Communication pubs.acs.org/IC
A Clear Insight into the Distinguishing CO2 Capture by Two Isostructural DyIII−Carboxylate Coordination Frameworks Xi Wang, Min Chen, and Miao Du* Henan Provincial Key Lab of Surface & Interface Science, Zhengzhou University of Light Industry, Zhengzhou 450002, People’s Republic of China S Supporting Information *
respectively (see Scheme 1). Remarkably, 477-MOF featuring an active triazinyl moiety shows superior capacity for CO2
ABSTRACT: Two isostructural the-type DyIII coordination networks were successfully constructed based on a pair of analogous tribenzoate bridging ligands with phenyl and triazinyl central spacers. Notably, the active triazinyl group can obviously enhance the capability and selectivity of CO2 sorption for the porous framework.
Scheme 1. Illustration of the [Dy2(HCOO)] SBU and the Interacting 8-fold Tecton (indicated by different colors) via BTB or TATB Ligands
The crystalline porous materials, namely metal−organic frameworks (MOFs), have attracted widespread interest for their fascinating structures and potential applications in many realms.1−5 In particular, the sorption performance is primary to MOF materials for their highly porous inherence, and in this context, the study on fuel storage,6 CO2 capture,7 hydrocarbon separation,8 et cetera has made great progress. As we all know, the properties of crystalline materials essentially rely on their intrinsic compositions and structures. Thus, the active domains within the host frameworks of MOF materials (e.g., open metal sites and exposed heteroatoms) will be critical to their gas sorption behaviors. Normally, the open metal sites are generated by the removal of coordination solvents, and the exposed heteroatoms can be achieved by introducing the functional substituents on ligands. However, in most cases, the exclusion of solvent ligands around metal centers will lead to the collapse of porous frameworks.7,9 Also, the sorption properties can be modified by adding the active groups on ligands,10 while the shrinkage of available space is inevitable under such circumstances, which in some respects is conflicting for the higher adsorbing capacity. In this connection, the modification of ligand backbones by heteroatom replacement can be a promising approach to improve the sorption properties of MOFs, where the inert phenyl or cyclopentadienyl groups are replaced with heterocycles as the active region.11−13 Significantly, the sorption investigation on such isostructural MOFs with almost identical voids but different active sites can offer a nice platform for profound understanding of the interactions between host frameworks and gas molecules, and the sorption mechanisms thereof. Unfortunately, just a tiny variation of ligands will lead to the formation of distinct crystalline products for MOFs,8,11,14 and as a result, only two sets of isostructural CuII-based MOFs have been constructed with this strategy thus far.12,13 Herein, we present two isostructural DyIII MOFs based on two similar trigonal building blocks, benzene-1,3,5-tribenzoate (BTB) and 4,4′,4″-s-triazine-2,4,6-triyltribenzoate (TATB), which are differentiating only in the central phenyl and triazinyl spacers, © XXXX American Chemical Society
capture and adsorption selectivity over other gases, in comparison to the phenyl-based 476-MOF. Colorless cubic crystals for {[Dy 2 L 8 / 3 (HCOO)](Me2NH2)3(DMF)7(CH3OH)7(H2O)7}n (L = BTB for 476MOF and TATB for 477-MOF) were similarly synthesized by solvothermal reactions of Dy(NO3)3 with the trigonal tectons in DMF-CH3OH solvent (see Supporting Information for details). Single crystal X-ray diffraction reveals that 476-MOF and 477-MOF are isostructural, crystallizing in the cubic Im3̅ space group (see Table S1). The dimeric [Dy2(HCOO)] secondary building unit (SBU) is afforded (see Scheme 1), in which the HCOO− bridge is disordered to adapt the high Th symmetry. Such a structural pattern is unprecedented according to a latest search of the Cambridge Structural Database (CSD). Each DyIII ion takes the 9-coordinated monocapped square prism geometry (Figure S2), completed by eight oxygen atoms from four chelating carboxylates and one oxygen atom from the bridging formate ion. All carboxylates of the organic ligands take the chelating mode with DyIII center to extend the [Dy2(HCOO)] SBUs into a 3D porous framework (see Figure S3). Furthermore, two such networks are entangled into each other to form a 2-fold interpenetrating architecture (see Figure S4). In each network, two types of cage-like voids are observed, where the mesoporous cage with a diameter of 2.3 nm is constituted by eight trigonal ligands and 12 DyIII centers (Figure 1a), while the microporous cage of 1.6 nm in diameter is constructed from eight ligands and six [Dy2(HCOO)] SBUs Received: March 27, 2016
A
DOI: 10.1021/acs.inorgchem.6b00760 Inorg. Chem. XXXX, XXX, XXX−XXX
Communication
Inorganic Chemistry
Figure 1. Crystal structures of 476-MOF/477-MOF. (a) Mesoporous and (b) nanoporous cage. (c) 2-fold interpenetrating motif formed by the inset of a nanoscale cage into a mesoscale cage. (d) Augmented version of the 2-fold interpenetrating the-type framework.
