Electron diffraction of pn–CO2–Mg2(dobpdc): 3D ED and 4D-STEM Raw Data and Structure Refinement Files

3D ED This repository contains 3D electron diffraction (3D ED) data acquired using a Ceta-S CMOS detector on a Thermo Fisher Scientific Themis TEM operating at 300 kV. Data of the CO₂-adsorbed phase were collected at room temperature after waiting > 1 h after sample loading into the TEM column; thermogravimetric analysis suggested these conditions were largely sufficient to desorb any weakly bound species (such as water). To access the activated phase, the sample was heated at 100 °C in the TEM column and left to equilibrate for > 1 h. The heating was then turned off and the sample was left to cool back to room temperature overnight. Data of the in situ activated phase were collected at room temperature under the same imaging conditions as the CO₂-adsorbed phase. Experimental parameters Parameter Value Accelerating Voltage 300 kV Selected area aperture 1.3 µm Exposure time 0.25 s Angular speed 2 degrees/s Temperature at data collection RT One directory contains raw data, in the form of .mrc files, from this in situ session. Datasets 2–4 were collected before activation, corresponding to the CO₂-adsorbed phase, and Datasets 5–8 were collected after heating, corresponding to the activated phase. Indexing, integration, scaling, and merging were performed using XDS and XSCALE. Although several nanocrystals were sampled, none diffracted cleanly to subangstrom resolution. Nevertheless, a single 74.9% complete movie (Dataset 1), which was from a separate data collection session of pn–CO₂–Mg₂(dobpdc) at higher resolution but with otherwise identical collection parameters to the in situ session, was sufficient to generate a viable ab initio solution at 1.0 Å resolution using SHELXT. This second directory contains the raw data in the form of an .mrc file, and separate sub-directories corresponding to a structure refined without restraints and a structure refined with geometric restraints. Each sub-directory contains a CIF and a RES file. Model-building into the ab initio electrostatic potential map and subsequent least-squares refinement were conducted in Olex2 using neutral electron scattering factors parameterized into four Gaussian functions. Given the lack of subangstrom resolution, all ADPs were refined isotropically, and precision in bond lengths was limited. Initial refinement of the atomic model was conducted without any geometric constraints or restraints to avoid injecting preconceived bias (overfitting). However, electrostatic potential corresponding to a single carbon atom in the 4,4′-dioxidobiphenyl-3,3′-dicarboxylate linker was visible only at a relatively low contour level of +1.5σ in the 2Fobs–Fcalc map (possibly due to lack of completeness). All other non-hydrogen atoms were visible at +3σ. See Fig1.jpg below for the image. Figure 1. Ab initio electrostatic potential map (2Fobs–Fcalc) contoured at a.) +1.5σ and b.) +3σ, with the refined atomic model superimposed. Carbon atoms are rendered in gray, oxygen atoms in red, magnesium atoms in blue, and hydrogen atoms in white. The blue arrow indicates the problematic carbon atom. This led to distorted phenyl rings when refined freely. Thus, standard geometric restraints (DFIX, FLAT, SADI) were applied to the carbon atoms in the phenyl ring, and hydrogen atoms were assigned using the riding model. All other atoms were allowed to refine freely. This led to a chemically reasonable model, albeit with an expectedly high B-factor on the problematic carbon atom. The SQUEEZE algorithm was used to account for residual electrostatic potential within the pores of the MOF. We include both unrestrained and restrained refinement results here for completeness. Refinement R-factors converged to values comparable to published 3D ED structures of MOFs. Initial refinement (no restraints) Final refinement (with restraints and H atoms) Empirical formula C42Mg6O18 C42Mg6O18H18 Temperature (K) 295 295 Formula weight (g mol-1) 938.27 956.42 Crystal system Trigonal Trigonal Space group* P3121 (152) P3121 (152) Point group 32 32 Laue symmetry 3m 3m Unit cell lengths a, b, c (Å) 20.9900(16), 20.9900(16), 7.000(3) 20.9900(16), 20.9900(16), 7.000(3) Unit cell angles α, β, γ (°) 90, 90, 120 90, 90, 120 Unit cell volume (Å3) 2670.9(12) 2670.9(12) Z 1 1 Wavelength (Å) 0.0197 0.0197 Accelerating voltage (kV) 300 300 Resolution (Å) 1.0 1.0 CC1/2 99.2 99.2 Rint 0.143 0.143 I/σ(I) 3.9 3.9 Completeness (%) 74.9 74.9 Number of reflections (observed) 2140 2140 Number of reflections (unique) 1128 1128 Data/restraints/parameters 1128/0/46 1128/22/46 Miller index ranges -14 ≤ h ≤ 14, -20 ≤ k ≤ 20, -6 ≤ l ≤ 6 -14 ≤ h ≤ 14, -20 ≤ k ≤ 20, -6 ≤ l ≤ 6 R1 (wR2) 11.52 (36.71) 16.26 (45.88) GooF 0.978 1.266 *No attempt was made to assign the absolute enantiomorph. Space group P3121 was selected over its enantiomorph P3221 arbitrarily. Isomorphous X-ray structures of Zn derivatives of the same MOF have been reported in both space groups P3121and P32211–3. 4D-STEM This repository also contains raw data from 4D-STEM experiments, which were carried out on the double-aberration-corrected TEAM 0.5 instrument, a modified FEI Titan TEM with a Gatan OneView camera and 4D Camera based on the Gatan K3 platform operating at 87000 fps. A 70 μm condenser aperture was used to produce a beam convergence semi-angle of 0.57 mrad (FWHM = 1.78 nm). For pn–CO2–Mg2(dobpdc) and Mg2(dobpdc), a probe step size of 1.56 nm and camera length of 144 and 180 mm, respectively, was used. For pn–Mg2(dobpdc) and CO2–Mg2(dobpdc), a probe step size of 2.24 nm and camera length of 180 mm was used. The probe current was approximately 100 pA.