UNIT CELLS

(UC)

MATERIALS SCIENCE
Part & A Learner’s Guide
of
ENGINEERING
AN INTRODUCTORY E-BOOK
Anandh Subramaniam & Kantesh Balani
Materials Science and Engineering (MSE)
Indian Institute of Technology, Kanpur- 208016
Email: anandh@iitk.ac.in, URL: home.iitk.ac.in/
~anandh
http://home.iitk.ac.in/~anandh/E-book.
htm
Important Note: the term “unit cell” in the current context refers to a crystallographic unit
cell.
Other geometrical shapes can be used as a “cell” or a “unit” of the lattice or crystal; but
technically they may not qualify to be called a unit cell.
❑ A unit cell (also sometimes causally referred to as a cell) is a representative
unit of the structure which when translationally repeated (by the basis
vector(s))gives the whole structure.
❑ The term unit should not be confused with ‘having one’ lattice point or motif
(The term primitive or sometimes simple is reserved for that).
❑ If the structure is a lattice, the unit cell will be unit of that (hence will have
points* only).
❑ If the structure under considerations is a crystal, then the unit cell will also
contain atoms (or ions or molecules etc.).
Note: Instead of full atoms (or other units) only a part of the entity may be
present in the unit cell (a single unit cell)
❑ The dimension of the unit cell will match the dimension of the structure**:
If the lattice is 1D → the unit cell will be 1D,
if the crystal is 3D → then the unit cell will be 3D,
if the lattice is nD → the unit cell will be nD.
Lattic Will contain lattice points
Unit of e only
cell a Cryst Will contain entities which decorate
al a point’ (andthe
* Strangely in crystallography often we even ‘split lattice
say that 1/8 th belongs to the UC).
** One can envisage other possibilities– e.g. a 2D motif may be repeated only along one direction (i.e. the crystal is 3D but the
repeat direction is along 1D)
 Why Unit
Cells?
Instead of drawing the whole structure I can draw a representative part and
specify the repetition pattern.
ADDITIONAL POINTS
❑ A cell is a finite representation of the infinite lattice/crystal.
❑ A cell is a line segment (1D) or a parallelogram (2D) or a parallelopiped (3D)
with lattice points at their corners → This is the convention.
❑ If the lattice points are only at the corners, the cell is primitive.
Hence, a primitive unit cell is one, wherein the lattice points are only at the
corners of the unit cell (or the ends of a line segment unit cell in 1D).
Primitive unit cell has one lattice point per cell (this is not defined w.r.t to atoms or other entities
of the motif).
Hence, it is wrong to say that a primitive unit cell is defined to have one atom per cell.
❑ If there are lattice points in the cell other than the corners, the cell is
non-primitive.
∞

Consider an infinite pattern made of squares

∞
This way the infinite information
∞ content of a crystal can be reduced
This can be thought of
to the information required to
as a single square specify the contents of a unit cell
repeated in x and y (along with the lattice translation
∞

directions
vectors).
Different types of
cells
❑ In general the following types of ‘cells’ can be defined:
Primitive unit cell
Non-primitive unit cells
Voronoi cells
Wigner-Seitz cells
Asymmetric Unit.
❑ As we shall see later, some of them are not crystallographic unit cells.
 Q& Define a Unit Cell (Crystallographic
A Unit Cell).
❑ An unit cell is an unit of the structure [is a line segment in 1D, a parallelogram
in 2D and a parallelepiped in 3D], such that lattice points are at the ends of the
line segment (1D) or at the vertices of the parallelogram (2D) or
parallelepiped (3D); which when repeated by the translational symmetry
vector(s) generates the whole structure (i.e. a lattice or a crystal).
❑ A primitive unit cell has only one lattice point per cell, which are at: the ends
of the line segment (1D) and at the vertices of the parallelogram (2D) or
parallelepiped (3D).
(A primitive unit cell need not have 1 atom per cell!!).
 Fu
nd Is the blue hexagon below a
a
Ch
unit cell?
▪ eck blue hexagon can tile the plane
The
without leaving any gaps.
▪ It can also serve as an unit for the
hexagonal lattice− but is not a
conventional crystallographic unit cell,
as it is not a parallelogram.
▪ This is because, this shape cannot be
used in conjunction with the lattice
translation vectors to generate the
whole structure; as there will be
overlaps.
❑ By the same token, the triangle below is also not a unit cell. As for the
hexagon, the triangle can be used monohedrally to tile the plane.
1
D
1
D
❑ Unit cell of a 1D lattice is a line segment of length = the lattice
parameter
this is the PRIMITIVE UNIT CELL (i.e. has one lattice point per cell).
Each of these lattice points contributes half a lattice point to
the unit cell
Primitive
UC