(Figure 1b). Further, each microporous cage is embedded into a mesoporous section from the other set of independent network, which will reduce the voids of the overall 3D framework (Figures 1c and S4). Strong aromatic interactions exist between the parallel central aromatic rings coming from two distinct networks, with the centroid-to-centroid distances of ca. 3.37 Å (476-MOF) and 3.41 Å (477-MOF), which will also contribute to the stability of interpenetrating patterns (Figure S5). The PLATON15 analysis reveals that the effective free volume is 63.1%/62.6% per unit cell in 476-MOF/477MOF, without considering the Me2NH2+ cations and solvent guests. Topologically, each ligand takes the 3-connected fashion to interact with three DyIII centers, while each DyIII can be viewed as a 5-connected node surrounded by four ligands and one formate anion. Therefore, the structure for 476-MOF or 477-MOF can be properly ascribed to a 2-fold interpenetrating binodal (3,5)-connected network with a point symbol of (48.64.812.104) (Figure S6). Alternatively, when [Dy2(HCOO)] is simplified as the node, linked by eight surrounding trigonal ligands, such 3D structures can be reduced to the binodal (3,8)connected the-type networks with 2-fold interpenetration (see Figure 1d). Similar reticular topology is also observed in TATBbased PCN-9 and PCN-17 MOFs, in which the two entangled networks are combined by bridging sulfate anions.11 The thermogravimetric analysis (TGA) curves of 476-MOF and 477-MOF (Figure S7) indicate that the first weight loss from 20 to 270 °C corresponds to a loss of solvent guests. The subsequent weight loss occurs in the 330−390 °C range, which can be ascribed to the exclusion of dimethylammonium cations. Beyond that temperature, pyrolysis of the host framework is observed. To evaluate the porosity for 476-MOF and 477MOF, gas adsorption experiments were carried out by using the activated materials at different temperatures (see Supporting Information for details). 476-MOF and 477-MOF display similar type I isotherms (see Figure 2) with saturated uptakes of 266 and 274 cm3 g−1 for N2 at 77 K and 1 atm. Fitting the Brunauer−Emmett−Teller (BET) equation to N2 adsorption isotherms gives the estimated surface areas of 898 and 902 m2 g−1 for 476-MOF and 477-MOF, respectively, which are higher than those for the reported TATB-based PCN-17 series (606− 820 m2 g−1).11 The H2 adsorption isotherm for 476-MOF
Figure 2. Gas adsorption isotherms for 476-MOF and 477-MOF at 77 and 273 K (closed/open shapes: adsorption/desorption).
indicates a moderate uptake of 1.60 wt % (179 cm3 g−1) at 77 K and 1 atm, while 477-MOF shows a slightly higher uptake of 1.77 wt % (198 cm3 g−1) under the same conditions. Also, the maximum O2 uptake is 302 cm3 g−1 (43.1 wt %) for 476-MOF, and a higher uptake of 330 cm3 g−1 (47.1 wt %) is for 477MOF. The gas sorption isotherms for 476-MOF and 477-MOF at 273 and 293 K (Figures 2 and S8) indicate their extremely poor capture capacities for N2, O2, and H2 (