An Experimental and Theoretical Study of

Transition Metal Complexes

Containing Redox Noninnocent ortho-Dithiolate Ligands

Dissertation for the degree of

Doktor der Naturwissenschaften

Fakultät für Chemie

Ruhr-Universität Bochum

Presented by

Flávio Luiz Benedito

Mülheim an der Ruhr, August 2008

This work was independently carried out between March 2005 and August 2008 at the Max-

Planck Institut für Bioanorganische Chemie, Mülheim an der Ruhr, Germany

Submitted on: 19th August 2008

Referent: Prof. Dr. Karl Wieghardt

Korreferent: Prof. Dr. Nils Metzler-Nolte

Acknowledgements I am grateful to many for the support and motivation they provided me through the challenges of this

work and the amazing years in Mülheim an der Ruhr. I would especially like to express my sincere

gratitude to the following:

Prof. Dr. Karl Wieghardt, for the opportunity to work within his research group and also for

showing me science in the most elegant and profound way. I am grateful for the motivation, guidance

and excitement in each step of the project. His encouragement and fascinating ideas will certainly have

a significant impact on my subsequent career.

Prof. Dr. Niels Metzler-Nolte, for kindly examining my thesis.

Dr. Thomas Weyhermüller, for interpretation of single crystal X-ray diffraction data and for being

one of my closest and precious friends.

Mrs. Heike Schucht, for collection of single crystal X-ray diffraction data and for the friendly

discussions.

Dr. Eckhard Bill, for his wise advices, continual patience in guiding me through spectroscopic data

and the kind support as a friend.

Dr. Eberhard Bothe and Mrs. Petra Höfer, for turning the time spent during the electrochemical

measurements into a real learning experience.

Mr. Frank Reikowski, Mr. Andreas Göbels, Mrs. Ursula Westhoff, and Mr. Jörg Bitter, for their

measurements of EPR, SQUID, GC-MS, and NMR.

Mr. Hans-Ulrich Pieper and Mrs. Rita Wagner, for their helpful hands and advice in the laboratory.

Prof. Dr. John Berry (“Johnny”), for sharing his large knowledge of science, but also for the

delightful musical experiences and the friendship I hope lasts forever.

My “amores” Dr. Nuria Aliaga-Alcade, Dr. Isabelle Sylvestre, and Mrs. Charlotte Creusen, for

their contagious passion for Chemistry, generosity, care, and friendship.

Dr. Melissa Koay (amore), for helping me in many difficult moments, but also for the joy and fun the

time we spent together.

Drs. Connie Lu, Jennifer Shaw, Geoff Spikes, and Corinna Hess, for endless help and advice in the

lab, as well as revision of the manuscript.

Dr. Stephen Sproules, for his endless patience and the wonderful time working together on the

rhenium project.

Dr. József Sándor Pap (Papito) and Mrs. Szabina Pap (Mammy), for giving me the opportunity to

share important moments of my life with such special friends.

Drs. Kallol Ray, for introducing me to the project in the very beginning, Taras Petrenko, for the help

with the DFT calculations, and Serena DeBeer George for the XAS measurements.

Special thanks to my office mates, Mr. Carsten Milsmann (Master Yoda) for his incredible help with

computational Chemistry and fruitful discussions. Dr. Shaun Presow (Obi Wan) for helping me with

the correction of the manuscript. These two best friends made my time in Mülheim unforgettable and

very special.

Drs. Krzysztof Chłopek, Prasanta Gosh, Ruta Kapre, Marat Khusniyarov, Nicoleta Muresan,

Shandan Mukherjee, Sumit Khanra, Swarnalatha Kokatam, Yu Fei Song, Peter Larsen,

Pryabrata Banerjee, Meenakshi Gosh, Marco Flores, Messrs. Nabarun Roy, Biplap Biswas,

Michael Nippe and Bram Pluijmaekers, for a sound academic and friendly life inside the laboratory.

Dearest Marc Herbrand, without whose encouragement, support and friendship this project would

never have been begun, let alone completed. I am greatly indebted to him and to his family.

Dr. Helenice Maida, my sister, confidant, friend, for the continuous help. My special thanks for

always believing in and loving me.

Dr. Oscar Barbosa de Souza Filho, for being present in the most important moments of my life and

proving me with special motivation since the beginning of my journey to Germany. Your support was

crucial to this work.

Dr. Norton Nóbrega, whose contribution is underestimated, if acknowledged through words. I am

endlessly grateful for his friendship.

Mrs. Regina Maria de Campos Rocha, for helping me settle in Mülheim, but mainly, for the years

of friendship and care.

Mrs. Henia and Mr. Dieter Seifert, for embracing me as a member of their family, my greatest

thanks for so much love and care over these years.

Messrs. Tereza Mattos, Maurício Virgens, Luzia and Elmar Griethe, for encouragement and solid

friendship over these years.

Prof. Dr. Shirley Nakagaki, for introducing me to science and for her precious friendship.

Dr. Alexander Straub and family, whose influence and support had strong impact on my scientific

work. I am grateful for the sincere friendship and motivation.

My dearest friends Viviane dos Santos Louro, Débora Cohen, Jefferson Princival, Gisele Afeche,

Danielle Juais, Mario Videira, Thiago Rodrigues, Lourdes and Henrique Rosa, Jailson Pacheco,

Mirna Mei e Miria Garcia, Melissa Koch, for being always present in my life, even over an ocean.

Without their encouragement nothing would be possible.

To my parents, Maria Darci Maragni and Mário Benedito, for their endless love and support. This

work is our victory and they are the most important people in my life.

Finally, I am grateful to the Max-Planck Gesellschaft (MPG) for financial support.

Agradecimentos

Gostaria de expresser meus sinceros agradecimentos a todos aqueles que me ajudaram, direta ou

indiretamente na realizacao deste trabalho e a todos que ficaram na torcida. Em especial meu muito

obrigado a:

Prof. Dr. Karl Wieghardt, pela oportunidade de fazer parte do seu grupo de pesquisa, pela orientação

exemplar e motivação a cada novo resultado.

Prof. Dr. Nils Metzler-Nolte, pela generosidade em revisar e analisar minha tese e fazer parte do

comitê de avaliação.

Dr. Thomas Weyhermüller, pelo auxílio na interpretação dos dados de difração de raio-X de

monocristal e pela preciosa amizade.

Heike Schucht, pela ajuda na coleta de dados de cristalografia e pelos momentos de descontração.

Dr. Eckhard Bill, pela paciência nas discussões diárias sobre EPR e magnetismo, conselhos e terna

amizade.

Dr. Eberhard Bothe e Sra. Petra Höfer, por tornar os momentos das análises de

espectroelectroquímica em momentos descontraídos de aprendizado.

Srs. Frank Reikowski, Andreas Göbels, Jörg Bitter e Sra. Ursula Westhoff, pelo auxílio nas

medidas de EPR, SQUID, RMN e GC-MS.

Sr. Hans-Ulrich Piper e Sra. Rita Wagner, pela ajuda e conselhos no trabalho prático do

laboratório.

Prof. Dr. John Berry (Johnny), por compartilhar não só seu vasto conhecimento sobre ciência, mas

também pelos momentos de muita música juntos que tornaram a vida em Mülheim muito mais

prazerosa. Meu profundo agradecimento pela amizade que espero durar por toda vida.

Dra. Nuria Aliaga-Alcade (carinho), pela ajuda de inestimável valor, a amizade preciosa e por me

acolher em seu coração gigantesco. Meu reconhecimento pela amiga e profissional que ela é.

Dr. Isabelle Sylvestre e Charlotte Creusen (amores), pelo carinho, incentivo e cuidado.

Dr. Melissa Koay (amore), pelo apoio nos momentos difíceis e pelos momentos de descontração.

Drs. Connie Lu, Jennifer Shaw, Geoff Spikes, and Corinna Hess, pela ajuda incansável e a

trabalhosa revisão da tese.

Dr. Stephen Sproules, pelo entusiasmo e o prazer de trabalhar em conjunto com os compostos de

Rênio.

Dr. József Sándor Pap (Papito) e Szabina Pap (Mammy), por compartilhar várias conquistas e

momentos de alegria.

Drs. Kallol Ray, pela ajuda no início do projeto e seu legado na Química de enxofre, Taras

Petrenko, pela ajuda com os cálculos de DFT, e Serena DeBeer George pelas medidas de absorção

de raio-X, (SLAC – Standford).

Especialmente aos meus colegas de escritório, Carsten Milsmann (Mestre Yoda) pela ajuda diária

em Química Computacional e ajuda em todo o projeto, Dr. Shaun Presow (Obi Wan), pela amizade

e ajuda na correção do manuscrito. Estes dois grandes amigos tornaram os dias em Mülheim an der

Ruhr inesquecíveis e muito especiais.

Drs. Krzysztof Chłopek, Prasanta Gosh, Ruta Kapre, Marat Khusniyarov, Nicoleta Muresan,

Shandan Mukherjee, Sumit Khanra, Swarnalatha Kokatam, Yu Fei Song, Peter Larsen,

Pryabrata Banerjee, Meenakshi Gosh, Marco Flores, Srs. Nabarun Roy, Biplap Biswas, Michael

Nippe e Bram Pluijmaekers, pelo ambiente de trabalho mais descontraído e prazeroso possível.

Querido Marc Herbrand, por acreditar no meu potencial, pela ajuda inestimável e por me ajudar a

tornar um sonho em realidade. Meu reconhecimento e gratidão a ele e toda sua família.

Helenice Maida, minha irmã, amiga, confidente e amiga. Obrigado pela ajuda e carinho de sempre.

Oscar Barbosa de Souza Filho, por estar sempre presente nos momentos mais importantes e pela

motivação em vencer sempre novos desafios. Sem sua ajuda, tudo seria mais difícil.

Meu querido amigo Dr. Norton Nóbrega, pela contribuição inestimável, impossível de ser expressa

em palavras. Meu mais profundo agradecimento pela torcida e incentivo diários, bem como a honra de

ser seu amigo.

Sra. Regina Maria de Campos Rocha, pela ajuda com toda a documentação necessária para o início

do doutorado, mas principalmente pela amizade sincera de todos esses anos, meu muito obrigado pelo

momentos inesquecíveis que passamos juntos. Esse trabalho não seria possível sem sua ajuda.

Henia e Dieter Seifert, por me acolherem como um verdadeiro filho, meu mais sincero

agradecimento pelo amor, carinho, ajuda e incentivo em todos esses anos.

Tereza Mattos, Amina, Warren Richardson e Maurício Virgens, por nunca deixarem “minha

peteca cair” e pelos momentos maravilhosos que compartilhamos nesses anos de luta.

Luzia e Elmar Griethe, pela amizade carinho e preocupação para que tudo desse certo desde o

começo.

Dr. Alexander Straub e família, por “abrir as portas” da Alemanha para mim, meu muito obrigado

pela instimável experiencia em um ano de Bayer CropScience, por acreditar nas minhas capacidades,

mas também pela amizade que guardo com apreço.

Meus queridos amigos e fiéis escudeiros(as), Viviane dos Santos Louro, Débora Cohen, Jefferson

Princival, Gisele Afeche, Danielle Juais, Mário Videira, Thiago Rodrigues, Jailson Pacheco,

Mirna Mei e Miria Garcia, por estarem sempre presentes em minha vida, mesmo através de um

oceano de distância.

Meus queridos Lourdes e Henrique Rosa, pela amizade sincera e todo o apoio moral e espiritual

indispensáveis para o início e conclusão desta jornada.

Melissa Koch, a grande incentivadora pela empreitada em terras distantes, registro aqui meu

reconhecimento e gratidão por todo carinho, compreensão, ajuda, motivação e pela honra de poder

compartilhar tantos altos e baixos com uma alma tão grandiosa. Este jornada não teria começado sem

seu apoio!

Meus queridos pais, Maria Darci Maragni e Mário Benedito, razão da minha existência, como

agradecer em palavras tanto amor? Este trabalho é o fruto do apoio e do carinho incondicionais de

vocês, as duas pessoas mais importantes da minha vida. Muito obrigado por tudo!

Finalmente, gostaria de agradecer ao Max-Planck Gesellschaft (MPG) e ao Governo Alemão pela

bolsa de estudos.

for Marc Herbrand

“It requires a very unusual mind to undertake the analysis of the obvious”.

Alfred North Withehead (*1861 – †1947)

CONTENTS

Chapter 1 Introduction 1

1.1 – General Introduction 3

1.2 – Objectives of this Work 12

1.3 – References 15

Chapter 2 Molecular and Electronic Structure of Square Planar Nickel, 18

Copper, and Gold Complexes with a New

ortho-Benezedithiolate Ligand

2.1 – Introduction 20

2.2 – Synthesis and X-ray Crystal Structures 21

2.3 – Electro- and Spectroelectrochemistry 29

2.4 – Magnetic Properties 35

2.5 – Theoretical Calculations 39

2.6 – Conclusions 45

2.7 – References 46

Chapter 3 Dimerization Processes of Square Planar 50

[PtII(tbpy)(dithiolate•)]+ Radicals

3.1 – Introduction 52

3.2 – Synthesis and X-ray Crystal Structures 55

3.3 – Sulfur K-edge X-ray Absorption Spectroscopy (XAS) 60

3.4 – Electro- and Spectroelectrochemistry 62

3.5 – X-Band EPR Spectroscopy 66

3.6 – Estimation of Equilibrium Constants 72

3.7 – Conclusions 75

3.8 – References 76

I

Chapter 4 Electronic Structure of Square Planar Cobalt and Rhodium 79

Complexes Containing a bis(ortho-Benzenedithiolate) Ligand

4.1 – Introduction 81

4.2 – Synthesis and X-ray Crystal Structures 85

4.3 – Electro- and Spectroelectrochemistry 89

4.4 – Magnetic Properties 95

4.5 – Preliminary Reactivity Studies 100

4.6 – Theoretical Calculations 103

4.7 – Conclusions 120

4.8 – References 121

Chapter 5 Synthesis and Characterization of Chromium Complexes 124

With ortho-Benzenedithiolate Based Ligands

5.1 – Introduction 126

5.2 – Synthesis and X-ray Crystal Structures 127

5.3 – Electro- and Spectroelectrochemistry 131

5.4 – Magnetic Properties 136

5.5 – Theoretical Calculations 139

5.6 – Conclusions 147

5.7 – References 148

Chapter 6 New tris(Dithiolate) Complexes of Rhenium – 151

A Radical Approach

6.1 – Introduction 153

6.2 – Synthesis and X-ray Crystal Structures 156

6.3 – Electro- and Spectroelectrochemistry 162

6.4 – Theoretical Calculations 170

6.5 – X-ray Absorption Spectroscopy (XAS) 184

6.6 – Conclusions 189

6.7 – References 190

II

Chapter 7 Experimental 193

7.1 – Physical Measurements 195

7.2 – Synthesis 199

7.3 – References 220

Chapter 8 Appendix 221

8.1 – Crystallographic Data 223

8.2 – Publication from this Thesis 230

III

List of abbreviations and symbols

A hyperfine coupling constant

Å angstrom

Ar aromatic

B applied magnetic field

B3LYP Becke 3-parameter (exchange), Lee, Yang and Parr (correlation; DFT)

BM Bohr magneton

Bpy bipyridine

Bu butyl

cm centimeter

Cn symmetry axis

°C degree Celsius

CV cyclic voltammetry

Cys cysteine

d doublet

dd double doublet

D axial zero-field splitting parameter

DFT density functional theory

DNA deoxyribonucleic acid

e electron

E½ half potential in electrochemistry

E/D rhombicity

eff effective

EI electron ionisation

ENDOR electron nuclear double resonance

EPR electron paramagnetic resonance

ESI electrospray ionisation

Fc Ferrocene

Fc+ Ferrocenium

fosc oscillator strength

g electron Lande factor

G gauss

GC gas chromatography

GHz gigahertz

IV

H hour

H Hamiltonian operator

His histidine

HOMO highest occupied molecular orbital

Hz hertz

I nuclear quantum number

IR infrared

iso isotropic

isop = i isopropyl

IVCT intervalence charge transfer

kred reduction rate constant

kox oxidation rate constant

K Kelvin

KOtBu Potassium-tert-butylate

l optical pathway length (cm)

L orbital quantum number

LLCT ligand-to-ligand charge transfer

LMCT ligand-to-metal charge transfer

LUMO lowest unoccupied molecular orbital

m meter (or multiplet in NMR)

mm millimeter

M molar = mol dm-3

MCD molecular circular dichroism

Me methyl

MeCN = CH3CN acetonitrile

MeOH methanol

MHz megahertz

min minute

MO molecular orbital

mV milivolt

nm nanometre

NMR nuclear magnetic resonance

OCT octahedral

pa anodic peak

V

pc cathodic peak

Ph phenyl

q quartet

Q quadrupole moment

r transition dipole operator

rt room temperature

s second (or singlet in NMR)

S local spin state (or spin quantum number)

S spin quantum number

SOC spin orbit coupling

SOMO singly occupied molecular orbital

SQUID superconducting quantum interference device

SWV square-wave voltammetry

t triplet

T Tesla (or temperature)

tert = t tertiary

TIP temperature independent paramagnetism

TMS tetramethylsilane or trimethylsilyl

TP trigonal prismatic

Tyr tyrosine

UV-vis ultraviolet-visible

V volt

vs versus

W Watt (or line width in EPR spectroscopy)

XAS X-ray absorption spectroscopy

βN nuclear magneton

δ isomer shift in NMR (or isomer shift in Mössbauer spectroscopy)

ΔEQ quadrupole splitting

ΔH0 enthalpy energy

ΔS0 entropy energy

Δg anisotropy in EPR spectroscopy

α covalency

σ standard deviation

ε extinction coefficient

VI

η asymmetry parameter

θ Weiss constant (or torsion angle)

λ wavelength

ζ spin-orbit coupling constant

Φ dihedral angle

μ dipole moment

μB Bohr magneton

μeff effective magnetic moment

ν frequency

VII

Chapter 1

Chapter 1

Introduction

1

Chapter 1

2

Chapter 1

1.1 - General introduction

The essential biological roles of transition metal ions in certain enzymes have been

recognized for many years. These metal centers provide binding sites and activate specific

bonds of the substrates. Transition metals can access a variety of oxidation states, and thus,

can act as a reservoir of electrons by accepting and donating electrons during redox cycles.

Metals can also stabilize reactive amino acid radicals, e.g., phenoxyl radicals in tyrosine

residues. Free radicals have emerged as a fundamental feature of biochemical catalysis,

associated with enzymes that have evolved strategies to take advantage of radical chemistry in

bond activation and molecular rearrangements.1-9 Radicals are known to play important roles

in biology, and although historically the initial focus was on their deleterious effects, there is

now abundant evidence that radicals are involved in many essential life processes including

DNA replications, respiration and photosynthesis. Free radicals have been recognized as key

elements in the mechanisms of a wide range of enzymes, including ribonucleotide

reductase,10,11 lysine-2,3-aminomutase,12,13 pyruvate-formate lyase,14 biotin synthase,15

prostaglandin H synthase,16 cytochrome c peroxidase,17 DNA photolyase,18 lipoyl synthase,19

and diol dehydrase,20 among others. An example of an extensively studied metalloenzyme is

galactose oxidase (Figure 1.1.1), in which the active site is comprised of a CuII ion

coordinated to a tyrosyl radical. The overall reaction (Equation 1.1.1) catalyzed by galactose

oxidase is the two-electron oxidation of a primary alcohol of galactose to the corresponding

aldehyde, coupled to the reduction of dioxygen to hydrogen peroxide as a byproduct.21

RCH2OH + O2 + 2e- → RCHO + H2O2 Eqn. 1.1.1

The active site of galactose oxidase is a shallow, exposed copper complex, in which the metal

is bound by four amino acid side chains: two tyrosines (Tyr272 and Tyr495) and two

histidines (His496 and His581), as shown in Figure 1.1.1.

3

Chapter 1

Figure 1.1.1 – Tertiary structure of Galactose Oxidase (left). The blue sphere corresponds to

the copper atom in the active site. Magnification of the active site (right). The structure of

Galactose Oxidase was obtained from the PDB (Protein Data Bank).

The residue Tyr272 has been found crystallographically to be chemically modified via cross-

linking with a nearby cysteine (see Figure 1.1.2a). The Tyr-Cys is a ligand coordinated to the

copper center, which becomes oxidized to a radical in the active form of the enzyme (Figure

1.1.2b). This free radical-couple copper complex is extremely stable, and in the absence of

reducing agents has been shown to persist for weeks at room temperature.22 However, it

reacts readly with a variety of electron donors, undergoing single-electron reduction to form a

catalytically inactive non-radical CuII complex (Figure 1.1.2c). Further reaction converts the

CuII center to CuI (Figure 1.1.2d), forming a fully reduced complex that is able to react with

O2 and represents a catalytic intermediate in the catalytic cycle.23

Figure 1.1.2 – a) Tyr272 cross-linked to Cys228 forming the Tyr-Cys moiety in galactose

oxidase. b) Active radical species. c) One-electron reduction product of (a), catalytically

inactive CuII site. d) Reduced active site, which reacts with O2 in the catalytic cycle.

a b c d

HO

SCys228

O

SCys228

Cu OCu

SCys228

OCu

SCys228

II II I

Tyr272Tyr272 Tyr272Tyr272

4

Chapter 1

The overall catalytic reaction expressed in Equation 1.1.1 can be written as separate

reduction and reoxidation steps, consistent with a ping-pong mechanism, as defined by

Whittaker et al.22 In the first (reductive) half-reaction, the oxidized radical species reacts with

a primary alcohol (with rate constant kred) to form two-electron reduced enzyme complex and

the aldehyde product (Figure 1.1.3a). In the second (oxidative) half-reaction, the reduced

enzyme reacts with dioxygen (with a rate constant kox), converting the active site to the radical

complex and forming hydrogen peroxide (Figure 1.1.3b).

a

b

O

SCys228

CuII

Tyr272

+ RCH2OH + RCHO + 2H+kred O

SCys228

CuI

Tyr272

O

SCys228

CuI

Tyr272

+ O2 + 2H+kox

O

SCys228

CuII

Tyr272

+ H2O2

Figure 1.1.3 – a) First half-reductive reaction catalyzed by galactose oxidase. b) Reoxidation

involving dioxygen. (kred = 0.8 – 2.7 x 104 M-1 s-1 and kox = 0.98 – 1.02 x 107 M-1 s-1).23

The most prominent sulfur-centered radical is the thiyl radical generated by hydrogen

atom abstraction from the corresponding thiol by hydroxyl or carbon-centered radicals.24-26

There are also several reports of reactivity at the iron in haemoglobin with thiol compounds

where thiyl radical involvement has been postulated. The oxidation of cysteine by horseradish

peroxidase in the presence of oxygen also forms a thiyl radical, which was demonstrated by

EPR spin-trapping and ENDOR techniques.27,28

Cobalt-thiyl radical interactions are related to ribonucleotide reductases. These

enzymes operate with adenosyl cobalamin as the precursor of a putative transient thiyl protein

radical.29 During the thiol-mediated oxidation of non-phenolic lignin model compounds by

5

Chapter 1

manganese peroxidase it was found that in the presence of MnII, H2O2, and thiols, the enzyme

converts alcohols (e.g. 3,4-dimethoxybenzyl, anisyl, or benzoyl alcohol), to their

corresponding aldehydes. It is suggested that the thiol is oxidized by MnIII to a thiyl radical,

which abstracts a hydrogen from the substrate and forms a benzylic radical. The latter reacts

with another thiyl radical to yield an intermediate that decays to the benzaldehyde product.30

Chromium toxicity appears to be related to thiyl radical interconversion. CrVI

carcinogenesis was assumed to depend on the presence of cellular redox components,

including thiols, which reduce the hexavalent metal ion into reactive species capable of

interacting with DNA.31,32 Likewise, in vivo toxicity of VV has been found to correlate with

the depletion of cellular glutathione and related nonprotein thiols. Thus, the oxidation of

glutathione, cysteine, N-acetylcysteine, and penicillamine by VV was investigated. In the

course of this process the corresponding thiyl radical and VIV complexes were generated. The

authors suggested that free radical reactions play a significant role in the depletion of cellular

thiols by VV and hence in its toxicity.33

General mechanisms and the intermediates involved in catalytic cycles of some

enzymes are not clear to date. Several intrinsic factors, such as strong coupling between a

coordinated radical and the metal center can make the detection of the radical species very

difficult by common techniques such as EPR spectroscopy and low-temperature

crystallographic measurements. Thus, the design and study of small complexes with redox-

active ligands is of great interest in order to understand the physical characteristics, bonding

properties, and electronic structure of coordinated ligand-based radicals. Such insights may

help elucidate the mechanisms and reactivity of metalloenzymes.

In order to understand the important features discussed above, the concept of an

oxidation number (state) must be taken into consideration. The formal oxidation state can be

defined as the charge left on the metal after all ligands have been removed in their normal

closed-shell configuration.34 Jørgensen proposed that an oxidation state, derived from a

known dn configuration, should be specified as the physical oxidation number,35 implying that

it is possible to measure the number using different spectroscopic methods. Often the formal

and physical oxidation states of a metal in a complex are identical. However the presence of

redox active ligands complicates the picture. For example, in [Co(NH3)6]3+, the low-spin d6

cobalt ion has both a formal and physical oxidation states of +III. Conversely, discrepancies

arise if we consider an O-coordinated phenoxyl radical complex of an iron ion with a d5

configuration. The formal oxidation number for the iron is +IV, after the closed-shell

phenolate anion is removed. On the other hand, Mössbauer and resonance Raman

6

Chapter 1

spectroscopies unequivocally proved the presence of a high-spin d5 electron configuration at

the metal ion. Thus, the iron ion has a physical oxidation number +III.36 As these examples

show, formal and physical oxidation numbers are not always synonymous.

In coordination chemistry the terms innocent and noninnocent ligands are widely used

to emphasize the fact that some ligands do not necessarily possess a closed-shell

configuration. These terms can only be used meaningfully in conjunction with the physical

oxidation state of the metal ion. Several noninnocent ligand classes can be identified and

interestingly, the different oxidation levels can be distinguished by using high quality X-ray

crystallography performed at cryogenic temperatures. Thus it is experimentally possible to

distinguish between the two electronic structures I and II shown in Figure 1.1.4.

BM

Az+

BM

A(z-1)+

I II

Figure 1.1.4 – Different oxidation states of noninnocent ligands. A and B can be O, NR or S.

In general, the C–A/B bond lengths vary systematically. In the N,S-coordinated ortho-

aminothiophenolate(1-) [(LNSAP)]1- ligand a C–N bond length of ~ 1.46 Å and a C–S bond

distances of ~ 1.76 Å are observed. In contrast, in ortho-imidothiophenolate(2-) [(LNSIP)]2-,

C–N and C–S bond distances average ~ 1.40 and ~ 1.75 Å, respectively.

NM

S

H HNH

MS

NM

S

H

1.761.46

1.751.40

1.721.361.42

1.431.38

1.42

1.36 1.41

av. C–C 1.39 ± 0.01 Å

[(LNSAP)]1-

av. C–C 1.39 ± 0.01 Å

[(LNSIP)]2-

[(LNSISQ)]1-

Figure 1.1.5 – Redox activity and selected bond lengths (Å) of ortho-aminothiophenolate

ligands.

7

Chapter 1

The C–N and C–S bond distances are intermediate between those of a single and double bond,

i.e. at ~ 1.36 Å and ~ 1.72 Å, respectively in the ortho-iminothiobenzosemiquinonate(1-)

[(LNSISQ)]1- π-radical ligand (Figure 1.1.5).37-41

Similar trends have been established for O,O-coordinated catecholate(2-) [(LCat)]2-,

benzosemiquininonate(1-) [(LSQ)]1- and benzoquinone [(LBQ)]0 (Figure 1.1.6),42 as well as in

the S,S-coodinated ortho-benzenedithiolate(2-) [(LSS)]2- and ortho-

dithiobenzosemiquinonate(1-) [(LSSSQ)]1- π-radical ligands (Figure 1.1.7).43-45

OM

O

OM

O

OM

O1.341.34

1.42

1.411.39

1.41

1.39 1.41 1.36

1.42

1.36 1.42

1.43

1.42

1.301.30 1.22

1.221.48

1.43

1.431.34

1.34

1.45

[(LCat)]2- [(LSQ)]1- [(LBQ)]1-

Figure 1.1.6 – Redox states and selected bond lengths (Å) of ortho-benzoquinone ligands.

In addition to the changes in the C–A bond distances, A,B-coordinated (LSQ)1- radicals

display a quinoid type distortion of the six-membered ring which is not observed in closed

shell analogues. This distortion involves two alternating short C–C distances of 1.37 ± 0.01 Å

and four longer ones of 1.415 ± 0.01 Å, whereas in the closed-shell aromatic mono- and

dianions, the six C–C bond lengths of 1.39 ± 0.01 Å are equidistant.

[(LSS)]2-

SM

S

SM

S1.7701.755

1.4021.383

1.390

1.398 1.426

1.401

1.374 1.391

1.744

1.7521.408

1.382

1.407

1.429

[(LSSSQ

)]1-

Figure 1.1.7 – Redox activity and selected bond lengths (Å) of ortho-benzenedithiolate

ligands.

The majority of bis(benzenedithiolate) based complexes have a square-planar

arrangement. Table 1.1.1 compiles typical geometries according to the d-orbital electronic

8

Chapter 1

configuration of the central metal. Distorted-tetrahedron arrangements only occur for d10

systems, whereas in d4, d8, and d9 systems square-planar arrangements occur exclusively. For

d5, d6, and d7 systems, square-planar arrangements are most common with some isolated

distorted-tetrahedral examples.46

Table 1.1.1 – Geometrical variety of metal bis(benzenedithiolate) complexes. The oxidation

state is given only for examples which the real dn configuration (or physical oxidation state)

has been elucidated by spectroscopic methods.

Distorted tetrahedral Square-planar Square-pyramidal

ZnII, CdII, HgII, AgI, CoII,

FeII, MnII

CuIII, NiII, PdII, PtII, AuIII,

Co, FeII, CrII, MnII

Pd, Pt, Co, Mn, FeIII

Although several different transition metals are listed, Cu, Ni, Pd, Pt and Co are

present in a large number of complexes, in multiple oxidation states and coordinated to a

variety of bis(dithiolate) ligands. Under specific conditions, the flat square-planar units can

sometimes form strongly joined dimers or trimers.47 In these cases, the coordination chemistry

about the central atom is best described as square pyramidal. These aggregates are held

together by strong intermolecular M–S or M–M bonds as shown in Figure 1.1.8. The

examples in Table 1.1.1 are part of a more expansive molecular stacking arrangement. The

nature of the crystal packing is important as it can be a strong indicator of whether a dithiolate

material may exhibit beneficial conductive properties.

9

Chapter 1

2-

Co(1)

Co(1)´

S(2)

S(1)

S(3)

S(4) S(1)´

S(2)´ S(3)´

Cl(1) Cl(2)

Cl(1)´ Cl(2)´

Bond lengths (Å) Angles (°)

Co(1)–S 2.189 S(1)–Co(1)–S(2) 90.4

S–C 1.759 S(3)–Co(1)–S(4) 90.7

C=C 1.406 Co(1)–S(3)–C 104.8

Co(1)–S(3)´ 2.405 S(1)–C=C 118.9

Co(1)•••Co(1)´ 3.104 Co(1)–S(3)–Co(1)´ 84.9

Pt(1)

Pt(1)´

S(1)

S(1)´

S(2)

S(2)´

S(3)

S(3)´

S(4)

S(4)´

Bond lengths (Å) Angles (°)

Pt(1)–S(1) 2.290 S(1)–Pt(1)–S(2) 99.7

S(1)–C 1.730 S(3)–Pt(1)–S(4) 89.9

S(2)–C 1.748 Pt(1)–S(3)–C 104.2

C=C 1.309 S(1)–C=C 124.5

Pt(1)–S(3)´ 2.294 S(1)–Pt(1)–Pt(1)´ 93.4

Pt(1)•••Pt(1)´ 3.015

Figure 1.1.8 – (Top) Selected example of the M–S dimeric structure [Co(tcdt)2]22- (tcdt2- =

3,4,5,6-tetrachlorobenzene-1,2-dithiolate). The dianion is on an inversion center. Selected

bond lengths and angles are listed. The monomeric units demonstrate distortions from

planarity (the angle λ characterized by the planes formed between S(3)–Co(1)–S(4) and S(2)–

Co(1)–S(1) is 22.1° and the Co(1) is out of plane by 0.331 Å).48 (Bottom) Selected example

of a M–M dimeric structure. (λ = 6.10°, Pt(1) is out of plane by 0.145 Å).49

10

Chapter 1

Compared to the large number of bis(dithiolate) complexes reported in the literature,

examples of tris(dithiolate) compounds are rarer. There are roughly 50 homoleptic

tris(dithiolate) complexes reported in the CSD (Cambridge Structural Database).50 The

majority of the six-coordinate complexes exhibit octahedral coordination geometries,

minimizing interligand steric interaction. The initial structural report by Eisenberg and

Ibers51,52 of [Re(S2C2Ph2)3] caught the attention of chemists, because it was the first example

of a molecular compound with trigonal prismatic geometry around the metal ion. Most of the

tris(dithiolate) complexes feature V, Mo and W, with very few examples containing Ti, Zr,

Nb, Ta, Cr, Tc, Ru and Os.46 In addition, there are also tris(dithiolate) complexes of Fe and

Co,53,54 elements that also form homoleptic complexes with two dithiolate ligands. Recently

in our group the electron transfer series of [Cr(LBu)3]1- was studied in detail revealing clearly

that the redox process is ligand-centered, and that the Cr ion remains in the oxidation state

+III.55

Another important class of compounds arises from the combination of dithiolate and

diimine chelating ligands (such as bipyridines) forming square planar species. Complexes

with d8 metals such as PdII, and PtII have been subject to extensive investigation and research

in the last few years. The combination of the π-acceptor diimine with the π-donor character of

the dithiolate results in unique luminescent properties and the ability to perform photo-

induced electron transfer.56,57 Figure 1.1.9 shows the simplest structure of a dithiolate-diimine

mixed-ligand complex of Pt.

N

N

PtS

S

Figure 1.1.9 – Schematic representation of a simple Pt(diimine)(dithiolate) complex.

Systematic variation in the nature of both the diimine and dithiolate ligands can be

used to “tune” the photoluminescent and excited-state properties. In order to understand the

molecular factors which influence the energy lifetime, emission quantum yield, and redox

potentials of the emissive excited state with the purpose of developing the

11

Chapter 1

Pt(diimine)(dithiolate) chromophore for use in light-driven reactions, a comprehensive study

of the system was reported.58

1.2 – Objectives of this work

The focus of this thesis is on complexes containing noninnocent bis(ortho-benzenedithiolate)

ligands, which are subjected to detailed characterization by a variety of spectroscopic

techniques to determine unambiguously the physical oxidation state of the transition metals.

The new ligand 3,6-bis(trimethylsilyl)benzene-1,2-dithiolate (LTMS) was synthesized and the

electron transfer series of homoleptic square-planar complexes of Ni, Cu, Au, Co, Rh and Cr

were prepared. The electron-rich trimethylsilyl substituents are believed to stabilize the sulfur

π-radical after oxidation of its closed-shell dianionic form, providing for easier isolation of

such species. (Figure 1.2.1).

S

S

Si

Si

S

S

Si

Si

e-

+ e-

(LTMS)2- (LTMS•)1-

Figure 1.2.1 – Redox activity of the noninnocent (LTMS)2- ligand.

It is known that in compounds of the type MII(diimine)(dithiolate) (where M = PdII or PtII)

both the diimine and dithiolate moieties are redox active. The reduction process of the

bipyridine moiety has been extensively studied.56,57,59 Conversely, the oxidation processes of

the dithiolate ligands still remain unclear. This class of compound could be considered the

easiest candidates to study the redox properties of a single coordinated dithiolate.

Surprisingly, during the oxidation process complicated dimerization events take place. The

full characterization of the monomeric and dimeric species, their intermediates and also the

determination of thermodynamic constants will be presented. Platinum complexes with two

different dithiolate ligands will be compared (Figure 1.2.2).

12

Chapter 1

Figure 1.2.2 – Representation of two heteroleptic Pt complexes.

Since their report in the early 1960s, the neutral rhenium tris(dithiolate) complex

(Figure 1.2.3) has intrigued the scientific community due to the unknown factors that lead to

the trigonal prismatic geometry. The redox chemistry of these Re compounds remains

unexplored to date.

Figure 1.2.3 – Crystal structure of Eisenberg´s trigonal prismatic [Re(S2C2Ph2)3]0 complex in

two different perspectives.51

In the last part of this work we will present the first examples of structurally characterized Re

tris(benzenedithiolate) monoanions for two different ligands. DFT calculations on the crystal

structure will be discussed, as well as the results for the electron transfer series of Re

compounds. The sulfur K-edge and Re L1-edge X-ray absorption spectra will be presented

N

N

Pt

S

S

Si

Si

N

N

S

Pt

S

[PtII(tbpy)(LTMS)] [PtII(tbpy)(LPh)]

Re

S

S

S

S

S S

13

Chapter 1

and compared with the simplest ortho-benzenedithiolate ligand in order to determine the

spectroscopic oxidation state of the rhenium and the ligands. Figure 1.2.4 shows the ligands

used for this purpose.

Figure 1.2.4 – Representation of the dianionic form of the ligands used in the study of

rhenium tris(benzenedithiolate).

S

S

S

S

Cl Si

Cl

S

S

Si

(L)2- (LCl)2- (LTMS)2-

14

Chapter 1

1.3 – References

1 Banerjee, R. Chem. Rev. 2003, 103, 2081-2093. 2 Frey, P. A. Chem. Rev. 1990, 1343-1357. 3 Frey, P. A. Curr. Opin. Chem. Biol. 1 1997, 347-356. 4 Frey, P. A. Annu. Rev. Biochem. 2001, 70, 121-148. 5 Marsh, E. N. Bioessays 1995, 431-441. 6 Pedersen, J. Z.; Finazzi-Agro, A. FEBS Lett. 1993, 325, 53-58. 7 Stubbe, J.; van der Donk, W. A. Chem. Rev. 1998, 98, 705-762. 8 Parsons, A. F. An Introduction to Free Radical Chemistry, Blackwell Science, Oxford. 2000. 9 Perkins, M. J. Radical Chemistry - An Introduction, Oxford University Press, Oxford. 2000. 10 Licht, S.; Gerfen, G. J.; Stubbe, J. Science 1996, 271, 477-481. 11 Sun, X.; Ollagnier, S.; Schmidt, P. P.; Atta, M.; Mulliez, E.; Lepape, L.; Eliason, R.; Graslund, A.; Fontecave, M.; Reichard, P.; Sjöberg, B.-M. J. Biol. Chem. 1996, 271, 6827-6831. 12 Banerjee, R. Chem. Rev. 2003, 103, 2094-2112. 13 Frey, P. A.; Booker, S. J. Adv. Protein. Chem. 2001, 58, 1-45. 14 Knappe, J.; Wagner, A. F. Adv. Port. Chem. 2001, 58, 277-315. 15 Berkovitch, F.; Nicolet, Y.; Wan, J. T.; Jarret, J. T.; Drennan, C. L. Science 2004, 303, 76-79. 16 Dorlet, P.; Seibold, S. A.; Babcock, G. T.; Gerfen, G. J.; Smith, W. L.; Tasi, A. L.; Un, S. Biochemistry 2002, 41, 6107-6114. 17 Huyett, J. E.; Dean, P. E.; Gurbiel, R.; Houseman, A. L. P.; Sivaraja, M.; Goodin, D. B.; Hoffman, B. M. J. Am. Chem. Soc. 1995, 117, 9033-9041. 18 Cheek, J.; Broderick, J. B. J. Am. Chem. Soc. 2002, 124, 2860-2861. 19 Chicchillo, R. M.; Iwig, D. F.; Jones, A. D.; Nesbitt, N. M.; Baleanu-Gogonea, C.; Souder, M. G.; Tu, L.; Booker, S. J. Biochemistry 2004, 43, 6378-6386. 20 Magnusson, O. T.; Frey, P. A. Biochemistry 2002, 41, 1695-1702.

15

Chapter 1

21 Firbank, S. J.; Rogers, M. S.; Wilmot, C. M.; Dooley, D. M.; Halcrow, P. F.; McPherson, M. J.; Phillips, S. E. V. PNAS 2001, 98, 12932-12937. 22 Whittaker, M. M.; Whittaker, J. W. Biochemistry 2001, 40, 7140-7148. 23 Whittaker, J. W.; Whittaker, M. M. J. Biol. Chem. 1988, 263, 6074-6080. 24 Asmus, K. D. Radioprotectors and Anticarcinogens 1983, Academic Press, New York, 23-42. 25 Asmus, K. D. Meth. Enzymol. 1990, 186, 168-180. 26 Mason, R. P.; Rao, D. N. R. Meth. Enzymol. 1990, 186, 318-329. 27 Harman, L. S.; Mottley, C.; Mason, R. P. J. Biol. Chem. 1983, 259, 5606-5611. 28 Sivaraja, M.; Goodin, D. B.; Smith, M.; Hoffman, B. M. Science 1989, 245, 738-740. 29 Mulliez, E.; Fontecave, M. Chem. Ber. Recueil 1997, 130, 317. 30 Wariishi, H.; Valli, K.; Renganathan, V.; Gold, M. H. J. Biol. Chem. 1989, 264, 14185-14191. 31 Aiyar, J.; Berkovits, H. J.; Floyd, R. A.; Wetterhahn, K. E. Chem. Res. Toxicol. 1990, 3, 595-603. 32 Aiyar, J.; Berkovits, H. J.; Floyd, R. A.; Wetterhahn, K. E. Environ. Health Perspect. 1991, 92, 53-62. 33 Shi, X.; Sun, X.; Dalal, N. S. FEBS Lett. 1990, 271, 185-188. 34 Hegedus, L. S. Transition Metals in the Synthesis of Complex Organic Molecules; 1994, University Science Books, Mill Valley, California. 35 Jörgensen, C. K. Oxidation Numbers and Oxidation States, 1969, Springer, Heidelberg - Germany. 36 Chaudhuri, P.; Verani, C. N.; Bill, E.; Bothe, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 2213-2223. 37 Bothe, E.; Verani, C. N.; Weyhermueller, T.; Chaudhuri, P.; Wieghardt, K. Inorg. Biochem. 2001, 86, 154. 38 Ghosh, P.; Bill, E.; Mueller, T. W.; Neese, F.; Wieghardt, K. J. Am. Chem. Soc. 2003, 125, 1293-1308. 39 Ghosh, P.; Bill, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2003, 125, 3967-3979. 40 Herebian, D.; Bothe, E.; Bill, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 10012-10023.

16

Chapter 1

41 Herebian, D.; Ghosh, P.; Chun, H.; Bothe, E.; Weyhermuller, T.; Wieghardt, K. Eur. J. Inorg. Chem. 2002, 1957-1967. 42 Pierpont, C. G. Coord. Chem. Rev. 2001, 216, 99 and references therein. 43 Ray, K.; Begum, A.; Weyhermueller, T.; Piligkos, S.; Van Slageren, J.; Neese, F.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 4403-4415. 44 Ray, K.; Bill, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 5641-5654. 45 Ray, K.; Weyhermueller, T.; Goossens, A.; Craje, M. W. J.; Wieghardt, K. Inorg. Chem. 2003, 42, 4082-4087. 46 Stiefel, E. I.; Schulman, J. M. Prog. Inorg. Chem. 2004, 52, 55-110. 47 Balch, A. L.; Dance, I. G.; Holm, R. H. J. Am. Chem. Soc. 1968, 90, 1139. 48 Baker-Hawkes, M. J.; Dori, Z.; Eisenberg, R.; Gray, H. B. J. Am. Chem. Soc. 1968, 90, 4253. 49 Pomarede, B.; Garreau, B.; Malfant, I.; Valade, L.; Cassoux, P.; Legros, J.-P.; Audouard, A.; Brossard, L.; Ulmet, J.-P. Inorg. Chem. 1994, 33, 3401. 50 Allen, F. H.; Kennard, O. Chemical Design Automation News 1993, 8, 31. 51 Eisenberg, R.; Ibers, J. A. J. Am. Chem. Soc. 1965, 87, 3776-3778. 52 Eisenberg, R.; Ibers, J. A. Inorg. Chem. 1966, 5, 411. 53 Hursthouse, M. B.; Short, R. L.; Clemenson, P. I.; Underhill, A. E. J. Chem. Soc., Dalton Trans. 1989, 1101. 54 Ren, X.; Xu, Y.; Meng, Q.; Hu, C.; Lu, C.; Wang, H. J. Chem. Cryst. 2000, 30, 91. 55 Kapre, R. R.; Bothe, E.; Weyhermueller, T.; DeBeer George, S.; Muresan, N.; Wieghardt, K. Inorg. Chem. 2007, 46, 7827-7839. 56 Zuleta, J. A.; Burberry, M. S.; Eisenberg, R. Coord. Chem. Rev. 1990, 97, 47-64. 57 Zuleta, J. A.; Chesta, C. A.; Eisenberg, R. J. Am. Chem. Soc. 1989, 111, 8916-17. 58 Bevilacqua, J. M.; Eisenberg, R. Inorg. Chem. 1994, 33, 2913-23. 59 Zuleta, J. A.; Bevilacqua, J. M.; Eisenberg, R. Coord. Chem. Rev. 1991, 111, 237.

17

Chapter 2

Chapter 2

Molecular and Electronic Structure of Square Planar

Nickel, Copper, and Gold Complexes

With a New ortho-Benzenedithiolate Ligand

18

Chapter 2

19

Chapter 2

2.1 Introduction

Nickel bis-dithiolate complexes were first reported in the 1960s, and they have been

studied extensively because of their ability to undergo facile one-electron transfer reactions

(Equation 2.2.1).1-4 This ability leads to the formation of a three-membered series of square

planar compounds in which the terminal members are diamagnetic and the monoanions show

S = ½ ground states.

[Ni(L)2]0 [Ni(L)2]1- [Ni(L)2]2-+e+e-e -e

Eq. 2.2.1 It is well established that the dianionic compounds contain a nickel(II) center coordinated by

two ortho-benzenedithiolate(2-) ligands (L), but the electronic structures of the monoanionic

and neutral counterparts are the subject of considerable debate.5 Sellmann et al.6 recently described the electron transfer series [Ni(LBu)2]z

(LBu = 3,5-di-tert-butylbenezene-1,2-dithiolate; z = 0, 1-, 2-) as a metal-centered oxidation

process based on crystallographic and spectroscopic studies. The authors ruled out the

possibility of coordinated ortho-dithiobenzosemiquinonate(1-) radical anions based on the

high stability of the aromatic phenyl rings. However, X-ray crystallographic studies at low

temperature show clear evidence of quinoid-type distortions observed in the neutral complex

with four long and two short C–C bond distances on the phenyl ring, which may imply redox

non-innocence of dithiolate ligands.6,7 The authors do not provide any alternative explanation

for the significant shortening of the C–S bonds in the solid-state structure of the neutral

compound compared to the dianionic species.

A number of low-quality crystal structures of compounds containing the [AuIII(L)2]1-

motif have been reported,8-12 such that large experimental errors in the C–C and C–S bond

lengths means the dianionic closed-shell form and ortho-dithiobenzosemiquinonate(1-)

radical anionic form are indistinguishable. The oxidized complex [Au(L)2]0 has been assigned

as a Au(IV) species with two closed-shell ligands. Interestingly, the medium-quality crystal

structure presented suggests the presence of one ortho-dithiobenzosemiquinonate(1-) radical.7

In contrast, the electronic structure of the [Cu(LMe)2]1- ((LMe)2- = toluene-3,4-

dithiolate) is described in the literature as a Cu(II) metal ion coordinated to one dithiolate(2-)

and one toluene-3,4-dithiosemiquinonate(1-) ligand.13 However, the structure of [Cu(L)2]1-

shows relatively long C–S bond lengths at 1.76 Å and the presence of aromatic phenyl rings

with nearly equidistant C–C bonds. In cases where both ortho-benzenedithiolate(2-) and

20

Chapter 2

ortho-benzenedithiosemiquinonate(1-) ligands are coordinated to the same metal ion, the

structural differences are expected to be small, and may be difficult to characterize

unambiguously.

