1.

4 Quantum mechanics basis

1.4.1 The Schrödinger equation

1.4.2 The Born-Oppenheimer approximation and the concept of electronic structure

1.4.3 The potential energy surface (PES)

1.5 Ab initio methods

1.5.1 HF self-consistent field method

1.5.2 Post-HF method-electron correlation

1.5.2.1 Configuration interaction

1.5.2.2 Multi-configuration self-consistent field method

1.5.2.3 MP2 perturbation method

1.6 Semi empirical Methods

1.7 Density Functional Theory

1.7.1 The Hohenberg-Kohn theorems

1.7.2 The Kohn-Sham methods

1.7.3 Exchange-correlation functionals

1.8 Basis set

1.9 Bonding Analysis

1.9.1 Molecular orbital analysis

1.9.2 Population analysis

1.9.2.1 Mulliken population analysis

1.9.2.2 Natural Bond-Order (NBO) analysis

1.9.2.3 Atoms in Molecules (AIM)

1.9.3 Electron density analysis

1.10 Solvent Model

1|Page
 General Introduction and Computational Methods

1.1 Catalysis

“A catalyst is a substance which alters the velocity of a chemical reaction without

appearing in the final products.“— Carl Wilhelm Wolfgang Ostwald

International Union of Pure and Applied Chemistry (IUPAC) Gold book describe
catalysis as “the action of a substance (i.e. catalyst) that increases the rate of a reaction

2|Page
 without modifying the overall standard Gibbs free energy change in the reaction”. A catalyst
accelerates a chemical reaction. It does so by forming bonds with the reacting molecules, and
by allowing these to react to a product, which detaches from the catalyst, and leaves it
unaltered such that it is available for the next reaction. In fact, we can describe the catalytic
reaction as a cyclic event in which the catalyst participates and is recovered in its original
form at the end of the cycle. A catalyst does not influence the thermodynamic equilibrium of
reactants and products but affects the rates of the chemical reactions i.e. the reaction Gibbs
free energy change (ΔG) in a catalyzed reaction remains unaffected whereas the activation
energy (Ea) changes as depicted in Figure 1.1.

Catalyzed rxn
proceeding through
an intermediate

Ea

Ea
c a ta ly ze d
G
Re a c ta nts
G

Produc ts

Re a c tion Coordina te

Figure 1.1. General representation of an arbitrary reaction profile for an uncatalyzed (blue
line) and catalyzed (red line) reactions pathways.

1.1.1 Computational study of catalysis

“It is nice to know that the computer understands the problem. But I would like to
understand it too.” -E. P. Wigner (quoted in Physics Today July 1993, p. 38)

3|Page
 Computational modelling of catalysis is one of the fundamental research areas of
chemistry which gained substantial importance at the beginning of 21st century.[1] Catalytic
processes are usually complicated with multistep reactions which are difficult to characterize
experimentally.[1] An efficient catalytic process should be fast, smooth and precise, hence,
the involved intermediates are difficult to isolate and characterize. Computational chemistry
can be applied to access detailed knowledge of the reaction mechanism, which is absolutely
necessary to design a new catalyst with better efficiency.

MLn
Pre-catalyst

P R1
Reactant Reactant-1
MLn-1
Catalyst

Inactive
MLn-1∙P MLn-1∙R species
Int-1
Int-3

R2
Reactant-2

MLn-1∙R1R2
Int-2

Scheme 1.1 General catalytic cycle.

1.1.2 The Catalytic Cycle

To regenerate the catalyst, we can combine all the steps involve in a catalytic reaction to
constitute a catalytic cycle as shown in Scheme 1.1. The isolated catalyst in general does not
act as an active species. In such case, the added catalysts are called precursors and they are
not included in catalytic cycle. The species that is responsible for catalytic process is
generally termed as active species and can be generated either by dissociation of a ligand or
by assistance from an activator. Once this active species is formed from the precursor, it
undergoes a series of transformations involving several intermediates and transition states,
and eventually giving rise to the final product along with its own regeneration. However, in

4|Page
some cases, there can appear competitive reactions giving rise to stable species, which causes
the catalyst deactivation and thus retards the catalytic process.

1.1.3 Quantification of the Activity and Efficiency of Catalytic processes

The performance of catalyst is determined by its activity for reaction and selectivity for a
product. The activity of a catalyst for a reaction at particular conditions is usually expressed
as the rate of reaction at that condition over the catalyst. The catalyst that shows higher rate
of reaction at the given conditions is said to have higher activity. For solid catalyzed reaction,
rate can be defined as, r = g mol of reacted /g of the catalyst used. Activity is also expressed
in terms of conversion of a reactant to product achieved. Higher the conversion at given
conditions, higher is the activity of the catalyst for that reaction. The conversion can be
defined as follows

𝑀𝑜𝑙𝑒𝑠 𝑜𝑓 𝑟𝑒𝑎𝑐𝑡𝑎𝑛𝑡 − 𝑀𝑜𝑙𝑒𝑠 𝑜𝑓 𝑝𝑟𝑜𝑑𝑢𝑐𝑡

𝐶𝑜𝑛𝑣𝑒𝑟𝑠𝑖𝑜𝑛(%) = × 100%
𝑀𝑜𝑙𝑒𝑠 𝑜𝑓 𝑟𝑒𝑎𝑐𝑡𝑎𝑛𝑡

In general, three quantities are very important in determining the efficiency of a catalytic
system. The first one is the turnover number (TON), second is the turnover frequency (TOF)
and third one is selectivity. TON is defined as the number of moles of substrate that a mole of
catalyst can convert before becoming inactivated. An ideal catalyst would have an infinite
turnover number, because it wouldn't ever be consumed, but in actual practice one often sees
turnover numbers which go from 100 up to 40 million

𝑚𝑜𝑙𝑒𝑠 𝑜𝑓 𝑝𝑟𝑜𝑑𝑢𝑐𝑡
𝑇𝑂𝑁 = (1. )
𝑚𝑜𝑙𝑒𝑠 𝑜𝑓 𝑐𝑎𝑡𝑎𝑙𝑦𝑠𝑡

Turnover frequency or TOF is defined as the number of molecules reacting per active
site per second (s-1).

𝑇𝑂𝑁
𝑇𝑂𝐹 = (1. )
𝑡𝑖𝑚𝑒

The next important parameter determining the performance of catalyst is selectivity for
a particular product. The selectivity of a product X can be defined as

5|Page
 𝑀𝑜𝑙𝑒𝑠 𝑜𝑓 𝑝𝑟𝑜𝑑𝑢𝑐𝑡 𝑋
𝑆𝑒𝑙𝑒𝑐𝑡𝑖𝑣𝑖𝑡𝑦 𝑜𝑓 𝑋 = × 100% (1. )
𝑇𝑜𝑡𝑎𝑙 𝑚𝑜𝑙𝑒𝑠 𝑜𝑓 𝑎𝑙𝑙 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠

Up to a few years ago, there was no simple method to calculate the efficiency of a
catalytic cycle (i.e. TOF), from a theoretically obtained energy profile. In 2006 onwards,
Kozuch and Shaik have introduced a method so-called energetic span model, which enabled
evaluating TOFs in a straightforward manner. [4] The formulation of this model is based on
Eyring’s transition state theory (TST) and corresponds to a steady-state regime. The
conclusions derived by the authors from this model is that in catalytic cycles there is no rate
determining step, but rate determining states.[Error! Bookmark not defined.]

1.2 Palladium-catalysed C-C cross-coupling reactions

“Carbon-carbon bond formation reactions are important processes in chemistry, because

they provide key steps in building complex, bio-active molecules developed as medicines and
agrochemicals. They are also vital in developing the new generation of ingeniously designed
organic materials with novel electronic, optical or mechanical properties, likely to play a
significant role in the burgeoning area of nanotechnology” – Akira Suzuki, Nobel Lecture
2010.

The C–C cross-coupling reactions are one of the most essential and valuable reactions in
modern day chemistry, since a large number of complex compounds can be synthesized from
easily available reactants. These reactions consist in the carbon-carbon bond formation
between an organic electrophile (R1–X) and an organometallic nucleophile (R2–m) in the
presence of a metal catalyst [M]. (scheme 1.2) Several named reactions viz. Suzuki-Miyaura,
Stile, Negishi, Hiyama may be performed by varying the nucleophile used in catalysis.

Scheme 1.2 General scheme for C–C cross-coupling reactions

6|Page
Among various metals, palladium catalysts have provided a plethora of new methodologies
for synthetic organic chemistry. Palladium-catalysed cross-coupling reactions are among the
most powerful synthetic tools routinely used in the synthesis of pharmaceuticals, fine
chemicals, and advanced materials.[6] The palladium-catalysed cross-coupling of aryl halides
(or halide analogues) with nucleophiles has been firmly established as one of the most
important methods available for carbon-carbon and carbon-heteroatom bond formation.
Palladium-catalysed cross-coupling reactions can be separated into three distinct classes:
those involving transmetallation, Hartwig-Buchwald type heteroatom crosscouplings, and the
Heck reaction. The general catalytic cycle of these coupling reactions are shown below
(scheme 1..)

Scheme 1. Pd catalyzed cross coupling reactions

The starting point of a classical cross-coupling reaction, for example a Suzuki–Miyaura

coupling, is the oxidative addition step followed by transmetalation, cis–trans isomerization,
and finally the reductive elimination.(Scheme 1.). [7] Till date, numerous experimental
studies have been dedicated to providing mechanistic insight into the Suzuki type C-C/C-X
bond formation reactions. [3]

7|Page
 1.2.1 Oxidative addition

When addition of ligand is accompanied by oxidation of the metal, then it is called an

oxidative addition reaction in which oxidation state of the metal increases by 2 units as well
as coordination number increases by 2 units. In this process, 2 new anionic ligands are added
to the metal.

The requirements for an oxidative reaction are:

 The metal should have available non-bonded electron density
 The metal complex must be coordinatively unsaturated i.e. there must be two vacant
coordination sites on the reacting complex
 The metal should have stable oxidation states separated by two units and the higher oxidation
state must be energetically accessible and stable. In other words, oxidative addition reactions
are most facile when there is a good two-electron redox couple. In other words, both the
starting and final oxidation states are relatively stable. For example, oxidative addition from
Ir(I) to Ir(III) is common but an oxidative addition from Fe(III) to Fe(V), while possible, is
generally unlikely.

Oxidative additions proceed via many pathways that depend on the metal center and the
substrates. Four distinct mechanisms are reported in the literature: a) concerted mechanism b)
SN2 mechanism c) radical mechanism and d) ionic mechanism (Scheme 1. )

8|Page
Scheme 1. : Reaction mechanisms for the oxidative addition of an A–B molecule to a
metal complex

In the concerted mechanism case the A–B molecule binds firstly to the metal center and then,
the cleavage of the A–B bond and the formation of the new M–A and M–B bonds take place
simultaneously through a three centered transition state. This mechanism is normally found in
oxidative additions of non-polar reagents and aryl halides.

On the other hand, the SN2 mechanism is an associative bimolecular process that consists in
two steps first, the ligand A is attacked by a metal electron pair and the anionic ligand B− is
expelled giving rise to the cationic species [LnM–A]+; subsequently, the two charged species
collapse to yield the oxidative addition product. Unlike the concerted mechanism, the relative
position of the ligands A and B in the final product via this mechanism can be either cis or
trans, depending on where the anionic ligand B ends up coordinated after the second step.
This SN2 mechanism is often found in oxidative additions of polar reagents and in polar
solvents, and results in the inversion of configuration of a stereogenic cente.

The third mechanism is radical mechanism also known as known as nonchain radical
mechanism. This non-chain variant is generally believed to operate in oxidative additions of
certain alkyl halides and consists in the one-electron oxidationof the metal by the A–B
molecule giving rise to the radical species [LnM–A]· and B·, which rapidly recombine to
yield the oxidative addition product [LnM(A)(B)]. Similar to the SN2 mechanism, the rate of
this radical mechanism increases as more basic is the metal, and the more easily the ligand A
is transferred to the metal. Experimental evidences for this type of mechanism can be the
significant changes in the reaction rate produced by the introduction of slight modifications
of the substrate, the metal complex, or the solvent. Another alternative to confirm this
mechanism is to use radical scavengers, such as RNO·.

This ionic mechanism adopted for A–B molecules are completely dissociated in solution, and
consists in the consecutive coordination of the two charged ligands to the metal complex.
9|Page
Thus, this mechanism can evolve through two possible variants. The first one involves the
initial coordination of the cation A+ to the metal complex and the subsequent coordination of
the anionic ligand B−. On the contrary, in the second alternative the coordination of the
anionic ligand B− to the metal complex occurs first, followed by the coordination of the
cation A+. In general, the first variant is the most common one and is favored by basic
ligands and a low oxidation state of the metal.

In a nutshell, oxidative addition is highly sensitive to the nature of the reagent A–B. Thus,
depending on the reagent one or another mechanism might be favoured. Other factors that
can affect this process are, for instance, the nature of the solvent, the metal-bound ligands or
the added additives.

1.2.2 Transmetallation

Transmetalation reactions consist in the transfer of an organic group R (e.g. alkyl,aryl, vinyl)
from one metal (M1) to another one(M2). Thus, these reactions entail the cleavage of a M1–
C bond and the concomitant formation of a new M2–C bond. Regarding the metal centers,
these can be both transition metals and main group elements. As a result, depending on the
type of the metal atoms involved in transmetalation we can distinguish between three
different processes: (i) organic ligand transfer between two main group elements; (ii) that
between one main group element and one transition metal; and (iii) transmetalation involving
two transition metals.

Scheme 1. : Mechanisms proposed for transmetalation reaction

The redox-type mechanism proposed consists of intermolecular transfer of the organic ligand
accompanied by the oxidation and reduction of the metal centers involved. The
thermodynamics of this type of transmetalation reactions is governed by the relative stability
of the metalcarbon bonds of the reactant and product.

In the second category of that classification we have those transmetalation reactions that
evolve through a metal exchange mechanism. As its name suggests, this mechanism entails
the metal-ligand exchange between an organometallic compoundM1–R and ametal complex
with a halogen or pseudo-halogen ligandM2–X (e.g. X = Cl, Br, CN). This type of
transmetalation is the most common among the three types of reactions and is characteristic

10 | P a g e
of cross-coupling reactions. On the other hand, the exchange reaction of alkyl or aryl ligands
between two organometallic compounds (i.e. X = R’) is also included within this category.

Lastly, within the third type of transmetalation reactions we find those that follow a
mechanism dubbed as “ate” complex mechanism. In this mechanism, the transfer of the
organic ligand from M1 to M2 gives birth to an ion-pair formed between the metal cation and
the anionic organometallic complex. This mechanism differs from the metal exchange
mechanism in that it only involves the transfer of the organic group from M1 to M2, and not
the simultaneous intermolecular transfer of both the organic and the halogen (or pseudo-
halogen) ligands between M1 and M2.

Given that the carbon-bonded ligand does not easily dissociate, all the above mentioned
reaction mechanisms involve transition states (or intermediates)where the organic group acts
as a bridging ligand connecting the two metal centers.

1.2.3 Reductive elimination

Reductive elimination is just the opposite of oxidative addition i.e. this step involves decrease
in the oxidation and coordination number by two units.

1.3 C-F bond activation

“Will fluorine ever have practical applications?

It is very difficult to answer this question. I may, however, say in all sincerity that I gave this
subject little thought when I undertook my researches, and I believe that all the chemists
whose attempts preceded mine gave it no more consideration.

A scientific research is a search after truth, and it is only after discovery that the question of
applicability can be usefully considered.”

― Henri Moissan, Proceedings of the Royal Institution (1897). In Annual Report of the
Board of Regents of the Smithsonian Institution to July 1897 (1898), 261.

Organofluoro compounds continue to draw attention both for their curious chemistry and for
their wide variety of applications since the first isolation of elemental fluorine by Henri
Moissan in 1886. [] Fluorinated organic compounds are widely used in applications related to
material sciences, polymer chemistry and agrochemical industries. Even after nearly 150
years of the first discovery of organofluro compounds by Alexander Borodin [12] these
compounds continue to draw attention both for their curious chemistry and for their wide
variety of applications.[] Such compounds can in principle be made from functionalization of
the corresponding perfluoroalkanes and this led to an increasing attention in the chemistry of
C-F bonds in the last few years. This area shares certain aspects with the related but older
problem of C-H bond activation[] of hydrocarbons, but an important difference is that the
11 | P a g e
enhanced value on going from the perfluorocarbon to the corresponding perfluorinated
functional derivative is often much greater than is the case for hydrocarbons and so makes it
practical to use more sophisticated reagents.

1.4 Basic Introduction of Quantum Mechanics

Classical mechanics provides a satisfactory description of the mechanical phenomena

in the macroscopic world that we can experience directly. However, in the late 19th century it
became apparent that Newtonian mechanics and Maxwell's electromagnetic theory the two
main branches of classical mechanics were unable to explain some experimental facts
concerning microscopic particles. In order to explain these experimental facts Planck,
Einstein, and Bohr independently combined the classical concepts with the so-called quantum
hypothesis which led to the development of quantum mechanics. Quantum mechanics was
developed by addressing two main aspects: i) the wave character of the particles and ii) the
probabilistic nature of our observations in microscopic realm.

In 1920 Neils Bohr formulated the correspondence principle which states that quantum
mechanical results must be equal to that of classical mechanics when the system becomes
very large. In a way this principle says that classical mechanics is a special case of quantum
mechanics i.e. through classical mechanics is valid for macroscopic bodies, we expect
quantum mechanics to give the same results as classical mechanics for macroscopic bodies
having uniform probability density.

1.4.1 The Schrödinger equation

In 1926, Erwin Schrödinger proposed a differential equation relating the energy of a

micro system to its space and time coordinates.

𝐻𝜓(𝑥, 𝑡) = 𝐸𝜓(𝑥, 𝑡) (1. )

In differential form the equation can be expresses as

𝜕𝛹(𝑥, 𝑡)
𝑖ħ = 𝐻𝛹(𝑥, 𝑡) (1. )
𝜕𝑡

12 | P a g e
 In this equation, ψ(x,t) is the wave function and H is a differential operator called the
Hamiltonian operator which describes the energy dependence of the wave function. It is a
postulate of quantum mechanics that the state of a system is fully described by its wave
function. Since ψ(x,t) can fully describe a system, so solving this equation gives information
about all the molecular properties of the system. If the Hamilton operator does not depend on
time, as is the case in the investigations described in this thesis, the above equation reduces to
the time-independent Schrödinger equation:

𝐻𝜓(𝑥) = 𝐸𝜓(𝑥)

The central purpose in computational chemistry is to solve this Schrödinger equation

for different chemically relevant systems.

1.4.2 The Born-Oppenheimer approximation and the concept of electronic structure:

The molecular Schrödinger equation is extremely complicated and its exact solution is
not possible.

The total Hamiltonian for a molecular system is given by,

𝐻𝑡𝑜𝑡 = 𝑇𝑡𝑜𝑡 + 𝑉𝑡𝑜𝑡 = (𝑇𝑒 + 𝑇𝑁 ) + (𝑉𝑁𝑒 + 𝑉𝑒𝑒 + 𝑉𝑁𝑁 )

Where TN stands for kinetic energy operator for the nuclei, Te stands for kinetic energy
operator for the electrons, Vee is the electron-electron repulsion term, VeN is the election-
nuclear attraction term and VNN is the nuclear-nuclear repulsion term.

