Rigid Nanoporous Urea-Based Covalent Triazine Frameworks for C2/C1 and CO2/CH4 Gas Separation

C2/C1 hydrocarbon separation is an important industrial process that relies on energy-intensive cryogenic distillation methods. The use of porous adsorbents to selectively separate these gases is a viable alternative. Highly stable covalent triazine frameworks (urea-CTFs) have been synthesized using 1,3-bis(4-cyanophenyl)urea. Urea-CTFs exhibited gas uptakes of C2H2 (3.86 mmol/g) and C2H4 (2.92 mmol/g) at 273 K and 1 bar and is selective over CH4. Breakthrough simulations show the potential of urea-CTFs for C2/C1 separation.


Introduction
The separation of C1 and C2 gases is a critical process in many industrial activities. For example, acetylene is an important industrial byproduct of petroleum and natural gas processing, which needs to be separated. In addition, there are other industrial processes wherein ethylene and acetylene are produced by the oxidative and non-oxidative coupling of methane [1]. However, quite often the methane conversion remains incomplete and recovering the unreacted methane is essential [2]. As another example, in the process of extracting natural gas, methane needs to be separated from carbon dioxide [3]. Natural gas consists of high amounts of carbon dioxide that must be removed to obtain pure methane, which can be used as an energy source for fuels and chemicals. As a final example, the separation of CO 2 in flue gases (typically containing about 75% nitrogen and traces of water (vapor)) is becoming an important process in carbon capture and utilization CCU strategies.
Metal organic frameworks (MOFs) have been studied for this purpose [4][5][6]. Although some show very high adsorption capacities and selectivities, they often lack longterm stability, an important factor for a potential adsorbent [7]. Hence, other types of porous adsorbents, such as porous organic polymers are also considered for such gas separation processes.
However, these harsh ionothermal synthesis conditions result in materials that are mostly amorphous. In some reports, CTFs with a partial crystallinity were obtained [10,12]. Due to carbonization at these high temperatures, the structural characterization of the material becomes arduous. Nonetheless, CTFs exhibit exceptional properties in comparison to other covalent organic framework (COFs) for the above noted applications. One of the most appealing properties of the ionothermally synthesized CTFs is their exceptionally high thermal, hydrothermal, and chemical stability. They can withstand temperatures up to 550 • C and extreme chemical environments, such as 1 M NaOH or 1 M HCl solutions, for an extended period. Such stability is important for "real life" gas adsorption/separation applications, wherein high temperatures and the acid/base poisoning of the gas streams are important considerations [30]. This encouraged us to design new CTFs, particularly with polar functional sites that can be beneficial for gas storage and separations. Recently, Yaghi et al., reported the first urea-linked ketoenamine COFs and highlighted their structural dynamics with respect to their flexibility [31]. However, the COFs were not highly stable in basic conditions (1 M NaOH). In order to develop porous materials that are stable under both strong acidic and basic conditions, we report herein the synthesis of ultra-stable ureabased CTFs using a dinitrile linker, (1,3-bis(4-cyanophenyl)urea) (Scheme 1). We studied their surface properties as well as their potential for C 2 H 2 , C 2 H 4 , and CO 2 separation over CH 4 . We report herein that urea-CTFs display high C 2 H 2 and C 2 H 4 uptakes and moderate CO 2 adsorption capacity in comparison to the existing CTFs. Moreover, the C2 hydrocarbon (C 2 H 2 and C 2 H 4 ) adsorption was selective compared to C1 hydrocarbon (CH 4 ). In addition, urea-CTFs also exhibited good selectivity for CO 2 over CH 4 .
