In plain words

The smallest block of a crystal which, repeated over and over in every direction, builds the whole thing – as one motif, repeated, covers a whole roll of patterned wallpaper. Its edge lengths and angles are the numbers quoted as a crystal’s lattice parameters.

Going deeper

Left: a two-dimensional lattice of points with one parallelogram highlighted as the unit cell, its two edge vectors a and b marked by arrows and the angle γ between them. Right: a side view of four stacked layers of a layered crystal; a bracket spanning two layers and their gaps marks the repeat distance c, and an arrow inside one layer marks the in-plane spacing a. a two-dimensional unit cell a b γ repeat it along a and b and it builds the whole sheet; a, b and the angle γ are the lattice parameters the cell of a layered crystal c a one layer van der Waals gap a and b are fixed by bonding inside a layer; c counts the layers in one repeat plus their gaps, so polytypes differ only along c
The unit cell is the smallest block that builds the whole crystal by repetition. In a layered material the in-plane lengths come from the bonding inside a layer, while the repeat along the stacking direction counts layers and the gaps between them.

The block that builds the crystal

A crystal is a lattice of repeated points with the same group of atoms attached to each. The unit cell is the parallelepiped that fills space by translation alone, described by three edge lengths a, b and c and the three angles α, β and γ between them, plus the positions of the atoms inside it. Those six numbers are the lattice parameters that any crystallographic description quotes.

The choice is not unique: many cells can generate the same lattice. Convention picks the smallest cell consistent with the symmetry, which is why a hexagonal crystal is described with a 120° angle rather than a rectangular box. The full set of conventions is codified in the International Tables for Crystallography.

Cells of layered materials

In a layered crystal the two in-plane parameters are set by strong bonding within a layer and hardly change between : graphene, graphite and graphene all share an in-plane spacing of about 2.46 Å. The third parameter is different in kind, because it counts how many layers pass before the pattern repeats, and each layer contributes its own thickness plus a .

That is why 2H-MoS2 has c of about 12.3 Å, two layers of about 6.15 Å each, while 3R-MoS2 has a c of three such layers. Polytypes share a and b and differ along c. Pressure squeezes c far more easily than a, and expands c while leaving a untouched.

Describing a single layer

A monolayer is not periodic perpendicular to itself, so it has no meaningful c. Its cell is given by a, b and γ, with the layer thickness stated separately as a measured or nominal value rather than a lattice parameter. Calculations, however, usually place the layer in a periodic box with vacuum above and below, and the vacuum height is a convergence parameter rather than a property of the material – an important distinction when comparing published numbers.

A between two twisted layers has a cell of its own, tens of nanometres across and containing thousands of atoms, which is why moiré calculations use effective models rather than treating every atom.

For specialists

The repeating parallelepiped that generates the lattice under translation, given by a, b, c and α, β, γ together with the positions of the atoms inside it. In a layered material the in-plane parameters are fixed by bonding while c depends on stacking and on the van der Waals gap, so polytypes share in-plane parameters and differ along c. A 2D cell quotes a, b and γ only, with the layer thickness stated separately.

Where this comes from

  1. International Tables for Crystallography International Union of Crystallography · Volume A: space-group symmetry (2006) cited by 829