Crystallography, Symmetry and Defects of Materials

Course Code: MSE8013

Course Time: 2022/2023 Semester A
Course Duration: 13 weeks (29 August 2022 – 21 November 2022)

Lecturer: Dr. Dangyuan LEI

• Room: R7125
• Phone: 3442 2487
• Email: [email protected]

Tutors: Mr Yang FU / Miss Xue MA

• Room: Y2440, AC1
• Phone: 6465 7596
• E-mail: Yang.Fu / [email protected] 1
 Course content
 A general introduction

 Structure and symmetry of the crystal

 Wave diffraction and the reciprocal lattice

 Crystal binding

 Point defects and dislocations

 Dislocations, elastic constant, and crystal vibration

2
Arrangements of atoms
Disorder
 No regular ordering of atoms
E.g.: inert monoatomic gases (Ar)

Short-range order
 Ordering only extends to the atom’s nearest
neighbors.
E.g.: water vapor (covalent bond between the hydrogen and
oxygen atoms, 104.5o), nitrogen gas, silicate glass (硅酸盐玻璃).

Long-range order
 Ordering extends over a large length scale
(typically > 100 nm) or even through the
whole material.
E.g.: metals, ceramics, some polymers
3
Classification of Materials in Terms of Atomic
Arrangement

Quasicrystal: ordered but not periodic

Daniel Shechtman, 2011 Nobel Prize in Chemistry
Fundamentals of Crystal
 An ideal crystal is a two- or three-dimensional infinite periodic
array of atoms or groups of atoms, usually with regular geometric
structure and symmetry.
 Macroscopic single crystals are usually identifiable by their
geometrical shape, consisting of flat facets with specific,
characteristic orientations.
 A crystal has the following characteristics:
• Uniformity, i.e. the macroscopic properties are the same over the
whole crystal;
• Anisotropy, i.e. different properties along different crystal
directions (i.e. directional properties);
• Ability to form polyhedron (多面體) spontaneously;
• Have a definite and obvious melting point;
• Have specific symmetry;
• Produce diffraction patterns with X-rays and electron beams.
5
Examples of crystal growth
 The elementary building blocks in (a) and (b) are identical, but
different external forms are developed due to different atomic
arrangement.
 Determination of crystal orientation helps cleave a crystal.

(a) Pyramid-like facets* (b) Stair-like facets*

6
*C. Kittel, Introduction to solid state physics (8th edition), Page 2
Bravais lattice: physical picture
 An ideal crystal is constructed by an infinite repetition of
identical groups of atoms. The group is called a basis. It may be a
single atom, or a group of atoms, molecules, or ions, etc.
 The set of mathematical points to which the basis is attached is
called the lattice or Bravais lattice. Therefore, a crystal structure
is the combination of basis and Bravais lattice.

Bravais lattice Basis Crystal structure*

(parallelogram) (containing two
different ions) 7
*C. Kittel, Introduction to solid state physics (8th edition), Page 4
Bravais lattice: physical definition
 We give two equivalent definitions of a Bravais lattice:
(a) A Bravais lattice is an infinite array of discrete points with an
arrangement and orientation that appears exactly the same,
from whichever of the points the array is viewed.
(b) A (three-dimensional) Bravais lattice consists of all points with
position translation vector 𝐑 of the form
𝐑 = 𝑢1 𝐚𝟏 + 𝑢2 𝐚𝟐 + 𝑢3 𝐚𝟑
where 𝐚𝟏 , 𝐚𝟐 and 𝐚𝟑 are three vectors not all in the same plane, and
𝑢1 , 𝑢2 and 𝑢3 are arbitrary integers.

 Both statements above represent the uniformity of a crystal and

are related to the invariance of crystal properties.

8
Primitive vectors and primitive cell
 The vectors 𝐚𝟏 , 𝐚𝟐 and 𝐚𝟑 defined in the last slide are the so-
called primitive vectors.
 A volume of space that, when translated through all the vectors
in a Bravais lattice, just fills all the space without either
overlapping itself or leaving voids is called a primitive cell*.
* The adjacent axes of a primitive cell are formed by primitive
vectors.
 A primitive cell defined by primitive axes is a minimum-volume
periodic cell. The volume of the parallelepiped (平行六面体)
defined by primitive vector 𝐚𝟏 , 𝐚𝟐 and 𝐚𝟑 in a three-dimensional
lattice is given as
𝑉c = 𝐚𝟏 ∙ 𝐚𝟐 × 𝐚𝟑

9
Primitive cell vs. Unit cell
 Unit cell: The smallest group of particles in the crystal that
constitutes the repeating pattern is the unit cell. The unit cell
completely reflects the symmetry and structure of the entire
crystal.
• In 3D, the unit cell is a parallelepiped whose vertices are lattice
points; In 2D it is a parallelogram whose vertices are lattice
points.
• A unit cell may be primitive or non-primitive.
= Primitive Cell

Two choices of unit cells in a 2D lattice 10

https://www.doitpoms.ac.uk/tlplib/crystallography3/unit_cell.php
Primitive vectors and primitive cell
 Sharing of atoms
Face-centered atom shared
between two unit cells

Corner atom shared by 8

units (1-4 in front, 5-8 in
back)

11
https://cnx.org/contents/[email protected]:EWNgknft@4/Lattice-Structures-in-Crystalline-Solids
Primitive vectors and primitive cell
 A primitive cell defined by primitive vectors
is usually a parallelogram (平行四边形) with
4 points, where each point is shared by
adjacent 4 cells for a two-dimensional lattice.