In this chapter the synthesis and characterization of the new ligand

3,6-bis(trimethylsilyl)benzene-1,2-dithiolate (LTMS)2- and its complexes with copper, gold and

the one-electron transfer series of nickel, are described. The ligand was designed with bulky

trimethylsilyl substituents in the 3 and 6 positions for two reasons: (1) to prevent dimerization

through the formation of intermolecular M–S bonds; and (2) to stabilize ligand π-radicals due

to their inherent electron-donation. A combination of structural, spectroscopic, and DFT

studies are applied in order to study these coordination compounds.

Results and Discussions:

2.2. Syntheses and X-ray Crystal Structures:

The new ligand 3,6-bis(trimethylsilyl)benzene-1,2-dithiolate has been synthesized in

five steps by using 1,2-benzenedithiol as a starting material (Scheme 2.2.1).

Scheme 2.2.1 – Synthesis of 3,6-bis(trimethylsilyl)benzene-1,2-dithiolate. (a) 1eq. H2L, 3 eq.

i-PrBr, 4 eq. NaOH, 0.03 eq. [MeN(n-Bu)3]Cl, 1:1 H2O/C6H6, HCl 10% (b), 5 eq.

n-BuLi/TMEDA, hexane, HCl 10%, 6 eq. Si(CH3)3Cl (c) 5 eq. n-BuLi/TMEDA, hexane, HCl

10%, 6 eq. Si(CH3)3Cl, (d) Na/NH3 at -78°C, HCl 10%, Et2O (e) 2 eq. KOtBu.

SiSH S

S

i-Pr

i-Pr

S

SSH

i-Pr

i-Pr

S

S

i-Pr

i-Pr

Si

Si

SH

SH

Si

Si

S

S

Si

Si

K

K

a, b c, d

e, f, g

h, ij

bdt (1)

(2)(3)

a b

H2L

c

e d

(1b) (1a) (1)

21

Chapter 2

In the first step, the sulfur groups were alkylated with isopropyl bromide in the

presence of NaOH and [MeN(n-Bu)3]Cl giving 1,2-bis(isopropylthio)benzene as a yellow oil

in high yields.14 According to 1H NMR analysis the 1,2-bis(isopropylthio)benzene has 93%

purity, thus was used without further purification.

Only one of the two ortho-hydrogen reacts with n-butyllithium even when the latter is

present in a large excess.15 The equimolar concentration or slight excess of TMEDA

(N,N,N´,N´-tetramethylethylenediamine) is crucial to the completion of reaction because the

chelated lithium complex is more reactive than n-butyllithium itself.16-19 In order to obtain the

bis(trimethylsilyl) compound 1, steps b and c were repeated (Scheme 2.2.1). The timing for

the generation of the lithium salt was optimized exactly to one hour. Periods longer than an

hour lead to the formation of several side-products, such as polysilanols (detected by

GC-MS); and reaction times less than an hour result in a very low yield of 1 (< 15%). The

product was isolated by crystallization from n-hexane solution. Single crystals of 1 suitable

for study by X-ray diffraction were obtained, and the structure diagram of the protected form

of the ligand 1 is shown in Figure 2.2.1 and bond lengths are summarized in Table 2.2.1.

In the final steps, the protecting groups were removed using the Birch reduction with

sodium-ammonia mixture at -78 °C followed by acidification with HCl solution. 20-24

C(1)C(2)

C(3)

C(4)C(5)

C(6)

C(7)

C(11)

C(12)

C(16)

C(18)

C(8)

C(9)

C(10)

Si(13)

S(1)

C(19) C(15)

C(14)

Si(17)

C(20)

S(2)

Figure 2.2.1 – Perspective view and numbering scheme of 1 with thermal ellipsoids at 50%

probability level. Hydrogen atoms are omitted for clarity.

22

Chapter 2

Table 2.2.1 – Selected bond distances (Å) in 1.

S(1)-C(1) 1.782(1) C(2)-C(3) 1.418(1)

S(1)-C(7) 1.843(1) C(3)-C(4) 1.399(1)

S(2)-C(2) 1.782(1) C(4)-C(5) 1.391(1)

S(2)-C(10) 1.844(1) C(5)-C(6) 1.405(1)

C(1)-C(6) 1.409(1) C(3)-Si(13) 1.896(1)

C(1)-C(2) 1.411(1) C(6)-Si(17) 1.897(1)

The impurities and side-products of the Birch reduction were removed by the isolation of the

ligand in its thiol form upon acidification with degassed HCl followed by extraction into

Et2O. Exposure to air gives a dimeric product formed by four-electron oxidation with

oxygen.25-27 The dimer is a trans-dibenzo-[1,2,5,6]-tetrathiocin derivative25 and comprises an

eight-membered ring as shown in the diagram for 1ox (Figure 2.2.2).

Reactions of the dimer using reductants such as metallic sodium, LiAlH4 or NaBH4

produced the desired monomer, though in poor yields.

C(2)

C(3) C(4)

C(5) C(6)

C(1)

S(1)

S(2)

Figure 2.2.2 – Perspective view and numbering scheme of compound 1ox with thermal

ellipsoids at 50% probability level. Hydrogen atoms are omitted for clarity. The bond

distances are similar to that of compound 1. Additional information: the S–S bond distance is

2.0625(3) Å.

23

Chapter 2

Under argon, the dithiol form 1a of the ligand can be deprotonated in situ with KOtBu to give

the potassium salt [K2(LTMS)] 1b which was subsequently used in metallation reactions.

The salt of ompound 2 [K2(μ-OHCH3)2(MeCN)2][Ni(LTMS)2] was synthesized under

argon by adding half an equivalent of NiCl2.6H2O to 1b in methanol followed by the addition

of the [N(n-Bu)4]I salt. Orange crystals of [K2(μ-OHCH3)2(MeCN)2][2] were obtained from a

MeOH/MeCN 1:1 mixture in 80% yield. In the solid-state structure of complex 2 there are

potassium ions solvated by MeOH and MeCN molecules and an absence of [N(n-Bu)4]+ ions.

The K+ ions make short contacts with the dithiolate sulfur atoms at 3.246(2) Å dorsal to the

planar [Ni(LTMS)2]2- moiety (Figure 2.2.3).

O

N

K

Ni S

Si

Figure 2.2.3 – Packing motif of [K2(μ-OHCH3)2(MeCN)2][2] in the unit cell.

Aerial oxidation of an orange solution of [K2(μ-OHCH3)2(MeCN)2][2], affords an

instantaneous color change to bright green, and the formation of the salt of complex 2a

[N(n-Bu)4][Ni(LTMS)2], which was obtained as green crystals from MeCN at -20 °C in 90%

yield. Figure 2.2.4 shows the X-ray crystal structure of the monoanions 2a.

The neutral complex [Ni(LTMS)2]0 2b was obtained by chemical oxidation of 2a with a

stoichiometric amount of tris-(4-bromophenyl)aminium hexachloroantimonate in CH2Cl2

affording purple crystals of 2b in 67% yield. The X-ray crystal structures of the square-planar

moiety of 2 and 2b are not shown due to overall similarity to that of 2a.

24

Chapter 2

1-

Ni(1)

C(1)

C(3)S(2)

S(1)

C(4)

C(6)

C(2)

C(5)

Figure 2.2.4 – Perspective view and numbering scheme of monoanions 2a [Ni(LTMS)2]1- with

thermal ellipsoids at 50% probability level. Hydrogen atoms are omitted for clarity. The

dianions 2 and neutral molecules in 2b are isostructural with the monoanions in 2a.

The corresponding copper and gold complexes were also synthesized. The salt of complex 3,

[N(n-Bu)4][Cu(LTMS)2] was obtained by adding one half of an equivalent of

Cu(CH3COO)2.H2O to 1b in MeOH followed by the addition of [N(n-Bu)4]I salt. The same

procedure was used for the synthesis of the salt of 4 [N(n-Bu)4][Au(LTMS)2], by using

Na[AuCl4].H2O. Both complexes gave a green-colored solution immediately upon exposure

to oxygen. From the corresponding green solutions, microcrystalline solids of [N(n-Bu)4][3]

and [N(n-Bu)4][4] were isolated in ~90% and ~70% yields, respectively. Attempts to prepare

other members of the respective electron-transfer series of Cu and Au were unsuccessful.

Chemical oxidation of 3 or 4 with tris(4-bromophenyl)aminium hexachloroantimonate in

CH2Cl2 solutions resulted in a color change from dark green to pale yellow. Single crystals

were obtained and characterized as the ligand dimer 1ox according to GC-MS and 1H NMR

analysis, suggesting that the complexes decomposed upon oxidation. Figure 2.2.5 shows the

crystal structures of the monoanions 3. Table 2.2.2 summarizes the principal structural

features of the five complexes presented in this chapter. The structures of all five compounds

are of good quality, with experimental errors of C–C and C–S bond distances in the order of

±0.01 Å (3σ).

25

Chapter 2

1-

Cu(1)

C(1)

C(2)

S(1)

S(2) C(3)

C(4)

C(5)

C(6)

Figure 2.2.5 – Perspective view and numbering scheme of the monoanions 3 [Cu(LTMS)2]1-

with thermal ellipsoids at 50% probability level. Hydrogen atoms are omitted for clarity. The

gold compound 4 is isostructural with complex 3.

Table 2.2.2 – Selected bond lengths in Å of the square planar complexes of nickel, copper

and gold.

[Ni(LTMS)2]2-

2

[Ni(LTMS•)(LTMS) ]1-

2a

[Ni(LTMS•)2]0

2b

[Cu(LTMS)2]-1

3

[Au(LTMS)2]-1

4

M(1)-S(1) 2.1684(5) 2.1475(3) 2.1291(2) 2.1783(7) 2.3049(4)

M(1)-S(2) 2.1610(5) 2.1521(3) 2.1230(2) 1.1786(7) 2.3074(4)

S(1)-C(1) 1.770(2) 1.750(1) 1.7237(9) 1.770(3) 1.774(1)

S(2)-C(2) 1.772(2) 1.748(1) 1.7208(9) 1.774(3) 1.774(1)

C(1)-C(2) 1.407(3) 1.414(1) 1.421(1) 1.406(4) 1.403(2)

C(2)-C(3) 1.414(3) 1.422(2) 1.431(1) 1.421(4) 1.404(2)

C(3)-C(4) 1.402(3) 1.395(2) 1.385(1) 1.398(4) 1.403(2)

C(4)-C(5) 1.395(3) 1.399(2) 1.417(1) 1.393(4) 1.387(3)

C(5)-C(6) 1.394(3) 1.395(2) 1.387(1) 1.398(4) 1.397(2)

C(6)-C(1) 1.419(3) 1.421(1) 1.430(1) 1.420(4) 1.409(2)

26

Chapter 2

In the dianionic species, 2, the average C–S bond lengths are relatively long at 1.770

Å. The six C–C distances of the phenyl ring are equidistant (av. 1.405 Å) within experimental

error (3σ), clearly indicate the presence of two closed-shell (LTMS)2- ligands bound to the

nickel in +II oxidation state.6,28,29

For neutral 2b, a pronounced semiquinoid type distortion is observed, with four long

C–C aromatic bonds at an average of 1.424 Å and two short C–C bonds (av. 1.386 Å).

The C–S bond distance shrinks to ~1.72 Å and implies the presence of two (LTMS•)1- π-radical

anions coordinated to a NiII (d8, S = 0) center, which has been observed for other bis-

dithiolate complexes.28,29 The diamagnetic, square planar complex 2b, cannot comprise a NiIV

(d6) ion with two closed-shell ligands as proposed by Sellmann et al.6 and others.30-32 In the

latter case, an S = 1 ground state, as for the isoelectronic34 [CoIII(LBu)2]1-, should be observed.

For the intermediate member of the series 2a [Ni(LTMS•)(LTMS)]1- a shortening of the

C–S bonds is observed, however the C–C bonds in the phenyl rings do not show clear

quinoid-like distortion as observed for similar compounds containing aromatic dithiolate

derivatives.29 In fact, two different electronic structures can be formulated for the

monoanionic complexes: either (1) a MIII center bound to two closed-shell ligands represented

as [MIII(LTMS)2]1-; or, (2) a MII center coordinated to one ligand π-radical and one closed-shell

ligand represented as [MII(LTMS•)(LTMS)]1- (Equation 2.2.1). In the second case, two resonance

structures are expected to have equal contribution due to electron-hopping process through the

metal ion.

[MII(LTMS)(LTMS•)]1- [MII(LTMS•)(LTMS)]1- Eqn. 2.2.1

The expected result will be a C–S bond length of about 1.747 Å, which is the

arithmetic average for the C–S bonds at 1.770 Å in aromatic dithiolates and 1.723 Å for

ligand π-radical. This is indeed observed for the [Ni(LTMS)2]1-, wherein the distortions of the

(LTMS)2- ligands are the arithmetic average of those in [Ni(LTMS)2]2- and [Ni(LTMS•)2]. This

pattern is present in several systems featuring noninnocent ligands with O-, S-, and N- atom

donors with different metal ions.28,29,33-40 The structural trends of the nickel series are

summarized in Figure 2.2.6.

27

Chapter 2

[NiII(LTMS)2]2- [2]2-

[NiII(LTMS•)(LTMS)]1- [2a]1-

[NiII(LTMS•)2]0 2b

Figure 2.2.6 – Schematic representation of bond lengths (Å) changes in the one electron

transfer series of nickel complexes.

In contrast, the bond lengths listed in Table 2.2.2 for the isoelectronic complexes 3 and

4 indicate that both ligands are in their closed-shell dianionic form, which would require a

trivalent metal center (CuIII and AuIII, d8, S = 0).29,41 The average C–S bond lengths is

1.773 ± 0.006 Å and no quinoid-like distortion in the phenyl ring is observed. This

interpretation differs from that of Sawyer et al.,13 who reported a room-temperature crystal

structure of the [N(n-Bu)4][Cu(LMe)2] with an error of ±0.1 Å (3σ) in the C–S bonds. Based

on electrochemical results, the electronic structure of the monoanion was formulated as

[CuII(LMe•)(LMe)]1-. Other reported structures such as the [PPh4][Cu(L)2] and

[PPh4][Cu(LMe)2] with large errors of ~0.02 and 0.03 Å (3σ) in the C–S bond distances are of

insufficient accuracy to determine unambiguously the oxidation level of the ligand.42

Rindorf7 and Schiødt43 have reported a good quality structure of the neutral [Au(L)2]0

complex, which was found to be square planar. Interestingly, the structural parameters are

significantly different from the monoanionic analogue. The average C–S bond of 1.735(6) Å,

is short and the C–C bonds show the alternating pattern of two shorter C=C bonds (average

1.380 Å) and four longer ones (average 1.406 Å). The Au-S bond length of 2.309 Å does not

show any significant changes compared to those in [Au(L)2]1- (average 2.306 Å) which is a

clear evidence that the oxidation process is not metal-centered. Consequently, the neutral

compound can be described as [Au(L•)(L)]0 ↔ [Au(L)(L•)]0.

28

Chapter 2

Recently, Wieghardt et al.41 reported the 197Au Mössbauer of both species in order to

elucidate the coordination numbers and oxidation states of the gold ions. The parameters

observed for [Au(L)2]1- were (δ = 3.36 mm s-1, ΔEQ = 2.92 mm s-1) and [Au(LBu•)(LBu)]

(δ = 3.20 mm s-1, ΔEQ = 3.06 mm s-1). These values are not significantly different to those

obtained when a clear change in the oxidation state of the gold is clearly characterized by 197Au Mössbauer spectroscopy. This is observed in the Au(II) and Au(III) complexes,

Me2C(Ph2PAuCl)2Br2 (δ = 2.01(1) mm.s-1, ΔEQ = 3.58(1) mm.s-1); Me2C(Ph2PAuCl)2Br4

(δ = 1.05(10) mm.s-1, ΔEQ = 1.20(10) mm.s-1).44

Thus, for the complexes [Au(L)2]0 and [Au(L)2]1- it was concluded that the coordination

sphere around the gold ion remains square planar in both oxidation states and after one

electron oxidation the electron configuration remains d8. This fact establishes that the

oxidation of [Au(L)2]1- is ligand-centered and corroborates with our assignment of the redox

processes of the neutral compound 4a (vide infra).

Further experimental support for the ligand oxidation process is based on IR and

Raman spectroscopies.45 For the system containing [M(LBu)(LBu•)]0 (M = Ni, Pd, Pt, and Au)

the phenylthiyl radical stretching ν(C=S•) is observed in the 1000 to 1100 cm-1 region. The

intensity of the band is highly dependent on the substitution pattern of the dithiolate ligand,

which decreases in intensity in the order (LBu•) > (LMe•) > (L•). Two tert-butyl substituents in

the 3,5-positions result in maximum intensity due to its inherent asymmetry.45 In the case of

(LTMS•), in which the two trimethylsilyl substituents are in the 3,6-positions, no similarly

intense bands are observed in the region of 1000 to 1100 cm-1. This is probably due to

cancellation of the vibrational mode imposed by the higher symmetry of the ligand.

2.3 – Electro- and Spectroelectrochemistry:

Figure 2.3.1 shows the cyclic voltammograms of 2a [Ni(LTMS•)(LTMS)]1-, obtained at

different scan rates in a dichloromethane solution with 0.10 M [N(n-Bu)4]PF6 as the

supporting electrolyte, using a glass carbon working electrode and Ag/AgNO3 reference

electrode. Ferrocene was used as an internal standard, and potentials are referenced versus the

ferrocenium/ferrocene couple (Fc+/Fc) and listed in Table 2.3.1.

The CV of the monoanions 2a and 3 [Cu(LTMS)2]1- feature two one-electron transfer

waves (according to coulometric measurements), which correspond to one reduction and one

oxidation process. The results are summarized in Table 2.3.1. An additional quasi-reversible

29

Chapter 2

oxidation at +0.73 V was observed for 4 [Au(LTMS)2]1- but it not featured in Figure 2.3.2. This

process was not studied in further detail due to its instability even at -25 °C.

Similar redox potential values have been observed for complexes containing the

(LBu)2-. Interestingly, the electrochemical analysis of the [Pd(LBu•)(LBu)]1- and [Pt(LBu•)(LBu)]1-

revealed very small differences (~15 mV) in the redox potentials, independent of the nature of

the central metal ion (Ni, Pd or Pt). This fact, coupled with the corresponding electronic

spectra, enabled the assignment of ligand-based redox processes shown in equation 2.3.1

where M = Ni, Pd, Pt.29

[MII(LBu•)2] [MII(LBu•)(LBu)]1- [MII(LBu)2]2- Eqn. 2.3.1 +e +e-e -e

It is important to note that for the [MIII(LTMS)2]1- species (M = Cu, Au) there is a large

separation in the reduction potential values (~ 0.8 V) moving from copper to the gold

compound, suggesting metal-based redox processes (Equation 2.3.2).

[MIII(LTMS)(LTMS•)] [MIII(LTMS)2]1- [MII(LTMS)2]2- Eqn. 2.3.2

+e

+e-e -e

The monoanion copper compound 3 was not studied in detail because the oxidation

process is electrochemically irreversible. According to coulometric measurements, compound

3 features a reversible one-electron reduction wave at -1.235 V corresponding to CuIII/CuII

redox couple.

30

Chapter 2

0.3 0.0 -0.3 -0.6 -0.9 -1.2 -1.5 -1.8

E (V) versus Fc+/Fc

200 mV.s-1

100 mV.s-1

50 mV.s-1

25 mV.s-1

5 μA

Figure 2.3.1 – Cyclic voltammogram of 2a [Ni(LTMS•)(LTMS)]1- recorded in CH2Cl2 solution

containing 0.10 M [N(n-Bu)4]PF6 as supporting electrolyte at scan rates of 200, 100, 50 and

25 mV s-1 at 25 °C. (Conditions: glassy carbon electrode; potentials referenced vs the

ferrocenium/ferrocene couple).

Figure 2.3.2 – Cyclic voltammogram of [MIII(LTMS)2]1- (M = Au 4 and Cu 3) species recorded

in CH2Cl2 solution containing 0.10 M [N(n-Bu)4]PF6 as supporting electrolyte at scan rates of

200, 100, 50 and 25 mV s-1 at 25 °C. (Conditions: glassy carbon electrode; potentials

referenced vs the ferrocinium/ferrocene couple).

4

3

0.5 0.0 -0.5 -1.0 -1.5 -2.0

5 μA

5 μA

E (V) versus Fc+/Fc

31

Chapter 2

Table 2.3.1 – Redox potentials of [M(LTMS)2]1- (M = Ni (2a), Cu (3), Au (4)) complexes in

CH2Cl2 solution containing 0.10 M [N(n-Bu)4]PF6 at 25 °C.

Complex E1½, V vs Fc+/Fc E2

½, V vs Fc+/Fc

2a -0.208 -1.094

3 +0.143 -1.235

4 +0.154 (irrev.) -2.050

Figure 2.3.3 displays the changes in the absorption spectra for the one-electron transfer series

of 2a; and, Figure 2.3.4 shows the UV-Vis spectrum of 4 and its one-electron oxidized

analogue 4a.

[Ni(LTMS) ]22- (2)

[Ni(LTMS•)(LTMS)]1- (2a)

[Ni(LTMS•)2]0 (2b)

300 450 600 750 900 10500.0

0.5

1.0

1.5

2.0

2.5

e, 1

04 M-1

cm

-1

λ, nm

Figure 2.3.3 – Electronic spectra of monoanionic nickel complex 2a, its electrochemically

generated one-electron oxidized and reduced forms in CH2Cl2 solutions containing 0.10 M

[N(n-Bu)4]PF6 at -25 °C.

The electronic spectrum of the monoanionic 2a [NiII(LTMS)(LTMS•)]1- complex in

CH2Cl2 solution shows an intense intervalence charge transfer (IVCT) band from ligand

(LTMS) to (LTMS•) in the near-IR region (896 nm).

32

Chapter 2

One-electron reduction of the monocationic species of 2a to 2 results in a drastic

reduction of this band, which was not driven to completeness due to the instability of the

dianionic species under the conditions used for cyclic voltammetry. The band at around 890

nm disappears completely since the reduction results in the formation of a second closed-shell

ligand and hence, the ligand-to-ligand charge transfer (LLCT) is no longer possible.45

400 450 500 550 600 650 700 7500

100

200

300

400

500

ε, M

-1cm

-1

λ, nm

4

Figure 2.3.4 – Electronic spectra of monoanionic gold compound 4 [Au(LTMS)2]1-, with

magnification (insert) and its electrochemically generated one-electron oxidized form 4a

[Au(LTMS)2]0 in CH2Cl2 solutions containing 0.10 M [N(n-Bu)4]PF6 at -25 °C.

The electronic spectrum of 4 displays two d–d transitions in the visible range at 641

nm (ε = 120 M-1 cm-1) and 420 nm (ε = 500 M-1 cm-1). No charge transfer transitions were

observed in the near-IR for this complex. Similar electronic spectra have been observed for

other diamagnetic square planar complexes of AuIII with d8 electronic configuration.11,41

Conversely, the UV-Vis spectrum of the electrochemically generated one-electron oxidized

analogue 4a displays a very intense absorption in the near-IR region at 1413 nm

(ε = 2.57 x 104 M-1 cm-1), along with rather weak maxima at 501 nm (ε = 4.9 x 103 M-1 cm-1)

and 1015 nm (ε = 2.0 x 103 M-1 cm-1). We tentatively assign the intense band at 1413 nm to an

intervalence transition of the type [AuIII(LBu)(LBu•)]0 ↔ [AuIII(LBu•)(LBu)]0 as suggested for

[Au(L)2]0 previously.46,47 These bands, as well as the one observed for complex

4a

400 600 800 1000 1200 1400 16000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

ε, 1

04 M-1

cm

-1

4

λ, nm

33

Chapter 2

[Ni(LTMS•)(LTMS)]1- 2a, are due to class III delocalisation based on the Robin-Day

classification.48 Notably, an intervalence transition in the near-IR region is absent in the

electronic spectrum of compound 3 (Figure 2.3.5), which is in good agreement with the

electronic description [CuIII(LTMS)2]1-. Three intense transitions at 246 nm

(ε = 5.2 x 104 M-1 cm-1), 352 nm (ε = 1.8 x 104 M-1 cm-1) and 405 nm (ε = 4.0 x 104 M-1 cm-1)

are observed. The band at 405 nm is assigned to the LMCT transition to the vacant dxy orbital

and the other two remaining bands in the UV region at 246 and 352 nm correspond to the

intraligand transitions, which are present in the free-ligand also.29 Table 2.3.2 summarizes the

features observed in the UV spectra of the presented complexes. More details about the

electronic structure will be discussed in the DFT calculation section (vide infra).

300 450 600 750 900 10500

1

2

3

4

5

ε, 10

4 M-1

cm

-1

λ, nm

Figure 2.3.5 – Electronic spectra of monoanionic copper compound 3, in CH2Cl2 solutions

containing 0.10 M [N(n-Bu)4]PF6 at -25 °C.

34

Chapter 2

Table 2.3.2 – Summary of the electronic spectra of the complexes at -25 °C in CH2Cl2

solutions.

Complex λ max., nm (ε, 104 M-1 cm-1)

2 289 (1.34), 316 (2.01), 446 (0.46)

2a 318 (1.85), 374 (0.64), 417 (sh. 0.26), 798 (sh. 0.35), 890 (0.96)

2b 311 (2.45), 347 (0.69), 559 (0.07), 859 (1.83)

3 352 (0.90), 405 (2.02), 613 (0.022)

4 420 (0.05), 631 (0.012)

4a 362 (3.78), 501 (0.49), 1015 (0.20), 1413 (2.57)

2.4 – Magnetic Properties:

According to SQUID measurements, complexes 2, 2b, 3 and 4 are diamagnetic and

have a singlet (S = 0) ground state configuration. In contrast, the compound

[N(n-Bu)4][Ni(LTMS•)(LTMS)]1- 2a is paramagnetic (Figure 2.4.1). Temperature dependent

(4 – 300 K) magnetic susceptibility measurements in an external field of 1.0 T, indicated a

temperature-independent (10 – 300 K) magnetic moment of 1.76 ± 0.01 µB, which is

consistent with the spin-only value expected for S = ½ systems.

0 50 100 150 200 250 3000.6

0.8

1.0

1.2

1.4

1.6

1.8

μ eff,

μB

Temperature, K

■ Experimental Calculated with: S = ½ g = 2.0 θ-Weiss: -0.636 K

Figure 2.4.1 – Temperature dependence of the effective magnetic moment of complex 2a

(4-300 K) measured with an applied field of 1.0 T.

35

Chapter 2

The doublet state of compound 2a was confirmed by the X-band EPR spectrum in

CH2Cl2 solution at 30 K shown in Figure 2.4.2. The EPR spectrum of the nickel compound 2a

(Figure 2.14) shows a rhombic signal with a large g anisotropy with g1 = 2.18, g2 = 2.04 and

g3 = 2.01.

300 320 340 360

-1.0

-0.5

0.0

0.5

1.0

1.5

2.3 2.2 2.1 2 1.9

dχ"

/ dB

B, mT

g values

Figure 2.4.2 – EPR spectrum of 2a in CH2Cl2 solution at 30 K. Conditions: frequency 9.430

GHz; modulation amplitude 10 G; power 100.6 µW. The parameters used for simulation are

g1 = 2.182; g2 = 2.045; g3 = 2.008; and isotropic line width W = 100 MHz. Black line

represents the experimental spectrum and the red corresponds to the simulation.

In the 1960s Holm et al.31 reported the EPR spectrum of the [Ni(mnt)2]1-

(mnt = maleonitriledithiolate). A detailed theoretical study prompted the authors to described

the nickel complex as NiIII (d7) because of the relatively high contributions of the metal to the

magnetic orbital. This description of a central NiIII ion was refuted a year later by Gray et al.49

who formulated the monoanionic complex as NiII with one π-ligand based radical.

Some years later Kirmse et al.50 analysed the single crystal EPR spectrum of 61Ni

enriched species [N(n-Bu)4][Ni(ortho-xylenedithiolate)2], which showed similar g values as

those observed for complex 2a. In addition to very strong hyperfine coupling of the unpaired

electron to the 61Ni nucleus (I = 3/2). The calculation of the spin-Hamiltonian parameters

36

Chapter 2

using the extended Hückel molecular orbital (MO) theory revealed that the spin density is

largely delocalized over the ligand atoms with most of the unpaired spin density localized on

sulfur. From the experimental and theoretical data, they concluded that for these complexes,

the SOMO has a b2g symmetry which is mainly a linear combination of the metal dxy orbital

(~30% metal character) and the out-of-plane 3px orbitals of the four sulfur atoms. Recently,

Wieghardt et al.29 interpreted the EPR spectrum of the nickel complex containing the LBu

ligand not in terms of NiIII (d7) ion containing two closed-shell ligands, but as NiII coordinated

to a ortho-dithiobenzosemiquinonate(1-) ligand radical with metal contributions of ~30% to

the magnetic orbital. This statement was made based on the combination of spectroscopic

methods (including sulfur K-edge X-ray absorption spectroscopy)5 and scalar relativistic

ZORA B3LYP DFT calculations, which will be described further in this chapter.

Figure 2.4.3 displays the EPR spectrum of the electrochemically generated compound

4a [Au(LTMS)2]0 in CH2Cl2 solution at 10 K, and is in agreement with a S = ½ ground state.

Unfortunately attempts to trap the chemically oxidized 4a were not successful due to the

instability of the compound. A rhombic signal with g1 = 2.071, g2 = 2.033 and g3 = 1.910 (giso

= 2.005) was observed without any detectable hyperfine splitting to the 197Au nucleus (I =

3/2, 100% natural abundance). This observation rules out the possibility of AuIV (SAu = ½) ion

with a low spin d7 electron configuration.

300 310 320 330 340 350 360 370

2.2 2.15 2.1 2.05 2 1.95 1.9 1.85

dχ"

/ dB

B, mT

g values

Figure 2.4.3 – X-band EPR spectrum of the electrochemically generated 4a [Au(LTMS)2]0 in

CH2Cl2 solution at 10 K. Conditions: frequency 9.4476 GHz; modulation 10 G; power 99.85

µW. For simulation parameters (g1 = 2.071; g2 = 2.033; g3 = 1.910; line widths Wx = 20.0;

Wy = 20.0; Wz = 40.0 MHz). Black line represents the experimental spectrum and the red

corresponds to the simulation.

37

Chapter 2

To our knowledge, no X-band EPR spectrum has been reported for an authentic AuIV

complex. However, one could expect to observe similar features to mononuclear AuII species

(d9, SAu = ½), in as much as there should be a large 197Au hyperfine splitting. X-band EPR

spectra of gold(II) complexes, such as [AuII(mnt)2]2-, [AuII(dialkyldithiocarbamato)2]0 and

[AuII([9]-aneS3)2](BF4)2, for example, have been reported.51-54 These spectra show large 197Au

hyperfine coupling and large g anisotropy due to spin-orbit coupling. Both of these factors

indicate a significant spin density at the Au center. The complexity of the EPR spectra

reported also arises from the large electric quadrupole (Q = 54.7 fm2) interactions of the Au

ion with a non-zero electric field gradient generated by the ligand field.54 More recently, a

stable, monomeric Au(II) complex has been reported with hematoporphyrin, in which the six-

coordinate Au center is bound to the hematoporphyrin macrocycle and two water molecules.

The EPR spectrum of this complex is dominated by an intense signal attributed to a stable

free-radical; a less intense signal containing nine lines due to the interaction of the unpaired

electron with the four coordinated N atoms is also observed.55

Although no hyperfine coupling was previously observed for neutral dithiolate gold

complexes, Wieghardt et al.56 reported a remarkable EPR spectrum of the [AuIII(LPh)2]

compound ((LPh)2- = 1,2-di(4-tert-butylphenyl)ethylene-1,2-dithiolate). The unusual X-band

EPR spectrum shows hyperfine splitting by the 197Au nucleus, which deviates from the

“normal” appearance of the most prominent multiplets due to the unusual spacings and

intensity distribution of the hyperfine lines. The simulation could be carried out only by

considering the mixing between magnetic and electric hyperfine interactions. DFT

calculations depicted a SOMO with less than 10% metal character. Thus, the complex

containing the ethylenedithiolate ligand can be described as Au(III) with one sulfur π-ligand

radical.56

Relativistic DFT calculations performed on the [Au(L)2]0 complex showed a metal

contribution of only 8% to the SOMO which is in agreement with the EPR spectrum.29 Thus

the EPR spectrum of our compound 4a is in accord with the assigned AuIII center with one

ortho-dithiosemiquinonate(1-) ligand radical. The isoelectronic Pd complex

[PdII(bpy)(LBu•)]1+ (g1 = 2.02, g2 = 2.01, g3 = 1.99, giso = 2.01) shows similar spectral features,

including no 105Pd (I = 5/2, 22.2%) hyperfine coupling.57

38

Chapter 2

2.5 – Theoretical Calculations:

The DFT calculations were carried out at the B3LYP level for the 2a [Ni(LTMS)2]1-

species. Its one-electron reduced and oxidized counterparts, as well as the gold and copper

complexes with (LBu)2- ligands have been previously reported.29 The DFT and spectroscopic

results obtained for the nickel system containing the unsubstituted (L)2- ligands are

qualitatively very similar to 2a. Thus, the results of the DFT calculations previously published

are incorporated to provide greater insight into the electronic structure of the compounds

under study in this chapter.

Structure Optimization:

The optimized geometry calculation of 2a is in good agreement with the experimental

data obtained by X-ray crystallography (Table 2.5.1). The small overestimation of the M–S

bond distances is typical for the B3LYP functionals.58-61 However, the intraligand distances

were accurately reproduced by the calculations within 0.02 Å.

Table 2.5.1 – Experimental and calculated (in parentheses) bond distance (Å).

SM

S

S

S1

23

4

56

1

23

4

56

Complex M–S C–S C1–C2 C2–C3 C3–C4 C4–C5 C5–C6 C6–C1

[Ni(L)2]2- a 2.173 (2.210)

1.762 (1.770)

1.426 (1.407)

1.398 (1.399)

1.390 (1.402)

1.383 (1.399)

1.402 (1.407)

1.401 (1.405)

[Ni(LTMS)2]1- b

2.150

(2.190) 1.750

(1.765) 1.421

(1.425) 1.395

(1.402) 1.399

(1.403) 1.395

(1.402) 1.422

(1.425) 1.414

(1.421)

[Ni(L)2]0 a 2.126 (2.158)

1.727 (1.744)

1.429 (1.412)

1.375 (1.382)

1.422 (1.412)

1.373 (1.382)

1.409 (1.412)

1.419 (1.424)

[Cu(L)2]1- a 2.168

(2.217) 1.768

(1.774) 1.419

(1.408) 1.404

(1.398) 1.405

(1.402) 1.366

(1.390) 1.415

(1.405) 1.394

(1.402)

[Au(L)2]1- a 2.310 (2.338)

1.764 (1.776)

1.397 (1.404)

1.386 (1.393)

1.392 (1.400)

1.382 (1.393)

1.402 (1.404)

1.397 (1.406)

[Au(L)2]0 a 2.300

(2.317) 1.735

(1.758) 1.402

(1.408) 1.374

(1.385) 1.415

(1.406) 1.384

(1.385) 1.420

(1.408) 1.406

(1.411) a Ref.29. b This work.

39

Chapter 2

Bonding Scheme and Ground State Properties:

A qualitative bonding scheme derived for 2a [NiII(LTMS•)(LTMS)]1- species is shown in

Figure 2.5.1, wherein the spin up and the spin down MOs are shown in order of decreasing

energy. The ground state electronic configuration of the [Ni(LTMS•)(LTMS)]1- and

[Au(LTMS•)(LTMS)]0 is predicted to be:

(1ag)2(1b3g)2(2ag)2(1b2g)2(1au)2(2b3g)2(1b1u)2(2b2g)1(1b1g)0

The calculated 2BB2g ground state concurs with the results from the extended Hückel

calculations on the [Ni(L)2] . The bonding scheme in Figure 2.5.1 identifies four metal d-

orbitals lower in energy relative to the ligand-based orbitals, similar to that observed by

Solomon et al. for [Ni(mnt)

1- 29

622] . These four doubly occupied orbitals, namely 1a1-

g (dx2-y

2),

1b3g (dyz), 2ag (dz2) and 1b2g (dxz) are predominantly metal-d in origin (over 70% metal

character); such that the valence states of the metals are best described as Ni and Au (d )

ions. The LUMO for both these complexes is the σ-antibonding combination of metal d

II III 8

xy and

the ligand 1b1u orbitals. Due to the square-planar geometry, the overlap between these two

orbitals is favourable, providing an effective pathway for ligand-to-metal σ-electron donation,

forming a highly covalent σ-bond. The SOMO comprises mainly the π* b

62

2g MO of the free-

ligand, which undergoes mixing with the low lying metal dxz orbital, conferring some metal

character to these orbitals. This interaction accounts for the metal-to-ligand π-electron

donation in these compounds. 29

The bonding scheme reported29 for [Au(L•)(L)]0 shows a considerable energy gap

(~4.3 eV) between the four metal d-orbitals and those of the ligand due to the higher effective

nuclear charge of the gold ion and its higher ionic charge +III. This feature is not observed for

the calculated [Ni(LTMS•)(LTMS)]1- species in Figure 2.5.1 which is consistent for a first row

transition metal in the +II oxidation state.

40

Chapter 2

Figure 2.5.1 – Unrestricted Kohn-Sham MOs and energy scheme of [Ni(LTMS•)(LTMS)]1- 2a

from B3LYP DFT calculations.

Spin up Spin down

1b1g

2b2g

1b1u

2b3g

1au

1b2g

2ag

1b3g

1ag-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

Ener

gy,e

V

1ag

1b3g

2ag

1b2g

1au

2b3g

1b1u

2b2g

1b1g 1b1g

2b2g

1b1u

2b3g

1au

1b2g

2ag

1b3g

1ag-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

S

S

SNi

S

TMS

TMSTMS

TMS

Y

X

1ag

1b3g

2ag

1b2g

1au

2b3g

1b1u

2b2g

1b1g

Ener

gy,e

V

41

Chapter 2

Table 2.5.1 – Percentage composition of selected molecular orbitals of [M(L´)2]z complexes

(M = Ni, Cu, Au; L´ = L, LTMS) obtained from B3LYP DFT calculations.

Complex MO M(ndyz) M(ndxz) M(ndxy) S(3pz) S(3px, y) C(2pz) C(2px, y)

a [Ni(L)2]2- 2b3g 71 17 8 2b2g 52 36 10 1b1g

50 40 2

b [Ni(LTMS )2]1- 2b3g 26 (55) 49 (30) 22 (8) 2b2g 38 (34) 43 (48) 16 (10) 1b1g

45 (50) 41 (43) 2 (2)

a [Ni(L)2]0 2b3g 49 34 12 2b2g 26 51 16 1b1g

43 48 3

a [Cu(L)2]1- 2b3g 17 53 30 2b2g 11 58 27 1b1g

33 58 4

a [Au(L)2]1- 2b3g 11 58 30 2b2g 9 64 26 1b1g

33 55 5

a [Au(L)2]0 2b3g 9 51 38 2b2g 8 61 30 1b1g 33 55 5

a Ref.29. b This work.

Spectroscopic Trends Based on DFT Calculations

Based on Figure 2.5.1 and Table 2.5.1, a more comprehensive understanding of the

electronic structures of the presented compounds can be obtained. In the case of 4a

[Au(LTMS•)(LTMS)]0, an intense and broad band at 1400 nm is observed (Figure 2.5.2). The

monoanion 2a also shows an intense transition at 890 nm and a shoulder at 798 nm. These

low-energy IVCT bands of 2a and 4a can be characterized as 1b1u → 2b2g and 1au → 2b2g

transitions and are spin and electric dipole allowed.29

The electronic spectrum of the compound 2b [NiII(LTMS•)2]0 generated by the

coulometric one-electron oxidation of the corresponding monoanion 2a, is dominated by the

spin- and dipole-allowed ligand-to-ligand charge transfer (LLCT) bands observed in the range

of 800-900 nm region. Interestingly, the LLCT band in the neutral complex is found to be

twice as intense as the most prominent IVCT band of 2a. Thus, the position and intensity of

the IVCT bands are, highly dependent on the effective nuclear charge at the metal ion.45

42

Chapter 2

Another feature observed for 2b is the presence of charge transfer bands in the UV

region (300 – 400 nm), which are exhibited by a large number of complexes containing 1,2-

benzenedithiolate derivatives.63 Those bands have been assigned as ligand-to-metal charge

transfer (LMCT) into the vacant dxy orbital.29

The NiII, PdII, PtII, and AuIII complexes containing the LBu ligand exhibit the same

behaviour to that described above. The IVCT band in the complex [AuIII(LBu)(LBu•)]0 shows

the greatest red shift (1400 nm) while [NiII(LBu•)(LBu)]1- the highest energy. Figure 2.5.2

compares the positions and relative intensities of IVCT bands of the complexes 2a, and 4a.

The increasing blue shift in the IVCT band together with its reduced intensity, moving from

Au to Ni can be attributed to the increasing LMCT contribution to the actual IVCT band. This

means that position and the intensity of the IVCT band is highly dependent on the metal

contribution of the acceptor orbital 2b2g, which decreases from Ni to Au, based on

computational data. The metal contribution was studied extensively for the similar system

containing LBu, showing 34% nickel character for the SOMO, while only 8% gold

contribution was obtained according to DFT calculations.29,45

400 600 800 1000 1200 1400 16000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

ε, 1

04 Μ−1

cm

-1

λ, nm

[Ni(LTMS•)(LTMS)]1-

2a

[Au(LTMS•)(LTMS)]0

4a

Figure 2.5.2 – Overlap of absorption spectra of compounds 2a and 4a.

The difference in metal contribution to the magnetic orbital is reflected in the EPR

spectra. Interestingly, X-band EPR studies involving thiyl radicals (R-CH2-S•) in proteins and

low molecular weight thiols such as cysteine, revealed unusual spectroscopic features caused

by the large spin-orbit coupling constant (ζS = 382 cm-1) of the sulfur atom.64 This is

43

Chapter 2

significant larger than the corresponding values for O, N and C (151, 76 and 28 cm-1,

respectively).64 As a consequence, thiyl radicals exhibit a large g anisotropy and a fast

relaxation behaviour which leads to broad line widths in the EPR spectra. According to DFT

calculations on cysteine thiyl radical, the singly occupied orbital is an almost pure sulfur 3p

orbital.65 This orbital is near-degenerate with a second lone-pair orbital, which has almost a

pure sulfur-p character. The near-degeneracy causes exceptionally large gx values in the EPR

spectrum.66

The anisotropy (Δg) for complexes 2a and 4a are 0.170 and 0.161, respectively. The

difference is surprisingly small, given the considerably larger spin-orbit coupling constant for

Au (ζAuIII = 2140 cm-1, ζNi

II = 630 cm-1).64 Though a larger Δg should be observed for the gold

compound, this effect appears to be countered by a significantly lower Au contribution to the

ground state.

44

Chapter 2

2.6 – Conclusions

The electronic structures of the monoanionic complexes [M(LTMS)2]z (M = Ni, Au, Cu;

z = 0, 1-, 2-) have been established. In the case of the nickel, each member of the one-electron

transfer series is shown to possess a common oxidation state of +II for the metal center.

Oxidation processes provoke a systematic shortening of the C–S bonds and quinoid-like

distortions of the phenyl rings, which were reproduced in the DFT calculations. The electronic

spectrum of the monoanionic nickel species displays IVCT bands at 800-900 nm,

characteristic of ortho-dithiobenzosemiquinonate(1-) radical anions coordinated to the metal

ion. These bands are considerably more intense for the neutral compound and non existent for

the dianionic species. Such bands are not present in the spectra of the isoelectronic gold and

copper compounds, which are formulated as CuIII and AuIII d8 coordinated to two closed-shell

(LTMS)2- ligands. The EPR spectra of 2a [Ni(LTMS•)(LTMS)]1- and 4a [Au(LTMS•)(LTMS)]0 are

interpreted as a sulfur radical spectrum with large g anisotropy resulting from spin-orbit

coupling of the sulfur.

Analogous to the complexes containing LBu, the radical character (or percentage S 3p

character in the b2g SOMO) decreases in the order [AuIII(LTMS•)(LTMS)]0 > [CuIII(LTMS)2]1- >

[NiII(LTMS•)(LTMS)]1-. This is attributed to variations in the d orbital energies relative to the

ligand, which is dependent on the effective nuclear charge and the relativistic potential that is

mainly influenced by the energies of the d-shell at the central metal ion. The near-IR band in

[Cu(LTMS)2]1- is absent because the acceptor dxz orbital is filled for CuIII ion (d8 configuration)

and the only possible acceptor orbital at the Cu is the σ-antibonding dxy based b1g orbital,

which is energetically inaccessible. Thus, despite the high effective nuclear charge of the

central CuIII, oxidation of the ligand does not occur.

The results from this chapter underscore how effective the combination of

spectroscopy and theoretical calculations are at defining the physical oxidation states of metal

and ligands.

45

Chapter 2

2.7 – References:

1 Eisenberg, R.; Ibers, J. A. Inorg. Chem. 1965, 4, 605-8. 2 Eisenberg, R.; Ibers, J. A.; Clark, R. J. H.; Gray, H. B. J. Am. Chem. Soc. 1964, 86, 113-15. 3 Schrauzer, G. N.; Mayweg, V. J. Am. Chem. Soc.1962, 84, 3221. 4 Schrauzer, G. N.; Mayweg, V. P. 1965, 87, 3585-92. 5 Ray, K.; George, S. D.; Solomon, E. I.; Wieghardt, K.; Neese, F. Chem. Eur. J. 2007, 13, 2783-2797. 6 Sellmann, D.; Binder, H.; Haussinger, D.; Heinemann, F. W.; Sutter, J. Inorg. Chim. Acta 2000, 300-302, 829-836. 7 Rindorf, G.; Thorup, N.; Bjoernholm, T.; Bechgaard, K. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1990, C46, 1437-9. 8 Cerrada, E.; Fernandez, E. J.; Gimeno, M. C.; Laguna, A.; Laguna, M.; Terroba, R.; Villacampa, M. D. J. Organomet. Chem. 1995, 492, 105-10. 9 Davila, R. M.; Staples, R. J.; Fackler, J. P., Jr. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1994, C50, 1898-900. 10 Gimeno, M. C.; Jones, P. G.; Laguna, A.; Laguna, M.; Terroba, R. Inorg. Chem. 1994, 33, 3932-8. 11 Mazid, M. A.; Razi, M. T.; Sadler, P. J. Inorg. Chem. 1981, 20, 2872-7. 12 Nakamoto, M.; Koijman, H.; Paul, M.; Hiller, W.; Schmidbauer, H. Z. Anorg. Allg. Chem. 1993, 619, 1341-6. 13 Sawyer, D. T.; Srivatsa, G. S.; Bodini, M. E.; Schaefer, W. P.; Wing, R. M. J. Am. Chem. Soc. 1986, 108, 936-42. 14 Sato, R.; Ohyama, T.; Kawagoe, T.; Baba, M.; Nakajo, S.; Kimura, T.; Ogawa, S. Heterocycles 2001, 55, 145-154. 15 Smith, K.; Lindsay, C. M.; Pritchard, G. J. J. Am. Chem. Soc. 1989, 111, 665-9. 16 Block, E.; Eswarakrishnan, V.; Gernon, M.; Ofori-Okai, G.; Saha, C.; Tang, K.; Zubieta, J. J. Am. Chem. Soc. 1989, 111, 658-65. 17 Figuly, G. D.; Loop, C. K.; Martin, J. C. J. Am. Chem. Soc.1989, 111, 654-8. 18 Figuly, G. D.; Martin, J. C. J. Org. Chem. 1980, 45, 3728-9.

