TUTORIAL REVIEW WWW.Q-CHEM.

ORG

DFT in a Nutshell
Kieron Burke[a,b] and Lucas O. Wagner*[a,b]

The purpose of this short essay is to introduce students and something new outside their area. Important questions
other newcomers to the basic ideas and uses of modern varying in difficulty and effort are posed in the text, and are
electronic density functional theory, including what kinds of answered in the Supporting Information. V C 2012 Wiley
approximations are in current use, and how well they work (or Periodicals, Inc.
not). The complete newcomer should find it orients them well,
while even longtime users and aficionados might find DOI: 10.1002/qua.24259

1 X 1
Electronic Structure Problem V^ee ¼ ; (4)
2 i6¼j
jri
rj j
For the present purposes, we define the modern electronic struc-
and the one-body operator is
ture problem as finding the ground-state energy of nonrelativistic
electrons for arbitrary positions of nuclei within the Born-Oppen- X
N
heimer approximation.[1] If this can be done sufficiently accurately V^ ¼ vðrj Þ: (5)
j¼1
and rapidly on a modern computer, many properties can be pre-
dicted, such as bond energies and bond lengths of molecules,
For instance, in a diatomic molecule, v(r) ¼
ZA/r
ZB/|r
R|.
and lattice structures and parameters of solids.
We use atomic units unless otherwise stated, setting
Consider a diatomic molecule, whose binding energy curve
e2 ¼
h ¼ me ¼ 1, so energies are in Hartrees (1 Ha ¼ 27.2 eV
is illustrated in Figure 1. The binding energy is given by
or 628 kcal/mol) and distances in Bohr radii (1 a0 ¼ 0.529 Å).
The ground-state energy satisfies the variational principle:

^
E ¼ min hWjHjWi; (6)
W

where the minimization is over all antisymmetric N-particle

wavefunctions. This E was called E0(R) in Eq. (1).*
Many traditional approaches to solving this difficult many-
body problem begin with the Hartree–Fock (HF) approxima-
tion, in which W is approximated by a single Slater determi-
nant (an antisymmetrized product) of orbitals (single-particle
wavefunctions)[2] and the energy is minimized.[3] These include
Figure 1. Generic binding energy curve. For N2, values for R0 and De are configuration interaction, coupled cluster, and Møller-Plesset
given in Table 1. [Color figure can be viewed in the online issue, which is
available at wileyonlinelibrary.com.] perturbation theory, and are mostly used for finite systems,
such as molecules in the gas phase.[4] Other approaches use
ZA ZB reduced descriptions, such as the density matrix or Green’s
Ebind ðRÞ ¼ E0 ðRÞ þ
EA
EB (1)
R function, but leading to an infinite set of coupled equations
where E0(R) is the ground-state energy of the electrons with that must somehow be truncated, and these are more com-
nuclei separated by R, and EA and ZA are the atomic energy mon in applications to solids.[5]
and charge of atom A and similarly for B. The minimum tells us More accurate methods usually require more sophisti-
the bond length (R0) and the well-depth (De), corrected by cated calculation, which takes longer on a computer. Thus,
zero-point energy (
hx=2), gives us the dissociation energy (D0). there is a compelling need to solve ground-state electronic
The Hamiltonian for the N electrons is structure problems reasonably accurately, but with a cost in

^
H^ ¼ T^ þ V^ee þ V; (2)
[a, b] K. Burke, L. O. Wagner
where the kinetic energy operator is Department of Chemistry, University of California, Irvine, California 92697
Department of Physics, University of California, Irvine, California 92697
1 X
N
E-mail: [email protected]
T^ ¼
r2j ; (3)
2 j¼1 *Explain why a vibrational frequency is a property of the ground-state of the
electrons in a molecule.
the electron–electron repulsion operator is C 2012 Wiley Periodicals, Inc.
V

96 International Journal of Quantum Chemistry 2013, 113, 96–101 WWW.CHEMISTRYVIEWS.ORG

 WWW.Q-CHEM.ORG TUTORIAL REVIEW

Kieron Burke is a professor of heretical physical and computational chemistry at the

