100% found this document useful (1 vote)
349 views31 pages

Density Functional Theory

The document provides an introduction to density functional theory (DFT). It discusses the key concepts of DFT including the Hohenberg-Kohn theorems, the Kohn-Sham approach, and solving the Kohn-Sham equations. It also describes common approximations to the exchange-correlation functional such as the local density approximation (LDA) and generalized gradient approximation (GGA), as well as hybrid functionals that incorporate Hartree-Fock exchange. DFT maps the interacting electron problem onto a fictitious non-interacting electron system and determines properties via the electron density rather than the many-body wavefunction.

Uploaded by

anusree s
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
100% found this document useful (1 vote)
349 views31 pages

Density Functional Theory

The document provides an introduction to density functional theory (DFT). It discusses the key concepts of DFT including the Hohenberg-Kohn theorems, the Kohn-Sham approach, and solving the Kohn-Sham equations. It also describes common approximations to the exchange-correlation functional such as the local density approximation (LDA) and generalized gradient approximation (GGA), as well as hybrid functionals that incorporate Hartree-Fock exchange. DFT maps the interacting electron problem onto a fictitious non-interacting electron system and determines properties via the electron density rather than the many-body wavefunction.

Uploaded by

anusree s
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 31

INTRODUCTION TO

DENSITY FUNCTIONAL THEORY

Anusree S
IMSc Physics
Electron density

Number of electrons per unit volume at a given point r


Hohenberg–Kohn Theorems

• First theorem:

𝑈𝑒𝑥𝑡

Number of electrons Nucleus


?
ρ (r)
• Total energy = kinetic energy of electron + interaction among electrons +
interaction between electron and nuclei
• Kinetic energy of electron and interaction among electrons are system
independent represent them as universal functional F[ρ(r)]
• External potential varies from 1 system to another depending on the kind of
nuclei
• Then different Hamiltonians differ only by their external potential
• If there are 2 different external potential that yields same ground state
electron density, it will contradict

Electron density alone defines external potential


First theorem: There is an unique external potential determines solely by the
ground state density
Second HK theorem

• Assume we have some well-behaved candidate density


• Density determines a candidate wave function and Hamiltonian
• Evaluate expectation value of energy

< ψ𝑐𝑎𝑛 |𝐻𝑐𝑎𝑛 |ψ𝑐𝑎𝑛 > = 𝐸𝑐𝑎𝑛 ≥ 𝐸0

• Repeat with different densities


• The lowest energy obtained will be closer to true ground state energy
Problems!

oHow to choose improved candidate density rationally?

ρ𝑐𝑎𝑛 ?
o𝐸𝑐𝑎𝑛 = < ψ𝑐𝑎𝑛 |𝐻𝑐𝑎𝑛 |ψ𝑐𝑎𝑛 > Not an improvement from
early First principle methods
Kohn-Sham approach

Interacting system Non-interacting system


of electrons of electrons

Hamiltonian Eigen functions = slator Eigen values =


= sum of 1 determinant of the sum of 1 electron
electron
individual 1 electron eigen values
hamiltonian
eigen functions
Schematic of the Kohn–Sham approach
Energy terms
• The interacting n-electron energy of the HF approach

• In KS approach all the interacting effects are identified as

• Regroup all the interacting terms together as exchange-correlation energy,


Exc

• Then, the final expression of the total energy


Energy terms
1-Non interacting kinetic energy 𝑛
𝑛𝑜𝑛 −1
𝐸𝑘𝑖𝑛 = ෍ ϕ∗𝑖 (𝑟) ∇2 ϕ𝑖 (𝑟)
2
𝑖=1

• ϕ𝑖 - KS orbitals
• Orbitals are directly related to electron density

ρ(𝑟) = ෍ |ϕ𝑖 𝑟 |2
𝑖

The name DFT is still valid


2-External energy

3-Hartree energy
4-Exchange-correlation energy

• The XC energy can be viewed easily with the introduction of the XC


hole in an electronic system

Exchange hole
XC
hole Correlation hole
• Correlation hole forms both
positive and negative directions
• Correlation hole changes shape
of XC hole
Treating XC Energy

Assumptions:

1. XC energy – Local or semi-local functionals of


electron density

2. XC hole is spherical in three dimensions


• The local XC energy per electron electrostatic interaction
energy of an electron at r
with XC hole density at r′

• XC Energy functional

• Thus, the treatment of the XC hole is basically the same as the Hartree
interaction
Kohn-Sham equations

• Total energy of the system

• Applying second HK theorem – variation principle


• Apply chain rule

• Substitute values
𝑛
of each term
𝑛𝑜𝑛 1
𝐸𝑘𝑖𝑛 = − ෍ ϕ∗𝑖 (𝑟) ∇2 ϕ𝑖 (𝑟)
2
𝑖=1

ρ = ϕ∗ ϕ
1 2
0= − 𝛻 + 𝑢𝑒𝑥𝑡 + 𝑢𝐻 + 𝑢𝑥𝑐 − λ𝑖 ϕ𝑖 (𝑟)
2
1 2
0 = − 𝛻 + 𝑢𝑒𝑓𝑓 + λ𝑖 ϕ𝑖 (𝑟)
2

• This is similar to Schrödinger equation with the corresponding KS


Hamiltonian,
1 2
− 𝛻 + 𝑢𝑒𝑓𝑓 𝑟 ϕ𝑖 (𝑟) = 𝜀𝑖 ϕ𝑖 𝑟
2

• For a given system, therefore, actual calculations are performed on this


Hamiltonian and it is required to solve this set of KS equations
simultaneously
Solving Kohn–Sham equations

Self-consistency

KS orbitals electron densities KS Hamiltonian

Therefore, to achieve meaningful results, we have to find a set of KS


orbitals that leads to a KS Hamiltonian whose solutions are the KS
orbitals we started with. This is called self consistency.
Iterative diagonalization
Local density approximation

Schematic of LDA in two dimensions. The actual electron-density (left)


distribution is approximated by four local elements of uniform electron
densities (right).
• Homogeneous electron gas – even distribution of electrons in a
uniform positive external potential, preserving the overall charge
neutrality
• The exchange energy and potential functionals are

• The correlation energy for a simple homogeneous electron gas at the


high density limit is

rs - Wigner–Seitz radius related to electron density


It is the radius of a sphere containing exactly one electron
• The values of per volume of constant ρ for all densities are
parameterized

• Then, the XC energy of the LDA can be calculated by multiplying


by the local electron density and integrating it over the space

• Corresponding XC potential of the LDA is


Drawbacks of LDA

• Overbinds (underestimates the lattice parameters) and thus overestimates


the cohesive energy and bulk modulus of solids
• Calculates the adsorption energies too high
• Underestimates the spin and orbital moments
• Calculates the band gaps ~50% smaller (e.g., Si) or even no gap (e.g.,
Ge)
• Cannot describe transition metals or strongly correlated systems. It
predicts, for example, incorrectly that the ground states of Fe
(ferromagnetic) and Cr (antiferromagnetic) are both nonmagnetic.
• Does not work well for materials that involve weak hydrogen bonds or
van der Waals attraction
Generalised gradient approximation

• Exc depends on the electron density and its gradient at a given point

• The general form of potential Uxc GGA[ρ(r)] is

• There is no simple functional form that correctly represents the GGA data
• Therefore, the general form of GGA is expressed based on the LDA
with an additional enhancement factor F(s) that directly modifies the
LDA energy

s depends on both electron density and its gradient

In solids, the typical range of s = 0 – 3


General trends of GGA

• GGA works very well with almost all systems, giving most structural
properties within a 1 to 3% error
• GGA corrects most of the overbinding problems of LDA, producing an
increase in lattice constants and a decrease in cohesive energies. In fact,
GGA slightly overcorrects bond length, resulting in low adsorption
energies
• GGA also calculates the band gaps approximately 50% smaller, or even
with no gap
Hybrid functional

• GGA-type functionals combined with some (~25%) of the accurate


exchange energy from the HF method
• HF method account exchange energy but does not account for the
correlation energy
• DFT exchange energies can reduce the errors if one introduces the HF
exchange energy into Exc typically in the form of
Reference

• Computational Material Science an Introduction - June Gunn Lee –


CRC Press
• Essentials of Computational Chemistry – Christopher J Cramer –
WILEY publishers – Second edition
THANK YOU

You might also like