d - Ruhr-Universität Bochum

Transcription

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 MaxPlanck 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
Chapter 2
Introduction
1
1.1 – General Introduction
3
1.2 – Objectives of this Work
12
1.3 – References
15
Molecular and Electronic Structure of Square Planar Nickel,
18
Copper, and Gold Complexes with a New
ortho-Benezedithiolate Ligand
Chapter 3
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
Dimerization Processes of Square Planar
50
[PtII(tbpy)(dithiolate•)]+ Radicals
52
55
3.3 – Sulfur K-edge X-ray Absorption Spectroscopy (XAS)
60
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
Chapter 5
81
85
89
95
4.5 – Preliminary Reactivity Studies
100
103
4.7 – Conclusions
120
4.8 – References
121
Synthesis and Characterization of Chromium Complexes
124
With ortho-Benzenedithiolate Based Ligands
Chapter 6
126
127
131
136
139
5.6 – Conclusions
147
5.7 – References
148
New tris(Dithiolate) Complexes of Rhenium –
151
A Radical Approach
153
156
162
170
6.5 – X-ray Absorption Spectroscopy (XAS)
184
6.6 – Conclusions
189
6.7 – References
190
II
Chapter 7
Chapter 8
Experimental
193
7.1 – Physical Measurements
195
7.2 – Synthesis
199
7.3 – References
220
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
0
quadrupole splitting
ΔH
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 crosslinking 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
II
Cu O
HO
S
Tyr272
Tyr272
Tyr272
II
Cu O
I
Cu O
S
S
Tyr272
S
Cys228
Cys228
Cys228
Cys228
a
b
c
d
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.
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).
Tyr272
II
Cu O
Tyr272
+ RCH2OH
I
Cu O
kred
+ RCHO + 2H+
S
S
a
Cys228
Cys228
Tyr272
Tyr272
I
Cu O
+ O2 + 2H+
kox
S
II
Cu O
+ H2O2
S
Cys228
Cys228
b
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 redoxactive 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.
A
M
A
z+
(z-1)+
M
B
B
II
I
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 orthoaminothiophenolate(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.
1.76
1.46
1.75
1.40
M
M
N
H
N
H
1.38
S
S
H
av. C–C 1.39 ± 0.01 Å
av. C–C 1.39 ± 0.01 Å
[(LNSAP)]1-
[(LNSIP)]2-
1.43
S
1.72
1.42 1.36
1.42
1.36
1.41
M
N
H
[(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
[(LSS)]2-
ortho-benzenedithiolate(2-)
and
ortho-
dithiobenzosemiquinonate(1-) [(LSSSQ)]1- π-radical ligands (Figure 1.1.7).43-45
1.39
1.41
1.39
1.41
1.42
1.34
O
1.41
1.42
1.36
[(LCat)]2-
1.43
1.34
O
1.43 1.30
1.30
1.42
M
1.34
1.42
1.36
O
1.45
M
O
O
1.22
1.48
1.22
1.34
[(LSQ)]1-
1.43
M
O
[(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.
1.398
1.426
1.770
1.401
1.755
1.390
1.383
1.402
1.382
S
1.407
1.408
M
S
1.429
1.374
S
1.744
1.752
1.391
M
S
[(LSSSQ )]1-
[(LSS)]2-
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,
CuIII, NiII, PdII, PtII, AuIII,
Pd, Pt, Co, Mn, FeIII
FeII, MnII
Co, FeII, CrII, MnII
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
S(3)
2-
Cl(1)
S(2)
Cl(2)
Co(1)
S(4)
S(1)´
S(1)
Cl(1)´
S(2)´
Cl(2)´
Co(1)´
S(3)´
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
S(1)
S(3)
Pt(1)
S(2)
S(4)
S(1)´
S(3)´
Pt(1)´
S(4)´
S(2)´
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 photoinduced electron transfer.56,57 Figure 1.1.9 shows the simplest structure of a dithiolate-diimine
mixed-ligand complex of Pt.
N
S
Pt
N
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).
Si
Si
S
S
S
e+ e-
S
Si
Si
(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
Si
S
N
S
N
Pt
Pt
N
N
S
S
Si
[PtII(tbpy)(LTMS)]
[PtII(tbpy)(LPh)]
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.
S
S
Re
S
S
S
S
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
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.
Cl
Si
S
S
S
S
S
S
Cl
(L)2-
Si
(LCl)2-
(LTMS)2-
Figure 1.2.4 – Representation of the dianionic form of the ligands used in the study of
rhenium tris(benzenedithiolate).
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
+e
-e
[Ni(L)2]1-
+e
-e
[Ni(L)2]2Eq. 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]1motif 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-3,4dithiolate) 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]1shows 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).
Si
SH
a,a b
SH
bdt
H2L
S
i-Pr
S
i-Pr
c,b d
S
i-Pr
S
i-Pr
(1)
e,cf, g
Si
Si
S K
S
Si
SH
ej
di
h,
SH
K
(1b)
Si
Si
Si
(3)
(1a)
S
i-Pr
S
i-Pr
(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.
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(14)
C(20)
C(5)
Si(17)
C(4)
C(6)
C(3)
C(19)
C(18)
C(8)
C(1)
C(2)
Si(13)
C(11)
S(2)
S(1)
C(16)
C(15)
C(10)
C(7)
C(12)
C(9)
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(5)
C(4)
C(6)
C(3)
S(1)
C(1)
C(2)
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
Si
S
Ni
K
N
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-
S(1)
C(6)
C(1)
C(5)
Ni(1)
C(4)
C(2)
C(3)
S(2)
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-
S(1)
C(6)
C(1)
C(5)
Cu(1)
C(4)
C(3)
C(2)
S(2)
Figure 2.2.5 – Perspective view and numbering scheme of the monoanions 3 [Cu(LTMS)2]1with 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-
[Ni(LTMS•)(LTMS) ]1-
[Ni(LTMS•)2]0
[Cu(LTMS)2]-1
[Au(LTMS)2]-1
2
2a
2b
3
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 bisdithiolate 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
197
Au 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
197
Au 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)]1revealed 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]
+e
-e
[MII(LBu•)(LBu)]1-
+e
-e
[MII(LBu)2]2-
Eqn. 2.3.1
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•)]
+e
-e
[MIII(LTMS)2]1-
+e
-e
[MII(LTMS)2]2-
Eqn. 2.3.2
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
5 μA
-1
200 mV.s
-1
100 mV.s
-1
50 mV.s
-1
25 mV.s
0.3
0.0
-0.3
-0.6
-0.9
-1.2
+
E (V) versus Fc /Fc
-1.5
-1.8
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).
4
5 μA
3
5 μA
0.5
0.0
-0.5
-1.0
-1.5
+
E (V) versus Fc /Fc
-2.0
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).
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
E2½, 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)2]2- (2)
2.5
[Ni(LTMS•)(LTMS)]1- (2a)
4 -1
-1
e, 10 M cm
2.0
[Ni(LTMS•)2]0 (2b)
1.5
1.0
0.5
0.0
300
450
600
750
900
1050
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
4.0
500
3.5
ε, M-1cm-1
400
cm-1
2.0
4
300
200
100
0
400
4
ε, 10 M
2.5
-1
3.0
450
500
1.5
550
600
650
700
750
λ, nm
1.0
4a
4
0.5
0.0
400
600
800
1000
1200
1400
1600
λ, nm
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
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).
5
3
ε, 10
4
-1
-1
M cm
4
2
1
0
300
450
600
750
900
1050
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.
1.8
1.6
μeff, μB
1.4
■
1.2
Experimental
Calculated with:
S=½
g = 2.0
θ-Weiss: -0.636 K
1.0
0.8
0.6
0
50
100
150
200
250
300
Temperature, 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.
2.3
2.2
g values
2.1
2
1.9
1.5
dχ" / dB
1.0
0.5
0.0
-0.5
-1.0
300
320
340
360
B, mT
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
61
Ni
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
61
Ni 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
197
Au 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.
g values
2.1
2.05
2
1.95
1.9
1.85
dχ" / dB
2.2 2.15
300
310
320
330
340
350
360
370
B, mT
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
197
Au 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 sixcoordinate 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
197
Au 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-)
II
ligand
radical.
The
Bu• 1+
[Pd (bpy)(L
isoelectronic
Pd
complex
)] (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]1species. 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).
6
5
3
[Ni(LTMS)2]1-
a
b
S
1
S
2
6
5
2
S
4
3
M–S
C–S
C1–C2
C2–C3
C3–C4
C4–C5
C5–C6
C6–C1
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)
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)
Complex
a
S
M
4
[Ni(L)2]2-
1
[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)
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 2B2g ground state concurs with the results from the extended Hückel
B
calculations on the [Ni(L)2]1-.29 The bonding scheme in Figure 2.5.1 identifies four metal dorbitals lower in energy relative to the ligand-based orbitals, similar to that observed by
Solomon et al.62 for [Ni(mnt)2]1-. These four doubly occupied orbitals, namely 1ag (dx2-y2),
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 NiII and AuIII (d8)
ions. The LUMO for both these complexes is the σ-antibonding combination of metal dxy and
the ligand 1b1u orbitals.62 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* b2g MO of the freeligand, 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
TMS
TMS
S
S
X
Ni
S
S
TMS
TMS
Y
Spin up
1.0
Spin down
1b1g
1b1g
0.5
2b2g
0.0
-0.5
1b1u
Energy, eV
-1.0
2b3g
-1.5
-2.0
2b2g
-2.5
1b1u
-3.0
-3.5
-4.0
-4.5
-5.0
1au
1b2g
2b3g
1au
1b2g
2ag
2ag
1b3g
1b3g
1ag
1ag
Figure 2.5.1 – Unrestricted Kohn-Sham MOs and energy scheme of [Ni(LTMS•)(LTMS)]1- 2a
from B3LYP DFT calculations.
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
a
b
[Ni(L)2]2-
[Ni(LTMS )2]1-
a
a
a
a
[Ni(L)2]0
[Cu(L)2]1-
[Au(L)2]1-
[Au(L)2]0
MO
M(ndyz)
2b3g
2b2g
1b1g
71
2b3g
2b2g
1b1g
26 (55)
2b3g
2b2g
1b1g
49
2b3g
2b2g
1b1g
17
2b3g
2b2g
1b1g
11
2b3g
2b2g
1b1g
a
29 b
Ref. . This work.
M(ndxz)
M(ndxy)
S(3pz)
S(3px, y)
17
36
52
50
40
45 (50)
41 (43)
43
48
33
58
33
9
33
4
30
26
55
51
61
8
3
30
27
58
64
9
2 (2)
12
16
53
58
11
2
22 (8)
16 (10)
34
51
26
C(2px, y)
8
10
49 (30)
43 (48)
38 (34)
C(2pz)
5
38
30
55
5
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 1,2benzenedithiolate 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
3.5
[Au(LTMS•)(LTMS)]0
4a
3.0
4 −1
-1
ε, 10 Μ cm
2.5
2.0
1.5
[Ni(LTMS•)(LTMS)]12a
1.0
0.5
0.0
400
600
800
1000
1200
1400
1600
λ, nm
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, ζNiII = 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, 53455360.
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.;
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
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, 44224436.
46
47
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
58
Becke, A. D. J. Chem. Phys. 1986, 84, 4524-9.
59
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 1,2benzenedithiolate(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
52
7
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
Si
Si
S
-e
S
S
+
e
S
N
N
Si
Si
(LTMS)2-
(LTMS•)1-
S
S
-e
+
S
(tbpy)
e
(LPh)2-
S
(LPh•)1-
P
(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
[PtII2(tbpy)2(LPh)(LPh•)]1+
5c
[PdII2(tbpy)2(LTMS•)2]2+
7c
[PtII2(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 blueblack 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
[PdII(tbpy)(LTMS)],
5
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.
Si(1)
S(1)
C(6)
N(1)
C(5)
C(1)
Pd(1)
C(2)
C(4)
C(3)
N(2)
S(2)
Si(2)
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.
55
Chapter 3
Si(1)
S(1)
C(6)
N(1)
C(1)
C(5)
Pt(1)
C(2)
N(2)
C(4)
C(3)
S(2)
Si(2)
Figure 3.2.2 – Perspective view and numbering scheme of compound 6 [PtII(tbpy)(LTMS)]
2+
S(2)
C(2)
C(1)
Pt(1)
S(1)
S(1A)
Pt(1A)
S(2A)
Figure 3.2.3 – Perspective view and numbering scheme of compound 7c [PtII2(tbpy)2(LPh•)2]2+
56
Chapter 3
S(1)
P(1)
C(1)
Pt(1)
C(2)
P(2)
S(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.
S(3)
S(1)
C(1)
C(3)
Pt(1)
C(2)
C(4)
S(4)
S(2)
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)]
[PtII(tbpy)(LTMS)]
[PtII2(tbpy)2(LPh•)2]2+
[Pt(PPh3)2(LPh)]
[PtII(LPh•)2]
5
6
7c
8
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
[Pt(edt)2]2
I
2.748
2.300
[Pt(dmit)2]2•TTF
I
2.935
2.315
7c [PtII2(tbpy)2(LPh•)2]2+
II
3.956
2.862
b
c
a
See text and Figure 3.2.3.
b
edt = (S4C4H4•)1-, ethylene-1,2-dithiolate(1-) radical
monoanion31. c H2(dmit) = 4,5-dimercapto-1,3-dithiol-2-thione, TTF = tetrathiafulvalene.
R
R
R
R
S
S
M
S
S
S
S
M
S
S
R
R
S
R
R
R
R
R
R
Type I
M–M dimer
S
M
S
S
S
S
M
S
S
R
R
R
R
Type II
M–S dimer
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
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 center, whereas in the dimer 7c [PtII2(tbpy)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•)2];47 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 [PtII2(tbpy)2(LPh•)2]2+. It is noteworthy that the geometry of the PtII(tbpy)
segment in 7c and in mononuclear 6 [PtII(tbpy)(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-. The species 7a [PtII(tbpy)(LPh•)]1+ (S = ½) dimerizes
in solution with generation of the dication 7c. The [PtII(tbpy)(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(tbpy)(LTMS)], 7 [PtII(tbpy)(LPh)], 7c [PtII2(tbpy)2(LPh•)2]2+, and 9
[Pt(LPh•)2]
II t
(measured
as
solids)
and
the
monomeric
paramagnetic
species
7a
Ph• 1+
[Pt ( bpy)(L )] (measured in CH2Cl2 solution at 25 °C).
The pre-edge features (marked P1 and P2 in Figure 3.3.1) correspond to transitions to
unoccupied antibonding orbitals with sulfur 3p character.
60
Chapter 3
6 [PtII(tbpy)(LTMS)]
E
2.0
P2
9 [PtII(LPh•)2]
1.5
P1
1.0
Norm. Abs.
0.5
0.0
P2
7 [PtII(tbpy)(LPh)]
2.0
7a [PtII(tbpy)(LPh•)]1+
E
1.5
II
t
Ph•
7c [Pt 2( bpy)2(L
2+
)2]
P1
1.0
0.5
0.0
2466
2468
2470
2472
2474
2476
Energy, eV
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 ligandbased 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
61
Chapter 3
between the dimer 7c [PtII2(tbpy)2(LPh•)2]2+ and the monomer 7a [PtII(tbpy)(LPh•)]1+ in CH2Cl2
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).
The cyclic voltammograms of compounds 5 [PdII(tbpy)(LTMS)], 6 [PtII(tbpy)(LTMS)], 7
[PtII(tbpy)(LPh)], and 8 [Pt(PPh3)2(LPh)] (Figure 3.4.1) were measured in CH2Cl2 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 (E2½) is split into two separate anodic (E2pa ~ +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(tbpy)(LTMS)], which shows two successive oneelectron redox waves: the first at –1.91 V (E1½) is reversible, while the second (E2½;
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(tbpy)(LPh)] exhibits two reversible one-electron transfer waves at –1.87 V
(E1½) and –0.10 V (E2½) and, in addition, a quasi-reversible wave at +0.72 V (E3½).
Table 3.4.1 – Spectroscopic data and redox potentials (E½ vs Fc+/Fc).
Complex
λ, nm (ε, M-1cm-1)a
λ, nm (ε, M-1cm-1)a
E1½
E2½
E3½
5
345 (9700)
486 (6700)
–1.91
c
–
6
342 (6000)
580 (8000)
–1.81
c
–
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
–0.01d
–
[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.
2
3.4.1): E
pa
= 0.12 V;
E2´pa
= 0.19 V;
E2´´pc
= –0.10 V;
peak potential is given.
62
E2´pc
c
Peak potential (Figure
= 0.11 V for 6. d Irreversible; the
Chapter 3
The E1½ waves in the cyclic voltammograms of 5, 6, and 7 each correspond to a oneelectron 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 = orthobenzenedithiolate(2), catecholate, ortho-aminophenolate and ortho- phenylenediamine).12,50
The E2½ waves of 5, 6 and 7 are due to ligand-centered one-electron oxidation of the (LTMS)
and (LPh) ligands, yielding the coordinated radicals (LTMS•) and (LPh•) (Scheme 3.4.1).
-e
[MII(tbpy•)(L)]1-
[MII(tbpy)(L)]0
-e
+e
+e
E 1½
E 2½
[MII(tbpy)(L•)]1+
E 3½
+e
-e
II
M = Pd or Pt
(L)2- = (LTMS)2-, (LPh)2(L•)1- = (LTMS•)1-, (LPh•)1ox
[MII(tbpy)(Lox)]2+
Ph-ox 0
(L ) = (L
)
Scheme 3.4.1 – Representation of redox processes in [Pt(tbpy)(L)] complexes.
This oxidation step (E2½) yields the monocations 6a [PtII(tbpy)(LTMS•)]1+ and 7a
[PtII(tbpy)(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(tbpy)(LTMS•)]1+ and 6 [PtII(tbpy)(LTMS)] dimerize in solution to form the
dinuclear species 6b [PtII2 (tbpy)2(LTMS)(LTMS•)]1+, which undergoes a second one-electron
oxidation yielding 6c [PtII2 (tbpy)2(LTMS•)2]2+.