Molecules 2021, 26, x FOR PEER REVIEW 2 of 8 produced materials show good properties for several applications in gas storage and separation [23], catalysis [24], electrocatalysis [25,26], photocatalysis [27,28], and batteries [29]. However, these harsh ionothermal synthesis conditions result in materials that are mostly amorphous. In some reports, CTFs with a partial crystallinity were obtained [10,12]. Due to carbonization at these high temperatures, the structural characterization of the material becomes arduous. Nonetheless, CTFs exhibit exceptional properties in comparison to other covalent organic framework (COFs) for the above noted applications. One of the most appealing properties of the ionothermally synthesized CTFs is their exceptionally high thermal, hydrothermal, and chemical stability. They can withstand temperatures up to 550 °C and extreme chemical environments, such as 1 M NaOH or 1 M HCl solutions, for an extended period. Such stability is important for "real life" gas adsorption/separation applications, wherein high temperatures and the acid/base poisoning of the gas streams are important considerations [30]. This encouraged us to design new CTFs, particularly with polar functional sites that can be beneficial for gas storage and separations. Recently, Yaghi et al., reported the first urea-linked ketoenamine COFs and highlighted their structural dynamics with respect to their flexibility [31]. However, the COFs were not highly stable in basic conditions (1 M NaOH). In order to develop porous materials that are stable under both strong acidic and basic conditions, we report herein the synthesis of ultra-stable urea-based CTFs using a dinitrile linker, (1,3-bis(4-cyanophenyl)urea) (Scheme 1). We studied their surface properties as well as their potential for C2H2, C2H4, and CO2 separation over CH4. We report herein that urea-CTFs display high C2H2 and C2H4 uptakes and moderate CO2 adsorption capacity in comparison to the existing CTFs. Moreover, the C2 hydrocarbon (C2H2 and C2H4) adsorption was selective compared to C1 hydrocarbon (CH4). In addition, urea-CTFs also exhibited good selectivity for CO2 over CH4. Scheme 1. Schematic representation of the synthesis and ideal structure of the urea-functionalized CTFs. 1,3-bis(4-cyanophenyl)urea is a flexible linker, and its possible conformations are listed.

Synthesis and Characterization of Urea-CTFs
For the synthesis of the targeted urea-based CTFs, the linker 1,3-bis(4-cyanophenyl)urea was synthesized from 4-aminobenzonitrile according to the reported procedure [32]. In general, urea-CTFs were obtained through ionothermal synthesis using ZnCl2 (5 eq.) both as a catalyst and a solvent at 400 °C (urea-CTF-400-5) and 500 °C (urea-CTF-500-5) (Scheme 1, ESI). The complete trimerization of the cyano (-CN) groups was confirmed through Fourier transform infrared (FTIR) analysis ( Figure S1) where the -CN peak at 2226 cm −1 of the monomer is no longer visible in the CTFs [10,14]. In addition, triazine peaks were observed around 1360 cm −1 and 1600 cm −1 , which further Scheme 1. Schematic representation of the synthesis and ideal structure of the urea-functionalized CTFs. 1,3-bis(4-cyanophenyl)urea is a flexible linker, and its possible conformations are listed.

Synthesis and Characterization of Urea-CTFs
For the synthesis of the targeted urea-based CTFs, the linker 1,3-bis(4-cyanophenyl)urea was synthesized from 4-aminobenzonitrile according to the reported procedure [32]. In general, urea-CTFs were obtained through ionothermal synthesis using ZnCl2 (5 eq.) both as a catalyst and a solvent at 400 • C (urea-CTF-400-5) and 500 • C (urea-CTF-500-5) (Scheme 1, ESI). The complete trimerization of the cyano (-CN) groups was confirmed through Fourier transform infrared (FTIR) analysis ( Figure S1) where the -CN peak at 2226 cm −1 of the monomer is no longer visible in the CTFs [10,14]. In addition, triazine peaks were observed around 1360 cm −1 and 1600 cm −1 , which further confirm the successful trimerization. Notably, a small broad peak around 1707 cm −1 was observed in the CTFs that are red-shifted from 1737 cm −1 of C(O) monomer and confirms the presence of urea groups in the resulting materials [31]. The observed lower wavenumber might be due to the decrease in the double-bond character of the C(O) bond of the urea functional group after the CTF formation.
