3D porous polymers for selective removal of CO2 and H2 storage: experimental and computational studies

In this article, newly designed 3D porous polymers with tuned porosity were synthesized by the polycondensation of tetrakis (4-aminophenyl) methane with pyrrole to form M1 polymer and with phenazine to form M2 polymer. The polymerization reaction used p-formaldehyde as a linker and nitric acid as a catalyst. The newly designed 3D porous polymers showed permanent porosity with a BET surface area of 575 m2/g for M1 and 389 m2/g for M2. The structure and thermal stability were investigated by solid 13C-NMR spectroscopy, Fourier-transform infrared (FT-IR) spectroscopy, and thermogravimetric analysis (TGA). The performance of the synthesized polymers toward CO2 and H2 was evaluated, demonstrating adsorption capacities of 1.85 mmol/g and 2.10 mmol/g for CO2 by M1 and M2, respectively. The importance of the synthesized polymers lies in their selectivity for CO2 capture, with CO2/N2 selectivity of 43 and 51 for M1 and M2, respectively. M1 and M2 polymers showed their capability for hydrogen storage with a capacity of 66 cm3/g (0.6 wt%) and 87 cm3/g (0.8 wt%), respectively, at 1 bar and 77 K. Molecular dynamics (MD) simulations using the grand canonical Monte Carlo (GCMC) method revealed the presence of considerable microporosity on M2, making it highly selective to CO2. The exceptional removal capabilities, combined with the high thermal stability and microporosity, enable M2 to be a potential material for flue gas purification and hydrogen storage.


