### UNIVERSIDAD ANDRES BELLO

### Facultad de Ciencias Exactas

### Programa de Licenciatura en Química

**“Molecular and Electronic Structure of Sandwich**

**Polyoxometalates type α-Keggin”.**

Unidad de Investigación para optar al grado académico de Licenciado en Química

Departamento de Ciencias Químicas Universidad Andrés Bello

**FRANCISCA PAZ CLAVERIA CADIZ**

Dr. Ramiro Arratia Perez - Dr. Desmond MacLeod-Carey DIRECTORES

### UNIVERSIDAD ANDRES BELLO

### Facultad de Ciencias Exactas

### Programa de Licenciatura en Química

**“Molecular and Electronic Structure of Sandwich**

**Polyoxometalates type α-Keggin”.**

** FRANCISCA PAZ CLAVERIA CADIZ**

**Comisión Evaluadora:**

**Prof. Dr. Jorge Martinez ** ** **
** **

**Prof. Dr. Juan Carlos Santos **

**ACTA DE CALIFICACIONES**

**NOTA DE ESCRITO DE UNIDAD DE INVESTIGACION**

Con fecha 17 de Agosto de 2015, se entregó el escrito de la Unidad de Investigación
realizada por la alumna **FRANCISCA PAZ CLAVERÍA CÁDIZ**, RUT: 15.935.724-4
titulada**“Molecular and Electronic Structure of Sandwich Polyoxometalates type**
**α-Keggin”.**

Las calificaciones obtenidas por la Sra. Francisca Clavería Cádiz en esta etapa son las siguientes:

Comisión Evaluadora:

Prof. Dr. Jorge Martinez ………..

Prof. Dr. Juan Carlos Santos ……….. Prof. Dr. Ramiro Arratia (Profesor Guía) .………. Prof. Dr. Desmond MacLeod-Carey (Profesor Guía) .……….

**Nota Final Escrito: …….**

Doy fe de la veracidad de esta información,

Dra. Verónica Paredes García Directora

**ACTA DE CALIFICACIONES**

** NOTA DE PRESENTACIÓN ORAL Y DEFENSA DE **

**UNIDAD DE INVESTIGACION**

Con fecha 18 de Enero de 2016, se realiza la presentación y defensa de la Unidad de
Investigación del (la) alumno(a) **FRANCISCA PAZ CLAVERÍA CÁDIZ**, RUT:
15.935.724-4 t i t u l a d a**“Molecular and Electronic Structure of Sandwich**
**Polyoxometalates type α-Keggin”.**

Las calificaciones obtenidas por la Sra. Francisca Clavería Cádiz en esta etapa son las siguientes:

**ComisiónEvaluadora: ** **Exposición Oral** **Defensa**

Prof. Dr. Jorge Martinez ……….. ………..

Prof. Dr. Juan Carlos Santos ……….. ………..

Prof. Dr. Ramiro Arratia Perez (Profesor Guía)

……….. ………..

Prof. Dr. Desmond MacLeod-Carey (Profesor Guía)

……….. ………..

**Nota Final Presentación y Defensa: **………..

Doy fe de la veracidad de esta información,

Dra. Verónica Paredes García Directora

**Abstract **

In the present thesis we carried out extensive relativistic scalar DFT calculations on six [XM12O40]q- polyoximetalates (POMs), α-Keggin heteropolyanions (where X =Sn, Sb, Te, and M = Mo, W) and the recently self-assembled synthesized [(TeW9O33)2Pd3]10- cluster.

The calculations included the structural characterization, electronic properties, the vibrational and optical properties of these POMs to increase our understanding of these exotic structures that can lead to the development of nanoscale materials.

We have focused our attention into the six [XM12O40]q- species and the electronic structure of the [(TeW9O33)2Pd3]10-. This latter cluster, synthesized at Cronin lab facilities has two {TeW9} units stabilized by a [Pd3]6+ triangular fragment.

The molecular structures described here have been characterized as the energy minima through calculations of normal mode vibrational frequencies, which were performed at the optimized geometries via second derivatives of the total energy with respect to the internal coordinates as implemented in the ADF code (Amsterdam Density Functional). The vibrational frequencies were calculated for the complete set of functionals, finding values in rationale agreement with already reported in literature data. We focused in the main intense signals in order to allow the comparison with experimentally characterizations.

**Resumen **

En la presente tesis se llevó a cabo extensos cálculos DFT escalares relativistas en seis polioximetalatos [XM12O40]q- (POMs), heteropolianiones α-Keggin (donde X =Sn, Sb, Te y M = Mo, W) y el reciente auto-ensamblado cluster sintetizado [(TeW9O33)2Pd3]10-.

Los cálculos incluyen la caracterización estructural, propiedades electrónicas, las propiedades vibracionales y ópticas de estos POMs para aumentar nuestra comprensión de estas estructuras exóticas que pueden conducir al desarrollo de materiales nanométricos.

Hemos centrado nuestra atención en las seis especies de [XM12O40]q- y la estructura electrónica de [(TeW9O33)2Pd3]10-. Este último cluster, sintetizado en las instalaciones del laboratorio de Cronin, tiene dos unidades de {TeW9} estabilizadas por un fragmento triangular de [Pd3]6+.

Las estructuras moleculares aquí descritas, han sido caracterizadas como los mínimos de la energía a través de cálculos de modos normales de frecuencias vibracionales, que se realizaron en las geometrías optimizadas a través de segundas derivadas de la energía total con respecto a las coordenadas internas como puesto en ejecución en el código ADF (Amsterdam densidad funcional). Las frecuencias vibracionales se calcularon para el conjunto completo de funcionales, encontrando valores de acuerdo a los datos reportados en la literatura. Nos hemos centrado en las principales señales de intensidad con el fin de permitir la comparación con caracterizaciones experimentales.

**Acknowledgements **

First, I would like to thanks my parents Mario Clavería and María Soledad Cádiz, for supporting me in my life with love possible. My sisters; Loreto, Paulina and Marisol for their unconditional support and special thanks to my husband Raúl Vásquez and my son Benjamín Vásquez Clavería to be the loves of my life.

My thanks to my aunt Mariana Cádiz for always believing in me as a mother.

First of all, I would like to express my sincere thanks to my advisors Dr. Ramiro Arratia Pérez and Dr. Desmond MacLeod-Carey for their continued support of my studies and undergraduate thesis for their patience, motivation, enthusiasm and immense knowledge. Their guidance helped me throughout the time of research and writing this thesis.

I would like to especially thank my friends Matías Reyes, Raúl Guajardo and Wilson Caimanque and for their support and friendship in this proccess.

I also thank the University Andres Bello and Department of Chemistry with their teachers.

**Contents **

2.2 General and Specific Goals 10

2.2.1 General Goals 10

2.2.2 Specific Goals 10

**Methodology **

**11**

3.1 Theoretical Background 12

3.1.1 Introduction 12

3.1.2 Density Functional Theory 13

3.2 !!!!-Representability and N-Representability 21 3.2.1 Fundamental Equation of Density Functional Theory 21

3.3 Relativistic effects 26

3.4 Time Dependent Density Functional Theory 28

3.5 Energy Decomposition Analysis 29

3.6 Computational Details 30

**Results and Discussion **

**32**

4.1 Density Functional Study of the Vibrational Frequencies, Molecular and Electronic Structure of a-Keggin heteropolyanions containing fifth period main group atoms.

