Local Deprotonation Enables Cation Exchange ... - ACS Publications

Article pubs.acs.org/crystal

Local Deprotonation Enables Cation Exchange, Porosity Modulation, and Tunable Adsorption Selectivity in a Metal−Organic Framework Published as part of a Crystal Growth and Design virtual special issue on Crystal Engineering of Nanoporous Materials for Gas Storage and Separation Jun-Hao Wang,†,§ Dong Luo,† Mian Li,† and Dan Li*,†,‡ †

Department of Chemistry and Key Laboratory for Preparation and Application of Ordered Structural Materials of Guangdong Province, Shantou University, Guangdong 515063, P. R. China ‡ College of Chemistry and Materials Science, Jinan University, Guangzhou 510632, P. R. China § Institute of Crystalline Materials, Shanxi University, Taiyuan 030006, P. R. China S Supporting Information *

ABSTRACT: This account demonstrates that under regulated synthetic conditions the protonated carboxyl sites in a neutral metal−organic framework (MOF), known as MOF-324 [formulated as Zn3OH(PzC)2(HPzC), H2PzC = 4-pyrazolecarboxylic acid], can undergo complete deprotonation, yielding NH4@ZnPzC [formulated as NH4·Zn3OH(PzC)3]. This modified, anionic framework with a pcu-g net is thus capable of postsynthetic cation exchange, which is highly modular, encompassing organic ammonium (Me3NH+, Et3NH+), main-group metal ions (Li+, Mg2+), and even lanthanide ions (Eu3+, Tb3+). The present approach is shown to be versatile and efficient in regulating porosity, fine-tuning gas adsorption properties, and endowing other functionalities such as liquid-phase adsorptive separation (benzene/cyclohexane). In particular, the selective adsorption behaviors of CO2 over N2 in this system have been studied in detail, targeting optimized CO2/N2 selectivity, which are evaluated through uptake capacity, isosteric heat, and ideal adsorbed solution theory. Besides the competent CO2/N2 selectivity (e.g., 50.8 for Li@ZnPzC at initial CO2/N2 = 15:85), the current approach, when compared with the popular strategy of amine-grafting on exposed metal sites, shows the advantages of (i) modular porosity and adsorption property, (ii) facile exchange processes which are quantifiable and would not severely block the pore windows, and (iii) mild isosteric heat which is required for subsequent CO2 release. It is worthy to note the regulation of local deprotonation can be regarded as a general approach for MOF chemistry and functionalization, given that there are plenty of reported MOFs bearing such “open protonated sites”.

■

The resulting adsorbents, mmen-Mg2(dobpdc)28,29 and mmenCuBTTri30 (mmen = N,N′-dimethylethylenediamine), are highly desirable, exhibiting high affinity with CO2 and exceptional CO2/N2 selectivity. However, this strategy still has some limitations. For example, the diffusion of amines into the interior pores is dependent on the binding strength of the amine to the metal site; those with a strong affinity might block some of the pore windows and thus severely reduce the adsorption ability.28 Therefore, the amines in use should be carefully selected, which means this grafting method is not modular in practice. An alternative approach for the regulation of adsorption affinity and selectivity of MOFs is through postsynthetic ion exchange.31−44 Compared with the above strategy of amine-

INTRODUCTION Metal−organic frameworks (MOFs),1 also known as porous coordination polymers (PCPs),2 are a promising class of hybrid crystalline porous materials3 with defined topologies4,5 and uniform pores, which exhibit widespread applications as gas reservoirs,6−8 catalysts,9−11 separators,12−14 and sensors.15−17 Taking gas adsorption properties, which are of utmost current interest due to energy and environmental concerns, as an example, MOF researchers have developed a variety of pre- and postsynthetic methods,18−20 aimed at enhancing the adsorptive ability and improving the separation performance with regard to carbon dioxide capture and storage.6,21−25 In terms of enhancing CO2/N2 adsorption selectivity, an intensely studied strategy was the use of open (or exposed) metal sites,6 notably demonstrated by Mg-MOF-7426 and CuBTTri.27 Such a strategy can drastically enhance the adsorptive preference of CO2 over N2, an advantage that can be further strengthened via postsynthetic alkylamine grafting. © 2017 American Chemical Society