If the molecule contains Nn-nuclei and Ne-electrons then in atomic unit

2 Nn
 2k
𝑇𝑁 = 
2

k 1 m k
electron j
electron i rij
rl i rkj -e, m
-e, mrki rl j
Ne
2
T e= 
2 me



1
2 nucleus k nucleus l

+Zk e +Zl e

13 | P a g e
 Ne Ne
e2
Vee=   
 1    40 | r  r |

Ne Nn
Zk e2
VeN=    
 1 k 1 40 | r  Rk |

Nn Nn
Zk Z je2
VNN=   
j 1 k  j 4 0 | R j  Rk |

Since the electrons are much lighter than the nuclei, they move faster in a molecule.
The electron carries out many cycles of motion in the time it takes the nucleus to move a
short distance. Calculation shows that the nuclei move only about 1 mm, while in the same
time, the electron travels through a distance of about 1 m. Therefore, we can consider the
nuclei to be almost fixed while the electron moves through the whole volume of the
molecule. Thus we can separate the Schrödinger equation for a molecule into two separate
equations, one depending upon the electronic motion and the other on the static nuclear
position. This approximation is known as the Born-Oppenheimer approximation. The nuclei
are considered fixed and only the electronic part is actually solved. In the Eq.1.3 the TN
operator is not affected by the electronic motion. Since the potential energy VNN due to
nuclear-nuclear repulsion is the constant quantity for a fixed inter-nuclear distance, the purely
electronic Hamilton(He) can be written as:

𝐻𝑒 = 𝑇𝑒 + 𝑉𝑒𝑁 + 𝑉𝑒𝑒

The Schrödinger equation for only electronic motion is:

(𝐻𝑒 + 𝑉𝑁𝑁 )𝜓𝑒 (𝑅, 𝑟) = (𝐸𝑒 + 𝑉𝑁𝑁 )𝜓𝑒 (𝑅, 𝑟) = 𝑈𝜓𝑒 (𝑅, 𝑟)

where U=(Ee+VNN)

Since VNN is a constant quantity, we can write

𝐻𝑒 𝜓𝑒 (𝑅, 𝑟) = 𝐸𝑒 𝜓𝑒 (𝑅, 𝑟)

where Ee is the electronic energy, ψe is the corresponding wavefunction and U is the

total energy of the nuclei and electrons. R and r denote the nuclear and the electronic
14 | P a g e
coordinates respectively.This approximation is valid as long as the various energy levels such
as electronic, vibrational and rotational are widely separated from one another. The Born-
Oppenheimer approximation leads to the idea of potential energy surface U= (Ee+VNN)

1.4.3 Potential Energy Surface

The potential energy surface (PES) is a key concept in computational chemistry. A

potential energy surface is a mathematical relationship connecting molecular structure and its
corresponding potential energy. A complete potential energy surface(PES), which is a
hypersurface with 3N-6 degrees of freedom for a non-linear N atomic molecule and 3N-5 for
linear molecules can provide sufficient predictions about the possible structures of a given
system, their relative energies, and stability.

15 | P a g e
 O + HCl  OH + Cl

first order
saddle point
Energy

Figure 1. A 3D-representation of an arbitrary PES.

The structure corresponding to the point on the PES at which the first derivative of the
energy with respect to each geometric parameter(gradient) is equal to zero and all the second
derivatives (Hessian eigenvalues) are positive is called a minimum. There can be two types of
minima, a global minimum, which is a structure with the lowest possible energy and local
minima that are structures having higher energies than the global minimum. First-order
saddle points (maxima along one coordinate and minima along all other directions) are other
important points on PESs at which the gradient is zero but one and only one of the Hessian
eigenvalues is negative, physically which means, there is one unique vibrational mode that

16 | P a g e
when followed leads downhill to an energy minimum by relaxing the system. These are also
called transition states, which provide lowest energy barriers between two different minima.
To locate a minimum on PES one needs to perform the process of structure optimization i.e.
to locate the lowest energy of a system with respect to the change of structural coordinates(R
and r).

1.5 Ab initio methods

Finding an exact solution for the Schrödinger Wave Equation(SWE) is computationally
not possible for most systems. Many mathematical approximation methods are developed
over the years. Ab-initio (Latin for ‘from the beginning') methods are one of the most
important approaches developed for molecular systems which employ the laws of quantum
mechanics to solve the Schrödinger wave equation. Ab-initio calculations are done directly
from theoretical principles with no inclusion of experimental data.

1.5.1 Hartree-Fock Self-Consistent-Field method

The Hartree-Fock method is the most common type of ab-initio calculation method.

The HF method also called the self-consistent field method (SCF) is based on the variational

principle. In HF, the primary approximation is that the effect of other electrons on the

calculated one electron is accounted for in a central mean-field i.e. Columb electron-electron

repulsion is taken into account by integrating the repulsion term. This gives the average effect

of the repulsion, but not the explicit repulsion interaction. The Hartree-Fock energy is an

upper bound of the exact energy and tends to a limiting value called the Hartree-Fock limit as

the basis set is improved. The advantage of HF is it transforms the many-electron

Schrödinger equation into many one-electron equations. Each one electron equation can be

solved to obtain one-electron wave function called an orbital and thus the total electronic

wavefunction is described as a single Slater determinant, an antisymmetrized product of one-

17 | P a g e
electron wave function. The HF wave function of N spin orbitals is given by the Slater

determinant

1 ( x1 )  2 ( x1 ) . . N ( x1 )
1 ( x2 ) . . .  N ( x2 )
1
ѰHF = . . . . .
N!
. . . . .
1 ( x N ) . . .  N (xN )

where x i includes the coordinates of space and spin.

The fundamental principle that enables us to obtain an approximate solution to the

Schroedinger Equation is the Variational Principle: E = <Ѱ|H|Ѱ>  E0. We vary the spin-

orbitals themselves in order to minimize the value of E, obtaining a quality approximation to

the ground state energy. Substituting the Hartree-Fock wavefunction into the expectation

value, we get,

EHF = <ѰHF |H|ѰHF >

N
1 N N 1 1
=     | h |      [    | |          | |    ]  Vnn
 1 2  1  1 r12 r12

Nn
1 Z
Where h    2   k is a one electron operator containing the electron kinetic
2 k 1 rk

energy and the electron-nucleus attraction energy The double integrals in the double

summation are the Coulomb and Exchange integrals, respectively. The antisymmetry

requirement makes the Exchange integral appear. The variational principle, applied to the

above equation yields the following equation that the spin-orbitals must satisfy:

{h  VHF ( x  )}      

N
VHF ( x  )   [ J ( x  ) K ( x  )]
 1

18 | P a g e
 dx2
J ( x1 )   |  ( x2 ) | 2
r12

dx2
K ( x1 )   ( x1 )  [  ( x2 ) *   ( x2 ) ] ( x1 )
r12

where the effective Hartree-Fock potential VHF is a operator containing both a Coulomb

(J) and a non-local Exchange operator (K) added over all electrons (the mean field). These

equations represent a fictitious system of non-interacting particles that approximates the

electron behavior. Furthermore, if we assume that upon removal of a given electron

(ionization) from the system, that the distribution of the rest of the electrons does not change

(which is not true), then the energy required to ionize the system from the orbital Ѱ0 is

precisely the eigenvalue E0. This is a statement of Koopman’s theorem and provides a

physical interpretation for the orbital eigenvalues

In practice, the orbitals are approximated by a linear combination of the basis functions.

The Hartree-Fock method satisfies the exchange requirements perfectly, but misses badly in

the rest of the electron correlation effects. To obtain chemical accuracy, post-Hartree-Fock

methods is used, such as Moller-Plesset perturbation theory, Configuration Interaction, or

Coupled Cluster methods.

References:

 Szabo, A.; Ostlund, N. S. (1996). Modern Quantum Chemistry. Mineola, New York:

Dover Publishing. ISBN 0-486-69186-1.

 Hinchliffe, Alan (2000). Modelling Molecular Structures (2nd ed.). Baffins Lane,

Chichester, West Sussex PO19 1UD, England: John Wiley & Sons Ltd. p. 186. ISBN

0-471-48993-X.

 Levine, Ira N. (1991). Quantum Chemistry (4th ed.). Englewood Cliffs, New Jersey:

Prentice Hall. p. 403. ISBN 0-205-12770-3

19 | P a g e
1.5.2 Post Hartree-Fock method:Electron correlation

When one electron moves, the positions of the other electrons in the molecule adjust to
minimize electron-electron repulsion. Therefore, the motions of the electrons are not
independent but are correlated. One of the limitations of HF calculations is that in HF
electon-electron repulsions are only averaged i.e. it do not include exact electron correlation
and therefore the calculated energy value is slightly higher than the true energy.[64 from BM]
Mainly there are methods to incorporate electron correlation into the wavefunction viz.
configuration interaction,[65] Møller-Plesset perturbation theory,[66] and coupled cluster
theory.[67] These so called post-HF method begin with HF approximation and then corrected
for correlation.
𝐸𝑐𝑜𝑟𝑟 = 𝐸𝑛𝑜𝑛𝑟𝑒𝑙 − 𝐸𝐻𝐹
The correlation energy (Ecorr) is the difference between the exact nonrelativistic energy
(Enonrel) and the HF energy (EHF). Usually, post-HF methods give more accurate results than
HF calculations, although the added accuracy comes with the price of additional
computational cost.
References:

 R. McWeeny, Int. J. Quant. Chem. 1967, 15, 351-359.

 (a) A. C. Hurley, Electron Correlation in Small Molecules, Academic Press, London,

1977; (b) S. Wilson, Electron Correlation in Molecules, Clarendon Press, Oxford,
1984; (c) K. Raghavachari, J. B. Anderson, J. Phys. Chem. 1996, 100, 12960-12973.
 C. Møller, M. S. Plesset, Phys. Rev.1934, 46, 618-622

 M. R. Hoffmann, H. F. Schaefer, Adv. Quantum Chem. 1986, 18, 207-279

1.5.2.1 Configuration Interaction

The contemporary best method for adjusting the molecular wave functions for electron
correlation is the configuration interaction (CI). CI calculations allow the ground state of a
molecule to mix with excited state of the molecule. For example, for the H2 molecule, the

20 | P a g e
ground state configuration and several excited state are shown below:

In CI calculations the ground state wave function is allowed to be a linear combination of all
of the possible excited states also called as configuration state functions (CSFs)
𝛹𝐶𝐼 = ∑𝑖=0 𝐶𝑖 𝛷𝑖 = 𝐶0 𝛷𝐻𝐹 + ∑𝑆 𝐶𝑆 𝛷𝑆 + ∑𝐷 𝐶𝐷 𝛷𝐷 + ∑𝑇 𝐶𝑇 𝛷𝑇 +∙∙∙∙∙∙∙∙
The advantage of mixing in excited state character to the ground state is that excited states
have nodes between the nuclei, which help electrons to avoid each other. The mixing of
ground state wave function and excited state wave function is done variationally. The
coefficients for each excited state configurations adjusted to give the minimum ground state
energy. This approach leads to techniques called CISD or CISDT. CISD includes single and
double excitations. CISDT includes single, double and triple excitations. The single
excitation contributes most in CISD calculations.
References:
 Cramer, Christopher J. (2002). Essentials of Computational Chemistry. Chichester:
John Wiley & Sons, Ltd. pp. 191–232. ISBN 0-471-48552-7.
 Sherrill, C. David; Schaefer III, Henry F. (1999). Löwdin, Per-Olov, ed. The
Configuration Interaction Method: Advances in Highly Correlated Approaches.
Advances in Quantum Chemistry. 34. San Diego: Academic Press. pp. 143–
269. ISBN 0-12-034834-9. doi:10.1016/S0065-3276(08)60532-8

1.5.2.2 Multi configuration self-consistent field method

A multi-configuration self-consistent field (MCSCF) is a combination of an SCF

computation with a full CI involving a subset of the orbitals. This subset is known as the

21 | P a g e
active space. In multi-configuration theory the wavefunction is represented by more than one

Slater determinant, also called configurations.

