Polyoxometalate-Based Ionic Frameworks for Highly Selective CO 2 Capture and Separation
2020; Chinese Chemical Society; Volume: 3; Issue: 11 Linguagem: Inglês
10.31635/ccschem.020.202000498
ISSN2096-5745
AutoresFengxue Duan, Xiaoting Liu, Di Qu, Bao Li, Lixin Wu,
Tópico(s)Carbon dioxide utilization in catalysis
ResumoOpen AccessCCS ChemistryRESEARCH ARTICLE1 Nov 2021Polyoxometalate-Based Ionic Frameworks for Highly Selective CO2 Capture and Separation Fengxue Duan, Xiaoting Liu, Di Qu, Bao Li and Lixin Wu Fengxue Duan State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University, Changchun 130012 , Xiaoting Liu State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University, Changchun 130012 , Di Qu State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University, Changchun 130012 , Bao Li *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University, Changchun 130012 and Lixin Wu *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University, Changchun 130012 https://doi.org/10.31635/ccschem.020.202000498 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail During the utilization of structural and functional advantages of polyoxometalates (POMs) for enhanced applications, a suitable assembly of these clusters in framework materials to act as binding nodes represents a promising approach. In contrast to well-developed coordination/covalent combinations, we have developed a convenient strategy to build porous structures of POMs with smaller-sized counterions as bridging ligands via ionic interactions to reinforce their capability in gas adsorption in parallel to metal–organic frameworks (MOFs)/covalent organic frameworks (COFs). With this goal, a series of POMs-based ionic frameworks (IFs) were constructed with triol-ligand modified Anderson–Evans-type clusters as building blocks, and their sodium counter-cations were used as linkers. The three-dimensional (3D) open-frameworks obtained displayed unusually selective CO2 capturing capability and efficient separation from their N2, H2, and CH4 mixtures under low pressure at room temperature. Among the synthesized IFs, the cobalt-centered cluster exhibited the best performance for the uptake and selective separation of CO2 over N2 and CH4 in a range of 0–1 bar, while the nickel-centered cluster displayed the highest selectivity over H2 at 1 bar. Breakthrough experiments based on real binary gas mixtures demonstrated that the cobalt-containing framework illustrated high performance in the actual gas separation and sustained stability against a simulated operating environment. Download figure Download PowerPoint Introduction Porous materials possessing large surface area and void open space within their bulk structures display various important functional features for extensive applications in adsorptions and separations, as different catalytic supports, as drug carriers, and so forth.1,2 Among the well-known porous materials, framework compounds such as inorganic zeolites, metal–organic frameworks (MOFs), and covalent organic frameworks (COFs) represent a typical class of crystalline microporous materials bearing precise porosity and chemical composition, with changeable structure modification and surface property through a rapid synthesis.3,4 Huge attention has been paid to the encyclopedically designed architectures to understand synthetic mechanisms and structure esthetics.5,6 One of the everlasting issues of this field focuses on the extension of driving forces and the resultant excess properties so that more desirable functions can be realized accordingly. In contrast to the coordination and covalent bonds used in the construction of MOFs/COFs by providing the convenient structural control, ionic interaction is less applied in the fabrication of frameworks, although such a driving force can also offer novel linking modes and potent combinations between building units and the unique chemical microenvironments.7–9 The main reasons for this lack may be ascribed to the few functional building units and the difficulties encountered in forming framework structures from common ionic components. However, once the intrinsic disadvantages are overcome effectively, superior profiles such as a suitable ionic environment could be expected for various potentials. Polyoxometalates (POMs) are types of typical