The quasi-reversible oxidation wave E3½ for 7 [PtII(tbpy)(LPh)] is probably a ligandcentered 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 E2½ and E3½ 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
observed for 7 [PtII(tbpy)(LPh)], although the spectator ligand (tbpy) in 7 is replaced by two
triphenylphosphines in 8.
5µ A
5
[PdII(tbpy)(LTMS)]
6
[PtII(tbpy)(LTMS)]
7
[PtII(tbpy)(LPh)]
8
[PtII(PPh3)2(LPh)]
5µ A
5µ A
5µ A
1.5
1.0
0.5
0.0
-0.5
-1.0 -1.5
E (V) versus Fc+/Fc
-2.0
-2.5
Figure 3.4.1 – Cyclic voltammograms of compounds 5–8 in CH2Cl2 0.10 M [N(n-Bu)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 CH2Cl2 with 0.10 M [N(n-Bu)4](PF6) 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.
6
5
6c
Absorbance
Absorbance
5c
5b
400
500
600
700
800
900
6b
1000 1100
400
500
600
λ, nm
7
700
800
900
1000
λ, nm
8
7 + 7a + 7b + 7c
Absorbance
Absorbance
8a
7b + 7c
400
500
600
700
800
900
1000 1100
400
λ, nm
500
600
700
800
900
1000
λ , nm
Figure 3.4.2 – Electronic spectra of compounds 5–8 at –25 °C in CH2Cl2 containing 0.10 M
[N(n-Bu)4](PF6) 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.
The oxidations of 5 [PdII(tbpy)(LTMS)] and 6 [PtII(tbpy)(LTMS)] in CH2Cl2 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 oneelectron oxidation of 6 [PtII(tbpy)(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 [PtII2(tbpy)2(LTMS)(LTMS•)]1+ (vide infra).
The
spectroelectrochemical
behaviour
upon
one-electron
oxidation
of
7
[PtII(tbpy)(LPh)] is similar that to that of 6, but clear evidence for the formation of 7b
[PtII2(tbpy)2(LPh)(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
[PtII2(tbpy)2(LPh)(LPh•)]1+ (S = ½).
3.5 – X-Band EPR Spectroscopy:
When CH2Cl2 solutions containing 0.10 M [N(n-Bu)4](PF6) as a supporting electrolyte
and millimolar amounts of 5 [PdII(tbpy)(LTMS)], 6 [PtII(tbpy)(LTMS)] or 7 [PtII(tbpy)(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+, 6c [PtII2(tbpy)2(LTMS•)2]2+
or 7c [PtII2 (tbpy)2(LPh•)2]2+, respectively.
The EPR spectrum of compound 5 after 50% oxidation by coulommetry shows an
axial signal with giso = 2.035. Hyperfine coupling with 105Pd nucleus (I = 5/2, 22.2%) was not
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 [PdII2(tbpy)2(LTMS)(LTMS•)]1+.
At 298 K, the EPR spectrum of dimer 7c [PtII2(tbpy)2(LPh•)2]2+ in CH2Cl2 displays the
signal of a spin doublet species with giso = 2.003 and 195Pt (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(tbpy)(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(tbpy)(LPh•)]1+ and dinuclear species 7c [PtII2(tbpy)2(LPh•)2]2+ establish an
equilibrium in solution. On the other hand, attempts to detect the corresponding mononuclear
radical 6a [PtII(tbpy)(LTMS•)]1+ in solution failed. Only the dinuclear, EPR silent species 6c
[PtII2(tbpy)2(LTMS•)2]2+ is present in solution between 25 °C and –90 °C.
Interestingly, electrochemically one-electron oxidized CH2Cl2 solutions of 8
[Pt(PPh3)2(LPh)] at 10 K display a rhombic signal with weak
195
Pt 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.
In order to detect directly the intermediates in the oxidations of 6 [PtII(tbpy)(LTMS)]
and 7 [PtII(tbpy)(LPh)], we have also carried out stepwise coulometric oxidations of their
CH2Cl2 (0.10 M [N(n-Bu)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
NaClO4, 10 mM HCl) measured under the same conditions.
67
2
2.1
2.15
dχ´´ / dB
1.95
Chapter 3
g values
2.05
310
320
330
340
350
B, mT
2.05
1.95
1.9
dχ´´ / dB
2.1
g values
2
320
330
340
350
B, mT
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.
68
Chapter 3
Figure 3.5.2 shows the EPR spectra of a 50% oxidized solution of 6 [PtII(tbpy)(LTMS)]
(top) and 7 [PtII(tbpy)(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
[PtII2(tbpy)2(LTMS•)2]2+ and 7c [PtII2(tbpy)2(LPh•)2]2+ exist at 30 K. This behaviour provides
evidence that the intermediates 6c and 7c are paramagnetic dinuclear species: 6b
[PtII2(tbpy)2(LTMS)(LTMS•)]1+ and 7b [PtII2(tbpy)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
195
Pt 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
195
Pt 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(tbpy)(LTMS)] or 7 [PtII(tbpy)(LPh)]); (2) the oxidized monomers (6a [PtII(tbpy)(LTMS•)]1+
7a [PtII(tbpy)(LPh•)]1+); (3) the singly oxidized dimers (6b [PtII2(tbpy)2(LTMS)(LTMS•)]1+ or 7b
[PtII2(tbpy)2(LPh)(LPh•)]1+); and (4) the doubly oxidized dimers (6c [PtII2(tbpy)2(LTMS•)2]2+ and
7c [PtII2(tbpy)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 [PtII2(tbpy)2(LTMS)2] or 7d [PtII2(tbpy)2(LPh)2],
and therefore, K3 must be very small (vide infra).
69
Chapter 3
g values
2.4
2.3
2.2
2.1
2
1.9
1.8
1.7
dχ´´ / dB
6b (%)
Oxidation conv. (%)
280
320
360
400
B, mT
2.4
g values
2
2.2
1.8
dχ´´ / dB
7b (%)
Oxidation conv. (%)
280
320
360
400
B, mT
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
195
Pt nuclei according to the statistical weights expected for symmetric Pt-dimers with total
spin S = ½ and
195
Pt (33.8 % natural abundance): 11.4 % double-labeled, 44.8% single
labelled, and 43.8 % without
195
Pt isotope. Insets represent the conversion of 6 or 7 to the
dimer 6b and 7b in a stepwise coulometric experiment.
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
2.035
2.065
2.065
1.975
Axx = Ayy = 50; Azz = 20
6bb
2.074
2.183
2.167
1.856
Axx = 200; Ayy = 153; Azz = 140
7ac
2.003
7bc
2.061
2.176
2.086
1.914
Axx = 103; Ayy = 112; Azz = 93
8aa
2.010
2.020
2.006
1.989
Axx = 70; Ayy = 20; Azz = 90
a
Aiso = 38e
In CH2Cl2 solution at 30 K. b In CH2Cl2 solution at 10 K. c In CH2Cl2 solution at 298 K.
d
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
|A0| = ⅓(Axx + Ayy + Azz).
K
K33
6, 7
6, 7
6d, 7d
6b, 7b
6a, 7a
6c, 7c
TMS
Ph
•
TMS•
Ph•
•(L
(L)(L)= =(L(L
) )or
)=
TMS
or(L
(LPh);); (L
) =(L(LTMS•) )or
or(L
(LPh•) )
Scheme 3.5.1 – Redox process cycle involved in Pt complexes 6 and 7.
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(tbpy)(LTMS)], shown in Figure 3.6.1. The
shapes of the voltammograms depend mainly on (1) the three equilibrium constants K1, K2, K3
and the three redox potentials E2½, E2´½, and E2´´½ and (2) the rate constants for dimerisation
(kf, M-1s-1) and dissociation (kb, s-1) 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 cm2s-1 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(tbpy)(LTMS)] to 6a [PtII(tbpy)(LTMS•)]1+ at E2pa (E2½), followed by rapid
dimerisation 6a + 6 → 6b [PtII2(tbpy)2(LTMS)(LTMS•)]1+ and further one-electron oxidation of
6b to the dinuclear dication 6c [PtII2(tbpy)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 K2. Reversing this sequence leads to the reduction 6c → 6b at E2´pc
(E2´½) which proceeds at the normal half wave potential E2½ (6c/6b). Since K2 is large, the
final reduction to 6 takes place at the more negative potential E2pc via the thermodynamically
less favourable pathway E2´´½ (6b/6d) with subsequent dissociation to 6 [PtII(tbpy)(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 cm2s-1 for all species)a
a
Kn, M-1
kf, M-1s-1
kb, s-1
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
K3
~0.5
~103
~103
Redox potentials vs Fc+/Fc: E2½, 0.167 V; E2´½, 0.160 V; E2´´½, –0.145 V. See Scheme 3.5.1.
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
[PtII2(tbpy)2(LTMS•)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
sim.
2'
E pc
2''
E pc
exp.
2
E pa
2'
E pa
200 mV.s-1
sim.
2'
E pc
2''
E pc
exp.