 For a three-dimensional lattice, the primitive

cell is a parallelepiped with 8 points and
every point in it is shared by 8 adjacent cells,
such that the total number of lattice points
in a primitive cell is one.

 The choice of primitive cells for a given

Bravais lattice is not unique. Different
choices for the primitive translations vectors
are possible. 12
Primitive vectors and primitive cell

Several possible choices of pairs of primitive vector except 𝐚′′′

𝟏
and 𝐚′′′
𝟐 because we cannot form the lattice translation T from
integral combinations of 𝐚′′′𝟏 and 𝐚 ′′′
𝟐 . And the parallelogram 1, 2,
3 are equal in area and any of them could be taken as a primitive
cell with a total lattice points of one.
The parallelogram 4 has twice the area of each primitive cell.
There are four points in parallelogram 4 shared by four cells and
two points in the middle of long edges shared by two cells, such
that the total number of lattice point is 2 and it is therefore not a
primitive cell. 13
*C. Kittel, Introduction to solid state physics (8th edition), Page 5
Primitive vectors and primitive cell
 Another way of choosing a primitive cell may follow this
procedure:
(1) draw lines to connect a give lattice point to all nearby
lattice points;
(2) at the midpoint and normal to these lines, draw new lines or
planes.
The smallest volume enclosed by the new lines or planes is a
Wigner-Seitz primitive cell and it includes only one lattice point.

2D Wigner-Seitz cell Wigner-Seitz cell for body- Wigner-Seitz cell for face-
centered cubic lattice centered cubic lattice
https://www.physics-in-a-nutshell.com/article/5/unit-cell-primitive-cell-and-wigner-seitz-cell 14
http://lamp.tu-graz.ac.at/~hadley/ss1/studentpresentations/script/Aschl09.pdf
2D lattices: 5 types
Restrictions on
Lattice type axes and angles Unit cell

Oblique 𝐚𝟏 ≠ 𝐚𝟐 ; 𝜑 ≠ 90°

Square 𝐚𝟏 = 𝐚𝟐 ; 𝜑 = 90°

Rectangular 𝐚𝟏 ≠ 𝐚𝟐 ; 𝜑 = 90°

Centered 𝐚𝟏 ≠ 𝐚𝟐 ; 𝜑 = 90°
rectangular

Hexagonal 𝐚𝟏 = 𝐚𝟐 ; 𝜑 = 120°
The primitive cell of the centered rectangular lattice is essentially an oblique cell such that it can be
included in the category of oblique lattice. Therefore, it may be classified to four systems for 2D 15
lattices in some textbooks.
3D lattice: 14 types Primitive Base-
(Simple) centered
Body- Face-
centered centered
𝐚𝟏 ≠ 𝐚𝟐 ≠ 𝐚𝟑
Triclinic 𝛼≠𝛽≠𝛾

𝐚𝟏 ≠ 𝐚𝟐 ≠ 𝐚𝟑
Monoclinic 𝛼 = 𝛾 = 90° ≠ 𝛽

Orthorhombic 𝛼𝐚𝟏= 𝛽≠ =𝐚𝟐𝛾 =≠ 90°

𝐚𝟑

𝐚𝟏 = 𝐚𝟐 ≠ 𝐚𝟑
Tetragonal 𝛼 = 𝛽 = 𝛾 = 90°

Cubic 𝐚𝟏 = 𝐚𝟐 = 𝐚𝟑
𝛼 = 𝛽 = 𝛾 = 90°

Rhombohedral 𝛼𝐚𝟏=≠𝛽 =𝐚𝟐𝛾 ≠ 𝐚𝟑

< 120°, ≠ 90°
𝐚𝟏 = 𝐚𝟐 ≠ 𝐚𝟑
Hexagonal 𝛼 = 𝛽 = 90° with 𝐚𝟏 =a, 𝐚𝟐 =b, 𝐚𝟑 =c
𝛾 = 120° 16
3D lattices: 7 systems
Restrictions on
System Number of lattices axes and angles
𝐚𝟏 ≠ 𝐚𝟐 ≠ 𝐚𝟑
Triclinic 1 𝛼≠𝛽≠𝛾
𝐚𝟏 ≠ 𝐚𝟐 ≠ 𝐚𝟑
Monoclinic 2 𝛼 = 𝛾 = 90° ≠ 𝛽
𝐚𝟏 ≠ 𝐚𝟐 ≠ 𝐚𝟑
Orthorhombic 4 𝛼 = 𝛽 = 𝛾 = 90°
𝐚𝟏 = 𝐚𝟐 ≠ 𝐚𝟑
Tetragonal 2 𝛼 = 𝛽 = 𝛾 = 90°
𝐚𝟏 = 𝐚𝟐 = 𝐚𝟑
Cubic 3 𝛼 = 𝛽 = 𝛾 = 90°
𝐚𝟏 ≠ 𝐚𝟐 ≠ 𝐚𝟑
Rhombohedral 1 𝛼 = 𝛽 = 𝛾 < 120°, ≠ 90°
𝐚𝟏 = 𝐚𝟐 ≠ 𝐚𝟑
Hexagonal 1 𝛼 = 𝛽 = 90°
𝛾 = 120°

17
Cubic lattice: unit cell
Unit cells of a three-dimensional lattice are not often primitive
cells. Here we discuss cubic lattice as an illustration.