Introduction
Global warming caused by the elevated levels of CO 2 has garnered significant attention in recent years.The elevated levels of CO 2 have become a serious problem due to their hazardous effects on the environment; these effects encompass a gradual increase in the temperature of the earth, resulting in droughts, fluctuations in the weather, and elevated oceanic water levels and ocean acidification (Feldman et al., 2015;Leal et al., 2018;Zakeri et al., 2022).During the last 40 years, the concentration of CO 2 has increased tremendously from 319 ppm to 414 ppm in 2021, setting a new record, and is estimated to increase to 800 ppm within the next 100 years if we continue relying on fossil fuels as a primary energy source (Mercer, 1978;Feldman et al., 2015;Abdelhakim et al., 2022).Fossil fuels, serving as energy sources, are typically divided into three types: natural gas, coal, and petroleum.Upon combustion, they release CO 2 , SOx, and NOx gases; mercury; and various particulates that cause pollution in the environment and have a large impact on human health (Khraisheh et al., 2020;Perera and Nadeau, 2022).Due to the major concern for the environment and human health, several methods and techniques have been identified to reduce the effect of CO 2 (Taylor et al., 2020;Long et al., 2021).These techniques and methods include finding new sources of energy, such as hydrogen gas, as an alternative energy source, reducing energy consumption by increasing energy efficiency, and finding new methods for capturing CO 2 (Kar et al., 2022;Paramati et al., 2022).Capturing CO 2 is one of these methods and has garnered considerable attention over the years.Two major methods have been used: chemisorption of CO 2 , which involves the formation of a chemical bond between CO 2 and the adsorbent.Such an example for chemisorption is the absorption of CO 2 by liquid amines which is the most commonly used method by refineries to capture CO 2 from natural gas streams, their operation is non-costeffective and requires high energy for regeneration.Furthermore, degradation of the liquid amines thermally and oxidatively causes corrosion in refinery setups (Bobek et al., 2016;Dey et al., 2017;Kong et al., 2019).The other major technique that is emerging is capturing CO 2 by physisorption.Physisorption is a process where CO 2 is bonded weakly with the adsorbent by weak van der Waals forces of attraction, which allows the sorbent to be capable of reversibly adsorbing CO 2 from flue gas streams by solid sorbents (Plaza et al., 2007;Oschatz and Antonietti, 2018;Kong et al., 2019).Solid sorbents have been developed through the years, and several key features should be included in the design of these sorbents for efficient CO 2 capture, such as i) high sorption capacity, ii) selectivity, and iii) adequate stability in the presence of contaminants (Zou et al., 2017;Abdelnaby et al., 2019;Khraisheh et al., 2020).Different classes of solid sorbents have emerged as promising materials for CO 2 reduction, such as metal-organic frameworks (MOFs), covalent organic frameworks (COFs), zeolites, carbonaceous materials, such as activated carbon, and porous organic polymers (POPs) (Cheung and Hedin, 2014;Gadipelli and Guo, 2015;Lohse and Bein, 2018;Zhao et al., 2018;Qasem et al., 2020).POPs are an interesting class of materials that possess excellent features such as low density, high surface area with a tunable pore size distribution, good thermal and chemical stability, and synthetic versatility (Zou et al., 2017;Gao et al., 2019;Gu et al., 2022).These features are considered requirements for the selective removal of CO 2 from flue gas and natural gas streams (Rufford et al., 2012;Ahmed et al., 2015;Alloush et al., 2022).In our endeavor to design and synthesize porous organic polymers for CO 2 capture and hydrogen storage, we demonstrate the design and synthesis of new 3D porous organic polymers with tuned porosity in this study.The synthesized 3D polymers were evaluated for their CO 2 and H 2 adsorption capabilities and for their selectivity of CO 2 over N 2 and CH 4 to assess their potential use in flue gas and natural gas treatment.
Methanol (MeOH, ≥99.9% purity) was acquired from Merck Millipore ™ .Except for pyrrole, which was distilled at 150 °C immediately before use, all chemicals were used as received.
Ultrahigh-purity-grade nitrogen (N 2 , 99.999%), helium (He, 99.999%), and high-purity carbon dioxide (CO 2 , 99.9%) gases were supplied by Abdullah Hashem Industrial Co., Saudi Arabia.Natural abundance solid-state 13 C-NMR spectra were collected using a Bruker 400 MHz spectrometer set to 125.65 MHz at room temperature (11.74 T).Samples were packed into 4 mm zirconium oxide rotors.Cross-polarization and high-power decoupling were used.The pulse delay was 2.5 s, and the magic angle spinning rate was 10 kHz.A PerkinElmer FT-IR spectrometer was used to obtain FT-IR spectra.FT-IR spectra were obtained in the range of 4,000-400 cm −1 using a PerkinElmer 16F PC FT-IR spectrometer and solid potassium bromide (KBr) pellets (mid-IR region).TGA was performed using the STA 429 ® (NETZSCH group, Germany) thermal analyzer.All gas uptake measurements were performed on the Quantachrome ® Autosorb IQ instrument, and isotherms were obtained at 273 K and 298 K.

Synthesis
In a typical experiment (Abdelnaby et al., 2018), tetrakis (4aminophenyl) methane (2.73 × 10 −3 mol, 1.0 g) and pyrrole (0.0109 mol, 0.73 g) were stirred in a 50-mL round-bottomed flask equipped with a magnetic bar containing 25 mL DMF until a homogeneous solution was obtained.p-Formaldehyde (0.02187 mol, 0.66 g) and nitric acid (10% of p-formaldehyde; 0.002187 mol, 0.199 g) were then added to the reaction mixture.The reaction mixture was flushed with N 2 gas and sealed and stirred for 24 h at 90 °C.Once the reaction was completed, the product was filtered and washed with methanol for 3 days with continuous exchange of methanol to ensure the removal of any monomers or unreacted materials left in the reaction.The product was vacuumdried at 90 °C for 24 h to get M1 as a fine black powder (yield % = 65%).M2 was obtained as a bright yellow powder (yield % = 49%) under similar reaction conditions with tetrakis (4-aminophenyl) methane, phenazine (M2), and p-formaldehyde taken at a molar ratio of 1:4:8 and 10 mol% of nitric acid relative to p-formaldehyde.The yield of the polymerization reaction was calculated as the mass of the product relative to the mass of all reactants.