E2pa
2'
E pa
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
E (V) versus Fc+/Fc
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
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 [PtII2(tbpy)2(LPh•)2]2+, whereas at 20 °C the mononuclear paramagnetic 7a
[PtII(tbpy)(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 K1. Values of K1 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.
A value of 660 M-1 was found for K1 of 7c [PtII2(tbpy)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 = –50 ± 1 kJ mol-1 and ΔS0 = –59 ± 2 kJ K-1 mol-1. Furthermore,
the increasing value of K1 with decreasing temperature (2.3 x 106 M-1 is calculated from
–90 °C in CH2Cl2) indicates that the concentration of paramagnetic 7a [PtII(tbpy)(LPh•)]1+ is
negligible at –90 °C and, therefore, becomes undetectable by EPR spectroscopy in frozen
solution.
1.0
7a
0.8
Absorbance
7c
238 K
0.6
0.4
0.2
306 K
0.0
400
500
600
700
800
900
1000
λ, nm
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(tbpy)(LTMS)], 6 [PtII(tbpy)(LTMS)], and 8 [PtII(PPh3)2(LPh)] were reported.
Electrochemical and chemical one-electron oxidation of compounds 5, 6 and 7 in CH2Cl2
solution afford the monomeric monocations 5a [PdII(tbpy)(LTMS•)]1+, 6a [PtII(tbpy)(LTMS•)]1+
and 7a [PtII(tbpy)(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
[PtII2(tbpy)2(LTMS)(LTMS•)]1+
and
7b
[PtII2(tbpy)2(LPh)(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
195
Pt 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
105
Pd nucleus. Complete one-electron oxidation of 6 and 7 yielded the
diamagnetic dicationic dimers 6c [PtII2(tbpy)2(LTMS•)2]2+ and 7c [PtII2(tbpy)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
[PtII2(tbpy)2(LPh•)2](PF6)2•3CH2Cl2. The structure revealed a centrosymmetric, lateral dimer
whose bridging part is a PtII2(μ2-S)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•)2]0
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)2(LPh)] 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
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, 199207.
23
Zhang, Y.; Ley, K. D.; Schanze, K. S. Inorg. Chem. 1996, 35, 7102-7110.
24
5641-5654.
25
13, 2783-2797.
26
Chem. 2003, 42, 4082-4087.
27
28
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, 315316.
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, 515.
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
Chapter 4
Electronic Structure of
Square Planar Cobalt and Rhodium Complexes
Containing a bis(ortho-Benzenedithiolate) Ligand
79
Chapter 4
80
Chapter 4
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 electronwithdrawing 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
Cl
NC
S
F 3C
S
NC
S
F 3C
S
2-
(mnt)
2-
(tdf)
Cl
S
S
S
S
S
Cl
S
S
S
S
S
Cl
(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.
81
Chapter 4
((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
82
Chapter 4
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 closedshell 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 orthobenzenedithiolates(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]1complex (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 fourcoordinate 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 shortlived 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
83
Chapter 4
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
NC
S
S
CN
2-
Rh
NC
S
S
CN
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.
84
Chapter 4
Results and Discussion
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-1,2dithiolate), 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.
85
Chapter 4
1-
S(1)
C(6)
C(1)
C(5)
Co(1)
C(4)
C(2)
C(3)
S(2)
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
86
Chapter 4
(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).
2-
S(1)
C(6)
C(5)
C(1)
Rh(1)
C(2)
C(4)
C(3)
S(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.
87
Chapter 4
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.
N
Si
N
S
Rh
Si
N
N
Si
S
S
S
N
Si
N
Figure 4.2.3 – Packing motif in crystals of complex [N(n-Bu)4]2[11]•4MeCN.
88
Chapter 4
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)
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
89
Chapter 4
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.
5 µA
0.5
0.0
-0.5
-1.0
-1.5
-2.0
E (V) versus Fc+/Fc
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).
10a
[CoIII(LTMS)(LTMS•)]0
S=½
10
+e
-e
[CoIII(LTMS)2]1S=1
10b
+e
-e
[CoII(LTMS)2]2S=½
10c
III
TMS
[Co (L )(LTMS•)]2
S=0
Eqn. 4.3.1
90
Chapter 4
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.
1.75
4
ε, 10 M
-1 cm-1
1.50
1.25
1.00
0.75
0.50
0.25
0.00
300
400
500
600
700
800
900
1000 1100
λ, nm
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.
91
Chapter 4
-1 cm-1
1.0
ε, 10 M
1.5
4
2.0
0.5
0.0
300
400
500
600
700
800
900
1000
λ, nm
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).
92
Chapter 4
5 µA
-0.2
-0.4
-0.6
-0.8
-1.0
-1.2
-1.4
-1.6
E (V) versus Fc+/Fc
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).
11a
III
TMS
[Rh (L )2]
S=0
1-
+e
-e
11
II
TMS
[Rh (L )2]
S=½
2-
+e
-e
11b
I
[Rh (LTMS)2]3S=0
Eqn. 4.3.2
Table 4.3.1 – Redox potentials (E2½) and peak potentials (E1p) 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
E2½, 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.
93
Chapter 4
2.5
-1 cm-1
1.0
4
1.5
ε 10 M
2.0
0.5
0.0
300
400
500
600
700
800
900
1000 1100
λ, nm
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)
94
Chapter 4
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 = gµ BBS+ D[Sz2-S(S+1)/3]+E/D[Sx2-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.
3.0
µeff/µB
Experimental
Calculated with:
■
2.5
S=1
g = 2.094
D = 35 cm-1
2.0
1.5
1.0
0
50
100
150
200
250
300
Temperature, K
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
95
Chapter 4
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.
96
Chapter 4
2.25
2.2
2.15
g values
2.1
2.05
2
1.95
1.9
1.2
dχ ´´ / dB
0.8
0.4
0.0
-0.4
-0.8
300
310
320
330
340
350
360
B (mT)
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).
97
Chapter 4
1.9
1.8
1.7
µeff/µ B
Experimental
Calculated with:
■
1.6
1.5
S=½
g = 2.046
θ-Weiss = -0.83 K
TIP = 646.3 x 10-6 emu
1.4
1.3
1.2
1.1
1.0
0
50
100
150
200
250
300
Temperature, K
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
103
Rh (I = ½, 100% natural
abundance) was observed. Very few examples of EPR spectra are found in the literature for
monomeric RhII complexes.
98
Chapter 4
2.6 2.5
2.4
2.3
g values
2.2
2.1
2
1.9
1.0
dχ ´´ / dB
0.5
0.0
-0.5
-1.0
260
280
300
320
340
360
B (mT)
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
99
Chapter 4
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
[Rh(mnt)2]2[Rh(mnt)2]
T(K)
Ref.
Sample prep.
g1
g2
g3
A 1a
A 2a
A 3a
77
26
polycrystalline
2.35
2.015
1.950
-
-
-
rt
27
single crystal
2.447
2.019
1.936
< 0.4
0.75
< 0.4
rt
38
powder
2.74
2.60
1.94
-
-
-
25
-
frozen solution
2.499
2.005
1.973
-
-
-
20
37
frozen solution
2.794
2.016
1.947
-
-
-
100
36
polycrystalline
2.45
2.45
1.96
66
66
62
77
39
frozen solution
2.96
2.58
1.85
4.7
5.6
5.6
b
2-
[Rh(C6Cl5)4]2-
c
11
II t
[Rh ( Bu2-boxate)2]
d
[Rh(TMPP)2(CNtBu)2]2+
e
Rh(2,4,6- Pr3C6H2)2(tht)2
f
i
a
A values for hyperfine coupling constants are given in 10-4 cm-1.
maleonitriledithiolate.
c
mnt2-:
The EPR spectra of a series of compounds containing
pentachlorophenyl derivatives have also been reported.40
butyl(bisoxazolinate))
b
e
(TMPP):
d
(tBu2-boxate)1- = di-tert-
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.
III
TMS
3[RhI(LTMS)2]3- + CH3I → [Rh CH3I(L )2]
100
Eqn. 4.5.1
Chapter 4
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%
III
of
12a
TMS(CH3)2
{Rh CH3I[L
cis-{RhIIII2[LTMS(CH3)2][LTMS(CH3)]}
and
30%
of
12b
cis-
TMS(CH3)
][L
]} 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.
101
Chapter 4
Si
C(2)
Si
I(1)
H
H
HC(1)
H
S(3) C(7)
S(1)
C(5)
Rh(1)
I(2)
C(8)
C(6)
S(4)
S(2)
C(3)
Si
C(4)
Si
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)
102
Chapter 4
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.
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.
TMS
TMS
6
5
1
S
2
S
S
M
4
3
S
TMS
TMS
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)
103
Chapter 4
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 dx2-y2 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 (1LNISQ)•- = ortho-diiminobenzosemiquinonate(1-)).
