COMPUTATION OF MATERIALS

VIA DENISTY FUNCTIONAL

THEORY METHODS
by

Dr Rashid Ahmed

1st Hands-on- workshop

Department of Physics, Faculty of Science, Universiti Teknologi Malaysia 23-
27 July 2017
 Pillars of scientific
investigations
 Experiment

 Theory

 Scientific

Simulation/Computation

 Big Data

Use math and simulation

to go from data to laws

2
 Scientific Simulation/Scientific Computing
• Scientific computing/computational science has been the stuff of fantasy
for generations (Galileo, Richardson, etc.). Because it was known since
1800s, many nonlinear problems in physics were not possible to solve within
traditional hand methods.

• Although scientific computation had been taken start with the work of
French naturalist Comte de Buffon, real progress started with John
Atanasoff paper from 1940 “Computing Machine for the Solution of
Large Systems of Linear Algebraic Equations” and with the
development modern computers.

• Late 1940s a few began to think that nonlinear problems of physics

might be cracked with computers.
 Scientific Simulation/Scientific Computing.
Computational science can be defined as an interdisciplinary
approach that uses concepts and skills from the science, computer
science and mathematics disciplines to solve complex problems in
the study of various phenomena.

Applied Science
Mathematics
Computational
Science

Computer
Science
 Scientific Simulation/Scientific Computing

 Theoretical experiments/virtual experiments/thought

experiments not an “other” but a hybrid and a platform for
integration of both theory and experiment

 a means of scientific discovery that employs a computer

system to simulate a physical system according to laws derived

from theory and experiment

 Use mathematical methods to go from theory to numerical

prediction
 Applied
Physics
radiation transport
supernovae
Environment Engineering
global climate crash testing
contaminant aerodynamics
transport

Biology Lasers & Energy

combustion
drug design Scientific ICF
genomics
Simulation

In these, and many other areas, scientific simulation is an

important complement to experiment.
6
 Applied
Physics
radiation transport
supernovae
Environment Engineering
global climate crash testing
contaminant aerodynamics
Experiments transport
controversial
Biology Lasers & Energy
combustion
drug design Scientific ICF
genomics
Simulation

In these, and many other areas, simulation is an

important complement to experiment.

7
 Applied
Physics
Experiments radiation transport
dangerous supernovae
Environment Engineering
global climate crash testing
contaminant aerodynamics
Experiments transport
controversial
Biology Lasers & Energy
combustion
drug design Scientific ICF
genomics
Simulation

In these, and many other areas, simulation is an

important complement to experiment.

8
 Experiments prohibited or
impossible
Applied
Physics
Experiments radiation transport
dangerous supernovae
Environment Engineering
global climate crash testing
contaminant aerodynamics
Experiments transport
controversial
Biology Lasers & Energy
combustion
drug design Scientific ICF
genomics
Simulation

In these, and many other areas, simulation is an

important complement to experiment.

9
 Experiments prohibited or
impossible
Applied
Physics
Experiments radiation transport Experiments difficult to
dangerous supernovae instrument
Environment Engineering
global climate crash testing
contaminant aerodynamics
Experiments transport
controversial
Biology Lasers & Energy
combustion
drug design Scientific ICF
genomics
Simulation

In these, and many other areas, simulation is an

important complement to experiment.

10
 Materials Science
• Materials science is an interdisciplinary and revolutionary field,
and the world is in the middle of a materials revolution.

• Materials science and engineering has transformed every aspect of

modern living.

• Advances in materials have preceded almost every major

technological leap since the beginning of civilization.

• Advances in engineered materials are crucial to the continued

vitality of countless industries.
• .To achieve these goals, understanding of scientific

fundamentals of materials, their design, their structure and

properties relation, their processing for real world applications

and science behind them etc is crucial.

• However to obtain optimized results for material engineering

over a wide range of operating environments, and length and

time scales is an expensive, time taking and tedious job.

 Computational Material Science
• Computational materials science, is one of the fastest growing
areas within the field of physics, chemistry and materials science.

• In the past, advances in materials were accomplished by

extensive laboratory testing combined with a healthy dose of
guesswork, a time-consuming and often costly approach.

• However with the development high performance computing

facilities like, supercomputers/parallel processing techniques,
multiscale computational codes, this trend is vastly changing
towards simulations.
13
 Solid is Quantum Many Body Problem
• Solid is a collection of heavy, positively charged (nuclei)
particles and lighter, negatively electrons.

