Ab initio machine learning simulation of calcium carbonate from aqueous solutions to the solid state

March 10, 2026 ยท View on GitHub

Authors: Pablo M. Piaggi, Julian D. Gale, and Paolo Raiteri

DOI arXiv Zenodo

Analysis and input files to reproduce simulations of manuscript "Ab initio machine learning simulation of calcium carbonate from aqueous solutions to the solid state" by Piaggi, Gale, and Raiteri

Description of folder contents:

  • molecular-dynamics: Input and output files of molecular dynamics simulations
  • notebooks: Jupyter notebooks with analysis
  • training-process: Training scripts for the DeePMD-kit
  • training-set: Training set in format suitable for the DeePMD-kit

Software needed to reproduce the simulations:

Simulations were performed using the following codes:

  • LAMMPS (23 Jun 2022)
  • DeePMD-kit v2.1.3
  • PLUMED 2.8.0

Jupyter notebooks require the following packages:

  • Python 3.10.4 (main, Mar 31 2022, 08:41:55) [GCC 7.5.0]
  • Matplotlib 3.5.3
  • Numpy 1.22.4
  • Scipy 1.7.3
  • Ase 3.23.0b1

Please e-mail me if you have trouble reproducing the results.