𝛹𝑒 = 𝐶0 𝛹0𝐻𝐹 + ∑ 𝐶𝑖 𝛹𝑖𝑒𝑥𝑐
𝑖=1

References:

 McWeeny, Roy (1979). Coulson's Valence. Oxford: Oxford University Press. pp.
124–129. ISBN 0-19-855145-2.
 Jensen, Frank (2007). Introduction to Computational Chemistry. Chichester, England:
John Wiley and Sons. pp. 133–158. ISBN 0-470-01187-4.
 Cramer, Christopher J. (2002). Essentials of Computational Chemistry. Chichester:
John Wiley & Sons, Ltd. pp. 191–232. ISBN 0-471-48552-7.

1.5.2.3 Møller-Plesset perturbation method

Møller-Plesset(MP) methods use a perturbation operator to introduce electron

correlation effects. The HF result is treated as first-order perturbation. The residual

correlation part is treated by a series of corrections including the excited configurations and

thus we get different order method viz. MP2, MP3, MP4 and so on. The accuracy of an MP4

calculation is roughly equivalent to that of a CISD calculation. MP5 and other higher order

MP methods are very rarely used due to high computational cost. Møller-Plesset(MP)

calculations are not variational. So, MP2 calculations sometimes give energies below the

actual total energy.

In general, ab initio calculations give very good qualitative results and can yield

increasingly accurate quantitative results as the molecules in the question become smaller.

However, computationally, these methods are very expensive, since it often takes enormous

amount of computer CPU time, memory and disk space. Hartree-Fock theory remains the

cornerstone of ab initio methods, while post-HF methods seek to improve the foundational
22 | P a g e
result of the HF formulism by the inclusion of electron correlation effects Moreover, the

Hartree-Fock method formally scales as N4, where N is the number of basis functions,

meaning that a calculation ten times as big takes ten thousand times as long to complete.

References:

 Head-Gordon, Martin; Pople, John A.; Frisch, Michael J. (1988). "MP2 energy

evaluation by direct methods". Chemical Physics Letters. 153 (6): 503–506.

Bibcode:1988CPL...153..503H. doi:10.1016/0009-2614(88)85250-3.

 Pople, J. A.; Seeger, R.; Krishnan, R. (1977). "Variational configuration interaction

methods and comparison with perturbation theory" (abstract). International Journal of

Quantum Chemistry. 12 (S11): 149–163. doi:10.1002/qua.56012082

 Krishnan, Raghavachari; Pople, John A. (1978). "Approximate fourth-order

perturbation theory of the electron correlation energy". International Journal of

Quantum Chemistry. 14 (1): 91–100. doi:10.1002/qua.56014010

 Raghavachari, Krishnan.; Pople, John A.; Replogle, Eric S.; Head-Gordon, Martin

(1990). "Fifth order Moeller-Plesset perturbation theory: comparison of existing

correlation methods and implementation of new methods correct to fifth order". The

Journal of Physical Chemistry. 94 (14): 5579–5586. doi:10.1021/j100377a033

1.6 Semiempirical Methods

Semi-empirical methods differ from ab-initio in that it uses experimental parameters to

simplify calculations. Approximate solutions to the Schrödinger equation are computed based

in part on experimental parameters available for a specific type of chemical system. Typically

based on the extended Hückel Theory, semi-empirical methods are differentiated by their

23 | P a g e
native sets of parameters. While the use of parameters reduces computational cost, this

practice may give results that are much less quantitatively correct.

References

 E. Hückel, Zeitschrift für Physik, 70, 204, (1931); 72, 310, (1931); 76, 628 (1932);

83, 632, (1933)

 Hückel Theory for Organic Chemists, C. A. Coulson, B. O'Leary and R. B. Mallion,

Academic Press, 1978.

 Andrew Streitwieser, Molecular Orbital Theory for Organic Chemists, Wiley, New

York, (1961)

 R. Hoffmann, Journal of Chemical Physics, 39, 1397, (1963)

 Ira Levine, Quantum Chemistry, Prentice Hall, 4th edition, (1991), pg 579–580

 J. Pople and D. Beveridge, Approximate Molecular Orbital Theory, McGraw–Hill,

1970.

 R. Pariser and R. Parr, Journal of Chemical Physics, 21, 466, 767, (1953)

 Michael J. S. Dewar & Walter Thiel (1977). "Ground states of molecules. 38. The

MNDO method. Approximations and parameters". Journal of the American Chemical

Society. ACS Publications. 99 (15): 4899–4907. doi:10.1021/ja00457a004

1.7 Density Functional Theory

Density functional theory (DFT) is currently the most widely utilized approach to

determine the electronic structure of molecule. Hohenberg and Kohn in 1964 originally

developed this theory. Walter Kohn was awarded with the Nobel Prize in Chemistry in 1998

for his development of the density functional theory.

The Hohenberg-Kohn Theorems

This theorem is the basic building block of the Density Functional Theory which states

that the energy of a system can be derived from its electron density i.e. ground state

properties of a molecular system can be obtained without explicitly constructing the very
24 | P a g e
complex wavefunction. In other words, for a given set of nuclear coordinates, the electron

density uniquely determines the energy and all properties of the ground state of a molecule.

The electron density[𝜌], for any molecular system, is a function of only three spatial

coordinates

Thus, the expectation value for the energy of the system can be written as:

𝐸[𝜌] = 𝑇[𝜌] + 𝑉𝑒𝑒 [𝜌] + 𝑉𝑛𝑒 [𝜌]

where 𝑇[𝜌] is electronic kinetic energy term, 𝑉𝑒𝑒 [𝜌] and 𝑉𝑛𝑒 [𝜌] are electron-electron
repulsion and electron-nuclear attraction respectively. If we separate the terms that are
dependent on the external potential 𝑉𝑒𝑥𝑡 (𝑟⃗) and those which are universal (independent on N,
R, Z), we get

𝐸[𝜌] = 𝑇[𝜌] + 𝑉𝑒𝑒 [𝜌] + 𝑉𝑛𝑒 [𝜌] = 𝐹𝐻𝐾 [𝜌] + ∫ 𝜌𝑟 𝑉𝑛𝑒 𝑑𝑟

𝐹𝐻𝐾 [𝜌] = 𝑇[𝜌] + 𝑉𝑒𝑒 [𝜌]

where the independent terms have been collected into a new quantity, the Hohenberg-Kohn
functional, FHK[ρ].

The above equation says that if we know FHK[ρ] we can solve the Schrödinger equation
exactly. This functional is completely independent on the system and can be applied equally
for any system. Development of this functional is one of the main areas of research in
theoretical chemistry now-a-days.

The applications and advantages of DFT methods are numerous. The accuracy achieved

by DFT calculations is often comparable to more expensive post Hartree-Fock methods,

while the computational costs are comparable to Hartree-Fock calculations.The total energy,

derived from the solution of the Schrödinger equation is the sum of kinetic energy, T,

electron-nuclear attraction, Ven and electron-electron interactions (the nuclear-nuclear

repulsion energy is constant for a given geometry), Vee. Electron-electron interaction consists

of the classical Columbic repulsion (J) and non-classical terms due to the electron correlation

25 | P a g e
and exchange effects. The Kohn-Sham formulation of DFT (KS-DFT) uses the kinetic energy

of non interacting electrons, which can be solved exactly. The energy difference between real

and non-interacting systems is included in the exchange correlation functional. Since the

exact functionals for DFT exchange and correlation functionals are known only for a free

electron gas, approximations such as the local density approximation (LDA) and generalized

gradient approximation (GGA) are used for molecular calculations. An LDA functional

depends only on the density at the coordinates where the functional is evaluated. Although

GGA is local (like LDA) it also considers the gradient of the density and hence GGA

functionals are considered to be more accurate than LDA functionals.Based on the

combination of exchange and correlation functionals used either a `pure’ or a `hybrid’ DFT is

obtained. `Pure’ DFT functionals results when just exchange and correlation functionals are

combined. ‘Hybrid’ DFT functionals result when HF exchange is added to the local DFT

exchange terms. In this project, the hybrid B3LYP Hamiltonian is the DFT method of

choice.The methods discussed are used in the evaluation of the electronic energy for fixed

nuclear positions as well as in the construction of potential energy surfaces (PESs). Gradient-

based geometry optimization techniques are used to obtain stationary points on the PESs.

Vibrational frequency calculations characterize the nature of the stationary points as minima

(no imaginary frequency) or transition state (one imaginary frequency), and the frequencies

thus obtained are used to calculate enthalpies and free energies.

