ABSTRACT
Previous high-pressure experiments have shown that pressure-transmitting fluids composed of small molecules can be forced inside the pores of metal organic framework materials, where they can cause phase transitions and amorphization and can even induce porosity in conventionally nonporous materials. Here we report a combined high-pressure diffraction and computational study of the structural response to methanol uptake at high pressure on a scandium terephthalate MOF (Sc2BDC3, BDC = 1,4-benzenedicarboxylate) and its nitro-functionalized derivative (Sc2(NO2-BDC)3) and compare it to direct compression behavior in a nonpenetrative hydrostatic fluid, Fluorinert-77. In Fluorinert-77, Sc2BDC3 displays amorphization above 0.1 GPa, reversible upon pressure release, whereas Sc2(NO2-BDC)3 undergoes a phase transition (C2/c to Fdd2) to a denser but topologically identical polymorph. In the presence of methanol, the reversible amorphization of Sc2BDC3 and the displacive phase transition of the nitro-form are completely inhibited (at least up to 3 GPa). Upon uptake of methanol on Sc2BDC3, the methanol molecules are found by diffraction to occupy two sites, with preferential relative filling of one site compared to the other: grand canonical Monte Carlo simulations support these experimental observations, and molecular dynamics simulations reveal the likely orientations of the methanol molecules, which are controlled at least in part by H-bonding interactions between guests. As well as revealing the atomistic origin of the stabilization of these MOFs against nonpenetrative hydrostatic fluids at high pressure, this study demonstrates a novel high-pressure approach to study adsorption within a porous framework as a function of increasing guest content, and so to determine the most energetically favorable adsorption sites.
ABSTRACT
Thermal desorption spectra of a number of metal-organic frameworks were studied using grand canonical Monte Carlo simulation. Our simulation results are in qualitative agreement with experimental results but also show that great care must be taken when choosing the force field to describe the hydrogen/framework interaction. As the simulations additionally yield the positions and potential energies of the adsorbed molecules it is straightforward to assign the peaks and features in the thermal desorption spectra to specific adsorption sites. We show that the location of the peaks is directly related to the hydrogen-framework interaction which is a complex function of the chemical and topological environment of the pore space, the pore size and the presence of specific interaction sites such as open metal sites. Finally, we demonstrate that an IRMOF-8 sample used to obtain an experimental thermal desorption spectrum must have indeed been catenated as previously suspected. Overall, molecular simulation is a useful tool to complement the interpretation of experimental thermal desorption spectra.
ABSTRACT
The crystal structure of the small pore scandium terephthalate Sc(2)(O(2)CC(6)H(4)CO(2))(3) (hereafter Sc(2)BDC(3), BDC = 1,4-benzenedicarboxylate) has been investigated as a function of temperature and of functionalization, and its performance as an adsorbent for CO(2) has been examined. The structure of Sc(2)BDC(3) has been followed in vacuo over the temperature range 140 to 523 K by high resolution synchrotron X-ray powder diffraction, revealing a phase change at 225 K from monoclinic C2/c (low temperature) to Fddd (high temperature). The orthorhombic form shows negative thermal expansivity of 2.4 × 10(-5) K(-1): Rietveld analysis shows that this results largely from a decrease in the c axis, which is caused by carboxylate group rotation. (2)H wide-line and MAS NMR of deuterated Sc(2)BDC(3) indicates reorientation of phenyl groups via π flips at temperatures above 298 K. The same framework solid has also been prepared using monofunctionalized terephthalate linkers containing -NH(2) and -NO(2) groups. The structure of Sc(2)(NH(2)-BDC)(3) has been determined by Rietveld analysis of synchrotron powder diffraction at 100 and 298 K and found to be orthorhombic at both temperatures, whereas the structure of Sc(2)(NO(2)-BDC)(3) has been determined by single crystal diffraction at 298 K and Rietveld analysis of synchrotron powder diffraction at 100, 298, 373, and 473 K and is found to be monoclinic at all temperatures. Partial ordering of functional groups is observed in each structure. CO(2) adsorption at 196 and 273 K indicates that whereas Sc(2)BDC(3) has the largest capacity, Sc(2)(NH(2)-BDC)(3) shows the highest uptake at low partial pressure because of strong -NH(2)···CO(2) interactions. Remarkably, Sc(2)(NO(2)-BDC)(3) adsorbs 2.6 mmol CO(2) g(-1) at 196 K (P/P(0) = 0.5), suggesting that the -NO(2) groups are able to rotate to allow CO(2) molecules to diffuse along the narrow channels.