33

4.1.1 Introduction 33

4.1.2 Molecular Structure 34

4.1.3 Vibrational Analysis 37

4.1.4 Electronic Structure 38

4.1.5 Energy Decomposition Analysis (EDA) 39

4.1.6 Conclusions 42

4.2 Electronic Structure of [(TeW9O33)2Pd3]10-.

A self assembled Nano-Object from a Polyoxometalate building

block. 43

4.2.1 Introduction 43

4.2.2 Computational Details 44

4.2.3 Results and Discussion 46

**Summary **

**53**

5.1 Overall Conclusions 54

**List of Figures **

### Figure 1 Polyoxometalates and their structural and size

### diversity.

### 3

### Figure 2 Octahedral MO

6### structral unit.

### 3

### Figure 3 Isomeric forms of Keggin type POMs.

### 4

### Figure 4 Keggin Lacunary species.

### 5

### Figure 5 Examples of sandwich Keggin type POMs.

### 8

### Figure 6 Molecular Structure.

### 34

### Figure 7 POM Cluster Position.

### 45

### Figure 8 Description Molecular Structure.

### 47

**List of Tables **

### Table 1 Calculated bond distance for

_{α}

### -Keggin

### heteropolyanions.

### 36

### 3

### 6

### Table 2 Calculated Vibrational Frequencies and their

### intensities for

_{α}

### -Keggin heteropyanions.

### 38

### Table 3 Energy Decomposition analysis of the fragments

### interaction for

### α

### -Keggin heteropyanions.

### 41

### Table 4 Calculated Select Bond Distance to Figure 6.

### 48

### Table 5 Calculated Select Bond Distances to Figure 7.

### 49

### Table 6 Calculated Vibrational Analysis

### 50

**List of Abbreviations **

**POM Polyoxometalate **

**Td Tetrahedric Geometry **

**DFT Density Functional Theory **

**VWN Vosko, Wilk and Nusair **

**STO Slater type orbital **

**MZ Morokuma–Ziegler **

**H-L HOMO-LUMO **

**GTO Gaussian type orbitals **

**MO Molecular Orbitals **

**EDA Energy Decomposition Analysis **

**GGA Generalized Gradient Approximation **

**HEG Homogeneous Electron Gas **

**LDA Local Density Approximation **

**PBE Perdew–Burke–Ernzerhof **

**RDFT Relativistic density functional theory **

**TD-DFT Time Dependent Density Functional Theory **

**Introduction **

**1.1 Generalities **

Polyoxometalates (POMs) are anionic metal oxide clusters of nanometrical size. Specifically, they can be described as aggregates of early-transition metal cations in high oxidation states (usually, their d0 species, V(V), Nb(V), Ta(V), Mo(VI) y W(VI)) bridged by oxo (O2-) ligands, which are formed through a self-assembly proccess. [1-6]

They can be represented by the general formula [MmOy]n-_{ in the case of the }
so-called Isopolyanions and [XxMmOy]n-_{ for the Heteropolyanions, where M }
represent the metals in the oxidation states early mentioned, X corresponds to a
heteroatom, which can be a main group element or a transition metal.
Subindexes x, m and y correspond to the number of each of them that can be
found in every different POM, and finally n- represents the charge of the anionic
cluster, which depends of the nature and number of M and X present in the
structure.

**1.2 Structural diversity **

!

**Figure 1 Polyoxometalates and their structural and size diversity **

!

For all of these five type of basic structures, the basic unit of construction is a MO6 octahedron, see Figure 2. According to the Lipscomb principle, none of the MO6 octahedrons can present more than two non-shared oxygen atoms (terminal oxo ligands).[9]

According to this, a MO6 octahedron are classified based on terminal oxo
ligands, as **Type I octahedron (mono-oxo) with one terminal oxygen atom per **
octahedron, and **Type II octahedron (cis-dioxo), where the number of terminal **
oxygen atoms are two in cis position.

!

**Figure 2 Octahedral MO6 structral unit. **

!

aditional electron falls in a non-bonding orbital. If the POM is built only by Type
**II octahedrons, the additional electron will be located in a bonding orbital, which **
leads to structural destabilization.[10]

!

**Figure 3 Isomeric forms of Keggin type POMs. **

! !

**1.3 Keggin type POMs **

{M3O13} units are rotated by π/3, the different isomers of the Keggin type POMs can be obtained, see Figure 3.[12]

**1.4 Lacunary species **

In addition to the isomeric forms of the POMs, certain Lacunary species can be synthesized via stoichiometric and pH control, through the remotion of one, two or three oxometallic vertex. In the case of the Keggin derivatives, the formation of the lacunary species, see Figure 4, can be accomplished by using the required quantities of bicarbonate or a sodium alcoxide as base.[13,14]

!

**Figure 4 Keggin Lacunary species **

!

The missing units at the lacunary sites, can be replaced by: i) other MO6 units;

ii) diverse organometallic fragments such as: M-η5-C5H5, M-η4-C8H8;

This chemical modification greatly expands the structural diversity of this type of POMs structures.[15]

It must be mentioned that the Keggin lacunary species are not hydrolitically neither thermodinamically stable, and can be spontaneusly reverted to their Keggin derivative.

**1.5 Dawson type POMs **

Within this context, the lacunary derivatives can react between them, giving rise to larger structures. In this way, though the symmetric union of two tri-lacunary [XM9O34]n- Keggin type species, the Dawson type POM derivative [X2M18O62]n- is obtained.[16] The Dawson type derivatives exhibit spectroscopic, redox and reactivity properties similar to the Keggin type derivatives.[17]

Hence, it is possible to obtain Dawson mono, di or tri-lacunary species, similar to the Keggin ones, which can be employed to prepare new derivatives with different morphologies and molecular properties. In this way, through the exhaustive stoichiometric and pH control of the complete and lacunary species of different POMs architectures, the species presented in Figure 1 can be selectively obtained, reaching an enormous size with a high degree of control.[15]

**1.6 POMs as building blocks **

On the other hand, technological advances continue to expand into unknown dimensions. Despite the progress in recent decades, it is uncertain whether the devices may work based on their bulk properties by maintaining their characteristic properties that allows operation, when its components reach nanometric dimensions. The best known case of this has to do with the development of Intel processors range i3, i5 and i7, where they had to go from a 2D design in older processors to 3D nanoscale (<15nm) in the last generation.[25] Given this situation, there is growing interest in developing molecular-based electronic materials. For example, a large number of active centers with redox molecules have been implemented in molecular/semiconductor hybrid architectures through covalent bonds for storage devices.[25]

!

**Hypothesis and **

**Goals **

**2.1 Hypothesis **

In this thesis we plan to explore the structural, electronic, vibrational and optical properties of several POMs to understand their main characteristics that may show the potential for interesting applications in materials science.

**2.2 General and specific goals **

**2.2.1 General goals **

To contribute to the study of complex species based on sandwich tri - lacunar Keggin type [XW9O33]n- and deepen the understanding of their structural, molecular, vibrational and optical properties through DFT calculations including relativistic corrections when necessary.