46

Chapter 2

19 Huynh, H. V.; Seidel, W. W.; Lugger, T.; Frohlich, R.; Wibbeling, B.; Hahn, F. E. Z. Naturforsch., B: Chem. Sci. 2002, 57, 1401-1408. 20 Ogawa, S.; Yomoji, N.; Chida, S.; Sato, R. Chem. Lett. 1994, 507-10. 21 Dirk, C. W.; Cox, S. D.; Wellman, D. E.; Wudl, F. J. Org. Chem. 1985, 50, 2395-7. 22 Goerdeler, J.; Kandler, J. Chemische Berichte 1959, 92, 1679-94. 23 McKinnon, D. M.; Lee, K. R. Can. J. Chem. 1988, 66, 1405-9. 24 Nacci, V.; Garofalo, A.; Fiorini, I. J. Heter. Chem. 1986, 23, 769-73. 25 Aromdee, C.; Cole, E. R.; Crank, G. Austr. J. Chem. 1983, 36, 2499-509. 26 Crank, G.; Makin, M. I. H. Austr. J. Chem. 1984, 37, 2331-7. 27 Sellmann, D.; Seubert, B.; Kern, W.; Knoch, F.; Moll, M. Z. Naturforsch., B: Chemical Scieces 1991, 46, 1435-48. 28 Herebian, D.; Bothe, E.; Bill, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 10012-10023. 29 Ray, K.; Weyhermueller, T.; Neese, F.; Wieghardt, K. Inorg. Chem. 2005, 44, 5345- 5360. 30 Huyett, J. E.; Choudhury, S. B.; Eichhorn, D. M.; Bryngelson, P. A.; Maroney, M. J.; Hoffman, B. M. Inorg. Chem. 1998, 37, 1361-1367. 31 Maki, A. H.; Edelstein, N.; Davison, A.; Holm, R. H. J. Am. Chem. Soc. 1964, 86, 4580-7. 32 Simao, D.; Alves, H.; Belo, D.; Rabaca, S.; Lopes, E. B.; Santos, I. C.; Gama, V.; Duarte, M. T.; Henriques, R. T.; Novais, H.; Almeida, M. Eur J. Inorg. Chem. 2001, 3119-3126. 33 Chun, H.; Weyhermuller, T.; Bill, E.; Wieghardt, K. Angew. Chem., Int.l Ed. 2001, 40, 2489-2492. 34 Verani, C. N.; Gallert, S.; Bill, E.; Weyhermuller, T.; Wieghardt, K.; Chaudhuri, P. Chem. Comm. 1999, 1747-1748. 35 Ghosh, P.; Begum, A.; Herebian, D.; Bothe, E.; Hildenbrand, K.; Weyhermuller, T.; Wieghardt, K. Angew. Chem., Int. Ed. 2003, 42, 563-567. 36 Herebian, D.; Bothe, E.; Neese, F.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2003, 125, 9116-9128. 37 Chaudhuri, P.; Verani, C. N.; Bill, E.; Bothe, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 2213-2223.

47

Chapter 2

38 Chun, H.; Verani, C. N.; Chaudhuri, P.; Bothe, E.; Bill, E.; Weyhermuller, T.; Wieghardt, K. Inorg. Chem. 2001, 40, 4157-66. 39 Lim, B. S.; Fomitchev, D. V.; Holm, R. H. Inorg. Chem. 2001, 40, 4257-4262. 40 Pierpont, C. G.; Lange, C. W. Prog. Inorg. Chem. 1994, 41, 331-442. 41 Ray, K.; Weyhermueller, T.; Goossens, A.; Craje, M. W. J.; Wieghardt, K. Inorg. Chem. 2003, 42, 4082-4087. 42 Mrkvova, K.; Kameni, J.; Sindela, Z.; Kvitek, L. Trans. Met. Chem., Marcel Dekkes, New York 2004, 29, 238. 43 Schioedt, N. C.; Bjoernholm, T.; Bechgaard, K.; Neumeier, J. J.; Allgeier, C.; Jacobsen, C. S.; Thorup, N. Phys. Rev. B: Condens. Matter 1996, 53, 1773-8. 44 Schmidbaur, H.; Wohlleben, A.; Schubert, U.; Frank, A.; Huttner, G. Chem. Ber. 1977, 110, 2751-7. 45 Petrenko, T.; Ray, K.; Wieghardt, K. E.; Neese, F. J. Am. Chem. Soc. 2006, 128, 4422- 4436. 46 Ray, K.; Begum, A.; Weyhermueller, T.; Piligkos, S.; Van Slageren, J.; Neese, F.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 4403-4415. 47 Ray, K.; Bill, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 5641-5654. 48 Robin, M. B.; Day, P. Advan. Inorg. Chem. Radiochem. 1967, 10, 247-422. 49 Shupack, S. I.; Billig, E.; Clark, R. J. H.; Williams, R.; Gray, H. B. J. Am. Chem. Soc. 1964, 86, 4594-602. 50 Kirmse, R.; Stach, J.; Dietzsch, W.; Steimecke, G.; Hoyer, E. Inorg. Chem. 1980, 19, 2679-85. 51 Ihlo, L.; Stosser, R.; Hofbauer, W.; Bottcher, R.; Kirmse, R. Z. Naturforsch., B: Chem. Sci. 1999, 54, 597-602. 52 Schlupp, R. L.; Maki, A. H. Inorg. Chem. 1974, 13, 44-51. 53 Ihlo, L.; Kampf, M.; Bottcher, R.; Kirmse, R. Z. Naturforsch., B: Chem. Sci. 2002, 57, 171-176. 54 Shaw, J. L.; Wolowska, J.; Collison, D.; Howard, J. A. K.; McInnes, E. J. L.; McMaster, J.; Blake, A. J.; Wilson, C.; Schroeder, M. J. Am. Chem. Soc. 2006, 128, 13827-13839. 55 Gencheva, G.; Tsekova, D.; Gochev, G.; Mehandjiev, D.; Bontchev, P. R. Inorg. Chem. Commun. 2003, 6, 325-328.

48

Chapter 2

56 Kokatam, S.; Ray, K.; Pap, J.; Bill, E.; Geiger, W. E.; LeSuer, R. J.; Rieger, P. H.; Weyhermueller, T.; Neese, F.; Wieghardt, K. Inorg. Chem. 2007, 46, 1100-1111. 57 Ghosh, P.; Begum, A.; Herebian, D.; Bothe, E.; Hildenbrand, K.; Weyhermuller, T.; Wieghardt, K. Angew. Chem., Int. Ed. 2003, 42, 563-567. 58 Becke, A. D. J. Chem. Phys. 1986, 84, 4524-9. 59 Becke, A. D. J. Chem. Phys. 1993, 98, 5648-52. 60 Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter 1988, 37, 785-9. 61 Neese, F. Orca, an Ab Initio, DFT and Semi empirical Electronic Structure Package, version 2.6-35, 2006. 62 Szilagyi, R. K.; Lim, B. S.; Glaser, T.; Holm, R. H.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. J. Am. Chem. Soc. 2003, 125, 9158-9169. 63 Beswick, C. L.; Schulman, J. M.; Stiefel, E. I. Prog. Inorg. Chem. 2003, 52, 55-110. 64 Mabbs, F. E.; Collison, D. Electron Spin Reson. 1994, 14, 88-129. 65 Lassmann, G.; Kolberg, M.; Bleifuss, G.; Graeslund, A.; Sjoeberg, B.-M.; Lubitz, W. Phys. Chem. Chem. Phys. 2003, 5, 2442-2453. 66 Van Gastel, M.; Lubitz, W.; Lassmann, G.; Neese, F. J. Am. Chem. Soc. 2004, 126, 2237-2246.

49

Chapter 3

Chapter 3

Dimerization Processes of Square Planar

[PtII(tbpy)(dithiolate•)]+ Radicals

50

Chapter 3

51

Chapter 3

3.1 Introduction

Square planar d8 complexes with dithiolate and diimine ligands are of interest due to

their unique charge transfer excited states.1,2 Many such complexes have been prepared by

using diimine-type ligands including 2,2´-bipyridine,3-5 biacetylbis(anil),6 1-10-

phenanthroline7 (and its derivatives1,2,5,8-11) and dithiolate ligands such as 1,2-

benzenedithiolate(2-)3,4,6,8,9,12,13 and 1,1-dithiocarbamate(1-),10,11,14 heterocyclic ligands with a

dithiolate chelating functional group,1,2,15-17 and 1,2-ethylenedithiolate(2-) derivatives.2,6,8,14

These species show an intense solvatochromic absorption band in the visible region, with

molar extinction coefficients (ε) of 0.5 - 1.0 x 104 M-1 cm-1, which shifts to higher energy with

increasing solvent polarity. A classical example of a complex where this band is observed is

[Pt(bpy•)(dithiolate)2]1- (bpy = bipyridine; (bpy•)1- is the corresponding radical anion), for

which this transition has been assigned as charge-transfer from a filled orbital of mixed Pt d

and dithiolate p composition (HOMO) to the LUMO of predominantly π* diimine

character.18,19 Extended Hückel15,20 and DFT calculations13 indicate that the HOMO in the

neutral complex is predominantly sulfur based with ca. 14% Pt 5d contribution.

[PtII(diimine)(dithiolato)]0 complexes are typical examples of compounds that show

unique photophysical properties.1,2 The absorption of visible light leads to a triplet excited

state, in which one electron from the dithiolate-based HOMO is excited to the π* orbital of the

diimine ligand. The excitation energy can be released by photon emission, which causes

solution-luminescence of these species, or by self-quenching reactions, which occur through

excited/ground state Pt•••Pt interactions.21 Interestingly, the excited state can undergo

photoinduced electron-transfer with reductive or oxidative quenching molecules to give the

corresponding ground state [Pt(diimine•)(dithiolate)]1- or [Pt(diimine)(dithiolate•)]1+

respectively.1,2 Several complexes undergo photoinduced reactions with dioxygen, (where the

reaction results in the addition of oxygen to the sulfurs of the complexed dithiolate) as a

consequence of the electron transfer.3,22,23 The photoinduced electron-transfer properties of

[PtII(diimine)(dithiolate)] complexes are due to both the diimine and dithiolate ligands being

redox active. The redox activity of coordinated ortho-benzenedithiolate ligands (LTMS)2- and

(LBu)2- was described in chapter 2 for the homoleptic [M(LTMS)2]z compounds (M = NiII, CuIII,

AuIII; z = 1+ → 2-) and the equivalent complexes containing the LBu ligand.24-28

In this chapter we describe the synthesis and characterization of the mononuclear

neutral complexes [PdII(tbpy)(LTMS)]0 5, [PtII(tbpy)(LTMS)]0 6, and [PtII(tbpy)(LPh)]0 7

52

Chapter 3

(tbpy = 4,4´-di-tert-butylbipyridine; LPh = 1,2-bis(4-tert-butylphenyl)ethylene-1,2-

dithiolate(2-)) and their oxidized derivatives. All ligands used are depicted in Figure 3.3.1.

The one-electron oxidized form of [MII(diimine)(dithiolate•)]1+ (M = Pd or Pt)

containing the ortho-dithiobenzosemiquinonate(1-) ligand radical had never been isolated in

the solid state or structurally characterized. Only the EPR spectrum of an electrochemically

generated species, [PdII(bpy)(LBu•)]1+, had been reported.12 This species was proposed to

rapidly dimerize in solution affording a diamagnetic dimer. A similar radical species,

[PtII(dpphen)(LBu•)]1+ (dpphen = 4,7-diphenyl-1,10-phenantroline), has recently been

generated in solution and characterized by EPR spectroscopy.9

The electrochemical oxidation of species 5, 6 and 7 revealed that dimerization

processes do take place yielding dinuclear paramagnetic monocationic intermediates and

diamagnetic dicationic dimers. The complex [PtII2(tbpy)2(LPh•)2](PF6)2 7a, was isolated in

crystalline form. Its crystal structure provides the first example of a mixed-ligand,

dithiolate(1-) π-radical containing complex. Conversely, the square planar analogue

[PtII(PPh3)2(LPh)]0 8, remains mononuclear upon oxidation to yield the [PtII(PPh3)2(LPh•)]1+ 8a.

The neutral diamagnetic compound [PtII(LPh•)2] 9 was also prepared and structurally

characterized to establish the differences in the C–S and C–C bond lengths in systems

containing a closed-shell dianionic ligand (LPh)2- or an one-electron oxidized radical (LPh•)1-.

A list of the complexes studied and their oxidation products is presented in Table 3.1.1. The

complexes in bold have been characterized by single crystal X-ray diffraction, whereas the

others are generated electrochemically or inferred.

53

Chapter 3

S

S

Si

Si

S

S

Si

Si

- e

+ e N N

S

S

- e

+ e

S

S P

(LTMS)2- (LTMS•)1-

(LPh•)1-(LPh)2-

(tbpy)

(PPh3)

Figure 3.1.1 – Schematic representation of the ligands.

Table 3.1.1 – List of complexes.

5 [PdII(tbpy)(LTMS)] 7 [PtII(tbpy)(LPh)]

5a [PdII(tbpy)(LTMS•)]1+ 7a [PtII(tbpy)(LPh•)]1+

5b [PdII2(tbpy)2(LTMS)(LTMS•)]1+ 7b [PtII

2(tbpy)2(LPh)(LPh•)]1+

5c [PdII2(tbpy)2(LTMS•)2]2+ 7c [PtII

2(tbpy)2(LPh•)2]2+

7d [PtII2(tbpy)2(LPh)2]

6 [PtII(tbpy)(LTMS)]

6a [PtII(tbpy)(LTMS•)]1+ 8 [PtII(PPh3)2(LPh)]

6b [PtII2(tbpy)2(LTMS)(LTMS•)]1+ 8a [PtII(PPh3)2(LPh•)]1+

6c [PtII2(tbpy)2(LTMS•)2]2+

6d [PtII2(tbpy)2(LTMS)2] 9 [PtII(LPh•)2]

54

Chapter 3

Results and Discussions

3.2 – Synthesis and X-ray Crystal Structures:

Complex 5 [PdII(tbpy)(LTMS)] was prepared under argon by adding [Pd(tbpy)Cl2] to

one equivalent of the salt 1b K2(LTMS) in MeCN. The solution instantaneously changed from

yellow to purple. Crystals of 5 were obtained upon cooling the solution to 0 °C. Compound 6

[PtII(tbpy)(LTMS)] was synthesized and crystallized following the same procedure using

[Pt(tbpy)Cl2] as the metal source. The reaction of [Pt(tbpy)Cl2] with the in situ generated

thiophosphoric ester20 of the ligand 1,2-bis(4-tert-butylphenyl)ethylene-1,2-dithiol [H2(LPh)]

in cold dioxane generated the blue crystalline complex 7 [PtII(tbpy)(LPh)]. A dimeric dication

is formed, and crystallized as a black hexafluorophosphate salt [PtII2(tbpy)2(LPh•)2](PF6)2 7c,

when a solution of 7 in CH2Cl2 was oxidised under argon using 1 equivalent of ferrocenium

hexafluorophosphate. The preparation of compounds 8 [Pt(PPh3)2(LPh)] and the neutral blue-

black complex 9 [Pt(LPh•)2] were performed as described in the literature.29,30 Crystals of 8

were obtained from a 1:1 CHCl3/n-hexane solution.

The crystal structures of 5 [PdII(tbpy)(LTMS)], 6 [PtII(tbpy)(LTMS)],

7c [PtII2(tbpy)2(LPh•)2](PF6)2•3CH2Cl2, 8 [Pt(PPh3)2(LPh)]0•3CHCl3, and 9 [PtII(LPh•)2]•toluene

have been determined at 100 (2) K. Selected bond distances and angles are listed in Table

3.2.1. The structures of complex 5 and 6, the dication of 7c, the neutral compounds 8•3CHCl3,

and 9•toluene are shown in Figures 3.2.1 to 3.2.5, respectively.

Figure 3.2.1 – Perspective view and numbering scheme of compound 5 [PdII(tbpy)(LTMS)]

with thermal ellipsoids at 50% probability level. Hydrogen atoms are omitted for clarity.

Pd(1)

S(1)

S(2)

N(1)

N(2)

C(1)

C(2)

C(3)

C(4)

C(5)

C(6)

Si(2)

Si(1)

55

Chapter 3

Pt(1)

S(1)

S(2)

N(1)

N(2)

C(1)

C(2)C(3)

C(4)

C(5)

C(6)

Si(2)

Si(1)

Figure 3.2.2 – Perspective view and numbering scheme of compound 6 [PtII(tbpy)(LTMS)]

with thermal ellipsoids at 50% probability level. Hydrogen atoms are omitted for clarity.

Pt(1)

Pt(1A)

S(1)

S(2)

S(2A)

S(1A)

C(1)

C(2)

2+

Figure 3.2.3 – Perspective view and numbering scheme of compound 7c [PtII2(tbpy)2(LPh•)2]2+

with thermal ellipsoids at 50% probability level. Hydrogen atoms are omitted for clarity.

56

Chapter 3

Pt(1)

S(1)

S(2)

P(1)

P(2)

C(1)

C(2)

Figure 3.2.4 – Perspective view and numbering scheme of compound 8 [Pt(PPh3)2(LPh)] with

thermal ellipsoids at 50% probability level. Hydrogen atoms are omitted for clarity.

Pt(1)

S(1)

S(2)

C(1)

C(2)

S(3)

S(4)

C(3)

C(4)

Figure 3.2.5 – Perspective view and numbering scheme of compound 9 [PtII(LPh•)2] with

thermal ellipsoids at 50% probability level. Hydrogen atoms are omitted for clarity.

57

Chapter 3

Table 3.2.1 – Selected bond lengths of the 5, 6, 7c, 8 and 9, distances are given in Å.

[PdII(tbpy)(LTMS)]

5

[PtII(tbpy)(LTMS)]

6

[PtII2(tbpy)2(LPh•)2]2+

7c

[Pt(PPh3)2(LPh)]

8 [PtII(LPh•)2]

9

M(1)-S(1) 2.2365(3) 2.2461(5) 2.2432(6) 2.3036(14) 2.243(2)

M(1)-S(2) 2.2397(3) 2.2430(5) 2.2304(6) 2.3034(12) 2.242(2)

M(1)-S(3) - - - - 2.259(2)

M(1)-S(4) - - - - 2.243(2)

M(1)-N(1) 2.0721(9) 2.0513(16) 2.0519(19) - -

M(2)-N(2) 2.0855(9) 2.0605(15) 2.0649(19) - -

M(1)-P(1) - - - 2.3004(13) -

M(1)-P(2) - - - 2.2829(14) -

Pt(1)-S(1A) - - 2.8619(6) - -

Pt(1A)-S(1) - - 2.8618(6) - -

S(1)-C(1) 1.7658(11) 1.760(2) 1.723(2) 1.764(5) 1.728(9)

S(2)-C(2) 1.7659(11) 1.766(2) 1.717(2) 1.771(6) 1.729(8)

S(3)-C(3) - - - - 1.709(10)

S(4)-C(4) - - - - 1.685(9)

C(1)-C(2) 1.4068(15) 1.411(3) 1.387(3) 1.354(8) 1.381(12)

C(2)-C(3) 1.4175(15) 1.414(3) - - -

C(3)-C(4) 1.3997(15) 1.398(3) - - 1.404(13)

C(4)-C(5) 1.3956(16) 1.395(3) - - -

C(5)-C(6) 1.4017(15) 1.400(3) - - -

C(6)-C(1) 1.4164(14) 1.418(3) - - -

The neutral complexes 5 [PdII(tbpy)(LTMS)] and 6 [PtII(tbpy)(LTMS)] are nearly square planar;

the dihedral angles between the planes S(1)–M(1)–S(2) and N(1)–M(1)–N(2) (M = Pd or Pt)

are 1.7° and 1.5° respectively. In both cases, the observed long C–S bond lengths are 1.76 Å

on average and the nearly equidistant C–C bonds of the dithiolate phenyl ring are typical for

closed-shell aromatic dithiolate dianions. Similar results have been reported for [MII(bpy)(L)]

complexes (M = Ni, Pd, Pt; L = ortho-benzenedithiolate)3 and [PtII(dpphen)(LBu)],9 (dpphen =

4,7-diphenyl-1,10-phenantroline). Compound 8 [Pt(PPh3)2(LPh)]0 is also square planar with

C–S bond lengths of 1.764(5) and 1.771(6) Å, corroborating the presence of a closed-shell

dianion (LPh)2-. The short C(1)–C(2) bond length of 1.354(8) Å is in agreement with a typical

value found for a C=C double bond.

A square planar geometry around the Pt center is also observed in the homoleptic

neutral compound 9 [Pt(LPh•)2]. In contrast to the complex 8 [Pt(PPh3)2(LPh)], the short

average C–S bond lengths of 1.713 Å and the relatively long average “olefinic” C–C bond

distance of 1.393 Å suggest a Pt center coordinated to two (LPh•) π-radical ligands, clearly

58

Chapter 3

illustrating the differences between a system containing a closed shell dianion (LPh)2- from

those containing a monoanionic radical (LPh•)1-.

The final structure isolated in this series is that of the dinuclear dication in 7c

[PtII2(tbpy)2(LPh•)2]2+ shown in Figure 3.2.3. In some cases, intermolecular interactions

between bis(dithiolate)metal units are sufficiently strong to form distinct, well-defined

molecular dimers. For these dimers, two distinct structural types are known: (I) a dimer

containing a M–M bond, for which only Pd31,32 and Pt31,33-35 complexes have been

characterized; (II) a centrosymmetric lateral M–S dimer with a M2(µ2-SR)2 rhomboid in

which the metal ion has an irregular five-coordinate geometry, typically involving first-row

transition metal ions such as Co,36-40 Fe,24,41-44 Mn,45 and Ni46. Figure 3.2.6 shows the two

possible structures for dimers and Table 3.2.2 compares specific bond distances for 7c with

examples reported in the literature for Pt dimers.

Table 3.2.2 – Comparison of selected averaged intermolecular bond lengths given in Å.

Complex Type of dimera Pt•••Pt Pt•••S b [Pt(edt)2]2 I 2.748 2.300

c [Pt(dmit)2]2•TTF I 2.935 2.315 II t7c [Pt 2( bpy)2(LPh•) ]2+ II 3.956 2.862 2

a See text and Figure 3.2.3. b • 1- edt = (S C H )4 4 4 , ethylene-1,2-dithiolate(1-) radical

monoanion31. c H (dmit) = 4,5-dimercapto-1,3-dithiol-2-thione, TTF = tetrathiafulvalene. 2

SM

S

S

S

SM

S

S

S

RR

RRR

R R

Figure 3.2.6 – Schematic structural representation of dimers type I and type II (see text).

It is quite revealing to compare the bond distances of the dithiolate moieties of the

approximately square planar [PtII(tbpy)(LPh•)]1+ building block in the dimer 7c with those of

R

SM

S

S

S

RRR

RS

MS

S

S

RRR

R

Type I Type II M–S dimer M–M dimer

59

Chapter 3

the neutral mononuclear unit in 8 [Pt(PPh3)2(LPh)]. The C–S and the “olefinic” C–C bond

lengths of the dithiolene ligand differ significantly. In the neutral species 8, a closed-shell

dianion (LPh 2-) is coordinated to a PtII II t center, whereas in the dimer 7c [Pt 2( bpy)2(LPh•) ]2+2 the

significantly shorter C–S bonds and the elongated C(1)–C(2) bond length indicate the

presence of a monoanionic π-radical (LPh• 1-) . The radical ligand has also been characterized by

X-ray crystallography in the mononuclear complex [PdII(LPh• 47) ];2 for which the C–S and

C(1)–C(2) bond lengths are identical within the experimental error (3σ) to those found here in

the dication, 7c [PtII t2( bpy)2(LPh•)2]2+. It is noteworthy that the geometry of the PtII t( bpy)

segment in 7c and in mononuclear 6 [PtII t( bpy)(LTMS)] are identical within experimental error

(3σ). The one-electron oxidation of 7 to 7c is therefore a ligand centered process, in which

(LPh)2- is oxidized to the radical (LPh• 1- II t) . The species 7a [Pt ( bpy)(LPh•)]1+ (S = ½) dimerizes

in solution with generation of the dication 7c. The [PtII t( bpy)(LPh•)]1+ moiety in the dimeric

dication is not planar. The N–Pt–N and S–Pt –S planes exhibit a dihedral angle of 16.5°.

As has been shown previously, the dithiolate monoanions are ligand centered radicals

where the spin density is predominantly localized in a 3p orbital of the sulfur atoms.24-27,48

The bridging Pt•••S bonds in the dication are very weak (average 2.862 Å) and may best be

described as a two-centered three-electron bond between a half filled 3p orbital at the sulfur

and a filled 5dz2 orbital at the Pt(II) center. The spins of the two singularly occupied molecular

orbitals (SOMOs) may then couple antiferromagnetically, yielding the observed diamagnetic

ground state.

3.3 – Sulfur K-edge X-ray Absorption Spectroscopy (XAS):

Figure 3.3.1 shows the normalized sulfur K-edge X-ray absorption spectral (XAS)

data for compounds 6 [PtII t( bpy)(LTMS)], 7 [PtII t( bpy)(LPh II t)], 7c [Pt 2( bpy)2(LPh•) ]2+2 , and 9

[Pt(LPh•)2] (measured as solids) and the monomeric paramagnetic species 7a

[PtII t( bpy)(LPh•)]1+ (measured in CH2Cl solution at 25 °C). 2

The pre-edge features (marked P and P1 2 in Figure 3.3.1) correspond to transitions to

unoccupied antibonding orbitals with sulfur 3p character.

60

Chapter 3

6 [PtII(tbpy)(LTMS)]

Figure 3.3.1 – Sulfur K-edge XAS spectra of 6 and 9 (top) and 7, 7a, and 7c (bottom).

The rising edge (marked E) corresponds to a sulfur 1s to 4p transition, the energy of

which reflects the effective nuclear charge on the sulfur. Complexes 6 and 7 exhibit an intense

single pre-edge feature at ~2472 eV. In the case of the singlet diradical compound 9

[Pt(LPh•)2], an additional pre-edge feature (P1) appears at lower energy (~2470 eV)

corresponding to a S 1s to 3p transition, which is consistent with the formation of a ligand-

based radical.25,49 This assignment is further supported by the increase in the rising edge

energy (E), which indicates that the sulfur is more oxidized. Similar changes are observed for

complex 7a, which indicates that a ligand-based oxidation has also occurred. The similarity

9 [PtII(LPh•)2]

7

7a [PtII(tbpy)(LPh•)]1+

7c [PtII

2(tbpy)2(LPh•)2]2+

[PtII(tbpy)(LPh)]

2466 2468 2470 2472 2474 24760.0

0.5

1.0

1.5

2.0Nor

m. A

bs.

Energy, eV

0.0

0.5

1.0

1.5

2.0

P1

P2

E

E

P2

P1

61

Chapter 3

between the dimer 7c [PtII t2( bpy)2(LPh•)2]2+ II t and the monomer 7a [Pt ( bpy)(LPh•)]1+ in CH Cl2 2

solution indicates that dimer formation does not significantly affect the sulfur radical

character indicating that the dimer involves a very weak Pt•••S interaction in accordance with

the crystal structure data (Table 3.2.1).

3.4 – Electro- and Spectroelectrochemistry:

II t The cyclic voltammograms of compounds 5 [Pd ( bpy)(LTMS II t)], 6 [Pt ( bpy)(LTMS)], 7

[PtII t( bpy)(LPh)], and 8 [Pt(PPh3)2(LPh)] (Figure 3.4.1) were measured in CH Cl2 2 solutions

with [N(n-Bu)4](PF6) as the supporting electrolyte. The redox potentials are showed in Table

3.4.1.

Complex 6 shows two successive one-electron redox waves: the first at –1.81 V (E1½)

is reversible whereas the second (E2 ) is split into two separate anodic (E2½ pa ~ +0.12 V and

E2´½ = +0.19 V) and cathodic peaks (E2´´

pc = –0.10 V and E2´pc = +0.11 V). Similar features

were observed for the Pd complex 5 [PdII t( bpy)(LTMS)], which shows two successive one-

electron redox waves: the first at –1.91 V (E1 ) is reversible, while the second (E2½ ½;

E2pa = +0.23 V) in this case is not split into two separate anodic peaks. Two cathodic peaks

were observed (E2´´pc = –0.08 V and E2´

pc = +0.14 V). In contrast, the cyclic voltammogram of

complex 7 [PtII t( bpy)(LPh)] exhibits two reversible one-electron transfer waves at –1.87 V

(E1½) and –0.10 V (E2 ) and, in addition, a quasi-reversible wave at +0.72 V (E3 ). ½ ½

+Table 3.4.1 – Spectroscopic data and redox potentials (E vs Fc /Fc). ½

a a-1Complex λ, nm (ε, M cm-1 -1) λ, nm (ε, M cm-1) E1½ E2 E3

½ ½

c5 345 (9700) 486 (6700) –1.91 – c6 342 (6000) 580 (8000) –1.81 –

7 347 (9300) 620 (7600) –1.87 –0.10 0.72

8 338 (6000) –0.01 0.69

[Pt(tbpy)(LMe)]b 563 (7200) –1.80 d–0.01 –

[Pt(tbpy)(mnt)] b 497 (5800) –1.67 0.54 – a In CH2Cl2 solution at 20 °C. b In dimethylformamide solution, ref.50. Abbreviations:

(LMe)2- = 3,4-toluenedithiolate, (mnt)2- = maleonitriledithiolate. c Peak potential (Figure

3.4.1): E d2pa = 0.12 V; E2´

pa = 0.19 V; E2´´pc = –0.10 V; E2´

pc = 0.11 V for 6. Irreversible; the

peak potential is given.

62

Chapter 3

The E1½ waves in the cyclic voltammograms of 5, 6, and 7 each correspond to a one-

electron reduction of the tbpy ligand generating a coordinated (tbpy• 1-) radical. This process

has been described for [PtII(bpy)Cl2]18 and other [PtII(bpy)(L)] complexes L = ortho-

benzenedithiolate(2), catecholate, ortho-aminophenolate and ortho- phenylenediamine).12,50

The E2 waves of 5, 6 and 7 are due to ligand-centered one-electron oxidation of the (LTMS½ )

and (LPh) ligands, yielding the coordinated radicals (LTMS•) and (LPh•) (Scheme 3.4.1).

[MII(tbpy•)(L)]1- [MII(tbpy)(L)]0 [MII(tbpy)(L•)]1+ -e +e

-e +e

E1½ E2

½

E3½ +e -e

[MII(tbpy)(Lox)]2+

MII = Pd or Pt

(L)2- = (LTMS)2-, (LPh)2-

(L•)1- = (LTMS•)1-, (LPh•)1-

(Lox) = (LPh-ox)0

Scheme 3.4.1 – Representation of redox processes in [Pt(tbpy)(L)] complexes.

This oxidation step (E2 II t½) yields the monocations 6a [Pt ( bpy)(LTMS•)]1+ and 7a

[PtII t( bpy)(LPh•)]1+, respectively. The fact that E2½ is split into two components indicates that

6a is not the sole product formed upon electrochemical one-electron oxidation of 6.

Compounds 6a [PtII t( bpy)(LTMS•)]1+ II tand 6 [Pt ( bpy)(LTMS)] dimerize in solution to form the

dinuclear species 6b [PtII t2 ( bpy)2(LTMS)(LTMS•)]1+, which undergoes a second one-electron

oxidation yielding 6c [PtII t2 ( bpy)2(LTMS•) ]2+. 2

The quasi-reversible oxidation wave E3 II t½ for 7 [Pt ( bpy)(LPh)] is probably a ligand-

centered process where the neutral 1,2-ethanedithione ligand, (LPh-ox)0, is produced. This

product is the least stable in solution and probably decomposes on the time scale of the cyclic

voltammogram measurement. The decomposition product resulted in new peaks at lower

potential (~ –1.2 and –1.6 V, see Figure 3.4.1). These low potential peaks are not observed

when the CV scans exclude E3½. The dithiolate ligand-based nature of the E2 and E3

½ ½ redox

waves is corroborated by the cyclic voltammogram of 8 [Pt(PPh3)2(LPh)] in Figure 3.4.1

which exhibits a quasi-reversible wave at –0.01 V (E2½), yielding [Pt(PPh3)2(LPh•)]1+ 8a and

an irreversible oxidation peak at +0.69 V (E3pa). These values are very similar to those

63

Chapter 3

II tobserved for 7 [Pt ( bpy)(LPh )], although the spectator ligand (tbpy) in 7 is replaced by two

triphenylphosphines in 8.

5µ A

5µ A

5µ A

5µ A

1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 -2.0 -2.5

5 [PdII(tbpy)(LTMS)]

6 [PtII(tbpy)(LTMS)]

8 [PtII(PPh3)2(LPh)]

7 [PtII(tbpy)(LPh)]

E (V) versus Fc+/Fc

Figure 3.4.1 – Cyclic voltammograms of compounds 5–8 in CH 0.10 M [N(n-Bu)Cl2 2 4](PF6),

glassy carbon working electrode, 20 °C, scan rate 200 mV s-1.

64

Chapter 3

The coulometric one-electron oxidations of 6–8 in CH with 0.10 M [N(n-Bu)Cl ](PF2 2 4 6) as

the supporting electrolyte have been followed spectroelectrochemically at –25 °C. The results

are shown in Figure 3.4.2. In each case three electronic spectra were recorded: (1) the

spectrum of the starting complexes (2) the spectrum after 50% of the one-electron oxidation

was completed and (3) the spectrum after complete oxidation.

400 500 600 700 800 900 1000 1100

Abs

orba

nce

λ, nm

7 7 + 7a + 7b + 7c

7b + 7c

400 500 600 700 800 900 1000 1100

Abs

orba

nce

λ, nm

5

5b

5c

400 500 600 700 800 900 1000

Abs

orba

nce

λ, nm

6

6b

6c

400 500 600 700 800 900 1000

λ , nm

Abs

orba

nce

8

8a

Figure 3.4.2 – Electronic spectra of compounds 5–8 at –25 °C in CH Cl2 2 containing 0.10 M

[N(n-Bu) ](PF4 6) as supporting electrolyte, glassy carbon as the working electrode. Black lines

represent the spectra of the initial complexes. The red lines represent the spectra after 50%

one-electron oxidation and blue lines after 100% oxidation.

Complex 8 [Pt(PPh3)2(LPh)] shows a simple transformation 8 → 8a [Pt(PPh3)2(LPh•)]1+

as indicated by the observation of a single, stable isosbestic point at 385 nm. In the visible

65

Chapter 3

range, the spectrum after “half” oxidation is composed of 50% of the fully oxidized form 8a,

[Pt(PPh3)2(LPh•)]1+, since 8 does not absorb strongly in the visible region. The fully oxidized

species 8a is paramagnetic and shows an S = ½ X-band EPR signal (100%) (vide infra). In

that manner, 8a does not dimerize in solution. Additionally, no indication of dinuclear

intermediates such as [PtII2(PPh3)4(LPh)(LPh•)]1+ were found.

II t The oxidations of 5 [Pd ( bpy)(LTMS II t)] and 6 [Pt ( bpy)(LTMS)] in CH Cl2 2 solution were

much more complex (cyclic voltammograms in Figure 3.4.1). The spectrum recorded after

50% oxidation exhibits four new absorption maxima at 343, 378, 505 and 1014 nm for

compound 5 and at 410, 520, 678, and 861 nm for complex 6 (red lines in Figure 3.4.2 top).

The maxima of 5 and 6 at 505 and 580 nm respectively, decreased in their intensities upon

50% oxidation. On the completion of the one-electron oxidation process (blue lines, Figure

3.4.2 top), a significant hypsochromic shift of the two isosbestic points is observed for

complex 6. After the first 50% oxidation, the isosbestic points at 492 and 638 nm shift to 477

and 605 nm during completion of the oxidation (50–100%), indicating clearly that one-

electron oxidation of 6 [PtII t( bpy)(LTMS)] involves an intermediate. This intermediate is

paramagnetic and, from simulations of its EPR spectrum, it is concluded that it is the

dinuclear species 6b [PtII t2( bpy)2(LTMS)(LTMS•)]1+ (vide infra).

The spectroelectrochemical behaviour upon one-electron oxidation of 7

[PtII t( bpy)(LPh)] is similar that to that of 6, but clear evidence for the formation of 7b

[PtII t Ph2( bpy)2(L )(LPh•)]1+ was not readily obtained. A small bathochromic shift of the

isosbestic point observed at 530 nm after the first 50% of the oxidation indicates the presence

of an intermediate which, as discerned from its EPR spectrum (vide infra), is probably 7b

[PtII t Ph2( bpy)2(L )(LPh•)]1+ (S = ½).

3.5 – X-Band EPR Spectroscopy:

When CH2Cl2 solutions containing 0.10 M [N(n-Bu) ](PF4 6) as a supporting electrolyte

and millimolar amounts of 5 [PdII t( bpy)(LTMS II t)], 6 [Pt ( bpy)(LTMS II t)] or 7 [Pt ( bpy)(LPh)] are

electrochemically or chemically fully oxidized by one-electron at 20 °C, the corresponding

frozen solutions are EPR-silent (at T < 80 K), which suggests the exclusive presence of the

diamagnetic dinuclear dicationic species 5c [PdII2(tbpy)2(LTMS•)2]2+ II t, 6c [Pt 2( bpy)2(LTMS•) ]2+

2

or 7c [PtII t2 ( bpy)2(LPh•) ]2+, respectively. 2

The EPR spectrum of compound 5 after 50% oxidation by coulommetry shows an

axial signal with g 105Pd nucleus (I = 5/2, 22.2%) was not iso = 2.035. Hyperfine coupling with

66

Chapter 3

observed (Figure 3.5.1 top). Due to the absence of hyperfine coupling no further structural

information could be obtained from the EPR, but according to similar spectroelectrochemical

features and changes observed with the Pt complex 6, we tentatively assign the intermediate

as 5b [PdII t2( bpy)2(LTMS)(LTMS•)]1+.

II t At 298 K, the EPR spectrum of dimer 7c [Pt 2( bpy)2(LPh•) ]2+ in CH Cl2 2 2 displays the

signal of a spin doublet species with g = 2.003 and 195iso Pt (I = ½, 33.8%) hyperfine coupling

of 38 x 10-4 cm-1 (Figure 3.5.1 bottom). This signal is due to the mononuclear 7a

[PtII t( bpy)(LPh•)]1+. The intensity of the spectrum (double integral) corresponds to a spin

concentration of about 80% of the chemical concentration of the compound. The mononuclear

species 7a [PtII t( bpy)(LPh•)]1+ II tand dinuclear species 7c [Pt 2( bpy)2(LPh•) ]2+ 2 establish an

equilibrium in solution. On the other hand, attempts to detect the corresponding mononuclear

radical 6a [PtII t( bpy)(LTMS•)]1+ in solution failed. Only the dinuclear, EPR silent species 6c

[PtII t TMS•2( bpy)2(L ) ]2+ is present in solution between 25 °C and –90 °C. 2

Interestingly, electrochemically one-electron oxidized CH Cl2 2 solutions of 8

[Pt(PPh3)2(LPh)] at 10 K display a rhombic signal with weak 195Pt hyperfine splitting in its

EPR spectrum (Table 3.5.1). The signal has been quantified (100%) and simulated. This EPR

signal is due to the presence of mononuclear, paramagnetic 8a [Pt(PPh3)2(LPh•)]1+. Therefore,

no dimer formation occurs in this case. II t In order to detect directly the intermediates in the oxidations of 6 [Pt ( bpy)(LTMS)]

and 7 [PtII t( bpy)(LPh)], we have also carried out stepwise coulometric oxidations of their

CH (0.10 M [N(n-Bu)Cl2 2 4](PF6)) solutions at –20 °C, recorded their electronic spectra, and

measured the corresponding EPR spectra at 30 K. Spin concentrations were obtained by

double integration of the signals compared to a standard 1 mM Cu(II) solution in H2O (2 M

NaClO , 10 mM HCl) measured under the same conditions. 4

67

Chapter 3

1.95 2 2.05 2.1 2.15

Figure 3.5.1 – X-band EPR spectra of 5a [PdII(tbpy)(LTMS•)]1+ (top) and 7a [PtII(tbpy)(LPh•)]1+

(bottom) in CH2Cl2 solution (0.10 M [N(n-Bu)4](PF6) at 30 K. Black lines represent the

experimental spectrum and red lines the simulation. Frequency, modulation amplitude, and

microwave power are 9.4335 GHz, 14 G, and 0.2 mW, respectively.

320 330 340 350

2.1 2.05 2 1.95 1.9

dχ´´

/ dB

B, mT

g values

310 320 330 340 350

dχ´´

/ dB

B, mT

g values

68

Chapter 3

II t Figure 3.5.2 shows the EPR spectra of a 50% oxidized solution of 6 [Pt ( bpy)(LTMS)]

II t(top) and 7 [Pt ( bpy)(LPh)] (bottom) and their respective simulations. The insets show the

increasing intensity of the signal with increasing oxidation level up to 50% (one electron

removed per two Pt ions). The intensity in the insets decreases after 50% until ultimately it is

EPR-silent at the 100% oxidation level, where only the diamagnetic dimers 6c

[PtII t TMS•2( bpy)2(L )2]2+ II t and 7c [Pt 2( bpy)2(LPh•) ]2+

2 exist at 30 K. This behaviour provides

evidence that the intermediates 6c and 7c are paramagnetic dinuclear species: 6b

[PtII t TMS2( bpy)2(L )(LTMS•)]1+ II t and 7b [Pt 2( bpy)2(LPh)(LPh•)]1+. A satisfactory simulation of the

two spectra in Figure 3.5.2 requires hyperfine interaction with two 195Pt nuclei as expected for

dimeric molecules with 33.8% natural abundance of 195Pt isotope. Details of the simulation

and parameters are given in the captions of Figure 3.5.2 and Table 3.5.1. Within experimental

accuracy the 195Pt nuclei appear to be equivalent, which indicates that the spin density has a

symmetric distribution, likely via the bridging Pt•••S interactions. The maximal spin

concentration (at 50% electrochemical conversion of the samples) reaches about 50% of the

Pt concentration for 6b but only about 14% for 7b (see insets in Figure 3.5.2), which is caused

by the chemical equilibrium of the following species: (1) the non-oxidized monomers (6

[PtII t( bpy)(LTMS II t)] or 7 [Pt ( bpy)(LPh II t)]); (2) the oxidized monomers (6a [Pt ( bpy)(LTMS•)]1+

II t7a [Pt ( bpy)(LPh•)]1+ II t TMS); (3) the singly oxidized dimers (6b [Pt 2( bpy)2(L )(LTMS•)]1+ or 7b

[PtII t Ph2( bpy)2(L )(LPh•)]1+ II t); and (4) the doubly oxidized dimers (6c [Pt 2( bpy)2(LTMS•) ]2+

2 and

7c [PtII t2( bpy)2(LPh•) ]2+

2 ). Since the concentration of the paramagnetic monomers is negligible

at low temperature (T < 80 K, vide supra) and 6c and 7c are diamagnetic, the spin

concentration of the measured spectra is controlled by the dissociation constants of 6c and 7c.

Thus, the spectroelectrochemical results (electronic absorption and EPR spectra) are in

agreement with the equilibria shown in Scheme 3.5.1. We have not found evidence for the

formation of the neutral dinuclear species 6d [PtII t2( bpy)2(LTMS II t)2] or 7d [Pt 2( bpy)2(LPh)2],

and therefore, K must be very small (vide infra). 3

69

Chapter 3

2.4 2.3 2.2 2.1 2 1.9 1.8 1.7

Figure 3.5.2 – X-band EPR spectra of 6b (top) and 7b (bottom) in CH2Cl2 solution (0.10 M

[N(n-Bu)4](PF6)) at 30 K. Black lines represent the experimental spectra and the red lines the

simulations. Frequency, modulation amplitude, and microwave power for 6b are 9.4335 GHz,

14 G, and 0.2 mW, respectively, and for 7b are 9.4213 GHz, 12G, and 1.0 mW, respectively.

The red lines are the superposition of three simulated subspectra with Gaussian lines and

anisotropic g values as given in Table 3.5.1 and hyperfine interactions with two, one and no 195Pt nuclei according to the statistical weights expected for symmetric Pt-dimers with total

spin S = ½ and 195Pt (33.8 % natural abundance): 11.4 % double-labeled, 44.8% single

labelled, and 43.8 % without 195Pt isotope. Insets represent the conversion of 6 or 7 to the

dimer 6b and 7b in a stepwise coulometric experiment.

280 320 360 400

2.4 2.2 2 1.8

dχ´´

/ dB

B, mT

g values

Oxidation conv. (%)

7b (%)

280 320 360 400

dχ´´

/ dB

B, mT

g values

Oxidation conv. (%)

6b (%)

70

Chapter 3

Table 3.5.1 – X-band EPR parameters of S = ½ compounds.

Complex giso gxx gyy gzz Hyperfine coupling constantsd (10-4 cm-1)

5ba A2.035 2.065 2.065 1.975 xx = Ayy = 50; Azz = 20

6bb A2.074 2.183 2.167 1.856 xx = 200; Ayy = 153; A = 140 zz

7ac 2.003 A = 38eiso

7bc A2.061 2.176 2.086 1.914 xx = 103; Ayy = 112; A = 93 zz

8aa A2.010 2.020 2.006 1.989 xx = 70; Ayy = 20; Azz = 90 a b c In CH2Cl2 solution at 30 K. In CH2Cl2 solution at 10 K. In CH2Cl solution at 298 K. d

2

The sign of the A values is not determined. e The value of A0 measured for 7a in fluid solution

compares well with the anisotropic A-tensor components of 7b, if one assumes that one of the

components (presumably Azz) has the opposite sign than the others (which is reasonable if as

usual the traceless spin-dipolar contribution is the main source of anisotropy of the A-tensor).

The isotropic part for 7b can be estimated by A0 = 41 x 10-4 cm-1, according to the relation

|A | = ⅓(A0 xx + Ayy + A ). zz

6, 7

K3

Scheme 3.5.1 – Redox process cycle involved in Pt complexes 6 and 7.

6d, 7d6, 7

6b, 7b

6a, 7a 6c, 7c

6, 7

(L) = (LTMS) or (LPh); (L ) = (LTMS•) or (LPh• )•

K3K3

6, 7 6d, 7d

6b, 7b

6a, 7a 6c, 7c

(L) = (LTMS) or (LPh); (L ) = (LTMS•) or (LPh• )•(L) = (LTMS) or (LPh); (L ) = (LTMS•) or (LPh• )•(L) = (LTMS) or (LPh); (L•) = (LTMS•) or (LPh•)

71

Chapter 3

3.6 – Estimation of Equilibrium Constants:

With the use of Scheme 3.5.1, it has been possible to simulate the current/voltage

profiles of the cyclic voltammograms of 6 [PtII t( bpy)(LTMS )], shown in Figure 3.6.1. The

shapes of the voltammograms depend mainly on (1) the three equilibrium constants K , K , K1 2 3

and the three redox potentials E2 2´½, E , and E2´´

½ ½ and (2) the rate constants for dimerisation

(k -1f, M s-1 -1) and dissociation (k , sb ) for the three processes in Scheme 3.5.1. The six

thermodynamic parameters (three K values and three E2½) are not completely independent of

each other since the sum of the free energy (ΔG0) values around the two reaction squares in

Scheme 3.5.1 must equate to zero. The diffusion coefficient (D) was set to 3 x 10-6 cm2 -1s for

all species involved. The fit parameters are summarized in Table 3.6.1.

