Dataset for "Intrinsic flexibility of the EMT zeolite framework under pressure"

Data collection method:

The high pressure X-ray diffraction data within this archive was collected on the ID15B and ID27 beamlines at the European Synchrotron Radiation Facility (ESRF) in Grenoble, France. The data collection process was as follows: The samples were first loaded into diamond-anvil cells (DACs), suspending in a non-penetrating pressure-transmitting medium (PTM) alongside a ruby chip. On the ID15B beamline, the PTM used was Daphne 7373 oil, the incident X-ray radiation was of wavelength 0.4113 A (Angstroms), and calibrations were performed using silicon. As for the ID27 beamline, the PTM used was silicone oil, the incident X-ray radiation of wavelength 0.3738 A, and calibrations performed with cerium dioxide (CeO2). The DAC pressure was increased in gradual steps, with three 2D diffraction images taken at each pressure step. The pressure was recorded by exciting the ruby chip with a laser and determining the shift of the R1 emission line. The pressure was recorded before and after each pressure point, with an average calculated. The sample was compressed until pressure-induced amorphisation was seen to be imminent. Following this the samples were decompressed to ambient conditions, with several diffraction images taken during the this cycle. Concerning the computational modelling, density-functional theory (DFT) calculations were performed on the empty FAU and EMT framework structures using the plane-wave pseudopotential code VASP. The PBEsol generalised-gradient approximation functional, including the semi-empirical DFT-D3 dispersion correction (PBEsol+D3) were used to determine the quantum-mechanical exchange and correlation. Plane-wave basis augmented-wave (PAW) pseudopotentials of the O 2s/2p and Si 3s/3p electrons in the valence shells were used to represent the electronic structures of the relevant atoms. The PAW projection was performed in real space, and the structures fully optimised. The influence of solvent molecules occupying the framework cages was calculated using the implicit-solvent VASPsol model. Bulk moduli was obtained by compressing/expanding the optimised structures by 1% volume increments, and re-optimising at fixed volumes before fitting the energy/volume curves to the Birch-Murnaghan equations of state. Lattice-dynamics calculations were performed on the optimised structures using the Phonopy package, with a finite-displacement step size of 10^-2 A. To calculate the phonon density of states (DoS) curves, the phonon frequencies were interpolated onto a regular Γ-centered grid of q-points, with 24x24x24 subdivisions. The simulated infrared spectra were calculated using the density-functional perturbation theory (DEPT) routines in VASP, within the open-source SpectroscoPy package. Calculations of the phonon frequencies, elastic constants and Born changes, the PAW projection was applied in reciprocal space.

Data processing and preparation activities:

Before analysis of the high-pressure diffraction data, the 2D diffraction images were processed in the following way. The three images taken at each pressure point were averaged using the FIT2D software, producing an average image. The area of the 2D image was subsequently integrated over in the Dioptas software, to produce a 1D diffraction pattern. This produced the .xy files found in this archive. The .xy files were used in the TOPAS Academic software using Pawley refinements in order to calculate the unit cell parameters at each pressure point. First the refinement at ambient conditions was performed, with subsequent refinements performed using the Batch mode. In this mode, the refinement process is iterative, meaning the input structure for each pressure point is the output from the automated refinement from the previous pressure point. For both the filled and empty zeolite EMC-2 samples the P63/mmc space group was used in the Pawley refinements. The bulk moduli were determined using the PASCal webtool, using data within the 0-2.2 GPa range. The data was fit to the 2nd order Birch-Murnaghan equation of state, weighted using the estimated 0.1 GPa pressure error within the DACs.

Additional information:

The relevant article and dataset include the flexibility window of the EMT framework (filled and empty) simulated using the GASP software. These geometric simulations have been performed previously, and are referenced in "Methodology link" below.

Methodology link: