Brief introduction to density  functional theory
and its applications .

NB: The loading might take a few minutes


Index
 
 I. Short introduction to density functional theory

 1. The total energy as a function of electron density
 2. Electron correlation

 II.Approximate DFT

 1. Local DFT
 2. Non-local DFT
 3.Ongoing Search for More Accurate Functional

III. Performance of Approximate DFT in Transition Metal Chemistry

 1. Calculated Geometries
 2. Calculated bond energies
 
3. Inorganic Spectroscopy

 3.1 Electronic Excitations
 3.2 Vibrational Spectroscopy
 

 

I. Short  introduction to density  functional theory
 back to index


1. The total energy as a function of electron density

Density functional theory is based on the notion that the total energy, E, of an
electronic system is determined by the electron density r, or E(r).

  Real.Audio sound


    This notion was first suggested by Fermi (1930) and later proven to be exact by W. Kohn
in 1964. Professor Kohn was given the  Nobel prize  for chemistry in 1998 as the founder of density functional theory (DFT)

        In the DFT theory of Kohn  ( see e.g. Ziegler, T. "Approximate Density Functional Theory as a Practical Tool in Molecular Energetics and Dynamics" Chem. Rev., 1991, 91,651. ) the total energy is written in terms of the energy of n non-interacting electrons and a term Eex that takes into account the complicated correlated motion of the electrons.


 
back to index
2. Electron correlation

    As the electrons move they try to avoid the repulsive interaction from each other.


by creating a hole (or no-fly zone) into which other electrons will not penetrate. This is called the exchange-correlation hole and gives rise to the term Eex.
(See Ziegler, T. "Approximate Density Functional Theory as a Practical Tool in Molecular Energetics and Dynamics" Chem. Rev., 1991, 91,651. Also   E.J.Baerends and O.V. Gritsenko Phys. Chem. A 1997,101,5383 )
 

    In electron theory the difficult part is to describe this hole or Eex. It follows from the work of Kohn that the shape of the hole or Eex in principle can be expressed as a function of the electron density. However, we do not know the exact form of this relationship. 
 
back to index
II.Approximate DFT
 

1. Local DFT

    The first generation of approximate DFT took the shape of the exchange-correlation hole function or Eex from the electron gas where the hole-function is known as a function of the electron density. This approximation became known as local DFT and it has been used extensively


 
back to index

2. Non-local DFT

    Axel Becke, John Perdew and others have improved on the simple local electron gas model where the hole function correspond to an electron density that is assumed to be constant throughout. In their non-local theory the hole function or Eex includes corrections from the fact that the electron density in atoms, molecules and solids changes with position.


back to index

 3.Ongoing Search for More Accurate  Functionals

Attempts to developed  better functionals in which the relation between the hole function ( or Eex) and the electron density is expressed more accurately  are still ongoing :

1. A. D. Becke, "A new mixing of Hartree-Fock and local density-functional theories," J. Chem. Phys. 98, 1372-1377 (1993).

2. . K. Burke, J. P. Perdew and Y. Wang, in Electronic Density Functional Theory: Recent Progress and New Directions, Ed. J. F.
Dobson, G. Vignale and M. P. Das (Plenum, 1998).

3.   J. P. Perdew, K. Burke and Y. Wang, Phys. Rev. B 54, 16533 (1996).

4. C. Adamo and V. Barone, Chem. Phys. Lett. 274, 242 (1997).

5. P. M. W. Gill, Mol. Phys. 89, 433 (1996).

6. A. D. Becke, J. Chem. Phys. 104, 1040 (1996).

7. A.D. Daniels, J. M. Milliam and G.E. Scuseria, J. Chem. Phys. 107, 425 (1997).

back to index

 

III. Performance of Approximate DFT in Transition Metal Chemistry
 For a general introduction see: Ziegler, Tom. "The 1994 Alcan Award lecture: Density Functional Theory as a Practical Tool in Studies of Organometallic Energetics and Kinetics. Beating the Heavy metal Blues with  DFT." Can.J.Chem  1995; 73, 743.
back to index


1. Calculated Geometries

In order to describe a chemical reaction from the reactant over the transition state to the product well , DFT would have to be able to determine structures accurately.
 