The failings of DFT are well documented and are still problematic within the current
generation of functionals. Furthermore, DFT also lacks a way to systematically converge on
the exact solution unlike wavefunction based methodologies. These issues arise from the
inadequate treatment of electron self-interaction and the exchange correlation
functionals.Energy states with different spin multiplicities, excited states in charge transfer
systems and odd charged electron systems are particularly dependent on the DFT functional
used. The inclusion of HF exchange energy in hybrid functionals has been found to improve
the modeling of such systems that favor high spin configurations. Another problematic issue
is the inability to describe weak to medium-range interactions such as van der Waals forces.
Dispersion forces are more pronounced in large molecules, such as the

26 | P a g e
organometalliccomplexes commonly used in catalysts. Several approaches to overcome the
inability to describe dispersion have been carried out over the years. Truhlar and Zhao have
empirically fitted functionals designed to better describe van der Waals forces. 38,40 These
functionals are called the Minnesota family of functionals (M06, M06-L, M06-2X, etc.) and
are commonly employed in catalyst design. Grimme and coworkers have implemented and
parameterized a correction to the energy to account for weak interactions. This correction is
denoted by a “-D” on the end of functional abbreviations (e.g., B3LYP-D).Since DFT has no
systematic way for improvement, extensive calibrations were carried out testing various
functionals and basis sets to ensure the level of theory utilized is adequate. (intro-5 references
page no 15)
1.8 Basis set

A basis set is a set of mathematical functions designed to describe the orbitals in an

atom. Molecular orbital is expressed as a linear combination of basis functions, the
coefficients for which are determined by the solution of the HF SCF equations. The full HF
wave equation is expressed as a Slater determinant formed from the individual occupied
MOs.

 i ( x1 )  j ( x1 ) ...  K ( x1 )
 i ( x2 )  j ( x 2 ) ...  K ( x 2 )
  ( N!) 1 / 2
   
 i ( x N )  j ( x N ) ...  K ( x N )

 i ( x j )   i (rj ) ( j )

For RHF wave functions

K
 i   c i 
 1

Where {μ} are a set of known functions

For UHF wave-functions two sets of coefficients are needed:

27 | P a g e
 K

i   ci  
 1 
K the same basis functions are used for  and  orbitals
 i   c i  
 

 1



Generally atomic orbitals are represented either by Slater functions or Gaussian

functions. Slater Type Orbitals (STO) introduced by John C. Slater in 1930 is similar to
atomic orbitals of the hydrogen atom and generally expressed in spherical coordinates in
normalized form as:

𝜑𝑆𝑇𝑂 (𝜁, 𝑛, 𝑙, 𝑚; 𝑟, 𝜃, 𝜑) = N𝑟 𝑛−1 𝑒 −𝜁𝑟 𝑌𝑙𝑚 (𝜃, 𝜑)

Where N is the normalization factor and n, l, m plays the role of quantum numbers; r is the
distance of the electron from the atomic nucleus. 𝑌𝑙𝑚 (𝜃, 𝜑) is called as the spherical
harmonics. ζ is Slater orbital exponent which is a constant related to the effective charge of
the nucleus being shielded by electrons and controls the size of the orbital which means large
ζ gives tight function and small ζ gives diffuse function.Slater functions for 1s, 2s and 2px are:

𝜁 1/2
𝜑1𝑠 = ( ) exp(−𝜁1 𝑟)
𝜋
1/2
𝜁25 𝜁1 𝑟
𝜑2𝑠 =( ) 𝑟𝑒𝑥𝑝(− )
96𝜋 2

1/2
𝜁25 𝜁2 𝑟
𝜑2𝑝𝑥 =( ) 𝑟𝑒𝑥𝑝(− )
32𝜋 2

Advantages of STOs are that the exponential dependence on distance(r) from the nucleus is
very close to the exact hydrogenic orbitals and ensures fairly rapid convergence with
increasing number of functions. Although STOs provide reasonable representations of atomic
orbitals however they are not well suited to numerical (fast) calculations (especially two-
electron integrals). So their use in practical molecular orbital calculations has been limited.

In 1950, S. F. Boys found that STOs could be approximated by a linear combination of

Gaussian Type Orbitals (GTOs) which makes the two electron integral calculation much
easier. Cartesian GTOs are defined as

28 | P a g e
 2
𝜑𝐺𝑇𝑂 (𝛼, 𝑙, 𝑚, 𝑛, 𝑓; 𝑥, 𝑦, 𝑧) = 𝑁𝑥 𝑙 𝑦 𝑚 𝑧 𝑛 𝑒 −𝛼𝑓𝑟

N is the normalization constant, α is the Gaussian orbital exponent and f is the scaling factor
which scale all exponents in the related Gaussians in molecular calculations and take into
account the contraction of basis functions in the molecular environment with respect to the
free atom. Here l, m, n are not quantum numbers L=l+m+n is used analogously to the angular
momentum quantum number for atoms to mark functions as s-type (L=0), p-type (L=1), d-
type (L=2) etc.

In spherical coordinates
2
𝜑𝐺𝑇𝑂 (𝜁, 𝑛, 𝑙, 𝑚, 𝑓; 𝑟, 𝜃, 𝜑) = N𝑟 2𝑛−2−1 𝑒 −𝛼𝑓𝑟 𝑌𝑙𝑚 (𝜃, 𝜑)

In both cases (STO and GTO), the angular dependence of the wavefunction is contained in
the spherical harmonics, where the l and m values determine the type of the orbital. One of
the main short come is GTOs do not resemble closely with the form of real AOs, and hence
they are not treated as eigen functions of the squared angular momentum operator L2.

GTO – uncontracted gaussian function (gaussian primitive)

 GTO - contracted gaussian function (gaussian contraction)

STO=  GTO

The first ten normalized gaussian primitives are:

 2 
3/ 4

g s ( , r )    exp( r 2 )
 

1/ 4
 128 5 
g x ( , r )    x exp( r 2 )
  
3

1/ 4
 128 5 
g y ( , r )    y exp( r 2 )
  
3

1/ 4
 128 5 
g z ( , r )    z exp( r 2 )
  
3

29 | P a g e
 1/ 4
 2048 7 
g xx ( , r )    x 2 exp( r 2 )
 9
3


1/ 4
 2048 7 
g yy ( , r )    y 2 exp( r 2 )
 9
3


1/ 4
 2048 7 
g zz ( , r )    z 2 exp( r 2 )
 9
3


1/ 4
 2048 7 
g xy ( , r )    xy exp( r 2 )
 
3


1/ 4
 2048 7 
g xz ( , r )    xz exp( r 2 )
 
3


1/ 4
 2048 7 
g yz ( , r )    yz exp( r 2 )
 
3


There are 6 possible d-type cartesian gaussians while there are only 5 linearly independent
and orthogonal d orbitals

The 6 d-type gaussian primitives may be combined to obtain a set of 5 d-type functions:

gxy  dxy

gxz  dxz

gyz  dyz

1
2 g zz  g xx  g yy   d z2
2
3
g xx  g yy   d x2  y2
4

The 6-th linear combination gives an s-type function:

g rr  51 / 2 ( g xx  g yy  g zz )  g s

30 | P a g e
In a similar manner, the 10 f-type gaussian primitives may be combined to obtain a set of 7 f-
type functions

The absence of rn-1 pre exponential factor restricts single Gaussian primitives to approximate
only 1s, 2p, 3d, 4f, ... orbitals. However, combinations of Gaussians are able to approximate
correct nodal properties of all atomic orbitals (2s, 3s, 3p,..)

GTOs are substandard to STOs in three ways: a) at the nucleus, the GTO has zero slope; the
STO has a cusp i.e. the behaviour near the nucleus is poorly represented by GTO. b) GTOs
diminish too rapidly with distance i.e. the ‘tail’ behaviour is poorly represented. On the
superior side analytical solutions of GTOs are possible.

STO, GTO and CGTO functions

General basis set classifications

Now-a-days there are lots of basis sets composed of GTOs. The choice of basis functions is
extremely important to balance accuracy and computational cost. In a quantum mechanical
calculation the total number of basis functions that are employed must be specified. There are
common sets of functions that are referred to as a minimal basis set, double zeta basis set,
triple zeta basis set or split valence basis set. In addition, there are add-on sets of basis
functions known as polarization and diffuse functions. These common basis set
classifications are described below.

Minimal Basis Set

31 | P a g e
The smallest of these basis set are called minimal basis sets composed of the minimum
number of atomic orbitals required to represent all of the electrons on each atom. For
example the electron configuration of carbon is 1s22s22p2. Thus, a minimal basis set for
carbon atom would consist of 1s, 2s, and 2px, 2py, and 2pz orbitals, a total of five basis
functions. The three 2p orbitals is often denoted 2p(3).

Since the electronic energy of a molecule decreases and approaches the exact value more
closely as the number of basis functions is increased in a calculation due to the Variation
Principle, it is often useful to construct basis sets that employ more than the minimal number
of basis functions. Due to this, minimal basis sets are not generally used.

Double and Triple Zeta Basis Set

The next one is the group of double and triple ζ which uses more two and three basis
functions for each atomic orbital respectively. For the carbon atom, a double zeta basis set
would consist of two 1s, two 2s, and two each of 2px, 2py, and 2pz orbitals, for a total of 10
basis functions. The two 1s orbitals employed in the basis set are not identical. Rather, they
have different orbital exponents. The same is true for the 2s and 2p orbitals. The basis
functions in the double zeta basis set are denoted 1s, 1sʹ, 2s, 2sʹ, 2p (3), and 2pʹ (3).

The Triple Zeta (TZ) basis set goes even further in increasing the size of the basis set (in an
effort to get even closer to the exact electronic energy). In this case, the triple zeta basis set
employs three basis function of each type occupied in the separated atoms. For the carbon
atom, a triple zeta basis set would consist of three 1s, three 2s, and three each of 2p x, 2py, and
2pz orbitals, for a total of 15 basis functions. The three basis functions of each type employed
in the basis set all have different orbital exponents. The basis functions in the triple zeta basis
set are denoted 1s, 1sʹ, 1sʺ, 2s, 2sʹ, 2sʺ, 2p (3), 2pʹ (3), and 2pʺ (3).