 Simple cubic (SC)

• Simplest stacking manner of atoms.

• The unit cell of a simple cubic lattice is

a primitive cell. Each lattice point is
shared among eight cells, such that
the total number of lattice points in
1
the cell is one: 8 × = 1.
8

18
https://cnx.org/contents/[email protected]:EWNgknft@4/Lattice-Structures-in-Crystalline-Solids
Cubic lattice: coordination number
 Simple cubic (SC)

• The coordination number, which is the

number of nearest neighbors of each
atom in the crystal, is 6. The distance of
nearest neighbors is 𝑎, where 𝑎 is the
edge length of the cube.

• Three primitive vectors are:

𝐚𝟏 = 𝑎𝒙; 𝐚𝟐 = 𝑎𝒚; 𝐚𝟑 = 𝑎𝒛 ,
and 𝒙, 𝒚, 𝒛 are unit vectors of Cartesian
coordinates.

19
https://cnx.org/contents/[email protected]:EWNgknft@4/Lattice-Structures-in-Crystalline-Solids
Body-centered cubic lattice: unit cell
 Body-centered cubic (BCC)

• Simplest cubic structure with one more

point in the body center.

• The unit cell of a body-centered cubic

lattice, which contains two lattice points
(each point in the vertex is shared
among eight cells, and the body-
centered point is independent, leading
1
to 8 × + 1 = 2 points), is not a
8
primitive cell.

20
https://cnx.org/contents/[email protected]:EWNgknft@4/Lattice-Structures-in-Crystalline-Solids
BCC lattice: coordination no.
 Body-centered cubic (BCC)
• The coordination number is 8. The distance
3
of nearest neighbors is 𝑎, where 𝑎 is the
2
edge length of the cube.
• Three primitive vectors are:
1
𝐚𝟏 = 𝑎 𝒙 + 𝒚 − 𝒛 ;
2
1
𝐚𝟐 = 𝑎 −𝒙 + 𝒚 + 𝒛 ;
2
1
𝐚𝟑 = 𝑎 𝒙 − 𝒚 + 𝒛 , x
2 y
where 𝒙, 𝒚, 𝒛 are unit vectors of Cartesian
coordinates.
• The shaded region is the primitive cell of
the BCC lattice. 21
*C. Kittel, Introduction to solid state physics (8th edition), Page 10
Face-centered cubic lattice: unit cell
 Face-centered cubic (FCC)

• Simplest cubic structure with six more

points in the centers of six faces.

• The unit cell of a face-centered cubic lattice,

which contains four lattice points (each
point in the vertex is shared among eight
cells, and each point in the face center is
1
shared among two cells, leading to 8 × +
8
1
6 × = 4 points), is not a primitive cell.
2

22
https://cnx.org/contents/[email protected]:EWNgknft@4/Lattice-Structures-in-Crystalline-Solids
Face-centered cubic lattice: primitive vectors
 Face-centered cubic (FCC)
• The coordination number is 12.
• Three primitive vectors are:
1
𝐚𝟏 = 𝑎 𝒙 + 𝒚 ;
2
1
𝐚𝟐 = 𝑎 𝒚 + 𝒛 ,
2
1
𝐚𝟑 = 𝑎 𝒛 + 𝒙 ;
2
where 𝒙, 𝒚, 𝒛 are unit vectors of Cartesian
coordinates.
• The shaded region is the primitive cell
of an FCC lattice.

https://www.physics-in-a-nutshell.com/article/11/close-packed-structures-fcc-and-hcp
Ventura, Michael James. Fabrication and characterisation of three-dimensional passive and active photonic crystals.
Swinburne University of Technology, Faculty of Engineering and Industrial Sciences, Centre for Micro-Photonics, 2008.
Cubic lattice: Hard sphere model
 Hard sphere model:
Atoms are thought of as being solid spheres having well-
defined diameters when describing crystal structures.