University of California in Irvine. He spent the last two decades of his life in functionals.
[Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Lucas Wagner began working in 2009 as a graduate student with Kieron Burke starting off
curious, plucky, and impressionable in the wide world of DFT. Lucas studies simple (mostly
1d) models and enjoys the combination of analytical and numerical studies involved. [Color
figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

computer time that does not become prohibitive as the enough for most properties of interest (for example, molecules
number of atoms (and therefore electrons) becomes large. do not bind[8]). For same-spin, noninteracting fermions in 1d,
the corresponding kinetic energy is
Basic DFT Z
p2
The electronic density n(r) is defined by the requirement that T TF 1d ½n ¼ dx n3 ðxÞ (11)
6
n(r) d3r is the probability of finding any electron in the volume
d3r around r. For a single electron with wavefunction f(r), it is and makes only a 25% error on the density of a single particle
simply |f(r)|2. In density functional theory (DFT), we write the in a box. Hours of endless fun and many good and bad prop-
ground-state energy in terms of n(r) instead of W. The first erties of functional approximations can be understood by
DFT was formulated by Thomas[6] and Fermi.[7] The kinetic applying Eq. (11) to standard text book problems in quantum
energy density at any point is approximated by that of a uni- mechanics,‡ and noting what happens, especially for more
form electron gas of noninteracting electrons of density n(r), than one particle.§
which for a spin-unpolarized system is:† But modern DFT began with the proof that the solution of
Z the many-body problem can be found, in principle, from a
T TF ¼ aS d3 r n5=3 ðrÞ; aS ¼ 3ð3p2 Þ2=3 =10: (7) density functional. To see this, we break the minimization of
Eq. (6) into two steps. First minimize over all wavefunctions
The interelectron repulsion is approximated by the classical
yielding a certain density, and then minimize over all densities.
electrostatic self-energy of the charge density, called the Har-
Because the one-body potential energy depends only on the
tree energy:
density, we can define separately[9,10]
Z Z
1 nðrÞ nðr0 Þ

U¼ d3 r d3 r0 : (8) F½n ¼ minhWj T^ þ V^ee jWi; (12)
2 jr
r0 j W!n

Because the one-body potential couples only to the density, where the minimization is over all antisymmetric wavefunc-
Z tions yielding a given density n(r).¶ This is transparently a
^ ¼
V ¼ hVi d3 r nðrÞ vðrÞ: (9) functional of the density, meaning it assigns a number to
each density, as was first proven by Hohenberg and
The sum of these three energies is then minimized, subject to Kohn.**[11] Then
the physical constraints:
Z
nðrÞ  0; d3 r nðrÞ ¼ N: (10) ‡
Calculate the TF kinetic energy for a 1d particle of mass m ¼ 1 in (a) a har-
monic well (v(x) ¼ x2/2) and (b) in a delta-well (v(x) ¼
d(x)). Give the % errors.
This absurdly crude theory gives roughly correct energies §
Deduce the exact energy for N same-spin fermions in a flat box of width 1
(errors about 10% for many systems) but is not nearly good bohr. Then evaluate the local approximation to the kinetic energy for N ¼ 1, 2,
and 3, and calculate the % error.
¶
Derive F[n] for a single electron. It has no electron–electron interaction, and is
known as the von Weizs€acker kinetic energy.
†
Evaluate the TF kinetic energy of the H atom and deduce the % error. Repeat **Technically, HK only proved fact this for v-representable densities, i.e. den-
using spin-DFT. sities which are the ground-state of some external one-body potential.

