Flexible (Breathing) Interpenetrated Metal-Organic Frameworks for CO 2 Separation Applications Praveen K. Thallapally,* ,† Jian Tian, ‡ Motkuri Radha Kishan, † Carlos A. Fernandez, † Scott J. Dalgarno, § Peter B. McGrail, † John E. Warren, | and Jerry L. Atwood ‡ Energy and EnVironment Directorate, Paciﬁc Northwest National Laboratory, Richland, Washington 99352, Department of Chemistry, UniVersity of Missouri-Columbia, Columbia, Missouri 65211, Department of Chemistry, School of Engineering and Physical Sciences-Chemistry, Heriot-Watt UniVersity, Edinburgh, U.K., and Daresbury Science & InnoVation Campus, Warrington, Cheshire, U.K. Received August 19, 2008; E-mail: Praveen.Thallapally@pnl.gov Carbon dioxide capture and separation is a major road block for the safe storage of carbon into deep geological formations. 1 The major source of CO 2 emissions are from coal ﬁred power plants where only 30% of CO 2 is captured, resulting in increased CO 2 emissions in the atmosphere. Conventional methods such as alkyl amine solutions are used to scrub the CO 2 from exhaust gas streams, which is costly and inefﬁcient. The major drawbacks of the existing technology include low CO 2 wt%, degradation of the solvent, and the high temperature required to regenerate the adsorbed gas. 2 Therefore alternative concepts based on chemical adsorption of CO 2 onto metal oxides and physical adsorption on activated carbons, silicas, zeolites, and nonporous calixarenes 3,4 were proposed, but microporous metal-organic frameworks (MOFs) have received considerable attention during the past couple of years. This may be due to the high mass ﬂux, thermal stability, adjustable chemical functionalities, extra high porosity, and availability of hundreds of well characterized materials reminiscent to zeolites. 5 Several research groups, including Yaghi, 6 Kitagawa, 7 Ferey, 8 Rosseinsky, 9 Zaworotko, 10 Hupp, and others, have developed a number of MOFs for gas storage and separation applications. 11-16 For example, Yaghi and co-workers have shown that MOF-177 exhibits a CO 2 sorption capacity of 1.4 g of CO 2 per gram of sorbent material. 17 This is a signiﬁcant improvement over commercially available zeolites sorbents. However, most MOFs have been deliberately designed with very large pore sizes or channels to achieve maximum loadings. While effective, the penalty is again lack of selectivity, which is required for gas separation applications. To approach this problem, we have designed and developed coordination solids based on a tetrahedral organic linker, tetrakis[4-(carboxyphenyl)oxam- ethyl]methane, 1 as a building block to generate MOFs with metal salts and organic pillars. Self-assembly of 1 with zinc nitrate hexahydrate and bipyridine in dimethyl formamide resulted in complex 3 (Supporting Informa- tion, Scheme S1). Crystallographic measurments on complex 3 conﬁrm the tetrahedral ligand was connected to two zinc atoms in a paddle-wheel fashion. The paddle wheels are further pillared by 4,4′-bipyridine molecules occupying the axial sites of the Zn 2 paddle wheels to form a three-dimensional (3D) structure (Figure 1). The overall structure of 3 is a pair of identical PtS nets of 3, which are mutually interpenetrated with each other to form doubly interpen- etrated frameworks (Scheme S1). The pores in 3 are partially ﬁlled with 4,4′-bipyridine molecules that are connected to paddle-wheel SBUs of the framework while the remaining channels are ﬁlled with solvent DMF and water molecules (SI, Figures S2-S4). Thermogravimetric analysis of the complex 3 shows 25 to 30% of weight loss between room temperature and 250 °C, which corre- sponds to the loss of DMF and water molecules (Figure S5). Prior to the sorption studies, the surface area of the activated sample was calculated using the N 2 adsorption at 77 K, which exhibits a typical type I isotherm with 1150 m 2 g -1 of surface area (Figure S6-S7). The surface area and the solvent accessible void space in this sample are signiﬁcantly reduced when compared to other open metal-organic frameworks due to the interpenetration of the 3D framework in the solid state. Gas sorption experiments using CO 2 were performed using an HPVA-100 volumetric device at room temperature. Sample 3 (∼80 mg) was placed in a sample chamber and activated at high temperature under vaccum for several hours. For low pressure experiments, 0.1 bar of CO 2 was dosed into the sample chamber every 10 min, and the volume adsorbed per gram of material was plotted against the pressure (Figure 2). At this pressure (1 bar) the calculated weight percentage was found to be close to 5 wt%, which is comparable to the other MOFs reported earlier. The absence of hysteresis during desorption of CO 2 is not surprising, and it has been found to be very common for materials with a pore size of 20 Å or less. The measurements were repeated several times by evacuating the sample and pressurizing with CO 2 again. The same weight percentage was obtained within (5%. The CO 2 sorption plot at 1 bar suggests that sample 3 does not reach a saturation point; therefore high pressure experiments with CO 2 were conducted at ambient conditions. Figure 2 shows the absorption isotherm of CO 2 at high pressures that indicate a type I relationship with a step in the absorption at ∼10 bar. Such a step during the absorption of gases in organics and MOFs is not common though it has been observed in similar systems before. 18,19 A number of mechanisms † Paciﬁc Northwest National Laboratory. ‡ University of MissourisColumbia. § Heriot-Watt University. | Daresbury Science & Innovation Campus. Figure 1. Representation of PtS network from 1 and Zn 2 paddlewheel connected by bipyridine molecules. Published on Web 11/24/2008 10.1021/ja806391k CCC: $40.75  2008 American Chemical Society 16842 9 J. AM. CHEM. SOC. 2008, 130, 16842–16843