• There are 1023 valence electron in each cubic centimeter

contributing to the bonding of a typical solid.

• These particles are so light, QM is needed. This makes

solids, a quantum many body/N-particles problems.

• Therefore, it requires a many body Schrödinger wave

equation.
 Theory of Electronic Structure
To find the wave function and other properties one
must solve the Schrödinger equation
  2 2 
i  (r , t )    (r , t )
t 2m x 2

  2  2  2  2 2 2  
i  ( r , t )    ( r , t )    2  2  2  ( r , t )
t 2m 2m  x y z 
 Why Approximations
• The quantum mechanical wave-function contains, in principle, all
the information about a given system.

• For the case of a simple 2-D square potential or even a hydrogen

atom, one can solve the Schrödinger equation exactly in order to
get the wavefunction of the system.

• And then determine the allowed energy states of the system.

• Unfortunately it is impossible to solve the Schrödinger equation

for a N-body system.

http://newton.ex.ac.uk/research/qsystems/people/coomer/dft_intro.html
 Why Approximate Methods?
• Because we are dealing with N+ZN particles and 3P+3N
coupled degrees of freedom.

• Schrödinger equation is impossible to be decoupled into a

set of independent equations.

• So it is impossible to solve such type of many body

Schrödinger equation exactly.

• To solve such problems, evidently, one must involve some

approximations to render the problem soluble albeit tricky.
Quantum mechanical approach for
solving many body problem
 The exact Hamiltonian For N-particle system is

Coulomb Coulomb Coulomb

K.E Operator for K.E Operator for interaction b/w interaction b/w interaction b/w
electrons nuclei electrons and nuclei and other nuclei and
other electrons nuclei electrons

The exact solution of such a problem is not possible. 18

 Born-Oppenhemier Approximation
• This approximation is based on the fact that the nuclei are much
heavier and therefore much slower than the electrons.
• Also this approximation is to treat the nuclei at fixed positions
and assume the electrons to be in instantaneous equilibrium with
them.
• After having applied this approximation on the nuclei do not
move any more, their kinetic energy is zero and the first term
disappears and third term reduces to constant.
• Also Born Oppenheimer approximation provides a method to
separate the ionic degrees of freedom from the electronic ones.
This shows that we are left with the problem of solving the
many-body electronic Schrödinger equation for a set of fixed
nuclear positions.
 Born – Oppenheimer approximation

By applying Born-Oppenheimer approximation, many body

Schrodinger equation transfer to the following form;

He  Te  Vee  Veion

S. Cottenier, Density Functional Theory and the Family of (L)APW-methods: a step-by-step introduction
, (Instituut voor Kern- en Stralingsfysica, K.U.Leuven, Belgium), 2002, ISBN 90-807215-1-4

20
 Electronic Many Particle Hamiltonian
 N 2K N,N
1 
     k   ( r1 ,........rN )
(0)
( rk ) W ( rk , r ' )
 k 1 2m k k ' 2 

e2
W(r, r ' ) 
4 0 r  r '

M
 ( 0)
(r )   Iion ( R I  r )
I 1

Problem is still many (for a typical solid : 1023) degrees of

freedom and is difficult to solve with analytical approaches. In
such situation, First principles methods are useful.
 Why ab initio Methods?
• Ab initio or first principles calculations means that there are no
parameters to adjust the result except for the information about the
constituent atoms, the atomic number.

• The calculations are based on the fundamental laws of physics rather

than on particular model considerations.

• If a system shows unusual phenomena that are not yet understood,

often relies on ab-initio calculations.

• Ab initio methods do not require any experimental knowledge to

carry out such calculations.
 Why ab initio Methods?
• Computational materials science based on ab initio

methods provides insight into the properties and

processes in atoms, molecules and condensed phases,

ranging from bulk crystals to surfaces, interfaces and

nanostructures.
 Ab initio Methods
• Hartree Fock Method
1. This is based on a wave function description.
2. Exchange is treated exactly but correlation effects are ignored by
definition.
3. This method is very useful for atoms and molecules but is less
accurate for solids

• Density Functional Theory

1. This is used from a single atom to a complex system containing
hundreds of atoms in a unit cell including transition metals.
2. This is a powerful and elegant electronic structure method for
calculating ground state properties of systems of interacting
electrons
3. In this method, many-body problem of interacting electrons and
nuclei is mapped onto a one- electron reference system with the
same density as the real system.
4. It is based on the electron density rather than on the wave functions
and treats approximately both exchange and correlation.
 Density Functional Theory
• DFT is based on two important theorems of Hohenburg and
Kohn;

• Theorem I : For any system of interacting particles in an external

potential Vext(~r), the density is uniquely determined
(in other words, the external potential is a unique functional of
the density).