Split Valence Basis Set

Because the core electrons of an atom are less affected by the chemical environment than the
valence electrons, they are sometimes treated with a minimal basis set while the valence
electrons are treated with a larger basis set (of double zeta or higher quality). This is known
as a split valence basis set. A basis consisting of a minimal basis set for the core electrons
and a double zeta basis set for the valence electrons would be called a split valence double

32 | P a g e
zeta basis set. Split valence triple zeta basis sets are also commonly used. Split valence basis
sets are also used for larger molecules because they reduce the computational cost of the
calculations. For the carbon atom, a split valence double zeta basis set would consist of a
single 1s orbital, along with two 2s and two each of 2px, 2py, and 2pz orbitals, for a total of 9
basis functions and are denoted 1s, 2s, 2sʹ, 2p (3), and 2pʹ(3).

Additions to the Basis Set: Polarization Functions

Another way to increase the size of the basis set in order to get closer to the exact electronic
energy and wavefunction is to include polarization functions in the basis set. Polarizations are
needed to improve the flexibility of the orbital basis set for describing angular space. A
polarization function is any higher angular momentum orbital used in a basis set that is not
normally occupied in the separated atom. For example, for the hydrogen atom, the only
orbital type that is occupied is s-type. Therefore, if p-type or d-type basis functions were
added to the hydrogen atom they would be known as polarization functions. For first row
elements like carbon, d-type and f-type basis functions would be considered to be
polarization functions. For transition metals with occupied d-type orbitals, only f-type or
higher functions would be considered polarization basis functions. Polarization functions are
included to improve the flexibility of the basis set, particularly to better represent electron
density in bonding regions.

Additions to the Basis Set: Diffuse Functions

Diffuse basis functions are extra basis functions (usually of s-type or p-type) that are added to
the basis set to represent very broad electron distributions. The orbital exponents of these
diffuse basis functions are very small. They are especially important in representing the
electron density in anions or in intermolecular complexes (where there may be particularly
long bonds with electron density spread over a large region).

Pople basis sets:

The most broadly used basis sets were developed by John A Pople and his group. These basis
sets are of the split valence double zeta and split valence triple zeta types, with optional
polarization and diffuse functions. Pople-style basis sets denoted by the general nomenclature
N-M1G or N-M11G, where N and M are integers. The first, N-M1G, is a split valence double

33 | P a g e
zeta basis set while the second, N-M11G, is a split valence triple zeta basis set. In Pople-style
basis sets, the basis functions are made to look similar to Slater Type Orbitals by representing
each basis function as a linear combinaton of fixed gaussian primitive functions.

For Pople basis sets, the ʺGʺ in the name simply indicates that gaussian primitives are used to
form the basis functions. The dash in the name indicates that the basis set is split valence in
type. If two numbers are listed after the dash (as in 3-21G or 6-31G), the basis set is split
valence double zeta. If three numbers are listed after the dash, the basis set is split valence
triple zeta.

Often polarization functions are also added to Pople-style basis sets. There are two common
methods for designating polarization functions included in a basis set. The first method is to
use * or ** after the Pople basis set name; for example, 6-31G* or 6-31G**. The single *
means that one set of d-type polarization functions is added to each non-hydrogen atom in the
molecule. The double ** means that one set of d-type polarization functions is added to non-
hydrogens and one set of p-type polarization functions is added to hydrogens.

The second method for including polarization functions in the basis set designation is more
general. It is indicated by the notation (L1,L2) following the Pople basis set name; for
example, 6-31G(d) or 6-31G(d,p). The first label indicates the polarization functions added to
non-hydrogen atoms in the molecule. The notations 6-31G(d) and 6- 31G(d,p) mean that one
set of d-type polarization functions is added to all non-hydrogens. The notation 6-311(2df)
means that two sets of d-type and one set of f-type polarization functions are added to non-
hydrogens. The second label in the notation (L1,L2) indicates the polarization functions
added to hydrogen atoms. The basis set 6-31G(d) has no polarization functions added to
hydrogen, while the basis 6-31G(d,p) has one set of p-type polarization functions added to
hydrogen atoms. The basis set 6-311G(2df,2pd) has two sets of p-type and one set of d-type
polarization functions added to hydrogen atoms.

The use of diffuse functions in a Pople basis set is indicated by the notation + or ++. The +
notation, as in 6- 31+G(d), means that one set of sp-type diffuse basis functions is added to
non-hydrogen atoms (4 diffuse basis functions per atom). The ++ notation, as in 6-31++G(d),
means that one set of sp-type diffuse functions is added to each non-hydrogen atom and one
s-type diffuse function is added to hydrogen atoms.

34 | P a g e
The basis sets optimized at the HF level are not optimal for correlated computations. Dunning
and coworkers developed correlation consistent basis sets with correlated configuration
interaction singles and doubles (CISD) wave functions. Those functions are called cc-pVXZ,
X = D, T, Q, 5, 6, 7 meaning the correlation consistent polarized valence X-zeta basis sets.
For the atoms after the third row, a pseudo potential (PP) or a relativistic effective core
potential (RECP) is used to the contribution of the core electrons to the total energy and those
basis functional are known as cc-pVXZ-PP. Scalar relativistic effects can be described by
using correlation consistent basis sets optimized for the Douglas-Kroll-Hess Hamiltonian
noted as cc-pVXZ-DK. The diffuse functions can be added into those basis sets denoted a
prefix of ‘aug’. The weighted core-valence correlation functions are also available for both
PP and DK basis functions known as cc- pwCVXZ-PP or cc-pwCVXZ-DK to differentiate
intra shell (core-core) and inter shell (core–valence) effects. One of the advantages of the
correlation consistent basis sets is that the sequence of basis sets is designed to converge
smoothly to the Complete Basis Set (CBS) limit.

References:

 E. R. Davidson, D. Feller, Chem. Rev. 1986, 86, 681-696.

 J. C. Slater, Phys. Rev. 1930, 36, 57-64
 (a) S. F. Boys, Proc. R. Soc. London Ser. A 1950, 200, 542-554; (b) P. M. W. Gill,
Adv. Quantum Chem. 1994, 25, 141-205.
 R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54, 724-728.
 T. H. Dunning, J. Chem. Phys. 1989, 90, 1007-1023.

1.9 Bonding Analysis

“The Nature of the Chemical Bond is the problem at the heart of all chemistry.” -
Bryce Crawford, Jr. (1953)

35 | P a g e
There are a number of ways of analysing a calculation to provide insight into the
interactions and bonding in molecules. These generally fall into several classes viz.
molecular orbital analysis, population analysis and electron density analysis(EDA).

1.9.1 Molecular orbital analysis

The goal of molecular orbital theory is to describe molecules in a similar way to how we
describe atoms i.e. in terms of orbitals, orbital diagrams, and electron configurations.
Molecular orbitals (or MOs) give us more specific and detailed information about the
electronic interactions occurring in a molecule. All of the molecular orbitals together must
contribute to the electron density. We will also look at the delocalised MOs output from a
standard quantum chemical calculation. We can determine MOs in more than one way, we
will also look at localised MOs. Each molecular orbital is expanded in terms of the atomic
orbital (AO) basis functions (STO's or (contracted) GTO's). The primary function of these
orbitals is twofold, one is as a mathematical basis, and the other by the fact that they are
based on "real" AOs, is to give us information about the atomic contributions to each
molecular orbital.

1.9.2 Population analysis

Although pictures of molecular orbitals are informative, sometimes we need numerical

values which can be easily quantified and compared. Population analysis is the chemist tool
in this regard. Population analysis is a mathematical way of segregating a wave function or
electron density into charges on the nuclei, bond orders, and other related information i.e. it is
the study of charge distribution within molecules These are probably the most widely used
results that are not experimentally observable. Population analysis delivers three important
results: 1) Determination of the distribution of electrons in a molecule 2) Creating orbital
shapes 3) Derivation of atomic charges and dipole (multipole) moments.

1.9.2.1 Mulliken population analysis

“The more accurate the calculations become, the more the concepts tend to vanish into
thin air.”-R. S. Mulliken, J. Chem. Phys. S2, 43 (1965)

Mulliken population analysis is the oldest and most popular method available in nearly every
software program for molecular modelling. Mullikan analysis is fast and simple method for
36 | P a g e
determination of electron distribution and atomic charges. It is based on the linear
combination of atomic orbitals and therefore the wave function of the molecule. The
electrons are partitioned to the atoms based on the nature of the atomic orbitals’ contribution
to the molecular wave function.

Major disadvantage of this method is the strong dependence of the results on the level of
theory (basis set or kind of calculation). Because of this very reason, Mulliken analysis is not
seriously considered in computational techniques.

The key features of this method depend upon the component of the matrix composed of
the elements Dαβ. The diagonal elements Dαα give the interaction of an AO α with itself
(summed over all MOs) while the overlap population of one AO with another is provided
from the off-diagonal elements. In the Mulliken analysis each of the contributing orbital is
assigned with half the overlap population, giving the total population of each AO. The gross
atomic population on a specific atom is now obtained by summing over all the atomic
orbitals.

References:

 Roberto Bochicchio, Robert Ponec, Luis Lain and Alicia Torre; J. Phys. Chem. A
2000, 104, 9130
 (a) R. S. Mulliken, J. Chem. Phys. 1955, 23, 1833-1840; (b) R. S. Mulliken, J. Chem.
Phys. 1965, 43, S2-11.
 A. E. Reed, R. B. Weinstock, F. Weinhold, J. Chem. Phys. 1985, 83, 735-746.

1.9.2.2 Natural Bond-Order (NBO) analysis

37 | P a g e
“One is almost tempted to say... at last I can almost see a bond. But that will never be, for a
bond does not really exist at all: it is a most convenient fiction which, as we have seen, is
convenient both to experimental and theoretical chemists. “

— Charles Alfred Coulson 'What is a Chemical Bond?', Coulson Papers, 25, Bodleian
Library, Oxford. In Mary-Jo Nye, From Chemical Philosophy to Theoretical Chemistry
(1993).