The porous properties of both the CTF materials were explored using argon sorption at 87 K ( Figure 1) and N 2 sorption measurements at 77 K ( Figure S2). Both urea-CTF_400_5 and urea-CTF_500_5 displayed a Type I isotherm typical for microporous materials, and the calculated BET surface areas were 555 m 2 g −1 and 928 m 2 g −1 , respectively. The detailed textural properties are described in Table S1. As seen in several reported CTFs [11], the microporosity content depends on the synthesis temperature, whereas, urea-CTF_400_5 shows a higher microporous-to-mesoporous volume ratio in comparison to Urea-CTF_500_5. The theoretical expected pore sizes are 0.7 nm and 1.4-1.5 nm as shown in the structure (Scheme S1). From the experimental argon pore-size distribution, 0.75/1.43 nm pores for urea-CTF_400_5 and 1.65/2.70 nm pores for urea-CTF_500_5 were obtained. The values for urea-CTF_400_5 correspond well with the expected pore size, whereas, for urea-CTF-500, the absence of the smallest pore (0.7 nm) and the appearance of a larger pore (2.70 nm) were observed. This is the result of thermal decomposition causing the fragmentation of the walls on top of the micropores, creating mesopores in urea-CTF_500_5 [33]. A higher synthesis temperature also causes a higher degree of carbonization [34], which is seen in the C/N ratio from the elemental analysis data. The presence of a sudden drop in the adsorbed volume in the desorption isotherm at P/P 0~0 .45 ( Figure S2) is due to the tensile strength effect leading to a forced closure of the hysteresis loop [35]. The powder X-ray diffraction (PXRD) analysis show the amorphous characteristics of the materials with a broad diffraction band at 2θ = 25.8 degrees ( Figure S3). The physicochemical stability of the urea-CTFs was analyzed using thermogravimetric analysis (TGA) which showed that the materials were stable up to 450 • C ( Figure S4). In addition, the chemical stability of the urea-CTF_400_5 and urea-CTF_500_5 material was studied by exposing them to boiling water (3 days), 6 M NaOH (3 days), and 6 M HCl (3 days). After each treatment, they were cleaned to remove the corresponding chemical traces, and N 2 sorption was performed ( Figures S5 and S6). In all cases, microporosity was retained, proving the permanent microporosity of the urea-CTF. Transmission electron microscopy (TEM) images show the two-dimensional stacking of the urea-CTFs (Figures S7 and S8). In addition, scanning electron microscopy (SEM) images show that urea_CTF_400_5 particles, are on average, larger than the urea_CTF_500_5 particles ( Figures S7 and S8). Lower temperature synthesis of the CTF created fewer defects, and hence, urea_CTF_400_5 had longer sheet morphology. confirm the successful trimerization. Notably, a small broad peak around 1707 cm −1 was observed in the CTFs that are red-shifted from 1737 cm −1 of C(O) monomer and confirms the presence of urea groups in the resulting materials [31]. The observed lower wavenumber might be due to the decrease in the double-bond character of the C(O) bond of the urea functional group after the CTF formation. The porous properties of both the CTF materials were explored using argon sorption at 87 K ( Figure 1) and N2 sorption measurements at 77 K ( Figure S2). Both urea-CTF_400_5 and urea-CTF_500_5 displayed a Type I isotherm typical for microporous materials, and the calculated BET surface areas were 555 m 2 g −1 and 928 m 2 g −1 , respectively. The detailed textural properties are described in Table S1. As seen in several reported CTFs [11], the microporosity content depends on the synthesis temperature, whereas, urea-CTF_400_5 shows a higher microporous-to-mesoporous volume ratio in comparison to Urea-CTF_500_5. The theoretical expected pore sizes are 0.7 nm and 1.4-1.5 nm as shown in the structure (Scheme S1). From the experimental argon pore-size distribution, 0.75/1.43 nm pores for urea-CTF_400_5 and 1.65/2.70 nm pores for urea-CTF_500_5 were obtained. The values for urea-CTF_400_5 correspond well with the expected pore size, whereas, for urea-CTF-500, the absence of the smallest pore (0.7 nm) and the appearance of a larger pore (2.70 nm) were observed. This is the result of thermal decomposition causing the fragmentation of the walls on top of the micropores, creating mesopores in urea-CTF_500_5 [33]. A higher synthesis temperature also causes a higher degree of carbonization [34], which is seen in the C/N ratio from the elemental analysis data. The presence