• Theorem II : A universal functional for the energy E[n] can be

defined in terms of the density. The exact ground
state is the global minimum value of this functional.
• Hohenburg and Kohn theorem states that “the density
of any system determines all ground-state properties of
the system”.

• It means that the total ground state energy of a many-

electron system is a functional of the density.

• So, if we know the electron density functional, we

know the total energy of our system.
 Density Functional Theory
• Thus a simplest definition of DFT is : A method of obtaining an
approximate solution to the Schrödinger equation of a many-body
system.
Density Functional Theory
The ground-state energy from Schrödinger’s equation is a
unique functional of the electron density.

The density contains as much information as the wave

function does, i.e. everything you could possibly know
about an atom, molecule or solid.

The concept of DFT, the 3N spatial coordinates

associated to a many body system has been reduced to
3 dimensions.

Nobel Prize in Chemistry 1998

28
 The DFT calculations are based on the quantum mechanics, but instead
of electron wave function it utilizes the electrons charge density as a
basic variable for solving the many body problems.

DFT calculations

Electrons
Wave function
density

Quantum Mechanics at work

3N
3 Variable
Variables
29
 The Kohn-Sham equation
 K-S equations turned DFT to one of the practical tool for the calculation of ground
state density.
 The Kohn-Sham total energy equation is,

No theoretical tool is
determined and hence
approximations are
needed to calculate this
energy

Coulomb energy
External Exchange and
Kinetic energy of due to electron-
potential due to correlation
electrons electron
nucleus energy term
interaction

W. Kohn, L.S. Sham, Phys. Rev. A, 140 (1965) 1133.

30

7/27/2017
 Kohn Sham Equations
To find the stationary point, we do variations at fixed N   dr n (r ) ,which
leads to
E (0)
 (Lagrange parameter)
n (r )

If we write the density as a sum over single-particle functions.

N
n (r )     j,k (r )
2

j1 kBZ

The variational principle E (0)[ * ] /  * ( r )  0 leads to the Kohn-Sham

equations
 2 
   Veff [n ](r )  j,k (r )   j,k j,k (r )
 2m 
e 2 n (r ' ) E XC [n ]
Veff [n ](r )  
With the effective potential ( 0)
(r )   dr ' 
4 0 r  r ' n (r )
RI
 How are Kohn Sham Equations Solved?
• These equations are solved iteratively
• Guess a starting charge density
• Set up a Hamiltonian for this charge density
• Diagonalize this approximate Hamiltonian
• Use the eigenvalues and wave functions to set up
a new charge density etc
• This procedure continues in a series till the final
density is obtained which is consistent with
Hamiltonian.
Description of the computational procedure
for the total energy calculation

Conjugate gradient
method
Molecular-dynamics
method
 DFT at Glance
• Density functional theory (DFT) has revolutionized in the

development of the quantum chemistry in the last 20 years.

• DFT allows to get information about the energy, the

structure and the molecular properties of materials at lower

costs than traditions approaches based on the wave function

use.
 DFT
• The Density Functional Theory was introduced in two seminal papers in the
60's:
• 1. Hohenberg-Kohn (1964): ~ 4000 citations
• 2. Kohn-Sham (1965): ~9000 citations
• The following figure shows the number of publications where the phrase
density functional theory appears in the title or abstract (taken from the ISI
Web of Science).

Density-functional theory (DFT) has

been established as the most widely
used and powerful tool for
investigations of physical properties
of materials.
 DFT
• DFT is presently the most successful (and also the most promising) approach to
compute the electronic structure of matter.