**2.2.2 Specific Goals **

1. To perform geometry optimization calculations with the most appropriate
methodology for sandwich complexes based on tri - lacunar Keggin type
[XW9O33]n- _{ species. }

2. To study their electronic structure calculations with the most appropriate methodology based sandwich Species tri - lacunar Keggin type [XW9O33]

**Methodology **

**3.1 Theoretical background **

**3.1.1 Introduction **

The molecular Hamiltonian is the operator representing the energy of the system consisting of electrons and nuclei of a molecule. This is a major operator in quantum mechanics, and can be found from the equation of eigenvalues and eigenfunctions, the energies and wave functions of different molecular electronic states. The molecular Hamiltonian contains terms that describe the kinetic energy of the electrons and the Coulomb interactions between electrons and between electrons and nuclei.[26]

The Born-Oppenheimer approximation disengages the movement of atomic nuclei and electrons. The Born–Oppenheimer approximation [27] takes note of the great difference in masses of electrons and nuclei. Because of this difference, the electrons can respond almost instantaneously to the displacement of the nuclei. Instead of trying to solve the Schrödinger equation for all the particles simultaneously, we regard the nuclei as fixed in position and solve the Schrödinger equation for the electrons in the static electric potential arising from the nuclei. Therefore it is possible to parameterize nuclear terms for a fixed configuration of the nuclei.[28]

atom, representing its kinetic energy as a function of the electron density and combining the classic relations electron-nucleus interactions, electron-electron, nucleus-nucleus. The original Hohenberg-Khon theorems were applied only for non-degenerate ground states in the absence of a magnetic field, although they have been generalized to degenerate ground states.[31,32] In DFT [33] the external potential produced for charges of nuclei acting on electrons of system, and number of electrons N is determined by the electron probability density !(r), hence the ground state wave function and energy.

**3.1.2 Density Functional Theory **

**3.1.2.1 Basics, Number of Electrons and External Potential **

The main objective is to solve the Schrödinger equation independent of time under a non-relativistic approach, which yields the energy of the system and the wavefunction !

Η!!=!Ε!! (3.1)

from which we can write the Schrödinger equation (non-relativistic) independent of time under the Born-Oppenheimer approximation

Η_{!"!#$}!_{!"!#$} !_{!};!_{!} = Ε_{!"!#$}!_{!"!#$}(!_{!};!_{!}) (3.4)

It is noteworthy that the variational theorem to calculate the energy of the ground state of a normalized wavefunction is

Ε ! = ! Η ! ≥!!! (3.7)

for which

!! =!"#!⟶!""#$%& ! Η ! (3.8)

From (3.3) we can define the external potential as

!_{!"#} = !_{!}
number of electrons and the external potential,

Ε=Ε[!,!_{!"#}] (3.10)

**3.1.2.2 Hartree-Fock methodology **

Η! !!,!!…!! = Ε! !!,!!…!! (3.11)

Due to the electron-electron repulsion term of the electronic molecular
Hamiltonian involves the coordinates of two different electrons, it is necessary to
reformulate it in an approximate way. Employing the Hartree Fock method, we
choose a subset of all possible wavefunctions, that is a product of N
monoelectronic antisymmetric orbitals !_{!}

!!" _{=} _{!"#} _{!}

!(!!)!!(!!)…!!(!!) =!!" (3.12)

where SD is Slater determinant and the Hartree Fock wave function !!"_{ is an }

eigenfunction of a Hamiltonian of non-interacting particles under a
monoelectronic effective potential !_{!}, which simulates the behavior of the
two-nuclearnuclear repulsion term are re-expressed as the sum of one-electron
operators, where

each of the two electrons in the !!!_{ orbital, and }

!!" = !!∗ !! !!∗ !! _{!} !

!!!!!! !! !! !! !!!!!! = !" !"

(3.18)

is the Exchange operator, defining the electron exchange energy due to the antisymmetry of the total N-electron wave function.

It must be pointed out that j=i can be interpreted as the self-interaction, with the
property!!!!!!_{!!} =!_{!!}.

By solving the variational theorem to calculate, the energy as

Ε! =!"#Ε {!!} = !"# !!" Η !!" (3.19)

and imposing !_{!}∗ _{!} _{!}

!!"= !!" while variying !!, we obtain a set of Hartree–

Fock one-electron wavefunctions, which is equivalent to solving the eigenfunction equation

orbitals. These atomic orbitals are called Slater-type orbitals. The one-electron wavefunctions can be expressed as a linear combination of atomic orbitals.[34]

!= !_{!}!_{!}!"
!

(3.21)

**3.1.2.3 Electron Density as Carrier of Information **

At the 1920 decade, Thomas and Fermi [29,30] developed a theory in where the electron density !(!) of the system was used as the basic carrier of information instead of the wave function.

! !_{!} =!! … !∗ _{!}

!,…,!! !! !!,…,!! !!!!!!…!!! (3.22)

Where, !(!_{!}) represents the probability of finding an electron (N) in a volume
!!!. Also holds that,

! !_{!} ≥ 0 (3.23)

! ! !!!= ! (3.24)

At this point, it can be drawn out that the purpose of the Density Functional Theory is likely to work by the electron density, meaning thereby not attempting to calculate the wave function itself. This theory is based on the theorems developed in 1964 by Pierre Hohenberg and Walter Kohn, which are presented below.

**3.1.2.4 First Theorem of Hohenberg-Kohn **

In 1964 Hohenberg and Kohn formally proved that the electron density can
indeed be used as the basic variable determining all atomic and molecular
properties.[31] They propose that the external potential !_{!"#}(!) is (to within a
constant) a unique functional of !(!); Since, in turn !!"#(!) fixes Η we see that
the full many particle ground state is a unique functional of ! ! .

Suppose two external potentials !_{!"#} and !_{!"#} differentiated in more than a
constant, but leading to the same electron density associated with the
nondegenerate ground state of N particles.

These two external potentials define two Hamiltonians associated with two wave

functions and two different ground state energies Ε_{!!}≠!Ε_{!}.!

!!"# ⟹!Η=!!+!!!!+!!"# ⟹!⟹Ε! (3.25)

!_{!"#} ⟹Η=!!+!!_{!!}+!_{!"#} ⟹!⟹Ε_{!} (3.26)

solve the Hamiltonian Η through the variational theorem

Ε_{!} < ! Η ! =! ! Η ! + ! Η−Η ! (3.27)

Thus,

Ε_{!} < Ε_{!}+ !_{!}(!) !_{!"#}−!_{!"#} !" (3.28)

Then, taking ! as a function test, we solve the Hamiltonian Η through the variational theorem

Ε_{!} < Ε_{!}+ !_{!}(!) !_{!"#}−!_{!"#} !" (3.29)

If we add both equations, we obtain the following contradiction

Ε_{!}+Ε_{!} <!Ε_{!}+Ε_{!} (3.30)
From the previous result, we can conclude that, there can’t be two external
potentials that provide the same electronic density, which means that the
electronic density of the fundamental state determines uniquely the external
potential

!_{!}_{!} ⟹{!,!_{!"#}}⟹ Η (3.31)
Since the Hamiltonian is a density functional, the energy of the ground state is a
functional

Ε_{!} = Ε[!_{!}] (3.32)

and their components as well

Ε_{!} = T !_{!} +Ε_{!!} !_{!} +V_{!"#}[!_{!}] (3.33)
specially the external potential, which corresponds to the electron-nucleus
attraction,

Ε! =T !! +Ε!! !! +E!"[!!] (3.34)
where Ε_{!} contains dependent and independent terms of the system

Ε_{!} = T !_{!} +Ε_{!!} !_{!} +E_{!"} !_{!} = T !_{!} +Ε_{!!} !_{!} + !_{!}(!)!_{!"}!" (3.35)

!!" _{!}

! =T !! +Ε!! !! (3.36)

It must be noted that, besides of the external potential, the density of the ground state must determine the number of electrons N. This is because we assume that is N representable.

The density determines the Hamiltonian and therefore all states of the system, the fundamental and excited ones. In principle, any observable, including energy, can be accurately calculated from the density of the ground state, that is, any observable is a density functional of the ground state. The one to one relationship between the density and the external potential is only possible if using the density of the ground state. We have analized the case of a non-degenerate ground state, but the theorem can be extended to non-degenerate states. The Hohenberg-Kohn functional, being independent of the system concerned is therefore universal. If it were known we could use it to solve any molecule regardless of size, from hydrogen atoms to proteins.