 

 Geometries calculated by the local DFT method (LDA) and the non-local method LDA/NL are in general in goodagreement with experiment. Bond distances are usually calculated with an accuracy of 0.02 angstrom or better with the LDA/NL method. However, exceptions exist

 Here is a full geometry optimization of Fe2CO9 compared to experiment


back to index
 


2. Calculated bond energies

A DFT description of a chemical reaction also requires accurate (relative) energies of the species involved. It would be useless if DFT predicted (green curve) the reaction to be exothermic with a small barrier that experimentally (red) was observed to be endothermic with a large barrier.


 
 

 Here are a number of calculated M-CO bond disociation energies (kJ/mol). Local DFT (LDA) affords bond dissociation energies that are much too high- it overestimates bond energies. With non-local corrections LDA/NL the bond energies are in good agreement with experiment. Non-local corrections are in general required to obtain good estimates of (relative) energies. The error margin for LDA/NL is around 5 Kcal/mol  or 20 kJ/mol.  (See :Ziegler, Tom. "The 1994 Alcan Award lecture: Density Functional Theory as a Practical Tool in Studies of Organometallic Energetics and Kinetics. Beating the Heavy metal Blues with  DFT." Can.J.Chem  1995; 73, 743. )
 
 Sigma bonds are also estimated with high accuracy by LDA/NL


 back to index

    We can conclude that both energies and geometries are well represented by  LDA/NL methods - to the point where LDA/NL can be used to study the potential energy surface of transition metal complexes.


 

3. Inorganic Spectroscopy

     Molecules in transition metal chemistry are characterized extensively by spectroscopic methods in which a molecule is subjected to light as well as  constant electric or magnetic fields. The response from this perturbation (the spectra) is used to deduce information about the composition or structure of the investigated species.

 DFT can be used to simulate and interprete experimental spectropic observations. Typical fields where this is possible are (a) ionizations; (b) electronic excitations; (c) molecular vibrations; (d) NMR and ESR spectroscopy

 
 back to index

3.1 Electronic Excitations

        Above the ground state potential energy surface are the potential surfaces of the excited states. They are important for photo-chemistry and other related fields. DFT is able to treat excited states on the same footing as the ground state, either by traditional methods (Ziegler, Rauk and Baerends TCA ,1977, 43, 261) or by schemes based on time-dependent perturbation theory (C. A. Ullrich and E. K. U. Gross, Phys. Rev. Lett. 74, 872 (1995)).


 
        The electronic spectrum of  the tetrahedral complex  MnO4-  (permagnate) has been studied extensively and used as a testing ground for new theoretical methods .  Below is a complete assignment of the electronic spectrum of that molecule. The excitations are from oxygen lone-pairs to empty d-orbitals of the formally d0 metal center. This study also examined the structure of the excited states and their distortion from tetrahedral geometry.
(see: Dickson, Ross M., and Tom Ziegler. "A Density Functional Study of the Electronic Spectrum of Permanganate." Int.J.Quantum Chem.1996,58,681. )
 


 back to index


 
3.2 Vibrational Spectroscopy

 DFT on the LDA and LDA/NL levels is able to calculate vibrational frequencies and normal modes of transition metal complexes with high accuracy. Calculated frequencies  can be used to identify intermediates (spectroscopic finger printing ). It is also required for evaluating the entropy of activation in elementary reaction steps. Finally, frequency calculations can be used to generate molecular mechanics force fields for transition metal complexes
 


 
 
 Force constants are the second derivatives of the molecular energy with respect to two nuclear displacements. They can be calculated directly from analytically derived expressions.

(See: Berces,A.; Ziegler,T. Application of Density Functional Theory to the Calculation of Force Fields and Vibrational Frequencies of Transition metal Complexes. Top.in Cur. Chem. 1996,182,42-85 )
 
 


 
 Here is the full assignment for the vibrational spectrum of ferrocene

 
 
 back to index