Received: March 9, 2017 Revised: April 13, 2017 Published: May 2, 2017 3387

DOI: 10.1021/acs.cgd.7b00346 Cryst. Growth Des. 2017, 17, 3387−3394

Crystal Growth & Design

Article

Figure 1. (a) Synthesis of NH4@ZnPzC and schematic postsynthetic cation exchange process. (b) Comparison of PXRD patterns of NH4@ZnPzC (bulk samples) and MOF-324 (simulated from single crystal data). Inset: IR spectrum of NH4@ZnPzC. (c) Comparison of PXRD patterns of NH4@ZnPzC (simulated from single crystal data) and cation-exchanged samples.

grafting on open metal sites, there are several advantages for the cation exchange approach for modifying anionic MOF adsorbents. (i) Unlike the grafting methods which rely on specific interaction between the metal sites and amines, the exchanged guests here can be considered modular: all charged species, given a suitable size, with a different chemical nature can be implemented. Therefore, the pore volume/window size and adsorption behaviors can be systematically tuned. (ii) Cation exchange which relies on nondirectional electrostatic interaction between the overall network and charged guests will facilitate the guest diffusion into the interior pores, and this process can be easily quantified. The prerequisite for utilizing the ion exchange approach is the MOF in question must have a charged framework.45,46 However, ionic MOFs are much rarer than neutral ones, and many existing ionic MOFs cannot maintain framework integrity during the ion exchange processes. An interesting example is two series of isoreticular (i.e., with same underlying topology) MOFs showing different charge states, namely, neutral [Ni3OH(L1)3(L2)1.5]·xguest (MCF-19 series)47,48 and cationic [In3OH(L1)3(L2)1.5](NO3) (ITC-n series).49 The present work reports the synthesis of an anionic ZnII MOF by introducing 1H-pyrazole-4-carboxylic acid (H2PzC) as the ligand under alkaline conditions, yielding NH4·Zn3OH(PzC)3

(denoted as NH4@ZnPzC). We noticed an isostructural MOF, known as MOF-324 [formulated as Zn3OH(PzC)2(HPzC)],50 was obtained through a different synthetic procedure and having a neutral framework host. The complete deprotonation in the local building units has warranted cation exchange for NH4@ZnPzC, with the advantages of tunable adsorption behaviors and adjustable CO2/N2 selectivity. The cationic entities selected are of a different chemical nature: organic ammonium (Me3NH+, Et3NH+), main-group metal ions (Li+, Mg2+), and trivalent lanthanide ions (Eu3+, Tb3+). The evaluations of CO2 and N2 adsorption behaviors include uptake capacity, isosteric heat, and ideal adsorbed solution theory (IAST) selectivity. To demonstrate the versatility of such a deprotonation-induced cation exchange approach, we also examine the performances of the cationexchanged products toward liquid-phase adsorptive separation (benzene/cyclohexane).14 Because there are a number of neutral MOFs in the literature bearing pendent, protonated carboxyl sites, one can propose the regulation of local deprotonation to be a general strategy for transforming neutral MOFs to anionic ones. 3388



Article

Figure 2. (a) Distorted cubic cage (pore diameter ca. 1.0 nm) viewed from of the diagonal direction of the unit cell of NH4@ZnPzC. (b) Top and (c) side views of the Zn3O SBU in NH4@ZnPzC, showing the multiple contacts between NH4+ and the SBU. Color codes: Zn, green polyhedra; O, red; N, blue; C, black, H, omitted.

■

(yield 90%). Elemental analysis (%): Found: C 24.6, H 2.21 and N 16.85. Calcd: C 25.67, H 1.97 and N 17.46 for C12N7O7H11Zn3. IR (cm−1): 3375 (br), 3122 (w), 2970 (w), 2790 (w), 2476 (w), 1550 (s), 1450 (s), 1332 (m), 1287 (s), 1193 (w), 1041 (s), 1005 (s), 886 (m), 790 (s), 617 (m). Single-Crystal X-ray Diffraction Study. Data collections of NH4@ZnPzC were performed on an Oxford Diffraction Gemini E (Enhance Cu X-ray source, Kα, λ = 1.5418 Å) equipped with a graphite monochromator and ATLAS CCD detector (CrysAlis CCD, Oxford Diffraction Ltd.) at room temperature. The data were processed using CrysAlis RED, Oxford Diffraction Ltd. (Version 1.171.34.44, release 25-10-2010 CrysAlis171.NET). The structure were solved by direct methods (SHELXTL-97) and refined on F2 using full-matrix last-squares (SHELXTL-97). All non-hydrogen atoms were refined with anisotropic thermal parameters, and all hydrogen atoms were refined with isotropic thermal parameters riding on those of the parent atoms. The PLATON SQUEEZES routine was applied to modify the reflection data because of the amount of disordered solvent present in the pores of the MOF. Parameters for data collection and refinement are summarized in Table S1 (see the Supporting Information). Selected bond lengths and angles for all complexes are given in Table S2 (see the Supporting Information). Cation Exchange Experiments. The extra-framework ammonium cations included in NH4@ZnPzC can be postsynthetically exchanged with Li+, Mg2+, La3+ (La = Eu3+, Tb3+), Me3NH+, and Et3NH+ cations, affording Li@ZnPzC, Mg@ZnPzC, La@ZnPzC, Me3NH@ZnPzC, and Et3NH@ZnPzC, respectively. In a typical experiment, the sample of NH4@ZnPzC was suspended in methanolic solution of metal nitrate salts (e.g., LiNO3, Mg(NO3)2, and La(NO3)3) or organic amines (e.g., Me3N and Et3N, concentration 0.1−0.5 M) for 8 days, with corresponding methanol solution refreshed every 4 days. All the exchanged solids were subsequently filtered, washed with