International Journal of Quantum Chemistry 2013, 113, 96–101 97

 TUTORIAL REVIEW WWW.Q-CHEM.ORG

 Z 
then vS;r ðrÞ ¼ vr ðrÞ þ vH ðrÞ þ vXC;r ðrÞ; (19)
E ¼ min F½n þ d3 r nðrÞ vðrÞ ; (13)
n

where
where the minimization is over all reasonable densities satisfy-
ing Eq. (10). Hohenberg and Kohn proved (i) that all properties dEXC
vXC;r ðrÞ ¼ : (20)
are determined by n(r), i.e., they are functionals of n(r),†† (ii) dnr ðrÞ
F[n] is a universal functional, independent of v(r), and (iii) the
exact density satisfies This is a formally exact scheme for finding the ground-state
dF energy and density for any electronic problem.*** In Figure 2,
¼
vðrÞ; (14)
dnðrÞ
where dF/dn(r) is the functional derivative of F with respect to
n(r). In fact, these days we use spin DFT,[12] in which all quanti-
ties are considered functionals of the up, n:(r), and down,
n;(r), spin densities separately. This makes approximations
more accurate for odd electron systems and allows treatment
of collinear magnetic fields. All functionals written without
spin dependence, such as the ones discussed thus far, are
assumed to be referring to a spin-unpolarized system.
The next crucial step in developing the modern theory came
from (re)-introducing orbitals. Kohn and Sham[13] vastly
improved the accuracy of DFT by imagining a fictitious set of
noninteracting electrons that are defined to have the same den-
sity as the interacting problem. They are still spin-12 fermions Figure 2. Exact KS potential of He atom along z axis; for neutral atoms,
the KS potential goes asymptotically to 1/r for large r (data from Umrigar
obeying the Pauli principle, so like in HF theory, their wavefunc-
and Gonze, Phys. Rev. A 1994, 50, 3827, V C American Physical Society, repro-
tion is (usually) a Slater determinant, an antisymmetrized prod- duced by permission). [Color figure can be viewed in the online issue,
uct of orbitals of each spin, fjr(r), j ¼ 1,…,Nr, r ¼ : , ;. These which is available at wileyonlinelibrary.com.]
KS electrons satisfy a noninteracting Schro €dinger equation:
  we emphasize the exactness of the KS scheme by plotting the
1

r2 þ vS;r ðrÞ /jr ðrÞ ¼
jr /jr ðrÞ; (15) exact KS potential for a He atom (which is trivial to find, once
2
the exact density is known from an accurate many-body calcu-
and the ejr are called KS eigenvalues and fjr(r) are KS lation[14]):††† Two noninteracting electrons, doubly occupying
orbitals.‡‡ By evaluating Eq. (3) on the KS Slater determinant, the 1s orbital of this potential, have a density that matches
the KS kinetic energy is the sum of the orbital contributions:§§ that of the interacting system exactly.‡‡‡ But in practical
Nr Z
calculations, we always use approximations to EXC and hence
1XX
TS ¼
d 3 r /jr ðrÞr2 /jr ðrÞ: (16) to vXC(r).
2 r¼";# j¼1 Traditionally, EXC is broken up into exchange (X) and correla-
tion contributions:§§§
If we write the energy in terms of KS quantities:
EXC ¼ EX þ EC : (21)
E ¼ TS þ U þ V þ EXC ; (17)

EXC is defined by Eq. (17) and called the exchange-correlation The exchange energy is V̂ee evaluated on the KS Slater deter-
(XC) energy.¶¶ KS showed that one could extract the unknown minant minus the Hartree energy, and typically dominates.¶¶¶
KS potential if one only knew how the terms depend on the In terms of the orbitals:
density. Writing the Hartree potential as
Z
dU nðr0 Þ
vH ðrÞ ¼ ¼ d3 r0 ; (18)
dnðrÞ jr
r0 j
***Why is a KS calculation much faster than direct solution of the Schr€
odinger
equation?
†††
Write the formula that extracts vS(r) from n(r) for the helium atom. Can you
explain why this does not tell us the vital vXC[n](r) for any spin-unpolarized
††
Again, HK proved this only for v-representable densities. two-electron systems?
‡‡ ‡‡‡
Is the sum of the KS eigenvalues equal to the total energy? Why is the helium KS potential less deep than the original potential,
2/r?
§§ §§§
Apply Eq. (12) to the KS system to define TS without ever mentioning vS(r). What is EXC for a one-electron system?
Then prove that T  TS always. ¶¶¶
What is the expectation value of the Hamiltonian (Ĥ ¼ T̂ þ V̂ þ V̂ee) eval-
¶¶
Give the signs of E, T, Vee, V, U, EX, and EC for real systems (i.e., atoms, mole- uated on the KS Slater determinant? Use this to prove that the DFT definition
cules, and solids). of correlation energy is never positive.