nanosized functional clusters consisting of early transition metal oxide subunits and possessing various structural architectures, morphologies, along with tunable negative surface charges.10,11 The published results demonstrate the formation of ionic frameworks (IFs) through suitable control of the preparation conditions and prove that the conductivity can be realized via the counterions surrounding the oxygen-rich surface of the nanosized clusters.12 Some POM-based IFs are synthesized following the strategies of hydrothermal or beaker reactions, and their high performances on proton conductivity, selective uptake, and exchange of cations, as well as catalysis have been achieved.13–15 However, unlike the broad investigations of POM-incorporated MOFs/COFs in catalysis, reports on essential properties of POM-based IFs such as gas adsorption and separations, significant in clean energy and environmental science, are still limited and not up to date. CO2 is the predominant greenhouse gas source, and its massive emission has caused one of the most severe issues of the environment globally.16 Therefore, treatment of CO2 mostly generated from the energy consumption of fossil fuels becomes a matter of the utmost urgency, and the highly efficient usage of the waste gas is one of the intrinsic driving forces. Thus, the collection and separation of CO2 as a carbon source for producing various chemical products and materials via cheap physical and chemical routes is desired.17,18 To realize the continuous utilization of CO2 recycling, highly efficient collection, separation, and transformation toward decreasing energy consumption are still critical challenges; thus, new materials are highly desired.19–21 Some POMs are demonstrated to be effective catalysts for transforming CO2 into various applicable organic products.22–24 Therefore, it will be significant if the catalytic properties of POMs can be combined with the trapping and separation of CO2 because such integration leads to multiple applications of a single material. Thus, it is inevitable to prepare POM materials possessing strong capabilities, as the first step, to collect and purify CO2. During exploring the methodology for organic modifications of Anderson–Evans-type POMs, some clusters are found to self-assemble into various packing forms in crystals.25–27 In this context, we report a beaker preparation of a series of cluster IFs, which can afford cluster properties in a new structural platform for the probability of the adsorption and separation properties regarding greenhouse gases. The IFs comprised organically modified Anderson–Evans-type POMs and named IF-M-Mes (where M stands for central heteroatom and Me denotes 2-(hydroxymethyl)-2-methyl-propane-1,3-diol). Interestingly, the results obtained showed a simplified IFs synthesis and built-up architecture with comparable advantages as most known IFs, with separation capabilities of more than three gases simultaneously. Noticeably, the present achievements demonstrate a useful principle that IF materials can have similar or even better gas adsorption and separation performance than many MOFs and COFs; meanwhile, other unique applications could be expected. Experimental Section Materials All chemicals were purchased from Aladdin Bio-Chem (Shanghai, China), and were used without further purification. Double-distilled water was used throughout the experiments. Measurements Fourier transform infrared (FT-IR) spectra were recorded on a Bruker Vertex 80v spectrometer (Bruker, Shanghai, China) equipped with a deuterated triglycine sulfate (DTGS) detector (32 scans) at a resolution of 4 cm−1 using KBr pellets. Organic elemental analysis for C, H, and N was carried out on a Vario MICRO cube from Elementar Trading Co. Ltd. (Shanghai, China). Inorganic elemental analysis for Fe, Co, Ni, Cu, Zn, Na, and Mo was taken on a PLASMA-SPEC (I) (Elemental Analysis Inc., Lexington, KY) inductively coupled plasma atomic emission spectrometry (ICP-AES) . Thermogravimetric analysis (TGA) was carried out on a TA Instruments Q500 Thermal Analyzer (TA Instruments, Shanghai, China) with a nitrogen flow, and the temperature was set from 30 to 800 °C under a heating rate of 10 °C/min. Single-crystal X-ray diffraction indexing and data collection were performed on a Bruker D8 Venture diffractometer (Bruker) with graphite-monochromated Mo Kα (λ = 0.71073 Å) at 293 K. All crystal structures were solved using SHELXTT and refined by full-matrix least-squares fitting on F2 via the SHELXTL software (SHELXTL Package, Tianjin, China). SQUEEZE function in the PLATON program of the SHELXTL software was used to remove the residual electron density, which could not be modeled precisely. Powder X-ray diffraction (PXRD) data were obtained from a Rigaku SmartLab X-ray diffractometer (Beijing, China) using Cu Kα radiation at a wavelength of 1.54 Å from 3° to 50°. The adsorption isotherms for CO2, N2, H2, and CH4 were measured using a Quantachrome Autosorb-IQ analyzer (Shanghai, China) with the ultra-high-purity gas (99.999% purity). Synthesis of IF-M-Mes For the synthesis of Na4/3(NH4)8/3{FeMo6O18[CH3C(CH2O)3]2}·1.2H2O (IF-Fe-Me), FeSO4·7H2O (0.56 g, 2.00 mmol), (NH4)6Mo7O24ˑ4H2O (2.47 g, 2.00 mmol), and CH3C(CH2OH)3 (0.48 g, 4.00 mmol) were dissolved in 20 mL of CH3COONa/CH3COOH buffer solution (pH 4.7) and the mixture was heated to 80 °C for 30 min under continuous stirring. The green solution obtained was cooled to room temperature, and light-green crystals were formed after 1 day of 68.6% yield based on Mo. Elemental analysis Calcd for IF-Fe-Me (M = 1254.12 g/mol): Fe, 4.45; Mo, 45.90; Na, 2.44; C, 9.58; H, 2.50; N, 2.98. Found: Fe, 4.46; Mo, 45.92; Na, 2.47; C, 9.56; H, 2.48; N, 2.97. IR (KBr, cm−1): 3448, 3163, 2917, 2855, 1620, 1405, 1108, 1039, 923, 900, and 630. A similar synthetic procedure was applied for the fabrication of Na4/3(NH4)8/3{CoMo6O18[CH3C(CH2O)3]2} (IF-Co-Me), as aforementioned except that CoSO4ˑ7H2O (0.56 g, 2.00 mmol) was used instead of hydrated Fe sulfate salt. Purple crystals can be obtained after 2 days in 60.8% yield based on Mo. Elemental analysis Calcd for IF-Co-Me (M = 1235.58 g/mol): Co, 4.77; Mo, 46.59; Na, 2.47; C, 9.72; H, 2.34; N, 3.03. Found: Co, 4.86; Mo, 46.96; Na, 2.51; C, 9.68; H, 2.45; N, 3.00. IR (KBr, cm−1): 3417, 3194, 2925, 2856, 1628, 1401, 1081, 1016, 931, 900, and 643. A similar procedure as mentioned above was used for the synthesis of Na4/3(NH4)8/3{NiMo6O18[CH3C(CH2O)3]2}·0.8H2O (IF-Ni-Me), except that NiSO4ˑ6H2O (0.53 g, 2.00 mmol) was used instead of hydrated Fe or Co sulfate salts. Cooling the resulting in dark green solution to room temperature generates dark-green crystals after 1 day in 55.0% yield based on Mo. Elemental analysis Calcd for IF-Ni-Me (M = 1249.77 g/mol): Ni, 4.70; Mo, 46.06; Na, 2.45; C, 9.61; H, 2.44; N, 2.99. Found: Ni, 4.75; Mo, 46.18; Na, 2.59; C, 9.70; H, 2.52; N, 2.91. IR (KBr, cm−1): 3479, 3160, 2990, 2848, 1629, 1416, 1108, 1031, 935, 900, and 639. During the synthesis of Na4/3(NH4)8/3{CuMo6O18[CH3C(CH2O)3]2}·1.3H2O (IF-Cu-Me), a similar synthetic procedure as mentioned above in IF-Co-Me was applied, except that CuSO4ˑ5H2O (0.50 g, 2.00 mmol) was used instead of hydrated Fe or Ni sulfate salts. Blue crystals were obtained from the reaction solution after 1 day in 57.8% yield based on Mo. Elemental analysis Calcd for IF-Cu-Me (M = 1249.60 g/mol): Cu, 5.03; Mo, 45.56; Na, 2.42; C, 9.50; H, 2.50; N, 2.96. Found: Cu, 5.12; Mo, 46.09; Na, 2.51; C, 9.60; H, 2.49; N, 2.99. IR (KBr, cm−1): 3467, 3160, 2925, 2851, 1643, 1403, 1101, 1024, 931, 900, and 646. For the synthesis of Na4/3(NH4)8/3{ZnMo6O18[CH3C(CH2O)3]2}·1.2H2O (IF-Zn-Me), a similar synthetic procedure as that of IF-Co-Me was applied, except that ZnSO4ˑ7H2O (0.58 g, 2.00 mmol) was used instead of hydrated Fe or Cu sulfate salts. Colorless crystals were obtained after 1 day in 57.8% yield based on Mo. Elemental analysis Calcd for IF-Zn-Me (M = 1263.04 g/mol): Zn, 5.18; Mo, 45.55; Na, 2.42; C, 9.50; H, 2.48; N, 2.96. Found: Zn, 5.12; Mo, 45.88; Na, 2.51; C, 9.60; H, 2.56; N, 2.88. IR (KBr, cm−1): 3460, 3144, 2917, 2840, 1632, 1413, 1104, 1024, 923, 900, and 651. Results and Discussion Preparation and characterization of IFs Based on the structural characterization, the prepared 3D IFs displayed a common chemical formula of Na4/3(NH4)8/3{MMo6O18[CH3C(CH2O)3]2} (shortened as IF-M-Me)·× H2O, in which FeII, CoII, NiII, CuII, and ZnII, played the roles of central heteroatoms, Ms, as described in the Experimental Section. Like the general synthesis of organically modified Anderson–Evans clusters, each IF-M-Mes was prepared by simply heating a