas and capacitance performances [34][35][36]. The pyrolyzed materials demonstrate excellent chemical and thermal resistance, which are clear advantages over the template materials (ZIFs). Thus, promising applications not possible with common MOFs (ZIFs) materials are now within reach. Inspired by those reasons, the functional carbon materials were applied as heterogeneous catalysts for CO 2 fixation. Furthermore, milder reaction conditions (near ambient pressure) were used to pursue a "green reaction" without solvents and a co-catalyst. A "green reaction" would reduce byproducts, pollution, and risks to human health. Consequently, a highly selective heterogeneous catalyst was made without the need for downstream processes. The catalyst also exhibited excellent recyclability behavior in catalytic performance and stability, confirming that a truly heterogeneous catalytic system was created.

Catalytic Studies
In this study, the porous carbon materials derived from ZIFs were applied as a catalyst for the CO 2 cycloaddition into epoxides to produce cyclic carbonates. The catalytic performance of the various N-doped porous carbon products was assessed using epichlorohydrin as a model substrate, as shown in Figure 6. The synthesized carbon materials exhibited a significantly different catalytic activity under the same reaction conditions. The Zn/CN revealed the highest catalytic performance among comparative catalysts. Whereas no activity was observed using reference carbon as a catalyst, similar to the blank reaction (without catalyst). These results show the significance of the catalyst in the model reaction. To illustrate the special properties of the synthesized N-doped porous carbon, commercial carbon and starting precursors were investigated (Table S3, entries 9-15). It must be mentioned that ZIF templates, when used as catalysts, achieved excellent catalytic performance (Table S3, entries 3-8); however, the decomposition, metal leaching, and organic ligand contamination in the product are severe drawbacks of those catalysts. Meanwhile, the metal species (ZIF-8, ZIF-67, ZnCo-ZIF, ZnO, ZnCO 3 , pyrolyzed Zn(NO 3 ) 2 ·6H 2 O, Co(NO 3 ) 2 ·6H 2 O), and pyrolyzed Co(NO 3 ) 2 ·6H 2 O), which might be present in pyrolyzed materials from the ZIF templates, were also applied as a catalyst in the model reaction. No or a minimal yield of cyclic carbonate (6% yield with ZnO) was obtained, see Table S3. The catalytic activity of commercial carbon materials was also studied. Nevertheless, no product was obtained with carbon, while a small product yield (5%) was acquired with carbon graphite oxide as a catalyst in a similar condition (Table S3). Nevertheless, impurities deriving from the synthesis of commercial carbon graphite oxide (such as iron) and the CO 2 adsorption ability of graphite oxide might provide some catalytic activity as well [37,38]. Remarkably, the 2-methylimidazole (2-MIM), the linker in the ZIF structure, performed excellently (100% Conv., 90% Yield). This result evidenced that the basic sites of 2-MIM can activate CO 2 (see discussion of the mechanism). However, it is impossible to recycle 2-MIM as a catalyst from the reaction mixture, which is the main drawback of applying 2-MIM as a catalyst. An alternative is the removal of the obtained cyclic carbonate by distillation, after which the remaining 2-MIM can be reused. However, the latter procedure is energy demanding. Therefore, all results reveal that the pyrolyzed carbon materials derived from ZIFs, especially Zn/CN, show excellent catalytic activities.
while, the metal species (ZIF-8, ZIF-67, ZnCo-ZIF, ZnO, ZnCO3, pyrolyz Zn(NO3)2·6H2O, Co(NO3)2·6H2O), and pyrolyzed Co(NO3)2·6H2O), which might be pres in pyrolyzed materials from the ZIF templates, were also applied as a catalyst in the mo reaction. No or a minimal yield of cyclic carbonate (6% yield with ZnO) was obtained, Table S3. The catalytic activity of commercial carbon materials was also studied. Nev theless, no product was obtained with carbon, while a small product yield (5%) was quired with carbon graphite oxide as a catalyst in a similar condition (Table S3). Nev theless, impurities deriving from the synthesis of commercial carbon graphite oxide (su as iron) and the CO2 adsorption ability of graphite oxide might provide some cataly activity as well [37,38]. Remarkably, the 2-methylimidazole (2-MIM), the linker in the Z structure, performed excellently (100% Conv., 90% Yield). This result evidenced that basic sites of 2-MIM can activate CO2 (see discussion of the mechanism). However, i impossible to recycle 2-MIM as a catalyst from the reaction mixture, which is the m drawback of applying 2-MIM as a catalyst. An alternative is the removal of the obtain cyclic carbonate by distillation, after which the remaining 2-MIM can be reused. Howev the latter procedure is energy demanding. Therefore, all results reveal that the pyrolyz carbon materials derived from ZIFs, especially Zn/CN, show excellent catalytic activit  Table S3. Particularly for commercial applications, the heterogeneous catalyst shows an ess tial advantage over the homogeneous catalysts, namely for catalyst separation and rec ery. Another major concern of the catalyst is its stability. The reusability of the catal was investigated by using Zn/CN. The spent catalyst was separated from the react mixture via centrifugal and washed with methanol five times before drying in a vacuu  Table S3. Particularly for commercial applications, the heterogeneous catalyst shows an essential advantage over the homogeneous catalysts, namely for catalyst separation and recovery. Another major concern of the catalyst is its stability. The reusability of the catalyst was investigated by using Zn/CN. The spent catalyst was separated from the reaction mixture via centrifugal and washed with methanol five times before drying in a vacuum oven at 100 • C overnight. The weight of the recovered catalyst was measured before reuse for each subsequent experiment (run 2-7). The perceived results indicate that each cycle's catalytic activity remains excellent (conversion and product yield), as shown in Figure 7.
Moreover, the catalyst weight was maintained after each recovery. The characteristic properties of the catalyst after reaction were determined, such as XRD, XPS, N 2 adsorption, porosity, and surface area analysis. The characteristic results were similar to the fresh catalyst ( Figures S10 and S11). These results evidence that Zn/CN is a robust catalyst that can be reused without losing catalyst weight and catalytic performance. Nevertheless, the conversion decreased slightly using a double amount of ECH moles under similar reaction conditions using Zn/CN (Table S3,  case of certain epoxides cannot be excluded, as was evidenced by some broad peaks in the 1 H NMR spectra (see Figures S19 and S21). However, a reduced yield (less than stoichiometric yield) is mainly associated with strong chemisorption of cyclic carbonates by the oxygen functionalities present on the pyrolyzed catalyst surface, or adsorption in the pores of the catalyst (high micro/mesopore porosity). Several catalytic systems have been reported for cycloaddition of CO2 into epoxides, including carbon-based materials.

