Water: molecules, hydrogen bonds and ice

Starting water model…LIQUID · SPACE

Model time0.000 ps

Water temperature— K

Structureanalysing…

Molecules—

Translation— K

Rotation— K

MD rate —

Start from

256 water molecules with room around the condensed sample, so molecules can evaporate into the surrounding space.

Surroundings temperature

300K

220 K360 K

Exchange energy with surroundings

Real water at about 1 atm: freezes at 273.15 K (0 °C) and boils at 373.15 K (100 °C). This simulation keeps a fixed periodic volume, so those temperatures do not act as switch points.

Show

Hydrogen bonds

Highlight ice-like molecules

Periodic box

One water molecule

H–O–H angle (model): 104.52°
O–H bond (model): 0.957 Å

A-level exam bridge

In the model

Water molecules rotate and rearrange in three dimensions. Dashed lines mark O–H···O arrangements that meet the hydrogen-bond geometry test.

For an A-level answer

Label O as δ− and H as δ+. Draw a dotted line from a δ+ hydrogen on one molecule to the δ− oxygen or a lone pair on another. Call the intermolecular attraction a hydrogen bond. Keep the covalent O–H bonds as solid lines.

Where this model stops

TIP4P/Ice holds every H–O–H angle at 104.52° because that geometry is part of the model. For normal A-level work, use about 104.5° for water. Do not learn 109° as a second H–O–H angle for ice from this simulation.

Current sample

Hydrogen bonds shown: —
Ice-like molecules: —
Ice Ih-like: —
Ice Ic-like: —
Ice interface: —
Force calculation: —

Model details and limitations

This model uses rigid TIP4P/Ice water molecules. Oxygen carries the Lennard-Jones site; two positive H charge sites and a negative massless M site give the model its fixed charge distribution. Molecular translation and full 3D rotation are integrated live.

Electrostatics use a browser WebAssembly particle-mesh Ewald research kernel with fourth-order B-spline charge assignment. The ice overlay uses the q₃ bond-correlation core of CHILL+.

In the free-surface samples, the larger cell gives molecules room to evaporate. The calculation is still periodic and does not represent a beaker at 1 atm. Small clusters have strong surface and finite-size effects.

Ice growth over nanosecond timescales is too slow for a normal classroom browser run. Prepared ice and the short live trajectories are calculated directly. A later ice-growth view will use a separately validated long molecular-dynamics trajectory.