98 International Journal of Quantum Chemistry 2013, 113, 96–101 WWW.CHEMISTRYVIEWS.ORG

 WWW.Q-CHEM.ORG TUTORIAL REVIEW

Z Z
1 X /ir ðrÞ /jr ðr0 Þ /ir ðr0 Þ /jr ðrÞ Although LDA is uniquely defined, there are two basic fla-
EX ¼
d3 r d3 r 0 : (22) vors of the more sophisticated functionals. There are those
2 jr
r0 j
r; i; j that are derived without fitting to reference data on atoms
occ and molecules, using information from only the slowly varying
This is precisely the same orbital expression given in HF, electron and known exact conditions on the functional. The
though it makes use of the KS orbitals (which are implicit standard GGA of this type is PBE,[18] while the hybrid is PBE0,
functionals of the density[15] rather than the HF orbitals.**** mixing 25% exact exchange.[19] Of the empirical type, B88 is
Then the correlation energy is everything else, i.e., defined to the standard GGA for exchange,[20] LYP for correlation,[21] com-
make Eq. (17) exact.†††† bining to form BLYP. The most commonly used functional
today is a hybrid called B3LYP.[22]
Real Calculations The original LDA became a standard tool in solid-state
physics, yielding excellent lattice parameters, and fairly good
Practical calculations use some simple approximation to bulk moduli. But LDA typically overbinds by about 1 eV/bond,
EXC[n:,n;]. The KS equations are started with some initial guess which is too large an error to be useful in quantum chemistry.
for the density, yielding a KS potential via Eq. (20). The KS GGAs reduce this error to about 0.3 eV/bond, and hybrids
equations are then solved and a new density is found. This reduce it another factor of 2. Table 1 shows typical results. Em-
cycle is repeated until changes become negligible, i.e., this a
self-consistent field calculation.
The standard approximations are very simple. The local Table 1. Differences in calculated [using Turbomole (http://
(spin) density approximation (often just called LDA) is[13]: www.turbomole.com/)[23]] versus experimental values (from Ref. [24]) for
well depth (De) in kcal/mol and equilibrium distance (R0) in pm of N2.
Z
LDA
EXC ¼ d3 r eunif
XC ðn"ð rÞ; n#ð rÞÞ (23) D ¼ calc
expt
HF LDA PBE BLYP PBE0 B3LYP Expt.
where eunif
XC (n:,n;)is the XC energy density of a uniform gas of De
110 40.4 16.2 13.0
1.73 1.06 227.0
spin densities n: and n;. The X contribution was first written R0
3.2
0.33 0.41 0.40
0.93
0.80 109.8
by Dirac,[16] and is found by inserting plane waves (the KS
orbitals of a uniform gas) into Eq. (22).‡‡‡‡ In the unpolarized
case, the result is§§§§ pirical parametrized functionals are usually about a factor of 2
better than nonempirical ones, but with less systematic
eunif 4=3
aX ¼ 3ð3p2 Þ1=3 =ð4pÞ errors.[25] Although hybrids are popular in chemistry, where HF
X ðnÞ ¼
aX n ; (24)
codes have been including exact exchange for decades, they
while eunif are much less popular for solids, due to the singularities in HF
C (n:,n;) has been calculated and accurately parame-
trized.[17] More accurate energetics are usually obtained with a for metals.
generalized gradient approximation (GGA), which includes de- In contrast, many limitations of these functionals have been
pendence on the gradient of the density identified. Perhaps the most well-known is the gap problem.
The fundamental gap is I
A, the difference between the ioni-
Z
GGA zation potential and the electron affinity of a system. This is
EXC ¼ d3 r eGGA
XC ðn" ðrÞ; n# ðrÞ; jrn" ðrÞj; jrn# ðrÞjÞ; (25)
usually larger than the KS gap, the difference between the KS
HOMO and LUMO energy eigenvalues. Calculations with LDA
where now the energy density is given by some approximate and GGA yield fundamental gaps of insulating solids close to
form. The third standard form of approximation is a hybrid, the KS gap, and so are too small (by about a factor of 2). This
which mixes in a fixed fraction of the exact exchange energy is related to self-interaction error (that is, the functionals are
incorrect for one electron). Hybrid functionals often do better.
hyb
EXC ¼ a ðEX
EXGGA Þ þ EXC
GGA
; (26) Other well known failures of these approximations include the
lack of asymptotic van der Waals forces, which can be impor-
where a is a universal parameter (often about 1/4) and EX is tant for soft matter.[26]
the exact exchange defined in Eq. (22). There are many suggestions on how to improve these stand-
ards. Presently, meta-GGAs, which use KS kinetic energy densities
****Derive a formula for EX[n:,n;] in terms of EX[n], evaluated on various den- to approximate EXC, can produce accurate energetics without
sities. Does the same formula apply to the KS kinetic energy, but with EX
using HF exchange.[27] Alternatively, including only the short-
replaced by TS? How about for EC?
††††
ranged contribution to exchange can also avoid the difficulties of
This definition differs slightly from that of quantum chemistry.
0
applying HF to solids.[28] But so far, none of these is even close to
‡‡‡‡
Show that if electrons repelled via a contact repulsion, d(r
r ), the replacing the standard functionals in amount of use.
exchange is given exactly by LDA, and give its expression, including spin-
dependence. Computer codes also need some basis set to represent the
§§§§ KS orbitals. Basis functions localized on atoms are used for
Use dimensional analysis to explain the powers of the density in the local
approximations of the kinetic energy in 1d and 3d, as well as the exchange molecular systems in quantum chemistry. Most codes now use
energy in 3d. Gaussians, but some use Slater-type orbitals (exponentials