of a sudden drop in the adsorbed volume in the desorption isotherm at P/P0~0.45 ( Figure S2) is due to the tensile strength effect leading to a forced closure of the hysteresis loop [35]. The powder X-ray diffraction (PXRD) analysis show the amorphous characteristics of the materials with a broad diffraction band at 2θ = 25.8 degrees ( Figure S3). The physicochemical stability of the urea-CTFs was analyzed using thermogravimetric analysis (TGA) which showed that the materials were stable up to 450 °C ( Figure S4). In addition, the chemical stability of the urea-CTF_400_5 and urea-CTF_500_5 material was studied by exposing them to boiling water (3 days), 6 M NaOH (3 days), and 6 M HCl (3 days). After each treatment, they were cleaned to remove the corresponding chemical traces, and N2 sorption was performed ( Figures S5 and S6). In all cases, microporosity was retained, proving the permanent microporosity of the urea-CTF. Transmission electron microscopy (TEM) images show the two-dimensional stacking of the urea-CTFs (Figures S7 and S8). In addition, scanning electron microscopy (SEM) images show that urea_CTF_400_5 particles, are on average, larger than the urea_CTF_500_5 particles ( Figures S7 and S8). Lower temperature synthesis of the CTF created fewer defects, and hence, urea_CTF_400_5 had longer sheet morphology.

Gas Storage and Separation
Although CTFs in general have high potential for gas storage and separation, their potential for C2 hydrocarbon storage and separation has only rarely been explored. Only recently, CTF-PO71 [36] and hexene-CTF [37] have been studied for C2 hydrocarbon storage and separation. The permanent microporosity and presence of urea/triazine functional groups make urea-CTFs excellent candidates for this purpose. To this end, C2 hydrocarbon storage capacity was tested for both urea-CTF_400_5 and urea-CTF_500_5. Among these samples, urea-CTF_400_5 showed the highest C 2 H 2 uptake (3.86 mmol/g) at 273 K and 1 bar pressure, which is higher than the previously reported CTFs (Figure 2a). affinity at 273 K and 298 K of the C2 hydrocarbons for the urea-CTFs was calculated by the Clausius-Clapeyron equation (Figure S11 and S12). The isosteric heat of adsorption (Qst) values are given in Table S3. As expected, in both cases, a higher affinity was observed in urea-CTF_400_5 because the lower synthesis temperature resulted in fewer defects. In addition to the storage capacity, selectivity is perhaps an even more important parameter for industrial utilization. First, we targeted C2H2/CH4 and C2H4/CH4 separation. The CH4 uptake isotherms at 273 K and 298 K are given in Figure 2b. Selectivity was estimated using the ideal adsorbed solution theory (IAST) (Table S4 and Figures S14-S17). The calculated selectivities of the urea-CTFs were within 16.9-20.2 and 8.9-12.4 for C2H2/CH4 and C2H4/CH4, respectively, which are promising results for C2/C1 hydrocarbon separation (Figure 2d, Table 1).  Interestingly, despite the higher surface area of urea-CTF_500_5, a similar C 2 H 2 uptake (3.78 mmol/g) at 273 K and 1 bar pressure was observed (Figure 2a, Table S2). This is due to the abundance of the micropores in both materials. However, for urea-CTF_400_5, the V micro /V tot (0.72) is slightly higher than for urea-CTF_500_5 (0.61) (Table S1). This results in a steeper increase of C 2 H 2 uptake at the lower-pressure regime for urea-CTF_400_5. In addition, similar trends were observed in C 2 H 4 uptake (2.89 mmol/g and 2.92 mmol/g for urea-CTF_400_5 and urea-CTF_500_5, respectively) ( Figure S9). The affinity at 273 K and 298 K of the C2 hydrocarbons for the urea-CTFs was calculated by the Clausius-Clapeyron equation (Figures S11 and S12). The isosteric heat of adsorption (Q st ) values are given in Table S3. As expected, in both cases, a higher affinity was observed in urea-CTF_400_5 because the lower synthesis temperature resulted in fewer defects. In addition to the storage capacity, selectivity is perhaps an even more important parameter for industrial utilization. First, we targeted C 2 H 2 /CH 4 and C 2 H 4 /CH 4 separation. The CH 4 uptake isotherms at 273 K and 298 K are given in Figure 2b. Selectivity was estimated using the ideal adsorbed solution theory (IAST) (Table S4 and Figures S14-S17). The calculated selectivities of the urea-CTFs were within 16.9-20.2 and 8.9-12.4 for C 2 H 2 /CH 4 and C 2 H 4 /CH 4 , respectively, which are promising results for C2/C1 hydrocarbon separation (Figure 2d, Table 1).
