An ultra-tunable platform for molecular engineering of high-performance crystalline porous materials

Metal-organic frameworks are a class of crystalline porous materials with potential applications in catalysis, gas separation and storage, and so on. Of great importance is the development of innovative synthetic strategies to optimize porosity, composition and functionality to target specific applications. Here we show a platform for the development of metal-organic materials and control of their gas sorption properties. This platform can accommodate a large variety of organic ligands and homo- or hetero-metallic clusters, which allows for extraordinary tunability in gas sorption properties. Even without any strong binding sites, most members of this platform exhibit high gas uptake capacity. The high capacity is accomplished with an isosteric heat of adsorption as low as 20 kJ mol−1 for carbon dioxide, which could bring a distinct economic advantage because of the significantly reduced energy consumption for activation and regeneration of adsorbents.

days. Pure green micro-crystals were obtained after cooling to room temperature. Pure sample was obtained by filtering and washing the raw product with DMF. The yield was about 55% based on Ni.

Synthesis of [Ni3(OH)(BDC)3(TPT)] (CPM-33a, Ni3-BDC) and [Ni3(OH)(DHBDC)3(TPT)]
(CPM-33b, Ni3-DHBDC). CPM-33a and 33b were prepared according to our former reported procedures. X-ray analysis, which was performed on a Bruker Smart APEX II CCD area diffractometer with nitrogen-flow temperature controller using graphite-monochromated MoKα radiation (λ = 0.71073 Å), operating in the ω and φ scan mode. The SADABS program was used for absorption correction. The structure was solved by direct methods and refined using SHELXTL 15 . All non-hydrogen atoms in the framework were refined with anisotropic displacement parameters. The Mg/V ratio was estimated from the occupancy refinement with single crystal X-ray diffraction data and further supported by the EDX analysis. The large volume fractions of solvents in the lattice pores could not be modelled in terms of atomic sites and were treated using the SQUEEZE routine in the PLATON software package 16 . Crystal data as well as details of data collection and refinements were summarized in Tables S9-S15.

Synthesis of [(CH3)2NH2][Zn3(OH)(OHBDC)3(TPT)] (CPM
Powder X-ray diffraction. Powder X-ray diffraction (XRD) data were collected on a Bruker D8  where R is the universal gas constant, q is the amount of CO2 loaded at pressure p and temperature T. These calculations are done through the "Heat of Adsorption" calculated function embedded in the software supplied by Micromeritics ASAP 2020M surface-area and pore-size analyzer machine.
Selectivity prediction for binary mixture adsorption. Ideal adsorbed solution theory (IAST) was used to predict binary mixture adsorption 17,18 from the experimental pure-gas isotherms.
To perform the integrations required by IAST, the single-component isotherms should be fitted by a proper model. There is no restriction on the choice of the model to fit the adsorption isotherms, however, data over the pressure range under study should be fitted precisely 19,20 . Several isotherm models were tested to fit the experimental pure isotherms for CH4 and CO2, and the Langmuir-Freundlich equation was found to be the best fit to the experimental data: Here, q is the adsorbed amount per mass of adsorbent (mol/kg), x is the pressure of the bulk gas at equilibrium with the adsorbed phase (kPa), b is the affinity coefficients of the sites interactions. Norm-conserving pseudopotentials were employed to account for the effects of core electrons. The unit cell configuration determined by XRD was used as the initial structure for the simulations. Hydrogen atoms that cannot be determined in XRD are added according to the nominal bond length and angle. Some of the sites have partial occupancy, and to account for this properly a supercell calculation would be desirable, but too costly in practice. Instead, a single unit cell was used and the partially occupied sites were modified to be either occupied or unoccupied so that the overall probability of being occupied is proportional to the actual occupancy. The atomic coordinates were relaxed with no imposed symmetry to allow minimization of the potential energy and the interatomic forces. The energy tolerance for the electronic structure calculations was 5 x 10 -7 eV, and the energy tolerance for ionic relaxation was 5 x 10 -6 eV. The tolerance for the interatomic forces was 0.01 eV/Å. After convergence was reached, the dynamical matrix was obtained using the linear response method, from which the phonon frequencies and vibrational modes were calculated. The electronic structure calculations and the phonon calculations were performed at the gamma-point only. The aClimax software 22 was used to convert the DFT-calculated phonon results to the simulated INS spectra.