• Its applicability ranges from atoms, molecules and solids to nuclei and quantum
and classical fluids.
• In its original formulation, the density functional theory provides the ground state
properties of a system, and the electron density plays a key role.
• An example: DFT predicts a great variety of molecular properties: molecular
structures, vibrational frequencies, atomization energies, ionization energies, electric
and magnetic properties, reaction paths, etc.
• The original density functional theory has been generalized to deal with many
different situations: spin polarized systems, multicomponent systems such as nuclei
and
• Electron hole droplets, free energy, superconductors with electronic pairing
mechanisms, relativistic electrons, time-dependent phenomena and excited states,
bosons, molecular dynamics, etc.
Applications in quantum chemistry.
Applications in quantum chemistry.
Applications in solid state physics.
Applications in Molecular Electronics.
Applications in conductance of a hydrogen
molecule.
 DFT Based Computational
Materials Science Methods
• Two DFT based approaches frequently in use for solving Kohn-
Sham equations, are pseudo-potential and full potential.
Pseudo-potential Methods (PP) & Full Potential
Methods (FP)
Pseudo-potential Methods (PP)
• The key idea behind this approach is “to replace the strong coulomb potential of
the nucleus and effects of core electron with an effective ionic potential”
• because only the valence electrons of atoms contribute to the bonding and other
properties of a system, where as core electrons just screen the charge of the
atomic nucleus.
• For calculating properties of molecules and solids, pseudo-potential calculated
in atomic calculation can be used as core states remains the same.
Full Potential Methods (FP)
• Pseudo-potential methods are excellent to measure valence electron properties.
• If we are interested in the properties dependent on the performance of core
electrons like core level excitations or we are in search of more efficient
unbiased basis set than plane waves that does not require pseudo-potential we
move to full potential methods.
• The main idea of full potential methods is to take into account all electrons of
an atom. There is no approximation used in choosing the form of a potential.
 Popular Codes
• VASP
• Castep
• ABINIT
• CP2K
• Siesta
• Quantum expresso
• EXCITING
• WIEN2k
• Gaussian
• Dmol
• Many others
Computational Codes used
within our group
• WIEN2k

• CASTEP

• DMOl

• Siesta

• Gaussian etc
WIEN2k
 Full-Potential Linearised Augmented Plane
Wave method and WIEN2k
• This is one of the more accurate method of solving Kohn-Sham single
particle equations for electronic band structure as implemented in
WIEN2k code
• This method is based on the so called muffin-tin model of the crystal
potential.
• WIEN2k is an all-electron scheme including relativistic effects
• The method is universal that can treat all materials and their
properties from first principles.
• In Wien2k a good web-based graphical user interface, called
w2web is used. Relativistic effects are taken into account within
the scalar-relativistic approximation.
 WIEN2k Package and w2web GUI
(graphical user interface)
• Structure generator
– space group selection
• Step by step initialization
– symmetry detection
– automatic input generation
• SCF calculations
– Magnetism (spin-polarization)
– Spin-orbit coupling
– Forces (automatic geometry
optimization)
• Guided Tasks
– Energy band structure
– DOS
– Electron density
– X-ray spectra
– Optics
SOME CALCULATIONS OF
WEIN2k
Al doped ZnO: Potential candidate for
optoelectronic devices
 TCOs based on ZnO
Doped ZnO is promising candidate for TCOs because pure ZnO is:
 Non toxic
 Can be easily grown with ordinary instrumentations.
 Direct and wide band gap of energy Eg ≈ 3.44 eV.
 Transparent materials.
 Cheap and abundantly available in the market.

53
 Al doped ZnO based TCOs in zinc-
blende phase
 Al:ZnO, is of particular importance for its low growth
temperature, high solubility of Al dopant, chemical stability, low
resistivity, and high transparency in the visible spectral range [1-
2]
 The literature available on Al: ZnO system is mostly focused on
wurtzite phase, and to the best of our knowledge no study has
been done in its zinc-blend phase.

• [1] H. Liu, V. Avrutin, N. Izyumskaya, Ü. Özgür and H. Morkoç,Transparent conducting oxides for
electrode applications in light emitting and absorbing devices Superlattices and Microstructures 48
(5), 458-484 (2010).
• [2] P. Chen, W. Lai, L.-C. Peng, C. Kuo, C.-L. Yeh, J. Sheu and C. Tun,GaN-based LEDs with AZO: Y upper
contact Electron Devices, IEEE Transactions on 57 (1), 134-139 (2010).

54
 Structural Properties
ZnO Al:ZnO (12.5% Al)

The constructed supercell of ZnO with 2x2x1 configuration containing 32 atoms

ZnO Al:ZnO
Energy E0(eV) -3742.7950 -3354.4872

Lattice constant (Å) 4.58 4.55

Lattice constant (Å) Al:ZnO is less

than ZnO because of smaller
7/27/2017 Al(125nm) atomic radius than Zn(135) 55
 Electronic Structure of ZnO
Electronic Strucutre Density of States

 Energy gap calculated with GGA= 1.20 eV.