104
Chapter 4
TMS
TMS
S
S
X
Co
S
S
TMS
TMS
Y
Spin up
Spin down
-1.0
1b1g (dxy + L)
-1.5
-2.0
1b1g
2b3g (dyz)
-2.5
-3.0
2b2g (dxz + L)
Energy, eV
-3.5
1b2g
-4.0
-4.5
-5.0
-5.5
-6.0
1b1u
1b2g
1b1u
1au
1b3g
1au
1b3g
-6.5
-7.0
-7.5
2ag
2b3g
2b2g
1ag
2ag (dx2-y2 + dz2)
1ag (dx2-y2 + dz2)
Figure 4.6.1 – Unrestricted Kohn-Shan MOs and energy scheme from B3LYP DFT
calculations of the monoanion 10 [CoIII(LTMS)2]1-.
105
Chapter 4
TMS
TMS
S
S
X
Co
S
S
TMS
TMS
Y
Spin up
Spin down
1.0
0.5
2b1u
2b2g
0.0
1b1g
1b1g (dxy + L)
3b2g (dxz)
-0.5
2b1u
-1.0
Energy, eV
-1.5
2b2g
-2.0
1b3g (dyz + L)
-2.5
2ag (dz2)
-3.0
-3.5
-4.0
-4.5
-5.0
1b2g
1b3g
1b2g
2ag
1b1u
1au
1b1u
1ag
1au
-5.5
-6.0
3b2g
1ag (dx2-y2)
calculations of the dianion 10b [CoII(LTMS)2]2-.
106
Chapter 4
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 closedshell ortho-benzenedithiolate(2-) ligands.13
107
Chapter 4
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
10 [Co(LTMS)2]1-
MO
2b3g
2b2g
(dyz)
82
*[Co(L)2]1-
2b3g
2b2g
76
10b [Co(LTMS)2]2-
1b3g
3b2g
2b2g
72
*[Fe(L)2]2-
2b3g
2b2g
90
11 [Rh(LTMS)2]2-
1au
1b1u
2b2g
11b [Rh(LTMS)2]3-
1b3g
1b2g
1au
(dxz)
71
S(pz)
7
22
65
10
28
10
08
71
19
18
10
3
9
12
69
82
3
8
2
6
26
47
26
39
42
10
15
30
9
13
11
87
28
61
71
58
(dxy)
S(px, y)
C(pz)
10
7
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
13
7.82
0.51
1.59
see text
2- 13
[Fe(L)2]
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
[Co(L)2]1-
108
Chapter 4
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-y2) 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]3are 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.
109
Chapter 4
TMS
TMS
S
S
X
Rh
S
S
TMS
TMS
Y
Spin up
Spin down
1.0
1b1g
1b1g (dxy + L)
0.5
3b2g
2b1u
3b2g
0.0
2b1u
-0.5
Energy, eV
-1.0
2b2g (dxz + L)
-1.5
1b3g (dyz + L)
-2.0
-2.5
-3.0
2ag (dz2)
2b2g
1b1u
-3.5
1b3g
-4.0
2ag
1b1u
1au
1au
1ag
1b2g
1b2g
1ag (dx2-y2)
-4.5
-5.0
-5.5
calculations of the dianion 11 [RhII(LTMS)2]2-.
110
Chapter 4
TMS
TMS
S
S
X
Rh
S
S
TMS
TMS
Y
2.5
1b1g
1b1g (dxy)
2.0
2b3g
1au
2b3g
2b2g
1au
1.5
1.0
2b1u
23b2g
Energy, eV
0.5
2b1u
0.0
-0.5
-1.0
-1.5
-2.0
2b2g (dxz + L)
1b2g
2
1b3g (dyz)
1b3g
2ag
2ag (dz2)
1ag (dx2-y2)
-2.5
-3.0
-3.5
1ag
1b1u
1b1u
1b2g
1b2g
Figure 4.6.4 – Restricted Kohn-Shan MOs and energy scheme from B3LYP DFT calculations
of the trianion 11b [RhI(LTMS)2]3-.
111
Chapter 4
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-.
112
Chapter 4
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
I
14890 (670)
19050 (525)
UV-Vis
14300 (700)
15900 (630)
II
16804 (595)
19342 (517)
18983 (527)
B1u (1b1u → 2b2g(dxz))
UV-Vis
3
B3u (1au → 2b2g(dxz))
MCD
19560 (511)
16007 (625)
IV
3
MCD
15000 (667)
III
assignment
UV-Vis
3
B3u (1b1u → 2b3g(dyz))
MCD
20200 (495)
18726 (534)
UV-Vis
3
B2u (1au → 2b3g(dyz))
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.
113
Chapter 4
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.
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
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
11b
calcd.
assignment
The oscillator strength can be calculated from the experimental absorption spectra using
equation 4.6.1.56
f 0→ I =
4.32 ⋅10−9
n
∫
Band
ε ( I ) (ν% )dν% ,
eqn. 4.6.1
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, ∫
represents the integral of the band area and v~ is the transition
Band
-1
energy in cm .
114
Chapter 4
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.
2.5
I
1.5
4
-1
ε, 10 M cm
-1
2.0
1.0
II
0.5
III
0.0
11000
12000
13000
14000
15000
16000
17000
Energy, 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)2ligands.13,50
115
Chapter 4
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
116
Chapter 4
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
initial d orbital
dxz
z
Final d orbital
-idxy
idx2-y2 – i 3 dz2
-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.
117
Chapter 4
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).
dxy
z2
6
x
2
xz
dxz
y
6
2
2
2
dyz
x2-y2
dz2
yz
8
y
z
x
2
xy
dx2- y2
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.
gi = ge
1+
niλ
∆i
Eqn. 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
118
Chapter 4
small, which agrees with the experimentally observed at 1.973. Thus, equation 4.6.3 can
written as:
gx = ge _ 2λ
∆xy
Eqn. 4.6.3
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-y2 and dz2 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:
gy = ge +
6λ
2λ
+
2
∆x2-y2
∆z
Eqn. 4.6.4
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.
gz = ge + 2λ
∆yz
Eqn. 4.6.4
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.
119
Chapter 4
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.
b1u
b1u
I
III
II
I
b2g
b2g
I
b1u
III / I
II
au
b1u
11000
13000
14000
15000
16000
17000
-1
Energy, cm
au
11
12000
11b
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 d8 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.
120
Chapter 4
4.8 – References:
1
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
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.
121
Chapter 4
21
Osborn, J. A.; Jardine, F. H.; Young, J. F.; Wilkinson, G. J. Chem. Soc. A 1966, 17111732.
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.
122
Chapter 4
41
Ni, Y. N.; Fitzgerald, J. P.; Carroll, P.; Wayland, B. B. Inorg. Chem. 1994, 33, 20292035.
42
43
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
51
5641-5654.
52
13, 2783-2797.
53
Ray, K.; Petrenko, T.; Wieghardt, K.; Neese, F. Dalton Transactions 2007, 15521566.
54
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.
123
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 5,10,15tris(pentafluorophenyl)chromium(III)-corrole.34
A
CrV
oxo
complex
containing
35
toluene-3,4-dithiolate ligands was reported by Gray et al. 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:
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 (superhydride) 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)
C(6)
S(3)
C(7)
C(1)
C(5)
Cr(1)
C(4)
C(8)
C(2)
C(3)
S(2)
S(4)
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
[N(n-Bu)4]2[13]•4MeCN.
127
the
packing
motif
of
13
Chapter 5
N
N
Si
Cr
Si
N
N
Si
S
N
S
Si
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
1O(1)
C(1)
C(2)
C(5)
C(4)
C(3)
C(11)
C(7)
C(8)
Cr(1)
C(6)
a)
C(12)
S(3)
S(1)
C(9)
C(10)
S(4)
S(2)
1-
Cr
b)
B
C
A
D
1-
c)
Φ1 = 144.6°
Φ2 = 175.8°
0.719 Å
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)
Φ2
175.8°
Φ1
144.6°
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 closed130
Chapter 5
shell (LTMS)2- ligands. These observations support the CrII and CrV oxidation state assignments
for the anions 13 and 14, respectively.
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-
– e-
[CrIII(LTMS)2(X)2]1- (X = MeCN).
131
Eqn. 5.3.1
Chapter 5
5 μA
-0.4
-0.6
-0.8
-1.0
-1.2
-1.4
-1.6
+
E (V) versus Fc /Fc
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
2.5
1.00
2.0
0.75
0.50
1.5
2
ε, 10 M
-1 cm-1
3.0
0.25
1.0
0.00
400
450
500
550
600
650
0.5
0.0
400
500
600
700
800
900
1000
1100
λ, nm
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 metalcentered redox chemistry.41
[CrIVO(LTMS)2]214a
-e
+e
[CrVO(LTMS)2]114
133
-e
+e
[CrVIO(LTMS)2]0
14b
Eqn. 5.3.2
Chapter 5
5 μA
0.5
0.0
-0.5
-1.0
-1.5
+
E (V) versus Fc /Fc
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
E2½, V vs Fc+/Fc
13
-1.050
-1.310
14
+0.119
-1.053
+0.150
-0.96
[CrVO(LMe)]1-
36
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 lowenergy 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.
2.00
[CrVO(LTMS)2]1- (14)
[CrIVO(LTMS)2]2- (14a)
1.50
1.25
4
ε, 10 M
-1
-1
cm
1.75
1.00
0.75
0.50
0.25
0.00
300
400
500
600
700
800
900
λ, nm
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.
a
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)
371 (1.21)
14a
Measurement performed at +25 °C.