Reaction Mechanism
Several studies in the literature report that acid sites originate from transition metals, while the basic sites derive from the N-species in the carbon catalyst made from MOFs/ZIFs. Both acid and basic sites are the active sites for the CO2 cycloaddition [22,39]. Moreover, it is well known that an organic base can activate CO2, and this activated species, in its turn, acts as a Lewis base to open the epoxide ring [40][41][42]. The catalytic mechanism describing the basic sites as the activating species for epoxides in the cycloaddition with CO2 to cyclic carbonates was recently reported for the N-doped carbon catalyst (carbon material) [42,43]. Herein, the catalytic performance of Zn/CN demonstrated the highest catalytic activity among the pyrolyzed catalysts (Co/CN and ZnCo/CN). To gain insight into the basic properties of the catalytic material, chemisorption analysis (CO2-TPD) Figure 7. The recyclable catalyst using 9.2 mmol ECH, catalyst Zn/CN (50 mg) at 90 • C, 1 bar CO 2 pressure for 24 h. Conversion and yields (isolated) were determined by 1 H NMR and using 1,3,5-trimethylbenzene as the internal standard.
The catalytic performance of the applied carbon materials with the reaction conditions and catalytic outcome are summarized in Table S4. Next, the performance of Zn/CN was investigated for different epoxides, and excellent conversions were obtained (Figure 6b, Table S3, entries [16][17][18][19][20][21][22][23][24][25]. The substrates containing electron-withdrawing substituents exhibit a higher conversion via an enhanced easiness of the ring-opening of the epoxide, allowing the nucleophilic attack followed by carbonate formation. The results confirm that a high selectivity is obtained using Zn/CN as a catalyst. Nonetheless, it needs to be mentioned that the formation of some amounts of oligomeric or polymeric carbonates in the case of certain epoxides cannot be excluded, as was evidenced by some broad peaks in the 1 H NMR spectra (see Figures S14-S18). However, a reduced yield (less than stoichiometric yield) is mainly associated with strong chemisorption of cyclic carbonates by the oxygen functionalities present on the pyrolyzed catalyst surface, or adsorption in the pores of the catalyst (high micro/mesopore porosity). Several catalytic systems have been reported for cycloaddition of CO 2 into epoxides, including carbon-based materials.