The functional Ε_{!!} !_{!} , which describes the electron-electron interaction,
contains the classic Coulomb electrostatic interaction whose expression is well
known and a residual non-classical part

!!!!!!!!!!!!!!Ε!! !! !=!_{!} !!!!_{!}! !!

!" !!!!!!+Ε!"!!!"#$%!#" !! !!!!!!!!!!!!

=! !_{!} +Ε_{!"!!}_{!"#$%!#"} !_{!}

(3.37)

**3.1.2.5 Second Hohenberg-Kohn Theorem **

Now, that the density has been recognized as a fundamental carrier of information, the next problem is how to get it, to solve this, Hohenberg and Kohn [31,32] introduce the variational principle into DFT through his second theorem, i.e. the groundstate energy can be obtained variationally: the density that minimises the total energy is the exact groundstate density.

The proof of the second theorem is straightforward: as just shown !(!) defined in which the external potential !(!) is unrelated to another density !(!),

Ε_{!} ! ! = ! ! !_{!"#} ! !"+! ! ! ! (3.41)

and the variational principle asserts,

! ! ! + !!_{!"#} ! > ! ! ! + !!_{!"#} ! (3.42)

where ! is the wavefunction associated with the correct groundstate ! ! .! This leads to,

! ! !!"# ! !"+![! ! ]> ! ! !!"# ! !"+![! ! ] (3.43)

and, in this way, the variational principle of the second Hohenberg-Kohn theorem is obtained,

**3.2. **

### !

_{!"#}

**-Representability and N-Representability **

The two theorems of Hohenberg-Kohn require densities that are associated with one or more wave functions as an external potential !!"#. The problem of !!"#

representability and N-representability it is to recognize those densities that meet both requirements.

A density ! is N-representable if there is a wave function ! such that:

! ! = ! … !∗ !_{!},…,!_{!} !! !_{!},…,!_{!} !!_{!}!!_{!}…!!_{!} (3.45)

the N-representable density therefore meets that:

! ! ≥ 0 (3.46)

! ! !" =! (3.47)

A density ! is !!"#- representable, if there is an external potential created by this
density. At the present date, we do not really know what conditions a density
must meet to be !_{!"#}-representable. However, in practice, it is only necessary
that a givenarbitrary density can be represented in the form presented before.
Thus, we have shown that the external potential and therefore, the Hamiltonian
of a system of N electrons is uniquely determined by the density. Consequently,
energy (and any expected value of any other operator) is a density functional
(first theorem). Moreover, the ground state energy is obtained when we
introduce the functional corresponding to the fundamental state density (second
theorem). Specifically, the second theorem is valid only if the exact functional is
unknown.

An approximate functional is not associated with the actual Hamiltonian system so the variational principle becomes invalid. Therefore, it is possible to obtain DFT energies below the exact value.

**3.2.1 Fundamental Equation of Density Functional Theory **

density, thus we can find the density that minimizes the energy (second Hohenberg-Kohn theorem). If now we minimize the electronic energy of equation (3.34) with respect to the electron density ! ! !with the additional contraint that during minimization, the electron density should at all times integrate to the total number of electrons N of the system. Attaching a Lagrange multiplier ! to the normalization constraint yields

!

!"(!)!!! !!( !! !"!!) !0 (3.48)

or

!!

!"!!!= 0 (3.49)

by using equations (3.34) the equation (3.48) can be rewritten as:

! ! +!(!! ! !!!![!! ]

!" ! !!= 0 (3.50)

which finally becomes

! ! +!!!"!

!"! =! (3.51)

This expression corresponds to the fundamental equation of Density Functional
Theory, and is often called the DFT analogue of the Schrödinger equation. If !_{!"}
were known exactly, this would give the exact solution for the ground state
density and associated energy. Unfortunately, large parts of the !_{!"} are
unknown and have to be guessed. As was pointed out in (3.37) the
electron-electron interaction, contains the classic Coulomb electrostatic interaction and a
residual unknown non-classical part. On the other hand, it is not possible to
carry out the same separation for the kinetic energy functional. This is bypassed
by using orbitals again through the Kohn-Sham method.[35,36]

**3.2.1.1 The Kohn-Sham Method **

As we saw, the energy of the ground state of an N-electron system can be expressed:

Ε!!min! !!" !! + !! ! !!" ! !" (3.52)

!!" _{!}

! =! ! +Ε!"!#$%&'# !! +!_{!}! !(!!_{!})!(!!)
!" !!!!!!

(3.53)

However, the Hohenberg - Kohn theorems tell us nothing about the explicit form of the functional or how to calculate the density without making first use of the wave function.

A model that has an explicit expression for the functional part of Hohenberg-Kohn is the Thomas-Fermi model, which provides an explicit expression for the kinetic energy functional: electron density as the exact, fully interacting system and additionaly this density mustintegrate to the total number of electrons in a single determinant as

!! ! =! !!(!)!!∗(!)
addition, we can use the variational principle (second Hohenberg-Kohn theorem)
considering that variations in density are now reduced to orbital variations. Here
!_{!} are the so-called Kohn-Sham orbitals.

The electron density is obtained exactly as

! ! = !_{!}(!) !
!

!!!

(3.57)

The final Khon-Sham energy density functional can thus be expressed as

Ε!" ρ =Ε!"!#$%&'#%$ ! +(! ! −!! ! ) (3.59) Whereby the Khon-Sham energy expression becomes

Ε_{!"} ρ =!_{!} ! +Ε_{!"} ! +! ! +Ε_{!"!#$%&'#%$} ! +Ε_{!"}[!] (3.60)
This energy expression would lead to an exact method if the exact expression of

Ε!"[!(!)] were to be known.

**3.2.1.2 Exchange-Correlation Functional Approximations **

It is now evident that the quality of the results that can be obtained using the KS
formalism depending on the approach we use forΕ_{!"}[!(!)]. There is no strategy
that can continue to obtain precisely the exchange correlation functional. In the
last 30 years several accurate approximations derived from first principle
considerations, using parametrizations, or combinations of both have become
available in several computational codes. At the present date, several atomic
and molecular properties can be computed very accurately. However, a number
of problems persist, including the accurate description of reaction barriers,
multiplet energies of transition metals, the accurate description of excited states,
dispersion and non-covalent interactions amongst others. In short, what
ultimately determines the usefulness of a functional is totally empirical, i.e.
compare the calculated results with a set of reference data. In some simple
cases we can compare the calculated results directly with potential precise
calculations derived from accurate wavefunction methodologies.[26]

**3.2.1.2.1 Local Density Approximation (LDA) **

The local density approximation is the oldest and most popular of the exchangecorrelation functionals, LDA [33] assumes a simple form which is a linear functional of the density

Ε_{!"}!"# _{!} _{=} _{!} _{!} _{!}

!" ! !" (3.61)

does not occur in atomic or molecular systems. Where "XC is the Exchange-correlation energy per particle in a homogeneous electron gas (HEG) of density

! ! .!The energy can now be divided into a contribution of exchange and other of correlation:

On the other hand, there is no an explicit form for the correlation functional, but
some analytical expressions can be interpolated from numerical simulations.
The most famous !_{!} ! ! are those proposed by Vosko, Wilk and Nusair known
as VWN.[37]

**3.2.1.2.2 Generalized Gradient Approximation (GGA) **

Since the electronic density in atoms and molecules is not homogeneous, we must add to LDA the information related to the gradient of the electronic density

∇! ! . The introduction the GGA [33] methods improved the results over LDA. The exchangecorrelation functional is then expressed as

Ε_{!"}!!" _{!}

At first place, the exchange functional can be expressed into the local and non-localcontributions. The later term can be expanded as follows for the Becke’s Exchange functional proposed in 1988.[38]

Ε_{!}!!" _{=}_{Ε}

where ! ! is a function of the reduced density gradient

! = ∇! !!!