135
Chapter 5
The magnetic susceptibility of [N(n-Bu)4]2[13]•4MeCN in the temperature range of 2290 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.
5.0
4.9
4.8
μeff/μB
4.7
4.6
■
4.5
4.4
Experimental
Calculated with:
S=2
g = 1.960
θ-Weiss = 1.00 K
D = 2.2 cm-1
TIP = 1056 x 10-6 emu
4.3
4.2
4.1
4.0
0
50
100
150
200
250
300
Temperature, K
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
1.8
1.7
■
μeff/μB
1.6
Experimental
Calculated with:
S=½
g = 1.998
θ-Weiss = -2.11 K
TIP = 351 x 10-6 emu
1.5
1.4
1.3
0
50
100
150
200
250
300
Temperature, K
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
g values
2
1.95
1.9
1.85
1.8
´´
dχ / dB
2.2 2.15 2.1 2.05
300
310
320
330
340
350
360
370
B, mT
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]1in 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 squarebased 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]2Cr–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)
Φ2
175.8°
(175.9°)
Φ1
144.6°
(142.0°)
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.
S
2-
S
X
Cr
S
S
Y
1b1g (dxy)
2b3g (dyz)
2b2g (dxz)
2ag (dz2)
1ag (dx2-y2)
1b2g
1b1u
1b3g
1au
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(dx2-y2), 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.
[CrII(L)2]2-
electrons-nd
4.91
electrons (n+1)s
0.49
142
Spin-nd
3.89
Metal oxdn. state
CrII
Chapter 5
TMS
O
TMS
1-
Cr
S
S
S
Z
TMS
S
TMS
Y
X
Spin up
1.5
1.0
0.5
Spin down
7a´ (dz2)
7a´
6a´´
6a´´ (dxy)
6a´
5a´´
6a´ (dxz)
0.0
5a´´ (dyz)
Energy, eV
-0.5
-1.0
5a´ (dx2-y2)
-1.5
4a´
-2.0
5a´
4a´´
-2.5
-3.0
-3.5
3a´´
2a´´
-4.0
3a´´
-4.5
2a´´
1a´´
2a´
1a´
-5.0
3a´
4a´
4a´´
3a´
1a´´
2a´
1a´
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-y2 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-y2 < 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 moderateto-weak equatorial ligand field. The 3dz2, 3dxz, and 3dyz orbitals are strongly destabilized by σand 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
3dx2-y2
3dxz
S
3dyz
3dxy
3dz2
C
3px,y
3pz
2px,y
2pz
2px,y
32
3
4a´
5
5
47
3
5a´
80
2
5
3
64
5a´´
6a´
66
50
6a´´
8
19
10
19
27
2pz
19
9
23
7a´
O
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-y2 character in contrast with the Mo analogue
with 70% 3dx2-y2 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-
80% dx2-y2
5a´
70% dx2-y2
5a´
Cr 3dx2-y2
Mo 4dx2-y2
S 3p
S 3p
S 3p
S 3p
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-y2 orbitals in both C2v and in Cs symmetries.
145
Chapter 5
O
Cs
a)
S
S
S
S
O
b)
C2v
S
S
S
S
Figure 5.5.4 – (a) Scheme of the 4a´ molecular orbital. Dashed lines represent the interaction
between the S pz and the Cr 3dx2-y2 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-y2, 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-y2 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-y2 orbital
becomes doubly occupied and in this case, the Cr 3dx2-y2 ↔ 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-y2 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 πsymmetry to the singly occupied Cr 3dx2-y2 orbital. The lowest band corresponds to the
4a´(Spπ) → 5a´ (3dx2-y2) transition.36
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, 177184.
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, 18251826.
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.;
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.;
41
Boyde, S.; Ellis, S. R.; Garner, C. D.; Clegg, W. J. Chem. Soc., Chem. Commun. 1986,
1541.
42
43
Chem. 2003, 42, 4082-4087.
44
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
5641-5654.
48
Enemark, J. H.; Cooney, J. A.; Wang, J.-J.; Holm, R. H. Chem. Rev. 2004, 104, 1175.
150
Chapter 6
Chapter 6
New tris(Dithiolate) Complexes of Rhenium –
A Radical Approach
151
Chapter 6
152
Chapter 6
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
153
Chapter 6
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
154
Chapter 6
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-y2 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-1,2dithiolate, 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- = 1,2ethanedithiolate) but forms a distorted octahedron.13-15
C3
C3
Twist angle, θ
Octahedron
Trigonal Prismatic
θ = 60°
a)
θ = 0°
c)
b)
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).
155
Chapter 6
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-,
[W(mnt)3]2-.7,18-20 Some authors support the hypothesis that the occurrence of
and
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 sp3
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 = orthobenzedithiolate). 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.
156
Chapter 6
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.
1-
1S(2)
S(3)
S(3)
S(5)
S(1)
S(4)
S(1)
S(6)
S(2)
S(4)
S(5)
S(6)
15
16
5
R
4
6
1
2
S1
R
17
16
S
18
6
S
2
1
Re
S
R
3
7
8
S5 S4
9
12
13
15
R
3
14
11
10
R
R
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).
157
Chapter 6
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)
158
Chapter 6
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.
1-
1-
1-
120
1-
16
15
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°)
159
Chapter 6
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.
Φ
S
S
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.
S
S
S
S
S
S
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.
160
Chapter 6
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°.
161
Chapter 6
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
TMS
[C8H16N][Re(L
)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.
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.
162
Chapter 6
[ReIV(LTMS)3]2-
-e
+e
-e
+e
[ReV(LTMS)3]1-
[ReV(LTMS•)(LTMS)2]0
15b
15a
15
S=½
S=0
S=½
[ReV(LTMS•)2(LTMS)]1+
15c
S=½
Eqn. 6.3.1
1-/2-
0/11+/0
5 µA
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
+
E (V) versus Fc /Fc
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]1in 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
163
Chapter 6
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-) πradical 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.
5
15 [ReV(LTMS)3]115a [ReIV(LTMS)3]215b [ReV(LTMS•)(LTMS)2]0
15c [ReV(LTMS•)2(LTMS)]1+
3
4
ε , 10 M
-1
-1
cm
4
2
1
0
300
400
500
600
700
800
900
1000
1100
λ, nm
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]1measured 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.
164
Chapter 6
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.
0/11+/0
5 µA
1.0
0.5
0.0
-0.5
-1.0
-1.5
E (V) versus Fc+/Fc
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,
165
Chapter 6
should disappear if the coulometric process was performed completely, as observed for the
redox process of complex 15 (vide supra).
6
16 [ReV(LCl)3]116a [ReIV(LCl)3]2-
5
4
100% conversion of 16 to 16a
3
16% convertion of 16 to 16b
4
ε , 10 M
-1
-1
cm
16b [ReV (LCl•)(LCl)2]0
2
Starting material 16
1
0
300
450
600
750
900
1050
λ, nm
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).
[ReIV(LCl)3]2-
-e
+e
[ReV(LCl)3]1-
-e
[ReV(LCl•)(LCl)2]1-
16a
16
16b
S=½
S=0
S=½
Eqn. 6.3.2
166
Chapter 6
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.
5 µA
1+/0
1.0
0.5
1-/2-
0/1-
0.0
-0.5
-1.0
-1.5
E (V) versus Fc+/Fc
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.
167
Chapter 6
5
17 [ReV(L)3]117a [ReIV(L)3]2-
-1 cm-1
2
17b [ReV (L•)(L)2]0
4
3
ε , 10 M
4
1
0
300
400
500
600
700
800
900
1000 1100
λ, nm
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.
[ReIV(L)3]2-
-e
+e
[ReV(L)3]1-
17a
17
S=½
S=0
-e
+e
[ReV(L•)(L)2]0
17b
S=½
[ReV(L•)2(L)]1+
17c
S=½
168
Eqn. 6.3.3
Chapter 6
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
E2½ 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)
169
Chapter 6
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.
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 electrontransfer 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-1,2benzenedithiolate(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 πradical (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.