EXPERIMENTAL SECTION

General Information. The ligand H2PzC (1H-pyrazole-4carboxylic acid), other reagents and solvents were commercially available and used as received. Powder X-ray diffraction (PXRD) experiments were performed on a D8 Advance X-ray diffractometer using CuKα radiation (1.5418 Å). Infrared spectra were recorded on a Nicolet Avatar 360 FTIR spectrometer in the range of 3800−400 cm−1 on KBr pellets. Elemental analysis (C, H, and N) was performed using a Vario EL III CHNS elemental analyzer. Thermogravimetric (TG) analysis was performed on a TA Instruments Q50 thermogravimetric analyzer under nitrogen flow of (40 mL·min−1) at a typical heating rate of 10 °C min−1. 1H NMR spectra were measured on a Bruker Spectrospin 300 spectrometer. For the determination of exchanged metal contents, the desolvated exchanged samples were digested in concentrated hydrochloric acid and measured through inductively coupled plasma-atomic emission spectrometry (ICP-AES) on a PerkinElmer Optima-4300 DV instrument. Synthesis. Single crystals of NH4·Zn3OH(PzC)3 (hereafter referred to as NH4@ZnPzC) were grown from a triple-layered solution consisting of a methanol solution (2 mL) of H2PzC (0.1 mmol) as the top layer, DMF (2 mL) as the middle layer, and Zn(NO3)2 (0.1 mmol) in 25% ammonia solution (2 mL) as the bottom layer. After three months, the crystals were grown and isolated by filtration in 60% yield. The bulk samples of NH4@ZnPzC were obtained by adding the ammonia solution of Zn(NO3)2 and then the methanol solution of H2PzC into DMF (H2O/CH3OH/DMF = 1:1:4, v/v/v), and reacting under refluxing for about 6 h. Refluxing time is crucial; longer than 10 h would result in impurities. White powders of NH4@ZnPzC were obtained and collected after filtrating, washing with DMF and CH3OH three times, respectively. The as-synthesized samples were activated with methanol (3 × 10 mL) over a three-day period before being dried 3389



Article

methanol three times, and later suspended in methanol for 24 h. After the exchanged samples were dried under a vacuum at 85 °C for 48 h, the degree of cation exchange was determined by ICP-AES for metal ions and 1H NMR analysis for organic amines. The data indicated that 100% of the ammonium cations were exchanged with Li+, Mg2+, La3+ (La = Eu3+, Tb3+), 85% for Et3NH+ and about 30% for Me3NH+ (see the Supporting Information for details). The unaltered PXRD patterns indicated the maintenance of the parent structure (see Figure 1c). Low-Pressure Gas Adsorption Measurements and Evaluations. The gas adsorption−desorption experiments were performed on a Micromeritics ASAP 2020 M surface area and pore size analyzer. All gases used were of 99.999% purity. Prior to the measurements, the samples were activated at 85 °C under a vacuum according to the TG analysis (Figure S1 in the Supporting Information) and then treated by using the “outgas” function of the surface area analyzer for 6 h at 120 °C. Adsorption isotherms for N2 were monitored at 77 and 273 K, respectively. CO2 adsorption isotherms were measured at 273, 298, 303, and 308 K, respectively. Surface area and pore size distribution were determined from the N2 gas isotherms at 77 K. Multipoint Brunauer−Emmett−Teller (BET) and the Langmuir surface areas were estimated by using the data recorded at P/P0 = 0.0001−0.1 atm. The pore size distribution was calculated from the Saito−Foley method51,52 using data recorded at P/P0 < 0.6 atm. All the adsorption data used for calculation of selectivity and the isosteric heat (Qst) were fitted using dual-site Langmuir−Freundlich equations.6,21,22 The Clausius−Clapeyron equation was employed to calculate the enthalpies of CO2 adsorption.6,21,22 The selectivity of CO2 over N2 was evaluated using ideal adsorbed solution theory (IAST, see the Supporting Information).6,22,53 Detailed adsorption data and analysis methods are given in the Supporting Information.