According to the simulation, the chain of events during an anodic scan is one-electron

oxidation of 6 [PtII t( bpy)(LTMS II t)] to 6a [Pt ( bpy)(LTMS•)]1+ at E2pa (E2

½), followed by rapid

dimerisation 6a + 6 → 6b [PtII t2( bpy)2(LTMS TMS•)(L )]1+ and further one-electron oxidation of

6b to the dinuclear dication 6c [PtII t2( bpy)2(LTMS•)2]2+ at E2´

pa (E2´½). In the cyclic

voltammogram, the oxidation 6 → 6a is cathodically shifted because of the kinetic effect of

the following equilibrium K . Reversing this sequence leads to the reduction 6c → 6b at E2´2 pc

(E2´½) which proceeds at the normal half wave potential E2 (6c/6b). Since K½ 2 is large, the

final reduction to 6 takes place at the more negative potential E2pc via the thermodynamically

less favourable pathway E2´´ II t½ (6b/6d) with subsequent dissociation to 6 [Pt ( bpy)(LTMS)].

The same chain of events occurs also for complex 7.

Table 3.6.1 – Parameter set for the simulation of the cyclic voltammogram of

compound 6 (concentration = 10-3 M) at 25 °C (D = 3 x 10-6 cm2 -1s for all species)a

-1 -1 -1 -1K k k , Mn f, M s b, s

K1 1.31 x 105 2.0 x 108 1.5 x 103

K2 1.0 x 105 1.2 x 106 1.2 x 102

3 3K ~0.5 ~10 ~103

a Redox potentials vs Fc+/Fc: E2½, 0.167 V; E2´ , 0.160 V; E2´´ , –0.145 V. See Scheme 3.5.1. ½ ½

Figure 3.6.1 shows that the parameter set in Table 3.4.1 adequately reproduces the cyclic

voltammogram of 6 at two different scan rates. Interestingly, at room temperature the stability

constants for the mono- and dicationic dimers 6b [PtII2(tbpy)2(LTMS)(LTMS•)]1+ and 6c

72

Chapter 3

II t TMS•[Pt 2( bpy)2(L )2]2+ are similar, whereas dimerization of 6 to 6d is not observed (K3 is very

small), and the redox potentials for the couples 6/6a and 6c/6b are almost identical.

1600 mV s-1

exp.

sim.Epc2'

Epc2''

Epa2'

Epa2

Figure 3.6.1 – Cyclic voltammogram of complex 6 [PtII(tbpy)(LTMS)] in CH2Cl2 solution

(0.10 M [N(n-Bu)4](PF6) using scan rates of 1600 (top) and 200 mV s-1 (bottom) at 20 °C.

Simulations (red lines) are shown using parameters given in Table 3.6.1.

The equilibrium constant K1 for the complex 7 [PtII(tbpy)(LPh)] was determined from

the absorption spectroscopy data. The electronic spectrum of 7c in CH2Cl2 is strongly

temperature dependent in the range from –35 to +33 °C as shown in Figure 3.6.2.

Simultaneously with the disappearance of the absorptions at 685 and 1050 nm, a new

1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 -2.0 -2.5

200 mV.s-1

Epc

E (V) versus Fc+/Fc

sim.

exp.

Epc2''

2'

Epa2' Epa

2

73

Chapter 3

absorption maximum grows at 728 nm upon increasing the temperature from –35 °C to

+33 °C. The process is fully reversible. The low-temperature spectrum is assumed to be that

of the dimer 7c [PtII t2( bpy)2(LPh•) ]2+

2 , whereas at 20 °C the mononuclear paramagnetic 7a

[PtII t( bpy)(LPh•)]1+ dominates, as elucidated by EPR spectroscopy (Figure 3.5.2).

These observations allow the assignment of the spectral changes to the temperature

dependence of equilibrium constant K . Values of K1 1 were calculated by using a molar

absorption coefficient of 12400 M-1cm-1 for 7c at 1015 nm, which was established from a

measurement at –50 °C where the dimer 7c is exclusively present. -1 II t A value of 660 M was found for K1 of 7c [Pt 2( bpy)2(LPh•) ]2+

2 at 25 °C. From the

temperature dependence of K1 the corresponding enthalpy and entropy values (ΔH0, ΔS0 at

295 K) were obtained: ΔH0 -1 = –50 ± 1 kJ mol and ΔS0 = –59 ± 2 kJ K-1 -1 mol . Furthermore,

the increasing value of K1 with decreasing temperature (2.3 x 106 M-1 is calculated from

–90 °C in CH II t2Cl2) indicates that the concentration of paramagnetic 7a [Pt ( bpy)(LPh•)]1+ is

negligible at –90 °C and, therefore, becomes undetectable by EPR spectroscopy in frozen

solution.

400 500 600 700 800 900 10000.0

0.2

0.4

0.6

0.8

1.0

Abs

orba

nce

λ, nm

7a

7c

238 K

306 K

Figure 3.6.2 – Electronic spectra of 7c in CH2Cl2 solution ([Pt]tot = 2.44 x 10-4 M, l = 0.5 cm)

recorded in the temperature range from –35 to +33 °C (the black arrows indicate the spectral

changes).

74

Chapter 3

3.7 – Conclusions:

The preparation and structural characterization of the neutral, square planar complexes

5 [PdII t( bpy)(LTMS II t)], 6 [Pt ( bpy)(LTMS II)], and 8 [Pt (PPh3) (LPh2 )] were reported.

Electrochemical and chemical one-electron oxidation of compounds 5, 6 and 7 in CH Cl2 2

solution afford the monomeric monocations 5a [PdII t TMS•( bpy)(L )]1+ II t, 6a [Pt ( bpy)(LTMS•)]1+

and 7a [PtII t( bpy)(LPh• 1+)] , respectively, which possess an S = ½ ground state. The cyclic

voltammograms of 5 and 6 show complex oxidation features, indicating the presence of more

than one species in solution. In fact the bulky ligands (LTMS 2-) and (LPh 2-) do not hinder the

dimerization of 6a and 7a intermediates under electrochemical conditions. The corresponding

spin doublet monocationic dimers 6b [PtII t2( bpy)2(LTMS)(LTMS•)]1+ and 7b

[PtII t Ph2( bpy)2(L )(LPh•)]1+ were electrochemically generated in solution (after 50% oxidation)

and identified by EPR spectroscopy. The X-band EPR spectra of the stepwise coulometrically

oxidized samples of 6 and 7 could be simulated considering: (1) spin concentration analysis;

and (2) a 195Pt hyperfine statistical contribution, which, in conjunction, gave reasonable

solutions for the spectra of the intermediate dimers 6b and 7b. In the case of complex 5a no

spectroscopical evidence of dimerization has been obtained, due to the absence of hyperfine

coupling with the 105Pd nucleus. Complete one-electron oxidation of 6 and 7 yielded the

diamagnetic dicationic dimers 6c [PtII t2( bpy)2(LTMS•)2]2+ II t and 7c [Pt 2( bpy)2(LPh•) ]2+

2 which

are in equilibrium with the corresponding paramagnetic monomers 6a and 7a in solution.

Evidence of dimer formation was obtained by X-ray analysis in crystals for 7c

[PtII t Ph•( bpy) (L ) ](PF )2 2 2 6 2•3CH Cl2 2. The structure revealed a centrosymmetric, lateral dimer

whose bridging part is a PtII (μ -S)2 2 2 rhomboid; the metal ions possess a square pyramidal

geometry. Solid-state sulfur K-edge X-ray absorption spectra of 7a, 7c and 9 [Pt(LPh•) ]02

showed clearly the presence of sulfur-centered radicals (LPh• 1-) which are absent in the neutral

complexes 6 and 7. One-electron oxidation of 8 [PtII(PPh3) (LPh2 )] afforded only the spin

doublet species 8a [PtII(PPh3)2(LPh•)]1+ and no dimer formation was detected. The equilibrium

constants have been determined from simulation of the cyclic voltammograms of 6 and

experimentally obtained for compound 7. In such systems containing [MII(bpy)(dithiolate)]

complexes (M = Pt or Pd), the formation of dimers upon oxidation must be taken into

consideration.

75

Chapter 3

3.8 – References: 1 Cummings, S. D.; Eisenberg, R. Prog. Inorg. Chem. 2003, 52, 315-367. 2 Paw, W.; Cummings, S. D.; Mansour, M. A.; Connick, W. B.; Geiger, D. K.; Eisenberg, R. Coord. Chem. Rev. 1998, 171, 125-150. 3 Connick, W. B.; Gray, H. B. J. Am. Chem. Soc. 1997, 119, 11620-11627. 4 Puthraya, K. H.; Srivastava, T. S. Polyhedron, 1985, 4, 1579-84. 5 Vogler, A.; Kunkely, H. J. Am. Chem. Soc. 1981, 103, 1559-60. 6 Vogler, A.; Kunkely, H.; Hlavatsch, J.; Merz, A. Inorg. Chem. 1984, 23, 506-9. 7 Zuleta, J. A.; Burberry, M. S.; Eisenberg, R. Coord. Chem. Rev. 1990, 97, 47-64. 8 Geary, E. A. M.; Yellowlees, L. J.; Jack, L. A.; Oswald, I. D. H.; Parsons, S.; Hirata, N.; Durrant, J. R.; Robertson, N. Inorg. Chem. 2005, 44, 242-250. 9 Weinstein, J. A.; Tierney, M. T.; Davies, E. S.; Base, K.; Robeiro, A. A.; Grinstaff, M. W. Inorg. Chem. 2006, 45, 4544-4555. 10 Zuleta, J. A.; Bevilacqua, J. M.; Rehm, J. M.; Eisenberg, R. Inorg. Chem. 1992, 31, 1332-7. 11 Zuleta, J. A.; Chesta, C. A.; Eisenberg, R. J. Am. Chem. Soc. 1989, 111, 8916-17. 12 Ghosh, P.; Begum, A.; Herebian, D.; Bothe, E.; Hildenbrand, K.; Weyhermuller, T.; Wieghardt, K. Angew. Chem., Int. Ed. 2003, 42, 563-567. 13 Makedonas, C.; Mitsopoulou, C. A.; Lahoz, F. J.; Balana, A. I. Inorg. Chem. 2003, 42, 8853-8865. 14 Bevilacqua, J. M.; Eisenberg, R. Inorg. Chem. 1994, 33, 2913-23. 15 Matsubayashi, G.; Hirao, M.; Tanaka, T. Inorg. Chim. Acta 1988, 144, 217-21. 16 Nakahama, A.; Nakano, M.; Matsubayashi, G.-e. Inorg. Chim. Acta 1999, 284, 55-60. 17 Smucker, B. W.; Hudson, J. M.; Omary, M. A.; Dunbar, K. R. Inorg. Chem. 2003, 42, 4714-4723. 18 McInnes, E. J. L.; Farley, R. D.; Macgregor, S. A.; Taylor, K. J.; Yellowlees, L. J.; Rowlands, C. C. J. Chem. Soc., Faraday Trans. 1998, 94, 2985-2991. 19 Zuleta, J. A.; Bevilacqua, J. M.; Proserpio, D. M.; Harvey, P. D.; Eisenberg, R. Inorg. Chem. 1992, 31, 2396-404. 20 Schrauzer, G. N.; Mayweg, V. P.; Heinrich, W. Inorg. Chem. 1965, 4, 1615-17.

76

Chapter 3

21 Connick, W. B.; Geiger, D.; Eisenberg, R. Inorg. Chem. 1999, 38, 3264-3265. 22 Shukla, S.; Kamath, S. S.; Srivastava, T. S. J. Photochem. Photobiol., A 1989, 50, 199- 207. 23 Zhang, Y.; Ley, K. D.; Schanze, K. S. Inorg. Chem. 1996, 35, 7102-7110. 24 Ray, K.; Bill, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 5641-5654. 25 Ray, K.; George, S. D.; Solomon, E. I.; Wieghardt, K.; Neese, F. Chem. Eur. J. 2007, 13, 2783-2797. 26 Ray, K.; Weyhermueller, T.; Goossens, A.; Craje, M. W. J.; Wieghardt, K. Inorg. Chem. 2003, 42, 4082-4087. 27 Ray, K.; Weyhermueller, T.; Neese, F.; Wieghardt, K. Inorg. Chem. 2005, 44, 5345- 5360. 28 Szilagyi, R. K.; Lim, B. S.; Glaser, T.; Holm, R. H.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. J. Am. Chem. Soc. 2003, 125, 9158-9169. 29 Bowmaker, G. A.; Boyd, P. D. W.; Campbell, G. K. Inorg. Chem. 1983, 22, 1208-13. 30 Kokatam, S.; Ray, K.; Pap, J.; Bill, E.; Geiger, W. E.; LeSuer, R. J.; Rieger, P. H.; Weyhermueller, T.; Neese, F.; Wieghardt, K. Inorg. Chem. 2007, 46, 1100-1111. 31 Browall, K. W.; Bursh, T.; Interrante, L. V.; Kasper, J. S. Inorg. Chem. 1972, 11, 1800-6. 32 Faulmann, C.; Legros, J. P.; Cassoux, P.; Cornelissen, J.; Brossard, L.; Inokuchi, M.; Tajima, H.; Tokumoto, M. J. Chem. Soc., Dalton Trans. 1994, 249-54. 33 Bousseau, M.; Valade, L.; Legros, J. P.; Cassoux, P.; Garbauskas, M.; Interrante, L. V. J. Am. Chem. Soc. 1986, 108, 1908-16. 34 Garreau, B.; Pomarede, B.; Cassoux, P.; Legros, J. P. J. Mater. Chem. 1993, 3, 315- 316. 35 Pomarede, B.; Garreau, B.; Malfant, I.; Valade, L.; Cassoux, P.; Legros, J.-P.; Audouard, A.; Brossard, L.; Ulmet, J.-P. Inorg. Chem. 1994, 33, 3401. 36 Alves, H.; Simao, D.; Santos, I. C.; Gama, V.; Henriques, R. T.; Novais, H.; Almeida, M. Eur. J. Inorg. Chem. 2004, 1318-1329. 37 Enemark, J. H.; Lipscomb, W. N. Inorg. Chem. 1965, 4, 1729-1734. 38 Fettouhi, M.; Ouahab, L.; Hagiwara, M.; Codjovi, E.; Kahn, O.; Constant-Machado, H.; Varret, F. Inorg. Chem. 1995, 34, 4152-9.

77

Chapter 3

39 Welch, J. H.; Bereman, R. D.; Singh, P.; Moreland, C. Inorg. Chim. Acta 1989, 158, 17-25. 40 Zurcher, S.; Gramlich, V.; Von Arx, D.; Togni, A. Inorg. Chem. 1998, 37, 4015-4021. 41 Hamilton, W. C.; Bernal, I. Inorg. Chem. 1967, 6, 2003-8. 42 Kang, B. S.; Weng, L. H.; Wu, D. X.; Wang, F.; Guo, Z.; Huang, L. R.; Huang, Z. Y.; Liu, H. Q. Inorg. Chem. 1988, 27, 1128-30. 43 Rodrigues, J. V.; Santos, I. C.; Gama, V.; Henriques, R. T.; Waerenborgh, J. C.; Duarte, M. T.; Almeida, M. J. Chem. Soc., Dalton Trans. 1994, 2655-60. 44 Schultz, A. J.; Eisenberg, R. Inorg. Chem. 1973, 12, 518-25. 45 Tamura, H.; Tanaka, S.; Matsubayashi, G.; Mori, W. Inorg. Chim. Acta 1995, 232, 51- 5. 46 Simao, D.; Alves, H.; Belo, D.; Rabaca, S.; Lopes, E. B.; Santos, I. C.; Gama, V.; Duarte, M. T.; Henriques, R. T.; Novais, H.; Almeida, M. Eur. J. Inorg. Chem. 2001, 3119-3126. 47 Kokatam, S.-L.; Chaudhuri, P.; Weyhermueller, T.; Wieghardt, K. Dalton Trans. 2007, 373-378. 48 Ray, K.; Petrenko, T.; Wieghardt, K.; Neese, F. Dalton Trans. 2007, 1552-1566. 49 Pap, J. S.; Benedito, F. L.; Bothe, E.; Bill, E.; George, S. D.; Weyhermueller, T.; Wieghardt, K. Inorg. Chem.) 2007, 46, 4187-4196. 50 Cummings, S. D.; Eisenberg, R. J. Am. Chem. Soc. 1996, 118, 1949-1960.

78

Chapter 4

79

Chapter 4

Electronic Structure of

Square Planar Cobalt and Rhodium Complexes

Containing a bis(ortho-Benzenedithiolate) Ligand

Chapter 4

80

Chapter 4

81

4.1 Introduction

An interesting series of bis(benzenedithiolate) metal complexes of cobalt with the

general formula [Co(L´)2]z (z = 0, -1, -2) displays diverse structural, magnetic and

spectroscopic properties depending on the overall charge of the complex and the electron-

withdrawing nature of the substituents at the ligands. The magnetic properties of the

monoanionic species show at first sight a rather confusing behaviour because some have been

reported to be diamagnetic, whereas others have an S = 1 ground state (Table 4.1.1). The

discovery that many of these presumed monomeric monoanions are actually dimeric dianions

led to a more consistent picture because they are diamagnetic (intramolecular

antiferromagentic coupling). The extent of dimerization is governed by the electronic nature

of the coordinating ligands and the overall oxidation state of the complex.1

Table 4.1.1 – Magnetic moments (BM) of cobalt complexes with varying ligand systems and

phases.

Phase Ligands

NC S

SNC

F3C S

SF3C

Cl

Cl

Cl

Cl

S

S

S

S

S

S

S

S

S

S

(mnt)2- (tdf)2- (tcdt)2- (L)2- (LMe2)2- (LMe)2- (LMe4)2-

solid Diamag.a Diamag.a Diamag.a 3.27 3.18 3.24 3.23

cyclohexanone Diamag.a - 3.14 - 3.29 - -

THF - - 3.18 - - - -

DMSO 2.81 - 2.37 - 3.39 - - a Dimeric complex.

An example that illustrates the effect of the oxidation state of the complex upon

molecular geometry is obtained by comparing the structures of the monomer2 [CoII(mnt)2]2-

(mnt2- = maleonitriledithiolate(2-)) and the two electron oxidized dimer3,4 [Co(mnt)2]22-. The

metal ion is displaced 0.37 Å out of the plane of the four basal sulfur atoms with axial

bonding to a fifth sulfur atom, as shown for the M–S type dimer in chapter 3 (Figure 3.2.6).

The oxidation of [Co(mnt)2]2- results in the removal of considerable electron density from the

vicinity of the cobalt and the Co•••S intermolecular interaction becomes favourable.

The effect of the electron-withdrawing nature of the substituents at the ligand is

observed by comparing [Co(LMe)2]1- and [Co(tcdt)2]

1- as reported by Gray et al.

Chapter 4

82

((LMe)2- = toluene-3,4-dithiolate; (tcdt)2- = 3,4,5,6-tetrachlorobenzene-1,2-dithiolate).5 In both

complexes, the cobalt atoms possess the same formal oxidation state and electronic

configuration but the magnetic susceptibility measurements are significantly different. The

[Co(LMe)2]1- is one of the most unusual benzenedithiolate complexes because of its stable

intermediate spin (S = 1) ground state, which is observed in different solid salts and in

solution.5 This remarkable compound synthesized in the 1960s became the first example of a

square planar complex having a spin triplet ground state. X-ray studies on

[AsMePh3][Co(LMe)2] revealed the presence of well defined monomeric square planar anions

with the cobalt centers being 10 Å apart.6 In contrast to [Co(LMe)2]1-, the presumed

[Co(tcdt)2]1- is diamagnetic in the solid state but exhibits a triplet ground state in solutions of

weakly coordinating solvents such as cyclohexanone and THF (see Table 4.1.1).7 The crystal

structure of the [N(n-Bu)4][Co(tcdt)2] shows the salt to consist of a dimer [Co(tcdt)2]

1- in the

solid state with a square pyramidal geometry at the cobalt ion. When this complex is

dissolved, dissociation in monomers is observed yielding the square planar [Co(tcdt)2]1- with a

spin triplet ground state similar to that of [Co(LMe)2]1-.7

However, the spin triplet ground state of [Co(LMe)2]1- suggested by Gray et al.5 was

contested by Hatfield et al.8 According to the authors, the complex has a spin singlet ground

state and the observed paramagnetism was explained as a result of mixing between the ground

state with a low lying excited triplet state. This theory was ruled out in the beginning of the

1970s by van der Put et al.9 based on magnetic susceptibility and far infrared measurements.

The authors confirmed the nature of the spin triplet ground state (S = 1), where the

degeneracy of the ground state is decreased by a zero-field splitting of 40 cm-1. No evidence

for the existence of an excited multiplet within 300 cm-1 of the ground state was found.

The correct assignment of the spectroscopic oxidation state for the central cobalt atom

in [Co(LMe)2]1- proved to be uncertain. Gray et al. initially proposed an intermediate spin

Co(III) configuration.5,6 Later, based on the electronic spectrum of the complex, which

showed an intense band at 630 nm (1.3 x 104 M-1 cm-1), Schrauzer et al.10 proposed a Co(I)

(d8) ion coordinated to two ortho-dithiobenzosemiquinonate(1-) radicals. Thus, the authors

supported the spin singlet ground state description proposed by Hatfield et al.8 More recently,

the crystal structure of [PMePh3][Co(L)2] was reported with average C–S bond lengths at

1.764 Å and no visible semiquinoid type distortion in the phenyl rings.11 These experimental

observations contradict the description of the complex possessing a central Co(I) (d8) ion

coordinated to two ortho-dithiobenzosemiquinonate(1-) radicals. Sawyer et al.,12 on the other

hand, proposed a Co(II) assignment for the complex thereby implying a ligand based

Chapter 4

83

oxidation for the [Co(LMe)2]2-/[Co(LMe)2]

1- couple. This assignment was made on the basis of

the enhanced susceptibility to oxidation of the [Co(LMe)2]2- species as compared to that of

[CoCl4]2-, in which the oxidation is undoubtedly metal centered.

The unambiguous assignment of metal or ligand oxidation based on oxidation

potentials is not reasonable when comparing complexes with totally different ligand systems.

The contributions of ligand and metal to the orbitals involved in the redox process vary

significantly with different substituents on the phenyl rings. As a consequence, the oxidation

potentials of the complexes are affected to a great extent. There are two different formulations

of the [Co(L)2]1- species ((L)2- = ortho-benzenedithiolate) and its analogues: (1) a Co(II) (d7)

metal ion coordinated to one closed-shell and one ortho-dithiobenzosemiquinonate(1-) ligand

radical, and (2) a complex consisting of a Co(III) (d6) metal center coordinated to two closed-

shell ligands. However, it is not possible to differentiate between the electronic structures of

[CoIII(L)2]1- and [CoII(L)(L•)]1- based on the available experimental data. The UV-Vis and

cyclic voltammetry results are contradictory to each other in their description of the electronic

structures of the complexes. Crystallographic data from systems containing ortho-

benzenedithiolates(2-) and ortho-dithiobenzosemiquinonate(1-) radicals do not show

significant change in the bond lengths due to delocalization of the free electron between the

two ligands. Recently, Wieghardt et al.13 reported a detailed study on the [Co(LBu)2]1-

complex (LBu = 3,5-di-tert-butylbenzene-1,2-dithiolate). Based on spectroscopic and DFT

data, the ground state of the complex was described as a combination of the following

resonance forms: [CoIII(L)2]1- ↔ [CoII(L)(L•)]1- ↔ [CoII(L•)(L)]1- with a greater weight for the

first resonance structure. In this chapter, the electronic structure of the electron transfer series

of a monoanionic cobalt complex containing the LTMS is described and the experimental

results are compared to DFT calculation data in order to assign the physical oxidation state of

the [Co(LTMS)2]1- species.

Belonging to the same group, rhodium complexes are of interest and the isoelectronic

species can be compared to the cobalt analogues. The majority of isolated monomeric four-

coordinate rhodium complexes have the metal center in the RhI oxidation state,14,15 whereas

RhII complexes have seldom been reported.15,16 Such species have been identified as short-

lived intermediates in flash photolysis studies and as intermediates in the stepwise reduction

of RhIII compounds in cyclic voltammetry experiments.17-19 Examples of RhII in square-planar

geometry are rare, with only seven complexes characterized by X-ray crystallography.20 One

of the most important square-planar Rh compounds is chloro-tris(triphenylphosphine)

rhodium(I), RhICl(PPh3)3, known as Wilkinson´s catalyst.21 This red-violet compound

Chapter 4

84

catalyses (1) hydrogenation of alkenes,22,23 (2) hydroboration of alkenes with catecholborane

or pinacolborane,24 and (3) selective reduction of α,β-unsaturated carbonyl compounds in

combination with triethylsilane.25 Since the beginning of the use of Wilkinson´s catalyst, a

small amount of paramagnetic impurity was detected by X-band EPR measurements. The

structure and the correct composition were not known until 1990, when Ogle et al.19

characterized the yellow impurity as being trans[RhIICl2(PPh3)2].

In dithiolate chemistry, [N(n-Bu)4]2[RhII(mnt)2] represents the first square-planar RhII

complex coordinated to a dithiolate ligand reported in the literature (Figure 4.1.1).26

Figure 4.1.1 – The proposed structure of the dianion [Rh(mnt)2]2-.

Based on analytical data, conductance and X-ray powder diffraction the authors were able to

establish indirectly the square-planar geometry at the rhodium ion. Magnetic susceptibility

measurements show clearly a µeff of 1.91 B.M. indicating only one unpaired electron, which

suggests a RhII (d7) metal ion.26 A few months later, Holm et al.27 reported the electronic

structure of [N(n-Bu)4]2[RhII(mnt)2] based on spectroscopic and theoretical data.

In this chapter, the synthesis and structural characterization of such rhodium

complexes with (LTMS)2- ligands are presented. A combination of experimental techniques and

DFT calculations allows the description of the electronic structures of the new complexes in

detail.

S

Rh

S

S

S

CN

CNNC

NC

2-

Chapter 4

85

Results and Discussion

4.2 – Synthesis and X-ray Crystal Structures:

When one equivalent of the dipotassium salt of the ligand 1b in degassed MeOH

containing [N(n-Bu)4]I reacts with half an equivalent of Co(CH3COO)2•4H2O, an air-sensitive

light green solution is obtained. Upon exposure to air the solution changes color to deep blue

generating the monoanion 10 [CoIII(LTMS)2]1-. Large rod-like crystals of

[N(n-Bu)4][10] were obtained in very good yields (94%) after removal of the solvent and

dissolution of the compound in MeCN. Attempts to isolate a salt containing the green dianion

10b [CoII(LTMS)2]2- were unsuccessful.

The crystal structure of compound [N(n-Bu)4][10] was determined at 100(2) K using

Mo Kα radiation, and shows two crystallographically independent monoanions 10 and two

cations in the unit cell. Figure 4.2.1 shows the important structural features and Table 4.2.1

summarizes the selected bond lengths. The coordination geometry at the central cobalt atom is

square planar with inter- and intra-ligand S•••S distances of 3.079 and 3.053 Å, respectively.

The average intramolecular Co–S distance at 2.168 Å is in full agreement to what is known

for other related monoanionic ortho-benzenedithiolate complexes of cobalt.2-4,13,28 In the

crystal structure, the anions are well separated with intramolecular Co•••Co and Co•••S

distances of 8.558 and 8.356 Å, respectively. Thus intermolecular interactions leading to spin

coupling phenomena are not likely. The six C–C bond lengths of the phenyl rings are 1.408 Å

on average and are equidistant within experimental errors (± 0.01 ≡ 3σ). The average C–S

bond is long at 1.77 ± 0.01 Å. These data indicate that both ligands are closed-shell dianions

(LTMS)2-, and, consequently, the central cobalt ion possesses a formal oxidation state of +III

with a d6 electron configuration in a square-planar field. Almeida et al. have described the

crystal structure of deep-blue dianion [CoII(LCN)2]2- ((LCN)2- = 4,5-dicyanobenzene-1,2-

dithiolate), and found that the average Co–S distance is 2.179(1) Å, which is longer by

0.011 Å than that in monoanion 10 at 2.168(1) Å. This is due to the change from low spin CoII

to intermediate-spin CoIII.

Chapter 4

86

Figure 4.2.1 – Perspective view and numbering scheme of the monoanion 10 in crystals of

[N(n-Bu)4][10] with thermal ellipsoids at 50% level. Hydrogen atoms are omitted for clarity.

When RhCl3•nH2O in glacial acetic acid and absolute ethanol solution is heated to

reflux under nitrogen in the presence of Na(CH3COO)•3H2O a dark green precipitate is

isolated. After dissolution in boiling methanol, blue-green crystals of

[Rh(CH3COO)2]2•2MeOH were obtained in 80% yield. The MeOH was removed under

vacuum at 45 °C for 20 hours and periodically monitored by infrared spectroscopy following

the reduction of the bands at 3400 and 1010 cm-1 characteristic for the O–H stretching

modes.29 Dark blue K4[Rh2(CO3)4]•2H2O was obtained in 89% yield after heating

[Rh(CH3COO)2]2 to reflux in a aqueous solution of K2CO3 (3 M).30 Treatment of

K4[Rh2(CO3)4]•2H2O in MeCN with an excess of a diethylether solution of HBF4 (54%)

results in a remarkably stable unbridged [Rh2(MeCN)10](BF4)4. This solid is hygroscopic as

evidenced by its facile conversion to the pink axial bis-water adduct when exposed to air or

undried solvents.14 After refluxing for three days, an orange solid was isolated and dried,

yielding orange rod-like crystals of [Rh2(MeCN)10](BF4)4. The procedure was improved by

using K4[Rh2(CO3)4]•2H2O and not the sodium salt, as the NaBF4 side product is soluble in

MeCN and KBF4 is not. This procedure simplifies the purification described by Dunbar

et al.14 The tetracation [Rh2(MeCN)10]4+ moiety is a rare example of an unbridged dimer with

a RhII–RhII single bond length of 2.624(1) Å, which is shorter than other unbridged examples.

The equatorial planes of the MeCN ligands are twisted with respect to each other

C(2)

C(1)

C(3)

C(4)

C(5)

C(6) S(1)

S(2)

Co(1)

1-

Chapter 4

87

(χav = 44.8(2)°) and the axial MeCN ligands deviate from linearity reducing the molecular

symmetry from an expected D4d to C2.14 The Rh–Rh bond can be easily broken by light.

According to Dunbar et al., the irradiation of [Rh2(MeCN)10](BF4)4 in methanol leads the

formation of paramagnetic RhII ions as well as RhI and RhIII species within the time scale of

one hour, but the dinuclear cation can be regenerated in essentially quantitative yields.31

Equation 4.2.1 shows the balanced equation for the synthesis of [Rh2(MeCN)10](BF4)4.

Eqn. 4.2.1

K4[Rh2(CO3)4] + xs. HBF4 + 10CH3CN [Rh2(MeCN)10](BF4)4 + 4CO2 + 4H2O + 4KBF4

The reaction of a quarter of an equivalent of the ligand 1b, with one equivalent of

[Rh2(MeCN)10](BF4)4 and two equivalents of [N(n-Bu)4]I in MeCN affords a red brown

solution. Rod-like crystals were obtained after storing the solution for a few days at -30 °C,

yielding the dark red crystalline salt of the dianion 11 [N(n-Bu)4]2[RhII(LTMS)2]•4MeCN in

67% yield (Figure 4.2.2).

Figure 4.2.2 – Perspective view and numbering scheme of the dianions 11 [RhII(LTMS)2]2- in

crystals of [N(n-Bu)4]2[11]•4MeCN with thermal ellipsoids at 50% probability level.

Hydrogen atoms are omitted for clarity.

2-

C(3)

C(1) C(5)

C(2) C(4)

C(6)

S(2)

S(1)

Rh(1)

Chapter 4

88

The crystal structure of [N(n-Bu)4]2[11]•4MeCN was determined at 100(2) K using Mo Kα

radiation. Figure 4.2.2 shows the dianion 11 [RhII(LTMS)2]2-, and the main structural features

are summarized in Table 4.2.1. The rhodium atom sits at a center of symmetry with Rh–S

bond lengths of 2.27 ± 0.01 Å on average. Two MeCN molecules are present in the crystal

structure with N•••Rh distances at 7.023 and 8.333 Å indicating clearly that the MeCN

molecules are not coordinated to the central metal ion. Figure 4.2.3 displays the packing motif

of compound 11 [N(n-Bu)4]2[RhII(LTMS)2]•4MeCN. The six C–C bond lengths of the phenyl

rings are 1.406 Å on average and are equidistant within experimental error (± 0.01 ≡ 3σ). The

average C–S bond is long at 1.75 ± 0.01 Å, which may indicate the presence of a π-ligand

radical, but DFT calculations and EPR spectroscopy support that both ligands are closed-shell

dianions (LTMS)2-, and consequently the central rhodium ion possesses a formal oxidation

state of +II with a d7 electron configuration in a square-planar geometry. Sellmann et al.32

reported the crystal structure of the neutral square planar compound [RhI(CO)(PPh3)(mtbt)]

(mtbt = ortho-methylthiobenzenethiolate(1–)) with a Rh–S bond distance at 2.332(2) Å.

Longer Rh–S bond distances are expected for the RhI compound as a consequence of the

lower oxidation state of the RhI metal ion. So far no examples of structurally characterized

RhIII complexes containing the ortho-benzenedithiolate ligand have been reported. Gray

et al.26 reported the X-ray powder diffraction of [N(n-Bu)4]2[RhII(mnt)2] which was found to

be isomorphous and presumably isostructural with [N(n-Bu)4]2[NiII(mnt)2]. Since the nickel

compound was already known to have a square planar geometry, the same geometry was

attributed to the rhodium complex.

Figure 4.2.3 – Packing motif in crystals of complex [N(n-Bu)4]2[11]•4MeCN.

Rh

S

N

N S

S

S

N

N

N

N

Si Si

Si

Si

Chapter 4

89

Table 4.2.1 – Selected bond lengths (Å) in anions 10 and 11.

Selected bonds 10 [CoIII(LTMS)2]1-

11 [RhII(LTMS)2]2-

M(1)-S(1) 2.1662(6) 2.2715(4)

M(1)-S(2) 2.1694(6) 2.2701(4)

S(1)-C(1) 1.772(2) 1.755(1)

S(2)-C(2) 1.773(2) 1.753(1)

C(1)-C(2) 1.408(3) 1.414(2)

C(2)-C(3) 1.416(3) 1.418(2)

C(3)-C(4) 1.402(3) 1.393(2)

C(4)-C(5) 1.395(4) 1.396(3)

C(5)-C(6) 1.404(3) 1.396(2)

C(6)-C(1) 1.424(3) 1.417(2)

4.3 – Electro- and Spectroelectrochemistry:

Figures 4.3.1 shows the cyclic voltammograms of 10 [CoIII(LTMS)2]1-, obtained in

dichloromethane solutions with 0.10 M [N(n-Bu)4]PF6 as the supporting electrolyte using a

glassy carbon working electrode and Ag/AgNO3 reference electrode. Ferrocene was used as

an internal standard. The potentials are referenced versus the ferrocenium/ferrocene couple

(Fc+/Fc) and summarized in Table 4.3.1.

Coulometric studies established that all redox processes correspond to one-electron

transfer reactions. The wave corresponding to the reduction at E½ = –1.402 V is reversible

and this process yields the dianion 10b [CoII(LTMS)2]2- shown in equation 4.3.1. The oxidation

waves at –0.230 and –0.602 V have a greater separation (> 0.5 V). However, they belong to

the same oxidation process, and this was confirmed by the absence of a reductive wave when

the scan was stopped and reversed before the onset of the oxidation wave. The large wave

separation is scan-rate dependent and increases with faster scan rates. Coulometric one

electron oxidation at +0.3 V (-25 °C) was possible, and after the completion of the

coulommetry the same cyclic voltammogram was observed, demonstrating that the pair of

waves represent a chemically reversible one-electron process. Surprisingly, the resultant

oxidized species is EPR silent, probably due to the dimerization of the initial oxidized

compound 10a [CoIII(LTMS)(LTMS•)]0 to complex 10c [CoIII(LTMS)(LTMS•)]2, shown in Equation

Chapter 4

90

4.3.1. The re-reduction wave at -0.07 V represents the two-electron reduction of 10c to 10.

The spectrum after the one-electron oxidation i.e. of the dimer 10c (blue line) is shown in

Figure 4.3.2.

Figure 4.3.1 – Cyclic voltammogram of 10 [Co(LTMS)2]1- in CH2Cl2 solution at 25 °C

containing 0.10 M [N(n-Bu)4]PF6 as supporting electrolyte and scan rate of 100 mV/s.

(Conditions: glassy carbon electrode; potentials referenced vs the ferrocenium/ferrocene

couple).

[CoIII(LTMS)(LTMS•)]0 [CoIII(LTMS)2]1-

-e

+e

S=½

[CoIII(LTMS)(LTMS•)]2

S=1

[CoII(LTMS)2]2-

S=½

S=0

10c

-e

+e

0.5 0.0 -0.5 -1.0 -1.5 -2.0

E (V) versus Fc+/Fc

5 µµµµA

10 10a 10b

Eqn. 4.3.1

Chapter 4

91

The electronic spectra recorded during the stepwise one-electron oxidation of the monoanion

10 show a well defined isosbestic point, indicating the presence of only two species, assigned

as the oxidized complex and the starting material (Figure 4.3.2). Compound 10 shows four

bands at 327, 370, 594 (sh.) and 668 nm. The two bands at the UV region are also present in

the free (LTMS)2- ligand and hence must originate from intra-ligand π→π* transitions. The

absence of any intervalence charge transfer band in the near infrared region, unlike 4a

[AuIII(LTMS)(LTMS•)]0 but similar that of 4 [AuIII(LTMS)2]1- and 3 [CuIII(LTMS)2]

1-, provides

evidence for the assignment of 10 as [CoIII(LTMS)2]1-. The band at 668 nm in the spectrum of

the reduced dianion 10b [CoII(LTMS)2]2- is absent (Figure 4.3.3). Only a residual band in this

region remains due to experimental conditions. New bands at 433 and 470 nm are also

observed for 10b.

Figure 4.3.2 – Absorption spectra of the coulometric one-electron oxidation of complex 10

[CoIII(LTMS)2]1- denoted in red. Changes on the spectrum are indicated by the arrows. The blue

spectrum represents the dimer 10c.

300 400 500 600 700 800 900 1000 11000.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

ε,

ε,

ε,

ε, 1

04

M-1

cm

-1

λ, λ, λ, λ, nm

Chapter 4

92

Figure 4.3.3 – Absorption spectra of the coulometric one-electron reduction of complex 10

[CoIII(LTMS)2]1- denoted in red. Changes on the spectrum are indicated by the arrows. The blue

spectrum represents the reduced form 10b.

Coulometric studies on 11 [RhII(LTMS)2]2- showed a reversible one-electron reduction

at -1.207 V and an irreversible oxidation at -0.357 V between 25 °C and -75 °C (Figure 4.3.4).

The reversible feature showed a peak separation of 68 mV, which was found to be similar to

the peak separation of 69 mV observed for the ferrocene/ferrocenium couple. The chemically

irreversible oxidation of 11 yields the monoanion 11a as shown in equation 4.3.2, the

reversible reduction produces an trianion 11b. As showed below both processes are

predominantly metal based (Equation 4.3.2).

300 400 500 600 700 800 900 10000.0

0.5

1.0

1.5

2.0ε,

ε,

ε,

ε,

10

4 M

-1 c

m-1

λ, λ, λ, λ, nm

Chapter 4

93

Figure 4.3.4 – Cyclic voltammogram of 11 [RhII(LTMS)2]2- at 25 °C in CH2Cl2 solution

containing 0.10 M [N(n-Bu)4]PF6 as supporting electrolyte and scan rate of 100 mV/s.

(Conditions: glassy carbon electrode; potentials referenced vs the ferrocenium/ferrocene

couple).

Table 4.3.1 – Redox potentials (E2½) and peak potentials (E1

p) of 10 [Co(LTMS)2]1- and 11

[Rh(LTMS)2]2- in CH2Cl2 solution 0.10 M [N(n-Bu)4]PF6 at 25 °C.

Complex E1p, V vs Fc+/Fc E2

½, V vs Fc+/Fc

10 -0.074 and -0.602 -1.402 (reversible)

11 -0.357 (irreversible) -1.207 (reversible)

The absorption spectrum of 11 [RhII(LTMS)2]2- (Figure 4.3.5) shows an intense transition at

752 nm which loses intensity when complex 11 is reduced to yield 11b [RhI(LTMS)2]3-. This

unusual behavior will be explained in more detail in the DFT section 4.6. Table 4.3.2

summarizes the observed transitions of these complexes.

Eqn. 4.3.2

[RhIII(LTMS)2]1- [RhII(LTMS)2]

2-

S = ½ S = 0 S = 0 -e

+e

-e

+e 11 11a 11b

[RhI(LTMS)2]3-

-0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6

E (V) versus Fc+/Fc

5 µµµµA

Chapter 4

94

Figure 4.3.5 – Absorption spectra of the coulometric one-electron reduction of 11

[RhII(LTMS)2]1- denoted in red. Changes on the spectrum are indicated by the arrows. The blue

spectrum represents the reduced trianion 11b.

Table 4.3.2 – Summary of the electronic spectra of the complexes in CH2Cl2 containing 0.10

M [N(n-Bu)4]PF6 solution at -25 °C.

Complex λ max., nm (ε, 104 M-1 cm-1)

10 328 (1.39), 370 (1.53), 534 (sh. 0.26), 594 (sh. 0.63), 630 (0.70), 668 (1.25)

10a 324 (1.67), 367 (sh. 1.71), 661 (0.73), 981 (0.20)

10b 316 (2.00), 342 (1.56), 381 (1.17), 433 (0.35), 470 (0.22)

11 388 (1.19), 500 (0.24), 752 (2.32)

11b 352 (1.12), 383 (1.02), 493 (0.36), 609 (0.56), 753 (1.20)

300 400 500 600 700 800 900 1000 11000.0

0.5

1.0

1.5

2.0

2.5

ε ε ε ε 1

04

M-1

cm

-1

λ,λ,λ,λ, nm

Chapter 4

95

4.4 – Magnetic Properties:

Figure 4.4.1 shows the magnetic behavior of [N(n-Bu)4][10] in the temperature range

of 2 to 290 K with an applied external field of 1T. The magnetic properties of the ground

state of a paramagnetic ion in a molecule can be described by a Hamiltonian proposed by

Abragam and Pryce33 as shown in Equation 4.4.1. The Hamiltonian considers the interaction

of the ion with the surrounding ligands.

Ĥ = gµBBS+ D[Sz2-S(S+1)/3]+E/D[Sx

2-Sy2] Eqn. 4.4.1

Where B is the applied magnetic field, g is the g-tensor, S the electronic spin, and D and E are

parameters which describe the effects of axial and rhombic ligand fields, respectively.

Figure 4.4.1 – Temperature dependence of the effective magnetic moment of complex

[N(n-Bu)4][10] (4-290 K) measured with an applied field of 1T.

Above 50 K the complex shows a temperature independent µeff value of 2.84 ± 0.01 µB,

indicating the presence of two unpaired electrons (S = 1 ground state) in the complex.

However, at temperatures lower than 50 K the µeff value decreases monotonically with the

temperature reaching a value of around 0.85 µB at 4 K. According to the X-ray diffraction

data, intermolecular Co•••Co and Co•••S distances (> 8 Å) in crystals of [N(n-Bu)4][10] are

too long for any kind of exchange coupling mechanisms involving Co•••S interactions in the

solid state. This behaviour (for S = 1 systems) can be explained by large zero field splitting

which splits the S = 1 ground state to Ms = 0 and Ms = ± 1 levels. The increasing population

0 50 100 150 200 250 300

1.0

1.5

2.0

2.5

3.0

µeff

/µB

Temperature, K

■ Experimental

Calculated with:

S = 1

g = 2.094

D = 35 cm-1

Chapter 4

96

of the Ms = 0 level is favoured at lower temperatures. The experimental magnetic

susceptibility data could be fitted with a zero field splitting value of |D| = 35 cm-1 and an

isotropic g value of 2.094 which are quite similar to the parameters obtained by van der Put et

al.9 for the corresponding [Co(LMe)2]

1- complex (g = 2.17; D = 34 cm-1) and by Wieghardt et

al.13 for the [Co(LBu)2]

1- species (g = 2.168; D = 32 cm-1). The sign of D can be determined

from the multi- field and temperature measurements and the spin Hamiltonian simulations of

the experimental data. This study has been already performed for [Co(LBu)2]1-

(D = + 32 cm-1). Far-infrared and magnetic circular dichroism (MCD) measurements on

[Co(LBu)2]1- were also conducted to observe the intratriplet transition between the Ms = 0 and

Ms = ± 1 components of the triplet ground state. The D values obtained by these techniques

for similar systems are consistent with other magnetic susceptibility results.13

According to the cyclic voltammogram of 10 [CoIII(LTMS)2]1- a precise electrode

potential for oxidation of the complex could not be obtained due to the large peak separation.

However, the potential was more negative than that of the ferrocenium/ferrocene couple and

the complex could be oxidized by the ferrocenium hexafluorophosphate salt. A stoichiometric

amount of the solid ferrocenium salt was added under argon to an EPR tube containing a

solution of compound 10 in CH2Cl2. The resultant solution was quickly frozen at 77 K to

avoid the dimerization (or decomposition) processes observed during the electrochemical

oxidation. The X-band EPR spectrum recorded at 10 K confirms the S = ½ ground state of

10a and shows hyperfine coupling with 59Co (I = 7/2, 100%, Figure 4.4.2). The g values in the

spectrum are very close to 2.0 and indicate that the unpaired electron must be located on the

ligand, which in turn requires a low spin state for the CoIII ion. Thus, the oxidation process is

followed by a change in spin state of the cobalt center from an intermediate to a low spin

state. The simulated parameters of the one-electron oxidized form of 10 [CoIII(LTMS)2]1-are

very similar to the other known Co(III) low spin complexes containing ligand π-radical, and it

would be expected an octahedral geometry around the central cobalt ion accoriding to Crystal

Field Theory considerations. Probably the change in 10 from intermediate to low spin after

one-electron oxidation is followed by a change in geometry from square-planar to octahedral,

whereby two water molecules (from the solvent) can occupy the axial coordination sites of the

octahedron forming [CoIII(LTMS•)(LTMS)(H2O)2]. Due to the instability of the oxidized species

the isolation of the proposed [CoIII(LTMS•)(LTMS)(H2O)2] complex was not possible.

Chapter 4

97

Figure 4.4.2 – X-band EPR spectrum of the presumed complex 10a in CH2Cl2 solution at 10

K. Conditions: frequency 9.45 GHz; modulation amplitude 1 G; power 2.52 x 10-6 mW.

Simulation parameters (g1 = 2.050; g2 = 2.032; g3 = 2.005; line widths Wx = 50.2 Hz;

Wy = 46.9 Hz; Wz = 67.1 Hz, Hyperfine coupling constants given in 10-4 cm-1: A1 = 7.31,

A2 = 54.50, A3 = -1.33). Black line represents the experimental spectrum and the red

corresponds to the simulation.

The EPR spectrum shown in Figure 4.4.2 shows similarities with a reported cobalt

complex [CoIII(L3•)]1+ ((L3•)1- = substituted triazacyclononane containing a thiyl radical).34

The giso of 2.022 and Aiso of 10.7 x 10-4 cm-1 observed for this compound is similar to that of

10 (giso = 2.023, Aiso = 21.0 x 10-4 cm-1), indicating that in both cases the unpaired electron is

located on the ligand.

The magnetic susceptibility data of [N(n-Bu)4]2[11]•4MeCN in the temperature range

of 2-290 K with an external applied field of 1 T showed an effective magnetic moment of

1.76 ± 0.01 µB, thus complex 11 possesses one unpaired electron (S = ½), suggesting the

presence of a RhII ion (d7) (Figure 4.4.3).