MD simulation procedure
Molecular dynamics simulations (Supplementary Material) were performed to reveal the underlying mechanism of adsorption of CO 2 , CH 4 , and N 2 gases by the polymers M1 and M2.The structural geometries of the polymers were built and optimized using the smart algorithm in the Forcite module of Materials Studio 8.0 software.The COMPASS II force field (Sun et al., 2016) was adopted, while the self-consistent field (SCF) convergence threshold, maximum force tolerance, and energy tolerance were set to 1.0 × 10 −5 Ha, 0.001 Ha/ A, and 1.0 × 10 −5 Ha, respectively.Thereafter, using the "Locate" task bar on the Sorption module, the suitable adsorption sites of the gases on M1 and M2 were identified, and the adsorption capacities were estimated based on the principle of simulated annealing using the GCMC method (Aljamaan et al., 2017;Song et al., 2018).The adsorption isotherms at 273.15, 298.15, and 313.15K were calculated using the Langmuir fitting equation: where a is the limit of adsorption capacity in mmol/g and b is the adsorption constant in MPa −1 .The estimated adsorption capacities SCHEME 1 Synthesis scheme for 3D porous polymers.
were given in the units of average molecules/cell and were converted to the amount of gas adsorbed in mmol/g using the following equation (Zhang et al., 2021): where Mw cell is the relative molecular mass of M1 and M2 polymers in the constructed supercell.
3 Results and discussion

Synthesis and characterization
This paper describes two new 3D porous amine-based polymers.The polymerization method was based on a modified Mannich polycondensation reaction, with tetrakis (4-aminophenyl) methane added as a common component in the polymers.The polymers were realized by polymerizing tetrakis (4-aminophenyl) methane with pyrrole to obtain M1 and phenazine to form M2. The polymerization reaction was conducted using DMF as a solvent and concentrated HNO 3 as a catalyst (Scheme 1).
The structural features of the polymers were characterized by solid 13 C-NMR, as shown in Figure 1A.The peaks residing from 100 ppm to 150 ppm correspond to the aromatic carbons of pyrrole, phenazine, and tetrakis (4-aminophenyl) methane.A peak at 65 ppm corresponds to the quaternary carbon of tetrakis (4aminophenyl) methane, linking the four aniline moieties.A peak at 55 ppm corresponds to the methylene linkage (-CH 2 -) between tetrakis (4-aminophenyl) methane and pyrrole or phenazine.A peak at ~30 ppm corresponds to the methylene linkage (-CH 2 -) present between pyrrole and pyrrole moieties (Luo et al., 2012; Abdelnaby  , 2018). Figure 1B represents the FT-IR spectra for the synthesized polymers.The figure shows a broadband in the region from 3,300 to 3,500 cm −1 resulting from the overlap between both the 1 °amine (-NH 2 ) stretching vibrations of the aniline moiety in the tetrakis (4-aminophenyl) methane monomer and the 2 °amine (-NH-) stretching band of the pyrrole moiety.The bands between 1,400 and 1,700 cm −1 correspond to the aromatic -C=C-and -C=N-stretching vibrations of phenazine, pyrrole, and aniline moieties.A band at 1,631 cm −1 attributed to -NH 2 scissoring can be observed overlapping with the -C=C-aromatic vibrational bands that appear in the same region.Both -NH 2 and -NH-wagging bands are at 694 and 755 cm −1 , respectively (Wu et al., 2002;Tian et al., 2009).Figure 1C shows the powder X-ray diffraction patterns of the 3D porous polymers.The powder X-ray diffraction patterns reveal the amorphous nature of the synthesized polymers with a broad signal at ~15 °2Ө with some degree of crystallinity shown by the signal at ~22 °2Ө present in M1 and M2 (Wei et al., 1992;Errahali et al., 2014).Figure 1D reveals the good thermal stability of the synthesized polymers, which could be related to the stiff cross-linked structures of M1 and M2.The thermograms in Figure 4D show a 5% weight loss of small, trapped molecules in M2 up to 200 °C, followed by a second degradation at ~500 °C, where the degradation of the polymer structure occurs by the loss of the methylene linkages, followed by the degradation of the polymer backbone.On the other hand, M1 begins to thermally degrade at ~300 °C up to 440 °C, which may be attributed to the loss of the methylene linkages between the moieties, followed by the complete degradation of the polymer structure at ~600 °C (Li et al., 2014;Yan et al., 2016).