Packing factor, which is the maximum proportion of the

available volume that can be filled with hard spheres, can
be calculated by
𝑁sphere𝑉sphere
𝑃𝐹 =
𝑉unit cell

https://www.researchgate.net/publication/251151526_Modelling_of_particle_wall- 24
deposition_in_two-phase_gas-solid_flows/figures
Calculation of packing factor
For a simple cubic lattice, the packing factor can be
calculated by
4 3
𝑁sphere𝑉sphere 𝑁 ⋅ 𝜋𝑟
𝑃𝐹 = = 3
𝑉unit cell 𝑎3
𝑎
where 𝑟 = is the maximum radius of hard spheres.
2
2
In an FCC lattice, the sphere radius 𝑟 = 𝑎, the number of
4
spheres 𝑁sphere = 4, the packing factor is
3
4 3 4 2
𝑁 ⋅ 3 𝜋𝑟 4 ∙ 3𝜋 4 𝑎
𝑃𝐹 = 3
= ≅ 0.74
𝑎 𝑎3

25
http://poh101.com/index3/mems.html
Cubic lattice
SC BCC FCC
Volume, unit cell 𝑎3 𝑎3 𝑎3

Lattice points per cell 𝑁 1 2 4

Volume, primitive cell 𝑎3 1 3

𝑎
1 3
𝑎
2 4

Number of nearest neighbors 6 8 12

Nearest neighbor distance 𝑑 𝑎 3

𝑎 = 0.866𝑎
2
𝑎 = 0.707𝑎
2 2

Number of nearest neighbors 12 6 6

Second neighbor distance 2𝑎 𝑎 𝑎

𝜋 𝜋 𝜋
Packing factor 6
= 0.524
8
3 = 0.680
6
2 = 0.740 26
Hexagonal close-packed structure (HCP)
• Spheres are arranged in a single closest-packed layer A by
placing each sphere in contact with six others in a plane. A
similar layer B may be added by placing each sphere of B in
contact with three spheres of layer A, as shown below. A third
layer may be added in two ways: repeating the position of layer
A (ABAB…), or over the C holes (see slide 29) in layer A that are
not occupied by B (ABCABC…).

The stacking manner of closest-packed layers The hexagonal close-packed structure

https://chem.libretexts.org/Bookshelves/Physical_and_Theoretical_Chemistry_Textbook_Maps 27
https://en.wikipedia.org/wiki/File:Hexagonal_close_packed.svg
Hexagonal close-packed structure (HCP)

28
https://pt.slideshare.net/kareemtharaa/crystallization-42686987
Hexagonal close-packed structure (HCP)
• If the sequence of close-packed layers goes in ABABAB…, the
structure is hexagonal close-packed. If it goes in ABCABCABC…,
the structure is face-centered cubic.
• The packing factors of HCP and FCC are both 0.74. No
structure has denser packing.

close-packed layer HCP structure FCC structure

29
https://www.slideserve.com/moana/crystalline-solids
Hexagonal close-packed structure (HCP)
Problem: How to calculate the packing factor of an HCP structure?
1 1
The total number of atoms in the hcp cell is 3 + × 2 + × 6 ×
2 6
2 = 6. As the cell is filled by hard spheres, the side length and
sphere radius have the relation of 𝑎 = 2𝑟, and the height 𝑐 =
2
4 𝑟. So the packing factor is
3
4 3 4 3
𝑁sphere𝑉sphere 6 ∙ 3 𝜋𝑟 6 ∙ 3 𝜋𝑟
𝑃𝐹 = = = ≅ 0.74
𝑉unit cell 3 3 2
𝑎 𝑐 3 3 (2𝑟)2 2 ∙ 4𝑟
2 2 3

The hexagonal close-packed structure 30

https://www.princeton.edu/~maelabs/mae324/glos324/hcp.htm
Some materials examples: CsCl
 Cesium Chloride Structure
• The unit cell of the CsCl crystal is a body-centered cubic
structure. It consists of one simple cubic formed by Cl- ions and
one Cs+ ion in the body center, or equivalently one cubic formed
by Cs+ ions and one Cl- ion in the body center.
• The Bravais lattice of the CsCl crystal is a simple cubic lattice
+ - 111
with the basis consisting of a Cs ion at 000 and a Cl ion at .
222
The displacement of two ions is one-half the body diagonal.
• Each ion has eight nearest neighbors of the opposite kind.

The unit cell of Cesium Chloride

31
*C. Kittel, Introduction to solid state physics (8th edition), Page 14
Some materials examples: NaCl
 Sodium Chloride Structure
• The Bravais lattice of the NaCl crystal is a face-centered cubic
+ 111
lattice with the basis consisting of one Na ion at and one Cl-
222
ion at 000 separated by one-half the body diagonal of a unit cube.
• There are four units of NaCl in each unit cube (where?).
• Each ion has six nearest neighbors of the opposite kind.