170
Chapter 6
5
R
4
6
2
S1
R
S
18
R
3
1
6
S
2
1
Re
S
R
3
7
S5 S4
12
13
11
R
R
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
[Re (LTMS)3]1V
Re(1)–S(1)
Re(1)–S(2)
Re(1)–S(11)
Re(1)–S(12)
Re(1)–S(21)
Re(1)–S(22)
S(1)–C(1)
S(2)–C(2)
S(3)–C(7)
S(4)–C(12)
S(5)–C(13)
S(6)–C(18)
C(1)–C(2)
C(2)–C(3)
C(3)–C(4)
C(4)–C(5)
C(5)–C(6)
C(6)–C(1)
2.374
(2.341(2))
2.373
(2.358(2))
2.372
(2.330(2))
2.372
(2.345(2))
2.372
(2.330(2))
2.372
(2.345(2))
1.763
(1.762(9))
1.764
(1.739(1))
1.762
(1.742(7))
1.763
(1.743(7))
1.762
(1.742(7))
1.763
(1.743(7))
1.409
(1.398(1))
1.423
(1.434(1))
1.403
(1.387(1))
1.423
(1.430(1))
1.403
(1.384(1))
1.422
(1.429(1))
16
[Re (LCl)3]1-
17
[Re (L)3]1-
17a
[Re (L)3]2-
17b
[Re (L•)(L)2]0
2.364
(2.3513(6))
2.364
(2.3373(6))
2.365
(2.3279(6))
2.364
(2.3390(6))
2.365
(2.3371(6))
2.365
(2.3476(6))
1.757
(1.745(3))
1.758
(1.737(3))
1.757
(1.738(2))
1.758
(1.740(2))
1.757
(1.740(2))
1.758
(1.735(2))
1.417
(1.400(4))
1.422
(1.407(4))
1.400
(1.382(4))
1.407
(1.387(5))
1.400
(1.381(4))
1.417
(1.403(4))
2.378
2.406
2.371
2.376
2.383
2.370
2.379
2.376
2.372
2.377
2.473
2.371
2.377
2.472
2.371
2.378
2.402
2.370
1.757
1.760
1.741
1.758
1.765
1.741
1.757
1.770
1.741
1.758
1.759
1.741
1.757
1.761
1.741
1.758
1.768
1.741
1.409
1.419
1.412
1.410
1.412
1.414
1.410
1.410
1.414
1.394
1.396
1.387
1.407
1.407
1.414
1.394
1.396
1.387
V
V
171
IV
V
Chapter 6
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.
Angle
S(1)–Re(1)–S(2)
S(3)–Re(1)–S(4)
S(5)–Re(1)–S(6)
S–Re–S av.
Twist
θ1
θ2
θ3
θav
Interligand
S(1)•••S(3)
S(1)•••S(5)
S(3)•••S(5)
S(2)•••S(4)
S(2)•••S(6)
S(4)•••S(6)
S•••S av.
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
81.0°
(81.7°)
80.9°
(80.7°)
80.9°
(80.7°)
80.9°
(81.0°)
83.1°
(83.2°)
83.1°
(83.1°)
83.1°
(83.1°)
83.1°
(83.1°)
82.7°
84.2°
82.7°
82.7°
82.6°
82.7°
82.7°
82.2°
82.7°
82.7°
83.0°
82.7°
0.01°
(0.00°)
0.02°
(0.7°)
0.03°
(0.7°)
0.02°
(0.46°)
22.8°
(26.5°)
22.9°
(25.5°)
22.9°
(22.3°)
22.9°
(24.8°)
24.9°
32.7°
0.35°
25.0°
33.1°
0.33°
24.8°
34.6°
0.34°
24.9°
33.5°
0.34°
3.121
(3.103)
3.121
(3.068)
3.125
(3.103)
3.129
(3.061)
3.129
(3.061)
3.131
(3.100)
3.119
(3.119)
3.123
(3.075)
3.123
(3.122)
3.131
(3.122)
3.128
(3.037)
3.133
(3.162)
3.157
3.344
3.081
3.163
3.259
3.079
3.165
3.216
3.081
3.171
3.301
3.087
3.167
3.325
3.085
3.168
3.206
3.086
3.126
(3.082)
3.126
(3.106)
3.165
3.275
3.083
172
Chapter 6
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.
πv
σ
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-y2 forming π-bonds.
4 – donation from the πv orbitals to the dxy and dyz giving π- and δ- bonds.
173
Chapter 6
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.
6p
dxz,yz
4e"
3πv
2a2'
A2 '
dx2-y2, xy
5e'
3πv
E'
6s
5d
4e'
3πv
3a1'
dz2
πh
2a2" 3e" 3e'
πh
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-y2 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.
174
Chapter 6
Another qualitative bonding scheme was proposed by Gray et al.10 and is shown in
Figure 6.4.3.
6p
dxz, yz
4e"
6s
dx2-y2, xy 5e'
3πv
5d
dz2
3a1'
3πv
2a2'
3πv
4e'
3πv
πh
πh
πh
πh
3e"
3e'
2a2"
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
175
Chapter 6
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.
6p
dxz, yz
4e"
dx2-y2, xy 5e'
3πv
3πv
2a2'
3πv
6s
4e'
SOMO
3πv
5d
dz2
3a1'
πh
πh
3e"
πh
πh
3e'
2a2"
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.
176
Chapter 6
5e´´
4e´´
(II)
5e´
(I)
2a´
2a´´1
3a´´1
3e´´
4e´
2a´´2
1a´´1
2e´´
Figure 6.4.5 – Unrestricted Kohn-Shan MO diagram of the monoanion 15 [ReV(LTMS)3]1from the spin unrestricted ZORA B3LYP DFT calculations.
177
Chapter 6
5e´´
4e´´
(II)
5e´
(I)
2a´2
3a´´1
2a´´1
3e´´
4e´
2a´´2
1a´´1
2e´´
Figure 6.4.6 – Unrestricted Kohn-Shan MO diagram of monoanion 16 [ReV(LCl)3]1- from the
spin unrestricted ZORA-B3LYP DFT calculations.
178
Chapter 6
5e´´
4e´´
(II)
5e´
(I)
2a´2
3a´´1
3e´´
4e´
2a´´1
2a´´2
3e´´
1a´´1
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.
179
Chapter 6
5e´´
5e´ (II)
5e´ (I)
3a´´1
2a´2
2a´2
4e´
2e´´
3e´´
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.
180
Chapter 6
5e´´
4e´´
(II)
5e´
(I)
2a´2
3a´1
2a´1
4e´
3a´´1
2a´´1
4e´´
3e´´2
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.
181
Chapter 6
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-y2 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 ligandbased 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.
182
Chapter 6
17b [ReV(L•)(L)2]0
17 [ReV(L)3]1-
5e´ (5d)
17a [ReIV(L)3]25e´ (5d)
5e´ (5d)
2a2´ (L)
3a1´ (5dz2)
2a2´ (L)
2a2´ (L)
3a1´ (5dz2)
3a1´ (5dz2)
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-y2
5dz2
26
7
5e´ (II)
48
29
2
63
2
30
30
7
9
30
7
9
5e´ (I)
41
4
4
41
66
2a2´
5e´ (I)
49
5e´ (II)
6
27
7
7
59
10
23
16
6
13
20
20
2
12
26
20
3
13
3a1´
52
27
5e´ (II)
18
13
11
66
2a2´
17b
5e´ (I)
5e´ (II)
30
7
50
5e´ (I)
5
27
2a2´
17a
22
47
5e´ (II)
17
2Cpx,y
5e´ (I)
2a2´
16
3Spz
74
2a2´
15
3Spx,y
48
49
183
31
30
6
6
28
7
6
Chapter 6
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
2
D0(S 1s → ψ*) = cte|〈S 1s|r|ψ*〉|2 = α h Is
3n
Eq. 6.5.1
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
184
Chapter 6
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.
Second Derivative
Normalized Absorption
17 [ReV(L)3]117a [ReIV(L)3]217b [ReV(L•)(L)2]0
2
1
1
2
2468
2470
2472
2474
2476
Energy (eV)
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.
185
Chapter 6
17 [ReV(L)3]117a [ReIV(L)3]2-
I
II
π* ligand
π* ligand
5e´
5e´
2a2´
2a2´
I
II
I
S 1s
II
S 1s
17 [ReV(L)3]1-
17a [ReIV(L)3]2-
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.
186
Chapter 6
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.
17 [ReV(L)3]1-
Normalized absorbance
16 [ReV(LCl)3]1-
2468
2470
2472
2474
2476
2478
Energy (eV)
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.
187
Chapter 6
Figure 6.5.4 shows the Re L1-edge of complexes 17, 17a, and 17b.
17 [ReV(L)3]1-
Normalized Absorbance
0.8
17a [ReIV(L)3]217b [ReV (L•)(L)2]0
0.6
0.4
0.2
0.0
12515
12520
12525
12530
12535
12540
Energy (eV)
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 metalcentered 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´
-
2470.9
12534.3
-
2470.9
12535.4
17b [ReV(L•)(L)2]0
2470.1
2471.2
12535.3
16 [Re(LCl)3]1-
-
2471.1
-
17 [ReV(L)3]1IV
17a [Re (L)3]
2-
188
Chapter 6
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 closedshell 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.
189
Chapter 6
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, 75047505.
19
Cowie, M.; Bennett, M. J. Inorg. Chem. 1976, 15, 1589-1594.
20
Cowie, M.; Bennett, M. J. Inorg. Chem. 1976, 15, 1595-1603.
190
Chapter 6
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
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
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.
191
Chapter 6
40
Wasinger, E. C.; deGroot, F. M. F.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. J.
Am. Chem. Soc. 2003, 125, 12894.
192
Chapter 7
Chapter 7
Experimental
193
Chapter 7
194
Chapter 7
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
1
H and
13
MHz). 1H and
13
C spectra were recorded with a Bruker ARX 250 spectrometer (1H at 400
C spectra (at 100 MHz) were referenced to TMS, using the
13
C 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.
195
Chapter 7
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.
196
Chapter 7
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 SpinHamiltonian.
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
197
Chapter 7
and maximum displacements of all atoms were smaller than 2 x 10-3 Bohr and 4 x 10-3 Bohr
respectively.
198
Chapter 7
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 1,2benzenedithiol (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.
199
Chapter 7
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 1,2bis(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).
1
H NMR (CDCl3 300K): δ = 0.34 (s, 18H), 1.12 (d, 12H), 3.95 (sept, 2H), 7.34 (s, 2H).
13
C 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).
200
Chapter 7
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.
201
Chapter 7
µ-OHCH3)2(MeCN)2][2]
Synthesis of the nickel complex [K2(µ
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
202
Chapter 7
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%).
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}+
203
Chapter 7
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%).
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
204
Chapter 7
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%).
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}+
205
Chapter 7
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%).
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
206
Chapter 7
Synthesis of the palladium complex 5 [PdII(tbpy)(LTMS)]
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%)
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
207
Chapter 7
Synthesis of the platinum complex 6 [PtII(tbpy)(LTMS)]
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%)
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}+
208
Chapter 7
Synthesis of the platinum complex 7 [PtII(tbpy)(LPh)]
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%).
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}+
1
H 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).
209
Chapter 7
Synthesis of the dimer 7c [PtII2(tbpy)2(LPh•)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%)
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%)
1
C58H56PtS2P2
%C
%H
%N
Calculated
64.9
5.3
-
found
65.0
5.3
H 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).
210
Chapter 7
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
211
Chapter 7
[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).
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).
212
Chapter 7
[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 redorange 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}-
213
Chapter 7
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%).
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
214
Chapter 7
Synthesis of 12a cis-{RhIIII2[LTMS(CH3)2][LTMS(CH3)]} + 12b cis-{RhIIICH3I[LTMS(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 Lithiumtrietylborohydride (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%).
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
215
Chapter 7
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%)
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}+
216
Chapter 7
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%)
1
H NMR (CDCl3, 300K): δ = 2.25 (s, 8H), 3.83 (s, 8H).
217
Chapter 7
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%).
Elemental analysis for [C8H16N][15]•2MeCN:
C48H82N3ReS6Si6
%C
%H
%N
Calculated
49.39
7.03
3.60
found
49.40
7.06
3.56
218
Chapter 7
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%).
Elemental analysis for [C8H16N][16]•acetone
C29H28Cl6NOReS6
%C
%H
%N
Calculated
37.07
2.98
1.49
found
37.18
2.87
1.54
219
Chapter 7
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, 12081213.
7
Schmidbaur, H.; Wohlleben, A.; Schubert, U.; Frank, A.; Huttner, G. Chem. Ber.
1977, 110, 2751-2757.
220
Chapter 8
Chapter 8
Appendix
221
Chapter 8
222
Chapter 8
8.1 – Crystallographic Data:
Compound
chem. formula
1
1ox
C18H34S2Si2
C24H40S4Si4
2
C30H54K2N2NiO2S
4Si4
-1
Fw [g.mol ]
370.75
569.16
852.26
Crystal size [mm]
0.12 x 0.09 x 0.03
0.24 x 0.20 x 0.15
0.17 x 0.08 x 0.07
Crystal system
Triclinic
Monoclinic
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)
d eg
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
100(2)
150(2)
100(2)
θ range for data collection 2.97 ≤ 2θ ≤ 50.00
3.35 ≤ 2θ ≤ 35.00
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 (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
c
223
Chapter 8
2a
2b
3
chem. formula
C40H76NNiS4Si4
C25H42Cl2NiS4Si4
C40H76NCuS4Si4
Fw [g.mol-1]
870.33
712.80
875.16
Crystal size [mm]
0.34 x 0.22 x 0.07
0.14 x 0.07 x 0.03
0.15 x 0.10 x 0.09
Crystal system
Triclinic
Monoclinic
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)
d eg
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
200(2)
100(2)
100(2)
3.11 ≤ 2θ ≤ 34.99
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]]
493 / 16
7695
[R(int) = 0.0314]]
184 / 3
13308
[R(int) = 0.0399]]
470 / 0
(Mo K), cm-1
0.655
1.097
0.711
0.0430 / 1.050
0.0270 / 1.035
0.0583 / 1.009
wR2 (all data)
0.1033
0.0727
0.1313
0.713 / -0.840
0.545 / -0.978
0.553 / -0.577
Compound
c
224
Chapter 8
Compound
4
5
6
chem. formula
C40H76NAuS4Si4
C30H44N2PdS2Si2
C30H44N2PtS2Si2
Fw [g.mol-1]
1008.58
659.37
748.06
Crystal size [mm]
0.09 x 0.08 x 0.07
0.20 x 0.18 x 0.13
0.24 x 0.16 x 0.04
Crystal system
Triclinic
Monoclinic
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
100(2)
100(2)
100(2)
3.25 ≤ 2θ ≤ 33.23
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]]
470 / 0
346 / 0
346 / 3
µ(Mo Kα), cm-1
3.147
0.796
4.558
0.0207 / 1.035
0.0239 / 1.038
0.0275 / 1.105
wR2 (all data)
0.0482
0.0576
0.0549
1.027 / -1.032
0.628 / -0.634
1.234 / -1.389
c
225
Chapter 8
Compound
9
7c
8
chem. formula
C86H106Cl6 F12N4P2Pt2 S4
C58H59Cl9P2PtS2
C51H60PtS4
Fw [g.mol-1]
2180.78
1432.28
996.32
Crystal size [mm]
0.25 x 0.20 x 0.10
0.09 x 0.04 x 0.02
0.20 x 0.15 x 0.10
Crystal system
Monoclinic
Triclinic
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)
d eg
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, Å
4645.0(2)
3043.1(3)
4777.7(5)
Z
2
2
4
T, K
100(2)
100(2)
100(2)
3.17 ≤ 2θ ≤ 27.50
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
557 / 28
682
479 / 1
(Mo K), cm-1
0.710
0.710
0.695
0.0298 / 1.035
0.0512
0.0525 / 1.011
wR2 (all data)
0.0602
0.1269
0.1198
0.105 / -1.024
3.792 / -2.402
3.09 / -1.83
3
c
226
Chapter 8
Compound
10
11
12a + 12b
chem. formula
C40H76NCoS4Si4
C64H124N6RhS4Si4
C27.3H49.9I1.7RhS4Si4
Fw [g.mol-1]
870.55
1321.20
937.41
Crystal size [mm]
0.21 x 0.14 x 0.07
0.26 x 0.23 x 0.12
0.12 x 0.12 x 0.11
Crystal system
Triclinic
Triclinic
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)
d eg
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, Å
2488.3(3)
1890.22(17)
3840.96(19)
Z
2
1
4
T, K
100(2)
100(2)
100(2)
2.92 ≤ 2θ ≤ 32.50
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]]
470 / 0
13615
[R(int) = 0.0445]]
370 / 0
20154
[R(int) = 0.0434]]
362 / 0
(Mo K), cm-1
0.635
0.439
2.171
0.0596 / 1.059
0.0450 / 1.063
0.0311 / 1.104
wR2 (all data)
0.1033
0.0991
0.0634
0.465 / -0.463
0.733 / -0.702
1.876 / -1.407
3
c
227
Chapter 8
Compound
13
14
15
chem. formula
C64H124N6CrS4Si4
C42H80Cl4NCrOS4Si4
C48H82N3ReS6Si6
Fw [g.mol-1]
1270.29
1049.47
1248.27
Crystal size [mm]
0.05 x 0.05 x 0.04
0.13 x 0.12 x 0.04
0.52 x 0.12 x 0.10
Crystal system
Triclinic
Monoclinic
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
100(2)
100(2)
100(2)
2.91 ≤ 2θ ≤ 27.50
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]
736 / 0
540 / 3
326 / 6
µ(Mo Kα), cm-1
0.357
0.663
0.796
0.0659/ 1.147
0.0440 / 1.037
0.0736/ 1.347
wR2 (all data)
0.1755
0.0886
0.1703
1.193 / -0.493
0.467 / -0.484
2.603 / -5.604
c
228
Chapter 8
Compound
16
chem. formula
C29H28Cl6NOReS6
Fw [g.mol-1]
997.78
Crystal size [mm]
0.12 x 0.10 x 0.09
Crystal system
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
100(2)
ρ calcd, g cm-3
1.890
refl. collected
91435
Unique refl.
12678
[R(int) = 0.0396]]
399 / 0
µ(Mo Kα), cm-1
4.309
0.0294/1.041
c
wR2 (all data)
0.0686
1.202 / -2.165
229
Chapter 8
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.
230

d - Ruhr-Universität Bochum

Transcription

Similar documents

DuraSIL-SL - Best Materials

Chem Comm 2011 - Research Repository UCD

THE LANTHANIDES

Supporting Information Metal-Ion Catalysis in Alkaline Ethanolysis of

Properties of Iron(ll) and Iron(lll) Perchlorate Complexes with

Synthesis and structure of a new example of nickel (II) coordinated

August 2012

- Tandberg Data

Chiral Semicorrins and Related Nitrogen Heterocycles as Ligands in