One of the most popular method of population analysis is the Natural Bond Order (NBO)
developed by Weinhold and co-workers. NBO is a calculated bonding orbital with maximum
electron density. This method is based on the theory of “Natural Orbitals” by Löwdin. NBO
classifies atomic orbitals into two distinct groups: NAOs, NBOs. NAOs are made up of basis
sets of single atoms (core, valence and Rydberg) and the NBOs are a combination of basis set
atomic orbitals of two atoms. Hierarchical Atomic/Molecular Orbitals in the NBO analysis
are:

Atomic Orbitals → NAO → NHO → NBO → NLMO → Molecular Orbitals

NAO: Natural Atomic Orbitals

NHO: Natural Hybrid Orbitals

NBO: Natural Bond Orbitals

NLMO: Natural (Semi-) Localized Molecular Orbitals

The general object of NBO methods is to translate accurate calculations into commonly
understood bonding concepts such as atomic charge, Lewis structure, bond type (covalent vs.
dative vs. ionic; σ vs. π, etc.), hybridization, bond order, charge transfer, resonance weights,
sterics and spectroscopic descriptors (e.g., of NMR chemical shielding or J-coupling).
(Eric D. Glendening, Clark R. Landis and Frank Weinhold Natural bond orbital
methods WIREs Comput Mol Sci 2012, 2: 1–42 doi: 10.1002/wcms.51)

In a general sense, the NBO localization method divides NBOs into core, bonding, anti-
bonding and Rydberg (remaining) orbitals. Instead of using molecular orbitals directly, the

38 | P a g e
natural orbitals (NOs) are used as the eigenfunctions of the first-order reduced density
operator Γ

𝛤 = 𝑁 ∫|𝛹(1,2, … , 𝑁)|2 𝑑𝜏2 𝑑𝜏3 … 𝑑𝜏𝑁

Mathematically the eigenvalue equation can be expressed as

𝛤𝜃𝑘 = 𝑝𝑘 𝜃𝑘 (𝑘 = 1, 2, … )

In this equation, the eigenvalue pk represents the population (occupancy) of the eigenfunction
θk for the molecular electron density operator Γ.
There are two parts of the methods: 1.Natural population analysis (NPA) to identify the
population numbers and 2.NBO which is the analysis of the bond order based on the electron
population obtained by NPA. Advantages of NBO are: smaller dependence on the basis set,
better reproducibility for different molecules and orientates itself at the formalism for Lewis
formulas. Moreover, from the resultant natural atomic orbital Wiberg bond indices (WBIs)
can be determined. For typical chemical bonds, WBI value is very close to formal bond order.
Wiberg bond index WAB, between two bonded atoms say A and B corresponds to the sum of
squares of non-diagonal elements of the density matrix of the corresponding atoms.

2
𝑊𝐴𝐵 = ∑ ∑|𝑃𝜇𝜈 |
𝜇∈𝐴 𝜈∈𝐵

But NBO is computationally more expensive and tends to predict larger charged. So NBO is
best used for comparing differences rather than absolute atomic charges.

References:

 (a) J. P. Foster, F. Weinhold, J. Am. Chem. Soc. 1980, 102, 7211-7218; (b) A. E.
Reed, F. Weinhold, J. Chem. Phys. 1983, 78, 4066-4073; (c) A. E. Reed, L. A.
Curtiss, F. Weinhold, Chem. Rev. 1988, 88, 899-926. (b) B. T. Sutcliffe, Int. J.
Quantum Chem. 1996, 58, 645-655
 (a) P. O. Löwdin, Phys. Rev. 1955, 97, 1474; (b) E. R. Davidson, Reduced Density
Matrices in Quantum Chemistry, Academic Press, London, 1976.

39 | P a g e
  K. B. Wiberg, Tetrehedron 1968, 24, 1083-1096
 F. Jensen, Introduction to Computational Chemistry John Wiley & Sons, New York
(1999).
 R. F. W. Bader, Encycl. Comput. Chem. 1, 64 (1998)
 W. D. Cornell, C. Chipot, Encycl. Comput. Chem. 1, 258 (1998)
 D. E. Williams, Rev. Comput. Chem. 2, 219 (1991).
 Frank Weinhold, Natural Bond Orbital Analysis: A Critical Overview of
Relationships to Alternative Bonding Perspectives Journal of Computational
Chemistry 2012, 33, 2363–2379
 Weinhold, Frank; Landis, Clark R. (2001). "Natural Bond Orbitals and Extensions of
Localized Bonding Concepts". Chemistry Education Research and Practice. 2 (2):
91–104. doi:10.1039/B1RP90011K
 Weinhold, Frank; Landis, Clark R. (2012). Discovering Chemistry With Natural
Bond Orbitals. New Jersey: John Wiley & Sons. pp. 132–133. ISBN 978-1-118-
22916-3.
 Eric D. Glendening, Clark R. Landis and Frank Weinhold Natural bond orbital
methods WIREs Comput Mol Sci 2012, 2: 1–42 doi: 10.1002/wcms.51

1.9.2.3 Atoms in Molecules (AIM)

“Science is based on observation. Among the most important of the quantities accessible to
measurement is the distribution of charge - nuclear and electronic - that constitutes matter and
determines its properties. We are indeed fortunate to live in an age where in the accurate
measurement of the charge distribution became a reality, with each year witnessing an
increase in our technical ability to determine its form and interpret its physical
consequences.”- Richard F.W. Bader

The quantum theory of atoms in molecules (QTAIM) is a model of molecular and

condensed matter electronic systems in which the principal objects of molecular structure -
atoms and bonds - are natural expressions of a system's observable electron density
distribution function. Basic motive of QTAIM is to exploit charge density or electron density
of molecules ρ(r; X) as a vehicle to study the nature of bonding in molecular systems. In
other words, the theory says about the topological analysis of the electron density ρ (r; X) and

40 | P a g e
behind this electron density hides the concepts of atoms, bonds, chemical structure and
structural stability.

According to this theory, an atomic boundary should satisfy this equation:

𝛻𝜌. 𝑛 = 0

Where n is the normal vector to the inter atomic surface.

A critical point (CP) in this theory is a point at which 𝛻𝜌 = 0.

Four different types of CP could be found in a molecular space of a molecule. Each CP can
be introduced by a rank and a signature, (r, s). Rank represents the number of non-zero
diagonal eigenvalues of the Hessian matrix of the Laplacian of electron density. Signature
denotes local curvatures of the CP with respect to the x, y and z directions.

Rank (r) Signature (s) Name

3 -3 Nuclear Attractor (NA)

3 -1 Bond Critical Point (BCP)

3 +1 Ring Critical Point (RCP)

3 +3 Charge Critical Point (CCP)

0 0 Critical Point at Infinity

Poincare-Hopf rule: NA –BCP + RCP–CCP = 1

A Nuclear Attractor or (3,-3) CP is a point in which electron density is maximum with

respect to all directions (x, y, z). A BCP or (3,-1) CP is a point in which electron density is
minimum with respect to the maximum density path, connecting two (3,-3) points but
maximum respecting to the rest of directions. A RCP or (3,+1) CP is a point on a surface
which is minimum in electron density on that surface but maximum with respect to the path
perpendicular to the surface. A CCP or (3,+3) CP is a point which is minimum with respect to
all other points around it.

41 | P a g e
Presence of BCP between two atoms was first introduced as the necessary and sufficient
conditions for presence of chemical bond. In QTAIM terminology each bond could be
classified either as Shared or Closed Shell. If in a BCP, 𝛻 2𝜌 < 0, then the corresponding bond
could be classified as Shared. In contrary, if in a BCP 𝛻 2𝜌 > 0, then the corresponding bond
is classified as a Closed Shell interaction. Shared interactions usually are found between
covalently bonded atoms. Closed Shell interactions usually correspond to ionic and van der
Waals bonds.

This theory has been effectively applied to understand the atom-atom interactions in
covalent and non-covalent bonding interactions in molecules, molecular clusters, small
molecular crystals, proteins, DNA base pairing and stacking. In addition to bonding, AIM
allows the calculation of certain physical properties on a per-atom basis, by dividing space
into atomic volumes containing exactly one nucleus, which acts as a local attractor of the
electron density. In QTAIM an atom is defined as a proper open system, i.e. a system that
can share energy and electron density which is localized in the 3D space.

Because QTAIM atoms are always bounded by surfaces having zero flux in the gradient
vector field of the electron density, they have some unique quantum mechanical properties
compared to other subsystem definitions, including unique electronic kinetic energy, the
satisfaction of an electronic virial theorem analogous to the molecular electronic and some
interesting variational properties. QTAIM has gradually become a method for addressing
possible questions regarding chemical systems, in a variety of situations hardly handled
before by any other model or theory in chemistry.

Reference:

 Bader, R. F. W. “Atoms in molecules: a quantum theory” Oxford University

Press:Oxford,1990
 R. F. W. Bader Atoms in Molecules Acc. Chem. Res. 1985,18, 9-15
 Bader, Richard (1994). Atoms in Molecules: A Quantum Theory. USA: Oxford
University Press. ISBN 978-0-19-855865-1.
 Bader, R. (1991). "A quantum theory of molecular structure and its
applications". Chemical Reviews. 91: 893¨C928. doi:10.1021/cr00005a013.

42 | P a g e
  Bader, R.F.W. (2005). "The Quantum Mechanical Basis for Conceptual
Chemistry". Monatshefte fur Chemie. 136: 819¨C854. doi:10.1007/s00706-005-
0307-x.
 Bader, R.F.W. (1998). "Atoms in Molecules". Encyclopedia of Computational
Chemistry. 1: 64–86.
 Hydrogen - Hydrogen Bonding: A Stabilizing Interaction in Molecules and
Crystals Cherif F. Matta, Jesus Hernandez-Trujillo, Ting-Hua Tang, Richard F.
W. Bader Chem. Eur. J. 2003, 9, 1940 ± 1951 doi:10.1002/chem.200204626
 Molecular Recognition in Organic Crystals: Directed Intermolecular Bonds or
Nonlocalized Bonding? Jack D. Dunitz and Angelo Gavezzotti Angew. Chem.
Int. Ed. 2005, 44, 1766–1787 doi:10.1002/anie.200460157
 Polycyclic Benzenoids: Why Kinked is More Stable than Straight Jordi Poater,
Ruud Visser, Miquel Sola, F. Matthias Bickelhaupt J. Org. Chem. 2007, 72,
1134-1142 doi:10.1021/jo061637p

1.9.3 Energy Decomposition analysis (EDA)

The energy decomposition analyses (EDAs) developed by Morokuma, Ziegler and Rauk are
powerful methods, which bridge the gap between elementary quantum mechanics and a
conceptually simple interpretation of the nature of the chemical bond.