The structure of NaCl crystal

32
http://www.metafysica.nl/turing/preparation_3dim_2.html
Diamond structure
• The diamond structure consists of two interpenetrating face-
centered cubic Bravais lattices, displaced along the body diagonal
of the cubic cell by one quarter the length of the body diagonal,
so that it contains 2 × 4 = 8 lattice points.
• This structure is a complex structure that would not be a Bravais
lattice if every atom in the structure is considered as a lattice
point. However, it becomes an FCC Bravais lattice if each pair of
111
two nearest, identical atoms is considered as a basis, 000 and .
444
• Each atom has 4 nearest neighbors and 12 next nearest neighbors.
The four nearest neighbors of each point form the vertices of a
regular tetrahedron (四面體).
Diamond Structure:
3D & 2D projection
(how to draw the
primitive cell?)
33
https://www.tf.uni-kiel.de/matwis/amat/iss/kap_4/illustr/i4_2_1.html
Some materials examples: ZnS
 Cubic Zinc Sulfide Structure
• The cubic zinc sulfide crystal is a diamond-like structure.
• The Bravais lattice of ZnS crystal is face-centered cubic lattice with
111
the basis consisting of one Zn atom at 000 and one S atom at
444
separated by one quarter of the body diagonal of the unit cube.
• There are four ZnS molecules per conventional cell. The four
nearest neighbors of each atom form the vertices of a regular
tetrahedron.
• There are only 4 lattice points (or molecules) in cubic
zinc sulfide crystal as the basis consists of two atoms
(while a diamond structure has 8 lattice points).
• The atomic arrangement of ZnS crystal is a diamond-
like structure but actually a face-centered cubic Crystal structure of
lattice so that it could have a primitive cell as a face- cubic zinc sulfide
centered cubic lattice does. 34
https://en.wikipedia.org/wiki/zinc_sulfide
Index systems for crystals: point coordinate
 Point coordinates
The coordinates ℎ 𝑘 𝑙 of a point in a crystal are a set of numbers
(not necessarily being integers) indicating its location, referred to
the axes.

Some point coordinates in a cubic crystal

35
Index systems for crystals: direction indices
 Indices of the direction
The indices [ℎ𝑘𝑙] of a direction in a crystal are a set of the smallest
integers (in a square bracket) that have the ratio of the coordinates
of a vector in the desired direction, referred to the axes.
The 𝐚𝟏 axis (i.e. the a axis in the below schematics) is the [100]
ത direction.
direction, and the −𝐚𝟐 axis (i.e. the b axis) is the [010]

Illustration of directional indices in two different crystal lattices

36
Index systems for crystals: Miller indices
 Indices of the plane (Miller indices)
Lattice constant: the physical dimension of a conventional unit
cell for a crystal lattice (not necessarily of a primitive cell).
The orientation of a crystal plane is specified by the indices
determined by the following rules:
• Find the intercepts of the plane on the axes in terms of the
lattice constant.
• Take the reciprocals of these numbers and then reduce to
three integers having the same ratio, usually the smallest
three integers.
• The result, enclosed in parentheses (ℎ𝑘𝑙), is called the indices
of the plane or Miller indices.

37
Index systems for crystals: Miller indices
 Indices of the plane (Miller indices)
As shown below, this shaded plane intercepts the axes 𝐚𝟏 , 𝐚𝟐 ,
𝐚𝟑 at 3𝐚𝟏 , 2𝐚𝟐 and 2𝐚𝟑 . The reciprocals of these numbers are
𝟏 𝟏 𝟏
, , . The smallest three integers having the same ratio are
𝟑 𝟐 𝟐
2,3,3, and thus the indices of the plane are (233).

The rules to specify the index of the plane

In general, the direction [ℎ𝑘𝑙] is not normal to the (ℎ𝑘𝑙) plane,

except in a cubic lattice.
38
*C. Kittel, Introduction to solid state physics (8th edition), Page 12
Index systems for crystals: Miller indices
 Indices of the plane (Miller indices)
The indices may denote a single plane or a set of parallel planes.
If a plane cuts an axis on the negative side of the origin, the
corresponding index is negative , indicated by placing a minus
ത
sign above the index: (ℎ𝑘𝑙).
Planes equivalent by symmetry may be denoted by a curly
bracket around the indices. For example, the cube faces of a
cubic crystal are (100), (010), (001), (100), (010), (001) , the set
of these cube faces is {100}.

Indices of some planes in a cubic crystal. The plane (200) is parallel to (100).
39
*C. Kittel, Introduction to solid state physics (8th edition), Page 12
Symmetry of crystal structure
 Point group
In geometry, a point group is a group of geometric symmetries
(isometries) that keep at least one point fixed in a Euclidean space.
In crystallography, a fixed number of dimensions has only a finite
number of point groups that are symmetric over some lattices or
grid of that number. Therefore, we may group together atoms that
possess the same symmetry elements and classify them according
to their symmetry. These are the so-called crystallographic point
groups.
 Crystallographic restriction theorem (晶體局限定理)
Crystals are modeled as discrete lattices, generated by a list of
independent finite translations. Because discreteness requires
that the spacings between lattice points have a lower bound, the
group of rotational symmetries for a lattice at any point must be a
finite group (有限群). 40
https://en.wikipedia.org/wiki/Point_group
Symmetry of crystal structure
 Point group
The strength of the theorem is that not all finite groups are
compatible with a discrete lattice in any dimension. Therefore, we
will only have a finite number of compatible groups.
• There are only two point groups for one spatial dimension, the
identity group (𝐶1) and the reflection group (𝐶2).
• 2D point groups can be classified into two infinite families:
(1) Cyclic groups (循环群) 𝐶𝑛 of n-fold rotation groups;
(2) Dihedral groups (二面体群) 𝐷𝑛 of n-fold rotation and
reflection groups.
• Three-dimensional point groups come into 7 infinite families of
axial or prismatic groups, and 7 additional polyhedral or Platonic
groups. Applying the crystallographic restriction theorem to
these groups yields 32 crystallographic point groups as shown
later. 41
Symmetry of crystal structure
 Symmetry operation
If an object can be (1) rotated, (2) reflected, or (3) inversed,
without changing the distances between its material points and so
that it comes into self-coincidence, then that object is symmetric.
A transformation of type (1), (2) or (3) or combinations thereof that
preserve distances and bring the object into coincidence is called a
symmetry operation.
There are three basic symmetry operations (point groups): rotation,
reflection and inversion.
The Bravais lattice of a crystal also has to fulfill the translation
symmetry.