Contributions to the unit cell: Left point = 0.5, Middle point = 1, Right point =
0.5. Total = 2

Doubly Non-
primitive UC

Triply Non-
primitive UC
❑ Unit cell of a 1D crystal will contain MotifsUnit cell of a 1D crystal will
contain Motifs in addition to lattice points.
❑ NOTE
The only kind of motifs possible in 1D are line segments.
Hence in ‘reality’ 1D crystals are not possible as Motifs typically have a
finite dimension (however we shall call them 1D crystals and use them for
illustration of concepts).
We could have 2D or 3D motifs repeated along 1D (hence periodicity and
‘crystallinity’ is only along 1D).
Correct unit Though the whole lattice point
cell is shown only half belongs to
the UC

Each of these atoms contributes ‘half-atom’ to

the unit cell

Though this is the correct

unit cell
Often unit cells will be drawn like this→ can be terms as
‘extended unit cell’
Or ‘ball and stick’ model of a unit cell
❑ Unit cell in 1D is described by 1 (one) lattice parameter: a
2
D
2 ❑ Unit cell in 2D is described by 3 lattice parameters: a, b, α
D b α
a
❑ Special cases include: a = b; α = 90° or 120°

❑ Unit Cell shapes in 2D → Lattice parameters

Square → (a = b, α = 90°)
Rectangle → (a, b, α = 90°)
120° Rhombus → (a = b, α = 120°)
Parallelogram (general) → (a, b, α)
2
D Infinite such
primitive unit
cells are possible
for the rectangle
lattice
UC- Note: basis vectors (& included
1 angle) will change based on the
‘unit cell’ chosen [which implies
that lattice parameters will change
α≠90 as well !]
°
Rectangl Area of a doubly non-
primitive UC is double
e that of the primitive
one (as space is
lattice dividied equally
among lattice points.

Note: these are

the basis vectors
(and included
angle) for UC-1
above

90
°

Note: Symmetry of the Lattice or the crystal is not altered by our choice
 IMPORTANT
Symmetry (or the kind) of the Lattice or
the crystal is not altered by our choice of
unit cell!!

You say this is obvious

I agree!
How to choose a unit
cell?
❑ When possible we chose a primitive unit cell
❑ The factors governing the choice of unit cell are:
Symmetry of the Unit Cell → should be maximum (corresponding
to lattice*)
Size of the Unit Cell → should be minimum
Convention → if above fails to resolve the issue we use some
convention. How does convention come into play in the choice of unit cell for
Orthorhombic lattices?
(We will see later - using an example- that convention is not without common sense!).
* The lattice may have higher symmetry than the crystal→ but in choosing the unit cell we
focus on the lattice.
E.g. if we decorate a square lattice with a triangle motif, we land up with a rectangle
crystal. But we prefer to chose a square unit cell for the crystal as well.
If the symmetry of the unit cell can be ‘commensurate’ with the lattice/crystal, this is the
best.
Centred square lattice = Simple square lattice
& the smaller square UC is preferred (which is
primitive).

This is nothing
but a square
lattice viewed at
45°!

Continued
…
❑ In this case the primitive (square) and the non-primitive square cell both
have the same symmetry
But the primitive square cell is chosen as it has the smaller size
❑ The primitive parallelogram cell is not chosen as it has a lower symmetry
The lattice has 4-fold symmetries as shown F
The square cells also have 4-fold symmetry u
l
The parallelogram cell does NOT have 4-fold symmetry l
(only 2-fold → lower symmetry)
s
y
m
m
e
t
r
y

i
s

4
m
m

Note these are symmetries of

the UC and not of the lattice!
Centred Rectangular
Lattice Lattice parameters: a, b, α =
90°
Note that the distribution of
symmetry elements has not
changed
(as compared to the Simple
Rectangular Lattice)

2m
m

Unit Cell of 2m
Lattice m

▪ The blue and the green unit cells have the same symmetry, which is 2mm.
Hence, we now have to go by size and chose the green UC as the preferred
one.
▪ However, by convention we chose the blue unit cell as the ‘preferred one’!!!
(What is the thought process behind this?)
▪ This is the reason (i.e. the choice of a reference rectangle UC) for callingContinued
the
 Simple rectangular Now the UC of the crystal will have a
Crystal motif
(Not a centred crystal) Part of the structure