International Journal of Quantum Chemistry 2013, 113, 96–101 99

 TUTORIAL REVIEW WWW.Q-CHEM.ORG

Table 2. Acronyms. Table 3. Notation of representative symbols.

Abbr. Expansion Symbol Meaning

B88 Becke 88 EX functional De Electronic well-depth (see Fig. 1)
BLYP GGA using B88 and LYP EC Correlation energy
B3LYP BLYP hybrid Ex Exchange energy [defined in Eq. (22)]
DFT Density functional theory eunif
XC Exchange-correlation energy density of uniform gas
GGA Generalized gradient approximation F[n] Hohenberg–Kohn universal functional (12)
HOMO Highest-occupied molecular orbital n
(r) Density of spin
electrons
KS Kohn-Sham R0 Equilibrium bond length (see Fig. 1)
LDA Local density approximation T True kinetic energy [expectation of Eq. (3)]
LUMO Lowest-unoccupied molecular orbital U Hartree energy [defined in Eq. (8)]
LYP Lee-Yang-Parr EC functional[21] Vee Electron–electron repulsion [expectation of Eq. (4)]
PBE Perdew-Burke-Ernzerhof GGA functional vS (r) KS potential (20)
PBE0 PBE hybrid Z Atomic charge
TF Thomas-Fermi
X Exchange
XC Exchange-correlation the kernel to be extracted from a ground-state XC approxima-
tion. Typical results give moderate accuracy for transition fre-
quencies (0.4 eV errors), but excellent results for the properties
centered on atoms) and localized basis sets. In solids and of molecules in excited states.[36] Dramatic failures of these
materials, periodic boundary conditions are convenient for approximations include double excitations, charge transfer,
extended systems, making plane waves a natural choice. Core and a missing contribution to the optical response of solids.[37]
electrons are often treated as frozen or by pseudo-potentials,
saving calculational effort. A few codes use numerical grids.[5] Useful Resources
Electronic KS-DFT has been implemented in a huge variety
of codes, and is being applied in many areas of science. Mod- Tables 2 and 3 give a list of acronyms and notation used in
ern computers allow routine calculation of hundreds of atoms, this essay. Perhaps the best pedagogical introduction of recent
far more than with traditional quantum chemical methods, but years is ‘ A Primer in DFT.’’[38] For physicists with some back-
with a reasonable (not very high) degree of accuracy and reli- ground in many-body theory, there’s nothing better than Drei-
ability. In traditional areas of condensed-matter physics and zler and Gross,[39] or the more recent Engel and Dreizler.[40]
quantum chemistry, many experimental groups run codes Analogously, if you are a chemist with knowledge of quantum
themselves, and many purely computational groups exist. For chemical methods, try Parr and Yang.[41] A more pragmatic
example, new catalysts have been found by calculation, then approach is that of Koch and Holthausen[42] who, in Part B,
demonstrated in the lab, patented, and are being further discuss the accuracy of different approximate functionals for
developed for industrial production.[29] But such calculations different molecular properties. For TDDFT, the standard refer-
and calculators are also appearing in materials science, geol- ence is that edited by Marques,[37] and there is a new intro-
ogy, and biochemical departments. Recent uses also include duction by Ullrich.[43]
protein science, soil science, and astrophysics. Liquids can be To check you have followed the essay, you should certainly
simulated using ab initio molecular dynamics (AIMD).[30] There try the questions posed in the text, which are collected and
is increasing interest in matter under high pressure and tem- answered in the Supporting Information.
perature (but not high enough to be pure plasma) and here
DFT simulation is making a large impact.[31]
Keywords: density functional theory, electronic structure, local
In its original form, electronic DFT focused purely on the
density approximation
ground-state energy and density. Several approaches exist to
tackle excited states, but the most popular is time-dependent
density functional theory (TDDFT), based on a formal proof for
time-dependent one-body potentials.[32] In the linear response How to cite this article: K. Burke, LO. Wagner, Int. J. Quantum
regime, this leads to a simple scheme for calculating transition Chem. 2013, 113, 96–101. DOI: 10.1002/qua.24259
frequencies which has become popular for molecules. The Additional Supporting Information may be found in the
response scheme uses the ground-state KS orbitals as its start- online version of this article.
ing point. In Figure 2, the (doubly occupied) 1s orbital is at