Reaction Mechanism
Several studies in the literature report that acid sites originate from transition metals, while the basic sites derive from the N-species in the carbon catalyst made from MOFs/ZIFs. Both acid and basic sites are the active sites for the CO 2 cycloaddition [22,39]. Moreover, it is well known that an organic base can activate CO 2 , and this activated species, in its turn, acts as a Lewis base to open the epoxide ring [40][41][42]. The catalytic mechanism describing the basic sites as the activating species for epoxides in the cycloaddition with CO 2 to cyclic carbonates was recently reported for the N-doped carbon catalyst (carbon material) [42,43]. Herein, the catalytic performance of Zn/CN demonstrated the highest catalytic activity among the pyrolyzed catalysts (Co/CN and ZnCo/CN). To gain insight into the basic properties of the catalytic material, chemisorption analysis (CO 2 -TPD) was performed, and the obtained profiles are depicted in Figure S9. The CO 2 desorption profile revealed that the adsorbed CO 2 mainly desorbs between 510 • C and 684 • C, confirming the strong basic properties of the synthesized materials. Interestingly, the CO 2 -TPD profile obtained from Zn/CN reveals that a significant part of the adsorbed CO 2 is desorbed at a lower temperature (510 • C) compared to ZnCo/CN (684 • C). In contrast, the smallest amount of desorption was observed in the profile obtained from Co/CN. The basic properties, generated from the N-species of the linker (2-MIM), could be enhanced by the metal elimination during the pyrolysis process [44]. Although the calculated distributions of the different N-types were obtained via XPS analysis, it is clear that Zn/CN possesses the highest amount of pyrrolic N (Figure 5b). This characteristic property (basicity) is significantly related to the increasing catalytic performance.
Moreover, the different starting compounds to construct the catalytic material were investigated, and complete conversion was observed using only the linker (2-MIM) as a catalyst (Table S3, entries 3-15). This high conversion is attributed to the basic sites present in 2-MIM. The catalytic reaction occurs at basic sites (N-doped porous carbon) together with the metals (acid sites, e.g., Co or Zn) as active centers. It is also well known that base molecules can adsorb CO 2 molecules, e.g., amines are used in commercial scrubbers to remove CO 2 from gas streams. In this study, 2-methylimidazole in the ZIF-8 structure was converted to porous carbon functionalized with N-species during pyrolysis. Those N-species doped in the porous carbon reveal basic properties of the synthesized materials, while the metal clusters (zinc) were partially removed from the material during the pyrolysis process. Consequently, the acid sites (zinc) and the basic sites present in the catalytic material take part in the catalytic reaction for the CO 2 cycloaddition reaction.
Based on the obtained experimental results and previously published reports, a plausible reaction mechanism was proposed for the CO 2 cycloaddition over metal embedded N-doped porous carbon catalysts (Figure 8). The reaction mechanism for the CO 2 cycloaddition can be attributed to the presence of the Lewis acid sites originating from metal species (Zn or Co) activating the epoxide substrate (ring-opening). Meanwhile, the lone pair electrons of the nitrogen act as a base, interacting with the electron-deficient carbon in CO 2 [45]. Nitrogen species could act as electron donors in carbon material, causing a shift in the Fermi level to the valence bands [46]. Next, a nucleophilic attack of N-species occurs on the carbon atom of CO 2 . The oxygen of the activated carbon dioxide binds with the carbon of the opened epoxide, forming an intermediate species. Finally, the cyclic carbonate molecule is formed via a nucleophilic attack of the oxygen from the opened epoxide during which the weak N-C dative bond will be disrupted, resulting in the product's easy desorption due to the nitrogen's conjugated nature in the carbon matrix, and the vacated sites for the following reaction cycle are generated. Moreover, the catalytic activity for the CO 2 cycloaddition into epoxides strongly depends on the type and relative abundance of the surface functional groups (here, pyrrolic N) [31,47,48]. The pyrrolic N is the most active site compared with the other N species for the CO 2 fixation reaction [22].

Materials and Synthesis
All chemical reagents for synthesis materials and catalytic reaction were purchased from Aladdin chemical and used as received without further purification. To prepare Ndoped porous carbonaceous materials, ZIFs (ZIF-8, ZnCo-ZIF, and ZIF-67) were firstly synthesized before using them as template material for the pyrolysis process ( Figure 9). The ZIFs were synthesized via a green method using water as a solvent at room tempera-