Porosity of the 3D porous polymers
As shown in Figure 2A, the nitrogen adsorption/desorption isotherms suggest that M1 and M2 are porous in nature.M1 polymer shows permanent porosity with a BET surface area of 575 m 2 /g.The BET isotherm of M1 suggests that the polymer exhibits Type I characteristics with a steep increase in nitrogen uptake at low relative pressure (P/P 0 < 0.05).The hysteresis in the M1 isotherm suggests the high interaction between M1 porous polymer and N 2 molecules, which could be due to the entrapment of N 2 molecules in the pores of M1 that leads to the hysteresis found in the adsorption/desorption isotherm (Li et al., 2022).The pore size distribution analysis based on density functional theory (DFT) calculations reveals two distinct regions in the M2 porous polymer.As shown in Figure 2B, there is a prominent peak at approximately 10 Å, indicating the presence of micropores, and another strong peak at an average pore width of approximately 33 Å, representing the mesoporous region.M2 polymer shows a permanent porosity with a BET surface area of 389 m 2 /g.The nitrogen adsorption isotherm of M2 suggests that the polymer is microporous in nature and exhibits Type I characteristics.Further examination using DFT calculations reveals that the apertures of M2 polymer are mainly in the range of micropores with pore widths less than 20 Å, as shown in Figure 2B.
The tuned pore size distribution, accompanied by the good surface areas, encouraged us to investigate the capabilities of M1 and M2 for CO 2 adsorption compared with CH 4 and N 2 for applications in natural gas purification and flue gas treatment.For the polymers to perform well, they should be selective toward CO 2 and that can be produced by enhancing the microporosity of the polymer.As shown in Figure 2B, M2 is microporous in nature, with pore size distributions falling in the microporous region less than 20 Å with a high intensity close to the kinetic diameter of CO 2 (3.3 Å).This is shown by the adsorption capacities observed in Figures 3A-D, where the adsorption capacity at 273 K of CO 2 is higher in M2 (2.1 mmol/g) compared to M1 (1.85 mmol/g).At 298 K, the adsorption capacities behave in a similar manner, where the adsorption capacity of M2 for CO 2 is 1.41 mmol/g, for CH 4 is 0.44 mmol/g, and for N 2 is 0.050 mmol/g, whereas the adsorption capacity of M1 for CO 2 is 1.24 mmol/g, for CH 4 is 0.32 mmol/g, and for N 2 is 0.08 mmol/g.Comparing the efficiency between M1 and M2, it is shown that the adsorption capacity of M2 was higher than M1, which is attributed to the microporous nature of the pores and the absence of mesopores in M2 (Song et al., 2022).The isosteric heat of adsorption (Q st ) of CO 2 shows the interaction energy between a sorbent and CO 2 gas.Figures 3E, F show the Q st vs uptake of CO 2 .The values of Q st decrease with the coverage of the surface of the polymer with CO 2 , indicating that the adsorption process occurred on a heterogeneous surface.The Q st values for the adsorption of CO 2 by M1 and M2 were found to be 33.1 kJ/mol and 33.6 kJ/mol, respectively.This indicates that the adsorption process is of physisorption in nature (Khosrowshahi et al., 2022;Ravi et al., 2023).Another feature that an adsorbent should possess is high selectivity.As shown in Figures 4A-D, the selectivity was investigated at 298 K to mimic ambient conditions, which is in agreement with post-combustion treatment conditions.By using the initial slope ratios of Henry's law constants at 298 K, the selectivity of M1 for CO 2 /N 2 is 43 and CO 2 /CH 4 is 9, whereas the selectivity of M2 for CO 2 /N 2 is 51 and CO 2 /CH 4 is 10.The selectivity of M2 was higher than that of M1 even though it has a lower surface area, which could be explained by the microporous nature of the polymer with a similar observation for CO 2 /CH 4 selectivity.As shown in Table 1, despite having lower surface areas, M1 and M2 exhibit superior adsorption capacity and selectivity for CO 2 /CH 4 and CO 2 /N 2 compared to reported porous materials with higher surface areas.
The molecular dynamics of the synthesized polymers were studied to correlate the experimental results with the theoretical   Frontiers in Chemistry frontiersin.org07 calculations (Supplementary Material).The adsorption of singlecomponent gases, CO 2 , CH 4 , and N 2 , on M1 and M2 polymers at 298.15 K was simulated in supercells of dimension 30 × 30 × 40 A, comprising of 20 repeating units of the polymer molecules, as presented in Figures 5A-C.The corresponding simulated adsorption isotherms are shown in Figures 5D, E. Both polymers demonstrated strong van der Waals attraction toward CO 2 molecules via the pyrrolic and pyridinic nitrogen atoms on M1 and M2, respectively.Moreover, CO 2 adsorption binding sites were located on both molecules, with fewer sites for CH 4 and N 2 gases.However, M2 demonstrated rapid uptake of CO 2 below 2 MPa (20 bar), indicating the presence of microporosity within the  polymer framework (Rizzuto et al., 2017), and the selectivity of the polymer to CO 2 gas is consistent with the experimental findings.
Using the Langmuir isotherm model (Table 2), the limit adsorption capacities of CO 2 at 298.15 K on M1 and M2 were estimated as 2.99 and 3.74 mmol/g, while for CH 4 and N 2 , the values were 0.44 and 0.17, and 0.98 and 0.20 mmol/g, respectively.Meanwhile, the corresponding theoretical isosteric heat of adsorption (Q st ) for CO 2 at 298.15 K was calculated as 43.1 and 43.9 kJ/mol on M1 and M2, respectively.While the theoretical values are slightly higher than the experimental values, which could be ascribed to the overestimation from the general assumptions input into the simulation software (Meng et al., 2018).The order of selectivity of the polymers is in good agreement with the experimental findings and revealed the preferential selectivity of M2 to CO 2 gas.The adsorption of CO 2 on M1 and M2 at temperatures of 273.15, 298.15, and 313.15K was further investigated, and the results are presented in Figures 6A, B. A slight decrease in the adsorption capacity of both polymers was observed with increasing temperature.This suggests that the adsorption of the gas molecules is strictly dependent on the van der Waals force of attraction between them and the active sites on the polymers, which tend to weaken with the increase in temperature due to the increase in the inherent kinetic energy of the gas molecules.Thus, M2 experiences less decline in the adsorption capacity, suggesting its greater adsorption preference for CO 2 gas.Meanwhile, the selectivity of the polymers toward CO 2 in multi-component gas streams, comprising CO 2 and CH 4 at a molar ratio of 50:50 and CO 2 and N 2 gases at a molar ratio of 20: 80, was further explored theoretically, as shown in Figures 6C-F.The presence of equimolar volumes of CH 4 and the abundance of N 2 gases did not impede the selectivity of the polymers to CO 2 gas, as the presence of quadrupole C=O bonds favors the van der Waals attraction to the polymers.On the other hand, CH 4 and N 2 molecules experienced a drastic decrease in adsorption affinity by the polymers due to the strong competition by CO 2 molecules, resulting in fewer molecular interactions, as shown in Figures 6C-F.In all cases, the M2 polymer rapidly adsorbs CO 2 molecules in the presence of competing gas molecules, affirming its selectivity and aligning well with the experimental findings.
In our endeavor to tap into the world of clean energy and seek alternatives to overcome the pollution of petroleum products, we assessed our polymers for their capability to store hydrogen gas.The results in Figure 7A revealed the adsorption capacity of M1 and M2 toward H 2 at 77 K and 1 atm to be 66 cm 3 /g (0.6 wt%) and 87 cm 3 /g (0.8 wt%), respectively.Interestingly, the absence of mesopores in M2 reflected the higher adsorption capacity toward H 2 compared to M1, which is consistent with Rong et al. (2021).The storage capacity is higher and comparable to that of porous polymers with similar or higher surface areas, as shown in Table 3.Finally, we simulate the potential of hydrogen storage on the polymer molecules by constructing supercells and conducting the grand canonical Monte Carlo simulation, as shown in Figures 7B, C. The theoretical H 2 uptake capacities of M1 and M2 at 77 K were calculated to be 61.7 and 83.2 cm 3 /g, respectively.These values are in good agreement with the experimental adsorption capacities of both polymers and corroborate the microporosity of M2, making it a potential material for H 2 storage.