(3.67)

As is defined, s considers the possible effects of a non homogeneous density.
Continuing with the example of the Becke’s functional, the !!!!_{(!}_{)}_{ function is }

On the other hand, the correlation functionals present analytic expresions even
more complex, as an example, the Perdew’s function [39] !!!"_{(}_{!}_{)}_{ proposed in }

The names of the exchange-correlation functionals are formed by linking functional acronyms first exchange and correlation later. In combination with the Becke Exchange functional proposed in 1988, for example, we can form: BP86, BLYP, BPW91, amongst others; which include the Perdew (BP86 proposed in 1986), Lee-Yang-Parr (LYP proposed in 1988) and Perdew-Wang (PW91 proposed in 1991) correlation functional respectively.

Considered overall, GGA considerably improves the results obtained by LDA and is adequate for most of the chemical needs. It was this improvement obtained by functional GGAs, what finally made DFT a useful methodology for computational chemistry.

**3.3 Relativistic effects **

its decoupled, where sv represents the Pauli matrix. The Dirac equation satisfies all requirements of special relativity and quantum mechanics. The ZORA Hamiltonian is the most commonly used relativistic corrections, this exhibit accurate results with less computational cost in comparision whith most accurate method as the Dirac equation four component equations given by matrix and p, the momentum operator. The elimination of the small component leads to,

remains < 1, and their regular expansion expansion of zero order leads to

electron with the magnetic field generated by the orbital angular momentum (L) around the core.

**3.4 Time Dependent Density Functional Theory **

In order to estimate the excitation energies with DFT, we use the Time Dependent Density Functional Theory (TD-DFT) based on time-dependent equations Khon-Sham [35,36] Schrödinger equation time-dependent:

!_{!"}!! ! = Η(!)!(!) (3.76)

In which the time-dependent Hamiltonian includes a time-dependent external potential:

Η=!+!+!(!) (3.77)

The Runge -Gross theorem [41] shows that the densities ! !,! !and !(!,!) of two systems which evolve from the same initial state to !=0!under the influence of the potential ! !,! and !(!,!), respectively , will always differ. These densities retain one to one correspondence between ! !,! and ! !,! and between !(!,!) and !(!,!). In the framework of Kohn-Sham Hamiltonian not time dependent interacting is given by:

Η_{!} ! =!+!+!_{!}(!) (3.78)

Determines the wavelength dependent function !!time, and therefore leads to obtain a set of time-dependent orbital N obey the time-dependent equations KS.

(−!_{!}∇!_{+}_{!"}_{(!,}_{!}_{))!}

!(!,!) =!_{!"}!!!(!,!) (3.79)

!_{!} !,0 =!_{!}(!) (3.80)

a Taylor series, in which the excitation energy and polarizability load depend
only first order in density which in turn depends on the linear response function
! that in a non-interactive system is !_{!} TD- KS. A detailed analysis of this
derivation and implementation can be found in reference. The success of the
TD-DFT analysis method by calculating the excitation energies of the transition
metal complex has been shown in recent studies.

**3.5 Energy Decomposition Analysis **

The Energy Decomposition (EDA) proposed by Morokuma [43] provides insights
intointermolecular interactions by separating the total interaction energy into
various terms such as electrostatic, exchange repulsion, and orbital overlapping.
In this analysis, we consider any host-guest system subject of this research as a
the entity AB with corresponding wavefunction !_{!"}, and total energy Ε_{!"}. AB is
the result of interaction between fragments Α!_{ and }_{Β}!_{ in their electronic and }
fragments is called the preparation energy ΔΕ_{!"#!}_{.}

ΔΕ_{!"#!} =Ε_{!}−Ε_{!}!_{+}_{Ε}

!−Ε!! (3.81)

The difference between the energy of AB Ε_{!"}, and the energies of the prepared
fragments Ε! and Ε! is termed Δ!"#$.

ΔΕ_{!"#} =Ε_{!"}−Ε_{!}−Ε_{!} (3.82)
Now, we will focus on the bond formation according to the EDA. In the first step,
the distorted fragments with frozen charge densities A and B are brought from
infinite separation to their position in the final system AB. This state is refered as

calculated subtracting the Coulomb integral of the fragmenents from that corresponding to the overall system.

ΔΕ_{!"!#$} =!_{!} !! !!

with energy Ε_{!"}. The associated energy lowering comes from the orbital mixing,
and thus, it can be identified as the covalent contribution. It is termed orbital
methodologies depending on the exchange-correlation functional employed.

**3.6 Computational Details **

triple-z Slater type orbital (STO) basis set.[48] The functional and basis set choices were based on the results of previous works performed on Lindqvist [49,50,51,52] and Keggin [53, 54,55,56] derivatives.

Solvation water effects were modeled by a conductor-like screening model for real solvents (COSMO).[57,58,73]

**Results and **

**discussion **

**4.1 Density Functional Study of the Vibrational **

**Frequencies, Molecular and Electronic Structure **

**of **

### α

**-Keggin heteropolyanions containing fifth **

**period main group atoms. **

**4.1.1 Introduction **

From the first report of [PMo12O40]3- by Berzelius in 1826 [44] Keggin POMs have played a central role in several fields as medicine, catalysis, analytical chemistry and technology.[45] The structure of α-Keggin heteropolyanions

consist of the assembly of twelve {MO6} octahedra that form a {M12O36} cage that encapsulates an internal {XO4} unit. The {MO6} units are edge shared into four {M3O13} groups, leading to an ideal Td point group symmetry, were all metal atoms are equivalent, resulting in the α-isomer Keggin structure. The 60º

rotation of one of the {M3O13} groups about a 3-fold axis of the α-isomer

produces the β-isomer. However, isomerizations on Keggin structures occur in

the β!α direction. Subsequent 60º rotations of the remaining triads give rise to

the three additional isomers γ, δ and ε, which are less stable than α and β

isomers. A large number of Keggin derivatives including a wide range of internal heteroatoms “X“ (e.g., X= P(V), Si(IV), As(V), Al(III), Ge(IV), Fe(II), Fe(III), Co(II), Co(III), etc) with addenda metal atoms “M” (e.g., M= Mo(VI), W(VI), etc) are well reported in literature. However, to our present knowledge, Keggin type POMs containing elements of groups 14–16 of the fifth period as internal heteroatoms are scarce, and available information in literature is vague and inaccurate, as is discussed below.

In this section, a serie of density functional theory (DFT) calculations on fully
oxidized α-Keggin heteropolyanions [SnMo12O40]4- **(SnMo), ** [SbMo12O40]3-

!