Materials and Synthesis
All chemical reagents for synthesis materials and catalytic reaction were purchased from Aladdin chemical and used as received without further purification. To prepare N-doped porous carbonaceous materials, ZIFs (ZIF-8, ZnCo-ZIF, and ZIF-67) were firstly synthesized before using them as template material for the pyrolysis process ( Figure 9). The ZIFs were synthesized via a green method using water as a solvent at room temperature with a slightly modified synthesis procedure from the previous report [23]. A typical synthesis of ZIF-8 and ZIF-67, 10 mmol of metal source (Zn(NO 3 ) 2 ·6H 2 O or Co(NO 3 ) 2 ·6H 2 O) and 80 mmol of ligand source (2-methylimidazole, Hmim) were separately dissolved in 100 mL of deionized water. For the bimetallic ZIF, a 50:50 ratio of Zn(NO 3 ) 2 ·6H 2 O and Co(NO 3 ) 2 ·6H 2 O with a constant total molar ratio (10 mmol) was applied for the ZnCo-ZIF synthesis. Triethylamine (5 mL) as a deprotonating agent was added to an aqueous Hmim solution before adding to the aqueous metal solution. The mixture was vigorously stirred at room temperature for 24 h, and then the precipitated solid was separated using a centrifuge (9500 rpm, 5 min). The obtained solids were washed with deionized water three times or until the obtained solution after centrifugal was clear. The solid products were dried at 100 • C under vacuum overnight before further use or analysis. After that, the ZIFs templates were pyrolyzed to synthesize carbonaceous materials at 1000 • C in a quartz tube furnace using a heating rate of 10 • C/min under 5% H 2 in Ar (50 cm 3 /min) for 1 h. After cooling to room temperature, the products were collected and stored in a desiccator to prevent moisture sorption. The obtained carbonaceous products were labeled as Zn/CN, ZnCo/CN, and Co/CN, deriving from template material of ZIF-8, ZnCo-ZIF, and ZIF-67, respectively. Furthermore, more details of synthesis materials are provided in the supporting information.
Catalysts 2022, 12, x FOR PEER REVIEW 1 Figure 9. Schematic presentation of the synthesis procedure of N-doped porous carbonaceo terials deriving from zeolitic imidazole frameworks.

Material Characterization
The crystal structure of synthesized materials was analyzed by powder X-ray d tion (Bruker D8 Advance diffractometer, Bragg-Brentano geometry, Cu Kα radiation data were typically collected in the 2θ range of 10-80°, scanning step 0.02°, and 0.2 step. Scanning electron microscope (FE-SEM, Zeiss Ultra Plus) and transmission el microscopy (TEM, JEM-2100F) with energy-dispersive X-ray spectroscopy (EDS), o ing at a high voltage of 200 kV, were carried out to investigate the morphology materials. The surface area and adsorption-desorption isotherm measurements we formed using a Micromeritics instrument (ASAP 2020) at 77 K using liquid nitroge coolant. Before the adsorption measurements, the samples (50-100 mg) were evacua 180 °C under a dynamic vacuum for about 3 h. The micropore surfaces were calc by the Brunauer-Emmett-Teller (BET) and Langmuir method (0.005 < P/P0 < 0.05 Horvath-Kawazoe method was applied for micropore analysis. X-ray photoelectron troscopy (XPS) measurements were conducted on a Kratos Axis Ultra DLD (dela detector) spectrometer equipped with a monochromatic Al Kα X-ray source (1486 All binding energies were referenced to the C1s peak at 284.9 eV of the surface ad tious carbon. The CasaXPS Processing software was applied to deconvolute the r Thermogravimetric analyses were carried out on a Netzsch (STA449c/3/G) instrume ing a heating rate of 10 °C min −1 under a nitrogen atmosphere. Fourier transformed red spectra (FT-IR) were recorded on a Nicolet 6700 FT-IR spectrometer with KBr in the wave range of 4000 to 400 cm −1 . The metals content was determined via induc

Material Characterization
The crystal structure of synthesized materials was analyzed by powder X-ray diffraction (Bruker D8 Advance diffractometer, Bragg-Brentano geometry, Cu Kα radiation). The data were typically collected in the 2θ range of 10-80 • , scanning step 0.02 • , and 0.2 s per step. Scanning electron microscope (FE-SEM, Zeiss Ultra Plus) and transmission electron microscopy (TEM, JEM-2100F) with energy-dispersive X-ray spectroscopy (EDS), operating at a high voltage of 200 kV, were carried out to investigate the morphology of the materials. The surface area and adsorption-desorption isotherm measurements were performed using a Micromeritics instrument (ASAP 2020) at 77 K using liquid nitrogen as a coolant. Before the adsorption measurements, the samples (50-100 mg) were evacuated at 180 • C under a dynamic vacuum for about 3 h. The micropore surfaces were calculated by the Brunauer-Emmett-Teller (BET) and Langmuir method (0.005 < P/P 0 < 0.05). The Horvath-Kawazoe method was applied for micropore analysis. X-ray photoelectron spectroscopy (XPS) measurements were conducted on a Kratos Axis Ultra DLD (delay-line detector) spectrometer equipped with a monochromatic Al Kα X-ray source (1486.6 eV). All binding energies were referenced to the C1s peak at 284.9 eV of the surface adventitious carbon. The CasaXPS Processing software was applied to deconvolute the results. Thermogravimetric analyses were carried out on a Netzsch (STA449c/3/G) instrument using a heating rate of 10 • C min −1 under a nitrogen atmosphere. Fourier transformed infrared spectra (FT-IR) were recorded on a Nicolet 6700 FT-IR spectrometer with KBr pellets in the wave range of 4000 to 400 cm −1 . The metals content was determined via inductively coupled plasma optical emission spectrometry (ICP-OES, Prodigy 7). The carbon and nitrogen elements were analyzed by CHN element analysis (Vario EL cube). The basic properties of the materials were examined via temperature-programmed desorption (TPD) (AutoChem II 2920) using NH 3 as the probe gases. The sample was first pre-treated at 150 • C under He flow for 3 h before TPD analysis. (Details are available in the Supplementary Materials).