MOF host. The exchanged samples can retain their crystallinity after thermal activation at 85 °C under a vacuum for 48 h (Figure S3 in the Supporting Information). However, they are sensitive to boiling water, as revealed by the new diffraction peaks in PXRD patterns (Figure S4 in the Supporting Information). Topological Analysis. We note that MOF-324 was described to have the same topology as MOF-554 with a primitive cubic (pcu) net,55 since each SBU is connected to six neighboring SBUs through PzC/HPzC linkers.50 However, the two structures differ at least in two notable aspects: (i) the composition and geometric shapes of the SBUs are distinct (octahedral Zn4O SBU for MOF-5, chairlike Zn3O SBU for MOF-324 and NH4@ZnPzC; see Figure 3); (ii) their crystal

■

RESULTS AND DISCUSSION Preparation and Deprotonation-Induced Cation Exchange. NH4@ZnPzC was prepared under alkaline conditions. The anionic framework [formula Zn3OH(PzC)3−] contains ammonium cations in the channels (Figure 1a), which is different from MOF-324 [formula Zn3OH(PzC)2(HPzC)] with partially deprotonated carboxylic moieties and a neutral framework.50 The complete deprotonation of the carboxyl sites in NH4@ZnPzC is confirmed by the disappearance of the strong band at 1658 cm−1 in the infrared spectra (Figure 1b, inset and Figure S2 in the Supporting Information). Except for the existence of extra-framework NH4+ cations, the crystal structure of NH4@ZnPzC is almost identical to that of MOF324, as indicated by PXRD patterns (Figure 1b). The NH4+ cation resides at the top of the Zn3O secondary building unit (SBU) and shows multiple contacts (dN3−O1 = 3.5625 Å; dN3−O3 = 2.7401 Å) with the COO− and OH− moieties in the SBUs, which are linked by the ligands into distorted cubic cages within the framework (Figure 2). The extra-framework NH4+ ions can be exchanged by other organic amine cations (Me3NH+ and Et3NH+) or metal ions (Li+, Mg2+, Eu3+, Tb3+, and mixed Eu3+/Tb3+). All the cation-exchanged samples retained their crystalline integrity, as shown by PXRD patterns (Figure 1c). TG analysis (Figure S1 in the Supporting Information) indicated NH4@ZnPzC showed better thermal stability (decomposition temperature Td = 400 °C) than MOF324 (Td = 250 °C).50 This is perhaps due to the existence of above-mentioned multiple contacts between NH4+ and the SBU. In comparison, Me3NH@ZnPzC and Et3NH@ZnPzC showed lower temperature to lose Me3N and Etz3N (150 and 230 °C, respectively), corresponding with the lack of hydrogenbonding-alike contacts. Mg@ZnPzC (Td = 440 °C) showed better thermal stability than that of the Li+-exchanged sample (Td = 390 °C), which is perhaps due to the higher valency of Mg2+ that results in stronger electrostatic interaction with the

Figure 3. Comparison of the M4O SBUs (M = Zn or Co) in three reported MOFs and Zn3O SBU in NH4@ZnPzC, as well as their simplified geometric shapes and underlying nets.