Conclusion
In this study, we report the synthesis of 3D porous polymers with tuned porosity.The choice of reaction conditions and monomers leads to polymers with microporous and meso-/ microporous structures.The produced polymers were found to be thermally stable up to temperatures of 400 °C.Analysis revealed the porous nature of polymers with a BET surface area of 575 m 2 /g for M1 and 389 m 2 /g for M2.The M1 polymer showed defined micropores of 7 Å and mesopores of 33 Å, whereas M2 exhibited micropores with a pore size distribution of <20 Å.The study also revealed the effect of microporosity on adsorption ability and selectivity.The results revealed superior performance of M2 in the absence of mesoporosity.The adsorption capacities at 273 K of CO 2 are higher in M2 (2.1 mmol/g) compared to M1 (1.85 mmol/g).In addition, at 298 K, the adsorption capacity of M2 for CO 2 was 1.41 mmol/g, for CH 4 was 0.44 mmol/g, and for N 2 was 0.050 mmol/g, while the adsorption capacity of M1 for CO 2 was 1.24 mmol/g, for CH 4 was 0.32 mmol/g, and for N 2 was 0.08 mmol/g.Furthermore, the absence of mesoporosity in M2 was evidenced by its superior performance in hydrogen storage.The molecular dynamics simulation confirmed the superior performance of M2 and coincided with the experimental values to prove the efficiency and capability of porous polymers to be a potential adsorbent for selective removal of CO 2 and H 2 storage.
FIGURE 1 (A) Solid 13 C-NMR CP/MAS spectra for 3D porous polymers.(B) FT-IR spectra for 3D porous polymers and identification of peaks.(C) Thermogravimetric analysis of 3D porous polymers.(D) Powder X-ray diffraction of 3D porous polymers.
FIGURE 4 (A) Adsorption of CO 2 , CH 4 , and N 2 by M1 at 298 K; (B) adsorption of CO 2 , CH 4 , and N 2 by M2 at 298 K; (C) initial slope fitting of M1 at 298 K; and (D) initial slope fitting of M2 at 298 K.
FIGURE 5 (A) CO 2 , (B) CH 4 , and (C) N 2 adsorption sites located on the M2 polymer packed in amorphous cells of dimension 30 × 30 × 40 A, comprising 20 repeating units.The red spheres represent the adsorbed gas molecules.The corresponding simulated adsorption isotherms for both polymers at 298.15 K are presented in (D) M1 and (E) M2.