300 310 320 330 340 350 360

-0.8

-0.4

0.0

0.4

0.8

1.2

2.25 2.2 2.15 2.1 2.05 2 1.95 1.9

dχχ χχ´´

/ d

B

B (mT)

g values

Chapter 4

98

Figure 4.4.3 – Temperature dependence of the effective magnetic moment of complex

[N(n-Bu)4]2[11]•4MeCN (4-300 K) measured with an applied field of 1T.

The doublet ground state of 11 [RhII(LTMS)2]2- was confirmed by the X-band EPR spectrum of

a frozen CH2Cl2 solution at 25 K (Figure 4.4.4). The EPR spectrum of 11 shows a rhombic

signal with large g anisotropy. No hyperfine coupling with 103Rh (I = ½, 100% natural

abundance) was observed. Very few examples of EPR spectra are found in the literature for

monomeric RhII complexes.

0 50 100 150 200 250 3001.0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

µµ µµe

ff/ µµ µµ

B

Temperature, K

■ Experimental

Calculated with:

S = ½

g = 2.046

θ-Weiss = -0.83 K

TIP = 646.3 x 10-6

emu

Chapter 4

99

Figure 4.4.4 – X-band EPR spectrum 11 in frozen CH2Cl2 solution at 25 K Conditions:

frequency 9.43 GHz; modulation 5 G; power 2.007 x 10-4 mW. For simulation parameters

(g1 = 2.4997; g2 = 2.0052; g3 = 1.9738; line widths Wx = 50.2 Hz; Wy = 46.9 Hz;

Wz = 67.1 Hz). Black line represents the experimental spectrum and the red corresponds to

the simulation.

X-band EPR studies were performed by Holm et al.27 on single crystals of the square

planar [N(n-Bu)4]2[Rh(mnt)2] complex diluted with [N(n-Bu)4]2[Ni(mnt)2]. A few months

later Gray et al.26 reported EPR data of polycrystalline samples of the same compound. As

shown in Table 4.4.1 there is a significant difference in the maximum principal g-value

between the data reported by Gray (2.35) and that of Holm (2.447). Normally, paramagnetic

resonance data of magnetically concentrated crystals may be suspect because of the possible

averaging of g-values of magnetically inequivalent sites by electron spin exchange. Such

averaging occurs in crystals of Cu(NH3)4SO4•H2O, for instance, containing two magnetically

inequivalent sites in the unit cell.35 Compared to the crystal packing of the isomorphous

[N(n-Bu)4]2[11] only one square-planar molecule per unit cell is observed, so averaging of

principal g-values by spin exchange is not expected to occur, even with a considerable spin

exchange rate.27

The organometallic compound [N(n-Bu)4]2[Rh(C6Cl5)4] shows an EPR spectrum with

g values of 2.74, 2.60 and 1.94 with no hyperfine coupling to the 103Rh nucleus. The unpaired

260 280 300 320 340 360

-1.0

-0.5

0.0

0.5

1.0

2.6 2.5 2.4 2.3 2.2 2.1 2 1.9

dχχ χχ´´

/ d

B

B (mT)

g values

Chapter 4

100

electron is mainly in a dz2 orbital with the z axes being perpendicular to the first coordination

plane of the rhodium atom. Conversely, the neutral compound trans-

[RhII(2,4,6-Pri3C6H2)2(tht)2] shows a remarkable rhombic EPR spectrum with large g

anisotropy and well-resolved hyperfine coupling.36 An example where large hyperfine

coupling constants (> 60 G) are observed is in the complex36 [Rh(TMPP)2(CNtBu)2]2+. The

pathway for this feature is the Fermi contact contributions where the inner s orbitals are

polarized, transferring spin density closer to the nucleus. The EPR spectrum of compound 11

[RhII(LTMS)2]2- shows some similar features observed in that of [RhII(tBu2-boxate)2],

((tBu2-boxate)1- = di-tert-butyl(bisoxazolinate))37 which indicates that in both cases the

unpaired electron is located in the same orbital.

Table 4.4.1 – X band EPR data of selected monomeric RhII square planar complexes.

Complex T(K) Ref. Sample prep. g1 g2 g3 A1a

A2a A3

a

[Rh(mnt)2]2- b 77 26 polycrystalline 2.35 2.015 1.950 - - -

[Rh(mnt)2]2- rt 27 single crystal 2.447 2.019 1.936 < 0.4 0.75 < 0.4

[Rh(C6Cl5)4]2- c rt 38 powder 2.74 2.60 1.94 - - -

11 25 - frozen solution 2.499 2.005 1.973 - - -

[RhII(tBu2-boxate)2] d 20 37 frozen solution 2.794 2.016 1.947 - - -

[Rh(TMPP)2(CNtBu)2]2+ e 100 36 polycrystalline 2.45 2.45 1.96 66 66 62

Rh(2,4,6-iPr3C6H2)2(tht)2 f 77 39 frozen solution 2.96 2.58 1.85 4.7 5.6 5.6

a A values for hyperfine coupling constants are given in 10-4 cm-1. b mnt2-:

maleonitriledithiolate. c The EPR spectra of a series of compounds containing

pentachlorophenyl derivatives have also been reported.40 d (tBu2-boxate)1- = di-tert-

butyl(bisoxazolinate)) e (TMPP): tris(2,4,6-tri-metoxyphenyl)phosphine; (CNtBu):

butyronitrile. f (2,4,6-iPr3C6H2): 2,4,6-triisopropylphenyl; (tht): tetrahydrothiophene.

4.5 – Preliminary Reactivity Studies:

Attempts to isolate the reduced trianion 11b [RhI(LTMS)2]3- in its crystalline form were

not successful. Reactions of 11 [RhII(LTMS)2]2- with Na/Hg amalgam in MeCN lead to a violet

solution. The absorption spectrum of this violet solution is similar to that of the one-electron

reduced species obtained after coulommetry. Some preliminary reactivity studies were

performed with complex 11b in reactions of oxidative addition to the metal with methyliodide

(MeI), shown in Equation 4.5.1.

[RhI(LTMS)2]3- + CH3I → [RhIIICH3I(L

TMS)2]3- Eqn. 4.5.1

Chapter 4

101

After reduction with Na/Hg amalgam the solution was filtered and MeI was added, resulting

in an instantaneous color change to orange. Interestingly, the same compound was obtained

upon addition of a stoichiometric amount and an excess of MeI. The sulfur atoms in 11b

[RhI(LTMS)2]3- are highly nucleophilic, and the crystallized complex was a mixture containing

70% of 12a cis-{RhIIII2[LTMS(CH3)2][LTMS(CH3)]} and 30% of 12b cis-

{RhIIICH3I[LTMS(CH3)2][LTMS(CH3)]} as shown in the crystal structure in Figure 4.5.1. The

selected bond lengths are given in Table 4.5.1. The separation of compounds 12a and 12b by

liquid chromatography with different stationary and mobile phases was not successful. When

the trianion 11b [RhI(LTMS)2]3- reacts with a stoichiometric amount of MeI, the mixture 12a +

12b is obtained in 15% and in reactions with an excess greater than 5 equivalents, better

yields are obtained (80%). According to the yields obtained independently of the

stoichiometry, we suspect that compounds 12a and 12b are not the only products formed, but

are the most insoluble products that crystallize from MeCN solutions after 3 days at -20°C.

The sequence of the oxidative addition to the metal or to the sulfurs remains unclear, as does

the step where the ligands twist in order to adopt a cis conformation. It is probable that the

MeI binds to the rhodium in a side-on fashion with posterior CH3–I heterolytic cleavage of the

bond between the methyl and the iodide. Three of the four sulfurs in 11b [RhI(LTMS)2]3- were

methylated through an SN2 reaction. A square planar geometry and RhI in 11b is required in

order to obtain the mixture of the two-electron oxidized complexes 12a and 12b. Studies with

dimeric tetraazoporphyrin derivatives of RhII, namely [(OETAP)Rh]2 (OETAP2- =

octaethyltetraazaporphyrinato(2-)), reveals that the reaction with MeI yields exclusively the

one-electron oxidized (OETAP)RhIII–I and (OETAP)RhIII–CH3 species in a 1:1 ratio.41 The

results obtained for [(OETAP)Rh]2 and 11 with MeI indicate different mechanisms, and more

than one pathway for the resulting mixture 12a + 12b can be proposed.

Chapter 4

102

Figure 4.5.1 – Perspective view and numbering scheme of the mixture of 12a and 12b with

thermal ellipsoids at 50% level. Hydrogen atoms are omitted for clarity, except for the methyl

group coordinated to the rhodium center.

Table 4.5.1 – Selected bond lengths of compounds 12a and 12b.

Selected bonds Bond lengths (Å) Selected bonds Bond lengths (Å)

Rh(1)-C(1) 2.057(6) S(1)-C(5) 1.789(1)

Rh(1)-I(1) 2.6697(2) S(2)-C(3) 1.811(1)

Rh(1)-I(2) 2.7154(2) S(2)-C(6) 1.793(1)

Rh(1)-S(1) 2.3632(4) S(3)-C(7) 1.766(1)

Rh(1)-S(2) 2.3136(4) S(4)-C(4) 1.821(1)

Rh(1)-S(3) 2.3409(4) S(4)-C(8) 1.790(1)

Rh(1)-S(4) 2.3187(4) C(5)-C(6) 1.398(1)

S(1)-C(2) 1.818(2) C(7)-C(8) 1.404(2)

H

C(3)

C(1)

C(5)

C(2)

C(6)

Si S(2)

S(1)

Rh(1)

S(3)

S(4)

I(2)

I(1) Si

Si

Si

C(4)

H H

H

C(7)

C(8)

Chapter 4

103

4.6 – Theoretical calculations:

DFT calculations have been carried out at the B3LYP level for 10 [CoIII(LTMS)2]1-, 11

[RhII(LTMS)2]2- and its reduced counterpart 11b [RhI(LTMS)2]

3-. Scalar relativistic corrections

have been taken into consideration for compound 11 and 11b, using the ZORA method (for

calculation of properties) with large uncontracted basis sets.

Structure Optimization:

The optimized geometry calculations of 10 and 11 are in a good agreement with the

experimental results obtained by X-ray crystallography (Table 4.6.1). The small

overestimation of the M–S bond distances is typical for the B3LYP DFT functional.42-45

However, the metrical parameters for the dithiolate ligand were accurately reproduced by the

calculations to within ~0.02 Å.

Table 4.6.1 – Experimental and calculated (in parentheses) metrical parameters in Å.

Complex M-S C-S C1-C2 C2-C3 C3-C4 C4-C5 C5-C6 C6-C1

10 [Co(LTMS)2]1- 2.166

(2.213) 1.772

(1.781) 1.408

(1.416) 1.416

(1.408) 1.402

(1.406) 1.395

(1.398) 1.404

(1.406) 1.424

(1.421)

10b [CoII(LTMS)2]2-

- (2.234)

- (1.782)

- (1.426)

- (1.424)

- (1.408)

- (1.400)

- (1.408)

- (1.423)

11 [RhII(LTMS)2]2- 2.271

(2.296) 1.755

(1.768) 1.414

(1.431) 1.418

(1.423) 1.393

(1.407) 1.396

(1.403) 1.396

(1.407) 1.417

(1.423)

11b [RhI(LTMS)2]3- -

(2.307) -

(1.781) -

(1.434) -

(1.424) -

(1.413) -

(1.401) -

(1.413) -

(1.424)

S

M

S

S

S

TMS

TMSTMS

TMS

1

23

4

5

6

Chapter 4

104

Bonding Scheme and Ground State Properties:

Qualitative bonding schemes derived from the unrestricted B3LYP DFT calculation of

10 [CoIII(LTMS)2]1- and 10b [CoII(LTMS)2]

2- are shown in Figures 4.6.1 and 4.6.2 respectively,

wherein the spin up and the spin down MOs are shown in order of decreasing energy. The

calculated 3B1g ground state is found to be in agreement with the results of the scalar

relativistic ZORA-B3LYP calculations on [Co(L)2]1-.13

The bond scheme in Figure 4.6.1 shows that the dz2 and the dx

2-y

2 orbitals are low in

energy and mix to form one pair of molecular orbital with ag symmetry. Additionally, four

doubly occupied orbitals with mainly ligand character and two singly occupied orbitals with

2b2g and 2b3g symmetries are found. The bonding scheme of the corresponding complex

[Co(1LN)2]1- (where (1LN) = ortho-phenylenediamine) compound was recently described in

detail.46 The electronic structure of this complex was difficult to rationalize, even

qualitatively, as the out-of-plane orbital 2b2g was revealed to have almost equal contribution

from the Co 3dxz orbital and the b2g fragment orbital of the ligands. The assignment of the

spectroscopic oxidation state of the cobalt ion was therefore not straightforward. In fact, two

possible electronic structures were proposed: 1) that the cobalt ion contains a d6 configuration

with a CoIII intermediate spin (S = 1) ion; or 2) that the 2b2g orbital possesses pure ligand

character, resulting in a CoII low-spin with d7 configuration coupled ferromagnetically with a

(1LNISQ)•- ligand π-radical (where (1LN

ISQ)•- = ortho-diiminobenzosemiquinonate(1-)).

Chapter 4

105

Figure 4.6.1 – Unrestricted Kohn-Shan MOs and energy scheme from B3LYP DFT

calculations of the monoanion 10 [CoIII(LTMS)2]1-.

S

Co

S

S

S

TMS

TMSTMS

TMS

Y

X

Spin up Spin down

1ag (dx2-y

2 + dz2)

1au

1b1u

1b3g

2b2g (dxz + L)

2b3g (dyz)

1b1g (dxy + L)

1b2g

2ag (dx2-y

2 + dz2)

-7.5

-7.0

-6.5

-6.0

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

En

erg

y,

eV

1ag

2b2g

2b3g

2ag

1b3g

1au

1b1u

1b2g

1b1g

1ag (dx2-y

2 + dz2)

1au

1b1u

1b3g

2b2g (dxz + L)

2b3g (dyz)

1b1g (dxy + L)

1b2g

2ag (dx2-y

2 + dz2)

-7.5

-7.0

-6.5

-6.0

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

En

erg

y,

eV

-7.5

-7.0

-6.5

-6.0

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

En

erg

y,

eV

1ag

2b2g

2b3g

2ag

1b3g

1au

1b1u

1b2g

1b1g

Chapter 4

106

Figure 4.6.2 – Unrestricted Kohn-Shan MOs and energy scheme from B3LYP DFT

calculations of the dianion 10b [CoII(LTMS)2]2-.

Spin up Spin down

S

Co

S

S

S

TMS

TMSTMS

TMS

Y

X

1ag (dx2-y

2)

1au

1b1u

1b2g

2ag (dz2)

1b3g (dyz + L)

2b2g

2b1u

3b2g (dxz)

1b1g (dxy + L)

-6.0

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

En

erg

y,eV

1b1g

2b1u

2b2g

1b2g

1b3g

2ag

1b1u

1au

1ag

3b2g 1ag (dx2-y

2)

1au

1b1u

1b2g

2ag (dz2)

1b3g (dyz + L)

2b2g

2b1u

3b2g (dxz)

1b1g (dxy + L)

-6.0

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

En

erg

y,eV

-6.0

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

En

erg

y,eV

1b1g

2b1u

2b2g

1b2g

1b3g

2ag

1b1u

1au

1ag

3b2g

Chapter 4

107

According to Table 4.6.2, the 2b2g and 2b3g orbitals of 10 [CoIII(LTMS)2]1- are

composed of 71 and 82% of Co 3dxz and Co 3dyz with 29 and 17% of ligand contribution,

respectively. The natural population analysis47-49 presented in Table 4.6.3 was obtained from

the B3LYP calculations and shows a d-population of 7.58 and a spin density of 1.78 at the

central cobalt ion. The excess over the formal d6 electron configuration arises from the

covalent population of the otherwise unpopulated Co dxy orbital due to strong σ-donation from

the ligand.13,50 However, the monoanion 10 appears to be more reduced than a typical CoIII

ion and more oxidized than a typical CoII ion. Taking these observations into consideration,

the electronic structure of compound 10 is ambiguous. Thus, the actual electronic structure

can be better represented by the following resonance forms [CoIII(LTMS)2]1- ↔

[CoII(LTMS•)(LTMS)]1- ↔ [CoII(LTMS)(LTMS•)]1- with a somewhat larger weight for the first

resonance structure. However, by calculation the CoIII character is more pronounced in

[Co(L)2]1- than in [Co(1LN)2]

1- as is evident from the enhanced metal character of the 2b2g

SOMO and larger spin density at the Co ion in [Co(LTMS)2]1-.

It is interesting to compare the electronic structure of 10 [CoIII(LTMS)2]1- with the

isoelectronic [Fe(L)2]2-, which exhibits the same 3B1g ground state. Despite this formally

identical ground-state electron configuration, there is an important difference in the

qualitative bonding schemes. In [Fe(L)2]2- the 2b3g SOMO is predominantly metal-centered

(82% metal 3dxz) in contrast with 10 (only 71% metal 3dxz). The predominance of the metal

character in the 2b2g level of [Fe(L)2]2- arises from the lower effective nuclear charge (Zeff) of

Fe as compared to Co, which raises the Fe 3d levels in energy and makes them less available

for back-bonding interactions with the ligand orbitals. The upper valence region of the iron

complex is therefore composed of two doubly occupied and two singly occupied orbitals that

are predominantly centered on the iron. The electronic structure of the complex can be

understood in terms of an intermediate spin (d6, S = 1) ferrous ion coordinated to two closed-

shell ortho-benzenedithiolate(2-) ligands.13

Chapter 4

108

Table 4.6.2 – Composition of selected molecular orbitals of [M(L)2]z complexes (%) as

obtained from the B3LYP DFT calculations. * Values taken from ref. 13.

Complex MO (dyz) (dxz) (dxy) S(pz) S(px, y) C(pz)

10 [Co(LTMS)2]1- 2b3g 82 7 10

2b2g

71 22 7

*[Co(L)2]1- 2b3g 76 10 10

2b2g

65 28 08

10b [Co(LTMS)2]2- 1b3g 72 18 9

3b2g 71 10 12 2b2g 19 3 69

*[Fe(L)2]2- 2b3g 90 3 2

2b2g 82 8 6

11 [Rh(LTMS)2]2- 1au 28 26 39

1b1u 47 42 2b2g

61 26 10

11b [Rh(LTMS)2]3- 1b3g 71 15 13

1b2g 58 30 11 1au 9 87

Table 4.6.3 – Comparison of the charge and spin populations at the metal ion resulting from a

natural population analysis of the one-electron density of the ground state obtained from

scalar relativistic ZORA-B3LYP DFT calculations.

Compound nd electrons (n+1)s electrons nd spin metal oxidn state

10 7.58 0.47 1.78 see text

[Co(L)2]1-

13

7.82 0.51 1.59 see text

[Fe(L)2]2-

13

7.11 0.49 1.92 FeII

10b 8.34 0.46 1.01 CoII

11 8.31 0.50 0.66 RhII

11b 9.21 0.59 0.00 RhI

Chapter 4

109

When compound 10 is reduced to 10b [Co(LTMS)2]2-, the extra electron populates the

1b3g molecular orbital which has 72% Co 3dyz and 27% ligand character. The 1b3g orbital

decreases considerably in energy compared to 10 and the 3b2g remains the SOMO, having

71% of Co 3dxz and 22 % ligand contribution. Thus, the reduction process is metal centered,

yielding 10b which can be described as a CoII (d7) ion. The population of the molecular

orbital 1b3g with an extra electron results in the destabilization of the 2ag orbital (mainly dz2)

which remains higher in energy, with a clear separation of the 1ag (dx2-y

2) orbital.

The dianion 11 [RhII(LTMS)2]2- shows a similar electronic structure to that of 10b and

the same 2B2g ground state (S = ½) shown in Figure 4.6.3. The unpaired electron resides in a

b2g orbital in both compounds 10b and 11, showing small differences in their orbital

contributions. In Table 4.6.2, the SOMO 2b2g of compound 11 possesses 61% Rh 4dxz and

36% ligand character. The 3b2g orbital in the isoelectronic complex 10b has 71% of metal and

22% of ligand contributions. The spin density on the d orbitals for 10b and 11 are 1.01 and

0.66, respectively, which indicates some RhI character to 11 due to high covalency.

The most notable changes in the electronic structures are observed when complex 11 is

reduced by one electron yielding the diamagnetic species 11b. Figure 4.6.4 shows the

molecular orbital scheme of the trianion [Rh(LTMS)2]3-. The d orbitals in 11b [RhI(LTMS)2]

3-

are higher in energy compared to that of 11 due to the lower effective nuclear charge of RhI.

Upon oxidation, an extra electron populates the previously singly occupied 2b2g orbital in 11.

This orbital in 11b is composed of an almost equal contribution of metal and ligand. In fact

the HOMO shows 58% of Rh 4dxz and 41% of ligand character and confers on the

diamagnetic compound the 1B2g ground state.

Chapter 4

110

Figure 4.6.3 – Unrestricted Kohn-Shan MOs and energy scheme from B3LYP DFT

calculations of the dianion 11 [RhII(LTMS)2]2-.

S

Rh

S

S

S

TMS

TMSTMS

TMS

Y

X

Spin up Spin down

1ag (dx2-y

2)

1b2g 1ag

1au

1au

1b2g

1b1u

1b1u

2ag

2ag (dz2)

1b3g (dyz + L)

1b3g

2b2g (dxz + L)

2b2g

2b1u

3b2g

3b2g

1b1g (dxy + L)1b1g

2b1u

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

En

erg

y,

eV

1ag (dx2-y

2)

1b2g 1ag

1au

1au

1b2g

1b1u

1b1u

2ag

2ag (dz2)

1b3g (dyz + L)

1b3g

2b2g (dxz + L)

2b2g

2b1u

3b2g

3b2g

1b1g (dxy + L)1b1g

2b1u

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

En

erg

y,

eV

1ag

1au

1au

1b2g

1b1u

1b1u

2ag

2ag (dz2)

1b3g (dyz + L)

1b3g

2b2g (dxz + L)

2b2g

2b1u

3b2g

3b2g

1b1g (dxy + L)1b1g

2b1u

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

En

erg

y,

eV

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

En

erg

y,

eV

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

En

erg

y,

eV

-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

En

erg

y,

eV

Chapter 4

111

Figure 4.6.4 – Restricted Kohn-Shan MOs and energy scheme from B3LYP DFT calculations

of the trianion 11b [RhI(LTMS)2]3-.

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Energ

y,eV

1b1u

1ag (dx2-y

2)

2ag (dz2)

1b3g (dyz)

2b2g (dxz + L)

2b1u

3b2g

1au

2b3g

1b1g (dxy)

1b1u

1ag

2ag

1b3g

1b2g

2b1u

2b2g

1au

2b3g

1b1g

1b2g1b2g

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Energ

y,eV

1b1u

1ag (dx2-y

2)

2ag (dz2)

1b3g (dyz)

2b2g (dxz + L)

2b1u

3b2g

1au

2b3g

1b1g (dxy)

1b1u

1ag

2ag

1b3g

1b2g

2b1u

2b2g

1au

2b3g

1b1g

1b2g1b2g

2

2

S

Rh

S

S

S

TMS

TMSTMS

TMS

Y

X

Chapter 4

112

The natural population analysis in Table 4.6.3 shows a total of 9.21 valence electrons

for the trianion 11b, which, excluding the σ-donation to the unpopulated ndxy, results in 8.1

electrons, which is in agreement with a RhI (d8) ion. For 11, the total number of valence

electrons without the σ-donation of dxy is 7.6. The natural population analysis corroborates to

the fact that 11 has some RhI character. Surprisingly, the LUMO in 11b is a π-antibonding

orbital with Cpz and Si pz contributions of 77% and 8%, respectively, and the next three

empty orbitals also have a high Cpz character. Most of the bonding and antibonding

combinations of the orbitals involving sulfur contributions are low in energy, compared to the

1b1u orbital represented in Figure 4.6.4. DFT calculations reported on a series of square-planar

[M(L)2]z complexes do not show those orbitals with high Cpz character, which are presumably

very much higher in energy.13,50-54 This feature is probably observed due to the low effective

nuclear charge of the RhI.

Spectroscopic Trends Based on DFT Calculations

Time-dependent DFT calculations (TD-DFT) for the [Co(L)2]1- species in CH2Cl2

solution and in vacuum have recently been reported.13 Compound 10 [CoIII(LTMS)2]1- shows a

very similar electronic structure to that of the [Co(L)2]1- species and the qualitative results

obtained for the calculation of [Co(L)2]1- corroborate the electronic structure of 10. Table

4.6.4 details the calculated and experimental results obtained for energy transitions of

[Co(L)2]1-.

Chapter 4

113

Table 4.6.4 – Analysis of the optical transitions for the complex [Co(L)2]1- of ref. 13.

Energy in cm-1 and in (nm)

Band Experimental Calculated Method assignment

I 14890 (670) 19050 (525) UV-Vis 3B1u (1b1u → 2b2g(dxz))

14300 (700) MCD

II 15900 (630) 19342 (517) UV-Vis 3B3u (1au → 2b2g(dxz))

15000 (667) MCD

III 16804 (595) 19560 (511) UV-Vis 3B3u (1b1u → 2b3g(dyz))

16007 (625) MCD

IV 18983 (527) 20200 (495) UV-Vis 3B2u (1au → 2b3g(dyz))

18726 (534) MCD

Four LMCT states with reasonable intensities were calculated for [Co(L)2]1- in the range

20000–12000 cm-1 (500–830 nm), which were in agreement with the experimental MCD and

absorption spectra.

I – The most intense band arises from the 1b1u → 2b2g transition, which corresponds to a

LMCT transition of 3B2u symmetry and is allowed in the X-polarization. This band

contributes to the observed blue color of the complex.

II – The second calculated transition is the spin and dipole allowed 1au → 2b2g

(3B3u symmetry, Y-polarized).

III – The next transition has a 3B3u symmetry and is assigned as the 1b1u → 2b3g excitation.

The corresponding band to this transition can become considerable intense by mixing with the

relatively intense band described in II.

IV – The final transition below 22000 cm-1 corresponds to the electric dipole allowed

1au → 2b3g (3B2u symmetry, X-polarized) transition. Since it has 3B2u symmetry it can also

gain intensity by mixing with the intense LMCT 1b1u → 2b2g band, which has the same

symmetry.

The TD-DFT calculations thus lead to an acceptable agreement with the

experimentally observed absorption spectra of [Co(L)2]1- and compound 10. However, it

appears that the calculation underestimates the mixing of the intense and weak bands, thereby

leading to too-low intensities for the weak and too-large intensities for the strong bands.

Chapter 4

114

On the basis of the success of the ground-state calculations, time-dependent DFT

calculations (TD-DFT) for 11 [RhII(LTMS)2]2- were performed applying the COSMO solvent

corrections for CH2Cl2 to interpret the striking differences in the absorption spectra of 11

[RhII(LTMS)2]2- and 11b [RhI(LTMS)2]

3-. Table 4.6.5 shows the comparison of experimental and

calculated results.

Table 4.6.5 – Analysis of the intervalence charge transfer bands in the [Rh(LTMS)2]z

complexes (z = 2- or 3-) following Gaussian deconvolution of the experimental data of 11

(in CH2Cl2 solutions) combined with the results obtained from TD-DFT (COSMO)55

calculations at the scalar relativistic ZORA-B3LYP level using CH2Cl2 as the solvent.

Energy, cm-1 (nm) Oscillator strength (f)

Compd. expt. calcd expt. calcd. assignment

11 13238 (750)

18708 (534)

0.105 0.163 1b1u (L) → 2b2g (dxz + L)

LMCT

14200 (704)

13297 (477)

0.037 0.026 1au (L) → 2b2g (dxz + L)

LMCT

16300 (614)

20829 (480)

0.027 0.077 2b2g (dxz + L) → 2b1u (L)

MLCT

11b 13297 (752)

12103 (826)

0.048 0.106 2b2g (dxz + L) → 2b1u (L)

MLCT

16420 (609)

17785 (562)

0.041 0.022 2b2g (dyz) → 1au (L)

MLCT

The oscillator strength can be calculated from the experimental absorption spectra using

equation 4.6.1.56

( )9

0 Band

4.32 10( )dI

Ifn

ε ν ν−

→

⋅= ∫ % % ,

where f0→I is the oscillator strength of the transition from the ground state to the Ith electronic

excited state, n is the refractive index (for CH2Cl2 it is 1.4242), ε(I) represents the extinction

coefficient of the band, ∫ Band represents the integral of the band area and v~ is the transition

energy in cm-1.

eqn. 4.6.1

Chapter 4

115

Three LMCT states with reasonable intensities are calculated in the range of 25000 – 8000

cm-1 for complex 11. The experimental spectrum of complex 11 shows the envelope of all

three transitions. Figure 4.6.7 shows the deconvolution of the experimental absorption

spectrum of the near-infrared band.

I – In the calculation the most intense band arises from the 1b1u → 2b2g transition, which

corresponds to a LMCT yielding a 3B2u excited state. This transition is calculated at 18708

cm-1 (534 nm, oscillator strength fosc = 0.163) and is observed in the region of 13238 cm-1

(755nm).

II – The second calculated transition is the spin and electric dipole allowed 1au → 2b2g

(2B3u symmetry) at 20958 cm-1 (478 nm) with a calculated oscillator strength of 0.026.

III – The third transition corresponds to a 2b2g → 2b1u, calculated at 20829 and observed at

16300 cm-1.

Figure 4.6.7 – Deconvolution of the absorption spectrum of 11 between 17000 and 11000

cm-1 (590 – 910 nm) with the assignments described in the text (vide supra).

Overestimation of 4000 cm-1 in calculated band energies compared with the experimental

absorption spectrum values are acceptable and are also observed for complexes with (L)2-

ligands.13,50

11000 12000 13000 14000 15000 16000 170000.0

0.5

1.0

1.5

2.0

2.5

III

II

ε,

ε,

ε,

ε, 1

04 M

-1cm

-1

Energy, cm-1

I

Chapter 4

116

When compound 11 [RhII(LTMS)2]2- is chemically or electrochemically reduced to

yield the [RhI(LTMS)2]3- species 11b, the intense band at 752 nm of the starting dianion 11

surprisingly remains with 50% of its original intensity. The TD-DFT calculations show two

distinct transitions well separated in energy.

I – The first calculated transition corresponds to the transition 2b2g (HOMO) → 2b1u (LUMO)

at 12103 cm-1 with an oscillator strength of 0.106. This transition involves the orbital 2b2g

(which has almost equal metal and ligand contributions), and the 2b1u with major Cpz

character, which is observed experimentally at 13297 cm-1. This metal-to-ligand charge

transfer band (MLCT) has not been observed before for ortho-benzenedithiolate complexes.

II – A band with a small intensity for the 2b2g → 1au transition was calculated at 17785 cm-1

(562 nm) with an oscillator strength of 0.023, which involves orbitals with the same

characteristics as that in I. In the experimental absorption spectrum this band presumably

corresponds to that at 16420 cm-1.

It is important to note that transition I in 11b corresponds to that of III in 11. When 11 is

reduced, bands I and II disappear as the 2b2g-acceptor orbital is now filled. Band III doubles

its intensity due to alpha and beta transitions from the 1b2g-donor to the 2b1u-acceptor orbital.

Although the calculations were not able to reproduce the correct energies for the bands in 11

and 11b, the intensities are well reproduced. In all probability, the transitions coincidentally

have the same energy in the experimental absorption spectra of 11 and 11b.

The g-tensors of the dianion 11 were calculated on the basis of the geometry optimized

structure using the DFT methodology. The g values obtained were gz = 3.06, gy = 2.20 and

gx = 2.04. The large deviation from the experimental values is due to calculation of the

molecule in the vacuum, but the trend of a large gz and two relatively close tensors (gy and gx)

is observed. This is in agreement with the ground state of 11 shown in Figure 4.6.3 and the

assignment of a 2B2g.

Spin-orbit coupling (SOC) effects are known to be factors that cause g values to

deviate from 2.00. The spin-orbit interaction is a phenomenon whereby the relative motion of

the electron and the charged nucleus results in the electron being exposed to a local magnetic

field arising from the nuclear charge. The SOC interactions can be expressed as λL•S, where

L•S is the combination of the orbital quantum number of the electron and the spin orbital

momentum on its axes (λ = ζ/2S, S can be ±½). The λL term can be understood as the

intensity of this local magnetic field. The local magnetic field can be added or subtracted from

Chapter 4

117

the applied field thereby shifting the g value from 2.0023. The SOC constant ζ, has the

dimensions of cm-1 as shown in chapter 2, section 2.5. In an isolated d orbital there is no

angular momentum because there is no pathway for the unpaired electron to move around the

applied magnetic field. The EPR spectrum of 11 [RhII(LTMS)2]2- is an example where the SOC

effect is observed. SOC yields mixing low-lying excited states with the ground state. The dxz

orbital is related to the other four d orbitals by rotations about one or more of the Cartesian

axes under the influence of the x, y, and z components of the L operator. These rotational

relationships are summarized in Table 4.6.6 where the individual entries represent the

consequences of rotating the dxz about the axis identified in the top row.

Table 4.6.6 – Rotational relationships of the dxz orbital.

For rotations about

x y z

initial d orbital Final d orbital

dxz -idxy idx2-y

2 – i 3 dz

2 -idyz

Thus, the unpaired electron in dxz can exploit SOC in all three directions for rotation about all

three axes converts dxz into one of the other d orbitals. The corresponding mixing of the

resulting orbitals into the ground state by SOC partially restores the orbital momentum and

lead to deviations of gx, gy, and gz from 2.0023. There is a barrier to SOC mixing that is

provided by the energy separation between dxz and the orbital into which it is being rotated. In

the case of 11, this separation is provided by ∆x, ∆y, ∆z for rotation about x, y and z axes

respectively. If the rotation moves the electron into a vacant orbital, as is the case here about

the x axis, then the current resulting from the circulation leads to a magnetic field that opposes

the applied field. Consequently, a larger applied field is required to meet the Zeeman

condition and the calculated g value will be smaller than that of the free electron

(i.e., < 2.002). Conversely, if the SOMO interacts with an occupied orbital then the magnetic

field produced by the circulation adds to the applied field. Now a smaller applied field is

required to meet the Zeeman condition and the calculated g value will be greater than that of

the free electron (i.e., > 2.002).

From these concepts together with Table 4.6.6 and the anticipated ordering of d

orbitals for rhodium, as shown in Figure 4.6.8, it is possible to predict the qualitative form of

the EPR spectrum for a RhII d7 (S = ½) and in any other system with the same multiplicity.

Chapter 4

118

The pentagon represents all possible rotational combinations of orbitals about the three axes.

The numeric values represent the matrix element (the entries in Table 4.6.6).

Figure 4.6.8 – Relative order of the energy levels of the five 3d orbitals for the dianion 11.

The electron circulation associated with the orbital contributions to the three g values are

depicted by the loops. The contribution of a loop to the orbital magnetism is inversely related

to its amplitude but the identity of the participating orbitals is also important (Table 4.6.6).

The pentagon represents all possible contributions for all five d orbitals.

A more quantitative approach requires using a result from quantum mechanical perturbation

theory, shown in equation 4.6.2.

where i = x, y and z; ge is the value of the free-electron (2.0023), n is the factor shown in the

pentagon, and ∆i is the relevant energy separation. In the case of 11, the rotation of the SOMO

dxz about the x axis results in the delocalization of the electron in the empty dxy orbital,

resulting in g value lower than ge. Because of the large ∆xy the deviation of g is expected to be

gi = ge niλ

∆i

1 +

Eqn. 4.6.2

dx2- y

2

dz2

dyz

dxz

dxy

x

y

dx2- y

2

dz2

dyz

dxz

dxy

x

y

x2-y2 xy

xz yz

z2

6 6

8

2 222

2y

zx

y

zx

Chapter 4

119

small, which agrees with the experimentally observed at 1.973. Thus, equation 4.6.3 can

written as:

Considering the rotation of dxz about the y axis, the interaction of the SOMO with the filled

orbital that results from the combination of the dx2-y

2 and dz

2 orbitals should lead to a g value

larger than ge. Again, the energy difference between the orbitals is large, the deviation of the

gy value is expected to be small. This is in good agreement with the experimental result of

2.005. Equation 4.6.2 can be written as follow:

Conversely, rotation about the z axis yielding the dyz orbitals results in the delocalization of

the electron hole between these two orbitals, leading to a large deviation from the ge value as

described by Equation 4.6.5.

In this situation, the gz is expected to be larger than ge, in agreement with the observed gz

value of 2.499 in the spectrum of 11 [RhII(LTMS)2]2-. The relatively large g deviation from

2.0023 can be explained by the small energy separation of dxz and dyz (∆yz). Thus, the

electronic structure deduced from DFT calculations is in good agreement with the results

obtained from EPR spectroscopy, confirming the 2B2g ground state of 11. This is in contrast to

the results of Gray et al.26 who report a similar EPR spectrum, but suggested a 2Ag ground

state based on the electronic structure of [Ni(mnt)2]2-. It is possible that the [Rh(mnt)2]

2- is

incorrectly characterized. From the experimental EPR evidence, a ground state similar to 11 is

more appropriate.

Eqn. 4.6.3

Eqn. 4.6.4

gx = ge ∆xy

_ 2λ

Eqn. 4.6.4

gz = ge ∆yz

+ 2λ

gy = ge + ∆x2-y2∆z

2

6λ+

2λ

Chapter 4

120

4.7 – Conclusions:

The monoanion 10 has a electronic structure very similar to that of previously reported

[Co(L)]1-.13 Compound [N(n-Bu)4]2[11] represents the first structurally characterized RhII

with ortho-benzenedithiolate ligands. The compound shows an intense LMCT absorption at

752 nm which is not observed for the isoelectronic complex 10b [CoII(LTMS)2]2-. The main

contribution of this chapter is the description of the electronic structure of the anionic

rhodium compounds 11 and 11b. Figure 4.7.1 shows the orbitals involved in the electronic

transitions of compounds 11 and 11b, respectively and the deconvoluted absorption spectrum

of the charge transfer band in 11.

Figure 4.7.1 – Schematic representation of the electronic transitions in 11 and 11b. The black

arrows in the absorption spectrum indicate the changes in the bands when 11 is reduced to

11b.

Transitions III in 11 and I in 11b involve the same orbitals. Bands II and III disappear

in complex 11b, as indicated in the absorption spectrum. The intensity of band I in 11b gains

intensity due to the additional beta transition to the 2b1u-acceptor orbital after reduction to a

RhI d

8 system. DFT calculations support the observed EPR results of 11, indicating a 2B2g

ground state where the unpaired electron lies in a mainly dxz orbital.

au

b1u

b2g

b1u

b2g

b1u

au

b1u

I

I III

II

11000 12000 13000 14000 15000 16000 17000

III / I II

Energy, cm -1

I

11 11b

Chapter 4

121

4.8 – References:

1 Stiefel, E. I.; Schulman, J. M. Prog. Inorg. Chem. 2004, 52, 55-110. 2 Forrester, J. D.; Zalkin, A.; Templeton, D. H. Inorg. Chem. 1964, 3, 1500. 3 Bellamy, D.; Connelly, N. G.; Lewis, G. R.; Orpen, A. G. Cryst. Eng. Comm. 2002, 4, 51. 4 Lewis, G. R.; Dance, I. J. Chem. Soc. Dalton Trans. 2002, 3176. 5 Gray, H. B.; Billig, E. J. Am. Chem. Soc. 1963, 85, 2019. 6 Eisenberg, R.; Dori, Z.; Gray, H. B.; Ibers, J. A. Inorg. Chem. 1968, 7, 741. 7 Baker-Hawkes, M. J.; Billig, E.; Gray, H. B. J. Am. Chem. Soc. 1966, 88, 4870. 8 Ollis, C. R.; Jeter, D. Y.; Hatfield, W. E. J. Am. Chem. Soc. 1971, 93, 547. 9 van der Put, P. J.; Schilperoord, A. A. Inorg. Chem. 1974, 13, 2476. 10 Schrauzer, G. N. Trans. Met. Chem., Marcel Dekkes, New York 1968, 4, 299. 11 Mrkvova, K.; Kameni, J.; Sindela, Z.; Kvitek, L. Trans. Met. Chem., Marcel Dekkes,

New York 2004, 29, 238. 12 Sawyer, D. T.; Srivatsa, G. S.; Bodini, M. E.; Schaefer, W. P.; Wing, R. M. J. Am.

Chem. Soc. 1986, 108, 936-42. 13 Ray, K.; Begum, A.; Weyhermueller, T.; Piligkos, S.; Van Slageren, J.; Neese, F.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 4403-4415. 14 Prater, M. E.; Pence, L. E.; Clérac, R.; Finniss, G. M.; Campana, C.; Auban-Senzier, P.; Jérome, D.; Canadell, E.; Dunbar, K. R. J. Am. Chem. Soc. 1999, 121, 8005-8016. 15 Pandey, K. K. Coord. Chem. Rev. 1992, 121, 1. 16 DeWitt, D. G. Coord. Chem. Rev. 1996, 147, 209. 17 Lilie, J.; Simie, M. G.; Endicott, J. F. Inorg. Chem. 1975, 14, 2129. 18 Ferraudi, G.; Grutsch, P. A.; Kutal, C. J. Chem. Soc., Chem. Commun. 1979, 15. 19 Ogle, C. A.; Masterman, T. C.; Hubbard, J. L. J. Chem. Soc., Chem. Commun. 1990, 1733-1734. 20 Seven mononuclear RhII complexes were found in the Cambridge Crystallography Database according to the general entry of tetracoordinate Rh complexes. For more details check the queries DENQAO, F., GIPYIN, JEZPUZ01, KUCFAP, RVWUO and SORJUE.

Chapter 4

122

21 Osborn, J. A.; Jardine, F. H.; Young, J. F.; Wilkinson, G. J. Chem. Soc. A 1966, 1711- 1732. 22 Birch, A. J.; Williamson, D. H. Organic Reactions 1976, 24. 23 James, B. R. Homogeneous Hydrogenation, 1973, John Wiley & Sons, New York. 24 Evans, D. A.; Fu, G. C.; Hoveyda, A. H. J. Am. Chem. Soc. 1988, 110, 6917-6918. 25 Ojima, I.; Kogure, T.; Nagai, Y. Tetrahedron Lett. 1972, 13, 5035-5038. 26 Billig, E.; Shupack, S. I.; Waters, J. H.; Williams, R.; Gray, H. B. J. Am. Chem. Soc.

1964, 86, 926-7. 27 Maki, A. H.; Edelstein, N.; Davison, A.; Holm, R. H. J. Am. Chem. Soc. 1964, 86, 4580-4587. 28 Alves, H.; Simao, D.; Santos, I. C.; Gama, V.; Henriques, R. T.; Novais, H.; Almeida, M. Eur. J. Inorg. Chem. 2004, 1318-1329. 29 Rempel, G. A.; Legzdins, P.; Smith, H.; Wilkinson, G. Inorg. Synth. 1972, 13, 90-91. 30 Wilson, C. R.; Taube, H. Inorg. Chem. 1975, 14, 405-409. 31 James, C. A.; Morris, D. E.; Doorn, S. K.; Arrington, C. A.; Dunbar, K. R.; Finnis, G. M.; Pence, L. E.; Woodruff, W. H. Inorg. Chim. Acta 1996, 242, 91-96. 32 Sellmann, D.; Fetz, A.; Moll, M.; Knoch, F. Polyhedron 1989, 8, 613-625. 33 Abragan, A.; Pryce, M. H. Proc. Roy. Soc. London 1951, A205, 135. 34 Kimura, S.; Bill, E.; Bothe, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 6025-6039. 35 Carlson, E. H.; Spence, R. D. J. Chem. Phys. 1956, 24, 471. 36 Dunbar, K. R.; Haefner, S. C. Organomettalics 1992, 11, 1431-1433. 37 Willems, S. T. H.; Russcher, J. A.; Budzelaar, P. H. M.; de Bruin, B.; de Gelder, R.; Smits, J. M. M.; Gal, A. W. Chem. Commun. 2002, 148-149. 38 García, M. P.; Jiménez, M. V.; Cuesta, A.; Siurana, C.; Oro, L. A.; Lahoz, F. J.; López, J. A.; Catalán, M. P. Organomettalics 1997, 16, 1026-1036. 39 Hay-Motherwell, R. S.; Koschmieder, S. U.; Wilkinson, G.; Hussain-Bates, B.; Hursthouse, M. B. J. Chem. Soc. Dalton Trans. 1991, 2821-2830. 40 García, M. P.; Jiménez, M. V.; Oro, L. A.; Lahoz, F. J.; Casas, J. M.; Alonso, P. J. Organomettalics 1993, 12, 3257-3263.

Chapter 4

123

41 Ni, Y. N.; Fitzgerald, J. P.; Carroll, P.; Wayland, B. B. Inorg. Chem. 1994, 33, 2029- 2035. 42 Becke, A. D. J. Chem. Phys. 1986, 84, 4524-9. 43 Becke, A. D. J. Chem. Phys. 1993, 98, 5648-52. 44 Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter 1988, 37, 785-9. 45 Neese, F. ORCA an ab initio, DFT and Semiempirical SCF-MO Package, Version 2.5,

Bonn University, (Germany) 2004. 46 Bill, E.; Bothe, E.; Chaudhuri, P.; Chlopek, K.; Herebian, D.; Kokatam, S.; Ray, K.; Weyhermueller, T.; Neese, F.; Wieghardt, K. Chem. Eur. J. 2005, 11, 204-224. 47 Reed, A. E.; Weinhold, F. J. Chem. Phys. 1983, 78, 4066. 48 Reed, A. E.; Weinhold, F.; Curtiss, L. A. Chem. Rev. 1988, 88, 899. 49 Reed, A. E.; Weinhold, F.; Weinstock, R. B. J. Chem. Phys. 1985, 83, 735. 50 Ray, K.; Weyhermueller, T.; Neese, F.; Wieghardt, K. Inorg. Chem. 2005, 44, 5345- 5360. 51 Ray, K.; Bill, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 5641-5654. 52 Ray, K.; George, S. D.; Solomon, E. I.; Wieghardt, K.; Neese, F. Chem. Eur. J. 2007, 13, 2783-2797. 53 Ray, K.; Petrenko, T.; Wieghardt, K.; Neese, F. Dalton Transactions 2007, 1552- 1566. 54 Ray, K.; Weyhermueller, T.; Goossens, A.; Craje, M. W. J.; Wieghardt, K. Inorg.

Chem. 2003, 42, 4082-4087. 55 Klamt, A.; Schurmann, G. J. Chem. Soc. Perkin Trans. 1993, 2, 793. 56 Perkampus, H.-H. Encyclopedia of Spectroscopy. Ed. Willey-VCH 1995, 307.

Chapter 5

Chapter 5

Synthesis and Characterization of

Chromium Complexes with ortho-Benzenedithiolate

Based Ligands

124

Chapter 5

125

Chapter 5

5.1 Introduction

The coordination chemistry of chromium represents a very broad area of research due

to the large range of oxidation states of this element (from –II to +VI).1 The majority of CrII

complexes synthesized since the 1980s are dimers with a quadruple Cr–Cr bond.2,3 More

recently, an interesting compound comprising of a dimer with a Cr–Cr quintuple bond was

reported with a very short Cr–Cr bond length of 1.83 Å.4 Monomeric square-planar CrII

complexes with N-, P-, O-, N-O-, and halide- donors are extensively reported in the

literature.5-15 Only two examples of CrII complexes containing dithiolates have been reported

to date and only their structural features were explored.16-18 Conversely, very few examples of

four-coordinate chromium complexes with S-donor ligands are reported; most are penta- or

hexacoordinated.15,19-21 Some square-planar CrII compounds can react with dioxygen forming

a five-coordinate [LCrV=O]1- compound. This species is known with a variety of macrocyclic

ligands, hydroxocarboxylate and perfluoropinacolate derivatives.22-27 Such CrV oxo

complexes are among the few types of metal complexes that cause oxidative DNA cleavage in

the absence of reactive oxygen species.28 The interactions of chromium oxo compounds with

DNA have been studied in detail and several possible mechanisms have been proposed.29,30

Several mechanistic studies on the reactions of [CrO(salen)]+ (salenH2 = N,N´-

bis(salicylidene)-1,2-ethanediamine) and related complexes with organic reductants have been

performed in relation to the roles of these complexes in CrIII–salen-catalyzed oxo-tranfer

reactions.31,32 Moreover, one of the most prominent reactivities of such complexes is their

ability to catalyse the epoxidation of alkenes by a formal oxygen atom transfer reaction,

which indicates that the oxo ligand may acquire considerable electrophilic character.

