Skip to content

A reactive force field molecular dynamics study on the inception mechanism of titanium tetraisopropoxide (TTIP) conversion to titanium clusters

Summary

ReaxFF reactive MD with a newly developed Ti/C/H/O potential maps titanium tetraisopropoxide (TTIP) droplet conversion to Ti-bearing clusters in 1000–2500 K with optional O₂, motivated by flame and aerosol synthesis routes where organometallic precursors first pyrolyze in gas phase before nucleating oxide nanoparticles. The study identifies intermediate Ti species and early decomposition pathways, and examines how temperature, O₂ concentration, and high-temperature residence time shift cluster inception and stoichiometry. Highlights include non-monotonic temperature effects (weaker Ti–O bonds at high \(T\)), earlier appearance of Ti\(_2\)O\(_x\)C\(_y\)H\(_z\) vs TiO\(_2\) in pyrolysis without O\(_2\), faster TiO\(_2\) formation and higher yield with ambient O\(_2\), and faster cluster formation when residence time is shortened (favoring TiO\(_2\) vapor condensation). Cluster growth is described as Ti–O bond formation to Ti–O\(_x\)C\(_y\)H\(_z\) moieties or clusters followed by C–O scission releasing hydrocarbon fragments.

Methods

  • Engine: ReaxFF in LAMMPS for gas-phase / precursor chemistry with bond-order updates each timestep.
  • Force field: New Ti/C/H/O ReaxFF parametrization for TTIP-derived chemistry (fitted to QM and related reference data as enumerated in the article, including dissociation and cluster energetics).
  • Systems: Isolated precursor droplet models in high-temperature environments; 1000–2500 K; with and without O\(_2\) to separate pyrolysis-limited and oxidation-accelerated pathways.
  • Analysis: Species tracking, reaction pathway identification, cluster size and composition metrics, and residence-time sweeps to connect vapor supersaturation to nucleation rates.

Atomistic protocol (full parameters in Chem. Eng. Sci.): LAMMPS reax/c reactive molecular dynamics; gas-phase TTIP droplet supercells with 3D PBC; NVT ensemble at 1000–2500 K; sub-fs timestep and Nose–Hoover-class thermostat as reported; psns equilibration and production; barostat N/A for the constant-volume gas-phase protocols summarized; hydrostatic pressure N/A in that sense; external electric field N/A; umbrella/metadynamics/replica exchange N/A; atom counts and O₂ partial pressures in the PDF.

Findings

  • Temperature: Very high pyrolysis temperature does not always maximize incipient clusters because Ti–O bonding can weaken at high \(T\), changing the balance between fragmentation and aggregation.
  • Stoichiometry vs. environment: Ti\(_2\)O\(_x\)C\(_y\)H\(_z\) appears before TiO\(_2\) in pyrolysis; with O\(_2\), TiO\(_2\) forms earlier and at higher concentration in the simulated window.
  • Residence time: Shorter high-\(T\) residence boosts clusters by condensing TiO\(_2\)-rich vapor before extensive organic decomposition dilutes oxide precursors.
  • Growth mechanism: Ti–O-mediated aggregation plus bond-breaking that ejects hydrocarbon pieces, consistent with a stepwise conversion from alkoxide ligands to oxide cores.

Limitations

Flame-synthesis models still often rely on simplified global chemistry; this paper focuses on atomistic inception rather than full reactor-scale coupling. Cluster counting thresholds and species definitions should be taken from the Chem. Eng. Sci. text when comparing to other Ti precursor studies.

Relevance to group

ReaxFF parametrization and application for Ti–O–C–H chemistry with van Duin co-authorship—relevant to oxide nanoparticle formation from organometallic precursors. Chem. Eng. Sci. readers comparing to continuum flame models should treat these trajectories as elementary-path libraries for nucleation rather than full reactor maps.

Citations and evidence anchors