EDA gives different energy interaction terms (attractive and repulsive) to explore the nature
of bonding.

∆𝐸 = ∆𝐸dist + ∆𝐸int

The distortion energy ΔEdist is the strain energy associated with deformation of the reactants
from their equilibrium geometry to the geometry they acquire in the respective complex. ΔEint
is the actual interaction energy between the prepared fragments to form the corresponding
complex. Moreover, this term can be further divided into three different physically
meaningful terms that can provide insight into the nature of the bonding interactions. These
terms are (1) the quasiclassical electrostatic interaction energy between the charge densities

43 | P a g e
of the fragments, Eelec, (2) the exchange repulsion between the fragments due to Pauli’s
principle, EPauli, and (3) the energy gain due to orbital mixing of the fragments, Eorb.

∆𝐸int = ∆𝐸elec + ∆𝐸orb + ∆𝐸Pauli

The component ΔEelec is a stabilizing term that corresponds to the classical electrostatic
interaction originates by for the ionic nature of connected fragments. Another stabilizing
term, ΔEorb calculates the energy gained when the fragment orbitals are able to relax into an
optimal form in the complex i.e. the orbital overlap to form covalent interaction. The only
destabilizing component, ΔEPauli arises due to repulsion between two electrons reside in same
spatial orbital.

The ΔEorb is further decomposed into pairwise contributions of interacting orbitals of the
two fragments using EDA-NOCV (EDA with natural orbitals for chemical valence) method,
developed by Mitoraj and Michalak. This method combines the energy terms from EDA with
the natural charge, partitioning schemes to decompose the deformation density (Δρ) into
different components of the chemical bond.

In this thesis we have performed EDA as implemented in the computational chemistry

package ADF (Amsterdam Density Functional).

There are quite a few desirable qualities that an Energy Decomposition Analysis should have:

1. An EDA should divide the total interaction energy as computed by a well-defined

electronic structure method into contributions from physically meaningful terms.
2. It should be descriptive, assigning an energy to as many useful concepts as possible
and avoiding the grouping of several effects into one term.
3. The EDA method should not put undue restrictions on the model chemistry such as
the requirement of a specific type of basis expansion or a density functional with a
specific form.
4. All terms should converge to a physically meaningful basis set limit. Together with
the last requirement i.e. an EDA for the decomposition of Kohn-Sham DFT
interaction energies should work for the exact functional and a complete basis set.
5. All terms with known asymptotic behavior should match this behavior within the
limits of the given model.
44 | P a g e
 6. The EDA should compute energies using only valid antisymmetric electronic
wavefunctions. Classical calculations are only valid in uninteresting limits and should
thus be avoided.
7. The energy terms that sum to the total interaction energy should be continuous if the
total is likewise continuous.
8. An EDA should define its terms variationally, allowing for the application of the
method to both weakly and strongly interacting system.
9. The decomposition of the interaction energy should not be significantly more
computationally expensive than the evaluation of the interaction energy itself

References:

 K. Morokuma, J. Chem. Phys. 1971, 55, 1236-1244.

 T. Ziegler, A. Rauk, Theor. Chim. Acta. 1977, 46, 1-10.
 (a) M. Mitoraj, A. Michalak, J. Mol. Model. 2007, 13, 347-355; (b) M. Mitoraj, A.
Michalak, Organometallics 2007, 26, 6576-6580; (c) A. Michalak, M. Mitoraj, T.
Ziegler, J. Phys. Chem. A 2008, 112, 1933-1939.
 Moritz von Hopffgarten and Gernot Frenking , WIREs Comput Mol Sci 2012, 2: 43–
62
 Morokuma K., Kitaura K. (1981) Energy Decomposition Analysis of Molecular
Interactions. In: Politzer P., Truhlar D.G. (eds) Chemical Applications of Atomic and
Molecular Electrostatic Potentials. Springer, Boston, MA

1.11 Solvent Model

“It cannot be overemphasized that solvation changes the solute electronic structure. Dipole
moments in solution are larger than the corresponding dipole moments in the gas phase.
Indeed, any property that depends on the electronic structure will tend to have a different
expectation value in solution than in the gas phase.”

-C. J. Cramer, “Essentials of Computational Chemistry,” 2002, John Wiley & Sons Ltd.
(ISBN 0-471-48551-9)

45 | P a g e
Most laboratory chemistry is done in solution where there is energy of interaction between
solute and solvent. Because of this, the solute properties reliant on energy, such as geometry,
vibrational frequencies, total energy, and electronic spectrum, depend on the solvent. The
effect on solvent may be short-range which typically concentrated in the first solvation sphere
viz. H-bonds, preferential orientation near an ion or long-range like polarization (charge
schreening).

Explicit Solvent Models

The explicit solvent model includes individual solvent molecules. The free energy of
salvation is calculated by simulating solute-solvent interactions. An explicit solvent model is
useful in Monte Carlo (MC) calculations, molecular dynamics (MD) simulations and involves
very lengthy calculations. Explicit model requires an empirical interaction potential between
the solvent and solute, and between the solvent molecules.

Implicit Solvent Models

In implicit model, solvent is treated as a polarizable continuum with a dielectric constant (),
instead of explicit solvent molecules.

46 | P a g e
 Fig: uniform polarizable medium of fixed dielectric constant e having a solute molecule M
placed in a suitably shaped cavity.

The expression of solvation free energy is:

ΔGsolv = ΔGelec + ΔGdisp + ΔGrep + ΔGcav

Here, the ΔGelec accounts for the polarization between solute and solvent induces charge
redistribution until self-consistent. This interaction is stabilizing in nature (lowers the
energy). The ΔGdisp and ΔGrep represent the dispersive (mainly van der Waals) and repulsive
interaction between solvent and solute respectively and are stabilizing in nature. The last term
ΔGcav is the energy needed to create a cavity in the medium. This process costs energy i.e
destabilization.

Diffrent implicit models differ from each other in five aspects a) size and shape of the solute
cavity b) method of calculating the cavity creation and the dispersion contributions c) how
the charge distribution of solute M is represented d) whether the solute M is described
classically or quantum mechanically and e) how the dielectric medium is described.

One of the first implicit models is Self-Consistent Reaction Field (SCRF) originally designed
by Lars Onsager in 1936 [] and subsequently developed by many others, which considers the
solvent as a uniform polarizable medium with a dielectric constant ε. In SCRF method, the
47 | P a g e
electric charge of the solute polarizes the medium to get electrostatic stabilization. The
solute-solvent interaction is represented by a solvent reaction potential introduced into the
Hamiltonian. Interactions must be computed self consistently and the process is done
repeatedly until the mutual polarization between the solute and solvent achieves self-
consistency.

One of the modern SCRF methods originally developed by Tomasi et al. to deal with implicit
solvation is the Polarizable Continuum Model (PCM). This model is based upon the idea of
generating multiple overlap in spheres for each of the atoms within the molecule inside of a
dielectric continuum where the shape of cavity is determined by shape of solute.

The SMD model, developed by Cramer and Truhlar,[] is an another modern technique to
calculate the free energy of solvation using electron density of the solute without determining
the partial charge. It is called universal method in the sense of its applicability to any charged
or neutral species. This method separates the solvation free energy into two main
components: electrostatic arising from SCRF treatment and other includes dispersion,
repulsion and cavitation terms. This method provides the best choice to calculate ΔGsol values
for a molecule transferring from gas-phase to solvent, since it actually calculate the energy of
cavitation and dispersion-repulsion energy. Mostly this SMD method is used in all our
calculations reported in this thesis.

Onether model called Conductor-like Screening Model (COSMO) in which solvent is

considered as a conductor except close to the cavity. Solute-solvent interactions is based on a
conducting polygonal surface and calculated numerically.

The implicit solvation model of course, is a rough approximation, but normally provides
reasonably appreciated results. The computational cost is very less compared to explicit
solvent models but it cannot model specific interactions such as hydrogen bonds. Hybrid
solvent models treat first solvation sphere explicitly while surrounding solvent by a
continuum model. These models usually treat inner solvation shell quantum mechanically and
outer solvation shell classically

48 | P a g e
 Schematic representation of a) explicit and b) implicit solvation model

Fig: Gaussian Output for PCM

49
References:

Cramer, Christopher J. (2002). Essentials of Computational Chemistry. John Wiley & Sons.
ISBN 0-470-09182-7.

Jensen, Frank (1999). Introduction to Computational Chemistry 2nd edition. John Wiley &
Sons. ISBN 0-470-01187-4.

H. D. Meyer; U. Manthe & L. S. Cederbaum (1990). "The multi-configurational time-

dependent Hartree approach". Chem. Phys. Lett. 165 (73). doi:10.1016/0009-
2614(90)87014-I

Chr. Møller & M. S. Plesset (October 1934). "Note on an Approximation Treatment form
Many-Electron Systems". Physical Review. The American Physical Society. 46 (7): 618–622.

Krishnan Raghavachari & John A. Pople (February 22, 1978). "Approximate fourth-order
perturbation theory of the electron correlation energy". International Journal of Quantum
Chemistry. Wiley InterScience. 14 (1): 91–100

Solvation Model References

K.I. Ramachandran, Deepa Gopakumar, and Krishnan Namboori. Computational Chemistry

and Molecular Modelling: Principle and applications. Springer Berlin Heidelberg 2008.
Pages 250-253

L. Onsager, Electric moments of molecules in liquids, J. Am. Chem. Soc. 1936, 58, 1486.

Mennucci, B., Tomasi, J., Cammi, R., Cheesman, J.R., et al., Polarizable Continuum
Model(PCM) Calculation of solvent effects on optical rotations of chiral molecules. J. Phys.
Chem. 2002. 106(25), 6102

Cramer, C. J.; Truhlar, D. G. (1999). "Implicit Solvation Models: Equilibria, Structure,

Spectra, and Dynamics". Chemical Reviews. 99 (8): 2161–2200.

50
Tomasi, J.; Mennucci, B., Cammi, R.; Cammi, Roberto (2005). "Quantum Mechanical
Continuum Solvation Models". Chem. Rev. 2005, 105, 2999

51