Experimental support for the crucial role played by reactive CrV oxo intermediates during the

epoxidation of olefins was initially provided by Groves et al.33 Some stable CrV oxo

compounds can be obtained by oxidation of air-sensitive CrIII complexes such as 5,10,15-

tris(pentafluorophenyl)chromium(III)-corrole.34 A CrV oxo complex containing

toluene-3,4-dithiolate ligands was reported by Gray et al.35 in 1966, however the chemistry of

this species was better explored only recently.36 An understanding of the electronic structures

of well-characterized CrV oxo complexes is essential for interpreting their spectroscopic and

catalytic properties.

In this chapter we present the synthesis and characterization of a CrII complex, namely,

[CrII(LTMS)2]2- 13, and its CrV oxo analogue [CrVO(LTMS)2]1- 14. These complexes have been

studied by cyclic voltammetry, absorption spectroscopy, EPR spectroscopy, SQUID

126

Chapter 5

measurements and DFT calculations. The innocent or noninnocent nature of LTMS has also

been investigated by crystallographic and spectroscopic methods.

Results and Discussions:

5.2 – Synthesis and X-ray Crystal Structures:

The salt of 13 [N(n-Bu)4]2[CrII(LTMS)2]•4MeCN was synthesized under argon by

adding half an equivalent of [CrII(CH3CN)4(BF4)2] to the potassium salt of ligand 1b in

MeCN followed by the addition of two equivalents of lithium-triethylborohydride (super-

hydride) and [N(n-Bu)4]I. The presence of the super-hydride provides a reducing environment

which is crucial in order to isolate pale orange crystals of [N(n-Bu)4]2[13]•4MeCN in 78%

yield. The geometry around the chromium ion is found to be square-planar, as shown in

Figure 5.2.1.

2-

S(1)

Cr(1)

S(2)

C(1)

C(2) C(3)

C(4)

C(5)

C(6) S(3)

S(4)

C(8)

C(7)

Figure 5.2.1 – Perspective view and numbering scheme of the dianion in crystals of

[N(n-Bu)4]2[13]•4MeCN with thermal ellipsoids at the 50% probability level. Hydrogen

atoms are omitted for clarity.

Four MeCN molecules are present in the crystal structure. The N•••Cr distances are

greater than 8 Å indicating clearly that the MeCN molecules are not interacting with the

central metal ion. Figure 5.2.2 displays the packing motif of 13

[N(n-Bu)4]2[13]•4MeCN.

127

Chapter 5

N

Cr

S

S

Si

Si N

N

N

Si

Si

N

N

Figure 5.2.2 – Packing motif of [N(n-Bu)4]2[13]•4MeCN in the unit cell.

Aerial oxidation of an orange CH2Cl2 solution of 13 affords an instantaneous color

change to purple and the formation of the salt of 14 [N(n-Bu)4][CrVO(LTMS)2]•2CH2Cl2 which

was obtained as purple crystals. Crystallizations from CH2Cl2, MeCN and THF solutions were

all successful. The crystal structure of [N(n-Bu)4][14]•2CH2Cl2 has been determined at 100 K

and three different perspectives of the monoanion 14 are shown in Figure 5.2.3. Table 5.2.1

summarizes the important bond distances and dihedral angles. The chromium monoanion

compound 14 has a approximately square-pyramidal geometry. The Cr ion lies 0.719 Å above

the square plane defined by the four sulfur atoms. An interesting structural feature is the

dihedral angle, Φ (see Figure 5.2.3), between the mean S–C–C–S trapezoidal plane and the

Cr–S–S plane which results in the bending of the dithiolate ligand about the S–S vector. The

extent of the folding is different in each of the ligands. In 14 the angle is 144.6° on one side

(angle Φ1 is A–B–Cr in Figure 5.2.3 (b)), while on the other side it is only 175.8° (angle Φ2

is C–D–Cr in Figure 5.2.3. (b)), indicating that the dithiolate ligand is almost coplanar to the

S(1)–C(1)–C(2)–S(2) trapezoidal plane.

128

Chapter 5

Cr(1)

S(1)

O(1)

S(2)

S(3)

S(4)

C(1) C(2)

C(3) C(4)

C(5)

C(6) C(7) C(8) C(9) C(10)

C(11) C(12)

1-

a)

Φ1 = 144.6° Φ2 = 175.8°

0.719 Å

1-

A B

Cr

CD

1-

b)

c)

Figure 5.2.3 – Perspective views of the monoanion 14. a) Numbering scheme with thermal

ellipsoids at 50% probability level. Hydrogen atoms are omitted for clarity. b) The points B

and C represent the centroids of S(1), S(2) and S(3), S(4) respectively. Similarly, the points A

and D represent the centroids of C(3)–C(6) and C(9)–C(12), respectively. The dihedral angles,

Φ1 and Φ 2, are defined by the angles A–B–Cr(1) and Cr(1)–C–D. c) Dihedral angles Φ1 and

Φ2 and the chromium displacement out-of-plane.

129

Chapter 5

Table 5.2.1 – Selected bond distances (Å) in 13 and 14.

13

Cr(1)–S(1) 2.365(1) S(4)–C(8) 1.770(4)

Cr(1)–S(2) 2.359(1) C(1)–C(2) 1.416(5)

Cr(1)–S(3) 2.368(1) C(2)–C(3) 1.409(5)

Cr(1)–S(4) 2.365(1) C(3)–C(4) 1.391(5)

S(1)–C(1) 1.764(4) C(4)–C(5) 1.403(5)

S(2)–C(2) 1.771(4) C(5)–C(6) 1.416(5)

S(3)–C(7) 1.781(3) C(6)–C(1) 1.418(5)

14

Cr(1)–S(1) 2.2781(7) C(2)–C(3) 1.411(3)

Cr(1)–S(2) 2.2872(7) C(3)–C(4) 1.400(3)

Cr(1)–S(3) 2.2768(7) C(4)–C(5) 1.397(3)

Cr(1)–S(4) 2.2803(7) C(5)–C(6) 1.390(3)

Cr(1)–O(1) 1.585(1) C(6)–C(1) 1.412(3)

S(1)–C(1) 1.759(2) C(6)–C(7) 1.411(3)

S(2)–C(2) 1.766(2) C(7)–C(8) 1.420(3)

S(3)–C(7) 1.771(2) C(8)–C(9) 1.391(3)

S(4)–C(8) 1.770(2) C(9)–C(10) 1.396(3)

C(1)–C(2) 1.413(3) C(10)–C(11) 1.393(3)

C(11)–C(12) 1.419(3)

Φ1 144.6° Φ2 175.8°

The dihedral angles, Φ, on either side of the chromium atom differ by ~31° (ΔΦ = Φ2 – Φ1).

DFT calculations presented later in this chapter indicate that the distortion is not a result of

the crystal packing, but electronic in origin.

The Cr–O bond distance of 1.585(1) Å is similar to other known CrV oxo bonds.36-38

The average Cr–S distance (2.280 Å) is slightly shorter than that observed for the Cr–S bond

length (2.30 Å) obtained from the EXAFS analysis of a series of CrV glutathione complexes.39

The C–C bond lengths of the phenyl rings are equidistant in 13 and 14 (within experimental

errors of ± 0.02 Å for 13 and ± 0.01 Å for 14, 3σ). The C–C distances of 1.409 (13) and 1.404

Å (14) are typical for aromatic phenyl rings. In particular, the average C–S bond length of

1.770 ± 0.01 Å (13) and 1.767 ± 0.01 Å (14) are long and indicate the presence of two closed-

130

Chapter 5

shell (LTMS)2- ligands. These observations support the CrII and CrV oxidation state assignments

for the anions 13 and 14, respectively.

5.3 – Electro- and Spectroelectrochemistry:

Figure 5.3.1 shows the cyclic voltammogram of 13 [CrII(LTMS)2]2- obtained at different

scan rates in a CH2Cl2 solution with 0.10 M [N(n-Bu)4]PF6 as the supporting electrolyte,

using a glassy carbon working electrode and Ag/AgNO3 reference electrode. Ferrocene was

used as an internal standard, and potentials are referenced versus the ferrocenium/ferrocene

couple (Fc+/Fc) and listed in Table 5.3.1.

The dianion 13 is extremely air-sensitive and coulometric studies could not be

performed successfully. The cyclic voltammogram shows two electron-transfer processes, the

nature of which could not be defined. It is probable that the CrII in 13 oxidizes to CrIII at very

low potential with a subsequential change in its coordination sphere as represented in equation

5.3.1. Generally, CrIII (d3) complexes with ortho-benzenedithiolate ligands are six-coordinate.

The structure of the monoanionic [CrIII(LBu•)2(LBu)]1- shows a distorted octahedral geometry.40

Thus, it is expected that upon oxidation 13 coordinates two MeCN molecules in the axial

positions of the octahedron.

[CrII(LTMS)2]2- [CrIII(LTMS)2(X)2]1- (X = MeCN). -

– e Eqn. 5.3.1

131

Chapter 5

-0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6

E (V) versus Fc+/Fc

5 μA

Figure 5.3.1 – Cyclic voltammogram of 13 [CrII(LTMS)2]2- recorded in CH2Cl2 solution at

25 °C containing 0.1 M [N(n-Bu)4]PF6 as supporting electrolyte at scan rates of 50 (blue),

100 (red) and 200 (black) mV s-1. (Conditions: glassy carbon electrode; potentials referenced

vs the ferrocenium/ferrocene couple).

Figure 5.3.2 shows the absorption spectrum of 13 in CH3CN at 25°C. The absorption

spectra of other known square-planar chromium(II) bis(dithiolate) complexes such as

[CrII(L)2]2- and [Cr(edt)2]2- (where edt2- = ethane-1,2-dithiolate) have not been reported.17,18 In

fact, the spectrum of 13 shows only a weak band (ε = 82 M-1 cm-1) at 471 nm which can be

attributed to d-d transitions. No bands in the near-infrared were found, supporting the

presence of two closed-shell ligands.

132

Chapter 5

400 500 600 700 800 900 1000 11000.0

0.5

1.0

1.5

2.0

2.5

3.0

ε, 10

2 M-1

cm

-1

λ, nm

400 450 500 550 600 6500.00

0.25

0.50

0.75

1.00

Figure 5.3.2 – Absorption spectrum of 13 [CrII(LTMS)2]2- in MeCN at 25 °C. The caption

shows the region between 400 and 650 nm.

The cyclic voltammogram shown in Figure 5.3.3 corresponds to complex 14

[CrVO(LTMS)2]1- and has also been recorded at 25 °C in CH2Cl2 with the same conditions as

described for 13. The redox potentials are summarized in Table 5.3.1. On the basis of

coulometric studies, the CV of 14 displays one fully reversible one-electron reduction at -

1.053 V, yielding the dianion 14a [CrIVO(LTMS)2]2- and a reversible one-electron oxidation at

+0.119 V, resulting in the formation of complex 14b [CrVIO(LTMS)2]0 as expressed in

Equation 5.3.2. These values are relatively close to the redox potentials found for

[CrVO(LMe)2]1- (see Table 5.3.1). The redox potentials of such species are highly dependent

on the nature of the central metal ion. For example, the reported [MoO(L)]1- shows two

successive one-electron reduction processes at -0.96 and -0.40 V, in agreement with metal-

centered redox chemistry.41

[CrIVO(LTMS)2]2- [CrVO(LTMS)2]1- [CrVIO(LTMS)2]0-e+e

-e Eqn. 5.3.2 +e

14a 14 14b

133

Chapter 5

0.5 0.0 -0.5 -1.0 -1.5

E (V) versus Fc+/Fc

5 μA

Figure 5.3.3 – Cyclic voltammogram of 14 [CrVO(LTMS)2]1- in CH2Cl2 solution at 25 °C

containing 0.10 M [N(n-Bu)4]PF6 as the supporting electrolyte at scan rates of (teal) 400,

(blue) 200, (black) 100 and (red) 50 mV s-1 (glassy carbon electrode, potentials referenced vs

the ferrocenium/ferrocene couple).

Table 5.3.1 – Redox potentials of 13 and 14 in CH2Cl2 solutions containing 0.10 M

[N(n-Bu)4]PF6 at 25°C.

Complex E1½, V vs Fc+/Fc E2

½, V vs Fc+/Fc

13 -1.050 -1.310

14 +0.119 -1.053

[CrVO(LMe)]1- 36 +0.150 -0.96

The one-electron reduced species 14a [CrIVO(LTMS)2]2-, is stable in CH2Cl2 solution and its

electronic spectrum was recorded during the coulometric studies. In contrast, the one-electron

oxidized species 14b decomposed during the coulometric process. Figure 5.3.4 shows the

spectrum of 14 with the respective spectral changes associated with its one-electron

electrochemical reduction. The absorption spectrum of 14 shows two moderately intense low-

energy LMCT transitions at 506 and 732 nm, which are not observed in the reduced dianion

14b. Both compounds 14 and 14b do not show any intense (> 104 M-1 cm-1) intervalence

ligand-to-ligand charge transfer bands in the near-infrared region, which has been previously

134

Chapter 5

discussed as a characteristic of the presence of ortho-dithiobenzosemiquinonate(1-) radicals.

This is an evidence that the redox processes are metal and not ligand centered. Table 5.3.2

summarizes the features observed in the absorption spectra of 13, 14 and 14a.

300 400 500 600 700 800 9000.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

ε, 1

04 M-1

cm

-1

λ, nm

[CrVO(LTMS)2]1- (14)

[CrIVO(LTMS)2]2- (14a)

Figure 5.3.4 – Stepwise one-electron coulometric reduction of complex 14 in CH2Cl2 solution

(0.10 M [N(n-Bu)4]PF6) at -25 °C.

Table 5.3.2 – Summary of the electronic spectra of the complexes at -25 °C in CH2Cl2

solutions.

Complex λ max., nm (ε, 104 M-1 cm-1)

13a 470 (82 x 10-4)

14 322 (1.02), 363 (0.57), 506 (0.42), 732 (0.21)

14a 371 (1.21) a Measurement performed at +25 °C.

135

Chapter 5

5.4 – Magnetic Properties:

The magnetic susceptibility of [N(n-Bu)4]2[13]•4MeCN in the temperature range of 2-

290 K with an external applied field of 1T showed an effective magnetic moment of

4.69 ± 0.01 μB, indicating that 13 possesses four unpaired electron (S = 2), thereby implying

the presence of a CrII ion d7 high-spin. This compound is EPR silent.

0 50 100 150 200 250 3004.04.14.24.34.44.54.64.74.84.95.0

μ eff/

μ B

Temperature, K

■ Experimental Calculated with:

S = 2 g = 1.960 θ-Weiss = 1.00 K D = 2.2 cm-1

TIP = 1056 x 10-6 emu

Figure 5.4.1 – Temperature dependence of the effective magnetic moment, μeff, of complex

[N(n-Bu)4]2[13]•4MeCN in an external applied field of 1 T.

The magnetic susceptibility of [N(n-Bu)4][14]•2CH2Cl2 shown in Figure 5.4.2 displays

an effective magnetic moment of 1.77 ± 0.01 μB, which is close to the expected spin-only

value for a system with one unpaired electron, supporting the assignment of a CrV (d1) central

metal ion.

136

Chapter 5

0 50 100 150 200 250 300

1.3

1.4

1.5

1.6

1.7

1.8

μ eff/

μ B

Temperature, K

■ Experimental Calculated with:

S = ½ g = 1.998 θ-Weiss = -2.11 K TIP = 351 x 10-6 emu

Figure 5.4.2 - Temperature dependence of the effective magnetic moment, μeff, of complex

[N(n-Bu)4][14]•2CH2Cl2 in an external applied field of 1 T.

The S = ½ ground state of 14 was confirmed by X-band EPR spectroscopy. Figure

5.4.3 shows the EPR spectrum and the corresponding simulation of 14 at 10 K. A rhombic

signal with g1 = 2.023, g2 = 1.997, g3 = 1.987 (giso = 2.002) was observed without hyperfine

coupling to 53Cr (I = 3/2, 10% natural abundance). The giso observed at 2.002 is very close to

the free electron g value of 2.0023 indicating a small spin-orbit contribution and thus the

excited state is well separated from the ground state.

137

Chapter 5

300 310 320 330 340 350 360 370

2.2 2.15 2.1 2.05 2 1.95 1.9 1.85 1.8

dχ´´ / d

B

B, mT

g values

Figure 5.4.3 – X-band EPR spectrum of complex 14 in CH2Cl2 at 10 K. Conditions:

frequency 9.44 GHz; modulation 4.0 G; power 1.263 μW. Simulation parameters: g1 = 2.023,

g2 = 1.997, g3 = 1.987, giso = 2.002; line widths W1 = 19.3, W2 = 30.3, W3 = 10.7 MHz.

138

Chapter 5

5.5 – Theoretical Calculations:

Complexes containing the unsubstituted (L)2- (ortho-benzenedithiolate(2-)) ligand

have basically the same electronic structure as that of the (LTMS)2- (for comparison see

refs.42-44). Thus, the molecule 13 was truncated by removing the trimethylsilyl substituents,

and spin unrestricted ZORA B3LYP DFT calculations were performed. Discrete differences

in the bond lengths between the truncated compound [CrII(L)2]2- and 13 [CrII(LTMS)2]2- are

expected because of the effect of the removed trimethylsilyl substituents.45

DFT calculations were carried out with the B3LYP functional for 14 [CrVO(LTMS)2]1-

in order to provide a greater insight into the electronic structure of this compound. In this

case, the trimethylsilyl substituents were taken into consideration in the calculations.

Structure optimization:

The calculated geometry of the truncated complexes [CrII(L)2]2-, and 14 are in good

agreement with the experimental data (Table 5.5.1). The Cr–S, C–S, and the C–C bond

lengths are, accurately reproduced in the calculations, with the error not exceeding ± 0.03 Å.

For both complexes, the C–S bond distances are predicted to be long (~1.77 Å) and the C–C

bond lengths within the phenyl ring are calculated to be essentially equivalent. Thus, the

calculation indicates that the ligands are in their fully reduced form, which is in agreement

with the assignments of CrII and CrV for 13 and 14, respectively. The approximate square-

based pyramidal geometry of 14 is also reproduced in the calculations. Moreover, there is a

reduction in the symmetry from C2v to Cs because of the folding of the C–S–S–C trapezoid

along the S–S vector on either side of the Cr atom. This feature is also observed in the

experimental data. The dihedral angles of the two ligands are reproduced well in the

calculations with errors of < 3°.

139

Chapter 5

Table 5.5.1 – Experimental bond lengths (in Å) of 13 compared with the calculated bond

lengths for the truncated complex [CrII(L)2]2- (in brackets). The calculated dihedral angles in

complex 14 are listed (in brackets).

[CrII(L)2]2-

Cr–S (av.) 2.364 (2.404)

C(3)–C(4) 1.391(5) (1.403)

S–C (av.) 1.776 (1.771)

C(4)–C(5) 1.403(5) (1.391)

C(1)–C(2) 1.416(5) (1.418)

C(5)–C(6) 1.416(5) (1.409)

C(2)–C(3) 1.409(5)

(1.416) C(6)–C(1) 1.418(5)

(1.410)

14

Cr–S (av.) 2.280 (2.314)

C(5)–C(6) 1.390(3) (1.405)

Cr(1)–O(1) 1.585(1) (1.560)

C(6)–C(1) 1.412(3) (1.421)

S–C (av.) 1.766 (1.778)

C(6)–C(7) 1.411(3) (1.421)

C(1)–C(2) 1.413(3) (1.417)

C(7)–C(8) 1.420(3) (1.420)

C(2)–C(3) 1.411(3) (1.421)

C(8)–C(9) 1.391(3) (1.406)

C(3)–C(4) 1.400(3) 1.405

C(9)–C(10) 1.396(3) (1.398)

C(4)–C(5) 1.397(3) (1.399)

C(10)–C(11) 1.393(3) (1.407)

C(11)–C(12) 1.419(3)

(1.419)

Φ1 144.6° (142.0°)

Φ2 175.8° (175.9°)

140

Chapter 5

Bonding Schemes and ground state properties:

Figure 5.5.1 shows the MO description for the truncated complex [CrII(L)2]2- within the D2h

point group.

1au

1b3g

1b1u

1b2g

SCr

S

S

S

Y

X

2-

1b1g (dxy)

2b3g (dyz)

2b2g (dxz)

2ag (dz2)

1ag (dx2-y

2)

Figure 5.5.1 – Qualitative MO diagram of compound [CrII(L)2]2- from the spin unrestricted

ZORA-B3LYP DFT calculations.

141

Chapter 5

The qualitative bonding scheme (Figure 5.5.1) is derived from the spin unrestricted scalar

relativistic BP86 DFT methods, and the ground state electronic configuration for the dianion

13 can be described as

(1au)2(1b3g)2(1b1u)2(1b2g)2(1ag)1(2ag)1(2b2g)1(2b3g)1(1b1g)0

As discussed previously, it was argued that the relative energies of the metal and the ligand

orbitals and the question of metal versus ligand oxidation in a series of transition metal

bis(dithiolate) complexes are essentially dictated by the effective nuclear charge of the central

metal ion involved. Thus, whereas the electronic structures of the 2a [Ni(LTMS•)(LTMS)]1- and

4a [Au(LTMS•)(LTMS)]0 were clearly consistent with the NiII and AuIII assignments, in 10

[Co(LTMS)2]1- the cobalt seems to have character between CoII (d7) and CoIII (d6) and could be

best represented by the resonance forms [CoIII(LTMS)2]1- ↔ [CoII(LTMS•)(LTMS)]1- ↔

[CoII(LTMS)(LTMS•)]1- with greater weight for the first structure.

The effective nuclear charge of the CrII ion is smaller than that of NiII and AuIII.

Correspondingly, the metal d orbitals are more destabilized relative to the ligand orbitals and

any kind of symmetry-allowed mixing between the ligand and the metal orbitals are

energetically forbidden. Thus, the bonding scheme of 13 shows four predominantly metal

based orbitals, namely 1ag(dz2), 2ag(dx

2-y

2), 1b2g(dxz) and 1b3g(dyz), which are singly occupied.

Therefore, the valence states of the metals are best represented as d4 CrII ion, with the support

of magnetization measurements. The ligand orbitals are lower in energy and all redox

processes are thus expected to be metal centered.

The d-population and spin density at the central metal ion obtained from the natural

population analysis of the B3LYP densities for 13 are summarized in Table 5.5.2, and are in

agreement with the valence nd4 electron configuration. The excess over the formal d4

electronic configuration arises from the covalent population of the otherwise unpopulated Cr

dxy orbital due to strong σ-donation from the ligand. The spin density of 3.89 for the Cr is

calculated to be located predominantly on the metal center supporting the assignment of a CrII

central ion.

Table 5.5.2 – Charge and spin population at the Cr ion resulting from a natural population

analysis of the one-electron density of the ground state obtained from scalar relativistic

ZORA-B3LYP DFT calculations.

electrons-nd electrons (n+1)s Spin-nd Metal oxdn. state [CrII(L)2]2- 4.91 0.49 3.89 CrII

142

Chapter 5

Spin up Spin down

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Ener

gy,e

V

1a´1a´

1a´´2a´

2a´

2a´´

1a´´2a´´

3a´´

3a´´3a´

3a´

4a´´

4a´´

4a´

4a´

5a´ (dx2-y

2)

5a´

5a´´

5a´´ (dyz)

6a´ (dxz) 6a´

6a´´ (dxy)6a´´

7a´ (dz2)7a´

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Ener

gy,e

V

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Ener

gy,e

V

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Ener

gy,e

V

1a´1a´

1a´´2a´

2a´

2a´´

1a´´2a´´

3a´´

3a´´3a´

3a´

4a´´

4a´´

4a´

4a´

5a´ (dx2-y

2)

5a´

5a´´

5a´´ (dyz)

6a´ (dxz) 6a´

6a´´ (dxy)6a´´

7a´ (dz2)7a´

1a´1a´

1a´´2a´

2a´

2a´´

1a´´2a´´

3a´´

3a´´3a´

3a´

4a´´

4a´´

4a´

4a´

5a´ (dx2-y

2)

5a´

5a´´

5a´´ (dyz)

6a´ (dxz) 6a´

6a´´ (dxy)6a´´

7a´ (dz2)7a´

CrS

S

S

S

O 1-

Z

Y

X

TMS

TMS

TMS

TMS

Figure 5.5.2 – Unrestricted Kohn-Sham MOs and energy scheme for the monoanion 14

[CrVO(LTMS)2]1- within the Cs group point obtained from the B3LYP DFT calculations.

143

Chapter 5

Figure 5.5.2 shows the bonding scheme of 14 [CrVO(LTMS)2]1-. Similar to 13, the Cr 3d

orbitals are all placed at energies much higher than the ligand orbitals. The electronic

structure configuration for Cr in compound 14 is formally d1, and this unpaired electron

resides in the nonbonding Cr 3dx2-y

2 orbital, leading to a doublet ground state. The other four

3d orbitals remain unoccupied. The d orbital splitting, as shown in Figure 5.5.2, is thus

predicted to be 3dx2-y

2 < 3dyz < 3dxz < 3dxy < dz2. This splitting is a result of: (1) the presence of

the terminal oxo ligand, which is a strong σ- and π-donor, and (2) the presence of a moderate-

to-weak equatorial ligand field. The 3dz2, 3dxz, and 3dyz orbitals are strongly destabilized by σ-

and π-antibonding interactions with the terminal oxo ligand to such an extent that they usually

tend to remain unoccupied. Beneath this d orbital manifold there is a set of eight orbitals

corresponding to the symmetry-adapted linear combinations of the sulfur ligand 3p lone pairs

(see Figure 5.5.2). Four of the sulfur 3p orbitals in Figure 5.5.2 are in π-symmetry with

respect to the Cr center. The remaining four posses σ-symmetry with respect to the metal.

Table 5.5.3 – Percentage composition of the selected orbitals of 14 as obtained from B3LYP

DFT calculation using large uncontracted Gaussian basis sets at the metal and uncontracted

all-electron polarized triple-ξ (TZVP) Gaussian basis sets for the remaining atoms.

Cr S C O

3dx2-y

2 3dxz 3dyz 3dxy 3dz2 3px,y 3pz 2px,y 2pz 2px,y 2pz

4a´ 5 5 47 3 32 3

5a´ 80 2 5 3

5a´´ 64 8 19 19

6a´ 66 10 19

6a´´ 50 27 9

7a´ 23 40

9

The composition of selected orbitals in 14 [CrVO(LTMS)2]1- is summarized in Table

5.5.3. Comparing the Cr–S covalency with the isoelectronic [MoVO(edt)2]1-

(edt = ethane-1,2-dithiolate),46 it is observed that complex 14 is less covalent than that of the

Mo analogue. The 5a´ orbital in 14 has 80% 3dx2-y

2 character in contrast with the Mo analogue

with 70% 3dx2-y

2 contribution as shown in Figure 5.5.3. This differences in the lowest-lying

acceptor orbitals are due to the higher effective nuclear charge of MoV compared to that of

CrV, which leads to a smaller mixing between ligand and metal orbitals in 14.

144

Chapter 5

[MoVO(edt)2]1- [CrVO(LTMS)2]1-

S 3p S 3p

5a´

Cr 3dx2-y

2

80% dx2-y

2

S 3p

5a´

70% dx2-y

2

S 3p

Mo 4dx2-y

2

Figure 5.5.3 – Schematic representation of the 5a´ orbital in MoV and CrV oxo complexes,

with their correspondent percentage of metal character.

The terminal oxo donor displaces the Cr ion in ~0.719 Å (calculated 0.72 Å) out of the

bis(dithiolate) plane composed by the four sulfur atoms, leading to significant less S(p)–Cr(d)

π-orbital overlap than other reported square planar bis(dithiolates) complexes.42-44,47 The

experimental distortion in 14 [CrVO(LTMS)2]1- is electronic in origin. In order to understand

the nature of the distortion in the ligands of 14, it is necessary to analyse the 4a´ orbitals in

two different models: in C2v and in Cs symmetry. Figure 5.5.4 shows the interactions between

the S pz with the Cr 3dx2-y

2 orbitals in both C2v and in Cs symmetries.

145

Chapter 5

Cs

SS

O

SS

a)

C2v

S S

O

S S

b)

Figure 5.5.4 – (a) Scheme of the 4a´ molecular orbital. Dashed lines represent the interaction

between the S pz and the Cr 3dx2-y

2 orbitals in Cs and C2v symmetries. The phenyl rings are

removed for clarity and the interactions are shown only with the sulfur atoms in the front. The

two figures in the middle represent different perspectives of the 4a´ orbital obtained by DFT

calculations. (b) Hypotetical model of the 4a´ orbital in C2v symmetry.

The better overlap between the S pz orbital with the half-filled Cr 3dx2-y

2, as indicated

by the dashed lines in Figure 5.5.4, leads to stabilization with the ligand folding. Calculations

on the [CrVO(L)2]1- showed that in fact, the Cs symmetry is energetically more favourable by

4.8 kcal mol-1 than that of C2v symmetry.36 Also the 4a´ π-type orbital is stabilized by 5.3 kcal

mol-1 due to the ligand folding allowing the delocalization of the unpaired electron in the Cr

146

Chapter 5

3dx2-y

2 over the S pz orbital. This type of distortion in the ligand has been observed in models

of molybdenum and tungsten oxotransferases.48 If the distortion occurs in both ligands it is

expected that the stabilizing effect would not be present anymore. Thus, the ΔΦ of ~31° for

14 represents the orientation of the ligand in which this stabilization is maximized, satisfying

then the π-acidity of the CrV ion. From the calculation, this is predicted to be 34°, in good

agreement with the experimental value.

When 14 is reduced to 14a [CrIVO(LTMS)2]2- the anti-bonding Cr 3dx2-y

2 orbital

becomes doubly occupied and in this case, the Cr 3dx2-y

2 ↔ S(p) π-interaction will now have a

destabilizing effect and the ligands will orient themselves to minimize this repulsive

interaction and the molecule will adopt a C2v symmetry. Although the crystal structure of 14a

[CrIVO(LTMS)2]2- is not available, the data of the reported [MoIVO(LTMS)2]2- analogue shows in

fact, a C2v geometry.41

The low symmetry for 14 [CrO(LTMS)2]1- makes all transitions from the eight doubly

occupied MOs in Figure 5.5.2 into the singly occupied 5a´ 3dx2-y

2 orbital and the four virtual

d-based orbitals dipole allowed. Such features have been calculated and also observed for

[CrVO(L)2]1-.36 The bands observed in the spectrum of 14 at 506 and 732 nm (see Figure 5.3.4

and Table 5.3.2) are LMCT transitions from the four out-of–plane sulfur p orbitals of π-

symmetry to the singly occupied Cr 3dx2-y

2 orbital. The lowest band corresponds to the

4a´(Spπ) → 5a´ (3dx2-y

2) transition.36

5.6 – Conclusions:

In this chapter we have discussed the characterization of a CrII square planar complex

coordinated to ortho-benzenedithiolate ligands, clarifying the ground state and the electronic

structure of this species. The sensitivity of compound 13 towards oxygen resulted in the

formation of compound 14, which has also been structurally characterized. The monoanion 14

shows different dihedral angles, φ, between the C–S–S–C and Cr–S–S planes on the two sides

of the chromium atom, which is a result of electronic effects. In both cases, the electronic

structure of the ground state was investigated, showing a quintet ground state for 13 (CrII, d4

high spin) and a doublet for 14. The d orbitals are higher in energy than those of the ligands,

as a consequence of the small effective nuclear charge of the Cr atom. The main contribution

of this study is the ground state characterization of compound 13, as analogues reported in the

literature have been only structurally characterized.

147

Chapter 5

5.7 – References

1 McCleverty, J. A.; Meyer, T. J. Comprehensive Coordination Chemistry II 2004, 4, 313-413. 2 Cotton, F. A.; Clerac, R.; Daniels, L. M.; Dunbar, K. R.; Murillo, C. A.; Pascual, I. Inorg. Chem. 2000, 39, 748-751. 3 Cotton, F. A.; Hillard, E. A.; Murillo, C. A.; Zhou, H.-C. J. Am. Chem. Soc. 2000, 122, 416-417. 4 Brynda, M.; Gagliardi, L.; Widmark, P.-O.; Power, P.; Roos, B. O. Angew. Chem., Int. Ed. 2006, 45, 3804-3807. 5 Dewan, J. C.; Edwards, A. J.; Guy, J. J. J. Chem. Soc. Dalton Trans. 1986, 2623-2627. 6 El-Sawy, N. M.; Al Sagheer, F. Eur. Polym. J. 2000, 37, 161-166. 7 Gayatri, R.; Rajaram, A.; Rajaram, R.; Govindaraju, K.; Rao, J. R.; Nair, B. U.; Ramasami, T. Proc. Indian Acad. Sci. Chem. Sci. 1997, 109, 307-317. 8 Gibson, V. C.; Newton, C.; Redshaw, C.; solan, G. A.; White, A. J. P.; Williams, D. J. J. Chem. Soc., Dalton Trans. 1999, 827-829. 9 Gulanowski, B.; Cieslak-Golonka, M.; Szyba, K.; Urban, J. BioMetals 1994, 7, 177- 184. 10 Hlavaty, J. J.; Nowak, T. Biochemistry 1998, 37, 8061-8070. 11 Liu, M.-H.; Zhang, X.-S.; Deng, Y.; H.-Y., Z. Water Environ. Res. 2001, 73, 322-328. 12 Mertz, W. Nutr. Rev. 1998, 56, 174-177. 13 Van Wart, H. E. Methods Enzymol. 1988, 158, 95-110. 14 Edema, J. J. H.; Gambarotta, S.; Spek, A. L. Inorg. Chem. 1989, 28, 811-812. 15 Larkworthy, L. F.; Leonard, G. A.; Povey, D. C.; Tandon, S. S.; Tucker, B. J.; Smith, G. W. J. Chem. Soc., Dalton Trans. 1994, 1425-1428. 16 Rao, C. P.; Dorfman, J. R.; Holm, R. H. Inorg. Chem. 1985, 24, 453-454. 17 Rao, C. P.; Dorfman, J. R.; Holm, R. H. Inorg. chem. 1986, 25, 428-439. 18 Sellmann, D.; Wille, M.; Knoch, F. Inorg. Chem. 1993, 32, 2534-2543. 19 Arif, A. M.; Hefner, J. G.; Jones, R. A.; Koschmieder, S. U. Coord. Chem. 1991, 23, 13-19.

148

Chapter 5

20 Jubb, J.; Larkworthy, L. F.; Leonard, G. A.; Povey, D. C.; Tucker, B. J. J. Chem. Soc., Dalton Trans. 1989, 1631-1633. 21 Jubb, J.; Larkworthy, L. F.; Povey, D. C.; Smith, G. W. Polyhedron 1989, 8, 1825- 1826. 22 Codd, R.; Levina, A.; Zhang, L.; Hambley, T. W.; Lay, P. A. Inorg. Chem. 2000, 39, 990-997. 23 Collins, T. J.; Slebodnick, C.; S., U. E. Inorg. Chem. 1990, 29, 3433-3436. 24 Judd, R. J.; Hambley, T. W.; Lay, P. A. J. Chem. Soc. Dalton Trans. 1989, 2205-2210. 25 Krumploc, M.; DeBoer, B. G.; Rocek, J. J. Am. Chem. Soc. 1978, 100, 145-153. 26 Meier-Callahan, A. E.; Gray, H. B.; Gross, Z. Inorg. Chem. 2000, 39, 3605-3607. 27 Nishino, H.; Kochi, J. K. Inorg. Chim. Acta 1990, 174, 93-102. 28 Levina, A.; Barr-David, R.; Codd, R.; Lay, P. A.; Dixon, N. E.; Hammershoi, A.; Hendry, P. Chem. Res. Toxicol. 1999, 12, 371-381. 29 Codd, R.; Dillon, C. T.; Levina, A.; Lay, P. A. Coord. Chem. Rev. 2001, 216-217, 533-577. 30 Levina, A.; Codd, R.; Dillon, C. T.; Lay, P. A. Prog. Inorg. Chem. 2003, 51, 145-250. 31 Rihter, B.; Masnovi, J. J. Chem. Soc., Chem. Commun. 1988, 35-37. 32 Sevvel, R.; Rajagopal, S.; Srinivasan, C.; alhaji, N. I.; Chellamani, A. J. Org. Chem. 2000, 65, 3334-3340. 33 Groves, J. T.; Kruper, W. J., Jr. J. Am. Chem. Soc. 1979, 101, 7613. 34 Mahammed, A.; Gray, H. B.; Meier-Callahan, A. E.; Gross, Z. J. Am. Chem. Soc. 2003, 125, 1162-1163. 35 Stiefel, E. I.; Eisenberg, R.; Rosenberg, R. C.; Gray, H. B. J. Am. Chem. Soc. 1966, 88, 2956. 36 Kapre, R. R.; Ray, K.; Sylvestre, I.; Weyhermueller, T.; DeBeer George, S.; Neese, F.; Wieghardt, K. Inorg. Chem. 2006, 45, 3499-3509. 37 Samsel, E. G.; Srinivasan, C.; Kochi, J. K. J. Am. Chem. Soc. 1985, 107, 7606. 38 Siddall, T. L.; Miyaura, N.; Huffman, J. C.; Kochi, J. K. J. Chem. Soc., Chem. Commun. 1983, 1185. 39 Levina, A.; Zhang, L.; Lay, P. A. Inorg. Chem. 2003, 42, 767.

149

Chapter 5

40 Kapre, R. R.; Bothe, E.; Weyhermueller, T.; DeBeer George, S.; Muresan, N.; Wieghardt, K. Inorg. Chem. 2007, 46, 7827-7839. 41 Boyde, S.; Ellis, S. R.; Garner, C. D.; Clegg, W. J. Chem. Soc., Chem. Commun. 1986, 1541. 42 Ray, K.; Begum, A.; Weyhermueller, T.; Piligkos, S.; Van Slageren, J.; Neese, F.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 4403-4415. 43 Ray, K.; Weyhermueller, T.; Goossens, A.; Craje, M. W. J.; Wieghardt, K. Inorg. Chem. 2003, 42, 4082-4087. 44 Ray, K.; Weyhermueller, T.; Neese, F.; Wieghardt, K. Inorg. Chem. 2005, 44, 5345- 5360. 45 Ray, K.; Benedito, F. L. unpublished results. 46 McMaster, J.; Carducci, M. D.; Yang, Y.-S.; Solomon, E. I.; Enemark, J. H. Inorg. Chem. 2001, 40, 687. 47 Ray, K.; Bill, E.; Weyhermueller, T.; Wieghardt, K. J. Am. Chem. Soc. 2005, 127, 5641-5654. 48 Enemark, J. H.; Cooney, J. A.; Wang, J.-J.; Holm, R. H. Chem. Rev. 2004, 104, 1175.

150

Chapter 6

151

Chapter 6

New tris(Dithiolate) Complexes of Rhenium –

A Radical Approach

Chapter 6

152

Chapter 6

153

6.1 Introduction

Synthesis and structural investigations of tris(dithiolate) complexes began in the early

1960s with the report of a neutral Mo(tfd)3 (tfd = bis-(trifluoromethyl)-dithietene) species

obtained from reaction of Mo(CO)6 with tdf.1 Since then, tris(dithiolate) metal compounds

have attracted considerable interest, specifically after the discovery that these complexes

undergo a series of facile one-electron redox reactions.2,3 Complexes containing Fe, V, Re,

Os, Ru, Cr, Mo, and W were reported by Schrauzer et al.4,5 One of the most astonishing

discoveries was the unexpected ReS6 geometry in the crystal structure of [Re(S2C2Ph2)3]0

reported by Eisenberg and Ibers in 1965.6 The geometry of this compound was found to

contain an almost perfect trigonal prismatic (TP) array of sulfur-donor atoms. In addition, this

was the first exception to the paradigm “six-coordinate ion means octahedral geometry”.

Figure 6.1.1 shows Eisenberg´s structure of [Re(S2C2Ph2)3]0.

Figure 6.1.1 – Crystal structure of [Re(S2C2Ph2)3]. The picture on the right hand side shows

more clearly the trigonal prismatic geometry at the rhenium atom, conferring a D3h point

group to the complex. In this type of complex, three sulfurs, one from each ligand form two

S3 planes, which are coplanar and the sulfur atoms of the two planes are fully eclipsed

(trigonal prismatic ReS6 polyhedron).

Subsequently, many other trigonal prismatic complexes have been structurally

characterized. To date, approximately 15 six-coordinated complexes containing dithiolate

ligands have been reported.7-10 On the basis of theoretical calculations, Gray et al. described

Chapter 6

154

the electronic structure of [Re(S2C2Ph2)3] as shown in Figure 6.1.2, and speculated about the

possible factors that stabilize the trigonal prismatic geometry.

Figure 6.1.2 – Bonding scheme proposed by Gray et al. for [Re(S2C2Ph2)3] in a D3h point

group. The metal-ligand planes are defined by the five-membered rings that radiate out with a

three-fold axis in a “paddle-wheel” fashion. The phenyl rings are omitted for clarity.

Labelling: πh are sp2 hybrids on sulfur whose orientation is at 120° to the σ orbitals. They

strongly overlap with dz2 (a1´) of the Re (this overlap is of both σ and π character). πv are the

four π-orbitals situated perpendicular to the plane of each of the three chelate rings and

delocalized over the S–C–C–S framework.

The interligand S–S distance in [Re(S2C2Ph2)3]0 is always close to 3.05 Å, indicating

that there are interligand bonding forces which are considerably stronger than in classical

octahedral, tetrahedral, or planar complexes. The compromise between these S–S bonding

interactions and the S–S repulsions leads to the ubiquitous 3.05 Å separation and cooperates

Chapter 6

155

to the stability of these nonclassical structures. One important factor, according to Gray, could

be the effective use of the three valence d orbitals not involved in σ bonding. For example,

strong involvement of the sulfur πh orbitals with the Re dz2 leads to a stable bonding orbital

2a2´, which is occupied. Another possible stabilizing influence for trigonal prismatic (TP)

coordination is the large interaction of the dxy, dx2-y

2 orbitals with the thoroughly delocalized

ligand 3πv level. This results in a stable 4e´ level, which is occupied.10 In this model, addition

of electrons to this orbital would destabilize the trigonal prism and cause a rotation of the S3

faces about the C3 axis toward octahedral symmetry. Figure 6.1.3 shows the changes in

geometry starting from octahedron to TP. Indeed, crystallographic investigations showed that

tris(dithiolate) complexes such as [Mo(mnt)3]2- (mnt2- = maleonitriledithiolate) and [Mo(L)3]

1-

(L = ortho-benzenedithiolate) adopt a geometry between TP and octahedral.11,12 However,

there are some notable exceptions such as [Mo(mdt)3]z (mdt2- = 1,2-dimethylethene-1,2-

dithiolate, z = 2-, 1-), where the Mo–S6 trigonal prism is maintained throughout the series.

Another noteworthy case is [Ta(L)3]1-, which is isoelectronic with TP [Mo(edt)3] (edt2- = 1,2-

ethanedithiolate) but forms a distorted octahedron.13-15

Figure 6.1.3 – Geometry changes from octahedron (OCT) (a) to trigonal prismatic (c) by

twisting systematically the C3 axis with consequent changes in the angle θ (in blue) from 60°

to 0° (b).

Twist angle, θ

Octahedron

θ = 60°

Trigonal Prismatic

θ = 0°

C3 C3

a)

b)

c)

Chapter 6

156

The trigonal twist angle, θ, between the sulfur atoms in a two dimensional projection along

the threefold axis (C3) (Figure 6.1.3) describes the coordination geometry of a structure

between TP (θ = 0°) and OCT (θ = 60°) extremes.16,17 The values of θ can be constrained

from reaching the full OCT limit (60°) by chelating ligands.

The argument proposed by Gray et al. based on interligand S–S interaction was

refuted by comparing the S–S distances in the trigonal prismatic complexes [Mo(L)3] and

[Nb(L)3]1- with those of the distorted complexes [Zr(L)3]

2-, [Mo(mnt)3]2-, and

[W(mnt)3]2-.7,18-20 Some authors support the hypothesis that the occurrence of different

dihedral angles is due to packing forces, which is dependent on the substituents in the

dithiolate ligand and the counter-ion involved.19 Schrauzer et al. speculate about the adoption

of trigonal prismatic geometry based on the fact that the highest occupied orbital has

predominantly ligand character, forcing the sulfur atoms into a state between sp2 and sp

3

hybridization, which could receive additional stabilization through intermolecular packing

effects.9 The factors that lead a complex to adopt a trigonal prismatic geometry remain

unclear, despite several previously reported proposals.

In this chapter the synthesis and characterization of two new Re complexes is

presented, namely, [Re(LTMS)3]1- 15, and [Re(LCl)3]

1- 16. These compounds were

characterized by X-ray crystallography and represent the first examples of monoanionic

rhenium complexes characterized by this technique. In order to gain a better understanding of

the electronic structures of tris(dithiolate) rhenium complexes, an electron-transfer series

consisting of [Re(L)3]1- 17, [Re(L)3]

2- 17a, and [Re(L)3]

0 17b was prepared (L = ortho-

benzedithiolate). Sulfur K-edge X-ray absorption spectra of these complexes were recorded

and DFT calculations were performed to unambiguously determine the electronic structure of

these Re tris(dithiolate) complexes.

Results and Discussions

6.2 – Synthesis and X-ray crystal structures:

The salt of 15, [C8H16N][Re(LTMS)3]•CH3CN was synthesized under argon by adding

one equivalent of ReCl5 to three equivalents of ligand 1b in MeCN followed by the addition

of an excess of anhydrous [C8H16N]Br (5-azonia-spiro[4,4]nonane bromide) in CH2Cl2. After

six days, crystals suitable for X-ray crystallography were isolated from the MeCN solution in

55% yield. Attempts to crystallize [Re(LTMS)]1- with other counter cations were unsuccesful.

Chapter 6

157

The [C8H16N]+ counter cation has advantages in crystallization due to its flexibility to adopt a

large variety of conformations, ranging from envelope to twisted geometry,21 which results in

a more efficient space-filling structure with less compact anions such as [Re(LTMS)3]1-.

The salt of 16, [C8H16N][Re(LCl)3]•acetone, was synthesized under argon using

commercially available 3,6-dichlorobenzenedithiol (H2LCl). Three equivalents of the ligand

were suspended in MeCN and deprotonated with KOtBu resulting in an orange solution. One

equivalent of solid ReCl5 was added to the solution of (LCl)2-. The resulting brown-green

solution was filtered and an excess of [C8H16N]Br dissolved in a minimum volume of CH2Cl2

was added. The solvent was removed and the material was redissolved in acetone. After five

days at 4 °C crystals were isolated in 24% yield. Crystallization attempts utilising other

solvents such as CH2Cl2, MeCN, CHCl3 and THF were unsuccessful. Compound 17b was

prepared according to the experimental procedure reported by Gray et al.10 complexes 17 and

17a were obtained upon reduction with 1 and 2 equivalents of n-butyllithium, respectively.

Figure 6.2.1 shows the crystal structure and the labelling scheme of the monoanions 15 and

16, respectively. Table 6.2.1 summarizes the structural features.

Figure 6.2.1 – Perspective view of the monoanions 15 and 16 with thermal ellipsoids at 50%

probability level. Hydrogen atoms are omitted for clarity. (R = TMS for 15, and Cl for 16).