structures has different cubic space groups (MOF-5 crystallized in Fm3̅m, whereas MOF-324 and NH4@ZnPzC in Pa3̅; note pcu has symmetry Pm3m ̅ ). According to a consistent approach for MOF topological analysis,4,5,56 here we call attention to the essence of such an approach, in which one is supposed to consider linking geometric shapes to form the underlying nets; that is to say, the augmented net (e.g., pcu-a, also known as cab) is a better than its basic net (e.g., pcu) for describing the structures of MOFs. Figure 3 illustrates the procedure for deconstructing some related MOFs into their underlying nets. Similar to the carboxyl group, the pyrazolyl group can form Zn4O SBUs, which has been widely documented in the literature for MOFs with single or mixed carboxylate/pyrazolate ligands, such as Zn4O(BDC)3 [MOF-5, BDC = 1,4-benzenedicarboxylate],54 Zn4O(BDC)(BPz)2 [Bpz = 3,3′,5,5′-tetramethyl-4,4′-bipyrazolate],57 Co4O(BDPz)3 [MFU-1, BDPz = 1,4-benzenebis(3,5-dimethylpyrazolate)],58 and Zn4O(DMPzC)3 [DMPzC = 3,5,-dimethyl-4-pyrazolecarboxylate].59 It is unambiguous that these three M4O SBUs (M = Zn or Co) can be simplified into an octahedron in a topological viewpoint, and the corresponding net is pcu-a (i.e., cab).55,56 Now turning back to the Zn3O SBU with PzC ligands in this work, one can see that although each SBU is also linked to six neighboring ones, its geometric shape is more similar to a chair conformation other than an octahedron. By considering this chair shape in the simplifica3390



Article

Figure 4. (a) N2 adsorption isotherms of NH4@ZnPzC and cation-exchanged samples at 77 K (filled square: adsorption; open square: desorption). (b) Pore size distribution calculated from Saito−Foley method using N2 adsorption data at 77 K recorded at P/P0 < 0.6.

tion, the augmented net for NH4@ZnPzC is pcu-g.55 Note that pcu-g has the intrinsic symmetry of Ia3̅, which is more compatible with that of the real structure of NH4@ZnPzC (space group Pa3,̅ compared with pcu-a in Pm3m ̅ symmetry). Using the augmented nets is useful in (i) describing the exact structural feature of MOF structures (shown above for the distinction of Zn4O and Zn3O SBUs), (ii) distinguishing MOF structures that are closely related but different in symmetry, and (iii) predicting new theoretical nets for the design of MOFs (many of this type are collected in RCSR).55 For (ii), another interesting example is [Zn3S(AmTAZ)3]·NO3·H2O (HAmTAZ = 3-amino-1,2,4-triazole),60 which exhibited a similar Zn3S SBU, but it has a cationic framework filled with chargebalancing nitrate anions. At first glance, the Zn3S SBU is also 6coordinated, and thus the framework has a pcu topology, but the framework has two types of embedded cages. By considering the SBU as the chair conformation as in NH4@ ZnPzC, the augmented net turns out to be pbp,55 which retains the structural feature of two types of cages and can be distinguished from pcu-a and pcu-g. Concerning (iii), in relation to the cubic pcu-g (symmetry Pa3̅), there is a predicted 55 net pcu-h (hexagonal symmetry R3m ̅ ) in RCSR, which is also based on the chairlike building unit and awaits to be discovered in a real MOF. Porosity Modulation. After thermal activation, the NH4@ ZnPzC and cation-exchanged samples were subjected to N2 adsorption at 77 K (Figure 4a). The profiles of the N2 sorption isotherms in this work differ significantly from that of MOF324.50 All the samples showed Type IV sorption behavior according to the IUPAC classification of physisorption isotherms,2,6 showing two-step adsorption and obvious hysteresis upon desorption. The first-step uptakes at lower pressures and subsequent horizontal plateau corresponding to type I isotherms indicate the presence of micropores accessible to N2 molecules. The second-step uptakes at higher pressures show doubled uptakes for Et3NH@ZnPzC and Mg@ZnPzC and almost quadrupled uptakes for NH4@ZnPzC and the other three exchanged samples. Such a phenomenon is usually observed in mesoporous materials, indicating that during the crystal growth or postsynthetic modification processes there may be mesopores forming inside the crystals. However, further experiments are needed to validate this explanation. The pore size distribution plots (Figure 4b) indicate the porosity of the

exchanged materials are properly regulated, and there are indeed a certain content of pores of other sizes forming in some of the samples (e.g., NH4@ZnPzC and Tb@ZnPzC). Through postsynthetic cation exchange, the pore size and adsorption behaviors of the material can be properly modulated, although they showed reduced uptake capacity in contrast to MOF-324 (ABET = 1600 m2/g)50 and other topical MOFs (see Table 1). The BET surface areas of NH4@ZnPzC and the cation-exchanged samples are all lower than 500 m2/g (listed in Table 1). Table 1. Comparison of Adsorption Performances of MOFs in This Work and the Literature MOFs