The presence of inherent triazine and urea functionalities in urea-CTFs also encouraged us to test CO 2 adsorption performance. The CO 2 adsorption and desorption isotherms were measured at 273 K and 298 K up to 1 bar. At 1 bar and 273 K, urea-CTF_400_5 and urea-CTF_500_5 showed 2.8 mmol/g and 3.1 mmol/g uptake respectively, which are moderate values in comparison to other CTFs ( Figure S10, Table S2). The heat of liquefaction of bulk CO 2 is 17 kJ/mol [38], and urea-CTF_500_5 shows an isosteric heat of adsorption of 48.57 kJ/mol ( Figure S12 and Table S3), which is much higher. In addition, the Q st values of urea-CTFs are much higher than several reported CTFs and higher than activated carbon at low CO 2 pressure (17.8 kJ/mol). This confirms the strong dipolar interactions between CO 2 and the N-basic sites, as well as the H-bonding interactions between the urea functional group and the CO 2 molecules. The selectivity of CO 2 over N 2 and CH 4 are also important factors for CCS applications. CO 2 /CH 4 and CO 2 /N 2 selectivity were calculated using IAST (Table S4), and the best values, 20.3 and 69.6, were respectively obtained for urea-CTF_400_5 at 273 K. Notably, the obtained CO 2 /N 2 selectivity of urea-CTF_400_5 is higher than several reported CTFs [14,[39][40][41]. To verify the performance of the adsorbents in a mixed component system, breakthrough simulations were performed. Urea-CTF_400_5 was selected for these simulations as it showed the best performance among the urea-CTFs in all gas separations. The affinity constants and maximal loadings at the corresponding temperatures were obtained from Langmuir adsorption isotherm fitting (ESI). With these values, the equilibrium data for a mixed-component system were simulated. The equilibrium plots for C 2 H 2 /CH 4 , C 2 H 4 /CH 4 , and CO 2 /CH 4 components with (i) constant gas composition and variable pressure and (ii) constant pressure and variable gas composition are shown in Figures S18-S20. As expected, even in a 50:50 mixtures, uptake is higher for C 2 H 2 , C 2 H 4 , and CO 2 as compared to CH 4 due to the higher affinity constants. Further breakthrough simulations were performed with defined height, diameter of the column, gas-flow rate, and mass of the adsorbent at 25 • C and 1 bar pressure. The breakthrough plots for C 2 H 2 /CH 4 , C 2 H 4 /CH 4 , and CO 2 /CH 4 are shown in Figure 2c and Figures S21 and S22. The results show promising C2/C1 and CO 2 /CH 4 separation using urea-CTF-5-400.

Conclusions
In conclusion, rigid and highly stable CTFs were synthesized using flexible ureabased linkers. These materials exhibit high surface areas with good C 2 H 2 , C 2 H 4 , and CO 2 adsorption properties. The calculated C 2 H 2 /CH 4 , C 2 H 4 /CH 4 , and CO 2 /CH 4 selectivity values demonstrate that these materials are promising for C2/C1 hydrocarbon separation, as well as for the separation of CO 2 in natural gas extraction.