S(1)

S(2)

S(3)

S(4)

S(5)

S(6)

15

S(1)

S(2) S(3)

S(4)

S(5)

S(6)

16

Re

S S

S

SS

S

R

R

R

R

R R

1

2

3

45

6

7

1 2

1

3

45

6

8

9

1011

1213

1415

16

1718

1- 1-

Chapter 6

158

Table 6.2.1 – Selected bond distances (Å) in 15 and 16.

15 16

Re(1)–S(1) 2.341(2) 2.3513(6)

Re(1)–S(2) 2.358(2) 2.3373(6)

Re(1)–S(3) 2.330(2) 2.3279(6)

Re(1)–S(4) 2.345(2) 2.3390(6)

Re(1)–S(5) 2.330(2) 2.3371(6)

Re(1)–S(6) 2.345(2) 2.3476(6)

S(1)–C(1) 1.762(9) 1.745(3)

S(2)–C(2) 1.739(1) 1.737(3)

S(3)–C(7) 1.742(7) 1.738(2)

S(4)–C(12) 1.743(7) 1.740(2)

S(5)–C(13) 1.742(7) 1.740(2)

S(6)–C(18) 1.743(7) 1.735(2)

C(1)–C(2) 1.398(1) 1.400(4)

C(2)–C(3) 1.434(1) 1.407(4)

C(3)–C(4) 1.387(1) 1.382(4)

C(4)–C(5) 1.430(1) 1.387(5)

C(5)–C(6) 1.384(1) 1.381(4)

C(6)–C(1) 1.429(1) 1.403(4)

C(7)–C(8) 1.398(9) 1.401(3)

C(8)–C(9) 1.426(1) 1.406(3)

C(9)–C(10) 1.386(1) 1.377(4)

C(10)–C(11) 1.381(1) 1.394(4)

C(11)–C(12) 1.391(1) 1.381(4)

C(12)–C(7) 1.414(1) 1.402(3)

C(13)–C(14) 1.398(1) 1.401(3)

C(14)–C(15) 1.426(1) 1.406(3)

C(15)–C(16) 1.386(1) 1.377(4)

C(16)–C(17) 1.381(1) 1.394(4)

C(17)–C(18) 1.391(1) 1.381(4)

C(18)–C(13) 1.414(9) 1.402(3)

Chapter 6

159

The coordination environment around the rhenium ion in complexes 15 and 16 is

relatively similar. Both compounds show a rhenium center coordinated by six sulfur atoms,

but the substituents in the phenyl rings are different between the complexes. Compound 15

has electron donating groups, in contrast to the electron withdrawing chlorine atoms in 16. In

principle, no significant structural differences were expected for both compounds but

surprisingly, the X-ray structure analysis revealed striking differences in their molecular

structures. Figure 6.2.2 shows the crystal structures of 15 and 16 in different orientations

showing clearly the differences in the structures.

120

Figure 6.2.2 – Two different perspective views of complexes 15 (left) and 16 (right). Note the

distortion from trigonal prismatic geometry for 16.

The structures in Figure 6.2.2 (top) are oriented along the C3 axis of the molecules. In

the case of complexes 15 and 16, θ values of 0.46° and 24.8° were found, respectively,

indicating an almost perfect TP geometry for complexes 15 and an intermediate structure

between TP and OCT for complex 16.

A number of methods can be used in order to describe the coordination geometry in

tris(dithiolate) complexes.11 In one such method, dihedral angles between ligand SMS planes

are indicative of structural tendencies between TP (120°) and octahedral (OCT, 90°)

16

1-

1-

1-

1-

15

Chapter 6

160

geometries.17 Some of the different methods applicable for the geometry determination of the

compounds can be described as follows:

I – Analysis of the dihedral angle between ligand SMS and SSS planes. The average of the

dihedral angle, Φ, between ligand SMS and trigonal SSS planes shown in Figure 6.2.3 has

been calculated for several tris(dithiolate) structures. The average Φ can be used as a direct

measure for the tendency towards TP (90°) or OCT (~55°) geometry.22,16,23 The Φ values

found for complexes 15 and 16 are 89.5° and 75.6°, respectively. This clearly shows that 15 is

TP and 16 adopts an intermediate conformation between TP and OCT.

Figure 6.2.3 – Diagram of one dihedral angle (Φ) between ligand SMS and trigonal SSS

planes. Due to the distortion from TP to OCT geometry, the ligands twist and Φ becomes

<90°.

II – Determination of geometry by the dihedral angles between SSS planes. Interesting

structural information regarding the geometry determination of tris(dithiolate) structures can

be obtained from θ, Φ and δ. The parameter δ can be described as the structural distortion in

the deviation from 0° of the dihedral angle between the two SSS planes, in which each plane

is defined by three sulfur atoms comprising a trigonal face, as shown in Figure 6.2.4.

Figure 6.2.4 – Diagram of the top and bottom trigonal SSS planes. The dihedral angle

between the planes (δ) is a measure of distortion in the complex.

Φ

S

S

S S

S

S S S

Chapter 6

161

The δ values for complexes 15 and 16 are 0.1° and 1.1°, respectively. Comparing these values

to other structures reported in the literature, it was found for [Sb(tdt)3]1- a dihedral angle of δ

= 5.7°, θ = 52° and Φ = 57°, which is practically octahedral.23

III – Determination of trans-SMS angles. A simpler and more straightforward method

reported in the literature involves measuring trans-SMS angles (SMStrans), where the two

sulfur atoms are from different dithiolate ligands and are nearly opposite each other in the

complex.20,24 Regular TP and OCT geometries have SMStrans values of 136° and 180°,

respectively. As in the case for twist angles, SMStrans may be constrained from reaching the

OCT limit by small chelate bite angles. The complement of the chelate angle (i.e. SMSintra)

must be approximately equal to the supplement of SMStrans.20 So [εcorr ≅ 180°–(90°–SMSintra)].

Where εcorr is the SMStrans value expected for an OCT geometry constrained by SMSintra. εcorr

only considers the geometrical constrains of the ligand in determining the OCT limit. A

compilation of those values is reported in the literature. To simplify the calculation, a

straightforward measure of coordination geometry (TP → OCT) can be calculated as:

TP → OCT = {[(SMStrans – 136°)/(εcorr – 136°)] x 100%}. Values of TP → OCT range

between 0 and 100%, which are representative of TP and OCT ligands, respectively.

Structural distortion can also mislead interpretation of TP → OCT values. A simple method of

estimating structural distortions (besides calculating values for δ) is to examine the range of

SMStrans values within a particular complex. If ∆SMStrans is large, significant structural

distortion is present and average values describing coordination geometry must be used with

caution. Effectuating this analysis, we obtain for monoanions 15 and 16 a TP → OCT of 1

and 45%, respectively. In the case of the neutral complex [Re(S2C2Ph2)3] shown in figure

6.1.1, the view down the C3 axis in the near D3h complex, shows the average twist angle to

near, though not exactly zero (θ = 6°). Other geometrical parameters (TP → OCT of 1% and

Φ of 86°) also suggest near TP geometry as in complex 15. One of the three ligands in

Eisenberg´s complex is twisted slightly, which results in an elevated distortion value

(∆SMStrans = 3.8°) similar to complex 15 with a ∆SMStrans = 3.4°.

Chapter 6

162

The crystal packing of the salt of 16 [C8H16N][Re(LCl)3]•acetone shows that the anions

are arranged in pairs, having a crystallographic inversion center between them.

The phenyl-phenyl ring distance at 3.479 Å indicates a weak Van der Waals interaction

between the two anionic units. The pairs are surrounded by solvent and counter cation

molecules. However, this feature is not observed for the salt of 15

[C8H16N][Re(LTMS)3]•CH3CN, in which the phenyl ring units are well separated with no

significantly short intermolecular interaction, probably due to steric effects caused by the

TMS substituents. Monoanions 15 and 16 show average Re–S bond lengths of 2.342 and

2.340 Å, respectively. The average C–S distances at 1.745 and 1.740 Å (3σ ≈ 0.01 Å), are

intermediate between a typical single and double bond length. In square-planar complexes

containing two radical ligands, the C–S• distance at 1.722 ± 0.003 Å (3σ) is observed for

compound 2b [Ni(LTMS•)2] (see chapter 2). In the complexes containing one radical ligand and

one dianionic ligand, the distance is in the range of 1.749 ± 0.003 Å (see chapter 2).

Comparing these values with the C–S bond lengths of 15 and 16 is possible to conclude that

two ligands are in their closed-shell form and one is in the ortho-dithiobenzosemiquinonate

(1-) state. However, it will be discussed on the basis of spectroscopic methods that no

evidence of ligand radical was found, thus the oxidation state assignment based on X-ray

crystallography is ambiguous in this case.

6.3 – Electro- and Spectroelectrochemistry

Figure 6.3.1 shows the cyclic voltammogram of 15 [Re(LTMS)3]1- obtained at 100 mV

s-1 in a CH2Cl2 solution with 0.10 M [N(n-Bu)4]PF6 as the supporting electrolyte, using a

glassy carbon working electrode and Ag/AgNO3 reference electrode. Ferrocene was used as

an internal standard, and potentials are referenced versus the Ferrocenium/Ferrocene couple

(Fc+/Fc) and listed in table 6.3.1. On the basis of coulometric studies, the CV of 15 displays

one fully reversible one-electron reduction at –1.886 V, yielding the dianion 15a, and two

reversible one-electron oxidations at –0.298 V and +0.271 V, resulting in the formation of

compounds 15b and 15c as expressed in Equation 6.3.1.

Chapter 6

163

Figure 6.3.1 – Cyclic voltammogram of 15 in CH2Cl2 solution at 25 °C containing 0.10 M

[N(n-Bu)4]PF6 as the supporting electrolyte at can rate of 100 mV s-1 (glassy carbon electrode,

potentials referenced vs the Ferrocenium/Ferrocene couple). The redox couples are given.

The absorption spectra of the one-electron transfer series of complex 15 [ReV(LTMS)3]1-

in CH2Cl2 solutions containing 0.10 M [N(n-Bu)4]PF6 at -25 °C is shown in Figure 6.3.2. The

absorption spectra are very complex. A large number of transitions are observed for all

species and the interpretation is not straightforward. Complexes 15 and 15a are very similar,

except for the band at 437 nm observed for 15, which is shifted to 473 nm in 15a. No

evidence for a π-radical ligand is observed in these spectra, supporting the assignment of a

-e

+e15a

S = ½

15b S = ½

15

S = 0

[ReV(LTMS)3]1- [ReIV(LTMS)3]

2- [ReV(LTMS•)(LTMS)2]0

[ReV(LTMS•)2(LTMS)]1+

-e

+e

15c

S = ½ Eqn. 6.3.1

1.0 0.5 0.0 -0.5 -1.0 -1.5 -2.0

E (V) versus Fc+

/Fc

5 µA

1+/0

0/1- 1-/2-

Chapter 6

164

ReV (d2) and ReIV (d3) for complexes 15 and 15a. In both compounds, the three ligands are in

their closed-shell configuration. In contrast, the spectra of 15 and 15b show significant

differences. The main feature in the absorption spectrum of 15b is the very intense band at

684 nm (ε = 3.89 x 104 M-1 cm-1) characteristic of an ortho-thiobenzosemiquinonate(1-) π-

radical ligand coordinated to the central rhenium ion. A similar transition at 661 nm is also

observed when the neutral compound 15b is oxidized to 15c [ReV(LTMS•)2(LTMS)]1+. The band

increases its intensity significantly, showing an extinction coefficient (ε) of 4.67 x 104 M-1

cm-1. It has been observed for the Ni complexes 2a [Ni(LTMS•)(LTMS)]1- and 2b [Ni(LTMS•)2]0

that the increase in the number of π-radical ligands coordinated to a metal ion leads to two

features: (1) higher extinction coefficients, and (2) blue shift of the band. Thus, the absorption

spectra 15b and 15c shown in Figure 6.3.2 are in agreement with a ligand based redox

process. The compounds can then be described as complexes containing a ReV (d2)

coordinated to one ligand π-radical in 15b and to two in 15c. The other ligands remain in their

closed-shell configuration.

Figure 6.3.2 – Absorption spectra of the one-electron transfer series of complex 15

[ReV(LTMS)3]1- in CH2Cl2 solutions containing 0.10 M [N(n-Bu)4]PF6 at -25 °C.

Figure 6.3.3 shows the cyclic voltammogram (CV) of complex 16 [ReV(LCl)3]1-

measured at 200 mV s-1 in a CH2Cl2 solution with 0.10 M [N(n-Bu)4]PF6 as the supporting

electrolyte, using a glassy carbon working electrode and Ag/AgNO3 reference electrode.

300 400 500 600 700 800 900 1000 11000

1

2

3

4

5

εε εε, 1

04 M

-1 c

m-1

λ,λ,λ,λ, nm

15 [ReV(LTMS)3]1-

15a [ReIV(LTMS)3]2-

15b [ReV(LTMS•)(LTMS)2]0

15c [ReV(LTMS•)2(LTMS)]1+

Chapter 6

165

Ferrocene was used as an internal standard, and potentials are referenced vs the

Ferrocenium/Ferrocene couple (Fc+/Fc) and are listed in table 6.3.1.

Figure 6.3.3 – Cyclic voltammogram of 16 in CH2Cl2 solution at 25 °C containing 0.10 M

[N(n-Bu)4]PF6 as the supporting electrolyte at can rate of 200 mV s-1 (glassy carbon electrode,

potentials referenced vs the Ferrocenium/Ferrocene couple).

In contrast to the CV of 15, the oxidation process is more complex in 16. The oxidation at

lower potential (peak potential of +0.45 V) seems to be coupled with a further irreversible

process with no reasonable peak separation. Equation 6.3.2 shows the redox series observed

for compound 16. According to coulometric studies, only a one-electron reduction to 16a was

observed. An attempt to generate complex 16b electrochemically was made. Less than 20% of

the coulommetry was completed before the sample decomposed, and the resulting spectrum is

shown in black in Figure 6.3.4. The absorption spectrum shows the changes in the equilibrium

of complex 16a, which begins conversion into the oxidized 16b. It is expected that the

intensity of the band at 680 nm increases considerably its intensity, due to the formation of a

ligand π-radical in the neutral complex 16b. In contrast, the starting material band at 724 nm,

1.0 0.5 0.0 -0.5 -1.0 -1.5

E (V) versus Fc+/Fc

5 µA

1+/0

0/1-

Chapter 6

166

should disappear if the coulometric process was performed completely, as observed for the

redox process of complex 15 (vide supra).

Figure 6.3.4 – Absorption spectra of the one-electron transfer series of complex 16

[ReV(LCl)3]1- in CH2Cl2 solutions containing 0.10 M [N(n-Bu)4]PF6 at -25 °C. The arrows

show the change in the spectrum of complex 16b discussed in the text (vide supra).

-e

+e16a

S = ½

[ReIV(LCl)3]2- [ReV(LCl)3]

1-

16

S = 0

-e [ReV(LCl•)(LCl)2]1-

16b

S = ½

Eqn. 6.3.2

300 450 600 750 900 10500

1

2

3

4

5

6

εε εε, 1

04

M-1

cm

-1

λ, λ, λ, λ, nm

16 [ReV(LCl)3]1-

16a [ReIV(LCl)3]2-

16b [ReV (LCl•)(LCl)2]0

Starting material 16

100% conversion of 16 to 16a

16% convertion of 16 to 16b

Chapter 6

167

Figure 6.3.5 shows the cyclic voltammogram (CV) and the square-wave

voltammogram (SWV) of complex 17b [ReV(L•)(L)2]0 in CH2Cl2. According to coulometric

studies, the CV shows one irreversible oxidation and two reversible reduction processes. The

oxidation of the neutral 17b (peak potential of +0.42 V) to the monocation 17c

[ReV(L•)2(L)]+1 is irreversible and does not show in the CV a reasonable peak separation from

the first reduction process. This process is more visible in the SWV measurement shown in

red.

Table 6.3.1 summarizes the redox potentials of the rhenium compounds. The effect of

the substituents in the phenyl ring is clearly reflected in the redox potentials. For example, the

redox potentials of the couple 0/1- increases in the order LTMS < L < LCl. Consequently, the

compounds can be easily oxidized by increasing the electron donating character of the

substituents.

Figure 6.3.5 – Cyclic voltammogram of 17 (black line) in CH2Cl2 solution at 25 °C

containing 0.10 M [N(n-Bu)4]PF6 as the supporting electrolyte at can rate of 100 mV s-1

(glassy carbon electrode, potentials referenced vs the Ferrocenium/Ferrocene couple). The red

line represents the square-wave voltammogram.

The absorption spectra of complex 17b and the one-electron reductions are shown in

figure 6.3.6.

1.0 0.5 0.0 -0.5 -1.0 -1.5

E (V) versus Fc+

/Fc

5 µA

1+/0 0/1- 1-/2-

Chapter 6

168

Figure 6.3.6 – Absorption spectra of the one-electron transfer series of complex 17b

[ReV(L•)(L)2]1- in CH2Cl2 solution containing 0.10 M [N(n-Bu)4]PF6 at 0 °C.

The intense band at 674 nm (ε = 3.89 104 M-1 cm-1) is characteristic of the compounds

containing π-radical ligand coordinated to the central metal ion, again in agreement with a

ReV (d2) electronic configuration. The spectra of the reduced species changes in the same

trend as observed for complex 16. Due to instability reasons, the generation of the monocation

17c was not successful. Equation 6.3.3 summarizes the redox processes observed for complex

17b.

-e

+e17a

S = ½

17b S = ½

17

S = 0

[ReV(L)3]1- [ReIV(L)3]

2- [ReV(L•)(L)2]0

[ReV(L•)2(L)]1+

-e

+e

17c

S = ½ Eqn. 6.3.3

300 400 500 600 700 800 900 1000 11000

1

2

3

4

5

εε εε , 1

04

M-1

cm

-1

λ,λ,λ,λ, nm

17 [ReV(L)3]1-

17a [ReIV(L)3]2-

17b [ReV (L•)(L)2]0

Chapter 6

169

Table 6.3.1 – Redox potentials of complexes 15, 16, and 17b in CH2Cl2 solutions 0.10 M

[N(n-Bu)4]PF6 at 25 °C.

1+/0 0/1- 1-/2-

Complex E1½ vs Fc+/Fc E2

½ vs Fc+/Fc E3½ vs Fc+/Fc

15 +0.271 -0.298 -1.886

16 - +0.293 -1.144

17b +0.420 -0.033 -1.441

Table 6.3.2 – Summary of the electronic spectra of complexes at -25 °C in CH2Cl2 solutions.

Complex λ max., nm (ε, 104 M-1 cm-1)

15 376 (2.54), 434 (1.26), 437 (1.15), 548 (0.87), 599 (0.85), 712 (1.09), 941 (0.24)

15a 377 (2.42), 434 (1.20), 473 (1.15), 548 (0.87), 599 (0.85), 729 (1.01), 941 (0.24)

15b 338(2.54), 459 sh. (0.87), 684 (3.89), 841 (0.23), 1026 (0.19)

15c 409 (2.45), 454 (1.67), 661 (4.67), 811 (0.27), 905 (0.22)

16 313 (5.90), 350 (2.07), 411 (2.22), 507 (0.56), 713 (1.81), 921 (0.57)

16a 315 (5.85), 360 (3.34), 429 (1.72), 603 (0.73), 658 (0.84), 737 (0.38)

16b 311 (5.55), 351 (1.95), 397 (2.09), 428 (1.79), 502 (0.52), 680 (1.48), 724 (1.46),

920 (0.44)

17 304 (3.19), 346 sh. (1.71), 429 (2.36), 502 (0.86), 742 (1.49), 942 (0.46)

17a 320 (4.28), 393 sh. (2.73), 429 sh. (2.36), 605 (0.83), 670 (0.90), 765 (0.44),

956 (0.12)

17b 390 (2.50), 442 sh. (1.54), 674 (3.89), 1062 (0.22)

Chapter 6

170

6.4 – Theoretical Calculations

DFT calculations were carried out with the B3LYP functional for complexes 15, 16,

17, 17a and 17b in order to provide a better understanding of the electronic structure of these

compounds. The substituents in the phenyl rings of complexes 15 and 16 were taken into

consideration in the calculations.

Structure Optimization:

The calculated Re–S, C–S, and C–C bond distances of complexes 15 and 16 are in

good agreement with the experimental values obtained from the X-ray crystallography data.

The bond distance errors do not exceed ± 0.03 Å. On the basis of the accuracy of the

calculated results for 15 and 16, it is expected that the calculated values for the electron-

transfer series of 17 reflect the real structural features in these molecules. Table 6.4.1 shows

the comparison between selected experimental (in parenthesis) and calculated bond distances.

The calculated C–S bond lengths of the monoanions 15, 16 and 17 are in average 1.76 Å, in

agreement with the experimental values of 1.74 Å, suggesting that all three complexes have

the same oxidation state. The relatively short C–S bond lengths between 1.74 and 1.76 Å are

also observed for other six-coordinated complexes. Recently Wieghardt et al.25 reported the

electronic structure of the [M(LBu)3]1-/0 couples (M = Mo or W; (LBu)2- = 3,5-di-tert-butyl-1,2-

benzenedithiolate(2-)). The monoanions are clearly comprised of a MV ion coordinated to

three closed-shell ligands with average C–S bond lengths of 1.76 Å. In contrast, the neutral

complexes of Mo and W possess a MV coordinated to two closed-shell ligands and one π-

radical (LBu•)1-. In this case, the average C–S bond lengths for the [Mo(LBu•)(LBu)2]0 and

[W(LBu•)(LBu)2]0 species are 1.74 and 1.75 Å respectively. Comparing the C–S bond

distances between the monoanions and the neutral complexes no significant differences are

observed when one ligand is oxidized. The shorter bond distance observed in systems

containing a coordinated C–S• is, in this case, averaged over the six sulfur atoms in complexes

15 and 16. The bond lengths from the crystal structures cannot be used to assign the oxidation

state of the ligands and metal unambiguously.

Chapter 6

171

Table 6.4.1 – Selected bond lengths (in Å) obtained by B3LYP DFT calculations.

Experimental values are given in parenthesis for comparison for complexes 15 and 16.

Crystallographic data for complexes 17, 17a and 17b are not available.

15 [ReV(LTMS)3]

1- 16

[ReV(LCl)3]1-

17 [ReV(L)3]

1- 17a

[ReIV(L)3]2-

17b [ReV(L•)(L)2]

0

Re(1)–S(1) 2.374 (2.341(2))

2.364 (2.3513(6))

2.378 2.406 2.371

Re(1)–S(2) 2.373 (2.358(2))

2.364 (2.3373(6))

2.376 2.383 2.370

Re(1)–S(11) 2.372 (2.330(2))

2.365 (2.3279(6))

2.379 2.376 2.372

Re(1)–S(12) 2.372 (2.345(2))

2.364 (2.3390(6))

2.377 2.473 2.371

Re(1)–S(21) 2.372 (2.330(2))

2.365 (2.3371(6))

2.377 2.472 2.371

Re(1)–S(22) 2.372 (2.345(2))

2.365 (2.3476(6))

2.378 2.402 2.370

S(1)–C(1) 1.763 (1.762(9))

1.757 (1.745(3))

1.757 1.760 1.741

S(2)–C(2) 1.764 (1.739(1))

1.758 (1.737(3))

1.758 1.765 1.741

S(3)–C(7) 1.762 (1.742(7))

1.757 (1.738(2))

1.757 1.770 1.741

S(4)–C(12) 1.763 (1.743(7))

1.758 (1.740(2))

1.758 1.759 1.741

S(5)–C(13) 1.762 (1.742(7))

1.757 (1.740(2))

1.757 1.761 1.741

S(6)–C(18) 1.763 (1.743(7))

1.758 (1.735(2))

1.758 1.768 1.741

C(1)–C(2) 1.409 (1.398(1))

1.417 (1.400(4))

1.409 1.419 1.412

C(2)–C(3) 1.423 (1.434(1))

1.422 (1.407(4))

1.410 1.412 1.414

C(3)–C(4) 1.403 (1.387(1))

1.400 (1.382(4))

1.410 1.410 1.414

C(4)–C(5) 1.423 (1.430(1))

1.407 (1.387(5))

1.394 1.396 1.387

C(5)–C(6) 1.403 (1.384(1))

1.400 (1.381(4))

1.407 1.407 1.414

C(6)–C(1) 1.422 (1.429(1))

1.417 (1.403(4))

1.394 1.396 1.387

Re

S S

S

SS

S

R

R

R

R

R R

1

2

3

45

6

7

1 2

1

3

45

6

11

1213

18

Chapter 6

172

The DFT calculations reproduced well other structural features such as the intraligand

S–Re–S angles (error < 1°), the twist angles θ (error < 3°), and the interligand S•••S bond

distances (error ± 0.03 Å). Table 6.4.2 summarizes the structural features of complexes 15,

16, 17, 17a, and 17b. It is very interesting that the experimental TP geometry of 15 and the

distorted TP of 16 are well reproduced by calculations. According to the calculations,

compounds 17 and 17a show geometry intermediate between TP and OCT, in contrast to the

calculated neutral species 17b that is almost TP.

Table 6.4.2 – Structural features of complexes obtained from DFT calculations. Experimental

values are given parenthesis for comparison.

15

[ReV(LTMS)3]1-

16

[ReV(LCl)3]1-

17

[ReV(L)3]1-

17a

[ReIV(L)3]2-

17b

[ReV(L•)(L)2]0

Angle

S(1)–Re(1)–S(2) 81.0° (81.7°)

83.1° (83.2°)

82.7° 84.2° 82.7°

S(3)–Re(1)–S(4) 80.9° (80.7°)

83.1° (83.1°)

82.7° 82.6° 82.7°

S(5)–Re(1)–S(6) 80.9° (80.7°)

83.1° (83.1°)

82.7° 82.2° 82.7°

S–Re–S av. 80.9° (81.0°)

83.1° (83.1°)

82.7° 83.0° 82.7°

Twist

θ1 0.01° (0.00°)

22.8° (26.5°)

24.9° 32.7° 0.35°

θ2 0.02° (0.7°)

22.9° (25.5°)

25.0° 33.1° 0.33°

θ3 0.03° (0.7°)

22.9° (22.3°)

24.8° 34.6° 0.34°

θav 0.02° (0.46°)

22.9° (24.8°)

24.9° 33.5° 0.34°

Interligand

S(1)•••S(3) 3.121 (3.103)

3.119 (3.119)

3.157 3.344 3.081

S(1)•••S(5) 3.121 (3.068)

3.123 (3.075)

3.163 3.259 3.079

S(3)•••S(5) 3.125 (3.103)

3.123 (3.122)

3.165 3.216 3.081

S(2)•••S(4) 3.129 (3.061)

3.131 (3.122)

3.171 3.301 3.087

S(2)•••S(6) 3.129 (3.061)

3.128 (3.037)

3.167 3.325 3.085

S(4)•••S(6) 3.131 (3.100)

3.133 (3.162)

3.168 3.206 3.086

S•••S av. 3.126 (3.082)

3.126 (3.106)

3.165 3.275 3.083

Chapter 6

173

Bonding Scheme and Ground State Properties:

The construction of qualitative MO schemes for the rhenium complexes can be

obtained taking into consideration the Re 5d, 6s, and 6p orbitals and 24 ligand orbitals, which

can have σ or π symmetry, as shown in Figure 6.4.1.

Figure 6.4.1 – σ- and π-ligand orbitals involved in bonding with the rhenium. The ligand is

simplified and only the ethenedithiolate moiety is represented.

When three of these ligands are around the rhenium central ion, four distinct bonding

combinations can be obtained by the:

1 – σ orbitals pointing towards dxz, dyz orbitals resulting in σ-bonds.

2 – σ orbitals interacting with the dz2 orbital leading to σ- and π-bonds.

3 – donation from the πv orbitals to the dxy and dx2-y

2 forming π-bonds.

4 – donation from the πv orbitals to the dxy and dyz giving π- and δ- bonds.

πv σ

Chapter 6

174

Three different electronic structures based on theoretical calculations have been proposed for

tris(dithiolate) complexes. In each model, distinct conclusions regarding the location of the

unpaired electron were obtained.

The first bonding scheme proposed by Schrauzer and Mayweg26,27 describes the

electronic configuration for the ground state of the complexes [M(LPh)3]1- (M = Cr, Mo, or

W). The bonding scheme derived from Hückel calculations is shown in Figure 6.4.2.

Figure 6.4.2 – Qualitative bonding scheme obtained from Hückel calculations (adaptation

from ref. 27 The red bar represents the SOMO.

Appling the same MO scheme for one of the neutral complexes 15b, 16b, or 17b the

corresponding configuration results in (3a1´)2(4e´)4(5e´)1. The 4e´ is a pair of orbitals resultant

of the linear combination between orbitals involving 41% of an antibonding π-ligand, 17% of

sulfur, 25% of d and 17% of p character from the metal (E´ symmetry). The unpaired electron

is located in the 5e´ MO, which results from the same combination of the 4e´ but with greater

dxy and dx2-y

2 metal character. In contrast, the LUMO 2a2´ is pure π-ligand and the relative

energies between the 5e´ and 2a2´ change significantly by changing the input parameters

employed in the Hückel calculations. EPR measurements on the [M(LPh)3]1- (M = Cr, Mo, or

W) showed significant metal character, which is in agreement with the ground state

configuration (3a1´)2(4e´)4(5e´)1. It is important to note that the authors ruled out the

configuration (3a1´)2(4e´)4(2a2´)

1 for the anionic complexes.

4e'

πh

6p

4e" dxz,yz

dx2-y2, xy

5e' 3πv

5d

E'

dz2 3a1'

2a2' 3πv

3πv

πh

3e' 3e" 2a2"

6s

A2'

Chapter 6

175

Another qualitative bonding scheme was proposed by Gray et al.10 and is shown in

Figure 6.4.3.

Figure 6.4.3 – Qualitative bond scheme adapted from ref. 10 The red bar represents the

SOMO.

The bonding scheme obtained by Gray to explain the electronic structure of the Eisenberg´s

[Re(LPh)3]0 complex is clearly different from that of Schrauzer. The ordering of the MOs are

different and the ground state configuration is described as (4e´)4(2a2´)2(3a1´)

1. The SOMO

3a1´ has a large Re 5dz2 contribution and the 4e´ is constituted by considerable 5d and

3πv- ligand character.

The EPR spectroscopy data and theoretical calculation results of [Re(LPh)3]0 and

[Re(LMe)3]0 performed by Porte et al.28 lead to a third description of the electronic structure.

The authors argued that the correct electronic structure is not consistent with either of the two

3πv

4e'

dz2

2a2' 3πv

3a1'

6s

4e" dxz, yz

5e'

3πv

5d

3πv

πh πh

πh πh

3e'

3e"

2a2"

6p

dx2-y2, xy

Chapter 6

176

first proposals, but is in agreement with the (3a1´)2(4e´)4(2a2´)

1 configuration rejected by

Schrauzer et al. Figure 6.4.4 shows the qualitative bonding scheme proposed by Porte. The

2a2´ in this case, is constituted by a pure nonbonding π-ligand orbital.

Figure 6.4.4 – Qualitative bonding scheme adapted from ref. 28 The SOMO of [Re(L)3]0 is

depicted on the right hand side. The phenyl rings are removed for clarity.

The presentation and discussion of the three distinct electronic structures proposed by

different authors is necessary in order to understand the results obtained by ZORA B3LYP

DFT calculations performed for the Re complexes discussed in this chapter. Figures 6.4.5 to

6.4.9 show the qualitative bonding schemes of compounds 15, 16, 17, 17a, and 17b,

respectively.

5d

3πv

4e'

dz2

2a2' 3πv

3a1'

6s

4e" dxz, yz

5e' 3πv

3πv

πh

πh

πh

πh

3e'

3e"

2a2"

6p

dx2-y2, xy

SOMO

Chapter 6

177

Figure 6.4.5 – Unrestricted Kohn-Shan MO diagram of the monoanion 15 [ReV(LTMS)3]1-

from the spin unrestricted ZORA B3LYP DFT calculations.

2e´´

1a´´1

2a´´2

4e´

3e´´

3a´´1

2a´´1

2a´ 5e´

4e´´

5e´´

(I) (II)

Chapter 6

178

Figure 6.4.6 – Unrestricted Kohn-Shan MO diagram of monoanion 16 [ReV(LCl)3]1- from the

spin unrestricted ZORA-B3LYP DFT calculations.

2e´´

1a´´1

2a´´2

4e´

3e´´

2a´´1

3a´´1

2a´2

5e´ (I) (II)

4e´´

5e´´

Chapter 6

179

Figure 6.4.7 – Unrestricted Kohn-Shan MO diagram of the monoanion 17 [ReV(L)3]1- from

the spin unrestricted ZORA B3LYP DFT calculations.

1a´´1

3e´´

2a´´2

2a´´1

4e´

3e´´

3a´´1

2a´2

5e´

4e´´

5e´´

(I) (II)

Chapter 6

180

Figure 6.4.8 – Unrestricted Kohn-Shan MO diagram of the dianion 17a [ReIV(L)3]2- from the

spin unrestricted ZORA-B3LYP DFT calculations.

3e´´

2e´´

4e´

2a´2

2a´2

3a´´1

5e´ (I)

5e´ (II)

5e´´

Chapter 6

181

Figure 6.4.9 – Unrestricted Kohn-Shan MO diagram of the neutral complex 17b

[ReV(L•)(L)2]0 from the spin unrestricted ZORA-B3LYP DFT calculations.

3e´´2

4e´´

2a´´1

3a´´1 4e´

2a´1

3a´1

2a´2 5e´

4e´´

5e´´

(I) (II)

Chapter 6

182

The calculation results obtained for the dianion 17a show a bonding scheme closer to that

reported by Schrauzer et al.27 In this case, the 2a2´ orbital is doubly occupied and its relative

energy is lower than the 3a1´ (dz2) as shown in Figure 6.4.8. In this paramagnetic species, the

unpaired electron occupies one of the 5e´ orbitals which have 65% of metal and 22 % of

sulfur contributions. This orbital is a result of mixing between 5dxy, 5dxz, and 5dx2-y

2 metal

orbitals, which can be explained in terms of the TP distortion (θav = 33.5°). Therefore, the

complex can be described as a ReIV (d3) metal ion coordinated to three closed-shell ligands.

When compound 17a is oxidized to the monoanionic complex 17 (Figure 6.4.7), the

electron is removed from the metal-centered 5e´ orbital. Due to oxidation, the effective

nuclear charge in the Re atom increases and the 3a1´ orbital (5dz2) is significantly stabilized

and a bonding scheme similar to that of Porte is obtained. Consequently, the electronic

structures reveal for the monoanionic compounds a 2a2´ ligand centered HOMO. These

nonbonding π-orbitals have approximately 91 to 96 % ligand character. The LUMOs are

composed of a pair of orbitals with 5e´ symmetry, which have almost equal metal and ligand

character. In the monoanionic compounds 15, 16, 17 (Figures 6.4.5 to 6.4.7) the Re 5dz2

orbital (3a1´ symmetry) is doubly occupied, compatible with the assignment of a ReV (d2)

central ion coordinated to three closed-shell ligands. The other four 5d orbitals remain

unoccupied and in the MO schemes represent the 5e´ and 5e´´ orbital pairs high in energy.

A further one-electron oxidation of the monoanionic species to the neutral complexes

15b, 16b, and 17b is, in this case, not metal but ligand centered, as exemplified clearly in the

bonding scheme of complex 17b in Figure 6.4.9. The oxidation process results in the removal

of one electron from the 2a2´ orbital (97% ligand character). The bonding scheme also reflects

the picture proposed by Porte et al.28, and the calculations support the assignment of a ligand-

based oxidation process shown by spectroelectrochemical studies in section 6.3. Figure 6.4.10

summarizes the features observed in the frontier orbitals. Table 6.4.1 summarizes the

composition of selected orbitals.

Chapter 6

183

Figure 6.4.10 – Representation of the frontier orbitals for complexes 17b, 17, and 17a.

Table 6.4.1 – Percentage composition of the selected orbitals of complexes 15, 16, 17, 17a

and 17b as obtained from B3LYP DFT calculations. The HOMO is written in bold.

Orbital 5dxy 5dxz 5dyz 5dx2-y

2 5dz2 3Spx,y 3Spz 2Cpx,y

2a2´ 74 22

15 5e´ (I) 47 26 7 5

5e´ (II) 48 29 2

2a2´ 63 2 30

16 5e´ (I) 41 4 30 7 9

5e´ (II) 4 41 30 7 9

2a2´ 66 30

17 5e´ (I) 49 27 7 6

5e´ (II) 50 27 7 7

2a2´ 59 10 23

3a1´ 52 16 6 13

17a 5e´ (I) 27 13 20 20 2 12

5e´ (II) 18 11 26 20 3 13

2a2´ 66 31

17b 5e´ (I) 48 30 6 6

5e´ (II) 49 28 7 6

2a2´ (L)

3a1´ (5dz2)

5e´ (5d)

3a1´ (5dz2)

2a2´ (L)

5e´ (5d)

3a1´ (5dz2)

2a2´ (L)

5e´ (5d)

17b [ReV(L

•)(L)2]

0 17 [Re

V(L)3]

1- 17a [Re

IV(L)3]

2-

Chapter 6

184

6.5 – X-ray Absorption Spectroscopy (XAS)

K- or L-edge X-ray absorption spectroscopy (XAS) is a powerful technique which

provides information about the oxidation state and the coordination chemistry about the

absorbing atom. The energy and intensity of the pre-edge transitions are sensitive to the

chemical form of the absorbing atom, thus fingerprinting chemical types. Solomon et al.

developed a methodology to provide a direct experimental probe of metal–sulfur bonding in

complexes and enzymes.29-33 The S 1s electron can be excited to unoccupied S 4p orbitals and

to the continuum by using tuneable synchrotron radiation at energy around 2471 eV, resulting

in an electric-dipole-allowed edge feature. Transitions to unoccupied metal-based orbitals are

typically at a lower energy than this feature and gain intensity through mixing with S 3p

orbitals. Specifically, the intensity of these pre-edge features, D0, is given by Equation 6.5.1.34

D0(S 1s → ψ*) = cte|⟨S 1s|r|ψ*⟩|2 = Is

Where r is the transition dipole operator, ψ* is the antibonding orbital corresponding to

metal–ligand bonding, ψ* = √1-α2|Md⟩ – α|S3p⟩, α2 is the covalency, i.e. amount of sulfur

character mixed into the metal d orbitals), h is the number of holes in the acceptor orbitals,

and n is the number of absorbing atoms. Is is the intensity of the electric dipole allowed

S 1s → 3p transition and has been shown to have a linear relationship with the S 1s → 4p

transition energy; therefore, Is can be estimated from experimental S K-edge data.35

Figure 6.5.1 show the S K-edge spectra of the monoanionic compounds 16, and 17.

Information about the oxidation state of the ligands can be obtained by analysing the region

between 2470 to 2472 eV. The absorption spectra of 17 and 17a show only one transition at

2470.9 eV, respectively, which is attributed to S 1s → 5e´ (LUMO) transition. The intensity

of 17a is less than 17 because the 5e´ orbital is partially filled in 17a, which decreases the

number of holes (h in Equation 6.5.1).36 In contrast, the spectrum of the neutral species 17b

[ReV(L•)(L)2]0 The first intense absorption (1) at 2470.1 eV is attributed to a transition from

the 1s to the singly occupied 2a2´ orbital (97% ligand character) and the large intensity is due

to a high sulfur contribution (66%) to this molecular orbital. The second absorption (2) at

2471.2 eV is basically the same S 1s → 5e´ (LUMO) transition observed in 17 and 17a but is

shifted to higher energy. Upon oxidation of compound 17, one electron is removed from the

2a2´ orbital with high sulfur character (66%). Consequently the effective nuclear charge on

the sulfur atom increases and the 1s orbital is more stabilized. Thus, more energy is required

α2h

3n Eq. 6.5.1

Chapter 6

185

for the transition. In an opposite situation, if the metal was oxidized, the effective nuclear

charge on the rhenium atom would increase, and the metal-based 5e´ orbitals are expected be

lower in energy, shifting the S 1s → 5e´ (LUMO) transition to lower energy.

Figure 6.5.1 – Sulfur K-edge spectra of the electron transfer series of complex 17 and the

corresponding second derivative. (1) and (2) represent the transitions of 17b discussed in the

text.

For all compounds shown in Figure 6.5.1, another intense transition is observed at

~ 2474 eV. This transition is attributed to a 1s → π*-ligand orbitals high in energy. DFT

calculations were performed for complexes 17 and 17a in order to confirm the transition

assignments. Figure 6.5.2 shows both experimental and calculated S K-edge spectra and a

schematic representation of the orbitals involved in the transitions.

17 [ReV(L)3]1-

17a [ReIV(L)3]2-

17b [ReV(L•)(L)2]0

2468 2470 2472 2474 2476

No

rma

lize

d A

bs

orp

tio

nS

eco

nd

Deri

vati

ve

Energy (eV)

1

2

1

2

Chapter 6

186

Figure 6.5.2 – S K-edge absorption spectra of complexes 17 and 17a. The lines indicate the

experimental spectra and the dashed lines represent the results obtained by DFT calculations

between 2469 and 2473.5 eV.

17 [ReV(L)3]1- 17a [ReIV(L)3]

2-

S 1s

2a2´

5e´

π* ligand

I II

S 1s

2a2´

5e´

π* ligand

I II

17 [ReV(L)3]1-

17a [ReIV(L)3]2-

II

I

Chapter 6

187

The S K-edge spectra of the monoanions 16 [Re(LCl)3]1- and 17 [Re(L)3]

1- are shown

in Figure 6.5.3. For compound 15, S K-edge spectra could not be obtained due to

decomposition of the sample in the beam line, probably caused by photoreduction. The

spectra are quite similar, what is expected comparing two compounds with similar structural

features and oxidation state. The first transition is observed at 2470.9 and 2471.0 for 16 and

17, respectively.

Figure 6.5.3 – S K-edge absorption spectra of the monoanionic complexes 16 and 17.

The metal L-edge absorption shows three distinct features assigned as L1, L2, and L3

edges. The L1-edge corresponds to the excitation of a 2s electron. The only observable feature

is the rising edge, which corresponds to an excitation to the continuum (ionization). The

energy of the rising edge is determined at the inflection point. No pre-edge features are

observed because the 2s → 5d transition corresponding to an excitation to the valence orbitals

is dipole forbidden.

The transitions involving the electrons of the metal 2p orbitals are split into L2- and

L3-edges due to the spin orbit coupling.37 The energies of the L2- and L3-edges vary

depending on the electronic structure of the site.38-40 A change in the effective nuclear charge

due to change in oxidation state or in coordination number, affects the energy of both the 2p

and 5d orbitals, while a change in the ligand field (which results in a change in the spitting of

the d-manifold) affects only the 5d orbital energies.

2468 2470 2472 2474 2476 2478

No

rma

lize

d a

bs

orb

an

ce

Energy (eV)

17 [ReV(L)3]1-

16 [ReV(LCl)3]1-

Chapter 6

188

Figure 6.5.4 shows the Re L1-edge of complexes 17, 17a, and 17b.

Figure 6.5.4 – Re L1-edge absorption spectra of the electron transfer series of compound 17.

The rising-edge in the region between 12525 and 12535 eV do not show significant

differences for compounds 17 and 17b. This behaviour indicates that the Re ion has the same

oxidation state in both complexes, which is in agreement with spectroscopic and DFT data. In

contrast, the difference of 1.5 eV between 17 and the dianion 17a is a result of a metal-

centered reduction. The reduction of ReV to ReIV decreases the effective nuclear charge at the

metal ion. Thus a shift in the rising edge to lower energy is expected. Table 6.5.1 summarizes

the energies of the Re L1-edge and the S K-edge absorptions. The transition energies of the Re

L1-edge were obtained by the first derivative of the spectra.

Table 6.5.1 – Summary of the transition energies of compounds 17, 17a, 17b, and 16.

S K-edge Re L1-edge

S 1s → 2a2´ S 1s → 5e´

17 [ReV(L)3]1- - 2470.9 12534.3

17a [ReIV(L)3]2- - 2470.9 12535.4

17b [ReV(L•)(L)2]0 2470.1 2471.2 12535.3

16 [Re(LCl)3]1- - 2471.1 -

12515 12520 12525 12530 12535 125400.0

0.2

0.4

0.6

0.8

No

rma

lize

d A

bs

orb

an

ce

Energy (eV)

17 [ReV(L)3]1-

17a [ReIV(L)3]2-

17b [ReV (L•)(L)2]0

Chapter 6

189

6.6 – Conclusions

Compounds 15 [ReV(LTMS)3]1- and 16 [ReV(LCl)3]

1- are the first monoanion rhenium

tris(dithiolate) complexes structurally characterized and they show distinctive geometries.

Compound 15 possess an almost perfect TP geometry, in contrast to 16, which has an

intermediate geometry between TP and OCT. The structural properties of 15 and 16, such as

the twist angle θ, the S–Re–S angles, and interligand S•••S distances were well reproduced in

the DFT calculations. It is noteworthy that the DFT calculations reproduced well the twist

angle θ for 15 and 16, indicating clearly that the geometry adopted by these compounds are

electronic in origin, and not necessarily a result of packing effects. The monoanionic

complexes 15, 16, and 17 are constituted by a ReV (d2) central ion coordinated to three closed-

shell ligands, as shown by UV-Vis, S K-edge spectroscopies in agreement with DFT

calculations. When these compounds are oxidized resulting in 15b, 16b, and 17b, the rhenium

remains in the +V oxidation state and the oxidation is ligand centered. These species show an

intense LMCT band at ~680 nm, characteristic for the presence of a coordinated π-ligand

radical. S K-edge of 17b shows an absorption at 2470.8 eV corresponding to 1s → 2a2´ (high

ligand character). This transition is not observed for 17 and 17a supporting the closed-shell

configuration of the ligands. In contrast, if the monoanionic compounds are reduced, the

redox process is metal centered and a ReIV (d3) is obtained. The Re L1-edge spectra show

clearly the difference in oxidation states. The main contribution of this chapter is the

experimental evidences for the presence of a π-ligand radical in the neutral species. Therefore,

Eisenberg´s complex [Re(S2C2Ph2)3]0 can be better described as [ReV(S2C2Ph2

•)(S2C2Ph2)2]0.

The investigation of this type of complexes and spectroscopic evidence for the presence of

coordinated π-ligand radical open new possibilities to the comprehension of the electronic

factors that lead to the adoption of a TP geometry.

Chapter 6

190

6.7 – References

1 King, R. B. J. Am. Chem. Soc. 1963, 2, 641-642. 2 Davison, A.; Edelstein, N.; Holm, R. H.; Maki, A. H. J. Am. Chem. Soc. 1964, 86, 2799. 3 Waters, J. H.; Williams, D. J.; Gray, H. B.; Schrauzer, G. N.; Finck, H. W. J. Am.

Chem. Soc. 1964, 86, 4198. 4 Schrauzer, G. N.; Mayweg, V.; Finck, H. W.; Müller-Westhoff, U.; Heinrich, W. Angew. Chem., Int. Ed. 1964, 3, 381. 5 Waters, J. H.; Williams, R.; Gray, H. B.; Schrauzer, G. N.; Finck, H. W. J. Am. Chem.

