Atomistic Mechanisms of Thermal Transformation in a Zr-Metal Organic Framework, MIL-140C

Summary¶

Thermal collapse of the Zr-based metal–organic framework MIL-140C is studied with ReaxFF molecular dynamics alongside transmission electron microscopy (TEM) of decomposed material, connecting atomistic pyrolysis chemistry to nanoscale oxide cluster morphologies. A Zr/C/H/O ReaxFF description is assembled from prior zirconia (YSZ-related) Zr/O/H training and glycine-derived C/H/O chemistry, then reoptimized against DFT data on Zr–oxo cluster models with truncated carboxylate linkers (formate/benzoate/BDC-like motifs), including charges, geometries, and reaction energies. The simulations follow heating-induced degradation, propose molecular pathways for carbon monoxide as a dominant gaseous product, and characterize residual ZrO\(_x\)-like clusters within disrupted organic matter.

Methods¶

1 — MD application (MIL-140C). Reactive molecular dynamics in LAMMPS on 3D PBC MIL-140C supercells (order 10⁴ atoms; cell and heating schedule in JPCL) uses NVT thermostat control, sub-fs timestep, and ps–ns equilibration/ramp/production to follow linker scission and cluster evolution at elevated temperature (K). Barostat / isotropic pressure coupling: N/A in the summarized constant-volume pyrolysis runs. N/A — external uniform electric field; N/A — umbrella / metadynamics in the standard runs.

2 — Force-field training. Zr/C/H/O ReaxFF merges parent Zr/O/H data (YSZ-related equation-of-state and surface energies for ZrO\(_2\), Mulliken charges in the source) with C/H/O organics, then reoptimizes against M06-2X/6-31G(d,p)+LANL2DZ DFT on Zr\(_6\)O\(_4\)(OH)\(_4\) cluster geometries with truncated linker models (formate/BDC-type). One-parameter parabolic extrapolation in the ReaxFF optimizer implements the fit; TEM of decomposed samples provides experimental validation of nanoscale morphology trends after heating comparable to simulation.

3 — Experiments (TEM). TEM on thermally treated MIL-140C is compared to MD snapshots for cluster size and texture (not a full in situ correlative benchmark of every trajectory).

Findings¶

Simulations and experiment both indicate very small ZrO\(_x\) clusters embedded in disrupted organic regions, consistent with high catalytic activity reported for decomposed MOFs in the discussion. The mechanism analysis highlights sequential loss of cross-linkers before vertically stacked linkers, CO-producing pathways, and oxidation states of evolved ZrO\(_x\) approaching tetragonal ZrO\(_2\)-like behavior in the model under aggressive heating.

Comparisons and sensitivity. TEM and MD are compared at the morphology level; thermal ramp rate and hold temperature are the main levers for which reaction channels are sampled. Corpus / PDF has full protocols.

Limitations¶

Cluster models and truncated linkers in the training set approximate full MIL-140C complexity; long-time mesoscale porosity collapse and gas transport are only partially accessible in nanosecond MD.

Relevance to group¶

Group co-authorship on reactive MD of MOF thermal chemistry with explicit ReaxFF parameterization to Zr-organic systems. Training clusters include Gaussian DFT single-point energies and relaxed geometries for Zr–oxo nodes with truncated linkers; reproduce those clusters before attempting parameter merges with unrelated Zr/O sets. Thermal ramp rates and total simulation times for MIL-140C degradation appear in the JPCL article and govern which reaction channels are sampled within nanosecond MD trajectories.

reaxff-family

Cross-link MOF thermal decomposition questions to [[theme-pyrolysis-combustion-organics]] and experimental TEM SI in the JPCL article when users need mesoscale morphology beyond atomistic snapshots from MD.