High and selective sorption of C₂ hydrocarbons in heterometal–organic frameworks built from tetrahedral units

Abstract

Alternate connection of tetrahedral [In(COO)₄]⁻ and [AgN₄]⁺ (or [CuN₄]⁺) units via nicotinate ligands leads to the formation of two metal–organic frameworks (MOFs) based on packing layers. The activated framework shows high C₂ hydrocarbons (C₂s) uptake and high selectivities of C₂s/CH₄.

---

C₂ hydrocarbons (C₂s) are usually considered to be important raw chemicals and are directly applied to some special fields, such as refrigerants, welding, ripeners and so on. Thus, the sorption and separation of C₂s from methane (C₁) is beneficial to not only the purification of natural gas, but also the recycling of C₂s.¹ Compared to traditional energy-consuming separation of C₂s from C₁, metal–organic frameworks (MOFs) as a relatively new class of microporous solid adsorbents have recently emerged their significantly potential application in the field of small hydrocarbons adsorption and separation under ambient conditions.² Current efforts on the sorption and separation of C₂s have shown that one of the most promising strategies is to deliberately enhance interaction of the adsorbent–adsorbate.³ Some factors that have been considered in this context include (1) creating open metal sites in the frameworks,^3a,b (2) controlling appropriate pore sizes,⁴ and (3) forming π-complexation between Ag(I) ions and the carbon–carbon double bonds of olefin.^3c,d Recently, some MOFs show higher uptakes and selectivity for this C₂s/C₁ separation under ambient conditions to significantly conserve energy.^2,3 For example, the well-known MOF-74 materials with a large amount of accessible metal sites on the hole-walls showed very high C₂H₂ uptake of 150 to 230 cm³ cm⁻³.^2d,3a,b Chen's group optimized C₂s uptakes and C₂s/C₁ selectivities of UTSA-33 and UTSA-34 by tuning their pore sizes and creating open metal sites.⁴ Ma's group first demonstrated that the introduction of π-complexation into a Ag(I) ion functionalized porous aromatic framework could afford a significant increase in ethylene uptake and ethylene/ethane selectivity.^3c

In this work, nicotinic acid (na) was chosen as a ligand with both N- and O-donor atoms for synthesizing heterometallic porous MOFs made of hard and soft metal ions, namely, In(III) and Ag(I) or Cu(I), which are apt to form tetrahedral [In(COO)₄] and [AgN₄] units.⁵ Thus, according to the “hard and soft acids and bases” (HSAB) interaction principle, two foreseeable heterometallic MOFs, namely, [InAg(na)₄]·1.5DMF (1; DMF = N,N-dimethylformamide) and [InCu(na)₄]·1.5DMF (2) were obtained and structurally characterized. Both compounds consisted of the alternate connection of equal amounts of [In(COO)₄]⁻ and [AgN₄]⁺ or [CuN₄]⁺, forming neutral 4-connecting frameworks containing 1D channels with a size of about 7 × 7 Ų. The activated empty phase of 1 had an excellent C₂s uptake in excess of 100 cm³ cm⁻³ at 273 K and 1 bar and showed a selectivity of 55.7 from C₂H₆ over CH₄ at a range of pressures up to 100 kPa.

Compound 1 was synthesized by mixing nicotinic acid, AgNO₃ and In(NO₃)₃·6H₂O in a 4 : 1 : 1 molar ratio in DMF solution at 100 °C for 12 hours. Single-crystal X-ray diffraction revealed it is formulated as [AgIn(na)₄]·1.5DMF (1). The phase purity of the bulk product was independently confirmed by powder X-ray diffraction (PXRD). Compound 1 crystallized in the chiral space group I222. Following the “soft–soft” and “hard–hard” (HSAB) interaction principle, each Ag center adopts a tetrahedral AgN₄ geometry by coordinating to four pyridine nitrogen atoms of na ligands; while each In center is chelated by four carboxylate groups of na ligands, giving another tetrahedral In(COO)₄ unit (Fig. 1a). Such two kinds of tetrahedral geometries are alternately linked each other through na ligands into a layer with an edged Ag⋯In distance of 8.0459(5) Å (Fig. 1b). From the viewpoint of structural topology, the layer can be topologically represented as a 4-connected sql net by reducing each metal center as a 4-connected node (Fig. 1b). These layers further stack into a 3D supramolecular architecture in an ABAB fashion, leaving 1D channels with a size of about 7 × 7 Ų along c directions (Fig. 1c). The V_void of 1 without guest DMF molecules is 40.1% of the unit cell volume as calculated by PLATON.

Fig. 1 (a) The coordination environment in 1; (b) the grid-like 4⁴ layers with sql topology; (c) view of the 3D framework packed from layers in 1 along the a-axis.

It is well-known that Cu(I) is more apt to adopt a tetrahedral geometry. When CuI was used to replace AgNO₃ during the synthetic process, compound 2 with CuN₄ and In(COO)₄ tetrahedral units was formed and crystallized in the achiral space group I4̄2m. The main difference between 1 and 2 is the slight change of bond length and angle. A solvent accessible volume of 38.9% was found in 2. In the structures of 1 and 2, the free channels are occupied by guest DMF molecules, being equal to about 1.5 DMF per unit formula of the host, as evidenced by thermogravimetric analysis (TGA).

The thermal stability of 1 and 2 was also confirmed by TGA and PXRD measurements. The TGA curves of 1 and 2 show that the trapped solvent DMF molecules can be drastically removed at high temperature (200 to 250 °C) (Fig. S1, ESI†). Such a high evacuated temperature goes against the activation of the sample for gas adsorption. To obtain a more complete evacuation and protect the structural integrality, DMF was exchanged with CH₂Cl₂ by soaking crystals of 1 and 2 in CH₂Cl₂ for one week, and then the hollow phases, 1a and 2a, could be obtained by putting CH₂Cl₂-exchanged 1 and 2 under high vacuum at 60 °C for 24 hours. Further PXRD experiments were carried out to verify that the framework of 1a can be retained (Fig. S2, ESI†). However, some peaks in the PXRD pattern of 1a were distinctly weakened and broadened, indicating the possible incline of the framework due to the easy changeable coordination geometry of the Ag⁺ center. However, the good agreement of the peaks in all diagrams demonstrates that the framework structure of 2a is more stable than that of 1a (Fig. S3, ESI†). Additionally, the second-harmonic generation (SHG) property of the microcrystalline sample reveals that the bulk materials for 1 display powder SHG efficiencies, which are approximately 0.63 time that of a potassium dihydrogen phosphate (KDP) powder, and further confirms the chiral property of the framework (Fig. S4, ESI†).

The permanent porosity of 1a and 2a was established by reversible gas sorption experiments by using N₂ and CO₂. Unexpectedly, the two compounds can hardly adsorb N₂ (Fig. 2c and S5a, ESI†). The low uptake of N₂ is similar to a microporous Ni-MOF reported by Kitagawa.⁶ This interesting fact is the case that some N₂ molecules strongly bind to the metal centers around pore windows of 1D channels to prevent the entering of other nitrogen molecules. It is sure to exclude the size-exclusion effect for such selective adsorption because the pore window size (∼7 Å) is much larger than the kinetic diameter of N₂ (3.64 Å). The same phenomenon can also be seen in two 3-fold diamond networks with the same effective pore window size to that of 1a and 2a.^5,7

Fig. 2 Gas sorption isotherms for 1a: CO₂ at 273 K (a) and 298 K (b), N₂ at 77 K (c), CH₄ at 273 K (d) and 298 K (e).

The adsorption isotherms of CO₂ for 1a and 2a were measured at 273 and 298 K, respectively. At 273 K and 1 bar, the uptake values of CO₂ of 1a and 2a were 94.0 cm³ cm⁻³ and 89.5 cm³ cm⁻³, respectively (Fig. 1a and b, S5b and S5c, ESI†). To better understand these observations, the adsorptive capability of 1a and 2a were also measured under 298 K, giving an uptake of 75.0 cm³ cm⁻³ for 1a and 66.7 cm³ cm⁻³ for 2a. It should be noted that the CO₂ isotherms of 1a show a two-step adsorption profile at 273 and 298 K, respectively. The CO₂ isotherm at 273 K shows a sudden increase at P = 0.04 bar and exhibits a hysteresis. Such observations are well documented for a number of previously reported flexible MOFs where pressure-dependent structure changes occur during adsorption–desorption processes.⁸ The atypical CO₂ adsorption behavior can also be seen at 298 K. However, the adsorption step and hysteresis move to higher absolute pressure (P = 0.20 bar). In contrast, the CO₂ sorption isotherms of 2 at 273 and 298 K show no adsorption step and hysteresis (Fig. S5b and S5c, ESI†). In addition, the enthalpies of CO₂ adsorption for 1a and 2a are 21.9 and 23.6 kJ mol⁻¹ at zero coverage, respectively (Fig. S6–S9, ESI†).