FIGURE 6
FIGURE 6 Theoretical adsorption isotherms of CO 2 at 273.15, 298.15, and 313.15K on (A) M1 and (B) M2.Adsorption isotherms of CO 2 /CH 4 in multicomponent streams with a molar ratio of 50:50 on (C) M1 and (D) M2 at 298.15 K; and CO 2 /N 2 with a molar ratio of 20:80 on (E) M1 and (F) M2 are also presented.
FIGURE 7 (A) H 2 adsorption/desorption isotherm of M1 and M2 at 77 K (filled circles refer to adsorption, and unfilled circles refer to desorption).(B) Supercell of dimension 30 × 30 × 40 A, comprising 20 repeating units of M2 for the simulation of H 2 storage capacity and (C) theoretical adsorption isotherms of H 2 at 77 K on M1 and M2.

TABLE 1
Comparison of porous materials with M1 and M2 with respect to surface area, CO 2 uptake at 273 K and 298 K, CO 2 /N 2 and CO 2 /CH 4 selectivity, and Q st (CO 2 ).

TABLE 2
Langmuir fitting parameters for the adsorption of CO 2 , CH 4 , and N 2 on M1 and M2.

TABLE 3
Comparison of porous materials to M1 and M2 for hydrogen storage.
a Data based on 77 K and 1 bar.