mixture of (NH4)6Mo7O24ˑ4H2O, transition metal salt, and a triol ligand in a HAc/NaAc buffer solution (1.0 M, pH 4.7). The buffer solution was a prerequisite for successfully synthesizing the IFs because the same products were unobtainable by simply adjusting the pH to 4.7 in the initial reaction solution. The Triol ligand is another critical factor, as no frameworks were obtained without its addition under the same synthetic reaction conditions. Also, we sought to synthesize the frameworks with single ammonium or sodium as a counter cation, but only a discrete cluster or two-dimensional (2D) network was obtained. Attempts for applying other alkali–metal ions as counter cations also failed to acquire the desired 3D frameworks. FT-IR spectra displayed absorption bands at 630–940 cm−1 and 1010–1110 cm−1, ascribed to the vibration modes of Mo=O, Mo–O–Mo and C–O–M bonds, indicating the formation of a complete polyanion cluster and a successful organic modification ( Supporting Information Figure S1 and Table S1).28,29 The single-crystal XRD analysis revealed that all five IF-M-Mes were isostructural and crystallized to the same space group of F m ¯ 3 ( Supporting Information Table S2). The 3D structures obtained (Figure 1a) were created from Anderson–Evans-type polyanions modified with triol-ligand on double-sides in δ/δ-isomer. All Na+ counterions served as the linker to connect the clusters through ionic bonds. Each of them was linked to three polyanions through an Na–O bond by localizing at the center of a heteroatom triangle bearing a side length of 10.26 Å (Figures 1b and 1c). Each polyanionic cluster was surrounded by four Na+ ions, forming a rectangular plane with a length of 9.59 Å and a width of 6.95 Å (Figures 1d and 1e). The two subunits co-assembled into a 3D cubic framework, as indicated in the crystal structure of IF-Co-Me (Figure 1f). Importantly, each cube with a length of 14.50 Å possessed an 18.38% void space in which the ammonium cation residues were excluded (Figure 1g). An identical pore size of 5.84 Å × 6.08 Å on each face of the cubic comprised three rectangular channels running through the 3D framework along the directions perpendicular to each other. The full detailed porosity data of all IF-M-Mes are summarized in Supporting Information Table S3. Figure 1 | Schematic illustration for the 3D structure of IF-M-Mes, which shows (a) the isolated building blocks {MMo6O18[CH3C(CH2O)3]2}4−, (b) Na+-centered trimer and (c) its steric sketch drawing, (d) position of Na+ tetramer and (e) its steric sketch drawing, (f) local 3D-packing structure of IF-Co-Me in ab plane and (g) sketch diagram together with trimer and tetramer locations in dashed lines corresponding to (c) and (e). M represents the center heteroatom, FeII, CoII, NiII, CuII, and ZnII. Download figure Download PowerPoint Though some of the Anderson–Evans-type clusters were applied in the fabrication of IFs via a hydrothermal route,12 the frameworks here showed differences of spatial structures and binding forms, compared with reported results. The detailed changes include the space group, the number of Na+ ions surrounding each polyanionic cluster, which caused an intensity improvement of electrostatic interactions, the pore shape being closer to a cube, and the larger void percentage of the whole cell volume ( Supporting Information Table S3). The differences in the starting materials and the reaction conditions provide a favorable architecture on the functional performance of obtained IFs in the following gas adsorption and separation. When 2-(hydroxymethyl)propane-1,3-diol, a smaller organic ligand, was applied instead, identical IFs to those modified with 2-(hydroxymethyl)-2-methyl-propane-1,3-diol were obtained as well. Although, it was presumed that this present framework structure was challenging to prepare via a hydrothermal approach due to the proposed unfavorable space between two methyl groups ( Supporting Information Figure S2a),12 a very little rotation of the cluster improved the local space to accommodate the two groups. In fact, in this present crystal structure shown in IF-Co-Me ( Supporting Information Figure S2b), the distance between a methyl carbon and a nearby oxygen atom reached 3.3 Å. An increased distance of 0.3 Å just matched the sum of the van der Waals radius of the