24.57 eV, precisely
I, the ionization potential for the system.
[1] M. Born, R. Oppenheimer, Ann. Phys. 1927, 389, 457.
The 2p orbital is at
3.429 eV, so the first allowed optical tran-
[2] J. C. Slater, Phys. Rev. 1929, 34, 1293.
sition of the KS electrons has frequency 21.14 eV.[33] TDDFT [3] D. R. Hartree, W. Hartree, Proc. R. Soc. London. Ser. A – Math. Phys. Sci.
corrects these transitions to the exact value 20.97 eV,[34] but 1935, 150, 9.
once again requires knowledge of an unknown functional, [4] A. Szabo, N. S. Ostlund, Modern Quantum Chemistry; MacMillan: New
York, 1982.
called the XC kernel, which depends on the perturbing fre-
[5] R. M. Martin, Electronic Structure: Basic Theory and Practical Methods;
quency.[35] Present applications all use the adiabatic approxi- Cambridge University Press, 2004.
mation, which ignores the frequency dependence and allows [6] L. H. Thomas, Math. Proc. Camb. Phil. Soc. 1927, 23, 542.

100 International Journal of Quantum Chemistry 2013, 113, 96–101 WWW.CHEMISTRYVIEWS.ORG

 WWW.Q-CHEM.ORG TUTORIAL REVIEW

[7] E. Fermi, Z. Phys. A: Hadrons Nucl., 1928, 48, 73. [31] S. Root, R. J. Magyar, J. H. Carpenter, D. L. Hanson, T. R. Mattsson,
[8] E. Teller, Rev. Mod. Phys. 1962, 34, 627. Phys. Rev. Lett. 2010, 105, 085501.
[9] M. Levy, Proc. Natl. Acad. Sci. USA 1979, 76, 6062. [32] E. Runge, E. K. U. Gross, Phys. Rev. Lett. 1984, 52, 997.
[10] E. H. Lieb, Int. J. Quantum Chem. 1983, 24, 243. [33] C. J. Umrigar, A. Savin, X. Gonze, Are Unoccupied Kohn-Sham Eigenval-
[11] P. Hohenberg, W. Kohn, Phys. Rev. 1964, 136, B864. ues Related to Excitation Energies? Plenum: New York, 1997.
[12] U. von Barth, L. Hedin, J. Phys. C: Solid State Phys. 1972, 5, 1629. [34] A. Kono, S. Hattori, Phys. Rev. A 1984, 29, 2981.
[13] W. Kohn, L. J. Sham, Phys. Rev. 1965, 140, A1133. [35] K. Burke, J. Werschnik, E. K. U. Gross, J. Chem. Phys. 2005, 123,
[14] C. J. Umrigar, X. Gonze, Phys. Rev. A 1994, 50, 3827. 062206.
[15] S. Kümmel, L. Kronik, Rev. Mod. Phys. 2008, 80, 3. [36] P. Elliott, F. Furche, K. Burke, Excited States from Time-Dependent
[16] P. A. M. Dirac, Math. Proc. Cambridge Philos. Soc. 1930, 26, 376. Density Functional Theory; Wiley: Hoboken, NJ, 2009; pp. 91–165.