**Figure 6 Molecular Structure **

**Description Figure 6: Molecular Structure and atom-labeling scheme for **!**-Keggin **

**heteropolyanions showing: i) Central yellow ball corresponds to X = Sn(IV), Sb(V) or **
**Te(IV) ii) Blue and light blue balls corresponds to M= Mo(VI) or W(VI) iii) Red and light red **
**balls corresponds to oxygen atoms. With labelling Ot= Oxygen terminal atoms; Ob= **
**Oxygen bridging metal atoms at the {M3O13} unit; Ob’ = Oxygen bridging metal atoms at **

**different {M3O13} units; Oi= Oxygen of the internal {XO4} unit. iv) {M3O13} unit is highlighted **

**with strong colors. **

**4.1.2 Molecular Structure **

The α-Keggin molecular structure and atom-labeling scheme employed along

or tellurium dodecamolybdate or dodecatungstate complexes which lack of
reliable information about the structure, physical or spectroscopic properties of
these compounds. A theoretical study carried out by Y. L. Si *et*. al. [62],
proposed that (C5H13N2O2)2(H3O)SbMo12O40 and (C5H13N2O2)2SnMo12O40 may
exhibit inhibitory actions to the human cancer cells just based on the
HOMO-LUMO gap calculated at BP86/LDA-ZORA level of theory. It must be mention
that this article does not report the molecular structure or any other relevant
information about this POMs. On the other hand, some catalytic properties for
the hydrotreatment of oil fractions are attributed to H8-X[X+X(Mo12O40)]·nH2O
(X=Sn,Sb) amongst a large serie of heteropolyanions, but the preparation
methods, supplier as well as the structural characterization of these two POMs
are not present in the text.[63] Into the same context, K6[α-SnW12O40] has been

reported in a scientific meeting [64,65] but no relevant data is informed, as expected from a book of proceedings, specially if the data is not yet published. Dodecamolybdoantimonate (V) salts have been reported several years ago,[66,67] with enough information to elucidate that its potassium, ammonium, caesium and rubidium salts possess Keggin structure, however, no structural data is available as we know at the present time.[67]!

A light of hope about the preparation of Na4[TeW12O40] is found at a report of the Vietnam Energy Comission [68], but once again, the synthetic procedure is only supported by a UV-Vis spectra of an ion-complex formed by Nile Blue, the tungstotelluric heteropolyanion and PVA. It must be mentioned that no further separation, purification or characterization of the complex was carried out.

Thus, as far as we know, this is the first time that structural data for these α

supported by the analysis of the M-Ob and M-Ob’ distances, which present insignificant differences while varying X, as can be observed in Table 1. It is known that DFT methodologies, as those employed here, reproduce very well the geometry of Keggin heteropolyanions, with the exception of M-Ot bond distances, which have shown deviations that overestimate these distances by about 0.004 Å when compared with the experimental ones.[55,56] Along the molybdenum and tungsten series, M-Ot bond distances does not show important differences between them. It must be noted that W-Ot distances are longer than Mo-Ot distances.

**Table 1 Calculated bond distance for **α**-Keggin heteropolyanions. **

!

**Description Table 1: Calculated bond distances (Å) for **_{α}

**-**

**Keggin heteropolyanions. Atom**

**labeling referred to Figure 6. **

**4.1.3 Vibrational Analysis **

Here we report for the first time, the vibrational frequencies and their assignement for the titled α-Keggin heteropolyanions. A complete vibrational

analysis of several [XMo12O40]q- Keggin anions has been carried out by A.J. Bridgeman [53,54] and X. López et.al.[55] In those articles, the authors indicate that DFT at LDA level, employing STO-TZP basis sets including ZORA corrections yield a better description of the vibrational spectra than employing quasi-relativistic effective core potentials to represent the atomic cores and Gaussian type orbitals (GTO) with double-z quality basis sets for the valence orbitals together with Becke’s three parameter hybrid functional (B3LYP). Thus, we have employed the methodology suggested by Bridgeman and López.

An α-Keggin ion [XM12O40]q- under *Td* symmetry posses 153 normal modes of

vibration that spans as:

!!"#= !!!! ! +!!!!+!"!!! ! +!"!!!+!!!!!(!",!)**!**

frequencies in comparison with SnM and TeM for each series (M=Mo, W) at the

**Description Table 2: Calculated Vibrational Frequencies (cm-1) and their intensities **

**(kcal/mol) for **!**-Keggin heteropolyanions. Atom labeling referred to Figure 6. **

**4.1.4 Electronic Structure **

yields a set of unoccupied frontier orbitals that conform the metallic band. Figure 9 corresponds to a molecular orbital (MO) diagram for the studied compounds, where some orbitals of the oxo and metallic bands are plotted as examples. The order and composition of the MOs does not depend of the metals (Mo or W). However, the HOMO-LUMO (H-L) gap is smaller in the case of molybdates, which results in a stronger oxidizing power, i.e. molybdenum clusters are more easily reduced. The other fundamental variable that plays an essential role in the oxidant power is the total negative charge of the anion. Pope and Varga rationalized the first cathodic potential dependence with the negative charge for Keggin anions.[77]

Since the total number of electrons increases by two from Sn to Te, these two extra electrons are located at an antibonding orbital, mainly composed by the tellurium moiety, stabilizing the corresponding 12a1 orbital, reducing the H-L gap from ~2.00 eV for SnM to ~0.40 eV for TeM heteropolyanions.

This suggest that **TeM POMs, are readily available for oxidation, which is **
centered at the tellurium heteroatom, resulting in stable clusters with a H-L gap
of about 2.00 eV.

**4.1.5 Energy Decomposition Analysis (EDA) **

In order to obtain a better understanding of the {XO4} encapsulation by the {M12O36} cage, we analyzed the interaction between this two subunits by means of the Morokuma-Ziegler energy decomposition analysis [43, 60, 61]. These results are presented in Table 3.

**Table 3 Energy Decomposition Analysis of the fragments interaction for **α

**-Keggin heteropyanions. **

!E**Int** !E**Pauli** !**EElstat** !**EOrb** !E**Solv**

**SnMo ** **LDA ** -1273.26 1816.34 -2023.81

(77%)

**SbMo ** **LDA ** -650.13 1489.15 -1436.21

(76%)

**TeMo ** **LDA ** -1223.95 1484.42 -1731.48

(77%)

**SbW ** **LDA ** -815.05 1369.66 -1496.93

(77%)

**4.1.6 Conclusions **

High level density functional calculations including scalar relativistic corrections have been used to predict the molecular and electronic structure of six [XM12O40]q- α-Keggin heteropolyanions, namely [SnMo12O40]4-, [SbM12O40]3-,

[TeMo12O40]4-, [SnW12O40]4-, [SbW12O40]3- and [TeW12O40]4-.

Geometry optimizations were carried out, showing that the cluster possess *Td *

symmetry, with longer M-Ot distances for W than Mo, and smaller variations inside a serie (Mo and W).

Also, the vibrational frequencies have been calculated and their most intense IR active signals have been assigned to the corresponding stretching and bending vibrational modes. It must be noted that the study of the vibrational spectra of these systems are intensely computationally demanding. Thus, they cannot be considered a current standard task for the study of these large, heavy element cluster anions.

Energy decomposition analysis of the {XO4}-{M12O36} interaction, shows a predominant ionic character (~75%), due to the high negative charge of the {XO4} anionic subunit which lift up in energy the orbitals available for this interaction.

**4.2 Electronic Structure of [(TeW**

**9**

**O**

**33**

**)**

**2**

**Pd**

**3**

**]**

**10-**

**. **

**A self assembled Nano-Object from a **

**Polyoxometalate building block. **

**4.2.1 Introduction **

In recent years, there has been an unprecedented rise in the synthesis and development of novel polyoxometalate (POM) clusters, most of them have been structurally characterized by single crystal X-ray crystallography. Since POMs are prepared with well defined structures, based on conserved building blocks, and due to the fact that they are formed via self assembly, they have great potential to bridge multiple length scales. Different building blocks can allow the assembly of new nano-objects with configurable architectures.[69,70] Into this context, we have focused our attention in the {XW9}, {X2W15}, {X3W21} lacunary

vacant species

(X = P(IV), S(IV), As(V), Te(VI). . . , see Figure 7), which have the ability to stabilize diverse inorganic and organometallic fragments.[71]

package.[46,59]

**4.2.2 Computational Details **

Density functional theory (DFT) calculations were carried out using the Amsterdam density functional (ADF) program.[59] Full geometry optimizations were done using the analytical gradient method implemented by Verluis and Ziegler.[46]

Spin-unrestricted calculations were performed for all the open-shell systems. Local electron correlation was treated through the local density approximation (LDA) within the Vosko, Wilk and Nusair parametrization (VWN),[37] where nonlocal corrections through the generalised gradient approximation (GGA) proposed by Becke [38] and Perdew [39] were incorporated into the exchange and correlation functionals respectively, and treated by a fully-self consistent method.