42
Rotation symmetry Reflection symmetry Inversion symmetry Translation symmetry
Symmetry operation: Transformation matrix
Generally, a transformation can be given by a transformation matrix
𝐴, where
𝑥′ 𝑥 𝑎11 𝑎12 𝑎13 𝑥
𝑦 ′ = 𝐴 𝑦 = 𝑎21 𝑎22 𝑎23 𝑦
𝑧′ 𝑧 𝑎31 𝑎32 𝑎33 𝑧
Obviously, either symmetry operation 𝐴 or 𝐴−1 (inverse matrix)
should preserve lengths of vectors and angles in between.
If we take 1 0 0 𝑇 (transposed matrix, row vector -> column
vector), 0 1 0 𝑇 and 0 0 1 𝑇 as examples, we have
2 2 2
𝑎11 + 𝑎21 + 𝑎31 = 12 + 02 + 02 = 1

2 2 2
𝑎12 + 𝑎22 + 𝑎32 = 02 + 12 + 02 = 1

2 2 2
𝑎13 + 𝑎23 + 𝑎33 = 02 + 02 + 12 = 1
43
Therefore, each column of 𝐴 is a unit vector.
Symmetry operation: Transformation matrix
Furthermore, if we take 1 1 0 𝑇 , 0 1 1 𝑇 and 1 0 1 𝑇
as examples, we can obtain (HW: Why?)
𝑎11 𝑎12 + 𝑎21 𝑎22 + 𝑎31 𝑎32 = 0
𝑎11 𝑎13 + 𝑎21 𝑎23 + 𝑎31 𝑎33 = 0
𝑎12 𝑎13 + 𝑎22 𝑎23 + 𝑎32 𝑎33 = 0
The three columns of 𝐴 are orthogonal unit vectors. It is easy to
obtain that the rows of 𝐴 are orthogonal unit vectors, too. In other
words, 𝐴 is an orthogonal matrix. As a result (HW: Why?),
𝐴𝐴𝑇 = 𝐼 ⇔ 𝐴𝑇 = 𝐴−1 .
All 𝑛 × 𝑛 orthogonal matrices form a group, 𝑂 𝑛 , known as the
orthogonal group. In a finite-dimensional space, an orthogonal
matrix represents an orthogonal transformation.
Orthogonal transformations in two- or three-dimensional Euclidean
space (欧几里得空间) includes only rigid rotations, reflections, or
combinations of a rotation and a reflection. 44
Symmetry operation: Rotation in 2D
 Rotation (𝑶 𝒏 )
Considering a two-dimensional (2D) space, a general point 𝑃 with
the position vector 𝐫 and coordinates (𝑥, 𝑦), is rotated to the
position 𝑃′, with position vector 𝐫′ and coordinates (𝑥′, 𝑦′). From
the drawing, we find that the coordinates of the new position
vector are given by:
𝑥 ′ = 𝑟 ′ cos 𝜃 + 𝛼
𝑦 ′ = 𝑟 ′ sin 𝜃 + 𝛼
where 𝑟 ′ = 𝐫 = 𝐫′ = 𝑟.
Obviously, we have
𝑥 = 𝑟 cos 𝛼
𝑦 = 𝑟 sin 𝛼
Therefore we obtain
𝑥 ′ = 𝑟 ′ cos 𝜃 + 𝛼 = 𝑟 ′ cos 𝜃 cos 𝛼 − sin 𝜃 sin 𝛼
= 𝑟 cos 𝛼 cos 𝜃 − 𝑟 sin 𝛼 sin 𝜃 = 𝑥 cos 𝜃 − 𝑦 sin 𝜃 45
Symmetry operation: Rotation in 2D
 Rotation (𝑪𝒏)
Similarly, we have
𝑦 ′ = 𝑟 ′ sin 𝜃 + 𝛼 = 𝑟 ′ sin 𝜃 cos 𝛼 + cos 𝜃 sin 𝛼
= 𝑟 cos 𝛼 sin 𝜃 + 𝑟 sin 𝛼 cos 𝜃 = 𝑥 sin 𝜃 + 𝑦 cos 𝜃
We can thus describe the transformation operation in matrix form
as:
𝑥′ cos 𝜃 − sin 𝜃 𝑥
= 𝑦
𝑦′ sin 𝜃 cos 𝜃

The rotation matrix 𝐶(𝜃) in a two-dimensional space is:

cos 𝜃 − sin 𝜃
𝐶 𝜃 =
sin 𝜃 cos 𝜃
(HW: double check whether the rows and columns in the above
matrix are orthogonal unit vectors.)
46
Symmetry operation: Schönflies notation
 Rotation (𝑪𝒏)
Because of the crystallographic restriction theorem, there are only
1-, 2-, 3-, 4-, 6-fold rotation axes are allowed in crystals, indicating
2𝜋 𝜋 𝜋
the rotation angle must be one of 2𝜋, 𝜋, , , .
3 2 3
We denote the rotation operation with Schönflies notation 𝑪𝒏,
where 𝑛 = 1, 2, 3, 4, 6 ( named after the German mathematician
Arthur Moritz Schönflies).