ABET (m2/g)a

NH4@ZnPzC Me3NH@ZnPzC Et3NH@ZnPzC Li@ZnPzC Mg@ZnPzC Tb@ZnPzC MOF-32450 Cu-BTTri27 en-CuBTTri27 mmen-CuBTTri30 Mg-MOF-7422,26 mmen-Mg2(dobpdc)28

132.35 97.33 479.61 56.15 359.25 71.23 1600 1770 345 870 1495 70

CO2 uptake (wt %)b

Qst (kJ/mol)c

CO2/N2 selectivityd

3.73 4.06 7.24 4.32 5.29 2.24

37.2−31.1 40.6−22.2 31.6−26.4 48.5−24.7 32.4−21.1 39.5−32.5

29.2 43.2 38.7 50.8 21.9 21.5

14.3 5.5 15.4 35.2 14.5

21 90 96 47 71

21e 25e 327e,f 148.1f,g 200e,f,h

a

BET surface area. bCO2 uptake capacity obtained at 298 K and 1 atm. Isosteric heat of adsorption for CO2 uptake. dIAST selectivity for 15:85 CO2/N2 at 273 K and 1 atm. eObtained at 298 K. fCO2:N2 15:75. gObtained at 323 K. hMolar selectivity. c

Tunable CO2/N2 Adsorption Selectivity. We focused on examining the influence of cation exchange on the selective adsorption of CO2 over N2, targeting optimized CO2/N2 selectivity. The CO2 and N2 adsorption isotherms (273 K) for NH4@ZnPzC and the cation-exchanged samples are compared in Figure 5a. As expected, all the samples showed a much higher uptake capacity for CO2 compared with N2, which is attributed to the greater quadrupole moment of CO2 facilitating its strong affinity with the extra-framework cations.6 Under ambient conditions (298 K, 1 atm), all exchanged 3391



Article

Figure 5. N2 and CO2 adsorption properties of NH4@ZnPzC and cation-exchanged samples. (a) Adsorption isotherms at 273 K. (b) Isosteric heats of CO2 adsorption calculated from isotherms at 298, 303, and 308 K. (c) IAST selectivity for a 15:85 CO2/N2 mixture at 273 K.

In particular, the highest zero-coverage Qst (48.5 kJ/mol) for Li@ZnPzC is consistent with the above observation on CO2 uptakes. This is reasonable because (i) the exchanged metal content for Li@ZnPzC is about twice or triple that for Mg@ ZnPzC and Tb@ZnPzC, respectively, due to charge balance, and (ii) Li+ has the smallest ionic radius of all exchanged cations and thus the strongest polarization ability to enhance the electrostatic potential interacting with CO2. Compared with several topical MOFs developed for CO2 capture (see Table 1), the zero-coverage Qst values (ranging from 31.5 to 48.5 kJ/mol) in this system can be considered appropriate for practical application, which are high enough for effective selective adsorption of CO2 over N2, but not as high as those of aminegrafting MOFs such as en-CuBTTri (90 kJ/mol), mmenCuBTTri (96 kJ/mol), and mmen-Mg2(dobpdc) (71 kJ/mol). This is beneficial for minimizing the energy penalty for regeneration.6 We then evaluate the CO2/N2 adsorption selectivity by applying IAST (see the Supporting Information for details).6,22,53 The results are depicted in Figure 5c and compared with literature in Table 1. Among all the samples, Li@ZnPzC showed the highest CO2/N2 selectivity (67.5 at 0.1 bar, 50.8 at 1 bar) over the entire pressure range, in accord with the highest zero-coverage Qst mentioned above. The modulation of CO2/ N2 selectivity through cation exchange is realized. Another material worthy of mentioning is Et3NH@ZnPzC, which combines the merits of competent CO2 adsorption capacity (12.7 wt % at 273 K, 7.24 wt % at 298 K), appropriate Qst values (31.6−26.4 kJ/mol) and considerable CO2/N2 selectivity (38.7 at 1 bar). More importantly, it exhibits almost constant CO2/N2 selectivity upon increasing the pressures

samples showed greater uptake amounts of CO2 than that of NH4@ZnPzC (see Table 1). Among them Et3NH@ZnPzC has the highest CO2 uptake capacity (12.7 wt % at 273 K, 7.24 wt % at 298 K). The CO2 uptake capacity in this system has limited relation with the surface areas. For example, Li@ZnPzC has the lowest BET surface area. Nevertheless, it exhibited the third highest CO2 adsorption capacity, and its low-pressure uptake (