!

**Figure 7 POM Cluster Position **

To simulate water solvent effects a conductor like screening model (COSMO) was applied.[57,58,73] The molecular structures described here have been characterized as the energy minima through calculations of normal mode vibrational frequencies, which were performed at the optimized geometries via second derivatives of the total energy with respect to the internal coordinates as implemented in the ADF code.[59,74,75]

**4.2.3 Results and Discussion **

In Figure 7, we present the molecular structure of the studied Pd3-POM system, the main selected distances are presented in Table 4. The W-O distances for the W=O and W-O-W moieties are in the range of those encountered for Keggin tungsten polyoxometalates, as well as the Te-O distances. The Pd-O distance is longer that the expected for a replacement of a W atom at the present positions. Into the same context, the large Pd-Pd distances (5.359 A) indicate that it does not exist a direct Pd-Pd interaction, i.e. there are not present metallophilic interactions. The calculated stretching modes for the vibrational frequencies are presented in Table 5. The calculated stretching modes for W=O and W-O-W are displaced c.a. 50 cm-1 with respect to the experimental values.[70,71] The electronic excitations were calculated including water solvent effects, and their results are given in Table 6.

!

**Figure 8 Description Molecular Structure **

**Description Figure 8: Front and side view of the Molecular Structure and atom-labeling **

**scheme for [(TeW9O33)2Pd3]10-: i) Yellow = Te(IV); Blue = W(VI); Red=O; Gray=Pd. ii) **

**Labeling corresponds to Wu = Tungsten cap metal centers; Wu = Tungsten belt metal **
**centers; Otu= Oxygen terminal atoms of cap centers; Otu= Oxygen terminal atoms of belt **
**centers; Obu= Oxygen bridging metal atoms at the cap metal centers; Obc= Oxygen **
**bridging metal atoms at the cap and belt metal centers; Obd=Oxygen bridging metal **
**atoms at the belt metal centers; OPd = Oxygen bridging palladium and tungsten atoms; **

**Table 4 Calculated Select Bond Distance to Figure 6. **

!

**Bond Type ** **Experimental ** **LDA/TZP ** **LDA/TZ2P ** **BP86/TZP **

**Pd-OPd **

**Table 5 Calculated Select Bond Distances to Figure 7 **

!

**Bond Type ** **BP86/TZ2P ** **PBE/TZP ** **PBE/TZ2P ** **PW91/TZP ** **PW91/TZ2P **

**Pd-OPd ** 2.005 2.019 2.005 2.018 2.003

**Wd-OPd ** 1.845 1.852 1.845 1.852 1.844

**Wd-Obd ** 1.973 1.981 1.973 1.980 1.971

**Wd-Ob’d ** 1.959 1.966 1.958 1.967 1.956

**Wd-Otd ** 1.766 1.770 1.765 1.768 1.765

**Wd-Obc ** 2.026 2.034 2.026 2.033 2.025

**Wu-Obc ** 1.922 1.927 1.921 1.928 1.921

**Wu-Obu ** 1.944 1.952 1.944 1.952 1.943

**Wu-Otu ** 1.757 1.762 1.757 1.760 1.756

**Te-Oi ** 1.939 1.957 1.932 1.964 1.937

**Wd-Oi ** 2.369 2.372 2.379 2.362 2.367

**Wu-Oi ** 2.519 2.512 2.523 2.511 2.516

**Description Table 5: Calculated selected bond distances (A) for [(TeW9O33)2Pd3]10-. Atom **

**Table 6 Calculated Vibrational Analysis **

!

**Symmetry ** **Assignation ** **Experimental ** **LDA/TZP ** **LDA/TZ2P **

**A2” ** uas(W-Otu) 954 939 (885) 948 (802)

**E’ ** uas(W-Otu) 934 (100) 943 (71)

**A” ** uas(Pd-OPd),

uas(W-Otu)

924 (598) 936 (436)

**E’ ** uas(Pd-OPd),

uas(W-Otu)

887 921 (1666) 931 (1606)

**E’ ** uas(Pd-OPd),

uas(W-Otu)

890 (2462) 899 (2525)

**E’ ** uas(Pd-OPd) 844 (2119) 853 (2167)

**A2” ** uas(Pd-OPd),

uas(W-Otu)

793 844 (3773) 851 (3668)

**E’ ** uas(W-Obu-W) 743 742 (218) 750 (226)

**E’ ** uas(W-Obu-W) 729 (17532) 735 (17088)

**A2” ** uas(W-Obu-W) 687 (12152) 694 (12409)

**E’ ** uas(W-Obu-W) 661 (216) 683 (529)

**A2” ** uas(W-Obu-W) 574 (77) 648 (118)

**Description Table 6: Calculated Vibrational Frequencies (cm-1) and their intensities **

**Table 7 Calculated Vibrational Frequencies **

!

**Symmetry ** **BP86/TZP ** **BP86/TZ2P ** **PBE/TZP ** **PBE/TZ2P **

**A2” ** 909 (1357) 918 (1232) 907 (1248) 917 (1186)

**E’ ** 898 (262) 908 (162) 899 (219) 908 (168)

**A” ** 892 (745) 900 (752) 891 (788) 901 (754)

**E’ ** 887 (1199) 897 (957) 887 (1189) 897 (975)

**E’ ** 844 (1349) 855 (1602) 844 (1382) 855 (1546)

**E’ ** 799 (3904) 808 (3984) 800 (3930) 807 (3979)

**A2” ** 794 (1738) 804 (2023) 794 (1598) 803 (1882)

**E’ ** 690 (18711) 701 (15376) 693 (17538) 704 (12056)

**E’ ** 683 (897) 692 (710) 683 (753) 690 (728)

**A2” ** 654 (1413) 675 (4369) 664 (2845) 681 (7917)

**E’ ** 640 (11770) 655 (9416) 643 (10827) 659 (7427)

**A2” ** 613 (582) 630 (3443) 621 (1756) 632 (5491)

**Description Table 7: Calculated Vibrational Frequencies (cm-1) and their intensities **

**(km/mol) between parenthesis for [(TeW9O33)2Pd3]10-. Atom labeling referred to Figure 7. **

**4.2.4 Conclusions **

Here we replaced W by Pd and the Pd-O distances are larger. The same occur for the Pd-Pd distance that are quite large indicating not a direct bonding interaction.

! !

!

! !

**Summary **

! ! !

**5.1 Overall Conclusions **

All calculations were performed using the software package Amsterdam Density Functional (ADF), within the framework of the relativistic density functional theory (RDFT), where relativistic effects scalar (SR) and spin-orbit (SO) which are incorporated by a two components Hamiltonian by regular zeroth approximation (ZORA). The use of different functional (LDA, GGA) in order to choose the most suitable method for post calculations are further explored. The solvent effect is modeled using a dielectric medium, by COSMO methodology, which is implemented in the ADF package. In the case of a-Keggin heteropolyanions [XM12O40] for the calculations of electronic density with scalar relativistic corrections they were used in order to predict the molecular and electronic structure of the cluster. Geometry optimizations showed Tetrahedric (Td) symmetry with small variations in the distances of the metal with oxygen terminal. With this, the most intense vibrational frequency signals demanded a computational study, because these large anionic cluster with heavy elements considered the predominant anionic heteropoly was achieved compared with total different loads. In the case of POM- Pd, the charge transfer is in the near-infrared (NIR) region of the electromagnetic spectrum. These POM-Pd NIR bands may have potential for technological applications.