Some 2D geometries
with 5 different
rotation axes.
47
https://www.tulane.edu/~sanelson/eens211/introsymmetry.htm
Symmetry operation: proof of rotation angle
 Proof of the limited number of rotation axes
Considering a Bravais lattice and the coordinates are referred to
primitive vectors, the coordinates must be integers for both the
original point and the rotated point. The coordinates of the
rotated point are given by:
𝑥 ′ = 𝑥 cos 𝜃 − 𝑦 sin 𝜃;
𝑦 ′ = 𝑥 sin 𝜃 + 𝑦 cos 𝜃;
𝑧′ = 𝑧 .
Since 𝑥 ′ + 𝑦 ′ + 𝑧 ′ must be an integer, we can obtain that
𝑥 − 𝑦 sin 𝜃 + 𝑥 + 𝑦 cos 𝜃 + 𝑧 must be an integer,
applicable to arbitrary 𝑥, 𝑦, 𝑧. Set 𝑥 = 𝑦 = 𝑧 = 1, then 1 +
2 cos 𝜃 should be an integer.
As −1 ≤ 1 + 2 cos 𝜃 ≤ 3, the integer must be one of -1, 0, 1, 2,
3. Correspondingly, the allowed rotation angles are limited to
2𝜋 𝜋 𝜋
those of 2𝜋, 𝜋, , , .
3 2 3
Symmetry operation: mirror/reflection
 Mirror (Reflection) (𝝈/𝒎)
Mirror operation is a point by point transformation as drawn
below.
The mirror operation (x-y mirror plane) can be described in the
matrix form as:
𝑎11 𝑎12 𝑎13 1 0 0
𝜎 = 𝑎21 𝑎22 𝑎23 = 0 1 0
𝑎31 𝑎32 𝑎33 0 0 −1

2D and 3D mirror(reflection) operations.

*We can use both 𝝈 and 𝒎 to denote the mirror operation.
49
*Bianchi, I. et al., Two Basic Components of Adults’ Perception and Mental Representation of Symmetry. Symmetry 9, 128.
Symmetry operation: inversion symmetry
 Inversion (𝒊)
In this operation, lines are drawn from all points on the object
through a point in the center, called an inversion center. We
make mirror operation for every point with respect to the plane
passing the inversion center and perpendicular to the line
linking the points and the inversion center.
The inversion operation can be described in the matrix form as:
𝑎11 𝑎12 𝑎13 −1 0 0
𝑖 = 𝑎21 𝑎22 𝑎23 = 0 −1 0
𝑎31 𝑎32 𝑎33 0 0 −1

2D and 3D inversion operations and their centers of inversion. 50

*https://www.pinterest.com/pin/336573772128273235/ **http://www.pci.tu-bs.de/aggericke/PC4e/Kap_IV/i-Op.html
Symmetry operation: inverse
 Inversion (𝒊)
As the 𝐶2 operation in the three-dimension space goes in the
matrix form:
cos 𝜋 −sin 𝜋 0 −1 0 0
𝐶2 = sin 𝜋 cos 𝜋 0 = 0 −1 0
0 0 1 0 0 1
We can obtain the relation of three basic symmetry operations
as:
−1 0 0
𝑖= 0 −1 0 = 𝜎𝐶2
0 0 −1
(HW: carry out matrix multiplication to prove the above
conclusion.)

51
Symmetry operation: combined operation
 Combined operation
In some crystals, there are symmetry operations based on a
combination of basic symmetry operations.
2𝜋
The denotation S𝑛 or 𝑛ത describes the rotation of followed by
𝑛
a reflection symmetry operation with respect to the plane
perpendicular to the rotation axis.
S𝑛 = 𝐶𝑛 𝜎
• The existence of combined symmetry operations in crystals
does not mean the existence of basic operations consisting
of the combined operations.
• On the contrary, the existence of basic operations means the
existence of their combined operations.