Cycloaddition Reaction
For the catalytic reactions, 9.2 mmol of epoxide and 50 mg of catalyst were charged in a reactor (high-pressure glass bottle, 15 mL) for the low-pressure experiment, and in a stainless steel high-pressure reactor (15 mL, XINGDA Company, Beijing, China) for the high-pressure experiments. Subsequently, the reactor was purged for a few seconds before pressuring with CO 2 (99.9%) via a pressure regulator and keeping a constant pressure along with the reaction time. The reactor was immersed in a pre-heat oil bath with a thermocouple (IKA RCT basic) under a setting reaction temperature and time. After that, the reactor was cooled to room temperature, and the excess gas was vented out. The 1,3,5-trimethylbenzene, as the internal standard (3 mmol), and CDCl 3 (1 mL) were added to the reaction mixture. The final product was analyzed by nuclear magmatic resonance (Bruker NMR 500 MHz: 1 H and 13 C). For a recycling experiment, the catalyst was separated from the reaction mixture via centrifugal and washed several times with methanol before drying under vacuum at 90 • C for 6 h before the subsequent use (Details are available in the Supplementary Materials).

Conclusions
Hierarchical porous N-doped carbon materials derived from zeolitic imidazolate frameworks as template materials were synthesized and characterized. Metal embedded in N-doped porous carbon materials was obtained via tuning structural ZIF templates (pyrolysis). The metallic dispersion (Zn, Co) with the N-species (pyrrolic-, pyridinic-, and ammonium-N) in hierarchically porous carbon demonstrated to be the key active sites over the embedded metal species to catalyze the cycloaddition of CO 2 with epoxides into cyclic carbonates. The pyrrolic N becomes dominant over pyridinic and ammonium N-species in Zn/CN at a higher pyrolysis temperature, demonstrating the importance of catalytic sites for the metal-and N-doped porous carbon catalysts. The lack of N-species in commercial carbon, metal sources, etc., resulted in no reaction products, confirming the significance of N-species for the catalytic reaction. Moreover, the metal de-coordination and partial elimination from the template framework provided different N-species, as well as metallic species and hierarchical pore structures that enhanced access to the inner active sites. The synthesized catalyst exhibited a high catalytic performance combined with thermal and chemical stability, as demonstrated by the recycling experiments (>7 cycles). Furthermore, the optimized reaction conditions, near ambient pressure, can be applied without solvent and co-catalyst, providing a green procedure based on a heterogeneous catalyst to produce cyclic carbonates by the synergistic activation of CO 2 and epoxide substrates. Therefore, two novelties can be highlighted. Firstly, in the catalyst synthesis, the template ZIFs were synthesized under room temperature and organic solvent-free (aqua-solution). A facile method was processed via the thermal treatment of template ZIFs to derive metal decoordination and encapsulation in N-doped carbon. Secondly, the synthesized materials were directly used as catalysts without post-treatment or additional functional requirements in the catalytic reaction. The catalysts showed good performance under mild conditions at 80-90 • C and atmospheric (1 bar of CO 2 pressure) without solvent and co-catalyst. The catalyst characterization reveals well-dispersed active sites in a hierarchical porous (micro and mesoporous) material. The catalysts' high surface area promotes the mass transfer to access the active species and substrates. The "green" catalyst is a robust heterogeneous catalyst that maintains excellent performance for multiple runs.
Author Contributions: Methodology, investigation, data curation by C.Q. and M.L.; writingoriginal draft by J.W.; conceptualization, methodology, investigation, writing-original draft by S.C.; writing-review by S.Z.; conceptualization, resources, supervision, writing-review and editing by F.V. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.