[1] A. Muller, E. Krickemeyer, H. Bogge, J. Meyer, F. Peters, W. Plass, E. Diemann, S. Dillinger, F. Nonnenbruch, M. Randerath, and C. Menke. Angew. Chem.Int. Ed. Engl, 2122:34, 1995.

[2] D. L. Long, E. Burkholder, and L. Cronin. Chem. Soc. Rev, 105:36, 2007.

[3] A. Muller, E. Beckmann, H. Bogge, M. Schmidtmann, and A. Dress. Angew. Chem, Int. Ed. Engl, 1162:41, 2002.

[4] A. Müller and S. Roy. Coord. Chem. Rev, 153:245, 2003.

[5] T.B. Liu, E. Diemann, H.L. Li, A.W.M. Dress, and A. Müller. Coord. Chem. Rev,

426:59, 2003.

[6] C.L. Hill. Chem. Rev, 1:98, 1998.

[7] J.J. Berzelius. Poggend. Ann. Phys. Chem, 6:369, 1826.

[8] A.P. Ginsberg. Early Transition Metal Polyoxoanions, Inorganic Syntheses, 71:27, 1990.

[9] M.C. Sneed. Comprehensive Inorganic Chemistry. Van Nostrand, New York,Tokio, 1953.

[10] M.T. Pope. Heteropoly and Isopoly Oxometalates. Springer-Verlag, Berlin, Heidelberg, New York, Tokio, 1983.

[11] J. Keggin. Nature, 908:131, 1933.

[12] L.C.W. Baker and J.S. Figgis. J. Am. Chem. Soc, 3794:92, 1970.

[13] A. Proust, R. Thouvenot, and P. Gouzerh. Chem. Commun, 1:1837, 2008.

[14] C. Bustos, D. MacLeod Carey, K. Boubekeur, R. Thouvenot, A. Proust, and P. Gouzerh. Inorg. Chim, 4262:363, 2010.

[15] D.-L. Long, R. Tsunashima, and L. Cronin. Angew. Chem., Int. Ed. Engl, 1736:49, 2010.

[16] B. Dawson. Acta Cryst, 113:6, 1953.

[17] C.R. Graham and R.G. Finke. Inorg. Chem, 3679:47, 2008.

[18] T. Yamase. Chem. Rev, 307:98, 1998.

[19] E. Coronado and C. J. Gómez-Garcia. Chem. Rev, 273:98, 1998.

[20] E. Coronado, C. J. Gómez-Garcia, and J. Mater. Chem. Rev, 66:15, 2005.

[22] P. Gomez-Romero. Adv. Mater, 163:13, 2001.

[23] D.L. Long and L. Cronin. Chem. A Eur. J, 2698:14, 2006.

[24] P. Gouzerh and A. Proust. Chem. Rev, 77:98, 1998.

[25] A. Kaneko, A. Yagishita, K. Yahashi, T. Kubota, M. Omura, K. Matsuo, I. Mizushima, K. Okano, H. Kawasaki, S. Inaba, T. Izumida, T. Kanemura, N. Aoki, K. Ishimaru, H. Ishiuchi, K. Suguro, K. Eguchi, and Y. Tsunashima. Electron Devices Meeting, Wasington DC, 1:100, 2005.

[26] Szabo and Ostlund. Introduction to Advanced Electronic Structure Theory. Dover Publications, 1996.

[27] M. Born and R. Oppenheimer. Zur Quantentheorie der Molekeln. Ann. Physik, 20:389, 1927.

[28] Sakurai. Advanced Quantum Mechanics. Addison Wesley, 1967.

[29] E. Fermi. Z. Phys., 48:73, 1928.

[30] L. H. Thomas. Proc. Cambridge Philos. Soc. Phys., 23:542, 1927.

[31] P. Hohenberg and W. Kohn. Phys. Rev., 136:864, 1964.

[32] P. Hohenberg and W. Kohn. Phys. Rev, 4:140, November 1965.

[33] W. Yang R. G. Parr. Density-Functional Theory of atoms and Molecules. Oxford University, New York, United State of America, 1989.

[43] K. J. Morokuma. Chem. Phys., 55:1236, 1971. Arratia-Pérez. J. Phys. Chem. A, 111:6563, 2007.

[51] A.J. Bridgeman and G. Cavigliasso. J. Phys. Chem., 107:4568, 2003.

[52] A.J. Bridgeman and G. Cavigliasso. Chem. Phys., 279:143, 2002.

[53] A. J. Bridgeman. Chem. Phys, 287:55, 2003.

[59] ADF 2000.01; Department of Theoretical Chemistry, Vrije Universiteit: Amsterdam. Baerends, E. J.; Ellis, D. E.; Ros, P. Chem. Phys., 2, 41-51,1973. Versluis, L.; Ziegler, T. J. Chem. Phys. 88, 322-328, 1988. Te Velde, G.; Baerends, E. J. J. Comput. Phys., 99, 84-98,1992. FonsecaGuerra, C.; Snijders, J. G.; Te Velde, G.; Baerends, E. J. Theor. Chem. Acc., 99, 391-403,1998.

[60] F. M. Bickelhaupt and E. J. Baerends. Reviews in Computational Chemistry. edited by K. B. Lipkowitz and D. B. Boyd, Wiley VCH, New York, 2000.

[61] T. Ziegler and A. Rauk. Theor. Chim. Acta, 46:1, 1977.

[62] Y. L. Si, H. L. Yu, X. Y. Zhao, C. A. Zhao, and L.J. Kang. Journal of Northeast Normal University Natural Science Edition, 43:84, 2011.

D. I. Ishutenko, and E.E. Vishnevskaya. Catalysis in Industry,7:30, 2015.

[64] R.P.Singh, R.P. Khatri, J. Dubois, S.S. Gaur, and M. Abe. J. Chem. Soc. Dalton Trans., 947, 1990.

[65] G.C. Bhattacharya and S.K. Roy. J. Indian. Chem Soc., 50:359, 1973.

[66] B. Deniau, J. M. M. Millet, S. Loridant, N. Christin, and J.L. Dubois. Journal of Catalysis, 260:30, 2008.

[67] J. Wang, A. Li, G. Bing, Y. Qiao, H. Zhao, and X. Wang. Chemical Industry and Engineering Progress, 29:2298, 2010.

[68] V.T.C. Hoa and et.al D.V. Dong. Study on Analytical Methods of Tellurium Content in Natriiodide (NA1311) Radiopharmaceutical Solution Produced in the Dalat Nueclar Reactor. in the Annual report for 2006. The Vietnam atomic energy commission, Hanoi, Vietnam, 233-239, December 2007.

[69] D. E. Katsoulis. Chem. Rev, 359388:98, 1998.

[70] J. Yan, D.L. Long, E. F. Wilson, and L. Cronin. Angew. Chem, 44404444:121, 2009.

[71] J. Yan, D.L. Long, and L. Cronin. Angew. Chem, 4211- 4214:122, 2010.

[72] S.J.A. Van Gisberger C. Fonseca Guerra E.J. Baerends J.G. Snijders T.J. Ziegler G. Te Velde, F.M. Bickelhaupt. Comput. Chem, 931:22, 2001.

[73] A. Berces, R.M. Dickson, H. Jacobsen L. Fan, D. Swerhone, and T. Ziegler. Comput. Phys. Commun, 247:100, 1997.