52
Symmetry operation: Schönflies notations
𝐶𝑛 : 𝑛-fold rotation axis.
𝐶𝑛ℎ : 𝐶𝑛 with the addition of a mirror (reflection) plane perpendicular to the axis of
rotation (horizontal plane).
𝐶𝑛𝑣 : 𝐶𝑛 with the addition of 𝑛 mirror planes containing the axis of rotation (vertical
planes).
𝑆𝑛 : 𝑛-fold rotation-reflection axis.
𝐶𝑛𝑖 : Rotoinversion axis. For odd 𝑛, 𝐶𝑛𝑖 = 𝑆2𝑛 ; for even 𝑛, 𝐶2𝑛𝑖 = 𝑆𝑛 = 𝐶𝑛ℎ .
𝐷𝑛 : 𝑛-fold rotation axis plus 𝑛 2-fold axes perpendicular to that axis.
𝐷𝑛ℎ : 𝐷𝑛 with a horizontal mirror plane and, therefore, also 𝑛 vertical mirror planes
each containing the 𝑛-fold axis and one of the 2-fold axes.
𝐷𝑛𝑑 : 𝐷𝑛 with 𝑛 vertical mirror planes passing between the 2-fold axes.
𝑇: rotation axes of a tetrahedron (three 2-fold axes and four 3-fold axes).
𝑇𝑑 : 𝑇 with diagonal mirror planes, each of which contains only one 2-fold axes and
passes between two other 2-fold axes.
𝑇ℎ : 𝑇 with three horizontal mirror planes, each of which contains two 2-fold axes and
is perpendicular to the third 2-fold axes.
𝑂: rotation axes of an octahedron or cube.
53
𝑂ℎ : 𝑂 with horizontal mirror planes. https://en.wikipedia.org/wiki/Schoenflies_notation
Symmetry operation
 32 crystallographic point groups
Crystal System Point groups (Schönflies notation)

Triclinic 𝐶1 , 𝐶𝑖 (𝑆2 )
Monoclinic 𝐶2 , 𝐶1ℎ , 𝐶2ℎ
Orthorhombic 𝐷2 , 𝐶2𝑣 , 𝐷2ℎ
Tetragonal 𝐶4 , 𝑆4 , 𝐶4ℎ , 𝐷4 , 𝐶4𝑣 , 𝐷2𝑑 , 𝐷4ℎ
Trigonal 𝐶3 , 𝑆6 , 𝐷3 , 𝐶3𝑣 , 𝐷3𝑑
Hexagonal 𝐶6 , 𝐶3ℎ , 𝐶6ℎ , 𝐷6 , 𝐶6𝑣 , 𝐷3ℎ , 𝐷6ℎ

Cubic 𝑇, 𝑇𝑑 , 𝑇ℎ , 𝑂, 𝑂ℎ

(HW: In combination with the 3D Bravais lattices in slide 17,

carry out some symmetry operations in slides 53 and 54. 54
Symmetry operation
 Neumann’s Principle
If a crystal is invariant with respect to certain symmetry operations,
any of its physical properties must also be invariant with respect to
the same symmetry operations, or otherwise stated, the symmetry
operations of any physical property of a crystal must include the
symmetry operations of the point group of the crystal.
 Tensor descriptions of physical properties of a crystal
If we have a symmetry operation 𝐴 = 𝑎𝑖𝑗 and a second-order
tensor 𝑇 as one of its physical properties, we can obtain
𝑇 ′ = 𝐴𝑇𝐴−1 = 𝐴𝑇𝐴𝑇 = 𝑇
i.e.
𝑇𝑖𝑗 = ෍ 𝑎𝑖𝑚 𝑎𝑗𝑛 𝑇𝑚𝑛 ,
𝑚𝑛
which requires a specific relationship between 𝑇𝑖𝑗 . 55
Symmetry operation example: Permittivity
For an anisotropic material, its permittivity is given by a 2nd-order
tensor
𝜀11 𝜀12 𝜀13
𝜀 = 𝜀21 𝜀22 𝜀23
𝜀31 𝜀32 𝜀33
For a cubic crystal, according to Neumann’s principle, the
permittivity is preserved after a specific symmetry operation.
If we rotate the crystal by 180 degrees around 𝑥 axis, we can obtain
1 0 0
𝜀 = 𝐴1 𝜀𝐴1𝑇 , 𝐴1 = 0 −1 0
0 0 −1
from which we can obtain 𝜀12 = 𝜀13 = 𝜀21 = 𝜀31 = 0 (HW: Why?),
that is
𝜀11 0 0
𝜀 = 0 𝜀22 𝜀23
0 𝜀32 𝜀33 56
Symmetry operation example: Permittivity
If we rotate the crystal by 90 degrees around 𝑧 axis, we can obtain
0 −1 0
𝜀 = 𝐴2 𝜀𝐴𝑇2 , 𝐴2 = 1 0 0
0 0 1
from which we have 𝜀23 = 𝜀32 = 0 and 𝜀11 = 𝜀22 (HW: Why?), that is
𝜀11 0 0
𝜀 = 0 𝜀11 0
0 0 𝜀33
If we rotate the crystal by 90 degrees around 𝑦 axis, we can obtain
0 0 1
𝜀 = 𝐴2 𝜀𝐴𝑇2 , 𝐴2 = 0 1 0
−1 0 0
from which we can obtain 𝜀11 = 𝜀33 (HW: Why?), that is
𝜀0 0 0
𝜀 = 0 𝜀0 0 = 𝜀0 𝛿𝑖𝑗
0 0 𝜀0 57