 The calculated energy gap value of ZnO is remarkably smaller than the experimentally
measured because of inherent self interaction (SIC) problem in GGA.
This SIC problem strongly underestimate the bind energies of d-electrons and wrongly
shift the d-band at energies closer to Fermi level.
Zn–d electrons are centered at -5 eV higher than real position (-7 eV)[]
Energy gap calculated with GGA+U= 2.7 eV.
R.A. Powell,W.E. Spicer, J.C. McMenamin, Phys. Rev. Lett. 27 (1971) 97.
K. Ozawa, K. Sawada, Y. Shirotori, K. Edamoto, J. Phys. Condens. Matter 17 (2005) 1271.
56
 Electronic Structure of Al:ZnO
GGA GGA+U

 GGA calculated electronic structure shows that Fermi level passes through C.B, showing
metallic character of Al:ZnO system.
 V.B is shifted to more negative energies.
 By applying U-parameter, the C.B minimum is shifted to close to Fermi level and system
shows nearly semi-mettallic n-type nature.
 In the presence of Al impurities, binding energies of Zn-d electrons has raised. d-
electrons are centered at -7 eV.
 Al-p electrons appear in V.B with a weak effect in energies -9 eV to -4.76 eV and are
major in C.B. 57
Ni doped ZnO: Potential candidate for
spintronic devices
Diluted Magnetic Semiconductors
Semiconductors are used for charge Magnetic materials are used
transfer in semiconducting devices. for data storage.

Semiconductor material Magnetic material

High frequency devices Integrated circuits Harddisk

59
 Diluted Magnetic Semiconductors
DMSs are the semiconducting materials where the host cation has
been substitutionally replaced by Magnetic ions [1].

Semiconductor material Magnetic material

Host Cation
Anion [1]. H. Ohno, Science 281, 951 (1998).
Magnetic ion
 Ni doped ZnO

Unit cell structure of Zn0.75Ni0.25O in zinc-

blende ZnO. Energy versus volume curve

Zn0.75Ni0.25O Calculated Values

Optimized Volume (V0) 628.9283 a.u3
Total Energy (E0) -14416.069 Ry
Lattice Constant (a) 4.532 A0
Bulk Modulus 148.80 Gpa
61
 Spin Polarized Band Structure

Transition of electrons takes place through Fermi level in both up and

down spin polarized band structures.

62
 Spin Polarized Density of States

 Zn-3d states are mostly originated from -7.5 to -4.8 eV.

 From -4.8 eV to -2.3 eV the states are contributed by Zn-3d and O-2p states
together with a weak effect on Ni-3d states.
 Incase of up spin DOS the peaks in energy range -2 to -1 eV are due to Ni-3d
states together with O-2p states.
 The d-band of Ni splits into eg and t2g states because of O-atoms.

63
Cr doped ZnO: Potential candidate for
spintronic devices
 Cr doped ZnO

Super cell of Zn0.75Cr0.25O in zinc-blende phase. Energy versus volume curve

Zn0.75Cr0.25O Calculated Values
Optimized Volume (V0) 656.6 a.u3 Smaller than ZnO lattice
Total Energy (E0) -13480.65 Ry constant(4.60A0) because of
difference in the ionic radii
Lattice Constant (a) 4.59 A0 of Zn(0.74A0) and
0
Cr(0.73A )
Bulk Modulus 123.14 Gpa
65
 Spin Polarized Band Structure

 States are passing through Fermi level in both up and down spin polarized
band structures.
66
 Spin Polarized Density of States

 Zn-3d states are mostly appear from -8.5 to -3.1 eV together with a week effect of
O-2p states.
 Zn-3d band exhibit a mirror symmetry.
 For up spin channel the DOS in energy range -1.4 to 0.23 are originated mainly from
Cr d- states.
 For spin down case Cr d-states appear from 1.71 to 2.62 eV.
67
 Spin Polarized Cr-d band

 The d-band of Cr splits into eg and t2g states because of O-atoms.

 The crystal field splitting energies calculated for up and down spin states
are 1.08 eV and 0.38eV respectively.
68
 Thank you

7/27/2017 69