In order to examine the utility of 1a as an adsorbent for the separation of industrially important small hydrocarbon, we examined the pure component sorption isotherms of 1a for various C₂s and CH₄ at both 273 and 298 K and 1 bar. As shown in Fig. 2 and 3, 1a uptakes different amounts of C₂H₆ (107.37 and 94.23 cm³ cm⁻³), C₂H₄ (102.52 and 92.00 cm³ cm⁻³), C₂H₂ (117.33 and 98.14 cm³ cm⁻³) and CH₄ (49.55 and 36.76 cm³ cm⁻³) at 273 and 298 K, respectively. Notably, there is small hysteresis during the adsorption–desorption processes of these C₂s, especially for C₂H₆. It is not common to many other MOFs and can be attributed to the strong adsorbent–adsorbate interaction.⁹ The enthalpies of C₂H₆, C₂H₄ and C₂H₂ adsorptions for 1a estimated from the sorption isotherms at 273 and 298 K using the virial equation are 43.76, 23.49 and 24.79 kJ mol⁻¹, respectively (Fig. 3, inset; Fig. S10–S12, ESI†). The high enthalpies of C₂H₆ for 1a bears a comparison with the well-known MOF-74, M₂(dobdc) (M = Mg, Mn, Fe, Co, Ni, Zn) that showed selective adsorption of small alkane over alkene.^3a The maximum uptakes of C₂ hydrocarbons for 1a is much better than the recently reported UTSA-33a (C₂H₆: 77.1 cm³ cm⁻³ at 273 K and 61.4 cm³ cm⁻³ at 296 K; C₂H₄: 83.9 cm³ cm⁻³ at 273 K and 60.5 cm³ cm⁻³ at 296 K; C₂H₂: 111.2 cm³ cm⁻³ at 273 K and 83.0 cm³ cm⁻³ at 296 K) and UTSA-34a (C₂H₆: 64.4 cm³ cm⁻³ at 273 K and 55.5 cm³ cm⁻³ at 296 K; C₂H₄: 69.5 cm³ cm⁻³ at 273 K and 52.4 cm³ cm⁻³ at 296 K; C₂H₂: 92.2 cm³ cm⁻³ at 273 K and 71.5 cm³ cm⁻³ at 296 K) under similar conditions.⁴ The maximum uptake (98.14 cm³ cm⁻³) of C₂H₂ for 1a at 298 K is inferior to those of MOF-74 materials (150 to 230 cm³ cm⁻³) with unsaturated metal ions, but equal to that for UTSA-34b (102 cm³ cm⁻³) and much better than many other MOFs and ZIFs, such as the famous MOF-5 (15 cm³ cm⁻³), MIL-53 (67 cm³ cm⁻³) and ZIF-8 (23 cm³ cm⁻³).^2e

Fig. 3 Gas sorption isotherms for 1a: C₂H₆ at 273 K (a) and 298 K (b), C₂H₄ at 273 K (c) and 298 K (d), C₂H₂ at 273 K (e) and 298 K (f). Inset: the Qst isotherms for C₂H₆ (a), C₂H₄ (b), and C₂H₂ (c).

The experimental data are fits of the single-site Langmuir–Freundlich mode (Fig. S13, ESI†) and the adsorption selectivities for equimolar mixture adsorption of different hydrocarbons with respect to CH₄ at 273 K were calculated using ideal solution adsorbed theory (IAST), as a widely adopted method to predict multicomponent isotherms from pure gas isotherms. The selectivities of C₂H₆, C₂H₄ and C₂H₂ components with respect to CH₄ at 273 K are in excess of 50.7, 20.2 and 17.3 for a range of pressures up to 100 kPa, respectively (Fig. 4). The C₂H₆/CH₄ separation selectivity of 1a is much higher than ZIF-8 (10), UTSA-33a and UTSA-34b (18–20).⁴ It is also worth to note that 1a with a high selective separation of C₂H₆ over CH₄ may be used for hydrocarbon separation and purification of natural gas.

Fig. 4 IAST predicted selectivity for C₂H₆/CH₄ (a), C₂H₄/CH₄ (b) and CO₂/CH₄ (c) in 1a at 273 K.

