DFT Functionals

 

DFT Functionals – Concept, Categories, and Detailed Study

Why Functionals Are Used in DFT

In Density Functional Theory, the total electronic energy of a system is written as:

Where:

• T[ρ] = kinetic energy
• V[ρ] = nuclear attraction
• J[ρ] = electron–electron repulsion
• E_xc​[ρ] = exchange–correlation energy

In DFT, everything depends on the exchange–correlation energy (E_xc).

Problem is we do not know the exact form of E_xc. So, we use mathematical approximations of exchange–correlation energy known as functionals.

Simple we can define.

“A functional is a mathematical expression used to approximate electron interactions in DFT.”

Categories of Functionals

DFT functionals are classified based on how they treat electron density.

·         Local Density Approximation

·         Generalized Gradient Approximation

·         Hybrid Functionals

·         Range-Separated Functionals

·         Dispersion-Corrected Functionals

 (1) Local Density Approximation (LDA)

• Depends only on electron density (ρ)
• Simplest approximation

Example:

• SVWN

✔ Advantage: Simple
❌ Disadvantage: Low accuracy

 (2) Generalized Gradient Approximation (GGA)

• Depends on:
• Density (ρ)
• Gradient of density (∇ρ)

Examples:

• PBE
• PW91

✔ Better than LDA
❌ Still lacks high accuracy

(3) Hybrid Functionals

• Combine:
• DFT exchange
• Hartree–Fock exchange

🔹 Examples:

• B3LYP
• MPW1PW91

✔ Much improved accuracy

(4) Range-Separated Functionals

• Treat:
• Short-range interactions
• Long-range interactions separately

🔹 Examples:

• CAM-B3LYP
• ωB97

✔ Important for excited states and charge transfer

 (5) Dispersion-Corrected Functionals

Include weak interactions like:

• van der Waals forces

🔹 Example:

• ωB97XD

✔ Important for large systems

Each functional tries to fix a specific weakness:

Problem in DFT	Functional Solution
Poor accuracy	Hybrid functionals
Weak long-range interactions	Range-separated
Missing dispersion	Dispersion-corrected
Bad excited states	CAM-type functionals

 Important Functionals

B3LYP functional

Hybrid functional that combines Becke exchange with LYP correlation using three fitted parameters.

B3LYP = Becke (3-parameter) + Lee–Yang–Parr

• Becke exchange (He developed the exchange functional)
• Lee–Yang–Parr correlation(They developed the correlation functional used in B3LYP)
• ~20% Hartree–Fock exchange

Why It Is Used

✔ Reliable for:

• Geometry optimization
• Vibrational analysis
• Organic molecules

Limitations

• Long-range charge transfer
• Excited states
• Band gap prediction

“Best general-purpose functional”

CAM-B3LYP Functional

 “CAM” stands for Coulomb-Attenuating Method.

•  “Coulomb” = electron–electron interaction
• “Attenuating” = modifying or splitting

Range-separated hybrid

• Splits interactions into:
• Short-range
• Long-range

Why It Is Used

✔ Excellent for:

• Excited states (TD-DFT)
• UV–Vis spectra
• Charge transfer systems

Advantage over B3LYP

✔ Corrects long-range error

Limitation

❌ Slightly more computational cost

“Best for optical and excited-state studies”

MPW1PW91 Functional

MPW1PW91 is a hybrid functional that uses modified Perdew–Wang exchange with one mixing parameter and PW91 correlation.

MPW” stands for

Modified Perdew–Wang exchange functional

  It is a modified version of the original Perdew–Wang exchange

  Designed to improve accuracy of exchange energy

One-Parameter Hybrid

1 means:

It uses one empirical mixing parameter

• This parameter controls how much Hartree–Fock exchange is mixed
• That’s why it is called a one-parameter hybrid functional

PW91 → Perdew–Wang 1991 (Correlation)

“PW91” refers to:

Perdew–Wang 1991 correlation functional

• Developed by:
• John Perdew
• Yue Wang

ωB97 Functional

Variants

• ωB97
• ωB97X
• ωB97XD (with dispersion)

ωB97XD = Range-separated (ω) + Becke (B97) + exact exchange (X) + dispersion (D)

ωB97XD is a modern DFT functional that includes range separation, exact exchange, and dispersion correction for high accuracy.

Key Concept

• Includes:
• Long-range correction
• Dispersion interaction

Why It Is Used

✔ Excellent for:

• Non-covalent interactions
• Biomolecules
• Large systems

Advantage

✔ Very accurate modern functional

Limitation

❌ Higher computational cost

“Best for weak interactions and modern studies”

Functional	Type	Best Use
B3LYP functional	Hybrid (global hybrid)	Geometry optimization, general organic molecules
CAM-B3LYP functional	Range-separated hybrid	Excited states, UV–Vis, charge transfer systems
MPW1PW91 functional	Hybrid (one-parameter)	Thermochemistry, reaction energies
ωB97XD functional	Range-separated hybrid + dispersion	Non-covalent interactions, weak forces, large systems