Unit Cell the

way it is
usually
shown

Though the
whole lattice
point is shown
only one fourth
belongs to the
UC

True
Unit Cell of
Crystal

Correct unit
cell
Note that the UC has entities of
the motif in parts!
The centres of only the green circles are
lattice points
(of course equivalently the centres of only the maroon
 nd
a Are the lattice translation vectors to be used with unit cells (to
Ch obtain the whole structure) unique?
❑ eck
Perhaps this one is a ‘no brainer’. Clearly the lattice translation vectors to be
used with unit cells depend on the unit cell chosen.
❑ For the two unit cells shown below (both primitive) (UC-1 & UC-2), the b
lattice translation vector is different (the a is same for both the unit cells).
❑ Also, this implies that the lattice parameters are not the same for the two unit
cells.
UC-1: a, b, α = 90°
UC-2: a, b 2, α ≠ 90°.

UC-
1

UC-
α≠90 2
°
 Click
here
Solved Choice of Lattice, Motif, UC, Symmetry Elements etc
Exampl are illustrated in the example
(try and understand those concepts with which you are familiar at this
e juncture and postpone the other concepts for a later discussion)
3
D
Cells-
3D
❑ In order to define translations in 3-d space, we need 3 non-coplanar
vectors.
❑ With the help of these three vectors, it is possible to construct a
parallelepiped called a UNIT CELL.
❑ Conventionally, the fundamental translation vector is taken from one
lattice point to the next in the chosen direction.
 Unit Cell shapes in
3D: Preferred unit cell Constraints on lattice
Constraints on lattice
Shape for parameters
parameters (angles)
_____ crystal system (distances)
Cube Cubic a=b=c α = β = γ = 90°

Square Prism Tetragonal a=b≠c α = β = γ = 90°

Rectangular Prism Orthorhombic a≠b≠c α = β = γ = 90°

120° Rhombic Prism Hexagonal a=b≠c α = β = 90°, γ = 120°

Parallopiped
Trigonal
(Equilateral, a=b=c α = β = γ ≠ 90°
(rhombohedral)
Equiangular)
Paralleogramic Prism Monoclinic a≠b≠c α = γ = 90° ≠ β
Parallopiped
Triclinic a≠b≠c α≠ β≠ γ
• (general)“≠” implies → “need not be equal to”.
The symbol
• Some common names of unit cells are given here → alternate names are also used
for these cells.
• As we have noted in many places, these are conventional unit cells chosen and
alternate unit cells are also possible for the structure (for which this unit cell
shape is chosen).
• Also, as we have noted elsewhere, these are unit cell shapes and not be confused
with the definition of crystal system (i.e. these unit cell shapes do not define the
Different kinds of
“CELLS”
Unit cell (Crystallographic Unit Cell)
A unit cell is a unit of the structure, which when translated by the lattice
translation vectors, gives rise to the entire lattice or crystal. (More details as given before).

Primitive unit cell

For each crystal structure there is a conventional unit cell, usually chosen to
make the resulting lattice as symmetric as possible. However, the conventional
unit cell is not always the smallest possible choice. A primitive unit cell of a
particular crystal structure is the smallest possible unit cell one can construct
such that, when tiled, it completely fills space.
Wigner-Seitz cell
A Wigner-Seitz cell is a particular kind of primitive cell, which has the same
symmetry as the lattice. We will see soon that it is not a crystallographic unit cell.

Asymmetric unit
The asymmetric unit in combination with the space group will tile the entire
space.
Tiling and Covering
One or more types of tiles* can be used to fill space. These tiles can act like ‘cells’
of the structure. Usually these tiles do not overlap or leave gaps. If overlaps are
allowed then the process is called covering.
* Akin to the tiles used in floors. Only that normal tiles are 2D, while we can think of 3D tiles for 3D crystals.
 Q&
How many ‘shapes’ of primitive unit cells are possible?
A
❑ 1D → one.
❑ 2D, 3D → Infinite (few examples in 2D given below) .

