Skip to content

5.2 Prepare liquid water without calling a packing a liquid

These are unexecuted teaching inputs and starting models. Original diagrams are schematics, not calculated results. Validate version-specific syntax, licensed or authorized data, numerical convergence and the scientific model before using this workflow.

5.2.1 Model, units and provenance

Keep basis/potential files and executable versions traceable. Grid controls use Ry; common energy/force outputs use hartree and hartree/bohr. Read the unit in each output heading.

Shared inputs, conventions and evidence

Original schematic: Prepare liquid water without calling a packing a liquid. No numerical results are claimed.
Original schematic: Prepare liquid water without calling a packing a liquid. No numerical results are claimed.

5.2.2 Unexecuted inputs and explicit deltas

Use the accompanying instructions to identify the parent calculation and placement of every delta; a snippet is not automatically a standalone input. Preserve all blank-line and file-provenance requirements.

5.2.2.1 Input block 1

! Replace inline COORD with SUBSYS/TOPOLOGY
&TOPOLOGY
  COORD_FILE_NAME water32.xyz
  COORD_FILE_FORMAT XYZ
  CONN_FILE_FORMAT OFF
&END TOPOLOGY
! SUBSYS/CELL
&CELL
  ABC [angstrom] 9.86 9.86 9.86
  PERIODIC XYZ
&END CELL

5.2.3 Worked investigation

Intuition. A list of water molecules at a chosen density is a starting configuration, not an equilibrated liquid. Overlaps and artificial orientational order can dominate early dynamics. Use an external packing tool or a documented equilibrated configuration with redistribution permission; record its provenance, molecule count and units. Classical pre-equilibration can reduce bad contacts, but switching to DFT changes the Hamiltonian and requires new equilibration.

Original input delta. For a 32-water teaching cell at a nominal density near 1 g/cm³, a cube of about 9.86 angstrom is a useful geometry-derived starting estimate, not a predicted DFT equilibrium volume. Replace A's coordinates with a real 96-atom XYZ file and use periodic electrostatics:

See input block 1 above.

Set POISSON PERIODIC XYZ and POISSON_SOLVER PERIODIC, retaining H/O basis and potentials. The coordinate file must exist; this course does not claim to supply an equilibrated liquid snapshot.

Workflow. 1. Count 32 O and 64 H, inspect intramolecular geometry and minimum intermolecular distances under periodic images. 2. Check that box volume and molecular masses reproduce the stated nominal density. 3. Perform a static DFT energy/force calculation; extreme forces indicate bad contacts or a geometry/data error. 4. If needed, relax carefully or use a documented preparation protocol without mistaking the relaxed structure for a liquid. 5. Start a short, conservative NVT equilibration from case 22, monitoring temperature, potential energy and structure. 6. Discard the transient based on stationarity and independent checks, not a fixed universal number of steps.

Interpretation and tests. A small periodic cell restricts long-wavelength fluctuations and limits accessible diffusion/structural statistics. The chosen functional and dispersion treatment affect water structure and density; an experimental density imposed in NVT does not prove the model predicts it. Repeating with another independently prepared configuration is more informative than simply extending an obviously biased start. Check SCF robustness on several snapshots rather than only the initial geometry.

Pitfalls and exercise. Randomly placing atomic H/O positions is not equivalent to packing intact water molecules. A crystalline-looking packing may remain ordered during a short expensive AIMD run. A visually disordered snapshot may still have unequilibrated energy or coordination statistics. Prepare an audit sheet with composition, nominal density, minimum distances, provenance, and the first observables used to diagnose equilibration. Design a larger-cell comparison without claiming the teaching cell is production-ready.

Diagram. Three states: packed molecules, early relaxation, equilibrated ensemble candidate. A gate between the last two lists stationarity checks. EN: preparation, transient, structural checks, independent start, production not yet assumed.

5.2.5 Sources and further reading