In summary, two heterometallic MOFs built from mixed tetrahedral units show porous layer-packing and high uptake capacities for C₂ hydrocarbons. These developed MOFs are potential materials for the separation of C₂ hydrocarbons.

Acknowledgements

This work was supported by 973 program (2012CB821705 and 2011CB932504), and NSFC (21425102, 21403235, 21221001).

Notes and references

1. R. W. Baker, Ind. Eng. Chem. Res., 2002, 41, 1393–1411.
2. (a) E. D. Bloch, W. L. Queen, R. Krishna, J. M. Zadrozny, C. M. Brown and J. R. Long, Science, 2012, 335, 1606–1610; (b) C. Y. Lee, Y.-S. Bae, N. C. Jeong, O. K. Farha, A. A. Sarjeant, C. L. Stern, P. Nickias, R. Q. Snurr, J. T. Hupp and S. T. Nguyen, J. Am. Chem. Soc., 2011, 133, 5228–5231; (c) Y. He, R. Krishna and B. Chen, Energy Environ. Sci., 2012, 5, 9107–9120; (d) R. B. Getman, Y.-S. Bae, C. E. Wilmer and R. Q. Snurr, Chem. Rev., 2012, 112, 703–723; (e) Y.-P. He, Y.-X. Tan and J. Zhang, Chem. Commun., 2013, 49, 11323–11325.
3. (a) S. J. Geier, J. A. Mason, E. D. Bloch, W. L. Queen, M. R. Hudson, C. M. Brown and J. R. Long, Chem. Sci., 2013, 4, 2054–2061; (b) Y. S. Bae, C. Y. Lee, K. C. Kim, O. K. Farha, P. Nickias, J. T. Hupp, S. T. Nguyen and R. Q. Snurr, Angew. Chem., Int. Ed., 2012, 51, 1893–1896; (c) B. Li, Y. Zhang, R. Krishna, K. Yao, Y. Han, Z. Wu, D. Ma, Z. Shi, T. Pham, B. Space, J. Liu, P. K. Thallapally, J. Liu, M. Chrzanowski and S. Ma, J. Am. Chem. Soc., 2014, 136, 8654–8660; (d) Y.-X. Tan, Y. Zhang, Y.-P. He, Y.-J. Zheng and J. Zhang, Inorg. Chem., 2014, 53, 12973–12976; (e) Y.-X. Tan, Y.-P. He and J. Zhang, Chem. Commun., 2011, 47, 10647–10649.
4. (a) Y. He, Z. Zhang, S. Xiang, F. R. Fronczek, R. Krishna and B. Chen, Chem.–Eur. J., 2012, 18, 613–619; (b) Y. He, Z. Zhang, S. Xiang, H. Wu, F. R. Fronczek, W. Zhou, R. Krishna, M. O'Keeffe and B. Chen, Chem.–Eur. J., 2012, 18, 1901–1904.
5. Y.-X. Tan, Y.-P. He, M. Wang and J. Zhang, RSC Adv., 2014, 4, 1480–1483.
6. T. K. Maji, R. Matsuda and S. Kitagawa, Nat. Mater., 2007, 6, 142–148.
7. Y.-X. Tan, Y.-P. He and J. Zhang, ChemSusChem, 2012, 5, 1597–1601.
8. (a) Z.-Q. Jiang, G.-Y. Jiang, F. Wang, Z. Zhao and J. Zhang, Chem.–Eur. J., 2012, 18, 10525–10529; (b) H. Li, W. Shi, K. Zhao, Z. Niu, H. Li and P. Cheng, Chem.–Eur. J., 2013, 19, 3358–3365; (c) S. Wang, Q. Yang, J. Zhang, X. Zhang, C. Zhao, L. Jiang and C.-Y. Su, Inorg. Chem., 2013, 52, 4198–4204; (d) P. Kanoo, R. Matsuda, R. Kitaura, S. Kitagawa and T. K. Maji, Inorg. Chem., 2012, 51, 9141–9143; (e) T. K. Maji, G. Mostafa, R. Matsuda and S. Kitagawa, J. Am. Chem. Soc., 2005, 127, 17152–17153.
9. S. Xiong, Y. He, R. Krishna, B. Chen and Z. Wang, Cryst. Growth Des., 2013, 13, 2670–2674.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, addition figures, TGA, powder X-ray patterns, adsorption isotherms and CIF file (CCDC 1029218 and 1029219). For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c4ra13058h

---