carbon (1.7 Å) and oxygen (1.5 Å) atoms. Thermal, humid, and solvent stabilities of IF-M-Mes TGA was carried out to evaluate the thermal stability of synthesized framework crystals, and the weight loss plot versus the temperature increase shows that the IF-M-Mes experience no apparent structure change below 105 °C except that a bit of weight loss was observed, attributed to the partial dehydration ( Supporting Information Figures S3–S7). Due to each void unit of the cubic framework, the calculated number of water molecules were 70% of the initial species, without suffering from humidity. These results indicated that IF-Co-Me did not experience a permanent loss of capacity for CO2 uptake in a humid environment. Notably, after the drying process, the uptake capacity recovered to the initial state of the same sample. This performance could be ascribed to the strong binding ability between POM anions and cations in IFs due to the high ionic bond energy by referring to that of coordination bonds in MOFs. Figure 6 | Histogram of CO2 uptake values for IF-Co-Me after it was treated in a humid environment of 90% RH for 0 to 5 days at 273 K and 1 bar. Download figure Download PowerPoint According to published results, the dynamic diameter of CO2 is 3.3 Å, smaller than both the sizes of N2 (3.6–3.8 Å) and CH4 (3.8 Å), but larger than that of H2 (2.8–2.9 Å) ( Supporting Information Table S7).34 Apparently, we found that the size scale of all used gases was smaller than the channels in the framework structure, which enabled these gases to obtain fluent access. Therefore, the molecular sizes seem not to be the determinant causing differences in gas adsorption. On the contrary, the excess volumes of N2 and CH4 over CO2 were also probably not the reason for such small uptake volumes concerning the final extent of adsorption of CO2. Subsequently, we checked the physicochemical parameters of the used gases. We found that in contrast to MOFs and COFs, IFs bore some unique features. The positive and negative charge separation occurred at the inner channel wall and outside surface, while the POM cluster was exposed mostly to the surfaces because sodium counterions yielding ionic interactions covered minimal surface area. As the abundant O atoms on the surface of clusters could play the role of a base, their interaction with acidic CO2 resulted in enhanced adsorption relative to other gases without the acidity. This result also implied stronger adsorption interaction of CO2 with the porous solid surface than with other CO2 molecules, leading to Type II adsorption, while other gases without acid could only afford weak surface adsorption in Type III adsorption. Besides, the IFs at charge separation state facilitated their surface interaction with polarized molecules like CO2 with the polarizability of 29.11 × 1025 cm3 than CH4, N2, and H2 with polarizabilities of 25.93, 17.40, and 8.04 × 1025 cm3.4,34 On the other hand, in the IFs, the potential local gradients allowed the surface to adsorb the molecules with higher quadrupole moments (CO2: 4.30 × 1026 esu·cm2, N2: 1.52 × 1026 esu·cm2, H2: 0.66 × 1026 esu·cm2, and CH4: 0) and apparently, CO2 had the superior quadrupole moment for adsorbing on the IF-M-Mes surface.4,34,35 These surface characteristics contributed to the surpassed the adsorption of CO2, and the selectivity was higher than the other three gases. Ideal adsorbed solution theory to predict selectivity of IF-M-Mes To determine the adsorption selectivity of CO2 from N2, H2, and CH4 in a binary mixture, the ideal adsorbed solution theory (IAST) was applied, which was suitable for two-component mixtures from experimental single-component isotherms.33 The adsorption selectivity could be described with the following equation36: S g 1 / g 2 = q 1 / q 2 p 1 / p 2 where g1 denotes CO2, g2 means N2, H2, or CH4, q1 and q2 are the quantities of the corresponding gases adsorbed, while p1 and p2 are the partial pressures of corresponding gases in the mixture. At a given volume fraction of 15/85 for CO2/N2, the selectivity values of IF-Fe-Me, IF-Co-Me, IF-Ni-Me, IF-Cu-Me, and IF-Zn-Me were calculated to be ∼225, 188, 137, 61, and 28 at 1 bar, as shown in Figure 7a. It was evident that IF-Co/Fe-Mes showed higher CO2/N2 selectivity from 0 to 1 bar at 273 K, while IF-Co-Me possessed
Referência(s)