[17] J. P. Perdew, Y. Wang, Phys. Rev. B 1992, 45, 13244. [37] M. A. L. Marques, N. T. Maitra, F. M. S. Nogueira, E. K. U. Gross, A.
[18] (a) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865; Rubio, Eds. Fundamentals of Time-Dependent Density Functional
(b) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1997, 78, 1396(E). Theory, Lecture Notes in Physics, Vol 837; Springer: Heidelberg, 2012.
[19] K. Burke, M. Ernzerhof, J. P. Perdew, Chem. Phys. Lett. 1997, 265, 115. [38] C. Fiolhais, F. Nogueira, M. Marques, A Primer in Density Functional
[20] A. D. Becke, Phys. Rev. A 1988, 38, 3098. Theory; Springer-Verlag: New York, 2003.
[21] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785. [39] R. M. Dreizler, E. K. U. Gross, Density Functional Theory: An Approach
[22] A. D. Becke, J. Chem. Phys. 1993, 98, 5648. to the Quantum Many-Body Problem; Springer–Verlag: Berlin, 1990.
[23] R. Ahlrichs, M. B€ar, M. H€aser, H. Horn, C. K€ olmel, Chem. Phys. Lett. [40] E. Engel, R. M. Dreizler, Density Functional Theory: An Advanced
1989, 162, 165. Course; Springer: Berlin, 2011.
[24] M. Ernzerhof, G. E. Scuseria, J. Chem. Phys. 1999, 110, 5029. [41] R. G. Parr, W. Yang, Density Functional Theory of Atoms and Molecules;
[25] D. Rappoport, N. R. M. Crawford, F. Furche, K. Burke, Computational Oxford University Press: Oxford, 1989.
Inorganic and Bioinorganic Chemistry, In Which functional should I [42] W. Koch and M. C. Holthausen, A Chemist’s Guide to Density
choose? E. I. Solomon, R. B. King, R. A. Scott, Eds.; John Wiley & Sons, Functional Theory, 2nd ed.; Wiley-VCH: Weinheim, 2002.
Inc., 2009. [43] C. A. Ullrich, Time-Dependent Density-Functional Theory; Oxford
[26] M. Dion, H. Rydberg, E. Schr€ oder, D. C. Langreth, B. I. Lundqvist, Phys. University Press: Oxford, 2012.
Rev. Lett. 2004, 92, 246401.
[27] J. P. Perdew, A. Ruzsinszky, G. I. Csonka, L. A. Constantin, J. Sun, Phys.
Rev. Lett. 2009, 103, 026403.
[28] J. Heyd, G. E. Scuseria, J. Chem. Phys. 2004, 121, 1187.
Received: 25 May 2012
[29] A. Nilsson, L. G. M. Pettersson, J. Nørskov, Eds. Chemical Bonding at
Revised: 25 May 2012
Surfaces and Interfaces; Elsevier: Amsterdam, 2008. Accepted: 4 June 2012
[30] R. Car, M. Parrinello, Phys. Rev. Lett. 1985, 55, 2471. Published online on 12 July 2012

International Journal of Quantum Chemistry 2013, 113, 96–101 101