1
D

2
D
s uch !
e
finit ilit ie s
In sib
s
po
 Fu
nd Q&
a +
Ch A
eck
Important Note: the term “unit cell” in the current context refers to a crystallographic unit
cell (CUC).
Other geometrical shapes can be used as a “cell” or a “unit” of the lattice or crystal; but
technically they may not qualify to be called a unit cell.
❑ We had started with the important note as above. We will now see some
examples of cells which are a “unit” of the structure but do not form an
crystallographic unit cell.
❑ Specific cases we will see are (wherein a “cell” is not a crystallographic unit
cell):
the vertices of the cell are not lattice points.
the shape of the cell is not a parallelogram (2D) or a parallelepiped (3D).
❑ In such cases the lattice translation vectors cannot fill space without either
overlaps or gaps.
❑ We have already noted that the CUC is not unique; this implies that for each
type of unit cell the appropriate lattice translation vectors have to be used to
fill entire space (without gaps or overlaps).
❑ We have also noted before that in some cases we draw a compound/
combination of cells, ‘keeping in view’ the symmetry of the lattice/crystal. In
that case, we can draw a combination of three cells for the hexagonal system.
Readers may also refer to the chapter on Miller Indices, wherein we use the
Miller-Bravais indices for hexagonal (and sometimes trigonal) systems,
 nd Should a Unit Cell have Lattice Points only at the
a
Corners?
Ch
eck
❑ The conventional unit cell chosen has lattice points at the ends of line segment/
corners/vertices …
1 2
D D

Conventional
UCs

❑ But in principle any unit cell like the ones below

(space filling) should work fine! (all the illustrated
cells fill space!)
❑ We had earlier seen that conventional choice of unit cells can ‘cut into’ the
lattice points (and may also cut into entities of motif) (as below).
❑ Choices of some “non-conventional” cells (like the ones drawn before) can
alleviate this problem of ‘cutting into’ lattice points.
❑ The new unit cell may still (or may not as below) cut into parts of the motif.
Corners
of the
unit cell
New choice of “non-
conventional” cell
still have
to be
lattice
The natural or preferred unit
cell points
By
Problem with this unit cell:
convention
UC has entities of the motif in
parts!

This is the natural or preferred unit cell as the symmetry

operators coincide with the lattice points. Symmetry
operators are “stuck’ to the crystal and not “affected” by our
choice of unit cell.
Wigner-Seitz Here we limit ourselves to the Voronoi tessellation
Cell of lattices
❑ The Wigner-Seitz cell is a primitive cell with the symmetry of the lattice.
❑ Technically it is not a unit cell as it does not have lattice points as vertices
and may not be a parallelogram/parallelepiped either.
❑ This is created by Voronoi tessellation of space.
❑ The region enclosed by the Wigner-Seitz cell is closer to a given lattice point
than to any other lattice point. Space is equally apportioned to lattice points.
❑ Wigner-Seitz cell is a primitive cell (in that it contains one lattice point per
cell).
❑ The Wigner–Seitz cell in the reciprocal space is the first Brillouin zone.
❑ In 2D the perpendicular bisector lines are drawn to obtain the Voronoi
tessellation. In 3D the perpendicular bisector planes are drawn.
 Wigner-Seitz Cell
(cotd..)
Square
lattice

m
o t a gra
N elo
ral
l Hexagonal Lattice (120°
pa rhombus)
And hence not a conventional
unit cell.
Wigner-Seitz
cells

Not a conventional unit cell as the vertices are not

lattice points

Centred rectangle
lattice
 Wigner-Seitz Cell
BCC Tetrakaidecahe (cotd..)
dron FCC Rhombic
Dodecahedron

ed
t a ip
No lep
le
ral
p a

ed
t a ip
N o l ep
le
ral
p a

• The Tetrakaidecahedron is a space filling

solid, which is semi-regular. This shape is
obtained by the truncation of the
octahedron.
• The rhombic dodecahedron has been
• Note: reciprocal lattice of FCC is BCC and
vice-versa. considered as the least ‘photogenic’ solid!
• This is also a ‘semi-regular’ space filling solid.
• This is the coordination polyhedron for the
BCC lattice.
Asymmetric
unit
❑ The asymmetric unit in combination with the space group gives rise to the
entire crystal.
❑ It is the smallest irreducible unit, which essentially does not contain any
symmetry elements related to the crystal.
❑ Let us understand the concept for the square crystal (Xal.1) with P4mm space
group (patch of the crystal shown as below). Starting with the triangle (Tr.1), the square (Sq.1)
can be generated using the vertical (mv), horizontal (mh) and diagonal (md )
mirrors. The triangle (Tr.1) has no symmetry elements related to the crystal
(though it has some symmetry still left!) → this is referred to as the asymmetric unit. This
in combination with the space group P4mm can give us the original crystal.
∞

Consider an infinite pattern made of squares

Xal.
∞ 1 ∞ Tr.
1

m
∞

This can be thought of

h m
as a single square Sq. m d
repeated in x and y 1 v
directions
This blue square in conjunction with the simple
square lattice will give rise to the whole crystal (so
add a ‘P’)