Soc. 1964, 86, 4198-4199. 6 Eisenberg, R.; Ibers, J. A. J. Am. Chem. Soc. 1965, 87, 3776-3778. 7 Cowie, M.; Bennett, M. J. Inorg. Chem. 1976, 15, 1584. 8 Kondo, M.; Minakoshi, S.; Iwata, K.; Shimizu, T.; Matsuzaka, H.; Kamigata, N.; Kitagawa, S. Chem. Lett. 1996, 489. 9 Schrauzer, G. N.; Mayweg, V.; Heinrich, W. J. Am. Chem. Soc. 1965, 87, 5798. 10 Steifel, E. I.; Eisenberg, R.; Rosenberg, R. C.; Gray, H. B. J. Am. Chem. Soc. 1966, 88, 2956. 11 Brown, G.; Stiefel, E. I. Inorg. Chem. 1973, 12, 2140-2147. 12 Cervilla, A.; Llopis, E.; Marco, D.; Pérez, F. Inorg. Chem. 2001, 40, 6525-6528. 13 Formitchev, D.; Lim, B. S.; Holm, R. H. Inorg. Chem. 2001, 40, 645-654. 14 Lim, B. S.; Donahue, J.; Holm, R. H. Inorg. Chem. 2001, 39, 263-273. 15 Martin, J. L.; Takats, J. Inorg. Chem. 1975, 14, 1358-1364. 16 Dymock, K. R.; Palenik, G. J. Inorg. Chem. 1975, 14, 1220. 17 Stiefel, E. I.; Brown, G. Inorg. Chem. 1972, 11, 434. 18 Bennett, M. J.; Cowie, M.; Martin, J.; Takats, J. J. Am. Chem. Soc. 1973, 95, 7504- 7505. 19 Cowie, M.; Bennett, M. J. Inorg. Chem. 1976, 15, 1589-1594. 20 Cowie, M.; Bennett, M. J. Inorg. Chem. 1976, 15, 1595-1603.

Chapter 6

191

21 Monkowius, U.; Nogai, S.; Schmidbauer, H. Z. Naturforsch., B: Chem. Sci. 2004, 59b, 259-263. 22 Brown, G.; Stiefel, E. I. Inorg. Chem. 1973, 12, 2140. 23 Stiefel, E. I.; Schulman, J. M. Prog. Inorg. Chem. 2004, 52, 55-110. 24 Stiefel, E. I.; Dori, Z.; Gray, H. B. J. Am. Chem. Soc. 1967, 89, 3353. 25 Kapre, R. R.; Bothe, E.; Weyhermueller, T.; DeBeer George, S.; Wieghardt, K. Inorg.

Chem. 2007, 46, 5642-5650. 26 Rosa, E. J.; Schrauzer, G. N. J. Phys. Chem. 1969, 73, 3132. 27 Schrauzer, G. N.; Mayweg, V. J. Am. Chem. Soc. 1966, 88, 3235. 28 Al-Molawi, A. H.; Porte, A. L. J. Chem. Soc. Dalton Trans. 1975, 250-252. 29 Glaser, T.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. Acc. Chem. Res. 2000, 33, 859. 30 Randall, D. W.; DeBeer George, S.; Holland, P. L.; Hedman, B.; Hodgson, K. O.; Tolman, W. B. J. Am. Chem. Soc. 2000, 122, 11632. 31 Shadle, S. E.; Penner-Hahn, J. E.; Schugar, H. J.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. J. Am. Chem. Soc. 1993, 115, 767. 32 Solomon, E. I.; Hedman, B.; Hodgson, K. O. J. Am. Chem. Soc. 1990, 112, 1643. 33 Solomon, E. I.; Hedman, B.; Hodgson, K. O.; Dey, A.; Szilagyi, R. K. Coord. Chem.

Rev. 2005, 249, 97. 34 Sarangi, R.; DeBeer George, S.; Rudd, D.; Szilagyi, R. K.; Ribas, X.; Rovira, C.; Almeida, M.; Hodgson, K. O.; Hedman, B.; Solomon, E. I. J. Am. Chem. Soc. 2007, 129, 2316-2326. 35 Szilagyi, R. K.; Lim, B. S.; Glaser, T.; Holm, R. H.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. J. Am. Chem. Soc. 2003, 125, 9158-9169. 36 Tenderholt, A. L.; Szilagyi, R. K.; Holm, R. H.; Hodgson, K. O.; Hedman, B.; Solomon, E. I. Inorg. Chem. 2008, 47, 6382-6392. 37 DuBois, J. L.; Mukherjee, P.; Solomon, E. I.; Stack, T. D. P.; Hodgson, K. O. J. Am.

Chem. Soc. 2000, 122, 5775. 38 Cramer, S. P.; deGroot, F. M. F.; Ma, Y.; Chen, C. T.; Sette, F.; Kipke, C. A.; Eichhorn, D. M.; Chan, M. K.; Armstrong, W. H. J. Am. Chem. Soc. 1991, 113, 7937. 39 George, S. J.; Lowery, M. D.; Solomon, E. I.; Cramer, S. P. J. Am. Chem. Soc. 1993, 115, 2968.

Chapter 6

192

40 Wasinger, E. C.; deGroot, F. M. F.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. J.

Am. Chem. Soc. 2003, 125, 12894.

Chapter 7

193

Chapter 7

Experimental

Chapter 7

194

Chapter 7

195

Experimental Section 7.1 – Physical Measurements

Elemental analysis

Elemental analyses were performed by H. Kolbe at the Microanalysis Laboratorium,

Mülheim an der Ruhr, Germany.

NMR spectra

1H and 13C spectra were recorded with a Bruker ARX 250 spectrometer (1H at 400

MHz). 1H and 13C spectra (at 100 MHz) were referenced to TMS, using the 13C or residual

proton signals of the deuterated solvents as internal standards.

Mass spectroscopy

Mass spectra in the Electron Impact mode (EI 70 eV) were recorded with a Finnigan

MAT 8200 mass spectrometer. Only characteristic fragments are given. The spectra were

normalized against the most intense peak, which therefore has intensity 100. Electron Spray

Interface (ESI) mass spectra were recorded with a Bruker Esquire 3000 instrument. The mode

(positive or negative) and the used solvents are given in parentheses.

UV-Visible spectrometry

UV-Visible spectra were recorded on either a Perkin-Elmer Lambda 19

spectrophotometer or on a Hewlett Packard HP 8452A diode array spectrophotometer in

various solvents. For UV-Vis spectroelectrochemical investigations, the HP 8452A diode

array spectrophotometer was used, by employing a coulommetry cuvette with [N(n-Bu)4]PF6

as supporting electrolyte and either MeCN or CH2Cl2.

Electrochemistry

Cyclic voltammograms and square wave voltammograms in the range of -25 to 25 °C

were recorded by using an EG&G Potentiostat / Galvanostat 273A. A three electrode cell was

employed with a glassy-carbon working electrode, a platinum-wire auxiliary electrode and a

Ag/AgNO3 reference electrode (0.01 M AgNO3 in CH3CN). Ferrocene was added as an

internal standard after completion of the measurements and potentials are referenced versus

the Fc+/Fc couple.

Chapter 7

196

Controlled potential coulometric measurements in a setup, which allows recording of

absorption spectra in situ during electrolysis, were performed by employing the same

potentiostat, but using a Pt-grid as a working electrode. A Pt-brush was used as counter

electrode and separated from the working electrode compartment by a Vycor frit. An

Ag/AgNO3 (0.01 M AgNO3 in CH3CN) reference electrode was employed again. Cyclic

voltammogram simulations were carried out using DIGISIM 3.03 from Bioanalytical Systems

– West Lafayette 2701 Kent Avenue, Indiana – USA.

EPR spectroscopy

Electron Paramagnetic Resonance spectra were recorded on a Bruker ELEXSYS E500

CW-Spectrometer, equipped with an ER 041 XK-D microwave bridge (X-band EPR spectra),

a helium flow cryostat (Oxford Instruments ESR 910) and a Hewlett Packard frequency

counter HP5253B. All EPR measurements were recorded on frozen solutions generated by

controlled potential coulometry or 10-3 M prepared solutions. The EPR spectra were

simulated with a self-written program (ESIM, by Dr. Eckhard Bill) for powder spectra from

spin S=½ systems with anisotropic g-tensor and Gaussian or Lorentzian line shape

distribution. Anisotropic magnetic hyperfine coupling was treated in first-order

approximation.

X-ray crystallography

Suitable crystals were coated with perfluoropolyether and picked up with glass fibers.

The specimens were immediately mounted in the nitrogen cold stream (100K) of a Nonius

Kappa-CCD diffractometer equipped with a Mo-target rotating-anode X-ray source and a

graphite monochromator (Mo Kα, λ = 0.71073Å). Final cell constants were obtained from

least-squares fits of all measured reflections. The Siemens ShelXTL software package was

used for solution and artwork of the structure. The refinement was performed using the

ShelXL97 program package. The structures were readily solved by Patterson methods and

subsequent difference Fourier techniques. All non-hydrogen atoms were refined

anisotropically. Hydrogen atoms placed at calculated positions and refined as riding atoms

with isotropic displacement parameters.

Chapter 7

197

Magnetic Susceptibility measurements

Measurements of temperature dependent magnetization of samples were performed in

the range 2 to 295K at 1 T on a Quantum Design SQUID Magnetometer MPMS. The samples

were encapsulated in spherical gelatine capsules. The response functions were measured four

times for each given temperature, yielding a total of 32 measured points. Diamagnetic

contributions were estimated for each compound using Pascal´s constants. The resulting

volume magnetization of the samples was than compensated for diamagnetic contributions

and recalculated as volume susceptibilities. The experimental results were fitted with JULIUS

program package, calculating through full-matrix diagonalization of the appropriate Spin-

Hamiltonian.

GC-MS analysis

GC of the organic compounds were performed either on HP 5890 II or HP 6890

equipments using RTX-1701 15m S-41 or RTX-5 Amine 13.5m S-63 columns respectively.

GC-MS analyses were performed using the above columns coupled with a HP 5973 mass

spectrometer with mass selective detector.

DFT calculations

The density functional theoretical calculations have been carried out by

employing B3LYP level of DFT. The all-electron Gaussian basis sets were developed by the

Ahlrichs group. For the calculated complexes, triple-ξ quality basis set TZV(P) with the one

set of polarization functions on the nickel, sulfur and silicon atoms. For the carbon and

hydrogen atoms, slightly smaller polarized split-valence SV(P) basis set were used that were

double-ξ quality in the valence region and contained a polarizing set of d-functions on the

non-hydrogen atoms. Auxiliary basis sets used to expand the electron density in the

resolution-of-the identity (RI) approach, where applicable, were chosen to match the orbital

basis. The SCF calculations were tightly converged (1 x 10-8 Eh in energy, 1 x 10-7 Eh in the

density change and 1 x 10-7 in maximum element of the DIIS error vector). The geometry

search for the complexes was carried out in redundant internal coordinates without imposing

the symmetry constraints. In this case the geometry was considered converged after the

energy change was less than 5 x 10-6 Eh, the gradient norm and maximum gradient were

smaller than 1 x 10-4 EhBohr-1 and 3 x 10-4 EhBohr-1, respectively, and the root-mean square

Chapter 7

198

and maximum displacements of all atoms were smaller than 2 x 10-3 Bohr and 4 x 10-3 Bohr

respectively.

Chapter 7

199

7.2 – Synthesis

Synthesis of 1,2-Bis(isopropylmercapto)benzene

The 1,2-Bis(isopropylmercapto)benzene was synthesized according to Sato´s

procedure with modifications.1 2-Bromopropane (2.63g, 21.1 mmol) was added dropwise

with stirring to a solution of of benzene (10 mL) and water (10 mL) containing 1,2-

benzenedithiol (H2L) (1.0 g, 7.0 mmol), NaOH (1.12 g, 28.1 mmol) and methyltributyl

ammonium chloride [MeN(n-Bu)3]Cl (75% solution in water, 96 mg, 0.21 mmol). After

stirring for 48 h at rt, the solution was treated with water and acidified with 20 mL of a 20%

HCl solution. The reaction mixture was extracted with benzene (4 x 10 mL). The organic

layer was washed with water, dried over MgSO4, filtered and concentrated under reduced

pressure yielding a light yellow oil.

Yield: 1.48 g (93%)

C12H18S2 = 226.24 gmol-1

Spectroscopic and microanalysis data agree with reported values.

Chapter 7

200

1,2-Bis(isopropylmercapto)-3,6-bis(trimethylsilyl)benzene (1)

The compound 1 was synthesized based on a procedure reported by Figuly and Martin

with modifications.2,3 1,2-Bis(isopropylmercapto)benzene (1.0 g, 4.4 mmol) in 5 mL of

n-hexane was added dropwise to a stirred mixture of n-butyllithium (1.04 g, 22 mmol, 2.5 M

in hexanes) and TMEDA (2.56 g, 22 mmol) at -50 °C. The cold bath was immediately

removed and 6 equiv. (2.87 g, 26.4 mmol) of freshly distilled trimethylsilanechloride

Si(CH3)3Cl was added. After 1 h of stirring at 25 °C, the solution was treated with 10%

aqueous HCl (30 mL). The reaction mixture was extracted with n-hexane, dried over Na2SO4

and concentrated under reduced pressure. This step yields 1.22 g (93%) of 1,2-

bis(isopropylmercapto)-3-trimethylsilylbenzene as a yellow oil and a small amount of the

disubstituted compound 1. The above procedure was repeated by using the crude oily product

to accomplish the further substitution. The final compound was recrystallized from n-hexane..

The synthesis was done in the fume hood and in closed containers due to the extremely

pungent odor. All glassware and apparatus were cleaned with a basic NaOCl or acidic H2O2

solutions.

Yield: 1.39 g (92%) colorless crystals of 1

Molecular weight = 370.255 g mol-1

Elemental analysis:

C18H34S2Si2 % C % H % S % Si

Calculated 58.3 9.2 17.3 15.2

found 58.6 9.5 17.0 15.2

GC-MS: m/z = 370(80), 355(28), 297(23), 255(100), 73(80).

1H NMR (CDCl3 300K): δ = 0.34 (s, 18H), 1.12 (d, 12H), 3.95 (sept, 2H), 7.34 (s, 2H).

13C NMR (CDCl3 300K): δ = 1.1 (Si–Cme), 22.7 (S–Cisop), 38.6 (CH3 isop), 133.5 (CAr), 146.4

(CAr), 149,2(CAr).

Chapter 7

201

Synthesis of 3,6-bis(trimethylsilyl)benzene-1,2-dithiol (1a)

Compound 1 (0.2 g, 0.54 mmol) was placed in a three-neck-round-bottom flask,

cooled at -78 °C under Argon and 60 mL of NH3 was liquefied. Under continuous stirring,

small pieces of metallic sodium were added until the dark blue color remains for at least 4 h.4

The reaction mixture was warmed up to room temperature to allow the complete evaporation

of NH3. The residual NH3 removed under vacuum and 60 mL of degassed HCl (10% solution)

was added dropwise. The desired compound was extracted from the aqueous phase with small

portions of degassed diethyl ether, dried over Na2SO4, filtered under Argon and dried under

vacuum yielding quantitatively compound 1a (colorless oil), which was used directly for

reactions with transition metals salts. The colorless compound 1a is extremely air-sensitive.

Traces of oxygen is enough to change the oil color to yellow, indicating that compound 1a is

being oxidized to 1ox.

Dipotassium-3,6-bis(trimethylsilyl)benzene-1,2-dithiolate (1b)

In a typical reaction, compound 1a (150 mg, 0.5 mmol) was suspended in degassed

MeOH or MeCN followed by the addition of solid KOtBu (56 mg, 1 mmol). The mixture was

stirred for 30 – 40 minutes resulting in an orange solution which was filtered through celite.

The filtrate was used directly for the synthesis of complexes.

Chapter 7

202

Synthesis of the nickel complex [K2(µµµµ-OHCH3)2(MeCN)2][2]

The ligand 1b (150 mg, 0.5 mmol) in degassed MeOH was added dropwise to a

solution of NiCl2 (34 mg, 0.26 mmol) and [N(n-Bu)4]I (210 mg, 0.57 mmol) in 5 mL of

degassed MeOH. After 2h of stirring under Argon a light red solution was formed. The

volume was reduced to half and 2 mL of MeCN was added. Light red crystals of

[K2(µ-OHCH3)2(MeCN)2][2] were obtained from concentrated MeOH/MeCN solutions at

-30°C.

Yield: 178 mg (81%).

Molecular weight: 852.26 g mol-1

Elemental analysis for [K2(µ-OHCH3)2(MeCN)2][2]:

C30H54K2N2NiO2S4Si4 % C % H % N

Calculated 42.28 6.38 3.29

found 41.89 6.22 3.18

Chapter 7

203

Synthesis of the nickel complex [N(n-Bu)4][2a]

Procedure I: The ligand 1b (150 mg, 0.5 mmol) was added dropwise to a solution of

NiCl2.6H2O (62 mg, 0.26 mmol) and [N(n-Bu)4]I (145 mg, 0.39 mmol) in 5 mL of degassed

MeOH and stirred for 2h under Argon yielding a light red solution. Upon exposure to a

stream of air the color of the reaction mixture changed to dark green. The solvent was

removed under vacuum and the solid was redissolved with MeCN. Dark green crystals of

[N(n-Bu)4][2a] were obtained from concentrated MeCN solutions at -20 or 4 °C.

Yield: 208 mg (92%).

Procedure II: MeOH/MeCN solutions of [K2(µ-OHCH3)2(MeCN)2][2] were exposed to air

yielding the dark green salt of 2a after slow evaporation of the solvent.

Yield: 200 mg (88%).

Molecular weight: 870.33 g mol-1

Elemental analysis for [N(n-Bu)4][2a]:

C40H76NNiS4Si4 % C % H % S

Calculated 58.3 9.2 17.3

found 58.6 9.5 17.0

ESI (in CH2Cl2 solution) : m/z = 626.4 -, 242.1 {M}+

Chapter 7

204

Synthesis of [2b]•CH2Cl2

The [N(n-Bu)4][2a] salt (100 mg, 0.15 mmol) was dissolved In 10 mL of CH2Cl2 and

tris-(4-bromophenyl)aminium hexachloroantimonate (122 mg, 0.15 mmol) was added and

stirred for 2 h. The color of the reaction mixture changed from dark green to purple with a

black precipitate. The residual precipitate was filtered and the solution was kept at -20°C.

Purple crystals of [2b]•CH2Cl2 were obtained from concentrated CH2Cl2 solutions.

Yield: 63 mg (67%).

Molecular weight: 626.4 g mol-1

Elemental analysis for [2]•CH2Cl2:

C25H42Cl2NiS4Si4 % C % H % S

Calculated 58.3 9.2 17.3

found 58.6 9.5 17.0

EI (in CH2Cl2 solution): m/z = 626.2

Chapter 7

205

Synthesis of the copper complex [N(n-Bu)4][3]

The ligand 1b in 10 mL of degassed MeOH (0.15 g, 0.5 mmol) was added dropwise to

a MeOH solution of Cu(CH3COO)2.H2O containing 145 mg (0.39 mmol) of [N(n-Bu)4]I. The

mixture was stirred for 2h under Argon yielding a light green solution. The solvent was

removed under vacuum and the solid was redissolved with CH2Cl2. Dark green crystals of

[N(n-Bu)4][3] were obtained from concentrated CH2Cl2 solutions at -20°C.

Yield: 205 mg (90%).

Molecular weight: 875.16 g mol-1

Elemental analysis for [N(n-Bu)4][3]:

C40H76CuNS4Si4 % C % H % N

Calculated 54.89 8.74 1.60

found 55.05 8.94 1.32

ESI (in CH2Cl2 solution) : m/z = 631.4 {M}-, 242.2 {M}+

Chapter 7

206

Synthesis of the gold complex [N(n-Bu)4][4]

To 150 mg (0.5 mmol) 1b suspended in 10 mL of degassed MeOH was added

dropwise a solution of Na[AuCl4].H2O (103 mg, 0.26 mmol) and [N(n-Bu)4]I (145 mg, 0.39

mmol) in 5 mL of degassed MeOH and stirred for 2h under Argon yielding a light green

solution. The solvent was removed under vacuum and the solid was redissolved with CH2Cl2.

Dark green crystals of [N(n-Bu)4][4] were obtained from concentrated CH2Cl2 solutions at

4°C.

Yield: 191 mg (73%).

Molecular weight: 1008.58 g mol-1

Elemental analysis for [N(n-Bu)4][4] :

C40H76AuNS4Si4 % C % H % N

Calculated 47.59 7.53 1.39

found 47.21 7.49 1.33

ESI (in CH2Cl2 solution) : m/z = 766.38 {M}-, 242.2 {M}+

Chapter 7

207

Synthesis of the palladium complex 5 [PdII

(tbpy)(L

TMS)]

To the dipotassium salt 1b (200 mg, 0.67 mmol) in 10 mL MeCN was added 300 mg

(0.67 mmol) of Pd(tbpy)Cl2 under vigorous stirring resulting in a purple color solution. The

reaction mixture was stirred for 6 h at rt and filtered through celite. The solvent was removed

under vacuum und the resulting purple solid redissolved in CH2Cl2. After 3 days purple

crystals of 5 suitable for x-ray crystallography were obtained.

Yield: 410 mg (93%)

Molecular weight: 658.78 g mol-1

Elemental analysis for 5 [PdII(tbpy)(LTMS)]:

C30H44N2PdS2Si2 % C % H % N

Calculated 54.81 6.43 4.26

found 55.05 7.08 4.02

EI (in CH2Cl2 solution) : m/z = 658.0

Chapter 7

208

Synthesis of the platinum complex 6 [PtII

(tbpy)(L

TMS)]

To 200 mg (0.67 mmol) of 1b in 10 mL MeCN were added to a solution of Pt(tbpy)Cl2

(358 mg, 0.67 mmol) under vigorous stirring, resulting in a purple solution. The reaction

mixture was stirred for 6 h at rt and filtered through celite. The solvent was removed under

vacuum and the resulting purple solid redissolved in CH2Cl2. After 3 days purple crystals

suitable for X-ray crystallography were obtained.

Yield: 470 mg (94%)

Molecular weight: 747.44 g mol-1

Elemental analysis for 6 [PtII(tbpy)(LTMS)]:

C30H44N2PtS2Si2 % C % H % N

Calculated 48.51 5.58 3.81

found 48.20 5.66 3.74

ESI (in CH2Cl2 solution) : m/z = 747.2 {M}+

Chapter 7

209

Synthesis of the platinum complex 7 [PtII

(tbpy)(L

Ph)]

To a 40 mL acetone/water (15:1) suspension of [PtII(tbpy)Cl2] (0.38 g, 0.71 mmol) a

cold dioxane solution of the ligand thiophosphoric ester5 was added (~0.6 mmol/L, 1.2 mL).

The reaction mixture was refluxed at 65 °C for 4 h under argon while its color changed from

yellow to greenish-blue. The precipitate product was filtered off and washed with high

amount of MeOH and some toluene. It was dried on the filter and washed with CH2Cl2. The

solution was evaporated to give 7 as a blue crystalline product.

Yield: 0.15 g (24%).

Molecular weight: 817.6 g mol-1

Elemental analysis for 7 [PtII(tbpy)(LPh)]:

C40H50N2PtS2Si2 % C % H % N

Calculated 57.6 6.0 3.3

found 57.6 6.0 3.1

ESI (in CH2Cl2 solution) : m/z = 817.4 {M}+

1H NMR (CDCl3 300K): δ = 1.28 (s, 18H), 1.42 (s, 18H), 7.24 (s, 4H), 7.46 (s, 4H), 7.58

(d, 2H), 7.83 (s, 2H), 9.25 (s, 2H).

Chapter 7

210

Synthesis of the dimer 7c [PtII

2(tbpy)2(L

Ph•)2](PF6)2

Ferrocenium hexafluorophosphate (19.8 mg, 0.06 mmol) was dissolved in CH2Cl2

(40 mL) under argon, and 7 (49 mg, 0.06 mmol) was added. The mixture was stirred for 30

min at rt. The solvent was evaporated, and the solid residue was washed with n-pentane

several times. Suitable crystals for X-ray analysis were obtained from CH2Cl2 solutions

layered with n-hexane.

Yield: 45 mg (75%)

Molecular weight: 1107.10 g mol-1

C80H100N4Pt2S4F12P2 % C % H % N

Calculated 49.9 5.2 2.9

found 50.3 5.2 2.8

Synthesis of the complex 8 [PtII

(PPh3)2(LPh

)]

This compound was synthesized according to modifications in the procedure described

by Bowmaker et al.6 starting with complex 9 [PtII(LPh•)2] (0.16 g, 0.18 mmol) and 0.95 g (3.6

mmol) of PPh3. The product was recrystallized from a CHCl3/EtOH mixture.

Yield: 0.153 g (80%)

Molecular weight: 1073.4 g mol-1

C58H56PtS2P2 % C % H % N

Calculated 64.9 5.3 -

found 65.0 5.3

1H NMR (CDCl3 300 K): δ = 1.18 (s, 18H), 6.95 (dd, 6H), 7.12 (t, 12H), 7.24 (dd, 8H), 7.50

(m, 12H).

Chapter 7

211

Synthesis of the cobalt complex [N(n-Bu)4][10]

To a solution of 1b (0.15 g, 0.5 mmol) in 25 mL of degassed MeOH was added

dropwise Co(CH3COO)2.4H2O (65 mg, 0.26 mmol) and [N(n-Bu)4]I (145 mg, 0.39 mmol) in

5 mL of degassed MeOH and stirred for 2 h under Argon. Upon exposure to air the light

green color of the reaction mixture changed to dark blue. The solvent was removed under

vacuum and the solid was redissolved in MeCN. Dark blue crystals of [N(n-Bu)4][10] were

obtained from concentrated MeCN solutions at -20 or 4 °C.

Yield: 213 mg (94%).

Molecular weight: 870.55 gmol-1

Elemental analysis for [N(n-Bu)4][10] :

C40H76CoNS4Si4 % C % H % N

Calculated 58.20 8.79 1.61

found 58.32 8.73 1.65

ESI (in CH2Cl2 solution) : m/z = 627.4 {M}-, 242.2 {M}+

Chapter 7

212

[Rh(CH3COO)2]2

1.0 g (3.69 mmol) of rhodium trichloride hydrate and 2.0 g (15 mmol) of sodium

acetate trihydrate in 20 mL glacial acetic acid and 20 mL absolute ethanol were gently

refluxed under nitrogen for 4 h. The initial red solution rapidly became green, and a green

solid was deposited. After cooling to rt the green solid was collected by filtration through a

Büchner funnel. The filtered supernatant was disposed of. The crude product is dissolved in

370 mL of boiling methanol and filtered; after concentration to about 80 mL the solution was

kept overnight at +4°C. After collection of the crystals, the solution is concentrated and

cooled to yield a further small amount of the methanol adduct [Rh(CH3COO)2]2.2CH3OH.

The blue-green adduct was heated under vacuum at 45°C for 20h to yield emerald-green

crystals of [Rh(CH3COO)2]2. Infrared spectra were performed periodically in order to check

the complete absence of methanol.

Yield: 653 mg (80% based on Rh).

Molecular weight: 220.9 g mol-1

K4[Rh2(CO3)4].2H2O

653 mg of [Rh(CH3COO)2]2 was suspended in 20 mL of 3 M potassium carbonate

solution. After 15 minutes the color changed to dark blue. The solution was heated to near

boiling and held at 100°C for 5h. Upon cooling and filtering, a precipitate formed. The

resulting solid was washed with cold water until it just started to dissolve to remove the

excess of carbonate. The dark blue solid was rinsed with small portions of methanol and

diethyl ether and air-dried overnight.

Yield: 684mg (89% based on Rh).

Molecular weight: 584.16 g mol-1

Chapter 7

213

[Rh2(CH3CN)10](BF4)4

Under nitrogen, a flask charged with 680 mg (1 mmol) of K4[Rh2(CO3)4].2H2O and

30 mL of CH3CN was stirred at rt for 30 min and 10 mL of a diethyl ether solution of HBF4

(54%) was added dropwise. The blue solution changed to pink upon addition of acid. The red-

orange solution was refluxed for 3 days. The orange supernatant was collected by filtration

and 320 mL of THF was added under stirring causing an orange precipitate to form. The

orange solid was collected by filtration and washed with THF and diethyl ether. The orange

compound was redissolved in a minimal amount of hot CH3CN and the orange solution was

then layered on top of THF. Needle-orange crystals of [Rh2(CH3CN)10](BF4)4 grew overnight.

The compound is stable in air, but under moisture two of the equatorial acetonitrile ligands

can be replaced by water yielding the pink compound [Rh2(CH3CN)8(H2O)2](BF4)4 detected

by mass spectroscopy.

Yield: 752 mg (92%)

Molecular weight of [Rh2(CH3CN)10](BF4)4 : 963.84 gmol-1

ESI (in CH2Cl2 solution) : m/z = 616.6 {M}+, 86.81 {M}-

Chapter 7

214

Synthesis of rhodium complex [N(n-Bu)4]2[11]•4MeCN

To a solution of 1b (0.15 g, 0.5 mmol) in 10 mL of degassed MeCN was added

dropwise to 240 mg (0.25 mmol) of finely powered [Rh2(CH3CN)10](BF4)4 and 145 mg

(0.39 mmol) of [N(n-Bu)4]I. the mixture was stirred for 4 h yielding a brown-redish solution.

The crude reaction mixture was filtered under celite and concentrated to 3 mL. Dark red

crystals of [N(n-Bu)4]2[11]•4MeCN were obtained from the concentrated MeCN solution at

-30°C.

Yield: 170 mg (67%).

Molecular weight: 1321.20 g mol-1

Elemental analysis for [N(n-Bu)4]2[11]•4MeCN:

C64H124N6RhS4Si4 % C % H % N

Calculated 58.13 9.45 6.36

found 58.12 9.45 6.34

Chapter 7

215

Synthesis of 12a cis-{RhIIII2[LTMS(CH3)2][LTMS(CH3)]} + 12b cis-{RhIIICH3I[L

TMS(CH3)2][LTMS(CH3)]}

To compound 11 (100 mg, 75.7 µmol) in 6 mL of MeCN was added Na/Hg amalgam

containing 20% of Na (18 mg, 80 µmol) and stirred at rt for 16 h under argon, yielding the

trianion 11b [RhI(LTMS)2]3-. The color of the reaction mixture changed to purple and the

excess of amalgam was removed through filtration under celite. MeI (11 µL, 75.6 µmol or

61 µL, 0.42 mmol) was added to the purple solution and stirred for 1 h. After 30 min the color

changed to orange which did not change when exposed to air. Orange crystals were obtained

by slow evaporation of the solvent at rt.

Synthesis of the square planar chromium complex [N(n-Bu)4]2[13]•4MeCN

To 150 mg (0.5 mmol) of ligand 1b in 10 mL of degassed MeCN was added 100 mg

(0.26 mmol) of finely powered [CrII(CH3CN)4(BF4)2], 1 mL (1 mmol) of Lithium-

trietylborohydride (super hydride® 1 M solution in THF) and 145 mg (0.39 mmol) of

[N(n-Bu)4]I. The mixture was stirred for 4 h yielding a green solution. The crude reaction

mixture was filtered and orange crystals of [N(n-Bu)4]2[13] were obtained from the

concentrated MeCN solution at -30°C.

Yield: 256 mg (78%).

Molecular weight: 1270.29 g mol-1

Elemental analysis for [N(n-Bu)4]2[13]•4MeCN:

C64H124N6CrS4Si4 % C % H % N % Cr

Calculated 60.62 9.72 6.70 4.12

found 60.58 9.74 6.68 4.10

Chapter 7

216

Synthesis of chromium [N(n-Bu)4][14]

50 mg (39 µmol) of compound [N(n-Bu)4]2[13]•4MeCN in 5 mL of MeCN was

exposed to a stream of air. The color changed immediately to purple and the solvent was

removed under vacuum. The crude material was redissolved in CH2Cl2. After 5 days at -20 or

4 °C purple crystals of [N(n-Bu)4][14] were obtained.

Yield: 38 mg (93%)

Molecular weight: 1049.47 g mol-1

Elemental analysis for [N(n-Bu)4][14]•2CH2Cl2:

C42H80Cl4CrNOS4Si4 % C % H % N

Calculated 48.02 7.68 1.33

found 48.10 7.55 1.21

ESI (in CH2Cl2 solution) : m/z = 636.2 {M}-, 242.2{M}+

Chapter 7

217

Synthesis of 5-azonia-spiro[4,4]nonane bromide

The quaternary ammonium salt was synthesized according to a procedure described by

Schmidbaur et al.7 with modifications. A mixture of pyrrolidone (17.7g, 0.25 mol),

1,4-dibromobutane (34 g, 0.25 mol) and potassium hydroxide water solution (1.3 M, 190 mL)

was heated under reflux for 4 h resulting in a yellow solution with a white precipitate. The

volume was reduced to approximately 80 mL and the precipitate was filtered off. The desired

compound was extracted from the filtrate with CH2Cl2 (3 x 50 mL), dried over sodium

sulphate and filtered. A highly hygroscopic white powder was obtained after evaporation of

the solvent and stored under dried conditions.

All analyses were identical to the results reported in the literature.

Yield: 39 g (77%)

Molecular weight: 206.8 g mol-1

ESI (in CH2Cl2 solution) : m/z = 80.0 {M}-, 126.5 {M}+

1H NMR (CDCl3, 300K): δ = 2.25 (s, 8H), 3.83 (s, 8H).

Chapter 7

218

Synthesis of the rhenium complex [C8H16N][15]•MeCN

To a suspension of ligand 1b (100 mg, 0.35 mmol) in 4 mL of MeCN, 76 mg

(0.68 mmol) of KOtBu was added and stirred for 30 minutes. To the resulting yellow solution

ReCl5 (33 mg, 9.1 x 10-5 mol) and [C8H16N]Br (20 mg, 0.18 mmol) were added and stirred for

1h yielding a brown solution. The crude reaction mixture was filtered through celite and

exposed to air. Brownish-green crystals suitable for X-ray analysis were obtained from

concentrated MeCN solutions.

Yield: 57 mg (55%).

Molecular weight: 1166.27 g mol-1

Elemental analysis for [C8H16N][15]•2MeCN:

C48H82N3ReS6Si6 % C % H % N

Calculated 49.39 7.03 3.60

found 49.40 7.06 3.56

Chapter 7

219

Synthesis of the rhenium complex [C8H16N][16]•acetone

To a suspension of 3,6-dichlorobenzenedithiol H2LCl (100 mg, 0.47 mmol) in 4 mL of

MeCN, 123 mg (1.1 mmol) of KOtBu was added and stirred for 30 min. To the resulting

orange solution were added ReCl5 (58 mg, 1.60 x 10-4 mol) and [C8H16N]Br (40 mg,

0.39 mmol) and stirred for 1 h yielding a brown-greenish solution. The crude reaction mixture

was exposed to air and filtered through celite. The solvent was removed under vacuum and

the material redissolved in acetone. Brownish-green crystals were obtained from concentrated

acetone solution.

Yield: 80 mg (53%).

Molecular weight: 939.78 g mol-1

Elemental analysis for [C8H16N][16]•acetone

C29H28Cl6NOReS6 % C % H % N

Calculated 37.07 2.98 1.49

found 37.18 2.87 1.54

ESI (in CH2Cl2 solution) : m/z = 802.8 {M}-, 126.2 {M}+

Chapter 7

220

7.3 – References

1 Sato, R.; Ohyama, T.; Kawagoe, T.; Baba, M.; Nakajo, S.; Kimura, T.; Ogawa, S. Heterocycles 2001, 55, 145-154. 2 Figuly, G. D.; Loop, C. K.; Martin, J. C. J. Am. Chem. Soc. 1989, 111, 654-8. 3 Figuly, G. D.; Martin, J. C. J. Org. Chem. 1980, 45, 3728-9. 4 Birch, A. J.; Williamson, D. H. Organic Reactions 1976, 24. 5 Schrauzer, G. N.; Mayweg, V. P.; Heinrich, W. Inorg. Chem. 1965, 4, 1615-1617. 6 Bowmaker, G. A.; Boyd, P. D. W.; Campbell, G. K. Inorg. Chem. 1983, 22, 1208- 1213. 7 Schmidbaur, H.; Wohlleben, A.; Schubert, U.; Frank, A.; Huttner, G. Chem. Ber.

1977, 110, 2751-2757.

Chapter 8

221

Chapter 8

Appendix

Chapter 8

222

Chapter 8

223

8.1 – Crystallographic Data:

Compound 1 1ox 2

chem. formula C18H34S2Si2 C24H40S4Si4 C30H54K2N2NiO2S

4Si4

Fw [g.mol-1]

Crystal size [mm]

Crystal system

370.75

0.12 x 0.09 x 0.03

Triclinic

569.16

0.24 x 0.20 x 0.15

Monoclinic

852.26

0.17 x 0.08 x 0.07

Monoclinic

space group P-1 No. 2 P21/c Nr. 14 C2/m No. 2

a, Å 9.0459(5) 10.4821(4) 23.9110(12)

b, Å 10.1596(5) 19.9186(6) 7.0459(3)

c, Å 12.4956(6) 7.4357(3) 17.3759(9)

���deg 76.235(3) 90.00 90.0

�, deg 86.399(3) 96.594(3) 131.472(3)

�, deg 79.359(3) 90.00 90.0

V, Å3 1096.00(10) 1542.2(1) 2193.44(18)

Z 2 2 2

T, K

θ range for data collection

100(2)

2.97 ≤ 2θ ≤ 50.00

150(2)

3.35 ≤ 2θ ≤ 35.00

100(2)

3.11 ≤ 2θ ≤ 34.99

� calcd, g cm-3 1.123 1.226 1.290

refl. collected 14103 6689 26257

Unique refl. 6929 [R(int) = 0.0894]

6209 [R(int) = 0.0310]

5134 [R(int) = 0.0381]

No. of params / restrains 209 / 0 151 / 0 148 / 1

�(Mo K), mm-1 0.349 0.710 0.959

R1 / GOOF [I>2σ(I)] 0.0380 / 1.036 0.0275 / 1.085 0.0454 / 1.056

wR2 c (all data) 0.1005 0.0743 0.1215

resid. density, e.Å-3 0.564 / -0.361 0.451 / -0.250 0.756 / -0.804

Chapter 8

224

Compound 2a 2b

3

chem. formula C40H76NNiS4Si4 C25H42Cl2NiS4Si4 C40H76NCuS4Si4

Fw [g.mol-1]

Crystal size [mm]

Crystal system

870.33

0.34 x 0.22 x 0.07

Triclinic

712.80

0.14 x 0.07 x 0.03

Monoclinic

875.16

0.15 x 0.10 x 0.09

Triclinic

space group P-1 No. 2 P21/c P-1 No.2

a, Å 11.9951(4) 11.9748(2) 11.9348(6)

b, Å 13.9977(4) 13.3369(3) 14.0058(6)

c, Å 17.1576(4) 12.3542(3) 17.0709(6)

���deg 67.070(3) 90.00 66.741(2)

�, deg 79.361(3) 117.435 (4) 78.951(3)

�, deg 79.142(3) 90.00 79.126(2)

V, Å3 2585.91(13) 1751.15(6) 2552.74(19)

Z 2 2 2

T, K

θ range for data collection

200(2)

2.94 ≤ 2θ ≤ 35.00

100(2)

3.11 ≤ 2θ ≤ 34.99

100(2)

3.00 ≤ 2θ ≤ 30.00

� calcd, g cm-3 1.118 1.352 1.139

refl. collected 89342 44211 15383

Unique refl. 22716 [R(int) = 0.0370]]

7695 [R(int) = 0.0314]]

13308 [R(int) = 0.0399]]

No. of params / restrains 493 / 16 184 / 3 470 / 0

�(Mo K), cm-1 0.655 1.097 0.711

R1 / GOOF [I>2σ(I)] 0.0430 / 1.050 0.0270 / 1.035 0.0583 / 1.009

wR2 c (all data) 0.1033 0.0727 0.1313

resid. density, e.Å-3 0.713 / -0.840 0.545 / -0.978 0.553 / -0.577

Chapter 8

225

Compound 4 5 6

chem. formula C40H76NAuS4Si4 C30H44N2PdS2Si2 C30H44N2PtS2Si2

Fw [g.mol-1]

Crystal size [mm]

Crystal system

1008.58

0.09 x 0.08 x 0.07

Triclinic

659.37

0.20 x 0.18 x 0.13

Monoclinic

748.06

0.24 x 0.16 x 0.04

Monoclinic

space group P-1 P21/c No.14 P21/c No.14

a, Å 11.8746(2) 15.2323(3) 15.2157(4)

b, Å 14.1084(2) 11.0835(2) 11.1370(4)

c, Å 17.1088(3) 20.1259(4) 19.9789(6)

α, deg 66.756(3) 90.00 90.00

β, deg 79.040(3) 107.208(4) 106.995(3)

γ, deg 79.431(3) 90.00 90.00

V, Å3 2566.85(7) 3245.70(11) 3237.72(17)

Z 2 4 4

T, K

θ range for data collection

100(2)

3.13 ≤ 2θ ≤ 31.00

100(2)

3.25 ≤ 2θ ≤ 33.23

100(2)

3.25 ≤ 2θ ≤ 35.00

ρ calcd, g cm-3 1.305 1.349 1.535

refl. collected 64370 101646 87925

Unique refl. 16346 [R(int) = 0.0386]

12372 [R(int) = 0.0465]

14241 [R(int) = 0.0567]]

No. of params / restrains 470 / 0 346 / 0 346 / 3

µ(Mo Kα), cm-1 3.147 0.796 4.558

R1 / GOOF [I>2σ(I)] 0.0207 / 1.035 0.0239 / 1.038 0.0275 / 1.105

wR2 c (all data) 0.0482 0.0576 0.0549

resid. density, e.Å-3 1.027 / -1.032 0.628 / -0.634 1.234 / -1.389

Chapter 8

226

Compound 7c 8 9

chem. formula C86H106Cl6 F12N4P2Pt2 S4 C58H59Cl9P2PtS2 C51H60PtS4

Fw [g.mol-1]

Crystal size [mm]

Crystal system

2180.78

0.25 x 0.20 x 0.10

Monoclinic

1432.28

0.09 x 0.04 x 0.02

Triclinic

996.32

0.20 x 0.15 x 0.10

Orthorhombic

space group P21/c n´ P-1 Pna2(1) Nr. 33

a, Å 10.8207(3) 11.7310(6) 11.9920(8)

b, Å 21.3485(6) 13.7027(7) 35.632(2)

c, Å 20.2643(6) 20.5167(12) 11.1812(6)

���deg 90.00 103.082(2) 90.00

�, deg 97.130(3) 105.661(4) 90.00

�, deg 90.00 96.134(4) 90.00

V, Å3 4645.0(2) 3043.1(3) 4777.7(5)

Z 2 2 4

T, K

θ range for data collection

100(2)

0.998 ≤ 2θ ≤ 31.00

100(2)

3.17 ≤ 2θ ≤ 27.50

100(2)

2.10 ≤ 2θ ≤ 29.00

� calcd, g cm-3 1.559 1.563 1.385

refl. collected 72493 12554 41461

Unique refl. 14772 13913 8615

No. of params / restrains 557 / 28 682 479 / 1

�(Mo K), cm-1 0.710 0.710 0.695

R1 / GOOF [I>2σ(I)] 0.0298 / 1.035 0.0512 0.0525 / 1.011

wR2 c (all data) 0.0602 0.1269 0.1198

resid. density, e.Å-3 0.105 / -1.024 3.792 / -2.402 3.09 / -1.83

Chapter 8

227

Compound 10 11 12a + 12b

chem. formula C40H76NCoS4Si4 C64H124N6RhS4Si4 C27.3H49.9I1.7RhS4Si4

Fw [g.mol-1]

Crystal size [mm]

Crystal system

870.55

0.21 x 0.14 x 0.07

Triclinic

1321.20

0.26 x 0.23 x 0.12

Triclinic

937.41

0.12 x 0.12 x 0.11

Monoclinic

space group P-1 No.2 P-1 P21/c No.14

a, Å 11.8504(8) 12.6822(6) 13.0640(4)

b, Å 13.8399(10) 13.0168(7) 20.3985(6)

c, Å 16.7034(10) 113.3739(7) 14.6090(4)

���deg 112.590(3) 69.060(4) 90.00

�, deg 91.552(3) 67.672(4) 99.387(3)

�, deg 98.818(3) 88.844(4) 90.00

V, Å3 2488.3(3) 1890.22(17) 3840.96(19)

Z 2 1 4

T, K

θ range for data collection

100(2)

2.98 ≤ 2θ ≤ 30.00

100(2)

2.92 ≤ 2θ ≤ 32.50

100(2)

3.32 ≤ 2θ ≤ 37.50

� calcd, g cm-3 1.162 1.161 1.621

refl. collected 43520 35348 161285

Unique refl. 14486 [R(int) = 0.0725]]

13615 [R(int) = 0.0445]]

20154 [R(int) = 0.0434]]

No. of params / restrains 470 / 0 370 / 0 362 / 0

�(Mo K), cm-1 0.635 0.439 2.171

R1 / GOOF [I>2σ(I)] 0.0596 / 1.059 0.0450 / 1.063 0.0311 / 1.104

wR2 c (all data) 0.1033 0.0991 0.0634

resid. density, e.Å-3 0.465 / -0.463 0.733 / -0.702 1.876 / -1.407

Chapter 8

228

Compound 13 14 15

chem. formula C64H124N6CrS4Si4 C42H80Cl4NCrOS4Si4 C48H82N3ReS6Si6

Fw [g.mol-1]

Crystal size [mm]

Crystal system

1270.29

0.05 x 0.05 x 0.04

Triclinic

1049.47

0.13 x 0.12 x 0.04

Monoclinic

1248.27

0.52 x 0.12 x 0.10

Orthorhombic

space group P-1 P21/c No.14 Cmca No.64

a, Å 14.4962(6) 26.038(2) 17.6026(5)

b, Å 14.7533(6) 12.0821(6) 22.7323(5)

c, Å 19.0960(6) 17.9404(8) 31.6701(8)

α, deg 80.243(4) 90.00 90.00

β, deg 81.856(4) 98.691(4) 90.00

γ, deg 74.004(4) 90.00 90.00

V, Å3 3849.6(3) 5579.01(6) 12672.7(6)

Z 2 4 8

T, K

θ range for data collection

100(2)

2.92 ≤ 2θ ≤ 26.00

100(2)

2.91 ≤ 2θ ≤ 27.50

100(2)

2.93 ≤ 2θ ≤ 30.00

ρ calcd, g cm-3 1.096 1.249 1.309

refl. collected 67365 37991 80046

Unique refl. 15106 [R(int) = 0.0593]

12788 [R(int) = 0.0536]]

9464 [R(int) = 0.0479]

No. of params / restrains 736 / 0 540 / 3 326 / 6

µ(Mo Kα), cm-1 0.357 0.663 0.796

R1 / GOOF [I>2σ(I)] 0.0659/ 1.147 0.0440 / 1.037 0.0736/ 1.347

wR2 c (all data) 0.1755 0.0886 0.1703

resid. density, e.Å-3 1.193 / -0.493 0.467 / -0.484 2.603 / -5.604

Chapter 8

229

Compound 16

chem. formula C29H28Cl6NOReS6

Fw [g.mol-1]

Crystal size [mm]

Crystal system

997.78

0.12 x 0.10 x 0.09

Monoclinic

space group P21/c No.14

a, Å 12.9802(3)

b, Å 13.9055(3)

c, Å 20.4137(5)

α, deg 90.00

β, deg 107.925(3)

γ, deg 90.00

V, Å3 3237.72(17)

Z 4

T, K

θ range for data collection

100(2)

3.07 ≤ 2θ ≤ 32.50

ρ calcd, g cm-3 1.890

refl. collected 91435

Unique refl. 12678 [R(int) = 0.0396]]

No. of params / restrains 399 / 0

µ(Mo Kα), cm-1 4.309

R1 / GOOF [I>2σ(I)] 0.0294/1.041

wR2 c (all data) 0.0686

resid. density, e.Å-3 1.202 / -2.165

Chapter 8

230

8.2 – Publication from this Thesis:

Pap, J. S., Benedito, F. L., Bothe, E., Bill, E., DeBeer George, S., Weyhermüller, T.,

Wieghardt, K., Inorg. Chem. 2007, 46, 4187–4196.