Ice growing into brine¶
The front is the height where the crystal slabs stop. CHILL+ labels
the waters. Slabs of 3.7 Å along z are written as a stack: H for
hexagonal, C for cubic, . for a slab that is neither. The front
is the first slab that is not crystal, times 3.7 Å. Ions are not
waters, so they stay out of that count. You still want them in the
box, because the liquid they sit in is the brine the ice is growing
into.
Packmol places six waters, one sodium and one chloride above a fixed GenIce ice Ih block. OpenMM then relaxes three stages. At each later stage the nearest brine oxygen is moved onto the next lattice plane and restrained. That move is the growth step. The relax is why the ions and the leftover liquid can answer it, instead of staying where packmol left them.
meson test docs_paper_brine checks
brine frame 0 hexagonal 31 water 38 front 7.40
brine frame 1 hexagonal 32 water 36 front 7.40
brine frame 2 hexagonal 32 water 35 front 7.40
The front stays at 7.40 Å. The stack remains HH..., so the new
crystal has not filled another 3.7 Å slab. The hexagonal count still
rises from 31 to 32, and the water count falls from 38 to 35. One
molecule has joined the ice. OVITO draws the last frame. The test
exits 77 when GenIce, packmol, OpenMM, or OVITO is absent.
Full run¶
The brine frame the tutorials name is
input/traj/nacl_front.lammpstrj: 864 waters and 9 ions, one frame.
meson test --suite slow docs_paper_brine_full reads that file.
CHILL+ is scored on the waters.
brine full input molecules 873 frames 1 file-frames 1 file input/traj/nacl_front.lammpstrj
brine full paper waters 864 ions 9 frames 1
brine full waters 864 ions 9 hexagonal 352 water 434
Hexagonal is 352, the same water census as the ion-free interface frame. OVITO draws the frame. The test exits 77 when OVITO or Snakemake is absent.